Benguela: Predicting a Large Marine Ecosystem
Large Marine Ecosystems – Volume 14 Series Editor:
Kenneth Sherman Director, Narragansett Laboratory and Office of Marine Ecosystem Studies NOAA-NMFS, Narragansett, Rhode Island, USA and Adjunct Professor of Oceanography Graduate School of Oceanography, University of Rhode Island Narragansett, Rhode Island, USA
On the cover The main cover picture illustrating the complexity of the Benguela Current Large Marine Ecosystem (BCLME) and adjacent regions is an AQUA MODIS level three, 4 km resolution, chlorophyll image for the week 2-10 February 2004, obtained from the NASA Oceancolor webpage: http://oceancolor.gsfc.nasa.gov/cgi/level3.pl The top picture, with the BCLME box inset, is the global map of average primary productivity and the boundaries of the 64 Large Marine Ecosystems (LMEs) of the world, available at www.edc.uri.edu/lme. The annual productivity estimates are based on SeaWIFS data collected between September 1998 and August 1999. The color enhanced image was provided by Rutgers University.
A list of recent publications in this series appears at the end of this volume.
Benguela: Predicting a Large Marine Ecosystem Edited by Vere Shannon Honorary Professor, Department of Oceanography University of Cape Town South Africa Gotthilf Hempel Science Advisor, Senate of Bremen, Germany Emeritus Professor, Bremen and Kiel Universities Germany Paola Malanotte-Rizzoli Professor, Department of Earth, Atmospheric and Planetary Sciences Massachusetts Institute of Technology Cambridge, Massachusetts United States Coleen Moloney Senior Lecturer, Department of Zoology University of Cape Town South Africa John Woods Emeritus Professor, Department of Earth Science and Engineering Imperial College London United Kingdom
Technical editor Sara P. Adams - Large Marine Ecosystem Program - Narragansett RI - USA
Amsterdam - Boston - Heidelberg - London - New York - Oxford - Paris San Diego - San Francisco - Singapore - Sydney - Tokyo
Elsevier Radarweg 29, PO Box 211, 1000 AE Amsterdam, The Netherlands The Boulevard, Langford Lane, Kidlington, Oxford OX5 1GB, UK First edition 2006 Copyright © 2006 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: 978-0-444-52759-2 ISBN-10: 0-444-52759-1 ISBN-13: 978-0-444-52760-8 ISBN-10: 0-444-52760-5 (CD-rom) ISSN: 1570-0461 For information on all Elsevier publications visit our website at books.elsevier.com Printed and bound in The Netherlands 06 07 08 09 10 10 9 8 7 6 5 4 3 2 1
v
Series Editor’s Introduction The world’s coastal ocean waters continue to be degraded from unsustainable fishing practices, habitat degradation, eutrophication, toxic pollution, aerosol contamination, and emerging diseases. Against this background is a growing recognition among world leaders that positive actions are required on the part of governments and civil society to redress global environmental and resource degradation with actions to recover depleted fish populations, restore degraded habitats and reduce coastal pollution. No single international organization has been empowered to monitor and assess the changing states of coastal ecosystems on a global scale, and to reconcile the needs of individual nations to those of the community of nations for taking appropriate mitigation and management actions. However, the World Summit on Sustainable Development convened in Johannesburg in 2002 in recognition of the importance for coastal nations to move more expeditiously toward sustainable development and use of ocean resources, declared that countries should move to introduce ecosystem-based assessment and management practices by 2010, and by 2015, restore the world’s depleted fish stocks to maximum levels of sustainable yields. At present, 121 developing countries are moving toward these targets in joint international projects supported, in part, by financial grants by the Global Environment Facility in partnership with scientific and technical assistance from UN partner agencies (e.g. UNIDO, UNEP, UNDP, IOC, FAO), donor countries and institutions and nongovernmental organizations including the IUCN (World Conservation Union). Many of these projects are linked to ecosystem-based efforts underway in Europe and North America in a concerted effort to overcome the North-South digital divide. The volumes in the new Elsevier Science series on Large Marine Ecosystems are bringing forward the results of ecosystem-based studies for marine scientists, educators, students and resource managers. The volumes are focused on LMEs and their productivity, fish and fisheries, pollution and ecosystem health, socioeconomics and governance. This volume in the new series, “Benguela: Predicting a Large Marine Ecosystem” progresses systematically and innovatively from studies that set the present scene, to studies constituting the “cutting edge” of forecasting changing states of the Benguela Current LME (BCLME) and move ahead to a fully integrated BCLME forecasting system. The authors are forward looking and quite deliberate in tightening up the linkages between science based assessments of the changing states of the BCLME and the socioeconomic benefits to be derived by the people of Angola, Namibia, and South Africa from the application of an LME forecasting system that is readily transparent and clearly adaptable to others of the world’s 64 LMEs. The volume provides an important LME baseline from which to support the stated goals of the Global Ocean Observing System (GOOS) and the Global Environmental Observing System of Systems (GEOSS). The volume is the fourteenth in the Elsevier Science New LME Series. Recent volumes in the Series are listed at http://www.elsevier.com/wps/find/bookseriesdescription.cws_home/BS_LME/description. Kenneth Sherman, Series Editor Narragansett, Rhode Island
ThisPage pageis intentionally This Intentionally left Leftblank Blank
vii
ThisPage pageis intentionally This Intentionally left Leftblank Blank
ix
THE PUBLICATION OF THIS BOOK HAS BEEN MADE POSSIBLE THROUGH THE GENEROUS SPONSORSHIP AND SUPPORT PROVIDED BY THE ABOVE ORGANISATIONS
ThisPage pageis intentionally This Intentionally left Leftblank Blank
xi
Foreword In recent years, there has been considerable international interest in ocean monitoring and operational oceanography to enable responsible management of marine and coastal resources and support a variety of other maritime activities through timeous provision of appropriate information, including forecasts. This is reflected in the strategies of the International Waters Programme of the Global Environment Facility (GEF) through its Large Marine Ecosystem (LME) initiatives and the Global Ocean Observing System (GOOS) of the Intergovernmental Oceanographic Commission (IOC), and endeavours are at present being made to develop close links between GOOS and the LMEs. The need for accurate forecasting, contingency planning and effective reporting mechanisms to managers and the public at large has more recently been highlighted in the wake of the devastating tsunami in SE Asia in 2005, hurricane Katrina in the USA and the broader global impacts of El Niño events and climate change. In developed countries and regions, significant advances have been made in ocean monitoring and observing systems and new generations of metocean buoys and satellite technology have been put in place which allow real-time monitoring and modeling of the processes taking place in the marine environment. This is not the case in many developing parts of the world, Africa in particular. The IOC, through its GOOS-Africa Programme and the network of African LME Programmes, is attempting to address this by developing affordable, implementable and sustainable ocean observing systems to service the needs of African countries, regions and the continent as a whole. Of major concern for the Benguela region and a key goal of the Strategic Action Plan (SAP) of the BCLME Programme has been the establishment of a viable and costeffective forecasting system that can provide resource and environmental managers and indeed the public at large in the region with early warning of catastrophic events. Benguela Niños, low oxygen anomalies and extensive harmful algal blooms periodically occur in the marine and coastal environment of the BCLME with devastating consequences on the living marine resources. This is particularly true for the northern Benguela. In November 2004, a four day International Workshop on Forecasting and Data Assimilation in the Benguela and Comparable Systems was held in Cape Town to address the key policy actions of forecasting environmental variability and its impacts on the BCLME. The information, knowledge, wisdom and advice resulting from this workshop are captured in this definitive peer-reviewed book entitled, Benguela – Predicting a Large Marine Ecosystem. This book will be a significant contribution to the BCLME Programme and its sustainable management objectives, and a blueprint for application
Foreword
xii
in other LMEs around the world and for fast-tracking the objectives of several international science organisations. The BCLME Programme is highly appreciative of the efforts of all the contributors to this volume and, in particular, I wish to thank Professors Vere Shannon, Gotthilf Hempel, Paola Malanotte-Rizzoli, Coleen Maloney and John Woods for so enthusiastically editing this volume and putting together the accompanying CD-ROM. Special thanks are due to Dr. Sara Adams, LME Program at the Northeast Fisheries Science Center’s Narragansett Laboratory in Rhode Island, USA, for her technical editing skills and preparing the book in camera ready format for publication. Finally, we of the Benguela LME community would like to dedicate this volume, the 14th in Elsevier’s Large Marine Ecosystems Series to Dr Kenneth Sherman, the founding father of the “Large Marine Ecosystems ” global movement which is now widely recognised as one of the most effective and practical strategies for operationalising the ecosystem approach to management of marine resources. His tireless efforts, support and guidance over the years in collaboration with Al Duda of the Global Environment Facility (GEF) have ensured the successful implementation of the BCLME Programme and of LME initiatives in other parts of the world. The publication of this volume was made possible through support of the BCLME Programme, the National Oceanic and Atmospheric Administration (NOAA) and the University of Cape Town. Michael John O’Toole Chief Technical Advisor BCLME Programme
xiii
Contributors Justin Ahanhanzo GOOS Africa Intergovernmental Oceanographic Commission IOC of UNESCO Paris, FRANCE Jürgen Alheit Baltic Sea Research Institute Institut für Ostseeforschung Warnemünde GERMANY Miriam Andrioli Maritime Division Forecasting Department Servicio Meteorologico Nacional 1002 Buenos Aires ARGENTINA Hernam Arango Institute of Marine and Coastal Sciences Rutgers University State University of New Jersey New Brunswick, New Jersey USA Claire Attwood Media Liaison Muizenberg 7945 Cape Town SOUTH AFRICA Awie Badenhorst SA Pelagic Fishing Industry Association SA Inshore Cape Town 8000 SOUTH AFRICA Geoff W. Bailey Department of Environmental Affairs and Tourism Marine and Coastal Management Roggebaai 8012 Cape Town, SOUTH AFRICA
Ray G. Barlow Department of Environmental Affairs and Tourism Marine and Coastal Management Rogge Bay 8012 Cape Town, SOUTH AFRICA Chris Bartholomae Ministry of Fisheries and Marine Resources National Marine Research and Information Centre Swakopmund NAMIBIA Eric D. Barton Spanish Council for Scientific Research Instituto de Investigaciones Marinas Vigo, SPAIN Stewart Bernard Department of Oceanography University of Cape Town Rondebosch 7701 Cape Town SOUTH AFRICA Geoff Brundrit Department of Oceanography University of Cape Town Rondebosch 7701 Cape Town SOUTH AFRICA Deidre Byrne School of Marine Sciences University of Maine Orono, Maine USA Rudi Cloete Ministry of Fisheries and Marine Resources National Marine Research and Information Centre Swakopmund, NAMIBIA
xiv Andy C. Cockcroft Department of Environmental Affairs and Tourism Marine and Coastal Management Rogge Bay 8012 Cape Town, SOUTH AFRICA R. J. M. Crawford Marine and Coastal Management (MCM) Department of Environmental Affairs and Tourism (DEAT) Roggebay 8012 Cape Town SOUTH AFRICA Philippe Cury Institut Recherche Développement (IRD) Centre de Recherche Halieutique Méditerranéenne et Tropicale Paris, Sète Cedex 10 FRANCE Antonio da Silva Instituto de Investigação Pesqueira Ministerio das Pescas Luanda, ANGOLA Hervé Demarcq Institut Recherche Développement (IRD) Centre de Recherche Halieutique Méditerranéenne et Tropicale Paris, Sète Cedex 10 FRANCE Chris M. Duncombe Rae Department of Environmental Affairs and Tourism Marine and Coastal Management (MCM) Roggebaai 8012 Cape Town SOUTH AFRICA Alex Fawcett Department of Oceanography University of Cape Town Rondebosch 7701 SOUTH AFRICA Katje Fennel Institute of Marine and Coastal Sciences Rutgers University New Brunswick, New Jersey USA
Contributors Wolfgang Fennel Baltic Sea Institute Institut für Ostseeforschung Warnemünde GERMANY Quilanda Fidel Instituto de Investigação Marinha Ministerio das Pescas Luanda ANGOLA John G. Field Zoology Department, Marine Biology Research Institute University of Cape Town Rondebosch 7701 Cape Town, SOUTH AFRICA Jim Fitzpatrick HydroQual Inc. MacArthur Blvd. Mahwah, New Jersey USA Pierre Florenchie Department of Oceanography University of Cape Town Rondebosch 7701 Cape Town, SOUTH AFRICA Regina Folorunsho Nigerian Institute for Oceanography and Marine Research Victoria Island Lagos, NIGERIA Peter J. S. Franks Scripps Institution of Oceanography University of California San Diego, California USA Pierre Fréon Institut Recherche Développement (IRD) Halieutique Méditerranéenne et Tropicale Paris, Cedex 10 FRANCE Ema Gomes Ministry of Petroleum of Angola 1279 C Luanda ANGOLA
xv
Contributors Paul Goodman Department of Geosciences University of Arizona, Tucson, AZ USA
Astrid Jarre Danish Institute for Fisheries Research North Sea Centre Hirtshals DENMARK
Leticia Greyling National Ports Authority of South Africa (NPA) Braamfontein, Jhb, 2017 SOUTH AFRICA
Ashley Johnson Department of Environmental Affairs and Tourism Marine and Coastal Management Rogge Bay 8012 Cape Town SOUTH AFRICA
Marten Grundlingh Council for Scientific and Industrial Research – CSIR Environmentek 7599 Stellenbosch SOUTH AFRICA Johannes Guddal Norwegian Meteorological Institute (DNMI Region W)) Bergen, NORWAY Dale B. Haidvogel Institute of Marine and Coastal Sciences Rutgers University New Brunswick, New Jersey USA Jenny A. Huggett Department of Environmental Affairs and Tourism Marine and Coastal Management Rogge Bay 8012 Cape Town SOUTH AFRICA Ian T. Hunter South African Weather Service South African Weather Bureau Pretoria 0001 SOUTH AFRICA Larry Hutchings Department of Environmental Affairs and Tourism Marine and Coastal Management Rogge Bay 8012 Cape Town SOUTH AFRICA David W. Japp Capricorn Fisheries and Monitoring Waterfront 8002 SOUTH AFRICA
Anél Kemp Council for Scientific and Industrial Research CSIR Environmentek 7599 Stellenbosch SOUTH AFRICA Souad Kifani Institut National de Recherche Halieutique Casablanca MOROCCO Vamara Koné Institut de Recherche pour le Développement (IRD) Centre de Recherches Halieutiques Mediterranéenne et Tropicale (CRH) 34203 Sète Cedex, FRANCE Anja Kreiner Ministry of Fisheries and Marine Resources National Marine Research and Information Centre Swakopmund NAMIBIA Raphael M. Kudela Ocean Sciences Department University of California Santa Cruz, California USA Uli Lass Institut für Ostseeforschung Institute for Baltic Sea Research Warnemünde, Rostock GERMANY
Contributors
xvi Deon Louw Ministry of Fisheries and Marine Resources National Marine Research and Information Centre Swakopmund NAMIBIA
Pedro M. S. Monteiro Coast Programme Council for Scientific and Industrial Research – CSIR Environmentek 7599 Stellenbosch SOUTH AFRICA
Lima Maartens De Beers Marine Namibia Windhoek NAMIBIA
Pat D. Morant Coast Programme Council for Scientific and Industrial Research – CSIR Environmentek 7599 Stellenbosch SOUTH AFRICA
Eric Machu Institut de Recherche pour le Développement (IRD) Centre de Recherches Halieutiques Mediterranéenne et Tropicale 34203 Sète Cedex, FRANCE Paola Malanotte-Rizzoli Department of Earth, Atmospheric and Planetary Science Massachusetts Institute of Technology (MIT) Cambridge, Massachusetts USA Thomas Malone Horn Point Laboratory Center for Environmental Science University of Maryland Cambridge, Maryland, USA Patrick Marchesiello IRD Institut Recherche Développement Halieutique Méditerranéenne et Tropicale Paris, Cedex 10 FRANCE Yukio Masumoto Department of Earth and Planetary Science Graduate School of Science The University of Tokyo JAPAN Coleen L. Moloney Zoology Department University of Cape Town Rondebosch 7701 SOUTH AFRICA
Kathie R. Peard Ministry for Fisheries and Marine Resources Lüderitz Marine Research Lüderitz NAMIBIA Patrick Penven Institut de Recherche pour le Développement (IRD) Centre de Recherches Halieutiques Mediterranéenne et Tropicale (CRH) 34203 Sète Cedex, FRANCE Pavitray Pillay SANCOR Secretariat Foundation for Research Development SOUTH AFRICA Grant C. Pitcher Research Aquarium, Sea Point Department of Environmental Affairs and Tourism Marine and Coastal Management Rogge Bay 8012 Cape Town SOUTH AFRICA Christopher J. C. Reason Department of Oceanography University of Cape Town Rondebosch 7701 Cape Town SOUTH AFRICA Mathieu Rouault Department of Oceanography University of Cape Town Rondebosch 7701 Cape Town SOUTH AFRICA
Contributors Jean-Paul Roux Ministry of Fisheries and Marine Resources Lüderitz Marine Research Lüderitz, NAMIBIA Claude Roy Institut Recherche Développement (IRD) Centre IRD de Bretagne BP 70 29280 Plouzané FRANCE Hidehary Sasaki Earth Simulator Center JAMSTEC Yokohama JAPAN Lynne J. Shannon Department of Environmental Affairs and Tourism Marine and Coastal Management (MCM) Rogge Bay 8012 Cape Town SOUTH AFRICA L. Vere Shannon Department of Oceanography University of Cape Town Rondebosch 7701 Cape Town SOUTH AFRICA Kenneth Sherman NOAA, National Marine Fisheries Service Northeast Fisheries Science Center Narragansett Laboratory Narragansett, Rhode Island USA Frank A. Shillington Department of Oceanography University of Capetown Rondebosch 7701 Cape Town SOUTH AFRICA Geoff G. Smith Coast Programme Council for Scientific and Industrial Research CSIR Environmentek 7599 Stellenbosch SOUTH AFRICA
xvii Neville Sweijd BENEFIT Secretariat c/o Ministry of Fisheries and Marine Resources National Marine Research and Information Centre Swakopmund, NAMIBIA Tomoki Tozuka COE Research Associate Depatrment of Earth and Planetary Science Graduate School of Science University of Tokyo JAPAN Roy C. van Ballegooyen Hydroodynamics Coast Programme and Marine and Estuarine Water Quality Council for Scientific and Industrial Research CSIR Environmentek 7599 Stellenbosch SOUTH AFRICA Carl D. van der Lingen Department of Environmental Affairs and Tourism Marine and Coastal Management Rogge Bay 8012 Cape Town SOUTH AFRICA Anja K. van der Plas Ministry of Fisheries and Marine Resources National Marine Research and Information Centre Swakopmund NAMIBIA Jennifer Veitch Department of Oceanography University of Cape Town Rondebosch 7701 Cape Town SOUTH AFRICA F. Vaz-Velho Instituto de Investigação Marinha Ministerio das Pescas Luanda ANGOLA
xviii Hans M. Verheye Department of Environmental Affairs and Tourism Marine and Coastal Management Rogge Bay 8012 Cape Town SOUTH AFRICA C. K. Wainman Institute for Maritime Technology Simon’s Town 7995 SOUTH AFRICA Scarla J. Weeks University of Queensland Centre for Marine Studies Brisbane AUSTRALIA
Contributors John Wilkin Institute of Marine and Coastal Sciences Rutgers University State University of New Jersey New Brunswick, New Jersey USA John D. Woods Complex System Modelling Department of Earth Science and Engineering Royal School of Mines Imperial College London London SW7 2AZ UK Toshio Yamagata Department of Earth and Planetary Science Graduate School of Science University of Tokyo JAPAN
xix
Contents Series Editor’s introduction Ministers’ page: Towards forecasting a changing ocean: An African Perspective (Salomão José Xirimbimbi, Minister of Fisheries, Angola; Abraham Iyambo, Minister of Fisheries and Marine Resources, Namibia; Marthinus van Schalkwyk, Minister of Environmental Affairs and Tourism, South Africa) Sponsorship page Foreword by Michael John O’Toole List of contributors
PART I: BY WAY OF INTRODUCTION 1. A plan comes together Vere Shannon
v vii
ix xi xiii
1 3
2. Forecasting within the context of Large Marine Ecosystem Programs Kenneth Sherman
11
3. The Global Ocean Observing System for Africa (GOOS Africa): Monitoring and Predicting in Large Marine Ecosystems Justin Ahanhanzo
35
PART II: SETTING THE SCENE Data, time series and models: What we think we know about variability in the Benguela and comparable systems.
47
4. Large scale physical variability of the Benguela Current Large Marine Ecosystem (BCLME) Frank A. Shillington, Chris J. C. Reason, Chris M. Duncombe Rae, Pierre Florenchie and Patrick Penven
49
5. Low oxygen water (LOW) variability in the Benguela system: Key processes and forcing scales relevant to forecasting Pedro M. S. Monteiro and Anja K. van der Plas
71
6. Variability of plankton with reference to fish variability in the Benguela Current Large Marine Ecosystem – An overview Larry Hutchings, Hans M. Verheye, Jenny A. Hugget, Hervé Demarcq, Rudi Cloete, Ray G. Barlow, Deon Louw and Antonio da Silva
91
xx
Contents
7. The variability and potential for prediction of harmful algal blooms in the southern Benguela ecosystem Grant C. Pitcher and Scarla J. Weeks
125
8. Resource and ecosystem variability, including regime shifts, in the Benguela Current system Carl D. van der Lingen, Lynne J. Shannon, Philippe Cury, Anja Kreiner, Coleen L. Moloney, Jean-Paul Roux and F. Vaz-Velho
147
9. Modelling, forecasting and scenarios in comparable upwelling ecosystems --California, Canary, Humboldt Pierre Fréon, Jürgen Alheit, Eric D. Barton, Souad Kifani and Patrick Marchesiello
185
PART III: HOPES, DREAMS AND REALITY Forecasting in the Benguela: Our collective wisdom
221
10. Influences of large scale climate modes and Agulhas system variability on the BCLME region 223 Chris J. C. Reason, Pierre Florenchie, Mathieu Rouault and Jennifer Veitch 11. Developing a basis for detecting and predicting long-term ecosystem 239 changes Astrid Jarre, Coleen L. Moloney, Lynne J. Shannon, Pierre Fréon, Carl. D. van der Lingen, Hans M. Verheye, Larry Hutchings, Jean-Paul Roux and Philippe Cury 12. The requirements for forecasting harmful algal blooms in the Benguela Stewart Bernard, Raphael M. Kudela, P. J. S. Franks, Wolfgang Fennel, Anél Kemp, A. Fawcett and Grant C. Pitcher
273
13. Low oxygen water (LOW) forcing scales amenable to forecasting in the Benguela ecosystem Pedro M. S. Monteiro, Anja K. van der Plas, Geoff W. Bailey, Paola Malanotte-Rizzoli, Chris M. Duncombe Rae, Deidre Byrnes, Grant Pitcher, Pierre Florenchie, Patrick Penven, Jim Fitzpatrick and H. Uli Lass
295
14. Forecasting shelf processes of relevance to living marine resources in the BCLME Carl D. van der Lingen, Pierre Fréon, Larry Hutchings, Claude Roy, Geoff W. Bailey, Chris Bartholomae, Andy C. Cockcroft, John. G. Field, Kathie R. Peard and Anja K. van der Plas
309
Contents
xxi
15. Environmental data requirements of maritime operations in the Benguela coastal ocean Marten.L. Gründlingh, Pat D. Morant, Roy C. van Ballegooyen, Awie Badenhorst, Ema Gomes, Leticia Greyling, Johannes Guddal, Ian T. Hunter, David W. Japp, Lima Maartens, Kathie R. Peard, Geoff G. Smith and C. K. Wainman
PART IV:
THE WAY AHEAD
349
373
16. Towards a future integrated forecast system Geoff Brundrit, Chris Bartholomae, Quilanda Fidel, Ashley Johnson and Johannes Guddal
375
17. Forecasting a large marine ecosystem John Woods
387
INDEX
401
ThisPage pageis intentionally This Intentionally left Leftblank Blank
xxiii
CD-ROM Contents Benguela: Predicting a Large Marine Ecosystem INSTRUCTIONS TO THE USER The main menu (FOLDERS A through F) allows the user to view documents on the CD, browse the BCLME Forecasting Workshop Webpage, install required software and browse the contents of the CD. Adobe Reader 7.0 is required to view the PDF documents directly from the menu system and the installation software is located in Folder F of the main menu. Although animations can be viewed directly from the CD menu, links to the QuickTime and IrfanView websites are also included in Folder F. Both Packages can be freely downloaded from the given websites and allow the user to have greater control when viewing mov, flic or gif animations. All the animations close automatically after running once, or can be closed with the escape <esc> key. The menu system is navigated by moving the mouse cursor over the folder icons. Menu titles will pop up and the folder letter will be highlighted as the mouse cursor moves over the folder. Click on the highlighted folder letter to access information on the CD. The CD menu system was developed for viewing in Windows XP and Windows 2000. INTRODUCTION PART 1 – (FOLDER A) provides comprehensive details about the International Workshop on Forecasting and Data Assimilation in the Benguela and Comparable systems, held in Cape Town, South Africa in November 2004, its planning, a persoal perspective about the Workshop in the form of a concluding summary by John Woods, and closing remarks by Gotthilf Hempel, the ‘Grandfather’ of the BCLME and BENEFIT Programmes. A: Workshop website B. Aspects of BCLME variability amenable to forecasting C. Presentation of workshop summary D. Closing remarks PART 2 – (FOLDER B) – Observations of model outputs - comprises outputs from observational work and models, including several animations which highlight the spatial and temporal variability in the Benguela. A. High resolution ocean general circulation model (OFES), provided by Tomoki Tozuka, Hidehary Sasaki, Yukio Masumoto and Toshio Yamagata.
xxiv
CD-ROM Contents
B. Forecasting anomalous climatic events in the tropical Atlantic sector using the NLOM Prediction System, provided by Pierre Florenchie. C. ROMS modelled surface chlorophyll a, provided by Eric Machu and Vamara Koné. This is a contribution of the IDYLE and ECO-UP Programmes of the IRD and of EUR-OCEANS. D. Monthly climatology of 18 Years of NOAA SST and 5 Years of SeaWiFS chlorophyll a pigment at 4 km resolution for the BCLME, provided by Hervé Demarcq. This is a contribution of the IDYLE and ECO-UP Programmes of the IRD and of EUR-OCEANS. E. Eight-day composites of ROMS modelled SST for the period 1992 – 2000 using realistic wind forcing. This is a contribution of the IDYLE and ECO-UP Programmes of the IRD and of EUR-OCEANS. F. Five-day SST composites of ROMS modelled for ten years using climatological wind forcing, provided by Patrick Penven. G. Five-day composites of AVISO altimetry around southern Africa, provided by Patrick Penven. H. Example of expert system model: Predicting anchovy recruitment PART 3 – (FOLDER C) – Supplementary material presented at the Benguela Forecast Workshop - contains some selected contributions presented at the Benguela Forecast Workshop, providing additional insights and inputs relevant to measuring, modelling and predicting. A. Requirements and needs of a viable observing and forecasting system in Angola, provided by Quilanda Fidel B. Developing operational forecasting capabilities for coastal GOOS, provided by Thomas Malone, Dir. Ocean, US Office, USA. C. Multi-scale modelling studies on the Northeast U.S. continental shelves, provided by Dale B. Haidvogel, John Wilkin, Katje Fennel, Hernam Arango and Paul Goodman D. Mechanisms and tools in oceanographic capacity building. Provided by Miriam Andrioli, Regina Folorunsho, Geoff Brundrit and Johannes Guddal. E. Ecosystem modelling approaches for South African Fisheries Management, provided by Lynne J. Shannon, Coleen L. Moloney, Carl D. van der Lingen, R.J.M. Crawford, Pierre Fréon F. Offshore oil and gas industry: Marine environment needs, provided by Pat D. Morant PART 4 – (FOLDER D) Information about programmes, network and data centre in the BCLME region - describes four major southern African regional initiatives, including the BCLME and BENEFIT Programmes. A. The BCLME Programme, prepared by Claire Attwood B. Benguela Environment Fisheries Interaction and Training (BENEFIT) Programme, prepared by Neville Sweijd C. SADCO: South African Data Centre for Oceanography, prepared by Marten Grundlingh D. South African Network for Coastal and Oceanic Research (SANCOR), prepared by Pavitray Pillay
CD-ROM Contents
xxv
PART 5 – (FOLDER E) - List of websites that authors and editors believe will be useful to readers. PART 6 - (FOLDER F) – Software installation and browse CD A. Launch Adobe Reader 7.0 installer B. Launch QuickTime Website C. Launch IrfanView Website D. Browse CD Contents
ThisPage pageis intentionally This Intentionally left Leftblank Blank
Part I: By Way of Introduction
This Page is Intentionally Left Blank
Large Marine Ecosystems, Vol. 14 V. Shannon, G. Hempel, P. Malanotte-Rizzoli, C. Moloney and J. Woods (Editors) © 2006 Elsevier B.V. All rights reserved.
3
1 A Plan Comes Together Vere Shannon
UNIQUE ENVIRONMENT The Benguela Current Large Marine Ecosystem (BCLME) is situated along the coast of south-western Africa, stretching from east of the Cape of Good Hope in the south equatorwards to the Angola (Cabinda) Front, near the northern border of Angola (Figure 1-1). It encompasses one of the four major coastal upwelling ecosystems of the world which lie at the eastern boundaries of the oceans. Like the Humboldt, California and Canary systems, the Benguela is an important centre of marine food production. The BCLME’s distinctive bathymetry, hydrography, chemistry and trophodynamics combine to make it one of the most productive ocean areas in the world. This high level of primary productivity of the BCLME supports an important global reservoir of biodiversity and biomass of zooplankton, fish, sea birds and marine mammals. Near-shore and off-shore sediments hold rich deposits of precious minerals (particularly diamonds), as well as oil and gas reserves. The natural beauty of the coastal regions, many of which are still pristine by global standards, have also enabled the development of significant tourism along parts of the coast. Pollution, poorly planned coastal developments, population pressure and near-shore activities such as mining are, however, resulting in rapid degradation of some vulnerable coastal habitats. The main area of coastal upwelling extends from the Angola-Benguela Front north of the Angola/Namibia border, southwards around the Cape of Good Hope, and intermittently as far east as Port Elizabeth (Figure 1-1). The upwelling system as we know it is about 2 million years old, and much of the adjacent land area is arid, e.g. the Namib Desert in Namibia. The principal upwelling centre near Lüderitz in southern Namibia, is one of the most concentrated and intense found in any upwelling regime (e.g. Shannon 1985). What also makes the Benguela upwelling system somewhat unique in the global context is that it is bounded at both northern and southern ends by warm water systems, viz. the tropical/equatorial Eastern Atlantic and the Indian Ocean’s Agulhas Current respectively. Sharp horizontal gradients (fronts) exist at the boundaries of the upwelling system, but these display substantial variability in time and in space – at times pulsating in phase and at others not.
4
Figure 1-1 Currents and boundaries of the Benguela Current Large Marine Ecosystem
V. Shannon
A plan comes together
5
Interaction between the BCLME and adjacent ocean systems occurs over thousands of kilometres. For example, much of the BCLME, in particular off Namibia, is naturally hypoxic –even anoxic – at depth partly as a consequence of subsurface flow southwards from the tropical Atlantic. (This hypoxia is compounded by depletion of oxygen from more localised biological decay processes.) The response of the South Atlantic to ENSO has been documented by Colberg et al. (2004) while links between the Benguela and processes in the North Atlantic may also exist. Moreover the southern Benguela lies at a major choke point in the “Global Climate Conveyor Belt” whereby on timescales of decades to centuries warm upper layer waters move from the Pacific via the Indian Ocean into the North Atlantic. (The South Atlantic is the only ocean in which there is a net transport of heat towards the equator.) As a consequence, not only is the Benguela at a critical location in terms of the global climate system, but it is also potentially extremely vulnerable to climate change and climate variability. TEN YEARS OF CLOSE REGIONAL COLLABORATION In mid-1995, recognising the need for a more holistic approach to the study and ultimately the sustainable management of the living resources of the Benguela region, the Namibian Ministry of Fisheries and Marine Resources hosted a Workshop/Seminar on Fisheries Resource Dynamics in the Benguela Current Ecosystem in the coastal town of Swakopmund. At this seminal meeting the seed was sown for two regional cooperative initiatives, BENEFIT and the BCLME Programmes. Country driven jointly by Angola, Namibia and South Africa, and with strong international encouragement and support, particularly from Norway, Germany and France, these Programmes have been instrumental in building goodwill, trust and close cooperation at all levels. Launched in 1997, BENEFIT (BENguela-Environment-FisheriesInteraction & Training) has as its overall goal the development of enhanced science capability required for the optimal and sustainable utilization of living resources of the Benguela by (a) improving knowledge and understanding of the dynamics of important commercial stocks, their environment and linkages between the environmental processes and the stock dynamics, and (b) building appropriate human and material capacity for marine science and technology in the countries bordering the Benguela ecosystem. Following its conception in mid-1995, an embryonic plan for the BCLME Programme was formulated by Kenneth Sherman, Les Clark, Michael O’Toole and Vere Shannon later that year. With enthusiastic country support and incremental funding from the Global Environment Facility (GEF), a comprehensive programme was developed over the next four years (see Anon 1999). Following the approval of the BCLME Project Document early in 2002, funds were released by the UNDP/GEF to enable the full implementation of the BCLME Programme, including reduction of uncertainty and improvement of predictability, and the Programme commenced in April 2002. Whereas the focus of BENEFIT is on science and science capacity building, the BCLME Programme is a broad-based multi-sectoral initiative aimed at sustainable and integrated management of the Benguela Current ecosystem as a whole, having as its developmental goal “to sustain the ecological integrity of the BCLME through integrated transboundary ecosystem management.”
6
V. Shannon
Key to the implementation of the BCLME Programme was the endorsement by the Governments of Angola, Namibia and South Africa in 1999/2000 of six main policy actions. These policy actions recognise the need to address both the impacts of anthropogenic factors and natural processes occurring in the ecosystem, including the highly variable nature of the environment of the BCLME. Fundamental to the above is knowledge and understanding of how the ecosystem, and components thereof, will respond to human actions (e.g. fishing, pollution, habitat alteration) and to natural environmental events and change (e.g. Benguela Niños, hypoxia/anoxia, harmful algal blooms i.e. HABs,). Important outputs of the GEF intervention are inter alia: • enhancement of sustainable management and utilisation of transboundary marine resources • assessment of environmental variability and its ecosystem impacts, and improvement of predictability for enhancing management of living marine resources • maintenance of ecosystem health and biodiversity and management of pollution to safeguard fisheries and other resources • increasing donor participation and co-financing throughout life of Programme and beyond Clearly there is a need to improve predictability of the natural and anthropogenic regimes, i.e. forecasting changes, major perturbations and ecosystem responses and impacts. OBSERVING AND PREDICTING IN THE BCLME WITHIN THE INTERNATIONAL CONTEXT There is considerable international interest in regional ocean forecasting as evident from the strategies of the International Association for the Physical Sciences of the Ocean (IAPSO) and the Global Ocean Observing System (GOOS) of the Intergovernmental Oceanographic Commission (IOC). In this respect it should be noted that the IAPSO strategy makes specific reference to promoting the creation of real-time forecasting strategies in developing countries, capitalising on expertise developed in the USA and Europe, and implemented through co-sponsorship of focussed workshops in targeted regions. Permanent, continuously operating real-time regional ocean prediction systems are increasingly required to support a variety of critical activities in the coastal environment, including navigation, fisheries and marine operations, response to oil and hazardous material spills, search and rescue, and prediction of harmful algal blooms and other ecosystem or water quality phenomena. The implementation of such systems in turn requires advanced technologies in sensors and observing systems, and numerical models and data assimilation, as well as the infrastructures necessary to jointly use them. Coastal ocean observation networks are now being constructed at numerous locations, and the USA and European networks can be prototypes for more extensive systems. Enabling technologies that make this possible include the rapid advances in sensor and platform technologies, multiple real-
A plan comes together
7
time communication systems for transmitting the data and the emergence of a universal method for the distribution of results via the World Wide Web. Future sensors that will expand observing capabilities include new ocean colour satellites, altimeters, HF radars and autonomous vehicles. Particularly important are the efforts of the Global Ocean Observing System to develop an observational network for the global ocean and that also meets the requirements for regional ocean observations and forecasting (c.f. GOOS-Africa, Abidjan Convention etc). Concurrently, hydrodynamic and ecological models for the regional systems have been developed and are beginning to show considerable skills. The crucial step allowing for real-time regional forecasting is the development, started in the late 80s, of oceanographic data assimilation, which is now becoming a reality. International thinking as reflected above is very much in keeping with the strategy and workplan of the BCLME Programme, where a key policy action is the assessment of environmental variability, ecosystem impacts and the improvement of predictability. Two cornerstones are the development of an environmental early warning system and the improvement of predictability of extreme events in the BCLME. To give effect to this, an Environmental Variability Advisory Group (EVAG) and associated Activity Centre (EVAC) were established in 2002, the terms of reference for the requisite projects were developed, contracts were awarded, and work on these is now progressing well. This builds on, and is being integrated with, ongoing modelling activities in the Benguela and comparable systems, which are being undertaken in partnership with overseas scientists and institutions – for example the French Institute for Research and Development (IRD). FAST-TRACKING THE DEVELOPMENT OF A REGIONAL OBSERVING SYSTEM AND PREDICTIVE CAPABILITY At its Strategy Task Group meeting in November 2002, IAPSO identified the Benguela region as a promising candidate site for development of a real-time forecasting capability. In view of the of the long history of ocean science in southern Africa, the coming into being of the BCLME Programme as well as international interest and regional needs, IAPSO decided that the most beneficial approach would be to host an international workshop in the region in partnership with other organisations, and involving specialists in the field of predicting and data assimilation. Accordingly, the concept was explored further in consultation with various regional and other international bodies and in November 2004 the International Workshop on Forecasting and Data Assimilation in the Benguela and Comparable Systems was held in Cape Town. Sponsored by the BCLME Programme, IAPSO and IOC-GOOS, together with seven other international, regional and national organisations, the Workshop addressed a key policy action of the BCLME, viz. the assessment of environmental variability, ecosystem impacts and the improvement of predictability. The Workshop was planned, developed and structured by an international Scientific Programme Committee and implemented by a regional Local Organising Committee. Participation in the four-day meeting was strictly by invitation, and over 100 leading international and regional experts, including Kenneth Sherman, the mastermind behind the global
8
V. Shannon
LME initiative and Gotthilf Hempel, the “Grandfather” of both BENEFIT and the BCLME Programmes attended the Workshop. An overarching objective of the Workshop was to contribute to BCLME management by improving assessment of variability and developing an effective and affordable forecasting capability for the region. In order to address this the Workshop had to assess what was known about variability in the BCLME and ascertain which aspects are amenable to forecasting of value, to review present status and recent advances in forecasting and data assimilation in the BCLME, to review advances in forecasting in comparable ecosystems (e.g. Humboldt, Canary, California), to specify minimum data, modelling and human capacity requirements for an early warning system and a blueprint for implementation, to transfer expertise and technology from leading overseas individuals and institutions to the BCLME region and promote collaboration, partnerships and networking, and to help improve numerical literacy skills of regional marine scientists and decision makers and generally build human capacity. The Workshop addressed a broad range of subjects (ocean and atmosphere physics, chemistry and biology – including ecosystem and resource dynamics) of importance for the development of a predictive capability for the greater Benguela Current region and comparable systems. Topics of relevance to forecasting on time scales ranging from hours to months, and even years and decades, were inter alia: • wind forcing on various scales • variability of the Angola (Cabinda) Front, Angola Current and Angola Dome • variability of the Angola-Benguela Front, Benguela Niños and other events of tropical origin • alongshore and cross-shelf process associated with the principal upwelling cell (Lüderitz cell) • intrusions of the Agulhas Current and Sub-Antarctic water into the Benguela • wind and waves and impacts on marine structures and maritime operations • hypoxia/anoxia and its impacts on fish stocks • harmful algal blooms and their predictability • advection and dispersal of pollutants, sulphur “eruptions” • modelling food chain dynamics, including regime shifts • environmental impacts on fish resources and environmental constraints on the distribution of fish vis-à-vis modelling and forecasting • impacts of environmental variability on the ecosystem on inter-annual and decadal time scales • developments in observing and forecasting in comparable systems (Humboldt Current, Canary Current etc) • models and data requirements • ocean observing system appropriate for the BCLME region and key elements of an early warning system for the BCLME. In order to make optimal use of time and the available knowledge and expertise, a programme was devised (somewhat similar to the successful “Dahlem” model)
A plan comes together
9
whereby definitive overviews of the state of knowledge and understanding of the BCLME variability were presented in plenary on the first day, followed by eight parallel specialist discussion, review and planning sessions on the remaining three days during which regular report-backs were made in plenary in order to inform and promote integration. Full particulars about the Workshop, its sponsors, the scientific programme, the overviews, specialist sessions and the organisers and the participants are provided in the accompanying CD-ROM to this book. ABOUT THIS BOOK This book draws on material presented at the Benguela Forecast Workshop and the specialist discussions, but it is not the proceedings per se. The idea for the book came from a meeting in 2003 among representatives of the BCLME Programme, GOOSAfrica and IAPSO, and was seen as a combination of an overview of the state of knowledge of the variability in the Benguela and the collective wisdom of experts about the predictability of the system. This concept was explored further with other interested international and regional organisations and the “father” of the global LMEs, Kenneth Sherman, for possible inclusion in the ongoing LME series published by Elsevier. The book is in four parts. Part I introduces the topic of prediction within the context of the international Large Marine Ecosystem initiative and GOOS-Africa. Part II sets the scene through a suite of five definitive overviews of aspects of Benguela variability and an overview of variability and change in comparable ecosystems such as the Guinea Current, Humboldt and California systems. Part III – titled “Hopes, Dreams and Reality” gets to the heart of the subject of forecasting and data assimilation, and captures the collective thinking on Benguela predictability. Part IV gives pointers to what is needed, what is possible and what should be done, and concludes with a vision in the form of an essay on modelling and forecasting in the BCLME and other eastern boundary upwelling systems. Complementary material, including model outputs, animations illustrating variability, graphical displays as well as some selected contributed papers, is included in the CD-ROM which accompanies this book. The editors and authors hope that the reader will find the book both a useful reference work and a source of inspiration as to how best to address the complex issue of predicting in the coastal oceans of developing countries and regions. We also hope that the book will provide a blueprint for managers and scientists for converting hopes and dreams into action in the BCLME and elsewhere. If it does, then the plan whose foundation was laid in Swakopmund, Namibia in 1995 and which has gathered momentum over the past decade through regional collaboration and international support, really will have “come together.”
10
V. Shannon
ACKNOWLEDGEMENT The introductory paragraphs of this chapter are based on previously published material which was prepared for the UNDP/GEF during the development of the BCLME Programme, e.g. Anon (1999), and which is now in the public domain. REFERENCES Anon. 1999. Benguela Current Large Marine Ecosystem Programme: Transboundary Diagnostic Analysis. UNDP/GEF/UNOPS. 51p. Colberg, F., C. J. C. Reason and K. Rogers. 2004. South Atlantic response to ENSO and induced climate variability in an OGCM. J. Geophys. Res. 100:15835-15847. Shannon, L. V. 1985. The Benguela Ecosystem Part I. Evolution of the Benguela, physical features and processes. In Barnes, M., ed. Oceanogr. Mar. Biol. Ann. Rev. 23:105-182.
Large Marine Ecosystems, Vol. 14 V. Shannon, G. Hempel, P. Malanotte-Rizzoli, C. Moloney and J. Woods (Editors) © 2006 Elsevier B.V. All rights reserved.
11
2 Forecasting Within the Context of Large Marine Ecosystems Programs Kenneth Sherman LME DEFINITION: DELINEATION AND MAJOR STRESSORS Large marine ecosystems are natural regions of ocean space encompassing coastal waters from estuaries to the seaward boundary of continental shelves and the outer margins of coastal currents. They are relatively large regions of 200,000 km2 or greater, the natural boundaries of which are based on four ecological criteria: bathymetry, hydrography, productivity, and trophically related populations. The concept that critical processes controlling the structure and function of biological communities can best be addressed on a regional basis (Ricklefs 1987) has been applied to the ocean by using large marine ecosystems as the distinct units for marine resources assessment, monitoring, and management. In turn, the concept of assessment, monitoring, and management of renewable resources from an LME perspective has been the topic of a series of ongoing national and international studies, symposia case studies and workshops initiated since 1984; in each instance, the geographic extent of the LME has been defined on the basis of bathymetry, hydrography, productivity, and trophodynamics. A list of peer reviewed published volumes of LME case studies is given in Table 2-1. The marine areas of the world most stressed from habitat degradation, pollution, and overexploitation of resources are the coastal ecosystems. Ninety percent of the usable annual global biomass yield of fish and other living marine resources is produced in 64 LMEs (Figure 2-1) identified within, and in some cases extending beyond, the boundaries of the EEZs of coastal states (Sherman 1994; Garibaldi and Limongelli 2003). Levels of primary production are persistently higher around the margins of the ocean basins than in the open-ocean pelagic areas (Figure 2-2). High population density characterizes these coastal ocean areas and contributes to the pollution that has its greatest impact on natural productivity cycles through eutrophication from high levels of nitrogen and phosphorus effluent from estuaries or air-born sources. The presence of toxins, harmful algal blooms, and loss of wetland nursery areas to coastal development are ecosystem-level problems that also need to be addressed.
12
K. Sherman
Table 2-1. Published Studies and volumes on LMEs
Author(s)
LME
7
Okemwa
East China Sea
5
Dwividi
Yellow Sea
7
Hazizi
East Bering Sea
1
West Greenland Shelf
8 3 5 10 2 4 5
Incze & Schumacher Livingston et al. Hovgård & Buch Blindheim & Skjoldal Rice Skjoldal & Rey Borisov Skjoldal
Kuroshio Current Sea of Japan
10
Dalpadado et al.
12 3 5
Matishov Ellertsen et al. Blindheim & Skjoldal Daan Reid McGlade Hempel
LME Somali Coastal Current Bay of Bengal
Barents Sea
Norwegian Shelf
Vol.
North Sea
1 9 10 12
Iceland Shelf
10
Faroe Plateau
10
Astthorsson, Vilhjálmsson Gaard et al.
1 3 5
Scully et al. Hempel Scully et al.
1 4 5
MacCall Mullin Bottom
Antarctic
California Current
Pacific American Coastal Humboldt Current Gulf of Thailand South China Sea Indonesian Sea Northeast Australian Shelf
12 8
Lluch-Belda et al. Bakun et al.
5 12 5 11
Bernal Wolff et al. Piyakarnchana Pauly & Chuenpagdee Christensen Zijlstra, Baars Bradbury & Mundy
5 3 2 5 8, 12
Kelleher Brodie
Oyashio Current Okhotsk Sea Gulf of Mexico
Southeast U.S. Shelf Northeast U.S. Shelf
Scotian Shelf Caribbean Sea Patagonian Shelf South Brazil Shelf East Brazil Shelf North Brazil Shelf Baltic Sea Celtic-Biscay Shelf Iberian Coastal Mediterranean Sea Canary Current Guinea Current
Benguela Current Black Sea
Vol. 8
Author(s) Chen & Shen
2, 5, 12 2
Tang Terazaki
8
Terazaki
2 5 2
Minoda Kusnetsov et al. Richards & McGowan
4 9 9 4
Brown et al. Shipp Gracia & Vasquez Baden Yoder
1
Sissenwine
4 6 10, 12 8 3 5 12
Falkowski Anthony Sherman
12
Ekau & Knoppers
12
Ekau & Knoppers
1 12 10
Kullenberg Jansson Lavin
2 10 5
Perez-Gandaras Wyatt & Porteiro Caddy
5 12
Bas Roy & Cury
5 11 11 11
Binet & Marchal Koranteng & McGlade Mensah & Quaatey Lovell & McGlade
11 11 2
Cury & Roy Koranteng Crawford et al.
12 5 12
Shannon & O’Toole Caddy Daskalov
Zwanenburg et al. Richards & Bohnsack Bakun Ekau & Knoppers
Forecasting within the LME programs context
Volume No. 1 2
3 4 5 6 7 8 9 10 11 12
13
Volume description 1986. Variability and Management of Large Marine Ecosystems. Sherman and Alexander, eds. AAAS Symposium 99. Westview Press, Boulder, CO. 319p 1989. Biomass Yields and Geography of Large Marine Ecosystems. Sherman and Alexander, eds. AAAS Symposium 111. Westview Press, Boulder, CO. 493p 1990. Large Marine Ecosystems: Patterns, Processes, and Yields. Sherman, Alexander and Gold, eds. AAAS Symposium. AAAS, Washington, DC. 242p 1991. Food Chains, Yields, Models, and Management of Large Marine Ecosystems. Sherman, Alexander and Gold, eds. AAAS Symposium. Westview Press, Boulder, CO..320p 1992. Large Marine Ecosystems: Stress, Mitigation and Sustainability. Sherman, Alexander and Gold, eds. AAAS Press, Washington, DC. 376 p. 1996. The Northeast Shelf Ecosystem: Assessment, Sustainability and Management. Sherman, Jaworski and Smayda, eds. Blackwell Science, Cambridge, MA. 564p 1998. Large Marine Ecosystems of the Indian Ocean: Assessment, Sustainability and Management. Sherman, Okemwa and Ntiba, eds. Blackwell Science, Malden, MA. 394p 1999. Large Marie Ecosystems of the Pacific Rim: Assessment, Sustainability and Management. Sherman and Tang, eds. Blackwell Science, Malden, MA. 455p 1999. The Gulf of Mexico Large Marine Ecosystem: Assessment, Sustainability and Management. Kumpf, Steidinger and Sherman, eds. Blackwell Science, Malden, MA. 736p 2002. Large Marine Ecosystems of the North Atlantic: Changing States and Sustainability. Skjoldal and Sherman, eds. Elsevier Science, New York. and Amsterdam.449p 2002. Gulf of Guinea Large Marine Ecosystem: Environmental Forcing and Sustainable Development of Marine Resources. McGlade, Cury, Koranteng, Hardman-Mountford, eds. Elsevier Science, Amsterdam and New York. 392p 2003. Large Marine Ecosystems of the World: Trends in Exploitation, Protection and Research. Hempel and Sherman, eds. Elsevier Science, New York and Amsterdam. 423p
Figure 2-1. Map showing 64 large marine ecosystems and linked watersheds
14
K. Sherman
Figure 2-2. Global map of average primary productivity and the boundaries of the 64 Large Marine Ecosystems (LMEs) of the world, available at www.lme.noaa.gov. The annual productivity estimates are based on SeaWiFS satellite data collected between September 1998 and August 1999, and the model developed by M. Behrenfeld and P.G. Falkowski (Limnol.Oceangr. 42(1): 1997, 1-20). The colorenhanced image provided by Rutgers University) depicts a shaded gradient of primary productivity from a high of 450 gCm2yr-1 in red to less than 45 gCm2yr-1 in purple.
Efforts are underway to meet the challenges of forecasting changing biotic and abiotic conditions within the boundaries of LMEs (USEO 2004; USOAP 2004; UN General Assembly 2001). Given the multi-sectoral and multi-disciplinary demand for timeseries data, consideration should be given to the use of standard and inter-calibrated protocols for measuring changing ecological states of the watersheds, bays, estuaries, and coastal water of LMEs. Long-term historical time series data on living marine resources (some up to 40-yr), coupled with measured or inferred long-term pollutant loading histories, have proven useful for relating the results of intensive monitoring to the quantification of ‘cause and effect’ mechanisms affecting the changing ecological states of LMEs. Temporal and spatial scales influencing biological production and changing ecological states in marine ecosystems have been the topic of a number of theoretical and empirical studies. The selection of scale in any study is related to the processes under investigation. An excellent treatment of this topic can be found in Steele (1988) (see also Denman et al. 1989). Steele indicates that in relation to general ecology of the sea, the best known work in marine population dynamics includes studies by Schaefer (1954), and Beverton and Holt (1957), following the earlier pioneering approach of Lindemann (1942). However, as noted by Steele (1988), this array of models is unsuitable for consideration of temporal or spatial variability in the ocean. A heuristic projection was produced by Steele (1988) to illustrate scales and ecosystem indicators of importance in monitoring pelagic components of the ecosystem including phytoplankton, zooplankton, fish, frontal processes, and short-term but large-area
Forecasting within the LME programs context
15
episodic effects (Figure 2-3). Advances in technology allow for cost effective methods for measuring the changing states of LMEs using suites of indicators including those depicted in Figure 2-3, supplemented with other modular suites of indicators.
Figure 2-3. A simple set of scale relations for the pelagic food web. (P) Phytoplankton, (Z) zooplankton, (F) fish, (MM) marine mammals, (B) birds. Two physical processes are indicated by (X) Predictable fronts with small cross-front dimensions, and (Y) weather events occurring over relatively large scales. (Adapted from Steele 1988)
LME INDICATOR MODULES A five-module indicator approach to the assessment and management of LMEs has been proven to be useful in ecosystem-based projects in the United States and elsewhere. The modules provide time-series data to support forecasting efforts. They are customized to fit the situation within the context of a transboundary diagnostic analysis (TDA) process and a strategic action plan (SAP) development process for the groups of nations or states sharing an LME. These processes are critical for integrating science into management in a practical way and establishing appropriate governance regimes. The five modules consist of 3 that are science-based indicators focused on: productivity, fish/fisheries, pollution/ecosystem health; the other two, socio-economics and governance, are focused on economic benefits to be derived from a more sustainable resource base and implementing governance mechanisms for providing stakeholders and stewardship interests with legal and administrative support for ecosystem-based management practices. The first four modules support the TDA process while the governance module is associated with periodic updating of the Strategic Action Plan. Adaptive management regimes are encouraged through periodic assessment processes (TDA updates) and updating of SAPs as gaps are filled (Wang 2004; Duda and Sherman 2002).
16
K. Sherman
Productivity Module Indicators Primary productivity can be related to the carrying capacity of an ecosystem for supporting fish resources (Pauly and Christensen 1995). Measurements of ecosystem productivity can be useful indicators of the growing problem of coastal eutrophication. In several LMEs, excessive nutrient loadings of coastal waters have been related to algal blooms implicated in mass mortalities of living resources, emergence of pathogens (e.g., cholera, vibrios, red tides, and paralytic shellfish toxins), and explosive growth of non-indigenous species (Epstein 1993). The ecosystem parameters measured and used as indicators of changing conditions in the productivity module are hydrography, nutrients, primary production, zooplankton biomass and species composition (Edwards et al. 2000a, 2000b). Plankton inhabiting LMEs have been measured over decadal time scales by deploying continuous plankton recorder systems monthly across ecosystems from commercial vessels of opportunity as well as from fixed stations. Advanced plankton recorders can be fitted with sensors for temperature, salinity, chlorophyll, nitrate/nitrite, petroleum, hydrocarbons, light, bioluminescence, and primary productivity, providing the means for in situ monitoring and for calibrating satellite-derived oceanographic data. Properly calibrated satellite data can provide information on such ecosystem aspects as physical state (i.e. surface temperature), nutrient characteristics, primary productivity and chlorophyll concentration (Berman and Sherman 2001; Aiken et al. 1999). Fish and Fisheries Module Indicators Changes in biodiversity and species dominance within fish communities of LMEs have resulted from excessive exploitation, naturally occurring environmental shifts due to climate change and coastal pollution. Changes in biodiversity and species dominance in a fish community can rise up the food web to apex predators and cascade down the food web to plankton components of the ecosystem (Frank et al. 2005; Choi et al. 2004; Pauly and Christensen 1995). The Fish and Fisheries Module includes both fisheries-independent bottom-trawl surveys and pelagic-species acoustic surveys to obtain time-series information on changes in fish biodiversity, population dynamics, and abundance levels. Standardized sampling procedures, when employed from small calibrated trawlers, can provide important information on changes in fish populations (NOAA 1993; NEFSC 1999, 2002) Sherman et al. 2002, 2003). Commercial fish catch provides biological samples for stock identification, stomach content analyses, age-growth relationships, fecundity, as well as data for preparing stock assessments and for clarifying and quantifying multispecies trophic relationships and pathological conditions. The survey vessels can also be used as platforms for obtaining water, sediment, and benthic samples for monitoring harmful algal blooms, diseases, anoxia, and changes in benthic communities.
Forecasting within the LME programs context
17
Pollution and Ecosystem Health Module Indicators In several LMEs, pollution and eutrophication have been important driving forces of change in biomass yields. Assessing the changing status of pollution and health of an entire LME is scientifically challenging. Ecosystem health is a concept of wide interest for which a single precise scientific definition is difficult. The health paradigm is based on multiple-state comparisons of ecosystem resilience and stability, and is an evolving concept that has been the subject of a number of meetings (NOAA 1993). To be healthy and sustainable, an ecosystem must maintain its metabolic activity level and its internal structure and organization, and must resist external stress over time and space scales relevant to the ecosystem (Costanza 1992). The Pollution and Ecosystem Health Module measures pollution effects on the ecosystem through the bivalve monitoring strategy of the U.S. Environmental Protection Agency’s (EPA’s) Mussel-Watch Program, through the pathobiological examination of fish; through the estuarine and nearshore monitoring of contaminants and contaminant effects in the water column, the substrate, and in selected groups of organisms through similar efforts. Where possible, bioaccumulation and trophic transfer of contaminants are assessed, and critical life history stages and selected food web organisms are examined for indicators of exposure to, and effects from, contaminants. Effects of impaired reproductive capacity, organ disease, and impaired growth from contaminants are measured. Assessments are made of contaminant impacts at both species and population levels. Implementation of protocols to assess the frequency and effect of harmful algal blooms, emergent diseases, and multiple marine ecological disturbances (Sherman 2000) are included in the pollution module. In the United States, the EPA has developed a suite of 5 coastal condition indicators: water quality index, sediment quality index, benthic index, coastal habitat index, and fish tissue contaminants index. The 2004 report, “National Coastal condition Report II,” includes results from EPA’s analyses of coastal condition indicators and NOAA’s fish stock assessments by LMEs aligned with EPA’s National Coastal Assessment (NCA) regions (USEPA 2001, 2004). Socioeconomic Module Indicators This module emphasizes the practical application of scientific findings to managing LMEs and the explicit integration of social and economic indicators and analyses with all other scientific assessments to assure that prospective management measures are cost-effective. Economists and policy analysts work closely with ecologists and other scientists to identify and evaluate management options that are both scientifically credible and economically practical with regard to the use of ecosystem goods and services. In order to respond adaptively to enhanced scientific information, socioeconomic considerations must be closely integrated with science. This component of the LME approach to marine resources management has recently been described as the human dimensions of LMEs. A framework has been developed by the Department of Natural Resource Economics at the University of Rhode Island for monitoring and assessment
18
K. Sherman
of the human dimensions of LMEs and for incorporating socioeconomic considerations into an adaptive management approach for LMEs (Sutinen et al. 2000). One of the more critical considerations, a method for economic valuations of LME goods and services, has been developed using framework matrices for ecological states and economic consequences of change (Hoagland et al. 2005). Governance Module Indicators The Governance Module is evolving, based on demonstration projects now underway in several ecosystems, such that ecosystems will be managed more holistically than in the past. In LME assessment and management projects supported by the Global Environment Facility for the Yellow Sea, the Guinea Current, and the Benguela Current LMEs, agreements have been reached among the environmental ministers of the countries bordering these LMEs to enter into joint resource assessment and management activities as part of building institutions. One of the major goals of the Benguela Current LME (BCLME) Programme is to establish a Benguela Current Commission which will enable Angola, Namibia and South Africa to engage constructively and peacefully in resolving the transboundary fisheries and environmental issues that threaten the integrity of the BCLME. A preliminary study has found that the establishment of a Benguela Current Commission (BCC) can be justified on several grounds. These include the need for an appropriate institution to implement an ecosystem-based management approach in the BCLME and the need to fulfill the international obligations and undertakings of the three countries of the Benguela. Other motives for the establishment of a regional commission include the need to develop a better understanding of the BCLME, to improve the management of human impacts on the BCLME, to facilitate regional capacity building and to increase the benefits derived from transboundary management and harvesting of fish stocks. A phased approach towards establishing a Benguela Current Commission has been recommended. The first priority would be to draft the necessary agreement between the three countries of the Benguela region. Thereafter, working groups and joint management committees could be brought into operation to address the most pressing transboundary concerns. An Interim Benguela Current Commission (IBCC) is seen as a preliminary step towards a permanent Commission. It would provide the three countries with an opportunity to test and strengthen the institutional structures that will be required for a permanent Commission. It is envisaged that the BCLME Programme’s existing structures would support the IBCC until new structures are made operational. Elsewhere, the Great Barrier Reef LME and the Antarctic LME are also being managed from an ecosystem perspective, the latter under the Commission for the Conservation of Antarctic Marine Living Resources. Governance profiles of LMEs are being explored to determine their utility in promoting long-term sustainability of ecosystem resources (Juda and Hennessey 2001). In each of the LMEs, governance jurisdiction can be scaled to ensure conformance with existing legislated mandates and authorities.
Forecasting within the LME programs context
19
APPLICATION OF INDICATOR MODULES TO LME MANAGEMENT SUPPORTED BY THE GLOBAL ENVIRONMENT FACILITY (GEF) Indicator data derived from spatial and temporal applications of the five modules is being applied by a growing number of nations in the assessment and management of LMEs with the financial assistance of the Global Environment Facility (GEF). Among the stressors affecting the sustainability of LMEs is the growing problem of coastal eutrophication. And the depletion of fish and fishery resources and biomass yields. Continued over-fishing in the face of scientific warnings, fishing down food webs, destruction of habitat, and accelerated nutrient loading, especially nitrogen export, have resulted in significant degradation to coastal and marine ecosystems of both rich and poor nations. Fragmentation among institutions, international agencies, and disciplines, lack of cooperation among nations sharing marine ecosystems, and weak national policies, legislation, and enforcement all contribute to the need for a new imperative for adopting ecosystem-based approaches to managing human activities in these systems in order to avoid serious social ad economic disruption. Following a three-year pilot phase (1991-1994), the Global Environment Facility (GEF) was formally launched to forge cooperation and finance actions in the con text of sustainable development—actions that address critical threats to the global environment from biodiversity loss, climate change, degradation of international waters, ozone depletion, and persistent organic pollutants. GEF-LME projects are implemented by UNDP, UNEP, and the World Bank. Expanded opportunities exist for participation by other agencies (GEF 1995, 2004). At present, 121 countries are in the planning and/or implementation phase of GEF/LME projects supported by $650 million in funding support for introducing an ecosystem-based approach to the assessment and management of marine resources and their environments (Table 2-2). SCIENCE-BASED ASSESSMENTS OF LME BIOMASS YIELDS The growing awareness that biomass yields are being influenced by multiple driving forces has broadened monitoring strategies to encompass food chain dynamics and the effects of environmental perturbations and pollution on living marine resources from an ecosystem perspective. To assist stewardship agencies in implementing ecosystembased assessment and management practices, Transboundary Diagnostic Analyses (TDAs) are being focused on the root causes of trends in LME biomass yields. In addition, information on principal driving forces of biomass yields from 29 LME case studies by marine resource experts has been analyzed. A list of the principal investigators, constituting the expert-systems analyses, appearing in 12 peer-reviewed and published LME volumes, is given in Table 2-1. The biomass yields in Table 2-3 are based on the mid-point value, 1995, of LME yields compiled by FAO for 19901999 (Garibaldi and Limongelli 2003). Biomass yield data for four LMEs not included in the FAO report were taken from published LME case studies. Based on expert systems analyses, principal and secondary driving forces were assigned to each LME using four categories (climate, fisheries, eutrophication, and inconclusive) and listed in descending order of catch level, as seen in Table 2-3.
20
K. Sherman
Table 2-2. 121* Countries Participating in GEF/Large Marine Ecosystem Projects
_______________________________________________________________ Approved GEF Projects
LME Gulf of Guinea (6)………………………… Yellow Sea (2)…………………………..... Patagonia Shelf/Maritime Front (2)………….. Baltic (9)………………………………………
Benguela Current (3)…………………………. South China Sea (7)…………………………… Black Sea (6)………………………………….. Mediterranean (19)……………………………. Red Sea (7)……………………………………. Western Pacific Warm Water Pool-SIDSa (13)…
Countries Benin, Cameroon, Côte d’Ivoire, Ghana, Nigeria, Togo China, Korea Argentina, Uruguay Denmark, Estonia, Finland, Germany, Latvia, Lithuania, Poland, Russia, Sweden Angola, Namibia, South Africa Cambodia, China*, Indonesia, Malaysia, Philippines, Thailand, Vietnam Bulgaria, Georgia, Romania, Russia*, Turkey, Ukraine Albania, Algeria, Bosnia-Herzegovina, Croatia, Egypt, France, Greece, Israel, Italy, Lebanon, Libya, Morocco, Slovenia, Spain, Syria, Tunisia, Turkey*, Yugoslavia, Portugal Djibouti, Egypt* Jordan, Saudi Arabia, Somalia, Sudan, Yemen Cook Islands, Micronesia, Fiji, Kiribati, Marshall Islands, Nauru, Niue, Papua New Guinea, Samoa, Solomon Islands, Tonga, Tuvalu, Vanuatu
Total number of countries: 70* GEF Projects in the Preparation Stage
Canary Current (7)............................................. Bay of Bengal (8)…………………………….. Humboldt Current (2)………………………… Guinea Current (16)…………………………..
Gulf of Mexico (3)……………………………. Agulhus/Somali Currents (8)…………………. Caribbean LME (23)…………………………..
Cape Verde, Gambia, Guinea, Guinea-Bissau, Mauritania, Morocco*, Senegal Bangladesh, India, Indonesia*, Malaysia*, Maldives, Myanmar, Sri Lanka, Thailand* Chile, Peru Angola*, Benin*, Cameroon*, Congo, Democratic Republic of the Congo, Côte d’Ivoire*, Gabon, Ghana*, Equatorial Guinea, Guinea*, Guinea-Bissau*, Liberia, Nigeria*, São Tomé and Principe, Sierra Leone, Togo* Cuba, Mexico, United States Comoros, Kenya, Madagascar, Mauritius, Mozambique, Seychelles, South Africa*, Tanzania Antigua and Barbuda, The Bahamas, Barbados, Belize, Colombia, Costa Rica, Cuba*, Grenada, Dominica, Dominican Republic, Guatemala, Haiti, Honduras, Jamaica, Mexico*, Nicaragua, Panama, Puerto Ricob, Saint Kitts and Nevis, Saint Lucia, Saint Vincent and the Grenadines, Trinidad and Tobago, Venezuela Total number of countries: 63*
*Adjusted for multiple listings Provisionally classified as Insular Pacific Provinces in the global hierarchy of LMEs and Pacific Biomes (Watson et al. 2003). b A self-governing commonwealth in union with the United States a
____________________________________________________________________
Forecasting within the LME programs context
21
Table 2-3. PRIMARY AND SECONDARY DRIVING FORCES OF LME BIOMASS YIELDS1 Based on published expert assessments in LME volumes listed in Table 1 LEVEL PRIMARY SECONDARY MMT
LME
Humboldt Current
climate
fishing
16.0
South China Sea
fishing
climate
10.0
East China Sea North Sea Eastern Bering Sea
fishing fishing -
climate climate -
3.8 3.5 2.1
Bay of Bengal
fishing
climate
2.0
Okhotsk Sea2
climate
fishing
2.0
Canary Current
climate
fishing
1.8
Norwegian Shelf
climate
fishing
1.5
Iceland Shelf
climate
fishing
1.3
Benguela Current
climate
fishing
1.2
Gulf of Thailand Mediterranean Sea of Japan3
fishing fishing climate
climate eutrophication fishing
1.1 1.1 1.0
Gulf of Mexico
fishing
climate
0.9
Guinea Current
climate
fishing
0.9
Baltic Sea
fishing
eutrophication
0.8
California Current
climate
fishing
0.7
U.S. Northeast Shelf
fishing
climate
0.7
Scotian Shelf
fishing
climate
0.7
Black Sea
eutrophication
fishing
0.5
EXPERT ASSESSMENTS Alheit and Bernal Wolff et al. Pauly and Christensen
5 12 5
Chen and Shen McGlade Schumacher et al. Dwividi Hazizi Kusnetsov et al. Roy and Cury Bas Ellertsen et al. Blindheim and Skjoldal Astthorsson and Vilhjálmsson Crawford et al. Shannon and O’Toole Pauly and Chuenpagdee Caddy Terazaki Richards and McGowan
8 10 12 5 7 5 12 5 3 5 10
Brown et al. Shipp Binet and Marchal Koranteng and McGlade Kullenberg Jansson MacCall Lluch-Belda et al. Sissenwine
Barents Sea
climate
fishing
0.5
Caribbean Sea Iberian Coastal Newfoundland-Labrador
fishing climate fishing
climate fishing climate
0.4 0.3 0.2
Yellow Sea4
fishing
climate
0.2
Great Barrier Reef
fishing
climate
0.1
West Greenland Shelf
climate
fishing
0.1
Faroe Plateau
climate
fishing
0.1
VOL REF
2 12 12 5 8 2 4 9 5 11 1 12 1 12 1
Murawski Sherman et al.
6 10
Zwanenburg et al. Zwanenburg Caddy Daskalov Skjoldal and Rey
10 12 5 12 2
Borisov Blindheim and Skjoldal Matishov et al.
4 5 12
Richards and Bohnsack Wyatt and Perez-Gandaras Rice et al. Tang Tang Brodie Hovgard and Buch Pederson and Rice Gaard et al.
3 2 10 2 12 12 3 10 10
22
K. Sherman
Notes to Table 2-3 1
Annual biomass yield levels based on 1990-1999 mid-decadal data (1995) from FAO 2003 Okhotsk Sea LME data from Kusnetsov et al. 1993 based on mid-decadal (1972) data on fishing yields from 1962 to 1982 1 Biomass yield data from Terazaki (1999) based on mid-decadal data (1985) from Sea of Japan 1980-1990. 1 Biomass yield data from Tang (2003) based on mid-decadal data for demersal species for the Yellow Sea, 1952 to 1992. 1
It would appear that the management regime for nearly half of this yield from the 29 case study LMEs (27.0 mmt) will need to focus efforts for forecasting biomass yields on the climate signal, whereas the management regime for slightly less of the biomass yield (24.8 mmt) will need to focus primarily on catch control and secondarily on the climate signal, to recover depleted fish stocks and achieve maximum sustainable yield levels. The influence of climate forcing in biomass yields for the California Current LME has been analyzed and illustrated by Lluch-Belda et al. (2003) for anchovy and sardine catches. They also provide a time-series indicator of regime shift (RIS) for sardine and anchovy populations off the coast of Japan, and for the California Current, Benguela Current and Humboldt Current LMEs (Figure 2-4).
Figure 2-4. Historic sardine (upper) and anchovy (middle) catches (1920-2000) from the California Current LME, and the Regime Indicator Series (RIS lower). Catch data were obtained from Schwarzlose et al. (1999). RIS is a composite series reflecting synchronous variability of sardine and anchovy populations of the Japan, California, Benguela and Humboldt currents. Modified from Lluch-Cota, D.B. et al. (1997). (Lluch-Belda et al. 2003)
Forecasting within the LME programs context
23
Evidence of climate forcing for the Humboldt Current LME has been given by Wolff et al. (2003) and for the Iceland Shelf LME by Astthorsson and Vilhjálmsson (2002). For the Benguela Current LME, the effects of climate forcing appear to be mediated through productivity, pollution-policy and fisheries interactions (Shannon et al. 2004). In contrast, the argument for urgent reduction in fishing effort is supported by the data in Sherman et al. (2003) for the US Northeast Shelf LME, and for the Gulf of Thailand based on the expert analysis of Pauly and Chuenpagdee (2003) (Figure 2-8). Of the 29 LME case studies, 13 were assigned to climate forcing as the principal driver of change in biomass yield, 14 were assigned to fisheries as principal driver, one was assigned to eutrophication, and the remaining one was deemed inconclusive. In all but the Mediterranean Sea LME, where climate forcing was the principal driver of changing biomass yield, fisheries forcing was the secondary driver. In the case of the Mediterranean Sea LME, the secondary driver was eutrophication (Caddy(1993). The contribution of the 29 LMEs to the annual global biomass yields amounts to 54.4 million metric tons (mmt) or 64% based on the average annual yield from 1995-1999 of 85 mmt (Garibaldi and Limongelli 2003). The observation that excessive fishing effort can alter the structure of the ecosystem, resulting in a shift from relatively highpriced, large-sized, long-lived demersal species, down the food chain toward lowervalued, smaller, short-lived pelagic species (Pauly and Christensen 1995), is supported by the LME data on species biomass yields.
Figure 2-5. A conceptual model of how climatic conditions in the Icelandic Shelf LME may affect production at lower trophic levels and eventually the yield from the Icelandic cod stock in Astthorsson and Vilhjálmsson (2002).
Evidence from the East China Sea, Yellow Sea, and Gulf of Thailand suggests that these three LMEs are approaching a critical state of change, wherein recovery to a
24
K. Sherman
previous ratio of demersal to pelagic species may become problematic. In all three cases, the fisheries are now being directed toward fish protein being provided by catches of smaller-sized species of low value (Chen and Shen 1999; Pauly and Chuenpagdee 2003; Tang 2003). The species change in biomass yields of the Yellow Sea, as shown in Tang (2003), represents an extreme case wherein the annual demersal species biomass yield was reduced from 200,000 mt in 1995 to less than 25,000 mt through 1980. The fisheries then targeted the pelagic anchovy and, between 1990 and 1995, landings of anchovy reached an historic high of 500,000 mt.
Figure 2-6. Major features of the Gulf of Thailand LME fisheries and trophic level. (A) Catches, by major species groups (excluding tuna and other large pelagics). Note stagnation and decline of demersal catches, following their rapid increase in the 1960s and 1970s. Also note increasing contribution of small and medium pelagics, and overall decline in the 1990s. (B) Trophic level (TL) trends in the catch of research trawlers (reflecting relative abundances in the ecosystems), and in the total landings (both series excluding large pelagics). Lower TL in 1977 to 1997 series is due to inclusion of small pelagics and other low-TL organisms caught by gears other than trawl. From Pauly and Chuenpagdee (2003).
RECOVERING FISHERIES BIOMASS The GEF/LME projects presently funded or in the pipeline for funding in Africa, Asia, Latin America and eastern Europe, represent a growing network of marine scientists, marine managers and ministerial leaders who are engaged in pursuing the ecosystem and fishery recovery goals. The significant annual global biomass yields of marine fisheries from ecosystems in the GEF-LME Network of almost half the world marine landings provides a firm basis for moving toward the 2004 Johannesburg World Summit on Sustainable Development (WSSD) goal for introducing an ecosystembased assessment and management approach to global fisheries by 2010 (Table 2-4).
Forecasting within the LME programs context
25
Table 2-4. LMEs With Relatively Stable Biomass Yields, 1990-1999
LME
LEVEL MMT
Arabian Sea
2.2
Bay of Bengal
2.3
Mediterranean
1.1
Indonesian Seas
1.6
Sulu-Celebes
0.8
North Brazil
0.1 TOTAL: 8.1
Percentage of Global Marine Biomass Yield: 9.5%
There is an international instrument supported by most coastal nations that could have immediate applicability to reaching Summit fishery goals. The FAO Code of Conduct for Responsible Fishery Practice of 2002 argues for moving forward with a “precautionary approach” to fisheries sustainability given a situation wherein available information can be used to recommend a more conservative approach to fish and fisheries total allowable catch levels (TAC) than has been the general practice over the past several decades (Holling 1973, 1986, 1993). Based on the decadal profile of LME biomass yields from 1990 to 1999 (Garibaldi and Limongelli 2003), it appears that the yields of total biomass and the biomass of 11 species groups of 6 LMEs have been relatively stable or have shown marginal increases over the decade (Sherman 2005). The yield of marine biomass for these 6 LMEs was 8.1 mmt, or 9.5 percent of the global marine fisheries yield in 1999 (Table 2-4). The countries bordering these six LMEs—Arabian Sea, Bay of Bengal, Indonesian Sea, Northeast Brazil Shelf, Mediterranean Sea and the Sulu-Celebes Sea—are among the world’s most populous, representing approximately one-quarter of the total human population. These LME border countries are increasingly dependent on marine fisheries for food security and for national and international trade. Given the risks of “fishing-down-the-food-web,” and in the absence of fishing-effort data, it would appear opportune for the stewardship agencies responsible for the fisheries of the bordering countries to consider options for mandating precautionary total allowable catch levels during a period prior to reductions in catch. Evidence for species recovery following a significant reduction in fishing effort through mandated actions is encouraging. Following management actions to reduce fishing effort, the robust condition of the U.S. Northeast Shelf ecosystem with regard to the average annual level of primary productivity (350 gCm2y), stable annual average levels of zooplankton (33 cc100m3), and a relatively stable oceanographic regime (Sherman et al. 2002), contributed to: (1) a relatively rapid recovery of depleted herring and mackerel stocks, with the cessation of foreign fisheries in the mid-1970s; and (2) initiation of the recovery of depleted yellowtail flounder and haddock stocks at the Georges Bank sub-area of the ecosystem following a mandated 1994 reduction in fishing effort (Figure 2-7).
26
K. Sherman
Figure 2-7. Increase in biomass of the Northeast Shelf LME, Georges Bank sub-area yellowtail flounder and haddock follows reduction in fishing effort (exploitation rate): http://www.nefsc.noaa.gov/nefsc/publications/crd/crd0216/
Three LMEs remain at high risk for fisheries biomass recovery expressed as a pre1960s ratio of demersal to pelagic species—the Gulf of Thailand, East China Sea and Yellow Sea. However, mitigation actions have been initiated by the People’s Republic of China toward recovery by mandating 60 to 90 day closures to fishing in the Yellow Sea and East China Sea during summer months (Tang 2003). The country-driven planning and implementation documents supporting the ecosystem approach to LME assessment and management practices can be found at www.iwlearn.org.
Forecasting within the LME programs context
27
NITROGEN OVER-ENRICHMENT OF LMEs Nitrogen over-enrichment has been reported as a coastal problem for two decades, from the southeast coast of the US as described by Duda (1982) twenty years ago, to the Baltic and other systems (Helsinki 2001). More recent estimates of nitrogen export to LMEs from linked freshwater basins are summarized in Jaworski 1999. These recent human-induced increases in nitrogen flux range from 4-8 in the US from the Gulf of Mexico to the New England coast while no increase was documented in areas with little agricultural or population sources in Canada (Howarth et al. 2000). In European LMEs, recent nitrogen flux increases of from 3-fold in Spain to 4-fold in the Baltic and 11-fold in the Rhine basin draining to the North Sea LME have been recorded (Howarth et al. 2000). Duda and El-Ashry (2000) described the origin of this disruption of the nitrogen cycle from the “Green Revolution” of the 1970s as the world community converted wetlands to agriculture, utilized more chemical inputs, and expanded irrigation to feed the world. As noted by Duda (1982) for the Southeast estuaries of the US, and by Rabalais (1999) for the Gulf of Mexico, much of the large increase in nitrogen export to LMEs is from agricultural inputs, both from the increased delivery of fertilizer nitrogen as wetlands were converted to agriculture and from concentrations of livestock as shown in Duda and Finan (1983) for eastern North Carolina, where the increase in nitrogen export over the forested situation ranged from 20- to 500-fold in the late 1970s. Industrialized livestock production in the last two decades increases the flux, the eutrophication, and the oxygen depletion even more as reported by the NRC (2000). The latest GESAMP Assessment (2001) also identified sewage as a significant contributor to the eutrophication from drainages from large cities; atmospheric deposition from automobiles/agricultural activities also contributes, depending on proximity to sources.
DIN Export by Rivers for World Regions 1990 and 2050 BAU Scenario 16 14
Tg N y-1
12 10 2050
8 6
1990
4 2 0
North South Africa Am erica Am erica
Europe
NE Eastern Asia Asia
Southern Asia
Figure 2-8. Model-predicted nitrogen [dissolved inorganic N in Tg (or 1012g) Ny-1] export by rivers to coastal systems in 1990 and in 2050 (based on a business-as-usual [BAU] scenario). Figure modified from Kroeze and Seitzinger (1998).
28
K. Sherman
The excessive levels of nitrogen contributing to coastal eutrophication constitute a new global environment problem that is cross-media in nature. Excessive nitrogen loadings have been identified as problems in the following LMEs that are receiving GEF assistance: Baltic Sea, Black Sea, Adriatic portion of the Mediterranean, Yellow Sea, South China Sea, Bay of Bengal, Gulf of Mexico, and Plata Maritime Front/Patagonia Shelf. In fact, preliminary global estimates of nitrogen export from freshwater basins to coastal waters were assembled by Seitzinger and Kroeze (1998) as part of a contribution to better understanding LMEs (2-8)). These preliminary estimates of global freshwater basin nitrogen export are alarming for the future sustainability of LMEs. Given the expected future increases in population and fertilizer use, LMEs may be, without significant N mitigation efforts, subjected to a future of increasing harmful algal bloom events, reduced fisheries, and hypoxia that further degrades marine biomass and biological diversity. LME MODELING AND DRIVING FORCES OF CHANGE Empirical and theoretical aspects of yield models for large marine ecosystems have been reviewed by several ecologists. According to Beddington (1986,1995; Daan (1986), Levin (1990) and Mangel (1991), published dynamic models of marine ecosystems offer little guidance on the detailed behavior of communities. However, these authors concur on the need for covering the common ground between observation and theory by implementing monitoring efforts on the large spatial and long temporal scales (decadal) of key components of the LMEs. The sequence for improving the understanding of the possible mechanisms underlying observed patterns in LMEs is described by Levin (1990) as: 1) examination of statistical analyses of observed distributional patterns of physical and biological variables, 2) construction of competing models of variability and patchiness based on statistical analyses and natural scales of variability of critical processes, 3) evaluation of competing models through experimental and theoretical studies of component systems, and 4) integration of validated component models to provide predictive models for population dynamics and redistribution. The approach suggested by Levin (1990) is consistent with the observation by Mangel (1991) that empirical support for the currently used models of LMEs is relatively weak, and that a new generation of models is needed that serves to enhance the linkage between theory and empirical results. Three models of ecosystem structure and function are being applied to LMEs with financial assistance from the GEF through one mid-sized proposal, “Promoting Ecosystem-based Approaches to Fisheries Conservation” (www.gefweb.org). (1) Estimates of carrying capacity using ECOPATH-ECOSIM food web approaches for the world’s 64 LMEs are being prepared in a collaboration between scientists of the University of British Columbia and marine specialists from developing countries. (2) A 24-month training project is being implemented by scientists from Rutgers University in collaboration with IOC/UNESCO to estimate expected nitrogen loadings for each LME over the next 50 years. (3) Scientists from Princeton University, under
Forecasting within the LME programs context
29
the direction of Simon Levin, and from the University of California at Berkeley, under the direction of Thomas “Zack” Powell, are modeling particle spectra pattern formation in LMEs, on the assumption that they represent complex adaptive systems in which ecological systems interact with socioeconomic ones. The goal is to describe and understand patterns of particle sizes emerging from multiple activation and inhibition processes that operate on diverse scales of space, time and organization. This work is informed by earlier studies that suggest, for example, the emergence of smaller size spectra of organisms following ecosystem-wide perturbations within LMEs (Chave and Levin 2003; Levin 2003; Cavendar-Bares et al. 2001; Gin Guo and Cheong 1998, Sheldon and Parsons 1967; Duplisea and Castonguay 2006). Additionally, the American Fisheries Society and the World Council of World Fisheries Societies are collaborating to create an electronic network that will expedite information access and communication among marine specialists participating in GEFsupported LME projects. There is a growing awareness among marine scientists, geographers, economists, government representatives, and lawyers of the utility of a more holistic ecosystem approach to resource management (Byrne 1986; Christy 1986; Alexander 1989; Belsky 1989; Crawford et al. 1989; Morgan 1989; Prescott 1989). On a global scale, the loss of sustained biomass yields from LMEs from mismanagement and overexploitation has not been fully investigated, but is likely very large. Effective management strategies for LMEs will be contingent on identification of major driving forces causing large-scale changes in biomass yields. Management of species responding to strong environmental signals will be enhanced by improving the understanding of the physical factors forcing biological change, thereby enhancing forecasts of El Niño-type events. In other LMEs, where the prime driving force is overfishing, options can be explored for reductions of fishing effort and implementing adaptive management strategies (Collie 1991). Further, remedial actions are required to ensure that the nitrogen overloading and the pollution of the coastal zone of LMEs is reduced and does not become a principal driving force in any LME. Recent reports explore the application of ecosystem-based research and modeling that is focused on management outcomes (Browman and Stergiou 2004) and on macroecology (Belgrano 2004; Hoagland et al. 2005; Edwards 2005; Grigalunas et al. 2005). Considerable effort has been focused on studies of modeling to improve forecasts of changing conditions in LMEs. For the Benguela Current LME, Shannon et al (2004) provide an excellent volume of ecosystem modeling studies pertinent to the Benguela Current LME. Based on 24 year time-series of fisheries and environmental data, Shannon et al. (2004) in their chapter on dynamic modeling concluded that fishing effects on fish stocks of the Benguela Current LME were less important than physical forcing on many of the important groups of fishes studied. Cury et al. (2000) consider the effects of physical forcing as triggering a critical predator-prey response between small pelagic fish and their zooplankton food base as an important consequence of forcing that is affecting biomass yields of the small pelagic fishes of the Benguela LME. The subsequent chapters in the present volume provide new insights on the components of physical forcing that could serve to improve forecasts of biomass yields as well as sustainable levels of yields for the Benguela Current LME.
30
K. Sherman
REFERENCES Aiken, J., R. Pollard, R. Williams, G. Griffiths, I. Bellan. 1999. Measurements of the upper ocean structure using towed profiling systems. In Sherman, K. and Q. Tang, eds. Large Marine Ecosystems of the Pacific Rim: Assessment, Sustainability, and Management. Malden: Blackwell Science, Inc. 346-362. Alexander, L.M. 1989. Large marine ecosystems as global management units. In Sherman, K., L.M. Alexander eds. Biomass Yields and Geography of Large Marine Ecosystems. American Association for the Advancement of Science (AAAS) Selected Symp. 111. Westview Press, Inc., Boulder, Colorado. 339-344. Astthorsson, O.S. and H. Vilhjálmsson. 2002. Iceland Shelf large marine ecosystem. In Sherman, K. and H.R. Skjoldal, eds. Large Marine Ecosystems of the North Atlantic: Changing States and Sustainability. Elsevier Science. Netherlands, London, New York, Tokyo. 219-244. 449p. Beddington, J.R. 1986. Shifts in resource populations in large marine ecosystems. In Sherman, K. and L.M. Alexander, eds. Variability and Management of Large Marine Ecosystems. AAAS Selected Symposium 99. Westview Press. Boulder, Colorado. 9-18. 319p. Beddington, J.R. 1995. The primary requirements. Nature 374:213-214. Behrenfeld, M. and P.G. Falkowski. 1997. Photosynthetic rates derived from satellite-based chlorophyll concentration. Limnol. Oceangr. 42(1): 1-20. Belgrano, A. coord. 2004. Theme Section: Emergent properties of complex marine systems: A macroecological perspective. Marine Ecology Progress Series 273:227-302. Belsky, M.H. 1989. The ecosystem model mandate for a comprehensive United States ocean policy and Law of the Sea. San Diego L. Rev. 26(3): 417-495. Berman, M.S. and K. Sherman. 2001. A towed body sampler for monitoring marine ecosystems. Sea Technology 42(9): 48-52. Beverton, R.J.H. and S.J. Holt. 1957. On the dynamics of exploited fish populations. Fish. Invest. Minist. Agric. Fish Food. (G.B.) Ser.II 19:1-533. Browman, H.I. and K.I. Stergiou, coord. 2004. Theme Section: Perspectives on ecosystem-based approaches to the management of marine resources. Marine Ecology Progress Series 274:269-298. Byrne, J. 1986. Large marine ecosystems and the future of ocean studies. In Sherman, K., L.M. Alexander, eds. Variability and Management of Large Marine Ecosystems. AAAS Selected Symp. 99. Westview Press, Inc. Boulder, Colorado. 299-308. Caddy, J.F. 1993. Contrast between recent fishery trends and evidence for nutrient enrichment in two large marine ecosystems: The Mediterranean and the Black Seas. In Sherman, K., L.M. Alexander and B.D. Gold, eds. Large Marine Ecosystems: Stress, Mitigation and Sustainability. AAAS Press. Washington, D.C. 376p. 137-147. Cavender-Bares, K.K, A. Rinaldo and S.W. Chisholm. 2001. Microbial size spectra from natural and nutrient enriched ecosystems. Limnology and Oceanography 46(4): 778-789. Chave, J. and S.A. Levin. 2003. Scale and scaling in ecological and economic systems. In Dasgupta, P. and K.-G. Möler, eds. The Economics of Non-convex Ecosystems, Special Issue, Environmental & Resource Economics 26:527-557. Chen, Ya-Qu and Xin-Qiang Shen. 1999. Changes in the biomass of the East China Sea ecosystem. In Sherman, K. and Q. Tang, eds. Large Marine Ecosystems of the Pacific Rim: Assessment, Sustainability and Management. Blackwell Science. Malden, Massachusetts. 221-239. 465p. Choi, J.S., K.T. Frank, W.C. Leggett, and K. Drinkwater. 2004. Transition to an alternate state in a continental shelf ecosystem. Can. J. Fish. Aquat. Sci. 61:505-510. Christy, F.T. 1986. Can large marine ecosystems be managed for optimum yields? In Sherman, K., and L.M. Alexander, eds. Variability and Management of Large Marine Ecosystems. AAAS Selected Symp. 99. Westview Press, Inc., Boulder, Colorado. 319p. 263-267. Collie, J.S. 1991. Adaptive strategies for management of fisheries resources in large marine ecosystems. In Sherman, K., L.M. Alexander, B.D. Gold, eds. Food Chains, Yields, Models, and Management of Large Marine Ecosystems. Westview Press, Inc. Boulder, Colorado. 320p. 225-242.
Forecasting within the LME programs context
31
Costanza, R. 1992. Toward an operational definition of ecosystem health. In Costanza, R., B.G. Norton, B.D. Haskell, eds. Ecosystem Health: New Goals for Environmental Management. Island Press, Washington DC. 239-256. Crawford, R.J.M., L.V. Shannon, P.A. Shelton. 1989. Characteristics and management of the Benguela as a large marine ecosystem. In Sherman, K., L.M. Alexander, eds. Biomass Yields and Geography of Large Marine Ecosystems. AAAS Selected Symp. 111. Westview Press, Inc., Boulder, Colorado. 493p. 169-219. Cury, P., A. Bakun, R. Crawford, A. Teichmann, R. Quiñones, 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. Daan, N. 1986. Results of recent time-series observations for monitoring trends in large marine ecosystems with a focus on the North Sea. In Sherman, K., L.M. Alexander, eds. Variability and Management of Large Marine Ecosystems. AAAS Selected Symp. 99, Westview Press, Inc., Boulder, Colorado. 319p. 145-174. Denman, K.L., H.J. Freeland and D.L. Mackas. 1989. Comparisons of time scales for biomass transfer up the marine food web and coastal transport processes. In Beamish, R.J. and G.A. McFarlane, eds. Effects of ocean variability on recruitment and an evaluation of parameters used in stock assessment models. Can. Spec. Publ. Fish. Aquat. Sci. 108. Dept. of Fisheries and Oceans, Ottawa. 379p. Duda, A.M. 1982. Mulicipal point sources and agricultural non-point source contributions to coastal eutrophication. Water Resources Bulletin 18(3): 397-407. Duda, A.M. and M.T. El-Ashry. 2000. Addressing the global water and environmental crises through integrated approaches to the management of land, water, and ecological resources. Water International 25:115-126. Duda, A.M. and D.S. Finan. 1983. Influence of livestock on nonpoint source nutrient levels of streams. Transactions of American Society of Agricultural Engineers 26(6): 1710-1726. Duda, A. and K. Sherman. 2002. A new imperative for improving management of large marine ecosystems. Ocean & Coastal Management 45(2002):797-833. Duplisea, D.E. and M. Castonguay. 2006. Comparison and utility of different size-based metrics of fish communities for detecting fishery impacts. Can. J. Fish. Aquat. Sci. 63:810-820. Edwards, S. 2005. Ownership of multi-attribute fishery resources in large marine ecosystems. In: Hennessey, T. and J. Suitinen, eds. Sustaining Large Marine Ecosystems: The Human Dimension. Elsevier. 368p. 137-154. Edwards, C.A., T.M. Powell, and H.P. Batchelder. 2000a. The stability of an NPZ model subject to realistic levels of vertical mixing. J. Mar. Res. 58:37-60. Edwards, C.A., H.P. Batchelder and T.M. Powell. 2000b. Modeling microzooplankton and mesozooplankton dynamics within a coastal upwelling system. J. Plankton Res.22:1619-1648. Epstein, P.R. 1993. Algal blooms and public health. World Resource Review 5(2): 190-206. FAO (The U.N. Food and Agriculture Organization). 2000. The State of the World Fisheries, Aquaculture. Rome. FAO. 142p. FAO Code of Conduct for Responsible Fisheries. 2002. www.fao.org/FI/agreem/codecond/ficonde.asp Frank, K.T., B. Petrie, J.S. Choi, W.C. Leggett. 2005. Trophic cascades in a formerly cod-dominated ecosystem. Science 308:1621-1623. Garibaldi, L. and L. Limongelli. 2003. Trends in oceanic captures and clustering of large marine ecosystems: Two studies based on the FAO capture database, as reported to the FAO by official national sources. FAO Fisheries Technical paper 435. Food and Agriculture Organization of the United Nations. Rome. 71p. GEF. 2004. Promoting Ecosystem-based Approaches to Fisheries Conservation. http://www.GEFweb.org,/ mid-sized proposals, March 23, 2004. GEF (Global Environment Facility). 1995. GEF Operational Strategy. Washington, DC: Global Environment Facility. GESAMP (Group of Experts on the Scientific Aspects of Marine Pollution). 1990. The state of the marine environment. UNEP Regional Seas Reports and Studies No. 115. Nairobi. Gin, K.Y.H., J. Guo, H-F Cheong. 1998. A size-based ecosystem model for pelagic waters. Ecological Modelling 112;53-72. Grigalunas, T.A., J.J. Opaluch, J. Diamantides and D-S Woo. 2005. Eutrophication in the Northeast Shelf large marine ecosystem: Linking hydrodynamic and economic models for benefit estimation. In
32
K. Sherman
Hennessey T., and J. Sutinen, eds. Sustaining Large Marine Ecosystems: The Human Dimension. Elsevier, Netherlands. 368p. 229-248. Helsinki Commission. 2001. Environment of the Baltic Sea Area 1994-1998. Baltic Sea Environment Proceedings No. 82A, Helsinki. 23p. Hoagland, P., D. Jin, E. Thunberg, and S. Steinback. 2005. Economic activity associated with the Northeast Shelf large marine ecosystem: Application of an input-output approach. In Hennessey, T. and J.Sutinen, eds. Sustaining Large Marine Ecosystems: The Human Dimension. Elsevier, Netherlands. 371p. 159-181. Holling, C.S. 1993. Investing in research for sustainability. Ecol. Applic. 3:552-555. Holling, C.S. 1986. The resilience of terrestrial ecosystems local surprise and global change. In Clark, W.C., and R.E. Munn, eds. Sustainable Development of the Biosphere. Cambridge Univ. Press, London. 292-317. Holling, C.S. 1973. Resilience and Stability of Ecological Systems. Institute of Resource Ecology. Univ. of British Columbia, Vancouver. Howarth, R., D. Anderson, J. Cloern, C. Elfring, C. Hopkinson, B. Lapointe, T. Malone, N. Marcus, K. McGlathery, A. Sharpley, D. Walker. 2000. Nutrient Pollution of coastal rivers, bays, and seas. ESA Issues in Ecology 7:1-15. Jaworski, N.A. 1999. Comparison of nutrient loadings and fluxes into the US Northeast Shelf LME with the Gulf of Mexico and other LMEs. In Kumpf, H., K. Steidinger, K. Sherman, eds. The Gulf of Mexico Large Marine Ecosystem: Assessment, Sustainability, and Management. Blackwell Science. Malden, Massachusetts. 704p. 360-371. Juda, L. and T. Hennessey. 2001. Governance profiles and the management of the uses of large marine ecosystems. Ocean Development and International Law. 32:41-67. Kroeze, C. and S.P. Seitzinger. 1998. Nitrogen inputs to rivers estuaries and continental shelves and related nitrous oxide emissions in 1990 and 2050: a global model. Nutrient Cycling in Agroecosystems 52:195-212. Kuznetsov, V.V., V.P. Shuntov, L.A. Borets. 1993. Food chains, physical dynamics, perturbations, and biomass yields of the Sea of Okhotsk. In Sherman, K., L.M. Alexander, and B.D. Gold, eds. Large Marine Ecosystems: Stress, Mitigation, and Sustainability. AAAS Press. 376p. 69-78. Levin, S.A. 1990. Physical and biological scales, and modeling of predator-prey interactions in large marine ecosystems. In Sherman, K., L.M. Alexander, B.D. Gold, eds. Large Marine Ecosystems: Patterns, Processes, and Yields. AAAS Press, Washington, DC. 242p. 179-187. Levin, S.A. 1993. Approaches to forecasting biomass yields in large marine ecosystems. In Sherman, K., L.M. Alexander, B.D. Gold, eds. Large Marine Ecosystems: Stress, Mitigation and Sustainability. AAAS Press, Washington, DC. 376p.36-39. Levin, S.A. 2003. Complex adaptive systems: Exploring the known, the unknown and the unknowable. Bulletin of the American Mathematical 40:3-19. Lindemann, R.L. 1942. The trophic dynamic aspect of ecology. Ecology 23:399-418. Lluch-Belda, D., D.B. Lluch-Cota and S.E. Lluch-Cota. 2003. Figure 9, p.212, Chapter 9, Interannual variability impacts on the California Current large marine ecosystem. In Hempel, G. and K. Sherman, eds. Large Marine Ecosystems of the World: Trends in Exploitation, Protection and Research. Elsevier Science, Amsterdam, Netherlands. 423p. Mangel, M. 1991. Empirical and theoretical aspects of fisheries yield models for large marine ecosystems. In Sherman, K., L.M. Alexander and B.D. Gold, eds. Food Chains, Yields, Models, and Management of Large Marine Ecosystems. Westview Press, Boulder, Colorado. 243-261. Morgan, J.R. 1989. Large marine ecosystems in the Pacific Ocean. In Sherman, K., L.M. Alexander, eds. Biomass Yields and Geography of Large Marine Ecosystems. AAAS Selected Symp. 111. Westview Press, Inc. Boulder, Colorado. 493p. 377-394. NEFSC. 2002. Assessment of 20 Northeast Groundfish Stocks through 2001: A Report of the Groundfish Assessment Review Meeting (GARM), Northeast Fisheries Science Center, Woods Hole, Massachusetts, October 8-11, 2002. This report is available at http://www.nefsc.noaa.gov/nefsc/publications/crd/crd0216/ NEFSC. 1999. Atlantic Herring. In: Report of the 27th Northeast Regional Stock Assessment Workshop (27th SAAW). Stock Assessment Review Committee (SARC) Consensus Summary of Assessments. Woods Hole Laboratory Reverence Document No. 98-15. NOAA (National Oceanic and Atmospheric Administration). 1993. Emerging theoretical basis for monitoring the changing states (Health) of large marine ecosystems. Summary report of two
Forecasting within the LME programs context
33
workshops: 23 April 1992, National Marine Fisheries Service, Narragansett, Rhode Island, and 11-12 July 1992, Cornell University, Ithaca, New York. NOAA Technical Memorandum NMFS-F/NEC100. NRC (National Research Council). 2000. Clean Copastal Waters: Understanding and reducing the effects of nutrient pollution. National Academy Press, Washington, DC. Pauly, D., V. Christensen. 1995. Primary production required to sustain global fisheries. Nature 374:255257. Pauly, D. and R. Chuenpagdee. 2003. Development of fisheries in the Gulf of Thailand large marine ecosystem: Analysis of an Unplanned experiment. In Hempel, G. and K. Sherman, eds. Large Marine Ecosystems of the World: Trends in Exploitation, Protection and Research. Elsevier Science. Amsterdam, Netherlands. 423p. 337-354. Prescott, J.R.V. 1989. The political division of large marine ecosystems in the Atlantic Ocean and some associated seas. In Sherman, K. and L.M. Alexander, eds. Biomass Yields and Geography of Large Marine Ecosystems. AAAS Selected Symp. 111. Westview Press, Inc. Boulder, Colorado. 395-442. Rabalais, N.N., R.E. Turner, W.J. Wiseman Jr. 1999. Hypoxia in the Northern Gulf of Mexico: Linkages with the Mississippi River. In Kumpf, H., K. Steidinger, K. Sherman, eds. The Gulf of Mexico Large Marine Ecosystem: Assessment, Sustainability, and Management. Malden, MA. Blackwell Science, Inc. 704p. 297-322. Ricklefs, R.E. 1987. Community diversity: Relative roles of local and regional processes. Science 235(4785):161-171. Schaefer, M.B. 1954. Some aspects of the dynamics of populations important to the management of the commercial marine fisheries. Bull. Inter-Am. Trop. Tuna Comm. 1:27-56. Schwartzlose, R.A., J. Alheit, T. Baumgartner, R. Cloete, R.J.M. Crawford, W.J. Fletcher, Y. Green Ruiz, E. Hagen, T. Kawasaki, D. Lluch-Belda, S.E. Lluch-Cotta, A.D. MacCall, Y. Matsuura, M.O. Nevárez-Martínez, R.H. Parrish, C. Roy, R. Serra, K.V. Shust, N.M. Ward and J.Z. Zuzunaga. 1999. Worldwide large-scale fluctuations of sardine and anchovy populations. S. Afr. J. Mar. Sci. 21:289347. Seitzinger, S.P. and C. Kroeze. 1998. Global distribution of nitrous oxide production and N inputs to freshwater and coastal marine ecosystems. Global Biogeochemical Cycles 12:93-113. Shannon, L.J., K.L. Cochrane and S.C. Pillar, eds. 2004. Ecosystem approaches to fisheries in the southern Benguela. African Journal of Marine Science 26. Republic of South Africa Department of Evironmental Affairs and Tourism: Marine and Coastal Management. Creda Communications, Cape Town, South Africa. 328p. Shannon, L.J., V. Christensen and C.J. Walters. 2004. Modelling stock dynamics in the southern Benguela Ecosystem for the period 1978-2002. In Shannon, L.J., K.L. Cochrane and S.C. Pillar, eds. Ecosystem approaches to fisheries in the southern Benguela. African Journal of Marine Science 26. 328p. 179195. Sheldon, R.W. and T.R. Parsons. 1967. A continuous size-spectrum for particulate matter in the sea. J. Rish. Res. Bd. Canada 235:909-915. Sherman, B. 2000. Marine ecosystem health as an expression of morbidity, mortality, and disease events. Marine Pollution Bulletin 41(1-6): 232-54. Sherman, K. 2005. The Large Marine Ecosystem Approach for assessment and management of ocean coastal waters. In Hennessey, T. and J. Sutinen, eds. Sustaining Large Marine Ecosystems: The Human Dimension. Elsevier. Amsterdam, Netherlands. 368p. 3-16. Sherman, K. 1994. Sustainability, biomass yields, and health of coastal ecosystems: An ecological perspective. Marine Ecology Progress Series 112:277-301. Sherman, K., J. O’Reilly and J. Kane. 2003. Assessment and sustainability of the U.S. Northeast Shelf Ecosystem. In Hempel, G. and K. Sherman. Large Marine Ecosystems of the World: Trends in Exploitation, Protection, and Research. Elsevier B.V., Amsterdam, Netherlands. 423p. 93-120. Sherman, K., J. Kane, S. Murawski, W. Overholtz and A. Solow. 2002. The U.S. Northeast Shelf large marine ecosystem: Zooplankton trends in fish biomass recovery. In Sherman, K. and H.R. Skjoldal, eds. Large Marine Ecosystems of the North Atlantic: Changing States and Sustainability. 449p. 195215. Steele, J.H. 1988. Scale selection for biodynamic theories. In Rothschild, B.J., ed. Toward a Theory on Biological-physical Interactions in the World Ocean. NATO ASI Series C: Mathematical and Physical Sciences, Vol 239. Kluwer Academic Publishers, Dordrecht. 513-526.
34
K. Sherman
Sutinen, J. ed. 2000. A Framework for Monitoring and Assessing Socioeconomics and Governance of Large Marine Ecosystems. NOAA Technical Memorandum NMFS-NE-158. 32p. Tang, Q. 2003. Figure 10, p.137 from Chapter 6, The Yellow Sea and mitigation action. In Hempel, G. and K. Sherman, eds. Large Marine Ecosystems of the World: Trends in Exploitation, Protection and Research. Elsevier Science. Amsterdam, Netherlands. 423p. Terazaki, M. 1999. The Sea of Japan large marine ecosystem. In Sherman, K. and Q. Tang, eds. Large Marine Ecosystems of the Pacific Rim: Assessment, Sustainability, and Management. Blackwell Science. 465p. 199-220. USEPA. 2004. National Coastal Condition Report. EPA-620-R-03/002 Washington, DC. http://www.epa.gov/owow/oceans/nccr/2005/nccr2-factsheet.html. USEPA. 2001. National Coastal Condition Report. EPA-620/R-01/005 Washington, DC. 204p. USEO. 2004. Executive Order 121704. Committee on Ocean Policy. http://www.whitehouse.gov/news/releases/2004. USOAP. 2004. U.S. Ocean Action Plan, Office of the President of the United States. 17 December 2004. http://ocean.ceq.gov/actionplan.pdf United Nations General Assembly. 2001. Report on the work of the United Nations Open-ended Informal Consultative Process established by the General Assembly in its resolution 54/33 in order to facilitate the annual review by the Assembly of developments in ocean affairs at its second meeting. Report A/56/121, 22 June, New York. 62p. Wang, H. 2004. An evaluation of the modular approach to the assessment and management of large marine ecosystems. Ocean Development & International Law 35:267-286. Watson, R., D. Pauly, V. Christensen, R. Froese, A. Longhurst, T. Platt, S. Sathyendranath, K. Sherman, J. O’Reilly, and P. Celone. 2003. Mapping fisheries onto marine ecosystems for regional, oceanic and global integrations. In Hempel, G., and K. Sherman. Large Marine Ecosystems of the World: Trends in Exploitation, Protection, and Research. Elsevier. 423p. Wolff, M., C. Wosnitza-Mendo and J. Mendo. 2003. The Humboldt Current LME. In: Hempel, G. and K. Sherman, eds. Large Marine Ecosystems of the World: Trends in Exploitation, Protection and Research. Elsevier Science, Amsterdam, Netherlands. 423p. 279-309.
Large Marine Ecosystems, Vol. 14 V. Shannon, G. Hempel, P. Malanotte-Rizzoli, C. Moloney and J. Woods (Editors) © 2006 Elsevier B.V. All rights reserved.
35
3 The Global Ocean Observing System for Africa (GOOS-Africa): Monitoring and Predicting in Large Marine Ecosystems Justin Ahanhanzo
INTRODUCTION The United Nations Conference on Environment and Development (UNCED) 1992 called for the creation of a global system of ocean observations to enable effective and sustainable management and development of seas and oceans, and prediction of future change. It is worth noting from a historical perspective that the Second World Climate Conference in 1990 also called for the establishment of such a system to provide the oceanographic data needed for the Global Climate Observing System (GCOS), which was proposed shortly afterward. Consequently, in response to these needs, the Twenty Fifth Session of the Executive Council of the Intergovernmental Oceanographic Commission of UNESCO formally initiated the Global Ocean Observing System (GOOS) in 1992, to which later the World Meteorological Organisation (WMO), the United Nations Environment Programme (UNEP) and the International Council of Scientific Unions (ICSU) committed themselves. GOOS is: (i) a sustained, coordinated international system for gathering data about the oceans and seas of the Earth, (ii) a system for processing such data, with other relevant data from other domains, to enable the generation of beneficial analytical and prognostic environmental information services, and (iii) the research and development on which such services depend for their improvement. Ultimately GOOS provides scientific monitoring and predicting for the implementation of the United Nations Framework Convention on Climate and the United Nations Convention on Biodiversity. THE LARGE MARINE ECOSYSTEM (LME) CONCEPT AND STRATEGY The coastal areas and the margins of the Pacific, Indian and Atlantic Oceans have been delineated into 64 Large Marine Ecosystems (LMEs), (Sherman and Alexander 1986; Sherman 1994; Sherman 1999). The LME boundaries were determined on the basis of four ecological criteria: (i) bathymetry, (ii) hydrography, (iii) productivity, and (iv) trophically dependant populations. Assessments of changing states of LMEs provide science-based information for the management of LME goods and services (Duda and Sherman 2002). Monitoring, assessment and management of the LMEs require long term and sustained ocean observations, measurements, data collection, processing, analysis, interpretation, and provision of products and services that may serve as
36
J. Ahanhanzo
decision making tools for governments, managers, planners and stakeholders. Marine data and information are crucial for monitoring, predicting and managing the LMEs and forecasting environmental change. The Global Ocean Observing System is conceived to provide these requirements on behalf of LMEs and other user communities. On the African coasts, there are six major LMEs including the Canary Current LME, Guinea Current LME and the Benguela Current LME on the Atlantic coast and the Red Sea LME, Somali Current and Agulhas Current LMEs in the Indian Ocean (Table 3-1) and (Figure 3-1). Research and investigations are ongoing to explore the criteria for the possibility of establishing a Mascarene LME.
Table 3-1. African LMEs and participating countries AFRICAN LMEs
Participating Countries
Agulhas Current
Comoros, Madagascar, Mozambique, South Africa
Benguela Current
Angola, Namibia, South Africa
Canary Current
Cape Verde, Gambia, Guinea, Guinea Bissau, Mauritania, Morocco, Senegal
Guinea Current
Angola, Benin, Cameroon, Congo, Democratic Republic of Congo, Equatorial Guinea, Gabon, Ghana, Guinea, Guinea Bissau, Ivory Coast, Liberia, Nigeria, Sao Tome and Principe, Sierra Leone, Togo
Red Sea
Djibouti, Egypt, Jordan, Saudi Arabia, Somalia, Sudan, Yemen
Somali Current
Kenya, Tanzania, Seychelles, (Somali not yet involved due to political issues)
THE RISE OF THE GLOBAL OCEAN OBSERVING SYSTEM IN AFRICA (GOOS-AFRICA) Based on several decades of experience, studies and sound expertise in oceanography and marine sciences, African institutions, marine scientists and stakeholders came to the conclusion that there is a need for an integrated approach to operational oceanography in support of the activities in LMEs and for other stakeholders. The initial GOOS-Africa information document was presented to the African Forum at the GCLME Seminar and Workshop in Abidjan, Côte d’Ivoire in early 1998. A core network of African scientists was established in 1998 to take forward the development of the Global Ocean Observing System for Africa. Upon the request of the Government of Mozambique, in the framework of the Pan-African Conference on
GOOS-Africa
37
Figure 3-1. Chlorophyll a biomass in African LMEs determined with ENVISAT/MERIS in June 2003 (courtesy of Coast Watch/ACRI)
Sustainable Integrated Coastal Management (PACSICOM), and with the support of African countries and institutions, UNESCO and its IOC, the first GOOS-Africa workshop was organized in Maputo in Mozambique, 18-20 July 1998. This workshop entitled, “Data for Sustainable Integrated Coastal Management, Global Ocean Observing System for Sustainable Integrated Coastal Management in Africa,” laid the foundations for the rise of the Global Ocean Observing System in Africa (GOOSAfrica). The programme addresses the fact that Africa can be impacted by extreme events such as El Niño, and La Niña, which affect rainfall and crops, as well as by floods, drought and tropical cyclones that have various causes in the atmosphere and ocean. In addition, the recent Indian Ocean tsunami also affected African countries on the East coast of the continent. A multidisciplinary approach to collecting observations, needed to forecast and interpret such events, becomes imperative because of the linkages between ocean, earth and meteorological processes and climate change: GOOS-AFRICA recognises that what happens at the coasts is commonly a complex function of earth, ocean and atmosphere on regional and global scales. Ocean processes can affect economic values of African investments in: (i) offshore and coastal oil and gas (ii) shipping and trade; (iii) offshore and coastal mining; (iv) coastal and offshore fisheries; (v) integrated coastal zone management; (vi) monitoring and predicting in large marine ecosystems; (vii) seaside tourism; (viii) public safety/health and protection of properties; (ix) early warning systems. The GOOS-Africa mandate includes a provision of a common platform of coastal and ocean services for monitoring and predicting dynamics of the large marine ecosystems through (i) assessing, (ii) hindcasting, predicting and forecasting, and (iii) establishing early warning systems. These will provide information on potential floods, sea level rise, regime shifts and their impacts on ecosystems and on the people who depend on them.
38
J. Ahanhanzo
GOOS-AFRICA STRATEGIC PARTNERSHIPS The GOOS-Africa Network recognized that a grassroots approach, with national and regional leadership and ownership is the key for building the long term institutional and scientific capacity required to get the desired results. In addition, positive synergy with reliable partners, in particular the regional financing institutions and industries, is needed to help to develop financial models enabling businesses and Governments to share the financial burden of the Regional Ocean Observing and Forecasting Systems for Africa (ROOFS-Africa) that will be the mechanism for implementing GOOS in Africa. THE AFRICAN LMES ARE CORE AND VITAL STRATEGIC PARTNERS FOR GOOS-AFRICA. GOOS-Africa is conceived to provide long term, sustained and systematic observations
both from in situ collecting devices and satellite remote sensing, combined with modelling, data assimilation and forecasting to support the monitoring, prediction and management of the resources of African LMEs (Figure 3-1) and (Table 3-1). Strategic partnerships have been established with relevant ongoing African and overseas programmes and specialized institutions. These programmes and partners include all ongoing African LMEs, the fisheries programmes, the UNESCO crosscutting project on the Applications of Remote Sensing for Integrated Management of Ecosystems and Water Resources in Africa, the African Monsoon Multidisciplinary Analysis (AMMA), the Ocean Data and Information Network for Africa (ODINAFRICA), the European Commission funded project for the GOOS Regional Alliances Network Development, the African Centre of the Meteorological Applications for Development (ACMAD); African Association of Remote Sensing of the Environment, (AARSE); National Oceanographic Center of Southampton, UK Meteorological Office, the European Space Agency (ESA), the US- National Oceanic and Atmospheric Administration (NOAA), the French Institute of Research for Development (IRD); the United Nations Agencies including the United Nations Environment Programme (UNEP); United Nations Industrial Development Organization (UNIDO), United Nations Development Programme (UNDP), United Nations Office of Outer Space Affairs (UNOOSA), and other multilateral and bilateral partners. In May 2005, GOOS-AFRICA was invited to sit as an observer and partner in the Met Ocean Committee of the International Association of Oil and Gas (OGP), given the considerable interest and the potential for mutual benefit between the GOOS-AFRICA Networks and the OGP members. GOOS-AFRICA CONTRIBUTION TO INTEGRATED MONITORING AND PREDICTING OF LARGE MARINE ECOSYSTEMS Based on a multidisciplinary and integrated strategy of five complementary work packages or modules, GOOS-Africa’s observing and forecasting system (ROOFSAfrica) will provide the observations and forecasting underpinning for ecosystembased management and rational use and exploitation of marine resources towards
GOOS-Africa
39
sustainable development of marine environment. The modules are listed in Table 3-2. To ensure flexibility, these modules are carried out in a complementary way. Their value will be maximised when all are integrated. Each module comprises a rationale, key objectives, key tasks, and a list of outputs, results and deliverables.
Table 3-2. Modular approach and strategy of GOOS-Africa
Modules
Title
M1
The African network of in situ ocean observing and monitoring systems including sea level records for monitoring coastal zones and impacts of global change in Africa
M2 M3
Remote sensing and satellite applications to marine and coastal environment Modelling, hindcasts/forecasts and data assimilation based on in situ and satellite data Effective involvement of different stakeholders at different stages of project implementation, and development of end-user interactive communication and information delivery system Industry and business partnerships towards reinforcing a Regional Ocean Observing and Forecasting System for Africa (ROOFS-Africa)
M4 M5
Module 1: The African network of in situ ocean observing and monitoring systems including sea level records In-situ ocean measurements are key elements of any coastal ocean observation system for Africa. These measurements must be made for the long term so as to provide the underpinning for accurate understanding and forecasting of water levels and water quality that are essential for a wide variety of uses and users including port management, shipping, fisheries, tourism, offshore and coastal installations and coastal erosion. This basic information is essential for the warnings and mitigation of natural disasters and the management of marine resources and offshore operations, including fish stocks and marine pollution. Measuring sea level and other ocean parameters provide a vital component of oceanographic observation programmes needed for immediate operational requirements of ships, navigation, and storm surge forecasting, for long-term monitoring and prediction of global sea level changes due to climate variations. In particular, tidal information is needed for addressing the following: (i) coastal erosion; (ii) flooding; (iii) salt water intrusion; (iv) sea-level topographic map production with satellite calibration; (v) assessment of ecosystem health; (vi) marine navigation and transportation; (vii) oil exploration and exploitation activities; (viii) marine pollution and oil spill mitigation; (ix) early warning systems.
40
J. Ahanhanzo
Module 2: Remote sensing and satellite applications to marine and coastal environment Satellite data, obtained synoptically every day from a number of different sensors together with in situ data necessary to establish sensor dependent error statistics, constitute a vital component of ROOFS-Africa. In addition, these data are required to constrain, force and initialize ocean model systems. Sea surface temperature, ocean colour (and derivative water quality parameters), surface wind speed, rain rates and distributions, solar radiation, sea surface height and state are priority data sets that should be ingested and archived in an operational manner to serve and foster the ROOFS for Africa. Synthetic aperture radar (SAR) and aircraft remote sensing measurements are also required for events (e.g., HAB, oil spill) requiring a rapid response. Remote sensing, as important as it is, will not on its own provide adequate answers to marine and coastal problems (Figure 3-2). Bencal 41 0m
Measured in-water particle size distribution data from the FRS Africana, showing the difference in size of the two algal assemblages despite their similar chlorophyll a concentrations
Bencal 37 0m Diatoms Pseudonitzschia sp. & Thalassiosira sp. dominant Chl a = 5.4 mg m-3 Deff = 12 μm
Dinoflagellates Alexandrium catenella & Prorocentrum micans dominant Chl a = 6.5 mg m-3 Deff = 24 μm
Size and Chl a data from the inverse reflectance algorithm - despite the speckling at low bio-mass, the size product demonstrates that it is possible to distinguish between a large cell size dinoflagellate community and a smaller sized diatom community
Figure 3-2. Application of the multispectral reflectance algorithm to SeaWiFS data. The SeaWiFS overpass is from the 15th of October 2003, during the BenCal bio-optical cal/val cruise in the southern Benguela. An extraordinary bloom of the toxic dinoflagellate Alexandrium catenella was reported in the Lamberts and Elands Bay vicinity several days later. The precursive expression of this bloom can be seen in the effective diameter image, which shows the presence of a large cell assemblage off Elands Bay. Note the ability of the algorithm to differentiate between water types dominated by differently sized phytoplankton at approximately equal chlorophyll concentrations (From Dr. Stewart Bernard et al., Oceanography Department University of Cape Town, South Africa. This study was funded and carried out under the framework of the UNDP/GEF BCLME project).
GOOS-Africa
41
Module 3: Modelling and forecasting based on in situ and satellite data Numerical modelling and forecasting based on in situ and satellite data constitute the basis for many services that provide forecasts or assessments useful to decision-makers working in the public or private sectors in the marine and coastal environment. The modelling and forecasting systems serve for coastal protection and better management of coastal erosion; pollution; marine transportation; fisheries; tourism; and pristine ecosystems. Module 4: Effective involvement of different stakeholders at different stages of project implementation and development of an end-user interactive communication and information delivery system This is a crosscutting package with the others. The main objective is to build up an end-user interactive information delivery system derived from active stakeholder participation in all stages of applications development and dissemination (from data gathering, to the packaging of information in forms that facilitate decision-making that will improve the life of societies). The ACMAD Communication System is the quantum leap in closing the information gap, providing missing links in the chain of development: ACMAD has developed a communication system (Radio and Internet – RANET) that involves uploading information to the AFRISTAR satellite from ‘editors’ such as ACMAD, downlink via solar powered receivers to local solar powered FM stations for rebroadcasts in local languages, with local interpretation to holders of wind-up radios in rural villages. Module 5: Industry and business partnerships towards reinforcing the Regional Ocean Observing and Forecasting System for Africa (ROOFSAfrica) The drivers consist of enhancing the transfer of meteorological and oceanographic (met ocean) information from the data providers to the users through establishing and fostering working partnerships between governments and industry. These partnerships will roadmap the role of met ocean information in the industry and business decision process and will establish the economic value of the information. There are four development and philosophical drivers for Industry and Business Partnerships: (i) The New Partnership for Africa’s Development (NEPAD) goals as exemplified in their existing projects; (ii) the WSSD goals as outlined in Africa (Johannesburg, South Africa, 2002); (iii) the Millennium Development Goals of the United Nations, particularly for the alleviation of poverty; and (iv) the Global Business Governance goals on sustainability as outlined in the “Triple Bottom Line” principles. Aligning industry and business partnerships with each of these goals can ensure the fulfilment of the sustainability mission.
42
J. Ahanhanzo
CONCLUDING REMARKS: SUCCESS STORIES
GOOS-Africa as an integral part of the African Renaissance is a viable scientific and technical framework for sustainable management of the Large Marine Ecosystems and for the protection of coastal and ocean environments. GOOS-Africa is the African contribution to the development and implementation of the Global Ocean Observing System (GOOS).
GOOS-Africa: Operational tool for Monitoring and Predicting the dynamics of African LMEs Successful arrangements concluded between GOOS-Africa and European Space Agency (ESA) enable free release of MERIS/ENVISAT data related to the African waters for the use by the members of the GOOS-Africa Network. The BCLME is the first GOOS-AFRICA partner that took advantage of these opportunities, in collaboration with the Department of Oceanography of the University of Cape Town (UCT) and the Marine and Coastal Management (MCM) of the Government of South Africa, combining remotely sensed MERIS and MODIS data with in situ measurements and observations to generate useful marine products and services for monitoring and predicting the health and status of the BCLME. Recently, in March 2005, responding to the call of African countries and Governments following the devastating Indian Ocean Tsunami, the GCLME in partnership with GOOS-Africa organized the Workshop on Coastal Dynamics in Integrated Areas Management and Early Warning Systems in Africa to identify the criteria for establishing an integrated Multi-hazards Early Warning and Mitigation System. GOOS-Africa has been called upon to contribute to the Project Development Phase of the CCLME and is participating in the Road Map Workshop that will develop the work programme for implementation. The joint leadership role of the BCLME, GCLME, GOOS–Africa, UCT and MCM to foster the development of operational oceanography in Africa shows the evidence of the positive synergy and complementarity between GOOS and the LMEs programmes, at least in the African context, towards integrated monitoring and predicting the environment and the ecosystem in LMEs. GOOS-Africa provided the observation background and substantial elements for the development of the ODINAFRICA-III project. This contribution enables installation of a number of modern tides gauges along African coasts reinforcing the networks of in situ measurements.
GOOS-Africa
43
GOOS-AFRICA FORWARD LOOK The positive synergy between GOOS-Africa and the African LMEs fosters a rapid implementation of capacity building and forecasting priorities and needs identified in the two previous chapters, 1 and 2, of this book through joint major initiatives in preparation, notably: (i)
Pan-African LMEs/GOOS-AFRICA Leadership Workshop on Operational Oceanography and Remote Sensing in Africa, Cape Town, South Africa 6-10 November 2006
(ii)
The Second Pan-African LMEs Forum, Cape Town, South Africa 13 November 2006
(iii)
The Third Forum of the GOOS Regional Alliances, Cape Town, South Africa, 14-17 November 2006-05-26
ACKNOWLEDGEMENTS The Author is grateful to Professor Vere Shannon, former Director of the Sea Fisheries Research Institute of South Africa for his continuous encouragement and patient pedagogical advice. Special thanks to Dr. Steward Bernard, Mr. Cristo Whittle (UCT) and Dr. Antoine Mangin (Coastwatch/ACRI) who provided part of the data supporting the illustrations of this Chapter. The Author thanks also Professor Geoff Brundrit, the past Chairman of GOOS-AFRICA; Dr Kwame Koranteng, Current Chairman of GOOS-AFRICA; Dr. Silvana Vallerga, the Chairperson of the Intergovernmental Committee of GOOS; Professor John Woods, Chairman of the first Experts Group on GOOS; Dr. Kenneth Sherman, US-NOAA founder of the LME Concept; Dr Brad Brown, former Director of the US-NOAA Southeast Sea Fisheries Center; Professor Gotthilf Hempel, former Director of the Institute of Tropical Ecology of Bremen; Professor Chidi Ibe, the Regional Director of the GCLME; Dr. Mike O’Toole, the Chief Technical Advisor of the BCLME; Mr. Mohammed Boulahya, first Director General, Co-Founder of ACMAD and Senior Expert Climate and Environment NEPAD Secretariat; Dr. Pierre-Philippe Mathieu, ESA; Dr. Mary Altalo, VicePresident, US-based Sciences Applications International; Dr. Colin Summerhayes, Former Director of the GOOS Project Office at IOC/UNESCO and current Executive Director, Scientific Committee on Antarctic Research (SCAR); and the Assistant Director General of IOC/UNESCO, Dr. Patricio Bernal. REFERENCES AAAS. 1994. Science in Africa: The Challenges of Capacity-Building, A forum organized by the AAAS Sub-Saharan Africa Program, American Association for the Advancement of Science, Washington, DC, May 10, 1994.
