Behavior of Marine Fishes: Capture Processes and Conservation Challenges

Behavior of Marine Fishes Capture Processes and Conservation Challenges Editor

Pingguo He

A John Wiley & Sons, Inc., Publication

Behavior of Marine Fishes Capture Processes and Conservation Challenges

Behavior of Marine Fishes Capture Processes and Conservation Challenges Editor

Pingguo He

A John Wiley & Sons, Inc., Publication

Edition first published 2010 © 2010 Blackwell Publishing Ltd. Chapter 8 remains with the U.S. Government. Blackwell Publishing was acquired by John Wiley & Sons in February 2007. Blackwell’s publishing program has been merged with Wiley’s global Scientific, Technical, and Medical business to form Wiley-Blackwell. Editorial Office 2121 State Avenue, Ames, Iowa 50014-8300, USA For details of our global editorial offices, for customer services, and for information about how to apply for permission to reuse the copyright material in this book, please see our website at www.wiley.com/wiley-blackwell. Authorization to photocopy items for internal or personal use, or the internal or personal use of specific clients, is granted by Blackwell Publishing, provided that the base fee is paid directly to the Copyright Clearance Center, 222 Rosewood Drive, Danvers, MA 01923. For those organizations that have been granted a photocopy license by CCC, a separate system of payments has been arranged. The fee codes for users of the Transactional Reporting Service are ISBN-13: 978-0-8138-1536-7/2010. Designations used by companies to distinguish their products are often claimed as trademarks. All brand names and product names used in this book are trade names, service marks, trademarks

or registered trademarks of their respective owners. The publisher is not associated with any product or vendor mentioned in this book. This publication is designed to provide accurate and authoritative information in regard to the subject matter covered. It is sold on the understanding that the publisher is not engaged in rendering professional services. If professional advice or other expert assistance is required, the services of a competent professional should be sought. Library of Congress Cataloging-in-Publication Data Behavior of marine fishes : capture processes and conservation challenges / editor Pingguo He. – 1st ed. p. cm. Includes bibliographical references and index. ISBN 978-0-8138-1536-7 (hardback : alk. paper) 1. Marine fishes–Behavior. 2. Fishery Management. 3. Marine fishes–Conservation. I. He, Pingguo. QL620.B44 2010 597.177–dc22 2009050959 A catalog record for this book is available from the U.S. Library of Congress. Set in 9.5 on 12 pt Times by Toppan Best-set Premedia Limited Printed in Singapore 1

2010

Contents Contributors

ix

Preface Clement S. Wardle

xi

Introduction Pingguo He

xiii

Part One: Locomotion and Sensory Capabilities in Marine Fish

3

Chapter 1

5

Chapter 2

Chapter 3

Swimming in Marine Fish John J. Videler and Pingguo He 1.1 Introduction 1.2 The Swimming Apparatus 1.3 Swimming-Related Adaptations 1.4 Styles of Swimming 1.5 Interactions between Fish and Water: Fish Wakes 1.6 Energy Required for Swimming 1.7 Swimming Speeds and Endurance 1.8 Concluding Remarks Fish Vision and Its Role in Fish Capture Takafumi Arimoto, Christopher W. Glass, and Xiumei Zhang 2.1 Introduction 2.2 Structure of the Fish Eye 2.3 Visual Function 2.4 Visual Capacity: Visual Acuity, Separable Angle, and Maximum Sighting Distance 2.5 Color and Appearance of Fishing Gear Underwater 2.6 Fish Vision and Its Application in Fish Capture 2.7 Concluding Remarks Hearing in Marine Fish and Its Application in Fisheries Hong Young Yan, Kazuhiko Anraku, and Ricardo P. Babaran 3.1 Introduction 3.2 Properties of Underwater Sound and Vibration 3.3 Underwater Sound Sources and Their Characteristics 3.4 General Morphology and Functions of Inner Ears and Ancillary Structures 3.5 Responses of Fish to Sound and Its Application in Fisheries 3.6 Concluding Remarks

v

5 5 8 12 13 14 18 19 25 25 25 27 32 35 36 40 45 45 45 47 48 53 60

vi

Contents

Part Two: Fish Behavior near Fishing Gears during Capture Processes

65

Chapter 4

67

Chapter 5

Chapter 6

Chapter 7

Chapter 8

Fish Behavior near Bottom Trawls Paul D. Winger, Steve Eayrs, and Christopher W. Glass 4.1 Introduction 4.2 Trawl Gear and Trawl Fisheries 4.3 Fish Behavior in the Pretrawl Zone (Zone 1) 4.4 Fish Behavior between Trawl Doors and in the Net Mouth (Zone 2) 4.5 Fish Behavior inside the Trawl Net and the Codend (Zone 3) 4.6 Factors Influencing Fish Behavior near Trawls 4.7 Concluding Remarks Fish Behavior in Relation to Longlines Svein Løkkeborg, Anders Fernö, and Odd-Børre Humborstad 5.1 Introduction 5.2 Worldwide Longline Fisheries 5.3 Description of the Gear 5.4 Chemoreception and Food Search—The Basis for Bait Fishing 5.5 Interactions between the Fish and the Longline Gear 5.6 Conservation Challenges and Potential Solutions 5.7 Concluding Remarks 5.8 Future Challenges Fish Pots: Fish Behavior, Capture Processes, and Conservation Issues Bjarti Thomsen, Odd-Børre Humborstad, and Dag M. Furevik 6.1 Introduction 6.2 Worldwide Use of Fish Pots 6.3 Fish Behavior in Relation to Pots 6.4 Conservation Challenges and Solutions 6.5 Concluding Remarks Large-scale Fish Traps: Gear Design, Fish Behavior, and Conservation Challenges Pingguo He and Yoshihiro Inoue 7.1 Introduction 7.2 Trap Fisheries and Trap Designs 7.3 Fish Behavior in and around Traps 7.4 Fish Behavior and Trap Designs 7.5 Size and Species Selectivity and Mortality of Escapees and Discards 7.6 Conservation Issues and Mitigation Measures in Trap Fisheries 7.7 Concluding Remarks Fish Behavior near Gillnets: Capture Processes, and Influencing Factors Pingguo He and Michael Pol 8.1 Introduction 8.2 Capture Mechanisms, Gear Designs, and Fishing Efficiency 8.3 Size Selectivity of Gillnets 8.4 Fish Behavior and Gillnet Fishing 8.5 Measures to Reduce Bycatch and Discards in Gillnets 8.6 Interaction of Marine Mammals, Seabirds, and Sea Turtles with Gillnets 8.7 Derelict Gillnets: Ghost Fishing Problems and Solutions 8.8 Concluding Remarks

67 67 69 75 82 89 95 105 105 106 108 110 114 123 130 132 143 143 143 146 150 154 159 159 159 165 170 172 174 178 183 183 184 187 189 192 195 197 198

Contents Chapter 9

Electric Senses of Fish and Their Application in Marine Fisheries Hans Polet 9.1 Introduction 9.2 Properties of an Electric Field in Water 9.3 The Electric Field 9.4 Application in Marine Fisheries 9.5 Conservation Issues 9.6 Concluding Remarks

Part Three: Contemporary Issues in Capture and Conservation in Marine Fisheries Chapter 10

Chapter 11

Chapter 12

Technical Measures to Reduce Bycatch and Discards in Trawl Fisheries Norman Graham 10.1 Introduction 10.2 Bycatch and Discard in World Fisheries 10.3 Cause of Bycatch and Discard 10.4 Technical Measures to Reduce Bycatch and Discard 10.5 Implementation of Discard Reduction Measures in Trawl Fisheries 10.6 Discussion 10.7 Concluding Remarks Mortality of Animals that Escape Fishing Gears or Are Discarded after Capture: Approaches to Reduce Mortality Petri Suuronen and Daniel L. Erickson 11.1 Introduction 11.2 Mortality of Discards and Escapees 11.3 Assessment of Mortality 11.4 Factors Causing Stress, Injury, and Mortality 11.5 Measures to Improve Survival 11.6 Concluding Remarks Effect of Trawling on the Seabed and Mitigation Measures to Reduce Impact Pingguo He and Paul D. Winger 12.1 Introduction 12.2 Review of Recent Studies on the Seabed Impact of Trawling 12.3 Description of Trawls and Their Operation 12.4 Pelagic and Semipelagic Trawls 12.5 Groundgear Modifications 12.6 Trawl Door Considerations 12.7 Other Trawl Gear Components 12.8 Beam Trawls 12.9 Concluding Remarks

Chapter 13 Measures to Reduce Interactions of Marine Megafauna with Fishing Operations Dominic Rihan 13.1 Introduction 13.2 Species and Fisheries Involved 13.3 Extent of Bycatch 13.4 Nature of the Problem

vii 205 205 206 209 219 228 231

237 239 239 240 242 242 257 258 259 265 265 266 269 276 283 286 295 295 295 296 299 301 305 306 309 310 315 315 316 318 320

viii

Contents 13.5 Regulatory Frameworks 13.6 Potential Mitigation Measures 13.7 Concluding Remarks

321 325 337

Appendix Species Names Mentioned in the Text

343

Index

347

Color Plate Section

Contributors

Norman Graham Marine Institute Rinville, Oranmore Galway, Ireland

Kazuhiko Anraku Faculty of Fisheries, Kagoshima University Kagoshima, Japan Takafumi Arimoto Fish Behavior Section Tokyo University of Marine Science and Technology Tokyo, Japan

Pingguo He University of Massachusetts Dartmouth School for Marine Science and Technology New Bedford, MA, USA

Ricardo P. Babaran College of Fisheries and Ocean Sciences University of the Philippines in the Visayas Miagao, Iloilo, the Philippines

Odd-Børre Humborstad Fish Capture Division Institute of Marine Research Bergen, Norway

Steve Eayrs Gulf of Maine Research Institute Portland, ME, USA

Yoshihiro Inoue Nakano-ku Tokyo, Japan

Daniel L. Erickson Oregon Department of Fish and Wildlife Marine Resources Program Newport, OR, USA

Svein Løkkeborg Fish Capture Division Institute of Marine Research Bergen, Norway

Anders Fernö Department of Biology University of Bergen Bergen, Norway

Michael Pol Massachusetts Division of Marine Fisheries New Bedford, MA, USA Hans Polet Institute for Agricultural and Fisheries Research Oostende, Belgium

Dag M. Furevik Fish Capture Division Institute of Marine Research Bergen, Norway

Dominic Rihan Irish Sea Fisheries Board Dublin, Ireland

Christopher W. Glass Institute for the Study of Earth, Oceans and Space University of New Hampshire Durham, NH, USA

ix

x Petri Suuronen Finnish Game and Fisheries Research Institute Helsinki, Finland Bjarti Thomsen Faroe Marine Research Institute Faroe Islands John J. Videler Dept. of Marine Biology Groningen University The Netherlands Clement S. Wardle Warkbraes, Craigievar Aberdeenshire, UK Paul D. Winger Fisheries and Marine Institute Memorial University of Newfoundland St. John’s, NL, Canada

Contributors Hong Young Yan Institute of Fishery Sciences National Taiwan University Sensory Electrophysiology Laboratory Institute of Cellular and Organismic Biology Academia Sinica (Taiwan National Academy of Science) Jiaoshi, Taiwan Xiumei Zhang College of Fisheries Ocean University of China Qingdao, China

Preface

the different fishing gears have been developed by the fishermen to make use of the details of fish behavior to catch or release the fish. Small fish pass through the open meshes of a trawl because they are soon exhausted when trying to swim with the moving net. Selective trawls work with precise knowledge of this scale effect. They are towed at just above the sustained swimming speed of the size and species of targeted fish. Larger fish, which are not exhausted, swim away when the net is lifted, whereas smaller fish are quickly exhausted and pass through the meshes or over herding ropes and wires. During the past 50 years, large amounts of data on fish behavior and fish capture processes have been collected. Interpretations of these data have generated a wealth of knowledge that has been applied to the catching and conservation of fish. The details of this developing knowledge tend to become scattered in refereed papers in different scientific journals and are often forgotten as time passes. It is important that every few years these papers are reexamined by experts and reinterpreted in relation to the current problems encountered in the world’s marine fisheries. The refereed papers, as collected in this book, give the reader, scientist, or fisherman authoritative views, both retrospective and forward looking, on the issues facing global marine fisheries. Not only does the behavior of target and bycatch species need to be understood, but all species of megafauna, which include many protected species, that may interact with the fish capture processes. The behavioral details must be studied and avoidance procedures developed. It is thus very relevant that an up-to-date collection of

For hundreds of years, nets were considered to sieve or filter the fish from the water. Then in 1952, a breakthrough was made during the first diving observations when fish were filmed near a towed fishing net. Instead of being passively filtered by the net, the fish were orderly swimming ahead of the groundgear of a Danish seine net, being herded through the net funnel and actively escaping through open meshes of the codend. The new technologies and underwater cameras allowed SCUBA divers to observe and record that fish were reacting to nets in many different ways that were related to the biological characteristics of fish (e.g., species, size, sex, and state of maturity), environmental conditions of the fishing ground (e.g., temperature, seasonal changes, day/night length, tide, current, wind, depth, changing color and light levels, clarity of water, and bioluminescence), and operational parameters (e.g., towing speed, colors of twines, gear and ship sounds). With so many variables, two hauls may never be the same, making causes and effects very difficult to investigate with limited tows during experiments at sea. Large aquaria and swimming tunnels were built so that many of these variables could be accurately controlled to allow experiments that measured the physiological limits of fish that were relevant when reacting to fishing gears under a variety of conditions. This research has included issues such as fish swimming capacity, schooling behavior, visual and hearing capability, bait odor distribution and chemoreception, electro senses, and learning and conditioning. Knowing the limits to the response abilities of fish gives us clues to help explain how

xi

xii

Preface

knowledge and a synthesis of fish behaviors, fish capture processes, and conservation issues are presented at a time when conservation is becoming the priority above and beyond the traditional values of marine resources. Currently and in the near future, all harvesting and extraction of our natural resources will need careful studies of the effects on the eco-

system and will involve detailed conservation measures in the capture fishery, including the avoidance or management of subjects like seabed damage, genetic selection due to size-selective fishing, damage to species diversity, and total fishing mortality. Clement S. Wardle

Introduction

stocks while avoiding depleted stocks became one of the major management tools. As a result, the fishing industry is allowed to catch only fish of certain species and sizes and at specific locations and times. Consequently, the aim of understanding fishing gear and fish behavior turned from catching more fish to catching selected fish. Development of selective fishing gears to reduce bycatch and discards became the mainstream of marine capture fishery research. As a result, research on fish behavior, especially fish reaction to fishing gears, has flourished since the 1970s. In a broad sense, fish behavior can be defined as the adaptation of fish to internal and external environments and to natural and artificial stimuli. In a narrower sense, related to fish capture, fish behavior can be considered the reaction of fish to physical and chemical stimuli associated with the fishing gear and its operation and the reaction to environmental changes in the forms of movement and distribution. Recognition of the importance of fish behavior in understanding and improving size and species selectivity and in realizing rational exploitation of the resource has encouraged applied fish behavior studies in the context of fish capture. As one of the few remaining wild resource harvesting activities, fishing operations have increasingly been criticized for collateral effects on other animals and the marine ecosystem. Discards in world marine fisheries represent huge wastes of resources. It has become a moral issue as well as a threat to global food security. Fish population decline and endangerment of some of charismatic

Humans might have started observing and interpreting fish behavior when they first threw rocks at a fish in a stream. Hunting has continued to this day, although increasingly sophisticated, in the wild capture fishery. During the course of fishing for various species under different conditions, a variety of fishing gears and methods have emerged. New materials and technologies have been invented and applied to design and construct gears and to study fish behavior, so as to capture fish more efficiently as well as to conserve fish for future use. Powered fishing boats and hauling equipment, synthetic gear materials, and echo-sounding equipment may be the most influential technologies in commercial marine capture fisheries, while echo-sounders and sonar, underwater cameras, and data storage tags have made great impacts on the study of fish behavior in the field. Modern fishing technologies, coupled with insatiable demands for seafood products, have caused great stress to fishery resources globally, which only a little over a century ago were still considered to be “inexhaustible.” Commercial fishing activity now occurs from the shoreline to depths exceeding several thousand meters, in the remotest of oceans, and even under polar ice caps. Dramatic collapse of some of the world’s major fish stocks, notably Norwegian herring and Chinese greater yellow croaker in the 1970s and Newfoundland northern cod in the early 1990s, stunned many in the fishing industry, fishery research, and management arenas. Because recovery can take a long time, if they do recover (as seen in Norwegian herring), selective fishing for healthy

xiii

xiv

Introduction

megafaunal species related to fishing operations have caused great concern and become a challenge for both the fishing and research communities. This book reviews and summarizes current understanding of fish behavior as it relates to capture processes in world marine fisheries and presents conservation challenges facing the fishing industry and fisheries researchers. It also illustrates potential technical solutions to the issues. The book is divided into three parts: (1) locomotive and sensory modalities relevant to capture; (2) fish behavior near fishing gears; and (3) conservation challenges in marine fisheries. The 13 chapters in the book were written by 22 leading researchers in the field and reviewed by 14 well-known experts, representing a total wisdom of 36 global scientists from 16 countries from Asia, Europe, and North America. The first part of the book, covering the locomotion and sensory capability of fish, consists of three chapters. Chapter 1 (by Videler and He) reviews and summarizes fish swimming mechanisms, styles, and capabilities. Chapter 2 (by Arimoto, Glass, and Zhang) describes fish vision, underwater light and the visual environment, and how and how well a fish can see and visually respond to a moving or stationary object. Chapter 3 (by Anraku, Yan, and Babaran) describes underwater acoustic properties and fish hearing capability, as well as presenting examples of the application of sound and acoustic stimuli in controlling fish behavior. Information presented in these three chapters is frequently been referred to in later chapters when the behavior of fish near fishing gears is described and interpreted. The second part of the book consists of six chapters that describe fish behavior near different fishing gears and the fish capture processes. The major commercial fishing gears included are otter trawl (Chapter 4 by Winger, Eayrs, and Glass), longline (Chapter 5 by Løkkeborg, Fernö, and Humborstad), fish pot (Chapter 6 by Thomsen, Humborstad, and Furevik), large-scale fish traps and setnets (Chapter 7 by He and Inoue), gillnet (Chapter 8 by He and Pol), and the theory and practice of fish catching using electricity (Chapter 9 by Polet). Each chapter starts with a review of the fishery and the fishing gear and describes fishing processes and fish behav-

ior near the fishing gear and then concludes with conservation issues and potential solutions related to the fishing gear type. The third and last part of the book illustrates specific conservation challenges and solutions across fishing gear types and fisheries. The conservation issues reviewed include bycatch and discard in trawl fisheries (Chapter 10 by Graham), mortality of discards and escapees (Chapter 11 by Suuronen and Erickson), seabed effects of trawling (Chapter 12 by He and Winger), and fishery interactions with megafauna species (Chapter 13 by Rihan). We have made great progress during the past 50 years in understanding fish behavior near fishing gears and in applying this knowledge in conservation-oriented fishing gear designs and operations. Yet there are still many things we do not know and we are still facing many challenges. As we enter the second decade of this century, conservation in marine fisheries will be placed in an even more prominent position, as greater attention is focused on the vast ocean, resulting in more rules and regulations to protect and preserve fisheries, ecosystems, and protected species. Further understanding of the behavior of fish and associated animals will be needed to achieve goals of conservation and sustainable utilization of marine fishery resources. This book is intended as a reference for fishery researchers, students, managers, and conservation enthusiasts. It can also be used as a textbook for fishery courses at undergraduate and graduate levels. I am grateful to Dr. Clem Wardle, the worldrenowned fish behaviorist and retired Aberdeen Marine Laboratory researcher who wrote the preface for the book. His insight on the subject is valuable even 10 years after his retirement. I would like to thank all authors for their commitments to writing chapters and their timely completion of the manuscript. I would like particularly to thank the following reviewers for their critical reviews and helpful suggestions: Michael Davis (USA), Steve Eayrs (USA), Michael Fine (USA), Daniel Foster (USA), Emma Jones (New Zealand), Sven-Gunnar Lunnaryd (Sweden), Bob van Marlen (the Netherlands), Henry Milliken II (USA), Barry O’Neill (UK), Michael Pol (USA), Craig Rose (USA), Al Stoner (USA), Anna-Liisa Toivonen

Introduction (Finland), and Timothy Werner (USA). Persistent understanding and encouragement of my commissioning editor at Blackwell Sciences, Justin Jeffryes, are fully appreciated. Last but not least, I am grateful to my wife, Miao, and daughters, Fiona and Joanna, for putting up with

xv

me over the past 3 years while I was writing chapters and editing the book. It would not have been possible without their support and understanding. Pingguo He Raynham, MA, USA

Behavior of Marine Fishes Capture Processes and Conservation Challenges

Part One Locomotion and Sensory Capabilities in Marine Fish

Chapter 1 Swimming in Marine Fish John J. Videler and Pingguo He

1.1 INTRODUCTION After 500 million years of natural selection, fish are extremely well adapted to various constraints set by the aquatic environment in which they live. In the dense fluid medium, they are usually neutrally buoyant and use body movements to induce reactive forces from the water to propel themselves. Animal movements are powered by contracting muscles, and these movements consume energy. The basic principles of fish locomotion are used by approximately 25,000 extant species. The variation in swimming styles, within the limits of these principles, is great. Swimming includes steady swimming at various speeds, accelerating, braking, maneuvering, jumping, diving downward, and swimming upward. Swimming behavior is different for every species and changes in each individual during growth from larva to adult. Speed, agility, and endurance maxima determine the chances for survival to a considerable extent. Peak performance in absolute terms is positively related to temperature and body length. Ultimately, performance affects the evolutionary fitness of each individual and is a significant factor that is directly related to capture by or escape from both active (e.g., trawls and seines) and passive (e.g., gillnet, longlines, and traps) fishing gear. This chapter reviews the current knowledge about fish swimming mechanisms and abilities to provide a background for discussions in later chapters.

fish into two lateral halves, and lateral longitudinal muscles that are segmentally arranged in blocks, or myotomes. The vertebral column is laterally highly flexible and virtually incompressible longitudinally. Consequently, contraction of the muscles on one side of the body bends the fish, and waves of curvature along the body can be generated by series of alternating contractions on the left and the right side (Videler and Wardle 1991). Fish vertebrae are concave fore and aft (amphicoelous) and fitted with a neural arch and spine on the dorsal side. In the abdominal region, lateral projections are connected with the ribs enclosing the abdominal cavity. The vertebrae in the caudal region bear a hemal arch and spine. Neural and hemal spines point obliquely backward. The number of vertebrae varies greatly among species—European eels (Anguilla anguilla) have 114 vertebrae, and the numbers in the large order of Perciformes vary between 23 and 40 (Ford 1937). The number is not necessarily constant within a species. Atlantic herring (Clupea harengus), for example, can have between 54 and 58 vertebrae (Harder 1975a, 1975b). The end of the vertebral column is commonly adapted to accommodate the attachment of the tail fin. Several vertebrae and their arches and spines are partly rudimentary and have changed shape to contribute to the formation of platelike structures providing support for the fin rays of the caudal fin. Most fish species have unpaired dorsal, caudal, and anal fins and paired pectoral and pelvic fins. Each fin is powered by intrinsic musculature. The lateral muscles are usually the main target of the fishing industry. Relatively short lateral muscle

1.2 THE SWIMMING APPARATUS Fish are aquatic vertebrates with a skull, a vertebral column supporting a medial septum that divides the

5

6

Locomotion and Sensory Capabilities in Marine Fish

Figure 1.1. (A) The myotomes and myosepts on the left side of the king salmon. Myotomes have been removed at four places to reveal the complex three-dimensional configuration of the lateral muscles. [Redrawn after Greene and Greene (1913).] (B) A cross section through the upper left quarter of the caudal region of a salmon. Red muscle fibers are situated in the dark area near the outside. The lines represent the myosepts between the complex myotomes. [Based on Shann (1914).]

fibers are packed into myotomes between sheets of collagenous myosepts. The myotomes are cone shaped and stacked in a segmental arrangement on both sides of the median septum (Fig. 1.1A). In cross sections through the caudal region, the muscles are arranged in four compartments. On each side is a dorsal and a ventral compartment; in some groups, they are separated by a horizontal septum. The left and right halves and the dorsal and ventral moieties are mirror images of each other. In cross sections, the myosepts are visible as more or less concentric circles of collagen. The color of the muscle fibers may be red, white, or intermediate in different locations in the myotomes (Fig. 1.1B), which was first described by Lorenzini in 1678 (Bone 1966). Red fibers are usually situated directly under the skin. The deeper white fibers form the bulk of lateral muscles, and in some species intermediately colored pink fibers are found between the two. The red fibers are slow but virtually inexhaust-

ible and their metabolism is aerobic. They react to a single stimulus owing to the high density of nerve terminals on the fibers. The white fibers are fast, exhaust quickly, and use anaerobic metabolic pathways. White fibers are either focally or multiply innervated. Pink fibers are intermediate in most aspects. The red muscles of some large tuna and shark species are positioned well inside the white muscle mass, an arrangement that can increase the muscle temperature by as much as 10°C during swimming (Carey and Teal 1969). The final paragraph describes how this halves the twitch contraction time of the white muscles and doubles the maximum swimming speed. Fish fins consist of two layers of skin, usually supported by fin rays that are connected to supporting skeletal elements inside the main body of the fish. Intrinsic fin muscles find their origin usually on the supporting skeleton and insert on the fin rays. The fins of elasmobranchs (sharks and rays) are

Swimming in Marine Fish

7

Figure 1.2. The structure of a typical teleost fin ray. (A) Dorsal or ventral view: the left and right fin ray halves are each other’s mirror image. (B) Lateral view; the size of the bony elements decreases to the right after each of the bifurcations. The position of the bifurcations in the various branches does not show a geometrically regular pattern. (C) Longitudinal section through the bony elements of the fin ray at a position indicated in (A). Note that a joint with densely packed collagenous fibers connects the elements. The collagenous fibers connecting the fin ray halves have a curly, serpentine appearance. [From Videler (1993).]

permanently extended and rather rigid compared with those of teleosts (bony fish). Elasmobranch fin rays consist of rows of longitudinally connected small pieces of cartilage in a juxtaposed arrangement. Intrinsic muscles on both sides of the rows running from the fin base to the edge bend these fins. Teleost fins can be spread, closed, and folded against the body. There are two kinds of teleost fin rays: spiny, stiff unsegmented rays and flexible segmented ones. Spiny rays stiffen the fin and are commonly used for defense. The flexible rays (Fig. 1.2) play an important role in adjusting the stiffness and camber of the fins during locomotion. They consist of mirror-image halves, each of which has a skeleton of bony elements interconnected by collagenous fibers. Muscles pulling harder on the fin ray head on one side will shift the two halves with respect to each other and bend the ray or increase the stiffness against bending forces from the water. In contrast to elasmobranch fins, there are no muscles on the fin itself. The body shape of fish may vary greatly among species, but the best pelagic swimmers have a common form. Their bodies are streamlined, with gradually increasing thickness from the point of the snout to the thickest part at about one-third of the length. From that point, the thickness gradually decreases toward the narrow caudal peduncle. A

moving body in water encounters friction and pressure drag. Friction drag is proportional to the surface area, and pressure drag is proportional to the area of the largest cross section. A spherical body has the lowest friction for a given volume; a needle-shaped body encounters minimal amounts of pressure drag. An optimally streamlined body is a hybrid between a sphere and a needle and offers the smallest total drag for the largest volume. It has a diameter-to-length ratio between 0.22 and 0.24. The best pelagic swimmers have near-optimal thickness-to-length ratios (Hertel 1966). The mechanically important part of fish skin is the tissue (the stratum compactum) underneath the scales, which consists of layers of parallel collagenous fibers (see Videler 1993 for a review). The fibers in adjacent layers are oriented in different directions, and the angles between the layers vary between 50 and 90 degrees, but the direction in every second layer is the same. The packing of layers resembles the structure of plywood, except that in the fish stratum compactum there are also radial bundles of collagen connecting the layers; the number of layers varies between 10 and 50. In each layer, the fibers follow left- and right-handed helices over the body surface. The angle between the fibers and the longitudinal axis of the fish decreases toward the tail. In some species, the

8

Locomotion and Sensory Capabilities in Marine Fish

stratum compactum is firmly connected to the myosepts in the zone occupied by the red muscle fibers; in other fish, there is no such connection. The strongest fish skins that have been tested are those of eel and shark. Values of Young’s modulus (the force per unit cross-sectional area that would be required to double the length) of up to 0.43 GPa (1 GPa = 109 N/m2) have been measured. This is about one-third of the strength of mammalian tendon, for which values of 1.5 GPa have been measured. Scales are usually found at the interface between fish and water. Several swimming-related functions have been suggested. Scales might serve to prevent transverse folds on the sides of strongly undulating fish, keeping the outer surface smooth. Spines, dents, and tubercles on scales are usually arranged to form grooves in a direction of the flow along the fish. Roughness due to microstructures on scales in general creates small-scale turbulence, which could delay or prevent the development of dragincreasing large-scale turbulence (Aleyev 1977). Fish mucus covering the scales is supposed to reduce friction with the water during swimming. This assumption is based on the idea that mucus shows the “Toms effect” (Lumley 1969), which implies that small amounts of polymers are released that preclude sudden pressure drops in the passing fluid. Measurements of the effects of fish mucus on the flow show contradictory results varying from a drag reduction of almost 66% (Pacific barracuda, Sphyraena argentea) to no effect at all (Pacific bonito, Sarda chiliensis) (Rosen and Cornford 1971). Experiments with rainbow trout showed that mucus increases the thickness of the boundary layer

(Daniel 1981). The layer of water around the fish is affected by the presence of the fish during swimming. Viscosity causes a layer of water close to the fish to travel with it at the speed of the fish. There is a gradient of decreasing water velocities in a direction away from the surface of the fish. The thickness of the boundary layer is defined as the distance from the surface of the fish to where the water is no longer dragged along. A thick boundary layer implies that the gradient is gradual, which reduces viscous friction. However, the penalty for a thicker boundary layer is that the fish has to drag along a larger amount of water. The conclusion might be that the effect of mucus is beneficial during slow-speed cruising but detrimental during fast swimming and acceleration (Videler 1995). 1.3 SWIMMING-RELATED ADAPTATIONS Some fish species are adapted to perform some aspect of locomotion extremely well, whereas others have a more general ability to move about and are specialized for different traits not related to swimming. Generalists can be expected to have bodies that give them moderately good performance in various special functions. Specialists perform exceptionally well in particular skills. Fast accelerating, braking, high-speed cruising, and complex maneuvering are obvious examples. The special swimming adaptations shown in Figure 1.3 are only a few of a wealth of possible examples. A closer study of the swimming habits of a large number of species will show many more specialist groups than the dozen or so described here (e.g., Lindsey 1978).

Figure 1.3. A representation of fish swimming specializations. See text for detailed explanation. (a) pike (Esox lucius), barracuda (Sphyraena argentea); (b) forkbeard (Phycis phycis), saithe (Pollachius virens); (c) bluefin (Thunnus thynnus), porbeagle shark (Lamna nasus); (d) angelfish (Pterophyllum scalare), butterfly fish (Chaetodon sp.); (e) sunfish (Mola mola), opah (Lampris guttatus), louvar (Luvarus imperialis); (f) swordfish (Xiphias gladius), sailfish (Istiophorus platypterus); (g) turbot (Scophthalmus maximus), skate (Raja batis); (h) moray eel (Muraena Helena), eel (Anguilla anguilla); (i) rainbow wrasse (Coris julis), sand eel (Ammodytes tobianus); (j) sea-horse (Hippocampus ramulosus), Atlantic flying fish (Cheilopogon heterurus), hatchet fish (Gasteropelecus sp.); cornet fish (Fistsularia sp.). [Modified from Videler (1993).]

h a

Pikc

Moray

Barracuda

Eel b

Forkbeard

Saithe

i c

Bluefin tuna

Porbeagle Rainbow wrasse

Sandeel

Anglefish d

j Butterflyfish Sea-horse

e Opah

Sunfish Louvar

e

k

Atlantic flying fish

f

Swordfish

Sailfish

Hatched fish

g

Turbot

Skate

l Cornet fish

9

10

Locomotion and Sensory Capabilities in Marine Fish

Specialists in accelerating, such as the pike (Esox lucius) and the barracuda (Fig. 1.3a), are often ambush predators. They remain stationary or swim very slowly until a potential prey is within striking distance. These species have a reasonably streamlined body and large dorsal and anal fins positioned extremely rearward, close to the caudal fin. Acceleration during the strike is achieved by the first two beats of the tail, the effect is enlarged by the rearward position of the dorsal and anal fins. The relative skin mass of the pike is reduced, compared with that of other fish, increasing the relative amount of muscles and decreasing the dead mass that has to be accelerated with the fish at each strike. Maximum acceleration rates measured for pike vary between 40 and 150 m/s2, which is 4 to 15 times the acceleration due to gravity (G = 9.8 m/s2). The highest peak acceleration value reported for pike is 25 times gravity (Harper and Blake 1990, 1991). Braking is difficult while moving in a fluid medium. Fish use the unpaired fins and tail, usually in combination with the pectoral and pelvic fins, for braking. Gadoids with multiple or long unpaired fins are good at braking. Forkbeard (Phycis phycis) swims fast and close to the bottom with elongated pelvic fins extended laterally for the detection of bottom-dwelling shrimp. The fish instantly spreads the long dorsal and anal fins and throws its body into an S-shape when a prey item is touched. In the process, the fin rays of the tail fin are actively bent forward. Braking is so effective that the shrimp has not yet reached the caudal peduncle before the fish has stopped and turned to catch it. The highest deceleration rate measured is 8.7 m/s2 for saithe (Pollachius virens) (Geerlink 1987). The contribution of the pectorals to the braking force is about 30%; the rest comes from the curved body and extended median fins (Geerlink 1987) (Fig. 1.3b). Cruising specialists (Fig. 1.3c) migrate over long distances, swimming continuously at a fair speed. Many are found among scombrids and pelagic sharks, for example. Cruisers have highly streamlined bodies, narrow caudal peduncles with lateral keels, and high-aspect-ratio tails (aspect ratio is the tail height squared divided by tail surface area). The bluefin tuna (Thunnus thynnus), for example, is an extreme long distance swimmer, crossing the

Atlantic twice a year. The body dimensions are very close to the optimum values, with a thickness-tolength ratio near 0.25 (Hertel 1966). Cruising speeds of 3-m-long bluefin tunas measured in large enclosures reached 1.2 L/s (260 km/d), where L equals body length (Wardle et al. 1989). Angelfish (Pterophyllum scalare) and butterfly fish (Chaetodon sp.) (Fig. 1.3d) are maneuvering experts with short bodies and high dorsal and anal fins. Species of this guild live in spatially complex environments. Coral reefs and freshwater systems with dense vegetation require precise maneuvers at low speed. Short, high bodies make very short turning circles possible. Angelfish make turns with a radius of 0.065 L (Domenici and Blake 1991). For comparison, the turning radius of a cruising specialist is in the order of 0.5 L, an order of magnitude larger. Sunfish (Mola mola), opah (Lampris guttatus), and louvar (Luvarus imperialis) are among the most peculiar fish in the ocean (Fig. 1.3e). They look very different from each other but they all have large body sizes. The sunfish reaches 4 m and 1500 kg, the opah may weigh up to 270 kg, and the louvar is relatively small with a maximum length of 1.9 m and weight of 140 kg. While little is known about the mechanics of their locomotion, they all seem to swim slowly over large distances. The opah uses its wing-shaped pectorals predominantly, and the louvar has a narrow caudal peduncle and an elegant high-aspect-ratio tail similar to those of the tunas. The sunfish has no proper tail, but the dorsal and ventral fins together form an extremely highaspect-ratio propeller. Sunfish swim very steadily, moving the dorsal and ventral fins simultaneously to the left side and, half a cycle later, to the right side. The dorsal and ventral fins have a cambered wing profile in cross section with a rounded leading edge and a sharp trailing edge. The intrinsic fin muscles fill the main part of the body and insert on separate fin rays, enabling the sunfish to control the movements, camber, and profile of its fins with great precision. Although there are no measurements to prove this as yet, it as appears that these heavy species specialize in slow steady swimming at low cost. Inertia helps them to keep up a uniform speed, while their well-designed propulsive fins generate just enough thrust to balance the drag as efficiently as possible.

Swimming in Marine Fish Swordfishes (Xiphiidae) and billfishes (Istiophoridae) (Fig. 1.3f) show bodily features unique to these fish—the extensions of the upper jaws, the swords, and the shape of the head. They are probably able to swim briefly at speeds exceeding those of all other nektonic animals, reaching values of well over 100 km/h (Barsukov 1960). The sword of swordfish is dorsoventrally flattened to form a long blade of up to 45% of the body length. The swords of billfish (including sailfish, spearfish, and marlin) are pointed spikes, round in cross section and shorter (between 14% and 30% in adult fish, depending on species) than those of swordfish. All swords have a rough surface, especially close to the point. The roughness decreases toward the head. One other unique bodily feature of the swordbearing fishes is the concave head. At the base of the sword, the thickness of the body increases rapidly with a hollow profile up to the point of greatest thickness of the body. The rough surface on the sword reduces the thickness of the boundary layer of water dragged along with the fish (Hertel 1966). This reduces drag. The concave head probably serves to avoid drag-enhancing large-scale turbulence. The caudal peduncle is dorsoventrally flattened, fitted with keels on both sides. These features and the extremely high-aspect-ratio tail blades with rearward-curved leading edges are hallmarks of very fast swimmers. The shape of the body of flatfish (Pleureonectiformes) and rays (Rajiformes) (Fig. 1.3g) offers the opportunity of hiding in the boundary layer of the seabed where speeds of currents are reduced. There is another possible advantage connected with a flat body shape. Both flatfish and rays can be observed swimming close to the bottom. These fish are negatively buoyant and, like flying animals, must generate lift (a downwash in the flow) at the cost of induced drag to remain “waterborne.” Swimming close to the ground could reduce the drag due to lift generation considerably, depending on the ratio between height off the ground and the span of the “wings” (Anderson and Eberhardt 2001). Only a few species can swim both forward and backward. Eels (Anguillidae), moray eels (Mutaenidae), and conger eels (Congeridae) (Fig. 1.3h) can quickly reverse the direction of the propulsive wave on the body and swim backward. The

11

common feature of these fish is the extremely elongated flexible body without a high tail fin. Swimming is usually not very fast; they prefer to swim close to the bottom and operate in muddy or maze-type environments. Rainbow wrasses (Coris julis) and sand eels (Ammodytes tobianus) (Fig. 1.3i) sleep under a layer of sand—sand eels in daytime and rainbow wrasse during the night. Both species swim headdown into the sand using high-frequency lowamplitude oscillations of the tail. If the layer of sand is thick enough, the speed is not noticeably reduced. Body shapes are similar—that is, slender with a well-developed tail. The wrasses use their pectoral fins for routine swimming and move body and tail fin during escapes and to swim into the sand (Videler 1988). Most neutrally buoyant fish species are capable of hovering in one spot in the water column. Some species can hardly do anything else. Seahorses (Fig. 1.3j) and pipefishes (Syngnathidae) rely on camouflage for protection from predators. They are capable of minute adjustments of the orientation of their body using high-frequency, low-amplitude movements of the pectoral and dorsal fins. Seahorses are the only fish with a prehensile tail. Flying fish (Exocoetidae) (Fig. 1.3k) have exceptionally large pectoral fins to make gliding flights out of the water when chased by predators. Some species are four-winged because they use enlarged pelvic fins as well. The lower lobe of the caudal fin is elongated and remains beating the water during takeoff. Hatchet fishes (Gasteropelecus sp.) actually beat their pectoral fins in powered flight. The pectoral fins have extremely large intrinsic muscles originating on a greatly expanded pectoral girdle. Many additional species occasionally or regularly leap out of the water but they are not specially adapted to fly. Cornet fishes (Fistularia sp.) (Fig. 1.3l) are predators of small fish in the littoral of tropical seas, most often seen above sea grass beds or sandy patches between coral reefs. They seem to have two tails; the first one is formed by a dorsal and anal fin, and the second one is the real tail. Beyond the tail is a long, thin caudal filament. These fish hunt by dashing forward in one straight line without any side movements of the head, using large-amplitude

12

Locomotion and Sensory Capabilities in Marine Fish

strokes of the two tail fins and the trailing filament (J.J.V., personal observations). It looks as though the double–tail fin configuration with the trailing filament serves to allow fast acceleration without recoil movements of the head. Precise kinematic measurements are needed to provide evidence for this assumption. 1.4 STYLES OF SWIMMING Most fish species swim with lateral body undulations running from head to tail; a minority use the movements of appendages to propel themselves. The waves of curvature on the bodies of undulatory swimmers are caused by waves of muscle activations running toward the tail with a 180-degree phase shift between the left and the right side (Videler and Wardle 1991). The muscular waves run faster than the waves of curvature, reflecting the interaction between the fish’s body and the reactive forces from the water. The swimming speed varies between 0.5 and 0.9 times the backward speed of the waves of curvature during steady swimming (Videler and Hess 1984). The wavelength of the body curvature of slender eel-like fish is about 0.6 L, indicating that there is more than one wave on the body at any time. Fast-swimming fish such as mackerel and saithe have almost exactly one complete wave on the body, and on short-bodied fish such as carp and scup, there is less than one wave on the length of the body during steady swimming. The maximum amplitude (defined as half the total lateral excursion) may increase toward the tail linearly, as in eels and lampreys, or according to a power function in other species (Wardle et al. 1995). The increase in maximum amplitude is concentrated in the rearmost part of the body in fast fish like tuna. The maximum amplitude at the tail is usually in the order of 0.1 L with considerable variation around that value. The period of the waves of curvature determines the tail beat frequency, which is normally linearly related to the swimming speed (Bainbridge 1958). The distance covered per tail beat is the “stride length” of a fish. It varies greatly between species but also for each individual fish. Maximum values of more than one body length have been measured for mackerel; the least distance covered per beat of the same individual was reported to be 0.7 L (Videler and Hess 1984; Wardle and He

1988). Many species reach values between 0.5 and 0.6 L during steady swimming bouts. Left–right undulations of the body from head to tail are used by fish varying in body shape from eel to tuna. However, the amplitudes of the undulations and the way in which they are generated greatly differ among species. In eel-like fishes, there is a fairly tight connection between the position of the backward-moving bend and muscle shortening on the concave side of the bend. One-sided waves of muscle contraction run from head to tail, causing the local bending of the body with a slight delay. There is always more than one complete wavelength present on the body of these fish between the head and tail (Wardle et al. 1995). Mackerel-type fishes, on the other hand, have only one wavelength on the body at all times, and the lateral muscles on one side of the body from head to tail are activated simultaneously (Wardle and Videler 1993). Simultaneous muscle activity would result in a C-shaped lateral bending of the body if it took place outside the water. However, during swimming, water reacts more strongly to sideward movements of the high tail than to the movements of the rounded side of the rest of the body. Lateral muscles in the anterior part of the body actually shorten when activated, but muscles in the rear part and especially in the caudal peduncle are active when they are being stretched by the sideward push of the water against the tail. Muscles there produce negative work by resisting being stretched. This type of muscle activity provides higher forces and power per unit crosssectional area. Therefore, the rear part of these fishes can be narrower (to contribute to a streamlined body) and still transmit the high forces from the anterior muscle mass to the tail blade (Wardle and Videler 1993; Wardle et al. 1995). Fish locomotion using paired and median fins was reviewed by Blake (1983). Swimming with appendages includes pectoral fin swimming and median fin propulsion. Pectoral fin movements of, for example, labrids (Labridae), shiner perches (Cymatogaster aggregata), and surfperches (Embiotocidae) make an elegant impression. The beat cycle usually consists of three phases. During the abduction phase, the dorsal rays lead the movement away from the body and downward. The adduction phase brings the fin back to the body surface led by

Swimming in Marine Fish horizontal movement of the dorsal rays. During the third phase, the dorsal rays rotate close to the body back to their initial position. Stride lengths vary with speed and may reach more than one body length at optimal speeds. Undulations of long dorsal and anal fins can propel fish forward and backward and are used in combination with movements of the pectoral fins and the tail. There is usually more than one wave on each fin (e.g., up to 2.5 waves on the long dorsal fin of the African electric eel, Gymnarchus niloticus). 1.5 INTERACTIONS BETWEEN FISH AND WATER: FISH WAKES Every action of the fins or the body of a fish will, according to Newton’s third law, result in an equal

13

but opposite reaction from the surrounding water. A swimming fish produces forces in interaction with the water by changing water velocities locally. The velocity gradients induce vortices, being rotational movements of the fluid. Vortices either may end at the boundary of the fluid or may form closed loops or vortex rings with a jet of water through the center (Videler et al. 2002). Furthermore, vortex rings can merge to form chains (Fig. 1.4). Quantitative flow visualization techniques have been successfully applied to reveal the flow patterns near fish using body undulations to propel themselves (Müller et al. 1997). The interaction between undulating bodies and moving fins and the water results in complex flow patterns along and behind the swimming animals (Videler 1993). A schematic

Figure 1.4. Schematic drawings of vortex ring structures. (A) A single vortex ring. The ring shaped centre of rotation is drawn as a line. The rotations of the vortex ring structure draw a jet of water through the centre of the ring, indicated by the arrow. (B) A 3-D reconstruction of a chain of three connected vortex rings. A resulting jet of fluid undulates through the centre of the vortex rings building the chain. [From Videler et al. (2002).]

