LONG-TERM ECOLOGICAL CHANGE IN THE NORTHERN GULF OF ALASKA
This page intentionally left blank
LONG-TERM ECOLOGICAL CHANGE IN THE NORTHERN GULF OF ALASKA Edited by Robert B. Spies
Elsevier Radarweg 29, PO Box 211, 1000 AE Amsterdam, The Netherlands The Boulevard, Langford Lane, Kidlington, Oxford OX5 1GB, UK
First edition 2007 Copyright © 2007 Elsevier B.V. All rights reserved No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means electronic, mechanical, photocopying, recording or otherwise without the prior written permission of the publisher Permissions may be sought directly from Elsevier’s Science & Technology Rights Department in Oxford, UK: phone (+44) (0) 1865 843830; fax (+44) (0) 1865 853333; email:
[email protected]. Alternatively you can submit your request online by visiting the Elsevier web site at http://elsevier.com/locate/permissions, and selecting Obtaining permission to use Elsevier material Notice No responsibility is assumed by the publisher for any injury and/or damage to persons or property as a matter of products liability, negligence or otherwise, or from any use or operation of any methods, products, instructions or ideas contained in the material herein. Because of rapid advances in the medical sciences, in particular, independent verification of diagnoses and drug dosages should be made Library of Congress Cataloging-in-Publication Data A catalog record for this book is available from the Library of Congress British Library Cataloguing in Publication Data A catalogue record for this book is available from the British Library ISBN-13: 978-0-444-52960-2 ISBN-10: 0-444-52960-8
For information on all Elsevier publications visit our website at books.elsevier.com
Printed and bound in Italy 07
08
09
10
11
10
9
8 7 6 5 4 3 2 1
We dedicate this book to the people of Alaska. We hope that it will help inspire both further support for research and wise human action in the Gulf of Alaska.
This page intentionally left blank
Acknowledgements
This book was made possible through the passionate work of government, academic and private scientists in the Northern Gulf of Alaska. Many of these hard-working scientists have dedicated their careers to some aspect of this productive ecosystem. Special thanks are due to the Exxon Valdez Trustee Council of 2000, who provided the funds for this project. This and previous councils recognized the wisdom of dedicating some small part of the trust funds to summarizing what we have learned since the 1989 oil spill about the ecosystem in which it occurred. Their vision for long-term monitoring and research is a great example for future political leaders. Graphic material for this book was prepared by Peter Veres. We are grateful to the many naturalist photographers who permitted use of their work.
This page intentionally left blank
Contributors
Kevin M. Bailey, Ph.D. Alaska Fisheries Science Center NOAA/NMFS 7600 Sand Point Way N.E., Building 4 Seattle, WA 98115
[email protected] Z. Morgan Benowitz-Fredericks, Ph.D. Department of Biology PO Box 351800 University of Washington Seattle, WA 98195
[email protected]. James L. Bodkin, Ph.D. Alaska Biological Science Center USGS 1011 E. Tudor Blvd. Anchorage, AK 99503 Wiebke J. Boeing, Ph.D, Department of Fishery and Wildlife Sciences 2980 South Espina, 132 Knox Hall PO Box 30003, Campus Box 4901 New Mexico State University Las Cruces, NM 88003-8003
[email protected]
Evelyn Brown, Ph.D. School of Fisheries and Ocean Sciences University of Alaska Fairbanks Fairbanks, AK 99775
[email protected] Theodore Cooney, Ph.D. Emeritus Professor School of Fisheries and Ocean Sciences University of Alaska Fairbanks, AK 99775-7220 PO Box 486 Choteau. Montana
[email protected] Sara J. Iverson, Ph.D. Department of Biology Dalhousie University 1355 Oxford Street Halifax, NS Canada B3H 4J1
[email protected] Gordon H. Kruse, Ph.D. President’s Professor of Fisheries University of Alaska, Fairbanks Juneau Center for Fisheries & Ocean Sciences 11120 Glacier Highway Juneau, AK 99801-8677
[email protected]
x
Contributors
John F. Piatt, Ph.D. Alaska Biological Science Center USGS Anchorage, Alaska Canada Fairbanks, AK Mailing address: Marrowstone Marine Station 616 Marrowstone Point Road, Nordland, WA 98358-9633
[email protected] Paul Reno, Ph.D. Department of Microbiology Oregon State University Coastal Oregon Marine Experiment Station Newport, OR Mailing address: Hatfield Marine Science Center 2030 SE Marine Science Drive Newport, OR 97365
[email protected] Stanley D. Rice, Ph.D. NOAA/NMFS Auke Bay Laboratory 11305 Glacier Highway Juneau, AK 99801-8626
[email protected]
Robert B. Spies PO Box 315 45100 Peterson St. Little River, CA 95456
[email protected] Alan M. Springer, Ph.D. Institute of Marine Sciences University of Alaska 1708 Marmot Hill Road Fairbanks, AK 99709
[email protected] Thomas Weingartner, Ph.D. Institute of Marine Sciences University of Alaska Fairbanks, AK 99775
[email protected]
Contents
1 Introduction
1
Robert B. Spies and Theodore Cooney
1.1. Why We Have Written This Book 7 1.2. Who Is the Audience for This Book, and What Is Its Scope? 8 1.3. Organization 8 References 9 2 Ecosystem Structure 11 2.1. Introduction 11 Robert B. Spies and Alan M. Springer
2.2. The Physical Environment of the Gulf of Alaska 12 Thomas Weingartner
2.2.1. Geomorphology 13 Local Atmospheric Forcing 17 Precipitation and Runoff 18 Winds 21 Heat Fluxes 24 2.2.2. Physical Oceanography 26 The Seasonal Cycle of Water Properties 26 Circulation over the Gulf of Alaska Shelf and Slope 30 2.2.3. Tides 41 2.2.4. Gulf of Alaska Basin 42 2.2.5. North Pacific Ocean 44 2.3. The Marine Production Cycle 47 Theodore Cooney
2.3.1. Introduction 47 2.3.2. The Annual Cycle of Production 48 Protected Inner Waters 55 The Open Ocean 57 Open Coastal and Shelf Waters 59
xii
Contents
2.4. The Transfer of Matter and Energy Through the Food Web 60 Theodore Cooney
2.4.1. Food-Web Structure – The Principal Forage Stocks 62 2.4.2. Efficient Foraging on Patches 70 2.4.3. Food-Web Complexity and Efficiency 71 2.5. Strategies for Survival 74 2.5.1. Introduction 74 Alan M. Springer
2.5.2. Introduction to Fishes
75
Theodore Cooney
2.5.3. Pink Salmon
76
Theodore Cooney
2.5.4. Pacific Herring
81
Theodore Cooney
2.5.5. Walleye Pollock
85
Kevin M. Bailey and Lorenzo Ciannelli
Introduction 85 Adaptations for Survival 87 Effect of Ecosystem Structure on Pollock Survival 91 Conclusions 92 2.5.6. Comparing Fish Life Histories 93 Theodor Cooney
2.5.7. Seabirds 94 Morgan Benowitz-Fredericks, A.S. Kitaysky and Alan M. Springer
Introduction 94 Seabirds in the Breeding Season 95 Focal Species 95 Phylogeny 96 Foraging Ecology and Reproductive Strategies 96 Foraging Ecology 97 Foraging Range and Habitat Use 100 Diet 101 Response to Changes in Prey 102 Life History Traits and Reproduction 103 Chick Strategies 104 Parental Behavior and Consequences for Survival 108 Other Factors Affecting Survival: Nesting and Predation 111 Summary 112 2.5.8. Marine Mammals 114 Sara J. Iverson, Alan M. Springer, and James Bodkin
Introduction 114 Offspring Constraints and Strategies 121
Contents
xiii
Trade-offs and Consequences: Energy and Population Considerations 123 Predator Avoidance 130 Summary 131 2.5.9. Crabs and Shrimps 135 Gordon H. Kruse
Introduction 135 Distribution and Habitats for Survival 135 Feeding Strategies 137 Predator Defense Strategies 138 Reproductive Strategies 140 Larval Survival: Strategies of Timing (or “Match the Hatch”) 143 Larval Survival: Strategies of Space (or “Location, Location, Location”) 144 Implications of Survival Strategies 145 References 145 Mini-Glossary 169 3 Agents of Ecosystem Change 171 3.1. Introduction 171 Robert B. Spies
3.2. Climate 171 Thomas Weingartner
3.2.1. Introduction 171 3.2.2. Climate Forcing 172 3.2.3. Physical Environmental Variability in the North Pacific Ocean 176 El Niño Southern Oscillation (ENSO) 176 The Pacific Decadal Oscillation (PDO) 178 3.3. Geophysical Mechanisms 180 Robert B. Spies
3.3.1. Introduction 180 3.3.2. Tectonics and Earthquakes 180 3.3.3. Sediment Slumping 183 3.3.4. Volcanism 183 3.3.5. Tsunamis 186 3.3.6. Glaciers 187 3.4. Species Interactions 187 Gordon H. Kruse
3.4.1. Introduction 187 3.5. Marine Mammal Harvest and Fishing 192 Gordon H. Kruse and Alan M. Springer
3.5.1. Introduction 192
xiv
Contents
3.5.2. Marine Mammal Harvests and Persecution 192 Great Whales 192 Pinnipeds 196 Sea Otters 199 3.5.3. Fisheries of the Northern Gulf of Alaska 201 Introduction 201 3.5.4. History of Fishing and Fishery Management in Alaska 203 Salmon 203 Herring 205 Groundfish 207 Shellfish 208 3.5.5. Direct and Indirect Effects of Fishing 211 3.5.6. Effects of Hunting and Fishing: Some Conclusions 218 3.6. Disease 219 Paul Reno
3.6.1. 3.6.2. 3.6.3. 3.6.4.
Introduction 219 Agents of Infectious Disease and their Lifestyles 220 Resistance to Infection and Disease 224 The Diversity of Infectious Agents Causing Diseases in Marine Animals 226 Diseases of Coral (http://www.coral.noaa.gov/coral_ disease/) 230 Diseases of Molluscs (http://www.pac.dfo-mpo.gc.ca/sci/shelldis/ title_e.htm) 230 Diseases of Crustaceans (http://www.pac.dfo-mpo.gc.ca/sci/shelldis/ pages/hematcb_e.htm) 233 Diseases of Teleosts (http://wfrc.usgs.gov/capfishhealth.htm) 233 Diseases of Reptiles (http://nationalzoo.si.edu/ConservationAndScience/ AquaticEcosystems/SeaTurtles/deem.cfm) 236 Diseases of Birds (http://www.nwhc.usgs.gov/pub_metadata/ index.html) 236 Diseases of Marine Mammals (http://www.nmfs.noaa.gov/ pr/health/) 237 3.6.5. Population Effects and Implications 238 3.7. Contaminants 241 Robert B. Spies and Stanley Rice
3.7.1. Significance of Life Stage at Time of Impact 246 3.7.2. Floating Oil and Surface-dwelling Animals 246 References 247 Mini-Glossary 256
Contents
4 Long-Term Change 259 4.1. Introduction 259 Robert B. Spies and Thomas Weingartner
4.2. Atmosphere and Ocean 265 Thomas Weingartner
4.2.1. Introduction 265 4.2.2. Gulf of Alaska Shelf 266 Winds 266 Variability in Surface Heating and Cooling and Shelf Temperatures 266 Variability in Runoff and Shelf Salinity 269 4.2.3. Gulf of Alaska Basin 273 4.3. Zooplankton 274 Theodore Cooney
4.4. History and Production Trends in Salmon 278 Theodore Cooney
4.4.1. 4.4.2. 4.4.3. 4.4.4. 4.4.5.
Introduction 278 Resource Use and Management 278 Broadscale Trends in Time and Space 280 Prince William Sound: A Pink Salmon Case History 284 What’s Behind the Large-Scale Patterns in Salmon Catch and Production? 287 4.5. Pacific Herring 290 Evelyn Brown
4.5.1. Resource Use and Management 290 4.5.2. Effects of Climate on Gulf of Alaska Herring 293 4.6. Groundfish 300 Wiebke J. Boeing, Michael H. Martin, and Janet T. Duffy-Anderson
4.6.1. 4.6.2. 4.6.3. 4.6.4. 4.6.5. 4.6.6.
Introduction 300 Walleye Pollock 303 Pacific Cod 303 Arrowtooth Flounder 305 Pacific Ocean Perch 305 Explaining Population Change 306 Climate forcing 306 Biological controls 308 Fisheries effects 309 Summary 310 4.7. Seabirds in the Gulf of Alaska 311 Alan M. Springer
xv
xvi
Contents
4.7.1. 4.7.2. 4.7.3. 4.7.4. 4.7.5.
Introduction 311 Long-term Changes in Common Murres 313 Long-term Changes in Tufted Puffins 320 Long-term Changes in Black-legged Kittiwakes 321 Long-term Changes in Other Species 323 Storm Petrels 323 Cormorants 323 Murrelets 324 Miscellaneous Species: PWS 324 4.7.6. Causes of Long-term Change in Seabirds 325 Changing Abundance 326 4.7.7. Changing Productivity 329 4.7.8. Conclusions 332 4.7.9. Summary 334 4.8. Population Ecology of Seabirds in Cook Inlet 335 John F. Piatt and Ann M.A. Harding
4.8.1. Introduction 335 4.8.2. The Cook Inlet Ecosystem 337 4.8.3. Response of Seabirds to Variability in Prey 339 Form of Response 339 Variability in Response 340 4.8.4. Population Dynamics of Seabirds in Cook Inlet 344 Population Parameter Indices 347 4.8.5. Long-term Changes in the Gulf of Alaska Marine Environment 350 4.9. Marine Mammal Populations 352 Alan M. Springer, Sara J. Iverson, and James L. Bodkin
4.9.1. Harbor Seals 352 4.9.2. Steller Sea Lions 355 4.9.3. Sea Otters 357 4.9.4. Potential Causes of Marine Mammal Population Change 361 4.9.5. Harbor Seal Decline 363 4.9.6. Steller Sea Lion Decline 368 4.9.7. Sea Otter Population Changes 374 4.9.8. Conclusions 375 4.10. Crabs and Shrimps 378 Gordon H. Kruse
4.10.1. Introduction 378 4.10.2. Long-term Dynamics of Shrimp Stocks 379 4.10.3. Long-term Dynamics of Crab Stocks 380
Contents
4.10.4. Climate Forcing 383 Is Climate Important to Invertebrate Populations? 383 Shrimp and the Match–Mismatch Hypothesis 385 Red King Crab and the Match–Mismatch Hypothesis 386 Reconciling the Match–Mismatch Hypothesis among Shrimp and Crabs 387 Other Potential Climate-forcing Mechanisms and Red King Crab 388 Climate Forcing and Tanner Crabs 388 4.10.5. Biological Controls 389 4.10.6. Fishing Effects 392 4.10.7. Conclusions 393 References 394 Mini-Glossary 418 5 The Exxon Valdez Oil Spill 419 Stanley D. Rice, Jeffrey W. Short, Mark G. Carls, Adam Moles and Robert B. Spies
5.1. 5.2. 5.3. 5.4.
Introduction 419 Pre-spill Conditions 422 History of the Spill 423 Oil Fate: Transport, Weathering, and Persistence 428 5.4.1. Contaminants in Prince William Sound Prior to the Exxon Valdez Oil Spill 432 5.4.2. Initial Fate of the Oil 433 5.4.3. Oil Cleanup Efforts 437 5.4.4. Long-term Oil Persistence 440 Early Surveys (1989–1993) 440 Post-1993 Surveys 442 5.4.5. Weathering and Bioavailability of Persistent Oil 445 5.5. Effects of the Spill on Aquatic Organisms 447 5.5.1. Acute Effects 449 Birds 450 Sea Otters 450 Seals 451 Sea Lions 452 Whales 452 5.5.2. Short-term Effects of the Spill 454 Pacific Herring 455 Pink Salmon Juveniles 457 Intertidal and Subtidal Communities 461 5.5.3. Long-term Effects of the Spill 466 Pink Salmon Embryos 466
xvii
xviii
Contents
Mussels 477 Sea Otters and Harlequin Ducks 480 5.6. Indirect Interactions 486 5.7. Expectations, Certainty, and Final Lessons from the Exxon Valdez Oil Spill 487 5.8. The Legacy of the Exxon Valdez Oil Spill 489 5.9. Appendices 491 Appendix A 491 Appendix B 503 Properties and Composition of Fresh and Weathered Alaska North Slope Crude Oil 503 References 506 Mini-Glossary 520 6 Long-Term Changes in the GOA: Properties and Causes 521 Robert B. Spies, Theodore Cooney, Alan M. Springer, Thomas Weingartner and Gordon H. Kruse
6.1. Introduction 521 6.2. Forces of Change 528 6.2.1. Climate 528 6.2.2. The Exxon Valdez Oil Spill 531 6.2.3. The 1964 Earthquake 534 6.2.4. Effects of Harvesting 535 6.3. Characteristics of Ecosystem Change 535 6.3.1. Understanding Productivity is Key 536 6.3.2. Responses to Climate Changes Vary at the Mesoscale 537 6.3.3. Animals with High Reproductive Rates Periodically Dominate 538 6.3.4. Top-down Forces Have been Underestimated 538 6.3.5. Bottom-up and Top-down Operate Together 543 6.3.6. Contaminants and Disease as Predators 548 6.3.7. Structure of the Ecosystem Affects Ecological Change 549 6.3.8. Interaction of Multiple Forcing Factors and Cascading Effects 549 6.3.9. Prediction of Ecological Change 550 References 555 Index 561
Chapter 1
Introduction Robert B. Spies and Theodore Cooney
Our species has had a long and complex relationship with the sea. It enabled us to reach all of earth’s continents during our ancient migrations, for modern commerce, and for warfare. Its fish, shellfish, algae, and mammals supply much of our food, and today, new pharmaceuticals. Its mineral resources are used for producing energy and modern manufactured goods, and its huge chemical engine replenishes atmospheric oxygen. Sadly, the sea has also become the most convenient repository for the complex waste streams of our modern economy. Through millions of years of our evolution on a planet with three-quarters of its surface covered by water, we have developed deep, enduring, practical and spiritual ties with the sea. Only in the past 50 years have we really begun to understand that the cumulative actions of billions of humans are changing the sea on a large scale. Yet, in the absence of a clear understanding of natural changes in oceanic waters, including cycles in temperature, salt content, current patterns, and plant and animal populations, we cannot determine how exactly we are affecting the sea. Today, global climate change threatens us through sea level rise, increasingly violent storms, new patterns of precipitation, and alteration of fish populations. A majority of the earth’s exploited fish stocks are in decline; polar ecosystems are contaminated with exotic persistent chemicals that could disrupt human and wildlife reproduction. Coastal habitats and their spiritual values are under siege, as much of the earth’s burgeoning population is located near the sea. Therefore, we need, more than ever, a deeper understanding of how the ocean works, so that we can limit our impacts. Fishermen have long known that the harvestable species populations are always changing. Since the dawn of modern science in the 1700s, we have studied the ocean to see why this is so. Progress has been frustratingly slow. We are generally unable to predict changes from first principles, even after several centuries of conceptualizing and grappling with an increasing awareness of the complexity of the sea. However, a few key insights are beginning to emerge. Our understanding of the marine ecosystem is probably at a similar level as our understanding of astronomy was in 1920. Nowhere is ecosystem change more profoundly apparent than in the Gulf of Alaska. The rise and fall in the populations of salmon in the twentieth century, the decline of commercial shrimp and crab populations, the near extinction of sea otters, and other changes all attest to the frequent, large-scale, fundamental shifts in the Long-Term Ecological Change in the Northern Gulf of Alaska Robert B. Spies (Editor) © 2007 Elsevier B.V. All rights reserved.
2
Long-Term Ecological Change in the Northern Gulf of Alaska
Gulf’s animal populations (see Box 1.1). Figures 1(A–G) depict examples of these dramatic changes, some of which last for centuries. There are powerful natural and anthropogenic forces that drive changes in the northern Gulf of Alaska. With regard to natural forces, the northern Gulf of Alaska has strong seasonal, annual, interannual, decadal, and longer-term cycles. The atmosphere and ocean are intimately linked. Meteorological changes find expression in the ocean and in coastal watersheds. In fact, the entire North Pacific Ocean appears to fluctuate broadly from east to west in quasi-long-term cycles, ranging from a few years to decades or even longer. The strength and geographic position of the Aleutian Low Pressure system, the dominant meteorological feature of the North Pacific Ocean, forces many of the ocean’s responses to climate change – on a range of time and space scales. In addition, the North Pacific Ocean warms during El Ni~ nos every 4 to 5 years, but the relationship to the Aleutian Low Pressure is not well understood. Contemporary studies have demonstrated that changes in the ocean can occur in as little as a day after atmospheric influences, but oceanic conditions also track the climate over extended periods – decades and even longer. In the Gulf of Alaska, single storms, abnormal seasons, the inter-annual El Ni~ no–La Ni~ na cycle, and 20- to 50-year climate fluctuations leave a variety of historical records. A recently assembled history
BOX 1.1: CHANGING ANIMAL POPULATIONS IN THE GULF by Robert Spies and Theodore Cooney Animal populations in the Gulf of Alaska have waxed and waned for thousands of years, but the only historical evidence for such long-term change comes from the sockeye salmon, recently revealed in lake sediments on Kodiak Island (Fig. 1.1). During the nineteenth century, the sea otter and whale harvests, and in the twentieth century the harvest of a half million whales in the North Pacific Ocean (e.g., Fig. 1.2), had direct effects and may have set in motion, cascading long-term changes in the Gulf ecosystem. In the twentieth century, several rearrangements of the Gulf ecosystem occurred, favoring different species in different decades. The twentieth century changes were first evident in fluctuations in the salmon fisheries (Fig. 1.3), and then the commercial fin and shellfish fisheries in the 1950s to 1970s (Fig. 1.4). Most recently, systematic studies of crustacean and fish populations in the 1970s set the foundation for recording a documented shift in the Gulf, beginning at about the middle of that decade (Fig. 1.5). There were also dramatic declines in marine mammal populations (Fig. 1.6). The huge Exxon Valdez oil spill in 1989 had major impacts on intertidal communities, seabirds, and marine mammals in parts of the Gulf.
Introduction
10 9
δ15N
8 7 6 5 4 3 −250
0
250
500
750 1000 1250 1500 1750 2000 YEAR AD
Figure 1.1: Long-term record of sockeye salmon (Onchorhynchus nerka) abundance in the northern Gulf of Alaska. (after Finney et al., 2002). The data are nitrogen isotope ratios (δ15N) from the sediment cores of Akalura Lake, Kodiak Island, plotted by year, with the ratio value increasing in proportion to the population size.
Figure 1.2: Fin whale (Balenoptera physalus) harvest in the North Pacific Ocean, 1924–1987. Harvested biomass in metric tons. Each dot represents the cumulative total biomass caught in a 100-km2 area (after Springer et al., 2003).
3
Long-Term Ecological Change in the Northern Gulf of Alaska
CATCH (millions of salmon)
80 70 60 50 40 30 20 10 0 1990
1985
1980
1975
1970
1965
1960
1955
1950
1945
1940
1935
1930
1925
YEAR
Figure 1.3: Commercial salmon catches in Alaska during the twentieth century (after Francis and Hare, 1994).
350000 PACIFIC OCEAN PERCH CATCH (metric tons)
4
300000 250000 200000 150000 100000 50000 0 1960
1965
1970
1975
1980 1985 YEAR
1990
1995
2000
Figure 1.4: Catch of Pacific ocean perch (Sebastes alutus) (after Hanselman et al., 2005).
Introduction
a
b
c
Figure 1.5: Trawl catches in Pavlov Bay, Alaska Peninsula, show the changes in the Gulf’s inner-shelf communities from (a) crustacean dominance in the 1960s and 1970s, (b) in transition with a mixture of shrimp and fish, 1977–1980, and (c) finally, to flatfish and gadid (cod-like fishes) dominance after 1981. (Photographs, courtesy of Paul Anderson, ret. NOAA, National Marine Fisheries Service. Also see Piatt and Anderson, 1996).
5
c b
a
Figure 1.6: The decline of the Steller sea lion (Eumetopias jubatus) stocks from the late 1970s through the 1980s. Photographs of the same beach at Ugamak Island taken in (a) 1969, (b) 1979, and (c) 1986 (courtesy of NOAA, National Marine Laboratory).
Figure 1.7: After the Exxon Valdez oil spill, thousands of birds were killed, and were washed up on Gulf beaches. These sea birds were mainly common murres (Uria aalge) that died offshore (Photograph, courtesy of the Exxon Valdez Trustee Council.) Several hundred thousand sea birds were killed by the spill (see Piatt and Ford, 1996).
Introduction
7
of sockeye salmon abundance during the last 3500 years reveals much about longerterm changes in production, some of several hundred years duration. The Gulf of Alaska has a highly productive food web, culminating in large stocks of crustaceans, fishes, birds, and marine mammals that are exploited for subsistence, commercial and sport fisheries, and form a basis for the growing tourist industry. The harvest of fish and marine mammals can have widespread impacts on the Gulf’s ecosystem. The people of Alaska have historically been an inextricable part of the marine ecosystem. Man, therefore, is both a top predator in the Gulf and, at the same time, uses its waters to transport large quantities of oil and other commodities, including forest products and coal. The results of scientific studies carried out in the northern Gulf of Alaska since the 1989 Exxon Valdez oil spill are refining our views about the response of this ecosystem, its structure, function, and long-term changes in the face of perturbations, both human and natural. Findings from oil spill studies have already advanced our understanding of key ecological relationships, pointed the way to future long-range studies, and contributed to a more informed management of our impacts. In marine research programs and in the development of scientific careers, opportunities for producing synthetic understanding of large ecosystems are relatively rare. Occasionally, after sufficient research has been completed, the time is ripe for synthesis and, possibly, revised concepts about the structure and function of marine ecosystems. The trustees of the publicly held settlement funds from the 1989 oil spill have afforded such an opportunity. So, while the aftermath of the grounding of the Exxon Valdez on Bligh Reef had the dimensions of an ecological and economic tragedy, it also ironically provided the unprecedented means to learn much that will serve the people of Alaska and elsewhere. It is this legacy that is a primary inspiration for our book.
1.1. Why We Have Written This Book Our intent is to synthesize recent information about historical changes in the northern Gulf of Alaska, including those forced by the Exxon Valdez oil spill of March 1989. Our analyses and conclusions are based on a large number of published reports and peer-reviewed manuscripts from a variety of scientific disciplines. We hope that bringing this material together will benefit professionals and non-professionals alike. While our grasp of the causes of change is compromised by the enormous complexity of marine ecosystems as they shift on scales of months, years, decades, and centuries, we believe that “ many small steps make a journey.” There are seemingly competing ideas and theories about the form and function of the marine ecosystem of the northern Gulf of Alaska that we present here. We hope that these ultimately lead to new hypotheses and understandings about the nature of change.
8
Long-Term Ecological Change in the Northern Gulf of Alaska
1.2. Who Is the Audience for This Book, and What Is Its Scope? The book is intended to serve marine ecologists as well as non-specialists who have a general scientific background and an interest in marine ecology, including, for example, resource managers, graduate students, policy makers, eco-tour leaders and visitors, industry representatives, environmental organizations, and researchers in allied disciplines. We have attempted to present the concepts in the most straightforward and understandable way possible, and to make the material maximally accessible to the wider public – those who do not read the specialized scientific journals where this information appears. At the same time, the authors have not watered down the scientific underpinnings of our knowledge, so it should also be a useful reference for marine scientists seeking a broad synthesis. We have written here about ecological change and what we know of its patterns and causes. This book is not a compilation of everything known about the Gulf of Alaska; there are other references available that are more comprehensive (e.g., Hood and Zimmerman, 1987); nor is it a textbook on oceanography or ecology. We have focused our efforts on ecosystem change as a necessary framework for understanding human impacts because change itself is a primary concern for the people who live next to, and rely on, the Gulf of Alaska.
1.3. Organization The synthesis is presented in five chapters: (1) How does the ecosystem work? In Chapter 2, the ecosystem and its seasonal changes are presented, along with the adaptations for survival in several portal species of fish, seabirds, and mammals. (2) What are the root causes of change? Chapter 3 describes the forces of change – climate, geophysics, species interactions, harvest, disease and contaminants. (3) How has the ecosystem changed in the past? Chapter 4 describes long-term ecosystem changes in physical and biological oceanography, as well as changes in the portal species introduced in Chapter 2 (Pacific herring, pink salmon, pollock, common murres, tufted puffins, black-legged kittiwakes, harbor seals, sea lions, and sea otters). (4) What were the effects of the Exxon Valdez oil spill? The history, fate, and effects of the 1989 spill are described in Chapter 5. (5) What are the reasons for the behavior of the whole system, and what are the emergent properties of ecosystem behavior over the long term? This discussion is presented in Chapter 6.
Introduction
9
References Finney, B. P., I. Gregory-Eaves, M. S. V. Douglas, and J. P. Smol. 2002. Fisheries productivity in the northeastern Pacific Ocean over the past 2,200 years. Nature 416: 729–733. Francis, R. C. and S. R. Hare. 1994. Decadal-scale regime shifts in the large marine ecosystems of the North-east Pacific: a case for historical science. Fish. Oceanogr. 3:4: 279–291. Hanselman, D., J. Heifetz, J. T. Fujioka, and J. N. Ianelli. 2005. Gulf of Alaska Pacific ocean perch. In: Appendix B: Stock Assessment and Fishery Evaluation Report for the Groundfish Resources of the Gulf of Alaska. North Pacific Fishery Management Council, Anchorage, Alaska. pp. 525–578. Hood, D. W. and S. T. Zimmerman (Eds). 1987. The Gulf of Alaska. Physical Environment and Biological Resources. US Minerals Management Service. pp. 655. Piatt, J. F. and P. Anderson.1996. Response of common murres to the Exxon Valdez oil spill and long-term changes in the Gulf of Alaska Ecosystem. In: S. D. Rice, R. B. Spies, D. A.Wolfe and B. A. Wright (Eds.), Proceedings of the Exxon Valdez Oil Spill Symposium, American Fisheries Society Symposium, Bethesda, MD. 18, pp. 720–737. Piatt, J. F. and R. G. Ford. 1996. How many birds were killed by the Exxon Valdez oil spill? In: Proceedings of the Exxon Valdez Oil Spill Symposium S. D. Rice, R. B. Spies, D. A. Wolfe and B. A. Wright (Eds.). American Fisheries Society Symposium. Bethesda, MD, 18, pp. 712–719. Springer, A. M., J. A. Estes, G. B. van Vliet, T. M. Williams, D. F. Doak, E. M. Danner, K. A. Forney, and B. Pfisterer. 2003. Sequential megafaunal collapse in the North Pacific Ocean: An ongoing legacy of industrial whaling? Proc. Natl. Acad. Sci. 100: 12223–12228.
This page intentionally left blank
Chapter 2
Ecosystem Structure 2.1. Introduction Robert B. Spies and Alan M. Springer Marine animals thrive in the Gulf of Alaska (GOA) for two basic reasons. First, the ecosystem is productive; a lot of food is available to animals living there. Second, the small number of successful species in the Gulf have made the most of the available energy by organizing themselves, i.e., evolving body forms, physiology, and behaviors that allow them to successfully withstand the rigors of the subarctic ocean, capture food, grow, and reproduce faster than they themselves are eaten, are harvested, or fall victims to contaminants or disease. In this chapter, we will look at the ecosystem from these two perspectives: how the sun’s energy is captured and passed from plants to animals and then from animals to animals; and the strategies that animals have evolved to capture a share of that energy, use it, and flourish in the Gulf of Alaska. The key to understanding long-term change rests on knowing how the ecosystem is “wired,” that is, what is connected to what (the nodes and the connections), but also what it is about each node, i.e., the strategies of successful species that allow them to prosper and dominate. First, to understand why the Gulf of Alaska is so productive, we start with the climate and physical structure of the ocean as it changes through the seasons. In the nine coldest and stormiest months of the year, prevailing winds in the northern Gulf spiral counterclockwise from the central Gulf, pushing surface water onshore and storms into the high ringing coastal ranges. The clouds lose their moisture as they rise over the mountains, dumping large amounts of snow. This precipitation is locked in the cold mountains and coastal plains until spring. As the weather warms, huge amounts of nutrient-poor, but iron-rich, freshwater run off the melting snowpack and glaciers into the nearshore ocean, feeding the Alaska Coastal Current (ACC). The freshwater floats as a layer over the saltier oceanic water near the coast and resists the mixing of deep nutrient-rich water with that at the surface during the summer and early autumn. At this time, the production cycle at most coastal and shelf locations becomes severely constrained by lack of essential nutrients, and the high rates of photosynthesis observed in the early spring decline to low levels. Thus, the northern Gulf of Alaska differs from marine systems off British Columbia and the Washington and Oregon coasts where coastal upwelling occurs over many months. At these more Long-Term Ecological Change in the Northern Gulf of Alaska Robert B. Spies (Editor) © 2007 Elsevier B.V. All rights reserved.
12
Long-Term Ecological Change in the Northern Gulf of Alaska
southerly locations, northerly winds, during the late spring, summer, and early fall, support seasonal coastal upwelling that strongly promotes photosynthesis during the same time the northern shelf in the Gulf is stratified and nutrient-poor in the upper layers. These different physical characteristics tend to predict a relatively unproductive biological system to the north, yet paradoxically the Gulf is very biologically productive. The apparent resolution of this contradiction means that there must be ways other than strong seasonal upwelling for getting deep-water nutrients to the surface to supply plant life at the base of the food web in the growing periods in spring and summer. Finding these ways is a central theme of current oceanographic research on the northern Gulf. As the earth’s orbit first tilts the northern hemisphere towards and then away from the sun, light and temperature increase and then decrease. Subarctic marine organisms have adapted to optimize the production and flow of energy within marine food webs in a widely fluctuating, but generally predictable, seasonal ocean climate. Better understanding the changes in Gulf production between years, decades, and centuries rests on our knowledge of the system viewed from two broad perspectives: (1) how the communities in the Gulf of Alaska are structured and function in response to this annual cycle of physical variability, and (2) how individual organisms use the environment and other organisms during their life cycles to maintain and increase their populations. So, this chapter will explore the changing structure of the ecosystem during a typical year. We will integrate related elements to form a series of stories that begin with climate as the force of seasonal change in the oceanography, move through the annual cycle of biological production, and conclude with survival strategies of key species of fish, birds, and mammals that tie them to their habitats.
2.2. The Physical Environment of the Gulf of Alaska1 Thomas Weingartner The Gulf of Alaska is the semi-enclosed basin comprising the northeast corner of the North Pacific Ocean. Extensive mountain ranges surround all but its southern boundary, which opens onto the North Pacific Ocean and is defined by the northern boundary of the subarctic front. In the 400-km-wide subarctic front, temperatures and salinities decrease rapidly northward (Yuan and Talley, 1996). Although the front shifts, its northern boundary typically lies along 45°N. The western end of the Gulf of Alaska is ill defined, but is 165°W for the purposes of this book. The southern 1
This chapter is based in part on an earlier contribution (Weingartner, 2005), but has been expanded and revised for this synthesis.
Ecosystem Structure 13
boundary of the Gulf thus extends westward from the North American coast for more than 3000 km, so that the Gulf covers an area of about 3.4 × 106 km2 (or 4% of the North Pacific Ocean area). Its small size and open connections to the North Pacific suggest that the Gulf is neither isolated nor insulated from atmospheric and oceanographic processes that occur elsewhere over the North Pacific Ocean. As will be seen, the effects of climatic processes occurring elsewhere in the North Pacific often move into the Gulf by the semi-permanent atmospheric and oceanic circulation pathways. On the other hand, the Gulf cannot be considered as simply a passive recipient of remotely generated climate signals, for these signals are modified regionally as a consequence of the Gulf’s geologic history and its high-latitude setting. Over time, geologic forces have carved a complex coastline and bathymetry that substantially affects the movements and modifications of water and air masses transported into the region from afar. The high-latitude Gulf has relatively cool air and water temperatures, which affects seawater density and the regional hydrologic cycle through partitioning of the atmosphere’s moisture load into snow or rain. Thus, an appreciation of both remote and regional atmospheric and oceanographic processes is required to properly understand the physical processes responsible for the present-day structure of the Gulf of Alaska ecosystem and its response to natural and anthropogenic perturbations. We will first describe the geomorphology of the Gulf of Alaska and then address regional atmospheric conditions. Then the major oceanographic features and the physical processes that influence its marine production will be described. The chapter concludes with a brief description of the atmosphere system of the North Pacific Ocean, emphasizing processes that influence the Gulf of Alaska.
2.2.1. Geomorphology The Gulf of Alaska encompasses several bathymetric provinces, including the abyssal plain, the continental slope, and the continental shelf (including adjoining bays and fjords). The abyssal plain underlies more than 75% of the Gulf’s area and shoals from 4000 m in the southwest Gulf to about 2500 m in the northern Gulf and along the continental rise (Fig. 2.1). Several fracture zones rupture the plain, and numerous seamounts or guyots (with some rising to within 1000 m of the surface) are scattered throughout this domain. Depths shoal rapidly from 2500 m to 250 m, over a distance of 100 km or less, across the steep continental slope, which connects the abyssal plain to the continental shelf. The shelfbreak is between 200 and 300 m deep and delineates the abrupt transition in bottom slope at the juncture of the continental slope and the more gradually sloping continental shelf. The area of the continental shelf is about 3.7 × 105 km2 (Lynde, 1986) or 12.5% of the shelf area of the United States (McRoy and Goering, 1974). The shelf varies in
14
Long-Term Ecological Change in the Northern Gulf of Alaska
Figure 2.1: The bathymetry of the Gulf of Alaska and the topography of the surrounding land (photograph of globe at Geophysical Institute, University of Alaska, courtesy of A. Springer). A = abyssal plain, AR = Aleutian Range, CI = Cook Inlet, CR = Cascade Range, CS = continental slope, KI = Kodiak Island, P = Prince William Sound, QC = Queen Charlotte Islands, SB = shelf break, and SH = Shumagin Islands. width from a minimum of about 5 km near the Queen Charlotte Islands in Southeast Alaska to a maximum of more than 200 km northeast of Kodiak Island. Shelf depths are typically between 150 and 250 m, and bottom depths are often this deep to within even a few kilometers of the coast. The Gulf of Alaska shelf thus differs substantially from most other continental shelves, where bottom depths are generally much shallower than 200 m and which generally shoal smoothly from the shelfbreak to the coast. The Gulf is one of the most tectonically active zones on earth (Jacob, 1987) because the region straddles the convergent Pacific and North American lithospheric plates. The seismic, tectonic, and volcanic activities associated with the intersecting plates are responsible for much of the regional geomorphology, including the system of mountains bounding the coast. These consist of the northward extension of the Cascade Range that arc northward from British Columbia and thence across south–central Alaska and the Aleutian Range along the Alaska Peninsula. Mountain elevations in the eastern and northern Gulf range from 3 to 6 km, whereas those in the Aleutian Range are about 1 km. The mountains are young, rugged, sparsely vegetated, and, because they easily erode, supply abundant sediments to the ocean. The seafloor and coastline are continuously reshaped through faulting, subsidence, landslides, tsunamis, and subsea turbidity flows. As recently as the Great Alaska Earthquake of
Ecosystem Structure 15
1964, portions of the shelf were uplifted by as much as 15 m (Malloy and Merrill, 1972; Plafker, 1972; von Huene et al., 1972). The mountains support numerous glaciers, covering nearly 20% of the Gulf of Alaska watershed (Royer, 1982), and make the region the third largest glacial field on the planet (Meier, 1984). Glacial mass varies through time; in some years, there is a net increase in glacial ice through accumulation of precipitation in the form of ice and/or snow, whereas in other years, glacial mass decreases as the accumulated snow and ice from the past melts and is released into the ocean. Glacial advance and retreat has occurred repeatedly throughout geologic time, but, today, the glacial fields surrounding the Gulf are retreating, and the retreat rate appears to have nearly doubled in the past 25 years (Arendt et al., 2002). Glacial scouring of the underlying bedrock is an important source of fine-grained sediments to the Gulf (Hampton et al., 1987). The major inputs of glacial silt are the Bering and Malaspina glaciers, the Alsek and Copper rivers in the northern Gulf, and the Knik, Matanuska, and Susitna rivers that empty into Cook Inlet in the northwest Gulf (Hampton et al., 1987). The sediment flux is enormous; for example, the Copper, Susitna, and Stikine rivers drain watersheds, which, in aggregate, amount to less than 4% of the area of the Mississippi River basin. Nevertheless, these rivers discharge nearly a third of the sediment load carried by the Mississippi (Wang et al., 1988). The bathymetry of the Gulf of Alaska reflects the diverse tectonic and glacial processes that have operated over the region for millions of years. The fracture zones of the deep basin are aligned along faults and spreading zones within the Pacific plate, and the seamounts are remnants of volcanoes. The shelf bathymetry is particularly complex; the numerous troughs and canyons that cross the shelf are potentially important pathways for water exchange between the shelf and the basin. Extensive channel systems thread through the island archipelago of Southeast Alaska and the Shumagin and Semidi island groups along the Alaskan Peninsula. Subsea embankments and ridges abound as a result of subsidence, uplift, and glacial moraines, and some of these, such as Portlock Bank northwest of Kodiak and Alsek Bank south of Yakutat, are important commercial fishing grounds. Similar geological processes carved the fjords, bays, and headlands along the Gulf’s immensely convoluted coastline. In addition to influencing the ocean circulation, these various geological features provide a diversity of biological habitats. Two of the more prominent embayments are Prince William Sound at the apex of the Gulf and Cook Inlet in the northwest (Fig. 2.2). The sound is about 60 km wide by 90 km long, and has characteristics of a small inland sea (Muench and Heggie, 1978). Numerous islands are scattered throughout this basin, and fjords that are fed glacially thread inland along the perimeter of the rugged coastline. The sound communicates with the shelf through Hinchinbrook Entrance in the east and through several passes in the southwest, with Montague Strait being the most prominent of these. Hinchinbrook Entrance connects the shelf with the sound’s central basin, where depths
16
Long-Term Ecological Change in the Northern Gulf of Alaska
Figure 2.2: The bathymetry of Cook Inlet: Shelikof Strait (left) and Prince William Sound (right). For Cook Inlet: K = Knik Arm, KE = Kennedy Entrance, MC = Middle Cook Inlet; LCI = Lower Cook Inlet; SS = Shelikof Strait; T = Turagain Arm, and U = Upper Cook Inlet. For Prince William Sound: HE = Hitchinbrook Entrance, KP = Knight Island Passage, MS = Montague Strait, and V = Valdez Arm.
exceed 350 m. The northern edge of this basin joins the upper sound through a 300-m-deep trough leading toward Valdez Arm. A second arm curves around to the northwest, where it broadens to form a smaller basin linking the upper sound to Knight Island Passage and the passes along the southwest boundary. Depths exceed 700 m in the small basin at the northern end of Knight Island Passage, and depths within this passage range from 300 to 600 m. In contrast, the shelf immediately south of the sound is shallower and bathymetrically simpler. For example, the shelf is about 120 m deep and flat toward the south of Hinchinbrook Entrance, except in Hinchinbrook Canyon, where depths exceed 200 m. The canyon extends inshore from the shelfbreak to Hinchinbrook Entrance, and, thus, provides a conduit through which continental slope waters can feed the deeper waters of the sound. Cook Inlet extends inland about 275 km from its mouth at the tip of the Kenai Peninsula to its head near Anchorage (Fig. 2.2b). The upper inlet is about 30 km wide and extends northward from the Forelands for 75 km before dividing into Turnagain and Knik arms, each of which protrudes inland an additional 70 km. The upper inlet, including both arms, has depths of 40 m or less and contains extensive tidal flats that are usually exposed at low tide. Lower Cook Inlet is nearly 70 km wide and contains a 100-m-deep channel along its central axis. The mouth of the inlet opens onto the continental shelf at Kennedy Entrance on its eastern edge and Shelikof Strait to the southwest. This strait is a 200-km-long by 50-km-wide rectangular channel, separating Kodiak Island from the Alaska Peninsula, with depths between
Ecosystem Structure 17
150 and 300 m. South of Kodiak Island, the strait’s deep channel veers southeastward and intersects the continental slope west of Chirikof Island. Further southwest, the shelf shallows (100–150 m) and is interrupted by the Shumagin and Semidi island groups. The shelf narrows at its far western end and effectively terminates in Unimak Pass. This narrow (15 km) and shallow (75 m) strait connects the Gulf of Alaska shelf to the Bering Sea (Schumacher et al., 1982; Ladd et al., 2005).
Local Atmospheric Forcing The storm systems that comprise the Aleutian Low Pressure System primarily determine atmospheric conditions in the Gulf of Alaska. Storm frequency and intensity vary seasonally, however, because of the variations in cyclogenesis (low-pressure formation) and under the influence of the Siberian and East Pacific High pressure systems (Wilson and Overland, 1987). Mean monthly maps of sea level pressure (Fig. 2.3) for January, April, July, and October illustrate the seasonality in the strength and
Figure 2.3: Mean monthly sea level pressure fields over the North Pacific Ocean (mb = millibar). The months selected are representative of winter (December–February), spring (March–May), summer (June–August), and fall (September–November).
18
Long-Term Ecological Change in the Northern Gulf of Alaska
position of these pressure systems. From fall through winter, the Aleutian Low deepens, and its center moves westward. Simultaneously, the East Pacific High weakens and migrates southward offshore of California, while pressure builds within the Siberian High as polar air masses accumulate over northeastern Siberia. Occasionally, frigid arctic air spills southward over the Bering Sea and/or the Alaska mainland, spawning storms in the Gulf (Winston, 1955; Businger, 1991). These “polar lows” storms can be particularly dangerous because they are short-lived and difficult to predict, but often include strong winds and sub-freezing air temperatures that can cause vessel icing. From May through August, both the Aleutian Low and Siberian High weaken, while the East Pacific High strengthens and migrates northward to about 40°N. As a consequence of these seasonal changes, storms tend to follow a more southerly path in winter and a more northerly path in summer and fall, with some continuing on into the Chukchi Sea and Arctic Ocean. In all seasons, however, the majority of storms enter the Gulf from the west, where they stall and dissipate because the coastal mountains contain them. Indeed, the frequency with which this occurs led Russian mariners to refer to the region as the “graveyard of lows” (Plakhotnik, 1964). The oceanographic characteristics of the Gulf of Alaska shelf result, to a large extent, from the interaction between the mountains and the storms moving into the Gulf of Alaska. Orographic effects profoundly influence many of the more important physical variables that mold this marine ecosystem, including precipitation rates and types, coastal runoff, wind velocity, and the heat exchange between the atmosphere and ocean.
Precipitation and Runoff Uplift and cooling of moist air by the mountains lead to condensation and heavy precipitation rates around the coast. Measured precipitation rates of 2 to 4 m year−1 are typical throughout the region, although rates in southeast Alaska and Prince William Sound can easily exceed 4 m year−1 (Fig. 2.4). Precipitation rates at higher elevations have not been measured, but are certainly greater because of greater cooling. With the exception of the Alaska Peninsula, coastal precipitation rates are much greater than rainfall rates over the basin and on the shelf. For example, precipitation at Middleton Island, which lies 100 km south of Prince William Sound on the northern shelf, are about 1.5 m year−1 (Danielson et al., in prep., unpublished data), and similar to those over the central Gulf (Baumgartner and Reichel, 1975). Unless stored as snow or incorporated into glaciers, the precipitation swiftly returns to the ocean via the many small streams that drain the steep coastal mountains. Royer (1982) estimates that, on annual average, approximately 24,000 m3 s−1 (750 km3 yr−1) of freshwater enters the Gulf of Alaska shelf between Ketchikan and Seward, with roughly 60% entering Southeast Alaska and the remainder along the south-central coast. The actual amount of discharge onto the shelf is undoubtedly greater because this estimate does
Ecosystem Structure 19
Figure 2.4: Circulation schematic for the Gulf of Alaska, including the basin current structure (North Pacific Current, Alaska Current, and Alaskan Stream) and the Alaska Coastal Current on the continental shelf. The vertical bars indicate the annual precipitation rate compiled from historical precipitation measurements.
not incorporate the discharge from rivers draining interior Alaska (Alsek, Copper, Susitna, etc.), the freshwater discharge along the British Columbian coast, or the altitudinal influence on precipitation rates. Nevertheless, the magnitude and the diffuse or distributed manner in which the discharge enters the Gulf of Alaska is extraordinary when compared to other continental shelf systems, where freshwater enters through one or a few, large “point” sources. In fact, the discharge into the Gulf of Alaska is much greater than the other mid- and high-latitude watersheds, and it exceeds that of the Mississippi River by 40% (Table 2.1). More striking is the disparity in yields; the yield for the Gulf of Alaska is 2–10 times greater than the other watersheds tabulated. Indeed, based solely on this criterion, the Gulf of Alaska coastal watershed has more in common with equatorial watersheds, such as the Amazon River, than it does with mid- and high-latitude drainage basins. Two other sources of freshwater to the continental shelf are the precipitation falling directly onto the shelf and glacial runoff. The former is estimated to be 47 km3 yr−1 (Weingartner et al., 2005), while recent measurements by Arendt et al. (2002) indicate that the glacial meltwater contributes 97 km3 yr−1 to the Gulf of Alaska shelf. In summary, estimates of the freshwater influx to the Gulf of Alaska shelf are uncertain, but probably exceed 1000 km3 yr−1. Moreover, the annual coastal freshwater flux is 50% greater than the total amount of precipitation falling over the Gulf of Alaska but is discharged directly onto the shelf, which contitutes only about 10% of the Gulf’s area.
20
Long-Term Ecological Change in the Northern Gulf of Alaska
Table 2.1: Discharge from some large watersheds.
Watershed Mackenzie Mississippi Yukon Columbia Amazon Gulf of Alaska*
Drainage Area (km2) 1,800,000 3,200,000 850,000 700,000 7,200,000 480,000
Mean Annual Discharge (km3 year–1) 237 536 221 221 5680 750
Yield = Discharge/Area (cm year–1) 13 17 26 32 80 160
*Royer, 1982.
The mean monthly values of freshwater runoff as estimated by Royer (1981a,1982) for the coastal segment between Ketchikan and Seward are given in Fig. 2.5. Discharge varies tremendously through the year with a maximum (~40,000 m3 s−1) in fall and a minimum (~10,000 m3 s−1) in winter when much of the precipitation is stored as snow. Discharge then increases with the onset of spring melt in May and increases more rapidly in August as the rainy season commences. Freshwater discharge to the ocean is important because it affects seawater density. Density is a function of temperature, salinity, and pressure, although the pressure influence is generally small over the depths of interest considered here and will be neglected hereafter. As will become evident later, density gradients (density changes over vertical or horizontal distances) are of fundamental importance in a variety of processes that affect marine ecosystems. Although both temperature and salinity affect density gradients, the salinity effect is generally of greater importance in the Gulf of Alaska because water temperatures vary little throughout the year, except near the surface in summer. Hence, the massive coastal freshwater discharge into the Gulf of Alaska modifies ocean salinities and, in so doing, profoundly affects vertical and horizontal density gradients. Vertical density gradients, or ocean stratification, affect biological production and mixing. Horizontal gradients are indicative of horizontal pressure gradients that force horizontal currents. (We will refer to such currents in this book as density gradient flows.) While the ocean’s density structure is strongly dependent on the discharge, winds and temperature changes induced by heat exchanges between the atmosphere and ocean also affect the density structure. Winds modify density gradients through vertical stirring (mixing) and through advection of waters from elsewhere.
Ecosystem Structure 21
Figure 2.5: Mean monthly coastal freshwater discharge (after Royer, 1981a, 1982) and mean monthly wind speed from Middleton Island on the northern Gulf of Alaska shelf. Negative wind speed values imply that the mean winds are westward.
Winds The winds associated with low-pressure systems over the Gulf of Alaska blow counterclockwise (with the coast to the right of the direction of the wind), and frictional coupling between wind and water (stress) forces the ocean circulation. Wind stress is proportional to the square of the wind speed and forces a surface Ekman transport (typically confined to the upper 10 to 30 m of the water column), which is directed 90° to the right of the wind in the Northern Hemisphere. On the continental shelf, the surface Ekman transport induces a sea level rise (relative to offshore) along the coast and sinking (downwelling) of surface waters within a narrow coastal band. The cross-shore sea level slope is small (amounting to a difference of only a few centimeters over 100 km), but it results in a cross-shore pressure gradient that, in conjunction with the earth’s rotation, propels a steady along-shelf current in the downwind direction throughout the water column. Because most of the surface Ekman transport sinks along the coast, there must be a compensatory offshore flow at depth (otherwise the sea level would continue to rise along the coast for as long as the wind blew). Under relatively steady wind forcing, the offshore flow is contained within a bottom Ekman layer established by frictional stress between the along-shelf flow and the seabed. An alongshore wind blowing with the coast to its left is an upwelling-favorable wind. Under such conditions, coastal sea level decreases, and the alongshore flow and surface and bottom Ekman transports are opposite in direction from the
22
Long-Term Ecological Change in the Northern Gulf of Alaska
downwelling case. Such winds also induce an uplift of deeper waters to the surface within a narrow coastal band; a process referred to as upwelling. As will be seen later, vertical and horizontal differences in water density alter this circulation structure in important ways, although the basic features outlined earlier are nonetheless retained. As an example of the annual cycle in wind forcing over the northern Gulf of Alaska, Fig. 2.5 shows the mean monthly along-shelf wind velocities measured at Middleton Island. The monthly averages, based on long-term measurements at Middleton, show that downwelling winds prevail throughout the year and are maximum in winter and minimum in summer. Spatial and temporal variations in wind stress complicate the circulation response outlined earlier, however, and the Middleton Island winds are not necessarily representative of conditions around the shelf. Unfortunately a detailed understanding of the temporal and spatial wind field over the Gulf of Alaska has yet to be achieved, because, with the exception of Middleton Island, direct, long-term wind measurements on the continental shelf and basin are scarce. Although there are numerous coastal communities in which winds have been measured for many years, these measurements are influenced by local orographic effects that make extrapolation of coastal measurements onto the shelf problematic. Our present understanding of spatial and temporal variations of the shelf wind field is instead derived from limited-duration measurements obtained from moored buoys and/or inferred from wind estimates generated by weather forecast models. In a general sense, these indicate that temporal variations are correlated over broad areas of the Gulf (especially in fall and winter, when winds are strongest) because storms entering the Gulf are comparable to the size of the basin itself (Livingstone and Royer, 1980) (Fig. 2.6). This result does not imply, however, that the wind field is spatially uniform around the shelf. Indeed, it appears that whereas, on monthly average, downwelling-favorable winds blow over the northern and eastern Gulf of Alaska year-round, the maximum in wind stress occurs in the northeast Gulf. Fall and winter wind directions are more variable in the western Gulf because storms entering the Gulf follow two predominant pathways: a west – east track across the Alaska Peninsula from the Bering Sea or a southwest – northeast track across the Gulf of Alaska. As a result, winter wind stress over the shelf between Kodiak and Unimak Pass appears to be upwelling favorable on average (Stabeno et al., 2004). Winds weaken throughout the Gulf in summer and, while downwelling winds prevail over the northern Gulf, upwelling-favorable winds occur along the British Columbian shelf and southwest Gulf of Alaska. While the mean monthly wind estimates provide a useful description of the annual cycle, these estimates belie the episodic nature of wind forcing, which is tied to storm systems and weather fronts propagating across the Gulf every 3 to 10 days. Particularly in summer, there are prolonged periods of relative calm or weak, upwelling-favorable winds that are interrupted by relatively strong downwelling winds associated with only a few storms of several days duration in each month.
Ecosystem Structure 23
Figure 2.6: Low-pressure system in the Gulf of Alaska. The MODIS instrument aboard NASA’s Aqua satellite captured this true-color image of a large low-pressure system spinning in the Gulf of Alaska on August 17, 2004. The spinning cloud systems of these lows express strong, persistent, low-pressure systems that form because of the persistent flow of semi-permanent pressure systems north (the polar easterlies) and south (the subtropical high) of the area. Image courtesy Jesse Allen: http://earthobservatory.nasa.gov/NaturalHazards/Arch.
Important wind stress variations also occur at smaller (mesoscale) spatial scales (30–100 km). These are especially prominent in winter and are closely tied to orographic effects and the channeling influence of straits and the mouths of bays and fjords. Mesoscale winds are enhanced in winter when the coastal mountains inhibit mixing between the cold, dry air of Alaska’s interior and the warmer, moister marine air. Mountain passes provide a conduit through which continental air pours across the shelf and collides with marine air (Macklin et al., 1988). These cold-air outbreaks produce offshore-directed gap winds at the mouths of coastal embayments, including Hinchinbrook Entrance, the Copper River delta (Macklin et al., 1988), Shelikof Strait (Lackmann and Overland, 1989; Macklin et al., 1984), lower Cook
24
Long-Term Ecological Change in the Northern Gulf of Alaska
Inlet, and the Barren Islands (Macklin et al., 1990). Under similar winter conditions, the coastal mountains can also enhance the alongshore wind component under downwelling-favorable situations. In this case, the shelf wind field includes a narrow (50–100 km wide) coastal barrier jet (Parish, 1982; Overland and Bond, 1995) having wind speeds several times greater than those further offshore. Mesoscale wind fields affect the coastal ocean in two significant ways. First, the barrier jets and the largergap wind events (such as those flowing through Shelikof Strait and/or offshore of Cook Inlet) can substantially affect the cross-shelf and along-shelf current structure. (The smaller-scale-gap winds at the mouth of bays are unlikely to have a significant influence on the shelf circulation, although they could be locally important in influencing exchange between the shelf and the bay or fjord.) Second, winter mesoscale winds are likely important in cooling the coastal ocean because the rate at which heat is lost from the sea is a function of both wind speed and the air–sea temperature difference. Although Overland and Heister (1980) suggest that these wind events are common in winter, the frequency of these phenomena and their importance in the coastal ocean has not been quantified. It also seems probable that the frequency and intensity of mesoscale wind phenomena will vary from year to year in accordance with atmospheric conditions in the Alaskan interior and the Gulf.
Heat Fluxes Changes in upper-ocean temperatures over both the basin and the continental shelf are largely controlled by the exchange of heat between the atmosphere and ocean. This exchange depends on a number of variables, including the air and sea temperatures and their difference, atmospheric humidity, wind speed and direction, and the sky or cloud cover. Sky cover affects the amount of radiation available to heat the upper ocean and sustain marine photosynthesis. Clouds form as moist air rises, expands, and cools, either during cyclogenesis or over the coastal mountain ranges. Sky cover over the Gulf of Alaska is extensive and varies little seasonally. Indeed, clouds obscure 60% or more of the sky more than half of the time, while a sky cover of 25% or less occurs only about 15% of the time (Brower et al., 1988). Because of its high-latitude location, the influx of solar radiation into the Gulf varies from a maximum of about 225 W m−2 at the summer solstice to about 10 W m−2 at the winter solstice (Fig. 2.7). (Over a column of water 30 m deep and 1 m2 in surface area, solar radiative heating at the summer solstice would lead to a 1.6°C rise in temperature in 10 days, whereas the winter solar radiative flux would increase temperatures by 0.08°C over the same period.) Interannual variations in skycover, especially in summer, can affect upper-ocean temperatures. As an extreme but not unrealistic example, a 25% reduction in summer cloud cover could increase temperatures in the upper 30 m of the water column (where most of the solar radiation is absorbed)
Ecosystem Structure 25
Figure 2.7: Mean monthly heat fluxes for short-wave, long-wave, latent, sensible, and the net or the sum of all terms for the Gulf of Alaska.
by 1–2°C. Although this temperature increase appears slight, it could be important to marine invertebrates whose metabolic rates are temperature dependent. Heat is lost from the ocean to the atmosphere by sensible and latent heat exchange and long-wave radiation. Each of these processes causes cooling in the Gulf of Alaska throughout the year (Fig. 2.7). The heat transfer rate for latent and sensible heat exchanges increases as both the air–sea temperature contrast and the wind speed increases. Long-wave, radiative heat loss occurs because the ocean surface emits radiation to the atmosphere at a rate proportional to the surface temperature, although the effectiveness of this cooling is reduced by clouds. As with much of the earth’s surface, the ocean radiates at infrared wavelengths that, on absorption by water vapor, carbon dioxide, and other gases, lead to atmospheric warming, e.g., the “greenhouse gas” effect. While long-wave heat loss is relatively invariant from month to month, the sensible and latent heat fluxes vary substantially throughout the year and are greatest in winter (when air–sea temperature differences and wind speeds are greatest) and least in summer. The net heat flux (based on the sum of all the heat fluxes) results in a distinct annual heating and cooling cycle that coincides with the solar cycle. Thus, the ocean gains heat between the spring and fall equinoxes and loses heat to the atmosphere over the rest of the year (Fig. 2.7). Averaged over the year, there is a net heat loss of 13 W m−2 over the northern Gulf of Alaska shelf to the atmosphere, as estimated from Middleton Island data. Most likely, the nearshore winter air temperatures are cooler
26
Long-Term Ecological Change in the Northern Gulf of Alaska
and wind speeds higher than those measured at Middleton, so that inner-shelf winter heat losses are probably appreciably greater. Consequently, inner-shelf waters (including Prince William Sound, Cook Inlet, and Shelikof Strait) are generally colder than outer-shelf waters in winter. Nevertheless, the annual imbalance between summer heat gain and winter heat loss based on Middleton Island data suggest an annual decrease in shelf water temperatures of at least 0.8°C over the upper 100 m. This implies that on long-term average, the ocean circulation must advect into the Gulf the heat required to balance that lost annually to the atmosphere.
2.2.2. Physical Oceanography The Seasonal Cycle of Water Properties The annual cycles of wind, freshwater inputs, and air–sea heat flux are reflected in seawater temperatures and salinities on the shelf. These changes can be seen in the 2.5-year record of hourly temperature and salinity values (Figs. 2.8 and 2.9) collected at the hydrographic station GAK 1, which lies at the mouth of Resurrection Bay offshore of Seward. The plots accompanying the mean monthly values show the amplitude and phase for the annual cycle as a function of depth based on a least squares fit to the annual cycle. (The leftmost plot shows the mean annual temperature and the range. The middle plot shows the phase or the time of the year when the maximum temperature or salinity occurs. The rightmost plot shows the variance accounted for by the annual period.) The annual period accounts for 90% (at the surface) to 50% (at the bottom) of the variance for both temperature and salinity (Weingartner et al., 2005). Mean sea-surface temperatures range from a minimum of 3.5°C in March to 14°C in August. The range of temperatures decreases with depth, however, and is less than 1°C annually at depths greater than 150 m. Although the lowest sea surface temperatures coincide with the end of the cooling season, the highest surface temperatures occur in August, about one month before the end of the heating season in late September (Fig. 2.7). Surface temperatures begin to decline in September because increasing wind-mixing and downwelling redistribute heat over the upper portion of the water column. These processes lead to the development of a subsurface temperature maximum in October. From fall through early winter, temperatures within the subsurface maximum temperature decrease, and the depth of this layer increases such that bottom water temperatures attain their annual maximum in January. Bottom water temperatures exceed those at shallower depths from January through April. Temperatures decline throughout the water column through April because of deep winter mixing and vigorous surface cooling. By late April or early May, near-surface temperatures begin to increase as day length increases, so that the water column becomes nearly isothermal in early May. Through midsummer, temperatures increase rapidly at the surface and more slowly at mid-depth, while temperature remains nearly constant at depths greater than 100 m.
Ecosystem Structure 27
Figure 2.8: Water temperatures (December 1999–March 2002) at selected depths (upper panel) at the hydrographic station GAK 1 near Seward, Alaska. The lower panel shows the results of a least squares fit to the annual harmonic for each calendar year (2000 and 2001) at each measured depth. The leftmost panel shows the mean (solid circle) and amplitude (horizontal bars), the middle panel, the phase (time of the maximum temperature at the specified depth), and the rightmost panel, the fraction of the total variance explained by the annual harmonic (from Weingartner et al., 2005).
28
Long-Term Ecological Change in the Northern Gulf of Alaska
Figure 2.9: Water salinities (December 1999–March 2002) at selected depths (upper panel) at hydrographic station GAK 1 near Seward, Alaska. The lower panel as in Fig. 2.8 (from Weingartner et al., 2005).
Ecosystem Structure 29
Surface water is saltiest in April and freshest in August. The range of surface water salinity is from 31 to about 25 PSU (practical salinity unit). Although surface salinity begins to increase in September, salinities in the upper 50 m decrease through early November because wind mixing and downwelling redistribute salt and freshwater over the upper layers. Consequently, the upper-layer salinities vary nearly in-phase with the annual discharge cycle. At depths greater than 100 m, however, minimum salinities occur in late winter following maximum downwelling, and maximum salinities occur in summer when downwelling is weak. Thus, the annual cycle in salinity over the lower half of the water is out-of-phase with the surface salinity cycle but in-phase with the annual cycle in along-shelf winds. The summer salinity increase below 150 m heralds the arrival of salty, nutrient-rich water from offshore that spreads inshore from the continental slope (Royer, 1975; Xiong and Royer, 1984; Weingartner et al., 2005). The near-bottom winter salinity minimum results from strong fall and winter downwelling, which gradually flushes deeper, more-saline waters offshore and/or mixes them with fresher surface waters. The annual temperature and salinity cycles described in the preceding text are characteristic of the Gulf of Alaska shelf in general, although there are important spatial differences. For example, at distances of more than 50 km from the coast, the annual range in salinity is generally about 1 PSU or less at all depths. Surface salinities in glacial fjords and bays (e.g., Cook Inlet) or near the mouths of rivers entering the shelf can be substantially fresher than elsewhere along the coast. In general, winter surface water temperatures are cooler along the coast than further offshore for the reasons discussed earlier. There are also alongshore gradients in these variables; for example, winter surface water temperatures decreasing along the coast and continental slope between Southeast Alaska and the Alaska Peninsula. Seasonal changes in temperature and salinity in the water column affect the density stratification, which in turn affects the extent of vertical mixing. A strongly stratified water column requires more energy to mix than a weakly stratified one. Seasonal changes in stratification affect photosynthesis by phytoplankton, which require both light and nutrients. In general, plant nutrient concentrations increase with depth, while the photosynthetically important wavelengths of light are rapidly attenuated with depth. When stratification is weak, deep mixing removes phytoplankton from the euphotic zone, the depth layer in which there is sufficient light for plant growth. On the other hand, this mixing supplies the euphotic zone with nutrients. The water column is weakly stratified from January through March, so deep mixing enriches the surface layers with nutrients. The decrease in wind speeds and the increase in runoff and solar heating in April and/or May trigger stratification. When the stratification is strong enough to prevent mixing to depths greater than the euphotic zone depth, a spring bloom of phytoplankton occurs (see Section 2.3 for a more detailed explanation and figures). Although stratification is building in early spring, storms can interrupt the process by inducing deep mixing. Nevertheless, by late May, the
30
Long-Term Ecological Change in the Northern Gulf of Alaska
shelf is generally stratified everywhere, with stratification strengthening through summer. Consequently, from late spring through early September, the shelf and basin consists of a shallow, well-stirred, surface-mixed layer about 20 m thick, over which salinity and temperature are nearly constant. The summer mixed layer is separated from deeper layers by a thin, vertical layer across which density changes very rapidly (the pycnocline). The summer pycnocline is typically between 15 and 40 m in depth and coincides with both a thermocline and a halocline. While nutrient concentrations are nearly depleted in the summer mixed layer due to phytoplankton consumption in spring, high concentrations are found just beneath the pycnocline. Stirring of the water column, by winds, entrainment of deeper water into a spreading plume of low-density water, or tidal mixing processes, can inject nutrient-rich water into the euphotic zone and restimulate plankton production. As winds intensify in fall, the stratification erodes due to both stronger vertical mixing and, at least near the coast, increased downwelling. There are important spatial variations in the onset of springtime stratification and in the structure of the seasonal pycnocline on the Gulf of Alaska shelf. The inner-shelf, bays, fjords, and Prince William Sound stratify first because coastal freshwater runoff is confined initially to nearshore regions and only gradually spreads offshore, if at all, through ocean advection and horizontal mixing processes. Solar heating provides additional surface buoyancy by warming the upper ocean layers over both the shelf and basin. However, thermal stratification often remains weak until May, so that the inner shelf stratifies earlier than the outer shelf and the northern Gulf of Alaska basin. While the springtime onset of stratification on the inner shelf depends primarily on freshwater runoff, surface heating primarily controls seasonal variations in stratification over the outer shelf, slope, and basin. Spatial or temporal variations in the timing, strength, or frequency of stratifying processes (warming, freshwater runoff) and destratifying processes (winds, tidal mixing) consequently affect phytoplankton production and, thus, the amount of food energy available for higher trophic levels. The annual cycle of nutrients is intimately linked to seasonal changes in water column stratification, temperature, and salinity. Childers et al. (2005) described the annual cycle of nitrate, silicate, phosphate, and ammonium across the shelf. These changes are similar to those at station GAK 1, as shown in Fig. 2.10. Uniform nutrient concentrations occur throughout the water column in March prior to the onset of the spring bloom. Except for ammonium, these are rapidly drawn down by phytoplankton consumption in the upper 50 m by late May and remain low through October. Upper-level ammonium concentrations increase in spring by zooplankton excretion. Deep nutrient concentrations tend to follow salinity changes, with maximum concentrations occurring in summer.
Circulation over the Gulf of Alaska Shelf and Slope In a broad sense, the major circulation features of the Gulf of Alaska circulation (shown schematically in Fig. 2.4) consists of a system of counterclockwise flows on
Ecosystem Structure 31
Figure 2.10: Seasonal cycle of the macronutrients: silicate (Si[OH]4, top); nitrate (NO3, middle); and phosphate (PO4, bottom) at selected depths at station GAK 1, at the mouth of Resurrection Bay, near Seward.
the shelf, slope, and central basin. There are, however, substantial differences in the structure of the currents and the physics controlling these various flows, and, for this reason, the circulation in each region is described separately in the following subsections. Flow on the inner shelf consists of the Alaska Coastal Current (ACC), while it includes the Alaska Current along the continental slope (in the eastern and
32
Long-Term Ecological Change in the Northern Gulf of Alaska
northeastern Gulf) and its transformation into the Alaskan Stream (in the northwestern Gulf). Both these current systems are swift (compared to flow in the central basin) and extend over a broad alongshore domain. By the nature of their geographical extent, these currents are likely important in conveying climate perturbations around the Gulf and, ultimately, into the Bering Sea. Although the circulation within the central basin is much feebler than that of shelf or slope, it involves a large mass of water and, hence, climate signals introduced into the basin interior may remain for a relatively long time. Each of these current systems and their advected water masses respond on seasonal and longer time scales to winds, coastal freshwater discharge, heat exchange with the atmosphere, and the advection of water masses formed elsewhere in the North Pacific Ocean. The water masses and circulation in the Gulf of Alaska shelf define three domains: the Alaska Coastal Current domain on the inner shelf, a mid-shelf region that extends from the offshore edge of the ACC to near the outer shelf, and an outer shelf region, which includes the shelfbreak and continental slope. These regions are separated from one another by frontal systems that act as semipermeable boundaries limiting exchanges of mass, material, and momentum. Frontal strength varies in accordance with the magnitude of horizontal density gradients, which, in the Gulf of Alaska, is primarily a consequence of salinity gradients. Fronts also support a three-dimensional flow field. However, the strongest component of flow is parallel to the front and oriented such that low-density water is located to the right of the flow direction in the Northern Hemisphere. As a preface to subsequent discussions, we present schematics of the circulation and frontal structure on the Gulf of Alaska shelf for fall through spring and late spring through early fall (Fig. 2.11). From fall through spring, relatively strong alongshelf winds impel an onshore Ekman transport of surface waters and downwelling near the coast. This process traps low-salinity waters (freshened by coastal runoff) on the inner shelf, forming a deep front within 35 km of the coast. The main axis of the ACC coincides with the front, and the region inshore of and including the front constitutes the ACC domain. A second front forms at the shelfbreak and separates the mid-shelf domain from the outer shelf, which includes the Alaska Current/Stream flowing along the continental slope. The shelfbreak front, which may or may not have a surface expression, is anchored to the shelfbreak and inclined in the offshore direction. Weak upwelling occurs on the inshore side and at the base of the front (Gawarkiewicz and Chapman, 1992; Pickart, 2000). In summer, when runoff is increasing but winds are weak, the ACC front is shallow and may spread 50 km or more offshore. With the reduction in downwelling (or an increase in upwelling), the alongshore flow over the shelf and slope weakens, causing the foot of the shelfbreak front to shoal and reattach at a shallower isobath inshore (Chapman, 2000; Weingartner et al., 2005). The cross-shore flow fields associated with surface and bottom Ekman transports also change seasonally. In winter, the surface onshore Ekman transport is compensated
Ecosystem Structure 33
Figure 2.11: Schematic of the circulation in fall, winter, and spring (left) and summer (right) over the shelf and slope of the northern Gulf of Alaska. The view is looking westward with the coast (and north) to the right. Cross-shelf flow and vertical flows are indicated by arrows. Flows over the middle shelf are variable, although weakly westward on average.
for by an offshore flow along the bottom and/or within the interior. In summer, the onshore Ekman transport is much weaker, but there is also an onshore flow at the bottom associated with the shoreward movement of the base of the shelfbreak front. A compensatory return flow presumably occurs in the interior, although how this is accomplished on the Gulf of Alaska shelf is not understood. Although these are the major seasonal differences in frontal structure over the Gulf of Alaska shelf, intraseasonal frontal variations can be substantial due to fluctuations in winds, alongshore variations in bathymetry, meanders in the ACC and slope flows, and the passage of large eddies along the continental slope. The magnitude of current speeds in the along-shelf, cross-shelf, and vertical directions are quite different from one another. Cross-shelf currents associated with Ekman transports are only a few cm s−1 and smaller (by a factor of 10–100) than the alongshore shelf currents. Vertical current speeds are even smaller, being a few meters or even less per day, except in the strong downwelling regions near the coast during intense storms when vertical speeds can be as large as 10–20 m/day.
34
Long-Term Ecological Change in the Northern Gulf of Alaska
The Alaska Coastal Current (ACC) The ACC is the most prominent aspect of the shelf circulation (Royer, 1981b; Johnson et al., 1988; Stabeno et al., 1995). Although slender, having a typical width of 35 km, it extends for over 2500 km from its probable origin on the British Columbian shelf, although in some months or years it might originate as far south as the Columbia River (Thomson et al., 1989; Hickey et al., 1991; Royer, 1998) to Unimak Pass in the western Gulf, where it leaves the Gulf of Alaska shelf and enters the Bering Sea (Schumacher et al., 1982, Ladd et al., 2005). The properties of the ACC in Southeast Alaska are poorly known, but it seems likely that much of it flows through the complex of channels threading through the islands. The current then continues westward along south-central Alaska until it reaches Prince William Sound. Much (if not most) of the ACC flow enters the sound through Hinchinbrook Entrance and exits through Montague Strait (Niebauer et al., 1994). The remainder of the current continues across the mouth of Hinchinbrook Entrance, southwestward along Montague Island, and thence westward after rounding the southern tip of this island. West of Montague Island, this branch of the ACC and the outflow from Montague Strait merge to continue westward along the south coast of the Kenai Peninsula. The ACC apparently splits northeast of Kodiak Island (Stabeno et al., 1995), with some of the current flowing southward along the shelf east of Kodiak Island. Most of the current curves around the mouth of Cook Inlet, however, before continuing southward through Shelikof Strait (Muench et al., 1978). As it arcs across the mouth of the inlet, bottom topography induces upwelling and locally strong tides mix salty, nutrient-rich water to the surface. Some of this upwelled water flows northward, supplying nutrients and salt (and possibly heat in winter), along the eastern shore of Cook Inlet. The inflow is gradually mixed by tides with freshwater from the rivers that enter along the sides of the inlet to form a dilute southward flow along the west side of Cook Inlet. This outflow then rejoins the ACC at the head of Shelikof Strait, with the mixture continuing flowing southward through this channel. At the lower end of Shelikof Strait, the current continues along the Aleutian Peninsula and through the Shumagin/Semidi island group en route to Unimak Pass. The seasonal transitions implied by the schematics (Fig. 2.11) are evident in the cross-shelf distribution of temperature and salinity properties shown in Fig. 2.12 for April, August, and October. (The data were collected on a cross-shelf transect on the northern Gulf of Alaska shelf offshore of Seward.) In winter (February through April), both the vertical and cross-shelf gradients of salinity and temperature are weak. The ACC front lies within about 10 km of the coast and extends from the surface to the bottom. In summer, the vertical stratification is large but cross-shelf salinity (and density) gradients are weak. At this time, the ACC front extends 50 km offshore and is usually less than 40 m deep. Vertical stratification weakens in fall, although the cross-shelf salinity gradients and the ACC front are stronger than at other times of the year. As along-shelf winds and coastal downwelling increase in
Ecosystem Structure 35
Figure 2.12: Cross-shore distributions of temperature (left) and salinity (right) in the northern Gulf of Alaska, contoured as a function of horizontal distance offshore (horizontal axis) and pressure (vertical axis). Pressure units are approximately equal to the depth in meters. The view is eastward with the coast (and north) on the left. Waters of lowest salinity are usually within 35 km of the coast and represent the Alaska Coastal Current.
fall, the front moves shoreward to within 30 km of the coast and steepens so that the base of the front intersects the bottom between the 50- and 100-m isobaths. The front deepens and steepens through January, but temperature and salinity gradients gradually weaken in response to wind mixing, surface cooling, and the reduction in runoff. A further notable feature in these figures is that cross-shelf temperature differences at all depths are generally small and rarely exceed 1–2°C. In contrast, vertical changes in temperature can be quite large, especially in summer, when a strong thermocline centered at about 25 m depth forms across the shelf. Temperatures decrease from 14–8°C across the summer thermocline. The structural changes in the ACC throughout the year are also accompanied by changes in the velocity distribution and transport. The large cross-shore density gradients
36
Long-Term Ecological Change in the Northern Gulf of Alaska
within the ACC front are associated with currents having substantial current shear. Speeds within the ACC front often exceed 30 cm s−1, but can approach 200 cm s−1 (Johnson et al., 1988) in fall, while speeds elsewhere in the ACC domain are generally 5–20 cm s−1. The shears in summer are primarily confined to the upper 50 m, with weaker (and possibly even reversed or eastward) flow throughout the rest of the water column. Along-shelf velocities increase in fall when substantial shears develop over the upper 100 m of the water column. In winter, when the current’s dynamics are primarily winddriven, vertical current shears are weak and speeds are more uniform throughout the ACC domain. Although maximum speeds are observed in the fall, it appears that maximum transport is in winter (Schumacher et al., 1989; Stabeno et al., 1995), while minimum transport occurs in summer. Stabeno et al. (1995) estimate that, on annual average, the ACC transports at least 800,000 m3 s−1, although transport variations of several days to a week can be large and exceed 3,000,000 m3 s−1. The complex vertical and cross-shore circulation cells within fronts often result in enhanced biological production and the trapping of materials within the front (Garrett and Loder, 1981; Yankovsky and Chapman, 1997; Chapman and Lentz, 1994; Chapman, 2000, Williams, 2003). This appears to be true for the ACC front as well. For example, surface drifters released seaward of the ACC front move onshore (in accordance with Ekman dynamics) and then westward upon reaching the front (Royer et al., 1979). Conversely, the surface layer spreads seaward on the inshore side of the front with the rate of offshore flow increasing with the discharge (Johnson et al., 1988; Williams, 2003). Together, these results suggest cross-frontal flow convergence arising from differing dynamics on either side of the ACC front. Freshwater runoff forces offshore surface flow on the inshore side of the front (with this influence varying throughout the year in accordance with the runoff), whereas wind forcing dominates offshore of the front. One consequence of this flow structure is that crossfrontal exchange of water and dissolved and suspended materials (including plankton) are inhibited (Csanady, 1984). A second consequence is that plankton might accumulate along the frontal boundary, possibly attracting foraging fish, seabirds, and marine mammals. The summer and fall current shears embedded in fronts might also affect predator–prey interactions. Phytoplankton, juvenile salmon, and forage fishes feed in the upper 20 m of the water column at this time of the year (Boldt, 2001; Coyle and Pinchuk, 2003) and drift with the current. However, the zooplankton, which feed on the phytoplankton and on which the fish prey, migrate diurnally over at least the upper 100 m. Hence, diurnally migrating zooplankton are unlikely to encounter the same phytoplankton patches and fish schools over a day because of the highly sheared flow.
Mid-shelf domain The mid-shelf domain is bracketed inshore by the ACC front and offshore by the shelfbreak front. The position of these fronts varies on seasonal and shorter time
Ecosystem Structure 37
scales, so that the dimensions of the mid-shelf domain also vary. The width of the mid-shelf domain will also vary in accordance with the shelf width. In the northwest Gulf of Alaska, the mid-shelf domain is about 100 km wide, but on the narrow shelf of the eastern Gulf, the mid-shelf domain might be very narrow and possibly nonexistent. Stratification over the middle of the shelf is always weaker than within the ACC, except in late winter, when both regions are weakly stratified. As with the ACC domain, salinity controls mid-shelf stratification from fall through late winter, but, in contrast to the ACC, temperature controls stratification in spring and summer. In general, crossshelf temperature and salinity gradients are relatively weak throughout the year, so that the density gradient component of flow over the mid-shelf is weaker than inshore, although still westward on average (Niebauer et al., 1981; Hermann et al., 2002). Flow over the mid-shelf domain is poorly understood, however, having received comparatively little study. Nevertheless, existing observations and models indicate that the mid-shelf domain contains energetic current variations, involving both current reversals and strong cross-shelf flows. The most likely sources of this variability are mesoscale (10–50 km) flow features and, less frequently, large (~150 km diameter) eddies that impinge on the continental slope. Mesocale variability originates in several ways, including, (1) separation of the ACC from coastal headlands with subsequent eddy formation (Ahlnäes et al., 1987; Cenedese, 2002), (2) eddy shedding or meandering by the ACC (Mysak et al., 1981; Bograd et al., 1994), (3) interactions of the shelf flow with variations in bottom bathymetry (Lagerloef, 1983), and (4) meandering of the shelfbreak front along the continental slope (Niebauer et al., 1981) induced by fluctuations in the Alaska Current/Stream. It is, however, extremely difficult to quantify mesoscale variability because it varies both spatially (depending on coastal landforms, bottom topography, and ambient currents) and temporally (depending on the winds and the shelf and slope density distribution). The transient but energetic mid-shelf currents are undoubtedly important to the shelf ecosystem by facilitating cross-shelf exchanges of organisms and dissolved and suspended materials. The mid-shelf thus serves as an important link between the inner shelf and the continental slope. Some of the eddies and current meanders include vigorous vertical motions capable of altering mixed layer depths and advecting nutrients into the euphotic zone. Moreover, until eddies dissipate, they do not readily exchange mass and materials with ambient waters and so may serve as important nurseries for some organisms. For example, in Shelikof Strait, the abundance of larval pollock is generally greater in eddies than outside them (Incze et al., 1989; Vastano et al., 1992; Schumacher and Kendall, 1991; Schumacher et al., 1993; Bograd et al., 1994). Moreover, juvenile pollock collected within eddies are in significantly better condition than those collected outside (Canino et al., 1991), consistent with these features enhancing production of both phytoplankton and zooplankton. Although the linkages between the inner and outer shelf are incompletely understood, the annual inshore incursion of the foot of the shelfbreak front is a robust
38
Long-Term Ecological Change in the Northern Gulf of Alaska
feature of the summer season, as discussed with respect to the schematic in Fig. 2.11. It is clearly evident in the August panel of Fig. 2.12 and delineated by the 32.8 salinity isopleth. The influx of high-salinity water probably exerts an important dynamical influence on the shelf circulation by modifying the stratification above the shelf bottom boundary layer (Gawarkiewicz and Chapman, 1992; Chapman, 2000; Pickart, 2000). The deep onshore flow might also serve as a pathway for deep oceanic zooplankton. Some of these organisms migrate diurnally over the full depth of the water column and only visit the surface to feed for a short period at night. Hence, the bottom flow that transports the high-salinity water shoreward might also result in a net shoreward flux of zooplankton in summer. The summertime inflow of saline water onto the inner shelf also allows the slope and basin interior to communicate directly with the nearshore, for (as discussed in the following text) this water is drawn from within the permanent halocline of the Gulf of Alaska. Finally, the deep summer inflow is a potentially important conduit for nutrients from offshore to onshore, although it is not the only means by which nutrient-rich offshore water supply the shelf. Other mechanisms include flow up canyons that intersect the shelfbreak (Klinck, 1996; Allen, 1996, 2000; Hickey, 1997), topographically-induced upwelling (Freeland and Denman, 1982), and shelfbreak eddies and flow meanders (Bower, 1991). Reed et al. (1987) and Schumacher et al. (1989) suggest that as the ACC flows southward through Shelikof Strait, it entrains deep salty and nutrient-rich water from the canyon depths into the surface layer. The deep waters within the canyon are then replenished by a deep inflow into the mouth of the canyon at the continental slope south of Kodiak.
Prince William Sound The circulation and water mass structure of Prince William Sound is intimately linked to the Gulf of Alaska shelf primarily through the ACC, although there are also conspicuous linkages to offshore waters, as discussed earlier. The basic circulation pattern within the sound appears to be as follows. Waters enter the sound through Hinchinbrook Entrance and then flow counterclockwise around the central basin. Some of this flow passes immediately into the western sound and exits through Montague Strait (and also perhaps along the western side of Hinchinbrook Entrance), and some of it continues into the northern sound (Schmidt, 1977; Royer et al., 1990; Niebauer et al., 1994; Gay and Vaughan, 2001). Northern sound waters flow southward through Knight Island Passage and reenter the shelf through the passes in the western sound, including Montague Strait. This circulation pattern varies seasonally in accordance with the seasonal cycle of winds and runoff and appears to be strongest in late fall and winter and weakest in summer. Indeed, the counterclockwise circulation pattern might even reverse occasionally, if not consistently through summer, with surface waters leaving through Hinchinbrook Entrance and entering through Montague Strait (Vaughan et al., 2001). Although there are no precise estimates of mass and material exchanges between
Ecosystem Structure 39
the sound and shelf, Niebauer et al. (1994) suggest that nearly half of the volume of the sound above 100 m depth is exchanged between May and September, whereas this same layer is renewed at least twice between October and April. There is also deep exchange of waters between the shelf and the sound. This occurs primarily through Hinchinbrook Entrance and also varies seasonally. A deep inflow of salty, nutrient-rich waters occurs in summer, coincident with the inshore migration of this slope water mass elsewhere across the shelf bottom. This flow is channeled northward across the shelf through Hinchinbrook Canyon and is probably an important contribution to the annual nutrient budget of the sound. In winter, fresher waters (formed by deep winter mixing) leave the sound in the lower layer (Niebauer et al., 1994; Vaughan et al., 2001). Exchanges of water between the main basins of the sound and the numerous fjords and bays within the sound occur in a variety of ways. As a general rule, the surface flow in a fjord is initiated by freshwater runoff at the fjord’s head. The fresh surface layer spreads along the fjord’s axis increasing its volume (by many times) and salinity through entrainment, induced by tidal currents and winds, of saltier subsurface waters. To compensate for the volume of deeper waters entrained into the outflow, subsurface waters flow inward toward the head of the fjord. This two-layer circulation, while present on average, is weak and can be easily masked in short-term observations by tidal and/or wind-driven flows. Although this generic circulation pattern holds for most fjords, there are substantial variations among fjords, owing to differences in geometry and bathymetry, winds, tides, and the salinity difference between the head and the mouth of the fjord. For example, in some seasons, the water at the mouth of the fjord might be fresher than that at the head (due to seasonal changes in river runoff into the fjord). In these situations, a reverse fjord circulation pattern can develop (Klinck et al., 1981). The renewal of deep fjord waters depends on fjord bathymetry and, in many cases, the depth of the fjord’s sill. Sills, which are usually located near the fjord’s mouth, restrict communication between the deep waters of the fjord’s inner basin and offshore waters. Deep-water renewal depends critically on sill geometry and ambient conditions. For some fjords, deep-water renewal may be infrequent and episodic, whereas for other fjords, complete or partial renewal occurs periodically either seasonally or, more frequently, through the interaction of stratified tidal flows with the sill. In such cases, tidal suction at the sill could draw deep offshore waters over the sill crest to resupply the inner basin of the fjord (Thompson and Golding, 1981; Thomson and Wolanksi, 1984). Strong tidal current– sill interactions can lead to a number of complex hydraulic effects that result in strong vertical mixing and exchange (Farmer and Smith, 1979; Freeland and Farmer, 1980).
Outer shelf and slope domain The third domain, waters over the shelf break and continental slope, includes the edge of the Alaska Current and the Alaskan Stream (west of about 150°W). These currents
40
Long-Term Ecological Change in the Northern Gulf of Alaska
are the northern portions of the North Pacific’s subarctic gyre (discussed in the following text) connecting the Gulf of Alaska shelf and the Pacific Ocean. The Alaska Current is about 250 km wide, and has relatively weak flows (5–15 cm s−1), with correspondingly small horizontal and vertical velocity shears. In contrast, the Alaskan Stream is a narrower, 75 km wide, much swifter (50–100 cm s−1), and with highvelocity shears in the upper 500 m (Reed and Schumacher, 1987). West of 150°, the Alaskan Stream flows along the continental slope south of Alaska Peninsula and Aleutian Islands and gradually weakens west of 180°W (Thomson, 1972). The transformation of the Alaska Current into the Alaskan Stream includes changes in the velocity and density gradients along and across the shelfbreak. These gradients, in conjunction with varying bottom topography, affect the exchange between the shelf and slope (Gawarkiewicz, 1991). Hence, the transition from the Alaska Current into the Alaskan Stream implies that shelfbreak exchange mechanisms are not uniform around the Gulf of Alaska shelf. The Alaskan Stream has a mean annual transport of 20 to 30 × 106 m3 s−1 (Reed and Schumacher, 1987; Musgrave et al., 1992; Ohtani et al., 1997; Reed and Stabeno, 1999), with the broad range reflecting the uncertainty of the relatively few direct measurements. Although seasonal transport variations are apparently small (Tabata, 1991), Thomson et al. (1990) find that the Alaska Current tends to be more concentrated along the continental margin in winter than in summer. Year-to-year transport variation can be as much as 30% (Royer, 1981b) of the mean. Surface salinities vary by only about 0.5 throughout the year, whereas the magnitude of the annual seasurface temperature cycle is comparable to that of the shelf (i.e., ~10°C). Nevertheless, except for the shallow summer thermocline, horizontal and vertical density gradients are controlled by the salinity distribution. Maximal stratification occurs between 100 and 300 m depth and is associated with the permanent halocline of the Gulf of Alaska. Halocline salinities range from 33–33.8, and temperatures range from 4–6°C (Tully and Barber, 1960; Dodimead et al., 1963; Reid, 1965; Favorite et al., 1976; Reed, 1984; Musgrave et al., 1992). The upper halocline, with salinities of 33–33.3 and temperatures of about 5°C, contributes to the deep waters that flood the shelf bottom each summer (Fig. 2.12). Although flow in the Alaskan Stream appears relatively steady (Royer; 1981b; Reed and Schumacher, 1987), large (150-km diameter), clockwise eddies, moving from the interior basin onto the slope and shelfbreak, can occasionally alter the circulation here (Musgrave et al., 1992; Crawford et al., 2000; Okkonen et al., 2003). These changes include flow reversals, deflection of the shelfbreak front, and vertical displacement of the pycnocline (Musgrave et al., 1992; Okkonen et al., 2003) during eddy passage. Upwelling occurs at the leading and trailing edges of the eddy, while the pycnocline is deflected downward as the eddy center passes. The eddies are long lived (2–3 years) and support a clockwise flow with speeds between 20–50 cm s−1 (Tabata, 1982; Musgrave et al., 1992; Okkonen, 1992; Crawford et al., 2000;
Ecosystem Structure 41
Okkonen et al., 2003). They form in winter in the eastern Gulf when wind stress is strong along the eastern boundary (Willmott and Mysak, 1980; Melsom et al., 1999; Meyers and Basu, 1999) and then propagate westward at about 2–10 cm s−1. Most of the eddies remain over the deep basin and far from the continental slope (Crawford et al., 2000); however, at least one usually forms in January or February in the northeast Gulf of Alaska. Once formed, these eddies tend to propagate westward along the continental slope for several months until they approach Kodiak Island, whereupon they translate southwestward (Okkonen et al., 2003). Eddies that impinge on the continental slope could substantially influence the shelf circulation and exchanges between the shelf and slope of salt, nutrients, plankton, fish eggs, and larvae in much the same manner as the Gulf Stream rings along the East Coast of North America (Houghton et al., 1986; Ramp et al., 1983; Joyce, et al., 1992). Indeed, in the eastern Gulf of Alaska, Whitney and Robert (2002) find that these eddies cause a net transport of nutrients from the shelf to the basin.
2.2.3. Tides The Gulf of Alaska has a mixed-type tidal regime with a dominant semi-diurnal M2 tide, the diurnal K1 tide usually being secondary in importance. Tidal amplitudes and velocities are strongly influenced by the complex bottom bathymetry and geometry of the shelf and coast. Consequently, there are large tidal differences across the Gulf coast. For example, Anchorage has the largest tidal range in the northern Gulf, the M2 and K1 tides being about 3.6 and 0.7 m, respectively. At Kodiak and Seward, the tides are about half as large. The cross-shelf flux of tidal energy onto the northwest Gulf shelf is very large and is dissipated by correspondingly high frictional forces (Foreman et al., 2000). The tidal dissipation rate in Kennedy Entrance is about half of the total dissipation of the M2 constituent in the Gulf, and nearly onethird of the energy of the K1 tide is dissipated within Cook Inlet. The tidal energy dissipation mixes the water vertically, bringing nutrient-rich water into the euphotic zone. As the tidal wave passes over the rough sea bottom, it can also generate diurnal shelf waves and residual or steady flows that can be locally important for transporting suspended and dissolved materials. Diurnal-period shelf waves are prominent features along the British Columbian shelf (Crawford, 1984; Crawford and Thomson, 1984; Flather, 1988; Foreman and Thomson, 1997; Cummins and Oey, 2000), where they displace the pycnocline and alter mixed layer depths. Foreman et al.’s (2000) model results indicate that diurnal shelf waves occur in the northwest Gulf and especially along the shelfbreak east of Kodiak Island. Water stratification facilitates vertical redistribution of tidal energy over the shelf by generating internal waves at the tidal period. (Internal waves move along density discontinuities and exist only in a stratified water column. They may not be noticeable
42
Long-Term Ecological Change in the Northern Gulf of Alaska
on the ocean surface, but the deeper wave can displace the pycnocline several meters.) Significant internal tides are likely generated at the shelfbreak in summer and fall when stratification is strong. The internal waves do not travel far (tens of kilometers), in contrast to the large scale (thousands of kilometers) of the generating tidal wave. Internal tide phases and amplitudes change with depth and seasonally with extent of stratification. While there have been no systematic studies of internal tides on the Gulf of Alaska shelf, Danielson (unpub. data) finds that the M2 tidal velocities in the ACC offshore of Seward in summer are approximately 20 cm s−1 at depths shallower than 30 m and about 5 cm s−1 below 100 m. In contrast, tidal velocities in winter are uniform over the water column at about 5 cm s−1. Internal tides can also displace the pycnocline with significant biological consequences, such as pumping nutrients into the euphotic zone, dispersing plankton and small fishes, and forming transitory and small-scale fronts that affect feeding behaviors (Mann and Lazier, 1996). Internal waves can also “break”, resulting in vertical mixing, a reduction in stratification, and the transport of nutrients into the euphotic zone. Regions of large bathymetric and tidal current gradients, such as found in Cook Inlet, often result in the formation of tidal fronts due to differential tidal mixing. These fronts typically parallel isobaths and delineate the boundary between waters that are well-mixed by tides from stratified regions. Tidal fronts can be very narrow (a few meters to hundreds of meters), and vary in strength and position over the fortnightly tidal cycle or seasonally in conjunction with changes in stratification. Rectified tidal currents and tidal fronts can also form around the perimeter of submarine banks. These regions are often highly productive and important feeding areas for fish. Although the classic example is Georges Bank off the New England coast (Horne et al., 1989), it seems likely that Portlock Bank northeast of Kodiak and Alsek Bank offshore of Yakutat behave similarly.
2.2.4. Gulf of Alaska Basin The circulation in the central Gulf of Alaska consists of the counterclockwise flow of the Alaska Gyre, which is part of the more extensive subarctic gyre of the North Pacific Ocean. The center of the Alaska Gyre is at about 53°N, and between 145 and 150°W. The gyre includes the Alaska Current and Stream and the eastward-flowing North Pacific Current along the southern boundary of the Gulf. Although some water from the Alaskan Stream apparently recirculates into the North Pacific Current, the strength and location of this recirculation is poorly understood and appears variable (Favorite et al., 1976; Emery et al., 1985). The gyral circulation pattern (illustrated schematically in Fig. 2.13) is, in fact, a three-dimensional circulation field that arises in response to the large-scale mean counterclockwise wind-stress distribution over the Gulf of Alaska. The divergence in the Ekman transports resulting from the
Ecosystem Structure 43
Figure 2.13: The major features of the Gulf of Alaska Gyre, including upwelling at mid-gyre, Ekman flow away from the center, and the currents at the shelf edge.
wind-stress distribution leads to upwelling at the center of the gyre and downwelling along the gyre boundaries. The long-term average upwelling rate is 10–30 m year−1 in the gyre’s center (Xie and Hsieh, 1995) and results in the permanent halocline being shallower here than around the edges of the gyre. It also provides a mechanism by which nutrient-rich deep waters are brought into the euphotic zone. Mean current speeds in the upper 150 m of the gyre (and far from the continental slope) are 2 to 10 cm s−1, although the variability is large (Thomson et al., 1990). Following Tully and Barber (1960) and Dodimead et al. (1963), the vertical temperature and salinity structure of the Alaska Gyre consists of: (1) a seasonally varying upper layer extending from the surface to about 100 m depth, (2) the permanent halocline between about 100 and 250 m, over which salinity increases from 33 to 33.8 and temperatures decrease from 6 to 4°C, and (3) a deep layer, extending from the bottom of the halocline to about 1000 m depth, over which salinity increases more slowly to about 34.4 and temperatures decrease from 4 to 3°C. Below 1000 m depth, salinity increases even more slowly to its maximum value of about 34.7 at the bottom. Seasonal changes in wind mixing and heat exchange with the atmosphere alter the properties of the upper layer. From October through March, cooling and strong wind-mixing establish a deep mixed layer with uniform temperature and salinity that extends to the top of the halocline. Winter mixed layer salinities range from 32.5 to 32.8, and temperatures range from 3.5 to 6°C, with the colder and fresher values
44
Long-Term Ecological Change in the Northern Gulf of Alaska
being in the northern Gulf. As wind speeds decrease and solar heating increases in spring, the upper layer freshens slightly and warms, ultimately forming a weak halocline and a strong thermocline at about 25 m. This seasonal pycnocline erodes, and upper layer properties revert to winter conditions as fall cooling and wind mixing ensues. The halocline is a permanent feature of the subarctic North Pacific Ocean and represents the deepest limit over which winter mixing penetrates. The halocline is a consequence of both the large-scale circulation and mixing process occurring over the North Pacific and the excess of precipitation and runoff over evaporation over the high-latitude North Pacific Ocean (Reid, 1965; Warren, 1983; Van Scoy et al., 1991). Although the halocline impedes vertical exchange between the upper layer and the nutrient-rich deep water below, it is relatively shallow compared to the main thermocline of the subtropical North Pacific. Consequently, sufficient exchange, driven by storm winds, occurs even in summer to maintain relatively high nutrient concentrations in the euphotic zone. The deep waters of the Gulf of Alaska also contain North Pacific intermediate water (formed in the northwestern Pacific Ocean) and, at greater depths, include contributions from the North Atlantic. Mean flows in the deep interior are feeble (1 cm s−1), with the flow dynamics governed by both the climatological wind stress distribution (Koblinsky et al., 1989) and the global-scale thermohaline circulation (Warren and Owens, 1985), subject to modifications by bottom topography. The global thermohaline circulation carries nutrient-rich waters into the North Pacific and forces a weak and deep upwelling throughout the region (Stommel and Arons, 1960a, 1960b; Reid, 1981; Broecker, 1991).
2.2.5. North Pacific Ocean The climate and climate variability at interannual and longer time scales of the Gulf of Alaska is linked to the North Pacific atmosphere–ocean system. Although this variability is addressed in Chapter 4, this section briefly outlines the main features of the North Pacific relevant to that subsequent discussion. The global atmosphere – ocean circulation system arises in response to the imbalance in radiative heating and cooling between the equator and the poles and the asymmetrical distribution of land and water over the globe. On annual average, this disparity is erased by heat transports from low to high latitudes provided by quasiorganized atmospheric and oceanic circulation systems. The term “quasi-organized” implies that these systems have both an identifiable mean structure (that includes seasonal variations) and some degree of randomness or unpredictability, which is an integral part of both fluids. The atmosphere and ocean are dynamically connected to one another through exchanges of heat, mass, and momentum. Hence, both fluids adjust to these exchanges, and perturbations in one part of the system can (and often do) propagate to another part. For example, regional sea surface temperatures and
Ecosystem Structure 45
temperature gradients are controlled by the amount of heat in the upper ocean and the transports and trajectories of the large-scale currents. Changes in upper-ocean heat occur by heating and cooling processes, primarily by exchanges with the atmosphere, but also through mixing and advection in the ocean, along the current’s path. The volume transport and pathway of currents are mainly determined by the large-scale wind stress pattern. On the other hand, changes in ocean temperatures and patterns of air–sea heat exchange ultimately affect atmospheric pressure patterns and the wind systems over the ocean. The spatial and temporal scales over which these adjustments occur depend on the scales, magnitude, and source of the perturbation, however. In some cases, these adjustments are immediate and obvious, while in other cases, the response might take years to decades and evolve along complex pathways. The principal atmospheric features of the North Pacific Ocean are associated with the sequence of low- and high-pressure systems distributed between the equator and the pole. These pressure cells are statistical composites based on temporal and spatial averages of the many individual pressure systems that build, translate, and decay over the North Pacific. High pressure prevails over the subtropical northeast Pacific (15–35°N; the East Pacific High) and polar regions (north of 65°N), while low pressure occurs along equatorial latitudes and between 40 and 60°N (the latter being the Aleutian Low, discussed earlier). The winds associated with these pressure systems are primarily zonal (east–west) and form alternating bands of surface easterlies (westward winds) over the equator (the Southeast Trade Winds), the subtropics (the Northeast Trade winds), and poles. Westerlies (eastward winds) prevail over the mid-latitudes with the jet stream centered at about 40°N and sandwiched between the East Pacific High and the Aleutian Low. The meridional (north–south) distribution of zonal wind stress is primarily responsible for the current structure in the upper 1000 m of the ocean. In the North Pacific, the wind stress distribution divides the basin into two massive counter-rotating gyres, with each gyre containing a distinct assemblage of currents (Fig. 2.14). The clockwise subtropical gyre encompasses the width of the Pacific Ocean and extends from 10–40°N. Its southern limb consists of the North Equatorial Current, which flows westward between 10 and 20°N across the breadth of Pacific. A portion of this current turns northward along the Asian continent, where it then narrows and intensifies to form the Kuroshio Current. The Kuroshio transports approximately 25 × 106 m3 s−1 of water from the tropics to mid-latitudes and is the major oceanic pathway by which warm, salty, tropical waters are transported northward in the North Pacific Ocean. The transport increases as the Kuroshio flows northward and, at about 40°N, it turns offshore from the Asian coast to form the Kuroshio–Oyashio Extension. Here, the transport increases to about 80 × 106 m3 s−1 from the addition of water flowing southward in the Oyashio Current. This mixture continues eastward between 40 and 50°N, and east of 170°E is called the North Pacific Current. West of North America, the North Pacific Current bifurcates, with part flowing northward into the Gulf of
46
Long-Term Ecological Change in the Northern Gulf of Alaska
Figure 2.14: The major surface circulation features of the North Pacific Ocean.
Alaska and the remainder turning southward in the California Current along the west coast of North America. The bifurcation latitude varies from about 45°N in winter to about 50°N in summer. Eventually, the California Current turns eastward to rejoin the North Equatorial Current, thereby completing the subtropical gyre. The equatorial current system consists of the westward flowing South Equatorial Current, which straddles the equator and the North Equatorial Countercurrent, which flows eastward between the North and South Equatorial currents. On annual average, the ocean gains heat south of about 30°N and loses heat to the atmosphere north of this latitude so, in aggregate, these currents affect a net northward transport of heat from the tropics. The oceanic sources of the atmospheric moisture that eventually enters the Gulf of Alaska vary seasonally (Emile-Geay et al., 2002). From spring through fall, most of this moisture is absorbed from the western tropical Pacific and the Indonesian seas and thence transported northward by the low-level winds of the Asian Monsoon. In winter, cold, dry winds spilling off the Asian mainland suck moisture into the atmosphere along the path of the Kuroshio and the Kuroshio–Oyashio Extension. Much of this latent heat exchange is controlled by the properties of the winter air masses flowing across Asia, suggesting that the strength and position of polar high-pressure systems possibly influence this process. The Subarctic Gyre consists of two small, counterclockwise-flowing gyres; the Alaskan Gyre (discussed earlier) and the Western Subarctic Gyre, which includes the Bering Sea basin. The gyres are linked to one another through the Alaskan Stream. Some of the Alaskan Stream enters the Bering Sea through the deeper passes of the
Ecosystem Structure 47
Aleutian Islands, then flows counterclockwise around the Bering Sea before leaving via a southward flow along the Kamchatka Peninsula. This outflow commingles with waters from the Sea of Okhotsk and the remnants of the Alaskan Stream to form the southward-flowing Oyashio Current. The Oyashio closes the Western Subarctic Gyre, for it feeds the Kuroshio–Oyashio Extension and eventually rejoins the North Pacific Current. The junction between the cool, fresh subarctic waters and the warm, salty subtropical waters that parallels the path of the Kuroshio–Oyashio Extension and the North Pacific Current forms the subarctic front. The warm waters, transported northward by the Kuroshio and the sea surface temperature contrast across the subarctic front, particularly within the Kuroshio–Oyashio Extension, where the sea surface temperature gradient is greatest, play a prominent role in the formation of the low-pressure systems that eventually enter the Gulf of Alaska. Most of these storms are generated in this region throughout the year, although cyclogenesis is more frequent and vigorous in fall and winter, when cold, dry air flows off Asia and encounters warmer ocean waters. Gradients in sea surface temperature result in similar gradients in sensible and latent heat fluxes, which can enhance cyclogenesis. Storms evolve as sea level pressures fall within the heated air masses and counterclockwise winds develop about the zone of falling pressure. Once formed, these storms intensify en route to the Gulf of Alaska as they continue to extract heat and moisture from the ocean (Roden, 1970).
2.3. The Marine Production Cycle Theodore Cooney 2.3.1. Introduction Photosynthesis establishes the base of the food web in most marine communities (Fig. 2.15). In the pelagic realm, phytoplankton and some bacteria are the dominant photosynthetic producers. Given sufficient light energy and nutrients (dissolved inorganic forms of nitrogen, phosphorus, silicon, and iron), these single-celled primary producers reduce carbon in the presence of sunlight and synthesize organic matter to form living plant populations. These populations are in turn grazed by a host of small animals – the zooplankton – ranging from tiny single-celled ciliate protozoans and flagellates to larger animals such as copepods and euphausiids (krill) that in turn become food for fishes, birds, and mammals. At the same time, detritus (primarily fecal material and the remains of dead plants and animals) is oxidized by bacteria as it falls toward the seabed, liberating nitrogen, phosphorus, silicon, iron, and other substances back into the water. Cycling of inorganic nutrients and carbon to living organic matter (a synthesis process) and back to dissolved inorganic forms (a remineralization process) defines the marine production cycle at lower trophic levels.
48
Long-Term Ecological Change in the Northern Gulf of Alaska
Figure 2.15: Organic matter cycling in the pelagic ocean. See text for description.
Unlike terrestrial ecosystems that are rich in stored carbohydrate and are relatively slow growing, the pelagic marine system is dominated by a protein ecology with little long-term energy storage but very rapid growth (Parsons, 1976). A hallmark of this system is that there is little carryover of plant matter from year to year.
2.3.2. The Annual Cycle of Production The rate at which organic matter is produced in the ocean depends on the cyclic availability of both nutrients and light (see Fig. 2.20). In mid-to-high latitudes, there is only enough daylight for significant phytoplankton growth during a part of the year – spring and summer into early fall. Within this period, plant growth can continue until nutrients are exhausted. During the dark days of winter, production falls to very low levels. At the same time, cooling of the surface waters and increasing winter storm activity deepens the wind-mixed layer, causing dissolved inorganic nutrients from deep sources to become entrained and brought to the surface. In this
Ecosystem Structure 49
BOX 2.1: PLANKTON by Theodore Cooney By definition, plankton is the community of tiny marine plants and animals whose distribution is influenced primarily by ocean currents. Phytoplankton (the photosynthetic plants and bacteria) (Fig. 2.16), zooplankton (Figs. 2.17 and 2.18) (the small drifting animals), and meroplankton (eggs and larval stages of larger organisms) in this community are generally most abundant near the surface of the ocean – the upper 200 m. Although their distributions are at the mercy of the currents, some zooplankters undergo extensive vertical migrations (up to several hundred meters) in response to seasonal cycles in their food abundance and/or reproductive strategies. Also, many zooplankters migrate daily – swimming up into the surface waters at night, where they feed or mate, and returning to depth during the day to digest their food and escape predators. The majority of marine plants are unicellular and can double their populations approximately every day under ideal conditions. In the seasonal ocean, the living mass of plankton varies over an order of magnitude or more each year, usually reaching a peak in the summer before falling to low levels in mid-to-late winter. The dramatic increase in biomass that begins early each spring is called the plankton bloom. Marine scientists use fine-mesh nets and water bottles to sample plankton (Fig. 2.19).
Figure 2.16: Phytoplankton: diatoms (photograph courtesy of NOAA).
50
Long-Term Ecological Change in the Northern Gulf of Alaska
Figure 2.17: The large pelagic copepod Neocalanus cristatus (photograph courtesy of R. Hopcroft, University of Alaska).
Figure 2.18: The pteropod Limacina sp., a pelagic mollusk (photograph courtesy of R. Hopcroft, University of Alaska).
Ecosystem Structure 51
Figure 2.19: Sampling with a bongo net (photograph courtesy of EVOS Trustee Council).
way, upper-layer nutrient concentrations are replenished each year prior to the next spring bloom (see Section 2.2). Sverdrup (1953) developed an elegant theory that explains the timing of onset of the plankton bloom in the spring. In the presence of sufficient nutrients, phytoplankton growth rates are primarily determined by light intensity. Since light is attenuated logarithmically with depth, a sinking cell will quickly reach a depth where available light permits just enough photosynthesis over 24 h to balance respiration – the compensation light intensity. This depth generally defines the seasonally changing base of the photic zone. Cells sinking below this depth receive less than the compensation light intensity and cease to grow. The critical depth is the depth above which the overall depth-integrated photosynthesis just equals the depth-integrated respiration over a 24-h period for a population of plant cells distributed uniformly by wind mixing. From this depth to the surface, the mixed-layer phytoplankton population experiences an average 24-h light intensity that just balances its respiration. When the mixed layer extends below the critical depth, cells in the population receive insufficient light to balance their respiration, and production ceases. Prediction of the bloom timing follows from these limiting conditions. In the winter, low surface light and deep vertical mixing establishes a mixed layer that is much deeper than the critical depth (Fig. 2.20). As a result, the rate of primary
52
Long-Term Ecological Change in the Northern Gulf of Alaska
Figure 2.20: The annual cycle of primary productivity including water, light, wind, and nutrient conditions that initiate the spring bloom, support periodic spring and summer production, and lead to the end of the production, season in the coastal Gulf of Alaska.
productivity – photosynthesis – is extremely low. As days lengthen in the spring, increasing sunlight extends the critical depth deeper into the water. At the same time, decreasing winds and fresher, warmer surface water cause the mixed layer to contract towards the surface. When the mixed layer becomes shallower than the critical depth, all plant cells in the mixed layer receive sufficient light energy to exceed their respiratory demands. The result is an immediate expansion of phytoplankton biomass, now growing rapidly on high concentrations of dissolved nutrients – the spring bloom. Because the timing of this explosive growth depends on weather-forced vertical mixing interacting with seasonally available light, the conditions for this event rarely occur on the same date each year, and the start of the bloom typically varies by a week or more between years. The main variables that appear to be responsible for the variation in timing are the degree of vertical stability determined mainly by winds and freshwater input, the amount of cloud cover, and the angle of incident radiation.
Ecosystem Structure 53
Once underway, the annual spring burst of primary production continues as long as adequate nutrients are available. This state is eventually terminated by increasing stratification as a warmer, less saline surface layer develops over a body of deeper, cooler, more dense water, dramatically reducing vertical exchange and nutrient renewal to the upper layers. Some production continues after this time, due to (1) episodic wind-mixing events that reinject nutrients into the surface layer and boost growth for short periods (see wind events in Fig. 2.20), (2) bacterial oxidation of detritus which recycles some inorganic nitrogen, mostly in the form of ammonia, and (3) nitrogen fixation by the cyanobacteria. Overall, primary production declines to low levels following stratification, and the resulting seasonal hiatus in plant growth is termed the nutrient-limited portion of the production cycle. In some locations where nutrients are continuously available in the surface waters (for example, through persistent upwelling from deep sources as occurs in the entrance to Cook Inlet, or in the open ocean), phytoplankton production can continue in proportion to light levels throughout the summer. In summary, over the course of the year, the general seasonal pattern in photosynthesis includes a period of severe light limitation during the winter, short episodes (a few weeks) of dramatic growth in the early spring (and possibly the fall), and nutrient limitation during the late spring, summer, and early fall. The examples of these patterns from the Gulf of Alaska and their regional exceptions will be discussed in the following text. The explosive spring bloom in plant plankton is followed quickly by a similar expansion in numbers of the various zooplankers that graze on it (Fig. 2.21). In this way their seasonal populations parallel that of the phytoplankton, but with varying time lags, depending on the species. For example, the small protozoans can keep pace reproductively with the plants, so their populations fluctuate in close synchrony. In contrast, the larger zooplankters take longer to reproduce – weeks or months – and most must first feed before they can produce offspring. The result is a time lag between the peak of the phytoplankton production and seasonal peaks in the large zooplankton community, which generally come later in the growing season. As the biomass of plant and animal plankton expands, the dominant species follow a predictable progression. At the beginning of the season, diatoms dominate in the cold, high-nutrient waters. These plants, many of them producing large colonies, have an external covering of silica and are often quite large – some cells greater than 50 microns in size. When nutrient concentrations decline in the late spring and early summer, the phytoplankton community shifts to smaller species with increased surface-area-to-volume ratios. Because nutrients are absorbed through the cell walls, increasing surface area relative to cell volume maximizes nutrient uptake for the plants. The result is that nutrient-poor summer waters typically host very small phytoplankton – some tiny diatoms (5 microns and smaller), and a larger proportion
54
Long-Term Ecological Change in the Northern Gulf of Alaska
Figure 2.21: The progression of various plankton populations, their trophic linkages, and the availability of nutrients during the spring and summer production season.
of microflagellates and medium-sized dinoflagellates. In addition to being small, the flagellated forms swim and can thus exploit more optimum levels of light and nutrients. The seasonal progression in size of the primary producers is mirrored by similar shifts in the zooplankton. The early-spring zooplankton community is dominated by large calanoid copepods and euphausiids – primarily filter feeders – which target the diatom populations. Some of these consumers (and sometimes their progeny) overwinter in the deep water and migrate back to the upper layers to feed in the spring. When the phytoplankton shifts toward smaller forms in response to nutrient exhaustion, the larger copepods and euphausiids are replaced, in part, by species that are more capable of feeding on particles 5 microns in size and smaller. During this time, the Pteropoda (mucus-feeding pelagic molluscs) and the Larvacea (very fine, filter-feeding pelagic tunicates) expand their populations. Also, during the summer and fall, carnivorous jelly plankton (Ctenophora and Cnidaria) become well established as the community shifts away from domination by herbivores toward a greater diversity in feeding types – more omnivores and carnivores.
Ecosystem Structure 55
The fundamental processes described in the preceding text form the basis of the marine production cycles in the northern Gulf of Alaska. However, there are some important exceptions and regional differences. In the following sections, we will compare plankton production cycles in three distinct areas: (1) protected inner waters, (2) open ocean, and (3) shelf and coastal waters. Figure 2.22 provides a comparison of important features of the production system in these domains.
Protected Inner Waters The production cycle of protected inner waters (sounds, fjords, and inlets) follows the classical picture described earlier. There is a well-developed spring burst of plant production and biomass accumulation (see Fig. 2.23) that, at first, features populations of large diatoms and then gradually gives way to smaller photosynthetic forms as the water column stratifies and nutrient supplies dwindle in the photic zone. This general pattern can be illustrated by measured changes in water fluorescence (a proxy measurement for chlorophyll a, the photosynthetic pigment for most algae) for Prince William Sound (Eslinger et al., 2001). Continuous measurements at a depth of 10 m over the year captured the diatom bloom that typically peaks in mid-April, followed by a rapid decline to summertime lows beginning in May and June. A second fall “event” is documented for October and November (Fig. 2.20). Similar patterns in
Figure 2.22: The contrast between inner- and outer-shelf plankton community structure, production, and how inshore–offshore differences in trophic efficiencies arise.
56
Long-Term Ecological Change in the Northern Gulf of Alaska
Figure 2.23: A false-color satellite image of the northern Gulf of Alaska taken on April 5, 2002, showing the high concentrations of chlorophyll a in the surface waters of Prince William Sound and some other isolated inshore and offshore locations. Red indicates the highest concentrations of chlorophyll a, and blue and green colors indicate the lowest (image courtesy of D. Musgrave and R. Potter, University of Alaska; data from NASA).
plant pigments have been reported for at least the late winter through summer periods in coastal Alaska (Goering et al., 1973; Iverson et al., 1974; Burrell, 1986; Larrance et al., 1977). Annual primary productivity estimates for protected waters range from 300 g C m−2 for lower Cook Inlet to 185 g C m−2 in northwestern Prince William Sound and 145 g C m−2 in Boca de Quadra in southeastern Alaska. These values are similar to those of protected marine waters in Norway, Sweden, Greenland, and Canada (Cooney and Coyle, 1988). Studies of nutrient availability in the upper layers of protected coastal waters document nitrogen enrichment during the light-limited period of the production cycle in winter. Nitrate and silicate concentrations decline rapidly after the initiation of
Ecosystem Structure 57
the spring boom (Goering et al., 1973; Heggie et al., 1977; Burrell, 1984; Ward, 1997), low levels of primary productivity being sustained through the summer by nutrient recycling in the photic zone (ammonia production), by nitrogen fixation, and by localized upwelling and vertical turbulence (Sambrotto and Lorenzen, 1987). During the period of relaxed coastal downwelling in each summer (more so in some years than in others), the intrusion of subsurface slope waters onto the shelf provides a mechanism to enrich the deeper coastal regions with plant nutrients. These nutrients find their way into the productive surface layers through deep convective and wind mixing in the following winter and early spring, and also by other processes such as tidal pumping. The phytoplankton of sounds and fjords is dominated by large diatoms during the spring bloom (Ward, 1997; Horner et al., 1973), while the summer plant assemblages are primarily smaller forms. Chaetoceros, Thalassiosira, and Skeletonema are among the most common large diatom genera, although their dominance ranking typically changes from year to year. The zooplankton in protected inner waters is a combination of oceanic and neritic forms. In the early spring, zooplankton community biomass in Prince William Sound is dominated by the large calanoid copepods Neocalanus spp., Calanus marshallae, and Metridia okhotensis, with barnacle nauplii and the early copepodites (C1 and C2) of Neocalanus being important contributors as well (Cooney et al., 2001). By early summer, the community biomass switches to dominance by the small calanoid, Pseudocalanus spp., the pterpod, Limacina helicina, and the larvacean, Oikopleura sp. Later in the summer, in the freshened and warm upper layers, Pseudocalanus, ctenophores, and arrow worms, Sagitta elegans, are dominant in most samples. Wetweight biomass in the upper 50 m averages about 10 mg m−3 in the winter and early spring but peaks near 600 mg m−3 in midsummer – late June and July. Copepods dominate the numbers and biomass of most net-caught samples during all seasons.
The Open Ocean The open ocean ecosystem differs remarkably from the protected inner waters. First, there is little or no observable spring bloom. Phytoplankton numbers generally remain low throughout the year; the only hint of a plant bloom may occur occasionally in the fall, not the spring; and the seasonal accumulation of plankton occurs as zooplankton, not phytoplankton. Second, inorganic nitrate, a principal limiting nutrient in shelf and inner waters, is only rarely exhausted in the surface waters of the open ocean. This apparent puzzle, i.e., abundant nitrate yet no phytoplankton bloom, has prompted a number of ideas about how the open ocean functions. Much of the story has been put together over the last 25 years for the North Pacific Ocean (Miller, 1993). Trophic structure and nutrient dynamics in the open ocean of the Gulf of Alaska are very similar to oceanic ecosystems elsewhere in the world that are termed
58
Long-Term Ecological Change in the Northern Gulf of Alaska
high-nutrient, low-chlorophyll (HNLC) areas. Compared to the classical picture observed in protected inner waters, there are four primary differences in these areas: (1) the main plants at the base of the food web are very small diatoms and flagellates (2–10 microns); (2) the primary grazers are small protozoans; (3) the dominant source of nitrogen for plant growth appears to be ammonia, not nitrate; and (4) in all HNLC areas investigated, plant growth has been shown to be iron limited. The lack of an observable spring phytoplankton bloom is probably due to the presence of very efficient grazers that eat the phytoplankton as quickly as the latter can grow and divide, even during the optimal conditions in the spring. So, the rapid spring growth of microflagellates responding to increasing day length is kept pace with by grazing ciliates and other protozoans. Because deep vertical winter mixing in the oceanic subarctic Pacific is restricted by a permanent halocline (steep gradient in salinity) at 100 m, plants and micrograzers are rarely diluted to levels that would cause a trophic uncoupling of the plants and tiny herbivores. Since the protozoans are capable of equaling the reproductive rates of the plants, there is rarely, if ever, a time during the year when the primary producers are free from grazing losses and capable of expanding their populations. The large calanoid copepods are important grazers in the open-ocean plankton communities, as in inner waters, but occupy a different trophic position. These copepods exploit the epipelagic zone (upper 200 m) for feeding and growth, but use the deeper mesopelagic zone (500–1000 m) for diapause and reproduction. In contrast to inshore waters, where they can flourish feeding on large phytoplankton, the large copepods cannot get enough energy from eating small primary producers (Dagg, 1993). Instead, they graze on the protozoans, which are the primary grazers. The year-round presence of nitrate in the photic zone of the open ocean may be partially explained by differences in nitrogen cycling. On the continental shelf, nitrate is the dominant form of nitrogen utilized by plants, but ammonia is much more important in the open ocean (Wheeler and Kokkinakis, 1990). Ammonia is the first oxidative product of the remineralization process, and because it is relatively high in energy, it is “preferred” by phytoplankton over nitrite and nitrate. Ammonia is produced by bacterial oxidation of organic matter, and also excreted by zooplankton (and other consumers). The highly developed protozoan community of the open ocean (along with other consumers) probably supplies large quantities of ammonia to the surface waters. As long as ammonia is present in the photic zone, nitrate uptake is negligible. Because nitrate, which ultimately comes from deep wintertime mixing and weak upwelling in the central gyre is not appreciably utilized, the remainder is presumably available for transport onto the shelf for utilization there by the diatomdominated bloom. Another potentially contributing factor to the nature of the offshore pelagic community is iron limitation. Iron is probably supplied to the surface waters of openocean regions around the world by dust blown in from surrounding land, sometimes
Ecosystem Structure 59
over great distances. This essential plant micronutrient greatly affects oceanic primary productivity in the subarctic Pacific (Martin and Fitzwater, 1988). There are several open-ocean regions that exhibit high nutrients and low chlorophyll (HNLC) conditions besides the subarctic Pacific, and all have demonstrated an increase in primary productivity when enriched with iron. Because of these findings, some have suggested that the subarctic Pacific might owe its characteristic small-sized pelagic flora and seasonally residual nitrate to oceanic iron deficiency (Miller, 1993). Thus, while some questions remain, the most recent studies suggest that, in the open ocean, the sustained nitrate levels and the dominance of small-sized plant cells is most likely due to a combination of iron limitation and ammonia production (Strom et al., 2000). In summary, the seasonal picture that emerges from plankton studies in the open ocean of the subarctic Pacific begins with light limitation in the winter and early spring. Winter mixing and central gyre upwelling enrich the surface layers with dissolved inorganic nutrients, while photosynthesis is at a minimum. With the advent of weak stratification in the spring (associated with a warming water column), photosynthesis begins in the plant communities. However, because the micrograzers can apparently crop the phytoplankton very efficiently, little or no accumulation of plant biomass occurs, and it is the grazing community that increases in stock size. This increase is particularly evident for the large copepods (easily sampled with nets) – Neocalanus spp. and other large forms. As the season progresses, inorganic nutrients are drawn down in the photic zone, but the rapid recycling of ammonia production by bacteria and other consumers, along with what appears to be a dissolved iron deficiency, seemingly prevents the nitrate levels from dropping to zero. Annual primary productivity for the oceanic subarctic Pacific is thought to exceed 100 g C m−2 (Welschmeyer et al., 1993). With the renewal of fall and winter cooling, wind-mixing, and decreasing light levels, the production cycle again enters its light-limited winter phase. Late fall, winter, and early spring domination of the northeastern Gulf of Alaska by the Aleutian low-pressure system produces a shoreward Ekman flow and coastal convergence for more than half the year. Although the shoreward transport is relatively weak (approximately 5 km day−1) and pulsed by storms, it is apparently sufficient, when operating over several months, to introduce sizable quantities of biomass of oceanic origin to shelf and coastal ecosystems (Cooney, 1988; Kline, 1999). How much of this material is imported is unknown, although some believe that it significantly enriches coastal and shelf food webs in some years (Kline, 1997).
Open Coastal and Shelf Waters Production in the open coastal and shelf waters is not as well characterized as the protected inner waters and the open ocean. There are only a few studies of rates of primary production, but this environment is now the focus of studies that should become available soon. Sambrotto and Lorenzen (1987) provide a thorough historical
60
Long-Term Ecological Change in the Northern Gulf of Alaska
review of phytoplankton and seasonal primary production for the Gulf of Alaska, including open coastal and shelf waters. The largely speculative estimates of KoblentzMishke et al. (1970) placed the production for shelf areas at 100–150 g C m−2 year−1. However, studies by Larrance (1971) of Adak Bay and shelf waters, by Larrance and Chester (1979) of outer Cook Inlet, and by Larrance et al. (1977) of the Kenai shelf provided annual estimates of primary production that ranged much higher; 300–330 g C m−2. In the absence of developed literature, it seems reasonable to suggest that the waters of the open coast and shelf of the northern Gulf of Alaska contain plant and animal plankton that is a mixture of organisms advected shoreward from the deep ocean to the coastal convergence, and those carried seaward with the seasonal offshore spread of freshened waters in the summer and early fall. Some work carried out in this region under Outer Continental Shelf Environmental Program sponsorship in the 1970s demonstrated the seasonal intrusion of oceanic zooplankton, presumably carried shoreward under active downwelling in the fall, winter, and spring months (Cooney, 1988). In addition, the landward margin of the shelf hosts the Alaska Coastal Current, which serves to distribute and mix plankton communities over 2000 km of shoreline, from northern British Columbia to the tip of the Alaska Peninsula. New studies of the shelf in the northern Gulf of Alaska sponsored by the North Pacific GLOBEC program focus on the spatially distributed attributes of the annual production cycle – nutrient sources and sinks, the role of microplankton, organic matter transfer processes, and the distribution, timing, and composition of plankton blooms in relation to factors influencing the survival of juvenile salmon entering this region from coastal nurseries. These studies are beginning to describe a much more complex shelf system than previously reported (Weingartner et al., 2002). Preliminary studies of the shelf break blooms seem to indicate that the dominant phytoplankton are small diatoms and that protozoans are important grazers (GLOBEC, unpublished). As more of these studies are reported, the structure and dynamics of the shelf ecosystem will emerge.
2.4. The Transfer of Matter and Energy Through the Food Web Theodore Cooney The marine production cycle in the northern Gulf of Alaska is strongly seasonal at the base of the food web, with periods of low productivity in the winter, an explosive spring bloom, but followed by often discontinuous growth in the remainder of the spring, summer, and fall months. The interplay between available light and inorganic plant nutrients – modified as we have seen by ocean physics – is primarily responsible
Ecosystem Structure 61
for the seasonal variability. The larger animals supported by seasonally shifting production at lower trophic levels must cope with a forage environment that cycles (only somewhat predictably) between feast and famine each year. The burst of production in the spring spreads quickly through the food web. For example, yellowfin sole (Pleuronectes asper), walleye pollock (Theragra chalcogramma), and Pacific cod (Gadus macrocephalus) replenish energy reserves (lost over the winter and from spawning) each spring in a short period of intensive feeding (a few weeks) (Paul and Smith, 1993; Paul 1997; Smith et al., 1988, 1990) (Fig. 2.24). Similarly, adult walleye pollock and Pacific herring feed on dense layers of near-surface macrozooplankters (primarily, large calanoid copepods, krill, and pteropods) to rebuild their post-spawning energy reserves during late April, May, and early June (Willette et al., 2001). The flow of organic matter in some parts of marine food webs is an intermittent process tied to the interlocking life histories, behaviors, and trophic requirements of predators and prey and modified by the seasonal dynamics of the production cycle. These habitat dependencies tie each species to the biological and physical vagaries of the marine environment that support them over time. As we will see, the food supply for fishes, birds, and mammals is often scattered and ephemeral, with most energy coming from a few key species.
1150
1050 (Calories g-1 wet wt.)
WHOLE BODY ENERGY
1100 J
1000
B
B
JH
J H
B J H
H B
J B H
950 900
H B J
850 800
J
B
males
J
females
H
juveniles
H B J
H B
750 Jan Feb Mar April May Jun
Jul
Aug Sept Oct Nov
MONTH
Figure 2.24: Seasonal changes in whole-body energy content of yellow sole Pleuronectes asper in the Gulf of Alaska (after Paul, 1997).
62
Long-Term Ecological Change in the Northern Gulf of Alaska
2.4.1. Food-Web Structure – The Principal Forage Stocks Pelagic and benthic food webs in the Gulf of Alaska are a complex array of hundreds of species linked in a poorly understood trophic process that moves matter and energy from primary producers to all consumer levels. Approximately 300 species of fishes, 147 species of birds, and 26 species of marine mammals feed in the Gulf of Alaska for at least some portion of each year (OCSEAP, 1986; DeGange and Sanger, 1986; Calkins, 1986). While the forage utilized by apex consumers is diverse overall, relatively small collections of invertebrates and small schooling fishes are believed to provide the bulk of the food (Springer and Speckman, 1997). These trophically important or key forage stocks include a few macrozooplankters (two genera of large calanoid copepods, a pteropod, and euphausiids), juvenile herring, salmon and pollock, capelin, sand lance, eulachon, and two or three mesopelagic fishes – fewer than 20 in all (Fig. 2.25). Most of these forage species exhibit schooling, layering, and/or swarming behaviors during some or all of their life histories, providing dense feeding opportunities for consumers who can exploit these patches. In this manner,
BOX 2.2: FOOD CHAINS AND WEBS by Theodore Cooney “Food chains” and “food webs” are terms used to convey word pictures describing some aspects of consumptive processes. Energy fixed by photosynthetic plants and bacteria is transferred to consumers through “trophic” or feeding relationships. Descriptions of forage dependencies, for example, by seabirds, provides a way to conceptually understand how some parts of a consumptive system operate. Early studies of energy flow through aquatic systems (marine and freshwater) made some simplifying assumptions about the structure of communities, electing to assign “trophic levels” to various consumers. This structure began with photosynthetic “primary producers” at level 1, herbivores (primarily zooplankton) at level 2, and first and secondary consumers at successive higher levels. As the sophistication of trophic studies grew, it became apparent that the consumptive process was quite complicated and actually more web-like, with some consumers, such as fishes, occupying several different “trophic levels” during different times in their life histories. Recent advances in numerical modeling studies are using this information to begin constructing parts of the trophic process for predictive and experimental purposes.
Ecosystem Structure 63
Figure 2.25: Waist species that funnel much of the matter and energy into the predatory fish, seabirds, and mammals. Upper row, left to right: calanoid copepods, euphausid, copepod, and pteropod; middle left: sand lance; lower left: capelin; and lower right: juvenile pollock [photographs courtesy of R. Hopcroft, University of Alaska (copepods, euphausid, and pteropod), NOAA (sand lance, capelin), and the Japan Agency for Marine-Earth Science and Technology, JAMSTEC (pollock)].
the food web in the northern Gulf of Alaska is dramatically constricted at intermediate levels for many apex consumers by the dominance of these few but locally and seasonally abundant forage stocks. An example of this “waist” can be seen in seabird diets (percentage weight in stomachs) in Prince William Sound that demonstrated (in descending order) the importance of some of these forage stocks for 14 different seabirds: (1) juvenile herring 22.4%; (2) sand lance 22.0%; (3) capelin 14%; (4) macrozooplankton 6.3%; (5) juvenile pollock 6.0%; and (6) juvenile salmon 1.0%. These six categories accounted for 72% of all foods (12 categories) by weight in the stomachs of the birds (Fig. 2.26) (Fisheries Center, 1998). Aggregations of macrozooplankton provide excellent forage. The surface-swarming and layer-forming large oceanic copepods – Neocalanus spp. and Calanus marshallae – exhibit peak biomasses in the spring and early summer (Cooney et al., 2001). This occurs just before the maturing older stages (adults and C5 copepodites) leave the upper layers to enter a diapause in the deep waters (>400 m). From mid-April through early June in protected coastal areas such as Prince William Sound, copepodite stages C4 and C5 occur in dense, near-surface layers that stretch unbroken for tens of kilometers. Abundances exceeding 500 individuals m−3 are common in
64
Long-Term Ecological Change in the Northern Gulf of Alaska
JUVENILE SALMON JUVENILE POLLOCK MACROZOOPLANKTON
JUVENILE HERRING
CAPELIN
SAND LANCE
Figure 2.26: The composition of major forage groups in seabird diets in the Gulf of Alaska (after Fisheries Center, 1998).
these layers (Kirsch et al., 2000). (Fig. 2.27) The C5 copepodites range from 3.0 to 5.0 mm in length, a size that makes them particularly vulnerable to planktivorous fishes (pollock and herring), some seabirds, and even humpback whales. The pteropod – Limacina pacifica – because of its size (5–10 mm), abundance, and swarming behavior also provides forage opportunities for near-surface feeders in late May and June (Cooney et al., 2001). In contrast to the seasonally dominant large copepods and other macrozooplankton, euphausiids – Euphausia pacifica and Thysanoessa spp. – are present near the ocean’s surface throughout the year at night (Fig. 2.28), and in extensive layers and swarms below the surface in shelf and coastal waters at about 100 to 120 m during the day. Thysanoessa spp. spawn at the surface during the spring diatom bloom in coastal and shelf waters. Adults older than about 18 months die after spawning, and their remains are sometimes washed up on area beaches. Spawning swarms provide excellent forage for whales, seals, and some birds. It is the small schooling fishes – capelin, Pacific sand lance, eulachon, Pacific herring, juvenile salmon, juvenile pollock and cod, and a few mesopelagic fishes – that are most often singled out as the dominant forage stocks in the Gulf of Alaska for larger fishes, birds, and mammals (Springer and Speckman, 1997). Their spawning aggregations often provide forage that is extremely important to their predators. Sand lance spawn in September and October near the shore on sandy or fine-gravel beaches (Robards et al., 1999a). Capelin, herring, and eulachon are all spring spawners; the former two laying eggs in the intertidal zone, while the latter is anadromous, spawing in local rivers.
Ecosystem Structure 65
Figure 2.27: A multichannel acoustic backscatter record (200 kHz) taken on the shelf off Seward, Alaska, in May 1998 at night, showing the dense surface layers of plankton. The red indicates the greatest backscatter and probably the densest aggregations of animals and the distance from shore (top) (image courtesy of K. Coyle, University of Alaska). Two calanoid copepods that contribute to these plankton spring layers, Calanus marshallae and Neocalanus cristatus (bottom left, right, respectively) (photographs courtesy of R. Hopcroft, University of Alaska).
The spawning of Pacific herring (usually in March/April) is a particularly momentous ecological event – tens of miles of shoreline turn white with milt, and portions of the intertidal zone are covered with half a meter or more of eggs (see Fig. 4.17). Birds, fishes, marine mammals, and large invertebrates gather to feed on the spawning adults and their eggs. Bishop and Green (2001), working in Prince William Sound in 1994 found that, five species of birds consumed 31% of the eggs deposited in the region that year. During a spawning event, stocks of adult herring are also targeted by
66
Long-Term Ecological Change in the Northern Gulf of Alaska
Figure 2.28: Three common euphausids in the gulf: Thysanoessa intermis (top), Thysanoessa longipes (middle), and Euphausia pacifica (bottom) (photographs courtesy of R. Hopcroft, University of Alaska).
Ecosystem Structure 67
humpback whales and Steller sea lions. Later in the summer, age-0 juvenile herring begin to appear in protected bays, inlets, and fjords termed nursery areas (Norcross et al., 2001). From this time on, the developing and mature herring represent one of the most significant forage bases in the coastal northern Gulf of Alaska (Brown, 2003). In Fig. 2.29, schools of herring and sand lance can be seen in nearshore areas of Prince William Sound.
Figure 2.29: Aerial photographs of nearshore schools of juvenile herring (top) and sand lance (bottom). Arrow indicates long, gray, arc-shaped school of sand lance in front of vessel (photographs courtesy of E. Brown, University of Alaska).
68
Long-Term Ecological Change in the Northern Gulf of Alaska
Little is known of the life history and seasonal distribution of capelin, an ephemeral but very important forage fish. Fished commercially in the northwestern Atlantic Ocean, this species is not exploited in the Gulf of Alaska. Capelin is a cold-water, pelagic species, inhabiting arctic and subarctic zones in both the Atlantic and Pacific. It can form very large schools in inshore waters. During the demise of the northwest Atlantic cod stocks during the 1990s, capelin increased, zooplankton decreased, and indices of phytoplankton rose (Carscadden et al., 2001). It was hypothesized that this rearrangement was the result of the significant downturn in cod population (because of over fishing and environmental causes), releasing the capelin from their major predator. Increasing capelin stocks placed greater pressure on zooplankton, which in turn grazed less phytoplankton. These changes were described as a top-down trophic cascade triggered by the dramatic declines in cod biomass during the 1990s. Spawning aggregates or capelin appear irregularly on shelf waters, sometimes in very large numbers – schools several kilometers in size. Large numbers of predatory fish, seabirds and marine mammals, including humpback whales, feed on these large capelin schools. In Fig. 2.30, thousands of seabirds, mainly shearwaters (Puffnus spp.), and many humpback whales are shown feeding on a school of capelin near the Barren Islands. Sand lance are schooling zooplanktivores that are common in the entire edge zone of the Gulf of Alaska (Springer and Speckman, 1997). They occur in abundance
Figure 2.30: Humpback whales and seabirds feeding on a very large school of capelin near the Barren Islands, lower Cook Inlet (photograph courtesy of A. Kettle, USFWS).
Ecosystem Structure 69
in eastern Prince William Sound, where a stock estimated to be around 70 tons occurs (E. Brown, pers. comm.). Sand lance aggregate in dense shallow-water schools, easily observed from low-flying aircraft. In samples of nearshore fishes taken with beach seines, sand lance were the most abundant fish taken off sandy beaches (Robards et al., 1999b). The maximum energy content of adult sand lance is reached in the spring, which then declines through the summer and into the fall. They spawn in the fall in Kachemak Bay in lower Cook Inlet (Robards et al., 1999a) and spawn relatively few yolky eggs that develop attached to the seabed just below the lowest tides. Sand lance burrow into soft-bottom sediments at night and perhaps over longer periods during the winter. Sand lance represent a major food item for a large number of seabirds in the Gulf of Alaska (DeGange and Sanger, 1986), and their presence or absence can be crucial to the reproductive success of some seabirds. Eulachon are anadromous pelagic zooplanktivores living on the outer shelf of the Gulf of Alaska but returning each spring to spawn in freshwater (Love, 1966). As a result, these smelts are only temporary residents of the coastal waters, where they are fished heavily for subsistence purposes. When they school on the inner shelf, eulachon are consumed by most of the same predators that eat capelin and herring. At some locations, juvenile pink salmon can be a substantial forage base. In Prince William Sound, the combined release of hatchery-reared juveniles and those entering naturally from streams and small rivers comes close to 730 million fish annually. Willette et al. (2001) calculated that, during the period 1994–1996, as many as 75% of these juveniles could be consumed by predators during their first 45 days in the ocean. If the juveniles enter saltwater at 0.3 g each and grow at 4% of their body weight per day, and further, if the mortality all occurs half-way through the 45-day period (for computational ease), predators eating these 547 million juveniles will conservatively consume about 400 MT of fry before the young fishes have a chance to adapt to their new environment. Local populations of seabirds, adult herring and pollock, and several juvenile gaddid, have been documented as the major juvenile pink salmon predators (Willette et al., 2001; Scheel and Hough, 1997). Finally, it is now recognized that all forage is not the same when it comes to energy content – some of the small schooling fishes and zooplankters are much more “fatty” than others. Feeding experiments comparing the growth of marine mammals and seabirds reared on foods of different energy densities have demonstrated that growth, survival, and reproductive capacity are all enhanced when the food is rich in lipids (see Springer and Speckman, 1997, for a review). Since the dominance patterns of forage stocks vary considerably over time, some “fatty” species such as herring will support enhanced consumer growth in some years, but not in others when “less-fatty” species (such as juvenile pollock) become dominant. For sand lance, their maximum energy content occurs soon after the spring plankton bloom and coincides with the breeding seasons of many of their predators, e.g., seabirds and some marine
70
Long-Term Ecological Change in the Northern Gulf of Alaska
mammals (Robards et al., 1999a). The effects of different energy content of prey can be partially modified by consumers who are able to switch to other forage sources when their preferred food is limiting. However, in the case of some seabirds, switching to another forage base may not be an option because of foraging range and the depth of feeding limitations (see Section 2.5.3).
2.4.2. Efficient Foraging on Patches Aggregating behavior of forage species probably makes the transfer of production through the food web relatively more efficient than if these aggregations did not occur. Supporting evidence comes from early attempts to model the marine ecosystem, including commercial stocks of pelagic and demersal fishes in the Bering Sea (Laevastu and Favorite, 1981). Modelers were surprised to discover that the levels of zooplankton required to sustain observed fish production were apparently much higher than reported from field studies. This discrepancy was eventually resolved (at least in part) when it was realized that plankton net samples integrate much of the important in situ small-scale patchiness that represents critical feeding opportunities in the ocean; the statistically averaged or operationally integrated abundances are only abstractions of the “real world,” not true characterizations of the actual feeding environment. Recent advances in acoustic and optical methods
BOX 2.3: IMPORTANCE OF TIMING by Theodore Cooney In a strongly seasonal ocean such as the northern Gulf of Alaska, conditions supporting productivity at the base of the food web are constantly changing. While the overall temporal pattern of the annual production cycle is generally similar from year to year, in reality the specifics of any 2 years are never exactly the same. Some years, the plant bloom is early, intense, and truncated, while for other years, it is late, moderate, and extended. Consumers living in this system have adapted their life cycles to address a kind of “general” seasonality in critical forage availability. This means that when conditions deviate markedly from the norm, parts of the food web become disconnected or less than optimally functional, and some consumers face failing conditions. One successful strategy adopted by some fishes and invertebrates to address this temporal and spatial “match/mismatch” problem is to extend the period during which they release eggs and larval forms, thereby broadening the temporal window to ensure that at least some progeny will find optimal survival conditions.
Ecosystem Structure 71
of census demonstrate the presence and trophic importance of a “grainy” feeding environment. In the northern Gulf of Alaska, “patch-dependent feeding” is fostered by the behaviors and life histories of the dominant forage species, especially reproductive aggregations, which, as we have seen, serve to transfer immense quantities of matter and energy through selected parts of the food web (Thomas and Thorne, 2003). Patchiness is also fostered by the physical environment where fronts and eddies serve (under some conditions) as collection sites for zooplankton and other weakly swimming pelagic forage populations. In this way, food sources are spatially distributed in relation to the physics and geology, while at the same time being temporally ordered in ways that reflect the unique characteristics of the annual production cycle at each time and place. The timing of the spring diatom bloom in Prince William Sound differs from year to year by as much as three weeks (Eslinger et al., 2001). This variability probably affects the coupling or trophic phasing between producers, first-order consumers, and higher trophic levels. However, little is known about the ecological consequences of early or late blooms. Also, there has been little attention paid to the trophic implications of renewed fall productivity in coastal and shelf waters. For some fishes that depend on late-season energy provisioning prior to winter – such as juvenile Pacific herring – the fall portion of the production cycle may be critically important to the success or demise of a year-class.
2.4.3. Food-Web Complexity and Efficiency Food chain length, or the number of steps between photosynthetic producers and the apex consumers of the production, is determined primarily by the physical size of the dominant primary producers (Ryther, 1969). The surface waters of most deep ocean environments are typically composed of small phytoplankton responding to nutrient limitations (such as iron). In these environments, more trophic exchanges are needed to generate stocks of suitably sized forage organisms than in shelf or coastal regions where larger primary producers (colonial and chain-forming diatoms) feed some forage stocks directly (see Fig. 2.31). For example, large diatoms feeding krill, which then immediately serve as food for some birds and mammals, defines a highly efficient two-step transfer mechanism. Conversely, tiny diatoms and flagellatefeeding protozoans, which are in turn consumed by copepods, then krill, small fishes, and squids, represent a system that passes on only about a tenth of the energy to apex consumers. Using this theoretical approach, Parsons (1986) described food-web transfers in different environments of the Gulf of Alaska and found that given similar rates of primary production, fjord and shelf waters were capable of supporting twice the apex production of estuarine regions, and approximately 100 times those found in the open ocean. Recent revisions (Welschmeyer et al., 1993) to estimates of
72
Long-Term Ecological Change in the Northern Gulf of Alaska
the open-ocean primary production (adjusted upward by a factor of at least two) still do not bring the deep-ocean apex production close to that on the shelf. Within the seasonal production cycle, food-web efficiency probably change in environments where nutrient limitation shifts the dominant size of primary producers from large to smaller forms in the summer and early fall. When springtime diatom populations are replaced by microflagellates during periods of nutrient limitation, the general trophic efficiency likely declines because a micro-consumer link is inserted into the transfer process. Thus, not only is primary productivity limited by nutrients in most shelf and coastal ecosystems during the summer and early fall, but a more complex food web probably becomes less efficient at transferring energy to higher- level consumers during this same time (Fig. 2.31). There are some exceptions. Larvaceans and pteropods – macrozooplankters adapted to feed directly on very small particles – continue to provide efficient trophic linkages to some apex consumers during the summer and fall. However, this forage base is rarely as abundant or as energy rich as are the large copepods and euphausiids occurring in the spring. To summarize, interaction between upper-layer physics and shelf diatom populations poised to bloom under improving light conditions each year establishes
Figure 2.31: Change in food-chain efficiency in passing energy to seabirds and mammals as dominant primary producers shift from large diatoms during the spring bloom to smaller phytoplankton in the late spring and summer.
Ecosystem Structure 73
the conditions for photosynthetic responses by the plants ranging from short intensive episodes of growth to prolonged periods of lesser and intermittent productivity (Eslinger et al., 2001; Sambrotto et al., 1986). When the upper layer of the ocean stabilizes early and strongly, the resulting pulse of primary production is intense but short-lived because of rapid nutrient diminishment. However, when the upper-layer stability is weak and disturbed by frequent storms, the bloom is much less intense but substantially extended over time. The consequences of these differences are: (1) Under the intense and temporally truncated bloom, the accumulated plant matter is poorly coupled to the pelagic food web, since grazers surviving winter require time to produce their annual broods. Under these “poorly coupled” conditions, a substantial portion of the organic matter sinks out of the surface layers; and (2) With the less intense but prolonged bloom, the pelagic food web is enriched because grazer populations have time to expand and much less organic matter falls to feed seabed consumers. In these quite different ways, organic matter is passed through the food web by the interactions between the living pelagic and benthic assemblages and the physical environment. That growth and reproduction occurs for many consumer stocks in the spring and early summer is evidence of the predictability of the annual diatom bloom
BOX 2.4: HOW DO WE KNOW WHO EATS WHOM? by Theodore Cooney Ever since questions about forage dependencies were first raised, food chains and webs have been constructed mostly by directly describing the stomach contents of consumers. This practice is relatively straightforward but by no means foolproof. For example, digested or partially digested food items are difficult if not impossible to identify. Also, the rate of digestion is not constant for all forage, so what is reported as food by percentage of stomach contents (weight or number) may not be at all correct. In the case of some seabirds, forage dependencies are determined by direct observation – telescopic monitoring of adults provisioning chicks with identifiable food items. More recently, stable isotope analyses have been used to assign “trophic status” using the rule that “you are what you eat,” and specific fatty acid signatures are often traceable to specific prey. These new methods are making it unnecessary to rely entirely on expensive and problematic quantitative stomach analyses. However, none of these techniques by themselves is comprehensive, so marine ecologists seeking information on feeding behaviors and forage dependencies use a suite of different methods.
74
Long-Term Ecological Change in the Northern Gulf of Alaska
BOX 2.5: FEEDING TERMINOLOGY by Theodore Cooney One of the ways that populations of consumers in the sea differentiate themselves is by prey selection. Those that target plankton are termed planktivores, those that select fishes are called piscivores, and those that subsist on a combination of different food types are termed omnivores or generalists. Assigning a species to one of these categories is complicated by the fact that forage dependencies may change seasonally and as an individual moves through its various life stages. Fishes, for example, are mostly planktivores in their larval and juvenile stages, but select other fishes and squids as they mature. Exceptions such as herring and pollock continue to feed on plankton (as adults) when it occurs abundantly (swarms and layers), but also supplement their diets by feeding on fishes when plankton sources are diminished seasonally.
and the short, efficient food webs operating at that time in shelf and coastal environments. But, from midsummer through fall (in all but the open ocean system that is always structurally inefficient), nutrient limitation of primary production and a more complex food-web likely defines less energetically efficient matter transfer – although clearly important to consumers such as the juvenile stages of some fishes – Pacific herring in particular. These seasonal forage shifts suggest planktivory as a primary foraging strategy for consumers early in the season, giving way to omnivory in the summer and fall months, and piscivory during the winter and early spring. In this kind of ecosystem, a generalist such as the walleye pollock is obviously superbly suited to thrive.
2.5. Strategies for Survival 2.5.1. Introduction Alan M. Springer All living things must heed the biological imperative to replace themselves during their reproductive life spans and thus survive as a species. To succeed, plants and animals have evolved life history characteristics, behaviors, and anatomical and physiological adaptations that allow them to find food, conserve energy, keep themselves from becoming the meals of things larger or fiercer, and produce viable offspring.
Ecosystem Structure 75
Strategies for survival vary between species, but all are designed to help individuals take advantage of environmental predictability, cope with environmental uncertainty, and avoid predators. When a strategy fails, it is usually because a species is overcome by forces of a magnitude outside the bounds of common variability and for which it is evolutionarily or circumstantially ill prepared. When this happens, it can place a population or even an entire species at risk. Examples of such strong forces include introduced predators, extreme weather such as El Niño, fisheries, diseases, and contaminants. Many times, causes of change are not conspicuous and are difficult to explain. Because high-level predators in an ecosystem represent an integration and culmination of responses below them, they can be sensitive indicators of ecosystem structure and change. In the following sections, we examine survival strategies of nine prominent species in the Gulf of Alaska, which have experienced major changes in abundance in the past several decades. These focal species – three fishes, three seabirds, and three marine mammals – all exploit the marine environment differently, and all have stories to tell about variability and the nature of change in the Gulf of Alaska. Besides belonging to three different phylogenetic groups, these species span a wide range of physiologies, life history patterns, and trophic levels. Thus, they can serve as sentinels of their environment and its changes. And, in these cases, the reasons their populations have increased or decreased are not entirely certain. By considering their recent histories in the context of their strategies for survival, we may be able to better understand why populations vary and, ultimately, how ecosystems work.
2.5.2. Introduction to Fishes Theodore Cooney In the marine ecosystem, fishes coexist within a matrix of physical constraints and competing biological populations in ways that define their respective niches. These strategies include unique reproductive, growth, and feeding activities that provide competitive leverage for forage resources and ecological space. By exploiting a specific range of physical tolerances and interactions, each species projects a unique “habitat dependency” that describes the conditions necessary for its long-term reproductive success. The life history strategies for marine fishes involve a variety of adaptations, including the life span, age at maturity, fecundity (number of eggs per female), aspects of the early-life history, forage and feeding habits, and reproductive behaviors. The range of adaptation for these features is remarkable. Rockfishes are among the longest living of all fishes – some exceeding 50–70 years of age. Rockfishes are also live bearers (ovoviviparous), but this is the exception among marine fishes. In contrast, a pink salmon is very short-lived – spending but 1 year in freshwater and 1 year in the ocean. The Pacific
76
Long-Term Ecological Change in the Northern Gulf of Alaska
cod (Gadus macrocephalus) is extremely fecund, and relatively long lived, up to 30 years; a female may spawn millions of eggs each year. The eggs of some fishes are pelagic (freely drifting), whereas for others, they are demersal (sticking to rocks or attached plants). Some fishes such as the salmonids and smelts are anadromous – reproducing in freshwater but growing and maturing in the ocean. Some species are planktivores, whereas others are piscivores or omnivores. It is not uncommon for the diets of different life-history stages to differ. Larval and juvenile pollock are principally planktivores, whereas the adults have a more catholic diet, including plankton, fishes, and invertebrates. Some species are even cannibalistic, eating their own young. One fairly common reproductive strategy for fishes is to release large numbers of eggs over a long period each year. Although the probability that any one egg will complete its life cycle to reproduce is vanishingly small, the probability that “some” eggs/larvae/juveniles will find suitable conditions for survival is a certainty. Because the potential for an unusually large survival is present every year for these species, environmental conditions will occasionally allow an “exceptional” number of survivors who will dominate the population for many years. In the case of the Gulf of Alaska, significant variability in year-class strength has been observed in many different fish stocks. In the following, we describe the life history characteristics for three dominant marine fishes – the pink salmon (Oncorhynchus gorbuscha), the Pacific herring (Clupea pallasii), and the pollock (Theragra chalcogramma) – to better understand how each has done in the past, or might do in the future under the various ocean climate regimes in the Gulf of Alaska.
2.5.3. Pink Salmon Theodore Cooney The pink salmon, Oncorhynchus gorbuscha, is the smallest and most numerous of the Pacific salmon. Odd and even year-classes do not interbreed, and with a strict 2-year life cycle, pinks spend about a year maturing in the ocean. Juvenile pink salmon enter the marine environment from freshwater natal areas as juveniles weighing about 0.3 g. After approximately 400 days, they return to spawn and die at an average weight of 1.7 kg; other species of Oncorhynchus grow to much larger adult sizes because of longer periods of feeding and maturation (both marine and freshwater). Over the geographic range of pink salmon in North America, odd and even year-classes trade dominance patterns (Rogers, 1986); runs in Washington and southern British Columbia are typically dominated by odd-year fishes. In southeastern and south-central Alaska, some stocks are dominant in the odd years, some in even years, and others demonstrate a mixture of year-class dominance over time. The very small runs of western Alaska pink salmon are almost exclusively even-year fish. These year-class patterns are unexplained.
Ecosystem Structure 77
In south-central Alaska, the largest runs of pink salmon occur in Prince William Sound, in lower Cook Inlet, and on Kodiak Island. In Prince William Sound, a surprising percentage (>50%) of both odd- and even wild year pink salmon spawn in the inter-tidal reaches of the many small streams and rivers that characterize their natal habitats (Helle, 1970). There is a tendency for odd-year spawners to use the higher regions of these streams and rivers. Each female pink salmon spawns about 1500 eggs before dying (Fig. 2.32). At least two eggs must produce surviving adults to sustain the population of males and females in the absence of a fishery. The large, yolky eggs are deposited in nests dug from the gravels by the females. Spawning usually begins in late June and concludes in September; most occurs in July and August. After a few weeks in the late summer and fall, the eggs hatch and the larval salmon – the alevins – burrow into the sediments to overwinter on yolk-sac reserves. By the following late March and early April, developing juveniles have absorbed most of their yolk and begin emerging, a process that lasts 60 days or more (Bailey, 1969; Taylor, 1988). The emerging juveniles’ transition to saltwater and the year class is entirely in the coastal zone by July (Cooney et al., 1995). There, the juveniles rear for 3–4 months in the shallow edge zone, feeding primarly on zooplankton but also on harpacticoid copepods, insects, polychaetes, and larval fishes (Parker, 1997). In late summer and early fall, surviving juveniles begin moving into the coastal current, initiating a feeding migration that terminates for survivors about 10 months later at their home streams or hatchery. The pink salmon survival strategy places a premium on timing and size. The relatively small number of large eggs per female (compared with many non-salmonid fishes) produce alevins and then juveniles that are much larger than the larvae of pollock and herring. The sedentary period of intertidal and freshwater rearing over the fall and winter months shelters most alevins from predators. However, stream bed scouring from heavy fall rains, freezing, and low oxygen during the winter are all factors that can reduce a year-class during this life stage (Alaska Department of Fish and Game). Pink salmon fry average about 30 mm in total length at entry into marine waters. Surviving juveniles exhibit rates of growth of about 3 to 5% of their body weight per day during early marine residence supported by the gradually warming springtime nearshore ocean and abundant food. A juvenile, entering at 0.3 g, will weigh between 4.5 and 27.0 g after only 90 days in the ocean. Larger size probably allows better escape predators – primarily birds and larger fishes (Willette, 2001). For most fishes, the greatest losses occur in the egg and larval stages. Studies of pink salmon demonstrate this is the case as well. Of the 1500 eggs placed in the redds by a female, some go unfertilized, some are consumed by predatory fishes and birds, some fail to mature correctly, and some of the overwintering alevins succumb to the rigors of the freshwater rearing environment. Heard (1991) estimated that only about 8% of the eggs deposited by female pink salmon develop into viable juveniles. So, hypothetically,
78
Long-Term Ecological Change in the Northern Gulf of Alaska
Figure 2.32: Life cycle of western Prince William Sound pink salmon with alevin rearing areas (1) and migration pathways for out-migrating juveniles (2–4) (top left) and returning adults (5) during the 2-year life cycle. Hypothesized open ocean distributions of maturing adults are shown at the top right, and the sizes and averaged numbers of different life stages surviving from the embryos of a single adult female are depicted in the bottom panel.
Ecosystem Structure 79
of the 1500 eggs deposited in redds by a female, only about 120 alevins survive the winter to enter the nearshore as juveniles, an immediate loss of 1380 potential adults. Perhaps 75% of juvenile pink salmon (wild and hatchery-reared) entering Prince William Sound will be eaten by predators during the first 45 days (Willette et al., 2001). So, after early marine residence and when the juveniles are leaving nearshore waters for the open ocean, just 30 juveniles remain from the original 1500 eggs. The estimated survival to adulthood is about five adults per female spawner in Prince William Sound over the period 1962–1997, according to the Alaska Department of Fish and Game. This level of production has provided for a sustained commercial fishery and generally met desired spawing escapements. The critical survival of juvenile pink salmon in nearshore marine rearing environments depends on food supply and losses to predators (Cooney et al., 2001; Willette et al., 2001). A remarkable concordance between the peak of the juvenile emergence into nearshore waters and the seasonal high in biomass of large calanoid copepods (mostly Neocalanus spp.) suggested that pink salmon in Prince William Sound have evolved a mechanism to enssure plentiful food supply each year for the critical juvenile stage (Cooney et al., 1995). However, further work has demonstrated an even more sophisticated aspect of this timing adaptation. Willette et al. (1999) reported that two of the major predators of juvenile pink salmon – adult walleye pollock and Pacific herring feed almost exclusively on near-surface swarms and layers of large copepods and krill at the same time the young salmon are entering the edge zone. Thus, a kind of predation shelter is established for a few weeks in late April and May that protects the juveniles while they exploit the growth conditions and relative safety of the shallows (Fig. 2.33). In years when the zooplankton layers are not well developed, pollock and herring supplement their diets by feeding more on small fishes, including salmon. Willette (2001) also found that the feeding behavior of juvenile pink salmon occasionally places them at high risk to predation. While there is little evidence that the growth rates of juveniles in Prince William Sound are severely limited by food – marked density-dependent growth is absent except at very high fry abundance – the young salmon apparently prefer the largest copepods available during the spring. When large-bodied Neocalanus or Calanus become scarce in the shallows, the juveniles begin searching over adjacent deeper water and, by doing so, place themselves at greater risk to encounters by adult pollock and herring. In this way, the pink salmon life history strategy for stocks in Prince William Sound protects the juvenile stages early in their marine residence. The apparent strategy is to match the entry timing of the juveniles with the planktonic feeding period of their most important predators – adult pollock and herring. While the larger fishes are feeding on dense surface layers of large calanoid copepods and other macrozooplankters during April, May, and early June, the fry grow in the relatively protected shallow edge of the sound. Water temperature is believed to exert an influence on fry
80
Long-Term Ecological Change in the Northern Gulf of Alaska
Figure 2.33: Relationships between juvenile pink salmon and their food, predators, and amounts of alternative prey for predators – adult pollock and herring. Better juvenile rearing conditions during early marine residence results in higher survivals and better year-class strength (left panels) than when the rearing conditions are poorer (right panels). Red arrows depict predation, with the relative width of the arrow representing importance.
growth at this time. In early June, when the large calanoids begin leaving the upper layers and are no longer available to pollock and herring, the surviving juveniles become alternative prey for other fishes. At this same time, increasing numbers of small gadids (cod and pollock) and adult salmon and Dolly Varden trout also begin invading the warming shallow fry nursery waters. During cool springs, body mass will increase relatively slowly, and the fry will be smaller and more vulnerable in early June. Conversely, when the spring is warmer than average, fry growth rates are expected to be higher. Size-dependent mortality has been demonstrated for pink salmon – the larger juveniles apparently being able to more efficiently avoid predation. In this complex way, temperature-dependent growth and predation sheltering, coupled with the feeding behaviors of the juveniles and their
Ecosystem Structure 81
dominant fish and bird predators determines the survival of a year-class early in the marine life history. The degree to which the early life history strategy exhibited by Prince William Sound pink salmon represents other regions in the northern Gulf of Alaska is unknown. However, since pollock and herring are common and abundant fishes in the coastal zone of the entire Gulf and large calanoids and other macrozooplankters are present in other fjords and sounds, it seems likely that these findings can be considered in a general way until further research proves otherwise. The interested reader is directed to Brodeur et al. (2003) for a thorough review of the latest research on the early life history of Pacific salmon.
2.5.4. Pacific Herring Theodore Cooney Pacific herring, Clupea pallasii, have immense ecological importance and commercial value. These small schooling fish are major contributors to food webs, supporting a wide variety of other fishes, seabirds, and marine mammals in coastal waters (University of Alaska Sea Grant, 2001). Herring are distributed in the eastern North Pacific Ocean from California to the northern Bering Sea (Hay et al., 2001). Separate spawning stocks occur in southeastern, central, and western Alaska. Spawning occurs in the spring, beginning in the south in March and concluding in Norton Sound in June. In Prince William Sound, spawning usually occurs in mid-April at water temperatures close to 4°C. (Fig. 2.34). Each female deposits many thousand small sticky eggs on intertidal and shallow subtidal substrates, including seaweeds and seagrasses. The annual herring spawning draws large numbers of birds, other fishes, and mammals, all targeting the eggs and massed adults in a nearshore feeding frenzy that may last (off and on) for 2–3 weeks. It is not unusual for a spawning population to stretch for tens of miles in the coastal zone, the milt of the males being clearly visible in the water from a distance. For 27 years from 1973 to 1999, the average total distance over which spawning occurred in Prince William Sound was about 90 km (Norcross and Brown, 2001). After spawning, adults feed aggressively to replenish their post-spawned energy reserves. Many move immediately into the surface waters to prey on layers of large calanoids (Neocalanus spp.), euphausiids, and pteropods. Adults supplement their plankton diets when necessary with small fishes, including juvenile salmon (Willette et al., 1999). Unhatched herring embryos are eaten by seabirds, migrating shorebirds, and a variety of fishes and invertebrates exploit the energy-rich egg masses that can also be exposed and further eroded by heavy wave action (Rooper, 1996; Bishop and Green, 2001). As many as 75% of all eggs deposited by spawners can be lost to predation in some years (Rooper et al., 1999).
82
Long-Term Ecological Change in the Northern Gulf of Alaska
Figure 2.34: Life cycle of Pacific herring in Prince William Sound, showing major spawning areas, some proposed larval dispersion paths, and overwintering areas for juveniles and adults (top panel). The sizes and timing of various stages are shown in the bottom panel.