44
J. Ahanhanzo
AAAS. 1995. New Directions for Science and Technology in South Africa: Opportunities for US Collaboration, A Seminar organized by the AAAS Sub-Saharan Africa Program American Association for the Advancement of Science Washington, DC, May 2, 1995. AAAS. 1996. Utilizing Africa's Genetic Affluence through Natural Products Research and Development, A Symposium at the 1996 Annual Meeting, American Association for the Advancement of Science, Washington DC 1996. Agenda 21. 1992. The United Nations Conference on Environment and Development, UNCED Document A/CONF. 151/4 (Parts I and II). The United Nations Convention on the Law of the Sea, United Nations, New York, 1983. Ahanhanzo, J. 1995. Integrated Development of Modern Oceanography and the Management of Oceanographic Resources in the Benguela Current Region and Comparable Eastern Boundary Upwelling Ecosystems. Ahanhanzo, J. 1998. The GOOS-AFRICA Concept, in the Guinea Current Large Marine Ecosystem, UNIDO, Editor Professor Ibe Chidi. Ahanhanzo, J. 2003. The Rise of GOOS-AFRICA In the IOC Annual Report, 2003. Chu, P.C. and J.C. Gascard, editors. 1991. Deep Convection and Deep Water Formation in the Oceans, Elsevier Oceanography Series 57, Amsterdam, 382 p. CNES. 1998. French space ambitions, Paris. Dahlin, H., N.C. Flemming, K. Nittis, S.E. Petersson, editors. 2003. Building the European Capacity in Operational Oceanography, Proceedings of the Third International Conference on EuroGOOS, Elsevier Oceanography Series 69, Amsterdam. Duda, A.M. and K. Sherman. 2002. A new imperative for improving management of large marine ecosystems. Ocean & Coastal Management 45:797-833. EuroGOOS. 1997, Annual Report 1995-96, EuroGOOS Publication No. 2, Southampton Oceanography Centre, Southampton UK. Flemming, N.C., S. Vallerga, N. Pinardi, H.W.A. Behrens, G. Manzella, D. Prandle, J.H. Stel, editors. 1999. Operational Oceanography: Implementation at the European and Regional Scales, Proceedings of the Second International Conference on EuroGOOS, Elsevier Oceanography Series 66, Amsterdam. Guymer, T.H., N.C. Flemming, J. Font, P. Gaspar, J. Johannessen, G.H. van der Kolff, C. le Provost, A. Ratier and D. Williams. 2001. EuroGOOS Conference on Operational Ocean Observations from Space, EuroGOOS Publication No. 16, Southampton Oceanography Centre, Southampton. Pinardi, N. and N.C. Flemming, editors. 1998. The Mediterranean Forecasting System Science Plan, EuroGOOS Publication No. 11, Southampton Oceanography Centre, Southampton. Sherman, K. and L.M. Alexander, editors. 1986. Variability and Management of Large Marine Ecosystems. AAAS Selected Symposium 99. Westview Press, Colorado. 319p. Sherman, K. 1994. Sustainability, biomass yields, and health of coastal ecosystems: An ecological perspective. Mar Ecol Prog Ser. 112:277-301. Sherman, K. and Q. Tang, editors. 1999. Marine Ecosystems of the Pacific Rim: Assessment, Sustainability, and Management. Blackwell Science, Inc., Malden, MA. 465p. Sherman, K. and A.M. Duda. 1999. An ecosystem approach to global assessment and management of coastal waters. Mar Ecol Prog Ser.190: 271-287. UNESCO/IOC. 1992, Twenty-fifth Session of the Executive Council, Paris, 10-18 March 1992, IOC Reports of Governing and Major Subsidiary Bodies. UNESCO/IOC. 1993, IOC Committee for the Global Ocean Observing System (GOOS), First Session, Paris, 16-19 February 1993, IOC Reports of Meetings of Experts and Equivalent Bodies. UNESCO/IOC. 2005, IOC-IUCN-NOAA Consultative Meeting on Large Marine Ecosystems (LMEs), Sixth Session, 29-30 March 2004, Paris, IOC Reports of Meetings of Experts and Equivalent Bodies. UNESCO/IOC 1998. “The GOOS 1998” IOC, Paris. U.S. GOOS. 1992. First Steps Towards A U.S. GOOS, Report of a Workshop on Priorities for U.S. Contributions to a Global Ocean Observing System, Woods Hole, Massachusetts, 14-16 October 1992, Joint Oceanographic Institutions Inc., Washington. GOOS Report 62, IOC-UNESCO Publications. Woods, J.D. 1991a. Oceanography on a global scale: The new challenge Phys. Ed 26:159-163, 168. Woods, J.D. 1991b. Global Ocean Observing and Climate Forecasting Science in Parliament 48 (3):4-10 Woods, J.D. 1992. Monitoring the ocean. In Cartledge, B. ed. Monitoring the Environment, Oxford University Press
GOOS-Africa
45
Woods, J.D. 1993. The case for GOOS (The Global Ocean Observing System) Intergovernmental Oceanographic Commission IOC /INF-915, Paris Woods, J.D. Dahlin, H., L. Droppert, M. Glass, S. Vallerga,. and N.C. Flemming. 1996. The Strategy for EuroGOOS. EuroGOOS Publication N.1, Southampton Oceanography Centre, Southampton. ISBN 0904175227 Woods, J.D., H. Dahlin, L. Droppert, M. Glass, S. Vallerga, and N.C. Flemming. 1997. The EuroGOOS Plan. EuroGOOS Publication N.1, Southampton Oceanography Centre, Southampton. ISBN 0904175226.
ThisPage pageis intentionally This Intentionally left Leftblank Blank
Part II: Setting the Scene Data time series and models: What we think we know about variability in the Benguela and comparable systems.
ThisPage pageis intentionally This Intentionally left Leftblank Blank
Large Marine Ecosystems, Vol. 14 V. Shannon, G. Hempel, P. Malanotte-Rizzoli, C. Moloney and J. Woods (Editors) © 2006 Elsevier B.V. All rights reserved.
49
4 Large Scale Physical Variability of the Benguela Current Large Marine Ecosystem (BCLME) F. A. Shillington, C. J. C. Reason, C. M. Duncombe Rae, P. Florenchie, and P. Penven
INTRODUCTION The Benguela Current Large Marine Ecosystem (BCLME) is situated off the west coast of Africa between 5-37ºS, 0-26ºE, and spans the three countries of Angola, Namibia and South Africa. It is one of the four major eastern boundary current upwelling systems of the world oceans (Hill et al. 1998), and although it has some similar characteristics to the other eastern boundary upwelling areas, a unique feature is that it is bounded on both the equatorial and poleward extremities by warm water current systems (the tropical warm Angola Current system in the north, and the Indian Ocean western boundary Agulhas Current System in the south; Shannon and Nelson, 1996; Shillington 1998; Shannon and O’Toole 2003). In the region between about 1537ºS, the surface currents are generally equatorward, with vigorous coastal upwelling cells, strong and narrow equatorward shelf edge jets (near Cape Town which is situated at 34ºS, 18ºE and off Lüderitz; 28ºS, 15ºE), and a poleward undercurrent along the shelf slope and bottom. The warm tropical Angola Current System (Ajao and Houghton 1998) generally has southward moving coastal currents which meet the Benguela Upwelling System at the Angola-Benguela Frontal Zone (ABFZ) at ~1517ºS (Shannon et al. 1987; Field and Shillington 2005; Monteiro and van der Plas, this volume; Veitch et al. 2006). The Angola Current is affected by input from the equatorial wave guide, the South Equatorial Current (SEC) and the South Equatorial Counter Current (SECC) at ~5°S (Peterson and Stramma 1991). Details of the circulation of the Angola Gyre and the nature of the Angola Dome are addressed by Monteiro and van der Plas (this volume), and by Reason et al. (this volume). At the centre of the BCLME region is an area of year-round coastal upwelling, 1530°S (Boyer et al. 2000); and a region of seasonal upwelling, 30-34°S. Coastal trapped waves have been observed to propagate polewards on the continental shelf at regular synoptic time scales (~3-10 day periods) from Walvis Bay in Namibia (20°S), and to continue around the Cape of Good Hope and up to 800 km east along the eastern coast of South Africa (Brundrit et al. 1987; Schumann and Brink 1990).
50
F.A. Shillington et al.
At the southern end the BCLME region, the Agulhas Bank (see Fig. 1-1 in Chapter 1 for the shelf topography of the BCLME region) is a very wide shelf region along the southern coast of Africa from 18-26°E, that has a highly vertically stratified water column in the west in summer, and a well mixed water column in the winter (Schumann 1998). Closer to the coast, there is summer upwelling of cool nutrient rich water at the major coastal embayments on the African coast between these longitudes. In the middle of the Agulhas Bank, there is a seasonal cold tongue apparent in surface and near-surface waters; the circulation around this feature appears to be cyclonic (Boyd and Shillington 1994). This feature is particularly visible as a ridge of elevated chlorophyll, occurring in the period from March-June (Demarcq et al 2003; HardmanMountford et al. 2003). The Agulhas Bank is very important for the spawning of pelagic fish such as anchovy and pilchard from September to March (Hutchings et al. 2002). After spawning, eggs and larvae drift northwards in the jet current past Cape Town, until juvenile fish recruitment occurs about 150 km north along the coast at ~32°S in St Helena Bay. Adult fish then make their way back to the Agulhas Bank to spawn in the following austral spring-summer (van der Lingen et al. this volume). Large-scale, multiyear climatic variations in the Benguela upwelling region have been observed from time to time and have been dubbed “Benguela Niños” as an analogue to the Pacific event (Shannon et al. 1986). The Benguela Niño, like its Pacific counterpart, has a strong effect on regional fisheries and this in turn has led to an effort to forecast these events. Benguela Niños have been observed/reported in 1934, 1963, (1972/3), 1984, 1995 (Shannon et al. 1986; Gammelsrød et al. 1998). Field measurements of the 1995 Benguela Niño were reported by Gammelsrød et al. (1998). More recently, Florenchie et al. (2003) and Florenchie et al. (2004) have examined the nature of the 1984 and 1995 Benguela Niños using an ocean general circulation model together with satellite derived sea surface temperature (SST) and sea surface height (SSH) data, to show how they can be related to local and remote wind forcing. Their results suggest that a possible forecast lead time of two months exists for anticipating strong positive SST anomalies propagating from the equatorial region, polewards beyond the Angola Benguela Frontal Zone (ABFZ). Benguela Niños represent the lowest frequency, largest-scale instance of variability in the BCLME. The main large scale physical features of the BCLME are summarised in the cartoon in Fig. 1-1 in Chapter 1. The main purpose of this review is to set the scene for a discussion of potentially forecastable aspects of major importance in the BCLME. Questions that are central to this discussion are formulated below. • What proportion of the BCLME large scale variability is associated with the seasonal forcing, and what part is related to inter-decadal events such as the Benguela Niño? • What is the role of other large scale modes such as ENSO and the Southern Annular Mode in driving variability in the BCLME region ?
Large scale physical variability of the Benguela Current LME
51
• Can monitoring the remote and/or local wind forcing in the western equatorial Atlantic region give an acceptable lead time for nowcasting/forecasting Benguela Niños at or south of the Angola Benguela Frontal Zone? • How is the Benguela Niño signal transmitted/influenced by the poleward flowing Angola Current and what is the nature of the interaction with the northern part of the Benguela Upwelling at the ABFZ at ~15-17ºS? • What is the nature and importance of the Angola Dome area for the large scale formation of low oxygen water, and is this water responsible for low oxygen water in the northern Benguela? • What is the effect and importance of the variability of the outflow of the main large rivers (e.g. Congo/Zaire and Orange/Gariep Rivers) on the BCLME? • Is it possible to nowcast/forecast changes in the extremely vigorous wind driven upwelling in the Lüderitz region, which tends to persist throughout the year? • What is the nature of the predictability of the remote and local forcing of the southern Benguela Upwelling System from the poleward end via the influence of the Agulhas Current and its retroflection? • Is there seasonality in the Agulhas Current and the shedding of Agulhas retroflection rings? Is the shedding of rings predictable or capable of being monitored sufficiently far “upstream in the Agulhas Current” to provide advance warning of interaction with the Benguela upwelling front? Major intrusions of sub-Antarctic water have been observed to interact with the outer boundary of the southern Benguela Ecosystem in 1987 (see Shannon et al. 1990 as cited in Hardman-Mountford et al. 2003). Are such intrusions of sub-Antarctic water into the southern Benguela System important sources of variability and are they predictable? MAJOR PHYSICAL PROCESSES IN THE BCLME The main dynamic processes in the BCLME are similar to other major eastern boundary upwelling systems (Hill et al. 1998). They include: • Dominant equatorward wind stress inducing Ekman offshore transport of surface water, which is replaced by cool, nutrient rich subsurface central water (see detailed section below on recently measured water masses in the central Benguela). The upwelling process leads to the surfacing of cool, coastal nutrient rich water; the subsequent growth and decay, and instability of oceanic fronts, filaments and frontal jets e.g. Shillington (1998). • A poleward undercurrent along the continental shelf break which later intrudes onto the shallower continental shelf in various places. The detailed mechanism responsible for this is not clear at present. • Poleward propagating coastal trapped waves on the shelf, which are easily detected in coastal tide gauge recordings of sea level, e.g. Brundrit et al. (1987). • Kelvin wave like disturbances travelling eastwards along the Atlantic Ocean equatorial waveguide, travel from South America to Africa, and later turn polewards along the Angolan coast. Temperature anomalies can give rise to either
52
F.A. Shillington et al.
local warm events, or in some cases, Benguela Niños that reach as far south as Walvis Bay (~22°S), e.g. Florenchie et al. (2004). • Agulhas ring formation after the Agulhas Current Retroflection south and west of Cape Town, and the subsequent interaction of these rings with the southern Benguela upwelling frontal system, e.g. Duncombe Rae et al. (1992), Shillington (1998). This process is unique to the BCLME, and not found in any of the other major eastern boundary current systems. • The seasonal and inter-annual meridional movement of the ABFZ, and the quasidecadal variability of Benguela Niños, e.g. Veitch et al. (2006). This process appears to be unique to the BCLME, and not found in any of the other major eastern boundary current upwelling systems. The main transboundary areas of the BCLME are: the northern Angola Current border with the equatorial currents; the ABFZ; the Lüderitz-Orange River cone area (Duncombe Rae 2005); the Agulhas Current-Benguela upwelling interaction at the southern boundary of the Benguela upwelling area; the coastal transition zone between the cold upwelling coastal and continental shelf region and the deeper ocean. ATMOSPHERIC FORCING OF THE BCLME The atmospheric circulation of the BCLME region is dominated by the South Atlantic subtropical anticyclone which gives rise to southerly wind stress near the west coast of Africa. To the south of Africa and the region, there is generally a westerly flow, with changes in wind direction associated with west to east travelling mid-latitude cyclones. During austral summer, surface heat induced low pressure systems develop over western South Africa, enhancing the zonal pressure gradient and leading to an intensification of the southerly wind stress off the west coast. A separate heat induced low pressure system develops over southern Angola/northern Namibia with an associated westerly windstress off the tropical SE Atlantic that feeds into the confluence between the ITCZ and the Congo air boundary. In winter, the major atmospheric circulation features shift north so that most of the BCLME region is dominated by low level southerlies. The exception to this is south of about 30°S, which is subjected to frequent atmospheric frontal activity (e.g. Hardman-Mountford et al. 2003). Superimposed on these seasonal changes in the low level winds is considerable mesoscale, synoptic, intra-seasonal, inter-annual and longer time scale variability. On synoptic time scales, the predominant anticyclonic equatorward wind flow along southern Namibian and South African west coast is perturbed by cold fronts, coastally trapped low pressure systems, “cut off lows”, and mesoscale features such as “berg winds” and sea breezes. Sometimes berg winds are followed by “coastal lows” (Reason and Jury 1990) which tend to significantly perturb the coastal wind fields. Cold fronts are most common in the winter half of the year whereas “cut off lows” may occur in any season but tend to be more likely in the austral spring and autumn.
Large scale physical variability of the Benguela Current LME
53
West coast troughs may affect the entire coast south of about 10°S but are more common in the north (south) during winter (summer). All these weather systems interrupt the upwelling-favourable winds at a variety of space and time scales. Risien et al. (2004) examined sixteen months of QuikScat satellite derived windstress data in the Benguela System, using an artificial neural network (the Kohonen self organising map) to divide the region into six discrete regions, and wavelet analysis to extract the spatial and temporal variability scales between four and sixty four days. Chelton et al. (2004) have located significant time independent narrow bands of cyclonic curl (see Fig. 4-1; negative in the Southern Hemisphere) with large alongshore scales, adjacent to the western coastline of southern Africa from an analysis of four years of Quikscat windstress. The detailed structures and evolutions of these nearshore curl and divergence features were previously poorly resolved by historical ship observations. The implication of the long term averaged cyclonic curl of windstress is that the shallow eastern boundary Atlantic Ocean thermocline will be elevated upwards towards the surface, while the divergence will modulate the upwelling along the coast.
Figure 4-1. Four year average windstress curl (left) and divergence (right) calculated from Quikscat. Units are N m-3 x 107. (After Chelton et al. 2004)
54
F.A. Shillington et al.
LARGE SCALE MODES OF VARIABILITY A number of large-scale modes of variability influence the atmospheric circulation over the South Atlantic, and hence the BCLME region. These include ENSO, which primarily influences the region via the Pacific-South America pattern (Mo and Paegle 2001; Colberg et al. 2004), the semi-annual oscillation (van Loon 1967), the Antarctic Oscillation or Southern Annular Mode (Kidson 1988), and large-scale modulations of the subtropical anticyclone which may be locally forced (Venegas et al. 1996; 1997; 1998) or related to near-hemispheric modulations of the subtropical high pressure belt (Jones and Allan 1998; Reason 2000). Modulations in the trades over the western tropical Atlantic may generate Benguela Ninos (Florenchie et al. 2004) which may then influence the atmospheric circulation over the northern Benguela region (Rouault et al. 2003). In addition, shifts in the atmospheric wave number three pattern can often produce dipole-like SST variability in the South Atlantic and South Indian Oceans (Fauchereau et al. 2003; Hermes and Reason 2005) that tends to occur during the summer. There are several other large-scale modes that are important for the tropical Atlantic (meridional gradient mode, zonal mode, North Atlantic Oscillation) and whose potential influence on the BCLME region needs to be assessed (see Chapter 10: Reason et al. this volume). WATER MASSES AND VERTICAL STRUCTURE OF THE BCLME The major oceanic influences on the Benguela upwelling system are derived from the equatorial Atlantic in the north and the South Atlantic/South Indian to the south. Direct water mass analysis in the BCLME (Figs. 4-2 and 4-3) can be used to discriminate the influence of tropical water entering from the Angola Basin and the northern Benguela, from that being upwelled in the southern Benguela. A recent comparative study of the historical record of nutrients and hydrographic properties of the Benguela has been made by Kearns and Carr (2003). Antarctic Intermediate Water (AAIW) that is formed at the surface in the sub-polar and polar regions has a salinity minimum deep in the water column, with distinct characteristics in the northern and southern Benguela (Shannon and Hunter 1988; Talley, 1996). From the Angola Basin a high (relative to the southern Benguela water type) salinity AAIW (HSAIW) enters the northern Benguela in a poleward undercurrent along the shelf edge. The southern Benguela has a low salinity AAIW (LSAIW) close to the Subtropical Front. Similarly the South Atlantic central water in the Benguela has a relatively High Salinity component (HSCW) originating in the tropical Angola Basin and relatively Low Salinity Central Water (LSCW) in the Cape Basin. Above the central waters there is higher salinity, warm Oceanic Surface Water (OSW). The surface water is subject to the influence of precipitation and continental runoff from rivers into the Angola Basin resulting in low salinities at the surface (Mohrholz et al. 2001). In the southern Benguela the run-off from the Orange River is intermittent
Large scale physical variability of the Benguela Current LME
55
and controlled by dams and therefore less evident in extent and persistence than in the north. The central water on the shelf is upwelled near the coast by the persistent equatorward component of the wind. Because of the atmospheric modification of temperature and salinity it is designated Modified Upwelled Water (MUW). In general terms, the intermediate, central, and upper waters can be summarised as having either (a) a high salinity, high temperature character indicating a tropical influence; or (b) a low salinity, low temperature character indicating an Antarctic or sub-Antarctic influence. Appropriate modifications of the surface and upwelled water occur during contact with the atmosphere due to solar heating and turbulent mixing processes.
Figure 4-2. Sampling stations on the BENEFIT cruises of RV Africana in 1999 (lines GG, WB) and 2002 (lines E, H, L), and from the ASTTEX deployment cruise of RV Melville in 2003 (line R). The dashed contour represents the 200 m isobath. Other isobaths are labelled.
56
F.A. Shillington et al.
Figure 4-3. Θ-S diagram of the water column profiles from the northern and southern Benguela. Note the clear salinity difference between the central waters of the two extremes. The water mass definitions used in the text are superimposed. The water mass definitions used in the text are superimposed. The water masses labelled are: ABW – Antarctic Bottom Water; NADW– North Atlantic Deep Water; LSAIW – Low Salinity Antarctic Intermediate Water; HSAIW – High Salinity Antarctic Intermediate Water; LSCW – Low Salinity Central Water; HSCW – High Salinity Central Water; MUW – Modified Upwelled Water; OSW–Oceanic Surface Water. The very low salinity seen in the surface water of some stations is due to continental run-off. Water masses below the isopycnal shown (σt= 27.75 kg.m.-3) are not discussed in detail in the text.
Figure 4-4. Monthly temperature fields in April at level 20 (surfaceleft panel) and at level 16 (approximating the mixed layer, right panel) from ROMS model.
Large scale physical variability of the Benguela Current LME
57
The waters of the South Atlantic Ocean thermocline (central water) layer originate in two source water mass regimes (Poole and Tomczak 1999): Eastern South Atlantic Central Water (ESACW) and Western South Atlantic Central Water (WSACW). The ESACW is derived from the Indian Central Water through the Agulhas Current, and the WSACW is derived from the Brazil Current through the Brazil/Malvinas Confluence in the western South Atlantic subtropical gyre. In the eastern basins of the South Atlantic, the WSACW is present in the Angola Basin while the ESACW is found in the Cape Basin. The characteristics of WSACW are modified in the region of the ABFZ from their source water characteristics by upper layers processes in the equatorial Atlantic (Mohrholz et al. 2001). Below the main thermocline, the AAIW on the west coast of southern Africa has a salinity of 34.35, rising to 34.50 near the ABFZ (Talley 1996; Duncombe Rae 1998; Shannon and Hunter 1988; Mohrholz et al. 2001). Higher salinities are found in the intermediate water on the east coast, and in the Agulhas Current, due to the influence of occasional intrusions of Red Sea Water (Gründlingh 1985). These latter high salinity AAIW sources, however, appear not to influence the intermediate water of the central Benguela. The broad circulation of the water masses (described by Shannon and Nelson, 1996, after Chapman and Shannon, 1985) is indicated in detail by the steric height anomaly at 500 dbar (Reid 1989) and shows two opposing gyres within the South Atlantic which have a confluence in the region of the Lüderitz upwelling cell (Mercier et al. 2003). In vertical sections across the shelf the high salinity water appears constrained to the shelf edge, consistent with a poleward undercurrent of Angola Basin origin. In the region of the Lüderitz upwelling centre, consistent with Monteiro (1996), the southward moving water in the poleward undercurrent appears directed off-shore at about the same level as a local oxygen minimum in the central water of the Cape Basin gyre. Discontinuity in water masses between this latitude and the Orange River Mouth suggests that the Lüderitz upwelling cell at 26°40’S diverts the southward movement of high salinity central water in the poleward undercurrent. As an indication of the extent of the exchange between the two kinds of central water, the proportion of the HSCW within the water column was determined as a fraction of the central water as defined above. The distribution of this proportion shows the exchange between the two extremes of the system occurring between Lüderitz and Cape Frio. As the vertical sections of water mass show, the water masses remain separable showing little mixing. It is only the extent of the denser high salinity portion that becomes less as the Lüderitz cell is approached. NUMERICAL OCEAN MODELLING IN THE BCLME The numerical modelling of the oceanic properties is a central aspect for oceanic forecasting in the BCLME. During the last 25 years, several models have been applied
58
F.A. Shillington et al.
to the Benguela Current System. While a number of these models concentrate uniquely on a limited portion of the ocean around southern Africa, others include ocean basins such as the South Atlantic or the Indian Ocean. With the tremendous increase in power of the supercomputers, Ocean General Circulation Models (OGCMs) might now be able to possess sufficient resolution to resolve the major processes in the Benguela (see the data and modelling animations in folder B of the CD ROM). Coastal models Van Foreest and Brundrit (1982) designed the first model for the South African west coast. The originality of their approach lay in the decomposition of the equations of motion into two vertical modes. The model domain extended from 70 km south of Cape Peninsula to North of St Helena Bay and to more than 150 km offshore. Open boundaries were applied at the connection with the open ocean. Although the model was forced with a constant wind and the duration of the simulation was brief (3 days), the solution showed some interesting spatial variability. The ocean modelling group at the CSIR is currently developing a high resolution model for the circulation in St Helena Bay (Monteiro and Kemp, personal communication). They use the Delft3D-FLOW ocean model with a variable grid resolution that ranges from a few kilometres offshore, to a few hundreds meters close to the coastline. The vertical grid is decomposed into 8 sigma layers and the model is forced by real time winds calibrated by a coastal weather station. This model is expected to resolve the coastal poleward flow during upwelling relaxations. To understand the retention of fish larvae in St Helena Bay, an idealized barotropic model (Penven et al. 2000) was designed and implemented during the VIBES-IDYLE project. The shallow water equations were solved on a 5 km resolution grid in a periodic channel forced by a constant alongshore wind. In the lee of Cape Columbine, the model produced a cyclonic recirculation that is able to retain biological elements. To extend the analysis to the different oceanic processes which might affect pelagic fish recruitment along the South Africa West Coast, a 3D regional configuration based on the Regional Ocean Modelling System (ROMS) has been implemented by Penven et al. (2001a). The model grid followed the coastline from Cape St Francis, 100 km west of Port Elizabeth to Lüderitz (see Fig. 1-1 in Chapter 1). The horizontal resolution ranges from 9 km at the coast to 18 km offshore. On the vertical, 20 sigma levels are stretched to keep a sufficient resolution close to the surface. The information at the open boundaries is provided by a basin scale model, and the atmospheric forcing was derived from the comprehensive ocean and atmosphere dataset (COADS) climatology. Using this model configuration, Blanke et al. (2002) quantified the wind contribution to inter-annual SST variability. They found that while the west coast is affected by mesoscale activity, the wind appears to be the dominant driving for the variability over the Agulhas Bank. By coupling the physical model to an individual based model, Parada et al. (2003) quantified the influence of eggs floatability on the transport from the Agulhas Bank to St Helena Bay, while Mullon et al. (2002) tested the "obstinate nature" hypothesis for the selection of the spawning zone, and Hugget et al. (2003) examined the influence of different
Large scale physical variability of the Benguela Current LME
59
environmental factors on the transport of fish eggs and larvae in the Southern Benguela. Larger scale models At a regional scale, Skogen (1999) adapted NORWECOM, a model based on the Princeton Ocean Model, to simulate the ocean around the whole south-western southern Africa (i.e. including the coasts of Angola, Namibia and South Africa). The resolution was 20 km, and the model was forced by the National Centre for Environmental Prediction (NCEP) winds, with a surface nudging of SST and no surface salinity flux. Below 500 m, a relaxation towards Levitus climatology prevented the solution from drifting numerically. The physical model has been coupled to a biogeochemical model and to an Individual Based Model to simulate the fate of sardine larvae in the Northern Benguela (Stenevik et al. 2003). Speich et al. (2004) have especially designed a model to explore the Agulhas Retroflection, and its influence on the BCLME. They used ROMS at 1/6° and 1/10° degree resolution, with 32 vertical levels, forced by a monthly wind climatology derived from QuickSCAT scatterometer data and OGCM data for the open boundaries. Their simulations show the sensitivity of the Agulhas Current to the bottom slope steepness and its variations. Basin scale and global models Barnier et al. (1998) performed one of the first basin scale experiments using a sigma coordinate model. They applied SPEM for the Southern Atlantic at 1.375° resolution. The model had 20 vertical sigma levels and was forced by the Hellerman and Rosenstein wind stress climatology. Although very coarse, this model was able to capture some of the large scale features in the Benguela region. Biastoch and Krauß (1999) took advantage of the curvilinear coordinate in MOM2 to design a model at coarse resolution over the South Atlantic and the South Indian Oceans, but with an increase of resolution in the South African waters (1/3°). The model was forced by ECMWF winds and used data from an OGCM for its lateral boundary conditions. From this simulation, Reason et al. (2003) derived a heat export into the Southern Atlantic at 20°E of 1 PW in winter and 0.7 PW in summer. The behaviour of the Agulhas retroflection has been extensively studied in the Fine Resolution Antarctic Model (FRAM) simulation (Lutjeharms and Webb, 1995). FRAM is based on the Bryan-Cox-Semtner ocean model, it encompasses the totality of the Southern Ocean from 24° S to the Antarctic at a resolution of 1/2° in longitude and 1/4° in latitude (i. e. approximately 27 km around 60°S). FRAM appeared to be able to reproduce several observed patterns of the Agulhas retroflection, but displayed too much regularity in the subsequent path of the Agulhas Rings into the Atlantic Ocean, compared with observations.
60
F.A. Shillington et al.
For the same period, simulation experiments of the basin scale circulation were also conducted by Florenchie and Veron (1998) with an eddy-resolving 1/6° quasigeostrophic model. By means of a nudging data assimilation procedure along satellite tracks, Topex/Poseidon and ERS1 altimeter measurements were introduced in the model to control the simulation. The assimilation procedure enabled to produce schematic diagrams of the circulation in which patterns ranging from basin-scale currents to mesoscale eddies were portrayed in a realistic way. Treguier et al. (2003) have analyzed the generation and the fate of cyclonic and anticyclonic eddies from the Agulhas retroflection in an eddy resolving simulation of the whole Atlantic Ocean (CLIPPER). The model employed is OPA, at 1/6° resolution, with 42 z-levels, and forced by ECMWF ERA15 data from 1979-1993. OGCMs now start to have sufficient resolution to be relevant for the Benguela region. The United States National Research Laboratory (NRL) Layered Ocean Model (NLOM) and the NRL Coastal Ocean Model (NCOM) are presently running globally in real-time at respectively 1/16° and 1/8° resolution (Rhodes et al. 2002). In Japan, the enormous computational power of the “earth simulator” made it possible to run a global simulation at a resolution of 1/10° (Masumoto 2004). The respective role for the transport of heat and salt of cyclonic and anticyclonic eddies that are generated in the Agulhas region has been quantified in a global simulation based on POCM (Parallel Ocean Circulation Model) at 1/4° resolution (Matano and Beier 2003). SCHEMATIC CIRCULATION DEDUCED FROM A NUMERICAL MODEL There is a dearth of observations in the northern BCLME. Therefore one of the BCLME projects has examined the output from a numerical ocean model such as ROMS (e.g. Fig 4-4) and CLIPPER. These model outputs could then be used as a cost effective method to test various hypotheses, and to guide the observational programme of monitoring the environment in this region of the BCLME. The CLIPPER numerical simulation model The most recent CLIPPER experiment is a simulation of the global Atlantic oceanic circulation (http://www.ifremer.fr/lpo/clipper/present.html) based on the OPA model (http://www.lodyc.jussieu.fr/opa/). From 1990-1992 (the period of the spin up), the model is forced by a windstress climatology based on the European Earth Resources Satellite (ERS) derived wind fields. For the period 1993-2000, the model is forced directly by the more realistic varying direct ERS wind field products. The European Centre for Medium Range Weather Forecasting (ECMWF) heat and freshwater fluxes are used in combination with the Reynolds sea surface temperature (SST) for the heat feedback term for the period 1990-2000. The model domain covers most of the Atlantic Ocean and extends from 60°S-60°N. The model output has been examined to determine aspects of the seasonal circulation from 0-30°S.
Large scale physical variability of the Benguela Current LME
61
Surface layers (0 to 30m) The modelled shelf circulation appears to be dominated by a narrow coastal current flowing northward all year long from about 30°S to the ABFZ near 18°S. Its intensity is higher during austral summer (January-March) and lower in winter (June to August). The coastal area between 26-22°S is somewhat different: the northward flow is less intense and it exhibits a weaker seasonal cycle. In fact there is an abrupt change in the current field immediately north of Lüderitz (28°S). From 30-26°S, the maximum velocity in the core of the current remains constant, with a value of about 25 cm s-1 in summer. At 26°S, the current speed maximum decreases abruptly to values of about 15 cm s-1 and the current intensity fluctuates as far southwards as 22°S. North of 22°S, the current speed increases once again. Figure 4-5 is a schematic representation of the model circulation at the surface (Fig. 4-5a – Lev 01) and at a 40m depth (Fig. 4-5b – Lev 04): Figure 4-5a is representative of the perennial modelled coastal circulation. However the northward coastal current (1 and 3 on the figure) is weaker during austral winter.
Figure 4-5. Schematic circulation reproduced by the CLIPPER model at the surface (a) and at 40 m (b).
The current is always stronger south of Lüderitz (branch 1). Then it bifurcates partly westward (branches 2) leading to a decrease of its transport and intensity north of 26°S. The orientation of the coast also changes near Lüderitz from northwest to north. Windstress in the area shows a regular northwestward direction all year long and it is stronger in summer south of 26°S. As a result, the surface discontinuity observed in the coastal current at 26°S might originate from the orientation of the coast, the wind field strength and its direction. The westward circulation in (4) does not seem to interact with the coastal circulation pattern. Circulation at 40m depth (lev04) The model circulation at 40m is globally identical to the surface, although weaker (Figure 4-5b). The maximum current speed in branch 1 is about 20 cm s-1 in summer. The northward coastal current experiences a similar seasonal cycle with minimum
62
F.A. Shillington et al.
intensity during austral winter. The discontinuity at 26°S is still present. The dotted line of branch 3 means that the current is not always clearly defined. The main change in Figure 4-5b concerns a new seasonal pattern in the circulation that occurs twice a year in February-March and October: a poleward current centred at 14°E develops off shore in the north and reaches a latitude of about 25-26°S at its maximum in October (branch 4). The dotted arrow indicates that the current is not permanent throughout the year. Its temperature is about 4°C higher than the coastal current with slightly higher salinity levels. It does not seem to interact much with the northward current. Circulation at 80m and 130m depths (lev07 and lev10) Figure 4-6 represents diagrams based on the model circulation at 80m and 130m depths for the same area. Dotted arrows indicate that the current is intermittent and shows some seasonal variability. It is at this model level in which the main major differences occur when a comparison is made with the surface layers. The northward coastal current is much weaker with maximum speeds of about 10 cm s-1 and it is not as clearly identifiable. It still shows a seasonal cycle but its intensity is higher in July and August, instead of summer. Branches feeding the current south of Lüderitz (1) are unstable and not well defined. Despite this, the northward current still reaches its maximum intensity in the Lüderitz area. The southward current (branch 4) intensifies in comparison with upper levels and is now noticeable from September-April. It exhibits two maxima: one in October and another in March, respectively. On these occasions the current meets the westward branch (2) of the coastal current near 26°S. It is more saline than the surrounding water and its temperature is about 2°C higher than the coastal water temperature. The circulation at model levels 08 through 10 (Figure 4-6b) reproduces the circulation patterns encountered at levels 04 and 07 with a marked seasonal shift. The cycle divides the whole area in two separate domains; September-March, the circulation is dominated by a southward flow north of Lüderitz (current 3). This flow develops along the coast as well and the northward current (branch 2) disappears. The southernmost extent of this flow occurs in February (26°S) and in October (27.5°S) with maximum speeds of about 10 cm s-1. From April to August, the situation is somewhat reversed. There is no more poleward flow. The cold northward current intensifies in the south (branch 1) with speeds of about 5 cm s-1. It reaches the Lüderitz area and its intensity north of 26°S remains very weak. At this depth, the Lüderitz area displays a natural border between two opposing seasonal regimes, a northern one associated with warmer and more saline waters flowing southward from October-March, a southern one concerning cold and fresher waters flowing northward from April/May-August. The cold pool shown on Figure 46b underlines the fact that along the coast the Lüderitz area constitutes a transition between the warm and the cold regimes. Temperature in the cold pool is about 1011°C whereas it is about 13°C north of 26°S. The cold pool appears to be a permanent feature with quite regular shape and size all year long.
Large scale physical variability of the Benguela Current LME
63
Figure 4-6. Schematic circulation reproduced by the CLIPPER model at 80m (a) and at 130 m (b) depths.
Figure 4-7. Schematic circulation reproduced by the CLIPPER model at 230m (a) and at 350 m (b) depths.
Figure 4-8. Schematic circulation reproduced by the CLIPPER model at 575m depth.
64
F.A. Shillington et al.
Circulation at 230m and 350m depths (lev13 and lev15) At deeper model levels, the circulation along the shelf is poleward almost all year long with maxima occurring in February and October. Speeds are of the order of a few centimetres per second. The two diagrams (Figure 4-7) illustrate the model circulation at 230m and 350m depths. The poleward current brings warmer waters south of the Lüderitz area until 28°S where it reaches the cold pool. At this depth this permanent feature is less developed and its temperature is about 2°C lower than the poleward flow temperature. The northward current along the shelf (branch 1) develops in winter with very low speeds. Once again the current represented by the branch 3 does not seem to interact with it. Circulation at 570m depth (lev17) The circulation is dominated by the permanent northwestward flow associated with relatively warmer water masses (Figure 4-8, red arrow). Along the coast, a poleward coastal current develops from July to September during the winter period. Its maximum southward extent occurs in August. It meets the northward dominant flow near 30 and retroflects northward. In terms of temperature the whole area can be divided in two persistent parts; north of the warm current (red arrow) the water is about 1.5°C colder compared to the south (respectively 5.5°C and 7°C). The temperature variability throughout the year is small.
DISCUSSION AND CONCLUSIONS Processes with forecasting potential (see also Monteiro and Van Der Plas , this Volume – Chapter 5) Good progress has been made recently in the study of the Benguela Current Large Marine Ecosystem, as evidenced by the reviews in part two of this volume. In particular, considerable effort has been invested in trying to understand the mechanisms underlying the formation and evolution Benguela Niños (Florenchie et al. 2003; Florenchie et al. 2004). From our present understanding, by using appropriate observing systems in the equatorial region, it may be possible to get a forecast lead time of about two months for major warm events arriving at, and progressing polewards beyond the Angola Benguela Frontal zone. It is expected that aspects of the large scale variability of the BCLME are likely to be amenable to near real time observation and/or short term forecast. The most likely processes that have been identified to have potential in an early warning/forecasting system in the BCLME are set out in Table 4-1. The table is divided into an area (Domain), the most important forcing component, the main type of forcing process responsible for the variability, the approximate time scale of variability, the potential for being able to observe the phenomenon in near real time, and the subjective forecast/early warning potential with present limited capacity and resources. A three point scale: poor, fair and good is used. The scheme notes that anomalous signals propagate both from the equatorial Atlantic Ocean and into the northern BCLME, and
Large scale physical variability of the Benguela Current LME
65
from the Agulhas Current in the Indian Ocean, into the southern BCLME. The discussion starts with the remote wind forcing in the western equatorial Atlantic Ocean, and its likely effects on the northern part of the BCLME (Angola and Namibia), via warm and cool anomalous signals propagating southwards along the Angolan coast. Severe warm SST anomalies at or south of the ABFZ are classed as Benguela Niños, the last well documented one occurring in 1995 (e.g. Florenchie et al. 2003). The main variability influence on the southern BCLME is from Agulhas Current ring shedding, early retroflection and intrusions of subantarctic cold water. The most difficult processes to forecast are the local influences on the upwelling centres at relatively short time scales of days-months. It is vital for the sustainable management of the BCLME, that extreme events (e.g. Roy et al. 2001) are recognised and understood, and if possible, forecast with a reasonable lead time. A regular state of the environment (SOE) reporting system, together with better communication for the BCLME would improve BCLME management advice. NUMERICAL MODELLING OF THE PHYSICAL PROCESSES IN THE BCLME In the past five years, there has been a sharp increase in the hydrodynamic modelling of the southern Benguela Upwelling ecosystem by implementing the 3-D ROMS numerical code and using seasonal wind forcing, (Penven et al. 2001a; Penven et al., 2001b) and then by refining the wind forcing with realistic winds from ERS (Blanke et al. 2002). The influence of the Agulhas Current shear edge instabilities on the southern border of the BCLME has been partially addressed (Lutjeharms et al. 2003). A ten-year model run with a time resolution of two days, and variable horizontal grid spacing from 9-18 km has provided the community with output for use of a number of individual based model (IBM) configurations (Field and Shillington 2005). With the advent of the BCLME, a dedicated group is presently modelling both the large scale influences on the BCLME, and using a nested approach to gain a better understanding of the local variability. Good synergy is maintained between the BCLME project and the IRD Upwelling Ecosystems project which is undertaking a comparative study of the Benguela, Canary and Humboldt Upwelling Systems. Project: SAfE (Southern Africa Experiment) Around the Southern African coasts, several different questions can be posed to the numerical ocean modeller. For example, how do the Benguela Niños propagate into the BCLME? Or, what is the role of Mozambique channel eddies in the shedding of Agulhas rings into the Atlantic Ocean, and their subsequent interaction with the BCLME? Or, why is there a cool ridge on the Agulhas Bank? To address each of these questions, the modeller needs a high degree of spatial resolution in the model region of interest, as well as a correct representation of the large scale ocean dynamics. To do this, a modelling platform under the auspices of the BCLME has been set up for
66
F.A. Shillington et al.
Table 4-1. Processes that are likely to have a cost effective observational capacity (satellite remote sensing or large scale in situ measurements). The forecasting potential is judged on a subjective scale of poor, fair and good. The forecasting potential of the large scale BCLME variability depends mainly on how well the linkages and processes that transfer the equatorial signals, and those from the Agulhas Retroflection, to the Benguela are understood.
Domain
Remote
Forcing System
Eastern Tropical South Atlantic
Processes
Scales of variability
Observing Potential
Forecast potential
Equatorial upwelling
Seasonal interannual
-
Good: Altimetry, Ocean Colour
Good
Intensity and timing of trade winds
Seasonal Interannual
-
Quikscat, PIRATA, GCM
Good
Equatorial stratification
Interannual – decadal (Benguela Niño)
Good: Ocean Buoys
Fair
Angola Current
Seasonal interannual
Fair: Altimetry and AVHRR
Poor
AngolaBenguela Frontal Zone
Twice annual
Good: SST, colour
Good
Good: Altimetry, SST
Good
Fair: Ocean Buoy
Fair
-
Remote
Agulhas Retroflection
Ring shedding
Episodic: few times per annum
Local
Upwelling centres
Benguela Poleward transport
Seasonal Interannual
Upwelling wind variability
Days - weeks
Good: wind forecasts
Fair
Relaxation events in the southern Benguela
Days - weeks
Good: SST, colour
Fair
-
the simulation of the ocean around Southern Africa (SAfE: Southern Africa Experiment). The model is based on ROMS and takes advantage of its nesting capabilities. The parent grid includes the ocean around Southern Africa at a reasonable resolution (i.e. ~20-25 km). Several levels of child grids can be embedded into the parent grid, to reach locally a resolution of a few kilometers to a few hundred meters (see for example the grid set up for the ABFZ in Figure 4-4. The Parent model is inexpensive to run: 30 hours of computing for 1 year of simulation on a PC
Large scale physical variability of the Benguela Current LME
67
workstation. Hence, it is possible to rapidly test new configurations and developments. Once the parent solution is satisfactory in its representation of the large scale solution, the presently one way nested high resolution “child” model configurations are added to provide the fine scale information. These simulations will be coupled to biogeochemical models (Monteiro 2005, pers. Com) in one of the BCLME projects. ACKNOWLEDGEMENTS The first author (FAS) acknowledges ongoing support funding from the NRF and UCT, while PF acknowledges funding from the BCLME project. PP has been seconded by the IRD (France) to Cape Town for the period 2004-2006. REFERENCES Ajao, E.A. and R.W. Houghton. 1998. Coastal ocean of Equatorial West Africa from 10°N to 10°S. 605631 in Robinson, A.R. and K.H. Brink, eds. The Sea, Vol. 11, The global coastal ocean, regional studies and syntheses Wiley, New-York. Barnier, B., P. Marchesiello, A.P. De Miranda, J.M. Molines, and M. Coulibaly. 1998. A sigmacoordinate primitive equation model for studying the circulation in the south Atlantic. Part I: Model configuration with error estimates.Deep-Sea Res. Part I, 45:543-572. Biastoch, A. and W. Krauß. 1999. The role of mesoscale eddies in the source regions of the Agulhas Current, J. Phys. Oceanogr. 29:2303-2317. Blanke, B., C. Roy, P. Penven, S. Speich, J. McWilliams, and G. Nelson. 2002. Linking wind and upwelling interannual variability in a regional model of the southern Benguela. Geophys. Res. Lett., 29, 2188, doi :10.1029/2002GL015718. Boyd, A.J. and F.A. Shillington. 1994. The Agulhas Bank: A review of the physical processes. S. Afr. J. Sci. 90:114-122. Boyer, D., J. Cole , and C. Bartholomae. 2000. Southwestern Africa: Northern Benguela Current Region. In Sheppard, C.R.C., ed. Seas at the Millenium: An environmental evaluation, Vol 1, Elsevier Science Ltd. 821-840. Brundrit, G. B., B.A. De Cuevas, and A.M. Shipley. 1987. Long-term sea-level variability in the eastern south Atlantic and comparison with that in the eastern Pacific, S. Afr. J. Mar. Sci. 5:73–78. Chapman, P. and L.V. Shannon. 1985. The Benguela Ecosystem Part II. Chemistry and related processes. Oceanogr. Mar. Biol. Ann. Rev. 23:183-251. Chelton, D.B., M.G. Schlax, M.H. Freilich, and R.F. Milliff. 2004. Satellite Measurements Reveal Persistent Short-Scale Features in Ocean Winds. Science, 303, Issue 5660: 978-983. Colberg, F., C.J.C. Reason, and K. Rodgers. 2004. South Atlantic Response to El Nino-Southern Oscillation induced Climate Variability in an Ocean General Circu;lation Model. J. Geophys. Res.,. 109, C12015, doi:10.1029/2004JC002301. Demarcq, H., R. Barlow, and F. A. Shillington. 2003. Climatology and variability of sea surface temperature and surface chlorophyll in the Benguela and Agulhas ecosystems as observed by satellite imagery. Afr. J. mar. Sci. 25: 363-372. Duncombe Rae, C.M., F.A. Shillington, J.J. Agenbag, J. Taunton-Clark and M. L. Grundlingh. 1992. An Agulhas ring in the South Atlantic Ocean and its interaction with the Benguela upwelling frontal system. Deep-Sea Research 39, 2009-2027. Duncombe Rae, C.M. 1998. Antarctic Intermediate and Central Waters in the Angola-Benguela Front region: results from the first BENEFIT cruise, April 1997. In: International Symposium on Environmental Variability in the South-East Atlantic, 30 March to 1 April 1998, Swakopmund, Namibia. p.14. Duncombe Rae, C.M. 2005. A demonstration of the hydrographic partition of the Benguela upwelling ecosystem at 26 40'.S .African Journal of Marine Science 27(3): 617-628.
68
F.A. Shillington et al.
Fauchereau, N., S. Trzaska, M. Rouault, and Y. Richard. 2003. Rainfall variability and changes in southern Africa during the 20th century in the global warming context. Natural Hazards 29: 139-154. Field, J.G. and F.A. Shillington. 2005. Variability of the Benguela Current System. 833-860 In Robinson, A.R. and K.H. Brink, eds. The Sea, Vol. 14, The global coastal ocean, Interdisciplinary regional studies and syntheses. Harvard University Press. Florenchie, P. and J. Verron. 1998. South Atlantic Ocean circulation: Simulation experiments with a quasi-geostrophic model and assimilation of Topex/Poseidon and ERS1 altimeter data, J. Geo. Res., Vol 103, NO. C11, 24737-24758. Florenchie, P., C.J.C. Reason, J.R.E. Lutjeharms, M. Rouault, C. Roy, and S. Masson. 2004. Evolution of Interannual Warm and Cold Events in the Southeast Atlantic Ocean. J. Climate 17: 2318-2334. Florenchie, P., J.R.E. Lutjeharms, C.J.C. Reason, S. Masson, and M. Rouault. 2003. The source of Benguela Niños in the South Atlantic Ocean, Geophys. Res. Lett. 30, (10), 1505, doi:10.1029/2003GL017172. Gammelsrød, T., C.H. Batholomae, D.C. Boyer, V.L.L. Filipe, and M.J.O’Toole. 1998. Intrusion of warm surface water along the Angolan-Namibian coast in February – March 1995: The 1995 Benguela Niño, S. Afr. J. Mar. Sci. 19: 51– 56. Gründlingh, M.L. 1985. Occurrence of Red Sea water in the south western Indian Ocean, 1981. J. phys. Oceanog. 15(2): 207-212. Hardman-Mountford, N.J., A.J. Richardson, J.J Agenbag, E. Hagen, L. Nykjaer, F.A. Shillington and C. Villacastin. 2003. Ocean Climate of the South East Atlantic observed from satellite data and wind models. Progress in Oceanography 59: 181-221. Hermes, J.C., and C.J.C. Reason. 2005. Ocean model diagnosis of interannual co-evolving SST variability in the South Indian and Atlantic Oceans. J. Climate 18: 2864-2882. Hill, A.E., B.M. Hickey, F.A. Shillington, P.T. Strub, K.H. Brink, E.D. Barton and A.C. Thomas. 1998. Eastern Ocean Boundaries. 29-68 in Robinson A.R. and K.H. Brink, eds. The Sea, Vol 11,The Global Coastal Ocean, Regional Studies and Syntheses. John Wiley and Sons, New York. Huggett, J., P. Freon, C. Mullon, and Penven. 2003. Modelling the transport success of anchovy (Engraulis encrasicolus) eggs and larvae in the southern Benguela: The effect of spatio-temporal spawning patterns. Marine Ecology Progress Series 250: 247-262. Hutchings, L., L.E. Beckley, M.H. Griffiths, M.J. Roberts, S. Sundby and C. van der Lingen. 2002. Spawning on the edge: spawning grounds and nursery areas around the southern African coastline. Mar. Freshwater Res. 53: 307–318. Jones, P.D. and R.J. Allan. 1998: Climate change and long-term climatic variability. In Karoly, D. and D. Vintcent, eds. Meteorology of the Southern Hemisphere. Amer. Meteorol. Soc., Boston, Massachussetts. 337-363. Kearns, E.J. and M-E. Carr. 2003. Seasonal climatologies of nutrients and hydrographicproperties on quasi-neutral surfaces for four coastal upwelling systems. Deep-Sea Research, II, 50, 3171-3197. Kidson, J.W. 1988. Interannual variations in the Southern Hemisphere circulation. J. Climate 1:11771198. Lutjeharms J.R.E., P. Penven and C. Roy. 2003. Modelling the shear edge eddies of the southern Agulhas Current. Continental Shelf Research 23:1099-1115. Lutjeharms, J.R.E., and D.J. Webb. 1995. Modelling the Agulhas Current system with FRAM (Fine Resolution Antarctic Model), Deep Sea Res. Part I, 42: 23-551. Masumoto, Y. 2004. Generation of Small Meanders of the Kuroshio South of Kyushu in a HighResolution Ocean General Circulation Model. Journal of Oceanography 60:13-320. Matano, R.P. and E.J. Beier. 2003. A kinematic analysis of the Indian/Atlantic interocean exchange, Deep-Sea Res. Part II, 50:229-249. Mercier, H., M. Arhan and J.R.E. Lutjeharms. 2003. Upper-layer circulation in the eastern Equatorial and South Atlantic Ocean in January-March 1995. Deep-Sea Res. I 50: 863-887. Mo, K.C. and J.N. Paegle. 2001. The Pacific-South American modes and their downstream effects. Int. J. Climatology 21(10): 1211-1229. Mohrholz, V., M. Schmidt, and J.R.E. Lutjeharms. 2001. The hydrography and dynamics of the AngolaBenguela Frontal Zone and environment in April 1999. S. Afr. J. Sci. 97:199-208. Monteiro, P.M.S. 1996. The oceanography, the biogeochemistry and the fluxes of carbon dioxide in the Benguela upwelling system. Ph.D. thesis, Univ.Cape Town, S. Afr, 354pp. Mullon, C., P. Cury, and P. Penven. 2002. Evolutionary individual-based model for the recruitment of the anchovy in the southern Benguela. Can. J. Fish. Aquat. Sci. 59: 910-922.
Large scale physical variability of the Benguela Current LME
69
Parada, C., C.D. Van der Lingen, C. Mullon and P. Penven. 2003. Modelling the effect of buoyancy on the transport of anchovy (Engraulis capensis) eggs from spawning to nursery grounds in the southern Benguela: An IBM approach. Fish. Oceanogr. 12:170-184. Penven, P., C. Roy, A. Colin de Verdiere and J. Largier. 2000. Simulation and quantification of a coastal jet retention process using a barotropic model. Oceanol. Acta 23:615-634. Penven, P., C. Roy, G.B. Brundrit, A. Colin de Verdière, P. Fréon, A.S. Johnson, J.R.E. Lutjeharms, and F.A. Shillington. 2001a. A regional hydrodynamic model of the Southern Benguela upwelling. S. Afr. J. Sci. 97: 472-475. Penven, P, J.R.E. Lutjeharms, P. Marchesiello, C. Roy, and S.J. Weeks. 2001b. Generation of cyclonic eddies by the Agulhas Current in the lee of the Agulhas Bank. Geophys. Res.Lett. 27:1055-1058. Peterson, R.G. and L. Stramma. 1991: Upper-level circulation in the South Atlantic Ocean. Progress in Oceanography 26, 1-73. Poole, R. and M. Tomczak. 1999. Optimum multiparameter analysis ofthe water mass structure in the Atlantic Ocean thermocline. Deep-Sea Res. Part I, 46:1895-1921. Reason, C.J.C., J.R.E. Lutjeharms, J. Hermes, A. Biastoch. 2003. Inter-ocean fluxes south of Africa in an eddy-permitting model, Deep-Sea Research Part II, 50:281-298. Reason, C.J.C. and M.R. Jury. 1990. On the generation and propagation of the southern African coastal low. Quarterly Journal of the Royal Meteorological Society, vol. 116 (495): 1133-1151. Reason, C.J.C. 2000. Multidecadal climate variability in the subtropics/midlatitudes of the Southern Hemisphere oceans. Tellus, 52A: 203-223. Reid, J.L. 1989. On the total geostrophic circulation of the South Atlantic Ocean: Flow patterns, tracers, and transports. Prog. Oceanog. 23: 149-244. Rhodes, R.C., H.E. Hurlburt, A.J. Wallcraft, C.N. Barron, P.J. Martin, E.J. Metzger, J.F. Shriver, D.S. Ko, O.M. Smedstad, S.L. Cross and A.B. Kara. 2002. Navy real-time global modeling systems. Oceanogr. 15: 29-43. Risien, C.M., C.J.C Reason, F.A Shillington, and D.B. Chelton. 2004. Variability in satellite winds over the Benguela upwelling system during 1999–2000. J. Geophys. Res. 109: C3, C0301010.1029/2003JC001880. Rouault, M., P. Florenchie, N. Fauchereau and C.J.C. Reason. 2003. South East Atlantic warm events and southern African rainfall. Geophys. Res. Lett. 30 (5): 8009, doi:10.1029/2002GL014840. Roy, C., S.J. Weeks, M. Rouault, G. Nelson, R. Barlow, and C.D. van der Lingen. 2001. Extreme oceanographic events recorded in the southern Benguela during the 1999-2000 summer season. S. Afr. J. Sci. 97:465-471. Schumann, E.H. and K.H. Brink. 1990. Coastal-Trapped Waves off the Coast of South Africa: Generation, Propagation and Current Structures. Journal of Physical Oceanography 20(8): 1206–1218. Schumann, E.H. 1998. The coastal ocean off southeast Africa, including Madagascar. 557-581 in Robinson, A.R. and K.H. Brink, eds. The Sea, Vol. 11, The global coastal ocean, regional studies and syntheses. Wiley, New-York. Shannon, L.V, A.J. Boyd, G.B. Brundrit, and J. Taunton-Clark. 1986. On the existence of an El-Niño type phenomenon in the Benguela system. J. Mar. Res. 44(3): 495-520. Shannon, L.V. and G. Nelson. 1996. The Benguela: Large scale features and processes and system variability. 163-210 in Wefer, G. W.H. Berger, G. Siedler, and D.J. Webb, editors. The South Atlantic Past and Present Circulation. Springer Verlag, Berlin, Heidelberg. Shannon, L.V., and M.J. O’Toole. 2003. Sustainability of the Benguela: ex Africa semper aliquid novi. In: K. Sherman and G. Hempel, Large Marine Ecosystems of the World – Trends in Exploitation, Protection and Research. Elsevier B.V. 227-253. Shannon, L.V. and D. Hunter. 1988. Notes on Antarctic Intermediate Water around southern Africa. S. Afr. J. mar. Sci. 6:107-117. Shannon, L.V., J.J. Agenbag, M.E.L. Buys., 1987. Large and mesoscale features of the Angola-Benguela front. In: Payne, A.I.L., J.A. Gulland and K.H. Brink, eds. The Benguela and Comparable Ecosystems, S. Afr. J. mar. Sci. 5:11-34. Shillington, F.A. 1998. The Benguela upwelling system off southwestern Africa. 583-604 in Robinson , A.R.and K.H. Brink. The Sea, Vol. 11, The Global Coastal Ocean, Regional Studies and Syntheses. Wiley, New-York. Skogen, M.D. 1999. A biophysical model applied to the Benguela upwelling system. S. Afr. J.. Mar. Sci. 21: 235-249.
70
F.A. Shillington et al.
Speich, S., P. Penven and B. Blanke. 2004. On the Cape Cauldron dynamics: some physical insights on the turbulent Indo-Atlantic exchange and impact of Agulhas waters in the Southern Africa upwelling region from a hierarchy of regional numerical simulations. European Geosciences Union 2004, Geophysical Research Abstracts. 6:4815. Stenevik, E.K., M. Skogen, S. Sundby, and D. Boyer. 2003. The effect of vertical and horizontal distribution on retention of sardine (Sardinops sagax) larvae in the Northern Benguela - observations and modelling. Fish. Oceanogr. 12(3): 185-200. Talley, L.D. 1996. Antarctic Intermediate Water in the South Atlantic. In G. Wefer, W.H.Berger, G. Siedler, and D.J. Webb, eds. The South Atlantic: Present and Past Circulation. Berlin Heidelberg: Springer-Verlag, pp. 219-238. Treguier, A.M., O. Boebel, B. Barnier, and G. Madec. 2003. Agulhas eddy fluxes in a 1/6 degree Atlantic model. Deep-Sea Res., Part II, 50: 251-280. Van Foreest, D. and G.B. Brundrit. 1982. A two mode numerical model with application to coastal upwelling, Prog. Oceanogr. 11:329-392. van Loon, H. 1967. The half-yearly oscillation in middle and high southern latitudes and the coreless winter. J. Atmos. Sci. 24: 472-486. Veitch, J.A., P. Florenchie and F.A. Shillington. 2006. Seasonal and interannual fluctuations of the Angola Benguela Frontal Zone (ABFZ) using 4.5 km resolution satellite imagery from 1982 to 1999. International Journal of Remote Sensing 27: 989-1000. Venegas, S.A., L.A. Mysak, and D.N. Straub. 1996. Evidence for interannual and interdecadal climate variability in the South Atlantic. Geophysical Research Letters 23(19): 2673-2676. Venegas, S.A, L.A. Mysak and D.N. Straub. 1997. Atmosphere-ocean coupled variability in the South Atlantic. J. Climate 10: 2904-2920. Venegas, S.A., L.A. Mysak and D.N. Straub. 1998. An interdecadal climate cycle in the South Atlantic and its links to other ocean basins. J. Geophys. Res. 103, No. C11, 24,723-24,736.
Large Marine Ecosystems, Vol. 14 V. Shannon, G. Hempel, P. Malanotte-Rizzoli, C. Moloney and J. Woods (Editors) © 2006 Elsevier B.V. All rights reserved.
71
5 Low Oxygen Water (LOW) Variability in the Benguela System: Key Processes and Forcing Scales Relevant to Forecasting Pedro M.S. Monteiro and Anja K. van der Plas INTRODUCTION Low oxygen water (LOW) is an endemic characteristic of the Benguela system (Chapman and Shannon 1985; Bailey 1991; Monteiro et al. 2004). The ecological impacts of LOW were identified in the early research work undertaken in the system (Copenhagen 1953; Pieterse and van der Post 1967) and its close association to the incidence of elevated sulphide concentrations was also noted in a qualitative sense (Marchand 1928; Copenhagen 1953; Hart and Currie 1960; Pieterse and van der Post 1967). Events that resulted in significant losses of both demersal and bottom species have occurred in both the central (Namibia) and southern (South Africa) Benguela system (see case study). At present, stock assessment models treat environmental factors as random sources of mortality that can be parameterised by a mortality factor - this assumes that there are no systematic shifts in the forcing and response to LOW. Similarly, ecosystem models are typically less than sensitive to environmental forcing which can impact fisheries and ecosystem behaviour, distribution and mortality (Shannon and Jarre-Teichmann 1999). The most recent time series data analysis supports the view that LOW variability is characterised by regime shifts in both remote forcing and local forcing factors which interact non-linearly to create LOW conditions and events of hitherto unpredicted magnitude (Monteiro et al. 2004). The BCLME Transboundary Diagnostic Analysis (TDA) identified LOW as one of the key environmental factors governing the variability and commercial viability of fisheries and ultimately the ecosystem (www.bclme.org). Its implementation plan requires that not only should the causes of LOW variability be understood but the BCLME should also invest in developing a forecasting capability which could assist the optimal ecosystem management, anticipate its impacts, provide better understanding of the underlying complexity and support fisheries management. The forecasting goal for LOW in the Benguela requires that the processes and the forcing scales that drive events and their variability be better characterized and understood.
72
P.M.S. Monteiro and A.K. van der Plas
The impact of LOW variability on hake fisheries: Namibia 1992 - 1994
Hake recruit mortalities off Namibia in 1992/3 (Woodhead et al. 1997a) and the hake recruitment failure off Walvis Bay in 1994 (Hamukuaya et al. 1998; Woodhead et al. 1997b) are examples of the decimating effect that LOW can have on the marine resources of the Benguela ecosystem. Although Cape hake Merluccius capensis have adapted both behaviourally and physiologically to tolerate hypoxic conditions to a degree (Woodhead et al. 1998) the severity and the prolonged duration of the hypoxic conditions over the central Benguela continental shelf between 1992 and 1994 is thought to have led to mass mortalities of the hake recruits during 1993 and 1994. In austral summer of 1992 to 1993 the juvenile hake were thought to have been trapped by the expansion of hypoxic conditions leading to loss of half the recruits (Woodhead et al. 1997a). During 1994 the juvenile Cape hake that did not succumb to the oxygendepleted waters sought to avoid the LOW offshore but cannibalism by the adults that frequent the deeper shelf waters as well as discarding by trawlers targeting these adults are thought to have lead to a recruit mortality of 70-84% (Hamukuaya et al. 1998; Woodhead et al. 1997b). Thus, at certain scales, LOW variability affects the abundance, distribution, availability and catchability of commercially fished stocks through modification of both behavioural and mortality responses. The non-random character of LOW impacts on fisheries also challenges the assumptions in fish stock assessment models of the relationship between mortality and environmental variability.
SYNTHESIS OF SYSTEM PROCESSES AND VARIABILITY The importance of an advection link between the tropical eastern Atlantic low oxygen reservoir and the Benguela was proposed by Moroshkin et al. (1970), and Bubnov (1972) who also suggested that the oceanic LOW reservoir was generated by productivity associated with the Angola Dome. This latter idea was challenged by the work of Voituriez and Herbland (1982) that pointed to the equatorial upwelling zone as the main source of production, which is consistent with remote sensing data (Monteiro and van der Plas 2004). The parallel models of remote forcing and local biological production as the main drivers for low oxygen variability were reviewed in detail by Chapman and Shannon 1985. Monteiro et al. (2004) have recently suggested that low oxygen variability in the Benguela system is forced by the interaction on varying scales of both large- (basin) and local- (shelf) scale forcing. The processes at these two scales form the core of the discussion. Furthermore, LOW variability in the Benguela system can be further divided into three physically characterised regimes: Northern (Angola): LOW variability is completely advection controlled and tightly coupled to upwelling that peaks in June – August.
Benguela low oxygen variability
73
Central (Namibia) LOW variability is governed by a complex interaction between the remotely forced shelf boundary conditions, seasonal thermocline variability and biogeochemical carbon fluxes. Southern (South Africa) LOW variability is largely driven by local seasonal wind characteristics and minimal remote forcing. This synthesis is based on the following two foundation papers: a review of BCLME LOW formation assessing the importance of both remote and local forcing for the Benguela region (Monteiro et al. 2004) and a paper on sediment vs. water column hypoxia coupling (van der Plas et al. 2005). As neither of these papers is yet published, the essence of the thinking in both is reflected in this synthesis. The physical oceanography focuses on processes that are directly relevant to the formation or advection of LOW. More general syntheses of large scale and shelf physical processes are provided elsewhere (Hardman-Mountford et al. 2003; Shillington et al. 2005). One of the key requirements of forecasting schemes is their ability to translate predicted LOW temporal and spatial characteristics into robust ecological risk categories. A revised set of categories using the most recent observational data is given in Table 5-1.
Table 5-1. Oxygen concentration thresholds that are of relevance to the linkages between predicted oxygen concentrations and their ecological consequences. These should be seen as guidelines to be interpreted more closely on a case by case basis because exposure times and frequencies are also relevant.
Oxygen State
Oxygen Concentrations
Impacts
Super Saturated
> 100% saturation
Out-gassing to the Atmosphere f (t,S): typical in high surface primary production
Saturated
100% Saturation
Equilibrium with the atmosphere f (t,S)
Under saturated
3 – 100% Saturation
Range over which biological responses should be insignificant
Depleted
2 - 3 ml l-1
Biological impacts felt at behavioural level
Critical Hypoxia
-1
1 – 2 ml l
Hypoxic
0.5 – 1 ml l-1
Anoxic
-1
< 0.5 ml l
Threshold that enables the system to go anoxic under a flux of bloom detritus. Organisms require physiological adaptation to survive Extreme stress and mortality in organisms. (denitrification) Respiration dominated by anaerobes and sulphide / methane fluxes
74
P.M.S. Monteiro and A.K. van der Plas
REMOTE FORCING: EASTERN TROPICAL SOUTHEAST ATLANTIC (ETSA – BENGUELA LINKAGE) (Further supporting information in CD-ROM: LOWCH5.htm)
The Eastern Tropical Southeast Atlantic (ETSA) region is recognised as the main reservoir of LOW in the region but its internal processes and its linkages to the Benguela are weakly understood (Chapman and Shannon, 1985; Voituriez and Herbland 1982). While the importance of temporal variability was understood early on (Chapman and Shannon 1987; Voituriez and Herbland 1982), limited progress has been made in understanding the processes that govern the scales of variability. The basin thermocline shallows eastwards and gets to within 50m of the surface in the ETSA zone, which is commonly referred to as the Angola gyre (Stramma and Schott 1999). It is here that occur the dominant processes of primary production, stratification and retention, which govern LOW formation, transport and ultimately the boundary conditions of the Benguela shelf. The following processes are essential to the formation and maintenance of the ETSA LOW reservoir: • The scale and variability of phytoplankton new production which provides the required electron donating capacity to the oxygen sink processes. • A thermocline that limits the downward flux of oxygen across the thermocline to below the biogeochemical uptake rate. • A retention zone that limits the rate of sub-thermocline ventilation by advected aerated water In order to understand the generation and change of LOW in the ETSA circulation zone it is essential to characterise the scales of spatial and temporal variability. The main features and flows of the Eastern Tropical South Atlantic (ETSA) – Benguela region that are relevant to LOW variability (Figure 5-1) are briefly noted below. The spatial and temporal characteristics of LOW oceanography in the tropical South Atlantic are governed by the cyclonic part of its circulation (Reid 1989). The core of LOW within the ETSA zone extends from the equatorial zone to two southern boundaries: one at 16 - 17°S and a second at 25 - 26°S (Figure 5-1). These two boundaries correspond to the southern edge of the Angola gyre and the southern edge of the sub-equatorial cyclonic circulation respectively. The sharp oxygen gradient across the latter boundary defines the transition between the two South Atlantic Central Water masses derived respectively from the hypoxic ETSA and the aerated Cape Basin (Figure 5-1). In the north, the equatorial divergence zone and its associated upwelled nutrient flux are driven by the seasonal easterly trade winds. This system supplies the main phytoplankton export production flux that creates the subthermocline oxygen demand within the Angola gyre. Although there are several intermittent divergent flow features here, such as the Angola dome, the only significant export flux of carbon is due to upwelling activity in the austral winter (Voituriez and Herbland 1982).
Benguela low oxygen variability
75
The current system in the region is fairly complex (Figure 5-1). The eastward flowing Equatorial Under Current (EUC) and South Equatorial Under Current (SEUC) converge, particularly during the seasonal (Dec - April) weakening in the trade winds, when the EUC intensifies to form the Guinea-Congo Under Current (GCUC; Stramma and Schott 1999). The combination of the GCUC and SEUC forms the southward flowing Angola Current (16 Sv), which is also the eastern boundary of the Angola gyre (Mohrholz et al. 2001; Mercier et al. 2003). The South Equatorial Counter Current (SECC) provides an additional inflow to the Angola Current during the austral winter.
Eastern Tropical South Atlantic System: ETSA
EUC
EDZ
GCUC
SEUC SECC
AC
sSEC
BPUC
Cape Basin SACW
Figure 5-1. A diagrammatic view of the main components of the Eastern Tropical South Atlantic System (ETSA) cyclonic circulation zone that are relevant to LOW variability in the Benguela. It shows the core cyclonic circulation, also known as the Angola gyre, supplied with three eastward flows, the Equatorial Under Current (EUC), South Equatorial UC (SEUC) and South Equatorial Counter Current (SECC). The eastern boundary comprises the seasonal Guinea-Congo UC (GCUC) (July – Sept), the southward coastal Angola Current (AC)(16 Sv), the Benguela southward extension as the Poleward Under Current (BPUC) (2 – 5 Sv) which defines the boundary conditions for the shelf upwelling system.
The Angola current splits into two flows, the main one (14 Sv) closing the Angola gyre while its southward extension becomes the Benguela Poleward Under Current (BPUC) along the Namibian shelf as far south as 27°S (Mercier et al. 2003). This southerly extension of ETSA-generated LOW establishes the boundary conditions for the northern and central Benguela system. The poleward undercurrent also feeds further south into the southern branch of the South Equatorial Current (SSEC), which
76
P.M.S. Monteiro and A.K. van der Plas
together with the Benguela Current, closes the basin scale cyclonic gyre of the South Atlantic thermocline waters (Stramma and England 1999). The BPUC also becomes the poleward undercurrent on the slope of the Benguela that acts as the main advective link for LOW between the ETSA and the Benguela systems. The closure of the Angola gyre by the main flow of the Angola current creates the recirculation retention zone that, together with its thermocline dynamics, establishes the conditions necessary for LOW formation. The main sources of ventilation for the sub-thermocline waters of the ETSA retention zone are the EUC, SEUC and SECC. These currents also may play an important role in the transport of the new production flux from the EDZ into the retention zone. Within the Angola gyre, the Angola dome is a seasonally transient feature with apparently only limited impact on low oxygen variability. Although previously thought to be the main source of divergent transport that supported phytoplankton production (Chapman and Shannon, 1985), its contribution to the overall oxygen demand is likely to be small compared to the upwelling at the equatorial divergence zone (EDZ). (Note that Chapman and Shannon did not consider the role of the EDZ in their paper.) At the southern end of the Angola gyre is a surface feature known as the AngolaBenguela Front (ABF). The spatial and intensity characteristics of the ABF are governed by the seasonal relaxation of the equatorial easterly winds in the late austral summer (Feb - April), which drives the eastward and southward propagation of warm surface water probably as a baroclinic Kelvin wave (Stramma and Schott 1999; Lass et al. 2000; Mohrholz et al. 2001). The relevance of this process to LOW variability in the Benguela is that the resulting intensification of the thermocline intensifies the poleward transport of LOW in the slope and on the shelf. Large perturbations of the Atlantic equatorial thermocline occur at approximately decadal intervals. These perturbations propagate eastward as a Kelvin wave and surface at the ABF. The effect is an anomalous warming of the surface layer, known as a Benguela Niño, that can then propagate onto the Namibian shelf (Florenchie et al. 2003). This process impacts LOW by intensifying the thermocline and increasing the poleward flow below it. Other associated but less well understood effects of the Benguela Niño include weakening of the equatorial thermocline and the EUC which impact on the ventilation of the ETSA and the ETSA Benguela linkage respectively. The Benguela Niño warming at the ABF should not be confused with the annual (late summer) warming that results from the seasonal relaxation of the equatorial easterly winds. These ETSA-derived features combine to drive the spatial and temporal characteristics of the LOW boundary conditions for the northern and central Benguela system, although not including the Lüderitz upwelling cell. The Lüderitz upwelling cell and the southern Benguela (28oS - 35oS) are defined by the SACW in the Cape Basin characterised by the well aerated boundary conditions driven from the sub-Polar domain.
Benguela low oxygen variability
77
BENGUELA SHELF VARIABILITY LOW variability in the Benguela (5oS – 35oS) has been separated into three recognizable domains according to the extent to which the variability is externally forced or locally generated. These are: • Northern Benguela: Congo – Angola sub-system • Central Benguela: Namibian sub-system • Southern Benguela: South African sub-system Whereas in the northern sub-system a narrow shelf results in a spatially extensive upwelling, in the central and southern sub-systems the slope – shelf link is at discrete sites also termed “gates” (Monteiro, 1996; Duncombe-Rae, 2004). Three main upwelling “gates” have been suggested to govern the slope – shelf exchange of SACW, Cape Frio (17 – 18°S), Lüderitz (25 – 26°S) and Oliphants Valley (33°S) (Monteiro 1996). Meridional and vertical shifts in the ETSA derived LOW core control the oxygen characteristics of upwelled water at the Lüderitz and the Cape Frio upwelling centres, the main slope – shelf exchange “gates” in the central Benguela system (Monteiro 1996; Duncombe-Rae 2004). Influxes of more oxygenated SACW at the Oliphants Valley zone will shift the southern Benguela shelf system away from hypoxic conditions even under intense upwelling derived new production fluxes. Northern Benguela: Congo – Angola sub-system The temporal variability of LOW in the narrow Congo – Angola shelf system (Figure 5-2) shows that it is strongly driven by the boundary conditions characteristic of the ETSA region along the shelf. The narrow shelf means that ETSA- LOW is upwelled from the slope onto the shelf along the entire coastal system and LOW seasonal variability is strongly correlated to temperature. This correlation indicates that the variability is governed by the advection of upwelled water rather than by any shelf domain processes. Seasonal variability in the northern sector of the system (Figure 52) shows that LOW intensifies in the 3rd and 4th quarter of the year linked to the intensification of the equatorial easterlies. Moreover, higher oxygen conditions of the system (Figure 5-2) are driven by downwelling and aeration linked to the southward advection of tropical warm water during the relaxation of the equatorial easterlies in the 1st quarter. For the remainder of the year the combined effects of the narrow shelf and proximity to the domed core of the ETSA LOW system mean that the oxygen concentrations are low and closely correlated to temperature (Figure 5-2). Central Benguela: Namibian sub-system LOW variability in the Central Benguela shelf is governed primarily by the boundary characteristics at the two main upwelling centres of Cape Frio and Lüderitz. The linkage between these boundaries and the ETSA is of key importance. We suggest that the relative contribution from these two sources of shelf water is strongly
P.M.S. Monteiro and A.K. van der Plas
78
Dep th (m)
Lobito (12°S, 110m to 150m bottom depth) 0 -50 -100 1995
a)
1996
1997
1998
1999
2000
2001
2002
2003
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 °C
De pth (m)
0 -50 -100 1996
1995
1996
b) Depth (m)
1997
1998
1999
2000
2001
2002
2003
2002
2003
36 PSU 35. 9 35 . 8 35. 7 35. 6 35. 5 35 . 4 35. 3 35. 2 35. 1 35 34 . 9 34 . 8 34. 7 34. 6 34 . 5 34 . 4 34. 3 34 . 2 34 . 1 34 33 . 8 33. 6 33 . 4 33. 2 33
1995
0 -50 -100
c)
1997
0
0.5
1998
1
2
1999
3
4
2000
5
6
2001
7 ml/l
Figure 5-2. Time series of temperature (a), salinity (b) and oxygen (c) variability on the Angolan shelf for the period 1994 – 2003. It highlights the strong relationship between the incidence of low oxygen waters (< 2mll-1) and cold upwelled water (< 16oC). Because of the narrow shelf the incidence of LOW is driven almost completely by the upwelling driven advection of ETSA LOW rather than any shelf based modification. In this part of the system oxygen behaves conservatively with temperature. Sampling periods and depths are indicated on the diagrams.
dependent on the characteristics and the poleward extent of the warm tropical surface water and the impact it has on the thermocline characteristics on the shelf. As stated above, the poleward extent of the warm tropical surface water governs the strength of the sub-thermocline poleward flow which regulates the spatial scale of the impact of the hypoxic waters upwelled at Cape Frio. Under conditions of weak stratification and south easterly wind stress, typical of the early upwelling season in the 3rd and 4th quarters of the year, the dominant flow on the Namibian shelf is equatorward, driven by the barotropic pressure gradient and a weak or non existent poleward flow on the shelf. When stratification intensifies, either as a result of
Benguela low oxygen variability
79
seasonal or interannual warm events, the sub-thermocline poleward flow strengthens due to the increasing forcing of the baroclinic pressure gradient. While the former condition favours a larger contribution of mostly aerated water (Cape Basin SACW) derived from the Lüderitz gate, the latter favours a greater magnitude from the hypoxic Cape Frio flux. Thus, we believe that the LOW environment on the Namibian shelf is modulated by the changing contributions of water from the two input fluxes driven indirectly by the strength of the warm water events. This dynamic is suggested to govern the magnitude of both the seasonal and the interannual LOW signal in this part of the system. Combining these ideas of stratification and shelf transport allows LOW variability over a 10 year period (1995 – 2004) within the central part of the Central Benguela to be better understood (Figure 5-3a-c). The time series of oxygen concentration at the outer shelf in the mid-Central Benguela (Figure 5-3) shows that the variability of the hypoxic water is driven by both the stratification as well as the LOW boundary conditions, with the strength of the stratification, which according to the model drives the poleward transport, modulating the boundary condition LOW signal on the shelf. In periods when the stratification weakens the hypoxic signal is also weakened because there is a greater contribution from water upwelled at Lüderitz and moving equatorward. This happens every year in the winter – spring upwelling period and occasionally, such as in 1997 – 1998, it covers an interannual scale when stratification remains weak and water column oxygen concentrations are relatively higher (< 2ml l-1 ; Figure 5-3). In this period salinities remained low, supporting the prediction that the system would under these conditions have a stronger forcing from Lüderitz. Salinities then increase as predicted from the result of the increasing contribution from the Cape Frio upwelling centre. The data shows that there are consistent differences in oxygen content between the inner and outer shelf areas of the central Benguela. The inner shelf concentrations are consistently lower. Differences in LOW variability between the inner and outer shelf zones are due to the lag effect caused by the biogeochemical oxygen demand driven by the respiration rates in the inner shelf mud belt where much of the surface derived new production accumulates (Monteiro et al. 2005; Monteiro and Roychoudhury 2005). The sediment ecosystem in the mud belt can exist in two redox states, aerobic and anaerobic. Both states create oxygen demand fluxes but whereas in the aerobic condition this is directly related to the metabolism of the flux of organic carbon, in the anaerobic condition it includes also the additional oxygen demand fluxes driven by reduced metabolic products such as HS-, CH4 and NH4 +. Once the system switches to the anaerobic condition, the lagged flux of reduced products driven by accumulated organic carbon maintains an oxygen sink that increases the persistence of hypoxic / anoxic conditions. The lag in the consumption of the electron donors as well as the flux of reduced products damp the variability in the inner shelf region of Namibia. However, it is not the upwelling-derived flux of organic carbon that governs the shift from aerobic to anaerobic conditions, but the boundary derived LOW signal of O2 < 1.5ml l-1. If this condition is not achieved, either because of boundary conditions at Cape Frio or an increased contribution from Lüderitz, the anaerobic fluxes weaken and
80
P.M.S. Monteiro and A.K. van der Plas
the system will after one or two seasons switch to an aerobic state ( e.g. 1997-1998). Despite the lag effect of the locally forced anaerobic conditions, LOW variability is still characterised by a seasonality where water column hypoxia is deepened in the later summer – autumn period and weakened in the winter – early spring period (Figure 5-3). Walvis Bay (23°S, 320m bottom depth)
Depth (m)
0 -100 -200 -300
1995
a)
1996 5
1997 6
7
8
1998
1999
2000
2001
2002
2003
2004
9 10 11 12 13 14 15 16 17 18 19 20 21 °C
Depth (m)
0 -100 -200 -300
1995
1997
1998
1999
2000
2001
2002
2003
2004
36 PSU 35.9 35. 8 35.7 35.6 35.5 35. 4 35 .3 35 .2 35 .1 35 34. 9 34.8 34 .7 34.6 34. 5 34.4 34.3 34.2 34. 1 34 33.8 33.6 33. 4 33.2 33
b)
1996
Depth (m)
0 -100 -200 -300 1995
c)
1996
1997 0
0.5
1998 1
2
1999 3
4
2000 5
6
2001
2002
2003
2004
7 ml/l
Figure 5-3. Variability of temperature (a), salinity (b) and oxygen (c) at an outer shelf location at 23oS in the Central Benguela between 1994 and 2004. The oxygen variability is modulated by both seasonal (summer / late summer) and interannual (1996 – 1999 vs 2000 – 2002) scales. The significant point is to link the period of enhanced LOW (2000 – 2002) to increased surface warming and higher salinities. In contrast, the 1996 – 1999 periods reflect weaker hypoxia. Sampling periods and depths are indicated on the diagrams.
The importance of this finding is that it supports the view that the toxic events driven by methane and sulphide are a response to boundary forcing rather than a forcing factor. The unexpected aspect is just how weak the local generation signal really is. It
Benguela low oxygen variability
81
controls persistence, intensity as measured by water column depth, and toxicity to local fauna but not the incidence. In summary, while LOW variability in the outer shelf is governed by both the boundary conditions and dynamic interaction of fluxes between the Lüderitz and Cape Frio upwelling centres, the variability in the inner shelf is the result of the same factors as well as the local biogeochemical processes. Measured oxygen concentrations reflect the spatially separated inputs from the Cape Frio and Lüderitz upwelling cells as well as the poleward transport of warm tropical surface water which exerts its impact through the baroclinic pressure gradient. Southern Benguela: South African sub-system In contrast to both the central and northern sub-systems, LOW variability in the southern sector (e.g. see Figure 5-4) is largely governed by a combination of local physical (stratification, recirculation-retention and advection) and biogeochemical processes (upwelling driven new production). Moreover, both northern and central sub-systems have shelf boundary conditions characterised by ETSA-derived LOW whereas the boundary conditions in the southern sub-system are those of aerated subAntarctic SACW (O2 > 4ml l-1) - see Chapman and Shannon 1987. Therefore, rather than being “primed” with remote sourced LOW, local formation has to rely on the physics of retention and stratification to bring down the oxygen concentrations of newly upwelled water. This is, in principle, the same set of processes that govern the ETSA zone on a larger spatial scale. The main LOW generation zone is the St Helena Bay retention zone (31 - 33oS) downwind from the Cape Columbine upwelling centre (Bailey and Chapman 1985; Penven et al. 2000). The hydrodynamics of this system drive a seasonal cyclonic circulation that gives rise to a strongly stratified two layer system sustained with cold upwelled water and a sun-warmed surface layer (Waldron and Probyn 1991). These conditions persist over the upwelling season (September – April) and support a highly productive nitrate-driven biological pump (Touratier et al. 2003; Monteiro et al. 2005; Monteiro and Roychoudhury 2005), which coupled to the physically driven nutrient fluxes, lead to high rates of sedimentation of POC (Bailey 1983). The remineralization of the POC coupled to fluxes of HS- creates an environment where, with strong stratification that reduces the aeration rates of sub-thermocline waters, the seasonal LOW is generated. The detailed interactive dynamics that govern LOW variability and which form the basis to a possible forecasting system are described in greater detail in the section dealing with LOW forecasting scales (see Chapter 13 this volume). LOW variability in the remainder of the southern Benguela sub-system shelf is the result of equatorward advection of LOW formed in St Helena Bay. The relatively low salinity values (S < 34.9) over a decade-long time series support the view that ETSA waters do not make a significant contribution to the water and LOW in the southern Benguela. The northward transport is depicted in the distribution of the integrated surface chlorophyll from St Helena Bay (see composite images in Figures 7-2 and 7-4
82
P.M.S. Monteiro and A.K. van der Plas
De pth (m)
Cape Columbine (100m to 150m bottom depth) 0 -50 -100 1959
1960
1961
1962
1963
1964
1965
1966
1967
Dep th (m)
0
-50 -100 -150 1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
De pth (m)
0 -50 -100 -150 1994
1995
1996
5
6
7
1998
8
1999
2000
2001
2002
2003
9 10 11 12 13 14 15 16 17 18 19 20 21 °C
0 -50
36 PSU 35.9 35 .8 35.7 35 .6 35.5 35 .4 35.3 35 .2 35.1 35 34.9 34 .8 34.7 34 .6 34.5 34 .4 34.3 34 .2 34.1 34 33.8 33 .6 33.4 33 .2 33
Dep th (m)
a)
1997
-100
1959
1960
1961
Dep th (m)
0 -50 -100 -150
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1993
1994
Depth (m)
0 -50 -100 -150 1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
De pth (m)
b) 0 -50 -100
1959
1960
1961
1962
1963
1964
1965
1966
1967
De pth (m)
0 -50 -100 -150
1984
1985
1986
1987
1988
1989
1990
1991
1992
De pth (m)
0 -50 -100 -150 1994
c)
1995
1996
1997
0
1998
0.5
1
1999
2
3
4
2000
5
6
2001
2002
2003
7 ml/l
Figure 5-4. The variability of temperature (a), salinity (b) and oxygen (c) at a mid-shelf position in the period 1984 – 2004. It shows a remarkable contrast in oxygen regimes between the 1980’s (aerated) and the 1990’s which were oxygen deficient / hypoxic. The explanation lies in the quasi-decadal scale changes in the upwelling wind regimes. The 1980s were characterised by relatively weak winds whereas the 1990’s by strong upwelling conditions. Hypoxia is related to changes in the retention characteristics of the St Helena Bay retention area. temperature. Its variability is driven exclusively by the variability of the ETSA characteristics
Benguela low oxygen variability
83
in Pitcher and Weeks – Chapter 7 this volume). LOW variability off Hondeklip Bay (30oS) is closely correlated to temperature (Monteiro et al. 2004) which is an expected outcome from advection-controlled variability. We agree with Johnson and Nelson (1999) that interannual LOW variability in the southern Benguela is governed mainly by the interannual variability in the equatorward component of the seasonal upwelling winds. This is in contrast to the controls on the boundary conditions of the northern and central Benguela system that are exerted by the seasonal, interannual and decadal shifts in the easterly equatorward winds. Summary of characteristics of LOW variability The characteristics of LOW variability in the Benguela can be summarised into three modes: Northern Benguela: The Angolan shelf system is directly coupled to the boundary conditions and the variability in LOW is largely predicted by its strong correlation to temperature. Its variability is driven exclusively by the variability of the ETSA characteristics. Central Benguela: LOW variability on the Namibian shelf is non-linear in respect of upwelling because it is dependent on a conjunction of processes and conditions that are not directly linked. The factors that govern LOW variability on the Namibian shelf are thought to be ETSA characteristics that set the boundary conditions, the incidence and strength of warm surface events, and upwelling rates at both Cape Frio and Lüderitz. These are amplified by the local production fluxes in the inner shelf. Southern Benguela: LOW variability in the Southern Benguela is largely governed by the interannual variability in the equatorward component of the seasonal upwelling winds. The importance of this characterisation is that it helps to define key scales of forcing and response that are sub-system specific and perhaps result in a more sensitive forecasting or at least predictive system.
PROCESSES REQUIRING DIAGNOSTIC ASSESSMENT The recently completed review identified a number of new possible processes that may govern low oxygen variability over a wide range of space and time scales (Monteiro et al., 2004). While these new proposed explanations were consistent with the data sets used in the review it is not certain whether the dynamics proposed to account for their impact on LOW variability are consistent.
84
P.M.S. Monteiro and A.K. van der Plas
The most important process uncertainties driving remote forcing are: • The combination of processes that govern the formation and variability of LOW in the ETSA zone • The coupling of warm surface flow and sub-thermocline poleward flow which may be the mechanism that transports LOW from the ETSA to the Benguela boundaries • The Slope – Shelf coupling which transports LOW onto the shelf at preferential sites such as the Cape Frio or Lüderitz upwelling centres • The coupling between warm surface flow and sub thermocline transport on the shelf through the strengthening of the thermocline • Coupling between remote and local forcing The dynamic consistency of these proposed mechanisms needs to be evaluated using appropriately set up hydrodynamic models through a set of modelling experiments. These do not need to be undertaken in simulation mode but in synthetic domains set up at scales that are comparable to the actual mechanism in question. This is the proposed approach in the follow up Chapter 13 that focuses on processes and scales that are amenable to forecasting. Coupled remote and local forcing The dependence of LOW variability on the coupling between remote and local forcing is a key finding which makes the forecasting potential of LOW variability and its impacts a possibility (van der Plas et al. 2006). This is because it is remote forcing that defines the regime modes that govern variability in the northern and central Benguela through the boundary conditions. Regime mode shifts on a basin scale that eventually impact on the Benguela boundaries may not only be forecast on a time scale of months but their impact on a time scale of years – decades may perhaps be evaluated through scenario modelling. However this forecasting potential depends sensitively on the proposed biogeochemical coupling between remote and local forcing (van der Plas et al. 2006). It is important that this hypothesised link be tested using a combination of modelling and observational data. The coupling was proposed using steady state assumptions and its incidence in a time varying sense needs to be tested (van der Plas et al. 2006). PROCESSES WITH FORECASTING POTENTIAL The table of processes below (Table 5-2) shows that the cost effective observational capacity of the individual processes is mostly good but the forecasting of LOW variability depends largely on how well the linkages that transfer the equatorial signal to the Benguela are understood and modelled.
Three different temporal or forecasting scales of LOW variability are evident from the table: • the short term events of a few days with localised impact
Benguela low oxygen variability
85
Table 5-2. Physical processes that have a bearing on the variability of Low Oxygen Waters in the Benguela and may be worth forecasting.
Domain
Forcing System
Processes
Scales of variability
Observational Potential
Worthwhile to Forecast
Remote
ETSA
Equatorial upwelling and new production
Seasonal interannual
Good: Ocean Colour
Maybe
Intensity and timing of trade winds
Seasonal Interannual
GCM
Yes
Equatorial stratification
Interannual – decadal (Benguela Niño)
Good: Ocean Buoys
Yes
Angola Current
Seasonal interannual
Good: Altimetry and AVHRR
Yes
LOW in the ETSA
Interannual decadal
Good: Ocean Buoy
Yes
Depth of the upper boundary of O2 < 2ml l-1
Interannual decadal
Good: Ocean Buoy
Yes
Depth range of O2 < 2ml l-1
Interannual decadal
Good: Ocean Buoy
Yes
Poleward transport of LOW into Benguela
Seasonal Interannual
Good: Ocean Buoy
Yes
Upwelling driven new production
Days - weeks
Good (ocean colour remote sensing)
Yes
Upwelling wind variability
Days - weeks
Good
Maybe
Relaxation events in the southern Benguela
Days - weeks
Good
Maybe
Spatial scales of depositional areas
10 – 1000km
Good
Yes
Transport and dispersion of LOW
10 – 1000km
Poor
Yes
Transport and dispersion of sulphide in water column
10 – 100km
Poor
Yes
Local
Upwelling centres
Ecosystem responses
86
• •
P.M.S. Monteiro and A.K. van der Plas
medium term events of a few months duration and with shelf wide impact long lasting LOW variability of year to decade scale and with system wide impact (scenarios)
Any future operational LOW forecasting system will need to combine modelling with data assimilation and verification platforms. It is envisaged that such a system would make use of “real time large scale models” running predominantly with data assimilation into which will be nested the LOW region specific model domains. These would derive their boundary conditions from the large scale models, advect the signal and drive the internal processes that govern LOW variability in critical habitat areas. Forecasting in region specific domains would be based on “free running” models rather than on data assimilation and would be verified against real time data sets. The LOW scales and processes that are most amenable to forecasting whether for scientific reasons or because they are relevant to ecosystem management perspectives are addressed in detail in the companion Chapter 13 (Monteiro et al. this volume). WHAT ARE THE GAPS? Time series observations The Lüderitz upwelling centre plays a pivotal role in forcing the system by supplying upwelled water to both the central and southern Benguela shelf. However, the attempts to understand this role are severely limited by the paucity of data from this area. The most important forcing point has the weakest data set. The temporal resolution of the data from the second most important upwelling “gate” in the Benguela, Cape Frio, is also quarterly at best. (Refer also to Shillington et al. and Reason et al. Chapters 4 and 10 respectively, this volume) Slope - Shelf exchange
An observational programme should be put in place that will elucidate the mechanisms of slope-shelf exchange of LOW. The proposed observationally based early warning system should make use of the understanding derived from both the literature review and the data based advances. The observational programme to support a first early warning system should aim to make use of existing freely available data products. These should include modelling, remote sensing and observational programmes that are already in place. The processes that need to be monitored include: • The thermocline characteristics in the ETSA area which governs the LOW characteristics for the Benguela; • Poleward advection of warm tropical water:
Benguela low oxygen variability
87
• This was shown to be one of the most sensitive indicators of the southward propagation of LOW • from the ETSA to the Benguela off the shelf • the shelf based southward displacement of tropical surface waters e.g.: ABF; and • The thermocline depth and strength on the slope (1000 – 2000m) and on the shelf (100m) The recommended additional observational programmes are the monitoring of oxygen and temperature in the ETSA region and at the two main upwelling centres that cover the Central and Southern Benguela. This is most likely best done using large ocean buoys with temperature records at 50 m intervals and oxygen observations at 50m intervals in the upper 200m and 100m spacing below 200m. The buoys should be located on the slope in the zone of the 1000 – 2000m depth range. These should also provide telemetry based data streams that allow the data quality to be assessed and test linkages with response scales at the monthly monitoring sites off Walvis Bay and St Helena Bay. It is recommended that, when this proposed programme is accepted, the BCLME commission be the regional facility to provide the products to the community. This first phase early warning system is expected to be operational for a period of two years by which time the modelling platforms for the BCLME should be operational and providing a second and later third phase forecasting. Remote - local coupling Advection vs. local formation: Local formation of hypoxic or anoxic LOW depends on the boundary conditions being at or below a critical threshold (approximately 1.5ml l-1 O2) at which the physical supply rate of dissolved oxygen falls below the biogeochemical demand and the system rapidly switches to anaerobic respiration (van der Plas et al. 2005). Thus the magnitude and persistence of LOW variability on the Benguela shelf is primarily the result of the degree of oxygen depletion in incoming water across the boundary and only secondarily the local oxygen demand driven by the sedimenting flux of upwelling-linked new production. The latter is, however, responsible for modulating the response of the system to boundary forcing (Monteiro et al. 2004). It has been proposed that state of environment (SOE) indicators be devised to monitor the LOW status over the Benguela shelf. The SOE indicator effort should be focussed on the areas where regular spatial monitoring can take place at least once a month. The indicators are a measure of the response of the system and should therefore also ideally be located in areas where measuring that response is of relevance. The present monthly monitoring lines in the southern, central and northern Benguela partially fulfil these requirements as the location of the lines is for historical reasons pragmatically close to the sponsoring institution. The following thresholds are suggested: depth of oxygen under- saturation in two zones of the shelf, namely at depth < 100m and depth > 100m.
88
P.M.S. Monteiro and A.K. van der Plas
SUMMARY LOW variability in the Benguela is governed by varying scales of remote and local forcing linked to both Equatorial and Cape Basin systems. The nature of these nonlinear interactions is not clearly understood because scales are large and their elucidation through observational programmes alone is not cost effective. Models are required to characterise the complexity of the most important forcing and response scales in both time and space. It will be necessary to approach this as a multi-phase process, beginning with a diagnostic emphasis which evolves to a forecasting system through hindcasting focussed specifically on large scale events of the past. It is clear that not all the variability scales are amenable to forecasting either because the driving process scales are too uncertain or because they are of little management of policy interest. Two scales were defined as being of interest to both these criteria: • Short term (7 day) scale related to forecasting conditions leading to the walkout or mortality of rock lobster in the southern Benguela • Medium term (2 month) forecasting of the intensification of the remote forcing of ETSA derived LOW which has a bearing on the Namibian hake fishery These two scales are discussed in detail in the companion Chapter 13, this volume. ACKNOWLEDGEMENTS We acknowledge the inputs from our collaborators Geoff Bailey and Quilanda Fidel as well as the constructive comments from Profs. Geoff Brundrit and Vere Shannon and Dr Piers Chapman of Louisiana State University. Our participation is made possible through the support of our respective institutions CSIR, South Africa and MFMR, Namibia and the work in general was only possible through the data contributions from both MCM, South Africa and IIM, Angola. REFERENCES Bailey, G.W. 1983. Pilot study of the vertical flux of POC and PON in St Helena Bay., South African Journal of Science. 79: 145-146. Bailey, G.W. 1991. Organic carbon flux and development of oxygen deficiency on the modern Benguela continental shelf south of 22°S spatial and temporal variability. 171-183 in Tyson, R.V. and T.H. Pearson, editors. Modern and Ancient Continental Shelf Anoxia. Bailey, G.W. and P. Chapman. 1985. Nutrient Status in the St Helena Bay region in February 1979. 125145 in Shannon, L.V., ed. The South African Ocean Colour and Upwelling Experiment. Cape Town, Sea Fisheries Research Institute. Bubnov, V.A. 1972. Structure and characteristics of the oxygen minimum layer in the southeastern Atlantic. Oceanology 12:193-200. Chapman, P. and L.V. Shannon. 1985. The Benguela Ecosystem Part II. Chemistry and Related Processes. Oceanography and Marine Biology: An Annual Review. 23:183-251. Chapman, P. and L.V. Shannon. 1987. Seasonality in the oxygen minimum layers at the extremities of the Benguela system. South African Journal of Marine Science. 5:85-94. Copenhagen, W.J. 1953: The periodic mortality of fish in the Walvis region; a phenomenon within the Benguela Current. Investigational Report of the Division of Fisheries South Africa 14, 35pp. Duncombe Rae, C. M. 2004. A demonstration of the hydrographic partition of the Benguela upwelling ecosystem at 26˚40'S. African Journal of Marine Science, (in press).
Benguela low oxygen variability
89
Florenchie, P., R.E. Lutjeharms, and C.J.C. Reason. 2003. The source of Benguela Niños in the South Atlantic Ocean. Geophysical Research Letters 30(10): 12. Hamukuaya, H., M. O’Toole and P.M.J. Woodhead . 1998. Observations of severe hypoxia and offshore displacement of Cape Hake over the Namibian shelf in 1994. In: Benguela dynamics: Impacts of variability on shelf-sea environments and their living resources. South African Journal of Marine Science 19:57-61. Hardman-Mountford, H.J. A.J. Richardson, J.J. Agenbag, E. Hagen,L. Nykjaer e, F.A. Shillington, C. Villacastin. 2003. Ocean climate of the South East Atlantic observed from satellite data and wind models. Progress in Oceanography. 59:181 – 221 Hart, T. J. and R. I. Currie. 1960. The Benguela Current. Discovery Reports 31: 123-298. Johnson, A. and G. Nelson. 1999. Ekman estimates of upwelling at Cape Columbine based on measurments of longshore wind from a 35 year time-series. South African Journal of Marine Science 99: 433 – 436. Lass, H.U., M. Schmidt, V. Mohrholz and G. Nausch. 2000. Hydrographic and current measurements in the area of the Angola-Benguela Front. Journal of Physical Oceanography 30: 2589-2609. Marchand, J. M. (1928). The Nature of the Sea-Floor Deposits in certain Regions of the West Coast. Fish and Marine Biological Survey, Department of Mines and Industries Annual Report 6(5): 1-11. Mercier, H., M. Arhan and J.R.E. Lutjeharms. 2003. Upper-layer circulation in the eastern Equatorial and South Atlantic Ocean in January-March 1995. Deep-Sea Research I 50: 863-887. Mohrholz, V., M. Schmidt and J.R.E. Lutjeharms. 2001. The hydrography and dynamics of the AngolaBenguela Frontal Zone and environment in April 1999. South African Journal of Science 97: 199-208. Monteiro, P.M.S. (1996) The oceanography and biogeochemistry of CO2 in the Benguela upwelling system. PhD Thesis, University of Cape Town, South Africa Monteiro, P.M.S., A.K. van der Plas, G.W. Bailey and Q. Fidel. 2004. Low oxygen variability in the Benguela ecosystem: a review and new understanding. CSIR Report (Internationally Peer Reviewed), ENV-S-C 2004-075, 67pp. Monteiro P.M.S. and A. Roychoudhury. 2005. Spatial Distribution of Trace Metals in an Eastern Boundary Upwelling Retention Area (St. Helena Bay, South Africa): A Hydrodynamic-Biological Pump Hypothesis. Estuarine, Coastal & Shelf Science 65:123-134. Monteiro, P.M.S., G. Nelson, A. van der Plas, E. Mabille, G.W. Bailey, E. Klingelhoeffer. 2005. Internal tide-shelf topography interactions as a potential forcing factor governing the large scale sedimentation and burial fluxes of particulate organic matter (POM) in the Benguela upwelling system. Continental Shelf Research 25:1864-1876. Monteiro, P.M.S., A. van der Plas, G.W. Bailey, P. Rizzoli, C. Duncombe Rae, D. Byrnes, G. Pitcher, P. Florenchie, P. Penven, J. Fitzpatrick, U. Lass. 2006. Low Oxygen Water (LOW) forcing scales amenable to forecasting in the Benguela ecosystem. Chapter 13, p67-89 in Shannon, V., G. Hempel, P. Malanotte-Rizzoli, C. Moloney and J. Woods, eds. The Benguela: predicting a large marine ecosystem. Elsevier (this volume). Moroshkin, K.V., V.A. Bubnov, and R.P. Bulatov. 1970. Water circulation in the eastern South Atlantic Ocean. Oceanology 10:27-37. Penven P., C. Roy, A. Colin de Verdiere, and J.L. Largier. 2000. Simulation of a coastal jet retention process using a barotropic model. Oceanologica Acta 23(5): 616-634 Pieterse, F. and D.C. van der Post. 1967. Oceanographical Conditions Associated with Red- Tides and Fish Mortalities in the Walvis Bay Region. Investigational Report of the Administration of South West Africa Marine Research Laboratory 14, 125pp. Pitcher, G.C. and S.J. Weeks. 2005. Variability and potential for prediction of harmful algal blooms in the southern Benguela ecosystem. (this volume) Reason et al., this volume Reid, J. L. 1989. On the total geostrophic circulation of the South Atlantic Ocean: Flow patterns, tracers, and transports. Progress in Oceanography 23:149-244. Shannon L.J. and A. Jarre-Teichemann. 1999. A model of trophic flows in the northern Benguela upwelling system during the 1980’s. South African Journal of Marine Science 21: 349 – 366. Shillington, F.A. et al. 2005. Large Scale Physical Processes. This volume 67-89. Stramma, L. and M. England. 1999. On the water masses and mean circulation of the south Atlantic ocean. Journal of Geophysical Research (Oceans), 104: 20863 – 20883. Stramma, L. and F. Schott. 1999. The mean flow field of the tropical Atlantic Ocean. Deep Sea Research II, 46: 279–303.
90
P.M.S. Monteiro and A.K. van der Plas
Touratier F., J.G. Field and C.L. Moloney. 2003. Simulated carbon and nitrogen flow of the planktonic food web during an upwelling relaxation period in St Helena Bay (southern Benguela ecosystem). Progress in Oceanography 58:1-41. van der Plas, A.K., P.M.S. Monteiro, and A. Pascall. 2005. The cross shelf biogeochemical characteristics of sediments in the central Benguela and its relationship to overlying water column hypoxia. Continental Shelf Research, in review. Voituriez, B. and A. Herbland. 1982. A Comparative Study of the productive systems of the tropical east Atlantic: Thermal Domes, Coastal Upwellings and Equatorial Upwelling. Rapp .P-v. Reun. Cons. perm. int. Explo. Mer. 180:114-130. Waldron, H.N. and T.A. Probyn. 1991. Short term variability during an anchor station study in the southern Benguela upwelling system: Nitrogen supply to the euphotic zone during a quiescent phase of the upwelling cycle. Progress in Oceanography 28:153 – 166 Woodhead, P.M., H. Hamukuaya, M.J. O’Toole, and M. McEnroe. 1998. Effects of oxygen depletion in shelf waters on hake populations off central and northern Namibia. In Shannon, L.V. and M.J. O’Toole, editors. International Symposium, Environmental variability in the South East Atlantic., 10 pp. NATMIRC, Namibia. Woodhead, P.M., H. Hamukuaya, M.J. O’Toole, T. Stroemme, G. Saetersdal and M. Reiss. 1997a. Catastrophic loss of two billion Cape hake recruits during widespread anoxia in the Benguela Current off Namibia. In: ICES International Symposium, Recruitment dynamics of exploited marine populations. Physical-biological interactions. pp. 105-106. Woodhead, P.M., H. Hamukuaya, M.J. O’Toole, T. Stroemme and S. Kristmannsson. 1997b. Recruit mortalities in Cape hake, following exclusion from shelf habitat by persistent hypoxia in the Benguela Current, Namibia. In: ICES International Symposium, Recruitment dynamics of exploited marine populations. Physical-biological interactions. pp. 26-27.
91
Large Marine Ecosystems, Vol. 14 V. Shannon, G. Hempel, P. Malanotte-Rizzoli, C. Moloney and J. Woods (Editors) © 2006 Elsevier B.V. All rights reserved.
6 Variability of Plankton with Reference to Fish Variability in the Benguela Current Large Marine Ecosystem – An Overview Larry Hutchings, Hans M. Verheye, Jenny A. Huggett, Hervé Demarcq, Rudi Cloete, Ray G. Barlow, Deon Louw, Antonio da Silva
ABSTRACT This article reviews the variability of plankton over time scales ranging from mesoscale upwelling events of a few days’ duration to decadal scale changes in the northern and southern subsystems in the Benguela Current. It focuses on the plankton that are considered important for fish, particularly the crustacean zooplankton. The southern Benguela is strongly pulsed over periods of 4-12 days with a series of upwelling events modulated by passing cyclonic weather systems. The northern Benguela is less pulsed with short-term variability linked to continental shelf waves. Upwelling is particularly active at seven major sites in the Benguela system. Dense phytoplankton blooms develop in the cool nutrient-rich plumes, which merge and blend with surrounding waters, creating a broad band of phytoplankton-rich water over the shelf. Species succession from small to large diatoms, dinoflagellates and small flagellates occurs as the waters mature after upwelling and generally move offshore, although numerous exceptions occur, with small-celled communities occasionally dominant in nearshore waters. Much regeneration and recycling of nutrients occurs, resulting in lower than expected f-ratios. Frontal zone aggregations provide important feeding opportunities in the transport phase of ichthyoplankton between the Agulhas Bank spawning grounds and the nursery grounds on the South African West Coast. The Angola-Benguela front in the northern Benguela is also an important region for pelagic fish spawning. Seasonal changes in wind forcing indicate maximum upwelling in spring and autumn throughout the Benguela, with a tendency for a summer maximum in the south. Lüderitz (25oS) and Cape Frio (17oS) are particularly active upwelling regions. Phytoplankton biomass, estimated as chlorophyll a, shows a winter maximum in the northern Benguela and a summer maximum in the southern Benguela. The Lüderitz area shows perennial phytoplankon minima, possibly due to strong turbulence. The central Namibian shelf and the South African west coast shelf have persistently high phytoplankton biomass. A seasonal intrusion of warm oligotrophic water from Angola in late summer (December to March) results in strong contrasts between winter and
92
L. Hutchings et al.
summer in the extreme northern Benguela. Zooplankton biomass shows different cycles along the coast, with spring, summer and autumn maxima in the south, and a slight maximum during the second half of the year (July to December) off central Namibia. The dominant fish spawning period is spring–summer throughout the region. Long-term changes in the southern Benguela include a significant increase in zooplankton over the past five decades, with a decline since 1995. Fish abundance has declined in the northern Benguela but remained reasonably stable in the southern Benguela until 2000, when pelagic fish biomass increased dramatically with concomitant declines in zooplankton biomass. A range of modelling exercises, including expert systems, statistical models and linked IBM-hydrodynamic models, has been compared to or derived from field data, and has stimulated new observational programmes at improved space and time scales. Observational data at pertinent time and space scales are lacking in the northern Benguela system, which will hamper validation of prognostic and diagnostic models. INTRODUCTION From space, viewed at a coarse scale (Demarcq et al. 2003), the Benguela Current looks like a cool, broad sluggish drift, the eastern portion of the South Atlantic gyre, which is characterized by a narrow belt of cold, phytoplankton-rich water along the coastline (Figure 6-1a). Sharp discontinuities occur at the southern boundary with the Agulhas Current (32˚-37˚S) and at the northern boundary in the Angola-Benguela Front region (12˚-18˚S). Intensive mixing and high variability characterize these boundary zones, matched only by the Brazil/Malvinas current interactions in the SW Atlantic (Bakun 1996). Within the major upwelling region of the Benguela, there are seven particularly active sites (Shannon and Nelson 1996), of which the Lüderitz cell (25˚-26˚S) is by far the most powerful in terms of Ekman transport and turbulent mixing (Parrish et al. 1983). Event-scale variability is often dominant, with significant seasonal modulation of upwelling winds in the northern and southern extremities. In a wind-driven upwelling system such as the Benguela, the plankton variability is driven by complex non-linear interactions between the driving forces of winds and solar radiation, and stabilization, sinking and the response times of the individual organisms. The time scales vary from hours and days through to inter-decadal shifts, with corresponding spatial scales. Fish populations that inhabit this ecosystem are subject to high variability in the triad of factors affecting recruitment, i.e. enrichment, retention and concentration (Bakun 1996), particularly during the early life-history stages. This paper is intended to give a brief description of the variability of the planktonic components within the Benguela Current and some of the problems associated with predicting future events, with particular reference to the dynamics of fish populations, such as growth, recruitment and fish condition. As such, this review focuses on the crustacean components, principally the copepods and the euphausiids. Pitcher and
Variability of plankton with reference to fish variability, BCLME
(a)
93
(b)
(e) (c)
13 15 17
Latitude (°S)
19
(d)
21 23 25 27 29 31 33
Figure 6-1. (a) Annual average chlorophyll concentration computed from monthly composites derived from Global Area Coverage (GAC) data at 4.5 km resolution from SeaWiFS ocean colour sensors between September 1997 and April 2002 and the annual mean position of the 1 mg m-3 offshore limit between 12 and 34˚S; (b) time-series of monthly averages of chlorophyll biomass integrated between the coast and this limit, showing seasonal and interannual variability and spatial patterns within the Benguela system, September 1997 – July 2003; (c) monthly variations in an enrichment index (chlorophyll content summed between the coast and the 1 mg.m-3 isoline) in the northern and southern Benguela, showing opposing trends, January 1998 – March 2002; mean seasonality of indices of (d) upwelling (upwellingfavourable winds, low-moderate-high) and (e) enrichment (mean integrated chlorophyll a) in the Benguela Current system (from Demarcq, unpubl. data).
94
L. Hutchings et al.
Weeks (2006) cover the major variability in phytoplankton, with particular reference to the formation of harmful algal blooms close inshore. However, fish populations such as anchovy (Engraulis encrasicolus), sardine (Sardinops sagax) and hake (Merluccius spp.) spawn over extensive areas of the continental shelf, and serially spawn for prolonged periods, possibly as an adaptation to the high mesoscale and microscale variability which characterizes the Benguela. As always, the major problem has been integrating small-scale events over a sufficiently long time and large area to provide indices of the suitability of a particular zone that is pertinent to fish population dynamics. EVENT-SCALE VARIABILITY Southern Benguela Characteristic time scales of phytoplankton and zooplankton in relation to the wind forcing and pelagic fish early life history were first described by Hutchings and Nelson (1985). Upwelling in the southern Benguela is typically pulsed over a period of approximately six days, varying between two and 12 days of upwelling-favourable winds interspersed with calms or wind reversals (Nelson 1992). The patterning of wind variability in the southern Benguela is best illustrated by Roy et al. (2001) (Figure 6-2a). Twelve upwelling events are seen over a six-month period, and each results in a “spring bloom!” For each wind event, cold source water originating from South Atlantic Central Water (Shannon 1985) rises at the upwelling site and moves offshore, warming, dispersing and mixing with existing surface waters. Essentially, this water injects new nitrogen into the euphotic zone, where it is retained by heating and stabilization of the upwelled water as a lens of warm water overlying cooler waters. Phytoplankton seed cells, either resting cells or viable cells present in source water or cells mixed laterally into the newly-upwelled water mass, grow and divide rapidly, forming dense blooms (Pitcher and Weeks, 2006). By following drogues placed in the coldest, newly-upwelled waters, the development and decline of phytoplankton in individual “boluses” of water could be followed over periods of 6-10 days (Figure 6-2b; Brown and Hutchings 1985). This period is short in relation to the 2-4 weeks development time of a dominant West Coast upwelling copepod, Calanoides carinatus, at temperatures typical of sun-warming upwelled waters (Figure 6-2c; Peterson and Painting 1990; Hutchings 1992), resulting in a potential for a major mismatch between phytoplankton and grazers. Only when chlorophyll a rises above 36 mg.m-3 do large herbivorous copepods respond in terms of increased egg production (Figure 6-2d; Armstrong et al. 1991) and somatic growth rates (Richardson and Verheye 1998; 1999). This further shortens the overlapping optimal time period between plants and grazers, since by the time the juvenile stages are developed the bloom has declined. Behavioural adaptations such as vertical migration may help to prolong the residence time of zooplankters in developing phytoplankton patches.
Variability of plankton with reference to fish variability, BCLME
95
Figure 6-2. Typical time scales for the Benguela region. (a) Daily time-series of north-south wind speed at Cape Columbine, 1 November 1999 – 30 April 2000 (top) and cumulative divergence per upwelling event for the same period (bottom); the episode number is indicated for each major upwelling event (after Roy et al 2001). (b) Development and decay of phytoplankton blooms over 6-13 days, tracked by drogues placed in newly upwelled water (after Brown and Hutchings 1985). (c) Development of Calanoides carinatus populations at 15ºC over 18-20 days, indicating a basic temporal mismatch with phytoplankton blooms in a strongly-pulsed upwelling system (after Peterson and Painting 1990 and Hutchings 1992). (d) Daily egg production by female Calanoides carinatus in response to variable chlorophyll a concentrations (after Armstrong et al. 1991 and Hutchings and Field 1997)
96
L. Hutchings et al.
Figure 6-3. Vertical sections of daily variation in (a) temperature; (b) nitrate concentration; (c) chlorophyll a concentration; (d) major phytoplankton groups and successional changes of dominant taxa; and (e) primary production during an anchor station time-series study in St Helena Bay, 20 March – 15 April 1987. Arrows at the top indicate the advection of newly-upwelled water (after Mitchell-Innes and Pitcher 1992)
Variability of plankton with reference to fish variability, BCLME
97
A 27-day anchor station study carried out in St Helena Bay (33˚S) in March-April 1987 demonstrated a sequence of two upwelling cycles as the upwelled water stabilized (Figure 6-3), and some succession of phytoplankton types, from small diatoms to large diatoms and then to small flagellates (Mitchell-Innes and Pitcher 1992). The same changes appear to occur with distance offshore on transects from the coast (Barlow et al. 2005). Using diagnostic pigment indicators, Barlow et al. (2005) showed diatoms and dinoflagellates were dominant in cool, maturing upwelled waters over the shelf, while small flagellates were dominant offshore but were prominent on occasions in the nearshore zone and likely to contribute significantly to the primary productivity of the Benguela. As copepods develop, there are ontogenetic changes in their vertical migration behaviour (Verheye and Field 1992; Huggett and Richardson 2000), which minimize predation and facilitate retention within the phytoplankton patch and within the nearshore zone (Verheye et al. 1992; Huggett 2003). The amplitude of migration is related to body size, developmental stage and food concentration (Verheye and Field 1992; Huggett 2003), and interspecific depthpartitioning has been demonstrated for some of the smaller copepod species in the southern Benguela (Stuart and Verheye 1991). Ontogenetic depth layering and migration have also been shown for local euphausiid communities (Figure 6-4a; Pillar et al. 1989), which, combined with differential flow patterns with depth, facilitate retention over the shelf. These life-history features provide important inputs to models of copepod or euphausiid abundance, as well as to individual-based models, such as that of Parada (2003), who modelled the vertical behaviour of anchovy larvae in the southern Benguela nursery area. Since peak upwelling occurs in summer, the frequency of upwelling events increases through the season, and successive plumes of stabilizing upwelled waters merge and mingle alongshore and across-shore, resulting in a broadening of the plankton-rich belt along the coast (Figure 6-1a). Much regeneration, recycling and secondary blooming (Figure 6-5) appear to occur in the mature upwelled waters, resulting in lower than anticipated f-ratios in the upwelling region (Probyn 1992). The seaward boundary of the inshore productive belt is a strong convergent front, where positively buoyant or phototropic organisms will aggregate in the gradually descending mature upwelled water as it sinks beneath the warm offshore water. Larval fish transported in the frontal jet currents over summer (Figure 6-6; Huggett et al. 1998) will therefore be afforded concentrated food organisms. The stronger the upwelling winds, the more convergence and concentration; essentially, this is a vertically orientated thermocline, which extends along much of the West Coast shelf as far as Lüderitz at about 25˚S.
Northern Benguela Upwelling appears to be less pulsed off Namibia, since the influence of the eastwardmoving cyclones is diminished with decreasing latitude (Shannon 1985). Upwelling centres are more dispersed and belt-like, so an inshore-offshore gradient in phytoplankton and zooplankton is apparent along most of the Namibian coast. Few
98
L. Hutchings et al.
Figure 6-4. (a) Weighted mean depth plots of eggs and early and late larval stages of Euphausia lucens in the southern Benguela (after Pillar et al. 1992). (b) Vertical distribution of Stylocheiron longicorne, Nematoscelis megalops, Thysanoessa gregaria, Euphausia gibboides, E. hanseni and E. americana. Data are averaged from nine night-time samples taken at a fixed station in the northern Benguela (after Barange 1990). (c) Conceptual three-dimensional model of cross-shelf circulation during upwelling in the northern Benguela. Encircled crosses denote equatorward flow, encircled dots indicate poleward flow (after Barange and Pillar 1992)
Variability of plankton with reference to fish variability, BCLME
99
Figure 6-5. Daily nitrogen transfers (mmol N m-3 day-1) for (a) a net-phytoplankton-dominated and (b) a pico- and nanophytoplankton-dominated system (after Probyn 1992). Nitrogen uptake rates and ammonia regeneration rates were measured during February 1991. Phytoplankton compartments are based on particulate N concentrations. Dissolved nitrogen fluxes are represented by shaded arrows and predatorprey transfers by clear arrows. Broken arrows indicate the sinking by phytoplankton and egestion by zooplankton. Fluxes are balanced assuming steady state to produce a potential zooplankton production P. Note the greatly increased mesozooplankton production and flux during the net-phytoplankton dominant state.
100
L. Hutchings et al.
Figure 6-6. Three-dimensional representation of mean monthly abundances and distribution of (a) anchovy eggs, (b) anchovy larvae, (c) sardine eggs, and (d) sardine larvae along the SARP transect off the Cape Peninsula from August 1995 to July 1996 (after Huggett et al. 1998). The transect extended 65km offshore with stations 01-04 situated over the shelf.
event-scale studies have been made here, a notable exception being repeated transects of 15 stations at 36-hour intervals off 20˚S by the R.V. Alexander von Humboldt in the spring of 1979, at 30-170 km offshore (Hagen 1985; Postel 1990). The physical variability appeared to peak at about 5.6 days up to 100 km offshore, driven by continental shelf waves; beyond that a 13-day interval was indicated. Phytoplankton peaked close inshore whereas zooplankton wet mass peaked at 100-130 km from the coast. Barange (1990) demonstrated differential vertical distribution among six species of euphausiid (Figure 6-4b), suggesting niche partitioning of potentially competitive species. In contrast to the one-celled, cross-shelf circulation model proposed for the dominant euphausiid (Euphausia lucens) in the southern Benguela, a two-celled circulation model comprising both nearshore and offshore cells was proposed to facilitate euphausiid retention and niche partitioning in the northern Benguela (Figure 6-4c; Barange and Pillar 1992). Only Verheye et al. (2005) have reported on shortterm changes in zooplankton in Angolan waters.
Variability of plankton with reference to fish variability, BCLME
101
SEASONAL CHANGES The clearest exposition of Benguela-wide seasonal variability in indices of upwelling (wind) and enrichment (chlorophyll a) is that of Demarcq (unpubl. data) (Figure 6-1b, c, d, e). Indices of enrichment were based on a time-series of Global Area Coverage (GAC) data at 4.5 km resolution from SeaWiFS ocean colour sensors between September 1997 and July 2003. Upwelling-favourable winds appear to have two maxima (in spring and autumn) throughout the Benguela upwelling region, tending towards a summer maximum in the extreme south. Stronger, more perennial winds blow at Lüderitz (26˚S) and Cape Frio (17˚S), with a slight winter maximum in northcentral Namibia. However, chlorophyll a maxima occur in the winter/spring months in the northern Benguela whereas the reverse pattern is seen in the southern Benguela, separated at Lüderitz where, paradoxically, a relative chlorophyll-minimum is evident throughout the year, possibly due to extreme turbulence. Persistent, strong upwelling also occurs at 17˚S (off the Cunene River), but a seasonal intrusion of warm Angola Current water in December-March overrides the phytoplankton signal. Off central Namibia and off the west coast of South Africa high phytoplankton concentrations persist throughout the year. Southern Benguela Another seasonal feature superimposed on the upwelling signal in the southern Benguela is a stabilization of the water column in early spring over the Agulhas Bank and offshore on the West Coast, after deep mixing during winter months (Shannon et al 1984). The nutrients isolated in the euphotic zone after stabilization generally do not result in extensive local phytoplankton blooms, since the water masses and phytoplankton are advected rapidly westwards and offshore, and surface waters are replaced by oligotrophic subtropical waters in filaments and eddies from the Agulhas Current. There is a general but very weak increase in phytoplankton in spring in the waters surrounding the upwelling region. This results in extremely food-poor environments immediately offshore of the upwelling zone over the summer months, unless eddies, filaments or shear-edge features increase productivity in the offshore zone adjacent to the frontal jet. This implies that any larval fish displaced offshore by whichever mechanism may have diminished survivorship, a central tenet of the recruitment hypotheses in the southern Benguela (Cochrane and Hutchings 1995). Seasonal cycles of mesozooplankton (primarily copepods) differ along the coast (Figure 6-7a). On the western Agulhas Bank (35˚S) zooplankton populations peak in late autumn and spring with a minimum in mid-summer. Off the Cape Peninsula (34˚S) the peak is in summer, with considerable monthly variability (Figure 6-7b), while in St Helena Bay (32˚S) the peak is in late summer with a marked decline in autumn. Verheye et al. (1992) suggested that the zooplankton annual cycle, although tightly coupled to the seasonality in upwelling and productivity (see e.g. Richardson et al. 2003a), is not solely driven from the bottom up, but may be altered, confounded or obscured by the top-down effect of predator-prey interactions. Thus, the distribution of different zooplanktivorous life-history stages (larvae, recruits, adults) of pelagic fish
102
L. Hutchings et al.
along different parts of the coast at different times of the year (Hampton 1992, Barange et al 1999)is likely to cause the observed spatial deviations from the expected zooplankton annual cycle. This makes both interpretation and prediction difficult.
Figure 6-7a. Seasonal variability in the biomass of copepods in the St Helena Bay area (top), the Cape Columbine-Cape Point area (centre), and the western Agulhas Bank (bottom); data points are 3-month running means between August 1977 and August 1978 (after Pillar 1986); peak recruitment and spawning seasons of anchovy are also shown (after Verheye et al. 1992).
Variability of plankton with reference to fish variability, BCLME
103
Figure 6-7b. Seasonal variation of mean zooplankton standing stock (top) in response to primary production (bottom) along an Upwelling Monitoring Line off the Cape Peninsula, October 1970 – March 1973 (after Andrews and Hutchings 1980).
Compared with 1977/78, the seasonal signal of zooplankton in St Helena Bay has amplified dramatically in recent years, with an increase of the summer to winter ratio from approximately 2:1 in the late 1970s to well over 5:1 in the early 2000s (Figure 68a), while phytoplankton seasonal variability is not that pronounced (Demarcq et al. 2003; Barlow et al. 2005). During those recent years, anchovy recruitment was at record levels and, as a consequence, winter levels of zooplankton along the west coast have been depleted due to intense predation by these record concentrations of anchovy (see below). Fluctuations in the abundance of Calanoides carinatus, one of the most characteristic copepods of the upwelled waters along the African west and northeast coasts, have been studied extensively (see Verheye et al. 1991). At lower latitudes, this species exhibits a striking seasonality; its life history is characterized by a deep-living state of temporary developmental arrest (dormancy or diapause) at the pre-adult copepodite
104
L. Hutchings et al.
stage C5, and its extremely reduced metabolism (Auel et al 2005) allows it to bridge unfavourable conditions during the prolonged, warm non-upwelling season. In contrast, in the southern Benguela it maintains large populations on the shelf perennially, with only slight seasonal variability in its abundance. It is uncertain whether these populations enter true diapause, as suggested by Verheye et al. (1991), although observations of pre-adult C5s in deep (>600 m) water (De Decker and Mombeck 1964; Borchers and Hutchings 1986) lend support to this hypothesis. This remains an area for future investigation. Likewise, in contrast to the northern Benguela, there is no distinct seasonality in euphausiid biomass in the southern Benguela, where Euphausia lucens is one of the dominant macrozooplankters (Pillar et al. 1992). Many dominant fish species spawn seasonally in the Benguela, resulting in strong seasonal changes in ichthyoplankton abundance. Anchovies have a distinct summer maximum, whereas sardines spawn through most of the year, with a slight minimum in winter months and slight maxima in early spring and late summer, bracketing the anchovy spawning. Round herring (Etrumeus whiteheadii) spawn in late winter/early spring. All three species, and several others, spawn on the Agulhas Bank and their eggs and larvae are advected to the West Coast offshore. By mechanisms that are not well understood, recruits appear on the west coast close inshore. The adult spawning, and juvenile nursery areas are clearly separated (Hutchings et al 1998). There have been marked shifts in the spatial location of spawning, from 32ºS on the West Coast to east of Cape Agulhas (Barange et al 1999, van der Lingen and Huggett 2003, ), with variable interannual recruitment success (Boyd et al ,1998, Hutchings et al, 1998). The utility of the Agulhas Bank as a nursery ground for sardine juveniles is currently being evaluated (Miller et al., in press). Little spawning occurs at the Lüderitz upwelling cell, a site of high offshore losses and turbulence, and low phytoplankton concentrations. Horse mackerel (Trachurus trachurus capensis) spawn in late winter/early spring on the central Agulhas Bank (Barange et al. 1998), with juveniles appearing in the West Coast nursery area over midsummer. Hake appear to spawn all year round, but the deep-spawned eggs have not been well sampled. Most of these species also appear to utilize the inshore West Coast area as a nursery ground (Hutchings et al. 2001). Northern Benguela Seasonal trends off central Namibia (Walvis Bay, 23ºS) indicate a broad division into seasons of maximum (April to December) and minimum upwelling. Phytoplankton are widespread across the shelf in winter months but maximum concentrations are usually found between 10 and 20 n. miles offshore (Louw and Barlow, unpubl. data). In summer, phytoplankton appear to be concentrated inshore as upwelling intensity declines and the surface waters stratify. Zooplankton off Walvis Bay (Figure 6-8b) increase slightly in the second half of the year but, unlike the southern Benguela (e.g. Richardson et al. 2003a), there is no clear signal coupling upwelling activity, phytoplankton concentration and zooplankton abundance in the northern Benguela.