14

Locomotion and Sensory Capabilities in Marine Fish

Figure 1.5. Artist’s impression of the flow behind a steadily swimming saithe. The tail blade is moving to the left and in the middle of the stroke. At the end of each half-stroke a column vortex is left behind when the tail blade changes direction. Tail tip vortices are shed dorsally and ventrally when the tail moves from side to side. Together the vortices form a chain of vortex rings (as shown schematically in Fig. 1.4) with a jet of water winding through the centers of the rings in the opposite swimming direction. [From Videler (1993).]

three-dimensional impression of the wake generated by the tail behind a steadily swimming fish is shown in Figure 1.5. This shows the dorsal and ventral tail tip vortices generated during the tail beat as well as the vertical stop–start vortices left behind by the trailing edge of the tail at the end of each half-stroke. During a half-stroke, there is a pressure difference between the leading side of the fin and the trailing side. Dorsal and ventral tip vortices represent the water escaping at the fin tips from the leading side, with high pressure to the trailing side where the pressure is low. At the end of the halfstroke, the tail changes direction and builds up pressure on the opposite side of the fin, leaving the previous pressure difference behind as a vertical vortex column. These vertical, dorsal, and ventral vortex systems form a chain through which a jet of water undulates opposite the swimming direction. If we concentrate on what happens in a mediofrontal plane through the fish and the wake, we expect to see left and right stop–start vortices with an undulating backward jet between them. The rotational sense should be anticlockwise on the right of the fish and clockwise on the left. Visualizations of the flow in the mediofrontal plane of swimming fish

reveal that this picture of the wake is correct (Fig. 1.6) (Müller et al. 1997). Flow patterns around fish using paired and unpaired fins, as well as the tail and those near maneuvering fish, are much more complex systems of jets and vortex rings. Such complex flow patterns have been published for a number of species during recent years (Lauder and Tytell 2006). 1.6 ENERGY REQUIRED FOR SWIMMING Swimming fish use oxygen to burn fuel to power their muscles. Carbohydrates, fat, and proteins are the common substrates. A mixture of these provides about 20 J/ml oxygen used (Videler 1993). Measurements of energy consumption during swimming are mainly based on records of oxygen depletion in a water tunnel respirometer. Respiration increases with swimming speed, body mass, and temperature and varies considerably among species. The highest levels of energy consumption measured in fish are about 4 W/kg (Videler 1993). Fast, streamlined fish can increase their metabolic rates up to 10 times resting levels during swimming at the highest sustainable speeds. Short-burst speeds

Swimming in Marine Fish

15

Figure 1.6. The wake of a continuously swimming mullet. The arrows represent the flow velocity in mm/s scaled relative to the 10 mm/s bar on the bottom. The shaded circles indicate the centers of the column vortices. The picture represents a horizontal cross-section half way down the tail through the wake drawn in Figure 1.5. [Based on Müller et al. (1997).]

powered by anaerobic white muscles can cost as much as 100 times resting rates. Most of the energy during swimming at a constant speed is required to generate sufficient thrust to overcome drag. The drag on a steadily swimming fish is proportional to the square of the swimming speed—the energy required increases as the cube of the speed. In other words, if a fish wants to swim twice as fast, it will have to overcome four times as much drag and use eight times as much energy. A fair comparison of the energy used requires standardization of the speed at which the comparison is made. The energetic cost of swimming is the sum of the resting or standard metabolic rate and the energy required to produce thrust. Expressed in watts (joules per second), it increases as a J-shaped curve with speed in m/s (Fig. 1.7) (Videler 1993). The exact shape of the curve depends mainly on the species, size, temperature, and condition of the fish. Owing to the shape of the curve, there is one optimum speed at which the ratio of metabolic rate over speed reaches a minimum. This ratio represents the amount of work a fish has to do to cover 1 m (J/s divided by m/s). To make fair comparisons possible, the

optimum speed (Uopt), where the amount of energy used per unit distance covered is at a minimum, is used as a benchmark. Series of measurements of oxygen consumption at a range of speeds provide the parameters needed to calculate Uopt and the energy used at that speed. The energy values are normalized by dividing the active metabolic rate at Uopt (in W = J/s = Nm/s) by the weight of the fish (in N) multiplied by Uopt (in m/s), to reach a dimensionless number for the cost of transport. Hence, COT represents the cost to transport one unit of weight over one unit of distance at Uopt. Available data show that Uopt is positively correlated with mass (M) and proportional to M0.17. The value of Uopt decreases, however, with mass if Uopt is expressed in L/s and is proportional to M−0.14. While there is great variation in measured Uopt, 2 L/s can serve as a reasonable first estimate of the optimum speed in fish. At Uopt the COT values are negatively correlated with body mass (M−0.38) (Fig. 1.8). Fish use an average of 0.07 J/N to swim their body length at Uopt. If the weight and the size of the animals are taken into account as well by calculating the energy needed to transport the body weight

16

Locomotion and Sensory Capabilities in Marine Fish

Figure 1.7. A theoretical curve of the rate of work as a function of swimming speed. SMR is the standard or resting metabolic rate at speed 0. The amount of work per unit distance covered (J/m) is at a minimum at Uopt. [From Videler (1993).]

over the length, the amount of energy used to swim at Uopt increases in proportion to M0.93 (Fig. 1.9) (Videler 1993). Burst-and-coast (or kick-and-glide) swimming behavior is commonly used by several species (Weihs 1974; Videler and Weihs 1982). It consists of cyclic bursts of swimming movements followed by a coast phase in which the body is kept motionless and straight. The velocity curve in Figure 1.10 shows how the burst phase starts off at an initial velocity (Ui) lower than the average velocity (Uc). During a burst, the fish accelerates to a final velocity (Uf), higher than Uc. The cycle is completed when velocity Ui is reached at the end of the deceleration during the coast phase. Energy savings in the order of 50% are predicted if burst-and-coast swimming is used during slow and high swimming speeds instead of steady swimming at the same average speed (Videler and Weihs 1982). The model predictions are based on the assumption that there is a three-fold difference in drag between a rigid gliding body and an actively moving fish.

COT

102

1

Undulatory P t l Pectoral Tuna Shark Salmon Larvae

10-2 10-4

10-2

1

Body mass (kg) Figure 1.8. Doubly logarithmic plot of dimensionless COT, being the energy needed to transport one unit of mass over one unit of distance during swimming at Uopt, related to body mass. The connected points indicate series of measurements of animal groups indicated separately; “undulatory” and “pectoral” refer to measurements of fish using body plus tail and pectoral fins respectively, for propulsion. [From Videler (1993).]

Swimming in Marine Fish

17

Figure 1.9. Doubly logarithmic plot of the energy needed by a swimming fish to transport its body weight over its body length as a function of body mass. Symbols as in Figure 1.7. [Based on Videler (1993).]

Figure 1.10. Part of a velocity curve during burst-and-coast swimming of cod. The average speed Uc was 3.2 L/s. The initial speed and the final velocity of the acceleration phase are indicated as Ui and Uf, respectively. [From Videler (1993).]

Schooling behavior probably has energy-saving effects (Weihs 1973). As seen in Figure 1.5, the wake of a steadily swimming fish shows an undulating jet of water in the opposite swimming direction through a chain of vortex rings. Just outside this system, water will move in the swimming direction. Theoretically, following fish could make use of this

forward component to facilitate their propulsive efforts (Weihs 1973). One would expect fish in a school to swim in a distinct three-dimensional spatial configuration in which bearing and distance among school members showed a distinct constant diamond lattice pattern and a fixed phase relationship among tail beat frequencies. This has not,

18

Locomotion and Sensory Capabilities in Marine Fish

however, been confirmed by actual observations. On the other hand, energetic benefits for school members have been confirmed by indirect evidence. It has been observed that the tail beat frequency of schooling Pacific mackerel (Scomber japonicus) is reduced compared with solitary mackerel swimming at the same speed (Fields 1990). In schools of sea bass, trailing individuals used 9% to 14% lower tail beat frequencies than fish in leading position. There is also some evidence showing that fast swimming fish in a school use less oxygen than the same number of individuals would use in total in solitary swimming at the same speed (Herskin and Steffensen 1998). 1.7 SWIMMING SPEEDS AND ENDURANCE The relationship between swimming speed and endurance is not straightforward due to the separate use of red, intermediate, and white muscle. Virtually inexhaustible red muscles drive slow cruising speeds; burst speeds require all-out contraction of

white muscles lasting only a few seconds. Endurance decreases rapidly when speeds are above cruising speeds. Maximum swimming speeds of fish are ecologically important for obvious reasons. However, slower swimming speeds and the stamina at these speeds represent equally important survival values for a fish. Figure 1.11 relates swimming speed, endurance, and the cost of swimming for a 0.18-m sockeye salmon (Oncorhynchus nerka) at 15°C (Videler, 1993). At low speeds, this fish can swim continuously without showing any signs of fatigue. The optimum speed Uopt is between 1 and 2 L/s. Limited endurance can be measured at speeds higher than the maximum sustained speed (Ums) in this case, somewhat less than 3 L/s. For these prolonged speeds, the logarithm of the time to fatigue (endurance) decreases linearly with increasing velocity up to the maximum prolonged speed (Ump) where the endurance is reduced to a fraction of a minute. Along this endurance trajectory, the fish will switch gradually from partly aerobic to totally anaerobic metabolism. The maximum burst

Figure 1.11. The metabolic rate (linear scale) and the endurance (logarithmic scale) of a 0.18-m, 0.05-kg sockeye salmon as functions of swimming speed in L per second. The water temperature was 15°C. The optimum swimming speed (Uopt), the maximum sustained speed (Ums), the maximum prolonged speed (Ump) and an estimate of the maximum burst speed (Umax) are indicated. [From Videler (1993).]

Swimming in Marine Fish speed in this case is in the order of 7 L/s for sockeye salmon (Brett 1964). A comparison of published data for some marine species reveals that values for Ums for fish varying in size between 10 and 49 cm are between 0.9 and 9.9 L/s, with larger fish achieving less Ums in L/s (Table 1.1). For example, 17-cm-long haddock (Melanogrammus aeglefinus) are capable of swimming at 2.6 L/s for longer than 200 min, but 41-cmlong haddock can only swim at 1.5 L/s for the same period of time (Breen et al. 2004). Bottom-dwelling demersal fish living in complex environments usually have shallower endurance curves than pelagic long-distance swimmers, which fatigue more quickly when they break the limit of the maximum sustained speed (Videler 1993). Endurance in fish swimming at prolonged speeds is limited by the oxygen uptake capacity. Higher speeds cause serious oxygen debts. The maximum burst speed in m/s increases with body length (Fig. 1.12). Average relative values for adult fish are between 10 and 20 L/s. Small fish larvae swim at up to 50 L/s during startle response bursts (Fuiman 1986). Speed record holders in m/s are to be found among the largest fish. Unfortunately, reliable measurements are usually not available. The maximum burst speed of fish depends on the fastest twitch contraction time of the white lateral muscles (Wardle 1975). For each tail beat, the muscles on the right and on the left have to contract once. Hence, the maximum tail beat frequency is the inverse of twice the minimum contraction time. The burst speed is found by multiplying the stride length by the maximum tail beat frequency. Muscle twitch contraction times halve for each 10°C temperature rise, and the burst speed doubles (Videler 1993). Larger fish of the same species have slower white muscles than smaller individuals. The burst swimming speed in L/s decreases with size with a factor of on average 0.89 for each 10 cm length increase. Estimates based on muscle twitch contraction times and measured stride length data for 2.26 m long blue fin tuna vary between 15 and 23 m/s (Wardle et al. 1989). Estimates for 3 m long swordfish exceed 30 m/s (Barsukov 1960). Measured values for burst speeds are difficult to find. The maximum swimming speed in terms of L/s ever recorded in captivity is that of a 30 cm mackerel

19

swimming at 18 L/s (or 5.4 m/s, Wardle and He 1988). At that speed, the tail beat frequency was 18 Hz and the stride length was 1 L.

1.8 CONCLUDING REMARKS Understanding fish swimming performance involves studies of the functional morphology of the swimming apparatus and requires insight in swimmingrelated adaptations. Undulations of body and tail propel the majority of species; others predominantly use movements of paired and unpaired fins. Hydrodynamic interactions between the moving fish and water represent the forces required. Visualization of the flow patterns reveals the vortex systems, pressure distributions, and forces. The energy used to swim is obtained by muscles burning fuel and can be as high as 4 W/kg. The metabolic rate increases exponentially with speed. Fair comparisons among species can be made by looking at the most economic swimming performance at the speed where the energy used per unit weight and unit distance is at a minimum. These dimensionless costs of transport decrease with body mass. The amount of work that needs to be done to transport the weight of a body over its length increases with body mass. Schooling behavior may reduce the costs of transport. Pelagic fish commonly use red lateral muscle during steady swimming at low speeds and white lateral muscle during burst swimming. The stride length of these fish is more or less constant and therefore swimming speed can be predicted from the tail beat frequency which in turn is directly related to the contraction times of the lateral the muscles. Swimming speeds can be classified as sustained speeds, prolonged speeds and burst speeds. Endurance of swimming is reduced at higher swimming speeds during prolonged swimming. Temperature has a profound effect on the swimming capacity with both endurance and swimming speed reduced at lower temperatures. Maximum swimming speeds double for every 10°C increase in temperature. The maximum swimming speed of many marine fish species is between 10 and 20 L/s. Swimming performance affects the evolutionary fitness of a species. For each individual, it is a significant factor directly related to capture by or escape from fishing gears.

Table 1.1. Maximum Sustained Swimming Speed (Ums) and Endurance at Prolonged Speeds of Some Marine Fish Species. Ums (cm/s)

Ums (L/s)

40 49

42 45

1.1 0.9

… …

FT

0.8 0.8

He (1991)

Atlantic cod Gadus morhua

36 36

75 90

2.1 2.5

logE = −0.99 • U + 3.99 logE = −1.13 • U + 4.96

FT

5 8

Beamish (1966)

Atlantic herring Clupea harengu

25

102

4.1

logE = −1.43 • U + 8.37

AT

13.5

He and Wardle (1988)

Atlantic mackerel Scomber scombrus

31

110

3.6

logE = −0.96 • U + 5.45

AT

11.7

He and Wardle (1988)

American shad Alosa sapidissima

42

logE = −1.78 • U + 19.02

FT

Haddock Melanogrammus aeglefinus

17

44

2.6

…

AT

24 31 34 41

53 58 57 60

2.2 1.9 1.7 1.5

… … … …

Jack mackerel Trachurus japonicus

14 21

90 90

6.4 4.3

logE = −7.2 • logU + 9.3

FT

19

Xu (1989)

Japanese mackerel scomber japonicus

10

99

9.9

logE = −0.62 • U + 4.38

FT

19

Beamish (1984)

Redfish Sebastes marinus

17 16 16

52 52 52

3.1 3.3 3.3

logE = −0.25 • U + 1.71 logE = −0.23 • U + 1.70 logE = −0.42 • U + 2.94

FT FT FT

5 8 11

Beamish (1966)

Saithe Pollachius virens

25

88

3.5

logE = −1.63 • U + 5.60

AT

14.4

He and Wardle (1988)

34 42 50

100 106 110

2.9 2.5 2.2

logE = −1.52 • U + 5.91 logE = −1.36 • U + 6.16 logE = −1.17 • U +5.95

Species

Length (cm)

Atlantic cod Gadus morhua

Striped bass Morone saxatilis

42–57

E–U Relation

logE = −0.69 • U + 10.65

Method

FT

T (°C)

Source

Castro-Santos (2005) 9.9

Breen et al. (2004)

Castro-Santos (2005)

Note. T, temperature; FT, flume tank; AT, annular tank; E, endurance (in min); U, swimming speed (in m/s).

20

Swimming in Marine Fish

21

Figure 1.12. Maximum swimming speed of some marine species in relation to their body length. Letter symbols, sources, and temperatures (when available) are as follows: AS— American shad, Alosa sapidissima, Castro-santos (2005); AW—Alewife, Alosa pseudoharengus, Castro-santos (2005); BH—blueback herring, Alosa aestivalis, Castro-santos (2005); CD—cod, Gadus morhua, 9.5° to 12°C, Blaxter and Dickson (1959); FH—flathead, platycephalus bassensis, 20°C, Yanase et al. (2007); HD—haddock, Melanogrammus aeglefinus, 12°C, Wardle (1975); HR—Atlantic herring, Clupea harengus, 10° to 15°C, Misund (1989); JK—jack mackerel, Trachurus japonicus, 23°C, Xu (1989); KW—kawakawa, Euthunnus affinis, 25°C, cited in Beamish (1978); MK—Atlantic mackerel, Scomber scombrus, 12°C, Wardle and He (1988); SB—striped bass, Morone saxatilis, Castro-santos (2005); SE—seabass, Dicentrarchus labrax, 20°C, Nelson and Claireaux (2005); SH1—saithe, Pollachius virens, Videler (1993); SH2— saithe,10.8°C, He (1986); SH3—saithe, 14° to 16°C, Blaxter and Dickson (1959); SK1—skipjack tuna, Katsuwonus pelamis, Walters and Fierstine (1964); SK2—skipjack tuna, cited in Magnuson (1978); SP—sprat, Sprattus sprattus, 12°C, Wardle (1975); WA—wahoo, Acanthocybium solandrei, >15°C, Fierstine and Walters (1968); WH—whiting, Gadus merlangus, 9° to 13°C, Blaxter and Dickson (1959); YT—yellowfin tuna, Thunnus albacares, Fierstine and Walters (1968). Lines for 20 L/s (solid), 10 L/s (dashed), and 5 L/s (dotted) are drawn to indicate swimming speed in L/s. (Modified from He, 1993).

REFERENCES Aleyev YG. 1977. Nekton. The Hague: Dr W. Junk. Anderson DF and Eberhardt S. 2001. Understanding Flight. New York: McGraw-Hill. Bainbridge R. 1958. The speed of swimming as related to size and to the frequency and amplitude of the tail beat. J. Exp. Biol. 35: 109–133. Barsukov VV. 1960. The speed of movement of fishes. Priroda 3: 103–104 (in Russian).

Beamish FWH. 1966. Swimming endurance of some northwest Atlantic fishes. J. Fish. Res. Bd. Can. 23: 341–347. Beamish FWH. 1978. Swimming capacity. In: Hoar WS and Randall DJ (eds). Fish Physiology, Vol. 7. Locomotion. pp 101–187. New York and London: Academic Press. Beamish FWH. 1984. Swimming performance of three southwest Pacific fishes. Mar. Biol. 79: 311–313.

22

Locomotion and Sensory Capabilities in Marine Fish

Blake RW. 1983. Median and paired fin propulsion. In: Webb PW and Weihs D (eds). Fish Biomechanics. pp 214–247. New York: Praeger. Blaxter JHS and Dickson W. 1959. Observations of the swimming speeds of fish. J. Cons. Perm. Int. Explor. Mer. 24: 472–479. Bone Q. 1966. On the function of the two types of myotomal muscle fibre in elasmobranch fish. J. Mar. Biol. Assoc. UK. 46: 321–349. Breen M, Dyson J, O’Neill FG, Jones E and Haigh M. 2004. Swimming endurance of haddock (Melanogrammus aeglefinus L.) at prolonged and sustained swimming speeds, and its role in their capture by towed fishing gears. ICES J. Mar. Sci. 61: 1071– 1079. Brett JR. 1964. The respiratory metabolism and swimming performance of young sockeye salmon. J. Fish. Res. Bd. Can. 21: 1183–1226. Carey FG and Teal JM. 1969. Regulation of body temperature by the bluefin tuna. Comp. Biochem. Physiol. 28: 205–214. Castro-Santos T. 2005. Optimal swim speeds for traversing velocity barrier: an analysis of volitional high-speed swimming behavior of migratory fishes. J. Exp. Biol. 208: 421–432. Daniel TL. 1981. Fish mucus: in situ measurements of polymer drag reduction. Biol. Bull. 160: 376–382. Domenici P and Blake RW. 1991 The kinematics and performance of the escape response in the angelfish (Pterophyllum eimekei). J. Exp. Biol. 156: 187– 204. Fields PA. 1990. Decreased swimming effort in groups of pacific mackerel (Scomber japonicus). Am. Soc. Zool. 30: 134A. Fierstine HL and Walters V. 1968. Studies in locomotion and anatomy of Scombroid fishes. Mem. South. Calif. Acad. Sci. 6: 1–31. Ford E. 1937. Vertebral variation in teleost fishes. J. Mar. Biol. Assoc. UK. 22: 1–60. Fuiman LA. 1986. Burst-swimming performance of larval zebra danios and the effect of diel temperature fluctuations. Trans. Am. Fish. Soc. 115: 143–148. Geerlink PJ. 1987. The role of the pectoral fins in braking in mackerel, cod and saithe. Neth. J. Zool. 37: 81–104. Greene CW and Greene CH. 1913. The skeletal musculature of the king salmon. Bull. U.S. Bur. Fish. 33: 21–60. Harder W. 1975a. Anatomy of Fishes. Part I, Text. Stuttgart: E. Schweizerbart’sche. Harder W. 1975b. Anatomy of Fishes. Part II, Figures and Plates. Stuttgart: E. Schweizerbart’sche.

Harper DG and Blake RW. 1990. Fast-start performance of rainbow trout Salmo gairdneri and northern pike Esox lucius. J. Exp. Biol. 150: 321–342. Harper DG and Blake RW. 1991. Prey capture and the fast-start performance of northern pike Esox lucius. J. Exp. Biol. 155: 175–192. He P. 1986. Swimming Performance of Three Species of Marine Fish and Some Aspects of Swimming in Fishing Gears. PhD thesis. University of Aberdeen, Aberdeen, UK. He P. 1991. Swimming endurance of the Atlantic cod, Gadus morhua L. at low temperatures. Fish. Res. 12: 65–73. He P. 1993. Swimming speeds of marine fish in relation to fishing gears. ICES Mar. Sci. Symp. 196: 183–189. He P and Wardle CS. 1988. Endurance at intermediate swimming speeds of Atlantic mackerel, Scomber scombrus L., herring, Clupea harengus L., and saithe, Pollachius virens L. J. Fish Biol. 33: 255–266. Herskin J and Steffensen JF. 1998. Energy savings in sea bass swimming in a school: measurements of tail beat frequency and oxygen consumption at different swimming speeds. J. Fish Biol. 53: 366–376. Hertel H. 1966. Structure-Form-Movement. New York: Reinhold. Lauder GV and Tytell ED. 2006. Hydrodynamics of undulatory propulsion. In: Shadwick RE and Lauder GV(eds).FishPhysiology,Vol.23.FishBiomechanics. pp 425–468. London: Academic Press. Lindsey CC. 1978. Form, function and locomotory habits in fish. In: Hoar WS and Randall DJ (eds). Fish Physiology, Vol. 7. Locomotion. pp 1–100. New York: Academic Press. Lumley JL. 1969. Drag reduction by additives. Annu. Rev. Fluid Mech. 3: 367–384. Müller UK, van den Heuvel BLE, Stamhuis EJ and Videler JJ. 1997. Fish foot prints: morphology and energetics of the wake behind a continuously swimming mullet (Chelon labrosus Risso). J. Exp. Biol. 200: 2893–2906. Misund OA. 1989. Swimming behavior of herring (Clupea harengus L) and mackerel (Scomber scombrus L) in purse seine capture situations. Proc. World Symp. Fish. Gear and Fish. Vessel Design. pp 541– 546. St. John’s, Newfoundland: Marine Institute. Nelson JA and Claireaux G. 2005. Sprint swimming performance of juvenile European sea bass. Trans. Am. Fish. Soc. 134: 1274–1284. Rosen MW and Cornford NE. 1971. Fluid friction of fish slimes. Nature. 234: 49–51.

Swimming in Marine Fish Shann EW. 1914. On the nature of lateral muscle in teleostei. Proc. Zool. Soc. Lond. 22: 319–337. Videler JJ. 1988. Sleep under sand cover of the labroid fish Coris julis. In: Koella WP, Obál F, Schultz H and Visser P (eds). Sleep ’86. pp 145–147. Stuttgart: Gustav Fischer. Videler JJ. 1993. Fish Swimming. London: Chapman and Hall. Videler JJ. 1995. Body surface adaptations to boundarylayer dynamics. In: Biological Fluid Dynamics. pp 1–20. Cambridge: The Society of Biologists Limited. Videler JJ and Hess F. 1984. Fast continuous swimming of two pelagic predators: saithe (Pollachius virens) and mackerel (Scomber scombrus). A kinematic analysis. J. Exp. Biol. 109: 209–225. Videler JJ, Stamhuis EJ, Müller UK and van Duren LA. 2002. The scaling and structure of aquatic animal wakes. Integr. Comp. Biol. 42: 988–996. Videler JJ and Wardle CS. 1991. Fish swimming stride by stride: speed limits and endurance. Rev. Fish Biol. Fish. 1: 23–40. Videler JJ and Weihs D. 1982. Energetic advantage of burst-and-coast swimming of fish at high speeds. J. Exp. Biol. 97: 169–178. Walters V and Fierstine HL. 1964. Measurements of swimming speeds of yellowfin tuna and wahoo. Nature. 202: 208–209.

23

Wardle CS. 1975. Limit of fish swimming speed. Nature. 255: 725–727. Wardle CS and He P. 1988. Burst swimming speeds of mackerel, Scomber scombrus L. J. Fish Biol. 32: 471–478. Wardle CS, Videler JJ and Altringham JD. 1995. Tuning in to fish swimming waves: body form, swimming mode and muscle function. J. Exp. Biol. 198: 1629–1636. Wardle CS, Videler JJ, Arimoto T, Franco JM and He P. 1989. The muscle twitch and the maximum swimming speed of giant bluefin tuna, Thunnus thynnus L. J. Fish Biol. 35: 129–137. Weihs D. 1973. Hydromechanics of fish schooling. Nature. 245: 48–50. Weihs D. 1974. Energetic advantage of burst swimming of fish. J. Theor. Biol. 48: 215–229. Xu G. 1989. Study on the Fish Swimming Movement and Its Application in Fishing by Trawls. PhD thesis. Tokyo University of Fisheries, Tokyo, Japan (in Japanese with English summary). Yanase K, Eayrs S and Arimoto T. 2007. Influence of water temperature and fish length on the maximum swimming speed of sand flathead, Platycephalus bassensis: implications for trawl selectivity. Fish. Res. 84: 180–188.

24

Locomotion and Sensory Capabilities in Marine Fish

SPECIES MENTIONED IN THE TEXT African electric eel, Gymnarchus niloticus alewife, Alosa pseudoharengus American shad, Alosa sapidissima angelfish, Pterophyllum scalare Atlantic cod, cod, Gadus morhua Atlantic flying fish, Cheilopogon heterurus Atlantic herring, Clupea harengu Atlantic mackerel, Scomber scombrus blueback herring, Alosa aestivalis bluefin tuna, Thunnus thynnus butterfly fish, Chaetodon sp. cornet fishes, Fistsularia sp. European eel, eel, Anguilla anguilla flathead, Platycephalus bassensis forkbeard, Phycis phycis haddock, Melanogrammus aeglefinus hatchet fish, Gasteropelecus sp. jack mackerel, Trachurus japonicus Japanese mackerel, Scomber japonicus kawakawa, Euthunnus affinis louvar, Luvarus imperialis moray eel, Muraena helena opah, Lampris guttatus

Pacific barracuda, barracuda, Sphyraena argentea Pacific bonito, Sarda chiliensis Pacific mackerel, Scomber japonicus pike, Esox lucius porbeagle shark, Lamna nasus rainbow wrasse, Coris julis redfish, Sebastes marinus sailfish, Istiophorus platypterus saithe, Pollachius virens sand eel, Ammodytes tobianus seabass, Dicentrarchus labrax seahorse, Hippocampus ramulosus shiner perch, Cymatogaster aggregata skate, Raja batis skipjack tuna, Katsuwonus pelamis sprat, Sprattus sprattus striped bass, Morone saxatilis sunfish, Mola mola swordfish, Xiphias gladius turbot, Scophthalmus maximus wahoo, Acanthocybium solandrei whiting, Gadus merlangus yellowfin tuna, Thunnus albacares

Chapter 2 Fish Vision and Its Role in Fish Capture Takafumi Arimoto, Christopher W. Glass, and Xiumei Zhang

2.1 INTRODUCTION Vision in vertebrates, including fish, has been extensively studied (see Crescitelli 1977; Douglas and Djamgoz 1990; Guthrie and Muntz 1993; Lythgoe 1979; Nicol 1989). While the structure of the eye is well known and mechanisms of vision have been described for a number of fish, many commercially important marine species have received little attention. Despite many years of research into the visual systems of fish, detailed knowledge and understanding of the role of fish vision in their reaction to fishing gears during capture processes are far from complete. Understanding visual characteristics of fish is an important component in understanding the fish capture process and interactions between fish and fishing gear. This chapter reviews the structure and function of the eyes of marine fishes as well as the underwater visual environment to assist in better understanding the reaction of fish to fishing gear. Some examples of using fish vision in the design and operation of fishing gear are also illustrated. Knowledge gaps are identified for future research in the field of fish vision and visual behavior in relation to fish capture processes.

Lythgoe 1979; Nicol 1989). A short review on the structure and function of fish eyes and their role in the fish capture process is provided here. Although there is great diversity in the details of individual component that reflects the variation in the light environment inhabited by fish, a typical fish eye is illustrated in Figure 2.1. The eye has two main functions: optics and accommodation (Fernald 1990).

2.2 STRUCTURE OF THE FISH EYE The form, function, and structure of the fish eye (Fig. 2.1) have been documented extensively in the past (Atema et al. 1988; Collin and Marshall 2003; Douglas and Djamgoz 1990; Kawamoto 1970;

Accommodation: The Focusing of the Image on the Retina In most fish, the image is focused on the retina by movement of the lens rather than by a change in the shape of the lens, as occurs in other vertebrates.

Optics: The Collection and Formation of an Image Both sensitivity and acuity depend on the brightness of an image reaching the retina, and this is affected by the properties of the eye. The fish pupil is usually immobile, and light control is performed by the retinomotor mechanism involving movement of melanin granules in the retinal pigment cells. Optical resolution depends on lens quality, receptor size, and density. Images are formed by the refractive properties of the lens as the cornea of most fish eyes has a refractive index almost identical to that of water and contributes little to the optics of the eye.

25

26

Locomotion and Sensory Capabilities in Marine Fish

Figure 2.1. Structure of fish eye. (Modified from Kawamoto 1970.)

2.2.1 Lens The lens is the main refracting component of the eye and is usually spherical in shape. Details of the lenses of fish eyes are discussed by Sivak (1990). The spherical lens of fish eyes has been shown to produce good resolution due in part to a refractive index gradient within the lens; a higher refractive index at the center of the lens decreases with radius in all directions (Fernald 1985). Differences in refraction may be obtained by a variation in the composition of the lens as indicated by the ratio of proteins and water in the lens, which varies among species (Nicol 1989). Accommodation in the fish eye is a result of movement of the lens (rather than of a change in its shape) or of altering the depth of the eyeball to change the lens–retina distance in some species. The lens is moved backward to focus an image in teleosts, moved forward in elasmobranches, and pushed backward by flattening the cornea in lampreys (Nicol 1989). Studies on accommodation have shown that the direction of lens movement and the region of highest cone density in the retina are related (Tamura 1957; Tamura and Wisby 1963).

2.2.2 Retina As with the general structure of the eye, retinas of fish display similar characteristics but differ in details, reflecting the unique visual-environmental conditions under which different fish species live, whether in shallow well-lit lagoons or the dark bottom of a deep ocean (see Guthrie and Muntz 1993; Nicol 1989). The generalized fish retina is composed of an outer sheet of pigment epithelium covering a layer of photoreceptors (visual cell layer) and an inner layer of nervous tissue (Fig. 2.2). Teleosts generally possess both rod and cone photoreceptor cells. Rods have only one pigment and are used for scotopic (dark-adapted) vision, whereas cones may have up to four pigments and are used for color or photopic (light-adapted) vision. Cones may be double or single, and some teleosts may have triple or quadruple cones. Where multiple cones exist, cones may be structurally different or indistinguishable and may even contain different pigments. Single and double cones are usually found in specific arrangements or mosaics, which vary among species. The number of cones may also change across the retina within a single species.

Fish Vision and Its Role in Fish Capture

27

Figure 2.2. Retinal structure of Pacific saury (Cololabis saira) indicating layers within the retina. (Hajar 2007.) For color detail, please see color plate section.

The neural layers of the retina are composed of a nuclear layer, a ganglion cell layer, and a plexiform layer. The signal from the retina passes from the photoreceptors to the bipolar cells and then to the ganglion cells. It is successively modified by the horizontal, amacrine, and inner plexiform cells. As with the structure of the eye as a whole, details of the structure of the retinal layer and its component cells are subject to great variations among species (Wagner 1990). 2.3 VISUAL FUNCTION 2.3.1 Color Vision Most fish species can distinguish color by the use of red-, green-, and blue-sensitive cones. At least two types of cones are required for color discrimination, while some freshwater and shallow-living marine species have the capability to detect ultraviolet radiation with a fourth type of cone. Many deep sea species are often referred to as color-blind,

because their retinas are composed of only rods with no cones. Color vision and color discrimination can be determined by behavioral conditioning techniques. However, care is needed in inferring true color discrimination capability from these experiments because characteristics of an apparently simple color can be confounded by subtle aspects such as hue, saturation, and brightness of the targets. Electroretinogram (ERG) is used to monitor the response of retina to stimulation by different wavelengths of light (i.e., color) and to determine spectral sensitivity of fish eyes. Figure 2.3 shows ERG wave patterns of walleye pollock (Theragra chalcogramma) in dark-adapted condition. Figure 2.4 provides the spectral sensitivity curve of walleye pollock showing the Purkinje phenomenon where the wavelength of highest sensitivity is shifted from 540 nanometer (nm) for the light-adapted eye (photopic vision) to 470 nm for the dark-adapted eye (scotopic vision) (Zhang 1992).

28

Locomotion and Sensory Capabilities in Marine Fish (Anthony 1981) or by monitoring schooling performance in obligate schooling species (Glass et al. 1986), and by electrophysiological monitoring of the retinal response to varying light intensities (Kobayashi 1962). Different fishing gears provide a different contrast image according to ambient light conditions, gear type, and the visual sensitivity of the fish. The contrast of an object against the water background appears to be more important than the brightness of the object (Wardle 1987, 1993).

Figure 2.3. Electroretinogram (ERG) amplitude in dark-adapted eyes of walleye pollock (Theragra chalcogramma) under different light intensities. (Zhang 1992.)

2.3.2 Light Vision Photosensitivity is the ability of fish eye to receive light and to get visual information in ambient light conditions. Light intensity varies with water depth, time of day, and transparency or turbidity of water. To adapt to the wide range of light intensities encountered at sea, functional changes between cone and rod cells are made through shifting of positions of visual cells according to the ambient light intensity. Rods are highly sensitive to low light intensities, while cones are used at high light intensities. This allows fish to function visually over a wide range of light intensities in the natural environment. The minimum light intensity threshold for fish to function visually has been determined for a number of fish species by cardiac condition response

2.3.3 Motion Vision A moving image is generally of more importance to an animal than a static image. Detection of movement is dependent on visual acuity and persistence time—the time taken to process the image. The frequency at which flickering images fuse to produce a continuous image is referred to as the flicker fusion frequency (FFF) or critical flicker frequency (CFF) and is dependent on light intensity, temperature, and flash duration. Research on the perception of flicker in fish has been periodically reviewed (Douglas and Hawryshyn 1990; Landis 1954; Nicol 1989; Protasov 1970). The ability of fish to detect a moving or flickering image is affected by the level of illumination. Fish can perceive motion at a wide range of light intensities from 10−7 to 10−4 lux (Protasov 1970). At low light intensities, detection of a target is only possible if sufficient light is received to activate the photoreceptors. Increased sensitivity to low light levels is facilitated by temporal summation, which is a low threshold frequency of flicker fusion. In other words, an image must persist for a longer duration if it is to be perceived under low light levels; therefore, fast-moving images will not likely be detected. As light intensity is increased, the sensitivity to, or detection of, an image is greatly enhanced. The amount of temporal summation necessary to perceive a moving or flickering image decreases as sufficient light is received in a shorter time period to activate the receptors (Douglas and Hawryshyn 1990). Simply put, decreasing light intensity leads to a decrease in capacity for perception of motion. As with spectral sensitivity, FFF can be determined using either behavioral or electrophysiologi-

Fish Vision and Its Role in Fish Capture

29

Figure 2.4. Relative electroretinographic amplitude in lightadapted eyes (open circle) and two dark-adapted eyes of different time, showing the phenomenon of Pukinje shift. (Zhang 1992.) For color detail, please see color plate section.

cal approaches. Values of FFF obtained by electroretinography have been shown to increase with light intensity, although the relationship is not linear (see Blaxter 1970; Loew and McFarland 1990; Protasov 1970; Zhang and Arimoto 1993a). Curves obtained by plotting FFF against log light intensity are generally double branched and the point of inflection is a result of change in receptors from rods to cones. The point of inflection differs among species and the switch in receptor type may be mediated to some degree by photochemical movements of photoreceptors and the pigment epithelium (Douglas and Hawryshyn 1990). Behavioral techniques have also been used to investigate FFF. The optomotor response—that is movement of the eyes, head, curvature of the body or trunk, or movement of the entire animal in response to follow a moving image–has often been used as a determinant of FFF and visual acuity (Harden-Jones 1963; Nicol 1989; Sbikin 1981).