Ecosystem Structure 83
Surviving embryos incubate for approximately 3 weeks in a gradually warming nearshore habitat (Biggs et al., 1992). Hatching begins in mid-to-late May, and the tiny (<10 mm total length) yolk-sac larvae are dispersed by tidal, wind-, and freshwaterdriven currents. Following a period of larval drift that may last for a few weeks, the surviving larvae metamorphose into juvenile herring by mid summer. While in their planktonic phase, the larvae are believed to subsist on small zooplankton – mostly copepods and copepod eggs and nauplii. The developing juveniles continue to exploit a primarily pelagic food source – small zooplankters – but will opportunistically feed on other sources as well, particularly when they enter shallow, nearshore rearing environments (Norcross et al., 2001; Foy and Norcross, 1999; Foy and Paul, 1999). In Prince William Sound, juvenile herring inhabit the nearshore embayments and fjords for 2–3 years before fully recruiting into the adult stock. Their distributions do not generally overlap with adults except during spawning. Herring in the Gulf of Alaska can live for 16 years and attain a body mass of 250 g (Hay et al., 2001). Adult herring aggregate in the fall, staying in schools through the spring spawning. Dense, prespawning layers of these fish can be located acoustically in areas adjacent to preferred spawning habitats. In some locations, these concentrations provide extensive prey for humpback whales, sea lions, and fishes foraging in the sound during the late fall and winter months (Thomas and Thorne, 2003). The herring survival strategy is to invest large amounts of energy in huge numbers of demersal eggs each year, betting that a small but sufficient number will incubate, hatch successfully, and give rise to dispersed populations of pelagic larvae, juveniles, and finally adults. There is evidence that spawning habitats are “selected” on the basis of their advantageous location to the subsequent extended period of larval drifting and juvenile, rearing (Lasker, 1985). In Prince William Sound, the timing of spawning and the duration of egg incubation occurs during the transition between the late fall/winter/early spring pattern of surface flow-through and the late spring/summer period of generally unorganized currents characterized by numerous reversals and eddies (Vaughan et al., 2001). This timing probably lessens the chance that large quantities of larvae will be washed from the sound, while at the same time providing a physical means to both retain and distribute the tiny fishes throughout the protected region. By August, age-0 juveniles school in peripheral bays and fjords. The juveniles feed mostly on zooplankton, storing lipid reserves for the coming winter (Foy and Norcross, 1999). July and August represent the warmest months in the coastal zone, and also the peak of the small copepod bloom (the genera Pseudocalanus, Acartia, Centropages; Cooney et al., 2001). The life history strategy utilizes the midsummerto-fall feeding period as a critical conditioning window for the juveniles. Extensive studies of the whole-body energy content (WBEC) of age-0 herring demonstrate the importance of summer and fall lipid provisioning and growth (Paul and Paul, 1998, 1999; Paul et al., 1998). In Fig. 2.35 the average energy content (in kilojoules/gram) is shown for age-0 herring (entering the first winter of their life)
84
Long-Term Ecological Change in the Northern Gulf of Alaska
5.5 Avg. WBE field coll. 12/95 WHOLE BODY ENERGY (Kl g-1)
5 4.5 4
Avg. WBE field coll. 3/96
3.5 Avg.WBE dying luv. 3 2.5 2 40
50
60 70 80 STANDARD LENGTH (mm)
90
100
Figure 2.35: Whole-body energy content (WBE) of age-0 herring from Prince William Sound in December 1995 (mean value, magenta line), dying starved specimens in a laboratory experiment (mean value, red line), and field collection in the March 1996 (mean value, green line).
in December 1995 (upper line) and the following March (middle line). An experimental starvation of the December fish yielded the spread of values of dying juveniles (blue dots) (Paul and Paul, 1999). Juveniles that fail to store sufficient energy during their summer growth may starve later in the winter (Patrick, 2000). Although there is little evidence for a complete fast during the plankton hiatus from December though February, the occasional ingestion of plankton during this time is believed to be of only minor physiological importance. Since water temperature controls the level of metabolic activity in overwintering juveniles, warmer-than-average conditions may exhaust fat reserves before the spring plankton bloom in March and April. Winter starvation can be a significant factor constraining the recruitment of herring to adult stocks in Alaska waters (Norcross and Brown, 2001). Many common nearshore fishes in Prince William Sound leave the cooling upper layers in the fall for warmer temperatures at depth and do not return until mid-spring each year (Rogers et al., 1987). This migration of potential predators to depth may benefit overwintering juvenile herring that stay behind in the cold upper layers. Unusually warm winters would both allow more predators to remain in the shallows, and would also accelerate juvenile metabolism – both factors detrimental to juvenile survival.
Ecosystem Structure 85
Juvenile herring remain in shallow coastal areas for about 3 years before joining adult winter aggregations. Apparently, for the older juveniles, the amounts of plankton in late March and April are also critical determinants of their year-class strength. Winter starvation continues to be a risk, but not as much as it is for the age-0 fish. For most fishes, losses are greatest in the youngest stages and decrease with age. For Pacific herring in Prince William Sound, it has been estimated that the percentage of survival is lowest for the period of the larval drift (1–7%), but most variable for the period of winter starvation (5–99%) (Norcross and Brown, 2001). All of the early life stages play some role in ultimately determining the recruitment strength of a year-class. In addition to the interaction of the usual limiting factors (the growth environment, predators, and human causes), disease can also play a big role in determining the production status of Pacific herring populations over time. In Prince William Sound, viral hemorrhagic septicima (VHS) and other pathogens have apparently prevented the recovery of stocks damaged by the oil spill of March, 1989 and resulted in reduced or cancelled fisheries in many years following that disaster (Marty et al., 1998).
2.5.5. Walleye Pollock Kevin M. Bailey and Lorenzo Ciannelli Introduction Walleye pollock, Theragra chalcogramma, typifies marine fish species that are highly fecund, producing millions of eggs per individual spawner, and which have highly variable mortality rates in early life. A consequence of this strategy is fluctuating annual recruitment (abundance of individual year classes in the fishery; Fig. 2.36) that must be buffered by the averaging effect of many age classes in the population. These life history adaptations allow walleye pollock to take advantage of episodic favorable environmental conditions and thrive by maximizing reproductive success and minimizing mortality. On the other hand, pollock employ a complex array of adaptations to their environment (Olla et al., 1996), only some of which are understood. Pollock has not always been the dominant species in North Pacific ecosystems, but its current success may be the product of a unique match between life history and a combination of environmental conditions. In contrast to the success of pollock in the North Pacific, its congener Theragra finmarchica is hardly abundant in the Atlantic Ocean, being a relatively rare catch in the Barents Sea. Two intriguing questions are: How have the opportunities for these congeneric species differed? Why are their abundances contrasting in the two oceans? Pollock is currently a key species in northern North Pacific ecosystems in the sense that its dynamics influence trophic levels above and below it. For example, declines of pollock have been linked by some scientists to the collapse of the Steller
86
Long-Term Ecological Change in the Northern Gulf of Alaska
4000 AGE-1 RECRUITS (millions)
3500 3000 2500 2000 1500 1000 500
00 20
95 19
90 19
85 19
80 19
75 19
70 19
19
65
0
YEAR
Figure 2.36: Year-class strength of walleye pollock in the western Gulf of Alaska, 1969–2000, determined from virtual population analysis (VPA) (from Bailey et al., 2003).
sea lions in the Gulf of Alaska (Merrick et al., 1997), bird mortalities (Springer, 1992), and to broader and more complex changes in ecosystem structure (Estes et al., 1998). As a major predator, pollock remove a lot of prey organisms, most noticeably competing with humans for valued species such as shrimp and juvenile salmon. In this sense, pollock may be a “waistband” species (Cury et al., 2000), whose dynamics regulate the structure of the ecosystem through both top-down and bottom-up effects (see Box 6.2 for a discussion of the oscillating control hypothesis). After an ancestral gadoid form invaded the North Pacific Ocean during the Pliocene period some 3 million years ago, pollock gained a strong foothold by adopting a generalist strategy, as opposed to niche specialization. Pollock (Fig. 2.37) thrive in temperatures from about 1 to 10°C in habitats as diverse as eelgrass beds in Puget Sound to the open ocean environment of the Aleutian Basin; it feeds on a variety of prey, from planktonic crustaceans and fishes to benthic clams and worms, and virtually everything taxonomically in between. Pollock inhabit and generally dominate ecosystems from Puget Sound to the northern Bering Sea, and across the Pacific Ocean from the Sea of Okhotsk to the Korean Peninsula. Of course, pollock populations wax and wane through time, so their own impact on ecosystem structure may be greater or less through time. Their abundance may be influenced by both natural environmental and biological factors and by
Ecosystem Structure 87
Figure 2.37: Juvenile walleye pollock Theragra chalcogramma (photograph courtesy of the Japan Agency for Marine-Earth Science and Technology, JAMSTEC).
human harvesting. Unfortunately, we have only observed one complete cycle of the pollock saga, in the Gulf of Alaska from historical lows in the 1960s to a high in the early 1980s, declining to another low in the 1990s through the present. Pollock in Puget Sound have undergone a more dramatic swing in amplitude, dominating groundfish populations in south Puget Sound in the 1980s and then becoming virtually extinct in the 1990s. Because the time series of pollock abundance is so short, our knowledge of how pollock populations respond to the environment and to self-regulation has a low degree of statistical confidence.
Adaptations for Survival Strategies that the species uses for survival must account for conditions that maximize reproduction and increase survival. The processes that maximize survival and reproduction interact with density-dependent forces within the pollock population to regulate its numbers.
Strategy: maximizing reproduction Reproductive strategy involves development of reproductive capacity and constraints. Pollock mature at 3–4 years of age, and every individual female produces millions of eggs every spawning season. Pollock are iteroparous (spawning successive annual batches), determinate (a fixed number of eggs spawned per season), and partially synchronous (multiple stages of eggs in the ovaries spawned in clutches). An individual
88
Long-Term Ecological Change in the Northern Gulf of Alaska
female may spawn her eggs in several clutches over a period of days and up to 14 clutches over the period of a month. Eggs are released into the water and fertilized by paired males after an elaborate courtship. Eggs are spherical, about 1.0 mm in diameter, take about 2 weeks to hatch at 6°C, and have enough yolk for hatched larvae to survive for about a week without feeding. Another strategic aspect of reproduction is selection of a spawning site. The spawning locations and the geographic range of the pollock distribution are probably bounded by the environmental conditions that allow the survival of pollock eggs and larvae. Some species have adapted to conditions at the spawning location to minimize the number of eggs and larvae being swept away by ocean currents by developing strong swimming capabilities (Leis and Carson-Ewart, 1997). However, pollock larvae are very weak swimmers for the first few weeks and are unable to effectively swim against the strong Alaskan currents. In order to make up for their weak swimming ability, coastal stocks of pollock broadcast their eggs in deep inshore bays or in sea valleys and canyons that penetrate the continental shelf, which tend to have oceanographic features that favor retention of eggs and larvae over the shelf and near favorable nursery sites (Bailey et al., 1999). In more oceanic stocks, such as in the eastern Bering Sea, they tend to spawn where currents are very weak, or where transport into favorable nursery areas is climatologically (on the average) probable. The spawning regions of pollock are noted for mixing of coastal and nutrientladen oceanic waters and stratification of the water column, leading to enhanced productivity; these conditions favor the survival of early life stages of pollock. In the Gulf of Alaska, pollock typically spawn during the last week in March and first week in April in the Shelikof Strait (Fig. 2.38). In this area, mixing of the Alaska Coastal Current, the Alaskan Stream, and coastal water, along with springtime increases in sunlight and water column stratification, leads to an intense spring bloom and reproduction of zooplankton. Zooplankton prey of pollock larvae are further concentrated by physical features, such as eddies and fronts (Napp et al., 1996), leading to favorable feeding conditions. Finally, late larvae and juvenile pollock are carried toward favorable nursery areas, such as the waters around the Shumagin Islands. How adult pollock find their way back to these established spawning areas is currently unknown and controversial (Bailey et al., 1996). Whether there is natal philopatry (return to birth location) or spawning philopatry (return to a previous spawning site used by adults), either through genetic adaptation, imprinting, or social facilitation (learning), should play a role in how stocks are managed. Natal philopatry is a strategy to maximize the probability that the offspring will find suitable and persistent nursery habitats. In contrast, spawning philopatry through social facilitation would allow the population to capitalize on more ephemeral environmental conditions. Thus, the prevalence of one (natal philopatry) over the other (spawning philopatry) homing mechanism may depend on the variability of the surrounding environment.
Ecosystem Structure 89
Figure 2.38: Life cycle of walleye pollock in the western Gulf of Alaska showing spawning (1) and juvenile rearing areas (2–5) in top panel. The sizes and timing of various stages are shown in the bottom panel.
We know that the bulk of pollock spawning in the Gulf of Alaska has historically been highly concentrated in the Shelikof Strait. However, in recent years, pollock spawning appears to have shifted in location, and the Shelikof Strait aggregation is no longer dominant. New and large pollock aggregations have been recently found near the Shumagin Islands and the Unimak Bight.
90
Long-Term Ecological Change in the Northern Gulf of Alaska
Strategy: minimizing mortality Mortality rates of pollock eggs and young larvae are very high, ranging from 4 to 40% per day, but decline as the pollock develop (Fig. 2.39). In fact, larval condition can vary from year to year and by location, and a high percentage of larvae in the ocean has been observed to be in poor feeding condition (Theilacker et al., 1996). Studies have shown that egg and early larval development and survival is suboptimal at temperatures below about 0∞C and above 10–12∞C. Extremely high and low temperatures can be lethal to eggs and larvae, but generally for the Gulf of Alaska population, which is in the central part of its distribution, higher temperatures (6–7°C) tend to favor better survival, perhaps through one or more indirect mechanisms (Bailey, 2000). Optimal prey levels for successful feeding depend on many different conditions, including larval size, temperature, light levels, turbidity, and turbulence (Porter et al., 2005), but they generally range between 20 and 40 prey/liter (Theilacker et al., 1996). Growth variation is another strategy to minimize mortality. Faster growth rates may lead to lower mortality under some conditions (Anderson, 1988). According to Houde’s (1987) stage duration hypothesis, faster growth through early stages that are vulnerable to mortality is beneficial. This concept has been difficult to show in pollock larvae. There is some evidence that strong year-classes of pollock have attained a greater size-at-age as early juveniles (Bailey et al., 1996). Pollock eggs and larvae are also lost by predation to many different types of invertebrate planktonic predators; unfortunately, these organisms masticate their prey, making the quantification of the
MORTALITY RATE (1/day)
0.45
0.35
0.25
0.15
0.05
-0.05 0
5
10
15
20
25
30
35
40
45
AGE (days)
Figure 2.39: Mean mortality rates of walleye pollock eggs and larvae determined in field studies as the difference in abundance of daily cohorts between survey periods (from Bailey, 2000).
Ecosystem Structure 91
predation rate difficult. Immunological detection of pollock as prey, using antipollock antibodies, has been employed to identify pollock eggs and early larvae in the guts of amphipods, euphausiids, and mysids. Visual inspection of predator guts has been used to detect pollock eggs and larvae in the guts of eulachon, capelin, and adult pollock. Adaptive strategies of pollock to reduce predation loss are numerous, including: transparency of eggs and early larvae, low swimming speeds that minimize encounters with predators, lack of motion making them less easily detected by visual and mechanoreceptive predators, and deep-water habitat where there is little light and few predators. At later stages, predation on juveniles is an important source of loss to the population. Piscivorous fishes, including halibut, cod, arrowtooth flounder, and flathead sole, contribute to significant mortality (Livingston, 1993). Juvenile pollock are also prey to marine mammals and birds. As described in the following section, changes in ecosystem structure and the abundances of these predators can have important consequences to pollock population dynamics. The high variability of pollock abundance may be an adaptive strategy to run the gauntlet of predators awaiting them, thereby overwhelming the predation capacity of the community when larvae are abundant. Ultimately, with a successful and dominating population such as pollock, densitydependent processes may either enhance or inhibit population increases. At high levels of abundance, pollock may outrun their prey supplies (Anderson et al., 2002), leading to slower growth, delayed maturity, and decreased reproductive success. At high densities of pollock, predators may also undergo changes (swarming, shift feeding behavior, increase recruitment, etc.) that lead to density-dependent mortality. Cannibalism is also an important regulative process in pollock, especially in the Bering Sea (Dwyer et al., 1987). In the Gulf of Alaska system, cannibalism is not as prevalent, and it may be minimized by the relatively nonoverlapping distributions of adults and juveniles (Shima et al., 2002). In the Bering Sea, the tolerance of juveniles to water <2°C may be an adaptation to provide them with a refuge from adult predators. Thus, there is a complicated scenario of predation and cannibalism in a complex landscape, such as in the Bering Sea, where the population may be self-regulating through cannibalism, but where there is an interaction with thermal refuges, prey availability (Sogard and Olla, 1996), and removal of large predators by fishing.
Effect of Ecosystem Structure on Pollock Survival Changes in environmental regimes and ecosystem structure may have important effects on factors regulating the recruitment of pollock. For example, in the late 1980s, as the Gulf of Alaska ecosystem apparently shifted from one dominated by pelagic forage fishes and shrimps to a community dominated by piscivorous flatfishes and gadids, a shift occurred in the life history stage at which recruitment of pollock is regulated (Bailey, 2000). Prior to the shift in ecosystem structure, recruitment was
92
Long-Term Ecological Change in the Northern Gulf of Alaska
correlated with early larval survival, but after the shift, juvenile mortality rates increased and recruitment was regulated by predation mortality of juveniles. Therefore, control points may change from year to year, and depend on longer-term changes in the environment and community structure, such as those occurring with environmental and biological regime shifts. Environmental and ecosystem structure shifts may also have indirect effects on pollock survival, such as causing changes in the operation of density-dependent mechanisms. For example, Ciannelli et al. (2004) found that the level of density-dependent mortality in juvenile pollock increases when water temperature and predation intensity are high. Statistical models have been successfully employed to describe how the density-dependence structure of the population interacts with high- and low-frequency factors to regulate recruitment. Changes in ecosystem structure may also influence the distribution of pollock, their predators, and prey with consequent effects on pollock production processes. For example, there has been an increase in the abundance of pollock in nearshore small-mesh trawl surveys since the regime shift in the late 1970s (Anderson and Piatt, 1999). At the same time, pollock stock abundance in the Gulf of Alaska has declined precipitously. One might infer from these seemingly conflicting observations that the distribution of juvenile pollock, the stage largely caught by the small mesh trawls, has shifted inshore, making them more vulnerable to the coastal assemblage of predators.
Conclusions Pollock is an opportunistic species that is able to expand and adapt quickly to different environments. On the other hand, the population is limited by finding and adapting to local conditions that favor successful spawning (maximizing reproduction) and survival (minimizing mortality) of the early life stages. Local populations of pollock respond differently to shifting environmental regimes, as warming periods have seen those stocks at the southern margins of the pollock distribution falter or fail. In the center of its distribution of mass in the eastern Bering Sea, pollock have been (if anything) favorably impacted by periods of environmental warming. In the Gulf of Alaska, the situation appears more complex, as pollock have been initially favored by a warm environmental regime (e.g., stock increase in the late 1970s and mid-1980s), but negatively impacted afterwards in connection with a sharp increase of predator biomass. Pollock spawn once per year, in an event that is highly concentrated in space and time. Given the fragility of eggs and larvae to environmental conditions, and their concentration in space and time, the survival of a whole year-class is vulnerable to the vagaries of the ocean and weather, such as storms passing through Shelikof Strait. On the other hand, pollock dynamics are buffered partly by multiple spawning stocks, spawning in different locales, and by multiple age classes in the population. Spawning in different locations moderates the effects of temporal variation in habitat suitability
Ecosystem Structure 93
by taking advantage of spatial variation. The long life span of pollock is an adaptation that tempers the high variation in year-class strength. A high abundance of predators on adults, as well as commercial fishing, which removes older age classes, reduces the age span over which the mean abundance is averaged (and perhaps other aspects of the contribution of older fish to the population’s viability). Consequently, the population will be more dependent on fewer age classes, hence contributing to overall stock variability (Longhurst, 2002).
Acknowledgements Section 2.5.5 is contribution FOCI-0491 to NOAA’s Fisheries-Oceanography Coordinated Investigations.
2.5.6. Comparing Fish Life Histories Theodore Cooney As we have seen reproductive strategies, food web dependencies, range, and tolerances to the physical environment are addressed by these three fishes in different ways. The remarkable 2-year life span of the pink salmon exploits freshwater and marine habitats, while the herring and pollock are more long-lived, living and reproducing entirely as marine fishes. The pink salmon invests energy in rapid growth and a relatively small number of very large eggs that produce large, sessile larval fish (the alevin), rearing in the protection of the substrates of coastal streams. The alevin is generally a nonfeeding stage, subsisting during the cool coastal winters primarily on yolk-sac reserves for many months. In this way, late summer and fall spawning by pink salmon has apparently evolved to produce a large (30 mm), free-swimming, juvenile stage that enters shallow marine waters during the spring plankton bloom (mid-April) at a time when water temperatures are warming and zooplankton stocks are increasing (Cooney et al., 2001). In contrast, both the pollock and herring produce large numbers of small eggs – pelagic in the case of pollock and demersal for herring. Pollock spawn in the late winter and early spring in selected shelf and deep-water coastal areas (>200 m), while the herring deposit their numerous sticky eggs on underwater vegetation and other intertidal and shallow subtidal substrates at locations along the coast in early to mid-spring. Larval pollock and herring are tiny (<10 mm), weak swimmers; their distributions are determined primarily by vertical and horizontal currents. Part of the reproductive strategy for pollock and herring is to produce huge numbers of early life stages (eggs and larvae), betting that a few will find the conditions necessary for survival to the juvenile and later stages. It is generally believed that pollock and herring spawn at definite times and in specific locations to place their drifting egg and/or larval forms in currents that usually distribute the developing stages to favorable rearing areas.
94
Long-Term Ecological Change in the Northern Gulf of Alaska
The sequencing of the reproductive cycles of these three dominant fishes in the Gulf of Alaska produces juvenile pink salmon beginning in April, juvenile pollock beginning in May and June, and juvenile herring in July and August. In this way, the surviving 0-age juvenile stages target slightly different portions of the seasonal marine production cycle each year. Juvenile pink salmon couple trophically to the large calanoid-dominated portion of the early diatom bloom, whereas surviving juvenile pollock and herring occur after the large calanoids have left the upper layers, but when the smaller copepods and other zooplankters are reaching seasonal highs. Juvenile pinks grow at an astonishingly rapid rate, reaching 1.5–1.7 kg per individual in the final 14 months of their shortened 2-year life cycle. This massive growth is accomplished during an oceanic feeding migration of thousands of kilometers that begins in midsummer and concludes about the same time the following year. After spawning, the adults die, and the biomass of their carcasses enters a variety of coastal food webs. Pollock and herring are long-lived fishes (compared to pink salmon), and their juvenile stages last at least 2–3 years. Once recruited into the spawning population, adult pollock and herring may reproduce annually for 12 or more years. Adults of both species range freely in coastal and oceanic waters, feeding during the summer and early fall. Their overwintering strategies differ, however. Beginning as juveniles, herring invest in whole-body energy lipid storage, rather than in large size. This allows them to bridge the time of low zooplankton densities in the late fall, winter, and early spring by drawing on energy reserves to sustain them when feeding on zooplankton is all but suspended. Pollock, in contrast, do not store energy reserves to any great degree and must therefore continue to feed (on and off) during the winter. The adults of both species seek upper-layer zooplankton forage immediately after spawning, and much of their annual growth occurs at this time. In these contrasting ways, populations of herring, pollock, and pink salmon coexist in the coastal and oceanic waters of the Gulf of Alaska. There, they share reproductive space and critical forage stocks, but at different times and at different locations. These unique habitat dependencies provide a way to evaluate their production responses to different conditions of ocean climate described for and expected in the Gulf of Alaska.
2.5.7. Seabirds Morgan Benowitz-Fredericks, A.S. Kitaysky, and Alan M. Springer Introduction Seabirds are abundant predators in the Gulf of Alaska. In general, they can serve as useful monitors of the marine environment because, in addition to tracking change in the ocean, they are relatively easy to observe in their breeding colonies and at sea
Ecosystem Structure 95
(Montevecchi, 1993; Furness and Camphuysen, 1997). We have learned a lot about how survival, behavior, and reproduction change in response to subtle shifts in marine ecosystems. As a consequence of variations in ecology and life history traits, species respond uniquely to change, and understanding these responses can tell us much about ecosystems that we cannot easily observe in other ways (Piatt, 2002). In this chapter, we focus on foraging ecology, life-history, and reproductive strategies as aspects of seabird strategies for survival that are most likely to reflect environmental change.
Seabirds in the Breeding Season For most of the year, seabirds range widely; then, in the spring and summer breeding seasons, adults forage within range of their colony. In order to reproduce, birds must establish and maintain nest sites at least long enough to mate, and lay and incubate at least one egg – a minimum of several weeks. Whereas some species produce precocial chicks that leave the nest shortly after hatching, most species are tied to their colonies for several more weeks (murres) to months (puffins and fulmars) until their chicks are ready to fledge. This is a challenging time for seabirds. Energy demands are high; adults must secure enough food to maintain themselves, commute between the food source and the colony, and carry meals to a growing chick. Yet, foraging is limited by the need for parents to return to the colony. Parents may risk losing chicks to starvation or predation if they leave the nest unattended for long, so there is pressure to restrict the duration (and, therefore, distance) of their foraging trips. Furthermore, adults may be more vulnerable to predation when commuting to and from nest sites than when at sea (Ydenberg, 1989). Many species may be most vulnerable to environmental change during nesting, when foraging ranges are limited, and energy demands and predation risk are high. Focal Species Three of the most common, abundant, and well-studied seabird species in the Gulf of Alaska are common murres (Uria aalge), tufted puffins (Fratercula cirrhata), and black-legged kittiwakes (Rissa tridactyla) (referred to hereafter as murres, puffins, and kittiwakes, respectively). Their breeding colonies are comparatively accessible, and we know quite a lot about their biology in the breeding season. They are often found in the same nesting and foraging areas, and they share some basic life history characteristics. Yet, conspicuous differences among them may have consequences for population trajectories. For example, murres and puffins have similar overall ranges during the reproductive season. However, since the 1970s, puffins have been declining or even extirpated in much of their southern range (Piatt and Kitaysky, 2002), while murre populations are increasing in most of their historical range (Carter et al., 2001). These divergent population trajectories are likely related to differing strategies for survival and the ability of each species to cope with changing environments.
96
Long-Term Ecological Change in the Northern Gulf of Alaska
Phylogeny Evolutionary relationships among species must be understood to appreciate differences in survival strategies (Felsenstein, 1985). Closely related species share more characteristics due to common ancestry. Murres, puffins, and kittiwakes are all Charadriiformes, one of several seabird orders. All three species are monogamous, have biparental care of semi-precocial chicks, and have high nest-site fidelity (individuals nest in the same spot year after year). Murres and puffins are most closely related and most similar. They belong to two tribes of the Alcid family, a group of pursuit-diving seabirds (Strauch, 1985). The murre tribe (Alcini) includes dovekies and razorbills, as well as murres. All puffins are in the Fraterculini tribe, a group of birds with unusual bills that develop bright colors during the breeding season and are laterally compressed (much taller than they are wide). Both murres and puffins lay a single egg. Murres nest on exposed ledges and cliff tops, laying their eggs on bare ground (Ainley et al., 2002), while puffins nest in sheltered burrows either dug out of the earth or in natural rock crevices (Piatt and Kitaysky, 2002) (Fig. 2.40). Black-legged kittiwakes are in a separate family, Laridae, which includes gulls and is closely related to the tern family. They are one of the smallest gulls, known for their obligate cliff nesting and, among the gulls, their highly pelagic lifestyle. They build nests on cliff ledges, where they rear from one to three chicks during a successful reproductive season.
Foraging Ecology and Reproductive Strategies Why focus on foraging ecology and reproductive strategies? Although seabirds are affected by many factors year round, populations of many species are strongly regulated by food availability during the nesting season and the ability of adults to raise chicks (Ashmole, 1971). As we discuss later in this chapter, seabirds are faced with a tradeoff between successfully rearing their current chicks and surviving long enough to produce chicks in the future. Food, and how adults allocate food between their chicks and themselves, is a primary mediator of this tradeoff. Food resources in the marine environment are patchily distributed, ephemeral and unpredictable (see Section 2.4), and the abundance of fish and invertebrate prey varies widely with oceanographic conditions (Hunt and Schneider, 1987). Understanding seabird strategies for coping with variable food supplies is an integral aspect of their reproductive strategies, and critical to interpreting the expression of environmental change in their populations. That is, the ways seabirds obtain food and then allocate it between themselves and their offspring might directly influence population dynamics. This is not to say, however, that natality is the only factor important to seabird population dynamics. As we shall see, murres and puffins are much better able to cope with fluctuations in food supply during the breeding season than are
Ecosystem Structure 97
Figure 2.40: The three focal species of this chapter: common murres (bottom left, photograph courtesy of John Schoen), black-legged kittiwakes (top, photograph courtesy of John Schoen), and tufted puffins (bottom right, photograph courtesy of Arthur Kettle). kittiwakes, and factors affecting mortality at other times of the year are probably of greater importance to them. For example, large die-offs of murres in winter due to starvation have been observed in the Gulf of Alaska (Piatt and Van Pelt, 1993; Mendenhall, 1997). Evaluating population fluctuations caused primarily by changes in productivity during summer or by mortality during winter may help us discover the nature of underlying ecosystem processes important not only to seabirds, but to other organisms, food webs, and communities as well.
Foraging Ecology Murres, puffins, and kittiwakes are flourishing in the Gulf of Alaska, in large part due to their capacity to cope with fluctuations in food supply. However, they differ in their
98
Long-Term Ecological Change in the Northern Gulf of Alaska
abilities to accommodate such variability. For example, a sharp decline of surfaceschooling forage fish would affect kittiwakes, which are able to capture prey only near the surface, while murres and puffins can dive deeply and, thus, could be unaffected if alternative prey are available at depth (Kitaysky et al., 2000). On the other hand, murres may require much higher prey densities for efficient foraging (Piatt, 1990), and an absence of densely aggregated prey may reduce reproductive performance. Important components of foraging ecology, in addition to prey abundance, availability, and foraging depths, include diet diversity, search and prey capture strategies, foraging ranges, and habitat use. Many of these aspects of foraging ecology are determined by the morphology of the three species, and this is thus a good starting point in a comparison of the different strategies.
Morphology The size and shape of seabirds provide clues to their foraging ecology. Biological features often scale predictably to body size, and allometric analyses can predict metabolic rate and energy demands, lifespan, growth, and reproduction (Calder, 1984). Murres are the largest of these three species (Ainley et al., 2002), followed by puffins (Piatt and Kitaysky, 2002) and kittiwakes (Baird, 1994). Absolute energy demands and egg size increase with body size, while vulnerability to predators decreases (Birkhead and Harris, 1985). For some variables, such as foraging depth and wing loading (body mass per wing area), comparisons of body size are most meaningful for animals with similar body plans – that is, the relative shape and size of different body parts. For example, the difference in size between murres and puffins, which have similar body plans, is an important factor affecting wing loading, which scales allometrically for alcids (Livezey, 1988). Likewise, the maximum foraging depths of murre and puffins probably scale to body size (Burger, 1991) and wing loading. In contrast, kittiwakes have a different body shape altogether (discussed latter), and the muchshallower foraging depth of kittiwakes is constrained by their body shape rather than their smaller size. Finally, body size may have consequences for the minimum prey density required for effective foraging, with smaller species able to more efficiently exploit low-density fish schools than larger species can (Piatt, 1990) (Fig. 2.41). In general, body plans reflect foraging strategies (Pennycuick, 1987). Murres and puffins forage in three dimensions. They capture prey by pursuit diving – literally flying underwater after their prey. Their body plan is well adapted to pursuit diving, with very stout wings and comparatively streamlined bodies. Pursuit divers can forage to a greater depth than surface feeders and, therefore, have access to more prey. However, the costs of this body plan are: (1) energetically expensive aerial flight because of high wing loading for underwater flight, and (2) reduced load-carrying capacity (Pennycuick, 1987; Gaston and Jones, 1998). Thus, one of the consequences of this body plan is that prolonged aerial searches for prey are energetically inefficient.
Ecosystem Structure 99
Figure 2.41: Comparison of body and wing shapes in tufted puffins (top), common murres (middle), and black-legged kitiwakes (bottom) (photographs courtesy of Michael Shultz).
How murres and puffins locate food is not completely known, but they may remember and return to reliable foraging areas (Davoren et al., 2003a,b) such as oceanic fronts, boundaries between water masses that are often associated with upwelling, mixing, and high concentrations of fish (Hunt and Schneider, 1987). This would tend to reduce flight time otherwise needed to locate prey. And because of the high energy cost of flight, murres and puffins spend 70–85% of their time at sea on the water rather than in the air (Cairns et al., 1990), and paddle more than kittiwakes. The body plan of kittiwakes is more related to their overall foraging strategy than to the way they capture prey (Pennycuick, 1987). Kittiwakes, with relatively long wings and low wing loading, are highly maneuverable and efficient flyers, but unable to propel themselves underwater. Consequently they can search widely and efficiently for good feeding opportunities (Suryan et al., 2000; Jodice et al., 2002) and carry more food back to the nest relative to their size. However, since their foraging is practically constrained to two dimensions, they will be limited at particular sites by how long fish stay at the surface (Suryan et al., 2000). Kittiwakes plunge-dive and surface-seize their prey in the upper meter of the ocean (Baird, 1994). If a fish school descends after an initial attack, kittiwakes can no longer access it and often forage for just a few minutes before moving on. In certain locations, oceanographic features yield reliable foraging sites that kittiwakes learn about and return to for
100
Long-Term Ecological Change in the Northern Gulf of Alaska
weeks at a time (Irons, 1992). Regardless, murres and puffins have the potential to exploit a particular patch for much longer by diving.
Foraging Range and Habitat Use Murres Although flight is energetically costly for murres, their high wing loading (the highest of any flying bird; Livezey, 1988) allows them to fly very fast. As a result, breeding birds may range far from breeding colonies despite the cost of flight. Overall, murres appear to forage within 60 to 80 km of colonies, but this varies by location, year, and breeding stage, undoubtedly in response to changing prey (Ainley et al., 2002; Cairns et al., 1990; Piatt, 2002). In one study in Newfoundland, estimated foraging ranges of breeding birds during the chick-rearing period were mostly <20 km from the breeding colony (Cairns et al., 1990), whereas at Cape Thompson (Chukchi Sea), murres regularly travel up to 135 km (Hatch et al., 1999). Though this range is similar to that observed for kittiwakes at colonies where both murres and kittiwakes breed, murres are more often found foraging in deeper offshore waters (Speckman, 2002). This is likely because their extremely high wing loading and streamlined bodies allow them to dive very deeply, down to at least 180 m (Piatt and Nettleship, 1985). Murres tend to forage in areas where oceanographic features concentrate prey – fronts, upwellings, and thermoclines (see review in Ainley et al., 2002). Puffins Foraging ranges of breeding puffins have not been directly measured, in part because puffins implanted with satellite transmitters tend to abandon the colony (Hatch et al., 1999), but some conclusions have been drawn from at-sea distributions. Unlike some other Alcids, puffins show no preference for shallow water (Ostrand et al., 1998), presumably because their capacity to dive regularly up to 60 m and maximally to greater than 100 m gives them access to preferred prey at a range of depths (Piatt and Kitaysky, 2002). Puffins prefer to forage near their colony, but are often found foraging farther offshore than murres and kittiwakes, in shelf and shelf-edge habitats (Speckman, 2002). They are considered the most pelagic alcid (Piatt and Kitaysky, 2002). They forage in small groups in comparison to murres (Piatt and Kitaysky, 2002), presumably because their smaller body size allows them to forage on low-density fish schools that would not support large feeding flocks (Piatt, 1990; Piatt, et al., 1997). Kittiwakes During the breeding season, kittiwakes forage closer to the coastline than murres or puffins (Speckman, 2002), usually within 80 km of the colony (Kitaysky et al., 2000; Suryan et al., 2000; Daunt et al., 2002; Davoren et al., 2003a). They are often
Ecosystem Structure 101
found in mixed-species foraging flocks in conjunction with alcids and terns (Baird, 1994). Similar to murres, they seek out oceanographic features that concentrate prey.
Diet Murres Adult murres generally eat the most abundant prey available among a suite of pelagic forage fishes (Piatt, 2002). Although they can dive deeply, they often forage near the surface. Thus, where murre and kittiwake foraging areas overlap, their diets are often similar (Shultz, 2002; Van Pelt and Shultz, 2002). Murres consume a larger proportion of fish than the other two species (Hobson et al., 1994), namely, a few species of forage fish such as sand lance, capelin, herring, and cods (Ainley et al., 2002; Van Pelt and Shultz, 2002). Provisioning chicks may require adult murres to forage more selectively, because only a single item can be returned to a chick at each feeding (Gaston and Jones, 1998; Davoren and Montevecchi, 2003b). As a result, murres appear to select high-energy prey for chicks. In Cook Inlet, adult murres that are feeding primarily on sand lance near the surface will often bring higher-energy capelin or smelt to their chicks (Van Pelt and Shultz, 2002; Piatt, 2002). With low feeding rates and small meals, the ability to select high-quality prey items for chicks may be critical.
Puffins Diets of puffin adults and chicks can differ even more than those of murres. Adult diets are often dominated by invertebrates (Hobson et al., 1994), especially squid and euphausiids in the Gulf of Alaska, while chicks are fed a wide variety of fish species, including prowfish, salmonids, herring, pollock, capelin, sand lance, and sandfish, in addition to squid (Baird, 1991; Hatch and Sanger, 1992; Piatt et al., 1997; Piatt and Kitaysky, 2002). Even where adult puffins consume substantial amounts of fish (sand lance, pollock, capelin), they are more selective when foraging for chicks – chicks receive larger fish with higher energy contents compared to those consumed by adults (Baird, 1991). Unlike murres, puffins can carry multiple items in their bill, and have been known to bring over two dozen small fish in a single load (Whele, 1983) (Fig. 2.42). Puffins are opportunistic feeders (Piatt and Kitaysky, 2002), and appear to have the fewest constraints on foraging. They eat a large array of prey, can capture prey at great depths, and can carry multiple prey back to chicks. Compared to murres, their smaller body size may allow them to utilize lower prey densities and still forage efficiently (Piatt, 1990). Despite this apparent flexibility, they are sensitive to changes in food availability. In some areas, small shifts in ocean conditions can disrupt the distribution of preferred prey with devastating effects on puffin reproduction (Gjerdrum et al., 2003; Kitaysky and Golubova, 2000).
102
Long-Term Ecological Change in the Northern Gulf of Alaska
Figure 2.42: A kittiwake feeding its chick (left, photograph courtesy of Shiway Wang, USFWS) and a tufted puffin with a bill load of small forage fish for its chicks (right, photograph courtesy of Arthur Kettle, USFWS).
Kittiwakes Kittiwakes usually feed on the most abundant prey available to them in surface waters, often sand lance, herring, capelin, and cods (Baird, 1994; Shultz, 2002; Suryan et al., 2002). Other locally abundant prey is also consumed. For example, in areas of the Bering Sea, kittiwakes feed heavily on lanternfish, euphausiids, squid, and other zooplankton, which either migrate to the water surface at dusk or are brought to the surface by convergence at tidal fronts (Schneider et al., 1990; Springer et al., 1996). Kittiwakes will also feed on salmon eggs in coastal areas and offal from fish processing (Cramp and Simmons, 1983). Despite this flexibility, kittiwakes have preferred prey, perhaps chosen for their nutritional value or ease of capture (Suryan et al., 2002; Robards et al., 1999; Van Pelt et al., 1997). In Cook Inlet, kittiwake chick and adult diets are similar (Shultz, 2002). In contrast to murres, kittiwakes need not be too size selective when foraging for chicks, as they can capture multiple prey in a single foray and store it in their fore-stomach, or proventriculus. Meals are then regurgitated at the nest in response to chick begging (Fig. 2.42). The ability to carry multiple prey items can potentially help adults compensate for low local food availability. If foraging trip durations are prolonged because birds have to travel farther to locate prey, they can compensate by bringing back fewer, larger meals for chicks.
Response to Changes in Prey When preferred prey decline, predators can abandon reproduction to save energy, work harder to capture the preferred prey, or switch to a more abundant prey (Suryan et al., 2000). If they switch and the new prey type is smaller, has poorer
Ecosystem Structure 103
nutritional content, or is more difficult to capture, foraging efficiency declines. Most predators have an abrupt prey density threshold, below which foraging and reproduction are compromised and above which there is no relationship between food abundance and reproduction (Cairns, 1987; Piatt, 1990). These thresholds have been clearly demonstrated for murres and kittiwakes in Cook Inlet (Piatt, 2002). What happens below the threshold is key to understanding the vulnerabilities of seabirds to environmental change (see Section 4.8). With murres, when prey availability declines during the breeding season, adults may decrease loafing time – the time when both parents are at the nest site. Since only one parent needs to be at the nest with the chick, loafing time appears to be an energy buffer for reproduction. A decrease in loafing time indicates increased foraging, and may allow murres to maintain similar, albeit relatively low, chick feeding rates from year to year, despite substantial changes in prey (Zador and Piatt, 1999; Burger and Piatt, 1990). In some cases, however, there is no change in loafing time, and the consequence of a reduction in availability of preferred prey is that chicks fledge in poor condition (Davoren and Montevecchi, 2003a). Murres also switch prey in response to declines in preferred species (e.g., Bryant et al., 1999; Boekelheide et al., 1990) – for example, a climate regime shift in the Gulf of Alaska in the 1970s led to a decline in capelin, a preferred forage fish, and in some areas, murres responded by switching to pollock (Anderson and Piatt, 1999). Little is known about puffin responses to prey decreases, but their highly variable diet suggests that they would respond by prey switching. However, if high-energy fish are not readily available, puffins are the most likely of the three species to abandon their chicks (Vermeer et al., 1979; Gjerdrum et al., 2003), although, as noted below, prey abundance must be very low before they will desert. Kittiwakes will increase foraging range and switch prey in response to food shortages, depending on individuals, colony location, and year (Irons, 1998; Suryan et al., 2000; Suryan et al., 2002). Sometimes they fly farther to obtain a preferred prey, other times they switch prey, and often they do both (Irons, 1998; Suryan et al., 2000). Since plunge-diving uses so much energy, flying further to find denser or more abundant surface schools may be beneficial when food becomes scarce locally (Jodice et al., 2002). Similar to murres, kittiwakes that are facing low availability may spend less time at the nest during chick-rearing (Gill et al., 2002).
Life History Traits and Reproduction Basic life history In general, seabirds live long lives with delayed sexual maturation (commonly, 3–4 years) and long reproductive seasons that yield few young (Ricklefs, 1990). Murres, puffins, and kittiwakes share these and other fundamental traits but exhibit some critical differences as well. One of the most striking is the difference in clutch size.
104
Long-Term Ecological Change in the Northern Gulf of Alaska
Murres and puffins lay single eggs (though they are both capable of relaying if the first egg is lost), whereas kittiwakes will lay up to three eggs (Baird, 1994; Ainley et al., 2002; Piatt and Kitaysky, 2002). It is not clear whether this difference evolved as a result of constraints during chick rearing or egg production. In general, alcids lay eggs that are large for their body size (Gaston and Jones, 1998), potentially precluding production of multiple eggs in a clutch. Puffins are capable of rearing artificially enlarged broods, which suggests that the one-egg clutch is probably a function of limitations on egg production (Whele, 1983). Regardless, the major implication of these differences in maximum clutch size is that kittiwakes can better capitalize on good years and produce up to three times as many offspring as either murres or puffins. Higher intrinsic growth potential should allow kittiwake populations to grow most rapidly when conditions are good.
Tradeoffs The life history strategies of seabirds, as with all animals, represent a balance between adult survival and reproductive success in a given year (Stearns, 1992). Food availability and the extent to which adults are willing or able to buffer their chicks against fluctuations in food determine how each species achieves that balance. In general, if parents increase foraging to feed their chicks when food is scarce, they do so at their own expense. In other words, the more parents buffer their chicks, the more they compromise their own survival and the potential for successful reproduction in the future. In cases where parents do not buffer, the burden is on chicks to cope with the consequences. Which generation absorbs the costs of a bad food year will affect adult survival and reproductive success rates. Chick Strategies Chicks have evolved developmental and behavioral strategies that increase their chances of survival when parents do not provide enough food. Developmental strategies Chicks of all three species are considered “semi-precocial” at hatching: they are covered with down and have open eyes, but are still dependent on their parents for food and warmth (Ricklefs and Starck, 1998) (Fig. 2.43). They are all fed by both parents at the nest for at least a few weeks, and all are sensitive to the quality (energy and lipid content) as well as quantity of the food delivered to them (Romano, 1999). However, developmental similarities end there. Developmental responses to food shortages and the stage of development at fledging vary among species in ways that have consequences for chicks’ abilities to withstand food shortages.
Ecosystem Structure 105
Figure 2.43: Nesting: A pair of murres at their nest with egg (top left, photograph courtesy of Arthur Kettle, USFWS); a murre egg that is pointed on one end (middle left, photograph courtesy of John Schoen); a murre with its chick (middle right; photograph courtesy of Shiway Wang, USFWS); a kittiwake nest with parent and chick (top right, photograph courtesy of John Schoen); a pair of tufted puffins near the entrance to their nest burrow (bottom right); and an older chick temporarily removed from a burrow (bottom left) (photographs of puffins courtesy of Arthur Kettle).
106
Long-Term Ecological Change in the Northern Gulf of Alaska
Murres Murres fledge much younger than puffins or kittiwakes, after attaining about one-quarter of adult size. They do not need to fly, but only to lose their down feathers to avoid becoming water-logged and develop sufficient plumage to descend safely from the nest site down to the water. This can take as little as 16 days. Murres are able to fledge at such an early stage of development because they remain dependent on the male parent to provision them at sea for weeks to months as they learn to forage and grow to adult size (Ainley et al., 2002). It has been proposed that this strategy evolved in response to the low provisioning capabilities of murre parents (due to their energetically costly flight and limits on load carrying) in comparison with the extremely high energy demands imposed by a fully developed, adult-sized chick (Sealy, 1973; Birkhead, 1977). Thus, murre chicks are buffered against fluctuations in food availability by the short period of time and the extremely low developmental threshold required to leave the colony and access food more readily at sea (see the following subsection “Behavioral strategies”). Puffins Although puffins and murres are closely related, puffin chicks have quite a different developmental strategy that provides them with a remarkable buffer against food shortages. The nestling stage is relatively long, development is slow, and chicks do not fledge until they are fully grown and capable of flight (Gaston and Jones, 1998). Parental provisioning can be intermittent and quite low at times, and chicks may be forced to fast for several days at a time (Whele, 1983). Puffin chicks cope with this unpredictable provisioning by reducing metabolic rates by as much as 47% in response to food shortages (Kitaysky, 1999). Development is slowed or arrested, and energy is allocated to maintenance needs. If provisioning increases, chicks will elevate their metabolic rates and resume growth. As a result, fledging age is tightly coupled to food intake: chicks remain in the burrow until they are sufficiently developed, often substantially prolonging the nestling period. In an experimental study, chicks with 50% reduction in caloric intake fledged 20 days later than controls (Kitaysky, 1999). In addition, puffin chicks can store copious quantities of fat and often undergo weight loss prior to fledging (Piatt and Kitaysky, 2002). Thus, the “wait-it-out” strategy of puffin chicks makes them extremely well suited to survive fluctuating food availability and prolonged food shortages. Kittiwakes As with puffins, kittiwake chicks must attain full adult body size before they fledge, and they utilize the same developmental buffer against low food availability of prolonging the nestling stage, although much less so than puffins (Kitaysky, 1999). Kittiwake chicks can further buffer themselves by returning to the nest for several weeks to receive supplemental provisioning from parents after they have learned to fly and are attempting to forage (Baird, 1994). Puffin chicks do not have this safety net and must be completely independent once they leave the nest (Gaston and Jones, 1998).
Ecosystem Structure 107
Behavioral strategies Many animals exhibit behavioral responses to food shortages: increasing foraging effort, changing foraging areas, switching diets, or not reproducing. However the behavioral repertoire of seabird chicks in the nest is limited, because they are not mobile and rely completely on parents to provide food. The three principal behavioral options they have are begging for food, outcompeting a sibling, and switching nests. Begging Chicks can obtain more food by increasing the frequency, duration, and intensity of begging. Chicks of all three species beg, but with varied results. Kittiwake parents are extremely responsive to begging chicks, nearly doubling prey delivery in some cases (Kitaysky et al., 2001). Because of their ability to store prey in their proventriculus, adults can regurgitate several sequential meals on returning to the nest. Begging not only increases adult foraging, but can also induce multiple or larger regurgitations, so adults forgo food they could digest for themselves. Murre parents deliver a single prey item to their chick at the end of each foraging trip (Ainley et al., 2002). Elevated begging may induce a slight increase in the number of foraging trips (Kitaysky, unpublished data), but parental response to increased chick demands is limited. Puffins deliver single bill loads containing multiple fish (Piatt and Kitaysky, 2002). The entire load is deposited in front of the chick, and because the parent departs immediately, begging may be least effective for them. Sibling competition Eldest (“alpha”) black-legged kittiwakes from two- chick and three- chick broods have another option – to increase food intake by eliminating the younger, smaller beta chicks and thereby reducing competition for a limited resource. Food stress increases chick aggression (Irons, 1992; Kitaysky et al., 2003), and the older, larger, more aggressive chicks will monopolize food or kill their smaller siblings outright (Braun and Hunt, 1983). Beta chick mortality, via either starvation or ejection from the nest, is substantial in years of poor food (Roberts and Hatch, 1993; Gill et al., 2002; Suryan et al., 2002). Nest switching Murre and kittiwake chicks occasionally leave their hatch sites for those of neighboring adults if they are not receiving enough food or brooding from parents. When this occurs, foster parents may feed and care for adoptees, sometimes in conjunction with true parents (Birkhead and Nettleship, 1984). This behavior is limited by terrain and proximity of neighboring nests, and also by the willingness of neighboring adults to accept the wandering chick. In two studies, adoption occurred in 8% of kittiwake nests (Roberts and Hatch, 1994) and 7% of murre nests (Birkhead and Nettleship, 1984). It is not clear whether this form of adoption is a product of parents’ imperfect ability to recognize their own chicks, as proposed for kittiwakes (Roberts and Hatch, 1994), or, as proposed for murres, either misdirected parental urges or the caring for a related neighbor (Birkhead and Nettleship, 1984). In addition,
108
Long-Term Ecological Change in the Northern Gulf of Alaska
it is not clear whether this is a helpful strategy. In kittiwakes, adoptees are often the smaller chick in the new nest and suffer aggression from their foster siblings (Roberts and Hatch, 1994). Although there are no data on the existence of adoptive care in puffins, intense aggression among captive puffin chicks suggests that hungry chicks may move among burrows and compete with resident chicks for food (A. Kitaysky, unpublished data). In summary, in response to decreases in provisioning, puffin chicks have a passive “wait-it-out” strategy facilitated by the ability to decrease metabolism; kittiwake chicks have an aggressive “obtain-more-food” strategy either via begging, sibling competition, or seeking foster care; while murres have a low developmental threshold required for successful fledging and may also seek foster care. These strategies have clearly coevolved with parental strategies.
Parental Behavior and Consequences for Survival Life history tradeoffs Energy allocated to one function is not available for another (Stearns, 1992). Thus, a life history strategy for any animal revolves around how they prioritize energy allocation. While chicks devote all their efforts towards maximizing their own growth and survival, parents must balance each chick’s growth and survival with their own and that of their other current and future chicks. It has been proposed that long-lived animals breeding every year are unlikely to invest in reproduction, to the extent that they compromise survival in a poor food year (Williams, 1966). Our three species are all long lived, so they would not be expected to compromise their own survival in any one breeding season. However, each species resolves the energy tradeoff differently and, as we shall see, this expectation does not hold true in all cases. Murres Murre parents may compensate for fluctuations in prey by reallocating “loafing” time to foraging (Burger and Piatt, 1990). In some situations, this results in similar chick provisioning rates between colonies with very different local food availability (Zador and Piatt, 1999). However, in others, parents increase foraging trip durations without reducing loafing time, and chicks suffer lower growth rates and poor fledging condition (Davoren and Montevecchi, 2003a). Similarly, adults of the closely related thick-billed murre do not forage more when prey declines during chick rearing, and chicks suffer reduced growth rates (Kitaysky et al., 2000b). Even when murres do buffer, they have a very limited capacity to increase provisioning in response to chick demands – they do not increase provisioning rates as chicks grow (Harris and Wanless, 1995), and older chicks receive less than half as much food from parents (30–60 g/day; Ainley et al., 2002) as they are capable of consuming (130 g/day; Coulson and Pearson, 1985). Finally, a study in Cook Inlet found no relationship between reproductive success and survival in common murres (Piatt, 2004).
Ecosystem Structure 109
This suggests that murres are unwilling to compromise their own survival to rear chicks, and there is no evidence of negative effects of reproduction on adults. Puffins Nothing is known about survival costs for adult puffins incurred by reproduction (Piatt and Kitaysky, 2002). This is partially due to puffins’ extreme sensitivity to disturbance – they will alter their behavior and even abandon disturbed burrows, making direct observations difficult (Piatt and Kitaysky, 2002). However, parents do not forage more when prey declines, implying that they do not compromise their survival for reproduction. Chicks bear the brunt of food shortages, sometimes enduring less than 1 feeding per day for several days in a row (Whele, 1983; Kitaysky, 1996). Chick growth rates vary enormously (Piatt and Kitaysky, 2002), by 107% in one comparison of several colony-years in Alaska (Whele, 1983), also suggesting that parents do not buffer chicks against prey fluctuations. This strategy is not without its costs. Consistently low provisioning rates result in prolonged nestling stages, binding parents to chicks and therefore the colony for longer periods. At some level, the energy benefits of foraging less each day are likely outweighed by the costs of the extended nesting period. While there are no data available for tufted puffins, in other puffin species parents will eventually abandon chicks if the nestling stage is prolonged too much (Johnsen et al., 1994). Also, adult tufted puffins will abandon chicks in extremely poor food years (Vermeer et al., 1979; Gjerdrum et al., 2003). Thus, puffins may reproduce successfully in the face of short-term fluctuations in food availability, but if poor conditions are prolonged, adult puffins will not compromise their own survival (for Atlantic puffins, see Erikstad et al., 1997). Finally, the relative safety and protection of the burrow greatly affects the puffin reproductive strategy. A prolonged nestling stage would be unlikely without a safe place for the chicks. The burrow also frees puffin parents from having to attend chicks, allowing more time to forage, which may help alleviate the costs of being tied to the nest for such a long time. For murres and kittiwakes, the presence of at least one parent at all times is crucial to chick survival – the exposed nest sites would leave unattended chicks vulnerable to predators and inclement weather (Regehr and Montevecchi, 1997). Puffin parents may be able to leave chicks unattended for days at a time without predation consequences while they attend to their own needs. Kittiwakes Of the three species, kittiwakes are the best understood. It is clear that the tradeoffs in kittiwakes vary in complex ways and depend on factors such as age and fitness. Kittiwake parents will forage more when food is scarce, due largely to increased begging by chicks. The foraging range of kittiwake parents and their energy expenditure vary greatly with fluctuations in prey, while chick growth rates often vary little (Coulson and Thomas, 1985; Kitaysky et al., 2000b; Gill et al., 2002; Shultz, 2002; Suryan et al., 2002). However, parents cannot always compensate for poor
110
Long-Term Ecological Change in the Northern Gulf of Alaska
conditions, and reproductive success varies greatly from colony to colony and year to year (e.g., Murphy et al., 1991). When food is too scarce, chicks will die from sibling rivalry, starvation, or predation while left alone by parents on prolonged foraging trips (Roberts and Hatch, 1993; Regehr and Montevecchi, 1997; Gill et al., 2002). Although there appears to be a threshold beyond which kittiwake parents will not assume further costs and will no longer provision chicks, adults generally compromise their own survival for reproduction (Golet et al., 2004). When clutches are experimentally enlarged, parents forage harder for the larger brood and have poorer survival (Jacobsen and Erikstad, 1995). At a colony in the Gulf of Alaska, kittiwakes raising chicks to fledging had poorer survival than those that did not (Golet et al., 1998). Estimated life spans for kittiwakes that are capable of breeding but do not are twice as long as yearly breeders (Golet et al., 1998), suggesting that the highest adult survival occurs in years of reproductive failure (Coulson, 2002). The reason for the inverse relationship between adult survival and productivity seems to be that when kittiwakes do not nest, or suspend their nesting attempt part way through, they lower their energetic demands and reduce damaging physiological stress. One other factor enters the relationship between successful reproduction and survival – quality of individuals. Kittiwakes that breed successfully actually have higher survival than birds that never attempt to breed (Cam et al., 1998; Golet et al., 2004). This does not contradict the conclusion that kittiwakes compromise their survival when they breed, but reveals another dimension of survival – the role of individual quality. Individual quality results in the overall fitness of the phenotype, and it is becoming increasingly evident in life-history studies that the variation from individual to individual is crucial (Cam et al., 2002). Some kittiwakes are fitter and better able to absorb the dual costs of maintaining themselves and reproducing than other, lower-quality birds (nonbreeders), which may struggle to survive even if they are not reproducing (Cam et al., 1998; Golet et al., 2004). Evidence is emerging that one source of compromised individual fitness may be nutritional stress during development (Kitaysky et al., 2003). Responses of kittiwake populations to food shortages may depend on individual quality, and populations in declining and increasing colonies probably respond differently to food shortages. Declining colonies have presumably experienced strong selection, with only the highest-quality phenotypes remaining. In contrast, increasing colonies should contain a far wider range of phenotypes, due to recruitment of a mixture of highly fit and less fit young birds, or immigration of low-quality adults that have failed in their breeding attempts elsewhere and are searching for better breeding sites (Danchin et al., 1998). So, severe food shortages might result in more drastic declines in colonies with a recent history of increase. Such colonies probably contain a larger proportion of lower-quality phenotypes than colonies that have been declining for some time, where all remaining individual are of high quality.
Ecosystem Structure 111
Summary of tradeoffs: who bears the burden of compensating for a food shortage? Puffins and kittiwakes represent two extremes. Puffin adults do not buffer chicks, which must cope with fluctuations in prey. This implies that puffin adult survival should be independent of reproduction. In contrast, kittiwake parents assume much of the burden of compensating for poor food availability by persisting with reproductive efforts in the face of food shortages (Oro and Furness, 2002) and compromising their survival as a result (Golet et al., 1998). Murres may be intermediate between puffins and kittiwakes.
Other Factors Affecting Survival: Nesting and Predation Nesting habitat and predation pressure also affect individual and population responses to environmental change. The nesting site can affect access to food, vulnerability to predators, and the amount of antagonism from neighbors. Nest sites may be limiting in some colonies. Degradation of nesting habitat via physical alteration, improved predator access, or deterioration of local food supplies will affect seabird colonies (see Section for an example from Middleton Island). Murres and kittiwakes both nest on rocky cliff ledges. Murres will also use cliff tops or, rarely, vegetated areas when terrestrial predators are absent (Parrish and Paine, 1996). While kittiwakes make nests of vegetation and mud with deep cups and can use extremely narrow, steeply angled ledges (Baird, 1994), murres require broader and lessangled ledges as they lay their eggs directly on rock (Ainley et al., 2002). Murres will nest in much higher densities than either kittiwakes or puffins, often touching neighbors. Puffins excavate burrows in soil or nest in natural rock crevices, which limits their options. In addition, because they nest in burrows and therefore often atop islands or cliffs (rather than on cliff faces), they are protected from aerial predators but are more vulnerable to terrestrial mammalian predators, who can wipe out entire colonies in a single season (Piatt and Kitaysky, 2002). In areas with terrestrial predators, puffins nest only in rocky crevices. Burrows can flood, so nesting can also be limited by rainfall, and a shortage of dry, predator-free nesting habitat may regulate puffin populations in Prince William Sound (Piatt et al., 1997). Adult murres and kittiwakes vigorously defend their nests against intruding birds. For kittiwakes, site selection has clear fitness repercussions. Birds lose fewer eggs and chicks to predators when they nest on sheer cliffs, under overhangs, and toward the center of the colony (Regehr et al., 1998). In fact, nest sites can be at such a premium that failed breeders will remain at the colony to defend their nest site for the duration of the breeding season (Baird, 1994). Artificial nesting ledges, whether incidentally or intentionally created, are readily used by kittiwakes (Kildaw, 1998; Gill and Hatch, 2002). However, the degree to which suitable nesting habitat is a limiting factor for murres and kittiwakes in the Gulf of Alaska is not clear.
112
Long-Term Ecological Change in the Northern Gulf of Alaska
Finally, predation can affect all three species. Common predators include foxes, rats, bears, eagles, falcons, corvids, and gulls (the latter two primarily consume eggs and chicks; Baird, 1994; Regehr et al., 1998; Ainley et al., 2002; Piatt and Kitaysky, 2002). All three species are vulnerable to aerial predators when commuting to and from the colony. Murres and kittiwakes are still at risk from aerial predators at the colony, though they are usually safe from terrestrial predators. In contrast, puffins are safe from aerial predators in the burrow, but far more vulnerable to terrestrial predators. Most seabird colonies in the Gulf of Alaska lack terrestrial predators, but eagles and falcons can cause significant mortality by preying on adults and by flushing adults from the colony, which leaves eggs and chicks vulnerable to corvids (crows and ravens) and gulls (Parrish et al., 2001). For murres in particular, predation pressure may have played a role in the evolution of synchronous colonial nesting (birds produce chicks all at the same time), which provides mutual vigilance and defense and swamps predators (Wittenberger and Hunt, 1985; Hatchwell, 1991; Murphy, 1995; Murphy and Schauer, 1996). Ecosystem changes that affect predators’ access to other prey can increase predation pressure on seabirds (Regehr and Montevecchi, 1997) – for example, in years when salmon run later than usual, the frequency of eagle attacks on murres rises at a colony in Cook Inlet (M. Shultz, pers. comm.).
Summary The marine ecosystem is subject to natural cycles on many timescales, and seabirds, similar to all marine organisms, have developed ways to cope with this variability. Fluctuations in available prey during the nesting season and the effects food shortage can have on individual fitness are important factors limiting most populations. That murres, puffins, and kittiwakes are common in the Gulf of Alaska is conclusive evidence that despite their differences, all three species have evolved successful survival strategies. Murre parents pass along to their chicks much of the responsibility for coping with variable food supply. Murre chicks buffer themselves and, in turn, their parents against food stress by leaving the nest at an early age and undergoing most of their growth at sea while being attended by their fathers. This frees the parents from costly flight and the constraint of returning to fixed locations, the colonies, and allows them to range widely, by swimming, while tracking prey. It also affords chicks food subsidies as they develop the critical skills needed to forage successfully. Adult murres may contribute by increasing foraging frequency when prey is scarce, a cost they can bear during the short time when chicks are in the nest while they are still small and their daily food requirements are low. Adult murres further contribute by selecting prey with high nutritional value to feed chicks. Puffin chicks buffer themselves and their parents in a different way than murre chicks do. They develop completely in their burrows before fledging, but have a
Ecosystem Structure 113
highly flexible metabolism, which they lower, along with growth rate, when food is scarce, reducing the need for parents to increase their foraging effort and providing them more time to feed for themselves. Flight is not as energetically expensive for puffins as for murres, and adults seem to be able to accommodate longer development periods of chicks, provided they do not work much harder each day. Thus, puffin parents may maintain a fairly irregular feeding frequency, which the chicks adjust to, but the parents only abandon their chicks if the nestling period becomes extremely long or prey becomes extremely scarce. Kittiwakes have yet another strategy. Kittiwake chicks help accommodate fluctuations in prey also, but less so than either murre or puffin chicks. When food is scarce, dominant individuals in broods of 2–3 chicks will practice siblicide, thus increasing the proportion of food they receive each time an adult returns from foraging and reducing energetic demands on adults. Kittiwake chicks will also alter their rate of development, but far less so than puffin chicks. But when hungry, kittiwake chicks, in 1, 2, or 3 chick broods, beg more insistently, and parents respond by foraging more frequently, for longer intervals, or over greater areas to obtain sufficient food to satisfy the high demands of rapidly growing young. A much lower wing loading than murres or puffins helps compensate by reducing the energetic costs of flight and allowing kittiwakes to carry a considerable quantity of prey that can provide several meals or larger meals to chicks after each foraging bout. The trade-offs between chicks and adults in buffering the effects of variable prey are reflected in other life history attributes. In years when prey is relatively scarce, adult murres and puffins incur few energy costs above those associated with breeding in any year, since murres increase their foraging effort little and puffins not at all. Thus, adult survival is not greatly jeopardized by excessive stress during the nesting season in poor-food years, and they are very long lived. As we shall see in the following section, interannual variability in murre productivity (laying to fledging) is very low, and breeding failures seldom occur, indicating that factors affecting population change operate primarily at the level of juvenile survival before birds recruit to the breeding population or on adult mortality during winter. In contrast, adult kittiwakes suffer the effects of taxing physiological stress during breeding in any year, particularly so in good-food years when they work hard, raising one or more chicks, and their survival is significantly compromised by the effort. To compensate for higher adult mortality and occasional-to-frequent breeding failures, kittiwakes can produce as many as three chicks per year in good-food years. On average, adult survival in food-stressed colonies with low reproductive success is higher than in colonies with ample food, due to the costs of raising chicks (Frederickson et al., 2005). The fundamentally different life history characteristics of murres and puffins, compared to those of kittiwakes, should lead to contrasting patterns of long-term population change. Murres and puffins are longer lived than kittiwakes, but have
114
Long-Term Ecological Change in the Northern Gulf of Alaska
a lower intrinsic growth potential. Murre and puffin populations should be more sensitive to adult mortality and slower to recover following decline, because they produce a maximum of just one chick per year. Such attributes should dampen the amplitude of variability in abundance over time. In contrast, kittiwake populations experience a more rapid turnover of individuals due to greater adult mortality and shorter life spans and a higher intrinsic growth potential. Kittiwakes are able to recover more quickly following a decline, and should experience higher-amplitude fluctuations in abundance than either murres or puffins. In the following sections, we shall see if these three species in the Gulf of Alaska conform to our predictions.
2.5.8. Marine Mammals Sara J. Iverson, Alan M. Springer, and James Bodkin Introduction Much as the previous chapters have focused on several key species, so this chapter contrasts the strategies for survival of three very important marine mammals in the northern Gulf of Alaska: harbor seals (Phoca vitulina), sea lions (Eumetopias jubatus), and sea otters (Enhydra lutris) (Fig. 2.44). Fundamental life history characteristics: similarities, differences, and gradations Harbor seals, sea lions, and sea otters share several similar life history characteristics and are markedly dissimilar in others. These are summarized in the following to provide a framework for understanding differing strategies and pressures faced by both individuals and populations. Similarities: ●
●
● ●
annual adult survival rates: high (up to 95%), with longevities of 20–30+ years in harbor seals and sea lions and 12–20 years in sea otters sexual maturity: delayed – average age at first pregnancy in the Gulf of Alaska is 5.6 years (range 4–9 years) for harbor seals, 4.8 years (range 3–8 years) for sea lions, and 3.0 years (range 2–5 years) for sea otters reproductive output: low – a maximum of one offspring is produced per year diets: diverse and flexible
Differences and gradations: ●
locomotion: slow with webbed feet in sea otters, faster with rear flippers in harbor seals, fastest with fore flippers in sea lions
Ecosystem Structure 115
Figure 2.44: A harbor seal towards the southern end of its range (top, photograph courtesy of J. Spies, Georgetown University); a sea otter and her pup share an octopus (middle, photograph courtesy of R. Davis, Texas A&M University). A group of Steller sea lions (bottom, photograph courtesy of Robert A. Pawlowski, NOAA).
116
●
● ● ● ●
● ●
●
● ● ●
Long-Term Ecological Change in the Northern Gulf of Alaska
thermoregulation: blubber in harbor seals (more) and sea lions (less) versus fur in sea otters stored energy: little in sea otters to abundant in harbor seals mating and reproduction: fully terrestrial in sea lions to fully aquatic in sea otters reproductive synchrony: low in sea otters to high in harbor seals and sea lions lactation strategy: primarily fasting with some foraging in harbor seals, foraging cycle in sea lions, continual feeding in sea otters neonatal dependency: shortest in harbor seals to longest in sea lions diet: benthic invertebrates in sea otters versus primarily fishes in harbor seals and sea lions diving: relatively shallow in sea otters to deep (usually >50 m) in harbor seals and sea lions movements: sedentary in sea otters to highly mobile in sea lions home range: small in sea otters to large in sea lions predator defense: in all three species – vigilance, relatively highly developed swimming skills by weaning, and the use of escape habitat including shallow water, lagoons, and complex reef structures; in the case of sea otters – residing in kelp beds; in both pinnipeds – residing in groups and predator avoidance by hauling out; in harbor seals (as in other phocids) – swimming at depth, thus minimizing vulnerable surface time; in sea lions – size, speed, and agility in defending against and alluding predators
Breeding patterns, parental investment, and energetic consequences Reproduction, a fundamental component of life histories and population dynamics, is energetically expensive in all mammals, and these costs are generally born most heavily by females. In marine mammals, males do not contribute to parental care and, thus, male reproductive costs are limited to those associated with resource defense and mate acquisition, both of which can be considerable. However, since all energetic costs of gestation and rearing of offspring until weaning are the sole responsibilities of the mothers, we shall focus our discussion on females. A summary of reproductive strategies for our three species of marine mammals is provided in Fig. 2.45. While gestation is certainly expensive in most mammals, the lactation period accounts for by far the greatest reproductive expenditure of energy. Females must balance the costs of both these events with those of other vital life processes and the resulting consequences for her own survival, long-term fitness, and continuing her lineage, i.e., the number of healthy, viable offspring she produces in her lifetime. For instance, in some species, females that reproduce for the first time at a younger age have lower survival rates than those that first breed at an older age, and females who raise offspring in one year may be less likely to raise offspring in the following year, especially if food resources are limited.
Figure 2.45: Top: schematic of generalized maternal lactation strategies in harbor seals, sea lions, and sea otters. Arrows above each bar indicate the approximate timing of birth and weaning. Below the bar, straight lines indicate the time mother spends on shore, and jagged line represents relative timing and frequency of maternal diving and foraging during lactation. Photographs courtesy of author (harbor seal), D. Withrow, NOAA (sea lion), and Randy Davis, Texas A&M University (sea otter). Bottom: comparison among species of estimated total maternal energy expenditures during lactation, including milk output, foraging and overhead costs, and comparison of estimated energy invested in total milk output over lactation as a function of body size (see text for explanation of estimates). Photographs courtesy of author (harbor seal), D. Withrow, NOAA (sea lion), and USGS (sea otter).
118
Long-Term Ecological Change in the Northern Gulf of Alaska
Maternal constraints and strategies Female harbor seals, sea lions, and sea otters have strategies to help meet and balance the requirements and costs of reproduction. One strategy is delayed implantation and some flexibility in the timing of implantation. For harbor seals and sea lions, this dissociation of the time from mating to the start of fetal growth allows parturition and subsequent mating to occur close together. This shortens the time that females and males must aggregate at haulouts, which are free from predators but where food resources in the adjacent ocean may not be optimal. The timing of implantation is generally synchronous and is likely most influenced by photoperiod, although maternal nutritional condition can also play a role (Boyd et al., 1999). In harbor seals, implantation can occur from 1.5 to 3 months after fertilization (Hoover, 1988a). Presumably, this gives a female some time to “decide” early in the cycle if she can proceed with a pregnancy, based on her physiological state. A related strategy is a variable rate of fetal development, depending on the availability of resources. Delaying parturition in the face of low food availability in order to extend fetal development has been suggested in harbor seals and other pinnipeds (Lunn and Boyd, 1993; Boyd, 1996; Bowen et al., 2003). The rate at which a female can divert energy from her own maintenance to her developing fetus during pregnancy will be influenced in large measure by the rate at which she can acquire energy. The result of such flexibility is, presumably, reduced physiological strain on the mothers. When foraging conditions are so poor that a female cannot support either the implantation of the embryo or its continued development, she may abort. This strategy is particularly prevalent in sea lions, for which the costs of a pregnancy could be detrimental to the female, the nursing pup, or both during the extended period of care, especially in view of their generally expensive energetic strategy (Pitcher et al., 1998), as discussed in the following. Although delayed implantation also occurs in sea otters, it does not appear to be linked to day length, and conception and birth can happen at any time of year if maternal and environmental conditions are favorable. Synchronous and seasonal breeding is likely not necessary in sea otters, given the small, overlapping home ranges of males and females, local and seasonal food resources (benthic invertebrates), and aquatic reproduction and nursing. A gestation strategy used by sea otters is known as “bet-hedging” (Monson et al., 2000). This strategy minimizes maternal investment in conception and fetal development through parturition in the face of immediate resource shortages, resulting in offspring with relatively small birth weights. The female can then potentially benefit from the ability to exploit episodic resources if they become available postpartum, without overly sacrificing her own fitness. If conditions subsequently improve, she can provision her pup well and compensate for the prepartum energy deficit. If not, pup survival will be reduced, but she will maintain greater fitness with a higher likelihood of future reproductive success.
Ecosystem Structure 119
A more important contrast between the pinnipeds and sea otter relates to the pattern and rate of energy transfer from mother to offspring in the form of milk and the length of neonatal dependency. Harbor seals, in particular, are able to rapidly transfer energy to their offspring because of the large amounts of energy stored as blubber prior to parturition. To a lesser extent, sea lions store energy in blubber to transfer to offspring, but they are also obliged to forage during most of a long 8-to24-month lactation period. In contrast, the female sea otter stores little energy and must offset the high costs of milk output entirely by increased food consumption throughout the 6-month lactation period. The differing strategies used by harbor seals, sea lions, and sea otters during lactation result in vastly different energy costs for the mother and equally divergent growth patterns and survival pressures in offspring. The breeding pattern of phocid seals has been described as energetically economical (Costa, 1993). Most or all of the energy needed by females to raise pups is acquired and stored as capital in the form of blubber prior to parturition, thus eliminating or greatly reducing the need for maternal foraging and reducing the overall length of the lactation period. The large-bodied phocid seals fast throughout brief lactation periods, during which mothers secrete milk with high fat content that results in rapid rates of pup growth. Although harbor seals are among the smaller phocids, females acquire a relatively thick blubber layer (~4 cm), which fuels the majority of energy costs associated with milk production. They secrete milk that is about 50% fat and 9% protein (Lang et al., 2005) and are able to support most (80%) of a relatively brief lactation period (~24 days) through blubber fat reserves (Bowen et al., 2001). Thus, the general lactation strategy of the harbor seal is similar to other phocids, although the smaller females must supplement stored energy with feeding during the lactation period (Bowen et al., 2001). Harbor seal mothers may forage regularly and increasingly after the first third of lactation; however, they typically remain with or relatively near their pup throughout most of the nursing period. Most foraging bouts of females last <8 h (Boness et al., 1994; Thompson et al., 1994; Bowen et al., 2001) because, unlike otariids (but like most other mammals), female harbor seals must regularly empty the mammary gland for sustained milk production (Lang et al., 2005). During the lactation period, females lose over 35% of their body mass and 70% of their body energy stores as blubber fat is rapidly mobilized to produce large quantities of high-fat milk. Females, which average about 85 kg at parturition, expend over 1300 MJ during the lactation period (estimated from Bowen et al., 2001), a large proportion of which is supplied by the energy stored in blubber, and the rest from supplemental food consumption. Of this total energy used, over half (743 MJ or 8.7 MJ/kg) is spent on the direct transfer of milk energy to her pup (Fig. 2.45). The rest, approximately 44%, is spent mostly on maternal “overhead,” or the amount of energy needed to fuel maternal expenditures while suckling on land, as well as costs of foraging during the lactation period.
120
Long-Term Ecological Change in the Northern Gulf of Alaska
This strategy contrasts in several important ways from the comparatively expensive strategy of sea lions and other otariids. Although female sea lions store energy in blubber, blubber represents a much smaller percentage of total mass than in harbor seals. A female remains with her pup continuously for only the first 7–10 days after birth, and thereafter must resume foraging. Even though sea lions are extremely large bodied, they are likely constrained by their early ancestral phylogeny to the otariid strategy of storing relatively less blubber, producing milk of lower fat content, and thus alternating fasting and foraging throughout a prolonged lactation period. This strategy may have evolved in early otariids in general through a combination of factors such as generally smaller body sizes and differences in types and proximity of prey resources (Schultz and Bowen, 2005). As a consequence of this strategy, female sea lions produce milk that contains about 24% fat and 9–10% protein and nurse their pups for at least 8–10 months, but up to 3 years, before they are weaned (Pitcher and Calkins, 1981). A female sea lion, averaging about 264 kg, is estimated to spend a total of about 50,240 MJ during the lactation period (estimated from Winship et al., 2002 and assuming a 12-month lactation period). Of this total energy used, about 40% (20,250 MJ or 76.7 MJ/kg) is used for direct transfer of milk energy and the rest, approximately 60%, is spent on maternal overhead while on land and foraging at sea (Fig. 2.45). Thus, a female sea lion expends a great deal more energy during lactation than does a female harbor seal in proportion to her size. Given the difficulty in comparing such different lactation periods and how they are coupled spatially and temporally with energy-acquisition periods, it is most straightforward to simply compare the total energy that must be transferred to successfully wean a pup. In order to wean a viable pup, a female sea lion spends about 27 times more energy, or 9 times more on a body size basis, than does a harbor seal female on total milk energy output. These costs are surely an underestimate, as the female sea lion will also incur overhead costs every time she comes back to shore to nurse for almost a year, whereas the harbor seal female incurs such costs for only 24 days. In contrast to either pinniped, the sea otter mother stores little, if any, of the energy needed for milk secretion. Instead, the female must resume an intensive foraging schedule within a day or two of giving birth. Females with small pups tend to spend less time foraging than other females, while females with large pups spend the most time foraging (up to ~12 h per day) (Riedman and Estes, 1990; Gelatt et al., 2002). Sea otter milk has a fat and protein content of about 23% and 10–11%, respectively (Jenness et al., 1981), similar to that of otariids, but has a higher fat content than their terrestrial mustelid cousins (8–10%). As with sea lions, sea otters exhibit a prolonged period of neonatal dependency, typically lasting about 6 months. Although this period is not as long as that of sea lions, the costs of rearing young in sea otters exceed those of either sea lions or harbor seals. One characteristic that also separates sea otter females from the pinnipeds is that the costs of rearing young do not reside solely in
Ecosystem Structure 121
the transfer of milk energy, but also include other forms of extensive maternal care. For instance, a sea otter mother may spend up to several hours per day grooming her pup in order to maintain the insulative properties of its fur until the pup can do so effectively on its own (about 2.5 months of age). Additionally, the mother carries her young on her chest for extended periods up to about 6 weeks of age and captures almost all solid food for the pup, as well as herself, until the pup is about 5 months old (Payne and Jameson, 1984; Riedman and Estes, 1990). The energetics of lactation in sea otters are not well documented. The only estimates suggest that nonlactating adults consume about 25% of their body mass per day, while lactating females consume about 50% of their mass per day (Tuomi and Williams, 1995). Since these values encompass all maternal expenses, they provide the best proxy for comparison with pinnipeds. If we assume that a typical female sea otter weighs about 20 kg and the prey averages about 3.35 MJ/kg (calculated from values for clams, mussels, and abalone), then a lactating female would consume about 33.5 MJ/day. As a minimum estimate, assuming that the female expends this full amount of energy in both milk output and other maternal care for ~5 months (20 week), her total expenditures are about 4686 MJ during the lactation period. Of this total energy used, at least 2343 MJ (or 117.2 MJ/kg) would be used for direct transfer of milk energy and other costs associated with maternal care (Fig. 2.45). This suggests that the costs (above overhead, on a body size basis) of the sea otter female rearing a viable pup are at least 1.5-fold those of the sea lion female and 13.5-fold those of the harbor seal female. Again, these likely represent minimum estimates, as most females probably lactate and continue to provide food to pups until they are weaned at 6 months of age (Riedman and Estes, 1990).
Offspring Constraints and Strategies Harbor seal pups are unusual among pinnipeds and other mammals in being born with substantial fat stored in blubber. This insulative layer, along with the lack of lanugo (the temporary white birth coat of most phocids responsible for early terrestrial thermoregulation), makes them capable of swimming and diving shortly after birth and certainly provides some immediate short-term energy reserves. Pups may also occasionally accompany their mothers on foraging trips from an early stage (e.g., Bowen et al., 1999). Such precocial behavior, coupled with a short maternal dependency, demands that pups be well developed at birth and that females rapidly transfer energy prior to weaning. At birth, harbor seal pups weigh about 11 kg (about 13% of maternal mass) and contain 11–14% fat stored in blubber (Bowen et al., 1992; Hammill, 2003). During the 24-day suckling period, the pups grow rapidly, doubling or tripling their weight. In the most comprehensively studied population (Sable Island, Nova Scotia, Canada), harbor seals are weaned at about 24 kg and contain 38 to 40% fat (Muelbert et al., 2003). However, data from a 4-year period
122
Long-Term Ecological Change in the Northern Gulf of Alaska
(1997–2000) in Prince William Sound suggests that harbor seals there are weaned both heavier (29 kg) and fatter (42% fat; Iverson et al., 2003). Although harbor seal pups begin to feed within days of weaning, they lose body mass for at least 4–6 weeks before they are able to attain positive energy balance (Muelbert et al., 2003). Hence the ample blubber reserve at weaning insulates the pup and provides an energy resource during the days and weeks it takes to learn to forage independently and efficiently after it is abruptly weaned. Thus, harbor seal pups have several strategies that likely enhance survival. They are buffered at birth by sizable fat stores deposited in utero and are also buffered during lactation if their mother can efficiently transfer milk energy rapidly. They are weaned abruptly with little foraging experience and only their blubber energy stores to rely on. However, unlike most other phocids, harbor seal pups have already acquired substantial swimming skills and likely also some foraging skills by weaning and, thus, may have an advantage over other newly weaned phocid pups. Nevertheless, acquiring these skills early perhaps places them at greater risk of being eaten by aquatic predators such as sharks (e.g., Bowen et al., 1999, 2003) or killer whales. At birth, sea lion pups weigh about 20 kg (about 8% of maternal mass) with little fat stored in blubber. This birth mass is greater than that of harbor seals on an absolute basis, but relative to maternal size is about 40% less than that of harbor seals. Many are weaned 8–12 months later at about 95 kg, a proportional weight gain twice that of harbor seals but over a time period of about 14 times longer. These values are all increased in sea lions that do not wean for 2 years or longer, as has recently been reported in sea lions from Southeast Alaska (K. Pitcher, pers. comm.). Sea lion pups grow much more slowly than harbor seal pups, in part because milk energy content and milk intake are lower and in part because nursing bouts are less frequent and interrupted, owing to the need for mothers to forage for extended periods during lactation. Every time the mother leaves to forage, the sea lion pup fasts, losing a substantial portion of the energy it just gained from nursing. This yo-yo pattern of energy intake and loss results in extremely slow growth rates, which are comprised primarily of lean tissue growth rather than just fattening as in the case of harbor seal and other phocid pups. Nevertheless, sea lion pups may contain as much as 27% fat at 10 months of age (Rea et al., 2003), providing them some energy resources during their gradual weaning process. Although sea lion pups grow slowly, they have the luxury of a prolonged (~1–3 years) and dependable food source (mother), while they learn the difficult lessons of how to feed themselves. By weaning, they have developed substantial bone and muscle mass and are highly coordinated, compared to most newly weaned phocid seals. Sea otter pups are born at about 2 kg, or about 8% of their mother’s mass (Kenyon, 1969). At weaning, a sea otter pup weighs about 12 kg (Riedman and Estes, 1990), nearly half the mass of an average adult female and representing a sixfold increase from birth weight. This compares to about a threefold gain in the harbor seal pup and
Ecosystem Structure 123
a fourfold gain in the sea lion pup by weaning. This difference in relative growth probably reflects contributions from prolonged maternal investment and the undeveloped stage of the sea otter pup at birth compared to the relatively precocial state of pinniped offspring at birth. In contrast to the sea lion, and even the harbor seal, a sea otter pup is in constant association with its mother until weaning, nursing up to six times per day (Riedman and Estes, 1990). While a harbor seal pup is a proficient diver the day after birth, a sea otter pup may not become a proficient diver until 3 to 4 months of age. There is good evidence that the dietary specialization for specific prey species exhibited by some adult female sea otters is taught to their offspring (Estes et al., 2002). Thus, sea otter pups benefit from prolonged and rather intense association with their mother. During this time, they not only are provided nutrients in the form of milk and captured prey, but they also learn to groom themselves, dive, and obtain food independently.
Trade-offs and Consequences: Energy and Population Considerations Patterns of energy acquisition and expenditure differ considerably among harbor seals, sea lions, and sea otters. Harbor seals demonstrate the greatest capacity to store energy, followed by sea lions, with sea otters having the least capacity. Female pinnipeds, particularly harbor seals, take more advantage of resources in the form of stored energy, but are tied to a strict reproductive calendar, while sea otters demonstrate temporal reproductive flexibility through the ability to conceive and pup at any time of the year, but lack significant energy storage capacity. These differences in energy acquisition are at least partially reflected in the relative rates at which maternal energy is transferred to offspring. Both sea lions and sea otters expend large amounts of energy compared to harbor seals overall, and this expenditure must necessarily be balanced by high rates of acquisition. At sea, metabolic rates of sea lions are about 6 times basal rates, while those of harbor seals are likely in the range of 1.5–3 times basal rates (Costa and Williams, 1999). Sea otter metabolic rates are even higher than those of pinnipeds, as they remain continually active and feed frequently to offset thermoregulatory costs. The use of heat generated from the digestion of food (heat increment of feeding) is in fact a major strategy used by sea otters for thermoregulation (Costa and Kooyman, 1984). One advantage of the economical reproductive strategy of harbor seals is that they are buffered more against fluctuations in prey availability during the lactation period than are sea lions or sea otters. In harbor seals, which store a substantial portion of the energy necessary for lactation as capital in blubber, reproductive success (raising a pup to weaning) is more a function of prey availability during the months prior to lactation. Maternal fitness also depends on the female’s ability to refuel following the severe body nutrient depletion that occurs during the intense-lactation period. However, at neither of these times is she restricted in time or space in terms of where
124
Long-Term Ecological Change in the Northern Gulf of Alaska
she can forage, as she has no offspring tied to her. Harbor seal females, as with most other phocids, appear physically spent at the end of lactation, having lost >35% of their body mass. However, there is good evidence that phocid females terminate lactation before actually compromising their physiological status and ability to recover (Mellish and Iverson, 2001). In contrast, sea lions and sea otters are income breeders and are thus almost completely dependent on food resources acquired during lactation. Sea lion and sea otter mothers must optimize the amount of time spent foraging versus nursing. If prey resources decline, foraging bouts must become longer or more intense. The trade-off of increased foraging times is decreased delivery rates of milk to pups, which means more of the energy delivered to pups is used for pup maintenance and less for its growth and energy storage. In the case of sea otters, there is good evidence that females in habitats where food and space resources are not limiting are in better condition and that their pups experience greater survival than in habitats where food is limiting (Monson et al., 2000). Thus, in sea otters, and to a large extent sea lions, reproductive success primarily reflects environmental conditions (e.g., energy or prey availability) over an extended period after birth, as opposed to environmental conditions over longer time periods during pregnancy for harbor seals. In addition, mothers of both sea lions and sea otters are constrained in their foraging by being tied to a pup – in the case of the sea otter, this may mean simply ceasing foraging bouts to feed her pup, but in the sea lion, this means returning to land to suckle. To help minimize these constraints, the sea otter mother takes her pup foraging and nurses it aquatically, while the sea lion mother can move her pup to a new terrestrial location in closer proximity to better resources. Nevertheless, both the sea otter and sea lion mothers are still constrained by the prolonged dependence of their offspring. Although little is known about the body composition and condition of female sea lions and sea otters at the time of weaning, given the gradual weaning process in both species, it is likely that successful females have been able to maintain or recover body condition to a reasonable degree. Life history strategies evolved by adult males and females of all three species undoubtedly provide substantial buffering against fluctuations in the environment and, as in most marine mammals, adult survival is generally high. In contrast, it is likely that juveniles are much more readily compromised in the face of variations in food supply, predation, and other environmental conditions over many scales. In general, the transition to nutritional independence is thought to be a critical period in the lives of most animals, as it usually represents the end of the period of parental care (e.g., Clutton-Brock, 1991) and a period of increased vulnerability. Little is known directly about this period and that of the first few years in the lives of juvenile marine mammals. However, juveniles are the age class most susceptible to changes in prey availability and are considered the group experiencing the highest mortality in these populations. Thus, in harbor seal, sea lion, and sea otter populations, juvenile
Ecosystem Structure 125
survival (principally dictated by starvation and predation pressures) is likely to be more sensitive, and hence more variable, than adult survival. However, it is important to note that changes in adult female survival will have even greater effects on demography. Nevertheless, juvenile survival can be enhanced in several ways. Although offspring of harbor seals, sea lions and sea otters have fundamentally different weaning strategies, body size at weaning in all three species is positively correlated with survival (Baker and Fowler, 1992; Le Boeuf et al., 1994; Hall et al., 2001), as it is in other mammals (Stearns, 1983). Larger body size confers greater energy storage capacity, greater oxygen storage, and, thus, dive capacity, and perhaps greater overall nutritional status. Better swimming capability and experience can also confer a greater ability to avoid predators. The offspring of harbor seals are forced to make the transition to nutritional independence abruptly and without the assistance of parents, and thus body size at weaning likely plays an important role in their survival. Harbor seals have an “insurance policy” in the form of substantial blubber reserves, especially if weaned at a large size. They partially are also adapted to fasting for prolonged periods, during which they partially spare critical body protein and use their fat to fuel most of their metabolism while learning to forage and during fluctuations in prey availability. In the case of sea lions, pups have already begun to learn how to forage during their gradual weaning process and can also carry some reserves of fat in blubber to buffer prey fluctuations. Finally, in the case of sea otters, offspring have been taught foraging strategies directly by their mothers and, if weaned at a viable body mass and in an area of moderate to abundant food resources, they can make the transition to complete nutritional independence fairly readily.
Diet and foraging In order to meet the energetic demands of existence, whether comparatively large or small, all animals have evolved feeding strategies that include the kinds of things they eat and how they go about finding it. As with basic life history characteristics described earlier, harbor seals, sea lions, and sea otters exhibit similarities, differences, and gradations in diets, foraging behaviors, and movements associated with foraging and prey distributions. Diet Prey typically consumed by harbor seals include a diverse array of pelagic, demersal, and benthic fishes and some invertebrates (Pitcher, 1980, 1981). In the Gulf of Alaska, harbor seals are known to eat at least 27 species of fishes from 13 families, in addition to octopus, squids, and shrimp. Prey of young harbor seals (<1 year old) may consist of the same species as those of adults, with individual prey being generally smaller. However, evidence from several phocids in the North Atlantic suggests that juveniles feed on different prey than adults (Bowen and Harrison, 1996).
126
Long-Term Ecological Change in the Northern Gulf of Alaska
For instance, newly weaned harbor seal pups apparently consume proportionally more small crustaceans, especially shrimp and mysids, than adults (Bigg, 1973; Muelbert and Bowen, 1993). Such prey may be relatively easy to capture and provide valuable nutritional benefits as pups, buffered by substantial energy reserves in blubber, develop the physical and behavioral capabilities to forage on more energyrich and cost-effective species. The suite of prey taken by sea lions overlaps greatly with that of harbor seals, and includes over 50 species of fishes and invertebrates (Pitcher, 1981; Calkins and Goodwin, 1988; Sinclair and Zeppelin, 2002). Sea lions tend to take larger individuals of common prey than harbor seals. Juvenile sea lions, similar to juvenile harbor seals, prey on similar species as adults, but also tend to take smaller individuals (Merrick and Calkins, 1996). Sea otter diets also are highly diverse – they are known to prey on more than 150 species (Estes and Bodkin, 2002). Diets are dominated by sessile or slowmoving, benthic invertebrates (e.g., clams) commonly found in a variety of habitats, including soft sediments, rocky reefs, and kelp canopies (Estes and Bodkin, 2002). Thus, sea otter diets are fundamentally different from those of the primarily piscivorous harbor seals and sea lions. Fish are sometimes important prey in the Aleutian Archipelago, where urchins are the dominant prey, but are essentially unknown in the diets outside the Aleutians. Fishes are proportionally more important in islands where sea otter densities are nearly in equilibrium with prey resources. Harbor seals, sea lions, and sea otters are all very flexible in their choice of prey, and only a few species, typically one to three, dominate diets at any given time and location. Prey selection presumably depends on the abundance of forage species, which varies between geographic regions, habitat types, seasons, and years, and it may also be influenced by profitability, e.g., search time, handling time, and energetic return (Bowen et al., 2002). In the case of sea otters at least, individual dietary specialization is also a factor (see Estes et al., 2003). Dietary flexibility and individual specialization make it somewhat difficult, therefore, to identify which prey are most important to any of them except at particular places at particular times. The prey of harbor seals and sea lions include species with varying amounts of fat. Fat content, and thus energy density, of prey has been equated with nutritional quality, since fewer energy-dense prey are needed to meet the energy needs of predators, and the return on foraging investment can be higher if such prey are abundant. However, nutritional quality is more complex than simple fat content, as protein and other ingredients in prey also are critical to the health of consumers, and, as long as predators can consume a mixture of prey types, it is likely that nutritional needs can be met. For instance, in Prince William Sound, juvenile harbor seals with more diverse diets were in better condition, independent of prey species eaten, provided that some high-quality prey were obtained to balance low-energy prey consumed (Iverson et al., 2003). Additionally, the behavior and digestive physiology of harbor seals are
Ecosystem Structure 127
adapted to compensate for diets of varying nutritional quality. When faced with low-quality diets, harbor seals eat more and enhance the assimilation of both protein and lipid, resulting in a digestible energy intake as much as 25% higher than that of high-quality diets (Trumble et al., 2003). Sea lions have made similar adaptations. Moreover, sea lions have evolved proportionally longer digestive tracts – gut length and increased surface area of the digestive tract increases retention time of food, allowing animals greater time to extract nutrients from poorer-quality food. It is also likely that individuals consuming low-quality food respond in real time by increasing gut length, as this is a somewhat plastic trait in animals. In contrast to both pinnipeds, sea otters tend to have diets consistently lower in fat and higher in protein, a likely consequence and adaptation of their evolution in the mustelid family. High-protein diets generate greater heat increments of feeding and thus should be important to the entire thermoregulatory processes of sea otters. Foraging and home range size Among pinnipeds, harbor seals are considered to be relatively sedentary and typically forage over small areas (Pitcher and McAllister, 1981; Iverson et al., 1997; Lowry et al., 2001). In Prince William Sound the distance from haul outs to at-sea locations (their foraging distance) is greater in fall–winter than in spring–summer and is greater for juveniles than adults. Foraging distances of juveniles average 10–25 km in winter and 5–12 km in summer, whereas adults travel straight-line distances averaging <11 km in winter and about 6 km in summer: about 90% of all locations at sea for both adults and juveniles are 0–25 km from haul outs. Home ranges of female harbor seals, both adults and juveniles, are considerably greater than those of males in winter in Prince William Sound, but are similar in spring–summer: female home ranges average about 600–1200 km2 in winter and 200 km2 in summer compared to male home ranges of about 200–800 km2 in winter and 200–300 km2 in summer. In comparison to harbor seals, seal lions have much larger foraging areas. In summer, home ranges of adult female sea lions in the Aleutian Islands and open Gulf of Alaska are about a third larger than those of female harbor seals, averaging just over 300 km2. Summer foraging trips typically last a few hours to a day and extend on average less than 20 km from rookeries, although trips can last 2–4 days and extend up to 250 km (Merrick and Loughlin, 1997; Loughlin et al., 1998). In winter, adult females usually forage for longer intervals and over greater distances than in summer. Winter foraging trips last on average about 8 days and extend 130 km from shore, but longer trips are common, with some females, presumably without pups, remaining at sea for as long as 24 days and traveling as far as 600 km from shore. Home ranges in winter average over 47,000 km2, compared to winter home ranges of female harbor seals of just 600–1200 km2. Young sea lions in their first winter forage over much smaller areas, forage for shorter intervals, dive to shallower depths, and exert less effort than adult females. However, as winter progresses, all of these parameters increase as
128
Long-Term Ecological Change in the Northern Gulf of Alaska
young develop physically and learn the skills and behaviors necessary to forage efficiently on their own. The long nursing period of sea lion pups affords them the luxury of such an extended time for their foraging abilities to mature. Sea otters are more sedentary than even harbor seals, and likely have the smallest home ranges among the marine mammals. Movements of sea otters differ between the sexes and ages, with males exhibiting greater movements than females, and juveniles moving more than adults. In Prince William Sound, mean home range size is just 4.6–11 km2 for males and 4.5–7.6 km2 for most females (Garshelis and Garshelis, 1984). Part of the reason sea otters are so sedentary is that they forage almost exclusively on sessile, or only slightly mobile, benthic invertebrates (Riedman and Estes, 1990; Estes and Bodkin, 2002; Bodkin, 2003) that generally do not exhibit the aggregating or schooling behavior typical of most of the highly mobile fish prey of pinnipeds, but are more uniformly distributed across habitats. Additionally, sea otters exhibit solitary feeding behavior and often have high fidelity to individual foraging locations, in some cases over periods of years (Estes and Bodkin, 2002). Harbor seals are able to dive to considerable depths, at least as great as 500 m, but most dives in Prince William Sound are in the range of 20–100 m (Frost et al., 2001). Still, this may underestimate the importance of foraging at even shallower depths, since 40–60% of the time that seals spent in the water between September and May was in depths <4 m; elsewhere, harbor seals have been found to forage regularly at such depths. Dives are typically deeper in winter, in conjunction with seals moving to offshore areas over the shelf of the Gulf of Alaska. Such movement is likely a response to shifts in prey distribution, as certain energy-rich prey (capelin, eulachon, and salmon) move offshore in winter. Also, pollock, another important winter prey species, is typically found at depths of 150–200 m. Most foraging appears to occur at night in Prince William Sound, although harbor seals are known to forage at all times of day and night (Frost et al., 2001). The foraging effort of adult female sea lions while at sea varies little between summer and winter, with animals diving for about 5 h/day. Most dives last 1–2 min. Dive depths are typically less than 20 m, but deeper, longer dives do occur, especially in winter – a few exceed 250 m in depth and last longer than 8 min (Merrick and Loughlin, 1997). As with harbor seals, sea lions forage at all times of the day and night, but most foraging appears to occur at night at many locations (Loughlin et al., 1998; Thomas and Thorne, 2001; Davis et al., 2002). Because sea otters feed on benthic species, their distribution and foraging behavior is limited by their diving ability. Although sea otters may dive deeper than any of the other otters, they are likely one of the shallowest diving of the marine mammals. Although maximum recorded sea otter dive depths are 100 m, fewer than 2% of more than 117,000 foraging dives recorded were greater than 55 m, whereas 84% were less than 30 m (Bodkin et al., 2004). Mean forage dive durations range from about 1–2.5 min and probably limit dive depths. Sea otters allocate more time to
Ecosystem Structure 129
foraging than most marine mammals because of their high metabolic requirement for thermoregulation, their inability to store large amounts of energy in fat reserves, and their lower assimilation efficiency compared to most mammals (Riedman and Estes, 1990; Costa, 1982). In areas where sea otter density is considered to be at or near equilibrium with prey resources, it may not be uncommon for them to allocate up to 45% of their daily time budget (11 h) to foraging, but where food is not limiting, as little as 25%, or 6 h may be required (Gelatt et al., 2002). Movements Adult harbor seals and sea lions with pups move little during summer because of the need to breed, molt, and suckle pups. At other times of the year, harbor seals move more widely, often utilizing several haul outs and ranging as far as 500 km in extreme cases, with juveniles, especially juvenile females, ranging greater distances than adults. Such movements, at times of the year when seals are not required to remain near rookeries and haul outs, allow them to exploit prey in a diversity of habitats and locations. The propensity of juveniles to range so widely is likely a dispersal behavior. However, despite such movements, tagging and genetic information indicate that most of the seals return to their breeding and molting areas by the following summer (Lowry et al., 2001). Sea lions also have winter haul outs at locations that are not necessarily breeding rookeries. The presumed advantage of such behavior, as for harbor seals, is that it allows them to exploit habitats where prey are abundant. An important winter haul out in the Gulf of Alaska is Cape St. Elias on Kayak Island where animals from as far west as Kodiak Islands come to spend all or significant portions of the winter. Juveniles travel as far as 1500 km from their rookeries after they gain independence from their mothers, but typically return to breed at their natal rookeries or at ones nearby. In contrast to harbor seals, male sea lions disperse more widely than females. Once a female has given birth at a particular rookery, it is highly unlikely that she will move to a different one (Raum-Suryan et al., 2002). While sea otter home ranges generally consist of a few to tens of km2, movements between centers of activity often exceed tens to >100 km, and may be common among adult males that hold seasonal breeding territories. Long-distance movements generally involve dispersal in the case of juveniles, reproductive opportunity in the case of adult males, and pup weaning in the case of adult females (Riedman and Estes, 1990). Where unoccupied habitat is available, sea otters are capable of exhibiting large-scale movements. For example, the abundance of sea otters in Glacier Bay grew from 0 in 1995 to about 1200 in 2002, with nearly all of the increase due to immigration (Bodkin et al., in press). Nevertheless, the generally sedentary nature of most sea otters would tend to constrain their populations in important ways. One would be a limited capacity to exploit increases in prey over spatial scales that exceed normal movement distances. Another would be to limit the ability of populations to recover from large-scale reductions in abundance. An example of the latter is available from the Exxon Valdez
130
Long-Term Ecological Change in the Northern Gulf of Alaska
oil spill. One potential means of recovery in Prince William Sound would have been the post-spill immigration of individuals from nearby unaffected areas into the spill area, where abundance was reduced by up to 90%, yet such redistribution was not observed (Bodkin et al., 2002). Thus, population recovery was delayed in part because of the limited mobility of sea otters and the fidelity they display to their comparatively small home range.
Predator Avoidance As Charles Darwin pointed out 150 years ago, “The amount of food for each species of course gives the extreme limit to which each can increase; but very frequently it is not the obtaining of food, but the serving as prey to other animals, which determines the average numbers of a species.” Although primary production does indeed set the ultimate upper limit on biomass yield at all higher trophic levels, standing stocks at all trophic levels are in dynamic equilibrium with food resources and with predators. Predator defenses in nature run the gamut from the production of toxic compounds, to cryptic coloration, speed, and elaborate behaviors. There are both terrestrial and marine predators of harbor seals, sea lions, and sea otters. All three species have effectively solved the problem of terrestrial predation: in the cases of harbor seals and sea lions, by locating rookeries and haul outs on islands free of predators or on exposed beaches and rocky headlands where access is difficult and approaching predators are conspicuous; and in the case of sea otters, by seldom coming ashore even to breed, give birth, or nurse. The only significant terrestrial predators of any of the three are eagles, which are large enough to snatch young otter pups from the sea and carry them back to land. Marine predators are quite another matter, however. All three species face, in varying degrees, the possibility that they will become meals of sharks or killer whales. The significance of the threat is illustrated by the dramatic collapse of harbor seals at Sable Island, off the coast of Nova Scotia, due to Greenland sharks (Bowen et al., 1999, 2003) and of sea otters in the Aleutian Islands due to killer whales (Estes et al., 1998). Such predation pressure has undoubtedly resulted in various forms of antipredator behaviors in harbor seals, sea lions, and sea otters, most of which consists primarily of predator avoidance strategies. Certainly, all three species are vigilant and have rapid escape responses to disturbance or predator attacks. They may also avoid aquatic predators by spending increased amounts of time hauled out or in very shallow waters near haul outs. Harbor seals and sea lions also reside in groups, which likely reduces the probability of a given individual being taken by a predator, and/or acts to increase overall vigilance, since all members of the group are performing the act (Wells et al., 1999). The manner in which they forage may also be linked to predator avoidance. Harbor seals, as with other phocid seals, forage almost entirely at depth and travel long distances by constantly diving, which minimizes the time spent at the surface where
Ecosystem Structure 131
they are much more vulnerable and less able to sense predators from below. At the same time, this may serve to maximize encounter rates with prey. Although sea lions travel frequently at the surface, they are especially large and extremely fast, which should increase their ability to elude predators. In the case of adult males, they may be large and aggressive enough to also discourage predators. In contrast to the pinnipeds, sea otters derive protection by living in kelp forests, which sharks do not enter, and perhaps by choosing shallow coves and lagoons inaccessible to killer whales. In the case of all three species, although newly weaned offspring have acquired highly developed swimming skills, juveniles are clearly the most vulnerable to predation due both to small body size and less agility at eluding predators.
Summary Table 2.2 contrasts and summarizes the main differences between the three marine mammals considered in this chapter. Harbor seals have evolved an economical, largely capital-based breeding strategy. They have a relatively low field metabolic rate and a thick blubber layer that insulates them and provides a substantial amount of the energy needed by females to produce rich milk during a brief 24-day lactation period. During lactation, the energetic overhead of female harbor seals is low, as they acquire most of the energy (capital) needed to nurse their pups during the months prior to birth. They are good swimmers, but not exceptionally fast. Consequently, their home ranges are typically smaller than those of sea lions, although they still have a large three-dimensional foraging range because they are able to dive deeply. Their diets are diverse and flexible, and numerous species of prey are utilized, most likely in proportion to their abundance at a given time and location, as well as the potential profitability of that prey. To cope with variability in the quantity and nutritional quality of prey, harbor seals have flexible behaviors and digestive physiologies – when faced with scarce or low-quality prey, they consume more and adjust their digestive efficiency to make the most of each meal. Female sea lions have a much more expensive, income-based breeding strategy. They have a higher field metabolic rate than harbor seals, a thinner blubber layer, and a prolonged lactation period. Much of the energy needed to support both the female and the pup during this interval, which can last as long as 3 years, must be obtained daily (income), resulting in a large energy overhead for females. But, sea lions are large and fast and able to search expansive home ranges for prey that are typically aggregated and ephemeral. They too have flexible feeding behaviors and digestive physiologies that maximize the return from consuming prey of varying abundance and nutritional quality. Moreover, they have long digestive tracts compared to other pinnipeds, which further enhances digestive efficiency.
132
Long-Term Ecological Change in the Northern Gulf of Alaska
Table 2.2: General size and life history characteristics of harbor seals, Steller sea lions, and sea otters. Character
Harbor Seal
Sea Lion
Sea Otter
Adult size (male/female, kg) Pup size (birth/weaning, kg) Dimorphism (male/female mass) Longevity (years) Insulation
97/75 11/27 1.3 35 Blubber
566/264 22/95* 2.0 30 Blubber
29/20 2/10 1.3 20 Air in fur
Reproduction Reproductive rate Maturation (years) Gestation (years) Offspring Dependency (years)
0.85 4 0.85 1 0.07
0.63 5 0.9 1 ≥1
0.9 3 0.5 1 0.5
Diving max (m) Consumption (% of mass/day)
Fish and invertebrates 500 0.07
Fish and invertebrates 328 0.06
Benthic invertebrates 100 0.25
Movements Foraging range (km)
10–100s km
10–1000s km
<5 km
Energy Expenditure Lactation period (MJ/kg)
9
77
117
Trophics Diet
Note: For many of these parameters, values are not fixed and will vary over time. *Estimate based on body masses at 10 months postpartum (L. Rea, pers. comm.). Sources: Loughlin, 2002; Boyd, 2003: Burns, 2002, 2003; Coltman et al., 1998; Ellis et al., 2000; Estes and Bodkin, 2003; Hoover, 1988a,b; Loughlin, 2002; Meulbert et al., 2003; NRC, 2003; Riedman and Estes, 1990; Iverson et al., 2003; Bowen et al., 2001.
Sea otters also have an income-based breeding strategy that is the most expensive of the three. They have no blubber, but insulate with fur, which is superior to blubber but requires nearly constant attention to maintain its insulating qualities. Their metabolic rate is very high, and because they have few energy reserves, must forage much of the time and consume prodigious quantities of food, relying on the heat increment of feeding for much of their thermoregulation. Maternal energetic overhead is very high year round and particularly during lactation. Sea otters are good swimmers, but slow, and cannot dive as deeply as pinnipeds. To help compensate for these burdens, they typically feed on sessile or slowly moving prey in predictable locations such that foraging efficiency is high, with prey captured on about 90% of
Ecosystem Structure 133
the feeding dives in Prince William Sound (Dean et al., 2002). Heat produced from both activity and digestion is appropriated for thermoregulation. Harbor seal pups are weaned abruptly at just over 3 weeks of age and must learn to forage independently. To compensate, they rapidly acquire a thick blubber layer that provides energy during the first few weeks of independence. They are capable swimmers within days of birth. In contrast, sea lion pups grow slowly by comparison, develop a thinner blubber reserve, but remain with their mothers and suckle for extended periods. During this time, they have the luxury of a dependable food supply while developing lean body mass and learning foraging and other critical survival skills, such as predator avoidance. Thus, while harbor seals have a less energetically demanding reproductive strategy than sea lions, they do not have the option of extending lactation, and when newly weaned pups have difficulty finding adequate food, they probably have high mortality. In comparison, juvenile sea lions can be buffered through difficult times by extended suckling. Even during tough times, harbor seals tend to produce pups each year but the pups have low survival, whereas sea lions may not have a pup every year but instead supply their young with additional milk as juveniles, which then survive at a higher rate. These different strategies lead the two species to a similar situation in the end. In contrast, sea otter pups are in constant contact with their mothers for several months, develop essentially no fat reserves, but are nursed, provided prey, and taught how to forage by the time they are weaned. Nevertheless, despite these different strategies for raising pups, juvenile mortality is high for all three species and is the key factor driving population dynamics under all but extreme circumstances. All three species are dietary generalists within certain limits – harbor seals and sea lions generalize primarily on a diverse suite of fishes, while sea otters generalize primarily on a diverse suite of invertebrates. Indeed, one of the principal ecological differences between sea otters and the two pinnipeds is the trophic webs they occupy. Harbor seals and sea lions forage primarily on prey in a food web based on carbon fixed by phytoplankton, or micro-algae. In contrast, sea otters occupy a position near the top of a food web supported at its base by carbon fixed by kelps, or macro-algae (Duggins et al., 1989). Such an ecological distinction is important to remember when considering the ways in which the primary producers and the productivity of food webs they fuel might respond to environmental change. Harbor seals, sea lions, and sea otters are all flexible in their choice of prey, which allows them to exploit species that have particular geographic distributions and to switch between species at particular locations depending on availability, which commonly varies seasonally, annually, and over longer intervals. Diverse, flexible diets buffer them against vagaries in the abundance of individual prey species, or suites of prey, that can be due to factors such as fisheries or natural environmental change, e.g., as forced by climate.
134
Long-Term Ecological Change in the Northern Gulf of Alaska
Sea lions forage over much larger areas than harbor seals, while sea otters are the most sedentary of the three. Harbor seals and sea lions require considerable flexibility in the distance they travel to forage and the duration of foraging bouts, because the fish they prey on are highly mobile. That is to say, they must be able to travel long distances and/or dive deeply, if necessary, to find prey. In contrast, sea otters travel distances of a few meters to several kilometers from resting areas to forage, and individuals tend to forage in consistent locations or habitats for periods of up to many years. Thus, sea otters rely on a spatially predictable prey base but one that probably does not have the capacity to increase (or decrease) over short time periods, such as with the transitory forage fishes central to the pinniped diet. These differences in foraging strategies may reflect other life history differences, primarily in the ability to store and transfer energy to offspring. Sea lions range widely during the nonbreeding season and commonly move to winter haul outs that may be long distances from summer rookeries. Harbor seals also range more widely in winter than in summer, but range less widely at either time than sea lions. Such behavior in the nonbreeding season when animals are not tied to shores where they molt and raise pups, either to weaning or to a size and ability that they can accompany their mothers to sea, presumably allows them to exploit prey resources having high spatial/temporal variability. Unlike harbor seals and sea lions, sea otters have no obligate ties to terrestrial habitats for resting, molting, or breeding. However, their limited diving ability and need for frequent feeding does require that they maintain close proximity to shallow benthic habitats. So, although pinnipeds require frequent access to predator-free terrestrial haul outs, they maintain the ability to travel widely in search of their mobile and often spatially clumped prey. While sea otters have no need for the haul out, their movements are instead constrained by physiological dive limits. In both cases, ties to their terrestrial past are reflected in life history attributes that may demonstrate their relatively recent membership in the community of marine mammals. Life in the sea frees all three species from most threats imposed by terrestrial predators. However, marine predators, primarily killer whales in Alaska, are a common threat throughout their ranges. Defensive mechanisms of sea lions include their vigilance, large size, and speed. Harbor seals commonly remain close to shore and spend most of their time at depth when traveling and foraging. Sea otters also remain close to shore, often in protective kelp forest canopies. Nevertheless, each of them is preyed upon to varying degrees, depending on the location and dietary proclivities of local killer whales. Thus, harbor seals, sea lions, and sea otters display an array of adaptations that reflect both their phylogenetic history and requirements imposed by a life in the marine environment. Similarities include delayed maturation, low reproductive output, long life spans, and diverse and flexible diets. Such life history traits typically buffer species possessing them against environmental uncertainty that can lead to large and rapid fluctuations in abundance. Yet, despite these buffers, and various
Ecosystem Structure 135
combinations of foraging strategies, energetics, and movements during the nonbreeding season that allow them each to exploit the marine environment in unique ways, populations of all three suffered widespread, rapid collapses in the Gulf of Alaska, Aleutian Islands, and Bering Sea in the past three decades (see Section 4.9). The common geographic extent, the similarity in magnitude, and the timing of the declines are suggestive of a common cause. Because there are such large fundamental differences in energetics, habitats, trophic positions, and food web structure among the three species, and especially between sea otters and the two pinnipeds, this is suggestive of causes independent of food availability. It seems apparent that the problems previously faced by these populations in recent decades include factors not accommodated by the repertory of life history adaptations common among them.
2.5.9. Crabs and Shrimps Gordon H. Kruse Introduction Invertebrate populations fluctuate widely in the Gulf of Alaska, often due to variable survival in early life. Crab and shrimp eggs, embryos, larvae, and young juveniles run a gauntlet – predators, winter storms, currents that sweep larvae away from nursery grounds, lack of adequate food quantity and quality – and can have poor survival for many consecutive years. It is a rare year when conditions align favorably and large numbers of eggs survive to adulthood, producing a “large year-class” or “strong recruitment” that sustains fisheries. In this chapter, we consider the strategies of red king crab (Paralithodes camtschaticus), Tanner crab (Chionoecetes bairdi), and northern (sometimes locally called “pink”) shrimp (Pandalus borealis) to maintain their populations in a highly uncertain environment. For crabs, we draw from publications of original research and reviews of biology and life histories, as well as workshops in which biological and physical processes important to survival were identified. The biology and life history of northern shrimp is not studied as well as the biology and life history of red king and Tanner crabs in Alaska. So, for shrimp, we also draw from research conducted on the same species in the North Atlantic Ocean. In Section 4.10, we consider the successes and failures of these survival strategies in populations of crabs and shrimps since the 1970s, and the mechanisms that lead to these patterns. Distribution and Habitats for Survival Northern shrimp are widely distributed in the North Atlantic and North Pacific Oceans. They are nektonic or water–water column species and live over a wide range of depths, from 20 to 1400 m. In the Gulf of Alaska, their greatest historical
136
Long-Term Ecological Change in the Northern Gulf of Alaska
abundances were around Kodiak and Shumagin Islands at depths of 110–220 m. The northern shrimp and the other major species of shrimp in the Gulf of Alaska are shown in Fig. 2.46. Unlike northern shrimp, Tanner crabs do not occur in the North Atlantic Ocean, but the closely related snow crabs (Chionoecetes opilio) occur in the Bering Sea and North Atlantic Ocean. Adult Tanner crabs are distributed throughout the northeast Pacific Ocean and Bering Sea at depths of 0–465 m, typically on muddy and sandy seabeds. Late in the pelagic larval stage, Tanner crabs change into megalopae, which can live both in the water (pelagic) and on the bottom (benthic) while searching for their new homes. Tanner crab megalopae seem to prefer silt and mud habitats in relatively deep, cool waters. Once the megalopae locate suitable habitat, they settle on the seabed and molt into early juveniles, resembling miniature adults. Red king crabs occur from British Columbia through the Gulf of Alaska, into the eastern and western Bering Sea, and along the northwest Pacific as far south as Japan. Adults live mainly on sand and mud bottoms at depths of 0–400 m. As an interesting
Figure 2.46: The major pandalid shrimp species of the western Gulf of Alaska. Clockwise from bottom right: Pandalus goniurus, P. borealis, P. tridens, Pandalopsis dispar, and Pandalus hysinotus (photograph courtesy of the Kodiak Fisheries Research Center).
Ecosystem Structure 137
side note, red king crabs from Kamchatka were transplanted into the Barents Sea in the northeast Atlantic Ocean in the 1960s. They have increased in Russian and Norwegian waters, providing a new fishery, but also becoming a bycatch nuisance to longline and gillnet fisheries for cod and other groundfish. The habitats of early juvenile stages of red king crab deserve special mention. Near the completion of the pelagic larval stage, red king crabs transform into intermediate forms, called glaucothoe (Fig. 2.47), akin to the megalopae of Tanner crabs. Just as with Tanner crab megalopae, once the glaucothoe locate suitable habitat, they settle and molt to become early juveniles. Habitat requirements of early juvenile red king crabs are nearshore, high-relief bottom types with high-profile sessile (attached, nonmobile) fauna, including areas with sea stars, anemones, ectoprocts (bryozoans), sea onions (stalked tunicates), soft coral colonies, shell hash, algae, rocks, and other structures providing cover.
Feeding Strategies The six major prey groups of adult red king crabs off Kodiak Island are annelids (i.e., polychaete worms), molluscs (mostly bivalves, such as clams, but also snails), arthropods (primarily crustaceans, such as barnacles, Tanner crabs, and shrimp), echinoderms (e.g., sea stars, brittle stars, and sea urchins), fish, and plant material. Bivalves and barnacles are the dominant prey off Kodiak Island; however, dominant prey types differ among areas. For example, barnacles dominate the diet in some bays and fishes in others. In lower Cook Inlet, the three most dominant prey were barnacles, bivalves, and hermit crabs. Male and female diets are similar. In general, red king crab stomachs contain more food in late spring and early summer than in other seasons, and crabs collected in shallow water (<25 m) have fuller stomachs than crabs
Figure 2.47: The glaucothoe stage of the red king crab is a transitional stage between the fourth zoea and the first crab stage. Although it has claws like the juvenile crab, it can swim using its abdominal appendages. This allows it to seek optimal shelter habitat for settlement (photograph courtesy of Brad Stevens, National Marine Fisheries Service).
138
Long-Term Ecological Change in the Northern Gulf of Alaska
collected in deep waters (Feder and Paul, 1980; Jewett and Feder, 1982). Although field studies do not indicate significant rates of cannibalism in king crabs, great care must be taken in the laboratory to prevent cannibalism, particularly among very young red king crabs, but also among molting adults. The diet of juvenile king crabs is composed of smaller prey items; newly settled crabs were found to eat mainly small crustaceans. Diets of adult Tanner crabs are similar to those of red king crabs. Off Kodiak, common Tanner crab prey are bivalves, arthropods (mostly shrimps and other Tanner crabs), and fish. In Cook Inlet, the top three prey were clams, hermit crabs, and barnacles. Interestingly, in contrast to the relative low rates of cannibalisms in fieldcaptured red king crabs, cannibalism is common in Tanner crabs collected near Kodiak, with 16% of all Tanner crab stomachs containing other Tanner crabs. Also in contrast to king crabs, Tanner crab stomachs tended to have greater fullness in winter, and Tanner crabs ate less food in shallow (26–50 m) than deep (126–150 m) waters (Jewett and Feder, 1983; Paul et al., 1979). Northern shrimp tend to be omnivorous, and they feed both as predators and scavengers. In lower Cook Inlet, they consumed 28 categories of prey, and the three most common prey were unidentified crustaceans, polychaetes, and diatoms (Rice et al., 1980). Other prey items include bivalves, foraminifera, and fish and plant remains. Elsewhere, they ate crustaceans, including copepods and euphausiids.
Predator Defense Strategies Northern shrimp are eaten not only by invertebrates, such as king and Tanner crabs, but also by Pacific cod (Gadus macrocephalus), arrowtooth flounder (Atheresthes stomias), walleye pollock (Theragra chalcogramma), spiny dogfish (Squalus acanthias), yellowfin sole (Pleuronectes asper), and Pacific halibut (Hippoglossus stenolepis), as well as by man (Homo sapiens). In particular, Pacific cod are notorious predators of northern shrimp (Albers and Anderson, 1985). Red king crabs are minor components in the stomachs of cod, halibut, sculpins, and other groundfish. Molting adult crabs are most vulnerable. In the eastern Bering Sea, most adult red king crabs in cod stomachs are molting female red king crabs (Livingston, 1989). Molting adult male red king crabs are rarely found in cod stomachs. However, definitive conclusions about cod predation rates on female versus male red king crabs must be tempered by the fact that most cod stomach samples come from late spring through early fall, a time period that encompasses the late spring–early summer molting period of females but not the late winter–early spring molting period of adult males. Likewise, few young-of-the-year red king crabs typically occur in groundfish stomachs, but groundfish are not collected for stomach analysis from the shallow, nearshore waters, where young-of-the-year red king crabs reside. Stomachs of some sampled flatfish, such as yellowfin sole, on occasion contain large numbers of young-of-the-year blue
Ecosystem Structure 139
king crabs (Paralithodes platypus), which sometimes occupy similar habitats as red king crabs. In contrast to red king crabs, groundfish stomach data clearly show that Tanner crabs are heavily preyed upon, particularly by Pacific cod. A wide range of sizes, including adults, are consumed, but most predation occurs on juvenile Tanner crabs in their first or second year of life. As cod body size increases, they tend to switch their diets from small invertebrates to larger fishes. Thus, cod consumption of crabs and shrimp may increase during periods of strong cod recruitment, during which the cod population is rejuvenated, and decrease during periods of weak recruitment associated with an aging cod population. Parasites are an additional factor. Tanner crabs suffer from bitter crab syndrome, a dinoflagellate parasite that causes 100% mortality in infected crabs (Meyers et al., 1987); see Section 3.6 for a discussion of bitter crab disease. Incidences are generally relatively low, but in some areas of Southeast Alaska, rates of infection exceed 95%. Different strategies for survival may explain the extent of predation on these three invertebrate species. Pandalid shrimp are soft bodied and may live above mud and sand bottoms without the advantages of rocky shelter. Therefore, shrimp appear to have relatively few defenses from predators, and many species eat them besides cod. In contrast to shrimp, adult Tanner and red king crabs are heavily armored against predation. Male red king crabs molt earlier in the year than females. Therefore as mating approaches, they are re-armored and provide females protection from predators while they molt prior to mating. Also, differences in survival strategies probably explain higher predation rates on juvenile Tanner crabs relative to juvenile king crabs. Tanner crabs live on muddy bottoms along with many groundfish, and they avoid these predators by burying. Juvenile red king crabs have two, perhaps better, strategies to guard against predation. In the early juvenile stage, king crabs are solitary animals, living in shallow water among high-relief structures, thus having the ability to hide from their comrades who might cannibalize them, as well as from groundfish predators whose abundances are higher in deeper waters where the commercial fisheries are conducted. Second, once king crab juveniles reach age 2, they begin to form intense aggregations of hundreds to thousands of individuals, called pods, often in shallow water (Powell and Nickerson, 1965, Fig. 2.48). Pods break up near dusk, as crabs forage on their own, and reform near dawn. Molting occurs only at night (Dew, 1990). The pods are probably a defense against predators. It is much more difficult for a predator to find one pod than to find a crab, if the thousands of crabs in the pod were instead widely distributed over the seafloor. Dew (1990) speculated that podding is initiated when king crabs grow too big to reside within the crevices of their environment. Molting at night is likely a defense against cannibalism of red king crabs by their siblings. If one crab molts while in the pod, the others would eat it. But if a crab molts alone at night, then the chances of being eaten by other crabs is much reduced.
140
Long-Term Ecological Change in the Northern Gulf of Alaska
Figure 2.48: Juvenile red king crabs form aggregations called pods of hundreds to many thousands of individuals in coastal waters of Alaska. Adults have also been found in some pods in Kodiak. Shown here is a small pod photographed in Auke Bay, Alaska (photograph courtesy of Lou Barr).
However, as demonstrated by Stone et al. (1993), podding behavior is not confined to juveniles. To date, most observed pods comprise juveniles in shallow water, easily accessible to scuba divers (Fig. 2.48). The extent to which mature adults form pods in deeper waters remains uncertain because of the lack of observations in deeper water.
Reproductive Strategies Crabs and shrimps use several other strategies besides predator defenses to increase the chances for survival of their progeny (see reviews in Kruse, 1993 and Orensanz et al., 1998). The reproductive strategies of these invertebrate species also represent an interesting set of solutions to the problem of maintaining viable populations in a highly variable marine environment. Considerations of similarities and contrasts in their reproductive tactics provide insights into causes of large fluctuations of their populations in the Gulf of Alaska. Commercially caught Alaskan crabs are relatively large bodied and long lived. In Kodiak, male red king crabs attain a maximum carapace length of 227 mm, weigh 11 kg, and reach a maximum age of >20 years. Males mature at a mean age of 7 years.
Ecosystem Structure 141
Male Tanner crabs may attain maximum carapace of width 200 mm, weigh 2.6 kg and reach a maximum age of 12 to 15 years. Males mature at a mean age of 6–7 years. Of course, shrimps are much smaller than crabs, but live longer than their counterparts in warmer waters. For instance, northern shrimp reach maximum carapace length of 25 mm, males mature at 2 years, and total life span is ~8 years (Anderson, 1991). In stark contrast to the crabs and shrimp off Alaska, along the midAtlantic coast of the U.S., blue crabs (Callinectes sapidus) may become mature as young as age 1 and typically live up to 2 years of age, although a few live up to 3 years. In the Gulf of Mexico, brown shrimp (Farfantepenaeus aztecus) mature and spawn at 7 months of age and live up to 2 years only. Longevity allows a species to reproduce annually over multiple years, an attribute called iteroparity. The longevity and multiple years of reproduction of Alaskan crabs and shrimp stand in sharp contrast to two warmer-water species, the brown shrimp and blue crabs, which reproduce only twice (brown shrimp) or at most three times (blue crabs). Ecological theory suggests that semelparity (reproduction only once in a life time) is favored by species in relatively stable, predictable environments. For such species, lesser energy is required for maintenance and more energy can be allocated to reproduction. On the other hand, iteroparity tends to be favored in less stable, unpredictable environments. Under such conditions, early-life survival is typically very low and uncertain. By reproducing many times during a long life span, Alaskan crabs and shrimp “do not put all eggs in one basket”. This is a bet-hedging strategy to spread the risk of reproductive failure over many attempts in the hopes of at least one reproductive success, i.e., a strong year-class. Female Tanner crabs can retain viable sperm from their male mates in receptacles, called spermathecae, for up to 2 years or more after insemination (Paul, 1984). Spawning can occur later in the absence of a male, when these reserves are used to fertilize subsequently extruded egg clutches. Female red king crabs lack these structures; hence, males must be present during egg extrusion for successful fertilization to occur (Powell and Nickerson, 1965). Thus, red king crabs may have greater sensitivity to sex ratio, as female red king crabs must mate in the year in which they carry embryos, whereas female Tanner crabs may carry sperm from a previous mating for use in current-year fertilization. In fact, a male Tanner crab can be on a dinner plate while still fathering progeny in the Gulf of Alaska. The northern shrimp is a protandric hermaphrodite that begins its life as a male, goes through a transition, and later becomes a female, a strategy shared with other shrimps. This may be an adaptation to promote successful reproduction (Fox, 1972). Males, being more numerous, may mate with many females, and a female can be mated more than once. Females, being larger, can produce more eggs than if they were the smaller sex of the species. The age at which this sex change occurs may vary with fluctuations in abundance and age or size distribution of breeding adults (Charnov and Anderson, 1989). Shrimp sex ratios may vary owing both to variations
142
Long-Term Ecological Change in the Northern Gulf of Alaska
in recruitment and to size-selective fisheries that differentially harvest the larger shrimp, most of which are females. Another reproductive strategy is to produce many young. Red king crabs produce 43,000–500,000 embryos and Tanner crabs produce 50,000–400,000 embryos, whereas northern shrimp produce only 200–3500 embryos per clutch. Pandalid shrimp carry substantially fewer embryos than crabs, partly owing to their smaller body size. Bear in mind that, on average over the long term, during their life time, each pair of mates produces only two progeny surviving to adulthood to replace their position in the population. So, high fecundity is yet another strategy to survive in an unpredictable environment in which the chances of survival of a single progeny approach nil. Female red king and Tanner crabs carry embryos for nearly 1 year before they hatch and larvae begin their independent lives. Female northern shrimp extrude eggs in fall, and hatching occurs in spring. This investment of maternal care is a mechanism to increase survival during incubation. Nonetheless, a successful outcome is not preordained. In the case of red king crabs, nemertean worms and amphipods feed on egg masses carried by females, potentially inflicting high rates of predation in some areas and years (Kuris et al., 1991). There may be trade-offs between growth and reproduction. Under adverse conditions (e.g., increasing temperatures), northern shrimp maintains somatic (body) tissues at the expense of egg production (Nunes, 1984). That is, females can skip reproducing until conditions improve and subsequent increased body size can support a larger clutch of embryos. This is yet another bet-hedging strategy in which the female bets on her ability to reproduce successfully at some later time against the risk that she will die before then. As red king crabs mature, they tend not to molt every year. Among mating pairs collected in the field off Kodiak Island, most (56–61%) males were oldshell or very oldshell animals, not having molted for one year or more (Powell et al., 2002). Males are larger than females in nearly all (95%) mating pairs, and the mean difference in carapace length between males and females was 33 mm. Taken together, this seems to suggest that it is a priority for male crabs to grow to large body size and to participate in mating in years when they do not molt. Under laboratory conditions, male Tanner crabs cannot mate for 100 days after molting, and oldshell males outperform newshells (i.e., molted within the current year) in contested matings with females (Paul et al., 1995). Taken together, these observations provide some evidence that there may be a tendency for male crabs to either mate or grow in a particular year, but not both. Tanner crab mating is even more complex, owing to a phenomenon known as bipartite (i.e., two part) breeding. Females that have not yet spawned for the first time (termed pubescent) molt to sizes of maturity and mate with relatively small males in shallow water during January to May. On the other hand, females that have their first egg clutch are called primiparous, and those with subsequent clutches are
Ecosystem Structure 143
called multiparous and tend to form mounds and mate with larger, oldshell males in deeper waters in April and May (Fig. 2.49).
Larval Survival: Strategies of Timing (or “Match the Hatch”) Relatively large phytoplankters, called diatoms, especially Thalassiosira spp., are the primary food of the early larval stages of red king crab larvae, whereas copepods, particularly young stages, called nauplii, of the genus Pseudocalanus, are primary food of Tanner crab larvae. The development rate of red king crab embryos is strictly a function of temperature, but hatching is apparently cued by the spring phytoplankton bloom (Shirley and Shirley, 1989; Shirley et al., 1990); availability of phytoplankton prey may have a strong influence on survival of newly hatched larvae (Paul and Paul, 1980). It is not known whether there are cues to synchronize red crab larval hatching and availability of their food. Mechanisms for synchronization in Tanner crabs, on the other hand, are rather well known. The mounds (Fig. 49), formed by primiparous and multiparous female Tanner crabs in spring, seem to be triggered by tidal rhythms associated with the highest spring tide in April or May (Stevens et al., 1999; Stevens, 2003). The timing of this event with spring tides suggests that mounds serve as larval launch pads to elevate larvae away from silty sea floor, thereby improving their chances of reaching upper layers of the ocean, where plankton concentrations are highest (Stevens, 2003). In the congener, snow crab (Chionoecetes opilio),
Figure 2.49: Female Tanner crabs forming a “mound” of about 500 crabs, at 150 m depth in Chiniak Bay. A large aggregation may contain up to 200 such mounds, in an area of approximately 2 ha (photograph courtesy of Brad Stevens, National Marine Fisheries Service).
144
Long-Term Ecological Change in the Northern Gulf of Alaska
hatching occurs in conjunction with sinking of metabolites associated with the phytoplankton bloom, which may trigger egg hatching, thus synchronizing crab larval life with that of their zooplankton prey (Starr et al., 1994). However, no laboratory experiments have tested the hatching cues. Environmental cues, at least for some species, may allow them to time the hatch to the availability of larval food. Larval northern shrimp can feed on a variety of phytoplankton, such as diatoms and dinoflagellates, and on zooplankton nauplii. Shrimp larvae survived better on diets of large diatoms than small diatoms in laboratory studies (Nunes, 1984). Whereas shrimp survived reasonably well when fed the large diatom, Thalassiosira, the preferred prey of larval red king crabs, survival was much higher when shrimp were fed dinoflagellates, or when fed a mixed diet of phytoplankton and zooplankton. The flexible feeding strategy implied by these findings may maximize production of successful shrimp larvae (Nunes, 1984). That is, some level of survival may result, regardless of whether larvae hatch before, during, or after the spring bloom.
Larval Survival: Strategies of Space (or “Location, Location, Location”) Female red king crabs move to shallow water in spring, where embryos hatch (Stone et al., 1992), and the glaucothoe prefer to settle on these structurally complex nearshore nursery areas (Stevens and Kittaka, 1998; Loher and Armstrong, 2005). Alongshore currents would tend to retain larvae in the vicinity of their settling habitats. However, nursery areas are patchily distributed along the Alaskan coastline, and there is little assurance that larvae will be carried to suitable areas for settlement at the end of their larval life. Tanner crab adults seem to prefer deep-water habitats during hatching of embryos, and Tanner crab megalopae also seem to prefer to settle in relatively deep-water habitats of silt, fine sand, and mud, which also tend to be more common offshore (Paul, 1982). However, the vagaries of open ocean currents may carry Tanner crab larvae over vast distances and, as with king crab larvae, the chances of larval settling in ideal nursery habitats may be small. Spatial distribution of northern shrimp and their young are not well studied in Alaska. Off British Columbia, newly hatched larvae can be found near spawned out adults (Berkeley, 1930). Later larval stages are found separated from adults in shallow (10–65 m) water, where they remain during their first summer. During their first winter, young shrimp migrate to deeper water to join the adults. It is not clear whether northern shrimp require specific habitats for nursery areas as the crabs do. In Kachemak Bay, northern shrimp make daily vertical migrations; they leave the bottom near dusk and return to the bottom near dawn (Barr, 1970). Perhaps this has something to do with feeding (upper water column) and predator avoidance (lower water column). In any case, shrimp seem to have greater flexibility in their location than Tanner and red king crabs, which cannot swim off the bottom as juveniles and adults.
Ecosystem Structure 145
Implications of Survival Strategies In summary, different life history strategies are employed by red king crabs, Tanner crabs, and northern shrimp to deal with the hazards of surviving in a hostile environment. From this review, some hints can be garnered about factors likely to cause the observed historical changes in crab and shrimp populations in the Gulf of Alaska. For instance, among the three invertebrate species, northern shrimp appear to have the fewest predator defenses, whereas red king crabs seem to have better defense strategies than Tanner crabs. Longevity of all three species, compared to many invertebrate counterparts in warmer waters, is a mechanism to deal with the rarity of good year-classes that sustain populations and fisheries. This strategy evolved over many generations and many eons. However, the success of these strategies may be severely compromised when Mother Nature or humans change the rules. For instance, the longevity strategy renders these species very vulnerable to fishing pressure, as fishing tend to truncate age structure, i.e., remove the oldest ages. So, the reproductive lifespan is reduced, and these populations have fewer years, perhaps half as many, to produce a successful year-class of progeny to replace themselves. Likewise, this strategy renders these species vulnerable to long-term regime shifts in climate, including global warming, when these climate changes are associated with poor early-life survival for time periods that exceed the population generation time. In Section 4.10, we will reconsider the survival strategies that we discussed here with respect to specific hypotheses about mechanisms that may explain historical changes in crab and shrimp populations and fisheries. As we will see, the great challenge is to attempt to separate the effects of humans from those of the natural environment. The fact that the combination of climate-driven regime shifts and intense fishing pressure can act synergistically to affect the viability of invertebrate populations means that this task is even more difficult than it may seem.
References Ahlnäes, K., T. C. Royer, and T. H. George. 1987. Multipole dipole eddies in the Alaska Coastal Current detected with Landsat thematic mapper data. J. Geophys. Res. 92: 13041–13047. Ainley, D. G., D. N. Nettleship, H. R. Carter, and A. E. Storey. 2002. Common murre (Uria aalge), In: A. Poole and F. Gill (Eds.), The Birds of North America. vol. 666, Philadelphia, PA: The Birds of North America, Inc. Albers, W. D. and P. J. Anderson. 1985. Diet of Pacific cod, Gadus macrocephalus, and predation on the northern pink shrimp, Pandalus borealis, in Pavlof Bay, Alaska. Fish. Bull. 83: 601–610. Allen, S. E. 1996. Topographically generated, subinertial flows within a finitie length canyon. J. Phys. Oceanog. 26: 1608–1632. Anderson, J. T. 1988. A review of size dependent survival during pre-recruit stages of fishes in relation to recruitment. J. Northwest Atlantic Fish. Sci. 8: 55–66.
146
Long-Term Ecological Change in the Northern Gulf of Alaska
Anderson, J. T., K. M. Bailey, and L. Cianelli. 2002. Consequences of a superabundance of larval walleye pollock, Theragra chalcogramma, in the Gulf of Alaska in 1981. Mar. Ecol. Prog. Ser. 243: 179–190. Anderson, P. J. and J. F. Piatt. 1999. Community reorganization in the Gulf of Alaska following ocean climate regime shift. Mar. Ecol. Prog. Ser. 189: 117–123. Anderson, P. L. 1991. Age, growth, and mortality of northern shrimp, Pandalus borealis Krˆyer, in Pavlof Bay, Alaska. Fish. Bull. 89: 541–553. Arendt, A. A., K. A. Echelmeyer, W. D. Harrison, C. S. Lingle, and V. B. Valentine. 2002. Rapid Wastage of Alaska Glaciers and their contribution to rising sea level. Science 297: 382–386. Ashmole, N. P. 1971. Sea bird ecology and the marine environment. In: J. K. D. S. Farner, K. C. Parks, (Eds.), Avian Biology, vol. 1. New York: Academic Press, pp. 224–286. Bailey, J. E. 1969. Alaska’s fishery resources – the pink salmon. Bur. Commer. Fish. Leafl. 169. U.S. Fish and Wildlife Ser., Washington, DC, 8 p. Bailey, K. M. 2000. Shifting control of recruitment of walleye pollock Theragra chalcogramma after a major climatic and ecosystem change. Mar. Ecol. Prog. Ser. 198: 215–224. Bailey, K. M., A. L. Brown, M. M. Yoklavich, and K. L. Mier. 1996. Interannual variability in growth of larval and juvenile walleye pollock Theragra chalcogramma in the western Gulf of Alaska, 1983–91. Fish. Oceanogr. 5(Suppl. 1): 137–147. Bailey, K. M., L. Ciannelli, and V. N. Agostini. 2003. Complexity and constraints combined in simple models of recruitment, In: I. Browman and A.B. Skiftesvik (Eds.). The Big Fish Bang: Proceedings of the 26th Annual Larval Fish Conference, HBergen, Norway: Institute of Marine Research. Bailey, K. M., T. J. Quinn, P. Bentzen, and W. S. Grant. 1999. Population structure and dynamics of walleye pollock, Theragra chalcogramma. Adv. Mar. Biol. 37: 179–255. Baird, P. H. 1991. Optimal foraging and intraspecific competition in the tufted puffin. The Condor 93: 503–515. Baird, P. H. 1994. Black-legged kittiwake (Rissa tridactyla). In: A. Poole and F. Gill, (Eds.), The Birds of North America, vol. 92. Philadelphia: The Academy of Natural Sciences. Baker J. D. and C. W. Fowler. 1992. Pup weight and survival of northern fur seals Callorhinus ursinus. J. Zool., London 227: 231–238. Barr, L. 1970. Diel vertical migration of Pandalus borealis in Kachemak Bay, Alaska. J. Fish.Res. Bd. Can. 27: 669–676. Baumgartner, A. and E. Reichel. 1975. The World Water Balance, Elsevier, New York, 179 pp. Berkeley, A. A. 1930. The post-embryonic development of the common pandalids of British Columbia. Contrib. Can. Biol. 10: 79–163. Bigg, M. A. 1973. Adaptations in the breeding of the harbor seal (Phoca vitulina). J. Reprod. Fertility Supplement 19: 131–142. Biggs, E. D., B. E. Haley, and J. M. Gilman. 1992. Historical data base for Pacific herring in Prince William Sound, Alaska, 1973–1991. Regional Information Report 2C91-11 Department of Fish and Game, Division of Commercial Fisheries, Anchorage, Alaska. Birkhead, T. R. 1977. Adaptive significance of the nestling period of guillemots Uria aalge. Ibis 119: 544–549. Birkhead, T. R. and D. N. Nettleship. 1984. Alloparental care in the common murre (Uria aalge). Can. J. Zool. 62: 2121–2124.
Ecosystem Structure 147
Birkhead, T. R. and M. P. Harris. 1985. Ecological adaptations for breeding in the Atlantic Alcidae. In: D. N. Nettleship and T. R. Birkhead (Eds.), The Atlantic Alcidae. San Diego, CA: Academic Press, Inc, pp. 205–232. Bishop, M. A. and S. P. Green. 2001. Predation on Pacific Herring (Clupea pallasi) spawn by birds in Prince William Sound, Alaska. Fish. Oceanogr. 10(Suppl. 1): 149–158. Bodkin, J. L. 2003. Sea otter. In: G. A. Feldhammer, B. C. Thompson and J. A. Chapman (Eds.), Wild Mammals of North America. Baltimore, MD, Johns Hopkins University Press, pp. 735–743I. Bodkin, J. L., B. E. Ballachey, T. A. Dean, A. K. Fukuyama, S. C. Jewett, L. M. McDonald, D. H. Monson, C. E. O’Clair, and G. R. VanBlaricom. 2002. Sea otter population status and the process of recovery from the Exxon Valdez oil spill. Mar. Ecol. Prog. Ser. 241: 237–253. Bodkin, J. L., G. G. Esslinger and D. H. Monson. 2004. Foraging depths of sea otters and implications to coastal marine communities. Mar. Mamm. Sci. 20: 305–321. Bodkin, J. L., B. E. Ballachey, K. A. Kloecker, G. G. Esslinger, D. H. Monson, H. A. Coletti (In press) In: J. Piatt and S. Gende (Eds.), Perspectives of an invading predator: Sea otters in Glacier Bay. Proceedings 4th Glacier Bay Science Symposium. USGS Tech. Rep. Boekelheide, R. J., D. G. Ainley, S. H. Morrell, H. R. Huber, and T. J. Lewis. 1990. Common murre. In: D. G. Ainley and R. J. Boekelheide, (Eds.), Seabirds of the Farallon Islands. Stanford, CA: Stanford University Press, pp. 245–275. Bograd, S. J., P. J. Stabeno, and J. D. Schumacher. 1994. A census of mesoscale eddies in Shelikof Strait, Alaska during 1989. J. Geophys. Res. 99: 18243–18254. Boldt, J. L. 2001. Ecology of Pink Salmon in the north Gulf of Alaska and Prince William Sound. PhD Thesis, University of Alaska Fairbanks, Alaska, 235 p. Boness, D. J., W. D. Bowen, and O. T. Oftedal. 1994. Evidence of a maternal foraging cycle resembling that of otariid seals in a small phocid, the harbor seal. Behav. Ecol. Sociobiol. 34: 95–104. Bowen, W. D. and Harrison, G. 1996. Comparison of harbour seal diets in two habitats of Atlantic Canada. Can. J. Zool. 74: 125–135. Bowen, W. D., D. Ellis, S. J. Iverson, and D. J. Boness. 2003. Maternal and newborn lifehistory traits during periods of contrasting population trends: implications for explaining the decline of harbour seals, Phoca vitulina, on Sable Island. J. Zool. London 261: 155–163. Bowen, W. D., D. J. Boness, and S. J. Iverson. 1999. Diving behaviour of lactating harbour seal females and their pups during maternal foraging trips. Can. J. Zool. 77: 978–988. Bowen, W. D., D. Tully, D. J. Boness, B. M. Bulheier, and G. J. Marshall. 2002. Preydependent foraging tactics and prey profitability in a marine mammal Mar. Ecol. Prog. Ser. 244: 235–245. Bowen, W. D., O. T. Oftedal, and D. J. Boness. 1992. Mass and energy transfer during lactation in a small phocid, the harbour seal. Physiol. Zool. 65: 844–866. Bowen, W. D., S. J. Iverson, D. J. Boness, and O. T. Oftedal. 2001. Foraging effort, food intake and lactation performance depend on maternal mass in a small phocid seal. Functional Ecol. 15: 325–334. Bower, A. 1991. A simple kinematic mechanism for mixing fluid parcels across a meandering jet. J. Phys. Oceanog. 21: 173–180. Boyd, I. L. 1996. Individual variation in the duration of pregnancy and birth date in Antarctic fur seals: The role of environment, age, and sex of fetus. J. Mamm. 77: 124–133.
148
Long-Term Ecological Change in the Northern Gulf of Alaska
Boyd, I. L. 2002. Pinniped life history. In: Perrin, W. F., Wursig, B. and Thewissen, H. G. M. (Eds.), Encyclopedia of Marine Mammals. Academic Press, San Diego, pp. 929–934. Boyd, I. L., C. Lockyear, and H. D. Marsh. 1999. Reproduction in marine mammals. In: J. E. Reynolds and S. A. Rommel (Eds.), Biology of Marine Mammals. Washington, DC, Smithsonian Institution Press, pp. 218–286. Braun, B. M. and G. L. Hunt. 1983. Brood reduction in black-legged kittiwakes. The Auk 100: 469–476. Brodeur, R. D., K. W. Myers, and J. H. Helle. 2003. Research conducted by the United States on the early ocean life history of Pacific salmon. N. Pac. Anadr. Fish. Comm. Bull. 3: 89–131. Broecker, W. S. 1991. The great ocean conveyor. Oceanography 4: 79–89. Brower, Jr., W. A., R. G. Baldwin, Jr., C. N. Williams, J. L. Wise, and L. D. Leslie. 1988. Climate Atlas of the Outer Continental Shelf Waters and Coastal Regions of Alaska, vol. 1, Gulf of Alaska. Asheville, NC: National Climatic Data Center. Brown, E. D. 2003. Stock structure and environmental effects on year class formation and population trends on Pacific herring, Clupea pallasi, in Prince William Sound, Alaska. PhD Dissertation, University of Alaska Fairbanks, Fairbanks, Alaska. Bryant, R., I. L. Jones, and J. M. Hipfner. 1999. Responses to changes in prey availability by common murres and thick-billed murres at the Gannet Islands, Labrador. Can. J. Zool. 77: 1278–1287. Burger, A. E. 1991. Maximum diving depths and underwater foraging in alcids and penguins. In: W.A. Montevecci and A. J. Gaston (Eds.), Studies of High-Latitude Seabirds. 1. Behavioral, Energetic and Oceanographic Aspects of Seabird Feeding Ecology. Occasional Papers, 68. Canadian Wildlife Service, Ottowa, ON, pp. 9–15. Burger, A. and J. F. Piatt. 1990. Flexible time budgets in breeding common murres: buffers against variable prey. Stud. Avian Biol. 14: 71–83. Burns, J. J. 2002. Harbor seal and spotted seal. In: Perrin, W. F., Wursig, B. and Thewissen, H. G. M. (Eds.), Encyclopedia of Marine Mammals. Academic Press, San Diego, pp. 552–560. Burrell, D. C. 1984. The biochemistry of Boca de Quadra and Smeaton Bay, southeast Alaska: summary report. Prepared for U.S. Borax and Chemical Corporation and Pacific Coast Molybdenum Company by the Institute of Marine Science, University of Alaska Fairbanks, Fairbanks, AK. 280 pp. Burrell, D. C. 1986. Interaction between silled fjords and coastal regions, In: D. W. Hood and S. T. Zimmerman (Eds.), Physical Environment and Biological Resources. The Gulf of Alaska. U.S. Minerals Management Service, pp. 187–220. Businger, S. A. 1991. Arctic Hurricanes. Amer. Scientist. 79: 18–33. Cairns, D. K. 1987. Seabirds as indicators of marine food supplies. Biol. Oceanog. 5: 261–271. Cairns, D. K., W. A. Montevecci, V. L. Birt-Friesen, and S. A. Macko. 1990. Energy expenditures, activity budgets and prey harvest of breeding Common murres. Stud. Avian Biol. 14: 84–92. Calder, W. A. 1984. Size, function and life history. Cambridge: Harvard University Press. Calkins, D. and E. Goodwin. 1988. Investigation of the decline of Steller sea lions in the Gulf of Alaska, Alaska Department of Fish and Game. Calkins, D. G. 1986. Marine mammals, pp. 527–595. In: D. W. Hood and S. T. Zimmerman (Eds.), The Gulf of Alaska. Physical Environment and Biological Resources. U.S. Dept. Commerce, Washington, DC.
Ecosystem Structure 149
Cam, E., J. E. Hines, J.-Y. Monnat, J. D. Nichols, and E. Danchin. 1998. Are adult nonbreeders prudent parents? The kittiwake model. Ecology 79: 2917–2930. Cam, E., W. A. Link, E. G. Cooch, J. Monnat, and E. Danchin. 2002. Individual covariation in life-history traits: Seeing the trees despite the forest. Amer. Natural. 159: 96–105. Canino, M. F., K. M. Bailey, and L. S. Incze. 1991. Temporal and geographic differences in feeding and nutritional condition of walleye pollock larvae Theragra chalcogramma in Shelikof Strait, Gulf of Alaska. Mar. Ecol. Prog. Ser. 79: 27–35. Carscadden, J. E., K. T. Frank, and W. C. Leggett. 2001. Ecosystem changes and the effects on capelin (Mallotus villosus), a major forage species. Can. J. Fish. Aquat. Sci. 58: 73–85. Carter, H. R., U. W. Wilson, R. W. Lowe, M. S. Rodway, D. A. Mauwal, J. E. Takekawa, and J. L. Yee. 2001. Population trends of the common murre (Uria aalge californica). In: D. A. Manuwal, H. R. Carter, T. S. Zimmerman and D. L. Orthmeyer (Eds.), Biology and Conservation of the Common Murre in California, Oregon, Washington and British Columbia. Washington, DC: U.S. Geological Survey, Biological Resources Division, Information and Technology Report, pp. 33–129. Cenedese, C. 2002. Eddy shedding from a boundary current around a cape and over a sloping bottom. J. Phys. Oceanog. 30: 1514–1531. Chapman, D. C. 2000. A numerical study of the adjustment of a narrow stratified current over a sloping bottom. J. Phys. Oceanog. 30: 2927–2940. Chapman, D. C. and S. J. Lentz. 1994. Trapping of a coastal density front by the bottom boundary layer. J. Phys. Oceanog. 24: 1464–1479. Charnov, E. L. and P. J. Anderson. 1989. Sex change and population fluctuations in Pandalid shrimp. Amer. Natural. 134: 824–827. Childers, A., T. E. Whitledge, and D. A. Stockwell. 2005. Seasonal and interannual variability in the distribution of nutrients and chlorophyll a across the Gulf of Alaska shelf: 1998–2000. Deep-Sea Res. 52: 193–216. Ciannelli, L., K.S. Chan , K.M. Bailey, and N.C. Stenseth. 2004. Non-additive effects of environmental variables on the survival of a large marine fish population. Ecology 85: 3418–3427. Clutton-Brock T.H. 1991. The Evolution of Parental Care. Princeton University Press, Princeton. Coltman, D., W. D. Bowen, S. J. Iverson, and D. J. Boness. 1998. The energetics of male reproduction in an aquatically mating pinniped: the harbour seal. Physiol. Zool. 71: 387–399. Cooney, R. T. 1988. Distribution and ecology of zooplankton in the Gulf of Alaska. Bull. Ocn. Res. Inst. Univ. Tokoyo, 26(1): 27–42. Cooney, R. T., J. R. Allen, M. A. Bishop, D. L. Eslinger, T. Kline, B. L. Norcross, C. P. McRoy, J. Milton, J. Olsen, V. Patrick, A. J. Paul, D. Salmon, D. Scheel, G. L. Thomas, S. L. Vaughan, and T. M. Willette. 2001. Ecosystem controls of juvenile pink salmon (Oncorhynchus gorbuscha) and Pacific herring (Clupea pallasi) populations in Prince William Sound, Alaska. Fish. Oceanogr. 10(Suppl. 1): 1–13. Cooney, R. T. and K. O. Coyle. 1988. Water column production. In: D. G. Shaw and M. J. Hameedi (Eds.), Lecture Notes on Coastal and Estuarine Studies. Springer-Verlag, New York, pp. 93–115. Cooney, R. T., K. O. Coyle, E. Stockmar, and C. Stark. 2001. Seasonality in surface-layer net zooplankton communities in Prince William Sound, Alaska. Fish. Oceanogr. 10(Suppl. 1): 97–109.
150
Long-Term Ecological Change in the Northern Gulf of Alaska
Cooney, R. T., T. M. Willette, S. Sharr, D. Sharp, and J. Olsen. 1995. The effect of climate on North Pacific pink salmon (Oncorhynchus gorbuscha) production: examining some details of a natural experiment. In: R. J. Beamish (Ed.), Climate Change and Northern Fish Populations. Can. Spec. Publ. Fish. Aquat. Sci. 121: 475–482. Costa, D. P. 1982. Energy, nitrogen, and electrolyte flux and sea water drinking in the sea otter Enhydra lutris. Physiol. Zool. 55: 35–44. Costa, D. P. 1993. The relationship between reproductive and foraging energetics and the evolution of the Pinnipedia. Symp. Zool. Soci. London 66: 293–314. Costa, D. P. and G. L. Kooyman. 1984. Contribution of specific dynamic action to heat balance and thermoregulation in the sea otter, Enhydra lutris. Canad. J. Zool. 57: 199–203. Costa, D. P. and T. M. Williams. 1999. Marine mammal energetics. In: J. E. Reynolds and S. A. Rommel (Eds.), Biology of Marine Mammals. Washington, DC, Smithsonian Institution Press, pp. 176–217. Coulson, J. C. 2002. Why do adult kittiwakes survive so long but breed so poorly in the Pacific? J. Avian Biol. 33, 111–112. Coulson, J. C. and C. S. Thomas. 1985. Changes in the biology of the kittiwake Rissa tridactyla: A 31-year study of a breeding colony. J. Anim. Ecol. 54, 9–26. Coulson, J. C. and T. H. Pearson. 1985. Food intake and weight increments of the common guillemot Uria aalge: the use of conversion rate. Ibis 127: 565–566. Coyle, K. O. and A. I. Pinchuk. 2003. Annual cycle of zooplankton abundance, biomass, and production on the northern Gulf of Alaska shelf, October 1997 through October 2000. Fish. Oceanog. 12: 327–338. Coyle, K. O. and A. I. Pinchuk. Seasonal cross-shelf distribution of major zooplankton taxa on the northern Gulf of Alaska shelf relative to water mass properties, species depth preferences and vertical migration behavior. Deep-Sea Res. (accepted). Cramp, S. and K. E. L. Simmons. 1983. The Birds of the Western Palearctic, Vol 3. Waders to Gulls. Oxford: Oxford University Press. Crawford, W. R. 1984. Energy flux and generation of diurnal shelf waves along Vancouver Island. J. Phys. Oceanog. 14: 1600–1607. Crawford, W. R. and R. E. Thomson. 1984. Diurnal period shelf waves along Vancouver Island: a comparison of observations with theoretical models. J. Phys. Oceanog. 14: 1629–164. Crawford, W. R., J. Y. Cherniawsky, and M. G. G. Foreman. 2000. Multiyear meanders and eddies in the Alaskan Stream as observed in TOPEX/Poseidon altimeter. Geophys. Res. Lttrs. 27: 1025–1028. Csanady, G. T. 1982. Circulation in the Coastal Ocean, D. Reidel, Boston, 279 pp. Csanady, G. T. 1984. The influence of wind stress and river runoff on a shelf–sea front. J. Phys. Oceanog. 14: 1383–1392. Cummins, P. F. and L.-Y. Oey. 2000. Simulation of barotropic and baroclinic tides off northern British Columbia. J. Phys. Oceanog. 27: 762–781. Cury, P., A. Bakun, J. M. Crawford, A. Jarre, R. A. Quinones, L. J. Shannon, and H. M. Verheye. 2000. Small pelagics in upwelling systems: patterns of interaction and structural changes in “wasp-waist” ecosystems. ICES J. Mar. Sci. 57: 603–618. Dagg, M. 1993. Grazing by the copepod community does not control phytoplankton production in the subarctic Pacific Ocean. Prog. Oceangr. 32: 163–183. Danchin, E., T. Boulinier, and M. Massot. 1998. Conspecific reproductive success and breeding habitat selection: implications for the study of coloniality. Ecology. 79: 2415–2428.
Ecosystem Structure 151
Daunt, F., S. Benvenuti, M. P. Harris, L. Dall’Antonia, D. A. Elston, and S. Wanless. 2002. Foraging strategies of the black-legged kittiwake Rissa tridactyla at a North Sea colony: evidence for a maximum foraging range. Mar. Ecol. Prog. Ser. 245: 239–247. Davis, R. W., T. C. Adams, E. A. Brandon, D. C. Calkins, and T. R. Loughlin. 2002. Female attendance, lactation, and pup growth in Steller sea lions. In: D. DeMaster and S. Atkinson (Eds.), Steller Sea Lion Decline: Is It Food II? Fairbanks, AK, University of Alaska Sea Grant, AK-SG-02–02, pp. 23–27. Davoren, G. K. and W. A. Montevecchi. 2003a. Consequences of foraging trip duration on provisioning behavior and fledging condition of common murres Uria algae. J. Avian Biol. 34: 44–53. Davoren, G. K. and W. A. Montevecchi. 2003b. Signals from seabirds indicate changing biology of capelin stocks. Mar. Ecol. Prog. Ser. 258: 253–261. Davoren, G. K., W. A Montevecci, and J. T. Anderson. 2003a. Distributional patterns of a marine bird and its prey: habitat selection based on prey and conspecific behavior. Mar. Ecol. Prog. Ser. 256: 229–242. Davoren, G. K., W. A. Montevecci, and J. T. Anderson. 2003b. The influence of fish behaviour on search strategies of common murres Uria aalge in the northwest Atlantic. Mar. Ornithol. 31: 123–131. Dean, T. A., J. L. Bodkin, A. K. Fukuyama, S. C. Jewett, D. H. Monson, C. E. O’Clair, and G. R. VanBlaricom. 2002. Food limitation and the recovery of sea otters in Prince William Sound, Enhydra lutris. Mar. Ecol. Prog. Ser. 241: 255–270. DeGange, A. R. and G. A. Sanger. 1986. Marine birds. In: D. W. Hood and S. T. Zimmerman (Eds.), The Gulf of Alaska. Physical Environment and Biological Resources. U.S. Dept. Commerce, Washington, D.C., pp. 479–524. Dew, C. B. 1990. Behavioral ecology of podding red king crab, Paralithodes camtschatica. Can. J. Fish. Aquat. Sci. 47: 1944–1958. Dodimead, A. J., F. Favorite, and T. Hirano. Salmon of the North Pacific Ocean. 1963. Part II. Review of oceanography of the subarctic Pacific region. Internat. North Pac. Fish. Comm. Bull. 13: 1–195. Duggins, D. O., C. A. Simenstad, and J. A. Estes. 1989. Magnification of secondary production by kelp detritus in coastal marine ecosystems. Science 245: 170–173. Dwyer, D. A., K. M. Bailey, and P. A. Livingston. 1987. Feeding habits and daily ration of walleye pollock (Theragra chalcogramma) in the eastern Bering Sea, with special reference to cannibalism. Can. J. Fish. Aquat. Sci. 44: 1972–1984. Ellis, S. L., W. D. Bowen, D. J. Boness, and S. J. Iverson. 2000 . Maternal effects on offspring mass and stage of development at birth in the harbour seal, Phoca vitulina. J. Mamm. 81: 1143–1156. Emery, W. J., T. C. Royer, and R. W. Reynolds. 1985. The anomalous tracks of North Pacific drifting buoys 1981 to 1983. Deep-Sea Res. 32: 315–347. Emile-Geay, J., M. A. Cane, N. Naik, R. Seager, A. C. Clement, and S. A. van Geen. 2002. Warren revisited: atmosphere freshwater fluxes and “Why is no deep water formed in the North Pacific”, J. Geophys. Res. 108: 3178. Erikstad, K. E., M. Asheim, P. Fauchald, L. Dalhaug, and T. Tveraa. 1997. Adjustment of parental effort in the puffin; the roles of adult body condition and chick size. Behav. Ecol. Sociobiol. 40: 95–100. Eslinger, D. L., R. T. Cooney, C. P. McRoy, A. Ward, T. C. Kline, E. P. Simpson, J. Wang, and J. R. Allen. 2001. Plankton dynamics: observed and modeled responses to physical conditions in Prince William Sound, Alaska. Fish. Oceanogr. 10(Suppl. 1): 81–96.
152
Long-Term Ecological Change in the Northern Gulf of Alaska
Estes, J. A. and J. L. Bodkin 2002. Marine otters. In: W. F. Perrin, B. Wursig, J. G. M. Thewissen, and C. R. Crumly (Eds.), Encyclopedia of Marine Mammals. Academic Press, pp. 843–858. Estes, J. A. and J. L. Bodkin. 2002. Otters. In: W. F. Perrin, B. Wursig and J. G. M. Thewissen (Eds.), Encyclopedia of Marine Mammals. New York, Academic Press, pp. 842–858. Estes, J. A., M. L. Reiman, M. M. Staedler, M. T. Tinker, and B. E. Lyon. 2003. Individual variation in prey selection by sea otters: Patterns, causes and implications. J. Anim. Ecol. 72: 144–155. Estes, J. A., M. T. Tinker, T. M. Williams, and D. F. Doak. 1998. Killer whale predation on sea otters linking oceanic and nearshore ecosystems. Science 282: 473–476. Farmer, D. M. and J. D. Smith. 1979. Generation of lee waves over the sill in Knight Inlet, pp. 259–270. In: H. J. Freelan, D. M. Farmer, and C. D. Levings (Eds.), Fjord Oceanography. NATO Conference on Fjord Oceanography, Victoria, BC, Plenum Press, New York, 715 pp. Favorite, F., A. J. Dodimead, and K. Nasu. 1976. Oceanography of the subarctic Pacific region, Intern. North Pacific Fish. Comm. Bull. 33: 1–187. Feder, H. M. and A. J. Paul. 1980. Food of the king crab, Paralithodes camtschatica and the Dungeness crab, Cancer magister in Cook Inlet, Alaska. Proceed. Nat. Shellfisheries Assoc. 70: 240–246. Felsenstein, J. 1985. Phylogenies and the comparative method. Amer. Natural. 125: 1–15. Fisheries Centre. 1998. In: T. A. Okey, and D. Pauley (Eds.), A trophic mass-balance model of Alaska’s Prince William Sound Ecosystem for the post-spill period 1994–1996. University of British Columbia, Draft Report. 128 pp. Flather, R. A. 1988. A numerical investigation of tides and diurnal-period continental shelf waves along Vancouver Island. J. Geophys. Res. 18: 115–139. Foreman, M. G. G., W. R. Crawford, J. Y. Cherniawsky, R. F. Henry, and M. R. Tarbotton. 2000. A high-resolution assimilating tidal model for the northeast Pacific Ocean. J. Geophys. Res. 105: 28629–28651. Foreman, M. G. G. and R. E. Thomson. 1997. Three-dimensional model simulations of tides and buoyancy currents along the west coast of Vancouver Island. J. Phys. Oceanog. 27: 1300–1325. Fox, W. W. 1972. Dynamics of exploited Pandalid shrimps and evaluation of management options. PhD Dissertation, University of Washington. 193 pp. Foy, R. J. and B. L. Norcross. 1999. Spatial and temporal in the diet of juvenile Pacific herring (Clupea pallasi) in Prince William Sound, Alaska. Can. J. Zool. 77: 697–706. Foy, R. J. and A. J. Paul. 1999. Winter feeding and changes in somatic energy content for age-0 Pacific herring in Prince William Sound, Alaska. Trans. Am. Fish. Soc. 128: 1193–1200. Frederiksen, M., M. P. Harris, and S. Wanless. 2005. Inter-population variation in demographic parameters: a neglected subject? OIKOS III: 209–214. Freeland, H. J. and D. M. Farmer. 1980. Circulation and energetics of a deep, strongly stratified inlet. Can. J. Fish. Aquat. Sci. 37: 1398–1410. Freeland, H. J. and K. L. Denman. 1982. Tographically controlled upwelling center off southern Vancouver Island, J. Mar. Res. 40: 1069–1093. Frost, K. J., M. A. Simpkins, and L. F. Lowry. 2001. Diving behavior of subadult and adult harbor seals in Prince William Sound, Alaska. Mar. Mamm. Sci. 17: 813–834.
Ecosystem Structure 153
Furness, R. W. and K. Camphuysen. 1997. Seabirds as monitors of the marine environment. ICES J. Mar. Sci. 54, 726–737. Garrett, C. J. R. and J. W. Loder. 1981. Dynamical Aspects of Shallow Sea Fronts, Phil. Trans.Roy. Soc. London A302: 563–581. Garshelis, D. L. and J. A. Garshelis. 1984. Movements and management of sea otters in Alaska. J. Wildl. Manag. 48: 665–678. Gaston, A. J. and I. L. Jones. 1998. The Auks. Bird Families of the World. New York: Oxford University Press. Gawarkiewicz, G. 1991. Linear stability models of shelfbreak fronts. J. Phys. Oceanog. 21, 471–488. Gawarkiewicz, G. and D. C. Chapman. 1992. The role of stratification in the formation and maintenance of shelf-break fronts. J. Phys. Oceanog. 22: 753–772. Gay III, S. M. and S. L. Vaughan. 2001. Seasonal hydrography and tidal currents of bays and fjords in Prince William Sound, Alaska. Fish. Oceanog.10(Suppl. 1): 159–193. Gelatt, T. S., D. B. Siniff, and J. A. Estes. 2002. Activity patterns and time budgets of the declining sea otter population at Amchitka Island, Alaska. J. Wildlife Manag. 66: 29–39. Gill, V. A. and S. A. Hatch. 2002. Components of productivity in black-legged kittiwakes Rissa tridactyla: Response to supplemental feeding. J. Avian Biol. 32: 113–126. Gill, V. A., S. A. Hatch, and R. B. Lanctot. 2002. Sensitivity of breeding parameters to food supply in black-legged kittiwakes Rissa tridactyla. Ibis 144: 268–283. Gjerdrum, C., A. M. Vallee, C. C. S. Clair, D. F. Bertram, J. L. Ryder, and G. S. Blackburn. 2003. Tufted puffin reproduction reveals ocean climate variability. Proceed. Nat. Acad. Sci. 100: 9377–9382. Goering, J. J., C. J. Patton, and W. E. Shiels. 1973. Primary Production, In: D. W. Hood, W. E. Shiels, and E. J. Kelley (Eds.), Environmental Studies of Port Valdez. Occasional Publ. 3, Inst. Mar. Sci., University of Alaska, Fairbanks, AK, pp. 251–279. Golet, G. H., J. A. Schmutz, D. B. Irons, and J. A. Estes. 2004. Determinants of reproductive costs in the long-lived black-legged kittiwake: A multi-year experiment. Ecol. Monogr. 74: 353–372. Golet, G., D. B. Irons, and J. A. Estes. 1998. Survival costs of chick rearing in blacklegged kittiwakes. J. Anim. Ecol. 67: 827–841. Hall A., B. McConnell, and R. Barker. 2001. Factors affecting first-year survival in greu seals and their implications for life history. J. Anim. Ecol. 70: 138–149. Hammill, M. O. 2003. Earless seals. In: W. F. Perin, B. Wursig and J. G. M. Thewissen (Eds.), Encyclopedia of Marine Mammals. New York, Academic Press, pp. 352–358. Hampton M. A., P. R. Carlson, H. J. Lee, and R. A. Feely. 1987. Geomorphology, sediment, and sedimentary processes, pp. 93–144. In: D. W. Hood and S. T. Zimmerman, (Eds.), The Gulf of Alaska, Physical Environment and Biological Resources. Alaska Office, Ocean Assessments Division, National Oceanic and Atmospheric Administration, U.S. Dept. Commerce, 655 pp. Harris, M. P. and S. Wanless. 1995. The food consumption of young common murres (Uria aalge) in the wild. Colonial Waterbirds 18: 209–213. Hatch, S. A. and G. A. Sanger. 1992. Puffins as samplers of juvenile pollock and other forage fish in the Gulf of Alaska. Mar. Ecol. Prog. Ser. 80: 1–14. Hatch, S. A., P. M. Meyers, D. M. Mulcahy, and D. C. Douglas. 1999. Seasonal movements and pelagic habitat use of murres and puffins determined by satellite telemetry. Condor 102: 145–154.