Variability of plankton with reference to fish variability, BCLME
105
a 12 SHBML 2000-2003 (< 200 m)
-2
BIOMASS (g C m )
10 8 6 4 2 0
CELP 1977/78
A M
J
J
A
S O
N
D
J
F M A M
J
J
A
S O
N
D
J
F M A M
J
J
A
S O
N
D
J
F M
o
b
2000
2001
Dec
Oct
Aug
Jun
Apr
Feb
Dec
Oct
Aug
Jun
Apr
Feb
Dec
Oct
Aug
Jun
Apr
2500 2000 1500 1000 500 0 -500 -1000 -1500 -2000 Feb
Nos/m³
Walvis Bay Monitoring Line - 23 S Febr. 2000 - Dec. 2002
2002
Figure 6-8. A comparison of the seasonal variability of zooplankton in the northern and southern Benguela. (a) Mesozooplankton (primarily copepods) biomass (gC m-2) on the continental shelf during monthly SHBML (St Helena Bay Monitoring Line) surveys in 2000-2003 in the St Helena Bay region, South Africa (from Koch and Hutchings, unpubl. data); data from monthly CELP (Cape Egg and Larvae Programme) surveys in 1977-1978 (calculated from Pillar 1986) are superimposed. (b) Copepod abundance (expressed as anomalies from the 2000-2002 time-series mean; No. m-3) on the monthly Walvis Bay Monitoring Line surveys at 23˚S, Namibia (from Cloete, unpubl. data).
In addition to the presence of an active population of C. carinatus throughout the year in the upper water column on the Namibian shelf (Unterüberbacher 1964; Timonin et
106
L. Hutchings et al.
al. 1992) and off southern Angola, there is a resting component of diapausal C5s in deep (300-1000 m) water offshore (Timonin et al. 1992; Arashkevich et al. 1996; Timonin 1997; Loick et al. 2005; Verheye et al. 2005). Whereas abundance of the neritic population varies considerably depending on the upwelling phase, there is very little seasonal variability in abundance of the deep-living animals (Timonin 1990, cited in Arashkevich et al. 1996), suggesting a permanent pool of diapausing animals throughout the year. The triggers that terminate diapause remain poorly understood (Verheye et al. 2005). Interestingly, despite the reversal between the southern and northern Benguela of seasonality in phytoplankton abundance (Figure 6-1c), pelagic fish in the northern Benguela have the same spawning seasons as in the southern Benguela; sardines are widespread spawners over the period September to March, while anchovy spawn in mid- to late-summer in the Angola-Benguela frontal region. Hake spawn in the September-October period from just north of the Lüderitz upwelling cell to northern Namibia. The central Namibian shelf appears to be an important nursery ground for juveniles, where low oxygen concentrations, warm water intrusions and strong upwelling are thought to exert major influences on recruitment success. Strong intrusions of very warm Angola Current water occur about every ten years with marked effects on the distribution and recruitment of Namibian fish stocks. A large intrusion of very low-oxygen water in 1994 appeared to have resulted in very poor hake recruitment and marked changes in the entire northern Benguela ecosystem (e.g. Boyer et al. 2001). INTERANNUAL AND DECADAL CHANGES In contrast to plankton variability, which has for long been regarded as trivial and not directly relevant to fisheries, decade-scale variability is a primary feature of fish stocks. For instance, several of the world’s productive upwelling regions have experienced extensive fluctuations in pelagic fish yields and regime shifts of fish populations, which are echoed in the sediment record of fish scale deposits over periods of 50-60 years (Lluch-Belda et al. 1989; 1992; Schwartzlose et al. 1999). Recent research (see Colijn et al. 1998) has, however, shown that long-term variability in plankton is closely linked to climate change and that foodweb changes are also manifested in long-term variations in the abundance, distribution and species composition of the plankton (see Perry et al. 2004 and references therein). There are three mechanisms that control trophic levels in marine ecosystems (Cury and Shannon 2004). The first mechanism is bottom-up control, the conventional trophic flow control that seems to dominate most ecosystems, with the environment as the controlling agent. A marked change in the environment will alter the primary productivity of the ecosystem and its availability to higher trophic levels. The second mechanism is top-down control, by which lower foodweb components are regulated by higher-level predators. Thus, the abundance of predators will determine that of lower trophic levels, leading to alternating up- and down patterns of abundance, or trophic cascades. The third mechanism is wasp-waist control, where environmental changes
Variability of plankton with reference to fish variability, BCLME
-3
CONCENTRATION (mg m )
b
3 2
Pseudo wind
1
-3
Mean surface
20 15 10 5 0 1950
1960
1970
1980
1990
2000
(c ) To ta l c o p e p o d s
7.0
LOG (ABUNDANCE)
1990
-2
25
(b)
c
1970
0 -1
1950
ANOMALY (100 m2 s-2)
a
107
6.0
5.0
4.0
3.0
2.0 1950
1960
1970
1980
1 990
2000
5.0
(d ) L a rg e c a la n o id s
d LOG (ABUNDANCE)
4.5
4.0
3.5
3.0
2.5
2.0 1950
1955
1960
1965
1970
1975
1980
1985
1990
1995
2000
Figure 6-9a to d. Bottom-up control of zooplankton in the southern Benguela: long-term time-series of: (a) an upwelling index, as annual averaged anomalies of equatorward pseudo wind stress on the west coast of South Africa, representing the southerly, upwelling-favourable component, 1951 – 1989 (redrawn from Shannon et al. 1992); (b) phytoplankton biomass, as annual (March – June) mean surface chlorophyll a concentration measured in St Helena Bay, 1974 – 1996 (from Verheye 2000); Mesozooplankton abundance, as annual mean abundance of total copepods (c) and large calanoid copepods (d) within ca 50 n. miles from the coast in the St Helena Bay region during autumn (March – June) of 1951-2004 (b, c and d updated from Verheye and Richardson 1998, Verheye et al. 1998, Verheye 2000).
108
L. Hutchings et al.
may initiate ecosystem changes via direct effects on dominant middle trophic level pelagic fish species, which in turn affect other pelagic fish species as well as higher and lower trophic levels. Southern Benguela Over the past 4-5 decades, declining trends in zooplankton have been observed in upwelling regions of most eastern boundary current systems (Cury et al. 1999). In contrast, in St Helena Bay on the west coast of South Africa, where pelagic fish recruit most intensely during the austral autumn/winter, zooplankton abundance has increased 100-fold since the 1950s (Verheye and Richardson 1998; Verheye et al. 1998). At the same time, there was a long-term shift in zooplankton community size structure, from large to small, coincident with a decade-scale change in dominance from sardine to anchovy, two planktivorous species with different prey size selectivities. Parallel long-term trends across consecutive lower trophic levels in the southern Benguela have been documented (Verheye 2000), which suggest upward-propagating effects of oceanographic and biological processes in response to increased wind stress (Figure 6-9a). This leads to intensified upwelling, nutrient enrichment and enhanced phytoplankton (Figure 6-9b) and zooplankton production (Figure 6-9c). Conversely, alternating up- and downward long-term trends observed across consecutive higher trophic levels (Verheye 2000) reveal trophic cascading effects of predator-prey interactions, from piscivorous apex predators (including fishing activities) to planktivorous small pelagic forage species to zooplankton. Beginning in the mid-1990s, there has been a reversal in the long-term trend in zooplankton of St Helena Bay (Figure 6-9c). Declining abundances are particularly evident in the large calanoid copepods (2-5 mm TL) (Figure 6-9d), which are selectively preyed upon by anchovy recruits (James, 1988; James and Findlay 1989). The biomass of these fish has increased substantially since the mid 1990s (see Figure 6-10a), with their predatory effect reducing prey abundances to below the time-series minimum of the 1950s (Figure 6-9d). That top-down mechanisms may indeed play an appreciable role in controlling coastal zooplankton populations is supported by negative relationships found between copepod abundance and pelagic fish stocks, both on the west coast recruit grounds and the south coast spawning grounds. Such predator-prey relationships are particularly evident since 1988, when hydro-acoustics were introduced for more accurate fish stock assessments and when zooplankton were concomitantly monitored along the coast (Figure 6-10). The best relationships were found between the abundance of large calanoid copepods in St. Helena Bay and the biomasses of total pelagic fish (r2 = 0.46) and anchovy recruits (r2 = 0.49; Figure 6-10b) on the west coast during autumn (March-June) of 1988 - 2004.
Variability of plankton with reference to fish variability, BCLME
a
109
Recruit surveys: Winter - West Coast
BIOMASS ('000 tons)
4000
RED EYE 3000
SARDINE
ANCHOVY
2000 1000
2500
4.5
2000
4.0
1500
3.5
1000
3.0
500
2.5
0
2004
ANCHOVY BIOMASS ('000 tons)
2002
2000
1998
1996
1994
1992
5.0
1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004
LOG ABUNDANCE LARGE CALANOIDS
b
1990
1988
0
c ANCHOVY BIOM AS S ('000 tons)
2500
y = -2609.6Ln(x) + 4167.7
2000
2
R = 0.7047 (excluding 2000 data point)
1500 1000 500 0 2.5
3.0
3.5
4.0
4.5
5.0
LOG ABUNDANCE LARGE CALANOIDS
Figure 6-10 a, b and c. Top-down control of zooplankton in the southern Benguela: (a) time-series of recruit biomass estimates of pelagic fish during annual hydro-acoustic stock assessment surveys on the west coast of South Africa, May – June of 1988 – 2004 (from Coetzee, unpubl. data); (b, c) relationship between large calanoid copepod abundance in St Helena Bay and anchovy recruit biomass on the West Coast, suggesting predator-prey interactions (from Verheye, unpubl. data);
110
L. Hutchings et al.
Spawner surveys: Summer - South Coast
d Biomass (000 tons)
7500
Round herring Sardine Anchovy
5000
2500
e
A N C H O VY SPA W N ER B IO M A S S ( '0 0 0 t o n s )
3500 3000 2500 2000 1500 1000 500 0
2002
2001
2000
1999
1998
1997
1996
1995
1994
1993
1992
1991
1990
1989
1988
1987
0
y = -986.69Ln(x) + 974.63 R2 = 0.6054
0.0
0.5
1.0
1.5
2.0
2.5
CALANUS BIOMASS (g dry wt m -2)
Figure 6-10 d and e. (d) time-series of spawner biomass estimates of pelagic fish during annual hydroacoustic stock assessment surveys on the south coast, November of 1988 – 2002 (from Coetzee, unpubl. data); (e) predator-prey relationship between anchovy spawner and Calanus agulhensis biomasses on the south coast for the period 1988 to 2000 (after Huggett 2003).
Exclusion of the data for 2000, a year when zooplankton levels remained comparatively high despite exceptionally high fish recruitment (the highest on record), improved the relationships substantially (r2 = 0.61 and 0.70 respectively; Figure 610c). Likewise, based on data from the south coast during summer (1988-2000), Huggett (2003) described significant predator-prey relationships between the biomasses of the dominant copepod Calanus agulhensis and adult anchovy on the western (r2 = 0.61) and central Agulhas Bank (r2 = 0.46) (Figure 6-10d, e), where these fish spawn each year and rely on this large calanoid copepod as a primary energy source (Richardson et al 1997, 1998). This copepod’s biomass was also negatively related to total pelagic fish biomass (r2 = 0.62 and 0.34 respectively). Interestingly, similar to its recruit biomass on the west coast, the spawner biomass of anchovy on the western Bank was extremely high during 2000. However, unlike recruit biomass, this peak spawner biomass did coincide with the lowest copepod prey biomass on record over the period 1988 – 2000.
Variability of plankton with reference to fish variability, BCLME
111
Northern Benguela The past decade has seen a series of warm events, but uncertainties in the early life history of the majority of Namibian fish stocks have not allowed more than speculation as to mechanisms causing changes in productivity of these stocks. Sardine stocks have been depleted since 1972 and an apparent recovery in the early 1990s was reversed abruptly when a combination of low-oxygen and warm waters occurred over a large area in central and northern Namibia, possibly constraining the spawning habitat of pelagic fish and impacting on the development and survival of their early life-history stages (e.g. Ekau and Verheye 2005). Sardine stocks are still depleted, despite stringent curtailment of fishing activities, in contrast to the southern Benguela stocks, which are at record highs. Shallow-water hake (Merluccius capensis) appeared to have experienced low recruitment through the late 1990s and early 2000s, while deepwater hake (M. paradoxus) have expanded their distribution range northwards in the Lüderitz region (van der Lingen et al. 2006). An extensive zooplankton sample repository (referred to as the SWAPELS collection – South West African Pelagic Egg and Larval Surveys) exists for the region between the Cunene River (17ºS) and Lüderitz (26ºS), collected monthly during the 1970s and 1980s. However, to date no comprehensive analysis of this enormous archive has been undertaken. It has, therefore, not been possible to quantify (suspected) long-term changes in the northern Benguela Current region. Recently however, Hansen et al. (2005) suggested long-term changes in zooplankton community structure since the early 1960s, based on sporadic published accounts of copepod species composition or species dominance in the region (Table 1). A pilot programme of retrospective analyses of some SWAPELS samples from late 1970s and early 1980s was recently initiated through the BENEFIT (Benguela Environment and Fisheries Interactions and Training) Programme (e.g. Tsotsobe et al. 2003; 2004; Mainoane, unpubl. data), and is currently being fast-tracked under the aegis of the BCLME Programme. It allows a preliminary reconstruction of decadescale variability of zooplankton since 1959 for the region off Walvis Bay, historically one of the main fish spawning areas. Although the data are only crude estimates of zooplankton biomass (expressed as settled volume), they do however suggest an overall increasing, long-term trend in coastal zooplankton abundance, as observed in the southern Benguela. After an initial decline over the first two decades, from roughly 200 ml m-2 in the late 1950s (calculated from Kollmer 1963) to <100 ml m-2 in the late 1970s (Tsotsobe et al. 2003; 2004) and early 1980s (Mainoane, unpubl. data), there appeared to be a marked increase in zooplankton over the next two decades to 400-600 ml m-2 in 2000 and thereafter (Cloete, unpubl. data).
Modelling Biological modelling studies in the Benguela Current have ranged from the empirical, through simulation of events after upwelling and comparison with real data, to size-
112
L. Hutchings et al.
Table 6-1. Long-term changes in dominance of calanoid copepod species off Namibia over the past four decades (from Hansen et al. 2005)
Years of sampling
1961, 1962
1976
1985, 1986, 1988
2000
Calanoides carinatus
Calanoides carinatus
Calanoides carinatus
Metridia lucens
2
Paracalanus “parvus”
Paracalanus “parvus”
Rhincalanus nasutus
Calanoides carinatus
3
Metridia lucens
Metridia lucens
Metridia lucens
Rhincalanus nasutus
4
Centropages brachiatus
Paracalanus scotti
Centropages brachiatus
Centropages brachiatus
5
–
–
–
Rank: 1
Reference
R. nasutus* C. brachiatus*
Unterüberbacher
Brenning
Timonin et al.
Hansen et al.
(1964)
(1985)
(1992)
(2005)
* both species equally dominant
based simulations where real data only confuse a beautiful picture. Cochrane et al. (1991) simulated a 27-day anchor station study where there were two cycles of upwelling followed by growth, decay, recycling and sedimentation (Figure 6-11). This simulation highlighted the lack of information on photosynthesis-irradiance curves or microbial activity. Moloney et al. (1991) and Painting et al. (1993) produced generalized size-based models of the theoretical course of events following upwelling and stabilization (Figure 6-12). Both field-based (Probyn 1992) and modelling-based analyses show the importance of recycled nutrients and the microbial loop to maintain phytoplankton growth on the shelf, despite the large volumes of new nutrients upwelled into the euphotic zone. In the vertical dimension, profiles of pigment distribution in relation to thermocline depth, chlorophyll a levels and area were derived from field data using neural network techniques (Silulwane et al. 2001; Richardson et al. 2002). More recently, E. Machu (IRD, pers. comm.) has developed a NPZ (NitrogenPhytoplankton-Zooplankton) model linked to the ROMS (Regional Oceanic Modelling System) model configuration for the southern Benguela (termed PLUME), where the outputs in terms of phytoplankton look similar to satellite-derived pigment
Variability of plankton with reference to fish variability, BCLME
113
a
b
c
Figure 6-11. Simulation of trends observed in plankton variability during the 1987 anchor station study in St Helena Bay: observed and computed biomasses of (a) diatoms, (b) mesozooplankton, and (c) macrozooplankton. P = mean computed mass for the 28 days, O = mean observed mass, and RMSPE = root mean square percent error between computed and observed (after Cochrane et al. 1991).
114
L. Hutchings et al.
Figure 6-12. Simulations of the succession of groups after upwelling, based on size-based models (after Painting et al. 1993).
distributions. However, productivity is underestimated relative to observations because cool-water upwelling is reduced and stratification enhanced in the physical model output. V. Kone (IRD, pers. comm.) is working on an NP2Z2 model to account for small and large phytoplankton and zooplankton. The main plankton (both phytoplankton and zooplankton) patterns, namely the very productive west coast of South Africa and the enriched South Coast, are retrieved by the 3D coupled physical/biogeochemical model. The vertical distribution of phytoplankton is also
Variability of plankton with reference to fish variability, BCLME
115
captured by the model. Surface chlorophyll a concentrations are about two times lower than measured by satellite. Validation of the simulated seasonal variability has been achieved through this study. C. Lett (IRD, pers. comm.) has also used the ROMS/PLUME model configuration with an IBM particle-tracking mode to delineate suitable areas of enrichment, concentration and retention of anchovy in the southern Benguela. This conceptual framework states that, from favourable spawning habitats, eggs and larvae would be transported to and/or retained in places where food originating from enrichment areas would be concentrated. Maps of enrichment, concentration and retention are of considerable interest (and debate!) for comparison with observed fish spawning areas, and larval transport and nursery areas. Using a similar approach, a potential nursery area for sardine on the south coast has been described by Miller et al. (in press). Results from these modelling exercises stimulated an investigation of the pre-recruit abundance of sardine and anchovy on the eastern-central Agulhas Bank in 2005. As the model outputs converge with observations, numerical experiments simulating changes in the current patterns likely to be associated with ENSO events or global warming can be conducted to improve scenario planning for future climate change. Models of feeding energetics of pelagic fish were derived from a combination of laboratory-based and field-derived estimates and showed a clear trophic distinction between anchovy and sardine (James et al. 1989; van der Lingen 2002). They demonstrated that, despite the obvious green colour and abundant phytoplankton cells in anchovy and particularly sardine stomach contents, the bulk of the carbon and nitrogen was derived from mesozooplankton or microzooplankton, with sardines feeding on smaller prey than anchovy. This has large implications for changes in fish distribution, climate and upwelling intensity. Plagányi et al. (1999) simulated the interactions of shoals of anchovy recruits, migrating southwards through the St Helena Bay region, feeding on a large copepod population, which grew in a pulsed, patchy environment at satellite-derived temperatures and phytoplankton concentrations (Figure 6-13). Using real data collected during a monthly monitoring transect off St Helena Bay and satellite data from SeaWiFS, R. Fairbairn (Southamptom Univ., pers comm.) has shown that the phytoplankton levels in St Helena Bay are so consistently high that at the scale of the model the zooplankton just grows exponentially. Thus for the model to show the observed decline of zooplankton through winter months, one has to consider fish predation, advective losses, size or micro-distribution of phytoplankton cells available to zooplankton, or migration to deep water. This assimilative model should help monitoring efforts to track growth conditions for pelagic fish recruits on the West Coast. The trophodynamic models based on production and consumption (Shannon and Jarrre-Teichman 1999) provide useful insights into future scenarios (Shannon et al. 2000). However, considerable lumping and tweaking of zooplankton and phytoplankton biomass and productivity levels are required to achieve balances. The biggest drawback appears to be the estimates of sedimentation, recycling and offshore losses, which are difficult to estimate in the field. The trophodynamic approach emphasizes interactions at the top end of the food web, and fluctuations at the lower
116
L. Hutchings et al.
trophic levels appear considerably dampened, so only a change of several orders of magnitude can disturb a system. Nevertheless the trophic balance approach and its dynamic versions are valuable tools for comparison of different regimes within (e.g. Shannon et al. 2003) and between (e.g. Jarre-Teichmann, A. and V. Christensen. 1998, Shannon and Jarre-Teichmann 1999; Moloney et al. 2005) upwelling systems, and highlight gaps or inconsistencies in the observational records. The well-defined life history of anchovy and sardine, with distinct spawning, transport and nursery areas has facilitated modelling efforts for recruitment forecasting. Bloomer et al. (1994) used a rule-based model, which incorporated wind frequency and velocity, SST and spawner biomass on the South and West coasts to estimate recruitment strength. The model worked well for all years between 1984 and 1991, except for 1989. An elaboration of this approach was the expert system approach (Cochrane and Hutchings 1995; Korrûbel et al. 1998; Painting and Korrûbel. 1998, Painting et al. 1998). This technique utilized a number of environmental and biological parameters on the western Agulhas Bank and on the West Coast. If selected parameters exceeded certain prescribed thresholds, the likelihood of average, betterthan-average and worse-than-average recruitment was estimated. The final likelihood was based on the proportion of positive indicators. This initiative has lapsed as it required considerable field effort and, more importantly, an eastward shift in distribution of anchovy to east of Cape Agulhas violated some of the assumptions underlying the Korrubel et al. 1998 model, which assumed that most of the anchovy spawning occurred on the western Agulhas Bank. New efforts utilizing satellitederived parameters and patterns (Richardson et al., 2003b; C. Roy, IRD, pers. comm.) and new assumptions (Miller et al. in press) indicate that further work on this aspect of forecasting is required, or we must fall back on the patently wrong but best-we-can-do assumption that the best predictor of the future is the past, and that environmental variability will result in random recruitment levels within certain bounds. The IBM approach, coupled to the ROMS/PLUME model configuration, has been used to examine assumptions and critical issues for survival during the early life history of pelagic fish (Mullon et al. 2002; 2003; Huggett et al. 2003, Miller et al.,in press). Skogen (2003) has applied his large- scale model of the Benguela region (Skogen 1999) to the drift of anchovy larvae in terms of variable flow or distribution of spawning to indicate suitable areas for spawning. Other modelling initiatives underway include the initiation and development of HABs (Pitcher and Weeks 2006), low oxygen on the shelf (Monteiro et al.,2006) and empirical expert system warnings of rock lobster walkouts (van der Lingen et al. 2006). CONCLUSIONS Forecasting and modelling requirements in the Benguela region have been concentrated in the SW Cape and there is a need to focus on changes in other parts of the Benguela, in tandem with basic field data to verify the model outputs. The
Variability of plankton with reference to fish variability, BCLME
117
Figure 6-13. Schematic diagram indicating the basic structure of a model of juvenile anchovy feeding on patchy copepod prey (after Plagányi 1995).
118
L. Hutchings et al.
Figure 6-14. (a) Potential utility of a recruitment expert system for assisting with a recommendation on total allowable catch (after Cochrane and Hutchings 1995). (b) Observed and predicted recruitment using (top) a deterministic and (bottom) probabilistic expert system (after Korrûbel et al. 1998).
Variability of plankton with reference to fish variability, BCLME
119
productivity on the Agulhas Bank is strongly influenced by wind mixing, internal waves and intrusions of Agulhas Current water (Shannon et al. 1984) or removals of shelf water during meanders of the Agulhas Current. The ROMS model indicates just how influential the Agulhas Current is on the West Coast, south of the Orange River at 30˚S, yet the behaviour of the Agulhas Current needs further attention, particularly its influence on the shelf regions. The northern Benguela is interesting in that the productive area in terms of a suitable food habitat is much wider than the current pelagic and demersal fish habitat. Regenerated nutrients fuel much of the inner shelf productivity so nitrate is available for new production far offshore, 100-200 km from the upwelling sources. The current depressed state of fish productivity relative to plankton productivity in the northern Benguela, in contrast to the southern Benguela, requires both modelling and extensive field investigations. Satellite imagery has greatly improved the spatial integration possible in the Benguela system with respect to phytoplankton, wind and temperature, and modelling of potential primary production is now possible. Assimilation of satellite data into hydrodynamic models and coupled IBM models allow simulation of the development of both zooplankton and ichthyoplankton populations. As realism improves relative to observational data, integration over time should see the initiation of prognoses for contrasting wind and thermal conditions. Simultaneously, the expert system approach should be pursued to enable short-term forecasts of pelagic fish recruitment to be made. REFERENCES Andrews, W.R.H. and L. Hutchings. 1980. Upwelling in the southern Benguela Current. Progress in Oceanography 9: 1-81. Arashkevich, E.G., A.V. Drits and A.G. Timonin. 1996. Diapause in the life cycle of Calanoides carinatus (Krøyer), (Copepoda, Calanoida). Hydrobiologia 320: 197-208. Armstrong, D.A., H.M. Verheye and A.D. Kemp. 1991. Short-term variability during an anchor station study in the southern Benguela upwelling system: Fecundity estimates of the dominant copepod, Calanoides carinatus. Progress in Oceanography 28: 167-188. Auel, H., W. Hagen, W. Ekau and H.M. Verheye. 2005. Metabolic adaptations and reduced respiration of the copepod Calanoides carinatus during diapause at depth in the Angola-Benguela Front and northern Benguela upwelling regions. Afr. J. Mar. Sci. 27:653-657. Bakun, A. 1996 – Patterns in the Ocean” Ocean Processes and Marine Population Dynamics. University of California Sea Grant Program, San Diego, California, USA, in cooperation with Centro de Investigaciones Biologicas de Noroeste, la Paz, Mexico: 323pp. Barange, M. 1990. Vertical migration and habitat partitioning of six euphausiid species in the northern Benguela upwelling system. J. Plankt. Res. 12:1223-1237. Barange, M. and S.C. Pillar. 1992. Cross-shelf circulation, zonation and maintenance mechanisms of Nyctiphanes capensis and Euphausia hanseni (Euphausiacea) in the northern benguela Upwelling system. Cont. Shelf Res. 12(9): 1027-1042. Barange, M., S.C. Pillar and I. Hampton. 1998. Distribution patterns, stock size and life-history strategies of Cape horse mackerel Trachurus trachurus capensis, based on bottom trawl and acoustic surveys. In Benguela Dynamics: Impacts of Variability on Shelf-Sea Environments and their Living Resources. Pillar, S.C., C.L. Moloney, A.I.L. Payne, and F.A. Shillington ,eds. South African Journal of Marine Science 19:433-447.
120
L. Hutchings et al.
Barange, M., I. Hampton and B.A. Roel. 1999. Trends in the abundance and distribution of anchovy and sardine on the South African continental shelf in the 1990s, deduced from acoustic surveys. South African Journal of Marine Science 21:367-391. Barlow, R., H. Sessions, M. Balarin, S. Weeks, C. Whittle and L. Hutchings. 2005. Seasonal variation in phytoplankton in the southern Benguela: pigment indices and ocean colour. Afr. J. Mar. Sci. 27(1): 275-287 Bloomer, S.F., K.L. Cochrane and J.G. Field. 1994. Towards predicting recruitment success of anchovy Engraulis capensis Gilchrist in the southern Benguela system using environmental variables: a rulebased model. South African Journal of Marine Science 14:107-119. Borchers, P. and L. Hutchings 1986 - Starvation tolerance, development time and egg production of Calanoides carinatus in the Southern Benguela Current. J. Plankt. Res. 8:855-874. Boyd, A.J., Shannon, L.J., Schülein, F.H. and J. Taunton-Clark. 1998. Food, transport and anchovy recruitment in the southern Benguela upwelling system off South Africa. In Global Versus Local Changes in Upwelling Systems. Durand, M.-H., P. Cury, R. Mendelssohn, C. Roy, A. Bakun and D. Pauly, eds.. Collection Colloques et Seminaires, ORSTOM, Paris: 195-209. Boyer, D.C., Boyer, H.J., Fossen, I. and A. Kreiner. 2001. Changes in abundance of the northern Benguela sardine stock during the decade 1990-2000, with comments on the relative importance of fishing and the environment. In A Decade of Namibian Fisheries Science. Payne, A.I.L., S.C. Pillar and R.J.M. Crawford, eds. S. Afr. J. mar. Sci. 23:67-84. Brenning, U. 1985. Structure and development of calanoid populations (Crustacea, Copepoda) in the upwelling regions off North West and South West Africa. Beiträge zur Meereskunde 52:3-33. Brown, P.C. and L. Hutchings. 1985. Phytoplankton distribution and dynamics in the southern Benguela Current. In International Symposium on the Most Important Upwelling Areas off Western Africa (Cape Blanco and Benguela)[, Barcelona, 1983] 1. Bas, C., R. Margalef, and P. Rubiés, eds.. Barcelona; Instituto de Investigaciones Pesqueras: 319-344. Cochrane, K.L., A.G. James, B.A. Mitchell-Innes, G.C. Pitcher, H.M.S. Verheye and D.R. Walker.. 1991 Short-term variability during an anchor station study in the southern Benguela upwelling system. A simulation model. Prog. Oceanog. 28: 121-152. Cochrane, K.L. and L. Hutchings. 1995. A structured approach to using biological and environmental parameters to forecast anchovy recruitment. Fisheries Oceanography 4(2): 102-127. Colijn, F., U. Tillmann and T. Smayda, eds. 1998. The temporal variability of plankton and their physicochemical environment. Proceedings of an ICES Symposium held in Kiel, Germany, 19-21 March 1997. ICES Journal of Marine Science 55(4): 557-823. Cury, P., A. Bakun, R.J.M. Crawford, A. Jarre-Teichmann, R.A. Quiñones, L.J. Shannon, and H.M. Verheye. 1999. Small pelagics in upwelling systems: patterns of interaction and structural changes in “wasp-waist” ecosystems. ICES Journal of Marine Science, Symposium Edition 57(3): 603-618. Cury, P. and L.J. Shannon. 2004. Regime shifts in upwelling ecosystems: Observed changes and possible mechanisms in the northern and southern Benguela. Prog. Oceanog. 60: 223-243. De Decker, A. and F.J. Mombeck. 1964. South African contribution to the International Indian Ocean Expedition. 4. A preliminary report on the planktonic copepoda. Investl Rep. Div. Sea Fish. 51: 10-67. Demarcq, H., Barlow, R.G. and F.A. Shillington. 2003. Climatology and variability of sea surface temperature and surface chlorophyll in the Benguela and Agulhas ecosystems as observed by satellite imagery. African Journal of Marine Science 25: 363-372. Ekau, W. and H.M. Verheye. 2005. Oceanographic fronts and low oxygen conditions controlling distribution patterns of early life stages of fish in the Benguela and southern Angola currents. Afr. J. Mar. Sci.27(3): 629-639 Hagen, E. 1985. Meso-scale upwellings off Namibian coast. In International Symposium on the Most Important Upwelling Areas off Western Africa (Cape Blanco and Benguela)[, Barcelona, 1983] 1. Bas, C., R. Margalef and P. Rubiés, eds. Barcelona; Instituto de Investigaciones Pesqueras: 161-179. Hampton, I. 1992. The role of acoustic surveys in the assessment of pelagic fish resources on the South African continental shelf. In Benguela Trophic Functioning. Payne, A.I.L., K.H. Brink, K.H. Mann, and R. Hilborn, eds. South African Journal of Marine Science 12:1031-1050. Hansen, F.C., Cloete, R.R. and H.M. Verheye. 2005. Seasonal and spatial variability of dominant copepods along a transect off Walvis Bay (23˚S), Namibia. Afr. J. mar. Sci. 27 (1): 55-63. Huggett, J.A. 2003. Comparative ecology of the copepods Calanoides carinatus and Calanus agulhensis in the southern Benguela and Agulhas Bank ecosystems. Ph.D. thesis, University of Cape Town, 232 pp. + 2 Appendices.
Variability of plankton with reference to fish variability, BCLME
121
Huggett, J.A., Boyd, A.J., Hutchings, L. and A.D. Kemp 1998 – Weekly variability of clupeoid eggs and larvae in the Benguela jet current: implications for recruitment. In Benguela Dynamics: Impacts of Variability on Shelf-Sea Environments and their Living Resources. Pillar, S.C., Moloney, C.L., Payne, A.I.L. and F.A. Shillington, eds.. South African Journal of Marine Science 19:197-210. Huggett, J. A., P. Freon, C. Mullon, and P. Penven. 2003. Modelling the transport success of anchovy Engraulis encrasicolus eggs and larvae in the southern Benguela: the effect of spatio-temporal spawning patterns. Mar Ecol. Prog. Ser. 250: 247-262. Huggett, J.A. and A.J. Richardson 2000 – A review of the biology and ecology of Calanus agulhensis off South Africa. ICES Journal of Marine Science 57:1834-1849. Hutchings, L. 1992 – Fish harvesting in a variable, productive environment – searching for rules or searching for exceptions? In Benguela Trophic Functioning. Payne, A.I.L., K.H. Brink, K.H. Mann, and R. Hilborn, eds. South African Journal of Marine Science 12:297-318. Hutchings, L. and J.G. Field 1997. Biological oceanography in South Africa, 1896-1996: observations, mechanisms, monitoring and modelling. In A Century of Marine Science in South Africa. Payne, A.I.L. and J.R.E. Lutjeharms, eds.. Transactions of the Royal Society of South Africa 52(1): 81-120. Hutchings, L. and G. Nelson 1985. The influence of environmental factors on the Cape pelagic fishery. In International Symposium on the Most Important Upwelling Areas off Western Africa (Cape Blanco and Benguela)[, Barcelona, 1983] 1. Bas, C., R. Margalef, and P. Rubiés, eds. Barcelona; Instituto de Investigaciones Pesqueras: 523-540. Hutchings, L., M. Barange, S.F. Bloomer, A.J. Boyd, R.J.M. Crawford, J.A. Huggett, M. Kerstan, J.L. Korrûbel, J.A.A. De Oliveira, S.J. Painting, A.J. Richardson, L.J. Shannon, F.H. Schülein, C.D.van der Lingen and H.M. Verheye. 1998. Multiple factors affecting South African anchovy recruitment in the spawning, transport and nursery areas. In Benguela Dynamics: Impacts of Variability on Shelf-Sea Environments and their Living Resources. Pillar, S.C., C.L. Moloney, A.I.L. Payne, and F.A. Shillington, eds.. South African Journal of Marine Science 19:211-225. Hutchings, L., C.D. van der Lingen, M. Griffiths,M.R. Roberts, L.E. Beckley, and S. Sundby. 2001. Spawning on the edge: Spawning grounds and nursery areas around the South African coast. Mar. Freshwat. Res. 53:307-318. James, A. G. 1988 - Are clupeid microphagists herbivorous or omnivorous? A review of the diets of some commercially important clupeids. S. Afr. J. mar. Sci. 7:161-177. James, A. G. and K. P. Findlay. 1989. Effect of particle size and concentration on feeding behaviour, selectivity and rates of food ingestion by the Cape anchovy Engraulis capensis. Mar. Ecol. Prog. Ser. 50(3):275-294. James, A.G., T.A. Probyn, and L. Hutchings. 1989. Laboratory-derived carbon and nitrogen budgets for the omnivorous planktivore Engraulis capensis Gilchrist. J. Exp. Mar. Biol. Ecol. 131:125-145. Jarre-Teichmann, A. and V. Christensen. 1998. Comparative modelling of trophic flows in four large upwelling ecosystems: global vs local effects. In Global vs Local Changes in Upwelling Systems. Proceedings of the First International CEOS Meeting, Monterey, USA, 6-8 September 1994. Durand, M.-H., P. Cury, R. Mendelssohn, C. Roy, A. Bakun, and D. Pauly, eds. ORSTOM; Paris: 423-443. Kollmer, W.E. 1963 – The pilchard of South West Africa. Notes on zooplankton and phytoplankton collections made off Walvis Bay. Investl Rep. Mar. Res. Lab. S. W. Afr. 8: 78 pp. Korrûbel, J.L., S.F. Bloomer, K.L. Cochrane, L. Hutchings and J.G. Field. 1998. Forecasting in South African pelagic fisheries management: the use of expert and decision support systems. In Benguela Dynamics: Impacts of Variability on Shelf-Sea Environments and their Living Resources. Pillar, S.C., C.L. Moloney, A.I.L. Payne, and F.A. Shillington, eds. South African Journal of Marine Science 19: 415-423. Lluch-Belda, D., R.J.M. Crawford, T. Kawasaki, A.D. MacCall, R.H. Parrish, R.A. Schwartzlose and P.E. Smith. 1989. World-wide fluctuations of sardine and anchovy stocks: the regime problem. South African Journal of Marine Science 8:195-205. Lluch-Belda, D., R.A. Schwartzslose, R. Serra, R.[H] Parrish, T. Kawasaki, D. Hedgecock, and R.J.M. Crawford. 1992. Sardine and anchovy regime fluctuations of abundance in four regions of the world oceans: a workshop report. Fish. Oceanogr. 1(4): 339-347. Loick, N., W. Ekau, and H.M. Verheye. 2005. Water body preferences of dominant calanoid copepod species in the Angola-Benguela frontal zone. Afr. J. Mar. Sci. 27(3):597-608. Miller, D.C.M., C.L. Moloney, C.D. van der Lingen, C. Lette, C. Mullon and J.G. Field. (accepted) Modelling the effects of physical-biological interactions and spatial variability in spawning and
122
L. Hutchings et al.
nursery areas on recruitment of sardine in the southern Benguela ecosystem. Journal of Marine Systems. Mitchell-Innes, B. A. and G. C. Pitcher 1992 - Hydrographic parameters as indicators of the suitability of phytoplankton populations as food for herbivorous copepods. In Benguela Trophic Functioning. Payne, A. I. L., Brink, K. H., Mann, K. H. and R. Hilborn, eds.. S. Afr. J. mar. Sci. 12:355-365. Moloney, C.L., J.G. Field and M.I. Lucas. 1991. The size-based dynamics of plankton 2. Simulations of three contrasting southern Benguela foodwebs. J. Plankt. Res. 13: 1039-1092 Moloney, C.L., A. Jarre, H. Arancibia, Y.-M. Bozec, S. Neira, S., J-P Roux and L.J. Shannon. 2005. Comparing the Benguela and Humboldt marine upwelling ecosystems with indicators derived from inter-calibrated models. ICES J. mar. Sci. 62: 493-502. Mullon, C. P. Cury and P. Penven 2002 – Evolutionary individual-based model for the recruitment of anchovy (Engraulis capensis) in the southern Benguela. Can. J. Fish. Aquat. Sci. 59(5): 910-922 Mullon, C., C. Parada, C.D. van der Lingen and J.A. Huggett.2003 – From particles to individuals: modelling the early stages of anchovy (Engraulis capensis/ encrasicolus) in the southern Benguela. Fish. Oceanogr., 12: 396-406. Nelson, G. 1992 - Equatorward wind and atmospheric pressure spectra as metrics for primary productivity in the Benguela system. In Benguela Trophic Functioning. Payne, A. I. L., Brink, K. H., Mann, K. H. and R. Hilborn (Eds). S. Afr. J. mar. Sci. 12: 19-28. Painting, S.J., Moloney, C.L. and M.I. Lucas 1993 – Simulation and field measurements of phytoplankton-bacteria-zooplankton interactions in the southern Benguela upwelling region. Mar. Ecol. Prog. Ser. 100: 55-69. Painting, S.J., Hutchings, L., Huggett, J.A., Korrûbel, J.L, Richardson, A.J. and H.M. Verheye 1998 – Environmental and biological monitoring for forecasting anchovy recruitment in the southern Benguela upwelling region. Fisheries Oceanography 7(3/4): 364-374. Painting, S.J. and J.L. Korrûbel. 1998. Forecasts of recruitment in South African anchovy from SARP field data using a simple deterministic expert system. In Benguela Dynamics: Impacts of Variability on Shelf-Sea Environments and their Living Resources. Pillar, S.C., C.L. Moloney, A.I.L.Payne and F.A. Shillington, eds.. South African Journal of Marine Science 19: 245-261. Parada, C.E. 2003 – Modeling the effects of environmental and ecological processes on the transport, mortality, growth and distribution of early stages of Cape anchovy (Engraulis encrasicolus) in the Benguela system. PhD thesis, University of Cape Town, 125 pp. Parrish, R.H., A. Bakun, D.M. Husby and C.S. Nelson. 1983. Comparative climatology of selected environmental processes in relation to eastern boundary current pelagic fish reproduction. In: G.D. Sharp and J. Csirke (eds). Proceeding of the Expert Consultation to Examine Changes in Abundance and Species Composition of Neritic Fish Resources, San Jose’, Costa Rica, April 1983. FAO Fish. Rept. 291(3):731-777. Perry, R.I, H.P. Batchelder, D.L. Mackas, S. Chiba, E. Durbin, W. Greve and H.M. Verheye. 2004. Identifying global synchronies in marine zooplankton populations: issues and opportunities. ICES J. mar. Sci. 61:445-456. Peterson, W. T. and S. J. Painting. 1990. Developmental rates of the copepods Calanus australis and Calanoides carinatus in the laboratory, with discussion of methods used for calculation of development time. J. Plankt. Res. 12(2): 283-293. Pillar, S.C. 1986 – Temporal and spatial variations in copepod and euphausiid biomass off the southern and south-western coasts of South Africa in 1977/78. South African Journal of Marine Science 4:219229. Pillar, S.C., V. Stuart, M. Barange and M.J. Gibbons. 1992. Community structure and trophic ecology of euphausiids in the Benguela ecosystem. In Benguela Trophic Functioning. Payne, A.I.L., K.H. Brink, K.H. Mann and R. Hilborn, eds. South African Journal of Marine Science 12: 393-409. Plagányi, É.E. 1995. A model of copepod population dynamics in the southern Benguela upwelling region. Ph.D. thesis, University of Cape Town, South Africa: 299 pp. Plagányi, É.E., L. Hutchings, J.G. Field and H.M. Verheye. 1999. A model of copepod population dynamics in the southern Benguela upwelling region. Journal of Plankton Research 21(9): 1691-1724. Postel, L. 1990 – Die Reaktionen des Mesozooplanktons, speziell der Biomasse, auf küstennahen Auftrieb vor Westafrika. Meereswissenschaftliche Berichte 1: 127 pp. Probyn, T.A. 1992 - The inorganic nitrogen nutrition of phytoplankton in the southern Benguela: new production, phytoplankton size and implications for pelagic foodwebs. In Benguela Trophic
Variability of plankton with reference to fish variability, BCLME
123
Functioning. Payne, A. I. L., Brink, K. H., Mann, K. H. and R. Hilborn, eds. S. Afr. J. mar. Sci. 12: 411-420. Richardson, A.J., Mitchell-Innes, B.A., Fowler, J.L., Bloomer, S.F., Verheye, H.M., Field, J.G., Hutchings, L. and S.J. Painting. 1998. The effect of sea temperature and food availability on the spawning success of Cape anchovy Engraulis capensis in the southern Benguela. In Benguela Dynamics: Impacts of Variability on Shelf-Sea Environments and their Living Resources. Pillar, S.C., C.L. Moloney, A.I.L. Payne, and F.A. Shillington, eds.. South African Journal of Marine Science 19: 275-290. Richardson, A.J., M.C. Pfaff, J.G. Field, N.F. Silulwane and F.A. Shillington. 2002. – Identifying characteristic chlorophyll a profiles in the coastal domain using an artificial neural network. Journal of Plankton Research 24: 1289-1303. Richardson, A.J. and H.M. Verheye. 1998. The relative importance of food and temperature to copepod egg production and somatic growth in the southern Benguela upwelling system. Journal of Plankton Research 20(12): 2379-2399. Richardson, A.J. and H.M. Verheye. 1999. Growth rates of copepods in the southern Benguela upwelling system: the interplay between body size and food. Limnology and Oceanography 44(2): 382-392. Richardson, A.J., H.M. Verheye, J.G. Field, S.M. Payne and E. Wright. 1997. Assessment of the food availability to Cape anchovy during their spawning season. South African Journal of Marine Science 18: 113-117. Richardson, A.J., H.M. Verheye, B.A. Mitchell-Innes, J.L. Fowler and J.G. Field. 2003a. – Seasonal and event-scale variation in growth of Calanus agulhensis (Copepoda) in the Benguela upwelling system and implications for spawning of sardine Sardinops sagax. Marine Ecology Progress Series 254: 239251. Richardson, A.J., C. Risien and F.A. Shillington. 2003b. Using self-organizing maps to identify patterns in satellite imagery. Prog. Oceanog. 59: 223-239. Roy, C., S. Weeks, M. Rouault, G. Nelson, R. Barlow and C. van der Lingen. 2001. Extreme oceanographic events recorded in the southern Benguela during the 1999-2000 summer season. South African Journal of Science 97:465-471. Schwartzlose, R.A., J. Alheit, A. Bakun, T.R. Baumgartner, R. Cloete, R.J.M. Crawford, W.J. Fletcher, Y. Green-Ruiz, E. Hagen, T. Kawasaki, D. Lluch-Belda, S.E. Lluch-Cota, A.D. MacCall, Y.Matsuura, M.O. Nevarez-Martinez, R.H. Parrish, C. Roy, R. Serra, K.V. Shust, M.N. Ward and J.Z. Zuzunaga. 1999. Worldwide large-scale fluctuations of sardine and anchovy populations. S. Afr. J. mar. Sci. 21:289-347. Shannon, L.J., P.M. Cury and A. Jarre. 2000. Modelling effects of fishing in the southern Benguela ecosystem. ICES Journal of Marine Science, 57(3): 720-722. Shannon, L. J. and A. Jarre-Teichmann. 1999. Comparing models of trophic flows in the northern and southern Benguela upwelling systems during the 1980s. In Ecosystem Approaches for Fisheries Management. Fairbanks; University of Alsaka Sea Grant, AK-SG-99-01: 527-541. Shannon, L.J., C.L. Moloney, A. Jarre and J.G. Field. 2003. Trophic flows in the southern Benguela during the 1980s and 1990s. J. mar. Syst. 39: 83-116. Shannon, L.V. 1985. The Benguela Ecosystem part 1. Evolution of the Benguela, Physical features and processes. Oceanogr. Mar. Biol. Ann. Rev. 23:105-182. Shannon, L.V., L. Hutchings, G.W. Bailey and P.A. Shelton. 1984. Spatial and temporal distribution of chlorophyll in southern African wasters as deduced from ship and satellite measurements and their implications for pelagic fisheries. S. Afr. J. mar. Sci. 2: 109-130. Shannon, L.V., R.J.M. Crawford, D.E. Pollock, L. Hutchings, A.J. Boyd, J. Taunton-Clark, A.Badenhorst, R. Melville-Smith, C.J. Augustyn, K.L. Cochrane, I. Hampton, G. Nelson, D.W. Japp and R.J.Q. Tarr. 1992. The 1980s: A decade of change in the Benguela ecosystem. In Benguela Trophic Functioning. Payne, A.I.L., K.H. Brink, K.H. Mann and R. Hilborn, eds. S. Afr. J. mar. Sci. 12:271-296. Shannon, L.V. and G. Nelson. 1996. The Benguela: large scale features and processes and system variability. In The South Atlantic: Present and Past Circulation. Wefer, G., W.H. Berger, G. Siedler, and D. J. Webb, eds. Berlin; Springer: 163-210. Shannon, L.V. and S.C. Pillar. 1986. The Benguela ecosystem. 3. Plankton. In Oceanography and Marine Biology. An Annual Review 24. Barnes, M., ed. Aberdeen; University Press: 65-170. Shelton, P.A. and L. Hutching.s 1990. Ocean stability and anchovy spawning in the southern Benguela Current region. Fishery Bull., Wash. 88(2): 323-338.
124
L. Hutchings et al.
Silulwane, N.F., A.J. Richardson, F.A. Shillington and B.I. Mitchell-Innes. 2001. Identification and classification of vertical chlorophyll patterns in the Benguela upwelling system and Angola-Benguela Front using an artificial neural network. In A Decade of Namibian Fisheries Science. Payne, A.I.L., S.C. Pillar and R.J.M. Crawford, eds. S. Afr. J. mar. Sci. 23: 37-51. Skogen, M.D. 1999. A biophysical model applied to the Benguela upwelling system. S. Afr. J. mar. Sci. 21:235-249 Skogen, M D., L.J. Shannon and J.E. Stiansen. 2003. Drift patterns of anchovy Engraulis capensis larvae in the southern Benguela, and their possible importance for recruitment. Afr. J. mar. Sci. 25: 37-47 Stuart, V. and H.M. Verheye. 1991. Diel migration and feeding patterns of the chaetognath, Sagitta friderici, off the west coast of South Africa. J. Mar. Sys. 49:493-515. Timonin, A.G. 1997. Diel vertical migrations of Calanoides carinatus and Metridia lucens (Copepoda: Calanoida) in the northern Benguela upwelling area. Oceanol. 37(6): 782-787. Timonin, A.G., E.G. Arashkevich, A.V. Drits and T.N. Semenova. 1992. Zooplankton dynamics in the northern Benguela ecosystem, with special reference to the copepod Calanoides carinatus. In: Benguela Trophic Functioning. Payne, A.I.L., K.H. Brink, K.H., Mann and R. Hilborn, eds. S. Afr. J. mar. Sci. 12: 545-560. Tsotsobe, S., H.M. Verheye and M. Gibbons. 2003. Seasonal, interannual and inter-decadal variability in total zooplankton biomass in the northern Benguela. Annual Symposium of the Zoological Society of South Africa (ZSSA)/South African Society of Aquatic Sciences (SASAqS), University of Cape Town, June 2003. Tsotsobe, S., Verheye, H.M. and M. Gibbons 2004 – Spatio-temporal variation in the abundance and community structure of the major copepod species in the northern Benguela. Annual Conference of the Australian Marine Sciences Association, Hobart, Tasmania, Australia, July 2004. Unterüberbacher, H.K. 1964 – Zooplankton studies in the waters off Walvis Bay with special reference to the Copepoda. Investl Rep. mar. Res. Lab. S.W. Afr. 11: 42 pp. + Plates 2-36. van der Lingen, C.D. 2002 – Diet of sardine Sardinops sagax in the southern Benuela upwelling ecosystem. S. Afr. J.mar. Sci. 24: 301-316. van der Lingen, C.D. and J.A. Huggett 2003. The role of ichthyoplankton surveys in recruitment research and management of south African anchovy and sardine. The Big Fish Bang. Proceedings of the 26th Annual Larval Fish Conference, H.I. Browman and A.B. Skiftesvik (Eds). Inst. Mar. Res. Bergen, Norway. ISBN 82-7461-059-8 Verheye, H.M. 2000 – Decadal-scale trends across several marine trophic levels in the southern Benguela upwelling system off South Africa. Ambio 29(1): 30-34. Verheye, H.M. and J.G. Field 1992 – Vertical distribution and diel vertical migration of Calanoides carinatus (Krøyer, 1849) developmental stages in the southern Benguela upwelling region. Journal of Experimental Marine Biology and Ecology 158:123-140. Verheye, H.M., L. Hutchings, J.A. Huggett and S.J. Painting. 1992. Mesozooplankton dynamics in the Benguela ecosystem, with emphasis on the herbivorous copepods. In Benguela Trophic Functioning. Payne, A.I.L., K.H. Brink, K.H., Mann and R. Hilborn, eds. South African Journal of Marine Science 12: 561-584. Verheye, H.M., L. Hutchings W.T. Peterson. 1991. Life-history and population maintenance strategies of Calanoides carinatus (Copepoda: Calanoida) in the southern Benguela ecosystem. South African Journal of Marine Science 11: 179-191. Verheye, H.M. and A.J. Richardson. 1998. – Long-term increase in crustacean zooplankton abundance in the southern Benguela upwelling region (1951-1996): bottom-up or top-down control? In The Temporal Variability of Plankton and Their Physico-Chemical Environment. Colijn, F., U. Tillmann, and T. Smayda, eds.. ICES Journal of Marine Science 55(4): 803-807. Verheye, H.M., A.J. Richardson, L. Hutchings, G. Marska and D. Gianakouras. 1998. Long-term trends in the abundance and community structure of coastal zooplankton in the southern Benguela system, 1951-1996. In Benguela Dynamics: Impacts of Variability on Shelf-Sea Environments and their Living Resources. Pillar, S.C., C.L. Moloney, A.I.L. Payne and F.A. Shillington, eds. South African Journal of Marine Science 19:317-332. Verheye, H.M., W. Hagen, H. Auel, W. Ekau, N. Loick, I. Rheenen, P.Wencke and S. Jones. 2005. Life strategies, energetics and growth characteristics of Calanoides carinatus (Copepoda) in the AngolaBenguela Front region. Afr. J. Mar. Sci. 27(3): 641-651.
Large Marine Ecosystems, Vol. 14 V. Shannon, G. Hempel, P. Malanotte-Rizzoli, C. Moloney and J. Woods (Editors) © 2006 Elsevier B.V. All rights reserved.
125
7 The Variability and Potential for Prediction of Harmful Algal Blooms in the Southern Benguela Ecosystem Grant C. Pitcher and Scarla J. Weeks ABSTRACT Harmful Algal Blooms (HABs) in the southern Benguela are usually attributed to dinoflagellate species, which constitute a regular component of normal phytoplankton populations. Fundamental to the success of HAB predictive systems is a sound knowledge of their variability. Although the Benguela remains poorly explored in terms of phytoplankton distribution, important biogeographic differences between the northern and southern Benguela, and the West Coast and Western Agulhas Bank have been reported and are reflected in the composition of HABs. The southern Benguela is characterized by clear seasonal trends, and high phytoplankton biomass and productivity during the latter months of the upwelling season can be attributed largely to dinoflagellate populations. Superimposed on the seasonal trend of increasing dinoflagellates and phytoplankton biomass are shorter successional patterns associated with spatial and temporal transitions in water column stratification driven by wind cycles and coastal topography. Understanding the mechanisms that control the transport, concentration and dissipation of dinoflagellate blooms is critical in predicting their coastal impact. For this purpose models of coastal wind-driven upwelling are required to reproduce both across-shelf and alongshore dynamics. Such information stands us in good stead in attempts to predict high biomass dinoflagellate blooms which impact the Benguela through low oxygen and hydrogen sulphide events. Less progress has been made on species-specific prediction fundamental to the prediction of toxin related events. INTRODUCTION The majority of Harmful Algal Blooms (HABs) in the southern Benguela ecosystem are attributed to one or another dinoflagellate species and the harmful impacts are associated with either their high biomass or toxic properties (Pitcher and Calder 2000). High biomass dinoflagellate blooms usually impact the coastal environment through their decay, resulting in the formation of low oxygen water (e.g. Pitcher and Cockcroft 1998), and in some cases the production of hydrogen sulphide (e.g. Matthews and Pitcher 1996). Some dinoflagellate blooms impact marine life through the production of toxins, which may act either directly or by transfer through the foodweb. Often
126
G.C. Pitcher and S.J. Weeks
referred to as red tides, Gilchrist (1914) listed these dinoflagellate blooms as one of the factors causing fluctuations in fish stocks in the Benguela. Toxic species accumulated by shellfish pose a serious health risk to humans usually in the form of Paralytic and Diarrhetic Shellfish Poisoning (Pitcher and Calder 2000). A misleading view regarding HABs is that they are typically irregular, unpredictable events when in fact many toxic species constitute a regularly occurring component of normal phytoplankton populations (Zingone and Wyatt 2005). There are no reasons to suppose that the population dynamics of harmful algal species are in any way distinctive, which means that the study of harmful species can illuminate the ecology of other species and, examples from non-harmful species may assist in explaining harmful bloom dynamics. To assess the potential for prediction it is important to appreciate that HABs involve a wide variety of organisms, physiological and ecological processes, and physical-biological interactions, and that HABs are not marginal biological phenomena. The potential for prediction of HABs in the southern Benguela will vary depending on the requirement to predict harmful impacts associated with high biomass blooms versus the requirement to predict species-specific toxic blooms. The prediction of high biomass blooms is more easily achieved as many major processes, which lead to the growth and decline of phytoplankton populations in response to upwelling have been identified (Hutchings et al. 1995). But attempts to predict named phytoplankton species, harmful or otherwise, pose a greater challenge (Zingone and Wyatt 2005). In this context, the respective roles of physics versus biology, endogenous versus exogenous control, and genetic versus phenotypic responses to driving the occurrence and seasonal succession of harmful algae are of extreme relevance. Harmful events have three requirements: the causative species must be present within a given area, they must reach a critical concentration and, they must manifest their harmfulness (Zingone and Wyatt 2005). Hence, the prediction of HABs must include three mains steps: the assessment of the geographic ranges of harmful species, the determination of the mechanisms underlying temporal fluctuations in the abundance of harmful species, and an understanding of the mechanisms of impact. The important role of environmental variables over a wide range of scales is clearly shown at each of these fundamental levels of HAB development. HAB mitigation operations in the southern Benguela presently consist of monitoring the distribution of high biomass blooms in order to pre-empt marine mortalities, and monitoring toxic organisms and/or toxins in shellfish, to ensure safe seafood consumption and trade (Probyn and Pitcher 2004). The next challenge is represented by the development of alert systems based on automated observations coupled with predictive models that can expand the lead-time to harmful events, allowing more effective mitigation operations. Knowledge of the scales of variability of HABs in the Benguela and of the processes determining such variability is fundamental to the success of these initiatives.
HABs, potential for prediction in southern Benguela ecosystem
127
THE SPATIAL [GEOGRAPHIC] DISTRIBUTION OF HABS The Benguela system is characterised by upwelling circulation along the entire west coast of southern Africa, and is unique in that it is bound at both the equatorward and poleward ends by warm water regimes, notably the Angola-Benguela front in the north and the Agulhas retroflection area in the south (Shannon and Nelson 1996). Cape Agulhas is considered the appropriate southern boundary of the Benguela as upwelling typically extends that far during summer and for this reason the Western Agulhas Bank is considered an integral part of the productive West Coast (Figure 7-1). Monitoring has indicated that the distribution of high biomass dinoflagellate blooms is clearly associated with the upwelling system, with few red tides reported east of Cape Agulhas (Pitcher and Calder 2000).
32°S
Lambert's Bay Elands Bay St Helena Bay Cape Columbine
-30.00
-32.00
33°S
Namaqua Region
West Coast
Cape Peninsula
34°S
False Bay 18.0°E
19.0°E
-34.00 Cape Agulhas
Western Agulhas Bank 15.00
17.00
19.00
Eastern Agulhas Bank
21.00
Figure 7-1. A map of the southern Benguela region. Upwelling extends as far south as Cape Agulhas and the Western Agulhas Bank forms part of the productive West Coast. The positioning of a 24 n.mile transect off Lambert’s Bay is depicted in the insert.
128
(a)
G.C. Pitcher and S.J. Weeks
(b)
Figure 7-2. A five year (July 1998 – June 2003) composite of (a) SST(ºC) and (b) chlorophyll (mg m-3) derived from daily high resolution (1 km) NOAA AVHRR and SeaWiFS ocean colour data respectively (adopted from Weeks 2004). Upwelling processes in the southern Benguela are influenced to a large degree by shelf bathymetry and local topography resulting in the Cape Peninsula, Cape Columbine and Namaqua upwelling cells, which clearly distinguish the southern Benguela. The greater St. Helena Bay region and False Bay are characterized by higher biomass owing to increased residence time and consequent retention. Both regions are known to suffer marine mortalities associated with high biomass blooms.
The distribution of HABs within the southern Benguela upwelling system is clearly related to variability in the physical environment. Whereas continental shelf bathymetry and upwelling winds provide the large-scale upwelling mechanism, local topography and meteorology create an alternating pattern of active and passive upwelling circulations along the coast. Alongshore variability may therefore equate to expected offshore changes. Three upwelling centres may be distinguished in the southern Benguela, all of which coincide with a narrowing of the shelf: the Cape Peninsula (34oS), the Cape Columbine (33oS) and the Namaqua (30oS) upwelling centres (Figure 7-2a; Shannon 1985). These upwelling cells are normally located near regions of cyclonic wind-stress curl and are associated with changes in the orientation of the coastline. The area downstream of Cape Columbine, incorporating the greater St. Helena Bay region, is clearly characterised by high phytoplankton biomass, as identified from remotely sensed ocean colour (Figure 7-2b). Here, the shelf is broad, favouring stratification that appears to be conducive to the development of harmful blooms and their negative impacts. An historical perspective of mortalities associated
HABs, potential for prediction in southern Benguela ecosystem
129
with red tide in this region is given by Cockcroft et al. (2000). False Bay is also identified from ocean colour as an area characterised by high phytoplankton biomass and is another area within the Benguela region known to suffer marine mortalities associated with red tide (e.g. Grindley and Taylor 1964; Brown et al. 1979). Both regions are thought to be characterised by higher residence times and retentive, nearsurface circulation patterns (Holden 1985; Graham and Largier 1997). Although the geographic distribution of harmful events does not necessarily coincide with the distribution of harmful species, information relating to the distribution of harmful species is clearly useful for risk assessment, planning and the prediction of HABs (Zingone and Wyatt 2005). Unfortunately, as with many other regions of the world’s oceans, the Benguela remains poorly explored in terms of phytoplankton distribution and lists of species are incomplete. The seasonality of many species, and their extreme patchiness in upwelling systems impair sampling effectiveness, especially for species that do not reach high concentrations, and the low taxonomic resolution of most phytoplankton monitoring programmes and studies severely limits our ability to determine the species present within the region. Important biogeographic differences between the northern and southern Benguela and the West Coast and Western Agulhas Bank have nevertheless been reported and are reflected in the composition of HABs (Pitcher and Calder 2000; Ruiz Sebastian et al. 2005). The northern Benguela is characterised by the fish-killing dinoflagellate Karlodinium micrum, previously described as Gymnodinium galatheanum, by Braarud (1957) from samples collected off Namibia by Steeman Neilson in 1950 during a fish kill. The distribution of this species does not however, appear to extend into the southern Benguela. Another fish-killing dinoflagellate is the newly described species Karenia cristata (Botes et al. 2003). Initially responsible for large abalone mortalities in the 1980s, this species appears to be restricted to the Western Agulhas Bank. This species shares several characteristics with Karenia brevis and the recently described species from New Zealand waters, Karenia brevisulcata, in that it also produces an aerosol toxin responsible for respiratory and skin disorders. Eighteen dinoflagellate species have been reported to form red tides in the southern Benguela. Only 7 of these species are known to form blooms on both the West Coast and Western Agulhas Bank (Table 7-1). Blooms of Alexandrium catenella, responsible for Paralytic Shellfish Poisoning (PSP) are restricted to the West Coast and have historically been the primary concern of marine scientists and managers. Although some of the earlier possible accounts of PSP in the Benguela date back to the 1880s, confirmed cases of PSP were only described in 1948 (Sapeika 1948). The northern extent of Alexandrium catenella is not clearly defined, but is considered to extend into southern Namibia. An undescribed Alexandrium species has recently been isolated from the Western Agulhas Bank, but is unlikely to render shellfish toxic as it does not appear to form substantial blooms and the cell toxin quota is low (Ruiz Sebastian et al. 2005). The distribution of these species is confirmed by the incidence of PSP-contaminated shellfish, with no recordings of PSP east of Cape Point, while the highest incidence of contaminated shellfish is found as expected in the Namaqua region downstream of the Cape Columbine upwelling cell (Pitcher et al. 2001).
130
G.C. Pitcher and S.J. Weeks
Maximum concentrations of A. catenella cysts in the sediments similarly correspond to this distribution (Joyce and Pitcher, in press). Cell densities of A. catenella in the Benguela can reach many million cells l-1 and these blooms not only render shellfish toxic to consumers, but are also seemingly responsible for fish and shellfish mortalities (Pitcher and Calder 2000). The other common form of shellfish poisoning in the Benguela is Diarrhetic Shellfish Poisoning (DSP), usually attributable to Dinophysis acuminata or Dinophysis fortii (Pitcher and Calder 2000). Dinophysis species often form relatively minor components of blooms dominated by other dinoflagellates, but can nevertheless attain high cell concentrations of the order of a million cells l-1 (Pitcher et al., in prep). For both D. acuminata and D. fortii, cell toxin quota data indicate that these species are only moderately toxic in the Benguela, with okadaic acid the primary toxin. However the region is characterised by a high incidence of DSP-contaminated shellfish owing to the ubiquitous distribution of Dinophysis species on both the West Coast (Figure 7-3) and Agulhas Bank (Pitcher et al., in prep). Table 7-1. Dinoflagellate species responsible for “red tide” on the West Coast and Western Agulhas Bank (Pitcher, unpublished data)
West Coast
Western Agulhas Bank
Akashiwo sanguineum
Yes
Yes
Alexandrium catenella
Yes
No
Alexandrium minutum
Yes
No
Ceratium dens
Yes
No
Ceratium furca
Yes
No
Ceratium lineatum
Yes
No
Dinophysis acuminata
Yes
Yes
Gonyaulax polygramma
Yes
Yes
Gyrodinium zeta
Yes
No
Karenia bicuneiformis
No
Yes
Karenia cristata
No
Yes
Noctiluca scintillans
Yes
Yes
Prorocentrum balticum
Yes
No
Prorocentrum micans
Yes
Yes
Prorocentrum rostratum
No
Yes
Prorocentrum triestinum
Yes
Yes
Protoceratium reticulatum
Yes
No
Scrippsiella trochoidea
Yes
Yes
HABs, potential for prediction in southern Benguela ecosystem
SST
131
D. acuminata
D. fortii
-29
-30
-31
-32
-33
-34
15
16
17
18
15
16
17
18
15
16
17
18
Figure 7-3. Surface distributions of SST, Dinophysis acuminata and Dinophysis fortii in the southern Benguela in March 2001 (scale range: 0 - 4675 cells l-1). Apart from the higher cell densities in St Helena Bay, the distribution of Dinophysis cells appears ubiquitous and confined only by the 18oC isotherm. Thus despite the apparent low toxin quota of Dinophysis cells in the southern Benguela, the widespread distribution of Dinophysis is responsible for a high incidence of DSP-contaminated shellfish in the region (from Pitcher et al., in prep).
SEASONAL INCIDENCE OF HABS The seasonal variation in phytoplankton biomass in the southern Benguela is evident from remotely sensed ocean colour data (Figures 7-4 and 7-5). Phytoplankton biomass is notably lower in the third quarter of the year when winter conditions tend to prevail. Biomass levels increase during the fourth quarter following the first upwelling events of the season, but the highest biomass is observed during the first and second quarters of the year, corresponding to the latter half of the upwelling season. Monitoring of HABs in the Benguela has demonstrated the seasonality of dinoflagellate blooms. Comparison of inshore chlorophyll concentrations on the southern Namaqua shelf with dinoflagellate concentrations illustrates the important contribution of this group of phytoplankton to the observations of high biomass during the latter part of the upwelling season (Figure 7-6). The high productivity of the southern Benguela has in the past generally been ascribed to diatoms (Shannon and Pillar 1986). However Mitchell-Innes et al. (2000) claim that highest productivity is observed in shelf waters on the west coast in late summer and can be attributed largely to dinoflagellate populations. A number of production estimates have been made within these high-biomass dinoflagellate blooms. Brown et al. (1979) estimated a productivity of 405 mg C m-3 h-1 within a Gymnodinium bloom on the south coast, Walker and Pitcher (1991) measured a productivity of 520 mg C m-3 h-1 in a Ceratium furca bloom in St Helena Bay and Mitchell-Innes et al. (2000) also reported
132
G.C. Pitcher and S.J. Weeks
a
b
c
d
Figure 7-4. Seasonal distribution of phytoplankton biomass (mg Chl m-3) as derived from composites of daily high resolution (1 km) SeaWiFS ocean colour data for the period (a) July – September of 1998 – 2002 (b) October – December of 1998 – 2002 (c) January – March of 1999 – 2003 and (d) April – June of 1999 – 2003 (Weeks, unpublished data). Phytoplankton biomass is notably lower in the third quarter of the year, a period dominated by winter conditions. Biomass tends to increase in the fourth quarter following some months of upwelling. The highest biomass is observed during the first and second quarters of the year corresponding to the latter half of the upwelling season.
HABs, potential for prediction in southern Benguela ecosystem
(a)
133
(b)
Figure 7-5. Hofmueller latitudinal plots generated from NOAA AVHRR and SeaWiFS ocean colour data (3 day means) showing the offshore (300 pixels – 288 kms) variation of (a) SST (ºC) and (b) chlorophyll (mg m-3) off Elands Bay over a 5 year period (July 1998 – June 2003) (Weeks, unpublished data). A clear seasonal signal is evident with lower chlorophyll concentrations typically confined to a 1-3 month period corresponding to the months of winter and the first months of the upwelling season. Chlorophyll concentrations exceeding 5 mg m-3 are generally confined to the inner 50 km of the shelf.
productivity estimates of >500 mg C m-3 h-1 for blooms of C. furca on the Namaqua shelf, all of which demonstrate the important role of dinoflagellates in contributing to the high productivity of the southern Benguela. The predictability of dinoflagellates in the southern Benguela is demonstrated by their increasing abundance during the course of the upwelling season as stratification intensifies in a predictable way owing to decreasing winds and increased solar irradiance. These dinoflagellates are normal components of the seasonal succession of phytoplankton and are selected as a response to a specific set of environmental
134
G.C. Pitcher and S.J. Weeks
parameters, which contribute to build up a habitat template. Since these templates are seasonally recurrent, dinoflagellates are recurrent too. It has been further argued that dinoflagellates are a diverse group in terms of their ecological requirements, and that distinct groups of dinoflagellate species may occupy different positions along a gradient of environmental parameters including nutrients, turbulence and euphotic zone depth (Smayda and Reynolds 2001). In these paradigms species succession is driven by exogenous forces, which select phylogenetic or functional groups, or lifeforms that may include harmful species. The formulation of these groups therefore provides the potential for improving our predictive capacity owing to the distinctive ecophysiological properties of these groups significant to their bloom dynamics, habitat selection and successions. However the prediction of single species belonging to these functional groups is less likely to succeed. Time series data of the incidence of selected bloom-forming dinoflagellates on the southern Namaqua shelf depicts the rather dramatic interannual variability in the dominance of any one particular species (Figure 7-7). A range of life-form types appear to be represented by the species known to form blooms on the southern Namaqua shelf indicating the possibility of a high diversity of habitats in this region incorporating nearshore, coastal and shelf environments, including fronts, coastal currents and upwelling events. The toxic species A. catenella and D. acuminata are well known in the region and appear to bloom during most years. Blooms of other species appear more erratic. For example, the appearances of the bloom-forming species Ceratium dens and Gyrodinium zeta during this period of monitoring represent the first recordings of these species in the southern Benguela. Community composition and the dominance of any one species may therefore, in practice, be impossible to predict as the predictability of individual species is consistent with stochastic selection, in that species are often selected as a result of being in the right place at the right time (Smayda and Reynolds 2003). Event related successional patterns The seasonal succession of phytoplankton in upwelling systems follows the general pattern of coastal temperate seas, with diatom dominance in spring, a progressive contribution of heterotrophic components during summer and a major contribution of dinoflagellates in late summer and early fall. However, this typical pattern shows considerable temporal and spatial heterogeneity caused by wind-forcing cycles and different hydrographic structures. The composition of phytoplankton communities is considered to reflect two fundamental selection features, notably life-form and species specific selection, and the importance of small-scale physical processes in effecting selection is observed in the dynamics of the surface mixed layer. Mixing and stratification have therefore been identified as habitat features operative in regulating spatial and temporal differences in phytoplankton assemblages by clustering taxa with shared features. Although the causes of shifts in phytoplankton community structure favouring flagellate taxa and their blooms have yet to be fully resolved (Smayda and Reynolds 2001), stratification
135
40 30 20 10 0
5e6 5e5 5e4 5e3 5e2 1 Jul 98
1 Jul 99
1 Jul 00
1 Jul 01
1 Jul 02
Dinoflagellate Concentration [cells.l -1]
Chlorophyll Concentration [mg.m3]
HABs, potential for prediction in southern Benguela ecosystem
1 Jul 03
Figure 7-6. A time series of (a) chlorophyll concentrations derived from SeaWiFS satellite data as determined by the mean value of a 5x5 pixel box at an inshore location off Elands Bay (Weeks, unpublished data) and (b) dinoflagellate concentrations determined from daily phytoplankton samples collected at a shore-based station at Elands Bay (Pitcher, unpublished data) confirm a seasonal pattern with low winter concentrations. Both chlorophyll and dinoflagellate concentrations tend to increase during the course of the upwelling season with maxima in late summer and autumn. A high degree of coincidence between the time series during the latter part of the upwelling season indicates that high chlorophyll concentrations at that time may be attributed to dinoflagellate populations. High event-scale variability is superimposed on these seasonal trends.
has been recognized as a precondition for the development of most harmful dinoflagellates which are considered particularly susceptible to turbulence, but well adapted to lower energy, stratified conditions (Cullen and MacIntyre 1998; Estrada and Berdalet 1998). Phytoplankton assemblages therefore often present a band-like distribution around upwelling centres. In a typical strong upwelling situation, there may be a phytoplankton-poor inshore region, followed downstream by a band of diatom dominance and by zones with an increasing contribution of other groups, such as dinoflagellates, better adapted to stratified conditions (Margalef 1978). This assemblage structure and species distribution will expand and contract in response to pulses of upwelling and relaxation. Consequently, phytoplankton succession in upwelling systems can be partially re-set in a fairly unique way, whereby upwelling
DINOFLAGELLATES
G.C. Pitcher and S.J. Weeks
5E6 5E5
A.catenella
5E4
5E6 5E5 5E4
C.dens
5E6 5E5 5E4
C.furca
5E6 5E5 5E4
D.acuminata
Cells l-1
Cells l-1
Cells l-1
Cells l-1
Cells l-1
136
5E6 5E5
G.zeta
5E6 5E5 5E4
P.triestinum
Cells l-1
Cells l-1
5E4
5E6 5E5 5E4
Jul92
Jul93
Jul94
Jul95
Jul96
Jul97
Jul98
Jul99
Jul00
Jul01
Jul02
Jul03
Figure 7-7. An eleven-year time series of daily dinoflagellate concentrations, including selected bloomforming dinoflagellates, off Elands Bay on the West Coast (Pitcher, unpublished data). The apparent reduction in the incidence of blooms from July 97 is considered to reflect a lower incidence of sampling owing to the fact that samples were no longer taken over weekends. Dinoflagellates are reliably predictable in that their concentrations peak toward the latter half of the upwelling season. Specific species are less predictable owing to a high stochastic component to the biological response at the specific level. The range of life-form types represented by the species capable of forming blooms indicates a high diversity of habitats in this region.
HABs, potential for prediction in southern Benguela ecosystem
137
leads to intermittent interruptions of succession. The phase to which the succession is re-set will depend on the combination of wind-forcing intensity, degree of water column stratification and the previous phase of succession (Estrada and Blasco 1979). Upwelling-favourable winds in the southern Benguela reach a maximum during spring and summer with modulation of upwelling at periods of 3-6 days in response to wind relaxation or reversal associated with the passage of cyclones south of the continent (Nelson and Hutchings 1983). Therefore superimposed on the seasonal trend of increasing dinoflagellates and phytoplankton biomass, are shorter cycles related to these wind reversals (Pitcher et al. 1992). Studies off the southern Namaqua coast have served to demonstrate these event-associated changes by relating phytoplankton community composition to the degree of stratification of the surface layer (Figure 7-8). The introduction of red tide into the coastal environment is closely associated with the relaxation or reversal of upwelling favourable winds and the warming of inshore waters. Community analyses may be linked to indices of water column stratification, with red tide corresponding to a very high index of stratification. Species successions within the Benguela tend therefore to be associated with spatial and temporal transitions from turbulent to stratified water columns in which there is a shift in the control of production dynamics from a dependence on new nitrogen to regenerated nitrogen (Hutchings et al. 1995). The growth of diatoms typically follows upwelling, but as stratification strengthens and persists, dinoflagellate blooms may follow and typically precede small-sized flagellate blooms after new nutrients become limited. Large blooms of coccolithophorids may characterise these late stages of succession (Weeks et al. 2004). THE TIMING TRANSPORT
OF
HABS:
ACROSS-SHELF
AND
ALONGSHORE
Physical processes transport, concentrate and dissipate HABs. The advection of populations is determined by physical properties which move spatial gradients of phytoplankton populations in the directions of x, y and z (Franks, in press). The importance of these terms to the local changes in the population depends on the strength of the velocities and the strength of the spatial gradients of the population. Interactions with the vertical swimming behaviour of the population may be important particularly in compressing plankton populations. Spatial gradients in phytoplankton populations are also influenced by diffusion, which tends to reduce spatial variability by mixing across gradients. The strength of mixing is determined by the strength of the spatial gradient of the population and the strength of the eddy diffusivities. The model of HABs in local areas of potential impact may therefore reduce to the terms of advection and diffusion. Areas of potential impact are typically inshore and their exposure to HABs may be dependent on the duration or frequency of particular advection events. Our ability to predict HABs in inshore waters may therefore be more dependent on understanding the mechanisms that control the transport, concentration and dissipation of blooms than on the dynamics of the harmful algae in offshore waters.
138
G.C. Pitcher and S.J. Weeks
14 12
Northward 10 8 6 4 2 0 -2 -4
Southward
-6
(a)
-8
Depth (m)
-10
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
22
23
24
25
26
27
28
29
30
20
40
Depth (m)
(b)
20
40
(c) 13
14
15
16
17
18
19
20
21
29
30
MARCH 2001
(d)
Figure 7-8. An inshore time series off Lambert’s Bay for the period 12-30 March 2001 of (a) hourly wind displacement, and (b) temperature and (c) chlorophyll a derived from daily profiles (Pitcher and Nelson (in prep) and (d) multi-dimensional scaling based on Bray-Curtis similarities.
HABs, potential for prediction in southern Benguela ecosystem
139
As depicted in figure 7-8, the first of the time series was dominated by southerly winds, and peaks in these upwelling favourable winds corresponded to particularly low sea surface temperatures inshore. The latter half of the time series was marked by an extended period of wind relaxation and reversal, which corresponded to warming of the surface waters and the development of red tide. Three groupings of surface phytoplankton samples could be distinguished by ordination of samples by MultiDimensional Scaling based on root-root transformed abundances and Bray-Curtis similarities (d). The first group (13, 17, 18, 20-22 March) reflected the assemblages present in cold water following upwelling, and was distinguished by low diatom cell densities in which Thalassiosira species were prominent. The second group of phytoplankton samples (14-16, 19, 23, 24 March) appeared to be characterized by intermediate cell composition and densities. Although species of Thalassiosira were still prominent the small dinoflagellate Gyrodinium zeta dominated the samples and was principally responsible for this grouping. The third group of phytoplankton samples (25-30 March) characterized the warm period during which red tide developed inshore. These samples were characterized by particularly high concentrations of G. zeta. The Thallassiosira species common in the former groups were virtually absent from this group. The community analyses were linked to the physical environment by superimposing an index of water column stratification on the ordination of daily phytoplankton samples (circles of increasing size reflect increasing stratification of the water column). The high degree of concordance between the stratification index and the groupings of phytoplankton clearly demonstrates the importance of the surface boundary layer in determining life-form selection and development in response to different regimes of turbulence. The Thalassiosira dominated community was characterized by a low index of stratification, the samples of intermediate composition were characterized by intermediate values and the red tide samples dominated by G. zeta were characterized by very high indices of stratification. The accumulation of blooms inshore and the formation of red tide in the southern Benguela is driven by synoptic weather patterns, which dictate the across-shelf and alongshore movement of blooms (Pitcher et al. 1995; Pitcher and Boyd 1996; Pitcher et al. 1998). Wind driven flows are well known to be important in the transport of HABs in many areas of the world and trying to understand the effects of temporally fluctuating and spatially variable winds in complicated coastal regions with capes, bays and subsurface topography is a challenging problem. Harmful blooms of motile algae have been observed to appear in stratified coastal waters of the southern Benguela far more rapidly, and to reach concentrations far higher, than can be explained by local growth. It is therefore important to examine these blooms from a quantitative population dynamics perspective that includes not only the population in the region of potential impact, but also the dynamics of the broader populations from which they may originate. In some cases it is appropriate to consider the local population in the region of interest as an integral part of a much larger population. In other cases it is useful to consider the population as being of limited spatial extent and to examine the advection of that population in a Lagrangian manner (Donaghay and Osborn 1997).
140
G.C. Pitcher and S.J. Weeks
The advective term is therefore an important source of variation with time and it is important not to interpret a rapid increase in local concentration as the rapid growth of a local population when it is in fact the result of advection from elsewhere. Transport processes that are likely to influence the population dynamics of HAB species are operative over a range of spatial and temporal scales. The specific hydrographic features, responsible for the advection of blooms need to be identified and quantified in order to predict when these events may be of potential coastal significance. For this purpose models of coastal wind-driven upwelling are required to reproduce both the across-shelf and alongshore dynamics that characterize such systems (Gan and Allen 2002). Across-shelf distribution and transport of blooms Dinoflagellate blooms on the southern Namaqua shelf form downstream of the Cape Columbine upwelling centre. The simultaneous measurement of currents with observations of blooms in this region has enabled across-shelf and alongshore currents associated with the formation of blooms to be identified and quantified. Under conditions of upwelling the surface drift downstream of the Cape Columbine upwelling centre is mainly northerly and dinoflagellate blooms are associated with a region of convergence associated with a weak frontal system demarcating an inshore band of upwelling (Figure 7-9). Clustering of phytoplankton assemblages with observed current patterns is evident and during upwelling the core of the bloom is associated with strong northward flow. Relaxation or reversal of the wind cause this frontal system and associated dinoflagellate bloom to move shoreward. The across-shelf transport of blooms has been observed by underway measurements of fluorescence and currents on consecutive days following a period of wind relaxation (Figure 7-10). This information and understanding will be useful in modelling bloom advection, which may afford prediction of the timing and duration of coastal blooms. Alongshore transport The poleward transport of inshore blooms in the southern Benguela has been demonstrated in a number of studies (Figure 7-11; Pitcher et al. 1998). This transport of blooms is to be expected under conditions of wind relaxation or reversal when an inshore counter current may result in the southward progression of blooms along the Namaqua shelf. This southward advection of red tide is thought to be associated with inshore barotropic flow generated by coastal-trapped waves over the southern over the southern Namaqua shelf (Lamberth and Nelson 1987, Probyn et al. 2000). The general southward progression of blooms during the latter part of the upwelling season results in latitudinal variation in the incidence of blooms (Pitcher and Calder 2000). In the Namaqua region the incidence of blooms increases steadily through the
HABs, potential for prediction in southern Benguela ecosystem
141
STATIONS 14
13
12
11
10
9
8
7
6
5
4
3
2
DEPTH (m)
20 40 60 80
a
DEPTH (m)
20 40 60
DEPTH (m)
80
b
-50
c
-100
d
S4 S2 S3 S6 S5 S7 S8 S13 S12 S14 S10 S9 S11
20
40
60
80
100
Similarity
Figure 7-9. A 24 n.mile transect offshore of Lambert’s Bay on 8th March 2000 depicting vertical sections of (a) temperature (b) chlorophyll a and (c) current vectors (Pitcher and Nelson, in prep). Surface phytoplankton samples were grouped by means of a (d) dendrogram, using group-average clustering from Bray-Curtis similarities on root-root transformed abundance data. Sampling followed the resumption of upwelling winds resulting in the cooling of inshore waters. A lens of warm surface water separated the cooler inshore stations from cooler offshore water. Clustering of phytoplankton assemblages with observed current patterns was evident. The inshore stations [2-8] were dominated by dinoflagellates, whereas diatoms dominated the offshore stations [9-14]. The core of the bloom was located offshore and associated with strong northward flow (>25 cm s-1) centered at stations 6 and 7. The highest biomass (>20 mg chl a m-3) was located in an area of convergence of onshore flow at the outer stations and offshore flow at the inner stations.
142
ST6
G.C. Pitcher and S.J. Weeks
ST5
ST4
ST3
ST2
30 25
DEPTH [m]
-10
(d)
20 15
-20 10 5
-30
(a) 0 ST 6
ST6
ST5
ST4
ST3
ST2
25
DEPTH [m]
-10
ST 4
ST 3
ST 2
(e)
20 15
-20 10 5
-30
(b) 0 ST 6
ST6
ST5
ST4
ST3
ST2
-10
ST 5
ST4
ST 3
ST 2
ST3
ST2
30 25
DEPTH [m]
ST 5
30
(f)
20 15
-20 10
-30
5
(c) 0 ST6
ST5
ST4
Figure 7-10. Measurements of the shoreward advection of a bloom dominated by Gyrodinium zeta, Scrippsiella trochoidea and Prorocentrum triestinum, as determined from ADCP and fluorescence data obtained from an 8 n. mile transect off Lambert’s Bay repeated on the 12th (a and d), 13th (b and e), and 14th (c and f) March 2002, following a period of wind relaxation. On the 12 March surface fluorescence data indicated the location of a bloom approximately 10.5 km offshore forming a band about 1.5 km wide. Current vectors derived from ADCP measurements indicated that the surface bloom was associated with onshore flow of between 20 and 30 cm s-1. Strong shear was observed at around 12 m with weak poleward undercurrents. Twenty-four hours later the bloom had moved to within 2.8 km of the coast and the current flow associated with the bloom was still onshore although current velocities were reduced to between 10 and 20 cm s-1. A day later the bloom had accumulated inshore and very high fluorescence values were observed.
upwelling season, reaching a peak in February and March. At this time blooms are absent further south, until they are advected from the north by currents associated with barotropic reversal during the latter part of the upwelling season (Probyn et al. 2000). Consequently, the highest incidence of HABs south of Cape Columbine is in April and May. The insight generated by these studies into the origin and transport of blooms allows the use of hydrodynamic models in predicting the spatial development and advection of blooms on the southern Namaqua shelf (Pitcher et al. 2004). These blooms are easily detected by means of satellite observations of ocean colour but hydrodynamic models are required in order to predict when and where these events may be of potential significance.
HABs, potential for prediction in southern Benguela ecosystem
143
STATION 5
4
3
2
10 DEPTH (m)
20
30
(a)
-10 DEPTH (m)
-20
-30
(b)
10 DEPTH (m)
20
30
(c)
Figure 7-11. An 8 n. mile transect off Lambert’s Bay on the 4 March 1997 serves to demonstrate the southward flow of an inshore Ceratium-dominated dinoflagellate bloom exceeding 6x106 cells l-1 (adapted from Pitcher et al., 1998). The thermal structure indicated a stratified environment with little across-shelf variability (a). The core of the dinoflagellate bloom as determined by fluorescence was observed at station 3 (c). Station 4 was situated outside of the bloom and a subsurface bloom was evident at Station 5. Current data indicated that the inshore surface water, characterized by the dinoflagellate bloom was flowing southwards while the surface water offshore of the bloom was flowing in a northerly direction (b). Current flow at depth indicated southward flow along the entire transect. Thus despite little variation in the across-shelf thermal structure a strong colour front separated an inshore, southbound, surface dinoflagellate bloom from an offshore northbound subsurface population.
144
G.C. Pitcher and S.J. Weeks
CONCLUSION: THE POTENTIAL FOR PREDICTION HABs are closely linked to the prevailing winds of the Benguela, as wind is a dominant factor at all spatial scales, having a direct influence on large-scale currents, local upwelling and frontal dynamics, and the dynamics of the surface mixed layer. This linkage offers the opportunity for HAB prediction based on our ability to predict weather and climate. The early warning and prediction of algal blooms requires observations to characterize algal variability in relation to environmental factors usually determined or influenced by wind, and models that relate algal population dynamics to the observed properties of the environment. Models can range from empirical predictions to detailed numerical forecasts based on simulations of algal growth and behaviour in hydrodynamic models. Predictive models can be developed and validated only if appropriate observations are available, thus physical, chemical and biological observation systems are essential to the early warning and prediction of algal blooms. Presently decisions relevant to harmful blooms often rely on empirical or conceptual models relating algal population dynamics to environmental forcing. Operational forecasts of algal bloom dynamics are not possible at this time, even though retrospective analyses have been successful at describing important factors that influence the distributions and persistence of algal blooms. However as better observations of algal distributions in relation to environmental factors are accumulated, better models will follow. Eventually predictive capabilities will develop and improve. A new generation of oceanographic instruments providing continuous measurements of many physical, chemical and biological properties from autonomous moorings is now able to provide the observations essential to test and refine these predictions. Progress in modeling HABs is however seriously hampered by the lack of knowledge on the basic mechanisms underlying the development of specific algal blooms. Much scientific effort has focused on the assessment of biogeochemical cycles and global carbon budgets where phytoplankton have been equated with bulk parameters like chlorophyll concentration, fluorescence or remotely sensed ocean colour. This information may stand us in good stead in our attempts to observe and model the high biomass dinoflagellate blooms, which impact the southern Benguela through low oxygen and hydrogen sulphide events. Much less progress has been made on speciesspecific contributions to biogeochemical cycles, on phytoplankton species succession, and on biological and ecological mechanisms, which control species abundance. Biogeochemical cycles seem to vary significantly in relation to the species-specific properties of the dominant organisms, therefore community shifts are likely to affect important marine ecosystem functions, such as carbon fluxes and food web structure, DMS production and climate regulation, nitrogen fixation and fisheries. A reasonably high risk exists that any attempt to formulate reliable budgets and to improve predictive capabilities may be ineffective, unless we develop an adequate knowledge of the organisms playing the games (Zingone and Wyatt 2005). These advances are fundamental to the prediction of toxin related events in the southern Benguela.
HABs, potential for prediction in southern Benguela ecosystem
145
REFERENCES Botes, L., S.D. Sym and G.C. Pitcher. 2003. Karenia cristata sp. Nov. and Karenia bicuneiformis sp. Nov. (Gymnodiniales, Dinophyceae): Two new Karenia species from the South African coast. Phycologia. 42: 563-571. Braarud, T. 1957. A red water organism from Walvis Bay (Gymnodinium galatheanum n. sp.). Galathea Deep Sea Exped. 1: 137-138. Brown, P.C., L. Hutchings and D. Horstman. 1979. A red-water outbreak and associated fish mortality at Gordon’s Bay near Cape Town. Fish. Bull. S. Afr. 11: 46-52. Cockcroft, A.C., D.S. Schoeman, G.C. Pitcher, G.W. Bailey and D.C. van Zyl. 2000. A mass stranding, or “walkout” of west coast rock lobster Jasus lalandii in Elands Bay, South Africa: causes, results and implications. In The Biodiversity Crises and Crustacea, Eds., J.C. Von Kaupel Klein and F.R. Schram, Crustacean Iss. 11: 362-688. Cullen, J.J. and J.G. MacIntyre. 1998. Behaviour, physiology and the niche of depth-regulating phytoplankton. In Physiological Ecology of Harmful Algal Blooms, Eds., D.M. Anderson, A.D. Cembella and G.M. Hallegraeff , Springer-Verlag, Berlin, NATO ASI Series, Vol. G 41, pp. 559-579. Donaghay, P.L. and T.R. Osborn. 1997. Toward a theory of biological-physical control of harmful algal bloom dynamics and impacts. Limnol. Oceanogr. 42:1283-1296. Estrada, M. and D. Blasco. 1979. Two phases of the phytoplankton community in the Baja California upwelling. Limnol. Oceanogr. 24: 1065-1080. Estrada, M. and E. Berdalet. 1998. Effects of turbulence on phytoplankton. In Physiological Ecology of Harmful Algal Blooms, Eds., D.M. Anderson, A.D. Cembella and G.M. Hallegraeff , Springer-Verlag, Berlin, NATO ASI Series, Vol. G 41, pp. 601-618. Franks, P.J.S. In Press. Physics and physical modeling of harmful algal blooms. In: Real-time coastal observing systems for ecosystem dynamics and harmful algal blooms, Eds., M. Babin and J. Cullen, UNESCO. Gilchrist, J.D.F. 1914. An enquiry into fluctuations in fish supply on the South African coast. Mar. Biol. Rep., Cape Town 2: 8-35. Graham, W.M. and J.L. Largier. 1997. Upwelling shadows as near-shore retention sites: the example of the northern Monterey bay. Cont. Shelf. Res. 17(5): 509-532. Gan, J.P. and J.S. Allen. 2002. A modeling study of shelf circulation off northern California in the region of the coastal ocean dynamics experiment-2. Simulations and comparison with observations. J Geophys. Res. 107: art. No. 3184. Grindley, J.R. and F.J. R. Taylor. 1964. Red water and marine fauna mortality near Cape Town. Trans. Roy. Soc. S. Afr., 37(2): 111-130. Holden, C.J. 1985. Currents in St Helena Bay inferred from radio-tracked drifters. South African Ocean Colour and Upwelling Experiment, Ed., L.V. Shannon, Cape Town; Sea Fisheries Research Institute 97-109. Hutchings, L., G.C. Pitcher, T.A. Probyn and G.W. Bailey. 1995. The chemical and biological consequences of coastal upwelling. In Upwelling in the Ocean: Modern Processes and Ancient Records, Eds., C.P. Summerhayes, K.-C. Emeis, M.V. Angel, R.L. Smith and B. Zeitzschel, John Wiley and Sons Ltd., pp. 65-81. Joyce, LB. and G.C. Pitcher. In press. The role of cyst germination in the initiation of Alexandrium catenella blooms on the West Coast of South Africa: characteristics of cyst germination. S. Afr. J. mar. Sci. Lamberth R. and G. Nelson. 1987. Field and analytical drogue studies applicable to the St Helena Bay area off South Africa’s west coast. In The Benguela and Comparable Ecosystems, Eds., A.I.L. Payne, J.A. Gulland and K.H. Brink, S. Afr. J. mar. Sci. 5: 163-169. Margalef, R. 1978. Phytoplankton communities in upwelling areas. The example of NW Africa. Oecologia Aquatica 3: 97-132. Matthews, S.G. and G.C. Pitcher. 1996. Worst recorded marine mortality on the South African coast. In Harmful and Toxic Algal Blooms, Eds., T. Yasumoto, Y. Oshima and Y. Fukuyo, Intergovernmental Oceanographic Commission of UNESCO. Mitchell-Innes, B.A., G.C. Pitcher and T.A. Probyn. 2000. Productivity of dinoflagellate blooms on the west coast of South Africa, as measured by natural fluorescence. S. Afr. J. mar. Sci. 22: 273-284. Nelson, G. and L. Hutchings. 1983. The Benguela upwelling area. Prog. Oceanog. 12: 333-356.
146
G.C. Pitcher and S.J. Weeks
Pitcher, G.C., P.C. Brown and B.A. Mitchell-Innes. 1992. Spatio-temporal variability of phytoplankton in the southern Benguela upwelling system. In Benguela Trophic Functioning, Eds., A.I.L. Payne, K.H. Brink, K.H. Mann and R. Hilborn, S. Afr. J. mar. Sci. 12: 439-456. Pitcher, G., J. Agenbag, D. Calder, D. Horstman, M. Jury and J. Taunton-Clark. 1995. Red tides in relation to the meteorology of the southern Benguela upwelling system. In Harmful Marine Algal Blooms, Eds., P. Lassus, G. Arzul, E. Erard, P. Gentien and C. Marcaillou. Lavoisier. Pitcher, G.C. and A.J. Boyd. 1996. Across-shelf and alongshore dinoflagellate distributions and the mechanisms of red tide formation within the southern Benguela upwelling system. In Harmful and Toxic Algal Blooms, Eds., T. Yasumoto, Y. Oshima and Y. Fukuyo, Intergovernmental Oceanographic Commission of UNESCO, Paris, pp. 243-246. Pitcher, G.C. and A. Cockcroft. 1998. Low oxygen, rock lobster strandings and PSP. In Harmful Algae News, Ed., T. Wyatt, Intergovernmental Oceanographic Commission of UNESCO, 17: 1-3. Pitcher, G.C., A.J. Boyd, D.A. Horstman and B.A. Mitchell-Innes. 1998. Subsurface dinoflagellate populations, frontal blooms and the formation of red tide in the southern Benguela upwelling system. Mar. Ecol. Prog. Ser. 172: 253-264. Pitcher, G.C. and D. Calder. 2000. Harmful algal blooms of the southern Benguela current: a review and appraisal of monitoring from 1989 to 1997. S. Afr. J. mar. Sci. 22: 255-271. Pitcher, G.C., J.M. Franco, G.J. Doucette, C.L. Powell and A. Mouton. 2001. Paralytic shellfish poisoning in the abalone Haliotis midae on the West Coast of South Africa. J. Shellfish Res. 20(2): 895-904. Pitcher, G., P. Monteiro and A. Kemp. 2004. The potential use of a hydrodynamic model in the prediction of harmful algal blooms in the southern Benguela. In Steidinger, K., J.H. Landsberg, C.R. Tomas and G.A. Vargo, eds. Harmful Algae 2002. Florida Fish and Wildlife Conservation Commission, Florida Institute of Oceanography, and Intergovernmental Oceanographic Commission of UNESCO. Pitcher, G.C., M.L. Fernandez and D. Calder. In prep. Observations of the bloom dynamics of okadaic acid producing Dinophysis species and the consequent contamination and depuration of shellfish in the southern Benguela upwelling system. Pitcher, G.C. and G. Nelson. In prep. Red tide on the southern Namaqua shelf of the Benguela upwelling system: the role of convergence, retention and advection. Probyn, T.A., G.C. Pitcher, P.M.S. Monteiro, A.J. Boyd and G. Nelson. 2000. Physical processes contributing to harmful algal blooms in Saldanha Bay, South Africa. S. Afr. J. mar. Sci. 22: 285-297. Probyn, T.A. and G.C. Pitcher. 2004. Monitoring of Biotoxins on the South African Coast. In Harmful Algae Management and Mitigation, Eds., S. Hall, S. Etheridge, D. Anderson, J. Kleindinst, M. Zhu and Y. Zou, Asia-Pacific Economic Cooperation (Singapore): APEC Publication #204-MR-04.2 Ruiz Sebastian, C., S.M. Etheridge, P.A. Cook, C. O’Ryan and G.C. Pitcher. 2005. Phylogenetic analysis of toxic Alexandrium (Dinophyceae) isolates from South Africa: implications for the global phylogeography of the Alexandrium tamarense species complex. Phycologia 44(1): 49-60. Sapeika, N. 1948. Mussel poisoning. S. Afr. med J. 22: 337-338. Shannon, L.V. 1985.The Benguela Ecosystem Part I. Evolution of the Benguela, Physical features and Processes, Ed., M. Barnes, Aberdeen University Press, Oceanogr. Mar. Biol. Ann. Rev. 23: 105-182. Shannon, L.V. and S.C. Pillar. 1986. The Benguela ecosystem Part III. Plankton, Ed., M. Barnes, Aberdeen University Press, Oceanogr. Mar. Biol. Ann. Rev. 24: 65-170. Shannon, L.V. and G. Nelson. 1996. The Benguela: Large scale features and processes and system variability. In The South Atlantic: Past and Present Circulation, Eds., G. Wefer, W.H. Berger, G. Siedler and D.J. Webb, Springer-Verlag, Berlin Heidelberg, pp. 163-210. Smayda, T.J. and C.S. Reynolds. 2001. Community assembly in marine phytoplankton: application of recent models to harmful dinoflagellate blooms. J. Plankton Res. 23: 447-461. Smayda, T.J. and C.S. Reynolds. 2003. Strategies of marine dinoflagellate survival and some rules of assembly. J. Sea Res. 49: 95-106. Walker, D.R. and G.C. Pitcher. 1991. The dynamics of phytoplankton populations, including a red-tide bloom, during a quiescent period in St Helena Bay, South Africa. S. Afr. J. mar. Sci. 10: 61-70. Weeks, S.J. 2004. Specific application of satellite remote sensing to the Benguela ecosystem. PhD, University of Cape Town. Weeks, S.J., G.C. Pitcher and S. Bernard. 2004. Satellite monitoring of the evolution of a coccolithophorid bloom in the southern Benguela upwelling system. Oceanography 17(1): 83-89. Zingone, A. and T. Wyatt. 2005. Harmful Algal Blooms: Keys to the understanding of phytoplankton ecology. 867-926 In The Global Coastal Ocean: Multiscale Interdisciplinary Processes, Eds., A.R. Robinson and K.H. Brink, Harvard University Press, The Sea 13.
Large Marine Ecosystems, Vol. 14 V. Shannon, G. Hempel, P. Malanotte-Rizzoli, C. Moloney and J. Woods (Editors) © 2006 Elsevier B.V. All rights reserved.
147
8 Resource and Ecosystem Variability, Including Regime Shifts, in the Benguela Current System C.D. van der Lingen, L. J. Shannon, P. Cury, A. Kreiner, C.L. Moloney, J-P. Roux, and F. Vaz-Velho ABSTRACT Interannual and decadal-scale variability in abundance, distribution and biological characteristics are described for important living marine resources of the Benguela Current system including small pelagic fish, horse mackerel, hakes, snoek, rock lobster, Cape fur seals, Cape gannets and African penguins. Variability at the ecosystem level for the northern and southern subsystems is also described using trophodynamic indices that track structural changes in the ecosystem. Current understanding and analysis of observed variability in both resources and the ecosystem is reviewed, and the knowledge required for predicting resource and ecosystem variability and the causal factors that need to be considered are discussed. We highlight the need to improve understanding of the processes that are important in the Benguela Current ecosystem, to identify what controls those processes, and to quantify such controls (particularly those acting on lower trophic levels) and the roles of important species in the ecosystem. The kinds of predictions considered possible in the Benguela Current system are examined, and a series of steps is suggested to improve understanding of ecosystem and fisheries dynamics and to monitor key aspects of the ecosystem. INTRODUCTION The Benguela Current system is one of the world’s major upwelling systems, and, as is typical of eastern-boundary upwelling systems, is highly dynamic in nature and characterized by high productivity. Located off the southwest coast of Africa, the Benguela is divided into northern and southern sub-systems that are separated by the permanent upwelling cell at Lüderitz (26°S). The northern Benguela extends from the Angola-Benguela front (usually located between 14°S and 16°S (Shannon et al. 1987) to the Lüderitz upwelling cell, whereas the southern Benguela extends from Lüderitz to the Agulhas Bank off South Africa’s south coast (see Figure 1-1 this volume). Conventionally, the geographic border between Namibia and South Africa at the Orange River mouth in the vicinity of 29°S is adopted as the division between the northern and southern Benguela ecosystems. Detailed reviews of various aspects of the Benguela Current system may be found in Shannon (1985), Chapman and Shannon (1985), Shannon and Pillar (1986), and Crawford et al. (1987). More recent studies for
148
C.D. van der Lingen et al.
the southern Benguela are reported in Payne and Lutjeharms (1997), and for the northern Benguela in Payne et al. (2001), and a historical overview of human activities and their impacts on marine life in the Benguela is provided by Griffiths et al. (2004). Aspects of physical and biological variability exhibited by the Benguela Current system across a wide variety of temporal scales are described by Field and Shillington (2006). The Benguela ecosystem supports large populations of living marine resources that are of substantial economic importance to the region. Fisheries for small pelagic fish species (including anchovy Engraulis encrasicolus, redeye round herring Etrumeus whiteheadi, sardine Sardinops sagax, and the sardinellas Sardinella aurita and S. maderensis), midwater fish species (Cape horse mackerel Trachurus trachurus capensis and Cunene horse mackerel T.t. trecae), demersal fish species (including the hakes Merluccius capensis and M. paradoxus), west coast rock lobster (Jasus lalandii), and fisheries targeting other species, have been well established in both the northern and southern Benguela for the past 50-100 years. Bottom trawling off South Africa was initiated in the early 1900s, and the demand for canned fish triggered by World War II provided the impetus for purse-seine fishing. Most of the catch was taken by local vessels before 1960, but distant-water fleets (e.g. from the USSR, Japan and Spain) that targeted hake and other demersal fish began fishing in the Benguela ecosystem thereafter. Whereas South Africa declared a 200-mile exclusive fishing zone in 1977, distant-water fleets continued to harvest large quantities of living marine resources from Angolan and Namibian waters until the late 1980s (Payne and Crawford 1989). Foreign fishing off Namibia for hakes and horse mackerel ceased after Namibian independence in 1990 (Boyer and Hampton 2001), but substantial fishing by foreign vessels still occurs off Angola. Currently, the fishing sector is important to the economies of all three countries bordering the Benguela, but particularly so for Namibia where this sector contributed 6.4% to that country’s GDP in 2001 (Molloy and Reinikainen 2003). In Angola, the fisheries sector is the third most important economic sector, contributing between 3% and 5% to GDP and providing products for both local and export consumption. Fishing contributed 0.4% to South Africa’s GDP in 1998. In addition to the species listed above, many other fish, bird and marine mammal species together comprise the abundant fauna of the Benguela Current ecosystem. The aim of this chapter is to provide an overview of resource and ecosystem variability in the Benguela ecosystem. For the purposes of this review we have extended the definition of the Benguela ecosystem in order to include important resources that occur or may be harvested outside the boundaries indicated above. For example, whereas only the southernmost part of Angola falls within the Benguela ecosystem, most Sardinella spp. occur and are caught to the north of the Angola-Benguela front, but catch data for this resource is not spatially explicit. Similarly, in South Africa, the bulk of shallow water hake M. capensis occurs and is caught off the south coast. Descriptions are provided of interannual and decadal scale variability exhibited by the economically-valuable marine resources listed above from 1950 to 2003, a period for which a fairly substantial dataset exists. Additionally, variability exhibited by a few of the top predator species in the system is described, including Cape gannet (Morus
Resource and ecosystem variability, including regime shifts, in the BCLME
149
capensis), African penguin (Spheniscus demersus), Cape fur seal (Arctocephalus pusillus pusillus), and snoek (Thyrsites atun). Descriptions of resource variability are grouped into three themes, namely variability in abundance, variability in distribution, and variability in biological characteristics, and examples are given where available. Variability exhibited at the ecosystem level for the northern and southern Benguela is described through indices that proxy ecosystem trophic structure, such as the ratio of demersal to pelagic catches and biomass, plots of catch versus the trophic level of the catch, and the Fishing-in-Balance (FiB) index (see below). Current understanding and analysis of observed variability of resources in both the northern and southern Benguela systems is reviewed, and the knowledge required for predicting resource and ecosystem variability and the causal factors that need to be considered discussed. The need to improve understanding of the processes that are important in the Benguela Current system, to identify what controls those processes, and to quantify such controls (particularly those acting on lower trophic levels) and the roles of important species in the ecosystem is highlighted. The kinds of predictions that are possible in the Benguela are examined, and examples of approaches to predict interannual variability in anchovy recruitment are briefly described. Finally, a series of steps is suggested to improve understanding of ecosystem and fisheries dynamics and to monitor key aspects of the Benguela Current system. RESOURCE VARIABILITY Variability in abundance Catches of most of the important resources of the northern Benguela have shown large reductions over the past 50 years. After peaking at over 1.5 million tons in the late1960s and sustaining levels of around 0.5 million tons between 1960 and 1980, catches of small pelagic fish species in the northern Benguela have dwindled to current levels of less than 100 000 tons (Figure 8-1), most of this being Sardinella spp. Rock lobster catches in the northern Benguela have also decreased, from over 5 000 tons annually before 1970, to around 2 000t annually until 1990, and a few hundred tons since then. Catches of hake in the northern Benguela peaked in the early 1970s (Figure 8-1) but decreased thereafter, and whilst not as severe a decline as those observed for small pelagics and rock lobster, annual hake catches made during the past decade have been less than a third of peak catches made in the early-1970s. However, hake catches have shown a slight increasing trend over the past decade. The large decreases for all these species are in contrast to catches of horse mackerel, which increased during the 1970s, peaked in the 1980s and after declining again have remained relatively stable since the 1990s at around 250-300 000 tons (Figure 8-1). Catches of Cunene horse mackerel are much reduced compared to the 1950s and 1960s (Figure 8-1). In the southern Benguela, rock lobster catches have shown a marked decline whereas catches of the other major resources have remained relatively stable (Figure 8-1). A replacement of sardine by anchovy in catches of small pelagics after the mid-1960s is
150
C.D. van der Lingen et al.
Figure 8-1. Variability in annual catches of important living marine resources in the northern (left panels) and southern (right panels) Benguela for small pelagic fish (anchovy, sardine and Sardinella spp in the northern, and anchovy, redeye round herring, and sardine in the southern Benguela); horse mackerel (Cunene and Cape horse mackerel catches are shown for the northern Benguela, and catches of Cape horse mackerel in the northern and southern are further divided into those taken by the pelagic [or purseseine] and midwater or demersal fleets); hakes (both species combined); snoek (hand-line and trawl catches are shown separately, except for the northern Benguela during 2000-2003 where catches are combined); and rock lobster. Plots are updates to those presented in Griffiths et al. (2004) using data supplied by NATMIRC, MCM, and IIM.
Resource and ecosystem variability, including regime shifts, in the BCLME
151
evident, as is a reduction in the relative contribution to horse mackerel catches by the pelagic fleet after 1970, and an increase in snoek catches in the trawl (demersal and midwater) fishery after 1975. Snoek catches are significantly higher in the southern than in the northern Benguela, whereas the reverse is true for horse mackerel catches, and catches of hakes were substantially lower in the southern compared to the northern Benguela but have been at similar levels for the past 15 years (Figure 8-1). Peak catches of small pelagics, horse mackerel and hakes landed in the southern Benguela were substantially lower than peak catches of those resources landed in the northern Benguela. Currently, however, only landings of horse mackerel are higher in the northern than in the southern Benguela, and whereas catches of small pelagics are currently minimal in the northern Benguela they are at close-to-record highs in the southern Benguela. Reliable and consistent biomass time-series estimated from surveys and/or population assessment models in both systems are available for sardine, hake, seals, and birds (Cape gannets and African penguins; Figure 8-2). Long-term trends in biomass have been consistent with fluctuations in catch for sardine and hake in both the northern and southern Benguela. A substantial decline in the biomass of northern Benguela sardine that occurred during the late-1960s and from which the population has not since recovered is apparent, although there were some signs of a recovery in the early 1990s (Figure 8-2). A decline in the 1960s of the southern Benguela sardine population, and a recovery in the late-1980s and 1990s, is also apparent, with this population currently at a size similar to that estimated before its collapse. Southern Benguela anchovy have shown moderate interannual variability in recruitment over most of the time series, with consequent interannual fluctuations in stock size, but strong recruitment that has led to the large population sizes observed since 2000 (Figure 8-2). Considerable reductions in the estimated biomass of hakes in both the northern and southern Benguela were observed soon after the development of the industrial trawl fisheries. These were initiated around 1965 in the northern Benguela and had already started in 1950 in the southern Benguela (Figure 8-2), and have resulted in current population sizes of around one third and one fifth of those estimated in 1950, for the northern and southern Benguela, respectively. Whereas stock assessment models indicate that hake biomass in the northern Benguela has remained at relatively constant levels around 1 million tons, estimates from swept area surveys (data not shown) show an increasing trend in the early 1990s, but due to adverse environmental conditions between 1993 and 1995 the biomass declined again and reached low levels in 1997 (van der Westhuizen 2001). The biomass of hake in 2003, although slightly improved from 1997, remains at low levels. Fur seal abundance, estimated from aerial censuses of pups, increased in the northern Benguela between the 1970s and the mid 1990s at an annual rate of between 2 and 4%, as a continuation of the recovery from historical overexploitation during the 18th and 19th centuries. Subsequent large interannual fluctuations (Figure 8-2) are linked to variability of food availability and environmental fluctuations (Roux 1998). By comparison, the seal population in the southern Benguela has increased slightly since 1970 and appears to have reached relatively stable levels since 1985.
152
C.D. van der Lingen et al.
Figure 8-2. Variability in the abundance of important living marine resources in the northern (left panels) and southern (right panels) Benguela, derived from direct observations and/or model estimates. For small pelagic fish (sardine in the northern Benguela), biomass estimates before the break are from virtual population analysis (Thomas 1986) and those after the break are from estimates made using acoustic survey data (updated from Boyer et al. 2001). For sardine and anchovy in the southern Benguela, biomass estimates before the break are from virtual population analysis (Armstrong et al. 1983) and those after the break are from estimates made using acoustic survey data (updated from Barange et al. 1999). Estimates of hakes (both species combined, and both time-series) are from population assessment models-Geromont et al. (2000) for the northern Benguela and Rademeyer and Butterworth (2003) for the southern Benguela. Top predators including Cape fur seals are estimated from the number of seal pups observed from aerial photograph censuses. Estimates of Cape gannets are based on the breeding area occupied by gannets at Ichaboe Island in the northern Benguela, and at Malgas Island and Lamberts Bay in the southern Benguela (Crawford 2005). Estimates of African penguins are based on the number of breeding pairs (Crawford and Whittington 2005).
Resource and ecosystem variability, including regime shifts, in the BCLME
153
African penguins and Cape gannets are both breeding species endemic to South Africa and Namibia, and both are classified as Vulnerable (IUCN criteria). Between 1956 and 2000, Cape gannets underwent a dramatic decline in the northern Benguela, as indicated by the tenfold contraction in the area they occupied at breeding localities (Figure 8-2). In the southern Benguela, there has been a steady increase in the gannet breeding area since the early 1980s (Figure 8-2). In the Benguela system overall, the penguin population decreased by about 90% between 1910 and the early 1990s (Crawford et al. 2001). The total number of adult birds, estimated to have been 220 000 in 1978, declined to 179 000 in the early 1990s and has increased again to 201 000 in 2000 (Crawford and Whittington 2005). In the northern Benguela, the penguin population declined by more than 74% between the mid 1950s and the late 1990s (Kemper et al. 2001) and has continued to decline at about 1.8% annually since the mid 1990s (J. Kemper, Avian Demography Unit, UCT, pers. comm.). In the southern Benguela, in contrast, the large decline experienced between the 1950s and the early 1980s was followed by a steady increase during the late1990s. The substantial historic reductions in catches and biomass observed for important resources in the Benguela Current system have been attributed primarily to overfishing, including the collapse of sardine and rock lobster populations, and the substantial decline in hake population size, in both the northern and southern Benguela (Figures 8-1 and 8-2; Griffiths et al. 2004). Many resources are currently at population sizes substantially smaller than those observed 50 years ago, although this is not the case for small pelagics in the southern Benguela. Fishing may also shorten the food chain by effectively removing certain species or age classes from the community, thereby reducing ecosystem resilience (Hutchings 2000) and increasing the time required for recovery of the community from a perturbation such as fishing (Vasconcellos et al. 1997). Mean trophic path length in the northern Benguela foodweb (see Heymans et al. 2004; Moloney et al. 2005) has decreased by 27% since the early 1980s, suggesting that the ecosystem has a reduced capacity to recover from perturbations in the current state in which it is functioning. Environmental impacts that contributed to resource collapse, or acted to retard the recovery of a population, have been hypothesized. Boyer et al. (2001) concluded that unfavourable environmental conditions (e.g. the Benguela Niño that occurred in 1995; see below) were important factors in the decline of the northern Benguela sardine stock observed during the 1990s, although this must have been exacerbated by heavy fishing pressure during 1994 and 1995 (Boyer et al. 2001). However, reduced fishing pressure since 2000 (TACs varying between 0t and 25 000t between 2000 and 2003) has not resulted in a recovery of the stock (see section below on Ecosystem Variability). Trophic modelling studies have suggested that the observed changes in pelagic fish stocks in the southern Benguela were also environmentally driven, with effects propagated up the foodweb via the availability of mesozooplankton prey to anchovy and sardine (Shannon et al. 2004c). Fitting a dynamic model to available catch and biomass data series for the southern Benguela from 1978-2002, Shannon et al. (2004b) estimated that fishing patterns (a form of top-down control) explained only 2-3% of the variability observed in the time-series examined, whereas an environmental forcing function (bottom-up control) applied to phytoplankton
154
C.D. van der Lingen et al.
production explained 4-12% of variability. For small pelagic fish in the southern Benguela, fishing mortality has been maintained at relatively low levels; since the mid1980s, anchovy and sardine fishing mortality ranged between 0.04 and 0.30 y-1 (Cunningham and Butterworth 2004a, b), with the small values occurring when biomass was high in the 2000s. In contrast to most ecosystems in the world, the southern Benguela ecosystem has been managed at moderate levels of fishing. Thus it is not surprising that trophic interactions rather than fishing are believed to be responsible for the observed changes in the southern Benguela ecosystem in the 1980s2000s. In particular the vulnerability of prey to predators could explain about 40% of the observed variability in the time series examined (Shannon et al. 2004b; trophic interactions and the processes controlling trophic flow are discussed in more detail in the section on Predicting Variability). Trophic interactions would be expected to be a dominant feature in upwelling ecosystems, because these systems are typically “waspwaist“ controlled (Cury et al. 2000). Some populations have recovered under conservative management strategies. For example the southern Benguela sardine has shown a remarkable recovery following implementation in the mid-1980s of a conservative fishing policy to rebuild the stock (De Oliveira 2002). However, the decadal-scale changes in abundance observed for sardine in the southern Benguela are characteristic of small pelagic species (Schwartzlose et al. 1999; Lehodey et al., in press), with cycles of alternate dominance of anchovy and sardine. Sardine in the California Current system show a 50-60 year time scale of population expansion and contraction (Baumgartner et al. 1996), a period somewhat longer than that observed for southern Benguela sardine (40 years). Decadal-scale changes in abundance of small pelagic fish and a variety of marine resources from a variety of systems have been linked to long-term changes in environmental forcing, and many studies have demonstrated “that climate-related variability of fish populations is the rule rather than the exception” (Lehodey et al., in press). The current situation of large population sizes of both anchovy and sardine in the southern Benguela seems contrary to the hypothesis of alternating periods of species dominance (Schwartzlose et al. 1999), since both are currently at record high levels. However, data for anchovy exist only from the mid-1960s and we do not know historical combined stock sizes. The southern Benguela has been characterised by large interannual variability in small pelagic fish biomass, and exceptionally high anchovy recruitment in 2000 and subsequent years (Figure 8-2). The strong anchovy recruitment in 2000 was linked to within-season variability in local forcing (SE wind), which minimized advective loss during the critical period for successful transport of eggs and larvae from the Agulhas Bank spawning grounds to the west coast nursery grounds, and then maximized their food environment there through sustained upwelling in late summer (Roy et al. 2001). However, those particular environmental conditions have not since been repeated, yet subsequent anchovy recruitment and hence population size has remained high (Figure 8-2). This suggests that processes other than environmental control of egg and larval survival may have become important determinants of recruitment success (Roy et al. 2002), such as the increased egg production arising from a substantially larger spawner stock or changes in spawner distribution. The continued high anchovy recruitment
Resource and ecosystem variability, including regime shifts, in the BCLME
155
since 2000 is not believed to be sustainable as density-dependent factors are likely to come into play, and evidence of reduced body lipid levels of anchovy recruits from 2000 and 2001 compared to previous years has been documented (van der Lingen and Hutchings 2005). Adverse environmental events that occur at interannual or shorter time scales also impact on resource abundance. Benguela Niños, which occur about once every 10 years and are observed as intrusions of warm, saline surface waters onto the northern Benguela shelf, have been associated with widespread mortalities of sardine, horse mackerel and kob (Argyrodomus inordorus) off the coasts of Angola and northern Namibia (Gammelsrød et al. 1998). Depletion of oxygen levels in the near-shore environment of the southern Benguela have caused rock lobster “walkouts” and have resulted in significant local mortality (Cockcroft 2001). Variability in distribution Small pelagic fish in the Benguela ecosystem have shown extensive variability in terms of the spatial extent of their populations, with a positive relationship between biomass and distributional area reported for southern Benguela anchovy, which expands its distribution at large stock sizes (Barange et al. 1999). Those authors found no such relationship for sardine, possibly arising from the relatively low sardine biomass levels observed during the period of their study. Further analysis that includes high sardine biomass levels has indicated that sardine also expand their spatial distribution at large stock size (J.C. Coetzee, MCM, pers. comm.). The distribution of sardine in the northern Benguela has become very patchy following the decline of the stock (Hampton 2003); historically, the stock was distributed more or less continuously between 25ºS and 16º30’S, but during the past decade the stock has been distributed in a few small patches along the northern part of the Namibian coast. Both seasonal and decadal-scale variability in the distribution of Sardinella spp. off the Angolan coast have been reported. Seasonally, adult S. maderensis move northwards at the beginning of the cold season (winter; June to October) and southwards at the beginning of the warm season (summer; February to May; FAO 2000). On a longer time-scale, the core of the S. aurita population was located off Angola and this species dominated landings during the 1970s, whereas from the mid1980s Angolan landings were dominated by S. maderensis (Binet et al. 2001). Interannual variability in the distribution of the major spawning areas of Benguela small pelagics is evident. In the southern Benguela the western Agulhas Bank was the major anchovy spawning area between the mid-1980s and 1995, but an eastward shift in the distribution of anchovy spawners over the Agulhas Bank has been observed during annual surveys since 1996 (Figure 8-3a), with the bulk of the population observed over the offshore regions of the central and eastern Agulhas Bank (van der Lingen et al. 2002). The spawning location of southern Benguela sardine has shown even larger-scale variability, with both the South African south and the west coasts comprising the major spawning grounds during different periods. Crawford (1981) noted that in the early 1960s, sardine spawned along the west coast as well as on the
156
C.D. van der Lingen et al.
south coast of South Africa, whereas sardine spawning was restricted to the south coast in 1965-1967, coinciding with the decline in sardine abundance. In 1987 and 1988, and from 1994-2000, sardine spawning in early summer occurred principally off the west coast, whereas the central and eastern Agulhas Bank were the principal spawning sites from 1989-1993 and since 2001 (Figure 8-3b; van der Lingen et al. 2001). Given that the distance between these two sites of intense spawning is 6001200 km, this variability represents a substantial spatial shift. The switch from spawning off the south coast to spawning off the west coast that occurred in the mid1990s was seen as a consequence of increased population size, reminiscent of the situation during the 1950s and early 1960s (van der Lingen et al. 2001). The increase in sardine spawning off the west coast was attributed to a change in the relative abundance of sardine and anchovy, which enabled sardine to escape the “school trap” (Bakun and Cury 1999) that previously had forced them to school with the more abundant anchovy and hence adopt their strategy of spawning on the south coast (van der Lingen et al. 2001). However, the recent return to spawning by sardine principally off the south coast, observed since 2001, is currently unexplained. This has occurred in tandem with what appears to be an eastward shift in the sardine population, as evidenced by commercial fishing patterns (Fairweather et al., submitted) and resource surveys (Coetzee et al. 2004). In the northern Benguela, egg distributions of small pelagics are currently contracted, and eggs are observed at substantially lower concentrations than was the case when populations of these species were larger than at present. The main areas of sardine spawning in the northern Benguela were in the vicinity of Walvis Bay in spring and Palgrave Point in summer (Figure 3c; Le Clus, 1990; O'Toole, 1977), and the sardine stock showed some indications of a partial separation, probably of younger and older spawners, into a northern Namibian sub-stock and a central Namibian sub-stock (O'Toole 1977; King 1977; Thompson and Mostert 1974). The production of eggs in the Walvis Bay region declined substantially following the collapse of the sardine stock and a reduction in the age structure of the population in the early 1970s, although spawning still occurred in the waters off Palgrave Point (Crawford et al. 1987). Ichthyoplankton surveys conducted in recent years have indicated that sardine egg and larval abundance in the vicinity of Walvis Bay is indeed very low, much lower than around Palgrave Point (Figure 8-3e), and consistent with the hypothesis that sardine spawning off Walvis Bay has declined in importance or possibly ceased. Anchovy eggs are now located substantially further north than was the case in 1971/72 (Figure 8-3d, f). The contraction in spawning area shown by anchovy and sardine in the northern Benguela at present low population sizes most likely reflects the relationship between abundance and the areal extent of the population’s distribution shown for these species in the southern Benguela. However, intense localized fishing pressure has also been implicated in changes in location of spawning by northern Benguela sardine. Historically, eggs and larvae spawned in the northern areas were believed to move south to a nursery area near Walvis Bay where they recruited into the fishery (Thomas 1986), following which survivors returned to the north where they first spawned. Older sardine returned back to Walvis Bay, while younger spawners remained in the north.
Resource and ecosystem variability, including regime shifts, in the BCLME
157
Figure 8-3. Variability through time in the distribution of small pelagic fish species in the Benguela ecosystem, showing (a) the relative distribution (% of total biomass) of southern Benguela anchovy spawners east and west of Cape Agulhas during pelagic spawner biomass surveys conducted by MCM each November over the period 1985-2003 (updated from van der Lingen et al. 2002); (b) the relative distribution of eggs of southern Benguela sardine during pelagic spawner biomass surveys conducted by MCM each November over the period 1984-2003; the x-axis represents the coastline, which has been “straightened” and divided into five standard strata (A is the west coast north of Cape Columbine, B is between Cape Columbine and Cape Point; C is the western Agulhas Bank between Cape Point and Cape Agulhas, D is the central Agulhas Bank between Cape Agulhas and Mossel Bay, and E is the eastern Agulhas Bank east of Mossel Bay; the thick black, line indicates the approximate position of Cape Point) and the percentage contribution to total egg abundance during the survey is shown for each strata in each year, with contouring used to interpolate between years and strata (updated from van der Lingen et al. 2001); and the distribution of northern Benguela sardine (c) and anchovy (d) eggs observed during Cape Cross surveys conducted monthly from September 1971 to April 1972 (the cumulative total number of eggs collected per station over the 8 consecutive surveys are shown; from King, 1977) compared to sardine (e) and anchovy (f) egg distributions observed during an ichthyoplankton cruise in January 2004 (the number of eggs per sample are shown (data from E. Stenevik, IMR, pers. comm.).
158
C.D. van der Lingen et al.