Behavioral results may differ from those obtained by electrophysiological techniques using electroretinography, possibly because the perceived FFF through behavioral techniques is a result of complex visual processing whereas an ERG is obtained in a relatively early stage of the process (Douglas and Hawryshyn 1990). Comparative studies on flicker fusion have often been presented as maximum photopic FFF. Elasmobranches generally have lower FFF than teleosts. Pelagic species living near the surface tend to have higher thresholds than do those dwelling in the bottom of deeper waters. Protasov (1970) was the first to relate FFF to ecology of the fish. For example, motion perception is important in detection of predators and prey; spurt reaction of prey will not be detected by a predator with a lower FFF. It is important to remember that many FFF values were determined from laboratory experiments where the background was usually black. In natural

30

Locomotion and Sensory Capabilities in Marine Fish

conditions, flickering or moving targets are often viewed against a nonuniform and changing background (Nicol 1989). The detection of movement has important implications in how fish react to fishing gears, in particular, trawl gears. The herding and optomotor reactions, by which fish hold station with the gear components such as bobbins, floats, ropes, and meshes until it becomes exhausted, are both a result of these images constantly moving relative to the background and relatively stationary to the fish. The higher FFF of a species, more acute is the sense of motion detection. In addition to FFF, fish have an ability to detect a moving object that is related to the pattern of the visual cells in the retina. A mosaic arrangement of cones confers better detection of moving objects and is found in species such as jack mackerel (Trachurus symmetricus) and chum salmon (Oncorhynchus keta) that rely on superior visual capabilities for predation and schooling (Wagner 1990). A random arrangement that confers lower motion detection capability is found in Japanese eel (Anguilla japonica), which has a lifestyle less reliant on the need for high-definition motion detection. Predatory species feeding on fast-moving preys require high temporal resolution motion vision. Higher FFF can be achieved with higher eye temperatures, which in most fish species is similar to ambient water temperature. For example, an increase of FFF from 5 Hz at 10°C to over 40 Hz at 20° was observed in swordfish (Xiphias gladius) (Fritsches et al. 2005). To increase higher temporal resolution for successful hunting of fast-moving prey, swordfish developed a specialized heating system to keep eyes and brain 10° to 15° above ambient water to increase temporal resolution (Fritsches et al. 2005). The increased capability of motion vision may also exist in other “warmbodied” tunas, which have a unique thermal regulation system (Sharp and Dizon 1978). Sharper vision, as well as faster swimming capacity, makes them superior predators. 2.3.4 Form Vision The lens quality and focal accommodation, together with the retinal resolving power, define the ability

Figure 2.5. Diagram showing binocular vision (1), monocular vision (2), and blind zone (3) in a typical teleost fish.

of fish to perceive details of a visual object. A pair of eyes generally located on opposite sides of the body gives a wide visual field with a monocular visual field on each side and a rather narrow binocular visual field facing forward (Fig. 2.5) (Wardle 1993). Different fish species have a different contour pattern of cone densities, with the highest density area being located in different regions. The position of highest cone density area defines the visual axis of each species (Shiobara and Arimoto 1999; Miyagi et al. 2001). The visual axis is related to the visual response, and in general it defines the direction in which the fish has greatest visual sensitivity. In some species like yellowtail (Seriola quinqueradiata), the concentration of high-density area is not obvious, indicating that this active predatory species has good all-round vision. The determination of the visual axis direction derived from the retinal histology is supported by the direction of movement of the lens during focal accommodation, which can be examined by electrical stimulation of the lens muscle. Visual acuity is the ability of fish to resolve and to see fine details of an object or a pattern. There are three different aspects of visual acuity:

Fish Vision and Its Role in Fish Capture

31

Table 2.1. Conversion Table for Visual Acuity Expressed in Minimum Separable Angle (MSA), Minimum Angle of Resolution (MAR), Snellen Notation, and Decimal Unit. MSD (m) for Target Size of MSA (Radian) 2.91 × 10−3 1.45 × 10−3 0.73 × 10−3 0.58 × 10−3 0.36 × 10−3 0.29 × 10−3 0.15 × 10−3

MAR (min)

Snellen Notation

Decimal Unit

0.5 cm

1 cm

2 cm

10 5.0 2.5 2.0 1.25 1.0 0.5

6/60 6/30 6/15 6/12 6/7.5 6/6 6/3.0

0.10 0.20 0.40 0.50 0.80 1.00 2.00

1.72 3.45 6.85 8.62 13.9 17.2 33.3

3.44 6.90 13.7 17.2 27.8 34.5 66.7

6.87 13.8 27.4 34.5 55.6 69.0 133.3

The corresponding maximum sighting distance (MSD) for different target sizes are estimated.

• Target detection—the ability to form an image on the retina • Gap detection—the ability to distinguish fine details such as the gap of the Landolt C Mark • Target recognition—the ability to recognize letters such as the Snellen Notation used in opticians’ clinics

and pigment index (P), which in turn are calculated from the cone position (c) and the thickness (A) of visual cell layer (Fig. 2.6):

Gap detection and target recognition are compatible with the minimum separable angle (MSA) or the minimum angle of resolution (MAR) (see also Section 2.4.1). The visual acuity can be expressed by several different units, which are derived from the MSA. These units can be converted from each other as shown in Table 2.1. The decimal unit is defined as the reciprocal of MAR in minutes of arc. The smaller MAR represents better visual acuity.

P=p A

2.3.5 Retinomotor Response The retinomotor response is a process by which the relative position of cones and rods in the retina changes in response to changes in ambient light intensity. Figure 2.6 shows cone positions of the retina of jack mackerel at different retinal adaptation stages from light, transitional, and dark adapted eyes. In light-adapted eyes, cones are moved to the surface level close to the outer limiting membrane in visual cell layers by shortening its myoid. There is an associated shift of pigments to cover the rods to protect them from strong light. Adaptation stages of the retina can be described by the cone index (C)

C=c A or from the shifting distance of pigment layer (p):

The retinomotor response follows a circadian rhythm. It is influenced by ambient light and can be modified by artificial light during light fishing. Identification of light/dark adaptation stages using cone index provides important information on visual response of target and nontarget species during their capture by fishing gears. In the case of walleye pollock captured by a trawl at a depth of 270 m during daytime, retina conditions were identified as transitional or dark-adapted stages, indicating a reduced visual response of fish to towed gears in deep waters (Zhang et al. 1993). 2.3.6 Optomotor Response The optomotor response, which is sometimes called the optomotor reaction, refers to the phenomenon that the fish maintains a relatively fixed position of a visual image on its retina. Optomotor response may be one of the most significant behavioral responses governing fish reaction to the surrounding stimuli during fish capture processes. The

32

Locomotion and Sensory Capabilities in Marine Fish

Figure 2.6. Changes in the position of layers in the retina during dark adaptation, transitional stage, and light adaptation. p, shifting distance of pigment layer; c, cone position; and A, thickness of visual cell layer. (Zhang 1992.) For color detail, please see color plate section.

response pattern explains a number of observed behaviors such as maintaining position in flowing water or in a school (Shaw 1965). The optomotor response has been used to control fish behavior for the purpose of measuring swimming speeds (Bainbridge 1958; He and Wardle 1988) and for studying spectral sensitivity using different light patterns as visual targets (Hasegawa 2006). Pavlov (1969) described station-holding swimming of fish at the wing of a trawl as being mediated by the optomotor response and discussed potential for its application in improving capture efficiency of the fishing gear. This was further elaborated in the laboratory (Inoue and Arimoto 1976; Inoue and Kondo

1972) and in the field during fish capture processes using direct underwater observation techniques (Wardle 1987).

2.4 VISUAL CAPACITY: VISUAL ACUITY, SEPARABLE ANGLE, AND MAXIMUM SIGHTING DISTANCE Visual acuity is represented by the MSA (α in radian), as defined by the following equation (Tamura 1957):

α=

1 ⎡ 2 × 0.1 × (1 + .025) ⎤ ⎦⎥ F ⎣⎢ n

Fish Vision and Its Role in Fish Capture

33

Minimum Separable Angle

Cone On

α

F

Off On

Landolt C

where F is the focal length of the lens, which is 2.55 times the radius of the lens, and n is cone density, which is the number of cones in an area of 0.01 mm2. The equation is derived from the concept of optical resolving capability by cones in the retina as shown in Figure 2.7, where at least three cones are required to identify the gap in the Landolt C Mark, with one cone for the gap between a pair of dots or between the two tip points of the Landolt C Mark. The visual information received by visual cells is integrated and modulated at ganglion cells that are located at the surface layer in the lining of retina and transmitted to the optic nerve in the retinal level. Collin and Pettigrew (1989) reasoned that the visual acuity is related to the density of ganglion cells, with the same concept of discriminating a pair of dots as the optical resolving capability. While the approach is theoretically feasible, it is difficult to identify and count the number of ganglion cells during histological examinations. Behavioral approaches to determine the MSA are possible using classic conditioning techniques by training fish to identify the direction of the opening of the Landolt C Mark. After the fish is conditioned with a larger Landolt C Mark, the mark size is systematically reduced until the fish cannot discern the direction of the opening. The MSA derived from behavioral approach is usually smaller than that obtained from histological examinations (Shiobara and Arimoto 2003).

Figure 2.7. Landolt C Mark, cone density, and the minimum separable angle (α). F, Focal length. (Hajar 2007.)

Figure 2.8 shows the MSA for a number of species and sizes of fish. Overall, visual acuity is affected by several optical factors, including ambient light intensity and contrast of visual targets against the background. Larger fish have smaller angles, indicating they have better visual acuity and are more capable of distinguishing smaller or finer details of visual objects. The maximum sighting distance (D) can be estimated from the MSA (α, in radian) and the size of the visual target (l): D=l α assuming ideal optical conditions, which include a high contrast target, in high illumination conditions, and in highly transparent water. Under these ideal optical conditions, the maximum sighting distance of walleye pollock in relation to the size of fish and the size of visual targets is shown in Figure 2.9. In general, larger fish have larger sighting distance and can detect fishing gear components or other underwater objects from farther away (for more details, see Arimoto and Namba 1996). The maximum sighting distance derived from the MSA may not fully represent the true capability of fish to detect a fishing gear or its components. The estimation is solely dependent on target size and cone density. Here the target size of floats, bobbins, or knots of netting is assumed to be the length between the top and bottom or left and right of the

Figure 2.8. Visual acuity as expressed by the minimum separable angle (in min) in relation to fish body length in yellowtail (Theragra chalcogramma) (Miyagi 2001), red sea bream (Pagrus major) (Shiobara et al. 1998), jack mackerel (Trachurus symmetricus), walleye pollock (Theragra chalcogramma) (Zhang 1992), and Pacific saury (Cololabis saira). (Hajar et al. 2008.)

Figure 2.9. Maximum sighting distance of walleye pollock (Theragra chalcogramma) of different body lengths when viewing object of 2- to 10-cm visual target. (Zhang and Arimoto 1993.)

34

Fish Vision and Its Role in Fish Capture

35

2.5 COLOR AND APPEARANCE OF FISHING GEAR UNDERWATER

variations in many factors, but a general classification of coastal and oceanic water types by color has been established by Jerlov (1964). In general, spectral absorption of light in the open oceans produces an effect, as shown in Figure 2.10, where light in the green/blue part of the spectrum transmits deeper into the water column than other wavelengths. This can have a profound effect on behavioral ecology of fish residing in different areas of the ocean, whether in the upper strata of the ocean, in shallow waters near the coastline, or in deep ocean bottoms. It should be noted, however, that much of the world’s commercial fishing operations are conducted in relatively deep waters either in the absence of visible light or in dim monochromatic conditions.

2.5.1 Spectral Properties of Seawater There is a great diversity in the underwater visual environment inhabited by fish. Many factors, including water property, nature of light source, and suspended particles, affect the distribution of light as it passes down and through the water column. The characteristics of underwater visual environment and their influence on fish vision and behavior have been described by a number of authors, notably Guthrie and Muntz (1993); Blaxter (1970, 1988), Lythgoe (1979), Loew and McFarland (1990), Nicol (1989), and Protasov (1970). Underwater natural light comes from either solar radiation or bioluminescence. As solar radiation passes through the water, it is refracted toward the vertical and progressively filtered and diminished due to absorption and scattering effect of water molecules, dissolved pigments, and particulate matter. This results in generally monochromatic conditions at depths at which fishing gears are usually operated, with the light coming more or less from directly above. Upwelling or horizontal transmission of light generated through bioluminescence, where it occurs, is of low intensity (approximately 5% of overall illumination) but has a significant effect on the visibility of fishing gear and may under certain circumstances result in directed behavioral reactions by fish in its vicinity. The spectral transmission curves of waters of oceans and coastal areas can vary greatly due to

2.5.2 Visual Contrast of Fishing Gear Blaxter et al. (1964) summarized the role of visual senses of fish when responding to fishing gears. Wardle (1983) stressed the importance of understanding visual contrast of fishing gear against the background and suggested that it was more important than the brightness of the gear itself. Wardle (1993) described and illustrated the herding effect induced by a high-contrast image presented by the otter boards and the trawl mouth and the entrapping effect induced by the less visible rear section of the net. The visual angle for detecting targets was compared with contrast perception for different parts of the gear as they were viewed from below against the bright surface and from above against the dark background (see Chapter 8, Figure 8.7). There is a complex relationship between color and contrast of gear components, ambient light intensity. and water quality. Kim and his colleagues (Kim 1998; Kim and Wardle 1998a, 1998b) modeled visual stimulus and contrast of fishing gear components and made some interesting predictions on fish behavior near fishing gear under visual conditions. But in general it has been demonstrated that light-colored netting panels are more difficult to detect against a bright background because there is little contrast between the image and its background. The reverse is also true for materials that contrast strongly with the background against which they are viewed. The contrast required to distinguish the target against its background can be defined by the

target, not the gap distance as used in calculating the MSA. Furthermore, the line acuity required for resolving mesh twines or grating resolutions for netting panels tends to be larger than the visual acuity for a point source target. Regardless, the maximum sighting distance calculated from the given formula is a useful tool for predicting visual capability of fish, especially for comparing different species and sizes of fish. It can also be used for estimating visibility of targets and visual range of fish and for manipulating the visual appearance of fishing gears and their components to make them more or less visible.

36

Locomotion and Sensory Capabilities in Marine Fish

Figure 2.10. Spectral absorption of light in the open oceans where light in the green/blue part of the spectrum transmits deeper into the water column than other wavelengths (reprinted from http:// ultramaxincorp. com/?p2=/modules/ ultramax/catalog. jsp&id=23). For color detail, please see color plate section.

following equation and has been referred to as the apparent contrast: C = (I − Ib ) Ib where I and Ib represent target and background radiance (W m−2) or irradiance (W sr−1 m−2). The contrast threshold was determined in several species by observing behavioral responses to food reward, including striped beak-perch (Oplegnathus fasicatus) (Arakawa et al. 2007), red sea bream (Pagrus major) (Miyagi 2001), and Japanese common squid (Tadorodes pacificus) (Siriraksophon et al. 1995; Siriraksophon and Morinaga 1996). The cardiac conditioning technique was also reported for Atlantic cod (Gadus morhua) (Anthony 1981). While the contrast threshold varies among species, target types, background conditions, and ambient light intensity, it generally showed higher contrast sensitivity in dark ambient light conditions due to function of the rod cell in scotopic vision and lower contrast sensitivity in brighter conditions mediated by the cone cell response in photopic vision.

The contrast of a net is determined by the apparent contrast of twine when viewed against the background. However, when viewing a netting panel, the twine of the netting panel comprises only a small portion of the overall area being viewed and this can have an impact on visibility of the netting and visual range of the fish. The projected area ratio, or netting solidity, is the ratio of the area of the mesh twine to the total area of the netting and is related to mesh size, twine thickness, knot type, and hanging ratio. More solid netting (i.e., netting with larger projected area ratios) is more visible and provides fish with a larger visual range. 2.6 FISH VISION AND ITS APPLICATION IN FISH CAPTURE 2.6.1 Herding and Capture of Fish by Trawl under Visual Conditions The importance of fish vision in relation to the capture process of trawl fisheries was first highlighted by Blaxter et al. (1964) and Parrish (1969) through a series of laboratory and field observa-

Fish Vision and Its Role in Fish Capture tions. Since these early studies, advances in underwater observation techniques with visual, photographic, and acoustic tools (Graham et al. 2004; Urquhart and Stewart 1993) have greatly increased our knowledge and understanding of the capture process of trawls—for example, the level of illumination required for fish to form visual images of the fishing gear (Wardle 1983, 1987, 1989, 1993; also see Chapter 4). The herding by the otter boards, trawl warps, and sand cloud is the first stage of the capture process, where the visual stimuli of the trawl gear assists with accumulating fish inside and toward the trawl mouth. Observations of fish avoidance reaction to the wing of the net, the netting of the trawl mouth, and other components have been described by Kim and Wardle (2003). In that study, they quantified optical characteristics such as visual contrast of gear components against water background as viewed by a fish. They modeled fish behavior in response to a suite of visual stimuli. Zhang and Arimoto (1993b) modeled escape distances for fish to avoid a trawl as a function of the maximum sighting distance and fish swimming speed. Despite these efforts, there is disconnect between our knowledge of visual physiology and how it relates to observed visual reactions to fishing gears, but this may be a fertile area for future research. This research may have important implications for developing conservation-oriented fishharvesting measures such as design of nets that avoid bycatch and reduce discard through manipulation of visual images of the gear. 2.6.2 Use of Light in Fishing Fishing with lights is one of the most advanced and successful methods for catching squids and other pelagic species (Ben-Yami 1976, 1988; Inada and Arimoto 2007). The technique has been successfully used to attract and aggregate fish for centuries. Started with fire torches, a wide range of light fishing equipment and methods has been developed for net gears such as purse seines and lift nets, as well as for hook-and-line methods for squid, mackerel, and jack mackerels. The method has been used in small-scale fisheries along the coast of Japan and some Southeast Asian countries, as well as in largescale offshore and oceanic fisheries. A wide variety of technologies and lighting equipment has been

37

developed over time to best match water quality at specific fishing grounds. Specific lighting parameters (e.g., color and intensity) have been applied according to the behavior of target species. While fishing techniques have evolved over time, there is little knowledge of why squids and fish are attracted to light, although many explanations have been offered, such as: • Schooling for feeding under the light • Conditioned responses to light intensity gradients • Curiosity behavior and other social behavior • Positive phototaxis making them orient to the light source • Optimum light intensity for feeding and other activities • Disorientation and immobilization due to localized high light levels in surrounding dark conditions It is understood, however, that attraction is likely due to a combination of the factors and almost certainly varies both temporally and spatially (BenYami 1976, 1988). Light fishing continues to be one of the largest and most productive fishing techniques globally and as such has generated a wide and diverse range of research and development activities. While much of this activity is focused on improving commercial productivity, there are many current initiatives to reduce energy costs associated with light production through use of light emitting diodes (LED) as the primary source of light and to tailor light quality (e.g., intensity and color) to different species aimed at species-selective fishing (Inada and Arimoto 2007). Both topic areas have important implications for conservation of energy and fish stocks for sustainable and profitable exploitation of fishery resources in the future. 2.6.3 Use of Light and Illusion in Guiding and Blocking As outlined earlier, many of the important behavioral reactions of fish to fishing gears are governed or mediated by the visual system. This has long been exploited, often unsuspectingly, by fishermen. For example, there are many opinions regarding the

38

Locomotion and Sensory Capabilities in Marine Fish

best or most appropriate color of fishing twine for commercial success, indicating that fishermen do indeed understand the role that color, contrast, and spectral qualities of water have on the behavior of fish in response to their fishing operations (Jones et al. 2005). Fishing gear may have evolved to exploit the optomotor response in the case of trawls or to exploit the visibility of twine in the case of monofilament gillnets. However, the role of light and manipulation of visual stimuli to modify natural behavior patterns of fish and to guide fish to desired directions or destinations has only recently been explored in new and innovative ways with regard to fish capture in commercial fishing operations. This field of research may yet prove to yield significant advances in improving fishing efficiency and selectivity, reducing bycatch and discard, and reducing the impact of fishing on the habitat. The use of fish behavior to guide animals has long received attention in rivers and waterways in many countries (Coutant 2001). Behavioral systems for guiding fish away from turbines or for facilitating fish migration through fish passages are generally attractive financially for users of a water resource as they can be less costly and easier to implement than structural systems (Coutant 2001). However, the notion of merely hanging a simple device such as a strobe light or other visual cue from an intake to “frighten” fish away or to attract fish along a different path has proved to be oversimplified and often misguided. Nevertheless, behavioral control systems have been studied extensively and are often considered an appropriate first step in problem solving (Coutant 2001). Within the field of marine commercial fisheries, while the importance of vision and the role of visual images of fishing gear components have been recognized for a long time, there have been, with a few notable exceptions, surprisingly few directed attempts to manipulate patterns of fish behavior. In the 1970s and 1980s, Wardle and a team of coworkers in Aberdeen, Scotland (Wardle 1983), developed an experimental, or concept, trawl with black-and-white striped patterns in its wings and main belly of the net (Fig. 2.11). The patterns were designed to maximize the effect of the optomotor response in the wings of the net and to guide exhausted fish more quickly to the codend for sub-

sequent capture. Later, a striped panel constructed from regular dark (blue) twine and bright Glownet material made by Nichimo of Japan was designed to provide visual guiding even when ambient illumination levels dropped below the visual threshold for image formation. Both nets proved successful in concept but were never further pursued toward commercialization. Nevertheless, they helped demonstrate how fish, even in commercial fishing conditions, could have their behavior manipulated by controlling the visual stimuli surrounding them. Subsequent studies by Wardle et al (1991) and Glass et al (1993, 1995) continued this exploration of behavior modification through manipulation of visual stimuli. Wardle et al (1991) demonstrated the importance of spectral quality of water in rendering monofilament gillnets visible or invisible to approaching fish (also see Chapter 8). In this study, they identified that the angle of the twine relative to the viewer was important in determining its visibility and that knots of the netting acted in a fashion similar to lenses, making knots the most visible part of the netting. These critical observations helped explain some of the observed behavioral reactions of fish to gillnet panels and prompted further investigation into the reactions of fish when surrounded by netting panels in the body and codends of towed fishing gears. Observations of fish in fishing gears have consistently shown that fish, under visual conditions, keep clear of the netting and are herded effectively by the panels of the wingends and the front regions of the net (Glass et al. 1993). Fish entering the extension and codend region of a net have also been shown to avoid the netting and do not attempt to pass through the open meshes around them. Different mesh configurations (diamond, square, hexagonal) present a different visual stimulus to fish and this stimulus is further affected by the color or contrast of the twine as viewed against the water background. Glass et al. (1993) also showed that fish reacted differently to component parts of a mesh. Horizontal bars elicited a different behavioral reaction to that shown when the fish were presented with vertically oriented bars. The importance of this study and the studies of Cui et al. (1991) and Wardle et al. (1991) is that they demonstrate that the natural behavior patterns of commercial fish species can be

Fish Vision and Its Role in Fish Capture

39

Figure 2.11. (A, B) A conceptual trawl design using optomotor response of fish to improve catch efficiency. (Courtesy Clem Wardle.)

altered by subtle changes in the surrounding visual stimulus. This led Glass et al. (1993) to postulate that by careful consideration of the behavior patterns of fish to surrounding panels of netting and by identification of the important components of the overall visual stimulus, it may be possible to create a system of visual illusions within a fishing net, which would stimulate the fish to approach and even penetrate meshes more readily. In a series of experiments in the laboratory and then at sea, Glass et al. (1995) and Glass and Wardle (1995) further investigated this concept to alter natural behavior patterns to encourage fish to escape. Throughout the fish capture process in trawls, there is little to encourage fish to escape or show avoidance responses. They postulated that a strong visual stimulus at the back of the net would encourage fish to attempt to escape through the meshes of the netting surrounding them rather than avoiding the meshes altogether. In developing the

strong visual stimulus that might be needed to elicit a response, they investigated the visual appearance of high-speed plankton samplers used to sample larval fish populations (Fig. 2.12). There was evidence of behavioral avoidance in larger-length classes, and this was thought to be induced by the high-contrast image of the approaching sampler, thereby allowing larger larval fish to react and swim out of its path (Glass et al. 1995). By recreating a similar high-contrast visual stimulus in laboratory experiments, Glass et al. (1995) identified that fish could, in fact, be induced to pass through meshes surrounding them when it appeared that alternative clear paths were either blocked or appeared to be blocked by visual patterns. To create the illusion that the path through the net was blocked, Glass and Wardle (1995) recreated a stimulus identical to that developed in laboratory experiments by placing a simple black tunnel in the net (Fig. 2.13). The fish were free to move through the tunnel but, when

40

Locomotion and Sensory Capabilities in Marine Fish

viewed from the front, the illusion of the path being blocked elicited a significant modification in the behavior. Fish were reluctant to pass through the tunnel, swam vigorously ahead of it, and responded by attempting to swim through open meshes just

ahead of the black tunnel. This illustrated how the knowledge of the underwater visual field and the behavior and sensory physiology of fish can be used in an applied manner to manipulate behavior of fish within the context of fishing operations.

Figure 2.12. Aberdeen Gulf III plankton sampler as viewed from the front. The engulfing black hole is believed to have induced larger plankton to escape by swimming out of the pass. (Wardle 1983.)

2.7 CONCLUDING REMARKS While the visual system of fish is well understood, there is a lack of research on its function and role in mediating behavior in commercial marine fish species during capture processes. Visual capacity of fish can be described by visual acuity, spectral sensitivity, and motion detection ability. Fish vision plays an important part in its reaction to fishing gears in catching a prey, and even more so in their attraction to light sources during light fishing. Many fish species have a two-gear system in vision provided by rod and cone cells in their retina, responsible for dark (scotopic) and light (photopic) conditions, respectively. Despite the earlier effort on modification and manipulation of visual stimulus and subsequent behavioral modification of fish within and around a net, this field of research has yet to achieve its full potential. With improvements in technology, there is tremendous scope for

Figure 2.13. Schematic illustration of the black tunnel (A) and underwater photograph of Atlantic cod (Gadus morhua) escaping just ahead of the black tunnel (B). (Crown copyright, reproduced with the permission of Marine Scotland.)

Fish Vision and Its Role in Fish Capture herding, guiding, and manipulating fish within and around a net, by use of their visual systems, in both light and dark conditions. Implications for improved fishing efficiency, conservation of fish stocks through reduction of bycatch and discard, reduction of interaction, and mortality of protected species may potentially revolutionize fishing operations. REFERENCES Anthony PD. 1981. Visual contrast thresholds in the cod Gadus morhua L. J. Fish Biol. 19: 87–104. Arakawa H, Watanabe T and Morikawa Y. 2007. Visual contrast threshold of striped beak-perch Oplegnathus fasciatus. Fish. Sci. 73: 469–471. Arimoto T and Namba K. (eds). 1996. Fish Behavior and Physiology for Fish Capture Technology. Tokyo: Koseisha-Koseikaku. 128 pp. (in Japanese). Atema J, Fay RR, Popper AN and Tavolga WN. 1988. Sensory Biology of Aquatic Animals. New York: Spring-Verlag. 936 pp. Bainbridge R. 1958. The speed of swimming of fish as related to size and to the frequency and the amplitude of the tail beat. J. Exp. Biol. 35: 109–133. Ben-Yami M. 1976. Fishing with Light: FAO Fishing Manuals. Surrey: Fishing News Books. 150 pp. Ben-Yami M. 1988. Attracting Fish with Light. FAO Training Series—14. Rome: FAO. 72 pp. Blaxter JHS. 1970. Light. Animals. Fishes. In: Kinne O (ed). Marine Ecology, Vol.1. Environmental Factors. pp 213–320. London: Wiley. Blaxter JHS. 1988. Sensory performance, behavior, and ecology of fish. In: Atema J, Fay RR, Popper AN and Tavolga WN (eds). Sensory Biology of Aquatic Animals. pp 203–232. New York: Springer-Verlag. Blaxter JHS, Parrish BB and Dickson W. 1964. The importance of vision in reactions of fish to drift nets and trawls. In: Modern Fishing Gear of the World 2. pp 529–536. London: Fishing News Books. Collin SP and Marshall NJ. 2003. Sensory Processing in Aquatic Environments. New York: SpringerVerlag. 446 pp. Collin SP and Pettigrew JD. 1989. Quantitative comparison of the limits on visual spatial resolution set by the ganglion cell layer in twelve species of reef teleosts. Brain Behav. Evol. 34: 184–192. Coutant CC (ed). 2001. Behavioral Technologies for Fish Guidance. Am. Fish. Soc. Symp. 26: 193 p. Crescitelli F (ed). 1977. Handbook of Sensory Physiology, Vol. VII/5: The Visual System in Vertebrates. New York: Springer-Verlag. 813 pp.

41

Cui G, Wardle CS, Glass CW, Johnstone ADF and Mojsiewicz WR. 1991. Light level thresholds for visual reaction of mackerel, Scomber scombrus L., to colored monofilament nylon gillnet materials. Fish. Res. 10: 255–263. Douglas RH and Djamgoz MBA (eds).1990. The Visual System of Fish. Devon: Chapman and Hall. 526 pp. Douglas RH and Hawryshyn CW. 1990. Behavioral studies of fish vision: an analysis of visual capabilities. In: Douglas RH and Djamgoz MBA (eds). The Visual System of Fish. pp 373–418. Devon: Chapman and Hall. Fernald RD. 1985. Growth of the teleost eye: novel solutions to complex constraints. Environ. Biol. Fish. 13: 113–123. Fernald RD. 1990. The optical system of fishes. In: Douglas RH and Djamgoz MBA (eds). The Visual System of Fish. pp 45–61. Devon: Chapman and Hall. Fritsches KA, Brill RW and Warrant EJ. 2005. Warm eyes provide superior vision in swordfishes. Curr. Biol. 15: 55–58. Glass CW and Wardle CS. 1995. Studies on the use of visual stimuli to control fish escape from codends. II. The effect of a black tunnel on the reaction behavior of fish in otter trawl codends. Fish. Res. 23: 165–174. Glass CW, Wardle CS and Gosden SJ. 1993. Behavioral studies of the principles underlying mesh penetration by fish. ICES Mar. Sci. Symp. 196: 92–97. Glass CW, Wardle CS, Gosden SJ and Racey DN. 1995. Studies on the use of visual stimuli to control fish escape from codends. I. Laboratory studies on the effect of a black tunnel on mesh penetration. Fish. Res. 23: 157–164. Glass CW, Wardle CS and Mojsiewicz WR. 1986. A light intensity threshold for schooling in the Atlantic mackerel (Scomber scombrus). J. Fish Biol. 29A: 71–81. Graham N, Jones EG and Reid DG. 2004. Review of technological advances for the study of fish behavior in relation to demersal fishing trawls. ICES J. Mar. Sci. 61: 1036–1043. Guthrie DM and Muntz WRA. 1993. Role of vision in fish behavior. In: Pitcher TJ (ed). Behavior of Teleost Fishes. 2nd ed. pp 89–128. Devon: Chapman and Hall. Hajar MAI. 2007. Visual Physiology of Fish for Understanding the Capture Process of Light Fishing. PhD thesis. Tokyo University of Marine Science and Technology, Tokyo, Japan.

42

Locomotion and Sensory Capabilities in Marine Fish

Hajar MAI, Inada H, Hasobe M and Arimoto T. 2008. Visual acuity of Pacific saury Cololabis saira for understanding capture process. Fish. Sci. 74: 461–468. Harden-Jones FR. 1963. The reaction of fish to moving backgrounds. J. Exp. Biol. 40: 437–446. Hasegawa E. 2006. Comparison of the spectral sensitivity of three species of juvenile salmonids. J. Fish Biol. 68: 1903–1908. He P and Wardle CS. 1988. Endurance at intermediate swimming speeds of Atlantic mackerel, Scomber scombrus L., herring, Clupea harengus L., and saithe, Pollachius virens L. J. Fish Biol. 33: 255–266. Inada H and Arimoto T. 2007. Trends on research and development of fishing light in Japan. J. Illum. Eng. Inst. Jpn. 91: 199–209. Inoue M and Arimoto T. 1976. On the optomotor reaction of fish relevant to fishing method. III. Experiments for fishing purpose. J. Tokyo Univ. Fish. 63: 9–16. Inoue M and Kondo T. 1972. On the optomotor reaction of fish relevant to fishing method. J. Tokyo Univ. Fish. 58: 9–16. Jerlov NG. 1964. Optical classification of ocean water. In: Tyler JE (ed). Physical Aspects of Light in the Sea. pp 45–49. Honolulu: Univ. Hawaii Press. Jones E, Glass C and Milliken H. 2005. The reaction and behavior of fish to visual components of fishing gears and the effect on catchability in survey and commercial situations. ICES CM 2004/B: 05, ref. ACE. Annex 2. pp 68–112. Kawamoto N. 1970. Fish Physiology. Tokyo: Koseisha Koseikaku. 554 pp. (in Japanese). Kim YH. 1998. Modeling on contrast threshold and minimum resolvable angle of fish vision. Bull. Kor. Soc. Fish. Technol. 33(3): 43–51. Kim YH and Wardle CS. 1998a. Modeling the visual stimulus of towed fishing gear. Fish. Res. 34: 165–177. Kim YH and Wardle CS. 1998b. Measuring the brightness contrast of fishing gear, the visual stimulus for fish capture. Fish. Res. 34: 153–166. Kim YH and Wardle CS. 2003. Optomotor response and erratic response: quantitative analysis of fish reaction to towed fishing gears. Fish. Res. 60: 455–470. Kobayashi H. 1962. A comparative study on electroretinogram in fish, with special reference to ecological aspects. J. Shimonoseki Coll. Fish. 11(3): 17–148.

Landis C. 1954. Determinants of the critical flickerfusion threshold. Physiol. Rev. 34: 259–286. Loew ER and McFarland WN. 1990. The underwater visual environment. In: Douglas RH and Djamgoz MBA (eds). The Visual System of Fish. pp 1–43. Devon: Chapman and Hall. Lythgoe JN. 1979. The Ecology of Vision. Oxford: Oxford University Press. 244 pp. Miyagi M. 2001. Study on Visual Physiology of Fish for Application in Net Gears. PhD thesis. Tokyo University of Fisheries, Tokyo, Japan. (In Japanese). Miyagi M, Akiyama S and Arimto T. 2001. The development of visual acuity in yellowtail Seriola quinqueradiata. Bull. Jpn. Soc. Sci. Fish. 67(3): 455–459 (in Japanese with English abstract). Nicol JAC. 1989. The Eyes of Fishes. Oxford: Oxford University Press. 308 pp. Parrish BB. 1969. A review of some experimental studies of fish reactions to stationary and moving objects of relevance to fish capture processes. FAO Fish. Rep. 62: 233–245. Pavlov DS. 1969. The optomotor reaction of fishes. FAO Fish. Rep. 62: 803–808. Protasov VR. 1970. Vision and near orientation of fish. Trans. by M. Raveh for Israel Program for Scientific Translations. Washington, DC: U.S. Dept of Commerce. 175 pp. Sbikin YN. 1981. The optomotor reaction and some characteristics of the vision of young sturgeon. J. Ichthyol. 21: 167–171. Sharp GD and Dizon AE. 1978. The Physiological Ecology of Tunas. New York: Academic Press. 485 pp. Shaw E. 1965. The optomotor response and the schooling of fish. ICNAF Spec. Publications 6: 753–755. Shiobara Y, Akiyama S and Arimoto T. 1998. Developmental changes in the visual acuity of red sea bream Pagrus major. Fish. Sci. 64: 944–947. Shiobara Y and Arimoto T. 1999. Behavioral analysis of feeding experiment on visual axis of red sea bream Pagrus major. Bull. Jpn. Soc. Sci. Fish. 65: 728–731(in Japanese with English abstract). Shiobara Y and Arimoto T. 2003. Change in visual acuity and retinal adaptation according to light intensity for red sea bream Pagrus mojor. Nippon Suisan Gakkaishi. 69: 632–636. Siriraksophon S and Morinaga T. 1996. Effect of background brightness on the visual contrast threshold of the Japanese common squid. Fish. Sci. 62: 534–537. Siriraksophon S, Nakamura Y and Matsuike K. 1995. Study on visual contrast threshold in Japanese

Fish Vision and Its Role in Fish Capture common squid Todarodes pacificus. Fish. Sci. 61: 574–577. Sivak JG. 1990. Optical variability of fish lens. In: Douglas RH and Djamgoz MBA (eds). The Visual System of Fish. pp 63–80. Devon: Chapman and Hall. Tamura T. 1957. A study on visual perception in fish, especially on resolving power and accommodation. Bull. Jpn. Soc. Sci. Fish. 22: 536–557. Tamura T and Wisby WJ. 1963. The visual sense of pelagic fishes especially the visual axis and accommodation. Bull. Mar. Sci. Gulf Caribb. 13: 433–448. Urquhart GG and Stewart PAM. 1993. A review of techniques for the observation of fish behavior in the sea. ICES Mar. Sci. Symp. 196: 135–139. Wagner H-J. 1990. Retinal structure of fishes. In: Douglas RH and Djamgoz MBA (eds). The Visual System of Fish. pp 109–157. Devon: Chapman and Hall. Wardle CS. 1983. Fish reaction to towed fishing gears. In: Macdonald AG and Priede IG (eds). Experimental Biology at Sea. pp 168–195. London: Academic Press. Wardle CS. 1987. Investigating the behavior of fish during capture. In: Bailey RS and Parrish BB (eds). Developments in Fisheries Research in Scotland. pp. 139–155. Surrey: Fishing News Books.

43

Wardle CS. 1989. Understanding fish behavior can lead to more selective fishing gears. Proc. World Symp. Fish. Gear and Fish. Vessel Design. pp 12– 18. St. John’s, Newfoundland: Marine Institute. Wardle CS. 1993. Fish behavior and fishing gear. In: Pitcher TJ (ed). Behavior of Teleost Fishes. 2nd ed. pp 609–643. London: Chapman & Hall. Wardle CS, Cui G, Mojsiewicz WR and Glass CW. 1991. The effect of color on the appearance of monofilament nylon underwater. Fish. Res. 10: 243–253. Zhang XM. 1992. Study on Visual Physiology of Fish for Applying Trawl Net Operation. PhD thesis. Tokyo University of Fisheries, Tokyo, Japan (in Japanese). Zhang XM, Akiyama S, Arimoto T, Inoue Y and Matsushita Y. 1993. Retinal adaptation of walleye pollock in trawl fishing ground of north Pacific. Bull. Jpn. Soc. Sci. Fish. 59: 481–485 (in Japanese with English abstract). Zhang XM and Arimoto T. 1993a. Electroretinogram critical fusion frequency of jack mackerel Trachurus japonicus by strobe light. J. Tokyo Univ. Fish. 80: 61–67. Zhang XM and Arimoto T. 1993b. Visual physiology of walleye pollock Theragra chalcogramma in relation to capture by trawl nets. ICES Mar. Sci. Symp. 196: 113–116.

44

Locomotion and Sensory Capabilities in Marine Fish

SPECIES MENTIONED IN THE TEXT Atlantic cod, Gadus morhua chum salmon, Oncorhynchus keta jack mackerel, Trachurus symmetricus Japanese common squid, Tadorodes pacificus Japanese eel, Anguilla japonica

Pacific saury, Cololabis saira red seabream, Pagrus major striped beak-perch, Oplegnathus fasicatus swordfish, Xiphias gladius walleye pollock, Theragra chalcogramma yellowtail, Seriola quinqueradiata

Chapter 3 Hearing in Marine Fish and Its Application in Fisheries Hong Young Yan, Kazuhiko Anraku, and Ricardo P. Babaran

ical processes involved in the acousticolateralis system and to make the best use of the knowledge in the designs of fishing gear and operation. This chapter reviews the fundamental knowledge of physical properties of sound and discusses the physiological characteristics of the ear. Methodologies involved in understanding fish hearing in terms of frequency range and hearing threshold are reviewed along with a discussion on how exposure to noise will impact the overall hearing abilities of targeted fishes. The last section of this chapter provides a few examples of practical applications of acoustic signals in fisheries either to attract or to expel fish.