154
Long-Term Ecological Change in the Northern Gulf of Alaska
Hatchwell, B. J. 1991. An experimental study of the effects of timing of breeding on the reproductive success of common guillemots (Uria aalge). J. Anim. Ecol. 60: 721–736. Hay, D. E., R. Toresen, R. Stephensen, M. Thompson, R. Claytor, F. Funk, E. Evshina, J. Jakobsson, T. Kobayashi, I. McQuinn, G. Melvin, J. Molloy, N. Naumenko, K. T. Oda, R. Parmanne, M. Power, V. Radchenko, J. Schweigert, J. Simmonds, B. Sjostrand, D. K. Stevenson, R. Tanasichuk, Q. Tang, D. L. Watters, and J. Wheeler. 2001. Taking stock and inventory of world herring stocks in 2000, In: F. Funk, J. Blackburn, D. Hay, A. J. Paul, R. Stephenson, R. Toresen, and D. Witherell (Eds.), University of Alaska Sea Grant College Program; AK-SG-01–04, Fairbanks, Alaska, pp. 381–454. Heard, W. R. 1991. Life history of pink salmon (Oncorhynchus gorbuscha), pp. 121–230. In: C. Groot and E. Margolis (Eds.), Pacific Salmon Life Histories. University of British Columbia Press, Vancouver, BC. Heggie, D. T., D. W. Boisseau, and D. C. Burrell. 1977. Hydrography, nutrient chemistry and primary productivity of Resurrection Bay, Alaska, 1972–75. Inst. Mar. Sci. Rept. R77–2, University of Alaska Fairbanks. 111 pp. Helle, J. H. 1970. Biological characteristics of intertidal and freshwater spawning pink salmon at Olsen Creek, Prince William Sound, Alaska, 1962–1963. U.S. Fish. Wildl. Ser. Spec. Sci. Rep. Fish. 602. 19 pp. Hermann, A. J., D. B. Haidvogel, E. L. Dobbines, and P. J. Stabeno. 2002. Coupling global and regional circulation models in the coastal Gulf of Alaska. Prog. Oceanog. 53: 335–367. Hickey, B. M. 1997. The response of a steep-sided narrow canyon to strong wind forcing. J. Phys. Oceanog. 27: 697–726. Hickey, B. M., R. E. Thomson, H. Yiu, and P. H. Leblond. 1991. Velocity and temperature fluctuations in a buoyancy-driven current off Vancouver Island. J. Geophys. Res. 96: 10507–10538. Hobson, K. A., J. F. Piatt, and J. Pitocchelli. 1994. Using stable isotopes to determine seabird trophic relationships. J. Anim. Ecol. 63: 786–798. Hoover, A. A. 1988a. Harbor seal. In: J. A. Lentfer (Ed.), Selected Marine Mammals of Alaska. Washington, DC, Marine Mammal Commission, pp. 125–157. Hoover, A. A. 1988b. Steller sea lion. In: J. A. Lentfer (Ed.), Selected Marine Mammals of Alaska. Washington, DC, Marine Mammal Commission, pp. 159–193. Horne, E. P. W., Loder, J. W., Harrison, W. G., Mohn, R., Lewsi, M. R., Irwin, B., and Platt, T. 1989. Nitrate supply and demand at the Georges Bank tidal front. Scienia Marina 53: 145–158. Horner, R. A., L. S. Dick, and W. E. Shiels. 1973. Phytoplankton studies. In D. W. Hood, W. E. Shiels and E. J. Kelley (Eds.), Environmental Studies of Port Valdez. Occasional Publ. 3, Inst. Mar. Sci., University of Alaska, Fairbanks, Alaska, pp. 283–294. Houde, E. D. 1987. Fish early life dynamics and recruitment variability. Amer. Fish. Soc. Symp. 2: 17–29. Houghton, R. W., D. B. Olson, and P. J. Celone. 1986. Observation of an anticyclonic eddy near the continental shelf break south of New England. J. Phys. Oceanog. 16: 60–71. Hunt, G. L. and D. C. Schneider. 1987. Scale-dependent processes in the physical and biological environment of marine birds. In: J. P. Croxall (Ed.), Seabirds: Feeding Ecology and Role in Marine Ecosystems. Cambridge: Cambridge University Press, pp. 7–41.
Ecosystem Structure 155
Incze, L. S., A. W. Kendall, Jr., J. D. Schumacher, and R. K. Reed. 1989. Interactions of a mesoscale patch of larval fish (Theragra chalcogramma) with the Alaska Coastal Current, Continent. Shelf Res. 9: 269–284. Irons, D. B. 1992. Aspects of foraging behavior and reproductive biology of the blacklegged kittiwake. PhD Dissertation, University of California, Irvine. Irons, D. B. 1998. Foraging area fidelity of individual seabirds in relation to tidal cycles and flock feeding. Ecology 79: 647–655. Iverson, R. L., H. C. Curl, H. B. O’Connor, and K. Zakar. 1974. Summer phytoplankton blooms in Auke Bay, Alaska, driven by wind mixing of the water column. Limnol. Oceanogr. 19: 271–278. Iverson, S. J., Frost, K. J. and Lowry, L. L. 1997. Fatty acid signatures reveal fine scale structure of foraging distribution of harbor seals and their prey in Prince William Sound, Alaska. Mar. Ecol. Progr. Ser. 151: 255–271 Iverson, S. J., K. J. Frost, and S. L. C. Lang. 2003. High body energy stores and condition are linked with diet diversity in juvenile harbor seals in Prince William Sound (PWS), Alaska: new insights from quantitative fatty acid signature analysis (QFASA). 15th Biennial Conference on the Biology of Marine Mammals, Greensboro, NC. Jacob, K. H. 1987. Seismicity, tectonics, and geohazards of the Gulf of Alaska regions, pp. 145–186, In: D. W. Hood and S. T. Zimmerman (Eds.), The Gulf of Alaska, Physical Environment and Biological Resources. Alaska Office, Ocean Assessments Division, National Oceanic and Atmospheric Administration, U.S. Dept. Commerce, 655 pp. Jacobsen, K.-O. and K. E. Erikstad. 1995. An experimental study of the costs of reproduction in the kittiwake Rissa tridactya. Ecology 76: 1636–1642. Jenness, R, T. D. Williams, and R. J. Mullin. 1981. Composition of milk of the sea otter (Enhydra lutris). Comp. Biochem. Physiol. 70A: 375–379. Jewett, S. C. and H. M. Feder. 1982. Food and feeding habits of the king crab Paralithodes camtschatica near Kodiak Island, Alaska. Mar. Biol. 66: 243–250. Jewett, S. C. and H. M. Feder. 1983. Food of the Tanner crab Chionoecetes bairdi near Kodiak Island, Alaska. J. Crustacean Biol. 3: 196–207. Jodice, P., D. D. Roby, R. M. Suryan, D. B. Irons, A. M. Kaufman, K. R. Turco, and G. H. Visser. 2002. Variation in energy expenditure among black-legged kittiwakes: Effects of activity-specific metabolic rates and activity budgets. Physiol. Biochem. Zool. 76: 375–388. Johnsen, I., K. E. Erikstad, and B. E. Saether. 1994. Regulation of parental investment in a long-lived seabird, the puffin Fratercula arctica: An experiment. Oikos 71: 273–278. Johnson, W. R., T. C. Royer, and J. L. Luick. 1988. On the seasonal variability of the Alaska Coastal Current, J. Geophys. Res. 93: 12423–12437. Joyce, T. M, J. K. B. Bishop, and O. B. Brown. 1992. Observations of offshore shelf water transport induced by a Warm Core Ring, Deep-Sea Res. 39(Suppl. 1A): 97–113. Kenyon, K. W. 1969. The sea otter in the eastern Pacific Ocean. N. Am. Fauna 68: 1–352. Kildaw, S. D. 1998. Population status and patterns of distribution and productivity of kittiwakes on St. George Island. PhD thesis. Fairbanks: University of Alaska Fairbanks. Kirsch, J., G. L. Thomas, and R. T. Cooney. 2000. Acoustic estimates of zooplankton distributions in Prince William Sound, spring, 1996. Fish. Res. 47: 245–260. Kitaysky, A. S. 1996. Behavioral, physiological and reproductive responses of alcids (Alcidae) to variation in food availability. PhD Thesis. Irvine: University of California Irvine.
156
Long-Term Ecological Change in the Northern Gulf of Alaska
Kitaysky, A. 1999. Metabolic and developmental response of Alcid chicks to experimental variation in food intake. Physiol. and Biochem. Zool. 72: 462–473. Kitaysky, A. S. and E. G. Golubova. 2000. Climate change causes contrasting trends in reproductive performance of planktivorous and piscivorous alcids. J. Anim. Ecol. 69: 248–262. Kitaysky, A., G. L. Hunt, E. N. Flint, M. A. Rubega, and M. B. Decker. 2000. Resource allocation in breeding seabirds: responses to fluctuations in their food supply. Mar. Ecol. Prog. Ser. 206: 283–296. Kitaysky, A. S., E. V. Kitaiskaia, J. F. Piatt, and J. C. Wingfield. 2003. Benefits and costs of increased levels of corticosterone in seabird chicks. Horm. Behav. 43: 140–149. Kitaysky, A. S., J. C. Wingfield, and J. F. Piatt. 2000. Corticosterone facilitates begging and affects resource allocation in the black-legged kittiwake. Behav. Ecol. 12: 619–625. Klinck, J. M. 1996. Circulation near submarine canyons: A modeling study. J. Geophys. Res. 101: 1211–1223. Klinck, J. M., J. J. O’Brien, and H. Svandsen. 1981. A simple model of fjord and coastal circulation interaction. J. Phys. Oceanog. 11: 1612–1626. Kline, T. C. 1997. Confirming forage fish food web dependencies in Prince William Sound using natural stable isotope tracers, In: Forage Fishes in Marine Ecosystems. Proceedings of the International Symposium on the Role of Forage Fishes in Marine Ecosystems. Alaska Sea Grant College Rept. 97–01. University of Alaska Fairbanks, pp. 257–269. Kline, T. C. 1999. Temporal and spatial variability of 13C/12C and 15N/14N in pelagic biota of Prince William Sound, Alaska. Can. J. Fish. Aquat. Sci. 56(Suppl. 1): 94–117. Koblentz-Mishke, O. J., V. V. Volkovinsky, and J. G. Kabanova. 1970. Plankton primary production of the world ocean. In: W. S. Wooster (Ed.), Scientific Exploration of the South Pacific. U.S. National Academy of Science, Washington, D.C., pp. 183–193. Koblinsky, C. J., P. P. Niiler, and W. J. Schmitz, Jr. 1989. Observations of wind-forced deep ocean currents in the North Pacific. J. Geophys. Res. 94: 10773–10790. Kruse, G. H. 1993. Biological perspectives on crab management in Alaska. In: G. H. Kruse, D. M. Eggers, R. J. Marasco, C. Pautzke, and T. J. Quinn II (Eds.), Proceedings of the International Symposium on Management Strategies for Exploited Fish Populations, University of Alaska Sea Grant College Program Report 93-02, Fairbanks, pp. 355–384. Kuris, A. M., S. F. Blau, A. J. Paul, J. D. Shields, and D. E. Wickham. 1991. Infestation by brood symbionts and their impact on egg mortality of the red king crab, Paralithodes camtschatica, in Alaska: geographic and temporal variation. Can. J. Fish. Aquat. Sci. 48: 559–568. Lackmann, G. M. and J. E. Overland. 1989. Atmospheric structure and momentum balance during a gap wind event in Shelikof Strait, Alaska. Month. Weather Rev. 117: 1817–1833. Ladd, C., G. L. Hunt, Jr., C. W. Mordy, S. A. Salo, and P. J. Stabeno. Marine Environment of the Eastern and Central Aleutians, Fish. Oceanog. 14: 22–38. Laevastu, T. and F. Favorite. 1981. Holistic simulation models of shelf-seas ecosystems, In: A. R. Longhurst (Ed.), Analysis of Marine Ecosystems. Academic Press, New York, pp. 701–728. Lagerloef, G. 1983. Topographically controlled flow around a deep trough transecting the shelf off Kodiak Island, Alaska. J. Phys. Oceanog. 13: 139–146. Lang, S. L. C., Iverson, S. J., and Bowen, W. D. 2005. Individual variation in milk composition over lactation in harbour seals (Phoca vitulina) and potential consequences of intermittent attendance. Can. J. Zool. 83: 1525–1531.
Ecosystem Structure 157
Larrance, J. D. 1971. Primary productivity and related oceanographic data, subarctic Pacific region. NMFS Data Rept. 50. 113 p. Larrance, J. D. and A. J. Chester. 1979. Source, composition and flux of organic detritus in lower Cook Inlet. Outer Continental Shelf Environmental Assessment Program. Final Reports of Principal Investigators 46: 1–71. Larrance, J. D., D. A. Tennant, A. J. Chester, and P. A. Ruffio. 1977. Phytoplankton and primary productivity in the northeast Gulf of Alaska and lower Cook Inlet: final report, Research Unit 425. Environmental Assessment of the Alaskan Continental Shelf, Annual Reports of the Investigators for the year ending March 1977, 10: 1–136. Lasker, R. 1985. What limits clupeoid production? Can. J. Fish. Aquat. Sci., 42: 31–37. Le Boeuf, B. J., P. Morris, and J. Reiter. 1994. Juvenile survivorship of Northern elephant seals. In: B. J. Le Boeuf and R. M. Laws (Eds.), Elephant Seals: Population Ecology, Behavior and Physiology. University of California Press, Berkeley, pp. 121–136. Leis, J. M. and B. M. Carson-Ewart. 1997. In situ swimming speeds of the late pelagic larvae of some Indo-Pacific coral-reef fishes. Mar. Ecol. Prog. Ser. 159: 165–174. Livezey, B. C. 1988. Morphometrics of flightlessness in the Alcidae. The Auk 105: 681–698. Livingston, P. A. 1989. Interannual trends in Pacific cod, Gadus macrocephalus, predation on three commercially important crab species in the eastern Bering Sea. Fish. Bull. 87: 807–827. Livingston, P. A. 1993. The importance of predation by groundfish, marine mammals, and birds on walleye pollock, Theragra chalcogramma, and Pacific herring, Clupea pallasi, in the eastern Bering Sea. Mar. Ecol. Prog. Ser. 102: 205–215. Livingstone, D. and T. C. Royer. 1980. Observed surface winds at Middleton Island, Gulf of Alaska and their influence on ocean circulation. J. Phys. Oceanog. 10: 753–764. Loher, T., and D. A. Armstrong. 2005. Historical changes in the abundance and distribution of ovigerous red king crabs (Paralithodes camtschaticus) in Bristol Bay (Alaska), and potential relationship with bottom temperature. Fish. Oceanogr. 14: 292–306. Longhurst, A. 2002. Murphy’s law revisited: longevity as a factor in recruitment to fish populations. Fish. Res. 56: 125–131. Loughlin, T. R. 2002. Steller sea lion. In: Perrin, W. F., Wursig, B. and Thewissen, H. G. M. (Eds.), Encyclopedia of Marine Mammals. Academic Press, San Diego, pp. 1181–1185. Loughlin, T. R., A. S. Perlov, J. D. Baker, S. A. Blokhin, and A. G. Makhnyr. 1998. Diving behavior of adult female Steller sea lions in the Kurile Islands, Russia. Biol. Conser. 1: 21–31. Love, M. 1966. Probably More Than You Want to Know About the Fishes of the Pacific Coast, 2nd edition. Really Big Press, Santa Barbara, CA, 381 pp. Lowry, L. F., K. J. Frost, J. M. Ver Hoef, and R. A. DeLong. 2001. Movements of satellite-tagged subadult and adult harbor seals in Prince William Sound, Alaska. Mar. Mamm. Sci. 17: 835–861. Lunn, N. J. and I. L. Boyd. 1993. Effects of maternal age and condition on parturition and the perinatal period of Antarctic fur seals. J. Zool. London 229: 55–67. Lynde, M. V. 1986. The historical annotated landing (HAL) database: documentation of annual harvest of groundfish from the northeast Pacific and eastern Bering Sea from 1956–1980, NOAA Tech. Memo. NMFS/NWC2. Macklin, S. A., N. A. Bond, and J. P. Walker. 1990. Structure of a low-level jet over Lower Cook Inlet, Alaska. Month. Weather Rev. 116: 2568–2578.
158
Long-Term Ecological Change in the Northern Gulf of Alaska
Macklin, S. A., G. M. Lackmann, and J. Gray. 1988. Offshore-directed winds in the vicinity of Prince William Sound, Alaska. Month. Weather Rev. 116: 1289–1301. Macklin, S. A., J. E. Overland, and J.P. Walker. 1984. Low-level gap winds in Shelikof Strait. In: Proceedings of the Third Conference on Meteorology of the Coastal Zone, Miami, Amer. Meteorol. Soc., Boston, MA, pp. 97–102 Malloy, R. J. and G. F. Merrill. 1972. Vertical crustal movement on the sea floor, In: The Great Alaska Earthquake of 1964, vol. 6: Oceanography and Coastal Engineering, National Research Council, National Academy of Sciences, Washington, DC, pp. 252–265. Mann, K. H. and J. R. N. Lazier. 1996. Dynamics of Marine Ecosystems, BiologicalPhysical Interactions in the Oceans, 2nd edition, Blackwell Science, Inc., Cambridge, MA, 394 p. Martin, J. H. and S. Fitzwater. 1988. Iron deficiency limits phytoplankton growth in the northeast Pacific subarctic. Nature 331: 341–343. Marty, G. D., E. F. Freiburg, T. R. Meyers, J. Wilcock, T. B. Farver, and D. E. Hinton. 1998. Viral hemorrhagic septicemia virus, Ichthyophonus hoferi, and other causes of morbidity in Pacific herring Clupea pallasi spawning in Prince William Sound, Alaska, USA. Dis. Aquat. Org. 32: 15–40. McRoy, C. P. and J. J. Goering. 1974. Coastal ecosystems of Alaska. In: H. T. Odum, B. J. Copeland, and E. H. McMahan (Eds.), Coastal Ecological Systems of the United States. vol. 3, The Conservation Foundation, Washington, DC, pp. 124–145. Meier, M. F. 1984. Contribution of small glaciers in global sea level. Science 226: 1481–1421. Mellish, J. E. and S. J. Iverson. 2001. Blood metabolites as indicators of nutrient utilization in fasting, lactating phocid seals: does depletion of nutrient reserves terminate lactation? Can. J. Zool. 79: 303–311. Melsom. A., S. D. Meyers, H. E. Hurlburt, E. J. Metzger, and J. J. O’Brien. 1999. El Niño induced eddies in the Gulf of Alaska, Earth Interactions, 3 [Available online at http:/EarthInteractions.org]. Mendenhall, V.M. 1997. Preliminary report on the 1997 Alaska seabird die-off. Unpublished report, U.S. Fish and Wildlife Serrvice, Anchorage, AK. Merrick, R. L. and D. G. Calkins. 1996. Importance of juvenile walleye pollock, Theragra chalcogramma, in the diet of Gulf of Alaska Steller sea lions, Eumetopias jubatus. U.S. Dep. Comm. NOAA Tech. Rep., NMFS 126: 153–166. Merrick, R. L. and T. R. Loughlin. 1997. Foraging behavior of adult female and youngof-the-year Steller sea lions in Alaskan waters Can. J. Zool. 75: 776–786. Merrick, R. L., M. K. Chumbley, and G. V. Byrd. 1997. Diet diversity of Steller sea lions (Eumetopias jubatus) and their population decline in Alaska: a potential relationship. Can J. Fish. Aquatic Sci. 54: 1342–1348. Meyers, S. D. and S. Basu. 1999. Eddies in the eastern Gulf of Alaska from TOPEX/POSEIDON altimetry. J. Geophys. Res. 104: 13333–13343. Meyers, T. R., T. M. Koeneman, C. Botelho, and S. Short. 1987. Bitter crab disease: A fatal dinoflagellate infection and marketing problem for Alaskan Tanner crabs Chionoecetes bairdi. Diseases Aquat. Org. 3: 195–216. Miller, C. B. 1993. Pelagic production processes in the subarctic Pacific. Prog. Oceanogr. 32: 1–15. Monson, D. H., J. A. Estes, J. L. Bodkin, and D. B. Siniff. 2000. Life history plasticity and population regulation in sea otters. Oikos 90: 457–468.
Ecosystem Structure 159
Montevecchi, W. A. 1993. Birds as indicators of change in marine prey stocks. In R. W. Furness and J. D. Greenwood (Eds.), Birds as Monitors of Environmental Change. London, Chapman and Hall, London, pp. 217–266. Muelbert, M. M. C. and W. D. Bowen. 1993. Duration of lactation and postweaning changes in mass and body composition of harbour seal pups. Can. J. Zool. 71: 1405–1414. Muelbert, M. M. C., W. D. Bowen, and S. J. Iverson. 2003. Weaning mass affects changes in body composition and food intake in harbour seal pups during the first month of independence. Physiol. Biochem. Zool. 76: 418–427. Muench, R. D. and D. T. Heggie. 1978. Deep water exchange in Alaskan subarctic fjords, In: B. Kjerfve (Ed.), Estuarine Transport Processes. B. Baruch Institute for Marine Biology and Coastal Research, University of South Carolina Press, Columbia, SC, pp. 239–267. Murphy, E. C. 1995. Seasonal declines in duration of incubation and chick periods of common murres at Bluff, Alaska in 1987–1991. The Auk 112: 982–993. Murphy, E. C. and J. H. Schauer. 1996. Synchrony in egg-laying and reproductive success of neighboring common murres, Uria aalge. Behav. Ecol. Sociobiol. 245–258. Murphy, E. C., A. M. Springer, and D. G. Roseneau. 1991. High annual variability in reproductive success of kittiwakes (Rissa tridactyla) at a colony in Western Alaska. J. Anim. Ecol. 60: 515–534. Musgrave, D., T. Weingartner, and T. C. Royer. 1992. Circulation and hydrography in the northwestern Gulf of Alaska. Deep-Sea Res. 39: 1499–1519. Mysak, L., R. D. Muench, and J. D. Schumacher. 1981. Baroclinic instability in a downstream varying channel: Shelikof Strait, Alaska. J. Phys. Oceanog. 11: 950–969. Napp, J. M., L. S. Incze, P. B. Ortner, D. L. W. Siefert, and L. Britt. 1996. The plankton of Shelikof Strait, Alaska: standing stock, production, mesoscale variability and their relevance to larval fish survival. Fish. Oceanogr. 5(Suppl. 1): 19–38. Niebauer, H. J., J. Roberts, and T. C. Royer. 1981. Shelf break circulation in the northern Gulf of Alaska J. Geophys. Res. 86: 13041–13047. Niebauer, H. J., T. C. Royer, and T. J. Weingartner. 1994. Circulation of Prince William Sound, Alaska. J. Geophys. Res. 99: 14113–14126. Norcross, B. L. and E. D. Brown. 2001. Estimation of first-year survival of Pacific herring from a review of recent stage-specific studies, In: F. Funk, J. Blackburn, D. Hay, A. J. Paul, R. Stephenson, R. Toresen, and D. Witherell (Eds.), Herring: Expectations for a New Millennium. University of Alaska Sea Grant College Program; AK-SG-01–04, Fairbanks, Alaska, pp. 535–558. Norcross, B. L., E. D. Brown, R. J. Foy, M. Frandsen, S. M. Gay, T. C. Kline, D. M. Mason, E. V. Patrick, A. J. Paul, and K. D. E. Stokesbury. 2001. A synthesis of the life history and ecology of juvenile Pacific herring in Prince William Sound, Alaska. Fish. Oceanogr. 10(Suppl. 1): 42–57. NRC (National Research Council) 2003. Decline of the Steller sea lion in Alaskan waters: Untangling food webs and fishing nets. Committee on the Alaska Groundfish Fishery and Steller Sea Lions, National Academic Press, Washington, DC. Nunes, P. 1984. Reproductive and larval biology of northern shrimp, Pandalus borealis Kr¯yer, in relation to temperature. PhD dissertation, University of Alaska, Fairbanks. 195 pp. OCSEAP, staff. 1986. Marine fisheries: resources and environments. In: D. W. Hood and S. T. Zimmerman (Eds.), The Gulf of Alaska. Physical Environment and Biological Resources. U.S. Dept. Commerce, Washington, D.C, pp. 417–458.
160
Long-Term Ecological Change in the Northern Gulf of Alaska
Ohtani, K., H. Onishi, N. Kobayashi, and G. Amma. 1997. Baroclinic flow referred to the 3000m reference level across the 180deg transect in the subarctic N. Pacific., Bull/ Faculty Fish., Hokkaido University, 48: 53–64. Okkonen, S. R. 1992. The shedding of an anticyclonic eddy from the Alaska Stream as observed by the GEOSAT altimeter. Geophys. Res. Letter 19: 2397–2400. Okkonen, S. R., T. J. Weingartner, S. L. Danielson, D. L. Musgrave, and G. M. Scmidt. 2003. Satellite and hydrographic observations of eddy-induced shelf-slope exchange in the northwestern Gulf of Alaska. J. Geophys. Res. 108: 3033. Olla, B. L., M. W. Davis, C. H. Ryer, and S. M. Sogard. 1996. Behavioral determinants of distribution and survival in early stages of walleye pollock, Theragra chalcogramma: A synthesis of experimental studies. Fish. Oceanogr. 5(Suppl. 1): 167–178. Orensanz, J. M., J. Armstrong, D. Armstrong, and R. Hilborn. 1998. Crustacean resources are vulnerable to serial depletion – the multifaceted decline of crab and shrimp fisheries in the greater Gulf of Alaska. Rev. Fish Biol. Fish. 8: 117–176. Oro, D. and R. W. Furness. 2002. Influences of food availability and predation on survival of kittiwakes. Ecology 83: 2516–2528. Ostrand, W. D., K. O. Coyle, and G. S. Drew. 1998. Selection of forage-fish schools by murrelets and tufted puffins in Prince William Sound, Alaska. The Condor 100: 286–297. Overland, J. E. and N. A. Bond. 1995. Observations and scale analysis of coastal wind jets. Month. Weather Rev. 123: 2934–2941. Overland, J. E. and T. R. Heister. 1980. Development of a synoptic climatology for the northeast Gulf of Alaska. J. Appl. Meteorol. 19: 1–14. Parish, T. R. 1982. Barrier wind along the Sierra Nevada mountains. J. Appl. Meteorol. 21: 925–930. Parker, D. G. 1997. A comparison of the feeding ecology and growth of juvenile pink salmon (Oncorhynchus gorbuscha) in northcentral and southwestern Prince William Sound, Alaska. MS Thesis, University of Alaska Fairbanks, Alaska. 176 pp. Parrish, J. K., M. Marvier, and R. T. Paine. 2001. Direct and indirect effects: interactions between bald eagles and common murres. Ecol. Applic. 11: 1858–1869. Parrish, J. K. and R. T. Paine. 1996. Ecological interactions and habitat modification in nesting common murres, Uria aalge. Bird Conserv. Internat. 6: 261–269. Parsons, T. R. 1976. Chapter 4. The structure of life in the sea. In: D. H. Cushing and J. J. Walsh (Eds.), The Ecology of the Seas. W.B. Saunders Company, Philadelphia. Parsons, T. R. 1986. Ecological relations. In: D. W. Hood and S. T. Zimmerman (Eds.), The Gulf of Alaska. Physical Environment and Biological Resources. U.S. Dept. Commerce, Washington, DC, pp. 561–570. Patrick, E. V. 2000. Evolution equation models for the advective transport during spring and the fasting physiology during winter for age-0 Pacific herring in Prince William Sound, Alaska. Results from projects J. U. T. and I. of the Sound Ecosystem Assessment Program. ISR Technical Report 2000–12, Institute for Systems Research, University of Maryland, College Park, MD. Paul, A. J. 1982. Distribution of juvenile Chionoecetes bairdi in Cook Inlet. In: Proceedings of the international symposium on the genus Chionoecetes. Alaska Sea Grant Report 82–10, Fairbanks, Alaska, pp. 173–180. Paul, A. J. 1984. Mating frequency and viability of stored sperm in the Tanner crab Chionoecetes bairdi (Decapoda, Majidae). J. Crustacean Biol. 4: 375–381.
Ecosystem Structure 161
Paul, A. J. 1997. The use of bioenergetic measurements to estimate prey consumption, nutritional status and thermal habitat requirements for marine organisms reared in the sea. Bull. Natl. Res. Inst. Aquacult Suppl. 3: 59–68. Paul, A. J. and J. M. Paul. 1980. The effect of early starvation on later feeding success of king crab zoea. J. Exper. Mar. Biol. Ecol. 44: 247–251. Paul, A. J. and J. M. Paul. 1998. Comparisons of whole body energy content of captive fasting age zero Alaskan Pacific herring (Clupea pallasi Valenciennes) and cohorts overwintering in nature. J. Exp. Mar. Biol. Ecol. 226: 75–86. Paul, A. J. and J. M. Paul. 1999. Interannual and regional variations in body length, weight and energy content of age-0 Pacific herring from Prince William Sound, Alaska. J. Fish. Biol. 54: 996–1001. Paul, A. J. and R. Smith. 1993. Seasonal changes in somatic energy content of yellowfin sole Pleuronectes asper Pallas. J. Fish. Biol. 43: 131–138. Paul, A. J., H. M. Feder, and S. C. Jewett. 1979. Food of the snow crab, Chionoecetes bairdi Rathbun, 1924, from Cook Inlet, Alaska (Decapoda, Majidae). Crustaceana Suppl. 5: 62–68. Paul, A. J., J. M. Paul and E. D. Brown. 1998. Fall and spring somatic energy content for Alaskan Pacific herring (Clupea pallasi Valenciennes 1847) relative to age, size and sex. J. Exp. Mar. Biol. Ecol. 223: 42. Paul, A. J., J. M. Paul, and W. E. Donaldson. 1995. Shell condition and breeding success in Tanner crabs. J. Crustacean Biol. 15: 476–480. Payne, S. F. and R. J. Jameson. 1984. Early behavioral development of the sea otter, Enhydra lutris. J. Mamm. 65: 527–531. Pennycuick, C. J. 1987. Flight of seabirds. In: J. P. Croxall (Ed.), Seabirds. Cambridge: Cambridge University Press, pp. 43–63. Piatt, J. F. 1990. The aggregative response of common murres and Atlantic puffins to schools of capelin. Stud. Avian Biol. 14: 36–51. Piatt, J. F. 2002. Response of seabirds to fluctuations in forage fish density: Can seabirds recover from effects of the Exxon Valdez oil spill? In: J. F. Piatt (Ed.), Final Report to Exxon Valdez Oil Spill Trustee Council (Restoration Project 00163M) and Minerals Management Service (Alaska OCS Region). Anchorage, Alaska: Alaska Science Center, U.S. Geological Survey, pp. 131–171. Piatt, J. F. 2004. Survival of adult murres and kittiwake in relation to forage fish abundance. In: Exxon Valdez Oil Spill Restoration Project Final Report (Restoration Project 00338). Anchorage, Alaska: U.S. Geological Survey. Piatt, J. F. and A. S. Kitaysky. 2002. Tufted puffin (Fratercula cirrhata). In: A. Poole and F. Gill (Ed.), The Birds of North America. vol. No. 708. Philadephia, PA: The Birds of North America, Inc. Piatt, J. F. and D. N. Nettleship. 1985. Diving depths of four alcids. The Auk 102: 293–297. Piatt, J. F., D. D. Roby, L. Henkel, and K. Neuman. 1997. Habitat use, diet and breeding biology of tufted puffins in Prince William Sound, Alaska. Northwest Natural 78: 102–109. Piatt, J.F. and T.I. van Pelt. 1997. Mass-mortality of guillemots (uria aalge) in the Gulf of Alaska in 1993. Mar. Pollut. Bull. 34: 656–662. Pickart, R. S. 2000. Bottom boundary layer structure and detachment in the shelfbreak jet of the Middle Atlantic Bight. J. Phys. Oceanogr. 30: 2668–2686. Pitcher, K. W. 1980. Food of the harbor seal, Phoca vitulina richardsi, in the Gulf of Alaska. Fish. Bull. 78: 544–549. Pitcher, K. W. 1981. Prey of the Steller sea lion, Eumetopias jubatus, in the Gulf of Alaska. Fish. Bull. 79: 467–472.
162
Long-Term Ecological Change in the Northern Gulf of Alaska
Pitcher, K. W. and D. C. McAllister. 1981. Movements and haulout behavior of radio-tagged harbor seals, Phoca vitulina. Can. Field-Naturalist 95: 292–297. Pitcher, K. W. and D. G. Calkins. 1981. Reproductive biology of Steller sea lions in the Gulf of Alaska. J. Mamm. 62: 599–605. Pitcher, K. W., D. G. Calkins, and G. W. Pendleton. 1998. Reproductive performance of female Steller sea lions: an energetics-based reproductive strategy? Can. J. Zool. 76: 2075–2083. Plafker, G. 1972. Tectonics, In: The Great Alaska Earthquake of 1964, vol. 6: Oceanography and Coastal Engineering, National Research Council, National Academy of Sciences, Washington, D.C., pp. 47–122. Plakhotnik, A. F. 1964. Hydrological description of the Gulf of Alaska. In: P. A. Moiseev (Ed.), Soviet Fisheries Investigations in the Northeast Pacific, Part II, 289 pp. Porter, S. M., L. Ciannelli, N. Hillgruber, K. M. Bailey, L. S. Chan, M. F. Canino, and L. J. Haldorson. 2005. Analysis of factors influencing larval walleye pollock Theragra chalcogramma feeding in Alaskan waters. Mar. Ecol. Prog. Ser. 302: 207–217. Powell, G. C. and R. B. Nickerson. 1965. Reproduction of king crabs, Paralithodes camtschatica (Tilesius). J. Fish. Res. Bd. Can. 22: 101–111. Powell, G. C., D. Pengilly, and S. F. Blau. 2002. Mating pairs of red king crabs (Paralithodes camtschaticus) in the Kodiak Archipelago, Alaska, 1960–1984. In: A. J. Paul, E. G. Dawe, R. Elner, G. S. Jamieson, G. H. Kruse, R. S. Otto, B. Sainte-Marie, T. C. Shirley, and D. Woodby (Eds.), Crabs in cold water regions: Biology, Management, and Economics. University of Alaska Sea Grant, AK-SG-02-01, pp. 225–245. Ramp, S., R. C. Beardsley, and R. Legeckis. 1983. An observation of frontal wave development on a shelf/slope warm-core ring front near the shelf break south of New England. J. Phys. Oceanogr. 13: 907–912. Raum-Suryan, K. L., K. W. Pitcher, D. G. Calkins, J. L. Sease, and T. R. Loughlin. 2002. Dispersal, rookery fidelity, and metapopulation structure of Steller sea lions (Eumetopias jubatus) in an increasing and a decreasing population in Alaska. Mar. Mamm. Sci. 18: 746–764. Rea, L. D., K. W. Pitcher, S. D. Farley, J. P. Richmond, and W. S. Dunlap-Harding. 2003. Percent total body lipid content increases in Steller sea lion (Eummetopias jubatus) pups throughout the first year of life in a similar pattern to other otariids pups. 15th Biennial Conference on the Biology of Marine Mammals, Greensboro, NC. Reed, R. K. 1984. Flow of the Alaskan Stream and its variations. Deep-Sea Res. 31: 369–386. Reed, R. K. and J. D. Schumacher. 1987. Physical Oceanography, In: D. W. Hood and S. T. Zimmerman (Eds.), The Gulf of Alaska, Physical Environment and Biological Resources. MMS/NOAA, Alaska Office, Anchorage, OCS Study MMS 86–0095, pp. 57–76. Reed, R. K., J. D. Schumacher, and L. S. Incze. 1987. Circulation in Shelikof Strait, Alaska. J. Phys. Oceanogr. 17: 1546–1554. Reed, R. K. and P. J. Stabeno. 1999. A recent full-depth survey of the Alaskan Stream. J. Phys. Oceanogr. 55: 79–85. Regehr, H. M. and W. A. Montevecchi. 1997. Interactive effects of food shortage and predation on breeding failure of black-legged kittiwakes: Indirect effects of fisheries activities and implications for indicator species. Mar. Ecol. Prog. Ser. 155: 249–260. Regehr, H. M., M. S. Rodway, and W. A. Montevecchi. 1998. Antipredator benefits of nest-site selection in black-legged kittiwakes. Can. J. Zool. 76: 910–915.
Ecosystem Structure 163
Reid, J. L. 1965. Intermediate Water of the Pacific Ocean, The Johns Hopkins Press, Baltimore. Reid, J. L. 1981. On the mid-depth circulation of the world ocean, In: B. A. Warren and C. Wunsch, (Eds.), Evolution of Physical Oceanography Scientific Surveys in Honor of Henry Stommel. MIT Press, Cambridge, MA, pp. 70–111. Rice, R. L., K. I. McCumby, and H. M. Feder. 1980. Food of Pandalus borealis, Pandalus hypsinotus and Pandalus goniurus (Pandalidae, Decapoda) from lower Cook Inlet, Alaska. Proceed. Nat. Shellfisheries Assoc. 70: 47–54. Ricklefs, R. E. 1990. Seabird life histories and the marine environment: Some speculations. Colonial Waterbirds 13: 1–6. Ricklefs, R. E. and J. M. Starck. 1998. Patterns of Development: the Altricial-Precocial Spectrum. In: J. M. Starck and R. E. Ricklefs (Eds.), Avian Growth and Development. New York: Oxford University Press, pp. 31–58. Riedman, M. L. and J. A. Estes. 1990. The sea otter (Enhydra lutris): Behavior, ecology and natural history. U.S. Fish and Wildlife Service Biological Report 90: 1–126. Robards, M. D., J. A. Anthony, G. A. Rose, and J. F. Piatt. 1999. Changes in proximate composition and somatic energy content for Pacific sand lance (Ammodytes hexapterus) from Kachemak Bay, Alaska relative to maturity and season. J. Exper. Mar. Biol. Ecol. 242: 245–258. Robards, M. D., J. F. Piatt, A. B. Kettle, and A. A. Brokkshire. 1999b. Temporal and geographic variation in fish communities of lower Cook Inlet, Alaska. Fish. Bull. 97: 962–977. Roberts, B. D. and S. A. Hatch. 1993. Behavioral ecology of black-legged kittiwakes during chick rearing in a failing colony. The Condor 95: 330–342. Roberts, B. D. and S. A. Hatch. 1994. Chick movements and adoption in a colony of black-legged kittiwakes. Wilson Bull. 106: 289–298. Roden, G. 1970. Aspects of the mid-Pacific transition zone. J. Geophys. Res. 75: 1097–1109. Rogers, D. E. 1986. Pacific salmon. In: D. W. Hood and S. T. Zimmerman (Eds.), The Gulf of Alaska. Physical Environment and Biological Resources. U.S. Dept. Commerce, Washington, D.C., pp. 461–476. Rogers, D. E., B. J. Rogers, and R. J. Rosenthal. 1987. The nearshore fishes. In: D. W. Hood and S. T. Zimmerman (Eds.), The Gulf of Alaska. Physical Environment and Biological Resources. U.S. Dept. Commerce, Washington, D.C., pp. 399–415. Romano, M. D. 1999. Effects of diet on growth and development of nestling seabirds. MS Thesis. Corvallis: Oregon State University. 59 pp. Rooper, C. N. 1996. Physical and biological factors affecting Pacific herring egg loss in Prince William Sound, Alaska. MS Thesis, University of Alaska Fairbanks, Juneau, AK. Rooper, C. N., L. J. Haldorson, and T. J. Quinn II. 1999. Habitat factors controlling Pacific herring (Clupea pallasi) egg loss in Prince William Sound, Alaska. Can. J. Fish. Aquat. Sci. 56: 1133–1142. Royer, T. C. 1975. Seasonal variations of waters in the northern Gulf of Alaska. Deep-Sea Res. 22: 403–416. Royer, T. C. 1979. On the effect of precipitation and runoff on coastal circulation in the Gulf of Alaska. J. Phys. Oceanogr. 9: 553–563. Royer, T. C. 1981a. Baroclinic transport in the Gulf of Alaska, Part II. Freshwater driven coastal current. J. Mar. Res. 39: 251–266. Royer, T. C. 1981b. Baroclinic transport in the Gulf of Alaska, Part I. Seasonal variations of the Alaska Current. J. Mar. Res. 39: 239–250.
164
Long-Term Ecological Change in the Northern Gulf of Alaska
Royer, T. C. 1982. Coastal freshwater discharge in the Northeast Pacific. J. Geophys. Res. 87: 2017–2021. Royer, T. C. 1998. Coastal Processes in the northern North Pacific, In: A. R. Robinson and K. H. Brink (Eds.), The Sea. John Wiley and Sons, NY, pp. 395–414. Royer, T. C., D. V. Hansen, and D. J. Pashinski. 1979. Coastal flow in the northern Gulf of Alaska as observed by dynamic topography and satellite-tracked drogued drift buoys J. Phys. Oceanogr. 9: 785–801. Royer, T. C., J. A. Vermisch, T. J. Weingartner, H. J. Niebauer, and R. D. Muench. 1990. Ocean circulation influencing the Exxon Valdez oil spill. Oceanography 3: 3–10. Ryther, J. H. 1969. Photosynthesis and fish production in the sea. Science 166: 72–76. Sambrotto, R. N. and C. J. Lorenzen. 1987. Phytoplankton and primary production. In: D. W. Hood and S. T. Zimmerman (Eds.), The Gulf of Alaska. Physical Environment and Biological Resources. US Minerals Management Service, pp. 249–282. Sambrotto, R. N., H. J. Neibauer, J. J. Goering, and R. L. Iverson. 1986. Relationships among vertical mixing, nitrate uptake, and phytoplankton growth during the spring bloom in the southeastern Bering Sea middle shelf. Cont. Shelf Res. 5: 161–198. Scheel, D. and K. R. Hough. 1997. Salmon fry predation by seabirds near an Alaskan hatchery. Mar. Ecol. Prog. Ser. 150: 35–48. Schmidt, G. M. 1977. The exchange of water between Prince William Sound and the Gulf of Alaska, MS thesis, 116 pp., University of Alaska, Fairbanks. Schneider, D. C., N. M. Harrison, and G. L. Hunt. 1990. Seabird diet at a front near the Pribilof Islands, Alaska. Stud. Avian Biol. 14: 61–66. Schultz, T. M. and W. D. Bowen. 2005. The evolution of lactation strategies in pinnipeds: A phylogenetic analysis. Ecol. Monogr. 75: 159–177. Schumacher, J. D. and A. W. Kendall, Jr. 1991. Some interactions between young walleye pollock and their physical environment in the western Gulf of Alaska. CalCOFI Rep. 32, Calif. Coop. Oceanic Fish. Invest., La Jolla, pp. 22–44. Schumacher, J. D., C. A. Pearson, and R. K. Reed. 1982. An exchange of water between the Gulf of Alaska and the Bering Sea through Unimak Pass. J. Geophys. Res. 87: 5785–5795. Schumacher, J. D., P. J. Stabeno, and S. J. Bograd. 1993. Characteristics of an eddy over a continental shelf: Shelikof Strait, Alaska. J. Geophys. Res. 98: 8395–8404. Schumacher, J. D., P. J. Stabeno, and A. T. Roach. 1989. Volume transport in the Alaska Coastal Current, Contin. Shelf Res. 9: 1071–1083. Sealy, S. G. 1973. Adaptive significance of post-hatching developmental patterns and growth rates in the Alcidae. Ornis Scandinavica 3: 113–121. Shima, M., A. B. Hollowed, and G. R. VanBlaricom. 2002. Changes over time in the spatial distribution of walleye pollock (Theragra chalcogramma) in the Gulf of Alaska, 1984–1996. Fish. Bull. 100: 307–323. Shirley, S. M. and T. C. Shirley. 1989. Interannual variability in density, timing and survival of Alaskan red king crab (Paralithodes camtschatica) larvae. Mar. Ecol. Progr. Ser. 54: 51–59. Shirley, T. C., S. M. Shirley, and S. Korn. 1990. Incubation period, molting and growth of female red king crabs: effects of temperature. In: Proceedings of the international symposium on king and Tanner crabs. Alaska Sea Grant Program Report 90-04, University of Alaska Fairbanks, pp. 51–63. Shultz, M. 2002. Response of seabirds to fluctuations in forage fish density: Black-legged kittiwakes. In: J. F. Piatt (Ed.), Final Report to Exxon Valdez Oil Spill Trustee Council (Restoration Project 00163M) and Minerals Management Service (Alaska OCS Region). Anchorage, Alaska: Alaska Science Center, U.S. Geological Survey, pp. 86–99.
Ecosystem Structure 165
Sinclair, E. H. and T. K. Zeppelin. 2002. Seasonal and spatial differences in diet in the western stock of Steller sea lions (Eumetopias jubatus). J. Mamm. 83: 973–990. Smith, R. L., A. J. Paul, and J. M. Paul. 1988. Aspects of energetics of adult walleye pollock, Theragra chalcogramma Pallas, from Alaska. J. Fish. Biol. 33: 445–454. Smith, R. L., A. J. Paul and J. M. Paul. 1990. Seasonal changes in energy and the energy cost of spawning in Gulf of Alaska Pacific cod. J. Fish. Biol. 36: 307–316. Sogard, S. M. and B. L. Olla. 1996. Food deprivation affects vertical distribution and activity of a marine fish in a thermal gradient: potential energy-conserving mechanisms. Mar. Ecol. Prog. Ser. 133: 43–55. Speckman, S. 2002. Response of seabirds to fluctuations in forage fish density: Pelagic seabird abundance and distribution in lower Cook Inlet. In: J. F. Piatt (Ed.), Final Report to Exxon Valdez Oil Spill Trustee Council (Restoration Project 00163M) and Minerals Management Service (Alaska OCS Region). Anchorage, Alaska: Alaska Science Center, U.S. Geological Survey, pp. 131–171. Springer, A. M. 1992. A review: Walleye pollock in the North Pacific. How much difference do they really make? Fish. Oceanog. 5: 205–223. Springer, A. M., J. F. Piatt, and G. V. Vliet. 1996. Sea birds as proxies of marine habitats and food webs in the western Aleutian arc. Fish. Oceanogr. 5: 45–55. Springer, A.M. and S.G. Speckman. 1997. A forage fish is what? Summary of the symposium. In: Forage fishes in marine ecosystems. University of Alaska Sea Grant Report 97–01, pp. 773–806. Stabeno, P. J, N. A. Bond, A. J. Hermann, N. N. Kachel, C. W. Mordy, and J. E. Overland. 2004. Meteorology and oceanography of the northern Gulf of Alaska. Contin. Shelf. Res. 24: 859–897. Stabeno, P. J., R. K. Reed, and J. D. Schumacher. 1995. The Alaska Coastal Current: continuity of transport and forcing. J. Geophys. Res. 100: 2477–2485. Starr, M., J.-C. Therriault, G. Y. Conan, M. Comeau, and G. Robichaud. 1994. Larval release in a sub-euphotic zone invertebrate triggered by sinking phytoplankton particles. J. Plankton Res. 16: 1137–1147. Stearns, S. 1992. The Evolution of Life Histories. Oxford: The Oxford University Press. Stearns, S. C. 1983. The influence of size and phylogeny on patterns of covariation among life-history traits in the mammals. Oikos 41:173–187. Stevens, B. G. 2003. Timing of aggregation and larval release by Tanner crabs, Chionoecetes bairdi, in relation to tidal current patterns. Fish. Res. 65: 201–216. Stevens, B. G., and J. Kittaka. 1998. Postlarval settling behavior, substrate preference, and time to metamorphosis for red king crab Paralithodes camtschaticus. Mar. Ecol. Prog. Ser. 167: 197–206. Stevens, B. G., J. A. Haaga, and W. E. Donaldson. 1999. Mound formation by Tanner crabs, Chionoecetes bairdi: Tidal phasing of larval launch pads?. In: J. C. von Voupel Klein and F. R. Schram (Eds.), Proceedings of the fourth international crustacean congress. A.A. Balkema. Rotterdam, The Netherlands, pp. 445–453. Stommel, H. and A. B. Arons. 1960a. On the abyssal circualtion of the world ocean – I. Stationary planetary flow patterns on a sphere. Deep-Sea Res. 6: 140–154. Stommel, H. and A. B. Arons. 1960b. On the abyssal circualtion of the world ocean – II. An idealized model of the circulation pattern and amplitude in oceanic basins. Deep-Sea Res. 6: 217–233. Stone, R. P., C. E. O’Clair, and T. C. Shirley. 1992. Seasonal migration and distribution of female red king crabs in a southeast Alaskan estuary. J. Crustacean Biol. 12: 546–560.
166
Long-Term Ecological Change in the Northern Gulf of Alaska
Stone, R.P., C.E. O’Clair, and T.C. Shirley. 1993. Aggregating behavior of ovigerous female red king crab, Paralithodes camtschaticus, in Auke Bay, Alaska. Can. J. Fish. Aquat. Sci. 50: 750–758. Strauch, J. G. J. 1985. The phylogeny of the Alcidae. The Auk 102: 520–539. Strom, S. L., Miller, C. B. and Frost, B. W. 2000. What sets lower limits on phytoplankton stocks in high-nitrate, low-chlorophyll regions of the open ocean? Mar. Ecol. Prog. Ser. 193: 19–31. Suryan, R. M., D. B. Irons, and J. Benson. 2000. Prey switching and variable foraging strategies of black-legged kittiwakes and the effects on reproductive success. The Condor 102: 374–384. Suryan, R. M., D. B. Irons, M. Kaufman, J. Benson, P. G. R. Jodice, D. D. Roby, and E. D. Brown. 2002. Short-term fluctuations in forage fish availability and the effect on prey selection and brood-rearing in the black-legged kittiwake Rissa tridactyla. Mar. Ecol. Progr. Ser. 236: 273–287. Sverdrup, H. U. 1953. On conditions for the vernal blooming of the phytoplankton. J. Cons. Inter. Explor. Mer. 18: 287–295. Tabata, S. 1982. The anticyclonic, baroclinic eddy of Sitka, Alaska, in the Northeast Pacific Ocean. J. Phys. Oceanogr. 1: 1260–1282. Tabata, S. 1991. Annual and interannual variability of baroclinic transports across Line P in the northeast Pacific Ocean. Deep-Sea Res. 38(Suppl 1A): 221–246. Taylor, S. G. 1988. Intra and inter-annual survival of pink salmon (Oncorhynchus gorbuscha) returning to Auke Creek, Alaska in 1986 and 1987. APPRISE Annual Report 1987. School of Fisheries and Ocean Sciences, University of Alaska Fairbanks, Alaska, Tech. Rep. 1: 547–571. Theilacker, G. H., K. M. Bailey, M. F. Canino, and S. M. Porter. 1996. Variations in larval walleye pollock feeding and conditions: a synthesis. Fish. Oceanog. 5(Suppl. 1): 112–123. Thomas, G. L. and R. E. Thorne. 2001. Night-time predation by Steller sea lions. Nature 411: 1013. Thomas, G. L. and R. E. Thorne. 2003. Acoustic-optical assessment of Pacific herring and their predator assemblage in Prince William Sound, Alaska. Aquat. Living Resour. 16: 247–253. Thompson, P. M., D. Miller, R. Cooper, and P. S. Hammond. 1994. Changes in the distribution and activity of female harbour seals during the breeding season: implications for their lactation strategy and mating patterns. J. Animal Ecol. 63: 24–30. Thompson, R. O. R. Y and T. J. Golding. 1981. Tidally induced ‘upwelling’ by the Great Barrier Reef. J. Geophys. Res. 86: 6517–6521. Thomson, R. E. 1972. On the Alaskan Stream. Ocean. J. Phys. Oceanogr. 2: 363–371. Thomson, R. E., Hickey, and P. H. LeBlond. 1989. The Vancouver Island Coastal Current: Fisheries barrier and conduit, In: R. J. Beamish and G. A. McFarlane (Eds.), Effects of Ocean Variability on Recruitment and an Evaluation of Parameters Used in Stock Assessment Models. Can. Spec. Public. Fish. Aquat. Sci. 108, pp. 265–296. Thomson, R. E., P. H. LeBlond, and W. J. Emery. 1990. Analysis of deep-drogued satellitetracked drifter measurements in the Northeast Pacific Ocean. Atmos. Ocean, 28: 409–443. Trumble, S. J., P. S. Barboza, and M. A. Castellini. 2003. Disgestive constraints on an aquatic carnivore: effects of feeding frequency and prey composition on harbor seals. J. Comp. Physiol. B 173: 501–509.
Ecosystem Structure 167
Thomson, R. E. and E. Wolanski. 1984. Tidal period upwelling with Raine Island Entrance Great barrier Reef. J. Mar. Res. 42: 787–808. Tully, J. P. and F. G. Barber. 1960. An estuarine analogy in the sub-arctic Pacific Ocean. J. Fish. Res. Bd Can. 17: 91–112. Tuomi, P. A. and T. M. Williams. 1995. Rehabilitation of pregnant sea otters and females with newborn pups. In: T. M. Williams and R. W. Davis, (Eds.), Emergency Care and Rehabilitation of Oiled Sea Otters. Fairbanks, AK, University of Alaska Press, pp. 121–132. University of Alaska Sea Grant. 2001. In: F. Funk, J. Blackburn, D. Hay, A. J. Paul, R. Stephenson, R. Toresen, and D. Witherell (Eds.), Herring: Expectations for a New Millennium. University of Alaska Sea Grant. AK-SG-01–04, Fairbanks, 800 pp. Van Pelt, T. I., J. F. Piatt, B. K. Lance, and D. D. Roby. 1997. Proximate composition and energy density of some North Pacific forage fishes. Comp. Biochem. Physiol. 118A: 1393–1398. van Pelt, T. I. and M. Shultz. 2002. Response of seabirds to fluctuations in forage fish density: Common murres. In: J. F. Piatt (Ed.), Final Report to Exxon Valdez Oil Spill Trustee Council (Restoration Project 00163M) and Minerals Management Service (Alaska OCS Region). Anchorage, Alaska: Alaska Science Center, U.S. Geological Survey, pp. 70–85. van Scoy, K. A., D. B. Olson, and R. A. Fine. 1991. Ventilation of North Pacific Intermediate Water: the role of the Alaskan Gyre. J. Geophys. Res. 96: 16801–16810. Vastano, A. C., L. S. Incze, and J. D. Schumacher. 1992. Environmental and larval pollock observations in Shelikof Strait, Alaska. Fish. Oceanogr. 1: 20–31. Vaughan S. L., C. N. K. Mooers, and S. M. Gay III. 2001. Physical variability in Prince William Sound during the SEA study (1994–1998). Fish. Oceanogr. 10(Suppl. 1): 58–80. Vermeer, K., L. Cullen, and M. Porter. 1979. A provisional explanation of the reproductive failure of tufted puffins Lunda cirrhata on Triangle Island, British Colombia. Ibis 121: 348–354. von Huene, R. W., G. G. Shor, Jr., and R. J. Malloy. 1972. Offshore tectonic features in the affected region, In: The Great Alaska Earthquake of 1964, vol. 6: Oceanography and Coastal Engineering, National Research Council, National Academy of Sciences, Washington, DC, pp. 266–289. Wang, Y., M. Ren, and J. Syvitski. 1988. Sediment transport and terrigenous fluxes, pp. 253–292, In: K. H. Brink and A. R. Robinson (Eds.), The Global Coastal Ocean Processes and Methods. John Wiley & Sons, Inc., N.Y., The Sea vol. 10, 604 pp. Ward, A. 1997. A temporal study of the phytoplankton spring bloom in Prince William Sound, Alaska. MS Thesis. University of Alaska Fairbanks, Fairbanks, Alaska. Warren, B. A. 1983. Why is no deep water formed in the North Pacific? J. Mar. Res. 41: 327–347. Warren, B. A. and W. B. Owens. 1985. Some preliminary results concerning deep northern-boundary in the North Pacific. Prog. Oceanogr. 14: 537–551. Weingartner, T. J. 2005. Physical and geological oceanography: Coastal boundaries and coastal ocean circulation, In: P. Mundy (Ed.), The Gulf of Alaska. Biology and oceanography. Alaska Sea Grant, Fairbanks, Alaska, pp. 35–57. Weingartner, T. J., K. Coyle, B. Finney, R. Hopcroft, T. Whitledge, R. Brodeur, M. Dagg, E. Farley, D. Haidvogel, L. Haldorson, A. Hermann, S. Hinkley, J. Napp, P. Stabeno, T. Kline, C. Lee, E. Lessard, T. Royer, and S. Strom. 2002. The northeast Pacific GLOBEC Program: Coastal Gulf of Alaska. Oceanography 15: 48–63.
168
Long-Term Ecological Change in the Northern Gulf of Alaska
Weingartner, T. J., S. L. Danielson, and T. C. Royer. 2005. Freshwater variability and predictability in the Alaska Coastal Current. Deep-Sea Res. 52: 169–192. Wells, R. S., D. J. Boness, and G. B. Rathbun. 1999. Behavior. In: J. E. Reynolds and S. A. Rommel (Eds.), Biology of Marine Mammals. Washington, DC, Smithsonian Institution Press, pp. 324–422. Welschmeyer, N. A., S. Strom, R. Goericke, G. DiTullio, M. Belvin, and W. Petersen. 1993. Primary production in the Subarctic Pacific Ocean: Project SUPER. Prog. Oceanogr. 32: 101–135. Welschmeyer, N. A., S. Strom, R. Goericke, G. Wheeler, P. A. and S. A. Kokkinakis. 1990. Ammonia recycling limits nitrate use in the oceanic Subarctic Pacific. Limnol. Oceanogr. 35: 1267–1278. Whele, D. 1983. The food, feeding and development of young tufted and horned puffins in Alaska. The Condor 85: 427–442. Whitney, F. A. and M. Robert. 2002. Structure of ‘Haida’ eddies and their transport of nutrients from coastal margins into the Northeast Pacific Ocean. J. Oceanogr. 58: 714–723. Willette, T. M. 2001. Foraging behavior of juvenile pink salmon (Oncorhynchus gorbuscha) and size dependent predation risk. Fish. Oceanog. 10(Suppl. 1): 110–131. Willette, T. M., R. T. Cooney, and K. Hyer. 1999. Predator foraging mode shifts affecting mortality of juvenile fishes during the subarctic spring bloom. Can. J. Fish. Aquat. Sci. 56: 364–376. Willette, T. M., R. T. Cooney, V. Patrick, D. M. Mason, G. L. Thomas, and D. Scheel. 2001. Ecological processes influencing mortality of juvenile pink salmon (Oncorhynchus gorbuscha) in Prince William Sound, Alaska. Fish. Oceanogr. 10(Suppl. 1): 14–41. Williams, G. C. 1966. Natural selection, the costs of reproduction and a refinement of Lack’s principle. Amer. Natur. 100: 687–692. Williams, W. J. 2003. Idealized modeling of seasonal variations in the Alaska Coastal Current, PhD thesis, University of Alaska, Fairbanks, 100 pp. Willmott, A. J. and L. A. Mysak. 1980. Atmospherically forced eddies in the northeast Pacific. J. Phys. Oceanogr. 10: 1769–1791. Wilson, J. G. and J. E. Overland. 1987. Meteorology, pp. 31–53, In: D. W. Hood and S. T. Zimmerman, (Eds.), The Gulf of Alaska, Physical Environment and Biological Resources. Alaska Office, Ocean Assessments Division, National Oceanic and Atmospheric Administration, U.S. Dept. Commerce, 655 p. Winship, A. J., A. W. Trites, and D. A. S. Rosen. 2002. A bioenergetic model for estimating the food requirements of Steller sea lions Eumetopias jubatus in Alaska, USA. Mar. Ecol. Progr. Ser. 229: 291–312. Winston, J. 1955. Physical aspects of rapid cyclogenesis in the Gulf of Alaska. Tellus 7: 481–500. Wittenberger, J. F. and G. L. Hunt. 1985. The adaptive significance of coloniality in birds. In: K. C. Parkes (Ed.), Avian Biology, vol. VIII. Orlando: Academic Press, pp. 1–78. Xie, L. and W. W. Hsieh. 1995. The global distribution of wind-induced upwelling. Fish. Oceanogr. 4: 52–67. Xiong, Q. and T. C. Royer. 1984. Coastal temperature and salinity observations in the northern Gulf of Alaska, 1970–1982. J. Geophys. Res. 8: 8061–8068. Yankovsky, A. E. and D. C. Chapman. 1997. A simple theory for the fate of buoyant coastal discharges. J. Phys. Oceanogr. 27: 1386–1401. Ydenberg, R. C. 1989. Growth-mortality trade-offs and the evolution of juvenile life histories in the alcidae. Ecology 70: 1494–1506.
Ecosystem Structure 169
Yuan, X. and L. D. Talley. 1996. The Subarctic frontal zone in the North Pacific: Characteristics of frontal structure from climatological data and synoptic surveys. J. Geophys. Res. 101: 16491–16508. Zador, S. G. and J. F. Piatt. 1999. Time-budgets of common murres at a declining and increasing colony in Alaska. The Condor 101: 149–152.
Mini-Glossary Advection: Transport by ocean currents from one region having a distinct set of water properties to another region where the water properties differ. Current advection can thus be an important agent of change by redistributing salinity, nutrients, heat, etc., in the ocean. Allometry: Quantitative relationship showing how the size of an animal’s structural (e.g. wing length) or physiological (e.g. metabolic rate) characteristics change in relation to overall body size. Facilitates predictions of an animal’s characteristics from its body size as well as meaningful comparisons among animals of different shapes and sizes. Congeneric: Different species within the same genus classification. Corvids: A family of birds including crows and ravens. Cyclogenesis: Generation or formation of a low-pressure (cyclonic) weather system. Density dependent: Effects whose intensity changes with increasing population density, such as mortality or growth rates. Fledge: When a chick leaves the nest. Front: A region of relatively rapid horizontal change in a water mass property, such as temperature or salinity. Gestation: Period of fetal development in mammals. Halocline: The horizontal layer in a water column across which salinity changes rapidly with depth. Implantation: Attachment of the fertilized egg or developing embryo to the uterine wall. Internal waves: Waves that form on the interface between ocean layers having two different densities, for example, the thermocline. More generally, however, internal waves propagate through a fluid where the density varies continuously throughout the water column. Unlike surface waves, internal waves can propagate vertically as well as horizontally. Internal waves have a variety of length and time scales, and their vertical displacements can be up to many tens of meters. Isopleth: A contour line along which the measured value is constant. Iteroparity: reproducing annually over multiple years Iteroparous: Spawning successive batches of eggs in a season, i.e. multiple spawning events. Lactation: Production of milk by the mammary gland. Latent heat exchange: Heat exchange at the sea surface, associated with evaporation. Megajoule (MJ): One million joules, where one joule is equal to 0.2390 calories. Natal philopatry: Return to birth location (usually for reproduction). Natality: Birth rate. Neonatal: Newly born. Ocean regime: A statistically defined multi-year ocean state that differs from other regimes and around which the ocean varies. Individual regimes are characterized by a unique set of physical features, e.g. degree of stratification and upwelling, that influence biological populations.
170
Long-Term Ecological Change in the Northern Gulf of Alaska
Omnivore: An animal that feeds on a large variety of prey. Orographic: The physical effect that mountains or mountain ranges exert on weather. Otariids: Family Otariidae, the eared seals or fur seals and sea lions. Parturition: Birth. Phocids: Family Phocidae, the earless seals or true seals. Piscivore: An animal whose main prey are fish. Planktivore: An animal whose main prey are plankton, e.g. copepods, fish larvae, pteropods. Postpartum: Following birth. Precocial: Chicks hatch and can immediately leave the nest and feed themselves. Prepartum: Prior to birth. Protandric hermaphrodite: a species that begins life as a male and then becomes a female in later life Pycnocline: The horizontal layer in a water column across which density changes most rapidly with depth. Typically, the pycnocline coincides with either the thermocline or halocline. Recruitment: The incoming strength of a year-class cohort entering the adult population. Semelparity: reproducing once in a life time Sensible heat exchange: Heat flows from warm to cold regions. Shear: A horizontal or vertical change in current velocity. Shelf waves: Large-scale, long-period waves whose existence depends on a coast, changes in bathymetry (most importantly, the offshore increase with depth), and the rotation of the earth. Shelf waves have periods of a day or longer and wavelengths varying from tens to hundreds of kilometers. Shelf wave amplitudes are generally largest at the coast and decay offshore. In the northern hemisphere, they propagate so that the coast is to the right, when looking in the direction of propagation. Shelf waves can cause the pycnocline to upwell or downwell as they pass through a region. They are typically excited by changes in alongshore winds. In addition, tides can propagate along continental shelves as shelf waves. Spawning philopatry: Return to a previous spawning site used by adults. Thermocline: The horizontal layer in a water column across which temperature changes rapidly with depth.
Chapter 3
Agents of Ecosystem Change 3.1. Introduction Robert B. Spies A typical year in the Gulf of Alaska encompasses great fluctuations in physical conditions, available biological energy and in the growth and reproduction of its dominant organisms (Chapter 2). The strategies that each of these species evolves in order to capture and utilize energy and maintain its status in the ecosystem allows them to expand their populations at some times, but do not work as well at other times, as we shall see in Chapter 4. The amount of production in the system and how it is passed through the food web changes in response to climate, and the effectiveness of these strategies for each species changes in response to climate, food supply, harvesting, predation, contaminants, disease, and the internal dynamics of the food web. So, many species are dominant in some decades, or for longer periods of time, and then decline. The various forcing factors may all be at work in any one year, but they do not have a constant or equal effect from one season to the next or one year to the next. These forcing factors have the potential to reverberate through the system with potentially profound effects. In this section, the factors that force changes in the ecosystem are discussed. We start with the physical changes in the system: climate (Section 3.2) and geophysical forces (Section 3.3). We then discuss biological interactions: interactions among species (e.g., competition and predation) (Section 3.4); harvesting and predator control by man (Section 3.5), effects of disease on populations (Section 3.6); and finally, the effects of man’s wastes in the ocean (Section 3.7). These root causes of change then set the stage for considering long-term changes in the ecosystem in Chapter 4.
3.2. Climate Thomas Weingartner 3.2.1. Introduction The Earth’s climate system includes the atmosphere, ocean, cryosphere (glaciers and sea ice), and land, and changes in one affect the others through transfers of matter Long-Term Ecological Change in the Northern Gulf of Alaska Robert B. Spies (Editor) © 2007 Elsevier B.V. All rights reserved.
172
Long-Term Ecological Change in the Northern Gulf of Alaska
(e.g., the sediment and chemical load in river runoff) and energy (e.g., heat and moisture transports in ocean currents and winds). The changes induced in one part of the system may or may not feed back to it. If feedbacks do occur, they may accelerate (positive feed back) or decelerate (negative feed back) the changes. Changes in any part of the system as well as the material and energy transfers between system components proceed along myriad pathways and hence over a broad range of time and space scales. So it may be difficult to isolate or identify the causative factors behind an observed change. For example, the Gulf of Alaska will respond to changes in atmospheric forcing locally and remotely; elsewhere in the North Pacific, or possibly even more distant regions. Such changes may be propagated into the Gulf via the motions of the atmosphere or the ocean. Interannual and interdecadal changes in the Gulf of Alaska ecosystem are primarily forced by the atmospheric or oceanic components of the climate system. For example, changes in the strength and position of the Aleutian Low will alter the intensity, frequency, and trajectory of storm systems entering the Gulf. These storm systems will, in turn, affect precipitation patterns and rates, cloud cover, heat fluxes between the ocean and atmosphere, and the wind stress distribution over the shelf and the basin. These “local atmospheric changes” will affect the circulation, stratification, and temperature structure of the shelf and basin. More remote changes can affect the large-scale current systems that deliver heat and organisms from lower latitudes into the Gulf. Mankind is presently engaged in a massive ecological experiment through the introduction of greenhouse gases into the atmosphere as well as other anthropogenic changes (such as land-use patterns) that will result in unprecedented changes in Gulf of Alaska’s marine ecosystem.
3.2.2. Climate Forcing Long-period (annual or longer) physical variations in the marine environment of the Gulf of Alaska are mediated by changes in the large-scale circulation of the North Pacific basin’s atmosphere and ocean. From analyses of both models and observations (largely gathered during the twentieth century), climatologists identify several characteristic time scales of ocean–atmosphere variability in the North Pacific that account for much of the annual and longer time period variance. These include interdecadal periods (of about 20 years), intra-decadal periods from 2 to 7 years, and the interannual or year-to-year. An important aspect of the longer-term variations are that these appear to be quasi-periodic meaning that the changes recur at approximately regular intervals, with the magnitudes and spatial structure of the variations being similar – but not identical – from one event to the next. Quasi-periodic attributes allow statistical characterizations of North Pacific climate variability and suggest that this variability is to some extent predictable.
Agents of Ecosystem Change
173
North Pacific climate variability at inter- and intra-decadal periods is accompanied by distinct sea surface temperature patterns that are used to classify climate state. Other variables can and are used as well, although the ease with which temperatures can be measured makes this a particularly attractive variable for climate diagnoses. The sea surface patterns are generally distinguished based on statistical analyses of temperature anomalies. An anomaly is defined as the deviation of a variable from a suitably defined mean. Thus, a monthly temperature anomaly is the temperature deviation for a specific month from the average monthly value estimated from a long-term data set. Using monthly anomalies effectively removes the annual cycle from the data and is particularly effective in revealing nonseasonal variability. The most prominent feature in the inter-decadal period is the Pacific Decadal Oscillation (PDO) described by Mantua et al. (1997). The PDO involves a warm (or positive) phase during one portion of its oscillation and a cold (or negative) phase during the other. During the warm phase, sea surface temperatures are above normal in the eastern and northeastern Pacific and over a broad band extending along the equator in the eastern and central Pacific and below normal in the central and northwest Pacific. The opposite pattern ensues during the cold phase. Figure 3.1 is a schematic of the anomaly patterns in sea surface temperature and northern North Pacific winds associated with the warm and cold phases of the PDO.
A
B
Figure 3.1: Broad scale patterns of sea surface temperature and wind anomalies over the North Pacific Ocean associated with the PDO warm phase (a) and cold phase (b). Sea surface temperature anomalies in the Pacific Ocean during the positive phase (left) and negative phase (right) of the Pacific Decadal Oscillation. The key to the anomalies is in degrees Centigrade. The black arrows indicate surface winds and their lengths are proportional to their average strength.
174
Long-Term Ecological Change in the Northern Gulf of Alaska
~o–Southern Variability at intra-decadal periods is mainly associated with the El Nin Oscillation (ENSO) cycle. While a number of indices have been developed to characterize ENSO, a particularly useful one is the cold tongue index (CTI), which is composed of the monthly sea surface temperature anomalies averaged over the eastern and central equatorial Pacific. Large positive values of the CTI indicate an active ~o, whereas large negative values denote a La Nin ~a, whose attributes are oppoEl Nin ~ site to El Nin o. Although the CTI is based solely on thermal conditions along the central and eastern equatorial Pacific, the broad scale patterns of temperature anomalies that subsequently develop across the North Pacific in response to an El Nin~o or ~a are very similar to the PDO patterns. During El Nin ~o, above-normal sea surface La Nin temperatures span the eastern and northeastern Pacific and along the equator, while below-normal temperatures span the central and northwestern Pacific. The reverse pattern develops during La Nin~a events, with the exception that strong anomalies generally do not develop along the west coast of the United States (Smith et al., 2001). Although the ENSO and PDO sea surface temperature anomaly patterns are somewhat similar, they differ in two major respects. First, the equatorial sea surface temperature anomalies associated with the PDO are smaller and more broadly distributed across the equator than those associated with the El Nin~o/La Nin~a. Second, the sea surface temperature anomalies in the northwest and central Pacific associated with the ~a. PDO are substantially larger than those occurring during El Nin~o/La Nin Time series of the PDO and CTI indices from 1930 to the present show how the various phases of the PDO and ENSO (El Nin~o versus La Nin~a) have varied from then to now (Fig. 3.2). Positive values of the PDO index indicate the warm phase and large positive values of the CTI indicate strong El Nin~o events. Negative values of each indicate the cold phase of the PDO and La Nin~a episodes. The figure also includes the time series of the North Pacific Index (NPI), which is an area-weighted average of mean sea level pressure over the region 30 to 65°N, 160°E to 140°W (Trenberth and Hurrell 1994). The NPI is computed for the November through March season of each year when the Aleutian Low is most prominently developed over the North Pacific (cf. Section 3.2.). It is evident from Fig. 3.2 that while there are distinct differences in the temporal behavior of these indices, there is also some co-variability among them. For example, the smoothed versions of the NPI and PDO (red lines in Fig. 3.2), indicate that the warm phase of the PDO is accompanied by a stronger Aleutian Low, while the PDO cold phase is associated with a weaker Aleutian Low. There is also a tendency for a stronger Aleutian Low and warm phase PDO to develop during Los Nin~os and a weaker low and cold phase PDO during La Nin~a events. Thus ENSO and PDO can constructively or destructively interfere with one another resulting in an amplified or diminished atmosphere–ocean response to either phenomenon at regional or broader scales. The correspondence among these indices is not consistent, however, suggesting
Agents of Ecosystem Change
175
Figure 3.2: Time series of several climate indices used to characterize the state of the North Pacific Ocean, including the North Pacific Index (top), the equatorial cold tongue index (middle), and the Pacific Decadal Oscillation (bottom), from 1930 to 2003. The blue lines are the annual values and the red lines are four-year running averages.
that random fluctuations or other, as yet unidentified, processes in the ocean–atmosphere system can modulate the expressions of longer period variability. Each mode of variability leaves its signature in the Gulf of Alaska via atmospheric and/or oceanic pathways. Changes in the strength and position of the Aleutian Low modify the “local” atmospheric forcing of the marine environment over the Gulf through air–sea heat exchanges, the hydrologic cycle, and wind forcing. “Remote” effects modify the ocean elsewhere in the North Pacific with the large-scale ocean circulation field eventually carrying these changes into the Gulf of Alaska. Because the ocean and atmosphere communicate with one another in many complex ways, ascribing changes solely to one medium or the other is difficult, if not impossible. Indeed, unraveling the mechanisms responsible for the evolution of these various modes of variability is a subject of active research and is central in developing the capability to predict climate variations and its effects upon marine ecosystems.
176
Long-Term Ecological Change in the Northern Gulf of Alaska
3.2.3. Physical Environmental Variability in the North Pacific Ocean El Nin~o Southern Oscillation (ENSO) Under normal conditions sea surface temperatures along the equator of the central and eastern Pacific Ocean are cooler relative to temperatures in the western Pacific and those immediately to the north and south of the equator. The low-temperature band is maintained by the westward-blowing Southeast Trade Winds, which induce equatorial ~o events are initiated in the equatorial Pacific upwelling of cold subsurface waters. El Nin Ocean when the Southeast Trade Winds collapse. This causes a reduction in equatorial upwelling and a warming in surface waters in the central and eastern portions of the equatorial basin. Under such conditions, the CTI is positive and indicative of above-normal ~o conditions. These adjustments are also accompasea surface temperatures and El Nin nied by a large change in atmospheric pressure between Australia and Tahiti (the Southern Oscillation) and marked changes in heat flux processes between the ocean and atmosphere over the equator. Internal adjustments within the tropical atmosphere–ocean system eventually strengthen the Southeast Trades and return the equatorial ocean to either normal conditions or, more typically, to the opposite phase of the ENSO cycle; a ~a. During Las Nin ~as, the Southeast Trades intensify, equatorial upwelling is La Nin enhanced, and unusually cold waters occur in the eastern and central equatorial Pacific ~o Ocean. The following paragraphs discuss the extra-tropical responses to an El Nin ~a being roughly opposite to those associated with event, with the responses to a La Nin ~ an El Nino. No two ENSO events are exactly alike in either their equatorial evolution or their extra-tropical signatures. ~o events trigger changes in tropical atmospheric convection patterns that affect El Nin the global atmospheric circulation. In the northern hemisphere, this “atmospheric bridge” (Lau and Nath 1994; Alexander et al., 2002a) usually involves a strengthening (decreased pressure and a negative tendency in the NPI) in the Aleutian Low and an increase in high pressure over central Canada. Consequently, the mid-latitude jet stream intensifies and extends southward across the central Pacific Ocean before veering northeastward into the Northeast Pacific. The jet stream adjustment causes southward deflection of storm tracks and the advection of cold, dry Asian air masses across the central Pacific and warm, moist air into the Northeast Pacific. ~o results in a deepening Large-scale adjustments in the equatorial ocean during El Nin of the equatorial thermocline that propagates slowly eastward along the equator as a long (thousands of km) wavelike disturbance. This disturbance, termed an “equatorial Kelvin wave”, is confined to within a few degrees of the equator because of the earth’s rotation. The equatorial Kelvin wave crosses the Pacific in about two months and leaves in its wake, a deepened thermocline and warmer sea surface temperatures. Upon encountering the South American continent, some of the Kelvin wave energy is reflected poleward as another type of Kelvin wave in which the propagating signal is largely confined to the shelf and continental slope. Signatures of the poleward propagating Kelvin wave
Agents of Ecosystem Change
177
include a deepening of the thermocline along the continental slope of western North America and anomalous northward currents that transport water from low to higher latitudes. The anomalous current patterns and the deepened thermocline contribute to elevated shelf water temperatures and coastal sea levels (Clarke, 1983; Clarke and Van Gorder, 1994). Hence, processes in both the atmosphere and the ocean link the equatorial ocean to extra-tropical regions. The ENSO events typically begin to develop at the equator in spring, but attain maximum expression in the fall. Extra-tropical ENSO manifestations occur in the northern North Pacific Ocean from 8 to 12 months after the initiation of an El Nin~o (or La Nin~a) event. Consequently, ENSO signals in the Gulf of Alaska have maximal expression in the fall and winter. The intensification of the Aleutian Low during an ENSO event induces significant changes in the upper ocean temperatures, heat and moisture fluxes, and mixed-layer depths. Enhanced latent and sensible heat loss over the central Pacific leads to cooler surface temperatures here. In the Gulf of Alaska, these heat fluxes diminish in winter and contribute to anomalously warm upper ocean temperatures. Modeling studies by Alexander et al., (2002a) suggest that increased vertical mixing (cool subsurface water brought to the surface) and southward, wind-driven, surface Ekman transport of cooler surface waters from northern latitudes also contribute to central Pacific cooling. In the Gulf of Alaska basin, the anomalous winds enhance upwelling of cooler water into the mixed layer, which buffer, but do not completely counteract, the warming tendency associated with the wintertime reduction in atmospheric heat loss. Along the west coast of North America, Los Nin~os also involve anomalous northward winds that suppress the coastal upwelling of deeper, cooler water. Diminished coastal upwelling thus enhances the warming tendency induced by the poleward propagating Kelvin wave. Reduced coastal upwelling also drastically reduces marine biological production by curtailing the vertical transport of nutrients into the euphotic zone. Consequently, the response of the Northeast Pacific to El Nin~o involves the ocean adjusting to both the “local” wind field and to the “remotely” generated Kelvin wave signal that propagates northward from the equator. The anomalous ocean surface temperature patterns that develop in the northern North Pacific are maximal during the winter following an ENSO event and tend to subside in summer with the development of the shallow summer thermocline (see Section 2.2.2). Nevertheless, ENSO thermal anomalies can persist at depths greater than the depth of the summer mixed layer and recur the following fall and winter after fall storms have eroded the summer stratification. Although variable in extent and magnitude, this “re-emergence mechanism” (Alexander and Deser, 1995; Alexander et al., 2002b) provides a means by which some El Nin~o (or La Nin~a) effects can persist in the extra-tropical ocean for 1– 2 years after its disappearance along the equator.
178
Long-Term Ecological Change in the Northern Gulf of Alaska
The Pacific Decadal Oscillation (PDO) Because of the similarity in sea surface temperature patterns between ENSO and PDO, Zhang et al. (1997) describe the PDO as “decadal ENSO-like variability” and suggest that the PDO might originate in the tropics and subsequently generate an extra-tropical response. As noted, the PDO response involves anomaly patterns in sea surface temperature and the Aleutian Low similar to those that accompany ENSO. Thus, during the PDO warm phase, the Aleutian Low deepens and moves southward over the central Pacific, whereas during the PDO cold phase, the Low weakens and is centered over the Northwest Pacific. A notable difference between ENSO and PDO is that the PDO lacks the Kelvin wave signals that occur during ENSO events. An outstanding problem is how these two independent expressions of ocean temperature and atmospheric variability have such similar patterns. This is because, while the dynamics of ENSO, and to a lesser extent its extra-tropical connections, are reasonably well understood (although there are significant issues still to be resolved (McPhaden, 2004), the mechanisms responsible for decadal scale variations are a subject of active debate. (Miller and Schneider (2000) provide the most recent review of candidate mechanisms.) Recent model studies suggest that the quasi-periodic nature of the PDO is not an inherent property of the mid-latitude North Pacific ocean–atmosphere system, but instead involves physical linkages to the equatorial region and/or more remote ocean–atmosphere dynamics (Seager et al., 2001; Schneider et al., 2002). There is also some suggestion that ice–ocean–atmosphere decadal scale oscillations in the Arctic (Thompson and Wallace, 1998) might also affect the northern North Pacific. The Arctic Oscillation, evident in sea level pressure over the polar cap, results in a decrease in mid-latitude westerlies when arctic sea level pressures are unusually low (Thompson et al., 2000). The trend since the 1970s has been toward decreasing arctic pressure, which might explain the stronger connections between the Arctic Oscillation and the PDO reported by Overland et al. (1999). While much remains to be done to understand the origins of North Pacific decadal scale climate variability and its connections to the tropics and poles, substantial progress is being made. In particular, considerable progress has been made in explaining the remarkable transition in oceanic conditions associated with the so-called “regime shift” of the mid-1970s. This climate shift was associated with the rapid change in the PDO from the cold to warm phase coincident with the southeastward displacement and intensification of the Aleutian Low (Trenberth and Hurrell, 1994) as evident in the change in sign of both the PDO and NP indices (Fig. 3.2). The regime shift resulted in an intensification of the Aleutian Low and an increase in the strength of mid-latitude westerlies (Hanawa et al., 1996; Parrish et al., 2000) over the north central Pacific. This induced cooling in the central North Pacific and warming in the eastern, northern, and southern portions of the basin (Nitta and Yamada, 1989) due to changes in wintertime atmosphere–ocean heat fluxes, deeper vertical mixing in the central Pacific and changes in wind-driven, surface Ekman
Agents of Ecosystem Change
179
transport (Miller et al., 1994). Mid-latitude transport in both the North Pacific Current and the Subtropical Gyre also increased (Miller et al., 1998; Deser et al., 1999; Parrish et al., 2000). There are some suggestions that transport in the western Subarctic Gyre slightly increased (Miller et al., 1998; Joyce and Dunworth-Baker, 2003) as well, while transport in the Alaskan Gyre appears to have decreased (Lagerleof, 1995; Parrish et al., 2000). The transport changes were accompanied by a southward shift in the position of both the Kuroshio–Oyashio Extension (Seager et al., 2001) and the Subarctic front (Joyce and Dunworth-Baker, 2003), intensification of the north–south temperature gradient across the Kuroshio–Oyashio Extension (Miller and Schneider, 2000), a deepening of the thermocline in the central Pacific and subsurface cooling in the Subarctic Gyre. The changes in the Kuroshio–Oyashio Extension area were not coincident with local atmospheric forcing (Xie et al., 2000), but instead, lagged the development of central Pacific cooling by about five years (Nakamura et al., 1997). The delayed response of the oceanic gyres to the intensification of the Aleutian Low reflects the longer time required for the deeper portions of the ocean to adjust to changes in the wind stress distribution associated with a sustained intensification of the Aleutian Low. Changes in mid-latitude winds, whether instigated by the ENSO or PDO cycle, lead to anomalous late winter– early spring mixed layer depths. During an El Nin~o or warm-phase PDO, mixed layer depths tend to deepen in the central Pacific and shoal in the Gulf of Alaska basin, with opposite responses developing during a ~a or cold-phase PDO. Because these responses are most prominent in late winter, La Nin mixed layer depth variations could affect the timing and (possibly) the magnitude of the spring phytoplankton bloom and eventually production at upper trophic levels (Polovina et al., 1995). During the most recent “regime shift”, it appears that the mixed layer deepened by as much as 70 m in the central Pacific, while it shoaled by from 10 to 30 m in the Gulf of Alaska (Polovina et al., 1995). Based on the results from a simple biological model, they argue that a mixed-layer deepening in the central Pacific should enhance biological production because deeper vertical mixing increases nutrient levels in the euphotic zone. (In contrast, mixed layer deepening at the higher latitudes of the Gulf of Alaska inhibits biological production because plant cells will spend more time at depths deeper than the euphotic zone where sub-optimal light levels will impede production.) Polovina et al. (1995) argue that mixed layer shoaling in the Gulf of Alaska should enhance primary production because phytoplankton are more likely to be retained in the euphotic zone where nutrients do not tend to be limiting. As will be discussed later, the Gulf of Alaska mixed layer does indeed shoal during Los Nin~os, however, this appears to negatively impact biological production because it has been accompanied by a longer-term trend of increasing nutrient depletion in the euphotic zone here. Although the ENSO and PDO patterns appear to be the dominant modes of North Pacific variability, there are possibly other modes operating as well.
180
Long-Term Ecological Change in the Northern Gulf of Alaska
For example, Bond et al. (2003) find that since 1998, sea surface temperatures and sea level pressures in the North Pacific resemble neither the cold nor warm phase pattern of the PDO. Instead, the more recent conditions show that the southeastern portion of the North Pacific has been subject to strong high pressure (akin to atmospheric conditions during the cold phase of the PDO), while the northern Pacific has been subject to an intensified Aleutian Low (similar to the warm phase of the PDO). These conditions have led to an increase in the transport of cool, fresh, sub-arctic waters along the west coast of North America (Huyer, 2003), with enhanced upwelling and primary production (Wheeler et al., 2003; Thomas et al., 2003). Murphree et al. (2003) find that both the North Pacific Current and California Current strengthened at this time. The response in the Gulf of Alaska is not known, although it seems likely that upwelling in the basin has also intensified. These changes are continuing and some believe that they herald the onset of a new regime shift, possibly different from that which occurred in the mid-1970s.
3.3. Geophysical Mechanisms Robert B. Spies 3.3.1. Introduction Earthquakes, tsunamis, and volcanic eruptions are sudden; the ecological adjustments can occur rather quickly or take as long as a decade or more. Past events suggest that the spatial extent of these alterations can be very local or up to several hundreds of kilometers, depending on the event. On large time scales, the gradual advance and retreat of glaciers, for example, will have effects on coastal and shelf habitat on scales of hundreds, thousands and millions of years. Marine animal populations can easily adjust to the earth’s gradually changing surface, but abrupt alterations of the seabed and shoreline can have pronounced, rapid effects and adjustments to the new habitat are more significant to populations. Organisms that depend on stable substrates, for example clams or nesting seabirds, are most susceptible to geophysical forces.
3.3.2. Tectonics and Earthquakes Tectonics in the Gulf, unrivaled in any other coastal area in the world, are at the root of many of these geophysical events (Fig. 3.3). The Pacific plate, generated at the Pacific mid-ocean ridge and moving at a rate of about 7–8 cm a year in a north to northeasterly direction into the Gulf of Alaska, abuts the North American plate along a 2000-km-long zone ringing the Gulf. Some of the Pacific plate accretes to the
Agents of Ecosystem Change
181
North American plate, building up the North American continent, while most of the Pacific Plate dives toward the liquid magma on which the continent floats. As the two plates collide and lock at fault structures, crustal pressure builds. A sudden release of pressure, as part of the plate slips, causes an earthquake, sometimes tilting the earth’s crust and drastically changing the elevation of habitats. The force of earthquakes varies about 10 orders of magnitude, measured on a logarithmic scale of 1–10 (the popularly labeled Richter scale). Frequent small earthquakes do not have drastic consequences; it is the occasional very large quake (about magnitude 7.5 and above) that causes sudden and widespread changes. In a severe earthquake, movement of large blocks of earth over hundreds of kilometers can displace significant amounts of coastal habitat laterally and vertically (see 1964 quake rupture zone, Fig. 3.3). This dislocation can alter intertidal communities, whose composition is finely tuned to depth, tidal exposure and substrate type, as happened in Olsen Bay, Prince William Sound in 1964 (Hubbard, 1971). Extensive effects on the distribution of rocky intertidal communities were seen 15–18 months after the 1964 earthquake in Prince William Sound (Haven, 1971). Moderate uplift (2–9 ft) reduced the abundance of most rocky shore intertidal organisms, but most mid-range intertidal zone communities resembled inferred pre-spill communities.
Figure 3.3: An overview of major tectonic features in the Gulf of Alaska. Major faults are indicated by orange lines; active volcanoes by red dots, movement of the Pacific Plate by blue arrows, fault blocks are in yellow with the dates of recent major quakes that moved them, and seismic gaps in transparent green.
182
Long-Term Ecological Change in the Northern Gulf of Alaska
The notable exception was a shift in abundance of the dominant species of barnacles, with post-spill recruitment dominated by Balanus balanoides, whereas before the spill, this species co-occurred with B. glandula. In areas of extreme uplift, almost total destruction of intertidal and subtidal communities occurred. In what was considered the Fucus zone in the post-earthquake environment, Porphyra, a red algae, was growing rather than the brown algae Fucus. Parts of Montague Island were uplifted 30 ft or more and this resulted in the death of large areas of subtidal communities that were easily seen from the air, as large numbers of calcareous remains of subtidal animals turned the new beaches white (Fig. 3.4). Surprisingly, in downthrust rocky shores, many healthy mid-intertidal organisms were living well at their new lower elevations. The decline of clam populations in Prince William Sound since the mid-1960s has been attributed partially to the effects of the 1964 earthquake (as well as to expanding sea otter numbers in the late 1960s–1980s). The earthquake was estimated to have killed 36% of the clams in areas of moderate uplift (3–9.7 ft) (Baxter, 1971). Earthquakes can damage anadramous fish streams, for example reducing escapements and alevin production in pink salmon (Oncorhynchus gorbuscha) on shorelines with the greatest changes in elevation (Roys, 1971). In some cases, the uplift resulted in impassable blockages, sharply curtailing or extirpating some runs. For example in 1964, extensive disruptions to salmon streams around Montague Island in Prince William Sound led to the extirpation of several pink and chum salmon (O. keta) sub-populations. The bed of one salmon stream in Herring Bay, Prince William Sound
Figure 3.4: An aerial view of the Cape Clear, Montague Island, Prince William Sound, which was uplifted in excess of 30 ft by the 1964 earthquake. The gently sloping surface to the water’s edge (a width of about 400 m) was previously underwater and the white is a result of dead, bleached calcareous organisms (Photograph, courtesy of USGS).
Agents of Ecosystem Change
183
was lifted away from the bay’s surface by more than 7 meters. In the same earthquake, a 2–4 meter uplift in the Copper River Delta greatly altered plant succession and available habitat for a host of bird species in this large and productive area (Thilenius, 1990). See Box 3.1 for an explanation of the way this earthquake affected nesting seabirds on Middleton Island. The probability of one or more earthquakes of magnitude 7.8 or greater along the Alaskan and Aleutian arcs in the next several decades appears to be over 50%. There are important seismic gaps in coastal Alaska, e.g., the Shumagin gap and the Yakataga gap, where there has been much lower than average seismic activity in the last century, and where the probabilities for a large quake are significantly greater than in other plate boundary segments (Jacob, 1984) (see Fig. 3.3). A quake in one or more of these areas could greatly affect large coastal areas of the Gulf and their biological populations.
3.3.3. Sediment Slumping The same crustal dynamics that give rise to earthquakes also build the mountains that ring the Gulf of Alaska, the highest coastal mountain range in the world. These high, young mountains contribute very large amounts of sediment to the adjacent continental shelf from glaciers such as the Bering, Malaspina and Columbia, and from large rivers such as the Copper, Knik, Matanuska and Susitna, that eventually form unconsolidated deposits on the adjacent shelf. There are large areas of the shelf that have inherently unstable fine-grained sediments that occasionally cleave off and slump into deeper waters, often apparently in response to earthquakes (Hampton et al., 1987). Areas of fine-grained clay–silt sediments with high water content, characteristic of glacial inputs, are the most unstable. Such areas are common on the continental shelf of the Gulf. Seismic-induced slumps are more likely to occur below 75 meters, whereas in shallower waters, storm waves are relatively more important sources of repeated forces that result in sediment slumping. There are slump features in some areas of the Gulf, e.g., near Yakutat, and the Copper River, that are 100 km2 or more in area. The effects on the benthic communities of large-scale sediment movement have not been studied, but they are potentially significant.
3.3.4. Volcanism Another result of the subduction of oceanic seafloor beneath the North American Plate is volcanism. There are two major zones of volcanism in Alaska. The Aleutian arc, which strectches from Aniak Island in the Aleutian chain to Denali in interior Alaska, borders the Gulf for thousands of kilometers, This arc contains 74 volcanoes, about 20 of them active during the present geological period (Fig. 3.3). A second
184
Long-Term Ecological Change in the Northern Gulf of Alaska
BOX 3.1: CHANGES IN SEABIRDS AT MIDDLETON ISLAND AFTER 1964 by Alan Springer The Good Friday earthquake of 1964 changed the face of coastal and offshore habitats at numerous sites around the rim of the Gulf of Alaska. One of the most notable alterations occurred on Middleton Island, where a fascinating sequence of biological changes resulted from the physical upheaval (Fig. 3.5). At Middleton, the immediate effect was an uplifting of the whole island, for a large block of the earth tilted up on a hinge line that extended from about the epicenter of the quake near Whittier in eastern Prince William Sound all the way to the southwest shore of Kodiak Island. The uplift instantly created more nesting habitat for common murres (Uria aalge) and black-legged kittiwakes (Rissa tridactyla). These two species, like nearly all birds, have “learned” to nest in locations protected from predators and over evolutionary time, have adopted nest sites high on cliff faces. On islands free from predatory mammals, murres nest on the sides and tops of unvegetated or sparsely vegetated hills and kittiwakes nest on steep, but not necessarily precipitous, bluffs. Thus, on Middleton Island, the newly uplifted cliffs, bluffs, and plant-free slopes instantly offered an abundance of additional nesting sites for the birds. And as the new habitat was occupied, the numbers of kittiwakes and murres burgeoned 10-fold over the next two decades. But seabirds were not the only species to pioneer the new lands. Plants began to invade, first on the slopes created by the earthquake and then on the newly eroding cliffs, which, being now set back beyond the ocean’s reach, were no longer being maintained sheer by the pounding surf. The advancing plant community, comprised mostly of grasses and sedges, transformed the face of the land into an ideal nesting habitat for glaucous-winged gulls (Larus glaucescens). (The gulls, unlike murres and kittiwakes, eschew bare ground and cliffs.) The gull population began to grow, from fewer than 1000 in the early 1970s to over 12,000 by the early 1990s. This was not without consequence for the other birds of the island, since gulls are voracious predators and have a particular taste for seabirds – adults of small species, the young of large species, and the eggs of all species. Now, young murre chicks leave their nests when they are only about 25–30% grown. At this age, they can flutter some distance but cannot fly far. Normally, from their typical cliff-face nest sites, the chicks can easily launch and splash safely down into the water below where their fathers
Agents of Ecosystem Change
185
9000 8000
Murres
7000
Count
6000 5000 4000 3000 2000 1000
1955 1957 1959 1961 1963 1965 1967 1969 1971 1973 1975 1977 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999
0
Year 90000 80000
Kittiwakes
70000
Count
60000 50000 40000 30000 20000 10000
1955 1957 1959 1961 1963 1965 1967 1969 1971 1973 1975 1977 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999
0
Year 14000 12000
Gulls
Count
10000 8000 6000 4000 2000
1955 1957 1959 1961 1963 1965 1967 1969 1971 1973 1975 1977 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999
0
Year
Figure 3.5: Views of the same area of Middleton Island in 1949 before the earthquake (top), in 1978, fourteen years after the earthquake with the extensive uplifting of the area below the cliffs visible (middle), and in 2000 with extensive development of vegetation on the uplifted bench (Photographs, courtesy of Scott Hatch, USGS).
186
Long-Term Ecological Change in the Northern Gulf of Alaska
wait for them. However, following the earthquake, many cliffs on Middleton Island were no longer sea-cliffs towering over the ocean, but instead, now had broad beaches before them. At first this was not too great an obstacle, for the chicks could flutter down to the beaches and simply run to the sea. Increasingly, however, the murre chicks faced a serious obstacle: the growing numbers of glaucous-winged gulls. In addition, ongoing changes in the plant community – larger and bushier willows and alder were beginning to replace grasses along the uplifted beaches–brought another difficulty for the chicks. With ever denser vegetation, the route to the sea was often obscured or blocked and many chicks became lost and died in the tangle before reaching the water. Gulls savor kittiwake chicks as well as murre chicks and are often able to pluck them from their nests, particularly if the parents are inattentive or stressed. So predation on kittiwakes also increased in proportion to the growth of the gull population. Thus, over time, the combination of primary plant succession, acting positively on gulls but negatively on murres, plus increasing predation by gulls on murre and kittiwake chicks, apparently led to profound declines in both murres and kittiwakes beginning in the late 1980s (see graph). Finally, in recent years, the gull population has also begun to fall, the likely result of now-diminishing prey resources.
group of volcanoes occurs in the Wrangell Mountains in the Alaskan arc. Ash from large eruptions, such as the 1912 Novarupta in the Katmai mountains in the Aleutian arc, forms distinct bands in buried sediments over a large area of the nearby continental shelf. That particular eruption was 10 times larger than the Mount St. Helens eruption in 1980, depositing 30 cm of ash on Kodiak Island. The effects of such ash deposition on marine life were undocumented but may well have included a substantial and prolonged period of reduced primary production in 1912, elevated temperatures, ash abrasion of delicate tissues such as fish gills, and burial of benthic communities.
3.3.5. Tsunamis Large earthquakes and some volcanic eruptions induce an extremely large force on the ocean, which can generate tsunamis. Tsunamis of up to 50 meters have inundated
Agents of Ecosystem Change
187
coastal areas, moving large amounts of sediments and reshaping coastal habitats. The 1964 earthquake in Prince William Sound created a huge tsunami that had widespread effects in conjunction with direct effects of the quake on landforms. This wave destroyed the native village of Chenega in Prince William Sound and killed many of its citizens. An eruption of the Augustine volcano in Cook Inlet in 1983 generated a tidal wave of nine meters at Port Graham on the Kenai Penisnsula.
3.3.6. Glaciers Over thousands of years, the numerous glaciers bordering the Gulf of Alaska have exerted dramatic effects on its shelf and coastal habitat. The substrate composition and topography of the continental shelf everywhere reflects glaciation to some degree. During the Pleistocene epoch, glaciers extended all the way to the edge of the current continental shelf at least once, and relict glacial sediments now litter their paths of advance and retreat. The current flora and fauna of the Gulf, including the distribution of species and their genetic makeup in some cases, reflects the huge influence of past glaciations as it divided populations and limited or prevented genetic interchanges. Even today, the continued retreat of remnant Pleistocene glaciers bordering the Gulf has provided new habitat, most notably for sockeye salmon which colonize newly formed lakes, but also for some species tied to tidewater glaciers whose populations are shrinking, e.g., Kittzlitz’s murrelets. Successive studies carried out in areas of rapidly retreating tidewater glaciers in Alaska reveal differing processes in terrestrial, freshwater and marine environments (Sharman et al., 1995). The marine communities developing on newly created habitat seem to develop to maturity in about a decade. In most cases, glacial effects occur very gradually. Animal populations can adjust to slow changes, although loss or gain of substantial habitat will have long-term effects. In some cases, however, the abnormally fast advance of a glacier has blocked fjords from tidal exchange and resulted in the stranding of marine animals, such as harbor seals, inside fresh, newly created lakes.
3.4. Species Interactions Gordon H. Kruse 3.4.1. Introduction Agents of ecosystem change not only include natural (e.g., climate and disease) and human (e.g., oil spills, fish harvests) perturbations, but also the inner workings of the ecosystem itself – species interactions. Exotic species can disrupt marine ecosystems.
188
Long-Term Ecological Change in the Northern Gulf of Alaska
For example, in some areas of the Mediterranean Sea and elsewhere, the exotic aquarium algae (Caulerpa taxifolia) now carpets the seafloor, radically altering the species composition of benthic communities. In many coastal water bodies, such as San Francisco Bay, many indigenous fauna have been replaced with newly introduced species. However, beyond such dramatic examples of ecosystem change through species interactions, changes in abundance and distribution of an existing species can cause changes in abundance of other species in the marine ecosystem through a set of complex and dynamic species interactions. In an ever-changing world, one important challenge for marine ecologists and fishery scientists is to attempt to separate the contributions of natural and human perturbations to the observed changes in the marine ecosystem so as to manage human harvest, waste disposal and other uses. This detective job is difficult owing to complex species interactions, which may dampen or amplify the magnitude of the effect, delay responses as changes are propagated through the system, or produce unanticipated outcomes through indirect effects. Our history of species introduction has shown us that the outcomes of species interactions are often unpredictable, and almost always negative. There are several main ways in which populations interact with each other (Pianka, 1974). Predation is one population eating another. The predator population benefits as prey consumption contributes to some combination of increased growth and reproduction and reduced mortality (less starvation). The predators generally affect the prey population adversely. Parasitism results in the same direction of interaction (positive for the parasite, negative for the host) as predation, but the predator is often exploited by the parasite over an extended period of time. Disease, a special case of parasitism, is treated separately in Section 3.6. Competition is when populations compete for the same resources (e.g., prey, habitat), which are in limited supply, with adverse effects to one or both competing populations. Whereas predation can be easily confirmed by sampling the gut contents of the predator for the prey, for example, proving that competition occurs is not so simple. If two predators eat the same prey, this is a necessary but insufficient condition for competition. Knowledge about whether prey are in short supply, particularly when predators eat multiple prey species, is the Achilles heel of many studies of prey competition. Commensalism is when the benefits accrue to only one species (the commensal), and the host is unaffected. For example, some species of juvenile fish, such as walleye pollock (Theragra chalcogramma), associate with jellyfish medusae, perhaps for protection from predators without adverse effects on the medusae. A better-known example is the remora, which attaches to large fishes, such as sharks for a free ride and food scraps when its host captures prey. Amensalism occurs when one species adversely affects another without itself being affected. Finally, neutralism occurs when neither population affects the other.
Agents of Ecosystem Change
189
The type of interaction between two populations or two species may depend upon the life stage. For example, the jellyfish medusae, which afford protection to some stages of juvenile pollock, eat younger (eggs) pollock and appear to compete with older pollock. Likewise, the role of who eats whom may reverse. Consider, for instance, Pacific cod (Gadus macrocephalus) and Pacific halibut (Hippoglossus stenolepis). Cod consume small (1–6 cm standard length) halibut (Lang et al., 2003), yet larger (>20 cm fork length) halibut consume Pacific cod (Yang, 1995). Pacific cod are well-known to eat a variety of invertebrates (e.g., shrimps, polychaetes worms, amphipods, crabs) and fish (e.g., pollock, flatfish), but their diet tends to shift from small invertebrates to fishes (i.e., they become more piscivorous) with increasing cod size (Clausen, 1981; Jewett, 1978). To complicate things further, species such as pollock, cod, red king crab, and Tanner crab, are cannibalistic – eating their own kind – thus providing positive and negative interactions to itself, depending on the life stage being considered. Marine ecosystems are comprised of communities of organisms, in which each pair of populations interact by the above mechanisms. To examine species interactions as agents of ecosystem change, a more holistic understanding of species connectedness must be first developed. Therefore, we briefly reintroduce some concepts that were described earlier in Section 2.4 for describing transfer of matter and energy. Here we introduce them in the context of the interaction between species. Marine ecosystems are composed of tangled food webs reflecting complex linkages of predators and prey. Food chains and food webs are convenient graphical depictions of the flow of energy, carbon and nutrients throughout an ecosystem. A simplified marine community model provides a conceptual basis for understanding the flow of this energy throughout the marine ecosystem (Fig. 3.6). Primary producers, called autotrophs, utilize nutrients and capture energy from the sun, which they can transform into plant biomass. In the sea, autotrophs include a diversity of species from microscopic plants (called phytoplankton) to large kelps that form impressive “forests” along rocky shorelines. The primary producers represent the first trophic level (level 1). Plant-eating species, called herbivores, are primary consumers and together constitute the second trophic level (level 2). Carnivores that eat only herbivores are secondary consumers (or primary carnivores) and form the third trophic level (level 3). Carnivores that eat primary carnivores represent the fourth trophic level, and so on. The wastes, excretions and carcasses are recycled to carbon dioxide, water and nutrients by decomposers. Taken together, all species other than the primary producers are termed heterotrophs. Real marine ecosystems are much more complex than that shown in Fig. 3.6, which groups organisms according to their relationship to the fixation of sunlight. Each box in the model is a complex mixture of populations, some of which occupy more than one box. Some predators eat a mixture of plants and herbivores and still others eat a
190
Long-Term Ecological Change in the Northern Gulf of Alaska
Figure 3.6: A conceptual model showing function and trophic level assignment of organisms in a simple marine community. The direction and width of arrows indicate the path and rate of energy flow, and the size of the boxes indicate the relative biomass of the different taxa. The figure was modified from Pianka (1974).
range of herbivores, primary carnivores, and secondary carnivores, thus making it very difficult to assign any given species to a particular trophic level. Moreover, marine ecosystems are not static as suggested by Fig. 3.6. Populations fluctuate in abundance and distribution, and these fluctuations may alter the strength and nature of the interactions of species pairs, with propagated implications on other connected species in the system. For instance, primary and secondary production is closely linked to physical oceanography, thus the changing climate can result in changes in the availability of certain types of lower trophic level prey for some portions of the food web. On the other hand, a period of warm ocean temperatures may allow a predator with a southerly distribution to extend its range north to facilitate predation of a prey population that would not be possible in cool years. In Chapter 6, we will discuss species interactions that may propagate changes in lower trophic levels to the upper trophic levels, changes in upper trophic levels that may be propagated downwards, and potential mechanisms for both types of effects that operate either synchronously or in an alternating pattern of dominance.
Agents of Ecosystem Change
191
Figure 3.7: Portion of the marine food web that includes harbor seals, sea otters and transient killer whales.
A portion of the complex food web in the Gulf of Alaska is depicted in Fig. 3.7, which includes some of the relationships between species related trophically to harbor seals. This figure shows both, some of the main species that pass energy to the harbor seals (e.g., copepods, herring, capelin and octopus) and their main predator (transient killer whales). One can see the potential for shifting dynamics within such a complex web of trophic relationships as various populations wax and wane and the changes reverberate in related populations as a function of shifting predation patterns. Understanding not only the internal dynamics of such complex webs, but also predicting the result in the changes of one population, for example resulting from a fishery for herring, is one of the most challenging aspects of both marine ecology and ecosystem-based management of large marine systems.
192
Long-Term Ecological Change in the Northern Gulf of Alaska
3.5. Marine Mammal Harvest and Fishing Gordon H. Kruse and Alan M. Springer 3.5.1. Introduction Alaska has abundant populations of many marine animals and, in recent times, management agencies have earned the reputation for their responsible management. Most groundfish and salmon stocks, the two most economically important groups for commercial fishing, have been sustained at high abundance levels since the late 1970s. In 2002, Alaska accounted for 5.1 billion pounds (54%) of a total 9.4 billion pounds of commercial fisheries landings into U.S. ports. Alaskan landings were worth $812 million (26%) of a total of $3.1 billion in ex-vessel value in the U.S. Landings of groundfish alone caught off Alaska accounted for 4.6 million pounds worth $566 million ex-vessel (paid to fishermen) and approximately $1.5 billion after primary processing. In addition, Alaska accounted for 523.1 million pounds (92%) of the U.S. total commercial landings of Pacific salmon. Despite these impressive statistics and healthy stocks of most groundfish and salmon, there have been dramatic declines of some fishes, crabs, shrimps, and marine mammals in Alaska that can be attributed directly to excessive harvests by man. In this section, we review the history of exploitation of living marine resources in the northern Gulf of Alaska, focusing on marine mammals, fish and invertebrates. For mammals, we highlight declines of great whales, harbor seals (Phoca vitulina), Steller sea lions (Eumetopias jubatus) and sea otters (Enhydra lutris) that were killed in vast numbers by fur traders, bounty hunters, fishermen, and others. Sea otters were the targets of fur traders beginning as early as the mid-1700s and great whales were exploited primarily for their oil for over a hundred years beginning in the mid-1800s. We recount the history of commercial fishing in Alaska, emphasizing groundfish, shellfish, salmon, and herring fisheries, and fishery management, including periods when fisheries were not well managed. Finally, we discuss direct effects of fishing on exploited populations and conclude with a discussion of more subtle, but potentially more important, indirect effects of fishing on marine species.
3.5.2. Marine Mammal Harvests and Persecution Great Whales Industrial whaling in the northern North Pacific began in the early 1840s when right whales (Eubalena glacialis) were discovered on their summer feeding grounds in the GOA. The number of American whaling ships working in the northern Gulf increased from just a few in 1840 to 108 by 1843, 292 by 1884, and 300–400 off the Kodiak
Agents of Ecosystem Change
193
grounds in 1846–1851 (Scarff, 1991; Gilmore, 1978). Some 11,000 were killed in the first decade of hunting and this onslaught proved disastrous for right whales. Only 3000 were taken in the next 10 years because the stock was so depleted and because bowhead whales were discovered in the northern Bering Sea as whalers began exploring elsewhere. Although right whales were afforded world-wide protection in 1935, they continued to be killed illegally, particularly by the Soviets, which drove them nearly to extinction (Rice, 1974; Wada, 1979; Gambell, 1976). By the early 1900s, whalers in the GOA were operating primarily out of shorebased stations in British Columbia, Southeast Alaska, Kodiak Island, and Akutan and were killing some of all the larger species, including humpback (Megaptera novaeangliae), blue (Balenoptera musculus), fin (Balenoptera physalus), sei (Balenoptera borealis), and sperm whales (Physeter macrocephalus). Harvests in many years were substantial but varied regionally. For example, it was estimated that 4000–5000 humpbacks were killed in Alaska and British Columbia between 1905 and 1910 (Rice, 1978). The harvest remained high during the 1910s, with over 3500 killed in the northeastern Pacific (Washington to Southeast Alaska, primarily British Columbia) and over 700 killed in the northern GOA. The take in British Columbia declined rapidly during the early 1920s to a total for the decade of about 650. However, this reduction in effort was offset by a dramatic increase in the mid-to late 1920s in Alaska to a total harvest of over 2100 for the decade (Rice, 1978). For other species, tens to low hundreds were taken annually through the 1930s until the beginning of World War II. As a result, stocks of some species declined noticeably, particularly those of humpbacks and blues, but not as disastrously as they would during the ensuing whaling era beginning in the late 1940s (Gregr et al., 2000; Springer et al., 2006). The coup de grace was applied to North Pacific great whales following World War II, when surplus vessels from Japanese and Soviet navies were converted to high-speed catcher boats and floating factory processors made it possible to kill and process vast numbers of whales at sea. On the order of 400,000 great whales were killed in the North Pacific between 1948 and 1979, including more than 250,000 sperm whales, 60,000 sei whales, and 45,000 fin whales. These totals were reported to the International Whaling Commission (IWC) by whaling nations, but are minimum estimates as it is known that the former Soviet Union under-reported their take, by as much as 60% in the case of sperm whales (Brownell et al., 2000). Whales were eventually protected by the IWC from pelagic whaling, but not until most species were greatly depleted. By the time blue and humpback whales were protected in 1965, there were only about 1000 to 1500 left in the entire northeastern Pacific (Rice, 1978; Mizrock et al., 1984). Fin and sei whales were not protected until 1975, and sperm and Bryde’s whales (Balenoptera edeni) in 1979, although whaling was closed in 1978 in the U.S. EEZ (Exclusive Economic Zone, or waters within 200 miles of land) with the enactment of the Magnuson Fishery Conservation and
194
Long-Term Ecological Change in the Northern Gulf of Alaska
Management Act. Estimates of the abundance of these species are less reliable, but there is no doubt that they were severely depressed. For example, in 1980, in an area of approximately 220,000 square kilometers of the northern GOA, which formerly supported thousands of whales, population estimates from a whale survey were: fin, 159; humpback, 364; sperm, blue, sei, and right, few (only 36 sperm whales and none of the other species were sighted) (Rice and Wolman, 1982). Although whales were harvested over a large area of the North Pacific, the majority were killed while concentrated on their summer feeding grounds. Thus, the removal of whales affected small geographic areas disproportionately (see Fig. 3.8). The greatest harvest of whales in the GOA in the modern era was in the 1960s. (Fig. 3.9).
Figure 3.8: The harvest of great whales in the North Pacific Ocean by species, 1924–1987. Reproduced from Springer et al., 2006.
Agents of Ecosystem Change
Figure 3.8: Continued.
195
196
Long-Term Ecological Change in the Northern Gulf of Alaska
Pinnipeds Sea lions and seals have been harvested in Alaska since indigenous peoples first inhabited coastal areas. Archeological sites reveal that Alaska Natives harvested sea lions for at least 3000–4000 years (Laughlin, 1980). Historical harvests have been documented in the northern Gulf of Alaska, including Prince William Sound, Kenai Peninsula, and Kodiak Archipelago (Haynes and Mishler, 1994). At four wellpreserved sites in the eastern Aleutian Islands, the estimated biomass of harvested animals was dominated by Steller sea lions (70.4%), followed by seals (12.2%), and sea otters (3%). Of course, the use of marine mammal species elsewhere depended on availability. Traditional uses included meat (food), hides (kayaks), flippers (boot soles), stomachs (boot uppers), intestines (rain coats), and bladders (fishing floats and sacks for oil storage). Since the settlement of Alaska by Russians and non-native Americans, Steller sea lions and seals (primarily harbor seals) were killed for various reasons. During the 1800s, large numbers of sea lions of the western stock were killed by Russians in the Aleutian Islands for their furs and skins. They were also killed on the Pribilof Islands for their furs and to reduce their abundance and open up space on beaches for fur seals. The later killing was apparently responsible for a large decline in sea lions by the end of the century (NRC, 2003). Another purpose for killing sea lions was to provide cheap food for fox farms. Fox farming in the GOA began in the Semidi Islands (southwest of Kodiak) in the 1880s, and by the 1890s, operations were established on islands in the Kodiak Archipelego and Prince William Sound. By 1919 there were 19 fox farms in Prince William Sound alone. At the height of the fox farming era, 485 farms were operating in coastal areas of Southeast and Southcentral Alaska, where salmon, sea lions, harbor seals, and porpoises provided plentiful feed. There is a long history of fishery conflicts with marine mammals in Alaska. At least as far back as the 1880s, seals and sea lions were reported to “prey upon cod, frequently taking them from the line” around Kodiak (Bean, 1887). Sea lions and seals take fish from fishing gear, such as sablefish and halibut longlines, salmon and herring gillnets, salmon troll gear, and groundfish trawls (Hoover, 1988a,b). In addition to taking catch, seals and especially sea lions damage nets and other gear. Sea lions puncture inflatable crab pot floats, causing them to sink, resulting in lost gear and catches. As a sea lion counter measure, solid-core Styrofoam “sea lion” buoys were developed to keep crab lines afloat. Because of the real and perceived losses of revenue due to damaged gear and depredations of commercial fish, the Territorial Legislature of Alaska started a predator control program for seals in 1927. As part of this program, a bounty ($2 during 1927–1938, $3 during 1939–1958) was paid for each seal scalp. In the 1930s and 1940s, the annual harvest of harbor seals in Alaska statewide ranged from 6000 to
Agents of Ecosystem Change
197
Figure 3.9: Geographic distribution of whale harvests (all species) in the North Pacific Ocean and Bering Sea from 1946 through 1976. Data (latitude/longitude and number of individuals). Data are from International Whaling Commission records of reported whale kills and are binned by sequential 7- to 8-year intervals to show temporal trends. Reproduced from Springer et al., 2003.
10,000, but increased from 12,000 to 24,000 in the late 1940s and early 1950s (Hoover, 1988a). During 1927–1958, $1.2 million in bounties was paid for a total of 358,023 “hair” seals, predominantly harbor seals (Lensink, 1958). In addition to bounties paid to members of the public, the Alaska Department of Fisheries, the precursor to the Alaska Department of Fish and Game (ADF&G), employed hunters to kill animals with rifles and “depth charges” (Fig. 3.10, see Box 3.2). During 1951–1958, 50,000 seals were killed at the Copper River Delta alone. The Federal government also was involved in predator control, particularly for Steller sea lions. The total estimates of sea lions killed are unavailable, but it appears that generally only a small fraction of the total population was removed. However, kills on individual rookeries in some years were large. For instance, agency hunters killed nearly all pups on Sugarloaf Island (central Gulf of Alaska) in two years, once in the 1950s and once in the 1960s (NRC, 2003).
198
Long-Term Ecological Change in the Northern Gulf of Alaska
By the late 1950s, bounty programs were abandoned as being expensive and ineffective and experimental harvests were attempted as a more cost-effective mechanism to control sea lion populations during 1959–1972. In 1959, 616 sea lions, mostly males, were killed from Kodiak to the eastern Aleutian Islands (Thorsteinson et al., 1961). During 1963–1972, 45,178 pups were killed (Merrick et al., 1987); of these, 16,763 and 14,180 were taken from Sugarloaf and Marmot Islands, respectively, in the central Gulf of Alaska. After the bounty program ended, fishermen still shot seals and sea lions, as well as harbor porpoises and Dall’s porpoises, while in transit to and from the fishing grounds (Ed Opheim, former cod and salmon fisherman, Kodiak, personal communication, September 2001). Some fishermen admit that large numbers of Steller sea lions were shot in association with the pollock fishery in Shelikof Strait during the years of the joint-venture, roe-stripping fishery that attracted sea lions to fishing areas. Crab fishermen shot sea lions in the 1960s and 1970s to prevent loss of crab floats, but also some fishermen reportedly used seal and sea lion meat as bait in crab pots. Salmon fishermen shot seals and sea lions in the belief that they were responsible for poor salmon runs during the 1930s–1950s, and many others shot them in attempts to prevent loss of salmon from gillnets. The number of marine mammals shot is sparsely documented. In the spring of 1954, it was estimated that salmon trap operators shot and killed 816 Steller sea lions in the Kodiak and Alaska Peninsula areas (Thompson et al., 1955). The Marine Mammal Protection Act of 1972 was an important landmark, although even after 1972, fishermen were allowed to continue to shoot animals that were destroying gear, and they did. For example, approximately 305 sea lions were shot during the drift gillnet fishery at the Copper River delta in 1978 (Angliss and Lodge, 2004). Not until 1990, when Steller sea lions were listed as “threatened” under the Endangered Species, did it become illegal under any circumstances to discharge firearms near sea lions. The Copper River fishery was monitored more recently and no mortalities were observed in 1990 and only two were recorded in 1991. Likewise, no deaths were found during observations on drift gillnet and set gillnet fisheries for salmon in Cook Inlet in 1999 and 2000 (Angliss and Lodge, 2004). Recent court cases and anecdotal information indicate that illegal shooting of Steller sea lions continues despite these laws, but shooting is much reduced now. Furthermore, marine mammals were also shot for sport. For instance, there are anecdotes of sea lions being shot by American war planes stationed in Alaska during the 1940s. According to Ed Opheim, “the big PBYs ... they’d come through the narrows ... a hundred feet off the water and they’d open up those big fifty caliber guns shooting at the sea lions on Triplet Islands. I was sitting there on the beach ... and here are these tracers hitting the sea lions and the bluffs with great big sparks flying. The sea lions were dropping off the top of the islands. And man, you talk about slaughter” (NRC, 2003).
Agents of Ecosystem Change
199
The negative effects of government sponsored control programs and shooting by fishermen and others were exacerbated during the 1960s when a commercial fur market developed, peaking in the middle of the decade with statewide harvests of harbor seals and other hair seals reaching 40,000–60,000 per year. The harvest of harbor seal pups at Tugidak Island alone totaled about 16,000 between 1964 and 1972 (Pitcher, 1990). Marine mammals also were inadvertently captured by purse seines, gillnets, and trawls (Hoover, 1988a). In 1978, 312 harbor seals died in the salmon gillnet fishery off the Copper River from a combination of entanglement and shooting of animals in the vicinity of gear. More recently, deaths are much rarer. Only two mortalities were documented in 1990 and one death in 1991. Generally, incidental fishery catches of harbor seals appear to be low. Based on observer data for gillnet, trawl and pot fisheries, an extrapolated total minimum annual mortality of harbor seals for the entire Gulf of Alaska was estimated recently to be 36 animals (Angliss and Lodge, 2004). Steller sea lions are captured by a variety of fisheries as well. Most reports of sea lion deaths involved past trawl fisheries. Based on a 10% observer coverage of foreign trawlers, the estimated annual mortality of sea lions was 724 during 1978–1981 (Loughlin and Nelson, 1986). This estimate does not include animals that may have been taken in domestic or joint-venture fisheries involving domestic fishing vessels and foreign processing vessels. In 1980, a joint-venture fishery was developed to harvest pollock in Shelikof Strait, and an estimated 1211–2115 sea lions were incidentally caught and killed over three years, 1982–1984 (Perez and Loughlin, 1991). During this time, at least two aspects of this pollock fishery led to exceptionally high numbers of sea lion deaths. First, the fishery involved roe-stripping and discarding carcasses, attracting sea lions to the area. Roestripping was banned in 1990. Second, trawls were retrieved by catcher vessels and were towed at or near the surface, capturing sea lions in the process. Aside from this particular example, incidental capture of sea lions by fisheries is low; a recent estimate is 26 animals for the region encompassed by the western stock of Steller sea lions, from the central Gulf of Alaska to the Aleutian Islands and Bering Sea (Angliss and Lodge, 2004).
Sea Otters The historic distribution of sea otters included shallow-water, ice-free habitats around the North Pacific Rim between northern Japan and central Baja California, Mexico. Factors that govern sea otter populations in the absence of human influence are poorly known. Evidence from human midden sites in the Aleutian Islands indicates that sea otter populations fluctuated markedly over a period of 2500 years, presumably in response to human exploitation (Simenstad et al., 1978).
200
Long-Term Ecological Change in the Northern Gulf of Alaska
BOX 3.2: PREDATOR CONTROL OF SEALS USING DEPTH CHARGES by Gordon H. Kruse Because the magnitude of the “seal problem” was so large that hunting by rifles proved ineffective, a novel method using dynamite “depth charges” was employed beginning 1951 (Fig. 3.10). Results were immediately successful, and the approach was used as part of the government’s predator control program for seals in the 1950s. According to The Alaska Fisheries Board (AFB) and Alaska Department of Fisheries (ADF) (1954), “When a seal herd was found hauled out on a bar or island, the skiff (with two men aboard) was loaded with depth charges and prepared for the bombing run. The best means of approaching the seals was decided upon, keeping in mind the characteristic behavior of the animals to quickly take fright and move into the water en masse as the boat advances toward them. The bombing run was then started, with the skiff traveling as fast as possible. When over the seals, which ideally were in fairly deep water but still well concentrated, the explosive charges were tossed overboard, the skiff continuing to move at a fair speed. The fuses, ignited by slip-on fuse lighters of the pull-wire type, were of such length that the charges sank for several seconds before detonating. Seals that were only injured by the blasts were promptly dispatched with shotgun fire when they surfaced. Most of the dead seals sank, but those which could be recovered were opened for stomach examination; their scalps were also removed and destroyed to avoid improper bounty claims.”
Figure 3.10: Design of depth charges for killing harbor seals. From AFB and ADF (Alaska Fisheries Board and Alaska Department of Fisheries). (1954 Annual Report. Alaska Fish and Game Commission and Alaska Department of Fish and Game, Report 6, Juneau, Alaska.)
Agents of Ecosystem Change
201
Sea otters were nearly exterminated from the North Pacific by the fur trade, beginning with the voyage of Vitus Bering in 1741 and continuing unchecked until 1911 (Kenyon, 1969). During that time, a minimum of 368,151 sea otter skins were exported from Alaska, with upper estimates on the order of 900,000 (Lensink, 1960). There were two distinct periods of harvest, with one reaching its peak about 1800 and averaging about 15,000 per year, and a second peak about 1870, averaging about 4000 per year. This bimodal harvest pattern probably represents two distinct periods of overexploitation separated by a brief period of population recovery. By the time sea otters were protected by the International Fur Treaty in 1911, there were only 11 geographically isolated populations between California and the Kuril Islands in Russia (Kenyon, 1969). At least two populations remained in the Aleutian Archipelago, three along the Alaska Peninsula, one at Kodiak Island and one in Prince William Sound. These remnant groups provided the nucleus from which the population recovered during the ensuing century. Important refugia in the GOA included Sanak I. and the Sandman Reefs in extreme western GOA; the Shumigan Is.; a few locations along the Alaska Peninsula, particularly Sutwick I.; the Kodiak I. Archipelago, particularly Afognak I., Shuyak I., and the Barren Is.; Southwest Prince William Sound; and possibly Kayak Island (Fig. 3.11).
3.5.3. Fisheries of the Northern Gulf of Alaska Introduction In their report to the nation, the Pew Oceans Commission (2003) noted that Alaska’s fisheries were “arguably, the best managed fisheries in the country. With rare exceptions, the managers have a record of not exceeding acceptable catch limits set by scientists. In addition, the North Pacific Fishery Management Council (NPFMC) and Alaska Board of Fisheries (BOF) have done more to control bycatch and protect habitat from fishing gear than any other region of the nation.” As an example, in contrast to the endangered populations of Pacific salmon of the Pacific Northwest and Atlantic salmon inhabiting waters of eastern North America and Europe, Alaskan salmon populations continue to be maintained at high and healthy levels since the early 1980s. In 2000, the Alaskan salmon fishery became the only salmon fishery in the world to be certified as “sustainable” according to the rigorous environmental standards by the Marine Stewardship Council. Additionally, in 2004 the Marine Stewardship Council certified the Alaska pollock fishery – the largest fishery in the U.S. – as sustainable. In making its recommendations, it was noted that the Bering Sea/Aleutian Islands pollock fishery is one of the best managed fisheries in the world.
202
Long-Term Ecological Change in the Northern Gulf of Alaska
Figure 3.11: Refugia (red dots) for remnant sea otter populations after the Russian harvests of the eighteenth and nineteenth centuries.
Yet, despite such praise and today’s conservative fishery management, such was not always the case. In this section, we provide a brief history of fishing in the northern Gulf of Alaska. We emphasize commercial fisheries, while recognizing the social and economic importance of subsistence and recreational fisheries to the region. Fishery management strategies have evolved and continue to evolve. The passage of the Magnuson Fishery Conservation and Management Act of 1976 (now known as the Magnuson–Stevens Fishery Conservation and Management Act) resulted in major changes in the management of many fisheries. A detailed history of fishery management practices is well beyond the scope of this review. However, in providing this short synopsis, we identify historical periods when overfishing occurred, and we discuss contemporary fishery management strategies and why they led to improved conservation of fishery resources of Alaska. Finally, we conclude with discussions about the direct and indirect effects of fishing that, taken together, indicate that a sense of complacency with existing levels of knowledge about the marine ecosystem and current management strategies could be quite premature.
Agents of Ecosystem Change
203
3.5.4. History of Fishing and Fishery Management in Alaska Alaska’s fisheries are often lumped into five broad categories: salmon, herring, groundfish (also called bottom fish or demersal fish), halibut, and shellfish. However, taxonomically, Pacific halibut (Hippoglossus stenolepis) are as much of a groundfish as any other flatfish (e.g., yellowfin sole (Limanda aspera), rock sole (Lepidopsetta bilineata), Greenland halibut (Reinhardtius hippoglossoides)) in the group. Actually, the invertebrates of the “shellfish group” are more taxonomically and ecologically diverse than the four other fish groups combined. Of course, the groupings are largely a matter of convenience, perhaps based mostly upon differences in fishery management governance, which we describe below under “Contemporary Fisheries Management.” Here, a brief history of fisheries and fishery management in the Gulf of Alaska is provided, drawing largely upon reviews provided by Alverson (1992); Alverson et al. (1964); Kruse et al. (2000); NMFS (2002); Rigby et al. (1995); Shirley and Kruse (1995); Thompson and Freeman (1930) and other sources.
Salmon Salmon have provided subsistence to Alaska Natives for thousands of years. Five Pacific salmon species occur in the Gulf of Alaska: pink (humpy) (Oncorhynchus gorbuscha), sockeye (red) (Oncorhynchus nerka), chum (dog) (Oncorhynchus keta), coho (silver) (Oncorhynchus kisutch), and chinook (king) (Oncorhynchus tshawytscha) salmon. Commercial fisheries began in the 1880s. Initially, most harvests were salted, but canning became prevalent with the turn of the twentieth century. When Alaska was purchased from Russia in 1867, the federal government assumed jurisdiction over salmon (and other) fisheries. At that time, most salmon were caught in large fish traps. Statewide salmon catches grew to over 102 million fish in 1918, but plummeted to just 36 million in 1921. Fishery conservation became an issue prompting the White Act of 1924 that required, among other conservation measures, 36-hour weekend closure of the salmon fishery and equal division of salmon runs into catch and escapement (fish allowed to spawn). A major problem at this time, however, was that few assessment programs were available to implement the 50% exploitation rate. Also, under federal management, seasons and areas fished were specified in the regulations and the management system was too cumbersome to reflect in-season changes in salmon distributions and run strength. Statewide salmon catches grew to 126 million fish in 1936, but after a couple of decades, decline ensued with only 25 million fish caught in 1959. A number of salmon runs are thought to have been overfished during the 1920s–1950s. The perceived need for state control of fishery management became the leading argument for statehood, which was enacted in 1959. As further indication of the importance of fisheries in Alaska becoming a state, one whole section of Alaska’s constitution is devoted solely to resource management.
204
Long-Term Ecological Change in the Northern Gulf of Alaska
After statehood, fish traps were banned. ADF&G implemented stock assessments, as well as a management system designed to assure escapement adequate to sustain salmon runs and fisheries. An effective state fish and wildlife enforcement program was created. Catches averaged 53 million fish during the 1960s. However, by the early 1970s, it was recognized that biological controls of harvest were insufficient and the Commercial Fisheries Entry Commission was created to limit fishing effort (number of permits). As an additional step to enhance natural production, a salmon hatchery program was initiated. In 2003, hatcheries released 1.5 billion young salmon and 50 million hatchery-reared adults were caught during commercial salmon fisheries in Alaska. That year, the hatchery catch was 28% of the total salmon harvest (in numbers of fish), but it must be kept in mind that hatchery runs are exploited at a much higher rate than wild runs. These conservation and enhancement efforts, combined with favorable climate, resulted in decades of healthy salmon populations and harvests, which averaged 149 million fish annually over 1980–1999. As an exception, in Prince William Sound runs of wild pink salmon declined in the late 1980s, coincident with large increases in hatchery runs. In recent years, these wild runs are relatively stable and at levels comparable to those prior to the late 1970s regime shift; hatchery fish account for about 80% of total pink salmon runs in Prince William Sound. Statewide, most salmon are caught by gillnet, purse seine (Fig. 3.12) and troll fisheries, and ADF&G manages these fisheries through time–area openings by emergency order to achieve escapement goals and allocations among users identified in fishery management plans approved by BOF. Stock assessments are based on in-season fishery monitoring and comprehensive salmon escapement enumeration programs. The latter include aerial counts, foot counts, weir counts, sonar counts, tower counts, and mark/recapture experiments. A sustainable salmon fisheries management policy, based on the attainment of escapement goals, underpins Alaskan salmon fishery management. The Pacific Salmon Commission, established by treaty between Canada and the U.S., provides regulatory advice concerning salmon originating in the waters of one country and which are subject to interception by the other. Also, the NPFMC has a federal fishery management plan that, in essence, defers much salmon fishery management to the state while extending state regulations to federal waters. West of Cape Suckling (144 W°, located between Yakutat and Prince William Sound), the federal plan closes federal waters to salmon fishing except for specific areas off the Copper River, in Cook Inlet, and south of the Alaska Peninsula. East of Cape Suckling, troll fisheries in federal waters are managed by the state under the auspices of the Pacific Salmon Treaty. One exception to this management paradigm is subsistence fisheries on federal lands, which are managed by the U.S. Fish and Wildlife Service. In contrast to the first half of the twentieth century, the most pressing issue today for Alaska’s salmon resources is low salmon prices, not resource conservation, largely due to greatly increased worldwide production of farm-reared salmon, and perhaps partly due to the hugely successful fishery management that, along with favorable ocean
Agents of Ecosystem Change
205
Figure 3.12: Salmon purse seiner operating in Galena Bay, Prince William Sound, Alaska, 1968. (Photograph courtesy of NOAA, J. M. Olson).
conditions since the mid-1970s has sustained high salmon abundances and catches year after year.
Herring Herring harvests by Alaska Natives predate recorded history. Herring are normally fished during spring when they spawn along shorelines. Traditional products include dried herring and harvests of spawned eggs on kelp, other seaweeds, and on weighted spruce and hemlock branches. Commercial fishing began in 1878. Most herring were rendered for oil at reduction plants from the late 1800s through the 1950s. Catches from Southeast Alaska dominated the reduction fishery through the early 1920s (Fig. 3.13). During the 1930s, harvests declined sharply in Southeast Alaska and the fishery shifted to Prince William Sound and the Kodiak area. The reduction fishery peaked at 142,000 mt in 1934 and thereafter catches generally declined through 1960, after which the reduction fishery was phased out, owing partly to displacement by the development of a huge reduction fishery for anchoveta off Peru. Although stock assessments were not conducted at the time, it is widely believed that there was a sequential (counterclockwise around the Gulf of Alaska) overfishing of herring during the first half of the twentieth century. A large foreign (Russian and Japanese) trawl and gillnet fishery was prosecuted in the eastern Bering Sea during 1960–1980. Sharp declines of herring in the eastern Bering Sea were attributed to this fishery, resulting in its phase out after the establishment of U.S. jurisdiction to 200 nautical miles (nm). The contemporary fishery was developed in the early 1970s. There are 25 distinct herring fisheries in Alaska, most of which catch pre-spawning fish by gillnets and
206
Long-Term Ecological Change in the Northern Gulf of Alaska
Figure 3.13: Abandoned herring fishery reduction plant, Prince William Sound. (Photograph courtesy of NOAA, Mandy Lindberg).
purse seines for their valuable roe for Asian markets. A few small food and bait fisheries are also conducted. Harvests from these roe fisheries were relatively stable during 1980 to the mid-1990s. However, in the late 1990s, herring biomass and catches declined sharply in Prince William Sound, Cook Inlet, and Kodiak areas. In Prince William Sound, a sharp decline occurred in 1993–1994. Generally, catch quotas are set for each fishery based on a 20% harvest rate, which is thought to represent a conservative yield. Also, fishery thresholds are commonly set at 25% of estimated virgin or unfished biomass (i.e., theoretical levels that existed before any fishing); no fishing is allowed when that stock falls below that threshold. Yet, despite these harvest control rules, estimates of spawning biomass exist for just a few areas (mostly in Southeast Alaska) where ADF&G conducts spawn deposition surveys in which they estimate the amount of spawned eggs and from this estimate the abundance of spawning females in the population based on their average fecundity. Spawn deposition surveys conducted after the Exxon Valdez oil spill in Prince William Sound have been discontinued owing to agency budget cuts and herring stock collapse. Such surveys have never been conducted in Lower Cook Inlet or around Kodiak Island in the northern Gulf of Alaska. So, in most areas, data come primarily from aerial surveys of herring run timing and age and length samples from the commercial fishery. These data do not allow estimation of herring biomass. As a result of low abundance, fisheries have not been prosecuted in Prince William Sound since 1998 and lower Cook Inlet since 1999.
Agents of Ecosystem Change
207
Groundfish Prior to western settlement of Alaska, Pacific halibut (Hippoglossus stenolepis) were one of the most important fishery resources of coastal Natives, who also harvested other groundfish, such as Pacific cod (Gadus macrocephalus). Indigenous peoples developed highly sophisticated wooden hooks with barbs that were admired for their effectiveness by the early western fishermen in Alaska. The early domestic groundfish fisheries were confined to longline and hook-and-line fisheries for Pacific halibut, sablefish (Anoplopoma fimbria), and Pacific cod. In the 1860s, a domestic schoonerbased dory fishery for Pacific cod was developed in the Gulf of Alaska (Bean, 1887). In the late 1890s, a longline fishery was developed for Pacific halibut and sablefish, but it was originally confined to the inside waters of Southeast Alaska. Overfishing of halibut stocks in the Atlantic led to a high demand for Pacific halibut, and the completion of the transcontinental railroad in 1887 and improved refrigeration techniques caused the exponential growth of the Pacific halibut fishery that could now supply markets along the east coast. The fishery began to shift from a nearshore to an offshore fishery in 1910, as some fishing grounds began to show evidence of overfishing. The eastern Gulf of Alaska became exploited during 1913–1922 after the construction of larger vessels capable of fishing more remote areas. The development of the diesel engine in 1921, along with technologies for direct longlining (rather than dory-based operation), facilitated further expansion throughout the northern and western Gulf of Alaska. The Halibut Convention of 1923 between the U.S. and Canada was a monumental achievement, because it was the first treaty in the world for the conservation of a depleted deep-sea fishery. Trawl fisheries for other groundfish species, such as walleye pollock (Theragra chalcogramma), flatfish (e.g., yellowfin sole), and rockfishes (Sebastes spp.) (e.g., Pacific ocean perch) were pioneered by foreign fleets. In the eastern Bering Sea, Japan conducted the first experimental fishing in 1929, and Japanese fisheries occurred starting in 1930 and Russian fisheries began in the 1950s. Both nations began to fish the northern Gulf of Alaska in the early 1960s. Foreign trawl vessels overfished yellowfin sole in the eastern Bering Sea in the 1960s, and in the Gulf of Alaska, Pacific ocean perch (Sebastes aleutis) was severely depleted by overfishing in the 1960s and 1970s (see Fig. 3.1D). For comparison, in the Gulf of Alaska, the peak catch of Pacific ocean perch by foreign fleets was 350,000 mt in 1965 whereas in 2003 total allowable catch was only 14,000 mt for Pacific ocean perch and only 236,000 mt for all groundfish species combined. After depletion of perch in the 1960s and 1970s, foreign fleets switched to other groundfish species. Trawl bycatch of halibut equaled one-third of the total domestic longline halibut harvest in 1965, causing much concern to domestic halibut fishermen. As foreign groundfish landings grew through the 1970s, gear conflicts escalated between foreign trawlers and mainly U.S. domestic longline and pot fisheries for halibut, cod, sablefish and crabs. In part, these conflicts were used to justify the Magnuson Fishery Conservation and Management
208
Long-Term Ecological Change in the Northern Gulf of Alaska
Act of 1976 as a vehicle to phase out foreign fishing in U.S. waters and to “Americanize” the groundfish fisheries off Alaska. The Americanization process was greatly facilitated by a provision of the Act that established a three-tiered allocation of the total allowable catch, giving top priority to U.S. processors, second priority to joint-venture operations, and last priority to foreign fishing. A transition occurred after 1978 when joint venture fisheries were initiated and domestic fishermen operated catcher vessels that delivered to foreign processing mother ships. Following the late 1970s, the joint-venture fishery grew rapidly until the mid-1980s, but then was phased out as U.S. domestic processing came on line. All foreign and joint-venture fisheries in Alaska were ended by 1991. Most groundfish fisheries are federally managed by the National Marine Fisheries Service (NMFS) and halibut are managed under a bilateral agreement with Canada through the International Pacific Halibut Commission. In some instances, ADF&G manages groundfish fisheries within state waters under management plans adopted by the Alaska Board of Fisheries. In addition, the state has been given lead management authority over lingcod (Ophiodon elongatus) and black and blue rockfish (Sebastes melanops and S. mystinus) throughout all waters of the territorial sea (<3 nm) and EEZ (3–200 nm). Groundfish are typically managed using a conservative exploitation rate and stock abundances are monitored by routine assessment surveys (Fig. 3.14) and a catch and bycatch monitoring program that includes observers aboard many commercial fishing vessels. In the Gulf of Alaska, major species in this groundfish fishery include walleye pollock, Pacific cod, sablefish, halibut, 30 species of rockfishes, and many species of flatfish, including yellowfin sole, rock sole and flathead sole (Hippoglossoides elassodon). Management of Alaskan groundfish fisheries is recognized among the best programs in the world. As a result of this stewardship, no groundfish stocks have been overfished since the establishment of U.S. domestic groundfish fisheries in Alaska.
Shellfish In the 1930s, Japanese fishermen pioneered crab fishing in Alaska, focusing in the eastern Bering Sea, but efforts were also expanded in the Gulf of Alaska. Catches of red king crabs (Paralithodes camtschaticus), were recorded from lower Cook Inlet and around Kodiak in the 1930s. A domestic fishery for red king crab started in earnest off Kodiak in the late 1950s and landings peaked at 45,000 mt in 1965. In the northern Gulf of Alaska, smaller king crab fisheries were prosecuted in lower Cook Inlet and Prince William Sound. Early fishing practices allowed tangle nets, but these were quickly banned as wasteful because they captured crabs of all sizes and sexes. Pots became the prevailing gear type. As red king crab fisheries declined, fisheries developed for Tanner crab (Chionoecetes bairdi) in the late 1960s and early 1970s. Waters off Kodiak supported
Agents of Ecosystem Change
209
Figure 3.14: A 16,000 pound catch of pollock coming aboard the MILLER FREEMAN. (Photograph courtesy of NOAA).
the largest Tanner crab fishery in the Gulf of Alaska, with landings that peaked at 13,000–15,000 mt in the mid-to late 1970s. Fisheries developed for Dungeness crabs (Cancer magister) also in the 1960s. Important Dungeness fisheries include Southeast Alaska, Yakutat and Kodiak. Smaller Dungeness crab fisheries were conducted in lower Cook Inlet and Prince William Sound through the 1970s. Crab management strategies have evolved over time. In the early years, fishing seasons lasted much of the year and there were few other regulations besides size, sex,
210
Long-Term Ecological Change in the Northern Gulf of Alaska
and season in which only males above a certain size could be retained outside of the molting and mating period. As fishing effort increased and stocks began to decline, seasons were shortened, and quotas were established and reduced over time. In the 1980s, stock assessment programs were established and the management program evolved to resemble that for Pacific herring with a conservative harvest rate and fishery threshold. Today, many of the crab fisheries in the northern Gulf of Alaska are closed because stock sizes are below thresholds for fishing. In the northern Gulf of Alaska, all red king crab fisheries have been closed since the early 1980s, and all Tanner (except part of Kodiak) and some Dungeness crab fisheries (e.g., Prince William Sound and lower Cook Inlet) have been closed since the 1980s. Whether crab stocks were overfished remains a topic of debate. There is some evidence of overfishing of red king crab stocks off Kodiak Island in the late 1960s when heavy harvests of males, especially large males, skewed sex ratios causing unprecedented incidences of barren (unmated) females. However, aside from this instance, clear cases of overfishing have not been demonstrated and periodic recruitment seems to be driven by environmental conditions. Commercial fisheries have existed for a number of other invertebrate species. One of the largest was a fishery for northern (locally called “pink”) shrimp (Pandalus borealis). An otter trawl fishery developed in the northern Gulf of Alaska in the late 1950s, and landings from the central and western Gulf of Alaska peaked in 1973 at about 68,000 mt, half of which was caught off Kodiak Island. This trawl shrimp fishery collapsed in the late 1970s to early 1980s and has been closed in most areas since 1983. In contrast, a smaller fishery using beam trawls in Southeast Alaska started in 1915 and landings have been relatively stable for the past 50 years. Small pot fisheries occurred for spot (P. platyceros) and coonstriped shrimps (P. hypsinotus) in Prince William Sound and lower Cook Inlet through the 1990s. Currently, the only ongoing shrimp fisheries in Alaska include a small trawl fishery for sidestriped shrimp (P. dispar) in Prince William Sound, and small pot fisheries for spot and coonstriped shrimp in Southeast Alaska and Yakutat. A small dredge fishery for weathervane scallops (Patinopectin caurinus) is prosecuted by just a few boats under a license limitation program. This fishery developed in the late 1960s and effort increased quickly. Landings declined in the early to mid1970s partly due to localized depletions off Yakutat and Kodiak Island as catch rates declined and age compositions shifted to young animals. Landings increased sharply in the late 1980s and early 1990s, due to improved stock conditions as well as a surge in participation. Since the mid-1990s, harvests have been maintained at moderate levels by a small fleet, most of which fish in a fishery cooperative. Owing to the collapse of many crab and shrimp fisheries, small fisheries for other invertebrates (e.g., sea cucumbers, Parastichopus californicus; sea urchins, Strongylocentrotus spp.; and geoduck clams, Panopea abrupta) developed in the midto late 1980s. Many of these dive fisheries developed in Southeast Alaska, but some
Agents of Ecosystem Change
211
occur throughout the northern Gulf of Alaska. Although low in volume, high prices are typically paid for seafood products from these dive fisheries. For most fisheries in Southeast Alaska, stock assessment surveys are conducted and management plans are designed to be precautionary, including management measures such as conservative harvest rates, rotational harvests, and closed areas. An exception may be the pinto abalone (Haliotis kamtschatkana) fishery in Southeast Alaska, which developed in the early 1970s. Landings have been declining since the early 1980s, perhaps due to a combination of overfishing, predation by a growing sea otter population, and low productivity from depleted stocks. Unlike the dive fisheries that developed in the late 1980s, conservation measures were not put into place until after harvests had already attained high levels.
3.5.5. Direct and Indirect Effects of Fishing Natural populations fluctuate owing to varying climate, species interactions and other factors. Whether a fish population increases, decreases, or remains stable from one year to the next depends on the balance of reproduction, mortality, growth, immigration, and emigration. In many fish and invertebrate populations, reproductive success is the most important factor in determining changes in population abundance. Reproductive success depends on adequate numbers of parents, but favorable environmental conditions for early life survival are crucial as well. As mortality is greatest early in life, the abundance of a fish cohort surviving to sizes at which they can be counted in surveys or caught by a fishery is called recruitment strength, cohort size, or year-class success. Fishing causes a number of direct effects on exploited populations. The most obvious effect is a reduction in the fish abundance (numbers) and biomass. Population dynamics theory suggests that, on average, reductions in parental fish abundance to moderate levels either has little effect on subsequent recruitment or may actually result in larger year classes. The reason for this is that, at large population sizes, densitydependent factors suppress further population growth. The mechanisms for this include limitation of spawning or rearing areas, limited food supplies, disease outbreaks at high densities, and other factors. Salmon provide a good example of spawning area limitation. A given stream has only so much spawning habitat, consisting of the stream bed with the correct gravel size, groundwater upwelling, and depth. If a salmon run is very large, eggs deposited in redds (a shallow depression in stream gravel where salmon lay eggs is called a “redd”) by females early in the run are later disturbed and washed downstream when subsequent spawning females come to the same site and excavate their redds. This process is called superimposition of redds. Alternatively, if salmon spawn in suboptimal sites, there is great risk that eggs will suffocate from low oxygen, be eaten by predators, or get washed downstream during storms.
212
Long-Term Ecological Change in the Northern Gulf of Alaska
Typically, groundfish stocks in the North Pacific are managed using a harvest rate that theoretically reduces spawning stock biomass to 40% of the unfished population size. Theoretical analysis and practical experience suggests that this is a conservative harvest rate for most fish populations. So, the expectation is that, while fishing reduces overall abundance of fish in the population, fish stock productivity (indexed as the number of recruits per spawner) is not adversely affected. Overfishing occurs when fish are harvested at too high a rate, so that spawning stock biomass is reduced to levels such that future recruitment is compromised. Conservative benchmarks of overfishing for Alaskan groundfish are generally set at harvest rates that theoretically reduce spawning stock biomass to 35% of the unfished biomass, and efforts are made to assure that fishing never approaches these levels. Stocks are defined as overfished, if they drop below a minimum threshold level, regardless of cause. So, both the rates of harvest as well as the level of fish population size relative to a threshold level are monitored to assure conservation of fishery resources. Many of the world’s fisheries have been overfished because of failure to monitor and constrain harvest rates to sustainable levels, like those currently used off Alaska. One recent analysis found that the median reduction in spawning stock size of exploited marine fish populations from known historic levels around the world was 83% (Hutchings and Reynolds, 2004). Much worldwide attention has been rightly focused on the maintenance of spawning stock biomass as a means to sustain fish stocks. Population dynamics theory focuses on biomass-based models of fish stock productivity, but considerations of fish biomass alone are not sufficient to avoid adverse effects of fishing. The effects of fishing on other features of the fish population, predator–prey interactions, and other attributes of the environment can be very important, as well. In the rest of this section, we explore other considerations, many of which are poorly studied. Aside from declines in fish biomass, the second most obvious direct effect of fishing is a reduction of age structure in exploited populations – i.e., the loss of older ages and largest sizes from the fished population. The reason that this happens is quite simple and analogous to what happened to human life expectancy since the Middle Ages. In the ninth to twelfth centuries, the life expectancy of humans at birth was 27–29 years. However, thanks to modern medicine that reduced mortality rate from diseases and accidents, the life expectancy of humans by the start of the twenty-first century was 77 years. In fished populations, mortality results from natural causes and fishing. As total mortality increases with increasing harvest rate, life expectancy declines. Thus, fishing truncates age structure – that is, the oldest and largest individuals in the population are lost. Whereas the reduction in the number of age classes of a fish population by fishing is an indisputable fact, its consequences are much less clear. Most fish population models are based on the premise that fish productivity is determined by the biomass of parental stocks, not their age structure. That is, if total spawning stock biomass levels are maintained at moderate to high levels, then stock productivity will be maintained.
Agents of Ecosystem Change
213
However, there are several reasons why this assumption may be faulty. First, larger, older female fish usually produce disproportionately more young than younger, smaller fish. However, depending on the relationship between biomass and fecundity (i.e., linear or exponential) equivalent harvests of mainly small or mainly large females may have different impacts on loss of population fecundity. Another factor is breeding experience as at least in some species, first-time spawners may not be as reproductively fit as repeat spawners (Wigley, 1999). More importantly, recent research on rockfishes found that larvae from older fish may have greater growth rates and higher survival rates than larvae from younger fish (Berkeley et al., 2004a). The effect of age was much greater than the effect of body size. The underlying mechanism appears to be related to the level of provisioning of larvae with lipids that increases the ability to withstand starvation. Second, in many fish stocks, older, larger fish spawn first. Protracted spawning times associated with a full age structure may bestow more opportunities for larvae to encounter favorable rearing conditions than synchronous spawning timing associated with populations with narrow age structures. Possibly, spawning of older fish is better timed with favorable conditions for survival of their young. Older, larger Pacific herring are the first to spawn along Alaska’s coastline and large Atlantic cod led formerly immense spawning aggregations off the coast of Newfoundland. The latter finding raised the possibility that migration routes are learned by younger fish from older fish (Rose, 1993), but more recent research has found that individual migration is more variable (Comeau et al., 2002). However, whether migration routes are learned for other species remains uncertain. In some instances, fisheries can create an imbalanced sex ratio that can cause reproductive failures not indexed by fish biomass alone. If males and females are spatially segregated during the fishing season, then disproportionate removal of one sex or the other can occur. Also, it is not uncommon for members of one sex to attain much larger sizes. For instance, female Pacific halibut grow much larger than males, and male red king crabs grow larger than females. In extreme instances like northern shrimp, animals begin their lives as males and end their lives as females. So, literally, all of the large shrimp are one sex (females). Naturally, fishing targets the largest animals. In the case of crabs, regulations only allow for the harvest of large males. Thus, sex ratios can be altered by fishing. For example, the experimental harvest of 29% of male white suckers from a lake resulted in a 38% increase in incidences of ovarian atresia (resorption of eggs) by female suckers (Trippel and Harvey, 1990). As an extreme example, trawl samples of red king crabs in Kaguyak Bay (Kodiak Island) in the spring of 1968 found that the female:male sex ratio was 72:1 and an anomalous 76.1% of adult females captured were not carrying eggs and went unmated (McMullen and Yoshihara, 1969). Beyond effects of skewed sex ratio on red king crab reproductive output, one might also wonder whether increased predation of females may result. Most red king crabs found in cod stomachs in the eastern Bering Sea are
214
Long-Term Ecological Change in the Northern Gulf of Alaska
softshelled adult females. Possibly, the pre- and post-mating embraces by male red king crabs may afford some protection from hungry cod predators. Spatial considerations are brought into traditional fisheries management when the “stock” is defined as a management unit. The definition of a stock may parallel the use of the term “population” when genetic methods are used to define a group of interbreeding individuals of the same species in the same area. However, for management purposes, other biological and fishery-related factors may go into the definition of stock. If stocks are defined over too broad a geographic area, then a harvest rate that may be appropriate if spread over a large region can be inappropriate and cause overfishing if concentrated in a small area occupied by a separate fish population. Intermingling of a small population of low productivity with another larger population of high productivity can also lead to overfishing when harvest rates are scaled to the more abundant population without recognizing the overlapping distribution with the smaller, less productive stock. Chronic overfishing can cause loss of genetic diversity, as groups of fish with different genetic makeup are extirpated. The genetics of each population may differ in ways favorable for their local existence. Berkeley et al. (2004b) postulate that failure to appreciate the value of old fish and fine-scale spatial dynamics of recruitment may have contributed to the overfished status of long-lived rockfishes along the U.S. west coast and other fishes elsewhere. Another issue regarding genetics is whether fishing can be a cause of evolution in fishes (Policansky, 1993). For life history traits to evolve as a result of fishing, heritable variation must exist and fishing must cause differential reproduction of genotypes. Decades ago, hatchery managers learned after the fact that by selecting gametes from the first salmon to return each year for their fish culture activities, they inadvertently shifted salmon run timing earlier by many weeks. As learned from animal husbandry, if certain traits are heritable, selective breeding, even if unintentional, is a powerful force. In recent years, studies of fishery-caused evolution have captured the interest of a number of researchers. In an experimental study of Atlantic silversides, disproportionate harvest of large fish caused selection for genotypes with slower rates of growth over just four generations of harvest (Conover and Munch, 2002). It is not unusual for fish populations that experience higher levels of mortality (such as predation) to adapt by maturing at younger ages and to devote more resources to reproduction (Reznick et al., 1996). Recent evidence suggests that, after a period of overfishing, Atlantic cod off Newfoundland and Labrador shifted to smaller sizes at a given age with fish reproducing at smaller sizes than seen historically (Hutchings and Reynolds, 2004). Given these and other recent studies, there is growing concern about fisheries-based evolution, because such outcomes could ultimately result in reductions of sustainable yields. Other aspects of fishing impacts are bycatch and discards. In general terms, bycatch is the capture of fish or shellfish other than the species for which the gear was deployed, whereas discards are those that are caught, but not kept. Discards may
Agents of Ecosystem Change
215
include fish of the target species too small to process profitably. Although strict catch quotas are set for most of Alaska’s fisheries, the utility of this management approach is only as good as the monitoring and accounting system. To keep total fishing mortality within limits defined by annual quotas for a given fish stock, a catch must be estimated as the sum of landings from the directed fishery, landed bycatch in non-targeted fisheries, and total mortality of discards from all fisheries. Additional sources of mortality, such as resulting from injuries of animals that are contacted by the gear but not captured or by “ghost fishing” by lost gear, are less well studied but mortalities from these sources are assumed to be low with one exception. Significant mortalities have been estimated for crabs caught in derelict pots in some instances, resulting in regulations requiring biodegradable exits, pot registration, and pot limits. Alaskan fisheries managers assure that catch and bycatch fall within quotas by use of a combination of reporting and monitoring systems. All fish processors are required to maintain detailed catch and production records; some are required to report daily, others weekly. Over 98% of the shoreside deliveries of Alaska groundfish are reported by a daily electronic reporting system. A state-run fish ticket system tallies the catches of all landings delivered within state waters. Some catcher vessels must maintain logbooks and other groundfish, scallop, and crab vessels are required to carry industry-funded onboard observers to monitor catch and bycatch. Finally, a vessel monitoring system is required on some vessels to enforce intricate time–area closures associated with protection measures for Steller sea lions. Fishing regulations are enforced by NOAA, the State of Alaska, and the U.S. Coast Guard. Although total fishing mortality complies with catch quotas, there is still a desire to reduce discards as a wasteful practice owing to its sheer volume in Alaska. In 1997, 258,000 mt of groundfish (15% of total groundfish catch) was discarded. By 2001, only 6.5% of the catch was discarded, partly due to implementation of a full retention requirement for pollock and cod fisheries. Similar retention requirements are being considered for flatfish fisheries in Alaska. Regardless of the requirements for any particular fishery, best estimates of total retained catch plus discard mortality are counted toward the total allowable catch limits used in fishery management. Conservation of habitats needed for fish reproduction, rearing, growth, and adult survival are essential for sustaining healthy populations of fish and invertebrates. For salmon, the State of Alaska recognized this many decades ago and prohibited the construction of hydroelectric dams on anadromous fish streams and implemented a Forest Practices Act requiring buffers along streams to prevent erosion, and an Anadromous Fish Act requiring approval of construction activities that may affect salmon spawning and rearing habitats. Failure to maintain pristine salmon habitats is one of the leading reasons for the collapse of Pacific salmon stocks in the Pacific Northwest. However, Alaska is not immune to political interests that may compromise fish habitats, as evidenced by the recent elimination of ADF&G’s Habitat Division.
216
Long-Term Ecological Change in the Northern Gulf of Alaska
The effects of fishing gears on seafloor habitats have been studied for many years in the North Atlantic, but have garnered research attention in Alaska only in the past decade. All fishing gears that contact the seafloor may affect benthic habitats. For example, longline gear and crab pots can uproot attached organisms, such as corals (Fig. 3.15). However, most attention has been focused on mobile bottom contact gear, such as groundfish and shrimp trawls and scallop dredges. Fishing effects on seafloor habitats has been the topic of many reviews, including a recent report by NRC (2002). These studies reached the following general conclusions: ● ● ● ● ●
●
●
trawling and dredging reduce habitat complexity; repeated bottom contact causes measurable changes in benthic animal communities; the magnitude of effects depends on the frequency of contact; fauna living in areas of low natural disturbance are most vulnerable to fishing impacts; gear type matters – e.g., among the mobile bottom-contact gear evaluated in the study, intertidal dredging tends to be most damaging and otter trawls the least damaging; effects vary with taxa – e.g., some fauna (e.g., corals, anemones, sea fans, crabs) are more vulnerable than others (e.g., clams, oligochaete worms, sea stars); and indirect effects on nutrient regeneration, community structure, and predator–prey relationships may occur.
Fishing impacts on centuries-old, cold-water corals have been of increasing concern in Alaska recently (NRC, 2002) owing to ongoing research by scientists at the Auke Bay Laboratory, National Marine Fisheries Service. However, lesser known are the very fine muddy habitats that have more subtle sedimentary (e.g., riffles) and
Figure 15: Deep sea coral landscape near the Aleutian Islands. (Photograph courtesy of NOAA, Alberto Lidner).
Agents of Ecosystem Change
217
biogenic (e.g., worm tubes) structures that divert flow of oxygen-laden bottom currents into sediments to support healthy aerobic communities of organisms and benthic food webs. Both of these quiescent environments are vulnerable to damage from mobile bottom contact gears. The principal management tools for seafloor habitat protection measures are: (1) reduced fishing effort (fewer trawl passes per year lessen the damage); (2) gear modifications to diminish the degree of bottom contact; and (3) closed areas to protect particularly vulnerable habitats from any contact with gear (NRC, 2002). Again, Alaskan fisheries managers and stakeholders deserve credit for the extensive habitat protection measures already in place through closed areas (e.g., trawl closures in state waters, federal waters in eastern Gulf of Alaska, Bristol Bay, etc.); reduced fishing capacity of many fleets (e.g., individual transferable quotas for halibut and sablefish, license limitations in scallop and groundfish fisheries); and gear modifications (e.g., bottom trawls are banned for pollock in the Bering Sea). In recent (February 2005) bold action, the NPFMC closed 95% of the Alentian Islands management are (~277100 nm2) to bottom trawling plus an additional 110 nm2 of especially dense coral and sponge habitat to all bottom contact gear, and another 10 areas (2086 nm2) of high relief, coral habitats were closed in the Gulf of Alaska to bottom trawls. However, there is very limited knowledge about the geographic distributions of bottom habitats and associated communities in the Gulf of Alaska. Extensive mapping of bottom habitats and fishing locations is necessary for informed decisions to be made on the need for additional protections. The NPFMC continues to evaluate new proposals to protect essential fish habitat and habitats of particular concern as new information becomes available. A worldwide issue is the notion of “fishing down” marine food webs associated with overfishing of higher trophic levels (Pauly et al., 1998, 2000). Analyses of global fisheries statistics revealed a shift in fisheries landings over the second half of the twentieth century from long-lived, high-trophic level demersal fishes to short-lived, low-trophic level invertebrates and pelagic fishes. This work spurred additional analyses that yielded conclusions consistent with this general pattern for worldwide predatory fishes, including pelagic sharks in the Gulf of Mexico (Myers and Worm, 2003; Baum and Myers, 2004). Although there has been considerable debate over the methods and conclusions of these studies, most fishery scientists agree that, in many areas of the U.S. and other areas of the world’s oceans, there has been substantial overfishing of many fish and invertebrates. Indeed, if top predators are removed, this could cause reorganizations of the marine ecosystem that persist over the long term. Marine food webs off Alaska do not show signs of having been fished down (Boldt et al., 2003). Despite an increase in the catch since the 1960s, the trophic level of fish and invertebrate catches in the Gulf of Alaska have remained stable. The same is true for the Aleutian Islands and eastern Bering Sea. Also, unlike other regions such as the Northwest Atlantic, Pauly’s “Fishery Is Balanced” index reveals that there is an
218
Long-Term Ecological Change in the Northern Gulf of Alaska
ecological balance to the catch patterns off Alaska. Moreover, an analysis of the size spectrum of groundfish sampled by annual trawl surveys in the eastern Bering Sea since 1979 indicates a slight increase in size spectrum (i.e., slightly more larger fish, slightly fewer smaller fish) over time rather than a trend toward smaller and smaller fish that could indicate a problem. These indications are not surprising as there have not been recent overfishing problems in the Alaskan region. Fishery management practices in Alaska hold up well when compared to other regions of the U.S. and world.
3.5.6. Effects of Hunting and Fishing: Some Conclusions Some conclusions can be drawn about the effects of hunting of marine mammals and harvesting of fish and invertebrates in the northern Gulf of Alaska. Sea otters were nearly exterminated from the North Pacific by the fur trade. Once protected, they slowly regained their former range in the GOA and were generally recovering in number until recently, when they collapsed across a broad region westward from the central GOA (see Section 4.9). Likewise, certain of the great whales are also recovering, particularly humpbacks which are comparatively abundant throughout coastal areas around the rim of the GOA. Fin whales are apparently increasing as well, as they are now common around Kodiak Island in summer. Sperm whales and blue whales, however, are still rare everywhere in the GOA and elsewhere in the northern North Pacific. Regarding other marine mammals, such as seals and sea lions, did historical killings affect changes in their populations? Subsistence levels are low and are not considered as a major contributing factor. However, historical incidental fishery catches and associated shootings by fishermen, government hunters, and others are difficult to evaluate because of very limited documentation. Even for fisheries with observers, extrapolated estimates of mortality are based on a set of assumptions, not the least of which is that the behavior of fishermen is not influenced by the presence of an observer. So, the actual historical number of kills is unknown but likely higher than official published estimates. Considerations of available information for Steller sea lions led the National Research Council to conclude that the mystery of the sea lion decline could, in part, be unreported and illegal takes (NRC, 2003). The effect of bounty programs and the fur trade on harbor seal abundance, as discussed in a following section, varied geographically, but was extreme in British Columbia and probably southeastern Alaska, where seals were greatly depleted in the 1950s and 1960s. Likewise, the number of harbor seals in the vicinity of the Copper River Delta must surely have declined because of the heavy toll taken by the control program. However, as we shall see in Section 4.9, effects of harvesting were apparently not so great in the northern GOA, although measurable nonetheless. In many ways, current fisheries management practices in Alaska are the shining light for fisheries management worldwide. In many fisheries, fishing effort is limited through license limitation programs or individual fishing quotas, quotas are restricted based on
Agents of Ecosystem Change
219
an allowable fraction of harvestable biomass, fisheries are closed when biomass falls below thresholds, bycatch is accounted for and managed within acceptable limits, and habitat protection measures are in place in many areas of the state. However, despite these current management practices, there were periods in the past in which overfishing occurred in the northern Gulf of Alaska. Some were caused by foreign fleets prior to extended jurisdiction (e.g., Pacific ocean perch in the 1960s and 1970s), but others were caused by domestic fishermen before the maturation of the discipline of fishery science and before effective fishery management institutions were created (e.g., halibut in the early 1900s, salmon during the 1920s–1950s, herring during the first half of the twentieth century). For currently depleted stocks of shellfish (e.g., crabs and shrimps), there is some evidence for periods of overharvest in some areas, but evidence for chronic overfishing is lacking. Nonetheless, there are real concerns for more subtle effects of fishing and recent research suggests that biomass-based fishery management strategies alone may be insufficient to maintain sustainable harvests of marine resources. Concerns exist regarding the application of current biomass-based strategies to long-lived species, need for more explicit spatial management, maintenance of age and size structure of exploited populations, genetic selection by fisheries, role of habitat in fish production and need for further habitat protection, and the ability of fishing to restructure marine food webs in undesirable ways. Much of this research is new, and generalizations are impossible. However, these ideas point toward new frontiers for fisheries research and additional considerations by fisheries managers. Alaska has much to be proud of, and undoubtedly, fisheries scientists and managers will continue to lead the way in considering these cutting edge challenges in the twenty-first century.