Since the collapse of the stock in the 1970s this migration pattern is believed to have broken down (Boyer et al. 2001), and it is likely that the previous intense spawning in the central region was mainly by older adults (Crawford et al. 1987). The decline in sardine spawning off Walvis Bay has also been attributed to the removal, via fishing, of larger, older fish (Daskalov et al. 2003), or to a selective change in migratory behaviour in response to heavy fishing pressure in the vicinity of Walvis Bay (Bakun 2001). The consequence is that the stock currently consists of few age-classes (usually dominated by a single cohort) of young fish whose natal area is believed to be in the northern part of the Benguela ecosystem. MacCall’s (1990) basin hypothesis suggests that spawning would be expected to be contracted to the most favourable habitat when abundance (and hence density dependence) is low, and the contraction in spawning by small pelagics in the northern Benguela may reflect this. However, the northern spawning ground of northern Benguela sardine appears to be sub-optimal compared to the southern spawning ground, at least in terms of likely offshore advective loss of spawn products, and hence could be considered as secondary spawning habitat (Bakun 2001). Similarly, modelling studies suggested that at low biomass levels anchovy and sardine in the southern Benguela spawned in sub-optimal areas, at least in terms of transport to the nursery area and possible offshore advection (Shannon 1998). These two examples contrast with MacCall’s (1990) hypothesis, and indicate that changes in the distribution of small pelagic fish at varying population levels are likely to also be influenced by changes in population age structure, and/or genetic makeup. Cape horse mackerel in the southern Benguela exhibited a shift in adult distribution patterns between the 1950-60s and after 1970. Adult horse mackerel comprised a significant portion of landings made by the pelagic fishery off South Africa’s west coast over the period 1950-1965, with annual catches of this species averaging 60 000t (Figure 8-1; Crawford et al. 1987). Now, however, only small (<10 000t) catches of juvenile horse mackerel are taken off the west coast, and the large schools of adult horse mackerel that were targeted there by the pelagic fishery have disappeared. Small catches of adult horse mackerel are caught by the mid-water and demersal trawl fisheries off the south coast, and current understanding of the life history of southern Benguela horse mackerel indicates that the west coast acts now as a nursery ground only, with fish migrating southwards and eastwards onto the Agulhas Bank as they grow (Barange et al. 1998). The adult population that supported moderate catches off the west coast during the initial years of South Africa’s pelagic fishery is assumed to have been a southern extension of the large northern Benguela population, and the local depletion of horse mackerel off the west coast which resulted in changed distribution patterns has been linked to spatially-concentrated fishing effort (Griffiths et al. 2004). Deep water hake Merluccius paradoxus in the northern Benguela have shown variability in their distribution in recent years, with this species now being found further to the north than previously (NORAD-FAO/UNDP 1992; Burmeister 2001; Burmeister et al. 2002). Whereas previous published reports indicated that the abundance of this species was low north of Lüderitz (27°S; Figure 8-4a; Payne 1989), high catches of this species are currently taken north of Walvis Bay, and M. paradoxus
Resource and ecosystem variability, including regime shifts, in the BCLME
159
Figure 8-4. Variability through time in the distribution of some living marine resources of the Benguela, showing (a) the distribution by density of deep-water hake in the northern Benguela from catch data collected between 1977 and 1989 (from Payne 1989) and (b) the distribution of hake (circles represent proportional CPUE [kg/hour] from all positive stations) along the Namibian coast during surveys conducted in 1992 (left; NORAD-FAO/UNDP 1992) and 2002 (right; Burmeister et al. 2002); and (c) the relative distribution per island (the % of the total number of breeding pairs on all of those islands is shown) of African penguin at six islands in the northern Benguela in 1956, 1978 and 1999 (R.J.M. Crawford, MCM, pers. comm.) and (d) the relative distribution per island (the % of the total number of breeding pairs on all of those islands is shown) of African penguin at four islands and one island group [those in Algoa Bay] in the southern Benguela in 1956, 1979 and 1999 (R.J.M. Crawford, MCM, pers. comm.).
160
C.D. van der Lingen et al.
now comprises the majority of the Namibian hake catch, whereas in the 1980s the majority of the catch was M. capensis (van der Westhuizen 2001). Survey maps from hake biomass surveys in the northern Benguela corroborate this (Burmeister 2001; Figure 8-4b). This northward expansion in the distribution of deep-water hake has been attributed to an increase in population size (Burmeister 2001), hence hake also appear to show population-size effects on their distribution. Shifts in the spatial distributions of predators have been apparent in both the northern and southern Benguela. Off Namibia, the large decline in African penguins has corresponded to a northward shift of breeding colonies; there has been a decrease (from 95% in 1956 to 18% in 1999) in the proportion of African penguins breeding at localities south of Lüderitz. Colonies at Ichaboe and Mercury Islands, north of Lüderitz, were increasing in the early 1990s (Crawford et al. 2001) but this trend was reversed in 1995 at Ichaboe Island (Kemper et al. 2001) and in 1998 at Mercury Island (J. Kemper, Avian Demography Unit, UCT, pers. comm). Colonies of African penguins along the west coast of South Africa decreased in size between 1956 and 1990 (Crawford et al. 2001), but some began increasing again towards the end of the century (e.g. Dassen Island, Figure 8-4d). New colonies of African penguins were established on Robben Island and at two land sites in False Bay (Crawford 1998) during the period of anchovy abundance in the mid-1980s. The colony at Dyer Island has undergone major declines since 1979, whereas colonies in Algoa Bay increased in size until the late 1990s (Crawford et al. 2001) but have recently decreased (Crawford and Whittington 2005). The eastward shift in pelagic fish distribution (Figure 8-3a) in the southern Benguela appears to have played a role in the spatial shifts of African penguins (Figure 8-4d; Crawford 1998). Seals in the northern Benguela have shown large spatial changes, whereas the distribution of seals in the southern Benguela has been less variable. The proportion of pups counted at colonies in the northern Benguela increased at sites north of 25oS from less than 20% until the early 1980s, to 30-34% between 1989 and 1993, and to more than 40% since the mid 1990s (unpublished data, MCM and MFMR). Environmental forcing is also strongly linked to changes in resource distributions at interannual and decadal scales. The occasional expansion of hypoxic bottom waters in the northern Benguela (so called low oxygen events) from their normal location on the inner shelf (Dingle and Nelson 1993; Kristmannsson 1994) to being widespread over the shelf apparently displaces juvenile Cape hakes offshore from their typical inshore habitat, subjecting them to increased mortality from predation by larger hake and by commercial trawling (Hamukuaya et al. 1998). In addition to resulting in faunal mortalities, Benguela Niños have been documented as impacting on the distribution of important northern Benguela resources, forcing Sardinella aurita off Angola (Binet et al. 2001) and sardine off Namibia (Gammelsrød et al. 1998) southwards. A large-scale movement of rock lobster in the southern Benguela, into the kelp forests between Cape Hangklip and Danger Point (on the western Agulhas Bank), was first noted in 1994 (Tarr et al. 1996), and coincided with the disappearance of the entire population of sea urchins Parechinus angulosus there. This massive increase in rock lobster numbers has occurred since the late 1980s, in an area stretching 150km east of
Resource and ecosystem variability, including regime shifts, in the BCLME
161
Cape Hangklip, the main fishing grounds for abalone (Haliotis midae). Mayfield and Branch (2000) verified the substantial increase in rock lobsters in this region through interviews with recreational fishers, diver surveys, and comparison with earlier surveys (in 1980), and reported a negative correlation between large (>68mm CL) rock lobster and sea urchin abundance, due to selective predation by the former on the latter. Reduction in sea urchin abundance has serious negative implications for abalone, since urchins provide shelter for and reduce predation on juvenile abalone (primarily by direct lobster predation) in areas where natural protection (crevices etc.) is minimal (Tarr 2000). Hence substantial changes in community structure have occurred from the mid-1980s to the present, apparently triggered by increases in rock lobster with consequent collapses of sea urchins and hence juvenile abalone. Largescale environmental forcing over the past decade or so has been implicated in the shift in southern Benguela rock lobster to the more southern fishing areas and the movement of rock lobster onto the Cape south coast (Cochrane et al. 2004). Variability in biological characteristics Long-term variability in reproductive characteristics has been reported for many of the Benguela’s living marine resources. Size at sexual maturity of southern Benguela sardine has changed in conjunction with population size, female fish maturing at a larger size at high population levels (1950-1963) and at a smaller size at low population levels (1965-87; Figure 8-5a; van der Lingen et al., submitted). A decrease in the age at maturity of northern Benguela sardine following the stock collapse was reported by Thomas (1986). Change in size at sexual maturity has also been reported for Cape horse mackerel in the northern Benguela, with a steady reduction in this parameter over the period 1977-83 (Wysokinsky 1984) and fish currently maturing at an even smaller length than was the case then (Figure 8-5b). A reduction in length at maturity of rock lobster in the southern Benguela in 1993/94 compared to earlier periods in the 1960s, 1970s and 1980s has also been reported (Figure 8-5c; Cockcroft and Goosen 1995). Another reproductive characteristic, the number of African penguin chicks fledged per nest, has shown an increase from around 0.4 to 1.0 over the period 1989-2001(Figure 5d; Crawford and Dyer 1995). The diets of top predators (snoek, gannets, and seals) in both systems have varied over time. A switch in the relative dominance of sardine and anchovy in the diet of snoek in the southern Benguela occurred around 1964; before this snoek diet was dominated by sardine and after this by anchovy (Figure 8-6a; Crawford et al., unpublished manuscript). Most recent observations (1994-97) showed equivalent dietary occurrence of the two prey species. Similarly, a change in the relative contribution of anchovy and sardine to the diet of Cape gannet in the southern Benguela occurred between 1987/88, before which anchovy clearly dominated gannet diet and after which sardine was generally the major dietary component (Schwartzlose et al. 1999). The diet of seals in the northern Benguela, which was dominated by sardine in the mid20thcentury, has changed dramatically after the collapse of this stock (Mecenero and Roux 2002). Between the mid-1970s and the mid-1980s seal diet was dominated by juvenile horse mackerel and pelagic goby (David 1987), and between 1994 and 2002 it
162
C.D. van der Lingen et al.
Figure 8-5. Variability in reproductive characteristics of some living marine resources of the Benguela ecosystem, showing (a) maturity ogives for female sardine in the southern Benguela for five periods between 1953 and 2003 (from van der Lingen et al., in press); (b) maturity ogives for horse mackerel in the northern Benguela over the period 1977-83 (redrawn from data in Wysokinsky [1984]) and 1999-2004 (data used to plot the ogive are from A. Kanandjembo, NatMIRC, pers. comm.); (c) maturity ogives for female rock lobster from four locations in the southern Benguela in 1993/94 and earlier periods (from Cockroft and Goosen, 1995; reproduced with permission); and (d) reproductive success of African penguin (numbers of chicks fledged per nest) over the period 1987-2001 (updated from Crawford and Dyer 1995; note that the data point for 2000 is excluded because of the negative impact the Treasure oil spill had on penguin breeding success that year).
was dominated by juvenile horse mackerel, juvenile hake and pelagic goby (Sufflogobius bibarbatus) (Mecenero et al. 2006; Figure 8-6c). Many other examples of biological variability in the Benguela Current system have been documented. The length distribution of both Sardinella species caught off Angola increased following a reduction in fishing pressure (Figure 8-7a, b; FAO, 2000). Rock lobster in the southern Benguela showed reductions in somatic growth rates during the 1990s (Pollock et al. 1997). Long-term changes in condition factor of both northern and southern Benguela sardine have been observed, with condition factor being higher
Resource and ecosystem variability, including regime shifts, in the BCLME
163
Figure 8-6. Variability in the diet of top predators of the Benguela ecosystem, showing (a) the relative contribution of anchovy and sardine in the diet of snoek in the southern Benguela at various times during the period 1950-2000 (% frequency of occurrence; data from M.H. Griffiths, formerly MCM, pers. comm.); (b) the relative contribution of anchovy and sardine in the diet of Cape gannets in the southern Benguela (Lambert’s Bay and Malgas Island) over the period 1978-2003 (% mass; updated from Schwartzlose et al. 1999); and (c) the seasonal and interannual variability of the teleost portion of the diet of seals in southern Namibia (numerical percentages), from 1994-2003 (J.-P. Roux, unpublished data).
164
C.D. van der Lingen et al.
Figure 8-7. Variability in various biological characteristics of some living marine resources of the Benguela ecosystem, including an increase in length frequency distributions of (a) Sardinella aurita and (b) S. maderensis caught by the industrial fishery off Angola during periods of heavy (1980-82) and light (1983-1990) fishing pressure (from FAO 2000); and long-term changes in condition factor of (c) northern Benguela sardine (CF anomaly data are from Le Clus 1987 and CF data are from Kreiner et al. 2001 and (d) southern Benguela sardine (from van der Lingen et al., in press).
Resource and ecosystem variability, including regime shifts, in the BCLME
165
at low population size and lower at high population size (Figure 8-7c,d; Le Clus, 1987; Kreiner et al. 2001; van der Lingen et al., in press). Northern Benguela sardine showed an increase in growth rate after the stock had collapsed (Thomas 1985), and also increased mortality rate and reduced age composition at low population size compared to high population size (Fossen et al. 2001). Changes in biological characteristics of living marine resources of the Benguela have been attributed to a variety of causes including fishing pressure, density dependence, and environmental effects including food availability. The removal of larger, older fish from the population through intense fishing pressure will impact fish size and age distributions. For example, the modal length of catches of northern Benguela sardine was 26cm TL between 1952 and 1957 (Matthews 1960), and only 21 cm TL between 1997 and 2000 (Boyer et al. 2001). Reports of compensatory responses by fish to fishing, including reductions in age-at-maturity and/or length-at-maturity, faster growth and increased fecundity, are becoming increasingly common for commercially exploited fish populations in many systems (Rochet 1998, 2000). Density-dependence appears to be strongly linked to variability in several biological parameters of sardine in the Benguela ecosystem, including growth rate (Thomas 1985), age- (Thomas 1986) and size- at maturity (Figure 8-5a; van der Lingen et al., submitted), condition factor (Figure 8-7c, d; Kreiner et al. 2001; van der Lingen et al., in press), and mortality rate (Fossen et al. 2001). The decline in male growth rate (Pollock et al. 1997) and reduction in female size at sexual maturity (Figure 8-6c) reported for southern Benguela rock lobster during the 1990s and shown to be a coast-wide phenomenon (Melville-Smith et al. 1995) has been ascribed to adverse environmental conditions. These were likely to be food limitation (Pollock 1982) and/or oxygen depletion of bottom waters (Pollock and Shannon 1987), and the strong correlation between the reduction in male growth rate and female brood sizes during 1987-1992 (Melville-Smith et al. 1995) suggests that the underlying cause of a slower growth rate was related to a diminished food supply and/or a decline in food quality (Pollock et al. 1997). The protracted and widespread nature of the phenomenon suggests that it was the result of a large-scale environmental perturbation in the Benguela ecosystem, and a decrease in both pelagic and benthic productivity in the southern Benguela associated with the El Niño years of 1990-1993 is believed to be the main cause of the phenomenon (Pollock et al. 1997). Food availability exerts substantial effects on biological characteristics of marine resources. The increased fledging success of African penguins in the southern Benguela over the past decade has been attributed to the increased abundance and availability of anchovy (Crawford and Dyer 1995). Similarly, the numbers of swift terns (Sterna bergii) breeding in the southern Benguela over the period 1987-2000 were significantly related to the combined biomass of anchovy and sardine there (Crawford 2003). Changes in diet of top predators typically reflect relative abundance of forage fish (Figure 8-6, with good correspondence observed between relative anchovy and sardine population sizes in the southern Benguela and their contribution to the diet of snoek and Cape gannet. Similarly, the strong correlation between hake recruitment (estimated from trawl surveys) and the contribution of juvenile hake to the
166
C.D. van der Lingen et al.
diet of seals at Atlas Bay (near Lüderitz) in the northern Benguela suggests that most of the interannual variability in dietary composition arises from variability in prey abundance (Mecenero and Roux 2002). ECOSYSTEM VARIABILITY High amplitude changes in community composition, species abundance and trophic structure are termed regime shifts and are thought to occur largely in response to oceanic and climatic changes (Collie et al. 2004). This broad definition does not cover temporal and spatial dimensions of a regime shift; alternative, more detailed definitions are provided in Table 1 of Jarre et al. (this volume). Changes in ecosystem states are not only environmentally driven but are also known to be responses to altered fishing pressure (Larkin 1996) and ecological and behavioural changes (Cury et al. 2003). Fish and other marine species respond to these changes by altering their population dynamics over 10-30 year periods (Beamish and Mahnken 1999). Despite their (possibly) expected similarities, the northern and southern Benguela have shown distinctly different ecosystem dynamics (Figure 8-1 and Figure 8-8). Differences between the northern and southern Benguela may be related to the fact that the southern Benguela upwelling region is bounded by the shallow Agulhas Bank, comprised of a diverse demersal fish assemblage. Biological resources are affected by dramatically different environmental perturbations in the two systems. For example, the northern Benguela is regularly affected by low oxygen events, and by large-scale warm water events such as Benguela Niños. In recent years, unusually large biomasses of pelagic fish have been attained in the southern Benguela, whereas the pelagic ecosystem in the northern Benguela has collapsed. The ecosystem structure and trophic functioning of the northern Benguela in recent years seems to differ from the way in which the system functioned in the 1970s (Heymans et al. 2004). Between the late 1960s and the late 1970s, anchovy and sardine stocks were replaced by a suite of zooplanktivorous fish including horse mackerel, mesopelagic fish and pelagic goby. It appeared as if sardine may have been starting to recover towards the end of the 1980s and early 1990s, but recovery was curbed by unfavourable environmental conditions (Boyer et al. 2001). Most fish stocks in the northern Benguela underwent large declines towards the end of the 1990s at the time of major environmental anomalies (Boyer and Hampton 2001; Roux 2003). On the other hand, jellyfish Chrysaora hysoscella and Aequorea aequorea have attained large abundances (Boyer et al. 2001) and may have changed the energy flow through the northern Benguela food web (Heymans et al. 2004). Since the 1980s, the northern Benguela appears to have undergone a regime shift (Figure 8-9a, b); unfavourable environmental effects have been exacerbated by heavy exploitation (Boyer et al. 2001), resulting in extreme modifications to the pelagic ecosystem there (Bakun and Weeks 2004). Exploitation of sardine, anchovy and subsequently hake and horse mackerel has taken place in the northern Benguela from 1950 onwards. Since Namibian independence in
Resource and ecosystem variability, including regime shifts, in the BCLME
167
Figure 8-8. Variability in selected trophic characteristics of the northern (left panels) and southern (right panels) Benguela Current ecosystem showing (a, b) the ratio of demersal catches to total catches over the period 1950-2003 (note that catches from Angola are not included for the northern Benguela, and that for the southern Benguela the corresponding demersal:total biomass ratio from a trophic model fitted to available time-series for 1978-2002 is also shown); (c, d) the mean trophic level of annual catches plotted against the Log(catch) for data from 1950-2003; and (e, f) the FiB index over the period 1950-2003 (updated from Cury et al. 2005).
1990, horse mackerel has continued to dominate total catches and there has been a moderate recovery of hake catches, but sardine catches have remained negligible. The combined species catch in each year can be represented on an ecosystem basis in terms of its mean trophic level. Trophic levels (TL) describe the positions of species in the food web. Primary producers are allocated to TL 1 and the TL of a consumer is calculated as the mean TL of the prey it consumes, weighted by the contribution of the prey species to the diet of the predator. Annual mean trophic levels of the catch in the
168
C.D. van der Lingen et al.
Figure 8-9. Changes in ecosystem structure and relative abundance of dominant living marine resources (a, c; small pelagic fish, horse mackerel, hakes, goby, jellyfish, Cape gannets and Cape fur seals are shown [not scaled to relative size] and the number of individuals approximately reflects relative abundance), and (c, d) average annual catches of small pelagic fish, horse mackerel and hakes by decade or half-decade (the last histogram shows average annual catches for the period 2000-2003) for the northern (upper figures) and southern (lower figures) Benguela. The two ecosystem states shown for the northern and southern Benguela approximate periods around 1970 (left hand side), and the current situation (right hand side).
Resource and ecosystem variability, including regime shifts, in the BCLME
169
northern Benguela reflected catches of sardine until the mid-1960s, increased from 1965 to 1972 as the hake (high TL) fishery developed and sardine catches decreased, and levelled off thereafter due to the continued decline in sardine and the partial replacement of hake catches by horse mackerel (Figure 8-8c). The relative and dimensionless FiB index (Pauly et al. 2000) remains constant when changes in the mean trophic level of the catch are matched by a corresponding increase/decrease in catch size to take account of the higher ecological cost of fishing at higher trophic levels. In general, an increase in the FiB indicates an expansion of the fishery whereas a decrease indicates a collapse of the underlying food web or a geographic contraction of the fishery. Off Namibia, the FiB index increased between 1960 and 1972 (Figure 8-8e), reflecting the rapid expansion of the pelagic fishery followed by growth in the trawl fishery as pelagic catches began to decline. Since 1972, there has been a small but significant decrease in the FiB, as a result of decreased total catches dominated by horse mackerel (having a higher TL than small pelagic fish). In both the northern and southern Benguela, there was an increase in the proportion of the total catches that were comprised of demersal fish since 1950 (Figure 8-8a, b), although the increase was much smaller in the southern Benguela. Since the collapse of the pelagic fisheries, Namibian catches are currently comprised almost entirely of horse mackerel and hakes. Because fish abundance declined in the northern Benguela, predation pressure on fish increased and predator biomass decreased between 1980 and 1999 (Cury et al. 2005a). Resources in the southern Benguela have also varied substantially between 1980 and 2004, but apparently without a shift to a completely new ecosystem state (Cury and Shannon 2004). Despite the differences in fish (particularly small pelagic) stock sizes and catches between 1980 and the mid 1990s, the trophic models used by Shannon et al. (2003) failed to reveal a change in the overall functioning of the southern Benguela between 1980 and 1997. Cury and Shannon (2004) argue that the southern Benguela experienced a pelagic species replacement rather than a clear regime shift to a different ecosystem state (Figure 8-9c, d) because ecosystem functioning remained relatively constant over that period. Those authors interpreted the observed changes as variability within the natural limits of population size and geographical distribution. However, the southern Benguela ecosystem may well have been in a different state prior to the start of industrialized fishing in the 1950s, after which a substantial increase in total catches occurred (see Figure 33 of Griffiths et al. 2004). Fishing can alter marine ecosystem structure (e.g. Pitcher 2000; Pitcher and Pauly 1998), and it is likely that the southern Benguela underwent a rapid shift from a pristine, or nearpristine, state following the advent of heavy fishing. Total catches in the southern Benguela have remained approximately stable since 1978. Sardine dominated catches until the mid 1960s, catches peaking in 1962 (Figure 8-2; De Villiers 1985). As pelagic fish catches declined, catches of hake in the demersal trawl fishery increased, and in the late 1970s, midwater trawls began to target horse mackerel (De Villiers 1985). Anchovy replaced sardine in the pelagic fishery in the 1970s and 1980s. There has been a small decline in the mean TL of catches and the FiB index during the 1990s (Figure 8-8f), reflecting the increase in sardine catches
170
C.D. van der Lingen et al.
as the stock began to recover. Since the late 1990s both anchovy and sardine have been abundant, supporting larger populations of some predators such as seabirds (Figure 8-2), although there is no evidence for a concomitant increase in primary productivity off South Africa between 1997-2003 (H. Demarqc, IRD, pers. comm.). Biomass of pelagic fish predators supported in the period 1990-1997 was 24% larger than that in 1980-1989. The sustainability of these high stock sizes of pelagic fish in the south is uncertain, as density-dependent effects are expected to come into play, and unpredictable environmental events may also occur, influencing resource and ecosystem variability. Large, dominant fish species are sustained at least partly because adult fish keep forage fish at sufficiently low abundances to limit competition with and/or predation on their own juveniles (Walters and Kitchell 2001). This has been termed the “cultivation effect”, and was considered by Barkai and McQuaid (1988) as “predatorprey role reversal.” Cury et al. (2005b) explain how fishing large predatory fish may prevent successful stock rebuilding by releasing top-down control, thereby increasing the depensatory effects of predation on, or competition between small forage fish. Indeed, the two hake species are important predators in both the northern (Roux and Shannon 2004) and southern Benguela (Shannon et al. 2003) ecosystems. When testing an indicator of interaction strength using ecosystem models, hakes emerged as the group having the strongest interaction strength, i.e. a change in the biomass of hakes caused large changes in abundances of other species in the southern Benguela (Shannon and Cury 2003). Similarly, some minimal realistic model (MRM) scenarios of cannibalistic interactions between the two hake species and predation by seals on hake in the southern Benguela suggested that reducing seal populations via culling would result in increased inter-hake cannibalism and hence fewer hakes overall (Punt and Butterworth 1995).
PREDICTING VARIABILITY We predict patterns using knowledge about processes. It is important to understand what processes are important in an ecosystem and what controls them, before being able to make predictions of resource and ecosystem variability. At present in the Benguela ecosystem we have imperfect understanding of past variability, although a number of hypotheses have been proposed to explain observed changes in species abundances. This uncertainty is not surprising, because different components of the ecosystem are likely to be affected in different ways at different times, depending on their state and the state of the ecosystem. For example, a population at a low biomass might respond differently to the same environmental forcing than when it is at a high biomass. For depleted stocks, the effect of recruitment variability is much more important than at high stock levels. As a consequence of these factors, non-stationarity and non-linearity should be expected in ecosystem control processes. In this section we discuss the knowledge required for predicting resource and ecosystem variability, the causal factors that need to be considered, and the kinds of predictions that are possible.
Resource and ecosystem variability, including regime shifts, in the BCLME
171
What controls ecosystem functioning? Identifying and quantifying the controls of ecosystem functioning and the roles of important species in the Benguela ecosystem is necessary because there is no general theory that can be ascribed to the functioning of marine ecosystems. Thus, and unlike in physics, our ability to predict ecosystem trajectories cannot depend on a few simple ‘laws of nature’; it is necessary to understand a complex set of processes that might operate differently under different conditions. There are several means of quantifying controls and their impact on marine systems. Ecosystem manipulations have been used in lakes (Carpenter et al. 2001), and in the marine environment they have been successfully applied to plankton ecosystems to examine the role of iron in controlling primary production (e.g. Boyd et al. 2000). Such manipulation experiments can be used to quantify processes that are assumed to be key for exerting controls. These could include quantifying predation processes through examination of stomach contents and isotopic studies to assign trophic levels to different species (e.g. Sholto-Douglas et al. 1991), or quantifying to what extent environmental effects are propagated up the food web by relating plankton dynamics to fluctuations in small pelagic fish, for example. Behavioural studies can test the extent and consequences of the school trap hypothesis, and reasons for changing migration patterns and distributions of key species. Retrospective studies allow patterns to be identified, and can also be used to test hypotheses. Such studies include meta-analysis (Worm and Myers 2003; Richardson and Schoeman 2004), empirical studies (Cury et al. 2000; Daskalov 2002), multivariate time-series analysis that considers both environmental variables and predator/competitor interactions (Stenseth et al. 2003), and spatial studies that use GIS to identify indicators of interactions (Drapeau et al. 2004; Fréon et al. 2005). The controls that operate in ecosystems can be further investigated using ecosystem models. Several simulation models exist of the Benguela ecosystem, including Ecopath with Ecosim models (Shannon et al. 2003; Heymans et al. 2004; Roux and Shannon 2004) that are based on the foraging arena hypothesis of predator-prey interactions (Walters et al. 1997), and a size-based predation model (OSMOSE; Shin and Cury 2001; Shin et al. 2004). Models of lower trophic levels are in development (Eric Machu, IRD, pers. comm.). Additional models can be constructed, such as dynamic trophic cascades (Herendeen 2004), and inverse models can be used with reconstructions of ecosystem flows (Vezina et al. 2004). Comparative studies among ecosystems (e.g. Moloney et al. 2005) can extend the data and knowledge base for understanding controls within a specific ecosystem. Causal factors influencing resource and ecosystem variability Environment-resource interactions can be synergistic or antagonistic, and changing the composition of the ecosystem will likely change the response of the ecosystem to environmental forcing (see above). It is important to understand controls of lower trophic levels, because these influence food quantity and quality, which can act as
172
C.D. van der Lingen et al.
controlling mechanisms at high trophic levels. Not only must there be a match between an organism and its food in time (Cushing’s [1990] match/mismatch hypothesis), but they should also match in space, and the food must be of a suitable quality (Beaugrand et al. 2003). For example, Winship and Trites (2003) reported that differences in the energy density and digestibility of prey (e.g. low energy gadids versus high energy forage fish) could have large effects on the prey biomass requirements of Alaskan Steller sea lion (Eumetopias jubatus) pups, such that they could starve in the midst of plenty. Cury et al. (2005a) have proposed that an environmental index be developed to quantify bottom-up effects of environmental perturbations on foodwebs. The future effects of climate change are not fully known, and there is insufficient knowledge about local impacts. Removing top-down controls through overexploitation of many demersal fish communities may change the way climate affects fisheries. Thus the interaction between global forces, i.e. the effects of climate and overexploitation, may be nonlinear and can act synergistically. As a result, the changing structure of marine ecosystems could lead to changing recruitment patterns of species that are valuable to fisheries (Beaugrand et al. 2003), and may enhance the overall impact of the environment on the ecosystem. When overexploited, ecosystems appear to be more frequently controlled by bottom-up forces, which makes it difficult to appreciate the true relative contribution of top-down versus bottom-up control forces in ‘balanced’ marine ecosystems. There are hypotheses about some large-scale impacts of climate change in the Benguela ecosystem, including altered wind stress that would enhance coastal upwelling (Bakun 1990; Bakun 1992; Shannon et al. 1996); an increased frequency of occurrence and intensity of Benguela Niños and the advection of warm tropical water (Siegfried et al. 1990); and an increase in sea surface temperature (Roux 2003). The possible impacts of climate change in the Benguela ecosystem can be addressed partly through modelling, by downscaling global models and linking these with threedimensional hydrodynamic models of the region (e.g. PLUME; Penven et al. 2001). Individual-based models can be used to test “scenario” cases and assess the impacts of oceanographic conditions on biology and ecology of key species, including their spatial dynamics (e.g. changes in spawning locality; Mullon et al. 2003). In addition, there might be unexpected interactions between fisheries exploitation and climate change, such as changes in a stock’s resilience to exploitation under different climatic scenarios. A regime shift has been documented for the northern Benguela (Cury and Shannon 2004). Regime shifts (in the northern Benguela and elsewhere) have generally been assumed to be controlled mainly by environmental forcing, but it is possible that they can also be induced by anthropogenic forcing (Steele 2004). For example, in the northern Benguela overexploitation of small pelagic fish (principally sardine) resulted in a radical restructuring of the food web and hence a very changed ecosystem (Cury and Shannon 2004; Roux and Shannon 2004). To date, the effects of pollution in the BCLME region have been mainly localised, although pollution is a serious threat for some species (e.g. oil pollution for seabirds).
Resource and ecosystem variability, including regime shifts, in the BCLME
173
In general, pollution has probably had very little impact on the Benguela ecosystem in the past (Griffiths et al. 2004), but there is concern that its possible impacts in the future are being ignored, because of reduced research effort (O’Donoghue and Marshall 2003).
MAKING PREDICTIONS Predictions in ecosystems, as for any complex systems, are difficult to make, and should rather be replaced by the analysis of risks associated with anthropogenic actions or natural trends that occur in marine ecosystems. A number of risks for the dynamics of marine ecosystems can be associated with the depletion of pelagic fish stocks, which can produce local and global extinction of marine bird populations and to a lesser extent seal populations. In some cases, but not all, such depletions can lead to dysfunctional ecosystems, such as is the case for the northern Benguela ecosystem at present (discussed above). Studies of the northern and southern Benguela illustrate that the identification of bottom-up as well as top-down controls are key for predicting population trajectories. Patterns in marine ecosystems are produced by global change and/or subtle mechanisms that act at different scales and with different strengths depending on the structure of the ecosystem. Thus, overexploitation as well as environmental changes strongly affect populations at different trophic levels, either directly in a predictable way, or indirectly in a largely unpredictable way. The environment plays a major role in influencing recruitment strength for pelagic fish. However, localized (in time and space) environmental events (e.g. those described by Roy et al. 2001) are difficult to predict but can play a major role in structuring (and restructuring) fish populations. In the case of the northern Benguela, it is difficult to predict when a favourable succession of environmental events could contribute to rebuild both sardine and anchovy populations. Daskalov et al. (2003) found that the functional response (through recruitment) of the northern Benguela sardine stock to environmental factors seems to have changed over time. This change could be linked to a shift between two environmental regimes between the mid- and late-1980s. Alternatively, it could be the result of fisheries-induced changes in the age structure of the stock that resulted in a change in spawning habitat. Both hypotheses however point to some non-linear biological response of the stock to either environmental variability, or fishery pressure, or both. Because of strong trophic interactions, drastic changes at the pelagic fish population level can be expected to produce changes at higher trophic levels, e.g. for marine birds and mammals. The roles of those large populations of pelagic fish at intermediate trophic levels that fluctuate radically in size have important consequences for the functioning of marine ecosystems. Small pelagic fish in upwelling systems have been shown to play a pivotal role in the food web, so-called “wasp-waist” control (Bakun 1996; Cury et al. 2000). In terms of prediction we can definitively say that the substantial reduction of pelagic fish will lead to the collapses or drastic diminution of a number of other populations that prey upon them. At the other end of the spectrum is the role of pelagic fish in controlling their zooplankton prey, sometimes with feedback effects. For example, Boyer et al. (2001) propose a mechanism that could
174
C.D. van der Lingen et al.
maintain the northern Benguela ecosystem in its current state. They propose that phytoplankton are significantly underutilized as a result of the small sardine stock size, and this will increase detritus loads and worsen the low oxygen situation, further negatively affecting sardine spawning. Marine bird numbers are strongly controlled by the abundance of pelagic fish; large pelagic fish populations often lead to improved survival for birds (and conversely). In this particular case predictions at the ecosystem level are possible and are reflected in the joint evolution of pelagic fish and top predators in the northern and southern Benguela. This prediction needs to be tuned spatially, knowing the distribution ranges of both predators and prey; in this case it is necessary to analyse the spatial distribution of the interactions, for example using a GIS (Drapeau et al. 2004; Fréon et al. 2005). Not everything is predictable at the level of an ecosystem. Cape fur seals, a protected species in the southern Benguela, have become abundant in the Benguela ecosystem overall (present population size about 1.5 to 2 million animals) and are killing substantial numbers of different species of seabirds --African penguins, Cape gannets, Cape cormorant Phalacrocorax capensis, bank cormorant P. neglectus and crowned cormorants P. coronatus (David et al. 2003). This predation process has been well known for a long time but was marginal in the past. Together with competition for breeding space, this predation process is now seen to endanger many protected bird populations. This example illustrates that long-term predictions can be made regarding predator-prey relationships but that new processes or novelties can emerge and modify the complex dynamics of interactions. In the northern Benguela the almost complete collapse of pelagic fish resources led to a drastic reduction in the abundance of penguins and gannets. It has also resulted in an outburst of pelagic gobies and a reorganisation of the food-web dynamics, through the collapse of the pelagic food chain. The resulting degraded marine ecosystem, combined with further increases in upwelling intensity, appears to be creating additional sources of greenhouse gas emissions (Bakun and Weeks 2004). Potential predictions can be made from ecosystem control theory (Cury et al. 2003), recognizing that those predictions can be challenged by the complexity of the interactions. Biological/predation controls on ecosystem functioning can occur by means of bottom-up (control by plankton affected directly by the environment), topdown (control by predators) or wasp-waist (discussed above) mechanisms. But why do we observe drastic changes in major components under relatively stable productivity? Why have both sardine and anchovy populations collapsed and not recovered in the northern Benguela? Questions such as these currently do not receive satisfactory answers, emphasizing the fact that prediction in marine ecosystems is in its infancy. That said, attempts to forecast anchovy recruitment strength in the southern Benguela using expert systems were developed around a decade ago. Expert systems take into account qualitative and quantitative data from a number of factors simultaneously, manage well with short data series, take the form of a number of multiple-choice questions and a set of rules that logically infers predetermined decisions based on the
Resource and ecosystem variability, including regime shifts, in the BCLME
175
answers to these questions. The rules take the form: IF (question, answers) THEN (decision). Four expert systems have previously been developed for anchovy recruitment in the southern Benguela region. The first was devised using three environmental variables: wind frequency, wind velocity and sea-surface temperature (Bloomer et al. 1994). This was followed by a survey of expert opinion that formed the basis of both deterministic and probabilistic expert systems for anchovy recruitment (Korrûbel et al. 1998), which predicted “very likely below average” anchovy recruitment for 1994 that was validated by the observation of recruitment that was slightly below long-term average. Then, a rule-based deterministic model to forecast anchovy recruitment was developed using field data from the South African Sardine and Anchovy Recruitment Programme (SARP) conducted in 1994 and 1995 (Painting and Korrûbel 1998). The predicted results compared very favourably with the final estimates of recruitment strength for 1994 and 1995. The most recent expert system (Miller and Field 2002) uses five predictors (covering anchovy spawning, transport and nursery areas) in a deterministic model to provide a qualitative forecast of anchovy recruitment, and this system correctly hindcast recruitment in 89% of an 18 year time-series (1985-2001), including two test years not used in model construction. Apart from that expert system however, none of the others has been applied to recent data, a situation that should be rectified given the recent exceptionally high recruitment estimated for southern Benguela anchovy. A proposed way to update these is given in van der Lingen et al. (this volume). Despite the success in hindcasting anchovy recruitment strength, however, success in making actual predictions remains elusive, suggesting that patterns in anchovy recruitment are not regular and/or that different factors impacting on recruitment may vary from year to year (Hutchings et al. 1998). In addition, the expert systems typically have used variables integrated over a year, whereas Roy et al. (2001) have shown that intra-annual processes can be important in affecting recruitment, and there is a need to develop indices that take that account of episodic events. These indications further emphasize the expectation of nonstationarity and non-linearity in ecosystem control processes. Much recent work on marine ecosystems has focused on the development of indicators that index some aspect of ecosystem state and/or trajectory and that are useful for fisheries management (Cury and Christensen 2005). Such indicators (e.g. FiB; see above) have the potential to simplify and quantify ecosystem complexity, and may enhance understanding of the dynamics of complex systems through their tracking of repeatable patterns (Cury 2004). The identification and development of further indicators for the Benguela Current ecosystem should be accorded a high priority, particularly given the move toward an ecosystem approach to fisheries currently underway in the Benguela current system (see Shannon et al. 2004a). Such indicators could then be used to monitor key aspects of the ecosystem, which may permit analysis of risks associated with anthropogenic actions or natural trends that occur in the Benguela Current system (Shannon et al. 2006). The potential for using indicators to describe and quantify changes in the state of the Benguela Current ecosystem is discussed in detail by Jarre et al. (this volume).
176
C.D. van der Lingen et al.
A WAY FORWARD We propose that five steps should be followed to permit an improvement in our ability to analyse risks associated with anthropogenic actions or natural trends that occur in the Benguela Current system, and possibly make predictions. They are: 1. Identify different possible states of the ecosystem, including current states, and evaluate their desirability on a socio-economic basis. Ecosystem states might include pristine, small pelagic fish biomass high, small pelagic fish biomass very reduced, etc. 2. Improve understanding of the control mechanisms that operate in different possible ecosystem states (as described above), through, for example, studies that quantify trophic interactions between species and the functional response of predators to varying prey fields. 3. Develop and test hypotheses about factors causing ecosystem changes, and assess if or how the relative importance of hypothesized mechanisms may vary in time and space. Similarly, spatio-temporal variability in the relative importance of hypothesized mechanisms impacting on resource variability should be assessed (e.g. Cochrane and Hutchings, 1995 for southern Benguela anchovy), although it is likely that in any one year several plausible, but different, factors impact on recruitment success, making it difficult to posit generalized hypotheses (Hutchings et al. 1998). 4. Identify the suite of indicators necessary to describe and quantify ecosystem changes; expert systems may also be useful in this regard. 5. Synthesise the information provided by the indicators (e.g. through an expert system approach that can capture many aspects of change) and thereby monitor key aspects of the ecosystem. By identifying changes in ecosystem state, improve our ability to analyse risks and possibly make predictions. CONCLUSIONS Marine ecosystems are complex dynamic systems which present stable or semi-stable states. Indeed, most of these systems worldwide have displayed remarkable resilience, despite having been subjected to large scale perturbations both of anthropogenic (fishing, pollution, habitat modification) and natural origin (climate and oceanic variability). This resilience and general lack of widespread chaotic behaviour in marine systems implies the existence of sets of controls or feedback mechanisms which tend to bring back the system towards its equilibrium state (or “local attractor”) following a disturbance (see Jarre et al., this volume). The most important of these feedback mechanisms are the demographic responses of the individual stocks to changes in abundance (density dependence effects, spawning stock/recruitment relationships) mediated by predator/prey relationships (trophodynamic controls). Our (imperfect) theoretical and empirical knowledge of these mechanisms forms the basis of our efforts to manage fisheries and predict fish stock and ecosystem trajectories. However, after the system has been subjected to a change of state, or regime shift, there is evidence that it is under the influence of an altered set of feedback mechanisms. Therefore, it is very unlikely that predictions at the ecosystem level in
Resource and ecosystem variability, including regime shifts, in the BCLME
177
one state could be inferred from observations of the properties of this system while in a previous state. Priority should therefore be given to understanding and predicting the limits to the stable equilibria we observe and the effects of the factors (natural as well as anthropogenic) that cause regime shifts and contribute to maintaining the system in an altered state. We consider it likely that the resource and ecosystem changes described above for the BCLME result from a combination of factors, including both fishing and environmental variability. Many different, plausible hypotheses have been proposed to explain the mechanisms responsible for resource and/or ecosystem variability; these are not necessarily contradictory. Mechanisms can operate simultaneously, but they can operate on different components of the ecosystem at the same or different times and places, and in isolation or synergistically. What we currently lack is a consistent approach that simultaneously combines and “tests” all hypotheses; our five-step approach aims to do just that. The importance of improving our understanding of the causes of variability in fish stocks and the need for further research into the factors that affect recruitment have been recognized, and the need to develop management advisory techniques that take account of dynamic recruitment fluctuations in a more effective way highlighted (Rothschild and Shannon 2004). Those authors note that addressing the implications of regime shifts in fisheries management is directly relevant in the move towards an ecosystem approach to fisheries management. ACKNOWLEDGEMENTS We would like to thank all those who supplied us with data and information used in the construction of the figures, including Robin Leslie (MCM) for updates to hake, horse mackerel and snoek catches from demersal and midwater trawling; Rob Crawford, Leshia Upfold and Bruce Dyer (all MCM) for data on seabird abundance, gannet diet composition and African penguin breeding success; Janet Coetzee (MCM) for data on pelagic fish biomass levels and the distribution of anchovy spawners; Herman Oosthuizen and Mike Mëyer (MCM) for seal counts; Steve Brouwer (MCM) for rock lobster data; Chris Wilkie (MCM) for hand-line snoek catches; Rebecca Rademeyer (MARAM) for hake biomass estimates; Angie Kanandjembo (NatMIRC) for recent information on Cape horse mackerel size at maturity and catch data; Kolette Grobler (MFMR) for updates of rock lobster catch data; Beau Tjizoo (NatMIRC) for updates of sardine and anchovy catch data; Stephanus Voges (NatMIRC) for updates of snoek catch data; and Erling Kåre Stenevik (IMR) for anchovy and sardine egg data from the 2004 Nansen cruise. We are extremely grateful to Cathy Boucher (MCM) for redrawing the figures. Comments on an earlier version of the manuscript by Dr K. Cochrane (FAO) and two anonymous referees are warmly acknowledged. This is a contribution of the IDYLE/ECO-UP and Upwelling Ecosystem programmes of IRD and EUR-OCEANS (a European Network of Excellence funded by the European Commission under the 6th Framework Programme - contract ref. 511106), respectively.
178
C.D. van der Lingen et al.
REFERENCES Armstrong, M.J., P.A. Shelton, R.M. Prosch and W.S. Grant. 1983. Stock assessment and population dynamics of anchovy and pilchard in ICSEAF Division 1.6 in 1982. Colln. Sci. Pap. Int. Comm. S.E. Atl. Fish. 10:7-25. Bakun, A. 1990. Global climate change and intensification of coastal upwelling. Science 247:198-201. Bakun, A. 1992. Global greenhouse effects, multidecadal wind trends, and potential impacts on coastal pelagic fish populations. ICES Mar. Sci. Symp. 195:316-325. Bakun, A. 1996. Patterns in the ocean. Ocean processes and marine population dynamics. California Sea Grant College System, National Oceanic and Atmospheric Administration, U.S.A., in cooperation with Centro de Investigaciones Biologicas del Noroeste, La Paz, BCS, Mexico. 323pp. Bakun, A. 2001 ‘School-mix feedback’: a different way to think about low frequency variability in large mobile fish populations. Prog. Oceanog. 49:485-511. Bakun, A. and P. Cury. 1999. The “school-trap”: a mechanism promoting large-amplitude out-of-phase population oscillations of small pelagic fish species. Ecology Letters 2(6): 349-351. Bakun, A. and S. Weeks. 2004. Greenhouse gas buildup, sardines, submarine eruptions and the possibility of abrupt degradation of intense marine upwelling systems. Ecology Letters 7:1015-1023. Barange, M., I. Hampton and B.A. Roel. 1999. Trends in the abundance and distribution of anchovy and sardine on the South African continental shelf in the 1990s, deduced from acoustic surveys. S. Afr. J. mar. Sci. 19: 367-391. Barange, M., S.C. Pillar and I. Hampton. 1998. Distribution patterns, stock size and life-history strategies of Cape horse mackerel Trachurus trachurus capensis, based on bottom trawl and acoustic surveys. S. Afr. J. mar. Sci. 19:433-447. Barkai, A. and C. McQuaid. 1988. Predator-prey reversal in a marine benthic ecosystem. Science 242:6264. Baumgartner, T., A. Soutar and W. Riedel. 1996. Natural time scales of variability in coastal fish populations of the California Current over the past 1500 years: Responses to global climate change and biological interaction. In Calif. Sea Grant Rep. 1992-1995, Calif. Sea Grant College, La Jolla, 31-37. Beamish, R.J. and C. Mahnken. 1999. Taking the next step in fisheries management. In: Ecosystem Approaches for Fisheries Management. University of Alaska Sea Grant, AK-SG-99-01, Fairbanks: 121. Beaugrand, G., K.M. Brander, J. Alistair Lindley, S. Soussi and P. C. Reid. 2003. Plankton effect on cod recruitment in the North Sea. Nature 426:661-664. Binet, D., B. Gobert and L. Maloueki. 2001. El Niño-like warm events in the Eastern Atlantic (6°N, 20°S) and fish availability from Congo to Angola (1964-1999). Aquat. Living Resour. 14:99-112. Boyd, P.W., A.J. Watson, C.S. Law, E.R. Abraham, T. Trull, R. Murdoch, D.C.E. Bakker, A.R. Bowie, K.O. Buessler, H. Chang, M. Charette, P. Croot, K. Downing, R. Frew, M. Gall, M. Hadfield, J. Hall, M. Harvey, G. Jameson, J. Laroche, M. Liddicoat, R. Ling, M.T. Maldonado, R.M. MCKay, S. Nodder, S. Pickmere, R. Pridmore, S. Rintoul, K. Safi, P. Sutton, R. Strzepek, K. Tanneberger, S. Turner, A. Waite and J. Zeldis. 2000. A mesoscale phytoplankton bloom in the polar Southern Ocean stimulated by iron fertilization. Nature 407:695-702. Boyer, D.C., H.J. Boyer, I. Fossen and A. Kreiner. 2001. Changes in abundance of the northern Benguela sardine stock during the decade 1990 to 2000 with comments on the relative importance of fishing and the environment. S. Afr. J. mar. Sci. 23:67-84. Boyer, D.C. and I. Hampton. 2001. An overview of the living marine resources of Namibia. S. Afr. J. mar. Sci. 23:5-35. Bloomer, S.F., K.L. Cochrane and J.G. Field. 1994. Towards predicting recruitment success of anchovy Engraulis capensis Gilchrist in the southern Benguela system using environmental variables: A rulebased model. S. Afr. J. mar. Sci. 14:107-119. Burmeister, L.-M. 2001. Depth-stratified density estimates and distribution of the Cape hake Merluccius capensis and M. paradoxus off Namibia deduced from survey data, 1990-1999. S. Afr. J. mar. Sci. 23:347-356. Burmeister, L.-M., D. Mouton and A. van der Plas. 2002. Cruise Report of the FV Blue Sea: Surveys of the hake stock, 16 January - 25 February 2002. Ministry for Fisheries and Marine Resources, Swakopmund, Namibia: 88pp.
Resource and ecosystem variability, including regime shifts, in the BCLME
179
Carpenter, S.R., J.F. Kitchell, D. Bade, T.E. Essington, J.N. Houser, J.J. Cole, M.L. Pace, J.R. Hodgson, K.L. Cottingham and D.E. Schindler. 2001. Trophic cascades, nutrients, and lake productivity: Wholelake experiments. Ecol. Monogr. 71:163-186. Chapman, P. and L.V. Shannon. 1985. The Benguela ecosystem Part II. Chemistry and related processes. Oceanogr. Mar. Biol. Ann. Rev. 23:183-251. Cochrane, K.L., C.J. Augustyn, A.C. Cockcroft, J.H.M. David, M.H. Griffiths, J.C. Groeneveld, M.R. Lipinski, M.J. Smale, C.D. Smith and R.J.Q. Tarr. 2004. An ecosystem approach to fisheries in the southern Benguela context. Afr. J. mar. Sci. 26:9-35. Cockcroft, A.C. 2001. Jasus lalandii “walkouts” or mass strandings in South Africa during the 1990s: an overview. Mar. Freshwat. Res. 52:1085-1094. Cockcroft, A.C. and P.C. Goosen. 1995. Shrinkage at moulting in the rock lobster Jasus lalandii and associated changes in reproductive parameters. S. Afr. J. mar. Sci. 16:195-203. Cockcroft, A.C. and A.I.L. Payne. 1999. A cautious fisheries management policy in South Africa: the fisheries for rock lobster. Mar. Policy 23:587-600. Coetzee, J., D. Merkle, J. Rademan, C. van der Lingen, L. Hutchings and J. Huggett. 2004. Results of the 2004 spawner biomass survey. MCM document WG/DEC04/PEL/04. Collie, J.S., K. Richardson and J.H. Steele. 2004. Regime shifts: Can ecological theory illuminate the mechanisms? Prog. Oceanogr. 60:281-302. Crawford, R.J.M. 1981. Distribution, availability and movement of pilchard Sardinops ocellata off South Africa, 1964-1976. Fish. Bull. S. Afr. 14:1-46. Crawford, R.J.M. 1998. Responses of African penguins to regime changes of sardine and anchovy in the Benguela system. S. Afr. J. mar. Sci. 19:355-364. Crawford, R.J.M. 2003. Influence of food on numbers breeding, colony size and fidelity to localities of Swift Terns in South Africa’s Western Cape, 1987-2000. Waterbirds 26(1):44-53. Crawford, R.J.M. 2005. Cape Gannet. In Roberts’ Birds of Southern Africa. Eds. P.A.R. Hockey, W.R.J. Dean, P.G. Ryan and S. Maree. John Voelcker Bird Book Fund, Cape Town. Crawford, R.J.M., J.H.M. David, L.J. Shannon, J. Kemper, N.T.W. Klages, J.-P. Roux, L.G. Underhill, V.L. Ward, A.J. Williams and A.C. Wolfaardt. 2001. African penguins as predators and prey – coping (or not) with change. S. Afr. J. mar. Sci. 23:435-447. Crawford, R.J.M and B.M. Dyer. 1995. Responses by four seabird species to a fluctuating availability of Cape anchovy Engraulis capensis off South Africa. Ibis 137:329-339. Crawford, R.J.M., L.J. Shannon, A. Kreiner, C.D. van der Lingen, J. Alheit, A. Bakun, D. Boyer, P. Cury, T. Dunne, M.-H Durand, J.G. Field, P. Fréon, M.H. Griffiths, E. Hagen, L. Hutchings, E. Klingelhoeffer, C.L. Moloney, D. Mouton, C. Roy, J.-P Roux, L.V. Shannon, F.A. Shillington, L.G. Underhill and H.M. Verheye Unpublished manuscript—Periods of major change in the structure and functioning of the pelagic component of the Benguela ecosystem, 1950-2000. Crawford, R.J.M., L.V. Shannon and D.E. Pollock. 1987. The Benguela ecosystem Part IV. The major fish and invertebrate resources. Oceanogr. Mar. Biol. Ann. Rev. 25:353-505. Crawford, R.J.M. and P.A. Whittington. 2005. – African Penguin. In Roberts’ Birds of Southern Africa. Eds. P.A.R. Hockey, W.R.J. Dean P.G. Ryan and S. Maree. John Voelcker Bird Book Fund, Cape Town. Cunningham, C.L., and D.S. Butterworth. 2004a. Base Case Bayesian Assessment of the South African Anchovy Resource. MCM document WG/APR04/PEL/01. Cunningham, C.L., and D.S. Butterworth. 2004b. Base Case Bayesian Assessment of the South African Sardine Resource. MCM document WG/APR04/PEL/02. Cury, P.M. 2004. Tuning the ecoscope for the Ecosystem Approach to Fisheries. In: Perspectives on ecosystem-based approaches to the management of marine resources. Mar. Ecol. Prog. Ser. 274:272275. Cury, P., A. Bakun, R.J.M. Crawford, A. Jarre-Teichmann, 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., Symposium Edition 57(3): 603-618. Cury, P.M. and V. Christensen. 2005. International Symposium of Quantitative Ecosystem Indicators for Fisheries Management. ICES J. mar. Sci. 62:307-310 Cury, P. and L.J. Shannon. 2004. Regime shifts in upwelling ecosystems : observed changes and possible mechanisms in the northern and southern Benguela. Prog. Oceanog. 60:223-243.
180
C.D. van der Lingen et al.
Cury, P., L.J. Shannon, J.-P. Roux, G. Daskalov, C.L. Moloney, A. Jarre, and D. Pauly. 2005a, Trophodynamic indicators for an ecosystem approach to fisheries. ICES J. mar. Sci. 62:430-442. Cury, P., P. Fréon, C.L. Moloney, L.J. Shannon, and Y.-J. Shin. 2005b. Processes and patterns of interactions in marine fish populations: An ecosystem perspective. Chapter 13, 475-554 in Robinson, A.R. and K. H. Brink, eds. The Sea. The global coastal ocean: multiscale interdisciplinary processes. Harvard University Press, Cambridge. Cury, P.M., L.J. Shannon and Y.-J. Shin.2003. The functioning of marine ecosystems: a fisheries perspective. In Responsible Fisheries in the Marine Ecosystem. Eds. M. Sinclair and G. Valdimarsson. FAO and CABI Publishing, UK, pp. 103-123. Cushing, D.H. 1990. Plankton production and year-class strength in fish populations: An update of the match/mismatch hypothesis. Adv. Mar. Biol. 26:249-293. Daskalov, G.M. 2002. Overfishing drives a trophic cascade in the Black Sea. Mar. Ecol. Prog. Ser. 225:53-63. Daskalov, G.M., D.C. Boyer and J.P. Roux. 2003. Relating sardine Sardinops sagax abundance to environmental variables in the northern Benguela. Prog. Oceanog. 59:257-274. David, J.H.M. 1987. Diet of the South African fur seal (1974-1985) and an assessment of competition with fisheries in southern Africa. S. Afr. J. mar. Sci. 5:693-713. David, J.H.M., P. Cury, R.J.M. Crawford, R.M. Randall, L.G. Underhill and M.A. Meyer. 2003. Assessing conservation priorities in the Benguela ecosystem: analysing predation by seals on threatened seabirds. Biol. Conserv. 114:289-292. De Oliveira, J.A.A. 2002. The development and implementation of a joint management procedure for the South African pilchard and anchovy resources. PhD thesis, University of Cape Town, iv + 319 pp. de Villiers. 1985. Simpsoa internacional sobre las areas de afloramiento mas importantes del oeste Africano (Cabo Blanco y Benguela). Int. Symp. Upw. W. Afr., Inst. Inv. Pesq., Barcelona 1985, 2: 1005-1039. Dingle, R.V. and G. Nelson. 1993. Sea-bottom temperature, salinity and dissolved oxygen on the continental margin off south-western Africa. S. Afr. J. mar. Sci. 13:33-49. Drapeau, L., L. Pecquerie, P. Fréon and L.J. Shannon. 2004. Quantification and representation of potential spatial interactions in the southern Benguela ecosystem. Afr. J. mar. Sci. 26:141-159. Fairweather, T.P., C.D. van der Lingen, L. Drapeau, A.J. Booth and J.J. van der Westhuizen, Submitted, Indicators of sustainable fishing for the South African sardine (Sardinops sagax) and management implications. Can. J. Fish. Aquat. Sci. FAO. 2000. Report of the workshop on small pelagic resources of Angola, Congo and Gabon. Luanda, Angola, 3-7 Nov. 1997. FAO Fish. Rep. 618, FIRM/SAFR/R618. FAO, Rome, 149pp. Field, J.G. and F.A. Shillington. 2006. Variability of the Benguela Current System. 833-861 in Robinson, A.R. and K.H. Brink, eds. The Sea, Vol. 14, Coasts. Harvard University Press, Cambridge. Fossen, I., Boyer, D.C. and H. Plarre. 2001. Changes in some key biological parameters of the northern Benguela sardine stock. S. Afr. J. mar. Sci. 23:111-121. Fréon, P., L. Drapeau, J. David, A. Fernandez Moreno, R. Leslie, H. Oosthuizen, L.J. Shannon and C.D. van der Lingen. 2005. Spatialised ecosystem indicators in the southern Benguela. ICES J. mar. Sci. 62:459-468. Gammelsrød, T., C.H. Bartholomae, D.C. Boyer, V.L.L. Filipe and M.J. O’Toole. 1998. Intrusion of warm surface water along the Angolan-Namibian coast in February-March 1995: the 1995 Benguela Niño. S. Afr. J. mar. Sci. 19:41-56. Geromont, H.F., D.S. Butterworth, L.-M. Burmeister, F. Dealy, A. van der Westerhuizen and L. Voges. 2000. Preliminary set of operating models proposed for the revised management procedure trials for the Namibian hake resource. BEN/NOV00/NH/3b, 35 pp. Griffiths, C.L., L. van Sittert, P.B. Best, A.C. Brown, P.A. Cook, R.J.M. Crawford, J.H.M. David, B.R. Davies, M.H. Griffiths, K. Hutchings, A. Jerardino, N. Kruger, S. Lamberth, R. Leslie, R. MelvilleSmith, R. Tarr and C.D. van der Lingen. 2004. Impacts of human activities on marine animal life in the Benguela – a historical overview. Oceanog. Mar. Biol. Ann. Rev. 42:303-392. Hamukuaya, H., M.J. O’Toole, and P.M.J. Woodhead. 1998. Observations of severe hypoxia and offshore displacement of Cape hake over the Namibian shelf in 1994. S. Afr. J. mar. Sci. 19:57-59. Hampton, I.. 2003. Harvesting the Sea. In Namibia’s Marine Environment. Eds. F. Molloy and T. Reinikainen. Directorate of Environmental Affairs, Ministry of Environment and Tourism, Namibia; 3169. Herendeen, R.A.. 2004. Dynamic trophic cascade. Ecol. Mod. 177:129-242. Heymans, J.J., L.J. Shannon and A. Jarre. 2004. Changes in the northern Benguela ecosystem over three decades: 1970s, 1980s and 1990s. Ecol. Mod., 172:175-195.
Resource and ecosystem variability, including regime shifts, in the BCLME
181
Hutchings, J.A.. 2000. Collapse and recovery of marine fishes. Nature 406:882- 885. Hutchings, L., M. Barange, S.F. Bloomer, A.J. Boyd, R.J.M. Crawford, J.A. Huggett, M. Kerstan, J.L. Korrubel, J.A.A. De Oliveira, S.J. Painting, A.J. Richardson, L.J. Shannon, F.H. Schulein, C.D. van der Lingen and H.M. Verheye. 1998. Multiple factors affecting South African anchovy recruitment in the spawning, transport and nursery areas. S. Afr. J. mar. Sci. 19: 211-255. Jarre, A., C.L. Moloney, L.J. Shannon, P. Fréon, C.D. van der Lingen, H.M. Verheye, L. Hutchings, J.-P. Rous and P. Cury. This volume, Chapter 11. Developing a basis for detecting and forecasting longterm ecosystem changes. Kemper, J., J.-P. Roux, P.A. Bartlett, Y.J. Chesselet, J.A.C. James, R. Jones, S. Wepner and F.J. Molloy. 2001. Recent population trends of African penguins Spheniscus demersus in Namibia. S. Afr. J. mar. Sci. 23:429-434 King, D.P.F. 1977. Distribution and relative abundance of eggs of the South West African pilchard Sardinops ocellata and anchovy Engraulis capensis, 1971/72. Fish. Bull. S. Afr. 9:23-31. Korrubel, J.L., S.F. Bloomer, K.L. Cochrane, L. Hutchings and J.G. Field. 1998. Forecasting in South African pelagic fisheries management: the use of expert and decision support systems. S. Afr. J. mar. Sci. 19:415-424. Kreiner, A., C.D. van der Lingen and P. Fréon. 2001. A comparison of condition factor and gonadosomatic index of sardine (Sardinops sagax) stocks in the northern and southern Benguela upwelling ecosystems, 1984-1999. S. Afr. J. mar. Sci. 23:123-134. Kristmannsson, S.S. 1994. Dissolved oxygen conditions on the shelf off Namibia in 1994. Rit Fiskideil. 16: 89-95. Larkin, P.A. 1996. Concepts and issues in marine ecosystem management. Rev. Fish Biol. Fish. 6:139164. Le Clus, F. 1987. Reproductive dynamics of female pilchard Sardinops ocellatus in the northern Benguela system, with particular reference to seasonality, the environment and fish condition. Ph.D. thesis, University of Port Elizabeth: 430pp. Le Clus, F. 1990. Spatial variability in the seasonal trends of ovary weight in the Namibian pilchard population, 1965-1979. S. Afr. J. mar. Sci. 9: 69-83. Lehodey, P., J. Alheit, M. Barange, T. Baumgartner, G. Beaugrand, K. Drinkwater, J.-M. Fromentin, S.R. Hare, G. Ottersen, R.I. Perry, C. Roy, C.D. van der Lingen and F. Werner. In press. Climate variability, fish and fisheries. J. Climate MacCall, A.D. 1990. Dynamic Geography of Marine Fish Populations. Seatle; University of Washington Press: 153 pp. Matthews, J.P. 1960. The pilchard of South West Africa (Sardinops ocellata): Size composition of the commercial catches in South West Africa, 1952-1957. Invest. Rep. No. 1, SWA Adm., Windhoek, 45pp. Mayfield, S. and G.M. Branch. 2000. Interrelations among rock lobsters, sea urchins, and juvenile abalone: implications for community management. Can. J. Fish. Aquat. Sci. 57:2175-2185. Mecenero, S. and J-P. Roux. 2002. Spatial and temporal variability in the diet of a top predator in the northern Benguela. GLOBEC Report 16: 62-64. Mecenero, S., J-P. Roux, L.G. Underhill and M.N. Bester. 2006. Diet of Cape fur seals Arctocephalus pusillus pusillus at three mainland breeding colonies in Namibia. Afr. J. mar. Sci.28(1).57-71. Melville-Smith, R., P.C. Goosen and T.J. Stewart, 1995, The spiny lobster Jasus lalandii (H. Milne Edwards, 1837) off the South African coast: inter-annual variations in male growth and female fecundity. Crustaceana 68(2):174-183. Miller, D.C.M. and J.G. Field. 2002. Predicting anchovy recruitment in the southern Benguela ecosystem: developing an expert system using classification trees. S. Afr. J. Sci. 98:465-472. Molloy, F. and T. Reinikainen. 2003. Introduction. 1-6 in Molloy, F. and T. Reinikainen., eds. Namibia’s Marine Environment. Directorate of Environmental Affairs, Ministry of Environment and Tourism, Namibia. Moloney, C.L., A. Jarre, H. Arancibia, Y.-M. Bozec, S. Neira, J.-P. Roux and L.J. Shannon. 2005. Comparing the Benguela and Humboldt marine upwelling ecosystems with indicators derived from inter-calibrated models ICES J. mar. Sci. 62: 493-502.
182
C.D. van der Lingen et al.
Mullon, C.P., P. Fréon, C. Parada, C.D. van der Lingen and J. Huggett. 2003. From particles to individuals: modelling the early stages of anchovy (Engraulis capensis/encrasicolus) in the southern Benguela. Fish. Oceanog. 12:396-406. NORAD-FAO/UNDP. 1992. Surveys of the fish resources of Namibia. Part 1: Surveys of the hake stocks 23 April - 22 May 1992. Ministry for Fisheries and Marine Resources, Swakopmund: 31pp. O'Donoghue, S. and D.J. Marshall. 2003. Marine pollution research in South Africa: a status report. S. Afr. J. Sci. 99:349-356 O'Toole, M. J., 1977, Investigations into some important fish larvae in the south east Atlantic in relation to the hydrological environment. Ph.D. Thesis. University of Cape Town, 299pp. Painting, S.J. and J.L. Korrubel. 1998. Forecasts of recruitment in South African anchovy from SARP field data using a simple deterministic expert system. S. Afr. J. mar. Sci. 19:245-262. Pauly, D., V. Christensen and C. Walters. 2000. Ecopath, Ecosim, and Ecospace as tools for evaluating ecosystem impact of fisheries. ICES J. Mar. Sci. 57:697-706. Payne, A.I.L., 1989, Cape hakes. In: Oceans of Life off Southern Africa. Eds. A.I.L. Payne and R.J.M. Crawford, Vlaeberg Publishers, South Africa, pp 136-147. Payne, A.I.L. and R.J.M. Crawford. 1989. The major fisheries and their management. In: Oceans of Life off Southern Africa. Eds. A.I.L. Payne and R.J.M. Crawford, Vlaeberg Publishers, South Africa, pp 5061. Payne, A.I.L. and J.R.E. Lutjeharms, eds. 1997. A Century of Marine Science in South Africa. Trans. Roy. Soc. S. Afr. 52(1). Payne, A.I.L., S.C. Pillar and R.J.M. Crawford, eds. 2001. A Decade of Namibian Fisheries Science. S. Afr. J. mar. Sci. 23 Penven, P., C. Roy, G.B. Brundrit, A. Colin de Verdière, P. Fréon, A.J. Johnson, J.R.E. Lutjeharms and F.A. Shillington. 2001. A regional hydrodynamic model of upwelling in the southern Benguela. S. Afr. J. Sci. 97:472-475. Pitcher, T.J. 2000. Rebuilding as a new goal for fisheries management: reconstructing the past to salvage the future. Ecol. Appl. 11:601-617. Pitcher, T.J. and D. Pauly. 1998. Rebuilding ecosystems, not sustainability, as the proper goal of fishery management. In Pitcher, T.J., Hart, P., and D. Pauly. 1998. Reinventing fisheries management. Kluwer Academic Publishers, London. Fish and Fisheries Series 23: 311-329. Pollock, D.E. 1982. The fishery for and population dynamics of West Coast rock lobster related to the environment in the Lambert’s Bay and Port Nolloth areas. Investl Rep. Sea Fish. Inst. S. Afr. 124: 57pp. Pollock, D.E., A.C. Cockcroft and P.C. Goosen. 1997. A note on reduced rock lobster growth rates and related environmental anomalies in the southern Benguela, 1988-1995. S. Afr. J. mar. Sci. 18:287-293. Pollock, D.E. and L.V. Shannon. 1987. Response of rock-lobster populations in the Benguela ecosystem to environmental change – a hypothesis. S. Afr. J. mar. Sci. 5: 887-899. Punt, A.E. and D.S. Butterworth. 1995. The effects of future consumption by the Cape fur seal on catches and catch rates of the Cape hakes. 4. Modelling the biological interaction between Cape fur seals Arctocephalus pusillus pusillus and the Cape hakes Merluccius capensis and M. paradoxus. S. Afr. J. mar. Sci. 16:255-285. Rademeyer, R.A. and D.S. Butterworth. 2003. Species disaggregated assessments of the South African hake resource. Unpublished Report, BENEFIT Workshop document, BEN/JAN04/SAH/3a. Richardson, A.J. and D.S. Schoeman. 2004, Climate impact on plankton ecosystems in the northeast Atlantic. Science 305:1609-1612. Rochet, M.-J. 1998. Short-term effects of fishing on life history traits of fishes. ICES J. mar. Sci. 55:371391. Rochet, M.-J., 2000, May life history traits be used as indices of population viability? J. Sea Res. 44:145157. Rothschild, B. and L.J. Shannon. 2004. Regime shifts and fishery management. Prog. Oceanog. 60:397402. Roux, J-P. 1998. The impact of environmental variability on the seal population. Namibia Brief; Spec. Issue 18: Focus on Fisheries and Research: 138-140. Roux, J-P. 2003. Risks. In Namibia’s Marine Environment. Eds. F.J. Molloy and T. Reinikainen. Directorate of Environmental Affairs, MET, Windhoek, Namibia, 137-152. Roux, J.-P. and L.J. Shannon. 2004. Ecosystem approach to fisheries management in the northern Benguela: the Namibian experience. Afr. J. mar. Sci. 26:79-93.
Resource and ecosystem variability, including regime shifts, in the BCLME
183
Roy, C., P. Fréon and C.D. van der Lingen. 2002. An empirical model of anchovy recruitment variability in the southern Benguela. GLOBEC Report 16: 52-54. Roy, C., S.C. Weeks, M. Rouault, G. Nelson, R. Barlow and C.D. van der Lingen. 2001. Extreme oceanographic events recorded in the Southern Benguela during the 1999-2000 summer season. S. Afr. J. Sci. 97:465-471. Schwartzlose, R.A., J. Alheit, A. Bakun, T.R. Baumgartner, R. Cloete, R.J.M. Crawford, W.J. Fletcher, Y. Green-Ruiz, E. Hagen, T. Kawasaki, D., Lluch-Belda, S.E. Lluch-Cota, A.D. MacCall, Y. Matsuura, M.O. Nevarez-Maryinez, H. Parrish, C. Roy, R. Serra, K.V. Shust, M.N. Ward and J.Z. Zuzunaga. 1999. Worldwide large-scale fluctuations of sardine and anchovy populations. S. Afr. J. Mar. Sci. 21:289-347. Shannon, L.J. 1998. Modelling environmental effects on the early life history of the South African anchovy and sardine: a comparative approach. S. Afr. J. mar. Sci. 19:291-304. Shannon, L.J., J.G. Field and C.L. Moloney. 2004c. Simulating anchovy-sardine regime shifts in the southern Benguela ecosystem. Ecol. Mod. 172:269-281. Shannon, L.J., V. Christensen and C. Walters. 2004b. Modelling stock dynamics in the southern Benguela ecosystem over the last 25 years. Afr. J. mar. Sci. 26:179-196. Shannon, L.J., K.L. Cochrane, C.L. Moloney and P. Fréon, eds. 2004a. Ecosystem Approaches to Fisheries in the Southern Benguela. Afr. J. mar. Sci. 26: 328pp. Shannon, L.J. and P. Cury. 2003. Indicators quantifying small pelagic fish interactions: application using a trophic model of the southern Benguela ecosystem. Ecological Indicators 3(4): 305-321. Shannon, L.J., C.L. Moloney, A. Jarre and J.G. Field. 2003. Trophic flows in the southern Benguela during the 1980s and 1990s. J. Mar. Sys. 39(1-2): 83-116. Shannon, L.J., G. Nelson, R.J.M. Crawford and A.J. Boyd. 1996. Possible impacts of environmental change on pelagic fish recruitment: modelling anchovy transport by advective processes in the southern Benguela. Global Change Biology 2: 407-420. Shannon, L.V., J.J. Agenbag and M.E.L. Buys. 1987. Large and mesoscale features of the AngolaBenguela front. S. Afr. J. mar. Sci. 5:11-34. Shannon, L.V. and S.C. Pillar. 1986. The Benguela ecosystem Part III. Plankton. Oceanogr. Mar. Biol. Ann. Rev. 24: 65-170. Shannon, L.V. 1985. The Benguela ecosystem Part I. Evolution of the Benguela, physical features and processes. Oceanogr. Mar. Biol. Ann. Rev. 23:105-182. Sholto-Douglas, A.D., J.G. Field, A.G. James and N.J. van der Merwe. 1991. 13C/12C and 15N/14N isotope ratios in the southern Benguela Ecosystem: indicators of food web relationships among different size classes of plankton and pelagic fish; differences between fish muscle and bone collagen tissues. Mar. Ecol. Prog. Ser. 78:23-31. Shin, Y.-J. and P. Cury. 2001. Exploring fish community dynamics through size-dependent trophic interactions using a spatialized individual-based model. Aquat. Liv. Res. 14(2): 65-80. Shin, Y.-J., L.J. Shannon and P.M. Cury. 2004. Simulations of fishing effects on the southern Benguela fish community using an individual-based model: learning from a comparison with ECOSIM. Afr. J. mar. Sci. 26:94-114. Siegfried, W.R., R.J.M. Crawford, L.V. Shannon, D.E. Pollock, A.I.L. Payne and R.G. Krohn. 1990. Scenarios for global warming induce change in the open-ocean environment and selected fisheries of the west coast of southern Africa. S. Afr. J. Sci. 86:281-285. Steele, J.H. 2004. Regime shifts in the ocean: Reconciling observations and theory. Prog. Oceanog. 60:135-141. Stenseth, N.C., G. Ottersen, J.W. Hurrell, A. Mysterud, M. Lima, K.-S. Chan, N.G. Yoccoz and B. Ådlandsvik. 2003. Studying climate effects on ecology through the use of climate indices; the North Atlantic Oscillation, El Niño Southern Oscillation and beyond. Proc. Royal Soc. Lond. B, DOI 10.1098/rspb.2003.2415. Tarr, R.J.Q.. 2000. The South African abalone (Haliotis midae) fishery: a decade of challenges and change. Can. Spec. Publ. Fish. Aquat. Sci. 130:32-40. Tarr, R.J.Q., P.V.G. Williams and A.J. MacKenzie. 1996. Abalone, sea urchins and rock lobster: a possible ecological shift may affect traditional fisheries. S. Afr. J. mar. Sci. 17:311-315. Thomas, R.M. 1985. Growth rate of the pilchard off South West Africa, 1971-1983. Investl Rep. Sea Fish. Res. Inst. S. Afr. 128: 41pp. Thomas, R.M. 1986. The Namibian pilchard: the 1985 season, assessment for 1952-1985 and recommendations for 1986. Colln. scient. Pap. Int. Commn. SE. Atl. Fish. 13(2): 243-269.
184
C.D. van der Lingen et al.
Thompson, D. and S. Mostert. 1974. Muscle esterase genotypes in the pilchard, Sardinops ocellata. J. Cons. perm.i nt. Explor. Mer 36:50-53. van der Lingen, C.D., J.C. Coetzee and L. Hutchings. 2002. Temporal shifts in the spatial distribution of anchovy spawners and their eggs in the Southern Benguela: Implications for recruitment. GLOBEC Report 16:46-48. van der Lingen, C.D., T.P. Fairweather, P. Fréon and J.J. van der Westhuizen. In press. Densitydependent changes in reproductive parameters and condition of southern Benguela sardine Sardinops sagax. Afr. J. mar. Sci. van der Lingen, C.D., P. Freon, L. Hutchings, C. Roy, G.W. Bailey, C. Bartholomae, A.C. Cockcroft, J.G. Field, K.R. Peard and A. van der Plas. This volume, Chapter 14. Forecasting shelf processes of relevance to living marine resources in the BCLME. van der Lingen, C.D. and L. Hutchings. 2005. Estimating the lipid content of pelagic fish in the southern Benguela by visual assessment of their mesenteric fat. Afr. J. mar. Sci. 27:45-53. van der Lingen, C.D., L. Hutchings, D. Merkle, J.J. van der Westhuizen and J. Nelson. 2001. Comparative spawning habitats of anchovy (Engraulis capensis) and sardine (Sardinops sagax) in the southern Benguela upwelling ecosystem. In Spatial Processes and Management of Marine Populations. Eds. G.H. Kruse, N. Bez, T. Booth, M. Dorn, S. Hills, R.N. Lipcius, D. Pelletier, C. Roy, S.J. Smith and D. Witherell, University of Alaska Sea Grant, AK-SG-01-02, Fairbanks: 185-209. van der Westhuizen, A. 2001. A decade of exploitation and management of the Namibian hake stocks. S. Afr. J. mar. Sci. 23:307-315. Vasconcellos, M., Mackinson, S., Sloman, K. and D. Pauly. 1997. The stability of trophic mass-balance models of marine ecosystems: a comparative analysis. Ecol. Mod. 100:125-134. Vezina, A.F., F. Berreville and S. Loza. 2004. Inverse reconstruction of ecosystem flows in investigating regime shifts: impact of the choice of objective function. Prog. Oceanog. 60:321-341. Walters, C., V. Christensen and D. Pauly., 1997. Structuring dynamic models of exploited ecosystems from trophic mass-balance assessments. Rev. Fish Biol. Fish. 7:139-172. Walters, C. and J.F. Kitchell. 2001. Cultivation/depensation effects on juvenile survival and recruitment: implications for the theory of fishing. Can. J. Fish. Aquat. Sci. 58:39-50. 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. 101:147-167. Worm, B. and R. A. Myers. 2003. Meta-analysis of cod-shrimp interactions reveals top-down control in oceanic food webs. Ecology 84:162-173. Wysokinsky, A. 1984. Length structure of the cape horse mackerel population and changes in the sexual maturity length in the Namibian region. Colln scient. Pap. int. Commn SE Atl. Fish. 11(II): 91-98.
Large Marine Ecosystems, Vol. 14 V. Shannon, G. Hempel, P. Malanotte-Rizzoli, C. Moloney and J. Woods (Editors) © 2006 Elsevier B.V. All rights reserved.
185
9 Modelling, Forecasting and Scenarios in Comparable Upwelling Ecosystems: California, Canary and Humboldt Pierre Fréon, Jürgen Alheit, Eric D. Barton, Souad Kifani, Patrick Marchesiello
ABSTRACT The three eastern boundary ecosystems comparable to the Benguela ecosystem (BCE) display differences and commonalities. The California (CalCE) and Humboldt Current (HCE) ecosystems are continuous topographically, whereas the Canary Current ecosystem (CanCE) is interrupted by the Gulf of Cadiz and the Canaries archipelago. All have similar regimes of equatorward flow over shelf and slope associated with upwelling and a subsurface poleward flow over the slope, though in the HCE multiple flows and counter-flows appear offshore. All systems exhibit year round upwelling in their centre and seasonal upwelling at their extremes as the trade wind systems that drive them migrate north and south, though the HCE is strongly skewed toward the equator. All systems vary on scales from the event or synoptic scale of a few days, through seasonal, to inter-decadal and long term. Productivity of each system follows the upwelling cycle, though intra-regional variations in nutrient content and forcing cause significant variability within regions. The CanCE is relatively unproductive compared to the CalCE and HCE as a result of differences in large scale circulation between the Pacific and Atlantic. The latter two systems are dominated by El Niño Southern Oscillation (ENSO) variability on a scale of 4-7 years. Physical modeling with the Princeton Ocean Model and the Regional Oceanic Modeling System has advanced recently to the stage of reproducing realistic mesoscale features and energy levels with climatic wind forcing. Operational forecasting by these models with assimilation of sea surface temperature and other data is successfully implemented in CalCE. On longer time scales, the Lamont-Doherty Earth Observatory model is able to hindcast El Niño variability over the long term up to 2 years in advance. Empirical ecological models in all three systems have attempted prediction of permissible catch level (fractional Maximum Sustainable Yield), recruitment, catches or onset of migration with lack of continued success, partly because discontinuous or inadequate observations hamper model implementation and assessment. Moreover, empirical models tuned to particular environments fail when fundamental regime shifts occur. One of the most successful approaches is that of intensive monitoring of catch and environmental parameters linked to an informal Operational Management Procedure
186
P. Fréon et al.
(OMP) to inform fisheries management off Peru. This OMP contributed to preservation of anchovy stock during the 1997-8 El Niño but remains to be formalized or tested under varying conditions. Prediction on time scales of global warming are uncertain because physical climate models still disagree on whether upwelling will intensify or weaken. Possible scenarios on decadal scale based on warming or cooling of waters in the Eastern Boundary Current systems can be proffered, albeit with little confidence at present. Future approaches for all systems, including the BCE, will in the long run likely combine coupled atmospheric/ocean models with biological process models. Judicious application of purely statistical modeling based on inherent time series properties will assist, though such techniques are unable to cope with regime shifts. INTRODUCTION The objective of this chapter is to review the variability and change in systems that can be compared to the Benguela system and to show what lessons can be learned from the modelling and forecasting activities in those systems. Eastern ocean boundary ecosystems can be classified into three zones (Mackas et al., in press): 1) mid & low latitude upwelling; 2) equatorial; and 3) high latitude, poleward surface flow and downwelling. The Benguela Current ecosystem (BCE) falls in the first group, which is characterized by local wind-driven upwelling, strong alongshore advection, a poleward undercurrent, high productivity of plankton and pelagic fish, seaward extension of the boundary current and biological system beyond the continental shelf, remote physical forcing by large scale teleconnections, very low to moderate precipitation and coastal freshwater inputs at least in the core area of the system, and highly dynamic systems displaying strong variability at all spatial and temporal scales. Here we focus on the other three ecosystems that belong to this first category (Figure 9-1): the California Current (CalCE), the Canary Current (CanCE) and the Humboldt Current (HCE) ecosystems. After briefly describing these three systems (part 1), we review efforts to model and forecast them (part 2), and finally we speculate on possible biological scenarios that represent the response of key populations to forecasted changes in the physical environment associated with global change (part 3). In part 1 we describe each system separately using three sub-sections: physical traits, productivity and fish and fisheries. In order to facilitate comparison between the three systems, in part 2 we present first a review of physical modelling and forecasting activities, then a review of ecology. In part 3, global scenarios are presented briefly, and their likely effect on each system is discussed. Although there is no consistent definition of regime shift, this terminology is frequently used in the present work as it is in the current literature. Here we followed de Young et al. (2004) in considering that a regime shift is an abrupt change from a quantifiable ecosystem state, representing substantial restructuring of the ecosystem persisting for long enough that a new quasi-equilibrium state can be observed.
Modelling forecasting and scenarios in upwelling LMEs
187
Figure 9-1. Schematic map of the eastern boundaries of the Pacific and Atlantic showing major current systems (CalCE = California Current Ecosystem; CanCE = Canary Current Ecosystem; HCE = Humboldt Current Ecosystem; BCE = Benguela Current Ecosystem) and their main features. (Longitude scale is compressed).
PART 1. BRIEF DESCRIPTIONS OF CALIFORNIA, CANARY AND HUMBOLDT CURRENT ECOSYSTEMS California Current Ecosystem (CalCE) Physical traits The collection of eastern boundary currents in the CalCE abutting the continental U.S. West Coast (often called the California Current System) has been extensively studied. Since 1950 the California Cooperative Ocean Fisheries Investigation (CalCOFI) has provided a large-scale time series of hydrographic measurements off Central and Southern California. A series of process experiments (Coastal Upwelling Experiment, CUE; Coastal Transition Zone, CTZ; Eastern Boundary Current, EBC; Coastal Ocean Dynamics Experiment, CODE; Coastal Ocean Processes, CoOP; Global Ocean Ecosystem Dynamics, GLOBEC) has been carried out with shipboard hydrographic
188
P. Fréon et al.
and Doppler current surveys, plus moored arrays and Lagrangian drifters to sample both large-scale and mesoscale currents. Satellite measurements of SST (sea-surface temperature) (AVHRR), sea-surface height (altimetry), and color (SeaWIFS) give extensive coverage, but are limited to the surface. Finally, theoretical and computational models have provided useful paradigms for understanding the dynamics of the CCS (Marchesiello et al. 2003). A recent review of observations, laboratory experiments, and model results on the CalCE can be found in Hickey (1998). Recent CalCE observations are reported in special issues of Deep Sea Research II (2000; 47, 761-1176) and Progress in Oceanography (2002; 53, Issue 2-4). Theoretical and modeling studies of the CalCE have traditionally focused on coastal upwelling and downwelling driven by strong alongshore winds over the narrow continental shelf. The system extends from about 47°N, where the West Wind Drift impinges on the coast to near 21°N at the tip of Baja California. Observations show that energetic near-shore currents respond to local wind forcing and remote influences transmitted along the coastal waveguide. During the upwelling season, strong southward jets develop along the upwelling front separating cool upwelled- from warm oceanic- waters, with significant topographically modulated variability. The persistent alongshore currents are unstable, and some separate from the near-shore region (Barth et al. 2000) to form offshore-flowing currents entraining cold upwelled water in the form of filaments (Brink and Cowles 1991). Upwelling is most intense along the central and northern California coast, and very seasonal to both north and south. At a depth around 300 m, a California Undercurrent flows northward along the slope all year (Collins et al. 2004). During winter a surface northward Davidson Current and downwelling develop nearshore along much of the coast (Strub and James 2002). The CalCE contains three characteristic water masses: Pacific Subarctic Water (low salinity S and temperature T; high oxygen and nutrients) is advected equatorward with the coastal current; North Pacific Central Water (high S, T, and nutrients; low oxygen) enters from the west with the West-Wind Drift; and Southern Water (high S and T; low oxygen and nutrients) comes from the south with the California Undercurrent. In general, S and T increase equatorward in the CalCE. Salinity also increases with depth in the CalCE, thereby enhancing stratification and baroclinicity; this constitutes a major and dynamically fundamental difference to all other eastern-boundary upwelling regimes. The intrinsic mesoscale variability is only weakly related to local wind stress fluctuations (Kelly et al. 1998). Instability of the coastal jets does not require topographic effects, but capes and ridges may promote locally enhanced upwelling centres and cross-shore transport (Narimousa and Maxworthy 1989; McCreary et al. 1991; Batteen 1997; Marchesiello et al. 2003). Eastern-boundary current systems are also affected by Rossby-wave dynamics, which transport energy westward to the open ocean (McCreary and Kundu 1985; Strub and James 2002). Productivity Satellite images often show sharp cross-shelf gradients in sea surface temperature (SST) and colour, barriers to material exchanges, which often develop into filamentary intrusions of cold, nutrient- and pigment-rich water forming a 300 km wide coastal transition zone. The shelf-flushing time of a few days, associated with cross-shelf
Modelling forecasting and scenarios in upwelling LMEs
189
transport by the cold filaments and associated mesoscale activity, thus is an important mechanism for shelf-ocean exchange of heat, nutrients, biota and pollutants (Mooers and Robinson 1984; Marchesiello et al. 2004). New production of phytoplankton in this system, largely caused by upwelling, forms the basis of a simple chain-like ecosystem along the coast characterized by large cell sizes (diatoms), mesozooplankton grazing, high biomass and nutrients. Upwelled nutrients are rapidly taken up by the growing coastal diatom populations, which can be advected over large distances offshore within mesoscale features. The formation of these near-shore phytoplankton blooms and their subsequent offshore advection in filaments is a striking feature of CalCE satellite ocean color images and has been used to demonstrate the tight coupling between biology and physics in the highly turbulent coastal region. A continental-shelf-resident zooplankton community is observed off the northern CalCE, dominated by the copepods Calanus marshallae, Pseudocalanus mimus, and Acartia longiremis, and the euphausiid Thysanoessa spinifera, whereas further to the south, a distinctive zooplankton assemblage has been observed (Mackas, in press; Jiménez-Pérez et al. 2000). In the coastal CalCE ecosystem, the mesozooplankton formed by the copepod community does not control the phytoplankton, because of the slow growing rates of the mesozooplankton. Along the coast, river run-off may affect the near-shore ecosystem, and the embayments provide an additional source of food for zooplankton and larval fish. At the same time, embayments are retention regions for the dominant large copepods and larval fish. Also, vertical migration is observed to be significant within the CalCE (Mackas et al. 1991; Batchelder et al. 2002), possibly providing an additional retention mechanism. Retention is crucial to the survival of zooplankton and larval fish in the CalCE, where particularly intense mesoscale activity leads to rapid offshore dispersal. Fish and fisheries Sinclair et al. (1985) analyzed El Niño impacts on larval success in decades of CalCOFI data and suggested that, despite reduced enrichment of coastal waters, El Niño provides a period of limited dispersal of fish eggs and larvae for certain species, which is favorable to later recruitment. Indeed, most of the biomass harvested from the CalCE is pelagic or semi-pelagic, mainly hake, market squid, anchovy, sardine, mackerel, jack mackerel, herring and, in the northern part of the system, salmon (Ware and McFarlane 1989; California Cooperative Oceanic Fisheries Investigations annual reports). These species occur throughout the continental shelf of the CalCE, although most display seasonal migration. Salmonids, demersal fishes, and some benthic invertebrates display narrower alongshore ranges. Several pelagic fish species spawn preferentially in the Southern California Bight (Parrish et al. 1981; Bakun and Parrish 1982). The Pacific sardine (Sardinops sagax) has displayed dramatic population changes: biomass estimates declined from 3 million tonnes in 1933 to <10,000 tonnes in 1975 then recovered to the present ~1 million tonnes (Smith and Moser 2003). In response to the collapse of the sardine fishery, the California Cooperative Oceanic Fisheries Investigations (CalCOFI) was initiated in 1947 and has since collected ichthyoplankton data (Figure 9-2) and many other oceanic data.
190
P. Fréon et al.
Figure 9-2. Examples of fish larval abundance 10m-2 obtained from CalCOFI data; surveys were triennial from 1966 to 1984 (Figure reprinted from Deep-Sea Research-II, 50, Smith and Moser, Long-term trends and variability in the larvae of Pacific sardine and associated fish species of the California Current region, pp 2519-2536, copyright 2003, with permission of Elsevier).
From 1988, use of the Continuous Underway Fish Egg Sampler (Checkley et al. 1997, Checkley et al. 2000) has confirmed that spawning of coastal pelagic species is broadly concentrated in and immediately north of the Southern California Bight. Spawning is frequently associated with the boundaries of eddies and upwelling plumes (Mackas et al., in press) and exhibits considerable interannual variation. Paleontological records of anoxic core sediment from the Santa Barbara basin indicate interdecadal variation in sardine and anchovy abundance prior to any substantial exploitation (Baumgartner et al. 1992). Catch rates, total biomass, and recruitment of most exploited species have varied 10100 fold at interannual and decadal time scales (Mantua et al. 1997; Bakun 2004; Chavez 2004; CalCOFI Reports series, http://www.calcofi.org/default.html), which is mainly attributed to variations in natural mortality. Similarly, several unexploited species of marine mammals and seabirds also show variability in population and
Modelling forecasting and scenarios in upwelling LMEs
191
reproductive success, alongshore migration, and seasonality and spatial zonation of their reproductive activity (e.g., Abraham and Sydeman 2004; Sydeman et al. 2001; Bertram et al. 2001; Veit et al. 1997). Canary Current Ecosystem (CanCE) Physical traits The CanCE, defined broadly by the eastern boundary of the North Atlantic subtropical gyre (Figure 9-1), extends from northern Iberia (43°N) to Guinea (10°N) (Barton 1998). The ecosystem is divided into the Iberian and the Northwest African areas, discontinuous across the Strait of Gibraltar. Reviews by Barton (1989), Mittelstaedt (1991), Barton (1998) and Arístegui et al., (in press) detail the physical oceanography and biogeochemistry of the region. Recent observations in the system are reported in volumes of Progress in Oceanography (2004, 62/2-4) and Deep-Sea Research (2002, 49). A weak Portugal Current flows southward offshore of Iberia. It and the Azores Current contribute to the Canary Current, which flows along the African coast to Cape Blanc (21°N), where the Canary Current separates to flow westward into the North Equatorial Current (Figure 9-1). Between Capes Blanc and Verde, a permanent cyclonic recirculation feeds water northwards along the coast. Part of this poleward flow continues beyond Cape Blanc as an undercurrent, possibly continuous as far north as Iberia. The north-south migration of the Azores High produces summertime upwelling off Iberia, year-round (more intense in summer) upwelling between 35° and 20°N, and winter upwelling south of Cape Blanc (Wooster et al. 1976). The upwelling is associated with formation of a strong temperature front and associated strong, equatorward jet-like flow near-shore. When wind forcing weakens or disappears, even briefly, this southward flow may be replaced by northward flow as the undercurrent extends to the sea surface, as occurs off Iberia in winter or Senegal in summer. The poleward flow may inject slope water into the open sea by shedding anticyclonic eddies (Peliz et al., 2003a, 2003b). Variability of upwelling occurs on decadal scales, in association with the North Atlantic Oscillation (NAO), but >70% of the variance is related to synoptic scale changes (<30 days) in wind forcing. Throughout the region, coastal upwelling filaments stretch offshore from capes and promontories to export waters rich in organic matter into the oligotrophic waters of the subtropical gyre (Arístegui et al. in press). Especially notable are the large filaments off Cape Ghir and off Cape Blanc, where the Canary Current separates from the coast. Pelegrí et al. (2005) suggest the Cape Ghir filament in autumn represents a separation of the main flow from the coast, similar to that at Cape Blanc, with a cyclonic recirculation extending south of the Canary Islands. This archipelago, unique in the eastern boundary current systems, introduces large mesoscale variability, mainly in the form of downstream vortices (Arístegui et al. 1994).
192
P. Fréon et al.
Productivity Varieties of North Atlantic Central Water (NACW) dominate the upper layers between Cape Finisterre and Cape Blanc. Eastern North Atlantic Central Water of subtropical (nutrient–poor) and subpolar (nutrient–rich) origin extends from the Canaries northward. High nutrient content South Atlantic Central Water (SACW) is separated from NACW by a front at 21°N (Cape Blanc). Upwelling related productivity is therefore a function of latitude as well as upwelling strength. Off Mauritania (~18°N), high surface nutrient concentrations are probably related to the doming in the cyclonic recirculation (Arístegui et al. in press). Generally, the region exhibits anomalously low productivity and chlorophyll concentrations relative to the observed nutrient concentrations (Minas et al. 1982). Compared to other upwelling regions, the source waters off NW Africa are poorer in nutrients but richer in oxygen as result of the global scale circulation (Codispoti et al. 1982), which may explain differences in productivity and regeneration rates found, for instance, between the NW Africa and the Peru ecosystems (Minas et al. 1982). Nutrient assimilation and regeneration ratios in the NW African coast are similar to other coastal upwelling regions, although significant regional differences within this system exist, particularly in silicon. Depending on geomorphological changes along the coast and the rate of regeneration, silicon can be either in excess or deficient (Herbland and Voituriez 1974; Minas et al. 1982). The interplay between Canary Islands eddies and upwelling filaments is likely to favour exportation of coastal enriched water to the oligotrophic open ocean. This would explain the strong imbalance between phytoplankton production and community respiration observed in the subtropical gyre (Arístegui et al., in press). Fish and fisheries The CanCE can be divided biogeographically into two areas that overlap at Cape Blanc, where the water masses transition between NACW and SACW. Sardine (Sardina pilchardus) is the predominant species off the Iberian and Moroccan coasts whereas sardinellas (Sardinella aurita and S. maderensis) and horse mackerels (Trachurus spp) are dominant further south off Mauritania and Senegal. At present, the CanCE overall produces ~3 million tonnes of marine catches. Small pelagic fish, mainly European sardine, represent ~44% of the catch, followed by medium-size pelagic fish, horse-mackerel and mackerel, (24%), semi-pelagic species like hake and blue whiting (5%) and cephalopods (mainly octopuses) (5%). Industrial exploitation along the West African coast is more recent than in the other upwelling ecosystems (Figure 9-3) and was initiated by European countries. In contrast, exploitation of pinnipeds started many centuries ago and almost led to extinction of the monk seal Monachus monachus, which shows no signs of recovery despite increasing protection of the few hundred individuals surviving in the southern part of the CanCE (CMS 1999). Another characteristic of the CanCE is that there are few marine bird colonies (Cushing 1969) and consequently little competition between fishing activities and bird predation.
Modelling forecasting and scenarios in upwelling LMEs
193
Figure 9-3. Landings of fish (excluding tunas) and shellfish from the Canary Current ecosystem for 1950-2001 and from the Iberian Peninsula for 1973-2001, (Source: FAO and ICES data).
Drastic fluctuations of abundance have been observed for sardines, sardinellas and other pelagic and bottom fish during the last fifty years. A downward trend shown by long-lived versus short-lived species ratios in the commercial fisheries (Figure 9-4) is caused by the partial replacement of longer-lived bottom fish species by short-lived small pelagics and cephalopods. Uncontrolled exploitation during the 1960s drastically reduced stocks of sparids and other ground fish but allowed growth of other fish stocks (Caddy 1983; Gulland and Garcia 1984, FAO 1986; Roy and Cury 2003). Increased cephalopod landings after the late 1960s partly reflect increased market demand. However, indirect evidence indicates that heavy fishing pressure removed the cephalopods’ predators (Caddy 1983, Caddy and Rodhouse 1998) while discards enhanced the abundance of some scavenger species that constitute cephalopod prey (Balguerias et al. 2000). In the 1970s, snipefish (Macrorhamphosus scolopax and M. gracilis) abundance increased dramatically in Iberian and Moroccan waters (Brêthes 1979) while triggerfish (Balistes carolinensis) spread geographically from Ghana to Mauritania, colonizing both pelagic and demersal ecosystems (Caverivière 1991; Fréon and Misund 1999). After attaining a peak biomass >1 million tonnes at the end of the 1970s, these populations collapsed in NW African waters (Sætersdal et al. 1999; Belvèze 1984). These events remain largely unexplained. Since 1969, the southern sardine stock and fishery have grown rapidly and spread further south (Holzlohner 1975; Barkova and Domanevsky 1976), while to the north sardine has come to dominate the small pelagic fish community (Gulland and Garcia 1984). This trend seemed to reverse in the mid-1990s when sardinellas extended north of Cape Barbas, and the sardine stock off Sahara crashed from ~5 to ~1 million tonnes
194
P. Fréon et al.
Figure 9-4. Evolution of the ratio of long lived to short-lived species in the landings from the Canary Current ecosystem from 1950 to 2001. in 1997 off Sahara. Sardines recovered gradually after 1997, whereas other small pelagic species remained abundant in the region (Anonyme 2003). These observed changes seem to be climate-driven. Quero et al. (1998) and Brander et al. (2003) noticed northward shifts of commercial and non-commercial fish distribution from southern Portugal to northern Norway since the late 1980s. As in the case of the triggerfish and the snipefish, the sardine collapse in 1997 does not appear to be linked to fishing pressure. Fish population abundance in the CanCE has been linked to climatic indices. Longterm changes in winds off Portugal in recent decades, related to NAO, modify Iberian upwelling patterns, and thus the annual catch of sardine (Borges et al. 2003). Roy and Reason (2001) found significant correlation between ENSO, NAO and upwelling intensity in the CanCE, and suggested that global environmental signals affect the fish populations through atmospheric teleconnections. On the other hand, sardine abundance fluctuates differently in different zones of the CanCE (Kifani 1998; ICES 2002a, 2002b; Borges et al. 2003; Carrera and Porteiro 2003). Decadal changes of some sardine populations in the CanCE parallel some in other eastern boundary currents, but within any region, different populations may be in or out of phase, which renders difficult any teleconnection hypothesis. Humboldt Current ecosystem (HCE) Physical traits The HCE extends from northern Peru (4°S) to central Chile (40°S). It stands out from other eastern boundary current systems because it extends very near to the equator. It is directly influenced by ENSO and displays the most extended, most superficial and most depleted minimum oxygen layer (MOL). The HCE is characterized by a complex set of flows and counterflows that persist between the coast and 1000 km
Modelling forecasting and scenarios in upwelling LMEs
195
offshore (Alheit and Bernal 1993). The oceanic sector is dominated by a sluggish, wide, equatorward flow of about 0.04 ms-1 that coincides with the boundary of the anticyclonic gyre of the Southeast Pacific, and is recognized as the Chile-Peru Oceanic Current. Nearer the coast, a southward counter-flow of about 0.06 ms-1, the Peruvian Oceanic Counter-current, has been identified at 79°W off Peru and between 76° and 77°W, 500 km off the Chilean coast. The faster (~0.18 ms-1) Humboldt Current is located somewhat closer to the coast, between 300 and 400 km offshore, with a core at about 200 m depth. Between the Humboldt Current and the coast, three seasonally varying and unstable branches can occasionally be distinguished. The most permanent branch is a counter-flow to the south centered at about 100 m depth off Peru and at 200 m depth off Chile. This Gunther Current or Peru-Chile Undercurrent is oxygen-poor and nutrient-rich. A spectrum of plumes, eddies, filaments and other transient structures has been observed (Montecino et al. in press). A thorough description of coastal ocean circulation off western South America is given in Strub et al. (1998). In Peru, coastal upwelling reaches a maximum during winter and a minimum during summer. In northern Chile (18°-30°S) it peaks during late spring and off Central Chile (30°-38°S) it peaks during late spring and summer. This temporal progression of coastal upwelling from the north to the south results from the displacement of the subtropical centre of high atmospheric pressure that intensifies and moves southward as summer progresses. Therefore, off Peru, upwelling occurs all year, whereas in central Chile it is restricted to spring and summer. Productivity Carr (2002) compared the productivity of the four eastern boundary upwelling systems using different satellite-borne colour sensors data from 1997 to 1999 and found that, despite higher fish productivity in the HCE, its productivity per unit area and its total production (rate x surface area) were lower than in the BCE and CanCE, in contrast with earlier works based on in situ measurements and estimated production areas. The lower values found by Carr (2002) are largely explained by a better estimation of production area for the Humboldt (half the size of the CanCE) and the occurrence of the strong El Niño event of 1997-98 during the study period. A time series of zooplankton volume started in 1961 indicates large interannual variations related to regime shift (Ayon et al. 2004). The species identification for this series promises to shed further light on this topic. In Chilean coastal waters, much chlorophyll biomass is found within 10-50 km of the coast. The maximum surface chlorophyll occurs in austral summer off both Peru and Chile despite the above-mentioned out of phase upwelling seasons between these two countries. The dominant zooplankton taxa are copepods (Calanus chilensis and Centropages brachiatus), euphausiids and the large holozooplankton such as salps, appendicularia (tunicata), siphonophores (cnidaria) and chaetognaths (Montecino et al., in press). Fish and fisheries The HCE supports extremely high fish production which is dominated by anchovy, although the usual mix of other pelagic fish stocks characteristic of eastern boundary
196
P. Fréon et al.
systems is also present: sardine, horse mackerel and mackerel. Since the beginning of industrial fisheries, catches were mainly dominated by anchovy whereas from 1977 to 1998 sardine catches were substantial (Figure 9-5).
Landings (Million tonnes)
Catches of horse mackerel and mackerel are less important. The most important demersal fish in the HCE is hake. The reason why the HCE has the highest fish yield whereas its satellite-based production estimates are lower than in the BCE and CanCE are still unclear. Among various hypotheses, Carr (2002) favoured the differences in trophic structure or spatial and temporal accessibility for different upwelling systems.
14 12
(b) Anchovy
Sardine
10 8 6 4 2 0 1955 1960 1965 1970 1975 1980 1985 1990 1995 2000
Figure 9-5. Alternating anchovy and sardine regimes in Humboldt Current ecosystem. (a) First principal component of normalized interdecadal SST time series from coastal stations (solid line) and Extended Reconstructed Sea Surface Temperature data (dashed line; Smith and Reynolds 2004; http://lwf.ncdc.noaa.gov/oa/climate/research/sst/sst.html) Montecinos, A., Purca, S. and Pizarro, O. Interannual-to-interdecadal sea surface temperature variability along the western coast of South America, Geophys. Res. Lett., 30 (11), 1570, doi:10.1029/2003GL017345. 2003. Copyright 2003 American Geophysical Union. Reproduced by permission of American Geophysical Union. (b) catches of anchovy and sardine in Peru (Miguel Niquen, IMARPE, Peru, pers. comm.) and Chile (Anuarios Estadísticos de Pesca, Servicio Nacional de Pesca de Chile, SERNAPESCA-Chile).
Modelling forecasting and scenarios in upwelling LMEs
197
The fisheries are sensitive to different scales of variability, especially the interdecadal variability that translates into changes in water mass characteristics (in terms of temperature, plankton structure, etc.) and the El Niño/ La Niña events (Chavez et al., 2003; Alheit and Niquen 2004; Bertrand et al. 2004a). Furthermore, the location and depth of the MOL plays a major role in the distribution and sometimes mortality of fish and ichthyoplankton (Mathiesen 1989; Ulloa et al. 2001; Montecino et al., in press).
PART 2. REVIEW OF MODELLING AND FORECASTING EFFORTS IN THE THREE ECOSYSTEMS PHYSICS California Current system Because near-shore and offshore currents have distinctive spatial scales (1-10 km nearshore and 100-1000 km offshore), they are usually measured and modelled with separate methods, implicitly assuming their interactions are weak. The few realistic regional modelling studies of the CalCE mostly have used simplified dynamics, domains, and forcing, with coarse spatial resolution and/or short integration times. These studies nonetheless have identified the primary mechanisms for seasonal, mesoscale, and sub-mesoscale variability of the CalCE (i.e. wind forcing, Kelvin waves, Rossby-wave propagation, and a large range of instability processes). Batteen’s (1997) model results indicate that, consistent with observations, the seasonal cycle in the CalCE is largely a deterministic response to the forcing, with phase and amplitude shifts due to Rossby waves. Strong intrinsic variability emerges from many numerical solutions (Ikeda et al. 1984; Auad et al. 1991; McCreary et al. 1991; Haidvogel et al. 1991; Pares-Sierra et al. 1993; Batteen 1997, Haney et al. 2001), which makes smallscale forecasting imprecise. Quasi-geostrophic models implicate baroclinic instability as the cause of variability of offshore currents; however, they lack the ability to produce sharp fronts and their associated instabilities. More recently, Marchesiello et al. (2003) used the Regional Oceanic Modeling System (ROMS) to simulate the mean-seasonal equilibrium CalCE with realistic dynamics and domain configuration. The level of eddy kinetic energy in their high resolution solutions is comparable to drifter and altimeter estimates. Since the model lacks transient forcing, they conclude that the dominant mesoscale variability in the CalCE is intrinsic. Eddy generation is mainly by baroclinic instability of upwelling, alongshore currents. There is progressive movement of mean-seasonal currents and eddy energy offshore and downwards into the oceanic interior in an annually repeating cycle. The associated offshore eddy heat fluxes essentially compensate near-shore cooling caused by transport and upwelling. The currents are highly non-uniform along
198
P. Fréon et al.
the coast; capes and ridges give rise to mean standing eddies and transient filaments and fronts. ROMS is being used operationally by the Jet Propulsion Laboratory (http://ourocean.jpl.nasa.gov) for CalCE, with nesting to zoom into the Monterey region. A VAR3D-type assimilation scheme (Li et al., in prep.) allows forcing with satellite as well as local in situ data. A similar approach has also been developed by the Office of Naval Research using the Princeton Ocean Model (POM). Canary Current system Analytical studies of the dependence of upwelling on wind forcing in the CanCE provide reasonable possibility of forecasting ocean response to atmospheric perturbation (Arístegui et al., in press). More accurate representation, which includes other factors (topography, boundary conditions, tides, etc.), can be expected from hydrodynamic models, although the capability of models is limited by the accuracy of wind forcings. There has been limited effort to model the hydrodynamics of the CanCE until very recently. A 2D tidal circulation model of the Atlantic continental shelf of the Iberian Peninsula using a finite element triangular grid of variable size produced output that compared well with coastal tide gauges, and provided a database for forecasting sea levels and currents in this area (Sauvaget et al. 2000). Johnson and Stevens (2000) modelled the region from Finisterre to the Canary Islands using monthly mean winds from the European Centre for Medium-range Weather Forecasts. They used the regional Modular Ocean Model with a horizontal resolution of 20 km and 36 vertical levels to reproduce many features of the circulation between the Canary Islands, the Azores and the Strait of Gibraltar, including a quasi-continuous slope undercurrent. Bateen et al. (2000) also showed that seasonal wind forcing was sufficient to produce eddy and filament-like features, especially in the presence of realistic coastline configuration. Neither model included realistic topography, however, and did not reproduce the smaller scale features of the CanCE. A similar scale hydrodynamics model of the Iberian region, coupled to a biological model including eight state variables from nutrients to zooplankton and detritus (Slagstad and Wassmann 2001), modelled the physics and phytoplankton production satisfactorily, but not the carbon export. More recently, an implementation of the ROMS hydrodynamics model in the same region (Peliz et al. 2003a, 2003b) investigated the poleward flow response to the interactions of two forcing mechanisms: an along-coast density gradient and wind forcing. On the larger scale, IRD (Institut de Recherche pour le Développement, France) and LPA (Laboratoire de Physique de l’Atmosphère, Senegal) are implementing a similar model for the whole CanCE, with two-level imbedded models allowing for fine scale study of key areas of ecological interest, e.g. retention areas (Marchesiello et al. 2004).
Modelling forecasting and scenarios in upwelling LMEs
199
The “parent” model has horizontal resolution of 25 km, and is aimed at describing the seasonal oscillation of the Azores and Canary Currents as well as the branches of the Equatorial counter-current. The first level of imbedding, presently implemented only off Senegal at a resolution of 5 km, will allow representation of mesoscale processes such as eddies and filaments. A second level of imbedding is currently implemented with 1 km grid cells in order to describe sub-mesoscale processes and circulation in bays and around capes arising from tidal and localised wind forcings. Humboldt Current system Basin scale hydrodynamic models represent the Pacific Ocean variability in temperature, salinity and current, using different codes such as POM, or the Ocean General Circulation modelling System (OPA). Some models are mainly process oriented, others are used for prediction. The latest version (5) of the Lamont-Doherty Earth Observatory (LDEO) ocean–atmosphere coupled model incorporates an assimilated SST field, which directly affects the surface wind field and has a persistent effect on the coupled system (Chen et al. 2004). The model’s internal variability generates a self-sustaining oscillation with El Niño-like periods and amplitudes. Using sea level, winds and SST for initialisation, the model satisfactorily forecasted monthly SST for the period 1857 to 2003 at lead times of up to two years (Figure 9-6), including all prominent El Niño events within this period. Penven et al. (2003) used the ROMS code to model the mean 3d-circulation of the Peruvian upwelling system, its seasonal cycle and its mesoscale dynamics at intermediate scales.
ECOLOGY California Current ecosystem Parrish and MacCall (1978) developed a forecast model for chub mackerel (Scomber japonicus) recruitment in the California Current region by incorporating environmental variables in the Ricker stock-recruitment relationship (Figure 9-7). Sea level and an index of transport collected from the late 1920s to the late 1960s were included into the model. Despite high r2 values (0.60 and 0.76) forecasting failed in recent years due to a regime shift. Conser et al. (2002) found that sardine productivity (fraction of mean sustainable yield) could be expressed as a quadratic function of the 3-year average temperature at
200
P. Fréon et al.
Figure 9-6. Hindcasts of El Niño and La Niña in the past 148 yr. using the LDEO5 model: (a) time series of SST anomalies averaged in the NINO3.4 region (58 S–58 N, 120–1708 W). The red curve is monthly analysis of ref. 12 and the blue curve is the LDEO5 prediction at 6-month lead; (b) Six of the largest El Niños since 1856. The thick red curves are observed NINO3.4 SST anomalies, and the thin curves of green, blue, magenta and cyan are predictions started respectively 24, 21, 18 and 15 months before the peak of each El Niño (reproduced from Chen et al. 2004).
Modelling forecasting and scenarios in upwelling LMEs
201
Figure 9-7. Observed and predicted recruitment of Pacific mackerel. (a) Ricker sea level model, (b) Ricker transport model (Reproduced from Parrish and MacCall, 1978 with permission from the “California Department of Fish and Game” and "NOAA Fisheries").
Scripps Pier, La Jolla, California. The Pacific Fishery Management Council adopted a policy to decrease sardine annual allowable catches from 15% to 5% of spawning biomass when average temperature falls below a threshold value. Based on temperatures observed since adoption of this procedure (1983-on), the indicated fraction of spawning biomass has consistently been 15% (Conser et al. 2004). Smith and Moser (2003) pointed out that while extrapolation from observed oscillations may provide useful short term predictions, the inherently unpredictable nature of regime shifts implies that vigilance and a cautious fishery regulation provide the best prospect. A good example of spatial prediction is provided by MacCall’s “basin model” which is based on the concept of density-dependent habitat suitability (MacCall 1990). During periods of low abundance, the population is restricted to the best-suited habitat, but the higher the biomass, the more extended is the population distribution. This model was successfully applied to the distribution of the Californian northern anchovy population Canary Current ecosystem A few models coupling physical forcing with fishery data at a yearly scale were developed in the region. For instance, one conceptual model off Portugal indicated that upwelling during winter spawning from 1993-1997 caused offshore transport of
202
P. Fréon et al.
larvae, impacting negatively on sardine (Sardina pilchardus) and horse mackerel (Trachurus trachurus) recruitment and catches the following year (Santos et al. 2001). However, these conclusions, based on few years of data, require further validation before being used for forecasting. As a first step in that direction, Santos et al. (2004) used a simplified 2D model and Lagrangian tracking to simulate the observed distributions of sardine eggs and larvae. Other models, based on linear or non-linear multiple regressions between fish abundance proxies (catches or catches per unit of effort (CPUE)) as a dependent variable and fishing effort and environmental variables, are largely empirical. Despite efforts to combine these models with existing surplus production models to reduce spurious correlations (Fréon et al. 1993), this approach is inherently limited. Nevertheless, practical applications of this kind of model have been implemented in Senegal, Morocco and Côte d’Ivoire, where the interannual variations of clupeoid species abundance was related to fishing effort and an upwelling index (Fréon 1988, 1989). The models explained between 72% and 94% of the CPUE variability of the 20-year time series (Figure 9-8). Lack of systematic CPUE data in recent decades has prevented the updating of these models (Do Chi 1994). A similar empirical model describes variations in octopus (Octopus vulgaris) recruitment in Senegal according to different upwelling or retention indices computed from wind or satellite data (Demarcq and Faure 2000; Caverivière and Demarcq 2002; Laurans et al. 2002). Here also, the forecasting capability of these models cannot be assessed for lack of updated systematic observations. Other empirical relationships have been found between, for instance, the time of arrival of adult sardinella on the Petite Côte of Senegal at the beginning of the year and the time of relaxation of upwelling off Mauritania, where they originate. This is suggested by negative correlations between monthly anomalies of CPUE and upwelling indices (Fréon 1986). Similarly, the migration of the emblematic thiof (Epinephelus aeneus) along the north coast of Senegal seems related to both the onset of the upwelling in Senegal and its relaxation in Mauritania (Cury and Roy 1988). In neither case was the relationship updated, and so evaluation of its forecasting ability is difficult. Humboldt Current ecosystem In Chile, empirical models similar to those developed for West Africa were used to link CPUE of pelagic species to fishing effort and/or environmental variables related to the upwelling intensity or turbulence (Yáñez et al. 2001). A similar model used on anchovy data from 1957 to 1977 considered the Peruvian and Chilean stocks as a single unit (Fréon and Yáñez 1995). Attempts to update this model were unsuccessful due to profound changes in the fisheries and a changed response of the ecosystem to environmental perturbation after the regime shift in the 1970s.
Modelling forecasting and scenarios in upwelling LMEs
203
Figure 9-8. Additive exponential model where the upwelling index (mean wind speed, V) influences the Sardinella spp stock abundance and catches (Y) in the Senegalese fishery, from 1966 to 1983. Upper graph: theoretical catches according to V values and fishing effort (solid curves) and observed catches (line with annual dots). Lower graph: time series of observed and predicted total catches (Fréon 1986, 1989).
In Northern Chile, operational models were used to predict favourable fishing grounds for small pelagic species (Silva et al. 2000; Nieto et al. 2001) and sailfish (Barbieri et al. 2000). These models couple real-time satellite data to an expert-system that learned from previous relationships between CPUE distribution and historical satellite data of SST and/or chlorophyll. In Peru, El Niño events impact the small pelagic fishery to such a large extent that IMARPE (Instituto del Mar del Perú) developed a series of tools to manage this pelagic fishery in real time according to environmental conditions and abundances
204
P. Fréon et al.
estimated from acoustic and fishery surveys (www.imarpe.gob.pe). Month-long scientific acoustic surveys by one or two research vessels are normally performed 2 to 4 times a year. These are augmented during El Niños with near-synoptic surveys (<48 hours) of the entire Peruvian continental shelf by up to 40 fishing vessels equipped with commercial echo-sounders. Furthermore, IMARPE observers on the vessels send real-time information on the catch composition (species and body length), SST, and ancillary data to a central office via satellite (Ñiquen et al. 2000; Bertrand et al. 2004b). On the basis of this information and previous experience, IMARPE advises the fishery minister on fisheries management. The advice has often translated into partial or total closure of the fishery with immediate application. Although no official operational management procedure (OMP) is formally documented, this method likely was one key factor in helping prevent collapse of the anchovy stock during the long and severe El Niño of 1997-1998 (Bouchón et al. 2000, Ñiquen et al. 2000; Bertrand et al. 2004a). The challenge in maintaining such OMPs is that not all El Niño events have similar effect, even when their temperature anomalies have the same amplitude (Bertrand et al. 2004a; Niquen and Bouchon 2004). TELECONNECTIONS BETWEEN ECOSYSTEMS Regime-like signals appear in many fisheries, which though distant may be linked through environmental teleconnections. These climatic oscillations can potentially induce synchrony among pelagic fish populations living in different ecosystems, as has been observed around the world (Kawasaki 1983; Klyashtorin 1998; Schwartzlose et al. 1999). Nonetheless, most of the observed synchronies are statistically insignificant, given that the series are only a few decades long and largely auto-correlated, and therefore have few degrees of freedom. In the absence of clearly identified processes linking the various ecosystems, one cannot yet firmly conclude that remote synchronies exist (Fréon et al. 2003), nor use them for prediction.
PART 3. DEVELOPMENT OF BIOLOGICAL SCENARIOS BASED ON FORECASTED PHYSICAL SCENARIOS Recently, a number of authors attempted to predict the reaction of eastern boundary upwelling ecosystems to global climate change (Bakun 1990, Mote and Mantua 2002, Snyder et al. 2003, Diffenbaugh et al. 2004, Bakun and Weeks 2004). Citing evidence from different eastern boundary systems, Bakun (1990) suggested that global greenhouse warming was deepening the continental low pressure centres, thereby enhancing the cross-shore atmospheric pressure gradient to intensify the alongshore sea surface wind stress and coastal upwelling. This was confirmed by Schwing and Mendelssohn (1997) and Snyder et al. (2003) for the California Current system, and Mendelssohn and Schwing (2002) demonstrated common trends in the California and Humboldt Current systems. Different models were applied to test the impact of future warming (Mote and Mantua 2002; Snyder et al. 2003; Diffenbaugh et al 2004; Bakun and Weeks 2004), with very different outcomes. Mote and Mantua (2002) used two
Modelling forecasting and scenarios in upwelling LMEs
205
“state-of-the-art climate models” and reported that, in each of the upwelling zones they modelled, the seasonal cycle of upwelling remains unchanged throughout the 21st century. In contrast, Snyder et al. (2003), who used a regional climate model to estimate changes in wind-driven upwelling, suggested that substantial changes of upwelling will occur under increased CO2 concentrations. However, different regions of the California Current system respond differently, resulting in changes in phase and an intensified peak season in the northern limb and a muted peak season in the southern limb (Figure 9).
April
May
June
July
Aug.
Sept
Figure 9-9. Difference of the monthly average modeled wind-stress curl (N/m3), calculated as: (19801999) – (2080-2099) (Snyder, M.A., L.C. Sloan, N.S. Diffenbaugh and J.L. Bell. Future climate change and upwelling in the California Current. Geophys. Res. Lett., 30: 1823-1826, 2003. Copyright 2003 American Geophysical Union. Reproduced by permission of American Geophysical Union).
Summing up, eastern boundary current upwelling systems (or parts of them) might react to global warming with either increased or decreased upwelling intensity. We will describe now how long-term dynamics of key populations in the three ecosystems could react to increased or decreased upwelling intensity, following suggestions of DeAngelis and Cushman (1990), but without using the modelling toolbox they suggested. California Current ecosystem A warming trend of about 1°C in SST from 1950 to 1999 has been observed along the Southern California coast (Bograd and Lynn 2003; Roemmich 1992). Increasing stratification and declining zooplankton have been linked with this warming trend
206
P. Fréon et al.
(McGowan et al. 2003; Roemmich and McGowan 1995). On the other hand, evidence has been found that upwelling-favourable winds have increased in recent decades (Schwing and Mendelssohn 1997). This increase in upwelling-favourable winds is consistent with the ideas of Bakun (1990) and also with recent regional climate modeling experiments (Snyder et al. 2003), which suggest that alongshore winds intensify as a response to warmer ocean temperatures. Upper ocean warming and increased upwelling-favourable winds may, at first, appear as two inconsistent scenarios for global warming effects. A recent regional modeling experiment using ROMS, (Di Lorenzo et al. 2005) shows that two opposed processes can explain this apparent inconsistency. On one hand, a cooling of about 0.5°C in SST is driven in their ocean model by the 50 year NCEP wind reanalyses, which contain a positive trend in upwelling favourable winds along the California coast. However a net warming trend of 1°C in SST occurs when the effects of surface heat fluxes are included as forcing functions. Previous analyses of observed surface heat fluxes (Cayan 1992) and ocean model hindcasts of the surface layer heat budget (Miller et al. 1994) confirm that along the California coast the long-term SST signal is indeed dominated by changes in surface heat fluxes. The ocean model simulations show that increased stratification associated with the warming reduces the efficiency of coastal upwelling in lifting subsurface waters to the ocean surface, masking any effects of the increased strength of the upwelling winds.
This scenario is consistent with a reduction of the nutrient supply at the coast, which may in turn explain the observed decline in zooplankton concentration. It is worthwhile noting that, in the simulations, increased static stability arising from upper ocean warming leads to increased mesoscale eddy variance and therefore increased cross-shore transport of material properties. This may provide an additional scenario for declining coastal planktonic communities due to warming trends. As a result, one can expect a decrease in abundance of small pelagic fish due to both dispersion of their larvae and decrease in prey for adults. Nonetheless this scenario remains largely speculative because we are ignorant of how this trend will interact (or not) with natural interdecadal variability of pelagic fish abundance (Baumgartner et al. 1992), whose underlying processes remain unknown. If prediction had to be based only on the longterm autocorrelation in sardine abundance observed in the time series of sediment cores, the recent recovery of the Californian sardine stock should last for a few decades. Canary Current ecosystem The northward shifts in distribution of commercial and non-commercial fish species that occurred from southern Portugal to northern Norway from the late 1980s has been related to an increase in water temperature (Quero et al. 1998; Brander et al. 2003). Although one should be extremely cautious in forecasting long-term scenarios, these changes in fish distribution and community could be accentuated by a further warming. Under this scenario, the northward extension of Mediterranean and NW Africa tropical or subtropical species could be exacerbated, whereas a scenario of cooling temperature could lead to a restoration of the previous situation, especially if
Modelling forecasting and scenarios in upwelling LMEs
207
the exploitation rate remains at a reasonable level for all compartments of the ecosystem. Prediction related to global changes might be hampered by independent (and poorly understood) decadal fluctuations in the abundance of small pelagic fish species and related changes in the ecosystem, as observed on the Iberian shelf (Borges et al. 2003; Cendrero 2002). Here and in similar upwelling ecosystems, pseudo-cycles of 40 to 60 years have been observed (Klyashtorin 1998; Schwartzlose et al. 1999). Based on empirical relationships between pelagic fish abundance and climatic indices (Klyashtorin 2001) or simply on autocorrelation (Fréon et al. 2005), one can infer that the present decline in sardine abundance should reverse during the next decade. Once more this type of prediction is highly uncertain due to its empirical nature. On the West African shelf, if there is no major change in the seasonal cycle of alongshore wind stress during the next decades, as supposed by Mote and Mantua (2002), the ecosystem should mainly undergo changes in the exploitation rates of the different compartments of the ecosystem. Nonetheless, population outbursts similar to those previously mentioned, but possibly in other species, are likely to occur unpredictably. Though triggered by largely unknown processes, such events seem favoured by an unbalanced exploitation of resources (Gulland and Garcia 1984; Caddy and Rodhouse 1998; Verity et al. 2002). Under the scenario of intensified upwelling, one expects enhanced southward migration of the central sardine stock, favouring the Moroccan fishery. The southern sardine stock should also display southward migration and extension of its habitat in Mauritania while sardinellas should retreat. In Senegal, the abundance of sardinella species should increase. Decadal changes in upwelling strength and SST as observed around the 1970s are likely to occur and are difficult to predict, despite some understanding of their functional link with the major atmospheric patterns, e.g. North Atlantic Oscillation (Arístegui et al., in press). The observation that warm ENSO (El Niño/Southern Oscillation) events in the Pacific lead positive SST anomalies in the southern CanCE by about 3 to 6 months during late winter and early spring (Enfield and Mayer 1997; Roy and Reason 2001) may allow prediction of fish population abundance. In several studies (Fréon and Stéquert 1976; Fréon 1989; Belvèze 1991; Binet 1997; Binet et al., 1998) high pelagic fish abundance and catches were related to enhanced coastal upwelling, therefore recruitment and population growth. Nonetheless, because the origin of this lag is not fully understood, associated predictions remain uncertain. Humboldt Current ecosystem As in other eastern boundary currents, the decadal-scale dynamics of the HCE are governed by alternating anchovy and sardine periods (Schwartzlose et al. 1999, Chavez et al. 2003, Alheit and Niquen 2004; Bertrand et al. 2004a), which reflect the entire restructuring of the ecosystem from phytoplankton to the top predators (JarreTeichmann 1998, Alheit and Niquen 2004) (Figure 9-5). These regime shifts are
208
P. Fréon et al.
associated with lasting warm or cool periods as warm subtropical oceanic waters approach or retreat from the Peru–Chile coast. Although fisheries might give a biased index of abundance due to the influence of anthropogenic factors, it seems that cool phases parallel anchovy regimes (1950s-early 1970s; 1985 on) while the warm period (early 1970s-1985) clearly favoured sardine recruitment, taking into account a 3-year lag due to the age structure of the catches of this species. Like an El Niño, the warm periods drastically change trophic relationships in the entire HCE exposing the Peruvian anchovy to a multitude of adverse conditions. Higher temperatures off Peru restrict the anchovy to the cooler upwelling zone at the coast, thereby decreasing their area of distribution and spawning. This concentration of the population favours egg and larval cannibalism and dramatically increases catchability. Increased spatial overlap between anchovies and the warmer water preferred by sardines permits heavy sardine predation on anchovy eggs. The anchovies’ diet of phyto- and zooplankton is limited to a narrower coastal zone of restricted upwelling in warm years. Both phytoplankton and zooplankton have lower volumes, the latter likely reflecting a diminution in the large copepods, their main food source. Horse mackerel and mackerel increase predation pressure on anchovies by invading the anchovy habitat in warmer years (Alheit and Niquen 2004). In this ecosystem, where "variability is normality" (M. Espino, personal communication), fish have developed adaptive strategies in space and time (Bertrand et al. 2004a) and trophic relationships are not the only parameters affecting population fluctuation. Thus, to understand the effect of any climatic event on fish, various factors occurring at different spatiotemporal scales (e.g., the inter-decadal regime; the ENSO situation; the population's condition before the event; fishing pressure and other predation; the adaptation of reproductive behaviour; the presence of local efficient upwelling) have to be considered. Such an integrated approach from fine to large spatio-temporal scales helps understanding of variations in fish population (Bertrand et al. 2004a). If the greenhouse gas build-up leads to increased upwelling, the HCE will support a higher phyto- and zooplankton biomass, the anchovy would be favoured and we could expect more and/or longer anchovy periods (Alheit and Niquen 2004). The size spectrum of copepods would shift towards larger specimens favourable for anchovy feeding. Predation on anchovy eggs, larvae, juveniles and adults would decrease because of less overlap with warm water predators and reduced density-dependent effects. Anchovy catches could increase to higher levels sustainable for longer periods, given application of an appropriate fishing policy. If, in contrast, future greenhouse gas build-up leads to decreased upwelling in the HCE, we could expect more and/or longer sardine periods resulting in reduced anchovy and augmented sardine catches. Nonetheless this kind of prediction, largely based on a bottom-up control, remains uncertain because top-down (Cushing 1971) or wasp-waist (Cury et al. 2001) controls, or selective and behavioural processes resulting in diet switches (Verity et al. 2002), can structure the functioning of pelagic ecosystems. In Peru for instance, recent plankton data question the exclusive scheme of bottom-up control (Ayon et al. 2004).