3.1 INTRODUCTION The acousticolateralis system of fish is composed of two major structures—the inner ear and the lateral line. The inner ear is responsible for the balance and detection of acoustic signals, whereas the lateral line detects water-borne vibration signals (Hawkins 1986). These mechanosensory functions are crucial for the survival of fish. In terms of anatomical structure, the functional units of the inner ear are sensory hair cells and are used to detect underwater acoustical signals, whereas the neuromasts of the lateral line detect waterborne lowfrequency vibrations caused by physical as well as biological forces. The sensation of water-borne sound and vibration offers fish a dual detection system to measure mechanical disturbances of their environment. In turn, fish can listen to sounds produced by either conspecifics or heterospecifics, and they can take corresponding actions such as retreating or escalating agonistic behavior or being attracted to the source if the sounds are courtship signals. Likewise, sensation from the lateral line informs recipients of the presence of obstacles, predators or prey. In conjunction with their eyes, the lateral line system also participates in the schooling behavior of fish (Pitcher 1979). The successful operation of fisheries whether at the commercial or subsistence level requires proper designs of fishing gears and methods. In light of how the auditory functions of ears and the sensation abilities of lateral line modulate the behaviors of fish, it would be useful to understand the physiolog-

3.2 PROPERTIES OF UNDERWATER SOUND AND VIBRATION An understanding of two major properties of sound in terms of sound pressure and water particle motion (displacement, velocity, and acceleration) is necessary to understand the responses of the ear and the lateral line system. The propagations of sound pressure and particle motion are complicated because of large differences in the attenuation level related to the type of sound source, its frequency, and the distance from the stimulus. 3.2.1 Sound Source and Sound Field Sound propagation speed in air (340 m/s) and water (1500 m/s) are different because the acoustic impedance values (ρc) of sound (where ρ represents the density of medium and c is the velocity of sound) in the two media are different due to differences in

45

46

Locomotion and Sensory Capabilities in Marine Fish

media density. The acoustic impedance is a measure of the total reaction of a medium to sound transmission (i.e., the easiness of sound passing through the specific medium). The acoustic impedances of air and water are about 39.6 g cm−2 s−1 and 150,000 g cm−2 s−1, respectively. Harris and van Bergeijk (1962) described the propagation of sound pressure and water particle motion generated by two types of sound source—a monopole and a dipole source. A monopole sound source is represented by a pulsating air bubble in the water that changes its volume. Sounds generated from underwater speakers and fish gas bladders are considered monopole sounds. Following the expansion of the air bubble, water particles move along the radial direction relative to the center of the bubble, and this movement is transferred from one particle to the next. Displacement of water particles decreases in an inverse proportion to the square of the distance from the sound source. Furthermore, a compression wave (i.e., pressure wave) is also induced during the propagation process, because water has a slight compressibility. As sound propagates, its pressure decreases inversely with increasing distance from the source. Within this sound field, the magnitude is larger for particle motion within the distance of λ/2π where λ is the wavelength, while the pressure is larger outside the distance of λ/2π. The sound field within the distance of λ/2π from the source is termed the near field and the sound field outside of the distance of λ/2π is termed the far field. In the case of dipole sound source, sound is generated by the motion of the object without changing its volume. Displacement of water particles is the largest along the axis of the motion and decreases with the cube of the distance from the sound source. The magnitude of sound pressure, on the other hand, decreases with the square of the distance from the sound source. Hence, the propagation properties of sound pressure and water particle motion are different depending on the sound source. In the meantime, sound frequency affects the velocity and acceleration of particle motion but not the displacement. For example, the velocity of the particle motion u is expressed as u = 2πfd, where f and d are sound frequency and displacement, respectively. Therefore, velocity of particle motion is

higher in high-frequency sounds. In terms of attenuation of a sound pressure level (SPL), high-frequency sounds lose their energy more rapidly than do low-frequency sounds because of the high rate of absorption—that is, the transformation of sound energy to thermal energy. Hence, low-frequency sounds remain in the medium for longer distances, and this is why low-frequency sounds dominate the underwater world (Hawkins 1986). 3.2.2 Sound Pressure Level SPL in water can be calculated with the following equation: SPL ( dB) = 20 log ( p p0 ) where p is the pressure level of the sound in μPa and p0 is the reference pressure level. For underwater sounds, the reference pressure level is 1 μPa; therefore, underwater sound pressure is usually expressed as “dB re 1 μPa.” For sounds in air, p0 is 20 μPa, which is the hearing threshold for human with 1000-Hz sound stimulus. Prior to 1990, underwater acoustic studies used “dB re 1 μbar” as the measurement unit. To convert “dB re 1 μbar” data to “dB re 1 μPa,” 100 dB is added to the “dB re 1 μbar” data—that is, dB re 1 μPa = dB re 1 μbar + 100. 3.2.3 Variations of Underwater Sound Pressure Levels Characteristics of underwater sound vary with the location in water. Hatakeyama (1996) reviewed underwater sound pressure in relation to various sound sources and approximated auditory thresholds of fishes (Fig. 3.1). One of the distinctive differences of the underwater sound is the noise levels in shallow waters compared with those in deep waters. In coastal marine waters, snapping sounds of the pistol shrimp (Alpheoidea spp.) are usually very loud, but such sounds have not been detected in deep water or freshwater (Urick 1983). Hearing thresholds are much lower in otophysan fishes (Ostariophysi: Otophysi), also known as hearing specialists, which account for 64% of the freshwater species (Nelson 1994), than are nonotophysan fishes, which are called as hearing generalists (details described later). Artificial sounds generated

Hearing in Marine Fish and Its Application in Fisheries

47

Figure 3.1. Underwater sound pressure levels of various types of sound sources. (Redrawn from Hatakeyama 1992.)

from underwater piling drilling and dynamite explosions result in high SPLs. In general, highintensity sound is considered aversive for fishes and can cause damage to fish (e.g., dynamite fishing [high sound pressure in coupling with compression waves]). A 200-dB SPL sound would cause an estimated 1010 μPa (or about 100 gf/cm2) force to the fish. For the lowest threshold level of otophysan fishes is 60 dB, the force is about 103 μPa (or about 0.01 mgf/cm2), whereas the lowest threshold level for nonotophysan fishes is 90-dB SPL, equivalent to 3.16 × 104 μPa (or about 0.3 mgf/cm2). 3.3 UNDERWATER SOUND SOURCES AND THEIR CHARACTERISTICS 3.3.1 Natural Underwater Ambient Sounds Underwater ambient noise covers a wide range of frequencies from 1 Hz up to about 100 kHz (NRC 2003; Urick 1983). Urick (1983) classified sources of ambient noise in the ocean into the following six categories:

• Those resulted from tides and hydrostatic pressure changes of relatively large amplitude and at the low-frequency end of the spectrum • Seismic disturbances that generate noise between 1 and 100 Hz • Oceanic turbulence in the form of irregular random water currents of large or small scales (Wenz 1962)—For instance, steady current at 1 knot can generate noise around 106 dB (re 1 μPa). • Ship traffic that generates frequency in the range of 50 to 500 Hz—Such noise can be detected at distances of 1000 miles or more from the site of measurement. • Surface waves that caused noise in the frequencies between 1 and 50 kHz (NRC 2003)—When below 5 to 10 Hz, the dominant ambient noise source was the nonlinear interaction of oppositely propagating ocean surface waves. • Noise caused by precipitation (rain, hail, and snow)—The spectrum of rain noise, for wind speeds below 1.5 m/s, showed a peak at 13.5 kHz

48

Locomotion and Sensory Capabilities in Marine Fish with a sharp cutoff on the low-frequency side and a gradual fall-off (7 dB per octave) on the high-frequency side.

3.3.2 Sounds Produced by Fishes Many fish produce calls as part of a specific behavioral pattern. Sounds are believed to elicit changes in the behavior of other individuals of the same or different species. Sounds vary in structure and characteristics depending on the mechanism used to produce them. In general, two major sound types are produced by fish. The stridulatory sounds result from rubbing hard parts of the body. For example, members of the grunt family, Pomadasyidae, produce a sharp, vibrant call by grating a dorsal patch of pharyngeal denticles against smaller ventral patches. Some catfish (Family: Siluridae, e.g., channel catfish Ictalurus punctatus) produce a squeak when the enlarged pectoral spines are moved against each other (Hawkins 1986; Fine et al. 1997). The second type of sound, the drumming sound, is produced by contraction of skeletal muscles along the body wall against a gas-holding structure, such as gas bladder. The notable cases are sounds produced in oyster toadfish (Opsanus tau), croaker (Micropogonius undulates) (Fine et al. 1997, 2001), and fish of the Family Gadiade, such as haddock (Melanogrammus aeglefinus), cod (Gadus morhua), pollock (Pollachius pollachius), and Tadpole-fish (Raniceps raniceps) (Hawkins 1986). The major difference between stridulation sound and drumming sound is that the former tends to have wider frequency bandwidth than the latter. Recently, fishermen have use the knowledge and new technology (e.g., hydrophone) to locate aggregation of spawners of sciaenid fish in the coastal area of central Taiwan, and this has resulted in a large-quantity catch of sciaenids. This has raised concerns that the wild stock of these sciaenids could be depleted within a short time if the new technology is not properly managed (Tu et al. 2004). 3.3.3 Sounds Produced by Fishing, Research, and Whale-Watching Vessels Modern fishing and research vessels use diesel engines in conjunction with high-thrust propulsion systems that can generate significant levels of noises that are radiated underwater. Field tests

revealed that a ship with the combination of a diesel engine and generator produced noise in the frequency range of 8 to 6 kHz with SPL between 110 and 140 dB (Mitson and Knudsen 2003). The frequency range and SPL of vessel noise are thus in the hearing range of many commercial important marine fish species such as cod (10–600 Hz, 60– 140 dB) and herring (50–1500 Hz, 55–150 dB) (Astrup and Møhl 1993; Chapman and Hawkins 1973; Blaxter et al. 1981; Enger 1967; Sand and Karlsen 1986; Schwarz and Greer 1984). Winger (2004) reported that Atlantic cod responded to an approaching vessel from as far as 1500 m. Over the past few decades, whale-watching has been promoted in many parts of the world and has become an important tourist industry (Hoyt 2000). A field study conducted in the Juan de Fuca Strait area of southern British Columbia and northwestern Washington, where killer whale watching is a significant business, revealed that SPLs of noise generated by the vessels were in the range of 145 to 169 dB (100–20 kHz). The recorded killer whale call source levels were 105 to 124 dB and the audiogram showed best frequency at 20 kHz (range 100– 100 kHz) with a hearing threshold of 40 dB. These data clearly indicated that noises generated from whale-watching vessels could be perceived by whales and most of the fish in the vicinity of the vessels (Erbe 2002). The long-term consequences of noise exposure generated by whale-watching vessels on dolphins and whales remain to be examined. Bottom trawl is carried out by towing a net over the bottom of the ocean (see Chapters 4 and 12). Such operations result in underwater noise when the fishing gear makes contact with the seabed. Buerkle (1968, 1977) reported that Atlantic cod were able to detect noise generated by a bottom trawl at a range of at least 2.5 km. Response of fish as indicated by a change in behavior due to noises from a combination of the trawler and the trawl it was towing has been reported (Winger 2004). 3.4 GENERAL MORPHOLOGY AND FUNCTIONS OF INNER EARS AND ANCILLARY STRUCTURES The inner ear of fish, including elasmobranches, consists of three semicircular canals (with associ-

Hearing in Marine Fish and Its Application in Fisheries

49

Figure 3.2. Anatomical structure of a typical fish (croaking gourmi [Trichopis pumila]) inner ear. The inner ear was stained with osmium tetraoxide to enhance contrast.

ated cristae ampullaris) and three otolithic end organs: the saccule, utricle, and lagena (Fig. 3.2). Despite some variations in the structure of otolithic organs in fishes, the basic functional morphology is essentially the same among fishes (Popper and Fay 1999). Lining some portions of the wall of the organ is a piece of sensory epithelium that contains sensory hair cells and supporting cells. Above the sensory hair cells, the sac also contains an otolith, a calcium carbonate structure that lies close to the sensory hair cells. Differences in density between the otolith and the adjacent sensory hair cells trigger relative movement between sensory hair cells and the otolith when a sound wave passes through the ear. Because of a denser mass of the otolith, its movement is smaller than that of the sensory hair cells, causing the bending of cilia bundles as well as kinocilia on top of the sensory hair cells. The shearing action between sensory hair cells and the otolith generates evoked potentials, which are then transmitted along ascending auditory neuronal pathways to the central hearing structures. In addi-

tion to hearing end organs, in some fish, some ancillary structures such as gas bladder or otic gas bladders also aid hearing by picking up the pressure component of the sound into the ear through direct or indirect contact with the hearing end organs. 3.4.1 Hearing Abilities of Fish Because of the similar acoustic impedance of the fish body and the surrounding water, the fish body is considered transparent to passing sound waves. Due to its physical nature, only lowfrequency sound with high energy can be perceived by such direct stimulation mechanism of sensory hair cells. Therefore, for most fishes that rely on hearing only through particle stimulation mechanism, their hearing ability is limited to a narrow frequency band (less than 1000 Hz) with high sound pressure threshold (as high as 120 dB at the best frequency). Such fishes are hence termed “hearing generalist” species. Certain species, however, evolved mechanisms to enhance their hearing through gas-containing structures that are coupled

50

Locomotion and Sensory Capabilities in Marine Fish

to the inner ears. The low-density gas that is enclosed inside the gas bladder changes volume when sound waves pass through the fish. It is generally believed that the passing sound waves lead to compression and expansion of the gas inside the gas bladder and omnidirectional sound is generated. The transmission of the resonant sound entering into the ears contributes to the hearing ability of fish. However, there are many studies showing exceptions to this generalization, as reported by Connaughton et al. (1997), Barimo and Fine (1998), Yan et al. (2000), and Fine et al. (2001, 2004), in which the gas bladder of weakfish, toadfish, goby, and gouramis does not contribute to auditory function. However, fishes in the superorder Ostariophysi (e.g., cyprinoids, characoids, and siluroids) have a specialized mechanical coupling structure (i.e., the Weberian ossicles) that connect the gas bladder to the inner ear (Furukawa and Ishii 1967). Hence, vibrations caused by the passing sound to the gas bladder are transmitted to the ears and hearing abilities are enhanced. Because of their extended hearing frequency range (up to 8000 Hz in certain catfish) and low thresholds (60 dB in goldfish), these fishes are called “hearing specialist” species. In addition, some species have either an otic gas bladder that is attached directly to the saccule or have embedded the inner ear adjacent to the suprabranchial chamber where a pocket of air is enclosed to pick up pressure component of the sound (Yan 1998; Yan and Curtsinger 2000, Yan et al. 2000). The radiographs of carp (Cyprinus carpio) (a hearing specialist), red sea bream (Pagrus major) (a hearing generalist), and bastard halibut (Paralichthys olivaceus) (a hearing generalist) show that the former two species have gas bladders and the latter does not (Fig. 3.3.). Not all fish with a gas bladder can be classified as a hearing specialist if they lack a mechanical coupling between the gas bladder and the inner ear (e.g., red sea bream). For these species, the pressure component of the passing sound cannot be picked up and transmitted into the inner ear and hence the hearing ability is limited. Figure 3.4 shows audiograms obtained from several commercially harvested species: bastard halibut (Fujieda et al. 1996), red sea bream (Ishioka et al. 1988), jacopever (Sebastes schlegeli) (Motomatsu et al. 1996), walleye pollock (Theragra

Figure 3.3. Soft radiographs of carp Cyprinus carpio (a hearing specialist) (A), red sea bream Pagrus major (a hearing generalist) (B), and bastard halibut Paralichthys olivaceus (a hearing generalist) (C). S indicates swimbladder.

chalcogramma) (Park and Iida 1998), Japanese jack mackerel (Trachurus japonicus) (Babaran et al. unpublished data), spotlined sardine (Sardinops melanostictus) (Akamatsu et al. 2003), and masu salmon (Oncorhynchus masou) (Kojima et al. 1992). Although sound frequencies were limited to 2000 Hz during tests, the audiograms clearly indicate that the most sensitive frequency range lies between 100 and 1000 Hz. Many other fish species, such as cichlid fish (Astronotus ocellatus) and European eel (Anguilla anguilla), are sensitive only to low-frequency sounds (Jerkøet al. 1989; Yan and Popper 1992). Some commercially important species such as dab (Limanda limanda) have a hearing frequency range of 30 to 200 Hz with the

Hearing in Marine Fish and Its Application in Fisheries

Figure 3.4. Auditory threshold curves (audiograms) of (Po) bastard halibut (Fujieda et al. 1996), (P) red sea bream (Ishioka et al. 1988), (S) jacopever Sebastes schlegeli (Motomatsu et al. 1996), (T) walleye pollock Theragra chalcogramma (Park and Iida 1998), (Tj) Japanese jack mackerel Trachurus japonicus (Babaran et al. unpublished), (Sm) Spotlined sardine Sardinops melanostictus (Akamatsu et al. 2003), and (O) masu salmon Oncorhynchus masou. (Kojima et al. 1992.)

best frequency at 100 Hz, whereas the hearing ability of Atlantic cod is limited to between 30 and 400 Hz with best frequency at 100 Hz. The Atlantic salmon (Salmo salar) has a hearing range of 30 to 300 Hz with the best frequency at 150 Hz (Hawkins 1986). Overall, these three important commercial species have very limited hearing frequency range and high hearing thresholds. Because of their limited hearing ability, they are not easily affected by ambient noises of low magnitudes. Hearing abilities of cetaceans are in the range of 50 Hz to 100 kHz with best hearing frequency around 10 kHz to 50 KHz and thresholds around 30 to 50 dB (NRC 2003). Because the echoes they produce have wavelengths greater than the diameter of the filaments or twines of a gillnet, the sound could pass through the filament without producing

51

an echo. Cetaceans relying on echo location can thus swim into the gillnet, resulting in bycatch and mortalities. Two methods of gear modifications have been tested to reduce small cetacean bycatch. Acoustic deterrents or pingers that emit highfrequency sounds were used to alert marine mammals of the presence of the fishing gear (Kraus et al. 1997; Goodson 1997; Mooney et al. 2007). The practices were applied to many European and North American fisheries and showed some successes. However, major drawbacks of using pingers are high cost, habituation of cetaceans, wide variability of success rates, and the unintended “dinnerbell” effect, which in fact resulted in unintended outcome of luring cetaceans to the nets to take fish caught in nets (Trippel et al. 2003). The second method is the use of an alternative gear materials by modifying the characteristics of the filaments of the gillnet to increase their acoustic reflectivity (i.e., target strength [TS]), making them easier to be detected by echo-locating odontocetes. Net alterations included air-filled monofilament nylon, multifilament nets, weighted filaments woven into nets, and adding a filler (e.g., barium sulfate) to the nylon to increase net density (Trippel et al. 2003). The latest tests showed that barium sulfate– and iron oxide–enhanced nets increased reflectivity compared with control nets, with the barium sulfate nets generating the highest TS values. The tested TS values indicated that dolphins should be able to detect these nets in time to avoid contact and entanglement, but porpoises, with typically lower source levels, may not detect nets at a range great enough to avoid entanglement (Mooney et al. 2007). These researches are ongoing regarding different mammals and in different fisheries. Further discussions on this topic can be found in Chapters 7, 8, and 13. 3.4.2 Research Techniques on Auditory Physiology of Fish Traditionally, studies of fish hearing have used behavioral or electrophysiological methods. Behavioral methods are based on conditioning the fish with acoustic signals in conjunction with either reward or punishment (Fay 1969; Yan 1995; Yan and Popper 1991). These psychoacoustical methods are time consuming, with experimental paces dictated by physical or cognitive conditions of the test

52

Locomotion and Sensory Capabilities in Marine Fish

subject. In addition, behavioral responses near the threshold fluctuated greatly even in a specific individual. Therefore, threshold determinations are made with the aid of mathematical probability paradigms. The electrophysiological methods record neuronal activities either from microphonics of auditory organs (Saidel and Popper 1987) or from single-unit recordings, which register single nerve fiber discharge patterns in response to acoustic signals (Enger and Anderson 1967). These two invasive electrophysiological methods have some common limitations. Preparations are rather complicated and invasive surgery is needed. The placement of electrodes is restricted to specific end organs or fibers; therefore, responses do not accurately represent the whole auditory pathways. Lately, a noninvasive auditory brainstem response (ABR) method was developed to measure responses of the whole auditory pathways in fish (Kenyon et al. 1998). Advantages of the ABR protocol include noninvasiveness, rapid completion, and measurement of relevant neuronal cells participated in the signal processing. These advantages are in stark contrast to qualitative psychoacoustical method or invasive electrophysiological method. The ABR method involves simple sedation of test fish, placement of recording electrodes on the cephalic region, and presentation of acoustic stimuli via either airborne sound or waterborne sound to the subject. Recordings of evoked potentials are through electrodes to a two-stage amplifier, averaged, and displayed on a computer screen. Determinations of hearing threshold (in terms of dB; re 1 μPa) of a particular frequency are made using either the traditional visual inspection method (Kenyon et al. 1998) or statistical method (Yan 1998). The whole process of obtaining an auditory tuning curve can be completed in less than 2 hours. The ABR system has been widely used by auditory research scientists around the world since its first publication in 1998. The ABR method has become the de facto standard method for fish auditory physiology research. 3.4.3 Effects of Noise Exposure on Hearing Ability of Fishes Urick (1983) chronicled noises that come from seismic disturbances, oceanic turbulence, ship traffic, surface waves, thermal noise, and coastal

water wave actions as well as biological sounds (e.g., calls of porpoises, noises of a mass of snapping shrimps). Snapping shrimps can produce sound ranging from 700 to 30,000 Hz with SPL as high as 70 dB (re 1 μPa), and the croaking sound of croakers (family: Sciaenidae) had a frequency range of 100 to 3000 Hz (dominant frequency around 200 Hz) with the highest sound pressure greater than 110 dB (re 1 μPa) (Barimo and Fine 1998; Fine et al. 2004; Urick 1983). The ambient noise caused by rain, as recorded in Long Island Sound, New York, showed a frequency range of 700 to 20,000 Hz with sound pressure as high as 85 dB (re 1 μPa) (Urick 1983). These natural or biological underwater sounds generally exert no harm to fish. However, anthropogenic sounds (i.e., man-made noises) have been increasingly become an issue that could harm the welfare of fish and other marine animals (Richardson and Würsig 1997). In a series of pioneering experiments, Scholik and Yan (2001, 2002a, 2002b) and Scholik et al. (2004) demonstrated that exposure to the noise (band width: 100–6000 Hz, with dominant frequency at 1300 Hz) generated by an 55-horsepower (hp) outboard engine hampered the hearing ability of a hearing specialist, the fathead minnow (Pimephales promelas), and led to elevation of the hearing threshold at 1, 1.5, and 2 kHz for 7.8, 13.5, and 10.5 dB, respectively. Further experiments by exposing fathead minnow to white noise (i.e., different frequencies of sound with equal energy) of 142 dB (band width 0.3–4 KHz) for either 1, 2, 4, 8, and 24 hours showed elevation of thresholds at 0.8, 1, 1.5, and 2 kHz. Even 14 days after exposure, hearing thresholds for 1.5 and 2 kHz were still significantly higher than the baseline data. The threshold shifts were, however, not observed in a hearing generalist species, the bluegill sunfish (Lepomis macrochirus) (Scholik and Yan 2002b). These results indicate that hearing specialist species are more vulnerable to negative impacts of prolonged exposure to noise than are hearing generalist species. A follow-up study by Scholik et al. (2004) showed that the negative effect of noise exposure to fathead minnow can be mitigated by feeding diets added with vitamin E at a dose of 450 mg/kg. The rationale of using vitamin E to offset the deleterious effect of reactive oxygen species (ROS) (i.e.,

Hearing in Marine Fish and Its Application in Fisheries free radicals) from the acoustic trauma is based on the chain-braking antioxidant effect of vitamin E to neutralize ROS (Chow et al. 1999). Following Scholik and Yan’s work, the study of noise effects has become a research topic of interest and a number of laboratories have examined the various effects of noise exposure on fish hearing. For examples, the effect of powerboat races in Alpine lake, exposure to laboratory-induced noises, exposure to seismic airguns, and ship noises (e.g., McCauley et al. 2003) have been examined. These studies provide further evidence that underwater noise exerts a negative impact on fish hearing, communication and can causes behavioral changes. Interestingly, two gobies living under the waterfall areas in Italy evolved hearing ability and produce sounds outside the frequency range of the background noise (Lugli et al. 2003). Yan et al. (2006) found that the levels of noises generated from various types of aerators used in aquaculture ponds were between 119 dB and 154 dB. These noise levels were found higher than hearing thresholds of 15 species of fish and shrimp commonly cultured in Taiwan. The long-term effects of noises generated from aerators remain to be investigated and attention should be paid to improve the design of aerators with less noise so as not to exert a negative effect on the growth and welfare of cultured fish and shrimp. 3.5 RESPONSES OF FISH TO SOUND AND ITS APPLICATION IN FISHERIES 3.5.1 Acoustic Attraction Underwater sound travels at a speed of about 1500m/s and may be used to control fish behavior over a longer distance compared with chemical or visual stimuli. Several applications of the use of sound in fishery operation to attract fish have been reviewed by Hashimoto and Maniwa (1964, 1966), and Maniwa and Hatakeyama (1970, 1975). For example, it was reported that fish schools can be driven into the set-net by the vocal sound of Risso’s dolphin (Grampus griseus) and that the yellowtail (Seriola quinqueradiata) could be attracted to the surface from a deep layer on the fishing ground by the swimming and feeding sounds of conspecifics.

53

One of Japanese traditional fishing methods, the “donburi” or “boko,” uses acoustical signals to attract fish to the desired fishing area. This technique is still being used in Shibushi Township, Kagoshima Prefecture, Japan, to harvest demersal fishes, such as red sea bream (Pagrus major). A donburi used in the Kagoshima area is a device consisting of a conical shaped lead, measuring 126 mm high and 47 mm at the bottom diameter and weighing about 880 g (Fig. 3.5). It has a 22-mm diameter and 48-mm-deep opening at the bottom and is held by a 7-m-long rope. donburi sizes vary among fishermen, but it is generally believed that larger ones are better because they can produce higher-intensity sounds than can smaller ones. The donburi is deployed while a fishing vessel is anchored and its engine is turned off. A fisherman throws the donburi so that it hits the water surface perpendicularly, to generate sound and to form a column of tiny bubbles while sinking. Air bubbles rise to the water surface for longer than 10 s. Nonperpendicular casts result in larger bubbles and a weaker sound. A fisherman casts donburi 10 to 20 times during one fishing operation. Fishing is carried out by using a handline while the donburi is cast. Catch rates increase gradually with fishing depth gradually raising from the bottom to a shallow layer by as much as 10 m. The donburi is effective in fishing grounds of 40 to 50 m in depth for red sea bream (Pagrus major), crimson sea bream (Evynnis japonica), threestripe tigerfish (Terapon jarbua), and sharpnose tigerfish (Rhyncopelates oxyrhynchus). To record sounds generated by the donburi, two replicas of fisherman’s gear shown in Figure 3.5 were made (K. Anraku, unpublished data)—one with an opening on the bottom and the other one without the opening. Sound recordings were made under quiet conditions in Lake Ikeda, Kagoshima Prefecture of Japan, at water depth of 30 m. The donburi produced sounds of the highest intensity when it hit the water surface, with sounds generated by air bubbles following (see sonograph in Fig. 3.6A; impact time at about 2.6 s mark). Power spectrum analyses indicate a broadband sound with frequencies ranging from 1 to greater than 10 kHz (Fig. 3.6B, spectrograph).The SPLs were measured (and averaged) at 1, 5, 10, 20, and 30 m, respectively,

54

Locomotion and Sensory Capabilities in Marine Fish

Figure 3.5. Fish attraction device used by Japanese fishermen, locally called donburi and used in Shibushi town, Kagoshima, Japan.

while the donburi was cast repeatedly (Fig. 3.5B). Even at 30-m water depth (Fig. 3.6C), both types of donburi produce sounds greater than 100 dB. Sounds produced by the donburi with a hole in the bottom of it (Fig. 3.6C) were 4.3 dB higher than sounds for the one without a hole. It appears that function of the bottom hole in a donburi is to produce louder sounds. 3.5.2 Use of Sound in Fish Guidance Devices The release of artificially raised fish seedlings into coastal waters has been an integral part of sea ranching operations in Japan for more than 40 years (Shishidou 2002). Proper introduction of seedlings into reef areas is crucial for their initial survival. The Research Institute of Oita Prefectural Government in Japan first applied the sound conditioning method in their marine ranching projects to prevent fish from dispersing and to enhance the recapture rate of released fish (Kamijyo 1998). Here we describe an active fish guidance method that uses an acoustic conditioning technique to transport fish over a long distance (Anraku et al. 2006a,

2006b). The method allows the transport of fish seedlings to a desired location in the sea without physical handling so to minimize physiological stress and physical injury. It combines acoustic, visual, and feeding stimuli to control fish during field transportation. The research involved red sea bream, which is one of the major sea ranching species in Japan. The following descriptions are mainly from work by Anraku et al. (2006a, 2006b). Conditioning About 50,000 juvenile red sea bream of about 50 mm in body length were held in a net cage for conditioning use. They were fed with artificial pellets while a sinusoidal tone burst (pulse width and pulse interval were both 0.75 s) with a frequency of 300 Hz was played to condition the fish with sound and food. By repeating such food and acoustic coupling conditioning, fish learned to associate feeding with acoustic stimuli. Once they were conditioned, the fish would respond to sound stimuli by swimming rapidly close to the water surface even before feed pellets were given. A visual target

Hearing in Marine Fish and Its Application in Fisheries

55

Figure 3.6. (A) Sound wave and (B) sound spectrograph recorded before and after a donburi is cast to sea (impact time at 2.6 s). (C) The mean sound pressure levels recorded at the different water depths, which were generated by a donburi with a hole. Error bar indicates standard deviation of means of 10 measurements. (K. Anraku, unpublished data.)

made of strings of blue plastic ribbons (each string was 15 cm in length) served as the aggregation point for fish. The underwater speaker was suspended in water at all times while the visual target was suspended only during conditioning. Feed pellets were given 0.5 to 1 min after sound was broadcasted. Each training session lasted between 5 and 10 min and a total of 20 to 30 training sessions were conducted each day for 9 days. The fishes’ response to the sound stimuli clearly changed within the first day but at least 3 days of training were required before dense aggregation around the visual target was observed. Once the fish were trained, fish would immediately respond to sound stimuli by increasing their swimming speed and swimming around the visual target in a schooling formation. The fish school followed the target even when its position was changed.

The Guiding Device The fish guidance device (FGD; Zeni Lite Buoy Co. Ltd., Tokyo, Japan) was made from an aluminum frame and equipped with a feeding device and an underwater speaker (Fig. 3.7). The feeding device was made of metal frames (2.5 m in length), PVC pipes, plastic hoses, and a water pump. This feeding apparatus dispenses feed pellets (particle size φ = 1.1–1.3 mm) through the holes (4 mm in diameter) drilled on the PVC pipes using a water pump. The blue plastic ribbons serving as visual targets for conditioning in the cage training stage were attached to the frame to aggregate the fish. Three underwater TV cameras were installed on the FGD to monitor fish behavior during the guidance experiments. An underwater speaker was attached to the towing boat to broadcast specific acoustic signals to the fish.

56

Locomotion and Sensory Capabilities in Marine Fish

Figure 3.7. (A) Structure of the fish guidance device (FGD). (B) The school of fish following the FDG. (C) Underwater camera view of the school following the FDG.

Field Trials in the Open Sea Fish guiding experiments with the use of FGD (Fig. 3.7) were conducted with acoustically conditioned juvenile red sea bream (55-mm mean body length), which were released into Kawashiri Port, Kagoshima, in Japan. Three separate guiding trials were conducted. The first two trials used 1000 conditioned fish, while the third trial used 700 individuals. In the first trial, the guiding distance was not prescribed and food was not given after leaving the port. In the following two trials, the guiding distances were prescribed to 1000 m and 3000 m, respectively, and food was given during the guidance trials. The feeding rate B (g/s) was calculated according to the following formula (Lovell 1977): B = 0.05 × N × W × V D

where N is the number of individuals following the FGD, W (g) is the average body weight of fish, V (m/s) is the towing speed of FGD, and D is the guiding distance (m). Earlier findings showed that satiated fish would not aggregate under the FGD and would not follow a moving FGD (K. Anraku, unpublished data). Daily satiation level of feeding was estimated as 5% of the fish body weight. The fish guiding experiment was terminated when either the distance traveled exceeded the prescribed guiding distance or the majority of the fish left the FGD. Figure 3.7B shows that fish actively swam toward feeding pipes and actively fed on pellets discharged. Fish formed a tightly packed school behind the pipes (Fig. 3.7C). Fish feeding at the front of the school were usually overtaken and then returned to

Hearing in Marine Fish and Its Application in Fisheries the trailing end of the school. Smaller fish with poor swimming ability usually fed on remains of pellets and were seldom seen at the front of the school. The number of fish remained with the FGD after leaving the port is shown in Figure 3.8. The guidance rate is the ratio of the number of fish with the FGD to the initial number. In the first trial (Figure 3.8A) during which no pellets were given, the initial total number of fish with the FGD was only 170 even

Figure 3.8. Changes in the numbers of fish (left y-axis) following the FGD (x-axis, in min) and changes in guidance rates (right y-axis) during guiding experiments in the open sea. (A) Guiding without the use of food pellets; (B) 1000-m guiding trial with food pellets given; (C) 3000-m guiding trial with food pellets given.

57

though a total of 1000 fish were released. This finding indicates that feeding remains an important incentive for fish to follow the FGD during the guidance. The guidance rate rapidly decreased with time elapsed and distance traveled (Fig. 3.8A). In trials B and C, a large number of fish remained with the FGD even after the targeted distance 1000 m was reached. Guidance rates rapidly decreased at the distance between 1000 and 2000 m, during which the towing boat was turning to change direction. The drop in guidance rate over the distance was likely the outcome of many fish that broke away from the FGD due to fluctuations in towing speed. The total distance traveled by the fish was very short in the first trial. Most fish were seen to swim down toward the bottom and formed various smaller schools. However, it was also observed that fish could be guided to a distance as far as 3000 m with a guidance rate as high as 40%. These guidance experiments showed that a large proportion of fish can reliably be guided for about 1000 m when combining acoustic, visual, and feeding stimuli. However, further experiments are needed to sort out the respective roles of acoustic, visual, and feeding cues in fish guiding.

3.5.3 Role of Sound and Hearing in Fish Aggregation Devices A traditional fish aggregation device (FAD) termed “payao” has long been used in certain parts of Philippine waters. A payao is an anchored FAD that is made up of a bamboo raft, anchoring rope, and a cement anchor weight (Fig. 3.9). The bamboo raft provides the buoyancy to the payao structure, serving as a marker to indicate its position and as the attachment for the series of palm fronds that play a role in fish aggregation. The anchor line is made of 14- to 16-mm φ polypropylene rope and tethers the raft to a 500-kg cement anchor. Similar structures have been used in Malaysia and Indonesia, where they are called unjam and rumpon, respectively (FAO; http://www.fao.org/fishery/equipment/fad/ en). Globally, the use of various kinds of FADs is widespread and covers many parts of the world’s oceans (Fonteneau et al. 1999; Kamijyo 1998). Fishing with various forms of FADs has long been practiced in Japan and in Mediterranean

58

Locomotion and Sensory Capabilities in Marine Fish

Figure 3.9. Schematic diagram of a fish aggregation device (payao in the Philippines).

regions, although most of the anchored structures are usually deployed in coastal waters. The practice of tuna fishing with anchored FADs started in the late 1960s when field tests were conducted with the deployment of drifting types of payaos in oceanic waters in the Philippines (Floyd 1985). The successful completion of that experiment led to the proliferation of payaos and a progressive increase in tuna production in the Philippines waters over the past three decades starting in the 1970s. Since then, similar structures were also deployed in Okinawa, Japan, Hawaii, and several other areas in the Pacific, Atlantic, and Indian Oceans. With the use of FADs, large tuna species (yellowfin tuna, Thunnus albacares; bigeye tuna, Thunnus obesus; and skipjack tuna, Katsuwonus pelamis), large pelagic species (dolphinfish, Coryphaena hippurus; blue marlin, Makaira mazarra; and sailfish Istioporus inducus), small tunas (bullet tuna, Auxis rochei; frigate tuna,

Auxis thazard; and kawakawa, Euthynnus affinis), and other small pelagic fish species (scads, Decapterus spp., and bigeye scad, Selar crumenophthalmus) have became regular targeted species (Fonteneau et al., 1999). There seems to be many incentives why fish prefer to associate with FADs. These include the search for food, the use of the FAD as shelter from predators, as a schooling companion, alternative environment, cleaning station, spatial reference, and meeting point of individuals to form schools, and the requirement of some fish to converge in biologically rich environments (Freon and Misund 1999). However, none of these hypotheses can adequately explain how and why fish are attracted to the FADs or how they remain aggregated within the effective area of the FADs. Because the understanding behind these processes is not yet clear, scientists are increasingly looking at the possibility that sound generated from FADs may provide an important acoustic sensory cue to explain the attraction or aggregation of fish near FADs like a payao. Westenberg (1953) previously suggested that the vibrations of the palm leaves of a rumpon in Indonesia may be responsible for the attraction of fish, and called for more research in this area of study. Until recently, however, there were no publications of the sounds generated by anchored FADs like a payao (Dempster and Tarquet 2004). Indeed, in payaos, surface waves and water currents acting on its parts may be responsible for the sounds generated with the structure itself (Dempster and Tarquet 2004). There are at least two possible sources of the sounds generated from the payao. One source is the raft itself, and the other is the anchor line. The mechanisms of sound generation by these two parts of the payao are different. The raft of a payao oscillates with the surface waves in the open sea, generating highly audible sounds. The sound is partly generated when the front end of the raft plows into incoming waves or when the rear end of the raft dips into the water with each passing wave. As it moves with the waves, the entire raft vibrates especially when it splashes down hard on the sea surface. However, field observations reveal that the ability of the raft to generate sound is not necessarily coupled to the big waves; even small waves acting

Hearing in Marine Fish and Its Application in Fisheries head-on against the hollow ends of the bamboo could generate a loud sound that is perceptible to the human ear at distances of greater than 50 m (authors’ personal observation, unpublished data). Meanwhile, the action of tidal currents on the anchor line of the payao also generates sound similar to that of the Aeolian tunes of telephone lines when blown by a consistently strong wind. The sound could also be generated by the anchor line resulting from its lateral vibrations due to the shedding of vortices when acted on by passing water currents. The first field recording of underwater sound ever made near an anchored FAD was made near a payao in the Philippines (Babaran et al., 2008). Payao-generated sounds were recorded with a hydrophone at various depths and distances downstream from the payao. The dominant peak depended on the conditions in the field at the time of recording. For example, during rough weather, the payao generated sound at a level as high as 145 dB close to the raft with a frequency centered around 63 Hz (Babaran et al., 2008). This sound, which was higher than the background noise by about 20 dB, had a limited range and attenuated rapidly with both increasing distances from the raft and increasing depths. The frequency of the gener-

59

ated sound matched with those of the raft’s pitching and rolling motions as recorded by accelerometers. Between these two motions, however, energy due to the former was much more dominant, apparently because of the tendency of the raft to orientate itself parallel to the wind’s direction. The sounds generated by the vibrations of the payao’s anchor line when acted by strong water current were distinct from the sounds generated by the raft’s motion. Measurements of underwater sound generated by the anchor line revealed a dominant peak at 49 Hz with SPLs ranging from 91 to 101 dB as measured 3 m from the anchor line at a depth of 15 m (Fig. 3.10). Moreover, within a range of 40 m from the raft’s position, the recorded SPL of this low-frequency sound did not vary with increasing distance from the payao but decreased with depth (Fig. 3.8) (Babaran et al. 2008). This finding was important because sound resulting from the vibration of the anchor line, which seemed more dominant than raft-generated sound, appeared to be the stimulus used by fish, particularly tuna, when they navigated between anchored FADs (Holland et al. 1990; Oohta and Kakuma 2005). It is interesting to note that the SPLs generated by a payao falls within the hearing range of the fish (as seen in Fig. 3.4).

Figure 3.10. Power spectrum of the sound generated by a fish aggregation device (payao in the Philippines). Sounds are recorded near the payao, at 3-m distance horizontally and 15-m depth, which is set in Panay Bay, Philippines.