3.6. Disease Paul Reno 3.6.1. Introduction Diseases can have a severe impact on the populations of animals and plants. Most of the evidence of population effects are however, for human diseases and the human population. The most thoroughly studied disease from a population perspective, for example, is measles, due primarily to its high level of contagiousness, virtual lack of mortality, and the acquisition of lifelong immunity in survivors (Anderson and May, 1979). Conversely, there is a distinct lack of studies on the effects of diseases on wild populations, especially in marine ecosystems. This is not because marine animals do not contract diseases, but more because the effects are inapparent to humans. Animals dying from disease are rapidly scavenged or while ill, eaten by predators. It is only when large mortalities occur in coastal areas that humans think more about the
220
Long-Term Ecological Change in the Northern Gulf of Alaska
impacts of infectious diseases on marine animal populations. In order to produce significant adverse effects on populations, a pathogen should have three characteristics. First, it must induce in the host a disease that results in direct or indirect mortality, or detrimentally affects fitness or fecundity. Second, it must affect a significant proportion of a population to reduce population size. Third, it should occur over a wide enough geographic range to engender population reduction over an extended range, rather than localized. This chapter provides an overview of the epidemiology and potential effects of infectious disease on marine animal populations. The potential detrimental effects of anthropogenic contaminants on disease susceptibility are also discussed.
3.6.2. Agents of Infectious Disease and their Lifestyles Infectious agents run the gamut in size and “lifestyle” from “simple” viruses to complex crustacean parasites. Epidemiologists have broadly categorized infectious agents into two main groups based upon their size and method of infection. Microparasites are small relative to their hosts, reproduce at high concentrations within or upon the bodies of their hosts, and generally confer long-lasting immunity in the host after recovery from infection. This category includes the viruses, bacteria, and protoctistan parasites (those formerly known as protozoans). Macroparasites are relatively large compared to their hosts, generally cannot multiply to high numbers in or on the host, and do not induce a long-lasting immunity in the host. Consequentially, macroparasites often only have adverse effects in large numbers. Conversely, even small numbers of microparasites can cause severe disease or death in infected individuals because of their ability to produce millions of progeny, but subsequent exposure will be less likely to cause disease again because of the protective effect of the induced immune response. Macroparasites often require several successive hosts to complete their infectious cycle, causing disease in one host but leaving another unaffected. Macroparasites with multiple hosts are seemingly less liable to have long-term prospect of surviving in a population because they must rely on more than one host being present in sufficient numbers to sustain infection, but from an evolutionary perspective, they have adapted well (sometimes too well!) to survival. As humans, we tend to think of infectious disease in terms of individuals rather than the potential population effects. The fear of contracting a disease as an individual supersedes the perception of population reduction. Even the scourge of AIDS in the US is not so much perceived as affecting our society, but rather how can I avoid getting it? The question arises as to how epidemics, or more appropriately referred to as epizootics in non-human animals, are initiated and subsequently diminish to nothing until another epizootic occurs at some time later. Expressed another way: why isn’t disease a constant in a population? The S-I-R model of disease that elucidates the
Agents of Ecosystem Change
221
characteristics and constraints on the dynamics of infection and disease in a population answers this question (Anderson and May, 1979). The model has 3 mutually exclusive classes of individuals in a population, relative to their infection status with a particular pathogen: the susceptible class (S), consisting of those animals that can become infected if exposed to a pathogen in an appropriate manner; the population of animals that are actively infected with the pathogen (I); individuals in a population that are removed from the infectious process either due to death or the development of immunity to infection (R). There is a dynamic interaction among the three classes of individuals that is exhibited in the waxing and waning of infection in an epizootic (Fig. 3.16). First, there is an accelerating increase in the proportion of the population that is becoming infected (the incidence), peaking at a time characteristic of a particular infectious agent, and then declining to the point of complete elimination from the population or a residual infection rate at significantly lower levels than during
Figure 3.16: The dynamics of infection in a population for a single episode of disease. Three mutually exclusive classes of animals are involved: S = those animals that are susceptible to infection; I = those animals that are infected; R = those animals that are removed from the infection process by either death or immunity. The graph depicts the dynamics of infection in an hypothetical population of 2000 susceptible animals at t = 0, starting with 1 infectious individual (I). As the infectious class size increases, the S class decreases proportionally. As more animals become infected there is an increase in the R class by virtue of individuals recovering and becoming immune to further infection and from disease-specific mortality. A point is reached, called the critical threshold density, where the density of individuals in the S class is insufficient to sustain transfer of the pathogen to other animals, the infection rate drops and the epizootic ceases.
222
Long-Term Ecological Change in the Northern Gulf of Alaska
an epizootic. The primary cause for the accelerated incidence is the increase of the number of infected individuals that serve as foci of infection; conversely, the decline in infection is due to a reduction of the susceptible population to a density less than the critical density needed to sustain infection.). There is a “cycle” of infection and disease in a population (Fig. 3.17). There are several notable characteristics of this model of the dynamics of infection. First, the model is based on the infection rate in populations, rather than either the disease or mortality rate, since the transfer of the pathogen is the elemental factor in the process at the population level. In fact, interestingly enough, if the disease-specific mortality is very high and rapid, the infection in the population will die out rapidly because the critical (threshold) density will be reached quickly and fewer individuals in the population will become infected than if the mortality rate is lower. On the other hand, if the infection results in low or no disease-specific mortality, the infection can spread and be maintained in a population for long periods, and even indefinitely. From an evolutionary standpoint, this is advantageous to the pathogen.
Figure 3.17: The cycle of disease. In the long term, the SIR model must take into account an influx of individuals as well as rates of change among the classes of individuals. The diagram depicts those relationships. Flows of animals into the susceptible (S) class of a spatially stable population are determined by the birth rate (δ). Outflows from the S class include those that die naturally at rate (µ) and those that are Infected at rate β. The Infected class has outflows to the Removed class at rate ρ, and to natural mortality (µ) and disease-related mortality. Reductions to class R are engendered by natural mortality and the loss of immunity (at rate υ) that return individuals to the Susceptible class.
Agents of Ecosystem Change
223
Another important factor is the time between when an individual is exposed to the pathogen and when it is capable of dispersing it to other individuals. That is, when an animal is exposed but not infectious (the E phase, not shown in Fig. 3.16). The duration of this phase can significantly alter the dynamics of the infection process: those pathogens with an extended E phase often produce chronic rather than acute infections and tend to remain in populations for long periods of time. Experiments in our laboratories and in others with viral and bacterial pathogens of salmonids have indicated that hosts begin to release the pathogen prior to developing external or behavioral signs of disease (Ogüt and Reno, 2004a,b). This is problematic because the exposed animals exhibit no signs of disease and a large portion of the population may be infected and become infectious before it is noticed. For example, Renibacterium salmoninarum, an unusual bacterial pathogen of anadromous salmon and trout, is a chronic infection that has become rampant in these salmonids everywhere because the pathogen is not manifest as disease until a large proportion of the population at risk has already become infected (Bullock and Herman, 2003). The removed class (R) comprises animals that have died and those that have gained immunity. In the mathematics of the model, animals that die are equivalent to those that survive and develop immunity, but obviously, from a population perspective, it is preferable to have a large proportion of the R class in the latter rather than the former group! Aspects of the relative proportion of the population in this class are dependent on the characteristics of the pathogen and how it affects the host. Acquisition of lifetime immunity is restricted to certain highly contagious but non-lethal viral diseases of humans. In other mammals and non-mammals, immunity is temporary and those survivors of an epizootic that are spared re-infection for some time period will return to the pool of susceptibles after the protective effects wear off. These models have been developed for human pathogens and have been most frequently utilized to portray the dynamics of an infection by human childhood viral diseases, due to the relative simplicity of the process of infection by these pathogens. However, as might be expected, there are significant differences between the infection dynamics of these human diseases and those that occur in marine animals. One of the major differences is that in human populations, there is a constant influx of births into the population during a year and, thus, a constant influx of new members of the S class. With nearly all marine animals, however, the reproductive cycle is seasonal and episodic. Spawning and calving seasons generate an influx of susceptible animals into a population within a short period of time that is characteristic of each species. Consequently, a new “bolus” of susceptible animals of a particular species is only available at certain times of the year. This influx of new susceptible animals during a short time period is advantageous to pathogens for several reasons. First, the progeny are found at the highest density at birth or spawning. Since most pathogens spread most easily and rapidly under conditions of high host density, this would enhance a pathogen’s ability to infect new hosts at this time. Second, often the
224
Long-Term Ecological Change in the Northern Gulf of Alaska
youngest animals in a population are most susceptible to infection, disease, and disease-related mortality. For example, aquatic birnaviruses, which infect a broad range of aquatic species ranging from flatfish to oysters and other shellfish, cause the highest mortality in animals less than a year of age (Reno, 1999), perhaps because they do not yet have their full repertoire of innate or acquired immune defenses. Additionally, like human urban populations, those marine animals that tend to aggregate in schools for protection from predators, such as krill, shrimp, clupeid fishes, and seals at haul out areas, increase their likelihood for pathogen transfer due to proximity of animals. Another difference is that water transports pathogens more effectively than direct contact or air. The bacterial concentration in seawater is in the range of 100,000 organisms mL−1, and although only a small proportion of these organisms have the potential to cause disease, there is certainly a high likelihood of marine animals coming in contact with those that may be present. When infected animals become infectious by releasing pathogens into the water through the portal appropriate to the particular pathogen, they can do so in prodigious quantities. For example, laboratory experiments have demonstrated that Aeromonas salmonicida, the agent of furunculosis in salmonids and other fishes can expel up to 104 bacteria mL−1 of water and that Pacific herring (Clupea pallasi) held in pens for spawn taking that are infected with the virus of hemorrhagic septicemia can produce 103 infectious doses of virus mL−1 of seawater, 2 m from the outside of the net pen (Kocan et al., 1998; Ogüt and Reno, 2004b). Considering the mass of seawater flowing through and around these net pens, this is an astonishing production of pathogens by their hosts.
3.6.3. Resistance to Infection and Disease All marine animals have innate resistance to the infectious agents with which they may come in contact. The efficiency and complexity of the protection is strongly correlated with the phylum to which the animals belong. For example, invertebrates possess a primarily cellular-based innate response that is simple and broad-based in its specificity; all vertebrates possess the hallmark cellular responses of the invertebrates as well as an induced cellular and humoral (serum-based) resistance that is quite specific in its effect and therefore more efficient than a nonspecific cellular based response. Teleosts have a simple specific response, while the marine mammals have an armamentarium similar to humans in its complexity and efficacy. The response of teleosts is relatively short-lived, whereas that of the mammals is much longer. Thus, reinfection of fish and invertebrates upon re-exposure would be likely but mammals would be less likely, as individuals would be resistant for much longer. The resistance of marine animals, especially poikilotherms, is temperature-dependent and generally much less rapid in its generation in animals with low body temperatures than in their avian and mammalian counterparts. Consequently, this is another
Agents of Ecosystem Change
225
detriment to resistance, given that most microbial agents replicate rapidly and a delay in the generation of the response can tip the balance in favor of the pathogen. Two other factors are germane to the focus of this chapter: stress and anthropogenic pollutants. Stress is always waiting in the wings for marine animals. The constant threat of predation induces a chronic stress pattern in all animals, and even during short periods of active escapement from predators, an acute stress response is induced. During these periods of stress, the pituitary and ancillary tissues control and secrete corticosteroids that, while of short-term benefit in the fight or flight response, can severely hamper the longer term specific and innate immune response. This class of compounds represses the immune response of all groups of marine animals. The inducers of stress are, unfortunately, not solely natural. A wide variety of anthropogenic materials elevate corticosteroid response and lead to a subsequent reduction in the effectiveness of the immune response, including heavy metals, polycyclic aromatic hydrocarbons, and petroleum-based compounds (Anderson et al., 1983; Dunier and Siwicki, 1993; Pipe and Coles, 1995; Le Moullac and Haffner, 2000; Jacobson et al., 2003). Immunosuppressors paint their victims with a broad brush. While many studies have been conducted on this topic, the variety of compounds that induce immunosuppression coupled with the broad range of hosts that are affected and the inherent complexity of defenses within each class of animals makes the pattern of information appear sparse indeed. One defense mechanism that seems to be generally compromised is phagocytsis, the ingestion and destruction of foreign particles by specialized cells of hemic origin – virtually always pathogens in natural situations – that forms the anchor of defense in all marine animals from the coelenterates to mammals. In vertebrates, the cells responsible for the specific immune response are often adversely affected by polycyclic aromatic hydrocarbons and other similar compounds. For example, T-lymphocytes (responsible for the cell-mediated immune response) and B-lymphocytes (responsible for the humoral or antibody response) normally proliferate in response to plant mitogens, which are used as in vitro surrogates for their reponses in vivo. The response of the T-lymphocyte mitogens is compromised by many pollutants, whereas in fewer cases, the B-lymphocytes lack responsiveness. In spite of the retention of a large number of serviceable antibody-producing cells, marine animals are less resistant to diseases of fungal, parasitic and bacterial origin after exposure to pollutants (Lowe et al., 1971; Anderson et al., 1983; Yevich and Barszcs, 1983). Thus, in marine animals that already have a rather tenuous resistance to infectious agents, it may not take much of a reduction in efficiency of the immune response to tip the balance in favor of the pathogen, with disease in the population elevated beyond the normal level as the consequence. In essence, referring to the S-I-R- model discussed earlier, pollutants and stress can reduce the proportion of animals residing in the R class and add them to the S class, where they can enable and exacerbate the disease process in the population.
226
Long-Term Ecological Change in the Northern Gulf of Alaska
3.6.4. The Diversity of Infectious Agents Causing Diseases in Marine Animals One could surmise that marine species with their vast range; wide variety of niches beneath the seas covering 75% of our planet; vertical ranges far exceeding terrestrial habitats; proclivity towards high fecundity; penchant for traveling in dense arrays almost designed to favor pathogens, would suffer grievously from the slings and arrows of disease. Apparently this is not the case, at least as far as we have little evidence to support this. We know of far fewer pathogens of wild marine animals than of humans or domestic animals, or even freshwater animals. There is significantly more information about the pathogens affecting cultured marine animals such as shellfish, crustaceans and teleosts, especially salmon, than of the other marine animals. It is interesting to note, for example, that there are virtually no references to epizootics of infectious disease in marine birds, aside from intoxications by botulism and chemical spills and reports of ectoparasites (lice and ticks). This disparity in knowledge results from intense monitoring of cultured marine animals coupled with the inherent economic interest in them. There is some evidence to indicate that the same pathogens that affect marine animals in culture also affect wild marine animals. In fact, these pathogens were present in wild marine animals and were transported to culture, as unwanted guests, along with those animals that were to be reared in culture. Examples of the pathogens afflicting marine animals are presented in Table 3.1. The pathogens of marine animals differ little from their terrestrial counterparts, perhaps on the whole being more disparate in their temperature range and in some, but not all, instances, adapted to a strictly saline life. A good example of the latter are the bacteria belonging to the genus Vibrio. These organisms are almost always detected in seawater samples and comprise a large portion of the bacterial flora in the marine and brackish environment. Several members of this genus can cause acute and chronic diseases in marine animals including rotifers, shellfish, crustaceans, teleosts, reptiles and marine mammals, as well as humans (cholera caused by Vibrio cholera; fatal wound infections by V. vulnificus; gastroenteritis by V. parahaemolyticus). Members of this genus mostly require salt for their growth and can survive in the temperature range of 5–38° C. Thus, they are capable of causing diseases in marine homeotherms as well as poikilotherms. Most of the organisms that are responsible for disease in marine animals, however, are not indigenous flora, but rather are organisms that are generally closely associated with their hosts: these are the pathogens that cause disease. Large metazoan parasites are the group of pathogens most often found associated with marine hosts by scientists (and commercial and recreational “anglers”). This is likely, in part, due to the nature of these beasts: (1) they are most often non-lethal to their hosts; (2) they often attach to their hosts externally and so, are
Agents of Ecosystem Change
227
readily observable; (3) they are generally large enough to be seen with the naked eye and are clearly noted as “abnormal”. Thus, external nematode and trematode parasites in the muscle tissue are often observed by recreational and commercial fishers and reported more often than microparasites such as protozoans, bacteria and viruses as detrimental agents of marine animals. As pointed out earlier, it is difficult to gauge the prevalence and impact of pathogens in marine animals because most affected animals are subject to increased predation because of their infirmities and are not seen by humans unless they are evident inshore, nearshore, or when caught. It is beyond the scope of this article to survey the pathogens of marine animals, but some insight into the breadth of the pathogens can be gained by seeking out the works of Sindermann (Diseases of marine fishes and shellfish, 1990) and the series of volumes by Kinne, 1980–1990 (who approaches the topic from the perspective of the host phylogeny rather than from the perspective of classes of pathogen). Space considerations preclude a detailed, or even superficial examination of the listed pathogens. Rather, several pathogens that will be discussed have been associated with marked population effects on their host in a natural environment. This will exclude those pathogens that are associated predominantly with cultured marine animals, although these pathogens are capable of literally decimating cultured animals in their captive state. It should be noted, however, that the pathogens affecting captive animals are not inherently different from those that cause disease in wild animals, since ultimately, they were originally present in wild animals prior to their cultivation by humans. Aquaculture has tipped the balance in favor of pathogens by rearing animals at the highest densities possible to maximize economic return. In doing so, aquaculturists are almost assuring that if a pathogen arrives at a culture facility, the stressed, and therefore immunocompromised animals, will be highly labile to infection and subsequent disease. By contrast, wild populations of marine animals tend to be present at significantly lower densities than their captive counterparts. Thus, they are less likely to become diseased. One possible exception to the above is that of schooling animals such as the clupeids (herring family), and disease among these populations could at times be as severe as occurs in culture situations. One of these will be among the diseases described below. It is interesting to note that not all marine animals have been reported with severe infectious diseases. For example, there are only a few diseases described for corals, although those that have been described have been very severe. Also, little has been described for the cartilaginous fishes and interestingly, from marine birds. It is difficult to ascribe the lack of reports to some hyper-resistance on the part of the animals belonging to these classes; rather it is more likely due to our not knowing about epizootics when they occur. Bearing in mind that the diseases to be discussed have all been found to produce significant population declines in local and temporally constrained periods, I will select several of the more interesting and important pathogens for discussion. The diseases
Bacteria
Corals
None reported
Molluscs (bivalvula)
Oyster velar virus disease
Molluscs (cephalopods) Crustaceans
Cartilaginous fishes Teleosts
Reptiles
Microparasites
Macroparasites
White spot disease and Aspergillosis brown spot disease Abyssomyces Vibriosis, hinge disease Norcardiosis, QPX disease
Flagellates (Trypanohis), ciliates MSX, Perkinsis, bonamiosis
Copepods
Iridovirus of octopus
Vibriosis
Cladosporium
Flagellates, coccidiosis
Taura syndrome, white spot disease, balculovirus of shrimp None reported
Shell disease, V. parahemolyticus shrimp rickettsia
Black mat Bitter crab disease of disease of tanner crabs, crabs, needle Lagenidium sp. disease of Dungeness lobster, shrimp crabs, None reported None reported
Viral hemorrhagic septicemia, lympocystis Grey patch disease
Vibriosis, streptococcal disease, flavobacteriosis Vibriosis, aeromonades, coccidiosis
None reported
Fungi
Ichthyophonus hoferi
Parvicapsula, Kudoa
Cladosporium
Heximita, coccidiosis
Digenetic trematodes, cestodes Gryodactylids, digenetic trematodes Carcinonemertes, carcinophila worm
Monogenetic trematodes, copepods Sea lice, countless others
Nematodes, digenetic trematodes
Long-Term Ecological Change in the Northern Gulf of Alaska
Virus
228
Table 3.1: Some reported pathogens and parasites of marine organisms.
Birds
Pinnipeds
Whales
Avian influenza, avian cholera Phocine distemper seal pox, seal lion calicivirus Dolphin distemper, pox, hepadnavirus
Vibrosis
Aspergillosis
Coccidiosis, trypanosomiasis
Renal trematodes
Dermatosis, hemorrhagic enteritis, bacterial pneumonia Vibriosis
Blastomyces, dermomycoses
None reported in wild animals
Asperigillosis, candidaisis
None reported
Digenetics trematodes gastric anasakids None reported
Agents of Ecosystem Change 229
230
Long-Term Ecological Change in the Northern Gulf of Alaska
are summarized in Table 3.2; where possible, I have provided an URL relevant to diseases of particular classes of marine animals.
Diseases of Coral (http://www.coral.noaa.gov/coral_disease/) During the 1970s, it was found that large areas of the amazing coral reefs in Florida and the Caribbean Sea were undergoing prominent degenerative changes (Green and Bruckner, 2000). Upon closer examination, it was determined that the cause of the disease was infectious rather than toxic, although at present, the relationship between the bacterial agents present and disease have not yet been fully proven to be causal (the fulfillment of Koch’s postulates necessary to prove an agent responsible for disease have not yet been attempted, Richardson, 1998). The disease was named coral plague and could move across the surface of the coral at rates of up to several centimeters per day. The disease agent is a new genus and species, Auriantimonas coralicida. The two major diseases were termed white band disease and brown or black band disease because of their primary visual affect on the coral. The agent of the diseases are bacteria, the former is likely to be caused by a vibrio (Vibrio carchariae which intriguingly was originally isolated from sharks) while the latter is known to be caused by cyanobacteria – formerly known as blue-green algae, but now classified as members of the archeobacterial group – and members of the genus Beggiotoa. The latter kills the coral tissue by production of high levels of sulphides, since their nutritional physiology is based on sulphate reduction, which is toxic to the coral. These diseases progress over the surface of the coral at rates of up to several centimeters per day, leaving denuded coral skeletons in their wake. As an example of the havoc wreaked by these diseases, white band disease has reduced the cover of the coral Acropora palmata from about 85% of the reefs in the US Virgin Islands to 5% since the 1980s (Gladfelter, 1991). Certain characteristics of this system do not portend well for the future resurrection of diseased corals. First, the diseases appear to be exacerbated, if not initiated by the stress placed on the reefs by anthropogenic pollutants, and there is little likelihood of amelioration of these hazards to the coral in the near future. Second, the corals are armed with a relatively simple defense capacity which is not efficient in protecting against invasion by infectious agents. This dooms the coral polyps in proximity to infected areas. In addition, the coral’s immobility contrasts with other marine animals that have the ability to move from an area affected by infectious organisms. Thus, it appears that the almost inexorable progress of disease destroying the coral is likely to continue unabated and offers a bleak outlook for regeneration of this essential component of the marine ecosystem in certain areas of the world. Diseases of Molluscs (http://www.pac.dfo-mpo.gc.ca/sci/shelldis/title_e.htm) Molluscan diseases have been more extensively studied than their coelenterate forebears since they have been considered of greater immediate economic value.
Table 3.2 Disease having a significant impact on populations of marine animals. Disease
Pathogen
Host(s)
Range
Impact
Coral plague, brown spot MSX
Auriantimonas Haplosporidium nelsoni Hematodiunium
Caribbean Eastern US Atlantic Coast Alaska
High Severe
Bitter crab disease VHS
Hemorrhagic septicemia virus Papilloma virus, herpesvirus Phocine morFbillivirus
Several species of coral Easter oyster (Crassostrea virginica) Tanner crab (Chionoecetes sp.) Clupeids, other marine teleosts Marine turtles Harbor seals (Phoca vitulina)
Moderate to severe Low to moderate Severe
Agents of Ecosystem Change
Grey patch disease, LETD Phocine distemper
Alaska, BC, Washington Caribbean, Hawaii, Florida Northeastern Europe, UK
Moderate
231
232
Long-Term Ecological Change in the Northern Gulf of Alaska
Beginning in the late 1950s, the Eastern oyster, Crassostrea virginica has been severely afflicted by a parasite that has literally decimated the population of wild and cultivated oysters. The disease is called MSX, which stands for multinucleated spore X (unknown), but the agent has been identified since then and is a sporozoan called Haplosporidium nelsoni (Haskin et al., 1966; Haskin and Andrews, 1988). It was originally noted in the lower reaches of Delaware Bay, but by the mid-1980s it has been detected in oysters from Maine to Florida. The disease has been especially catastrophic to oyster populations in the Chesapeake Bay, an area renowned for its oysters and other seafoods. This disease alone has cost the oyster industry tens, if not hundreds of millions of dollars, over the last 50 years. The wild oyster harvest in the 1960s of about 35 million pounds per year in Chesapeake Bay declined to merely 350,000 pounds in 2004, the lowest ever; this represents a 95% decline in the oyster harvest. While overfishing and other factors have contributed to the reduction, disease plays a major role in this decline. The mortality caused by the disease is frequently greater than 50% in a year, and is elevated during the summer months, and generally a smaller epizootic in the fall–winter. One primary factor noted in the distribution of the disease is that it does not cause mortality when oysters are held in salinities of 10 parts per thousand (ppt), and is markedly more lethal at salinities of 15 ppt (Ford et al., 1985). The parasite kills the oyster within a few months after the initial exposure, by replicating to profusion in virtually all the tissues. One of the most interesting, and exasperating characteristics of MSX is that although the parasite has been identified and cataloged for its taxonomic relationships with other organisms, its mechanism of transmission remains cryptic. Laboratory transmission of the parasite has been uniformly unsuccessful, and the epizootiologic characteristics of the disease in nature imply the presence of an as yet unknown intermediary host. Thus, this disease along with another parasitic disease termed “Dermo” in reference to the causative agent Dermocystidium, has caused a literal decimation of oyster populations in the formerly productive Chesapeake and Delaware Bay oyster beds, and there is no indication that the diseases will abate, or will soon be ameliorated by management and manipulation (Ford and Haskin, 1988). Like the coral diseases discussed above, the consequences of MSX infection appear to be spreading inexorably and at this time, managers and researchers are contemplating problematic solutions to the problem, such as introducing native oysters from Japan, in spite of the fact that this is purported to be the original source of MSX (Anderson and Hedgecock, 2004). Like the coral, the immobile oysters are unable to evade the parasites that affect them, and because researchers have not yet been able to identify a possible intermediate host of the pathogen, it is unlikely that the disease can be managed very well. This is an instance in which disease has had a severe, long-term impact on its host population, with no apparent relief in the foreseeable future.
Agents of Ecosystem Change
233
Diseases of Crustaceans (http://www.pac.dfo-mpo.gc.ca/sci/shelldis/pages/ hematcb_e.htm) It is said that everything and everybody loves crustaceans, from human consumers to predatory consumers of every stripe. This includes pathogens, of course. There are numerous diseases of crustaceans, caused by a wide variety of pathogens. Most of those investigated by researchers affect cultivated marine species, especially shrimp and lobsters. There are, however, several diseases of wild crustaceans that have had a severe impact on populations, notably the bitter crab disease of Tanner crabs (Chionoecetes bairdi) of the Pacific coast of the USA. Bitter crab disease, also called bitter crab syndrome (BCS), was first noted and the etiologic agent identified by Ted Meyers and his colleagues in Alaska in 1987 (Meyers et al., 1987). While sometimes thought of as a bacterial or viral disease, it is caused by a dinoflagellate (an algae-like organism) belonging to the genus Hematodinium, although it is distinct from other species that infect species other than Tanner and snow crabs (Chionoecetes spp.). The parasite causes the hemolymph of the crab to turn milky white, rather than its normal clear appearance, and the hemolymph is replete with excessive numbers of the feeding stage of the parasite. As its name implies, the muscle of cooked crabs infected with the bitter crab disease parasite has an unpleasant astringent taste. The musculature becomes depleted, likely leading to enhanced predation upon infected crabs that cannot readily evade predators. The distribution of the parasite is from Alaska to the Bering Sea and south to British Columbia; it has also been recently detected in Atlantic Canada. The prevalence of disease can be up to 80% in some areas of southeast Alaska, but it is generally much lower than that elsewhere. Tanner crabs sampled in Alitak Bay (Kodiak Island) between 1991 and 2000 averaged 10% infection (Urban and Byersdorfer, 2002). Prevalence is tenfold lower in the snow crab in the Atlantic. Morado et al. (2000) determined that BCS prevalence was significantly higher in juvenile snow and Tanner crabs than in legal size crabs, indicating that affected young crabs may succumb to disease prior to reaching legal size. Thus, recruitment into the harvestable cohort may be diminished significantly in areas of high prevalence. The disease is invariably fatal to the host, although it may take up to a year for the disease to take its ultimate toll. Under these circumstances, this disease has taken a considerable toll on the commercial crab population in Alaska. Thus, BCS may be a significant factor in declines of some Tanner and snow crab stocks in Alaska.
Diseases of Teleosts (http://wfrc.usgs.gov/capfishhealth.htm) Most of the known diseases of marine animals are found in teleost fishes, due to their long history of cultivation. More than 10,000 parasites, hundreds of bacteria, and nearly 100 viruses have been reported from fishes. However, far fewer have been detected in wild fishes and, interestingly, few have been reported from the cartilaginous fishes, the
234
Long-Term Ecological Change in the Northern Gulf of Alaska
sharks and rays. Of considerable import to Alaska and the Pacific Northwest, however, are the diseases viral hemorrhagic septicemia (VHS) and phoma, a fungus. The disease VHS is caused by a virus related to rabies (Meyers and Winton, 1995). First isolated in the 1960s in Danish hatchery trout (Jensen, 1965), it became an impediment to European trout culture, but remained confined to the UK and the European continent (Wolf, 1988). In 1976, Jensen isolated VHS and an unrelated virus from skin ulcers in a wild Atlantic cod (Gadus morhua), this being the first isolation of the virus from a strictly marine fish. Since the virus belonged to the same serotype as the trout virus, and it was from an unusual species, some fish pathologists thought that the isolation of the virus was the result of a laboratory contamination. No further isolations were made in marine species until 1989, when apparently healthy coho and chinook salmon (Onchorhynchus kisutch and O. tschawytscha) returning to hatcheries on the Olympic Peninsula in Washington were found to harbor VHS; the isolated virus was about 90% identical to the European virus. This occurred despite rigorous importation requirements for fish from Europe to exclude the infection of North American trout with VHS. The target hatcheries were depopulated and disinfected. Laboratory studies indicated that the isolate from Pacific salmon was not virulent in any of the salmonid species tested, including rainbow trout. The overarching question was how did this purportedly indigenous European virus of fish arrive in the Eastern rim of the Pacific? The answer was inapparent at the time it was first reported. Ted Meyers in Alaska detected VHS in the ulcer of a single Pacific cod (Gadus macrocephalus) brought to the Alaska Department of Fish and Wildlife Fish Pathology Lab in 1989 (Meyers and Winton, 1995). This finding appeared to be similar to that of Jensen in Europe more than a decade earlier. However, further examination of Pacific Herring in Alaska in 1990 revealed that the virus was also present in this essential forage fish (Fig. 3.18) and laboratory studies indicated that the virus was virulent in the host species and in others. Further work by Kocan et al. (1998) indicated that when herring held in net pens with seaweed that served as spawning mats, a commercial usage practiced in sheltered Alaskan waters for a period of two weeks were tested for the prevalence of VHS before and after holding, that the infection level rose two-fold. Even more disconcerting was that, when seawater two meters away from the pens was tested for virus, up to 1000 infectious doses per mL were detected. Given the volumes of seawater that pass through and around the nets, this is an astonishing amount of virus. Since these initial findings, several other gadid and clupeid species have been found to harbor the virus and its dispersal has been documented from Prince William Sound Alaska to British Columbia, Puget Sound, and Coos Bay Oregon, as well as from a variety of other fishes in Europe (Kocan et al., 2001; Meier et al., 1994). The ability of the virus to kill herring and other fish, coupled with its distribution throughout a wide geographic area in the Pacific Northwest implies that it has dispersed very rapidly or has been present for a long time without being noted. The latter is a more likely case, since monitoring of these marine forage species for the
Agents of Ecosystem Change
235
Figure 3.18: A 7-year-old, female, Pacific herring sampled from Rocky Bay in Prince William Sound on April 27, 1995. Lesions are typical of those found during the outbreaks of viral hemorrhagic septicemia virus (VHSV).
presence of any viruses has not been extensive. It was only when this pathogen of trout was detected in Pacific salmon, most likely as a consequence of ingesting or contacting infected herring, that efforts were directed towards closer investigation of wild nonsalmonids in the area. Further, the presence of lesions on herring following a precipitious decline in the herring population of Prince William Sound in 1993–1994 motivated further testing of herring, revealing the presence of VHS. Its effects on wild gadids (e.g., cod) and clupeids (e.g., herring) populations are likely to be significant, although it is difficult to assess mortality. Since herring exhibit a marked schooling behavior, the virus has a high probability of being passed from fish to fish in a school, thereby elevating the prevalence over time. Its ability to pass from one school to another, either directly or indirectly, and to be passed to predator species by consumption or other mechanisms has not yet been tested. However, these are the most logical mechanisms by which the pathogen presence has become so extensive in the Pacific Northwest. In addition to manifestation of VHS in the herring population in Alaska and the Pacific Northwest, clupeids from this area have been afflicted with the fungal pathogen Ichthyophonus hoferi, which may also have an adverse affect on the herring population. This pathogen has been associated with epizootics and population declines in eastern US, the Gulf of St Lawrence and the North Sea (Daniel, 1933; Sindermann, 1958; Mellergård and Spanggård, 1997). The pathogen tends to be present in many herring populations, but at relatively low prevalence (<5%). However, it was found that in a survey of herring from Prince William Sound, Alaska in 1994 after a significant decline in herring harvest, that the prevalence was 29% (Marty et al., 1998). However, in a recent analysis of the data and development
236
Long-Term Ecological Change in the Northern Gulf of Alaska
of a model of the estimated importance of factors involved in the decline of the herring population, it was shown that while the fungal parasite was present at elevated levels, its impact on the herring population was not significant. Rather, the model indicated that VHS, even though present at lower levels, was significantly associated with the population decline (Marty et al., 2003). This example shows the potential for disease to affect fish populations significantly. It also, however, indicates the complex nature of pathogens and the populations they afflict. In this instance, the effects of the multiple pathogens present were not correlated with the prevalence of infection, but rather the severity of the disease associated with the pathogen.
Diseases of Reptiles (http://nationalzoo.si.edu/ConservationAndScience/ AquaticEcosystems/SeaTurtles/deem.cfm) Because several species of sea turtles are found on the US Endangered Species list, there is considerable concern for any impediment to their survival. There are several diseases of these species that have been detrimental to local populations. Two viral diseases have been especially problematic: fibro-papillomatosis (FP) and the respiratory lung–eye–trachea disease (LETD). Both of these diseases appear to be herpes virus-associated, although causality has not yet been definitively determined (Haines et al., 1974). The herpes virus of endangered Kemp–Ridley (Lepiochelys kempii) and other sea turtle species was first noted in the Caribbean and off the coast of Florida. In one survey, 21% of the Florida green and loggerhead turtles (Chelonia mydas and Caretta caretta) were seropositive for virus antibodies, indicating exposure to the virus. Testing of plasma from nesting adult turtles indicated that 11/13 were seropositive, indicating an extensive exposure to the virus. Similarly, approximately 30% of green turtles sampled in Hawaii have been found to suffer from clinical fibropapillomatosis. However, in keeping with the nature of these viruses, they are difficult or impossible to grow outside the host’s body and therefore, little laboratory work has been done on these agents. There is also little information about the actual impact of the disease on the populations of afflicted turtles, although the high prevalence levels reported in some populations would suggest a potentially severe effect.
Diseases of Birds (http://www.nwhc.usgs.gov/pub_metadata/index.html) As mentioned above, there have been few documented cases of infectious diseases in marine birds. Epizootics have been noted, to be sure, but they have been either found inland, such as the significant outbreaks of avian botulism, cholera, avian pox, and Newcastle disease of waterfowl, including double crested cormorants (Phalocrocorax auritus) in the Salton Sea, CA and Canada (O’Meara and Witter, 1971; Karstad, 1971;
Agents of Ecosystem Change
237
Wobeser et al., 1993; Wobeser, 1997), or have been found to be associated with toxins rather than being infectious by nature. Aspergillosis, a fungal disease, is a common infection in birds that are captured for rehabilitation after injury or oil spills, and in wild waterfowl that feed on contaminated grain (Friend, 2001). However, none of these diseases has been known to be of sufficient magnitude to have effects on populations of marine birds in the marine environment itself.
Diseases of Marine Mammals (http://www.nmfs.noaa.gov/pr/health/) The marine mammals reside at the pinnacle of both the marine food chain and the human concern for marine creatures. Consequently, much attention has been paid to factors, including infectious disease, that reduce populations of both cetaceans and pinnipeds. It is in this group of marine animals that the most amazing and alarming disease to date has occurred: phocine distemper. Abruptly, in late 1988 and 1989, significant incidents of spontaneous abortion occurred in grey seals (Halichoerus grypus) in the western end of the Baltic Sea, just off the coast of Denmark (Deitz et al., 1989). By the next year, from April to August, more than 18,000 seals had succumbed to a new disease designated as phocine distemper. The disease is caused by a paramyxovirus, the morbillivirus, a relative of the viruses that induce mumps in humans and canine and feline distemper (Crosby et al., 1988). Molecular studies have indicated that the virus is closely related to, but distinct from, canine distemper and was unknown until this serious outbreak of disease. The signs of disease are primarily respiratory in nature, and the mortality at most haul-out sites exceeded 50%. It is estimated that approximately half of the European population of harbor seals died directly or indirectly from the disease during the 1989 season. The most astonishing facet of the outbreak was the rapidity with which it spread. In a period of only a few months, the disease spread from its initial site to Norway, Sweden, the Belgian–German coast of the Waddell Sea to England, Scotland and Ireland, an area of more than 1500 km2. (Fig. 3.19) The mortality was somewhat lower in the British Isles than in the continent, and this may have been due to genetic differences in susceptibility to disease in populations of seals in various areas of the North Sea. The rapid spread of this catastrophic disease, that was undetected anywhere in the world previously, from one area to another in northern Europe is troubling and offers an intriguing insight into the potentially devastating effects of emerging diseases. One of the most intriguing scientific studies to arise from this short, but brutal epizootic, was that of Grenfell and his colleagues (1992). These investigators developed a model of the disease that incorporated all of its salient characteristics and predicted that, given that the population of survivors was likely resistant to reinfection for life, it would take approximately a decade for the population of newly recruited susceptible seals to build to a density sufficient to support another epizootic. In 2002, another epizootic of phocine distemper occurred in the same population, arising from the
238
Long-Term Ecological Change in the Northern Gulf of Alaska
Figure 3.19: Distribution of phocine distemper in harbor seals in northern Europe, 19881989. Initial report of mortality recorded at Anholt (Red star) in April, 1988. Further reports of mortality indicated by colored dots according to month of first report.
same location that had been initially noticed 13 years earlier (Jensen et al., 2002). While it was not quite as extensive or devastating as the initial epizootic, it killed nearly one quarter of the population of seals in the same area. It remains to be seen if this decadal oscillation will occur repeatedly on an extended temporal basis. It remains unclear what the source of infection was in 2002, since the virus was not detected in seal populations in the inter-epizootic period. This is similar to measles and mumps in humans where biennial epidemics are the norm when a sufficient density of susceptible individuals has been achieved by new births and a re-introduction of the pathogen is necessary to initiate a new epidemic because survivors are immune and non-contagious.
3.6.5. Population Effects and Implications The examples cited in detail above portray the sometimes drastic effects that infectious diseases can have on the populations of marine animals. The populations affected may suffer from direct mortality or may be affected indirectly such as by increased susceptibility to predation when physically compromised during or after the disease episode, or by reduced fecundity. The extent of the effects are determined by a number of factors, including the severity of the disease itself, the resiliency of the affected population and the trophic level of the diseased organisms.
Agents of Ecosystem Change
239
The severity of the disease is determined by the amalgamation of host, pathogen, and environmental factors. Diseases inducing a high mortality rate are often acute in nature, lasting only a short time because they rapidly deplete the population of susceptible animals and the infectious agent is eliminated until the pathogen is re-introduced to a new, sufficiently large population of susceptible animals. This is the case for phocine distemper as well as its human equivalent, mumps, which has been extensively studied by those epidemiologists tracking disease patterns (including Grenfell who has worked on morbillivirus infections in humans and seals). It may seem as though the increase of the population by recruitment of new generations may serve only as fodder for the new generation of disease that is initiated when a sufficient population of susceptible animals is present. This same pattern of epizootics separated by long time intervals may also be seen in fungal diseases of Atlantic herring (Clupea harengus) which seems to cause little effect during the inter-epizootic period. By contrast, a disease like bitter crab disease in Alaska, MSX in Eastern oysters (Crassostrea virginica), and perhaps VHS in Pacific herring, tends to be present at relatively constant prevalence levels over long periods of time, oscillating in magnitude, but always present at some enzootic level. Diseases of this nature may more effectively compromise the population than, for example, phocine distemper, because the population will be truncated constantly and lead to a smaller overall harvestable population because of the constant reduction of the population “ceiling” due to disease. This chronic, persistent type of disease may also have a more substantial effect on the ability of the animal population to “rebound” from disease. Even in a fecund species such as the Eastern oyster, each of which produces millions of progeny per year, the constant pressure of disease mortality on the recruited year class as well as the parental cohorts will prevent populations from reaching normally expected levels. Thus, the oyster population of the Chesapeake Bay area and Mid-Atlantic coast of the US has been depressed since the 1950s due in part to the toll taken by the MSX parasite and others. With the severe harvest pressure on these animals, the mortality due to disease, and a lack of full understanding of the mechanisms of pathogen transmission, it is unlikely that the population will achieve a significant rebound in the near future. The ramifications of disease in a discrete population of marine animals can be minimal or severe depending, to some extent, upon the trophic level of the affected animals. It is an example of the aphorism “For want of a nail, the shoe was lost; for want of a shoe, the horse was lost ...” For example, while the phocine distemper outbreaks may have had a considerable impact on the population of harbor seals in Europe, and extracted considerable emotional attention from humans, there may have been few detrimental impacts on the rest of the animal population of the North Sea region. Because the hosts are high on the food chain, and aside from orcas and a few other species, the seals are not consumed by many others, there might have been some short-term positive effects for species that serve as prey to these seals, but none has
240
Long-Term Ecological Change in the Northern Gulf of Alaska
been noted. There could also be some prey switching by seal predators that might have negative effects on other prey of harbor seal predators. On the other hand, when species low on the trophic scale die in large numbers, significant impacts can accrue to predator species. For example, corals provide nourishment and shelter for a truly broad array of marine animals. The demise of large patches of corals which leave only denuded calcareous accretions behind can have obvious adverse affects on the many species dependent on them for sustenance and protection. And, it appears that the diseases of corals are long term, thereby affecting their dependent species for extended periods. Likewise, herring lost to disease are also lost to the array of fishes, birds and mammals that readily consume them with relish. The geographic dispersal of a pathogen has considerable affect on the extent of the detriment to a host population. If the host is limited in its distribution and the pathogen broad, the impact can be severe. Most of the pathogens discussed have distributions that are limited relative to the distribution of the hosts. While VHS in marine fishes is constrained to northern Europe and the Northeastern Pacific rim, the host species, predominantly herring and cod are widely distributed and present in high numbers. Similarly, the agent of phocine distemper in harbor seals, while locally devastating will have little effect on the worldwide population of these pinnepeds which are distributed throughout the Northern Atlantic and Pacific Oceans. One of the frustrating facets of diseases of marine animals is the distinct lack of our ability to manipulate the process. Thus, in a situation such as the variety of coralline diseases mentioned above, the host animals are stationary and susceptible to “predation” by the bacteria and various other pathogens in their proximity. The magnitude of the affected area coupled with the severe environmental strictures placed on any treatment regimes precludes any human amelioration of the situation. An obvious counter measure to the effects of infectious disease on marine animals is to reduce anthropogenic contamination. Given the direct and indirect impacts that these substances can have on all marine life, this admittedly difficult task would likely improve the survival of all creatures inhabiting our oceans. Some management potential exists for altering the balance of disease for marine species by considering aquaculture practices. At the outset, it should be pointed out that no causal association has been proven between aquaculture and direct infection of nearby wild marine animals leading to epizootics, and that one cannot conclude that the sources of infection for the diseases discussed were in any way directly associated with aquaculture. If any connection can be made, it is that when cultured animals experience disease it is often more severe than those experienced by wild marine animals due to the inherent high densities at which cultured species are held, and because they are confined in static containers. However, the larger impact may be the dispersal of pathogens with movement and shipment of animals from one culture facility to another. This method of geographic dispersal of pathogens “leapfrogs” the spread of the pathogen by natural means, which would generally range from an original
Agents of Ecosystem Change
241
central focus to the immediate population nearby from where it would then inexorably lead outward. This rate, however, is much less rapid than dispersal with intentional or unintentional movement by humans. Although there is no evidence that this has resulted in direct infection of wild animals, this may be the case at sometime in the future. Thus, by utilizing extant surveillance techniques for known pathogens prior to the movement of cultured animals from one location to another, the risk of transferring novel pathogens to susceptible populations of wild marine animals can be avoided. While we can make efforts that will lead to some reductions in the impact of infectious disease on marine animals, it is an unfortunate truth that the dance of death between infectious agents and their hosts is the natural way of things. Short-term impacts of disease, such as phocine distemper and MSX, are part and parcel of the ebb and flow of life in the seas. Each member of the disease diorama seeks competitive advantage; each will have short-term victory and defeat. Inexorably, however, equilibrium will be restored by long-term or short-term changes, albeit perhaps with altered proportions of host and agent. Our best efforts should be aimed to avoid disturbing the resonance of the system and allowing nature to take its course.
3.7. Contaminants Robert B. Spies and Stanley Rice Introduction The Gulf of Alaska is sparsely populated by humans and lacks heavy manufacturing industry and agriculture, but toxic chemicals affect even this remote part of the world ocean. Heavy metals, organic, and organo-metallic contaminants occur only in trace amounts in the Gulf waters, except in the case of an accidental spill, such as the 1989 Exxon Valdez oil spill. However, many anthropogenic chemicals from distant sources, most notably some of the persistent chlorinated hydrocarbons (e.g., polychlorinated biphenyls, PCBs; 1,1,1-trichloro-2,2-bis-(4′-chlorophenyl)ethane, DDT and hexachlorobenzene (HCB), and alkylated metals (e.g., methyl mercury), are taken up and concentrated in organisms orders of magnitude over their concentrations in sea water, and further concentrated in food chains, so that they can reach potentially toxic concentrations in wildlife and human consumers. These toxins can cause damage, initially in the cells of marine organisms, and then affect their physiology as expressed by changes in fitness, growth and reproduction. Contaminants act like pathogens or predators, removing organisms and limiting the productivity and recovery of populations. Developing embryos and young animals are generally the most vulnerable life stages to the effects of contaminants accumulated in their tissues. Spilled oil also has direct physical effects on sea birds, mammals and intertidal organisms.
242
Long-Term Ecological Change in the Northern Gulf of Alaska
Nearly every chemical is toxic at some concentration, so dosage is the most important consideration in the toxicology of marine organisms. Also, the ultimate yardstick of harm is whether a contaminant affects the population.
Contaminants in Alaska: Types and Sources The major sources of contamination in the Gulf of Alaska fall into two categories: (1) point-source oil spills, sewage and industrial discharges; and (2) non-point chronic inputs of metals and organic compounds. There are scattered point sources in Alaska, e.g., from municipal waste discharges, boat harbors and the Alyeska pipeline terminal, but contaminant concentrations are typically not high enough more than a kilometer from the source to be a concern. The Exxon Valdez oil spill in 1989 was a rare, large chemical spill. The spill demonstrated the scales of risk: acute effects over months and hundreds of kilometers to chronic effects over tens of kilometers for a decade or more (see Chapter 5 on the impacts of the Exxon Valdez oil spill). Spills also occurred when oil storage tanks on the shores of PWS ruptured during the 1964 earthquake and when the Selendeng Ayu broke up near Unalaska in the Aleutian Islands in 2004. There is also wide-spread, diffuse, low-level contamination throughout the arctic and sub-arctic. This source includes organic compounds, such as pesticides, PCBs, and organo-metallic compounds like methyl mercury that have a broad-scale distribution, persistence, limited degradation, and can accumulate in upper-level predators. Many of these compounds are transported to Alaska through cycles of evaporation and condensation, entrapped in sub-arctic and arctic ecosystems with cold temperatures, accumulated in lipid-rich organisms and passed through the food web in ever increasing concentrations (AMAP, 2004). At low tissue concentrations, these compounds do not cause acute toxicity but may cause long-term chronic effects such as depressed growth and fitness, reproductive problems, and cancers in apex predators where they accumulate. The implications of high concentrations of chlorinated hydrocarbons in predators such as killer whales in the Gulf of Alaska (e.g., Hayteas and Duffield, 2000; Ylitalo et al., 2001) are not known. However, polar bears (Ursus maritimus) in the Arctic are more amenable to study, and hormonal disorders have been linked to contaminants in them (Oskam et al., 2003, 2004). Another source of coastal contamination is the return of spawning anadramous fish to their natal streams. Their decaying bodies release the loads of contaminants accumulated in their life at sea. Specifically, sockeye salmon (Oncorhynchus nerka) returning to spawn in the Alaskan lakes resulted in grayling (Thymallus arcticus) with elevated concentrations of PCBs compared to grayling in lakes without spawning sockeye (Ewald et al., 1998). The magnitude of this source to the food web has not been compared to PCBs that are deposited in coastal areas from atmospheric fall out. Most of the persistent organic pollutants (POPs), e.g., pesticides, flame retardants and PCBs, are manufactured elsewhere in the northern hemisphere and are alien to the
Agents of Ecosystem Change
243
marine environment. These compounds have half-lives of over two months in the environment, bioaccumulation factors greater than 5000, the potential for long-range transport, and are potentially toxic. Such compounds include halogenated hydrocarbons such as pesticides, PCBs, dioxins, and furans. In addition to POPs, heavy metal contaminants are present in Alaskan habitats, including mercury, cadmium, chromium, arsenic, lead and silver. One study of Alaskan marine mammals sampled between southeastern Alaska and the Pribilof Islands found PCBs, DDT, chlordanes, hexachlorocyclohexanes (HCHs), hexachlorobenze (HCB), dieldrin, butyltins, arsenic, mercury, cadmium, and lead in their tissues (Barron et al., 2003). With over 100,000 chemicals on the market and an additional 1000 to 2000 new ones introduced annually, there are likely new toxic compounds in the environment whose concentrations are increasing. In addition, combustion and industrial processes result in the inadvertent production of unregulated chemicals (AMAP, 2004). Oils, their byproducts and polynuclear aromatic hydrocarbons (PAH) are present in trace amounts discontinuously throughout Alaska as urban runoff, automobile and vessel exhaust, and fuel oil spills as well as in natural sources such as coal deposits; but it is the aerial transport of POPs into Alaska from foreign sources that are of the greatest concern. Trends in contamination of the GOA are being tracked in three programs that analyze organism tissues. Two of these, the National Mussel Watch Program (http://ccma.nos.noaa.gov/cit/data/mw_monitoring.html) and the Regional Citizens Advisory Council’s Long-Term Environmental Program (LTEMP) (http://www. pwsrcac.org/projects/EnvMonitor/Itemp.html) analyze contaminants in intertidal mussels. The former program has three monitoring sites in the GOA and analyzes over 100 individual contaminants, including PAH, PCBs, pesticides, chlorinated hydrocarbons and metals. The latter program has 10 sites in PWS and analyzes hydrocarbons. A suite of contaminants (including chlorinated hydrocarbons, pesticides, mercury are also being monitored in the eggs of five species of sea birds around the coast of Alaska in the Seabird Tissue Archival and Monitoring Project (STAMP) (http://www.absc. usgs.gov/research/ammtap/stamp.htm). In this latter program, concentrations of organic contaminants (DDE, PCBs and Hg) are 4–5 times lower in Alaska than in the lower 48 states but generally higher in the Gulf of Alaska than in the Bering Sea (Vander Pol et al., 2006). In the LTEMP program, concentrations of PAH are highest at the station near the Valdez Oil Terminal.
Mechanisms of Contaminant Action Toxins act in many different ways, but most alter molecular processes in cells (Klaasen, 2001). They either react directly with molecules or receptors in an abnormal way or they alter the cellular environment. Some compounds are toxic as they are (e.g., ionic copper), but some are only toxic after chemical transformation (e.g., PAH). Some toxins bind to the active site of enzymes and render them ineffective, e.g., carbon monoxide, or the toxin binds to other sites and changes the conformation
244
Long-Term Ecological Change in the Northern Gulf of Alaska
of the enzyme. Other toxins (phenols, carbon tetrachloride) lose or gain an electron, transforming to free radicals that react with DNA, proteins, and lipids (see Livingstone, 2003). Other toxins are structurally similar to hormones (o,p-DDE resembles estrogen), bind to hormone receptors, and either block natural signaling by hormones or cause signaling to occur at the wrong time (Spies and Thomas, 1997). Some contaminants, i.e., spilled petroleum, have direct physical effects on organisms, diminishing their insulation from cold water in the case of birds and mammals (Kooyman and Costa, 1978), or directly bury animals that are fixed in the intertidal zone, e.g., barnacles and seaweeds. Highly weathered oil has been shown to interfere with the operation of ion channels after transformation by P4501A in Pacific herring (Incardona et al., 2004). Alteration of biochemical pathways have been well documented for some contaminant responses, but it is not always clear what the ecological relevance of the response is unless the molecular change can be tied to long-term changes in fitness, growth, or reproduction. One of the best contaminant biomarkers is the production of xenobiotic detoxification enzymes (cytochome P450s, usually the P4501A form is measured as it is particularly responsive to contamination). In vertebrates, these are biomarkers for exposure to oil, PCBs, and some other organic compounds, because the enzyme is manufactured in the cell in response to some of these compounds. During the transformation of waterinsoluble aromatic compounds by these enzymes to water-soluble products, reactive metabolites are produced which may bind to DNA and proteins, impairing cellular processes. The net benefit from enzyme induction depends on the extent of any damage from oxidative reactions of the intermediate products and the extra energy required for their metabolism versus the advantage of ridding the tissues of the PAH. The induction of these enzymes is not necessarily indicative of damage. However, in some organisms P450 induction has been correlated with negative effects, e.g., on reproductive success in flounder (Spies and Rice, 1988), and immune suppression in trout (Springman et al., 2005). One example of an apparent mechanism leading to deformed Pacific herring embryos by weathered oil is as follows. The PAH in weathered petroleum is metabolized by P4501A, the PAH transformation products interfere with ion transport channels, the circulatory system goes out of ionic balance, edema develops in the heart and impaired circulation leads to morphological deformities, followed by death or impairment in the adult (see Incardona et al., 2004). So, molecular changes and repair can alter the physiology of an organism and its fitness. That is, toxic chemicals compromise the ability of an organism to maintain itself, migrate, feed, reproduce, and pass on its genes. While toxins interfere with the smooth functioning of the machinery of life, as we have seen above, they also often induce a response that requires energy. The fitness of the individual is directly related to the energy available to the organism and the environmental stresses it faces: toxicants, disease challenges, and predator escape. These stresses directly draw on the
Agents of Ecosystem Change
245
energy of the organism and diminish its ability to maximally respond to other needs and demands. Measures of growth or reproductive success are common measures of fitness, either at the individual level, or at the population level. It can be expected that fitness, as measured by these processes, will be at risk from contaminants that either damage the normal machinery of cells or invoke additional energy-demanding metabolic responses. For example, wild pink salmon fry collected in 1989 from nearshore waters in oiled areas grew at half the rate of fry from reference areas (Wertheimer and Celewycz, 1996). The reduced growth in the juveniles was the net result of contaminant-induced changes, including costs of oil metabolism, because the difference between oiled and un-oiled areas was primarily the oil, and not food availability or other factors. Supporting evidence comes from laboratory tests, where decreased growth was observed in pink salmon fry exposed to sublethal concentrations of aromatic hydrocarbons in the water column (Moles and Rice, 1983) as energy for growth was diverted to hydrocarbon metabolism and excretion (Thomas and Rice, 1979). A conceptual model that shows the energy demands on organisms that maintain a constant high body temperature (e.g., sea otters) from contaminant exposure is shown in Fig. 3.20. In this figure, the period of net energy acquisition (spring and summer) is contrasted with winter conditions where there is a net loss of energy and storage is drawn down. So, the energy demanding response to contaminants (enzyme production and contaminant metabolism) will deplete energy storage which could hasten the arrival of a survival bottle neck in the winter. Reproductive competence has perhaps the highest ecological relevance of all these physiological responses. The success of bird colonies, for example, is the net result of transferring energy to eggs, and later to chicks, spending the least amount of time away from the nest foraging, fending off predators, and this could
Figure 3.20: Hypothesized relative cost of energy for contaminant responses to P450IA inducers compared to other demands on homeotherms in Alaskan waters; E = energy intake, G = growth, R = reproduction.
246
Long-Term Ecological Change in the Northern Gulf of Alaska
be compromised for either the adults or the developing chicks by accumulated contaminants. In summary, the outcome of sublethal contaminant exposure can be a reduction of fitness from biochemical alterations and from energy demand. Contaminant exposure reduces the fitness and efficiency of highly adapted organisms and so compromises their ability to survive the rigors of their environment, grow, and reproduce. Small reductions in fitness are at least additive and are expressed on the population level as fewer fit individuals to carry on the population.
3.7.1. Significance of Life Stage at Time of Impact Early life stages are generally more vulnerable than adults to the effects of toxic compounds. In a developing embryo, cellular damage may occur without an apparent effect. The embryo may develop normally for some time, as it may not need the damaged tissue until a later life stage, thus the consequences of early damage may not be apparent for some time. This was eloquently demonstrated in a controlled laboratory exposure of salmon embryos to low levels of crude oil, which were then tagged and released to the marine environment. A variety of symptoms were evident in many of the exposed larvae, including mortalities, deformities, and reduced growth in normalappearing animals. Delayed impacts on growth were evident at very low doses of oil (parts per billion), and the ecological consequences were evident when fewer adults returned from the oil-exposed groups than from the controls (Heintz et al., 2000). The early life stages are vulnerable to predation and, when exposed to low doses of oil reduced survival from increased predation become even more severe.
3.7.2. Floating Oil and Surface-dwelling Animals Surface dwelling and intertidal animals are triply vulnerable to oil spills. First, they take up petroleum in their food and from the water as do sub-surface dwellers. Second, the air-breathing marine mammals and birds inhale oil fumes in the first days of the spill, which contain the more volatile toxic components: benzene, toluene, and xylene. These fast-acting monoaromatic hydrocarbons enter the lungs and, are distributed in the blood to the nervous system, possibly leading to acute narcosis. Third, the surface-dwelling animals will also be physically coated with the oil when a slick enters their habitat. Oil on the fur of mammals or feathers of birds causes two further problems. Preening is stimulated, leading to direct ingestion of oil. In addition, oil destroys the insulating properties of pelage leading to hypothermia and death. For example, if a third of a sea otter’s coat is oiled, it will become hypothermic and dies (Kooyman and
Agents of Ecosystem Change
247
Costa, 1978). Except for the marine mammals with a heat-retaining fat layer, like seals and sea lions, higher vertebrates rely on very active metabolism and their pelage to maintain a differential of up to 100°F between their body and the surrounding water. To maintain a high rate of metabolism sea otters, for example, must eat 20% of their body weight a day. So for the smaller animals, like sea otters and birds, staying warm in the Gulf waters takes a lot of energy, and an oil coating on pelage is a serious threat. The death of more than two hundred thousand sea birds and more than 2000 sea otters in the cold waters of Alaska in the spring of 1989 attests to these combined risks. For the intertidal invertebrates with little or no ability to move away from surface contamination, oil toxicity may be insignificant compared to suffocation from oil coating. Daily tidal excursions can re-expose these animals to recurring oil slicks and coatings. Intertidal animals are often among the heartiest and most tolerant to stress, but they will die if an oil coating prevents exchange of oxygen and wastes with the seawater. In addition, intertidal communities can be exposed to harsh clean-up procedures. In some bays inundated with Exxon Valdez oil in 1989, clean-up was repeated several times on the same beach over the course of several weeks, often with hot water that killed many of the intertidal animals.
References AFB and ADF (Alaska Fisheries Board and Alaska Department of Fisheries). 1954. 1954 annual report. Alaska Fish and Game Commission and Alaska Department of Fish and Game, Report 6, Juneau, Alaska. Alexander, M. A. and C. Deser. 1995. A mechanism for the recurrence of wintertime midlatitude SST anomalies. J. Phys. Oceanog. 25: 122–137. Alexander, M. A., I. Blade, M. Newman, J. R. Lanzante, N.-C. Lau, and J. D. Scott. 2002a. The atmospheric bridge: the influence of ENSO teleconnections on air-sea interaction over the global ocean. J. Clim. 15: 2205–2231. Alexander, M. A., M. S. Timlin, and J. D. Scott. 2002b. Winter-to-winter recurrence of sea surface temperature, salinity and mixed layer depth anomalies. Prog. Oceanogr. 49: 41–61. Alverson, D. L. 1992. A review of commercial fisheries and the Steller sea lion (Eumetopias jubatus): the conflict arena. Rev. Aquat. Sci. 6: 203–256. Alverson, D. L., A. T. Pruter, and L. L. Ronholt. 1964. A study of demersal fishes and fisheries of the Northeastern Pacific Ocean. H. R. MacMillan Lectures in Fisheries, University of British Columbia. 190 pp. AMAP. 2004. AMAP Assessment 2002: Persistent Organic Pollutants in the Arctic. Arctic Monitoring and Assessment Programme (AMAP), Oslo, Norway. xvi+310 pp. Anderson, J. and D. Hedgecock. 2004. Non-native oysters in the Chesapeake Bay, National Academies Press, Washington, DC, 2001, 344 pp. Anderson, R. M. and R. M. May. 1979. Population biology of infectious diseases: Part I. Nature 280: 361–367.
248
Long-Term Ecological Change in the Northern Gulf of Alaska
Anderson, R. S., C. Giam, L. Ray, and M. Tripp. 1983. Effects of environmental pollutants on immunological competency of the clam Mercenaria mercinaria: Impaired bacterial clearance. Aquat. Toxicol. 1: 187–195. Angliss, R. P. and K. L. Lodge. 2004. Alaska marine mammal stock assessments, 2003. NOAA Technical Memorandum NMFS-AFSC-144. 230 p. Barron, M. G., R. Heintz, and M. M. Krahn. 2003. Contaminant exposure and effects in pinnipeds: implications for Steller sea lion declines in Alaska. Sci. Total Environ. 311: 111–133. Baum, J. K. and R. A. Myers. 2004. Shifting baselines and the decline of pelagic sharks in the Gulf of Mexico. Ecol. Lett. 7: 135–145. Baxter, R. 1971. Earthquake effects on clams of Prince William Sound, In: The Great Alaskan Earthquake of 1964: Biology. NAS Publ. 1604, National Academy of Sciences, Washington, DC. Bean, T. H. 1887. The cod fishery of Alaska. In: G. B. Goode (Ed.), Fisheries and Fishery Industries of the U.S., Section 5, Vol. 1, History and Methods of the Fisheries, US Government Printing Office, Washington, DC, pp. 198–226. Berkeley, S. A., C. Chapman, and S. M. Sogard. 2004a. Maternal age as a determinant of larval growth and survival in a marine fish, Sebastes melanops. Ecology 85: 1258–1264. Berkeley, S. A., M. A. Hixon, R. J. Larson, and M. S. Love. 2004b. Fisheries sustainability via protection of age structure and spatial distribution of fish populations. Fisheries 29: 23–32. Boldt, J. 2003. Ecosystem consideration for 2004. Appendix C. Stock assessment and fishery evaluation report for the groundfish resources of the Bering Sea/Aleutian Islands Regions. North Pacific Fishery Management Council, Anchorage, AK. Bond, N. A., J. E. Overland, M. Spillane, and P. Stabeno. 2003. Recent shifts in the state of the North Pacific, Geophys. Res. Lett. 30: 2183. Brownell, R. L., Jr., A. V. Yablokov, and V. A. Zemsky. 2000. USSR pelagic catches of North Pacific sperm whales, 1949–1979: conservation implications. In: A. V. Yablokov and V. A. Zemsky (Eds.), Soviet Whaling Data (1949–1979). Center for Environmental Policy Moscow, pp. 123–130. Bullock, G. L. and R. L. Herman. 2003. Bacterial kidney disease of salmonids caused by Renibacterium salmoninarum. USFWS, Fish Disease Leaflet # 78, Kearneysville, WV. Clarke, A. J. 1983. The reflection of equatorial waves from oceanic boundaries. J. Phys. Oceanogr. 13, 1095–1125. Clarke, A. J. and S. Van Gorder. 1994. On ENSO coastal currents. J. Phys. Oceanog. 24: 661–680. Clausen, D. M. 1981. Summer food of Pacific cod, Gadus macrocephalus, In: Coastal Waters of Southeastern Alaska. Fish. Bull. 78: 968–973. Comeau, L. A., S. E. Campana, and M. Costonguay. 2002. Automated monitoring of a large-scale cod (Gadus morhua) migration in the open sea. Can. J. Fish. Aquat. Sci. 59: 1845–1850. Conover, D. O. and S. B. Munch. 2002. Sustaining fisheries yields over evolutionary time scales. Science 297: 94–96. Crosby, S. L. and 11 others. 1988. Characterization of a seal morbillivirus. Nature 336: 115–116. Daniel, G. 1933. Studies on Ichthyophonus hoferi, a parasitic fungus of the herring, Clupea harengus. I. The parasite as it is found in the herring. Amer. J. Hyg. 17: 262–276.
Agents of Ecosystem Change
249
Deitz, R., M-P. Heide-Jφrgensen, and T. Härkönen. 1989. Mass deaths of harbor seals (Phoca vitulina) in Europe. Ambio 5: 258–264. Deser, C., M. A. Alexander, and M. S. Timlin. 1999. Evidence for a wind-driven intensification of the Kuroshio Current Extension from the 1970s to the 1980s. J. Clim. 9: 1840–1855. Dunier, M. and A. K. Siwicki. 1993. Effects of pesticides and other organic pollutants in the aquatic environment on immunity of fish: a review. Fish & Shellfish Immunol. 3: 423–438. Ewald, G., P. Larsson, H. Linge, L. Okla, and N. Szarzi. 1998. Biotransport of organic pollutants to an inland Alaska lake by migrating sockeye salmon (Oncorhynchus nerka). Arctic 51: 40–47. Ford, S. E. 1985. Effects of salinity on survival of the MSX parasite Haplosporidium nelsoni in oysters. (Haskin, Stauber and Mackin) J. Shellfish Res. 5: 85–90. Ford, S. E. and H. H. Haskin. 1988. Management strategies for MSX (Haplosporidium nelsoni) disease in eastern oysters. In: W. S. Fisher (Ed.), Disease Processes in Marine Bivalve Mollusks. Amer. Fish. Soc. Special Publ. 18: 249–256. Friend, M. 2001. Aspergillosis. In: M. Friend and J. C. Franson (Tech. Eds.), E. A. Ciganovich (Ed.), Field Manual of Wildlife Diseases: General Field Procedures and Diseases of Birds. Bio. Res. Div. Info. and Tech. Report 1999–2001. U.S.D. Interior and U.S.G.S., Washington, DC, pp. 129–133. Gambell, R. 1976. World whale stocks. Mamm. Rev. 6: 41–53. Gilmore, R. M. 1978. Right whale. In: D. Haley (Ed.), Marine mammals of eastern North Pacific and Arctic waters. Pacific Search Press, Seattle. pp. 62–69. Gladfelter, W. B. 1991. Population Structure of Acropora palmata on the windward fore reef, Buck Island National Monument, St. Croix, U.S. Virgin Islands. U.S. Virgin Islands: US Department of the Interior, National Park Service. 172 pp. Green, E. and A. W. Bruckner. 2000. The significance of coral disease epizootiology for coral reef conservation. Biol. Conser. 96: 347–361. Gregr, E. J., N. Nichol, J. K. B. Ford, G. Ellis, and A. W. Trites. 2000. Migration and population structure of northeastern Pacific whales off coastal British Columbia: an analysis of commercial whaling records from 1908–1967. Mar. Mamm. Sci. 16: 699–727. Grenfell, B. T., M. E. Lonergan, and J. Harwood. 1992. Quantitative investigations of the epidemiology of phocine distemper virus (PDV) in European common seal populations. Sci Total Environ. 115: 15–29. Haines, H. G., A. Rywlin, and G. Rebell. 1974. A herpes virus disease of farmed green turtles (Chelonia mydas). J. Avault, (Ed.), Workshop of the World Mariculture Society, Charleston, SC (USA). Hampton, M. A., P. R. Carlson, H. J. Lee, and R. A. Feely. 1987. Geomorphology, sediment and sedimentary processes. In: D. W. Hood and S. T. Zimmerman (Eds.), The Gulf of Alaska. MMS publication number: OCS study, MMS 86-0095, pp. 93–143. Hanawa, K., S. Ishizaki, and Y. Tanimoto. 1996. Examination of the strengthening of the wintertime mid-latitude westerlies over the North Pacific in the 1970s. J. Meteorol. Soc. Jpn. 74: 715–721. Haskin, H. H. and J. D. Andrews. 1988. Uncertainties and speculations about the life cycle of the eastern oyster pathogen Haplosporidium nelsoni (MSX). In: Fisher, W. S. (Ed.), Disease Processes in Marine Bivalve Mollusks. Amer. Fish. Soc. Special Publ. 18: 5–22. Haskin, H. H., L. Stauber, and J. Macklin. 1966. Minchinia nelsoni n. sp. (Haplosporida, Haplosporidiidae) causative agent of the Delaware Bay oyster epizooitc. Science 153: 1414–1416.
250
Long-Term Ecological Change in the Northern Gulf of Alaska
Haven, S. B. 1971. Effects of land-level changes on intertidal invertebrates, with discussion of post-earthquake ecological succession. In: The Great Alaskan Earthquake of 1964: Biology. NAS Publ. 1604, National Academy of Sciences, Washington, DC, pp. 82–128. Haynes, T. L. and C. Mishler. 1994. The subsistence harvest and use of Steller sea lions in Alaska. Alaska Department of Fish and Game, Division of Subsistence, Technical Paper 198. Hayteas, D. I. and D. A. Duffield. 2000. High Levels of PCB and p,p′-DDE Found in the Blubber of Killer Whales (Orcinus orca). Mar. Pollut. Bull. 40: 558–561. Heintz, R. A., S. D. Rice, A. C. Wertheimer, R. F. Bradshaw, F. P. Thrower, J. E. Joyce, and J. W. Short. 2000. Delayed effects on growth and marine survival of pink salmon Oncorhyncus gorbusca after exposure to crude oil during embryonic development. Mar. Ecol. Prog. Ser. 208: 205–216. Hoover, A. A. 1988a. Harbor seal. In: J. W. Lentfer (Ed.), Selected marine mammals of Alaska: species accounts with research and management recommendations. Marine Mammal Commission, Washington, DC, pp. 125–157. Hoover, A. A. 1988b. Steller sea lion. In: J. W. Lentfer (Ed.), Selected Marine Mammals of Alaska: Species Accounts with Research and Management Recommendations. Marine Mammal Commission, Washington, DC, pp. 159–193. Hubbard, J. D. 1971. Distribution and abundance of intertidal invertebrates at Olsen Bay in Prince William Sound, Alaska one year after the 1964 earthquake. In: The Great Alaskan Earthquake of 1964: Biology. NAS Publ. 1604, National Academy of Sciences, Washington, DC, pp. 137–157. Hutchings, J. A. and J. D. Reynolds. 2004. Marine fish population collapses: consequences for recovery and extinction risk. BioSci. 54: 297–309. Huyer, A. 2003. Preface to special section on enhanced Subarctic influence in the California Current, 2002. Geophys. Res. Lett. 30: 8019. Incardona, J., T. K. Collier, and N. L. Scholz. 2004. Defects in cardiac function precede morphological abnormalities in fish embryos exposed to polycyclic aromatic hydrocarbons. Toxicol. Appl. Pharmacol. 196: 191–205. Jacob, K. H. 1984. Estimates of long-term probabilities for future great earthquakes in the Aleutians. Geophys. Res. Lett. 11: 295–298. Jacobson, K. C., M. R. Arkoosh, A. N. Kagley, E. R. Clemons, T. K. Collier, and E. Casillas, 2003. Cumulative effects of natural and anthropogenic stress on immune function and disease resistance in juvenile Chinook salmon. J. Aquat. Anim. Health. 15: 1–12. Jensen, M. H. 1965. Research on the virus of Egtved disease. Ann. N. Y. Acad. Sci. 126: 422–426. Jensen, T. M., M. Van de Bildt, H. H. Dietz, T. H. Andersen, A. S. Hammer, T. Kuiken, and A. Osterhaus, 2002. Another Phocine Distemper Outbreak in Europe. Science 297: 209. Jewett, S. C. 1978. Summer food of the Pacific cod, Gadus macrocephalus, near Kodiak Island. Fish. Bull. 76: 700–706. Joyce, T. M. and J. Dunworth-Baker. 2003. Long-term hydrographic variabililty in the Northwest Pacific Ocean. Geophys. Res. Lett. 30: 1043. Karstad, L. l97l. Pox, In: Davis, J. W. and others (Eds.), Infectious and Parasitic Diseases of Wild Birds. Ames, Iowa, Iowa State University Press. pp. 34–4. Kenyon, K. W. 1969. The sea otter in the eastern Pacific Ocean. N. Amer. Fauna 68: 1–352.
Agents of Ecosystem Change
251
Kinne, O. (Ed.), 1980. Diseases of Marine Animals. v. 1 General Aspects Protozoa to Gastropoda; v II. Bivalvia to Scaphopoda; v. III. Cephalopoda, Annelida, Crustacea, Chaetognatha, Echinodermata, Urochordata; v. IV, part 1, Pisces; v. IV, part 2, Reptilia, Aves, Mammalia. Biologische Anstalt Helgoland, Hamburg, Germany. Klaasen, C. D. (Ed.), 2001. Cassarrett & Doull’s Toxicology. The Basic Science of Poisons. 6th Edition. McGraw Hill, San Francisco, 1236 pp. Kocan, R. M., P. K. Hershberger, and J. R. Winton. 1998. Controlled field and laboratory studies of VHSV and Ichthyophonus in Pacific herring. In: G. Marty et al. (Eds.), Exxon Valdez Oil Spill Restoration Project Annual Report, Section II. Investigations of Disease Factors affecting declines of Pacific Herring Populations in Prince William Sound. 21 pp. Kocan, R. M., P. K. Hershberger, N. E. Elder, and J. R. Winton. 2001. Epidemiology of Viral Hemorrhagic Septicemia among Juvenile Pacific Herring and Pacific Sand Lances in Puget Sound, Washington. J. Aquat. Anim. Health 13: 77–85. Kooyman, G. and D. Costa. 1978. Effects of oiling on temperature regulation in sea otters. In: Wolfe, D. A. (Ed.), Marine Biological Effects of OCS Petroleum Development. NOAA/ERL; Boulder, CO (USA). NOAA-TM-ERL-OCSEAP-1.; NOAA Tech. Memo. pp. 216–22. Kruse, G. H., F. C. Funk, H. J. Geiger, K. R. Mabry, H. M. Savikko, and S. M. Siddeek. 2000. Overview of state-managed marine fisheries in the central and western Gulf of Alaska, Aleutian Islands, and Southeastern Bering Sea, with reference to Steller sea lions. Alaska Department of Fish and Game, Division of Commercial Fisheries, Regional Information Report 5J00-10, Juneau. Lagerloef, G. S. E. 1995. Interdecadal variations in the Alaska Gyre. J. Phys. Oceanog. 25: 2242–2258. Lang, G. M., C. W. Derrah, and P. A. Livingston. 2003. Groundfish food habits and predation on commercially important prey species in the eastern Bering Sea from 1993 through 1998. National Marine Fisheries Service, Alaska Fisheries Science Center, AFSC Processed Report 2003-04, Seattle, WA. Lau, N.-C. and M. J. Nath. 1994. A modeling study of the relative roles of tropical and extratropical SST anomalies in the variability of the global atmosphere-ocean system. J. Clim. 9: 2036–2057. Laughlin, W. S. 1980. Aleuts: survivors of the Bering Land Bridge. Holt, Rinehart and Winston, New York. Le Moullac, G. and P. Haffner. 2000. Environmental factors affecting immune responses in crustacea. Aquacult. 191: 121–131. Lensink, C. J. 1958. Predator investigation and control. In: Alaska Fish and Game Commission and Alaska Department of Fish and Game, 1958 Annual Report, Report 10, Juneau, Alaska, pp. 91–104. Lensink, C. J. 1960. Status and distribution of sea otters in Alaska. J. Mammal. 41: 172–182. Livingstone, D. R. 2003. Oxidative stress in aquatic organisms in relation to pollution and aquaculture. Rev. Med. Vet. 154: 427–430. Loughlin, T. R. and R. Nelson, Jr. 1986. Incidental mortality of northern sea lions in Shelikof Strait, Alaska. Mar. Mamm. Sci. 2: 14–33. Lowe, J. I., P. Wilson, A. Rick, and A. Wilson Jr. 1971. Chronic exposure of oysters to DDT, toxaphene and parathion. Proc. Nat. Shellfish Assoc. 61: 71–79. Mantua, N. J., S. R. Hare, Y. Zhang, J. M. Wallace, and R. C. Francis. 1997. A Pacific decadal climate oscillation with impacts on salmon production. Bull. Amer. Meteorol. Soc. 78: 1069–1079.
252
Long-Term Ecological Change in the Northern Gulf of Alaska
Marty, G. D., E. F. Freiberg, T. R. Meyers, J. Wilcock, T. B. Farver, and D. E. Hinton. 1998. Viral hemorrhagic septicemia virus, Ichthyophonus hoferi, and other causes of morbidity in Pacific herring Clupea pallasi spawning in Prince William Sound, Alaska, USA Dis. Aquat. Org. 32: 15–40. Marty, G. D., T. J. Quinn, G. Carpenter, T. R. Meyers, and N. H. Willits. 2003. Role of disease in abundance of a Pacific herring (Clupea pallasi) population. Can. J. Fish. Aquat. Sci. 60: 1258–1265. McMullen, J. C. and H. T. Yoshihara. 1969. Effects of delayed mating on the reproduction of king crab, Paralithodes camtschatica. J. Fish. Res. Bd. Can. 26: 2737–2740. ~o. Bull. Amer. Meteorol. Soc. 84: McPhaden, M. J. 2004. Evolution of the 2002/03 El Nin 677–695. Meier, W., M. Schmitt, and T. Wahli. 1994. Viral hemorrhagic septicemia (VHS) of nonsalmonids. Ann. Rev. Fish Dis. 4: 359–373. Mellegård, S. and B. Spanggård. 1997. An Ichthyophonus hoferi epizootic in herring in the North Sea, the Skagerrrrak, the Kattegat and the Baltic Sea. Dis. Aquat. Org. 28: 191–199. Merrick, R. L., T. R. Loughlin, and D. G. Calkins. 1987. Decline in abundance of the northern sea lion, Eumetopias jubatus, in Alaska, 1956–86. Fish. Bull. 83: 351–365. Meyers, T. R. and J. Winton. 1995. Viral Hemorrhagic septicemia in North America. Ann. Rev. Fish Diseases 6: 3–24. Meyers, T. R., T. M. Koeneman, C. Botelho, and S. Short. 1987. Bitter crab disease: a fatal dinoflagellate infection and marketing problem for Alaskan Tanner crabs Chionoecetes bairdi. Dis. Aquat. Organ. 3: 195–216. Miller, A. J. and N. Schneider. 2000. Interdecadal climate regime dynamics in the North Pacific Ocean: Theories, observations, and ecosystem impacts, Prog. Oceanog. 27: 257–260. Miller, A. J., D. R. Cayan, and W. B. White. 1998. A westward-intensified decadal change in the North Pacific thermocline and gyre-scale circulation. J. Clim. 11: 3112–3127. Miller, A. J., D. R. Cayan, T. P. Barnett, N. E. Graham, and J. M. Oberhuber. 1994. Interdecadal variability of the Pacific Ocean: Model response to observed flux and wind stress anomalies. Clim. Dyn. 9: 287–302. Mizrock, S. A., D. W. Rice, and J. M. Breiwick. 1984. The blue whale, Balaenoptera musculus. Mar. Fish. Rev. 46: 15–19. Moles, A. and S. D. Rice. 1983. Effects of crude oil and naphthalene on growth, caloric content and fat content of pink salmon juveniles in seawater. Trans. Amer. Fish. Soc. 112: 205–211. Morado, J. F., T. R. Meyers, and R. S. Otto. 2000. Distribution and prevalence of bitter crab syndrome in snow (Chionoecetes opilio) and tanner (C. bairdi) crabs of the Bering Sea, 1988–1996. J. Shellfish Res. 19: 646–647. (Abstract). Murphree, T., S. J. Bograd, F. B. Schwing, and B. Ford. 2003. Large scale atmosphere-ocean anomalies in the northeast Pacific during 2002. Geophys. Res. Lett. 30: 8026. Myers, R. A. and B. Worm. 2003. Rapid worldwide depletion of predatory fish communities. Nature 423: 280–283. Nakamura, H., G. Lin, and T. Yamagata. 1997. Decadal climate variability in the North Pacific during recent decades. Bull. Amer. Meteorol. Soc. 78: 2215–2225. Nitta, T. and S. Yamada. 1989. Recent warming of tropical sea surface temperature and its relationship to the northern hemisphere. J. Japan Meteorol. Soc. 67: 375–383.
Agents of Ecosystem Change
253
NMFS (National Marine Fisheries Service). 2002. Annual report to Congress on the Status of U.S. Fisheries–2001. US Department of Commerce, National Oceanic and Atmospheric Administration, National Marine Fisheries Service, Silver Spring, MD. NRC (National Research Council). 2002. Effects of Trawling and Dredging on Seafloor Habitat. National Academy Press, Washington DC, 126 pp. NRC (National Research Council). 2003. Decline of the Steller sea lion in Alaskan waters: untangling food webs and fishing nets. National Academy Press, Washington DC, 204 p. O’Meara, D. C. and J. F. Witter. 1971. Aspergillosis, In: J. W. Davis et al. (Eds.), Infectious and Parasitic Diseases of Wild Birds. Ames, Iowa, Iowa State University Press, pp. 153–162. Ogut, H. and P. W. Reno. 2004a. Early kinetics of infectious hematopoietic necrosis virus (IHNV) infection in rainbow trout, Oncorhynchus mykiss. J. Aquat. Anim. Health 16: 152–160. Ogut, H. and P. W. Reno. 2004b. Prevalence of furunculosis depends on density of the host (Oncorhynchus tshawytscha) exposed by cohabitation. N. Amer. J. Aquacult. 66: 191–197. Oskam, I. C., E. Ropstad, E. Dahl, E. Lie, A. E Derocher, O. Wiig, S. Larsen, R. Wiger, and S. Skaare. 2003. Organochlorines affect the major androgenic hormone, testosterone, in male polar bears (Ursus maritimus) at Svalbard. J. Toxicol. Environ. Health, A. 66: 2119–2139. Oskam, I. C., E. Ropstad, E. Dahl, E. Lie, A. E Derocher, O. Wiig, S. Larsen, R. Wiger, and S. Skaare. 2004. Organochlorines affect the steroid hormone cortisol in freeranging polar bears (Ursus maritimus) at Svalbard, Norway. J. Toxicol. Environ. Health, A. 67: 959–977. Overland, J. E., J. M. Adams, and N. A. Bond. 1999. Decadal variability of the Aleutian Low and its relation to high-latitude circulation. J. Clim. 12: 1542–1547. Parrish, R. H., F. B. Schwing, and R. Mendelssohn. 2000. Mid-latitude wind stress: The energy source for climatic shifts in the North Pacific Ocean. Fish. Oceanog. 9: 224–238. Pauly, D., V. Christensen, J. Dalsgaard, R. Froese, and F. Torres, Jr. 1998. Fishing down marine food webs. Science 279: 860–863. Pauly, D., V. Christensen, R. Froese, and M. L. Palomares. 2000. Fishing Down Aquatic Food Webs. Am. Sci. 88: 46–51. Perez, M. A. and T. R. Loughlin. 1991. Incidental catch of marine mammals by foreign and joint venture trawl vessels in the U.S. EEZ of the north Pacific, 1973–1988. US Department of Commerce, Seattle. NOAA Technical Report NMFS 104. 57 pp. Pew Oceans Commission. 2003. America’s living oceans: charting a course for sea change. Pew Oceans Commission, Arlington, Virginia, 144 pp. Pianka, E. R. 1974. Evolutionary ecology. Harper and Row, New York. 356 pp. Pipe, R. K. and J. A. Coles. 1995. Environmental contaminants influencing immune function in marine bivalve mollusks. Fish. Shellfish Immunol. 5: 581–596. Pitcher, K. W. 1990. Major decline in number of harbor seals, Phoca vitulina richardsi, on Tugidak Island, Gulf of Alaska. Mar. Mamm. Sci. 6: 121–134. Policansky, D. 1993. Fishing as a cause of evolution in fishes. In: T. K. Stokes, J. M. McGlade, and R. Law (Eds.), The Exploitation of Evolving Resources, Lecture Notes in Bio-Mathematics 99. Springer-Verlag. Berlin, pp. 2–18. Polovina, J. J., G. T. Mitchum, and G. T. Evans. 1995. Decadal and basin-scale variations in mixed layer depth and the impact on biological production in the Central and North Pacific, 1960–88. Deep-Sea Res. 42: 1701–1716.
254
Long-Term Ecological Change in the Northern Gulf of Alaska
Reno, P. W. 1999. Infectious pancreatic necrosis and associated birnaviruses, In: P. T. Woo, and D. Bruno (Eds.), Parasites and Diseases of Fishes, v 3, Bacterial and Viral Pathogens. CABI, London, pp. 1–55. Reznick, D. N., M. J. Butler, F. H. Rodd, and P. Ross. 1996. Life-history evolution in guppies (Poecilia reticulate) 6. Differential mortality as a mechanism for natural selection. Evolution 50: 1651–1660. Rice, D. W. 1974. Whales and whale research in the eastern North Pacific, In: W. E. Schevill, (Ed.), The Whale Problem. Harvard University Press. Cambridge, pp. 170–195. Rice, D. W. 1978. The humpback whale in the North Pacific: distribution, exploitation, and numbers. In: K. S. Norris and R. R. Reeves (Eds.), Report on a Workshop on Problems Related to Humpback Whales (Megaptera novaeangliae) in Hawaii, Mar. Mamm. Comm. Rep. No. MMC-77/03. pp. 29–44. Rice, D. W. and A. A. Wolman 1982. Whale census in the Gulf of Alaska June to August 1980. Rep. Int. Whal. Comm. 32: 491–497. Richardson, L. L. 1998. Coral diseases: what is really known? Trends Ecol. and Conserv. 13: 438–443. Rigby, P. W., D. R. Ackley, F. Funk, H. J. Geiger, G. H. Kruse, and M. C. Murphy. 1995. Management of marine fisheries resources of Alaska: A report to the Northern Forum. Alaska Department of Fish and Game, Commercial Fisheries Management and Development Division, Regional Information Report 5J95-04, Juneau. Rose, G. A. 1993. Cod spawning on a migration highway in the north-west Atlantic. Nature 366: 458–461. Roys, R. S. 1971. Effect of deformation on pink salmon runs in Prince William Sound. In: The Great Alaskan Earthquake of 1964: Biology. NAS Publ. 1604, National Academy of Sciences, Washington, DC, pp. 220–237. Scarff, J. E. 1991. Historic distribution and abundance of the right whale (Eubalaena glacialis). In: The North Pacific, Bering Sea, Sea of Okhotsk and Sea of Japan from the Maury Whale Charts. Rep. Int. Whal. Comm. 41: 467–489. Schneider, N., A. J. Miller, and D. W. Pierce. 2002. Anatomy of North Pacific Decadal Variability. J. Clim. 15: 586–605. Seager, R., Y. Kushnir, N. Naik, M. A. Cane, and J. A. Miller. 2001. Wind-driven shifts in the latitude of the Kuroshio-Oyashio Extension and generation of SST anomalies on decadal timescales. J. Clim. 14: 4249–4265. Sharman, L. C., A. M. Milner, F. S. Chapin III, and D. R. Engstrom. 1995. Qualitative successional models in Glacier bay: A comparison of terrestrial, marine, stream and lake ecosystems, In: D. R. Engstrom, (Ed.), Proceedings of the Third Glacier Bay Science Symposium. National Park Service, Anchorage, Alaska, pp. 190–195. Shirley, S. M. and G. H. Kruse. 1995. Development of the fishery for weathervane scallops, Patinopecten caurinus (Gould, 1850), in Alaska. J. Shellfish Res. 14: 71–78. Simenstad, C. A., J. A. Estes, and K. W. Kenyon. 1978. Aleuts, sea otters, and alternate stable-state communities. Science 200: 403–411. Sindermann, C. 1958. An epizootic in the Gulf of St. Lawrence fishes. Trans. N. A. Wildlife Conf., 23: 349–360. Sindermann, C. 1990. Principal diseases of marine fishes and shellfish, 2nd Ed. v. I, Diseases of Marine Fish; v II, Diseases of marine shellfish. Academic Press, San Diego, CA. Smith, R. L., A. Huyer, and J. Fleischbein. 2001. The coastal ocean off Oregon from 1961 ~os? Prog. Ocean. 49: to 2000: Is there evidence of climate change or only of Los Nin 63–93.
Agents of Ecosystem Change
255
Spies, R. B. and D. W. Rice, Jr. 1988. The effects of organic contaminants on reproduction of starry flounder, Platichthys stellatus (Pallas) in San Francisco Bay. Part II. Reproductive success of fish captured in San Francisco Bay and spawned in the laboratory. Mar. Biol. 98: 191–202. Spies, R. B. and P. Thomas. 1997. Reproductive and endocrine status of mature female kelp bass Paralabrax clathratus from a contaminated site in the Southern California Bight and estrogen receptor binding of DDTs, Chapter 9. In: R. M. Rolland, M. Gilbertson, and R. E. Peterson (Eds.), Chemically-induced Alterations in Functional Development and Reproduction of Fishes, Society of Environmental Contamination and Toxicology, Technical Publication Series, SETAC Press, Pensacola, Fl. Springer, A. M., G. B. Van Vliet, J. F. Piatt, and E. M. Danner. 2006. Whales and whaling in the North Pacific: oceanographic insights and ecosystem impacts. In: J. A. Estes, R. L. Brownell, D. P. DeMaster, D. P. Doak, and T. M. Williams (Eds.), Whales, Whaling, and Ocean Ecosystems, University of California Press, Berkeley, CA. Springer, A. M., J. A. Estes, G. B. van Vliet, T. M. Williams, D. F. Doak, E. M. Danner, K. A. Forney, and B. Pfisterer. 2003. Sequential megafaunal collapse in the North Pacific Ocean: An ongoing legacy of industrial whaling? Proc. Natl. Acad. Sci. 100: 12223–12228. Springman, K. R., G. K. James, J. Anderson, and J. M. Emlen. 2005. Contaminants as viral cofactors: assessing indirect population effects. Aquat. Toxicol. 71: 13–23. Thilenius, J. F. 1990. Woody plant succession on earthquake-uplifted coastal wetlands of the Copper River Delta, Alaska. Forest Ecol. Manag. 33/34: 439–462. Thomas, A. C., P. T. Strub, and P. Brickley. 2003. Anomalous satellite-measured chlorophyll concentrations in the northern California Current in 2001–2002. Geophys. Res. Lett. 30: 8022. Thomas, R. E. and S. D. Rice. 1979. The effect of exposure temperature on oxygen consumption and opercular breathing rates of pink salmon fry exposed to toluene, naphthalene, and water-soluble fractions of Cook Inlet crude oil and No.2 fuel oil. In: W. B. Vernberg, A. Calabrese, F. P. Thurberg, and F. J. Vernberg (Eds.), Marine Pollution: Functional Responses, Academic Press Inc., New York, pp. 39–52. Thompson, D. W. J. and J. M. Wallace. 1998. The Arctic Oscillation signature in the wintertime geopotential height and temperature fields. Geophys. Res. Lett. 25: 1297–1300. Thompson, D. W. J., J. M. Wallace, and G. C. Hegerl. 2000. Annular modes in the extra-tropical circulation. Part II: Trends. J. Clim. 13: 1018–1036. Thompson, W. F. and N. L. Freeman. 1930. History of the Pacific halibut fishery. International Fisheries Commission, Report 5. Vancouver, BC. 61 pp. Thompson, W. F., F. V. Thorsteinson, and D. E. Bevan. 1955. Present status of sea lion investigations in the Alaska Peninsula and Kodiak Island areas. Circular 75 (Peninsula Memorandum 4). Seattle, WA. Thorsteinson, F. V., R. W. Nelson, and D. F. Lall. 1961. Experimental harvest of the Steller sea lion in Alaska waters. US Fish and Wildlife Service, Special Scientific Report–Fisheries 371. Washington, DC Trenberth, K. E. and J. W. Hurrell. 1994. Decadal Atmosphere–Ocean Variations in the Pacific. Clim. Dyn. 9: 303–319. Trippel, E. A. and H. H. Harvey. 1990. Ovarian atresia and sex ratio imbalance in white sucker, Catostomus commersoni. J. Fish Biol. 36: 231–239. Urban, D. and S. C. Byersdorfer. 2002. Bitter crab syndrome in Tanner crab (Chionoecetes bairdi), Alitak Bay, Kodiak, Alaska, 1991–2000. In: A. J. Paul, E. G. Dawe, R. J. Elner,
256
Long-Term Ecological Change in the Northern Gulf of Alaska
G. S. Ameison, G. H. Kruse, R. S. Otto, B. Saint-Marie, T. C. Shirley, and D. Woodby. (Eds.), Crabs in Coldwater Regions: Biology, Management and Economics. University of Alaska Sea Grant AK-SG-02-01. Fairbanks, AK, pp. 401–403. Vander Pol, S. S., P. R. Becker, R. D. Day, M. B. Ellisor, R. S. Pugh, D. G. Roseneau, K. S. Simac and G. K. Chen. 2006. Banking and use of seabird eggs for the Seabird Tissue Archival and Monitoring Project. Poster, Marine Science in Alaska, January 22–25, 2006, Anchorage, AK. Wada, S. 1979. Indices of abundance of large-sized whales in the North Pacific in the 1977 whaling season. Rep. Int. Whal. Comm. 29: 253–264. Wertheimer, A. C. and A. G. Celewycz. 1996. Abundance and growth of juvenile pink salmon in oiled and non-oiled locations of western Prince William Sound after the Exxon Valdez oil spill. In: Rice, S. D., R. B. Spies, D. A. Wolfe, and B. A. Wright (Eds.), Proceedings of the Exxon Valdez Oil Spill Symposium. American Fisheries Society Symposium 18. American Fisheries Society, Bethesda, Maryland, pp. 518–532. Wheeler, P., A. Huyer, and J. Feishbein. 2003. Cold halocline, increased nutrients and higher chlorophyll off Oregon in 2002. Geophys. Res. Lett. 30: 8021. Wigley, S. E. 1999. Effects of first-time spawners on stock-recruitment relationships for two groundfish species. J. Northwest Atlantic Fish. Sci. 25: 214–218. Wobeser, G. A. 1997. Avian cholera, In: Diseases of Wild Waterfowl (2nd Ed.): New York, N. Y., Plenum Press. pp. 57–69. Wobeser, G. F., F. A. Leighton, R. Norman, D. J. Meyers, D. Onderka, M. J. Pybus, J. L. Neufeld, G. A. Fox, and D. J. Alexander. 1993. Newcastle disease in wild water birds in western Canada. 1990. Can. Veter. J. 34: 353–359. Wolf, K. 1988. Fish Viruses and Fish Viral Diseases, Cornell University Press, London. Xie, S. P., T. Kunitani, A. Kubokawa, M. Nonaka, and S. Hosada. 2000. Interdecadal thermocline variability in the North Pacific 1958–97: A GCM simulation. J. Phys. Oceanog. 30: 2798–2813. Yang, M.-S. 1995. Food habits and diet overlap of arrowtooth flounder (Atheresthes stomias) and Pacific halibut (Hippoglossus stenolepis) in the Gulf of Alaska. Proceedings of the International Symposium on North Pacific Flatfish. University of Alaska Sea Grant Program, Report 95-04, Fairbanks, pp. 205–223. Yevich, P. and C. A. Barszcz. 1983. Histopathology as a monitor for marine pollution. Results of histopathological examinations of the animals collected for the 1976 Mussel Watch Program. Cons. Int. Explor. Mer. 182: 96–102. Ylitalo, G. M., C. O. Matkin, J. Buzitis, M. M. Krahn, L. L. Jones, T. Rowles, and J. E. Stein. 2001. Influence of life-history parameters on organochlorine concentrations in free-ranging killer whales (Orcinus orca) from Prince William Sound, AK. Sci. Total Environ. 281: 183–203. Zhang, Y., J. M. Wallace, and D. S. Battisti. 1997. ENSO-like interdecadal variability: 1900–93. J. Clim. 10: 1004–1020.
Mini-Glossary Benthic: on the bottom of the sea, on the sea floor. Contaminant biomarker: a molecular, biochemical or morphological change caused by contaminants.
Agents of Ecosystem Change
257
Corticosteroids: Steroids produced by the cortex of the adrenal glands that are involved in regulation of wide range of physiological processes, including the stress and immune responses, control of inflammation, carbohydrate metabolism and protein breakdown. Cryosphere: The domain of frozen water, snow, and ice. Escapement: the number or proportion of returning anadromous fish that is not caught and enter freshwater for spawning. Koch’s postulates: The four necessary criteria to show a connection between a particular pathogen and a disease it causes: (1) The pathogen must be found in all animals suffering from the disease; (2) The pathogen must be isolated from the diseased animal and grown in culture; (3) The isolated pathogen should cause the disease when introduced into a healthy animal; and (4) The pathogen must be re-isolated from the experimentally infected animal. Mitogens: Substances that cause a cell to start dividing. Monoaromatic: hydrocarbons with one aromatic ring, that is, a 6-member carbon ring with 3 double bonds in it. Poikilotherms: Animals that have a variable body temperature that fluctuates with changes in environmental temperature, as opposed to homeotherms, which maintain a relatively constant body temperature. Piscivorous: fish eating. Redd: a shallow depression in stream gravel where salmon lay eggs. Seismic gap: an area along fault or fault system that has recently had relatively little significant seismic activity Seropositive: Finding a host-produced antigen for a disease agent in the serum of the host. Subduction: movement of one crustal plate beneath another
This page intentionally left blank
Chapter 4
Long-Term Change 4.1. Introduction Robert B. Spies and Thomas Weingartner After surviving a harsh winter of long nights, cold, and severe storms, organisms in the Gulf of Alaska (GOA) would have depleted their energy reserves as they enter the growing season and, for most, the reproductive season. Yet, they are a part of an ecosystem poised to capture and distribute the increasing solar energy warming the upper layers of the subarctic ocean. Every year is different in the Gulf of Alaska, and the seeds of longer-term variability lie in how each year differs from the last in climate, in other natural factors, and man’s effects. To be sure, the major ecological phenomena each year are similar: late winter stratification of the upper layers of the ocean; a spring phytoplankton bloom in April–May followed by a zooplankton bloom; and predation on the bloom by larger zooplankton and forage fish, which in turn are consumed by seabirds and marine mammals. Each year is a fresh start and change, irregularly expressed over the periods of years, comes from the press of climate, disease, contaminants, predators, and competitors on populations of organisms with unique adaptations. Foraging strategies, energy storage, reproductive strategies, size, and life span provide an array of “choices” made by each species in the course of its evolution that involves multiple trade-offs. It is the play of a changing ecosystem against these “choices” minus the effects of man that ultimately determine the success and, therefore, the population trajectory and long-term change in a species. That ecological change occurs over different timescales is a truism of ecology. In this book, we devote considerable resources to changes that occur over several years to decades rather than longer periods simply because this is where we have the most data. Even on this scale, we must recognize that ecological data in the Gulf as with most other regions is particularly sparse pre-1975. The first barrier to understanding change is that we do not have the long-term data sets, and those that we have seldom encompass the scales and include the variables that are required to quantify changes, much less understand their causes. For example, the instrumental record of physical data spans barely more than a century in the North Pacific Ocean. The longest term physical data records consist of easily measured variables, such as sea surface temperature, salinity, air pressure, temperature, and sea level. Long-Term Ecological Change in the Northern Gulf of Alaska Robert B. Spies (Editor) © 2007 Elsevier B.V. All rights reserved.
260
Long-Term Ecological Change in the Northern Gulf of Alaska
Long-term measures of other physically important variables i.e., salinity, precipitation, nutrients, stratification, currents, and coastal wind fields) are however relatively rare. For the biological data needed to track and understand long-term change, the situation is even worse. Typically, our records consist only of catch data of commercially important species or, more rarely, population estimates, which are sometimes needed by managers. Noncommercial populations are less well studied, particularly pre-1970. The uneven treatment of biological populations in this book is to a large extent the result of the lack of data for many populations. For example, virtually nothing is known about long-term changes in two of the most important components of the ecosystem, phytoplankton and benthic communities. However, with regard to climate, the paleoecological record is replete with evidence of changes in ecosystems coincident with climate variations. Indeed, when viewed from the perspective of geological time, climate, and ecosystem changes appear to be the rule rather than the exception. Nevertheless, climate change, whether of natural or anthropogenic origin, brings with it uncertainty and the potential to upset existing economic and social structures. It is precisely this uncertainty that provides the impetus for current research efforts directed at understanding how marine ecosystems function and respond to climate variability. The second barrier to understanding long-term changes in the Gulf of Alaska is the complex nonlinear structures that characterize these systems. Perturbations to these systems often yield unexpected behaviors that can affect not only the mean state of the system but, perhaps more importantly to organisms, the variance of the system (see Box 4.1). Examples of this abound in both the paleoclimate record and model predictions of the evolution of present-day climate when perturbed by the increases in greenhouse gases. Among the many predicted effects are increases in the frequency and magnitude of extreme or episodic events and alterations in the annual cycle of precipitation and runoff. Conceivably these changes have a much greater influence on ecosystem processes than mean conditions might suggest. In addition, climate responses might not be uniform over the globe but may instead be accelerated (or delayed) or amplified (or damped) in one region relative to another. Another potential example is the effect that whaling may have had on gulf megafuana for decades (Springer et al., 2003). Given these difficulties, scientists have often tried to establish a statistical correlation between physical and biological variables. Although occasionally successful, this approach has potential pitfalls. In particular, the space and timescales of physical and biological variables are different from one another and the relation between a physical change and a biological response might be indirect, nonlinear, and involve several other (unmeasured and also varying) system components. An additional problem is that the limited observations do not allow unequivocal conclusions to be made with respect to physically induced causes of biological change, especially when considering decadal (10–50 years) and longer timescales. In such cases, there are too few representations of the covariability to permit statistically discriminating real causal
Long-Term Change 261
BOX 4.1: NONLINEAR AND CHAOTIC RELATIONSHIPS by Robert B. Spies The relationships between organisms or between organisms and their environment is key to understanding or even trying to predict long-term change. Ecological relationships are often described as nonlinear, although linear relationships can also occur. In mathematics, a nonlinear relationship or function is where a change in one variable (m) elicits a response in a second variable (n), but the response instead of being linear (Fig. 4.1.A.2) is not proportional throughout its range (Fig. 4.1.A.3, and Fig. 4.1.A.4), taking a variety of forms (sigmoid, asymptotic, etc.). In real ecosystems even “nonlinearity” often fails to capture the complexity of relationships. A more realistic abstraction of the relationship between two variables might look like Fig. 4.1.B.1. Here, a relationship between m and n is affected by many more factors acting on n. If the other factors hold steady, then there might be a defined relationship, e.g., as in Fig. 4.1.B.2. However, instead of a nonlinear relationship between m and n, there is more likely a chaotic relationship (Fig. 4.1.B.3). Chaotic does not mean that m and n are not related, but that only the relationship is affected by other factors and cannot be predicted based solely on interactions between these two variables. The nature of real ecosystems is such that doing “experiments” where the relationship between two variables, e.g., oil and population size, is studied by contrasting settings where one of the factors varies is so challenging. Unmeasured variables and their fluctuations and interactions tend to move the relationships towards chaotic ones. sigmoid
linear
A m
n
m
m n
1
m n
2 sigmoid
asymptotic
n
3 chaotic
4
B m
m m
n n 1
2
n 3
Figure 4.1: Some simple types of relationships between ecosystem variables (m,n): A. When m and n vary independent of other ecosystem variables – (1) the dependent relationship, (2) linear, (3) nonlinear, sigmoid, and (4) nonlinear, asymmpotic. B. When one variable (n) also varies due to other factors: (1) multiple factors affecting n, (2) nonlinear, sigmoid, and (3) chaotic.
262
Long-Term Ecological Change in the Northern Gulf of Alaska
relationships from apparent correlations, trends, or “state” shifts that can emerge between two random time series (e.g., Wunsch, 1999). Indeed, long-period climate fluctuations represent the “gray area of climate variability” (Karl, 1988), which are poorly described by short instrumental records. In addition, factors that correlate under some state conditions may not correlate in another state when the controlling factors may shift (see Box 4.2). There are, however, alternative approaches to the constraints imposed by the short observational record. One technique is to examine proxy data, such as obtained from paleoenvironmental studies. In some cases, these provide a reasonable extension of the instrumental record for many hundreds of years into the past. An alternate approach is to use numerical models based upon realistic approximations to the physics and biology. The observations are then used in conjunction with models to evaluate model performance and/or to constrain model predictions and hindcasts. While both approaches
BOX 4.2: CORRELATION OR CAUSE AND EFFECT? by Robert B. Spies The history of fisheries science abounds in correlations between various environmental factors and fish populations. Almost inevitably the correlations work for a while but then dissolve, leaving scientists to conclude that the link was either not causative or that there are new factors that have become the determinants of the population trajectory. There are several examples in the Gulf of Alaska where there have been observations supporting hypothesized causes of ecological changes that appeared to have changed as the ecosystem entered a new phase or state. We present two here: 1. Upwelling and pink salmon production There was a relationship between strength of the Bakkun upwelling index reported near the Hinchinbrook Entrance to Prince William Sound (a proxy measure for onshore flow) and the springtime standing stock of zooplankton measured by PWSAC at the AFK hatchery each spring (April–May average) as settled volume of zooplankton (Fig. 4.2). This relationship held in the 1980s but “dissolved” in the 1990s, despite the fact that the range in upwelling indices for April–May did not change in the 1990s. What did change was the zooplankton – only about half the volume in comparable samples from 1991–1998. The number of pink salmon adults coming back to spawn was in turn correlated with the AFK zooplankton (r2 > 0.5) (Cooney et al., 2001). 2. A relationship between young-of-the-year Pacific herring abundance and kittiwake productivity There was a strong relationship in the 1990s (kittiwake reproductive success (lagged 2 year) vs. log of age-3 herring abundance. That is, if age-3 Pacific
0
5
−5
4
−10
3 −15 2 −20
1
BAKUN UPWELLING INDEX
6
−25
0 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998
ZOOPLANKTON (settled volume, mL/m3)
Long-Term Change 263
YEAR
Figure 4.2: Relationship between upwelling index and zooplankton.
9
0.6
herring
AGE 3 HERRING BIOMASS (tonnes, Log10)
8 0.5 7 0.4
6 5
0.3
kittiwake
4
0.2
3 2
0.1 1 0 2001
2000
1999
1998
1997
1996
1995
1994
1993
1992
1991
1990
1989
1988
1987
1986
0
KITTIWAKE REPRODUCTIVE SUCCESS (chicks/nest)
herring abundance (the earliest indication of the strength of the age class) is lagged 3 years to account for the strength of the age class at the time it was recruited, black legged kittiwakes, which feed on age-0 herring, have reproductive success that correlates with how many young herring are in the shallow waters in which they feed. The comparable data from the 1980s did not show this relationship (see Fig. 4.3) (Suryan et al., 2000).
YEAR
Figure 4.3: Relationship between kittiwake productivity and juvenile herring abundance.
264
Long-Term Ecological Change in the Northern Gulf of Alaska
have limitations, they are becoming increasingly effective in quantifying long-term variability over a range of space and timescales and in deciphering mechanistic relationships between climate change and ecosystem response. Thus, while marine scientists are still at an early stage in explaining ecosystem variability, substantial progress has been made in the last two decades in understanding and quantifying this variability. Two principal reasons account for this progress. First, new technologies are allowing a more comprehensive and systematic collection of observations. Moreover, technological developments have allowed marine science to undertake interdisciplinary programs that are yielding new insights into marine ecosystem structure and function. Second, advances in computational resources and a more sophisticated understanding of atmosphere–ocean dynamics are improving physical model reliability. Continued efforts along these lines suggest that our understanding of climate–ecosystem variations will accelerate even more rapidly in the near future. Still, these efforts are much better at explaining what happened in the past, hindcasting, than in forecasting. For example, there are mathematical models of the Gulf of Alaska that have been constructed based on how freshwater inputs influence buoyancy, the way that wind moves water, and how currents are generated by differences in water properties over distance (Wu and Hseih, 1999; Herman et al., 2002; Murray et al., 2001; Wang et al., 2001). These models draw conclusions and make predictions of system behavior that are largely deductive. There are also some deductive models that predict behavior of the biological components of the system, e.g., models exist that predict the intensity and duration of plankton blooms in Prince William Sound, based on water properties at various depths and wind strength (Eslinger et al., 2001). The forcing factors are well understood and strong for high-latitude plankton blooms, but subtler interactions of many forces may be at work in plankton populations at other times of the years and at higher trophic levels. So, linking forcing factors (e.g., fishing, changes of the winds or cloud cover, and contamination) with outcomes, i.e., changes in populations, presents significant challenges, especially for species at higher trophic levels. Therefore, the value of deductive logic and deterministic models is limited when dealing with the manifold interacting processes that determine the trajectories of fish, bird, and mammal populations. Induction can contribute explanations that are consistent with the observations, but are difficult to establish with a high degree of rigor that is the standard for experimental science. The challenges of establishing cause and effect lie at the heart of understanding our impact on complex marine ecosystems. Resource users, e.g., commercial fishers and waste dischargers, can and usually do expect a high level of proof when their activities are regulated. That is, they expect challenges to their privileges to use the marine environment, based on perceived negative consequences of their activities, to be accompanied by rigorous demonstrations of cause and effect. Government resource agencies responsible for the maintenance of healthy populations and ecosystems, are,
Long-Term Change 265
by contrast, usually much more conservative in protecting the ecosystem in the face of uncertainty and prefer an approach that considers the weight of evidence, the precautionary approach. The tension revolving around cause and effect is crucial background for understanding the conflicting claims about the effects of the Exxon Valdez oil spill (Chapter 5) and other resource management issues in the Gulf of Alaska. Change is a fundamental ecological reality in the Gulf of Alaska that is neither constant in its rate nor predictable with our present understanding. Here, in Chapter 4 we examine ecosystem changes longer than 1 year. In Chapter 2, a foundation for longer-term changes was laid by exploring strongly interrelated seasonal physical, chemical, and biological cycles and survival strategies for some major vertebrate species. Chapter 3 described root forcing mechanisms of long-term change, both natural and anthropogenic. This leads us to consider here what the long-term changes have been in the Gulf of Alaska and what may have caused them. In the first part of Chapter 4 records of long-term physical, chemical, and plankton fluctuations are presented, and the potential causes for change are discussed. This is followed by a series of case histories of the species first introduced in Chapter 2. The possible causes for their population changes are then discussed. This section provides the foundation for Chapter 5, which synthesizes the studies on the Exxon Valdez oil spill, and Chapter 6 that discusses long-term ecosystem change holistically.
4.2. Atmosphere and Ocean Thomas Weingartner 4.2.1. Introduction This section outlines some of the substantial variability observed over recent decades in the physical environment of the Gulf of Alaska. Variations have been detected in winds, surface heat fluxes and runoff, water temperature and salinity, mixed-layer depth, nutrient supply, and circulation properties. Most of the data forming the basis of this presentation have been collected on the northwest Gulf of Alaska shelf and along “Line P,” which extends westward from the British Columbian shelf to 50°N, 145°W in the southern Gulf of Alaska. Line P, which has been sampled by Canadian oceanographers since the mid-1950s, spans the bifurcation zone of the North Pacific Current as it approaches the west coast of North America (see Section 2.2). The physics of the bifurcation region and the shelf differ considerably from each another, however, so the spatial extent over which observed changes along Line P are representative of changes in the central basin is not readily apparent. We, therefore, discuss each region separately. Nevertheless, there are a number of consistent trends observed in both
266
Long-Term Ecological Change in the Northern Gulf of Alaska
areas, suggesting that these changes might be spatially extensive, although the magnitude of the changes might vary in different portions of the shelf and basin.
4.2.2. Gulf of Alaska Shelf Winds Stabeno et al. (2004) analyzed interannual variability in shelf winds, produced by the National Oceanic and Atmospheric Administration–National Center for Environmental Prediction (NOAA–NCEP) weather forecast models. Although the modeled winds are likely biased because of the influence of the coastal mountain ranges (Section 2.2), they are the primary source of long-term wind variations for this shelf. Stabeno et al. (2004) find that most of the variability is on the interannual timescale (rather than the interdecadal or intradecadal scales discussed in Section 3.2). They do suggest, however, that both anomalously low winter wind speeds (important for vertical mixing) and along-shore wind stress (important for coastal downwelling and along-shore shelf transport) prevailed from the mid-1960s to mid-1970s and above average values from the late 1970s through the 1980s. Although these differences appear to coincide with the mid-1970s regime shift, there is no significant correlation between wind variations in the northern Gulf and the Pacific Decadal Oscillation (PDO) index (Royer, 2005). They did find a weak (but significant) correlation between eastern Gulf winds and the PDO, however, with a positive PDO index (warm-phase) associated with stronger than normal downwelling-favorable wind stress. Stronger downwelling in the eastern Gulf would tend to strengthen flow in the Alaska Coastal Current (ACC) in the northern Gulf, because the coastal flow at a particular location responds to the integrated response to both the local and the upstream along-shelf wind field. Stabeno et al. (2004) also found substantial year-to-year variability in the frequency and strength of summer wind-mixing events and upwelling-favorable winds, with these variations being largest over the shelves in the southwest and southeast Gulf of Alaska. This is a potentially important finding because upwelling or wind-mixing events entrain nutrient-rich subsurface waters into the nutrient-depleted surface layer. Upwelling events are intermittent so year-to-year variations in event frequency, duration, and strength would likely yield similar variations in the total summer primary production.
Variability in Surface Heating and Cooling and Shelf Temperatures The temperature of the ocean at a given location is controlled by heat exchange with the atmosphere, mixing with waters of different temperatures, and the advection of heat into or out of a region by ocean currents. In most regions of the ocean, including the Gulf of Alaska shelf, air–sea heat exchange exerts the primary control over upper
Long-Term Change 267
ocean temperatures. Changes in mixing and advection are difficult to assess in the absence of routine measurements of ocean currents and temperatures at several locations, but surface heat flux estimates can be estimated from routine meteorological measurements made over the ocean or from weather forecast models. The interannual variability in the net seasonal air sea heat fluxes over the northern Gulf of Alaska shelf from 1950 through 2000 is shown in Fig. 4.4. The figure shows the cumulative net heat exchange for fall–winter (late September through late March) when cooling occurs and for summer (April through September) when heating occurs based on NOAA–NCEP weather forecast models. These values are shown for each year, as an 11-year running mean, and for the 50-year average. The northern Gulf of Alaska generally loses more heat in winter than it gains in summer (Section 2.2), although there are some years when heat loss and gain nearly balance each other and one year (2000) when the ocean gained more heat in summer than it lost in winter. There is substantial interannual variability in these fluxes, with the variations being much larger in winter than in summer. Heat flux variations in summer are less than winter because summer heat fluxes are primarily due to solar radiation (Section 2.2). Variability in radiative heating is primarily controlled by cloud cover, which, in general, varies little from summer to summer. In winter, the latent and sensible heat fluxes dominate air–sea heat exchange, and changes in these heat fluxes depend
Figure 4.4: Annual cumulative heat loss in winter (red) and heat gain in summer (blue) from 1950 to 2001. The horizontal blue line is the mean for all summers, and the horizontal red line is the mean for all winters. The thin colored lines are the annual values, and the thicker colored lines are the smoothed versions of the annual values to highlight decadal variability.
268
Long-Term Ecological Change in the Northern Gulf of Alaska
on wind speed and the air–sea temperature difference. Consequently, winter heat loss primarily determines upper ocean water temperatures in spring and is likely to affect the metabolic rates of many marine organisms, including zooplankton and immature fish at a critical phase in their life history. For illustrative purposes, the winter cooling rates in Fig. 4.4 can be expressed in terms of a corresponding temperature change over the upper 100 m of the shelf water column during the cooling season. On average, this results in a temperature change of −4.4°C, with the range being −6.5°C (1970) to −2.5°C (2000). For summer, the mean temperature increase is 3.8oC, with the range being 4.2°C (1958) and 2.8°C (1998). Winter cooling rates decreased substantially (~20%) in the mid-1970s coincident with the shift of the PDO from the cold to the warm phase and the intensification of the Aleutian Low. Prior to the regime shift winter cooling rates would have induced a 5°C decrease over the upper 100 m of the ocean versus a 3.8°C decrease since then. These changes are consistent with a warming of nearly 1.0°C in the upper 100 m of the water column since 1970 based on temperature observations made at hydrographic station GAK 1 near Seward (Royer, 2005). He found a temperature increase of similar magnitude between 100–200 m over this same period. Hence, the warming over the whole water column exceeds that due to air–sea heat exchange alone and suggests that oceanic transport processes have also contributed to the warming trend. The most likely source of this oceanic warming is the alongshore transport of heat from southerly latitudes by the ACC on the shelf and by the Alaska Current along the continental slope. Indeed this is consistent with the findings of Freeland and Whitney (2000) of a similar increase in upper ocean temperatures over the British Columbian shelf during this period. The warming trend, along with suggestions of increased along-shelf transport due to the winds and changes in runoff (discussed next) suggest that there has been an increase in along-shore mass transport over the shelf. Royer’s (2005) analysis of the GAK 1 record also shows that there are substantial ENSO-induced temperature variations on the northern Gulf of Alaska shelf. Observed ENSO-induced temperature changes at GAK 1 range between 0.5 and 1.7°C and occur in winter some 7–10 months after the equatorial onset of ENSO. However, ENSO-associated temperature perturbations are generally short-lived on the Gulf of Alaska shelf, usually appearing in early winter and disappearing by late spring. Although brief in duration, the timing of these perturbations could be of biological significance because they occur in spring when during the early life history of many invertebrates and fish. Interestingly, Royer (2005) finds that the ENSO response is statistically significant at depths between 50 and 150 m, but not at shallower depths. Conceivably the lack of a significant relationship between the equatorial ENSO signal and upper ocean temperatures on the Gulf of Alaska shelf occurs because random local variability in air–sea heat fluxes blur the ENSO signal in the upper ocean. The presence of the ENSO-temperature signal at depth does suggest that ocean advection is important, however. Indeed, Royer (2005) argues that ENSO warming in
Long-Term Change 269
the Gulf of Alaska is primarily associated with oceanic processes that displace the large-scale coastal temperature gradient northward into the Gulf of Alaska. This interpretation is consistent with observations made during the 1997–1998 ENSO event, which showed anomalously large increases in the flow along the Alaskan continental slope (Strub and James, 2002), within the ACC (Weingartner et al., 2005), and in the northward flux of heat along the British Columbian continental slope (Freeland, 2002). Royer (2005) also finds that the PDO (positive or warm phase is associated with above normal temperatures) is significantly and uniformly correlated with monthly water temperature anomalies over the entire shelf water column, although the PDO signal accounts for only about 25% of the temperature variability. Nevertheless, his result indicates that the PDO is a useful indicator of long-period, deep temperature variations for the northern shelf. Given the uniform response of the shelf water column to the PDO, it seems probable that both the atmosphere and the ocean influence the thermal structure on the shelf. Since the PDO index reflects winter, sea-surface temperature variations over the winter mixed layer, which is typically 150 m throughout the northern North Pacific, these anomalies could be advected onto the shelf by ocean circulation processes. Air–sea heat exchange is also important however, because winter cooling tends to decrease during the warm phase of the PDO and increase during the cold phase.
Variability in Runoff and Shelf Salinity As discussed in Section 2.2, runoff onto the Gulf of Alaska shelf affects salinity, which in turn critically controls both horizontal and vertical ocean density gradients. The long-term runoff record for the Gulf of Alaska based on Royer’s (1982, 2005) simplified hydrology model captures monthly anomalies in coastal freshwater discharge since 1931 (Fig. 4.5). The monthly anomalies can be enormous; several times greater (or smaller) than the mean monthly value. Indeed, anomaly magnitudes can be as large as 25,000 m3 s−1, which is comparable to the mean annual runoff. Although the monthly and interannual variability dominates the signal, there is considerable longer-period variability in this record as well (Fig. 4.6) where the blue curve is a smoothed (3-year low-pass filter) version of Fig. 4.5 along with an alternative index of runoff variability (red) discussed below. (Note for both curves the anomalies are normalized by their standard deviation.) Abnormally large runoff occurred in the 1940s and from the mid-1980s to the mid-1990s and abnormally low runoff occurred in the 1960s–1970s. The red line is a proxy time series for runoff based on the significant correlation between runoff and the atmospheric sea level pressure difference between Seward and Ketchikan (Weingartner et al., 2005). Positive values of this difference are believed to represent atmospheric transport of warm, moist air into the northern Gulf of Alaska, while negative values reflect outbreaks of cold, dry continental air.
270
Long-Term Ecological Change in the Northern Gulf of Alaska
Figure 4.5: Monthly anomalies of coastal freshwater discharge into the Gulf of Alaska since 1931 [adapted from Royer (1982, 2005)].
Because these two time series are significantly correlated, the sea-level pressure record serves as a surrogate for runoff and can be extended back to the beginning of the twentieth century (although pressure estimates prior to 1910 are less certain than those afterwards). Monthly anomalies of ACC salinity and the transport component forced by crossshore density gradient are correlated with runoff from November through May (Weingartner et al., 2005). Hence the time series in Fig. 4.6 suggest how these ACC parameters have varied at interdecadal timescales. Accordingly, the first decade of
Figure 4.6: Long-term freshwater runoff anomalies in the Gulf of Alaska. The blue curve is based on the model of Royer (1982), and the red curve is based on the atmospheric sea-level pressure difference between Seward and Ketchikan. Adapted from Weingartner et al. (2005).
Long-Term Change 271
the twentieth century and the period from 1960 to mid-1975 were more saline and had weaker ACC transport than the period from 1920 through 1945 and the 1980s. The GAK 1 record indicates a salinity decrease of about.07 in the upper 100 m since 1970, which is consistent with the increase in runoff over the same time period (Royer, 2005). The implied increase in ACC transport from 1970 through the mid-1990s is also consistent with the previous discussion that the warming trend observed at GAK 1 since 1970 must have involved an increase in along-shelf transport of heat into the northern Gulf of Alaska. Although direct measurements of mixed-layer depth variations are not available for the northern Gulf of Alaska shelf, the warming and freshening trend in the upper ocean reported by Royer (2005) also imply that winter shelf stratification is strengthening. This tendency appears to be Gulfwide as similar trends have been reported for the British Columbian shelf and over the southern portion of the basin by Freeland and Whitney (2000). Although changes in runoff, upper ocean salinities, and transport of the ACC seem consistent with the “regime shift” change in the PDO from the cold to the warm phase in the mid-1970s, these variables are only weakly correlated with the PDO index (Weingartner et al., 2005; Royer, 2005) as is the correlation between ENSO and salinity (Royer, 2005). It is not entirely clear why this is the case, although Dettinger et al. (2001) find that, while precipitation and river discharge increase during the PDO warm phase and El-Niños over south central Alaska, these variables tend to decrease over the Pacific Northwest and British Columbia. Since the salinity and ACC transport in the northern Gulf are a consequence of runoff along the entire coast, the spatially out-of-phase patterns in runoff between the northern Gulf of Alaska and the Pacific Northwest would tend to degrade the correlation between salinity and the PDO and ENSO climate indices. There are occasions, however, when ENSO-related salinity affects can be substantial. The best-documented example of this in the northern Gulf of Alaska is the comparison between the El Ni˜no winter of 1997–1998 and the La Ni˜no winter of 1998–1999 (Weingartner et al., 2002, 2005). March and April shelf temperatures over the upper 100 m of the water column were nearly 2.0°C greater and salinities were about 0.3 lower in 1998 than for the comparable period in 1999. These changes were consistent with a 30% reduction in atmosphere–ocean heat loss during the winter of 1997–1998 compared to 1998–1999 and anomalously large late summer–early fall discharges in the Fraser River (British Columbia) and Columbia River (Oregon) and winter freshwater discharge into the Gulf of Alaska. The effects of the increased runoff in 1997–1998 were further enhanced by unusually strong downwelling-favorable wind stress in the northeastern Gulf of Alaska. In aggregate these anomalies resulted in the density-driven component of ACC transport in 1998 being double that of 1999. The increased transport and the above normal temperatures over the British Columbian shelf in 1997–1998 imply that oceanic heat advection was also greater in 1997–1998 than in 1998–1999 and contributed to the ocean temperature differences between
272
Long-Term Ecological Change in the Northern Gulf of Alaska
these two winters. However, the contribution of ocean heating to the observed warming has not been evaluated yet. The runoff differences between these years were also accompanied by an earlier onset of stratification in spring (Weingartner et al., 2005) and lower nitrate concentrations over the upper ocean (Childers et al., 2005) in 1998 compared to 1999. The large coastal freshwater influx was the primary cause of the early onset of stratification in 1998, which occurred nearly a month earlier than in 1999. The low surface nutrient concentrations are consistent with the increased runoff, although larger-scale affects might also have been operating since surface nitrate concentrations were also diminished over the northern North Pacific (Goes et al., 2001) and the southern Gulf of Alaska (Freeland and Whitney, 2000). The difference in the timing of early spring stratification between these two years provides a hint of the complexities in the development of stratification and the spring bloom on the inner portion of the Gulf of Alaska shelf. Springtime phytoplankton blooms can begin in the sheltered waters of Prince William Sound in late March or April and shortly thereafter in the ACC. In contrast, the spring bloom on the outer shelf develops about mid-May. As discussed by Weingartner et al. (2005), stratification does not develop uniformly in space or time over the Gulf of Alaska shelf because stratification mechanisms on the outer shelf are different from those within the ACC (and Prince William Sound). Seaward of the ACC, upper-ocean warming is primarily responsible for initiating stratification in spring and this is governed by wind-mixing and solar heating. In contrast, stratification inshore depends upon three-dimensional circulation and mixing processes associated with freshwater dispersal and winds. Many variables contribute to inner-shelf stratification including the fraction of winter precipitation delivered as snow and rain, the timing and rate of spring snowmelt, and the wind velocity. The relevant timescales range from a few days (storm events) to the seasonal or longer, with the longer scales associated with advection of freshwater by the ACC from distant upstream regions. Given the number of parameters involved, large interannual variations in the onset of stratification on the inner shelf are expected. This also implies that the application of Gargett’s (1997) optimal stability window hypothesis to the ACC is far more complicated than in regions where vertical heat fluxes primarily control upper-ocean stability. The stratification differences between 1998 and 1999 also suggest how climate warming might affect the Gulf of Alaska ecosystem. Projections of future climate response to increased greenhouse gas concentrations (IPCC, 2001) indicate an increase in atmospheric warming and moisture over much of the globe, including the Gulf of Alaska. This implies greater winter rainfall and runoff but less snow accumulation in the coastal mountain ranges. Consequently, an earlier onset in spring melt is to be expected. If wind mixing does not increase proportionately, then stratification and the spring bloom may occur earlier on the inner shelf than it does presently. Conceivably, these changes might lengthen the time lag between bloom development
Long-Term Change 273
on the inner shelf and the outer shelf, which could affect the community structure and recruitment success of zooplankton and fish over these different portions of the shelf. The IPCC (2001) projections also indicate larger year-to-year variations in precipitation so that interannual variability in the timing of the spring bloom on the inner shelf might also increase.
4.2.3. Gulf of Alaska Basin Based upon data collected from 1968 to 1990, Lagerloef (1995) and Hunt (1996) concluded that the Alaska gyre underwent a large transition coincident with the regime shift when the PDO switched from the cold phase to the warm phase. They show that during the cold phase (prior to the mid-1970s), the center of the Alaska gyre was shifted northeastward, the gyre circulation was stronger, and cooler sea surface temperatures prevailed over the central Gulf compared to after the transition. The transition to the warm-phase PDO mode led to a west-southwest displacement of the gyre center and a reduction in upwelling and gyre transport in the central and eastern gulf. There are some indications that on ENSO timescales the Alaska gyre and the California Current vary out-of-phase (Chelton and Davis, 1982; Tabata, 1991; Kelly et al., 1993) such that more water from the North Pacific Current enters the Gulf of Alaska when the gyre strengthens, while more of it is deflected southward when the gyre weakens. Under such conditions, Van Scoy and Druffel (1993) find that subpolar waters carried by the North Pacific Current leak into the California Current. As previously noted, Polovina et al. (1995) concluded that the regime shift was accompanied by changes in mixed-layer depth. Freeland et al. (1997), Freeland and Whitney (2000), and Whitney and Freeland (1997) expanded on Polovina et al.’s work by closely examining changes in mixed-layer depth and properties along Line P. They find that the winter mixed-layer depth has decreased since 1956 at a rate of about 47 m/century and suggested that there was a possible step-change to a decrease in winter mixed-layer depth coincident with the regime shift. However, they found that the shallowing of the winter mixed-layer was related to an increase in upper ocean stratification brought about by both a freshening and warming of the surface layers. Superposed on this long-term trend, are ENSO-related variations that corroborate the model results of Alexander et al. (2002). These workers find that shallower, winter mixed-layers are associated with El Niños, while deeper mixed layers occur during La Niñas. The shoaling mixed-layer observed along Line P decreases the winter re-supply of nitrate and silicate to the euphotic zone (Whitney et al., 1998; Whitney and Freeland, 1999). Although the decrease does not appear to affect the magnitude of the spring bloom presently, it increases the likelihood of nutrient exhaustion through summer. Indeed, this decrease was so severe during the 1998 El Niño event that it led to the first ever report of nitrate depletion in the surface waters if the Gulf of Alaska basin.
274
Long-Term Ecological Change in the Northern Gulf of Alaska
4.3. Zooplankton Theodore Cooney Records from the Canadian Weather Ship program at Ocean Station P(OSP) (1956–1980; Fulton, 1983) located at the southern margin of the Gulf of Alaska provide a starting point for documenting and analyzing long-term patterns of change in zooplankton stocks in the Gulf of Alaska. Frost (1983) analyzed these records (approximately 2500 samples; vertical tows in the upper 150 m) and discovered a weak, positive statistical relationship between the annual average (February to August) standing stock of zooplankton and surface water salinity. A similar relationship had been reported previously by Wickett (1967), who hypothesized that increased salinity should accompany nutrient upwelling, higher primary productivity, and greater zooplankton growth. Frost was able to examine that possibility by looking for coherence between zooplankton growth (extracted from a portion of the OSP time series) and estimates of primary production provided by Parslow (1981). Oddly, Frost found no relationship between these data sets, suggesting something other than bottom-up forcing of the production cycle was responsible for driving the interannual patterns in zooplankton abundance recorded at OSP. Because there were no comparable annual observations for other locations in the Gulf of Alaska, there is no way to determine just how representative the OSP observations were of the broader Gulf between 1956 and 1980. Brodeur and Ware (1992) used data collected from the Joint U.S.–Canadian High Seas survey, 1956–1962, and from later Japanese surveys (R/V Oshoro maru of Hokkaido University; 1981–1986, 1988–1990) of the Gulf of Alaska to reconstruct the spatial distribution and stock size of net zooplankton caught in vertical tows from 150 m to the surface during these two different decades. The results of this analysis are dramatic; zooplankton was distributed more broadly and was more than twice as abundant in the 1980s. Also in the 1980s, the biomass of zooplankton was found to be greatest around the “edge” of the Gulf – a kind of bathtub ring – rather than in the interior of the region as was the case in the 1950s and early 60s (Fig. 4.7). The authors suggested two reasons for these patterns: (1) increased center gyre wind-forced upwelling (demonstrated for the 1980s) generated more nutrients (including iron) from deep water and thus increased the overall productivity of plankton and (2) predation on primary producers by zooplankton was enhanced by deeper vertical mixing during the 1980s when the Aleutian Low was stronger, enabling greater production of grazers (Parsons, 1988). In addition, Miller et al. (1992) reported that the size of individual species tended to be smaller in the interior of the oceanic realm rather than at the edges of the system. Neocalanus plumchrus and N. flemingeri were often 25–80% less massive at OSP than when encountered along the shelf edge, presumably a response to enhanced feeding opportunities nearer the continental margin. Brodeur and Ware (1992) also looked at the OSP zooplankton time series and plotted the average biomass between June 15 and July 31 each year from 1957 to 1980;
Long-Term Change 275
Figure 4.7: Comparison of zooplankton biomass in the Gulf of Alaska from two periods. After Brodeur and Ware (1992).
they found a statistically significant linear increase (approximately a factor of 2.0), although there was also an apparent cyclic progression to the observations as well (see Fig. 4.5; Brodeur and Ware, 1992). Zooplankton biomass was lowest in the early 1960s and 1970s, and higher in the early 1950s and late 1970s. There was no explanation given for these apparent cycles. In a parallel analysis of 14 selected fish and mollusk species (Fig. 4.8) (salmonids, nonsalmonid fishes, and cephalopods) from the Northeast Pacific Ocean, Brodeur and Ware (1995) were able to demonstrate that catches of all but one species were greater in the decade of the 1980s than in the mid1950s and early 1960s. Salmonid biomass increased by a factor of 2.0 in concert with the doubling of the zooplankton stock. The authors suggested that the carrying capacity of the region can vary from decade to decade or over longer periods, and warn that current high levels of salmon production could decline markedly if the system reverted
Figure 4.8: Comparison of nekton biomass in the Gulf of Alaska from two periods. After Brodeur and Ware (1995).
276
Long-Term Ecological Change in the Northern Gulf of Alaska
Figure 4.9: Changes in settled zooplankton volume at AFK hatchery in northern Prince William Sound in spring (May–June) since 1981–1998. Data from Cooney et al. (2001).
to one supporting fewer zooplankton. Cooney et al. (2001) report that wild pink salmon production in PWS and springtime zooplankton levels (upper 20 m) declined by about a factor of 2.0 from highs in the 1980s to lows in the 1990s (Fig. 4.9). These observations tracked similar changes in zooplankton noted by Sugimoto and Tadokoro (1997, 1998) for the broader Gulf of Alaska and Bering Sea. Brodeur et al. (1996) reported on gradients in zooplankton biomass between northern reaches of the Gulf of Alaska and the waters of Oregon and Washington. Stock production in these different areas appears to vary inversely over time – when zooplankton is abundant in the Washington–Oregon shelf waters, it tends to be lower in the northern shelf waters of the Gulf of Alaska and vice versa. This inverse relationship also holds for the status and production of salmon populations; cycles of high production in the northern Gulf of Alaska coincide generally with lower production for stocks originating in the Pacific Northwest and southern British Columbia. One of the more curious aspects of long-term zooplankton change is reported by Mackas et al. (1998) in a study that examined interannual variability in the developmental timing of Neocalanus plumchrus in the North Pacific Ocean. Using samples from the Station P weathership program and Canadian GLOBEC studies, the investigators were able to document a long-term trend in the biomass maximum of this large and dominant calanoid copepod. Over a period of nearly 30 years, the maximum rose gradually from early mid-May (1958) to a much later maximum in late July (1974) and
Long-Term Change 277
Figure 4.10: Changes in biomass maximum of Neocalanus plumchrus at Ocean Station P, 1955–1994 (after Mackas et al., 1998). Neocalanus plumchrus (photograph Courtesy of Dr. Russel Hopcroft, University of Alaska).
then returned to a much earlier maximum, again in mid to early May (1994) (Fig. 4.10). The reasons for this cycling are believed to involve shifts in the temperature-dependent developmental timing of the different naupliar and copepodite stages and changes in the survival of early versus later portions of each annual cohort. These changes were linked statistically to similar shifts in ocean climate over the same time span. These and other studies are beginning to paint an overall picture of long-term change in zooplankton communities in the subarctic Pacific (see Mackas and Tsuda, 1999, for a comprehensive review). There is evidence that in addition to the prominent north/south gradient in stock production that shifts its sign over time in relation to cycling ocean climate, there is also an east–west gradient; the western gyre being more productive than the Alaska gyre. As we have seen above, changes in biomass by a factor of 2–3, marked spatial sifts that involve the whole of the Gulf of Alaska, and apparent relationships to similar changes in pelagic nekton all point to a highly dynamic system at lower trophic levels responding in poorly understood ways to
278
Long-Term Ecological Change in the Northern Gulf of Alaska
climate forcing. Presently missing, but now under intensive study, is how open ocean changes on a variety of time and space scales influence the more productive shelf and coastal areas in the Gulf of Alaska, regions where almost all of the reproductive ecology of apex consumers occurs. In addition to the intrusion of oceanic-derived zooplankters into the shelf and coastal zones of the Gulf (as noted in an earlier chapter), the propensity for these intruders to grow to larger absolute size likely has important trophic consequences for consumers (see Section 2.4).
4.4. History and Production Trends in Salmon Theodore Cooney 4.4.1. Introduction The long-term record of salmon abundance comes from two sources – the historical record of the fisheries and geochemical records in sockeye salmon lake sediments (Finney et al., 2002). This section begins with a record of the exploitation of the salmon populations, examines the longer-term historical record extracted from lake sediments in the northern Gulf of Alaska sediments, and then discusses the role of climate and other factors in salmon population fluctuations.
4.4.2. Resource Use and Management Since early times, Pacific salmon populations have supported human kind in the coastal and inland waters of the Gulf of Alaska. Archeological studies of Prince William Sound and other nearby regions indicate man’s presence dates back at least 3000–4000 years (Lethcoe and Lethcoe, 1994). When the Europeans arrived in the middle of the eighteenth century, there were eight groups of Chugach (Alutig) natives living around the sound. These ancient peoples subsisted primarily on sea mammals, fishes, shellfish, and birds. Salmon was immensely significant in their culture, as it is today among their survivors. Commercial exploitation of Gulfwide salmon stocks began in the late 1800s primarily pursued by companies operating in San Francisco and Seattle. The full-scale commercial exploitation of salmon in Prince William Sound began with the incorporation of the Alaska Packers Association (APA) in San Francisco in 1892 (Lethcoe and Lethcoe, 1994). By 1900, the APA owned two-thirds of the canneries in Alaska, and accounted for 72% of the annual pack. During the early days of commercial harvest, fishing practices were ruthless, and waste was condoned in the absence of any substantial regulation of the industry. When Alaska was made a territory in 1918, Alaska fishermen demanded the territorial
Long-Term Change 279
government abandon the use of ecologically damaging fish traps owned by the industry to catch salmon. However, the salmon industry lobby was so strong that fish traps were still being used at some locations until statehood in 1959. Alaska Department of Fish and Game assumed responsibility for managing salmon resources in 1960. The State mandated that harvests be regulated to sustain salmon stocks at levels that would both conserve the resource and provide a reasonable livelihood for fishers and the industry. To this end, policies and regulations were put in place to assure that sufficient numbers of spawning adults would be permitted to escape the fishery each year to sustain a strong reproductive stock. In 1974, following a history of significant year-to-year variability in salmon catches associated with natural disasters (the 1964 earthquake and several severe winters in the late 1960s and early 1970s), the Alaska legislature passed the Private Salmon Hatchery Act as a means to help rehabilitate salmon resources. Private, non-profit regional aquaculture corporations were authorized. The concept was ocean ranching – rearing eggs and alevins in a protective hatchery environment and then releasing the juveniles in the ocean to grow and return later with wild fish to a common-property fishery. In the winter of 1974–1975, stakeholders, including fishermen, processors, city governments, and native corporations formed the Prince William Sound Aquaculture Corporation (PWSAC), with offices in Cordova, Alaska (Olsen, 1994). PWSAC would eventually administer four large hatcheries in Prince William Sound that produced all species of salmon (Fig. 4.11). Together with a hatchery near Valdez (Valdez Fisheries Development Association),
Figure 4.11: Photo of AFK Hatchery, Prince William Sound Aquaculture Association (photograph courtesy of Mandy Lindberg, NOAA).
280
Long-Term Ecological Change in the Northern Gulf of Alaska
MILLIONS OF FRY RELEASED
350 300 250 200 150 100 50
1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003
0
YEAR
Figure 4.12: History of releases of hatchery-raised pink salmon fry from three Prince William Sound hatcheries operated by the William Sound Aquaculture Corporation (Armin F. Koernig, Cannery Creek, and Wally Noerenberg).
the combined release of juveniles would exceed 500 million fry and smolts annually by the early 1990s (Fig. 4.12). Current hatchery returns of 15–20 million adults to Prince William Sound each year show the success of the hatcheries in producing harvestable salmon.
4.4.3. Broadscale Trends in Time and Space Analyses of the voluminous information and data bases for northeastern Pacific salmon with some proxy values extending back in time more than 2000 years illustrate four major time–space patterns: (1) very-low-frequency production variability on a millennial scale for sockeye salmon (Oncorhynchus nerka) (Finney et al., 2002) (see Fig. 4.1.A), (2) a pronounced north/south production gradient in the northeastern Pacific Ocean that changes its sign on decadal scales – the so-called “inverse production regime” (Hare et al., 1999), (3) statistical relationships between annual catches and several indices of climate forcing that are common across species and regions in the North Pacific (Beamish et al., 1999), and (4) evidence of decreasing adult body size for adults during the past two decades (Bigler et al., 1996). Finney et al. (2002) examined temporal patterns in a stable isotope of nitrogen in salmon-derived nutrients (SDN) sampled in sediment cores from Karluk, Akalura,
Long-Term Change 281
and Frazer lakes on Kodiak Island, Alaska. From this information, they reconstructed a millennial-scale record of adult sockeye salmon returning to spawn in these lakes that extends back to about 200 BC (see Fig. 4.1.A.). Over roughly 2200 years, the record depicts declining escapements from a high at the beginning of the record to a low around AD 100, followed by mostly steady increases to high values again in AD 1200 and beyond. The last 100 years (or so) of the time-series document rapid declines, which are probably associated with the development of the modern red-salmon fishery. Since contemporary records illustrate a general tracking of population change among most salmonid stocks in the northern Gulf of Alaska, the very long-term reconstructed production history for sockeye salmon may also apply to pink (O. gorbusca), chum (O. nerka), silver (O. kisutch), and king salmon (O. tshawytscha). Hare et al. (1999) used a sophisticated statistical procedure to demonstrate an inverse production response to shifting ocean climate for salmon stocks in the northern Gulf of Alaska and those found off Washington and Oregon. They discovered that under certain conditions of ocean climate, the southern stocks did less well than those occurring further north, and that these different production regimes appeared to change their sign in response to climate forcing described later as the Pacific Decadal Oscillation (PDO; Mantua et al., 1997) (Fig. 4.13). Following a dramatic “regime shift” in the late 1970s, most stocks of salmon in Alaska underwent unprecedented expansion whereas those in Washington and Oregon languished or declined. This pattern was reversed in the decade prior to 1977–1978, and at present there is evidence that yet another change has occurred or is occurring (Sugimoto and Tadokoro 1997, 1998). Recently, the returns of salmon to southern coastal regions (Washington and Oregon) have been strengthening under cooler temperatures and higher plankton production. The surprise
Figure 4.13: The relationship between the Pacific Decadal Oscillation and salmon harvests in the Gulf of Alaska (after Mantua et al., 1997).
282
Long-Term Ecological Change in the Northern Gulf of Alaska
has been that northern Gulf stocks have generally continued to produce at high levels during this same period rather than reverting to significantly lower levels as our understanding of the inverse production regime would predict. Obviously, there are some general conditions that favor elevated levels of production in both the southern and northern reaches of the Gulf that need to be investigated further. Hare and Francis (1995) used intervention analyses to determine the dates of temporal shifts in ecosystem state associated with long-term records of pink and sockeye salmon originating in three different regions of Alaska – western, central, and southeast. For records of catch extending back to the 1920s, these analyses demonstrated a remarkable concurrence between shifts in production from high to low and back to high across both species and the three different regions. The shifts occurred in the late 1940s and in the late 1970s, the latter associated with a well-documented change in ocean climate. Records for air temperature at Kodiak, Alaska, and the time-series of the North Pacific Index (NPI) demonstrated similar dates for changes in state. Rogers (1984) hypothesized that increased catches of sockeye salmon during warm ocean periods reflected better marine survivals caused by an altered (shortened) migratory pathway for returning adults that lessened their vulnerability to predation by marine mammals. Brodeur and Ware (1992) suggested that increased salmon production during this same period was caused by much improved foraging opportunities for juveniles. Their work demonstrated that during the late 1950s and early 1960s, oceanic zooplankton was much less abundant and dramatically more patchy in its distribution than in the 1980s. The warm period of the 1980s was one of more intense Aleutian low pressure and greater center gyre upwelling that presumably improved nutrient supply and increased primary productivity supporting higher zooplankton stocks. Although different mechanistically, both ideas linked salmon survival to meteorological forcing of ocean climate. Concern over the possibility that growth-limited adult size may be occurring was articulated by Helle and Hoffman (1995) and Bigler et al. (1996), who demonstrated a trend toward smaller adults for all five species of Pacific salmon since the late 1970s. This phenomenon occurred during the same time that Alaskan stocks were undergoing unprecedented increases. It was believed that density-dependent growth was forced by competition for limiting forage during the annual feeding migrations of these fish. Cooney and Brodeur (1998) reviewed the problem and used a simple bioenergetics model to demonstrate that most of the consumption of forage occurs in the oceanic rather than coastal feeding areas. This does not mean that density-dependent growth does not occur in the abundant younger stages, but that, in an absolute sense, the daily ration and overall forage demand is much greater for the larger, older fishes feeding in the oceanic environment. Smaller adult size is associated with fewer and smaller eggs per female and a general reduction in reproductive fitness on the spawning grounds. Some have speculated that the growing contribution of hatchery-reared
Long-Term Change 283
PINK SALMON CATCH (metric tonnes)
200000 180000 160000 140000 120000 100000 80000 60000 40000 20000 2000
1995
1990
1980 1985
1975
1965 1970
1960
1955
1945 1950
1940
1935
1925 1930
1920
1910 1915
1905
1900
0
YEAR
Figure 4.14: The history of pink salmon catches in the Gulf of Alaska, 1900–2001 (after Brodeur et al., 2003).
salmon to an overall Gulf population is stressing present levels of oceanic carrying capacity for salmonids (and other consumers) in the northern Gulf of Alaska, despite evidence of significant increases in salmon forage after the late 1970s. Brodeur et al. (2003) portray the historical catch (in metric tons) of pink salmon from Alaska over the period 1900–2001 (Fig. 4.14). Between 1900 and 1940, the catch generally reflected the increasing development of the fishery. Between 1940 and 1960, declines in catch are believed to be associated with overfishing and adverse environmental conditions. The harvest recovered somewhat in the 1960s, only to fall on hard times again in the early 1970s. Closure of the Japanese high-sea drift net fishery, improving environmental conditions, and the establishment of the private, non-profit hatchery program led to a period of unprecedented salmon production with catches increasing at a steady rate from about 1974–2001. Thus, over a century, the pink salmon industry in Alaska grew from landings of about 10,000 mt in the early 1900s to approximately 200,000 mt today. Rogers (1986) reported that pink salmon in Central Alaska (mostly from Prince William Sound and Kodiak Island) accounted for about 35% of all Alaska pinks between 1950 and 1977, but the growing contribution of hatchery-reared pink salmon after that time raised the percentage substantially in the late 1980s and 1990s (Fig. 4.15). Between 1950 and 1984, Central Alaska pink salmon returned an average of 22.5 million adults each year although the range of returns was broad; 5–59 million fish.
284
Long-Term Ecological Change in the Northern Gulf of Alaska
ADULT RETURNS (millions)
40 35 30 25 20 15 10 5 0 1960 1965 1970
1975 1980 1985
1990 1995 2000
YEAR
Figure 4.15: The returns of wild and hatchery pink salmon to Prince William Sound.
4.4.4. Prince William Sound: A Pink Salmon Case History Alaska Department of Fish and Game in Cordova, Alaska began compiling production statistics for pink salmon in Prince William Sound immediately after being declared a state in 1959. This information included spawning escapements, catches, total returns for both wild and (later) hatchery stocks, marine survivals for hatchery fish, and preemergent wild-fry indices. The preemergent index was derived from a quantitative census of alevins surviving in streams and intertidal natal areas a few weeks before the beginning of the peak of fry entry into marine waters each spring (this program was discontinued in 1996). The production of pink salmon in Prince William Sound has varied over the years (Fig. 4.15). In the 1960s, overall returns averaged about 7 million adults annually, perhaps in part caused by the loss of spawning habitat associated with the largest earthquake in Alaska recorded history in 1964 (see also Section 3.3.1). The epicenter of the quake was in northwestern Prince William Sound. In the 1970s, the average return increased to about 10 million annually, primarily driven by production increases late in the decade. Production grew dramatically in the 1980s and 1990s to an average return of 25 million adults in recent years. A large part of this increase was associated with the growing contribution of hatchery-produced pink salmon that began to dominate the return of adults in the late 1980s. Part of the production increase has also been attributed to an improvement in marine rearing conditions that began in 1978 during the widespread oceanographic regime shift that began at that time and lasted through the 1980s. Pink salmon stocks in both southeastern and central Alaska regions
Long-Term Change 285
responded positively to a generally warmer and wetter ocean climate following the shift (Hare and Francis, 1995). The time series of wild and hatchery pink salmon stock production in Prince William Sound has been interpreted differently by those who believe hatchery practices eventually lead to the demise of natural stocks (Hilborn and Eggers, 2000), and those who contend that these different populations can exist together without unduly influencing each other (Wertheimer et al., 2001). Those who see hatchery stocks as detrimental to the sustainability of wild populations point to Prince William Sound as a prime example of hatchery pink salmon displacing the wild production. This displacement allegedly began in the mid to late 1980s and continues to this day. As evidence, opponents of hatcheries point to the increasing proportion of hatchery fish to the overall returns from 1985 to the present. The opposing view uses the same time series to demonstrate that declines in wild production following the elevated production years in the 1980s merely represent a return of wild populations to pre-regime-shift levels after a period of elevated escapements and unusually good ocean survival conditions. Sugimoto and Tadokoro (1997) provide evidence for another shift in ocean climate in the north Pacific that occurred in the early 1990s. After the salmon hatcheries reached their permitted levels of production in the late 1980s, they have been consistently placing about three times more fry into Prince William Sound than do the wild stocks each year. The return-per-alevin index can be used as a proxy value for the marine survival of wild pink salmon in Prince William Sound. For return years beginning in 1962 and continuing through 1996, anomalies of this value demonstrated a period of generally below average marine survivals from 1962 through 1978 followed by seven years of consistently above average survivals from 1979 through 1985 (Fig. 4.16). After that time, the anomalies of marine survival have bounced back and forth between positive and negative values. This pattern supports the view that a dramatically improved ocean climate stimulated elevated wild-stock production following the regime shift in 1977–1978. Other factors probably contributed as well. Wild pink salmon escapement increased over the same period and this may have contributed to the observed increase in alevins at this time. In the sound, higher levels of alevins generally result in larger adult returns (Alaska Department of Fish and Game, unpublished). The Sound Ecosystem Assessment (SEA) program in Prince William Sound investigated mechanisms influencing the mortality of wild and hatchery-reared pink salmon during early marine residence, the life stage that most believe determines the strength of the adult return (Willette et al., 2001). SEA discovered that water temperature affecting growth rates, amounts of fry food, the kinds and abundances of fry predators (primarily other fishes and birds), and the kinds and abundances of alternative prey for these same predators all interacted in ways that influenced the survival of fry each year. As noted previously, two facultative planktivores, adult pollock and Pacific herring early in the season, followed by juvenile gadids later, comprised the
286
Long-Term Ecological Change in the Northern Gulf of Alaska
RETURN/FRY INDEX ANOMALY (x1000)
40 30 20 10 0 −10 −20 −30 1960
1965
1970
1975 1980 YEAR
1985
1990
1995
Figure 4.16: Anomalies in the return/fry index (a proxy for marine survival) for Prince William Sound pink salmon, brood years 1962–1996.
major predators. Pelagic rock fishes, Dolly Varden trout, adult salmon, and some marine birds were also implicated (Scheel and Hough, 1997). A numerical model of fry loss incorporating the feeding physiology of the fry and their predators, and observed spatial distributions of zooplankton and water temperature simulated the loss process from releases at one of the northern pink salmon hatcheries. The simulation followed the fry to locations in the southwestern part of the sound where survivors stage before leaving on their oceanic feeding migration (Willette et al., 2001). Model results suggested that, in the immediate vicinity of a hatchery following releases of fry, losses to fish predators (adult pollock and herring) were associated with a complex predator consumption process. In the megaschools of fry (many millions of small salmon), predator swamping (or not) determined the loss rate. If fry densities provided sufficient food to satiate the large fish, fewer fry were taken than under conditions where adult pollock and herring could feed continuously. It was also discovered that in the immediate area of the releases, feeding on fry was almost always more energetically beneficial to predators than feeding on zooplankton. However, as the fry schools dispersed from hatchery sites during a southward migration, alternative prey for pollock and herring could occur at levels sufficient to
Long-Term Change 287
“switch” the predator consumptive process to euphausiids, large calanoids and/or pteropods (Willette et al., 1999). In contrast, it seems unlikely that the entry of wild fry over a 60-day period each spring (from 800 or so small streams and rivers) would ever approach the densities of juveniles found in megaschools around hatcheries, so “zooplankton sheltering” may be a key factor influencing the losses of wild stocks during early marine residence in most years. Cooney et al. (2001) noted that, during the 1980s, zooplankton stocks and adult returns of wild pink salmon in Prince William Sound (PWS) were approximately twice those occurring later in the 1990s. Given what is presently understood about pink salmon in Prince William Sound and the interacting factors that account for high mortalities during early ocean residence (perhaps as many as 75% of the total fry entry), the historical pattern of wild stock production suggests the following: (1) a period from 1960 through the middle 1970s where marine and freshwater survivals sustained the stocks at levels of 5–10 million per year (2) a period of approximately 10 years of elevated production – 20–30 million wild and hatchery adults per year during the 1980s when annual marine survivals were often well above average – and when overescapement may have been partly the cause of increased alevin densities in the spring, and (3) a diminishment of elevated wild stock production in the early 1990s back to pre-1980 levels while hatchery stocks remain high. The cause of the wild production decline continues to be debated.
4.4.5. What’s Behind the Large-Scale Patterns in Salmon Catch and Production? It seems obvious from the above that salmon production is influenced by periodic if not strictly cyclic shifts in ocean climate on several different time and space scales (Francis and Sibley, 1991) and by management and harvest strategies that include stock enhancement (hatcheries). Most agree that run-strength is established early in the marine life history in near-shore waters. Most also agree that predation, rather than starvation is the most important factor. In this regard, there is evidence that mortality during the early marine stages is size-dependent, the smaller fishes being at much greater risk to predation than the larger. As a consequence, growth rates apparently mediate these early losses; both water temperature and food have been shown to affect juvenile growth. Fry growing slowly remain in the smallest, most vulnerable sizes proportionally longer than do fry that encounter optimal growth conditions. This paradigm has been challenged recently by Beamish and Mahnken (2001) who provided evidence that at least for coho (silver) salmon (Oncorhynchus kisutch) there is reason to believe that unless the smolting juveniles are sufficiently prepared (physiologically) for their year in the ocean, they may not survive. Physiological preparation seems
288
Long-Term Ecological Change in the Northern Gulf of Alaska
to be associated with achieving a certain minimal size. Juveniles failing to reach this “critical size” apparently do not survive a winter at sea. It has not been demonstrated convincingly that other salmonids in the system are affected in a similar way. Shifts in ocean state in the northeastern Pacific seemingly affect salmon production according to where a particular stock resides on the north/south production gradient between Oregon/Washington and Alaska. An ocean state that exhibits warmer and generally wetter than average conditions in the north apparently favors production there (Mueter et al., 2002; Hare et al., 1999). Under the “northern warm regime” (positive PDO) increased center gyre upwelling and a deeper upper mixed layer in the Gulf responding to an intensified Aleutian low may improve growing conditions for plankton (juvenile salmon forage) by increasing nutrient supplies and plankton production, and through increased runoff, creating a more stable photic zone (Gargett, 1997). Also, an open ocean distribution of zooplankton that strengthens population levels at the continental margins (Brodeur and Ware, 1992) may “leak” more ocean-derived zooplankton landward, reinforcing shelf and coastal stocks that serve as food for juveniles and alternative forage for salmon predators. Further, when (for whatever reason) zooplankton stocks are elevated, the overall trophic status of the ecosystem probably shifts toward planktivory, the result being that generalists derive more of their energy by feeding on zooplankton rather than eating small fishes (such as juvenile salmon) (see Section 4.4.5). Conversely, when zooplankton stocks collapse, generalists that rely heavily on plankton must supplement their diets by feeding more on small fishes and squids, and the ecosystem will reflect a more piscvorous status. These feeding mode shifts have been observed and described (Willette et al., 1999). Finally, a warmer and wetterthan-average northern system probably favors greater freshwater survivals of early life stages. Heavy snowfall tends to insulate salmon eggs and alevins from freezing conditions, and periods of melting maintain stream flows necessary to provide sufficient oxygen for respiring embryos. Warm spring conditions promote rapid runoff and the flushing of juveniles into shallow coastal waters in relatively high numbers. These large outmigration events may create local predator swamping and enhanced overall survival (Cooney et al., 1995). At the same time, a warmer ocean climate in southern waters probably means greater stratification and reduced coastal upwelling resulting in less plankton production perhaps leading to an ecosystem adjustment toward piscivory and greater losses to populations of small fishes, including juvenile salmon. Warmer conditions in the south may also allow warm-water predators (such as Pacific hake) to invade a region from more southerly locations. Hare et al. (1999) suggest that conditions off Washington and Oregon over the last two decades (detrimental to salmon production) may have in some measure been responsible for the lackluster performance of hatchery stocks of coho salmon rearing in coastal waters. If this was the case, a “return” to a higher production regime in southern
Long-Term Change 289
waters – believed by some to be underway now – could improve the contribution of southern hatchery stocks to local fisheries, a largely unrealized regional strategy for many years. The strength of wild salmon populations in the northern Gulf of Alaska has been attributed in part to management activities designed to protect the reproductive potential of stocks. In an odd twist of fortune, the economic value of this immense and healthy resource has recently been strongly and negatively impacted by globally available farmed salmon. The devaluation of wild stocks and ocean-ranched salmon is affecting the harvest sector and thus one of the major “tools” that managers use to create effective spawning escapements each year. When the price per pound offered to fishermen declines substantially, harvesters may decline to fish, or may demand that the hatcheries produce more fish to increase the fishable stock size. Taken to extremes, the result may be many more low value fish competing for forage resources on the high sea, maturing at a later age and returning as smaller adults with compromised reproductive potential (Cooney and Brodeur, 1998). Aggressive marketing programs are presently underway to restore the economic value of non-farmed salmon populations in the Gulf of Alaska. Only time will tell whether these actions will succeed (Box 4.3).
BOX 4.3 THE ROLE OF SPAWNING HABITAT by Theodore Cooney Natural limitations are placed on the production of wild salmon stocks by the distribution and amounts of spawning habitat available to adults each year. Clear-water rivers and streams with suitably porous, gravelly sediments provide substrates that can successfully incubate the eggs and host the developing larval salmon. The strict dependence on these “spawning habitats” places upper bounds on the numbers of spawning adults that can be successfully accommodated each year. Fully utilized spawning habitats under ideal conditions will produce a given number of juveniles whose survival to adulthood is determined primarily (but perhaps not always) by losses to predation during the ocean feeding and maturation phase of the life history. When the adult returns are particularly large, natal habitats may be overrun by adults and the capacity of a region to produce juveniles seriously degraded. Because of this, resource managers attempt to regulate the harvest so that an optimal number of adults are allowed to “escape” the fishery each year. The sustained productivity of wild salmon in the northern Gulf of Alaska has been attributed, in large measure, to the success of this management strategy.
290
Long-Term Ecological Change in the Northern Gulf of Alaska
4.5. Pacific Herring Evelyn Brown 4.5.1. Resource Use and Management Pacific herring (Clupea pallasi), as with salmon, have been an important human food for thousands of years (Lethcoe and Lethcoe, 1994). Herring eggs were especially important to Alaskan natives as they were easily collected on beaches after spring spawning (Fig. 4.17). Eggs provided an early boost of high fat and protein, a welcome repast after depleted winter food supplies. Herring spawn on spruce boughs is a delicacy, and boughs were left in spawning beds during low tide. The laden bows were dipped briefly in boiling water and then in seal oil. To the native peoples in the region, the arrival of vast schools of herring in the spring was a sign of an abundant harvest of salmon and marine mammals during the season as these predators were known to accompany the herring schools whereever they went. In this way, herring were harbingers of good fortune, plentitude, and a reason to celebrate. Commercial fisheries for Pacific herring began in 1882 in Alaska (Reid, 1971), expanded rapidly, and catches were in the thousands of tons by the early 1900s. At first, they were dry-salted for oriental markets (1904–1934). From 1935 to 1967, herring were reduced for meal and oil, then sold domestically and overseas (Rousenfell and Dahlgren, 1932; Skud et al., 1960; Reid, 1971). From 1971 to the present, roe and spawn on kelp (both for oriental markets), food (mainly overseas), and bait (domestic) (Hourston and Haegele, 1980) were the main uses. The history of GOA catches and the estimated biomass of herring is shown in Fig. 4.18. Herring fisheries in the Gulf of Alaska were second only to salmon in tonnage. Catches during the early
Figure 4.17: A thick layer of herring spawn on a beach in Prince William Sound (photograph courtesy of Evelyn Brown, University of Alaska).
Long-Term Change 291
reduction fishery days were not controlled, and fishermen sometimes took a large percentage of the population, likely severely reducing or eliminating some populations (Rounsefell, 1929; Reid, 1971). Alaska Department of Fish and Game (ADF&G) took over the stock assessment programs from the Federal Department of the Interior (Bureau of Commercial Fisheries) in 1960. This ongoing program consisted of estimating harvest by the three main Gulf of Alaska (GOA) districts (Prince William Sound, Kodiak, and Southeast Alaska; Fig. 4.19), and sampling the catch for size and age from the mid-1920s on. There were also tagging studies to assess migrations and vertebral counts as a way to assess population discreteness (Rounsefell, 1929; Reid, 1971). Catch quotas were first introduced in 1940 but later abandoned in 1952, a mixed result of integrity of biological assessment data, politics, and economics (Reid, 1971). By the late 1950s, aircraft were regularly used to measure the extent of spawning and the size of the spawning population. Aerial surveyors provided the first real indices of abundance for regional populations (Grice and Wilimovsky, 1957) The aerial surveys, estimates of catch and catch age–weight–length (awl) sampling were continued by ADF&G from the 1960s on as the mainstays for herring stock assessment and fisheries management.
Figure 4.18: The two types of fishery data used in this analysis. The solid black line is total annual Gulf of Alaska fishery catches (Skud et al., 1960; Reid, 1971; Gretsch et al., 1989; Prokopowich, 1989; Schroeder and Morrison, 1989; Brady et al., 1991; Larson and Minicucci, 1991; Funk and Harris, 1992; unpublished catch records from Fritz Funk, ADF&G, Juneau, Alaska). The grey line represents the annual biomass estimates for Prince William Sound in the northern GOA (Funk and Harris, 1992; unpublished stock assessment records from Fritz Funk, ADF&G, Juneau, Alaska).
292
Long-Term Ecological Change in the Northern Gulf of Alaska
Figure 4.19: The Gulf of Alaska region covered in this chapter.
The demand for herring reduction fishery products was in decline during the 1960s, the harvest declined as a result, and fishermen began to experiment with other herring products and markets. The markets artificially controlled harvest levels until the early 1970s, when Japanese demand for herring roe products increased dramatically. The rapid increase in herring roe price and fishery expansion led to a need for fishery quotas and more accurate stock assessment tools. The ADF&G herring program increased dramatically as a result and a quota of 20% of measured stock level was set by 1973. In 1982, ADF&G initiated a hydroacoustics project to assess if that method may provide better information than the aerial counts of miles of spawn (Gaudet, 1984). Acoustics were used for about 5 years, and then discontinued as there were problems interpreting and separating populations of adults and juvenile herring, which were mixed at times and separate at other times. Modeled after a Canadian survey method, ADF&G initiated egg deposition surveys with divers sampling egg density along randomly placed transects (Willette et al., 1998). This method provided the most accurate estimate of the herring spawning biomass and was used from 1985 on in Southeast Alaska and from 1988 to 1998 in Prince William Sound. During that same period, virtual population analysis was used to develop an age-structured assessment (ASA) model to track survival of adult herring by age (Funk and Harris, 1992). Population estimates from egg deposition and aerial surveys along with regional age structure information were inputs for the ASA models. From the late 1980s on,
Long-Term Change 293
ASA modeling has been the primary tool used by ADF&G for setting fishing quotas and estimating population size (pers. comm., Fritz Funk, Div. of Comm. Fish. Manage., ADF&G, Juneau, Alaska; Funk and Harris, 1992). Since the late 1990s, demand for herring roe products and bait has been in the decline. Ground prices have fallen off and fishery participation has declined as a result. Herring stock assessment programs have also suffered because commercial fish taxes and commercial test fisheries (that help fund the program) have been reduced. Herring research and assessment programs around the GOA are operating on minimum budgets, and there is no commercial or political pressure to initiate new or improved herring research projects. Given the ecological importance of the species, this situation is unfortunate.