Modelling forecasting and scenarios in upwelling LMEs
209
Bakun and Weeks (2004) recently suggested a different scenario resulting from increased upwelling intensity. The world’s strongest eastern boundary upwelling zone, the Lüderitz cell in the Benguela Current ecosystem off Namibia, is characterized by widespread hypoxia and massive eruptions of noxious, radiatively active gases. The rapid offshore transport of the ocean surface layer (plus its organic and inorganic contents) in the Lüderitz cell prevents herbivorous copepods from maintaining substantial populations within and adjacent to the strongly divergent upwelling zone. This allows massive build-up of phytoplankton biomass, much of which sinks unutilised to accumulate as thick deposits of unoxidized organic matter on the sea floor. Generally increased upwelling could lead to extensive sea bed areas of low or zero dissolved oxygen concentrations, which could eventually lead to eruptions of poisonous gases from the sediment with detrimental effects to fish and fisheries. Bakun and Weeks (2004) suggest a hypothetical biological buffering process that might counter such a scenario. Sardines, because of their very fine-meshed gillraker structures, which allow them to filter and consume microscopic phytoplankton directly, could constrain phytoplankton growth by heavy grazing. In such a way, increased upwelling, in contrast to the scenario developed above, might eventually lead to a sardine regime, providing that the stock is not overexploited as is presently the case. CONCLUSIONS AND FURTHER GENERAL THOUGHTS The three ecosystems briefly presented here are all upwelling systems physically forced by local winds and influenced by larger scale teleconnections. All are characterised by strong alongshore advection, a poleward undercurrent, generally weak tidal mixing, weak winter convective mixing, zero to moderate precipitation and coastal freshwater input, and, finally, high productivity of plankton and pelagic fish resulting in the dominance of pelagic vs demersal consumers. All these properties are shared with the Benguela ecosystem. Another feature common to the four ecosystems is the strong variability at interannual to decadal time scales. This variability manifests at all levels: hydrodynamics and water properties, primary and secondary production, abundance and species composition, and fisheries (Mackas et al., in press), which renders forecasting a high priority, although a challenge. These systems also differ in a number of ways which depend on which ocean they are located in, on geographical and topographical characteristics, and on seasonal and shorter term temporal variability related to wind forcing, atmospheric fronts and waves, coastal trapped waves, and stratification (Table 9-1). Major differences are the great width of the Pacific basin, which allows ENSO development, and the related existence of “older”, nutrient-rich (therefore low oxygen) sub-pycnocline water in the Pacific. The South Atlantic coastal boundary also ends about 15° further north than the Pacific, so that the BCE is open to both Antarctic and Indian Oceans at its southern limit. The CanCE is characterised by the gap at the Gibraltar Strait and the interruption presented by the Canary Islands off NW Africa. Eddies and filaments are less important and frequent in the HCE than in the three other systems.
210
P. Fréon et al.
Table 9-1. Comparison of Pacific and Atlantic Eastern Boundaries (summarized from three larger tables from Mackas et al. in press; Ecosystem Feature added).
Latitude range
California Current LME
Canary Current LME
Humboldt Current LME
23-51°N
15-44°N (incl. Iberia 36-44°N)
5-45°S
f = 0.6 to 1.1 x 10
Shelf width
-4
Mainly <50 km but up to 100 km in places
f = -0.1 to -1.0 x 10
Benguela Current LME -4
-4
f = 0.9 to 1.0 x 10 25-75 km off Iberia, varies off NW Africa between 10 and 120 km.
Often <25 km but up to 100 km. Bordered by deep trench.
Strong ENSO. Sheltered Southern California Bight. Filaments off capes. CTW. Aeolian dust near 30N.
ENSO+NAO effects. Sheltered Gulf of Cadiz/Gibraltar Strait, Canaries archipelago. Filaments off capes. CTW? Ria systems in north. Aeolian dust off Sahara
Strong ENSO. Sheltered Gulf of Guayaquil in north. Few filaments? Strong CTW. Hypoxia and OML. High trophic efficiency & fish production
Wind stress & stratification
23-32°N all year, >32°N summer. Strong in S, Moderate in N. Low in S, high in N (Columbia).
29-44°N summer, 20-28°N all year, 15-20°N winter. Moderate. Mainly near zero, but significant at >37°N and at 15°N.
5-32°S all year, >32°S summer. Weak in N, Moderate-strong to S. Very low
Large variation in pelagic stock abundance and long period of collapse of the sardine stock Importance of salmonid stocks (northern part, coastal zone)
Dominance of sardine (north) or sardinella (south) compared to anchovy. Frequent burst of secondary species. Partial replacement of longer lived bottom fish species, by short lived small pelagics and cephalopods, in the commercial fisheries. Low abundance of top predators (birds, large sharks and mammals)
Most extended system. Most extended, most superficial and most depleted MOL. Dominance of anchovy versus sardine. Frequent burst of secondary species. Strong effect of ENSO events on all living organisms
Ecosystem features
f = -0.4 to -0.9 x 10
Physical features
Freshwater input
15-37°S -4
Mainly narrow off west coast (<25 km) but up to 250 km on zonal Agulhas Bank. ENSO + “Benguela Ninos”. Wide Agulhas Bank & Agulhas influence. Filaments & rings. Strong CTW. Hypoxia locally. Low fish production. 15-30°S all year, >30°S summer. Strong in S, moderate to N Generally low, with occasional incursions of Congo R. water, and seasonal incursions of Orange, Cunene and Angolan river waters. Alternation between anchovy dominated and sardine dominated regimes Spawning area of many species located upstream of their nursery area. Observed mortality of fish and shellfish due to H2S and/or low oxygen concentration and/or sulfide emissions.
The abundance of some top predators (pinnipeds, birds) is presently low in the CanCE compared to the other three and this ecosystem is also characterised by the dominance of sardine and sardinella, whereas anchovies are also important in the other
Modelling forecasting and scenarios in upwelling LMEs
211
ecosystems. The CanCE and the HCE display more frequent outbursts of secondary or rare species than the two other systems. Because many rivers flow into the northern CalCE, several species of salmon occur in coastal waters of this system. Finally, the biological variability in the two Pacific systems is largely dominated by El Niño/La Niña events, whereas in both Atlantic regions El Niño has weaker influence complemented by NAO in the north. In all regions, coastal trapped waves play a significant if not fully understood role. The way in which climate changes affect ecosystems, on both land and ocean, is complex and difficult to forecast, although efforts in meta-analysis, process studies and modeling with data assimilation are promising (Drinkwater and Myers 1987; DeAngelis and Cushman 1990; Bakun and Broad 2002). Uncertainty persists in the physical effects of climate change. Model based scenarios of the impact of global warming on the physical components of the earth system still remain coarse and uncertain. The poor resolution of these simulations introduces major uncertainties when trying to resolve regional scales. Further work is needed before global climate change scenarios will be directly applicable to the regional responses of the upwelling and other ecosystems. Furthermore, under any given scenario, additional uncertainty surrounds the biological responses to physical forcing. The effects of anthropogenic forcing, particularly increasing fishing pressure but also habitat invasions, eutrophication and diseases, complicate the issue (Verity et al. 2002). Increased fishing effort, and consequent increased fish mortality, will likely increase the relative abundance of low trophic levels in these ecosystems (Pauly et al. 1998) and might favour or exacerbate regime shifts and possibly population outbursts as observed in the CanCE. The most promising example of environmental forecasting from a model including data assimilation is provided by Chen et al. (2004) on El Niño/La Niña predictions. There is a clear need to develop similar environmental models coupling atmospheric and oceanographic processes in the other upwelling systems, with data assimilation to make them more realistic. One can expect that this kind of physical prediction will help in forecasting biological responses of the ecosystem based on our understanding of processes. In the mean time, biological or fishery related predictions can be made simply by using autoregressive properties of time series (e.g. Stergiou et al. 1997) but the resulting forecast is likely to fail whenever a major change (regime shift, modification of the exploitation pattern) occurs (Ulltang 1996). Simple or multiple regression models suffer from the same limitation, aggravated by the poor predictive power of these models due to limited degrees of freedom, the difficulty of selecting the right explanatory variables and the uncertainty in the functional form of variable relationships (Fréon et al. 2005). Furthermore, for purely mathematical reasons, when r2<0.65 the predicting value is limited (Prairie 1996). Nonetheless the positive side of this approach is that it identifies significant variables for inclusion in conventional population dynamics models (e.g. Fréon et al. 1993) or simple regressions. Processbased models, like the Bakun (1998) triad and the optimal environmental window (Cury and Roy 1989), are more satisfactory, assuming that the processes are correctly
212
P. Fréon et al.
identified and lasting. Our major challenge, of understanding and forecasting regime shifts, will likely require several decades of research to acquire a basic knowledge of the nature of the controlling factors and to unravel the complexity and variability in predation that complicate prediction (Bax 1998; Verity et al. 2002). In the near future, a two-level (short- and long-term) management strategy should be investigated for the small pelagic species that are exploited (Fréon et al. 2005). The first level could be a conventional adaptive management approach incorporating new ecosystem-based thresholds or Limit Reference Points as much as possible. The second level of the strategy should address the problem of inter-decadal variations in the abundance of pelagic fish that induce counterproductive investments in the fishing sector. Based on the pseudo periodicity of around 40 to 60 years, as observed in many large stocks, some management action could limit long-term investment in fishing units and related infrastructure, especially once a turning point is passed and therefore uncertainty decreases (Figure 9-10). This suggested approach requires further investigation to quantify the risks and benefits.
Figure 9-10. Schemes of typical variation in the abundance, effort and investment in pelagic fisheries: (a) pseudo-cyclic variation; (b) long periods of collapse; (c) investment strategies related to pseudo-cyclic variation. The size of question marks reflects the uncertainty about the trend in abundance over the next 10-year period (from Fréon P., Cury, P., Shannon, L. and Roy, C., Sustainable exploitation of small pelagic fish stocks challenged by environmental and ecosystem changes: a review. Fishery Bulletin, 2005).
Modelling forecasting and scenarios in upwelling LMEs
213
ACKNOWLEDGMENTS We are grateful to Arnaud Bertrand for his contribution to the Humboldt Current section of this work and to Dave Checkley for reviewing the California Current section. We acknowledge the comments of two anonymous referees. This is a contribution of the IDYLE and Upwelling ecosystem programmes of IRD and of the EUR-OCEANS network of excellence (sixth Framework Programme of the European Community) REFERENCES Abraham, C. and W.J. Sydeman. 2004. Ocean climate, euphausiids, and auklet nesting: Inter-annual trends and variation in phenology, diet and growth of a planktivorous seabird. Mar. Ecol. Prog. Ser., 274: 235-250. Alheit, J., and P. Bernal. 1993. Effects of physical and biological changes on the biomass yield of the Humboldt Current ecosystem. In Large marine ecosystems – Stress, mitigation and sustainability (pp. 53–68) K. Sherman, L.M. Alexander and B.D. Gold, eds. American Association for the Advancement of Science, Washington. Alheit, J.,and M. Niquen. 2004. Regime shifts in the Humboldt Current ecosystem. Progress in Oceanography 60: 201-222. Anonymous. 2003. Survey of the pelagic fish resources off North West Africa. Part III. Morocco 19 November-19 December 2003. NORAD/FAO PROJECT GCP/INT/730/NOR. Arístegui, J., P. Sangrà, S. Hernández-León, M.Cantón, A. Hernández_Guerra, and J.L. Kerling. 1994. Island-induced eddies in the Canary Islands. Deep-Sea Res. I, 41: 1509-1525. Arístegui, J., X.A. Alvarez-Salgado, E.D. Barton, F.G. Figueiras,S. Hernández-León, C. Roy, and A.M.P. Santos. Oceanography and fisheries of the Canary Current/Iberian region of the eastern North Atlantic. In The Sea, Vol 14. Robinson, A.R. and K.H. Brink, eds,. In press. Auad, G., A. Pares-Sierra and G.K. Vallis. 1991. Circulation and energetics of a model of the California Current system. J. Phys. Ocean. 21:1534-1552. Ayón, P., S. Purca, and R. Guevara-Carrasco. 2004. Zooplankton volume trends off Peru between 1964 and 2001. ICES J. mar. Sci. 61:478-484. Bakun, A. 1990. Global climate change and intensification of coastal ocean upwelling. Science, 247: 198201. Bakun, A. 1998. Ocean triads and radical interdecadal stock variability: bane and boon for fisheries management. Pages 331-358. In Reinventing fisheries management T. Pitcher, P.J.B. Hart and D. Pauly, eds.. Chapman and Hall, London. Bakun, A. Regime shifts. In The Sea, Vol. 13. Robinson, A.R. and Brink, K. eds. Harvard University Press, Cambridge, MA, in press. Bakun, A. and K. Broad, eds. 2002. Climate and Fisheries: Interacting Paradigms, Scales, and Policy Approaches. International Research Institute for Climate Prediction, New York: 67 p. Bakun, A. and R.H. Parrish. 1982. Turbulence, transport, and pelagic fish in the California and Peru current systems. Calif. Coop. Oceanic Fish. Invest. Resports, 23:999-112. Bakun, A. and S.J. Weeks. 2004. Greenhouse gas buildup, sardines, submarine eruptions and the possibility of abrupt degradation of intense marine upwelling ecosystems. Ecology Letters. 7:1015– 1023. Balguerias, E., E. Quintero and C.L. Hernandez-Gonzalez, C.L. 2000. The origin of the Saharan Bank cephalopod fishery. ICES Journal of Marine Science, 57:15-23. Barbieri, M.A., K. Nieto, C. Silva, and E. Yáñez. 2000. “Evaluation of potential swordfish (Xiphias gladius) fishing grounds along central Chile, by the use of remote sensed sea surface temperature satellite data”. In Environment and Development in Coastal Regions and Small Islands Project. Programa CSI, UNESCO. Barkova, N.A. and L.N. Domanevsky. 1976. Some peculiarities of sardine (Sardina pilchardus) distribution and spawning along the Northwest Africa. ICES. C.M. 1976/J: 6 Pelagic fish (Southern) Committiee.-15p.
214
P. Fréon et al.
Barth, J.A., S.D. Pierce and R.L. Smith. 2000. A separating coastal upwelling jet at Cape Blanco, Oregon and its connection to the California Current System. Deep-Sea Research Part Ii-Topical Studies In Oceanography 47(5-6): 783-810. Barton, E.D. 1989. The poleward undercurrent on the eastern boundary of the Subtropical North Atlantic. In Poleward flows along Eastern Ocean Boundaries. Ed S.Neshyba, R.L., Smith, C.N.K.Mooers. pp 82-95. Springer-Verlag Lecture Note Series. Barton, E.D., 1998. Eastern boundary of the North Atlantic: Northwest Africa and Iberia. Coastal segment (18,E). In The Sea, Vol. 11. Robinson, A.R. and K.H. Brink, eds. John Wiley and Sons, Inc., New York, pp. 633-657. Barton, E.D. and J. Arístegui. 2004. The Canary Islands Transition – upwelling, eddies and filaments. Progress in Oceanography, 62(2-4):67-69. Batchelder H.P., C.A. Edwards and T.M. Powell. 2002. Individual-based models of copepod populations in coastal upwelling regions: Implications of physiologically and environmentally influenced diel migration on demographic success and nearshore retention. Progress in Oceanography, 53:307-333. Batteen, M.L. 1997. Wind-forced modeling studies of currents, meanders, and eddies in the California Current system. J. Geophys. Res., 102: 985-1010. Batteen, M.L., J.R. Martinez, D.W. Bryan, E.J. Buch. 2000. A modeling study of the coastal eastern boundary current system off Iberia and Morocco. Journal of Geophysical Research 105 (C6): 1417314195 JUN 15 Baumgartner, T.R., A. Soutar & V. Ferreira-Bartrina. 1992. Reconstruction of the history of Pacific sardine and northern anchovy populations over the past two millennia from sediments of the Santa Barbara Basin, California. CalCOFI Reports, 33:24-40. Bax, N.J. 1998. The significance and prediction of predation in marine fisheries. ICES Journal of Marine Science, 55: 997-1030. Belvèze, H. 1984. Biologie et dynamique des populations de Sardina (Sardina pilchardus) peuplant les cotes atlantiques marocaines et propositions pour un aménagement des pêches. Doc. d’Etat. Université de Bretagne Occidental: 532 pp. Belvèze, H. 1991. Influence des facteurs hydroclimatiques sur la pêcherie marocaine de petits pélagiques côtiers. In Les pêcheries ouest-africaines: variabilité, instabilité et changement. Cury, P. and Roy, C. eds., ORSTOM, Paris. Bertrand S., E. Díaz and M. Ñiquen. 2004b. Interactions between fish and fisher’s spatial distribution and behaviour: an empirical study of the anchovy (Engraulis ringens) fishery of Peru. Ices. J. Mar. Sci., 61:1127-1136. Bertram, D.F., D.L. Mackas and S.M. McKinnell. 2001. The seasonal cycle revisited: Interannual variation and ecosystem consequences. Progr. Oceanogr., 49: 283-307. Bertrand A., M. Segura, M. Gutiérrez and L. Vásquez. 2004a. From small-scale habitat loopholes to decadal cycles: a habitat-based hypothesis explaining fluctuation in pelagic fish populations off Peru. Fish and Fisheries, 5 (4): 296–316. Binet, D. 1997. Climate and pelagic fisheries in the Canary and Guinea currents 1964-1993: the role of trade winds and the southern oscillation. Oceanol. Acta, 20:177-190. Binet, D., B. Samb, M.T. Sidi, J.J. Levenez and J. Servain. 1998. Sardine and other pelagic fisheries changes associated with multi-year trade wind increases in the southern Canary current. In Global versus local changes in upwelling systems, Durand, M.H., Cury, P. Mendelssohn, R. Roy, C. Bakun, A. and Pauly, D. eds., ORSTOM Editions, Paris, pp. 211-233. Bograd, S.J. and R.J. Lynn. 2003. Long-term variability in the Southern California Current System. Deep Sea Research Part II: Topical Studies in Oceanography, 50(14-16): 2355-2370. Borges, M.F., A.M.P. Santos, N. Crato, H. Mendes and B. Mota. 2003. Sardine regime shifts off Portugal: a time series analysis of catches and wind conditions. Sci. Mar., 67(Suppl.1): 235-244. Bouchón, M., S. Cahuín, E. Díaz, and M. Ñiquen. 2000. Captura y esfuerzo pesquero de la pesquería de anchoveta peruana (Engraulis rigens). Boletín del Instituto del Mar del Perú, 19:109-115. Brander, K., G. Blom, M.F. Borges,K. Erzini, G. Henderson, B.R. MacKenzie, H. Mendes, J. Ribeiro, A.M.P. Santos and R.Toresen. 2003. Changes in fish distribution in the eastern North Atlantic; are we seeing a coherent response to changing temperature?. ICES Marine Science Symposia, 219:261-270. Brêthes J.C. 1979. Contribution à l’étude des populations de Macrorhamphosus scolopax (L, 1759) et Macrorhamphosus gracilis (Lowe, 1839) des côtes atlantiques marocaines. Bulletin de l’Institut des Pêches Maritimes. N°24 (juillet 1979).
Modelling forecasting and scenarios in upwelling LMEs
215
Brink, K.H. and T.J. Cowles. 1991. The Coastal Transition Zone Program Journal Of Geophysical Research-Oceans, 96 (C8): 14637-14647. Caddy, J.F., ed. 1983. Advances in assessment of world cephalopod resources. FAO technical paper 231: 452 p. Caddy, J.F. and P.G. Rodhouse. 1998. Cephalopod and groudfish landings: evidence for ecological change in global fisheries. Rev. Fish Biol. Fish., 8:431-444. Carr, M.-E. 2002. Estimation of potential productivity in Eastern Boundary Currents using remote sensing. Deep-Sea Research II, 49(1-3): 59-80. Carrera P. and C. Porteiro. 2003. Stock dynamic of the Iberian sardine (Sardina pilchardus. Walb) and its implication on Galicia (NW spain). Sci. Mar., 67 (Suppl. 1): 245-258. Caverivière, A. 1991. Quelques considérations méthodologiques sur les campagnes scientifiques de chalutage de fond en Afrique Occidentale. ICES Statutory Meetings: Vol. C.M. 1991/D:10 Ref. Sess. U. Caverivière A. and H. Demarcq. 2002. Indices d'abondance du poulpe commun et intensité de l'upwelling au Sénégal. In Le poulpe Octopus Vulgaris, Sénégal et côtes nord-ouest africaines. Caverivière, A. Thiam, M. Jouffre D. Eds. Coll. Colloques et séminaires, 143-156. Paris, 2002. Cayan, D.R. 1992. Latent and Sensible Heat-Flux Anomalies over the Northern Oceans - Driving the SeaSurface Temperature. J. Phys. Oceanogr., 22: 859-881. Cendrero, O. 2002. Sardine and anchovy crises in northern Spain: natural variations or an effect of human activities?. ICES Marine Science Symposia, 215: 279-285. Chavez, F.P. 2004. Biological consequences of interannual to multidecadal variability. In The Global Coastal Ocean, Multiscale Interdisciplinary Processes. In The Sea, Vol 13. Robinson, A.R. and K.H.Brink, eds. Chapter 16. Harvard University Press. 1018pp. (In press). Chavez, F.P., J.T. Pennington, R. Michisaki and J.P. Ryan. 2004. The Monterey Bay Ocean Time-Series and Observatory (MOTO) Sheds Light on Multi-Decadal Basin-Scale Fluctuations of Anchovies and Sardines, MBNMS Ecosystem Observations, 20-21. Chavez, F.P., J. Ryan, S.E. Lluch-Cota. and M. Niquen. 2003. From anchovies to sardines and back: multidecadal change in the Pacifc Ocean. Science. 299: 217-221. Checkley, D.M., R.C. Dotson and D.A. Griffith. 2000. Continuous, underway sampling of eggs of Pacific sardine (Sardinops sagax) and northern anchovy (Engraulis mordax) in spring 1996 and 1997 off southern and central California. Deep-Sea Research II, 47: 1139–1155. Checkley, D.M., P.B. Ortner, L.R. Settle and S.R. Cummings. 1997. A continuous, underway fish egg sampler. Fisheries Oceanography, 6:58-73. Chen, D, M.A. Cane, A. Kaplan, S.E. Zebiak and D. Huang. 2004. Predictability of El Niño over the past 148 years. Nature 428: 733-736. CMS. 1999. Rapport du conseil scientifique de la convention sur la conservation des espèces migratrices appartenant à la faune sauvage sur les travaux de sa neuvième reunion. UNEP/CMS/ScC9/Doc. 11, 11 novembre 1999: 31 p. Codispoti, L.A., R.C. Dugdale and H.J. Minas. 1982. A comparison of the nutrient regimes off Northwest Africa, Peru, and Baja California. Rapp. P.-V. Reun. Cons. Int. Explo. Mer, 180: 184–201. Collins, C.A., L.M. Ivanov, O.V. Melnichenko and N. Garfield. 2004. California Undercurrent variability and eddy transport estimated from RAFOS float observations. Journal of Geophysical ResearchOceans 109 (C5): art. no. C05028. doi:10,1029/2003JC002191. Conser, R.J., K.T. Hill, P.R. Crone, N.C.H. Lo and D. Bergen. 2002. Stock assessment of Pacific sardine with management recommendations for 2002. Executive summary submitted to Pacific Fishery Management Council, Portland, OR, 11pp. Conser, R.J., K.T. Hill, P.R. Crone, N.C.H. Lo, and D. Bergen. 2004. Stock assessment of Pacific sardine with management recommendations for 2004. Executive summary submitted to Pacific Fishery Management Council, Portland, OR, 17pp. http://swfsc.nmfs.noaa.gov/frd/Coastal%20Pelagics/Sardine/03Full%20Paper3.pdf). Cury, P. and C. Roy. 1988. Migration saisonnière du thiof (Epinephelus aeneus) au Sénégal : influence des upwellings sénégalais et mauritanien. Oceanologica Acta, 11(1): 25-36. Cury, P. and C. Roy. 1989. Optimal environmental window and pelagic fish recruitment success in upwelling areas. Can. J. Fish. Aquat. Sci., 46(4): 670-680. Cury P., Shannon L. and Shin Y-J 2001. The Functioning Of Marine Ecosystems. Reykjavik Conference on Responsible Fisheries in the Marine Ecosystem 7. Reykjavik, Iceland, 1-4 October 2001.
216
P. Fréon et al.
Cushing, D.H. 1969. Upwelling and fish production. FAO Fisheries Technical Paper No. 84 (Distribution restricted; http://www.fao.org/docrep/89204E/89204e00.htm#Contents): 1-40. Cushing, D.H. 1971. Upwelling and the production of fish. Adv. Mar. Biol. 9: 255-334 DeAngelis, D.L. and R.M. Cushman. 1990. Potential application of models in forecasting the effects of climate changes on fisheries. Transactions of the American Fisheries Society, 119:224-239. Demarcq, H. and V. Faure. 2000. Coastal upwelling and associated retention indices from satellite SST. Application to Octopus vulgaris recruitment. Oceanol. Acta, 23: 391-408. deYoung, B., R. Harris, J. Alheit, G. Beaugrand, N. Mantua, L. Shannon. 2004. Detecting regime shifts in the ocean: Data considerations. Progress in Oceanography 60(2-4): 143-164. Di Lorenzo, E., A.J. Miller, N. Schneider and J.C. McWilliams. 2005. The warming of the California Current: Dynamics and ecosystem implications. Journal of Physical Oceanography, in press. Diffenbaugh, N.S., M.A. Snyder and L.C. Sloan. 2004. Could CO2-induced land-cover feedbacks alter near-shore upwelling regimes? Proc. Natl. Acad. Sci. USA, 101: 27–32. Do Chi, T. 1994. Groupe de travail ad hoc sur les sardinelles et autres espèces de petits pélagiques côtiers de la zone nord du COPACE. COPACE/PACE Series 91/58: 295 p. Drinkwater, K.F. and R.A. Myers. 1987. Testing predictions of marine fish and shellfish landings from environmental variables. Canadian Journal of Fisheries and Aquatic Sciences, 44: 1568-1573. Enfield, D. B., and D. A. Mayer. 1997. Tropical Atlantic sea surface temperature variability and its relation to El Nino–Southern Oscillation. J. Geophys. Res., 102: 929–945. FAO. 1986. Report of the first ad hoc working group on seabream (sparidae) stocks in the northern cecaf zone. CECAF/ECAF SERIES 86/38. Fréon, P. 1986. Réponses et adaptations des stocks de Clupéidés d'Afrique de l'Ouest à la variabilité du milieu et de l'exploitation : analyse et réflexion à partir de l'exemple du Sénégal. Thèse doctorat d'Etat es sciences. Université d'Aix-Marseille II, 287 p. Fréon, P. 1989. Introduction of climatic variables into global production models. In International Symposium on Long Term Changes in Marine Fish Populations (pp. 481-528). Larrańeta, M.G.; Wyatt, T. Eds. Consejo Superior de Investigaciones Cientificas, Spain. Fréon, P. and B. Stéquert. 1976. Note sur la présence de Sardina pilchardus (Walb.) au Sénégal : étude de la biométrie et interprétation. Cybium, 6: 65-90. Fréon, P. and O.A. Misund. 1999. Dynamics of pelagic fish distribution and behaviour: effect on fisheries and stock assessment. London, Blackwell, Fishing News Book. Fréon, P., and E.R. Yáñez. 1995. Influencia del medio ambiente en evaluación de stock: una aproximación con modelos globales de producción. Investigaciones Marinas., 23: 25-47. Fréon, P., C. Mullon and G. Pichon. 1993. CLIMPROD: Experimental interactive software for choosing and fitting surplus production models including environmental variables. FAO Computerized Information Series Fisheries, 5: 76 pp. Fréon, P., C. Mullon and B.Voisin. 2003. Investigating remote synchronous patterns in fisheries Fish. Oceanogr., 12(4):1–15. Fréon, P., P. Cury, , L. Shannon and C. Roy. 2005. Sustainable exploitation of small pelagic fish stocks challenged by environmental and ecosystem changes: a review. Fishery Bulletin, 76(2): 385-462. Gulland, J.A. and S. Garcia. 1984. Observed patterns in multispecies fisheries. In Exploitation of marine communities. R.M. May, ed., Dahlem Konferenzen. Berlin, Heidelberg, New-York, Tokyo, SpringlerVerlag, pp. 155-190. Haidvogel, D.B., A. Beckmann and K.S. Hedstrom. 1991. Dynamical Simulations of filament formation and evolution in the coastal transition Zone. J. Geophys. Res. 96:15017-15040. Haney, R.L., R.A. Hale and D.E. Dietrich. 2001. Offshore propagation of eddy kinetic energy in the California Current. J. Geophys. Res., 106: 11709-11718. Herbland, A. and B.Voitiurez. 1974. La production primaire dans l’upwelling mauritanien enmars 1973. Cah. ORSTOM, sér. Océanogr., 12: 187-201. Hickey, B.M. 1998. Coastal oceanography of western North America from the tip of Baja California to Vancouver Island. In The Sea, 11 Robinson, A.R. and K.H. Brink, eds. 345-393. Holzlohner, S. 1975. On the recent stock developement of Sardina pilchardus Walb. Off Spanish Sahara. ICES CM 1975/J:13. ICES. 2002a. Report of the Working Group on Ecosystem Effects of Fishing Activities (WGECO). ICES CM 2002/ ACE: 03. ICES. 2002b. Report of the Working Group on the Assessment of Mackerel, Horse Mackerel, Sardine and Anchovy. ICES CM 2002/ACFM, 6: 474 pp.
Modelling forecasting and scenarios in upwelling LMEs
217
Ikeda, M., L.A. Mysak and W.J. Emery. 1984. Observation and modelling of satellite-sensed meanders and eddies of Vancouver Island. J. Phys. Oceanogr. 14: 3-21. Jarre-Teichmann, A. 1998. The potential role of mass balance models for the management of upwelling ecosystems. Ecological Application 8:93-103. Jiménez Pérez, L.C., B.E. Lavaniegos, and T.R. Baumgartner. 2000. Biomasa y estructura del zooplancton en la parte sur de la Corriente de Califonia durante Julio de 1998. Crucero IMECOCAL 9807. B/O Francisco de Ulloa. 15-30 de Julio de 1998. Informe Tecnico. Comunicaciones Académicas, Serie Ecologia CICESE 41: 40 p. Johnson, J.A., I.G. Stevens. 2000. A fine resolution model of the eastern North Atlantic between the Azores, the Canary Islands and the Gibraltar Strait. Deep-Sea Research I, 47: 875-899. Kawasaki, T. 1983. Why do some pelagic fishes have wide fluctuations in their numbers? Biological basis of fluctuations from the viewpoint of evolutionary ecology. FAO Fish. Rep. 293(3): 1065-1080. Kelly, K.A., R.C. Beardsley,R. Limeburner, K.H. Brink, J.D. Paduan, K.T. Chereskin. 1998. Variability of the near-surface eddy kinetic energy in the California Current based on altimeter, drifter, and moored current data. J. Geophys. Res., 103: 13067-13083. Kifani S. 1998. Climate dependant fluctuatiosn of the Moroccan sardine and their impact on fisheries. In Global versus local changes in upwelling systems. Durand M.H., P. Cury, R. Mendelssohn, C. Roy, A. Bakun and D. Pauly, eds. Editions ORSTOM Paris, pp.235-248. Klyashtorin, L.B. 1998. Long-term climate change and main commercial fish production in the Atlantic and Pacific. Fish. Res. 37: 115-125. Klyashtorin, L.B. 2001. Climate change and long term fluctuations of commercial catches: the possibility of forecasting. FAO Fish. Tech. Pap. 410: 86 p. Laurans M., D. Gascuel, A. Caverivière. 2002. Application d'un modèle global avec effet de l’environnement au stock de poulpe du Sénégal. In Le poulpe commun Octopus vulgaris. Sénégal et côtes nord-ouest africaines. Caverivière A., M. Thiam M., D. Jouffre, éds. Editions IRD, Paris, Colloques et séminaires, pp. 255-267. Li, Z., Y. Chao. and J.C. McWilliams. (in prep.) A three-dimensional variational data assimilation scheme for the Regional Ocean Modeling System: I, Theoretical Formulation. MacCall, A. 1990. Dynamic Geography of Marine Fish Populations. University of Washington Press, Seattle, 153 p. Mackas, D. Interdisciplinary Oceanography of the Western North American Continental Margin: Vancouver Island to the tip of Baja California. In The Sea. Robinson, A.R.and K. Brink,eds. Harvard University Press, Cambridge, MA, in press. Mackas, D.L., L. Washburn and S.L. Smith. 1991. Zooplankton community pattern associated with a California Current cold filament. J. Geophys. Res., 96: 14781-14797. Mackas, D., P.T. Strub, A.Thomas and V. Montecino. Eastern ocean boundaries pan-regional overview. In The Sea. Robinson, A.R.and Brink, K., eds. Harvard University Press, Cambridge, MA, in press. 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. Meteorol. Soc., 78: 1069-1079. Marchesiello, P., J.C. McWilliams and A. Shchepetkin. 2003. Equilibrium structure and dynamics of the California Current System. J. Phys. Oceanogr. 33: 753-783. Marchesiello, P., S. Herbette, L. Nykjaer and C. Roy. 2004. Eddy-driven dispersion processes in the Canary Current upwelling system: comparison with the California system. GLOBEC International Newsletter, 10(1): 5-7. Mathiesen, O.A. 1989. Adaptation of the anchoveta (Engraulis ringens) to the Peruvian upwelling system. In The Peruvian upwelling ecosystem: dynamics and interactions, Pauly, D., P. Muck, J. Mendo and I. Tsukayama, eds. ICLARM Studies and Reviews, Manilia, 18: 220-234. McCreary, J.P., and P.K. Kundu. 1985. Western boundary circulation driven by an alongshore wind: with application to the Somali Current system. J. Mar. Res., 43: 493-516. McCreary, J.P., Y. Fukamachi and P.K. Kundu. 1991. A numerical investigation of jets and eddies near an eastern ocean boundary. J. Geophys. Res. 96:2515-2534. McGowan, J.A, S.J. Bograd, R.J. Lynn and A.J. Miller. 2003. The biological response to the 1977 regime shift in the California Current. Deep Sea Research Part II: Topical Studies in Oceanography, v. 50(1416): 2567-2582. Mendelssohn, R. and F.B. Schwing. 2002. Common and uncommon trends in SST and wind stress in the California and Peru-Chile current systems. Prog. Oceanogr. 53:141–162.
218
P. Fréon et al.
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 Heat-Flux and Wind Stress Anomalies. Clim. Dyn. 9:287-302. Minas, H.J., L.A. Codispoti and R.C. Dugdale. 1982. Nutrients and primary production in the upwelling region off Northwest Africa. Rapp. Proc. Reun. Cons. Int. Expl. Mer, 180: 148-183. Mittelstaedt, E. 1991. The ocean boundary along the Northwest African coast: circulation and oceanographic properties at the sea surface. Progress in Oceanography 26: 307-355. Montecino, V., P.T. Strub, F. Chavez, A.Thomas, J. Tarazona and T. Baumgartner. Bio-physical interactions off Western South-America. In The Sea. Robinson, A.R. and K. Brink, eds. Harvard University Press, Cambridge, MA. In press. Montecinos, A., S. Purca and O. Pizarro. 2003. Interannual-to-interdecadal sea surface temperature variability along the western coast of South America, Geophys. Res. Lett. 30 (11): 1570, doi:10.1029/2003GL017345. Mooers, C.N.K. and A.R. Robinson. 1984. Turbulent jets and eddies in the California Current and inferred cross-shore transports. Science 223:51-53. Mote, P.W. and N.J. Mantua. 2002. Coastal upwelling in a warmer future. Geophys. Res. Lett., 29: 2138– 2141. Narimousa, S., and T. Maxworthy. 1989. Application of a laboratory model to the interpretation of satellite and field observations of coastal upwelling. Dyn. Atmos. Oceans, 13:1-46. Nieto, K., E. Yáńez, C. Silva and M.A. Barbieri. 2001. Probable fishing grounds for anchovy in the northern Chile using an expert system. In Proceeding of the IGARSS 2001: International Geoscience and Remote Sensing Symposium, Sidney, Australia, 9-13 July 2001. Ñiquen, M. and M. Bouchon. 2004. Impact of El Niño events on pelagic fisheries in Peruvian waters. Deep-Sea Research., II, 51: 563–574. Ñiquen, M., M. Bouchón, S. Cahuín and E. Díaz. 2000. Pesquería de anchoveta del mar Peruano 1950 1999. Boletín del Instituto del Mar del Perú, 19:117-123. Pares-Sierra, A., W.B. White and. C.-K. Tai. 1993. Wind driven coastal generation of annual mesoscale eddy activity in the California Current system: A numerical model. J. Phys. Oceanogr. 23:1110-1121. Parrish, R.H. and A.D. MacCall. 1978. Climatic variation and exploitation in the Pacific mackerel fishery. Calif. Dep. Fish. Game Fish. Bull. 167: 109p. Parrish, R.H., C.S. Nelson, and A. Bakun, A. 1981. Transport mechanisms and reproductive success of fishes in the California Current. Biol. Oceanogr. 1:175-203. Pauly, D., V. Christensen, J. Dalsgaard, R. Froese and F. Torres Jr. 1998. Fishing down marine food webs. Science., 279 (6 February 1998):860-863. Pelegrí, J.L. J. Arístegui, L. Cana, M. González-Dávila, A. Hernandez-Guerra, S.Hernández-León, A. Marrero-Díaz,M.F. Montero, P. Sangrá and M. Santana-Casiano. 2005. Coupling between the open ocean and the coastal upwelling region off northwest Africa: water recirculation and offshore pumping of organic matter. Journal of Marine Systems 54(1-4): 3-37. Peliz, A., J. Dubert, D. Haidvogel. 2003a. Subinertial response of a density driven eastern boundary poleward current to wind forcing. Journal of Physical Oceanography, 33(8): 1633–1650. Peliz, A., J. Dubert, D. Haidvogel, B. Le Cann. 2003b. Generation and unstable evolution of a densitydriven Eastern Poleward Current: the Iberia poleward current. Journal of Geophysical Research. 108 (C8), 3268 (doi:10.1029/2002JC001443). Penven, P.,J. Pasapera, J. Tam and C. Roy. 2003. Modeling the Peru Upwelling System seasonal dynamics Mean circulation, seasonal cycle and mesoscale dynamics, GLOBEC International Newsletter 9:23-25. Prairie, Y.T. 1996. Evaluating the predictive power of regression models. Canadian Journal of Fisheries and Aquatic Sciences 53:490-492. Quero, J-C., M-H. Du Buit and J.-J. Vayne. 1998. Les observations de poissons tropicaux et le réchauffement des eaux dans l'Atlantique européen. Oceanol. Acta. 21:345-351. Roemmich, D. 1992. Ocean Warming and Sea Level Rise Along the Southwest U.S. Coast. Science, 257 (5068): 373-375. Roemmich, D. and J. McGowan. 1995: Climatic Warming and the Decline of Zooplankton in the California Current. Science 267(5202): 1324-1326. Roy, C. and P. Cury. 2003. Decadal environmental and ecological changes in the Canary Current Large Marine Ecosystem and adjacent waters: Patterns of connections and teleconnection. In Large Marine
Modelling forecasting and scenarios in upwelling LMEs
219
Ecosystems of the World - Trends in Exploitation, Protection and Research, Sherman, K. and Hempel, G. Eds.: 255-277. Roy, C. and C. Reason,. 2001. ENSO related modulation of coastal upwelling in the eastern Atlantic. Progress in Oceanography 49:245-255. Sætersdal, G., G. Bianchi, T. Strømme and S.C. Venema. 1999. The DR. FRIDTJOF NANSEN Programme 1975-1993. Investigations of fishery resources in developing countries. History of the programme and review of results. FAO Fisheries Technical Paper. 391: 434p. Santos, A.M.P., M.F. Borges, and S.B. Groom. 2001. Sardine and horse mackerel recruitment and upweling off Portugal. ICES Journal of Marine Science 58:589-596. Santos A.M.P., A. Peliz, J. Dubert, P.B. Oliveira, M.M. Angélico and P. Ré. 2004. Impact of a winter upwelling event on the distribution and transport of sardine (Sardina pilchardus) eggs and larvae off western Ibéria: A retention mechanism, Cont. Shelf Res. 24:165-194. Sauvaget, P. E. David and C. Guedes Soares. 2000. Modelling tidal currents on the coast of Portugal. Coastal Engineering 40:393–409. Schwartzlose, R.A., J. Alheit, A. Bakun, T.R. Baumgartner, R. Cloete, R., R.J.M. Crawford, W.J. Fletcher, Y. Green-Ruiz, E. Hagen, T. Kawasaki, D. Lluch-Belda, S.E. Lluch-Cota, A.D. MacCall, , Y. Matsuura, M.O. Nevarez-Martinez, R.H. Parrish, C. Roy, R. Serra, K.V. Shust , M.N. Ward and J.Z. Zuzunaga. 1999. Worldwide large-scale fluctuations of sardine and anchovy populations. S. Afr. J. Mar. Sci. 21:289-347. Schwing, F.B. and R. Mendelssohn. 1997. Increased coastal upwelling in the California Current System. Journal of Geophysical Research, 102:3421–3438. Silva, C., E. Yáńez, M.A. Barbieri, K. Nieto, V. Mimica, F. Espindola and J.Acevedo. 2000. Exploring the association between small pelagic fisheries and SeaWiFS chlorophyll and AVHRR sea surface temperature in the north of Chile. In Proceeding of the Sixth International Conference on Remote Sensing for Marine and Coastal Environments, Charleston, South Carolina, 1-3 May 2000. ERIM International, Vol. II: 81-88. Sinclair M., M.J. Tremblay and P. Bernal, 1985: El Nino events and variability in a Pacific Mackerel (Scomber japonicus) Survival Index: support for Hjort’s second hypothesis. Can. J. Fish. Aquat. Sci. 42:602-608. Slagstad, D. and P. Wassmann. 2001. Modelling the 3-D carbon flux across the Iberian margin during the upwelling season in 1998. Progress in Oceanography. 51:467–497. Smith, P. and H.G. Moser. 2003. Long-term trends and variability in the larvae of Pacific sardine and associated fish species of the California Current region. Deep-Sea Research-II, 50: 2519-2536. Smith, T.M. and R.W. Reynolds. 2004. Improved Extended Reconstruction of SST (1854-1997). Journal of Climate. 17:2466-2477. Snyder, M.A., L.C. Sloan, N.S. Diffenbaugh, and J.L. Bell. 2003. Future climate change and upwelling in the California Current. Geophys. Res. Lett. 30:1823-1826. Stergiou, K.I., E.D. Christou and G. Petrakis. 1997. Modelling and forecasting monthly fisheries catches: comparison of regression, univariate and multivariate time series methods. Fisheries Research 29:5595. Strub, P.T. and C. James. 2002. Altimeter-derived surface circulation in the large-scale NE Pacific Gyres. Part 1. seasonal variability. Progress In Oceanography 53 (2-4): 163-183. Strub, P.T., J.M. Mesias,V. Montecino, J. Rutllant and S. Salinas. 1998. Coastal ocean circulation off western South America. 273-313 In Robinson, A.R. and K.A Brink, eds. The Sea. Harvard University Press, Cambridge, MA: Sydeman, W.J., M.M. Hester, J.A. Thayer, 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. Progress In. Oceanography. 49:309-329. Ulloa, O., R. Escribano, S. Hormazabal, R.A. Quiñones, R.R. González and M. Ramos. 2001. Evolution and biological effects of the 1997-98 El Niño in the upwelling ecosystem off northern Chile. Geophysical Research Letters. 28:1591-1594. Ulltang, O. 1996. Stock assessment and biological knowledge: can prediction uncertainty be reduced? ICES J. Mar. Sci., 53: 659-675. Veit, R.R., J.A., McGowan, D.G. Ainley, T.R. Wahls, and P. Pyle. 1997. Apex marine predator declines ninety percent in association with changing oceanic climate. Global Change Biol. 3:23-28. Verity, P.G., V. Smetacek and T.J. Smayda. 2002. Status, trends and the future of the marine pelagic ecosystem. Environmental Conservation. 29:2077-237.
220
P. Fréon et al.
Ware, D.M. and G.A. McFarlane. 1989. Fisheries production domains in the northeast Pacific Ocean. In: Effects of ocean variability and an evaluation of parameters used in stock assessment models. Beamish, R.J. and G.A. McFarlane, eds. Can. Spec. Publ. Fish. Aquat. Sci. 108:359-379. Wooster, W.S., A. Bakun and D.R. McLain. 1976. The seasonal upwelling cycle along the eastern boundary of the North Atlantic. J. Mar. Res. 34:131-141. Yáñez, E., M.A. Barbieri, C. Silva, K. Nieto and F. Espíndola. 2001. Climate variability and pelagic fisheries in northern Chile. Progress in Oceanography 49:581-596.
Part III: Hopes, Dreams and Reality Forecasting in the Benguela: Our Collective Wisdom
ThisPage pageis intentionally This Intentionally left Leftblank Blank
Large Marine Ecosystems, Vol. 14 V. Shannon, G. Hempel, P. Malanotte-Rizzoli, C. Moloney and J. Woods (Editors) © 2006 Elsevier B.V. All rights reserved.
223
10 Influences of Large Scale Climate Modes and Agulhas System Variability on the BCLME Region C.J.C. Reason, P. Florenchie, M. Rouault, J. Veitch
INTRODUCTION In this chapter we review the external forcing of the BCLME region, for example, via large-scale modes such as the Benguela Niño or the El Niño Southern Oscillation (ENSO) or via large-scale anomalies in the regional atmospheric circulation. Unlike other subtropical eastern boundary currents, the Benguela Current is bounded by warm upper ocean conditions at both its poleward and equatorward margins and may be impacted by anomalous oceanic conditions both to the north and to the south. The termination of Africa at relative low latitudes not only leads to an ocean connection to the south via filaments and rings shed off the Agulhas Current retroflection zone, but also allows rapid communication of atmospheric signals to the BCLME from the midlatitudes of the South Atlantic, since the lack of landmass to the south means that atmospheric blocking events are short-lived by comparison to those in the Northern Hemisphere. Atmospheric variability emanating over the Indian and Pacific Oceans may also be communicated to the Benguela, either directly via atmospheric Rossby waves (e.g., Cook 2001) or easterly waves (e.g., Reason and Keibel 2004) impacting on the winds over the BCLME, or indirectly, via their impact on the Agulhas Current. The latter is the strongest western boundary current in the Southern Hemisphere and exhibits pronounced meander and eddy activity in its southern regions which may then interact with the southern Benguela (Boebel et al. 2003).
ATMOSPHERIC VARIABILITY OF THE BCLME REGION Shillington et al. (2006) described the annual cycle of the winds over the BCLME region together with the synoptic and mesoscale weather systems that modulate the typical upwelling-favourable winds along the southwestern coast of Africa. In this section, an overview of the variability on intraseasonal to interdecadal scales is given. Foltz and McPhaden (2004) showed evidence of 30-70 day oscillations in the winds over the tropical Atlantic (as far south as about 20oS) and connected these with Madden–Julian oscillations originating in the Indian Ocean. Although the strongest signals were in the central tropics of the basin, their data suggests that there may also
224
C.J.C. Reason et al.
be an impact on the northern part of the BCLME region. Indeed, analysis of QuikSCAT data for the 1999/2000 period by Risien et al. (2004) also found evidence of significant intraseasonal variability of the winds in this region, although the timing of this variability led these authors to suggest linkages with the pulsing of the convection over the Congo basin and in the West African monsoon. In the eastern tropical Atlantic Ocean, a semi-annual signal is apparent in the trade winds (Philander and Pacanowski 1986). Figure 10-1, taken from eight years of satellite scatterometer winds analysed by Veitch (2004), shows that the zonal wind stress over the equatorial South East Atlantic is strongest in May with a second peak in November and weakest in February with another weaker minimum in September. Veitch (2004) presented evidence from the OPA-TOTEM and CLIPPER ATL6 models that this semi-annual variation in the local winds may influence the currents and thermocline depth in this region. It is not known how this local wind variation might impact on the northern BCLME region or if it interacts with the manifestation of the Benguela Niño.
Figure 10-1. Annual cycle of the zonal wind stress over the equatorial South East Atlantic derived from ERS winds for 1992-1999 (after Veitch 2004).
On the interannual scale, the dominant mode globally is ENSO and it has been shown that this mode impacts on both the atmospheric circulation and the SST of the South Atlantic region via the so-called Pacific South America (PSA) pattern (Mo and Paegle 2001; Colberg et al. 2004). The PSA involves atmospheric Rossby wave propagation from the tropical western Pacific to the southeast and into the mid- to high latitude South Atlantic where it modulates the regional winds and hence SST during the mature phase of ENSO as well as the preceding austral spring and following autumn. Colberg et al. (2004) analysed output from a global model forced by 50 years of National Center for Environmental Prediction (NCEP) re-analyses to show that the SST anomalies induced in the South Atlantic mainly evolve via changes to the surface heat flux exchange but with sizeable contributions in certain areas from modulations to the meridional Ekman heat transport and Ekman pumping. A lag of one season between
Influences of large scale climate modes and Agulhas variability on the BCLME
225
the SST response and the modulations to the surface heat fluxes was observed. In the southeastern Atlantic, the largest ENSO signal was observed to extend northwestwards from the central and southern Benguela region. Thus, although an EOF or SVD analysis of South Atlantic SST may not produce the ENSO-induced patterns as the leading mode, the work of Colberg et al. (2004) shows that there is a robust and significant ENSO signal over large areas of the South Atlantic, particularly the southern and central BCLME. A recent example concerns the strong influence of the 1999/2000 La Niña event on the southern Benguela (Roy et al. 2001). Central to the ENSO-induced variability of the South Atlantic are modulations to the trade winds and midlatitude westerlies of the basin. Sustained weakening or strengthening of the trade winds over the tropical central to western South Atlantic during the late austral spring or early summer may lead to Benguela Niños or Niñas respectively, perhaps the most well known of the modes of variability over the BCLME region (Shannon et al. 1986; Florenchie et al. 2003). Smaller warm or cold events may be related to less intense weakenings or strengthenings of the trades (Florenchie et al. 2004). Although the oceanic response to these wind modulations and the subsequent evolution of substantial SST anomalies along the Angolan and northern Namibian coast has been studied by these and other authors, little is known about what causes the wind changes themselves. The latter is important if the apparent predictability of Benguela Niños is to be extended beyond the roughly 1-2 month lag between these trade wind fluctuations and the SST anomalies. Given that these wind anomalies are located in the tropics, they may arise through a variety of mechanisms such as equatorial Kelvin and Rossby modes associated with ENSO or other tropical SST modes, Madden-Julian oscillations (e.g., Foltz and McPhaden 2004), or links with convection over the Amazon, Congo or West African monsoon regions for example. In addition to Benguela Niños and the impact of ENSO on the South Atlantic, interannual dipole-like patterns in South Atlantic SST often occur during the austral summer (Fauchereau et al. 2003; Hermes and Reason 2005). These patterns involve an anomaly of one sign in the northeast (i.e., in, and extending offshore from, the central and northern Bengeula region) and the opposite sign in the midlatitude southwest Atlantic and appear to be related to modulations of the subtropical anticyclone and the circumpolar trough to the south. For some events, there is a similar dipole-like SST pattern in the South Indian Ocean as well but its evolution lags that in the Atlantic by one or two months (Hermes and Reason 2005). In these cases, there is a hemispheric modulation of the wave number 3 or 4 pattern in the Southern Hemisphere superimposed on an Antarctic Oscillation type pattern (Kidson 1988) such that when there are cool (warm) SST anomalies in the southwest (northeast) of the subtropics / midlatitudes of each basin, high (low) pressure anomalies are evident over Antarctica (the midlatitudes) (Hermes and Reason 2005). For the reverse SST dipole anomaly pattern, the wave 4 high pressure anomalies in the midlatitudes are pronounced but the Antarctic Oscillation pattern is less evident. The results of Fauchereau et al. (2003) and Hermes and Reason (2005) indicate that there is often a linkage between variability in the South Atlantic and elsewhere in the subtropics / midlatitudes of the Southern Hemisphere achieved via coherent atmospheric forcing on a hemispheric scale.
226
C.J.C. Reason et al.
High latitude forcing The South Atlantic is open to the Antarctic region to the south and therefore it may be impacted by modes of variability originating in the high latitudes. In terms of the atmosphere, the leading mode in the circulation south of 20oS is the Antarctic Oscillation (AAO), Southern Annular mode, or ‘high latitude’ mode (Kidson 1988). This mode is characterized by pressure anomalies of one sign centered in the Antarctic and anomalies of the opposite sign centred at about 40-50oS but in recent decades there has been a trend towards positive polarity in which high pressure anomalies occur in the midlatitudes with low pressure anomalies over Antarctica. Although the dominant signal is zonally symmetric, there is also some signature of wave number 2 or 3 superimposed upon it. The AAO is apparent throughout the year, but tends to be more active in austral spring (Thompson and Wallace 2000). Using observational, model and NCEP re-analysis data, Reason et al. (2002) and Reason and Rouault (2005) showed that the AAO modulates the surface heat flux exchange and atmospheric circulation over the midlatitude South Atlantic region with associated changes in winter rainfall over western South Africa. Given that these studies show sizeable wind anomalies over the southeastern Atlantic, there may well be an AAO influence on the Benguela system. In addition, analysis of a long coupled model run by Hall and Visbeck (2002) indicates a significant proportion of the ocean variability south of 30oS is due to the AAO and that this mode both drives changes in poleward heat transport and modulates the currents and wind forcing throughout the mid- to high latitudes of the Southern Hemisphere. As a result, it is likely that the AAO may influence both the Agulhas and Benguela Currents – however, detecting such an influence will be difficult given the large variability of the region. It should also be mentioned that observations taken over 1948-2003 at Marion Island, located some 1500 km to the southeast of South Africa, suggest that the AAO may influence the rainfall, temperature and local circulation there (Rouault et al. 2005). White and Peterson (1996) presented evidence for an interannual wave 2 anomaly pattern in SST, winds, sea level pressure and sea-ice extent that is advected by the Antarctic Circumpolar Current around the Southern Hemisphere and which they termed the Antarctic Circumpolar Wave (ACW). The ACW was suggested to project substantially into the subtropics of each of the three ocean basins. More recently, Simmonds (2003) has suggested that the ACW is difficult to detect outside the 19801995 period, although there are concerns about the quality of the data prior to 1980. This work suggested that stationary or westward (i.e. opposite to the ACW) propagating anomalies may occur during periods before 1980. To date, no evidence has been presented of a direct influence of the ACW on the BCLME region although White and Peterson’s (1996) work did show an impact of this mode on the broader subtropical South Atlantic. However, as for the AAO, detecting an ACW signal in the BCLME region will be difficult. A prominent aspect of Southern Hemisphere climate in the 35-65oS zone is the semiannual oscillation (SAO) in sea level pressure, winds and rainfall (van Loon 1967). For example, George on the south coast of South Africa at 34oS receives rainfall all year round but it has weak peaks in April and October. These semi-annual peaks are
Influences of large scale climate modes and Agulhas variability on the BCLME
227
more pronounced at stations located further south in the Southern Hemisphere (e.g., Melbourne, Australia). As such, the SAO may be expected to influence the winds over the southern Agulhas and Benguela regions and hence potentially these currents as well as the upwelling along the southwestern Cape coast. For example, Reason et al. (2003) and Hermes et al. (2005) show that there is a semi-annual signal in the volume transport of the Agulhas Current through 20oE south of South Africa. A wavelet analysis of the NCEP re-analysis meridional wind averaged west and southwest of the South African coast (32-37oS, 15-18oE) shows substantial power at the semi-annual time scale, particularly during the 1950-1985 period. This result points to the decadal variability previously found in the SAO (van Loon et al. 1993; Meehl et al. 1998). In addition, the phasing of the SAO appears to have changed since 1980 so that its peak is later in autumn; this phase change has been suggested by Rouault et al. (2005) to help account for the changes observed in the summer climate of Marion Island in recent decades. Decadal scale variability An EOF analysis of 1948-2003 NCEP re-analysis mean sea level pressure data over the South Atlantic for the austral summer indicates that around 22 % of the variance is associated with a roughly decadal scale pulsing of the South Atlantic anticyclone. Using SVD analysis on SST and sea level pressure for a shorter period, Venegas et al. (1996) also found a quasi-decadal signal in the South Atlantic anticyclone which they determined to be the leading mode. On bi-decadal time scales, Reason (2000) used UKMO HadSLP data to find evidence of modulations of the South Atlantic anticyclone which tended to vary in phase with those in the subtropical anticyclone in the South Indian Ocean, and to lesser extent, the South East Pacific. Thus, on both decadal and interdecadal scales, variations in the strength and position of the South Atlantic anticyclone are important and likely to influence the surface winds and SST of the basin (Venegas et al. 1996; Reason 2000; Wainer and Venegas 2001) as well as the BCLME region itself. In addition to this variability, Hermes and Reason (2005) noted that there was a decadal pulsing of the dipole-like SST patterns discussed above such that the timing of the events in the South Atlantic and South Indian Oceans moved in and out of phase with each other on this time scale. Evidence exists that ENSO-like decadal modes influence the atmospheric circulation, SST and rainfall of the southern African region (Allan 2000; Reason and Rouault 2002; Allan et al. 2003). Work by Garreaud and Battisti (1999), Allan (2000), Allan et al. (2003) and others has shown that an ENSO-like pattern is the leading mode of nearglobal scale variability on a range of interannual, decadal and multidecadal scales. Given the evidence of its projection onto southern African rainfall and atmospheric circulation (Allan 2000; Reason and Rouault 2002), it seems likely that at least part of the decadal-multidecadal variability in SST and winds over the BCLME region may be related to these ENSO-like decadal modes. Other decadal-multidecadal signals in the region concern those in the SAO (van Loon et al. 1993; Meehl et al. 1998) and the tendency in recent decades for the AAO to
228
C.J.C. Reason et al.
assume positive polarity (i.e., negative height anomalies over Antarctica and positive over the midlatitudes) (Thompson and Wallace 2000). These lower frequency signals may influence the slowly varying environment in which higher frequency modes evolve and hence modulate the impact of the latter over the region. A quasi-decadal scale signal may also exist in the manifestation of the Benguela Niño in the northern BCLME region. Although warm and cold events occur quite frequently off the Angolan coast, the large events or so-called Benguela Niños and Niñas only evolve roughly every decade or so. The next section considers these events in more detail.
BENGUELA NIÑOS AND SST VARIABILITY IN THE TROPICAL EASTERN ATLANTIC OCEAN The Benguela Niño is the most prominent mode of low frequency variability in the South East Atlantic and hence it is described in some detail. Attention is also paid to its relationships with the so-called Atlantic Niño (Zebiak 1993) and ENSO. Using optimally interpolated SST (Reynolds and Smith 1994), Florenchie et al. (2004) showed that the maximum standard deviation in SST over the South Atlantic and equatorial North Atlantic occurs between 10 and 20oS and from 8°E to the African coast, termed the ABA (Angola/Benguela area). This area also shows the largest SST seasonal signal amplitude, but unlike the tropical Pacific Ocean, the range of interannual variability is smaller compared to the seasonal one (Servain and Arnault 1995). Within the ABA is the Angola-Benguela frontal zone (ABFZ), where warm and saline water from the equator flows poleward and meets the much cooler and nutrient-rich water from the northern Benguela upwelling system (Shannon et al. 1987, Meeuwis and Lutjeharms 1990). Meridional movements of the frontal zone could be responsible for the strong SST anomalies in the region (Veitch et al. 2006). Along and just south of the equator, a weaker maximum in variability spreads in the area of the cold tongue, roughly between 20°W and 5°W. The timing of the SST anomalies in the equatorial and Angolan coastal regions is considered in the final sub-section. The variability over the basin shows seasonality with pronounced maxima in March in the ABA (Florenchie et al. 2004) and in June/July along the equator (Carton and Huang 1994). Time series of NOAA extended SST (Smith and Reynolds 2004) averaged over the ABA area indicate the periodicity, magnitude and duration of anomalies over the last few decades (Figure 10-2). Extreme warm and persistent events develop in the ABA roughly every 10 years or so (Shannon et al. 1986; Walker 1987) and were termed Benguela Niños by Shannon et al. (1986) because they showed some similarities with the Pacific Niños. However such episodes are less frequent and less intense than their Pacific counterparts.
Influences of large scale climate modes and Agulhas variability on the BCLME
229
During Benguela Niños, the Angola-Benguela front tends to be displaced southward following an intrusion of warm and saline water of equatorial origin as far as 25°S (Mohrholz et al. 2001; Shannon et al. 1986). These anomalous events can induce significant rainfall anomalies (Rouault et al. 2003) and can drastically modify fish distribution and abundance (Boyer et al. 2001). The collapse of the Namibian sardine stock after 1974 followed a protracted warm event during 1972-74, the effect of which was aggravated by overfishing (Boyer and Hampton 2001). More recently, the 1995 event had a drastic impact on the whole ecosystem with a 4-5° southward shift of the sardine population, associated with high mortality and poor recruitment of the major pelagic fish species (Boyer and Hampton 2001). Benguela Niños occurred in 1934, 1949, 1963, 1984 (Shannon et al. 1986) and more recently in 1995 (Gammelsrød et al. 1998). The moderate but very persistent warm episode during 1972-74 seems also to be generally considered as a Benguela Niño. Being in phase with late summer (March/April), these events induce very high sea temperatures that affect the biota. By interacting with the atmosphere via modulating the surface evaporation and the moisture flux convergence in the Angolan low, high SST anomalies may lead to well above average late summer rainfall over western Angola and Namibia, sometimes with local flooding.
Figure 10-2. ER-SST anomalies – ABA index
Some minor warm (1986, 1988, 1997/98) and cold (1969, 1982/83) events in the southeast Atlantic have been partly documented (Carton and Huang 1994; Walker 1987; Boyd et al. 1987). Other warm events also developed in the ABA in May 1991, in April 1996 and from January 1998 to March 2001 while strong cold events were observed in April 1978, 1992, 1997 and 2004. Figure 10-2 also suggests a warming tendency in the ABA since the beginning of the 1980s. This warming tendency is also noticeable over most of the tropical southeast Atlantic Ocean, although weaker than that in the ABA.
230
C.J.C. Reason et al.
Source of Benguela Niños To explain the origin of climate anomalies in the eastern tropical Atlantic, Hirst and Hastenrath (1983) suggested a causality chain of atmospheric-oceanic anomalies across the basin. Benguela Niños are generated by specific wind stress events in the west-central equatorial Atlantic, and progress from there as subsurface temperature anomalies that eventually outcrop only on reaching the south-west African coast (Florenchie et al. 2003). At present, the driving mechanism for these wind stress anomalies is not well understood. Figure 10-3 indicates that the warm pool associated with the 1995 Benguela Niño along the Angola and Namibia coast led to local SST anomalies of greater than 5°C.
Figure 10-3. OI-SST anomalies during the mature phase of the 1995 Benguela Niño
There is indeed a strong correlation between SST anomalies in the ABA and interannual zonal wind anomalies over the equatorial western Atlantic basin. Two numerical simulations of the tropical Atlantic Ocean using the OPA model over the periods 1982-1999 and 1992-2000 confirmed the equatorial origin of Benguela Niños and suggested a mechanism based on equatorial and coastal trapped waves to explain the equatorial origin of most episodes (Florenchie et al. 2004). Local sea-air heat flux exchanges do not seem to pre-condition the sea surface in the Angola-Benguela region prior to the arrival of an event. Subsurface anomalies are attributed to vertical shifts of the thermocline under the action of propagating equatorial Kelvin waves initially triggered by zonal wind variations. On reaching the African coast, poleward traveling coastally trapped waves transport the signal towards the ABA. Temperature anomalies are more or less visible at the surface depending on various factors like the strength of
Influences of large scale climate modes and Agulhas variability on the BCLME
231
the event, the fluctuations of the depth of the thermocline or the strength of the local upwelling-favourable winds. The development of equatorial subsurface anomalies could be detected in advance via local measurements (e.g., PIRATA) so that it may be possible to predict the occurrence of these disruptive events with a lead-time of about 2 months. Atlantic Niños and Benguela Niños The equatorial Atlantic can develop an interannual phenomenon similar but weaker to ENSO, sometime known as the Atlantic Niño or zonal mode. It is less robust and cannot be self-sustained (Zebiak 1993). Many authors suggest that the eastern equatorial Atlantic is primarily governed by remote wind stress effects through equatorial wave dynamics (Picaut 1985; Huang and Shukla 1997). Warm/cold anomalies are linked to a relaxation/increase of zonal wind stress in the western equatorial Atlantic. Most of the variability along the equator takes place in the central / eastern Atlantic. SST interannual fluctuations in the ABA and the cold tongue area (CTA) appear to be strongly related. An index has been defined to compare the interannual variability over these two areas. Figure 10-4 shows two time series of detrended optimally interpolated (OI-SST) (Reynolds and Smith 1994) and extended (ER-SST) (Smith and Reynolds 2004) SST anomalies averaged respectively over the ABA and the CTA (15°W-5°W/2.5°S-2.5°N) for the periods 1982-2004 and 19602003.
Figure 10-4a: OI-SST anomalies - ABA index (black) - CTA index (red)
Figure 10-4b: ER-SST anomalies – ABA index (black) - CTA index (red)
232
C.J.C. Reason et al.
Both Figure 10-4a and 10-4b show strongly correlated signals significant at the 99 % level (the maximum correlations are 0.64 and 0.60 respectively with a one-month lag for the CTA). However, anomalies in the ABA have much larger magnitudes than those in the CTA. This difference may occur since the temperature anomalies outcrop on reaching the ABA whereas they tend to remain beneath the surface in the CTA. According to Figure 10-4a, seven significant and persistent warm events (Atlantic Niños) with SST anomalies exceeding 0.8°C developed over the CTA index during the 1982-2004 period: 1984, 1987, 1988, 1991, 1995, 1996, 1999. They tend to peak in June/July, a few months after a warm episode in the ABA (March/April). There are some exceptions - the June 1991 warm event in CTA occurred only one month after the May 1991 warm episode in the ABA whereas the 1987 warm event has no equivalent in the ABA. Five cold events exceeding 0.8°C are observed in the CTA during 1982, 1983, 1992, 1997 and 2004. Once again, these CTA events tend to be preceded a few months earlier by similar cold events in the ABA that have larger amplitude. Since SST anomalies in the ABA and the CTA are both thought to be triggered by wind anomalies along the equator, similarities between the two signals are not surprising. Anomalies in the CTA are related to the vertical structure of the thermocline and its fluctuations under the surface. The fact that the ABA mode leads the CTA one could be explained by the spatial and temporal scales of Benguela Niños. During Benguela Niños or strong warm events, subsurface warm anomalies spread from the equator to the ABA at the depth of the thermocline for a few months. The seasonal lifting of the thermocline during austral winter could bring these warm anomalies near the surface in the CTA, a few months after their occurrence in the ABA. However, two warm episodes in the ABA have no equivalent in the CTA during the 1982-2004 period: the 1986 and 2001 warm events. Various factors like local wind regimes, heat flux anomalies, coupling between the atmosphere and the ocean with subsequent positive or negative feedbacks, modulations to the Angola Current, and Kelvin wave reflections at the eastern side of the basin could also play a key role in modulating the SST variability in the ABA and CTA. ENSO and Tropical Atlantic Variability In the tropical North Atlantic, the impact of ENSO is transferred via the atmosphere wave-train generated in the tropical Pacific (Huang 2004). These teleconnections locally modify the northeasterly trade winds, which then influence the air-sea heat fluxes and induce warm anomalies in the tropical North Atlantic a few months after the mature phase of ENSO. The North Atlantic Oscillation (NAO) can also influence the region and modulate the impact of ENSO (Melice and Servain 2003). Along and south of the equator, a clear relationship between ENSO and Atlantic Niños or Benguela Niños is not apparent. For instance, the 1984 warm event in the tropical Atlantic is often cited as a result of the 1982/83 El Niño. However, inconsistent results may occur when using different SST products at different periods for different events. Saravanan and Chang (2000) described a surface wind stress signal over the western
Influences of large scale climate modes and Agulhas variability on the BCLME
233
equatorial Atlantic associated with ENSO. The observational analyses of Enfield and Mayer (1997) suggest that the mean easterly winds are intensified in the western equatorial Atlantic in response to El Niño but that coherent local changes in SST are less obvious. Curtis and Hastenrath (1995) concluded that warm events in the Pacific induce only a modest warming in the western tropical Atlantic whereas Ruiz-Barradas et al. (2000) and Zebiak (1993) found no significant impact of ENSO on the equatorial Atlantic Ocean variability. In the South Atlantic, composite maps of SST and wind anomalies derived for 14 strong events during the past century (Reason et al. 2000) suggest that warm (cool) anomalies evolve in certain large areas of the subtropical / midlatitude southeast Atlantic during the transition and peak phases of El Niño (La Niña) in association with stronger (weaker) tradewinds. Colberg et al. (2004) used output from the ORCA2 ocean model forced with 50 years of NCEP re-analyses to show that SST over most of the South Atlantic responds to ENSO via modulations to the surface heat fluxes induced by the anomalous winds with about a one season lag. Taken together, this previous work suggests that there is an ENSO impact in the southeastern Atlantic but the exact relationship between this signal and other modes such as Benguela Niños remains to be clarified. In their study of the meridional gradient mode, in which the tropical North and South Atlantic display opposite signed SST anomalies, Melice and Servain (2003) found that the anomalies in the South Atlantic seem to lead the SOI index, implying a precursory event to SOI in the region. Wright (1986) presented evidence of three largely independent precursors of changes in the SOI, one of them in the equatorial Atlantic. Jury et al. (2002) found that upper zonal winds over the equatorial Atlantic led the Niño3 index in the Pacific. In fact, the ABA index appears to be anti-correlated with OI-SST anomalies over the central Pacific basin, south of the equator, during the years 1982-1985 and 1995-1997. A sliding correlation has been calculated between the two series with a window of 36 months and the strongest negative coefficients are -0.7 to 0.75 for the 1982-1985 and 1995-1997 periods (significant at the 90 % level). It is worth noting that these two periods include major cold and warm events in the ABA. For other times, there is sometimes an in-phase relationship between the two series but it is not significant. The corresponding time lag is 8 months with the leading signal in the ABA. Different leads and lags were tried but the 8 month lag gave the strongest results. Figure 10-5 depicts two time series of OI-SST anomalies over the ABA and the PAC index (160°W-130°W, 8°S-2°N). In Figure 10-5, the ABA signal is shifted by 8 months so, for example, the April 1997 cold event appears in December 1997. During the last two decades, the two major cold events in the ABA (namely, 1982 and 1997) preceded the two strongest El Niños (1982/83, 1997/98). This result could be explained by a quick response of the equatorial Atlantic winds to SST anomalies in the Pacific via atmospheric teleconnections. The limited width of the Atlantic Ocean could also contribute to a faster response in the ABA. The 1984 and 1995 Benguela Niños were followed by moderate cold events in the Pacific. However, since Figure 10-5 only relates to data after 1982, it is possible that the relationship between the ABA and the PAC SST indices could also change considerably from one decade to the next.
234
C.J.C. Reason et al.
Figure 10-5. Detrended OI-SSTA over the ABA index (black) and the PAC index (red). A lag of 8 months is introduced in the ABA time series.
INFLUENCE OF VARIABILITY IN THE SOUTHERN AGULHAS SYSTEM ON THE BCLME REGION The retroflection region of the southern Agulhas Current is characterised by substantial mesoscale variability in the form of rings, eddies and meanders as well as by shifts in the position of the retroflection itself on a range of time scales. Both model and observations (de Ruijter et al. 1999; Reason et al. 2003; Treguier et al. 2003) suggest that there is significant variability of the heat and volume transport from the Agulhas system into the South East Atlantic on mesoscale, seasonal and interannual time scales. Rings shed off the retroflection transport heat and salt into the Cape Basin of the South East Atlantic, thereby influencing the water masses of the southern BCLME region. Lutjeharms et al. (1992) suggested that the Agulhas retroflection loop moves westwards at around 12 km a day until a ring is shed, with its westernmost limit being at around 5°E. A ring is formed when the westward flowing Agulhas Current in the retroflection loop ‘touches’ the eastward flowing Agulhas Return Current, thereby ‘short-circuiting’ the system and forming a new retroflection loop with the old loop being occluded as a ring. In this process, cold Antarctic Surface Water is able to sometimes flow northward through the gap that is formed between the ring and the new loop (Lutjeharms and Fillis 2003) and this flow may also impact on the southern Benguela region. After a ring has been shed, the current then makes an abrupt ‘jump’ back to its preferred position at around 20°E. Molinari et al. (2002) showed a correlation between ring shedding events and peaks in the transport into the South Atlantic Ocean, with between 4 to 7 rings being shed per year. Substantial interactions between various rings and eddy features occur in the southeastern part of the Cape Basin, termed the Cape Cauldron by Boebel et al. (2003), leading to a complex influence on the water mass properties of the southern Benguela. Since the anticyclonic features track to the northwest whereas the cyclonic eddies tend to move more to the west-southwest (Boebel et al. 2003), the former are more likely to influence the southern Benguela system. These authors estimated that most of the
Influences of large scale climate modes and Agulhas variability on the BCLME
235
intermediate level transport into the southeastern Cape Basin comes from the east (i.e., the Agulhas system), with only a small direct inflow from the Atlantic. Evidence exists (van Leeuwen et al. 2000) that variability in the retroflection region may be related to Natal pulses, mesoscale features that develop in the Natal Bight north of Durban. This result suggests that the transboundary influence on the southern Benguela may be related to processes occurring in the northern Agulhas current. On a larger scale, Schouten et al. (2002) suggested that eddy variability in the Agulhas system could be traced back to that occurring in the equatorial Indian Ocean via Rossby and Kelvin wave propagation of the signals across the basin. These authors related this variability to 90-day oscillations in the monsoonal winds. As a result, the possibility exists of a potential influence of large scale modes such as ENSO and the Indian Ocean Zonal Mode on the southern Agulhas, and hence on the southern BCLME, via this wave propagation mechanism. SUMMARY This chapter has reviewed the major large scale modes that exist in the South Atlantic as well as those external to the region such as ENSO and the Antarctic Oscillation that may influence the BCLME region. Most attention has been paid to the Benguela Niño which is the most well known and dominant mode in the South East Atlantic. In addition to these large scale modes, the highly variable southern Agulhas Current also influences at least the southern part of the BCLME region. Some of the Agulhas Current variability develops locally due to instability processes but at least part may be linked to that evolving in the tropical Indian Ocean. Although the most recognizable feature of the BCLME region is its upwelling, the strength and timing of this process and its related SST expression is modulated by ENSO and likely also by other large-scale modes of variability. In the north of the BCLME region, the Benguela Niño impacts on the Angola – Benguela frontal zone and on SST to as far south as about 25ºS. The proximity of the BCLME region to the Southern Ocean and the South West Indian Ocean, due to the termination of Africa in the sub-tropics, means that the Benguela upwelling system tends to display greater variability than do the Humboldt, Canary or California Current upwelling systems. Better understanding of this variability is fundamental for assessing its potential predictability and for developing appropriate management strategies of its rich ecosystems. A concern is the relative sparseness of in situ observations in the region. These include not only ocean data but also land surface and atmospheric data which, amongst other applications, are needed to improve the reliability of the NCEP re-analyses often used to diagnose modes of variability over the region. In particular, prediction efforts for the region are significantly hindered by a lack of data with which to validate model outputs. Close collaboration between the observing system and modelling communities is therefore needed in order to make progress on better understanding the variability of the BCLME region and working towards prediction.
236
C.J.C. Reason et al.
ACKNOWLEDGEMENTS Partial funding for this work from the BCLME and the South African Water Research Commission is gratefully acknowledged. Three anonymous reviewers provided useful comments which helped us to improve the chapter. REFERENCES Allan, R.J. 2000: ENSO and climatic variability in the last 150 years. In El Niño and the Southern Oscillation: Multiscale variability and its Impacts on Natural Ecosystems and Society, Diaz, H.F., and Markgraf, V. (Eds.). Cambridge University Press, Cambridge, U.K. Allan, R.J., C.J.C. Reason, J.A. Lindesay, T.J. Ansell. 2003 Protracted ENSO episodes over the Indian Ocean region. Deep-Sea Res. II, Special Issue on Physical Oceanography of the Indian Ocean: From WOCE to CLIVAR, 50, 2331-2347. Boebel O., J.R.E. Lutjeharms, C. Schmid, W. Zenk, T. Rossby and C. Barron. 2003. The Cape Cauldron: A regime of turbulent inter-ocean exchange. Deep-Sea Res. II 50: 57-86. Boyd, A.J., J. Salat and M. Masó. 1987. The seasonal intrusion of relatively saline water on the shelf off northern and central Namibia, The Benguela and Comparable Ecosystems, S. Afr. J. Mar. Sci. 5, 107120. Boyer, D. C. and I. Hampton. 2001. An overview of the living marine resources of Namibia. S. Afr. J. Mar. Sci. 23: 5-35. Boyer, D.C., H.J. Boyer, I. Fossen and A. Kreiner. 2001. Changes in abundance of the northern Benguela sardine stock during the decade 1990-2000, with comments on the relative importance of fishing and the environment. S. Afr. J. Mar. Sci. 23:76-84. Carton, J.A. and B. Huang. 1994. Warm events in the Tropical Atlantic, J. Phys. Oceanogr. 24:888-903. Colberg, F., C.J.C. Reason, and K. Rodgers. 2004: South Atlantic response to ENSO induced climate variability in an OGCM. J. Geophys. Res., 109, C12015, 10.1029/2004JC002301. Cook, K.H. 2001: A Southern Hemisphere wave response to ENSO with implications for Southern Africa precipitation. J. Atmos. Sci. 58:2146–2162. Curtis, S., and S. Hastenrath. 1995. Forcing of anomalous sea surface temperature evolution in the tropical Atlantic during Pacific warm events, J. Geophys. Res. 100:15835-15847. de Ruijter, W.P.M., A. Biastoch, S.S. Drijfhout, J.R.E. Lutjeharms, R.P. Matano, T. Pichevin, P.J. van Leeuwen, W. Weijer. 1999. Indian-Atlantic inter-ocean exchange: dynamics, estimation and impact. J. Geophys. Res. 104:20885-20911. Enfield, D.B. and D.A. Mayer. 1997. Tropical Atlantic sea surface temperature variability and its relation to El-Niño-Southern Oscillation, J. Geophys. Res. 102: 929-945. Fauchereau, N., S. Trzaska, Y. Richard, P. Roucou and P. Camberlin. 2003: Sea surface temperature covariability in the Southern Atlanitc and Indian Oceans and its connections with the atmospheric circulation in the southern Hemisphere. Int. J. Climatol. 23:663-677 Florenchie, P., J.R.E. Lutjeharms, C.J.C. Reason, S. Masson and M. Rouault. 2003. The source of Benguela Niños in the South Atlantic Ocean, Geophy. Res. Lett., 30, 10, 1505, doi: 10.1029/2003GL017172. Florenchie, P., C.J.C. Reason, J.R.E. Lutjeharms, M. Rouault, C. Roy and S. Masson. 2004. Evolution of interannual warm and cold events in the Southeast Atlantic ocean, J. Climate 17: 2318-2334. Foltz G.R., M.J. McPhaden. 2004. The 30–70 day oscillations in the tropical Atlantic. Geophys. Res. Lett., 31, L15205, doi:10.1029/2004GL020023. Gammelsrød, T., C.H. Bartholomae, D.C. Boyer, V.L.L. Filipe and M. J. O’Toole. 1998. Intrusion of warm surface water along the Angolan-Namibian coast in February-March 1995: the 1995 Benguela Niño, Benguela Dynamics. S. Afr. J. Mar. Sci. 19: 51-56. Garreaud, R. and D. Battisti. 1999: Interannual (ENSO) and interdecadal (ENSO-like) variability in the Southern Hemisphere tropospheric circulation. J. Climate. 12:2113–2123. Hall, A. and M. Visbeck. 2002: Synchronous variability in the Southern Hemisphere atmosphere, sea ice, and ocean resulting from the Annular Mode. J. Climate 15:3043–3057.
Influences of large scale climate modes and Agulhas variability on the BCLME
237
Hermes, J.C., and C.J.C. Reason. 2005: Ocean model diagnosis of interannual co-evolving SST variability in the South Indian and Atlantic Oceans. J. Climate, 18, 2864-2882. Hermes, J.C., C.J.C. Reason and J.R.E. Lutjeharms. 2005: Modelling the variability of source regions of the Agulhas Current. J. Climate, submitted. Hirst, A.C., and S. Hastenrath. 1983. Atmosphere-ocean mechanisms of climate anomalies in the Angolatropical Atlantic sector. J. Phys. Oceanogr. 13: 1146-1157. Huang B. H. and J. Shukla. 1997. Characteristics of the interannual and decadal variability in a general circulation model of the tropical Atlantic Ocean. J. Phys. Ocean 27: (8) 1693-1712. Huang, B. 2004. Remotely forced variability in the tropical Atlantic ocean. Clim. Dynamics 23:133-152, doi: 10.1007/s00382-004-0443-8. Jury, M.R., D.B. Enfield, J.-L. Melice. 2002. Tropical monsoons around Africa: stability of ENSO associations and links with continental climate, J. Geophys. Res. 107 (C10): 15-1, 15-17 Kidson, J.W. 1988: Interannual variations in the Southern Hemisphere circulation. J. Climate 1:11771198. Lutjeharms, J.R.E, W.P.M. de Ruijter and R.G. Peterson. 1992: Interbasin exchange and the Agulhas retroflection; the development of some oceanographic concepts. Deep-Sea Research A 39(10A): 17911807. Lutjeharms, J.R.E. and C.S. Fillis, 2003: Intrusions of sub-Antarctic water across the Subtropical Convergence south of Africa. S. Afr. J. Sci., 99 (3-4): 173-176. Madec, G., P. Delecluse, M. Imbard, and C. Levy (1998), OPA 8.1 Ocean General Circulation Model reference manual. Notes de Modelisation de I’ Institut Pierre-Simon Laplace, 11, 91pp. Meehl, G.A., J.W. Hurrell and H. van Loon. 1998. A modulation of the mechanism of the semiannual oscillation in the Southern Hemisphere. Tellus, 50A: 442-450. Meeuwis, J.M., and J.R.E. Lutjeharms.1990. Surface thermal characteristic of the Angola-Benguela front, S. Afr. J. Mar. Sci. 9:261-279. Melice, J.-L. and J. Servain. 2003. The tropical Atlantic meridional SST gradient index and its relationships with SOI, NAO and Southern ocean. Clim. Dynamics 20:447-464, doi: 10.1007/s00383002-0289-x. Mo, K.C. and J.N. Paegle. 2001.: The Pacific-South American modes and their downstream effects. International Journal of Climatology 21:1211-1229. Mohrholz, V., M. Schmidt and J.R.E Lutjeharms, 2001, The hydrography and dynamics of the AngolaBenguela Frontal Zone and environment in April 1999, S. Afr. J. Sci., 97, 199-208. Molinari, R.L., R. Lusic, S.L. Garzoli, M. O. Baringer and G. Goni. 2002: Benchmarks for Atlantic Ocean Circulation. CLIVAR Exchanges, 7(3-4): 35, 6-9. Philander, S.G.H. and R.C. Pacanowski. 1986. A model of the seasonal cycle in the tropical Atlantic Ocean. J. Geophy. Res., 91 (14):192-14,206. Picaut, J., 1985: Major dynamics affecting the eastern tropical Atlantic and Pacific Oceans. CalCOFI Report, Vol. 26. Reason, C.J.C., 2000: Multidecadal climate variability in the subtropics / midlatitudes of the Southern Hemisphere oceans. Tellus 52A:203-223. Reason, C.J.C., R.J. Allan, J.A. Lindesay and T.J. Ansell. 2000. ENSO and climatic signals across the Indian Ocean basin in the global context: Part I, Interannual composite patterns, Int. J. Clim. 20:12851327. Reason, C.J.C., M. Rouault, J.-L. Melice and D. Jagadeesha. 2002: Interannual winter rainfall variability in SW South Africa and large scale ocean-atmosphere interactions. Met. Atmos. Phys., Special Issue on Atmosphere- surface interactions 80 (1-4):19-29. Reason, C.J.C., J.R.E. Lutjeharms, J. Hermes, and A. Biastoch. 2003. Inter-ocean fluxes south of Africa in an eddy-permitting model, Deep-Sea Research, Part II, 50:281-298. Reason, C.J.C. and A. Keibel. 2004. Tropical Cyclone Eline and its unusual penetration and impacts over southern Africa. Weather and Forecasting 19 (5), 789-805. Reason, C.J.C. and M. Rouault. 2002. ENSO-like decadal patterns and South African rainfall. Geophys. Res. Lett. 29 (13), doi:10.1029/2002GL014663. Reason, C.J.C. and Rouault, M. 2005. Links between the Antarctic Oscillation and winter rainfall over western South Africa. Geophys. Res. Lett., 32(7), L07705 , doi 10.1029/2005GL022419. Reynolds, R.W., and T.M. Smith. 1994. Improved global sea surface temperature analyses using optimum interpolation. J. Climate 7:929-948.
238
C.J.C. Reason et al.
Risien, C., C.J.C. Reason, F. Shillington, D.B. Chelton, 2004: Variability in satellite winds over the Benguela upwelling system during 1999-2000. J. Geophys. Res., 109, C3, C03010, doi10.1029/2003JC001880. Rouault, M., P. Florenchie, N. Fauchereau and C.J.C. Reason. 2003: South East Atlantic warm events and southern African rainfall. Geophys. Res. Lett. 30 (5): 8009, doi:10.1029/2002GL014840. Rouault, M., Melice, J-L., Reason, C.J.C. and J.R.E. Lutjeharms. 2005. Climate variability at Marion Island, Southern Ocean, since 1960. J. Geophys. Res. 110, C05008, doi10.1029/ 2004JC002492. Roy, C., S. Weeks, M. Rouault, G. Nelson, R. Barlow, and C. van der Lingen. 2001: Extreme oceanographic events recorded in the southern Benguela during the 1999-2000 summer season. S. Afr. J. Sci. 97:465-471. Ruiz-Barradas, A., J.A. Carton and S. Nigam. 2000. Structure of interannual-to-decadal climate variability in the tropical Atlantic sector. J. Climate 13:3285-3297. Saravanan, R., and P. Chang. 2000. Interaction between tropical Atlantic variability and El-Niño-Southern Oscillation, J. Climate 13:2277-2292. Schouten, M.W., W.P.M. de Ruijter, P.J. van Leeuwen and H.A. Dijkstra. 2002: A teleconnection between the equatorial and southern Indian Ocean. Geophys. Res. Lett, 29:doi:10.1029/2001GL014542. Servain, J., and S. Arnault. 1995. On forecasting abnormal climatic events in the tropical Atlantic Ocean, Ann. Geophysicae 13:995-1008. Shannon, L.V., A.J. Boyd, G.B. Bundrit and J. Taunton-Clark. 1986. On the existence of an El Niño-type phenomenon in the Benguela system J. Mar. Sci. 44:495-520. Shannon, L.V., J.J. Agenbag and M.E.L. Buys, 1987, Large and mesoscale features of the AngolaBenguela front, S. Afr. J. Mar. Sci. 5:11-34. Shillington, F.A., C.J.C. Reason, C.M. Duncombe Rae, P. Florenchie, and P. Penven. 2006. Large scale physical variability of the Benguela Current Large Marine Ecosystem (BCLME). This volume, Chapter 4, 47-68. Simmonds I. Modes of atmospheric variability over the Southern Ocean, J. Geophys. Res. 108 (C4): 8078, doi:10.1029/2000JC000542, 2003. Smith, T.M., and R.W. Reynolds. 2004. Improved extended reconstruction of SST (1854-1997). J. Climate 17:2466-2477. Thompson, D.W.J. and J.M. Wallace. 2000: Annular modes in the extratropical circulation. Part I: monthto-month variability. J. Climate 13:1000-1016. Treguier, A.M., O. Boebel, B. Barnier and G. Madec. 2003. Agulhas eddy fluxes in a 1/6 degrees Atlantic model. Deep-Sea Res., Part II 50: 251-280. van Leeuwen P. J., W.P.M. de Ruijter and J.R.E. Lutjeharms. 2000. Natal Pulses and the formation of Agulhas rings. J. Geophys. Res. 105:6425-6436. van Loon, H. 1967: The half-yearly oscillations in middle and high southern latitudes and the coreless winter. J. Atmos. Sci. 24:472-486. van Loon, H., J.W. Kidson and A.B. Mullan. 1993: Decadal variation of the annual cycle in the Australian dataset. J. Climate 6:1227-1231. Veitch, J. 2004: Numerical model investigation of near-surface circulation features of the Angola Basin. Unpublished MSc thesis, University of Cape Town, 130 pp. Veitch, J., P. Florenchie and F. A. Shillington. 2006. Seasonal and interannual fluctuations of the Angola Benguela Frontal Zone (ABFZ) using 4.5 km resolution satellite imagery from 1982 to 1999, Int. J. Rem. Sensing 27(5):987-998. Venegas, S.A., L.A. Mysak and D.N. Straub. 1996: Evidence for interannual and interdecadal climate variability in the South Atlantic. Geophys. Res. Lett. 23:2673-2676. Wainer, I. and S.A. Venegas. 2001. South Atlantic multidecadal varibility in the Climate System Model. J. Climate 15:1408-1420. Walker, N.D. 1987. Interannual sea surface temperature variability and associated atmospheric forcing within the Benguela system. S. Afr. J. Mar. Sci., 5, 121-132. White W.B., and R.G. Peterson. 1996: An Antarctic circumpolar wave in surface pressure, wind, temperature and sea-ice extent. Nature 380:699-702. Wright, P.B. 1986. Precursors of the Southern Oscillation. Int. J. Climatol. 6:17-30 Zebiak, S.E. 1993. Air-sea interaction in the equatorial Atlantic region. J. Climate 6:1567-1586.
239
Large Marine Ecosystems, Vol. 14 V. Shannon, G. Hempel, P. Malanotte-Rizzoli, C. Moloney and J. Woods (Editors) © 2006 Elsevier B.V. All rights reserved.
11 Developing a Basis for Detecting and Predicting Long-Term Ecosystem Changes A. Jarre, C.L. Moloney, L.J. Shannon, P. Fréon, C.D. van der Lingen, H.M. Verheye, L. Hutchings, J.-P. Roux, P. Cury
ABSTRACT Long-term ecosystem changes in the Benguela region include species alternations and regime shifts, which are sometimes obscured by large intra- and inter- annual variability in the ecosystem. This chapter proposes that no single model or approach can resolve this variability and effectively detect and predict long-term ecosystem changes; a coherent, robust, transparent and reproducible synthesis framework is required. Indicators and models are described that can be used to identify some aspects of the current state of ecosystem structure and to detect and monitor long-term change. A short-term challenge is to synthesize these varied sources of multidisciplinary (and sometimes contradictory) information in a logical and consistent fashion. An expert system approach is proposed to do this, consolidating results of different indicators and models within a dynamic process that uses feedbacks to validate predictions of the expert system, and to improve it. It is suggested that such an approach should be initiated in the short term, even as models and indicators are being developed further. In parallel, multivariate statistical tools should be refined and applied to existing time series, to identify past periods of ecosystem change. Current data gaps should be filled, including time series of primary production and the abundance of gelatinous zooplankton. In the medium term, the expert system model should evolve to a point where its results can be used to inform various management groups about the state of the ecosystem. Part of this evolution requires that ecosystem indicators be presented with error estimates or formal assessments of quality. INTRODUCTION The detection and prediction of ecosystem states and changes in those states is at the very heart of ocean observation programmes globally (IOC/GOOS 2003), and one of the key policy actions of the Benguela Current Large Marine Ecosystem (BCLME) Programme (O’Toole et al. 2001). Two cornerstones of this policy action are the development of an early warning system for monitoring major environmental events within the Benguela LME, and the improvement of the predictability of extreme events and their impacts. While extreme events can have severe impact in the region in the
240
A. Jarre et al.
short and medium term (e.g., Chapters 4, 5 and 7, this volume), this contribution focuses on detecting and monitoring changes in the Benguela LME region in the long term. The Benguela LME consists of three main sub-regions, (i) the subtropical shelf ecosystem north of the Angola Front, (ii) the northern Benguela ecosystem, a typical eastern boundary upwelling system bounded by the Angola-Benguela Front in the north and the Lüderitz upwelling cell in the south, and (iii) the southern Benguela ecosystem, which extends along South Africa's west and south coasts until approximately East London, and shows characteristics of both an eastern boundary current system and a temperate shelf ecosystem. The ecosystems in each of these subregions have their own characteristics and dynamics, and assessing and forecasting change is therefore a complex topic. It is unlikely that a single approach can be used across these systems. Indeed, it is unlikely that a single approach would be effective within any ecosystem, because of the range of scales involved, and the complex interactions that exist among living and non-living components of ecosystems. We focus instead on highlighting a suite of approaches to detect and monitor long-term change, and methodology to synthesise the results of different approaches. We emphasise a general procedure that should be applicable to any subsystem in the Benguela LME region. Our focus on the long term includes analyses of causes and effects of species alternations (sensu Schwartzlose et al. 1999) as well as regime shifts (e.g., Cury and Shannon 2004). Consequently, our management concerns are strategic (i.e., on the time scale of 4-7 years) as opposed to tactical (1-3 years); we anticipate that the objectives of tactical management, typically single-species or fisheries-based, will be fashioned on the basis of strategic thinking guided by long-term ecosystem considerations. Similarly important on a strategic basis are changes that affect or manifest among communities (e.g. zooplankton communities, kelp-bed communities, benthic shelf communities) or changes that affect species with long population cycles (e.g., seabirds, cetaceans, predatory reef-fish), if there is an established link to ecosystem-level changes. Van der Lingen et al. (this volume) propose a way forward in addressing long-term ecosystem change in the Benguela Current region. They emphasize the need to identify and understand different states of the ecosystem, the controls operating within ecosystems, and the processes by which change occurs. They further recommend that suites of ecosystem indicators be used to describe and quantify ecosystem change, and that these indicators be synthesized to allow predictions. The main objective of this chapter is to expand on these proposals. The chapter aims to answer a number of questions about the complex task of detecting and predicting long-term ecosystem changes. These include the kinds of changes that should be considered, and the many ways in which they might be measured and modelled. Composite indicators rather than single variables are believed to be most useful for depicting many ecosystemlevel attributes, and the chapter asks which of these indicators will be most useful, and what models can be used to derive and test them. Finally, the chapter aims to outline a feasible way of combining models and measurements of key characteristics of
Developing a basis for detecting and predicting long-term ecosystem changes
241
ecosystems, by integrating the varied approaches that will probably be used in an “expert system model”. ECOSYSTEM CHANGES TO BE MONITORED What is ecosystem change? There is neither a consistent definition of regime shift, nor is there a consistent approach to determining ecosystem state, changes in the state, or the relation between the physical, chemical and biological environment that might trigger a regime shift or ecosystem change (de Young et al. 2004). Several proposed definitions of regime shifts are given (Table 11-1). We adopt the definition that a regime shift is a rapid change from a quantifiable state, representing substantial restructuring of the ecosystem, acting over large spatial scales and persisting for long enough that a new quasi-equilibrium state can be observed (de Young et al. 2004). If ‘ecosystem state’ were to be defined, a change in state would need to be measurable. However, unlike (closed) freshwater/lake systems (e.g. Scheffer et al. 2001, Scheffer and van Nes, 2004), determination of ecosystem state in (large and open) marine ecosystems proves difficult, remaining an unresolved, imprecise problem (see e.g. Longhurst 1998). Therefore, at present, we have to be satisfied with a broader, and perhaps less ecologically precise, definition of “regime shift” in marine ecosystems. Nevertheless, appropriate statistical analyses need to be developed and applied before concluding whether or not a regime shift has occurred (or is occurring). Throughout this chapter we distinguish between bottom-up environmental forcing (e.g. by changes in winds, ocean currents, temperatures, oxygen concentrations, etc) and anthropogenic forcing (which can be bottom-up through e.g. pollution, or top-down through e.g. fishing). The response of the ecosystem to environmental or anthropogenic forcing will depend on the ecosystem state and its functioning, underlining the need to understand the inherent characteristics of the ecosystem to be able to predict probable shifts or changes. Collie et al. (2004a,b), similarly also de Young et al. (2004), exemplify three ways in which regime shifts occur: 1. Gradual change e.g. shifts between dominance of coral and macroalgae around Jamaica, which is generally reversible, 2. Abrupt change e.g. the cod collapse in the North Sea, which is not necessarily reversible, and 3. Discontinuous shift e.g. the increased abundance of jellyfish off Namibia, caused by fishing or environmental forcing and which is unlikely to be easily reversible. Following Mantua’s (2004) definition (see Table 11-1) and the stability landscape model of Scheffer et al. (2001), two diagrammatic representations are presented to assist in visualising the idea of stable states and attractors (Figures 11-1 and 11-2). Unlike during regime shifts, ecosystem structure and functioning are not necessarily altered during replacements or alternations (see Table 11-1) of species at similar
242
A. Jarre et al.
Table 11-1. Some definitions of regime, regime shift, and species replacement/alternation important to the BCLME.
Reference
Definition
Regime Mantua (2004)
A period of quasi-stable biotic or abiotic system behaviour where temporal variations in key state variables are concentrated near distinct dynamical attractors, or stability wells, within phase space.
Lluch-Belda et al. (1989,1992)
Prolonged periods of high or low abundance of species.
Isaacs (1976)
Distinct climatic and/or ecosystem states and is multifarious, involving biology or climate, or oceanography, or migrations, temperature, or weather, or combinations of these.
Regime shift Bakun (2004)
Persistent radical shift in typical levels of abundance or productivity of multiple important components of marine biological community structure, occurring at multiple trophic levels and on a geographical scale that is at least regional in extent.
Cury and Shannon (2004)
Sudden shift in structure and functioning, which affect several living components and which result in an alternate state.
Wooster and Zhang (2004)
Abrupt change in a marine ecosystem and its abiotic environment from one stationary state to another.
Polovina (2005)
High-amplitude changes in community composition, species abundance and trophic structure, thought to be a response to shifts in the oceanic and atmospheric climate, and therefore relatively coherent with climate changes.
de Young et al. (2004)
Changes in marine system structure and functioning that are relatively abrupt, persistent, occurring at large spatial scales, observed at different trophic levels, and related to climate forcing.
Mantua (2004)
Relatively brief time period in which key state variables of a system are transitioning between different quasi-stable attractors in phase space.
Mantua and Hare (2002)
Abrupt change in relation to the duration of a regime, from one characteristic behaviour to another.
Reid et al. (2001)
Large decadal-scale switches in the abundance and composition of plankton and fish.
Miller and Schneider (2000)
Change from a persistent and relatively stable period of biological productivity after a similarly stable period in physical oceanographic variables.
Caddy and Garibaldi (2000)
“Punctuated equilibria” involving fundamental changes in ecosystems and reflecting ecological change.
Developing a basis for detecting and predicting long-term ecosystem changes
Beamish and Mahnken (1999)
The process whereby a large marine ecosystem that is climate-linked, undergoes a shift in state over a 10-30 year period, and to which fish and other marine biota respond by changes in their dynamics;
Steele (1996, 1998)
Concurrent change in several stocks at longer time scales, and causally connected. Implies a coherent response, at the community level, to external stresses.
Lluch-Belda et al. (1989,1992)
Dramatic and long-lasting switches between periods of sardine and anchovy-dominated states in upwelling systems of eastern boundary current systems.
243
Species replacement or alternation Cury and Shannon (2004)
Species composition of an ecosystem changes, but ecosystem is not necessarily altered in terms of its structure (e.g., food-web, size composition) and functioning.
Lluch-Belda et al. (1992)
Negative correlation observed between similar species (e.g. sardine and anchovy) in the same ecosystem
levels, where only species composition changes (Cury and Shannon 2004). Nevertheless, species alternations (e.g. replacement by a commercially less desirable species) may have severe socio-economic implications, and are important for fisheries management. Species alternation is often associated with changes in spatial distribution of fish (e.g., van der Lingen et al., this volume) and hence availability to fisheries as well as to top predators such as commercially valuable large pelagics and vulnerable seabird species. Processes triggering regime shifts or species replacements may include both environmental changes and anthropogenic effects, e.g., fishing, which may act synergistically or antagonistically (Cury and Shannon 2004, van der Lingen et al. this volume). What changes in the ecosystem might be caused by fishing, pollution, environment, or climate? Probably the most well-known changes in the upwelling ecosystems are the decadalscale species alternations/regime shifts involving sardines and anchovies, which have been observed worldwide (Lluch-Belda et al. 1989). These changes have important management implications as they may alter the structure and functioning of ecosystems and the way in which they respond to fishing (Rothschild and Shannon 2004). It is often difficult to disentangle the possible drivers of these and other changes. Anthropogenic impacts (fishing, pollution) cause change from the pristine situation, and these changes can have undesirable consequences for the ecosystem but might be desirable for humans. For example, overfishing predators off West Africa (Caddy and Rodhouse 1998) resulted in a lucrative octopus fishery being supported. "Natural" impacts (environment, climate) can have links to human activities, but the anthropogenic effect is indirect. Often, in these cases, the forcing for the change is density-independent (displaying synchrony between biological populations on a global
244
A. Jarre et al.
or basin scale) and external to the biological ecosystem, usually forced through the physical climate system (de Young et al. 2004) and triggering a series of concomitant physical and biological processes. For instance, in the northern hemisphere, largescale climate cycles and warming trends (e.g. Hare and Mantua 2000, Beaugrand et al. 2002) – in addition to anthropogenic disturbance (e.g. eutrophication, e.g., Brander et al. 2003) – have been implicated in long-term changes in plankton abundance and community structure. Changes at these lower trophic levels have been shown to propagate up the foodweb, causing changes at higher trophic levels (e.g. Beaugrand et al. 2003), including harvested fish populations (e.g. Reid et al. 2003).
Figure 11-1. Illustration of external (natural and/or anthropogenic) forcing leading to regime changes in an ecosystem. Under weak to moderate external forcing, the ecosystem maintains its structure and functions largely in an unchanged way (A), whereas major external forcing can cause functional changes in the ecosystem that may manifest themselves in another period of weak to moderate forcing into a new regime (B). Within any regime, switching between “pseudo-states” (e.g., species dominance patterns) is generally reversible. New major forcing events may cause completely new regimes, or return the system into previous ones.
Figure 11-2. Illustration of regime change using solar systems as analogy. Any regime is assumed to be represented by a solar system with one single major attractor (A or B). Various "pseudo-states" (e.g., species dominance patterns) can exist, these are represented by sub-attractors (A1 & A2 versus B1 and B2). The mechanism by which the major attractor may change from A to B is unclear, but could be based on major external forcing as illustrated in Fig. 11-1.
Developing a basis for detecting and predicting long-term ecosystem changes
245
Although environmental forcing typically acts from the bottom-up (via phytoplankton or zooplankton resource limitation, e.g., Verheye et al. 1998, Hutchings et al. this volume), it can also manifest itself in a wasp-waist manner (via direct effects on pelagic fish recruitment, Cury et al. 2003). Often, wasp-waist structures appear through pelagic fish structuring the dynamics of their ecosystem, by controlling species at both higher and lower trophic levels (Cury et al. 2003). Overfishing of one particular pelagic fish can alter the abundance, composition and distribution of others as well as other components of the pelagic community, inducing drastic changes of state. Top-down control (predation) is considered to be the most important source of mortality for exploited species, and can affect the whole ecosystem because predation tightly connects species. Fishing, acting as a top predator, targets preferentially large fish species and, in a top-down structure, causes a shift from a large-predatory fish dominated ecosystem to one dominated by small pelagic fish, which are more sensitive to environmental change. Such changes are generally not likely to be reversible since most large fish populations are not very resilient. Fishing may not only be a cause of species alternation, but it may also be a source of additional variability (over and above natural variability) and may hasten stock collapses or slow down stock recoveries (Beverton 1990). Land-based and airborne pollution is probably less of a problem for the Benguela than for other shelf areas (e.g., the European waters), but local effects are observed (Pitcher and Weeks, this volume). Pollution by shipping and sea-based structures can be severe and may have severe system consequences, e.g., by affecting top predators in the system (Gründlingh et al. this volume) It is uncertain whether ecosystem manipulation is a viable possibility for future ecosystem management. What is certain is that it is very difficult to manipulate Benguela LME regions by virtue of their open boundaries and complex, dynamic ecological interactions, offering no guarantees that the desired ecosystem state will be reached (e.g., Moloney et al. 2004). For instance, attempts have failed to encourage a commercially valuable species to increase in abundance by fishing more heavily on its competitors. What are the causal processes driving ecosystem changes? The major driving forces in the pelagic marine ecosystems in the three Benguela LME sub-regions are winds and ocean currents. Changes in the wind forcing occur either as a change in magnitude (optimal environmental window), a change in direction, or a change in patterning (seasonal or event scale), affecting the stability of the water column and nutrient supply to the euphotic zone. This, in turn, determines the proportion of large vs. small cells in the plankton, and primary and secondary productivity. Here, only persistent changes over prolonged periods (i.e. large scale features such as changes in the South Atlantic high pressure cell off the southern African west coast) are likely to lead to regime shifts.
246
A. Jarre et al.
The large-scale current systems operating in the Bengulea region include (Shannon, 1985): (i) the South Equatorial Counter Current and Angolan Current at the northern boundary; (ii) the Agulhas Current in the south; (iii) the Benguela Current, which includes the broad drift as part of the South Atlantic gyre and the upwelling belt along the eastern edge; and (iv) the northward penetration of subantarctic water masses from the subtropical convergence in the south. Thus, currents at its boundaries set one of the main physical conditions for the ecosystem, the three others being atmospheric forcing, solar radiation and the bottom/coastline topography. Changes in the current systems affect the distribution of particular organisms or processes, such as fronts, eddies or transport. Again a persistent change is required to alter ecosystem state. Considerable feedback and secondary interactions between wind, currents and radiation can be expected to accompany changes over decades, requiring long-term monitoring programmes. The physical processes described above might cause, or at least play a role in species alternations in the Benguela LME, and notably those involving anchovies and sardines. These species alternations are potentially important for the ecosystem and consequently, to the management of human activites in the ecosystem. Species alternations might be mediated through subtle changes in feeding niches (James 1987, 1988, van der Lingen 2002) and spawning/recruitment habitat preferences (van der Lingen et al. 2001, 2002, van der Lingen and Huggett 2003) or, possibly, external factors like changes in the bio-chemical properties of the water masses entering the ecosystem. However, mechanisms maintaining the persistence of one species over the other are not yet clear (Hutchings et al. 1998, Schwartzlose et al. 1999). Fishing could also drive, or add to, the factors leading to species alternations. In the southern Benguela, the bycatch inflicted upon sardine juveniles during anchovy-directed purse seine operations when anchovies are dominant, is also important for sardine dynamics (De Oliveira et al. 1998). In the northern Benguela, the area of suitable spawning habitat appears to have remained the same in terms of temperature and phytoplankton, but the removal of “southern” sardine spawners by overfishing, the increased frequency of warm water intrusions across the Angola-Benguela front and the role of large jellyfish, which may be predators of fish eggs or larvae or competitors with sardines for zooplankton prey, are all possible mechanisms contributing to the decline in sardines. Horse mackerel increased in abundance in the northern Benguela and may play a role in suppressing sardine by preying on sardine and anchovy larvae or by enmeshing sardines in "school traps." These occur when a species at low abundance (e.g., sardine) subordinates its specific needs to those of a more abundant species by aggregating in a mixed school, potentially affecting individual fitness and reducing the population's chances of recovery (Bakun and Cury 1999). Fréon and Misund (1999) concluded that, for small pelagic fish, the urge to become a member of a school of similarly-sized fish of similar body form, regardless of species, is a dominating aspect of behaviour, and anchovy, sardine and horse mackerel have frequently been observed to aggregate in mixed schools. The midwater trawl fishery could have a significant bycatch (e.g. 5-10%) of adult sardines together with the 200,000 to 400,000 tons of horse mackerel caught annually. Sardines, currently at a much reduced population size, are probably only utilising a small fraction of the suitable habitat, and so are
Developing a basis for detecting and predicting long-term ecosystem changes
247
subject to severe predation by predators (snoek, seals, hake) which can also utilise alternative prey to maintain high population densities. What ecosystem changes is it desirable and/or possible to monitor? Monitoring upwelling areas to detect ecosystem changes is a daunting task, as boundary effects dominate the narrow, ribbon-like features on the eastern edges of ocean basins. A number of easily measured, remotely sensed parameters are obvious choices, which include surface temperature, water-leaving irradiance, wind strength and sea surface height. From these measurements a number of indices based on the habitat preferences of the dominant organisms can be derived (e.g., HardmanMountford et al. 2003, Daskalov et al. 2003, Agenbag et al. 2003, and also see references in Table 11-2). These tend to cover the spatial attributes of populations as related to boundary or average conditions. There are a few important parameters which require ground-truthing (e.g. primary production) or which cannot be measured remotely, which include zooplankton (micro-, meso- and macro-), extent of oxygendepleted water, water column stratification and aggregations of prey organisms at interfaces. These parameters represent a much more difficult task as high mesoscale variability needs to be integrated over much longer space and time scales to be pertinent for ecosystem changes. Remote sensing, buoy deployment, regular shipboard monitoring and widespread but infrequent coverage during cruises of opportunity, such as fish survey cruises, need to be integrated using dynamic ecosystem models and simulated using coupled hydrodynamic/ IBM approaches. Feedback between theory and observations allows experiments to be conducted which simulate extreme conditions, determine thresholds, and minimise sampling redundancy. A number of critical areas and processes have already been identified in the southern Benguela and theorised in the northern Benguela (Moloney et al. 2004, Hutchings et al. 2002). In the southern Benguela the SARP line off the SW Cape monitors the transport of pelagic eggs and larvae spawned on the western Agulhas Bank to the west coast. However the shift in spawning further eastwards to the central and eastern Agulhas Bank has confounded this time series to some extent. The St. Helena Bay Monitoring Line was designed to monitor: • the feeding conditions of early recruits as they entered the west coast ecosystem; • the feeding conditions of pelagic recruits inshore on the west coast as they grow and build up fat reserves for the migration southwards to the Agulhas Bank; • the seasonal changes in the phytoplankton, including harmful algal blooms and micro- and mesozooplankton across the west coast shelf; • the extent of low oxygen water on the inshore part of the shelf; • the distribution of epipelagic and mesopelagic fish and other sound-scatterers across the shelf; • the feeding patterns of large pelagic fish across the shelf; and
248
A. Jarre et al.
•
the seasonal changes in the hydrodynamic structure and fertility of the west coast. Areas not covered by current monitoring programmes in the southern Benguela include the western, central and eastern Agulhas Bank, including the cold ridge, the area off Port Elizabeth and off the tip of the Agulhas Bank, which Hutchings et al (2002) identified as critical transport gateways and the Lüderitz/Orange River cone area, i.e., the northern boundary of the southern Benguela subsystem.
Table 11-2. Indicators relevant to detecting long-term change in the BCLME
Indicator Indicator examples class and type Environmental and habitat indicators - Indices related to effects of global warming - Pollution indices, e.g., number of pollution Pressure events by category - SST, wind stress, offshore extension index; upwelling index - wind pulses; - thermogradient, stratification indices (e.g. depth of the thermocline) - intensity and position of fronts, deviation from mean position of the front; - other indices of large scale forcing, e.g., State Benguela Niño; - LOW indices; - Phytoplankton biomass and productivity; - Size distribution or spectra of phytoplankton and zooplankton; - Ratio large/small zooplankton; - Total zooplankton abundance. Single species indicators 1 - Number of non-target species caught by method, area and season or year - Exerted effort, total capacity in suitable Pressure categories - Abundance/ biomass of pollution indicator species - Abundance of gelatinous organisms - Abundance, condition, breeding success/ recruitment of sensitive species (e.g. seals, gannets, penguins) and/or top predators (e.g., hake) and/or indicator plankton forms (if any) - Dominance and distribution of species indicative State of environmental change, e.g., sardine / anchovy ratio (aimed at tracking alternation between species) - Species diversity (by community) - Genetic diversity (by species) - Abundance of exotic species - Spawner biomass & recruitment trends, age (or
Key references for the BCLME
Bakun (1990) Reinikainen & Molloy (2003)
Richardson et al. (1998) Roy et al. (2001) Demarcq et al. (2003) Reinikainen & Molloy (2003) Hardman-Mountford et al. (2003) Shillington et al. (this volume) Monteiro et al. (this volume) Painting et al. (1998) Verheye et al. (1998))
Nel et al. (2003)
Roux & Shannon (2004) Kemper et al. (2001) Roux & Mercenero (2004) Hutchings et al. (1998) van der Lingen et al. (2001) Crawford et al. (1985) Reinikainen & Molloy (2003) Korrûbel et al. (1998) Barange et al. (1999) Kreiner et al. (2001)
Developing a basis for detecting and predicting long-term ecosystem changes
length) at first maturity, growth rate trends, condition factor of trends species sensitive to fishing. Size-based indicators - Minimum authorized mesh-size on fishing gears - Minimum authorized fish size per species or Pressure group of species
− State
− −
1
Mean and maximum length of populations sensitive to ecosystem change Size at maturity, condition at size of selected (not necessarily target) populations Medium and maximum length of a community, Slope of the size spectrum
Trophodynamic indicators − FiB Pressure − Mean trophic level of catch (incl. bycatch) − Catch and biomass ratios, production and consumption ratios of selected groups/guilds (e.g., pelagic vs. demersal, planktivores vs. piscivores) State − Primary production required to sustain production of selected groups/guilds − Mixed trophic impact and similar indices of trophic dependency Spatial indicators − Mean ratio between exploited area and Pressure distribution area by species, − Exploited fraction of the ecosystem − Spawning distribution of adults and eggs of populations exploited by the fishery State − Total distribution area of the stocks, fraction of habitat actually used
249
Fairweather et al. (in prep.a,b)
Fairweather et al. (in prep.a,b), Shin et al. (2005) Yemane et al. (2004)
Cury et al. (2005a)
Cury et al. (2005a) Moloney et al. (2005)
Fréon et al. (2005b)
Drapeau et al. (2004) Pecquerie et al. (2004)
Defined here as species that are important for ecosystem structure and functioning, not necessarily limited to fisheries’ target species.
In the northern Benguela system, the extent of low oxygen water and the intensity and location of the Angola-Benguela Front are considered to be critical components. The intense upwelling zone at Lüderitz is also important as a boundary zone separating oxygenated Atlantic Central Water originating from the Cape Basin in the south from less-oxygenated Central water moving southwards from Angola. The inshore area at 20-25 oS is important as a nursery region for recruitment of hake, horse mackerel and sardine to the northern Benguela stocks. A Marine Oceanographic Monitoring (MOM) transect at 23oS is currently monitored for oceanographic and biological variables each month, with a supplementary transect at 20 oS sampled every alternative month. A further transect at 15 oS in southern Angola covers the northern boundary of the Angola-Benguela front, but logistical problems have resulted in infrequent, irregular sampling. Continuous, underway fish egg sampling should commence in Namibia in 2005, to complement the CUFES surveys undertaken in the southern Benguela. These surveys will demarcate the spatial extent of pelagic fish spawning over extensive areas in the Benguela Current. There are plans to extend oceanographic
250
A. Jarre et al.
monitoring throughout the subtropical system off Angola in the near future. For all Benguela LME sub-regions, the existing monitoring of fishing activities, and resource status, needs continuation. However, in order to assess structure, dynamics and changes at the ecosystem level, it is necessary that monitoring extend beyond abundance/biomass alone. As an example, continued monitoring of diet compositions of predators is essential for detecting and understanding changes in trophic structure (e.g. MacQueen and Griffiths 2004). APPROPRIATE ECOSYSTEM INDICATORS AND MODELS Because there is no general theory that can describe the whole functioning of marine ecosystems, the management decision process must be based on several different tools, analyses, models and indicators. An indicator generally is defined as a variable, pointer or index, whose fluctuation reveals key characteristics of a system. The position and trend of the indicator in relation to reference points or values indicate the present state and dynamics of that system, and in this respect, indicators can link observations on the one hand to management goals and objectives on the other (Slocombe 1999, FAO 1999). For this contribution, we use the term "ecosystem indicator" as a measurable characteristic of an ecosystem, which can provide feedback to the question of whether or not long-term ecosystem change is occurring. Indicators are fundamental to wider objectives in the management of human activities in the ocean, as e.g., under an Ecosystem Approach to Fisheries (Sinclair and Valdimarsson 2003, FAO 2003). Because of the link to management objectives, it is important to keep in mind that the relevance of any indicator will not only depend on the specific objectives for its use, but also on the particular group of people that is to be informed by it (FAO 2003, Degnbol and Jarre 2004). Much research on properties and applicability of marine ecosystem indicators has recently been carried out by the SCOR/IOC Working Group 119 on "Quantitative ecosystem indicators for fisheries management" (www.ecosystemindicators.org), and the proceedings of a dedicated international conference have just been published (for an overview, see Cury and Christensen 2005). In this context, "ecosystem" science in the Benguela LME is at the international forefront (Cury et al. 2005a,b, Moloney et al. 2005, Fréon et al. 2005a,b, Underhill and Crawford 2005, Yemane et al. 2005, but also see the contributions in Shannon et al. 2004c).
It is often questioned whether indicators really improve our ability to detect/predict resource and ecosystem changes beyond the detection/prediction capability that single species/resource trajectories alone can offer. The danger of relying only upon single species indicators (e.g. survey or catch records) is that one misses capturing the effects of interactions between these resources, and catch data in particular may not necessarily reflect what is happening at the community or ecosystem level. Indeed, simulations carried out to formally evaluate the performance of a large suite of indicators, show that community-based indicators may hold most promise in a management context (Fulton et al. 2005), and that it is necessary to use a variety of indicators simultaneously, capturing several key functional groups.
Developing a basis for detecting and predicting long-term ecosystem changes
251
These recent results support our general experience that the net effects of the various direct and indirect interactions in an ecosystem are often unexpected and sometimes counterintuitive. Consequently, a formal mechanism is required (and proposed below) that combines the signals from the various indicators, for use in a management context. With respect to management, several frameworks for the use of indicators exist; here we base our discussion on the DPSIR framework as used, inter alia, by GOOS and the European Community (Smeet and Weterings 1999, IOC/GOOS 2003: p.87). Contributions from the natural sciences focus on the pressure (P) and state (S) categories, whereas the indicators of drivers (D), impact (I), and response (R) often would be rooted in the social sciences. Selected indicators for the Benguela are summarised in Tables 11-2 and 11-3. What ecosystem indicators can be used to help detect change in the Benguela LME? Indicators scrutinised by the SCOR-IOC Working Group 119 were grouped into the categories "single species and habitat", "species-based", "size-based", "trophodynamic", "spatial" and "integrated". We follow this structure to introduce some indicators for detecting change in the Benguela LME. Habitat indicators for the Benguela are to a large extent derived from physical oceanography, and are mostly GOOS variables (IOC/GOOS 2003) or derivatives. Indicators of productivity and characteristics of the productivity chain provide an indication of the primary production of the ecosystem (Richardson et al. 2003 a,b) and its major characteristics in terms of structure, especially the short versus the long food chains of plankton (Demarcq et al. 2003). They can be derived from in situ or satellite observations (Carr 2002; Carr and Kearns 2003). Species-based indicators integrate various ecosystem signals in time and space while still having relatively fast response times. In the Benguela LME, they have principally been used for sensitive or threatened species (e.g., cormorants, gannets, penguins) or for top predators (e.g., seals, whales). Recently, however, a suite of indicators for both anchovy and sardine derived primarily from fishery-dependent data has been developed, including length at maturity, mean length of the catch, centre of gravity of catches, exploitation rate and others (Fairweather et al. in prep. a,b). Size-based indicators have the advantage of a good theoretical basis, and are relatively cheap to obtain, as size is easy and cheap to measure (Shin et al. 2005). Trophodynamic indicators measure the interaction strength between species and help to track structural changes in the ecosystem caused by fishing or environmental forcing. Cury et al. (2005a) reviewed indicators derived from trophic models and catch records, and applied eight selected trophic indicators to the northern and southern Benguela ecosystems. They were found useful for detecting large ecosystem changes and for understanding ecosystem and fisheries dynamics. However, good understanding of trophic interactions in the system is crucial for their meaningful
252
A. Jarre et al.
application (MacQueen and Griffiths 2004, Moloney et al. 2005). Both size-based and trophodynamic indicators appear conservative, and suitable reference points generally are still lacking. Their signal is relevant for strategic planning, not for management decisions on the short term. Several spatial indicators have been developed from a GIS that covers the areas fished by three major fleets in the southern Benguela, the foraging areas of three top predators and the distribution of 15 important fish species (Fréon et al. 2005b). These spatial indicators can be used to monitor changes in the ecosystem and the effectiveness of fisheries management in an ecosystem context. Similarly, Drapeau et al. (2004) used a GIS to explore potential spatial interactions between 13 important resources (including small pelagic fish, horse mackerel, and hake) in the southern Benguela and to quantify their spatial overlap. After the incorporation of information on the diet of different species and from trophic models, the main trophic interactions between those resources were identified and mapped, complementing conventional trophic models that are not spatially resolved. True co-operation between natural and social scientists in management of human activities in the ocean, including the detection of change relevant to the ecosystem, is still in its infancy. Indicators integrating information from both research realms are, therefore, only now starting to become available. Recent collaboration between natural and social scientists to examine stakeholder perceptions about the status of small pelagic resources off South Africa (as part of the Knowledge in Fisheries Project [EU/INCODEV KNOWFISH]) indicated convergence between resource users and natural scientists in those perceptions (Fairweather et al. in prep. c). Baseline economic and socio-economic data for South Africa’s fisheries have been compiled and used to provide socio-economic indicators for each (Mather et al. 2003; Sauer et al. 2003, Table 11-3). What threshold levels or turning points can be used to define different ecosystem states? Specifying the threshold levels or turning point indicators that can be used to define different ecosystem states is a complex issue that requires detailed discussion, and only general considerations are presented here. Univariate analysis can be used to determine if, or when, an indicator statistically passes a threshold value that can be assigned according to the variance of the time-series examined (typically a departure of two standard deviations from the mean), but this approach is constrained by problems associated with the length of the time-series, and its distribution function. Obviously, the threshold value must be passed (above or below) for a minimum number of years before suggestions of a change in ecosystem state or a regime shift can be entertained.
Developing a basis for detecting and predicting long-term ecosystem changes
253
Table 11-3. Social and economic indicators relevant to detecting long-term change in the BCLME.
Indicator class and type
Indicator examples
Pressure or Impact
Social 1 - Population density along coast (by area) - Land-use patterns along the coast - Employment in harvesting sector by area and fleet - Dependence of coastal communities on fishing (by area) - Degree of literacy in coastal population (by area) - Lifestyle value (by area) - Cultural value (by area) - Use of goods & services by sector & area, e.g., harvesting effort, beach tourism, ecotourism - Tradition and potential of using artisanal exploitation (spatialised indices as far as possible) - Reservoir of unemployed people for low-qualification fishing employment (e.g. beach seine) Economic 2 - Fleet structure (by gear and area) - Degree of industrialisation of fisheries (by area) - Degree of poverty in coastal population (by area) - Importance of harvests as source of nutrition for resident population: e.g., consumption of (local) marine products per capita in coastal communities; - Importance of harvests as source of income, .e.g, Mean, range and variance of per capita revenue (spatialised as far as possible) - Net economic return for fishery, profit to harvesting sector, - Governmental subside to the fishing sector - Royalties (absolute value or trend) for forein fleets - Price (absolute value or trend) of fishing products at the regional, national or global scale when relevant (e.g. fishmeal, fish oil, frozen products) - Price (absolute value or trend) of soya meal (as a substitute or complement to fishmeal) - Balance between the use of fishmeal versus soya meal in farming (poultry, pork, aquaculture) - Fraction of income generated from eco-labelled marine products and services; - Importance of non-consumptive use as source of income (e.g., leisure activities, tourism, aquaculture) - Structure of non-consumptive use modes (by activity) - Relative importance of mode of ecosystem use, by activity (e.g., generated income) - Resources (personnel & monetary) available for monitoring, control and surveillance of ecosystem use - National budgetary allocation to research (e.g, into the ability to detect change) within the BCLME (by region) - External economic inflows to research (e.g., into the ability to detect change) into the BCLME (by region) Measures to increase ecological sustanability of harvesting1 - Number (and fraction) of fisheries with well-developed management plans, including indicators and references points (or directions); - Fraction of co-management arrangements (of all management arrangements) implemented and successful - Fraction of government officials involved in ocean management trained in conflict resolution and/or change management - Fraction of management arrangements that are considered legitimate among the majority of stakeholders involved in the corresponding arrangement Degree of compliance with management arrangements
Response
Indicators derived to monitor change in - National policies on employment /unemployment management in the fishing sector; - National policies and allocated resources on compliance with rules of resource access and use; - National policies on aquaculture development; - National policies on access rights (open or closed); - National policies on incentives to control and modify overall fishing capacity in the different segments of the fishery (industrial, small-scale, etc.).