60

Locomotion and Sensory Capabilities in Marine Fish

3.5.4 Aversive Sound to Reduce Fish Entrapment in the Cooling Water Intakes Much research to develop techniques to control fish behavior to prevent their entry into water intakes of the power plants has been carried out worldwide (Carlson and Popper 1997). Conventional coalburning and nuclear power plants have to draw in a large volume of water to cool either steam turbines or reactors. Inevitably, some fish are either impinged or entrapped (Grimes 1975; Hanson et al. 1977; Stanford et al. 1982). The compositions of species and economic losses due to the impingement and entrapment of fishes have been assessed, and a case study in Taiwan showed that US$2.5 million yearly economic loss could be due to the intake in the Nuclear Power Plant I (Shao et al. 1990). Additionally, many fish deformities caused by the thermal plume of cooling water discharge have been observed in Taiwan. Four species of marine fish—Jarbua tarpon (Terapon jarbua), large scale mullet (Liza macrolepis), milkfish (Chanos chanos), and grey mullet (Mugil cephalus)—were attracted to the thermal plume at the discharge outlet site of a Nuclear Power Plant II in northern Taiwan, which caused extensive lordosis of vertebral columns. Prolonged exposure to high water temperature caused deficiency of ascorbic acid in the fish’s muscle, which then led to mismatched growth between vertebrates and therefore muscles that resulted in lordosis formation (Shao et al. 1990). Various methods, including mechanical screening devices, electric barriers, strobe light, and sound (EPRI 1992, Humbles 1993), have been deployed to reduce either impingement or entrapment (see summary in Ross and Dunning 1996). Work by Ross and Dunning (1996) demonstrated that by broadcasting high-frequency sound (122– 128 KHz) at a pressure level of 190 dB, alewives (Alosa pseudoharengus) could be driven from the water intake of the James A. FitzPatrick nuclear power plant in Ontario, Canada. In Taiwan, a study using aversive underwater sound to drive fish from intake areas has been undertaken since 2006. First, the audiograms of 11 species of the most frequently entrapped fish were obtained with the aforementioned ABR protocol. Based on their best hearing frequencies and threshold data, randomized pulsed low-frequency sound

(100–2000 Hz interval; 100 Hz duration; 1 s in each frequency) was digitized with a function generator and amplified through an Industrial Power Amplifier (IPA 300T) and broadcast with a underwater speaker (Lubell Labs LL9162; sound pressure of 187 dB, measured with an Okidata SW-1030 hydrophone placed 10 m from the speaker) at the water intake of Nuclear Power Plant II in northern Taiwan coast (Wu et al. 2009). A total of 17 field tests were conducted from November 13, 2006, to February 26, 2008, each with sound on or off for a period of 24 h. During periods when the sound was off, a total of 17 species (1076 individuals) were entrapped. During the periods when sound was on, 10 species (572 individuals) were entrapped. The results indicated that sound significantly reduced the entrapment rate by almost 50% (Wu et al. 2009). The promising results prompted the planning and execution of the long-term use of underwater aversive sound to repel fish from the cooling water intakes and discharge sites of nuclear power plants in Taiwan. 3.6 CONCLUDING REMARKS The underwater world is full of sound and vibration. Fish evolved to have a mechanosensory system to detect both sound and vibration. Hearing generalist fish have a narrower hearing frequency range (less than 1500 Hz) and higher hearing threshold (above 100 dB, re 1 μPa) than do hearing specialist fish (up to 8 kHz and down to 60 dB re 1 μPa). For communication purposes, some fish evolved to produce sound. Unwanted noises generated by fishing, research, and whale-watching vessels can inevitably affect commercially important fish species. Prolonged exposure to noise results in reduced hearing abilities of fish. Fish can be conditioned by coupling the sound with food and visual cues. Devices utilizing this knowledge can be used to guide the fish over a long distance for underwater transport purposes to minimize physiological stress. A donburi fishing method that uses sound generated underwater to attract fish has been used in the Kagoshima area of Japan. Many FADs using sound as the main attraction feature have been deployed by fishermen around the world. Artificial underwater noise has been widely used by power plant operators to drive fish away from cooling water

Hearing in Marine Fish and Its Application in Fisheries intake areas to reduce unwanted impingement and entrapment. REFERENCES Akamatsu T, Nanami A and Yan HY. 2003. Spotlined sardine Sardinops melanostictus listens to 1-kHz sound by using its gas bladder. Fish. Sci. 69: 348–354. Anraku K, Kawamura G, Nakahara M, Shigesato N and Archdale MV. 2006a. Fish behavior controlmethods in marine ranching in Japan-I. Development of conditioning method on the basis of hearingability and auditory behavior. INOC-UMS/BMRI, ICCOSMA. pp 391–397. Anraku K, Makino T, Okawa F, Watanabe K, Masu S, Ozono H, Takeshita H, Kawamura G and Archdale MV. 2006b. Fish behavior control methods in marine ranching in Japan–III. Development of fish guidance device. INOC-UMS/BMRI, ICCOSMA. pp 405–413. Astrup J and Møhl B. 1993. Detection of intense ultrasound by the cod (Gadus morhua L.). J. Exp. Biol. 182: 31–42. Babaran RP, Anraku K, Ishizaki M, Watanabe K, Matsuoka T and Shirai H. 2008. Sound generated by a payao and comparison with auditory sensitivity of jack mackerel Trachurus japonicus. Fish. Sci. 74: 1207–1214. Barimo JF and Fine ML. 1998. Relationship of swimbladder shape to the directionality pattern of underwater sound in the oyster toadfish. Can. J. Zool. 76: 134–143. Blaxter JHS, Gray JAB and Denton EJ. 1981. Sound and startle responses in herring shoals. J. Mar. Biol. Assoc. UK. 61: 851–869. Buerkle U. 1968. Relation of pure tone thresholds to background noise level in the Atlantic cod (Gadus morhua). J. Fish. Res. Bd. Can. 25: 1155–1160. Buerkle U. 1977. Detection of trawling noise by Atlantic cod (Gadus morphua L.). Mar. Freshw. Behav. Physiol. 4: 233–242. Carlson TJ and Popper AN. 1997. Using sound to modify fish behavior at power-production and water-control facilities: A workshop. December 12–13, 1995. Portland State University, Portland, OR. Phase II: Final Report to Bonneville Power Administration, Contract No. 1986BP62611, Project No. 199207101, 360 pp. (BPA Report DOE/ BP-62611-11.) Chapman CJ and Hawkins AD. 1973. A field study of hearing in the cod (Gadus morhua L.). J. Comp. Physiol. 85: 147–167.

61

Chow CK, Ibrahim W, Wei Z and Chan AC. 1999. Vitamin E regulates mitochondrial hydrogen peroxide generation. Free Rad. Biol. Med. 27: 580– 587. Connaughton M, Fine ML and Taylor MH. 1997. The effects of seasonal hypertrophy and atrophy on fiber morphology, metabolic substrate concentration and sound characteristics of the weak fish sonic muscle. J. Exp. Biol. 200: 2449–2457. Dempster T and Tarquet M. 2004. Fish aggregation device (FAD) research: gaps in current knowledge and future directions for ecological studies. Rev. Fish Biol. Fish. 14: 21–42. Enger PS. 1967. Hearing in herring. Comp. Biochem. Physiol. 22: 527–538. Enger PS and Anderson R. 1967. An electrophysiological field study of hearing in fish. Comp. Biochem. Physiol. 22: 517–525. EPRI. 1992. Evaluation of strobe lights for fish diversion at the York Haven hydroelectric project. Electric Power Research Institute EPRI-TR-101703. Final Report. Boston, MA. 120 pp. Erbe C. 2002. Underwater noise of whale-watching boats and potential effects on killer whales (Orcinus orca), based on an acoustic impact model. Mar. Mam. Sci. 18: 394–418. Fay RR. 1969. Behavioral audiogram for the goldfish. J. Audit. Res. 9: 112–121. Fine ML, Friel JP, McElroy D, King CB, Loesser KE and Newton S. 1997. Pectoral spinelocking and sound production in the Channel Catfish Ictalurus punctatus. Copeia. 1997: 777–790 Fine ML, Malloy KL, King CB, Mitchell SL and Cameron TM. 2001. Movement and sound generation by the toadfish swimbladder. J. Comp. Physiol. (A).187: 371–379. Fine ML, Schrinel J and Cameron TM. 2004. The effect of loading on disturbance sounds of the Atlantic croaker Micropogonius undulates: air vs. water. J. Acoust. Soc. Am. 116: 1271–1275. Floyd JM. 1985. Development of the Philippine tuna industry. Pacific Islands Development Program, East-West Center, Honolulu, Hawaii. 60 pp. Fonteneau A, Pallares P and Pianet R. 1999. A worldwide review of purse seine fisheries on FADs. In: Tuna Fishing and Fish Aggregating Devices Symposium. pp.13–25. Caribbean, Martinique, October 15–19, 1999. Freon P and Misund OA. 1999. Dynamics of Pelagic Fish Distribution and Behavior: Effects on Fisheries and Stock Assessment. Cambridge: Fishing News Books. 360 pp.

62

Locomotion and Sensory Capabilities in Marine Fish

Fujieda S, Matsuno Y and Yamanaka Y. 1996. The auditory threshold of the bastard halibut Paralichthys olivaceus. Nippon Suisan Gakkaishi. 62: 201–204. Furukawa T and Ishii Y. 1967. Neurophysiological studies on hearing in goldfish. J. Neurophysiol. 30: 1377–1403. Goodson AD. 1997. Developing deterrent devices designed to reduce the mortality of small cetaceans in commercial fishing nets. Mar. Freshw. Behav. Physiol. 29: 211–236. Grimes CB. 1975. Entrapment of fishes on intake water screens at a stream electric generating station. Chesapeake Sci. 16:172–177. Hanson CH, White JR and Li WH. 1977. Entrapment and impingement of fishes by power plant coolingwater intakes. Mar. Fish. Rev. 1266: 7–17. Harris GG and van Bergeijk WA. 1962. Evidence that the lateral-line organ responds to near-field displacements of sound sources in water. J. Acoust. Soc. Am. 34: 1831–1841. Hashimoto T and Maniwa Y. 1964. Research on the luring of fish schools by utilizing underwater acoustical equipment (1). Tech. Rep. Fish. Boat. 19: 1–12. Hashimoto T and Maniwa Y. 1966. Research on the luring of fish schools by utilizing underwater acoustical equipment (2). Tech. Rep. Fish. Boat. 20: 1–5. Hatakeyama Y. 1992. Hearing abilities of fish. Fish. Eng. 28: 111–119. (In Japanese) Hatakeyama Y. 1996. Hearing abilities of fish and its responses to underwater sound (II). J. Mar. Acoust. Soc. Jpn. 23: 132–139 (in Japanese). Hawkins AD. 1986. Underwater sound and fish behavior. In: Pitcher T (ed). The Behavior of Teleost Fishes. pp 114–151. Baltimore, MD: The Johns Hopkins University Press. Holland KN, Brill RW and Chang RKC. 1990. Horizontal and vertical movements of yellowfin and bigeye tuna associated with fish aggregating devices. Fish. Bull. 88:493–507. Hoyt E. 2000. Whale watching 2000: worldwide tourism numbers, expenditures, and expanding socioeconomic benefits. Yarmouthport, MA: International Fund for Animals Welfare. 157 pp. Humbles G. 1993. Evaluation of Smith-Root type electric fish barrier. Yakima Project, Yakima, WA. 84 pp. Ishioka H., Hatakeyama Y and Sakaguchi S. 1988. The hearing ability of the red sea bream Pagrus major. Nippon Suisan Gakkaishi. 54: 947–951. Jerkø K, Turunen-Rise I, Enger PS and Sand O. 1989. Hearing in the eel (Anguilla anguilla). J. Comp. Physiol. (A).165:455–459.

Kamijyo Y. 1998. Example of marine ranching system (Kaiyou Bokuzyou no Jitsurei). In: Soeda H, Hatakeyama Y and Kawamura G (eds). Hearing Physiology in Fishes (Gyorui no Choukaku Seiri). pp 359–367. Tokyo: Kouseisya Kouseikaku. Kenyon TN, Ladich F and Yan HY. 1998. A comparative study of hearing ability in fishes: the auditorybrainstem response approach. J. Comp. Physiol. (A).182: 307–318. Kojima T, Shimamura T, Yoza K, Okumoto N, Hatakeyama Y and Soeda H. 1992. W-shaped auditory threshold curves of masu salmon Oncorhynchus masou. Nippon Suisan Gakkaishi. 58: 1447–1452. Kraus SD, Read AJ, Solow A, Baldwin K, Spradlin T, Anderson E and Williamson J. 1997. Acoustic alarms reduce porpoise mortality. Nature. 388: 525. Lovell RT. 1977. Fish nutrition: energy. Comm. Fish Farmer. 2: 29–30. Lugli M, Yan HY and Fine ML. 2003. Acoustic communication in two freshwater gobies: the relationship between ambient noise, hearing thresholds and sound spectrum. J. Comp. Physiol. (A).189: 309–320. Maniwa Y and Hatakeyama Y. 1970. Research on the luring of fish schools by utilizing underwater acoustical equipment (3). Tech. Rep. Fish. Boat. 24: 1–5. Maniwa Y and Hatakeyama Y. 1975. Research on the luring and driving away of fish schools by utilizing underwater acoustical equipment (4)–Experiments and practical use in the squid fishing. Tech. Rep. Fish. Boat. 28: 1–22. McCauley RD, Fewrell J and Popper AN. 2003. High intensity anthropogenic sound damages fish ear. J. Acoust. Soc. Am. 113: 638–642. Mitson RB and Knudsen HP. 2003. Causes and effects of underwater noise on fish abundance estimation. Aquat. Living Resour. 16: 225–263. Mooney TA, Au WWL, Nachtigall PE and Trippel EA. 2007. Acoustic and stiffness properties of gillnets as they relate to small cetacean bycatch. ICES J. Mar. Sci. 64: 1324–1332. Motomatsu K, Hiraishi T, Yamamoto K and Nashimoto K. 1996. Auditory threshold and critical ratio of black rockfish Sebastes schlegeli. Nippon Suisan Gakkaishi. 62: 785–790. Nelson JS. 1994. Fishes of the World, 3rd. ed. New York: Wiley and Sons. 624 pp. NRC. 2003. Ocean noise and marine mammals. Washington, DC: National Research Council. 204 pp. Oohta I and Kakuma S. 2005. Periodic behavior and residence time of yellowfin and bigeye tuna associated with fish aggregating devices around Okinawa

Hearing in Marine Fish and Its Application in Fisheries Islands, as identified with automated listening stations. Mar. Biol. 146:581–594. Park Y and Iida K. 1998. Walleye pollock, In: Soeda H, Hatakeyama Y and Kawamura G (eds). Hearing Physiology in Fishes (Gyorui no Choukaku Seiri). pp 223–233. Tokyo: Kouseisha-Kouseikaku. Pitcher TJ. 1979. Sensory information and the organization of behavior in a shoaling cyprinid. Anim. Behav. 27: 126–149. Popper AN and Fay RR. 1999. The auditory periphery in fishes. In: Fay RR and Popper AN (eds). Comparative Hearing: Fish and Amphibians. pp 43–100. New York: Springer-Verlag. Richardson WJ and Würsig B. 1997.Influences of man-made noise and other human actions on cetacean behavior. Mar. Freshw. Behav. Physiol. 29: 183–209. Ross QE and Dunning DJ. 1996. Reducing impingement of alewives with high frequency sound at a power plant in Ontario. N. Am. J. Fish. Manag. 16: 548–559. Saidel WM and Popper AN. 1987. Sound perception in two anabantoid fishes. Comp. Biochem. Physiol. 88: 37–44. Sand O and Karlsen HE. 1986. Detection of infrasound by the Atlantic cod. J. Exp. Biol. 125: 197–204. Scholik AR and Yan HY. 2001. Effects of underwater noise on auditory sensitivity of a cyprinid fish. Hear. Res. 152: 17–24. Scholik AR and Yan HY. 2002a. Effects of boat engine noise on the auditory sensitivity of the fathead minnow, Pimephales promelas. Environ. Biol. Fish. 63: 203–209. Scholik AR and Yan HY. 2002b.The effects of noise on the auditory sensitivity of the bluegill sunfish, Lepomis macrochirus. Comp. Biochem. Physiol. (A).133: 43–52. Scholik AR, Lee US, Chow CK and Yan HY. 2004. Dietary vitamin E protects the fathead minnow, Pimephales promelas, against noise exposure. Comp. Biochem. Physiol. (C).137: 313–323. Schwarz AL and Greer GL. 1984. Responses of Pacific herring, Clupea harengus pallasi, to some underwater sounds. Can. J. Fish. Aquat. Sci. 41: 1183–1192. Shao KT, Lin CP, Ho LT and Lin PL. 1990. Study on the fish communities from northern and southern waters of Taiwan by analyzing the impingement data. J. Fish. Soc. Taiwan 17:73–90. Shishidou H. 2002. Stocking effectiveness of red sea bream, Pagrus major, in Kagoshima Bay, Japan. Fish. Sci. 68(suppl. 1): 904–907. Stanford RM, Jordan SW, Talhelm DR, Liston CR, Korson C and Steinmuller MH. 1982. The bioeco-

63

nomic impact of impingement and entrapment on yellow perch in Lake Erie. N. Am. J. Fish. Manag. 2: 285–293. Trippel EA, Holy NL, Palka DL, Shepard TD, Melvin GD and Terhyne JM. 2003. Nylonbarium sulphate gillnet reduces porpoise and seabird mortality. Mar. Mam. Sci. 19: 240–243. Tu C, Wei RC and Chan HS. 2004. Passive acoustic localization for sciaenid habitat in coastal water of Taiwan. J. Acoust. Soc. Am. 115: 2474. Urick RJ. 1983. Principles of underwater sound. Los Altos, CA: Peninsula Publishing. 423 pp. Wenz GM. 1962. Acoustic ambient noise in the ocean: spectra and sources. J. Acoust. Soc. Am. 34: 1936–1956. Westenberg J. 1953. Acoustical properties of some Indonesian fisheries. J. Mar. Sci. 18:311–325. Winger P. 2004. Effect of Environmental Conditions on the Natural Activity Rhythms and Bottom Trawl Catchability of Atlantic Cod (Gadus morhua). PhD thesis, Memorial University of Newfoundland. 151 pp. Wu YH, Yu HY, Shao IT, Lee ZC, Lin ST,Yan HY, Hsu CH, Lee CP, and Jiang HH. 2009. The method using underwater sound in reducing fish entrainment and impingement at the cooling water inlets of nuclear power plants in northem Taiwan. TaiPower Eng. Month. 733: 108–117. Yan HY. 1995. Investigations of fish hearing ability using an automated reward method. In: Klump GM, Dooling RJ, Fay RR and Stebbins WC (eds). Methods in Comparative Psychoacoustics. pp 263– 276. Basel, Switzerland: Birkhauser Verlag. Yan HY. 1998. Auditory role of the suprabranchial chamber in gourami fish. J. Comp. Physiol. (A).183: 325–333. Yan HY and Curtsinger WS. 2000. The otic gas bladder as an ancillary auditory structure in a mormyridfish. J. Comp. Physiol. (A).186: 595–602. Yan HY and Popper AN. 1991. An automated positive reward method for measuring acoustic sensitivity in fish. Behav. Res. Method. Instru. Compu. 23: 351–356. Yan HY and Popper AN. 1992. Auditory sensitivity of the cichlid fish Astronotus ocellatus (Cuvier). J. Comp. Physiol. (A).171: 105–109. Yan HY, Fine ML, Horn NS and Colon WE. 2000. Variability in the role of the gas bladder in fish audition. J. Comp. Physiol. A. 186: 435–445. Yan HY, Lin SH, Yu HY and Lee WJ. 2006. Analyses on noises generated by aerators and its possible impacts on cultured fish and shrimp. Proc. TaiwanFrance Symp. Marine Biodiver. Sustain. Fish. Maricult. pp. 95–108. Kao-Hsiung, Taiwan.

64

Locomotion and Sensory Capabilities in Marine Fish

SPECIES MENTIONED IN THE TEXT alewife, Alosa pseudoharengus Atlantic salmon, Salmo salar bastard halibut, Paralichthys olivaceus bigeye scad, Selar crumenophthalmus bigeye tuna, Thunnus obesus blue marlin, Makaira mazarra bluegill sunfish, Lepomis macrochirus bullet tuna, Auxis rochei carp, Cyprinus carpio channel catfish, Ictalurus punctatus cichlid fish, Astronotus ocellatus cod, Gadus morhua crimson sea bream, Evynnis japonica croaker, Micropogonius undulates dab, Limanda limanda dolphinfish, Coryphaena hippurus European eel, Anguilla anguilla fathead minnow, Pimephales promelas frigate tuna, Auxis thazard grey mullet, Mugil cephalus jacopever, Sebastes schlegeli

Japanese jack mackerel, Trachurus japonicus Jarbua tarpon, Terapon jarbua kawakawa, Euthynnus affinis large scale mullet, Liza macrolepis masu salmon, Oncorhynchus masou milkfish, Chanos chanos oyster toadfish, Opsanus tau pistol shrimp, Alpheoidea spp. pollock, Pollachius pollachius red sea bream, Pagrus major Risso’s dolphin, Grampus griseus sailfish, Istioporus inducus scads, Decapterus spp. sharpnose tigerfish, Rhyncopelates oxyrhynchus skipjack tuna, Katsuwonus pelamis spotlined sardine, Sardinops melanostictus tadpole-fish, Raniceps raniceps threestripe tigerfish, Terapon jarbua walleye pollock, Theragra chalcogramma yellowfin tuna, Thunnus albacares yellowtail, Seriola quinqueradiata

Part Two Fish Behavior near Fishing Gears during Capture Processes

Chapter 4 Fish Behavior near Bottom Trawls Paul D. Winger, Steve Eayrs, and Christopher W. Glass

4.1 INTRODUCTION A bottom trawl is a towed fishing gear that is designed to catch fish, shrimp, or other target species that live on or in close proximity to the seafloor (Fig. 4.1). The process by which fish enter and are retained involves a complex sequence of fish behaviors in response to the fishing vessel and the various components of the trawl. Observing and understanding these behavior patterns represent a critical step in the effective design of mobile trawling systems. Underwater observations of fish behavior in relation to trawls began as early as the 1960s by researchers in Canada, Scotland, and Russia (Beamish 1966a, 1969; Martyshevskii and Korotkov 1968; Parrish et al. 1969) and soon the elaborate relationship between trawl design and fish behavior began to be articulated (Okonski 1969) together with mathematical models of fish behavior (Foster 1969). Nearly 40 years later, the field continues to grow with much of our current understanding of fish behavior near bottom trawls coming about through the technological advancement of various observational techniques including scuba diving, towed vehicles, underwater cameras, telemetry, and hydroacoustics. For a review of these techniques, see Urquhart and Stewart (1993), Godø (1998), and Graham et al. (2004). In this chapter, we review the current knowledge of fish behavior in relation to bottom trawls, building on the earlier valuable reviews by Wardle (1983, 1986, 1993), Laevastu and Favorite (1988), Engås (1994), Godø (1994), and Glass and Wardle (1995a). We attempt to distil more than 100 studies since the 1960s on fish behavior in response to

visual and auditory stimuli produced by the vessel, doors, sand clouds, sweeps, footgear, and trawl netting. We review the typical patterns of behavior as well as several extrinsic and intrinsic factors known to influence behavior. We also equally emphasize where possible the high degree of between-individual variability in behavioral expression that is often observed. We interpret this variability as differences in tradeoffs (Fernö 1993) at the individual level that minimize costs and maximize benefits. Finally, we broadly extend the application of the economic hypothesis of antipredator behavior (Ydenberg and Dill 1986) initially introduced to the field of fish capture by Fernö and Huse (2003). 4.2 TRAWL GEAR AND TRAWL FISHERIES The historical development of a bottom trawl can be traced back to the early use of beam trawls during the fourteenth century with a series of later technological strides during the Industrial Revolution and post–World War II years (see Graham 2006 for review). It is the principal technique by which most demersal fish and shrimp are captured, accounting for approximately 22% of the world’s fish production (Kelleher 2005). Trawl fishing is practiced by nearly all of the world’s coastal states and can be found in estuaries, coastal regions, and the high seas to depths of 2000 m or more. In many regions of the world, the types of bottom trawls used and their operational techniques have not changed much over the past 50 years. But for

67

68

Fish Behavior near Fishing Gears during Capture Processes

Figure 4.1. Schematic drawing of a complete bottom trawl fishing system. Design, shape, and size of individual components vary depending on the fishery and operation

other regions, particularly those of developed countries, the trawls used today barely resemble those of even 20 years ago (Walsh et al. 2002). Modern designs are more advanced and sophisticated as a result of increasing fuel costs, the need for species and size selectivity, stringent bycatch restrictions, and the necessity to minimize the impact on the environment. Meeting these challenges has led to significant improvements in the way bottom trawls are designed and tested, including advances in computer design, simulation, and physical modeling (Winger et al. 2006). A bottom trawl is designed and engineered as a system of parts that work together with predictable geometry and performance (Fig. 4.1). Depending on the trawl design and targeted species, towing speeds range from 2.0 to 5.0 knots (1.0 m/s to 2.5 m/s). The otter boards or doors provide horizontal spreading force through a combination of hydrodynamic lift and frictional shear with the seabed. They also help keep the trawl net on the seabed and generate a sand cloud to assist herding of fish into the net. Wire sweeps and bridles connect the doors to the wingends of the trawl net and vary in length depending on the fishery, often short (less than 40 m) or non-

existent for shrimp fisheries and long (approximately 350 m) for flatfish fisheries. The trawl body consists of a series of netting panels selvaged together. The design, shape, and dimensions of these panels help define the overall shape and performance of the net. The vertical opening of the mouth of the net is usually achieved by attaching positive buoyancy (e.g., floats) or hydrodynamic kites to the headline, although in some shrimp fisheries direct attachment of the headline to the otterboard means the height of the net is equivalent to the height of the otterboard. To maintain seabed contact, chain, rubber discs, or steel bobbins are attached to the fishing line and/or bolsh line of the net. The process by which fish enter and are retained by a net involves a complex sequence of fish behaviors. A helpful approach for understanding and describing this process is to compartmentalize it into three zones as shown in Figure 4.2. Zone 1 describes fish behavior in the pretrawl zone, including their initial detection and reaction to the lowfrequency noise produced by the vessel, as well as avoidance behavior in response to the trawl warps. Zone 2 describes fish behavior in response to the trawl doors, sweeps, and net mouth. And, finally,

Fish Behavior near Bottom Trawls

69

Figure 4.2. Schematic drawing of a bottom trawl illustrating the different zones that fish may encounter during the capture process. (Adapted from Walsh 1996.) Hatched areas represent the “sweep zones”—fish in these zones must be herded or guided into the net path to become vulnerable to capture.

Zone 3 describes fish behavior once inside the trawl body. 4.3 FISH BEHAVIOR IN THE PRETRAWL ZONE (ZONE 1) 4.3.1 Underwater Radiated Noise The fish capture process begins well ahead of the vessel, where fish initially detect and respond to low-frequency noise produced by the vessel, warps, doors, and trawl. The combination of these sounds produces an underwater-radiated noise signature that is highly specific to each vessel and trawling operation. In most cases, this noise signature is dominated by low-frequency tones (10–10,000 Hz) produced mainly by the vessel. In fact, detailed measurements of vessel-radiated noise alone have shown extreme variation in the frequency spectrum and pressure levels between vessels (Mitson 1993; Mitson and Knudsen 2003). Factors that determine a vessel’s noise signature include cylinder firing rate, engine load, gearbox configuration, and propeller pitch (see Anon 1995 in press for review). Additional noise is also produced by the vibrations of trawl warps, door contact with the seabed, and

trawl components as they pass through the water (Buerkle 1977; Handegard et al. 2003; Handegard and Tjøstheim 2005; Mitson 1993), all of which contribute to the overall noise signature. For many species of fish, these low-frequency sounds occur directly within their hearing range. Over the past few decades, audiograms of hearing sensitivity have been collected for a number of fish species (see Anon 1995 in press; Fréon and Misund 1999; Popper 2003 for reviews). Figure 4.3 illustrates a few examples of the relationship that exists between hearing threshold (dB) and sound frequency (Hz). Popper et al. (2004) recently argued that fish can be grouped as hearing “specialists” or hearing “nonspecialists” (i.e., generalists). The hearing generalists detect sounds up to 1500 Hz and have lower sensitivity (higher thresholds). Hearing specialists, by comparison, detect sounds of 3000 Hz or above and have better sensitivity (lower thresholds). While this species-specific variation in hearing capability is probably the result of variation in selection pressure over time (Manley et al. 2004), it has immediate and important relevance to how well fish detect approaching fishing gear, particularly vessels and trawls.

70

Fish Behavior near Fishing Gears during Capture Processes

Figure 4.3. Audiograms of fish hearing sensitivity for several species, including Atlantic cod (Gadus morhua), Atlantic salmon (Salmo salar), American shad (Alosa sapidissima), herring (Clupea harengus), North Sea plaice (Pleuronectes platessa), common dab (Limanda limanda), North Sea pollack (Pollachius pollachius), and goldfish (Carassius auratus auratus). (Data from Mitson 1993; Popper 2003.)

Once the dB–frequency relationship is known for a species (e.g., Fig. 4.3), it is possible to calculate the theoretical distance that the species can detect vessel/trawl noise (see Mitson 1993). For example, Atlantic cod (Gadus morhua) have one of the lowest thresholds (Fig. 4.3) and are therefore capable of relatively long-range detection (Sand and Karlsen 1986) for certain frequencies. It has been estimated that cod can detect trawling noise from a minimum distance of 3.2 km during summer, reducing to 2.5 km during winter due to increased ambient noise in the sea at that time of year (Buerkle 1977). The main question remains, how well can fish discriminate the direction and distance from which a vessel/trawl is approaching? This cannot be an easy task given the diffuse and complex sounds emitted from an approaching vessel and trawl. Several studies have attempted to investigate different aspects of sound source localization in haddock (Melanogrammus aeglefinus), pollack (Pollachius

pollachius), ling (Molva molva), walleye pollock (Theragra chalcogramma), and cod (Gadus morhua) (Chapman 1973; Chapman and Johnstone 1974; Hawkins and Sand 1977; Mann et al. 2009; Olsen 1969; Schuijf 1975). Using elegant behavioral or electrophysiological methods, these studies have demonstrated rather convincingly the capacity for sound source localization in many species, in both the near- and far-field. But these have mainly been in response to pure-tones under laboratory conditions, not the diffuse and complex sounds of a vessel and bottom trawl. Fay (2005) reviews several theories as to “how” fish collect and process sounds from their environment and concludes that our knowledge of sound source localization in fish remains incomplete, thus identifying a key field for future investigation. 4.3.2 Reaction Distance A significant body of evidence suggests there is large difference between the distance at which fish

Fish Behavior near Bottom Trawls first “detect” approaching sounds and the distance at which they “react” to approaching sounds. In other words, their detection thresholds and reaction thresholds are very different. Detection thresholds are determined by the physics of the inner ear and are likely to be predictable within a species, size class, ambient noise in the water, and physical property of the water. Response thresholds by comparison, are the outcome of a behavioral tradeoff. They are the behavioral expression made by individual fish that are attempting to minimize costs and maximize benefits in response to an approaching threat (Fernö 1993; Godin 1997). It was recently predicted (Fernö and Huse 2003) that the timing of reaction by fish to an approaching vessel and trawl should follow similar economics to prey fleeing from predators. If this is true, it opens the tantalizing opportunity to apply basic principles from established predator–prey theory and, in particular, the growing field of “optimal escape theory,” initiated by Ydenberg and Dill (1986). Since their influential model was published more than 20 years ago, dozens of behavioral studies have found the model to be robust in its ability to predict relative reaction distances to

71

threats across species and taxa (see reviews by Cooper and Frederick 2007; Stankowich and Blumstein 2005). In general, when a fish detects a threat, usually an attacking predator, it must sequentially decide (1) whether and when to flee, (2) in which direction to flee, (3) how fast to flee, and (4) how far to flee. In Ydenberg and Dill’s (1986) economic (costbenefit) model, a fish under threat of a predator continually chooses between two behavioral options, staying where it is (and perhaps continuing with an ongoing activity) or fleeing, as the distance between it and the predator shrinks. The distance at which the fish flees (i.e., reaction distance) is determined by a balance between the costs of these two options (Fig. 4.4). The costs of fleeing (F) are assumed to increase linearly and the costs (or risk) of remaining (R) to decrease proportionately, with increasing distance to the predator. Timing is critical as fleeing too early may result in lost opportunities (high F) and remaining too long may result in injury or death (high R). A fish under threat is assumed to reassess its choice of action moment by moment as the distance between itself and the threat changes and should opt for the behavior with the

Figure 4.4. Ydenberg and Dill’s (1986) economic model of reaction distance (D) for fish under the threat of a predator. (Adapted from Godin 1997.) We broadly extend this model to help develop predictions about response thresholds and corresponding optimal reaction distances for fish in response to a bottom trawl. See text for details.

72

Fish Behavior near Fishing Gears during Capture Processes

lowest cost. The animal should opt for remaining when F > R and opt for fleeing when R > F. The optimal reaction distance (D*) is defined as the intersection of the two curves (see Fig. 4.4). It is the point where the costs of remaining just balance the costs of fleeing (i.e., escaping). The model predicts that the optimal reaction distance should increase with increasing cost of remaining (D2* → D3*) and decrease with increasing cost of fleeing (D2* → D1*). As a trawl approaches an aggregation of fish in the wild, fish in the pretrawl zone (Zone 1) are faced with the important decision of whether and when to react. The optimal reaction distance (D*) should be determined by a balance between the costs of fleeing (F) and remaining (R). Factors that define the R cost curve will be related to the “apparent risk” of the vessel and trawl (e.g., type, age, speed, engine load, and propeller pitch, trawl rigging, performance, and operation). The F cost curve will be defined by lost opportunities (e.g., foraging and spawning) and energy expenditure. Layered over this, of course, are then the several extrinsic and intrinsic factors that are expected to modify behavioral expression (see Section 4.6). It is quite plausible to expect that each individual fish could have a unique intersection of the curves and therefore a unique optimal reaction distance (D*). Several studies have attempted to empirically measure the reaction distance of demersal species in response to approaching vessels and bottom trawls. While most of the early research used traditional echo-sounders to monitor the response of whole aggregations (e.g., Olsen et al. 1983a; Ona and Godø 1990), recent approaches now monitor the movements of individual fish using either splitbeam echo-sounders (e.g., De Robertis et al. 2008; Handegard et al. 2003; Handegard and Tjøstheim 2005; Hjellvik et al. 2003; Jørgensen et al. 2004; McQuinn and Winger 2003; Ona et al. 2007) or acoustic telemetry of individual fish tagged with acoustic transmitters (Engås et al. 1998; HardenJones et al. 1977; Winger 2004). Both have proven highly effective at estimating reaction distances and determining avoidance patterns. The individual reaction distances of acoustically tagged cod were studied by Engås et al. (1998) and Winger (2004) in response to an approaching

vessel and bottom trawl. Both studies used an acoustic positioning system (Voegeli et al. 2001) to monitor the behavior of individual fish for the periods before, during, and after an encounter (Fig. 4.5). Beginning at a distance of 1 to 3 km the array, a vessel course was plotted to take the vessel and trawl directly over the fish with the closest possible precision. The vessel then exited the opposite end of the array and took up position at a distance of 1 to 3 km. Engås et al. (1998) found that cod were capable of initiating avoidance responses at distances ranging from 470 to 1470 m. These results were the first of their kind, demonstrating that reaction distances in some cases can occur at distances greater than 1 km. The results were verified in a similar study by Winger (2004), which found reaction distances between 206 and 1512 m. Figure 4.6 illustrates various changes in swimming speed observed in response to an approaching vessel at different speeds, as well as a vessel towing a trawl. The corresponding right axis shows the distance between the fish and the approaching threat when the change in swimming speed occurred (i.e., reaction distance). Together, these studies indicate that cod are capable of detecting and reacting to an approaching trawler from considerable distances and that the optimal reaction distance can be highly variable depending on the behavioral tradeoff chosen by individual fish. Further discussions on extrinsic and intrinsic factors that are known to affect reaction distance are given in Section 4.6.

4.3.3 Avoidance Patterns Once a fish has detected and decided to react at some optimal distance to the threat of a fishing vessel (see earlier), it must decide in which direction and how fast to swim. The traditional view has long been held that fish react with a graduated response, beginning first with a slow adjustment in swimming direction away from the approaching stimulus (Olsen et al. 1983a). Several studies have provided empirical data to support the theory, including those of Ona and Godø (1990) and Nunnalle (1991), which both documented a density draining phenomenon in advance of an approaching vessel, indicative of horizontal avoidance. However,

Fish Behavior near Bottom Trawls

73

Figure 4.5. Illustration of the experimental setup used by Engås et al. (1998) and Winger (2004) to investigate the individual reaction distances of acoustically tagged cod in response to an approaching vessel and trawl.

recent evidence suggests that for some vessels, certain species may actually swim in toward the vessel path as it approaches (Handegard and Tjøstheim 2005). For the individual fish, the decision on which direction to swim in response to an approaching vessel/trawl is complicated by the nonuniform propagation of noise intensity. As a vessel and trawl move along a given trajectory, high and low areas of noise intensity systematically form and collapse in response to the moving hull, warps, and trawl. The hull’s ability to shadow propeller

cavitations produces large lobes of high-intensity noise on the vessel’s port and starboard sides. The resulting “butterfly pattern” (see Anon in press; Misund 1994) has been shown in some studies to induce intense outward horizontal movement toward the port and starboard (Engås 1994; Soria et al. 1996), to attract fish inward toward the vessel track in other studies (Handegard and Tjøstheim 2005), and even to trap fish into swimming in the forward direction in line with the vessel’s track, known as the pursuit effect (Misund 1994; Misund et al. 1996).

74

Fish Behavior near Fishing Gears during Capture Processes

Figure 4.6. Examples of Atlantic cod behavior in response to an approaching vessel (top and middle) and vessel plus trawl (bottom). Top, Increase in swimming speed in response to a vessel approaching at 10.0 knots. Middle, Decrease in swimming speed in response to a vessel approaching at 4.5 knots. Bottom, Increase in swimming speed in response to a vessel and trawl approaching at 3.0 knots.

In addition to horizontal movement, fish occurring in the pelagic zone may also dive vertically toward the seabed in response to vessel-radiated noise. This has been documented for several species including capelin, herring, anchovy, common sardine, haddock, and cod (Gerlotto et al. 2004; Handegard et al. 2003; Hjellvik et al. 2003; Olsen et al. 1983b; Ona and Godø 1990; Michalsen et al. 1999; Vabø et al. 2002). In a recent study, Handegard and Tjøstheim (2005) used a free-floating splitbeam echo-sounder system to examine the simultaneous horizontal and vertical displacements of individual fish in response to an approaching vessel and trawl. The authors provide an impressive threedimensional model of the directional behavior of gadoids built using multiple years of experimental data. One of the key findings was that vertical displacements of fish (diving) tended to occur at the start of the tow when vessel noise drops markedly. They also noted the strongest and sharpest responses toward the warps. These findings contradict earlier indications that the timing of diving occurs in response to the gradual rise in vessel noise caused by an approaching vessel.

Finally, avoidance patterns in the pretrawl zone are often characterized by a change in swimming speed. According to the avoidance model proposed by Olsen et al. (1983a), changes in swimming speed should occur when a change in swimming direction alone has been insufficient in reducing the approaching threat. Field studies investigating the swimming speeds of fish in response to an approaching vessel/ trawl have commonly reported increases in swimming speed (e.g., Handegard et al. 2003; Handegard and Tjøstheim 2005; McQuinn and Winger 2003; Michalsen et al. 1999; Olsen et al. 1983a; Winger 2004) but also in rare occasions observed decreases in speed (Engås et al. 1998; Winger 2004). Figure 4.6 illustrates a few examples of individual cod modifying their swimming speed in response to an approaching vessel at different speeds, as well as a vessel towing a trawl. Figure 4.6, A and C, reveals a more than doubling of swimming speed, whereas Figure 4.6B illustrates an example of a fish that reduced its swimming speed from roughly 20 cm/s to near zero, well in advance of the approaching threat (approximately 1500 m). Both of these avoidance patterns are considered to be adaptive from a

Fish Behavior near Bottom Trawls predator-evasion point of view (Lima and Dill 1990) and are probably reflective of differences in distance to shelter/cover for these particular fish, hence affecting the steepness of the F-cost curve, and their likely reaction distance (Fig. 4.4).