4.5.2. Effects of Climate on Gulf of Alaska Herring All Pacific herring stocks respond to climate similarly to salmon but differently than other forage fish species. A negative correlation exists between southern British Columbia (BC) herring year-class strength and warm conditions; warm conditions appear to reduce zooplankton food resources and increase predation by other fish on herring (Ware, 1992) (see also Sections 2.5 and 4.1). The same inverse relationship is reported by Hollowed and Wooster (1995) with higher average recruitment for Vancouver Island herring during cool years associated with a weakened winter Aleutian low pressure. However, the opposite effect occurs in northern BC and the Gulf of Alaska with increased herring production during warm years associated with an intensified winter Aleutian low pressure (Hollowed and Wooster, 1995). Pacific herring appear to have a similar north versus south opposing response to that observed in Pacific salmon (Beamish, 1993). In Alaska, recruitment of Southeast Alaska Pacific herring is positively associated to warm, wet climate conditions (Zebdi and Collie, 1995). Furthermore, synchronicity in herring recruitment patterns corresponds to hydrographic domains in the northeast Pacific (Zebdi and Collie, 1995) as originally hypothesized by Ware and McFarlane (1989). In the GOA, Pacific herring may be out of phase with other forage species, including capelin (Mallotus villosus) and shrimp, which appear to do better during the cool phases associated with a weakened Aleutian low (Anderson and Piatt, 1999). Periods of high herring population have generally been associated with the positive phases of the significant climate indices (Brown, 2003). An index of Prince William Sound spawning biomass was significantly correlated to four major climate indices, the Pacific Decadal Oscillation (PDO; Mantua et al., 1997), the Aleutian Low Pressure Index (ALPI; Beamish and Bouillon, 1993), the Pacific Inter-Decadal Oscillation (PIDO; Enfield and Mestas-Nuñez 1999), and the Atmospheric Forcing Index (AFI; MacFarlane et al., 2000) (Fig. 4.20). An index representing all Gulf of Alaska herring stocks (including Kodiak and Southeast Alaska) was also significantly
294
Long-Term Ecological Change in the Northern Gulf of Alaska
Figure 4.20: A 4-year moving-average (ma) transformation of the PWS biomass index compared to the AFI (lagged 4 years), the ALPI (lagged 4 years), and the winter PDO (no lag) for the period of 1973–1993.
correlated with the same four variables (Fig. 4.21) which all measure a similar phenomenon in different ways. The relationship of GOA herring to climate indices is similar to the relationships for salmon production and climate (Mantua et al., 1997; Hollowed et al., 1998) and GOA zooplankton production (Brodeur and Ware, 1992; Brodeur et al., 1996). In the subarctic, a strong Aleutian low causes above-average fall and winter water column mixing, with a high influx of nutrients, followed by above-average spring to summer stability creating conditions that optimize primary and secondary production. The scenario is a direct application of the optimal environmental window theory (Cury and Roy, 1989) where a domed shaped relationship exists between wind-induced upwelling and recruitment due to nutrient limitation of primary production during periods of low upwelling versus excess turbulence during intensified upwelling. This may be the mechanism involved in the positive response of zooplankton and Alaskan Pacific salmon stocks to a positive PDO signal (Gargett, 1997) (see Sections 4.2 and 4.4). Because GOA herring react similarly to climate, the mechanism may be the
(a)
(b)
Figure 4.21: A 5-year moving average (ma) of the Gulf of Alaska index of herring abundance, created by combining catch and biomass, compared to a 5-year ma of the Pacific Inter-Decadal Oscillation (a); Enfield and Mestas-Nuñez (1999) and 5-year ma of the GOA index, AFI, ALPI, and winter PDO plotted for the period of 1902 – 1995 (b).
296
Long-Term Ecological Change in the Northern Gulf of Alaska
same for herring. Enhanced zooplankton resources will likely result in higher growth and possibly better survival for early life stages of herring. Periods of high herring production in the GOA, during the 1930s and 1980s, correspond with periods of positive PDO and AFI index values but not in the 1960s (Figs. 4.20 and 4.21). The two peaks in abundance in. However, This may be due to poorly documented overall stock size information during the 1960s rather than a lack of relationship. Fishery catches were low due to weak herring markets (Reid, 1971). It is possible that herring were more abundant in the Gulf than indicated, as they were abundant in the Bering Sea during this period (Wespestad, 1991). If so, the overall correlation between population abundance and climate could have been even stronger. Changes in climate also have a dramatic effect on herring body size at age. Within the GOA region, PWS herring size-at-age trends oscillated with a maximum in spectral density at a period of 13 years for all ages, but strongest in 3–5 year-old herring
Figure 4.22: Size-at-age by weight (g) of age 3–8 Pacific herring from PWS are significantly correlated with peak zooplankton density anomalies (from southwestern PWS) for the period of 1973–1999 of the size-at-age data plotted with a 4-year moving- average transformation of peak zooplankton anomaly.
Long-Term Change 297
(Brown, 2003) (Fig. 4.22) Size-at-age trends were also significantly correlated with peak zooplankton density lagged 1 year, corresponding to the food available to herring juveniles at age-1 (Fig. 4.22). Zooplankton biomass was, in turn, significantly correlated to the winter PDO. Although size-at-age did not decrease with population size, indicating a lack of density-dependence at the population level, the close relationship with zooplankton and fish size indicates that food limitation may occur. There are also climatic affects on the timing of events in herring life history, which may in turn, affect organisms that rely on these large aggregations of energy rich fish. In PWS, mean spawn dates were highly variable and spawning had been progressively earlier since the 1980s but similar to spawn dates in the 1970s (Fig. 4.23) (Brown, 2003). The mean spawn date anomaly was weakly correlated to the September–October. SST anomaly the fall prior to spawning but mean spawn date was not significantly correlated to any other climate variable. The relationship between SST 6 months prior
Figure 4.23: The smoothed series for the September–October (6 months prior to spawning or cohort year) mean sea surface temperature (to 20 m) anomalies plotted with the mean date of spawning anomaly for PWS for the period of 1973–1999. The time series correlation is weak, but significant.
Figure 4.24: The ln R/S index (lagged 1 yr) is also significantly correlated to the proportion of spawning (by lineal coverage on the beach) occurring in eastern PWS for the period of 1973–1995 (a). The natural log transformation of the recruit per spawner (ln R/S) index for PWS is significantly correlated with with average zooplankton density anomaly (a corresponding to age-1 herring around their first birthday (b).
Long-Term Change 299
to spawn and the mean spawn date agrees with other studies on Pacific herring. Spawn timing is a function of the temperature exposure history of the adults (Hay, 1985; Hay and Kronlund, 1987; Ware and Tanasichuk, 1989; Wespestad, 1991) as well as the age composition since older, larger herring spawn earlier (Hay, 1985; Ware and Tanasichuk, 1989). Because of the relationship between spawn timing and age composition, spawn timing should be highly variable due to variations in SST and the arrival of a recruiting cohort every few years. Herring exhibit plasticity in maturation rates and spawn timing as an adaptive process in response to changing ocean conditions (Lasker, 1985; Winters and Wheeler, 1996; Sinclair, 1988). Therefore, spawn timing should generally follow ocean conditions, especially temperature. Climateinduced shifts in herring spawn timing could well have effects on migrating shorebirds and sea ducks, which stopover and replenish their energy on herring spawn in Prince William Sound. Sea lions and whales as well as numerous other predators also rely on herring but how they are affected by an altered timing of herring spawning is not known.) Fish production is often described in terms of an index of recruits surviving per spawning individual or group of individuals and climatic impacts were also found on this index for PWS herring. The log-transformed recruit per spawner (ln R/S) index for age-4 herring was highly variable but exhibited a general downward over the last 26 years (Fig. 4.24(a)) (Brown, 2003). During this period, spawning area use had also shifted from mainly eastern to western PWS. The ln R/S was significantly correlated to average zooplankton settled volume anomaly lagged 1 year (Fig. 4.24(b)). As with the correlation to size-at-age, this lag corresponds to secondary production during age 1 or just prior to the second birthday of the cohort. This ln R/S was significantly correlated to the proportion of eastern spawning (Fig. 4.24(a)). During the 1980s, when the population was at a peak, spawn was more evenly spread among eastern, western, and northern spawning beaches than in the 1970s or 1990s. In PWS, the links between the declining R/S trend, the east to west spawn area shift and climatedriven trends in ocean conditions are not well understood. The shift in spawning region may simply represent a random switch. Alternatively, the shift in spawning area could be a function of climate-driven changes in ocean conditions favoring the choice of one area over another. The actual cause of the decline deserves further examination. In summary, GOA herring populations are responding to climate-driven changes in ocean conditions. The implication for fishery management is that stock protection or building measures can only operate up to stock levels dictated by climate. Because herring play an important role in the northeast Pacific ecosystem, changes in population levels could potentially impact apex predator or herring competitor population levels. Ecosystem modeling of process-oriented responses of herring to ocean conditions e.g., the model proposed in Brown, 2003) could be a useful tool for interpreting the climatic relationships described in this chapter.
300
Long-Term Ecological Change in the Northern Gulf of Alaska
4.6. Groundfish Wiebke J. Boeing, Michael H. Martin, and Janet T. Duffy-Anderson 4.6.1. Introduction In this chapter, long-term changes are presented for several ecologically and economically important groundfish species in the Gulf of Alaska: walleye pollock (Theragra chalcogramma), Pacific cod (Gadus macrocephalus), arrowtooth flounder (Atheresthes stomias), and Pacific ocean perch (Sebastes alutus). Walleye pollock and Pacific cod are in the family gadidae; arrowthooth flounder, pleuronectidae, and Pacific ocean perch, scorpaenidae. Various factors that might have impacted each of these populations and caused the observed changes are discussed. These species are generally representative of major fish species complexes of the GOA, and historical (20+ years) estimates of adult and larval abundances are used to examine population changes. Total biomass of each groundfish species over the last 42 years (25 years for Pacific cod) has been estimated with species-specific age-based or length-based models (A’mar et al., 2003). Briefly, for walleye pollock, data integrated by the model include fishery catch and age composition, National Marine Fisheries Service (NMFS) bottom trawl and Echo-Integration Trawl survey estimates of age composition and biomass, and egg production estimates of spawning biomass (among other factors). The model for Pacific cod utilizes a length-structured model and includes commercial catch-data (biomass, size composition), NMFS survey data (size composition and abundance), and length–weight relationships to estimate total biomass (among other parameters). The model used for arrowtooth flounder includes survey estimates from the International Pacific Halibut Commission in addition to the NMFS groundfish trawl survey data, and size composition estimates for selected years. Fishery catch and size data are also incorporated. Finally, for Pacific ocean perch, the model used is a generic rockfish model that integrates information on age structure, size composition, fishery catch data, and survey biomass estimates (when available) to estimate total biomass. For all species, modeled results have been used instead of actual biomass estimates, because annual estimates are limited to survey data and some applied correction factors, whereas the models provide longer, more complete time series with retrospective analyses to estimate population biomass. The NMFS survey, conducted by the Resource Assessment and Conservation Engineering Division (RACE) of the Alaska Fisheries Science Center (AFSC), has been collecting data on a triennial basis since 1984 and biennially since 1999. The survey uses a high-opening Poly-Nor’eastern bottom trawl (127 mm mesh) with rubber bobbin roller gear. Surveys average 800 tows (range: 489–929 tows) and span the entire GOA (133–170°W). The GOA is divided into the Eastern (133–147°W), Central (147–159°W),
Long-Term Change 301
Figure 4.25: Three regions of the Gulf of Alaska sampled by the triennial NMFS survey. The survey takes place on the continental shelf (light blue).
and Western (159–170°W) regions (Martin 1997; Fig. 4.25). The modeled estimates of population biomass often exceed the results from the surveys due to catch efficiencies of the gear. Also, as new information on fish populations become available each year from fisheries or scientific surveys, estimates of previous years may change retrospectively. Therefore, estimates are sensitive and subject to change due to potential immigration and emigration, as well as the possibility of ageing errors. There are discrepancies in trends between the models and NMFS surveys (Fig. 4.26) and the small-mesh trawl surveys conducted by the Alaska Department of Fish and Game (ADF&G) (Anderson and Piatt, 1999), especially for pollock and cod. For example, from the 1980s – 1990s, gadid biomass fell according to the modeled estimates, while the ADF&G small-mesh survey shows an increasing trend. These incongruities are best explained by differences in sampling gear, region, and depths sampled. The smallmesh trawl surveys primarily sample some nearshore areas in the central and western GOA (150–163°W) and the smaller mesh size (32 mm) and high opening is designed to catch shrimp and juvenile fishes. Since the 1970s, fish larvae (ichthyoplankton) dynamics have been monitored in the GOA (145–165°W, Shelikof vicinity) by the Recruitment Processes Program of the AFSC. Data are collected in oblique tows with a 60-cm bongo net (333 and 505-µm mesh) (Matarese et al., 2003). Most of this sampling was conducted in May. Results from these surveys document year-to-year changes in larval abundance. The relationship between climate forcing factors and biological responses (see Table 4.1) were investigated with linear and nonlinear methods. The climate forcing factors were temperature and salinity data from the GAK 1 time series in Resurrection
302
Long-Term Ecological Change in the Northern Gulf of Alaska
Figure 4.26: Biomass estimates of some groundfishes by model (columns) and survey (triangles) compared to fisheries catches (red lines; on left vertical axis).
Long-Term Change 303
Bay (http://www.ims.uaf.edu/gak1/), National Centers for Environmental Prediction reanalysis of surface temperature and the Pacific Decadal Oscillation (ftp://ftp.atmos. washington.edu/mantua/pnw_impacts/INDICES/PDO.latest). Wind-mixing data were provided by Nick Bond (pers. comm.; http://www.pfeg.noaa.gov: 16080/ products/las/docs/), and the retention index was estimated by Mick Spillane (pers. comm.) from a semispectral primitive equation model (SPEM) of Shelikof Strait (http://www.pmel.noaa.gov/foci/spem-ibm.html).
4.6.2. Walleye Pollock Walleye pollock (Theragra chalcogramma) are widely distributed in the Pacific Ocean north of California. Adult pollock form schools in the open ocean and are close to the bottom over the continental ocean shelf. They reach maturity at around 4 years and live up to 15 years. For the last two decades, it has been one of the most abundant groundfish in the GOA and supported a large commercial fishery since the early 1970s (Megrey, 1989). The GOA biomass of adults (3+ years) was estimated to be 500,000 metric tons (t) during the 1960s and early 1970s, and the model shows a steady, strong increase until the population peaked in 1982 at just below 4 million t (Fig. 4.26(A)). The rapid decrease following 1982 was interrupted in the early 1990s before it continued downward to a minimum value of 570,000 t in 2000–2001. This is comparable to the values recorded in the 1960s. Since 2001, the biomass has increased again due to a strong recruitment in the 1999 year-class (Livingston, 2003), and the current estimate is just below 1 million t. Although triennial survey data indicate that walleye pollock were fairly steady from 1984 to 1999 at around 750,000 t, they exhibited a significant decrease over time (r2 = 0.54) due to the low estimates of the last two surveys. Traditionally pollock were most abundant around Kodiak Island, but survey catches in recent years have increased west of the Shumagin Islands along the Alaska Peninsula (Fig. 4.25). In the GOA, pollock spawn in Shelikof Strait mainly in late March and early April, with females producing up to 1.2 million eggs (Matarese et al., 2003), and their larvae are more abundant than any other species in the spring (Matarese et al., 2003). Larval abundance is highly variable from year to year, although average catches in May stayed typically below 1000 larvae per 10 m2, except for the extremely high abundances observed in 1981 (coinciding with maximum adult pollock biomass) and 1996 (Fig. 4.27).
4.6.3. Pacific Cod Pacific cod (Gadus macrocephalus) is a benthic gadid living along the continental shelf. This species matures at 2–3 yrs and lives as long as 13 years. Cod show significant
304
Long-Term Ecological Change in the Northern Gulf of Alaska
Figure 4.27: Larval abundance of selected groundfish species in the Gulf of Alaska.
Long-Term Change 305
migration between the GOA and Bering Sea (Shimada and Kimura, 1994), and genetic studies have been unable to separate them into distinct stocks (Grant et al., 1987). According to model estimates, biomass of Pacific cod showed a slight increase through the early 1990s (from 500,000 t in 1978 to 860,000 t), followed by a gradual decline (540,000 t in 2003) (Fig. 4.26(B)). Similar to walleye pollock, the survey biomass estimates have been consistently below the modeled estimates, but they also show a decreasing trend since 1984 (r2 = 0.52). Between 1984 and 1996, biomass values were above 400,000 t and below 300,000 t since 1999. Pacific cod are winter–spring spawners, and their larvae are most abundant in April and May. Females produce up to 3 million demersal eggs. Highest larval abundance is found west of Kodiak Island along the Alaska Peninsula in spring (Matarese et al., 2003). There was a stepwise increase of larval abundance from the 1980s (7 larvae per 10 m2) to the values after 1989 (31 larvae per 10 m2) (Runs test, p = 0.001; Fig. 4.27) despite the decreasing trend of adult biomass.
4.6.4. Arrowtooth Flounder Arrowtooth flounder (Atheresthes stomias) range from the Bering Sea to central California. They are commonly found along the continental shelf and slope and prefer soft muddy bottoms. They mature at approximately 3–5 years of age (Zimmermann and Goddard, 1996; Zimmermann, 1997) and may live as long as 23 years (Eschmeyer et al., 1983). Adult biomass was estimated to be around 335,000 t between 1961 and 1971. Subsequently, the population in the GOA steadily increased to 2,400,000 t in 2003 (Fig. 4.26(c). It is currently the most abundant groundfish in the GOA. Survey estimates of arrowtooth flounder biomass show an increasing trend over time, which is significant for the eastern (r2 = 0.70) and western (r2 = 0.51) but not the central GOA. Spawning takes place after September in the GOA, and larvae are found from January until June with highest abundance along the shelf edge (Matarese et al., 2003). Larval abundance is also clearly higher in the 1990s as compared to the 1980s (Fig. 4.27), although recent years have shown a decrease.
4.6.5. Pacific Ocean Perch Pacific ocean perch (Sebastes alutus) is the most abundant rockfish along the continental slope from California to the Bering Sea. They are very slow-growing, mature at approximately 7 or 8 yrs of age, and may live more than 90 years (Leaman, 1991). Biomass estimates were high in the early 1960s with a peak of 1,150,000 t in 1963. Afterwards, biomass dramatically declined to about 70,000 t in the late 1970s and early 1980s. Since the mid-1980s, the population biomass has slowly recovered and
306
Long-Term Ecological Change in the Northern Gulf of Alaska
is currently estimated at around 290,000 t (Fig. 4.26(D)). Estimates from the triennial survey reflect the recovery since the mid-1980s (r2 = 0.52), and are above the modeled estimates. Unusually large survey catches, especially of the 1996–2001 surveys, have made these estimates highly uncertain and may indicate that the annual biomass estimates from the model are imprecise (Hanselman et al., 2003). Spawning occurs in early winter, and species of the genus Sebastes are livebearers. Larvae are released in April–May, but can rarely be identified to species except by molecular analysis (Gharrett et al., 2001), so species-specific larval abundance estimates are not available.
4.6.6. Explaining Population Change There have been three major shifts in population biomasses among key species in the GOA: Pacific ocean perch reached their peak biomass in the mid-1960s and declined rapidly thereafter; gadids (walleye pollock and Pacific cod) were the dominant biomass in the GOA throughout most of the 1980s; and the biomass of arrowtooth flounder, and to a lesser extent of other flatfishes has continually increased since the mid-1970s, recently (early 1990s) surpassing walleye pollock as the dominant fish biomass in the region. Changes in population biomass in the GOA may be due to physical factors, such as climate forcing, biological factors (such as densitydependent competition and predation), fishing mortality, or, more likely, to some combination of these effects, influencing each species at particular life history stages.
Climate Forcing The North Pacific Ocean appears to oscillate between warm and cold regimes on a multidecadal timescale (Hare and Mantua, 2000). The shift in 1976–1977 from a cold to a warm regime is believed to be the cause for the extensive restructuring of the marine community in the GOA, and in particular, to the rise of groundfish and flatfish stocks over forage fishes and shrimps (Anderson and Piatt, 1999). Temperature could directly affect adult or larval mortality. It is also possible that the dramatic changes in the marine community of the northern GOA in the mid-1970s are caused by interaction of increased precipitation, stronger wind, and higher temperatures that are manifested through changes in the distribution and abundance of planktonic production and mediated through the food web. The data presented here suggest little relation between immediate shifts in adult groundfish biomass and climate indices. Such a result was not entirely unexpected, and it seems unlikely that mortality of adult fishes would be directly impacted by shifts in the abiotic environment, given that adult fish are able to compensate for suboptimal climate conditions by altering their metabolism and/or behavior (Bryan et al., 1990; Thurston and Gehrke, 1993). Populations of long-lived
Long-Term Change 307
organisms are buffered from or can lag dramatic climatic changes to the extent their populations integrate over longer periods of time than short-lived animals. Shifts in biomass of adult fishes may be more closely tied to shifts in distributional abundance since adult fish are highly mobile and able to follow optimal clines in temperature, salinity, or food availability. Juvenile mortality is also less likely to be directly impacted by abiotic stresses for many of the same reasons that adult mortality is largely unaffected. Larval stages, however, are far more responsive to shifts in the physical environment and, in fact, increases in recruitment in several groundfishes were observed after the regime shift (Hollowed et al., 2001). Larvae are underdeveloped at hatching; they are weak swimmers that depend on favorable water currents to transport and retain them in suitable nursery areas. Furthermore, larvae have very high energy needs for their size and no energy reserves (Brett and Groves, 1979), so they are at increased risk of starvation compared to adult and juvenile stages, making their survival particularly vulnerable to even subtle physical changes i.e., water temperature; Houde and Zastrow, 1993), salinity (Ottesen and Bolla, 1998), turbulence (Bailey and Macklin, 1994), or advection to unsuitable nursery areas (Bailey and Picquelle, 2002). Therefore, larvae are probably the most vulnerable life stage to direct impacts of climate variations, while mortality among juvenile and adult stages are probablly modulated more by indirect biotic effects (see the following text). However, there are many climate factors (Table 4.1) that may confound discovering simple relationships (see the section titled “Biological controls”). Of the examined species, walleye pollock larval abundance was positively correlated with higher retention (years of high eddy occurrence on the continental shelf). Temperature and wind mixing negatively affected abundance of Pacific cod larvae; however, recruitment had a positive relationship to temperature. Similarly, arrowtooth flounder had increased larvae abundances in years of high retention, although recruitment was negatively affected by that relationship (Table 4.1). These examples demonstrate that total larval abundance is not
Table 4.1: Impact of environmental variables on larval abundance and recruitment. Values represent significant correlation coefficients. Walleye pollock
Pacific cod
Arrowtooth flounder
Larvae Recruitment Larvae Recruitment Larvae Recruitment Temperature Salinity Wind mixing Retention index
n.s. n.s. n.s. 0.64
n.s. 0.62 n.s. n.s.
−0.77 0.76 −0.67 n.s.
0.68 n.s. n.s. −0.70
−0.44 n.s. n.s. 0.82
n.s. −0.46 n.s. −0.55
308
Long-Term Ecological Change in the Northern Gulf of Alaska
connected to survival but that both, abundance and survival can be directly or indirectly affected by climate variables.
Biological Controls There are can be three major categories of biological controls: parasites/disease, competition, and predation. Parasites and diseases negatively affect fitness. Even closely related species, for example northern and southern rock sole in the GOA (Lepidopsetta polyxystra and L. bilineata, respectively), can have very different parasite infection rates, and parasites may significantly reduce fish weight (Zimmermann et al., 2001). Often, parasite abundance and richness correlates positively with temperature (Poulin and Rohde, 1997) and, thus, should have a larger impact on some fish species within the community during warmer regimes. Density-dependent mortality, potentially induced by changes in physical factors, probably contributes to most of the fluctuation in population size among juveniles and adults (Bailey, 2000). For example, food limitation produced by climate shifts may lead to heightened inter- and intra-specific competition among juveniles and adults. Such a mechanism is suggested by the lower-than-average weight of pollock during strong year-classes, indicating competition for food and/or space at advanced stages (Fig. 4.28). In exceptional cases, it has been shown that larval mortality can also be locally enhanced due to food competition (Duffy-Anderson et al., 2002). Predation on older fish is another important cause of population change. Juvenile and adult fishes are eaten by other adult groundfish, marine birds, and mammals
Figure 4.28: Mean weight of age-4 pollock in the Shelikof Strait echo integration–trawl survey. Strong year-classes are indicated by the large red symbols. Adapted from Dorn et al. (2003).
Long-Term Change 309
(Brodeur and Wilson, 1996). The control of pollock recruitment to the adult population changed from the larval to the juvenile stage after the 1976–1977 regime change (Bailey, 2000), an excellent example of the effects of predation. This shift in control of recruitment from the larval to the juvenile stage is presumably due to increased predation by large numbers of predatory fishes on juvenile pollock (see Chapter 6 for a discussion of top-down controls of fish recruitment). Predators also eat large numbers of eggs (Brodeur et al., 1996) and larvae (Fancett and Jenkins, 1988; Bailey and Houde, 1989), though it is generally believed that survival among larvae is primarily controlled by density-independent factors (Bailey et al., 1996). According to the oscillating control hypothesis (OCH) (Hunt et al., 2002), a close connection between climate and biological controls exists in the North Pacific. A warm regime is believed to favor high zooplankton production, providing a plentiful food source for larvae and juvenile fishes. Such a scenario would support strong recruitment to the adult population of predatory fishes (e.g., walleye pollock and arrowtooth flounder), which in turn exerts a strong top-down regulation (predation) on smaller fish (e.g., forage species and their own juveniles). Further, the OCH predicts that a cold regime will lead to bottom-up regulation. Low temperatures will limit zooplankton production, enhancing competition and reducing larval and juvenile fish survival. The hypothesis predicts that the adult piscivorous fish community would decline, releasing small fish species from predation pressure. Adult piscivorous fish are also hypothesized to more susceptible to fishing pressure during cold climate regimes. (Also see the discussion on OCH in Chapter 6).
Fisheries Effects Climate shifts and their indirect repercussions in the food web certainly have contributed to observed changes in fish biomass in the GOA due to shifts in recruitment (Hollowed et al., 2001), but the effects of fishing should not be overlooked. In contrast to climate forcing, fisheries have little or no direct effect on larvae or juvenile mortality. Also species of low commercial value such as arrowtooth flounder (Greene and Babbitt, 1990; Porter et al., 1993) are only minimally impacted by fishing efforts (i.e., bycatch). However, fisheries can significantly contribute to mortality of adult fish populations, especially species with low growth rates and an advanced age at maturity (sexual reproduction). The fishery for Pacific ocean perch provides a prime example. In the 1960s, up to 35% of the Pacific ocean perch biomass was caught on an annual basis, and it is largely believed that this overfishing was responsible for the precipitous crash of the Pacific ocean perch population in the GOA in the late 1960s and early 1970s (Kramer and O’Connell, 1995; Hanselman et al., 2003). Quotas regulating fisheries activity started in 1986, and management efforts seem to be successful in aiding species recovery (Fig. 4.26(D)). It is interesting to speculate that the rise in gadid populations in the 1980s may have been precipitated not by climate
310
Long-Term Ecological Change in the Northern Gulf of Alaska
shifts alone, but also by a release from competition for food due to the diminished numbers of Pacific ocean perch. We caution, however, that the effects of fisheries on population biomass in general are somewhat obscure due to the confounding effects of directed fishing effort, and to the superimposed effects of climate variation.
Summary Modeled long-term dynamics and sampling data of several groundfish species in the Gulf of Alaska were analyzed. Pacific ocean perch (Sebastes alutus) dominated in the 1960s, the gadid walleye pollock (Theragra chalcogramma) in the 1980s, and arrowtooth flounder (Atheresthes stomias) is presently the most abundant species in the fish community. Although inter-decadal changes in the groundfish community in the Gulf of Alaska are well documented, the exact causes remain elusive. Simultaneously varying forcing factors obscure the relationship between forces of change (climate, competition, predation, fishery, and parasites or disease) and groundfish population dynamics. Most likely, the major driving force for fish larvae and its food supply are climatic factors; for juvenile fishes, biological factors, and for the adult populations; biological factors and fishing pressure. Climate change, indirect biological controls, and fishing mortality exert considerable influence. Disease and parasities can also contribute to population fluctuation. It seems most likely that these and other factors act in concert over all life history stages to affect population abundance of groundfishes in the GOA. We found evidence that fishing in the 1960s and 1970s caused the decline of Pacific ocean perch. Biological and climate forcings combined with a high population doubling time probably delayed the recovery of the population. The population dynamics of gadids are complex, and top-down and bottom-up factors seem to alternate as the driving force. Favorable climate possibly contributed to the increase in biomass of walleye pollock in the 1980s, and biological top-down factors may have led to the decline. Arrowtooth flounder is only minimally impacted by fisheries, and we expect that biological forcings – such as release from competition and predation by gadids- and climate forcings dominate population dynamics. The GOA is an enormously complex and dynamic ecosystem, and continued effort directed toward long-term monitoring is essential to credibly forecast future population trends.
Acknowledgements We would like to thank Sandra Lowe, Martin Dorn, Dana Hanselman, Grant Thompson, Jack Turnock, Jeff Napp, Kevin Bailey, and Mark Wilkins for providing data and constructive comments. This contribution is partially funded by the Joint
Long-Term Change 311
Institute for the Study of the Atmosphere and Ocean under NOAA Cooperative Agreement No. NA178RG1232.
4.7. Seabirds in the Gulf of Alaska Alan M. Springer 4.7.1. Introduction Approximately 8 million seabirds of 26 species nest at some 800 colonies around the rim of the GOA, and an additional 10 million birds and 2 species nest in the Aleutian Islands (Table 4.2 and Fig. 4.29). Clearly, it is not possible to monitor population trends of all species or of any individual species at all locations where they are found. Many nest underground, are nocturnal, or both, and are especially difficult to study. Others nest away from the coast or in forests and are similarly inaccessible. Species about which we know the most in the GOA are those that are abundant, widespread, and conspicuous. These are common murres (Uria aalge) and blacklegged kittiwakes (Rissa tridactyla), and, to a lesser extent, tufted puffins (Fratercula cirrhata). (Fig. 4.30). On the order of 600,000 murres and 700,000 kittiwakes are found in the northern Gulf of Alaska, with nesting colonies located on exposed cliff faces and bluffs in numerous locations. As a result, more is known about these two species, and their closely related congeners, thick-billed murres (Uria lomvia) and redlegged kittiwakes (Rissa brevirostris), in the Aleutians and elsewhere than about any other species of seabirds in Alaska. Tufted puffins score highly on the first two criteria – about 1,000,000 birds are widely distributed among many colonies in the GOA, but they nest in burrows and crevices and thus are more difficult to study than murres and kittiwakes. Still, some things are known about them, and because their strategies for survival differ from murres and kittiwakes in several important ways, they are included here as focal species for what they may tell us about ecosystem change in the GOA. Although less is known overall about most other species, we can learn from then about the nature of change in the GOA. Among these other species are Leach’s and fork-tailed storm petrels (Oceanodroma furcata and O. leucorhoa), marbled and Kittlitz’s murrelets (Brachyramphus marmoratus and B. brevirostris), and pelagic and red-faced cormorants (Phalacrocorax pelagicus and P. urile) (Fig. 4.31). Murres and kittiwakes are conspicuous and comparatively easy to count, yet care must be taken to count in ways that allow comparisons between years and locations (Hatch and Hatch, 1988, 1989; Byrd, 1989). This is especially important for murres, which have attendance patterns at colonies that vary throughout the day, between days, and between stages of the nesting cycle, and which often nest in such huge aggregations that counting all of them is not feasible. Therefore, monitoring protocols
312
Long-Term Ecological Change in the Northern Gulf of Alaska
Table 4.2: Estimated abundances (in 1000s) of nesting seabirds in the Gulf of Alaska and Aleutian Islands. Data from U.S. Fish and Wildlife Service, seabird colony database: marbled murrelet in Gulf of Alaska from Piatt and Ford (1993).
English Name
Scientific Name
Northern fulmar Fork-tailed storm petrel Leach’s storm petrel Double-crested cormorant Brandt’s cormorant Pelagic cormorant Red-faced cormorant Unidentified cormorant Mew gull Herring gull Glaucous-winged gull Black-legged kittiwake Red-legged kittiwake Arctic tern Aleutian tern Unidentified tern Common murre* Thick-billed murre Unidentified murre1 Pigeon guillemot Marbled murrelet Kittlitz’s murrelet Ancient murrelet Cassin’s auklet Parakeet auklet Least auklet Whiskered auklet Crested auklet Rhinoceros auklet Tufted puffin Horned puffin Total
Fulmarus glacialis Oceanodroma furcata Oceanodroma leucorhoa Phalacrocorax auritus Phalacrocorax penicillatus Phalacrocorax pelagicus Phalacrocorax urile Phalacrocorax spp. Larus canus Larus argentatus Larus glauscescens Rissa tridactyla Rissa brevirostris Sterna paradisaea Sterna aleutica Sterna spp. Uria aalge Uria lomvia Uria spp. Cepphus columba Brachyramphus marmoratus Brachyramphus brevirostris Synthliboramphus antiquum Ptychoramphus aleuticus Cerorhinca monocerata Aethia pusilla Aethia pygmaea Aethia cristatella Cyclorrhynchus psittacula Lunda cirrhata Fratercula corniculata
*Essentially all common murres.
Gulf of Alaska
Aleutian Islands
440 640 1067 3.3 0.086 21 20 15 15 1 185 675 − 8.9 9.4 1.7 589 55 1197 24 200 + 164 355 58 0.02 − 46 170 1093 773 7826
510 2354 2483 1.2 − 6 26 9.1 − − 57 60 13 0.28 0.41 0.05 43 109 66 15 + + 54 118 86 2278 6.5 873 0.03 1267 91 10,527
Long-Term Change 313
Figure 4.29: Locations of seabird colonies in the GOA and elsewhere in Alaska that are discussed in the text.
were developed that improved the accuracy and precision of census data. These protocols specify time in the nesting season and time of day and involve replicate counts of individuals on well-defined census plots within colonies over several days. However, in the formative years of seabird studies in Alaska, such protocols were not in place and censuses were often conducted in ways and at times that rendered them unreliable for comparison to later counts. For example, murres on cliffs, in the air, and on the water were sometimes counted and added together as the estimate of abundance for a colony. Many of those counts were adjusted (commonly doubled) to include estimated numbers of birds at sea. Counts were sometimes done early in the nesting season when attendance at colonies is particularly erratic from day to day. As a result, we have very little reliable historic data on the abundance of murres and kittiwakes with which to evaluate longer-term trends in populations. In this chapter, we have not used early information unless it is known to be comparable to standardized counts of later years.
4.7.2. Long-term Changes in Common Murres Although most murres in the GOA are common murres, thick-billed murres also nest there in small numbers at several colonies, and the two species are not differentiated in most colony counts. Therefore, in the following accounts, unless noted otherwise,
314
Long-Term Ecological Change in the Northern Gulf of Alaska
Figure 4.30: Black-legged kittiwakes (top, photograph courtesy of J. Schoen), common murres (middle, photograph courtesy of J. Schoen), and tufted puffins (bottom, photograph courtesy of A. Kettle, USFWS).
Long-Term Change 315
Figure 4.31: Leach’s storm petrel (upper left, photograph courtesy of N. Konyukhov), forked-tail storm petrel (upper right, photograph courtesy of N. Konyukhov), marbled murrelet (middle left, photograph courtesy of Gus Van Vliet), Kittlitz’s murrelet (photograph courtesy of Gus Van Vliet), common cormorant (lower left, photograph courtesy of N. Konyukhov), and red-faced cormorant (lower right, photograph courtesy of N. Konyukhov).
316
Long-Term Ecological Change in the Northern Gulf of Alaska
Figure 4.32: Population trends of murres at colonies in the GOA. Data from U.S. Fish and Wildlife Service (unpublished data), Dragoo et al. (2004), and Hatch (2003).
the term murre is generic and refers to mixtures of mainly common murres with some thick-billed murres. The histories of some murre populations are known in general where counts have been made for comparatively long intervals, but none is known in great detail, particularly in the earlier portions of the records when gaps of many years typically separate successive counts. Histories of several other populations are known in detail but over short, recent intervals. The longest time series comes from Middleton I. (Fig. 4.32), where the first estimate of numbers was made in 1956 (Rausch, 1958). This history was recounted earlier in Section 3.3, but briefly, the population grew about fifteen-fold from 1956 to the early 1970s, beginning presumably after the Good Friday earthquake of 1964 created extensive new nesting habitat. Numbers fluctuated considerably for nearly 20 years from the early 1970s to the late 1980s, but declined by about 80% in the following decade (Fig. 4.32, Table 4.3). The next longest time series began in the early to mid-1970s at five colonies in a variety of oceanographic settings. In lower Cook Inlet, murres at Chisik I. have undergone a long, steady decline of about 90%, while increasing by at least eight-fold just across the inlet at Gull I. in Kachemak Bay. Murres declined by 75% at Puale Bay on the mainland coast of the Alaska Peninsula beginning perhaps as early as the mid1970s or before. However, given the high interannual variability that can occur in murre abundance at any given colony and the long intervals between the first three counts (1976, 1981, 1989), the exact time when the decline began cannot be determined. Abundance during the 1990s at Puale Bay was stable. As we will see, getting the correct timing of the start of long population trends is crucial to an understanding
Table 3: Population trajectories of kittiwakes, murres, and puffins at colonies in the Gulf of Alaska and Bering Sea. Data from Dragoo et al. (2003) and USFWS seabird colony database. Interval 1956, 1960, 1972, 1973, 1974–1977 Colony
Aleutian Is. Buldir I.c Aiktak I. Bogoslof I.d
− +
0
− +
0
1989–1998
Murres Puffins Kittiwakes − + − + +
+ −
− − + 0 + + +
+ +
+
Murres Puffins Kittiwakes − − 0 + + + − − + 0
0 0
− − − − 0 0 + 0
0 + + +
0 + Continued
Long-Term Change 317
GOA Chisik I.a Middleton I.b Puale Bay Chowiet I. Gull I. Chiniak Bay PWS Barren Is. Chiswell Is. St. Lazaria I.
Murres Puffins Kittiwakes
1977–1989
318
Interval 1956, 1960, 1972, 1973, 1974–1977 Colony
Murres Puffins Kittiwakes
Bering Sea St. George I. (COMU) St. George I. (TBMU) St. Paul I. (COMU) St. Paul I. (TBMU) Cape Peirce Bluff Chukchi Sea Cape Thompsone Cape Lisburne a
1972. 1956. c 1974. d 1973. e 1960. b
−
−
1977–1989
1989–1998
Murres Puffins Kittiwakes 0 − − −
−
0 − +
Murres Puffins Kittiwakes +
+
0 + − − − 0
+ 0
+ +
− +
−
− − −
Long-Term Ecological Change in the Northern Gulf of Alaska
Table 3:—cont’d.
Long-Term Change 319
of cause and effect. At Chowiet I., in the Semidi Is. group, the number of murres was variable but mostly stable in the mid-1970s to early 1980s but increased by about 40% from then through the late 1990s. Over more recent intervals, murre numbers at the Barren Is. (East Amatuli I., Nord I., and Light Rock) increased during the 1990s. Numbers at the Chiswell Is. near the mouth of Resurrection Bay were stable in 1989–1992 and declined sometime between then and the next count in 1998. At St. Lazaria I. in southeast Alaska, murres declined somewhat from 1995 through at least 2000. Thus, no overall, broadscale trends in murre abundance in the GOA can be identified from the reliable information at hand. No regional patterns are apparent either, although the sample size is small and not sufficient for a rigorous analysis. Rather, murres have increased, decreased, and remained stable over various intervals at the colonies that have been monitored. With this said, it must be pointed out that a previous analysis of change in murre abundance in the GOA reached a different conclusion by using counts at these and many additional colonies over the years (Piatt and Anderson, 1996). In their analysis, Piatt and Anderson found evidence of decline at 15 of 16 colonies counted at least twice from before the Exxon Valdez oil spill and at least once afterward through 1994. The overall decline was greater at colonies outside the spill path than in the path, suggesting that factors other than the oil spill were responsible for the apparent declines. Elsewhere to the west in the Aleutian Islands and Bering Sea, there are similar mixed trends. In the Aleutians, murres have increased conspicuously at Buldir I. in the western Aleutians since the early 1970s, and they increased by a factor of about four at Koniuji I. in the central Aleutians since the early 1990s (Dragoo et al., 2004). At nearby Kasatochi I., there was no trend from 1980 to 1997, although numbers varied by a factor of up to two between years. However, murres disappeared from Kasatochi I. in a matter of just 4 years, falling from over 2000 in 1997 to 50 in 1998 and 0 in 2001. At a third nearby colony, Ulak I., numbers were variable between years but without trend from 1997–2004. In the eastern Aleutians at Aiktak I., abundance has been extremely variable since 1980, but also without a convincing long-term trend. Numbers have apparently declined since the mid-1990s. On the Pribilof Islands, both species of murres declined steadily on St. Paul I. after the mid-1970s. Thick-billed murres declined on St. George I., but only until the mid-1980s, after which time they returned to their former abundance. Numbers of common murres at St. George I. have not changed and have exhibited little interannual variability as well. At coastal colonies in the eastern Bering Sea, murres at Bluff declined during the late 1970s (the first count was in 1975) but increased slightly since then, and have fluctuated and declined somewhat in Bristol Bay since 1990, when the first count was made there. Even to the north in the Chukchi Sea there is no overall trend, with murres at Cape Thompson declining from the mid-1970s through the mid-1980s and then
320
Long-Term Ecological Change in the Northern Gulf of Alaska
increasing in the following decade, whereas just 100 km north at Cape Lisburne they increased steadily over this interval, approximately doubling in abundance. At Cape Thompson, the abundance of murres in 1976 was approximately 40% lower than when they were first systematically counted in 1960–1961 (Murphy et al., 1980).
4.7.3. Long-term Changes in Tufted Puffins Because of the habit of puffins to nest underground, censuses are typically conducted along transect lines in areas of nesting habitat and numbers of burrows are recorded during painstaking searches in usually thick vegetation, or they are counted at sea. As a result, systematically collected data on trends are available for only three locations in the GOA, and a fourth at Aiktak I. in the far eastern Aleutians (Fig. 4.33, Table 4.3). At St. Lazaria I. in southeast Alaska, six counts between 1994–2001 were variable but without trend; in PWS, a pelagic count in 1972 was nearly twice as high as any of the six counts between 1989 and 1998, which apparently increased each year between 1989 and 1991 but have been variable and without trend since; at east Amatuli I. (Barren Is.), eight counts between 1995 and 2002 were stable or slightly decreasing; and at Aiktak I., 11 counts between 1989 and 2002 showed a slight increase (Dragoo et al., 2004). Thus, based on this information, in the past 10–15 years, numbers of tufted puffins in the GOA have been essentially stable or slightly increasing. 20000 10000 Prince William Sound
INDEX VALUE
1000 100
E. Amatuli Is.
10 St. Lazaria Is.(blue) 1 Aiktak Is. 1970 1972 1974 1976 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002
0.1
YEAR
Figure 4.33: Population trends of tufted puffins at colonies in the GOA. Data from U.S. Fish and Wildlife Service (unpublished data) and Dragoo et al. (2004).
Long-Term Change 321
The significance of the higher count in PWS in 1972 is difficult to evaluate, as the survey methodology employed that year differed from the standardized protocol used after 1989 (Klosiewski and Lang, 1994). Tufted puffins in the Aleutian Is. have been increasing notably throughout the archipelago (Dragoo et al., 2004). In contrast, puffins from British Columbia to central California have in general declined over intervals spanning the past two to three decades (Piatt and Kitaysky, 2002). Thus, for tufted puffins there are conspicuous, broad patterns of population change, with abundance increasing notably throughout the Aleutian Islands, perhaps slightly increasing in the GOA in the past 15 years, and decreasing at colonies in British Columbia and southward through California.
4.7.4. Long-term Changes in Black-legged Kittiwakes As with murres, there are no compelling broadscale trends in the abundance of kittiwakes at colonies in the GOA. ●
●
●
●
●
●
At Middleton I., kittiwake abundance paralleled that of murres in the early years, where they both increased dramatically after 1956 (Fig. 4.34, Table 4.3). By the early 1970s, Middleton I. had become the largest kittiwake colony in the GOA, with some 70,000 nesting pairs. It grew even further during the next 10 years to over 80,000 pairs, but then began a steep, steady decline in the early 1980s to its current size of about 10,000 pairs (see also Box 3.1). In lower Cook Inlet, kittiwakes, as with murres, have experienced a sustained decline since the early 1970s at Chisik I., but increased considerably between the mid-1970s and mid-1980s nearby at Gull I. Systematic counts of kittiwakes at the Barren Is. began in 1993 and reveal no apparent change in abundance through the last count in 1999. Kittiwakes might have declined in the late 1990s at Puale Bay on the Alaska Peninsula, but counts have been infrequent and are difficult to interpret. At Chowiet I., where murres increased steadily over the past 25 years, the kittiwake population declined abruptly between 1990 and 1991 – the nesting population was stable before then and has been since. Total kittiwake abundance at the 23 distinct colonies in Chiniak Bay on Kodiak I. has been variable between years since 1975, but exhibits no compelling long-term trend. The total number of kittiwakes in PWS in the early 1980s was similar to a single count in 1972, but has increased overall since then despite major variability at the many individual colonies, e.g., some have been founded and grown dramatically, some have declined and disappeared, etc.
Black-legged kittiwakes in the Aleutian Is. have increased notably since the early 1970s at the two colonies with the longest-term records, Bogoslof I. and Buldir I. at
322
Long-Term Ecological Change in the Northern Gulf of Alaska
Figure 4.34: Population trends of black-legged kittiwakes at colonies in the GOA. Data from U.S. Fish and Wildlife Service (unpublished data), Dragoo et al. (2004), and Hatch et al. (1993).
Long-Term Change 323
either end of the chain (Dragoo et al., 2004). The increase was dramatic at Buldir I. between the mid-1970s and late 1980s and has been stable since. At Koniuji I., kittiwakes were stable from the early 1980s through the late 1990s and then might have declined sometime between 1998 and 2001: however, with only one count indicating a decline (2001) this is not certain. Trends in numbers of kittiwakes generally parallel those of murres on the Pribilofs also, with a large, steady decline at St. Paul from the mid-1970s through the late 1990s, and a similar decline through the late 1980s followed by recovery at St. George. Patterns of change in abundance of red-legged kittiwakes in the Aleutians and Pribilofs have mirrored those of black-legged kittiwakes. Red-legged kittiwakes nest only in the Aleutians and Pribilofs. Black-legged kittiwake abundances at coastal colonies in the eastern Bering Sea have been similar to those of murres – slight increase in Norton Sound in the 1980s and significant decline during the 1990s at Cape Peirce in Bristol Bay. In the Chukchi Sea, they increased at both Cape Thompson and Cape Lisburne since the mid-1970s. Thus, in terms of larger-scale pattern, the Aleutian Islands stand out as the only region where all three species, murres, puffins, and kittiwakes, show generally consistent change.
4.7.5. Long-term Changes in Other Species Trends in abundance of several other species have been monitored in the GOA and elsewhere over various intervals and together further inform us about the nature of change in seabird populations and in the ecosystem.
Storm Petrels Storm petrels are difficult to count because they nest underground and are nocturnal. They have been monitored at just four colonies in Alaska, and the two species (Leach’s and fork-tailed) have generally not been differentiated in the counts. Storm petrels have been increasing sharply at Aiktak I. (extreme eastern Aleutians) since 1990, St. Lazaria I. (southeast Alaska) since 1994, and east Amatuli I. (Barren Is.) since 1998, the years they were first counted at those sites. There has been no trend since 1975 at Buldir I. in the western Aleutians (Dragoo et al., 2004). Cormorants Cormorants also are difficult to monitor because they commonly move nesting sites within colonies, such that numbers on specific census plots can vary without real change in the colony as a whole. And, as is the case for the other species, interannual variability can be high and there are occasional long gaps between successive population counts. With this in mind, there still appears to have been a pervasive decline
324
Long-Term Ecological Change in the Northern Gulf of Alaska
in the abundance of both pelagic and red-faced cormorants (Dragoo et al., 2004). The pattern and timing of declines vary substantially from one colony to another, however. For example, at Middleton I. pelagic cormorants increased from the early 1970s through 1990, but then collapsed by about 80% by 1992: they have recovered somewhat since. At Chowiet I., pelagic cormorants apparently collapsed by 90% between 1979 and 1986 and have remained low since.
Murrelets Marbled and Kittlitz’s murrelets do not aggregate in conspicuous colonies in summer but nest solitarily or in loose associations and in situations that make counting nesting pairs impossible. Marbled murrelets typically nest high in the boughs of mature conifers in old growth forests, although in parts of their range where there are no trees they will also nest on the ground. Kittlitz’s murrelets nest on the ground, typically in scree fields in high alpine areas. Both species often nest many kilometers from the coast, and population trends are thus based on counts of birds on the water where they forage. Numbers of both species have been monitored in three locations in the GOA over various intervals. The longest time series exists for PWS, where murrelets and other marine birds were first censused by pelagic surveys in spring and summer of 1972 and spring of 1973 (Haddock et al., unpublished data, Agler et al., 1999; Stephensen et al., 2001; Kuletz, 2005). Those counts were not repeated until 1989, but have been undertaken in 8 years since. Although survey designs differed between the 1970s and later surveys, modeling exercises indicated that numbers of marbled murrelets declined 62% since 1989, when surveys were standardized, and 85% since 1972. Kittlitz’s murrelets have declined even more drastically – by 92% since 1989 and 99% since 1972. The trend for Kittlitz’s murrelets in PWS characterizes trends elsewhere – they have declined by over 80% since 1976 in the Kenai Fjords area and by over 60% in Glacier Bay between 1991 and 2000 (Robards et al., 2003; Van Pelt and Piatt, 2003). Gulfwide trends for marbled murrelets are not so uniform – they declined by about 75% in Glacier Bay in the 1990s, but increased dramatically in that decade in the Kenai Fjords after declining during the previous 15 years. It should be noted, however, that the total abundance of marbled murrelets in Kenai Fjords is an order of magnitude smaller than in either PWS or Glacier Bay.
Miscellaneous Species: PWS In addition to counting tufted puffins and murrelets at sea in PWS during the pelagic surveys mentioned above, all other species of seabirds and marine birds (e.g., sea ducks, mergansers, loons, and grebes) were also counted. Several species, notably piscivores (species that prey primarily on fishes), such as red-throated and Pacific loons (Gavia stellata and G. pacifica), cormorants, Bonaparte’s and glaucous-winged gulls (Larus philadelphia and L. glaucescens), black-legged kittiwakes, arctic terns
Long-Term Change 325
(Sterna paradesaea), pigeon guillemots (Cepphus columba), parakeet auklets (Cyclorhynchus psittacula), and horned puffins (Fratercula corniculata), were apparently less numerous, in some cases much less numerous, in 1989 than they were when first counted in 1972 (Agler et al., 1999; Stephensen et al., 2001). Overall, of the 17 taxa classified as piscivorous by Agler et al., 14 declined between 1972 and 1989. Moreover, some species continued to decline through the 1990s. In contrast, several species that feed primarily on benthic invertebrates, such as harlequin ducks (Histrionicus histrionicus) and Barrow’s goldeneyes (Bucephala islandica), did not decline or increased. Overall, five out of eight species classified as nonpiscivores increased. The declines of piscivores between 1972 and 1989 have been cited as additional evidence of the effect of the mid-1970s regime shift on forage fishes and their predators. The continuing declines through the 1990s have been explained as effects of the oil spill, as well as possible continuing food limitation caused by climate change. Interpreting the significance of these pelagic data is not easy. One confounding factor is that the survey methods in 1972 differed substantially from those in subsequent years. Also, of particular note is that for kittiwakes, the decline in abundance estimated from the surveys of birds at sea is opposite to the increase determined from the much more precise counts of birds at the nesting colonies. The pelagic data suggest that kittiwakes declined from nearly 107,000 in 1972 to about 60,000 in 1989–1993 (−45%), and from there they continued to fall to about 28,000 by 2000 (−53%), for an overall change of –76%. In contrast, the colony counts indicate that kittiwake abundance changed little from 1972 to the early 1990s (~16,000 in 1972 and 1989–1993) and increased during the ensuing decade to about 23,000 in the early 2000s (+44%).
4.7.6. Causes of Long-term Change in Seabirds Clearly, seabirds are sensitive to food supply, as demonstrated through aspects of their breeding biology and natural histories, and summer observations at their nesting colonies (see Section 2.5.3). The various species have a range of survival strategies to accommodate fluctuations in prey and maintain robust populations. These strategies are not always up to the challenges of life in the northern Gulf of Alaska, however, as in the case of murres and kittiwakes at Chisik I. in western Cook Inlet, where both species have declined steadily and considerably in the past three decades in association with food limitation (Piatt and Harding, Section 4.8). Food limitation is not the only factor causing population change in seabirds, as predation can be a strong influence on population size. Seabirds have adaptations to thwart predators, but they are not always successful, as in the case at Middleton I. and elsewhere, as we shall see. In order to correctly identify causes of change in ecosystems, one must carefully consider matters of scale in time and space. That is, we must be mindful of when
326
Long-Term Ecological Change in the Northern Gulf of Alaska
change began and where it is occurring. In the GOA and elsewhere in Alaska, we have very little information on seabird biology prior to 1976 and the beginning of the Outer Continental Shelf Environmental Assessment Program (OSCEAP). OCSEAP funded numerous studies of marine ecosystems in Alaska, including seabirds. Thus, substantial long-term records of seabird population abundance and productivity began at most colonies in the mid-1970s. It is sometimes forgotten that trends in seabird populations may have begun even before then. As it happened, a potent change in climate also occurred in the mid-1970s that had large effects on the physical environment and the biota across the North Pacific (Ebbesmeyer et al., 1991; Francis and Hare, 1994; Mantua et al., 1997; Francis et al., 1998). Subsequent events occurred in 1989 and 1998 that also led to conspicuous changes in the ocean (Springer, 1998; Hare and Mantua, 2000; Bond et al., 2003). By extension, these “regime shifts” have often been invoked to explain much, or most, of the significant change at high trophic levels in marine ecosystems of the GOA and elsewhere in the North Pacific (Agler et al., 1999; Anderson and Piatt, 1999; Trites and Donnelly, 2003). In the majority of cases, a particular emphasis has been placed on perceived Gulfwide declines of pelagic forage fishes, and of species that prey upon them, that were precipitated by the regime shift of the mid-1970s. Moreover, the Exxon Valdez oil spill, which occurred in 1989 as well, killed an estimated 250,000 seabirds, primarily murres. But do these regime shifts and the oil spill really explain all change, or just some change? And if just some, which? And what changes really occurred? In the case of seabirds, the roles of climate regime shifts, oil, and predation in causing change in populations must be carefully considered.
Changing Abundance As noted earlier, getting the timing right of just when change began is critical to an understanding of why it began. For seabirds, there are only two colonies in the GOA where we have reliable information on abundance from well before the regime shift of the mid-1970s. At Middleton I., the nesting murre population grew by nearly 15-fold between 1956 and 1974, and an additional 30% between 1974 and 1988. Kittiwakes increased by five- to six-fold in the first interval and by an additional 13% by the early 1980s. The initial increases, at least, occurred apparently in response to new nesting habitat created by the earthquake in 1964 (see Section 3.3). At Chisik I., both species were in rather steep decline through the early to mid-1970s, apparently in response to food shortage that reduced productivity and recruitment and increased emigration (see Piatt and Harding, Section 4.8). We do not know when the declines at Chisik I. actually began – all we know with reasonable certainty is that they were well underway prior to the mid-1970s regime shift. By way of comparison, in the Chukchi Sea, murres at Cape Thompson declined some 40% from 1960 to 1976.
Long-Term Change 327
At colonies where counts did not begin until the mid-1970s, at about the time of the regime shift, and continued through the duration of that regime (officially 1977–1989), the following changes were seen: ● ● ● ● ●
Puale Bay kittiwakes increased, but murres declined; Chowiet I. kittiwakes did not change, but murres increased slightly; Gull I. kittiwakes and murres both increased; Chiniak Bay kittiwakes increased; murres are not present; Prince William Sound kittiwake abundance in 1984 was similar to a single earlier count in 1972 and numbers increased through the remainder of the regime; murres are not present.
Thus, if there was an effect of the mid-1970s regime shift on abundance of seabirds at these colonies with reliable census data from the mid-1970s and before, the effect was generally positive, opposite to the prevailing paradigm. But, still we do not know when the trends began, and it seems ill advised to assume that they began with the first counts. The next regime shift occurred in 1989, and evidence for a possible effect on seabird abundance comes from: ●
●
Middleton I., where murres declined abruptly, although kittiwakes continued a long steady decline from the early 1980s; Chowiet I., where kittiwakes declined abruptly, but murres continued to increase from the early 1980s. At two additional colonies with much less resolution:
●
●
Puale Bay kittiwakes, but not murres, were fewer in the late 1990s than in the late 1980s and early 1990s; murre abundance was stable after 1989; Chiswell Is. murres were fewer in the late 1990s than in the late 1980s and early 1990s.
In both these cases, however, long stretches between successive counts in this interval (1991–1999 at Puale Bay and 1992–1997 at the Chiswell Is.) complicate interpretations of change. Elsewhere: ●
● ●
●
Gull I. kittiwakes did not change, owing to saturated nesting habitat for them, and murres continued to increase from the early 1980s; Chiniak Bay kittiwakes continued to increase from the earlier period; Barren Is. kittiwakes did not change between the first count in 1994 and the last count in 1999, whereas murres increased markedly after counts began in 1989; Prince William Sound (PWS) kittiwakes continued to increase from the earlier period (1972–1989).
328
Long-Term Ecological Change in the Northern Gulf of Alaska
Thus, there is some evidence that the 1989 shift might have had an effect on abundance (two abrupt declines, two other declines that might have been associated with the shift, and one possible inflection point). There has not been sufficient time since 1998 to know if this most recent event was consequential. Contrasting trends in populations of kittiwakes and murres at some colonies, e.g., Puale Bay and Chowiet I., further confuse simple explanations. Although Puale Bay and the Chiswell Is. were in the path of spilled oil, Piatt and Anderson (1996) concluded that data were insufficient to separate possible effects of climate change and oil mortality on murre abundance. Whatever effects oil might have had on murres were localized and ephemeral. Trends prior to the oil spill and other changes in the GOA ecosystem in general, however, supported climate change as the more important factor. Effects of oiling on the abundance of other species were much less, including kittiwakes, which were hardly affected at all. Moreover, there is no evidence that oil had broadscale negative effects on the abundance of forage fishes. Local, temporary effects in PWS might have occurred, especially in the case of sand lance, and these might have been of significance to seabirds (Golet et al., 2002). Tufted puffins have been monitored only since the early to mid-1990s. In that time, there have been no substantial changes. Increases of tufted puffins in the Aleutian Is. are attributable to the reduction of introduced foxes on many of the islands, a ban on salmon driftnets in the North Pacific, and possibly other reasons such as climate change (Byrd et al., 1992). At one of the best-known colonies in the Pacific Northwest, Tatoosh I., Washington, the decline of tufted puffins is thought to have been caused by the loss of nesting habitat rather than a change in prey abundance (R. Paine, pers. comm.). Trends in abundance of other species in the GOA are mixed, further confusing our understanding about population change relative to climate change or other causes. Storm petrels have been increasing dramatically in the past 15 years at three widely spaced colonies from Southeast Alaska to the Aleutians. Cormorants declined precipitously in the late 1970s and 1980s at several colonies, but increased at Middleton I. from the early 1970s through the 1980s before collapsing by approximately 80% between 1990 and 1992. Marbled murrelets have declined steadily since the early 1970s in PWS and during the 1990s in Glacier Bay, but apparently increased dramatically in the Kenai Fjords in the 1990s after declining in the previous 15 years. Kittlitz’s murrelets have collapsed by 60–99% in Glacier Bay, PWS, and the Kenai Fjords. The trends in marine birds in PWS, based on pelagic surveys that indicate declining numbers of most piscivorous species, are imprecise and equally confusing for three reasons – the very long interval between the first count in 1972 and the second count in 1989, the fact that survey methods differed in the 1972 count from all subsequent counts, and the apparent decline in kittiwake abundance indicated by the pelagic counts is opposite to the increases revealed by the much more precise counts at colonies.
Long-Term Change 329
Despite the paucity of evidence that regime shifts have precipitated widespread trends in abundance of kittiwakes, murres, or other species in the GOA, change has certainly occurred on smaller scales. And, changes in prey abundance appear to have been involved in some of it. For example, overall kittiwake abundance in PWS has gradually increased since the early 1980s, and there has been a conspicuous change in distribution within the sound, with colonies in the northern, more inner regions increasing at the expense of colonies in more southern, outer regions (Suryan and Irons, 2001). This redistribution is attributed primarily to movements of birds from poorer to better foraging areas – kittiwakes at the better colonies feed mainly on juvenile herring that have increased, whereas kittiwakes at the poorer colonies feed mainly on sand lance and capelin, which may have declined there as did capelin in the Kodiak I. area and Alaska Peninsula in the late 1970s–1980s (Anderson and Piatt, 1999). Likewise, the long-term decline of kittiwakes and murres at Chisik I. and corresponding increases across Cook Inlet at Gull I. appear to have been driven in large measure by movements of birds in response to contrasting foraging conditions in the two regions, as well as to differential productivity and recruitment (Piatt and Harding, Section 4.8). But not all movements of seabirds and changes in abundance are driven by changes in prey. We have already examined the case of Middleton I. and the major effect terrestrial habitat quality and gull predation appear to have had on numbers of kittiwakes and murres. At Kodiak I., kittiwakes in Chiniak Bay have been relocating from older colonies generally on larger more exposed cliffs to smaller more secure sites as the abundance of bald eagles (Haliaeetus leucocephalus) and peregrine falcons (Falco peregrinus) has increased over the past 30 years (Kildaw et al., 2005). The higher quality, larger cliffs preferred by kittiwakes are also attractive to eagles and falcons as roosting and hunting perches, and their added presence has increased disturbance of adult kittiwakes at the colonies and allowed greater access to their eggs by northwestern crows and black-billed magpies, both inveterate egg thieves. Predation also is an important detriment to productivity of kittiwakes and pigeon guillemots in PWS that might have colony-scale effects (Oakley and Kuletz, 1997; Suryan et al., 2006). And, declines of Kittlitz’s murrelets have been attributed to the recession of tide water glaciers in the GOA (Kuletz et al., 2003). In this case, the mechanism is not certain, although it is suspected that a major factor is the loss of critical foraging habitat, rather than a loss of prey per se, as glaciers shrink with continued global warming.
4.7.7. Changing Productivity As we have seen, productivity of kittiwakes is particularly sensitive to food supply, and variability in productivity can reveal variability in prey at numerous spatial scales. At a regional scale, such as PWS, productivity at colonies that have increased in
330
Long-Term Ecological Change in the Northern Gulf of Alaska
abundance in the past 20 years has been markedly higher than at colonies that have declined: 0.24 chicks per nest (SE = 0.018, n = 152 colony-years) versus 0.054 chicks per nest (SE = 0.0075, n = 279 colony-years). The difference, as noted above when explaining changes in abundance, is likely due primarily to prey availability, although predation can be a major factor as well (Suryan and Irons, 2001; Suryan et al., 2006). On a broader scale, the data can be pooled in ways which indicate that productivity of kittiwakes in the GOA, and an even larger region of the North Pacific, has changed in the last three decades. Hatch et al. (1993) lumped all productivity data for kittiwakes from Alaska for all years through 1989 (162 estimates of productivity from 28 colonies). They found that productivity declined steadily during the 1970s and 1980s, from about 0.5 chicks fledged per nesting attempt in the early 1970s to fewer than 0.2 chicks per nest in the late 1980s. At the same time, total nesting failures increased. Piatt and Harding (Section 4.8, Fig. 4.44) found similar broad, decadalscale changes in kittiwake productivity in the GOA and Bering Sea through the 1990s. When the data are pooled somewhat differently, by colony and regime (the duration of the mean state of each regime, except the most recent one that is apparently ongoing), a contrasting picture emerges (Fig. 4.35). Although differences between interval means are not generally significant, there is a common pattern of higher productivity
1 0.9
Kittiwake colonies
Chowiet I.
Middleton I.
Chiniak Bay
PWS
Barren Is.
Gull I.
0.8 PRODUCTIVITY
0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 1970s−1980s
1990s DECADE(S)
2000s
Figure 4.35: Productivity (chicks/nest) of black-legged kittiwakes at colonies in the GOA. Productivity averaged by decade, approximately the duration of climate regimes as indexed by the PDO. Data from Dragoo et al. (2004), D. Irons (unpublished data), and D. Kildaw et al. (2005).
Long-Term Change 331
in the mid-1970s through the 1980s, lower productivity in the 1990s, and higher productivity again in the 2000s. The exception is at Chiniak Bay, where productivity in the 1990s was higher than it was before or after. But is it so simple? Not really. At kittiwake colonies where annual resolution is high (including colonies in the GOA and Bering Sea), we can see that productivity goes through cycles at each one, but the cycles are not in phase, have different periods, and the periods are sometimes irregular at individual colonies (Fig. 4.36). For example in PWS, as exemplified by the Shoup Bay colony (the largest colony in PWS; it has grown steadily since 1972), productivity has risen and fallen with a period of about 8 years during the past 20 years. So too at Chiniak Bay, but there the period of the first 1.5 cycles was about 13 years, whereas the most recent one is apparently just about half that. Moreover, it is roughly 180 degrees out of phase with Shoup Bay. Chiniak Bay more closely resembles St. George I. (Pribilof Is., Bering Sea), which has a similar period but slightly different phase. Other colonies with evidence of cyclic productivity are Middleton I., although productivity there is so strongly influenced by predation that it is likely that it is not entirely reflective of whatever factors are driving the cycles; and Bluff, a coastal colony in the northeastern Bering Sea.) In none of these examples are the waves of productivity consistently organized around regimes, as indexed by the PDO, or possible shift triggers, such as El Niños or La Niñas. A strong La Niña in 1975–1976 followed by El Niño in 1976–1977
0.8 KITTIWAKE COLONIES
0.7
Shoup Bay Chiniak Bay
PRODUCTIVITY
0.6 0.5 0.4 0.3 0.2
2005
2003
2001
1999
1997
1995
1993
1991
1989
1987
1985
1983
1981
1979
1977
1975
0.1 St. George Is. 0
YEAR
Figure 4.36: Annual trend (smoothed) in productivity (chicks/nest) of black-legged kittiwakes at two colonies in the GOA and at St. George I. in the Bering Sea. Data from Dragoo et al. (2004), D. Irons (unpublished data), and D. Kildaw (unpublished data).
332
Long-Term Ecological Change in the Northern Gulf of Alaska
marked the mid-1970s regime shift; an El Niño in 1986–1987 followed by a strong La Niña in 1988–1989 marked the next shift; and a La Niña in 1995 followed by a strong El Niño in 1997–1998 marked the most recent shift. The strong El Niño in 1982–1983 without a corresponding La Niña was not associated with a regime shift. Rather, the beats in the various data sets appear to be independent of climate events or regimes.
4.7.8. Conclusions There is little compelling evidence that the mid-1970s regime shift was of great importance to seabird populations in the GOA, certainly not as an agent of widespread decline or reproductive failure. Despite changes in the structure of fish communities, the ability of murres, puffins, and kittiwakes to buffer themselves against such changes through their natural history and survival strategies seems to have prevailed. Or, alternatively, there may not have been significant change in the prey assemblage important to seabirds. The notion that the regime shift of 1977 and community reorganization led to a collapse of the assemblage of forage fishes essential to seabirds, i.e., especially those with a high lipid content, hinges on just a single species. Although shrimp and a few species of fishes commonly referred to as forage fish declined in abundance at various times from the early to mid-1970s through the mid-1980s in the western and central GOA (Bechtol, 1997; Piatt and Anderson, 1996; Anderson and Piatt, 1999; Mueter and Norcross, 2000), capelin is the only one that was of particular importance to seabirds (DeGange and Sanger, 1986; Piatt and Anderson, 1996). Sand lance was thought to have declined in PWS (Kuletz et al., 1997), but evidence from elsewhere in the GOA does not support a conclusion of widespread change (Golet et al., 2002). Despite the high nutritional value of capelin and the fact that it is usually consumed by seabirds when it is present, its decline, as revealed by changes in seabird diets and fishery survey results, is itself not evidence that seabirds were food limited – that is, limited to the extent that their populations fell. They still had available to them gadids, particularly walleye pollock and sand lance (a comparatively lipid-rich species), and elsewhere murres and kittiwakes subsist quite well, at least during the nesting season, on diets dominated by gadids and sand lance (Springer et al., 1984, 1987). Thus, in this case the lack of capelin in seabird diets simply demonstrates that a decline in abundance of a common forage species can be seen at higher trophic levels. Moreover, increases, rather than decreases, in abundance of murres, puffins, and kittiwakes at many colonies in the GOA during the 1970s and 1980s indicate that food shortage was not a universal characteristic in the environment at that time. The abundance of murres and kittiwakes did change at several colonies during the years following the first counts in the mid-1970s, and if the changes began then it
Long-Term Change 333
might be fair to assume that the mid-1970s regime shift was an important factor. But we do not know whether changes really did begin in the mid-1970s rather than earlier, as they did at Chisik I. in the GOA and at Cape Thompson in the Chukchi Sea. Nor do we know whether changes in abundance, whenever they began, were necessarily driven by changes in prey. Many likely were, but predation and habitat quality also can have major effects on seabird populations, as in the cases of murres and kittiwakes at Middleton I. and Chiniak Bay, Kittlitz’s murrelets in glacially influenced fjords, and kittiwakes and guillemots in PWS. This is not to say that seabirds are insensitive to prey availability, or that there have not been broadscale responses in seabird populations to regime shifts or other climate ~o that most likely involved food. They are sensitive to change events such as El Nin prey, and they have recorded signals of the effect climate has on ocean food webs – for example, changes in breeding parameter values described above by Piatt et al. that were associated with the regime shifts of 1977 and 1989; large episodic die offs of seabirds in the GOA and Bering Sea apparently caused by food shortage and some~o and La Nin ~a events (Nysewander and times, though not always, following El Nin Trapp, 1984; Hatch, 1987; Piatt and Van Pelt, 1993); and adjustments kittiwakes have made to short-term fluctuations in prey availability in PWS (Suryan et al., 2002). Moreover, patterns of colony size and distribution of the many kittiwake colonies in PWS has likely been shaped to a large degree by prey distribution and abundance (Ainley et al., 2003). And, if abundance data of murres used by Piatt and Anderson (1996), and discussed in the subsequent chapter, are more or less accurate, then there may well have been a broadscale decline in murres beginning in the early 1970s, or before, and continuing into the late 1980s. Furthermore, if pooling kittiwake productivity data across all colonies, different as they are (in terms of the oceanographic settings in which they are found), is informative, then those data can also be used in support of the idea of broadscale patterns in seabird biology in the GOA and Bering Sea. Finally, if capelin really is the crucial forage species for seabirds in the GOA, and if its overall decline was caused by the regime shift but lagged it by five or so years, i.e., did not really begin to decline until the early 1980s, and the timing of decline and other changes in fish and shellfish assemblages varied substantially from place to place in the GOA (Mueter and Norcross, 2000), then this too might support the scenario that the regime shift was important in some cases. But El Niños, La Niñas, and regime shifts do not align well with much of the variability in seabird abundance in the GOA or elsewhere in Alaska. For example, in colonies with the longest histories, the abundances of murres in most of them have exhibited directional change over extended intervals that crossed one or more regime boundaries. Such long-period fluctuations may be driven by climate change, and if so the time frame is greater than the decadal-scale PDO-type oscillation and may reflect other modes of variability in marine ecosystems, as described by Minobe (1997).
334
Long-Term Ecological Change in the Northern Gulf of Alaska
Likewise, variability in kittiwake productivity, which is thought to be particularly sensitive to prey availability and quality, does not match the PDO regime pattern. Productivity in many colonies shows wave patterns, although the waves are commonly out of phase (even at nearby colonies) and the periods vary. There is no consistent relationship to major climate signals, but instead they seem to be driven by intrinsic properties of each population or of local environments, food webs, or prey populations supporting them. The phase shifts and variable periods could lead to perceptions that pooled inter-regime means have significance relative to those regimes, when in reality they may not. Effects of physical forcing on marine production patterns and processes are conspicuous at low trophic levels; however, they are attenuated more and more at higher and higher trophic levels due to the survival strategies of individual species. The survival strategies of murres, kittiwakes, and puffins, varied as they are, appear to have succeeded in buffering most populations against adverse changes in prey, if they did in fact occur. For others, notably Chisik I., they were not successful in saving kittiwakes and murres from apparent changes in prey, beginning in the early 1970s or before. Nor were survival strategies entirely successful in buffering some populations against the effects of predators. For seabirds, therefore, an alternative conclusion to the one most commonly pronounced could be that there is rather weak support for the notion that the regime shift of the mid-1970s and the restructuring of the fish and shellfish communities in the GOA had broadscale, negative effects on abundance. Rather, if changes in prey of significance to seabird populations did occur, they were more at local scales and were both positive and negative. The lack of organization of kittiwake productivity at many individual colonies around regimes and regime shifts further argues against a pervasive, broadscale, adverse effect of climate change on seabirds at those timescales in the GOA. At the local and even regional scale, chaos may be a more apt description of seabird population biology in the North Pacific, and this may reflect complex interactions between microscale oceanographic conditions, local responses of individual species to physical forcing, and predator–prey relationships. This contrasts in many important ways from patterns in marine mammal biology, which will be described in Section 4.9.
4.7.9. Summary Unfortunately, very little is known about seabird abundance or numerical trends in the GOA prior to the mid-1970s. In the two cases we have, both murres and kittiwakes at Chisik I. in Lower Cook Inlet were in decline in the early 1970s, while both species were increasing at Middleton I. We do not know when the declines began at Chisik I., but the increases at Middleton I. likely began following the 1964 earthquake that increased nesting habitat.
Long-Term Change 335
After the mid-1970s and through the 1980s, there are examples of increasing and declining populations, but for most of those that changed following the first counts, it is again not known when they began to change. The same holds true for the 1990s. This is not to say we know nothing about the timing of change. One exception is Gull I., where murres apparently began to increase in the mid-1980s. And, there are examples of abrupt changes in the early to mid-1990s. Overall, there are just two apparent broadscale patterns in seabird abundance in the GOA – that of tufted puffins, which have been stable or increased slightly, and Kittlitz’s murrelets, which have declined dramatically at all locations where they have been counted. For other species, there are examples of increases, decreases, and no change over intervals spanning one to three decades. Moreover, for murres and kittiwakes, about which we know the most, trends at individual colonies where both species nest were not always in the same direction. This lack of pervasive pattern suggests that a variety of factors affecting populations have been at play, or that if there is only a single factor, its expression has differed markedly in the spatially heterogeneous environment of the GOA.
4.8. Population Ecology of Seabirds in Cook Inlet John F. Piatt and Ann M.A. Harding 4.8.1. Introduction Many seabird colony populations in the Gulf of Alaska have fluctuated in recent decades, and they have declined markedly at a few sites (Hatch and Piatt, 1995; Dragoo et al., 2000). The Exxon Valdez oil spill of 1989 had an immediate impact on some seabirds (Piatt et al., 1990; Piatt and Ford 1996), adding to other anthropogenic factors influencing seabird populations in Alaska (e.g., gill-net mortality, introduced predators, etc.; Hatch and Piatt, 1995). However, evidence accumulated during the 1990s that background variability in the marine environment had an even greater effect on seabird populations over annual and decadal timescales. Most notably, a “regime shift” occurred in the Gulf of Alaska during the late 1970s, apparently causing marked changes in seabird diets, lower reproductive success, occasional wrecks (large-scale die-offs), and long-term declines in some marine bird and mammal populations (Piatt and Van Pelt, 1997; Piatt and Anderson, 1996; Francis et al., 1998). Fish communities on the continental shelf of the Gulf of Alaska also changed dramatically during that time period (Anderson and Piatt, 1999). Coincident with cyclical fluctuations in seawater temperatures, the abundance of small forage fish species such as capelin (Mallotus villosus) declined precipitously in the late 1970s, while populations of large predatory fish such as walleye pollock (Theragra chalcogramma) and
336
Long-Term Ecological Change in the Northern Gulf of Alaska
cod (Gadus macrocephalus) increased dramatically. Correspondingly, capelin virtually disappeared from seabird diets in the late 1970s, and were replaced by juvenile pollock and other species in the 1980s (Piatt and Anderson, 1996). Seabirds and marine mammals exhibited several signs of food stress (population declines, reduced productivity, and die-offs) throughout the 1980s and early 1990s (Merrick et al., 1987; Piatt and Anderson, 1996). In part because of these observations, the Exxon Valdez Oil Spill Trustee Council initiated the Apex Predator Ecosystem Experiment (APEX) in 1994 to assess whether post-spill environmental conditions could influence the recovery of seabirds from the oil spill. Initially focused on Prince William Sound, APEX studies expanded in 1995 to include Cook Inlet, where it was possible to conduct seabird and forage fish studies around three seabird colonies (Chisik, Gull, and Barren Islands) (Fig. 4.37). Oceanography, plankton, forage fish ecology, and seabird distribution at sea were studied in waters around each colony (Drew, 2002; Drew and Piatt, 2002; Piatt, 2002a; Robards et al., 1999a,b,c; Robards, 2000; Robards et al., 2002; Abookire et al., 2000; Abookire and Piatt, 2004; Litzow et al., 2000; 2004a,b, Speckman, 2004;
Figure 4.37: The location of the three study colonies in lower Cook Inlet.
Long-Term Change 337
Speckman et al., 2005). Seabird foraging behavior, diets, time budgets, chick growth rates, physiological condition, reproductive success, population trend, and adult survival were studied concurrently at each of the three colonies (Roseneau et al., 1997, 2000; Zador et al., 1997; Kitaysky et al., 1998, 1999, 2001, 2005; Piatt et al., 1999; Zador and Piatt, 1999; Litzow, 2000; Litzow et al., 2000; Romano, 2000; Van Pelt, 2000; Harding, 2001; Harding et al., 2002, 2003; Litzow and Piatt, 2003; Piatt, 2004). Most data were collected on common murres (Uria aalge; Van Pelt et al., 2002) and black-legged kittiwakes (Rissa tridactyla, Shultz et al., 2002), which breed at all three study sites in lower Cook Inlet. Results of APEX work are still being published, but here we summarize some of the main findings of studies in lower Cook Inlet. After a brief review of the Cook Inlet ecosystem, we consider some of the similarities and differences in the way murres and kittiwakes respond to fluctuations in food supply. We then consider murre and kittiwake population dynamics in Cook Inlet and elsewhere in Alaska and what this may reveal about long-term changes in the marine environment of the Gulf of Alaska and Bering Sea.
4.8.2. The Cook Inlet Ecosystem We set out from the beginning to study seabirds and prey resources at colonies known from historical work to be chronically failing (Chisik Island), thriving (Gull Island), and possibly stable or recovering from the oil spill (Barren Islands). Our hope was that historic differences in bird biology were indeed the result of regional differences in food supplies and that, by studying all three colonies for five years, we would obtain enough data to characterize favorable and unfavorable environmental settings for seabirds (Piatt, 2002b). The engine that drives productivity in lower Cook Inlet is the persistent upwelling of cold, nutrient-rich water at the entrance to Cook Inlet (Fig. 4.38). A plume of cold Gulf of Alaska water that extends from the Barrens to Kachemak Bay (large bay on the east side of lower Cook Inlet) persists throughout summer in all years (Drew and Piatt, 2002; Speckman et al., 2005) and provides nutrients that fuel extraordinary primary production (Drew, 2002). Waters on the west side of lower Cook Inlet are oceanographically distinct (warmer, less saline, weakly stratified, turbid, and outflowing), and much less productive. The east–west difference in oceanography and primary production is reflected at all higher trophic levels. The abundance of zooplankton (Drew, 2002; Speckman et al., 2005), forage fish offshore (Abookire and Piatt, 2004; Speckman, 2004) and nearshore (Robards et al., 1999b), and seabirds (Speckman, 2004) is in all cases 1–2 orders of magnitude greater on the east side of lower Cook Inlet. The growth rate of resident forage fish such as sand lance (Ammodytes hexapterus) is significantly lower in Chisik waters than in Kachemak Bay
338
Long-Term Ecological Change in the Northern Gulf of Alaska
Figure 4.38: Sea surface temperature in the northern Gulf of Alaska on July 3, 1999. Note the colder (blue) surface water that results from upwelling in lower Cook Inlet and east of Kodiak Island in contrast to the warmer (red–orange–yellow), stratified waters in Prince William Sound and shelf waters south of the Kenai Peninsula. Data is from NOAA AVHRR satellite.
(Robards et al., 2002). Thus, all evidence suggests that lower Cook Inlet is segregated into distinct oceanographic domains, with striking differences in productivity and biology among them (Speckman et al., 2005). In all years, the well-mixed, cold waters in the lower inlet – particularly offshore – were dominated by juvenile pollock and capelin (Abookire and Piatt, 2004), important prey for murres (Van Pelt et al., 2002). Sand lance and herring (Clupea pallasi) were the most common prey for kittiwakes (Shultz et al., 2002), and were most abundant in stratified coastal waters of the Kenai Peninsula and Kachemak Bay (Robards et al., 1999b; Abookire and Piatt, 2004). Fish were markedly variable in the vertical dimension as well. Most acoustic biomass was concentrated in the upper 30 m in all areas, but in Chisik and Barren island waters, schools were also concentrated at depths of 60–100 m (Speckman, 2004). There was also a clear segregation of species by water depth; sand lance and herring dominated above depths of 40 m, whereas pollock and capelin dominated below 60 m. Diets of adult murres and kittiwakes reflected food supplies around each colony. Whereas more than 90 species of fish were caught near shore and 40 species were caught in offshore trawls, communities were overwhelmingly dominated (>95%) by
Long-Term Change 339
four species: sand lance, herring, Pollock, and capelin (Robards et al., 1999b, 2002; Abookire and Piatt, 2004). Diets of adult murres and kittiwakes were dominated by the same species in similar proportions to local abundance except that herring were generally eaten less and capelin eaten more in proportion to their relative abundances (Van Pelt et al., 2002; Shultz et al., 2002). Sand lance dominated murre and kittiwake diets at both Chisik and Gull, while pollock comprised a much larger proportion in diets of birds from the Barrens. The size classes of prey eaten by adults was similar to the size classes caught in trawls and seines. Taken together, the evidence suggests that adult murres and kittiwakes generally eat what is most available to them within foraging range of their colonies. In contrast to adult diets, however, chick diets were poor indicators of relative prey availability because adults choose to feed their chicks prey that are oily and rich in calories, a behavior with obvious adaptive value. The breeding biology of seabirds differed markedly among colonies owing to the persistent geographic differences in forage fish availability described earlier. Birds at Chisik Island struggled to reproduce, while those at Gull and Barren islands usually had few problems rearing young (Kitaysky et al., 1998, 1999; Zador and Piatt, 1999; Shultz et al., 2002; Van Pelt et al., 2002). Within each colony, breeding and behavioral parameters varied among years to a lesser degree than among sites. Breeding success in all species was lower in 1998 than in other years; presumably a lingering effect of the previous winters’ El Niño event (Piatt et al., 1999). Population censuses revealed that seabirds at Chisik Island continued in a long-term decline, whereas populations at Gull and Barren islands were increasing (Zador et al., 1997; Roseneau et al., 1997, 2000; Piatt, 2002a). Behavioral studies revealed that seabirds worked harder (longer foraging trips, less discretionary time) at colonies where nearby fish densities were lower (Zador and Piatt, 1999; Piatt, 2002a). Overall, the studies show that seabird parameters (breeding success, foraging effort, population trend, etc.) varied most between islands and to a much lesser degree between years. We attribute this regional variability and temporal stability in seabird biology to distinct, persistent oceanographic regimes around each colony that determined the availability of fish to birds within those areas (Abookire and Piatt, 2004; Speckman et al., 2005).
4.8.3. Response of Seabirds to Variability in Prey Form of Response We predicted that – just like other vertebrates (Holling, 1959; Murdoch and Oaten, 1975; Piatt, 1990) – murres and kittiwakes would exhibit nonlinear functional relationships with food supply (Piatt, 2002b). Of the 25 relationships that we examined between parameters of seabird biology, behavior or physiology e.g., clutch size,
340
Long-Term Ecological Change in the Northern Gulf of Alaska
Figure 4.39: Behavioral and reproductive responses of common murre and blacklegged kittiwake to variation in their food supply.
body condition, chick growth, hatching success, feeding rate, etc.) and prey density, more than half (14) were nonlinear, only one was linear, and the rest were not described by any significant model (Piatt, 2002a). For example, the attraction of foraging murres and kittiwakes to prey schools at sea (the “aggregative response”) was sigmoidal for both species (Speckman, 2004) (Fig. 4.39). In recent years, breeding success in several seabird species has been shown to be a curvilinear function of food density (Arctic skua, Philips et al., 1996; Atlantic puffin, Anker-Nilssen et al., 1997; Arctic tern, Suddaby and Ratcliffe, 1997), and we demonstrated a similar relationship for black-legged kittiwakes (Fig. 4.39). Common murres did not exhibit such a relationship because breeding success was usually buffered by the ability to increase time spent foraging in the face of declining food supply. Instead, we observed a curvilinear relationship between discretionary time spent at the colony and food density in murres (Fig. 4.39). These and other functional response curves in murres and kittiwakes (Piatt, 2002a) suggests that food supplies at Gull and Barren islands – but not at Chisik – are above threshold limits and adequate to support recovery of losses from the Exxon Valdez oil spill.
Variability in Response We found that neither hatching, fledging, nor breeding success in common murres was correlated with food supply (Shultz et al., 2002; Piatt, 2002a). Murres appeared to have trouble fledging chicks in only 2 of 14 colony-years of study (Fig. 4.39). On these two occasions, murres exhibited unusually low hatching success (52% vs.
Long-Term Change 341
70–90%), fledging success (<45% vs. 64–92%), and consequently low overall breeding success (<30% vs. 53–81%). In other years, however, murres were never limited by food supplies. It appears that they were able to compensate for low food abundance by reducing time spent attending their nest site (Fig. 4.39) and devoting that time to foraging (Burger and Piatt, 1990; Zador and Piatt, 1999). Because of this buffering capacity, murre fledging success (CV = 28%), and breeding success (CV = 29%) were about three times less variable than kittiwake fledging success (CV = 81%) and breeding success (CV = 87%) during 14 colony-years of study in Cook Inlet (Piatt, 2002a). Despite a robust capacity to deal with variation in prey abundance, however, murre breeding success cannot be completely independent of food supply. Functional response curves for aggregation and discretionary time spent at the colony (Fig. 4.39) indicate that murres – as with kittiwakes – have a foraging threshold at about 0.013 g/m3 of fish biomass. Perhaps we failed to characterize a functional reproductive response curve because food supplies were never so low that they could not be behaviorally buffered to prevent breeding failure. Consequently, while we frequently observed kittiwake breeding failures in Cook Inlet, we never observed total breeding failure in murres. One explanation for the difference may by that when they are faced with the same prey fields around a colony, murres can fly further and search nearly twice as much surface area than kittiwakes in the same time, dive far below the surface in search of prey, and they have more discretionary time to divert to foraging (Piatt, 2002a). How representative are these data from Cook Inlet for murres and kittiwakes elsewhere? We can examine variability in breeding success of murres and kittiwakes from long-term (1975–1999) data collected throughout Alaska (Gulf of Alaska, Aleutians, Bering, and Chukchi Seas) in a variety of monitoring and research programs (Hatch et al., 1993; Dragoo et al., 2000). From these data (Fig. 4.40, Table 4.4), which include an extreme – but natural – range of environmental conditions for breeding (Hatch et al., 1993), we find that common murres (n = 14 colonies, n = 99 colonyyears) have rarely exhibited complete breeding failure (0 chicks/pair on only 4% of occasions), and on only 26% of the occasions was breeding success indicative of a limiting food supply (i.e., below 0.40 chicks/pair; see earlier text). Remarkably, common murres were successful (>0.40 chicks/pair) about three-quarters of the time (Fig. 4.40) and variability in breeding success was quite low (CV = 40%) and similar to that observed in Cook Inlet (CV = 28%). In contrast, kittiwakes (n = 17 colonies, n = 235 colony-years) had complete failures (0 chicks/pair) 18% percent of the time, and showed signs of food limitation (breeding success <0.46 chicks/pair) 77% of the time. On only 23% of occasions did kittiwakes appear to be unrestricted by food supply. Kittiwakes showed high levels of variability in breeding success in Cook Inlet (CV = 87%) and Alaska (CV = 110%). A similar contrast in murre and kittiwake breeding success has been noted elsewhere (Table 4.4). Common murres observed during 54 colony-years at a variety of
342
Long-Term Ecological Change in the Northern Gulf of Alaska
Figure 4.40: Frequency of different levels of breeding success for common murres and black-legged kittiwakes in Alaska (data from Hatch et al., 1993; Dragoo et al., 2000; Roseneau et al., 2000; Piatt, 2002a). Values are the upper end of the range, i.e., 0.2 is the frequency of breeding success between 0 and 0.2 chicks per pair. The murre graph includes 96 colony-years of data, and the kittiwake graph includes 232 colony-years, where a colony-year comprises data collected at any one colony in 1 year.
colonies in the North Atlantic (Birkhead, 1976; Hedgren, 1980; Birkhead and Nettleship, 1987; Harris and Wanless, 1988, unpublished data; Burger and Piatt, 1990; Bryant et al., 1999) averaged 0.75 ± 0.09 S.D. chicks/pair. Despite the span of years (1963–2001) and colonies (n = 11), variability in breeding success was low (CV = 12%), with only one occasion where breeding success was less than 0.4 chicks/pair (0.26 chicks/pair, remaining values ranged from 0.52–0.88 chicks/pair). At the Isle of May, murres never failed in 21 years of study (range 0.63 to 0.81 chicks/pair; M. Harris and S. Wanless, unpublished data). Likewise, in 29 years of study at the Farallon Islands, California, common murres failed (<0.4 chicks/pair) only three times, all in association with strong El Niño (ENSO) events (Sydeman et al., 2001). Otherwise, breeding success ranged between 0.61 and 0.91 chicks/pair.
Long-Term Change 343
Table 4.4: Variability in breeding success of black-legged kittiwakes and common murres in different geographic areas. Breeding Success Species
Location
Kittiwakes
Cook Inlet Gulf of Alaska Aleutians Bering Sea Chukchi Sea Alaska Newfoundland Vedoy I., Norway Hornoya I., Norway Isle of May, UK North Sea 1986 North Sea 1987 North Sea 1988 West Coast UK Atlantic Ocean Cook Inlet Gulf of Alaska Aleutians Bering Sea Alaska California Newfoundland Isle of May, UK Europe Atlantic Ocean
Murres
N (years)
mean
15 113 20 84 18 235 7 20 17 17 15 20 21 31 148 14 34 13 52 99 29 14 21 19 54
0.31 0.24 0.27 0.24 0.82 0.29 0.86 0.69 0.93 0.59 1.09 0.96 0.61 0.62 0.77 0.61 0.54 0.41 0.50 0.50 0.74 0.76 0.78 0.70 0.75
CV (%) 87 110 84 94 65 110 65 41 27 69 29 49 85 56 53 29 34 76 35 41 29 9 7 20 12
S.D. 0.27 0.26 0.22 0.24 0.54 0.32 0.56 0.28 0.25 0.41 0.32 0.47 0.52 0.35 0.39 0.18 0.18 0.31 0.17 0.20 0.22 0.07 0.06 0.14 0.09
(See text for sources of data).
Kittiwake breeding success measured in the Atlantic during 143 colony-years (42 colonies, 1973–2001; Birkhead and Nettleship, 1988; Harris and Wanless, 1990, unpublished data; Hamer et al., 1993; Erikstad et al., 1995; Barrett, 1996, unpublished data; Anker-Nilssen et al., 1997) averaged 0.77 ± 0.39 chicks/pair and variability (CV = 53%) was more than four times greater than that observed in Atlantic murres (12%, Table 4.4). Indeed, it appears that under a wide range of conditions, kittiwake productivity is always more variable than murre productivity (Fig. 4.41). Furthermore, for both species, variability is high when productivity (and
344
Long-Term Ecological Change in the Northern Gulf of Alaska
Figure 4.41: Coefficient of variation (CV) in breeding success versus average breeding success in Common Murres (y = −134x + 114; r2 = 0.79) and black-legged kittiwakes (y = −80.3x + 118; r2 = 0.77) among different colony-years within various regions and subregions around the world (see Table 4.4).
presumably food density) is low. This suggests that factors controlling seabird populations operate over moderate to large regional scales such as those arbitrarily selected in Table 4.4. It appears that in “good times,” high-density prey aggregations are available to most colonies in a region, most colonies do well, and variability is therefore low. In “bad times,” prey aggregations are more patchy and available to fewer colonies, more colonies fail, and variability is therefore high.
4.8.4. Population Dynamics of Seabirds in Cook Inlet Our study was designed to provide contrasting data from a “food-poor” colony (Chisik), where murre and kittiwake populations were known to have been declining at rates of 4–9% per annum for the past 30 years (Fig. 4.42), and a “food-rich” colony (Gull), where murre and kittiwake populations grew at rates of 9% per annum at some point during the past 25 years (Zador et al., 1997; Piatt, 2002a). Kittiwakes increased rapidly on Gull Island during the 1980s, but populations leveled off in the 1990s and remained at the same level throughout the course of our study. Evidence suggests that this was due to saturation of nesting habitat on the island; otherwise, kittiwakes would still be increasing at the rate observed prior to the plateau, and at a rate similar to that observed for murres (which were not limited by nesting habitat). Trends at the Barren Islands were unknown prior to the Exxon Valdez oil spill, but both murres and kittiwakes exhibited modest increasing trends (Fig. 4.42) during the 1990s (Roseneau et al., 1997, 2000).
Long-Term Change 345
Figure 4.42: Populaton trends for populations of common murres and black-legged kittiwakes at Chisik, Gull, and Barren islands, with average annual percentage changes. Data for Gull and Chisik from Piatt 2002, and Zador et al., 1997. Data for Barren Islands from Roseneau et al., 1997 and Roseneau et al., 2000.
Historical measures of productivity in kittiwakes (Fig. 4.43) are parallel to population trends. Kittiwakes have failed chronically at Chisik for more than 30 years, averaging 0.05 chicks/pair during that time and only 0.02 chicks/pair during our study in 1995–1999 (Zador et al., 1999; Shultz et al., 2002). Kittiwakes averaged 0.44 chicks/ pair at Gull Island since 1984 and 0.48 chicks/pair during our study. This is higher average productivity than has been observed at any other colony in Alaska except Cape Lisburne, where populations have been increasing during the past couple of
346
Long-Term Ecological Change in the Northern Gulf of Alaska
Figure 4.43: Historical productivity of black-legged kittiwakes at Chisik, Gull, and Barren islands, 1979–1999. Data from Piatt, 2002a; Zador et al., 1997; Roseneau et al., 2000 and Dragoo et al., 2000. Years when surveys were conducted but productivity was zero are indicated with a “0” above the bar.
decades (Dragoo et al., 2000). Productivity was more variable at the Barrens, averaging 0.29 chicks/pair during the past decade and 0.43 chicks/pair during 1995–1999 (Dragoo et al., 2000; Roseneau et al., 2000). In contrast, murre productivity does not correlate with population trends. Despite having markedly different population trends at the three colonies (Fig. 4.42), breeding success was high on Chisik (average of 0.56 chicks/pair), Gull (0.54 chicks/pair), and the Barren Islands (0.73 chicks/pair) during five years of study in 1995–1999 (Roseneau et al., 2000; Van Pelt et al., 2002). There are no prior historical reproductive data for murres except from the Barrens in 1989–1993 (Nysewander et al., 1993; Roseneau et al., 1997; Boersma et al., 1995). When included, these data suggest an average productivity of 0.54 chicks/pair at the Barrens during the past 12 years.
Long-Term Change 347
Results from Chisik beg the question: Why is the murre population there declining by 9% per annum (pa) when reproductive success appears normal? A similar question has been asked about murres in the Shetlands, which continued to have high breeding success even as numbers at colonies declined in apparent response to a crash in food stocks (Furness and Camphuysen, 1997). Preliminary results of a survival study on Gull and Chisik islands (Piatt, 2004) may help answer this question. This study suggested that there are marked differences in population parameters of murres on Gull and Chisik islands. On Chisik, annual adult mortality (9.2% pa) exceeds slightly the rate of population decline (−8.9% pa), suggesting a small balancing rate of 0.3% pa increase due to recruitment (or possibly immigration). This indicates that survival of murre chicks to breeding at Chisik is neglible. Indeed, survival of chicks is likely to be much lower if fledgling chicks are underweight (Hatch, 1983; Sagar and Horning, 1998) – as they frequently are at Chisik (Van Pelt et al., 2002). Furthermore, recruitment at a declining murre colony is likely to be less than 20% (Hudson, 1985) and possibly less than 5% (Hatchwell and Birkhead, 1991). This would explain how Chisik murres can maintain such high breeding success and yet experience serious population declines. In contrast, the high rate (9.1% pa) of murre population increase at Gull Island can be explained by a low rate (6% pa) of adult mortality that is more than offset by high rates of recruitment and/or immigration. Kittiwake population parameters are more straightforward (Piatt, 2004). At Chisik, recruitment has to be virtually zero because productivity is negligible. Thus, the population decline (−4.3% pa) is explained entirely by adult mortality (6.7% pa), offset slightly by immigration. At Gull Island, a much higher adult mortality rate (14.5% pa) is balanced by much higher levels of productivity, recruitment, and immigration. The differences in survival and productivity between Gull and Chisik seem to support the hypothesis that long-lived seabirds trade off the costs of reproduction with adult survival (Erikstad et al., 1998; Golet et al., 1998, 2004).
Population Parameter Indices The question posed by the Exxon Valdez Oil Spill Trustee Council was whether or not recovery of seabirds was limited by environmental conditions in the Gulf of Alaska – particularly at the Barren Islands, whose murre populations were hard hit by the spill (Piatt et al., 1990; Roseneau et al., 1997). But how do we compare the robustness of seabird populations among colonies in Cook Inlet, or among regions in Alaska? As illustrated in the preceeding text, any one parameter we choose to examine may be biased depending on the form of its relationship to food supply and whether it is highly variable or relatively constant in the face of environmental change. Similarly, any one species may be more sensitive to different aspects of environmental change than another species.
348
Long-Term Ecological Change in the Northern Gulf of Alaska
Figure 4.44: Normalized deviations from average of seabird breeding and behavioral parameters at Chisik, Gull, and Barren islands, 1995–1999. Deviations have been arbitrarily ranked by magnitude from most positive (left) to most negative (right). Data from Piatt, 2002a and Roseneau et al., 2000.
One way to assess and compare the performance of seabirds among colonies in Cook Inlet is to compare the deviation of parameter values at any one colony with the average of all three colonies combined (Fig. 4.44). For example, the average breeding success of kittiwakes in 15 colony-years (three colonies in 1995–1999) of study was 0.312 chicks/pair (Shultz et al., 2002). Success of kittiwakes at the Barrens was lower than this in 2 years of study and higher in 3 years. Success was higher than this average in all 5 years at Gull Island, and much lower than this average in all 5 years at Chisik. Similarly, we calculated deviations from average in many different behavioural, reproductive and physiological parameters (e.g., attendance, feeding rate, growth rate, fledging success, etc., for both murres and kittiwakes; from tables in Shultz et al., 2002; Van Pelt et al., 2002), standardized the deviations, and arbitrarily ranked them from largest to smallest at each colony so that we could examine them
Long-Term Change 349
all together (Fig. 4.44). In total, we can compare 266 parameter deviations (about 20 species parameters by year, colony, with some missing values). This provides an integrated assessment of how well seabirds were doing at each colony during the 5 years of study. Analysis reveals (Fig. 4.44) that seabirds at Gull Island do better than average most of the time (mean deviation = +0.24), while those at Chisik do poorly most of the time (mean deviation = −0.37). At the Barrens, measured parameters were above average slightly more often than they were below average (mean deviation = +0.09). Judging from the range of parameter values we observed at Chisik and Gull islands, and in comparing these with values obtained in similar studies on murres and kittiwakes conducted elsewhere under a wide range of conditions (e.g., Hamer et al., 1993; Uttley et al., 1994; Monaghan et al., 1994; Dragoo et al., 2000; Gill, 1999 etc.), it is clear that Chisik and Gull exemplify the extremes of failing and thriving colonies, respectively, in the North Pacific and Atlantic oceans. Therefore, this analysis provides a calibration for seabird performance at the Barren Islands and suggests that murres and kittiwakes there are doing modestly well. This conclusion is corroborated by data on population trends (Fig. 4.42), and once again suggests that murre populations at the Barren Islands should not be limited by food in their recovery from the oil spill. Indeed, there is a strong correlation between our indices of population health and observed population trends (Fig. 4.45). This relationship seems intuitively obvious. Our parameter indices integrate a suite of values that include behavioral, physiological and biological measurements. The combination of these parameters – and many more unmeasured parameters – is ultimately what determines whether a population
Figure 4.45: Average parameter index (from Fig. 4.44) versus population trend (from Fig. 4. 42) for common murres and black-legged kittiwakes at Chisik, Gull, and Barren islands.
350
Long-Term Ecological Change in the Northern Gulf of Alaska
will increase or decrease. Similarly, population trend represents an integration of all factors influencing population biology. The parameter index also offers an instantaneous measure of the health of seabird populations, whereas population trend data, by definition, needs to be collected over many years to establish a trend. One year’s sampling may be all that is needed to assess the status of populations (e.g., parameter indices for any one year in 1995–1999 at Chisik were always negative, ranging from −0.24 to −0.62; whereas at Gull they were always positive, ranging from +0.19 to +0.29). In contrast, census data can be highly variable among consecutive years, and may need to be collected for a decade or longer to establish trends (Dragoo et al., 2000). Finally, census information will be misleading if study plots used for census purposes become saturated with breeding birds, or breeding habitat on a colony is saturated (as we observed on Gull Island, see “Population Dynamics” above).
4.8.5. Long-term Changes in the Gulf of Alaska Marine Environment Are environmental conditions in lower Cook Inlet typical of those found elsewhere in the Gulf of Alaska or Bering Sea? In the absence of quantitative long-term data on food supplies, one way to answer that question is to compare seabird parameter indices from Cook Inlet to those gathered in other areas of Alaska. Unfortunately, the full suite of parameters measured in our studies has rarely been surveyed at other colonies. However, one parameter that has been widely reported is breeding success of black-legged kittiwakes (Table 4.4). We have established that kittiwake breeding success exhibits a strong, sigmoidal response to prey density (Fig. 4.39) and from that quantitative relationship we also know that kittiwake breeding success above 0.46 chicks/pair represents asymptotic reproduction unlimited by food supply, while breeding success below 0.015 chicks/pair represents asymptotic failure to reproduce under conditions of severe food deprivation (Piatt, 2002a). Breeding success that ranges between 0.015 and 0.46 chicks/pair represents reproduction that is limited to some degree by food supplies that hover around the threshold (Piatt, 2002a). Using these criteria, we can indirectly assess the historical status of food supplies for seabirds in Alaska by examining the breeding success of kittiwakes in past years (Fig. 4.46, data from Hatch et al., 1993; Dragoo et al., 2000). Prior to the regime shift that occurred in the late 1970s (Francis et al., 1998), kittiwake productivity in both the Gulf of Alaska and Bering Sea was very similar: only a small proportion (5–6%) of colony-years were food-deprived failures, slightly more than half (55–56%) were limited to some degree by food supply, and a large proportion (38–40%) were unlimited by food. After the regime shift, but mostly prior to the Exxon Valdez oil spill in 1989, there was a marked change in kittiwake productivity (Fig. 4.46). The frequency of food-deprived failures in the 1980s increased six-fold
Long-Term Change 351
Figure 4.46: Historical breeding success (brs) of black-legged kittiwakes in the Gulf of Alaska (GOA) and Bering Sea (BS), categorized by functional relationships with food supply as “Deprived” (brs <0.015 chicks/pair), "Limited" (0.015 brs <0.46 chicks/pair), and "Unlimited" (brs >0.46 chicks/pair). Data from Hatch et al., 1993, Zador et al., 1997, Dragoo et al., 2000, and Roseneau et al., 2000.
352
Long-Term Ecological Change in the Northern Gulf of Alaska
(to 37%) in the Gulf of Alaska, while the frequency of unlimited production declined by half (to 17%). A similar, but less pronounced, shift occurred in the Bering Sea. In the 1990s, there was a significant improvement in feeding conditions in the Gulf of Alaska; whereas the frequency of unlimited production remained the same (at 17%), food-deprived failures decreased from 37 to 30% while production that was food-limited to some degree increased from 46 to 54%. By comparison, however, conditions in Cook Inlet had improved substantially more than the Gulf as a whole (in which Cook Inlet data are included). Of 24 colony-years of production in the 1990s, only 21% were food-deprived failures, 50% were limited to some degree, and 29% were unlimited by food. Conditions in the Bering Sea improved even more than in the Gulf of Alaska. In summary, this analysis supports the hypothesis that a regime shift in the late 1970s reduced food availability to seabirds in the 1980s and 1990s, resulting in widespread population declines, lower breeding success, and mass mortality events (Piatt and Anderson, 1996; Francis et al., 1998). The evidence further suggests that there was a slight improvement in feeding conditions in the 1990s and that conditions in Cook Inlet are better than those in the Gulf as a whole. In general, however, current conditions continue to be depressed compared to the 1970s. While seabirds in Cook Inlet colonies may have already recovered numerically to levels observed prior to the Exxon Valdez oil spill, it is still not clear whether conditions elsewhere in the Gulf of Alaska would have supported similar rates of recovery during the 1990s.
4.9. Marine Mammal Populations Alan M. Springer, Sara J. Iverson, and James L. Bodkin 4.9.1. Harbor Seals Harbor seals are the most widely distributed pinniped in the world, occurring in both the North Pacific and North Atlantic oceans. Their range is nearly continuous around the rim of the North Pacific from San Ignatio Lagoon, Mexico (27°N) to Hokkaido, Japan (43°N), and extends into the eastern Bering Sea as far north as Kuskokwim Bay (60°N) (Fig. 4.47). Today, there are some 36,000 harbor seals in the northern GOA between Kayak I. and the Copper River Delta in the east to False Pass at the end of the Alaska Peninsula in the west (Boveng et al., 2003). This total is much lower than in the middle of the past century because of the eradication programs and commercial harvests described in Section 3.5, and unexplained losses since the early 1970s (Fig. 4.48).
Long-Term Change 353
Figure 4.47: Harbor seal distribution (in yellow) in the North Pacific Ocean.
These more recent declines are best documented at Tugidak I., a small island southwest of Kodiak I. (in the Kodiak Archipelago), which was formerly the largest harbor seal rookery in the Gulf of Alaska, perhaps the largest in the world. The abundance of seals there prior to 1964 and the inception of the commercial harvest was estimated to have been approximately 20,500 (Pitcher, 1990). A population simulation model 100 Prince William Sound
90 % MAXIMUM COUNT
80 70 60 50 40 30 Tugidak Island
20 10
1962 1964 1966 1968 1970 1972 1974 1976 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000
0
YEAR
Figure 4.48: Trends in harbor seal populations on Tugidak Is. and in Prince William Sound. Data from Pitcher (1990), Frost et al. (1999), and Jemison et al. (2006).
354
Long-Term Ecological Change in the Northern Gulf of Alaska
indicated that the commercial harvests of some 16,000 pups in 1964–1972 should have caused a 30% decline to about 15,000 animals. Following protection, numbers should have stabilized and then begun to grow during the mid to late 1970s. Yet, the abundance of harbor seals at Tugidak I. in 1976, when systematic counts were initiated, was only about 9300, or just 60% of the modeled estimate. The population continued to collapse at a precipitous rate of 19% per year through the end of the decade, before slowing to about 7% per year in the 1980s and early 1990s (Pitcher, 1990; Jemison et al., 2006). By 1993, there were fewer than 800 seals left at Tugidak I. Between 1993 and 1998, the population recovered somewhat (about 10% per year) before leveling off at about 1400 animals (Fig. 4.48). The collapse of harbor seals at Tugidak I., at least the unexplained portion from the early 1970s onward, and subsequent partial recovery appears to fairly represent trends at nearby rookeries on Kodiak I., where seals declined by an average of 66% between the mid-1970s and early 1990s, and have since then increased by nearly 40% (Lewis et al., 1996; Small et al., 2003). The status of seals on Kodiak I. prior to 1975 is not known because systematic counts were not made. Yet, despite increases at several sites in recent years, the overall abundance of harbor seals in the Kodiak Archipelago remains at just 15–35% of that in the mid-1970s. And, compared to 50–60 years ago, today’s abundance is certainly a much smaller percentage. Some sense of the magnitude of this difference can be gained from the fact that the current abundance of harbor seals at Tugidak I. is still less than 10% of that prior to exploitation and the ensuing unexplained declines. Likewise, the decline of harbor seals in the northwestern GOA typifies trends farther west. On Otter I. (Pribilof Is., east Bering Sea), they declined by about 40% between 1974 (first count) and 1978 and by an additional 70% between 1978 and 1996 (Jemison et al., 2006); seals declined at a rate of about 3.5% per year along the north side of Alaska Peninsula between 1975 (first count) and 1995 (Withrow and Loughlin, 1996); and despite an annual increase in Nanvak Bay (Bristol Bay, east Bering Sea) of about 2.1% per year from 1990 to 2000, the maximum count in 2000 was just 20% of the highest count in 1975 (first count) (L. Jemison in Small et al., 2003). Systematic counts of harbor seals have not been made in the Aleutian Islands, but conspicuous declines have been observed by scientists working there during the past 30 years (J. Estes, unpublished obs.). Harbor seals in eastern and central PWS declined also – by 63% between 1984 and 1997 and by about 3% per year since then (Frost et al., 1999; Ver Hoef and Frost, 2003). The lack of systematic counts prior to 1984 makes it impossible to know whether the decline began before then, perhaps concurrent with declines in the Kodiak I. region. Notably, the decline in PWS was underway by the time of the 1989 oil spill, and the magnitude of the decline from 1984 to 1990 (50%) was similar to that in the same time period at Tugidak I. (60%). Numbers in PWS continued to decline after 1990, but increased at Tugidak I.
Long-Term Change 355
Scientists believed initially that about 320 harbor seals were killed in PWS by oil from the spill (Frost et al., 1994). That estimate was based on changes in numbers at 25 haul-out sites along a standard trend-count route established in the early 1980s. Subsequent monitoring of seals at those sites led to the further conclusion that the population in the sound continued to decline through the 1990s (Frost et al., 1999). However, this interpretation was challenged based on more widespread surveys that included additional areas of PWS, particularly the western sound and glacial haul-out sites (Hoover-Miller et al., 2001). These scientists contend that there is no evidence of significant direct mortality of seals by oil in 1989, and that surveys in 1989 and since then along the original trend-count route failed to account for the mobility of seals – they suggest that many seals likely moved to other parts of the sound to avoid oiled habitats. In support of their belief, they point out the large interannual variability in numbers of seals at individual haul-outs, suggesting short-term local movements, and that since 1991 numbers of seals at glacial sites have tended to increase while numbers at land sites have tended to decrease. However, studies of radio-tagged seals in the spill area did not show adult movements into the northern fjord areas of Prince William Sound where counts have been increasing. Also, dead harbor seals often sink, as may have happened following the spill. In contrast to the widespread declines of harbor seals throughout the northern and western GOA, Aleutian Is., and Bering Sea since the 1970s, trends in abundance in southeastern Alaska and British Columbia have been variable but generally opposite. In Glacier Bay in northern southeastern Alaska, a dramatic increase occurred between the mid-1970s and mid-1980s, followed by declines beginning in the early 1990s (Mathews and Pendleton, 1997, 2000). Harbor seals near Ketchikan in southern southeastern Alaska increased rapidly from the early 1980s through the mid-1990s, then slowed through the end of the decade, whereas seals near Sitka increased little if any over this same period (Small et al., 2003). In British Columbia, harbor seals had been reduced to about 9000–10,500 by 1970 because of irresponsible bounty programs and heavy commercial harvests, particularly in the 1950s and 1960s. Following protection in 1970, harbor seals increased at the rate of about 12% per year (approximately the maximum intrinsic growth rate), and they numbered 75,000–88,000 by 1988 (Olesiuk et al., 1990; Olesiuk, 1999). The growth rate then slowed in the early 1990s, and the population numbered about 108,000 by the end of the decade (Olesiuk, 1999).
4.9.2. Steller Sea Lions Steller sea lions also are found around the entire rim of the North Pacific Ocean, from central California to Japan. They have recently been separated into three genetically distinct stocks: the Eastern Stock along the west coast of North America from California to southeastern Alaska; the Central Stock from Kayak Island west to the
356
Long-Term Ecological Change in the Northern Gulf of Alaska
Figure 4.49: Steller sea lion distribution (in yellow) in the North Pacific Ocean. The red dotted line separates the range of the Western Stock from that of the Eastern Stock.
Commander Islands at the western end of the Aleutian Archipelago; and the Western Stock from the Commander Islands and western Bering Sea south to Japan (Bickham et al., 1996; Baker et al., 2005) (Fig. 4.49). Official recognition of the distinction between the Central and Western Stocks has not yet occurred, and they are still grouped together as the “Western Stock” under the Endangered Species Act. In the following discussion, “Western Stock” refers to animals of the Western Stock in Alaska. The direct and incidental killing of sea lions in the 1950s through the first half of the 1980s (described in Section 3.5) likely had effects on local abundances, but in the mid-1970s the Western Stock was considered to be at or near the pre-exploitation level of abundance (Interagency Task Group, 1978). Since then, however, it has collapsed. Although the Western Stock was formerly the largest stock by far, with about 170,000 individuals in the mid-1970s, it is just 30,000–40,000 today and was listed as “Endangered” under the Endangered Species Act in 1997. The overall rate of decline of the Western Stock from the late 1970s to the mid-1980s was about 6% per year. It accelerated rapidly to nearly 16% per year during the second half of the 1980s, when as many as 8000–25,000 animals disappeared from the population each year, before dropping to just over 5% per year throughout the 1990s (Paine et al., 2003). Beginning in the late 1990s, numbers may even have increased slightly at some rookeries in the GOA. Within the range of the Western Stock, which in Alaska spans some 2000 km from Kayak I. to the Near Islands, the pattern of population decline from the 1980s on has varied, although precise details for most locations are lacking. One problem is that numbers of sea lions at rookeries change considerably from month to month, and
Long-Term Change 357
unless counts are made on similar dates between years, the results may not be comparable. This is a particular issue with counts from early years that were not always made at the optimal time. The other, larger problem is that counts at all rookeries were infrequent in the 1960s through the middle of the 1980s, so the time when declines began is uncertain for some regions. It is reasonably certain that in the GOA, from the Kenai Peninsula to Unimak Pass, sea lion abundance was stable from the 1950s through the 1970s and collapsed in the 1980s (Fig. 4.50(a)). The picture is less clear for the eastern GOA, but it can be safely said that sea lions at the only two rookeries in that region did not decline until about 10 years later. However, whereas non-pup abundance fell, the number of pups produced each year did not, suggesting that perhaps the “resident” population remained stable while numbers of visiting sub-adults from other regions of the GOA declined (K. Pitcher, pers. comm.). This interpretation is supported by counts at two prominent haul outs, The Needle in Montague Strait (PWS) and Cape St. Elias on Kayak I., where numbers plummeted by an average of at least 80% in the interval 1989–1990 to 1996 (Fig. 4.50(a)). The population decline in the Aleutian Is. apparently followed a pattern similar to that in the central and western GOA, except that in the eastern Aleutians it may have begun somewhat earlier, perhaps by the early to mid-1970s. However, the perception that the unexplained portion of the decline of the Western Stock began in the eastern Aleutians may be colored somewhat by changes that likely resulted from the massive harvests of pups (the entire annual production for three straight years) and associated disturbance of adults at the largest of the rookeries in that region in 1970–1972. In contrast to these trends in the Western Stock, the Eastern Stock in southeast Alaska, although historically much smaller than elsewhere, has been increasing at 1–2% per year over the past 40 years and now numbers about 20,000, the highest in recorded history (Fig. 4.50(b)). Likewise, sea lions of the Eastern Stock in British Columbia have also increased considerably in the past three decades, recovering from effects of control programs there in the 1950s and 1960s that reduced their number from about 14,000 when the first survey was conducted in 1913 to about 3400 by the early 1970s (Bigg, 1985; Olesiuk, 2003). Since being protected in 1970, the Steller sea lion population in British Columbia has increased at a rate of about 3% per year and by 2002 numbers on rookeries had grown to about 8700 (DFO, 2003; Olesiuk, 2003).
4.9.3. Sea Otters The historic distribution of sea otters included shallow-water, ice-free habitats around the North Pacific Rim between northern Japan and central Baja California, Mexico (Fig. 4.51). Factors that govern sea otter populations in the absence of human influence
358
Long-Term Ecological Change in the Northern Gulf of Alaska
% maximum count
(a) 100 80 60 40
1995 1997 1999 2001
1987 1989 1991 1993
1977 1979 1981 1983 1985
1967 1969 1971 1973 1975
1959 1961 1963 1965
0
1957
20
Year
% maximum count
(b)
100 80 60 40 20
1957 1959 1961 1963 1965 1967 1969 1971 1973 1975 1977 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2001
0
Year
Figure 4.50: (a) Trends in Steller sea lion populations in the central and western GOA (blue) and the northeastern GOA (red). (b) Trends in sea lion pup counts in southeastern Alaska. Values for the central and western GOA are proportional annual means (±SE) of counts made during late June to July at the rookeries Clubbing Rock, Pinnacle Rock, Chernabura Is., Atkins Is., Chowiet Is., Chirikof Is., Marmot Is., Sugarloaf Is., Outer Pye Is., and Chiswell Is.; and for the northeastern GOA at the rookeries Wooded Is. and Seal Rocks. Data from the National Marine Mammal Laboratory, Seattle, WA. Data for southeastern Alaska from Bigg (1985), Calkins et al. (1999), D. Calkins and K. Pitcher (unpublished data).
Long-Term Change 359
are not well known. Evidence from human midden sites in the Aleutian Islands indicates that sea otter populations fluctuated markedly over a period of 2500 years, presumably in response to human exploitation (Simenstad et al., 1978) or other predators, as appears to have been the case most recently (see below). By the time sea otters were protected from the commercial fur harvest by the International Fur Treaty in 1911, only 11 geographically isolated populations persisted between California and the Kuril Islands in Russia, from which the species recovered during the following century (see Section 3.5). After protection, sea otters expanded their range and abundance throughout Alaska during much of the twentieth century. Although relatively few surveys were conducted during the early phases of recovery, estimated rates of growth among remnant populations range from about 9–13% in the GOA and Aleutian Is. (Bodkin et al., 1999). The population at Amchitka I. was likely the first to attain full recovery, reaching an initial equilibrium density by 1945 (Estes et al., 1989). Probably the best information on the process of recovery of a population in the GOA is from PWS, where a small number of animals, probably fewer than 50 in the early 1900s, survived the fur harvest (Lensink, 1960). By 1959, the population had grown to about 1000 animals, although it may have been reduced to about half that number by the Good Friday earthquake in 1964 (Kenyon, 1969). Throughout most of the twentieth century, the population demonstrated a fairly uniform growth trajectory, with an average annual increase of about 10% (Bodkin et al., 1999), resulting in an estimated population size of 9284 animals in 2003.
Figure 4.51: Sea otter distribution (in yellow) in the North Pacific Ocean.
360
Long-Term Ecological Change in the Northern Gulf of Alaska
Three stocks of sea otters are now recognized in Alaska (Gorbics and Bodkin, 2001). The Southwest Stock includes the Aleutian Archipelago and the Alaska Peninsula east to Kamishak Bay in lower Cook Inlet; the Southcentral Stock includes the Kenai Peninsula, PWS, and south to Cape Yakataga; and the Southeastern Stock ranges south from Cape Yakataga. Information on population sizes and trends indicates a severely declining Southwest Stock, a stable Southcentral Stock, and an increasing Southeast Stock (Table 4.5). ●
●
●
Southwest Stock: Despite a long period of steady expansion and population growth through the 1980s, dramatic declines began by the that of that decade, averaging about 17–25% per year throughout the Aleutian Is. (Estes et al., 1998; Doroff et al., 2003). Declines of similar magnitude appear to have extended eastward into the GOA, from the Alaska Peninsula to near Kodiak I. (USFWS stock assessment reports). Southcentral Stock: While sea otter mortality in PWS was extensive as a result of the 1989 Exxon Valdez oil spill, population recovery was evident by increases in abundance in the spill area (Bodkin et al., 2002), although in some heavily oiled areas increases were not evident. Four sound-wide surveys conducted from 1994 to 2003 indicated that the population has been stable at about 8500 individuals. Such stability appears to be characteristic throughout much of the range of this stock, although increases are evident locally where unoccupied habitat occurs such as in lower Cook Inlet. (Fig. 4.52). Southeast Stock: The southeast Alaska population resulted from translocation of 412 sea otters from Amchitka I. in the Aleutians and PWS between 1965 and 1969 (Jameson et al., 1982). Subsequent surveys indicated an average annual growth rate of about 18% through 1995. A survey of southeast Alaska completed in 2003 indicated reduced growth since 1995 (J. Bodkin, unpublished data), except in Glacier Bay where the population increased from zero in 1995 to more than 1200 in 2002, undoubtedly a consequence of immigration from nearby occupied habitats in Icy Strait and Cross Sound.
Table 4.5: The status of Alaskan sea otter populations. Stock
Status
Southwest Southcentral Southeast
Declining Stable Increasing
Estimated Number 41,500 16,500 10,000 68,000
Sources USFWS, USGS USGS, USFWS USGS, USFWS
Long-Term Change 361
Figure 4.52: Estimated size of the Prince William Sound sea otter populations from four aerial surveys, 1994–2003. Data from J. Bodkin (USGS, unpublished data).
4.9.4. Potential Causes of Marine Mammal Population Change The unexplained collapse of the Western Stock of Steller sea lions in the northern North Pacific Ocean beginning in the late 1970s has caused a great deal of legal, scientific, and conservation activity in recent years because of three things: (1) the magnitude of the decline, about 80%, that led to classification of the stock as “Endangered” under the Endangered Species Act in 1997; (2) the widely held belief that the decline was caused by food shortage; and (3) the perception that sea lions and commercial fisheries compete for common prey, particularly walleye pollock. One of the most startling outcomes was the appropriation of $120,000,000 of public money to study the problem between 1999 and 2003. But sea lions are not the only species of marine mammal of concern in the North Pacific – harbor seals and sea otters have also collapsed by 80–90% over a similar geographic scale in the GOA, Aleutian Islands, and Bering Sea. In addition, the fur seal herd on the Pribilof Is. has declined by nearly two thirds since the 1960s and is continuing to fall more than 6% per year since 1998 (York and Hartley, 1981; Trites and Larkin, 1989; Towell, 2004). Among the many species of marine mammals in the North Pacific, including great whales, dolphins, porpoises, and other pinnipeds, these four species stand out because of the magnitudes and rates of their declines. None of the other species are known to have undergone such changes, and some, such as certain whales, increased in number during this time.
362
Long-Term Ecological Change in the Northern Gulf of Alaska
Disease, contamination, redistribution, subsistence use, and commercial harvests have been discounted as important causes of any of the major declines of pinnipeds or sea otters in the North Pacific since the 1970s (Estes et al., 1998; Loughlin and York, 2000; DeMaster and Atkinson, 2002; Paine et al., 2003), except in the particular case of sea otters in PWS. As noted above, direct and incidental killing were significant causes of mortality for harbor seals and sea lions in the 1950s and 1960s. This likely had broadscale effects on the abundance of harbor seals and sea lions in the GOA, particularly in southeast Alaska and British Columbia, and local effects at several rookeries in the Kodiak Archipelago and eastern Aleutian Is. Also, a misguided experimental harvest of fur seals on the Pribilof Is. in 1958–1965 caused much of the initial decline throughout the 1960s and the first half of 1970s, and entanglement in marine debris in the 1970s and 1980s may have further contributed to early declines (Fowler, 1987; Swartzman et al., 1990). Since the early to mid-1970s, however, intentional and accidental killing probably have not had effects on populations of any of these species locally or rangewide that were large enough to explain the magnitude and rate of decline they all suffered since then. Nutritional limitation, because of general lack of prey or because of the lack of prey of high nutritional quality, has been the leading hypothesis for the unexplained portions of the pinniped declines that began with harbor seals in the early 1970s (Alverson, 1992; Anonymous, 1993; Trites and Donnelly, 2003). Initially, it was thought that food shortage had two effects that combined to cause the declines – starvation and reduced productivity. Although there is some evidence for lowered productivity of sea lions, population models indicate that this was not sufficient to explain the collapse without greatly elevated mortality of juveniles and adults (Pascual and Adkison, 1994; York, 1994; Pitcher et al., 1998; Paine et al., 2003). Yet, despite the huge annual loss of sea lions during the period of greatest decline in the mid to late 1980s, when as many as 25,000 animals disappeared from the population each year, a lack of reports of distressed or emaciated sea lions (or of harbor seals or fur seals), on rookeries or haulouts in summer, of thin animals taken by subsistence hunters in fall, winter, and spring, or unusual numbers of dead sea lions or seals on beaches in any season, argued against starvation or disease as a cause of high mortality (Hoover, 1988). Mortality from nutritional limitation, disease, and pollution typically results in large numbers of stranded carcasses, such as occurred during the morbillivirus outbreak in pinnipeds in the North Sea (see Section 3.5) and during El Niño events in the North Pacific when many animals starve. Still, little is known about what happens to pinnipeds at sea at many times of the year, especially in the juvenile age class, which is likely to be hit hardest by food shortage. Yet, juvenile mortality alone could not account for the magnitude and rate of decline that was observed in harbor seal, sea lion, or sea otter populations in the 1970s, 1980s, and 1990s.
Long-Term Change 363
Lack of support for the starvation idea led to an important modification of the nutritional stress hypothesis, applied to the particular case of sea lions. The modified explanation was that nutrition was not so poor as to cause sea lions to die outright, but was sufficiently poor to lower productivity and individual fitness and make them much more susceptible to other sources of mortality such as disease and predation (Trites and Donnelly, 2003). Trites and Donnelly further modified the hypothesis to say that sea lions were physiologically compromised not because of an overall lack of prey, but because of a lack of prey with high fat content. Because disease apparently was not important, that leaves predation as the most plausible “other source,” or at least a dominant source among a suite of possible factors. The recent review by the National Research Council (Paine et al., 2003) also failed to support the notion that food shortage or disease was to blame for patterns of decline in the 1990s, but instead implicated top-down processes such as predation. Paine et al. did not elaborate on the particular mechanisms of the sea lion collapse in the 1980s, beyond saying it likely involved multiple agents. Their review also failed to support the idea that commercial fisheries were responsible for perceived food shortages during the decline. More recently still, predation by killer whales has been proposed as a competing, unified hypothesis to nutritional stress to explain much of the decline of all three species of pinnipeds (Springer et al., 2003; Estes et al., 2004; Williams et al., 2004). It follows from compelling arguments and the weight of evidence that killer whale predation was responsible for the collapse of the sea otter population in the Aleutian Is. (Estes et al., 1998). In that case, thorough studies were made of disease, redistribution, direct mortality, food availability, and other factors that might have been important to the otter population, but they were rejected as contributing in any important way to the decline. According to the hypothesis, elevated killer whale predation was not caused by lowered individual fitness of pinnipeds and otters, and thus increased susceptibility to predation, but by a decline in the abundance of other prey of killer whales, the great whales, that was brought on by industrial whaling following the end of World War II. So, what is the evidence for these two ideas – lack of food or predation, or both? Are there similarities between observations of the various species? And, what are the timelines? As we shall see, getting the timing right of when and where events occurred and trends began is crucial to an understanding of the cause of the collapse of these species of megafauna in the northern North Pacific.
4.9.5. Harbor Seal Decline The commercial and bounty harvests of harbor seals in the GOA in the 1950s and 1960s varied geographically in important ways. Very little seal killing occurred on the
364
Long-Term Ecological Change in the Northern Gulf of Alaska
south side of the Alaska Peninsula or in the Aleutian Is. Pups were targeted in summer in the eastern Bering Sea (primarily Port Moller and Port Heiden on the north side of the Alaska Peninsula); on Tugidak I., where 90% of the seals killed were pups; and elsewhere in the northern GOA, including Yakutat, PWS, and Kodiak I. (K. Pitcher, pers. comm.). In southeast Alaska, however, most seal hunting was done by fishermen in winter, the off-season for fishing, and older seals including adults were taken. Likewise, all age classes were killed in British Columbia. Although juvenile mortality accounts for most mortality in populations, harbor seal populations, like those of other pinnipeds and sea otters, are particularly sensitive to adult mortality. Therefore, effects of killing pups versus adults on population demographics and stability must be born in mind when trying to understand the highly variable patterns in abundance of seals from British Columbia to the Bering Sea in the past 3–4 decades. In British Columbia, and probably throughout the broader region including southeast Alaska, reduction programs and commercial and bounty harvests decimated seals in the 1950s and 1960s, and abundances were very low by the beginning of the 1970s. In contrast, harbor seal populations in the northern GOA were not affected as greatly, even though substantial numbers of seals were killed. Harvests were localized at particular rookeries and targeted newly born pups, many of which would not have survived their first year of independence anyway, let alone survived to adulthood. The Tugidak Is. case discussed above illustrates this difference nicely – the capture of some 16,000 pups over a period of 9 years should have resulted in a population decline of less than 30%. At Tugidak Is., this initial portion of the harbor seal decline is therefore considered to be “explained” by the pup harvest. However, the continuation and acceleration of the decline after 1972 and the cessation of the harvest remains “unexplained.” Of particular note is that this portion of the overall decline began well in advance of the climate regime shift in the second half of the 1970s. Elsewhere in the GOA and Bering Sea, harbor seals were clearly in decline from at least 1975–1976 onward, the period when counts at most rookeries were first made. The idea that unexplained declines began earlier than the mid-1970s is further supported by counts at Otter I. (Pribilof Is.), which registered a 40% drop in abundance between 1974 and 1978. Information on other aspects of harbor seal biology over time in the GOA is limited to pupping dates at Tugidak I. Pupping dates there, as indexed by the date of maximum pup numbers in the pupping period, were about 10 days later in 1976–1979 than in either 1964 or 1994–1998 (~22–25 June compared to 11–13 June; Jemison and Kelly, 2001). The authors suggested two possible explanations – differences in timing of when food was available to pregnant females and overall differences in the quantity or quality of prey. A third factor that can influence pupping dates, age structure of the population (older females tend to pup earlier than younger ones), was discounted: Jemison and Kelly argued that the large commercial pup harvests at Tugidak Is. in the 1960s and early 1970s should have skewed the population towards older
Long-Term Change 365
females by the mid-1970s and, all other things being equal, they should have given birth earlier on average than a younger population. Unfortunately, there is no information on the actual age structure of the population at any time during this interval. The conclusion Jemison and Kelly reached in their study at Tugidak Is. was that nutritional stress explained the later pupping dates in the mid to late 1970s. So, what is the likelihood, based on information on harbor seals from Alaska and elsewhere, as well as on their survival strategies, that (1) differences as great as this in pupping dates would be due to differences in food availability, (2) such differences in pupping dates would have demographic relevance because of related differences in the fitness of pups, and (3) the later pupping dates in the 1970s reflected decreased fitness of adult females, and indeed the population in general (adults, juveniles, and young of year) at other times of the year so severe that individuals would have been less likely to reproduce, to die of starvation, or to succumb to disease or predators? In aggregate, could these possible effects explain the population collapse? These questions are not easily answered, given the limited information available from the GOA. Indeed, we know practically nothing about many aspects of harbor seal biology during difficult and critical times of the year outside the breeding season, for example, during winter for all age classes and during the first few months of independence for pups when they must learn to forage for themselves and learn other survival skills necessary to live in an extreme and uncertain environment. However, we can look at data on harbor seals from another population that is well studied as an example. On Sable Is., Nova Scotia, the harbor seal population has been declining for 15 years, plummeting from about 625 animals in 1989 to fewer than 32 by 1997 and about 10 today (Lucas and Stobo, 2000; Bowen et al., 2003; W. D. Bowen and S. Iverson, unpublished data). The decline is similar in relative magnitude and duration to that at Tugidak Is., except that the population at Sable Is., now at an all-time low, shows no sign of recovery. Over this time at Sable I., there were no indications that food shortage during the breeding season (or in the few months prior) contributed to the population decline. For example, no interannual or long-term trends were seen in maternal postpartum mass, pup birth mass, the rate of mass gain by pups during the lactation period, or the mass of pups at weaning (Bowen et al., 2003). But, as with Tugidak I., there were differences in pupping dates, which increased by 7 days over the course of the 1990s, suggesting that implantation in females became progressively later, perhaps due to nutritional stress (Bowen et al., 2003). Support for nutritional stress of harbor seals at Sable Is. is found in information on dietary overlap and competition. Harbor seals on Sable consume primarily sand lance, flatfishes, cods, and herring, the same suite of prey consumed by the coexisting, larger grey seals (Halichoerus grypus) (Bowen and Harrison, 1994, 1996). The Sable Is. grey seal population has been growing exponentially for the past 40 years, and, whereas grey seals outnumbered harbor seals by about 20:1 in the late 1980s, they
366
Long-Term Ecological Change in the Northern Gulf of Alaska
outnumbered them in the late 1990s by more than 500:1 (Bowen et al., 2003). Thus, grey seals may have limited prey available to harbor seals, at least during the fall, when females are beginning to implant and gestate. The observations from Sable Is. suggest that the answer to question 1 above is that nutritional status of female harbor seals likely is the explanation for variability in pupping dates on Tugidak Is. However, the answers to questions 2 and 3 are that the likelihood is low that later pupping dates led to reduced fitness of pups at weaning or was related to debilitating physiological condition of females in the months prior to or during the lactation period. Although harbor seals are buffered by their economical, largely capital-based energetic strategy during the breeding season, nutritional stress at other times of the year might have affected not only implantation date on Sable Is., but also female fecundity and, thus, may have contributed to the population decline (Bowen et al., 2003). However, even the complete cessation of pupping at Sable Is. would not have explained the coincident collapse of the population. In the GOA, reduced fecundity was apparently not the cause of the population collapse on Tugidak I., as the pregnancy rate of mature females in the GOA in the mid-1970s was 92%, similar to rates observed in healthy populations elsewhere (Pitcher, 1990). Another change in the population at Sable Is. was a decline in recruitment, indicating a decline in juvenile survival. This too might have been caused by nutritional stress, but in this case it was more likely caused by predation. Indeed, the bulk of the decline in harbor seals on Sable Is. has been attributed to predation by sharks, which take animals of both sexes and all age classes, including adult females, which are killed in disproportionate numbers (Lucas and Stobo, 2000; Bowen et al., 2003). There was an obvious decline in recruitment at Tugidak Is. as well, and it too might have been caused by reduced fitness of pups and juveniles that was caused by nutritional stress. The transition to independence by pups is a critical time, and mortality of young-of-year and juvenile harbor seals and other pinnipeds in general is high. There is no indication of fitness being so low that young harbor seals in the GOA died outright, but there are few observations of seals in winter that would permit an evaluation of physiological condition. The only indication of the health of seals during winter is the lack of reports of weak or thin animals taken by subsistence hunters or of distressed or dead seals at haul outs or elsewhere. If fitness of juvenile harbor seals in the GOA was compromised by a shortage of prey, it perhaps made them more susceptible to predation. Although there is the possibility of an interaction between the two, the example at Sable Is. suggests that lowered fitness is not a prerequisite for increased predation on juveniles, as sharks regularly take healthy adult females that should be less vulnerable than even the fittest juveniles. Evidence of nutritional stress of harbor seals in the GOA based on diet information is speculative. In general, harbor seals eat the usual forage species available in the
Long-Term Change 367
North Pacific, primarily including pollock, Pacific herring, capelin, eulachon, sand lance, Pacific cod, salmon, octopus, and squid, as well as numerous miscellaneous species (Pitcher, 1980; Iverson et al., 2003). But as noted above, the importance of the various species to harbor seals changes considerably from place to place and from season to season. For example, in both the Kodiak Is. region and in PWS in the mid-1970s, salmon were important in summer, but were absent at all other times of the year; Pacific herring were important in all seasons in PWS but never around Kodiak; and octopus were important in all seasons in both places. Likewise in PWS in the 1990s, Pacific herring and pollock were important to harbor seals in all locations sampled, but the seals were eating large individuals in the southern sound and small individuals in the northern and eastern sound. Thus, harbor seals are clearly generalist feeders, selecting from a diversity of prey and apparently choosing among them those that are most abundant or energetically profitable at any time and place. The diets and body condition of juvenile harbor seals in PWS were documented in the late 1990s (1997 through 2000) and demonstrated the degree to which they make use of diverse prey species (Iverson et al., 2003). Across all 4 years, newly weaned pups were consistently heavy with very high body fat contents (42%) compared to any other harbor seal populations studied, indicating that lactating females were not food limited. Yearlings and 2–3-year-old juveniles also had comparatively and consistently high body fat contents across years, suggesting no food limitation, at least in spring and early summer. Diets of these yearlings and juveniles were dominated by mediumfat herring, low-fat pollock, and flatfish; however, they also included eulachon, sandlance, shrimp, and cephalopods. In other words, their diets were quite diverse. Additionally, yearlings and juveniles with more diverse diets were in better condition independent of prey species eaten, provided that some high-quality prey (such as herring or very-high-fat eulachon) were obtained to balance the low-energy prey consumed (Iverson et al., 2003). Information on diets of harbor seals over longer intervals of time, such as decades, is not available for most of the GOA. This is unfortunate as there is no way to evaluate the food limitation hypothesis for them based on possible change in diet. The only information on long-term variability in harbor seal diets comes from PWS, where herring and pollock were overall the most common prey in both the 1970s and 1990s. It has been hypothesized that the change in composition of fish communities in the GOA following the mid-1970s regime shift reduced the availability of nutritionally high-quality prey to the point that harbor seals and sea lions were physiologically compromised to such a degree that their populations collapsed for one reason or another. However, flexible diets and flexible digestive physiology should buffer harbor seals against variability in prey abundance, and, as we have seen, the decline of harbor seals predated the regime shift by several years.
368
Long-Term Ecological Change in the Northern Gulf of Alaska
Thus, the evidence that nutritional limitation of harbor seals caused their populations to collapse in the northern GOA and Bering Sea is weak. This is not to say that there is no evidence that harbor seals responded to a change in prey quantity or quality – later pupping dates in the 1970s demonstrated a sensitivity of seals to environmental conditions. If delayed pupping dates indicated nutritional limitations at certain times of the year that translated into large negative impacts on juvenile foraging success, or were accompanied by changes in individual fitness and population productivity, and if these persisted for several years, a decline in abundance would hardly be surprising. For example, despite the capital-based energy strategy of harbor seals, whereby most energy needed during lactation is acquired in the months prior to parturition and stored as blubber, females still exhibit effects of variable prey abundance, or in their abilities to obtain it. On Sable Is., heavier females give birth to larger pups that grow more rapidly and have higher survival than pups born of lighter females (Bowen et al., 2001). Over time, populations of seals, like other species in the sea, surely wax and wane as environmental conditions vary. But the buffers seals have in their life history characteristics generally keep populations from undergoing radical fluctuations in abundance of the kind witnessed in Alaska, except in exceptional circumstances such as outbreaks of disease, e.g., the morbilovirus outbreak in the northeast Atlantic (see Section 3.5); starvation, as occurs during El Niño; and predation, as occurred on Sable Is. Harbor seals are known to be preyed upon by killer whales, but unfortunately little is known about the distribution and food habits of killer whales at many times of the year or how this may have influenced the collapse of harbor seals in the northern GOA, Aleutian Is., and Bering Sea. Evidence from other regions has demonstrated that killer whales can significantly affect the distribution and behavior of harbor seals and that in one case in British Columbia, transient killer whale pods foraged for much of their time near harbor seal haulouts and were almost 100% successful in 138 capture attempts (Baird and Dill, 1995). Of all transient killer whale attacks observed since the mid-1970s in British Columbia and Washington, about 62% were on harbor seals, and harbor seals are known to be one of the most common pinnipeds taken by killer whales around the world (Weller, 2002). Clearly, killer whale predation could have resulted in substantial mortality in some areas or populations in the North Pacific.
4.9.6. Steller Sea Lion Decline Sea lions have been the subject of disproportionately more research in the past three decades than harbor seals. Studies in the field and laboratory have been especially intense in the years since the Western Stock was listed as endangered. Concerns that sea lions might be in conflict with commercial fisheries led to restrictions on fishing
Long-Term Change 369
activities around rookeries and haulouts, economic hardships for commercial fisheries, and huge Federal appropriations for sea lion – fisheries research. Despite the attention and major advances in understanding of sea lion biology and ecology that have come from all of those studies, little more is known about them in terms of population change than is known about harbor seals. As with harbor seals, sea lions were subjected to control programs in the 1950s and 1960s, and there was a commercial harvest of pups. These efforts apparently caused less change in sea lion abundance than they did for harbor seals. Still, it is possible that direct killing and incidental killing in commercial fishing gear contributed to small localized and possibly broader declines by the 1970s (Paine et al., 2003). However, the beginning of the collapse of the Western Stock in the northern and western GOA apparently did not begin until about 1980, several years after the beginning of the harbor seal collapse. The notion that a lack of food caused the collapse was based originally on three reports from field studies suggesting that (1) sea lion diets changed from prey with high fat content to prey with low fat content after the 1970s, the so called “junk food hypothesis”; (2) sea lions grew more slowly in the 1980s than in the 1970s; and (3) fewer female sea lions carried their pregnancies full term in the 1980s (Alverson, 1992; Merrick and Calkins, 1996; Calkins et al., 1998; Pitcher et al., 1998). All of these changes, as well as the decline in general, are considered by many people to have resulted from the regime shift of the mid-1970s and the effect it had on the structure of the fish community in the GOA (Francis et al., 1996; Trites and Donnelly, 2003). Laboratory studies of captive sea lions in the 1990s confirmed that they are sensitive to diet quantity and quality (see Trites and Donnelly, 2003); however, these studies tested individuals fed on only a single low-fat prey species, a situation likely not faced by sea lions in the wild with a diverse assemblage of prey. Several thorough reviews of the evidence of food limitation and other factors (DeMaster and Atkinson, 2002; Paine et al., 2003; Trites and Donnelly, 2003) failed to discover a sound explanation for the abrupt and rapid decline, and a recent review of the junk food hypothesis as applied to the case of sea lions rejected the notion (Fritz and Hinckley, 2005). Full accounts of those reviews will not be repeated here – we will, however, examine several main issues in the context of sea lion life histories and strategies for survival and the likelihood that nutritional stress could have caused the precipitous collapse of the Western Stock of Steller sea lions. The fundamental underpinning of the junk food hypothesis is diet. Diets of sea lions in the GOA have been estimated in several periods since the 1940s from analyses of stomach contents and scats. A small sample of sea lions was collected from a broad region of the GOA in the summers of 1945 and 1946 (Imler and Sarber, 1947); large samples were collected from southeast Alaska and the western GOA over four summers from 1958 to 1962 (Mathisen et al., 1962; Thorsteinson and Lensink, 1962; Fiscus and Bain, 1966); and large samples were collected from a broad region of the northern GOA in all four seasons in the mid-1970s and in spring–fall in the mid-1980s (Pitcher, 1981;
370
Long-Term Ecological Change in the Northern Gulf of Alaska
Calkins and Goodwin, 1988). In all of these studies, diets were determined by examining stomach contents. Lethally collecting animals for biological studies was discontinued in the 1990s, and instead of using stomach contents, diets were estimated by identifying prey remains in feces or scats (Merrick et al., 1997; Sinclair and Zeppelin, 2002). This method is superior to the analysis of stomach contents in that very large numbers of samples can be obtained, but otherwise is greatly inferior since the digestive tract of sea lions is highly efficient, owing in large part to its unusual length (as described in a preceding section), and few prey remains are passed in scats. Recognizing that there are biases in estimates from these methods, it nevertheless appears that diets of the sea lions collected in the mid-1940s in southeast Alaska and the northern GOA (Chiswell Is., Barren Is., and Kodiak Is.) were dominated by pollock, with important amounts of flatfishes, particularly arrowtooth flounder, and lesser amounts of salmon, Pacific cod, and other species. Sea lions collected from southeast Alaska to Unimak Pass in the late 1950s and early 1960s had consumed few pollock or flatfishes, but instead had been feeding primarily on capelin, sandlance, and rockfish, and on lesser amounts of sculpins, octopus, and various other species. Pollock was once again the predominant prey overall by the mid-1970s, a position it continued to hold in the mid-1980s and throughout the 1990s. By the mid-1980s and through the 1990s, flatfishes, again particularly arrowtooth flounder, also were common prey. Thus, over this span of more than half a century in the GOA, diets in the late 1950s to early 1960s stand out as anomalous. This suite of prey overlaps broadly with prey eaten by harbor seals in the GOA. Another characteristic sea lions share with harbor seals is dietary flexibility – within years, sea lion diets vary considerably between regions and seasons. For example, in common with harbor seals, sea lions appear to consume many more salmon and capelin in summer than winter, and they consume more herring and squids in PWS than elsewhere (Pitcher, 1981). And, as with harbor seals, this flexibility complicates efforts to ascribe importance to particular prey. The conspicuous difference between diets in 1958 to 1962 and later years, as well as a small increase in the proportions of pollock and flatfishes in diets in the 1980s and 1990s compared to the 1970s, have been proposed as evidence that a change in prey quantity or quality was the cause of the sea lion decline. The argument hinges on the fact that forage species such as capelin have a higher fat content than pollock and flatfish, and thus their reduced abundance in diets led to lowered fitness, reduced productivity, and population decline (Merrick et al., 1997; Calkins et al., 1998; Pitcher et al., 2000; Trites and Donnelly, 2003). The principal evidence for food limitation and reduced individual fitness is that sea lions appeared to grow more slowly in the late 1970s and early 1980s than they did prior to the 1970s regime shift, and the proportion of adult females that carried pregnancies to full term declined in the 1980s. While it does appear that growth was slower in the 1980s than the 1970s, growth in the 1970s was actually slower still than in the late 1950s. Calkins et al. (1998)
Long-Term Change 371
measured age 9+ female sea lions and found that animals were significantly larger (standard length and axillary girth) in the 1970s compared to the 1980s. Since body size is influenced most during the first 8 years of life, Calkins et al. (1998) backdated 8 years from the mid-1980s to determine the break point for the reduction in size – the late 1970s – and concluded that these data supported the nutritional stress hypothesis. This conclusion has been generally interpreted to support the hypothesis that the sea lion decline resulted from the mid-1970s regime shift in the North Pacific and the detrimental effect it had on critical forage species, despite the caution by the authors that evidence of a direct link between suboptimal nutrition and the decline was lacking. Calkins et al. (1998) further reviewed similar growth rate data obtained from a sample of sea lions collected in 1958 from within the range of the 1970s and 1980s collections (Fiscus, 1961; Mathisen et al., 1962). Age 9+ female sea lions (age range 9–22 years) in 1958 were significantly larger than in the 1970s, and the authors raised the possibility that growth rates had been in decline since well before the 1970s (Fig. 4.53). Backdating 9 to 22 years from 1958, to see when growth was important in setting the size of the older females collected then, yields 8-year intervals of 1936–1944 as the critical years for the oldest females and 1949–1957 for the youngest. Thus, female sea lions collected in 1958 grew to large sizes during a period when diets, for at least a portion of the interval, apparently consisted primarily of gadids and flounders. Applying the same procedure to the size data from the mid1970s yields 8-year growth intervals of approximately 1959–1967 for the oldest (16 years) and 1968–1976 for the youngest. Thus, the oldest animals from the 1970s
2500 standard length
CENTIMETERS
2300 2100 1900 1700 1500
auxillary girth
1983 1985
1981
1979
1975 1977
1973
1969 1971
1967
1965
1963
1961
1959
1957
1955
1300
YEAR
Figure 4.53: Historic Calkins et al. (1998).
changes
in
sea
lion
growth
rates.
Data
from
372
Long-Term Ecological Change in the Northern Gulf of Alaska
underwent their 8 critical growth years during a period of what is thought by some to have been prey nirvana, yet they were conspicuously smaller than animals from the preceding “junk food” era of the 1940s. Two conclusions emerge from this analysis. The first one is that since growth rate likely is an indication of nutritional state, then sea lions in the GOA experienced suboptimal nutrition during a large portion of the twentieth century. The second conclusion is that the significance of the 1977 regime shift on growth rates of sea lions in the GOA is lost. The other evidence from field studies that sea lions in the Western Stock were nutritionally compromised is the decline in the ratio of females that conceived to those that gave birth in the 1980s compared to the 1970s (Pitcher et al., 1998). Although this too is compelling and a likely indicator of suboptimal nutrition, the difference in productivity that would result from the change is far too small to account for the collapse of the population that occurred over the short interval of about a decade. Furthermore, as the complete record of declining growth rates tells us, there is no reason to suppose that declining productivity did not begin prior to the 1970s as well. Diets of sea lions throughout the northern GOA and Aleutian Is. in the 1990s were much the same as in the 1980s (Sinclair and Zeppelin, 2002). They were dominated by pollock and flatfishes in the northern GOA and by Atka mackerel in the Aleutians. In contrast, sea lions in southeast Alaska enjoy a greater diversity of prey and higherquality diet than animals in the Western Stock (Winship and Trites, 2003), and this has been offered as a reason for the opposite population trends. Nevertheless, the physical condition in the past decade of adult females and pups in the Western Stock has been excellent and equal to or superior to that of females and pups in the Eastern Stock in southeast Alaska (Merrick et al., 1995; Davis et al., 1996, 2002; Adams, 2000; Rea et al., 1998, 2003). Likewise, Western Stock maternal attendance patterns and foraging trip durations in summer and attendance patterns at haul outs in winter are similar to those in southeast Alaska (Brandon, 2000; Milette and Trites, 2003). Therefore, it appears that despite an expensive, income-based energey budget, the suite of sea lion survival strategies has effectively buffered animals of the Western Stock from theoretically adverse effects on individual fitness of existing in an environment replete with “junk food.” Given their large body size and proportionately even larger digestive tract, it is probable that, as with other large mammals such as bears, sea lions can compensate for a lack of food quality by consuming poorer-quality foods in abundant quantity e.g., Schaller et al., 1985; Stirling and Derocher, 1990; Welch et al., 1997). Despite demonstrations that captive sea lions can be put into negative energy balance when fed single-species diets of low nutritional quality and will lose weight steadily until switched to high-quality diets (Rosen and Trites, 2000), there is no evidence of such radical effects on sea lions in the wild in Alaska, nor of any individuals consuming diets comprising single prey species. The small declines in growth rates from the 1950s to the 1970s to the 1980s and in full-term pregnancy rates from the1970s to the 1980s were subtle by comparison, and are of the kind that might help
Long-Term Change 373
maintain a population in dynamic equilibrium with prey resources such that abundance fluctuates somewhat about a long-term mean. The nature of sea lion life history characteristics, as described above, supports this view. None of this supports the idea that nutritional limitation precipitated the collapse of sea lions in the 1980s any more than evidence of a subtle response of harbor seals to suboptimal prey availability supports nutritional limitation as the cause of their population collapse. Yet, something did cause the collapse of the Western Stock of sea lions, just as something caused the collapse of harbor seals several years earlier. The conclusion of the National Research Council was that at least since the 1990s the decline bears characteristics of top-down forcing, or predation (Paine et al., 2003). And, if predation is the cause of declining numbers in recent years, why should it not be considered a factor in the collapse? Indeed, sea lions in the GOA are preyed upon in substantial numbers by killer whales (Heise et al., 2003). In one famous case, flipper tags (identification markers applied to the flippers of sea lion pups) from 14 sea lions were found in the stomach of one killer whale that beached in PWS in 1992. Those animals had been marked at Marmot Is. (near Kodiak Is.) 2–5 years earlier. Killer whales are known to prey on sea lions throughout the GOA as, for example, at the Chiswell Islands, where one was recorded on remote video link circling a haul out (Fig. 4.54). Plausibility analyses using energetics models of killer whales, demographic models of sea lions, and estimates of the abundance of transient killer whales in the GOA indicate that killer whales would have been capable of causing the collapse of sea lions in the 1980s
Figure 4.54: A transient killer whale stalking a group of Steller sea lions in the Chiswell Islands. Photo taken from video monitor (photograph courtesy of J. Maniscalo, Alaska SeaLife Center).
374
Long-Term Ecological Change in the Northern Gulf of Alaska
(Springer et al., 2003; Williams et al., 2004). The lack of evidence of severe food shortage during the period of rapid decline in the 1980s, and the ability of sea lions in recent times to easily cope with prey considered by some to be nutritionally poor further argues that predation could have been a major factor in sea lion population dynamics.
4.9.7. Sea Otter Population Changes Our view of coastal marine communities and the role of sea otters in those communities has been shaped to a large degree by what we have observed during the twentieth century, as sea otter numbers grew from a few hundred animals distributed over thousands of miles of shoreline to many tens of thousands of individuals occupying more than half of their historic range. Suitable, unoccupied habitat remains primarily in southeast Alaska, and it would appear likely that in the absence of increased mortality sea otter populations there will continue to recover. Annual adult female sea otter reproductive rates are nearly invariant (85–90%) among those populations studied, with most females producing a single offspring annually (Monson et al., 2000). In contrast to generally high and constant reproductive output, survival appears to be the life history character resulting in variation in population size (Estes and Bodkin, 2002). Young sea otters are subjected to two periods of elevated mortality during their first year of life. The first immediately follows birth and is likely highest among pups of first-time mothers. The second follows weaning at about 6 months of age. Survival of young sea otters is likely influenced by the age and experience of the mother and the availability of food resources (Bodkin, 2003). Where food and space are abundant (e.g., recently recolonized areas), females will be in better condition (greater mass per unit length) and their offspring will benefit through increased survival. Conversely, in long-established populations, where food may be a limiting resource, females will weigh less per unit length on average and their offspring may suffer elevated mortality. If a juvenile survives through its second year of life, survival probability remains high until about the age of 10 (Monson et al., 2000). As with harbor seals and sea lions, sea otter population sizes usually appear to be regulated primarily by mortality (principally starvation) of young during the first year of life (Estes and Bodkin, 2002). In California and the Aleutian Islands, while at or near equilibrium densities, only about half of the pups reach weaning age, and fewer than half of those survive to their first birthday. At Kodiak Is. during early recolonization, when females were in relatively good condition, post-weaning survival was nearly 90% (Monson et al., 2000), demonstrating the importance of food resources in population regulation. We know little about how sea otter prey populations respond to variability in primary production in the Gulf of Alaska. Diets of sea otters in the soft sediment habitats of the GOA (PWS and Kodiak Island) consist predominantly of clams (>70%) (Dean et al., 2002), although in PWS dungeness crabs were important prey until
Long-Term Change 375
depleted by otters (Garshelis et al., 1986). On Kodiak Is., mussels become more abundant in otter diets in areas where clams have been depleted by otters (Doroff and DeGange, 1994). Among rocky reef habitats throughout their range, urchins, crabs, snails, and mussels are common prey (Estes and Bodkin, 2002). However, the nature of community structure, survival strategies of invertebrate prey species, and the relatively small magnitude of interannual variability in macrophyte production in the benthic food web tend to dampen fluctuations in productivity compared to pelagic food webs. Over longer intervals, increases in prey available to sea otters should eventually have positive effects on population sizes. Conversely, as prey becomes less available, there is evidence that individuals respond through declining mean weights and reduced survival (Bodkin et al., 2000). As local population densities reach equilibrium with food and space resources, large, abrupt declines in those populations might occur under stressful conditions, such as during a harsh winter. For example, recently at Bering Is. (Commander Is., Russia), sea otters experienced a major winter die-off and the population declined from about 5000 animals to 3000 in a single year (Bodkin et al., 2000). Nearly 80% of the recovered carcasses were male, leaving the surviving population with a higher proportion of females. Similar phenomena might be expected as recovery continues in regions where sea otters have not recently declined, particularly on individual islands where emigration might be limited, or in local mainland areas where the distance to unoccupied habitat exceeds the dispersal distance of sea otters. For example, in southeast Alaska, as unoccupied habitat shrinks and growing populations come into equilibrium with their food resources, it seems unlikely that we will continue to see the high growth rates we observed in the twentieth century. However, the catastrophic, widespread declines of sea otters of the magnitude witnessed in the Aleutian Is. and Alaska Peninsula in the past decade were unexpected and appear to have been unrelated to food resources, traditionally considered to be a primary factor regulating sea otter populations. Nowhere were there observations of unusual numbers of stranded carcasses as there were during the recent decline in the Commander Is. Nor is there evidence that direct or indirect human removals in modern times affected sea otter populations. There is strong evidence, at least for the Aleutian Is., that predation by killer whales was the principal cause of the collapse (Estes et al., 1998), which began about 10 years after the start of the sea lion collapse. Similar information is not available from the Alaska Peninsula, but predation is certainly a contributing factor to sea otter mortality throughout the species range.
4.9.8. Conclusions Harbor seal, sea lion, and sea otter populations in Alaska have all been reduced at times by irresponsible killing by people. In the case of sea otters, they were nearly
376
Long-Term Ecological Change in the Northern Gulf of Alaska
exterminated 200 years ago. More recently, in the 1950s and 1960s, bounty programs, commercial hunting, and other control efforts had local and regional effects on the abundance of harbor seals and sea lions, with the greatest depletions in southeast Alaska and British Columbia but with much less effect in the northern GOA and Aleutian Is. By the early 1970s, the abundance of all three species was relatively high in the northern and western GOA and low in the eastern and southern GOA. Since the 1970s, we have witnessed a series of unexplained population collapses throughout the region of former abundance. The first species to go from much of its range were harbor seals, beginning in the early 1970s. Their collapse was followed by that of sea lions, and in another 10 years, of sea otters. The leading, long-standing explanation for the collapses of harbor seals and sea lions has been nutritional limitation. It was based on observations that pupping dates of harbor seals were delayed and that sea lions grew more slowly and females achieved fewer successful pregnancies during the periods of decline in the 1970s and 1980s. Furthermore, it was thought the regime shift of the mid-1970s was to blame for these changes because of the negative effect it had on the structure of forage fish communities in the GOA. But, the problem with the nutritional limitation hypothesis is that the apparent changes in individual fitness were not of a magnitude sufficient to cause starvation, and there is no other evidence that starvation did occur. The more subtle effects on fitness that were observed in adults were not sufficient to cause a population collapse otherwise. Still, we must remember that there is much we do not know about harbor seals or sea lions in winter and during the first 1–2 years of independence when most juvenile mortality occurs, likely at sea. The problem with the regime shift hypothesis for these collapses in the northern North Pacific marine mammal populations is that the unexplained portion of the harbor seal collapse on Tugidak Is., as well as the beginning of the unexplained portion of the recent decline of fur seals in the Bering Sea, began well in advance of the regime shift. Had food limitation caused a reduction in juvenile survival, and thus recruitment, that precipitated the collapse of the whole population, it would have had to begin many years prior to the onset of the collapse, such that the adult population would have had time for senescence and begin dying en masse by the early 1970s. There is no evidence for this scenario either. A reduction in juvenile recruitment certainly could have occurred, and while it is a significant regulator of populations, there is no evidence that it could have been so severe as to cause such a precipitous population collapse. Furthermore, the collapse of sea lions in the eastern GOA and of sea otters in the Aleutian Is. and western GOA occurred in the 1990s, long after the mid-1970s regime shift. In neither case is there evidence that food limited those populations. The mid-1970s regime shift has also been invoked to explain the increases in abundance of harbor seals and sea lions in southeast Alaska and British Columbia for a related reason – the positive effect it had in that region on forage fish species and communities. This notion is based on observations of inverse production regimes in the
Long-Term Change 377
broad region of the northern GOA and Bering Sea compared to the region from southeast Alaska to the Pacific Northwest. That is, when conditions are good in the north and west, they are poor in the east and south, and visa versa, because of the ways ocean physics in these regions are, or may be, differentially influenced by variable meteorological forcing over the North Pacific (Wickett, 1966; Chelton and Davis, 1982; Hollowed and Wooster, 1992; Gargett, 1997; Francis et al., 1998; Hare et al., 1999). While evidence for broadscale, east–west physical asymmetry in the GOA is compelling, as is evidence that it affects biology at lower and intermediate trophic levels, such a complex regime shift–phase shift mechanism is hardly needed to explain the increasing pinniped populations in southeast Alaska and British Columbia. A more likely explanation is simply that protections enacted in Canada and the US in 1970 and 1972, respectively, against the irresponsible commercial and bounty programs of the 1950s and 1960s reversed the fortunes of pinnipeds and set the severely depleted populations on courses of recovery. The only question is why recovery has been so slow? Could it be poor nutrition and/or predation? The ability of pinniped life history strategies to successfully buffer populations against apparent effects of food limitation was recently demonstrated for elephant seals, the largest phocid seal, and relative of harbor seals, in the North Pacific. Elephant seals breed on islands off the coast of California, but feed in the central and northern GOA. Beginning in the mid-1970s (or earlier – the record does not begin until 1975), weaning weights of pups declined in association with increased foraging trip durations of adult females and a decline in the mass females gained during their post-breeding foraging trip, both indications of reduced foraging success, i.e., reduced prey availability (Le Boeuf and Crocker, 2005). All trajectories changed sign in the late 1990s, at about the time of the 1998 regime shift, indicating improved prey availability. Yet, despite the negative effects on individuals of apparently suboptimal nutrition from the mid-1970s through the mid-1990s, the population nevertheless grew substantially during that era, increasing at the Año Nuevo colony from 605 pups born in 1975 to 2500 in 2004. None of this is to say that regime shifts and climate change that alter the nature of physical forcing are not important in structuring marine ecosystems. They clearly are, particularly at lower trophic levels, and this provides a mechanism for bottom-up regulation of populations at higher trophic levels. Nor is it to say that marine mammals are immune to the possible effects of fluctuations in abundance, distribution, or quality of prey, and the role of nutritional limitation in the North Pacific in the past half century. On the contrary, they are sensitive as indicated by changing pupping dates of harbor seals and changing growth rates and rates of fetal abortion in sea lions. But, radical effects of environmental change on population dynamics of species at lower trophic levels is attenuated at higher and higher levels in the food chain because of the buffers provided by strategies for survival of progressively longer-lived, more k-selected species. The comparatively mild individual responses to suboptimal
378
Long-Term Ecological Change in the Northern Gulf of Alaska
nutrition of pinnipeds are of the kind that would reasonably be expected to adjust long-term population dynamics and maintain abundances in equilibrium with prey, not of the magnitude to explain population collapses. Nutritional limitation was never offered as an explanation for the sea otter collapse, which has been blamed instead on predation by killer whales. Population dynamics of harbor seals, sea lions, and sea otters are believed to be driven generally by recruitment success, i.e., by juvenile mortality. However, there are few empirical data on juvenile survival for most species. In turn, juvenile mortality may be due to poor nutrition and starvation, or disease and predation, both of which could be exacerbated by poor nutrition. However, population dynamics of all three species during their respective crashes also had to necessarily include high adult mortality, for which there was no evidence of starvation or disease. Therefore, a more likely explanation is that, while there could have been a suite of factors affecting individuals, the strategies for survival that these species have evolved to cope with uncertain environments, different as they are, were successful in buffering populations against the most common kind of variability, that of prey abundance, but may have been inadequate to save them from effects of a highly unusual change in predation pressure. Similarities and patterns in the magnitude (~89–90%), timing (7–10 years between events), and spatial extent (western GOA, Aleutian Is., southeast Bering Sea) of harbor seal, sea lion, and sea otter declines, in addition to weak support for nutritional limitation as causing their collapse, further imply that top-down forcing effects, not previously considered as important in the dynamics of marine mammal populations, must be considered. Whichever is the case, nutritional limitation or predation, these are unprecedented events in the recorded history of pinniped populations of the North Pacific.
4.10. Crabs and Shrimps Gordon H. Kruse 4.10.1. Introduction Some of the species with the most widely fluctuating populations in the Gulf of Alaska are invertebrates. Invertebrates are an extremely large and diverse group of organisms that include copepods, clams, sea urchins, squid, abalone, and many others. As with many groundfishes and unlike many marine birds and mammals, invertebrates typically offer little to no parental care for their young. However, some (e.g., crabs) carry a clutch of developing embryos for nearly one year until hatching, when larvae become free swimming. It is generally believed that invertebrate population fluctuations are largely driven by factors affecting larval and juvenile survival, including advection by ocean currents, food availability, and predation.
Long-Term Change 379
In this chapter, long-term changes in populations of red king crab (Paralithodes camtschaticus), Tanner crab (Chionoecetes bairdi), and northern (sometimes locally called “pink”) shrimp (Pandalus borealis), are examined, as well as some potential responsible mechanisms. Fisheries take place for other crab and shrimp species in the Gulf of Alaska, but changes are most well documented for these three species. These three provide the best case studies of invertebrate population dynamics in the Gulf of Alaska, because of a relatively long history of stock assessment surveys by state and federal fishery management agencies owing to the high commercial value of these species. In contrast to many of the groundfishes, most crab and shrimp stocks in the Gulf of Alaska crashed to low levels of abundance in the late 1970s and early 1980s. Even after 25 years, the causes of these spectacular collapses remain a matter of much debate. Because of their divergent life history strategies and population trajectories, crabs and shrimps provide revealing contrasts with other taxa discussed so far. In this chapter, we will examine several of the leading theories for these declines.