1 2
Based on Sowman et al. (2003), Reinikainen and Molloy (2003) and FAO (2003) Based on Mather et al. (2003), and Sauer et al. (2003)
254
A. Jarre et al.
Using absolute threshold values may not be particularly useful given the high variability observed in many biological time-series for the region (e.g. small pelagics; see Figure 8-3 of van der Lingen et al., this volume). Because the definition of a change in ecosystem state will remain largely empirical, it is suggested that multivariate statistical tests or indices be developed, which will allow more confidence to be given to cases where several indicators coincide in their detection of a change in ecosystem state. Time-series of population descriptors of anchovy and sardine in the Benguela LME have been compiled, including data on egg distributions and the seasonal pattern of spawning, larval abundance, recruitment and stock size, condition factor, the contribution of these species to the diet of selected predators, and annual landings (Crawford et al., unpublished manuscript). However, methods to objectively detect turning points that may be indicative of ecosystem changes or regime shifts in the Benguela LME have not been applied to these time-series, which should be reevaluated using one or more of the multivariate techniques for detecting regime shifts reviewed by Mantua (2004). For example, Hare and Mantua (2000) applied Principal Components Analysis (PCA) and the Average Standard Deviates compositing approach to empirical evidence from the North Pacific and identified regime shifts in 1977 and 1989. Whereas those authors reported relative clarity in the regime shift of 1989 as indexed by biological data, they found a lack of clear change expressed by climate indices, leading Mantua (2004) to suggest that biotic and abiotic time-series should be analyzed separately in order to isolate ecosystem behaviour from other influences such as environmental change. Whereas PCA provides an attractive means of investigating regime shifts since it requires no a priori assumptions about candidate years, its major limitations include its inability to identify non-linear relationships between input variables, and the requirement for further time-series analysis using methods such as Intervention Analysis in order to identify statistically significant shifts in the principal components scores (Mantua 2004). PCA and additional time-series analysis should be applied to data from the three sub-systems of the Benguela LME in order to derive new timeseries that can permit the identification of particular ecosystem states (or regimes; see above) and key state variables in each sub system, and also the identification of the time periods during which those key state variables are in transition between different quasi-stable attractors (i.e. regime shifts sensu Mantua 2004; see Table 11-1). What models can be used to help predict ecosystem change? It is likely that many different models will be needed to help detect and predict ecosystem change. Model development should be guided by specific objectives related to different aspects of detecting and forecasting long-term ecosystem changes. Rather than first try to reproduce the real world in a model and then use it to make predictions, appropriate models should be developed to address specific questions, and the results should be combined in a sensible way. In this section we describe different kinds of models that address different components of the ecosystem at different temporal and spatial scales.
Developing a basis for detecting and predicting long-term ecosystem changes
255
3D hydrodynamic models forced by realistic winds, solar radiation and boundary conditions can be used for now-casting or forecasting of environmental events that might exert bottom-up controls on ecosystems, especially if they incorporate data assimilation (e.g. Chen et al. 2004). The same models forced by scenarios of longterm climatological changes, such as those predicted in relation to global warming, can be used for the “what-if” type of forecasting sensu Woods (this volume). The major challenge here is to get realistic scenarios at large temporal and spatial scales. Conflicting views can result from the use of different models (Fréon et al., this volume). Furthermore, it is difficult to disentangle the interactions between the recent global trend of warming and natural inter-decadal climatic oscillations. Nonetheless, these scenarios can be useful to indicate the expected ranges of magnitude in sea temperature and currents and, by elimination, exclude unlikely situations. Marine plankton can integrate meteorological variability, and because of their environmental sensitivity, short life cycles and inability to escape their environment, they make excellent indicators of environmental change and are invaluable in the mapping of the environmental consequences of climate change in the marine environment (Reid et al. 1998). Single or multi-species plankton stage-resolved and spatially-explicit models are seldom used but could help to predict the effect of large climatic changes on the plankton community and to improve the parameterisation of biogeochemical models. However, a major difficulty is to simulate and predict behavioural changes in feeding and vertical migration. Biogeochemical models (e.g. NPZD models), especially those with enough compartments to distinguish short from long plankton trophic chains, can be used for “what-if” forecasting. The present state of the art allows satisfactory simulation of observed phytoplankton abundance and distribution but is not yet sufficiently evolved to fully reproduce the complexity of zooplankton spatial and temporal dynamics. Therefore one cannot expect too much precision in zooplankton long term forecasting. Conventional single species models for fish stock assessment can provide estimates of change in abundance according to different exploitation levels, but these models do not incorporate the effects of changes in the abundance of prey and predator of the considered species, and they usually assume constancy in population parameters. Multispecies models of population dynamics incorporating trophic relationships allow for variability in population parameters, but are usually constrained by data availability and their results have increased uncertainty. Age-structured or surplus production models incorporating an environmental variable can be cautiously used for short-term prediction, especially for low trophic level species, but are currently unable to take into account interdecadal changes such as the alternation between sardine and anchovy. Because the processes driving these interdecadal changes are not understood, only empirical models can be used at present (Klyashtorin 2001; Fréon et al. 2005a). Fishery GIS can be used to simulate changes in fish distribution and spawning area according to predictions of environmental changes and knowledge of habitat preferences (e.g. temperature). The difficulty here is to describe adequately fish habitat according to realistic proxies of forcing factors. Finally, trophic box models like EwE (Pauly et al. 2000) or dynamic and spatially-resolved individual-based
256
A. Jarre et al.
models like Osmose (Shin and Cury 2001) can help to predict the effect of drastic exploitation or climate change on the structure and functioning of ecosystems (Shannon et al. 2003 a, b, Shin et al. 2004). All these models have a role to play in understanding the dynamic processes in an ecosystem. Some of the models have an empirical statistical basis, whereas others are based on “first principles” (sensu Schneider 1992). The empirical statistical models depend on historical data, and can produce predictions with estimates of probability. However, the models are generally limited to predicting scenarios that have been observed in the past, or that do not depend on new processes. They are probably most useful for short-term predictions. Models that are based on first principles use equations and relationships that represent the main processes, and are well suited to what-if predictions. However, for these models it is difficult to validate their results, and they are probably most useful for identifying possible ecosystem states, for eliminating unlikely ones, and for identifying potential indicators of change. Ideally, we would hope to combine the results of all models in a structured and logical fashion, to make best use of available data and untested hypotheses. In conclusion, no real ecosystem model can be used for prediction yet. There are possibilities of models being adaptable (e.g. growth rate controlled by temperature) but this often does not include all effects and interactions. The bottlenecks are our poor understanding of how systems function and how species will adapt to drastic changes in their habitat. Future advances in these fields will allow better parameterisation or changes in assumptions; modelling of long term changes must be viewed as a dynamic process. However, it is also likely that the more we learn, the more we realise we do not know! Most of the models reviewed in this section are already available for the Benguela region (Table 11-4) and form part of the “ecoscope” toolbox (Shannon et al. 2004b; CD this volume). At this stage we suggest that it would be productive to develop synthesis tools to make best use of the available modelling expertise, taking into account different degrees of uncertainty in the models. It is expected that conflicting outcomes will emerge from different tools, but not only does one learn from the models, one also learns from the model errors – at least as much. DESIRED END PRODUCTS AND DATA REQUIREMENTS Integrated data management and communication sub-systems are seen as the primary integrator of ocean observing systems, "the 'life-blood' of the system that links all of its components" (IOC/GOOS 2003: p. 102). To detect and monitor ecosystem changes in the Benguela LME region in the long term, we propose that integrating tools be developed for the Benguela LME that allow the interpretation and synthesis of different ecosystem indicators. Past experience indicates that knowledge bases (sensu Starfield and Louw 1986) that are formalised within “rule-based models” (Starfield and Louw 1986; Ferrar 1986; Liao 2005) are well suited to such an application. Rulebased models have been constructed to predict recruitment strength of anchovy in the southern Benguela (e.g., Korrûbel et al. 1998; Miller and Field 2002, see CD (this
Developing a basis for detecting and predicting long-term ecosystem changes
257
Table 11-4. Examples of models available for sub-regions in the BCLME than can be used for scenario exploration in a comparative fashion
Class of model
Physical-Biogeochemical 3D hydrodynamic NPZ(&D) Zooplankton population models
Environmental Processes / Bakun's triad IBM
Benguela implementation
Reference
PLUME, RIGA and SAFE NPZD and N2P2Z2D2 No zooplankton models being applied yet at an ecosystem scale Enrichment and retention SB
Penven et al. (2001) Koné et al. (submitted) Moloney (pers. comm.)
Anchovy recruitment SB
Mullon et al. (2002, 2003), Parada et al. (2003, submitted), Huggett et al. (2003) Miller et al. (in press)
Sardine recruitment SB
Process-based multispecies Size-based ecosystem OSMOSE Trophic ecosystem EwE
Analytical empirical* Fish stock assessment models, single species
Multispecies
Bayesian assessments, agestructured production models, virtual population analyses.
Minimum realistic models
Lett et al. (in press)
Shannon et al. (2003 a,b), Shannon et al. (2004a)
Cunningham & Butterworth (2004 a,b), Johnston & Butterworth (2004), Rademeyer & Butterworth (2004). Punt & Butterworth (1995)
* In contrast to process-based multispecies and ecosystem models, analytical models do not model predator dynamics in their own right, and consequently, predation is only used as a forcing fuction for modelling prey dynamics. It is generally recognised that results such of analytical models often form the basis for algorithms and parameters used in process-based multispecies models (e.g., Whipple et al. 2000).
_____________________________________________________________________ volume) for an example application of this approach using the software of Quadling and Quadling 1995). We argue that the flexibility of this approach has not been exhausted, and that it can be applied profitably to the detection of long-term change in the Benguela LME. Because a multitude of information sources needs to be considered, this approach can facilitate assessment of whether ecosystem change is taking place in a logical, defendable and transparent way. Expert systems typically contain a high level of expertise in a form that makes it accessible. The expert system models should provide an effective means of communication between scientists and end users (Starfield and Louw 1986), and they
258
A. Jarre et al.
have the potential to inform management groups within the region about the current and possible future state of the ecosystem in a consistent fashion. Such knowledge would be useful for resource management, for long term planning within different fisheries sectors, and for environmental managers. In addition, because expert system models capture expertise from a variety of specialists, they provide an important interdisciplinary information source for local, regional and international scientific communities, including academics, practitioners, decision-makers and students. Decision support tools in general, and expert systems in particular, have evolved considerably during the past two decades (Guimarães-Pereira et al. 2005). The problems of integrating disparate kinds and sources of information are encountered in many arenas. For example, Roetter et al. (2005) describe a system for land use planning in Asia, Power and Bahri (2005) describe an improved system for coordinating operational tasks in industrial plants, and Guimarães Pereira et al. (2005) show how an innovative information tool is applied to a groundwater governance issue in France. This last study emphasized that knowledge tools are useful for initiating and informing debates, rather than simply for legitimising decisions. What is an expert system? Expert systems capture and organise knowledge in a database (Starfield and Louw 1986). They provide a formal means of synthesis, as opposed to analysis, and they provide an operational language (IF-THEN rules) that is equivalent to mathematics as the language of analysis (Starfield and Louw 1986). In the context of ecosystem change, rule-based models can synthesize different ecosystem indicators so that, as a group, the indicators are interpreted effectively and consistently to identify the probability of long-term ecosystem change. A simplified draft template is shown in Figure 11-3 to illustrate the approach. A variety of models and observations is used by experts to produce ecosystem indicators. Each indicator typically and individually might suggest one of six (for example) possible states of the ecosystem: • No indication of long-term change, current state neither identified as good or bad; • No indication of long-term change, current state good; • No indication of long-term change, current state bad; • Indication of long-term change occurring, direction of change neither identified as good or bad; • Indication of long-term change occurring, direction of change identified as good; • Indication of long-term change occurring, direction of change identified as bad. In identifying possible states, the assessment of whether each is good, bad or neutral requires both ecological and socio-economic criteria. The expert system therefore provides a vehicle for multi-disciplinary integration. Individual indicators are likely to focus only on aspects of the ecosystem. The interpretation of some indicators may be ambiguous (e.g. sardine is increasing in abundance while anchovy is decreasing),
Developing a basis for detecting and predicting long-term ecosystem changes
259
whereas others may give clear signals (e.g. jellyfish appear to have increased in abundance by two orders of magnitude). It will be necessary to garner expert opinion on the interpretation of the indicators when viewed as a group, and at this stage the rules are constructed (Figure 11-3), and expertise is captured within the expert system. While straightforward in structure and seemingly "simple" (see also van der Lingen et al. this volume), expert systems in practice rapidly acquire a degree of complexity that underlines the usefulness of such a formal approach in decision support.
Figure 11-3. Outline of the expert system approach recommended for integrating the information of various indicators of long-term change in the entire BCLME or its sub-regions. Note that the development and calculation of indicators resides in the scientific domain only, whereas stakeholder expertise should be part of the process of definition of rules underlying the expert system model.
260
A. Jarre et al.
Developing and maintaining an expert system for detecting and predicting ecosystem change will be a dynamic process. In the initial, pre-operational stages of development, the focus will be in the scientific arena (Figure 11-3), where the system is tested using what-if scenarios and modified. Guimarães-Pereira et al. (2005) pointed out that broad participation of all stakeholders in designing and exercising the knowledge-based system corresponded to a peer-review process, providing elements of quality assurance. The system can then be moved into a pilot phase, where two sequential routes for the end-product (the “traffic light” signal) are suggested (Figure 11-3). The first route is from the “traffic light” to the research community. The purpose of this step is to alert scientists to possible change that might be occurring, or to inconsistencies in the results, where indicators are providing contradictory signals. This might provide guidance for future short-term research and/or monitoring (e.g. carry out more intensive sampling in specified regions), or indicate the need to check models and even the expert system as necessary. If the signal is incorrect, the expert system should be updated and refined (expert systems should “learn” through experience), with detailed documentation. If the signal is correct, the second route is followed, where the end product (an indication of change or no-change) is directed at management groups, to help inform them in the decision-making process. Requirements of expert systems At present, many research groups within the Benguela LME are producing indicators as part of their normal activities. For the medium term, we can probably assume that existing data will continue to be needed and will also be available in the future. However, there are currently some important data gaps, and these should be discussed and prioritized in the short term. Some examples of data gaps include comprehensive measurements of sub-surface variables (e.g. changes in depth of the thermocline), primary production (from field measurements and remote sensing), and integrative variables (over large spatial areas, from remote sensing and towed undulator technology). For an expert system to be developed and tested, time series of indicators are required. Because these time series are likely to be short relative to the time scale being considered, it would be useful also to use a comparative approach, where data are standardized across sub-systems and scales, allowing comparisons among different sub-systems within the Benguela LME region, and other LMEs. The major inputs to the expert system are the various indicators. At present, many of these are produced without qualifiers or errors. The outputs of the expert system will depend on a weighted assessment of the inputs (the indicators), and this assessment should be informed by the skill of the indicators in providing reliable values. For some indicators the skills level can be represented by confidence intervals and error estimates, whereas others might require more qualitative reliability scores. This is an area of research that needs to be tackled in the short term.
Developing a basis for detecting and predicting long-term ecosystem changes
261
What are the appropriate time and space scales for data and predictions? The expert system will serve as an early warning system for long-term changes including, but not restricted to, regime shifts. The key characteristic of a regime shift is that the time-scale for the change between states is much shorter than the time within alternate states. This pragmatic definition can be applied, or tested, by measuring the rate of change of time-series (de Young et al 2004), and this will provide guidance for the interpretation of different indicators in developing the expert system. For the end-products of the expert system, the time-scales for depicting ecosystem states are likely to be of the order of a decade or longer, and spatial scales probably also will be large (sensu IOC/GOOS 2003), probably incorporating all three main sub-systems (off South Africa, Namibia and Angola). Other levels of ecosystem organisation might also be considered, such as benthos versus pelagial, inshore versus offshore, and coastal gradients. In general, data will be required on all time scales up to annual and possibly longer, depending on the indicators that are used. The time scales for end products should be annual, although this would need to include the recognition that trends are being analysed. To develop and maintain an operational system for detecting and predicting ecosystem change, organisational structure and infrastructure are required (IOC/GOOS 2003). Of great importance for a sustainable system is the need to improve data management in the Benguela LME region, including systems for quality assurance and quality control, and good communication (IOC/GOOS 2003). Previous experience in the North Atlantic is that an optimum staff complement is needed to ensure effective database design and maintenance, and for timely provision and analysis of national and regional data (e.g., ICES 1999, OSPAR 2000). There is an urgent short-term need to address data management and communication issues within the Benguela LME region, and to foster strong institutional partnerships that will facilitate this. SCHEDULE FOR IMPLEMENTATION Environmental changes related to regime shifts are not yet taken into account in fishery management (Sinclair and Valdimarsson 2003, ICES 2004, Rothschild and Shannon 2004). Detecting and predicting changes and finding ways of incorporating this information into fishery management advice are highly desirable, particularly in the light of the “Reykjavik Declaration on Responsible Fisheries in the Marine Ecosystem”, to which Angola, Namibia and South Africa are signatory and thus obliged to ensure that the declaration is upheld and fully considered in the work of their fisheries scientists, in the functioning of their fishing industries and in their respective fisheries management approaches. Further, the targets agreed upon at the World Summit for Sustainable Development (WSSD), held in Johannesburg in 2002, include the following undertakings: •
“Encourage the application by 2010 of the ecosystem approach, noting the Reykjavik Declaration on Responsible Fisheries in the Marine Ecosystem and decision 5/6 of the Conference of Parties to the Convention on Biological Diversity” (WSSD Chapter 4, paragraph 29 d)
262
A. Jarre et al. •
“Develop and facilitate the use of diverse approaches and tools, including the ecosystem approach, the elimination of destructive fishing practices, the establishment of marine protected areas …. and time/area closures for the protection of nursery grounds and periods…” (WSSD Chapter 4, paragraph 31 c)
What are the priorities for detecting and forecasting change? In the Benguela LME region as a whole, high priority for detecting changes will be given to issues that require immediate management attention. In addition, due to the fact that this is a developing region, issues with social and economic implications (e.g. food supply and employment) generally will be given first priority. Several marine taxa endemic or migrating to the Benguela LME region are vulnerable or endangered and could be affected drastically by ecosystem changes. Many species of seabirds and some cetaceans are highly sensitive to changes in abundance and distribution of their prey (small pelagic fish in particular) as well as being affected by industrial fisheries through incidental catches. In order to comply with international agreements, countries in the region might have to implement immediate remedial actions to improve and maintain the conservation status of those species. Due to different ecological and socio-economic contexts, it is expected that priorities will differ between the different sub-regions of the Benguela LME. In the southern Benguela, large fluctuations in abundance, species alternations and shifts in distributions of small pelagics (sardine and anchovy) have had significant economic impacts on the fisheries sector as well as important ecological implications (Cury and Shannon 2004, Shannon et al. 2003b) and constitute probably the highest priority there. Ecosystem changes in the inshore communities involving abalone, kelp, west coast rock lobster and sea urchins (Cochrane et al. 2004) are also of high priority due to the high market value of the products of the fisheries involved and their impacts on the local communities. In the northern Benguela, the collapse of the small pelagic stocks (sardine in particular) in the early 1970s and mid 1990s has had profound ecological (e.g., Cury and Shannon 2004) and economic (Armstrong and Thomas 1995) implications. The reasons for the lack of recovery of the sardine stock are still not completely understood (van der Lingen et al. this volume) and warrant the highest priority in Namibia in order to attempt rebuilding this stock and restoring the degraded pelagic ecosystem. Forecasting environmental anomalies (Benguela Niño and low oxygen events) and detecting longer term ecological changes are important as they may have long term impacts on the whole ecosystem and the fisheries and require timely management mitigating actions (Roux 2003, Roux and Shannon 2004, van der Plas et al. this volume). In Angola, the ecosystem effects of habitat degradation (effect of fishing gear on benthic habitats and degradation of mangroves in particular) have been highlighted as
Developing a basis for detecting and predicting long-term ecosystem changes
263
a priority due to their potential impacts on some key commercial resources (shrimps) as well as the sustainability of the multispecies demersal fishery important for local food supply (Cochrane 2004). Ecosystem considerations linked to the horse mackerel industrial fishery and the depletion of the stock, as well as changes in abundance and distribution of Sardinella sp. are also important both in their ecological and socioeconomic implications. What practical steps can be taken? What are realistic time frames for implementation? Short term (1-3 years) There is a need for improved understanding of current ecosystem states in the subregions of the Benguela LME, and evaluation of the extent to which they are ecologically, economically and socially desirable. The capability to predict (or hypothesize) whether and how long-term changes occur would be a giant leap forward in being able to manage fisheries in an ecosystem context. A necessary step would be adaptation and development of multivariate statistical tools for the analysis of available time series (catch, spawner biomass, recruitment, egg and larval abundances, fish condition factor, proportion of pelagic fish in predators’ diets, etc.), and this is seen as an important area of study. A proposed starting point is the time series collated and examined during a workshop to identify “turning points” in the Benguela ecosystem (Alheit et al. 2001, Crawford et al. unpublished MS). We recommend that these data be revisited with preliminary statistical analyses of catch and biomass series, and with an analysis using a methodology that has been developed more recently (see above section, "What threshold levels and turning points can be used to define different ecosystem states?"). Generic indicators for detecting and monitoring ecosystem changes can be indentified through a comparative approach to establish which indicators are likely to be the most sensitive to detecting ecosystem changes across a range of possible ecosystem states and driving forces. By comparison with other ecosystems, an attempt to identify indicators or early warning signals of ecosystem change would be valuable. For example, early warning indicators of small pelagic fish stock collapses have been proposed as a high priority for management of the South African pelagic fisheries. Single species indicators should be examined in conjunction with ecosystem/integrated indicators and environmental indicators, and the most appropriate set of indicators should be selected for each fishery. It would be important to ensure that these indicator sets are regularly updated to inform management and to be used by the proposed expert system (see medium-term actions below). Establishing reliable, taxon-specific, spatially extensive and area-explicit estimates of primary production for developing long time series are important short-term targets for understanding ecosystem changes in the Benguela LME region. Backward projections of existing time-series (likely to extend into medium-term activities) enable quantification of variability of bottom-up forcing over short periods (years, e.g., Carr
264
A. Jarre et al.
2002, Demarcq et al. 2002). Trophic models can be used to estimate what should have been required to generate the observed dynamic fluctuations within the southern and northern Benguela ecosystems (e.g. Shannon et al. 2004a), but quantitative estimates of primary production over the whole Benguela LME region (including off Angola) still require refinement and ground-truthing, and longer time-series are needed. Another major “gap” would be filled by estimates of the biomass of gelatinous zooplankton in the northern Benguela. Trophic impacts of gelatinous zooplankton on predators and prey need quantification to assess the trophic roles of gelatinous zooplankton in the northern versus southern Benguela ecosystems. Results might be expected in the short term, with refinements in the medium term. Medium term (4-7 years) An important medium-term target is the development of methodology to quantify uncertainty related to both inputs and outputs of ecosystem models (e.g., confidence intervals). Other practical steps should include completion of (or new) analyses of sediment deposits (fish scales, plankton) and linkage of these records to historical ocean climate, and possibly investigation of the potential value of ecosystem modifications (importantly, with necessary caveats: see discussion at the end of section “What changes in the ecosystem might be caused by fishing, pollution, environment, or climate?”). There is also a need for evaluation of the usefulness of Marine Protected Areas (MPAs) as control systems for purposes of comparison with fished/uncontrolled ecosystems subject to the same environmental effects. This could assist in distinguishing the effects and/or driving forces for ecosystem change exerted by anthropogenic versus natural (environmental, biological) processes. In addition, MPAs provide opportunities for the validation of community indicators (Trenkel and Rochet 2003), so that the some of the sets of ecosystem indicators identified (see shortterm activities) could be validated and their effectiveness at capturing ecosystem changes tested. A high-priority, over-riding activity would be the development and implementation of the proposed expert system of ecosystem indicators to monitor ecosystem state, identify ecosystem changes, evaluate the effectiveness of adopted management strategies and their underlying strategic objectives and, where appropriate, identify/recommend actions to be taken.
CONCLUSIONS 1. The detection and prediction of ecosystem states and changes is central to ocean observation programmes globally, and one of the key policy actions of the BCLME Programme.
Developing a basis for detecting and predicting long-term ecosystem changes
265
2. We highlight a suite of approaches to detect and monitor change in the long-term in one of several of the Benguela LME sub-regions. We elaborate on the many ways in which ecosystem changes may be measured and modelled. Suites of composite indicators, rather than single variables, appear to be the most useful for depicting ecosystem-level attributes. 3. We suggest an expert system of ecosystem indicators as a general and feasible methodology (i) to synthesize the results of these different approaches, (ii) that should be applicable to any sub-system of the Bengulea LME region, and (iii) which will support long-term ecosystem considerations in the management of human activities in the Benguela LME. 4. Priorities for detecting and predicting long-term ecosystem change need to be given to issues that require immediate management attention in the different sub-regions of the Benguela LME. These issues include fluctuations in the abundance and shifts in the distributions of small pelagic species (South Africa and Namibia), rebuilding of collapsed stocks (Namibia), effects of habitat degradation (Angola), ecosystem change in marine inshore communities (South Africa), and population sizes of vulnerable and endangered species (South Africa and Namibia). 5. Important practical steps towards a basis for detecting and predicting long-term ecosystem changes are highlighted. In the short term (1-3 years), important steps are the re-analysis of existing time series with up-to-date multivariate tools, the identification of early warning indicators, the improved understanding of patterns of and fluctuations in primary production in the region, and the improved understanding of the abundance and ecological role of gelatinous zooplankton. 6. Important practical steps in the medium term (4-7 years) include the development of methodology to quantify uncertainty related to inputs and outputs of ecosystem models, analyses of sediment deposits and linkage of these to historic ocean climate, and the evaluation of the usefulness of Marine Protected Areas. The development and implementation of the proposed expert system of ecosystem indicators is seen to be a pratical step to be awarded high priority. ACKNOWLEDGEMENTS This chapter was drafted at the Specialist Session C, “Detecting and forecasting longterm ecosystem changes” of the Benguela Forecasting Workshop, 8-11 November 2004, in Cape Town. The Session was convened by Dr. C. Moloney (UCT), and cochaired by Drs. C. Moloney and H. Verheye (MCM). Dr. A. Jarre (DIFRES) acted as session rapporteur. All authors gratefully acknowledge the contributions during the workshop session, and comments on the draft manuscript, by (in alphabetical order) Dr. J. Alheit (IOW), Prof. D. Butterworth (UCT), Dr. A. Cockcroft (MCM), Mr. H. Demarcq (IRD), Mr. L. Drapeau (IRD), Dr. H. Hamukuaya (BCLME), Dr. M. Kasu (INAMET), Dr. S. Kifani (INRH), Dr. A. Kreiner (MFMR-NatMirc), Dr. R. Leslie (MCM), Dr. T. Malone (OCEAN), Mr. S. Neira (UCT), Dr. E. Plagányi (UCT), Mr. S.
266
A. Jarre et al.
Shikongo (MET), Dr. K. Sherman (NOAA), Dr. A. Da Silva (IIM), Dr. Y. Spitz (OSU), Dr. E. Urban (SCOR), and Prof. J. Woods (ICL). We thank two anonymous reviewers for helpful comments on the draft of the manuscript. This is a contribution of the IDYLE and ECO-UP joint programmes between SA and IRD-France ; and by EUR-OCEANS, a European Network of Excellence under the CEC 6th Framework Programme (Contract Ref. 511106). REFERENCES Agenbag, J.J., A.J. Richardson, H. Demarcq, P. Fréon, S. Weeks and F.A. Shillington. 2003. Estimating environmental preferences of South African pelagic fish species using catch size and remote sensing data. Prog. Oceanogr. 59:275–300. Alheit, J., L.J. Shannon and R.J.M Crawford. 2001. Workshop of the Decadal Changes WG of SPACC on “Major Turning Points in the Structure and Functioning of the Benguela Ecosystem,” GLOBEC International Newsletter 7(1): 20-21. Armstrong, M.J. and R.M. Thomas. 1995. Clupeoids. 104-121 (Chapter 11) in A.I.L Payne and R.J.M. Crawford, eds. Oceans of Life off Southern Africa. Halfway House, Vlaeberg, South Africa. 380 pp. Bakun, A. 1990. Global climate change and the intensification of coastal upwelling. Science 247 (4939): 198-201. Bakun, A. 2004. Regime shifts, Chapter 25 In A.R. Robinson and K. Brink, eds. The Sea (Volume 13). Harvard University Press, Cambridge, Massachusetts, USA. Bakun, A. and P. Cury. 1999. The “school trap”: a mechanism promoting large-amplitude out-of-phase population oscillations of small pelagic species. Ecology Letters 2: 349-351. Barange, M., I. Hampton and B.A. Roel. 1999. Trends in the abundance and distribution of anchovy and sardine on the South African continental shelf in the 1990s, deduced from acoustic surveys. S. Afr. J. mar. Sci. 19:367-391. Beamish, R.J. and C. Mahnken. 1999. Taking the next step in fisheries management. 1-21 in Ecosystem Approaches for Fisheries Management. University of Alaska Sea Grant, AK-SG-99-01, Fairbanks, Alaska, USA. Beaugrand, G., F. Ibanez, J.A. Lindley and M. Edwards. 2002. Reorganization of north Atlantic marine copepod biodiversity and climate. Science 296:1692-1694. Beaugrand, G., K.M. Brander, J.A. Lindley, S. Souissi and P.C. Reid. 2003. Plankton effect on cod recruitment in the North Sea. Nature 426:661-664. Beverton, R.J.H. 1990. Small marine pelagic fish and the threat of fishing; are they endangered? Journal of Fish Biology 37 (supplement A): 5-16. Brander, K.M., R.R. Dickson and E. Edwards. 2003. Use of Continuous Plankton Recorder information in support of marine management: Application in fisheries, environmental protection, and in the study of ecosystem response to environmental change. Progress in Oceanography 58:175-191. Caddy, J.F. and L. Garibaldi. 2000. Apparent changes in the trophic composition of world marine harvests: The perspective from the FAO capture database. Ocean and Coastal Management 43:615655. Caddy, J.F. and P. Rodhouse. 1998. Cephalopod and groundfish landings: evidence for ecological change in global fisheries? Rev. Fish Biol. Fisheries 8:431-444. Carr, M.-E. 2002. Estimation of potential productivity in Eastern Boundary Currents using remote sensing. Deep-Sea Research II 49(1-3): 59-80. Carr, M.-E. and E.J. Kearns. 2003. Production regimes in four Eastern Boundary Current systems. DeepSea Research II 50: 3199 –3221. Chen D, M.A. Cane, A. Kaplan, S.E. Zebiak and D. Huang. 2004. Predictability of El Niño over the past 148 years. Nature 428:733-736. Cochrane, K.L. 2004. Ecosystem Approaches for Fisheries (EAF) Management in the BCLME. Report of the first regional workshop, 21-24 Sept. 2004, Windhoek, Namibia. BCLME. Cochrane, K.L., C.J. Augustyn, A.C. Cockcroft, J.H.M. David, M.H. Griffiths, J.C. Groeneveld, M.R. Lipinski, M.J. Smale, C.D. Smith and R.J.Q. Tarr. 2004. An ecosystem approach to fisheries in the southern Benguela context. Afr. J. mar. Sci. 26:9-35.
Developing a basis for detecting and predicting long-term ecosystem changes
267
Collie, J.R., K. Richardson and J. Steele. 2004a. Regime shifts: Can ecological theory illuminate these mechanisms? Progr. Oceanogr. 60:281-302. Collie, J.R., K. Richardson, J. Steele and L.T. Mouritsen. 2004b. Physical forcing and ecological feedbacks in marine regime shifts. ICES C.M. 2004/M:06. 26p. Crawford, R.J.M., R.A. Cruickshank, P.A. Shelton and I. Kruger. 1985. Partitioning of a goby resource amongst four avian predators and evidence of altered trophic flow in the pelagic community of an intense, perennial upwelling system. S. Afr. J. mar. Sci. 3:215-228. Crawford, R.J.M., L.J. Shannon, A. Kreiner, C.D. van der Lingen, J. Alheit, A. Bakun, D. Boyer, P. Cury, T. Dunne, M.-H. Durand, J.G. Field, P. Fréon, M.H. Griffiths, E. Hagen, J. Hutchings, E. Klingelhoeffer, C.L. Moloney, D. Mouton, C. Roy, J-P. Roux, L.V. Shannon, F.A. Shillington, L.G. Underhill and H.M. Verheye. Periods of major change in the structure and functioning of the pelagic component of the Benguela ecosystem, 1950-2000. Unpublished manuscript. Cunningham, C.L and D.S. Butterworth. 2004a. Base case Bayesian assessment of the South African anchovy resource. Unpublished Report. Marine and Coastal Management, Cape Town, South Africa WG/PEL/APR04/01. 19 pp. Cunningham, C.L. and D.S. Butterworth. 2004b. Base case Bayesian assessment of the South African sardine resource. Unpublished Report. Marine and Coastal Management, Cape Town, South Africa WG/PEL/APR04/02. 22 pp. Cury, P.M., L.J. Shannon and Y-J. Shin. 2003. The functioning of marine ecosystems: A fisheries perspective. 103-123 in Sinclair, M., and G. Valdimarsson. Responsible Fisheries in the Marine Ecosystem. FAO and CABI Publishing, UK. 426pp. Cury, P. and L.J. Shannon. 2004. Regime shifts in upwelling ecosystems: Observed changes and possible mechanisms in the northern and southern Benguela. Progress in Oceanography 60(2-4): 223-243. Cury P. and V. Christensen. 2005. Quantitative ecosystem indicators for fisheries management: an introduction. ICES Journal of Marine Science 62:307-310. Cury, P., L.J. Shannon, G.M. Daskalov, J-P. Roux, A. Jarre, C.L. Moloney and D. Pauly. 2005a. Trophodynamic indicators for an ecosystem apporach to fisheries. ICES Journal of marine Science 62:430-442. Cury, P., C. Mullon, S.M. Garcia and L.J. Shannon. 2005b. Viability theory for an ecosystem approach to fisheries. ICES Journal of Marine Science 62(3): 577-584. Daskalov, G.M., D.C. Boyer and J-P. Roux. 2003. Relating sardine Sardinops sagax abundance to environmental indices in northern Benguela. Prog. Oceanogr. 59:257–274. Degnbol, P. and A. Jarre. 2004. Indicators in fisheries management: A development perspective. Afr. J. mar. Sci. 26:303-326. Demarcq, H., B. Mitchell-Innes and C. van der Lingen. 2002. Developing satellite derived indices of ecosystem productivity at scales relevant to pelagic fish, 2002 SAMSS conference, Swakopmund, Namibia. Abstract only. Demarcq, H., R. Barlow and F.A. Shillington. 2003. Climatology and variability of sea surface temperature and surface chlorophyll in the Benguela and Agulhas ecosystems as observed by satellite imagery. Afr. J. mar Sci. 25:363-372. De Oliveira, J.A.A., D.S. Butterworth, B.A. Roel, K.L. Cochrane and J.P. Brown. 1998. The application of a management procedure to regulate the directed and bycatch fishery of South African sardine (Sardinops sagax). S. Afr. J. mar. Sci. 19:449-469. de Young, B., R. Harris, J. Alheit, G. Beaugrand, N. Mantua and L. Shannon. 2004. Detecting regime shifts in the ocean: Data considerations. Progress in Oceanography 60:143-164. Drapeau, L., L. Pecquerie, P. Fréon and L. Shannon. 2004. Quantification and representation of potential spatial interactions in the southern Benguela ecosystem. Afr. J. mar. Sci. 26:141-159. Fairweather, T.P., C.D. van der Lingen, L. Drapeau, A.J. Booth and J.J. van der Westhuizen. In prep.a. Indicators of sustainable fishing for the South African sardine (Sardinops sagax) and management implications. Fairweather, T.P., C.D. van der Lingen, and J.J. van der Westhuizen. In prep.b. Indicators of sustainable fishing for the South African anchovy (Engraulis encrasicolus) and management implications. Fairweather, T.P., M. Hara, C.D. van der Lingen, J. Raakjær Nielsen, L. Shannon, G.G. Louw, P. Degnbol and R.J.M. Crawford. In prep.c. The knowledge base for management of the capital-intensive fishery for small pelagic fish off South Africa. Book Chapter in Knowledge in Fisheries Management.
268
A. Jarre et al.
FAO. 1999. Indicators for sustainable development of marine capture fisheries. FAO Technical Guidelines for Responsible Fisheries (8), Food and Agricultural Organisation of the United Nations. Rome, Italy.68 pp. FAO. 2003. The ecosystem approach to fisheries. FAO Technical Guidelines for Responsible Fisheries. (4, Suppl. 2), Food and Agricultural Organisation of the United Nations. Rome, Italy. 127 pp. Ferrar, T. 1986. Expert systems for resource management in southern Africa. South African Journal of Science 82: 555. Fréon, P., and O.A. Misund. 1999. Dynamics of pelagic fish distributino and behaviour: Effect on fisheries and stock assessment. Blackwell, London, UK. 348 pp. Fréon, P., P. Cury, L. Shannon and C. Roy. 2005a. Sustainable exploitation of small pelagic fish stocks challenged by environmental and ecosystem changes. Bulletin of Marine Science 76(2): 385-462. . Fréon, P., L. Drapeau, J.H.M. David, A. Fernández Moreno, R.W. Leslie, W.H. Oosthuizen, L. Shannon and C.D. van der Lingen. 2005. Spatialized ecosystem indicators in the Southern Benguela. ICES Journal of Marine Science 62(3): 459-468. Fréon, P., J. Alheit, E.D. Barton, S. Kifani, P. Marchesiello. 2006. Modelling, forecasting, and scenarios in comparable upwelling ecosystems: California, Canary and Humboldt (this volume). Fulton, E.A., A.D.M. Smith, A.E. Punt. 2005. Which ecological indicators can robustly detect effects of fishing? ICES J Mar.Sci. 62:540-551. Gründlingh, M. P.D. Morant, R.C. van Ballegooyen, A. Badenhorst, E. Goemes, L. Greyling, J. Guddal, I.T. Hunter, D.W. Japp, L. Maartens, K.R. Peard, G.G. Smith, and C.K. Wainman. 2006. Maritime operations in the Benguela coastal ocean (Chapter 15, this volume). Guimarães Pereira, Â., S. Corral Quitana , and S. Funtowicz. 2005. GOUVERNe: new trends in decision support for groundwater governance issues. Environmental Modelling & Software 20:111-118. Hardman-Mountford, N.J., A.J. Richardson, J. Agenbag, E. Hagen, L. Nykjaer, F.A. Shillington and C. Villacastin. 2003. Ocean climate of the South East Atlantic observed from satellite data and wind models. Progress in Oceanography 59(2-3): 181-221. Hare, S.R. and N. Mantua. 2000. Empirical evidence for North Pacific regime shifts in 1977 and 1989. Progr. Oceanogr. 47:103-145. Huggett J., P. Fréon, C. Mullon and P. Penven. 2003. Modelling the transport success of anchovy Engraulis encrasicolus eggs and larvae in the southern Benguela: The effect of spatio-temporal spawning patterns. Mar. Ecol. Progr. Ser. 250:247-262. Hutchings, L., L.E. Beckley, M.H. Griffiths, M.J. Roberts, S. Sundby and C.D. van der Lingen. 2002. Spawning on the edge: Spawning grounds and nursery areas around the southern African coastline. Mar. Freshwat. Res. 53:307-318. Hutchings, L., H.M. Verheye, J.A. Huggett, H. Demarcq, R. Cloete, R.G. Barlow, D. Louw, A. da Silva. 2006. Variability of plankton in the Benguela Current Large Marine Ecosystem: An overview. (Chapter 6, this volume). Hutchings, L., M. Barange, S.F. Bloomer, A.J. Boyd, R.J.M. Crawford, J.A. Huggett, M. Kerstan, J.L. Korrûbel, J.A.A. de Oliveira, S.J. Painting, A.J. Richardson, L.J. Shannon, F.H. Schülein, C.D. van der Lingen and H.M. Verheye. 1998. Multiple factors affecting South African anchovy recruitment in the spawning, transport and nursery areas. S. Afr. J. mar. Sci. 19:211-225. ICES. 1999. Data handling, p. 128-134. In Report of the ICES Advisory Committeee on the Marine Environment 1999, ICES Cooperative Research Report No. 239. 277 p. ICES. 2004. Report of the ICES Advisory Committee on Fishery Management and Advisory Committee on Ecosystems 2004. ICES Advice, Volume 1, Number 2, 1544 pp. Available at http://www.ices.dk/products/icesadvice.asp. IOC/GOOS. 2003. The Integrated Strategic Design Plan for the Coastal Ocean Observations Module of the Global Ocean Observing System, GOOS Report No. 125; IOC Information Documents Series No. 1183. UNESCO, Paris, France. Available at http://ioc.unesco.org/goos/docs/doclist.htm. Isaacs, J.D. 1976. Some ideas and frustrations about fishery science. California Cooperative Oceanic Fisheries Investigations Reports 18:34-43. James, A.G. 1987. Feeding ecology, diet and field-based studies on feeding selectivity of the Cape anchovy Engraulis capensis Gilchrist. S. Afr. J. mar. Sci 5:673-692. James, A.G. 1988. Are clupeid microphagists herbivorous or omnivorous? A review of the diets of some commercially important clupeids. S. Afr. J. mar. Sci 7:161-177.
Developing a basis for detecting and predicting long-term ecosystem changes
269
Johnston, S.J. and D.S. Butterworth. 2004. The South African horse mackerel assessment using an agestructured production model, with future biomass projections, Unpublished Report, BENEFIT Workshop document, BEN/DEC04/HM/SA/4a: 24 pp. Kemper, J., J.-P. Roux, P.A. Bartlett, Y.J. Chesselet, J.A.C. James, R. Jones, S. Wepener and F.J. Molloy. 2001. Recent population trends of African penguin (Spheniscus demersus) in Namibia. S. Afr. J. mar. Sci. 23:429-434. Koné, V., E. Machu, P. Penven, V. Andersen, V. Garçon, P. Fréon, and H. Demarcq. 2005. Modeling the primary and secondary productions of the southern Benguela upwelling system: A comparative study through two biogeochemical models. Global Biogeochem. Cycles 19, GB4021, doi:10.1029/2004GB002427: 1-22. Korrûbel, J.L., S.F. Bloomer, K.L. Cochrane, L. Hutchings and J.G. Field. 1998. Forecasting in South African pelagic fisheries management: The use of expert and decision support systems. S. Afr. J. Mar. Sci. 19:415-423. Klyashtorin, L. B. 2001. Climate change and long term fluctuations of commercial catches: The possibility of forecasting. FAO Fish. Tech. Pap. No. 410. Kreiner, A., C.D. van der Lingen and P. Fréon. 2001. A comparison of condition factor and gonadosomatic index of sardine (Sardinops sagax) stocks in the northern and southern Benguela upwelling ecosystems, 1984-1999. S. Afr. J. mar. Sci. 23:123-134. Liao, S-H. 2005. Expert system methodologies and applications – a decade review from 1995 to 2004. Expert Systems with Applications 28:93-103. Lett, C., C. Roy, A. Levasseur, C.D. van der Lingen and C. Mullon. Simulation and quantification of enrichment and retention processes in the southern Benguela upwelling ecosystem. Fish. Oceanog. (in press) Lluch-Belda, D., R.J.M. Crawford, T. Kawasaki, A.D. MacCall, R.H. Parrish, R.A. Schwartzlose and P.E. Smith. 1989. World-wide fluctuations of sardine and anchovy stocks: The regime problem. S. Afr. J. mar. Sci. 8:195-205. Lluch-Belda, D., R.A. Schwartzlose, R. Serra, R. Parrish, T. Kawasaki, D. Hedgecock and R.J.M. Crawford. 1992. Sardine and anchovy regime fluctuations of abundance in four regions of the world oceans: A workshop report. Fisheries Oceanography 1:339-347. Longhurst, A. 1998. Ecological geography of the sea. Academic Press, New York, USA. 398 pp. MacQueen, N. and M.H. Griffiths. 2004. Influence of sample size and sampling frequency on the quantitative dietary descriptions of a predatory fish in the Benguela ecosystem. Afr. J. mar. Sci. 26:205-217. Mantua, N. 2004. Methods for detecting regime shifts in large marine ecosystems: A review with approaches applied to North Pacific data. Progr. Oceanogr. 60 (2-4): 165-182. Mantua, N. J., and S.R. Hare. 2002. The Pacific decadal oscillation. Journal of Oceanography 58:35–44. Mather, D., P.J. Britz, T. Hecht and W.H.H. Sauer. 2003. An Economic and Sectoral Study of the South African Fishing Industry, Volume 1, Economic and regulatory principles, survey results, transformation and socio-economic impact. Report prepared for Marine and Coastal Management, Department on Environmental Affairs and Tourism, by Rhodes University, Grahamstown, South Africa, iii+300pp. Miller, D.C.M. and J.G. Field. 2002. Predicting anchovy recruitment in the southern Benguela ecosystem: developing an expert system using classification trees, S. Afr. J. Sci. 98:465-472. Miller, D.C.M., C.L. Moloney, C.D. van der Lingen, C. Lett, C. Mullon and J.G. Field, 2005, Modeling the effects of physical-biological interactions and spatial variability in spawning and nursery areas on recruitment of sardine in the southern Benguela system, J. Mar. Sys. (in press) Miller, A.J. and N. Schneider. 2000. Interdecadal climate regime dynamics in the North Pacific Ocean: Theories, observations and ecosystem impacts. Prog. Oceanogr. 47 (2-4): 355-379. Moloney, C., A. Jarre, H. Arancibia, Y.-M. Bozec, S. Neira, J.-P. Roux and L.J. Shannon. 2005. Comparing the Benguela and Humboldt marine upwelling ecosystems with indicators derived from inter-calibrated models. ICES Journal of Marine Science 57(3): 493-502. Moloney, C.L, C.D. van der Lingen, L. Hutchings, and J.G. Field. 2004. Contributions of the Benguela Ecology Programme to pelagic fisheries management in South Africa. Afr. J. mar. Sci. 26:37-51. Monteiro, P. and A.K. van der Plas. 2006. Forecasting low oxygen waters (LOW) variability in the Benguela system. (Chapter 5, this volume). Mullon, C., P. Cury and P. Penven. 2002. Evolutionary individual-based model for the recruitment of the anchovy in the southern Benguela. Can. J. Fish. Aquat. Sci. 59:910-922.
270
A. Jarre et al.
Mullon, C., P. Fréon, C. Parada, C. D. van der Lingen and J.A. Huggett. 2003. From particles to individuals: Modelling the early stages of anchovy (Engraulis capensis/encrasicolus) in the southern Benguela. Fish. Oceanog. 12:396-406. Nel, D.C., R.J.M. Crawford and N. Parson. 2003. The conservation status and impact of oiling on the African Penguin. 1-7 in Nel, D.C., P.A. Whittington, B. von Bodungen and R.K. Turner, eds. Rehabilitation of oiled African Penguins: A conservation success story. Birdlife South Africa and Avian Demography Unit, Cape Town, South Africa. 31pp. OSPAR. 2000. Quality Status Report 2000. OSPAR Commission, London, 108 + viii p. O’Toole, M.J., L.V. Shannon, V. de Barros Neto and D.E. Malan. 2001. Integrated management of the Benguela Current region. 229-251 in B. von Bodungen and R.K. Turner, eds. Science and Integrated Coastal Management. Dahlem University Press. Painting, S.J., L. Hutchings, J.A. Huggett, J.L. Korrubel, A.J. Richardson and H.M. Verheye. 1998. Environmental and biological monitoring for forecasting anchovy recruitment in the southern Benguela upwelling region. Fish. Oceanogr. 7 (3-4): 364-374. Parada, C., C.D. van der Lingen, C. Mullon, and P. Penven. 2003. Modelling the effect of buoyancy on the transport of anchovy (Engraulis capensis) eggs from spawning to nursery grounds in the southern Benguela: An IBM approach. Fish. Oceanogr. 12:170-184. Parada, C., C. Mullon, C. Roy, P. Fréon, L. Hutchings and C.D. van der Lingen. Modeling the contribution of vertical migration to pre-recruitment success of Cape anchovy early stages in the southern Benguela ecosystem. Fish. Oceanog. Submitted. Pauly, D., V. Christensen and C. Walters. 2000. Ecopath, Ecosim, and Ecospace as tools for evaluating ecosystem impact of fisheries. ICES J. Mar. Sci. 57 (3): 697-706. Penven P., C. Roy, G. Brundrit, A. Colin de Verdière, P. Fréon, A. Johnson, J. Lutjeharms and F. Shillington. 2001. A regional hydrodynamic model of the Southern Benguela upwelling. South African Journal of Science 97:472-475. Pecquerie, L., L. Drapeau, P. Fréon, J.C. Coetzee, R.W. Leslie and M.H. Griffiths. 2004. Distribution patterns of key fish species of the southern Benguela ecosystem: An approach combining fisherydependent and fishery-independent data. Afr. J. mar. Sci. 26:115-139. Pitcher, G. and S. Weeks. The variability and potential for prediction of harmful algal blooms in the southern Benguela ecossytem. Chapter 7, this volume. Polovina, J. 2005. Climate variation, regime shifts and implications for sustainable fisheries. Bulletin of Marine Science 76(2): 233-244. Power, Y. and P.A. Bahri. 2005. Integration techniqes in intelligent operational management: A review. Knowledge-based Systems 18:89-97. Punt, A. E. and D.S. Butterworth. 1995. The effects of future consumption by the Cape fur seal on catches and catch rates of the Cape hakes. 4. Modelling the biological interaction between Cape fur seals Arctocephalus pusillus pusillus and the Cape hakes Merluccius capensis and M. paradoxus. S. Afr. J. mar. Sci. 16:255-285. Quadling, H. and M. Quadling. 1995. WinExp , a small expert system for Windows v2.11. University of Minnesota, Minneapolis, MN, USA. Rademeyer, R.A. and D.S. Butterworth. 2004. Revised assessments of the Merluccius paradoxus and M. capensis resources for the south and west coasts combined. Unpublished Report. Marine and Coastal Management, Cape Town, South Africa WG/11/04/D:H:25. 34 pp. Reid, P.C., B. Planque and M. Edwards. 1998. Is observed variability in the long-term results of the continuous plankton recorder survey a response to climate change? Fisheries Oceanography 7:282288. Reid, P.C., M.F. Borges, and E. Svendsen. 2001. A regime shift in the North Sea circa 1988 linked to changes in the North Sea horse mackerel fishery. Fisheries Research 50(1): 163-171 Reid, P.C., J.M. Colebrook, J.B.L. Matthews, J. Aiken and Continuous Plankton Recorder Team. 2003. The Continuous Plankton Recorder: Concepts and history, from the Plankton Indicator to undulating recorders. Progress in Oceanography 58:117-173. Reinikainen, T. and F.J. Molloy. 2003. Indicators. 153-162 in Molloy, F.J. and T. Reinikainen, eds. Namibia’s marine environment. Directorate of Environmental Affairs, Ministry of Environment and Tourism, Windhoek, Namibia. 167 pp. Richardson, A.J., B.A. Mitchell-Innes, J.L. Fowler, S.F. Bloomer, H.M. Verheye, J.G. Field, L. Hutchings and S.J. Painting. 1998. The effect of sea temperature and food availability on the spawning success of Cape anchovy Engraulis capensis in the southern Benguela. S. Afr. J. Mar. Sci. 19:275-290.
Developing a basis for detecting and predicting long-term ecosystem changes
271
Richardson, A.J., C. Risien, and F.A. Shillington. 2003a. Using self-organizing maps to identify patterns in satellite imagery. Prog. Oceanogr. 59:223–239. Richardson, A.J., N.F. Silulwane, B.A. Mitchell-Innes and F.A. Shillington. 2003b. A dynamic quantitative approach for predicting the shape of phytoplankton profiles in the ocean. Prog. Oceanogr. 59:301–319. Roetter, R.P., C.T. Hoanh, A.G. Laborte, H. Van Keulen, M.K. Van Ittersum, C. Dreiser, C.A. Van Diepen, N. De Ridder and H.H. Van Laar. 2005. Integration of systems network (SysNet) tools for regional land use scenario analysis in Asia. Environmental Modelling & Software 20:291-307. Roux J-P. 2003. Risks. 137-152 in Molloy, F.J. and T. Reinikainen, eds. Namibia’s Marine Environment, Directorate of Environmental Affairs, Ministry of Environment and Tourism, Windhoek, Namibia, 167p. Roux, J-P. and L. Shannon. 2004. Ecosystems approach to fisheries management in the nothern Benguela: the Namibian experience. Afr. J. mar. Sci. 26:79-93. Roux, J-P. and S. Mercenero. 2004. The diet of a top predator, the Cape fur seal, as an indicator of recruitment variability and juvenile growth rates of Cape hake in Namibia. Poster, presented at the SCOR-IOC Symposium on "Quantitative Ecosystem Indicators for Fisheries Management", 31 March – 3 April 2004, Paris, France. Abstract available at http://www.ecosystemindicators.org. Rothschild, B.J. and L.J. Shannon. 2004. Regime shifts and fisheries management. Progr. Oceanogr. 60:397-402. Roy, C., S. Weeks, M. Rouault, G. Nelson, R. Barlow and C.D. van der Lingen. 2001. Extreme oceanographic events recorded in the Southern Benguela during the 1999-2000 summer season. S. Afr. J. mar. Sci. 97:465-471. Sauer, W.H.H., T. Hecht, P.J. Britz and D. Mather. 2003. An Economic and Sectoral Study of the South African Fishing Industry. Volume 2, Fishery profiles. Report prepared for Marine and Coastal Management, Department on Environmental Affairs and Tourism, by Rhodes University, Grahamstown, South Africa. 308pp. Scheffer, M., S.R. Carpenter, J.A. Foley, C. Folke and B. Walker. 2001. Catastrophic shifts in ecosystems. Nature 413:591-596. Scheffer, M., and E.H. van Nes. 2004. Mechanisms for marine regime shifts: Can we use lakes as microcosms for oceans? Progress in Ocenaography 60 (2-4): 303-319. Schneider, S.H. 1992. Introduction to climate modelling. 3-26 in Trenberth, K.E., ed. Climate system modelling. Cambridge University Press, Cambridge. 788pp. Schwartzlose, R.A., J. Alheit, A. Bakun, T.R. Baumgartner, R. Cloete, R.J.M. Crawford, W.J. Fletcher, Y. Green-Ruiz, E. Hagen, T. Kawasaki, D. Lluch-Belda, S.E. Lluch-Cota, A.D. MacCall, Y. Matsuura, M.O. Nevárez-Martínez, R.H. Parrish, C. Roy, R. Serra, K.V. Shust, M.N. Ward and J.Z. Zuzunaga. 1999. Wordwide large-scale fluctuations of sardine and anchovy populations. S.Afr. J. mar. Sci. 21:289-347. Shannon, L.J, C.L. Moloney, A. Jarre, and J.G. Field. 2003a. Trophic flows in the Southern Benguela during the 1980s and 1990s. Journal of Marine Systems 39 (1-2): 83-116. Shannon, L.J., C.L. Moloney and J.G. Field. 2003b. Simulating anchovy-sardine regime shifts in the southern Benguela ecosystem. Ecological Modelling 172 (2-4): 269-281. Shannon, L.J., V. Christensen and C.J. Walters. 2004a. Modelling stock dynamics in the southern Benguela ecosystem for the period 1978 – 2002. Afr. J. Mar. Sci. 26:179 – 196. Shannon, L.J., C.L. Moloney, P.M. Cury, C.D. van der Lingen, R.J.M. Crawford, P. Fréon, and K.L. Cochrane. 2004b. Ecosystem modelling approaches for South African Fisheries Management. Poster presented at the Fourth World Fisheries Congress, 3-6 May 2004, Vancouver, Canada. Shannon, L.J., K. Cochrane and S.C. Pillar, eds. 2004c, Ecosystem approaches to fisheries in the southern Benguela, African Journal of Marine Science 26, 326 p. Shannon, L.J., C.L. Moloney, P.M. Cury, C.D. van der Lingen, R.J.M. Crawford, P. Fréon, and K.L. Cochrane. 2005. Ecosystem modelling approaches for South African Fisheries Management, Proceedings of the Fourth World Fisheries Congress: Reconciling Fisheries with Conservation: The Challenge of Managing Aquatic Ecosystems (in press) Shannon, L.V. 1985. The Benguela Ecosystem, Part I. Evolution of the Benguela, Physical features and Processes, Oceanogr. Mar. Biol. Ann. Rev. 23:105-182. Shillington, F., C.J.C. Reason, C.M. Duncombe Rae, P. Florenchie and P. Penven. 2006. Large-scale physical variability of the Benguela Current Large Marine Ecosystem (BCLME). Chapter 5, this volume.
272
A. Jarre et al.
Shin, Y.J. and P. Cury. 2001. Exploring fish community variability through size dependent trophic interactions using a spatialized individual-based model. Aquatic Living Resources 14(2): 65-80. Shin, Y.J., L.J. Shannon, and P. Cury. 2004. Simulations of fishing effects on the southern Benguela fish community using an individual-based model: Learning from a comparison with Ecosim. Afr. J. mar. Sci. 26:95-114. Shin, Y.J., M.-J. Rochet, S. Jennings, J.G. Field and H. Gislason. 2005. Using size-based indicators to evaluate the effects of fishing. ICES J. mar. Sci. 62(3): 384-396. Sinclair, M. and G. Valdimarsson, eds. 2003. Responsible Fisheries in the Marine Ecosystem, Food and Agriculture Organization of the United Nations, Rome, and CABI Publishing, UK. 426 pp. Slocombe, D.S. 1999. Defining goals and criteria for ecosystem-based management. Environmental Management 22(4): 483-493. Smeet, E. and R. Weterings. 1999. Environmental Indicators: Typology and overview. Technical Report (No. 25), European Environment Agency, Copenhagen, Denmark.19 pp. Sowman, M., M. Hauck and G. Branch. 2003. Lessons learned from nine coastal and fisheries comanagement case studies. 299-340 in M. Hauck and M. Sowman, eds. Waves of change: Coastal and fisheries co-management in South Africa. University of Cape Town Press, South Africa. 358 pp. Starfield, A.M. and N.J. Louw. 1986. Small expert systems: As perceived by a scientist with a computer rather than a computer scientist. S. Afr. J. Sci. 82:552–555. Steele, J.H. 1996. Regime shifts in fisheries management. Fisheries Research 25:19-23. Steele, J.H. 1998. Regime shifts in marine ecosystems. Ecological Applications 8(S1):S33-S36. Trenkel, V. and M.J. Rochet. 2003. Performance of indicators derived from abundance estimates for detecting the impact of fishing on a fish community. Can. J. Fish. Aquat. Sci. 60:67-85. Underhill, L.G. and R.J.M. Crawford. 2005. Indexing the health of the environment for breeding seabirds in the Benguela ecosystem. ICES Journal of Marine Science 62(3): 360-365. van der Lingen, C.D. 2002. Diet of sardine Sardinops sagax in the southern Benguela upwelling ecosystem. S. Afr. J. mar. Sci 24: 301-316. van der Lingen, C.D. and J.A. Huggett. 2003. The role of ichthyoplankton surveys in recruitment research and management of South African anchovy and sardine. 303-343 in Browman, H.I. and A.B. Skiftesvik, eds. The Big Fish Bang: Proceedings of the 26th Annual Larval Fish Conference, Institute of Marine Research, Bergen, Norway, ISBN 82-7461-059-8. van der Lingen, C.D., J.C. Coetzee, and L. Hutchings. 2002. Temporal shifts in the spatial distribution of anchovy spawners and their eggs in the Southern Benguela: Implications for recruitment, GLOBEC Report 16:46-48. van der Lingen, C.D., L. Hutchings, D. Merkle, J.J. van der Westhuizen and J. Nelson. 2001. Comparative spawning habitats of anchovy (Engraulis capensis) and sardine (Sardinops sagax) in the southern Benguela upwelling ecosystem. 185-209 in Spatial Processes and Management of Marine populations. Alaska Sea Grant College Program, Fairbanks, Alaska, USA. van der Lingen, C.D., L.J. Shannon, P. Cury, A. Kreiner, C.L. Moloney, J.-P. Roux, and F. Vaz-Velho, Resource and ecosystem variability, including regime shifts, in the Benguela Current system. (Chapter 8, this volume). Monteiro, P.M.S., A.K. van der Plas, G.W. Bailey, P. Malanotte-Rizzoli, C.M. Duncombe Rae, D. Byrnes, G. Pitcher, P. Florenchie, P. Penven, J. Fitzpatrick, H.U. Lass. 2006. Forecasting shelf processes in respect of low oxygen water – hypoxia / anoxia etc. (Chapter 13, this volume) Verheye, H.M., A.J. Richardson, L. Hutchings, G. Marska and D. Gianakouras. 1998. Long-term trends in the abundance and community structure of coastal zooplankton in the southern Benguela system, 1951-1996. S. Afr. J. Mar. Sci. 19:317-332. Whipple, S.J., J.S. Link, L.P Garrison and M. Fogarty. 2000. Models of predation and fishing mortality in aquatic ecosystems. Fish and Fisheries1(1):22-40. Woods, J. A vision for modelling and forecasting. 2006. Chapter 17, this volume. Wooster, W.S. and C.I. Zhang. 2004. Regime shifts in the North Pacific early indications of the 19761977 event. Progress in Oceanography 60(2-4):183-200. Yemane, D., J.G. Field and M.H. Griffiths. 2004. Effects of fishing on the size and dominance structure of linefish of the Cape region, South Africa. Afr. J. mar. Sci. 26:161-177. Yemane, D., J.G. Field and R.W. Leslie. 2005. Exploring the effects of fishing on fish assemlages using Abundance Biomass Comparison (ABC) curves. ICES Journal of Marine Science 62(3): 374-379.
273
Large Marine Ecosystems, Vol. 14 V. Shannon, G. Hempel, P. Malanotte-Rizzoli, C. Moloney and J. Woods (Editors) © 2006 Elsevier B.V. All rights reserved.
12 The Requirements for Forecasting Harmful Algal Blooms in the Benguela S. Bernard, R.M. Kudela, P. Franks, W. Fennel, A. Kemp, A. Fawcett and G.C. Pitcher
INTRODUCTION The Benguela system suffers from the frequent occurrence of a variety of harmful algal blooms (HABs) (Pitcher and Calder 2000). These blooms can have severe negative impacts on local marine ecosystems and communities, in addition to commercial marine concerns such as rock lobster and aquaculture operations. Harmful impacts of HABs are associated with either the toxigenicity of some species, or the high biomass such blooms can achieve. Collapse of high biomass blooms through natural causes such as nutrient exhaustion can lead to low oxygen events, which in extreme cases result in hypoxia and the production of hydrogen sulphide, frequently causing dramatic mortalities of marine organisms. Effective coastal management requires the characterisation of HABs as ecologically prominent phenomena, the means of monitoring critical ecosystem locations in real-time and, ultimately, the operational forecasting of both HABs and their impacts. This document outlines the feasibility and requirements for establishing an operational HAB monitoring and forecasting system in the southern Benguela based on the current state of understanding of the variability of HABs within the region (Pitcher and Weeks, this volume). HAB forecasts are likely be derived primarily from the output of sub-ecosystem models. The structure of a potential forecasting system is thus dictated to a large degree by the effectiveness of coupled physical-biological models. There is a high degree of uncertainty associated with the biological components of such models, particularly any species level aspect of prediction, as discussed in greater detail below. A central tenet of any regional forecasting system is thus the use of real-time observations to effectively replace the need to model biological processes associated with HAB development. Algal blooms classified as potentially harmful in the Benguela additionally have a highly variable taxonomic composition (see Pitcher and Weeks, this volume), and for the purposes of forecasting are best characterized by their impacts. Distinct in their nature, these impacts are associated with either the toxicity of some species present in the assemblage, or hypoxia resulting from the shoreline retention and collapse of high biomass blooms. The requirements for the forecasting of HABs in the Benguela are dictated primarily by these two modes of impact, which both require prediction of shoreline impact and retention.
274
S. Bernard et al.
Ideally, real-time observation techniques would address both of those concerns, offering assemblage identification, toxicological status, biomass quantification, and spatial bloom delineation using multiple techniques at the required variety of temporal and spatial scales. In addition, HAB prediction and effective risk assessment in a physically dominated system such as the Benguela requires meteorological and hydrophysical data pertaining to the transport, maintenance and potential retention of blooms. An ideal observation network would thus utilise high frequency data from multi-sensor coastal observation platforms situated at locations critical to both preliminary bloom detection and resultant advective transport, in addition to synoptic satellite derived data. Products from such a network would be required at several levels: geophysical input data for use with physical, ecophysiological or probabilistic impact assessment models; and high level data products suitable for immediate use by coastal management agencies. The specific aims of a regional operational forecasting system are also an obvious and important determinant of the structure of any real-time observational system. Given the dependence on real-time observation rather than ecophysiological modeling, effective HAB forecasting in the Benguela in the immediate future is thus likely to take two forms. The first of these is a probabilistic “ecological window” or fuzzy logic model, whereby the probability of the succession of broad taxonomic groups and HAB occurrence is determined from observations of physical, chemical and biological conditions in real-time. Such a structure allows the complexities of physicalbiological couplings to be reduced to operationally attainable indices, e.g., identifying the likelihood of diatom or dinoflagellate dominance based on the expected turbulence regime and water column stratification. The second and more important form of forecast is likely to be limited to a short term prediction of impact based on a combination of real-time bloom detection and transport prediction, likely in turn to be based on a concomitant meteorological forecast. The core product of any realisable forecasting system in the short term is therefore likely to be a prediction of HAB shoreline impact and retention, highly dependent upon observational ability, the temporal limitations of meteorological forecasts, and the dynamic circulation models needed to simulate the transportation of HABs in local flow fields. The need for appropriately configured and located coastal moorings is thus absolutely critical to all forms of HAB forecasting, as is the need for routine access to appropriate satellite derived geophysical products. Also vital to such a forecasting capability are dynamic circulation models able to integrate data from real-time observation systems. Simulation of the shoreward transport of blooms should be considered the primary aim of these sub-ecosystem circulation models. The ability to assimilate real-time physical data, and utilise real-time biological data for particle tracking and bloom dispersion is also necessary for operational forecasting. These requirements are considered here with the aim of establishing a set of prerequisites for establishing an operational circulation model for HAB forecasting. Mechanistic forecasting of HABs in the Benguela is unlikely in the immediate future: the complex biological processes underlying HAB development are simply not well enough understood to model effectively. Nevertheless, the physical processes
Requirements for forecasting harmful algal blooms in the Benguela
275
underlying the concentration, dispersion and advection of blooms are relatively well understood, and furthermore can be simulated using dynamic circulation models. The use of these circulation models in conjunction with real-time observation networks, and relatively simple physical-biological indices, offer the structure around which an operational HAB forecasting system in the Benguela can be developed. PHYSICAL-BIOLOGICAL COUPLINGS UNDERLYING HABS Harmful algal blooms of phytoplankton deserve special recognition relative to other more benign phytoplankton assemblages and blooms because of their obvious importance to humans. It is often difficult, however, to identify physical, chemical or biological characteristics of HAB species (other than toxicity) that would allow us to treat these organisms as a unique group or subset of phytoplankton. All phytoplankton by nature are intimately connected to their physical environment, since phytoplankton are by definition subject to horizontal transport by the physical environment. Therefore, any discussion of physical-biological couplings must ultimately be generally applicable to phytoplankton. This is especially true of upwelling systems such as the Benguela, where the dominant influence is wind-driven transport, both from local forcing and from the subsequent development of characteristic larger-scale forcing such as coastal jets and countercurrents. Theoretical frameworks for HAB species succession In identifying the important physical-biological couplings underlying HABs, it is useful to start with a theoretical framework for patterns of biological succession, with emphasis on upwelling regions. Margalef et al. (1979) described the general succession of phytoplankton functional groups based on an ecological space defined by a continuum from r-selected (diatoms) to K-selected (dinoflagellates) organisms, largely controlled by turbulence, which provides a proxy for nutrient and light availability. According to his Mandala, red-tide events of high biomass, a primary HAB event in the Benguela, were somewhat of an anomaly from the expected successional patterns, and could be treated as unusual or alternate sequences. Margalef identified these red tides as being associated with low turbulence, highly stratified waters with an excess of nutrients. More recently, Smayda and Reynolds (2001) revisited Margalef’s Mandala, and updated it based on a classification scheme which divides phytoplankton into three groups: C, or colonizing organisms, characterized by small size and motility, and an opportunistic growth strategy; S, or stable organisms, adapted to be stress-tolerant, with large, slow-growing forms dominant (Margalef’s K-selected dinoflagellates); and R, or ruderal, organisms, typically disturbance and turbulence tolerant, autotrophic, with an affinity for high nutrient levels and often associated with upwelling fronts (Margalef’s r-selected diatoms). Based on this scheme, Smayda (2002) further refined this for dinoflagellates, subdividing this group into nine categories based on the
276
S. Bernard et al.
ecological niches they occupy as defined by preference for nutrients and tolerance to decreasing light and mixing. Smayda (2002) also generally classified diatoms and dinoflagellates, the two main groups of bloom-forming (and HAB) organisms in coastal upwelling systems such as the Benguela, as having differing strategies for growth. Diatom blooms tend to be annual, of high species diversity, exhibit a fairly stable successional pattern as a group relative to other phytoplankton, and have adopted a “sink” strategy when faced with sub-optimal environmental conditions. Dinoflagellate blooms tend to exhibit both lower diversity and truncated succession patterns, and exhibit a “swim” strategy (detailed below). It is also important to note that dinoflagellate blooms and red tides, while relatively unpredictable in comparison to diatom blooms, are not unusual or anomalous, and are a normal part of the successional pattern occurring in upwelling systems. As noted by Pitcher and Weeks (this volume), the switch from diatom to dinoflagellate dominated algal communities can be considered succession in its simplest form, and as such the most likely succession pattern amenable to forecasting. In terms of growth strategy, dinoflagellates as a group are generalists, with representative genera found in all major ecosystem types. Although dinoflagellates are often associated with highly stratified, low-nutrient environments (e.g. K-selected), they do not appear to be particularly well adapted to a low-nutrient environment (i.e. they exhibit moderate to low growth rates, moderate to low affinities for nutrients, etc.). To overcome this, four strategies have developed: vertical migration, mixotrophy, allelochemical competition, and allelopathic grazing deterrents (Smayda 2002). Although dinoflagellates as a group can be considered generalists, there is considerable species- and probably sub-species level selection and adaptability, introducing a great deal of stochasticity in blooms; additionally, within a given ecosystem such as the Benguela, it is not uncommon to find three or more of Smayda’s (2002) nine subdivisions of dinoflagellates simultaneously in a small spatial region. Despite the stochasticity evident at the species-level in bloom development, there are clear successional patterns in upwelling systems. For example, in the California Current System, from year to year pennate diatoms dominate early in the year accompanying the onset of upwelling-favorable conditions, succeeded in the summer and fall by larger centric diatoms, then by dinoflagellates in the fall and early winter, and finally by a small flagellate and picoplankton community in the winter (with a substantial decrease in total biomass). What is unusual about upwelling systems compared to other regions is that, within these seasonal patterns, the biological patterns can be “reset” to some degree in response to changes in physical forcing (upwelling/downwelling), which typically occur on the order of 5-10 day timescales. It is clearly possible to identify time periods and conditions favorable for one group versus another, generally related to the physical setting. It is at the more detailed level (a particular year, or an attempt to predict particular species) that forecasting becomes difficult. This leads to a gap in predictive forecasting skill of HAB events, where it should be possible to identify and predict the formation of group- or species-specific bloom events several days in advance (on the timescale of weather forecasting), and it
Requirements for forecasting harmful algal blooms in the Benguela
277
is possible to predict the general seasonal patterns, but it is extremely difficult to forecast events from weeks to months in advance. Identification of physical-biological interactions Within this larger pattern of diatoms versus dinoflagellates and seasonal succession, there are several important physical-biological interactions that influence the development or suppression of high-biomass or toxic dinoflagellate blooms. At the cellular level, the most obvious example is the ability of dinoflagellates to swim. This behavior allows them to maintain position in the water column, to seek out optimal growth conditions (e.g., to cross the nutricline in nutrient-depleted waters or to optimize their light harvesting), and to overcome or avoid the often-turbulent conditions associated with frontal systems and upwelling regions in general. Despite the importance of motility, there is no evidence that HAB-forming dinoflagellates exhibit particularly weaker or stronger swimming behavior (see the review by Smayda 2000), although chain-forming dinoflagellates such as Alexandrium catenella, one of several HAB species found in the Benguela, do exhibit strong swimming behavior. There is also substantial variability amongst HAB species; recent laboratory experiments conducted by Sullivan et al. (2003) showed that the dinoflagellate Lingulodinium polyedrum exhibits enhanced growth and decreasing cell size in response to turbulence, possibly a mechanism for coping with shear stress. At the same turbulence levels, A. catenella exhibits decreased growth and increased cell size, suggesting that one organism (L. polyedrum) might be stimulated by turbulence, while another, A. catenella, would likely use its swimming behavior to congregate in lowturbulence regions. Turbulence has also been shown to result in enhanced toxicity in A. fundyense (Juhl et al. 2001). Turbulence-avoidance, as hypothesized for A. catenella, is also consistent with the observed formation of subsurface accumulations, often forming thin layers, by this and other HAB organisms. These subsurface layers may accumulate due to behavioral and/or purely physical mechanisms (Donaghay and Osborne 1997; McManus et al. 2003). Evolutionarily, a possible benefit for concentration of HAB species into thin layers may be relief from grazing pressure. Some zooplankton grazers may cease feeding and starve rather than consume toxic algae (Turner and Tester 1997), and zooplankton have also been observed to avoid layers with high concentrations of toxic phytoplankton (Fiedler 1982). Sub-surface layers of HAB organisms are of obvious importance in terms of monitoring and prediction, since these layers may serve as refugia or sources for the apparently sudden surface expression of HABs, and cannot easily be detected from the surface using methods such as remote sensing. In addition to physical-biological interactions at the cellular level, interactions with mesoscale physical features are also of importance in upwelling systems. Some HAB dinoflagellates found in the Benguela produce cysts (e.g., A. catenella), and the persistence and seasonal reoccurrence of these organisms is almost certainly related to the formation of cyst beds. Many other HAB dinoflagellates do not encyst, however, and it is not clear what the source of initial seed population is, particularly in
278
S. Bernard et al.
upwelling systems which are by nature highly advective. Previous work in the Benguela has identified the formation of pelagic seed banks (Pitcher and Boyd 1996, Pitcher et al. 1998, Smayda 2002); potentially leading to dinoflagellate concentration in offshore coastal fronts, and ultimately coastal impact through physically-induced onshore transport. Potential for forecasting HAB biological characteristics HAB organisms should be considered part of the normal successional patterns of species in upwelling ecosystems, despite the seemingly random selection at the species level among HAB organisms which occupy similar ecological niches. Although upwelling systems are unique in that successional patterns can be at least partially “reset” by the cyclic occurrence of upwelling/relaxation events, at a gross level ecological windows conducive to HAB formation should be identifiable. On larger physical-biological scales, there are several examples of HAB predictability based on statistical correlations between changes in the mean physical or environmental structure and the occurrence of HAB groups. In Hong Kong, Yin (2003) has shown that severe red tides are often associated with monsoonal wind conditions. In Galicia, the upwelling index, in combination with remotely sensed sea surface temperature and chlorophyll, has provided a robust index of Pseudo-nitzschia blooms (Sacau-Cuadrado et al. 2003). Along the Baja Peninsula in Mexico, Ochoa (2003) identified HABpromoting conditions related to the ENSO index, and there is some evidence that L. polyedrum in the California Current System is at least loosely correlated to both ENSO and the Pacific Decadal Oscillation (Kudela, unpublished). There is also potential for the forecasting of the initiation of high biomass HAB events from remotely sensed physiological proxies such as quantum efficiency of fluorescence. Synoptic fluorescence quantum yield data can be derived from the use of analytical reflectance algorithms with space-based ocean colour data (Bernard 2005). Preliminary indications are that such data may be used to establish algal growth phase variations (e.g., Young and Beardall 2003). In summary, the difficulties of ecophysiological modeling, and the inherent stochasticity of algal succession and bloom dynamics mean that the most pragmatic current means of predicting biological HAB attributes in the Benguela is the use of real-time physical, chemical, and biological data in conjunction with a statistical “ecological window” approach. It must also be realised that high biomass HAB events in the Benguela are primarily forced by mesoscale physical factors: identification of the processes underlying these factors, and their effects on algal concentration, dispersion and advection, are critical to establishing a forecasting capability.
Requirements for forecasting harmful algal blooms in the Benguela
279
IDENTIFICATION OF THE PHYSICAL PROCESSES IMPORTANT TO BLOOM CONCENTRATION AND TRANSPORT The ability to predict HABs in coastal upwelling systems depends fundamentally on understanding not only causative biological processes, but also the physical environment in which the HAB species grows. From a management perspective, it is important to be able to anticipate the movements of a HAB, particularly the intersection of an offshore HAB with the coast. Thus understanding the factors that determine the advection of a HAB will be critical elements in forecasting the impacts of HABs. The basic aspects of the physical dynamics common to many coastal upwelling systems are amenable to real-time measurement and short-term forecasting. In particular, an understanding must be sought of the factors controlling the motions of waters containing HABs and the waters that may stimulate the formation of HABs. In the following pages models of upwelling/downwelling systems are therefore examined, with the intention of identifying important transport pathways and essential dynamics that may support the initiation, transport, or dispersal of HABs. An investigation is made of a simple 2D 2-layered upwelling system, 3D effects in upwelling systems, and 2D and 3D flows in downwelling and relaxation conditions. The investigation concentrates on identifying where water comes from, where it goes, and the forces that underlie these motions, in addition to dispersion and concentration mechanisms.
Principal physical processes and their role in shoreward bloom transport Upwelling A 2D model looks only at variations in two dimensions (cross shore and vertical), but ignores the alongshore changes. In Ekman’s simple 2D 2-layered model of winddriven coastal upwelling, an alongshore wind stress causes an offshore flux of surface water, the Ekman transport. This offshore flow must be replaced by an onshore flow of deeper water which drives upwelling of deep water at the coast with subsequent uplifting of the pycnocline. The assumption of two well-mixed layers obscures details of the origin and vertical structure of the deep return flow, in particular whether it originates from the interior of the water column or the bottom boundary layer. Convergent or divergent structures at the front, also obscured in a 2D structure, can only be resolved in models with vertical stratification. Allen et al. (1995) explored a 2D primitive-equation model with a turbulence-closure mixed-layer model whose bathymetry was representative of the shelf off the coast of Oregon. The model had a gradual vertical stratification, characteristic of the water column of that region, and was forced with a steady upwelling-favorable wind. In the base case (no heating), the surface wind stress caused the formation of a surface mixed layer that deepened with time, as well as the offshore Ekman transport of surface water. The offshore surface flux was replaced by water from a thin near-bottom
280
S. Bernard et al.
Ekman layer, not waters from the interior of the domain. As the wind blew, isopycnals were pushed upward and finally broke the surface, forming a front that moved offshore. A convergence zone formed at the front, as inshore water subducted beneath it. A strong along-front jet formed at the front, and moved offshore with the front. Heating tended to reduce the surface turbulence and the depth of the surface mixed layer. Doubling the wind stress led to a doubling of the offshore surface Ekman transport, and a halving of the time taken for the frontal jet to reach a given speed. Decreasing stratification led to a decrease in the strength of the frontal jet and caused the cross-shelf flow to be distributed through a greater portion of the water column. Austin and Lentz (2002) performed a similar modeling study, though their model was configured for a broad shelf (such as the east coast of the US), and with a sharp pycnocline at about 10 m depth. Their results were largely similar to those of Allen et al. (1995), with some interesting additional details. During upwelling, Austin and Lentz showed that as the isopycnals rose and surfaced to form the front, a vertically well-mixed region formed inshore of the front where the surface and bottom Ekman layers met. This shallow inner-shelf region extended to a depth of about 15-20 m (10 km offshore with their bathymetry). Because of the lack of vertical stratification, this inner shelf region did not support vertical shear, so onshore-offshore motions could not penetrate to the coast. Rather, the upwelling was confined to a zone near the front, and propagated offshore with the front, leaving a region with no horizontal motions inshore of the front. Tracers introduced into this model either in the bottom layer offshore, or at the coast, showed very little across-shore motion during upwelling. The tracer in the bottom layer moved slightly onshore until the front rode over it, when it became vertically well mixed and ceased to move horizontally. The tracer at the coast was almost immediately trapped within the vertically mixed inner shelf region, and showed no across-shelf movement. Thus in this scenario, delivery of a HAB to the coast would be unlikely, no matter where the HAB was in the water column. Downwelling During active downwelling, an alongshore wind (poleward on a west coast in the southern hemisphere) forces surface Ekman transport onshore, causing the pycnocline to bow downwards, intersecting the bottom and moving offshore at depth. This front is associated with an increasingly intense along-front jet in the direction of the wind. Allen and Newberger (1996) used the same numerical model as Allen et al. (1995) to explore the dynamics of coastal downwelling in a 2D domain configured for the Oregon coast. In their base case (no heat flux), the isopycnals behaved as predicted by Ekman theory, bending downwards and moving offshore. A well-mixed inner shelf region formed as the front moved offshore and deeper, with an along-front jet associated with strong convergence and downwelling. Austin and Lentz (2002) also explored downwelling-forced flows in a region with a broad shelf and a sharp pycnocline. Their results echoed those of Allen and Newberger (1996), with the formation of a strong front at the bottom that propagated offshore, associated with a strong along-front jet and a vertically mixed inner-shelf region suppressing vertical shear and horizontal circulations between the coast and the
Requirements for forecasting harmful algal blooms in the Benguela
281
front. This was clearly demonstrated by the motions of tracers introduced into the surface layer, and at the coast in the model. The tracer in the surface layer moved slightly onshore as the wind blew, but became horizontally stationary once the front had reached and passed the patch. After the front had passed by, the patch of tracer was vertically homogeneous, and no longer moved onshore or offshore in the inner shelf. The tracer introduced at the coast remained trapped at the coast by the wellmixed inner shelf. These results suggest that onshore transport of organisms embedded in these flows is minimal. Even a patch of tracer at the surface moved onshore only minimally under forcing from a downwelling-favorable alongshore wind. Tilburg (2003) explored this problem further by investigating the effects of across-shore (onshore) wind stresses on the transport of a tracer in the surface layer of a 2D linearly stratified model. He showed that a wind with an onshore component had a substantial effect on the onshore movement of the surface tracer and could transport it through the well-mixed inner shelf to the coast. Using these results it is possible to predict that onshore movement of a surface HAB would be driven by winds with an onshore component. Relaxation An incorrect generalisation of upwelling systems is that when the wind ceases, the upwelling front propagates onshore under the influence of the across-shore pressure gradient associated with the uplifted pycnocline. In a purely 2D system, however, this will not happen, as conservation of the potential vorticity results in a geostrophic balance which maintains the position of the offshore front and the along-front jet until the system is forced by changes in the across shore pressure gradient such as new wind events or propagating coastally trapped waves. On the other hand, it is a common observation that warm water reaches the coast during relaxation (e.g., Send 1987), and it is important to identify both the source of this warm water and the underlying transport processes. There are several mechanisms that could lead to the appearance of warm water at the coast after an upwelling-favorable wind stops (see Figure 12-1). One process, as discussed above, is active downwelling, which can cause onshore movement of water if there is an onshore component to the wind stress. Other processes are more complex, and include the formation of eddies (baroclinic instability) and 3D dynamics associated with complex topography such as capes or canyons. Baroclinic instabilities form when small along-front perturbations are magnified by the across-front shear and the slope of the pycnocline. As the small perturbations grow, they can roll up into eddies that can detach from the front and intersect the coast. These eddies may concentrate organisms and deliver them to inshore regions, causing the sudden appearance of warm water and HABs. These eddies are often associated with coastal topography and may impinge on the coast at predictable locations. The size and growth rate of the eddies is related to the bottom slope relative to the pycnocline slope. Steeper bottom slopes tend to reduce the formation of small eddies and increase the time taken for the eddies to form.
282
S. Bernard et al.
Alongshore topography has been recognized to play a role in upwelling relaxation for many years (e.g., Send 1987); however, the details of the mechanisms are still somewhat obscure. Gan and Allen (2002) explored the dynamics of flows near capes embedded in a coastal upwelling system using a 3D primitive-equation model. Their model clearly shows the appearance of warm water at the coast around a cape during upwelling relaxation and their analyses shed some light on the underlying mechanism. During upwelling, the frontal jet moves offshore, where the jet can separate from the coast and move into open waters downstream of the cape. Under steady upwellingfavorable alongshore winds, water downstream of the cape is also accelerated alongshore in the direction of the wind. Two regions of strong alongshore pressure gradients build up around the cape: on the upwind side, the pressure gradient is oriented toward the cape (i.e., in the direction of the wind), while on the downwind side of the cape the pressure gradient again points toward the cape, counter to the direction of the wind. The force balance seems to indicate that the upwind pressure gradient is balanced by a strong geostrophic onshore flow of deep, cold water, while the downwind pressure gradient is balanced by nonlinear forces associated with the movement of the jet around the cape. When the wind subsides, the upwind pressure gradient dissipates, causing no alongshore flows. The downwind pressure gradient does not dissipate, however, and forces water to move alongshore opposite to the former wind direction. This water comes from offshore due to some recirculation around the cape during upwelling, causing a warm water mass to advect toward and around the cape during relaxation, in the opposite direction of the previous upwelling jet. These dynamics may account for the appearance of warm water at the coast during relaxation and the sudden appearance of HABs in bays downwind of a cape, or along the coast upwind of a cape during relaxation.
Forcing mechanisms underlying the prediction of shoreward bloom transport Wind stress curl As previously discussed, the main reason for coastal upwelling is the divergence of the Ekman transport caused by the coast. Another mechanism able to generate divergence of the Ekman transport are wind stress curls, i.e., wind variation perpendicular to the wind direction. The wind stress curl is not associated with a wave guide. Thus contrary to areas close to the coast where upwelling diminishes after a while owing to coastally trapped waves, upwelling driven by the wind stress curl will not be reduced by wave processes (Fennel 1999). Whilst bands of high wind stress curls have been observed in upwelling areas (Bakun and Nelson 1991), measurements of these curls were relatively sparse. Recent work (e.g., Chelton et al. 2004, Koracin et al. 2004) has shown that the wind stress contains a great deal of structure that is usually not measured or used to force models. Koracin et al. show that the curl of the wind stress, the aspect of wind stress that drives upwelling and downwelling, can have scales as small as a few kilometers across shore, particularly in regions of coastal topography such as mountains or capes. This has important implications for the structure of the coastal flow patterns, since the
Requirements for forecasting harmful algal blooms in the Benguela
283
Figure 12-1. Schematic of principal physical processes responsible for shoreward bloom transport on the southern Benguela Namaqua shelf region.
vertical flows forced by the wind will tend to have spatial scales similar to the spatial scales of the wind stress curl. Thus, we might expect multiple fronts to form, with complex patterns of transport along and among the frontal jets. This will have important implications for the transport of HAB species to nearshore locations. These results reinforce the notion that wind-driven flows in upwelling systems are highly three-dimensional, particularly in regions of complex coastal topography. Coastally trapped waves and undercurrents Although the basic process of coastal upwelling, the divergence of the Ekman transport near the coast, can be understood on the basis of 2D models, it is important to note that the dynamics of an upwelling system are basically three dimensional. The coastal or along-front jets are significantly stronger than the cross-shore motion. The coast is known to act as a guide for coastally trapped waves. The finite extent of the wind fields and irregularities of the coastline are the main reason for the formation of waves, in response to temporal changes of the wind fields. A model case, which can be treated analytically and provides understanding of the basic features, is a stratified coastal ocean with a straight coast and flat bottom (McCreary 1981, Fennel 1988). In this case the relevant wave processes are Kelvin waves, which propagate with the coast to the right (in the northern hemisphere) and switch the dynamical balance from
284
S. Bernard et al.
linearly growing upwelling and accelerating coastal jets to a regime where the upwelling stops, and the speed of the coastal jet will be arrested. Behind the Kelvin waves an alongshore pressure gradient develops, which is driving a coastal undercurrent. The Ekman transport then no longer drives the vertical flow but is fed into the undercurrent. The Kelvin waves are the simple non-dispersive extreme case. In a more realistic consideration, the topography must be included, which makes the wave properties much more complex. The associated waves are continental shelf waves which are dispersive and can even have frequency- wavenumber combinations with vanishing group velocity. The theory of these processes is complex and cannot easily be outlined. However, some of the basic features of the Kelvin wave regime are retained with gradual modifications due to the shelf topography. With regard to predictions of the motion of HABs the undercurrent is of obvious importance. Small upward or downward vertical migration of algal cells can change their direction of transport: from a downwind direction with the surface coastal jet to a movement in the opposite direction with the nearshore coastal undercurrent. Whilst traditionally concerns have focused on across-shelf flows during upwelling, the along-shelf flows actually dominate the transport of organisms. The formation of an inner shelf circulation cell may substantially inhibit across-shelf transport of organisms to the coast, though this may be overcome by three-dimensional circulation patterns such as baroclinic instabilities and onshore winds. REAL-TIME OBSERVATION OF HABS In-situ observation technologies Technological, logistical, and financial restraints will all impact upon the ability to construct an observational network capable of allowing the observations required for the prediction of all potential impacts of harmful algae. Technological restraints are likely to change rapidly, given the emergent nature of autonomous instrumentation in the field; and whilst some discussion must necessarily be made concerning the current nature of these restraints, it should be realised, and indeed hoped, that such discussion will rapidly lose relevance. Only techniques of potential utility for real-time autonomous operation on moorings will be discussed here, although traditional sampling and analysis methods, such as cell counts and toxin assays, will obviously continue to be extremely important for the foreseeable future. Species identification and toxicological observations Identification of phytoplankton to a species or genus level with remote autonomous instrumentation is a significant technological challenge, albeit one that is being rapidly surmounted. Autonomous species level identification has already been demonstrated using rRNA probes with the Environmental Sample Processor (ESP) developed at MBARI (Scholin et al. 1999) At the time of publication, a second-generation ESP system, which would be more widely available to end users, was being developed with
Requirements for forecasting harmful algal blooms in the Benguela
285
support from the U.S. National Science Foundation. Similar molecular probe and other techniques showing promise for automated application are currently being tested and used in field studies (e.g., Scholin et al. 2003). Identification to a species or genus level is also possible using the autonomous FLOWCAM instrument, a combined flow cytometer and microscope recently made commercially available (Sieracki 1998). Recent reviews of these techniques discuss them in greater detail (Glasgow et al. 2004, Sellner et al. 2003). Autonomous determination of cellular toxicological data is also extremely desirable, given the variability of cellular toxicity in response to environmental conditions (e.g., Bates 1998). The autonomous ESP platform has also been used for conducting toxin assays with immunological probes. While at present it appears that this is the only technology suitable for such measurements, there is potential use for in situ toxin measurements using miniaturized autonomous mass spectrometers (Langebrake 2003), while other technologies applicable to HAB monitoring are also being developed (Daly et al. 2004). It should be realised that whilst the systems discussed above are of great potential utility, they are emergent technologies using complex, specialised, high maintenance instrumentation that is unlikely to be cost effective (or even commercially available) soon. As such, these systems are unlikely to provide routine algal monitoring data in the Benguela in the near future. Bio-optical observations In comparison to the systems described above, a wide variety of relatively simple, robust, and cost effective bio-optical instrumentation is commercially available that is well suited to autonomous deployment (e.g., Cullen et al. 1997). However, while such instrumentation offers data that is able to provide suitably accurate determinations of algal biomass, and in some cases a limited description of assemblage type (Roesler et al. in press), it cannot offer the level of specificity described above, i.e., species level identification or toxicity information. Nevertheless, bio-optical systems offer a currently realisable means of obtaining real-time data relating to the algal assemblage using a variety of platforms, sensor systems, and processing techniques (Cullen et al. 1997, Sellner et al. 2003) ranging from the empirical to the use of sophisticated analytical inversion algorithms (e.g., Roesler and Boss 2003) – a recent synopsis of available instrumentation and techniques can be found in Oceanography (Vol. 17(2):June 2004). One particular approach will be focused on here, the use of passive radiometric sensors in conjunction with analytical reflectance inversion algorithms. While offering a robust, cost effective moored system allowing the derivation of algal biomass, size and accessory pigment descriptors (Bernard 2005), the reflectance algorithm approach also offers the ability to derive concentrations of other water constituents, such as algal degradation products, which may be of utility in assessing bloom growth phase. The approach also offers a significant advantage in that it allows the derivation of equivalent geophysical products from both in situ, airborne, or space-based sensors. This offers the ability to use analogous multi-platform derived data measured on a
286
S. Bernard et al.
Figure 12-2. Mooring time series data and MERIS chlorophyll a data showing the detection and wide spatial extent of a bloom of the small dinoflagellate Prorocentrum triestinum from 2 to 5 April 2004, in the Namaqua shelf region. The bloom appears at the mooring ~ 4 hours after the satellite overpass, as warm high biomass bloom waters are advected shoreward in the easterly surface flow. Satellite chlorophyll a data, derived through an experimental red band algorithm designed for high biomass application, show the widespread and complex distribution pattern of the bloom. Data such as these are likely to form the basis of a southern Benguela HAB observation system.
Requirements for forecasting harmful algal blooms in the Benguela
287
variety of temporal and spatial scales, i.e., Eulerian high frequency mooring-derived data and daily synoptic satellite-derived data. Physical and chemical observations The dominance of physical forcing mechanisms in the Benguela, the critical role of bloom transport in HAB forecasting, and the desire to use data assimilation techniques in hydrodynamic models underline the need to gather physical in conjunction with biological data. The importance of spatial variability in wind stress as an influence on shelf transport processes emphasises the need for wind measurements in as many across– and along-shelf locations as is feasible. Temperature profiles, indicative of both potential ecological niches and the stratification changes likely to lead to conditions favouring dinoflagellate blooms, are also desirable. Surface currents or current profiles offer both a Eulerian indication of immediate bloom transport and data suitable for assimilation into hydrodynamic models. In this regard, there would be obvious advantages to synoptic surface current data from a coastal high frequency radar system, e.g., CODAR (Glenn et al. 2004), to greatly improve predictions of surface bloom transport. Additional sensors offering the capability of indicating the onset of wind reversals leading to HAB favourable conditions, owing to longshore bloom transport, are paired pressure sensors installed around capes, as discussed here on p. 276. Such sensor pairs also represent a low-cost, low-maintenance forecasting tool, which may be more easily achievable than installation and maintenance of an HF radar system. Chemical parameters of advantage that may be measured autonomously are the inorganic nutrients nitrate, silicate, phosphate, and ammonium: the immediate utility of such data would be as input to a probabilistic “ecological window” prediction model. While strictly falling under the remit of other chapters in this volume, real-time bottom oxygen data is of great importance to the short term prediction of high biomass bloom impact in critical locations such as Elands Bay in the southern Benguela. Real-time bottom water oxygen sensors can be considered the single most important segment of an operational network for forecasting the locally devastating impact of HABs on organisms vulnerable to hypoxic events. Satellite based observations A variety of satellite derived data, allowing description of both biological and physical variables, is required for HAB forecasting. Ocean colour-based sensors offer the most immediately useful HAB-specific data, providing synoptic maps of surface algal biomass on a near real-time basis. In addition to traditional empirically derived chlorophyll a products, the use of satellite ocean colour data in conjunction with reflectance inversion algorithms offer experimental geophysical products that are specifically designed for HAB monitoring application. These include more accurate estimates of algal biomass in high biomass waters and products that allow some form of assemblage description independent to biomass. These include assemblage size descriptors such as the cellular effective diameter and a fluorescence quantum yield
288
S. Bernard et al.
product of potential utility as a physiological proxy (Bernard 2005). While these and other ocean colour products offer tremendous potential for incorporation into an operational HAB forecasting system, several logistical and scientific hurdles will need to be overcome for routine operational use. These include enabling routine access to full spectral data from sufficient sensors to ensure daily coverage for the region (e.g., MERIS and MODIS) and development of optimal flagging and processing routines for application in the high biomass waters of the Benguela, a system complex with regard to both in-water bio-optics and atmospheric correction. HAB forecasting in the Benguela necessitates that satellite data pertaining to the dominant physical forcing mechanisms are also routinely available. Such data are essential to determining bloom pre-conditioning, transport, and longevity and are likely to play an important role in any probabilistic forecasting model in addition to potential assimilation into hydrodynamic models. Sea surface temperature (SST) and the surface wind field are considered the most important variables and are available from a range of space based platforms. These include the NOAA AVHRR and NASA MODIS missions for SST data (daily, ~ 1 km spatial resolution) and the NASA QuikSCAT mission for surface winds (daily, ~ 25 km spatial resolution, no coverage within 50 km from coast). The availability of new or improved remotely sensed products is also to be expected. For example, a mission that has the potential to provide valuable high frequency SST data is the ESA MeteoSat Second Generation (MSG) mission (15 minutes sampling frequency, 3 km spatial resolution). Platforms and observation systems The establishment of permanent multi-sensor moorings, providing the biological and physical data necessary for bloom detection and transport prediction, is critical to establishing a HAB forecasting capability in the Benguela. Mooring platform payloads have been discussed above; an additional and important consideration is the location of the mooring(s). In the southern Benguela, HAB formation and retention is in response to generally well defined physical processes (Probyn et al. 2000, Pitcher and Weeks, this volume) and mooring location should be dictated primarily by the shelf circulation patterns associated with the various stages of bloom development, maintenance, and decay. A potential complementary approach to the use of fixed platforms is the use of autonomous underwater vehicles (AUVs), sampling platforms that are becoming increasingly more effective as instrument systems are miniaturized, deployment times are extended, and costs are lowered. These platforms, in particular low power gliders, have the capability to provide high frequency spatial data pertaining to HAB detection. However, whilst the effectiveness of such platforms has been demonstrated elsewhere (Glenn et al. 2004), logistical demands in their use are still considered too high for operational use in the Benguela in the near future. Current HAB related efforts in the southern Benguela have focused on the provision of a lightweight multi-sensor platform servicing the Namaqua shelf, carrying paired hyperspectral radiometers, a fluorometer, thermistor chain, and an Acoustic Doppler Current Profiler (ADCP). The system is designed as a lightweight, low cost, multiple
Requirements for forecasting harmful algal blooms in the Benguela
289
buoy mooring employing GSM telemetry and designed to be serviceable from small boats. An example of the effectiveness of the approach relative to bloom detection is presented in Figure 12-2, detailing the detection of a bloom dominated by the dinoflagellate Prorocentrum triestinum, observed in data derived from both a prototype mooring and the Medium Resolution Imaging Sensor (MERIS) ocean colour sensor. An operational observation network in the southern Benguela, effective for the prediction of HABS in the Greater St. Helena Bay region (see Pitcher and Weeks, this volume), can be briefly considered as an example. The network is considered to consist of a minimum of three moorings in combination with the satellite data discussed above. Two inshore moorings, configured in a similar manner to the example platform discussed above, should be located in ~50 m water depth offshore from the Olifants river mouth and Lamberts Bay, allowing the early detection of surface blooms moving in a typical southerly direction during quiescent periods (Probyn et al. 2000). A third midshelf mooring located at mid-latitude relative to the inshore moorings is intended to provide meteorological and physical data for ecological window assessment, data assimilation, and modeling validation studies. A schematic of the system can be seen in Figure 12-3. Finally, it must be realised that there is still a great deal of uncertainty regarding the mechanisms underlying HAB formation: the research value of an effective coastal observation network providing multi-sensor data across a wide range of spatial and temporal scales cannot be understated. NUMERICAL MODELLING AND PREDICTION OF HAB DYNAMICS The availability of appropriately configured and validated circulation models, specific to each sub-ecosystem region of impact, are vital to establishing an operational HAB forecasting capability. Accurate simulation of the processes responsible for shoreward bloom transport and coastal retention are the most important criteria for establishing model efficacy. These processes can be summarised as the response to onshore winds, baroclinic instabilities, poleward pressure driven flows, and non-linear internal waves. Preliminary modelling efforts in the region have focussed on the southern Benguela, employing a 3D numerical hydrodynamic model using Delft3D-FLOW (Pitcher et al. in press; WL|Delft Hydraulics, 2003a) for the Greater St. Helena Bay region. Initial validation studies appear to show that the temporal and spatial characteristics of physical variability in the St. Helena Bay – Namaqua system can be simulated with uncertainties approximating 20%. While demonstrating the feasibility of simulating the primary circulation characteristics of the system, the study also reveals the complex dynamical characteristics of circulation and stratification and aids in establishing model requirements for operational forecasting. Current modelling efforts require improvement through the use of dynamic boundary conditions, e.g., the nested ROMS (Regional Ocean Model System, Song and Haidvogel 1994) model approach and more extensive validation data from both a
290
S. Bernard et al.
Figure 12-3. Schematic of proposed HAB forecasting system for the southern Benguela.
temporal and spatial perspective. Validation data should not be restricted to surface fields and should include data from throughout the water column. Simulation and process oriented studies should be conducted with validated models to identify the dominant dynamics underlying the initiation and transport of HABs. Effective data assimilation requires the establishment of critical thresholds in observation; achieving such a critical threshold is likely to require considerable additional investment in observation network structure in order to provide data with a high enough spatial density. However, the first aim of a simple operational model is to provide a prediction of bloom shoreline impact and retention through a bloom tracking approach, given initial offshore bloom magnitude and delineation from near real-time buoy and satellite data. This capability, where real-time observations are utilised to determine initial model conditions for a bloom tracking simulation, should be distinguished from a true data assimilation capability, which is considerably more complex and observationally resource intensive. Further modelling studies should therefore include particle and bloom tracking studies, supported by similar efforts in the field. It is recommended that platform-related sources of uncertainty identified in initial southern Benguela modelling studies be addressed by transferring to the ROMS
Requirements for forecasting harmful algal blooms in the Benguela
291
platform, where the required boundary condition resolution can be achieved through nesting. There is also much to gain by considering several model structures, ranging in complexity from the simple 2D to complex 3D formulations. A simple alternative to the use of complex 3D models running in real-time is the use of a pre-generated offline catalog of circulation scenarios; such a facility used in conjunction with available real-time data would offer a simple and robust first approximation of potential shoreline impact. CONCLUSIONS The substantial complexity of the biological response amongst HAB organisms (e.g., variability in swimming, growth, mortality, nutrient and light acquisition, etc.) and the inherent stochasticity of species selection and bloom events means that it is unlikely that realistic predictions of species-level bloom formation can be achieved in the near future. The forecasting scheme outlined here provides an alternative approach by utilising real-time observation systems to replace the need to model biological aspects of HAB development. It appears that that the most pragmatic means for the prediction of biological HAB attributes in the Benguela at this time is a statistical “ecological window” approach, whereby physical, chemical, and biological real-time data are used to predict broad taxonomic groups (e.g., diatom versus dinoflagellate dominance, based on turbulence criteria). The future availability of species-targeting real-time observation techniques would substantially improve the ability to identify and predict taxonomic HAB aspects. Until such time, routine identification of toxic HAB blooms will require traditional sampling methods (direct cell counts and toxin assays), since these HAB events fall into the category of species-level forecasting. The principal considerations with regard to the effects of physical processes on HAB forecasting can be summarised as follows. Three dimensional coastal morphology, and the spatial and temporal structure of the wind field, are the principal determinants of initial physical conditions. Shoreline impact of HABs, a key forecasting parameter, is determined principally by baroclinic instabilities, onshore winds associated with wind reversals, and southerly alongshore transport near capes during relaxation. With regard to surface transport in upwelling systems, alongshore transport dominates across-shore transport. Consideration must be made of variability in both local and remote forcing, e.g., coastally trapped waves, and the impact of these on HAB transport in local flow fields. New sets of observations are also required, most importantly high spatial resolution wind stress, and the deployment of pressure sensors around capes, given the importance of wind reversals in the vicinity of these features. Finally, it can be concluded that while bloom and retention are amenable to prediction, the processes underlying precondition and formation are unlikely to be sufficiently characterised to enable prediction in the near future. An operational, real-time observation network is critical to the success of regional HAB forecasting schemes and a combination of multi-sensor coastal moorings and satellite data are required. Offering both biological and physical data, dedicated HAB moorings should provide hyperspectral reflectance data allowing the derivation of
292
S. Bernard et al.
algal assemblage and physiological descriptors, surface fluorometer-derived algal biomass, and profiles of temperature and currents from thermistor chains and ADCPs. Bottom oxygen sensors are important in locations prone to hypoxic events. Additional chemical and meteorological sensors would be highly beneficial, as would be pressure sensors in the vicinity of capes, and the availability of high frequency coastal radar. Greater continuity and greater spatial coverage are also needed. At present there are a maximum of two moorings in the entire Benguela providing real-time HAB related data. Routine availability of appropriately processed satellite data is essential and should consist of multi-sensor ocean colour, SST and wind data. Species-targeting real-time observation techniques, needed for the prediction of toxin related impacts, are unlikely to be routinely available within the next few years. The establishment of appropriately configured and validated circulation models, specific to each sub-ecosystem region of impact, is vital to any HAB forecasting capability. The first aim of a simple operational model is to provide a prediction of bloom shoreline impact and retention through a bloom tracking approach, given initial offshore bloom magnitude and delineation from near real-time buoy and satellite data. Simple offline circulation catalogs can be used to provide initial first approximation predictions of shoreline impact. Ultimately, it would be hoped to establish a forecasting structure operating in near real-time, utilising probabilistic ecological models and complex nested 3D physical models in conjunction with data from a variety of multi-sensor observation platforms. REFERENCES Allen J.S., P.A. Newberger and J. Federiuk. 1995. Upwelling circulation on the Oregon continental shelf. Part 1: Response to idealized forcing. J. Phys. Oceanogr. 25(8): 1843-1866. Allen J.S., and P.A. Newberger. 1996. Downwelling circulation on the Oregon continental shelf. Part 1: Response to idealized forcing, J. Phys. Oceanogr. 26(10): 2011-2035. Austin J.A., and S.J. Lentz. 2002. The inner shelf response to wind-driven upwelling and downwelling. J. Phys. Oceanogr. 32(7): 2171-2193. Bakun, A., and C.S. Nelson. 1991. The seasonal cycle of the wind-stress curl in the sub-tropical eastern boundary current regions. J. Phys. Oceanogr. 21:1815-1834. Bargu, S., K. Lefebvre, and M.W. Silver. In press. Feeding responses of Euphausia pacifica on non-toxic Pseudo-nitzschia pungens in the presence of added dissolved domoic acid. Bates S.S. 1998. Ecophysiology and metabolism of ASP toxin production. 405–426 in: Anderson DM, A.D. Cembella, G.M. Hallegrae, eds. Physiological ecology of harmful algal blooms. NATO ASI series, vol G41, Springer, Berlin Heidelberg New York. Bernard, S. 2005. The bio-optical detection of harmful algal blooms. PhD thesis, Department of Oceanography, University of Cape Town, Private Bag, Rondebosch 7701, Cape Town, South Africa, 232p. Chelton, D.B., M.G. Schlax, M.H. Freilich and R.F. Milliff. 2004. Satellite measurements reveal persistent small-scale features in ocean winds. Science 303:978-983. Cullen, J.J., A.M. Ciotti, R.F. Davis, M.R. Lewis. 1997. Optical detection and assessment of algal blooms. Limnol. Oceanogr. 42(5): 1223-1239. Daly, K.L., R.H. Byrne, A.G. Dickson, S.M. Gallager, M.J. Perry, and M.K. Tivey. 2004. Chemical and biological sensors for time-series research: current status and new directions. Mar.Tech. Soc. Journ. 33(2): 121-143. Donaghay, P.L. and T.R. Osborn. 1997. Toward a theory of biological-physical control of harmful algal bloom dynamics and impacts. Limnol. Oceanogr. 42:1283-1296.
Requirements for forecasting harmful algal blooms in the Benguela
293
Fennel,W. 1988. Analytical theory of the steady-state coastal ocean and equatorial ocean. J. Phys. Oceanogr. 18:834-850. Fennel,W. 1999. Theory of the Benguela upwelling system. J. Phys. Oceanogr. 20:177-190. Fiedler, P.C. 1982. Zooplankton avoidance and reduced grazing responses to Gymnodinium splendens (Dinophyceae). Limnol. Oceanogr. 27:961-965. Gan J.P. and J.S. Allen. 2002. A modeling study of shelf circulation off northern California in the region of the Coastal Ocean Dynamics Experiment: Response to relaxation of upwelling winds. J. Geophys. Res. 107 (C9): art. no. 3123. Glasgow H.B., J.M. Burkholder, R.E. Reed, A.J. Lewitusb and J.E. Kleinman. 2004. Real-time remote monitoring of water quality: a review of current applications, and advancements in sensor, telemetry, and computing technologies. J. Exp. Mar. Biol. Ecol. 300:409–448. Glenn, S., O. Schofield, T.D. Dickey, R. Chant, J. Kohut, H. Barrier, J. Bosch, L. Bowers, E. Creed, C. Haldeman, E. Hunter, J. Kerfoot, C. Mudgal, M. Oliver, H. Roarty, E. Romana, M. Crowley, D. Barrick, and C. Jones. 2004. The expanding role of ocean color and optics in the changing field of operational oceanography. Oceanography 17(2), Special Issue: Coastal Ocean Optics and Dynamics, 85-95. Juhl, A.R., V.L. Trainer, M.I. Latz. 2001. Effect of fluid shear and irradiance on population growth and cellular toxin content of the dinoflagellate Alexandrium fundyense. Limnol. Oceanogr. 46:758–764. Koracin, D., C.E. Dorman, and E.P. Dever. 2004. Coastal perturbations of marine layer winds, wind stress, and wind stress curl along California and Baja California in June 1999. J. Phys. Ocean. 34:1152-1173. Langebrake, L.C. 2003. AUV sensors for marine research. In: Griffiths, G.. ed. Technology and Applications of Autonomous Underwater Vehicles. Taylor and Francis, London, pp. 245– 277. Margalef, R., M. Estrada, and D. Blasco. 1979. Functional morphology of organisms involved in red tides, as adapted to decaying turbulence. 89-94 in Taylor, D.L., and H.H. Selige, eds. Toxic Dinoflagellate Blooms.. New York, Elsevier. McCreary, J.P. 1981. A linear stratified ocean model of the coastal undercurrent. Phil. Trans. Roy. Soc., A 302:385-413. McManus, M.A., A.L. Alldredge, A.H. Barnard, E. Boss, J.F. Case, T.J. Cowles, P.L. Donaghay, L.B. Eisner, D.J. Gifford, C.F. Greenlaw, C.M. Herren, D.V. Holliday, D. Johnson, S. McIntyre, D.M. McGhee, T.R. Osborn, M.J. Perry, R.E.Pieper, J.E.B. Rines, D.C. Smith, J.M. Sullivan, M.K. Talbot, M.S. Twardowski, A. Weidemann and J.R. Zaneveld. 2003. Characteristics, distribution and persistence of thin layers over a 48 hour period. Mar. Ecol. Progr. Ser. 261:1-19. Ochoa, J.L. 2003. ENSO phenomenon and toxic red tides in Mexico. Geofisica Internacional 42:505-515. Pitcher, G.C., and A.J. Boyd. 1996. Cross-shelf and along-shore dinoflagellate distributions and the mechanisms of red tide formation within the southern Benguela upwelling system. In Harmful and Toxic Algal Blooms. Yasumoto, T., Y. Oshima, and Y. Fukyo, eds. Paris, Intergovernmental Oceanographic Commission of UNESCO, 243-246. Pitcher, G.C. and D. Calder. 2000. - Harmful Algal Blooms of the southern Benguela current: A review and appraisal of monitoring from 1989-1997. S. Afr. J. mar. Sci. 22:255-271. Pitcher, G., P. Monteiro and A. Kemp. In Press – The potential use of a hydrodynamic model in the prediction of harmful algal blooms in the southern Benguela. In Harmful and Toxic Algal Blooms. Steidinger, K., ed. Intergovernmental Oceanographic Commission of UNESCO. Pitcher, G.C. and S. J. Weeks. 2006 – The variability and potential for prediction of harmful algal blooms in the southern Benguela ecosystem. Chapter 7 in Shannon, V., G. Hempel, P. Rizzoli, C. Moloney, and J. Woods,, eds. Benguela: Predicting a Large Marine Ecosystem., Elsevier.(this volume) Pitcher, G.C., A.J. Boyd, D.A. Horstman, and B.A. Mitchell-Innes. 1998. Subsurface dinoflagellate populations, frontal blooms and the formation of red tide in the southern Benguela upwelling system. Mar. Ecol. Prog. Ser. 172:253-264. Probyn, T.A., G.C. Pitcher, P.M.S Monteiro, and G. Nelson. 2000. Physical processes contributing to harmful algal blooms in Saldanha Bay, South Africa. S. Afr. J. mar. Sci. 22:285-297 Roesler, C.S., and E. Boss. 2003. Spectral beam attenuation coeffecient retrieved from ocean colour inversion. Geophys. Res. Lett. 30(9). Roesler, C.S., S.M. Etheridge and G.C. Pitcher. In Press – Application of an ocean color algal taxa detection model to complex red tides in the southern Benguela. In Harmful and Toxic Algal Blooms. Steidinger, K., ed. Intergovernmental Oceanographic Commission of UNESCO.
294
S. Bernard et al.
Sacau-Cuadrado, M., P. Conde-Pardo, and P. Otero-Tranchero. 2003. Forecast of red tides off the Galician coast. Acta Astronautica. 53:439-443. Sellner K.G., G.J. Doucette and G.J. Kirkpatrick. 2003. Harmful algal blooms: causes, impacts and detection, J. Ind. Microbiol. Biotechnol. 30:383–406. Send U., Beardsley R.C. and Winant C.D. 1987. Relaxation from upwelling in the coastal ocean dynamics Experiment. J. Geophys. Res. 92 (C2): 1683-1698. Sieracki C.K., M.E. Sieracki, C.S. Yentsch. 1998. An imaging-in-flow system for automated analysis of marine microplankton. Mar. Ecol. Progr. Ser. 168:285-296. Scholin CA, R. Marin III, P.E. Miller, G.J. Doucette, C.L. Powell, P. Haydock J. Howard and J. Ray J., 1999. DNA probes and a receptor-binding assay for detection of Pseudo-nitzschia (Bacillariophyceae) species and domoic acid activity in cultured and natural samples. J. Phycol. 35:1356–1367. Scholin, C., E. Vrieling, L. Peperzak, L. Rhodes, and P. Rublee. 2003. Detection of HAB species using lectin, antibody and DNA probes. In: Hallegraeff, G.M., D.M. Anderson, and A.D. Cembella, eds., Manual on Harmful Marine Microalgae. UNESCO. 793 p. Smayda, T.J. 2000. Ecological features of harmful algal blooms in coastal upwelling ecosystems. S. Afr. J. mar. Sci. 22:219-253. Smayda, T.J. 2002. Turbulence, watermass stratification and harmful algal blooms: an alternative view and frontal zones as “pelagic seed banks.” Harmful Algae 1:95-112. Smayda, T.J. and C.S. Reynolds. 2001. Community assembly in marine phytoplankton: application of recent models to harmful dinoflagellate blooms. J. Plankton Res. 23:447-461. Song, Y.T. and D.B. Haidvogel. 1994. A semi-implicit ocean circulation model using a generalized topography-following coordinate system. J. Comp. Phys. 115:228-244. Sullivan, J.M., E. Swift, P.L. Donaghay and J.E.B. Rines. 2003. Small-scale turbulence affects the division rate and morphology of two red-tide dinoflagellates. Harmful Algae 2:183-199. Tilburg C.E. 2003. Across-shelf transport on a continental shelf: Do across-shelf winds matter?, J. Phys. Oceanogr. 33:2675. Turner, J.T., and P.A. Tester. 1997. Toxic marine phytoplankton, zooplankton grazers, and pelagic food webs. Limnol. Oceanogr. 42:1203-1214. WL|DELFT HYDRAULICS (2003a). DELFT3D-FLOW User Manual 2003 Version 3.10. WL|Delft Hydraulics, Delft, The Netherlands. Yin, K. 2003. Influence of monsoons and oceanographic processes on red tides in Hong Kong waters. Mar. Ecol. Prog. Ser. 262:27-41. Young, E.B. and J. Beardall. 2003. Rapid ammonium- and nitrate-induced perturbations to Chl a fluorescence in nitrogen-stressed Dunaliella Tertiolecta (Chlorophyta). J. Phycol. 39:332–342.
Large Marine Ecosystems, Vol. 14 V. Shannon, G. Hempel, P. Malanotte-Rizzoli, C. Moloney and J. Woods (Editors) © 2006 Elsevier B.V. All rights reserved.
295
13 Low Oxygen Water (LOW) Forcing Scales Amenable to Forecasting in the Benguela Ecosystem P .M. S. Monteiro, A. K. van der Plas, G. W. Bailey, P. Malanotte-Rizzoli, C. M. Duncombe Rae, D. Byrnes, G. Pitcher, P. Florenchie, P. Penven, J. Fitzpatrick, H.U. Lass
INTRODUCTION Episodic Low Oxygen Water (LOW) events (day – seasonal time scales) in the Benguela system impact on ecosystem properties such as habitat suitability for the different life cycle stages of economically important marine resources (Monteiro et al. 2004; van der Lingen et al. Chapters 8 and 14 this volume). The timing, persistence and spatial extent of such events in the recent past have led to major mortality and shifts in distribution of economically important living marine resources (hake in Namibia 1992-1994; rock lobster in South African west coast: 1990s) which impacted on fisheries yields and continue to be a source of uncertainty in forecasting fisheries total allowable catches (TACs). Stock assessment models, used to forecast fishery TAC and the potential future economic yields from a fishery, incorporate the effects of environmental uncertainty such as LOW as a random contribution to mortality. This may have been justified while there was a weak understanding of the actual scales of LOW variability and how they are governed by complex interacting processes. Recent work on long term data sets indicates that LOW variability is not a random effect but can be characterised by specific frequency, persistence and intensity scales (Monteiro et al. 2004). The incidence of event scale (days) seasonal LOW variability (weeks to months duration) and its ecosystem impacts responds to longer time scales (decadal) of remote forcing. Moreover, improved understanding of the key processes and their linkages and response to short and long term forcing scales is beginning to improve the feasibility of reliable forecasting (Monteiro et al. 2004; Monteiro and van der Plas, Chapter 5 this volume). A requirement for forecasting LOW variability and its impact on the economic value of ecosystem services and living marine resources is an ability to deal with the extreme range of scales involved. Thus, while basin scales govern shelf boundary conditions on seasonal to decadal time scales, local event scales modulate the variability by
296
P.M.S. Monteiro et al.
amplifying or weakening the oceanic signal. The biogeochemical response that drives habitat suitability appears to be sensitively dependent on the linkages across these scales. Thus the ecologically important responses depend critically on the interaction of these largely uncoupled remote and local time and space scales of forcing. The challenge is to deal with the implicit complexity of linkages and deliver a system that is able to serve the aggregated scale needs of ecosystem management without losing its scientific validity. Globally, there is increasing concern related to the apparent unpredictability of natural LOW events and their impact on economically important ecological services (Turner and Rabalais 2001; Diaz and Rosenberg 1995). Of particular concern is the possibility that changes to the LOW related oceanic physical and biogeochemical processes may be related to global change (Grantham et al. 2004). Recent experiences from other upwelling systems indicate that the scales of LOW variability are essential to understanding and predicting changes in ecosystem characteristics (Grantham et al. 2004; Service 2004) The focus of this working group discussion synthesis is to investigate time scales relevant to forecasting LOW events and not the ecological or biological impacts of the variability. The working group relied significantly on the knowledge setting for low oxygen variability (Chapter 5 this volume), Harmful Algal Blooms (HAB)(Chapter 12 this volume), and physical oceanographic processes (Chapter 4 this volume). The ecological impact issues are addressed in a companion study by van der Lingen et al.(Chapter 14 this volume). SCALES OF LOW VARIABILITY AMENABLE TO FORECASTING LOW forecasting depends sensitively on an adequate degree of confidence in respect of the model outputs over the time scale of interest. This in turn depends on the extent to which the model set up stably integrates the required processes, the boundary conditions and responds to the forcing scales. While statistical checks of model reliability are essential good practice, verification through the hind casting of known events is a required test of confidence. Hind casting of significant LOW events in the Benguela, especially in the physically more complex central and southern sub-systems is an essential part of the proposed approach. The approach to hind casting in the Benguela would concentrate on the 1992-1994 LOW events in Namibia and the large rock lobster walkouts in early 2002 in the southern Benguela for which good verification data exists. Assuming that it is possible to hind cast significant historical events, the following scales were considered to be amenable to prediction: • Short time scale: 7 day lead events linked to “harmful algal blooms (HABs)” • Medium Time Scale: 2 month lead events linked to the “eastern tropical south Atlantic (ETSA)” • Decadal Time Scale: in ‘what if’ scenarios where the uncertainties are large in respect to forecasting forcing variability.
Low oxygen water forcing scales amenable to forecasting in the BCLME
297
Three different regimes of LOW variability were recently recognized in the BCLME (Monteiro and van der Plas, Chapter 5 this volume): • Northern (4o – 17oS): Narrow shelf, LOW variability directly coupled to boundary conditions (ETSA) • Central (17o– 27oS): Complex dynamics between Cape Frio and Lüderitz upwelling centres and different boundary conditions • Southern (27o– 35oS): variability depends strongly on long term (decadal) local equatorward wind regimes as well as but the incidence and collapse of high biomass Harmful Algal Blooms governs the event scale. The northern BCLME is most susceptible to variability in the decadal scale of LOW characteristics in the ETSA system. This arises because of the combination of proximity of the ETSA LOW boundary to the narrow shelf and the seasonality of upwelling which advects the LOW onto the shelf (Monteiro and van der Plas, Chapter 5 this volume). LOW variability in the ETSA domain was shown to be dominated by a decadal time scale (Monteiro and van der Plas, Chapter 5 this volume). Forecasting LOW variability in the northern BCLME is therefore largely linked to the coupling of seasonal upwelling fluxes that bring the LOW onto the shelf and the scenario based forecasting of LOW in the equatorial system. In the central BCLME system (Namibian shelf), a combination of remote forcing of low oxygen waters (high salinity) from the Eastern Tropical South Atlantic oxygen minimum zone and local formation from deposition of planktonic detritus throughout the year combine to control low oxygen water formation. The remote forcing part is thought to be strongly influenced by the southward penetration of the warm equatorial surface water driven by seasonal and ENSO related relaxation of equatorial easterly winds (Monteiro et al. 2004). The propagation time of the baroclinic wave results in a 2 month forecasting time scale (Florenchie et al. 2003). The forecasting time scale relates to the strengthening of the sub-thermocline poleward flow that carries the LOW signal with it and sets up the initial conditions for the seasonal shelf hypoxia / anoxia. The preferential exchange between the slope and the shelf at the main upwelling centres of Cape Frio and Lüderitz which are also regions of maximum meridional wind stress (Figure 13-1) on the shelf. The region off Lüderitz (26oS) has been identified as a barrier (Monteiro 1996) to advection into the southern BCLME of remotely-sourced Central Waters of a northern and therefore higher salinity content (Duncombe Rae 2005)(Figure 13-2). Later research conducted by van der Plas et al. (2004) and others have stressed the role of sediment oxygen demand in the more severely oxygendepleted bottom waters on the shelf of the northern Benguela system. The organic rich anoxic mud belt, which is maintained by a combination of physics and biological production, drives a flux of reduced compounds such as hydrogen sulphide, ammonium and methane from interstitial waters into the overlying water column and contributes to maintaining hypoxia in bottom waters (Monteiro et al. 2005). The most useful forecast scale for the seasonal and decadal (Benguela Niño) LOW variability in the central Benguela is suggested to be related to the 2 month lag between the relaxation of equatorial winds and the intensification of poleward flow in the central Benguela.
298
P.M.S. Monteiro et al.
Figure 13-1. Map showing the spatial distribution of the QuickScat averaged meridional wind stress component in the BCLME in the summer of 2003-2004. It shows that the meridional wind stress peaks close to shore at the location of the Cape Frio (17oS), Luderitz (27oS) and Cape Columbine upwelling centres. This spatial distribution supports the role of these discrete sites in the slope-shelf flux of upwelled water. QuikScat data are produced by Remote Sensing Systems and sponsored by the NASA Ocean Vector Winds Science Team. Data are available at www.remss.com
The southern BCLME is thought to be most amenable to the 7-day forecasting time scale due to the importance of the combined effects of local wind forcing regimes and HAB related event scale (Monteiro and van der Plas, Chapter 5 this volume; Bernard et al., this volume; Shillington et al. Chapter 4 this volume).
REMOTE EQUATORIAL FORCING: 2 MONTH FORECASTING SCALE The basis for the 2 month forecasting time scale of the equatorial forcing signal is derived from the hypothesised coupling of the warm surface waters and sub thermocline (Monteiro et al. 2004). The effect of this interaction is the enhancement of the poleward component of sub-thermocline waters that carry LOW along the slope and on the shelf. The two month time scale is the time taken for an equatorial baroclinic Kelvin wave to reach the Benguela (Florenchie et al. 2003).
Low oxygen water forcing scales amenable to forecasting in the BCLME
299
Figure 13-2: A spatial distribution of the fractional contribution by high salinity tropical central waters to the general body of SACW in the BCLME. It shows that the Lüderitz region is a boundary between the polar and tropical central water inputs. While Angola and the region north of Cape Frio are characterised by >90% water of tropical origin, the Namibian shelf is a mix (20 – 80%) of tropical and Cape basin SACW. Luderitz forms an effective boundary between the tropical and sub-Antarctic SACW both on the slope and the shelf. The southern Benguela is largely characterized by SACW of sub-Antarctic origin. This salinity distribution is a good predictor of source water oxygen concentrations boundary conditions for the Benguela Shelf.