4.4 FISH BEHAVIOR BETWEEN TRAWL DOORS AND IN THE NET MOUTH (ZONE 2) As a trawl begins to move through an aggregation of fish, a proportion of these fish will enter between the doors. Fish located in this zone will be either directly in the path of the net itself or in one of the two sweep zones between the wings of the net and trawl doors (see Fig. 4.2). The net path is defined as the area swept between the wingends of the trawl net. Fish located in this zone are directly available for capture by the trawl net. However, fish in the

75

two sweep zones must be herded (or guided) into the net path to become available for capture. 4.4.1 Herding Patterns: Roundfish Roundfish that are located near the seafloor in advance of an approaching trawl tend to respond to trawl doors as soon as their presence is sensed visually. Naturally, this is dependent on both the visual field and the visual range of the species. Observations have shown that fish tend to choose the path of least resistance in this situation. That is, they swim in a manner that maintains the threat at the edge of their visual range, keeping at least one eye on the door at all times as it passes by. This behavior results in what is known as the “fountain maneuver” (Fig. 4.7). The process unwittingly guides many of the fish directly into the path of the net, increasing their vulnerability to capture (Hall et al. 1986; but see Wardle 1993 for detailed description).

Figure 4.7. Typical fountain maneuver of roundfish in response to trawl doors and their subsequent herding behavior into the trawl mouth. Dotted lines represent the limit of the visual range of fish, which varies across species and, most important, with the underwater light field. This behavioral model is based on a series of observations from several studies conducted over time and illustrates common behavior patterns observed at high light intensities.

76

Fish Behavior near Fishing Gears during Capture Processes

Fish that swim around the outside of the doors effectively escape capture and are unlikely to reenter the capture zones. Fish that swim around the inside of the doors tend to immediately enter the path of the net and begin swimming toward the net mouth, maintaining position between the sweeps and sand clouds. In many cases, the visual stimuli of the netting, floats, and footgear are still outside the visual range, presenting what appears to be a clear route to safety. It is not until these components enter the visual range of the fish (denoted by the second dotted line in Fig. 4.7) that fish tend to alter course and begin swimming in the direction of tow in the net mouth. Factors affecting the fountain maneuver as well as the degree of variation within and between species are not fully understood. Experiments have demonstrated that the herding efficiency of some species can be improved by increasing the length of sweep between the door and wingend (i.e., sweep length). Engås and Godø (1989a) found that catch rates for large cod and haddock generally increased with increasing sweep length from 20 to 120 m. They also found that smaller fish (less than 29.5-cm length) were generally underrepresented with increasing sweep length, indicating the process can also be size selective for roundfish (see discussion by Engås 1994). Somerton (2004), by contrast, was unable to duplicate this finding for either Pacific cod (Gadus macrocephalus) or walleye pollock and concluded that these species have weak herding responses. Co-related to sweep length is the angle of attack of the sweep to the direction of tow (sweep angle), which is typically between 10 and 20 degrees. Strange (1984) found that the catching efficiency for cod and haddock was reduced at sweep angles greater than 20 degrees. At such large angles of attack, the sand clouds are likely to deviate from the sweep and trail well outside of the net. This leaves only the relatively inconspicuous sweep to herd fish toward the wing-end, reducing overall herding effectiveness. 4.4.2 Herding Patterns: Benthic Species For benthic species such as flatfish, monkfish, and skates, herding is typically induced after direct or near contact with the doors, sand clouds, and sweeps (see review by Ryer 2008). Reaction distances are

generally short, and in most cases fish are seen swimming in close proximity to the sweep (Fig. 4.8). Under sufficient light conditions, animals will typically select a swimming trajectory that is approximately perpendicular (90 degrees) to the advancing sweeps. This results in the fish slipping along the sweep toward the path of the net (Fig. 4.9A), increasing their vulnerability to capture. However, not all fish will behave this way, and swimming trajectory can be somewhat variable among individuals and among species (S.J. Walsh and P.D. Winger, unpublished observations). Individuals may select trajectories that do not direct them into the path of the net, thus evading capture (Fig. 4.9B). With the exception of Reid et al. (2007), empirical estimates of swimming trajectory and its variability have not been adequately estimated for many species, highlighting another key area for future investigation. Behavioral observations have also revealed that the choice of swimming behavior can be somewhat variable (Harden-Jones et al. 1977; Hemmings 1969, 1973; High 1969; Main and Sangster 1981; Reid et al. 2007; Ryer and Barnett 2006). Once disturbed from the seafloor, individuals may choose to (1) swim slower than the speed of the advancing sweep, in which case they would be overtaken and escape from the gear, (2) swim continuously at the same speed as the sweep, “keeping station” a certain distance ahead of the stimulus, or (3) swim at a speed greater than the sweep for a period of time and then slow down or settle onto the seafloor. Only in the latter two cases do fish progressively slip along the sweep toward the mouth of the trawl, increasing their vulnerability to capture as they move toward the net path. Several authors have attempted to mathematically model these different behavior patterns (Foster 1969; Foster et al. 1981; Fuwa 1989; Fuwa et al. 1988; Reid et al. 2007; Tanaka et al. 1991). The models demonstrate that herding efficiency is highly sensitive to subtle changes in behavior (i.e., swimming continuously versus periodic settling), indicating that the choice of swimming behavior probably determines the likelihood of capture. Functional explanations for the three behavior patterns are still unresolved and the degree of variation within and between species is unknown, although recent laboratory observa-

Fish Behavior near Bottom Trawls

77

Figure 4.8. Typical herding behavior of benthic species such as flatfish and skates in response to the sweep of a bottom trawl. Reaction distances are often low and most animals are seen swimming approximately perpendicular (90 degrees) to the sweep.

tions suggest species-specific differences are likely and that extrinsic factors such as ambient light level and water temperature are important proximate factors (Ryer and Barnett 2006). Finally, for herding to be effective, these fish must also have sufficient endurance to reach the net path. Assuming most fish swim perpendicular (90 degrees) to the advancing sweep, the “herding speed” will be determined by the forward towing speed of the trawl and the sweep angle (i.e., angle of attack of the sweep to the direction of tow). Depending on trawl rigging and operation, herding speeds usually range between 0.2 and 0.6 m/s. Given that all fish are stimulated to swim at the

same speed, fish of different sizes will operate at different levels within their performance range and exhibit different gaits (Winger et al. 2004). The distance required to swim to reach the net path is determined by the sweep angle and the position along the sweep where the fish initially encounter the gear. The probability of successfully swimming this distance is thought to be size and temperature dependent (see Section 4.6). 4.4.3 In the Trawl Mouth Video and photographic observations have repeatedly shown that behavioral diversity is greatest in the trawl mouth. Different species and sizes of fish

78

Fish Behavior near Fishing Gears during Capture Processes

Figure 4.9. Under sufficient light conditions, most flatfish and skates will select a swimming trajectory that is perpendicular (90 degrees) to the advancing sweeps. This results in the fish slipping along the sweep toward the path of the net (A). In some cases, individuals may select swimming trajectories that never increase their vulnerability to capture (B). Unless these individuals change their swimming trajectory, they might never transition into the path of the net.

all arrive here in varying physiological conditions depending on their avoidance behavior in Zone One, as well as their subsequent herding behavior in Zone Two. The trawl mouth is the melting pot where everything comes together at high speed and quick decisions have deliberate outcomes. Because of the high variability in behavior in this region of the gear, several authors have attempted to analyze the behavior by coding or classifying patterns of behavior into meaningful categories (Albert et al. 2003; DeAlteris et al. 1992; Eayrs and Piasente 2006; Piasente et al. 2004; Walsh and Hickey 1993). Once this is done, the frequency and duration of the patterns can be quantified, making interpretation considerably easier. For many species, the most common response is to orient in the direction of tow and keep station with the advancing trawl (Fig. 4.10). This behavior

occurs at high light intensities and appears to be an optomotor reflex in response to the visual cues (contrast) produced by the surrounding footgear and netting panels (see Glass and Wardle 1989; Main and Sangster 1981; Walsh and Hickey 1993; Wardle 1993; also see Chapter 2). This behavior is similar to that observed in the laboratory where fish can be induced to follow moving stripes and patterns projected on the floor and walls of a tank (Breen et al. 2004; Harden-Jones 1963; He and Wardle 1988). For some species, this reflex produces a strong motivation to maintain position (or station) within the mouth of a moving trawl, whereas for other species, the response is less apparent (or nonexistent), producing a more erratic behavior in the trawl mouth (Kim and Wardle 2003). As light intensity drops, ordered patterns of behavior eventually cease. Glass and Wardle (1989)

Fish Behavior near Bottom Trawls

79

Figure 4.10. (A, B) Typical behavior of fish in the mouth of a bottom trawl at high light intensities. Fish are seen swimming forward in the direction of tow, immediately ahead of the footgear. (Photograph from Glass and Wardle 1989.)

and later Walsh and Hickey (1993) observed several demersal species at various angles to the approaching gear, showing little evidence of reaction at low light intensities (Fig. 4.11). Under these conditions, reaction thresholds are relatively high and the corresponding reaction distances short, in some cases even resulting in collision with the gear (Fig. 4.12). See Section 4.6 for a detailed discussion on the role of light level, color, and contrast. Naturally, not all species are created equal in their swimming capability and performance. The period of time that fish are capable of swimming in the trawl mouth, and hence their vulnerability to

capture, is heavily dependent on the towing speed of the trawl, which is carefully chosen to match the sustained swimming speed range of the targeted species (see Chapter 1). Several studies have managed to visually observe the endurance of different species swimming in the mouth of a trawl under different contexts. Reported values range from as low as a few seconds for jack mackerel (Trachurus japonicus) (Martyshevskii and Korotkov 1968), 3 to 4 s for skates (Raja sp.) (DeAlteris et al. 1992), 5 to 6 s for Spanish sardine (Sardinella anchovia) (Korotkov 1970), 1 to 10 s for Greenland halibut (Reinhardtius hippoglossoides) (Albert

80

Fish Behavior near Fishing Gears during Capture Processes

Figure 4.11. (A, B) Typical behavior of fish in the mouth of a bottom trawl at low light intensities. Fish are oriented at various angles to the approaching gear, showing no evidence of reaction. (Photograph from Glass and Wardle 1989.)

et al. 2003), 2 to 12 s for various flatfish in the Gulf of Alaska/Bering Sea (Bublitz 1996), up to 10 s for blue grenadier (Macruronus novaezelandiae) (Piasente et al. 2004), up to 1.0 min for tiger flathead (Neoplatycephalus richardsoni) (Piasente et al. 2004), up to 1.5 min for dogfish (Sqaulus acanthias) (DeAlteris et al. 1992), up to 2.0 min for sand flathead (Platycephalus bassensis) (Yanase et al. 2009), 2.5 min for haddock (Main and Sangster 1983), 3.0 min for squid (Loligo pealeii) (Glass et al. 1999), up to 8.0 min for Pacific halibut

(Hippoglossus stenolepis) (Rose 1995), and up to 15 min for saithe (Pollachius virens) (Main and Sangster 1983). However, in situ estimates of swimming endurance in the trawl mouth are often costly to obtain and difficult to interpret. With this in mind, several studies have investigated the endurance of commercially targeted species under controlled laboratory conditions for the direct application to trawling (Beamish 1966b; Breen et al. 2004; Chandler 1967; Dogˇanyilmaz-Özbilgin et al. 2006; He 1991; He

Fish Behavior near Bottom Trawls

81

Figure 4.12. (A, B) Collision and escapement of fish under the footgear of a bottom trawl. The process is known to be species and size selective as well as dependent on ambient light intensity. (Photograph from Walsh and Hickey 1993.)

and Wardle 1988; Winger et al. 1999, 2000). Using swimming flumes or large annular tanks, researchers have investigated the period of time that fish are capable of swimming at specific speeds before fatiguing. Once the endurance relationships are defined, the probability of swimming a given period of time can be predicted for various swimming/ towing speeds. Figure 4.13 illustrates various endurance probability curves for Atlantic cod at speeds comparable to those experienced in the trawl mouth. These curves show that the probability of cod achieving a given endurance decreases rapidly with increasing swimming speed. For example, the

probability that cod has endurance greater than 2 min in the trawl mouth is 85% at 1.1 m/s, dropping to 35% at 1.3 m/s, and 0% at speeds equal to or greater than 1.5 m/s. The findings demonstrate that for this species, endurance is expected to be highly sensitive to changes in towing speed, and that subtle changes in the speed of the trawl through the water could have a significant effect on the turn-over rate of cod in the trawl mouth (see Winger et al. 2000 for more details). The point at which fish cease swimming in the trawl mouth is triggered by a behavioral decision. In most cases, we assume it is coupled with the

82

Fish Behavior near Fishing Gears during Capture Processes

Figure 4.13. Estimated probability curves for the swimming endurance of Atlantic cod at swimming speeds comparable to those experienced in the trawl mouth. (Data from Winger et al. 2000.)

onset of metabolic exhaustion as it is often evidenced by a change in gait from steady to unsteady swimming just prior to falling back into the trawl (e.g., Main and Sangster 1983). However, recent studies have clearly shown that “exhaustion” and “fatigue” may not necessarily be one and the same (see Peake and Farrell 2006). Exhaustion refers to the state in which a fish has fully depleted its stored energy, whereas fatigue is a behavioral decision that can occur in advance of energy depletion. Breen et al. (2004) found under laboratory conditions that seemingly exhausted haddock were in fact not exhausted at all, just “unwilling” to continue swimming. This suggests that the mechanism associated with the point at which the fatigue decision is invoked while swimming in the trawl mouth is undoubtedly complex and cannot be entirely explained by metabolic exhaustion alone. This is, of course, supported by the fact that fish are often observed swimming actively once they have fallen back into the trawl (e.g., Eayrs and Piasente 2006; Grimaldo et al. 2007; He et al. 2008; Jones et al. 2008; Piasente et al. 2004). For further discussion, see Ryer (2008). Once the decision to cease swimming occurs, the majority of fish will turn and fall back into the trawl

(Zone 3), but some will in fact escape over the headline or under the footgear. Those that escape under the footgear, such as flatfish, skates, and cod (Wardle 1984), may do so through active escapeseeking or simply through accidental collision with the footgear. Underwater observations and experiments using mini-sampling nets behind the footgear have together demonstrated that escapement under the trawl is often species specific and size dependent (Albert et al. 2003; DeAlteris et al. 1992; Engås and Godø 1989b; Godø and Walsh 1992; Ingólfsson and Jørgensen 2006; Korotkov 1970; Langeland 2005; Walsh 1992; Weinberg and Munro 1999). 4.5 FISH BEHAVIOR INSIDE THE TRAWL NET AND THE CODEND (ZONE 3) 4.5.1 Entry and Orientation The entry of fish into the trawl net is highly variable within and between species and dependent on a host of extrinsic and intrinsic factors (see Section 4.6). Attempts to characterize entry behavior have ranged from the highly descriptive (e.g., Main and Sangster 1981) to the highly quantitative (e.g., Albert et al. 2003; Bublitz 1996; Castro et al. 1992).

Fish Behavior near Bottom Trawls For most species of fish, entry into the trawl net marks a transition. They were, just moments ago, swimming in the net mouth under the influence of the optomotor reflex. Suddenly a behavioral change has occurred, triggered by either fatigue, the onset of metabolic exhaustion, collision with another fish, social facilitation, or some combination of these factors. Some species, such as sandeels (Ammodytidae) and haddock (Main and Sangster 1981; Wardle 1993), rise upward as they enter the trawl net and swim directly and purposely toward the codend; depending on ambient light levels and net length, these fish see a clear passage surrounded by netting disappearing into the gloom (Wardle 1993; Watson 1988). Cod, flatfish, and saithe, on the other hand, typically turn toward the trawl net and may even attempt escape between the discs of the footgear (e.g., Engås 1994; Wardle 1993) or dive through large meshes if available (e.g., Beutel et al. 2008; Milliken and DeAlteris 2004). Some species make burst-swimming maneuvers in seemingly random directions, striking netting and other fish as they pass through the trawl mouth, while others with limited swimming capability, such as New Zealand dory (Cyttus novaezelandiae) (Piasente et al. 2004), remain motionless and do not respond unless they collide with netting or other fish. The height at which fish rise over the footgear and enter the trawl net is species dependent and can range from a few centimeters to several meters. Benthic species such as skates and flatfish typically enter very low (e.g., Bublitz 1996; Rose 1995), whereas gadoids, small pelagics, and squid tend to rise higher and enter at greater heights (e.g., Glass et al. 1999; Main and Sangster 1981, Thomsen 1993). Several examples exist where gear technologists have successfully used these species-specific behavioral differences to reduce bycatch of nontargeted species (see reviews by Glass 2000; Graham 2006). On reaching the relatively confined funnel section of netting immediately ahead of the codend (sometimes called the extension or intermediate section), some species, including haddock and cod, will turn and swim in the towing direction (Grimaldo et al. 2007; He et al. 2008). As exhaustion sets in, each individual will slow and eventually enter the exten-

83

sion, where increased crowding may disrupt their orderly behavior and elicit randomly orientated burst-swimming behavior. This behavior is likely to cause collision with netting or other fish, or even the escape of small individuals through the side and upper meshes of the trawl net. It is in this location where square-mesh panels or exit windows may be inserted to increase rates of fish escape (e.g., Broadhurst et al. 2002; Graham et al. 2003; Krag et al. 2008). The vertical location of fish during passage through the extension of a trawl net is often highly variable and sometimes species specific. Main and Sangster (1981) observed whiting (Merlangius merlangus) swimming directly toward the codend close to the bottom of the extension and Atlantic mackerel (Scomber scombrus) doing likewise before turning and swimming out of the trawl mouth. Blue grenadier were observed by Piasente et al. (2004) passing through the extension at all heights, burst-swimming in random directions and often colliding with netting or other fish (these collisions usually elicited further bouts of highspeed swimming and additional collisions). Atlantic bumper (Chlorscombrus chrysurus) and anchovies (Ancoa hepstus) passed toward the codend through the upper half of the extension, while red snapper (Lutjanus campechanus) passed through the lower half, with smallest individuals closest to the bottom of the trawl net (D. Foster, National Marine Fisheries Service, Mississippi Labs, personal communication). The behavior of flatfish in the extension can also be highly variable and may range from dynamic to passive. Sand flathead may swim in any direction, often at high speed, or lie motionless on the lower panels of netting (Yanase et al. 2009), particularly on inclined panels where hydrodynamic forces hold them in position (S. Eayrs, personal observations). These fish may remain in place for several minutes or longer, until altering hydrodynamics, collision with passing fish, or another stimulus causes them to rise upward and swim toward the codend. In contrast, He et al. (2008) observed that over 80% of flounders (mainly yellowtail flounder [Limanda ferruginea]) swam slowly close to the bottom of the extension and that most of these individuals were facing the codend. Tiger flathead (Neoplatycephalus

84

Fish Behavior near Fishing Gears during Capture Processes

richardsoni) pass through the net cruise-swimming and enter the codend either facing toward the codend or trawl mouth (Piasente et al. 2004), while rock sole (Pleuronectes bilineatus) show little ability to maneuver or slow their progress toward the codend (Rose 1995). 4.5.2 Fish Behavior inside a Codend Most fish that enter a codend are considered to be exhausted and have limited capability for sustained swimming, having endured to a greater or lesser degree forced swimming to minimize or avoid contact with the trawl gear and other fish. The passage of fish from the net mouth to the codend may take several seconds or longer, depending on fish condition (stamina and available reserves of energy), swimming direction, towing speed, and length of trawl net. Using a multicamera system, Yanase et al. (2009) noted that motionless sand flathead took 7 s on average to pass from the mouth of a fish trawl to the codend, while individuals swimming vigorously in the towing direction took 17 s on average to reach the codend. In a shrimp trawl, juvenile red snapper took 15 to 20 s to swim from the net mouth to a turtle excluder device, a distance of 9 m (Engås et al. 1999). Stronger swimming fish can be expected to take a similar duration to these reported times, or even longer, to enter a codend under similar circumstances. Upon reaching the codend, individuals may become part of the accumulated catch, or turn and swim ahead of the catch for a period of time (Watson 1988; Wardle 1992; O’Neill et al. 2003). Small fish may be passively filtered through the netting, whereas those still capable of active swimming may be able to detect, orientate, and swim through open meshes of the codend and escape. Species with poor swimming capability, such as whiptails and New Zealand dory, appear motionless as they travel from the net mouth to the codend, although contact with other fish or netting often results in a burst-swimming response (Eayrs and Piasente 2006). Strong swimming species, such as spotted warehou (Seriolella punctata), can maintain cruiseswimming behavior within the trawl net for long periods before gradually falling back into the codend, although some of these individuals are also able to swim the entire length of the trawl net during

haulback and escape through the trawl mouth (Piasente et al. 2004). The construction of the codend and the amount of accumulated catch are dominant factors affecting the swimming duration of fish and their likelihood of escape from this part of the trawl net. As the codend is towed through the water, netting gathered at the rear of the codend generates water turbulence and eddies, some of which is pushed forward while some is displaced laterally through open meshes in the codend. This turbulence, together with vesselinduced pulsing movement of the codend, may assist fish swimming by reducing the speed required to maintain station within the codend (Broadhurst et al. 1999; Rose 1995) and serve to provide fish an additional opportunity to escape through the codend meshes (Jones et al. 2008; O’Neill et al. 2003). Main and Sangster (1991) compared the swimming duration of fish in codends constructed from diamond-mesh or square-mesh netting and found a three-fold increase in swimming duration in diamond-mesh codends. The open-square meshes resulted in a 25% increase in relative water flow through the codend and, despite having higher filtering potential—due to consistency of mesh opening—fish with poor swimming capability may struggle to swim against this higher flow and may need to make repeated escape attempts before successfully orientating themselves and swimming through a square mesh. Thus, while the use of square-mesh codends seems intuitively a more appropriate option to allow certain species to escape, higher water flow may to some extent be counterproductive and hamper or delay fish escape from these codends. Further influencing water movement in the codend is catch-induced turbulence. As the catch accumulates in the codend, fish block the open meshes and further dam water movement through the codend. Additional water is now displaced forward and will pass laterally through the codend meshes ahead of the catch. Using a flume tank, Wakeford (2004) characterized the flow field within a codend fitted with a bycatch reduction device (Fig. 4.14). Turbulent flow was particularly apparent ahead of a simulated catch, including a reversal in flow immediately ahead of the catch along the upper section of the codend. There was also a sig-

Fish Behavior near Bottom Trawls

85

Figure 4.14. Typical flow field observed throughout a standard extension/codend containing a bycatch reduction device. (Data from Wakeford 2004.) Variation in water speed through the trawl is known to modify fish behavior. Under certain conditions, turbulent flow and eddies are expected, creating areas of flow reversal as shown here near the top of the codend. For color detail, please see color plate section.

nificant reduction in flow velocity extending approximately 3 m ahead of the catch. Placing a bycatch reduction device (BRD) in this region would have a high likelihood of success because displaced water aids fish escapement through the device (Broadhurst et al. 1999; Eayrs 2007). However, if the BRD is located too far ahead of the catch, fish will be unable to swim forward and reach the device and escape. Bycatch reduction will therefore be comprised and consist mainly of individuals with the capability to respond and escape through the device as they initially pass through the trawl extension. Further hampering determination of the optimum location of a BRD is inconsistent catch volume within and between individual tows; the efficacy of these devices is therefore variable and their precise location in a codend is difficult to determine. In shrimp-trawl fisheries, where a BRD that is located too close to the accumulated catch will result in substantial shrimp loss, fishermen usually take a conservative approach and locate the device as far from the codend drawstrings as is legally permitted (Eayrs 2007). An option to overcome this problem is to stimulate fish to seek escape before they reach the codend. Efforts to achieve this have usually

been based on partial blockage of the codend and include tapered funnels (Watson 1988), hummer bars (Brewer et al. 1998; Watson 1992), deflector grids (Watson 1988), netting cones (Engås et al. 1999) or floats, or the use of contrasting materials such as colored selvedges, dyed netting, or black cylinders/tunnels (e.g., Glass and Wardle 1995b; He et al. 2008; Madsen et al. 1998; Wardle 1992). To date, there has only been limited uptake of these devices by fishermen, suggesting varying and/or limited operational performance. 4.5.3 Fish Behavior in Response to Bycatch Reduction Devices Numerous BRDs1 have been developed over the past few decades for different trawl fisheries around the world (see reviews by Eayrs 2007; Glass 2000; Watson et al. 1999). See Figure 4.15 for some common examples. In some instances, the behavior

1

In the broadest sense a bycatch reduction device is any addition or modification to a trawl net designed to allow the escape of unwanted animals. In many fisheries, including shrimp-trawl fisheries, this term specifically refers to devices designed to allow the escape of fin-fish and other small animals such as crabs and jellyfish.

86

Fish Behavior near Fishing Gears during Capture Processes

Figure 4.15. Examples of bycatch reduction devices (BRDs), including (A) the downward excluding Super Shooter TED, (B) Radial Escape Section (RES) for escape of strong swimming fish from the codend, (C) square-mesh codend for escape of small fish from the codend, and (D) square-mesh window (or escape panel) for the escape of small fish from the codend. (Images from Eayrs 2007.)

of fish in response to these devices has been thoroughly evaluated and is known to be in be influenced by a variety of intrinsic factors such as physiological condition, motivational state, fish size, and visual ability, as well as extrinsic factors such as ambient light levels, water temperature, and BRD design, position, and orientation. This, however, is the exception rather than the rule, and a thorough evaluation of the behavior of most species in relation to BRDs awaits further research. Inclined separator grids are primarily designed to mechanically filter and separate the catch by size (e.g., Fig. 4.15a), although the behavior of some species in response to a grid can also aid their escape from the trawl. Several factors are known to affect the hydrodynamic performance of these devices (Riedel and DeAlteris 1995), their catching performance, and associated fish behavior (see a review by Eayrs 2007). As fish pass through the extension of a trawl net toward a grid, they are faced with the option of either turning and maintaining station within the extension, swimming through the escape opening located ahead of the grid, or attempting passage through the bars of the grid. Variation in behavior is often high and many nontarget species

seem particularly tenacious at seeking areas of reduced flow and resisting escape from the trawl net. Engås et al. (1999) reported that red snapper turned to swim with the trawl net and avoid passage through a netting funnel located immediately ahead of a turtle excluder device (TED) and that this response was presumably due to visual stimuli presented by the device. As these fish tired, they passed through the funnel tail first and then encountered the TED. Most of these fish passed through the TED and continued toward the codend; however, many took up station either behind the TED or swam forward (back through the TED) and took up station adjacent the funnel where it was attached to the extension of the trawl net. These fish swam with low tail-beat frequencies, suggesting reduced water flow in this region of the trawl net, and apparently were not inclined to leave for the remainder of the tow. Given the high variation in behavior reported by these authors, it would seem that new or different strategies are required to facilitate the escape of this species from the trawl net. Currently, the relationship between fish escape and the orientation of a grid or TED is not well understood. Anecdotal evidence indicates that

Fish Behavior near Bottom Trawls upward excluding grids improve fish exclusion rates because down welling light is reflected from the bars of the grid and this increases the distance at which the grid first becomes visible. This reflected light may also improve the visibility of the escape opening and further contribute to fish escape. In contrast, the bars of a downward excluding grid will be in shadow to an approaching fish and the escape opening may be less visible. This explanation is certainly plausible; however, the superior performance of upward excluding grids may simply reflect the preferred direction of fish escape, and these conflicting explanations highlight a need to better understand the interplay between fish response to visual stimuli and their natural escape direction. A dramatic example of the importance of grid orientation was recently demonstrated in a southeastern U.S. shrimp fishery. Aware that several nontargeted finfish species such as red snapper resisted vertical displacement by an upward excluding grid, the researchers tested an innovative sideopening grid and observed an estimated 80% increase in the escapement of nontargeted fish. In response to the horizontal bars of the grid, most red snapper turned to face the trawl mouth orientated perpendicular to the grid, then passed along the face of the grid before swimming through the escape opening (D. Foster, National Marine Fisheries Service, Mississippi Labs, personal communication). The orientation and size of a fish may also play a role in their response as they approach an inclined separator grid. Hannah et al. (2003) reported that fish facing an inclined grid often swam up the grid and through the escape opening in a very short period, whereas fish facing away from the grid often took longer to escape. Large fish in particular often responded by maintaining their position a short distance ahead of the grid where turbulent water induced by the funnel or guiding panel reduced the effort required to swim. As these fish became exhausted, many drifted back, then moved upward to avoid the grid and swam through the escape opening (Hannah et al. 2003). Some large fish have also been observed passing through a grid and entering the codend, only to reemerge later and escape through the trawl mouth (S. Eayrs, personal observations). In contrast, small fish, flatfish, and

87

others with relatively poor swimming capability are sometimes held hydrodynamically against the bars of a grid. These fish are usually dislodged after a short period by sudden grid movement or by contact with passing fish (Eayrs 2007). Small fish have also been observed attempting to maintain station immediately behind the bars of a grid (Engås et al. 1999; Eayrs 2007) or even deliberately lodging themselves behind the bars where they are attached to the trawl netting (S. Eayrs, personal observations). This behavior is considered adaptive in that it reduces energy expenditure; in this region, the swimming speed required to maintain station with the grid may be reduced by 50% or more (Fig. 4.14). This behavior may continue for several minutes until the individual is overcome with exhaustion or contacted by other passing fish. As flatfish pass through a grid, they may attempt to rest on netting distorted by the circumference of the grid. In this position, these fish are orientated at an angle to the apparent water movement and thus held in place against the netting; these fish may take several minutes to reach the codend (Eayrs 2007). The escape of fish through a BRD relies on their ability to detect, orientate, and swim through the escape openings of the device. Özbilgin and Wardle (2002) reported that haddock typically used one of two methods to escape through a square-mesh panel (e.g., Fig. 4.15d)—either slowly working their way through a mesh with a single tail-beat followed by motionless behavior or with rigorous acceleration through a mesh followed by continuous high-speed swimming away from the trawl net. Grimaldo et al. (2007) observed that loose or slack rigging of such panels caused them to “flutter,” and this helped provoke undersized haddock to actively escape and improve escape success. This suggests that both the visual capability of fish and ambient light level are important parameters influencing fish escape. In the absence of vision, fish may not be able to respond to a BRD and only passive filtration is possible, and then only if correctly orientated (Gabr et al. 2007). For fish swimming ahead of accumulated catch in the codend (having missed the opportunity to escape as they initially passed through the extension and codend), reaching a BRD requires swimming at a speed greater than the apparent speed of water in

88

Fish Behavior near Fishing Gears during Capture Processes

the codend. While this will be less than the towing speed, many individuals by this point may simply not have the metabolic reserves to power their escape. Parsons and Foster (2007) argue that the ideal water flow in and around a BRD should be little more than 0.4 m/s to permit successful escape of certain bycatch species. In many tropical shrimp trawl fisheries, bycatch is dominated by small fish that are similar in size or larger than the shrimp. These fish generally have superior swimming capabilities compared with shrimp, and BRDs designed to allow fish to escape by swimming through one or more escape openings are required to reduce catches of these fish. Many of these BRDs are designed to be placed in close proximity to a region of turbulent water flow or generate turbulent water flow themselves (Eayrs 2007). Fish are able to detect turbulent water through their lateral line system and will attempt to remain within an area of turbulent flow, thus increasing their likelihood of encountering a BRD and escaping from the trawl net. Several BRDs have proved particularly successful and have been mandated in certain fisheries, including the fisheye, extended funnel, and Jones/Davis device (see reviews by Watson 2007; Watson et al. 1999). Admittedly, however, none of the BRDs in use today are completely successful in eliminating fish bycatch. In many cases, fish can be guided toward the escape opening of a BRD but make no attempt to escape through the device (Eayrs 2007; Parsons and Foster 2007). The precise reason for this is unknown, but it is thought to be linked to the optomotor reflex and a desire to maintain station within the trawl net. Escapement of these more resistive individuals often occurs only during the haulback procedure, when reduced speed, pulsing codend movements, and slack netting together overcome this reflex and enable fish to swim more readily through the escape openings of the device (Broadhurst et al. 1996; Eayrs 2007; Engås et al. 1999; Madsen et al. 2008; Watson et al. 1999). The key to improved BRD designs therefore seems in the first instance to be linked to deliberate generation of water turbulence to entice fish toward a BRD and, second, by developing methods to overcome the optomotor reflex so fish voluntarily depart from the trawl net throughout the entire tow.

However, efforts to manipulate water movement and improve BRD design have usually been based on a trial-and-error approach to testing and development, and this has meant that the design of many BRDs has remained largely unchanged for several decades. A relatively recent innovation, the fishbox BRD, uses one or more metal plates (or foils) to induce water turbulence adjacent to escape openings. Despite having been tested successfully in shrimp fisheries in the United States and Australia (Eayrs 2007) and outperforming most other BRDs, particularly during daytime, an intimate understanding of fish behavior to variations in foil design and orientation awaits to be fully realized. Overcoming the optomotor reflex to facilitate fish escapement has to date not been effectively achieved and also requires further research. Reducing the visual stimulus (contrast) of the trawl extension, codend, and BRD is an obvious first step but requires a greater understanding of the visual capabilities of fish under a range of ambient light levels encountered in the fishery and their response under these various conditions. Meanwhile, until this is achieved, the influence of the optomotor reflex will dominate fish behavior and BRD performance will not be optimized. The immediate future of BRD research clearly rests in greater knowledge of fish behavior and their response to trawl stimuli, particularly in low-light conditions. The need for divers or cameras to observe fish behavior in close proximity to BRDs remains a significant limitation. The use of flume tanks to assess water flow, test ways to generate turbulence, and develop new concepts is an important component of BRD research (Brewer et al. 1998). These facilities provide a controlled testing environment and have the necessary instrumentation to quantify water velocity and direction in and around these devices and codends (Winger et al. 2006). This knowledge can be used to develop new or improved designs that can be applied to at-sea studies. Watson (1988) argues that deliberately altering water flow is the most promising technique for developing selective shrimp trawls, in combination with knowledge of fish response to trawl stimuli. Wakeford (2004) successfully used a flume tank to assess the effect of a fisheye BRD on water flow in and around a full-sized shrimp-trawl codend,

Fish Behavior near Bottom Trawls and despite only being able to record water flow in two dimensions, the author reported that the device reduced flow in excess of 80% near the upper panel of the codend. Just how this knowledge might be used to improve the performance of the fisheye remains unclear, but it does provide a first step in stimulating new ideas and ways to elicit fish movement toward such devices and escapement from a trawl net. 4.6 FACTORS INFLUENCING FISH BEHAVIOR NEAR TRAWLS 4.6.1 Extrinsic Factors Variation in bottom trawl catchability is known to be associated with spatial and temporal changes in environmental conditions. Ambient light intensity, water temperature, and fish density are generally considered to be three of the most influential extrinsic factors governing fish behavior and fish capture; each is described next. Light Level, Contrast, and Color It is well known that most species of fish have welldeveloped and efficient visual systems that are particularly well adapted to detect very small differences in contrast in the generally monochromatic underwater environment in which they exist (Jones et al. 2004). Many species, particularly fast-swimming pelagic ones, have been demonstrated to have excellent visual acuity and low light sensitivity. In addition, many commercially important fish species can form visual images at light intensities well below those at which human visual systems cease to perform. In some species these minimum light intensity thresholds occur at extremely low levels (less than 10−6 lux) (Glass and Wardle 1989; Walsh and Hickey 1993) and can be difficult to detect and measure reliably. Consequently, environmental conditions, such as underwater light field, can have a significant effect on fish availability (Jones et al. 2004), catchability (Wileman et al. 1996), and reaction behavior (Ryer and Barnett 2006) within and around the net. Observations on the reaction behavior of fish to towed fishing gears have been conducted for more than 60 years and have shown that when light is present, fish are not sieved from the sea but react in quite subtle ways to the complex

89

visual stimuli presented by components of the gear. However, despite the advanced performance of fish visual systems, there will clearly be times and/or depths at which their visual systems cease to perform and this will manifest itself in a different suite of behavioral responses. To date, this aspect has received little attention, perhaps due to the difficulty of “observing” behavior under these dark conditions. The visibility of an object such as a netting panel, trawl float, door, or other component of a net depends largely on its contrast with the background and is therefore dependent on the background against which it is viewed, the properties of the water, as well the direction and intensity of the illumination (Kim and Wardle 1998a, 1998b). All of this combines to produce a unique visual background against which a fish must view its world. The first and perhaps most obvious characteristic about each unique visual background is its visible range. Where water clarity is good and light intensity is high, fish may see an approaching net from afar and may react to its approach by rising in the water column and allowing the net to pass underneath, thus avoiding capture. Conversely, in dark or turbid conditions, visible range may be very short and fish may have little time to react to the approaching net. Under these circumstances, response thresholds are generally expected to be higher, reaction distances shorter, and collisions with the gear more likely (Glass and Wardle 1989; Walsh and Hickey 1993). However, under most circumstances where the fish are able to form visual images, they will react in a predictable manner to the visual stimuli presented by the gear and its components. The headrope and its floats will be seen as a high-contrast silhouette, as will the sweeps and footgear as viewed against the sediment underneath. The netting surrounding the fish appears as a moving pattern, and this induces an optomotor response, causing the fish to turn and swim in the direction of the tow at the same speed as the moving netting (see Section 4.4). Aside from the visual background against which an object is viewed, the nature of the object itself is also important, particularly its color. Netting panels used in bottom trawls may be dyed any number of colors (green, blue, red, orange, yellow), and harvesters rarely agree on the combination of

90

Fish Behavior near Fishing Gears during Capture Processes

panel colors necessary to achieve optimal catch rates and bycatch reduction. Certain panels may be colored to increase their visibility and assist with herding, whereas other panels may be altered to reduce their conspicuous to assist with escapement of nontargeted species or sizes (for review, see Glass 2000; Jones et al. 2004; Wardle 1993). Visual illusions may also be used to encourage fish to actively select certain behavior patterns. Laboratory experiments have demonstrated that fish will choose to penetrate meshes when alternative routes are blocked or appear blocked due to visual illusion (Glass et al. 1995). In this case, the visual stimulus was a black tunnel that was believed to appear like the looming open mouth of a large predator. Subsequent sea trials were carried out using black PVC-coated canvas laced into the extension of a trawl to create a black tunnel effect. Video observations of fish showed a strong aversion to entering the tunnel and vigorous attempts to penetrate square-mesh escapement panels located immediately ahead of the perceived “threat” (Glass and Wardle 1995b). Trials in the Gulf of Maine have also demonstrated promising results of black tunnels at separating target from nontarget species (He et al. 2008), although other attempts to duplicate the findings have been less successful (Brewer et al. 1998). Water Temperature With the exception of certain scombrids and sharks, water temperature is known to be an important environmental constraint on the behavior of most species of fish. It affects almost every aspect of an individual’s activities, including metabolism, muscle performance, growth rate, activity level, foraging routine, vigilance to predators, and even behavior toward fishing gear. It is generally thought that temperature influences the catchability of a bottom trawl in two ways: (1) by affecting response threshold and (2) by affecting swimming capability. In Section 4.3, it was noted that an individual’s response threshold toward an approaching threat is the outcome of a behavioral tradeoff that minimizes costs and maximizes benefits. If we assume the costs of fleeing (F) are related in some way to water temperature (probably through energy expendi-

ture), we would expect individuals to adjust their optimal reaction distance (D*) so as to minimize these costs (see Fig. 4.4). In other words, we expect that at unfavorable water temperatures (either too warm or cold), the F-cost curve should rise more steeply, shifting the intersection of the curves to the left, resulting in a reduced reaction distance (D*) to the threat. Field data collected by Winger (2004) support this argument. Atlantic cod exhibited a reduced likelihood of responding to an approaching trawler in winter compared with summer, as well as a corresponding reduction in reaction distance at the lower water temperatures. Under environmentally extreme conditions, we would expect an even higher response threshold, near-zero reaction distance, and an inability for the trawl to elicit a behavioral response whatsoever. It is possible that the passive drifting of walleye pollock into a trawl net under low water temperatures, recorded by Inoue et al. (1993), may be representative of this kind of situation. The second way that temperature affects trawl capture is through its influence on swimming capability. At the most basic level, it alters the rate at which muscle fiber is physically capable of contracting. This limits muscle power and in turn the swimming speeds at which fish are capable of swimming, essentially shrinking the performance range. Although empirical field observations are still largely lacking, laboratory evidence suggests water temperature has a profound effect on the fish capture process, including limiting the herding efficiency of trawl sweeps (Ryer and Barnett 2006; Winger et al. 1999), limiting swimming speed and endurance in the trawl mouth (He 1991, 1993; Yanase et al. 2007), and limiting the ability to escape once inside the net (Özbilgin 2002; Özbilgin and Wardle 2002). For a detailed review, see Chapter 1. Fish Density Variation in distribution, patch size, and fish density are common. Some species maintain aggregations almost continuously, while others do so infrequently or only on occasion. At low densities, the behavior of individual fish is largely independent of neighbors, while at higher densities social interactions with conspecifics are likely to occur. This

Fish Behavior near Bottom Trawls variation in the local density of fish is known to influence aspects of fish capture behavior from the net mouth to the codend. Godø et al. (1999) provides evidence for densitydependent behavior in the net mouth based on a comparative analysis of Norwegian and Canadian underwater video observations. The authors report qualitative differences in behavior at varying densities for cod, haddock, and American plaice. At low densities (one or two fish), cod and haddock displayed “loner behavior,” characterized as kick-andglide swimming while criss-crossing the net mouth. Turnover rates (number of fish entering the net per unit time) were high and escapement under the footgear was common. As fish density increased (more than five fish), cod and haddock typically exhibited “schooling behavior.” This was characterized as a more-uniform, less-frightened behavioral response, resulting in a lower turnover rate and reduced escapement under the footgear. In the case of plaice, low densities were characterized by zigzag behavior and increased escapement under the footgear. At higher densities, the zigzag behavior was often disrupted and collisions with other fish were common, resulting in reduced escapement under the footgear. Evidence that variation in fish density in the net mouth actually corresponds to differences in trawl efficiency has been examined by Walsh (1996) and Godø et al. (1999). Both studies examined the catch of fish in the trawl net compared with fish caught in smaller bag nets that were attached underneath the main trawl, designed to catch downward escaping fish that pass through the footgear. Albeit highly variable, both studies showed a general tendency toward lower efficiency of the footgear (i.e., more fish escaping underneath) at lower densities and greater efficiency (i.e., less fish escaping underneath) at higher densities. Fish density is also expected to affect the performance of BRDs installed within the net. When large pulses of fish are encountered, devices such as selection windows, sorting grids, or separator panels may be temporarily masked by neighboring conspecifics. This reduces the probability of fish encountering the devices and thus reduces the potential sorting efficiency. For example, a BRD that is designed to reduce the capture of small fish

91

may sustain reduced performance at high catch rates (i.e., high densities), resulting in a larger proportion of small fish in the catch, lowering the overall L50 for the haul. Evidence of a negative correlation between L50 and catch rate has been documented by Kvamme and Isaksen (2004) and Jørgensen et al. (2006). Finally, recent evidence suggests that the degree of accumulated catch and subsequent density of fish in the codend affects the duration of station keeping as well as the likelihood of active escape seeking behavior in some species. Jones et al. (2008) collected underwater video observations of haddock in the codend of a commercial whitefish trawl. The authors found that at low densities, haddock exhibited a reduced preference for station keeping, resulting in an increased rate of contact with the netting and subsequent escape. As the catch accumulated in the codend and the density increased, station keeping behavior increased and the rate of contact with the netting decreased. The findings are consistent with similar observations in the net mouth (Godø et al. 1999), suggesting an increased preference in some species to maintain the status quo, or togetherness, when surrounded with the company of other fish. 4.6.2 Intrinsic Factors In addition to the many extrinsic factors that modify fish behavior during capture, there are also a number of intrinsic factors—that is, conditions or states wholly belonging to the individual—that are also known to modify behavioral expression and therefore trawl catchability. The best understood of these is fish size, but there are other lesser studied factors, including an individual’s motivational state, physiological condition, learning, and experience. Fish Size One of the most obvious and well-studied factors affecting fish capture by trawls is fish size. The effect is manifested primarily as length-dependent swimming capability. Both the maximum swimming speed that an animal can achieve and its endurance are closely related to its body length (see Chapter 1). This is particularly important during the herding of benthic species located in the sweep zone as well as for all fish swimming in the trawl

92

Fish Behavior near Fishing Gears during Capture Processes

mouth (see Section 4.4). In both cases, every fish, regardless of size, is stimulated to swim at a certain speed. As a result, fish of different sizes must operate at different levels within their performance range. Small fish are required to swim vigorously with high tail-beat frequencies and will operate at the upper end of their performance range, analogous to sprinting. Larger fish, by comparison, are capable of higher swimming speeds and therefore operate at a lower point in their performance range, analogous to jogging. To effectively swim at these different levels, teleost fish exhibit more than one gait (Alexander 1989; Peake and Farrell 2004; Winger et al. 2004). Gaits represent a multigear system (synonymous with walking, jogging, sprinting) for the fractionation of the swimming performance range (Webb 1994a). Similar to terrestrial animals, individual gaits in fish work over a part of the performance range and a series of gaits together cover the entire performance range. Each gait is defined on the basis of its muscle use, propulsor type, and propulsor kinematics (Webb 1994b). Looking at the trawl mouth as an example, fish of different sizes are observed using different gaits in an effort to produce sufficient thrust to power their swimming and avoid falling back into the trawl. Small fish are characterized by high tail-beat frequencies coupled with burst-and-coast motions. These sprintlike bursts rely heavily on the fast contraction rates of white anaerobic muscle fibers to produce sufficient mechanical power. Medium-sized fish, by comparison, are usually characterized by continuous swimming punctuated with occasional large-amplitude kicks of the tail for extra thrust. As fish size continues to increase, a transition from unsteady to steady swimming occurs, with the largest fish swimming effortlessly in the mouth of the trawl. This final gait transition coincides with a change from prolonged to sustained swimming activity. Fish at these sizes have sufficient cross-sectional muscle area to produce the power for steady cruising and do not recruit white muscle fiber to the extent that smaller fish must (see reviews by Videler 1993; Videler and Wardle 1991; Webb 1994ab). Motivational State Ecological theory tells us that animals are routinely (if not constantly) balancing conflicting demands.