4.10.2. Long-term Dynamics of Shrimp Stocks The trawl fishery for northern shrimp in the central and western Gulf of Alaska and eastern Bering Sea is an excellent example of a boom-bust fishery. Small recorded landings were first recorded in the late 1950s, but the fishery did not take off until the late 1960s (Fig. 4.55). Peak landings of approximately 68,000 mt were taken in 1973 and a second peak of 59,000 mt in 1977. Thereafter, the fishery collapsed to just 15,000 mt by 1980 and was closed in most areas by 1983. Commercial harvest is not always indicative of population size, so it is difficult to infer trends in shrimp abundance from catch records alone. Whereas landings declined during the late 1970s to early 1980s because of declining abundance, landings are unlikely to be indicative of changes in stock abundance during the development phase of the fishery in the 1950s and 1960s. Thus, fisheryindependent surveys are preferred methods for estimating patterns in shrimp abundance. Small-mesh trawl surveys have been conducted in the central and western Gulf of Alaska by ADF&G and NMFS since 1953. Primary survey locations included Prince William Sound, lower Cook Inlet, Kodiak Island, and the south side of the Alaska Peninsula, including Pavlof Bay. These data were analyzed for trends in northern shrimp survey catches for Kachemak Bay in lower Cook Inlet (Bechtol, 1997) and all areas combined (Anderson et al., 1997; Anderson and Piatt, 1999; Anderson, 2000). Unfortunately, survey gear and methods were not standardized during the early years, so catch rates, measured as catch-per-unit-effort (CPUE), among locations and agencies are comparable only since 1973. Analyses of aggregated survey data indicate that the proportion of shrimp in the total (all taxa) survey catches declined slightly during 1972–1976, and then declined exponentially through the late 1980s (Fig. 4.56). Examination of aggregate CPUE
380
Long-Term Ecological Change in the Northern Gulf of Alaska
Figure 4.55: Historical yearly shrimp landings in from Prince William Sound, Cook Inlet, Kodiak, Alaska Peninsula, Chignik, Aleutian Islands, and eastern Bering Sea (from Kruse et al., 2000).
data since 1973 indicates that relative abundance of northern shrimp remained high in 1974–1975, and the sharp decline began in 1976. The analysis was conducted by aggregating CPUE data across all survey areas, and the data were further smoothed using 3-year running averages (Fig. 4.56). Obviously, this approach blurs area-specific differences. A length-based population analysis of northern shrimp in Kachemak Bay alone reveals a more complicated trajectory (Fu et al., 1999). This stock increased from a moderate level in 1971 to high levels during 1973–1975, declined to moderate levels during 1976–1977, returned to high levels during 1978–1979, and declined exponentially thereafter (Fu et al., 1999). This contrasts with trends in commercial catch in Cook Inlet; harvests peaked in 1980–1981, and the fishery did not fully collapse until 1983–1984 (Bechtol, 1997).
4.10.3. Long-term Dynamics of Crab Stocks King crab fisheries in the Gulf of Alaska were developed by Japanese fishermen in the late 1930s. Small domestic catches were taken since the late 1930s, but landings were
Long-Term Change 381
Figure 4.56: Proportion of shrimp and nonshrimp species in small-mesh survey catches in the northern and western Gulf of Alaska during 1953 to 1997 (from Anderson, 2000). Proportions represent 3-year running averages of aggregates across all survey areas.
not recorded from Cook Inlet and Kodiak until the 1950s (Fig. 4.57). In lower Cook Inlet (Kachemak and Kamishak Bays), landings peaked at nearly 4000 mt in 1962–1963 and then declined in the mid to late 1960s owing to damage of fish processing facilities from the 1964 earthquake. Landings improved somewhat in the 1970s, but then stock abundance declined in the early 1980s, and this fishery has been closed since 1983. Landings in Kodiak peaked at nearly 43,000 mt in 1965–1966, stabilized at 5000–11,000 mt in the late 1960s until the early 1980s when landings declined sharply, and the fishery has been closed since 1983. Most Tanner crab fisheries in the Gulf of Alaska began in the late 1960s or early 1970s, largely as a result of increased interest triggered by the decline of fisheries for red king crabs. The Tanner crab fishery in Prince William Sound peaked at 6300 mt in 1972–1973, associated with improved prices following king crab declines. Landings in the sound declined to just 215 mt in 1988 before it was closed (Fig. 4.58). The domestic fishery for Tanner crabs in Kodiak began in 1967, peaked at 15,000 mt in 1977–1978, and declined until closed in 1993–1994. In the northern and western Gulf of Alaska, assessment surveys have been conducted on a number of stocks of red king and Tanner crabs. The longest time series
382
Long-Term Ecological Change in the Northern Gulf of Alaska
Figure 4.57: Historical landings of red king crabs in Cook Inlet, Kodiak, Aleutian Islands, and Bristol Bay (from Kruse et al., 2000). Catches for all areas except Bristol Bay are listed by fishing season (e.g., 1970 refers to the 1970–1971 winter fishing season).
Figure 4.58: Historical Tanner crab landings in Prince Williams Sound, Cook Inlet, Kodiak, Chignik, south Alaska Peninsula, and eastern Bering Sea (from Kruse et al., 2000). Years represent either calendar year or fishing season (e.g., 1970 means 1970–1971 winter season), depending on area and year.
Long-Term Change 383
covers Kodiak red king crab, which is also the best-studied crab stock in the gulf. Pot surveys were conducted during 1973–1986, and trawl surveys have been conducted annually since 1986. Collie and Kruse (1998) estimated historical abundance and recruitment of Kodiak red king crab using catch-survey analysis, a population estimation model that links survey and catch data to estimate catchability of crabs by pots. Estimated abundance of legal males was high (7–11 million crabs) during 1973–1975, declined to moderate levels (~4 million) by 1978, increased slightly (5–6 million) during 1979–1981, and fell to very low levels (<1 million) thereafter. The number of recruits (those crabs that just molted to legal size) were moderate in 1973, peaked in 1974, declined to moderate levels in 1978, increased slightly during 1979–1981, and declined thereafter. Zheng and Kruse (2000) examined recruitment patterns (trends in the size of cohorts or year-classes) of 15 Alaskan crab stocks from southeast Alaska to the Bering Sea, including 8 crab stocks in the northern and western Gulf of Alaska – red king crabs from Kamishak Bay, Kachemak Bay, Kodiak, and south Alaska Peninsula, and Tanner crabs from the same locations. Recruitment was lagged to year of egg hatch (brood year). After an increase in the late 1960s, year-class strength declined sharply during the early 1970s for all red king crab stocks in the northern Gulf of Alaska (Fig. 4.59). Outside the northern Gulf of Alaska, trends diverge from this “gulf” pattern. Although the pattern for Bristol Bay red king crab is similar, patterns for red king crabs in southeast Alaska and Norton Sound differ markedly. Recruitment patterns for Tanner crabs are more intricate (Fig. 4.59). Similar to red king crabs, Tanner crab recruitment declined from the late 1960s to early 1970s and has been generally low since the 1980s. Unlike red king crabs, Tanner crab recruitment was mixed in the 1970s, depending on the stock. The pattern for Kamishak Bay was most divergent, with some relatively strong year-classes in the late 1970s and 1980s. Although recent surveys suggest some slight improvement in Tanner crab recruitment in some areas, only the Kodiak Tanner crabs have improved sufficiently to warrant a small fishery during 2001–2003 of less than 375 mt annually. Otherwise, fisheries for red king and Tanner crabs are closed in the central and western Gulf of Alaska because of low abundance. Bering Sea crab stocks (e.g., Tanner, snow, and blue king crabs) had very divergent historical trends from these patterns in the Gulf of Alaska. Similarly, in their analysis of 100 climate and biological data sets, Hare and Mantua (2000) also noted some differences in species patterns among the Gulf of Alaska and eastern Bering Sea.
4.10.4. Climate Forcing Is Climate Important to Invertebrate Populations? The climate regime shift in the late 1970s has been proposed to have caused a community reorganization that included a decline of northern shrimp and some forage fishes
384
Long-Term Ecological Change in the Northern Gulf of Alaska
Figure 4.59: Time series of normalized (log-transformed) recruitment indices for eight crab stocks from Zheng and Kruse (2000). Red king crab and Tanner crab stocks and their abbreviations are South Peninsula (SPRK, SPTC), Kodiak (KODRC, KODTC), Cook Inlet – Kamishak District (CIKRK, CIKTC), and Cook Inlet – Southern District (CISRK, CISTC).
in the Gulf of Alaska (Anderson and Piatt, 1999). Pacific salmon experienced a general increase in abundance subsequent to the regime shift; one hypothesis for this increase is an intensification of the Aleutian low-pressure system in winter, which promoted increased phytoplankton and zooplankton production and standing stocks that led to enhanced feeding conditions and survival of outmigrating salmon smolts. If this mechanism was responsible for the increase in salmon, at least at face value it does not seem to explain declines in other taxa, such as forage fish species, shrimps, and crabs. Most dynamics of exploitable crab and shrimp stocks appear to be explained by variability in recruitment, although parental stock size has some effects on recruitment in some stocks, at least at low stock size (i.e., strong year-classes rarely result from small parental stocks). Recruitment is periodic, not regular, and it seems that strong year-classes likely result when a number of factors affecting survival all line up favorably, an apparently rare event. Crab and shrimp stocks wax and wane as these occasional large year-classes grow in body size and recruit to the adult population and then age as natural and fishing mortality take their tolls on the survivors.
Long-Term Change 385
Koslow (1984) suggested that positive correlations in recruitment among stocks within species or among closely related species over large geographic areas are indicative of a role of large-scale physical forcing. Likewise, Zheng and Kruse (2000) interpreted the “gulf pattern” – the relatively strong correlation of recruitment trends among king and some Tanner crab stocks in the northern and western Gulf of Alaska – as indicative of a relatively strong role of climate-forced effects on crab stocks as opposed to other factors. For stocks with the gulf pattern, the declining trend in abundance since the early 1970s was similar to indices of the strength of the Aleutian low, seemingly consistent with regime-shift theory (Francis et al., 1998). However, one irony is that their decline in recruitment actually started in the early 1970s, 3–6 years prior to the regime shift in 1976–1977. In the rest of this chapter, we consider potential mechanisms for climate control of crab and shrimp populations in greater detail, along with other potential explanations for the observed dynamics of these invertebrates.
Shrimp and the Match–Mismatch Hypothesis Because swings in crab and shrimp abundance seem to be driven largely by recruitment, most attempts to unravel the mystery of invertebrate population dynamics involve theories about factors affecting early-life survival. Anderson and Piatt (1999) invoked Cushing’s (1995) match–mismatch hypothesis to attempt to explain why some fishes and invertebrates declined in the face of a general increase in ocean productivity (see Box 4.4). They noted that biomass peaks of Neocalanus copepods are 1–2 months later during cold years than warm years. Larvae of most shrimp and crabs hatch in late spring or early summer, as well as larvae of forage fishes such as capelin. Thus, shifts to earlier plankton blooms, prior to shrimp egg hatch, could result in a mismatch between shrimp larvae and their zooplankton prey. On the contrary, earlier plankton blooms could benefit groundfishes, many of which are winter spawners. However, as cooler temperatures delay the spawning time of northern shrimp (Apollonio et al., 1986) and prolong embryo incubation (Nunes and Nishiyama, 1984), the match–mismatch hypothesis would apply only if temperature differentially shifts the timing of the plankton bloom relative to the shift in larval shrimp hatching. Field studies are required to carefully evaluate potential changes in the match of the spring bloom and shrimp hatch timing. The potential role of the match–mismatch hypothesis is intriguing, but climate can influence shrimp stocks through other mechanisms. For instance, a decline of 2–3°C in bottom temperature was associated with a 40% increase in fecundity of northern shrimp in the Gulf of Maine (Apollonio et al., 1986). Thus, recruitment could increase due to direct effect of temperature on reproductive output. Also, within temperatures of 3–9°C, total hatch and viable hatch was highest at 3°C, and upon yolk exhaustion, these larvae were also larger compared to viable larvae reared at warmer temperatures
386
Long-Term Ecological Change in the Northern Gulf of Alaska
(Nunes and Nishiyama, 1983). So, larger larvae, associated with colder incubation temperatures, are expected to have greater chances of survival. So, temperature can regulate shrimp reproduction and survival by climate forcing, regardless of potential additional effects on a match or mismatch between shrimp larvae and their prey.
BOX 4.4: MATCH–MISMATCH HYPOTHESIS by Gordon H. Kruse David Cushing’s match–mismatch hypothesis suggests that the magnitude of individual year-classes varies, in part, due to year-to-year differences in the timing of fish spawning with respect to timing of the spring bloom of primary production (Cushing, 1995). Cushing (1969) noted that the timing of spawning of many fishes in the North Atlantic Ocean is virtually fixed at the same time each year, varying at most by one week. Noting that the onset of the spring plankton bloom varied by 4–6 weeks in the North Atlantic Ocean, he surmised that the chances of larvae finding food must vary from year to year. Cushing argued that, if climate change is to affect fish recruitment, it would have the greatest impact during the larval stage through variations in the availability of food. The mechanism linking climate to the timing of the spring bloom is often called the Gran effect or the Sverdrup mechanism. During winter, cooling of the surface waters and strong winter winds create a well-mixed ocean in which nutrients at depth are brought to the surface. Despite the high concentrations of nutrients near the surface, a bloom cannot occur because phytoplankton cells are regularly mixed to depths too dark for photosynthesis to occur. However, as spring approaches, the frequency of strong winter winds subsides and more intense solar radiation creates a stratified water column with warm water at the surface and cool water at depth. This stratification causes mixing to be confined to the upper illuminated layers of the ocean, thereby allowing the bloom to occur as phytoplankton cells now experience high concentrations of nutrients and ample sunlight (see also Section 2.3). Climate regulates the timing of this stratification, and hence the timing of the spring bloom, by modulating the frequency of storms, and their associated winds (mixing) and cloud cover (incident sunlight). When the spring bloom is poorly matched with the relatively fixed spawning time of fishes, poor survival will lead to a weak year-class. If the timing is good, then fewer larvae will starve, and the chances for a strong year-class are much improved.
Red King Crab and the Match–Mismatch Hypothesis The match–mismatch hypothesis motivated an intensive study of red king crab recruitment dynamics in Auke Bay, southeast Alaska, in 1985 to 1989 called APPRISE,
Long-Term Change 387
Association of Primary Production and Recruitment in Subarctic Ecosystems (Shirley and Shirley, 1989, 1990). However, incident light, not temperature, drives the timing of the spring bloom in Auke Bay (Ziemann et al., 1990). The timing of the primary bloom is nearly fixed and begins during the first or second week of April, whereas secondary blooms in May and June are triggered by increases in nutrient concentration associated with wind-mixing events. Under controlled laboratory conditions, the incubation period of red king crab eggs is highly dependent on temperature; hatching occurs after 207 days at 12°C and 305 days at 3°C (Shirley et al., 1990). In the APPRISE study, there was no relationship between larval survival and temporal synchrony of crab larvae with the spring bloom, so the match–mismatch hypothesis was rejected in this case (Shirley and Shirley, 1989, 1990).
Reconciling the Match–Mismatch Hypothesis among Shrimp and Crabs On face value, these red king crab findings for Auke Bay seem contrary to the match–mismatch hypothesis for shrimp, but recall that Anderson and Piatt’s (1999) hypothesis concerned the role of temperature on the timing of the bloom of zooplankton, not phytoplankton. In fact, Anderson and Piatt did not mention phytoplankton in their match–mismatch hypothesis. Although the phytoplankton bloom in Auke Bay depends primarily on incident solar radiation and not temperature, zooplankton blooms in Auke Bay are related to temperature. For example, in Auke Bay Pseudocalanus populations peaked in late April in 1986, late May/early June in 1987, and late March or early April in 1988 and 1989 (Coyle and Paul, 1990). Coolest late winter/early spring temperatures occurred in 1986 and warmest in 1988, a trend fairly consistent with the progressively earlier dates of Pseudocalanus population peaks from 1986 to 1989. Noting that Auke Bay is a small, partially closed embayment in a system of fjords in southeast Alaska, it may not be possible to generalize these results to the entire Gulf of Alaska. A comparison of the ecology of crab and shrimp species may shed more light on the veracity of the match–mismatch mechanism. As mentioned in Section 2.5.5, diatoms, especially Thalassiosira spp., appear to be important prey of red king crabs, whereas copepods, particularly nauplii of Pseudocalanus, are common food of Tanner crabs. Tanner crab zoeae are larger and faster swimmers than early red king crab zoeae, so their ability to capture larger, more elusive prey is not very surprising. So, even if a temperature-driven mechanism is discounted for red king crab larvae because temperature does not regulate the timing of the phytoplankton bloom, it remains plausible for crab larvae that eat copepod nauplii, such as Tanner crabs owing to the apparent positive effect of temperature on copepod reproduction and development (Smith and Vidal, 1984). So, where do northern shrimp fit in? Early-stage northern shrimp larvae reportedly prey mainly on diatoms, although some copepods are also eaten (Stickney and Perkins, 1981). Given this, one might expect that it may be
388
Long-Term Ecological Change in the Northern Gulf of Alaska
more important for shrimp larvae to match the phytoplankton bloom (as with red king crab), rather than the zooplankton bloom (as with Tanner crab). So, the match–mismatch hypothesis begs further scrutiny for shrimp.
Other Potential Climate-forcing Mechanisms and Red King Crab Zheng and Kruse (2000) sought to identify a mechanism to explain the similarity between Aleutian low-pressure dynamics and recruitment patterns of red king crabs (and some Tanner crabs) in the Gulf of Alaska (Fig. 4.59). They noted that red king crab larvae must feed within 2–6 days of hatching in order to survive (Paul and Paul, 1980), and growth of the survivors is directly related to concentrations of Thalassiosira (Paul et al., 1990). In the 5-year APPRISE study in Auke Bay, phytoplankton species composition was greatly influenced by short-term weather (Bienfang and Ziemann, 1995). In years when winds were light and the water column was stable, Thalassiosira dominated the spring bloom of phytoplankton. In years of stronger vertical mixing, such as might be expected during intensified Aleutian lows, a more well-mixed water column supported a spring bloom comprising a more diverse phytoplankton community. Zheng and Kruse (2000) set these observations within the regime shift context by suggesting that years of intensified Aleutian lows (e.g., post 1976–1977) may be associated with stronger water-column mixing in late winter/early spring and lower Thalassiosira abundance, thus adversely affecting feeding success and survival of larval red king crab. That is, total primary production may be less relevant to red king crab larval survival than phytoplankton species composition – in particular, the predominance of Thalassiosira diatoms. This differs from the match–mismatch hypothesis in that it is the species composition of the spring bloom, not the timing of the bloom, which may determine larval crab survival. Of course, red king crab larvae may feed largely on Thalassiosira simply because they are generally dominant in the spring bloom, not because they are preferred prey. In any case, this hypothesis seems consistent with the limited observations in Auke Bay, but empirical testing of broader applicability to the northern gulf remains to be carried out.
Climate Forcing and Tanner Crabs An emerging understanding is that Tanner crab recruitment patterns (Fig. 4.59b) and their causative mechanisms are more reflective of local, rather than regional, conditions than previously appreciated. The most detailed study of Tanner crab recruitment was conducted in Bristol Bay, eastern Bering Sea (Rosenkranz, 1998; Rosenkranz et al., 2001). There is evidence that warmer sea surface temperatures during May–June are associated with years of stronger Tanner crab year-classes. Warmer temperatures favor growth of Pseudocalanus populations (Smith and Vidal, 1984), so warmer sea temperatures during spring may favor feeding success of Tanner crab larvae because
Long-Term Change 389
of higher prey densities. Moreover, it appears that two other mechanisms operate: warmer bottom temperatures favor Tanner crab gonadal development and egg incubation, and winds blowing from the northeast along the Alaska Peninsula during spring promote coastal upwelling while advecting larvae offshore to soft bottom sediments used as nursery areas. Regression models with wind and bottom temperature explained about 50% of the recruitment variability of Tanner crabs in Bristol Bay. However, as with any exploratory analysis, these results should be revisited to determine if they stand the test of time. Nonetheless, it seems that multiple factors and local conditions (e.g., wind direction relative to coastline orientation) may foil attempts to find simplistic one-size-fits-all explanations to Tanner crab declines over the broad region of the northern Gulf of Alaska. The greater diversity of recruitment patterns of Tanner crabs compared to red king crabs (Fig. 4.59) is consistent with a more important role of local, rather than regional, factors. However, as discussed in the rest of this section, there may be other explanations for these recruitment patterns.
4.10.5. Biological Controls As described in the section on the Cascade Hypothesis, top-down controls through predation can have major impacts on ecosystem structure. As a corollary hypothesis to their match–mismatch hypothesis, Anderson and Piatt (1999) suggested that predation contributed to the declines of forage species subsequent to the late 1970s regime shift. They suggested that a negative correlation between groundfish predators e.g., Pacific cod, arrowtooth flounder, and halibut) and forage species (e.g., shrimp, capelin) supports the notion that the Gulf of Alaska is moderated by predation. Is there other evidence for the role of groundfish predation on declines of shrimp abundance in the northern Gulf of Alaska? There is some evidence from surveys and commercial fishery data that there was a change in the abundance and distribution of Pacific cod in the 1970s. Cod biomass increased steadily from the late 1970s to the late 1980s and since then has steadily declined to late 1970s levels (Thompson et al., 2003). Strong recruitment in the late 1970s was at least partly responsible for the increase in biomass. Aside from changes in cod abundance, changes in distribution may have occurred. During the early 1970s, cod were infrequently observed in inshore bays occupied by northern shrimp. By the late 1970s, they were frequently encountered in these nearshore areas and remained common in nearshore waters even though cod and shrimp biomass have both declined. Anderson et al. (1997) suggested that warmer temperatures allowed cod to remain in inshore bays throughout winter, instead of migrating offshore after temperatures became cooler nearshore. At GAK1, temperature anomalies up to 4°C below the long-term average were common prior to 1977; since 1978, anomalies more than 1°C than average have been rare. Pacific cod are winter spawners, and both cod spawning activity and egg hatch success are
390
Long-Term Ecological Change in the Northern Gulf of Alaska
associated with temperatures of 2.5–8.5°C (Alderdice and Forrester, 1971). So, if nearshore temperatures dropped below this range, cod would be expected to move to warmer waters for spawning. Studies elsewhere seem to support this proposition. For instance, the degree of overlap between Pacific cod and their prey (including northern shrimp) was significantly correlated with temperature in the eastern Bering Sea (Kihara and Shimada, 1988). Also, the extent of cod migrations and their spatial distribution are maximized when stock abundance is high and Bering Sea temperatures are warm (Stepanenko, 1997). So, effects of warmer temperature on the movement of Pacific cod into nearshore bays, and associated increases in shrimp predation, seem quite plausible for the northern Gulf of Alaska. Additional support for the role of predation on shrimp declines comes from a detailed study of the Kachemak Bay stock of northern shrimp. Natural mortality was the most important factor controlling this stock during the late 1970s and early 1990s (Fu et al., 1999). In particular, a strong increasing trend of natural mortality seems to have been responsible for the crash of the stock in the 1980s. This trend virtually parallels increasing trends in cod abundance from trawl surveys in Kachemak Bay, indicating that intensified predation by cod was likely to be responsible for the shrimp decline (Fu and Quinn, 2000). Moreover, moderate levels of recruitment during the mid-1980s failed to stem the decline, presumably because predation remained high. Similarly, an examination of Pacific cod stomachs led Albers and Anderson (1985) to suggest that cod predation was the cause of the decline of a lightly fished stock of northern shrimp in Pavlof Bay, Alaska. Additional evidence for the role of predation on shrimp population dynamics comes from the North Atlantic Ocean. Worm and Myers (2003) conducted a metaanalysis of data from nine regions in the Northwest and Northeast Atlantic from the Gulf of Maine to the Barents Sea to test whether northern shrimp and Atlantic cod (Gadus morhua) are consistent with top-down (e.g., predation) or bottom-up (e.g., climate forcing) hypotheses. The top-down view emerged as the leading hypothesis, because eight of the nine regions exhibited inverse correlations between shrimp and cod. Exceptions occurred only near the southern range limits of both species on both sides of the Atlantic. Shrimp biomass was strongly negatively related to cod biomass but not to temperatures. Cod biomass was positively related to temperature but the strength of the shrimp–cod relationship weakened at higher temperatures. Another meta-analysis of the relationships between temperature and recruitment in nine North Atlantic cod stocks revealed negative relationships for stocks located in warm water and positive relationships for stocks in cold water (Planque and Frédou, 1999). Detailed studies of specific shrimp populations in the Atlantic Ocean indicate that relationships with cod and temperature may be more complex, but predation was invoked as a partial explanation for shrimp population dynamics in most areas. For instance, the western Scotian Shelf contains broad areas of suitable fine-sediment habitat for shrimp, but distribution and abundance of shrimp is typically restricted by patchy areas of
Long-Term Change 391
marginally suitable temperatures (Koeller, 2000). Shrimp abundance increases to commercially exploitable levels only after ocean temperatures decline to favorable levels. In addition to habitat and temperature, predation by cod contributed to trends in shrimp abundance in this region since 1977 (Koeller, 2000). On the northeast Newfoundland shelf, it appears that the initial increase of shrimp biomass in the early to mid 1980s was unrelated to cod, but a subsequent, larger increase in shrimp in the 1990s was related to the collapse of the cod stock (Lilly et al., 2000). In the Barents Sea, the abundance of cod and shrimp are significantly correlated over 1982 to 1998, and cod stomach data suggest that age 3–6-cod have the strongest impacts (Berenboim et al., 2000). Similar to Worm and Myers (2003), Zheng and Kruse (2000) tested top-down and bottom-up hypotheses involving five crab and three groundfish stocks in the eastern Bering Sea. Only one statistically significant relationship between predator and prey was found, and this involved Pacific cod predation of 1-year-old red king crabs. Using updated datasets, statistical support was found for a predation relationship between yellowfin sole and red king crab, as well (Zheng and Kruse, 2006). However, analyses of Pacific cod stomachs in the eastern Bering Sea do not indicate that cod eat large numbers of young red king crabs. Interestingly, molting adult female red king crabs are consumed by cod in spring in the Bering Sea, but consumption rates cannot explain the crab stock declines (Livingston, 1989). Likewise, in the Gulf of Alaska, field sampling of groundfish stomachs generally do not indicate heavy predation on red king crabs. During 12 years of red king crab pot surveys, only 77 (0.6%) of 46,125 cod stomachs contained red king crabs, some of which may have consumed the crab while in the crab pot (Blau, 1986). A review of multiple studies of the feeding habits of groundfish in the gulf indicated typically low rates of king crab predation (Tyler and Kruse, 1996). However, definitive conclusions are difficult, because of a relative lack of stomach samples from cod sampled in shallow waters where young-of-theyear king crab settle. Also, it must be remembered that most groundfish stomachs are sampled starting in early summer, whereas most adult male red king crabs molt in late winter and early spring. So, the paucity of king crabs in cod stomachs may result from a sampling bias regarding time (adults) and space (young). Conversely, a stronger case exists for the role of predation on Tanner crabs owing to higher consumption rates of ages 0 and 1 crabs (Lang et al., 2003). Despite this empirical evidence for a heavy predation of young Tanner crabs by cod, conclusive evidence of the role of predation on Tanner crabs at the population level remain to be demonstrated. In the Bering Sea, crab population dynamics seem to be largely driven by fluctuations in recruitment, with two notable exceptions. Sharp increases in mortality of juvenile and adult crabs were associated with the declines of red king crabs in Bristol Bay in the early 1980s (Zheng et al., 1995a,b) and blue king crabs off St. Matthew Island in the late 1990s (Zheng and Kruse, 2002). It remains unclear whether these increases in mortality were attributable to predation or other factors, but high rates of
392
Long-Term Ecological Change in the Northern Gulf of Alaska
predation on adult and subadult crabs are unprecedented. Moreover, although fishing is highly unlikely to be involved in the decline off St. Matthew Island, the potential for fishing effects (e.g., discards in the directed crab fishery and/or bycatch in trawl fisheries) cannot be discounted in the case of elevated mortality of red king crabs in Bristol Bay. Mass mortality from disease and sharp changes in environmental conditions (e.g., temperature) are alternative speculations.
4.10.6. Fishing Effects Kruse et al. (1996) calculated an “overfishing” biological reference point for Kodiak red king crabs based on F30% (equating to a 41.1% harvest rate of legal males), the fishing mortality rate that reduces spawning stock biomass of male crabs to 30% of the unfished levels. If one accepts this definition, then it appears that overfishing occurred during 1966–1968 and 1980–1982. Abundance of legal male king crabs declined coincident with overfishing from 30 million crabs in 1964–1965 to 6 million in the late 1960s. During the autumn 1966, Kodiak fishermen began to report significant numbers of mature barren females (McMullen and Yoshihara, 1969). Much sampling during 1968 found large concentrations of unmated mature females on the southeast side of Kodiak Island (McMullen, 1968; McMullen and Yoshihara, 1969). In April–May 1968, trawl samples revealed that 76% of mature females in Kaguyak Bay were not carrying eggs and had not mated – a most unusual observation. Female:male sex ratio was 72:1. This unusual percentage of unmated females was attributed to heavy fishing (Powell et al., 1973; McMullen and Yoshihara, 1969). In Kodiak, harvest rates were reduced to 14–33% during 1969–1972 and legal males rebounded to 14 million crabs in 1972. However, the stock crashed to <1 million legal males in 1983 after severe overfishing in the early 1980s (harvest rates peaked at 80% in 1982). In a more formal assessment of stock abundance and harvest rates, Collie and Kruse (1998) likewise concluded that harvest rates <40% generally corresponded to periods of stable or increasing abundance and periods of harvest >40% corresponded to periods of stock decline and that the highest harvest rates (>80% for Kodiak) corresponded to the sharpest declines. Although there is evidence for periods of overfishing, red king crab population dynamics are very complex and cannot be explained by fishing alone (Kruse et al., 1996). Orensanz et al. (1998) went beyond these findings for Kodiak red king crab and concluded that evidence of overfishing of Alaskan crustacean stocks in general was compelling: “From a managerial perspective, the pattern and magnitude of the collective rise and fall of the crustacean fisheries of Alaska are such that overfishing has to be considered as the default working scenario.” The motivation for this conclusion was the serial pattern of collapse, starting with the most valuable species closest to ports to the most-distant, least-valuable species.
Long-Term Change 393
While this argument is intriguing, Orensanz’s conclusions are controversial and, clearly, declines of all Alaskan crustacean stocks cannot be explained by fishing alone. Year-class strength of red king crabs in Bristol Bay began declining after 1970, well before the heaviest catches in the late 1970s and 1980. Also, some lightly fished northern shrimp stocks experienced similar declines to shrimp stocks that were fished (Anderson, 1991). Similarly, a simulation study of northern shrimp in Kachemak Bay showed that no alternative harvest strategies could have altered the fate of the stock, because natural mortality, not fishing mortality, was the most important contributor to its collapse (Fu et al., 1999). If fishing had a greater role in crab and shrimp declines, operative mechanisms would seem to have to involve effects of fishing that have yet to be demonstrated, such as genetic selection, disruption of mating systems, or deleterious effects on stock productivity by habitat disturbance. Research has begun to investigate such subtle but potentially insidious effects of fishing in the North Atlantic, and similar studies should be conducted in Alaska.
4.10.7. Conclusions Causes of crab and shrimp declines in the late 1970s and early 1980s remain uncertain, but some conclusions can be drawn about probable factors. For northern shrimp, cod predation appears to have played an important role in the declines. A regime shift of the late 1970s was not only associated with several strong cod year-classes, but warmer temperatures may have facilitated an influx of cod into bays occupied by shrimp. Laboratory experiments and field studies in the North Atlantic on the same species suggest that the deleterious effect of warmer temperatures on shrimp reproduction may have triggered a series of year-class failures at the same time that predation increased in the late 1970s. Evidence is weak, but equivocal, for a prominent role of a match–mismatch hypothesis in shrimp declines. Regarding red king crabs, some connection to Aleutian low-Pressure dynamics, as indexed by the PDO, seems likely, given the similarity of trends of red king crabs stocks over the broad expanse of the northern Gulf of Alaska. However, the fact that king crab recruitment began to decline in the early 1970s and continued through the 1980s, suggests a progressive deterioration in conditions favorable to their survival, rather than a simple shift from a regime with one set of environmental conditions in the late 1970s to another regime with a different set of conditions, as suggested for other species. The match–mismatch hypothesis was tested and rejected for red king crabs. Instead, field studies suggest that phytoplankton species composition may be much more important than timing of the zooplankton bloom. In particular, red king crab survival may be favored by predominance of Thalassiosira diatoms in the spring bloom. This diatom is associated with relatively calm winds and a stable water column, presumably associated with the cool-phase of the PDO. Yet, this and other causative
394
Long-Term Ecological Change in the Northern Gulf of Alaska
mechanisms remain speculative. Other beneficial effects of cool ocean temperatures on red king crab cannot be discounted. It seems unlikely that predation played much of a role in the decline of red king crab, but definitive conclusions are not possible owing to the lack of sampling of groundfish predators in shallow, nearshore waters, where young-of-the-year red king crabs settle and during late winter, when adult males molt. Instances of overfishing have been demonstrated for red king crabs, but even in these cases, overfishing cannot fully explain crab stock declines. For Tanner crabs, it seems clear that recruitment is affected by local rather than regional conditions. Causes of Tanner crab population changes are not simplistic, and they seem to involve multiple factors at multiple timescales. Three mechanisms seem to combine to explain most of the variability in Tanner crab recruitment in Bristol Bay: (1) warmer bottom temperatures favor Tanner crab gonadal development and egg incubation, (2) winds blowing from the northeast along the Alaska Peninsula during spring promote coastal upwelling while advecting larvae offshore to soft bottom sediments used as nursery areas, and (3) warmer sea surface temperatures during May–June favor growth of calanoid copepods that are preyed upon by Tanner crab larvae. Cod prey on young Tanner crabs at high levels, but strong associations between cod abundance and Tanner crab recruitment are not readily apparent, at least in the Bering Sea. Studies of Tanner crab stocks in the northern Gulf of Alaska are required to determine the generality of the three mechanisms that may be operating in Bristol Bay. In general, simple, single-factor hypotheses are unlikely to fully explain the remarkable changes in invertebrate populations observed in the northern Gulf of Alaska. Development of comprehensive life-stage-based models (e.g., Tyler and Kruse, 1996a,b, 1997, 1998) is one step toward understanding the causes of shrimp and crab population changes in the northern Gulf of Alaska. Such models are useful, not only because of their ability to evaluate the consistency of alternative hypotheses with available empirical evidence within a logical framework, but they provide a means to organize existing information and to identify gaps in knowledge that require laboratory experiments and field studies. A better understanding of crab and shrimp population dynamics in the northern Gulf of Alaska remains a challenge for invertebrate researchers in the twenty-first century. What ecosystem and regulatory changes must happen for these valuable invertebrate species to recover, and can similar collapses be avoided in the future? Time may provide these answers.
References A’mar, T. (and 39 co-authors). 2003. Stock Assessment and Fishery Evaluation Report for the Groundfish Resources of the Gulf of Alaska. Compiled by the Plan Team for the Groundfish Fisheries of the Gulf of Alaska, North Pacific Fishery Management Council, 605 W. 4th Avenue, Suite 306, Anchorage, AK 99501. Abookire, A. A., J. F. Piatt, and M. D. Robards. 2000. Nearshore fish distributions in an Alaskan estuary in relation to stratification, temperature and salinity. Estuarine, Coastal and Shelf Science 51: 45–59.
Long-Term Change 395
Abookire, A. A. and J. F. Piatt. 2004. Structure of forage fish communities in relation to oceanographic conditions in lower Cook Inlet, Alaska. Mar. Ecol. Progr. Ser. 287: 229–240. Adams, T. C. 2000. Foraging differences and early maternal investment in adult Alaskan Steller sea lions (Eumetopias jubatus). PhD thesis, Galveston, TX, Texas A&M University. Agler, B. A., S. J. Kendall, D. B. Irons, and S. P. Klosiewski. 1999. Declines in marine bird populations in Prince William Sound, Alaska coincident with a climatic regime shift. Waterbirds 22: 98–103. Ainley, D. G., R. G. Ford, E. D. Brown, R. M. Suryan, and D. B. Irons. 2003. Prey resources, competition, and geographic structure of kittiwake colonies in Prince William Sound. Ecology 84: 709–723. Albers, W. D., and P. J. Anderson. 1985. Diet of Pacific cod, Gadus macrocephalus, and predation on the northern pink shrimp, Pandalus borealis, in Pavlof Bay, Alaska. Fish. Bull. 83: 601–610. Alderdice, D. F., and C. R. Forrester. 1971. Effects of salinity, temperature, and dissolved oxygen on early development of the Pacific cod (Gadus macrocephalus). J. Fish. Res. Board Can. 28: 883–902. Alexander, M. A., I. Blade, M. Newman, J. R. Lanzante, N.-C. Lau, and J. D. Scott. 2002. The atmospheric bridge: the influence of ENSO teleconnections on air–sea interaction over the global ocean. J. Clim. 15: 2205–2231. Alverson, D. L. 1992. A review of commercial fisheries and the Steller sea lion (Eumetopias jubatus): the conflict arena. Rev. Aquat. Sci. 6: 203–256. Anderson, P. J. 1991. Age, growth, and mortality of the northern shrimp, Pandalus borealis, in Pavlof Bay, Alaska. Fish. Bull. 83: 601–610. Anderson, P. J. 2000. Pandalid shrimp as indicators of ecosystem regime shift. J. Northwest Atlantic Fish. Sci. 27: 1–10. Anderson, P. J. and J. F. Piatt. 1999. Community reorganization in the Gulf of Alaska following ocean climate regime shift. Mar. Ecol. Prog. Ser. 189: 117–123. Anderson, P. J., J. E. Blackburn, and B. A. Johnson. 1997. Declines of forage species in the Gulf of Alaska, 1972–95, as indicator of regime shift. In: Proceedings of the international symposium on the role of forage fishes in marine ecosystems. Alaska Sea Grant College Program Report 97-01, University of Alaska, Fairbanks, pp. 531–543. Anker-Nilssen, T., R. T. Barrett, and J. V. Krasnov. 1997. Long- and short-term responses of seabirds in the Norwegian and Barents Seas to changes in stocks of prey fish. In: Proceedings of the International Symposium on the role of Forage Fishes in Marine Ecosystems, University of Alaska Sea Grant Rep. 97-01. University of Alaska, Fairbanks, pp. 683–698. Anonymous. 1993. Is it food? Addressing marine mammal and seabird declines. Fairbanks, AK, Alaska Sea Grant College Program. 59 pp. Apollonio, S., D. K. Stevenson, and E. E. Dunton, Jr. 1986. Effects of temperature on the biology of the northern shrimp, Pandalus borealis, in the Gulf of Maine. NOAA Technical Report NMFS 42. Bailey, K. M. 2000. Shifting control of recruitment of walleye pollock, Theragra chalcogramma, after a major climatic and ecosystem change. Mar. Ecol. Prog. Ser. 198: 215–224. Bailey, K. M. and E. D. Houde. 1989. Predation on eggs and larvae of marine fishes and the recruitment problem. Adv. Mar. Biol. 25: 1–83.
396
Long-Term Ecological Change in the Northern Gulf of Alaska
Bailey, K. M. and S. A. Macklin. 1994. Analysis of patterns in larval walleye pollock Theragra chalcogramma survival and wind mixing events in Shelikof Strait, Gulf of Alaska. Mar. Ecol. Prog. Ser. 113: 1–12. Bailey, K. M. and S. J. Picquelle. 2002. Larval distribution of offshore spawning flatfish in the Gulf of Alaska: Potential transport pathways and enhanced onshore transport during ENSO. Mar. Ecol. Prog. Ser. 236: 205–217. Bailey, K. M., R. D. Brodeur, and A. B. Hollowed. 1996. Cohort survival patterns of walleye pollock, Theragra chalcogramma, in Shelikof Strait, Alaska: A critical factor analysis. Fish. Oceanogr. 5: 179–188. Baird, R. W. and L. M. Dill. 1995. Occurrence and behaviour of transient killer whales: seasonal and pod-specific variability, foraging behaviour and prey handling. Can. J. Zool. 73: 1300–1311. Barrett, R. T. 1996. Egg laying, chick growth, and food of kittiwakes Rissa tridactyla at Hopen, Svalbard. Polar Res. 15: 107–113. Beamish, R. J. 1993. Climate and exceptional fish production off the west coast of North America. Can. J. Fish. Aquat. Sci. 50: 2270–2291. Beamish, R. J. and C. Mahnken. 2001. A critical size and critical period hypothesis to explain the natural regulation of salmon abundance and the linkage to climate and climate change. Prog. Oceanogr. 49: 423–427. Beamish, R. J. and D. R. Bouillon. 1993. Pacific salmon production trends in relation to climate. Can. J. Fish. Aquat. Sci. 50: 1002–1016. Beamish, R. J., D. J. Noakes, G. A. McFarlane, L. Klyashtorin, V. V. Ivanov, and V. Kurashov. 1999. The regime concept and natural trends in the production of Pacific salmon. Can. J. Fish. Aquat. Sci. 56: 516–526. Bechtol, W. R. 1997. Changes in forage fish populations in Kachemak Bay, Alaska, 1976–95. In Proceedings of the international symposium on the role of forage fishes in marine ecosystems. Alaska Sea Grant College Program Report 97-01, University of Alaska, Fairbanks, pp. 441–455. Berenboim, B. J., A. V. Dolgov, V. A. Korzhev, and N. A. Yaragina. 2000. The impact of cod on the dynamics of Barents Sea shrimp (Pandalus borealis) as determined by multispecies models J. Northwest Atlantic Fish. Sci. 27: 69–75. Bickham, J. W., J. C. Patton, and T. R. Loughlin. 1996. High variability for control region sequences in a marine mammal; implications for conservation and maternal phylogeny of Steller sea lions (Eumetopias jubatus). J. Mammol. 77: 95–108. Bigg, M. A. 1985. Status of the Steller sea lion (Eumetopias jubatus) and California sea lion (Zalophus californianus) in British Columbia. Can. Spec. Pub. Fish. Aquat. Sci. 77: 1–20. Bigler, B. S., D. W. Welch, and J. H. Helle. 1996. A review of size trends among North Pacific salmon (Oncorhynchus spp.). Can. J. Fish. Aquat. Sci. 53: 455–465. Birkhead, T. R. 1976. Breeding biology and survival of Guillemots Uria aalge. Unpublished PhD thesis, University of Oxford, England. Birkhead, T. R. and D. N. Nettleship. 1987. Ecological relationships between common murres, Uria aalge, and thick-billed murres, Uria lomvia, at the Gannet Islands, Labrador. II. Breeding success and site characteristics. Can. J. Zool. 65: 1630–1637. Birkhead, T. R. and D. N. Nettleship. 1988. Breeding performance of black-legged kittiwakes, Rissa tridactyla, at a small, expanding colony in Labrador. Can. Field Naturalist 102: 20–24. Blau, S. F. 1986. Recent declines of red king crab (Paralithodes camtschatica) populations and reproductive conditions around the Kodiak Archipelago, Alaska. In: G. S. Jamieson
Long-Term Change 397
and N. Bourne, (Eds.), North Pacific workshop on stock assessment and management of invertebrates. Canadian Special Report of Fisheries and Aquatic Sciences 92, pp. 360–369. Bodkin, J. L. 2003. Sea otter. In: G. A. Feldhammer, B. C. Thompson and J. A. Chapman (Eds.), Wild mammals of North America. Baltimore, MD, Johns Hopkins University Press, pp. 735–743. Bodkin, J. L., A. M. Burdin, and D. A. Ryazanov. 2000. Age and sex-specific population structure in sea otters. Mar. Mam. Sci. 16: 201–219. Bodkin, J. L., B. E. Ballachey, M. A. Cronin and K. T. Scribner. 1999. Population demographics and genetic diversity in remnant and translocated populations of sea otters. Conser. Biol. 13: 1378–1385. Bodkin, J. L., B. E. Ballachey, T. A. Dean, A. K. Fukuyama, S. C. Jewett, L. McDonald, D. H. Monson, C. E. O’Clair, and G. R. VanBlaricom. 2002. Sea otter population status and the process of recovery from the 1989 ‘Exxon Valdez’ oil spill. Mar. Ecol. Prog. Ser. 241: 237–253. Boersma, P. D., J. K. Parrish, and A. B. Kettle. 1995. Common Murre abundance, phenology and productivity on the Barren Islands, Alaska: The Exxon Valdez oil spill and longterm environmental change. In: P. G. Wells, J. N. Butler, and J. S. Hughes (Eds.), Exxon Valdez Oil Spill: Fate and effects in Alaskan waters. American Society for Testing and Materials, Philadelphia, PA. ASTM STP 1219, pp. 820–853. Bond, N. A., J. E. Overland, M. Spillane, and P. Stabeno. 2003. Recent shifts in the state of the North Pacific. Geophys. Res. Lett. 30: 2183–2186. Boveng, P. L., J. L. Bengtson, D. E. Withrow, J. C. Cesarone, M. A. Simpkins, K. J. Frost, and J. J. Burns. 2003. The abundance of harbor seals in the Gulf of Alaska. Mar. Mam. Sci. 19: 111–127. Bowen, W. D. and G. Harrison. 1994. Offshore diets of grey seals (Halichoerus grypus) near Sable Island, Canada. Mar. Ecol. Prog. Ser. 112: 1–11. Bowen, W. D. and G. Harrison. 1996. Comparison of harbour seal diets in two habitats of Atlantic Canada. Can. J. Zool. 74: 125–135. Bowen, W. D., D. Ellis, S. J. Iverson, and D. J. Boness. 2001. Maternal effects on offspring growth rate and weaning mass in harbour seals. Can. J. Zool. 79: 1088–1101. Bowen, W. D., D. Ellis, S. J. Iverson, and D. J. Boness. 2003. Maternal and newborn lifehistory traits during periods of contrasting population trends: implications for explaining the decline of harbour seals, Phoca vitulina, on Sable Island. J. Zool. London 261: 155–163. Brandon, E. A. 2000. Maternal investment in Steller sea lions in Alaska. Ph.D. thesis, Galveston, TX, Texas A&M University. Brett, J. R. and D. D. Groves. 1979. Physiological energetics. VII: bioenergetics and growth. In: Hoar, W. S. and D. J. Randall (Eds.), Fish Physiology. Academic Press Inc., pp. 261–342. Brodeur, R. D. and D. M. Ware. 1992. Long term variability in zooplankton biomass in the subarctic Pacific Ocean. Fish. Oceanogr. 1: 32–38. Brodeur, R. D. and D. M. Ware. 1995. Interdecadal variability in distribution and catch rates of epipelagic nekton in the Northeast Pacific Ocean. In: R. J. Beamish (Ed.), Climate Change and Northern Fish Populations. Can. Spec. Publ. Fish. Aquat. Sci. 121: 329–356. Brodeur, R. D. and M. T. Wilson. 1996. A review of the distribution, ecology and population dynamics of age-0 walleye pollock in the Gulf of Alaska. Fish. Oceanogr. 5: 148–166.
398
Long-Term Ecological Change in the Northern Gulf of Alaska
Brodeur, R. D., B. Frost, S. R. Hare, R. C. Francis, and W. J. Ingerham. 1996. Interannual variations in zooplankton biomass in the Gulf of Alaska and covariations with the California Current zooplankton biomass. CalCOFI Rept. 37: 80–99. Brodeur, R. D., K. W. Myers, and J. H. Helle. 2003. Research conducted by the United States on the early ocean life history of Pacific salmon. In: P. Symons (Ed.), A Review of the Research on the Early Marine Period of Pacific Salmon by Canada. Japan, Russia, and the United States. N. Pac. Anadrom. Fish. Comm. Bull. 3: 89–131. Brodeur, R. D., R. C. Francis, B. W. Frost, W. J. Ingraham, Jr., and S. R. Hare. 1996. Interannual variations in zooplankton biomass in the Gulf of Alaska, and covariation with California Current zooplankton biomass. Scripps Institute, University of California, San Diego, CalCOFI Report 37: 80–99. Brodeur, R. D., S. J. Picquelle, D. M. Blood, and N. Merati. 1996. Walleye pollock egg distribution and mortality in the western Gulf of Alaska. Fish. Oceanogr. 5: 92–111. Brown, E. D. 2003. Stock structure and environmental effects on year class formation and population trends on Pacific herring, Clupea pallasi, in Prince William Sound, Alaska. PhD Dissertation, University of Alaska Fairbanks, Fairbanks, Alaska. Bryan, J. D., S. W. Kelsch, and W. H. Neill. 1990. The maximum power principle in behavioral thermoregulation by fishes. Trans. Am. Fish. Soc. 119: 611–621. Bryant, R., I. L. Jones, and J. M. Hipfner. 1999. Responses to changes in prey availability by common murres and thick-billed murres at the Gannet Islands, Labrador. Can. J. Zool. 77: 1278–1287. Burger, A. E. and J. F. Piatt. 1990. Flexible time budgets in breeding common murres: Buffers against variable prey availability. Studies in Avian Biology 14: 71–83. Burn, D. M. and A. M. Doroff. 2005. Decline in sea otter populations along the Alaska Peninsula, 1986–2001. Fish. Bull. 103: 270–279. Byrd, G. V. 1989. Seabirds in the Pribilof Islands, Alaska: trends and monitoring methods. MS Thesis, University of Idaho, Moscow, ID. Byrd, G. V., J. C. Williams, and R. Walder. 1992. Status and biology of the tufted puffin in the Aleutian Islands, Alaska after a ban on salmon drift nets. U.S. Fish and Wildlife Service, Adak, AK. Calkins, D. and E. Goodwin. 1988. Investigation of the decline of Steller sea lions in the Gulf of Alaska. Anchorage, AK, Alaska Department of Fish and Game, 76 pp. Calkins, D. C., D. C. McAllister, K. W. Pitcher, and G. W. Pendleton. 1999. Steller sea lion status and trend in southeast Alaska: 1979–1997. Mar. Mam. Sci. 15: 462–477. Calkins, D. G., E. F. Becker, and K. W. Pitcher. 1998. Reduced body size of female Steller sea lions from a declining population in the Gulf of Alaska. Mar. Mam. Sci. 14: 232–244. Chelton, D. and R. W. Davis. 1982. Monthly mean sea-level variability along the west coast of North America. J. Phys. Oceanogr. 12: 757–779. Chelton, D. B. and R. E. Davis. 1982. Monthly mean sea-level variability along the west coast of North America. J. Phys. Oceanogr. 12: 757–784. Childers, A., T. E. Whitledge, and D. A. Stockwell. Seasonal and interannual variability in the distribution of nutrients and chlorophyll-a across the Gulf of Alaska shelf, 2005, Deep-Sea Res. II 52: 193–216. Collie, J. S., and G. H. Kruse. 1998. Estimating king crab (Paralithodes camtschaticus) abundance from commercial catch and research survey data. In: G. S. Jamieson and A. Campbell, (Eds.), Proceedings of the North Pacific symposium on invertebrate stock assessment and management. Canadian Special Publication of Fisheries and Aquatic Sciences 125, pp. 73–83.
Long-Term Change 399
Cooney, R. T. and R. D. Brodeur. 1998. Carrying capacity and North Pacific salmon production: stock enhancement implications. Bull. Mar. Sci. 62(2): 443–464. Cooney, R. T., J. R. Allen, M. A. Bishop, D. L. Eslinger, T. Kline, B. L. Norcross, C. P. McRoy, J. Milton, J. Olsewn, V. Patrick, A. J. Paul, D. Salon, D. Scheel, G. L. Thomas, S. L. Vaughan, and M. Willette. 2001. Ecosystem controls of juvenile pink salmon (Oncorhynchus gorbuscha) and Pacific herring (Clupea pallasi) populations in Prince William Sound, Alaska. Fish. Oceanogr. 10(Suppl. 1): 1–13. Cooney, R. T., T. M. Willette, S. Sharr, D. Sharp, and J. Olsen. 1995. The effect of climate on North Pacific pink salmon (Oncorhynchus gorbuscha) production: examining some details of a natural experiment. In R. J. Beamish (Ed.), Climate Change and Northern Fish Populations. Can. Spec. Publ. Fish. Aquat. Sci., 121: 475–482. Coyle, K. O., and A. J. Paul. 1990. In: D. A. Ziemann and K. W. Fulton-Bennett (Eds.), APPRISE – Interannual variability and fisheries recruitment. The Oceanic Institute, Honolulu, HI, pp. 179–228. Cury, P. and C. Roy. 1989. Optimal environmental window and pelagic fish recruitment success in upwelling areas. Can. J. Fish. Aquat. Sci. 46: 670–680. Cushing, D. H. 1969. The regularity of the spawning season of some fishes. International Council for the Exploration of the Sea, Journal du Conseil 33: 81–92. Cushing, D. H. 1995. Population production and regulation in the sea: a fisheries perspective. Cambridge University Press, Cambridge, UK. Davis, R. W., E. A. Brandon, T. C. Adams, T. M. Williams, M. Castellini, T. R. Loughlin, and D. C. Calkins. 1996. Indices of reproductive effort, pup condition and pup growth for Steller sea lions (Eumetopias jubatus) in Alaska. Alaska Dept. of Fish and Game, Wildlife Tech. Bull. 13: 53–68. Davis, R. W., T. C. Adams, E. A. Brandon, D. C. Calkins, and T. R. Loughlin 2002. Female attendance, lactation, and pup growth in Steller sea lions. In: D. DeMaster and S. Atkinson (Eds.), Steller sea lion decline: is it food II? Fairbanks, AK, University of Alaska Sea Grant. AK-SG-02-02, pp. 23–27. Dean, T. A., J. L. Bodkin, A. K. Fukuyama, S. C. Jewett, D. H. Monson, C. E. O’Clair, and G. R. VanBlaricom. 2002. Food limitation and the recovery of sea otters in Prince William Sound. Mar. Ecol. Prog. Ser. 241: 255–270. DeGange, A. R. and G. A. Sanger 1986. Marine birds. In: D. W. Hood and S. T. Zimmerman (Eds.), The Gulf of Alaska. NOAA/MMS, Anchorage, Alaska. pp. 479–524. DeMaster, D. and S. Atkinson (Eds.). 2002. Steller sea lion decline: Is it food II. University of Alaska Sea Grant, AK-SG-02-02, Fairbanks, AK, 80 pp. Dettinger, M. D., D. S. Battisti, G. J. McCabe, C. M. Bitz, and R. D. Garreaud. 2001. Interhemispheric effects of interannual and decadal ENSO-like climate variations on the Americas. pp. 1–22. In: V. Markgraf (Ed.), Interhemispheric Climate Linkages. Academic Press, New York, 454 pp. DFO. 2003. Steller sea lion (Eumetopias jubatus). Nanaimo, B.C., Canada, Department of Fisheries and Oceans, Stock Status Report 2003/037. Dorn, M., S. Barbeaux, M. Guttormsen, B. Megrey, A. Hollowed., M. Wilkins, and K. Spalinger. 2003. Assessment of walleye pollock in the Gulf of Alaska. In Stock Assessment and Fishery Evaluation Report for the Groundfish Resources of the Gulf of Alaska. Compiled by the Plan Team for the Groundfish Fisheries of the Gulf of Alaska, North Pacific Fishery Management Council, 605 W. 4th Avenue, Suite 306, Anchorage, AK 99501.
400
Long-Term Ecological Change in the Northern Gulf of Alaska
Doroff, A. M. and A. R. DeGange. 1994. Sea otter Enyhdra lutris prey, composition and foraging success in the northern Kodiak Archipelago. Fish. Bull. 92: 704–710. Doroff, A. M., J. A. Estes, M. T. Tinker, D. M. Burn, and T. J. Evans. 2003. Sea otter population decline in the Aleutian Archipelago. J. Mammol. 84: 55–64. Dragoo, D. E., G. V. Byrd, and D. B. Irons. 2000. Breeding status and population trends of seabirds in Alaska in 1999. U.S. Fish and Wildlife Service Report AMNWR 2000/02. Homer, AK. Dragoo, D. E., G. V. Byrd, and D. B. Irons. 2004. Breeding status, population trends and diets of seabirds in Alaska, 2002. U.S. Fish and Wildlife Service Report AMNWR 04/15. Drew, G. 2002. Primary and Secondary Production in Lower Cook Inlet. In J. F. Piatt (Ed.), Response of seabirds to fluctuations in forage fish density. Final Report to Exxon Valdez Oil Spill Trustee Council (Restoration Project 01163M) and Mineral Management Service (Alaska OCS Study MMS 2002-068). Alaska Science Center, U.S. Geological Survey, Anchorage, Alaska, pp. 27–32. Drew, G., and Piatt, J. F. 2002. Oceanography of Lower Cook Inlet. Pages 18–26 In: J.F. Piatt (ed.). Response of seabirds to fluctuations in forage fish density. Final Report to Exxon Valdez Oil Spill Trustee Council (Restoration Project 01163M) and Mineral Management Service (Alaska OCS Study MMS 2002-068). Alaska Science Center, U.S. Geological Survey, Anchorage, Alaska. Duffy-Anderson, J. T., K. M. Bailey, and L. Ciannelli. 2002. Consequences of a superabundance of larval walleye pollock, Theragra chalcogramma, in the Gulf of Alaska in 1981. Mar. Ecol. Prog. Ser. 243: 179–190. Ebbesmeyer, C. C., D. R. Cayan, F. H. McLain, D. H. Peterson, and K. T. Redmond. 1991. 1976 step in the Pacific climate: Forty environmental changes between 1968–1975 and 1977–1985. In: J. L. Betancourt and V. L. Tharp (Eds.), Proceedings of the Seventh Annual Pacific Climate Workshop. California Department of Water Resources Interagency Ecological Studies Program Technical Report 26: pp. 129–141. Enfield, D. B. and A. M. Mestas-Nuñez. 1999. Multiscale variabilities in global sea surface temperature and their relationships with tropospheric climate patterns. J. Clim. 12: 2719–2733. Erikstad, K. E., P. Fauchald, T. Tveraa, and H. Steen. 1998. On the cost of reproduction in long-lived birds; the influence of environmental variability. Ecol. 79: 1781–1788. Erikstad, K. E., T. Tveraa, and R. T. Barrett. 1995. Adult survival and chick production in long-lived seabirds: a 5-year study of the kittiwake Rissa tridactyla. In: Skjoldal, H. R. C. Hopkins, K. E. Erikstad, and H. P. Leinaas (Eds.), Ecology of Fjords and Coastal Waters. Elsevier Science, London, pp. 471–477. Eschmeyer, W. N., E. S. Herald, and H. Hammann, 1983. A Field Guide to Pacific Coast Fishes of North America. Houghton Mifflin Company. Eslinger, D. L., R. T. Cooney, C. P. McRoy, A. Ward, T. C. Kline, Jr., E. P. Simpson, J. Wang, and J. R. Allen. 2001. Population dynamics: observed and modeled responses to physical conditions in Prince William Sound, Alaska. Fish. Oceanogr. 10(Suppl. 1): 81–96. Estes, J. A. and J. L. Bodkin. 2002. Marine Otters. In: W. F. Perrin, B. Wursig, J. G. M. Thewissen and C. R. Crumly (Eds.), Encyclopedia of Marine Mammals. Academic Press, pp. 843–858. Estes, J. A., D. O. Duggins, and G. B. Rathbun. 1989. The ecology of extinctions in kelp forests. Cons. Biol. 3:252–264. Estes, J. A., E. M. Danner, D. F. Doak, B. Konar, A. M. Springer, P. D. Steinberg, M. T. Tinker, and T. M. Williams. 2004. Complex trophic interactions in kelp forest ecosystems. Bull. Mar. Sci. 74: 621–638.
Long-Term Change 401
Estes, J. A., M. T. Tinker, T. M. Williams, and D. F. Doak. 1998. Killer whale predation on sea otters: linking oceanic and nearshore ecosystems. Science 282: 473–476. Fancett, M. S. and G. P. Jenkins. 1988. Predatory impact of scyphomedusae on ichthyoplankton and other zooplankton in Port Phillip Bay. J. Exp. Mar. Biol. Ecol. 116: 63–77. Finney, B. P., I. Gregory-Eaves, M. S. V. Douglas, and J. P. Smol. 2002. Fisheries productivity in the northeastern Pacific Ocean over the past 2,200 years. Nature 416: 729–733. Fiscus, C. H. 1961. Growth in the Steller sea lion. J. Mammol. 47: 195–200. Fiscus, C. H. and G.A. Baines. 1966. Food and feeding behavior of Steller and California sea lions. J. Mammol. 47: 195–200. Fowler, C. W. 1987. Marine debris and northern fur seals: a case study. Mar. Poll. Bull. 18: 326–335. Francis, R. C. and S. R. Hare. 1994. Decadal-scale regime shifts in the large marine ecosystems of the North-east Pacific: a case for historical science. Fish. Oceanogr. 3/4: 279–291. Francis, R. C. and T. H. Sibley. 1991. Climate change and fisheries: what are the real issues? Northwest Environ. J. 7: 295–307. Francis, R. C., S. R. Hare, A. B. Hollowed, and W. S. Wooster. 1998. Effects of interdecadal climate variability on the oceanic ecosystems of the NE Pacific. Fish. Oceanogr. 7: 1–21. Francis, R.C., L.G. Anderson, W.D. Bowen, S.K. Davis, J.M. Grebmeier, L.F. Lowry, I. Merculieff, N.S. Mirovitskaya, C.H. Peterson, C. Pungowiyi, T.G. Royer, A.M. Springer, and W.S. Wooster. 1996. The Bering Sea Ecosystem. Washington D.C., The National Academies Press, 307 pp. Freeland, H. 2002. The Heat Flux across Line P 1996–1999. Atmosphere–Ocean 40: 81–89. Freeland, H. and F. Whitney. 2000. Climate Changes: Gulf of Alaska. In: C. Sheppard (Ed.), Seas at the Millennium: An Environmental Evaluation. Elsevier Science Ltd., pp. 179–186. Freeland, H., K. Denman, C. S. Wong, F. Whitney, and R. Jacobs. 1997. Evidence of change in the winter mixed layer depth in the Northeast Pacific Ocean. Deep-Sea Res. 44: 2117–2129. Fritz, L.W. and S. Hinckley. 2005. A critical review of the regime shift—”junk food”— hypothesis for the decline of the western stock of Steller sea lion. Mar. Mam. Sci. 21: 476–518. Frost, B. W. 1983. Interannual variation of zooplankton standing stock in the open Gulf of Alaska. In: W. S. Wooster (Ed.), From Year to Year. Interannual Variability of the Environment and Fisheries of the Gulf of Alaska and the Eastern Bering Sea. Washington Sea Grant Program. University of Washington, Seattle, pp. 146–157. Frost, K. J., L. F. Lowry, and J. M. Ver Hoef. 1999. Monitoring the trend of harbor seals in Prince William Sound, Alaska, after the Exxon Valdez oil spill. Mar. Mam. Sci. 15: 494–506. Frost, K. J., L. F. Lowry, E. H. Sinclair, and J. M. Ver Hoef. 1994. Impacts on distribution, abundance, and productivity of harbor seals. In: T. R. Loughlin (Ed.), Marine mammals and the Exxon Valdez. San Diego, CA, Academic Press, pp. 97–118. Fu, C., and T. J. Quinn, II. 2000. Estimability of natural mortality and other population parameters in a length-based model: Pandalus borealis in Kachemak Bay, Alaska. Can. J. Fish. Aquat. Sci. 57: 2420–2432. Fu, C., T. J. Quinn, II, and G. H. Kruse. 2000. Analyses of harvest strategies for pandalid shrimp populations. J. Northwest Atlantic Fish. Sci. 27: 247–260. Fu, C., T. J. Quinn, II, and M. D. Adkison. 1999. Retrospective projection using Monte Carlo simulation: an application of a length-based model to Kachemak Bay pink shrimp. In: Proceedings of the international symposium on the role of forage fishes in
402
Long-Term Ecological Change in the Northern Gulf of Alaska
marine ecosystems. Alaska Sea Grant College Program Report 97-01, University of Alaska Fairbanks, pp. 59–77. Fulton, J. 1983. Seasonal and annual variations of net zooplankton at Ocean Station “P”, 1956–1980. Can. Data Rep. Fish. Aquat. Sci. 374. Funk, F. and M. Harris. 1992. Preliminary forecasts of catch and stock abundance for 1992 Alaska herring fisheries. Alaska Dept. Fish and Game, Division of Commercial Fisheries, Information Report No. 5J92-04. Juneau, Alaska. Furness, R. W. and C. J. Camphuysen. 1997. Seabirds as monitors of the marine environment. ICES J. Mar. Sci. 54: 726–737. Gargett, A. 1997. The optimum stability ‘window’: a mechanism underlying decadal fluctuations in North Pacific salmon stocks? Fish. Oceanogr. 6: 109–117. Garshelis, D. L., J. A. Garshelis, and A. T. Kimker. 1986. Sea otter time budgets and prey relationships in Alaska. J Wildlife Manage. 50: 637–647. Gaudet, D. M. 1984. Prince William Sound herring survey, 1982. Alaska Department of Fish and Game, Prince William Sound Technical Data Report No. 1984-2. Gharrett, A. J., A. K. Gray, and J. Heifetz. 2001. Identification of rockfish (Sebastes spp.) from restriction sites analysis of the mitochondrial NM-3/ND-4 and 12S/16S rRNA gene regions. Fish. Bull. 99: 49–62. Gill, V. A. 1999. Breeding performance of black-legged kittiwakes (Rissa tridactyla) in relation to food availability: a controlled feeding experiment. M.S. Thesis, University of Alaska, Anchorage. Goes, J. I., H. R. Gomes, A. Limsakul, W. M. Balch, and T. Saino, 2001: El Niño related interannual variations in biological production in the North Pacific as evidenced by satellite and ship data. Progr. Oceanogr., 49: 211–225. Golet, G. H., D. B. Irons, and J. A. Estes. 1998. Survival costs of chick rearing in blacklegged kittiwakes. J.Anim. Ecol. 67: 827–841. Golet, G. H., J. A. Schmutz, D. B. Irons, and J. A. Estes. 2004. Determinants of reproductive costs in the long-lived black-legged kittiwake: a multiyear experiment. Ecol. Monogr. 74: 353–372. Golet, G. H., P. E. Seiser, A. D. McGuire, D. D. Roby, J. B. Fisher, K. J. Kuletz, D. B. Irons, T. A. Dean, S. C. Jewett, and S. H. Newman. 2002. Long-term direct and indirect effects of the ‘Exxon Valdez’ oil spill on pigeon guillemots in Prince William Sound, Alaska. Mar. Ecol. Progr. Ser. 241: 287–304. Gorbics, C. S. and J. L. Bodkin. 2001. Stock structure of sea otters (Enhydra lutris kenyoni) in Alaska. Mar. Mam. Sci. 17:632–647. Grant, W. S., C. I. Zhang, and T. Kobayashi. 1987. Lack of genetic stock discretion in Pacific cod (Gadus macrocephalus). Can. J. Fish. Aquat. Sci. 44: 490–498. Greene, D. H. and J. K. Babbitt. 1990. Control of muscle softening and proteaseparasite interactions in arrowtooth flounder, Atheresthes stomias. J. Food Sci. 55: 579–580. Grice, G. D. and N. J. Wilmovsky. 1957. Herring spawning surveys of Alaska with particular reference to Southeast Alaska. Department of Interior, Bureau of Commercial Fisheries, Alaska Region, Marine Fisheries Investigations, Operation Report-1957. Hamer, K. C., P. Monaghan, J. D. Uttley, P. Walton, and M. D. Burns. 1993. The influence of food supply on the breeding ecology of kittiwakes Rissa tridactyla in Shetland. Ibis 135. Hanselman, D., J. Heifetz, J. T. Fujioka, and J. N. Ianelli. 2003. Gulf of Alaska Pacific ocean perch. In: Stock Assessment and Fishery Evaluation Report for the Groundfish Resources of the Gulf of Alaska. Compiled by the Plan Team for the Groundfish
Long-Term Change 403
Fisheries of the Gulf of Alaska, North Pacific Fishery Management Council, 605 W. 4th Avenue, Suite 306, Anchorage, AK 99501. Harding, A. A., T. van Pelt, J. F. Piatt, and A. S. Kitaysky. 2002. Reduction of provisioning effort in response to experimental manipulation of chick nutritional status in the horned puffin Fratercula corniculata. Condor 104: 842–847. Harding, A. M. 2001. The breeding ecology of horned puffins Fratercula corniculata. M.Sc. Thesis, University of Durham, England. 74 pp. Harding, A. M. A, J. F. Piatt, and K. C. Hamer. 2003. Breeding ecology of horned puffins (Fratercula corniculata) in Alaska: Annual variation and effects of El Niño. Can. J. Zool. 81: 1004–1013. Hare, S. R. and R. C. Francis. 1995. Climate change and salmon production in the northeast Pacific Ocean. In: R. J. Beamish (Ed.), Climate Change and Northern Fish Populations. Can. Spec. Publ. Fish. Aquat. Sci. 121: 357–372. Hare, S. R., and Mantua, N. J. 2000. Empirical evidence for North Pacific regime shifts in 1977 and 1989. Progr. Oceanogr. 47(2–4): 103–146. Hare, S. R., N. J. Mantua, and R. C. Francis. 1999. Inverse production regimes: Alaskan and West Coast Pacific salmon. Fisheries 24: 6–14. Harris, M. P. and S. Wanless. 1988. The breeding biology of guillemots Uria aalge on the Isle of May over a six year period. Ibis 130: 172–192. Harris, M. P. and S. Wanless. 1990. Breeding success of British kittiwakes Rissa tridactyla in 1986-1988: Evidence for changing conditions in the northern North Sea. J. Appl. Ecol. 27: 172–187. Hassell, M. P., J. H. Lawton, and J. R. Beddington. 1977. Sigmoid functional responses by invertebrate predators and parasitoids. J. Anim. Ecol. 46: 249–262. Hatch, S. A. 1983. The fledging of common and thick-billed murres on Middleton Island, Alaska. J. Field Ornith. 54: 266–274. Hatch, S. A. 1987. Did the 1982–83 El Niño-Southern Oscillation affect seabirds in Alaska? Wilson Bulletin 99: 468–474. Hatch, S. A. and M. A. Hatch. 1988. Colony attendance and population monitoring of black-legged kittiwakes on the Semidi Islands, Alaska. Condor 90: 613–620. Hatch, S. A. and M. A. Hatch. 1989. Attendance patterns of murres at breeding sites: implications for monitoring. J. Wildlife Manag. 53: 483–493. Hatch, S. A., and J. F. Piatt. 1995. Seabirds in Alaska, pp. 49–52. In: E. T. LaRoe, G. S. Farris, C. E. Puckett, P. D. Doran, and M. J. Mac (Eds.), Our living resources: A report to the nation on the distribution, abundance, and health or U.S. plants, animals, and ecosystems. U.S. Department of the Interior, National Biological Service, Washington, DC., 530 pp. Hatch, S. A., G. V. Byrd, D. B. Irons, and G. L. Hunt. 1993. Status and ecology of kittiwakes (Rissa tridactyla and R. brevirostris) in the North Pacific. In: K. Vermeer, K. T. Briggs, K. H. Morgan, and D. Siegel-Causey (Eds), The Status, Ecology, and Conservation of Marine Birds of the North Pacific. Special Publication, Canadian Wildlife Service, Ottawa, pp. 140–153. Hatch, S.A. 2003. Statistical power for detecting trends with applications to seabird monitoring. Biol. Conserv. 111: 317–329. Hatchwell, B. J. and T. R. Birkhead. 1991. Population dynamics of common guillemots Uria aalge on Skomer Island, Wales. Ornis Scandinavica 22: 55–59. Hay, D. E. 1985. Reproductive biology of Pacific herring (Clupea harengus pallasi). Can. J. Fish. Aquat. Sci. 42: 111–124.
404
Long-Term Ecological Change in the Northern Gulf of Alaska
Hay, D. E. and A. R. Kronlund. 1987. Factors affecting the distribution, abundance, and measurement of Pacific herring (Clupea harengus pallasi) spawn. Can. J. Fish. Aquat. Sci. 44: 1181–1194. Hedgren, S. 1980. Reproductive success of guillemots Uria aalge on the island of Stora Karlso. Ornis Fennica 57: 49–57. Heise, K. A., L. G. Barrett-Leonard, E. L. Saulitis, C. O. Matkin, and D. Bain. 2003. Examining the evidence for killer whale predation on Steller sea lions in British Columbia and Alaska. Aquat. Mam. 29: 325–334. Helle, J. H. and M. S. Hoffman. 1995. Size decline and older age at maturity of two chum salmon (Oncorhynchus kewta) socks in western North America, 1972–92. In: R. J. Beamish (Ed.), Climate Change and Northern Fish Populations. Can. Spec. Publ. Fish. Aquat. Sci. 121: 245–260. Hermann, A. J., D. B. Haidvogel, E. L. Dobbins, and P. J. Stabeno. 2002. Coupling global and regional circulation models in the coastal Gulf of Alaska, Prog. Oceanogr. 53: 335–367. Hilborn, R. and D. Eggers. 2000. A review of hatchery programs for pink salmon in Prince William Sound and Kodiak Island, Alaska. Trans. Amer. Fish. Soc. 129: 333–350. Holling, C. S. 1959. The components of predation as revealed by a study of small mammal predation of the European pine sawfly. Can. Entomol. 91: 293–320. Hollowed, A. B. and W. S. Wooster. 1992. Variability of winter ocean conditions and strong year classes of Northeast Pacific groundfish. ICES Mar. Sci. Symp. 195: 433–444. Hollowed, A. B. and W. S. Wooster. 1995. Decadal-scale variations in the eastern subarctic Pacific: II. Response of northeast Pacific fish stocks. Can. Spec. Pub. Fish. Aquat. Sci. 121: 373–385. Hollowed, A. B., S. R. Hare, and W. S. Wooster. 1998. Pacific basin Climate variability and patterns of northeast Pacific marine fish production, In: (G. Holloway, P. Muller, and D. Henderson, Eds.), Proceedings on the 10th ‘Aha Huliko’a Hawaiian Winter Workshop on Biotic Impacts of Extratropical Climate Variability in the Pacific. U.S. Department of Commerce, National Oceanic and Atmospheric Administration, NOEST Special Publication Award No. A67RJ0154. pp. 89–104. Hollowed, A. B., S. R. Hare, and W. S. Wooster. 2001. Pacific Basin climate variability and patterns of Northeast Pacific marine fish production. Prog. Oceanogr. 49: 257–282. Hoover, A. A. 1988. Steller sea lion, In: J.A. Lentfer (Ed.), Selected Marine Mammals of Alaska. Washington, D.C., Mar. Mammal Comm. pp. 159–193. Hoover-Miller, A., K. R. Parker, and J. J. Burns. 2001. A reassessment of the impact of the Exxon Valdez oil spill on harbor seals (Phoca vitulina Richardsi) in Prince William Sound, Alaska. Mar. Mam. Sci. 17: 111–135. Houde, E. D., and C. E. Zastrow. 1993. Ecosystem-specific and taxon-specific dynamic and energetics properties of larval fish assemblages. Bull. Mar. Sci. 53: 290–335. Hourston, A. S and C. W. Haegele. 1980. Herring on Canada’s Pacific coast. Dept. Fisheries and Oceans, Ottawa, Can. Spec. Pub. Fish. Aquat. Sci. 48. 23 pp. Hudson, P. J. 1985. Population parameters for the Atlantic Alcidae. In: D. N. Nettleship and T. R. Birkhead (Eds.), The Atlantic Alcidae. Academic Press. New York, pp. 233–261. Hunt, G. L., Jr., P. Stabeno, G. Walters, E. Sinclair, R. D. Brodeur, J. M. Napp, and N. A. Bond. 2002. Climate change and control of the southeastern Bering Sea pelagic ecosystem. Deep-Sea Res. II 49: 5821–5853. Hunt, H. D. 1996. Normal Mode Analysis of Dynamic Height in the Gulf of Alaska, 1968–1990. MS thesis, University of Alaska, 66 pp.
Long-Term Change 405
Imler, R. H. and H. R. Sarber. 1947. Harbor seals and sea lions in Alaska, U.S. Fish and Wildlife Service, Spec. Sci. Rep. No. 28, 28 pp. Interagency Task Group. 1978. Final environmental impact statement. Consideration of a waiver of the moratorium and return of management of certain marine mammals to the State of Alaska. National Marine Fisheries Service and U.S. Fish and Wildlife Service, Washington, DC. IPCC, 2001. J. T. Houghton, Y. Ding, D. J. Griggs, M. Noguer, P. J. van der Linden, X. Dai, K. Maskell, and C. A. Johnson (Eds.), Climate Change 2001: The Scientific Basis, Contribution of Working Group I to the Third Assessment Report of the Intergovernmental Panel on Climate Change, Cambridge University Press, Cambridge, United Kingdom and New York, NY, U.S.A, 881 pp. Iverson, S. J., K. J. Frost, and S. L. C. Lang. 2003. High body energy stores and condition are linked with diet diversity in juvenile harbor seals in Prince William Sound (PWS), Alaska: new insights from quantitative fatty acid signature analysis (QFASA). 15th Biennial Conference on the Biology of Marine Mammals, Greensboro, NC. Jameson, R. J., K. W. Kenyon, A. M. Johnson, and H. M. Wight. 1982. History and status of translocated sea otter populations in North America. Wildlife Soc. Bull. 10: 100–107. Jemison, L. A. and B. P. Kelly. 2001. Pupping phenology and demography of harbor seals (Phoca vitulina richardsi) on Tugidak Island, Alaska. Mar. Mam. Sci. 17: 585–600. Jemison, L. A., G. W. Pendleton, C. A. Wilson, and R. J. Small. 2006. Long-term trends in harbor seal numbers at Tugidak Island and Nanvak Bay, Alaska. Mar. Mam. Sci. 22: 1–22. Karl, T. R. 1988. Multi-year fluctuations of temperature and precipitation: the gray areas of climate change. Clim. Change 12: 179–197. Kelly, K. A., M. J. Caruso, and J. A. Austin. 1993. Wind-forced variations in sea surface height in the Northeast Pacific Ocean, J. Phys. Oceanogr. 23: 2392–2411. Kenyon, K.W. 1969. The sea otter in the eastern Pacific Ocean. North American Fauna 68. 352 pp. Kihara, K., and A. M. Shimada. 1988. Prey-predator interactions of Pacific cod Gadus macrocephalus and water temperature. Nippon Suisan Gakkaishi, Japanese Soc. Sci. Fish, Bull. 54: 2085–2088. Kildaw, S.D., D.B. Irons, D.R. Nysewander, and C.L. Buck. 2005. Formation and growth of new seabird colonies. The singnificance of habitat quality. Mar. Ornith. 33: 49–58. Kitaysky, A. S., J. C. Wingfield, and J. F. Piatt. 1998. Dynamics of food availability, body condition and physiological stress response in breeding black-legged kittiwakes. Functional Ecol. 13: 577–584. Kitaysky, A. S., J. F. Piatt, J. C. Wingfield, and M. Romano. 1999. The adrenocortical stress-response of black-legged kittiwake chicks in relation to dietary restrictions. J. Comp. Physiol. (B): 303–310. Kitaysky, A., J. Wingfield, and J. Piatt. 2001. Corticosterone facilitates begging and affects resource allocation in the black-legged kittiwake. Behav. Ecol. 12: 619–625. Kitaysky, A.S., M.D. Romano, J.F. Piatt, J.C. Wingfield, M. Kikuchi. 2005. The adrenocortical response of tufted puffin chicks to nutritional deficits. Horm. Behav. 47: 606–619. Klosiewski, S. P. and K. K. Lang. 1994. Marine bird populations of Prince William Sound, Alaska, before and after the Exxon Valdez oil spill. U.S. Fish and Wildlife Service, Anchorage, AK. Koeller, P. A. 2000. Relative importance of abiotic and biotic factors to the management of the northern shrimp (Pandalus borealis) fishery on the Scotian Shelf. J. Northwest Atlantic Fish. Sci. 27: 21–33.
406
Long-Term Ecological Change in the Northern Gulf of Alaska
Koslow, A. J. 1984. Recruitment patterns in Northwest Atlantic fish stocks. Can. J. Fish. Aquat. Sci. 41: 1722–1729. Kramer, D. E. and V. M. O’Connell. 1995. Guide to northeast Pacific rockfishes. Genera Sebastes and Sebastolobus. Alaska Sea Grant Program, Marine Advisory Bulletin No. 25. Kruse, G. H. 1993. Biological perspectives on crab management in Alaska. In: G. H. Kruse, D. M. Eggers, R. J. Marasco, C. Pautzke, and T. J. Quinn II (Eds.), Proceedings of the International Symposium on Management Strategies for Exploited Fish Populations, University of Alaska Sea Grant College Program Report 93-02, Fairbanks, pp. 355–384. Kruse, G. H., and J. Zheng. 1999. Are changes in Bering Sea crab and groundfish populations related?. In: Ecosystem approaches for fisheries management. University of Alaska Sea Grant Report 99-01, Fairbanks, pp. 143–148. Kruse, G. H., F. C. Funk, and J. Zheng. 1996. Were Alaskan red king crabs overfished? In: Proceedings of the international symposium on biology, management, and economics of crabs from high latitude habitats. University of Alaska Fairbanks, Alaska Sea Grant Report 96-02, pp. 295–299. Kruse, G. H., F. C. Funk, H. J. Geiger, K. R. Mabry, H. M. Savikko, and S. M. Siddeek. 2000. Overview of state-managed marine fisheries in the central and western Gulf of Alaska, Aleutian Islands, and southeastern Bering Sea, with reference to Steller sea lions. Alaska Department of Fish and Game, Division of Commercial Fisheries, Regional Information Report 5J00-10, Juneau. Kuletz, K. J. 2005. Foraging behaviour and productivity of a non-colonial seabird, the marbled murrelet (Brachyramphus marmoratus) relative to prey and habitat. Ph.D. Dissertation. University of Victoria, Victoria, British Columbia. Kuletz, K. J., D. B. Irons, B. A. Agler, J. F. Piatt, and D. C. Duffy. 1997. Long-term changes in diets and populations of piscivorous birds and mammals in Prince William Sound, Alaska. In Forage fishes in marine ecosystems. Alaska Sea Grant College Program Fairbanks, AK, AK-SG-97–01, pp. 703–706. Kuletz, K. J., S. W. Stephensen, D. B. Irons, E. A. Labunski, and K. M. Brenneman. 2003. Changes in distribution and abundance of Kittlitz’s murrelets Brachyramphus brevirostris relative to glacial recession in Prince William Sound, Alaska. Mar. Ornith. 31: 133–140. Lagerloef, G. S. E. 1995. Interdecadal variations in the Alaska Gyre, J. Phys. Oceanogr. 25: 2242–2258. Lang, G. M., C. W. Derrah, and P. A. Livingston. 2003. Groundfish food habits and predation on commercially important prey species in the eastern Bering Sea from 1993 through 1996. National Marine Fisheries Service, AFSC Processed Report 2003–04, Seattle, WA. Lasker, R. 1985. What limits clupeiod production. Can. J. Fish. Aquat. Sci. 42: 31–37. Le Boeuf, B. J. and D. C. Crocker. 2005. Ocean climate and seal condition. BMC Biol. 3: 1–10. Leaman, B. M. 1991. Reproductive styles and life history variables relative to exploitation and management of Sebastes stocks. Environ. Biol. Fishes 30: 225–243. Lensink, C.J. 1960. Status and distribution of sea otters in Alaska. J. Mammol. 41:172–182. Lethcoe, J. and N. Lethcoe. 1994. A History of Prince William Sound, Alaska. Prince William Sound Books, Valdez, Alaska. 275 pp. Lewis, J. P., G. W. Pendleton, K. W. Pitcher, and K. M. Wynne. 1996. Harbor seal population trends in southeast Alaska and the Gulf of Alaska. Douglas, AK, Alaska Department of Fish and Game.
Long-Term Change 407
Lilly, G. R., D. G. Parsons, and D. W. Kulka. 2000. Was the increase in shrimp biomass on the Northeast Newfoundland Shelf a consequence of a release in predation pressure from cod? J. Northwest Atlantic Fish. Sci. 27: 45–61. Litzow, M. A. 2000. Food limitation in a generalist seabird: reproductive consequences of food quality and prey switching in the Pigeon Guillemot (Cepphus columba). M.Sc. Thesis, University of California, Santa Cruz, CA. 36 pp. Litzow, M. A. and J. F. Piatt. 2003. Variance in prey abundance influences time budgets of breeding seabirds: evidence from pigeon guillemots. J. Avian. Biol. 34: 54–64. Litzow, M. A., J. F. Piatt, A. A. Abookire, A. K. Prichard, and M. D. Robards. 2000. Monitoring temporal and spatial variability in sandeel (Ammodytes hexapterus) abundance with pigeon guillemot (Cepphus columba) diets. ICES J. Mar. Sci. 57: 976–986. Litzow, M. A., J. F. Piatt, A. A. Abookire, and M. Robards. 2004a. Energy density and variability in abundance of pigeon guillemot prey: support for the quality-variability tradeoff hypothesis. J. Animal Ecol. 73: 1149–1156. Litzow, M. A., J. F. Piatt, A. A. Abookire, S. G. Speckman, M. L. Arimitsu, and J. D. Figurski. 2004b. Spatiotemporal predictability of schooling and non-schooling prey of Pigeon Guillemots. Condor 106: 410–414. Livingston, P. (Ed.). 2003. Ecosystem Considerations for 2003. Reviewed by the Plan Team for the Groundfish Fisheries of the Bering Sea, Aleutian Islands, and Gulf of Alaska, North Pacific Fishery Management Council, 605 W. 4th Avenue, Suite 306, Anchorage, AK 99501. Livingston, P. A. 1989. Interannual trends in Pacific cod, Gadus macrocephalus, predation on three commercially important crab species in the eastern Bering Sea. Fish. Bull. 87: 807–827. Loughlin, T. R. and A. E. York. 2000. An account of the sources of Steller sea lion (Eumetopias jubatus) mortality. Mar. Fish. Rev. 62: 40–45. Lucas, Z. and W. T. Stobo. 2000. Shark-inflicted mortality on a population of harbor seals (Phoca vitulina) as Sable Island, Nova Scotia. J. Zool., London 252: 405–414. MacFarlane, G. A., J. R. King, and R. J. Beamish. 2000. Have there been recent changes in climate? Ask the fish. Prog. Oceanog. 47: 147–169. Mackas, D. L. and A. Tsuda. 1999. Mesozooplankton in the eastern and western subarctic Pacific: community structure, seasonal life histories, and interannual variability. Prog. Oceanogr. 43: 335–363. Mackas, D. L., R. Goldblatt, and A. G. Lewis. 1998. Interdecadal variation in developmental timing of Neocalanus plumchrus populations at Ocean Station P in the subarctic North Pacific. Can. J. Fish. Aquat. Sci. 55: 1878–1893. Mantua, N. J., S. R. Hare, Y. Zhang, J. M. Wallace, and R. C. Francis. 1997. A Pacific interdecadal climate oscillation with impacts on salmon production. Bull. Amer. Meteor. Soc. 78: 1069–1079. Martin, M. H. 1997. Data report: 1996 Gulf of Alaska bottom trawl survey. U.S. Dep. Commer., NOAA Tech. Memo. NMFS-AFSC-82, 235 pp. Marty, G. D., E. F. Freiberg, T. R. Meyers, J. Wilcock, T. B. Farver, and D. E. Hinton. 1998. Viral hemorrhagic septicemia virus, Ichthyophonus hoferi, and other causes of morbidity in Pacific herring (Clupea pallasi) in Prince William Sound, Alaska, U.S.A. Dis. Aquat. Org. 32: 15–40. Marty, G. D., T. J. Quinn II, G. Carpenter, T. R. Meyers, and N. H. Willits. 2003. Role of disease in abundance of a Pacific herring (Clupea pallasi) population. Can. J. Fish. Aquat. Sci. 60: 1258–1265.
408
Long-Term Ecological Change in the Northern Gulf of Alaska
Matarese, A. C., D. M. Blood, S. J. Picquelle, and J. L. Benson. 2003. Atlas of abundance and distribution patterns of ichthyoplankton from the northeast Pacific Ocean and Bering Sea ecosystems based on research conducted by the Alaska Fisheries Science Center (1972–1996). U.S. Department of Commerce, NOAA Professional Paper NMFS 1, 281 pp. Mathews, E. A. and G. W. Pendleton. 1997. Estimation of trends in abundance of harbor seals from at terrestrial and glacial haulouts in Glacier Bay National Park, Southeast Alaska. In Harbor seal investigations in Alaska. Anchorage, AK, Alaska Department of Fish and Game, pp. 57–76. Mathews, E. A. and G. W. Pendleton. 2000. Declining trends in harbor seal (Phoca vitulina richardsi) numbers at glacial ice and terrestrial haulouts in Glacier Bay National Park, 1992–1998. Juneau, AK, University of Alaska Southeast. 24 pp. Mathisen, O. A., R. T. Baade, and R.J. Lopp. 1962. Breeding habits, growth and stomach contents of the Steller sea lion in Alaska. J. Mammol. 43: 469–477. McMullen, J. C. 1968. Investigation of king crab ocean reproduction and brood stock composition, Kodiak Island. Alaska Department of Fish and Game, Informational Leaflet 126, Kodiak. McMullen, J. C., and H. T. Yoshihara. 1969. Fate of unfertilized eggs in king crabs, Paralithodes camtschatica (Tilesius). J. Fish. Res. Board Can. 26: 2737–2740. Megrey, B. A. 1989. Exploitation of walleye pollock resources in the Gulf of Alaska, 1964–1988: portrait of a fishery in transition. Proceedings of the International Symposium on the Biology and Management of Walleye Pollock, Lowell Wakefield Fisheries Symp., Alaska Sea Grant Rep. 89–1, 33–58. Merrick, R. I., D. G. Brown, D. C. Calkins, and T. R. Loughlin. 1995. A comparison of Steller sea lion, Eumetopias jubatus, pup masses between rookeries with increasing and decreasing populations. Fish. Bull. 93: 753–758. Merrick, R. L. and D. G. Calkins. 1996. Importance of juvenile walleye pollock, Theragra chalcogramma, in the diet of Gulf of Alaska Steller sea lions, Eumetopias jubatus. U.S. Dep. Comm. NOAA Tech. Rep., NMFS 126: 153–166. Merrick, R. L., M. K. Chumbley, and G. V. Byrd. 1997. Diet diversity of Steller sea lions (Eumetopias jubatus) and their population decline in Alaska: a potential relationship. Can. J. of Fish. Aquat. Sci. 54: 1342–1348. Merrick, R. L., T. R. Loughlin, and D. G. Calkins. 1987. Decline in abundance of the northern sea lion, Eumetopias jubatus, in Alaska, 1956–86. Fish. Bull. 85: 351–365. Milette, L. L. and A. W. Trites. 2003. Maternal attendance patterns of lactating Steller sea lions (Eumetopias jubatus) from a stable and a declining population in Alaska. Canadian J. Zool. 81: 340–348. Miller, C. B., J. Fulton, and B. W. Frost. 1992. Size variation of Neocalanus plumchrus and Neocalanus flemingeri in a 20-year sample series from the Gulf of Alaska. Can. J. Fish. Aquat. Sci. 49: 389–399. Minobe, S. 1997. A 50–70 year oscillation over the North Pacific and North America. Geophys. Res. Letters 24: 683–686. Monaghan, P., P. Walton, S. Wanless, J. D. Uttley, and M. D. Burns. 1994. Effects of prey abundance on the foraging behaviour, diving efficiency and time allocation of breeding Guillemots Uria aalge. Ibis 136: 214–222. Monson, D. H., J. A. Estes, J. L. Bodkin, and D. B. Siniff. 2000. Life history plasticity and population regulation in sea otters. Oikos 90: 457–468.
Long-Term Change 409
Mueter, F. J. and B. L. Norcross. 2000. Changes in species composition of the demersal fish community in nearshore waters of Kodiak Island, Alaska. Can. J. Fish. Aquat. Sci. 57: 1169–1180. Mueter, F. J., R. M. Peterman, and B. J. Pyper. 2002. Opposite effects of ocean temperature on survival rates of 120 stocks of Pacific salmon (Oncorhynchus spp.) in northern and southern areas. Can. J. Fish. Aquat. Sci., 59: 456–463. Murdoch, W. W. and A. Oaten. 1975. Predation and population stability. Adv. Ecol. Res. 9: 1–125. Murphy, E. C., M. I. Springer, D. G. Roseneau, and A. M. Springer. 1980. Monitoring population numbers and productivity of colonial seabirds. MMS/NOAA OCSEAP Annual Report 1: 142–272. Murray, C. P., S. L. Morey, and J. J. O’Brien. 2001. Interannual variability of upper ocean vorticity balances in the Gulf of Alaska. J. Geophys. Res. 106: 4479–4491. Nunes, P., and T. Nishiyama. 1984. Effects of temperature on the embryonic development of the northern pink shrimp Pandalus borealis Krøyer. J. Shellfish Res. 4: 96–97. Nysewander, D. R. and J. L. Trapp. 1984. Widespread mortality of adult seabirds in Alaska, August–September 1983, U.S. Fish and Wildlife Service, Anchorage, Alaska. Nysewander, D., C. H. Dipple, G. V. Byrd, and E. P. Knudtson. 1993. Effects of the T/V Exxon Valdez oil spill on murres: A perspective from observations at colonies. NRDA Bird Study No. 3 Final Rep., U.S. U.S. Fish and Wildlife Service, Homer AK. 40 pp. Oakley, K. L. and K. J. Kuletz. 1996. Population, reproduction, and foraging of pigeon guillemots at Naked Island, Alaska, before and after the Exxon Valdez oil spill. In: Rice, S. D., R. B. Spies, D. A. Wolfe, and B. A. Wright (Eds.), Proceedings of the Exxon Valdez Oil Spill Symposium American Fisheries Society Symposium 18. American Fisheries Society, Bethesda, Maryland, pp. 759–769. Olesiuk, P. F. 1999. An assessment of the status of harbor seals (Phoca vitulina) in British Columbia. Nanaimo, B.C., Canada, Department of Fisheries and Oceans, Research Document 99/33. Olesiuk, P. F. 2003. Recent trends in the abundance of Steller sea lions (Eumetopias jubatus) in British Columbia. Nanaimo, B.C., Canada, Department of Fisheries and Oceans, Working Paper 2003–11. Olesiuk, P. F., M. A. Bigg, and G. M. Ellis. 1990. Recent trends in the abundance of harbour seals, Phoca vitulina, in British Columbia. Can. J. Fish. Aquat. Sci. 47: 992–1003. Olsen, J. 1994. Development of the pink salmon enhancement program in Prince William Sound, Alaska. Proc. 16th Pink and Chum Salmon Workshop. Alaska Sea Grant College Program Rep. 94–02, Fairbanks, Alaska, pp. 57–68. Orensanz, J. M., J. Armstrong, D. Armstrong, and R. Hilborn. 1998. Crustacean resources are vulnerable to serial depletion – the multifaceted decline of crab and shrimp fisheries in the greater Gulf of Alaska. Rev. Fish . Biol. 8: 117–176. Ottesen, O. H. and S. Bolla. 1998. Combined effects of temperature and salinity on development and survival of Atlantic halibut larvae. Aquacult. Int. 6: 103–120. Paine, R. T., D. W. Bromley, M. A. Castellini, L. B. Crowder, J. A. Estes, J. M. Grebmeier, F. M. D. Gulland, G. H. Kruse, N. J. Mantua, Two Crow, D. B. Siniff, and C. J. Walters. 2003. Decline of the Steller sea lion in Alaskan waters. Washington, DC, The National Academies Press. 204 pp. Parslow, T. R. 1981. Phytoplankton-zooplankton interactions: Data analysis and modeling (with particular reference to Ocean Station P [50°N, 145°W] and controlled ecosystem experiments). Ph.D. Dissertation, University of British Columbia, 400 pp.
410
Long-Term Ecological Change in the Northern Gulf of Alaska
Parsons, T. R. 1988. Trophodynamic phasing in theoretical experimental and natural pelagic ecosystems. J. Oceanogr. Soc. Japan. 44: 94–101. Pascual, M. A. and M. D. Adkison. 1994. The decline of the Steller sea lion in the northeast Pacific: demography, harvest or environment? Ecol. Appl. 4: 393–403. Paul, A. J., and J. M. Paul. 1980. The effect of early starvation on later feeding success of king crab zoea. J. Exper. Mar. Biol. Ecol. 44: 247–251. Paul, A. J., J. M. Paul, and K. O. Coyle. 1990. Growth of stage I king crab larvae of Paralithodes camtschatica (Tilesius) (Decapoda: Lithodidae) in natural communities. J. Crustacean Biol. 10: 175–183. Philips, R. A., Caldow, R. W. G., and Furness, R. W. 1996. The influence of food availability on the breeding effort and reproductive success of Arctic Skuas Stercorarius parasiticus. Ibis 138: 410–419. Piatt, J. F (Ed.), 2002a. Response of seabirds to fluctuations in forage fish density. Final Report to Exxon Valdez Oil Spill Trustee Council (Restoration Project 00163M) and Minerals Management Service (Alaska OCS Report 2002-068). Alaska Science Center, U.S. Geological Survey, Anchorage, Alaska. 454 pp. Piatt, J. F. 1990. Aggregative response of Common Murres and Atlantic Puffins to their prey. Studies in Avian Biol. 14: 36–51. Piatt, J. F. 2002b. Introduction and Project Overview. In: J. F. Piatt (Ed.), Response of seabirds to fluctuations in forage fish density. Final Report to Exxon Valdez Oil Spill Trustee Council (Restoration Project 01163M) and Mineral Management Service (Alaska OCS Study MMS 2002-068). Alaska Science Center, U.S. Geological Survey, Anchorage, Alaska, pp. 1–17. Piatt, J. F. 2004. Survival of Adult Murres and Kittiwakes in Relation to Forage Fish Abundance. Exxon Valdez Oil Spill Restoration Project Final Report (Restoration Project 00338), U.S. Geological Survey, Anchorage, Alaska. Piatt, J. F. and A. S. Kitaysky 2002. Tufted puffin (Fratercula cirrhata). In: A. Pool and F. Gill (Eds.), The birds of North America, No. 708., The Birds of North America Inc., Philadelphia, PA. Piatt, J. F. and P. Anderson. 1996. Response of common murres to the Exxon Valdez oil spill and long-term changes in the Gulf of Alaska marine ecosystem, In: Rice, S. D., R. B. Spies, D. A. Wolfe, and B. A. Wright (Eds.), Proceedings of the Exxon Valdez Oil Spill Symposium. American Fisheries Society Symposium 18. American Fisheries Society, Bethesda, Maryland, pp. 720–737. Piatt, J. F. and R. G. Ford. 1993. Distribution and abundance of marbled murrelets in Alaska. Condor 95: 662–669. Piatt, J. F. and R. G. Ford. 1996. How many seabirds were killed by the Exxon Valdez oil spill? In: Rice, S. D., R. B. Spies, D. A. Wolfe, and B. A. Wright (Eds.), Exxon Valdez Oil Spill Symposium Proceedings. American Fisheries Society Symposium 18, Bethesda, Maryland, pp. 720–737. Piatt, J. F. and T. Van Pelt. 1993. A wreck of common murres (Uria aalge) in the northern Gulf of Alaska during February and March of 1993. U.S. Fish and Wildlife Service, Anchorage, AK. Piatt, J. F. and T. Van Pelt. 1997. Mass-mortality of guillemots (Uria aalge) in the Gulf of Alaska in 1993. Mar. Poll. Bull. 34: 656–662. Piatt, J. F., C. J. Lensink, W. Butler, M. Kendziorek, and D. Nysewander. 1990. Immediate impact of the Exxon Valdez oil spill on marine birds. Auk 107: 387–397.
Long-Term Change 411
Piatt, J. F., G. Drew, T. I. Van Pelt, A. Abookire, A. Nielsen, M. Shultz, and A. Kitaysky. 1999. Biological effects of the 1997/1998 ENSO event in lower Cook Inlet, Alaska. PICES Scientific Report No. 10: 93–100. Pitcher, K. 1980. Food of the harbor seal, Phoca vitulina richardsi, in the Gulf of Alaska. Fish. Bull. 78: 544–549. Pitcher, K. W. 1981. Prey of the Steller sea lion, Eumetopias jubatus, in the Gulf of Alaska. Fish. Bull. 79: 467–472. Pitcher, K. W. 1990. Major decline in number of harbor seals, Phoca vitulina richardsi, on Tugidak Island, Gulf of Alaska. Mar. Mam. Sci. 6: 121–134. Pitcher, K. W., D. G. Calkins, and G. W. Pendleton. 1998. Reproductive performances of female Steller sea lions from the Gulf of Alaska: an energetics-based reproductive strategy? Can. J. Zool. 76: 2075–2083. Pitcher, K. W., D. G. Calkins, and G. W. Pendleton. 2000. Steller sea lion body condition indices. Mar. Mam. Sci. 16: 427–436. Planque, B., and T. Frédou. 1999. Temperature and the recruitment of Atlantic cod (Gadus morhua). Can. J. Fish. Aquat. Sci 56: 2069–2077. Polovina, J. J., G. T. Mitchum, and G. T. Evans. 1995. Decadal and basin-scale variations in mixed layer depth and the impact on biological production in the Central and North Pacific, 1960–88. Deep-Sea Res. 42: 1701–1716. Porter, R. W., B. J. Kouri, and G. Kudo. 1993. Inhibition of protease activity in muscle extracts and surimi from Pacific whiting, Merluccius productus, and arrowtooth flounder, Atherestes stomias, off Washington. Fish. Bull. 93: 127–138. Poulin, R., and K. Rohde. 1997. Comparing the richness of metazoan ectoparasite communities of marine fishes: Controlling for host phylogeny. Oecologia. 110: 278–283. Powell, G. C., B. Shafford, and M. Jones. 1973. Reproductive biology of young adult king crabs Paralithodes camtschatica (Tilesius) at Kodiak, Alaska. Proc. Natl. Shellfish Assoc. 73: 77–87. Rausch, R. 1958. The occurrence and distribution of birds on Middleton Island, Alaska. Condor 60: 227–242. Rea, L. D., K. W. Pitcher, S. D. Farley, J. P. Richmond, and W. S. Dunlap-Harding. 2003. Percent total body lipid content increases in Steller sea lion (Eummetopias jubatus) pups throughout the first year of life in a similar pattern to other otariids pups. 15th Biennial Conference on the Biology of Marine Mammals, Greensboro, NC. Rea, L. D., M. Castellini, B. S. Fadely, and T. R. Loughlin. 1998. Health status of young Alaska Steller sea lion pups (Eumetopias jubatus) as indicated by blood chemistry and hematology. Comp. Biochem. Physiol. A 120: 617–623. Reid, G. M. 1971. Age composition, weight, length and sex of herring, Clupea Pallasii, used for reduction in Alaska, 1929–66. U.S. Department of Commerce, National Oceanic and Atmospheric Administration, National Marine Fisheries Service, Tech. Rpt. NMFS SSRF 634. Seattle, Washington. Robards, M. D. 2000. Ecology and demographics of Pacific sand lance, Ammodytes hexapterus Pallas, in lower Cook Inlet, Alaska. M.Sc. Thesis, Memorial University of Newfoundland, St. John’s, Canada. 105 pp. Robards, M. D., G. A. Rose, and J. F. Piatt. 2002. Growth and abundance of Pacific sand lance, Ammodytes hexapterus, under differing oceanographic regimes. Environ. Biol. Fish. 64: 429–441.
412
Long-Term Ecological Change in the Northern Gulf of Alaska
Robards, M. D., J. Anthony, J. F. Piatt, and G. Rose. 1999c. Changes in proximate composition and somatic energy content for Pacific sand lance (Ammodytes hexapterus) relative to maturity and season in Kachemak Bay, Alaska. J. Exper. Mar. Biol. Ecol. 242: 245–258. Robards, M. D., J. F. Piatt, A. B. Kettle, and A. A. Abookire. 1999b. Temporal and geographic variation in fish communities of lower Cook Inlet, Alaska. Fish. Bull. 97: 962–977. Robards, M. D., J. F. Piatt, and G. A. Rose. 1999a. Maturation, fecundity and intertidal spawning of Pacific Sand Lance (Ammodytes hexapterus) in the northern Gulf of Alaska. J. Fish Biol. 54: 1050–1068. Robards, M., G. Drew, J. Piatt, J. M. Anson, A. Abookire, J. Bodkin, P. Hooge, and S. Speckman. 2003. Ecology of selected marine communities in Glacier Bay: zooplankton, forage fish, seabirds and marine mammals. U.S.G.S, Alaska Science Center, Anchorage, AK. Rogers, D. E. 1984. Trends in abundance of northeastern Pacific stocks of salmon. In: W. G. Pearcy (Ed.), The Influence of Ocean Conditions on the Production of Salmonids in the north Pacific. Publ. ORESU-W-83-001, Oregon State University Sea Grant College Program, Corvallis, Oreg. pp. 100–127. Rogers, D. E. 1986. Pacific salmon, In: D. W. Hood and S. T. Zimmerman (Eds.), The Gulf of Alaska. Physical Environment and Biological Resources. NOAA, U.S. Department of Commerce, Springfield, Va. pp. 461–478. Romano, M. D. 2000. Effects of diet on growth and development of nestling seabirds. M.Sc. Thesis, Oregon State University, Corvallis OR. 59 pp. Rosen, D. A. S. and A. W. Trites. 2000. Pollock and the decline of Steller sea lions: testing the junk food hypothesis. Can. J. Zool. 78: 1243–1250. Roseneau, D. G., A. B. Kettle, and G. V. Byrd. 1997. Common murre population monitoring at the Barren Islands, Alaska, 1996. Exxon Valdez Oil Spill Restoration Ann. Rep. (Restoration Project 96144), U.S. Fish and Wildlife Serv., Alaska Maritime NWR, Homer, Alaska. Roseneau, D. G., A. B. Kettle, and G. V. Byrd. 2000. Barren Islands seabird studies, 2000, Exxon Valdez Oil Spill Restoration Project Final Report (Reostoration Project 00163J), U.S. Fish and Wildlife Service, Alaska Maritime National Wildlife Service, Homer, Alaska. Rosenkranz, G. E. 1998. Statistical modeling of Tanner crab recruitment in Bristol Bay, Alaska. Master’s thesis, University of Alaska Fairbanks, 119 pp. Rosenkranz, G. E., A. V. Tyler, and G. H. Kruse. 2001. Effects of water temperature and wind on recruitment of Tanner crabs in Bristol Bay, Alaska. Fish. Oceanogr. 10: 1–12. Rosenkranz, G. E., A. V. Tyler, G. H. Kruse, and H. J. Niebauer. 1998. Relationship between wind and year class strength of Tanner crabs in the Southeastern Bering Sea. Alaska Fish. Res. Bull. 5: 18–24. Rounsefell, G. A and E. H. Dahlgren. 1932. Fluctuations in the supply of herring (Clupea pallasii) in Prince William Sound, Alaska. Bull. Bureau of Fisheries Vol. XLVII, Bulletin No. 9, U.S. Dept. Commerce Bureau of Fisheries Doc. No. 9, Washington, DC. Rounsefell, G. A. 1929. Contribution to the biology of the Pacific herring (Clupea pallasii), and the condition of the fishery in Alaska. Bulletin of the Bureau of Fisheries Vol. XLV, U.S. Dept. Commerce, Bureau of Fisheries Doc. No. 1080, Washington DC. Royer, T. C. 1982. Coastal freshwater discharge in the Northeast Pacific. J. Geophys. Res. 87: 2017–2021. Royer, T. C. 2005. Hydrographic responses of the coastal Gulf of Alaska to seasonal and interannual forcing. Deep-Sea Res. Part II. 52: 267–288.
Long-Term Change 413
Sagar, P. and D. S. Horning. 1998. Mass related survival of fledging sooty shearwaters Puffinus griseus at the Snares, New Zealand. Ibis 140: 329–339. Schaller, G. B., H. Jinchu, P. Wenshi, and Z. Jing. 1985. The Giant Pandas of Wolong. University of Chicago Press, Chicago, IL. Scheel, D. and K. R. Hough. 1997. Salmon fry predation by seabirds near an Alaskan hatchery. Mar. Ecol. Prog. Ser. 150: 35–48. Shimada, A. M. and D. K. Kimura. 1994. Seasonal movements of Pacific cod (Gadus macrocephalus) in the eastern Bering Sea and adjacent waters based on tag-recapture data. Fish. Bull. 92: 800–816. Shirley, S. M., and T. C. Shirley. 1989. Interannual variability in density, timing and survival of Alaskan red king crab Paralithodes camtschatica larvae. Mar. Ecol. Prog. Ser. 54: 51–59. Shirley, T. C., and S. M. Shirley. 1990. Planktonic survival of red king crab larvae in a subarctic ecosystem, 1985–1989, In: D. A. Ziemann and K. W. Fulton-Bennett (Eds.), APPRISE – Interannual Variability and Fisheries Recruitment. The Oceanic Institute, Honolulu, HI, pp. 263–285. Shirley, T. C., S. M. Shirley, and S. Korn. 1990. Incubation period, molting and growth of female red king crabs: effects of temperature, In: Proceedings of the international symposium on king and Tanner crabs. Alaska Sea Grant Program Report 90-04, University of Alaska, Fairbanks, pp. 51–63. Shultz, M., Harding, A. M. A., and A. B. Kettle. 2002. Black-legged Kittiwake Biology in lower Cook Inlet. In: J. F. Piatt (Ed.), Response of seabirds to fluctuations in forage fish density. Final Report to Exxon Valdez Oil Spill Trustee Council (Restoration Project 01163M) and Mineral Management Service (Alaska OCS Study MMS 2002-068). Alaska Science Center, U.S. Geological Survey, Anchorage, Alaska, pp. 87–100. Simenstad, C. A., J. A. Estes, and K. W. Kenyon. 1978. Aleuts, sea otters, and alternate stable state communities. Science 200: 403–411. Sinclair, E. H. and T. K. Zeppelin. 2002. Seasonal and spatial differences in diet in the western stock of Steller sea lions (Eumetopias jubatus). J. Mammol. 83: 973–990. Sinclair, M. 1988. Marine Populations: An Essay on Population Regulation and Speciation, 2nd edition. Washington Sea Grant Program, University of Washington Press, Seattle. Skud, B. E., H. M. Sakuda, and G. M. Reid. 1960. Statistics of the Alaska herring fishery 1878–1956. U.S. Department of the Interior, Fish Wildlife Service, Statistical Digest 48. Small, R. J., G. W. Pendleton, and K. W. Pitcher. 2003. Trends in abundance of Alaska harbor seals, 1983–2001. Mar. Mam. Sci. 19: 344–362. Smith, S. L., and J. Vidal. 1984. Spatial and temporal effects of salinity, temperature, and chlorophyll on the communities of zooplankton in the southeastern Bering Sea. J. Mar. Res. 42: 221–257. Speckman, S. G. 2004. Characterizing fish schools in relation to the marine environment and their use by seabirds in Lower Cook Inlet, Alaska. PhD thesis, University of Washington, Seattle. Speckman, S. G., J. F. Piatt, C. V. Minte-Vera, and J. K. Parrish. 2005. Parallel structure among environmental gradients and three trophic levels in a subarctic estuary. Progr. Oceanogr. 66: 25–65. Springer, A. M. 1998. Is it all climate change? Why marine bird and mammal populations fluctuate in the North Pacific. In: G. Holloway, P. Muller and D. Henderson (Eds.), Biotic Impacts of Extratropical Climate Variability in the Pacific. Proceedings ‘Aha Huliko’a Hawaiian Winter Workshop. Honolulu, HI, University of Hawaii, pp. 109–119.
414
Long-Term Ecological Change in the Northern Gulf of Alaska
Springer, A. M., D. G. Roseneau, E. C. Murphy, and M. I. Springer. 1984. Environmental controls of marine food webs: food habits of seabirds in the eastern Chukchi Sea. Can. J. Fish. Aquat. Sci. 41: 1202–1215. Springer, A. M., E. C. Murphy, D. G. Roseneau, C. P. McRoy, and B. A. Cooper. 1987. The paradox of pelagic food webs in the northern Bering Sea – I. Seabird food habits. Cont. Shelf Res. 7: 895–911. Springer, A. M., J. A. Estes, G. B. Van Vliet, T. M. Williams, D. F. Doak, E. M. Danner, K. A. Forney, and B. Pfister. 2003. Sequential megafaunal collapse in the North Pacific Ocean: An ongoing legacy of industrial whaling? Proc. Natl. Acad. Sci. 100: 12223–12228. Stabeno, P. J., N. A. Bond, A. J. Hermann, N. B. Kachel, C. W. Mordy, and J. E. Overland. 2004. Meteorology and oceanography of the northern Gulf of Alaska. Conti. Shelf Res. 24: 859–898. Stepanenko, M. A. 1997. Variations from year to year in the spatial differentiation of the walleye pollock, Theragra chalcogramma, and the cod, Gadus macrocephalus, in the Bering Sea. J. Ichthyol. 37: 14–20. Stephensen, S. W., D. B. Irons, S. J. Kendall, B. K. Lance, and L. L. Mcdonald. 2001. Marine bird and sea otter population abundance of Prince William Sound, Alaska: trends following the T/V Exxon Valdez oil spill, 1989–2000. U.S. Fish and Wildlife Service, Anchorage, AK. Stickney, A. P., and H. C. Perkins. 1981. Observations on the food of the larvae of northern shrimp, Pandalus borealis Krøyer (Decapoda, Caridea). Crustaceana 40: 36–49. Stirling, I. and A. E. Derocher. 1990. Factors affecting the evolution of the modern bears. International Conference on Bear Research and Management 8: 189–205. Strub, P. T. and C. James. 2002. Altimeter-derived surface circulation in the large-scale northeast Pacific gyres, Part II: Seasonal variability. Progr. Oceanogr. 54: 185–214. Suddaby, D. and N. Ratcliffe. 1997. The effects of fluctuating food availability on breeding Arctic terns (Sterna paradisaea). Auk 114: 524–530. Sugimoto, T. and K. Tadokoro. 1997. Interannual-interdecadal variations in zooplankton biomass, chlorophyll concentration and physical environment in the subarctic Pacific and Bering Sea. Fish. Oceangr. 6: 74–93. Sugimoto, T. and K. Tadokoro. 1998. Interdecadal variations in plankton biomass and physical environment in the North Pacific. Fish. Oceanogr. 7: 289–299. Suryan, R. M. and D. B. Irons. 2001. Colony and population dynamics of black-legged kittiwakes in a heterogeneous environment. Auk 118: 636–649. Suryan, R. M., D. B. Irons, J. Benson, and M. Kaufman. 2000. Foraging ecology, reproductive biology, and population dynamics of black-legged kittiwakes (Rissa tridactyla) in Prince William Sound, Alaska. Appendix E. In: APEX project: Alaska Predator Ecosystem Experiment in Prince William Sound and the Gulf of Alaska. B. A. Wright and D. C. Duffy (compilers). Exxon Valdez Oil Spill Restoration Project Final Report, (Project 00163E), Exxon Valdez Oil Spill Trustee Council, Anchorage, Alaska. Suryan, R. M., D. B. Irons, E. D. Brown, P. G. R. Jodice, and D. D. Roby 2006. Sitespecific effects on productivity of an upper trophic-level marine predator: Bottom-up, top-down, and mismatch effects on reproduction in a colonial seabird. Progr. Oceanogr. 68: 303–328. Suryan, R. M., D. B. Irons, M. Kaufman, J. Benson, P. G. R. Jodice, D. D. Roby, and E. D. Brown. 2002. Short-term fluctuations in forage fish availability and the effect on prey selection and brood-rearing in the black-legged kittiwake Rissa tridactyla. Mar. Ecol. Progr. Ser. 236: 273–287.
Long-Term Change 415
Swartzman, G. L., C. A. Ribic, and C. P. Haung 1990. Simulating the role of entanglement in northern fur seal, Callorhinus ursinus, population dynamics. In: R. S. Shomura and M. L. Godfrey (Eds.), Proceedings of the Second International Conference on Marne Debris. Honolulu, HI, U.S. Department of Commerce. NOAA Technical Memorandum NMFS-SWFSC-154. Sydeman, W. J., M. M. Hester, F. Gress, P. Martin, and J. Buffa. 2001. Climate change, reproductive performance, and diet composition of marine birds in the southern California Current system, 1969–1997. Progr. Oceanogr. 49: 309–329. Tabata, S. 1991. Annual and interannual variability of baroclinic transports across Line P in the northeast Pacific Ocean, Deep-Sea Res. 38(Suppl. 1): S221–S245. Thompson, G. G., H. H. Zenger, and M. W. Dorn. 2003. Chapter 2: Assessment of the Pacific cod stock in the Gulf of Alaska. In: Appendix B, Stock assessment and fishery evaluation report for the groundfish resources of the Gulf of Alaska. North Pacific Fishery Management Council. Anchorage, Alaska, pp. 149–241. Thorsteinson, F. V. and C. J. Lensink. 1962. Biological observations of Steller sea lions taken during an experimental harvest. J. Wildl. Manage. 26: 353–359. Thurston, R. V. and P. C. Gehrke, 1993. Respiratory oxygen requirements of fishes: description of OXYREF, a data file based on test results reported in the published literature, In: R.C. Russo and R.V. Thurston (Eds.), Fish Physiology, Toxicology, and Water Quality Management. Proceedings of an International Symposium, Sacramento, California, U.S.A., September 18–19, 1990. U.S. Environmental Protection Agency EPA/600/R-93/157, pp. 95–108. Towell, R. 2004. 2004 northern fur seal pup production on the Pribilof Islands, Alaska. Seattle, National Marine Mammal Laboratory. Trites, A. W. and C. P. Donnelly. 2003. The decline of Steller sea lions in Alaska: a review of the nutritional stress hypothesis. Mammal. Rev. 33: 3–28. Trites, A. W. and P. A. Larkin. 1989. The decline and fall of the Pribilof fur seal (Callorhinus ursinus): A simulation study. Can. J. Fish. Aquat. Sci. 46: 1437–1445. Tyler, A. V., and G. H. Kruse. 1996a. Conceptual modeling of brood strength of red king crabs in the Bristol Bay region of the Bering Sea. In: Proceedings of the International Symposium on Biology, Management and Economics of Crabs from High Latitude Habitat. University of Alaska Fairbanks, Alaska Sea Grant Program Report 96-02, pp. 511–543. Tyler, A. V., and G. H. Kruse. 1996b. Process modeling and red king crab year-class strength in the Bristol Bay region of the Bering Sea. PICES Press 4: 2–7. Tyler, A. V., and G. H. Kruse. 1997. Modeling workshop on year-class strength of Tanner crab, Chionoecetes bairdi. Alaska Department of Fish and Game, Commercial Fisheries Management and Development Division, Regional Information Report 5J97-02, Juneau. Tyler, A. V., and G. H. Kruse. 1998. A comparison of year-class variability of red king crabs and Tanner crabs in the eastern Bering Sea. Memoirs of the Faculty of Fisheries, Hokkaido University, 45: 90–95, Hakodate, Japan. U.S.F.W.S. 2004. Seabird colony catalog. Available from U.S. Fish and Wildlife Service, 1011 E. Tudor Road, Anchorage, AK. Uttley, J. D., P. Walton, P. Monaghan, and G. Austin. 1994. The effects of food abundance on breeding performance and adult time budgets of guillemots Uria aalge. Ibis 136: 205–213. Van Pelt, T. I. 2000. Reproductive costs and their expression in the common guillemot Uria aalge. MSc. Thesis, University of Glasgow, Scotland. 119 pp.
416
Long-Term Ecological Change in the Northern Gulf of Alaska
Van Pelt, T. I. and J. F. Piatt. 2003. Population status of Kittlitz’s and marbled murrelets, and surveys for other marine bird and mammal species in the Kenai Fjords area, Alaska. U.S.G.S, Alaska Science Center, Anchorage, AK. Van Pelt, T. I., M. Shultz, and A. Kettle. 2002. Common murre Biology in lower Cook Inlet. In: J. F. Piatt (Ed.), Response of seabirds to fluctuations in forage fish density. Final Report to Exxon Valdez Oil Spill Trustee Council (Restoration Project 01163M) and Mineral Management Service (Alaska OCS Study MMS 2002-068). Alaska Science Center, U.S. Geological Survey, Anchorage, Alaska, pp. 72–86. Van Scoy, K. and E. R. M. Druffel. 1993. Ventilation and transport of thermocline and intermediate waters in the Northeast Pacific during recent El Niños. J. Geophys. Res. 98: 18083–18088. Ver Hoef, J. M. and K. Frost. 2003. A Bayesian hierarchical model for monitoring harbor seal changes in Prince William Sound, Alaska. Environ. Ecol. Statis. 10: 201–209. Wang, J., M. Jin, E. V. Patrick, J. R. Allen, D. L. Eslinger, C. N. K. Mooers, and R. T. Cooney. 2001. Numerical simulations of the seasonal circulation patterns and thermohaline structures of Prince William Sound, Alaska. Fish. Oceanogr. 10(Suppl. 1): 132–148. Ware, D. M. 1992. Climate, predators and prey: behaviour of a linked oscillating system. In: T. Kawasaki, S. Tanaka, T. Yoahiaki and A. Taniguchi (Eds.), Long-term variability of pelagic fish populations and their environment, Pergamon Press, Sendai, Japan, pp. 279–291. Ware, D. M. and G. A. McFarlane. 1989. Fisheries production domains in the Northeast Pacific Ocean. Can. J. Fish. Aquat. Sci. 108: 359–379. Ware, D. M. and R. W. Tanasichuk. 1989. Biological basis of maturation and spawning waves in Pacific herring (Clupea harengus pallasi). Can. J. Fish. Aquat. Sci. 46: 1776–1784. Weingartner, T. J., K. Coyle, B. Finney, R. Hopcroft, T. Whitledge, R. Brodeur, M. Dagg, E. Farley, D. Haidvogel, L. Haldorson, A. Hermann, S. Hinckley, J. Napp, P. Stabeno, T. Kline, C. Lee, E. Lessard, T. Royer and S. Strom. 2002. The Northeast Pacific GLOBEC Program: Coastal Gulf of Alaska. Oceanography 15: 48–63. Weingartner, T. J., S. Danielson, and T. Royer. 2005. Freshwater Variability and Predictability in the Alaska Coastal Current. Deep-Sea Res. II 52: 169–192. Welch, C. A., J. Keay, K. Kendall, and C. T. Robbins. 1997. Constraints on frugivory by bears. Ecology 78: 1105–1119. Weller, D. W. 2002. Predation on Marine Mammals. In: W. F. Perrin, B. Würsig, and H. Thewissen (Eds.), Encyclopedia of Marine Mammals. Academic Press, San Diego, CA, pp. 985–994. Wertheimer, A. C., W. W. Smoker, T. L. Joyce, and W. R. Heard. 2001. Comment: a review of the hatchery programs for pink salmon in Prince William Sound and Kodiak Island, Alaska. Trans. Amer. Fish. Soc. 130: 712–720. Wespestad, V. G. 1991. Pacific herring population dynamics, early life history, and recruitment variation relative to eastern Bering Sea oceanographic factors. Doctoral dissertation, University of Washington, Seattle. Whitney, F. A. and H. J. Freeland. 1999. Variability in upper-ocean water properties in the Northeast Pacific. Deep-Sea Res. Part II, 46: 2351–2370. Whitney, F. A., Wong, C. S., and P. W. Boyd. 1998. Interannual variability in nitrate supply to surface waters of the Northeast Pacific Ocean. Mar. Ecol. Progr. Ser. 170: 15–23.
Long-Term Change 417
Wickett, W. P. 1966. Ekman transport and zooplankton concentration in the North Pacific Ocean. J. Fish. Res. Bd. Canada 24: 581–593. Wickett, W. P. 1967. Ekman transport and zooplankton concentration in the North Pacific Ocean. J. Fish. Res. Bd. Canada 24: 581–594. Willette, T. M., G. Carpenter, K. Hyer, and J. Wilcock. 1998. Herring Natal Habitats. Exxon Valdez Oil Spill Trustee Council Final Report for Restoration Project 97166. Anchorage, Alaska. Willette, T. M., R. T. Cooney, and K. Hyer. 1999. Predator foraging mode shifts affecting mortality of juvenile fishes during a subarctic spring bloom. Can. J. Fish. Aquat. Sci. 56: 364–376. Willette, T. M., R. T. Cooney, V. Patrick, D. M. Mason, G. L. Thomas, and D. Scheel. 2001. Ecological processes influencing the mortality of juvenile pink salmon (Ocorhynchus gorbuscha) in Prince William Sound, Alaska. Fish. Oceanogr. 10(Suppl. 1): 14–41. Williams, T. M., J. A. Estes, D.F. Doak, and A.M. Springer. 2004. Killer appetites: assessing the role of predators in ecological communities. Ecology 85: 3373–3384. Winship, A. J. and A. W. Trites. 2003. Prey consumption of Steller sea lions (Eumetopias jubatus) off Alaska: how much prey do they require? Fish. Bull. 101: 147–163. Winters, G. H. and J. P. Wheeler. 1996. Environmental and phenotypic factors affecting the reproductive cycle of Atlantic herring. ICES J. Mar. Sci. 53: 73–88. Withrow, D. E. and T. R. Loughlin. 1996. Abundance and distribution of harbor seals (Phoca vitulina richardsi) along the north side of the Alaska Peninsula and Bristol Bay during 1995. Annual Report to the MMPA Assessment Program, Office of Protected Resources, NMFS, NOAA, Silver Springs, MD. Worm, B., and R. A. Myers. 2003. Meta-analysis of cod-shrimp interactions reveals topdown control in oceanic food webs. Ecology 84: 162–173. Wu, J. Q. and Hsieh, W. W. 1999. A modeling study of the 1976 climate regime shift in the North Pacific Ocean. Can. J. Fish. Aquat. Sci. 56: 2450–2462. Wunsch, C. 1999. The interpretation of short climate records, with comments on the North Atlantic and Southern Oscillations. Bull. Amer. Meteorol. Soc. 80: 245–255. York, A. E. 1994. The population dynamics of Northern sea lions. Mar. Mam. Sci. 10: 38–51. York, A. E. and J. R. Hartley. 1981. Pup production following harvest of female northern fur seals. Can. J. Fish. Aquat. Sci. 38: 84–90. Zador, S. and J. F. Piatt. 1999. Time-budgets of common murres at a declining and increasing colony in Alaska. Condor 101: 149-152. Zador, S., A. Harding, J. F. Piatt, L. Ochikubo, and A. Nielsen. 1997. Monitoring Seabird Populations in Areas of Oil and Gas Development on the Alaskan Continental Shelf: Monitoring Populations And Productivity Of Seabirds At Colonies In Lower Cook Inlet, Alaska, 1995. OCS Study, MMS 95-013. Zebdi, A. and J. S. Collie. 1995. Effect of climate on herring (Clupea pallasi) population dynamics in the northeast Pacific Ocean. Can. J. Fish. Aquat. Sci. 121: 277–290. Zheng, J. M. C. Murphy, and G. H. Kruse. 1995b. An update of the length-based population model and stock-recruitment relationships for red king crab, Paralithodes camtschaticus, in Bristol Bay, Alaska. Alaska Fish. Res. Bull. 2: 114–124. Zheng, J., and G. H. Kruse. 1998. Stock-recruitment relationships for Bristol Bay Tanner crab. Alaska Fish. Res. Bull. 5: 116–130.
418
Long-Term Ecological Change in the Northern Gulf of Alaska
Zheng, J., and G. H. Kruse. 2002. Assessment and management of crab stocks under uncertainty of massive die-offs and rapid changes in survey catchability. In: Paul, A. J., E. G. Dawe, R. Elner, G. S. Jamieson, G. H. Kruse, R. S. Otto, B. Sainte-Marie, T.C. Shirley, and D. Woodby (Eds.), Crabs in cold water regions: biology, management, and economics. University of Alaska Sea Grant, AK-SG-02-01, Fairbanks, pp. 367–384. Zheng, J., and G. H. Kruse. 2003. Stock-recruitment relationships for three major Alaskan crab stocks. Fish. Res. 65: 103–121. Zheng, J., and G. H. Kruse. 2006. Recruitment variation of eastern Bering Sea crabs: climate forcing or top-down effects? Progr. Oceanogr. in press. Zheng, J., M. C. Murphy, and G. H. Kruse. 1995a. A length-based population model and stock-recruitment relationships for red king crab, Paralithodes camtschaticus, in Bristol Bay, Alaska. Can. J. Fish. Aquat. Sci. 52: 1229–1246. Zheng. J., and G. H. Kruse. 2000. Recruitment patterns of Alaskan crabs and relationships to decadal shifts in climate and physical oceanography. ICES J. Mar. Sci. 57: 438–451. Ziemann, D. A., L. D. Conquest, K. W. Fulton-Bennett, and P. K. Bienfang. 1990. In: D. A. Ziemann and K. W. Fulton-Bennett (Eds.), APPRISE – Interannual variability and fisheries recruitment. The Oceanic Institute, Honolulu, HI, pp. 129–170. Zimmerman, M. 1997. Maturity and fecundity of arrowtooth flounder, Atherestes stomias, from the Gulf of Alaska. Fish. Bull. 95: 598–611. Zimmerman, M., and P. Goddard. 1996. Biology and distribution of arrowtooth, Atherestes stomias, and Kamchatka, A. evermanni, flounders in Alaskan waters. Fish. Bull. 94: 358–370. Zimmermann, M., R. C. Harrison, and A. F. Jones. 2001. Differential parasitism by Naobranchia occidentalis (Copepoda: Naobranchiidae) and Nectobrachia indivisa (Copepoda: Lernaeopodidae) on northern rock sole (Lepidopsetta polyxystra Orr and Matarese, 2000) and southern rock sole (L. bilineata Ayres, 1855) in Alaskan waters. Fish. Bull. 99: 371–380.
Mini-Glossary: By-catch: The portion of a fishing catch that is discarded as unwanted or commercially unusable. Demersal: Dwelling at or near the bottom of a body of water. Depot lipids: Fat that is being stored within a body as an energy reserve. Ichthyoplankton: The newly hatched, larval stage of fish that may undergo metamorphosis, differing markedly in form and appearance from the adult.
Chapter 5
The Exxon Valdez Oil Spill Stanley D. Rice, Jeffrey W. Short, Mark G. Carls, Adam Moles and Robert B. Spies
5.1. Introduction The tanker vessel Exxon Valdez ran hard aground on Bligh Reef in Prince William Sound (PWS) on March 24, 1989, spilling 42 million liters of Alaska North Slope crude oil into the northern Gulf of Alaska (Fig. 5.1). Winds and currents carried floating oil hundreds of kilometers from Bligh Reef, inundating much of western PWS, parts of the outer Kenai Coast and Lower Cook Inlet, some beaches on Kodiak Island, and portions of the Alaska Peninsula as it was carried through Shelikof Strait and west into the Gulf of Alaska (Fig. 5.2). The oil eventually washed ashore discontinuously on over 2000 kilometers of shoreline, the equivalent of a spill in New York Harbor reaching South Carolina. This was the largest marine oil spill in the United States and the largest spill in a sub Arctic ecosystem; however, the damage from the Exxon Valdez oil spill (EVOS) was due as much to when and where it happened as it was to the size of the spill. Although there were more than 50 previous marine spills that had been larger elsewhere in the world, the EVOS was more devastating than many larger spills because it affected so much wildlife. As the spilled oil spread, the spring plankton bloom and the reproductive and growth seasons for many animals were about to start. The oil inundated seabird, sea otter, and harbor seal habitat just prior to their breeding seasons and that of many other vulnerable species as well. The oil contaminated the mouths of numerous streams where pink and chum salmon were about to spawn and the nearshore environment where fry from the previous year’s spawning were to feed for their first several months of marine life. The acute effects of the spill were immediate and obvious, but there was surprising oil persistence and attendant lingering effects. In the first few weeks, the acute effects on wildlife were dramatic, with hundreds of dead sea otters and tens of thousands of bird carcasses washed ashore, deformed herring larvae developing on beaches, and intertidal communities coated with oil. Despite the massive cleanup in 1989–1991 Long-Term Ecological Change in the Northern Gulf of Alaska Robert B. Spies (Editor) © 2007 Elsevier B.V. All rights reserved.
420
Long-Term Ecological Change in the Northern Gulf of Alaska
Figure 5.1: Oil continued to leak from the Exxon Valdez as it was transferred to the Exxon Baton Rouge (photograph, courtesy of the Exxon Valdez Trustee Council).
and natural weathering, oil stranded on many beaches was surprisingly persistent. This lingering oil has lead to unexpected chronic effects that had not been documented in previous spills. The fate and effects of the spill for the first several years had been summarized and reported in two symposium volumes (Rice et al., 1996; Wells et al., 1995), so we focus here on the longterm, up to 15 years after the spill. The studies of the EVOS provide us with insights into the effects of a major oil spill, and more generally of oil pollution. While every spill is different, investigation of this spill led to many insights: the decades-long persistence of oil in intertidal sediments, the highly toxic nature of weathered oil to developing fish embryos, and the potential for long-term depression of seabird and marine mammal populations that feed and nest in oiled habitat. Scientists working for the United States and Alaskan governments and those funded by the Exxon Corporation have extensively studied the spill since 1989. Although the results of the studies have, in many cases, been similar, profound disagreements on the interpretation of findings emerged as both groups published
The Exxon Valdez Oil Spill 421
Figure 5.2: The cumulative extent of the Exxon Valdez oil slick (in light brown), which generally moved southwest across Prince William Sound, and along the Kenai and Alaska Peninsulas (Data from Gundlach et al., 1990).
papers on their findings. We will focus on the government studies, which came to the following conclusions: (1) oil persisted beyond a decade in surprising amounts and in toxic forms; (2) the residual oil was sufficiently bioavailable to induce chronic biological exposures to nearshore species; and (3) oil had both short-term and long-term effects on a wide variety of species, with prolonged effects on species associated with oiled sediments. The Exxon Valdez spill is the most thoroughly studied oil spill in history. This section will: 1. provide a history of events in 1989, set against the background of pre-spill ecological knowledge of Prince William Sound (PWS); 2. describe the fate of the oil in the months and years following the spill, and how much oil remains;
422
Long-Term Ecological Change in the Northern Gulf of Alaska
3. describe the acute, short-term, and long-term effects of the oil on key species; and 4. discuss the recovery of the ecosystem from spill effects.
5.2. Pre-spill Conditions Prince William Sound and the northern Gulf of Alaska are remote areas with a small coastal human population, and the status of the natural resources in the region, in general, were poorly documented prior to the spill. Some harvested species, such as pink salmon and herring, were tracked very well for a number of years previously to help regulate fisheries, but most other species were not tracked and their population status was largely unknown. The nearshore and intertidal communities, which were greatly affected, were poorly understood. Biologists were not prepared for predicting the potential impacts of a large spill. In the mid-1970s, following large spills in England (the Torrey Canyon), and the United States (the Platform A blowout in the Santa Barbara Channel), the Arab oil embargo, and with the growing desire of the United States to achieve energy independence, the Nixon administration started a national program to promote off-shore leasing and energy independence. The Outer Continental Shelf Environmental Assessment Program (OCSEAP) was established to evaluate biological resources and risk on the continental shelves, and this program initiated comprehensive studies throughout the state of Alaska. The Alaskan portion of the program, because of its oil and gas potential, was particularly large, with an annual budget of about $20 million at the height of the program. This effort resulted in a massive leap forward in scientific understanding throughout all of Alaska, including the Gulf of Alaska, and made possible the landmark synthesis “The Gulf of Alaska” (Hood and Zimmerman, 1986). There was precious little pre-spill ecological data for affected areas and species, but, without the results from OCSEAP, there would have been much less quantitative data or taxonomic information on spill area resources. The boat survey data from the Fish and Wildlife Service were particularly useful for estimating the pre-spill abundances of seabirds and sea otters in PWS and murres (Uria spp.) along the shallow costal areas and islands west of PWS – all hard hit by the spill. At about the same time that OCSEAP was underway, pioneering concepts for characterizing contamination in U.S. coastal waters in a mussel watch program were being implemented (Butler, 1973; Goldberg et al., 1983; O’Connor, 1994, 2002). Mussels concentrate many contaminants, including components of crude oil that can affect marine organisms, and their analysis provides an integrated picture of average water quality. Scientists from the Auke Bay laboratory anticipated the likelihood of a spill in PWS where the primary traffic of oil tankers originates at the oil terminal in Valdez, and
The Exxon Valdez Oil Spill 423
initiated several years of pre-spill sampling of mussels and sediments to establish pre-spill chemical baselines (Karinen et al., 1993). Besides the previously mentioned USFWS boat-based surveys of seabirds and sea otters, other useful sources of pre-spill data in Prince William Sound included: 1. surveys of nesting adult bald eagles carried out by the Fish and Wildlife Service; 2. pink salmon counts in spawning streams carried out by the Alaska Department of Fish and Game; 3. herring test fishery results and aerial surveys of herring spawn carried out by the Alaska Department of Fish and Game; and 4. harbor seal aerial surveys carried out by Alaska Department of Fish and Game.
5.3. History of the Spill The 300-m long (986 ft) T/V Exxon Valdez, laden with 197 million liters (52 million gallons) of Alaska North Slope Crude oil, left the Alyeska Pipeline terminal shortly after 9 PM on March 23, 1989. A pilot guided the ship along the established route southeasterly down the center of upper Valdez Arm. After being piloted through the Valdez Narrows, the tanker entered lower Valdez Arm. The route ahead and to the west was littered with icebergs from the nearby Columbia Glacier. The tanker captain ordered the helmsman to take a more southerly route to avoid the ice and then relinquished the wheelhouse to a junior officer, with instructions to return to the shipping lane at a particular point. However, the ship did not return to the designated route. Despite the last-minute course corrections of a new helmsman, at 12:04 AM on Friday, March 24, 1989 (Good Friday), the Exxon Valdez grounded on Bligh Reef, resulting in a large tear along the underside of its hull and a rupture of eight of the eleven oil compartments, and 20 million liters (12 million gallons) of Alaska North Slope oil spilled into the largely pristine waters of PWS. This set the stage for an extensive cleanup operation, a large-scale scientific effort, and it marked a milestone in the environmental consciousness of the American public. Ironically, it was on a Good Friday 25 years earlier that one of the largest earthquakes in the recorded history of North America rearranged thousands of kilometers of shoreline in PWS and northern Gulf of Alaska, sending a huge tsunami through Port Valdez, PWS, the northern Gulf of Alaska, and beyond. For three days following the grounding, the seas were calm, and the spilled oil pooled near Bligh Reef, mainly to the southwest. During that time, attempts were made to contain the oil, but proper equipment was not available. Some dispersants were tried, but with little apparent effect without wind mixing. A test burn was even
424
Long-Term Ecological Change in the Northern Gulf of Alaska
attempted. The highest priority was to offload the remaining oil from the Exxon Valdez to prevent an even worse spill. Although approximately 80% of the cargo was successfully removed, the 20% lost was the largest spill in U.S. waters. Then, on Monday, March 27, a northerly gale (with 70-knot winds) descended on the sound and dispersed the oil to the southwest for about 100 km, and any hope of containing the oil was lost. The oil blew along the water’s surface and followed the prevailing currents in a roughly counterclockwise direction in PWS. The extensive crenulated coastline of the sound with its many islands and narrow passages to the Gulf of Alaska in the southwestern sector ensured that much of the oil would be intercepted by the predominately cobble and rock beaches of these islands. Northeastfacing beaches were particularly hard hit. Crude oil accumulated up to a foot or more on some beaches, pushing far up the beaches on high tides. Some of the oil retreated with ebbing tides, but some of it also followed the dropping water level between the rocks and cobbles and became entrained into the finer material deep within the beach–to persist for many years. In some bays, floating oil continued to rise and fall with the tides for days or weeks before it was left stranded or flushed out to other parts of the sound or to the Gulf of Alaska. The oil came ashore along portions of the mainland on the western side of the sound, e.g., Eshamy Bay and Main Bay, and other areas east of Whittier (see Fig. 5.2). Islands in the path of the oil within the sound included those in the Naked Island complex: Perry Island, Lone Island, Block Island, Ingot Island, Eleanor Island, Smith Island, Knight Island, and Green Island. The spill also affected parts of the northern end of Montague Island, but much of that island, greatly affected by the 1964 earthquake, remained unoiled. The oil heavily inundated the islands defining the southwest passages to the gulf: Latouche, Erlington, Evans, and Bainbridge, as it spread towards the open Gulf of Alaska. The floating oil was discontinuous as it exited the sound, floating in linear rafts of oily mousse. On entering the Gulf of Alaska, the slick followed the buoyant nearshore Alaskan Coastal Current, carrying oil to the outer parts of the southeastern Kenai Peninsula, and then entering Cook Inlet. In Cook Inlet, a small amount of oil was carried by strong tides north of Homer, but most of the oil was carried past the Barren Islands and south through Shelikof Strait. Some shorelines on the western side of the Strait along the Alaska Peninsula, e.g., Tonsina Bay, and parts of Kodiak Island, were heavily oiled as well. Coating of beaches outside the sound was discontinuous and varied from heavy to light or none. The remaining floating oil was carried southwesterly out of Shelikof Straight and back into the Gulf of Alaska, where the smaller and discontinuous masses of oil presumably met the fate of oil left behind by the advancing front: evaporation, photolysis, dissolution, biodegradation, and sinking. The post-spill cleanup operations occurred over a 3-year period, 1989–1991, and were restricted to the late spring and summer, and with a decreasing effort each year. The early cleanup was very widespread. In the few weeks after the spill, a small
The Exxon Valdez Oil Spill 425
but significant proportion of spilled oil was retrieved from the water’s surface by skimmers, but most of the cleanup effort, which eventually totaled more than 20 million man-hours, was expended on the beaches (Fig. 5.3). On shorelines, oil was picked up with shovels, or manually wiped from rocks, removed by hosing, flushed with seawater, or bioremediated with application of nutrient-laced liquid fertilizers (Mearns, 1996). Heavy equipment was also deployed to expose oil buried in beaches and berms to natural weathering. Deluging with seawater was done extensively to clean rocky beaches. About 600 commercial pressure washers sprayed hot water (usually at a temperature of 60°C) onto the beaches. A large number of moderately to heavily oiled shorelines were sprayed with hot seawater, killing much of the marine intertidal life (Lees et al., 1996). Then, in 1997, under pressure from Native Groups in PWS, further cleanup of Sleepy Bay on the northern end of LaTouche Island was undertaken (Brodersen et al., 1999). Damage assessment by state and federal scientists was immediately undertaken to determine the spill’s effects, especially to animals in the hard-hit southwestern sound. The task of spill assessment was enormous, and scientists had to be diverted from their normal activities. Agencies with resource management responsibilities needed to know what the effects were, how extensive and long lasting, but the information was also in support of legal action against the Exxon Corporation; hence, the studies were “litigation sensitive” (i.e., secret), and the designs and objectives were often
Figure 5.3: Oil was removed from shorelines by shoveling, manual wiping, hosing, or bioremediated with application of nutrient-laced liquid fertilizers. In this case, oil was moved downslope by hot water, trapped within a boom, and picked up with a skimmer (photograph, courtesy of Adrian Celewycz, NOAA).
426
Long-Term Ecological Change in the Northern Gulf of Alaska
influenced by this legal process. There was a rush to put studies into the field in order to document spill damages soon after they occurred. Wildlife biologists with little first-hand knowledge of spills had to learn about oil toxicity, behavior, and sampling to support chemical analyses. Immediate field sampling included gathering seabird, bald eagle, and sea otter carcasses; taking samples of mussels, seawater, and sediments in the path of the spill; and mapping the full extent of the spill. This was just the beginning of a massive effort, ramping up to at least 40 separate studies by the state and federal governments in 1989 alone. Resource agencies responded as they saw the need, but with little coordination, peer review, or oversight in the first year. There just was not enough time. In the second year, peer review and coordination between agencies were emerging in the process, but the legal team had the ultimate say on what studies were chosen and which were eliminated. The Exxon Valdez Trustee Council evolved a robust process in subsequent years for choosing studies and managing the direction of the restoration science after the 1991 settlement – an interesting story in itself, but beyond the scope of this section. As it became clear what the immediate damage was, these biological studies evolved toward understanding the long-term effects, especially to populations, and how to restore the damage. However, to track recovery from damages and to conduct restoration, the extent of damages had to be known. The paucity of pre-spill data made this a challenging goal. The assumptions made and the approaches taken to estimate the extent of damages and recovery from the spill contributed to the deep disagreements between the governments (state and federal) and the Exxon Corporation (see Box 5.1).
BOX 5.1: DUELING SCIENTISTS by Robert B. Spies The controversy over the effects of the Exxon Valdez spill is similar to those over cigarette smoking, global warming, or other controversial health and environmental issues in which the industry has a stake and governments have a role in the greater public interest. Controversies arise when government scientists and policy makers come to conclusions different from corporations and their consultants. The opposing views of the governments and the Exxon Corporation in this case were aired widely in the mass media and in various technical forums and journals. A quick read of the abstracts from two scientific meetings held in 1993, a government-sponsored symposium in Anchorage, Alaska (Rice et al., 1996), and an Exxon-sponsored symposium in Atlanta, Georgia (Wells et al., 1995), reveal stark contrasts in conclusions between the two groups of scientists studying spill damages and recovery. This may be puzzling to environmental scientists who were not involved, let alone the public. How could these widely differing accounts arise from one reality?
The Exxon Valdez Oil Spill 427
There was a great deal to be learned in the aftermath of the spill, but the atmosphere was not dispassionate. On both sides, legal concerns affected study designs, levels of financial support, and goals. Government scientists evaluated damage to the coastal ecosystem of the northern Gulf of Alaska to support management decisions and, subsequently, a claim for damages from Exxon Corporation. Exxon scientists designed studies to defend against government claims and discount government findings. Consequently, while both sides studied the impacts of the spill, the designs were different in concept, statistical rigor, and choice of study sites, and these differences led directly to the differing conclusions (Peterson et al., 2001). In addition to differences in study designs, there was also a marked difference in standards of proof. Exxon would not accept much less than a complete causal chain linking spilled oil to its ultimate effects, while the governments (State of Alaska and U.S.) used a weight-of-evidence standard. Since a complete causal chain is difficult or practically impossible to establish in the wake of accidents in poorly characterized ecosystems, it was easy for Exxon scientists and consultants to reject government conclusions on damage and recovery, and create controversy and doubt. Such differences in standards of proof are common in similar controversies, e.g., with regard to the links between cigarette smoking and lung cancer. A good example of the discrepancy of conclusions is the question of what effect the spill had on pink salmon juveniles. There were juveniles with coded wire tags in their noses, released from hatcheries in western PWS in the spring of 1989. Some of the juveniles were recovered in the oiled waters around Knight Island, and their growth was compared with tagged juveniles captured in unoiled areas (Willette, 1996). The juveniles collected in the unoiled areas had grown significantly more than those captured in the oiled areas, and there were similar findings in a second study (Wertheimer and Celewycz, 1996). Juvenile salmon captured in the latter study also had induced P4501A, a molecular marker of oil exposure (Carls et al., 1996a), as did some pre-emergent fry captured in oil-affected streams (Wiedmer et al., 1996). Combined with a known relationship between rate of growth and survival to adulthood, it was possible to model the effect of reduced growth in the juvenile stage on numbers of returning adults in 1990 (Geiger et al., 1996) and led government trustees to conclude that return of pink salmon to western PWS in 1990 had been reduced by the spill (EVOS Trustee Council, 1994). Exxon-sponsored scientists countered that, as there was no record of where the juveniles were between release from the hatchery and collection in the oiled waters, the slower growth of the juveniles found in the oiled areas may have been caused by exposures to colder temperature, and on this basis they rejected a link between the oil, slower growth in the juveniles, and adult returns (Brannon and Maki, 1996). The Exxon-sponsored
428
Long-Term Ecological Change in the Northern Gulf of Alaska
concern sounded plausible, except that the government studies had accounted for possible temperature effects, and subsequent government research demonstrated growth depression from oil exposure in the laboratory (Carls et al., 1996b) that confirmed the size of the growth effect attributed to oil exposure in the field. Science progresses through the articulation of differing interpretations of the same phenomena and their resolution through experimentation and debate. By itself, presentation of alternative explanations is an essential ingredient of the scientific process, ensuring that all the pertinent facts are considered. Eventually, apparent differences among interpretations are either resolved, or identify research opportunities that are often unexpectedly fruitful. Unfortunately, this scholarly endeavor is compromised when forced into confrontations in a legal arena, especially when there is substantial financial risk that may result and causing doubt may be sufficient to avert this risk. “Causing doubt” has been described by Michaels (2005) as a growing use of private consultants by industry on drug and environmental health issues. To stave off government litigation or regulation, some industry groups manufacture uncertainty about solid scientific findings. Use of study designs by Exxon scientists that were underpowered to detect significant differences, or where study sites were not picked in impact areas (Peterson et al., 2001), suggests that causing doubt is a strategy that Exxon has employed. A clear motive for this is evident, given the ongoing litigation over the Exxon Valdez spill. The civil suit by fishermen and Alaska natives has been judged against Exxon, but has been in appeal for nearly a decade, with payments of up to $5 billion potentially at risk. Further, at the time of this writing, the state and federal governments are weighing the issue of the “re-opener clause” of the 1991 settlement, and litigating with Exxon for an additional $100 million for damages above those known at the time of the 1991 settlement. Consequently, there is significant economic motivation for Exxon to support further studies that “cause doubt” relative to long-term persistence of oil and long-term damage both in Alaska and elsewhere.
5.4. Oil Fate: Transport, Weathering, and Persistence Only the early phases of transport and transformation of petroleum hydrocarbons followed expectations. Many of the effects of the oil spill were due to the way the oil behaved in the months and years following the spill. After the first few weeks, the oil was transported, transformed, and retained in ways that were often unexpected. In order to fully appreciate how and why the petroleum hydrocarbons (the compounds that make up oil) altered the ecology of PWS, it is helpful to examine the types of hydrocarbons present in the sound prior to the spill, the initial
The Exxon Valdez Oil Spill 429
transport of the oil into the environment, the effect of cleanup efforts (both natural and man-made), the transformation of the hydrocarbons in the oil, and the long-term persistence of some toxic compounds in the environment. Finally, we will examine how available these residual toxic hydrocarbons were to the animals of the sound.
BOX 5.2: BTEX AND PAH: PETROLEUM’S TOXIC CULPRITS by Jeffrey W. Short Oil and petroleum products are mostly made up of hydrocarbons, and are a major source of an important class of toxic chemicals known as aromatic hydrocarbons. Hydrocarbons are chemicals containing only carbon and hydrogen. The carbon atoms in aromatic hydrocarbons are joined together to form rings and are bonded to fewer hydrogen atoms than they could be (i.e., they are unsaturated). To be aromatic, the carbon rings must have clouds of electrons above and below them that contain a particular number of electrons (i.e., 6, 10, 14, …), in addition to the electrons between the carbon atoms that bind them together more directly. These delocalized electrons add bonding strength, making aromatic hydrocarbons much more stable (and hence persistent in the environment), than saturated hydrocarbons, and they are also the cause of the toxicity of these molecules. The simplest aromatic hydrocarbon is benzene, which contains six carbon atoms and six hydrogen atoms (Fig. 5.4). Chemists use a kind of shorthand to depict these molecules, in which two lines meeting at an angle indicate a carbon atom, a circle within a hexagonal ring of benzene stands for the six delocalized electrons, and the hydrogen atoms at the edge of the ring are not shown because there can only be one for each carbon atom. The alkanes are another important class of hydrocarbons in petroleum (Fig. 5.4), and consist of carbon skeletons that are bonded to the maximum number of hydrogen atoms possible (i.e., they are saturated). Simple alkanes include methane, ethane, propane, and two kinds of butane (see structures). When one of the hydrogen atoms of these compounds is replaced by something else, the name of the remaining hydrocarbon part is modified by replacing the –ane with –yl, as in methyl, ethyl, propyl, etc. (see structures), which are referred to generally as alkyl groups. Any one of the hydrogen atoms on the benzene ring may be replaced by an alkyl group, which may contain any number of carbon atoms. Addition of a methyl group (one carbon), produces toluene (see structure), addition of two methyl groups produces xylene (see structures), and there are other combinations of one or more groups, each of which could contain one to several carbons atoms. Note that there are three different kinds of xylene, depending on the geometric arrangement of the two methyl substituents. These six monocyclic aromatic
430
Long-Term Ecological Change in the Northern Gulf of Alaska
hydrocarbons are often abbreviated as “BTEX” (benzene, toluene, ethyl benzene, and toluene), and they constitute about 1.6% of fresh Alaska North Slope crude oil, the type spilled in PWS (see Appendix B). They are very volatile and pose a serious inhalation hazard during oil spills. Benzene in particular is a potent carcinogen. More complicated monocyclic aromatic hydrocarbons may be formed through more alkyl substitution (Fig. 5.4). Note that as the number and complexity of these added hydrocarbons increase, so does the number of structurally distinct compounds, giving rise to rapidly increasing numbers of isomers (compounds with the same numbers of atoms of each element that are bonded together differently). Polycyclic aromatic hydrocarbons (PAH) are formed when two or more aromatic rings (each containing six carbons) are joined together (Fig. 5.4). The compound biphenyl results when two benzene rings replace a hydrogen atom on each other (Fig. 5.4), while naphthalene results when two benzene rings fuse (Fig. 5.4). As with the benzene compounds, the number of PAH isomers increases rapidly with the number and complexity of saturated hydrocarbons that replace the hydrogen atoms. Hence, methylnaphthalene has 2 distinct isomers, dimethylnaphthalene has 10, and trimethylnaphthalene has 11. Even more isomers result when the substituting groups are different (e.g., methylethylnapthalene with 14). Environmental chemists usually combine results for these PAH isomers according to the number of carbon atoms in the substituting groups, for example, the ten dimethylnaphthalenes are lumped together under the label “C2-naphthalenes,” where “C2-” denotes that the substituting alkyl groups include two carbon atoms (i.e., two methyl groups or one ethyl group: “C3-naphthalenes” includes the 11 trimethylnaphthalenes, 14 methylethylnaphthalenes, and 4 propylnaphthalenes). The presentation of the PAH composition of Alaska North Slope crude oil follows this convention in Appendix B. Other important PAH in Alaska North Slope oil include fluorene and alkylsubstituted fluorenes, the phenanthrenes, pyrenes, chrysenes, and other 4- and 5-ring PAH (Fig. 5.4). The delocalized electrons cause PAH to be toxic in two fundamental ways. First, they are capable of reacting with biomolecules and thereby disrupting their function (whereas alkane hydrocarbons are nearly inert except when oxidized). For example, PAH may be transformed by cells into forms that react with DNA and RNA, interfering with cell division and with protein production. Second, certain PAH may catalyze the formation of a highly reactive form of molecular oxygen within cells when exposed to ultraviolet radiation (called photoinduced toxicity). Photoinduced toxicity may occur at partper-billion (ppb) PAH concentrations within affected cells, and translucent biota exposed to intense sunlight are at greatest risk.
The Exxon Valdez Oil Spill 431
Figure 5.4: Structure examples of aromatic, heterocyclic, and alkane hydrocarbons present in Exxon Valdez crude oil.
432
Long-Term Ecological Change in the Northern Gulf of Alaska
5.4.1. Contaminants in Prince William Sound Prior to the Exxon Valdez Oil Spill Most of the region affected by the Exxon Valdez oil spill was pristine wilderness. About 7000 people lived in PWS in 1989, almost all of them in Cordova, Valdez, and Whittier. Less than 200 people resided within the spill trajectory inside the sound, mainly Alaska natives in the village of Chenega, and residents of a fish hatchery at Sawmill Bay. Few people lived along the coasts of the Kenai and Alaska peninsulas and Kodiak and adjacent islands, except for the ∼8000 residents in the city of Kodiak. The primary industry in the spill region was commercial fishing, which introduced negligible contamination to shoreline sediments and biota. Small-scale mining and land-based fish processing were important industries historically within PWS, but these were few and scattered (Lethcoe and Lethcoe, 1994). Contaminant effects of these activities on shorelines were attenuated by the 1964 Great Alaska Earthquake, which uplifted most shorelines within the spill trajectory from 1 to 10 m into the supratidal (the zone immediately above the highest reach of the tides). Although contaminants from these human activities were occasionally found in inter- and shallow subtidal sediments, they were quite localized, affecting a very small fraction of the shoreline (Karinen et al., 1993). Human activities occur on ∼0.2% of the shoreline of PWS (Boehm et al., 2004). Prior to the spill, the most likely sources of hydrocarbons on shorelines and shallow subtidal sediments within the spill region were asphalt and fuels from storage tanks in Valdez and elsewhere that ruptured during the 1964 earthquake (Kvenvolden et al., 1995), and a natural “regional background” of hydrocarbons from eroded organic-rich shales and siltstones east of PWS. The high viscosity of the asphalt (>10,000 centipoise) prevented it from penetrating into subsurface intertidal sediments, so patches became stranded by high tides on surface rocks. Small asphalt patches may still be found firmly adhered to cobbles, boulders, and bedrock above +3 m tidal elevation, but they were a small amount (<3%) relative to the Exxon Valdez oil remaining in PWS by 2001 (Short et al., 2004a). Natural oil seeps were proposed as the source of natural hydrocarbons found throughout the shelf sediments of the northern Gulf of Alaska (Bence et al., 1996; Page et al., 1995), but appear now to be negligible sources, with eroded shelf rock being the major source (Short et al., 2004b). These hydrocarbons are not bioavailable because they are sequestered within coal or rock matrixes (Short et al., 2004b). The hydrocarbon “source rocks” are eroded by streams and by glaciers from outcrops of the Kulthieth and Poul Creek formations along the southern coast of the Gulf of Alaska from Katalla to Yakutat Bay (Van Kooten et al., 2002). Finely eroded sediments become entrained by the Alaska Coastal Current and transported to PWS and westward, where they settle on subtidal sediments. Concentrations of source rock PAH tend to increase with depth, ranging from less than ∼100 ng g−1 dry sediment in
The Exxon Valdez Oil Spill 433
the intertidal of the spill-affected region to ∼1500 ng g−1 in benthic sediments of the deepest parts of the sound (O’Clair et al., 1996; Page et al., 1995). Deposition of PAH from forest fires on Kenai Peninsula has also been reported (Page et al., 1999), but hydrocarbon signatures indicative of combustion sources are at trace levels except near former or present human habitation sites (Carls et al., 2004a), which are widely scattered.
5.4.2. Initial Fate of the Oil The fate of the oil was largely determined by the weather during the first few weeks of the spill. Oil discharge from the grounded vessel began spilling shortly after midnight on Friday, March 24, 1989, under light winds and calm seas, and a seawater temperature of about 5°C. Winds were variable and below 5 m s−1 for the next 2.5 days, and the oil slick was a compact and roughly circular pool southwest of Bligh Island. During this period, the oil slick spread at rates up to 2000 m2 s−1 to an area of nearly 400 km2, with thicknesses that decreased from generally less than 1 mm after the first few hours to perhaps 0.1 mm by Sunday. The enormous increase in the surface area of the oil accelerated evaporation of the most volatile components, leading to losses of perhaps 15% by weight, including nearly all of the benzene, toluene, ethylbenzene, and xylenes (BTEX), and of the saturated hydrocarbons with vapor pressures greater than that of dodecane (Payne et al., 1991). A chemical dispersant (Corexit 9527) was applied to a small portion of the slick on Friday (the day of the spill), but it was ineffective as the sea was dead calm and the slick was too thick. A test burn of about 100 m3 of oil was successful on Saturday, at slick thicknesses of 1–3 mm, demonstrating that evaporation rates were sufficient to permit ignition 40 h after the spill. A larger burn was planned for the following day, but was precluded by a storm entering the Gulf of Alaska. Oil slicks were distributed by wind and currents in the weeks and months following the spill (Fig. 5.5). Gale force northeast winds up to 35 m s−1 disrupted the slick and drove it shoreward during the next 3 days beginning Sunday afternoon (March 26). The first landfall of the oil slick occurred on Monday morning with breaking waves of 1–3 m on beaches of Naked Island and nearby islands. The oil continued to move along the east and west coasts of these islands towards Smith and Eleanor Islands, and onward toward the southwest along the coasts of Knight Island and islands adjacent. Beaches on a complex of islands near the southwestern margin of PWS, including Latouche, Evans, and Bainbridge Islands, were affected by the oil on the last day of the storm (Wednesday, March 29). The leading edge of the slick exited the sound the following day as seas and winds diminished. The storm dispersed the slick as small oil droplets in the water, and also promoted water incorporation into the oil. Oil droplet dispersion occurred at least to depths of
434
Long-Term Ecological Change in the Northern Gulf of Alaska
Figure 5.5: Oil transported as surface slicks drifted along coastlines. Critical areas were sometimes protected with floating booms; in this case the boom did not prevent all the oil from entering the cove (photograph, courtesy of Alex Wertheimer, NOAA).
25 m (Short and Harris, 1996), and probably considerably deeper because the water column in PWS is not stratified in late winter (Vaughan et al., 2001). The numerous small oil droplets allowed the more soluble oil components to dissolve in seawater, leaving PAH concentrations as high as the low parts per billion (ppb) weeks after the storm had passed (Short and Harris, 1996). Water incorporation into the surface slick along with continued evaporative weathering led to the formation of a “mousse” with viscosities ranging from several hundred to a few thousand centipoise (Payne et al., 1991; Bragg and Yang, 1995). The high winds thickened the slick as it approached beaches, and the waves pushed the oil up the beaches, stranding substantial accumulations during falling tides. The upper ∼4 m (vertical) of the intertidal zone initially accumulated most of the oil (Fig. 5.6). On porous beaches, pools of stranded oil percolated into subsurface sediments as the water table lowered on the ebb tide. Even the fine-grained sediments of these beaches are very permeable (Carls et al., 2003; Bragg and Yang, 1995), so the water table in the beach followed the tide closely, and permitted the more viscous oil several hours to seep into the underlying sediments, to depths of 1 m in some locations (Neff et al., 1995). Subsequent rising tides failed to completely remove this subsurface oil because the increased oil viscosity inhibited displacement from finer-grained sediments. Rock, cobble, and coarse-grained sediments often protected subsurface
The Exxon Valdez Oil Spill 435
Figure 5.6: An oil-coated beach in Snug Harbor, Prince William Sound (photograph, courtesy of Mark Carls, NOAA).
reservoirs, slowing the processes of physical dispersion and weathering. Fine-grained sediments often lay beneath coarser sediments on PWS beaches because much of the sound affected by the spill was uplifted by the 1964 Alaska earthquake, which elevated fine-grained subtidal sediments into the intertidal zone that were subsequently covered by rock transported from the new upper intertidal to the lower intertidal by erosion and wave action. This initial percolation and retention of oil into subsurface sediments helped set the stage for long-term persistence. For weeks following the storm, the oil in PWS was redistributed by wind, tide, and currents, while the oil that escaped the sound became entrained in the Alaska Coastal Current, and was carried southwest along the Kenai Peninsula. About 16,500 m3 of oil ultimately beached within PWS, contaminating patches along 783 km of shoreline, compared to about 3700 m3 of oil that contaminated patches of beaches outside the sound along another 1300 km of shoreline (Wolfe et al., 1994; ADNR, 1991; Neff et al., 1995). The redistribution of oil within the sound occurred as successive tides refloated stranded oil in the intertidal and transported it to other beaches or exported it to the Gulf of Alaska during the ensuing months (Fig. 5.7). The viscosity of the oil continued to increase with weathering (mostly evaporative), and when the oil became too viscous to penetrate into beaches, it formed surface pavements and tar mats that were often extensive, covering several hundred m2 at thicknesses of up to several centimeters. These surface oil deposits were usually located in the upper intertidal zone
436
Long-Term Ecological Change in the Northern Gulf of Alaska
Figure 5.7: Oil retention in intertidal beaches. Oil slicks encroached at high tide; stranded oil penetrated the highly porous sediment during the hours of ebb tide and became trapped in interstices. Subsequent tidal cycles removed oil slowly; liquid oil is still present in some beaches (2004), as illustrated in the fourth photograph (photographs, courtesy of NOAA).
after initial stranding at high tides, in contrast to the subsurface oil, which was most frequent at the mid-tide elevations (Short et al., 2004a), a fact that was not fully appreciated for 12 years. Oil escaping into the Gulf of Alaska also became more viscous, but still penetrated into subsurface sediments on some beaches (Irvine et al., 1999). Boulder–cobble beaches are more common on the outer coast of the Gulf because of the frequent large waves,
The Exxon Valdez Oil Spill 437
and the more porous beaches could absorb the more viscous oil. Surface oil deposits also formed on outer coast beaches, where the beach porosity did not allow oil penetration. By the end of May 1989, the oil slick had transited Shelikof Strait and entered the southwestern margin of the Gulf of Alaska as scattered patches. Tarballs were reported near Chignik (Neff et al., 1995), with perhaps 30% of the oil ultimately dispersing into the Gulf as tarballs that slowly degraded from physical and biological processes (Wolfe et al., 1994). These processes included continued evaporation, and dissolution of the more labile components, photo-oxidation, and microbial degradation. About 20% of the dispersed oil (i.e., ∼3000 m3) ultimately became associated with enough sediment to sink to the seafloor of the Gulf (including PWS), while the remainder joined the tarball population of the northern Pacific Ocean. About half the oil discharge volume (∼20,500 m3) remained on the beaches of PWS and the Gulf of Alaska, and unprecedented human efforts were undertaken to remove it.
5.4.3. Oil Cleanup Efforts Exxon Corporation sponsored the most extensive shoreline cleanup ever attempted. Involving more than 10,000 people and $2 billion, this cleanup used several methods over the first three years. During the summer of 1989, manual, hydraulic, and bioremediation methods were used on 396, 486, and 292 of 1060 beach segments, respectively. Usually, the more heavily oiled beaches were treated by multiple methods (Mearns, 1996). Successively smaller cleanups were carried out during 1990 and 1991. The most efficient cleanup method was collection of oil floating on the sea surface by skimming operations (Fig. 5.8). About 2900–3500 m3 of the original spill volume was collected by oil skimming vessels within PWS, primarily during the first month (Wolfe et al., 1994). The amount of oil recovered by surface skimming was comparable with the amount recovered from oiled beaches over the next three years, which accounted for most of the cleanup expense. Manual methods involved removing surface tar mats, oiled debris, and oiled vegetation by hand, application and recovery of oil absorbance pads, and shoveling thick lenses of oil. These methods were sometimes complemented by the use of oilabsorbent materials attached to cables set parallel to the shoreline (“snare booms”) that moved over portions of beaches driven by tidal movement. Mechanical tilling on 54 beaches exposed subsurface oil to more rapid weathering through dispersion. The hydraulic methods were more controversial because of their greater effect on beaches. High-pressure washing (usually 345–827 kPa), especially by hot water (approximately 60°C), had traumatic effects on intertidal life, and it was discontinued after 1989 (Fig. 5.9). The high velocity of the water from the high-pressure washing dislodged intertidal biota through scouring, and the high temperatures exceeded the
438
Long-Term Ecological Change in the Northern Gulf of Alaska
Figure 5.8: Collection of floating oil by skimming was an efficient clean-up method (photograph, courtesy of Adrian Celewycz, NOAA).
Figure 5.9: High-pressure beach washing damaged intertidal organisms, particularly when hot water was used (photograph, courtesy EVOS Trustee Council).