In terms of the warm surface water-enhanced poleward flow model, the southward propagation of the Kelvin wave does not in itself transport LOW. Rather, the enhancement of the thermocline strength by the warm water strengthens the poleward flow below the thermocline which carries the LOW signal. The propagation of the equatorial baroclinic Kelvin wave has two main time scales of variability: • Annual – Seasonal: the seasonal relaxation of the equatorial trade winds in the eastern Atlantic which coincides with the transit of the intertropical convergence zone (ITCZ) over the equatorial area. This results in two relaxation periods; the stronger in the austral late summer (Feb – Mar) and the weaker (Sept – Oct) (Stramma and Schott 1999) • Interannual – decadal Benguela Niño: the relaxation of the easterly winds in the western Atlantic Ocean (Florenchie et al. 2003).
300
P.M.S. Monteiro et al.
It is presently unclear how these two scales interact but both have been separately investigated (Shannon and Nelson 1996; Florenchie et al. 2003). While the forcing of the Benguela Niño has been investigated using a combination of both modelling and observational data (Florenchie et al. 2003), the seasonal scale has only been monitored through its expression as a poleward surface warm water plume in the late summer and its impact on the Angola – Benguela front (Shannon and Nelson 1996). The two components of the LOW variability that are most susceptible to this forcing are: • Impact that it has on the poleward propagation of LOW on both shelf and slope • Impacts that it has on the LOW characteristics at the Lüderitz upwelling centre which is also the boundary between tropical hypoxic Central water and the polar aerated Central water (Monteiro et al. 2004; Duncombe Rae 2005) The likely corollary is that Benguela Niño periods that are characterised by the strong southward propagation of warm equatorial water will also be periods of enhanced LOW. The dependency of this forecasting capability on the validity of the hypothesis on the interaction between the southward propagation of warm water and the semipermanent LOW ETSA pool makes it necessary to develop this capability through an intermediate step which explores its validity. It was proposed that a modelling approach be adopted to test the various hypotheses through sensitivity analyses of a range of forcing and response scales. The proposed modelling approach makes use of an idealised basin domain and two layer schematization to test the dynamics behind the generation and propagation of the equatorial baroclinic Kelvin wave (Figure 13-3). The proposed schematization includes the main Equatorial undercurrents (Equatorial Undercurrent (EUC), South Equatorial Undecurrent (SEUC), South Equatorial Countercurrent (SECC)) prescribed as boundary conditions. The remainder of the boundary conditions are proposed to be “sponge” type with relaxation to climatological t, S characteristics. The domain covers the South-eastern Atlantic Ocean basin and extends from 0 – 40oS and 10oW to the South-west African coastline. Three modelling experiments are proposed to test the interaction and its possible impact on LOW transport in the Benguela. • Modelling Experiment 1: Objective: to understand the forcing scales and propagation that govern the characteristics and impacts of the seasonal (austral later summer / spring) equatorial baroclinic Kelvin waves. • Modelling Experiment 2: Objective: to test the hypothesis that the seasonal and Benguela Niño warm water intrusions strengthen the sub-thermocline poleward flow on both the slope and the shelf. • Modelling Experiment 3: Objective: to run a number of event scale, seasonal and interannual scenarios with fully coupled remote and local forcing to investigate the critical scales that govern their linkage.
Low oxygen water forcing scales amenable to forecasting in the BCLME
301
Figure 13-3. A schematised model set up which will form the basis for the modelling experiments to test the relationship between equatorial forcing and LOW variability in the BCLME.
The successful completion of this elucidation process would provide a basis to set up an operationalised 2-month forecasting system that could provide management with an index of the intensity and spatial extent of summer hypoxia characteristics. Coupled to improved understanding of the behavioural responses of the planktonic and fish life stages, this could drive an optimisation of an ecosystem exploitation strategy. Improved understanding of ecosystem level responses and more reliable forecasting could provide useful support to risk management strategies for both regulators and industry. This could improve the effectiveness of mitigation plans and contribute to both socio-economic resilience and adaptation in the context of climate variability and change. SHELF SCALE FORCING: 7 DAY FORECASTING SCALE It is on the 7-day event scale, that local formation of low oxygen water in association with the dynamics and collapse of harmful algal blooms is most important and perhaps more demanding, for forecasting purposes. This is particularly relevant in the Lamberts Bay / Elands Bay region where Cockcroft et al. (2000), Cockcroft (2001) and van der Lingen et al. (Chapters 8 & 14, this volume) - have described the incidence of low oxygen water-induced rock lobster mortalities, most of which occurred in
302
P.M.S. Monteiro et al.
association with collapse of a harmful algal bloom. The classical cycle of HAB development is a cessation in SE wind forcing, HAB development (Pitcher and Boyd 1996), further depletion of water column dissolved oxygen by organic matter sedimenting from the HAB and finally exacerbation of low oxygen waters on the bottom by collapse of the bloom (Bernard et al. Chapter 12 this volume). A two phase model of low oxygen water development on the Benguela shelf has been proposed (Figure 13-4) consisting of low oxygen water “reservoir” which varies on the seasonal scale on the mid-shelf and a HAB accumulation and collapse zone which operates on a much shorter time scale of around 7 days on the inner shelf.
Onshore and poleward Transport of HAB Bloom HAB Bloom accumulation and collapse
Seasonal low oxygen water reservoir
Hydrodynamically driven reduction of suitable bottom
THE TWO PHASE MODEL FOR ROCK LOBSTER WALKOUTS IN THE ST HELENA BAY SYSTEM
Figure 13-4. Shows the essential characteristics of the two phase coupled model of LOW which results in the episodic rock lobster walkouts in St Helena Bay. The seasonal reservoir developed in the retention zone and its eastern boundary moves onshore and offshore in accordance with the upwelling cycles. The rock lobster habitat gets gradually compressed in the inshore region. In the late summer the poleward transport associated with the narrow inshore flow transports HABs into the inshore zone which then collapses due to nutrient depletion. The local biogeochemical oxygen demand overwhelms the physical supply and mortalities and walkouts result.
In terms of this model, a seasonal LOW reservoir develops in response to the combined effects of hydrodynamic retention, stratification and upwelling forced flux of carbon production (Monteiro and Roychoudhury 2005). The LOW reservoir moves closer to and further from the near shore zone driven by the response of the
Low oxygen water forcing scales amenable to forecasting in the BCLME
303
thermocline to upwelling and resulting density gradients (Monteiro and Largier 1999; Probyn et al. 2000). The inshore transport of the seasonal LOW reservoir narrows the suitable habitat for the demersal and bottom species. During the late summer the upwelling relaxation events result in the transport and concentration of HAB along a narrow inshore strip directly over the narrowed habitat. The sedimentation of such a large carbon biomass results in a short term oxygen demand that overwhelms the physically driven flux of atmospheric oxygen into the upper layers (Figure 13-5). This can be clearly seen as a sharp drop in oxygen concentration that occurred following two large HAB events in February and March 2002 which also led to rock lobster walkouts and mortality in excess of 1000 tons. The resultant impact on the resources is therefore not driven singly by the seasonal LOW reservoir but by the coupling of the seasonal and event scales. The seasonal signal forces a behavioural response that increases the vulnerability of the herded organisms to the HAB event scale. This new understanding explains why the time series hitherto used to explain the rock lobster behaviour and mortality, only describe variability of the seasonal reservoir which accounts for the habitat reduction, but not for the actual walkout and mortalities (Figure 13-6).
St Helena Bay: Q1: 2002 2.5
Rock Lobster Walkouts
Oxy (ml/l)
2 1.5 1 0.5 0
26-Mar-02
19-Mar-02
12-Mar-02
05-Mar-02
26-Feb-02
19-Feb-02
12-Feb-02
05-Feb-02
29-Jan-02
22-Jan-02
15-Jan-02
08-Jan-02
01-Jan-02
Date
Figure 13-5. An oxygen concentration time series from St Helena Bay in 2002 showing the relationship between the seasonal LOW reservoir (0.5 – 1mll-1) and the HAB driven events that drive the actual walkouts. The seasonal LOW reservoir brings the oxygen concentrations to < 1mll-1 and the HAB driven event scale governs the variability below this threshold. The instrument was located at a depth of 22m in the western side of the Bay.
304
P.M.S. Monteiro et al.
Figure 13-6. A depth – time contour plot of oxygen concentrations in St Helena bay for a 20 year period 1983 – 2003. It shows that the seasonal reservoir has both a seasonal and a decadal character. The 1980’s were largely well aerated whereas the 1990’s were persistently hypoxic. The main difference between the regimes lies in the equatorward wind strength. They were weak in the 1980’s and strong in the 1990’s.
The system where the short term forecasting was thought to be of greatest use was the St Helena Bay – Namaqua region of the southern Benguela system. This region is known to be susceptible to event scale LOW variability with significant impacts on local populations of bottom dwelling rock lobster and in extreme events water column based resources as well (Cockcroft et al. 2000). The magnitude of the impacts of LOW events has been linked to the intensity of the events in respect of HSconcentrations in the bottom layer. IMPORTANCE OF COUPLED MECHANISMS One of the key concepts which have emerged from the recent advances in understanding and the discussions of the working group is the importance of an explicit treatment of system complexity to arrive at a forecasting capability. Specifically, it is the role of coupled mechanisms that provides the required understanding linked to both remote and local forcing (Monteiro et al. 2004; Monteiro and van der Plas, Chapter 5 this volume; van der Plas et al. in press). A shift from a correlation to a process based forecasting capability will rely on the apparent randomness of LOW variability to be explained by the non-linear coupling of forcing and response on a wide range of scales. Of particular importance are the coupling and linkages of: • Equatorial easterly wind forcing and ETSA upwelling, thermocline strength and propagation of baroclinic Kelvin waves
Low oxygen water forcing scales amenable to forecasting in the BCLME
305
•
Coupling between changes in the southward transport of warm equatorial surface water and sub-thermocline poleward transport of LOW • Coupling of shelf boundary LOW with shelf based biogeochemical processes. Coupling of shelf based seasonal production fluxes and HAB events that govern near shore hypoxia / anoxia. One of the additional important points to emerge is that the complexity needs cannot be addressed by observational programmes alone. They will require the close alignment of both modelling and observational networks. OBSERVATIONAL PROGRAMME A modelling based LOW forecasting system needs to be supported by an observational programme to achieve the following three objectives: 1. to verify that the model set up reflects the process and variability adequately and that the process parameters are realistically calibrated 2. to verify that given an adequate model set up, it still provides outputs that reproduce observational time series. 3. to provide long term ongoing assessment that the underlying assumptions on which the model is built remain valid. An observational programme was proposed to support model verification rather than long term forecasting requirements. This is a necessary intermediate step to consolidate the required process level understanding, prior to advancing into a forecasting programme. Equatorial Domain It was proposed that the best approach would be to make use of an existing PIRATA buoy that is stationed at 0°; 0°, although it does not measure dissolved oxygen. However, the ideal placement for a current meter and oxygen sensor mooring array would be at 3 - 5°S. Benguela Slope - Shelf The proposed observational strategy to establish the circulation over the Benguela shelf system requires 3 mooring arrays at 400 - 500m depth. The objective of these moorings would be to clarify the variability of the poleward flow on the slope that forms the boundary conditions to the main upwelling centres at Cape Frio, Lüderitz and Columbine. The seasonal and interannual character of the poleward flow in response to equatorial forcing should become apparent. These moorings were considered essential. • 1 mooring at ~130m off Walvis Bay (existing IOW mooring) • 2 moorings at ~100m depth: (Lüderitz and St Helena Bay)
306
P.M.S. Monteiro et al.
These three moorings are required to test the shelf scale responses to the combined effects of variable and remotely forced boundary conditions and shelf based oxygen demand (local generation). It is essential that historic poleward undercurrent measurements collected between 1980 and 2005 be re-investigated in the light of new process and variability understanding. It was agreed that the Quickscat wind data was sufficient to provide wind forcing information, as long as it was continuously available. In addition, long wind time series exist for both Lüderitz and Cape Columbine. Remote sensing data (e.g. microwave data) is also readily available. SUMMARY Three scales of LOW variability were found to be amenable to forecasting as well as being suitable to provide information that would be of use to management response plans. • Short time scale: 7 day lead events linked to local wind variability and “HABs” • Medium Time Scale: 2 month lead events linked to “ETSA” easterly equatorial winds • Decadal Time Scale: in ‘what if’ scenarios where the uncertainties are large in respect to forecasting forcing variability. The areas where LOW forecasting can be of benefit include: • Input to stock assessment and ecosystem management models to provide a more sensitive indicator of the mortality / behavioural modification risk particularly when the variability is modulated by regime shifts. • Direct management response to a threat to stocks: rock lobster where it is practically possible to operate a contingency plan. • Ecosystem scale assessment of the risk posed by LOW to habitat suitability and trends in TAC size. This applies to a capability of assessing indirect impacts on stocks through impacts on the food web or on underlying services. • Climate change: the uncertainties linked to climate change are of importance because the time scales match those of the depreciation of capital investments in fisheries. Assessment of the risk of natural ecosystem degradation through increased impact of LOW from the equatorial system. Forecasting will only be a reality when the key gaps in process level understanding are addressed. This is underway and it is likely that hind casting will be possible by 2007 with true operational forecasting linked to ecosystem effects available by 2012.
ACKNOWLEDGEMENTS The authors acknowledge the support of their respective institutions (CSIR, University of Cape Town, NatMIRC, MCM, MIT, University of Maine, HydroQual, and Institute for Baltic Sea Research) that enabled their participation in the workshop. PMSM and AKvdP participation was funded through BCLME contract EV/LOW/02/04.
Low oxygen water forcing scales amenable to forecasting in the BCLME
307
REFERENCES Bernard, S., R.M. Kudela, P. Franks, W. Fennel, A. Kemp, A. Fawcett and G.C. Pitcher. (this volume). The requirements for forecasting harmful algal booms in the Benguela (Chapter 12). Cockcroft, A.C. 2001. Jasus lalandii “walkouts” or mass strandings in South Africa during the late 1990’s: an overview. Mar. Freshwater Res. 52:1085-1093. Cockcroft, A., D.S. Schoeman, G.C. Pitcher, G.W. Bailey, and D.C. van Zyl. 2000. A mass stranding of west coast rock lobster Jasus lalandii in Elands Bay, South Africa: Causes, results and applications. In: Von Kaupel Klein J.C. and F.R. Schram, eds. The biodiversity crises and crustaceans. Vol.11: 362 – 368 (Crustacean Issues). Diaz, R. J. and R. Rosenberg. l995. Marine benthic hypoxia - Review of ecological effects and behavioral responses on macrofauna. Oceanography and Marine Biology, Annual Review. 33:245-303. Duncombe Rae, C. M. 2005. A demonstration of the hydrographic partition of the Benguela upwelling ecosystem at 26°40'. African Journal of Marine Science 27:617-628.. Florenchie, P., R.E. Lutjeharms and C.J.C. Reason. 2003. The source of Benguela Niños in the South Atlantic Ocean. Geophysical Research Letters 30(10) [20 May]:12. Grantham, B.A., F. Chan, K.J. Nielsen, D.S. Fox, J.A. Barth, A. Huyer, J. Lubchenco,and B. Menge. 2004. Upwelling driven nearshore hypoxia signals ecosystem changes in the north Pacific. Nature. 429:749 – 754 Monteiro, P.M.S. 1996. The oceanography and biogeochemistry of CO2 in the Benguela upwelling system, PhD Thesis, University of Cape Town, South Africa. Monteiro, PMS and J.L. Largier. 1999. Thermal stratification in Saldanha Bay, South Africa, and subtidal, density driven exchange with coastal waters of the Benguela upwelling system. Estuarine, Coastal and Shelf Science 49(6): 877-890 Monteiro, P.M.S., G. Nelson, A. van der Plas, E. Mabille, G.W. Bailey, E. Klingelhoeffer.2005. Internal tide-shelf topography interactions as a potential forcing factor governing the large scale sedimentation and burial fluxes of particulate organic matter (POM) in the Benguela upwelling system. Continental Shelf Research 25: 1864- 1876. Monteiro, P.M.S. and A. Roychoudhury. 2005. Spatial distribution of trace metals in an eastern boundary upwelling retention area (St. Helena Bay, South Africa): A hydrodynamic-biological forcing hypothesis. Estuarine Coastal and Shelf Science 65:123-134. Monteiro, P.M.S., A.K. van der Plas, G.W. Bailey, and Q. Fidel. 2004. Low oxygen variability in the Benguela ecosystem: A review and new understanding. CSIR Report ENV-S-C 2004-075, Stellenbosch, South Africa Monteiro P.M.S. and A.K.van der Plas. (this volume). Forecasting Low Oxygen Water (LOW) variability in the Benguela System. Benguela: Predicting a Large Marine Ecosystem. Elsevier Series, Large Marine Ecosystems, Part II: Chapter 5. Pitcher G and A.J. Boyd. 1996. Across shelf and alongshore dinoflagellate distributions and mecahnsims of red tide formation within the Benguela upwelling system. In: Harmful and toxic algal blooms (Yasumoto, T., Oshima, Y. and Fukuyo, Y. eds.) Intergovernmental Oceanographic Commission of UNESCO, Paris. Pp 243 – 246. Probyn TA, G. Pitcher, P.M.S. Monteiro, A.J. Boyd and G. Nelson. 2000. Physical processes contributing to harmful algal blooms in Saldanha Bay, South Africa. South African Journal of Marine Science 22:285-298 R.E. Turner and N.N. Rabalais. 2001. Commonality and the future. In Rabalais, N.N. and R.E. Turner, eds. Coastal hypoxia: Consequences for living resources and ecosystems Coastal and Estuarine Studies 58:451-454. Service, R.F. 2004. New dead zone off Oregon coast hints at sea change in current. Science 305:1099 Shannon, L.V.and G. Nelson. 1996. The Benguela: Large scale features and processes and system variability. 163-210 In Wefer, G., W.H. Berger, G. Siedler and D.J. Webb, eds. The South Atlantic: Present and Past Circulation,. Berlin Heidelberg: Springer-Verlag. Shillington, F.A., C.J.C. Reason, C.M. Duncombe Rae, P. Florenchie, and P. Penven. 2006. Large scale physical variability of the Benguela Current Large Marine Ecosystem (BCLME). Chapter 4 this volume) Stramma, L. and F. Schott. 1999. The mean flow field of the tropical Atlantic Ocean. Deep Sea Research II. 46:279-303.
308
P.M.S. Monteiro et al.
van der Lingen, C.D., P. Fréon, L. Hutchings, C. Roy, G. Bailey, C. Bartholomae, A.C. Cockcroft, J.G. Field, K.R. Peard and A. van der Plas 2006. Shelf process effects on marine living resources. Part III, Chapter 14 this volume. van der Lingen, C.D., L. J. Shannon, P. Cury, A. Kreiner, C.L. Moloney, J-P. Roux, and F. Vaz-Velho. 2006. Resource and ecosystem variability, including regime shifts, in the Benguela Current System. Chapter 8 this volume. van der Plas, A.K., P.M.S. Monteiro and A. Pascall. (in press) The cross shelf biogeochemical characteristics of sediments in the central Benguela and their relationship to overlying water column hypoxia. African Journal of Marine Science.
Large Marine Ecosystems, Vol. 14 V. Shannon, G. Hempel, P. Malanotte-Rizzoli, C. Moloney and J. Woods (Editors) © 2006 Elsevier B.V. All rights reserved.
309
14 Forecasting Shelf Processes of Relevance to Living Marine Resources in the BCLME C.D. van der Lingen, P. Fréon, L. Hutchings, C. Roy, G.W. Bailey, C. Bartholomae, A.C. Cockcroft, J.G. Field, K.R. Peard and A.K. van der Plas
ABSTRACT This chapter focuses on describing, discussing and evaluating the feasibility of forecasting selected shelf processes considered to be of relevance in terms of their impacts on commercially important living marine resources of the BCLME. The impact of shelf processes is examined with regard to both the availability of resources to fishing and their abundance. Three shelf processes, namely low oxygen water events, mesoscale processes, and boundary processes, are examined separately and in detail. For each of these processes, the resource impacted and its response, the type of forecast considered appropriate and feasible, the requirements for making such forecasts, and case studies illustrating examples of forecasting systems already in place, are provided. Other processes that may have significant impacts on living marine resources are briefly discussed. The technology for forecasting low oxygen water events is available and therefore the feasibility of making such predictions is good, although at present there are insufficient moored instruments dedicated to inshore oxygen monitoring in relevant areas, either in the northern or southern Benguela regions. Indices of mesoscale processes have been used in attempts to forecast anchovy recruitment variability in the southern Benguela, and indices of boundary processes to hindcast hake recruitment variability in the northern Benguela. For southern Benguela anchovy, the wealth of studies relating environmental variability to recruitment variability and the insights gained from simulating the incorporation of predictive models into management procedures for this stock, should allow the development of recruitment prediction models that can feasibly be incorporated into management procedures. However, the incorporation of environmentally-based recruitment or stock size prediction models into management procedures should take account of assumptions and uncertainties associated with such models, and their potential for utility to management should be tested through simulation. INTRODUCTION Living marine resources of the Benguela Current Large Marine Ecosystem (BCLME) exhibit seasonal, interannual and decadal-scale variability in their abundance,
310
C.D. van der Lingen et al.
distribution, and certain biological characteristics (see van der Lingen et al. Chapter 8, this volume). Much of this variability has been attributed to anthropogenic forcing, for example the collapse of sardine and rock lobster populations in both the northern and southern sub-systems through over-fishing (Griffiths et al. 2004), and changes in distribution patterns of sardine in the northern Benguela and horse mackerel in the southern Benguela arising from intense localized fishing pressure (van der Lingen et al. Chapter 8, this volume). However, some of this variability may be attributable to environmental effects arising from the highly dynamic nature of the Benguela system. This chapter focuses on describing, discussing and evaluating the feasibility of forecasting selected shelf processes considered to be relevant to their impacts on commercially important living marine resources of the BCLME, namely hakes (Merluccius capensis and M. paradoxus), small pelagic fish (anchovy Engraulis encrasicolus, sardine Sardinops sagax and Sardinella spp), and rock lobster Jasus lalandii. The impact of shelf processes was examined in relation to the availability of resources to fishing and their abundance. A number of shelf processes that may be important in the BCLME, the resources on which they may impact, and the responses of those resources, are listed in Table 14-1. Three of these processes are considered particularly important and are dealt with in detail here, namely low oxygen water events, mesoscale processes, and boundary processes, although the distinction between these three sometimes appears fuzzy. For each of the three shelf processes we briefly refer to the physical/biological forcing of the process (see other chapters in Part I of this volume where appropriate, e.g. Monteiro and van der Plas, Chapter 5 for low oxygen events); identify the living marine resources most likely to be impacted and describe their response to the process and the time-scale over which the impact occurs; identify the appropriate type of forecast (i.e. hindcast, forecast, what-if-scenarios); assess the feasibility of producing such a forecast; estimate the economic value of such a forecast; list the requirements needed to make such a forecast and identify gaps in knowledge and understanding, and in data collection; identify the timescale of such a forecast; indicate the next steps required in order to develop such forecasting capability; and provide a case study which describes forecasting systems either already in place or approaches that are considered likely to improve predictive ability with regard to that particular process. Finally, problems considered likely to negatively affect our ability to make predictions that are useful for management are identified, and future research directions are suggested. The possible impacts of processes other than the major three processes are then briefly discussed, and the chapter concludes by summarizing pertinent issues regarding forecasting of important shelf processes. LOW OXYGEN WATER EVENTS Low oxygen water events are amongst the most pronounced environmental features that have an effect on the living marine resources of the BCLME. This is particularly evident for rock lobster Jasus lalandii because of the spectacular walkouts and subsequent mortality of this resource that occur during low oxygen water events, and the persistently slow growth rates of rock lobster in the northern parts of the Benguela.
Forecasting shelf processes of relevance to living marine resources in the BCLME
311
Table 14-1. List of processes that are initiated or occur over the shelf and are considered to impact on living marine resources in the BCLME. NB and SB refer to the northern and southern Benguela regions, respectively, and key references for the processes and resource responses are indicated.
PROCESS
REGION
IMPACTED RESOURCE AND TIMESCALE
RESOURCE RESPONSE
Low oxygen events (Chapman and Shannon 1985; Bailey 1991; Monteiro and van der Plas Chapter 5, this volume)
NB + SB
Rock lobster–immediate effect (walkouts)
Rock lobster–death via hypoxia/walkout (Cockcroft 2001)
Hakes–immediate and shortterm effects
Hakes–death via hypoxia, shifts in distribution (Hamukuaya et al. 1998)
-Meso-scale processes (enrichment, concentration and retention; advective loss; etc; Shannon and Nelson 1996; Field and Shillington 2005; Hill et al. 1998; Shillington et al. Chapter 4, this volume)
NB + SB
All, with small pelagic fish particularly well documented –seasonal and interannual effect
Enrichment and concentration have impacts on food availability to fish resources, and may affect adult fish spatial distribution and school structure (Bakun 1996)
Sulphide eruptions (Weeks et al. 2002)
NB
All–immediate effect
Death via hypoxia and/or intoxication by sulphide (Weeks et al. 2004)
Harmful Algal Blooms (Pitcher and Weeks Chapter 7, this volume)
SB (NB?)
Rock lobster–immediate effect
Rock lobster–death via hypoxia (localized red-tides result in low oxygen; Cockcroft 2001)
Unknown processes
NB + SB
Pelagic species, mainly sardine; interdecadal-scale variability in abundance
Large interdecadal variability in recruitment success, larger than interannual variability in the case of sardine (Schwartzlose et al. 1999)
Retention may be linked to east coast recruitment of small pelagics in the SB; reduced upwelling linked to reduced condition, reduced reproductive output and reduced recruitment; reduced population viability if low upwelling sustained (Hutchings 1992; Lett et al. in press) Increased advective loss linked to reduced anchovy and sardine (and other species?) recruitment success (Roy et al. 2001; van der Lingen and Huggett 2003)
312
Boundary processes (southward movement of the AB Front; Benguela Niños; permeability of the Lüderitz upwelling cell; Agenbag and Shannon 1988; Shannon and Nelson 1996; Florenchie et al. 2004)
C.D. van der Lingen et al.
NB + SB
Pelagic and demersal species, seasonal and interannual effect
1. AB Front and Benguela Nińos shifts in distribution, reduced reproductive effects, mortality due to high temperature (Luyeye 1995; Binet 2001; Boyer and Hampton 2001) 2. Permeability of the LUC– colonisation of the NB by SB stocks (Hewitson 1988) 3. Behaviour of the Agulhas Current– increased mixing/reduced stabilization linked to reduced quality in food environment for pelagic fish and consequent reproductive output; “predation” of eggs and larvae by Agulhas rings (Duncombe Rae et al. 1992; Hutchings et al. 2002a;)
The development of low oxygen waters in the BCLME has been described by Hart and Currie (1960), Stander (1964), De Decker (1970), Andrews and Hutchings (1980), and Bailey et al. (1985), and the larger scale development of the oxygen minimum zone in the tropical SE Atlantic by Moroshkin et al. (1970) and Bubnov (1972). Chapman and Shannon’s (1985) review on the chemistry of the Benguela system summarised current thinking at that time concerning low oxygen water, which was that there was both a large-scale, remotely-sourced origin and more localized sources of hypoxia in the Benguela system. The relative magnitudes of these vary in both time and space, as has been advanced by Bailey (1991), Monteiro and van der Plas (Chapter 5, this volume) and Monteiro et al. (Chapter 13, this volume). The schematic produced by Chapman and Shannon (1985; see Figure 14-1) conveys the following ideas: • Local production in the Angola Basin and the semi-closed, clockwise, gyral circulation formed by the Angola current, the offshore Benguela current and the equatorial counter current, govern the formation and persistence of the basin-wide oxygen minimum zone to the north of the Angola-Benguela Front. • A combination of remotely derived hypoxic water originating in the Angola Basin that flows southwards onto the Namibian shelf, and locally formed low oxygen water, is responsible for the frequent low oxygen events occurring in the northern Benguela region, with occasional decadal-scale more extreme anoxic events, such as in 1994. Low oxygen waters on the inner shelf of the southern Benguela region are mostly formed locally as a result of decay of planktonic detritus arising from the high levels of productivity. Using a sequence of five time-series oxygen sections constructed for 100m positions off Walvis Bay, Lüderitz, the Orange River Mouth, Roodewal Bay and Cape Columbine, Bailey (1991) demonstrated that there is a progressive southward
Forecasting shelf processes of relevance to living marine resources in the BCLME
313
Figure 14-1. Conceptual model showing areas of low oxygen water formation in the BCLME and South East Atlantic (from Chapman and Shannon 1985).
increase in the seasonal development of hypoxia in the water column on the Benguela shelf. In the north, off Walvis Bay, the sub-thermocline is perennially hypoxic whereas off Cape Columbine there is a definite maximum in the extent and severity of hypoxia in late summer and a minimum in winter, when winter mixing and reduced surface primary production are thought to play a role in ventilating the shelf waters. At the time of publication of Bailey’s (1991) paper, it was suggested that the
314
C.D. van der Lingen et al.
permanence of the bottom hypoxia off Walvis Bay might be a reflection of the southward advection of low oxygen bottom waters derived from the Angola current. Recent findings have suggested the chemical oxygen demand exerted by reduced compounds such as hydrogen sulphide, which is abundant in this area, may also play a role (Emeis et al. 2004). Periodic low-oxygen induced mortality of fish in the Walvis Bay region was reported by Copenhagen (1953), who described an almost annual mortality of fish that took take place between December and March in the “Walvis region” off central Namibia. He noted the permanent lack of dissolved oxygen in the bottom layers of the water column, but cautioned that under certain conditions, hydrogen sulphide and dinoflagellates belonging to the genus Gymnodinium, both of which can be toxic to fish, were also present. Severely oxygen-depleted bottom waters that occurred inshore on the northern Benguela shelf in 1994 displaced juvenile Cape hake Merluccius capensis offshore, apparently increasing mortality caused by cannibalism and the discarding of juveniles caught during trawling (Hamakuaya et al. 1998). This mortality was estimated at 70% (Hamakuaya et al. 1998) which would have substantially reduced recruitment, although a published time-series suggests that hake recruitment was in fact above average in 1992-1994 (Voges et al. 2002). However, the abundance of M. capensis off Namibia estimated from demersal surveys shows a decrease in biomass over that period (Burmeister 2001), providing support for the contention of reduced recruitment by Hamakuaya et al. (1998). The impacts of low oxygen water events on rock lobster populations are felt along the whole Namibian coast, and along the South African west coast from the Orange River to approximately St Helena Bay. In the northern Benguela the areas most severely impacted are located from north of Lüderitz to Easter Cliffs (although there is insufficient information concerning reefs between Lüderitz and the Orange River), and mortalities have been recorded between Walvis Bay and the Ugab River. The impact of low oxygen water events on rock lobster in Namibia is largely restricted to the recreational fishery, due to the location of events causing mortalities. Whereas low oxygen water is a permanent feature in the northern Benguela it is eventdriven in the southern Benguela, where low oxygen water events impact in “nodes” along the area between Port Nolloth to St Helena Bay, particularly around Elands Bay (see Figure 14-2). Low oxygen events result in an onshore migration of rock lobster, leading to stranding and subsequent mortality, or death prior to washing up (which occurs mostly in Namibia). Walkouts range from single strandings on a scale of a day to multiple strandings over a period of a month, the number of stranding events depending on the severity and duration of the low oxygen event. Off South Africa, the short-term impact has highly negative consequences to the local fishery (Cockcroft 2001), and since the response of the resource in the area is directly related to the severity of event, may result in a substantial loss of income to fishing companies and local communities. The long-term impact of low oxygen water events on local rock lobster populations depends on the severity of the event in the area affected, and whether or not hydrogen
Forecasting shelf processes of relevance to living marine resources in the BCLME
315
sulphide (H2S) was produced (Cockcroft 2001). Where no H2S was produced the impact on the benthos is minimal, and recovery of lobster abundance is relatively rapid since re-colonisation occurred within four to 12 months, leading to a relatively minor direct impact on the lobster fishery in the area although growth rates may have been affected. Low oxygen water events during which H2S was produced do have a major impact on the benthic community, and the recovery in lobster abundance takes around one to three years as re-colonisation is slower. However, the full impact of low oxygen water events may take even longer to transmit through the lobster population as a whole; because the bulk of stranded lobster are undersize females (based on South African data; Cockcroft 2001), the effects on egg production, recruitment and fishable biomass might be delayed by five to seven years. In addition to causing walkouts, low oxygen water events also exert sublethal effects that affect a variety of lobster life stages. Reductions in growth rates, egg production, and larval quality have been reported (see references in Cockcroft 2001) with most information available on growth rates of adults, size at sexual maturity and egg production. These sub-lethal effects are observed in all Namibian commercial fishing grounds and the recreational fishery in northern Namibia, and off the west coast of South Africa are most evident in the northern area between Port Nolloth and Hondeklip Bay.
Figure 14-2. Time-series of variation in dissolved oxygen levels with depth at a station in St Helena Bay from March 1983 to December 2004. Sampling dates and depths are indicated by the small circles overlaid on the contour plot, and periods of rock lobster Jasus lalandii walkouts are indicated by the vertical bars at the bottom of the figure.
316
C.D. van der Lingen et al.
LOW OXYGEN WATER EVENTS case study: An operational forecasting method to provide early warning of rock lobster Jasus lalandii walkouts in St Helena Bay, South Africa.
An oceanographic transect and an oxygen-monitoring buoy positioned in a water depth of 18m provide an operational method for providing early warning of low oxygen water events and possible rock lobster walkouts in Elands Bay. This is complemented by direct surveys by divers, and monitoring of catch rates of lobsters by commercial dinghy fishers in the kelp beds, and used to assess the likelihood and time of a walkout. A decision rule process is used to assign different levels of alertness, depending on the answers to the questions posed. If answers are affirmative, the identified level of alert is attained. The following decision rules have been formulated for different levels of alertness:
•
•
•
•
•
•
•
•
Do data from the St Helena Bay Monitoring Line (a monthly monitoring transect that extends from the coast to the shelf edge off St Helena Bay; Hutchings et al. 2002b) indicate that the extent of low oxygen water on the shelf is well-developed and that dissolved oxygen concentrations are declining to <1 ml/l at the inner two stations off Elands Bay and St Helena Bay (located 2 and 5 nautical miles off the coast, respectively)? → LEVEL 1 ALERT Does real-time data from the monitoring buoy off Elands Bay indicate a sustained drop at 18m depth in dissolved oxygen concentrations to <1 ml/l? → LEVEL 2 ALERT Have the catch rates of the hoop-net fishery for rock lobster increased indicating that lobsters have been herded shore-wards? → LEVEL 3A ALERT Is there a dense bloom of phytoplankton or a HAB, aggregated by calm/gentle onshore winds (indicating that decay of the bloom is likely to exacerbate oxygen demand in the water column)? → LEVEL 3B ALERT Has the wind changed to offshore (indicating shore-ward movement of bottom water)? → LEVEL 3C ALERT Is there no significant wave action (i.e. no aeration of the subtidal zone)? → LEVEL 3D ALERT Have the catch rates by the hoop-net fishery suddenly declined (indicating that the rock lobsters have become affected by low dissolved oxygen levels and have reduced feeding intensity)? → LEVEL 3E ALERT Do divers report massed lobsters in the shallow sub-tidal zone? → FULL ALERT, INVOKE CONTINGENCY PLAN
Once invoked, the contingency plan aims to collect and return as many live rock lobsters as possible to the ocean in areas unaffected by low dissolved oxygen levels. The contingency plan becomes very expensive in the final stages as large trucks, helicopters, cooling facilities and the local community have to be mobilised to move 100s to 1000s of tons of rock lobsters to oxygenated waters within a few hours. The contingency plan was invoked following a lobster walkout event in 2000 in which approximately 200t of live lobster were collected and returned to the sea from an estimated 1600t that had been stranded. In early 2005 a Level 3C alert was attained, during which lobsters were observed massing in shallow water but did not walk out, possibly due to wave action that aerated the subtidal zone, and/or sinking of oxygen-rich surface waters during periodic, onshore wind-induced downwelling events.
Forecasting shelf processes of relevance to living marine resources in the BCLME
317
The management response in South Africa to both lethal and sublethal effects of low oxygen water events on lobster populations includes a contingency plan for walkout events (see Low Oxygen Water Events case study), and the incorporation of data regarding biomass loss during mortalities and reductions in adult growth rates into management decisions concerning total allowable catch (TAC). No response to low oxygen water events is made by managers of the lobster fishery in Namibia. Forecasting at a timescale of months to days, or even hours before a low oxygen water event is considered appropriate for predictions of lethal effects (walkouts). Data required to make such forecasts include hourly to daily in situ measurements of bottom dissolved oxygen content at key sites (from a moored array), coastal wind speed and direction, observations on harmful algal bloom (HAB) distribution and biomass, and observation of massing of lobster in shallow waters (via SCUBA). Forecasting sublethal impacts of low oxygen water events on lobster on a monthly / annual to decadal scale will require hourly to daily in situ bottom dissolved oxygen content at key sites (from a moored array) for information relevant to the fishery (effect on daily catch rate) and population dynamics (growth rate and possibly egg production). The feasibility of, and requirements for, large-scale forecasting of basin-scale development of low oxygen events in the tropical SE Atlantic are discussed by Monteiro and van der Plas (Chapter 5, this volume). The local-scale development of low oxygen water on the shelf is predictable using the early warning / contingency plan system tabulated (see Low Oxygen Water Events case study), and is what managers of this fishery require. Studies are needed to examine dissolved oxygen levels in relation to lobster catchability and growth rate, and how these may affect TAC recommendations; these may lead to environmental inputs to biomass models. To fine-tune forecasting requires research on lethal effects via controlled experiments on critical oxygen concentration and time of exposure for key life history stages, and on sublethal effects via controlled experiments on critical oxygen concentration and time of exposure in terms of important biological parameters such as growth rate, egg production etc. In South Africa, the economic value of forecasting low oxygen water events and their impacts on rock lobster is high, and would lead to better management of walkout events and improved management of the resource for commercial exploitation. For example, the approximately 1700 tons of rock lobster that died during the walkout that occurred in 1997 (Cockcroft 2001) is estimated to have had a landed value of R100 200 million (one US$ approximates to seven South African Rands). For Namibia, the economic value of forecasts is seen more in terms of potentially improved management of the resource. Successful forecasting could lead to the incorporation of environmental parameters into resource management models and the development of an ecosystem approach to management of the rock lobster fishery. Technology is available and therefore the feasibility of making such predictions is good, although at present there are insufficient moored instruments dedicated to inshore oxygen monitoring in relevant areas. For South Africa, requirements include three moored instruments to provide real time measurement of dissolved oxygen,
318
C.D. van der Lingen et al.
temperature and bottom currents in hotspot areas (e.g. Port Nolloth, Elands Bay and Saldanha Bay). For Namibia, moored instruments that provide real time measurement of dissolved oxygen, temperature and bottom currents should be distributed in the southern, central and northern commercial grounds. Additionally, true colour satellite images should be available to detect hydrogen sulphide events (e.g. Weeks et al. 2002; 2004). Implementation of forecasting would be possible as soon as moored arrays are in place. The next step required is to obtain funding for moored arrays and appropriate research. Forecasts of local-scale development of low oxygen water may also be useful to management of the Namibian hake fishery, if low oxygen water events do impact negatively on recruitment success. During the late 1990s there appeared to be a number of warmer than usual events in the northern Benguela, evidenced as southward anomalies of the Angola Benguela Front (Florenchie et al. 2004; Shillington et al. Chapter 4, this volume), and persistent reduced availability of older shallow water hake (M. capensis) beyond the 200m isobath, which resulted in increased targeting of larger deep water hake (M. paradoxus) in the Lüderitz region. The relationship between hake distribution and warm events in the northern Benguela has previously been examined; anomalously warm conditions appear to induce hake to concentrate closer to the seabed (Macpherson et al. 1991), and may also lead to an increased density of hake on the fishing grounds (Shannon et al. 1988). MESOSCALE PROCESSES Physical and biological mesoscale processes of relevance to living marine resources are processes ranging from several hundred meters to a few hundred kilometres, and the corresponding time scales range from a few hours to several months (Figure 14-3). Mesoscale processes are of major interest when studying environmental processes of relevance to living marine resources because they impact directly on the habitat of individuals, schools or clusters, or other aggregations. However, developing projects to investigate the link between mesoscale environmental processes and resource dynamics has been limited by cost and the technical difficulties in developing synoptic sampling at those scales. Remote sensing techniques as well as models (hydrodynamic models and lagrangian tools coupled to individual-based models) are today the favoured tools to investigate the coupling between mesoscale processes and resources dynamics. Like other upwelling regions, the BCLME is characterised by intense mesoscale activity over the shelf and along the transition zone between the cold, nutrient-rich coastal waters and the warmer, oligotrophic waters of the offshore domain (Shannon and Nelson 1996; Shillington 1998; Field and Shillington 2005; Shillington et al. Chapter 4, this volume). The wind-induced coastal upwelling and its interaction with the topography and with open ocean circulation features, such as the Agulhas Current or tropical waters off Angola, are the major forcing factors of mesoscale processes in the BCLME. In the southern Benguela, the upwelling cell along the Cape Peninsula
Forecasting shelf processes of relevance to living marine resources in the BCLME
319
Figure 14-3. Time-horizontal space scale diagram illustrating the major physical and biological processes operating in the ocean (from Dickey 2003)
and the upwelling plume at Cape Columbine are particularly dynamic transient structures mostly controlled by wind forcing (Nelson and Hutchings 1983). Further north, the Namaqua and Lüderitz upwelling cells have a much broader spatial extent (Figure14-4a,b). Satellite images provide direct evidence of the surface manifestation of mesoscale processes such as plumes, fronts, eddies and filaments (Figure 14-4a,b). These surface structures result from instabilities that develop at the interface between the cold water on the shelf and the warmer offshore water, as well as from interactions between the upwelling process and the local topography (Brink 1983; Barth 1989a, 1989b; Chapman and Lentz 1994; Smith 1995). Surface flow, as well as vertical water movement on the continental shelf, are strongly modulated by those mesoscale structures (see, for example, Figure 9 in Barange and Pillar 1992). Cold water filaments extending off the shelf enhance offshore flow in the form of narrow but intense jets (Brink and Cowles 1991; Shillington et al. 1992; Hill et al. 1998). Fronts are also places where vertical movements are enhanced, resulting in a complex dynamic of both upward and downward exchanges (Simpson and Hunter 1974).
320
C.D. van der Lingen et al.
Figure 14-4. Daily SST images for the northern (a) and southern (b) Benguela regions. Courtesy OceanSpace CC.
Shelf break upwelling or upwelling plumes can play a major role in controlling the cross-shore exchanges and the spatial distribution of plankton (Barange and Pillar 1992; Graham and Largier 1997; Penven et al. 2000). In the southern Benguela, the sheltering effect of the Cape Columbine upwelling plume often results in enhanced chlorophyll concentrations in St Helena Bay while a chlorophyll minimum is recorded in the core of the plume (Figure 14-5). This pattern should be compared with the structure observed further north along the Namaqua cell off Hondeklip Bay: the nearshore domain is occupied by newly upwelled water low in chlorophyll and the maximum phytoplankton biomass is located mid-shelf (Figure14-5). With such a configuration, phytoplankton is subject to exchanges with the offshore domain. The sheltering effect provided by mesoscale structure is thought to have a major effect on fish habitat by maximising food availability in relatively small areas but also by reducing potential offshore losses for small organisms such as fish larvae and juveniles (see the Boundary Processes section for details on the onshore movements of larvae and earlier transport of eggs and larvae by the jet current off the Cape Peninsula). Processes relevant to the survival and recruitment of early life stages of pelagic fishes have been synthesized through Bakun’s fundamental triad as enrichment, concentration and retention (Bakun 1996). The three elements of this triad involve a
Forecasting shelf processes of relevance to living marine resources in the BCLME
321
Figure 14-5. Daily images of (a) SST and (b) sea surface chlorophyll a in the southern Benguela. Courtesy OceanSpace CC.
Figure 14-6. Output maps of simulated enrichment intensity (left panel; log scale, white is zero; averaged over the period 1992-1999) and simulated retention (right panel; proportion of particles retained; averaged over the period 1992-1999) from a particle-tracking model developed for the southern Benguela. After Lett et al. (in press).
322
C.D. van der Lingen et al.
set of mesoscale processes that will be specific to each ecosystem. In an upwelling region such as the Benguela, enrichment will be mainly derived from the wind-induced Ekman divergence along the coast or at the shelf break. Concentration will develop in areas where vertical stratification is enhanced or along surface structures such as thermal fronts. Retention will be enhanced within eddies or within areas of the shelf where recirculation structures are able to develop. Quantifying the triad processes in the real world remains a major challenge, and its contribution to observed variability in fish recruitment has not yet been tested. An empirical method for quantifying triad processes using satellite-derived ocean colour data and surface wind stress as a proxy for flow divergence, convergence, and Ekman transport, has been used to identify favourable reproductive habitats for anchovy (E. encrasicolus) in the Mediterranean (Agostini and Bakun 2002). Additionally, recent advances in hydrodynamic modelling based on Lagrangian tracking of particles might provide some useful insights on how the triad processes are related to the variability of the environment, and how they might impact on recruitment (Lett et al. in press). Such modelling approaches can be used to illustrate mesoscale events in a dynamical manner (Figure 14-6) and, if proved to be successful, may provide a useful step towards an integrated way to quantify the effect of mesoscale processes on recruitment variability and to develop the ability to quantify mesoscale processes in an operational way. Substantial research effort aimed at improving understanding of the causes of variability in recruitment and population size of anchovy in the southern Benguela has resulted in the development of a variety of approaches to predicting recruitment strength of this important resource, including expert systems and empirical approaches (see van der Lingen et al. Chapter 8, this volume; and Mesoscale Processes case study below), in addition to survey data. Initially (from 1964), annual fluctuations in anchovy recruitment and spawning biomass were estimated by virtual population analysis (VPA) based on the age structure of commercial catches (Armstrong et al. 1983; Butterworth 1983). Subsequently (from 1983), annual estimates of anchovy population size were based primarily on hydroacoustic and daily egg production surveys conducted during the peak (November) spawning season, followed (in 1985) by surveys that were implemented in May each year to obtain an estimate of the abundance of anchovy recruits (Hampton 1987). From 1991 onwards, annual surveys were conducted in March to assess the abundance and distribution of pre-recruit anchovy and sardine off South Africa’s west coast. These surveys were designed to obtain an indication of forthcoming anchovy recruitment strength (O’Toole and Hampton 1989), and timed so that such information would be obtained sufficiently early in the fishing season to be useful for management (Cochrane and Starfield 1992). However, whereas the pre-recruit surveys have provided useful knowledge on early life history stages and the recruitment process, they did not prove useful in predicting anchovy or sardine recruitment success (van der Lingen and Huggett 2003). Most recently (since 1996), attempts to predict recruitment strength from field observations have used data collected along the Sardine/Anchovy Recruitment Project (SARP) Monitoring Line, a transect across the jet current off the Cape Peninsula (Huggett et al. 1998). Ichthyoplankton collected along this transect primarily reflect spawning on the western Agulhas Bank, and a significant correlation between an index of anchovy egg abundance (over the September to March spawning season) and subsequent anchovy
Forecasting shelf processes of relevance to living marine resources in the BCLME
323
recruitment strength indicates the potential of this index for use in forecasting anchovy recruitment strength 2-3 months before recruitment survey estimates are available (van der Lingen and Huggett 2003). The relevance of forecasting shelf processes to predict variability in marine populations is still an open question. To date, no study has truly been able to link the variability of physical processes to variability in populations (Werner and Quinlan 2002). MESOSCALE PROCESSES case study: Predicting recruitment of anchovy Engraulis encrasicolus in the southern Benguela using environmental indices.
Because recruits comprise up to 70% of annual anchovy landings in the southern Benguela and because recruitment shows substantial interannual variability, correct predictions of anchovy recruitment strength could result in substantial increases in average annual catches without an associated increase in risk (Cochrane and Starfield 1992). Prediction of anchovy recruitment therefore has important implications for management, and substantial research effort has been directed toward this in the southern Benguela. Expert systems and empirical approaches that use indices of mesoscale processes as input terms have been developed, such indices being primarily related to upwelling intensity (e.g. wind stress and sea surface temperature) or the strength of the frontal jet current that transports anchovy eggs and early larvae from the south coast spawning grounds to the nursery grounds off the west coast (see references in van der Lingen and Huggett 2003, for details). Most of these approaches used environmental indices that were spatio-temporally averaged over the anchovy spawning season (September to March) and area (the western Agulhas Bank), but more recent work has examined within-season variability in mesoscale oceanographic processes and incorporated environmental indices derived at a finer spatio-temporal resolution.
Roy et al. (2001; 2002) proposed that reduced upwelling in December off the Cape Peninsula would limit offshore loss of anchovy eggs and larvae during the transport phase, and that moderate upwelling off the west coast in January would enhance retention and provide sufficient food for larvae. Those authors combined observed relationships between these indices and anchovy recruitment strength into an empirical model for the period 1985-1994 and used that model to hindcast anchovy recruitment success over the period 1995-2000, and observed a reasonable fit between hindcast and observed anchovy recruitment strength. Following that work, Richardson et al. (2003) used a Self-Organizing Map (SOM) approach to extract patterns from satellite-derived SST anomaly data, and defined a characteristic state that represented the processes highlighted by Roy et al. (2002); i.e. a pattern of a positive SST anomaly off the Cape Peninsula indicative of reduced upwelling and a negative SST anomaly north of Cape Columbine indicative of strong upwelling. Richardson et al. (2003) related the frequency of occurrence of this state during the spawning season to subsequent anchovy recruitment over the period 1985-1999 and reported a significant correlation between the two, but not between recruitment and any other of the 15 characteristic SST anomaly states identified by their SOM. However, despite substantial research effort and the achievements attained in hindcasting anchovy recruitment strength, real prediction remains elusive and predictions of anchovy recruitment have yet to be incorporated into management procedures. Simulations to investigate the benefits of using environmental indices to set appropriate Total Allowable Catches for southern Benguela anchovy have recently been published, and suggest that such indices need to explain 50% or more of the total variation in recruitment before the current management procedure would start to show benefits in terms of risk (of the population falling below a specified level) and/or average catch (De Oliveira and Butterworth 2005).
324
C.D. van der Lingen et al.
There are many reasons for this, including difficulties in sampling both the relevant physical processes and organisms such as larvae and juveniles at the right scale, as well as the cost of ensuring long term monitoring. Another difficulty is that variation in populations results not from a single process but from an integrated set of processes affecting individuals at different time and space scales, depending on their development stages. In a given system, the existence of major bottlenecks should limit the number of processes requiring investigation, and identifying those bottlenecks is a necessary step before the setting-up of monitoring and forecasting systems. Hutchings (1992) identified a set of potential environmental contributors to the variability of fish recruitment that provides a good basis to elaborate forecasting capabilities in the southern Benguela region. Presently, biological individual based models (IBMs) coupled to hydrodynamic models are being used to explore and better understand the processes (including mesoscale processes) responsible for recruitment fluctuations (e.g. Huggett et al. 2003; Mullon et al. 2003; Parada et al. 2003; Skogen et al. 2003; Miller et al. in press). Results from these IBMs have matched current knowledge based on field observations, providing confidence that inferences made from these IBMs are meaningful. Such experimental simulations are likely to improve our understanding of the key processes responsible for anchovy recruitment success in relation to their spawning strategy, a major step in predicting fluctuations in stock size, and may be used in the near future to hindcast recruitment. Additionally, these IBM simulations may be used to test a variety of what-if scenarios, including examining the likely impacts of climate change on recruitment variability. Additional data and analysis of these and historical data are required, and particular attention should be paid to the three dimensional validation of hydrodynamic and productivity models. This could be achieved through moored ADCP arrays and by continuous vertical profilers (such as the YOYO-ANAIS; Provost et al. 1998; Thouron et al. 2003) in addition to ongoing monitoring lines for measuring physical and plankton parameters. In order to test the hypothesis of food limitation for fish larvae, the determination of the RNA/DNA ratio as a proxy for starvation (e.g. Clemmesen et al. 1997) would be useful. Furthermore, core sediments in anoxic areas would provide an idea of natural, multi-decadal variation of fish abundance in the absence of exploitation, as has been done for small pelagic fish populations in the northern Benguela (Baumgartner et al. 2004) and elsewhere (e.g. Baumgartner et al. 1996). To be pertinent the sampling design should cover, as far as possible, a substantial part of the fish habitat (see basin theory of MacCall 1990). Requirements differ for now-casting or hindcasting compared to forecasting or what-if scenario predictions. In the former, historical and near-real-time data on wind and surface fluxes within the domain of the hydrodynamic model, and water current and Nutrient/Phytoplankton/Zooplankton/Detritus (NPZD) data at the boundaries of the domain are required. In the latter, forecasts at an adequate resolution of the forcing factors within the model domain and at its boundaries are needed, in addition to the existing monitoring lines of physical and plankton parameters plus possibly a few additional ones. An important need that is not presently satisfied is the analysis of
Forecasting shelf processes of relevance to living marine resources in the BCLME
325
plankton data collected on monitoring lines. Because current techniques of collection and analysis of plankton are time consuming, it is suggested that modern techniques such as automated identification and size measurement of plankton from scanned images of samples coupled to image recognition and in situ Optical Particle Counter (OPC) and multi-frequency acoustic identification of organisms, are adopted. Data collection and analysis (especially of historical records) should prioritize and focus on key biological stages and key spatio-temporal data such as spawning and transport areas. A general comment is that the value of data from monitoring systems is incremental in time. Additionally, the need for correcting the imbalance in monitoring and modelling capabilities was identified in the Benguela region, as well as the need for improving communication and coordination in environment monitoring systems. There is a need to move from empirical-based predictions that often prove to fail sooner or later (see Myers 1998) to process-based predictions that should be more robust, although one cannot exclude changes in the key processes related to changes in the state of the ecosystem. Coupled hydrodynamic and individual based models will be particularly useful to understand these key processes, once identified, and allow for sensitivity analysis of the impact of the environment on living marine resources, such as the effect of changes in wind pattern or on the flow of the Agulhas Current. Presently, models must be seen as exploratory tools rather than predictive tools, because of the need of validation in three dimensions. Using such models to predict resource responses to environmental forcing should follow a two-phase approach; the first being in the short-term and comprising a 3-5-year learning period during which tentative predictions are made, and the second phase being the implementation of forecasting in management procedures over the mid-term (5-10 years). In the case of short-lived species such as pelagic fish, shrimps and squids, the expected major economic benefits of successful predictions concerning the impacts of mesoscale processes on such resources would be the prevention of stock collapse that would in turn lead to fishery collapse, with obvious economic and employment problems. BOUNDARY PROCESSES A number of boundaries exist within the BCLME, some of which are well defined and prominent, such as the Angola-Benguela Front, the Lüderitz upwelling cell or the Agulhas Current, while a number of the offshore frontal boundaries are highly variable and range from diffuse to very distinctive on short-term and seasonal time scales. Most of the major boundary features described below are shown in Figure 1-1 of this volume (Chapter 1), and the following text provides a brief description of these boundaries and their ecological roles, progressing from the north to the south.
326
C.D. van der Lingen et al.
Angola-Equatorial Front/Congo River mouth This is a very seasonal front about which little is known, and much more data are needed to properly evaluate the nature of the boundary separating the subtropical Angolan coastal waters from the truly tropical Gulf of Guinea ecosystem. The Angola-Equatorial Front is characterised by a sharp discontinuity in salinity and is heavily influenced by the Congo River outflow. It is also the zone where equatorial counter-currents impinge on the coast and where Kelvin waves and coastally trapped waves are first detected close to the coast. This region is the centre of oil production in Angola, and operations are currently expanding into deeper water to the south. BOUNDARY PROCESSES case study: Predicting recruitment of hake Merluccius capensis off Namibia from environmental indices.
Recent research has used multinomial logistic regression analysis to calculate the probability of strong, average or weak recruitment of Cape hake off Namibia (Voges et al. 2002), and to identify environmental conditions that may be related to recruitment strength and could possibly explain some of the interannual variability in recruitment strength observed for this species. The model includes environmental indices (satellite-derived SST data) that describe the extent of warm water intrusion (southerly penetration of the Angola Current arising from a southward movement of the Angola-Benguela Front) as well as upwelling strength over the Namibian shelf, both during the spawning season and in the subsequent two years. The rationale for using these indices is that if warm water is widespread during the SeptemberMarch spawning season and upwelling during the following May-September is reduced, eggs and early juveniles would be retained in favourable nursery areas which should increase the probability of strong recruitment. Intensified upwelling the following year (i.e. two years after spawning) should provide sufficient food for late juveniles, also increasing the probability of strong recruitment. Hake recruitment strength predicted from this model (as either strong, average or weak) corresponds well with the observed recruitment index derived from biomass surveys, and the model accounts for 79% of the variance in historical recruitment strength. The model correctly predicted strong and average recruitment 50 and 71% of the time, respectively, and weak recruitment 100% of the time. This last record is considered particularly important, since it permits anticipation of the worst-case scenario (i.e. weak recruitment) for the hake stock which could enable an appropriate adjustment of the management policy.
Whilst the feasibility of predicting hake recruitment strength using this model appears promising, it is based on a relatively short (14 years) time series and has only been tested for a limited number of years. For example, predictions of this model have indicated strong probabilities that good recruitment would result from the 2002 and 2003 year-classes. The resultant 2002 cohort was strong as predicted but results of the 2005 hake survey indicated a weak 2003 year-class. However, initial signs from seal scat information also pointed towards an average to above average 2003 year-class, but numbers dwindled as the year (2004) progressed. There is now strong evidence that the demise of the 2003 year-class resulted from cannibalism by the exceptionally strong 2002 cohort during their coexistence in the pelagic zone (2004). Incorporation of such information into this model is planned and should be tested to see if it improves the predictive capabilities of the model. However, without such updates and extensive testing, incorporation of the model in its present state for management procedures should be approached with caution.
Forecasting shelf processes of relevance to living marine resources in the BCLME
327
There is probably a wealth of information utilised for operational purposes in the oil company databases. Pelagic fish such as Sardinella spp. are distributed on each side of the Angolan Front (Baptista 1977) although Sardinella aurita tend to avoid the Congo River plume of low-salinity and turbid water, as well as similar water masses from the Bay of Biafra that flow southward during certain years (Guinean Niños). This low tolerance of S. aurita to these conditions result in its becoming trapped between the coast and the Congo River plume during Benguela Niños (Binet et al. 2001). In contrast, S. maderensis prefers lower salinity waters, and in that respect the Angolan Front and the leakage of a tongue of coastal water from the Bay of Biafra appears to function ecologically as conduits, rather than as barriers, to this species and for some other organisms. Angola-Benguela Frontal region The Angola-Benguela Front represents the major boundary that separates the upwelling-dominated northern Benguela from the subtropical waters of Angola. This region is characterised by a sharp density front at the surface, complex mixing, and high variability in the short term. The Front shows regular latitudinal movement of about 2-3° on a seasonal basis, forced by the atmosphere, with a maximum northward movement in late austral winter (August) that coincides with maximum upwelling (Shannon et al. 1987). The Angola-Benguela Front is a convergent front between cool and warm coastal currents, and is associated with strong offshore (zonal) flow that carries coastal water rich in chlorophyll and often containing fish eggs and larvae far offshore into the South Atlantic. Phytoplankton from just south of the AngolaBenguela Front may sink offshore in the Angola Dome region, contributing to the oxygen deficit in subsurface waters there. Despite suitable temperatures, the high offshore losses and low retention restrict the area suitable for spawning to a small area adjacent to the coastline. The Angola-Benguela Front acts as a barrier, rather than a conduit, with major shifts in the dominant species composition of marine communities on either side. The nature of this barrier for deepwater species such as red crab (Chaceon maritae) or hakes (Merluccius capensis/polli) is uncertain. Moderate interannual variability in the location of the Angola-Benguela Front is likely driven by local seasonal wind forcing (Shannon et al. 1987), and the extent of southerly penetration by Angola Current water onto the Namibian shelf has been used as an input parameter in a model aimed at predicting hake recruitment from environmental indices (Voges et al. 2002; see Boundary Processes case study). This model has potential for incorporation into management procedures since it could reduce by one year the 3-year delay between hake recruitment and its estimate from trawl surveys. An example of how an environmental stock-recruitment relationship could be incorporated into assessment and management procedures of an important demersal species (Pacific cod Gadus macrocephalus) has been provided by Sinclair and Crawford (2005). Marked decadal scale variability is apparent in the location of the Angola-Benguela Front, with the Front moving further south than usual approximately every 10 years.
328
C.D. van der Lingen et al.
This results in a major warm-water intrusion penetrating into the northern Benguela upwelling region, with important ecological consequences, particularly for pelagic fish. These so-called Benguela Niños occur over the summer months, and the resultant warm-water anomalies change the community structure of the pelagic assemblage in the northern Benguela. The Sardinella aurita stock, which is at the southern limit of its distribution in southern Angolan waters, is driven towards Namibia, and whilst catches of this species off Angola are depressed (Luyeye 1995) significant landings may be made off Namibia (Thomas 1984). Benguela Niños also drive sardine (Sardinops sagax) southwards from Angolan to Namibian waters and closer to the main fishing harbour at Walvis Bay, and increases in the Namibian sardine catch were observed during and after the warm events of 1968 and 1973/74 (Hewitson et al. 1989). Shoreward intrusions of warm water that follow Benguela Niños also increase the catchability of horse mackerel (Trachurus spp.) on the Angolan shelf (Binet et al. 2001). These changes in distribution patterns indicate that fishery managers need to distinguish between changes in fish availability caused by displacements in water masses and genuine population increases, and should act conservatively in setting catch levels in years of southward intrusions of warm water. Management is further complicated by the fact that Benguela Niños appear to be linked to poor recruitment conditions in the normal nursery grounds off north-central Namibia for many important species, including anchovy, sardine, hake and horse mackerel (Boyer and Hampton 2001), which means that poor recruitment may follow greatly improved CPUE statistics. The northward penetration of cold Benguela water is important for feeding conditions for pelagic and demersal species in southern Angola. Benguela Niños appear to be remotely forced from events in the west equatorial Atlantic off the Brazil coast (Shannon et al. 1986; see also Shillington et al. Chapter 4, this volume; Reason et al. Chapter 10, this volume; Brundrit et al. Chapter 16, this volume). The southward movement of the Angola-Benguela Front is predictable by about 2 months ahead at this stage (Florenchie et al. 2004), and links to the Pacific may extend these predictions. Information from the PIlot Research moored Array in the Tropical Atlantic (PIRATA) (Servain et al. 1998) on subsurface heat content and thermocline depth is considered a potentially valuable source of data for prediction purposes, as would be sea level recorders along the Angolan coast. Central Namibian Oceanic boundary This diffuse boundary zone (Figure 14-7) is up to 400 km wide (in terms of phytoplankton) from 14-22oS, and 150-200 km wide at 22-25oS. Hence productive waters extend well beyond the current domain of important fish stocks that are harvested on or close to the shelf. The Central Namibian Oceanic boundary is a potential refuge for early life history stages of neritic fish stocks, and has a high potential for enhancing their survival and subsequent recruitment, provided that young fish can return to the neritic zone as juveniles. There is a seasonal shoreward
Forecasting shelf processes of relevance to living marine resources in the BCLME
329
Figure 14-7. Montage of the frontal boundary of the BCLME for summer (December 1983-February 1984; top panel) and winter (June-August 1984; bottom panel) derived from METEOSAT II thermal infrared data and by superimposing the upwelling fronts as observed in each five-day period during the respective three month season. The 200, 1000, 2000 and 4000m depth contours are shown (thin lines) and thick broken lines indicate the main protrusion axes of the upwelling front. From Lutjeharms and Stockton (1987), reproduced with permission.
330
C.D. van der Lingen et al.
movement of this boundary region in summer months (C. Bartholomae, MFMR, unpublished data) which is predictable, and occasional shoreward movement that is associated with a lack of winds from the south. Such shoreward movement is relevant in terms of biogeography and life history strategies for population maintenance of euphausiids (Barange et al. 1992), and has also been suggested as driving the availability to fishing of demersal resources including hake (Macpherson et al. 1991). Subsurface current structures are much less documented than surface features, but are thought to be of major importance in controlling the position and location within the continental shelf of small organisms such as fish eggs or larvae (Sundby et al. 2001; Batchelder et al. 2002). In the northern Benguela, sardine eggs were found to be most abundant within the upper layers and sardine larvae were found deeper (Stenevik et al. 2001). This was explained as a behavioural adaptation to the vertical circulation structure over the shelf: by migrating at a depth deeper than the wind-induced offshore-moving surface layer, sardine larvae make use of the subsurface upwellingcompensatory flow to avoid advective loss and to enhance transport to and retention within the nearshore area (Stenevik et al. 2003). Lüderitz/Orange River Cone boundary region The perennial cell of intense upwelling situated off Lüderitz separates the northern and southern Benguela subsystems, and the Lüderitz Upwelling Centre / Orange River Cone (LUCORC) region experiences considerable offshore losses in the surface layers, strong surface mixing and low phytoplankton concentrations. It is considered a barrier to small pelagic fish such as anchovy and sardine, and their eggs and larvae, primarily because of high levels of wind-induced turbulence in the region of Meob Bay (24°30’S; Agenbag and Shannon 1988). This is probably less of a barrier to mesopelagic species such as lanternfish (Lampanyctodes hectoris), redeye (Etrumeus whiteheadi) or gobies (Sufflogobius bibarbatus), all of which are capable of more extensive vertical migration than the small epipelagic fish. The LUCORC boundary is characterised by changes in source water to the shelf region, differing in oxygen, salinity and a few zooplankton species. Subsurface currents that flow southwards in the northern Benguela off central Namibia, flow offshore near Lüderitz and are replaced by more oxygenated Eastern Atlantic Central water from the Cape Basin, which flows southwards. The LUCORC region therefore constitutes a subsurface as well as a surface boundary zone, and may impinge dramatically on the alongshore movement of demersal species such as hake, as well as on pelagic species. The LUCORC barrier occasionally breaks down during periods of reduced southerly, upwelling-favourable wind, permitting the movement of biota between the northern and southern Benguela subsystems. For example, the record catch of 376 000t of anchovy that was taken off Namibia in 1987 was considered to have originated, in part, from the southern Benguela (Hewitson 1988). Anomalously warm sea surface conditions were recorded north of the Orange River in early 1987, suggesting that the LUCORC barrier had collapsed sufficiently to enable anchovy larvae and post-larvae to migrate from the southern to the northern Benguela and recruit to the population
Forecasting shelf processes of relevance to living marine resources in the BCLME
331
targeted by the Namibian pelagic fishery (Hewitson 1988). Northward penetration of the LUCORC barrier by early stages of anchovy appears to depend in part upon recruitment strength, the distribution of anchovy spawning (whether confined to the Agulhas Bank or extending up the west coast, as was the case in 1987; van der Lingen et al. 2001) and the alongshore flow during the transport phase to advect larvae sufficiently far northwards along the South African west coast. Possible southward movement of biota across this boundary appears to be more likely to occur during austral autumn (April-June), a period when the distribution of low wind speeds (the cube of which is roughly proportional to turbulence) extends further south than during the rest of the year (Agenbag and Shannon 1988). The recent prolonged lack of winds in the late 1990’s and early 2000’s, which are likely to have weakened the LUCORC barrier, are in contrast to expectations of increased wind stress arising from climate change (Bakun 1990; Shannon et al. 1996). The LUCORC boundary is the southern boundary of spawning habitat for shallow water hake, Merluccius capensis; (Sundby et al. 2001) and small pelagic species (King 1977) in the northern Benguela. Deep-water hake (M. paradoxus) have penetrated further north into the northern Benguela over the past two decades (see van der Lingen et al. Chapter 8 this volume), but apparently do not spawn off Namibia (Sundby et al. 2001). Identifying the nature of these changes is part of a continuing research program to understand and predict the factors affecting movement of deep-water hake. The cross-shelf circulation, and particularly the strong offshore advection, is associated with the strength of the southerly winds in the Lüderitz (25oS) region, which is in turn associated with the mean position of the South Atlantic high pressure cell off the coast of southern Africa. The modes of variability of the South Atlantic are not well understood at present and a St Helena Island Index (HIX; Feistel et al. 2003) indicates an approximate 13 to 18-year mode but appears to be a poor indicator of local winds at Lüderitz. Unless empirical and theoretical modelling studies currently underway produce an advance in understanding of the mechanisms underlying the wind variability in the Lüderitz upwelling region, the probablility of forecasting remains poor. The subsurface boundary between the oxygen-poor water of Angolan origin, carried southward in the shelf-edge poleward undercurrent and the more oxygenated Central waters originating from the Cape basin in the SE Atlantic, are currently not well understood and more observations and modelling are required before feasible scenarios can be made. At best real time measurements should allow nowcasting or “what-if” scenario development. West Coast Oceanic boundary This boundary is more defined than its equivalent in the northern Benguela, and occurs between cool, productive, food-rich mature upwelled water inshore and warm, foodpoor waters comprising a variable mixture from the Agulhas Current, the Agulhas Bank or offshore Atlantic surface water (Figure 14-7). Modelling studies indicate that much of the interannual signal in offshore temperature is derived from contributions from the Agulhas Bank or Current, rather than from local upwelling-favourable winds
332
C.D. van der Lingen et al.
(Penven et al. 2001; Blanke et al. 2002; 2005). The boundary is diffuse in winter but more defined in summer and autumn, when it is associated with alongshore, northward jet currents. Alongshore transport dominates over across-shore movement, but mesoscale eddies and filaments can alter the pathway and the feeding conditions, and mesoscale processes are considered particularly important in this region. Larvae and pre-recruits of pelagic fish which spawn further south on the Agulhas Bank and the shelf edge must cross over this boundary to enter the inshore nursery areas off the west coast (Hutchings et al. 2002a), either by actively swimming towards the shore or by being carried passively as winds abate at the end of summer and surface waters move shorewards, or by vertical migration using subsurface shoreward moving currents as modelled by Parada et al. (submitted). Studies of plankton productivity and abundance are needed to assess and predict the energy supply for young fish in this boundary region, and the role of the seasonal wind decline in autumn and its contribution to shoreward movement of pre-recruits needs evaluation. Furthermore, measurements or models of shoreward transport success are needed in order to estimate or predict anchovy recruitment success. Anchovy prerecruit abundance is not well correlated with recruitment measured three months later (van der Lingen and Huggett 2003), and the timescale of the process for pre-recruits to move onshore is considered to be of the order of 1-3 months. An anchovy larva of 20mm TL swimming at a cruising speed of 1 body length.s-1 (Hunter 1972) for 12 h.d-1 would take 105 days to swim the 180 km from offshore of the continental shelf to the coast, if water movement either toward or away from the coast were zero. However, this is probably too long a period for active swimming to be solely responsible for onshore movement of larvae. Pre-recruits may respond to offshore Ekman surface flow and food availability by altering their vertical distribution patterns through diel vertical migration, and field observations indicate that they show Type 1 vertical migration, being deeper (around 20-40m) during day and in surface waters at night (van der Lingen and Hutchings unpublished data). The mechanisms involved in transporting pre-recruits from offshore to onshore require further investigation. Suitably-located wind measurements and hydrodynamic numerical modelling may help improve our understanding of processes for nowcasting, but forecasting shoreward transport (and hence anchovy recruitment success) over longer periods is probably best approached in terms of testing scenarios of varying wind strength or Agulhas Bank water intrusions. The feasibility of making such forecasts is considered moderate. The economic value of such a forecast is modest, since anchovy recruitment starts in April and recruitment strength is estimated three months later through hydroacoustic surveys, and field data for such a forecast is collected by March. Requirements for forecasting include high resolution wind measurements, a set of hydrodynamic, biogeochemical and IBM models (see Penven et al. 2001; Mullon et al. 2003; Koné et al. submitted), information on the behaviour of prerecruits, and indices of food availability and the strength of Agulhas Bank water intrusions on to the west coast. The timescale of forecast possible is about one month, possibly extended up to six months, with limited feasibility of applying this forecast in management.
Forecasting shelf processes of relevance to living marine resources in the BCLME
333
Agulhas Bank to West Coast Frontal Jet Boundary A strong, northward-flowing jet current from Cape Point along the shelf-edge is associated with a marked thermal front between cold upwelled water and warm oceanic water in this region (Bang and Andrews 1974; Shelton and Hutchings 1990). Resources impacted by this boundary include small pelagic fish species, horse mackerel and hake, all of which spawn on the Agulhas Bank or the shelf slope in the extreme south of the Benguela but recruit to the inshore west coast nursery grounds (Hutchings et al. 2002a). Eggs and larvae of pelagic fish are advected rapidly alongshore between the Agulhas Bank and the west coast via the jet current during October-March, and recruitment strength should be proportional to transport “success”. If the drift of eggs and larvae is purely passive, anchovy recruitment should be inversely proportional to pathway length. However, measurements of seasonal variability in jet strength and convolutions, which would impact on pathway length, are needed from models and observations. The location of eggs and larvae in relation to the jet core is important in terms of offshore losses and feeding conditions within the convergent front (Hutchings et al. 1998). Furthermore, the strength and position of the jet relative to the coast would obviously impact on transport success. While little is known about the control of wind forcing on the behaviour of the jet, it is quite likely that wind-induced variability of the jet current contributes to explaining the inverse correlation that has been observed between southeast wind intensity and anchovy recruitment (Boyd et al. 1998). Forecasts of transport/recruitment success of anchovy based on the first half of the spawning season, from October to December, would be most useful for predicting relative recruitment strength six months later. Hindcasts to verify various models have been attempted on a number of occasions, utilising transport, feeding conditions and fish condition factors (e.g. Bloomer et al. 1994; Boyd et al. 1998; Korrûbel et al. 1998; Painting et al. 1998; see Mesoscale Processes case study), while Shannon et al. (1996) has described scenarios of increased or decreased upwelling on transport success from the Agulhas Bank to the west coast. The feasibility of the forecast is considered to be moderate, as the jet current is a very persistent feature both in hydrodynamic models of the region (Penven et al. 2001) and observations. But other factors are likely to impact on recruitment during (cannibalism and predation) and after (offshore advection, starvation, predation) the transport phase. The economic value of the forecast is considered to be substantial, as a prediction based on the first half of spawning season would allow a 6-month forecast. Replacing the currently assumed median anchovy recruitment value with one determined in part by environmental conditions during the transport phase and the size and location of the spawning stock, can easily be implemented into the operational management plan (De Oliveira and Butterworth 2005), but longer-term forecasts are not feasible until predictions of seasonal climate variability improve.
334
C.D. van der Lingen et al.
Boundaries associated with the Agulhas Bank The Agulhas Bank is an important spawning area for a large number of valuable resources, including small pelagic fish, horse mackerel, kingklip (Genypterus capensis), South Coast rock lobster (Palinurus gilchristi), and squid (Loligo vulgaris reynaudii). In addition, the Bank itself is an important nursery area for several organisms which spawn upstream on the narrow eastern coastal shelf and whose eggs and larvae are advected southwards onto the Agulhas Bank (Beckley 1993; Hutchings et al. 2002a). Three boundary processes associated with the Agulhas Bank are briefly considered here, including warm water intrusions from the Agulhas Current, the drawoff of coastal waters through premature retroflection of the Agulhas Current, and the “cool ridge” (see below). As the austral summer advances, warm water of Agulhas Current origin penetrates onto the shallow shelf of the Agulhas Bank, deepening the thermocline and reducing productivity. Internal waves produce oscillations in the thermocline, temporarily exposing nutrient-rich water to higher light levels, and increased winds may erode the thermocline deeper, mixing nutrients into the euphotic zone. During the austral winter months, the diminished heating effects and strong winds break down the water stratification to depths of 80-100m, while re-establishment of the stratification usually occurs in the October-December period (Shannon et al. 1984). Occasionally, large meanders of the Agulhas Current associated with the Natal Pulse (van Leeuwen et al. 2000) result in strong intrusions of Agulhas Current water across the Bank, or the removal of large quantities of shelf water and associated biota into the south west Indian Ocean interior. Additionally, retroflection of the Agulhas Current occasionally results in the formation of Agulhas rings (Duncombe Rae 1991) that move into the south Atlantic. These rings may entrain frontal water and draw it off from the Benguela system, possibly removing eggs, larvae and pre-recruits of small pelagic fish and hence negatively impacting on recruitment (Duncombe Rae et al. 1992). A cool, subsurface ridge often occurs on the central Agulhas Bank (Swart and Largier 1987; Boyd and Shillington 1994), driven by a complex interactions of coastal upwelling with meanders in the Agulhas Current along the shelf edge. The ridge is visible seasonally in satellite ocean colour images (Demarcq et al. 2003), and is characterised by shallow thermoclines and raised phytoplankton (Probyn et al. 1995) and zooplankton (Peterson et al. 1992) concentrations. This in situ production supplements the energy reserves of spawning fish to allow sustained serial spawning over the prolonged summer period from October to March. Cyclonic circulation patterns associated with the cool ridge (Boyd and Shillington 1994) may retain early stages of copepods over the inshore regions of the central and eastern Agulhas Bank, providing an important food source for both fish larvae (Hutchings et al. 2002a) and squid paralarvae (Augustyn et al. 1994). Additionally, the inshore, eastward-moving part of the cyclonic circulation pattern may be important in the eastward migration of horse mackerel (Barange et al. 1998). The response of the fish to changes in these boundaries processes that occur in the vicinity of the Agulhas Bank is unknown as the influence of the Agulhas Current on
Forecasting shelf processes of relevance to living marine resources in the BCLME
335
shelf water processes is not well understood. At this stage only hindcast and simulation models are feasible for comparison with observations, particularly with an expanded hydrodynamic model configuration (known as the Southern Africa Experiment or SAFE; Pierrick Penven, IRD, pers. com.). However, should forecasting of these processes and their impacts on the living marine resources be achieved, the potential economic value of such is considered to be high, particularly for squid and small pelagic fish species. Regional model simulations need to be verified with available observations on the Agulhas Bank. The data requirements are linked to an explicit understanding of the links between Agulhas Current rings and meanders, inshore eddies, early and late retroflections and intrusions onto the Bank in relation to the circulation on the Bank itself. The effect of local winds on the circulation and retention on the Bank itself, relative to offshore losses, is also important, particularly at the eastern and southern extremities of the Bank. Monitoring the Agulhas Current upstream of the Agulhas Bank will prove useful, particularly as the suspected sources of variability in the Mozambique Channel and south of Madagascar propagate southwards 1-2 months later. Further knowledge regarding biological interactions on the Agulhas Bank is also important. In particular, the role of predation and cannibalism on eggs and larvae of small pelagic fish that spawn over the Agulhas Bank during the summer months, and predation by medium-sized predatory fish on those small pelagic spawners, require investigation. The energy reserves of small pelagic fish required for sustained serial spawning and extended migration is an important biological forcing function, currently estimated via fish condition indices (see van der Lingen and Hutchings 2005), but requires further study. OTHER SHELF PROCESSES Other shelf processes not detailed above are also of major importance for the BCLME (Table 14-1), and some of them could be forecasted in the near future. Extensive sulphide eruptions are often observed in the northern Benguela, and their occurrence is likely to increase with the intensification of the trade winds in response to global greenhouse warming (Bakun and Weeks 2004). Such eruptions have a direct toxic impact as well as a secondary effect of depleting oxygen from the water, so that marine organisms suffer from severe hypoxia and anoxia. Moreover, additional mortality from predation can result from sulphide events because of exclusion of fish from their favoured near-coastal habitat by lingering anoxia, an example being the loss in the austral summer of 1992-93 of about half of the recruit population of Namibian Cape hake thought to have died as a result of being trapped by widespread anoxia in shelf bottom waters (Hamukuaya and O’Toole 1994). As a result, the size of the entire hake population of central Namibia is estimated to have fallen to less than 30% of its abundance level of the previous year (Hamukuaya et al. 1998). Sulphide eruptions are detected by ocean colour from satellite (Weeks et al. 2002; 2004), and although little can be done to prevent their negative impact, the development of continuous monitoring of the northern Benguela via satellite could facilitate the implementation of
336
C.D. van der Lingen et al.
management measures designed to limit the detrimental interaction of exploitation activities with eruptions. A challenging question in modern fishery science is to understand the processes driving the large interdecadal variability in recruitment success of some species, mainly the pelagic and semi-pelagic species, a typical example being the sardine stocks in the northern and southern Benguela subsystems. This long-term variability is often larger than the interannual variability (Schwartzlose et al. 1999), and some degree of regional or even global synchrony appears although the latter might be spurious (Fréon et al. 2004). Although the process(es) driving this interdecadal variability are largely unknown, except for some evidence of empirical correlation with environmental variables (Klyashtorin 1998), the fact that the signal is a long-term one (pseudo-periodicity of 40 to 60 years) opens the door for a certain degree of predictability and consequently the implementation of management actions. This may be particularly appropriate for resources such as sardine, which have been documented as showing a 50-60 year time scale of population expansion and contraction in other systems (Baumgartner et al. 1996 for sardine in the California Current). Fréon et al. (2005) suggested investigating a two-level (short- and long-term) management strategy, in which the second level takes into account this predictability by trying to limit the negative effects of overcapitalization in the fishing industry. Harmful algal blooms (HABs), often referred to as red tides, are mostly attributed to dinoflagellate species. Their negative impacts are due either directly through the production of toxins, or indirectly when a high biomass of dinoflagellates impacts the coastal environment and, through their decay, results in the formation of low oxygen water and, in some cases, the production of hydrogen sulphide (Pitcher and Weeks, Chapter 7, this volume). Bernard et al. (Chapter 12, this volume) explore the feasibility of forecasting HABs based on real-time observations to effectively overcome the present difficulty in modelling biological processes associated with HAB development (e.g. stochasticity of species selection, species variability in swimming, growth, mortality, nutrient and light acquisition, etc.). Those authors propose the implementation of an observation network that would utilise high frequency data from multi-sensor coastal observation platforms situated at locations critical to both preliminary bloom detection and resultant advective transport, in addition to synoptic satellite-derived data. Forecasting will be based on two complementary methods: firstly, a probabilistic “ecological window” or fuzzy logic model whereby the probability HAB occurrence is determined from observations of physical, chemical and biological conditions in real time, and secondly via short term predictions of impact and transport prediction, likely in turn to be based on a concomitant meteorological forecast (Bernard et al. Chapter 12, this volume; Pitcher et al. in press). Another area where transport prediction is essential is forecasting the transport of unexpected release of contaminants, such as oil spills and other hazardous or criminal wastes of chemicals which result in harmful exposure of marine organisms (including top predators like sea-birds, mammals and their land-based habitat, farmed species of fish and shelfish) and require expensive cleanup of the coastal environment. The response to harmful wastes depends on projections of where winds and currents will
Forecasting shelf processes of relevance to living marine resources in the BCLME
337
take the contaminants. Spill forecasting requires a detailed knowledge of contaminant release dynamics, oceanography, meteorology, contaminant chemistry, and slick observations when feasible as in the case of oil spills that can be tracked by remote sensors. Some national agencies, such as NOAA or Meteo France, implement dedicated software to track oil spills and to forecast their drift according to weather scenarios (Daniel 1996; Beegle-Krause 2001). Such an operational feature-tracking model could be implemented in the Benguela to predict the shoreline impact and retention of spills, but it will be highly dependent upon observational ability and the temporal limitations of meteorological forecasts. A major issue will be to merge realtime surface currents given by hydrodynamic models with the wave and wind-induced drift calculated by such feature-tracking models, which are mostly wind and wave driven and would not be able to incorporate features such as the Agulhas Current and eddies or filaments. Dynamic circulation models ideally should assimilate real-time physical data, in particular wind data, SST and sea level information provided by remote and land-based sensors. Remote sensing products considered most useful. in this respect include the NOAA AVHRR and NASA MODIS sensor for daily SST data at ~ 1 km spatial resolution, the NASA QuikSCAT sensor for daily surface winds at ~ 25 km spatial resolution (but without coverage within 50 km from coast), and the ESA MeteoSat Second Generation (MSG) sensor that provides 3 km spatial resolution and a 15 minutes sampling frequency allowing efficient de-clouding. The physical processes underlying the concentration, dispersion and advection are relatively well-known and should allow identifying important transport pathways and essential dynamics that may support the transport and dispersal of contaminants, or similarly, sediment plumes generated by offshore mining (Bernard et al. Chapter 12, this volume; Grundlingh et al. Chapter 15, this volume). Pollution tracking, if associated with early warning and preventative actions, could allow a saving on cleaning costs and may reduce damage to sensitive estuarine or near shore areas. Furthermore, colonies of bird and mammal species, some of which are classified as endangered or vulnerable, may also be protected. There are also direct and indirect benefits that can be expected on tourism activities (cleaner beaches, eco-tourism, etc.). DISCUSSION AND CONCLUSIONS In this chapter we focused on examining shelf processes that impact on commercially important living marine resources of the BCLME, particularly rock lobster, demersal fish and pelagic fish species. We examined three processes in detail (low oxygen events, mesoscale processes, and boundary processes) and, although treated separately, these processes are not necessarily distinct: changes in boundary conditions and/or advection may be implicated in the evolution of major low oxygen events in the northern Benguela, and features arising from mesoscale processes such as upwelling plumes, eddies etc. are important in particular boundary regions, such as the oceanic boundaries. The all-encompassing influence of physical forcing in determining the dynamics of the BCLME and other upwelling systems means that there is a high degree of correlation between the various physical processes.
338
C.D. van der Lingen et al.
Impacts on biomass may be direct, e.g. mass mortalities of rock lobster following “walkout” events due to low oxygen events; or they may be indirect, e.g. increased (or decreased) recruitment of pelagic and demersal fish that arise from a particular set or sequence of shelf processes, and reductions in growth rates etc. of rock lobster exposed to low oxygen events. Impacts on availability are direct, in that shelf processes may cause stocks to move into or away from a particular region. This may be of importance to fishery managers if such stocks cross national boundaries, as is the case with Sardinella spp. moving into Namibian waters during Benguela Niño events, and highlights the mismatch between static geopolitical boundaries and dynamic ecological boundaries. This problem will be exacerbated if climate change effects increase the likelihood and/or extent of such movements of transboundary stocks, or induces their latitudinal shifts. However, in some instances it may not be easy to discriminate between changes in availability and changes in abundance, and doing so effectively is a significant challenge that fishery managers will have to meet. Confusing the two can have serious consequences, e.g. if an increase in availability is taken as an increase in population abundance and fishing pressure is then increased, this may expose the stock to excessive fishing mortality (Fréon and Misund 1999). Being able to forecast the impacts of shelf processes on living marine resources will have obvious positive consequences for management of commercially important resources of the BCLME. The likely large-scale impacts of climate change in the Benguela region are briefly discussed by van der Lingen et al. (this volume). Altered wind stress leading to increased coastal upwelling will increase primary productivity which, in turn, increases organic loading and hence the likelihood (in terms of both frequency and intensity) of low oxygen events and sulphide eruptions (Bakun and Weeks 2004). Climate change is also likely to affect the intensity of a boundary (i.e. thermal gradient), and may also result in a consistent displacement in the modal position of a boundary. Alternatively, the modal position may stay the same but there could be a greater spatial extent and hence variability in the boundary’s location. The impacts of climate change on shelf processes, and the responses of living marine resources in the BCLME to these impacts, may best be examined through scenario testing using hydrodynamic models coupled to ecological models. Even if a good understanding of the likely impacts on living marine resources of oceanographic processes is available, and even if forecasting of these processes is feasible, addressing the question of how to incorporate such predictions into management procedures is critical. Environmental indices that have been directly used in the management of fish stocks are generally rare. For a review of the use of environmental indices in the management of small pelagic fish populations see Barange (2001; 2003). Simulation studies for South Africa’s pelagic fishery have indicated that environmental indices need to explain 50% or more of the total variation in anchovy recruitment before showing benefits in terms of management procedures (De Oliveira and Butterworth 2005). Additionally, the benefits of using environmental indices to predict recruitment may be compromised by uncertainties related to the real degrees of freedom, the mode of selection of explanatory variables, and errors in the values of explanatory variables (co-linearity, spurious correlation, etc.) and in the functional forms assumed for environment-recruitment relationships.
Forecasting shelf processes of relevance to living marine resources in the BCLME
339
An example of how an environment-recruitment relationship may change through time (i.e. show non-stationarity) has been provided by Daskalov et al. (2003), who related recruitment strength of northern Benguela sardine to two environmental indices (sea surface temperature in the tropical Atlantic and coastal wind stress at Lüderitz) over the period 1960-2000. Those authors found that sardine recruitment was positively correlated with SST and negatively correlated with wind prior to the mid-1980s, but that these relationships were reversed (recruitment being negatively correlated with SST and positively correlated with wind) thereafter. Daskalov et al. (2003) proposed two hypotheses to explain this reversal, the first of which suggested a switch between environmental regimes that occurred during the mid-1980s from an environment characterized by weak stratification and strong enrichment to one characterized by strong stratification and weak enrichment. Their second hypothesis suggested that the observed reversal could be attributed to changes in population structure, distribution and migration, arising from severe overfishing of the sardine stock and resulting in a change in spawning habitat from the vicinity of Walvis Bay to further north (see van der Lingen et al. Chapter 8, this volume, for further details on changes in northern Benguela sardine spawning habitat). The above example clearly illustrates the difficulty regarding the making of forecasts when not all processes are clearly identified. By contrast, it is technically relatively simple to implement an early warning signal of major changes in the environmental forcing used in the model of Daskalov et al. (2003; see for instance a pre-project of a Benguela Niño warning system set up as the result of this international workshop at: http://drmroull.sea.uct.ac.za/). The unsuccessful attempt to use a significant relationship between an upwelling index and recruitment determined for Bay of Biscay anchovy (Borja et al. 1998) to set total allowable catches for this species (ICES 2000; 2001) has been partially attributed (De Oliveira and Butterworth 2005) to a failure to take proper account of risks associated with the uncertainties listed above, although a lack of clear management objectives for that fishery, and conflicting requirements between French and Spanish fishermen, is considered the major issue (M. Barange, GLOBEC IPO, pers. comm.). Nonetheless, the wealth of studies that have examined anchovy recruitment variability and its predictability in the Southern Benguela (Hutchings et al. 1998; Mullon et al. 2003; van der Lingen and Huggett 2003), and the insights gained from simulating the incorporation of predictive models into management procedures for this stock, should allow the development of recruitment prediction models that can feasibly be incorporated into management procedures. A suggested way forward is to update previous approaches to predicting anchovy recruitment by including recent data (such as the eastward shift in spawning intensity on the Agulhas Bank; van der Lingen et al. 2002) and then combining these individual approaches into an expert system in which the recruitment predictions are weighted according to quantitative (e.g. r2 value of the environment-recruitment value, where available) and/or qualitative (e.g. perceptions of the “realism” of the approach and its ability to capture the underlying process) criteria in order to arrive at a single most likely prediction. Regarding sardine recruitment forecasting, the long-term autocorrelation observed in catch and abundance time series (Baumgartner et al. 1996; Fréon et al. 2005) could be incorporated into management procedures in the absence of further evidence that the observed pseudo-periodicity is environmentally driven (Klyashtorin 1998). However, periods of high and low
340
C.D. van der Lingen et al.
productivity (as evidenced by estimates of instantaneous surplus production rate [ISPR] and annual surplus production [ASP]) of sardine populations from a variety of regions have been linked to environmental variability (Jacobson et al. 2001), and an environmentally responsive management approach that adaptively controls catch by shifting fishing effort in response to changing stock productivity has been suggested for Japanese sardine (Yatsu et al. 2005). Finally, we emphasize that predictions should be based on a solid understanding of the processes involved and their spatio-temporal location in terms of the life history of the marine resource on which they impact, rather than solely on statistical correlations between process and resource that are poorly explained and/or not understood, and often break down after a number of years (see Myers 1998), although inter-decadal changes might be an exception to this rule. ACKNOWLEDGEMENTS The issues presented in this chapter were discussed at Specialist Session F, “Forecasting shelf processes relevant to living marine resources” of the Benguela Forecasting Workshop. The Session was convened by Drs P. Fréon (IRD) and C. van der Lingen (MCM), and chaired by Prof. J.G. Field (UCT). We gratefully acknowledge contributions made by participants during the workshop session, and wish to thank Cathy Boucher (MCM) for redrawing some of the figures. Comments on earlier versions of the manuscript by Dr M. Barange (GLOBEC IPO) and two anonymous referees are warmly acknowledged. This is a contribution of the IDYLE/ECO-UP and Upwelling Ecosystem programmes of IRD and EUR-OCEANS (a European Network of Excellence funded by the European Commission under the 6th Framework Programme - contract ref. 511106), respectively. REFERENCES Agenbag, J.J. and L.V. Shannon. 1988. A suggested physical explanation for the existence of a biological boundary at 24°30’S in the Benguela System. S. Afr. J. mar. Sci. 6:119-132. Agostini, V.N., and A. Bakun. 2002. 'Ocean triads' in the Mediterranean Sea: physical mechanisms potentially structuring reproductive habitat suitability (with example application to European anchovy, Engraulis encrasicolus). Fish. Oceanogr. 11:129–142. Andrews, W.R.H. and L. Hutchings. 1980. Upwelling in the Southern Benguela Current. Prog. Oceanogr. 9:181. Armstrong, M.J., P.A. Shelton, R.M. Prosch and W.S. Grant. 1983. Stock assessment and population dynamics of anchovy and pilchard in ICSEAF Division 1.6 in 1982. Colln. Sci. Pap. Int. Comm. S.E. Atl. Fish. 10: 7-25. Augustyn, C.J., M.R. Lipinski, W.H.H. Sauer, M.J. Roberts and B.A. Mitchell-Innes. 1994. Chokka squid on the Agulhas Bank: life history and ecology. S. Afr. J. Sci. 90: 143-154. Bailey, G.W., C.J.deB. Beyers, and S.R. Lipschitz. 1985. Seasonal variation of oxygen deficiency in waters off southern South West Africa in 1975 and 1976 and its relation to the catchability and distribution of the cape rock lobster Jasus lalandii. S. Afr. J. mar. Sci. 3:197-214. Bailey, G.W. 1991. Organic carbon flux and development of oxygen deficiency on the modern Benguela continental shelf south of 22°S: spatial and temporal variability. In Tyson, R.V. and T.H. Pearson, eds. Modern and Ancient Continental Shelf Anoxia., Geological Society Special Publication 58:171-183.
Forecasting shelf processes of relevance to living marine resources in the BCLME
341
Bakun, A. 1990. Global climate change and intensification of coastal upwelling. Science 247:198-201. Bakun, A. 1996. Patterns in the Ocean: Ocean Processes and Marine Population Dynamics. University of California Sea Grant Program, San Diego, California, USA, in cooperation with Centro de Investigaciones Biológicas del Noroeste, La Paz, Mexico. California Sea Grant College, University of California, La Jolla, 323 p. Bakun, A. and S. Weeks. 2004. Greenhouse gas buildup, sardines, submarine eruptions and the possibility of abrupt degradation of intense marine upwelling systems. Ecology Letters 7:1015-1023. Bang, N.D. and W.R.H. Andrews. 1974. Direct current measurements of a shelf-edge frontal jet in the southern Benguela system. J. Mar. Res. 32:405-417. Baptista, S.R.da F. 1977. The distribution and movements of the sardinellas (Sardinella aurita Val. and Sardinella eba Val.) off the Angolan coast. Colln scient. Pap. int. Commn SE Atl. Fish. 4:21-24. Barange, M., ed. 2001. Report of the 1st meeting of the SPACC/IOC Study Group on “Use of environmental indices in the management of pelagic fish populations (3-5 September 2001, Cape Town, South Africa)”. GLOBEC Special Contribution No. 5, 122 pp. Barange, M., ed. 2003. Report of the 2nd meeting of the SPACC/IOC Study Group on “Use of environmental indices in the management of pelagic fish populations (9-11 December 2002, Paris, France)”. GLOBEC Special Contribution No. 6. 156 pp. Barange M. and S. C. Pillar. 1992. Cross-shelf circulation, zonation and maintenance mechanisms of Nyctiphanes capensis and Euphausia hanseni (Euphausiacea) in the northern Benguela upwelling system. Cont. Shelf Res. 12:1027-1042. Barange M., S. C. Pillar and I. Hampton. 1998. Distribution patterns, stock size and life-history strategies of Cape horse mackerel Trachurus trachurus capensis, based on bottom trawl and acoustic surveys. S. Afr. J. mar. Sci. 19:433-447. Barange M., S. C. Pillar and L. Hutchings.1992. Major pelagic borders of the Benguela upwelling system according to euphausiid species distribution. S. Afr. J. mar. Sci. 12:3-17. Barth, J.A. 1989a. Stability of a coastal upwelling front, 1, Model development and a stability theorem. J. Geophys. Res. 94:10844-10856. Barth, J.A. 1989b. Stability of a coastal upwelling front, 2, Model results and a comparison with observations. J. Geophys. Res. 94:10857-10883. Batchelder H. P., C. A. Edwards and T. M. Powell. 2002. Individual-based models of copepod populations in coastal upwelling regions: implications of physiologically and environmentally influenced diel vertical migration on demographic success and nearshore retention. Prog. Oceanogr. 53(2-4): 307-333. Baumgartner, T., U. Struck and J. Alheit. 2004. GLOBEC Investigation of Interdecadal to MultiCentiennial Variability in Marine Fish Populations. PAGES News 12:19-21. Baumgartner, T., A. Soutar and W. Riedel. 1996. Natural time scales of variability in coastal fish populations of the California Current over the past 1500 years: Responses to global climate change and biological interaction. In Calif. Sea Grant Rep. 1992-1995, Calif. Sea Grant College, La Jolla, 31-37. Beckley, L.E. 1993. Linefish larvae and the Agulhas Current. In: Fish, Fishers and Fisheries – Proceedings of the Second South African Linefish Symposium. L.E. Beckley and R.P. van der Elst,eds. Special Publication, Oceanic Research Institute 2:57-63. Beegle-Krause, C.J. 2001. General NOAA Oil Modeling Environment (GNOME): a new spill trajectory model. In Proc. 2001 Int. Oil Spill Confer., Tampa FL. API, Washington DC, 865-871. Bernard, S., R.M. Kudela, P. Franks, W. Fennel, A. Kemp, A. Fawcett and G.C. Pitcher, This volume. The requirements for forecasting harmful algal blooms in the Benguela. Part III: Chapter 12 In: Benguela: Predicting a Large Marine Ecosystem. Elsevier Series, Large Marine Ecosystems. Binet, D., B. Gobert, and L. Maloueki. 2001. El Niño-like warm events in the Eastern Atlantic (6°N, 20°S) and fish availability from Congo to Angola (1964-1999). Aquat. Living Resour. 14:99-113. Blanke, B., C. Roy, P. Penven, S. Speich, J. McWilliams and G. Nelson. 2002. Linking wind and interannual upwelling variability in a regional model of the southern Benguela. Geophys. Res. Lett. 29: 2188, doi:2110.1029/2002GL015718. Blanke, B., S. Speich, A. Bentamy, C. Roy and B. Sow. 2005. Southern Benguela upwelling and QuikSCAT wind variability. J. Geophys. Res. 110: C07018, doi:10.1029/2004JC002529, 2005 Bloomer, S.F., K.L. Cochrane and J.G. Field. 1994. Towards predicting recruitment success of anchovy Engraulis capensis Gilchrist in the southern Benguela system using environmental variables: A rulebased model. S. Afr. J. mar. Sci. 14:107-119.
342
C.D. van der Lingen et al.
Borja, A., A. Uriarte, J. Egaña, L. Motos and V. Valencia. 1998. Relationships between anchovy (Engraulis encrasicolus) recruitment and environment in the Bay of Biscay (1978-1996). Fish. Oceanogr. 7:375-380. Boyd, A.J., L.J. Shannon, F.H. Schülein and J. Taunton-Clark. 1998. Food, transport and anchovy recruitment in the southern Benguela upwelling system off South Africa. In: Durand, M.H., P. Cury, R. Mendelssohn, C. Roy, A. Bakun and D. Pauly, eds. Global versus Local Changes in Upwelling Systems., ORSTOM editions, Paris: 195-210. Boyd, A.J. and F.A. Shillington. 1994. Physical forcing and circulation patterns on the Agulhas Bank. S. Afr. J. Sci. 90:114-122. Boyer, D.C. and I. Hampton. 2001. An overview of the living marine resources of Namibia. S. Afr. J. mar. Sci. 23:5-35. Brink, K.H. 1983. The near-surface dynamics of coastal upwelling. Prog. Oceanogr. 12:223-257. Brink, K.H. and T. Cowles. 1991. The Coastal Transition Zone Program. J. Geophys. Res. 96:1463714647. Bubnov, V.A. 1972. Structure and characteristics of the oxygen minimum layer in the Southeastern Atlantic. Oceanology 12(2): 193-201. Burmeister, L.-M. 2001. Depth-stratified density estimates and distribution of the Cape hake Merluccius capensis and M. paradoxus off Namibia deduced from survey data, 1990-1999. S. Afr. J. mar. Sci. 23:347-356. Butterworth, D.S. 1983. Assessment and management of pelagic stocks in the southern Benguela region. In: Proceedings of the Expert Consultation to Examine Changes in Abundance and Species Composition of Neritic Fish Resources, San José, Costa Rica, April 1983. G.D. Sharp and J. Csirke, eds. FAO Fisheries Report 291:329-405. Chapman, D.C. and S.J. Lentz. 1994. Trapping of a coastal density front by the bottom boundary layer. J. Phys. Oceanogr. 24:1464-1479. Chapman, P. and L.V. Shannon. 1985. The Benguela Ecosystem. Part II. Chemistry and Related Processes. Oceanogr. Mar. Biol. Ann. Rev. 23:183-251. Clemmesen, C., R. Sanchez, and C. Wongtschowski. 1997. A regional comparison of the nutritional condition of SW Atlantic anchovy larvae, Engraulis anchoita, based on RNA/DNA ratios. Archiv Fisch. Meeres. 45:17-43. Cochrane, K.L. and A.M. Starfield. 1992. The potential use of predictions of recruitment success in the management of the South African anchovy resource. S. Afr. J. mar. Sci. 12:891-902. Cockcroft, A.C. 2001. Jasus lalandii “walkouts” or mass strandings in South Africa during the 1990s: An overview. Mar. Freshwat. Res. 52:1085-1094. Copenhagen, W.J. 1953. The periodic mortality of fish in the Walvis Region. S.A. J. Sci. 49:330-331. Daniel, P. 1996. Operational Forecasting of Oil Spill Drift at Météo-France. Spill Science & Technology Bulletin. 3(1/2): 53–64. Daskalov, G.M., D.C. Boyer and J.P. Roux. 2003. Relating sardine Sardinops sagax abundance to environmental variables in the northern Benguela. Prog. Oceanogr. 59:257-274. De Decker, A.H.B. 1970. Notes on an oxygen-depleted subsurface current off the west coast of South Africa. S. A. Div. Sea Fish. Invest. Rep. 84, 27pp. Demarcq, H., R.G. Barlow and F.A. Shillington. 2003. Climatology and variability of sea surface temperature and surface chlorophyll in the Benguela and Agulhas Ecosystems as observed by satellite imagery. Afr. J. mar. Sci. 25:363-372. De Oliveira, J.A.A. and D.S. Butterworth. 2005. Limits to the use of environmental indices to reduce risk and/or increase yield in the South African anchovy fishery. Afr. J. mar. Sci. 27(1): 191-203. Dickey, T.D. 2003. Emerging ocean observations for interdisciplinary data assimilation systems. J. Mar. Syst. 40-41: 5-48. Duncombe Rae, C.M. 1991. Agulhas retroflection rings in the South Atlantic Ocean: an overview. S. Afr. J. mar. Sci. 11:327-344. Duncombe Rae, C.M., A.J. Boyd and R.J.M. Crawford. 1992. “Predation” of anchovy by an Agulhas ring: A possible contributory cause of the very poor year-class of 1989. S. Afr. J. mar. Sci. 12:167-173. Emeis, K.-C., V. Bruchert, B. Currie, R. Endler, T. Ferdelman, A. Kiessling, T. Leipe, K. Noli-Peard, U. Struck and T Vogt. 2004. Shallow gas in shelf sediments of the Namibian coastal upwelling ecosystem. Cont. Shelf Res. 24:627-642. Feistel, R., E. Hagen and K. Grant. 2003. Climatic changes in the subtropical Southeast Atlantic: the St. Helena Island Climate Index (1893-1999). Prog. Oceanogr. 59:321-337.
Forecasting shelf processes of relevance to living marine resources in the BCLME
343
Field, J.G. and F.A. Shillington. 2006. Variability of the Benguela Current System. In The Sea, Vol. 14, Coasts Eds. A. R. Robinson and K. H. Brink, Harvard University Press, Boston, USA: 833-861. Florenchie, P., C.J.C. Reason, J.R.E. Lutjeharms, M. Rouault, C. Roy, and S. Masson. 2004. Evolution of interannual warm and cold events in the southeast Atlantic Ocean. J. Climate 17:2318-2334. Fréon, P. and O.A. Misund. 1999. Dynamics of Pelagic Fish Distribution and Behaviour: Effects on Fisheries and Stock Assessment. Fishing News Books, Oxford, United Kingdom, 348 pp. Fréon, P., C. Mullon and B. Voisin. 2004. Investigating remote synchronous patterns fisheries. Fish. Oceanogr. 12: 443-457. Fréon, P., P. Cury, L. Shannon and C. Roy. 2005. Sustainable exploitation of small pelagic fish stocks challenged by environmental and ecosystem changes: a review. Bull of Mar. Sci. 76(2): 385-462(78). Graham, W.M. and J.L. Largier. 1997 Upwelling shadows as nearshore retention sites: The example of northern Monterey Bay. Cont. Shelf Res. 17:509-532 Griffiths, C.L., L. van Sittert, P.B. Best, A.C. Brown, P.A. Cook, R.J.M. Crawford, J.H.M. David, B.R. Davies, M.H. Griffiths, K. Hutchings, A. Jerardino, N. Kruger, S. Lamberth, R. Leslie, R. MelvilleSmith, R. Tarr and C.D. van der Lingen. 2004. Impacts of human activities on marine animal life in the Benguela –A historical overview. Oceanog. Mar. Biol. Ann. Rev. 42:303-392. Grundlingh, M., P. Morant, R. van Ballegooyen, A. Badenhorst, E. Gomes, L. Greyling, J. Guddal, I. Huntr, D. Japp, L. Maartens, K. Peard, G. Smith, C. Wainman. This volume. Environmental requirements of maritime operations in the Benguela coastal ocean. In: Shannon, V., G. Hempel, P. Malanotte-Rizzoli, C. Moloney and J. Woods, eds. Benguela: Predicting a Large Marine Ecosystem. Elsevier Series, Large Marine Ecosystems. Part III: Chapter 15. Hampton, I. 1987. Acoustic study on the abundance and distribution of anchovy spawners an recruits in South African waters. S. Afr. J. mar. Sci. 5:901-917. Hamukuaya, H. and M.J. O’Toole. 1994. Survey of juvenile hake in the central Namibian region, March 1994. Internal Report, National Marine Information and Research Centre, Swakopmund, Namibia: 24 pp. Hamukuaya, H., M.J. O’Toole, and P.M.J. Woodhead. 1998. Observations of severe hypoxia and offshore displacement of Cape hake over the Namibian shelf in 1994. S. Afr. J. mar. Sci. 19:57-59. Hart, T.J. and R.I. Currie. 1960. The Benguela Current. "Discovery" Rep. 31:123-297. Hewitson, J.D. 1988 Catch trends in the multispecies pelagic fishery off Namibia in 1987. Colln scient. Pap. int. Commn SE Atl. Fish. 15(II): 7-17. Hewitson, J.D., Y.C. Melo, and R.M. Cooper.1989. The Namibian pelagic fishing resource during 1988. Colln scient. Pap. int. Commn SE Atl. Fish. 16(II): 119-131. Hill, A.E., B.M. Hickey, F.A. Shillington, P.T. Strub, K.H. Brink, E.D. Barton and A.C. Thomas. 1998. Eastern Ocean Boundaries. 29-68 In: Robinson, A.R. and K.H. Brink, eds. The Sea. Volume 11. The Global Coastal Ocean Regional Studies and Syntheses. J. Wiley and Sons, New York. Huggett, J. A., A.J. Boyd, L. Hutchings and A.D. Kemp. 1998.Weekly variability of clupeoid eggs and larvae in the Benguela jet current: implications for recruitment. S. Afr. J. mar. Sci. 19:197-210. Huggett, J.A., P. Fréon, C. Mullon and P. Penven. 2003. Modelling the transport success of anchovy Engraulis encrasicolus eggs and larvae in the southern Benguela: The effects of spatio-temporal spawning patterns. Mar. Ecol. Prog. Ser. 250:247-262. Hunter, J.R. 1972. Swimming and feeding behaviour of larval anchovy Engraulis mordax. Fish. Bull. US. 70:821-838 Hutchings, L. 1992. Fish harvesting in a variable, productive environment - searching for rules or searching for exceptions? S. Afr. J. mar. Sci. 12: 297-318. Hutchings, L., M. Barange, S.F. Bloomer, A.J. Boyd, R.J.M. Crawford, J.A. Huggett, M. Kertsan, J.L. Korrûbel, J.A.A. De Oliveira, S.J. Painting, A.J. Richardson, L.J. Shannon, F.H. Schülein, C.D. van der Lingen and H.M. Verheye. 1998. Multiple factors affecting South African anchovy recruitment in the spawning, transport and nursery areas. S. Afr. J. mar. Sci. 19:211-225. Hutchings, L., L.E. Beckley, M.H. Griffiths, M.J. Roberts, S. Sundby and C.D. van der Lingen. 2002a. Spawning on the edge: spawning grounds and nursery areas around the southern African coastline. Mar. Freshwater Res. 53:307-318. Hutchings, L., C. Duncombe-Rae, B. Mitchell-Innes, G. Bailey, H. Verheye and T. Tanci. 2002b. South African west coast ravaged by predatory hordes. In: Symposium Handbook. Southern African Marine Science Symposium 2002, 1-5 July, Swakopmund, Namibia, p 40. ICES. 2000. Report of the Working Group on the Assessment of Mackerel, Horse Mackerel, Sardine and Anchovy. ICES, CM 2000/ACFM:05.
344
C.D. van der Lingen et al.
ICES. 2001. Report of the Working Group on the Assessment of Mackerel, Horse Mackerel, Sardine and Anchovy. ICES, CM 2001/ACFM:06. Jacobson, L.D, J.A.A. De Oliveira, M. Barange, M.A. Cisneros-Mata, R. Felix-Uraga, J.R. Hunter, J.Y. Kim, Y. Matsuura, M. Ñiquen, C. Porteiro, B. Rothschild, R.P. Sanchez, R. Serra, A. Uriarte and T. Wada. 2001. Surplus production, variability, and climate change in the great sardine and anchovy fisheries. Can. J. Fish. Aquat. Sci. 58:1891-1903. King, D.P.F. 1977. Distribution and relative abundance of eggs of the South West African pilchard Sardinops ocellata and anchovy Engraulis capensis, 1971/72. Fish. Bull. S. Afr. 9:23-31. Klyashtorin, L.B. 1998. Long-term climate change and main commercial fish production in the Atlantic and Pacific. Fish. Res. 37:115-125. Koné, V., E. Machu, P. Penven, V. Andersen, V. Garçon, H. Demarcq, and P. Fréon, Submitted. Modelling primary and secondary productions of the Southern Benguela upwelling system: a comparative study through two biogeochemical models, Global Biogeochemical Cycles. Korrûbel, J.L., S.F. Bloomer, K.L. Cochrane, L. Hutchings and J.G. Field. 1998. Forecasting in South African fisheries management: The use of expert and decision support systems. S. Afr. J. mar. Sci. 19:415-423. Lett, C., C. Roy, A. Levasseur, C.D. van der Lingen and C. Mullon. In press. Simulation and quantification of enrichment and retention processes in the southern Benguela upwelling ecosystem. Fish. Oceanogr. Lutjeharms, J.R.E. and P.L. Stockton. 1987. Kinematics of the upwelling front off southern Africa. S. Afr. J. mar. Sci. 5:35-49. Luyeye, N. 1995. Distribution and abundance of small pelagic fish off Angolan coast. ICES CM1995/H:32. ICES Copenhagen. MacCall, A.D.. 1990. Dynamic Geography of Marine Fish Populations. Seattle; University of Washington Press: 153 pp. Macpherson, E., M. Masó, M., Barange and A. Gordoa. 1991. Relationship between measurements of hake biomass and sea surface temperature off southern Namibia. S. Afr. J. mar. Sci. 10:213-217. Miller, D.C.M., C.L. Moloney, C.D. van der Lingen, C. Lett, C. Mullon and J.G. Field. In press. Modelling the effects of physical-biological interactions and spatial variability in spawning and nursery areas on transport and retention of sardine eggs and larvae in the southern Benguela ecosystem. J. Mar. Sys. Monteiro, P.M.S. and A.K. van der Plas. This volume. Forecasting Low Oxygen Water (LOW) variability in the Benguela System. In: Benguela: Predicting a Large Marine Ecosystem. Elsevier Series, Large Marine Ecosystems Part II: Chapter 5. Monteiro, P.M.S, A.K. van der Plas, G.W. Bailey, P. Rizzoli, C. Duncombe Rae, D. Byrnes, G. Pitcher, J. Fitzpatrick, and U. Lass. This volume. Low Oxygen Water (LOW) forcing scales amenable to forecasting in the Benguela ecosystem. In: Benguela: Predicting a Large Marine Ecosystem. Elsevier Series, Large Marine Ecosystems Part III: Chapter 13. Moroshkin, K.V., V.A. Bubnov, and R.P. Bulatov. 1970. Water circulation in the eastern South Atlantic Ocean. Oceanology 10(1): 27-34. Mullon, C., P. Fréon, C. Parada, C.D. van der Lingen and J.A. Huggett. 2003. From particles to individuals: Modeling the early stages of anchovy (Engraulis capensis/encrasicolus) in the southern Benguela. Fish. Oceanogr. 12:396-406. Myers, R.A. 1998. When do environment-recruitment correlations work? Rev. Fish Biol. Fish. 8:285-305. Nelson G. and L. Hutchings. 1983. The Benguela upwelling area. Prog. Oceanogr. 12:333-356. O’Toole, M. and I. Hampton. 1989. New prospects for surveying and sampling anchovy pre-recruits. SA Shipp. News Fish. Ind. Rev. October 1989: 32-33. Painting, S.J., L. Hutchings, J.A. Huggett, J.L. Korrûbel, A.J. Richardson and H.M. Verheye. 1998. Environmental and biological monitoring for forecasting anchovy recruitment in the southern Benguela upwelling region. Fish. Oceanogr. 7:364-374. Parada, C., C. Mullon, C. Roy, P. Fréon, L. Hutchings and C.D. van der Lingen. Submitted. Modeling the contribution of vertical migration to pre-recruitment success of anchovy early stages in the southern Benguela ecosystem. Fish. Oceanogr. Parada, C., C.D. van der Lingen, C. Mullon and P. Penven 2003. Modeling the effect of buoyancy on the transport of anchovy (Engraulis capensis) eggs from spawning to nursery grounds in the southern Benguela: An IBM approach. Fish. Oceanogr. 12:1-15.
Forecasting shelf processes of relevance to living marine resources in the BCLME
345
Penven, P., C. Roy, G. Brundrit, A. Colin de Verdière, P. Fréon, A. Jonhson, J.R.E. Lutjeharms and F.A. Shillington. 2001. A regional hydrodynamic model of upwelling in the Southern Benguela. S. Afr. J. Sci. 97:472-475. Penven P., C. Roy, A. Colin de Verdiere and J. Largier. 2000. Simulation of a coastal jet retention process using a barotropic model. Oceanologica Acta 23:615-634 Petersen, W.T., L. Hutchings, J.A. Huggett and J.L. Largier. 1992. Anchovy spawning in relation to the biomass and the replenishment rate of their copepod prey on the western Agulhas Bank. S. Afr. J. mar. Sci. 12: 487-500. Pitcher, G., P. Monteiro and A. Kemp. In Press. The potential use of a hydrodynamic model in the prediction of harmful algal blooms in the southern Benguela. In: Steidinger, K., ed. Harmful and Toxic Algal Blooms. Intergovernmental Oceanographic Commission of UNESCO. Pitcher, G.C. and S. Weeks. This volume. The variability and potential for prediction of harmful algal blooms in the southern Benguela ecosystem. Part II: Chapter 7 In: Benguela: Predicting a Large Marine Ecosystem. Elsevier Series, Large Marine Ecosystems Probyn, T.A., B.A. Mitchell-Innes and S. Searson. 1995. Primary productivity and nitrogen uptake in the subsurface chlorophyll maximum on the Eastern Agulhas Bank. Cont. Shelf Res. 15:1903-1920. Provost, C., R. Lampitt, D. Stuben, A. Meirier, M. Garcier, V. Garçon, A. Cruzado and A. Weeks. 1998. YOYO 2001: Ocean Odyssey. Third European marine science and technology conference (MAST conference), Lisbon, 23-27 May 1998: Project synopses Vol. 4: Advanced Systems. Vol. 4: 14291432. Richardson, A.J., C. Risien and F.A. Shillington. 2003. Using self-organizing maps to identify patterns in satellite imagery. Prog. Oceanogr. 59:223-239. Roy, C., P. Fréon, and C.D. van der Lingen. 2002. An empirical model of anchovy recruitment in the Southern Benguela. GLOBEC Report 16:52-54. Roy, C., S. Weeks, M. Rouault, G. Nelson, R. Barlow and C.D. van der Lingen. 2001. Extreme oceanographic events recorded in the Southern Benguela during the 1999-2000 summer season. S. Afr. J. Sci. 97:465-471. Schwartzlose, R.A., J. Alheit, A. Bakun, T.R. Baumgartner, R. Cloete, R.J.M. Crawford, W.J. Fletcher, Y. Green-Ruiz, E. Hagen, T. Kawasaki, D. Lluch-Belda, S.E Lluch-Cota, A.D. MacCall, Y. Matsuura, M.O. Nevarez-Martinez, R.H. Parrish, C. Roy, R. Serra, K.V. Shust, M.N. Ward and J.Z. Zuzunaga. 1999. Worldwide large-scale fluctuations of sardine and anchovy populations. S. Afr. J. mar. Sci. 21:289-347. Servain, J., A.J. Busalacchi, M.J. McPhaden, A.D. Moura, G. Reverdin, M. Vianna, and S.E. Zebiak. 1998. A Pilot Research Moored Array in the Tropical Atlantic (PIRATA). Bull. Amer. Meteor. Soc. 79:2019-2031. Shannon, L.J., G. Nelson, R.J.M. Crawford and A.J. Boyd. 1996. Possible impacts of environmental change on pelagic fish recruitment: modelling anchovy transport by advective processes in the southern Benguela. Global Change Biology 2:407-420. Shannon, L.V., J.J. Agenbag, and M.E.L. Buys. 1987. Large- and mesoscale features of the AngolaBenguela Front. S. Afr. J. mar. Sci. 5:11-34. Shannon, L.V., A.J. Boyd, G.B. Brundrit and J. Taunton-Clark. 1986. On the existence of an El Niño-type phenomenon in the Benguela system. J. Mar. Res 44 (3): 495-520. Shannon, L.V., R.J.M., Crawford, G.B. Brundrit and L.G. Underhill. 1988. Responses of fish populations in the Benguela ecosystem to environmental change. J. Cons. Int. Explor. Mer 45:5-12. Shannon, L.V., L. Hutchings, G.W. Bailey and P.A. Shelton. 1984. Spatial and temporal distribution of chlorophyll in southern African wasters as deduced from ship and satellite measurements and their implications for pelagic fisheries. S. Afr. J. mar. Sci. 2:109-130. Shannon, L. V., and G. Nelson. 1996. The Benguela: Large scale features and processes and system variability. 163-210 In: Wefer, G., W.H. Berger, G. Siedler, and D.J. Webb, eds. The South Atlantic past and present circulation. Springer Verlag, Berlin, Heidelberg: Shelton, P.A. and L. Hutchings. 1990. Ocean stability and anchovy spawning in the southern Benguela Current region. Fish. Bull. 88: 323-338. Shillington, F.A. 1998. The Benguela upwelling system off southwestern Africa. The Sea 11:583-604. Shillington, F.A., L. Hutchings, T.A. Probyn, H.N. Waldron, and W.T. Peterson. 1992. Filaments and the Benguela frontal zone: Offshore advection or recirculating loops? S. Afr. J. mar. Sci. 12:207-218 Shillington, F. C.J.C. Reason, C.M. Duncombe Rae, P. Florenchie, and P. Penven. This Volume. Large scale physical variability of the Benguela Current Large Marine Ecosystem (BCLME) In: Shannon, V.
346
C.D. van der Lingen et al.
G. Hempel, P. Malanotte-Rizzoli, C. Moloney and J. Woods, eds. Benguela: Predicting a Large Marine Ecosystem. Elsevier Series, Large Marine Ecosystems Part II: Chapter 4. Simpson, J.H. and J.R. Hunter. 1974. Fronts in the Irish Sea. Nature 250:404-406. Sinclair, A.F. and W.R. Crawford. 2005. Incorporating an environmental stock-recruitment relationship in the assessment of Pacific cod (Gadus macrocephalus). Fish. Oceanogr. 14:138-150. Skogen, M.D., L.J. Shannon and J.E. Stiansen. 2003. Drift patterns of anchovy (Engraulis capensis) larvae in the southern Benguela and their possible importance for recruitment. Afr. J. Mar. Sci. 25:37– 47. Smith, R.L. 1995. The physical processes of coastal ocean upwelling systems. 39-64 In: Summerhayes, C.P., K.-C. Emeis, M.V. Angel, R.L. Smith and B. Zeitzsche, eds. Upwelling the Ocean: Modern processes and Ancient Records. J. Wiley & Sons, Inc., Chichester. Stander, G.H. 1964. The Benguela Current off South West Africa. Investl Rep. mar. Res. Lab. S.W. Afr. 12: 43 pp. + Plates 5-81. Stenevik E.K., M. Skogen, S. Sundby and D. Boyer. 2003. The effect of vertical and horizontal distribution on retention of sardine (Sardinops sagax) larvae in the Northern Benguela - observations and modelling. Fish. Oceanogr. 12:185-200. Stenevik, E.K., S. Sundby and R. Cloete. 2001. Influence of buoyancy and vertical distribution of sardine Sardinops sagax eggs and larvae on their transport in the northern Benguela ecosystem. S. Afr. J. mar. Sci. 23:85-97. Sundby, S., A.J. Boyd, L. Hutchings, M.J. O’Toole, K. Thorisson and A. Thorsen. 2001. Interaction between Cape hake spawning and the circulation in the Northern Benguela upwelling ecosystem. S. Afr. J. mar. Sci. 23:317-336. Swart, V.P. and J.L. Largier. 1987. Thermal structure of Agulhas Bank water. S. Afr. J. mar. Sci. 5:243253. Thomas, R.M.. 1984. Catches of Sardinella aurita off South West Africa in 1983. Colln scient. Pap int. Comm. SE Atl. Fish. 11(II): 87-90. Thouron, D., R. Vuillemin, X. Phillopon, A. Lourenco, C. Provost, A. Cruzado and V. Garcon. 2003. An autonomous nutrient analyser for oceanic long-term in situ biogeochemical monitoring. Anal. Chem. 75:2601-2609. van der Lingen, C.D., J.C. Coetzee and L. Hutchings. 2002. Temporal shifts in the spatial distribution of anchovy spawners and their eggs in the Southern Benguela: Implications for recruitment. GLOBEC Report 16:46-48. van der Lingen, C.D. and J.A. Huggett. 2003. The role of ichthyoplankton surveys in recruitment research and management of South African anchovy and sardine. In: The Big Fish Bang. Proceedings of the 26th Annual Larval Fish Conference. H.I. Browman and B.A. Skiftesvik, eds. Institute of Marine Research, Bergen, Norway: 303-343. van der Lingen, C.D. and L. Hutchings. 2005. Estimating the lipid content of pelagic fish in the southern Benguela by visual assessment of their mesenteric fat. A. J. mar. Sci. 27(1): 45-53 van der Lingen, C.D., L. Hutchings, D. Merkle, J.J. van der Westhuizen and J. Nelson. 2001. Comparative spawning habiutats of anchovy (Engraulis capensis) and sardine (Sardinops sagax) in the southern Benguelas upwelling ecosystem. 185-209 In: Kruse, G.H. N. Bez, A. Booth, M.W. Dorn, S. Hills, R.N. Lipcius, D. Pelletire, C. Roy, S.J. Smith and D. Witherell, eds. Spatial Processes and Management of Marine Populations. University of Alaska Sea Grant, AK-SG-01-02, Fairbanks. van der Lingen, C.D., L.J. Shannon, P. Cury, A. Kreiner, C.L. Moloney, J.P. Roux, and F. Vaz-Velho. This volume. Resource and ecosystem variability, including regime shifts, in the Benguela Current System. Part II: Chapter 8. In: Shannon, V., G. Hempel, P. Malanotte-Rizzoli, C. Moloney and J. Woods, eds. Benguela: Predicting a Large Marine Ecosystem. Elsevier Series, Large Marine Ecosystems. van Leeuwen, P.J., W.P.M. de Ruijter and J.R.E. Lutjeharms. 2000. Natal pulses and the formation of Agulhas rings. J. Geophys. Res. 105:6425-6436. Voges, E., A. Gordoa, C.H. Bartholomae and J.G. Field. 2002. Estimating the probability of different levels of recruitment for Cape hakes Merluccius capensis off Namibia, using environmental indices. Fish. Res. 58:333-340. Weeks, S.J., B. Currie and A. Bakun. 2002. Satellite imaging: Massive emissions of toxic gas in the Atlantic. Nature 415:492–493.
Forecasting shelf processes of relevance to living marine resources in the BCLME
347
Weeks, S.J., B. Currie, A. Bakun and K. Peard. 2004. Hydrogen sulphide eruptions in the Atlantic Ocean off southern Africa: implications of a new view based on SeaWiFS satellite imagery. Deep-Sea Res. I, 51: 153–172. Werner, F.E. and J.A. Quinlan. 2002. Fluctuations in marine fish populations: physical processes and numerical modelling. ICES Marine Science Symposia 215:264-278. Yatsu, A., T. Watanabe, M. Ishida, H. Sugisaki and L.D. Jacobson. 2005. Environmental effects on recruitment and productivity of Japanese sardine Sardinops melanostictus and chum mackerel Scomber japonicus with recommendations for management. Fish. Oceanog. 14:263-278.
ThisPage pageis intentionally This Intentionally left Leftblank Blank
Large Marine Ecosystems, Vol. 14 V. Shannon, G. Hempel, P. Malanotte-Rizzoli, C. Moloney and J. Woods (Editors) © 2006 Elsevier B.V. All rights reserved.