Assuming fish cannot maximize all things all the time, the situation inescapably arises where two or more demands cannot be satisfied simultaneously, and a tradeoff must be made. Usually, this includes things such as balancing the need to eat against the need to avoid being eaten (e.g., Nøttestad et al. 1996). Albeit relatively modern from an evolutionary perspective, mobile trawls also represent a tangible threat, and fish are expected to make behavioral tradeoffs at the individual level that minimize costs and maximize benefits. One of the best ways to observe this tradeoff is to witness changes in response threshold (i.e., reaction distance) under different scenarios (see Section 4.3). One particularly interesting tradeoff scenario is the harvesting of spawning aggregations. Under these conditions, the threat of capture is pitted against maximizing reproductive fitness/output. In this situation, we would predict the cost of fleeing (F) to be high (that is a steeper line than usual) because of lost reproductive opportunities. Hence, vigilance toward approaching threats should be low, raising the response threshold, and lowering the optimal reaction distance (see Fig. 4.4). Empirical evidence to support this argument is limited but not entirely lacking. Olsen (1971) described capelin (Mallotus villosus Müll.) as insensitive to vessel noise during the spawning season. Mohr (1971) compared the avoidance behavior of prespawning aggregations of Atlantic herring to midwater trawls against those of spawning aggregations. He described prespawning herring to be “skittish” but, as the spawning time approached, herring became “sluggish and indolent.” At peak spawning season, almost no reaction to the vessel and trawl could be observed whatsoever. More recently, Skaret et al. (2005) also reported no avoidance of spawning aggregations of herring toward the noise of a large research vessel in shallow water (30–40 m). These observations are atypical during the rest of the year when herring are typically riskaverse and react strongly to such threats (e.g., Misund 1994; Vabø et al. 2002), supporting the argument that motivational state can modify behavioral expression. In general, motivational state is poorly understood and our ability to accurately observe and record tradeoffs under different motivational condi-

Fish Behavior near Bottom Trawls tions is limited. In many cases, such things cannot be adequately teased apart under field conditions due to confounding differences with other extrinsic factors (e.g., water temperature, light intensity, depth, predator–prey field, etc.), which may produce unexpected results at times (e.g., Skaret et al. 2006). Further research is required to illuminate this field. Laboratory experiments, albeit limited in their application, may be the only way to investigate variation in behavioral expression under different motivational states. Physiological Condition The condition of fish is known to vary seasonally in association with things such as feeding, migrating, and reproductive cycles. It can also vary with environmental conditions and food availability. For comparative purposes, it is often assessed on the basis of morphometrics or somatic indices or with the use of biochemical indicators (e.g., Dutil et al. 1998; Martínez et al. 1999). Although empirical evidence is limited, changes in fish condition are suspected to affect trawl catchability in at least three ways. First, physiological condition may affect trawl capture through its influence on response threshold and motivational state (see earlier discussions). Clearly, individuals with poor nutritional status will make different tradeoffs as they balance the conflicting demands of fleeing in response to an approaching trawl or continuing with some existing activity. Even more complicating would be if they are actively engaged in something important such as a feeding opportunity, migration, or spawning. Under this scenario, there would be serious lost opportunities associated with fleeing. This should push the F-cost curve upward, shifting the intersection of the curves to the left, resulting in a decreased reaction distance (D*) to the threat (Fig. 4.4). This is, of course, speculative; empirical evidence to support this hypothesis is completely lacking. Second, physiological condition affects trawl capture through its influence on swimming capability. As fish condition deteriorates, lipid and glycogen reserves are typically mobilized first, but eventually so are white muscle proteins. This changes white muscle composition, reduces muscle mass, limits power output, and shrinks the perfor-

93

mance range. Laboratory studies have demonstrated that starved cod suffer both reduced sprint speeds (Martínez et al. 2002) and reduced swimming endurance (Martínez et al. 2003) compared with their fed counterparts. Although field data are lacking, we would expect this reduced performance to limit the herding efficiency of trawl sweeps, limit swimming speed and endurance in the trawl mouth, and limit escape ability once inside the trawl. Finally, physiological condition is expected to affect trawl capture through its influence on mesh penetration capability. Özbilgin et al. (2006) and Ferro et al. (2008) investigated temporal variation in codend size selection for North Sea haddock over a period of 1 year. The authors found seasonal variation in retention characteristics of the trawl based on differences in fish girth. Interestingly, for fish of a given length, an increase in girth lead to increased retention in the winter, whereas it led to a decrease in the fall. They explain this difference on the basis that the increased girth in winter was associated with gonad development, whereas in the fall it was associated with muscle volume. They speculate that the latter probably had better swimming performance (for a given length and girth) resulting in increased capacity to penetrate and escape open meshes in the codend. Learning and Experience Models of fish reaction behavior often assume that all fish react in a similar manner and that all fish are naïve when they encounter a bottom trawl. These are both erroneous assumptions. In Section 4.6, we explored how motivational state can have a significant effect on the reaction behavior of fish but it is also reasonable to assume that, in many cases, fish may encounter fishing gear on more than one occasion and that their behavioral reactions may be modified through a process of learning from past experience. More than 50 years ago, Golenchenko (1955) raised the possibility that fish could have conditioned responses to active fishing gear. Using the Russian made hydrostat, Kiselev (1968) documented some of the first field observations of fish behavior in response to the repeated sounds of a trawler. He reported that cod were initially “agitated” but, after a few repetitions of the sound, soon

94

Fish Behavior near Fishing Gears during Capture Processes

habituated and showed “no response.” Direct observations of cod and haddock in response to repeated trawl hauls were later conducted during the 1980s using the towed submersible “Tetis” (Zaferman and Serebrov 1989). The authors documented reduced catchability of conditioned fish, especially smaller fish that had escaped the trawl during previous encounters. Using acoustic telemetry, Pyanov (1993) demonstrated that experienced fish will actually move out of the path of an approaching trawler and that the learned behavior could be observed as much as 9 days after the initial encounter. How rapid can fish learn? While this is likely to vary between species and individuals (Hunter and Wisby 1964; Zhuykov and Pyanov 1993), evidence suggests that fish have the capacity to learn quickly, particularly in response to aversive stimuli such as fishing gear. This is because learning about fishing gear is a risky business, and there is little room for mistakes or extensive training. Laboratory studies have demonstrated rather convincingly that fish can learn to avoid approaching nets after just one or two experiences (Hunter and Wisby 1964; Pyanov 1993; Pyanov and Zhuykov 1993) and that this can lead to noticeable reductions in catching efficiency during repeated tows (for a review of the Russian literature, see Pavlov and Kasumyan 1995). However, despite multiple demonstrations of the ability of fish to avoid approaching trawlers, fishing gears continue to capture fish even in heavily fished areas where there is a likelihood that individual fish may encounter and pass through the meshes on more than one occasion. Özbilgin and Glass (2004) demonstrated in a laboratory setting that haddock were reluctant to pass through a mesh panel but learned quickly to penetrate the meshes when a food reward was presented as a conditioned stimulus. Once this response was learned, the fish penetrated the meshes with ease when presented with the conditioned stimulus while still reluctant to do so in the absence of the stimulus. They concluded that haddock were capable of modifying their behavior based on prior experience and argued that fish that encounter a trawl gear on more than one occasion may be capable of “escaping” more effectively on subsequent encounters with a net or that, at the very least, their behavior may be modified in some way.

Can learned behavior be transmitted to naïve individuals? In formal terms, this is referred to as “social learning” and is defined as the process by which individuals acquire new behavior or information about their environment via the observation or interaction with other animals (see review by Brown and Laland 2003). Two remarkable studies demonstrated social learning among fish in response to approaching nets under laboratory conditions (Brown and Warburton 1999; Hunter and Wisby 1964). In these experiments, nets were towed along a tank and the ability of fish to locate and escape through a small hole was evaluated. Failure to locate the hole resulted in a negative/punitive experience (trapped) and success resulted in a positive experience (escape). Both studies showed that groups of fish solved the problem faster than isolates or smaller groups and that escape latency decreased with repeated exposure. The findings demonstrate the potential value of social interactions in avoiding active fishing gear. In these examples, fish were more likely to make the “right” decision sooner and with greater accuracy when they were able to monitor the behavior of several fellow conspecifics, benefiting from a discovery made by any one of them. Other supporting evidence for social learning in fish has been documented for herring (Soria et al. 1993) as well as haddock (Özbilgin and Glass 2004). In the latter study, naïve haddock learned to penetrate mesh more quickly when in the company of experienced fish. One clear implication of such behavioral modification through learning is that the efficiency and selectivity of a net may not be the same for all individuals that encounter the net and that “experienced” fish may have a higher probability of escapement, and hence potential survivability, than naïve fish. This may have important implications for surveys and other scientific studies such as those that seek to determine selective efficiency of fishing gears where the assumption is that all fish react in a similar manner to the net. The effect of learned avoidance behavior in fishing gears has also been discussed in detail by Fréon et al. (1993) and Soria et al. (1993), who concluded that there could be a long-term decrease in catchability in exploited stocks where fish are relatively more experienced than for fish of unexploited stocks.

Fish Behavior near Bottom Trawls Although a body of evidence exists in support of the role of learning and experience in behavioral modification of reactions to fishing gears, this remains an area that may yet yield important insights through further research. 4.7 CONCLUDING REMARKS This chapter reviewed the current knowledge of fish behavior in relation to bottom trawls, with discussions starting in the pretrawl zone ahead of the vessel and finishing with behavior in the codend. The entire capture process has been discussed along with key extrinsic and intrinsic factors that are known to influence behavior. We have described typical patterns of behavior but equally emphasized the high degree of variability in behavioral expression that is often observed. We have extended the application of the economic hypothesis of antipredator behavior (Ydenberg and Dill 1986) initially introduced to the field of trawl capture by Fernö and Huse (2003) and found that, in many cases, it is helpful in formulating predictions of fish behavior in response to vessels and bottom trawls. ACKNOWLEDGMENTS The authors are grateful to Emma Jones (NIWA, New Zealand) and Daniel Foster (NMFS, Mississippi, USA) for their helpful comments and suggestions on the manuscript. REFERENCES Albert OT, Harbitz A and Høines ÅS. 2003. Greenland halibut observed by video in front of survey trawl: behavior, escapement, and spatial pattern. J. Sea Res. 50: 117–127. Alexander RM. 1989. Optimization and gaits in the locomotion of vertebrates. Physiol. Rev. 69: 1199–1227. Anon. 1995. Underwater noise of research vessels: review and recommendations. Ed. by RB Mitson. ICES Coop. Res. Rep. 209: 61 pp. Anon. In press. Causes and consequences of fish reactions to fisheries research vessels. Ed. by F. Gerlotto, ICES Coop. Res. Rep. Beamish FWH. 1966a. Reaction of fish to otter trawls. Fish. Can. 19(5): 19–21. Beamish FWH. 1966b. Swimming endurance of some Northwest Atlantic fishes. J. Fish. Res. Bd. Can. 23: 341–347.

95

Beamish FWH. 1969. Photographic observations on reactions of fish ahead of otter trawls. FAO Fish. Rep. 62(3): 511–521 Beutel D, Skrobe L, Castro K, Ruhle P Sr, Ruhle P Jr, O’Grady J and Knight J. 2008. Bycatch reduction in the Northeast USA directed haddock bottom trawl fishery. Fish. Res. 94: 190–198. Breen M, Dyson J, O’Neill FG, Jones E and Haigh M. 2004. Swimming endurance of haddock (Melanogrammus aeglefinus L.) at prolonged and sustained swimming speeds, and its role in their capture by towed fishing gears. ICES J. Mar. Sci. 61: 1071–1079. Brewer D, Rawlinson N, Eayrs S and Burridge C. 1998. An assessment of bycatch reduction devices in a tropical Australian prawn trawl fishery. Fish. Res. 36: 195–215. Broadhurst MK, Kennelly SJ and Eayrs S. 1999. Flowrelated effects in prawn trawl codends: potential for increasing the escape of unwanted fish through square-mesh panels. Fish. Bull. 97: 1–8. Broadhurst MK, Kennelly SJ and Gray CA. 2002. Optimal positioning and design of behavioral-type by-catch reduction devices involving square-mesh panels in penaeid prawn-trawl codends. Mar. Freshw. Res. 53: 813–823. Broadhurst MK, Kennelly SJ and O’Doherty G. 1996. Effects of square-mesh panels in codends and of haulback delay on bycatch reduction in the oceanic prawn-trawl fishery of New South Wales. Aust. Fish. Bull. 94: 412–422. Brown C and Laland KN. 2003. Social learning in fishes: a review. Fish Fish 4: 280–288. Brown C and Warburton K. 1999. Social mechanisms enhance escape responses in shoals of rainbowfish, Melanotaenia duboulayi. Environ. Biol. Fish 56: 455–459. Bublitz CG. 1996. Quantitative evaluation of flatfish behavior during the capture by trawl gear. Fish. Res. 25: 293–304. Buerkle U. 1977. Detection of trawling noise by Atlantic cod (Gadus morhua L.). Mar. Behav. Physiol. 4: 233–242. Castro KM, DeAlteris JT and Milliken HO. 1992. The application of a methodology to quantify fish behavior in the vicinity of demersal trawls in the Northwest Atlantic, USA. Mar. Technol. Soc. Conf. Proc. 1992: 310–315. Chandler RA. 1967. Swimming endurance of haddock. Fish. Res. Bd. Can. Manuscr. Rep. Ser. 930: 14 pp. Chapman CJ. 1973. Field studies of hearing in teleost fish. Helgoländer wiss. Meeresunters 24: 371–390.

96

Fish Behavior near Fishing Gears during Capture Processes

Chapman CJ and Johnstone ADF. 1974. Some auditory discrimination experiments on marine fish. J. Exp. Biol. 61: 521–528. Cooper WE Jr and Frederick WG. 2007. Optimal flight initiation distance. J. Theor. Biol. 244: 59–67. DeAlteris JT, Castro KM and Milliken HO. 1992. Development of an underwater video camera and recording system for observing fish behavior in the vicinity of a bottom trawl and a methodology to quantitatively analyze the resulting data. Final Report, Rhode Island Sea Gant Program, DOC Award No. NA-89AA-D-56082, 20 pp. De Robertis A, Hjellvik V, Williamson NJ and Wilson CD. 2008. Silent ships do not always encounter more fish: comparison of acoustic backscatter recorded by a noise-reduced and a conventional research vessel. ICES J. Mar. Sci. 65: 623–635. Dogˇanyilmaz-Özbilgin Y, Özbilgin H and Bas¸aran F. 2006. Relationship between critical and maximum swimming speeds. ICES WGFTFB Report, pp 145–153. Dutil J-D, Lambert Y, Guderley H, Blier PU, Pelletier D and Desroches M. 1998. Nucleic acids and enzymes in Atlantic cod (Gadus morhua) differing in condition and growth rate trajectories. Can. J. Fish. Aquat. Sci. 55: 788–795. Eayrs S. 2007. A guide to bycatch reduction in tropical shrimp-trawl fisheries. Revised edition. Rome: FAO. 105 pp. Eayrs S and Piasente M. 2006. In situ examinination of fish behavior in response to a demersal trawl net. Presented at ICES Symposium on Fishing Technology in the 21st Century, Oct 30-Nov 3, 2006. Boston, MA. Engås A. 1994. The effects of trawl performance and fish behavior on the catching efficiency of demersal sampling trawls. In: Fernöö A and Olsen S (eds). Marine Fish Behavior in Capture and Abundance Estimation. pp 45–68. Oxford: Fishing News Books Ltd. Engås A and Godø OR. 1989a. The effect of different sweep lengths on the length composition of bottomsampling trawl catches. J. Cons. Int. Explor. Mer. 45: 263–268. Engås A and Godø OR. 1989b. Escape of fish under the fishing line of a Norwegian sampling trawl and its influence on survey results. J. Cons. Int. Explor. Mer. 45: 269–276. Engås A, Foster D, Hataway BD, Watson JW and Workman I. 1999. The behavioral response of juvenile red snapper (Lutjanus campechanus) to

shrimp trawls that utilize water flow modifications to induce escapement. Mar. Technol. Soc. J. 33(2): 43–50. Engås A, Haugland EK and Øvredal JT. 1998. Reactions of cod (Gadus morhua) in the pre-vessel zone to an approaching trawler under different light conditions: preliminary results. Hydrobiologia 371/372: 199–206. Fay RR. 2005. Sound source localization by fishes. In: Popper AN and Fay RR (eds). Sound Source Localization. pp 36–66. New York: Springer. Fernö A. 1993. Advances in understanding of basic behavior: consequences for fish capture studies. ICES Mar. Sci. Symp. 196: 5–11. Fernö A and Huse I. 2003. Fish avoidance of survey vessels and gear: can predictions based on the response of fish to predators explain the observed variations? Presented at the ICES Symposium on Fish Behavior in Exploited Ecosystems, June 23– 26, 2003. Bergen, Norway. Ferro RST, Özbilgin H and Breen M. 2008. The potential for optimizing yield from a haddock trawl fishery using seasonal changes in selectivity, population structure, and fish condition. Fish. Res. 94: 151–159. Foster JJ. 1969. The influence of fish behavior on trawl design with special reference to mathematical interpretations of observations on the swimming speeds of fish and results of C.F. experiments. FAO Fish. Rep. 62(3): 731–773. Foster JJ, Campbell CM and Sabin GCW. 1981. The fish catching process relevant to trawls. Can. Spec. Publ. Fish. Aquat. Sci. 58: 229–246. Fréon P, Gerlotto F and Misund OA. 1993. Consequences of fish behavior for stock assessment. ICES Mar. Sci. Symp. 196: 190–195. Fréon P and Misund OA. 1999. Dynamics of Pelagic Fish Distribution and Behavior: Effects on Fisheries and Stock Assessment. Oxford: Fishing News Books. 348 pp. Fuwa S. 1989. Fish herding model by ground ropes considering reaction of fish. Nippon Suisan Gakkaishi. 55: 1767–1771. Fuwa S, Sato O, Nashimoto K and Higo N. 1988. Fish herding model by ground rope. Nippon Suisan Gakkaishi 54: 1155–1159. Gabr M, Fujimori Y, Shimizu S and Miura T. 2007. Behavior analysis of undersized fish escaping through square meshes and separating grids in simulated trawling experiment. Fish. Res. 85: 112–121. Gerlotto F, Castillo J, Saavedra A, Barbieri MA, Espejo M and Cotel P. 2004. Three-dimensional

Fish Behavior near Bottom Trawls structure and avoidance behavior of anchovy and common sardine schools in central southern Chile. ICES J. Mar. Sci. 61: 1120–1126. Glass CW. 2000. Conservation of fish stocks through bycatch reduction: a review. Northeast Naturalist 7: 395–410. Glass CW and Wardle CS. 1989. Comparison of the reactions of fish to a trawl gear, at high and low light intensities. Fish. Res. 7: 249–266. Glass CW and Wardle CS. 1995a. A review o fish behavior in relation to species separation and bycatch reduction in mixed fisheries. In: Solving Bycatch: Considerations for Today, and Tomorrow. pp 243–250. Alaska Sea Grant College Program Report No. 96–03, University of Alaska Fairbanks. Glass CW and Wardle CS. 1995b. Studies on the use of visual stimuli to control fish escape from codends. II. The effect of a black tunnel on the reaction behavior of fish in otter trawl codends. Fish. Res. 23: 165–174. Glass CW, Sarno B, Milliken HO, Morris GD and Carr HA. 1999. Bycatch reduction in Massachusetts inshore squid (Loligo pealeii) trawl fisheries. Mar. Technol. Soc. J. 33(2): 35–42. Glass CW, Wardle CS, Gosden SJ and Racey DN. 1995. Studies on the use of visual stimuli to control fish escape from codends. I. Laboratory studies on the effect of a black tunnel on mesh penetration. Fish. Res. 23: 157–164. Godin J-G. 1997. Evading predators. In: Godin J-G (ed). Behavioral Ecology of Teleost Fishes. pp 191– 236. Oxford: University Press. Godø OR. 1994. Factors affecting the reliability of groundfish abundance estimates from bottom trawl surveys. In: Fernö A and Olsen S (eds). Marine Fish Behavior in Capture and Abundance Estimation. pp 169–199. Oxford: Fishing News Books. Godø OR. 1998. What can technology offer the future fisheries scientist—possibilities for obtaining better estimates of stock abundance by direct observations. J. Northw. Atl. Fish. Aquat. Sci. 23: 105–131. Godø OR and Walsh SJ. 1992. Escapement of fish during bottom trawl sampling—implications for resource assessment. Fish. Res. 13: 281–292. Godø OR, Walsh SJ and Engås A. 1999. Investigating density-dependent catchability in bottom trawl surveys. ICES J. Mar. Sci. 56: 292–298. Golenchenko AP. 1955. Speech. In: Proceedings of the Conference on Fish Behavior and Searching. pp 53–54. Ichthyological Commission, 5, Academy of Science, USSR, Moscow. 236 pp (in Russian).

97

Graham N. 2006. Trawling: historic development, current status and future challenges. Mar. Technol. Soc. J. 40: 20–24. Graham N, Jones EG and Reid DG. 2004. Review of technological advances for the study of fish behavior in relation to demersal fishing trawls. ICES J. Mar. Sci. 61: 1036–1043. Graham N, Kynoch RJ and Fryer RJ. 2003. Square mesh panels in demersal trawls: further data relating haddock and whiting selectivity to panel position. Fish. Res. 62: 361–375. Grimaldo E, Larsen RB and Holst R. 2007. Exit windows as an alternative selective system for the Barents Sea demersal fishery for cod and haddock. Fish. Res. 85: 295–305. Hall SJ, Wardle CS and MacLennan DN. 1986. Predator evasion in a fish school: test of a model for the fountain effect. Mar. Biol. 91: 143–148. Handegard NO, Michalsen K and Tjøstheim D. 2003. Avoidance behavior in cod (Gadus morhua) to a bottom-trawling vessel. Aquat. Liv. Resour. 16: 265–270. Handegard NO and Tjøstheim D. 2005. When fish meet a trawling vessel: examining the behavior of gadoids using a free-floating buoy and acoustic split-beam tracking. Can. J. Fish. Aquat. Sci. 62: 2409–2422. Hannah WR, Jones SA and Matteson KM. 2003. Observations of fish and shrimp behavior in ocean shrimp (Pandalus jordani) trawls. Newport, OR: OR Dept. Fish. Wildl. Mar. Res. Prog. 28 pp. Harden-Jones FR. 1963. The reaction of fish to moving backgrounds. J. Exp. Biol. 40: 437–446. Harden-Jones FR, Margetts AR, Greer Walker G and Arnold GP. 1977. The efficiency of the Granton otter trawl determined by sector-scanning sonar and acoustic transponding tags. Rapp. P.-v. Réun. Const. Int. Explor. Mer. 170: 45–51. Hawkins AD and Sand O. 1977. Directional hearing in the median vertical plane by cod. J. Comp. Physiol. 122: 1–8. He P. 1991. Swimming endurance of Atlantic cod, Gadus morhua L., at low temperatures. Fish. Res. 12: 65–73. He P. 1993. Swimming speeds of marine fish in relation to fishing gears. ICES Mar. Sci. Symp. 196: 183–189. He P and Wardle CS. 1988. Endurance at intermediate swimming speeds of Atlantic mackerel, Scomber scombrus L., herring, Clupea harengus L., and saithe, Pollachius virens L. J. Fish Biol. 35: 255–266.

98

Fish Behavior near Fishing Gears during Capture Processes

He P, Smith T and Bouchard C. 2008. Fish behavior and species separation for the Gulf of Maine multispecies trawls. J. Ocean Technol. 3(2): 59–77. Hemmings CC. 1969. Observations on the behavior of fish during capture by the Danish seine net, and their relation to herding by trawl bridles. FAO Fish. Rep. 62(3): 645–655. Hemmings CC. 1973. Direct observation of the behavior of fish in relation to fishing gear. Helgoländer wiss. Meeresunters 24: 348–360. High WL. 1969. SCUBA diving, a valuable tool for the investigating the behavior of fish within the influence of fishing gear. FAO Fish. Rep. 62(2): 253–667. Hjellvik V, Michalsen K, Aglen A and Nakken O. 2003. An attempt at estimating the effective fishing height of the bottom trawl using acoustic survey recordings. ICES J. Mar. Sci. 60: 967–979. Hunter JR and Wisby WJ. 1964. Net avoidance behavior of carp and other species of fish. J. Fish. Res. Bd. Can. 21: 613–633. Ingólfsson OA and Jørgensen T. 2006. Escapement of gadoid fish beneath a commercial bottom trawl: relevance to the overall trawl selectivity. Fish. Res. 79: 303–312. Inoue Y, Matsushita Y and Arimoto T. 1993. The reaction behavior of walleye pollock (Theragra chalcogramma) in a deep/low-temperature trawl fishing ground. ICES J. Mar. Sci. 196: 77–79. Jones EG, Glass CW and Milliken HO. 2004. The reaction and behavior of fish to visual components of fishing gears and its effect on catchability in survey and commercial situations. ICES FTFB Working Group meeting, Gdynia, Poland. April 20–23, 2004. 58 pp. Jones EG, Summerbell K and O’Neil F. 2008. The influence of towing speed and fish density on the behavior of haddock in a trawl cod-end. Fish. Res. 94: 166–174. Jørgensen R, Handegard NO, Gjøsæter H and Slotte A. 2004. Possible vessel avoidance behavior of capelin in a feeding area and on a spawning ground. Fish. Res. 69: 251–261. Jørgensen T, Ingólfsson OA, Graham N and Isaksen B. 2006. Size selection of cod by rigid grids. Is anything gained compared with diamond mesh codends only? Fish. Res. 79: 337–348. Kelleher K. 2005. Discards in the world’s marine fisheries. An update. FAO Fish. Tech. Pap. 470: 131 pp. Kim Y-H and Wardle CS. 1998a. Measuring the brightness contrast of fishing gear, the visual stimulus for fish capture. Fish. Res. 34: 151–164.

Kim Y-H and Wardle CS. 1998b. Modelling the visual stimulus of towed fishing gear. Fish. Res. 34: 165–177. Kim Y-H and Wardle CS. 2003. Optomotor response and erratic response: quantitative analysis of fish reaction to towed fishing gears. Fish. Res. 60: 455–470. Kiselev ON. 1968. Visual observations of fish behavior in natural conditions. pp. 18–22. In: Fish Behavior and Fishing Techniques. Ministry of Fisheries of the USSR, Murmansk. Korotkov VK. 1970. The speed and endurance of fishes escaping from a trawl. J. Ichthyol. 10: 832–836. Krag LA, Frandsen RP and Madsen N. 2008. Evaluation of a simple means to reduce discard in the Kattegat-Skagerrak Nephrops (Nephrops norvegicus) fishery: commercial testing of different codends and square-mesh panels. Fish. Res. 91: 175–186. Kvamme C and Isaksen B. 2004. Total selectivity of a commercial cod trawl with and without a grid mounted: grid and codend selectivity of north-east Arctic cod. Fish. Res. 68: 305–318. Laevastu T and Favorite F. 1988. Fishing and Stock Fluctuations. Farnham: Fishing News Books. 239 pp. Langeland MR. 2005. Escapement of Fish under a Survey Trawl: The Effect of Ground Gear Configuration. MSc thesis, University of Bergen, Norway. 77 pp. Lima SL and Dill LM. 1990. Behavioral decisions made under the risk of predation: a review and prospectus. Can. J. Zool. 68: 619–640. Madsen N, Moth-Poulsen T and Lowry N. 1998. Selectivity experiments with window codends fished in the Baltic Sea cod (Gadus morhua) fishery. Fish. Res. 36: 1–14. Madsen N, Skeide R, Breen M, Krag LA, Huse I and Soldal AV. 2008. Selectivity in a trawl codend during haul-back operation: an overlooked phenomenon. Fish. Res. 91: 168–174. Main J and Sangster GI. 1981. A study of the fish capture process in a bottom trawl by direct observations from a towed underwater vehicle. Scot. Fish. Res. Rep. 23: 1–23. Main J and Sangster GI. 1983. Fish reactions to trawl gear: a study comparing light and heavy ground gear. Scot. Fish. Res. Rep. 27: 1–17. Main J and Sangster GI. 1991. A study of haddock (Melanogrammus aeglefinus (L.)) behavior in diamond and square mesh codends. Scot. Fish. Work. Pap. 19/91.

Fish Behavior near Bottom Trawls Manley GA, Popper AN and Fay RR. 2004. Evolution of the Vertebrate Auditory System. New York: Springer. 415 pp. Mann DA, Wilson DW, Song J and Popper AN. 2009. Hearing sensitivity of the walleye pollock. Trans. Am. Fish. Soc. 138: 1000–1008. Marténez M, Couture P and Guderley H. 1999. Temporal changes in tissue metabolic capacities of wild Atlantic cod Gadus morhua (L.), from Newfoundland. Fish Physiol. Biochem. 20: 181–191. Marténez M, Guderley H, Nelson JA, Webber D and Dutil J-D. 2002. Once a fast cod, always a fast cod: maintenance of performance hierarchies despite changing food availability in cod (Gadus morhua). Physiol. Biochem. Zool. 75: 90–100 Marténez M, Guderley H, Dutil J-D, Winger PD, He P and Walsh SJ. 2003. Condition, prolonged swimming performance and muscle metabolic capacities of cod Gadus morhua. J. Exp. Biol. 206: 503–511. Martyshevskii VN and Korotkov VK. 1968. Behavior of some fish in the zone of action of the trawl. In: Fish Behavior and Fishing Techniques. pp. 73–79. Murmansk: Ministry of Fisheries of the USSR. McQuinn IH and Winger PD. 2003. Tilt angle and target strength: target tracking of Atlantic cod (Gadus morhua) during trawling. ICES J. Mar. Sci. 60: 575–583. Michalsen K, Aglen A, Somerton D, Svellingen I and Øvredal JT. 1999. Quantifying the amount of fish unavailable to a bottom trawl by use of an upward looking transducer. ICES CM. 1999/J:08. 19 pp. Milliken HO and DeAlteris JT. 2004. Evaluation of a large-mesh panel to reduce the flatfish bycatch in the small-mesh bottom trawls used in the New England silver hake fishery. N. Am. J. Fish. Manag. 24: 20–32. Misund OA. 1994. Swimming behavior of fish schools in connection with capture by purse seine and pelagic trawl. In: Fernö A and Olsen S (eds). Marine Fish Behavior in Capture and Abundance Estimation. pp 84–106. Oxford: Fishing News Books. Misund OA, Øvredal JT and Hafsteinsson MT. 1996. Reactions of herring schools to the sound field of a survey vessel. Aquat. Liv. Resour. 9: 5–11. Mitson RB. 1993. Underwater noise radiated by research vessels. ICES Mar. Sci. Symp. 196: 147–152. Mitson RB and Knudsen HP. 2003. Causes and effects of underwater noise on fish abundance estimation. Aquat. Liv. Resour. 16: 255–263. Mohr H. 1971. Behavior patterns of different herring stocks in relation to ship and midwater trawl. In:

99

Kristjonsson M. (ed). Modern Fishing Gear of the World, Vol. 3. pp 368–371. London: Fishing News Books. Nøttestad L, Aksland M, Beltestad A, Fernö A, Johannessen A and Misund OA. 1996. Schooling dynamics of Norwegian spring spawning herring (Clupea harengus L.) in a coastal spawning area. Sarsia 80: 277–284. Nunnalle EP. 1991. An investigation of the avoidance reactions of Pacific whiting (Merluccius productus) to demersal midwater trawl gear. ICES CM 1991/B:5. Okonski S. 1969. The influence of fish behavior on the choice of the trawl net shape and size. FAO Fish. Rep. 62(2): 389–407. Olsen K. 1969. A comparison of acoustic threshold in cod with recordings of ship-noise. FAO Fish. Rep. 62(2): 431–438. Olsen K. 1971. Influence of vessel noise on behavior of herring. In: Kristjonsson M (ed). Modern Fishing Gear of the World, Vol. 3. pp 291–294. London: Fishing News Books. Olsen K, Angell J and Løvik A. 1983b. Quantitative estimations of the influence of fish behavior on acoustically determined fish abundance. FAO Fish. Rep. 300: 139–149. Olsen K, Angell J, Pettersen F and Løvik A. 1983a. Observed fish reactions to a surveying vessel with special reference to herring, cod, capelin, and polar cod. FAO Fish. Rep. 300: 131–138. Ona E and Godø OR. 1990. Fish reaction to trawling noise: the significance for trawl sampling. Rapp. P.-v. Réun. Const. Int. Explor. Mer. 189: 159–166. Ona E, Godø OR, Handegard NO, Hjellvik V, Patel R and Pedersen G. 2007. Silent research vessels are not quiet. J. Acoust. Soc. Am. 121: 145–150. O’Neill FG, McKay SJ, Ward JN, Strickland A, Kynoch RJ and Zuur AF. 2003. An investigation of the relationship between sea state induced vessel motion and cod-end selection. Fish. Res. 60: 107–130. Özbilgin H. 2002. Effect of temperature change on maximum swimming speed of whiting, Merlangius merlangus (Linnaeus, 1758). Turk. J. Zool. 26: 255–262. Özbilgin H and Glass CW. 2004. Role of learning in mesh penetration behavior of haddock (Melanogrammus aeglefinus). ICES J. Mar. Sci. 61: 1190–1194. Özbilgin H and Wardle CS. 2002. Effect of seasonal temperature changes on the escape behavior of haddock, Melanogrammus aeglefinus, from the codend. Fish. Res. 58: 323–331.

100

Fish Behavior near Fishing Gears during Capture Processes

Özbilgin H, Ferro RST, Robertson JHB, Holtrop G and Kynoch RJ. 2006. Seasonal variation in trawl codend selection of northern North Sea haddock. ICES J. Mar. Sci. 63: 737–748. Parrish BB, Blaxter JHS, Pope JA and Osborn RH. 1969. Underwater photography of fish behavior in response to trawls. FAO Fish. Rep. 62(3): 873–884. Parsons GR and Foster DG. 2007. Swimming performance and behavior of red snapper: its application to bycatch reduction. Am. Fish. Soc. Symp. 60: 59–75. Pavlov DS and Kasumyan AO. 1995. A review of Russian studies on the behavior and sensory systems of fishes. J. Ichthyol. 35(1): 124–138. Peake SJ and Farrell AP. 2004. Locomotory behavior and post-exercise physiology in relation to swimming speed, gait transition and metabolism in freeswimming smallmouth bass (Micropterus dolomieu). J. Exp. Biol. 207: 1563–1575. Peake SJ and Farrell AP. 2006. Fatigue is a behavioral response in respirometer-confined smallmouth bass. J. Fish Biol. 68: 1742–1755. Piasente M, Knuckey IA, Eayrs S and McShane PE. 2004. In situ examination of the behavior of fish in response to demersal trawl nets in an Australian trawl fishery. Mar. Freshw. Res. 55: 825–835. Popper AN. 2003. Effects of anthropogenic sounds on fishes. Fisheries 28(10): 24–31. Popper AN, Plachta DTT, Mann DA and Higgs D. 2004. Response of clupeid fish to ultrasound: a review. ICES J. Mar. Sci. 61: 1057–1061. Pyanov AI. 1993. Fish learning in response to trawl fishing. ICES Mar. Sci. Symp. 196: 12–16. Pyanov AI and Zhuykov AY. 1993. Conditioned reflex in fishes for avoidance of active fishing gear. J. Ichthyol. 33(8): 40–50. Reid DG, Allen VJ, Bova DJ, Jones EG, Kynoch RJ, Peach KJ, Fernandes PG and Turrell WR. 2007. Anglerfish catchability for swept-area abundance estimates in a new survey trawl. ICES J. Mar. Sci. 64: 1503–1511. Riedel R and DeAlteris JT. 1995. Factors affecting hydrodynamic performance of the Nordmøre grate system: a bycatch reduction device used in the Gulf of Maine shrimp fishery. Fish. Res. 24: 181–198. Rose CS. 1995. Behavior of North Pacific groundfish encountering trawls: applications to reduce bycatch. pp. 235–241. In: Solving Bycatch: Considerations for Today, and Tomorrow. Alaska Sea Grant College Program Report No. 96–03, University of Alaska Fairbanks.