The Exxon Valdez Oil Spill 439
short-term tolerances of many species, in effect partially cooking them (Peterson, 2001). These methods were often coupled with spraying large volumes of seawater onto the upper intertidal, flushing the surface of the treated beach. These methods also removed fine-grained sediments, which inhibited recolonization of some species for a decade or more (Driskell et al., 1996; Houghton et al., 1996; Lees et al., 1996; Peterson, 2001). Some of the oil-laden fine-grained sediments transported oil to the shallow subtidal (Short et al., 1996). The high temperatures made the oil less viscous and, therefore, better able to penetrate the subsurface sediments on some beaches. Bioremediation usually consisted of an application of nutrients (mainly nitrate and phosphate) in a water-or lipid-soluble carrier to promote microbial decomposition of oil. Very limited addition of microbes and enzymes tailored for oil decomposition was employed on an experimental basis (Fig. 5.10). These methods often accelerated removal of surface oil (Bragg et al., 1994), but were less effective at removing subsurface oil. Chemical removal agents were tested only during 1989, but were not approved for widespread use. Bioremediation and manual removal methods were continued in 1990 and 1991 on beaches where oil appeared to persist. To accelerate weathering and dispersion, sediment accumulations left in the upper intertidal zones were broken up to expose oil in the sediments. Hydraulic spot-washing was used on 58 beach segments in 1990 and on one segment in 1991.
Figure 5.10: Bioremediation, that is, the promotion of microbial decomposition of oil by application of nutrients, often accelerated removal of surface oil, but was less effective at removing subsurface oil (photograph, courtesy EVOS Trustee Council).
440
Long-Term Ecological Change in the Northern Gulf of Alaska
By the end of 1992, 5–8% (2040–4080 m3) of the initial spilled oil had been recovered from beaches (Wolfe et al., 1994). Beach cleaning also helped disperse the fine-grained oiled sediments, which were ultimately deposited on bottom sediments over a broad area of the coastal shelf of the northern and western Gulf of Alaska. Beach cleaning remains controversial. Some methods effectively remove surface oil, but they also destroy beach flora and fauna. The long-term persistence of oil, which was not appreciated in 1989–1991 to the extent it is today, adds another layer of uncertainty to judgments of how aggressively to clean beaches. Harsh beachcleaning methods may be very harmful to intertidal fauna, but may be the best course of action for the protection of vertebrate predators such as birds and harbor seals that use the rocks as resting areas, and in some cases as foraging grounds. We do not know how to strike a balance between not treating to preserve intertidal life and cleaning to lessen exposure of foraging animals.
5.4.4. Long-term Oil Persistence Early Surveys (1989–1993) The massive beach cleanup effort required comprehensive monitoring to allocate cleanup resources and to evaluate efficacy. The initial location and extent of oiled shorelines was documented by low-altitude, low-airspeed color videotaping (Teal, 1991). These surveys defined the general area to be evaluated and monitored by groundbased “shoreline cleanup assessment teams” (SCAT). The SCAT teams included at minimum a geomorphologist, an ecologist, and an archaeologist, who comprehensively inspected the entire shoreline within the potential affected region during beach walks or from boats when beach access was impractical (Neff et al., 1995). The SCAT teams received similar training and employed uniform criteria and forms for recording observations (Owens, 1999). These observations provided the basis for segregating the shoreline into contiguous series of beach segments up to 2.5 km long that were bounded by readily identifiable landmarks in the field. Visual assessment of oiling intensity on each of these segments was carefully and consistently documented during the spring and summers of 1989 through 1992 (Fig. 5.11). Subsurface oil was also monitored after 1989 by excavation of thousands of pits to assess persistence. The SCAT survey results showed that surface oil on PWS beaches dispersed quickly, but corresponding results for oiled beaches outside the sound have not been reported. The cumulative length of visibly oiled beach segments inside PWS decreased from 783 km in 1989 to 10 km by 1992. The cumulative area of oiled beach was substantially less than what might be inferred from these results because the beaches were often not entirely coated by oil, especially after 1989. Very approximate estimates of oiled beach area may be calculated from the oiled shoreline lengths and percentages of intertidal area covered by oil given by Neff et al. (1995), which
The Exxon Valdez Oil Spill 441
Figure 5.11: Visual assessment of shoreline oiling, completed in fall 1989 by the Alaska Department of Environmental Conservation, after the first summer’s cleanup efforts (Gundlach et al., 1990).
indicate a decline from 240 ha in 1989 to about 20 ha by 1990, and less than 1 ha by 1992. A total of 2.7 ha of oiled beach area was reported from the 1993 SCAT survey (Gibeaut and Piper, 1998). Subsurface oil appeared to decline more slowly than surface oil. The surface area of beaches contaminated by moderate to very heavy subsurface oiling exceeded 5 ha in 1991, and decreased by nearly 70% from 1991 to 1992 on beaches surveyed both years, but the total extent of subsurface oiling was not reported for either year (Neff et al., 1995). In 1993, the total beach area contaminated by subsurface oil was estimated as 3.4 ha, and this oil was suspected to be more firmly associated with the sediments, and hence less easily dispersed compared with prior years (Hayes et al., 1991; Gibeaut and Piper, 1998). The amount of oil remaining on PWS beaches in 1992 was estimated to be about 2% of the volume spilled, or about 817 m3 (Wolfe et al., 1994).
442
Long-Term Ecological Change in the Northern Gulf of Alaska
There were no long-term appreciable accumulations of subtidal oil. Although fluxes of oil were substantial to the shallow subtidal during the first year following the spill (much of it aided by scouring from the cleanup efforts), these did not usually result in accumulations because the subtidal sediments were continually resuspended and dispersed after initial oil deposition (Short et al., 1996). Some of the oil formed clay–oil flocs were of near-neutral buoyancy, promoting dispersion over a wide area of the northern Gulf of Alaska (Bragg and Yang, 1995). Oiled sediments did accumulate in the shallow subtidal adjacent to heavily oiled beaches when trapped in basins, for example, oil collected behind a terminal sill at the mouth of Northwest Bay. Subtidal sediment PAH concentrations on the order of 1000 ppb, and decreasing with depth, occurred in 1990. In succeeding years, these concentrations declined rapidly (O’Clair et al., 1996). A resurvey of subtidal sediments during 2001 failed to detect hydrocarbons from the Exxon Valdez near any of the five beaches that had been heavily oiled (Short et al., 2003). The rapid loss of oil from PWS beaches determined by the SCAT surveys implied that the remaining oil would soon disappear (Boehm et al., 1995). It was suggested that clay–oil floc formation would lead to rapid natural removal of subsurface oil (Bragg and Yang, 1995), although the low concentrations of clay-sized sediments in the surface seawaters of PWS (≤1 ppm) likely constrained the significance of this process (Payne et al., 1989, 2003). Comparison of the volume of oil remaining on these beaches in fall 1992 with the estimated volume that originally beached in 1989 (∼42% of the spill volume, or ∼17,200 m3) (Wolfe et al., 1994), and assuming first-order dispersion kinetics during the 3.5-year interval, leads to an instantaneous dispersion rate of −0.87 yr−1, or 58% less oil every year. If this were true, only about 60 m3 of oil would have remained by the fall of 1995. Unfortunately, such predictions were overly optimistic.
Post-1993 Surveys Studies conducted after 1993 supported the conjecture of Gibeaut and Piper (1998) that the remaining oil would be less easily dispersed and degraded by natural forces. Based on repeated surveys of fixed sampling sites from 1989 through 1997, Hayes and Michel (1998, 1999) found substantial and relatively unweathered subsurface oil at 8 of the 11 stations in 1997. These occurred at sediment depths of 25–50 cm beneath cobble/boulder armor on coarse-grained gravel beaches in PWS. Irvine et al. (1999) found subsurface oil at five stations on the Alaska Peninsula that were sampled in 1994. These were beaches that were more exposed and were more heavily armored by boulders. The PAH composition of this oil had changed little since 1989. Remnant oil also persisted beneath mussel beds in PWS through 1999 (Carls et al., 2001; 2004b). Oil remained so persistent at Sleepy Bay on northern LaTouche Island that another cleanup effort was made in 1997, and substantial subsurface oil deposits were found there (Brodersen et al., 1999). So, there was a surprising persistence of relatively unweathered oil in beaches observed in several relatively small-scale studies, and this
The Exxon Valdez Oil Spill 443
realization prompted a major effort to estimate the true amount of oil remaining and the extent of beach area affected during summer 2001 inside PWS. The methods for the 2001 estimate of oil in PWS (Short et al., 2004a) were fundamentally different from the SCAT surveys. The 2001 survey had different objectives, namely to provide an unbiased estimate of the amount of oil remaining on the beaches. In contrast, the SCAT surveys were used to direct the cleanup efforts during 1989–1991. The 2001 study employed stratified random sampling instead of attempting a comprehensive assessment so that the study could be completed during one summer field season. Random sampling was used to select beaches among the heavily and moderately oiled beaches previously identified for evaluation and to locate sampling quadrats on the beaches selected. This approach has the advantage of providing a confidence interval (an estimate of the error) for the quantities estimated, at the disadvantage of not locating all of the remaining oil. The sampling was stratified according to 0.5 m tidal elevation intervals in the upper half of the intertidal. The lower intertidal zone was only sampled occasionally to delineate the size of discovered oil patches, because the prior SCAT studies indicated that the persistent oil was almost entirely in the upper intertidal (Neff et al., 1995; Gibeaut and Piper, 1998). The 2001 survey would unfortunately demonstrate otherwise. The results from the 2001 study confirmed that the oil was more persistent than predicted. The cumulative areas of beach contaminated by surface and subsurface oil were 4.13 ha (95% confidence interval: 2.07–7.17 ha) and 7.80 ha (95% CI: 4.06–12.7 ha), respectively (Short et al., 2004a). These estimates of remaining oiled areas in 2001 exceeded those based on the SCAT surveys of 1992 or 1993 (Neff et al., 1995, Gibeaut and Piper, 1998). The distribution of surface oil was largely disjointed from the subsurface oil distribution, with surface oil found throughout the upper intertidal, but subsurface oil most frequently found near the mid-intertidal zone (Fig. 5.12). Occasional sampling confirmed that subsurface oil occurrences extended into the more biologically productive lower intertidal, nearly to 0 m tidal elevation. The volume of subsurface oil estimated to remain in the upper intertidal in 2001 was 58 m3 (95% CI: 27.5–99.4 m3 (Short et al., 2004a), assuming a density of 0.95 for weathered oil (Payne et al., 1991). The total amount of oil remaining was probably about twice this amount, after accounting for surface oil and for subsurface oil in the lower intertidal, which was excluded from the quantitative estimate. Hence, the amount of oil remaining on PWS shorelines in 2001 was around twice the amount predicted to remain by 1995. The discrepancy between the 2001 survey results and expectations based on the earlier SCAT surveys is not only because the oil was more persistent than anticipated. The SCAT surveys systematically underestimated the extent and amount of subsurface oiling, with biases that increased annually during the early 1990s. The SCAT survey efforts were focused on the upper intertidal, where oil was incorrectly assumed to be most persistent (Jahns et al., 1991; Neff et al., 1995; Gibeaut and Piper, 1998;
444
Long-Term Ecological Change in the Northern Gulf of Alaska
Figure 5.12: Twelve years after the spill, the distribution of surface and buried oil in beaches was disjointed; surface oil was found throughout the upper intertidal, but subsurface oil was most frequently found near the mid-intertidal zone (photographs, courtesy of Mandy Lindeberg, NOAA).
The Exxon Valdez Oil Spill 445
Hayes and Michel, 1999). Because the upper intertidal zone was drier, it was thought to be conducive to increased oil adhesion (Owens, 1991), so subsurface oil in the middle and lower intertidal zone were often overlooked. Surface oil was also less visually apparent after 1989 because of surface weathering, and these biases contributed to the overly optimistic predictions of rapid dispersion and degradation in the early 1990s. Observation of long-term bioavailability, presented in Section 5.5.3, corroborates the 2001 survey estimates. The 2001 surveys showed that oil was much more abundant on PWS shorelines than anticipated in the 1990s, with commensurate increase in risk of exposure to fish and wildlife. The estimated volume of oil that initially affected beaches (∼16,500 m3) is based on the initial discharge volume and predictions of the NOAA On-Scene Spill Model (Galt, 1991a,b), which accounts for evaporation, oil that escaped to the Gulf of Alaska, and dispersion, and may therefore be regarded as relatively precise. About 80% of the oil was estimated to have dispersed from shorelines during the first year (Michel et al., 1991), and this seems plausible considering the intensive cleanup in 1989, high-energy waves in winter storms, and that much of the oil was only loosely associated with these beaches. Assuming 3300 m3 of oil remained on shorelines by 1990 and first-order exponential dispersion thereafter to ∼100 m3 of oil by 2001 (accounting for surface oil and oil in the lower-intertidal) leads to an instantaneous dispersion rate of −0.32 year−1. At this rate, about 1750 m3 of oil would have remained by 1992, and 670 m3 by 1995, in contrast to the ∼60 m3 anticipated by 1995 on the basis of the SCAT surveys of 1990–1992 (Fig. 5.13). Hence, oil was likely more abundant than anticipated by factors of ten or more through the mid- to late1990s, and the subsurface proportion was often well preserved, with a proportional complement of PAH, the most toxic fraction of oil, that was only slightly smaller than what was present in the oil initially (Short et al., 2004a).
5.4.5. Weathering and Bioavailability of Persistent Oil Unlike the n-alkanes, which are readily degraded by microbes, losses of PAH from oils (and hence PAH bioavailability) are determined primarily by physical factors – evaporation and dissolution. The two factors that limit these processes, and hence the bioavailability of PAH in petroleum, are the ratio of surface area to volume of the oil, and the extent of weathering. Tiny droplets or thin films of oil release PAH rapidly until the supply of PAH is exhausted. Loss rates of individual PAH depend on molecular size, with small molecules being lost more readily. Whereas the dynamics of PAH dissolution losses are usually modeled by equilibrium thermodynamics (e.g., Shiu et al., 1988), an even simpler approach may be used for the very sparingly soluble alkyl-substituted PAH, especially those with three or more aromatic rings or extensive alkyl substitution. The larger PAH dissolve into water or evaporate into air very slowly, and, in the environment, the volumes of water or air available to dilute
446
Long-Term Ecological Change in the Northern Gulf of Alaska
Figure 5.13: Projections of oil persistence based on SCAT survey estimates of remaining oil in 1990–1992 (blue line), and on backcasting from a systematic survey of actual amounts of subsurface oil in 2001 (red line).
PAH that escape from oil are usually greater than the volume of oil from which PAH are lost by many orders of magnitude. This means that the ambient water or air are effectively infinite sinks for PAH lost from oil, and that losses of PAH from oil are almost always final, meaning equilibrium is never established, and hence PAH losses are determined solely by kinetic lossrates. The absolute PAH loss rates are affected by a number of environmental factors (e.g., temperature, wind speed, ratio of oil surface area to volume, etc.), making them very difficult to measure. However, their relative loss rates (i.e., loss rates relative to each other, or to any particular PAH selected as a reference rate) are more tractable. Such relative rates are determined primarily by the molecular surface area, and the pattern of relative PAH losses from oil reflects the cumulative effects of all the environmental factors that affect the absolute PAH loss rates. Application of these insights to the patterns of PAH losses from oiled sediments and mussels analyzed for the Exxon Valdez oil spill indicated that nearly all of the 1500 samples evaluated showed the same pattern of PAH weathering losses (Short and Heintz, 1997). This pattern is described by:
Pi,j,t = Pi,j,0 exp (−kj wj), where Pi,j,t is the concentration of the jth PAH in the ith sample of oil at time t, kj is a relative dissolution rate constant for the jth PAH, and wi is a weathering parameter describing the extent of evaporative and dissolution losses. The weathering parameter, w, provides a convenient means of quantitatively comparing PAH compositional changes induced by physical weathering. Examples of the
The Exxon Valdez Oil Spill 447
relation between w and the PAH composition of weathered oil are depicted in Fig. 5.14. Values of w < 2 correspond with PAH compositions dominated by naphthalenes, values of 2 < w < 6 by three-ring PAH (especially phenanthrenes), and values of > 6 by four-ring PAH (especially chrysenes). These ranges also correspond with oil weathering states that may be categorized as light, moderate, and heavy, with viscosities that increase from readily mobile fluids to slightly mobile fluids to immobile asphalt-like crusts. Using this weathering scale, analysis of oiled sediments and biota during the first few years following the spill showed values that range from near zero to >10 each year, with only a very weak trend of increased values with time (Short and Heintz, 1997). This is because the oil was distributed among an array of environments, from surface layers of heavily weathered oil to compact subsurface deposits that are little changed 12 years after deposition (Short et al., 2004a). These persistent subsurface PAH sources were potentially available to biota that associate with the intertidal in those portions of PWS that were most heavily oiled initially, including clams, pink salmon, sea otters, and some sea ducks.
5.5. Effects of the Spill on Aquatic Organisms The spill affected organisms and their communities acutely (within days and weeks), short-term (up to 2 years), and long-term (up to 15 years after the spill), and this subsection is organized into those three periods of effect. Effects ranged from severe and immediate toxicity and smothering from large doses of oil and hypothermia, to longer-term chronic effects and indirect effects. The effects of the spill on intertidal communities, fish, birds, and mammals varied depending on the severity and length of exposure. Within days and weeks, surface oil combined with BTEX compounds in the water and air and killed large numbers of animals over a huge area. The late winter and early spring are very stressful times for many animals as their stored energy has been depleted by the long winter and food is not yet very abundant prior to the start of the production season. Many organisms were therefore living on the energetic edge in the early spring of 1989, emerging from a cold winter and beginning migration and reproductive cycles. For the next few years, weathered oil was still present in habitats and in some organisms that had been initially exposed to oil. These effects, which we refer to as short-term effect, although not immediate, were similar to those seen in many previous oil spills where response and recovery takes time. Finally, the persistence of bioavailable oil in beach sediments resulted in long-term effects, both direct and indirect, that had been rarely (if ever) documented in previous oil spills. Some species recovered relatively quickly; others were affected for years following the spill. Full recovery, as best as we can determine, of all the aquatic species has
448
Long-Term Ecological Change in the Northern Gulf of Alaska
Figure 5.14: Changes in polynuclear aromatic hydrocarbon composition as Exxon Valdez oil weathers, summarized by weathering coefficient w, as estimated by the model of Short and Heintz (1997). Letters on bottom refer to the individual PAH and numbers to the number of carbon atoms attached to the aromatic ring. For example No = naphthalene, N1 = C1 naphthalene, etc. BP = biphenyl, AY = acenaphthaylene, AE = acenaphthene, F = flourene, D = dibenzanthracene, P = phenanthrene, AN = anthracene, FLO = flouranthene, BA = benzanthracene, C = chrysene.
The Exxon Valdez Oil Spill 449
still not happened at the time of this writing, 15 years after the spill. Here, we discuss why some species in some environments were more affected than others. First, we describe the initial mortality among seabirds, sea otters, and seals due to direct oil contact, inhalation, and coating. Then we turn our attention to animals such as Pacific herring and pink salmon juveniles that were exposed to oil in the short term (up to 2 years). This also includes a summary of the mortality among intertidal life, which was largely the result of smothering and cleanup mortality. Finally, we focus on long-term effects (>2 years) on pink salmon embryos, mussels, sea otters, and Harlequin ducks, where lingering effects were tied to lingering oil in the intertidal. Species that live, spawn, or forage in the intertidal are vulnerable to long-term effects from lingering oil. An overall summary of the biological effects of the spill documented by government studies is presented in Appendix A. Much of this information is summarized in the volume from the 1993 symposium (Rice et al., 1996). As mentioned above, not all of these results are discussed in this section; rather, our emphasis is on the longer term effects of the spill.
5.5.1. Acute Effects The initial effect of the spill was partially reflected in the estimated death toll: 250,000 seabirds, 2800 sea otters, 300 harbor seals, 250 bald eagles, and potentially billions of fish larvae such as pink salmon and herring (Table 5.1). About 2000 kilometers of intertidal habitat were also inundated with oil. This estimated toll is based on carcass counts from beaches and population censuses from before and after the spill. Among the species of seabirds killed by the spill were black oystercatchers (Hemitopus bachmani), common loons (Gavia immer), common murres (Uria aalage), cormorants (Phalacrocorax spp.), Harlequin duck (Histionicus histionicus),
Table 5.1: Recorded and estimated mortalities of aquatic organisms as a result of the Exxon Valdez oil spill. Species Sea otters Harbor seals Bald eagles Seabirds Pink salmon Pacific herring
Carcasses Recovered Estimated Mortalities 994 151 35,000
2800 300 250 250,000 1,900,000 12,980,000,000
Reference Garrott et al., 1993 Frost et al., 1999 Bowman et al., 1995 Piatt and Ford, 1996 Geiger et al., 1996 Brown et al., 1996
450
Long-Term Ecological Change in the Northern Gulf of Alaska
marbled and Kittletz’s murrelets (Brachyrampus marmoratus and B. brevirostrus), and pigeon guillemot (Cepphus columba).
Birds Seabird spill damage had less to do with the amount of oil than with the proximity of slicks to large bird congregations and the foraging habits of the birds. Some seabirds swim extensively on the surface of the water, where the potential of oiling is greatest. For example, oil near murre colonies in the Gulf of Alaska resulted in large numbers of these diving seabirds being coated with oil and dying quickly; probably more than 100,000 murres died in this way in the first month after the spill. In addition, migratory seabirds aggregate and many of them may be exposed together. They may be exposed while feeding. For example, oystercatchers forage in the intertidal zone, where large quantites of oil were present on the beach surface for the first year. More than 1600 Harlequin ducks were among the quarter million seabirds killed by the Exxon Valdez oil spill, but their foraging habits throughout the intertidal zone exposed them to residual oil for years (Esler et al., 2000). These and five other species of seabirds may not have recovered 15 years after the spill (EVOSTC, 2002). Whether this is due to continuing oil exposure or to other factors is unclear. Coating of feathers and acute oil toxicity initially killed birds; and, as we shall discuss later, long-term damage was also evident in shorebirds that encountered lingering oil in their habitat. Just after the spill, preening birds ingested the oil coating their feathers. When oil coating was extensive and feathers lost their shape, cohesion, and insulating ability, many birds lost their ability to fly and, most importantly, their ability to keep warm. Without such insulation, birds quickly become hypothermic and die. The timing of the Exxon valdes spill was unfortunate for migratory birds as it occurred during the migratory season before the arrival of summer breeding populations. This affected more than just resident birds, so the spill had an effect beyond the geographic boundaries of the slick.
Sea Otters Perhaps our most persistent collective memory of the oil spill is the dead and dying sea otters. Being the only sea mammal without blubber, the otter’s dense fur and high metabolism are their only protection from the cold subarctic waters. The fur and its air bubbles also help keep them afloat. Effective insulation from the cold is lost, and an otter dies when more than 20% of its body is coated with oil (Costa and Kooyman, 1982). Prior to the spill, the estimated sea otter population in Alaska exceeded 100,000, with more than one-third living in the path of the oil spill (Bodkin et al., 1995). Nearly 1000 carcasses were recovered from the spill area, their fur matted with oil. Many more than 1000 sea otters may therefore have died and sunk, or the carcasses scavanged,
The Exxon Valdez Oil Spill 451
their carcasses never recovered (DeGange et al., 1994). These dead otters, which did not sink, had weakened lung membranes and damaged kidneys, indicating possible trouble diving for food, leading in turn to starvation. Hypothermia combined with emphysema and starvation killed an estimated total of 750–2650 sea otters during the first year after the spill (Garrott et al., 1993). In addition, 347 live sea otters were taken to rehabilitation centers in 1989, and 116 of these died at the centers from a likely combination of toxicity and shock (Rebar et al., 1995). Age-at-death data from spring carcass collections revealed that otter survival was lower in the oiled portion of the sound and declined rather than increased after 1989 (Monson et al., 2000). Mortality of recently weaned juveniles was also higher in oiled areas than in unoiled areas during the winter of 1990–1991 (Rotterman and Monnett, 1995).
Seals Harbor seals were also susceptible, for in PWS they did not avoid oil slicks, or oilcoated intertidal algal mats. In May 1989, 81% of 585 harbor seals observed in oiled areas of PWS were coated with oil, and many were heavily contaminated (Frost et al., 1994). Only 14 dead seals, mostly pups, were recovered in PWS after the spill because carcass retrieval and counts were difficult – dead seals often sink. Some of these carcasses could be from natural causes. Coupled with a 40% decline in harbor seal abundance in the 5 years before the spill, modeling of population trends was needed to estimate the mortality due to oil rather than natural factors. Using the counts from annual aerial surveys of haulout sites, Frost et al. (1994) estimated that 300 harbor seals died from the spill. That the number was not higher may be due to a survival strategy – a layer of blubber that buffers them against hypothermia. So, despite having their hair coated with oil, seals could still stay warm, providing some time for recovery from acute exposures. Harbor seals in the spill area were also, however, lethargic. They are normally wary on land and enter the ocean when people approach or when aircraft fly over them at low altitudes. In the first 2 to 3 months after the spill, many oiled seals could be approached on foot to within a few meters and were described as sick or unusually tame (Lowry et al., 1994). In contrast to this unusual behavior, interactions between oiled mother–pup pairs appeared normal – they seemed to remain bonded and pups nursed, even on heavily oiled mothers. The physical condition of pups appeared normal as well, indicating that nursing and milk provisioning were successful. Monoaromatic hydrocarbon fumes are suspected of affecting seals, but there were no direct measurements of these volatile hydrocarbons made just above the slicks where the seals breathe. Of 27 seals collected and immediately examined, all had several types of external and internal lesions that were consistent with oil vapor exposure (Spraker et al., 1994). In most cases, the lesions were mild and likely reversible. Lesions in the brain of some seals, however, may explain the disorientation and
452
Long-Term Ecological Change in the Northern Gulf of Alaska
lethargy noted in numerous heavily oiled seals in early summer (Lowry et al., 1994), and may have been responsible for the deaths of the most severely affected animals. Live oiled seals taken to rehabilitation centers did not display behavioral or clinical signs of organ dysfunction, and 15 of 18 animals survived to release (Williams et al., 1994). It was clear from field observations that many seals with oiled pelts survived and lost the oil rather quickly once they did not directly contact oil.
Sea Lions Acute oil effects were equivocal for Steller sea lions (Eumetopias jubatus) that inhabited PWS. Oil did not persist as long on the pelage of sea lions as harbor seals (Calkins et al., 1994). Although histological examination of adult sea lions did not reveal significant oil-related damage (Calkins et al., 1994), before-and-after comparisons to estimate spill mortality were not possible due to migration.
Whales Several cetacean species frequented the spill area and were even photographed in or near the oil slicks. In some cases, numbers definitely declined after the spill, but carcasses were not recovered (hence no autopsies or contaminant estimates), so the linkage to the oil spill is equivocal. Mysticete and odontocete whales are able to detect oil but apparently do not avoid it (Geraci, 1990); killer, humpback, and gray whales and Dall’s porpoises were observed in oiled waters (Harvey and Dahlheim, 1994; Matkin et al., 1994; von Ziegesar et al., 1994). The case for an oil effect on cetaceans was strongest for two pods of killer whales in PWS; the “resident” fish-eating AB pod, and the “transient” mammal-eating AT1 pod. Both pods lost about 40% of their members by 1990, an unprecedented decline for any North American pod of killer whales. When the AB pod was first sighted after the spill, seven fewer whales were seen than had been sighted in the previous year, and six more disappeared by June 1990 (Matkin et al., 1994). These losses, of about 20% per year, were much higher than any previously known natural rate, and could not be explained by emigration to another pod or formation of a secondary pod (Matkin et al., 1994). Missing individuals were primarily juveniles and reproductive females. Although killer whales in AB pod had interfered with the sablefish fishery in the mid-1980s and some of them carried bullet scars, there was no fishery in 1989, so there was no potential conflict with the fisheries in that year. Also, the dorsal fins of two AB males began to collapse in 1989, a possible sign of poor health, and these whales remained permanently disfigured (Matkin et al., 1994) (Fig. 5.15). The “transient” AT1 pod suffered similar losses in the year following the spill, but this pod eats marine mammals and does not interact with the fishery. This pod ranges further, and resightings were less frequent, so it took several years before several members were
The Exxon Valdez Oil Spill 453
Figure 5.15: Two male killer whales from AB pod of killer whales had collapsed dorsal fins. believed to be permanently missing. Other fish-eating pods in PWS and in southeast Alaska increased during the decade after the spill, indicating that there was enough to eat and that food was probably not limiting or affecting the survival of the AB pod. Therefore, it was suggested that the spill accounted for this unusually high loss (Dahlheim and Matkin, 1994; Matkin et al., 1994). Mortality in other resident fisheating killer whale pods was not unusual, but they were not observed in the spill like AB and ATl. No distressed killer whales were observed, and no carcasses were found for pathologic or toxicologic studies, because these whales usually sink after dying (St. Aubin and Geraci, 1994; Loughlin et al., 1996). If the spill were responsible for killer whale death, the exposure route may have been inhalation of hydrocarbon vapors (Matkin and Saulitis, 1997); the pod was photographed in and near slicks in the early days of the spill. Inhaled hydrocarbons may inflame mucous membranes, cause lung congestion, lead to pneumonia, and cause neurological damage and liver disorders (Geraci, 1990). High-volatile hydrocarbon concentrations (>100 ppm) or long exposures may be required to cause damage (Geraci, 1990; Matkin; Saulitis, 1997), and post-spill studies suggested that concentrations did not reach this level (Matkin and Saulitis, 1997). On the other hand, the brain pathology found in harbor seals, presumably caused by inhaling petroleum vapors, offers a plausible toxic mechanism of exposure for killer whales (St. Aubin and Geraci, 1994). Recovery of both AB and AT1 pods have been poor. The AB pod has had successful recruitment, but the rate of increase is less than other fish-eating pods in PWS,
454
Long-Term Ecological Change in the Northern Gulf of Alaska
and, after 16 years, there are still 25% fewer members than there were in 1989. The loss of adult females from these matriarchially organized family groups led to suppressed reproduction in AB pod for multiple years (Loughlin et al., 1996). This species, with a generation time of decades, will likely be one of the last species to recover from the spill. The AT1 pod has declined even further in the 16 years past the spill, indicating that other factors are now limiting recovery of this pod. This pod has not had a successful recruitment into the pod in many years and has less than 10 individuals remaining. The pod, will likely become extinct. For other cetacean species, there were even fewer observations to assess the spill effects. Changes in humpback whale abundance, calving rates, seasonal residency time, and mortality were not significant; von Ziegesar et al. (1994) concluded that there were insufficient data to detect oil-related effects. Dall’s and harbor porpoises may have been affected by the spill, but little study effort was directed toward them (Loughlin et al., 1996). Harvey and Dahlheim (1994) noted that three gray whales swimming in a moderate amount of oil appeared lethargic, while Loughlin (1994) did not detect alterations in swimming speed, direction, and breathing behavior. The number of gray whale carcasses discovered in 1989 exceeded those in other years, but was not unusual compared to other regions and could be explained by coincidental timing of the survey and migration. The cause of gray whale deaths could not be determined (Loughlin, 1994), and concentrations of potentially toxic chemicals in carcasses (PAHs, chlorinated hydrocarbons, and toxic metals) were low compared to concentrations in tissues of marine mammals feeding on higher trophic levels (Varanasi et al., 1994).
5.5.2. Short-term Effects of the Spill Beyond the initial mortality due to the physical contact and chemical toxicity of the crude oil, some seabirds, sea otters, and fishes experienced physiological or population changes for two years after the spill. Among seabirds, the black oystercatchers on Knight Island (oiled) produced fewer chicks per nest than birds on Montague Island (unoiled) in the first few years after the spill (Andres, 1997). For a few years after that, chicks gained weight more slowly and had oil in their feces on Knight Island (Andres, 1999). Common murres exhibited lower breeding success as well as delayed breeding by several weeks in colonies on oiled islands in the years immediately following the spill (Piatt and Anderson, 1996). Losses from the spill could not be separated from losses due to climatic factors, however. Populations of pigeon guillemots in oiled areas declined more than populations from unoiled areas in the year following the spill (Oakley and Kuletz, 1996). Laboratory studies suggest that oiling could have affected the energy available to river otters to find and capture prey due to subtle alterations of blood chemistry and
The Exxon Valdez Oil Spill 455
energetic fitness. River otters fed on oiled food developed anemia, had greater oxygen consumption rates during exercise, and made fewer dives to capture fish (Ben-David et al., 2001). In field studies, river otters from oiled sites two years after the spill had elevated levels of blood haptoglobin, interleukin-6, amino transferases, and creatine kinase (Duffy et al., 1994a). Otters in the oiled areas also had lower body weight, abandoned latrine sites at a greater rate (Duffy et al., 1993,1994a; Blajeski et al., 1996), had larger home territories, and a less diverse diet of fish (Bowyer et al., 1994, 1995) than otters from unoiled areas. By the spring of 1992, differences in blood chemistry were not present (Duffy et al., 1994b). The full significance of these findings from field studies has yet to be established by laboratory experimentation with the biochemical markers.
Pacific Herring The Exxon Valdez oil spill occurred at a critical time for Pacific herring (Clupea pallasi), a keystone species of great ecological and commercial importance. In late March 1989, adult herring were entering the nearshore habitat in PWS to spawn on intertidal vegetation and rocks. Eggs were spawned in early to mid-April and incubated for about 24 days (Biggs and Baker, 1993); thus, the critical embryonic development period coincided with the highest aqueous concentrations of oil (Brown et al., 1996a; Carls et al., 2002). All available evidence points to water as the primary source of oil contamination of herring eggs. Low concentrations of dissolved PAH were present in water (Neff and Stubblefield, 1995; Short and Harris, 1996), and accumulated in both transplanted and native mussels in herring-spawning areas (Brown et al., 1996a; Short and Harris, 1996). Combining site-specific estimates of exposure from mussel data and spawning surveys with laboratory effects thresholds, Carls et al., (2002) concluded that approximately 25–32% of the developing embryos were exposed to potentially damaging oil concentrations in 1989. However, the effect of lost production on the herring population could not be estimated because natural recruitment processes are poorly understood and not predictable. Herring embryos were damaged by oil exposure (Fig. 5.16). Compared to herring embryos from unoiled sites, oil-exposed embryos hatched earlier, were longer but weighed less, and had skeletal, craniofacial, and fin-fold abnormalities (Hose et al., 1996, Brown et al., 1996a; Marty et al., 1997a). The severity of skeletal abnormalities, certain types of craniofacial defects (jaw abnormalities, microphthalmia, and missing otic capsules) and the total severity index were significantly correlated to logtransformed total PAH concentrations in mussels collected nearby (Hose et al., 1996). Observed responses were confirmed with laboratory study; total PAH water concentrations between 0.4 and 9.1 ppb damaged embryos (Carls et al., 1999). In PWS, April and May 1989 concentrations detrimental to herring embryos in the laboratory were well within documented aqueous total PAH concentrations, in the low parts per billion (Neff and Stubblefield, 1995; Short and Harris, 1996).
456
Long-Term Ecological Change in the Northern Gulf of Alaska
Figure 5.16: Exposure to oil, damages developing Pacific herring embryos in characteristic ways, including edema, spinal deformities, and heart defects. The abnormal embryo was exposed to oiled water for just 4 days (Carls et al., 1999).
Pacific herring larvae were also adversely affected by oil in 1989, and this was confirmed in laboratory studies. Major oil-associated effects in larvae captured from oiled sites in spring 1989 included small size, ascites (fluid accumulation), pericardial edema (swelling around the heart), delayed development, and genetic damage (Marty et al., 1997a). Many of these larval abnormalities were likely caused by sublethal exposure to oil during early embryonic development. Microscopic lesions were consistent with decreased growth and increased mortality of herring larvae collected near oiled beaches (McGurk and Brown, 1996), and were similar to those observed in laboratory studies of herring eggs and larvae exposed to Alaska North Slope crude oil (Kocan et al., 1996; Marty et al., 1997b; Carls et al., 1999). Growth of older larvae from offshore areas also decreased throughout PWS in 1989, but a direct link to oil exposure was not possible because these larvae likely originated
The Exxon Valdez Oil Spill 457
from a mixture of oiled and unoiled areas (Norcross et al., 1996). In contrast, the frequency of genetic defects was low and jaw size was within normal limits in PWS larvae 6 years after the spill (Norcross et al., 1996). These observed effects in egg and larval herring that were still alive at the time of capture are likely lethal within a short time. Adult Pacific herring in the path of the slick in 1989 were exposed to hydrocarbons and exhibited tissue lesions consistent with such exposure, while fish from reference sites did not (Moles et al., 1993; Marty et al., 1999). Adults surface in the evening, often gulping air to recharge their swim bladders. If an oil slick were present, exposure would be expected. Fish from oiled sites had hepatic necrosis and elevated PAH concentrations (primarily naphthalenes) in their tissues (Marty et al., 1999). Naphthalenes were also preferentially accumulated in muscle tissue in laboratory exposures of adult herring to PAH in water (Carls et al., 2000). These fish also had evidence of oil metabolism (Thomas et al., 1997). Herring from oil-exposed areas had fewer nematode parasites in their body cavities than did fish from reference sites (apparently nematodes migrated into muscle tissue), and the link to sublethal oil exposure was confirmed through laboratory study (Moles et al., 1993). In later laboratory studies, exposure of wild herring to concentrations of crude oil similar to those that were encountered following the oil spill depressed immune functions and allowed expression of a viral disease, viral hemorrhagic septicemia virus (VHSV), in concert with increased liver lesions (Carls et al., 1998) (Fig. 5.17). In 1990 and 1991, fish sampled from oiled sites had neither oil-related lesions nor significant PAH concentrations in their tissue (Marty et al., 1999).
Pink Salmon Juveniles Pink salmon (Oncorhynchus gorbuscha) provide an important Gulf fishery, especially PWS, and were the most extensively studied species after the spill. The fish, which have a 2-year life cycle, go out to sea shortly after emerging from the gravel as fry, and return as adults the following summer. With 24.5–49.5 million fish caught annually (hatchery-raised and wild) prior to the spill (Sharr et al., 1995), the pink salmon resource is a mainstay of the fisheries economy in PWS (49.5 million fish in 2003). Pink salmon are also a major part of the PWS ecosystem: approximately 1 billion wild and hatchery fry enter PWS each spring, and adult returns of up to 50 million, with 1–20 million wild adults escaping into streams to spawn. Effects from oil exposure were thoroughly documented at all the life stages, often at very low oil concentrations (Fig. 5.18). At the time of the spill, fry were emerging from the spawning gravels, feeding along the shoreline, and were soon going to migrate toward the Gulf of Alaska. The primary fry migration route from PWS to the Gulf of Alaska is the same one the oil followed, the Alaska Coastal Current – through the western sound and southwest
458
Long-Term Ecological Change in the Northern Gulf of Alaska
Figure 5.17: Exposure of wild herring to concentrations of crude oil similar to those that were encountered following the oil spill, resulted in the expression of a viral disease, viral hemorrhagic septicemia virus (VHSV). There was a linear relationship between increasing oil concentration and greater prevalence of experimental herring with the virus expressed (Carls et al., 1998).
passages – thus exposing fry to oil (Carls et al., 1996a; Wertheimer and Celewycz, 1996; Willette, 1996). Despite high mixing energy and large amounts of oil released, the inherent low solubility of oil in water resulted in only low concentrations (up to 6.2 ppb) (Short and Harris, 1996). These low concentrations of oil in water were dominated by polycyclic aromatic hydrocarbons (PAH) rather than the monocyclic aromatics, which had largely evaporated into the atmosphere. Caged mussels deployed below the surface for 2–8 weeks accumulated hydrocarbons matching the composition of the spilled Exxon Valdez oil (Short and Heintz, 1997). Pink salmon fry were not killed outright, as hydrocarbon concentrations in water were far too low and declined with evaporation and dilution. However, growth in 1989 was reduced in both wild (Wertheimer et al., 1996) and hatchery-raised fry (Willette, 1996) entering the sound and migrating through oiled areas (Fig. 5.19). Wild pink salmon fry collected from nearshore waters in oiled areas between April 10 and June 26, 1989, grew at half the rate of fry from reference areas (Wertheimer and Celewycz, 1996). These differences were noted primarily in migration corridors rather than in the bays where the fry spent little time. Willette (1996) also measured lower growth in hatchery-reared pink salmon collected in oiled areas than in reference sites in 1989. These fry were coded-wire tagged and measured at release, giving very accurate determinations of marine growth (Fig. 5.20). Fry depend on rapid growth soon after emergence to escape predation (Parker, 1971; Healy, 1982); thus, indirect
Figure 5.18: Low-level exposure to oil in water reduced pink salmon embryo survival (a); caused abnormalities in emergent fry (b); and induced a mixed function oxygenase enzyme, cytochrome P4501A (c); also measured at emergence. Abnormalities observed at emergence included ascites, bulging eyes, malformed heads, short opercular plates, external hemorrhaging, mouth or jaw malformations, and deformed caudal fins. Growth of emergent fry, cultured for 6 months in clean water, was depressed by prior exposure to oil (d). Although growth was predictably depressed in most oil treatments, only the most vigorous fish survived the highest oil treatment and these survivors outgrew all others, an example of natural selection. Vertical bars are ±1 standard error (Carls et al., 2005).
460
Long-Term Ecological Change in the Northern Gulf of Alaska
Figure 5.19: In 1989, wild and hatchery pink salmon fry grew slower in oiled areas of Prince William Sound than in non-oiled areas; asterisks indicate significant differences between reference and oiled fry. Rates of growth in these same areas were not significantly different a year after the spill.
mortality likely occurred as a result of retarded growth (Geiger et al., 1996). These wild and hatchery pink salmon fry were exposed to Exxon Valdez oil in 1989, as evidenced by tissue PAH concentrations, induction of the biomarker enzyme cytochrome P4501A, tissue hydrocarbons (Carls et al., 1996a; Willette, 1996; Wiedmer et al., 1996) (Fig. 5.21), and oil globules observed in the stomachs and intestines of fry (Sturdevant et al., 1996). In 1990, fry grew at similar rates in oiled and reference portions of the sound, with no evidence of increased P4501A enzyme induction or petroleum hydrocarbons in their tissue, so oil-reduced growth of pink salmon fry in the marine environment was restricted to 1989. Growth reduction in oil-exposed fry was probably caused in part by a shift in net energy from growth to hydrocarbon metabolism (Fig. 5.22). In laboratory tests, decreased growth was observed in pink salmon fry exposed to sublethal concentrations of aromatic hydrocarbons in water (Moles and Rice, 1983) or in food (Carls et al., 1996b). The oil fractions having the greatest effect were the three- and four-ringed aromatics. Food consumption was not reduced in fry captured in oil-contaminated
The Exxon Valdez Oil Spill 461
Figure 5.20: Pink salmon juveniles with coded wire tags in their noses.
locations in PWS (Sturdevant et al., 1996), or in laboratory experiments with oiled food, but the same concentrations of oil in the water column that reduce growth are also capable of increasing the respiration rate (Thomas and Rice, 1979), suggesting that some energy for growth was diverted to hydrocarbon metabolism and metabolite excretion. Measuring population declines for pink salmon required modeling due to the tendency of pink salmon to sometimes return to streams other than the one they occupied as fry (straying). Natural straying rates in PWS are very high, sometimes as high as 50%, and could have swamped any variations in fry-to-adult survival in oiled streams (Sharp et al., 1994). Precipitous population declines among adult salmon occurred in 1992 and 1993 at the same time that plankton production was relatively low, but any linkage to long-term oil effects is speculative at best. Loss estimates were calculated using growth reductions from the field studies of Wertheimer et al. (1996) and Willette (1996). Modeling estimated these losses to be 1.9 million fewer adult pink salmon returning in 1990 due to the spill (Geiger et al., 1996). The combination of growth reductions and modeling indicates that fewer fish returned to oiled streams from the 1989 year-class even though there were very high overall returns to the region, mainly due to very good hatchery production.
Intertidal and Subtidal Communities Severe damage to most intertidal and subtidal communities lasted only about two years. The intertidal areas sustained extensive and widespread damage over hundreds of
462
Long-Term Ecological Change in the Northern Gulf of Alaska
Figure 5.21: Photomicrograph of a cross section of a 53-day-old pink salmon embryo exposed to weathered crude oil since fertilization stained for P4501A (red). The black box outlines the part of the kidney shown at high magnification in the inset. Staining is prominent in the yolk sac (y), skin (external surface), and in endothelial cells (arrowheads) of large blood vessels (v) and small sinusoids of the kidney. Staining is less intense in kidney tubules (t), and blood vessels of the gill (g) and brain (b). Also labeled are the retina (r), otolith chamber (o), pectoral fin (f), and skeletal muscle (s) (Carls et al., 2005).
The Exxon Valdez Oil Spill 463
Figure 5.22: In laboratory tests, decreased growth was observed in pink salmon fry exposed to sublethal concentrations of aromatic hydrocarbons in food (Carls et al., 1996a). Asterisks indicate significantly shorter groups.
kilometers of shoreline. Oil slicks that came ashore were often deposited in a thick layer that smothered intertidal animals and plants (Fig. 5.23). Oil was also chemically toxic to many intertidal organisms, and its cleanup was clearly damaging (Lees et al., 1996). Oiling and the cleanup changed the habitat structure, by removing the algal canopy (cover from predation and protection from dehydration), as well as the removal of fine sediments during the cleanup effort. All types of shorelines were affected by the oil, including rocky and cobble beaches, sand and gravel beaches, and marsh. However, the most common beaches and the majority of affected shorelines were rocky with boulder and cobble. During the two years following the oil spill, most of the taxa on the intertidal beaches were less numerous than on the unoiled beaches. Limpets, barnacles, mussels, and snails were all present in lower abundance and biomass on affected beaches (Highsmith et al., 1996; Stekoll et al., 1996). Many of the beaches that were cleaned lost mussels and clams; species that take time to recolonize (Lees et al., 1996). Although many of these beaches were at least partially recovered by 1992, recovery of infauna (Driskell et al., 1996) on sites that had been cleaned lagged behind recovery at oiled but uncleaned sites. Fucus gardneri, the dominant intertidal brown alga, was greatly affected (Fig. 5.24). Also, the longer-lived barnacles, Semibalanus balanoides and Balanus glandula, and
464
Long-Term Ecological Change in the Northern Gulf of Alaska
Figure 5.23: Oil slicks were deposited in a thick layer that smothered intertidal animals and plants. The most commonly affected shorelines were rocky with boulder and cobble (photo courtesy of Mark Carls, NOAA).
mussels were greatly reduced on many rocky shorelines throughout the spill area, especially where aggressive oil cleanup techniques were used (Highsmith et al., 1996; Houghton et al., 1996, 1997; Stekoll et al., 1996). The remaining Fucus lost fertility (Stekoll and Deysher, 2000), young gametophytes (juveniles), when they did settle and grow, dried out and died during low tides without the shading effect of the resident canopy, or many young plants settled and grew on barnacles and were washed off by wave action (Van Tamelen et al., 1997). Algal recruitment was reduced by tar coatings on rocks (Duncan and Hooten, 1996). The canopy loss also set in motion indirect interactions, which apparently included settlement of opportunistic algae that are normally excluded from growing and settling, settlement of a more opportunistic barnacle, Chathalamus dalli, and increased oystercatcher predation on snails (Peterson, 2001). Experimental removal of Fucus canopy in the Herring Bay region of Knight Island resulted in loss of limpets (Tectura persona) and the periwinkle (Littorina sitkana) (Highsmith et al., 1996), so both of these indirect effects were likely as well. Fucus canopy reestablishment required several years – at least until 1994 for portions of the higher intertidal zone (Stekoll and Deysher, 1996, 2000). Not only were invertebrates affected, there were significant declines in 1990 in the
The Exxon Valdez Oil Spill 465
Figure 5.24: One of the most impacted members of the intertidal community was Fucus gardneri, a dominant perennial brown alga; heavily oiled (top), clean (bottom) (photo courtesy of Mark Carls, NOAA).
466
Long-Term Ecological Change in the Northern Gulf of Alaska
numbers of small intertidal fish sampled at oiled and unoiled sites at three habitat types in PWS: sheltered rocky, exposed rocky and coarse textured (Barber et al., 1995); however, intertidal fish recovered by 1991. On finer-grained beaches, there were more oligochaetes (Stekoll et al., 1996); these annelid worms are able to feed in chronically oiled sediments and incorporate carbon from petroleum in their proteins and carbohydrates (Spies and DesMarais, 1983; Bauer et al., 1990). Harpacticoid copepod populations were also higher on some oiled beaches (Wertheimer et al., 1996). Experiments with meiofauna (animals less than 0.4 mm in size) demonstrated reduced migration and density of harpacticoids during initial colonization and migration to oiled sediments, suggesting that there could have been short-term effects on these small but important components of the fauna on soft sediment beaches (Fleeger et al., 1996). Of all the intertidal species, the one that seems not to have recovered after 1991 is the clam Protothaca staminea. As of 1997, the bivalve fauna had not recovered on beaches that had been cleaned by vigorous washing and had lost their fine sediments (Driskell et al., 1996). In a reciprocal transplant experiment (Fukuyama et al., 2000), clams transferred from an unoiled site to an oiled site grew less than other treatments. The subtidal communities (down to 20 m) were less affected by the oil spill because of the rapid loss of oil from these deeper soft sediments in the two years following the spill (Dean and Jewett, 2001; Wolfe et al., 1994). Studies around Knight Island in PWS, in the heart of the spill-affected area, showed little effect on kelps (Dean et al., 1996), some potential effects on flowering of eel grass (Dean et al., 1998), and that some amphipod crustaceans, of the families Phoxocephalidae and Isaeidae, were less numerous in oiled areas. These amphipods are known to be particularly susceptible to oil pollution. There were other differences in the abundances of subtidal organisms such as seastars (Jewett et al., 1999), but the differences were not nearly of the magnitude that occurred in the adjoining intertidal zone. The oiling doses and the physical disturbance from cleaning in the subtidal zone were never comparable to intertidal zone.
5.5.3. Long-term Effects of the Spill Pink Salmon Embryos The weight of evidence indicates that the Exxon Valdez oil spill resulted in significant damage to pink salmon in Prince William Sound. This damage lasted for as many as five years following the spill, as a result of reduced growth of the 1989 year-class (and, consequently, reduced ocean survival) coupled with reduced survival of embryos in oiled stream deltas through 1994. About 31% of the streams used as spawning habitat by pink salmon were oiled (Geiger et al., 1996). Oil-contaminated water drained from the beaches into the streams, exposing developing embryos to PAH (Wiedmer et al., 1996; Carls et al., 2003) (Figs. 5.25 and 5.26). Cytochrome P4501A
The Exxon Valdez Oil Spill 467
Figure 5.25: Buried oil on stream deltas can contaminate salmon redds by leaching into streams during ebb tide. Water that passes through oiled sediment becomes contaminated with polynuclear aromatic hydrocarbons and drains downslope during falling tides, thus transferring contaminants to incubating embryos (illustration courtesy of Christine Brodersen, NOAA). enzyme induction in pre-emergent fry was noted through 1991, indicating that embryos were exposed to PAH from persistent intertidal oil deposits in some streams for at least two years following the spill (Wiedmer et al., 1996). Embryo mortality was consistently greater in oiled streams than in reference streams through 1993 (Bue et al., 1996, 1998; Craig et al., 1995). Pink salmon exposed in the laboratory to low part-per-billion concentrations of PAH from oil-coated gravel can result in a cascade of long-term effects on marine survival and growth (Marty et al., 1997b; Heintz et al., 1999, 2000; Carls et al., 2005).
Figure 5.26: Drainage of interstitial water from banks into stream beds was demonstrated by release of fluorescent dye (Carls et al., 2003).
468
Long-Term Ecological Change in the Northern Gulf of Alaska
Long-term pink salmon egg mortality The most dramatic evidence of long-term effect of low concentrations of weathered oil occurred in the intertidal reaches of salmon streams where pink salmon eggs incubate. These effects appear to be the result of the combination of lengthy exposure coupled with the sensitivity of embryos to contaminants. Up to 75% of pink salmon in PWS deposit their eggs in the intertidal reaches of streams between mid-July and September (Helle et al., 1964) (Fig. 5.27). Eggs deposited in the intertidal gravel hatch between late October and mid-December and the larvae, termed alevins, continue to incubate in the gravel, subsisting on their yolk sac until April. When the yolk supplies are depleted in April, the fry emerge from the gravel and move immediately to the marine environment to feed. Eggs and alevins have a potential for 8–9 months of exposure to oil in the intertidal environment. There was elevated egg mortality in oiled streams from 1989 through 1993, four years after the oil spill (Bue et al., 1994, 1996, 1998). Mortality rates varied each year, but relative to reference streams, the oiled streams had significantly elevated mortality rates each year (Fig. 5.28). Embryos were hydraulically sampled from spawning gravels at 10 oiled and 15 reference streams from 1989 to 1997. At each stream, embryos were removed from 14 locations along transects within three intertidal zones and one zone above mean high water. Mortality of embryos was significantly greater in oiled streams through 1993, but not significant in the years thereafter, except 1997.
Figure 5.27: Incubating pink salmon eggs. The top center egg is infertile and was mechanically damaged about one minute before it was photographed (photo courtesy of Mark Carls, NOAA).
The Exxon Valdez Oil Spill 469
Figure 5.28: Egg mortality in oiled streams was elevated from 1989 through 1993, four years after the oil spill (Bue et al., 1994, 1996, 1998). Mortality of embryos was significantly greater in oiled streams through 1993, but not significant in the years thereafter except 1997. Significant differences are indicated by solid symbols.
470
Long-Term Ecological Change in the Northern Gulf of Alaska
Exxon researchers disagreed (Brannon and Maki, 1996), and concluded that time of spawning and collection methods employing mechanical disturbance were confounding factors for concluding that there was an effect of oil on egg mortality. However, Craig et al. (2002) continued to find oil-induced mortality when time of spawning was included as a covariate in 1991, the only year in which spawn timing in individual streams was accurately monitored. This confirmed oil as a causal factor, at least in 1991. Egg mortality was consistent with drainage of oil-contaminated water from the surrounding sediment (Carls et al., 2003), and the declining differences between oiled and unoiled streams was consistent with declining oil levels on the beaches that the streams crossed. Egg mortality was again greater in oiled streams in 1997 (Craig et al., 2002). The differences between oiled and reference streams in 1997 were primarily due to elevated mortalities in three of the ten oiled streams, and these transected the beaches with the greatest oiling. Possibly these three streams may have shifted into new zones of oil-contaminated sediment deposits. Such shifts in intertidal stream channels are not unusual (Michel and Hayes, 1994; Carls et al., 2003). The effects of oil on salmon embryo mortalities were initially surprising, because it was obvious that there were no acute exposures directly to oil; freshwater flow in stream channels at low tide prevented the direct oiling of the spawning redds. The oil stranded along the sides of these streams was not suspected as a source of oil contamination until the elevated embryo mortalities were consistently measured for several years. Murphy et al. (1999) documented the lingering oil along the banks of several streams in 1995, confirming the probable oil effects on embryos measured by Bue et al. (1996) and the elevated P4501A observed in 1989 by Weidmer et al. (1996). In 1999, a decade after the spill, lingering oil was detected in some previously oiled streams, but in low concentrations (Carls et al., 2004a). In 1999, PAH were significantly elevated in two of the six previously heavily oiled streams in a pattern again consistent with stream drainage of oil-contaminated water, and chemical fingerprinting confirmed that the lingering oil matched with the spilled Exxon Valdez oil. Concentrations of PAH in sediment and induction of cytochrome P4501A enzymes in pink salmon embryos were correlated with PAH concentrations in streams (Carls et al., 2004a) (Fig. 5.29). So, PAH was bioavailable a decade after the spill, but only in the most severely oiled stream mouths. Interstitial (below the beach surface) water flow within the stream beds is driven by a combination of tidal pumping and incoming precipitations and supplies the salmon larvae with oxygen. Flow from oiled sediments on the stream banks to salmon redds was confirmed when fluorescent tracer dyes injected into beaches during ebb tide were observed throughout most of the adjacent intertidal zone, including surface and subsurface stream water where salmon eggs incubate (Carls et al., 2003). Mean horizontal groundwater flow through intertidal gravel was rapid (4–7 m per hour). Therefore, even though oil is deposited at some distance from the stream bed, interstitial water can transport PAH dissolved from subsurface oil deposits high in the
The Exxon Valdez Oil Spill 471
Figure 5.29: Correspondence between PAH in water and sediment of an intertidal stream and P4501A enzyme induction in wild pink salmon alevins exposed as embryos (Carls et al., 2004a). These data were collected a decade after the Exxon Valdez oil spill from Sleepy Creek, a stream still affected by oil: PAH concentrations in water were measured with passive samplers known as low-density polyethylene membrane devices (PEMD). The staining product is a measure of P4501A induction (intensity × occurrence). Oil was unlikely in the uppermost zone.
472
Long-Term Ecological Change in the Northern Gulf of Alaska
intertidal zone to eggs incubating lower in the intertidal zone (Carls et al., 2003). This mechanism was applicable to the intertidal streams of PWS, where the average tidal range is 3 m and most of the lingering oil was subsurface (Short et al., 2002). Pink salmon eggs and alevins were thus exposed to oil for months, which had effects later in the life cycle and apparently on the subsequent generation, as outlined below. Stream mortality of salmon eggs was estimated on eggs removed by hydraulic pumps from the stream gravel, which can, during certain periods of development, mechanically damage (shock) a portion of the live eggs, resulting in an appearance similar to dead eggs. This collection artifact may have affected some of the results. However, estimates of potential observer bias were not sufficient to explain poorer embryo survival in oiled streams. In a modeling study designed to retrospectively explore the accuracy of the original egg classifications, potential egg misclassification errors by Bue et al. (1996, 1998) did not explain observed survival differences between oiled and reference streams (Thedinga et al., 2005). This modeling was based on experimentally determined egg misclassification rates by novice and experienced observers (Carls et al., 2004c). The best evidence of reproductive impairment in adult pink salmon from oil exposure was a controlled experiment in 1993 on pink salmon (Bue et al., 1998). Observations in 1991 of elevated embryo mortality in the freshwater sections of oiled streams above high tide, (above where the oil could have exposed incubating salmon embryos in the stream) led to questions of possible reproductive impairment of returning adults (Bue et al., 1998). Because oil contamination in this exclusively freshwater zone was unlikely, any reproductive dysfunction would have been due to the returning adults that had been exposed to oil as embryos in 1989–1990. Reproductive success of returning adults was tested in 1993 and 1994, when fertilized eggs from a series of oiled and unoiled streams were incubated in a clean hatchery environment. In 1993, the last continuous year of an oil effect on embryo mortality rates, the embryos from adults collected in oiled streams had elevated embryo mortality rates compared to embryos from adults collected from unoiled streams. In 1994, the experiment was repeated, but embryo mortality rates were the same in incubated eggs from oiled and unoiled streams. These results were consistent with field results in 1994 (Bue et al., 1998) (Fig. 5.30). Elevated oil-related mortality in the 1993 hatchery experiment paralleled elevated mortality in oiled streams observed during annual surveys using hydraulic pumping.
Laboratory studies confirm long-term effects Effects on pink salmon embryos of long-term exposure to low concentrations of PAH were documented in an extensive series of laboratory studies. These multigenerational studies were designed to determine if continued egg mortality could occur despite significant declines in oil concentrations in PWS beaches in the 1990s. Embryonic exposure to very low concentrations of PAH altered morphology and compromised reproductive capability and marine survival of pink salmon.
The Exxon Valdez Oil Spill 473
Figure 5.30: In 1993, but not 1994, mortality of embryos produced in the hatchery from adults collected in oiled streams was higher than those in unoiled streams, which was consistent with the results of field collections of embryos from oiled and unoiled streams (Bue et al., 1998). Solid circles = oiled streams; open circles = unoiled streams. Each data point represents approximately 27,000 eggs from 900 fish. Thus it served as the reference; solid symbols highlight data significantly different from the references.
474
Long-Term Ecological Change in the Northern Gulf of Alaska
The first study established the sensitivity of pink salmon embryos to long-term weathered oil exposure (Marty et al., 1997b). Larvae exposed to various concentrations of oil in gravel were sampled four weeks before, during, and after emergence and examined for histopathological abnormalities and induction of cytochrome P4501A. Some of the PAH on gravel dissolved into water, and uptake by eggs was mediated by water, not direct contact with oiled gravel. Significant adverse biological effects were measured at aqueous concentrations as low as 4.4 ppb of total PAH. Pink salmon eggs accumulate PAH rapidly to concentrations that may exceed those in water by 9000 times (Carls et al., 2004d). Maximum total PAH concentrations in tissue likely occurred within the first two weeks of embryonic development and were controlled by the capacity of eggs to accumulate PAH and declining total aqueous PAH concentrations. Clearly, the chorion was sufficiently porous for high-molecularweight PAH to pass through and did not protect against toxicity. Embryos exposed to weathered PAH had induced cytochrome P4501A, ascities (swelling due to excess fluid retention), retarded development, and increased mortality (Marty et al., 1997b; Heintz et al., 1999, 2000; Carls et al., 2005) (Fig. 5.31). The fry that emerged from oiled gravel had more yolk and hepatocellular glycogen, increased apoptosis (programmed cell death) of gonadal cells and midventral skin cells, and less
Figure 5.31: Ascities, a swelling due to excess fluid retention, was frequent in larvae exposed to weathered crude oil as embryos; control (top), oil-exposed larva (bottom) (Marty et al., 1997b).
The Exxon Valdez Oil Spill 475
food in the gastrointestinal tract than did control fish of the same age and emergence date. Histopathological lesions occurred in larvae four weeks before emergence, at emergence, and 13 days after emergence, including significant gonadal apoptosis, in turn suggesting a possible mechanism for the reproductive impairment observed in field-sampled fish by Bue et al. (1998). Several of these oil-related changes were indicative of premature emergence. The long-term sensitivity of embryos was confirmed again in subsequent trials using oil-coated gravel that had been washed for 9 months and sat in the open for an additional three months prior to the exposure tests (Heintz et al., 1999). Eggs exposed from fertilization through hatching to an initial PAH concentration of 1.0 ppb of this highly weathered oil resulted in significantly greater mortality than in controls. Sublethal effects included spinal deformities, opercular hypoplasia, and ascites. The biological effects were the same whether embryos were incubated in oiled gravel or in water effluent from oiled gravel, demonstrating that PAH exposure could be via water and that contact with oiled gravels was not necessary to produce the effects. The study results suggested that the low concentrations of weathered oil produced by the stream-side deltas could cause reproductive effects observed in PWS. These and earlier results (Moles et al., 1979) show that pink salmon embryos are far more sensitive to long-term exposure to low concentrations of heavily weathered PAH than they are to short, high-concentration exposures of water-accommodated fractions of low-molecular-weight aromatic hydrocarbons. Despite the inherent low solubility of PAH in water, toxic hydrocarbons leached from the gravel can effectively be sequestered in internal tissues of these lipid-rich embryos. Transfer of hydrocarbons from water to embryo is the result of kinetic processes rather than solubility. Transfer kinetics result in more rapid PAH accumulation by smaller embryos, such as Pacific herring because the surface-area-per-unit volume of eggs increases inversely with the egg radius and the uptake rate is directly proportional to this surface-to-volume ratio. Comparison of the adverse effects caused by exposure to less-weathered and more-weathered oil suggests that smaller PAH (e.g., naphthalenes) contributed little to the toxicity (Carls et al., 1999). Consistent with this observation, comparison of three toxic mechanisms (narcosis, aryl hydrocarbon receptor agonism, and alkyl phenanthrene toxicity) in Pacific herring and pink salmon suggested alkyl phenanthrenes explained most of the toxicity (Barron et al., 2003). The concentration of large-molecular-weight PAH, which persist longer in the environment, was similar in both less- and more-weathered exposures, with similar toxic outcomes. Developing fish embryos are clearly at risk from environmentally persistent PAH in the low parts-per-billion concentrations. Oil can also cause delayed effects on growth and marine survival (Heintz et al., 2000; Carls et al., 2005). In experiments designed to test this possibility, pink salmon embryos were exposed to water contaminated with four different concentrations of PAH, again produced by percolation of water through columns of gravel coated with
476
Long-Term Ecological Change in the Northern Gulf of Alaska
different amounts of weathered Alaska North Slope crude oil. Most surviving alevins appeared healthy, and fish from each exposure group were transferred to saltwater net pens for feeding. A delayed effect on growth was measured in juvenile salmon that survived embryonic exposure to an aqueous total PAH concentration of 18 ppb. Over 200,000 fry from the experiment were released directly to the marine environment with coded-wire tags (see Fig. 5.20). Tag-bearing fish were counted upon return to their southeast Alaska natal stream one year later. Overall marine survival rates were 0.8% (in the 19-ppb exposure group), a nearly 40% reduction compared to the 1.3% marine survival rate in control fish (Fig. 5.32). Marine survival in the 5-ppb exposure group was intermediate, 1.1%, and significantly different from the control group. In a later study, cytochrome P4501A induction was correlated with growth inhibition, other adverse responses, and lower adult returns, suggesting induction can be used to predict long-term negative responses (Carls et al., 2005). Both field and laboratory experiments show that exposure to PAH at low partper-billion concentrations impairs development of pink salmon. Embryonic toxicity was not a narcotic effect associated with acute exposure, but the result of other
Figure 5.32: Sublethal exposure of embryos to oil had delayed effects on marine survival. Asterisks indicate significantly reduced survival (from Heintz et al., 2000).
The Exxon Valdez Oil Spill 477
mechanisms operating during embryonic development. Reduced growth and survival were noted long after exposure had stopped and long after PAH was eliminated from tissue, demonstrating that long-lasting or permanent damage had occurred. Roy et al. (1999) described mutagenic effects of oil on pink salmon embryos exposed to weathered Exxon Valdez oil, confirming that weathered oil is capable of causing somatic genetic damage. Marty et al. (1997b) argue that abnormal histology of gonads in pink salmon alevins after lengthy low-level exposure to PAH might be sufficient to explain subsequent reproductive damage. The cumulative effect of these injuries could decrease population productivity. Such decreases would be from the interplay of defects acquired during embryonic exposure, the increased energy needed for hydrocarbon metabolism (cytochrome P4501A is induced in pink salmon embryos at total PAH concentrations less than 4 ppb; Carls et al., 2005), and delayed effects that decrease growth and reduce survival after exposures have ended. Not only are fewer fry produced from a set of exposed eggs (Bue et al., 1996; Heintz et al., 1999), but when coupled with a 40% reduction in marine survival, a modest-sized run that produces 10 million eggs would result in 46,000 fewer adult fish returning if those eggs were exposed to 19 ppb total PAH (Rice et al., 2001). Such concentrations were still present in the interstitial waters of a number of salmon streams in 1995 (Murphy et al., 1999). Fortunately, by 1999, most pink salmon habitat was apparently recovering or had recovered (Carls et al., 2004a).
Mussels Mussels stabilize sediment, provide physical structure for a host of intertidal organisms, and are food for various predators, so, recognizing their ecological value, aggressive shoreline cleanup with hot-water washes and mechanical disturbance were not used in these areas (Babcock et al., 1996; Mearns, 1996). As a result, oil was retained longer than expected in the intertidal and was bioavailable to both mussels (Mytilus trossulus) and their predators for over a decade (Fig. 5.33) (Babcock et al., 1996; Carls et al., 2001). The slow leakage of residual oil from uncleaned mussel beds along with oil from other areas of western PWS had biological consequences (Fig. 5.34). In a study of heavily oiled mussel beds, oil was estimated to persist in the sediment underlying the beds for 5–30 years (Carls et al., 2001), a timescale consistent with other spill experiences (Teal and Howarth, 1984; White et al., 2005). Physiological stress due to chronic oil exposure was observed in mussels and other bivalves (Protothaca staminea and Mya arenaria) 5–11 years after the Exxon Valdez spill (Thomas et al., 1999; Fukuyama et al., 2000; Downs et al., 2002). Bivalve mollusks are good indicators of local petroleum hydrocarbon contamination because they do not move and have a high capacity to bioaccumulate hydrocarbons with little effect on metabolism (Fossato and Canzonier, 1976; Vandermeulen and
478
Long-Term Ecological Change in the Northern Gulf of Alaska
Figure 5.33: Exxon Valdez oil persisted for many years in mussel beds (Carls et al., 2001).
Penrose, 1978; Stegeman, 1985; Livingstone et al., 1989). Mussels were often sampled to determine site-specific biological availability of EVOS oil. For example, Brown et al. (1996a,b) used PAH concentrations in mussels as surrogates for Pacific herring egg exposure at specific spawning sites, and Wertheimer et al. (1994) compared hydrocarbons in mussels to those in migrating pink salmon fry to interpret their
Figure 5.34: Possible mechanism for movement of oil trapped in sediment into mussels.
The Exxon Valdez Oil Spill 479
results on sublethal effects on fry. The National Mussel Watch Program exploits the ability of these bivalves to act as environmental samplers (O’Connor, 2002). Habitat type determined availability of oil to the intertidal community, and, in some habitats, the oil remained for years. Persistence of oil in mussels was short (months) on shorelines not coated by oil and where mussels were only exposed to oil dissolved in water (Carls et al., 2002). Persistence of oil in mussels was long (6–10 years or more) on beaches where oil remained in soft sediment (Babcock et al., 1996; Carls et al., 2004b; Page et al., 2005). In a study of heavily oiled beds, total PAH concentrations in mussels from 10 of 23 beds sampled remained above background concentrations in 1995 (Carls et al., 2001). A decade after the spill, total PAH concentrations in mussels were indistinguishable from background in 11 of the 12 of these beds (Carls et al., 2004b). This indicator of declining biological availability of oil, hence habitat recovery, occurred despite lingering oil in sediment. The general decline of oil concentrations in the upper sediment shows that this reservoir of oil is being depleted, allowing habitat recovery (Carls et al., 2004b) (Fig. 5.35), and decreasing risks to higher level consumers. The persistence of oil in intertidal sediment, mussel beds, and associated fauna may help explain the decadal exposure of predators such as Harlequin ducks
Figure 5.35: Total PAH concentrations in mussels typically dropped in the decade after the Exxon Valdez spill, as was the case in Herring Bay, where oil retained by intertidal sediment served as the contaminant reservoir. Data illustrated are means ± standard error; horizontal dashed lines indicate estimated background concentrations (B).
480
Long-Term Ecological Change in the Northern Gulf of Alaska
(Histionicus histionicus) and sea otters (Enhydra lutris) to hydrocarbons. Biochemical indicators of exposure to PAH persisted in sea otters and Harlequin ducks inhabiting oiled areas for at least 12 years (Bodkin et al., 2002; Esler et al., 2002). Both species forage in nearshore environments, and exposure to oiled sediment during foraging and consumption of oil-exposed fauna are likely the principal routes of exposure for these predators. We do not know the relative contributions of oil in mussels on other prey species or oil in sediment to predator exposure, but recognize that consumption of oiled mussels is not likely the sole route of exposure because PAH concentrations in mussels generally declined to background before exposure ceased in predators.
Sea Otters and Harlequin Ducks Sea otters and Harlequin ducks had not fully recovered from the oil spill as of 2003, long after many species have been declared either recovered or on the road to recovery. In addition to suffering heavy losses in 1989, populations of these two species have been chronically exposed to oil for up to 13 years after the spill in western PWS. Sea otters still have not regained pre-spill abundances in the areas around Knight Island in PWS, and population effects on Harlequin ducks are evident still – from poor overwinter survival of adult females in the oiled western PWS (Esler et al., 2000, 2002). Their poor recovery is probably due to foraging for intertidal invertebrates in oiled sediments. The elevated liver enzymes (P4501A) in both sea otters and Harlequin ducks indicate chronic oil exposure in the Knight Island area, where the adverse population effects continue to be observed. As of 2003, P4501A induction levels in oiled areas declined to reference levels for Harlequin ducks (D. Esler, personal communication). Enzyme levels are also approaching parity for sea otters, which may signify imminent sea otter population recovery from spill effects (B. Ballachey and J. Bodkin, personal communications). Sea otters The PWS sea otter populations was reduced by as much as 18% by the Exxon Valdez spill and did not recover as quickly as expected (Fig. 5.36). The initial losses were predictable because the surface dwelling otters were very vulnerable to spilled oil. The lack of recovery in the hardest hit areas of PWS, around northern Knight Island, however, was unexpected. Their poor recovery in some areas of the sound through the mid-1990s stimulated a long-running research program to determine why. Recovery rates of sea otter populations within PWS varied significantly, and were poorest in the hardest hit areas such as northern Knight Island. Recovery of sea otter populations averaged about 4% per year after 1993 throughout the western (oiled) portion of the sound (Bodkin et al., 1999), and by 2000, the number of sea otters in the spill area increased by about 600 to nearly 2700 (Bodkin et al., 2002). However, for
The Exxon Valdez Oil Spill 481
Figure 5.36: Sea otter population sizes in Western Prince William Sound (WPWS) and the heavily oiled northern Knight Island. The estimated size pre-spill 1989, is indicated by the red bar (Bodkin et al., 2002).
northern Knight Island, population numbers remained static; averaging half the population size they had been prior to the spill and no signs of recovery were evident for at least a decade after the spill (Dean et al., 2000). On a smaller scale, Herring Bay on Knight Island had a minimal population of about 38 animals (38 oiled carcasses were recovered in Herring Bay in 1989), yet there are no resident sea otters using the bay, even 15 years after the spill. This contrasted sharply with the 10% per year recovery rate experienced after the Russian trade in sea otter pelts stopped in the nineteenth century (Bodkin et al., 1999) (see also Section 4.9), and with the population at a unoiled location in the sound, Montague Island, which doubled in the three years between 1995 and 1998. The doubling was probably largely due to immigration from the large population in the nearby Copper River Delta. The poorest recovery occurred at the sites where initial oiling and mortality was greatest, but the reasons for these lingering effects were unclear. Continuing oil exposure was not obvious, and it was difficult to understand how the exposures in 1989–1990 were still having an effect in the mid-1990s. Modeling indicated that survival of sea otters, even including those born after 1989 in oiled portions of the sound, declined following the spill (Monson et al., 2000). While reproduction was not impaired following the spill, juvenile mortality was relatively high (Monson et al., 2000), and a disproportionately higher number of prime-age sea otters died in western PWS after the spill than before (Monson et al., 2000). Food, or more specifically, caloric intake, was thought to possibly be the constraint. Sea otters operate energetically close to the edge, and survive in cold water through
482
Long-Term Ecological Change in the Northern Gulf of Alaska
metabolism and heat generation, rather than through the use of fat insulation like harbor seals do. Consequently, they must consume about 20–25% of their body weight per day. Food supplies and other demographic factors were examined as possible root causes for the lack of recovery of sea otters at Knight Island (along with other apex predators), from 1996–1998 during the Nearshore Vertebrate Predator study (Peterson and Holland-Bartels, 2002). Food availability, consumption, and foraging success of sea otters at Knight Island and at unoiled Montague Island were equivalent and condition factor of sea otters at Knight Island was equal to or better (Dean et al., 2002). So, food limitation was discounted as the primary factor constraining recovery (Bodkin et al., 2002). Additional evidence of continuing Exxon Valdez oil exposure emerged from research in the late 1990s. Sea otter liver P4501A activities were elevated around northern Knight Island relative to those at Montague Island (Fig. 5.37). Several prey items such as mussels (Carls et al., 2001) and clams (Fukuyama et al., 2000; Peterson, 2001) were found to be contaminated with PAH. However, the contamination levels found in prey tissues were unlikely to lead to debilitating PAH loads in sea otters.
Figure 5.37: A measure of cytochrome P4501A activity (from mRNA) in livers of sea otters from unoiled Montague Island and oiled northern Knight Island from 1996 through 2001.
The Exxon Valdez Oil Spill 483
Surprising quantities of PAH were still present in intertidal sediments of PWS (Short et al., 2003), and much of that was detected in the lower intertidal where the mussels and clams are present (Short et al., 2004a). The highest sediment concentrations (in the hundreds of parts per thousand with liquid oil present) and the highest rates of encounter of oiled habitat were in the northern Knight Island area. This study suggested that the lingering effects in sea otters were not either from the original exposures in 1989–1990, or from contaminated food (in the parts per million), but was from the oiled habitat. Their foraging behavior, that is, digging of pits for clams, was the most probable route of oil exposure. While foraging for invertebrates to obtain large amounts of their prey, sea otters dig about 100–300 underwater pits per day. On average, about 7% of these pits are dug in the intertidal zone during flood tide. The probability of encountering oil in a pit when foraging in the northern Knight Island area is increased by this foraging behavior. There is considerable variability in the foraging of otters in intertidal zones; some never dig there (and would presumably have low exposure), while others forage as much as 50% of the time in this zone (shallow dives are easier). Although the weight of evidence supports an oil-related causation of poor recovery of sea otter populations in the northern area of PWS, sea otter recolonization also involves social factors, such as males setting up territories, which may play some as yet undefined role, perhaps interacting with oil exposure in constraining recovery.
Harlequin ducks About 1000 Harlequin ducks died in 1989 by direct exposure to the oil slick (Piatt et al., 1990); the direct mortalities in 1989 were probably the result of ducks foraging in the oiled nearshore and intertidal zone (e.g., in mussel beds). Harlequin ducks overwinter in PWS, foraging in a very limited home range (Fig. 5.38). In the spring, most birds disperse north, as far as Siberia, Alaska North Slope, and Northwestern Canada and then return to their overwintering sites with specific fidelity to these limited home ranges. There are some Harlequin ducks that are year-round PWS residents and breed in its watersheds. If the home range was oiled, then these birds were very vulnerable to the oil. The mortalities in the first years were predictable, but the continuing struggle of this species to recover in outlying years was surprising. Like the sea otters, this species was one of the targeted species in the Nearshore Vertebrate Predator studies of the mid-late 1990s. The dominant hypothesis at the start of the study was that food sources were limiting. Harlequin ducks feed directly in the intertidal zone where most of the initial oil was stranded and where oil has persisted for more than a decade. The ducks that overwinter in PWS need to feed daily to survive. Feathers are a marginal insulator at best (compared to the fat-encased bodies of seals), and survival through the winter is dependent on feeding success, particularly when temperatures are at the lowest and sufficient light to forage is very limited as well.
484
Long-Term Ecological Change in the Northern Gulf of Alaska
Figure 5.38: Julie Morse from USGS with two Harlequin ducks captured in Prince William Sound.
During the nearshore vertebrate predator studies, food supplies available to Harlequin ducks were judged to be adequate at both oiled and unoiled sites, yet tracking radio-tagged female ducks through the winter continued to document greater mortality rates in western PWS in the winter (Esler et al., 2000). During the three winters of 1995–1998, approximately 100 birds in each winter were captured in the fall, radio-tagged, and tracked weekly with over-flights; movement and survival were noted (Fig. 5.39). Given that the winter home ranges are about 1–2 km of beach, tracking was accurate. Weekly aerial surveys confirmed a slow decline of numbers in both oiled and unoiled areas as the winter progressed. Then in December–January, female birds in the oiled areas had an increased rate of mortality relative to those in unoiled areas for about a month. The overwinter survival differences between the females from oiled and unoiled areas was not huge (78% compared to 83%), but population recovery is very sensitive to female abundance. The decline in Harlequin duck numbers in oiled areas (fall survey by ADFG) between 1995 and 1997,
The Exxon Valdez Oil Spill 485
Figure 5.39: Scaled P4501A values in livers of harlequin ducks from 1998 through 2002, compared to overwinter mortality rates of females (% mortalities at top of graph).
compared to stable numbers in the unoiled areas of PWS (Rosenberg and Petrula, 1998), support the likely importance of these small differentials in overwinter mortality. Like the sea otters, elevated P4501A in ducks from the oiled areas strongly suggested a lingering exposure to oil (Esler et al., 2000; Trust et al., 2000). The SCAT survey of 2001 (Short et al., 2003), documented the area and the quantity of oil remaining in PWS, and identified the Northern Knight Island area as an area where much of the remaining oil persisted. Like the sea otters, foraging biology intersected with chemical persistence. As with sea otters, P4501A levels declined as the contaminated habitat in the northern Knight Island area diminished. By 2002–2003, overwinter survival rates and P4501A levels in Harlequin ducks in oiled and unoiled areas converged and were no longer significantly different (Esler, personal communication) (Fig. 5.39). However, in 2004–2005, there were again significant differences between oiled and unoiled areas in P4501A activities, not only in Harlequin ducks, but also in Barrrow’s goldeneye ducks (Bucephla islandica) (D. Esler, personal communication). The increased mortality rate among Harlequin ducks during the coldest period of the winter clarifies the significance and importance of indirect mechanisms of oil toxicity. The mortalities were not caused by acute narcosis due to oil exposure but possibly the additional drain of energy needed to cope with the exposure. Elevated metabolism to fuel the P4501A detoxification pathway, usually advantageous to rid the body of a toxicant, combined with the thermal stress of coping with the cold that proved to be too much of an energetic burden for the ducks. There were insufficient daylight hours for foraging during December–January in the oiled areas to compensate for the extra energetic burden (see also Fig. 3.20).
486
Long-Term Ecological Change in the Northern Gulf of Alaska
5.6. Indirect Interactions The direct mortality of so many animals and plants after this massive spill had indirect effects on other species. All animals depend on other species, as food, as prey, or for shelter. The reduction in primary or secondary consumers, reduced abundance of a top predator, or a change in competition for limited resources among predators sharing a common prey item are the most common ecological interactions. This spill that differentially affected plants and animals would induce a cascade of indirect effects as well. The reduction of keystone species such as sea otters, which so vigorously crop subtidal invertebrates, or the common brown intertidal algal species Fucus, which provides shelter for other invertebrates, has cascading effects. Indirect effects can be as important as direct trophic interactions in structuring communities. Cascading indirect effects are delayed in expression because they are mediated through changes in an intermediary species or habitat. Perhaps this was nowhere more evident than in the long recovery process of the rocky shorelines where changes in living habitat altered the ecosystem for the residents of these intertidal areas (Peterson, 2001). As discussed above, the dramatic initial loss of Fucus cover in the intertidal following the spill and subsequent cleanup triggered a cascade of indirect effects. Freeing the space on the rocks and the losses of important grazers such as limpets, periwinkles, and predators such as whelks combined to promote initial blooms of ephemeral green algae in 1989 and 1990 and an opportunistic barnacle in 1991 (Peterson, 2001). The absence of structural algal canopy lead to declines in invertebrates and recruitment of young Fucus; young Fucus is more easily dislodged during storms because the attachment to the barnacle is not as permanent as it is to rocks (van Tamelen and Stekoll, 1997). After the apparent recovery of Fucus, previously oiled shorelines exhibited another mass rockweed mortality in 1994, a cyclic instability possibly caused by simultaneous senility of a single-aged stand (Driskell et al., 2001). The general sequence of succession on rocky intertidal shores extending over a decade after the Exxon Valdez oil spill closely resembles the dynamics after the Torrey Canyon oil spill in the UK (Hawkins and Southward, 1992). Expectations of rapid recovery based on short generation times of most intertidal plants and animals proved naïve because interspecific interactions led to a sequence of delayed effects for years. The loss of more than 50% of sea otters around Knight Island in PWS apparently resulted in a greater proportion of large urchins, although the lack of pre-spill data from the oiled and unoiled sites makes the cause of this difference somewhat uncertain (Dean et al., 2000). Furthermore, the sea otter loss was apparently not enough to induce a dramatic trophic cascade. Fleeger et al. (2003) reviewed 150 papers on indirect effects of contaminants, including three on oil in the marine environment. In addition to altered trophic cascades and changes in competition, they identified changes in foraging behavior, such
The Exxon Valdez Oil Spill 487
as reduced predation rates and increased susceptibility to predation. Noted were changes in habitat use as species abundance declined and the implications when multiple stressors were present. Although the authors admit that indirect effects of contaminants are more rare in marine pelagic environments than in freshwater environments due to fewer documented trophic cascades, this too may be a function of the difficulty in studying the marine ecosystems. Certainly changes in habitat utilization have been noted for benthic fish and invertebrates when oil is present (Pinto et al., 1984; Moles et al., 1994; Moles and Stone, 2002), potentially increasing the crowding and competition in the remaining habitat. As we have seen, indirect and cascade effects did occur following the Exxon Valdez oil spill.
5.7. Expectations, Certainty, and Final Lessons from the Exxon Valdez Oil Spill The Exxon Valdez oil spill was the largest spill in U.S. waters, and the most studied spill in history. Yet, did we learn that much from this oil spill? Large oil spills are accidents, and there is seldom the kind of detailed and up-to-date pre-spill data to measure their impacts with great certainty. Yet, some established phenomena were confirmed in this spill, and new phenomena were documented as well. The short-term acute responses were well known from other spills. Oil stranded on shorelines, the massive cleaning of beaches, and the oiled carcasses of birds and sea otters, vivid images for the first summer of the spill – all were relatively predictable from past spills. As the effects extended from months to years, the conceptual models based on laboratory tests and previous oil spills proved inadequate for describing or predicting the outcome of the Exxon Valdez oil spill. The Alaskan marine environment was relatively pristine prior to the spill and understudied in ways important to determining the long-term effects of the spill. Data on non-commercially important species were mostly nonexistent, and the various ecosystem interactions and dependences were poorly understood. Climatic forcing of ecosystems grew rapidly in appreciation in the 1990s and the details of how natural forces would play out over time could only be guessed. Consequently, it proved difficult to tease apart the effects caused by the spill and other ecosystem dynamics. Consequently, there were many surprises and events that may never be fully understood. The crash of the herring population in winter 1992–1993 and the lack of recovery of the AB pod of killer whales are prime examples of outcomes in these complex systems that are not fully understood. There is a lack of evidence to directly tie either of these events to the spill, yet no other region in Alaska has suffered these crashes or the lack of recovery over an extended period of time (15 years), and no other
488
Long-Term Ecological Change in the Northern Gulf of Alaska
region has had an oil spill of this magnitude. Linkage is suggestive at best, but certainly not conclusive. The growing herring population all through the 1980s combined with the oil spill and poor plankton production in the early 1990s could have led to increased disease susceptibility, resulting in this crash. Or, the disease could have been secondary to the population crash from food shortages and growing interspecific competition. The role of the spill in either of these events will likely remain uncertain; likewise, the lack of recovery in herring and killer whales may not be understood. Recovery, or the lack of it, can be tracked, but that is a long way from understanding the complexity of factors that add to or subtract from the recovery process. Although we have not answered all the questions about the spill, we have learned a lot about the continued persistence of oil, the long-term effects, and the ecosystem into which the oil was spilled; this is the true scientific legacy of the spill. Exxon Valdez spill is the most studied large oil spill, both in intensity and duration. Although nature confounded the tracking and observed changes over time, this spill was not burdened with the layering of other spills, industrial effluents, or urban development and sewage. If this spill had occurred in Puget Sound or Chesapeake Bay, with the multitude of other influences, we would know less about its consequences. Except for the immediate catastrophic effects, it would be nearly impossible to tease out the long-term spill effects in these highly perturbed environments. For example, what environments could serve as uncontaminated comparison or reference sites? As difficult as the task was in PWS, it was easier scientifically to separate out these various factors in Alaska, and assess which were more important than others. The most significant findings are the long-term, intertidal oil persistence, and how these discontinuous subsurface patches of oil probably affected both pink salmon survival as well as the long-term recovery of sea otters and harlequin ducks. The evidence here is strong: the chemical evidence of long-term persistence of oil in these patches is linked geographically, and biochemically to these affected species. Biological availability has been demonstrated, through direct contamination and through elevated P4501A enzymes. Other causes of P4501A induction have been largely discounted. Effects at the individual level and at the population level have been documented. The linkage is supported by the observation that as the oil diminished in later years, so did the effect (for example, pink salmon survival in oiled and unoiled stream beds was equal after four years). The signal of elevated P4501A enzymes in sea otters and Harlequin ducks in the oiled areas diminished in recent years, when the oil patches also diminished. The mechanisms of long-term toxicity, with effects on fitness rather than acute narcosis, have been demonstrated in the long-term studies of released pink salmon fry that were exposed as embryos and in the overwinter mortalities of Harlequin ducks in the oiled areas when stressed with cold and short feeding periods. The controlled long-term exposures to pink salmon embryos demonstrated that effects on fitness, such as reductions in returning adults after release of the exposed fry, can occur with PAH exposures in the low parts-per-billion concentrations if the exposure is for months.
The Exxon Valdez Oil Spill 489
All of these observations are unique to the Exxon Valdez spill; all are unprecedented in scientific literature; and combined, are the best contributions to the scientific literature on long-term effects of PAH contamination. Implications from this spill are certainly applicable to other environments, including the more urbanized areas such as Puget Sound or Chesapeake Bay. The long-term persistence and effects of oil exposure on fitness may well apply to the species there. As the spill area recovered through the late 1990s, the chronic exposure and toxicity processes emerged as possible models for chronic urban pollution. The suite of toxicants are different, but the chronic and intermittent exposures are analogous to the processes in an urban waterway, and the toxicity mechanisms affecting organisms are as applicable to urban environments as they were to PWS.
5.8. The Legacy of the Exxon Valdez Oil Spill Despite the long-term damage from the Exxon Valdez oil spill, there have been some hopeful developments as well. In particular, the institutions that responded were changed by this experience, and our knowledge of the impacted ecosystem took a great leap forward. First, to deal with this disaster, it became necessary to modify the way that ecosystems are studied and assessed in Alaska. No single agency, nor any single approach, would have been sufficient to meet the challenges of assessing damage, tracking recovery, or prescribing restoration of the spill. Six government agencies joined to form the Exxon Valdez Oil Spill Trustee Council, a single body for research planning, reviews, study coordination, and funding. The council sponsored large multidisciplinary studies, pulling together the government, academic, and private scientists with the training and experience to unravel the dynamics of the spill area ecosystem. For well over a decade, they worked over a large portion of the northern Gulf of Alaska to define and understand the changes in this system. Chemists, toxicologists, physical and biological oceanographers, modelers, fisheries scientists, wildlife biologists, invertebrate biologists, and algologists worked to very high standards under often difficult and challenging conditions. Large multidisciplinary studies had the effect of breaking down the barriers of cooperation between various government agencies and universities and eventually to broader appreciation of ecosystem processes by the participating agencies. It would yield an even greater dividend if this level of cooperation could continue in the future. The second positive outcome of the spill was the dividends in human knowledge of the spill area ecosystem. Much of what appears in other sections of this book is either the direct result of studies of the spill-impacted ecosystem since 1989, or owes some debt to this extensive scientific enterprise.
490
Long-Term Ecological Change in the Northern Gulf of Alaska
The largest challenge was to assess the recovery of the sound. To do this, scientists were asked to separate the human footprint from the natural cycles of climate and change. How can we know what the population trajectories and ecological interactions of various species would have been, had the spill not happened? The brightest minds from the legal, economic, and scientific world are yet to agree on what recovery is in an ever-changing ecosystem. Much of the rest of this book is devoted to changes and forces of change in this ecosystem, aside from the spill itself. It is readily apparent that the timescale and the spatial scales of spill impact are exceeded by other forces, such as climate. The spill occurred on top of changes brought about by the expression of climate change and human harvest in the northern Gulf of Alaska. Today, the visible evidence of the spill has faded, and we have come to understand that the hydrocarbons that are so much a part of our daily lives can be more toxic and persistent than we thought. The evidence suggests that Prince William Sound has yet to completely recover. We are left to marvel at how well the ecosystem, with some notable exceptions, was able to heal. Nature’s own cycles alter the Gulf of Alaska daily and seasonally, and with time, mitigate the effects from both the earthquake and the spill. As we gain new perspective on the abilities of nature and the scientists, perhaps it is we, and not the sound, who have changed the most since 1989.
5.9. Appendices Appendix A Summary of the major studies and findings on the biological impacts from the Exxon Valdez Oil Spill. Species/Habitat
Study
Bald eagles
Carcass recovery
Comparisons*
PrS/PstS
Reproduction
O/UO, PstS
Survival
O/UO, PstS
Carcass recovery
Comments
References
151 Carcasses
Total spill mortality estimate 250
Bowman et al., 1993 Bowman et al., 1993
Population increased since spill to pre-spill numbers O < UO Chick production Bernatowicz et al., was lower in WPWS 1996 in 1989, but recovered in 1990; effect may have been due to disturbance. O = UO Survival of radio-tagged Bowman et al., eagles equivalent in 1995 WPWS and unoiled areas. Life table model suggests bald eagle population would have recovered by 1995. 9 carcasses recovered. Continued
The Exxon Valdez Oil Spill 491
Population (adults)
Findings*
492
Appendix A. Cont’d. Study
Comparisons*
Findings*
Comments
Black oystercatchers
Repro/feeding
O/UO, PstS
O
Expos/repro
O/UO, PstS
O
% pairs nesting Andres, 1997 decreased (1989–1991); decreased feeding rates. Oil in chick feces; Andres, 1999 chicks gained weight more slowly.
Clams Common loons
See intertidal communities below. Carcass recovery
Population (adults)
Common murres
References
216 Carcasses
O/U PrS/PstS
Carcass recovery
Severe declines in other loon species (>50%) Total estimated mortality: 185,000
Population (adults)
O/U PrS/PstS
Population declines from mid-1970s to 1991
Reproduction
O/U, PstS
O
No species-specific expansion factors have been specified to estimate total losses. Pre-spill data: 1972. Agler et al., 1999 Climate shifts appear to have affected fish-eating birds. About 22,200 of the Piatt and Ford, found carcasses were 1996 murres whose deaths could be attributed to the spill. Available data cannot Piatt and Anderson, distinguish losses 1996 from spill and those due to climate change. Lower breeding Piatt and Anderson,
Long-Term Ecological Change in the Northern Gulf of Alaska
Species/Habitat
var. GOA colonies
Carcass recovery
Cutthroat trout
Growth and survival
Dolly varden
Growth and survival
1996;
838 carcasses
No species-specific expansion factors have been specified to estimate total losses.
O/U, PstS E vs W PWS
O
Slower growth by 36–43% for oil exposed pop. and differences persisted 2 yrs. After spill. Survival was not different.
O/U, PstS E vs W PWS
O
Slower growth by Hepler et al., 1996 22–24% for oil exposed population for 1 yr. only. Survival was not different.
Hepler et al., 1996
Continued
The Exxon Valdez Oil Spill 493
Cormorants (red faced, pelagic and double crested)
success in oiled colonies; breeding delayed several weeks in affected colonies; evidence for these effects outside the spill region as well. Available data cannot distinguish losses from spill and those due to climate change. Return to pre-spill conditions in late 1990s.
Study
Harbor seals
Carcass recovery
Population (adults)
Harlequin ducks
Intertidal communities
Carcass recovery Oil exposure
Comparisons*
O/U; PrS/PstS
O/U, PstS E vs W PWS
Findings*
Comments
A few carcasses recovered; difficult to know if this was normal mortality or not O
U O
Female survival
O/U, PstS E vs W PWS
Population and community
O/U, PstS O
Greater induction of P4501A in ducks from oiled areas. Poorer winter survival in oiled areas for winters of 1995–1996 through 1997–1998.
Decreases in 1990 and 1991 of most major
References
Frost et al., 1994
Piatt and Ford, 1996 Trust et al., 2000
Esler et al., 2000. Also see Esler et al., 2002 for synthesis of injury and summary of population study results. Highsmith et al., 1996; Stekoll et al.,
Long-Term Ecological Change in the Northern Gulf of Alaska
Species/Habitat
494
Appendix A. Cont’d.
in spill zone
O/U, PstS Herring Bay PWS
O
Clam (Protothaca)
O/U
O
Invertebrate recruitment
O/U experimental O
Intertidal fish
O/U, PstS O
Continued
The Exxon Valdez Oil Spill 495
Fucus population
taxa (e.g., Fucus, 1996 barnacles, snails, mussels). Increases in oligochaetes. Impacts on adult Van Tamelen and population through Stekoll, 1996 1990; recovered by 1992. Reproductive capacity has not recovered by 1993. Populations in areas cleaned with hot water were more affected. Effects found on Fukuyama et al., growth in clams from 2000 an oiled transferred to an unoiled site. Limpets and litorine Duncan and snails had lower Hooten, 1996 recruitment on tarred setting plates than controls. In 1990 diversity was Barber et al., 1995 equivalent but numbers of fish were greater in unoiled areas. Effects were not detected in 1991.
496
Appendix A. Cont’d. Study
Comparisons*
Findings*
Comments
References
Subtidal communities
Population and community
O/U, PstS Herring Bay PWS
O
Jewett et al., 1999
Killer whales
Population and pod PrS-PstS, oilexposed pods
Very high mortality in AB transient pod
Marbled murrelets
Carcass recovery
528 identified carcasses
Negative effects on some amphipod populations; enhancement of some polychsetes at oiled sties. Natural variability may have influenced findings. Loss of 13 individuals in 2 years out of the original 36 members of AB pod. Pod had history of harassing fishermen and had bullet scars. Injury is circumstantial. Cause(s) of death uncertain. Total estimated mortality: 4984 (another 3143 unidentified murrelets) There was a negative relationship between boat activity and abundance that
Population
PrS/PstS Naked Island PWS
PstS < PrS in 1989 PstS >PrS in 1990
Matkin et al., 1994
Kuletz et al., 1995
Kuletz, 1996
Long-Term Ecological Change in the Northern Gulf of Alaska
Species/Habitat
effected the population in 1989. Kittzlitz’s murrelets
Carcass recovery
Pacific herring
Egg/larval impacts
Pigeon guillemots
Kuletz et al., 1995
O
Brown et al., 1996a,b
About half the deposited eggs were exposed. Larval deformities and abnormalities were elevated in oiled areas. It was estimated that larval production was reduced by about 3 orders of magnitude in WPWS.
More than 50 oiled mussel bed identified in 1993. Carcass recovery
136 carcasses in PWS
Karinen et al., 1993; Badcock et al., 1996 Total spill mortality highly uncertain. May have been as high as 1–15%
See Oakley and Kuletz, 1996
Continued
The Exxon Valdez Oil Spill 497
Mussels
O/UO, PstS
51 identified 255 total estimated carcasses. Some mortality. unidentified murrelet carcasses may have been this species.
Pink salmon
Study
Comparisons*
Findings*
Population
PrS/PstS Naked Island PWS
O
Larval mortality
O/UO, PstS
O
Larval exposure
O/UO, PstS
O >U
Comments
References
of the spill area population or much lower depending on assumptions. Post-spill population Oakley and Kuletz, around Naked Island 1996 was 43% less than in the late 1970s, but declines were area wide. Oiled areas declined greater than unoiled areas in 1989–1990. Surveys showed that Bue et al., 1996 embryonic fish (before hatching) had poorer survival in oiled streams than in unoiled stream around Knight Island in 1989–1992. In 1989 and 1990 Wiedmer, 1996 pre-emergent alevins had P450IA induction in a variety of tissues in oil-exposed, but not unoiled areas.
Long-Term Ecological Change in the Northern Gulf of Alaska
Species/Habitat
498
Appendix A. Cont’d.
O/UO, PstS
O
Juvenile exposure
O/UO, PstS
O >U
Adult survival
O/UO, PstS
O
Body mass, blood haptoglobin
O/UO, PstS
O > U haptoglobin; O
Juveniles that were Willette, 1996; tagged with coded Wertheimer and wires and recovered Celewycz, 1996 in oiled areas had less growth than those recovered in unoiled areas. Second study with untagged juveniles had similar results. Juveniles in oiled areas of PWS had induced P4501A. Juveniles captured in PWS Carls et al., had greater induction 1996a of P4501A than those captured in non-oiled areas in 1989, but 1990 Modeling study indicated Geiger et al., that about 1.6 million 1996 pink salmon did not return as adults to spawn in 1990 due to the effects of oil exposure, mainly on juvenile growth Body mass lower in Duffy et al., 1993 oiled area and haptoglobin elevated Continued
The Exxon Valdez Oil Spill 499
River otters
Juvenile growth
Study
Comparisons*
Findings*
Comments
References
Blood/enzyme chemistry
O/UO, PstS
O >U
Duffy et al., 1994a
Fecal porphyrins
O/UO, PstS
O >U
Diving physiology
O/U, laboratory exposure
O >U
Biomarker responses Blood/enzyme
O/U, laboratory exposure O/UO, PstS
O=U
Elevated blood haptoglobins, interleuken 6ir, asparate amino transferase, alanine aminotransferase and creatine kinase in 1991. Otters in oiled areas abandoned latrine sites at greater rates than in non-oiled areas. Elevated fecal porphyrins in 1990. Otters in oiled areas abandoned latrine sites at greater rates than in nonoiled areas. Oil doses of 0.5 and 0.05 g/d caused anemia and increased O2 consumption during exercise. Oil doses did not affect haptoglobins. Blood haptoglobins
O >U
Blajeski et al., 1996
Ben-David et al., 2000
Ben-David et al., 2000 Duffy et al., 1994b
Long-Term Ecological Change in the Northern Gulf of Alaska
Species/Habitat
500
Appendix A. Cont’d.
chemistry
interleuken 6ir, no longer different by 1992.
Adult mortality
PstS
Floating carcasses of rockfish reported after spill.
Sea otters
Summary of the O/UO, PrS/PstS 2650 total estimated Estimate of total major studies mortality on Prince mortality had wide and findings William Sound confidence limits. on the biological impacts from the Exxon Valdez Oil Spill. Adult mortality Intersection 500–700 total model estimate estimated mortality on Kenai Peninsula Pathology Examination of Oil caused dead oiled pulmonary otters from emphysema, the wild and followed by stress, rehabilitation gastric erasion and centers. hemorrhage, hepatic and renal lipidosis, and hepatic necrosis. Delayed mortality O/UO, PrS/PstS Life table model A greater proportion based on carcass of prime aged animals recoveries died after the spill than before.
Garrott et al., 1993
Bodkin and Utewitz, 1994 Lipscomb et al., 1993
Monson et al., 2000
Continued
The Exxon Valdez Oil Spill 501
Rockfish
Sockeye salmon
Study
Comparisons*
Findings*
Comments
Biomarker responses
O/UO, PstS
O >U
Biomarker responses O/UO, PstS
O >U
Elevated P4501A Bodkin et al., 2002 induction in oiled otters at least through 2000. Elevated liver Ballachey et al., transaminase 2001 enzymes (e.g., GGT).
Large excapements up rivers
Larger than usual numbers of fish went up several Alaskan rivers as a result of fishing closures; there was a threat of overgrazing by the developing juveniles in the lakes Scale growth Spill year related effects assessment revealed were seen for 2–4 interactive effects years in the Kenai among various brood River system and years of juvenile at Red and Akalura sockeye salmon. Lakes on Kodiak Island.
Large excapements PrS/PstS up rivers
O = oiled, UO = unoiled, PrS = pre-spill, PstS = post spill
References
Schmidt et al., 1996
Ruggeroni et al., 2003
Long-Term Ecological Change in the Northern Gulf of Alaska
Species/Habitat
502
Appendix A. Cont’d.
The Exxon Valdez Oil Spill 503
Appendix B
Properties and Composition of Fresh and Weathered Alaska North Slope Crude Oil Alaska North Slope oil is a brown-black medium crude oil with a calculated API gravity of 30.89. Following are properties and compositional data for fresh oil and for artificially weathered oil that had 30.5% of the initial mass removed by heating (data from Wang et al., 2003): Properties
Unweathered
30.5% Weathered
Sulfur, % Flash Point, °C Density (at 0 °C) Pour Point, °C Viscosity, cP (at 0 °C) Viscosity, cP (at 15 °C)
1.11 <−8 0.878 −32 23.2 11.5
1.5 115 0.946 −6 4230 625
Cumulative Weight Fraction (%) Boiling Point (°C) 40 60 80 100 120 140 160 180 200 250 300 350 400 450 500 550 600 650
0% (Weathered) 2.5 3.9 6.5 10.0 13.4 16.6 19.8 22.6 25.2 32.6 40.7 49.5 57.7 66.0 72.8 79.0 84.1 88.4
30.5% (Weathered)
0.5 7.5 18.7 31.1 42.8 54.5 64.2 72.8 79.9 85.8 Continued
504
Long-Term Ecological Change in the Northern Gulf of Alaska
Concentration (Weight %) Component Saturates Aromatics Resins Asphaltenes Waxes
0% (Weathered) 75.0 15.0 6.1 4.0 2.6
64.8 18.5 10.3 6.4 3.6 Concentration (µg/g oil)
Component
0% (Weathered)
Benzene Toluene Ethylbenzene Xylenes† C3-Benzenes‡ Total BTEX
2866 5928 1319 6187 5620 16300
Alkylated PAH
0% (Weathered)
Naphthalene C0-N C1-N C2-N C3-N C4-N Sum Phenanthrene C0-P C1-P C2-P C3-P C4-P Sum Dibenzothiophene C0-D C1-D C2-D C3-D Sum
30.5% (Weathered)
30.5% (Weathered)
0 0 0 0 30 0 Concentration (µg/g oil) 30.5% (Weathered)
261 1015 1800 1702 815 5594
167 1288 2716 2575 1174 7919
209 666 710 486 296 2368
295 932 988 707 432 3354
122 225 318 265 931
174 319 456 362 1312 Continued
The Exxon Valdez Oil Spill 505
Fluorene C0-F 142 C1-F 328 C2-F 447 C3-F 379 Sum 1295 Chrysene C0-C 48 C1-C 74 C2-C 99 C3-C 84 Sum 306 Total 10493 2-m-N/1-m-N 1.49 (3+2-m/phen)/(4-/9-+1m-phen) 0.76 4-m:2/3m:1-m-DBT 1:0.65:0.34 Other PAHs Biphenyl 134.71 Acenaphthylene 12.03 Acenaphthene 13.03 Anthracene 2.88 Fluoranthene 2.88 Pyrene 8.40 Benz(a)anthracene 4.64 Benzo(b)fluoranthene 5.14 Benzo(k)fluoranthene 0.50 Benzo(e)pyrene 10.28 Benzo(a)pyrene 2.26 Perylene 3.01 Indeno(1,2,3cd)pyrene 0.13 Dibenz(a,h)anthracene 0.63 Benzo(ghi)perylene 3.13 Total 204
197 449 647 525 1819 68 107 141 115 430 14834 1.41 0.76 1:0.65:0.34 176.9 18.43 20.02 4.55 3.81 11.92 8.11 7.49 0.70 14.74 3.69 4.42 0.25 1.02 4.91 281
506
Long-Term Ecological Change in the Northern Gulf of Alaska
References ADNR (Alaska Department of Natural Resources). 1991. “Exxon Valdez” Oil spill: Technical Services #3. GIS Mapping and Technical Services. Alaska Department of Environmental Conservation and US Fish and Wildlife Service, Anchorage, Alaska. Agler, B. A., S. J. Kendall, D. B. Irons, and S. P. Klosiewski. 1999. Declines in marine bird populations in Prince William Sound, Alaska coincident with a climatic regime shift. Waterbirds. 22: 98–103. Andres, B. A. 1997. The Exxon Valdez oil spill disrupted the breeding of black oystercatchers. J. Wildl. Manag. 61: 1322–1328. Andres, B. A. 1999. Effects of persistent shoreline oil on breeding success and chick growth in black oystercatchers. The Auk 116: 640–650. Babcock, M. M., G. V. Irvine, P. M. Harris, J. A. Cusik, and S. D. Rice. 1996. Persistence of oiling in mussel beds three and four years after the Exxon Valdez oil spill. In: Rice, S. D., R. B. Spies, D. A. Wolfe, and B. A. Wright (Eds.), Proceedings of the Exxon Valdez Oil Spill Symposium. American Fisheries Society Symposium 18. American Fisheries Society, Bethesda, Maryland, pp. 286–297. Ballachey, B. E., J. L. Bodkin, S. Howlin, K. A. Kloecker, D. H. Monson, A. H. Rebar, and P. W. Snyder. 2001. Hematology and serum chemistry of sea otters in oiled and unoiled areas of Prince William Sound, Alaska, from 1996–98. Appendix BIO-0, In: Nearshore Vertebrate Predators Final Project Report. Exxon Valdez Oil Spill Trustee Council, Anchorage, Alaska. Barber, W. E., L. L. McDonald, W. P. Erickson, and M. Vallarino. 1995. Effect of the Exxon Valdez oil spill on intertidal fish: a field study. Trans. Amer. Fish. Soc. 124: 461–476. Barron, M. G., M. G. Carls, J. W. Short, and S. D. Rice. 2003. Photoenhanced toxicity of aqueous phase and chemically dispersed weathered Alaska North Slope crude oil to Pacific herring and larvae. Environ. Toxicol. Chem. 22: 650–660. Bauer, J. B., R. B. Spies, J. S. Vogel, D. E. Nelson, and J. R. Southon. 1990. Radiocarbon evidence of fossil-carbon cycling in sediments of a nearshore hydrocarbon seep. Nature 348: 230–232. Bence, A. E., K. A. Kvenvolden, and M. C. Kennicutt, Jr. 1996. Organic geochemistry applied to environmental assessments of Prince William sound, Alaska, after the Exxon Valdez oil spill–a review. Org. Geochem. 24: 7–42. Ben-David, M., T. M. Williams, and O. A. Ormseth. 2000. Effects of oiling on exercise physiology and diving behavior of river otters: a captive study. Canad. J. Zool. 78: 1380–1390. Ben-David, M., L. K. Duffy, and R. T. Boyer. 2001. Biomarker responses in river otters experimentally exposed to oil contamination. J. Wildl. Dis. 37: 489–508. Bernatowicz, J. A., P. F. Schempf, and T. D. Bowman. 1996. Bald eagle productivity in south-central Alaska in 1989 and 1990 after the Exxon Valdez oil spill. Amer. Fish. Soc. Symp. 18: 785–797. Biggs, E. D. and T. T. Baker. 1993. Studies on Pacific herring Clupea pallasi spawning in Prince William Sound following the 1989 Exxon Valdez (oil spill 1989–1992). Draft for Natural Resource Damage Assessment Fish/Shellfish Study Number 11, Exxon Valdez Oil Spill Trustee Council, Anchorage, Alaska. Blajeski, A., L. K. Duffy, and R. T. Bowyer. 1996. Differences in fecal levels of phorphrin among river otters exposed to the Exxon Valdez oil spill. Biomarkers 1: 262–266.
The Exxon Valdez Oil Spill 507
Bodkin, J. L. and M. S. Udevitz. 1994. An intersection model for estimating sea otter mortality along the Kenai Peninsula. In: T. R. Loughlin, (Ed.), Marine mammals and the Exxon Valdez. Academic Press, San Diego, CA, pp. 81–95. Bodkin, J. L., R. J. Jameson, and J. A. Estes. 1995. Sea otters in the North Pacific Ocean, In: LaRoe, E. T., G. S. Farris, C. E. Puckett, P. D. Doran, and M. J. Mac (Eds.), Our Living Resources, US Department of the Interior, National Biological Service, Washington, DC, pp. 353–356. Bodkin, J. L., B. E. Ballachey, M. A. Cronin, and K. T. Scribner. 1999. Population demographics and genetic diversity in remnant and translocated populations of sea otters. Conserv. Biol. 13: 1378–1385. Bodkin, J. L., B. E. Ballachey, T. A. Dean, A. K. Fukuyama, S. C. Jewett, L. McDonald, D. H. Monson, C. E. O’Clair, and G. R. Van Blaricom. 2002. Sea otter population status and the process of recovery from the 1989 Exxon Valdez oil spill. Mar. Ecol. Prog. Ser. 241: 237–253. Boehm, P. D., D. S. Page, E. S. Gilfillan, W. A. Stubblefield, and E. J. Harner. 1995. Shoreline ecology program for Prince William Sound, Alaska, following the Exxon Valdez oil spill: Part 2 – Chemistry and toxicology, In: Wells, P. G., J. N., Butler, J., S. Hughes (Eds.), Exxon Valdez Oil Spill: Fate and Effects in Alaskan Waters. American Society for Testing and Materials Pub. 1219: Philadelphia, Pennsylvania, pp. 347–397. Boehm, P. D., D. S. Page, J. S. Brown, J. M. Neff, and W. A. Burns. 2004. Polycyclic aromatic hydrocarbon levels in mussels from Prince William Sound, Alaska, USA, document the return to baseline conditions. Environ. Toxicol. Chem. 23: 2916–2929. Bowman, T. D., P. F. Schempf, and J. A. Bernatowicz. 1993. Effects of the Exxon Valdez oil spill on bald eagles. Final Report, Natural Resources Damage Assessment. Bird Study 4. US Fish and Wildlife Service, Juneau, Alaska. Bowman, T. D., P. F. Schempf, and J. A. Bernatowicz. 1995. Bald eagle survival and population dynamics in Alaska after the Exxon Valdez oil spill. J. Wildl. Manage. 59: 317–324. Bowyer, R. T., J. W. Testa, J. B. Faro, C. C. Schwartz, and J. B. Browning. 1994. Changes in diets of river otters in Prince William Sound, Alaska: effects of the Exxon Valdez oil spill. Can. J. Zool.: 970–976. Bowyer, R. T., J. W. Testa, and J. B. Faro. 1995. Habitat selection and home ranges of river otters in a marine environment: effects of the Exxon Valdez oil spill. J. Mammal. 76: 1–11. Bragg, J. R., R. C. Prince, E. J. Harner, and R. M. Atlas. 1994. Effectiveness of bioremediation for the Exxon Valdez oil spill. Nature 368: 413–418. Bragg, J. R. and S. H. Yang. 1995. Clay-oil flocculation and its role in natural cleaning of Prince William Sound following the Exxon Valdez oil spill. In: Wells, P. G., J. N., Butler, J. S. Hughes (Eds.), Exxon Valdez Oil Spill: Fate and Effects in Alaskan Waters. American Society for Testing and Materials Pub. 1219: Philadelphia, Pennsylvania, pp. 178–214. Brannon, E. L. and A. W. Maki. 1996. The Exxon Valdez Oil Spill: Analysis of impacts on the Prince William Sound pink salmon. Rev. Fish. Sci. 4: 289–337. Brodersen, C., J. Short, L. Holland, M. Carls, J. Pella, M. Larsen, and S. Rice. 1999. Evaluation of oil removal from beaches 8 years after the Exxon Valdez oil spill. In: Proceedings of the Twenty-second Arctic and Marine Oil Spill Program (AMOP) Technical Seminar. Environment Canada, Ottawa, Ont., Canada, pp. 325–336.
508
Long-Term Ecological Change in the Northern Gulf of Alaska
Brown, E. D., T. T. Baker, J. E. Hose, R. M. Kocan, G. D. Marty, M. D. McGurk, B. L. Norcross, and J. Short. 1996a. Injury to the early life history stages of Pacific herring in Prince William Sound after the Exxon Valdez oil spill. In: Rice, S. D., R. B. Spies, D. A. Wolfe, and B. A. Wright (Eds.), Proceedings of the Exxon Valdez Oil Spill Symposium. American Fisheries Society Symposium 18. American Fisheries Society, Bethesda, Maryland, pp. 448–462. Brown, E. D., B. L. Norcross, and J. W. Short. 1996b. An introduction to studies on the effects of the Exxon Valdez oil spill on early life history stages of Pacific herring, Clupea pallasi, in Prince William Sound, Alaska. Can. J. Fish. Aquat. Sci. 53: 2337–2342. Bue, B. G., S. Sharr, S. D. Moffitt, and A. Craig. 1994. Assessment of injury to pink salmon embryos and fry. In: Rice, S. D., R. B. Spies, D. A. Wolfe, and B. A. Wright (Eds.), Proceedings of the 16th Northeast Pacific Pink and Chum Salmon Workshop. AK-SG-94-02. pp. 173–176. Bue, B. G., S. Sharr, S. D. Moffitt, and A. Craig. 1996. Effects of the Exxon Valdez oil spill on pink salmon embryos and pre-emergent fry. In: Proceedings of the Exxon Valdez Oil Spill Symposium. American Fisheries Society Symposium 18. American Fisheries Society, Bethesda, Maryland, pp. 619–627. Bue, B. G., S. Sharr, and J. E. Seeb. 1998. Evidence of damage to pink salmon populations inhabiting Prince William Sound, Alaska, two generations after the Exxon Valdez oil spill, Trans. Amer. Fish. Soc. 127: 35–43. Butler, P. A. 1973. Residues in fish, wildlife, and estuaries. Pest. Monit. Bull. 6: 238–262. Calkins, D. G., E. Becker, T. R. Spraker, and T. R. Loughlin. 1994. Impacts on Steller sea lions. In: T. R. Loughlin (Ed.), Marine mammals and the Exxon Valdez. Academic Press, San Diego, California, pp. 119–139. Carls, M. G., A. C. Wertheimer, J. W. Short, R. M. Smolowitz, and J. J. Stegeman. 1996a. Contamination of juvenile pink and chum salmon by hydrocarbons in Prince William Sound after the Exxon Valdez oil spill. In: Rice, S. D., R. B. Spies, D. A. Wolfe, and B. A. Wright (Eds.), Proceedings of the Exxon Valdez Oil Spill Symposium. American Fisheries Society Symposium 18. American Fisheries Society, Bethesda, Maryland, pp. 593–607. Carls, M. G., L. Holland, M. Larsen, J. L. Lum, D. G. Mortensen, S. Y. Wang, and A. C. Wertheimer. 1996b. Growth, feeding, and survival of pink salmon fry exposed to food contaminated with crude oil. In: Rice, S. D., R. B. Spies, D. A. Wolfe, and B. A. Wright (Eds.), Proceedings of the Exxon Valdez Oil Spill Symposium. American Fisheries Society Symposium 18. American Fisheries Society, Bethesda, Maryland, pp. 608–618. Carls, M. G., G. D. Marty, T. R. Meyers, R. E. Thomas, and S. D. Rice. 1998. Expression of viral hemorrhagic septicemia virus in pre-spawning Pacific herring (Clupea pallasi) exposed to weathered crude oil. Can. J. Fish. Aquat. Sci. 55: 2300–2309. Carls, M. G., S. D. Rice, and J. E. Hose. 1999. Sensitivity of fish embryos to weathered crude oil: part 1. Low-level exposure during incubation causes malformations, genetic damage and mortality in larval Pacific herring (Clupea pallasi). Env. Toxicol. Chem. 18: 481–493. Carls, M. G., J. E. Hose, R. E. Thomas, and S. D. Rice. 2000. Exposure of Pacific herring to weathered crude oil: assessing effects on ova. Environ. Toxicol. Chem. 19: 1649–1659. Carls, M. G., M. M. Babcock, P. M. Harris, G. V. Irvine, J. A. Cusick, and S. D. Rice. 2001. Persistence of oiling in mussel beds after the Exxon Valdez oil spill. Mar. Environ. Res. 51: 167–190.
The Exxon Valdez Oil Spill 509
Carls, M. G., G. D. Marty, and J. E. Hose. 2002. Synthesis of the toxicological impacts of the Exxon Valdez oil spill on Pacific herring (Clupea pallasi) in Prince William Sound, Alaska, U.S.A. Can. J. Fish. Aquat. Sci. 59: 153–172. Carls, M. G., R. E. Thomas, M. R. Lilly, and S. D. Rice. 2003. Mechanism for transport of oil-contaminated water into pink salmon redds. Mar. Ecol. Prog. Ser. 248: 245–255. Carls, M. G., S. D. Rice, G. D. Marty, and D. K. Naydan. 2004a. Pink salmon spawning habitat is recovering a decade after the Exxon Valdez oil spill. Trans. Am. Fish. Soc. Carls, M. G., P. M. Harris, and S. D. Rice. 2004b. Restoration of oiled mussel beds In: Prince William Sound, Alaska. Mar. Environ. Res. 57: 359–376. Carls M. G., J. F. Thedinga, and R. E. Thomas. 2004c. Observer classification of live, mechanically damaged, and dead pink salmon eggs. Trans. Am. Fish Soc. 133: 245–251. Carls, M. G., L. Holland, J. W. Short, R. A. Heintz, and S. D. Rice. 2004d. Monitoring polynuclear aromatic hydrocarbons in aqueous environments with passive low-density polyethylene membrane devices. Environ. Toxicol. Chem. 23: 1416–1424. Carls M. G., R. A. Heintz, G. D. Marty, and S. D. Rice. 2005. Cytochrome P4501A induction in oil-exposed pink salmon Oncorhynchus gorbuscha embryos predicts reduced survival potential. Mar. Ecol. Prog. Ser. 301: 253–265. Costa, D. P. and G. L. Kooyman. 1982. Oxygen consumption, thermoregulation, and the effects of fur oiling and washing on the sea otter, Enhydra lutris. Can. J. Zool. 60: 2761–2767. Craig A. K., M. Willette, D. G. Evans, and B. G. Bue. 1995. Injury to pink salmon embryos in Prince William Sound B Field monitoring. Restoration Report 98191A-1 Exxon Valdez Oil Spill Final Report. Alaska Department of Fish and Game, Anchorage. Craig A. K., M. Willette, D. G. Evans, and B. G. Bue. 2002. Injury to pink salmon embryos in Prince William Sound: field monitoring. Restoration Report 98191A-1 Exxon Valdez Oil Spill Final Report. Alaska. Department of Fish and Game, Anchorage. Dean, T. A., J. L. Bodkin, A. K. Fukuyama, S. C. Jewett, D. H. Monson, C. E. O’Clair, and G. R. VanBlaricom. 2002. Food limitation and recovery of sea otters following the Exxon Valdez oil spill. Mar. Ecol. Prog. Ser. 241: 255–270. Dean, T. A., M. S. Stekoll, and R. O. Smith. 1996. Kelps and oil: Effects of the Exxon Valdez oil spill on subtidal algae. in Rice, S. D., R. B. Spies, D. A. Wolfe, and B. A. Wright (Eds.), Proceedings of the Exxon Valdez Oil Spill Symposium. American Fisheries Society Symposium 18. American Fisheries Society, Bethesda, Maryland, pp. 593–607. Dean, T. A., M. S. Stekoll, S. C. Jewett, R. O. Smith, and J. E. Hose. 1998. Eelgrass (Zostera marina L.) in Prince William Sound, Alaska: effects of the Exxon Valdez oil spill. Mar. Poll. Bull. 36: 201–210. Dean, T. A., J. L. Bodkin, S. C. Jewett, D. H. Monson, and D. Jung. 2000. Changes in sea urchins and kelp following a reduction in sea otter density as a result of the Exxon Valdez oil spill. Mar. Ecol. Prog. Ser. 199: 281–291. Dean, T. A. and S. C. Jewett. 2001. Habitat specific recovery of shallow subtidal communities following the Exxon Valdez oil spill. Ecol. Appl. 11: 1456–1471. DeGange, A. R., A. M. Dorff, and D. H. Monson. 1994. Experimental recovery of sea otter carcasses at Kodiak Island, Alaska, following the Exxon Valdez oil spill. Marine Mammal Science, 10(4): 492–496. Downs C. A., G. Shigenaka, J. E. Fauth, C. E. Robinson, and A. Huang. 2002. Cellular physiological assessment of bivalves after chronic exposure to spilled Exxon Valdez crude oil using a novel molecular diagnostic biotechnology. Environ. Sci. Technol. 36: 2987–2993.
510
Long-Term Ecological Change in the Northern Gulf of Alaska
Driskell, W. B., A. K. Fukuyama, J. P. Houghton, D. C. Lees, A. J. Mearns, and G. Shigenaka. 1996. Recovery of Prince William Sound intertidal infauna from Exxon Valdez oiling and shoreline treatments, 1989 through 1992. In: Rice, S. D., R. B. Spies, D. A. Wolfe, and B. A. Wright (Eds.), Proceedings of the Exxon Valdez Oil Spill Symposium. American Fisheries Society Symposium 18. American Fisheries Society, Bethesda, Maryland, pp. 362–378. Driskell, D. C., J. L. Ruesink, D. C. Lees, J. P. Houghton, and W. N. Lindstorm. 2001. Long-term signal of disturbance: Fucus gardneri after the Exxon Valdez oil spill. Ecol. Applic. 11: 815–827. Duffy, L. K., R. T. Bowyer, J. W. Testa, and J. B. Faro. 1993. Differences in blood haptoglobin and length–mass relationships in river otters (Lutra canadensis) from oiled and nonoiled areas of Prince William Sound, Alaska. J. Wildl. Dis. 29: 353–359. Duffy, L. K., R. T. Boyer, J. W. Testa, and J. B. Faro. 1994a. Chronic effects of the Exxon Valdez oil spill on blood and enzyme chemistry of river otters. Environ. Toxicol. Chem. 13: 643–647. Duffy, L. K., R. T. Boyer, J. W. Testa, and J. B. Faro. 1994b. Evidence for recovery of body mass and haptoglobin values of river otters following the Exxon Valdez oil spill. J. Wildl. Dis. 30: 421–425. Duncan, P. B. and A. J. Hooten. 1996. Influence of residual and applied oil on intertidal algal recruitment. In: Rice, S. D., R. B. Spies, D. A. Wolfe, and B. A. Wright (Eds.), Proceedings of the Exxon Valdez Oil Spill Symposium. American Fisheries Society Symposium 18. American Fisheries Society, Bethesda, Maryland, pp. 238–248. Esler, D., J. A. Schmutz, R. L. Jarvis, and D. M. Mulcahy. 2000. Winter survival of adult female harlequin ducks in relation to history of contamination by the Exxon Valdez oil spill. J. Wildl. Manag. 64: 839–847. Esler, D., T. D. Bowman, K. A. Trust, B. E. Ballachey, T. A. Dean, S. C. Jewett, and C. E. O’Clair. 2002. Harlequin duck population recovery following the Exxon Valdez oil spill: process and constraints. Mar. Ecol. Prog. Ser. 241: 271–286. EVOS Trustee Council. 1994. Exxon Valdez Oil Spill Restoration Plan. EVOSTC Office, Anchorage, Alaska. EVOS Trustee Council. 2002. Exxon Valdez Oil Spill Restoration Plan. Update on injured Resources and Services, EVOSTC Office, 441 W. Fifth Avenue, Suite 504, Anchorage, Alaska, 29 pp. Fleeger, J. W., T. C. Shirley, M. G. Carls, and M. A. Todaro. 1996. Meiofaunal recolonization experiment with oiled sediments. In: Rice, S. D., R. B. Spies, D. A. Wolfe, and B. A. Wright (Eds.), Proceedings of the Exxon Valdez Oil Spill Symposium. American Fisheries Society Symposium 18. American Fisheries Society, Bethesda, Maryland, pp. 271–285. Fleeger, J. W., K. R. Karman, and R. M. Nisbet. 2003. Indirect effects of contaminants in aquatic ecosystems. Sci. Total Environ. 317: 207–233. Fossato, V. U. and W. J. Canzonier. 1976. Hydrocarbon uptake and loss by the mussel Mytilus edulis. Mar. Biol. 36: 243–250. Frost, K. J., L. F. Lowry, E. H. Sinclair, J. ver Hoef, and D. C. McAllister. 1994. Impacts on distribution, abundance, and productivity of harbor seals. In: T. R. Loughlin, (Ed.), Marine mammals and the Exxon Valdez. Academic Press, San Diego, CA, pp. 97–118.
The Exxon Valdez Oil Spill 511
Frost, K. J., L. F. Lowry, and J. ver Hoef. 1999. Monitoring the trend of harbor seals in Prince William Sound, Alaska, after the Exxon Valdez oil spill. Mar. Mammal Sci. 15: 494–506. Bird 11. Fukuyama, A. K., G. Shigenaka, and R. Z. Hoff. 2000. Effects of residual Exxon Valdez oil on intertidal Protothaca staminea: mortality, growth, and bioaccumulation of hydrocarbons in transplanted clams. Mar. Poll. Bull. 40: 1042–1050. Galt, J. A., W. J. Lehr, and D. L. Payton. 1991a. Fate and transport of the Exxon Valdez oil spill. Environ. Sci. Technol. 25: 202–209. Galt, J. A., G. Y. Watabayashi, D. L. Payton, and J. C. Petersen. 1991b. Trajectory analysis for the Exxon Valdez: hindcast study, In: Proceedings, 1991 International Oil Spill Conference. American Petroleum Institute Publication No. 4529. American Petroleum Institute, Washington, DC, pp. 629–634. Garrott, R. A., L. L. Eberhardt, and D. M. Burns. 1993. Mortality of sea otters in Prince William Sound following the Exxon Valdez oil spill. Mar. Mam. Sci. 9: 343–359. Geiger, H. J., B. G. Bue, S. Sharr, A. C. Wertheimer, and T. M. Willette. 1996. A life history approach to estimating damage to Prince William Sound pink salmon caused by the Exxon Valdez oil spill. In: Rice, S. D., R. B. Spies, D. A. Wolfe, and B. A. Wright (Eds.), Proceedings of the Exxon Valdez Oil Spill Symposium. American Fisheries Society Symposium 18. American Fisheries Society, Bethesda, Maryland, pp. 487–498. Geraci, J. R. 1990. Physiologic and toxic effects on cetaceans in Sea mammals and oil: confronting the risks, Geraci J. R. and St. Aubin J. (Eds.), Academic Press, San Diego, California, pp. 167–197. Gibeaut, J. C. and E. Piper. 1998. Shoreline oiling assessment of the Exxon Valdez oil spill. EVOS Restoration Project Final Report 93038. Exxon Valdez Trustee Council, Anchorage, AK. Goldberg, E. D., M. Koide, V. Hodge, A. R. Flegal, and J. Martin. 1983. US Mussel Watch: 1977–1978 results on trace metals and radionuclides. Estuarine, Coastal and Shelf Sci. 16: 69–93. Gundlach, E. R., J. Bauer, R. Bayliss, S. Provant, and M. Kendziorek 1990. Response to the Exxon Valdez by the Alaska Department of Environmental Conservation. In: M. L. Spaulding and M. Reed (Eds.), Oil spill management and legislative implications. Proceedings of the Conference, May 15–18, 1900, Newport, R. I. American Society of Civil Engineers, New York, NY, pp. 471–478. Harvey J. T. and M. E. Dahlheim. 1994. Cetaceans in oil, In: T. R. Loughlin, (Ed.), Marine Mammals and the Exxon Valdez. Academic Press, San Diego, California. 395 pp. Hawkins, S. J. and A. J. Southward. 1992. The Torrey Canyon oil spill: recovery of rocky shore communities. In: G. W. Thayer (Ed.), Restoring the Nation’s Marine Environment. Maryland Sea Grant Publication, College Park, Maryland, pp. 584–631. Hayes, M. O., J. Michel, and D. C. Noe. 1991. Factors controlling initial deposition and long-term fate of spilled oil on gravel beaches, In: Proceedings, 1991 International Oil Spill Conference. American Petroleum Institute publication No. 4529. American Petroleum Institute, Washington, DC, pp. 453−460. Hayes, M. O. and J. Michel. 1998. Evaluation of the condition of Prince William Sound shorelines following the Exxon Valdez oil spill and subsequent shoreline treatment: 1997 geomorphological monitoring survey. National Oceanic and Atmospheric Administration Technical Memorandum NOS ORCA 126, Washington, DC, 1998, 115 pp. Hayes, M. O. and J. Michel. 1999. Factors determining the long-term persistence of Exxon Valdez oil in gravel beaches. Mar. Pollut. Bull. 38: 92–101.
512
Long-Term Ecological Change in the Northern Gulf of Alaska
Healey, M. C. 1982. Timing and relative intensity of size-selective mortality of juvenile chum salmon (Oncorhynchus keta) during early sealife. Can. J. Fish. Aquatic Sci. 39: 952–957. Heintz, R. A., S. D. Rice, A. C. Wertheimer, R. F. Bradshaw, F. P. Thrower, J. E. Joyce, and J. W. Short. 2000. Delayed effects on growth and marine survival of pink salmon Oncorhynchus gorbuscha after exposure to crude oil during embryonic development. Mar. Ecol. Prog. Ser. 208: 205–216. Heintz, R. A., J. W. Short, and S. D. Rice. 1999. Sensitivity of fish embryos to weathered crude oil: Part II. Incubating downstream from weathered Exxon Valdez crude oil caused increased mortality of pink salmon (Oncorhynchus gorbuscha) embryos. Environ. Toxicol. Chem. 18: 494–503. Helle, J. H., R. S. Williamson, and J. E. Bailey. 1964. Intertidal ecology and life history of pink salmon at Olsen Creek, Prince William Sound, Alaska. US Fish Wild Serv. Spec. Sci. Rep. Fish 483. Hepler, K. R., P. A. Hanse, and D. R. Bernard. 1996. Impact of oil spilled from the Exxon Valdez on survival and growth of Dolly varden and cutthroat trout in Prince William Sound. In: Rice, S. D., R. B. Spies, D. A. Wolfe, and B. A. Wright (Eds.), Proceedings of the Exxon Valdez Oil Spill Symposium. American Fisheries Society Symposium 18. American Fisheries Society, Bethesda, Maryland, pp. 645–658. Highsmith, R. C., T. L. Rucker, M. S. Stekoll, S. M. Saupe, M. R. Lindeberg, R. N. Jenne, and W. P. Erickson. 1996. Impact of the Exxon Valdez oil spill on intertidal biota. In: Rice, S. D., R. B. Spies, D. A. Wolfe, and B. A. Wright (Eds.), Proceedings of the Exxon Valdez Oil Spill Symposium. American Fisheries Society Symposium 18. American Fisheries Society, Bethesda, Maryland, pp. 212–237. Hood, D. W. and S. T. Zimmerman. 1986. The Gulf of Alaska. Physical Environment and Biological Resources, Alaska Office, Ocean Assessments Division, Anchorage, Alaska, 665 pp. Hose, J. E., M. D. McGurk, G. D. Marty, D. E. Hinton, E. D. Brown, and T. T. Baker. 1996. Sublethal effects of the Exxon Valdez oil spill on herring embryos and larvae: morphologic, cytogenetic, and histopathological assessments, 1989–1991. Can. J. Fish. Aquat. Sci. 53: 2355–2365. Houghton, J. P., D. C. Lees, W. B. Driskell, S. C. Lindstrom, and A. J. Mearns. 1996. Recovery of Prince William Sound intertidal epibiota from Exxon Valdez oiling and shoreline treatments, 1989 through 1992. In: Rice, S. D., R. B. Spies, D. A. Wolfe, and B. A. Wright (Eds.), Proceedings of the Exxon Valdez Oil Spill Symposium. American Fisheries Society Symposium 18. American Fisheries Society, Bethesda, Maryland, pp. 379–411. Houghton, J. P., R. H. Gilmour, D. C. Lees, W. B. Driskell, S. C. Lindstrom, and A. J. Mearns. 1997. Prince William Sound intertidal biota seven years after – has it recovered?, Oil Spill Conference, paper 260. American Petroleum Institute, Washington, DC. Irvine, G. V., D. H. Mann, and J. W. Short. 1999. Multi-year persistence of oil mousse on high energy beaches distant from the Exxon Valdez spill origin. Mar. Poll. Bull. 38: 572–584. Jahns, H. O., J. R. Bragg, L. C. Dash, and E. H. Owens. 1991. Natural cleaning of shorelines following the Exxon Valdez spill, In: Proceedings, 1991 International Oil Spill Conference. American Petroleum Institute publication No. 4529. American Petroleum Institute, Washington, DC, pp. 167–176.
The Exxon Valdez Oil Spill 513
Jewett, S. C., T. A. Dean, R. O. Smith, and A. Blanchard. 1999. Exxon Valdez oil spill: impacts and recovery in the soft-bottom benthic community in and adjacent to eelgrass beds. Mar. Ecol. Prog. Ser. 185: 59–83. Karinen, J. F., M. M. Babcock, D. W. Brown, W. D. MacLeod Jr., L. S. Ramos, and J. W. Short. 1993. Hydrocarbons in intertidal sediments and mussels from Prince William Sound, Alaska, 1977–1980: characterization and probable sources. NOAA Technical Memorandum NMFS-AFSC-9. US Department of Commerce, National Oceanic and Atmospheric Administration, National Marine Fisheries Service, Alaska Fisheries Science Center. Kocan, R. M., J. E. Hose, E. D. Brown, and T. T. Baker. 1996. Pacific herring (Clupea pallasi) embryo sensitivity to Prudhoe Bay petroleum hydrocarbons: laboratory evaluation and in situ exposure at oiled and unoiled sites in Prince William Sound. Can. J. Fish. Aquat. Sci. 53: 2366–2375. Kuletz, K. J., D. K. Marks, N. L. Naslund, and M. B. Cody. 1995. Marbled murrelet activity relative to forest characteristics in the Naked Island area, Prince William Sound, Alaska, Northwestern Naturalist 76: 4–11. Kuletz, K. J. 1996. Marbled murrelet abundance and breeding activity at Naked Island, Prince William Sound, and Kachemak Bay, Alaska, before and after the Exxon Valdez oil spill. In: Rice, S. D., R. B. Spies, D. A. Wolfe, and B. A. Wright (Eds.), Proceedings of the Exxon Valdez Oil Spill Symposium. American Fisheries Society Symposium 18. American Fisheries Society, Bethesda, Maryland, pp. 770–784. Kvenvolden, K. A., F. D. Hostettler, P. R. Carlson, J. B. Rapp, C. N Threlkeld, and A. Warden. 1995. Ubiquitous tarballs with a California-source signature on the shorelines of Prince William Sound, Alaska. Environ. Sci. and Technol. 29: 2684−2694. Lees, D. C., J. P. Houghton, and W. B. Driskell. 1996. Short-term effects of several types of shoreline treatment on rocky intertidal biota in Prince William Sound. In: Rice, S. D., R. B. Spies, D. A. Wolfe, and B. A. Wright (Eds.), Proceedings of the Exxon Valdez Oil Spill Symposium. American Fisheries Society Symposium 18. American Fisheries Society, Bethesda, Maryland, pp. 329–348. Lethcoe, J. and Lethcoe, N. 1994. A history of Prince William Sound Alaska. Prince William Sound Books, Valdez, Alaska. Lipscomb, T. P., R. K. Harris, R. B. Moeller, J. M. Pletcher, R. J. Haebler, and B. E. Ballachey. 1993. Histopathologic lesions in sea otters exposed to crude oil. Vet. Pathol. 30: 1–11. Livingston, D. R., M. A. Kirchen, and A. Wiseman. 1989. Cytochrome P450 and oxidative metabolism in molluscs. Xenobiotica 19: 1041–1062. Loughlin, T. R. 1994. Tissue hydrocarbon levels and the number of cetaceans found dead after the spill, In: T. R. Loughlin (Ed.), Marine mammals and the Exxon Valdez. Academic Press, San Diego, California, pp. 359–370. Loughlin, T. R., B. E. Ballachey, and B. A. Wright. 1996. Overview of studies to determine injury caused by the Exxon Valdez oil spill to marine mammals. In: Rice, S. D., R. B. Spies, D. A. Wolfe, and B. A. Wright (Eds.), Proceedings of the Exxon Valdez Oil Spill Symposium. American Fisheries Society Symposium 18. American Fisheries Society, Bethesda, Maryland, pp. 798–808. Lowry, L. F., K. J. Frost, and K. W. Pitcher. 1994. Observations of oiling on harbor seals in Prince William Sound. In: T. R. Loughlin (Ed.), Marine Mammals and the Exxon Valdez. Academic Press, San Diego, CA.
514
Long-Term Ecological Change in the Northern Gulf of Alaska
Marty, G. D., J. E. Hose, M. D. McGurk, E. D. Brown, and D. E. Hinton. 1997a. Histopathology and cytogenetic evaluation of Pacific herring larvae exposed to petroleum hydrocarbons in the laboratory or in Prince William Sound, Alaska, after the Exxon Valdez oil spill. Can. J. Fish. Aquat. Sci. 54: 1846–1857. Marty, G. D., J. W. Short, D. M. Dambach, N. H. Willits, R. A. Heintz, S. D. Rice, J. J. Stegeman, and D. E. Hinton. 1997b. Ascites, premature emergence, increased gonadal cell apoptosis, and cytochrome P4501A induction in pink salmon larvae continuously exposed to oil-contaminated gravel during development. Can. J. of Zool. 75: 989–1007. Marty, G. D., M. S. Okihiro, E. D. Brown, D. Hanes, and D. E. Hinton. 1999. Histopathology of adult Pacific herring in Prince William Sound, Alaska, after the Exxon Valdez oil spill. Can. J. Fish. Aquat. Sci. 56: 419–426. Matkin, C. O., G. M. Ellis, M. E. Dahlheim, and J. Zeh. 1994. Status of killer whales in Prince William Sound, 1985–1992. In: T.R. Loughlin, (Ed.), Marine mammals and the Exxon Valdez. San Diego, CA: Academic Press, pp. 141–162. Matkin, C. and E. Saulitis. 1997. Killer whale, Orcinus orca. Restoration Notebook, Exxon Valdez Oil Spill Trustee Council. Anchorage, AK. 11 pp. McGurk, M. D. and E. D. Brown. 1996. Egg-larval mortality of Pacific herring in Prince William Sound, Alaska, after the Exxon Valdez oil spill. Can. J. Fish. Aquat. Sci. 53: 2343–2354. Mearns, A. J. 1996. Exxon Valdez shoreline treatment and operations: Implications for response, assessment, monitoring, and research, In: Rice, S. D., R. B. Spies, D. A. Wolfe, and B. A. Wright (Eds.), Proceedings of the Exxon Valdez Oil Spill Symposium. American Fisheries Society Symposium 18. American Fisheries Society, Bethesda, Maryland, pp. 309–328. Michaels, D. 2005. Doubt is their product. Sci. Amer. 292: 96–101. Michel, J., M. O. Hayes, W. J. Sexton, J. C. Gibeaut, and C. Henry. 1991. Trends in the natural removal of the Exxon Valdez oil spill in Prince William Sound from September 1989 to May 1990, In: Proceedings, 1991 International Oil Spill Conference: Prevention, Behavior, Control, Cleanup, American Petroleum Institute, Washington, D.C. pp. 181−187. Michel, J. and M. O. Hayes. 1994. Evaluation of the condition of Prince William Sound shorelines following the ExxonValdez, oil spill and subsequent shoreline treatment. Vol. II: 1994 Geomorphological monitoring survey. NOAA Tech Memo NOS ORCA 91. Moles, A., S. D. Rice, and S. Korn. 1979. Sensitivity of Alaskan freshwater and anadromous fishes to Prudhoe Bay crude oil and benzene. Trans. Amer. Fish. Soc. 108: 408–414. Moles, A. and S. D. Rice. 1983. Effects of crude oil and naphthalene on growth, caloric content and fat content of pink salmon juveniles in seawater. Trans. Amer. Fish. Soc. 112: 205–211. Moles, A., S. D. Rice, and M. S. Okihiro. 1993. Herring parasite and tissue alterations following the Exxon Valdez oil spill. 1993 Proceedings on the International Oil Spill Symposium, American Petroleum Institute, Washington DC, pp. 325–328. Moles, A., S. D. Rice, and B. Norcross. 1994. Non-avoidance of hydrocarbon laden sediments by juvenile flatfishes. Netherlands J. Sea Res. 32: 361–367. Moles, A. and R. P. Stone. 2002. Habitat preferences of juvenile Tanner and red king crabs: substrate and crude oil. In: Proceedings of the 2001 Lowell Wakefield Symposium: Crabs in Cold Water Regions: Biology, Management, and Economics. Fairbanks, Alaska. AK-SG-02-01, pp. 631–644.
The Exxon Valdez Oil Spill 515
Monson, D. H., D. F. Doak, B. E. Ballachey, A. M. Johnson, and J. L. Bodkin. 2000. Long-term impacts of the Exxon Valdez oil spill on sea otters, assessed through agedependent mortality patterns. Proc. Natl Acad. Sci. USA 97: 6562–6567. Murphy, M. L., R. A. Heintz, J. W. Short, M. L. Larsen, and S. D. Rice. 1999. Recovery of pink salmon spawning after the Exxon Valdez oil spill. Trans. Amer. Fish. Soc. 128: 909–918. Neff, J. F., E. H. Owens, S. W. Stoker, and D. M. McCormick. 1995. Shoreline oiling conditions in Prince William Sound following the Exxon Valdez, pp. 312–346. Neff, J. M. and W. A. Stubblefield. 1995. Chemical and toxicological evaluation of water quality following the Exxon Valdez oil spill. In: P. G. Wells, J. N. Butler, and J. S. Hughes (Eds.), Exxon Valdez oil spill: Fate and effects in Alaskan Waters. ASTM STP 1219, American Society for Testing and Materials, Philadelphia, pp. 141–177. Norcross, B. L., J. E. Hose, M. Frandsen, and E. D. Brown. 1996. Distribution, abundance, morphological condition and cytogenetic abnormalities of larval herring in Prince William Sound, Alaska following the Exxon Valdez oil spill. Can. J. Fish. Aquat. Sci. 53: 2376–2387. Oakley, K. L. and K. J. Kuletz. 1996. Population, reproduction, and foraging of pigeon guillemots at Naked Island, Alaska, before and after the Exxon Valdez oil spill, In: Rice, S. D., R. B. Spies, D. A. Wolfe, and B. A. Wright (Eds.), Proceedings of the Exxon Valdez Oil Spill Symposium. American Fisheries Society Symposium 18. American Fisheries Society, Bethesda, Maryland, pp. 759–769. O’Clair, C. E., J. W Short, and S. D. Rice. 1996. Contamination of intertidal and subtidal sediments by oil from the Exxon Valdez in Prince William Sound. In: Rice, S. D., R. B. Spies, D. A. Wolfe, and B. A. Wright (Eds.), Proceedings of the Exxon Valdez Oil Spill Symposium. American Fisheries Society Symposium 18. American Fisheries Society, Bethesda, Maryland, pp. 61–93. O’Connor, T. P. 1994. The NOAA National Status and Trends Mussel Watch Program: National monitoring of chemical contamination in the coastal United States, In: C. R. Cothern and N. P. Ross (Eds.), Environmental Statistics, Assessment and Forecasting, Lewis Pub. Boca Raton, FL, pp. 331–349. O’Connor, T. P. 2002. National distribution of chemical concentrations in mussels and oysters in the USA. Mar. Environ. Res. 53: 117–143. Owens, E. H. 1991. Shoreline conditions following the “Exxon Valdez” oil spill as of fall 1990. In: Proceedings, 14th Arctic and Marine Oil Spill Program (AMOP) Technical Seminar, Environment Canada, Ottawa, ON, pp. 579−606. Owens, E. H. 1999. SCAT–A Ten-Year Review. Proceedings 22nd Arctic and Marine Oilspill Programme (AMOP) In: Proceedings, 22nd Arctic and Marine Oil Spill Program (AMOP) Technical Seminar, Environment Canada, Ottawa, ON, pp. 337−360. Page, D. S., P. D. Boehm, G. S. Douglas, and A. E. Bence. 1995. Identification of hydrocarbon sources in the benthic sediments of Prince William Sound and the Gulf of Alaska following the Exxon Valdez oil spill, In: Wells, P. G., Butler, J. N., and Hughes, J. S., (Eds.), Exxon Valdez Oil Spill: Fate and Effects in Alaskan Waters, American Society for Testing and Materials Pub. 1219: Philadelphia, Pennsylvania, pp. 312−346. Page, D. S., P. D. Boehm, G. S. Douglas, A. E. Bence, W. A. Burns, and P. J. Mankiewicz. 1999. Pyrogenic polycyclic aromatic hydrocarbons in sediments record past human activity: A case study in Prince William Sound, Alaska. Mar. Poll. Bull. 38: 247–260. Page D. S., P. D. Boehm, J. S. Brown, J. M. Neff, W. A. Burns, and A. E. Bence. 2005. Mussels document loss of bioavailable polycyclic aromatic hydrocarbons and the return to baseline conditions for oiled shorelines in Prince William Sound, Alaska. Mar. Environ. Res. 60: 422–436.
516
Long-Term Ecological Change in the Northern Gulf of Alaska
Parker, R. R. 1971. Size selective predation among juvenile salmonid fishes in a British Columbia inlet. J. Fish. Res. Bd. Canada 28: 1503–1510. Payne, J. R., J. R. Clayton, Jr., G. D. McNabb, Jr., B. E. Kirstein, C. L. Clary, R. T. Redding, J. S. Evans, E. Reimnitz and E. W. Kempema. 1989. Oil-ice-sediment interactions during freezup and breakup. US Department of Commerce, National Oceanic and Atmospheric Administration (NOAA), OCSEAP Final Report 64: 1–382. Payne, J. R., J. R. Clayton, G. D. McNabb, Jr., and B. E. Kirstein. 1991. Exxon Valdez oil weathering fate and behavior: model predictions and field observations. In: Rice, S. D., R. B. Spies, D. A. Wolfe, and B. A. Wright (Eds.), Proceedings of the Exxon Valdez Oil Spill Symposium. American Fisheries Society Symposium 18. American Fisheries Society, Bethesda, Maryland, pp. 641–654. Payne, J. R., J. R. Clayton, Jr., and B. E. Kirstein. 2003. Oil/suspended particulate material interactions and sedimentation. Spill Sci. Technol. Bull. 8: 201–221. Peterson, C. H. 2001. The Exxon Valdez oil spill in Alaska: Acute, indirect and chronic effects on the ecosystem. Adv. Mar. Biol. 39: 1–103. Peterson, C. H., L. L. McDonald, R. H. Green, and W. P. Erickson. 2001. Sampling designs begets conclusions: The statistical basis for detection of injury to and recovery of shoreline communities after the Exxon Valdez oil spill. Mar. Ecol. Prog. Ser. 210: 255–283. Peterson, C. H. and L. Holland-Bartels. 2002. Chronic impacts of oil pollution in the sea: risks to vertebrate predators. Mar. Ecol. Progr. Ser. 241: 235–236. Piatt, J. F., C. J. Lensink, W. Butler, M. Kendziorek, and D. R. Nysewander. 1990. Immediate impact of the Exxon Valdez oil spill on marine birds. Auk 107: 387–397. Piatt, J. F. and P. Anderson. 1996. Response of common murres to the Exxon Valdez oil spill and long-term changes in the Gulf of Alaska marine ecosystem, In: Rice, S. D., R. B. Spies, D. A. Wolfe, and B. A. Wright (Eds.), Proceedings of the Exxon Valdez Oil Spill Symposium. American Fisheries Society Symposium 18. American Fisheries Society, Bethesda, Maryland, pp. 720–737. Piatt, J. F. and R. G. Ford. 1996. How many seabirds were killed by the Exxon Valdez oil spill? In: Rice, S. D., R. B. Spies, D. A. Wolfe, and B. A. Wright (Eds.), Proceedings of the Exxon Valdez Oil Spill Symposium. American Fisheries Society Symposium 18. American Fisheries Society, Bethesda, Maryland, pp. 712–719. Pinto, J. M., W. H. Pearson, and J. W. Anderson. 1984. Sediment preferences and oil contamination in the Pacific sand lance Ammodytes hexapterus. Mar. Biol. 83: 193–204. Rebar, A. H., T. P. Lipscomb, R. K. Harris, and B. E. Ballachey. 1995. Clinical and clinical laboratory correlates in sea otters dying unexpectedly in rehabilitation centers following the Exxon Valdez oil spill. Vet. Pathol. 32: 346–350. Rice, S. D., R. B. Spies, D. A. Wolfe, and B. A. Wright (Eds.), 1996. Proceedings of the Exxon Valdez Oil Spill Symposium. American Fisheries Society Symposium 18. American Fisheries Society, Bethesda, Maryland. Rice, S. D., R. E. Thomas, R. A. Heintz, A. C. Wertheimer, M. L. Murphy, M. G. Carls, J. W. Short, and A. Moles. 2001. Impacts to pink salmon following the Exxon Valdez oil spill: persistence, toxicity, sensitivity, and controversy. Rev. Fish. Sci. 9: 165–211. Rosenberg, D. H. and M. J. Petrula. 1998. Status of harlequin ducks in Prince William Sound, Alaska after the Exxon Valdez oil spill 1995–1997. Exxon Valdez Oil Spill Project Final Report, Restoration Project 97427, AK. Dept. Fish and Game, Anchorage, Alaska. Rotterman, L. M. and C. Monnett. 1995. Mortality of sea otter weanlings in eastern and western PWS, AK, during the winter of 1990–91. Exxon Valdez Oil Spill State/Federal
The Exxon Valdez Oil Spill 517
Natural Resource Damage Assessment Final Report, Marine Mammal Study No. 6–18, United States Fish and Wildlife Service, Anchorage, Alaska. Roy, N. K., J. Stabile, J. E. Seeb, C. Habicht, and I. Wirgin. 1999. High frequency of K-ras mutations in pink salmon embryos experimentally exposed to Exxon Valdez oil. Environ. Toxicol. Chem. 18: 1521–1528. Ruggerone, G. T. and D. E. Rogers. 2003. Multi-year effects of high densities of sockeye salmon spawners on juvenile salmon growth and survival: a case study from the Exxon Valdez oil spill. Fish. Res. (Amsterdam) 63: 379–392. Schmidt, D. C., K. E. Tarbox, B. E. King, L. K. Brannian, G. B. Kyle, and S. R. Carlson. 1996. Kenai River sockeye salmon: an assessment of overescapements as a cause of the decline. In: Rice, S. D., R. B. Spies, D. A. Wolfe, and B. A. Wright (Eds.), Proceedings of the Exxon Valdez Oil Spill Symposium. American Fisheries Society Symposium 18. American Fisheries Society, Bethesda, Maryland, pp. 628–638. Sharp, D., S. Sharr, and C. Peckham. 1994. Homing and straying patterns of coded wire tagged pink salmon in Prince William Sound. Proceedings of the 16th Northeast Pink and Chum Salmon Workshop. University of Alaska Sea Grant Program, Report 94-02: 77-82. Fairbanks, Alaska. Sharr. S., C. J. Peckham, D. G. Sharp, L. Peltz, J. L. Smith, T. M. Willette, D. G. Evans, and B. G. Bue. 1995. Coded wire tag studies on Prince William Sound salmon 1989–1991. Fish/Shellfish Study Number 3, Exxon Valdez Oil Spill Trustee Council Natural Resource Damage Assessment Final Report, Anchorage, Alaska. Shiu, W. Y., A. Maijanen, A. L. Y. Ng, and D. MacKay. 1988. Preparation of aqueous solutions of sparingly soluble organic substances: II. Multicomponent systems–hydrocarbon mixtures and petroleum products. Environ. Toxicol. Chem. 7: 125–137. Short, J. W. and P. M. Harris. 1996. Chemical sampling and analysis of petroleum hydrocarbons in near-surface seawater of Prince William Sound after the Exxon Valdez oil spill. In: Rice, S. D., R. B. Spies, D. A. Wolfe, and B. A. Wright (Eds.), Proceedings of the Exxon Valdez Oil Spill Symposium. American Fisheries Society Symposium 18. American Fisheries Society, Bethesda, Maryland, pp. 17–28. Short, J. W., D. M. Sale, and J. C. Gibeaut. 1996. Nearshore transport of hydrocarbons and sediments after the Exxon Valdez oil spill. In: Rice, S. D., R. B. Spies, D. A. Wolfe, and B. A. Wright (Eds.), Proceedings of the Exxon Valdez Oil Spill Symposium. American Fisheries Society Symposium 18. American Fisheries Society, Bethesda, Maryland, pp. 40–60. Short, J. W. and R. A. Heintz. 1997. Identification of Exxon Valdez oil in sediments and tissues from Prince William Sound and the northwestern Gulf of Alaska based on a PAH weathering model. Environ. Sci. and Technol. 31: 2376–2384. Short, J. W., M. R. Lindeberg, P. M. Harris, J. Maselko, and S. D. Rice. 2002. Vertical oil distribution within the intertidal zone 12 years after the Exxon Valdez oil spill in Prince William Sound, Alaska. In: Proceedings of the Twenty-fifth Arctic and Marine Oilspill Program (AMOP) Technical Seminar. Environment Canada, Ottawa, Ontario, pp. 57–72. Short, J. W., M. R. Lindeberg, P. M. Harris, J. M. Maselko, J. J. Pella, and S. D. Rice. 2003. Evaluation of oil remaining in Prince William Sound from the Exxon Valdez oil spill. Final Report, Exxon Valdez Trustee Council Project 02543. Exxon Valdez Trustee Council, Anchorage, Alaska. Short, J. W., M. R. Lindeberg, P. M. Harris, J. M. Maselko, J. J. Pella, and S. D. Rice. 2004a. Estimate of oil persisting on beaches of Prince William Sound, 12 after the Exxon Valdez oil spill. Environ. Sci. Technol. 38: 19–25.
518
Long-Term Ecological Change in the Northern Gulf of Alaska
Short, J. W., J. J. Kolak, J. R. Payne, and G. K. Van Kooten. 2004b. Geochemical investigation of hydrocarbon fate and transport in the Northern Gulf of Alaska fluvio-marine environment. Final Report, Exxon Valdez Trustee Council Project 00599. Exxon Valdez Trustee Council, Anchorage, Alaska. Spies, R. B. and D. J. DesMarais. 1983. Natural isotope study of trophic enrichment of marine benthic communities by petroleum seepage. Mar. Biol. 73: 67–71. Spraker, T. R., L. F. Lowry, and K. J. Frost. 1994. Gross necropsy and histopathological lesions found in harbor seals. In: Loughlin, T. R. (Ed.), Marine mammals and the Exxon Valdez, Academic Press, San Diego, California, pp. 281–312. St. Aubin D. J. and J. R. Geraci. 1994. Summary and conclusions, In: Loughlin, T. R. (Ed.), Marine mammals and the Exxon Valdez. Academic Press, San Diego, California, pp. 371–376. Stegeman, J. J. 1985. Benzo[a]pyrene oxidation and microsomal enzyme activity in the mussel (Mytilus edulis) and other bivalve mollusc species from the western north Atlantic. Mar. Biol. 89: 21–30. Stekoll, M. S. and L. Deysher. 1996. Recolonization and restoration of upper intertidal Fucus gardneri (Fucales, Phaeophyta) following the Exxon Valdez oil spill. Hydrobiologia, 326/327: 311–316. Stekoll, M. S., L. Deysher, R. C. Highsmith, S. M. Saupe, Z. Guo, W. P. Erickson, L. McDonald, and D. Strickland. 1996. Coastal habitat injury assessment: intertidal communities and the Exxon Valdez oil spill. In: Rice, S. D., R. B. Spies, D. A. Wolfe, and B. A. Wright (Eds.), Proceedings of the Exxon Valdez Oil Spill Symposium. American Fisheries Society Symposium 18. American Fisheries Society, Bethesda, Maryland, pp. 177–192. Stekoll, M. S. and L. Deysher. 2000. Response of the dominant alga Fucus gardneri (Silva) (Phaeophyceae) to the Exxon Valdez oil spill and cleanup. Mar. Poll. Bull. 40: 1028–1041. Sturdevant, M. V., A. C. Wertheimer, and J. L. Lum. 1996. Diets of juvenile pink and chum salmon in oiled and non-oiled nearshore habitats in Prince William Sound, 1989 and 1990, In: Rice, S. D., R. B. Spies, D. A. Wolfe, and B. A. Wright (Eds.), Proceedings of the Exxon Valdez Oil Spill Symposium. American Fisheries Society Symposium 18. American Fisheries Society, Bethesda, Maryland, pp. 578–592. Teal, A. R. 1991. Shoreline cleanup – Reconnaissance, evaluation, and planning following the Valdez oil spill. In: Proceedings, 1991 International Oil Spill Conference. American Petroleum Institute publication No. 4529. American Petroleum Institute, Washington, DC, pp. 149–152. Teal J. M. and R. W. Howarth. 1984. Oil spill studies: A review of ecological effects. Environ. Manag. 8: 27–44. Thedinga, J. F., M. G. Carls, J. M. Maselko, R. A. Heintz, and S. D. Rice. 2005. Resistance of naturally spawning pink salmon eggs to mechanical shock. Alaska Fishery Research Bulletin 11: 37–43. Thomas, R. E. and S. D. Rice. 1979. The effect of exposure temperature on oxygen consumption and opercular breathing rates of pink salmon fry exposed to toluene, naphthalene, and water-soluble fractions of Cook Inlet crude oil and No. 2 fuel oil, In: W. B. Vernberg, A. Calabrese, F. P. Thurberg, and F. J. Vernberg (Eds.), Marine Pollution: Functional Responses. Academic Press Inc., New York, pp. 39–52. Thomas, R. E., M. G. Carls, S. D. Rice, and L. Shagrun. 1997. Mixed function oxidase induction in pre- and post-spawn herring (Clupea pallasi) by petroleum hydrocarbons. Comp. Biochem. Physiol. 116C: 141–147.
The Exxon Valdez Oil Spill 519
Thomas, R. E., P. M. Harris, and S. D. Rice. 1999. Survival in air of Mytilus trossulus following long-term exposure to spilled Exxon Valdez crude oil in Prince William Sound. Comp. Biochem. Physiol. Part C 122: 147–152. Trust, K. A., D. Esler, B. R. Woodin, and J. J. Stegeman. 2000. Cytochrome P4501A induction in sea ducks inhabiting nearshore areas of Prince William Sound Alaska. Mar. Poll. Bull. 40: 397–403. Vandermeulen, J. and W. R. Penrose. 1978. Absence of aryl hydrocarbon hydroxylase (AHH) in three marine bivalves. J. Fish. Res. Bd. Can. 35: 643–647. Van Kooten, G. K., J. W. Short, and J. J. Kolak. 2002. Low-maturity Kulthieth Formation coal: A possible source of polycyclic aromatic hydrocarbons in benthic sediment of the northern Gulf of Alaska. Environ. Forensics 3: 227–242. Van Tamelen, P. G. and M. S. Stekoll. 1996. Population response of the brown alga Fucus gardneri and other algae in Herring Bay, Prince William Sound, to the Exxon Valdez oil spill. In: Rice, S. D., R. B. Spies, D. A. Wolfe, and B. A. Wright (Eds.), Proceedings of the Exxon Valdez Oil Spill Symposium. American Fisheries Society Symposium 18. American Fisheries Society, Bethesda, Maryland, pp. 193–211. Van Tamelen, P. G. and M. S. Stekoll. 1997. The role of barnacles in the recruitment and subsequent survival of the brown alga, Fucus gardneri (Silva). J. Exper. Mar. Biol. Ecol. 208: 227–238. Van Tamelen, P. G., M. S. Stekoll, and L. Deysher. 1997. Recovery processes of the brown alga, Fucus gardneri (Silva), following the Exxon Valdez oil spill: settlement and recruitment. Mar. Ecol. Prog. Ser. 160: 265–277. Varanasi, U., J. E. Stein, K. L. Tilbury, J. P. Meador, C. A. Wigren, R. C. Clarc, and S. L. Chan. 1994. Chemical contaminants in gray whales (Eschrichtius robustus) stranded along the west coast of North America. Sci. Total Environ. 145: 29–53. Vaughan, S. L., C. N. K. Mooers, and S. M. Gay, III. 2001. Physical variability in Prince William Sound during the SEA study (1994–98). Fish. Oceanogr. 10(Suppl. 1): 58–80. von Ziegesar, O., E. Miller and M. E. Dahlheim. 1994. Impacts on humpback whales in Prince William Sound. In: T.R. Loughlin (Ed.), Marine mammals and the Exxon Valdez. Academic Press, San Diego, California, pp. 173–192. Wang, Z., B. P. Hollebone, M. Fingas, B. Fieldhouse, L. Sigouin, M. Landriault, P. Smith, J. Noonan, and G. Thouin. 2003. Characteristics of Spilled Oils, Fuels, and Petroleum Products: 1. Composition and Properties of Selected Oils. U.S. Environmental Protection Agency, Athens, GA. Publication No. EPA/600/R-03/072. Wells, P. G., J. N. Butler, and J. S. Hughes (Eds.), 1995. Exxon Valdez Oil spill: Fate and effects in Alaskan waters. American Society for Testing and Materials, Philadelphia; ASTM Special Technical Publication 1219. Wertheimer, A. C., A. G. Celewycz, M. G. Carls, and M. V. Sturdevant. 1994. Impact of the oil spill on juvenile pink and chum salmon and their prey in critical nearshore habitats. Fish/Shellfish 4 State/Federal Natural Resource Damage Assessment Final Report to the Exxon Valdez Oil Spill Trustee Council. Wertheimer, A. C. and A. G. Celewycz. 1996. Abundance and growth of juvenile pink salmon in oiled and non-oiled locations of western Prince William Sound after the Exxon Valdez oil spill, In: Rice, S. D., R. B. Spies, D. A. Wolfe, and B. A. Wright (Eds.), Proceedings of the Exxon Valdez Oil Spill Symposium. American Fisheries Society Symposium 18. American Fisheries Society, Bethesda, Maryland, pp. 518–532. Wertheimer, A. C., N. J. Bax, A. G. Celewycz, M. G. Carls, and J. H. Landingham. 1996. Harpacticoid copepod abundance and population structure in Prince William Sound,
520
Long-Term Ecological Change in the Northern Gulf of Alaska
one year after the Exxon Valdez oil spill. In: Rice, S. D., R. B. Spies, D. A. Wolfe, and B. A. Wright (Eds.), Proceedings of the Exxon Valdez Oil Spill Symposium. American Fisheries Society Symposium 18. American Fisheries Society, Bethesda, Maryland, pp. 551–563. White H. K., L. Xu, A. L. C. Lima, T. I. Eglinton, and C. M. Reddy. 2005. Abundance, composition, and vertical transport of PAHs in marsh sediments. Environ. Sci. Technol. 39: 8273–8280. Wiedmer, M., M. J. Fink, J. J. Stegeman, R. Smolowitz, G. D. Marty, and D. E. Hinton. 1996. Cytochrome P-450 induction and histopathology in preemergent pink salmon from oiled spawning sites in Prince William Sound, In: Rice, S. D., R. B. Spies, D. A. Wolfe, and B. A. Wright (Eds.), Proceedings of the Exxon Valdez Oil Spill Symposium. American Fisheries Society Symposium 18. American Fisheries Society, Bethesda, Maryland, pp. 509–517. Williams, T. M., G. A. Antonelis, and J. Balke. 1994. Health evaluation, rehabilitation, and release of oiled harbor seal pups. In: T. R. Louglhin (Ed.), Marine Mammals and the Exxon Valdez. Academic Press, New York, pp. 227–242. Willette, M. 1996. Impacts of the Exxon Valdez oil spill on migration, growth, and survival of juvenile pink salmon in Prince William Sound. In: Rice, S. D., R. B. Spies, D. A. Wolfe, and B. A. Wright (Eds.), Proceedings of the Exxon Valdez Oil Spill Symposium. American Fisheries Society Symposium 18. American Fisheries Society, Bethesda, Maryland, pp. 533–550. Wolfe, D. A., M. J. Hameedi, J. Galt, G. Watabayashi, J. Short, C. O’Clair, S. Rice, J. Michel, J. R. Payne, J. Braddock, S. Hanna, and D. Salel. 1994. Fate of the oil spilled from the Exxon Valdez. Environ. Sci. Technol. 28: 561A–568A.
Mini Glossary Agonist: A compound that triggers a response in a receptor. Edema: Swelling of tissues with water, a common outcome of PAH exposure. Histopathology: An abnormality in cells, tissues or organs that is visible and usually detected by microscopic examination of the tissue. PAH: Polynuclear aromatic hydrocarbons. Multi-ringed aromatic hydrocarbons, the most toxic class of hydrocarbons found in crude oil.