349
15 Environmental Data Requirements of Maritime Operations in the Benguela Coastal Ocean M.L. Grundlingh, P.D. Morant, R.C. van Ballegooyen, A. Badenhorst, E. Gomes, L. Greyling, J. Guddal, I.T. Hunter, D.W. Japp, L. Maartens, K.R. Peard, G.G. Smith, C.K. Wainman
INTRODUCTION The discoverers of the Cape route in the fifteenth century, or even the intrepid Phoenicians who may have circumnavigated the tip of Africa almost two millennia earlier, were the first to become aware of the often adverse maritime conditions off the south and west coast of southern Africa. These would have been manifested by a combination of storms (wave intensity progressively getting worse on the southward journey), winds (reaching gale force in the vicinity of the Cape of Good Hope), fog (the origin of many wrecks on the Skeleton Coast) and currents in the AgulhasBenguela mixing area (c.f. Rennell 1832). While initial scientific surveys of the Benguela region were all exploratory by nature, and operated from a relatively low information base, they started unveiling the existence of the slow northwest-flowing Benguela Current, upwelling, poleward undercurrent, Agulhas Current, Agulhas rings and their advection, to name but a few (Shannon 1985). Superimposed on these are geochemical features (Chapman and Shannon 1985), and an extensive range of biological processes (e.g. Payne et al. 1987). Local studies in the 20th century were largely focussed on the biology of the region (Lutjeharms and Shannon 1997), but these gradually transformed to include sensitivity to, and understanding of, the interaction between maritime activities and the environment (e.g. when oil and gas exploration moved offshore in the second half of that century). The relevant industries have since then become recognised stakeholders in the sustainable use of the Benguela ecosystem. Over the past 15 years or so, this process has been further modified to include aspects of forecasting, and this was supported by advances in numerical modelling as well as by the wealth of information generated through large marine programmes such as the Benguela Ecology Programme (BEP: Shannon et al. 1988), the Benguela Environment Fisheries Interaction and Training (BENEFIT: www.benefit.org.na), BEST (Benguela Source and Transport; Garzoli et al. 1996) and Benguela Current Large Marine Ecosystem (BCLME).
350
M.L. Grundlingh et al.
Focussing on interactions between maritime operations and the environment, the aim of this chapter is to identify the (forecast) environmental information needs of the southern African maritime industry. The inclusion of a chapter on maritime issues in a presentation dealing mainly with biogeochemical aspects of the Benguela is quite novel and the opportunity is therefore taken to introduce readers to some of the characteristics of the main maritime industries of the region. Forecasting is understood here to be limited to operational requirements on the time scale of hours to days, and which can benefit from operational weather forecasting. Items considered to fall outside the limited scope of this chapter are therefore longerterm prediction of stock assessment, environmental predictions for design purposes (e.g. 1-in-50 year wave conditions), prediction of long-term morphological and bathymetric evolution, etc. OIL AND GAS INDUSTRY ENVIRONMENTAL INFORMATION NEEDS In terms of daily commercial turnover, the oil industry is probably the single largest industry in the Benguela, producing revenue of about US$50 million per day. The largest oil and gas exploration and production activities in the Benguela region are located off Angola, where those activities account for the largest percentage (between 77 % and 89 %) of the Angolan economy. Oil production has increased dramatically since 1974 from a maximum of 173 000 barrels a day to a present peak of 1 100 000 barrels a day, with a planned increase to more than 1 500 000 barrels a day by 2006. Angola has three coastal geological Basins: the Lower Congo Basin, the Kuanza Basin and the Namibe Basin (Figure 15-1). Although petroleum was discovered in Angola in the 18th century, the first commercial field in the offshore domain was only discovered in 1966 in the shallow waters of Lower Congo Basin. After 1978 the offshore area was divided into shallow blocks number 0-13 (less than 200m), deep offshore blocks number 14-30 (200m to 1500m water depth), and ultra deep offshore blocks number 31-34 (see Figure 15-1). In 1996 the giant fields Girassol and Dalia (which have estimated reserves of 800 million to 1 billion barrels of oil) were discovered in the deep waters of the Lower Congo Basin in Block 17. Other very important discoveries have been made in Blocks 15 (same magnitude as Block 17), 14 and 18. The drilling of appraisal wells in the deep waters of Kuanza Basin started in 2001, while the potential of Namibe Basin has not yet been tested. In contrast to Angola, Namibia and South Africa are minor producers of offshore oil and gas. The only commercial discovery in Namibia, the Kudu gas field off the extreme southern coast, is currently being brought into production to provide fuel for a major electricity generating station at Oranjemund. In South Africa both oil and gas in commercial quantities have been discovered and brought into production on the eastern Agulhas Bank, and the Ibhubesi gas discovery on the Cape west coast is being appraised to determine its commercial viability.
Environmental data requirements of maritime operations in the BCLME
351
The two main activities undertaken by the offshore oil and gas industry are exploration and, in the event of a commercially viable discovery, production. Exploration comprises geophysical (seismic) surveys and exploration drilling, and production consists of appraisal and production drilling, installation of oil and gas production infrastructure, and the operation of product export facilities. Oil and Gas Industry - Geophysical surveying Marine geophysical surveys use high energy, low frequency sound sources towed behind a ship. The noise signal from the source array is directed downwards into the earth, and the reflected signals provide insight into the various geological formations (McCauley 1994). Oil and Gas Survey—Campaign planning Historic metocean data, such as sea state and the ocean current regime, are used to optimise the selection of the seismic survey vessel with the conditions likely to be encountered in the survey area (Table 15-1 on p.353 here). The data are used to estimate the duration of the survey campaign and to make allowance for time lost as a result of unfavourable conditions. Oil and Gas Survey—Seismic data quality Seismic survey vessels usually acquire water column current profile (ADCP) and CTD data while conducting the survey, and these data are used to account for the effect of the density stratification in the received signal. Oil and Gas Survey—Other operational considerations Apart from the possible interference between geophysical surveying vessels towing kilometres-long arrays of sensors and fishing vessels dragging similarly long lengths of gear, potential acoustic impacts (biological and physiological) of geophysical survey operations are of concern (e.g. McCauley 1994; McCauley et al. 2000a and 2000b, Turnpenny and Nedwell 1994). This holds particularly for baleen whales since they are considered to have very good low frequency hearing which overlaps the output range of seismic sources. The main mitigation measure is to avoid seismic surveying when the whales are likely to be present. Oil and Gas Industry - Exploration, appraisal and production drilling The same process is used for exploration, appraisal and production drilling, the only difference being the number of wells drilled. Should there be evidence of commercial quantities of hydrocarbons being present, appraisal drilling is undertaken to establish the viability of the find. Production drilling consists of further wells in order to optimise the production of the oil or gas. Data requirements here relate to:
352
M.L. Grundlingh et al.
Oil and Gas Exploration—Campaign planning As is the case for geophysical surveying historic metocean data are used for the selection of the drilling unit and estimation of campaign duration. In recent years exploration drilling on the continental slope in depths of 500 m to 2000 m has become commonplace, particularly off northern Angola. Surface current, sea state and wind data are used to estimate the amount of fuel required to keep the drilling unit on station, while the current profile of the entire water column is used to determine the forces acting on the drill string and casing which is unsupported between the drilling unit and the seafloor (Table 15-1). Oil and Gas Exploration—Impact assessment The only potentially significant environmental impact of drilling a trouble-free well is the discharge of drilling muds and cuttings. Cuttings are discharged overboard throughout the drilling operation whereas the surplus mud is usually discharged only on completion of the well. The water column current profile is a key input in the determination of the “footprint” on the seafloor of these discharges. Risks associated with potential oil spills, to date have been addressed mainly by oil spill modelling using a surface origin for the spill. However, as oil exploration has progressed into greater water depths (1000-2000 m) a surface release point for modelling a deepwater blow-out has become unrealistic (Bishnoi and Mainik 1979). Modelling a deepwater release requires data on the density of the spilled oil, and the current, salinity and temperature profile of the water column (Table 15-1). Oil and Gas Industry - Production Production involves the installation of the infrastructure which permits the controlled release of the oil or gas from the subsea reservoir and its transport to a loading facility or refinery. The infrastructure would typically consist of: a production facility which, depending upon the water depth, could either be floating (moored) or standing on the seabed; and undersea pipelines to a tanker loading facility or to a refinery ashore. Environmental data requirements are related to: Oil and Gas Industry Production—Engineering design and infrastructure installation Metocean data are required for engineering design purposes and for planning infrastructure installation e.g. laying pipelines, installing production platforms, etc. Typically the structures are designed to withstand the most severe oceanic conditions likely to be encountered during the life of the oil or gas field. Oil and Gas Industry Production— Impact Assessment The main impacts of production activities on the environment are their effects on the ecology (caused by the physical disturbance, noise, and operational discharges and waste, accidental oil spills) and the atmosphere (emissions of the gas turbines, gas flaring, and from the diesel engines).
Environmental data requirements of maritime operations in the BCLME
353
NAMIBIA
Figure 15-1. Layout of the Angolan oil fields and the associated subdivision into Blocks 0–34. The three basins, Bacua do Congo, Bacua do Kwanz and Bacua do Namibe are indicated. Used here with permission.
354
M.L. Grundlingh et al.
The most important issue under “operational discharges” is the impact of produced water which is usually the largest single aqueous discharge from offshore production platforms and reaches volumes of up to 20 000 m3 per day (Sakhalin-1 1994). The main source of produced water, sometimes known as produced formation water, is fossil water trapped within the oil-bearing rock formation which comes to the surface during oil production. Produced water may contain chemicals which could have an adverse effect on marine life, particularly larval stages. Metocean data are required for modelling the trajectory and fate of produced formation water discharged from a production platform (Table 15-1). Information on potential receptor organisms and activities e.g. plankton, fishes, spawning, etc. is required in order to assess the potential impact of the discharge of produced formation water. Currently the prediction of the effects of operational and accidental discharges tends to focus on physical impacts while biogeochemical effects of some of the mud constituents adsorbed onto clay and other fine particles have been given little attention. In terms of atmospheric emissions improper combustion from the flare stack releases amounts of hazardous chemicals (toxins including carcinogens), metals (mercury, arsenic and chromium), nitrogen oxides and sour gas (with hydrogen sulfide and sulfur dioxide). At a global level it is important to mention the emission of carbon dioxide and volatile organic compounds (VOCs). Oil and Gas Industry - Legislative aspects The petroleum sector (globally and in the Benguela region) is very advanced in terms of environmental protection, pursues the use of modern and less polluting technologies, materials and equipment, and has adopted advanced laws and standards. Most of the legislation applied to the sector is based on the experience of other countries, or accords with the minimum acceptable standards proposed by international conventions and practices. DIAMOND MINING Africa is the world’s largest producer of diamonds, accounting for about 50% of global production. Offshore diamond mining in southern Africa is mainly located off Namibia (see Figure15-2) where it constitutes about 45% of all diamond-mining operations in that country. As coastal on-land diamond sources are becoming depleted, mining operations are migrating seaward. The remaining on-land diamond reserves will be mined as close to the high water mark as can be safely achieved. The feasibility of mining on a large scale in the inter-tidal, surf and the near-shore zones, is under investigation. In addition to these activities, deep-water mining by means of drills and remote vehicles continues further offshore.
Environmental data requirements of maritime operations in the BCLME
355
Figure 15-2. Namdeb Diamond Corporation’s concessions off southern Namibia. De Beers Marine Namibia mines the Atlantic 1 Mining License (ML) on exclusive contract to Namdeb Diamond Corporation.
Coastal diamond mining Coastal mining operations include (a) construction of massive sea walls of sand and/or rock on the shoreline and even in the surf zone to protect mining operations on their landward side, and (b) dredging, whereby large volumes of overburden sediment is removed by means of a dredger floating in a coastal pond. Such operations are susceptible to waves overtopping the beach and possible consequent flooding of areas being mined. Erosion by storm waves could result in breaching of protective sand (beach and seawall) barriers and possible consequent direct wave attack on mining equipment. Forecasts of erosive wave and water-level conditions will allow vulnerable seawalls to be timeously bolstered with sand or rock by means of earthmoving vehicles. The type of forecast data that would benefit coastal mining includes wave-heights, water levels and volumes of erosion of the beach and/or seawall (Table 15-1). At
356
M.L. Grundlingh et al.
present, a warning system for mining operations in southern Namibia employs forecasts of wave conditions from global models, as well as advance warning of approaching storms based on the delay (for most wave directions) in arrival of storm waves between Cape Point (where real-time wave data are collected) and southern Namibia. An example of prediction of long-term environmental change that could aid in planning of coastal mining is the prediction of shoreline change over a period of several years. This has been demonstrated to be predictable by means of morphological models (Smith et al. 2002). Nearshore diamond mining Mining companies are investigating the feasibility of large-scale mining operations in this region, employing self-elevating “walking” platforms. It is intended that such platforms will be capable of mining the intertidal and surf zones, as well as the relatively shallow water regions seaward of the surf zone. Given adequate forecasts, operators of self-elevating walking platforms will be able to take evasive action; anticipate shut-down action; or anticipate start-up action as a storm subsides. The primary types of data that would be required are total water-level (resulting from tides, waves, wave set-up, wind set-up, atmospheric pressure and infragravity wave action) across the nearshore region; wave heights and periods in the same area; and surf zone currents. Offshore diamond mining Present mining operations in deep water involve the deployment of large diameter drills combined with airlift pumps and the deployment of remote-controlled crawler vehicles (see Figure 15-3). Future deep-water mining may involve the deployment of large dredgers, which can dredge sediments to depths of 130 m. The feasibility of dredging material offshore and pumping it ashore via a pipeline is under investigation. Deep-water mining operations involve vessels that are moored by means of a 4-anchor system over an area of about 1 km2 at a time. During adverse swell conditions, only a restricted area of this square kilometre can be accessed, while during very severe storms, shut-down may be necessary. Accurate forecasts can therefore provide decision support for short-term planning (optimising the mining operations); for shutdown warning due to approaching storms (maximising operations without compromising safety); and for start-up operations after a storm has passed. Deep-water mining, using Trailing Suction Hopper Dredgers (TSHDs) may involve overburden removal and dumping, as well as the actual mining of material. The mined material is to be pumped to shore via a pipeline and accurate forecasts will allow optimisation and safety of operations.
± 100 - 120m
Environmental data requirements of maritime operations in the BCLME
357
Drill String
Drill Bit
Crawler
Figure 15-3. Diamond mining technologies for deep-water operations (~ 130 m). Image: De Beers Marine Namibia, used here with permission)
The primary data required for all of the above activities are: swell wave heights, periods, directions; sea wave heights, periods, directions; and wind speed and direction. All of these can be provided by means of large scale forecast models, the Cape Point wave prediction system as mentioned above, and computational models to translate the offshore forecast data to the nearshore region. SHIPPING In terms of the amount of maritime traffic carried along international trade routes, the Cape route is not seen as a “choke point” in the same sense as the Panama Canal, Suez Canal or the Strait of Hormuz. Nevertheless, the number of vessels and amount of cargo transported around the Cape of Good Hope and by implication through the Benguela area is quite significant. Approximately 1000 bulk carriers, 1000 cargo vessels, 400 tankers and 1000 container vessels, as well as a number of other vessels pass along the route (eastbound and
358
M.L. Grundlingh et al.
westbound combined) per year (State of the Environment 2001). The value of each of these vessels, plus their cargo, is roughly estimated at US$ 200 million, resulting in a total “investment” of about US$ 700 billion. In the event of an oil spill, or some other maritime incident, the costs per incident can be up to $50 million, in addition to the loss or damage of the vessel and its cargo, and such events seem to occur at a rough frequency of one every 3 years. The loss of life or damage to the environment has not been included in the above estimate. On a global scale transportation accounts for about 24% of oil pollution inputs into the marine environment (Etkin et al. 1998), although tanker accidents contribute only 6 percentage points to this (the rest by operational discharges, bilge water, also inside ports). A considerable implicit and explicit stake is nevertheless held by the local countries as well as the trading partners in other parts of the globe in the shipping route traversing the Benguela. This section focuses on environmental issues associated with vessels at sea, while the Ports section below deals largely with the environmental issues related to ports. Impact of environment on the shipping industry The effect of adverse weather on a vessel can vary from simple delay (in loading or offloading, leaving or entering port, receiving supplies via helicopter, and reduction of steaming speed), to structural damage (to the engines or hull), loss of contents (cargo damage, container loss), or complete loss of the vessel and its contents including environmental damage (e.g. from oil spills). The marine variables involved here include waves, wind, currents and tides (water level), and a forecast 1-5 days ahead is essential (Table 15-1). International weather service responsibilities to the shipping industry The Global Maritime Distress Services System (GMDSS) in cooperation with the World Meteorological Organization and the International Maritime Organization, provides a harmonised, background framework for marine weather forecasting. The baseline for this system is a 1–3 days ahead, general weather forecast, including severe weather warnings. Lead national meteorological services have taken regional responsibilities within this system, and the South African Weather Service (SAWS) is the lead forecasting agency for the ocean adjacent to southern Africa (METAREA VII, see Figure 15-4). Internationally there are several new developments underway, such as: • • • •
Extension of forecast period to 10 days Inclusion of routing advice Enhanced inclusion of general sea state and sea ice information Specific services in support of oil spill mitigation
Environmental data requirements of maritime operations in the BCLME
359
• More modern modes of information dissemination, including adding forecast information on top of electronic sea charts, web based services, and radio broadcasts with ‘interruptive warnings’. • Enhanced coastal navigation services, including real-time coastal current information, especially in narrow approach channels • Better information on, and warnings of, extreme and damaging waves • The global services in support of oil spill mitigation include both the share of responsibility areas as well as the exchange of tools such as numerical oil spill models.
Figure 15-4. Ocean forecasting areas and country responsibilities
Increased emphasis on operational oceanography has led to regional models capable of forecasting short-term ocean circulation, sea temperature variations, salinity and other parameters. Extensive use is also made of satellite information (e.g. QuickScat), although these are not operational yet in the Benguela region. Local weather service responsibilities to the shipping industry The responsibility of the SAWS to provide metocean forecasts for METAREA VII (Figure 15-4) has been guaranteed to international shipping as a result of South Africa being a signatory to the SOLAS (Safety of Life at Sea) Convention. SAWS is also responsible for providing support services in the case of a spill or other maritime emergencies. Such services are disseminated via radio and fax.
360
M.L. Grundlingh et al.
Finally, investigations of extreme and damaging waves have shown that these phenomena require more research and also greater attention to operational safety. Some ocean areas are notorious for such waves, and there are discussions among marine designers about how to account for these events while still in the design and planning stage for new vessels. PORTS The three countries bordering the Benguela region have ports that operate against the background of rapidly expanding trade resulting from economic growth and globalisation. The South African ports are the largest in the region in terms of vessel traffic, volume of freight handled and facilities. This section is confined to activities within the port limits, while environmental issues of vessels on high seas have been addressed in the Shipping section of this paper. Port operational need for real-time environmental and biogeographical data First, the ports’ requirements for real-time environmental information are intended to minimise the risk to ships entering/leaving the port and while moored inside the port, and to enhance the safety and speed of loading and other operations. The environmental data required for this are real-time wave, wind, tides and currents. Present real-time services at major South African ports include in situ wave, wind and tide measurements, with forecast information on waves and tides also included. Where particular services do not yet exist in the BCLME region, there is no reason why the existing network could not be expanded to other ports as well as to specific locations of interest to other maritime industries (e.g. mining and energy). Environmental information (waves, tides) is also required in analyses of stability of breakwaters and other structures, as well as in the port-associated impact on sediment movement and beach erosion. Second, information is also required to support the biogeochemical diversity of ports. In most countries commercial ports are situated in river mouths and estuaries at the land-sea interface. It is generally recognised that these sensitive natural systems provide critical ‘free’ environmental goods and services, in the form of nutrient recycling, biodiversity protection and erosion control, to the surrounding urban and semi urban areas. Ports also play a role in the global effort to maintain biodiversity, either through rehabilitation of negatively impacted areas or stringent environmental controls and management procedures of resources available to us now. The following is a list of environmental information needs and their applications: • Dredging is required to maintain necessary draught/depth for the safe berthing of ships. At the same time, research needs to be conducted into economically viable alternatives for dumping dredge spoil at sea.
Environmental data requirements of maritime operations in the BCLME
361
• There are environmental concerns relating to ship repair and scrapping activities and general dry dock management, and environmental impacts, development of appropriate mitigation measures/standards and policy formulation should be researched. • Environmental information is required to manage waste arising from storm water run-off (within port boundaries and from surrounding municipalities/areas); sewage discharges into harbours; and shipping (galley, illegal dumping). • The quality of fresh water (surface), marine water and sediment (dredging) needs to be monitored. Strategic Environmental Assessments (SEAs) need to be conducted to guide planning, development and monitoring requirements. • Environmental information, in the form of real-time data on waves, wind and currents, is needed for emergency responses and preparedness to avoid pollution due to oil spills and hazardous waste. • Environmental information is required for management of biodiversity; the characteristics of sensitive areas in/close to ports; for rehabilitation for already impacted areas and management of marine alien species (incoming vessels).
FISHING The South African fishing industry is largely located in the Benguela Current Large Marine Ecosystem, and tries to ensure sustainability of the industry through, inter alia, better understanding of the marine environment. Although the commercial fisheries have been active since about 1900, a more recent trend has been the development of high seas fishing as local and international operators seek alternative resources using new technologies. The present section deals largely with the activities of fishermen, rather than those of resource managers which should implicitly be covered elsewhere in this book. Fishing Area and technology The broad operational area of the fishing within the BCLME area is shown in Figure 15-5. The domestic commercial fisheries are regulated by the particular state, while the “High Seas” area beyond the economic fishing zone require flag vessels to obtain flag state permission to fish in the area. Many foreign flag vessels also fish in this area and land catches in local ports when granted permits to do so. High Seas fishing within the BCLME boundaries falls under the management of two Regional Fisheries Management Organisations (RFMO’s)—these being the South East Atlantic Fishing Organisation (SEAFO) and the International Commission for the Conservation of Atlantic Tunas (ICCAT) (Figure 15-5). In the Southern Ocean, the
362
M.L. Grundlingh et al.
Commission for the Conservation of Antarctic Marine Living Resources (CCAMLR) is the responsible management authority, but this is located beyond the BCLME boundaries. Management approach to fisheries The 2001 Reykjavik Declaration on Responsible Fisheries in the marine ecosystem directly and specifically addressed the issue of introducing more ecosystem considerations into conventional fisheries management. The South African fishing industry has recognized the need for the conservation and long-term sustainable utilization of the resources they exploit. An initiative in this regard was the certification of the South African hake trawl fishery by the Marine Stewardship Council as a sustainably managed fishery. This certification embraces three principles, namely Management, Resource assessment and Ecosystem effects. Conditions of certification include bycatch considerations, effects on seabirds and marine mammals.
S 0
0
CONGO
5
0
Cabinda
Seychelles
TANZANIA
Congo River
100
ANGOLA UE
Comores
0
250
ICCAT
0
SOUTH
30
ATLANTIC OCEAN
Mauritius Reunion
MA DA
M
GA
OZ AM
NAMIBIA
0
20
SC A
BI
R
Q
15
Maputo INDIAN OCEAN
Durban AFRICA
0
35
200 nm Fishing zones 0
40
SEAFO
450 0
50
CCAMLR 0
30
0
25
0
20
0 150 10
5
0
00
5
0
0
10
0
15
0
20
0
25
0
30
0
0 35 40
0
45
500
550 E
Figure 15- 5. Operational areas (Fishing) and Regional Fisheries Management Organisations within the BCLME boundaries
Environmental data requirements of maritime operations in the BCLME
363
Fishing operations Purse seine (small pelagics) The largest fishing sector with respect to volumes landed is the purse seine fishery targeting small pelagics such as anchovy and Pilchard. The fleet comprises about 100 purse seiners that can be from 15-30 m in size and are limited to about 50 km operations from the shore. The annual catch varies considerably from 200 000-600 000 tons (horse mackerel and redeye pilchard combined). Hake trawling This fishery uses large trawlers, catching about 140 000 tons of hake annually with 7080 vessels up to 70 m in length. A similar fishery for hake is found in the northern part of the BCLME area in Namibian waters (about 120 boats) and exploits similar stocks. In the south, on the Agulhas Bank, a small inshore trawl fishery targets hake and sole. These vessels operate mostly within the 200 m depth zone and number about 30 boats not exceeding 30 m in length (Fig. 6). Demersal longliners In South Africa there are about 150 hake longline vessels ranging in size from 15 – 35 m that shoot from 4000-20 000 hooks per day and catch approximately 10 % of the hake total allowable catch. The hake longline fleet in Namibia comprises vessels of similar operational characteristics, but are fewer in number. Operations are confined to the continental shelf. Hake handline A substantial handline fleet directs effort on hake mostly in the Mossel Bay to Port Elizabeth area, operating on a daily basis and is restricted to close inshore (<80 m water depth). There are about 100–150 hake handline vessels mostly of the open deck ski-boat type from 5-15 m in length. Pelagic fishery The South African pelagic fishery is based upon the three forage species namely sardine, anchovy and round herring. Two of these are limited by total allowable catches, allocated as rights to some 114 rights holders. Other commercial fishing sectors A small directed trap fishery for South Coast Rock Lobster operates on the southern and eastern limits of the Agulhas Bank i.e. the eastern extremity of the BCLME area (8 boats). A second trap fishery for West Coast Rock Lobster operates on the West Coast close inshore out to about 60 m water depth (300 boats). A significant pole and line fishery mostly for albacore tuna is based in Cape Town and Hout Bay on the West Coast. These vessels range in size up to 30 m and many are used in other fishery sectors, e.g. hake longline.
364
M.L. Grundlingh et al.
Demersal Trawling
Trawl warps (Steelwire rope) Up to 550 m (Deepwater Trawling up to 1000 m)
Demersal Trawl Net
Doors (2000 kg) Spread 100m+
Figure 15-6. Basic gear configuration of hake-directed trawlers in the BCLME area (Figure by A.P. van Dalsen of Marine and Coastal Management, used here with permission).
Examples of environmental data requirements Offshore fishing Despite much potential promise, remote sensing of environmental predictors of fish availability are not utilised by the pelagic fishing industry to optimise fishing. A recent study showed a severe temporal and spatial mismatch between areas of abundance and concentration of fishing effort (Agenbag et al. 2002). It also investigated the correlation between fish catches and environmental parameters (e.g. sea surface temperature, as derived from satellite imagery), and concluded that there are no scientifically proven relationships between sea surface temperature and pelagic fish distributions. This, however, does not exclude the usefulness of other environmental parameters. A more useful indicator than sea surface temperature would be upwelling intensity and the use of satellite-derived ocean colour images to provide indications of phytoplankton abundance. The stakeholder survey, mentioned in the Maritime Forecasting section below, revealed that knowledge of distinct subsurface temperatures could be beneficial to some components of the fishing industry. Inshore fishing An example of inshore fishing is the lobster industry. The environmental parameters that influence the biology and population dynamics of the species are seasonality in bottom dissolved oxygen levels, low oxygen due to hydrogen sulphide emission from
Environmental data requirements of maritime operations in the BCLME
365
Table 15-1. Summarised environmental information requirements of the maritime industry operations
Parameter
Currents, weather, waves (or selection thereof)
Waves, water levels, tides
Seawater density, air temperature Biological characteristics Atmospheric characteristics Wind, Air temperature; Relative humidity; Cloud cover; Visibility Stormwater discharges; water quality; sediment quality Ballast water information Waves, oxygen, subsurface temperature, upwelling index, colour Seawater density; Fronts; Mixing and eddies; Visibility; Internal waves; Water quality Depth; Bottom texture; absorption/reflectivity Sand grain size and compaction, underwater noise and scattering layers; Bio fouling; Species distribution and toxicity; Underwater visibility; Bio luminescence Depth; Tidal streams; Changes in coastline properties
A*
Data mode C* B*
Industry
Application
3
2
1
Oil & gas
3 3 2
3 1 1
1 1 1
Diamond Naval Shipping
3
3
2
Ports
3
3
2
Diamond
3
3
3
Ports
3
2
1
Oil & gas
1
1
1
Oil & gas
Survey planning, exploration, design of operational structures Operational decision support Sea-keeping ability, diving Operations, safety Ship operations, emergency response Erosion of beach wall, platforms Navigation, cargo transfer Oil spill modelling, produced water Receptor organisms
2
1
1
Oil & gas
Impacts of flaring
3
2
1
Naval
Infrared, radar, laser performance Communications
2
2
1
Ports
Pollution monitoring
2
2
1
Ports
Alien species management
2
1
1
Fishing
Operations
3
1
1
Naval
Sonar, Submarine operations; Infrared sensor; Diving
1
1
1
Naval
Navigation; Sonar; Sea mines, diving
2
1
1
Naval
Navigation, Amphibious operations
* A Historic; B Real-time; C Forecast * 3 = data available; 2 = data only moderately available; 1 = data not available, or status unknown
the sediment, or accumulation thereof in the water column due to phytoplankton bloom decay. The lobster vessels set traps and nets in the shallow inshore reef environment and catches are also affected by surface sea conditions such as wind and swell.
366
M.L. Grundlingh et al.
Instrumentation that is currently used in South Africa to monitor oxygen and temperature through the water column should input into a forecasting system for low oxygen events and ultimately to the mobilisation of contingency measures in the event of lobster walkouts. In Namibia, satellite true colour images in real time allow prediction of low oxygen events due to hydrogen sulphide eruptions from the sediment, and a moored oxygen logger allows bottom oxygen levels to be related to catches. It therefore seems that the environmental data needs of the fishing industry (in terms of data type and availability) are not yet catered for adequately, and that research is also required to identify the right parameters and how to obtain them.
SOVEREIGNTY AND RESOURCE PROTECTION Strategic issues Navies are traditionally not classified as “maritime industries.” However, their presence in the maritime domain and (especially) their collection of and need for environmental information is undisputed. It is therefore appropriate to discuss their environmental needs as important stakeholders in the Benguela region. Southern Africa is experiencing considerable change in terms of the improvement in regional collaboration as well as with countries further abroad. Exchange of information, combined exercises and interoperable communications has become key concepts to improve regional support and mutual confidence. Additional to national safety and defence mandates there are important political and economical responsibilities, such as protection of submarine cables, fishing and mineral resources, maritime traffic, pollution and scientific research. In addition, the hydrographic offices are responsible for surveying the ocean topography within their areas of jurisdiction. The enormous monetary value of protectable resources (fishing, oil and gas) and infrastructure (shipping) thus places a huge burden on the limited facilities of the naval forces of Angola, Namibia and South Africa. Operational issues Although many design, planning, tactical, operations and safety information requirements have parallels with other maritime activities, the scope, scale and diversity is significantly greater for a navy. This may range from weather, surface characteristics, subsurface stratification, to sub-seafloor characteristics for harbour security. This is summarised in Table 15-1, where it is evident that naval environmental requirements are the most extensive of all the maritime role players.
Environmental data requirements of maritime operations in the BCLME
367
Individual regional naval operational roles may differ but may be broadly defined as: protection of marine resources, combating maritime crime, monitoring of shipping, search and rescue, hydrographic services and sharing of expertise, disaster relief assistance, interoperability, building inter-regional marine forces For a modern navy to achieve its mission, all possible variables need to be considered and this includes environmental variability. With equal access to weapons, machines, ammunition and training, a navy’s only real advantage in the battle space may be a superior knowledge of the environment. As above and below water combat technologies improve, so sensors, guidance systems, stealth technology and operators demand ever increasing data about the environment in which they are expected to operate. Aganist these strategic demands for environmental information it is recognised that some of the largest waves, strongest currents, and most extreme sea temperature differences exist in the three oceans (Atlantic, Indian and Southern) that southern African navies are expected to operate in. Mobilising naval resources in a region of such maritime environmental extremes and contrasts over vast areas can be significantly improved by the provision of appropriate, timely and cost effective information solutions. Naval environmental data needs and method of delivery The environmental data under consideration includes currents, weather, waves, tides seawater density patterns (for sound propagation, anti-submarine warfare and mine detection), and presence of internal waves (for submarine operation). For technologyempowered navies and decision makers this information needs to be available at a mouse-click, even when at sea. In contrast, the extremely limited technical capacity of “coastguard-type’’ navies with financially constrained options have their own unique environmental data requirements dictated by their limited area of influence. At this end of the spectrum, even a basic need for maritime charts, with accompanying tide, weather and current information to provide for safe navigation, may suffice. Although large volumes of marine data have been collected in southern Africa over the past century, much of the data lacks the added value of graphic enhancement. In a quest to pool, enhance, and apply available information to naval advantage a spatial basis using GIS (Geographic Information Systems) technology was implemented for the South African Navy. This system allows a diversity of data types, standards and volumes to be centrally stored and accessed, with long-term continuity. On an operational basis, specific historic information can be provided on a GIS platform using commercial off the shelf software for an area of interest. Near real-time modelled or monitored output information in the form of text (e.g. weather or sea state reports), maps (e.g. synoptic wind and swell maps) or images (e.g. port or estuary approach) can be provided as spatial geo-referenced compressed files for unpacking on
368
M.L. Grundlingh et al.
board a ship at sea. The benefit of this utility is its remote update operation on a common digital backdrop map of historic information (e.g. coastlines, depth contours, vessel location etc) for easy comparison and interpretation by the user. MARITIME FORECASTING IN SUPPORT OF RISK MANAGEMENT It is clear that the current investment in and return from maritime activities in the Benguela is large and growing rapidly, particularly in the energy and mining sectors due to growing energy needs in the region, increased investment in the exploitation of oil and gas reserves and major technological advances in the exploitation of marine resources, e.g. new diamond mining techniques. Improved assessment and management of risk in both the coastal and offshore zones of the Benguela region, is obviously called for. Most maritime industries of the region have indicated a need for improved environmental information to ensure operational efficiency/effectiveness and to manage risks that may originate from: • The impact of the marine environment (i.e. metocean conditions) on maritime operations: In the maritime transport industry the risks (vessel damage or loss, loss of human life, loss of cargo, need for salvage or assistance, delays, etc) may be substantial as evidenced by a vigorous salvage industry in the region. The same holds for most other industries (e.g. the high annual investment to keep ports operational by dredging unwanted sediment; the environmental damage to breakwaters; sensitivity of diamond mining to environmental impacts, etc). These risks may be reduced by establishing more accurate, comprehensive and accessible maritime forecasts of metocean conditions, currents, and waves (see also the Shipping section, here). • The impact of maritime operations on the environment: This comprises risks to the marine environment associated with shipping casualties (e.g. oil spills, toxic cargoes, etc) or failed maritime operations (e.g. deep-sea blowouts, spills during refuelling operations, etc) where the probability of the risk event occurring and the associated potential environmental impacts can often be minimised (or even avoided) by appropriate risk management based on real-time or forecast metocean conditions. Longer term impacts of maritime activities on the environment due to oil exploration, the impact of ports on the environment (interruption of sediment movement, effluent and pollution, introduction of alien species, etc) also pose potentially huge risks and could, e.g., result in similar litigation as the more dramatic and more traditional claims associated with oil spills or maritime incidents. In contrast to the risks involved with the environmental impact on maritime operations, the converse risks are mostly not moderated through the type of metocean forecasting mentioned in this chapter but rather by a better understanding and longer term predictions of various processes in the marine environment as described in a number of other chapters in this book.
Environmental data requirements of maritime operations in the BCLME
369
Risk management in the marine environment of the BCLME The availability of appropriate environmental information to ensure adequate risk management and operational efficiency/effectiveness will make the Benguela an area of extremely attractive financial investment, especially in the energy, mining and fishing sectors where the potential returns can be high. Existing support in managing operational risks in maritime operations range from relatively well-developed, more generalised prediction services based on global forecast of winds and waves to smaller scale, highly customised forecast services based more on local measurements and experience. Real-time Services Real-time services, where they exist in the region, are focussed on the ports and specific offshore operations (e.g. oil rigs) (see Ports section). Expansion of these services to provide increased spatial coverage is presently under development. A system of “virtual wave buoys” is presently under development in Table Bay and will combine wave measurements with localised model simulations of wave conditions to provide enhanced spatial coverage. This initiative can easily be extended to other activity hubs and in future may include real-time current observations and predictions. Nowcast/forecast Services Existing efforts to provide nowcast/forecast systems include a trial WEB-based forecast system for oil spills, navigational safety and search and rescue operations for Saldanha Bay (http://coastalmodels.csir.co.za). In addition a number of ad hoc forecast services have been attempted, most notably those provided during the Treasure oil spill incident in Table Bay. These endeavours highlight the need to develop a 24-hour operational forecast system for high risk regions to ensure effective operational support. Based on this indicated need and a rapid improvement in the global modelling and real-time monitoring system capabilities, a forecasting system comprising real-time measurements, operational models and appropriate and readily accessible information dissemination and decision support systems has been designed to provide accurate forecast data in high activity regions or hubs, e.g. Table Bay. Stakeholders expected to benefit from such an enhanced maritime forecast of metocean conditions around these high activity centres are the shipping companies, insurers, national ports authorities, regulatory agencies ensuring maritime safety, sea rescue services, the salvage industry, offshore mining and mineral exploration, and construction companies. Further afield (i.e. at the offshore location of rigs, offshore mining and fishing operation, etc.) typically more specialised forecasts are required both to improve safety and the efficiency of their operations. Other important stakeholders are the national navies and agencies tasked with maritime policing and protection of natural resources, all of which need real-time or forecast metocean information to support their operations.
370
M.L. Grundlingh et al.
This need was confirmed when representatives of most of these industries in South Africa were surveyed by the CSIR in 2002, and expressed themselves in favour of a real-time and forecast/predictive capability that is accurate, reports on weather, waves, currents (with other more specialised requirements such as thermocline prediction and water quality), is updated hourly to once a day, is highly customisable and simple (as it is typically core to their operations but not part of their core functions), that provides simple integrated decision support (i.e. combine a number of data streams at appropriate time and space scales) that is robust, easily accessible and has a rapid and reliable response time. Based on this stakeholder survey and a technology survey focussed on the technological and commercial feasibility of a forecast facility that will meet these requirements, a HUB mode operational modelling capability (i.e. modelling capability providing information in and around high activity hubs) has been initiated as a pilot project in Table Bay. In the proposed numerical modelling system communication is primarily via the internet but with a strong emphasis on cell-phone technology when delivering to the user. This provides the flexibility to accommodate the often widely varying infrastructure and capabilities of the user groups in developing and particularly African countries. SUMMARY AND CONCLUSION The environmental information needs of the southern African maritime industry is presented in a concise fashion in Table 15-1, where a differentiation has been made between historic, real-time and forecast data, and also whether these data are available. In terms of these three categories it seems that the management of historic data is reasonably well in hand: • • •
The South African Data Centre for Oceanography contains large amounts of surface and subsurface data, including temperature, salinity, nutrients, oxygen, surface waves, weather, etc. (sadco.csir.co.za). Data are also collected and stored by the large marine mineral industrial organisations (diamonds, oil and gas). Research information is also collected and stored by the University of Cape Town, the Institute for Maritime Technology, National Marine Information and Research Centre (NATMIRC), Namibia and Marine and Coastal Management, South Africa, the latter being the largest marine organisation in the Benguela region.
However, the available data and services for real-time and forecast modes diminish. Especially in the forecast domain only isolated capabilities and delivery modes exist. The maritime industry indicates a strong need for operational forecasts particularly in the shipping, naval, ports, fishing and the offshore and coastal mining industries. The oil and gas industry seems to have a demand for historic and real-time information, but little need for forecast data (this is probably offset by the design of platforms and the nature of the industry). Sovereignty and resource protection requires an expanded set
Environmental data requirements of maritime operations in the BCLME
371
of forecast and data dissemination services. The fishing industry indicated that, while potential exists for using such forecasts operationally, the efficacy of existing products is somewhat tenuous. In addition, most maritime industries indicated a strong requirement for a better understanding and longer term predictions of the marine environment within which they operate, to better assess the potential environmental impacts associated with their maritime activities. ACKNOWLEDGEMENTS A number of scientists participated in the Maritime session of the BCLME Forecasting Workshop, November 2004, and contributed to the science, spirit and success of the session. REFERENCES Agenbag, J., A.J. Richardson, H. Demarcq, P. Fréon, F.A. Shillington and S.J. Weeks. 2002. (extended Abstract). Relating the distribution of South African pelagic fish species to environmental variables using a novel quantitative approach. In: Report of a GLOBEC-SPACC/IDYLE/ENVIFISH workshop on Spatial Approaches to the Dynamics of Coastal Pelagic Resources and their Environment in Upwelling Areas (6-8 September 2001, Cape Town, South Africa). van der Lingen, C.D., C. Roy, P. Fréon, L. Castro, M. Gutierrez, L. Nykjaer, and F. Shillington,eds. GLOBEC Report 16: 81-83. Bishnoi, P.R. and B.B. Mainik. 1979. Laboratory study of behaviour of oil and gas particles in salt water, relating to deepwater blowouts. Spill Technology Newsletter 4(1): 24-36. Chapman, P. and L.V. Shannon. 1985. The Benguela Ecosystem Part II. Mar. Biol. An. Rev. 23,183-251. Etkin, D.S., P. Wells, M. Nauke, J. Campbell, C. Grey, J. Koefoed, T Meyer, S Reddy. 1998. Estimates of oil entering the marine environment in the past decade: GESAMP Working Group 32 Project. Proceedings of the Twenty-first Arctic and Marine Oilspill Program Technical Seminar: p 903 – 910. Garzoli, S.L., A.L. Gordon, D. Kamenkovich, D, Pillsbury and C. Duncombe Rae. 1996. Variability and sources of the south eastern Atlantic circulation. J. mar. Res. 54:1039-1071. Lutjeharms, J.R.E. and L.V. Shannon. 1997. A century of physical oceanography in South Africa; In search of the legacy of John D. Gilchrist. In A.I.L Payne and J.R.E. Lutjeharms,editors. A century of Marine Science in South Africa. Sea Fisheries research Institute, Royal Society of South Africa, p 1730. (also Trans. Royal Soc. S. Afr. 52 (1): 17-30). McCauley, R.D., 1994, Seismic surveys. 19-122 in Environmental implications of offshore oil and gas development in Australia – The findings of an Independent Scientific Review. Sean, J.M J.M. Neff and P.C. Young, APEA, Sydney, Australia. 695pp. McCauley, R.D., J. Fewtrell, A.J. Duncan, C. Jenner, M-N. Jenner, J.D. Penrose, R.I.T. Prince, A. Adhitya, J. Murdoch and K. McCabe. 2000a, Marine seismic surveys – A study of environmental implications. APPEA Journal 2000: 692-708. McCauley, R.D., J. Fewtrell, A.J. Duncan, C. Jenner, M-N.Jenner, J.D. Penrose, R.I.T. Prince, A. Adhitya, J. Murdoch and K. McCabe. 2000b, Marine seismic surveys: Analysis and propagation of air-gun signals; and effects of air-gun exposure on humpback whales, sea turtles, fishes and squid. Report produced for the Australian Petroleum Production Exploration Association, 198 pp. Payne, A.I.L., J.A. Gulland and K.H. Brink. 1987. The Benguela and comparable ecosystems. S. Afr. J. Mar. Res. 5. 957 pp. Rennell, J. 1832. An investigation of the currents of the Atlantic Ocean and those which prevail between the Indian Ocean and the Atlantic. J.G and F. Rivington, London. 359 pp.
372
M.L. Grundlingh et al.
Sakhalin-1. 1994. Technico-economical calculations for the project Sakhalin-1. 11: Preliminary environmental impact assessment and suggestions on environmental protection. EXXON, Sodeko, SMNG, 180 pp (Russian). Shannon, L.V. 1985. The Benguela Ecosystem Part I. Evolution of the Benguela, physical features and processes. Mar. Biol. An. Rev. 23:105-182. Shannon, L.V., L.Y. Shackleton and W.R. Siegfried. 1988.The Benguela Ecology Programme: the first five years. S. Afr. J. Sci. 84:472-475. Smith, G.G., G.P. Mocke, R. van Ballegooyen and C. Soltau. 2002. Consequences of Sediment Discharge from Dune Mining at Elizabeth Bay, Namibia. J. Coast. Res. 18(4):776-791. State of the Environment Report, 2001. http://www.environment.gov.za/soer/nsoer/index.htm Turnpenny, A.W.H. and J.R. Nedwell. 1994. The effects on marine fish, diving mammals and birds of underwater sound generated by seismic surveys. Report from Fawley Aquatic Research Laboratories Ltd, 40 pp + 9 pp appendices.
Part IV: The Way Ahead
ThisPage pageis intentionally This Intentionally left Leftblank Blank
Large Marine Ecosystems, Vol. 14 V. Shannon, G. Hempel, P. Malanotte-Rizzoli, C. Moloney and J. Woods (Editors) © 2006 Elsevier B.V. All rights reserved.
375
16 Towards a Future Integrated Forecast System G. Brundrit, C. Bartholomae, Q. Fidel, A. Johnson and J. Guddal
SUMMARY Getting useful and effective marine forecasts to the community is a key objective of the Benguela Current Large Marine Ecosystem Programme. This chapter examines recommendations from the Workshop sessions which might be incorporated into the design of a future integrated forecast system for the BCLME. Five candidate predictive capabilities are identified for priority action, and assessed for their relevance and value to the region. The system requirements needed to realise these predictive capabilities are reviewed, and additional infrastructural capacity noted. Finally, the elements of a future integrated Benguela Regional Forecast Facility are set down as a challenge for joint action by the countries of the region. INTRODUCTION This Chapter brings together the recommendations made in the various chapters of Part III of this volume on our collective wisdom on forecasting in the Benguela, and seeks to incorporate those recommendations into the design of a future integrated forecast system for the Benguela Current Large Marine Ecosystem (BCLME). A pragmatic view of the challenges in capacity building and institutional strengthening in the region, necessary for the implementation of any envisaged forecast system, are addressed. As a start, it will be important to examine the existing forecasting efforts in the region, so as to understand the context for forecasting, the needs that have driven these efforts, and the potential for expanding these activities into a more valuable forecasting system. The three national fisheries agencies, the Instituto Nacional Investigação Pesqueira in Angola, the Ministry of Fisheries and Marine Resources in Namibia and Marine & Coastal Management in South Africa, have the responsibility to advise their respective governments on the proper management of their living marine resources. Given the fluctuating nature of many of these fish stocks and the danger of over-fishing to their long-term sustainability, forecasting is recognised as a valuable management tool. However, forecasts of what might happen to a particular fish stock can only be built on a detailed understanding of, at least, the life history of that species, and an effective
376
G. Brundrit et al.
stock monitoring programme to provide reliable estimates of recruitment success and current biomass. These estimates, and the resulting forecasts, are typically turned into management advice for setting an annual Total Allowable Catch for that particular stock by the national fisheries agency responsible. The necessary monitoring and understanding are only available for a few of the economically important fish stocks of the region. Much needs to be done. Other national agencies are responsible for providing operational forecasts of environmental conditions in the Exclusive Economic Zones of each country. The national meteorological agencies provide regular weather forecasts for coastal waters, whilst the South African Weather Service, through the World Meteorological Organisation, is responsible for forecasts in the South-East Atlantic Ocean. National hydrographic agencies monitor sea level, and publish tidal predictions, particularly for the ports of each country. Increasing industrial activities in the Exclusive Economic Zones of the region make use of specialised forecasts, derived from the public good monitoring, and delivered by national and international marine environmental enterprises. For example, remote sensing of sea surface temperature is used to identify the position and strength of temperature fronts. These, in turn, are the proxies for advice to the fishing industry for fish stock targeting, to the shipping industry for ship routing, and to the offshore oil and gas industry for warnings of high currents. Open ocean wave forecasts are reported to port authorities, the diamond industry and the offshore oil and gas industry, to assist in their operational planning. In summary, the existing forecasting activities are somewhat fragmented, with responsibilities spread over the public and private sectors and little contact between them. Forecasting is regionally under-developed, with initiatives in Angola having to cope with very poor infrastructure. In the future, there is much to be gained from taking an integrated approach to what are common needs and priorities in the region as a whole. Such an approach, consistent with the objectives of the Benguela Current Large Marine Ecosystem, would involve, inter alia, • A regional monitoring effort based on a common infrastructure, • A forecasting approach based on ecosystem (process) models, • The operational use of the forecasts for better management advice, and • The enhancement of the socio-economic benefits to the countries of the region from the improved sustainability of offshore industries. Progress will require the consideration of forecasts of value that are relevant, doable and cost-effective for the region. Recommendations have come from earlier chapters as to which forecasts are important for the region, and how they might be realised. A selection of these high priority forecasts are now introduced as candidate predictive capabilities with a view to their incorporation into a future Benguela Regional Forecast Facility.
Towards a future integrated forecast system
377
CANDIDATE PREDICTIVE CAPABILITIES FOR THE BCLME Abundance Related Predictions of Living Marine Resources (Chapter 14, this volume) The long term sustainability of the fishery depends on progress towards a complete ecosystem definition, with knowledge of its sensitivity and robustness. A thorough exploration of these will await the use of prognostic primitive circulation models incorporating a full description of the relevant biology. Such models will enable the variability within the ecosystem to be investigated, and the critical states, which lead to a regime shift, to be identified. Modelling, with a view to forecasting, can also be valuable in the exploration of the detailed interactions of an ecosystem. Processes associated with the life history of specific fish stocks, and the manner in which they are crucial to recruitment success, can provide valuable insights into the workings of the ecosystem.
Pilot Study on Environmental Controls of Pelagic Fish Stocks The purpose of the pilot study is to identify the environmental factors controlling the annual recruitment of economically important pelagic fish stocks in the Benguela. • Factors critical to spawning success • Factors controlling transport out of spawning areas • Factors ensuring access to and retention in nursery areas • Factors encouraging growth in the nursery areas
The approach is to establish an annual climatology of the circulation in the region, using a circulation model such as ROMS, with surface and boundary forcing monitored from historic in situ and remotely sensed observations. The strength and variability resulting from these environmental controls can then investigated through, for example, individual based modelling incorporating details of the life history of the pelagic stock.
The output of the study would be a comparison of the controls operating at the different stages of the life history of the pelagic stock, together with indicators that can act as proxies for these controls. These indicators can then be used as input to an expert system, to improve the estimates, from direct monitoring, of annual recruitment available (Chapter 11, this volume).
Pelagic stocks amenable to this approach are found in both the northern and the southern Benguela. Further detail can be found in Chapter 14 of this volume on forecasting of living marine resources.
The Benguela Niño (Chapter 10, this volume) The Angola-Benguela Front forms the northern boundary of the BCLME. The Front shifts its position seasonally, but in some years the Front is displaced far southward and tends to persist for a few months in this extreme position. This is the Benguela
378
G. Brundrit et al.
Niño. Its occurrence means that the shelf waters offshore of Angola and northern Namibia are warm equatorial waters rather than the cold waters of the BCLME. These waters are also nutrient poor and are characterised by low oxygen conditions. All of this has serious implications for the living marine resources in the region. The annual life history of the economically important fish stocks can be disrupted, leading to recruitment failure and long term impacts on the stock biomass. More immediately, the fish stocks are displaced southwards with the shift of the Front and may not be available to the fishery. It is strongly suspected that the occurrence of a Benguela Niño is remotely forced by changes in the Equatorial Atlantic Currents. As such, it may be possible to forecast the shift of the Angola-Benguela Front and the occurrence of the anomalies of Angola and northern Namibia with a two month advanced warning.
Proof of Concept Study Advanced Warning of the Benguela Niño The purpose of the proof of concept study is to confirm that the Benguela Niño is remotely forced from the Equatorial Atlantic, and that the southward shift of the Angola-Benguela Front and the ocean anomalies on the shelf of Angola and Northern Namibia can be forecast with a two month advanced warning.
The approach is to set up a wind-driven circulation model of the Equatorial and South-East Atlantic Ocean, and to confirm the dynamics of the propagation of sub-surface anomalies from the equatorial region into the northern Benguela over a period of two months. This model, driven by historic winds in the Equatorial Atlantic, can then be used to hindcast the occurrence and strength of Benguela Niños over the past twenty five years, thereby validating the proof of concept.
This Proof of Concept Study can then be converted into a pre-operational forecast model based on altimeter observations from the Topex-Poseidon satellite, in situ observations from an extended PIRATA equatorial ocean buoy network, and sea surface temperature observations from the TRMM satellite. Whilst the pre-operational forecast model may be an extension of the original dynamic model, it is also possible to use an empirical-statistical model, underpinned by the climatology, for the forecasting and management advice.
Pre-conditioning of Low Oxygen Water (Chapter 13, this volume) Sub-surface Low Oxygen Water (LOW) is present throughout the subtropical eastern Atlantic. Its presence reflects the “age” of this water, with its long-term exposure to biological activity and its slow replenishment in the lee of the South Atlantic Gyre. The Low Oxygen Water leaks southward in the poleward undercurrent, on both the shelf and slope in the BCLME. It is thought to be the source water of the hypoxic subsurface layer which occasionally comes onto the shelf of Namibia, and even continues south of Lüderitz into the shelf waters of the west coast of South Africa. The Low Oxygen Water then comes into contact with belts of high organic sediments
Towards a future integrated forecast system
379
on the shelf, which can strip off the little oxygen remaining and initiate sulphur blooms in the surface waters. Such conditions are associated with fish kills and rock lobster walk-outs. There appear to be at least two stages in the leakage of LOW into the BCLME, with serious implications for the fisheries. Both appear to be amenable to cost-effective forecasting. The first is the remotely forced push of the LOW polewards and onto the BCLME shelf. The second is the local wind forced biological enhancement associated with the development of Harmful Algal Blooms during the relaxation phase of the upwelling cycle.
Proof of Concept Study Low Oxygen Water Pre-conditioning The purpose of this proof of concept study is to identify the environmental triggers that control the leakage of Low Oxygen Water into the BCLME, and to confirm that this pre-conditioning can be forecast with a two month advanced warning.
The approach is to use a similar wind-driven circulation model of the Equatorial and SouthEast Atlantic Ocean, as for the advanced warning of Benguela Niños, with additional biogeochemical capabilities. The focus of this study will, however, be the remote forcing of the poleward undercurrent carrying LOW from the pool in the Angola Gyre and into the BCLME region.
It will then be a priority to identify a proxy measurement to act as a trigger for the entire process. This proxy is likely to be associated with conditions in the Angola Gyre, and any preoperational model may well require observations from a south-east extension to the PIRATA moored buoy network.
The pre-operational model, either as a dynamic forecast model or as an empirical statistical model, will then be investigated to give a two-month advanced warning of LOW preconditioning in the BCLME.
Harmful Algal Blooms (Chapter 12, this volume) Algal blooms are to be found intermittently everywhere in the BCLME. Problems arise when the algal species involved is toxic in some way or when the intensity of the bloom leads to completely anoxic conditions throughout the water column. The consequences are particularly severe when they occur in rock lobster habitats and localities favoured for stock recruitment. The mariculture industry is only economically viable if high water quality standards are met; harmful algal blooms are a distinct threat to those standards. It is possible to give a five-day warning of the onset of a Harmful Algal Bloom, and for management to receive advice on the shoreline impact in the locality. This lead
380
G. Brundrit et al.
time would be of value in arranging for mitigating actions, such as the closure of the mariculture facility to the open sea, which would safeguard the certification of the facility.
Pilot Study on the Prediction of Harmful Algal Blooms The purpose of this pilot study is to establish the physical and biogeochemical controls on the evolution of harmful algal blooms within the BCLME, with a view to their prediction over the five day upwelling cycle.
Before any locality comes under threat from HABs, there needs to have been an influx of Low Oxygen Water into the region, and the possibility of further degradation of this pre-conditioned water. The approach is to investigate how the local meteorological forcing brings preconditioned water off the shelf and close inshore. A small-scale high-resolution threedimensional circulation model, with biogeochemical capabilities, is used to describe the onset of the HAB, to delineate the spatial extent of the bloom and to characterise its evolution and dispersion over time.
Starting from the initial Low Oxygen Water conditions, and with real time local winds, such a model can provide a five day nowcast of algal growth and decay with details of any drastic changes to the low oxygen environment. In situ optical measurements from a moored buoy, and remotely sensed colour imagery from satellite, will assist in confirming the development of the bloom. Local experience may lead to the identification of any toxic species present.
Operational Forecasting of Coastal Sea States (Chapter 15, this volume) The Benguela region is subject to extreme sea states. In the south, the significant wave height on the shelf can exceed ten metres on an annual return period. Whilst these heights are reduced in the north, the prevailing high wind conditions lead to hazardous sea states on a regular basis. Forecasts of sea state are essential for safe and efficient fishing operations in the coastal seas of the BCLME. The same operational considerations occupy the attention of other economically important offshore activities, such as the port authorities, coastal diamond mining and the offshore oil and gas industries. Extreme sea states are a recognised risk factor for all coastal and offshore investments in the BCLME. Sea state forecasting can mitigate the consequences of extreme events through real time operational planning, leading to economic savings from the prevention of infrastructure damage and the reduction of down-time. Forecasts of remotely forced ocean waves entering the BCLME are routinely available from global agencies such as NOAA, and value adding vendors such as www.buoyweather.com. These take the form of three-day forecasts at one degree intervals along the ocean boundary of the BCLME. The need is to bring these three-day forecasts into coastal waters, right up to the particular localities occupied by the offshore industries at risk.
Towards a future integrated forecast system
381
Operational Nowcasting of Coastal Sea States The purpose of this study is to set up operational nowcasting of sea state at those coastal and shelf sites where offshore industries are active.
This can be accomplished through dynamic downscaling of the deep-water directional wave spectrum, to give the local directional wave height spectrum at sites on the shelf. The approach requires the use of a shallow water wave model such as SWAN, forced by the local wind, and with bottom dissipation and topographic refraction incorporated into the wave model. If needed, a surf zone model can be attached to deal with beach sites. Repeated computations will provide the structure of an empirical transfer function, taking each deep water wave height spectrum and producing the corresponding inshore wave height spectrum.
Given the nowcasts of deep water sea state, validated from in situ observations, and the predetermined transfer function, the output of this operational nowcast would be the directional wave height spectrum with a three day lead time, at a “virtual buoy” located at relevant inshore sites.
SYSTEM REQUIREMENTS CAPABILITIES
FOR
THE
CANDIDATE
PREDICTIVE
Candidate predictive capabilities, which might form the basis of a regional forecasting effort, have been discussed from the viewpoint of their relevance as regional priorities. The emphasis now turns to their realisability, starting with the infrastructure support that would be needed. The present monitoring network will be explored, and compared with the needs of the candidate predictive capabilities. The feasibility of adding any new monitoring effort will then be assessed. Better knowledge of stock life histories and ecosystem dynamics is a prerequisite before a full application of an ecosystem approach to fisheries can be attempted. These are gradually being assembled, so that much of the forecasting will remain for the future. In this long term, two approaches appear relevant; qualitative expert systems based on robust indicators of the state of the ecosystem and prognostic dynamic modeling of the ecosystem based on versatile extensions of existing models. Continued research exposure to these approaches will help the national fisheries agencies ensure their rapid implementation in the region. State of Environment forecasting will rely heavily on adequate observations from the regional monitoring effort. Whilst physical variables may be approaching adequacy, the biological monitoring remains seriously under-sampled. There are also national contrasts in adequacy within the region, and there exists a real need for better quality control of data. In Namibia, the impact of extreme environmental events in the coastal ocean on their living marine resources has been recognised, and regular State of Environment Reports to the Ministry of Fisheries and Marine Resources have been instituted. This should be extended to all the countries of the region, and would
382
G. Brundrit et al.
provide the context for an effective regional early warning system. Several of the candidate predictive capabilities depend on the assembly of reliable climatologies of environmental variables and these, in turn, depend on proper quality control of observational data. A regional BCLME initiative using the Southern African Data Centre for Oceanography would assist in ensuring progress in this endeavour. To turn to the specific proposals of forecasts of the Benguela Niño and LOW preconditioning, these have some common requirements. The proposed Proof of Concept Study of the influence of the Equatorial Atlantic is achievable using an existing dynamic model, and hindcasting with readily available re-analysis of past weather, such as ERA40. The extension of the model to include the necessary biogeochemical dynamics of the propagation of the LOW signal will be a challenge (Chapter 17, this volume). The identification of a remote trigger will probably rely on observations from an extension of the PIRATA moored buoy network towards the Benguela region, a development that will require a proper cost-benefit analysis. The eventual preoperational forecasting models for the Benguela Niño and the LOW pre-conditioning will also make use of readily available satellite imagery. Once these models have been defined and tested, the objective of providing two-month forecasts appears feasible. The introduction of three day forecasts of localised HABs will, of necessity, follow the LOW preconditioning study. The forecasts will require wind observations from in situ moorings and the validation of local area wind climatologies, leading to the availability of five-day forecasts of local wind conditions for the forecast models. Optical measurements will also be needed from the moorings, backed up with colour satellite imagery. This approach appears to be feasible, but would benefit from a pilot study before widespread expenditure is made. The sea state forecasting makes use of a mature and readily available open ocean forecasting system. The extension of the forecasts through to the coastal virtual buoys uses available software, the same local wind observations as the HAB forecasting, and requires local high-resolution bathymetric charts. To move to the operational phase is entirely possible, but requires time and the necessary investment. An important imperative for the BCLME is to develop a common regional approach to the forecasting effort and to utilise a common monitoring platform. In this, the BCLME can draw on experiences gained elsewhere, such as the New Jersey shelf (refer to contribution by Haidvogel et al. in the accompanying CD-ROM) and the forecasting envisaged in the coastal regions of GOOS (Malone, also in the CD-ROM), but the detail needs to be adapted to local priorities and capabilities. Given the system requirements arising from the regional candidate predictive capabilities, a comprehensive list of the regional infrastructure requirements can be assembled. The requirements fall into several categories. The first consists of the in situ monitoring requirements and their associated forecasting capabilities for the ocean environment and living marine resources, which are natural extensions of the present responsibilities of the national fisheries agencies. The other categories consist of new initiatives, often utilising internationally available observations and software, where
Towards a future integrated forecast system
383
Regional Infrastructure Requirements for Forecasting Capabilities In situ moored buoys PIRATA extension deep sea mooring Local metocean buoys, for wind, sea state, temperature, salinity, air pressure Local “green” buoys, for nutrients, oxygen, chlorophyll Coastal ocean laboratories Coastal ocean radar systems Offshore monitoring lines State of Environment reporting LMR stock abundance estimates, life history processes Regional models, expert systems Ecosystem states, ecosystem dynamics, regime shifts Prognostic models, expert systems
Quality control of observational data Archiving of historic data Local and regional climatologies, physical and biogeochemical Empirical-statistical forecasting
Satellite remote sensing products: Winds, e.g. Quikscat Sea surface temperature, e.g. TRMM Sea surface altimetry, e.g. Topex-Poseidon Ocean colour, e.g. MODIS
Local and regional bathymetries Sea State forecast models Open ocean forecasts, e.g. WW3: http://polar.ncep.noaa.gov/waves/main_int.html Coastal forecasts, e.g. SWAN: www.wldelft.nl/soft/swan Ocean circulation models Equatorial Atlantic, with biogeochemical capabilities Benguela region, with IBM capabilities Local HAB, with biogeochemical capabilities, e.g. www.hab.org.za
further investment is needed to enhance the regional forecasting capability. In line with the regional approach of the BCLME, this forecasting infrastructure should be embodied in a regional data management centre, perhaps building on the existing Southern African Data Centre for Oceanography, a BCLME Remote Sensing Facility, and a BCLME Ocean Modeling Platform. These new facilities will also require properly trained personnel, and a Regional Training Centre to undertake this necessary capacity building. It is not the case that the national fisheries authorities already have the capacity to undertake the monitoring and forecasting in the first category of this list. The required infrastructure will have to be built up off the existing base. This varies markedly between BCLME countries. The country in the region most in need of infrastructure and human capacity building is Angola where improvements to the infrastructure base
384
G. Brundrit et al.
in both equipment and personnel are required. Moreover, Angolan national laboratories are in urgent need of additional technical support, finding it difficult to maintain the equipment for essential services (refer to contribution by Fidel in the accompanying CD-ROM). There would appear to be little point in investing in new sophisticated equipment without addressing these basic needs. Fortunately, the prospect of rapid economic development in Angola, on the back of the burgeoning offshore oil and gas industry, is good. Thus national and regional investment in basic infrastructure, and scientific, technical and management skills is beginning to accelerate, and plans for future economic development can now be made. The common regional approach to an integrated forecast system for the BCLME as a whole will therefore play a key role in fast-tracking Angola’s development in this area. Offshore economic development is also vital in Namibia and South Africa. The need for a parallel investment in infrastructure and an associated capacity building programme is important to all three countries, and to many other regions worldwide. Capacity building must take on an “operational” profile, enabling nations to implement the marine services required by society, maintaining the vital links to science, technical infrastructure and international cooperation. It must be based on identified priorities as well as the utilisation of shared observation and data resources and shared technical and scientific service tools. Not all of these background conditions are adequately present today. However, experience from existing oceanographic services, the availability of freely exchangeable data and sophisticated numerical models, and the expanding use of Internet technology provide the prospect of rapid implementation of ocean forecasting systems. The capacity building activities must find a balance between front-running high technology, and the realism needed for robust, sustained observing and prediction systems. The aim must be to make nations optimally selfsufficient in using ocean and observing systems to protect the economic needs of society in the coastal ocean (see contribution by Andrioli et al. in the accompanying CD-ROM.).
CONCLUSIONS AND RECOMMENDATIONS The implementation of a marine forecasting system is recognised to be of value in improving the quality of management advice on living marine resources, and in safeguarding offshore industries operating in hazardous coastal waters. The three countries of the Benguela region, Angola, Namibia and South Africa, have common needs and priorities in fisheries, diamond mining, oil and gas production, and shipping in their Exclusive Economic Zones. They believe that advantage can be gained from a co-operative regional approach to addressing these needs, rather than each country attempting its own independent effort. As far as the implementation of a regional forecasting system is concerned, the lead can be taken by the BCLME to establish the regional facilities which will form the basis of such a system. The various national fisheries and environmental agencies will need to be heavily involved. They will be the prime local monitoring agencies, supplying in situ
Towards a future integrated forecast system
385
observations to the new regional forecasting facility, and they will be the principal clients, using the forecast products as the basis for management advice to their respective government departments. Their guidance will continue to be essential in setting the priorities for action by the regional forecast facility. An attractive approach will be to target forecasts of value using common high quality observations for multiple purposes across prioritised application areas. It is recommended that the Benguela Regional Forecast Facility should be established with responsibility for the development and dissemination of forecasts of value to the Benguela community. This should be done through proof of concept studies to confirm the feasibility of new forecasts, and pilot studies for operational forecasts in prioritised application areas. As well as strong links to the national agencies, the Benguela Regional Forecast Facility will require the founding of several new facilities with distinct responsibilities within the integrated forecast system. • An Extended Observation Network will be needed, mainly under national control, to gather in situ observations using a common infrastructure throughout the region. This may involve standard coastal observation laboratories at high priority sites in each country. • A Remote Sensing Facility will be needed to analyse relevant satellite imagery and to extract the information essential for input into regional forecasting models. • A Data Management Centre will be needed for quality control of all observational data, for archiving of historic data, and for the generation of climatologies to place the forecasts into a reliable context. • An Ocean Modeling Platform will be needed to house the computing power and modeling software for the development of (prognostic) dynamic models and (diagnostic) empirical-statistical models. • A Benguela Training Centre will be needed to ensure the on-going availability of properly trained scientific, technical and management staff to generate and utilise the forecasts of value from the Benguela Regional Forecast Facility. This training should be carried out in a hands-on manner, with personnel being responsible for the success of the various studies mounted in the Benguela Regional Forecast Facility.
ThisPage pageis intentionally This Intentionally left Leftblank Blank
Large Marine Ecosystems, Vol. 14 V. Shannon, G. Hempel, P. Malanotte-Rizzoli, C. Moloney and J. Woods (Editors) © 2006 Elsevier B.V. All rights reserved.
387
17 Forecasting a Large Marine Ecosystem John Woods SUMMARY The Benguela project is unique among LMEs in seeking to go beyond diagnosis to useful forecasting. That is a daunting challenge. This essay shows how 21st century modelling practice might be applied to forecasting LMEs. It addresses signal-to-noise and predictability in modelling the marine ecosystem. That leads to four classes of forecast systems to provide information about changes on short, medium and long term, and for planning. Each is limited by predictability: the first in the atmospheric weather (Nowcasting), the second in the oceanic weather (Forecasting) and the third in the ecosystem itself (Climate prediction). The fourth class, What-if? Prediction, avoids these limits by combining hindcast boundary conditons with scenarios for exogenous events. The essay closes with a vision of the future that extends L.F.Richardson’s pioneering work in meteorology to operational prediction of LMEs. INTRODUCTION The Benguela project has the unique goal among LMEs of forecasting rather than merely diagnosing aspects of the ecosystem that are important for mankind. It brings ecology into operational oceanography, which has hitherto been concerned almost exclusively with predictions based on marine physics. Operational oceanography Operational oceanographers deliver descriptions and sometimes predictions about aspects of the sea that are important to paying customers. To do so they use tools that have been developed by oceanographers to explore the sea. And they use modelling techniques that oceanographers have developed to understand what they discover. Every customer has a responsibility that requires information about the sea. It may be a statutory responsibility of government arising from national law or international treaty, or it may be a commercial responsibility arising from exploitation of the sea. A customer is different from other individuals concerned about the problems of the sea: he or she has access to funds to pay for the information needed to perform his/her duties. The source may be taxes or grants from external bodies such as development banks, or the source may be the operating budgets of commercial enterprises. Often the funds are guaranteed only for a few years, whereas the operational system needed
388
J. Woods
to collect the information needs a long-term commitment. This is the chronic problem of monitoring. The traditional solution in education, health and welfare, to replace renewable funding by prior endowment, has not yet found its charitable sponsor for the marine environment. Meanwhile, operational oceanography leads a hand-to-mouth existence subject to the whims of politicians and investors. But operational oceanography does exist. Making it work is the task of service providers, who generate the information needed by customers. They exist in public and private sectors. Sometimes a government ministry acts both as customer and inhouse service provider. But increasingly the two tasks are separated, with the customer remaining in the ministry (or commercial headquarters) while the service provider acts independently as a government-owned agency or commercial company. Both draw on the know-how and technologies developed in research communities located in universities, and public or private oceanographic institutions. Often the service providers commission researchers to improve the flow of information by extending the knowledge base or to develop new tools. The prospects for improvement result from dialogue between researcher, service provider and customer. The Benguela Workshop provided an opportunity for exploring such opportunities. Operational oceanography is not a recent development. It has a long and distinguished history, from mapping observations, (e.g. Rennel’s 1795 map of the Agulhas current), to 19th century predictions of tides at ports, and 20th century predictions of wind waves. Dynamical models are routinely used operationally to predict storm surges and surface currents for a few hours/days ahead in support of the offshore energy industry, and the drift of flotsam. They play an important part in predicting the coastal flooding and the movement of oil pollution, harmful algal blooms and disabled vessels. On longer times scales, ocean dynamical models have been coupled to atmospheric models for operational forecasting of the El Niño climate and for experimental prediction and global warming due to greenhouse pollution. The last of these has seen the introduction of simple models of the plankton ecosystem to compute oceanic uptake of carbon dioxide. It is widely recognised that modelling the ecosystem is equally important for operational oceanography related to coastal problems, such as the development (rather than mere advection) of pollution events, for fisheries, and for human diseases like cholera that have a marine origin. The first steps towards ecosystem modelling were taken in the latter part of the 20th century, but the subject remains in its infancy. We shall see below that current practice lags that in meteorology by decades. One of the goals of this essay is to consider what it will take to introduce into LME prediction the philosophy and practice that has made weather forecasting so successful. The motivation is to identify priorities for investment that will lead to useful forecasting in the Benguela LME. This chapter assesses that challenge and identifies solutions that have worked in forecasting other systems. It highlights two key factors: the source of forecasting errors and the limits to predictability. Weather forecasting has been revolutionized by addressing those factors. They guide priorities for investing to improve the forecasting
Forecasting a large marine ecosystem
389
system, so that each segment is in balance, whether the mathematical models, the observations, or data assimilation. MODELLING PRACTICE IN THE 21ST CENTURY Classical method of induction In the 20th century most consulting engineers adopted the method of induction to predict how a system would respond to various scenarios for exogenous forcing. The method uses mathematical equations fitted to observations of the bulk properties that are of concern to the customer. Ecologists and epidemiologists have adopted the same method. In the 21st century there is a growing tendency for best practice to be based on models that do not depend on empirical relationships between the forecast products. This new approach uses the methods of complexity science. Complexity science In complex system modelling the scientific equations do not describe interactions between the forecast products but processes at a finer-grained level. Integrating such models creates a virtual system. The bulk variables of interest to the customer are computed as emergent properties of the virtual system. They are not constrained by empirical relationships with each other and with exogenous properties. This new approach has a number of advantages. First, the scientific equations for finegrained processes can often be derived from reproducible experiments performed under controlled conditions. That makes them more trustworthy than empirical relations between bulk properties. Second, the whole system tends to adjust more gracefully (and more realistically) to changes in exogenous forcing. Third, it avoids the artificial feedback between bulk properties that can excite chaotic fluctuations. Avoiding artificial chaos is a pre-requisite for useful forecasting. Plankton ecosystem models based on demographic equations are prone to chaotic instability. It is avoided by the Lagrangian Ensemble method, which simulates the life histories of millions of individual plankters whose behaviour and physiology are governed by biological primitive equations (Woods 2002, 2005; Woods, Perilli and Barkmann 2005). Signal and noise One of the prerequisites for any forecasting system is to have the signal (i.e. the forecast product) greater than the noise (i.e. the uncertainty in the product). The R&D programme in forecasting seeks to reduce the noise. Priorities are determined by
390
J. Woods
sensitivity tests, which relate the errors in each component (model, data, method of integration, etc.) to the noise in the products. Assessing and controlling errors in marine ecosystem models is a relatively young subject. It has two main components. The first is the inherent noise in the ecosystem due to natural variability in the environment, especially turbulence, which produces demographic noise in the plankton populations, and (through biofeedback) in the physical and chemical environment. The magnitude of this noise can be assessed by ensemble simulations. The second component of persistent model noise arises from competition among species represented in the model. The emergence of a ranking of relative abundance in balance with the ambient forcing often takes many years. Indeed, an ecosystem forecast normally reports the state of emergent properties in an ecosystem that is in a transient state of adjustment to the multi-year history of upstream forcing. In most cases the greatest source of noise in observations lies not in instrumental, but sampling error. This is particularly true for the resource data (especially nutrient concentrations) used to initialize the integration of the ecosystem model. It is usually necessary to invest in collecting resource data specifically in support of each forecasting system. Non-resource data are not strictly needed to initialise a forecast, because the ecosystem adapts to an attractor that is independent of them. However, they can be used to reduce the adjustment time. Data assimilation The aim of data assimilation is to accelerate the adjustment of the forecast system to its attractor. The ecosystem adjusts so slowly that data assimilation is essential for LME forecasting. But research on ecological data assimilation has not yet yielded useful operational procedures. The focus has been on observations that are easy to make, like ocean colour, i.e. phytoplankton pigments in the mixed layer. However they provide a relatively soft handle on the ecosystem, which tends to be controlled by the zooplankton. Verification The quality of the forecast is assessed by comparing its products with independent observations. Needless to say, this process of verification is valid only if the noise levels in the forecast and observed versions do not mask the difference between them. In practice most observations of opportunity, such as ocean colour, fail that criterion. It is best practice to collect observations with low noise specifically for model verification.
Forecasting a large marine ecosystem
391
Circulation modelling The latest ocean circulation models use an adaptive mesh to minimize computation in problems that require very high spatial resolution at unpredictable sites in the ocean (Ford et al. 2004a,b)1. The method is ideal for predicting the occurrence of local spots of mesoscale upwelling and deep convection. It also offers the very high resolution needed to model the complex bathymetry and flow in coastal waters. Using an adaptive mesh (rather than a uniform mesh with very high resolution everywhere) can reduce computation by a factor of one hundred. Ecological modelling The goal of ecological modelling is to predict bulk properties of the ecosystem that are important to customers. The strategy is to compute them as emergent properties of a complex system, which is controlled by finer-grained processes. The challenge for the modeller is to simulate these fine-grained processes realistically. In the case of ecosystem modelling they are the biological processes by which individual living organisms interact with their environment and with other organisms. The first step is to model the biological processes occurring in plankton, which by definition cannot swim fast enough to change their ambient environment usefully by migrating horizontally. Their behaviour is essentially one-dimensional, with strong diel and seasonal vertical migration. The biological processes of a planktonic organism (plankter) can be expressed mathematically by fitting curves to data from experiments on cultures performed under controlled conditions. These reproducible experiments provide the biological bedrock for ecosystem modelling. They yield biological primitive equations that are just as reliable as those describing physical and chemical processes in the ecosystem (Woods 2002). The challenge for ecosystem modellers is to design models in which these physical, chemical and biological processes of individual plankters are combined to predict bulk properties, which may be physical (mixed layer depth, ocean colour), chemical (concentration of pollutants) or biological (a harmful bloom or epidemic). The criterion for a well-designed ecosystem model is that it is globally stable, i.e. produces a stable attractor for all parameter values and resource levels. The properties of that attractor are not influenced by the initial conditions used to integrate the model. If, as is often necessary, the forecaster uses initial conditions that are not on the attractor, then the ecosystem takes some time to adjust to the attractor. The adjustment time scale is governed by the interval between successive reproductions in the slowest growing species. For plankton, these are often the larger zooplankton (e.g. copepods), which often reproduce only in a brief annual growing season each year in high latitudes. An ecosystem model that behaves stably in this way is a pre-requisite for ecosystem forecasting. The Lagrangian Ensemble metamodel (Woods et al. 2005) is the first to do so. By way of contrast, Popova et al (1997) have shown that the 1 See the ICOM website: http://amcg.ese.ic.ac.uk/cgi-bin/index.pl?page=home.htm.
392
J. Woods
common Eulerian metamodel (as used, for example, in ERSEM, the European Regional Seas Ecosystem Model, Baretta et al. 1995) is not globally stable: its predictions follow a strange attractor, with large inter-annual uncertainty in upwelling conditions. SHORT-TERM LME FORECASTING Operational oceanography embraces a wide range of services, designed to support: safety at sea, offshore operations, coastline protection, pollution control, fisheries and aquacultures, human health and mitigation of diseases from the sea. These applications are featured in earlier chapters of this book, so there is no need to describe them in detail. Here we classify the various forecasting services in terms of how far ahead in time they are needed (see Table 17-1). This allows the designer to take account of the limits to predictability of each component of the forecasting system. It helps to establish realistic expectations about the likely errors involved in forecasting for each service. And it focuses research onto attainable objectives. Table 17-1 Potential for LME forecasting on different time scales
Nowcasting
Seasonal forecasting
Climate prediction
What-if? prediction
Pollution
advection
development
-
remedial action
Toxic blooms
advection
development
occurrence
remedial action
Eutrophication
advection
development
occurrence
remedial action
Fisheries
-
recruitment
competition
fisheries policy
Health
-
epidemiology in plankton
occurrence
health planning
Alien species
advection
development
competition
remedial action
Climate change
-
-
community change
fisheries policy
Time scale: Problem:
sea level
Offshore operations
impact on structures search & rescue
-
-
planning and design of structures
Every forecasting system involves design compromise. The goal is to focus investment in modelling and observations onto those aspects that can yield useful reduction in the uncertainty of each forecast product. Every component of the system must earn its keep in maintaining an acceptable signal-to-noise level. Here I discuss some of the design issues that need to be addressed in LME forecasting. I start by
Forecasting a large marine ecosystem
393
considering the prospects for forecasting as a classical initial-value problem in which the exogenous initial and boundary conditions are determined by observations. The model includes only plankton, whose functions are described by biological primitive equations, which can be determined by experiment. It excludes fish and other higher organisms. That avoids the large uncertainty in equations for learning and choice of behaviour, which affect the demography and biofeedback in higher organisms. Atmospheric predictability The predictability of this simple LME forecasting system is not limited by its internal physical, chemical and biological processes. It is limited by an exogenous factor, the weather, which provides the boundary conditions for the forecast system. Meteorologists know that a weather forecast loses predictability after about one week. If the weather model is integrated beyond that limit it continues to generate weather patterns that are physically realistic, and compatible with the exogenous climate forcing (the sun, greenhouse gases, etc.), but its storms, fronts and clouds are no longer controlled by the initial conditions. When an ecosystem model is integrated beyond about one week the emergent properties will exhibit errors due to the deviation of the forecast weather from what actually occurs. Weather forecasting works because the legacy of initialization errors decays rapidly, leaving the variables in the simulated atmosphere to achieve a balanced state. The mathematical process of data assimilation accelerates that decay. Once the balanced state is achieved, the signal-to-noise ratio remains acceptable for the next few days. Ecosystem adjustment The legacy of initialization errors in the marine ecosystem decays far more slowly. It can take years for a simple food chain model to adjust to its attractor. A food web model with several competing species can take decades. This means that ecological adjustment has barely begun before the atmospheric forecast passes its limit to predictability. The forecast system must be designed to minimize errors arising from the use of an ecosystem that is not in balance with the boundary conditions. The ideal solution is to initialize the integration with a balanced ecosystem. Where does the forecaster obtain such information? There are two possibilities. The first is climatology, i.e. from the normal state of the ecosystem on the day of the year when the integration will be started. That normal state is derived from statistical analysis of observations taken in previous years. There are two sources of error, non-stationarity and sampling. The former can be assessed from the inter-annual variability in the historic data, provided the sampling was dense enough. Normally, however, ecological sampling is sparse, and the sampling error is even larger than that due to inter-annual variation. Using climatology to initialize short-term forecasts (i.e. for up to a week ahead) cannot guarantee that the initial state of the ecosystem is in balance with the weather.
394
J. Woods
The second possibility is to base the initial state of the ecosystem on synoptic observations collected for that purpose. The challenge is to achieve a sufficient sampling density to ensure the required signal-to-noise ratio. The principal source of error lies in aliasing mesoscale patchiness. The problem can be overcome in principle by raster sampling with an array of fast autonomous unmanned vehicles (AUVs, Griffiths 2003) such as AutoSub, each carrying instruments to measure temperature, light, turbulence, nutrient concentrations, phytoplankton pigments, and zooplankton species (perhaps by optical holography). Such platforms and instruments all exist, but they have not yet been used together to map the mesoscale structure of the ecosystem as input for LME forecasting. Trials are needed to assess the sampling errors in such observations. Advection by ocean currents The development of the ecosystem during the one-week forecast period depends on the mesoscale velocity field. Meandering jets, with associated upwelling patches a few kilometres across, are particularly important. This velocity field is exogenous, i.e. it is not influenced by the ecosystem. It can, in principle, be mapped by current meters mounted on the AUVs used to map the initial state of the ecosystem. The development of the velocity field can then be computed as an initial value problem, in parallel with the ecosystem forecast. The computation will include the Ekman flow driven by the overlying wind. That requires an ocean circulation model with a mesh size of, say, one kilometre in the horizontal and one metre in the vertical. The various components of the ecosystem are advected by this velocity field. Plankters follow three-dimensional trajectories that can be computed by Lagrangian integration of the velocity field without significant error. It is more difficult to control the errors in the Eulerian integration of field variables, i.e. the seawater properties like nutrient concentration. Errors in computing field advection can be a limiting factor in the forecast signal-to-noise ratio. LME nowcasting In the simple forecasting system we are considering, limited to one week ahead, it is reasonable to assume that the three-dimensional velocity field produced by ocean currents does not change significantly during the forecast period. That assumption is familiar in meteorology, where it is exploited in Nowcasting. That method describes the sequence of weather as a front that is advected over a site on the ground. We can use the term LME Nowcasting to describe the changes in ecological bulk properties encountered at a site in the ocean as the currents advect the mesoscale structure through it. The prediction can extend for a few days into the future, with the development in the ecosystem being computed from the surface boundary conditions, determined by a weather forecast. The oceanic velocity field is assumed to be unchanged through this period. Such LME Nowcasting can provide a useful service
Forecasting a large marine ecosystem
395
for prediction of the short-term, local impact of eutrophication, harmful blooms and pollution events. MEDIUM-TERM LME FORECASTING Many customers for LME forecasting are concerned with time-scales longer than one week, but shorter than one year. That is long enough for biological processes to produce major changes in plankton abundance and (through biofeedback) to the physical and chemical environment. Information of this kind is needed by customers concerned with such phenomena as fisheries recruitment, eutrophication, toxic blooms, pollution and plankton-borne diseases. The challenge for the LME forecaster is to predict the development of these phenomena from first principles, using an ecosystem model driven by the weather and ocean currents. The technical problem is that these exogenous boundary conditions cannot themselves be forecast, because the time scale lies beyond the limits of predictability of the weather both in the atmosphere and in the ocean (the latter being mesoscale turbulence). The consequence is that the boundary conditions used to drive the ecosystem model will inevitably contain errors, which reduce the signal-to-noise ratio in the forecast products. The forecaster must work closely with the customer to discover what is the minimum lead-time needed for the forecast product to remain useful. If it is less than the limit to predictability of mesoscale turbulence in the upper ocean (say one month) the strategy will be to use a high resolution model (such as ICOM2) to forecast the changing mesoscale velocity field as a four-dimensional boundary condition for the ecosystem model. This can be appropriate when the phenomenon to be forecast occurs within a critical period lasting a few weeks. This can be the case for eutrophication and harmful blooms. Such an approach can also predict the development of a pollution event that occurs during the summer, or the development of a cholera epidemic in zooplankton. In all these examples, the atmospheric forcing will be based on hindcasting, i.e. selecting a weather sequence from the same season in a previous year. Hindcasting We have seen that running an atmospheric model beyond one week produces a climatologically realistic sequence of weather events, even though they are not constrained by the initial state of the weather. That may be sufficient for longer-term LME forecasting. However, running a global atmospheric model is expensive. An alternative is to use historical descriptions of the weather, created by feeding archived observations to a modern weather forecasting model using modern data assimilation techniques. The ERA40 data set, created in this way by the European Centre for Medium-range Weather Forecasting, contains the global state of the weather at sixhour intervals at a resolution of one degree of latitude and longitude for the period 1963-2002. The LME modeller can select from ERA40 the weather over any period of 2
ICOM = Imperial College Ocean Model (Ford et al. 2004a,b)
396
J. Woods
time at any location around the world. The period may extend to several years, which is sufficient for the ecosystem to adjust to its attractor. Fisheries recruitment The critical period for fisheries recruitment in most cool, temperate ecosystems occurs in spring when the mortality of fish larvae peaks.3 The challenge is to predict the mortality with a signal-to-noise level that is sufficient to resolve inter-annual variation due to exogenous factors such as the weather, ocean currents and the trophic cascade. The information is not needed in real time, so the model can be run after the critical period is ended. That allows the recruitment forecaster to use the sequence of weather that actually occurred at the site during the critical period of a few weeks. Thus fisheries recruitment forecasting can run in hindcast mode, and still deliver the product on time for the customer. LONG-TERM LME FORECASTING We now consider variation in the LME from years to decades. This raises two issues: natural changes in the ecosystem when the climate has a stationary annual cycle, and the response to climate change. On these time scales the key ecosystem phenomenon is competition leading to a community of plankton with thousands of species. The challenge is to predict the mean relative abundance of the more important species (for a customer) at any given location, and its inter-annual variability. This is the new science of prognostic biodiversity, which will become an important discipline within biological oceanography in the 21st century.4 It will solve Hutchinson’s famous paradox of the plankton: why do we find so many species when there seem to be so few niches in the upper ocean? The answer has come from modelling the ecosystem with many competing species. Numerical experiments with sixteen competing species of diatom in a virtual ecosystem have shown that it is possible to compute their changing relative abundance as a function of ambient climate (Dr. R.L. Wiley, personal communication). Natural selection favours the species best fitted to the marine environment created by that climate. The other species decline in abundance at a rate that depends on their ranked competitive advantage (an emergent property of the computation). Highly ranked species decline very slowly: their emergent extinction time is many years, even decades. A small set of species with biological attributes close to the best fitted one exhibit neutral competition, with first one then another becoming most abundant over a The approach described here will need extending to predict recruitment in upwelling systems such as the Benguela where spawning may extend over months. However, the basic approach remains valid. 4 Prognostic biodiversity is the science of predicting the relative abundance of species by primitive equation modelling. The science is in its infancy, with the aim being to achieve useful signal-to-noise in relative demography predicted by Lagrangian Ensemble simulation of competition among species (Woods 2005). 3
Forecasting a large marine ecosystem
397
period of decades. This is the result of demographic noise due to turbulence in the mixed layer. Finally, if the climate is artificially changed during the numerical experiment the ranking of competitive advantage adjusts, leading some species that had become nearly extinct to recover and become well-fitted. Now we consider the circulation in the Southern Ocean. Investigations with eddyresolving ocean circulation models have revealed the complex trajectories of water parcels passing southern Africa. The combination of permanent currents and transient eddies mixes water parcels with a time scale of decades. The plankton ecosystem in a water parcel is in a permanent state of transient adjustment to the ambient climate experienced along its track. At any location along that track the plankton community comprises many competing species with a ranking of relative abundance that reflects their changing competitive advantage established during many years upstream under changing ambient climate. The challenge for prognostic biodiversity is to predict this geographically-Lagrangian attractor for the plankton community at any location. The computation must also take account of lateral mixing, which modifies the dominant process of upstream adjustment. The goal of long-term forecasting of an LME is to use the method of prognostic biodiversity to predict how the relative abundance of key species of plankton varies with natural climate variability, such as the El Niño and North Atlantic Oscillation, and how it will respond to IPCC scenarios for future climate change. The key species include those that provide food for commercial fish species, or produce harmful algal blooms, or host cholera or other diseases. Prognostic biodiversity provides a longterm forecasting procedure that is one of the pre-requisites for implementing the Reykjavik declaration, which requires fisheries policy to be based on an ecosystem approach. WHAT-IF? PREDICTION Customers need forecasting support when they are dealing with a crisis. They also need to make plans for how best to deal with future crises whenever they will occur. One of the most important roles of LME operational oceanography is to support such planning. It is done by computing the ecological consequences of a variety of scenarios for exogenous events that can provoke a crisis. This is called “What-if? Prediction”. It uses the same ecosystem models as those used in forecasting. The scenarios have three parts. The first uses the method of hindcasting to describe natural forcing by ocean circulation and weather. The second part adds one or more events. These might be pollution by, say, oil or alien species from ballast water, or run-off from farming or effluent from coastal development; or it might be a change in fish stocks. The third part comprises an option for remedial action. There are two goals. The first is to discover how the proposed remedial action will help alleviate the unwanted consequences of the event. The second is to discover the (often unexpected) side effects of the remedial action. Information from What-If?
398
J. Woods
Prediction can be used to plan the optimal response to natural disasters, and to license human activities, such as fishing and coastal development. A VISION OF THE FUTURE In 1922 Lewis Fry Richardson published a book “Weather prediction by numerical process”, which laid the foundations for modern operational meteorology. In his book he briefly considered the possibility of predicting changes in the sea surface temperature, which provides the marine boundary condition for the atmosphere. He famously remarked that it might be done by the same process as he had outlined for the atmosphere, but he hoped that “it would not come to that.” Eighty years later, climatologists use coupled ocean-atmosphere models in which sea surface temperature and the fluxes between the ocean and atmosphere are computed by his method. If Richardson were writing a new edition of his book today I imagine he would consider the possibility of forecasting the marine ecosystem. He might hesitate, as he did for sea surface temperature in the first edition. But there are powerful motivations for proceeding, because biological processes strongly affect many properties of the sea that mankind wants to exploit sustainably. So in the 21st century it is no longer a question of whether it can be done, but how. In this essay I have presented a personal vision of what it will take to make useful predictions of the Benguela LME. The key is to choose a metamodel for the ecosystem that is globally stable, so that its bulk properties adjust to an attractor that is independent of initialization errors (other than resources). That metamodel can be used to create specific models with plankton species and nutrients appropriate for each application. In many cases the biological primitive equations needed for such models already exist; where they do not, marine biologists will be commissioned to derive them from culture experiments. Atmospheric boundary conditions can be obtained from weather forecasting models, directly for nowcasting, or indirectly for hindcasting. Oceanic circulation can be obtained from a model driven by the same atmospheric data. The model must have sufficient resolution to simulate mesoscale jets with their patterns of upwelling, which control plankton patchiness; adaptive mesh modelling as used by ICOM promises computational economy. The ecological initial conditions depend critically on the resource concentrations: it will be essential to establish a permanent monitoring operation to map nutrients in the Benguela LME. The performance of the forecasting system can be assessed routinely by the Ecological Turing Test, which determines whether the prediction of some feature of the ecosystem is statistically distinguishable from an observation of that feature, after taking account of the errors in each (Woods 2002). Data must be collected explicitly for such validation; it is not good practice to rely on data of opportunity, which seldom satisfy the strict criteria for sampling.
Forecasting a large marine ecosystem
399
Finally, while LME forecasting is in its infancy, every effort should be made to establish a prototype forecasting system for the Benguela LME, which satisfies the technical criteria listed above, and provides the products needed by the customer. That should be done in the spirit of pre-operational experimentation, with the goal of measuring forecast errors, and identifying and correcting their causes. The Benguela project offers the first opportunity to go down the road of LME forecasting. The road may be long, but there is every prospect of success. REFERENCES Baretta, J.W., W. Ebenhöh and P.Ruardij, eds. 1997. The European Seas Ecosystem Model II. Journal of Sea Research 38(3/4). Ford, R., C.C. Pain, M.D. Piggott, A.J.H. Goddard, C.R.E. de Oliveira and A.P. Umpleby. 2004. A nonhydrostatic finite-element model for three-dimensional stratified oceanic flows. Part I: Model formulation. Monthly Weather Review 132 (12), 2816-2831, (2004). Ford, R., C.C. Pain, M.D. Piggott, A.J.H. Goddard, C.R.E. de Oliveira and A.P. Umpleby, “A nonhydrostatic finite-element model for three-dimensional stratified oceanic flows.” Part II: Model validation'. Monthly Weather Review 132 (12), 2832-2844, (2004). Griffiths, G., ed. 2003. Technology & Applications of Autonomous Underwater Vehicles. London, Taylor & Francis, 342pp. Popova, E.E., M. J. R. Fasham, Osipove, A.V. and V.A. Ryabchenko. Chaotic behaviour of an ocean ecosystem model under seasonal external forcing. Journal of Plankton Research 19(10): 1495-1515. Richardson, L.F. 1922. Weather Prediction by numerical process. Cambridge, 236pp. Woods. J.D. 2002. Primitive equation modeling of plankton ecosystems. Ch.18, 375-425 in Pinardi, N. and J.D.Woods, eds. Ocean Forecasting. Springer-Verlag.. Woods, J.D. 2005. The Lagrangian Ensemble metamodel for simulating plankton ecosystems. Progress in Oceanography 67:84-159. Woods, J.D., A. Perilli, and W. Barkmann. 2005. Stability and Predictability of a Virtual Plankton Ecosystem created with an individual-based model. Progress in Oceanography 67:43-83.
ThisPage pageis intentionally This Intentionally left Leftblank Blank
401
INDEX Abalone, 129,161,262 Abidjan Convention, 7 Acoustic survey, 16,119,120,152,178,204,266, 332,341 Adaptive management, 15,18,29,212 Agulhas Bank, 50,58,65,67,69,101,102,104,110,115,116, 119, 125,127,129,130,147,155-158, 160, 166,210, 247,248,322,323,331-333,339, 340,342,345, 346,350,363 boundaries associated with, 334-335 Agulhas Current, 3,8,36,49,51,52,57,59,65,68,69,92, 119,223,227,234,235,246,312,318, 325,331,334,335,337,341,349,389 Agulhas Current Large Marine Ecosystem, 36 Agulhas Retroflection, 59,60,127,234,343 Agulhas Ring, 52,59,65,67,238,312, 334,342, 346,349 Albacore, 363 Alert system, 126 Alexandrium catenella, 40,129,130,277 Algae, see also Harmful Algal Blooms 126,137,139,146,241,277,284,294 Algal blooms, see Harmful Algal Blooms Alien species, see Introduced species Algoa Bay, 159,160 Anchovy, 22,33,50,68,69,94,97,102,106, 108-110,115-117,120-124,148-158,160, 161,163,165,166,169,170,173-184,186, 189, 190, 195,196,201,202,204,207,210, 214-216,218, 219,243,246,248,251,254258,262,266-272, 309-311,328,330,331, 338-346,363 eggs, 100,208,323 forecasting, 323 juvenile, 117 landings, 323 pelagic, 24 Peruvian, 208 predicting, 339 recruitment, 103,149,154,174,175, 257,309,322-324,332,333,338,339 Angola, see also Fisheries, Northern Benguela 3,65,77,91,148,155,160,229,230, 249,250,299,350,362,376,378,383 map, oil and gas, 353
oil and gas, 350 Angola Basin, 54,57,238,312 Angola-Benguela Front, 3,8,49,67,89, 91, 106,119,121,127,147,148,228,229, 237,240,246,247,312,325-328,377, 378 Angola Current, 8,49,51,52,75,76,85, 106,314,326,327 Angola Dome, 8,49,51,72,74,76,327 Angola Front, 240 Angola Gyre, 49,74-76,379 Anoxia, 6,8,16,88,90,272,297,305 Anthropogenic forcing, 172,241,310 Aquaculture, 31,253,273,391 Atlantic Current, 378 Atlas Bay, 166 Atmospheric circulation, 52,54,223,224, 227 Atmospheric forcing, 52,58,225,238,395 Atmospheric/ocean model, 186 Bakun’s triad, 320 Baltic Sea, 12,21,32,306 BCLME Programme, 5-9,18,111 BENEFIT, see Benguela-EnvironmentFisheries-Interaction & Training, Benguela Current Commission, see Interim Benguela Current Commission Benguela Current Large Marine Ecosystem, see also Central Benguela, Northern Benguela, Southern Benguela atmospheric forcing, 52-53 atmospheric variability, 223-228 Benguela environment, 3-5 Benguela shelf, 77-83 boundaries, 4 currents, 4 decadal change, 106-111 fisheries agencies, 375 food-web, 166 interannual variability, 106-111 maps, 4,313, ocean modelling, 57-64 physical processes and modelling, 51-52,65-67 pollution, 245 seasonal variability, 101-106 water masses, 54-57 Benguela-Environment-Fisheries-Interaction
402 &Training (BENEFIT), 5,15,55,67,111, 182,269,349 Benguela Niños, 6,8,50-52,54,64, 65,76,89,153, 155,160,166,172, 210,223-225,228-233,235,236,248, 262,297,299,300,307,312,327,338, 339,377-379,382 and SST anomalies, 230 Benthic communities, 315 Biogeochemical oxygen demand, 79,302 Biological variability, 148,162,211 Biomass loss, 317 Biomass recovery, 24 Birds, see Sea birds Blooms, see Harmful Algal Blooms Bottom species, 71,303 Boundary processes, 325-335 and Bakun’s triad, 320 Benguela boundaries, 4 Bray-Curtis, 138,139,141 Brazil Current, 57 By-catch, 246,249,268,362 Calanus, see Copepod California Current Large Marine Ecosystem, Fish and fisheries, 189-191 Modelling and forecasting, ecology, 199-201 Modelling and forecasting, physics, 197-198 Physical traits, 187-188 Productivity, 188-189 Canary Current Large Marine Ecosystem, fish and fisheries, 192-194 modelling and forecasting, ecology, 201-202 modelling and forecasting, physics, 198-199 physical traits, 191 productivity, 192 Cannibalism, 72,170,208,314,326,333,335 Cape Agulhas, 104,116,127,157 Cape Basin, 54,57,74,75,79,88,249,299, 330,331,334,335 Cape Columbine, 58,81,82,89,95, 127-129,140,142,157,298,312,313, 319,320,323 Cape Frio, 57,77-79,81,83,84,86,91, 101,297-299,305 Cape Hangklip, 160,161 Cape of Good Hope, 3,49,349,357
Index Cape Peninsula, 58,100,101,103,128, 318,320,322,323 Cape St. Francis, 58 Carrying capacity, 16,28 Catch per unit effort, 159,201,202,328 Central Benguela, 51,57,72,75-77,79, 80,83,90,297,308 Cetaceans, see Whales Chile, 20,194-196,202,203,208,213, 217-220 Chlorophyll a, 37,40,91,93-96,101,107, 115,123,135,138,141,219,286,287, 320,321,327 Chromium, 354 Circulation models, 50,58,60,68,100, 198,237,274,275,289,292,294, 337,377-380,383,391,394 Circulation patterns, 61,62,129,284 Climate change, 5,16,19,68,106,115, 172,178,204,205, 211, 213,216, 217,219,255,256,266,269, 270,306, 324,338,341,344,392,396,397 Climate forcing, 22,23,242 Climate models, 186,205,271 CLIPPER model, 60,61,63,224 Cod, 23,178,241,327,346 Congo River, 326,327,362 Continuous Plankton Recorder (CPR), 68,266,270, Copepods, abundance, 105 biomass, 102 calanoid, 107-110,112,121 egg production, 123 herbivorous, 94,122,124,209 Cormorants, 174,251 Côte d’Ivoire, 20,36,202 CPR, see Continuous Plankton Recorder CPUE, see Catch per unit effort Crab, 327 Cunene River, 101,111,148-150,210 Currents, see Agulhas Current, Angola Current, Atlantic Current, Benguela Current, Brazil Current, California Current, Canary Current, Eastern boundary currents, Humboldt Current, Somali Current, South Atlantic Current, South Equatorial Current
Index Cyclones, 37,97,137,237 Cyclonic circulation, 74,75,334 Dahlem Model, 8,216,270 Danger Point, 160 Dassen Island, 160 Delft3D-FLOW ocean model, 58,289,294 Demersal species, 22,23,312,327,328,330 Diamonds, 3,354-357,365,368,370,376,384 mining technologies, 357 Diarrhetic Shellfish Poisoning (DSP), 126, 130,131 Diatoms, 40,91,97,113,131,134,135,137,139, 141,189,274-277,291,396 Dinoflagellates bloom forming, 136,143,272 fish-killing, 129 harmful 135,146,294 high biomass, 125,127,131,144 toxic, 40,293 Downwelling, 77,186,188,276,279-282, 292,316 Dredging, 355,356,360,361,368 Dyer Island, 160 Early Warning System, ,8,37,39,42,86,87,239, 262,382 Eastern boundary current systems, 52,108,186, 191,194,266 Ecological modelling, 272,391-392 Ecopath-Ecosim model, 28,151,255,257 Ecosystem change, see Ecosystem variability Ecosystem goods and services, 18,35,360 Ecosystem indicators, 239,240,250-256, 258,264,265 definition, 250 Ecosystem management, 5,71,86,181,245, 296,306 Ecosystem variability, 147-184 defining ecosystem change, 241-243 predicting variability, 170-175 species replacement or alternation, 243 Eddies, 60,65,67-69,101,190-192,195,198,199, 209,234,246,284,319,322,332,335,337,365, 397 EEZs, see Exclusive Economic Zones Ekman Transport, 92,279,280,282,284,322 Elands Bay, 40,127,133,135,136, 145,287,301,307, 314,316,318 El Niño, see also Benguela Niños
403 29,37,67,165,178,183,185,186,189, 195-197,199,200,203,204,207,208, 211,223,232,233,388,397 El Niño-Southern Oscillation (ENSO), 5,10,50,54,115,185,194,207-210, 223-225,227,228,231-233,235-237, 278,297 ENSO, see El Niño-Southern Oscillation Environmental forcing, 13,71,144,153, 154,160,161,170-172,241, 245,325, 339 Equatorial divergence zone, 74 Equatorial forcing, 298-301, 305 Euphausiids, 92,97,100,104,119, 122,189,195,213, 330,341 Eutrophication, 11,16,17,19,22, 27,28,31,311,344,392,395 Exclusive Economic Zones (EEZs), 11,376,384, Expert system model, 239,241,257-261 Extreme events, 7,65,239,304,380 FAO, see Food and Agriculture Organization FAO Code of Conduct for Responsible Fishery Practice (2002), 25,31 False Bay, 128,129,160 Fish, see Bottom species, Demersal species, Neritic fish, Pelagic species age class, 153 eggs and larvae, 50,59,68,104, 115, 121,154,156,189,247, 268,312,320, 327,330,333, 335,343 fish indicators, 16 fish kill, 129,379 nursery areas, 97,104,115, 154,332,334,377 population dynamics, 92,94 predation, 115,208 resource variability, 149-166 spawning grounds, 108,154 Fisheries, see also Industrial fishery, Recreational fishing, Reykjavik Declaration on Responsible Fisheries Angolan, 162,164,167,328 fishery collapse, 325 Namibian, 88,148,160,180,184, 229,261,265,315,317,318, 326,328,331,362,375
404 South African, 148,158,175,252, 263, 265,295,314,316,322, 338,361-363 Fisheries management, 71,121,175, 177-179,181,182,186,204,213,252, 261, 266,272,274,276,277,344,361, 362 fisheries dynamics, 147,149,251 management organizations in BCLME, 362 multi-species approach, 255 single-species approach, 240 Fishing, see also By-catch, Overfishing fishing technology, 361-362 inshore fishing, 364-366 offshore fishing, 364 Fishing down the food web, 19,25 Fishing effort, 23,25,26,29,158,202, 203,211,340,364 Fishing gear, see also Purse seine 249,363 Food and Agriculture Organization (FAO) 19,22,31,155,158 Food chain, 13,23,30,32,174,393 food chain dynamics, 8,19 Food web, 16,28,31,115,144,153,171-173, 183,184,306 Forcing, see Anthropogenic forcing, Atmospheric forcing, Climate forcing, Equatorial forcing, Forcing scales, Meteorological forcing, Ocean variability, Physical forcing, Wind forcing Forcing scales, 89,295-297,299-301,303, 305,344 Forecasting, see also Maritime forecasting, Weather forecasting forecasting systems, 38,41,309,310, 324,384 hindcasting, 37,88,175,323,324,382, 395,397,398 nowcasting, 51,331,332,381,387, 392,394,398 what if scenario, 255,256,260, 296,306, 310,324,329,387,392,397 Foreign vessels, 148 FRAM (Fine Resolution Antarctic Model) 59,68 Bryan-Cox-Semtner ocean model, 59 Frontal dynamics, 144 Frontal system, 52,140,277 Fronts, see Angola-Benguela Front,
Index Thermal front Gannets, 161,177,248,251 Cape gannets, 147,148, 151-153,163,165,168,174 GEF, see Global Environment Facility Geographic Information Systems, see GIS Ghana, 20,36,193 GIS, 171,174,252,255,367 Global Environment Facility (GEF), 5,6,19,20,24,28,29,31 Global Ocean Observing System (GOOS),6,35-45,251,382 Global warming, 68,115,183,186, 205,206,248,255, 388 Gobies, 161,162,166,168,174,267,330 GOOS, see Global Ocean Observing System GOOS-Africa, 7,9,36-39,41-43 Governance indicators, 18 Greenhouse effect, 178 Greenhouse gas, 174,208,393 Guinea Current Large Marine Ecosystem, 9,18,20,21,36,44,216 Gulf of Thailand Large Marine Ecosystem, 21,23,24,26,33 Gyres, see also Angola Gyre 57,76,92,191,192,195,219 South Atlantic Gyre, 92,246, 378 Habitat,
benthic, 262 coastal, 3,17,335 critical, 86 degradation, 11 fish habitat, 119,256,320,324 food habitat, 119 habitat alteration, 6 habitat diversity, 134 habitat dynamics, indicators, 249,252 reproductive, 340 spawning, 111,173,184,246,272, 331,339 HABs, see Harmful Algal Blooms Haddock, 25,26 Hake, 94,104,111,147,148,150,152, 158,166,168,177, 178,192,196,247249,252,295,310,311,318, 328,330,
405
Index 331,333 biomass, 151,170,344 cannibalism, 170 Cape hake, 89,90,160,180,182,270, 271,326,335,342-343,346 catch, 149,151,169 decline, 153 hake trawling, 362,363 juvenile, 72,162,165,343 predicting recruitment, 326,327 recovery, 167 recruitment,72,106,165,309,314,326 spawning, 106 Harmful Algal Blooms (HABs), see also Red tides 6,8,11,16,17,28,89,94,125-146, 247,273-294,296,297,301,302,317,336, 379-381379-380 seasonal incidence, 131-137 Herring, 25,32,189 Round herring, 104,110,148,150,363 Hout Bay, 363 HUB mode operational modelling, 370 Humboldt Current Large Marine Ecosystem, fish and fisheries, 195-197 modelling and forecasting, ecology, 202-204 modelling and forecasting, physics, 199 physical traits, 194-195 productivity, 195 Hydro-acoustics, 108,110,322,332 Hydrogen sulphide, 125,144,273,297, 314,318,336,364,366 Hypoxia, 5,6,8,28,73,80,82,209,210, 273,297,301,305,311-314,335 IBM-hydrodynamic model, 92 Icelandic Shelf Large Marine Ecosystem, 23 Ichaboe Island, 152,160 Ichthyoplankton, 91,104,119,156, 157,189,197,322 Indicators, see Ecosystem indicators, Fish, Pollution, Productivity, Socioeconomic trends, Governance environmental and habitat indicators, 248,251,262,272 EPA indicators, 17 LME indicators, 15-18 single species indicators, 248, 251,264
size-based indicators,248,249, 251,272 socioeconomic indicators, 252, 253 spatial indicators, 248,249,251, 252, trophodynamic indicators, 248, 249,251,252 Industrial fishery, 164,196,262,263 Institute of Research for Development (IRD),7,38,65,177,198,213,266 Intergovernmental Oceanographic Commission (IOC), 6 Interim Benguela Current Commission, 18 International Union for the Conservation of Nature and Natural Resources (IUCN),153 Introduced species, 361,365,368,392,397 IUCN, see International Union for the Conservation of Nature and Natural Resources Jellyfish, 166,168,241,246,259 Kelp, 160,240,262,316 Kelvin wave, 51,76,230,232,235,283, 284, 298-300,304,326 Kob, 155 La Niña, 37,197,200,211,225,397,399 Lagrangian model, 139,188,202,318, 322,389,391,394,396,397, 399 Lambert’s Bay, 127 Lamont-Doherty Earth Observatory model, 185,199 Large Marine Ecosystems (LMEs), see LME driving forces, LME ecological criteria LME definition, 1-15 LME management, 19 LME modelling, 28-29 LMEs, see Large Marine Ecosystems LME driving forces, 17,19,21,28,92, 245,264,265 LME ecological criteria, 11,35 Lobster, 181,315,317,364,365 lobster walkout, 116,296,302, 303,316,366 Rock lobster, 88,116,145,146,148150,153,273,295,296,301-304,
406 306,307,310,311,314-317,334,337, 338,340,363,379 Low Oxygen Water, 51,71,78,89,125,247,249,269,295,297, 299, 301-303,305,307,309,310,312318,344,378-380 effect on hake fisheries, 72 and physical processes, 85 Lüderitz, 3,8,49,51,52,57,58,61,62,64,76, 77,79,81,83,84,86,91,92,97,101,104,106, 111,147,158,160,166,209,240,248,249,297300,305,306,312,314,318, 319,325,330, 331,339,378 MacCall’s basin model, 158,199,201,324 Mackerel, 25,189,201 Chub mackerel, 199 Horse mackerel, 104,147,148151,158, 161,162,166169,177,178,192,196,202,246, 249,252,263,310,328,334,363 Malgas Island, 152,163 Management, see Adaptive management, Ecosystem management, Fisheries management, Large Marine Ecosystems, Risk management Mariculture, 379,380 Marine mammals, 15,148,190,362 Marine pollution, see also Oil pollution 31,33,39,182 Marine resources, see Diamonds, Fishing, Oil and Gas Marion Island, 226,227 Maritime forecasting, 368-370,375,384 Maritime industry, see also Diamonds, Oil and gas, Ports, Shipping, 350,365,370 Maritime safety, 369 navigational safety, 369 Mauritania, 20,36,192,193,202,207 Maximum Sustainable Yield (MSY), 22,185 Mercury, 354 Mercury Islands, 160 MERIS, 37,42,286,288,289 Mesoscale processes, 199,309,310,318-325, 332,333,337 Southern Benguela, 323 Metals, 354 Meteorological forcing, 380 Meteorological services, 358,376 Millennium Development Goals, 41 Minerals, 3,366,369,370
Index Minimal realistic model, 257 Mining, see also Diamonds 3,37,354-357,360,368-370,380, 384 offshore mining, 337,369 Models, see Atmospheric/ocean model, Bakun’s triad, Bray-Curtis, Circulation models, Climate models, CLIPPER, Dahlem, Delft3D-FLOW ocean model, Ecological modelling, Ecopath-Ecosim, Expert system model, FRAM, HUB mode operational modelling, Hydrodynamic model, IBM-hydrodynamic model, Lagrangian model, Lamont-Doherty Earth Observatory model, MacCall’s basin model, Minimal realistic model, NCOM, NLOM, NORWECOM, NPZ model, Numerical modelling, Ocean-atmosphere model, Ocean circulation model, OPA, ORCA2 ocean model, Parada model, Parent model, Physical modelling, PLUME, POCM, Population assessment model, Predation model, Predictive models, Princeton Ocean Model, Regional Ocean Modelling System, Ricker models, SPEM, Statistical modelling, Stock assessment models, Tidal circulation model, Trophic model, Trophodynamic model, van Foreest and Brundrit MODIS, 42,288,337,383 Morocco, 20,36,202 Mossel Bay, 157,363 MSY, see Maximum Sustainable Yield Namaqua, 127-129,131,133,134,137, 140,142, 283,286,288,289,304,319, 320 Namibia, see also Central Benguela, Fisheries, Northern Benguela 3,5,6,9,18,20,36,49,52,59,71,73,92, 97,101,104-106,112,129,147,148, 153, 159,166,169,249,261,296,318, 328, 338, 366,370,381,384
407
Index Benguela Niño, 230 boundary processes, 326 central Namibia, 91,101,104, 106,156,314,328,330,335 diamond mining, 354-355 hake, 326 northern Namibia, 52,111,156,315,378 oceanic boundary, 328-330 oil and gas, 350 regime shift, 241 sardines, 229,261 southern Namibia, 52,129,163,355,356 upwelling, 209 Namibian Shelf, 75,76,78,79,83,91,105, 106,312,326,327,397,399 NAO, see North Atlantic Oscillation National Centre for Environmental Prediction, 59 National Oceanic and Atmospheric Administration (NOAA), 16,38,43,128, 133,201,228,288,337,380 NCOM (U.S. Coastal Ocean Model), 60 Neritic fish, 106,328 Neural networks, 112 Nitrate, 16,81,96,119,287 Nitrogen, see also NPZ model 11,19,27-29,32,90,94,99,112,115, 137, 144,354 model-predicted, 27 nitrogen export, 19,27,28 NLOM (U.S. Layered Ocean Model), 60 NOAA, see National Oceanic and Atmospheric Administration Non-indigenous species, see also Introduced Species,16 North Atlantic Oscillation (NAO), 54,191,194,210,211,207,232,397 Northeast Shelf Large Marine Ecosystem, 13,21,23,25,26 Northern Benguela, 51,54,59,67,77, 83,87,91,92,97-101,104-106,111-119, 129,147-153,155-162, 165-167,169, 172-174,228,240,246,247,249,262, 264,267,297,309,310,314,318,327, 328,330,331,335,337,339,378 NORWECOM model, 59 NPZ (Nitrogen-Phytoplankton-Zooplankton) model, 112,255,257,324 Numerical modelling, 6,41,57,60,65,280,
289-291,332,349,370,384 Nutrient enrichment, 108 Nutrient loading, 16,19,32 Ocean-atmosphere model, 398 Ocean circulation, 195,318,359,383 Ocean circulation model, 60,391,394, 397 Ocean color, 189 Ocean variability, 199,226,233 Offshore operations, 39,369,392 Oil and gas, 3,37,38,349-352,354,366, 368,370, 371,376,384 Oil pollution, 172,358,388 Oil spills, 6,39,40,336,337,352,358,359, 361,365,368,369 Treasure oil spill, 162,369 OPA model, 60,230 Orange River, 52,54,57,119,147,248,312, 314,330 ORCA2 ocean model, 233 Overfishing, 19,29,153,229,243,245, 246,310,339,375 Oxygen state, see also Low Oxygen Water 73 Palgrave Point, 156 Parada model, 58,97,257,324,332 Paralytic Shellfish Poisoning (PSP), 16, 129 Parent model, 66 Pelagic species, 23,26,154,169,190,192, 194, 202,203,212,265,311,330,331, 336 Penguins, 147,149,151-153 Phosphorus, 11 Physical forcing, 29,201,211,276,287,288, 337 Physical modelling, 59,185,186,257,292 Physical variability, 49-70,289 Phytoplankton, see also NPZ model abundance, 115,106,255,364 biomass, 91,107,115,125,128,129, 131,132,137,209,248,320 categories, 275 decay, 95 decline, 94,126 distribution, 114,125 near-shore, 189 scales, 94 seasonal variation, 131
408 shifts, 134 surface phytoplankton, 139,141 toxic phytoplankton, 277 Phytoplankton blooms, 91,95,101,189,275, 316,365 Pilchard, 50,192,202,363 Plankton, see Ichthyoplankton, Phytoplankton, Zooplankton Plankton variability, 92,106,113 PLUME, 112,115,116,172,257 POCM (Parallel Ocean Circulation Model), 60 Pollution, see Marine pollution, Oil pollution pollution indicators, 17 Population assessment model, 152,211 Port Elizabeth, 3,58,248 Ports, 358,360-361,365,368-370,376,388 Precautionary approach to fisheries sustainability, 25 Predation, fish predation, 115 lobster predation, 161 models, 171,250 predator-prey interactions, 29,101,108-110,171,174 sardine predation, 208 Predators, apex or top predators, 16,108,152, 174,207,210,243,245,248,251,252, 336 biomass, 169 diet, 161,163 Predictive models, 126,144,309,339 Primary productivity, 3,14,16,25,338 Princeton Ocean Model, 59,185,198 Pristine environments, 3,41,243 Productivity, see Primary productivity productivity indicators, 16 Purse seine, 363 Pycnocline, 209,279-281 QuikSCAT, 53,59,66,224,288,298,306,337, 359,383 Recreational fishing, 161,314,315 Red Sea Large Marine Ecosystem, 36 Red tides, see also Harmful Algal Blooms 129,137,139,140 Regime shift, 22,37,71,106,147-185,195, 199,202,239-272,306,377,383 abrupt change, 241 Canary Current, 211
Index definition, 186, 241, 242, 261 discontinuous shift, 241 gradual change, 241 Humboldt Current, 207 modelling, 8 Regional Oceanic Modelling System (ROMS), 5 Remote Sensing, see also Satellite remote Sensing, 38-40,43,72,86,247,260,277, 298,306,318,337,264,376,383,385 Research Vessels, 204 RV Africana, 40,55 RV Alexander von Humboldt, 100 RV Melville, 55 Reykjavik Declaration on Responsible Fisheries (2001), 261,362,397 Ricker models, 199,201 Rings, 52,210,223,234,335 Agulhas ring, 59,65,312, 334,349 retroflection, 51 Risk, 73,212,306,323,339,380 health risk, 126 oil spill risk, 352 risk analysis, 129,173,175,176,274 risk management, 301,368-370 shipping, 360 Robben Island, 160 Rossby waves, 197,223,224 Safety at Sea Convention (SOLAS), 359, 392, St. Helena Bay, 50,58,81,82,87-90,96,97, 101-103,105,107-109,113,115,128,131, 247,289, 302-305,314-316,320 Salinity distribution, 299 Sardinella, 148-150,155,160,162,164,192, 193,202,203,207,210,263,310,327,328,338
Sardine,
abundance, 115,156,207 Benguela sardine, 151,153-155, 157, 158,161,162,164,165,173, 339 biomass, 155 catch, 22,151,167,169,194,196, 208,328 collapse, 153,194,310 decline, 156,158,169,207,246 dynamics, 246 females, 162 juvenile, 246
Index larvae, 100,156,202,330 Pacific sardine, 189,190 recruitment, 257,322,339 sardine eggs, 100,156,202 spawning, 104,155,156,158,174, 246,339 Sardine and Anchovy Recruitment Programme (South Africa), 175 Satellite imagery, 119,364,382,385 Satellite remote sensing, 38,383 Satellite technology, see also Remote Sensing 287-288 AFRISTAR, 41 Scale factors, see Teleconnections, Scales of variability, Seasonal Variability Scales of variability, 28,66,74,126,197 decadal, 82,91,148,160,186,191, 227,297,327 interannual, 79,224 interdecadal, 223,227 seasonal, 300,302 Seabirds, 3,170,172,174,177,190,240, 243,262,362 Sea level change, 39 Sea Surface Temperature (SST),40,50,60,139, 172,185,188,196,278,288,323,364,376, 378,383,398 Sea Urchin, 160,161,262 Seals, 151,160,161,163,166,170,173,177, 192,247,248,251,326 Cape fur seals, 147,149,152,168,174 Seasonal variability, 62,77,101-106,115,224 Sea-viewing Wide Field of View Sensor, see SeaWiFs Search and Rescue, 6,367,369 SeaWiFS, 14,40,93,101,115,128,132,133, 135,188 Senegal, 20,36,191,192,198,199,202,203,207 Shelf processes, 8,309-347 Shellfish, see also Paralytic Shellfish Poisoning 126,129,130,131,193,210 Shipping, 37,39,245,357-361,365-370,376,384 Shrimp, 263,325 Single-species analysis, 240,248,251, 255,257,263 Snoek, 147,149-151,161,163,165,177,247 Socioeconomic trends, 15,17,18,29,176, 243,252,258,262,263,301,376 socioeconomic indicators, 17-18 Sole, 363
409 Somali Current Large Marine Ecosystem, 20,36 South Africa, see also Southern Benguela 5,6,18,20,36,40,43,49,52,59,71,73, 77,81, 88,91,101,105,107-109,114, 147,148,153,155,156, 160,170,226, 227,236,240,261,265,331,358,359,366, 367,376,378,384 anchovy, 175 Data Centre for Oceanography, 370, 383 demersal longliners, 363 fishing industry, 148,158,252, 361, 362 management, 317,370,375 marine resources, 295 modelling, 58 oil and gas, 350 pelagic fishery, 252,263,265, 338, 363, ports, 360 rock lobster, 295,314,316 sardines, 175,322 South Equatorial Counter Current, 49,75, 246, 300 South Equatorial Current, 75 Southern Benguela, 5,51,54,56,65,76,77,81,83,86-94,97, 98, 100,101,104,106-109,111,112, 119-170, 173-176,223,225, 234,240, 247,248,252,257, 264, 273,283,286290,296,299, 304, 309-312,314, 318, 320-324,330,336,339, 340 SPEM model, 59 Spills, see also Oil spills spill forecasting, 337 Squid, 189,325,334,335 SST, see Sea Surface Temperature Statistical modelling, 92,186,256,378, 385 Stock assessment models, 71,151,295 Storms, 39,349,355,356,361,388,393 stormwater, 365 Strategic Action Plan (SAP), 15 Stratification, equatorial, 66,85 seasonal, 133 vertical, 279,280,322 water column, 125,137,139,247, 274
410 Sulphide, see also Hydrogen sulphide 71,73,80,85,311,315,335,338 Sulphur, 8,379 Sustainability, fisheries, 25 long-term, 18,375 TAC, see Total allowable catch Teleconnections, 186,194,204-209,232, 233 Temperature variability, see also Sea Surface Temperature 64,96,196 Thermal front, 322,333 Thermocline, 53,57,73,74,76,78, 79,81,84,86,87, 97,112,224,230-232, 248,260, 297-300,303-305, 313,328, 334,370 Tidal circulation model, 198 Tides, 42,198,356,358,360,365,367,388 Total Allowable Catch (TAC), 25,118, 295,317,363,376 Tourism, 3,37,39,41,253,337 Toxic events, 80 Transboundary Diagnostic Analysis (TDA) 15,19,71 Transition zone, 52,187,188,318 Trawling, 148,160,177,314,363,364 Triggerfish, 193,194 Trophic levels high, 172 intermediate or medium, 108,173 low, 211,255 Trophic cascades, 171,396 Trophic model, 60,153,167,169,197,251, 252,264 Trophodynamic model, 115 Tsunami, 37,42 Tuna, 361,363 UNCED, see United Nations Conference on the Environment and Development (1992) UNEP, see United Nations Environment Program UNIDO, see United Nations Industrial Development Organization United Nations Conference on theEnvironment and Development (UNCED, 1992), 35 United Nations Convention on Biodiversity, 35 United Nations Environment Program (UNEP), 19,35,38
Index United Nations Industrial Development Organization (UNIDO), 38 Upwelling, 279-280 eastern boundary upwelling, 9,49, 51,147,188,195,209,240 equatorial, 72,85 mesoscale, 91,391 models, 279-280 Peruvian, 199 seasonal, 49,83,185,297 southern Benguela, 51,65,128,166 wind-driven, 92,140,186,205 van Foreest and Brundrit Model, 58 Variability, see also Scales of variability decadal, 82,91,148,160,186,191, 227, 297,327 interannual, 79,224 interdecadal, 223,227 multidecadal, 227 resource abundance, 149-155,176 resource biology, 161-166 resource distribution, 155-161 Walvis Bay, 49,52,72,80,87,104,105, 111,156,158,305,312-314,328,339 “Wasp waist”, 106,173,174,208,245 Wave, see also Kelvin wave 283-284 Weather forecasting, 60,276,350,358,388, 393,395,398 Wind stress curl, 282-283 Wind forcing, 8,50,51,65,91,154,185,188, 191,197-199,209, 226,245,298,302, 304,306,319,327,333,379 Whales, 251,351 World Conservation Union, see International Union for the Conservation of Nature and Natural Resources (IUCN) World Bank, 19 World Summit on Sustainable Development (WSSD), 24,41,261,262 Yellowtail flounder, 25,26 Zooplankton, see also NPZ model biomass, 92,111,208
Large Marine Ecosystems Series Large Marine Ecosystems of the North Atlantic Changing States and Sustainability Edited by: K. Sherman and H.R. Skjoldal, ISBN 0444510117
The Gulf of Guinea Large Marine Ecosystem Environmental Forcing and Sustainable Development of Marine Resources Edited by: J.M. McGlade, P. Cury, K.A. Koranteng and N.J. Hardman-Mountford, ISBN 0444510281
Large Marine Ecosystems of the World Trends in Exploitation, Protection, and Research Edited by: G. Hempel and K. Sherman, ISBN 0444510273
Sustaining Large Marine Ecosystems The Human Dimension Edited by: T. Hennessey and J. Sutinen, ISBN 0444510265
Benguela Predicting a Large Marine Ecosystem (Book + CD-ROM) Edited by: V. Shannon, G. Hempel, P. Malanotte-Rizzoli , C. Moloney, and J. Woods, ISBN 0444527591
ThisPage pageis intentionally This Intentionally left Leftblank Blank