Chapter 6
Long-Term Changes in the GOA: Properties and Causes
Robert B. Spies, Theodore Cooney, Alan M. Springer, Thomas Weingartner and Gordon H. Kruse
6.1. Introduction Ecological change in the Gulf of Alaska occurs on all temporal and spatial scales. In this book we have addressed temporal scales ranging from single anomalous years, groups of successive or near-successive years with similar behavior, decades, multiple decades, centuries, and millennia. Spatially, we have considered scales from local (e.g., a bay or fjord), through meso-scales (e.g., Prince William Sound or Lower Cook Inlet), to the basin-wide scale (the whole Gulf of Alaska). Of course the largest, most longlasting changes that affect our lives and the ecosystem we depend on are our greatest concerns. Temporal changes range from those with periodic behavior to those with no apparent repeating patterns at all. On one end of the spectrum, the statistically defined Aleutian Low Pressure system (ALPS) and the North Pacific High (NPH) repetitively trade dominance in the northern Gulf with the change of the seasons. For roughly 8 months of each year, October to May, winds responding to the location and magnitude of the ALPS drive a surface onshore flow and downwelling over the northern shelf. During the rest of the year, June through September, the strength of the ALPS lessens and its location shifts, and the coastal/shelf downwelling environment changes to allow occasional weak upwelling under coastal westerlies associated with the presence of the NPH. Longer-term quasi-periodic changes, such as the multi-annual ENSO and decadal oscillations in the ALPS that are described as the PDO and the Victoria Pattern, do not occur as strictly regular temporal and spatial patterns. That is, they are not lock-step repetitions. On the other end of the spectrum are the terminal, non-repeating changes (e.g., the extinction of the Steller sea cow in the nineteenth century from human harvest). Long-Term Ecological Change in the Northern Gulf of Alaska Robert B. Spies (Editor) © 2007 Elsevier B.V. All rights reserved.
522
Long-Term Ecological Change in the Northern Gulf of Alaska
BOX 6.1: MARINE ECOREGIONS OF ALASKA by John F. Piatt and Alan M. Springer Any attempt to understand why marine organisms fluctuate in abundance is confounded by spatial scale. At larger scales, we are not surprised to observe seabird populations increasing rapidly at one colony in the Gulf of Alaska (GOA) while plummeting at a colony 1500 km away in the Eastern Bering Sea (EBS) – after all, these bodies of water are discrete “Large Marine Ecosystems” (LMEs, Sherman et al., 1990) and march to the beat of different climatic drummers (Hare and Mantua, 2000). But at smaller scales, we are often perplexed to find contrasting population dynamics in nearby colonies (Section 4.5). This mesoscale spatial heterogeneity probably reflects oceanographic processes occurring at the same scale (Speckman et al., 2005), or differential predator–prey dynamics (Kildaw et al., 2005), but in any case it complicates our efforts to understand how marine fish, bird, and mammal populations respond to, or are altered by, changes in their environment. To further cloud interpretation, many species operate over a wide range of spatial and temporal scales in a lifetime. For example, seabirds forage over tens of kilometers daily, but capture and consume prey in a matter of meters and seconds. Populations fluctuate over basin and decadal scales, but migrate and aggregate to breed over scales of hundreds of kilometers and months. Therefore, it would be useful to characterize spatial heterogeneity in marine systems at scales smaller than those of LMEs. One approach is to adopt methods used to characterize terrestrial ecosystems (Bailey, 1980; Demarchi, 1996). Alaska has at least 20–30 terrestrial “ecosystem provinces” that are defined by topography, vegetation, climate, and other measurable features (Bailey, 1998; Nowacki et al., 2001). We know that there is similar spatial heterogeneity in marine systems, and we can use characteristics such as bathymetry, currents, temperature, and primary production to define marine regions (Bailey, 1998). The degree to which habitats can be subdivided – and the classification system for naming each division – are somewhat arbitrary. For convenience, we followed Demarchi (1996) who developed an hierarchical system for classifying terrestrial and marine ecosystems of British Columbia, and who used the term “ecoregion” to describe a marine ecoregion unit as an area with major physiographic and minor oceanographic variation at a regional spatial scale. For example, slope, shelf, and coastal areas would be segregated because of “major physiographic variation.” Within a shelf area, we might further segregate waters with major oceanographic boundaries, such as fronts associated with the 50-m and 100-m isobaths on the EBS shelf (Coachman, 1986); topographic irregularities, e.g., islands such as Kodiak I. that creates persistent oceanographic differences
Long-Term Changes in the GOA: Properties and Causes
523
upstream and downstream of the island; or currents that significantly alter shelf production regimes, such as the Anadyr Current in the northern Bering Sea (Springer et al., 1989; Springer and McRoy, 1993). For this analysis, we considered only features present during the summer months in Alaska. Presumably, mesoscale features could be quite different in winter – they are probably significant to species that do not migrate south, but are beyond detailed description here due to the paucity of winter-time data on habitat characteristics. There are many sub-regions within the LMEs that can be identified by their similarities in oceanography and biology, and these mesoscale ecoregions often cross LME boundaries (Favorite et al., 1976; Piatt and Springer, 2003; Batten et al., 2005). One may also find regional patterns in geochemistry between and within LMEs (Longhurst 1998; Schell et al., 1998). After examining such mesoscale patterns in the distribution of biological indicators and considering topographic, bathymetric, and oceanographic features (such as persistent fronts, e.g., Belkin and Cornillon, 2003), we have tentatively identified 30 marine ecoregions in Alaska (see Fig. 6.1, Table 6.1). For example, cross-shelf differences in bottom depth and oceanography in the GOA and EBS create welldescribed heterogeneity in plankton, fish, and bird communities across these shelves, reflected in discrete coastal, inner shelf, outer shelf, slope, and oceanic species assemblages (e.g., Cooney, 1981; Doyle et al., 2002; Piatt and Springer, 2003; Lanksbury et al., 2005; see also Section 2.3). Within the enclosed waters of southeastern Alaska and Cook Inlet, gradients in oceanographic conditions create discrete distributional boundaries for some plankton and fish species (Johnson et al., 2005; Speckman et al., 2005). Open ocean areas may be defined by large-scale oceanographic processes, such as the Alaska Gyre in the GOA (Favorite et al., 1976), which appears to spatially structure plankton, fish, bird, and mammal populations (Brodeur and Ware, 1992; Springer et al., 1999). Strong along-slope currents carrying nutrient-rich waters create productive habitats along the edges of continental shelves in the GOA, EBS and Beaufort Sea, creating narrow bands of high primary productivity (e.g., the Bering Sea “Green Belt,” Springer et al., 1996). These support a high abundance of organisms, some of which are tightly associated with those shelf-edge ecoregions (e.g., sablefish and Pacific ocean perch, Fritz et al., 1998; myctophids and squid, Sinclair and Stabeno, 2002; albatross, Piatt et al., 2006). A remarkable synthesis of papers on the Aleutians (Schumacher et al., 2005) reveals how spatial variation in topography and oceanography (Ladd et al., 2005) results in marked segregation of some fish (Logerwell et al., 2005), bird (Jahncke et al., 2005) and mammal (Call and Loughlin, 2005; Sinclair et al., 2005) populations into three distinct ecoregions along the Aleutian Archipelago.
524
Long-Term Ecological Change in the Northern Gulf of Alaska
Table 6.1: Names of Alaska marine ecoregions. Number 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30
Ecoregion name Alaska Gyre Center North Pacific Current – Alaska Stream Loop Eastern Gulf of Alaska Transitional Eastern Gulf of Alaska Slope Prince of Wales Shelf and Inside Waters Chichagof Shelf and Inside Waters Northern Gulf of Alaska Slope Northern Gulf of Alaska Shelf Prince William Sound Inside Waters Western Cook Inlet – Shelikof Strait Southeastern Cook Inlet – Kodiak Upwelling Alaska Peninsula Coastal and Shelf Western Gulf of Alaska – Alaska Stream Eastern Aleutians Central Aleutians Western Aleutians Aleutian Arc – Alaska Stream Bering Sea – Bowers Basin Bering Sea – Aleutian Basin Bering Sea Shelf Edge – Green Belt Eastern Bering Sea – Outer Domain Eastern Bering Sea – Middle Domain Eastern Bering Sea – Inner Domain Eastern Bering Sea – Alaska Coastal Northern Bering – Chukchi Sea – Anadyr Stream Western Bering Sea – Shelf Beaufort–Chukchi Coastal – Shelf Beaufort–Chukchi Sea – Barrier Island-Lagoon System Beaufort–Chukchi Sea – Shelf Edge Arctic Ocean – Basin
In these examples, some indicator species were found to occupy only one or a few ecoregions (and thereby helped define them), whereas many more species showed no apparent affinity for any one ecoregion or its boundaries. This highlights the subtlety of ecoregional structuring: it is important enough to explain the distribution patterns of some taxa, but may only serve as a source of
Long-Term Changes in the GOA: Properties and Causes
525
background variability to species that are adapted for living in a broader range of habitats. Either way, it may be useful to consider the population ecology of marine animals in Alaska in light of ecoregion patterns (Fig. 6.1). If boundaries are stable, perhaps these patterns can help explain the distribution of some taxa. If ecoregions have different production regimes and patterns of variability, perhaps we can better explain spatial and temporal variability in status and trends of widely distributed taxa (e.g., Dragoo et al., 2003), and even the contrasting population dynamics of adjacent colonies that are actually situated in different ecoregions (e.g., see Piatt and Harding, Section 4.8). This is an initial effort; the boundaries, shapes, and number of ecoregions will no doubt be refined. For the present, we can draw a few conclusions: (1) The size and shape of marine ecoregions differ one from another, but usually extend along one axis that is determined mostly by bottom topography and current flow; (2) There is much greater heterogeneity in coastal-shelf environments than in the open ocean; (3) Across-shelf boundaries between ecoregions are fairly conspicuous, being defined by persistent fronts or strong topographic gradients between
Figure 6.1: Marine ecoregions of Alaska.
526
Long-Term Ecological Change in the Northern Gulf of Alaska
shelf, slope, and oceanic habitats; (4) Along-shelf boundaries are more subtle, and are often resolved by patterns in animal distribution more so than by physical characteristics; (5) More analyses are needed to better resolve marine ecoregions in Alaska using both physical and biological datasets.
Indeed, anthropogenic changes tend to be highly aperiodic (marine mammal harvests), chronic and continual (e.g., contamination by persistent organic pollutants, greenhouse gas emissions), or terminal and non-repeating. Large spills like the Exxon Valdez are hopefully in the latter category. Capturing the variability in this dynamic environment requires sampling over a range of spatial and temporal scales that are matched to the dimensions of change and replicated sufficiently to account for random error. In this regard the Gulf of Alaska is greatly undersampled. So, any synthesis about the Gulf of Alaska must acknowledge that there is simply not enough sampling to fully describe, let alone understand its ecological changes. Our longest climate records are about 100–150 years, except for some proxy measurements, such as several hundreds of years from tree rings and several thousands of years from sediment cores. The long-term oceanographic records also are few and very widespread, that is at Ocean Station P in the extreme southeastern part of the Gulf on the southern edge of the sub-arctic gyre that go back into the 1950s, and at the GAK-1 station in the far northern Gulf of Alaska near Seward, Alaska, that began in 1970. The lack of long-term physical oceanographic records in most places in the Gulf has been addressed recently through the ARGO program that has deployed water column profiling instruments throughout the Gulf. Biological records are typically of shorter duration; the longest is the time series of phytoplankton and zooplankton samples at Ocean Station P, which started in 1956 and ran for about 25 years. Otherwise, we have rather poor historical records on the phytoplankton and the benthic communities and virtually nothing can be said with regard to changes in these important groups of organisms, mirroring the lack of data in other ocean basins (e.g., Reid et al., 1998 on North Atlantic phytoplankton). Some new plankton studies are now being analyzed as part of the EVOS GEM Program and GLOBEC-supported research, including a continuous plankton recorder. These studies have recently identified a shift in the plankton communities in the North Pacific with the initiation of a cold-water event in 1998–1999 (Batten and Welch, 2004), pointing to the importance in continuing such measurements for the long term. Many commercially important fish populations are better sampled due to stock assessments by management agencies, however, the very important small pelagic forage fishes and the more cryptic benthic invertebrates that are key to success of larger predatory fish, sea birds, and marine mammals are not generally sampled and we know little of their
Long-Term Changes in the GOA: Properties and Causes
527
historical changes. Juvenile herring and pollock are exceptions, because the adults are fished commercially and we obtain some measures of the status of the juvenile stocks through various monitoring activities. Seabirds, because of their visibility in breeding colonies, are relatively well sampled since the mid-1970s but not earlier. Likewise, marine mammals are relatively well sampled over the last thirty years only. In nearly all cases, however, there are large gaps in the geographical coverage and in the years sampled. The lack of sampling in the Gulf explains why some ecosystem components are woefully under-represented in this book: not much on the intertidal and subtidal communities and very little on the benthos. We simply do not know if and how these components of the ecosystem fluctuate in response to natural or human forces, since they do not have much recognized immediate human importance (as do fisheries or meteorology), and consequently do not have a historical record in our literature. One tantalizing, recently discovered example of long-term change in the benthos linked to climate change is the change in echinoderm populations over a fourteen-year period in abyssal depths (4000 m) in the northeastern Pacific Ocean, apparently in response ~a cycle (Ruhl and Smith, to the changes in food production during the El Nin~o-La Nin 2004). This suggests that even very deep communities apparently well isolated from surface water production do, in fact, receive a clear climate change signal. Therefore, long-term changes in pelagic and nearshore production might well be reflected also in the bottom communities that are so important in the recycling of nutrients to the surface waters of the Gulf. The studies of the intertidal communities carried out after the Exxon Valdez oil spill and the work around the Valdez oil terminal (Blanchard et al., 2002) are two notable examples of intertidal and subtidal research that document the effects of man on these ecosystem components and begin to approach the status of long-term data sets. There are several major themes in this chapter. First, we discuss the relative importance of various major forces of change in the Gulf as we now understand them. This first part includes an explicit discussion of the Exxon Valdez oil spill. Second, we make some generalizations about the Gulf ecosystem with regard to its behavior over the long-term.
6.2. Forces of Change On time scales of ten thousand years and longer, geophysical forces (glaciers, earthquakes and eruptions) and climate have been the two largest influences on the marine ecosystems of the Gulf of Alaska, but it is likely that human effects will be as great a factor in the future on this time scale. In fact, we have entered what is proposed as the Anthropocene era (Finnegan, 2003), where our activities are a dominant force in the biosphere, including the oceans. The two largest influences in causing change on the time
528
Long-Term Ecological Change in the Northern Gulf of Alaska
scale of human history are climate and human harvests. On this same time scale, earthquakes and oil spills are the next most important in the Gulf, due to two large episodes.
6.2.1. Climate There is an underlying low-frequency decadal and multi-decadal forcing that is expressed in the atmosphere as changes in pressure, winds, temperature, and precipitation, and translated into the ocean as changes in spatial patterns of water temperature (central vs. northeast and south), mixing depth, upwelling, and downwelling. The major atmospheric variable is the position of the ALPS, which along with the eastern Pacific High pressure system, defines the state of the Pacific decadal oscillation, the PDO. In addition, there is another mode of variability that has occurred since about 1999 in the North Pacific, the so-called Victoria pattern. This is a pattern in which the atmospheric and oceanographic conditions in the Gulf of Alaska have continued to resemble the post-1977 situation, but the eastern Pacific has returned to pre-1977 conditions (Bond et al., 2003). Predominant meteorological and oceanographic phenomena are expressed in the biological components of the ecosystem, particularly in the lower trophic levels that presumably respond most directly to the physics. Plankton and the fry and juveniles of most fish, such as salmon, herring, and groundfish that have early life history stages in the plankton, are also more directly tied to the variation in the physical variables. For example, the major populations of Pacific salmon apparently benefit greatly from positive phases of the PDO (Francis and Hare, 1994; Hare and Mantua, 2000) and changes in herring abundance are linked to changes in both atmospheric pressure and plankton populations (see Section 4.5.2). This expression is quite variable in space and time, i.e., grainy, and is greatly diluted, inconsistently expressed, and often masked by other forcing in the upper trophic levels, e.g., predation by marine mammals or commercial fishing. That is, the bottom-up forcing of the system is most evident at the bottom of the trophic system and less evident at the top. However, an analysis of long-term biological records of a variety of species at numerous trophic levels in the Gulf of Alaska and Bering Sea found that a strong component of the biological variability is correlated with the PDO (Hare and Mantua, 2000), including the restructuring of a large part of the fish and shellfish community in response to the climate regime shift of the mid-1970s (Anderson and Piatt, 1999). Physical oceanographic changes forced by the PDO probably drive the fluctuations of these ecosystem components, but the exact mechanisms that may be involved are speculative. The new mode of variability, the Victoria pattern, has not occurred for very long and there are not yet any similar broad-scale analyses of ecological responses to it. Changes in winds, nutrient supply to the photic zone, temperature, and primary production that affect food supply are likely involved in the PDO and other modes of physical variability.
Long-Term Changes in the GOA: Properties and Causes
529
The effects of changes in the atmospheric–oceanic system are assumed to be expressed through bottom-up forcing. That is, some aspect of the change in climate that accompanies a shift of ALPS results in altered production at the base of the food web. It is not difficult to imagine that when the ocean climate favors greater or lesser rates of primary production, that higher-level consumer stocks will respond in kind, but perhaps on different time and space scales. For example, zooplankters that produce one-to-many generations per year are capable of an almost “instantaneous” response to shifting food resources, whereas long-lived fishes, and marine birds and mammals with lower reproductive rates, are buffered from rapid population changes by multiple age groups and slower production of young. However, these long-lived species are susceptible to sudden losses that affect many age groups simultaneously, such as overharvest, pollution or lethal disease outbreaks. Figure 6.2 shows the conceptual basis for the buffering effect of life span and reproductive rate on response to bottom up forcing. This figure represents the potential for most kinds of change except sudden lethality in longer-lived animals with multiple age classes and low reproductive rates. While bottom-up influences – nutrient supplies and concentrations, and light – limit ecosystem productivity, the distributional (spatial) aspects (vertical and horizontal) of this “production” – particularly of key forage stocks – can presumably govern the degree to which the system is actually enriched. Concentration of enhanced production in the form of one or a few highly mobile forage species can lead to uncertain outcomes in large fish, seabirds and mammals depending on the particular forage species and how they are distributed temporally and spatially. We suggest that marine science has little or no predictive capability when it comes to the distribution of pulses of new production within middle and higher levels of the food web. That is, in this complex web of trophic and other interactions predicting which species populations will receive new energy and carbon and the potential, therefore, to expand is less
Figure 6.2: Diagram of the proposed relationship between potential for population change, reproductive output and life span of Gulf of Alaska animals.
530
Long-Term Ecological Change in the Northern Gulf of Alaska
predictable than changes at whole trophic levels. The reader can refer to Figs. 3.6 (trophic levels) and 3.7 (trophic web) to visualize how predictability for trophic levels may be greater than for trophic webs. The top-down forces that presumably negate, obscure, or modulate these bottomup changes are not well understood, but there is an emerging appreciation of the role of predation in regulation of marine populations. Top down control and trophic cascades are treated later in this chapter. There is another higher frequency pattern of change that influences much of the GOA – the El Nin~o/Southern Oscillation (ENSO). The major quasi-periodic states of the ocean, the ENSO and the PDO, are expressed in nearly identical spatial patterns of ocean water temperature, but the temporal scales are obviously different. The warming of the GOA mixed layer resulting from ENSO events has been associated with altered patterns of recruitment of some groundfish (Hollowed et al., 2001) and may have contributed to the ascendancy of gadoid fishes that occurred in the late 1970s. Die-offs of common murres and kittiwakes are occasionally caused by these warm-water episodes that temporarily depress prey availability (Piatt and Van Pelt, 1993). We surmise that, because the PDO and ENSO events have different time scales (15 to 30 versus 3 to 7 years, respectively), the PDO and other mode(s) of multi-decadal variability have greater cumulative effects on populations of organism in the GOA, especially those with life cycles longer than an ENSO event. The third mode of variability is global warming. Global warming may be expected to have amplified effects closer to the poles because of the increased solubility of water vapor per degree rise at lower temperatures. In the northern GOA, it is likely to be expressed as warmer and wetter conditions, especially in winter, that may result in earlier springtime stratification and a more strongly stratified water column, and therefore, a difference in the timing and duration of the spring plankton bloom. This would be a fundamental change that could affect the ecosystem in many ways, such as selection of different dominance patterns in the plankton and the forage stocks. Given the physiological tolerances of GOA organisms, significant further warming of the North Pacific Ocean could have profound effects, for example on the distribution of sockeye salmon in the Alaska Gyre. About half of the extra greenhouse gases that have gone into the atmosphere since the Industrial Revolution and contributed to global warming have been sequestered in the ocean (Sabine et al., 2004). The continuing accumulation of CO2 may eventually have profound repercussions by shifting the carbonate equilibrium of sea water that many plants and animals require to precipitate their carbonate-base skeletons (Feeley et al., 2004). Therefore, as with the PDO, more factors than simply temperature are involved in the oceanic–atmospheric response to climate change – in this case pH of sea water. As heat balance and distribution change, the positions and intensities of high and low pressure systems, storm tracks, prevailing winds, precipitation, ocean
Long-Term Changes in the GOA: Properties and Causes
531
currents and nutrient renewal to surface waters will all likely change and affect the ecosystem in potentially profound ways.
6.2.2. The Exxon Valdez Oil Spill The oil spill had immediate acute effects on scales of hundreds of kilometers that peaked within days and weeks. As much of the oil was dissipated, weathered, biodegraded, and photooxidized, and, as the injured populations began to replace their lost members, the affected area shrank. Within weeks and months, this retraction allowed the open water environment to recover from acute effects. Then, after several years, the shallow benthic communities probably recovered – damage to the deepwater benthos (below 20 m) was not established and it appears that the effects of this spill were limited to remarkably shallow depths compared to other large spills. A great deal of recovery had taken place in the intertidal communities within several years, but probably continued at a slower pace for many more years. The loss of oil in low-energy environments is key to understanding recovery and the rate of loss in such environments is a process of diminishing returns. That is, the rate continually decreases so that after several years, it is a fraction of what it was just after the spill. The result is that reduction of oil to background concentrations of hydrocarbons is a long-term process, and low-level exposure of some animals associated with the low energy environments is possible for years. The recent discovery of subsurface deposits of liquid oil in the beach at McClure Bay from a tank ruptured in the 1964 earthquake strongly suggests that similar deposits of Exxon Valdez oil will persist in several locations of the PWS for many more decades with attendant exposure risks to animals. The recovery of injured populations from acute effects of the spill was quite uneven among species. Some started to recover fairly quickly, while others showed no apparent recovery for years. As far as it could be determined, complete recovery only occurred in the populations of a few species, such as bald eagles and perhaps a few populations in the lower intertidal zone, around 1995. It should be remembered that recovery has two aspects – the return of a decimated species to its expected abundance in the absence of further exposure, and recovery under continuing low-level oil exposure, which may not be completed until exposure ceases. It is now fairly apparent that the speed of recovery in particular populations is due to the life history characteristics of the species (e.g., reproductive rate), environmental conditions, and oil exposure affecting the fitness of the recovering population, i.e., sublethal effects. The sublethal effects of the spill occurred in a variety of species, and in some of them continued in excess of ten years. The sublethal effects in the first year or two included genetic and morphological abnormalities (e.g., herring), reduced growth of juveniles (e.g., pink salmon), and reproductive impairment (e.g., the seaweed Fucus and bald eagles).
532
Long-Term Ecological Change in the Northern Gulf of Alaska
The indications for long-term sublethal effects came mainly from the presence of biomarkers (i.e., the cytochrome P-4501A protein) produced by exposure to the aromatic hydrocarbon fraction of crude oil. The presence of these markers alone indicates potential damage from conversion of some aromatic compounds to toxic metabolites that oxidize cellular components. In most instances, there were no data to indicate whether there was enough damage to affect growth and reproduction, and hence affect the recovery of populations damaged immediately after the spill. However, there are a number of indications strongly suggesting that sublethal effects on growth and survival were linked to the presence of aromatic hydrocarbons: 1. Some species showing long-term P4501A induction utilize the intertidal habitat where the residual oil is located. These are primarily intertidal and nearshore fish, harlequin ducks and sea otters. 2. The recovery of over-winter survival rates of harlequin ducks in oiled areas to levels comparable to those in unoiled areas was coincident with the reduction of P4501A in tissues of birds in oiled areas approaching levels observed in birds inhabiting unoiled areas. The apparent pronounced sensitivity of sea otters and harlequin ducks to oil exposure in Prince William Sound is consistent with a view that these homeotherms are living on the energetic edge. Metabolism of the oil to which they are exposed in the course of their day-to-day activities adds an extra energy demand that may be a critical factor for their survival at key points in their life history. The effect of a sub-arctic winter, in conjunction with additional stressors, was demonstrated with radio-tagged harlequin ducks. Harlequin ducks in western PWS which were exposed to oil, showed higher P4501A induction, and experienced a sudden increase in mortality in mid-winter, while those from unoiled areas had normal levels of P4501A and normal mortality patterns. Such sensitivity is also plausible for sea otters, whose high metabolic rate is necessary for maintenance of their body temperature. The energy demands required for the long Alaskan winters must be quite high, likely making sea otters vulnerable to additional demands that oil metabolism requires. This hypothesis of energetic demand is offered not knowing the energetic demand associated with small amounts of oil exposure. It may be that the energetic demands from low-level oil exposure may not be significant relative to overall metabolism. So, this response is speculative and requires further evaluation as a plausible explanation of the problems that the nearshore predators have experienced following the spill. Because the weight-of-evidence supports a long-term impact of spilled oil retained in the beaches of Prince William Sound on marine birds and mammals, the findings from the post-spill studies are unique. Long-term impacts of spills have been previously established for salt marsh invertebrates, e.g., after the 1969 Florida spill of No. 2 fuel oil (Sanders et al., 1980; Krebs and Burns, 1977; Teal et al., 1992), and
Long-Term Changes in the GOA: Properties and Causes
533
for mangroves after the Galeta spill in Panama (Levings et al., 1994), among others. In each case, oil was retained in the sediments of lower-energy environments longer than was initially anticipated. Just how severe was the damage from the oil spill? Based on the results of the many post-spill studies, the GOA ecosystem was not massively altered for a long period of time – that is, there was no likely irreversible damage. But this requires some qualification with regard to potential indirect effects that may have occurred but are not proven. It is becoming more widely understood that changes in ecosystems are not linear or smooth, but under duress they bend so far, and then break. For example, there are cascading effects of overfishing in tropical reef ecosystems (Knowlton, 1992) where severe fluctuations in the abundance of predators, herbivores, and algal populations have converted systems into other states from which they may not easily recover. That is, a new quasi-stable equilibrium in the community is established that differs from the pre-impact state. The GOA ecosystem began to manifest some of these changes in the intertidal communities and in one case, a part of the system, the herring food web, may have been altered in this way by the spill. In the intertidal zones there was a succession of green and red algae on rocky surfaces left bare by the loss of Fucus cover after the spill and possible increased predation on limpets in these areas by oystercatchers. There is little evidence that these changes persisted. One possible persistent effect of the spill was on the herring population, which has still not returned to pre-spill conditions. There was a documented effect of the spill on the developing embryos in the eggs in 1989, perhaps damaging a large portion of the 1989 cohort in PWS (Chapter 5). The link between the spill and the subsequent 1993–1994 herring population crash is uncertain. However if oil contributed to, or was the main cause of the crash, then the apparent lack of recovery of this key forage fish population may be a case of a partial breakdown of ecosystem structure that may not be recoverable in the foreseeable future. That is, while the herring population in PWS and the GOA appears to respond strongly to bottom up climatic forcing (see Section 4.5.2), additional controlling factors, i.e., predation by marine mammals and losses to two diseases (Section 3.6.4) that acted differently on the population after the spill than before may have kept the population from recovery. In any case the oil spill was not as severe a perturbation as the 1977 or 1989 climate-driven regime shifts, the former one altering the composition of the shelf pelagic ecosystem on a basin scale for at least 20 years (Piatt and Anderson, 1996; Hare and Mantua, 2000). There is abundant evidence presented in Chapter 5 that full recovery from the spill damage took 15 years at least, and may still be occurring. In general, but with several important exceptions, as time passes it becomes more difficult to distinguish recovery from ecosystem fluctuations due to other causes, and this problem makes decisions on
534
Long-Term Ecological Change in the Northern Gulf of Alaska
recovery increasingly uncertain. Yet, we do not accept the position of the Exxon Corporation and some of the scientists who have claimed that the spill area was virtually recovered within several years. The damages inflicted on nearshore organisms, e.g., sea otters and harlequin ducks, are supported by a weight-of-evidence approach, 10 years after the spill, but it seems that this is not a high enough standard for full acceptance by some members of the scientific community. Following the spill, the ecosystem was also adjusting to shifts in climate as it was recovering. It is become more apparent since 1989 that the pelagic portions of the ecosystem were altered late in the 1970s (Piatt and Anderson, 1999), again in 1989 (Chavez et al., 2003), and once again in 1999 (Batten and Welch, 2004). For example, the standing stock of plankton in Prince William Sound in the early 1980s was much greater than in the early 1990s after the spill (Eslinger et al., 2001). Whether the intertidal and nearshore vertebrates that were hard hit by oil in 1989 and which depend on the shallow benthos and intertidal zones were affected by the shifting climate is a question that we have virtually no data to answer. It is not known, nor is there a way to predict, what another large spill might do cumulatively to this ecosystem before it is recovered. The measures, such as some provisions of the Oil Pollution Act of 1990, taken towards prevention of another spill of this sort are well warranted for this reason.
6.2.3. The 1964 Earthquake Large earthquakes can have positive effects in some types of habitat, cause the loss of other portions of some habitats (e.g., salmon streams) and displacement of the subtidal and intertidal communities that may have long-lasting effects (e.g., clams in PWS). The 1964 earthquake had effects on an area tens of thousands of square kilometers – on the order of the area affected by the oil spill but mainly confined to the intertidal and sea bottom. The time course of recovery from nearshore and intertidal studies was in the order of years for some of the damaged resources, while some salmon runs on the uplifted margin of Montague Island in Prince William Sound were extirpated. Effects on sea bottom communities were not documented beyond the uplift of some subtidal communities into the intertidal zone. The known effects of the 1964 earthquake are still cascading through the populations of colonial seabirds on Middleton Island. Although follow-up studies were not carried out on the initial descriptions of intertidal and subtidal community damage by the earthquake, it seems reasonable to assume that it took several years, perhaps a decade or more, for these communities to recover. It has been suggested that the uplifting of the subtidal habitat within Prince William Sound had a devastating effect on clam populations. However, the recolonization of the sound by sea otters, whose diet consists mainly
Long-Term Changes in the GOA: Properties and Causes
535
of clams, in the 1960s, 1970s and 1980s probably also contributed to this depletion. Certainly, return of the shallow subtidal clam populations to pre-earthquake status, which has not yet occurred, has been slowed by the sea otter predation. So, the earthquake’s effects in space and time resembled the effects of the oil spill, with initial damage occurring over tens of thousands of square kilometers and taking several years to recover: some effects in the ecosystem still persist after a decade or more. The geophysical alterations from the earthquake however were more permanent than the fading toxic effects caused by the oil spill.
6.2.4. Effects of Harvesting Man’s harvesting at the top of the trophic pyramid, i.e., marine mammals, has had long-lasting effects in the Gulf of Alaska. Certainly the extinction of the Steller sea cow is a permanent change in the Aleutians. Harvesting of sea otters, great whales, harbor seals and sea lions have played roles in the long-term population dynamics of these species. These harvesting activities have caused or led to changes that have lasted multiple decades and sometimes centuries in the Gulf of Alaska. In the case of sea otters, it was not until the 1980s that sea otter populations rebounded from the intense harvests of the eighteenth and nineteenth centuries. The bounty programs for harbor seals, and to a much lesser extent sea lions, implemented in the 1930s through 1960s in the hope of reducing competition with commercial fisheries, may have initiated or contributed to the multi-decadal decline of these species that we are currently experiencing.
6.3. Characteristics of Ecosystem Change Our collective experience leads us to some general observations about changes in the Gulf of Alaska marine ecosystem. These emerged from extended discussions among the core authors and relate to particular properties of the Gulf of Alaska including: high productivity of the Gulf, response to perturbations, the effects of humans in the ecosystem, contaminants as agents in long-term ecological change, life history characteristics of animals that drive ecological change, how the trophic structure might make the pelagic ecosystem vulnerable to change, how we might better conceptualize the system, and to what extent we can hope to develop predictability for ecosystem change. We offer some of these features as the bases for working hypotheses that should help guide efforts to document change, its causes and what might be expected in the future.
536
Long-Term Ecological Change in the Northern Gulf of Alaska
6.3.1. Understanding Productivity is Key Significant progress in understanding ecological change will come from a deeper understanding of Gulf productivity. The ecosystem appears to be very productive based on the following observations: a. The deep water of the GOA has very high concentrations of nutrients, e.g., nitrate, silicate and phosphate, as there is very slow exchange with other basins. The deepwater carbon is also very old, which is consistent with slow horizontal and net vertical advection. How the rate of supply of deep-water nutrients to the surface waters might be affected by long-term climatic changes is the subject of much current research. b. The shelf, which is where most primary production occurs, is relatively deep compared to other ocean shelves. The shelf break is often below the permanent pycnocline in the Gulf of Alaska, which allows the deep-water nutrients to reach far inshore in the late summer and fall when downwelling is minimal. The shelf is also relatively broad in the northern apex and in the northwestern sector of the Gulf. In the world’s oceans, shelf production is usually higher than open ocean production. c. The coast is rough and this roughness causes both horizontal and vertical turbulence, promoting nutrient transport from the deep and its distribution over the shelf. d. There are strong winds and deep wind mixing that bring nutrients to the surface. Also, the strong winds induce significant Ekman transport of nutrients and organisms inshore. e. There is a very large tidal range, which interacts with the topography on the shelf to produce vigorous vertical advection. The entrance to Cook Inlet is subject to a great deal of tidal-induced mixing that likely brings nutrients to the surface even when conditions would otherwise favor stratification. f. Energy capture and exchange through the food web occur in a relatively short interval each year so that the higher trophic levels are fueled soon after solar energy is fixed at the lowest trophic level. Energy transfer in the lower trophic levels is very efficient, because little energy gets consumed in maintenance metabolism. At the highest trophic levels, pinnipeds, porpoises and killer whales have quite large amounts of maintenance metabolism. So it is not so evident that this pattern of energy capture is as predominant in the spring and summer. g. The amount of energy reaching higher-level consumers is also a function of the complexity of the food web. When the web is supported by large diatoms, as over the shelf during spring, fewer steps/exchanges are needed to move fixed carbon to higher levels than is the case when the primary producers are very small. Thus, an increase in food for fishes, birds and mammals may be facilitated by simplifying the structure of the consumptive process – fewer exchanges translate into more energy reaching the consumers.
Long-Term Changes in the GOA: Properties and Causes
537
Inevitably, long-term climate changes, the major force affecting productivity in the Gulf of Alaska, are driven by heat balance and distribution. When the amount of heat trapped on the earth’s surface and in the atmosphere changes, concomitant changes will occur in the transport of heat in the atmosphere or in ocean currents; this will be manifested in altered rates of vertical turbulence, rates of nutrient renewal to the surface waters and stability of the upper water column. As was outlined in Chapter 2, these are important factors in determining the amount of solar energy that is fixed by plant photosynthesis and ultimately sets limits on the food and energy supply to the higher trophic levels. At this time active, process-oriented research in the GLOBEC program is investigating how climate change may affect nutrient renewal in the Gulf of Alaska. Little more can be said with any degree of certainty until these studies have been completed.
6.3.2. Responses to Climate Changes Vary at the Mesoscale Changes in productivity seem likely to occur with alterations of climate, whatever the specific mechanisms are, and it is clear from studies of change over the past 30 years that populations of the Gulf of Alaska will respond. Such forcing of populations may result in basin-scale, meso-scale, or local changes. The spatial scale of the changes may provide clues about the causes of the change. Common meteorological forcing across the Gulf finds different expression in different habitats, and these differences sometimes provide a means to interpret cause and effect. One example of this phenomenon is the change in Cook Inlet seabird colonies over the past thirty years. These changes provide an interesting contrast within Cook Inlet itself as well as with changes in Prince William Sound during the same period. Basically, populations of common murres and black-legged kittiwakes have increased at the Gull Island colony in Kachemak Bay, Lower Cook Inlet, but they have been decreasing at Chisik Island on the western side. From anecdotal accounts, it appears that there is decreased salinity in the area around Chisik Island in the year 2000 than 30 years earlier. Satellite photos taken during the summer show a clear influence of turbid freshwater outflow from upper Cook Inlet to the western side of middle Cook Inlet. We speculate (in the absence of long-term stream gauge data in the Susitna and Matanuska Rivers) that freshwater inputs to Cook Inlet have increased with global warming and the strong positive phase of the PDO starting in the mid-1970s. The general circulation and salinity patterns in lower Cook Inlet appear to result from the combined intrusion of warmer (in the summer), fresher, turbid and nutrient-poor water from upper Cook Inlet and the intrusion of cold, nutrient-rich, saline water from the Gulf carried in the Alaska Coastal Current (see Fig. 4.38). The latter properties are probably intensified by topographically and tidally induced upwelling in the
538
Long-Term Ecological Change in the Northern Gulf of Alaska
entrances to the inlet. We suggest that the balance between these opposing water masses and their relative strengths set the productivity field. Our interpretation is that Chisik Island bird colonies have suffered from the greater influence of the unproductive, oligotrophic, freshwater from the northern, upper Cook Inlet since at least the early 1970s and that some aspects of the PDO acted to make the marine waters coming into the Inlet more productive over part of this time. Therefore, because sea bird colony productivity often reflects the productivity of adjacent marine habitat, we infer that productivity at Chisik Island decreased, and increased at Gull Island, as a result of climate change at one scale or another (see Chapter 4). This interpretation is speculative, but what has happened in Cook Inlet may be a good example of how common forcing on a basin scale can be expressed in different environments on a mesocsale, and thereby offers clues to possible mechanisms of change.
6.3.3. Animals with High Reproductive Rates Periodically Dominate The physical and biological conditions in the Gulf of Alaska allow animals with large annual reproductive capacities to occasionally produce very large numbers of young in a few consecutive years, which will then lead to dominance of the system by this species on decadal scales. Examples of this phenomenon include some year classes of Alaska king crab (see Section 4.10) and groundfish, such as pollock (Hollowed et al., 2001). Because of our short historical record of surveys for these populations, it is not known how often these occasional successful years occur, but there continues to be outstanding examples of this year-class phenomenon in the Gulf and it appears to be an important strategy for some invertebrate and fish species. Because they are infrequent, large increases in some populations may be anomalies rather than the norm. It is clear that the reproductive strategy of many invertebrates and fishes is to flood the environment with large numbers of eggs and larvae. This adaptation occasionally allows above-normal survival when other elements of the system are appropriately predisposed. These “anomalous” conditions must surely effect the food web at certain times and places, and for commercial species, they may support a fishery for many years.
6.3.4. Top-down Forces Have been Underestimated The importance of predation has been underestimated in traditional approaches to biological oceanography, with primary importance being attached to biogeochemical cycling and treating the individual populations in the food web as functional nodes,
Long-Term Changes in the GOA: Properties and Causes
539
or black boxes, for transfer of energy rather than individual organisms whose unique adaptations, in fact, can have important effects on energy transfer (Verity and Smetacek, 1996). The role of top-down forces by predators has been well studied in terrestrial (e.g., wolves) and freshwater (e.g., bass) lake systems, but in the marine environment, most studies have been confined to intertidal areas where direct experimental manipulations of predators (e.g., starfish) are possible. Because traditional oceanographic approaches do not involve experimentation for obvious reasons, and because predator control of populations is not manifested in most observations of marine ecosystems, bottom-up forcing has been over-emphasized. We maintain that the adaptations of individual species are key to understanding long-term ecological change and that predation is a powerful structuring force in the GOA, especially in relation to marine mammal populations. The effects of sea otters on nearshore marine communities are very well documented and provide excellent examples of top-down control of marine populations: removal of predation by a loss of sea otters allows sea urchin populations (Strongylocentrotus spp.) to flourish and denude kelp forests (Macrocystis pyrifera), thus creating urchin barrens (Estes and Palmisano, 1974; Estes and Duggins, 1995). A trophic cascade (see Box 6.2), a term coined by Paine (1980), refers to such top-down effects across multiple trophic levels. The cascade does not stop with
BOX 6.2: TROPHIC CASCADES by Gordon H. Kruse Trophic cascades are often used to describe changes in primary producers owing to changes in carnivores, mediated through intermediary changes in herbivores. Under a relatively simple trophic cascade, reductions of carnivores lead to increases in herbivores and subsequent reductions in plants. The ability of sea otters to structure the nearshore kelp community through their predation on sea urchins provides an excellent Alaskan example. Likewise, experimental removal of starfish from rocky intertidal communities allows for mussels to quickly expand their distribution to effectively monopolize all available space in the intertidal zone at the expense of other species that require space, such as barnacles (Paine, 1980). Because the presence or absence of sea otters or starfish determines the diversity of nearshore and intertidal communities, respectively, they are called keystone species (Paine, 1969). Trophic cascade theory has been widely applied to terrestrial, freshwater, and nearshore marine ecosystems, such as rocky intertidal communities and coastal marine reefs. Applications to continental shelves and deep seas are less common,
540
Long-Term Ecological Change in the Northern Gulf of Alaska
because it is much more practical to experimentally manipulate lakes, land parcels, and local areas of rocky reefs and intertidal beaches. Though not universal, reviews and meta-analyses (i.e., analyses of the collective results of multiple studies) of such experiments have found that trophic cascades are relatively common phenomena (e.g., Schmitz et al., 2000; Carpenter et al., 1985; Carpenter and Kitchell, 1996). It has been argued that trophic cascades are more common in water than land (e.g., Strong, 1992), and a recent meta-analysis appears to confirm that top-down predator control of plant biomass is stronger in aquatic (including marine) ecosystems than terrestrial ecosystems (Shurin et al., 2002). Another “cascade hypothesis” has been proposed for Alaska by Merrick (1995, 1997) and NRC (1996, 2003). Although it has been termed a “cascade hypothesis,” strictly speaking, a cascade hypothesis involves top-down control through two or more trophic levels to the plants, whereas this broader application also involves effects propagated from the bottom up, which are more correctly termed “indirect food web interactions.” Because the term “cascade” has been associated with the hypothesis, we will continue to use the term “cascade hypothesis” in this instance. Also, whereas this hypothesis was mainly developed for the eastern Bering Sea, it has also been considered appropriate to the Gulf of Alaska, owing to some shared species trends among the regions (NRC 1996, 2003). This “cascade hypothesis” contends that changes in the North Pacific over the past 50 years are attributable both to the history of human exploitation of living resources and environmental changes. The hypothesis interconnects changes in some populations of marine mammals, seabirds, fish and invertebrates. In overview, the cascade hypothesis for the North Pacific Ocean involves the following key events (see Fig. 6.3). Intensive commercial whaling greatly reduced the populations of blue (Balaenoptera musculus), sei (B. borealis), fin (B. physalus), and humpback whales (Megaptera novaeangliae) off Alaska during the 1950s–1970s. In the 1960s and 1970s (see Section 4.9), large foreign trawl fisheries overfished stocks of Pacific Ocean perch (Sebastes alutus) in the Gulf of Alaska, Aleutian Islands and Bering Sea. Large reduction fisheries for herring are thought to have been associated with declines in herring stocks. In the Gulf of Alaska, a large domestic fishery was prosecuted for illegal fishing of herring in the 1930s to 1950s, while a large foreign trawl fishery was prosecuted in the eastern Bering Sea during 1960–1980. On the other hand, walleye pollock (Theragra chalcogramma) experienced a huge concurrent increase in biomass from 1960 to the early- to mid-1980s in both the Gulf of Alaska and eastern Bering Sea. All of these species have significant diet overlap: euphausiids (Thyanoessa spp.), calanoid copepods, and small schooling fishes. This cascade
Long-Term Changes in the GOA: Properties and Causes
541
Figure 6.3: The normalized changes in populations of major species in the Gulf of Alaska, including humans for several recent decades. hypothesis contends that the depletion of whales, Pacific Ocean perch and herring decreased the competition for prey for the pollock. As a result, pollock biomass demonstrated a positive numerical response to the increase in food availability. An additional mechanism leading to the increase in pollock in the Bering Sea may be a reduction in a pollock predator, the northern fur seal (Callorhinus ursinus) at the Pribilof Islands, owing to high seal harvests during 1950–1970. Whether this mechanism helps to explain the continued (since mid-1980s) high abundance of pollock in the eastern Bering Sea compared to the Gulf of Alaska is unclear. However, Steller sea lions, common to both areas but more abundant in the Gulf of Alaska, may have had comparable or complimentary effects to northern fur seals. Conversely, large increases in pollock may have increased predation on forage fishes, thus causing intense competition with other mammals, such as Steller sea lion (Eumetopias jubatus). The western stock of Steller sea lions, which occupies the Gulf of Alaska, Aleutian Islands and eastern Bering Sea, declined by approximately 80% during 1970–2000. On the contrary, recent Ecopath/Ecosim modeling of the eastern Bering Sea was unable to explain declines in Steller sea lions by trophic dynamics alone, so it appears that additional mechanisms need to be invoked to explain their decline (NRC, 2003). The recent decline in sea otters may be part of a long-term sequential megafaunal collapse triggered by industrial whaling after World War I and is probably another example of the cascade hypothesis (see Section 4.9). The hypothesis is
542
Long-Term Ecological Change in the Northern Gulf of Alaska
consistent with the sequential decline of these species in Alaskan waters, but whether top-down control by killer whales is the operative mechanism remains an open question. Regardless, the possibility of such delayed effects of whaling leads to a sobering implication about ways that one perturbation on the system can cause far-reaching effects on other components of the ecosystem over many decades.
the kelp. Kelp is important for the survival of the juveniles of some species of fishes (e.g., rockfishes, Sebastes spp.) and a whole associated group of fauna. Rockfish can, in turn, graze down barnacle nauplii (Gaines and Roughgarden, 1987). So, removal of sea otters can result in increases in the next lower trophic level (sea urchins), decreases in the level below sea urchins (kelp) and, perhaps, increases in barnacle larvae. Therefore, trophic and habitat dependencies involved in such cascades can ripple through a marine community in many, and perhaps unexpected, ways from removal of a top predator. The converse can also occur, so the expansion of remnant sea otter populations to areas beyond their historical range (e.g., Prince William Sound) is thought to be responsible for the depletion of Dungeness crabs and the elimination of an established fishery. Evidence indicates that this is now occurring in portions of Southeast Alaska. Reductions of sea urchins (Strogylocentrotus spp.), abalone (Haliotis spp.), Pismo clams (Tivela stultorum), and crab populations (Cancer spp.) occurred in California as remnant sea otter populations expanded. A review of the record reveals that the status of many fish, bird and mammal stocks of high value have been conditioned by centuries of human exploitation, the most intense occurring since the 1700s. The Steller sea cow was quickly driven to extinction, while sea otters, seals, and the great whales were systematically harvested to low levels without regard to conservation principles. In more recent times, Pacific Ocean perch, cod, and walleye pollock were targeted by large European and Asian distantwater factory fleets, and North American salmon were fished by Japanese driftnet fisheries on the high seas. Efficient factory trawling also accounted for a huge bycatch and mortality of non-targeted species, as well as losses of juvenile stages of targeted fishes. While not sought directly outside of traditional subsistence harvests, some marine birds are also vulnerable to fishing techniques (surface trawls, drift nets and long-lines) and thus impacted by these activities. Large numbers of seabirds were also captured by the Japanese high-seas driftnet fisheries for salmon. Surprisingly, it was the repercussions associated with the bycatch of Dall’s porpoise, and not the interception of illegal fishing of North American salmon, that led to the final demise of
Long-Term Changes in the GOA: Properties and Causes
543
these fisheries. In these and other ways, harvest mortalities must be accounted for in any attempt to describe the nature of change of exploited stocks in the Gulf of Alaska. Thus, we see no way to rule out the activities of modern humans as major players in affecting system structure and function. What seemed to be boundless and immune to human tampering, is now recognized as finite and fragile, particularly in the areas of resource utilization and environmental contamination. We see that the magnificent robustness and resilience of the natural system is vulnerable to events on rather short time scales for which evolutionary processes are ineffective – a “problem” of time.
6.3.5. Bottom-up and Top-down Operate Together There is a need to bring both top-down and bottom-up perspectives together into an integrated approach if we are really going to make greater progress toward understanding the dynamics of marine systems. The oscillating control hypothesis (OCH) offers one such approach for the eastern Bering Sea (see Box 6.3), which, perhaps, could have some application to the Gulf of Alaska, as well. Regardless, we believe that great advances can be made by studying the unique ways in which important species of predators, their prey, and mankind interact. We feel that more attention must be paid to predation, in the broad sense, and more sophisticated approaches employed to unite bottom-up and top-down forcing. One useful way to view this dichotomy in our conceptualization is that bottom-up forcing sets the theoretical limits of upper trophic level biomass and turnover, but the distribution of energy flow among species is an outcome of other factors, such as specific adaptations, predator–prey relationships, habitat dependencies, competition and prey-switching.
BOX 6.3: THE OSCILLATING CONTROL HYPOTHESIS by Gordon H. Kruse The oscillating control hypothesis (OCH) offers an explanation of changes in forage and piscivorous (i.e., fish-eating) fishes, marine birds, and pinnipeds in the southeastern Bering Sea (Hunt et al., 2002). The OCH asserts that the pelagic ecosystem is controlled by food supply in cold regimes and predation in warm regimes. Although the extent of winter sea–ice formation is a key driver in the application of this hypothesis for the Bering Sea, other mechanisms (e.g., timing of springtime upper water column stratification) could render the OCH applicable to the Gulf of Alaska where sea ice is not a factor.
544
Long-Term Ecological Change in the Northern Gulf of Alaska
Some of the major changes in the southeastern Bering Sea in the last 30 years (NRC, 1996; Hunt et al., 2002) parallel those in the Gulf of Alaska. For example, the most abundant groundfish, walleye pollock (Theragra chalcogramma), increased in both regions in the early- to mid-1980s due to the spawning of strong cohorts in the late 1970s, but, since then, pollock have steadily declined in the Gulf of Alaska (Dorn et al., 2002), whereas the eastern Bering Sea stock remains high (Ianelli et al., 2002). There are also some parallels in the behavior of some crab, salmon, seabird and marine mammal populations. (Dragoo et al., 2001; Kruse, 1998; Zheng and Kruse, 2000). The OCH contends that the couplings between forage fish and zooplankton and between forage fish and large predatory fish, such as pollock, arrowtooth flounder (Atheresthes stomias), and Pacific cod (Gadus macrocephalus), may be either top-down owing to predation or bottom-up from control of food supply (Hunt et al., 2002). The abundance of large piscivorous fish is crucial to the kind of control that predominates in the lower parts of the ecosystem. When large piscivorous fishes are sufficiently abundant, they exert top-down control of forage fishes, and the abundance of forage fishes is uncoupled, or independent, from zooplankton production. In this case, zooplankton may be abundant and respond more directly to phytoplankton production than when zooplankton populations are constrained by forage fish predation. On the other hand, when piscivorous fish are rare, they consume fewer larval fish and forage fish. In such a scenario, recruitment of both predatory and forage fishes is determined during the early life stages, and zooplankton abundance exerts bottom-up control on fish recruitment. However, temperature is another key to the hypothesis. Because cold temperatures inhibit zooplankton growth and reproduction, sea temperature conditions during a phytoplankton bloom affect the coupling of primary (i.e., phytoplankton) and secondary (i.e., zooplankton that feed on phytoplankton) producers as well as fish larvae and therefore fish larval survival. For instance, zooplankton and phytoplankton biomass and productivity are likely to be uncoupled during cold periods, because zooplankton are unable to increase their abundance and grazing pressure quickly enough to control the growth of phytoplankton. However, forage fish are able to respond to increased abundance and productivity of zooplankton through feeding success during warm periods only if large predatory fish are less in number and do not control forage fish abundance. The OCH connects these trophic linkages with climate regime shifts, as shown schematically in Fig. 6.4 (Hunt et al., 2002). During cold regimes (Fig. 6.4, top panel), sea ice is extensive, retreats late (April–May), and ice-edge plankton blooms occur in cold water associated with the formation of a halocline (Stabeno et al., 2001). Cold temperatures limit zooplankton productivity; so early life survival of fishes will
Long-Term Changes in the GOA: Properties and Causes
545
Figure 6.4: The oscillating control hypothesis. be low owing to short food supply. Presumably, most primary production falls ungrazed to the seafloor, thus enhancing benthic production. During this regime, fish recruitment will be weak, in general, owing to bottom-up constraints. Populations of pinnipeds and piscivorous birds may do well, if the distribution of forage fish, such as capelin, that prefer cold water shift their distribution toward rookeries and colonies. When a shift occurs to a warm climate regime (Fig. 6.4, second panel) as in the late 1970s, sea ice retreats early (prior to mid-March), and the spring bloom
546
Long-Term Ecological Change in the Northern Gulf of Alaska
is delayed until May or June in association with thermal stratification of the water column rather than due to haline stratification (Stebeno et al., 2001). Under these conditions, there is high zooplankton production, because the spring bloom occurs in warm water favorable to copepod reproduction. In the eastern Bering Sea, growth and reproduction of copepods are more strongly related to warm temperatures than to food availability (Smith and Vidal, 1984); copepod abundance increases two- to thirteen-fold in association with increases of mean water column temperatures of about 3°C in spring (Coyle and Pinchuk, 2002). High abundance of zooplankton and low abundance of adult fish predators favor large year classes of some fish stocks. Indeed, an increased frequency of strong year classes among a number of fish species occurred during the 1976–1977regime shift (Hollowed et al., 1987; Hollowed and Wooster, 1992). As the warm regime continues (Fig. 6.4, third panel), the biomass of piscivorous fish species becomes large owing to the accumulation of large year classes as adults. These predators exert top-down control of forage fishes (Hunt and Stabeno, 2002), providing more zooplankton to other consumers, such as jellyfish (Brodeur et al., 1999), chaetognaths, salmon and baleen whales. Abundance and productivity of piscivorous seabirds and pinnipeds declines owing to the reduced availability of forage fishes. Finally, with a shift to a cold regime (Fig. 6.4, bottom panel), cold temperatures limit zooplankton productivity and much of the primary production enters the benthos. Fish recruitment is severely depressed owing to both reduced availability of zooplankton and high rates of predation by the still-large biomass of fish predators. Owing to a lack of sea ice, there is much less variability in the timing of the spring bloom in the Gulf of Alaska. However, much more study of the spring bloom in the Gulf is sorely needed before a deeper understanding of its variability can be achieved. Although the timing of stratification of the water column through either the formation of a halocline (via sea ice melt) or thermocline (via solar heating) and its interplay with photoperiod determines the spring bloom timing in the Bering Sea, we can only speculate on the roles of wind mixing and/or freshwater to control water stratification and thus timing of the spring bloom in the Gulf. Both the wind field and freshwater inputs to the Gulf vary with the PDO, so hypotheses about their potential effects on the timing of the spring bloom are testable. In his original match–mismatch hypothesis, Cushing (1990) suggested that the phase of the North Atlantic Oscillation determined the timing of the spring bloom in the North Sea, depending on the direction and intensity of the winds that determined whether the water column would stratify (early bloom) or continue to mix (late bloom). As Atlantic cod egg-hatching time is relatively fixed,
Long-Term Changes in the GOA: Properties and Causes
547
variability in the spring bloom is speculated to determine the match between primary and secondary production and cod larvae. For the North Sea, Cushing postulated that cold winds from the northeast delayed the spring bloom, promoting a better match with hatching cod eggs. The applicability of this hypothesis to the Gulf of Alaska could set into place the bottom-up portion of the OCH. As we discussed in Section 2.5.5, there is considerable evidence for the role of temperature and cod predation on shrimp populations in the North Atlantic. We can envision that decadal shifts in the PDO can lead to shifts between topdown and bottom-up control in the Gulf. Periods of low cod abundance may foster bottom-up control of shrimp populations. On the other hand, climatic shifts that lead to bursts in cod recruitment (such as the strong year classes in the late 1970s) will lead to periods of top-down control owing to predation by cod on shrimp and other forage species. Because young cod feed on invertebrates (like shrimp and juvenile crabs) but old cod feed more on fishes, deterioration of conditions favorable to cod recruitment will lead to a shift toward older cod in the population and reduced predation on invertebrates. Thus, regardless of cause, shifts in periods of bottom-up and top-down control are plausible for the Gulf. Indeed, as in the Bering Sea, recruitment data on many groundfish stocks in the Gulf of Alaska indicate that strong year classes predominated when climate shifted from cold to warm regimes in the late 1970s. In modeling recruitment processes, Bailey (2000) showed that it was likely that there had been a shift from bottom-up control of pollock larval survival in the early 1970s to top-down control of juvenile pollock after the regime shift. As a result, the biomass of fishes, such as pollock and cod increased in the Gulf of Alaska, through the early(pollock) to mid- (cod) 1980s as strong year classes were recruited to spawning stocks, but abundance of both species declined with poorer recruitment as the warm regime persisted through the late 1980s and 1990s (Dorn et al., 2002; Thompson et al., 2002).
Moreover, the interplay of bottom-up and top-down controls must be viewed in a broader context than the Gulf of Alaska. Atmospheric teleconnections link the Gulf of Alaska with the Bering Sea and other regions of the Northern Hemisphere. So, changes in one area are linked to changes in the other. Thus, climate forcing of bottom-up processes should not be treated in isolation. Also, there are a number of strong connections between predators causing top-down control over a region much broader than the Gulf of Alaska. Killer whales are distributed throughout the Gulf of Alaska, Aleutian Islands and eastern Bering Sea. The great whales roam
548
Long-Term Ecological Change in the Northern Gulf of Alaska
throughout this region and beyond. Stocks of predatory fishes, such as Pacific halibut, are broadly distributed throughout the region, as well. Many larvae from halibut spawning in the Gulf drift into the eastern Bering Sea over their long (~6 month) larval life. In what is termed “counternatant” movement, many juvenile halibut migrate with ontogeny (age) from the eastern Bering Sea back to the Gulf of Alaska where they were spawned. So, halibut in the Gulf of Alaska and eastern Bering Sea are highly connected and are managed as one stock. For these reasons, at least some of the top-down and bottom-up processes can only be understood by recognizing the connections between the Gulf of Alaska and the Bering Sea.
6.3.6. Contaminants and Disease as Predators Contaminants are usefully viewed as another sort of predation on Gulf species. This is particularly so for the chronic, long-lasting contaminants in the system. Immediate low level, but growing industrial contamination and pollutant spills are threats worth examining as their occurrence becomes more common and their effects more widespread. As horrible as it was, the spill of crude oil from the grounded tanker Exxon Valdez into Prince William Sound was confined in its effects to a relatively small region of the Gulf. Just how serious the lingering negative influences (natural and cultural) of this incident will be, remains to be determined. Certainly at “ground zero” the region may see chronic influences for years or may never completely recover. One of the greatest legacies of the EVOS tragedy will be the published record describing how this disaster played out over time – both culturally and biologically. Contaminants lead to biochemical responses of the affected organisms – responses that detract from the allocation of metabolic energy to maintenance, growth and reproduction. Contaminants can also serve as direct toxins and as receptor-mediated interference in the normal communication within organisms as they self-organize through physiological adaptations to deal with changes in the system. Because contaminants often exert their greatest effects in the early life history stages of the organism, they are best viewed as “predators” acting on that stage. Once we understand how rates of predation on early life history stages affect population trajectories, we will also better understand the population effects of contaminants. The demands for energy among competing physiological processes within an organism might lead to weakened immune responses that rely on energy and protein as well. In this connection, continued global warming and potentially increasing levels of contamination in the North Pacific Ocean might lead to greater disease expression. Conceivably, warmer winter ocean temperatures could elevate basal metabolism of poikilotherms, and in the absence of significant primary production
Long-Term Changes in the GOA: Properties and Causes
549
and increased food supplies, cause them to deplete their energy reserves. Loss of energy reserves could lead to increased disease expression.
6.3.7. Structure of the Ecosystem Affects Ecological Change There are several aspects of the trophic structure of North Pacific food webs that have implications for long-term change. First, a very large part of the energy that is transferred through the food web passes through relatively few species. Studies of complex systems demonstrate that these types of assemblages, i.e., domination of energy or information transfer by a few nodes, are not as robust to perturbations as are those that more equitably distribute energy or information (Barabási and Bonabeau, 2003). Remarkably, ecosystems, cellular metabolic machinery, and computer systems seem to share common properties that are scaleless. The study of these properties may lead to new insights about the behavior of complex systems, which in turn might be applied to understanding the response of the Gulf of Alaska to physical or other forcing. Second, as one moves higher up the food web, the direct connection to ocean physics becomes more tenuous and other influences, such as interspecies competition and predation become proportionally more important in population regulation. On this basis it is expected that top trophic level populations, while ultimately limited by the amount of energy that is fixed by plants at the base of the food web, will more likely reflect other influences, including their own adaptive processes. Third, there are specific times/places during the year where large percentages of energy transfer occur very efficiently (short food webs; high rates of productivity). As noted earlier, this makes the transfer of energy to the top of the food web more efficient because of the short half-life of energy at some trophic levels, and is also likely to reduce the foraging effort of predators that can greatly benefit from aggregations of prey (Section 2.4). On the other hand, mismatches between predators and their prey may result from such bursts of production.
6.3.8. Interaction of Multiple Forcing Factors and Cascading Effects It can be seen from the examples discussed in the first part of this section that the largescale perturbations interact with other changes in the system. This makes any simple interpretation of the impact and recovery of a system disturbed by a single large event, an unrealistic model for understanding what is really happening. The marine ecosystems of the Gulf of Alaska are constantly changing and there will be interactions between the various forcing functions. The examples are abundant. One is the interaction between climate-induced changes in the productivity of Prince William Sound, as evidenced by
550
Long-Term Ecological Change in the Northern Gulf of Alaska
differences in springtime standing stocks of plankton between the 1980s and 1990s, and the recovery of impacted populations from the 1989 oil spill. For some recovering species, e.g., Pacific herring, there was simply less to eat after the spill than in the previous decade. In addition, the night-time feeding of sea lions on herring aggregations in Montague Strait (where most of the PWS herring congregate in winter) could have slowed or reversed the recovery – perhaps another non-linear aspect of ecosystem dynamics). So, to ask simple questions about the time to recover from a large oil spill to suit human needs, is assuming that the ecosystem has simpler dynamics than it probably does. Cascading and indirect effects are likely to be important and quite large, especially in consideration of the role of humans in the marine ecosystem. There are several important examples in the Gulf of Alaska that we touched on in preceding chapters. a. Removal of sea otters by the spill altered prey abundance. There were other changes in intertidal ecology, as disruptions caused by the oil spill set into action a series of cascades due to changes in predation and sheltering habitat (see Peterson, 2000; Peterson et al., 2003). b. Extensive historical whaling in the North Pacific Ocean may well have been responsible for the shift of killer whales from predating great whales to other important marine mammals, resulting in their serial collapse throughout the Gulf of Alaska. The multi-factorial aspect of ecosystem forcing and response leads us to a further conclusion: the pulsing of the system by perturbations, either natural or anthropogenic, will not always have the same biological response; the outcome depends on the state of the system at the time. For example, the conditioning of the spring mixed layer by winter physical conditions sets up primary productivity to be either intense and short or delayed and extended. Likewise, there is no evidence that different regimes return to exactly the same states after the perturbation has switched back to an earlier condition. Different outcomes from similar perturbations at different times are the hallmarks of complex system behavior.
6.3.9. Prediction of Ecological Change It is unlikely that an overall simulation model for forecasting ecosystem behavior and future trajectories of key species will succeed because of (1) mismatches in scale between climate, physics and biology; (2) lack of data on energy transfer coefficients between species; and (3) non-linear and chaotic relationships caused by feedbacks in the system that are most likely poorly characterized or unknown. An alternative to understanding and managing man’s activities in the Gulf ecosystem is a comprehensive monitoring strategy that provides a picture of current conditions and a history of past conditions.
Long-Term Changes in the GOA: Properties and Causes
551
However, we emphasize that all aspects of the ecosystem are not too complicated to understand and that progress is being made on how to best conceptualize the way it works. That is, we can probably understand many of the important coupling relationships/ processes and determine cause and effect, but may not be able to rely much on forecasts of future states (the forcing is “complex” and the system response is also “complex”). Nonetheless, we should be capable of broad predictions of what will happen, given a range of scenarios in future meteorological conditions. With regard to anthropogenic impacts, we have considerable experience with oil spills and can probably predict some of their major effects, if the type and volume of oil, the conditions under which it is spilled, and the nature of the shoreline it encounters are known in advance. Ecosystems are complex, but their behavior is not necessarily complicated. What we mean by this is that marine ecosystems have myriad interactions and feedbacks that make some kinds of predictions very difficult if not impossible; their behavior is complex. However, on some time and space scales they are predictable, as past and present behavior may be able to explain to a certain degree. Correlations through several climate cycles, for instance, might yield repeating patterns that make some future outcomes predictable within wide bounds. This certainly seems to hold for some features of the North Atlantic for example, where there are long-term records and the fluctuations of some fish stocks in the northeastern Atlantic are broadly correlated with the North Atlantic Oscillation (Alheit and Hagen, 1997). Likewise in the North Pacific, many salmon stocks have fluctuated in response to the PDO during the past hundred years. However, there are severe limitations on the predictability of ecological change, given the multiple factors potentially acting in linear and non-linear fashions both directly and indirectly on marine populations (see Box 6.4).
BOX 6.4: IS THE GULF OF ALASKA A SELF-ORGANIZED CRITICAL SYSTEM? by Robert B. Spies and Gordon H. Kruse Before defining what we mean by this question, we first provide a little background for its motivation. Many scientists are coming around to the view that, while it may be possible to develop a retrospective understanding of the behavior and dynamics of complex systems, it is highly unlikely that future changes can be predicted with any precision. This asymmetry of being able to understand the past, but not the future very well, is taking root among scientists and mathematicians from diverse disciplines, including geology, astrophysics, and ecology. The classical reductionist approach, which led to such spectacular advances in our understanding of nature since Sir Isaac Newton, has not been nearly as fruitful to scientists studying complex systems. One view is that, if we just study more of the details – that is, everything about the bits and pieces and
552
Long-Term Ecological Change in the Northern Gulf of Alaska
how each is linked to another – we will ultimately be able to develop sufficient understanding of complex systems to make predictions possible. For example, if the trophic linkages are better known then perhaps we can better predict how changes in primary production will be manifested in the production of a particular species of fish. The opposing view is that understanding ever-smaller scales and more details will never allow us to predict the behavior of a complex system because of unpredictable internal changes within the system itself. Support for the latter view comes from a growing list of futile attempts to develop ever more complicated models that do not behave like nature. Although there are examples in which ecologists have developed models of populations and their interactions with predators and prey with some measure of predictive success, the outlook is much more dismal for whole ecosystems. Now back to defining our question: What is a “self-organized critical system” anyway? In the late twentieth century, physicists discovered that systems may be “self organized critically” – a concept that relates to the study of complexity in nature. A critical point is a point at which the system radically changes its structure or behavior often in response to what appears to be only a small incremental change in external conditions. An example is the observed change in structure and behavior as one applies heat to turn an ice cube into water. Ice and liquid water have very different properties. In this case, the heat was applied to the ice. Self-organized critical systems attain a critical state by virtue of internal dynamics of the systems in response to pulsing by an external stimulus. Some complex systems apparently self-organize to criticality without external forcing. At some point the next pulse of the system results in a critical change in the system. The classic example results from the dropping grains of sand to form a pile. Eventually, the pile reaches a height such that the intermittent flow of grains down the slope maintain a balance in which the outflux over the edge of the pile on average equals the rate at which new grains are added, but the changes are not a linear response to the external force. So, the pile actually self-organizes itself into a state where avalanches of all sizes may occur. The avalanches take on a range of sizes with a size–frequency spectrum expressed by what mathematicians and statisticians refer to as a power function. This concept of self-organized criticality has entered into the prediction of forest fires, earthquakes, and other natural phenomena, including the appearance of species in the fossil record (Bak, 1997). Following the sand pile analogy, it has been proposed that the successive introduction of new bird species to the Hawaiian Islands has built up repeatedly to a critical number of species resulting in “avalanches” of species’ extinctions over the past century (Keitt and Marquett, 1996). Just as the avalanches of sand grains take on a range of sizes,
Long-Term Changes in the GOA: Properties and Causes
553
Figure 6.5: The distribution of extinction event sizes in introduced bird fauna of the Hawaiian Islands. The trend line is a power function with an exponent of 0.91. After Keitt and Marquet, 1996.
likewise, the resultant cascades of extinction have included varying numbers of species that suggest a power function (Fig. 6.5). In these self-organized critical systems, the history of perturbations plays a key role in change. Successive perturbations lead the system to self-organize so system changes are critical failures with the characteristic size–frequency spectrum that can be represented by a power function. When perturbations are applied, these complex systems do not change much at first. However, later as the system evolves under additional perturbations, more and more individual elements of the system become modified as a result of the perturbations and the critical behavior begins to emerge. So, we ask: Is the northern Gulf of Alaska ecosystem a self-organizing critical system? And, the corollary question is: Do the ecosystem changes follow the characteristic power distribution of systems that are self-organized critically? Unfortunately, it seems that we cannot answer these questions at this time, because of inadequate data. If we consider the annual changes in the northern Gulf of Alaska, then our records of marine ecosystem change, spanning 30 years or so, are simply not long enough to determine such behavior. If we focus on the ecosystem perturbations from climatic sources, we have observed one relatively large regime shift in the late 1970s, a moderate shift in 1989, another apparent one ~o-scale (every 2–7 years) and in 1999 and some smaller perturbations from El Nin
554
Long-Term Ecological Change in the Northern Gulf of Alaska
other annual changes in the climate. So, we do not have enough repeat observations of the northern Gulf of Alaska to know whether such perturbations have led to any emergent ecosystem behaviors. In further reflecting on this question for the northern Gulf of Alaska, we draw an analogy to the field of meteorology. Meteorologists have developed the largest monitoring program on the planet involving continuous observations of synoptic weather at countless locations around the globe and an equally impressively huge computer modeling system. Even with this level of detail, the system is rather weakly predictable up to about 5–10 days at most, with only general predictions (wetter than normal, warmer than normal) that sometimes characterize expectations for months ahead. However, even with this limited ability to predict the future state of atmospheric weather conditions, society still deems this expense to be well worth its price tag. Indeed, weather forecasts serve a wide variety of human services. So, our task to understand the northern Gulf of Alaska continues, as we strive to assure that it will continue to provide commercial, recreational, educational, and spiritual ecosystem services to future generations. Whether or not this marine ecosystem is self-organized critically remains to be seen, but our investments in monitoring and research will not only assure that marine resource managers have the best near-term forecasts available for their decision making, but also these same observations will help to document future emergent system behaviors over the long term.
By approaching ecological changes through the adaptations of a handful of the most numerous fish, bird and mammal species and their historical changes, we have emphasized processes on the population level. Classical population dynamics tells us that the state of a consumer stock at any time represents a balance between the additions through reproduction and the numbers of individuals lost (deaths) over some period of time. Because the reproductive potential for most fishes and invertebrates is very high, small differences associated with reproductive successes can translate into significant additions or losses to stocks each year. For example, an adult female herring can spawn 6000 eggs, so a 1% change (up or down) in survival of these eggs results in plus/minus 60 adults added or subtracted per female to/from the stock. An adult female cod can spawn 1 million eggs, so a similar shift in survival – plus/minus 1 % – can result in 10,000 adults added to or subtracted per female. These insights demonstrate that only tiny differences in reproductive success can significantly tilt a stock survival trajectory up or down when averaged over the entire population, particularly if, for example, one year of success follows another so additions compound one another. Large fluctuations in the abundance of strongly interacting species, such as
Long-Term Changes in the GOA: Properties and Causes
555
pollock and arrowtooth flounder, can cause large effects in ecosystems. As the abundance of arrowtooth flounder is at its highest level since stock assessments began three decades ago, and continuing to increase, the direct and indirect effects of this predator on the Gulf of Alaska are beyond the range of historical observations. While some modelers are beginning to grapple with the dynamics of change as they relate to the abundance of species, most are still engaged in predicting biochemical cycling and trophic level biomass dynamics, Thus, our ability to understand systemwide changes is continually challenged. In summary, our synthesis of available information on the northern Gulf of Alaska provides much new insight about changes of this complex marine ecosystem. It is our hope that others, following up with new observations and ideas, will find this a useful conceptual basis for further refinement. We encourage additions to the richness of an evolving discussion and understanding in the hope that, in time, it will yield greater predictability and application. This type of work is tedious and slow, yet ultimately rewarding as we begin to understand in a more comprehensive way, the nature of ecosystem change that occurs on many interacting temporal and spatial scales. This glimpse of organization dissolves any vestige of simple interpretation and points future studies toward the hard work necessary to make further progress. The first tentative steps have been taken – it remains to be seen where this pathway will lead.
References Alheit, J. and E. Hagen. 1997. Long-term climate forcing of Europe herring and sardine populations. Fish. Oceanogr. 6: 130–139. Anderson, P. J. and J. F. Piatt. 1999. Community reorganization in the Gulf of Alaska following ocean climate regime shift. Mar. Ecol. Progr. Ser. 189: 117–123. Bailey, R. G. 1980. Description of the Ecoregions of the United States. United States Department of Agriculture Miscellaneous Publications No. 1391, Washington DC, USA, 77 pp. Bailey, R. G. 1998. Ecoregions: The Ecosystem Geography of the Oceans and the Continents. Springer-Verlag. New York. Bailey, K. M. 2000. Shifting control of recruitment of walleye pollock Theragra chalcogramma after a major climatic and ecosystem change. Mar. Ecol. Progr. Ser. 198: 215–224. Bak, P. How Nature Works: The Science of Self-Organized Criticality. 1997. Oxford University Press. 205 pp. Barabási, A.-L. and E. Bonabeau. 2003. Scale-Free Networks. Sci. Amer. 288: 60–69. Batten, S. D., and D. W. Welch. 2004. Changes in oceanic zooplankton populations in the north-east Pacific associated with the possible climatic regime shift of 1998/1999. Deep Sea Res. Part. II: 863–873. Batten, S. D., K. D. Hyrenbach, W. J. Sydeman, K. H. Morgan, M. F. Henry, P. P. W. Yen, and D. W. Welch. 2005. Characterising meso-marine ecosystems of the North Pacific. Deep-Sea Res. II 53: 270–290.
556
Long-Term Ecological Change in the Northern Gulf of Alaska
Belkin, I. M. and P. C. Cornillon. 2003. SST fronts of the Pacific coastal and marginal seas. Pacific Oceanogr. 1(2): 90–113. Blanchard, A. L., H. M. Feder, and D. G. Shaw. 2002. Long-term investigation of benthic fauna and the influence of treated ballast water disposal in Port Valdez, Alaska. Mar. Pollut. Bull. 44: 367–382. Bond, N. A., J. E. Overland, M. Spillane, and P. J. Stabeno. 2003. Recent shifts in the state of the North Pacific. Geophys. Res. Lett. 30: 2183. (Abstract) Brodeur, R. D. and D. M. Ware. 1992. Long-term variability in zooplankton biomass in the subarctic Pacific Ocean. Fish. Oceanogr. 1: 32–38. Brodeur, R. D., C. E. Mills, J. E. Overland, G. E. Walters, and J. D. Schumacher. 1999. Evidence for a substantial increase in gelatinous zooplankton in the Bering Sea, with possible links to climate change. Fish. Oceanogr. 8: 296–306. Call, K. A. and T. R. Loughlin. 2005. An ecological classification of Alaskan Steller sea lion (Eumetopias jubatus) rookeries: a tool for conservation/management. Fish. Oceanogr. 14: 212–222. Carpenter, S. R., J. F. Kitchell, and J. R. Hodgson. 1985. Cascading trophic interactions and lake productivity. Bioscience 35: 634–639. Carpenter, S. R. and J. F. Kitchell, (Eds.). 1996. The Trophic Cascade in Lakes. Cambridge University Press. Cambridge, United Kingdom. Chavez, F. P., J. Ryan, S. E. Lluch-Cota, and M. C. Niquen. 2003. From anchovies to sardines and back: Multidecadal change in the Pacific Ocean. Science 217–221. Coachman, L. K. 1986. Circulation, water masses, and fluxes on the southeastern Bering Sea shelf. Cont. Shelf Res. 5: 23–108. Cooney, R. T. 1981. Bering Sea zooplankton and micronekton communities with emphasis on annual production. In: D. W. Hood and J. A. Calder (Eds.), The Eastern Bering Sea Shelf: Oceanography and Resources. Vol. 2, NOAA, Juneau, Alaska, pp. 947–974. Coyle, K. O. and A. I. Pinchuk. 2002. Climate-related differences in zooplankton density and growth on the inner shelf of the southeastern Bering Sea. Progr. Oceanogr. 55: 177–194. Cushing, D. H. 1990. Plankton production and year-class strength in fish populations: an update of the match/mismatch hypothesis. Adv. Mar. Biol. 26: 249–292. Demarchi, D. A. 1996. An introduction to the ecoregions of British Columbia. Ministry of Environment, Lands and Parks, Victoria, British Columbia. Available on the web at: Dorn, M., S. Barbeaux, M. Guttormsen, B. Megrey, A. Hollowed, E. Brown, and K. Spalinger. 2002. Assessment of the walleye pollock stock in the Gulf of Alaska, In: Stock assessment and fishery evaluation report for the groundfish resources of the Gulf of Alaska. North Pacific Fishery Management Council, Anchorage, AK, pp. 33–88. Doyle, M. J., K. L. Mier, M. S. Busby, and R. D. Brodeur. 2002. Regional variation in springtime ichthyoplankton assemblages in the northeast Pacific Ocean. Prog. Oceanogr. 53: 247–281. Dragoo, D. E., G. V. Byrd, and D. B. Irons. 2001. Breeding status, population trends and diets of seabirds in Alaska, 2000. US Fish and Wildlife Service, Report AMNWR 01/07, Homer, Alaska. Dragoo, D. E., G. V. Byrd, and D. B. Irons. 2003. Breeding status, population trends and diets of seabirds in Alaska, 2001. U.S. Fish and Wildl. Serv. Report AMNWR 03/05. Estes, J. A. and J. F. Palmisano. 1974. Sea otters: their role in structuring nearshore communities. Science 185: 1058–1060.
Long-Term Changes in the GOA: Properties and Causes
557
Estes, J. A. and D. O. Duggins. 1995. Sea otters and kelp forests in Alaska: generality and variation in a community ecological paradigm. Ecol. Monogr. 65: 75–100. Favorite, F., A. J. Dodimead, and K. Nasu. 1976. Oceanography of the subarctic Pacific region, 1960–1971. Bull. Inter. North Pacific Fisheries Commission, 33: 1–187. Feely, R. A., C. L. Sabine, K. L. Lee, W. Berelson, J. Kleypas, V. J. Fabry, and F. J. Millero. 2004. Impact of Anthropogenic CO2 on the CaCO3 System in the Oceans, Science 305: 362–366. Finnegan, J. 2003. Earth system science in the early anthropocene era. IGBP Newsletter 55: 8–11. Francis, R. C. and S. R. Hare. 1994. Decadal-scale regime shifts in the large marine ecosystems of the North-east Pacific: A case for historical science Fish. Oceanogr. 3: 279–291. Fritz, L. W., A. Greig, and R. Reuter. 1998. Catch-per-unit-effort, length, and depth distributions of major groundfish and bycatch species in the Bering Sea, Aleutian Islands and Gulf of Alaska regions based on groundfish fishery observer data. NOAA Technical Memorandum NMFS-AFSC-88. Seattle, Washington. Gaines, S. D. and J. Roughgarden. 1987. Fish in offshore kelp forests affect recruitment to intertidal barnacle populations. Science 235: 479–480. Hare, S. R. and N. J. Mantua. 2000. Empirical evidence for North Pacific [climatic] regime shifts in 1977 and 1989. Prog. Oceanogr. 47: 103–145. Hollowed, A. B., K. M. Bailey, and W. S. Wooster. 1987. Patterns in recruitment of marine fishes in the Northeast Pacific Ocean. Biol. Oceanogr. 5: 99–131. Hollowed, A. B. and W. S. Wooster. 1992. Variability of winter ocean conditions and strong year classes of Northeast Pacific groundfish. ICES Mar. Sci. Symp. 195: 433–444. Hollowed, A. B., S. R. Hare, and W. S. Wooster. 2001. Pacific Basin climate variability and patterns of Northeast Pacific marine fish production. Prog. Oceanogr. 49: 257–282. Hunt, G. L. Jr. and P. J. Stabeno. 2002. Climate change and the control of energy flow in the southeastern Bering Sea. Prog. Oceanogr. 55: 5–22. Hunt, G. L. Jr., P. Stabeno, G. Walters, E. Sinclair, R. D. Brodeur, J. M. Napp, and N. A. Bond. 2002. Climate change and control of the southeastern Bering Sea pelagic ecosystem. Deep-Sea Res. II 49: 5821–5853. Ianelli, J. N., S. Barbeaux, T. Honkalehto, G. Walters, and N. Williamson. 2002. Eastern Bering Sea walleye pollock stock assessment. Pages 330120 In: Stock Assessment and Fishery Evaluation Report for the Groundfish Resources of the Bering Sea/Aleutian Islands Regions. North Pacific Fishery Management Council, Anchorage, AK. Jahncke, J., K. O. Coyle, and G. L. Hunt, Jr. 2005. Seabird distribution, abundance and diets in the eastern and central Aleutian Islands. Fish. Oceanogr. 14: 160–177. Johnson, S. W., A. D. Neff, and J. F. Thedinga. 2005. An atlas on the distribution and habitat of common fishes in shallow nearshore waters of southeastern Alaska. U.S. Dep. Commer., NOAA Tech. Memo. NMFS-AFSC-157. 89 p. Keitt, T. H. and P. A. Marquet. 1996. The introduced Hawaiian avifauna reconsidered: Evidence for self-organized criticality? J. Theor. Biol. 182: 161–167. Kildaw, S. D., D. B. Irons, D. R. Nysewander, and C. L. Buck. 2005. Formation and growth of new seabird colonies: the significance of habitat quality. Mar. Ornith. 33: 49–58. Knowlton, N. 1992. Thresholds and multiple stable states in coral reef community dynamics. Am. Zool. 32: 674–682. Krebs, C. T. and K. A. Burns. 1977. Long-term effects of an oil spill on populations of a salt marsh crab Uca pugax. Science 197: 484.
558
Long-Term Ecological Change in the Northern Gulf of Alaska
Kruse, G. H. 1998. Salmon run failures in 1997–1998: a link to anomalous ocean conditions? Alaska Fish. Res. Bull. 5: 55–63. Ladd, C., G. L. Hunt, C. Mordy, S. Salo, and P. Stabeno. 2005. Marine environments of the central and eastern Aleutian Islands: physical mechanisms, nutrient availability, and primary production. Fish. Oceanogr. 14: 22–38. Lanksbury, J. A., J. T. Duffy-Anderson, K. L. Mier, M. T. Wilson. 2005. Ichthyoplankton abundance, distribution, and assemblage structure in the Gulf of Alaska during September 2000 and 2001. Estuar. Coast. Shelf Sci. 64: 775–785. Levings, S. C., S. D. Garrity, and K. A. Burns. 1994. The Galeta oil spill. 3. Chronic reoiling, long-term toxicity of hydrocarbon residues and effects on epibiota in the mangrove fringe. Estuar. Coast. Shelf Sci. 38: 365–395. Logerwell, E. A., K. Aydin, S. Barbeaux, E. Brown, M. E. Conners, S. Lowe, J. W. Orr, I. Ortiz, R. Reuter, and P. Spencer. 2005. Geographic patterns in the demersal ichthyofauna of the Aleutian Islands shelf. Fish. Oceanogr. 14: 93–112. Longhurst, A. R. 1998. Ecological Geography of the Sea. Academic Press, San Diego, 398 p. Merrick, R. L. 1995. The relationship between the foraging ecology of Steller sea lions (Eumetopias jubatus) to their population decline in Alaska. University of Washington, Ph.D. dissertation, Seattle. 175 pp. Merrick, R. L. 1997. Current and historical roles of apex predators in the Bering Sea ecosystem. J. Northwest Atlantic Fish. Sci. 22: 343–355. Nowacki, G., P. Spencer, M. Fleming, T. Brock, and T. Jorgenson. 2001. Ecoregions of Alaska, U.S. Geological Survey Open-File Report 02-297. NRC (National Research Council). 1996. The Bering Sea ecosystem. National Academy Press, Washington, DC. NRC (National Research Council). 2003. Decline of the Steller sea lion in Alaskan waters: untangling food webs and fishing nets. National Academy Press, Washington, DC. Paine, R. T. 1969. A note on trophic complexity and community stability. The American Naturalist 103: 91–93. Paine, R. T. 1980. Food webs: Linkage, interaction strength and community infrastructure. J. Anim. Ecol. 49: 667–685. Peterson, C. H. 2000. The Exxon Valdez Oil Spill in Alaska: Acute, indirect and chronic effects on the ecosystem. Adv. Mar. Biol. 39: 3–84. Peterson, C. H., S. D. Rice, J. W. Short, D. Esler, J. L. Bodkin, B. E. Ballachey, and D. B. Irons. 2003. Long-Term Ecosystem Response to the Exxon Valdez Oil Spill. Science (Wash.) 302: 2082–2086. Piatt, J. F. and P. J. Anderson. 1996. Response of common murres to the Exxon Valdez oil spill and long-term changes in the Gulf of Alaska marine ecosystem, In: S. D. Rice, R. B. Spies, D. A. Wolfe, and B. A. Wright (Eds.), Exxon Valdez oil spill Symposium Proceedings, Amer. Fish. Soc. Symp. 18. Bethesda, MD. pp. 720–737. Piatt, J. F. and A. M. Springer. 2003. Advection, pelagic food webs, and the biogeography of seabirds in Beringia. Marine Ornithol. 31: 141–154. Piatt, J. F., G. Drew, T. Van Pelt, A. Abookire, J. L. Nielsen, M. Shultz, and A. Kitaysky. 1999. Biological effects of the 1997/98 ENSO in Cook Inlet, Alaska Proceedings of the 1998 Science Board Symposium on the Impacts of the 1997/98 El Nino event on the North Pacific Ocean and its marginal seas. Sidney, BC, Canada, PICES Sci. Rep. 10: 93–99.
Long-Term Changes in the GOA: Properties and Causes
559
Piat, J. F. and T. I. Van Pelt 1993. A wreck of common murres (Uria aalge) in the northern Gulf of Alaska during February and March of 1993. US Fish and Wildlife Service, Anchorage, Alaska. Piatt, J. F., J. Wetzel, K. Bell, A. R. DeGange, G. R. Balogh, G. S. Drew, T. Geernaert, C. Ladd, and G. V. Byrd. 2006. Predictable hotspots and foraging habitat of the endangered short-tailed albatross (Phoebastria albatrus) in the North Pacific: Implications for conservation. Deep Sea Res. II 53: 387–398. Reid, P. C., M. Edwards, H. G. Hunt, and A. Warner. 1998. Phytoplankton change in the North Atlantic. Nature 391: 546. Ruhl, H. A. and K. L. Smith, Jr. 2004. Shifts in deep-sea community structure linked to climate and food supply. Science 305: 513–515. Sabine, C. L., R. A. Feely, N. Gruber, R. M. Key, K. Lee, B. Kitack, J. L. Bullister, R. Wanninkhof, C. S. Wong, D. W. R. Wallace, B. Tilbrook, T. Bronte; F. Millero, T-H. Peng, A. Kozyr, T. Alexander, Ono, and A. Rios. 2004. The oceanic sink for anthropogenic CO2. Science 305: 367–371. Sanders, H. L., J. F. Grasle, G. R. Hampson, L. S. Morse, S. Garner-Price, and C. C. Jones. 1980. Anatome of an oil spill: Long-term effects from the grounding of the barge Florida off West Falmouth, Massachusetts. J. Mar. Res. 38: 265–380. Schell, D. M., B. A. Barnett, and K. A. Vinette, 1998. Carbon and nitrogen isotope ratios in zooplankton of the Bering, Chukchi and Beaufort seas. Mar. Ecol. Prog. Ser. 162: 11–23. Schmitz, O. J., P. A. Hambäck, and A. P. Beckerman. 2000. Trophic cascades in terrestrial systems: a review of the effects of carnivore removal on plants. Amer. Naturalist 155: 141–153. Schumacher, J. D., G. H. Kruse, and S. A. Macklin (Eds.). 2005. The Aleutian ecosystem: processes controlling variability in productivity and ecosystem structure. Fish. Oceanogr. 14: 1–306. Sherman, K., L. M. Alexander, and B. D. Gold (Eds.). 1990. Large Marine Ecosystems: Patterns, Processes and Yields. Amer. Assoc. Adv. Sci., Washington D.C., 242 p. Shurin, J. B., E. T. Borer, E. W. Seabloom, K. Anderson, C. A. Blanchette, B. Broitman, S. D. Cooper, and B. S. Halpern. 2002. A cross-ecosystem comparison of the strength of trophic cascades. Ecol. Lett. 5: 785–791. Sinclair, E. H., and P. J. Stabeno. 2002. Mesopelagic nekton and associated physics of the southeastern Bering Sea. Deep Sea Res. II. 49: 6127–6145. Sinclair, E. H., S. E. Moore, N. A. Friday, T. K. Zeppelin, and J. M. Waite. 2005. Do patterns of Steller sea lion (Eumetopias jubatus) diet, population trend and cetacean occurrence reflect oceanographic domains from the Alaska Peninsula to the central Aleutian Islands? Fish. Oceanogr. 14: 223–242. Smith, S. L. and J. Vidal. 1984. Spatial and temporal effects of salinity, temperature, and chlorophyll on the communities of zooplankton in the southeastern Bering Sea. J. Mar. Res. 4: 221–257. Speckman, S., J. F. Piatt, C. Minte-Vera, and J. Parrish. 2005. Parallel structure among environmental gradients and three trophic levels in a subarctic estuary. Prog. Oceanogr. 66: 25–65. Springer, A. M. and C. P. McRoy. 1993. The paradox of pelagic food webs in the northern Bering Sea: III. Patterns of primary production. Cont. Shelf Res. 13: 575–599. Springer, A. M., C. P. McRoy, and K. R. Turco. 1989. The paradox of pelagic food webs in the northern Bering Sea: II. Zooplankton communities. Cont. Shelf Res. 9: 359–386.
560
Long-Term Ecological Change in the Northern Gulf of Alaska
Springer A. M., C. P. McRoy, and M. V. Flint. 1996. The Bering Sea Green Belt: shelfedge processes and ecosystem production. Fish. Oceanogr. 5: 205–223. Springer, A. M., J. F. Piatt, V. P. Shuntov, G. B. van Vliet, V. L. Vladimirov, A. E. Kuzin, and A. S. Perlov. 1999. Marine birds and mammals of the Pacific Subarctic Gyres. Prog. Oceanogr. 43: 443–487. Stabeno, P. J., N. A. Bond, N. B. Kachel, S. A. Salo, and J. D. Schumacher. 2001. On the temporal variability of the physical environment over the south-eastern Bering Sea. Fish. Oceanogr. 10: 81–98. Strong, D. R. 1992. Are trophic cascades all wet? Differentiation and donor control in speciose ecosystems. Ecology 73: 747–754. Teal, J. M., J. W. Farrington, K. A. Burns, J. J. Stegeman, B. W. Tripp, B. Woodin, and C. Phinney. 1992. The West Falmouth oil spill after 20 years: Fate of fuel oil compounds and effects on animals. Mar. Poll. Bull. 24: 607–614. Thompson, G. G., H. H. Zenger, and M. W. Dorn. 2002. Assessment of the Pacific cod stock in the Gulf of Alaska. In: Stock Assessment and Fishery Evaluation Report for the Groundfish Resources of the Gulf of Alaska. North Pacific Fishery Management Council, Anchorage, AK, pp. 89–168. Verity, P. G. and V. Smetacek. 1996. Organism life cycles, predation, and the structure of marine pelagic ecosystems. Mar. Ecol. Prog. Ser. 130: 277–293. Zheng. J. and G. H. Kruse. 2000. Recruitment patterns of Alaskan crabs and relationships to decadal shifts in climate and physical oceanography. ICES J. Mar. Sci. 57: 438–451.
Subject Index abalone (Haliotis spp.), 378, 542 abiotic stresses, 307 Acartia, 83 acenaphthaylene, 448 ADF&G herring program, 292, 379 adult euphausiids, 64 adult puffins, 101, 109 adult survival, 104, 110–114, 124, 125 advection, 20, 30, 32, 45, 169, 176, 267, 307, 378 water masses, 32 AFK hatchery, 276 Age-Structured Assessment (ASA), 292, 293 Age-Weight-Length (AWL), 291 aggregating behavior, 70 agonist, 520 AIDS, 220 air–sea heat flux, 26 Akalura lake, 3, 280 Akutan, 193 Alaska Board of Fisheries (BOF), 201, 204, 208 Alaska Coastal Current (ACC), 11, 19, 31–40, 42, 60, 88, 266, 268–232, 424, 432, 435, 457, 537 front, 32, 34, 36 Alaska Department of Fish and Game (ADF&G), 77, 79, 197, 201, 206, 208, 215, 279, 284, 301, 423 small-mesh survey, 301 Alaska Department of Fisheries (ADF), 197, 200 Alaska Fisheries Board (AFB), 200
Alaska Fisheries Science Center (AFSC), 300, 301 Alaska North Slope, 419, 430, 456, 476, 483 Alaska Packers Association (APA), 278 Alaska Peninsula, 5, 14, 16, 18, 22, 29, 40, 60, 198, 303, 305, 316, 321, 329, 352, 354, 360, 364, 375, 379, 380, 382, 383, 389, 394, 419, 421, 432, 442 precipitation, 11, 15, 18, 20, 44 Alaska, State of, 214 South Central Alaska, 196 Southeast Alaska, 193, 204, 209, 210 Alaskan continental slope, 269 Alaskan natives, 196, 203, 204, 290, 428, 432 Alaskan Stream, 19, 32, 39, 40, 42, 46, 47, 88 alcids, 98, 100, 101 Aleutian Archipelago, 126, 356, 360 Aleutian Basin, 86 Aleutian Low(s), 17, 18, 45, 59, 172, 178–180, 268, 274, 293, 294, 382, 388 pressure, 293 Aleutian Low Pressure Index (ALPI), 287–289 Aleutian Low Pressure System (ALPS), 2, 17, 59, 378, 387, 521, 528, 529 Aleutian Peninsula, 34 Aleutian Range, 14 alevins, 182, 279, 284, 282, 288, 468
562
Subject Index
algal canopy, 463 blue-green algae, 230 Caulerpa taxifolia, 188 alkane hydrocarbons, 429–431 alkyl groups, 429 alkylated metals, 241 methyl mercury, 241 alkyl phenanthrene toxicity, 475 alkyl-substituted fluorenes, 430 alkyl-substituted PAH, 445 allocate food, 96 allometric analyses, 98 allometrically, 98 allometry, 169 along-shelf velocities, 36 alpha, 107 Alyeska Pipeline, 242, 422 amensalism, 188 amino transferases, 455 ammonia, 30, 53, 57–59 production, 57, 59 amphipod crustaceans, 91, 142, 466 anadromous, 64, 69, 76, 182, 215, 242 pelagic zooplanktivores, 69 Anadromous Fish Act, 214 anchorage, 16, 41, 426 animal husbandry, 214 annelid worms, 466 annual herring spawning, 81 annual primary productivity, 56, 59 Año Nuevo, 377 anomalies, 291 anthracene, 448 anthropocene era, 527 anthropogenic, 550 changes, 172, 526 contaminants, 220, 225, 240 factors, 335 forces, 2 pollutants, 225, 230 anthropogenic chemicals, 241 chlorinated hydrocarbons, 241, 242 DDT, 241, 243 hexachlorobenzene (HCB), 241, 243
polychlorinated biphenyls (PCBs), 241–243 trichloro-2, 2-bis-(4`-chlorophenyl) ethane, 241 apex consumers, 62, 63, 71, 72, 278 apex predator, 299 Apex Predator Ecosystem Experiment (APEX), 336, 337 apoptosis, 474, 475 aquaculture, 227, 240, 279 archeobacterial, 230 Beggiotoa, 230 Arctic Ocean, 18 Arctic oscillation, 178 Arctic terns (Sterna paradesaea), 324 ARGO program, 526 aromatic hydrocarbon, 245, 429–431, 460, 463, 475, 532 aromatic rings, 430, 445 arrow worms, 57 arrowtooth flounder (Atheresthes stomias), 91, 138, 300, 305–307, 309, 310, 370, 389, 544, 555 Pleuronectidae, 294 aryl hydrocarbon, 475 ascites, 456, 474, 475 Asian monsoon, 46 assessment surveys, 208 Association of Primary Production & Recruitment in Subarctic Ecosystems (APPRISE), 386–388 AT1 pod, 452–454 atka mackerel, 372 Atlanta, 426 Atlantic Canada, 233 Atlantic Ocean, 349, 390 Atlantic silversides, 214 atmospheric bridge, 176 convection, 176 atmospheric forcing, 172 inter-decadal, 172, 172 intra-decadal, 172, 173
Subject Index 563
Atmospheric Forcing Index (AFI), 293–296 Auke Bay Laboratory, 216 Australia, 176 bacterial oxidation, 53, 58 Baja California, 199, 357 bald eagles (Haliaeetus leucocephalus), 329, 423, 426, 449, 491, 531 Baltic Sea, 237 Barents Sea, 85, 390, 391 barnacle/barnacle nauplii, 57, 463, 539, 542 Balanus balanoides, 182 Balanus glandula, 182, 463 Chthalamus dalli, 464 barrier jets, 24 Barrow’s goldeneyes (Bucephala islandica), 325, 485 basin-scale, 533, 537, 538 basin-wide scale, 521 bathymetric provinces, 13 abyssal plain, 13, 14 continental rise, 13 continental shelf, 13, 16, 19, 21, 22, 24, 58, 60, 88 continental slope, 13, 14, 16, 17, 29, 31–33, 37–41, 43 fjords, 13, 15, 23, 29, 30, 39, 55, 57, 67, 81, 83 guyots, 13 seamounts, 13, 15 shelfbreak, 13, 14, 16, 32, 33, 36–38, 40, 41, 42 bathymetry, 13–16, 33, 37, 39, 41 bay Alitak, 233 Auke, 140, 386–388, 422 Bristol, 217, 319, 321, 323, 327, 329, 331, 333, 354, 382, 383, 388, 389, 391–394 Chesapeake, 232, 239, 488, 489 Chiniak, 143 Coos, 234 Delaware, 232
Eshamy, 424 Galena, 212 Glacier, 324, 328, 355, 360 Herring, 464, 479, 481 Kachemak, 69, 144, 316, 337, 338, 379–381, 383, 390, 393, 537 Kaguyak, 213, 392 Kamishak, 360, 380, 383 Kuskokwim, 352 Main, 424 McClure, 531 Nanvak, 354 Northwest, 442 Olsen, 181 Pavlof, 5, 379, 390 Puale, 316, 321, 327, 328 Resurrection, 26, 31, 301–303, 319 Rocky, 235 San Francisco, 188 Sawmill, 432 Shoup, 331 Sleepy, 425, 441 Tonsina, 424 Yakutat, 15, 42, 209, 210, 364, 432 begging for food, 107 behavioral strategies, 104, 106, 107 Belgian/German, 237 benthic, 62, 73, 86, 116, 118, 125, 126, 128, 132, 134, 136, 216 communities, 526, 531 food web, 62, 217, 375 habitats, 134, 256, 325 benzanthracene, 448 benzene, 246, 429, 430, 433, 505 benzene ring, 429 Bering Sea, 17, 18, 22, 32, 34, 46, 47, 70, 81, 86, 88, 91, 92, 102, 135, 136, 138, 187, 193, 195, 201, 207, 212, 227, 237, 276, 296, 305, 319, 323, 330, 331, 333, 337, 343, 350–352, 354–356, 361, 364, 368, 376–380, 382, 383, 388, 390, 391, 394, 522–524, 528, 540, 541, 543, 544, 546–548
564
Subject Index
bioaccumulation, 243, 477 bioavailability, 445 bioavailable, 421, 432, 447, 470, 477 biodegradation, 424 biodegraded, 531 bioenergetics, 282 biogeochemical cycling, 538 biological availability, 488 bottom-up, factors, 310 controls, 308, 309 forcings, 310 top-down, factors, 310 biomarker enzyme, 460 bioremediated, 425 bioremediation, 437, 439 bipartite, 142 biphenyl, 430, 448 birna viruses, 224 birth rate, 222 bivalve mollusks, 477 black legged kittiwakes (Rissa tridactyla), 8, 95–97, 107, 184, 331, 334, 336, 338–340, 343–345, 537 Bligh Reef, 7, 419, 423 blood haptoglobin, 455 bloom, 51, 60, 64, 69–74, 83, 84, 88, 93, 94, 143, 144 diatom bloom, 55, 64, 71, 74, 94 diatom-dominated bloom, 58 phytoplankton bloom, 57, 58, 143, 144 , 387, 388 plankton bloom, 49, 51, 60, 69, 84, 93, 385 plant bloom, 57, 70 primary, 381 secondary, 387 shelf break blooms, 60 spring bloom, 29, 30, 51–53, 57, 60, 72, 88, 144, 272, 273, 385–388, 393, 546 spring plankton bloom, 419, 530 spring phytoplankton bloom, 58, 143
springtime phytoplankton bloom, 272, 386 zooplankton, 387, 388, 393 bluff, 319 bobbin roller, 300 Boca de Quadra, 56 bottom trawl, 300 bottom-up, 543 control, 543 effects, 86 forcing, 274, 528, 529, 539, 543 hypothesis, 391 regulation, 309, 377 breeding colonies, 94, 95, 100, 527 breeding habitat, 350 breeding success, 339–344, 346–348, 350–352, 454 British Columbia (BC), 11, 14, 19, 22, 34, 41, 60, 76, 136, 144, 193, 233, 234, 265, 268, 269, 271, 276, 293, 321, 355, 357, 362, 364, 368, 376, 377 British Columbian shelf, 22, 34, 41 BTEX (benzene, toluene, ethyl benzene and toluene), 429, 430, 433, 447 butane, 429 butyl tins, 243 bycatch, 214, 218, 392, 418, 542 monitoring program, 208 C2-naphthalenes, 430 C3-benzenes, 504 C3-naphthalenes, 430 calanoid copepods, 54, 57, 58, 61–63, 65, 79, 80, 81, 94, 287, 394, 540 California (CA), 18, 46, 81, 199, 236, 303, 305, 321, 342, 355, 359, 374, 377 California current, 180, 273 Canada, 56, 176, 204, 236 Canadian GLOBEC, 276 weather ship, 274 Cancer spp., 542 cannibalistic, 76, 189
Subject Index 565
canyons, 15, 38, 88 Cape Lisburne, 320, 323, 345 Cape Peirce, 323 Cape St. Elias, 129, 357 Cape Suckling, 204 Cape Thompson, 100, 319, 320, 323, 326, 333 Cape Yakataga, 360 capelin (Mallotus villosus), 62–64, 68, 69, 91, 101–103, 128, 191, 293, 329, 332, 335, 336, 338, 339, 367, 370, 385, 389, 545 carbon tetrachloride, 243 carbonate equilibrium, 530 Caribbean Sea, 230 carnivores, 54, 189, 539 primary 190 secondary, 190 carnivorous jelly plankton, 54 carrying capacity, 275, 283 cartilaginous fishes, 227, 233 cascade effects, 487 cascade hypothesis, 389, 540, 541 cascade range, 14 catch efficiencies, 301 catch-per-unit-effort (CPUE), 379, 380 causative mechanisms, 388 centropages, 83 cephalopods, 275, 367 cetacean, 237, 454 chaetoceros, 57 chaetognaths, 546 charadriiforms, 96 chemical fingerprinting, 470 chemical spill, 242 Chenega, 432 chick strategies, 104 chicks fledge, 103 Chignik, 380, 382, 437 Chisik murres, 336–340, 344–350 chlordane, 243 chlorinated hydrocarbons, 454 chorion, 474 chrysenes, 430, 448, 505
Chugach (Alutig) natives, 278 Chukchi Sea, 18, 100, 319, 323, 326, 333, 341 ciliate protozoans, 47 clams, 374, 375, 378, 447, 463, 466, 482, 483 clay-oil flocs, 442 climate change(s), 1, 2, 307, 325, 326, 333, 334, 377, 386, 489, 527, 530, 537, 538 changes, driven, 299 control, 385 fluctuations, 2 forced, 385 forcing(s), 278, 280, 281, 301, 306, 309, 310, 383, 386, 388, 487, 547 indices, 293, 294, 306 perturbations, 32 processes, 13 regime shift(s), 178, 309, 330, 364, 383, 528, 544 shifts, 308, 309, 326 shifts, induced, 299 signals, 13, 32, 334 clines, 307 cloud cover, 24, 52 Clubbing Rock, 358 clupeid fishes, 224, 227, 234, 235 cnidaria, 54 coal deposits, 243 coarse-grained gravel/sediments, 434, 442 coastal contamination, 242 convergence, 59, 60 downwelling, 266 ecosystem, 427 habitats, 187 natives, 207 oceanic waters, 88 plankton communities, 60 runoff, 18, 32 upwelling, 288, 389, 394 zones, 278
566
Subject Index
coastal areas Southeast, 196 coastal watershed, 19 Amazon, 19, 20 Columbia, 20, 34 Mackenzie, 20 Mississippi, 15, 19, 20 Yukon, 20 cod, 198, 389–391, 393 Pacific cod (Gadus macrocephalus), 61, 76, 101, 102, 138, 139, 189, 207, 208, 234, 300, 303–307, 336, 365, 367, 370, 389–391, 544, 547, 554 Atlantic, wild (Gadus morhua), 213, 234, 390, 546 Northwest Atlantic cod, 68 coho (silver) salmon (Oncorhynchus kitsutch), 234, 287, 288 cohort, 211, 277, 533, 544 cold regime(s), 543, 544, 546 Cold Tongue Index (CTI), 174, 176 Columbia Glacier, 423 commensalism, 188 Commercial Fisheries Entry Commission, 204 common loons (Gavia immer), 449, 492 common murres (Uria aalge), 6, 95, 97, 99, 108, 449, 454, 495, 530, 537 communities intertidal, 181 competition, 188 inter-specific, 308 intra-specific, 308 compounds organic, 242 organo-metallic, 242 congeneric species, 85, 169 contaminant biomarker, 244 contaminants, 244, 548 accumulated, 245 exposure, 245 contemporary fisheries management, 203
continental margin, 40, 274 continental shelf(ves), 13, 14, 16, 19, 21, 22, 24, 58, 60, 88, 170, 301, 303, 305, 335, 422, 523, 539 slope(s), 13, 14, 16, 17, 29, 31–33, 37–41, 43, 176, 305 Cook Inlet, 14–16, 24, 26, 29, 34, 41, 42, 53, 56, 60, 68, 69, 77, 101–103, 108, 112, 137, 138, 316, 321, 325, 329, 335–332, 341, 343, 347, 348, 350, 352, 360, 379–382, 384, 424, 523, 524, 536–538 copepodite, 57, 63, 64, 277 copepod(s), 58, 191, 378, 385, 387, 394, 540, 546 Calanus marshallae, 57, 63, 65 copepod nauplii, 387 Neocalanus copepods, 385 coral, 216 Acropora palmata, 230 coral plague, 230 Auriantimonas coralicida, 230 Cordova, 279, 284, 432 cormorants (Phalacrocorax spp.), 323, 324, 328, 449, 493 common, 315 pelagic (Phalacrocorax pelagicus), 311, 324 red-faced (Phalacrocorax Urile), 311, 315, 324 double crested (Phalocrocorax auritus), 236 corticosteroids, 225 corvids, 112, 169 counternatant, 548 crab(s), 1, 210, 213, 215, 375, 378, 379, 384, 385, 387, 393, 394, 544 Alaskan crab, 140, 141, 383 blue crab (Collimates sapidus), 141 blue king crab (Paralithodes platypus), 139, 383, 391 crab pots, 216
Subject Index 567
Dungeness (Cancer magister), 209, 210, 542 king crab, 381, 385, 391 Kodiak red king crab, 383, 392 Kodiak Tanner crabs, 383 male king crabs, 392 Pacific, 234, 235 red king crab (Paralithodes camtschaticus), 135, 136–142, 144, 145, 189, 208, 210, 213, 379, 381–384, 386–389, 391–393, 394, 538 red king crab zoeae, 387 snow crab (Chionoecetes opilio), 136, 143, 233 snow king crab, 383 Tanner crab (Chionoecetes bairdi), 135–139, 141–143, 145, 189, 208, 210, 233, 379, 381–385, 387–389, 391, 394 Tanner crab zoeae, 387 creatine kinase, 455 critical forage stocks, 94 critical point, 552 critical size, 288 Cross Sound, 360 cross-shelf flux, 41 cross-shelf temperature, 35, 37 cross-shore circulation, 36 cross-shore distributions, 35 cross-shore flow fields, 32 crustacean(s), 2, 5, 7, 226, 233, 392, 393 cryosphere, 171 ctenophora, 54 ctenophores, 57 current meanders, 37 current reversals, 37 current shear, 36 current variations, 37 cutthroat trout, 493 cyanobacteria, 53 cyclogenesis, 17, 24, 47, 169 cytochrome P4501A, 459, 460, 466, 470, 474, 476, 477, 482, 532
decadal shifts, 547 decomposers, 189 deep mixed layer, 43 deep-water benthos, 531 deep water renewal, 39 demersal, 418 eggs, 83 fishes, 70 Denmark, 237 density-dependence, 282, 297 density dependent, 79, 87, 91, 92, 169, 211, 306, 308 forces, 87 mechanisms, 92 mortality, 91, 92 processes, 91 density gradient(s), 20, 32, 34, 40, 269, 270 density independent, 309 density stratification, 29 depot lipids, 418 detritus, bacterial oxidation of, 53, 58 developmental strategies, 104 diapause, 58, 63 diatoms, 387, 388, 393, 536 dibenzanthracene, 448 dibenzothiophene, 504 dieldrin, 243 digestive physiologies, 131 dimethylnaphthalene, 430 dinoflagellate, 54, 144, 233 hematodinium, 233 discards, 214, 392 diseases, 11, 165, 187, 219, 222, 225, 226, 227, 230, 232, 233, 236, 237, 238–240, 361, 363, 365, 368, 378, 392, 488, 529, 533, 548, 549 aspergillosis, 237 avian botulism, 236 avian pox, 236 bacterial origin, 225 bitter crab disease, 233, 239 black band, 230 botulism, 226
568
Subject Index
diseases (Continued) challenges, 244 cholera, 236 cyanobacteria, 230 fibro-papillomatosis (FP), 236 fungal, 225, 234 infectious, 220, 223, 224, 226, 230, 237, 238, 241 lung-eye-trachea Disease (LETD), 236 measles, 219 morbillivirus, 237–239 multinucleated spore X (MSX), 232, 239, 241 New Castle disease, 236 paramyxovirus, 237 parasitic origin, 225 phocine distemper, 237, 238, 239–241 phoma, 234 outbreaks, 211 vibrio carchariae, 230 viral hemorrhagic septicemia (VHS), 234–236, 239, 240 viral, 223 white band, 230 dispersion, 439, 445 kinetics, 442 rate, 445 dissolution, 424, 437, 445, 446 dissolved inorganic forms/nutrients, 47, 48, 59 diurnal K1 tide, 41 diurnal period shelf waves, 41 diurnal shelf waves, 41 dodecane, 433 dolphins, 361 downthrust, 182 downwelling, 21, 22, 24, 26, 29, 30, 32–34, 43, 57, 60, 170, 521, 536 favorable, 22, 24, 271 drift net fishery(ies), 283, 542 E phase, 223 eagles, 112, 130 attacks, 112
early life history strategy, 81 early life stages, 85, 88, 92, 93 East Bering Sea, 354 East Coast of North America, 41 East Pacific High, 17, 18, 45 East Pacific High pressure systems, 17 eastern oysters (Crassostrea virginica), 232, 239 eastern stock, 355–357, 360, 372 Echo-Integration Trawl survey, 300, 308 ecological change, 8 ecological interactions, 489 ecosystem change, 112, 187, 311 Gulf, 2 open ocean, 57 shelf, 37, 60 management, 191 dynamics, 487 ectoparasite, 226 eddies, 33, 37, 38, 40, 41, 71, 83, 88 edema, 456, 520 eel grass, 466 eruption Augustine volcano, 187 Chenega, 187 Cook Inlet, 187 Kenai Peninsula, 187 Mount St. Helens, 1980, 186 Port Graham, 187 EEZ (Exclusive Economic Zone), 193, 194 eggs and larvae, 41, 88, 90–92, 93 Ekman dynamics, 36 Ekman flow, 43, 59 Ekman layer, 21 Ekman transport(s), 21, 32, 33, 42, 177, 536 El Niño, 2, 75, 174, 179, 268, 269, 271, 273, 331–333, 362, 368, 521, 530 El Niño-La Niña cycle, 2, 527 El Niño scale, 553
Subject Index 569
El Niño-Southern Oscillation (ENSO), 173, 176–178, 339, 342, 530 cycle, 176, 179 thermal anomalies, 177 elephant seals, 377 elevated P4501A enzymes, 470, 485, 488 embayments, 15, 23, 83 emergence, 459, 474, 475 emergent fry, 459 emphysema, 451 endangered species, 198 threatened, 198 Endangered Species Act, 1997, 356, 361 energetic burden, 485 energy allocated, 108 energy allocation, 108 England, 237 entrainment, 30, 39 environmental change, 95, 96, 103, 111, 133, 347, 377 environmental regimes, 91, 92 enzootic, 221 epipelagic zone, 58 epizootic, 220, 222, 223, 226, 227, 232, 235, 236–240 epizootiologic, 232 Equatorial Kelvin wave, 176 equinoxes, 25 Erlington, 424 escapement(s), 203, 204, 281, 284–287, 289 estuarine regions, 71 ethane, 429 ethylbenzene, 504 etiologic, 233 eulachon, 62, 64, 69, 91, 128, 367 Euphausia pacifica, 64, 66 euphausiids (Thyanoessa spp.), 287, 540 euphotic zone, 29, 30, 37, 41–44, 273 Europe, 201, 234, 240 Evans and Bainbridge, 424, 433
evaporation, 424, 433, 437, 445 evaporative weathering, 434, 435 EVOS GEM program, 526 Exxon Corporation, 426, 427, 437, 534 Exxon Valdez Oil Spill Trustee Council, 6, 336, 347, 426, 489 False Pass, 352 fatty acid signatures, 73 fecund, 239 fecundity, 75, 142, 206, 220, 226, 238, 366, 385 Federal Department, Interior, 291 feedback negative, 172 positive, 172 feeding strategies, 125, 137 feeding success, 388, 483 filter feeders, 54 fine-grained oiled sediments, 440 fine-grained sediments, 434, 435, 439 fine-sediment habitat, 390 fine sediments, 466 finer-grained beaches, 466 Fish and Wildlife service, 422 fish life histories, 93 fish stocks, 1 fisheries cod, 215 dive, 211 dredge, 210 hook-and-line, 207 longline, 207 pot, 199, 210 roe, 206 trawl, 199, 207, 210 troll, 204, 205 fisheries managers, 217 fishery(ies) management, 194, 203, 204, 213, 218, 291, 299, 379 fjord’s sill, 39 fjords, 13, 15, 23, 29, 30, 39, 55, 57, 67, 81, 83, 355 flagellates, 47, 58
570
Subject Index
flatfish(es), 5, 207, 208, 224, 306, 365, 367, 370, 372 flathead sole (Hippoglossoides elassodon), 208 fledging/fledge(d) , 95, 103, 106, 169, 330, 340, 341 success, 341, 348 age, 106 Florida, 230, 236 flounder(s), 244, 371 flouranthene, 448 flourene, 430, 448, 501 fluorescence, 55 flow features, 37 flow meanders, 38 focus, 553 food chain(s), 189, 191, 237, 239, 241, 377 food chains and webs, 62, 73 food limitation, 325, 341, 367, 370, 482 hypothesis, 367 food web(s), 7, 60, 62, 63, 71, 72, 74, 165, 189, 191, 217, 242, 309, 333, 334, 529, 533, 536, 538, 540, 549 structure, 62 transfers, 71 foraging areas, 95, 99, 101, 107, 127 behavior, 337, 483, 486 depth, 98 ecology, 95–98 efficiency, 103, 132 effort, 339 habitat, 329 range, 70, 100, 103, 109, 131, 132 strategy(ies), 74, 98, 99, 125, 134, 135 success, 368 factors, 310 forelands, 16 Forest Practices Act, 215 fork-tailed storm petrel (Oceanodroma leucorhoa), 311, 315, 323 fracture zones, 13, 15 Fraterculini tribe, 96 frazer lakes, 281
freshwater coastal discharge, 19, 20, 21, 32, 269 dispersal, 272 flow, 470 influx, 19 inputs, 26, 546 outflow, 537 runoff, 11, 20, 30, 36, 39 zone, 472 freshwater-driven currents, 83 fry, 280, 284–287, 419, 427, 457, 459, 460, 461, 468, 476, 477 Fucus gardneri, 463, 464, 465, 486 fulmars, 95 fungal pathogen Icthyophonus hoferi, 235 fur seals, 36, 376 furunculosis, 218 Aeromonas salmonicida, 224 gadid(s), 5, 69, 80, 234, 235, 285, 301, 303, 306, 309, 310, 332, 371 Gadidae, 300 gadoid fishes, 530 GAK 1, 26–28, 30, 31, 268, 271, 301, 389, 526 Galeta, 533 gametophytes, 464 gear longline, 216 bottom-contact, 216 generalist feeders, 367 genetic damage, 456 geoduck clamps (Panopea abrupta), 211 geophysical events, 180 geophysical forces, 180, 527 gestate, 366 gestation, 116, 118, 132, 169 ghost fishing, 214 gillnet, 198, 204, 205, 335 gillnet fishery, 205 drift gillnet, 198 salmon gillnet, 199 glacial haul-out sites, 355
Subject Index 571
glacially influenced fjords, 333 glaciers, 11, 15, 18, 180, 183, 187, 329, 432 Bering, 15, 183 Columbia, 183 fine-grained sediments, 15 glacial advance, 15 glacial field, 15 glacial fjords, 29 glacial ice, 15 glacial meltwater, 19 glacial moraines, 15 glacial process, 15, 19, 29 glacial runoff, 19 glacial scouring, 15 glacial silt, 15 Malaspina, 15, 183 glaucothoe, 137, 144 global climate change, 1 global thermohaline circulation, 44 global warming, 329, 426, 530, 537, 548 global-scale thermohaline circulation, 44 GLOBEC program, 537 GLOBEC supported research, 526 gradients cross-shelf, 34 cross-shore density, 35 cross-shelf salinity, 34 gray whales, 452, 454 grayling (Thymallus arcticus), 242 grazers, 486 grazing ciliates, 58 Great Alaska earthquake, Good Friday, 1964, 14, 180–182, 184, 186, 242, 279, 284, 316, 334, 359, 381, 424, 432, 434, 534, 535 grebes, 324 greenhouse gas(es), 25, 172, 272, 526, 530 Greenland, 56 Greenland sharks, 130 grey seals (Halichoerus grypus), 365, 366 groundfish(es), 192, 196, 207, 208, 214, 300, 302–308, 310, 378, 379, 385, 389, 391, 394 groundwater flow, 470
growth inhibition, 476 guillemots, PWS, 333 Gulf of Alaska Basin, 30, 42 Gulf of Maine, 385, 390 Gulf of St. Lawrence, 235 gulf pattern, 383, 385 Gulf Stream, 41 gulls Bonaparte’s, (Larus philadelphia), 324 glaucous-winged gulls (Larus glaucescens) , 184, 324 Gull Island colony, 537 gyre, 179, 273, 274, 277 Alaska, 42, 43, 46, 179, 276, 530 Gulf of Alaska, 43 subarctic, 40, 42, 46, 47, 179 subtropical, 45, 46, 179 upwelling, 59, 282, 288 western subarctic, 16, 46, 47 habitat dependency(ies), 61, 75, 94 habitat disturbance, 393 halibut, 196, 208, 389 Greenland (Reinhardtius hippoglossoides), 203 Pacific (Hippoglossus stenolepis) , 138, 189, 196, 207, 214, 548 Halibut Convention, 1923, 207 haline stratification, 546 halocline, 30, 38, 40, 43, 44, 58, 169, 544, 546 halogenated hydrocarbons, 243 dioxins, 243 furans, 243 harbor porpoises, 454 harbor seals (Phoca vitulina), 8, 116–131, 187, 191, 192, 196, 197, 199, 218, 238, 352–355, 361–370, 373–378, 419, 440, 449, 451–453, 494, 535 pinniped(s), 352, 361–364, 366, 368, 377, 378 harlequin ducks (Histrionicus histrionicus), 325, 449, 450, 479, 480, 483–485, 488, 494, 532, 534
572
Subject Index
harpacticoid copepod, 466 harvest strategies, 393 hatchery(ies), 279, 280, 282–289, 432 hatchery-raised fry, 458 hatchery-reared pink salmon, 458, 460 hatchery-reared salmon, 283, 285 hatchery stocks, 284 hatching success, 340 haul out(s), 118, 127, 129, 130, 134, 237, 355, 357, 366, 372, 373, 451 heart defects, 456 heat exchange, 18, 25, 32, 43, 45, 46 heat flux(es), 24, 25, 47, 172, 176, 177, 267, 272 heat transfer rate, 25 heavy metal contaminants, 243 arsenic, 243 cadmium, 243 chromium, 243 lead, 243 mercury, 243 silver, 243 heavy metals, 225 hemorrhagic septicemia, 224 hepatic necrosis, 457 hepatocellular glycogen, 474 herbivore(s), 54, 58, 62, 189, 533, 539 herpes virus, 236 herring (Clupea pallasi), 61–65, 67, 69, 71, 74, 76, 77, 79–83, 101, 191, 192, 205, 206, 234, 235, 240, 338, 339, 365, 367, 370, 419, 422, 455–458, 487, 488, 527, 528, 531, 533, 540, 541, 550, 554 herring, 285, 286, 290–299 Atlantic herring (Clupea harengus), 233, 239 eggs, 290 embryos, 81 food web, 533 life history, 297 PWS, 296, 299
roe, 292, 293 spawn, 290 stocks, 540 heterotrophs, 189 hexachlorocyclohexanes (HCHs), 243 high energy prey, 101 high level predators, 75 high nest-site fidelity, 96 high nutrient low chlorophyll (HNLC), 58 high-pressure washing, 437 higher-energy capelin, 101 higher trophic levels, 536, 537 Hinchinbrook Canyon, 16, 39 Hinchinbrook Entrance, 15, 16, 23, 34, 38, 39 histology, 477 histopathological abnormalities, 474 histopathological lesions, 475 histopathology, 520 Hokkaido, 352 home ranges, 118, 127–129, 131, 483 homeotherms, 226, 246, 532 Homer, 424 human harvesting, 87 humoral, 224 humpback(s), 217, 452, 454 hydraulic methods, 437 hydraulic spot-washing, 439 hydroacoustics, 292 hydrocarbons, 429, 432, 433, 442, 451, 453, 457, 458, 460, 477, 480, 489, 531, 532 hydrologic cycle, 175 hypothermia, 246, 447, 451 hypothermic, 450 ichthyoplankton, 418 Icy Strait, 360 immune responses, 548 immune suppression, 244 immuno-compromised, 227 immunosuppression, 225 immunosuppressors, 225 implantation, 118, 169
Subject Index 573
incubate, 83, 95 incursion, 37 index return/fry, 286 return-per-alevin, 285 indicators, 75 indigenous people, 196 indirect effects, 486, 487, 533, 550, 555 indirect food web interactions, 540 industrial discharges, 242 Industrial Revolution, 530 infected, 221 infected, 222 infectious agents, 226, 230, 241 infrared, 25 inner and outer shelf plankton, 55 inorganic nitrate, 57 inorganic nutrients, 61 insemination, 141 inshore migration, 39 inter decadal, 270, 310 interleukin-6, 455 intermediary host, 232 internal tide, 42 internal waves, 169 International Fur Treaty, 1911, 201, 359 International Pacific Halibut Commission, 208, 300 gillnets, 190 International Whaling Commission (IWC), 193 interspecies competition, 549 interspecific competition, 488 interspecific interactions, 486 interstitial water, 470, 477 intertidal community(ies), 2, 461, 473, 527, 531, 533, 534, 539 intertidal zone(s), 531, 533, 534, 539 intra-decadal, 266 intrinsic growth potential, 104, 113, 114 inverse production regime, 280, 282, 376 response, 281
invertebrates, 25, 62, 65, 70, 76, 81, 101, 116, 118, 125, 126, 128, 132, 133, 138, 139 Ireland, 237 iron deficiency, 59 iron limitation, 58, 59 Isaeidae, 466 islands Afognak, 201 Aiktak, 319, 320, 323 Alaska, 198, 204 Aleutian archipelago, 201 Aleutian, 196, 199, 201, 218, 240, 311, 319–321, 323, 328, 354–357, 359–364, 368, 372, 374–376, 378, 380, 382, 535, 541, 547 Amchitka, 359, 360 Atkins, 358 Barren, 24, 68, 201, 319–321, 323, 327, 336–340, 344–349, 370, 424 Bering, 375 Bligh, 433 Block, 424 Bogoslof, 321 Buldir, 319, 321, 323 Chernabura, 358 Chirikof, 17, 358 Chisik, 316, 321, 325, 326, 329, 333–335, 337, 339, 347, 537 Chiswell, 319, 327, 328, 358, 370, 373 Chowiet, 319, 321, 324, 327, 328, 358 Commander, 356, 375 Cook inlet, 198, 204, 206, 208, 210 E. Amatuli, 319, 320, 323 Eleanor, 424, 433 Farallon, 342 Green, 424 Gull, 316, 321, 327, 329, 335, 336, 337, 339, 340, 344–350 Hawaiian, 552, 553 Ingot, 424 Isle of May, 342, 343 Kasatochi, 319 Kayak, 129, 201, 352, 355–357
574
Subject Index
islands (Continued) Kenai Peninsula, 196 Knight, 424, 427, 433, 454, 464, 466, 480–483, 485, 486 Kodiak Archipelago, 196 Kodiak, 2, 3, 13–17, 22, 29, 31–34, 37–43, 77, 129, 136–138, 140, 142, 184, 193, 196, 201, 205, 206, 208, 210, 281–283, 303, 305, 321, 329, 338, 353, 354, 364, 367, 370, 373–375, 379–382, 384, 390, 419, 432 Koniuji, 319, 323 Kuril, 201, 359 LaTouche, 424, 425, 433, 442 Lone, 424 Marmot, 198, 358, 373 Middleton, 18, 21, 22, 25, 26, 111, 184, 316, 317, 321, 324–329, 331, 333, 334, 534 Montague, 34, 424, 454, 481, 482, 534 Naked, 424, 433 Near, 356 Nord, 319 Otter, 354, 364 Outer Pye, 358 Peninsula, 198 Perry, 424 Pribilof, 196, 243, 319, 323, 331, 354, 361, 362, 364, 541 Prince William Sound, 196, 201, 203–206, 208, 210, 374 Queen Charlotte, 14 Sable, 121, 130, 365, 366, 368 Sanak, 201 Semidi, 15, 17, 34, 196, 319 Shumagin, 14, 15, 17, 34, 88, 89, 136, 201, 303 Shuyak, 201 Smith, 424, 433 Southwest Prince William Sound, 201 St. George, 319, 323, 331 St. Lazaria, 319, 320, 323
St. Matthew, 391 St. Paul, 319, 323 Sugarloaf, 197, 358 Sutwick, 201 Tatoosh, 334 Triplet,198 Tugidak, 199, 353, 354, 364–366, 376 Ugamak, 6 Ulak, 319 US Virgin Islands, 230 Vancouver, 293 Whittier, 184, 424, 432 Wooded, 358 isobaths, 35, 42 isomers, 430 isopleth, 169 iteroparity, 141, 169 iteroparous, 87, 169 Japan, 199, 232, 355–357 jellyfish, 546 Joint U.S.–Canadian High Seas, 274 junk food, 372 junk food hypothesis, 369 juvenile herring, 329 juvenile pollock, 76, 189 Kamchatka Peninsula, 47, 137 Kamishak District, 384 karluk, 280 Katalla, 432 Katmai, 186 Novarupta, 1912, 186 kelp forest canopies, 134 kelp (Macrocystis pyrifera), 466, 539 Kelvin wave, 177, 178 Kenai Coast, 419 Kenai Fjords, 324, 328 Kenai Peninsula, 16, 34, 338, 357, 360, 421, 424, 432, 433, 435 Kennedy Entrance, 16, 41 Ketchikan, 18, 20, 269, 270, 355 key species, 550 keystone species, 455, 486, 539 killer whales, 452, 453, 487, 488, 496
Subject Index 575
kinetic processes, 475 king crab fisheries, 380 kittiwakes, 95–104, 106–114, 337–350, 530, 537 black-legged (Rissa tridactyla), 311, 313, 320, 321, 323, 325–337 red-legged kittiwakes (Rissa brevirostris), 311, 323 Knight Island Passage, 16, 38 Knik Arm, 16 Koch’s postulates, 230 Kodiak Archipelago, 353, 354, 362 Korean Peninsula, 86 krill, 47, 61, 62, 71, 79, 224 k-selected species, 377 Kulthieth, 432 Kuroshio, 45–47 Kuroshio-Oyashio Extension, 45–47, 179 La Niña(s), 174, 179, 271, 273, 331–333 Labrador, 214 lactating harbor seals, 367 lactation, 116, 117, 119–124, 131–133, 365, 366, 368 landslides, 14 Lanugo, 121 large copepods, 58, 59, 64, 72, 79 large calanoids, 94 larval pollock, 76 large-scale wind stress pattern, 45 Laridae, 96 Larvacea, 54 larvacean, 57, 72 latent, 177 latent heat exchange, 25, 46, 169 latent heat fluxes, 25, 47 lesions, 451, 456, 457 less-weathered, 475 life history(ies), 62, 268, 287, 306, 310, 368, 369, 373, 374, 379, 532 adaptations, 85, 135 characteristics, 74, 76, 95, 113, 114, 125, 132, 531, 535
patterns, 75 stages, 528, 548 strategies, 75, 104, 124, 145, 379 traits, 214 genotypes, 214 heritable variation, 214 selective breeding, 214 studies, 110 tradeoffs, 108 life stage, 246 life stage-based models, 394 light levels, 53, 59, 90 light limitation, 53, 59 light-limited period, 56 Limacina helicina, 57 Limacina pacifica, 64 Limanda aspera, 203 limited resources, 486 limiting factor, 111 limiting nutrient, 57 limiting resource, 374 limpets (Tectura persona), 463, 464, 486, 533 line P, 265, 273 lingcod (Ophiodon elongatus), 208 live-bearers (ovoviviparous), 75 loafing time, 103, 108 lobsters, 233 local atmospheric forcing, 17 long-lines, 542 long-term ecological change(s), 277, 535, 539 patterns, 274 trend, 276 zooplankton, 276 Long-Term Environmental Program (LTEMP), 243 long-wave radiation, 25 loons red-throated (Gavia stellata), 324 Pacific (Gavia pacifica), 324 Los Niños, 174 loss of immunity, 222
576
Subject Index
Lower Cook Inlet, 16, 23, 56, 68, 69, 77, 137, 138, 334, 336–338, 419, 521, 537 lower trophic levels, 528, 536, 542 Lower Valdez Arm, 423 lows, Graveyard of, 18 macroparasites, 220 macrozooplankters, 61, 62, 72, 79, 81 Magnuson Fishery Conservation/Management Act, 194, 202, 207 Magnuson-Stevens Fishery Conservation/ Management Act, 202 Maine, 232 man (Homo sapiens), 138 management strategy(ies), 202, 289 mangroves, 533 marine ecology, 8 ecologists, 8 Marine Mammal Protection Act of 1972, 202 marine stewardship council, 201 match–mismatch hypothesis, 385–389, 393, 546 mechanism, 387 matriarchially-organized, 454 Mediterranean Sea, 188 Megalopae, 136, 137, 144 meiofauna, 466 mergansers, 324 mesopelagic, 58, 62, 64 mesopelagic fishes, 62, 64 mesopelagic zone, 58 mesoscale, 23, 24, 37, 521, 537 metazoan parasites, 226 nematode, 227 trematode, 227 meteorological forcing, 282, 377, 537 methane, 429 methyl group, 429
methylethylnaphthalenes, 430 methylnaphtalene, 430 Metridia okhotensis, 57 Mexico, 199, 352 Mexico, Gulf of, 217 microbial decomposition, 439 microbial degradation, 437 microflagellates, 54, 58, 72 micrograzers, 58, 59 microparasites, 220 bacteria, 220 protoctistan parasites, 220 protozoans, 227 viruses, 220 microphthalmia, 455 microplankton, 60 mid-shelf currents, 37 mid-shelf domain/region, 32, 36, 37 minimum prey density, 98 mitogens, 225 T-lymphocyte, 220 mixed layer, 30, 37, 41, 43, 48, 51, 52, 177, 269, 273, 288, 530, 550 depth, 177, 179, 265, 271, 273 mixing depth, 528 mixing events, 387 monoaromatic hydrocarbon(s), 246, 451 Montague Strait, 15, 16, 34, 38, 357, 550 morbillivirus, 362 more-weathered, 475 morphological abnormalities, 531 morphology, 98 mortality, 222 multidecadal, 306 multi-disciplinary studies, 489 multiparous, 143 multiple age classes, 93 murre(s), 95–97, 103, 112, 184, 311, 313, 316, 319–321, 323, 325–329, 332–335, 337–349, 422, 450, 530, 537
Subject Index 577
Atlantic murres, 343 common murres (Uria aalge), 311, 313, 314, 319, 337, 340–345, 349 fledging, 341 thick-billed (Uria lomvia), 311, 313, 319 murrelets kittlitz’s (Brachyramphus brevirostris), 187, 311, 315, 324, 328, 329, 333, 335, 450, 496 marbled (Brachyramphus marmoratus), 311, 315, 324, 328 mussels Mytilus trossulus, 375, 422, 426, 442, 446, 449, 455, 458, 463, 477, 478, 479, 480, 482, 483, 497, 539 Mya arenaria, 477 Protothaca staminea, 466, 477 mussel beds, 472, 479 National Mussel Watch Program, 243, 422, 479 mysids, 91, 126 mysticete, 452 Neocalanus flemingeri, 274 n-alkanes, 445 naphthalene, 430, 447, 448, 457, 475, 504 narcosis, 475 natal, 284, 476 natal philopatry, 88, 169 natality, 96, 169 National Marine Fisheries Service (NMFS), 207, 216, 300, 301 groundfish trawl survey data, 300 survey(s), 300, 301 National Oceanic and Atmospheric Administration (NOAA), 215 National Oceanic and Atmospheric Administration-National Center for Environmental Prediction (NOAA-NCEP), 266, 267 National Research Council, 218, 363, 373 Native Groups, 425 natural hydrocarbons, 432
naupliar, 277 Nearshore Vertebrate Predator, 482, 483, 484 nekton/nektonic, 135, 275 nematode parasites, 457 Neocalanus cristatus, 50, 65 Neocalanus plumchrus, 274, 276, 277 Neocalanus spp., 50, 57, 59, 63, 65, 79, 81 neonatal, 116, 119, 120, 169 neritic forms, 57 nesting habitat, 111, 326, 328, 344 nestling period, 106, 113 net heat flux, 25 neutralism, 188 New York Harbor, 419 Newfoundland, 100, 213 Newfoundland shelf, 391 Niche-specialization, 86 nitrate, 30, 31, 56–59, 272, 439 nitrate levels, 59 nitrogen cycling, 58 nitrogen fixation, 53, 57 NMFS, 379 NOAA On-Scene Spill Model, 445 non-point chronic inputs, 242 North America, 180, 201, 355, 452 North American plate, 180, 183 North Atlantic Ocean, 44, 125, 135, 136, 342, 352, 353, 355, 356, 359, 361, 386, 390, 393 North Atlantic Oscillation (NAO), 546, 551 North Equatorial Countercurrent, 46 North Pacific Current, 19, 42, 45, 47, 179, 265, 273 North Pacific Fishery Management Council (NPFMC), 201, 204, 217 North Pacific GLOBEC, 60 North Pacific High (NPH), 521 North Pacific Index (NPI), 174, 282 North Pacific Intermediate Water, 44 North Pacific Ocean, 2, 3, 269, 272, 280, 352, 353, 355–357, 359, 361–363, 367, 368
578
Subject Index
North Pacific Rim, 199, 240, 357 North Sea, 235, 239, 362, 546 North Slope Crude oil, 423 North Western Canada, 483 Northeast Atlantic, 390 Northeast Pacific, 176 Northern Atlantic, 240 Northern fur seal (Callorhinus ursinus), 541 northern warm regime, 288 Northwest Atlantic, 217, 390 Norton Sound, 81, 323, 383 Norway, 56, 237 Nova Scotia, 121, 130, 365 NP indices, 178 Nuclear Regulatory Commission (NRC), 216 nutrient(s), 11, 12, 29, 30, 34, 37, 38, 41, 42, 47, 48, 51–54, 57, 59, 61, 72, 123, 127 availability, 56 depletion, 123 limited portion, 53 sources and sinks, 60 nutrient-laden oceanic waters, 88 nutrient-rich deep waters, 43 nutrient-rich waters, 39, 44 nutritional limitation, 71, 362, 368, 373, 376–378 nutritional stress, 110, 363, 365, 366, 369 hypothesis, 363, 371 ocean climate, 281, 282, 285, 287, 288 regime, 169 stratification, 20 oceanic copepods, 63 forms, 57 plankton communities, 60 oceanic fronts, 99, 169 cross-frontal exchange, 36 cross-frontal flow convergence, 36 frontal boundary, 36
shelfbreak front, 32, 33, 36, 37, 40 small-scale fronts, 42 sub-arctic front, 12, 179 Ocean station P (1956), 526 Ocean Station P (OSP), 274, 277 oceanographic regimes, 339 octopus, 191, 367, 370 odontocete, 452 offshore pelagic community, 58 Oikopleura sp., 57 oil spills England (Torrey Canyon), 422, 486 United States (Santa Barbara Channel), 422 Arab Oil embargo, 422 Okhotsk, Sea of, 47, 86 oligochaetes, 466 Olympic peninsula, 234 omnivore(s), 54, 74, 76, 170 open coasta/shelf waters, 59, 60 open ocean, 53, 55, 57–59, 72, 74, 78, 79, 86, 144 plankton community, 57 primary production, 72 opercular hypoplasia, 475 opportunistic feeders, 101 opportunistic species, 92 Oregon, 11, 271, 276, 281, 288 organic contaminants DDE, 243 Hg, 243 organic matter, 47, 48, 58, 60, 61, 73 orographic effects, 18, 22, 23, 170 Oscillating Control Hypothesis (OCH) , 309, 543, 545 otariids, 119, 120 otter, 357–364, 374–376, 378 out-competing, 107 Outer Continental Shelf Environmental Assessment Program (OCSEAP), 60, 326, 422 outer shelf, 26, 30, 32, 37, 39, 55, 69 shelf/slope domain, 39 ovarian atresia, 213
Subject Index 579
overfished, 203, 208 overfishing, 68, 202, 207, 210, 212–214, 217, 218, 232, 283, 309, 392, 394 Oyashio current, 45, 47 oyster(s), 224, 232 oystercatcher(s), 450, 464, 532 black (Hemitopus bachmani), 449, 454, 492 P4501A enzyme, 244, 456, 460, 471, 480, 485, 532 induction, 471, 488 levels, 485 Pacific Decadal Oscillation (PDO), 173, 178–180, 266, 268, 269, 271, 273, 281, 288, 293–297, 309, 330, 331, 334, 393, 528, 530, 537, 546, 547, 551 index, 266 PDO-type oscillation, 333 Pacific hake, 288 Pacific herring (Clupea pallasi), 8, 61, 64, 65, 71, 74, 76, 79, 81, 82, 85, 210, 224, 234, 239, 244, 245, 285, 290, 293, 296, 299, 367, 449, 455–457, 475, 478, 497, 550 Pacific Inter-Decadal Oscillation (PIDO), 293, 295 Pacific/North American Lithospheric Plates, 14, 15, 40, 42, 45, 46, 58–61, 64, 65, 68, 86 seismic, 14 tectonic, 14, 15 volcanic, 14 Pacific northeast, 275 Pacific northwest, 201, 215, 234, 235, 271, 276, 377 Pacific ocean perch (Sebastes alutus), 4, 207, 540–542 Scorpaenidae, 300 Sebastes alutus, 300, 305, 306, 309, 310 Pacific Oceans, 240 Pacific plate, 15, 180, 181
Pacific Salmon Commission, 204 Pacific Salmon Treaty, 204 Pacific sand lance, 64 Pacific subarctic, 277 PAH isomers, 430 Panama, 533 parakeet auklets (Cyclorhynchus psittacula), 325 parasite infection rates, 308 parasites/disease, 308, 310 parasitic disease 232 Dermocystidium, 232 parasitism, 188 parental provisioning, 106 partially synchronous, 87 parturition, 118, 119, 368 patch-dependent feeding, 71 pelagic, 47, 48, 50, 54, 58, 59, 62, 68–71, 73, 76, 83, 91, 93, 96, 100, 101, 125, 136, 137, 193, 320, 324–326, 328, 487 alcid, 100 ecosystem, 533, 534 fishes, 217 food web(s), 73, 375 forage fishes, 91, 101 larvae, 83 lifestyle, 96 mollusk, 50 nekton, 277 rock fishes, 286 sharks, 217 pericardial edema, 456 periwinkle (Littorina sitkana) , 464, 486 persistent organic pollutants (POPs), 242, 526 pesticides, 242 flame retardants, 242 Peru, 205 pesticides, 242 petroleum hydrocarbon(s), 428, 460 contamination, 477 petroleum-based compounds, 225
580
Subject Index
Pew Oceans Commission, 2003, 201 phagocytsis, 225 T-lymphocytes, 225 B-lymphocytes, 225 phenanthrenes, 430, 447, 504 phenols, 243 phenotype(s), 110 phocid seal, 119, 122, 130, 116, 119, 121, 122, 124, 125, 170, 377 phosphate, 30, 31, 439, 536 photic zone, 51, 55, 57–59, 288, 528 photoinduced toxicity, 430 photolysis, 424 photo-oxidation, 437 photo-oxidized, 531 photosynthesis, 11, 12, 24, 29, 47, 51–53, 59, 537 photosynthetic forms, 55 photosynthetic producers, 47, 71 phoxocephalidae, 466 phylogeny, 96, 120 phylogenetic groups, 75 physical forcing, 334, 377, 385 physical weathering, 446 physiological stress, 477 physiologies, 75, 131 phytoplankton, 189, 384, 386–388, 393, 526 bloom, 544 consumption, 30 production, 30, 53 pigeon guillemot (Cepphus columba), 325, 329, 450, 454, 497 Pinnacle Rock, 358 pinnipeds, 116, 118–123, 127, 128, 131–135, 196, 237, 240, 536, 545, 546 life history strategies, 377 pinto abalone (Haliotis kamtschatkana), 211 piscivores, 324, 325 piscivorous fishes/flatfishes, 91, 189, 257, 309, 328, 543, 544, 546
piscivory, 74, 288 pismo clams (Tivela stultorum), 542 planktivores/ planktivory, 285, 288 planktivorous fishes, 64 plankton, 274, 281, 288, 336, 461, 523, 526, 528, 534, 550 blooms, 544 production, 488 planktonic phase, 83, 306 plant succession, 183 Pleistocene, 187 plunge dive, 99 pneumonia, 453 poikilotherms, 224, 226, 548 point-source oil spills, 242 polar bears (Ursus maritimus), 242 Polar lows, 18 pollock, 198, 285, 286, 527, 538, 541, 544, 547, 555 adult, 61, 79, 80, 88, 91, 94 Alaska, 201 alevin, 77, 78, 79 93 eggs, 90, 91 fry, 77 juvenile fish, 77 juvenile pinks, 94 juvenile, 36, 60, 63, 64, 81, 86 larvae, 37, 88, 90, 93 larval salmon, 77 larval survival, 92, 143, 144 life stage, 77 redds, 77, 79 spawn, 92, 93 spawning, 18, 61, 64, 65, 68, 77, 81–83, 87–89, 92–94 Theragra finmarchica, 85 walleye pollock (Theragra chalcogramma), 8, 61, 74, 76, 79, 85–87, 89, 90, 138, 188, 207, 208, 300, 303, 305–310, 332, 335, 336, 338, 339, 361, 367, 370, 372, 540, 542, 544
Subject Index 581
pollutants, 224 Pollution Act of 1990, 534 polycyclic aromatic hydrocarbons (PAH), 224, 429, 430, 432– 434, 442, 445–447, 454, 455, 457, 458, 460, 466, 467, 470– 472, 474–481, 483, 488, 489 polyethylene membrane devices (PEMD), 471 polynuclear aromatic hydrocarbons (PAH), 243, 244, 429, 448 population dynamics, 554 Porphyra, 182 porpoises, 196, 361, 536 Dall’s, 198, 452, 454, 542 Harbor, 198 Port Valdez, 423 post-partum, 118, 132, 170, 365 pot surveys, 383, 391 Poul Creek Formation, 432 precipitation, 11, 15, 18–20, 44 rates, 18, 19 precocial chicks, 95, 96, 121, 123 pre-emergent, 284 preening, 246 pre-partum, 118, 170 pressure gradients coastal freshwater flux, 19 cross-shore, 21 horizontal, 20 prey density threshold, 103 primary consumers, 486 producers, 47, 54, 58, 62, 71, 72, 133, 189, 274, 536, 539 production, 53, 59, 60, 71–74, 130, 266, 274, 294, 337, 374, 386–388, 528, 536, 545, 546, 548, 552 productivity, 52, 56, 57, 59, 72, 274, 282, 550 primiparous, 142, 143
Prince William Sound (PWS), 14–16, 18, 26, 30, 34, 38, 55–57, 63, 65, 67, 69, 71, 77–79, 81–85, 111, 122, 126–128, 130, 133, 181, 182, 184, 187, 234, 235, 242, 272, 276, 278–280, 283–287, 290–294, 296–299, 320, 321, 324, 327–329, 331–333, 336, 338, 339, 353–355, 357, 359–362, 364, 367, 370, 373, 374, 379–381, 419, 421–423, 427, 428, 430, 432, 434, 435, 437, 440–443, 445, 447, 451, 452, 453, 451, 452, 453, 460, 461, 466, 468, 472, 475–477, 480, 481, 483–486, 488, 489, 521, 532, 534, 537, 542, 548, 549 Prince William Sound Aquaculture Corporation (PWSAC), 279 private Salmon Hatchery Act, 279 production regimes, 281 propane, 429 propylnaphthalenes, 430 protandric hermaphrodite, 141, 170 protected bays, 67 protected inner/marine waters, 55–59 protozoans, 47, 53, 58, 60, 71 proventriculus, 102, 107 Pseudocalanus, 57, 83, 143, 387, 388 pteropod(s), 50, 54, 57, 61–64, 72, 81, 170, 287 puffin(s), 95, 96, 319, 323, 332, 334 chicks, 106, 108, 112, 113 horned (Fratercula corniculata), 325 tufted (Fratercula cirrhata), 8, 95, 97, 99, 105, 109, 311, 314, 320, 321, 324, 328, 335 Puget Sound, 86, 87, 234, 488, 489 purse seines, 199, 204 pursuit-diving, 96, 98 seabirds, 96 pycnocline, 30, 40–42, 44, 170, 549 pyrenes, 430
582
Subject Index
quake rupture zone, 180 quasi-organized, 44 quasi-periodic states, 172, 178, 530 radiative heat loss, 25 radiative heating, 24, 44, 267 rainfall rates, 18 rays, 234 receptor agonism, 475 reciprocal transplant experiment, 466 recruit per spawner, 298, 299 recruiting cohort, 299 Recruitment Processes Program, 301 recruitment, 84, 85, 91, 92, 110, 135, 139, 142, 211, 273, 293, 303, 307, 326, 329, 347, 376, 383–385, 388–391, 393, 394, 454, 455, 464, 486, 544, 547 dynamics, 386 success, 378 recruits, 383, 384 recruits, 299 rectified tidal currents, 42 redds, 211, 470 superimposition, 211 regime change, 309 regime shift(s), 92, 103, 145, 178, 204, 266, 271, 273, 281, 284, 285, 307, 331, 334, 326, 327, 329, 332–335, 350, 352, 364, 367, 369–372, 376, 377, 383–385, 388, 389, 393, 533, 546, 553 hypothesis, 376 regional hydrologic cycle, 13 remineralization, 47, 58 removed class, 221–223, 225 reproductive aggregations, 71 competence, 245 fitness, 282 impairment, 472, 475, 531 stock, 279 strategies, 49, 76, 87, 93, 95, 96, 109, 116, 123, 133, 140, 142, 538
success, 69, 75, 85, 104, 108, 110, 118, 123, 124, 141, 211, 335, 337, 472, 554 resident, 452 Resource Assessment and Conservation Engineering Division (RACE), 300 rivers Columbia, 271 Fraser, 271 Matanuska, 15, 183 Yakutat, 183 Alsek, 15, 19, 42 Columbia, 34 Copper, 15, 19, 23, 183 Knik, 15, 16, 183 Matanuska, 537 Mississippi, 15, 19 Stikine, 15 Susitna, 15, 19, 183, 537 river banks Georges, 42 Portlock, 15, 42 Alsek, 15, 42 river delta Copper, 23, 198, 199, 204, 218, 352, 481 river discharge, 271 river otters, 454, 455, 499 river runoff, 166 rock fish, 75, 300, 305, 370, 501, 542 black (Sebastes melanops), 208 blue (Sebastes mystinus), 208 Lepidopsetta billineata, 203 Sebastes spp., 207, 208 rock sole Lepidopsetta bilineata, 308 Lepidopsetta polyxystra, 308 roe-stripping, 198, 199 rookeries, 127, 129, 130, 134, 197, 353, 354, 356–358, 362, 364, 369, 545 natal, 129 rotifers, 226 runoff, 18–20, 29, 30, 32, 35, 36, 38, 39, 44 Russia, 201, 215, 359
Subject Index 583
sable fish, 196, 207, 208, 452 Sagitta elegans, 57 salinity gradients, 32, 34, 35, 37 salinity isopleth, 38 salmon, 2–4, 7, 8, 192, 196, 198, 203, 215, 272–277, 280–283, 284, 288, 367, 370, 384, 427, 449, 470, 472, 476, 528, 542, 544, 546 adult, 286 Alaskan, 201 anadramous salmon (Renibacterium salmoninarum), 223 Atlantic, 201 chinook (king) (Oncorhynchus tshawytscha), 203, 234, 281 chum (dog) (Oncorhynchus keta), 182, 203, 281, 419 coho (silver) (Oncorhynchus kisutch), 203, 281 embryos, 246 farmed/farm-reared, 204, 289 habitats, 215 hatchery pink, 284, 285 hatchery(ies) , 204, 285, 286, 427, 457, 460, 461, 472, 473 juvenile, 36, 60, 63, 64, 81, 86 North American salmon, 542 ocean-ranched, 289 Pacific salmon, 278, 280, 282, 293, 384 pink (humpy) (Oncorhynchus gorbuscha), 8, 76, 182, 281–287, 419, 422, 427, 447, 449, 457, 459, 461, 462, 463, 466, 467, 468, 470, 472, 474, 475, 476, 477, 488, 498, 531 pink salmon alevins, 477 pink salmon fry, 244, 245, 280, 458, 460, 478, 488 redds, 281, 470 roe, 290 runs, 204, 534 Salmon-Derived Nutrients (SDN), 280
sockeye (red) (Oncorhynchus nerka) , 2, 3, 7, 203, 242, 278, 280–282, 501, 530 stocks, 192 Western Alaska pink salmon, 76 wild pink salmon alevins, 471 wild pink salmon fry, 458 wild pink salmon, 276, 289 salmon driftnets, 328 salmonid(s), 76, 101 , 223, 224, 226, 234, 275, 281, 283, 288 Salton Sea, 236 San Francisco, 278 San Ignatio Lagoon, 352 sand lance (Ammodytes hexapterus), 62–64, 67–69, 101, 102, 328, 329, 332, 337–339, 365, 367, 370 sandfish, 101 Sandman reefs, 201 scallop, 215 dredges, 216 fur, 196 problem, 200 Scotian shelf, 390 Scotland, 237 sculpins, 370 sea cucumbers (Parastichopus californicus), 210 sea ducks, 324, 447 sea lion(s) (Eumetopias jubatus) , 8, 114, 196–199, 218, 299, 355–358, 361–363, 367–378, 452, 535, 550 Steller (Eumetopias jubatus), 6, 67, 115, 132, 192, 196–199, 218 sea otter liver P4501A, 482 sea otter (Enhydra lutris), 1, 2, 8, 114, 182, 191, 199, 201, 211, 419, 423, 426, 447, 449, 450, 451, 454, 480, 481, 482, 483, 485, 486, 487, 488, 501, 532, 534, 539, 541, 542, 550 sea surface extension, 47 sea surface temperature, 26, 40, 44, 273, 297, 338, 388, 394 gradient, 47
584
Subject Index
sea turtles, 236 sea urchins (Strongylocentrotus spp.), 210, 375, 378, 539 Seabird Tissue Archival and Monitoring Project (STAMP), 243 Seal Rocks, 358 seals (Sebastes spp.), 196, 218, 449, 451, 539 grey (Halichoerus grypus), 237 Sebastes alutus, 300, 305, 306 seals, haul out, 224 harbor, 237, 239, 240 seamounts, 15 seasonal pycnocline, 30, 44 seastars, 466 Seattle, 278 seawater density, 13, 20 gradients, 20, 32, 34, 40 structure, 20 vertical density gradients, 20, 40 secondary consumers, 62, 486 sediment fines, 463 Selendeng Ayu, 242 self-organized critical system, 551, 552 semelparity, 141, 170 Semibalams balanoides, 463 semi-diurnal M2 tide, 41 semi-precocial chicks, 96, 104 Semispectral Primitive Equation Model (SPEM), 303 sensible heat exchange, 170, 177 sentinels, 75 seropositive, 236 serotype, 234 sewage, 242 Seward, 18, 20, 26–28, 31, 34, 41, 42, 65, 269, 270 sharks, 234, 366 shear, 170 shearwaters, 68 shelf, 278 shelf/coastal waters, 55, 64 bathymetry, 15 bottom boundary layer, 38
break, 14 circulation, 24, 34, 38, 41 edge, 305 waves, 170 shelfbreak, 13, 14, 16, 32, 33, 36–38, 40, 41, 42 eddies, 38 exchange mechanisms, 40 Shelikof Straight, 16, 23, 24, 26, 34, 37, 38, 88, 89, 92, 198, 303, 308, 419, 424, 437 shell fish, 1, 2, 203, 208, 224, 226, 333, 334 Shetlands, 347 Shoreline Cleanup Assessment Teams (SCAT), 440, 441, 443, 445, 446, 485 shrimp, 210, 224, 233, 293, 306, 332, 367, 379–381, 383–391, 393, 394, 547 brown shrimp (Farfantepenaeus aztecus), 141 coonstriped (Pandalus hypsinotus), 210 northern shrimp (Pandalus borealis), 1, 5, 135, 136, 138, 141, 142, 144, 145, 210, 213, 379, 380, 383, 385, 389, 390, 393 Pandalus tridens, 136 Pandalid shrimp, 136, 139, 142 Pandalus goniurus, 136 Pandalus hysinotus, 136 “pink”, 210 sidestriped (Pandalopsis dispar), 136, 210 spot (Pandalus platyceros), 210 Siberia, 18, 483 Siberian High, 18 siblicide, 113 sibling competition, 107, 108 silicate, 30, 31, 56, 536 SIR model, 220, 222, 225 foci, infection, 222 infected, 221 infection rate, 222
Subject Index 585
infection(R), 221, 223 mortality rate, 222 susceptible class(S), 221 Sitka, 355 size, 90 size-dependent, 287 size-dependent mortality, 80 skeletonema, 57 Sleepy Creek, 471 slumping, 183 continental shelf, 183 slumps, 183 small calanoid, 57 small diatoms, 58, 60, 144 small phytoplankton, 53, 71 small protozoans, 53, 58 small zooplankton, 83 small-mesh trawl surveys, 301 smelt(s), 69, 76, 101 smolting, 280, 287, 384 snails, 375, 463, 464 snare booms, 437 Snug Harbor, 435 solar heating, 29, 30, 44 solar radiation, 267, 386, 387 solar radiative flux/heating, 24 somatic genetic damage, 477 Sound Ecosystem Assessment (SEA), 285 South Carolina, 419 South Equatorial Current, 46 South Peninsula, 384 Southcentral Stock, 360 Southeast Alaska Pacific herring, 293 Southeast Alaska, 14, 15, 18, 29, 34, 122, 139, 291, 292, 319, 320, 323, 327, 355, 357, 358, 383, 386, 387, 453, 542 Southeastern Stock, 360 Southern District, 384 southern oscillation, 176 Southwest Stock, 360 spawning areas, 82, 88 euphausiids, 72
habitats, 289, 466 philopatry, 88, 170 site, 88 species interactions, 187, 211 spilled petroleum, 243 spinal deformities, 456 spiny dogfish (Squalus acanthias), 138 sporozoan 232 Haplosporidium nelsoni, 232 spruce boughs, 290 squid(s), 101, 102, 288, 367, 370, 378 SST, 297, 299 stable isotope, 280 analyses, 73 standards of proof, 427 standing stock(s), 384, 550 station P, 276 Steller sea cow, 521, 535, 542 Steller sea lion (Eumetopias jubatus), 355–358, 361–363, 367–378, 452, 541 stock, 214 assessments, 204, 210, 291–293, 526 fish, 210 parental, 210 spawning, 81, 93, 210, 392, 547 surveys, 379 stomach analyses, 73 contents, 73 storm petrels, 329, 334 fork-tailed (Oceanodroma furcata), 317, 329 Leach’s (Oceanodroma leuorhoa), 317, 321, 329 stratification, 20, 29, 30, 34, 37, 38, 40–42, 53, 59, 88, 172, 177, 271–273, 288, 386, 530, 536, 543, 546 stratified, 338, 386, 433 stratified-random sampling, 443 straying, 461 strong cross-shelf flows, 37 strong recruitment, 135
586
Subject Index
sub-arctic, 242 ecosystem, 419 gyre, 526 marine organisms, 12 ocean, 11 pacific, 58, 59 zones, 68 subduction, 257 subsea embankments, 15 turbidity flows, 14 subsidence, 14, 15 subsistence, 7, 69 harvests, 542 hunters, 362, 366 use, 362 substantial shears, 36 subsurface deposits, 447, 531 oil, 439, 440, 441, 442, 443, 470 temperature, maximum, 26 subtidal communities, 461, 466, 527, 534 habitat, 534 oil, 442 sulphides, 230 summer heat gain, 26 mixed layer, 30 solstice, 24 surface cooling, 26, 35 Ekman transport, 21 feeders, 64, 98 mixed layer, 30 trawls, 542 survey small mesh trawl, 379 small-mesh, 381 survival strategies, 12, 75, 96, 112, 135, 139, 143–145, 325, 332, 334, 365, 372, 375 susceptible animals, 221– 223
sustainable, 201 Sweden, 56, 237 switching nests, 107 Tahiti, 176 tangle nets, 208 tectonic process, 14, 15 tectonics, 180, 181 teleosts, 224, 226, 233 temperature, 178 anomalies, 173, 269 anomaly, 173 gradient, 179, 269 temperature dependent, 224, 277 growth, 80 temperature gradients, 45 temperature-driven mechanism, 387 terns, 101 Thalassiosira spp., 57, 143, 144, 387, 388, 393 thermal stratification, 30, 546 thermal stress, 485 thermocline(s), 30, 35, 40, 44, 100, 170, 176, 546 thermodynamics, 445 thermoregulation, 116, 121, 123, 129, 132, 133 Thysanoessa inermis, 66 Thysanoessa longipes, 66 Thysanoessa spp., 64, 66 tidal amplitudes, 41 tidal dissipation rate, 41 mixing processes, 30 period, 41 suction, 39 velocities, 42 tidal-induced mixing, 536 tidal current(s), 39, 42 gradients, 42 tiny diatoms, 53, 71 toluene, 246, 429, 433, 504 top predator, 486
Subject Index 587
top-down, 543 control, 309, 389, 539, 540, 542, 544, 546, 547 effects, 68, 86 forces, 530, 538 forcing effects, 378 forcing, 373, 543 hypothesis, 390, 391 predator control, 540 processes, 363 regulation, 309 trophic, 68 Total BTEX, 504 toxic chemicals, 241, 244 heavy metals, 241 organic, 241 organo-metallic, 241 toxicants, 244 toxic metal, 454 transient, 452 trawl trawl, 199 beam, 210 bycatch, 207 fishery, 379, 540 surveys, 383 otter, 216 shrimp, 216 trimethylnaphthalene, 430 trophic, 30, 47, 54, 55, 57, 58, 61, 62, 68, 71–73, 75, 85, 130, 133, 135, 278, 288 cascade(s), 486, 487, 530, 539, 540 dynamics, 541 efficiency, 72 euphotic zone, 177, 179 interactions, 486 level(s), 61, 179, 189, 217, 238, 239, 277, 320, 326, 328, 337, 377, 454, 528, 540, 542, 549, 555 linkages, 54, 72, 544, 552 nutrient depletion, 179 phytoplankton, 180
primary production, 180 process, 62 pyramid, 535 relationships, 191 spring phytoplankton bloom, 179 status, 73 structure, 535, 549 system, 528 web, 530 trophically, 191 troughs, 15 trout, 244 Dolly Varden, 80, 286, 493 North American, 234 rainbow, 234 Trustee Council, 489 tsunamis, 14, 180, 186, 423 tunicates, 54, 137 Turagain Arm, 16 turbidity, 14, 90 turbulence, 57, 90 turtle Florida green (Chelonian mydas), 236 green turtles, 236 Kemp-Ridley (lepiochelys kempii), 236 loggerhead (Caretta caretta), 230 two-step transfer mechanism, 71 U.S. Coast Guard, 215 U.S. Fish and Wildlife Service, 204 United States (US), 174, 419, 420, 204, 208, 218, 220, 233 UK, 234, 486 ultraviolet radiation, 430 unalaska, 242 Unimak Bight, 89 Unimak Pass, 17, 22, 34, 357, 370 United States Fish and Wildlife Service (USFWS), 423 un-weathered, 442 uplifting, 15, 18, 22, 181, 184 Cape Clear, 180 Herring Bay, 182
588
Subject Index
uplifting (Continued) Montague Island, 182 Prince William Sound, 182 Upper Valdez Arm, 423 upwelling, 12, 176, 177, 274, 294, 337, 338, 521, 524, 537 coastal, 12, 177 favorable, 21, 22, 266 equatorial, 176 seasonal, 12 urban runoff, 243 urchin barrens, 539 Valdez Arm, 16 Valdez Narrows, 423 Valdez Oil Terminal, 243 Valdez, 279, 432 vertical current shears, 36 vertical density gradients, 20, 40 vertical mixing, 29, 30, 39, 42, 51, 52, 177, 274, 388 vertical stratification, 34 vertical turbulence, 57 Vibrio cholera, 226 Vibrio parahaemolyticus, 226, 230 Vibrio vulnificus, 226 Victoria Pattern, 521, 528 viral disease, 457, 458 viral hemorrhagic septicemia virus (VHSV), 85, 235, 457, 458 Vitus Bering, 1741, 201 volcanic eruptions, 180, 187 volcanoes, occurrence of, 15, 183 Alaskan arc, 186 Wrangell mountains, 186 volcanism Aleutian arcs, 183 Aleutian chain, 183 Aniak island, 183 Denali, 183 Waddell Sea, 237 waistband species, 86 warm climate regime, 545
warm regime(s), 543, 546, 547 Washington, 11, 193, 234, 276 Washington–region shelf waters, 276 water-column mixing, 388 weathered/weathering, 21, 447, 468, 474, 475, 476, 477, 421, 435, 437, 439, 443, 445, 446, 447, 531 oil, 443, 447 PAH, 474, 475 weathers, 448 weathervane scallops (Patinopectin caurinus), 210 Western Prince William Sound (WPWS), 481 western stock, 356, 357, 368, 369, 372, 373 whales, 2, 3, 192, 194, 299, 361, 541 AB pod, 452–454, 487 baleen whales, 546 blue ( Balenoptera musculus), 193, 194, 218, 540 bowhead, 193 Bryde’s (Balenoptera edeni), 193 fin whale, (Balenoptera physalus), 3, 193, 194, 540 great whales, 193, 361, 535 humpback (Megaptera novaeangliae), 64, 67, 68, 83, 193, 194, 540 killer whales, 122, 130, 131, 134, 191, 242, 363, 368, 373, 375, 378, 536, 542, 547, 550 Orcas, 239 right ( Eubalena glacialis), 198–199 sei (Balenoptera borealis), 199–200, 540 sperm (physeter macrocephalus), 193, 194, 218 whelks, 486 whole body energy content (WBE), 61, 83, 84 William Sound Aquaculture Corporation, 279, 280
Subject Index 589
winds, 11, 12, 18, 20–24, 29, 30, 32–34, 37–39, 44–47, 52 along-shelf, 22 downwelling, 22 mesoscale wind fields, 24 mixing, 26, 43, 53, 59, 266 spatial wind field, 22 Southeast Trade Winds, 45 Northeast Trade winds, 45 Westerlies, 45 stress, 21–23, 41–45, 166, 271 forcing, 21, 22, 36, 169 alongshore, 21, 24 field, 22, 24, 171 gap, 23, 24 mesoscale, 23, 24 wind stress distribution, climatological, 44 wind-forced, 274 wind-mixed layer, 48 wing loading, 98–100, 113
winter heat loss, 26 mesoscale, 24 solstice, 24 xylene, 246, 429, 505 yellowfin sole (Pleuronectes asper), 61, 138, 139, 391 young-of-the-year, 138, 139, 394 zonal wind stress, 45 zone Fucus, 182 intertidal, 181 zooplankters, 49, 53, 69, 94, 278, 529 zooplankton sheltering, 287 zooplankton, 30, 36–38, 47, 49, 53, 54, 57, 58, 60, 62, 68, 70, 71, 77, 79, 83, 88, 93, 94, 102, 144, 268, 273–277, 282, 286–288, 293, 294, 296–299, 309, 337, 384, 385, 387, 388, 393, 526, 544, 546
This page intentionally left blank