Ryer CH. 2008. A review of flatfish behavior relative to trawls. Fish. Res. 90: 138–146. Ryer CH and Barnett LAK. 2006. Influence of illumination and temperature upon flatfish reactivity and herding behavior: potential implications for trawl capture efficiency. Fish. Res. 81: 242–250. Sand O and Karlsen HE. 1986. Detection of infrasound by the Atlantic cod. J. Exp. Biol. 125: 197–204. Schuijf A. 1975. Directional hearing of cod (Gadus morhua) under approximate free field conditions. J. Comp. Physiol. 98: 307–332. Skaret G, Axelsen BE, Nøttestad L, Fernö A and Johannessen A. 2005. The behavior of spawning herring in relation to a survey vessel. ICES J. Mar. Sci. 62: 1061–1064. Skaret G, Slotte A, Handegard NO, Axelsen BE and Jørgensen R. 2006. Pre-spawning herring in a protected area showed only moderate reaction to a surveying vessel. Fish. Res. 78: 359–367. Somerton DA. 2004. Do Pacific cod (Gadus macrocephalus) and walleye pollock (Theregra chalcogramma) lack a herding response to the doors, bridles, and mudclouds of survey trawls? ICES J. Mar. Sci. 61: 1186–1189. Soria M, Fréon P and Gerlotto F. 1996. Analysis of vessel influence on spatial behavior of fish schools using a multi-beam sonar and consequences for biomass estimates by echo-sounder. ICES J. Mar. Sci. 53: 453–458. Soria M, Gerlotto F and Fréon P. 1993. Study of learning capabilities of tropical clupeoids using an artificial stimulus. ICES Mar. Sci. Symp. 196: 17–20. Stankowich T and Blumstein DT. 2005. Fear in animals: a meta-analysis and review of risk assessment. Proc. R. Soc. (B) 272: 2627–2634. Strange ES. 1984. Review of the fishing trials with Granton and Saro deep sea trawl gear 1963–1967. Scot. Fish. Work. Pap. 8/84. Tanaka E, Matuda K and Hirayama N. 1991. A simulation model of gear efficiencies of trawlers for flatfish. Nippon Suisan Gakkaishi 57: 1019–1028. Thomsen B. 1993. Selective flatfish trawling. ICES Mar. Sci. Symp. 196: 161–164. Urquhart GG and Stewart PAM. 1993. A review of techniques for the observation of fish behavior in the sea. ICES Mar. Sci. Symp. 196: 135–139. Vabø R, Olsen K and Huse I. 2002. The effect of vessel avoidance of wintering Norwegian spring spawning herring. Fish. Res. 58: 59–77. Videler JJ. 1993. Fish Swimming. London: Chapman and Hall. 260 p.

Fish Behavior near Bottom Trawls Videler JJ and Wardle CS. 1991. Fish swimming stride by stride: speed limits and endurance. Rev. Fish Biol. Fish. 1: 23–40. Voegeli FA, Smale MJ, Webber DM, Andrade Y and O’Dor RK. 2001. Ultrasonic telemetry, tracking and automated technology for sharks. Environ. Biol. Fish. 60: 267–281. Wakeford J. 2004. An investigation into the influence of the Super-Shooted TED and several types of bycatch reduction devices on water flow through a prawn trawl codend. In: Brewer DT, Heales DS, Eayrs SJ, Taylor BR, Day G, Sen S, et al. (eds). Assessment and improvement of TEDs and BRDs in the NPF: a co-operative approach between fishers, scientists, fisheries technologists, economists and conservationists. Final Report on FRDC Project 2000/173. CSIRO Cleveland. 412 pp. Walsh SJ. 1992. Size-dependent selection at the footgear of a groundfish survey trawl. N. Am. J. Fish. Manag. 12: 625–633. Walsh SJ. 1996. Efficiency of bottom sampling trawls in deriving survey abundance indices. NAFO Sci. Coun. Stud. 28: 9–24. Walsh SJ and Hickey WM. 1993. Behavioral reactions of demersal fish to bottom trawls at various light conditions. ICES Mar. Sci. Symp. 196: 68–76. Walsh SJ, Engås A, Ferro R, Fonteyne R and van Marlen B. 2002. To catch or conserve more fish: the evolution of fishing technology in fisheries science. ICES Mar. Sci. Symp. 215: 493–503. Wardle CS. 1983. Fish reaction to towed fishing gears. In: MacDonald AG and Priede IG (eds). Experimental Biology at Sea. pp 167–196. London: Academic Press. Wardle CS. 1984. Fish behavior. Trawl efficiency and energy saving strategies. FIIT Working Paper. Rome: FAO Fishing Technology Service. Sept. 1984. Wardle CS. 1986. Fish behavior and fishing gear. In: Pitcher TJ (ed). Behavior of Teleost Fishes. pp 463–495. London: Chapman and Hall. Wardle CS. 1992. Studies of fish behavior leading to a reduction of bycatch. In: Jones RP (ed). Proceedings of an International Conference on Shrimp Bycatch. May 24–27, 1992, Lake Buena Vista, Florida. pp 163–184. Tallahassee, FL: Southeastern Fisheries Association. Wardle CS. 1993. Fish behavior and fishing gear. In: Pitcher TJ (ed). Behavior of Teleost Fishes. 2nd ed. pp 609–643. London: Chapman and Hall. Watson JW. 1988. Fish behavior and trawl design: Potential for selective trawl development. Proc.

101

World Symp. Fish. Gear and Fish. Vessel Design. pp 25–29. St. John’s, Newfoundland: Marine Institute. Watson JW. 1992. Status of knowledge in the United States relating fish behavior to the reduction of bycatch. In: Jones RP (ed). Proceedings of an International Conference on Shrimp Bycatch. May 24–27, 1992, Lake Buena Vista, Florida. pp 185– 196. Tallahassee, FL: Southeastern Fisheries Association. Watson JW. 2007. Reconciling fisheries with conservation through programs to develop improved fishing technologies in the Unites States. In: Kennelly SJ (ed). By-catch Reduction in the World’s Fisheries. pp 23–36. The Netherlands: Springer. Watson JW, Foster D, Nichols S, Shah A, ScottDenton E and Nance J. 1999. The development of bycatch reduction technology in the Southeastern United States shrimp fishery. Mar. Technol. Soc. J. 33(2): 51–56. Webb PW. 1994a. Exercise performance of fish. In: Jones JH (ed). Comparative Vertebrate Exercise Physiology: Phyletic Adaptations. pp 1–49. San Diego: Academic Press. Webb PW. 1994b. The biology of fish swimming. In: Maddock L, Bone Q and Rayner JMV (eds). Mechanics and Physiology of Animal Swimming. pp 45–62. Cambridge: Cambridge University Press. Weinberg KL and Munro PT. 1999. The effect of artificial light on escapement beneath a survey trawl. ICES J. Mar. Sci. 56: 266–274. Wileman DA, Ferro RST, Fontaine R and Millar RB. 1996. Manual of methods of measuring the selectivity of towed fishing gears. ICES Coop. Res. Rep. 215: 126 pp. Winger PD. 2004. Effect of Environmental Conditions on the Natural Activity Rhythms and Bottom Trawl Catchability of Atlantic Cod (Gadus morhua). PhD thesis, Memorial University of Newfoundland. 151 pp. Winger PD, DeLouche H and Legge G. 2006. Designing and testing new fishing gears: the value of a flume tank. Mar. Technol. Soc. J. 40(3): 44–49. Winger PD, He P and Walsh SJ. 1999. Swimming endurance of American plaice (Hippoglossoides platessoides) and its role in fish capture. ICES J. Mar. Sci. 56: 252–265. Winger PD, He P and Walsh SJ. 2000. Factors affecting the swimming endurance and catchability of Atlantic cod (Gadus morhua). Can. J. Fish. Aquat. Sci. 57: 1200–1207.

102

Fish Behavior near Fishing Gears during Capture Processes

Winger PD, Walsh SJ, He P and Brown JA. 2004. Simulating trawl herding in flatfish: the role of fish length on behavior and swimming characteristics. ICES J. Mar. Sci. 61: 1179–1185. Yanase K, Eayrs S and Arimoto T. 2007. Influence of water temperature and fish length on the maximum swimming speed of sand flathead, Platycephalus bassensis: implications for trawl selectivity. Fish. Res. 84: 180–188. Yanase K, Eayrs S and Arimoto T. 2009. Quantitative analysis of the behavior of the flatheads (Platycephalidae) during the trawl capture process

as determined by real-time multiple observations. Fish. Res. 95: 28–39. Ydenberg RC and Dill LM. 1986. The economics of fleeing from predators. Adv. Study Behav. 16: 229–249. Zaferman ML and Serebrov LI. 1989. On fish injuring when escaping through the trawl mesh. ICES CM.1989/B:18. 14 pp. Zhuykov AY and Pyanov AI. 1993. Differences in behavior of fish with different learning ability as demonstrated with a model of a trap net. J. Ichthyol. 33(9): 141–147.

Fish Behavior near Bottom Trawls SPECIES MENTIONED IN THE TEXT American shad, Alosa sapidissima Atlantic bumper, Chlorscombrus chrysurus Atlantic cod, cod, Gadus morhua Atlantic mackerel, Scomber scombrus Atlantic salmon, Salmo salar anchovies, Ancoa hepstus blue grenadier, Macruronus novaezelandiae capelin, Mallotus villosus Müll. common dab, Limanda limanda dogfish, Sqaulus acanthias goldfish, Carassius auratus auratus Greenland halibut, Reinhardtius hippoglossoides haddock, Melanogrammus aeglefinus, herring, Clupea harengus jack mackerel, Trachurus japonicus ling, Molva molva New Zealand dory Cyttus novaezelandiae

103

North Sea plaice, Pleuronectes platessa, North Sea pollack, pollock, Pollachius pollachius, Pacific cod, Gadus macrocephalus Pacific halibut, Hippoglossus stenolepis red snapper, Lutjanus campechanus rock sole, Pleuronectes bilineatus saithe, Pollachius virens sand flathead, Platycephalus bassensis sandeels, Ammodytidae sp. skates, Raja sp. Spanish sardine, Sardinella anchovia spotted warehou, Seriolella punctata squid, Loligo pealeii Tiger flathead, Neoplatycephalus richardsoni whiting, Merlangius merlangus walleye pollock, Theragra chalcogramma yellowtail flounder, Limanda ferruginea

Chapter 5 Fish Behavior in Relation to Longlines Svein Løkkeborg, Anders Fernö, and Odd-Børre Humborstad 5.1 INTRODUCTION Fishing by hook and line, including handlining, rod and reel fishing, jigging, trolling, and longlining, is a method used all over the world with a very wide range of vessels, from small artisanal boats to large mechanized longliners. This traditional fishing method has been one of the most important fish capture techniques used since the Stone Age (Bjordal and Løkkeborg 1996). Archaeological excavations show that various materials such as stone, bone, sea shells, and horn have been used for hooks throughout history, and some of these hooks, used on a handline, are believed to be more than 4000 years old. Longlining is a more recent development that was not used on a large scale until industrialization made large numbers of hooks available at a reasonable cost. The use of longlines in Norwegian waters can be dated back to the early 1700s, and Spanish sources indicate that the use of longline gear in the Mediterranean dates back to at least the 1500s (Bjordal and Løkkeborg 1996). A longline is a passive and stationary form of fishing gear. The principle of longline fishing is to attract fish to ingest hooks by using odor-releasing bait that entices the fish to ingest the hook-bait combination. Although longline gear is thus a simple fishing method, there are wide variations in gear construction and mode of operation. Moreover, as the gear can be set either on the bottom or drifting in the water column, a large variety of species can be targeted, from bottom-dwelling flatfish to highly migratory tunas. Longline fishing is therefore a very versatile fishing method.

Understanding the principle behind longline fishing requires knowledge of how fish search for and capture food. The literature on feeding behavior and interactions between fish and longline gear parameters has been synthesized in earlier reviews (Bjordal and Løkkeborg 1996; Løkkeborg 1994). In the course of the past decade, however, a number of behavioral studies have provided new knowledge of great relevance to the performance of longlining, especially regarding food search, feeding motivation, and effects of environmental variables. This literature is reviewed here with the aim of providing a more comprehensive understanding of the interactions between foraging fish and baited hooks. Until recently, concerns about sustainable exploitation of marine resources have focused mainly on the proper management of commercially valuable fish stocks, and longlining has been considered as an environmentally friendly fishing method. However, fishing causes incidental mortality of many nontarget species, and of particular concern are long-lived species such as seabirds, sea turtles, elasmobranches, and marine mammals (Heppell et al. 2005). Thus, gradually more attention is being paid to the general impact of fishing activities on the marine ecosystem (Gislason, 1994). With regard to longline fishing, most attention has been paid to bycatch of seabirds and the development of mitigation measures to solve this problem. The failure of sea turtle populations to recover (six of the seven species are listed as endangered) has also been attributed in part to incidental capture by fishing gear, including longlines. A third important

105

106

Fish Behavior near Fishing Gears during Capture Processes

conservation issue related to longline fishing is shark management [see the U.N. World Food and Agriculture Organization (FAO)’s International Plan of Action for the Conservation and Management of Sharks]. With regard to the principles of ecologically sustainable fishing, this chapter thus focuses on interactions with seabirds, sea turtles, and sharks and discusses how our current knowledge of behavior and gear operation may help to mitigate these impacts. 5.2 WORLDWIDE LONGLINE FISHERIES Baited longlines are used in all oceans and seas, and large proportions (15% to 90%) of several important fish resources are caught by longlines (Bjordal and Løkkeborg 1996). Demersal (bottom) longline fishing takes place on the relatively shallow waters of the continental shelves and slopes down to depths of 3000 m. Many demersal longline fishing operations are carried out in cold waters at relatively high latitudes and target a large variety of fish species. Pelagic (drifting) longlining takes place in deep waters, generally off continental shelves. This fishing method is operated in all oceans from temperate to tropical waters and targets mainly tuna and swordfish (billfish) species. The most important longline fisheries in terms of tonnage of landed catches are described below and summarized in Table 5.1. More detailed information on effort and landings has been provided by Bjordal and Løkkeborg (1996) and Brothers et al. (1999a). 5.2.1 Northeastern Atlantic Ocean Demersal Fisheries Nearly all longline fisheries in the northeast Atlantic region target demersal species. The Norwegian, Icelandic, and Faeroese fleets dominate the longline fishery in this region. The vessels move seasonally between fishing grounds targeting different species, mainly Atlantic cod (Gadus morhua), haddock (Melanogrammus aeglefinus), tusk (Brosme brosme), ling (Molva molva), wolffish (Anarhichas lupus), and Greenland halibut (Reinhardtius hippoglossoides). The longline fleets operate on the shelf and shelf edge on both coastal and high-sea

Table 5.1. Regions and Main Target Species for the World’s Most Important (in Terms of Landings) Longline Fisheries. Region Demersal longlining: Northeastern Atlantic Northwestern Atlantic Northeastern Pacific Northwestern Pacific Southern Ocean Pelagic longlining: Pacific Ocean Atlantic Ocean Indian Ocean

Main Target Species

Cod, haddock, tusk, ling Cod, haddock, spiny dogfish, white hake Cod, halibut, sablefish Walleye pollock, cod Patagonian toothfish Bigeye tuna, yellowfin tuna, albacore Bigeye tuna, albacore, swordfish Yellowfin tuna, bigeye tuna

fishing grounds from west of the British Isles to the Barents Sea. Most coastal vessels are relatively small, and their longlines are baited onshore by hand and coiled into tubs (baskets). Larger vessels using autoline systems operate farther offshore and on high seas fishing grounds. These vessels remain at sea for several weeks and may set more than 30,000 hooks per day. 5.2.2 Northwestern Atlantic Ocean Demersal Fisheries Demersal longlining in this region is carried out by Canadian vessels fishing off Nova Scotia, Newfoundland, and Labrador and in the Gulf of St. Lawrence and by U.S. vessels fishing in the Gulf of Maine and on the Georges Bank. Groundfish species caught include Atlantic cod, haddock, spiny dogfish (Squalus acanthias), Greenland halibut, tusk (cusk), American plaice (Hippoglossoides platessoides), saithe (Pollachius virens), and white hake (Urophycis tenuis). In 1992 a moratorium on groundfish (primarily for cod) came into force in

Fish Behavior in Relation to Longlines Canadian Atlantic waters, but in 1997 a limited longline fishery for cod recommenced around Newfoundland. The Canadian and U.S. longline fleets are dominated by small vessels (less than 45 feet) that bait their lines onshore by hand. The larger vessels (which are small compared with Norwegian and Icelandic vessels) are not mechanized but may remain at sea for several days, baiting their lines by hand.

5.2.3 Northeastern Pacific Ocean Demersal Fisheries This region is divided into three areas—Alaska (Gulf of Alaska and Bering Sea), Canada (British Colombia), and Washington-Oregon-California— and the longline fleet consists entirely of domestic vessels from the United States and Canada. Fishing activity is highest in the Gulf of Alaska and the Bering Sea, where the most important fishing grounds are located. Landings from this area are dominated by Pacific cod (Gadus macrocephalus), Pacific halibut (Hippoglossus stenolepis), and sablefish (blackcod) (Anoplopoma fimbria). Alaskan fisheries account for all of the cod landings and approximately 80% of halibut landings from the northeastern Pacific region. The vessels range in size from relatively small boats to large factory longliners using mechanized longline systems. The Pacific halibut fishery in Alaska was an open access fishery until 1995, when an individual vessel quota system was put into effect. As a result, the number of vessels fell, and the active fishing period changed from a few days to several months.

5.2.4 Northwestern Pacific Ocean Demersal Fisheries Demersal longlining in the northwest Pacific region is carried out by vessels from Japan, South Korea, China, and Taiwan. The vessels operate on both coastal and high-sea fishing grounds, and both demersal longlines and longlines floated off the seabed are used. The longline fleet is dominated by small boats that operate on the coastal fishing grounds. The most important species in this region

107

are walleye pollock (Theragra chalcogrammus), Pacific cod, and Pacific halibut. 5.2.5 Southern Ocean Demersal Fisheries Demersal longline fishing in the Southern Ocean commenced in 1988 and is primarily targeting Patagonian toothfish (Dissostichus eleginoides). This valuable species has a circumpolar distribution south of 55 degrees S, and longline fishing is carried out off the southern part of South America (Argentina, Falkland Islands, and Chile) and around many of the islands of the Southern Ocean, especially those of southern Atlantic and Indian Oceans (e.g., South Georgia, Prince Edward, Crozet, and Kerguelen Islands). The fishery is carried out in very deep waters (down to 3000 m) by large mechanized vessels from several countries. Annual regional quotas are set by the Commission for the Conservation on the Antarctic Marine Living Resources (CCAMLR), but illegal, unreported, and unregulated (IUU) fishing is a problem. 5.2.6 Pelagic Longline Fisheries A large fleet of tuna vessels operates with pelagic longlines in the Indian Ocean and in the central and southern Atlantic and Pacific Oceans. The home ports of most of these vessels are in Japan, Taiwan, and South Korea, which are the major operators in the tuna longline fisheries and land the great majority of the catches. The most important species by weight is bigeye tuna (Thunnus obesus), followed by yellowfin tuna (T. albacares) and albacore (T. alalunga). Catches of bluefin tuna (T. thynnus) are relatively small, but the prices for this species on the Japanese sashimi (raw fish) market are much higher than for the other species. There is also an important pelagic longline fishery targeting swordfish, but this fishery is much smaller than the tuna fisheries (about 10%). The Atlantic Ocean is the most important region for the swordfish longline fishery, and the European and North American fleets are the dominant operators. The tuna longline fisheries have changed over time, driven by economic and market forces. Traditional tuna longlines were set at a maximum depth of about 170 m, and the Japanese longline fleet targeted mainly yellowfin tuna and albacore to supply the U.S. canning industry. In the 1970s,

108

Fish Behavior near Fishing Gears during Capture Processes

a “deep longline” was developed to target bigeye tuna, which is distributed deeper in the water column (more than 200 m) than yellowfin tuna and albacore (less than 200 m). Bigeye tuna fetch a higher price on the sashimi market, and the fleet gradually switched to target this species. The substantial increase in fuel and labor costs in the 1970s forced the Japanese fleet to reduce their export to the low-value canning market and concentrate on high-value products for the sashimi market in Japan. Later the same market-driven shift was seen in the South Korean tuna fleet. 5.3 DESCRIPTION OF THE GEAR Longlining is a versatile fishing method, and there are numerous ways of rigging longlines according to target species and fishing conditions (e.g., bottom type and topography, depth, current). However, all longline gear used in the worldwide fishery is based on the same basic unit, which consists of four parts (Fig. 5.1): • The mainline (groundline); a longline (rope), which has given the gear its name and to which the snood and hook are attached at intervals • The snood (gangion, branchline); a thinner line attached to the main line at one end and to the hook at the other • The hook (large varieties in shape and size) • The bait (cut in pieces or whole finfish, shellfish, or squid/octopus) The unit in practical rigging and use of longlines, however, is the basket, tub, skate, or, in mechanized longlining, magazine. A certain length of mainline with a given number of hooks is baited and coiled into a tub (basket) prior to setting the longline. In mechanized longlining, this unit is replaced by a magazine holding the hooks with the mainline hanging in coils. During setting, several tubs/ baskets or magazines are linked to make a fleet of longlines, which may vary in length from a few hundred meters in small-scale coastal fishing to more than 50 km in high seas mechanized longline fisheries. The mainline and snood vary in length, dimension, and type of material according to the type of fishery. Variations are also found in spacing

Figure 5.1. The basic unit of longline gear consists of four parts: mainline, snood, hook, and bait.

between neighboring snoods/hooks (hook spacing) and in the way in which the snood is attached to the mainline. The widest variations in longline gear characteristics are seen in hook types, in terms of both shape and size. The type of bait used varies according to the food preference of the target

Fish Behavior in Relation to Longlines species, availability, and price, and different baits may be used on the same longline in multispecies fisheries. The most common materials used for mainline and snood are polyamide (nylon) and polyester, both of which have a higher density than seawater and thus sink. More recently, gear manufacturers have developed ropes made from a mixture of different types of materials to obtain longlines with improved properties—in particular, higher breaking strength. Multifilament and monofilament lines have different properties and are used accordingly. Due to their high resistance to chafing, multifilament lines are commonly used for demersal longlines, but monofilament lines may be used for snoods. Monofilament lines are preferred in pelagic and semipelagic longlining because they have higher catching efficiency, probably because they are less visible. The length of the snood varies from relatively short (less than 1 m) in demersal longlining to very long in tuna longlining (up to 30–40 m). The snood may be attached to the mainline in three different ways. It has traditionally been tied to the mainline by a simple knot. More recently, swivels have been used, first on monofilament longlines and later also on demersal multifilament longlines. Swivels prevent twisting of the snood and have been shown to improve catching efficiency. The third way of attachment is by a metal clip that holds the snood and is manually attached to the mainline as it is deployed. This method is commonly used in pelagic tuna longlining and in the fishery for Pacific halibut. With this method, the hook spacing needs to be relatively wide, as the snood is attached during the setting operation. There are wide variations in the types of hook used in longline fisheries. The J-shaped hook has traditionally been most common and entirely dominated the fisheries for groundfish until the mid1980s. During the 1980s, several new hook designs were developed and tested in comparative fishing experiments. These hooks yielded substantial increases in catch rates for many demersal species compared with traditional J-hooks, and marked shifts in hook types were seen in several longline fisheries. For examples, an almost total conversion from J-hooks to circle and EZ-baiter hooks occurred,

109

respectively, in the Pacific halibut and Northeast Atlantic autoline fisheries. Furthermore, using circle hooks instead of traditional J-hooks has proved to be an efficient mitigation measure to reduce incidental bycatch of sea turtles in pelagic longlining (see Section 5.3). Bait type is an important factor affecting the species selectivity of longlining. This is explained by the fact that feeding attractants and stimulants are species specific. There are thus large variations in bait types among different longline fisheries, and fishermen have learned from experience which bait types are most effective for a particular species. Price and availability are other factors that determine the fishermen’s choice of bait type. The bait specimens are usually cut into pieces in demersal longlining, whereas whole specimens are used to target large pelagic species such as tunas. Much effort has been put into the development of artificial baits, but to date no alternative bait has been widely adopted by commercial vessels. There are three main methods of setting longlines, with a wide range of variation in each. Bottom-set or demersal longlines are the traditional and most common setting method, in which the baited lines are laid on the seabed. Demersal longlines are set either as a single mainline or using the double line system (also called the Spanish system). The latter system has a second safety or mother line. This line is unweighted and floats above the mainline to which it is attached at several points. The double-line system is designed to be used in areas with rough bottoms and strong currents and allows the mainline to be retrieved even if it breaks. Bottom-set longlines are used to target demersal species such as cod, halibut, hake (Merluccius spp.), ling, and Patagonian toothfish. Pelagic longlines are suspended from floats at the surface and drift freely in the sea as they are not anchored to the bottom. This setting method is most frequently used in high-seas fisheries targeting tunas, swordfish, and sharks. The fishing depth of the baited hooks is adjusted by varying the length of the float lines according to the depth inhabited by the target species. Surface longlines targeting swordfish may be set very shallow (10–20 m), whereas pelagic lines targeting big eye tuna are set at a few hundred meters depth.

110

Fish Behavior near Fishing Gears during Capture Processes

The third longline setting method, semipelagic longlining, is an intermediate between demersal and pelagic longlining. The mainline is anchored to the seabed but floated off the bottom. Usually, there are floats and sinkers alternating at specific intervals along the mainline. The sinkers are attached to the mainline by a vertical line, and the length of this line is adjusted to ensure that the baited hooks are suspended at the depth inhabited by the target species. Semipelagic longlining is used for hake and other demersal species that seasonally display semipelagic distribution, such as cod, Pollock, and haddock. 5.4 CHEMORECEPTION AND FOOD SEARCH—THE BASIS FOR BAIT FISHING 5.4.1 The Chemosensory Senses and Chemical Stimuli Properties Fishing with bait exploits the feeding behavior of fish and is based on one of the most fundamental activities in an animal’s life—searching for and capturing food (Løkkeborg 1994). Thus, the effectiveness of baited fishing gear ultimately depends on the behavior of the target species and their activity rhythms, feeding motivation, and locomotory abilities (Stoner 2003). Studying the behavior of foraging fish may therefore assist in better understanding the interactions between the target species and the baited gear. However, vulnerability to baited gears varies because changes in activity, feeding, and locomotion are all affected by environmental variables such as temperature, light level, current velocity, ambient prey density, and the presence of potential bait competitors (Stoner 2004). While many aspects of foraging behavior, such as olfaction and taste, feeding stimulants and thresholds, activity rhythms, feeding motivation, foodsearch strategies, and ability to locate prey, have been well studied in some species (e.g., Caprio 1978; Carr 1982; Hara 1992; Ishida and Kobayashi 1992; Johnstone 1980; Løkkeborg et al. 1989; Løkkeborg and Fernö 1999; Stoner 2003; Sutterlin 1975), research on how feeding biology is affected by environmental variables is still in its infancy (Stoner 2004). A starting point for explaining the feeding behavior in fish is optimal foraging theory (Krebs 1978), which assumes that feeding behavior

is determined by the balance between costs and benefits. Fish detect chemical stimuli through at least two different channels of chemoreception: olfaction (smell) and gustation (taste) (Hara 1992, 1993). The olfactory organ in fish is basically paired structures situated in the snout, each of which has an anterior inlet and posterior outlet for a current of water. The receptor cells of the olfactory epithelium or mucosa are located in the floor of the nasal cavity, where they are arranged in folds or lamella to form a rosette (Hara 1992, 1993; Laberge and Hara 2001). Neuron receptor cells carry external information directly through cranial nerve I (olfactory) to the olfactory bulb of the brain, where they terminate and make synaptic contact with second-order bulbar neurons in the form of glomeruli. Onion- or pearshaped taste buds make up the structural basis for the gustatory organ and consist of gustatory microvilliated receptor and supporting and basal cells. Dependent on the species, taste buds can be found on barbels, fins, the oral cavity, pharynx, and gills and over the entire body surface. Chemical information detected by epithelial receptor cells is transmitted by neurons of cranial nerve VI (facial), IX (glossopharyngeal), or X (vagus) to the central nervous system (Marui and Caprio 1992). The olfactory and gustatory organs of fishes exhibit considerable diversity, reflecting adaptations to different environmental conditions and ecological strategies. The vital function, however, is the same—to extract information about the chemical environment, to control behavior. Molecules dissolved in water mediate both senses, and it can be difficult to identify the specific role that each system plays in any particular behavior. However, there are distinct differences in the anatomy of the olfactory and gustatory organs in fish, and these distinctions are also evident in the functional characteristics of these chemosensory systems (Kasumyan and Døving 2003). The gustatory system tends to be more selective than the olfactory system (i.e., a chemical substance that induces feeding behavior need not possess high palatability), and the gustatory spectra of effective taste substances are more species specific than the olfactory spectra. Because of its lower detection thresholds (Ishida and Kobayashi 1992; Johnstone

Fish Behavior in Relation to Longlines 1980), it is believed that the olfactory system initiates the feeding process (Hara 1993), whereas the gustatory system provides the final sensory evaluation of potential food items (Kasumyan and Døving 2003). Nearly all fish use olfaction for distant prey detection, and fish have the capability of locating prey from a distance well beyond their visual range by means of their chemosensory system (Atema 1980). Chemical stimuli are different from visual and acoustic stimuli with regard to two properties that are crucial for their use in attraction of fish to fishing gear. First, a chemical stimulus has long range and can be detected from very long distances (several hundred meters; Løkkeborg 1998), whereas visual and acoustic (except low frequency sounds) stimuli attenuate rapidly. Second, a chemical stimulus lasts for a long period of time (several hours), whereas the two other types of stimuli fade rapidly after being transmitted. These two properties form the basis of a capture principle that has a long range in both time and space and thus make baits and other types of chemical attractants unique when used to attract target species to fishing gears. The spatial range of chemical attractants means that baits can attract fish from a long distance. The area in which a fish will detect and respond to attractants released from an odor source is called the active space (Bossert and Wilson 1963; Elkinton et al. 1984; McQuinn et al. 1988). Because the rate of diffusion in water is very low, the current is the most important agent for dispersion of chemicals in seawater. Turbulence dilutes attractants released from baits, and the concentration tends to decrease with increasing distance from the odor source (but see Section 5.5.3). The distance over which an odor source may attract fish is thus determined by the initial release rate of feeding attractants, the rate of dilution (i.e., turbulence), and the chemosensory capability and response thresholds to feeding attractants of the target fish. The temporal range of an odor source is determined by the change over time in release rate. The rate of release of feeding attractants from traditional mackerel bait decreases rapidly within the first 1.5 h of immersion in flowing seawater and after 2 to 3 h is only one third of the initial rate (Løkkeborg 1990a). Due to this decrease in release rate over

111

time, the length of active space has been estimated to be halved at 24 h compared with 1 h (Løkkeborg et al. 1995). Thus, the development of a system that extends the period of time over which the attractants are released at high concentrations (i.e., longlasting baits) has great potential in the longline fishery. 5.4.2 Mechanisms to Locate an Olfactory Source Aquatic animals that are chemically stimulated need to move upstream to locate an odor source. Olfactory arousal is therefore often followed by upstream swimming toward the chemical source (Atema 1980). There are two possible mechanisms that can be used to locate an odor source. During rheotaxis, the chemical stimulation releases upstream swimming. Although orientation via rheotaxis is not very accurate (Atema 1980), the animal will gradually approach the source. During a gradient search, the animal reacts to a concentration gradient. Given that the concentration increases as the distance to the odor source is reduced and the animal is able to detect the differences in concentration and orient in relation to these, searching based on gradients should be more accurate. However, the gradients are often weak and inconsistent as turbulence can break up the regular pattern, making it very difficult for animals to locate an odor plume by equalizing the sensory input from bilateral olfactory receptors (Webster et al. 2001, Webster and Weissburg 2001). Fish might alternatively sequentially compare concentrations at different points of space based on time-averaged sampling, but the large spatial and temporal variations in concentrations within chemical odor plumes may make it difficult to detect concentration differences even after long periods of sampling (Webster and Weissburg 2001).Chemically stimulated rheotaxis is therefore regarded as the most likely mechanism used by fish and crustaceans to locate odor sources. 5.4.3 Feeding Attractants Many studies have been carried out to identify the chemical nature of feeding attractants for fish. Amino acids comprise the most important group of compounds identified as feeding attractants and stimulants for fish and crustaceans (see reviews by

112

Fish Behavior near Fishing Gears during Capture Processes

Atema 1980; Carr 1982; Carr et al. 1996; Hara 1975; Jones 1992; Mackie 1982). Nearly all studies of chemically stimulated feeding behavior have shown that food extracts lose their stimulatory capacities when their amino acids have been eliminated (Carr and Derby 1986). Ishida and Kobayashi (1992) concluded that the simple and neutral amino acids alanine and glycine play a role as feeding attractants in several fish species. Of a variety of compounds examined electrophysiologically, amino acids also stand out as a highly stimulatory class of compounds, and this sensitivity to amino acids in fish may have evolved as a food-finding mechanism (Ishida and Kobayashi 1992; Sutterlin 1975). However, other groups of compounds also elicit food search. Carr and Derby (1986) state that the major stimulants in tissue extracts of preys that elicit feeding behavior in marine fish and crustaceans are “common metabolites of low molecular weight that include amino acids, quaternary ammonium compounds, nucleosides and nucleotides, and organic acids.” Polyamines have been shown to be potent olfactory stimulants for goldfish (Carassius auratus), and behavioral assays have indicated that polyamines elicit feeding behavior similar to that elicited by amino acids (Rolen et al. 2003). Interestingly, behavioral observations showed that juvenile sablefish responded to squid odor at dilutions of amino acids far below ambient concentrations, indicating that substances other than amino acids are the primary feeding stimulants in this species (Davis et al. 2006). Studies of the role of specific chemicals as feeding stimulants have demonstrated that a mixture of substances acting in concert rather than a single dominant substance is required to yield extracts with stimulatory capacities similar to those of the total tissue extracts of preys (Carr and Derby 1986; Ellingsen and Døving 1986; Johnstone and Mackie 1990; Jones 1992). Furthermore, reviews of the characteristics of substances that act as feeding stimulants have revealed that different species respond to different substances in food extracts (e.g., Carr and Derby 1986; Mackie 1982). Studies using the same squid extract showed that the major stimulants of feeding behavior were different for turbot (Scophthalmus maximus), rainbow trout

(Oncorhynchus mykiss), and plaice (Pleuronectes platessa) (Adron and Mackie 1978; Mackie 1982; Mackie and Adron 1978). Carr (1982) compared the stimulatory capacity of extracts from different organisms and found that their relative effectiveness differed in pinfish (Lagodon rhomboides) and pigfish (Orthopristis chrysopterus). Similarly, field studies showed that four species of marine fish were attracted by different amino acids (Sutterlin 1975). Fishing experiments in commercial longlining and experiences of fishermen have also shown species-selective effects of baits. When several bait types were compared in fishing trials, squid was found to be the most effective bait for capturing cod and hake (Merluccius sp.), whereas mackerel appeared to be more effective for haddock (Martin and McCracken 1954). Bjordal (1983) found that squid bait caught twice as many ling as mackerel but only 9% more tusk. Comprehensive studies of Japanese tuna longlining have shown the speciesspecific effects of bait type on captures of tuna and marlin (Imai 1972; Imai and Shirakawa 1972; Shimada and Tsurudome 1971). Experiments with artificial baits have also demonstrated the effect of bait types on species selectivity (e.g., Løkkeborg 1991; Yamaguchi et al. 1983). For the commercially important cod, several studies have aimed at identifying the primary stimuli that elicit the search for food. Bottom food search behavior in cod was used to determine the feeding stimulants present in shrimp (Ellingsen and Døving 1986). The amino acid glycine was the most potent single component, followed by alanine. There was a synergistic effect among glycine, alanine, proline, and arginine, and a mixture of these substances was more efficient than the total amino acid mixture in the shrimp extract. Two unidentified substances were also found to be highly potent. Laboratory investigations were carried out to identify the feeding stimulants in squid for juvenile cod (Johnstone and Mackie 1990). The study confirmed that amino acids are the major feeding stimulants for cod. A mixture of non–amino acid components was inactive on its own, but there was a synergistic interaction between these components and the amino acid mixture. These results were in general agreement with those obtained in a later

Fish Behavior in Relation to Longlines study based on choice experiments to distinguish preference between feeding stimulants (Franco et al. 1991). Studies aimed at identifying the chemical nature of feeding attractants thus indicate that there is great potential for using baits or extract mixtures to attract specific species to an odor source, to develop species-selective fishing methods. Moreover, as mentioned in Section 4.1, the efficiency of baited gears may be substantially improved by developing long-lasting baits that release feeding attractants at high concentrations over a long period of time. 5.4.4 Chemosensory Thresholds Several laboratory experiments have attempted to determine detection and response thresholds in fish. Electrophysiological techniques have been employed to obtain information on detection thresholds in many freshwater and marine fish species (e.g., Belghaug and Døving 1977; Goh et al. 1979; Nikonov et al. 1990; Sutterlin and Sutterlin 1971), and behavioral methods have been used to study response thresholds (e.g., Løkkeborg et al. 1995; Pawson 1977). Olfaction is generally the more sensitive distant chemoreceptor in fish (Ishida and Kobayashi 1992), but for some species, in particular the genus Ictalurus, the gustatory receptors may be a more acute chemical sense for certain substances (Bardach et al. 1967; Caprio 1978). Johnstone (1980) presented a summary of some comparable detection threshold determinations for freshwater and marine species. Electrophysiological studies of salmon (Salmo salar) demonstrated that the most effective olfactory stimulus tested, alanine, gave a detection threshold of 3.2 × 10−9 M (Sutterlin and Sutterlin 1971). Belghaug and Døving (1977) determined the electrophysiological thresholds to amino acids in char (Salvelinus alpinus) and found that arginine had the highest stimulatory capacity, with a threshold of 2.5 × 10−8 M. Similar studies on coho salmon (Oncorhynchus kisutch), rainbow trout, and white catfish (Ictalurus catus) demonstrated detection thresholds of 10−8 to 10−6 M to several individual amino acids (Hara 1972, 1973; Suzuki and Tucker 1971). Electrophysiological studies of chemosensory capacity have also been carried out on marine fish.

113

The detection thresholds in Atlantic stingray (Dasyatis sabina) and sea catfish (Arius felis) to alanine and cysteine, respectively, were determined at between 10−8 and 10−6.5 M (Silver et al. 1976). Among a range of amino acids, the electrophysiological thresholds to glutamine were estimated at 10−7 M in red sea bream (Chrysophyrys sp.) and below 10−8 to 10−9 M in conger eel (Conger sp.) (Goh et al. 1979). Nikonov et al. (1990) estimated extremely low threshold values of 10−12 to 10−14 M to the most stimulatory amino acids in Black Sea skate (Raja clavata). The detection thresholds in the algivorous rabbitfish (Siganus fuscescens) to 19 amino acids ranged from 10−10 to 10−5 M, with the lowest detection threshold for alanine (Ishida and Kobayashi 1992). A classic conditioning method was used to determine the sensitivity of cod to several amino acids (Johnstone 1980). The amino acids with lowest thresholds for detection, in order of effectiveness, were tyrosine, cysteine, phenylalanine, and glycine, with mean threshold levels ranging from 2.5 × 10−8 M to 6.7 × 10−8 M. This study also investigated the effect of raising the background level of glycine on the response threshold level for glycine. These results indicated that to detect a specific amino acid against a background level of the same substance, the difference in level for detection needs to be proportionally greater with higher background concentrations. These electrophysiological and conditioning studies indicate the threshold concentrations at which the animal is capable of detecting the olfactory stimulus tested. However, attracting fish to an odor source relies on the response of the whole animal (the response threshold), and the threshold at which an animal detects a food extract differs from the threshold for food-searching behavior (Pearson and Olla 1977; Zimmer-Faust and Case 1983). In red hake (Urophycis chuss), the threshold for detection of clam extract was shown to be an order of magnitude lower than the threshold for food searching (Pearson et al. 1980). Furthermore, response thresholds are affected by feeding motivation, and increasing food deprivation and higher temperature have been shown to increase responsiveness to bait odor in sablefish (Løkkeborg et al. 1995; Stoner 2004; Stoner and Sturm 2004). Studies

114

Fish Behavior near Fishing Gears during Capture Processes

on response thresholds are therefore more relevant than those on detection thresholds for various aspects of bait attraction. Unfortunately, few behavioral studies have been carried out to determine response thresholds to food odors. Behavioral response thresholds to glycine were determined to be below 10−7 M in cod and whiting (Merlangius merlangus) (Pawson 1977). Herring (Clupea harengus) larvae were shown to respond to glutamic acid at a threshold concentration of 1.5 × 10−6 M (Dempsey 1978). Response thresholds to the mixture of amino acids and other compounds released from squid bait have been determined in sablefish kept on three different feeding regimes (Løkkeborg et al. 1995). The response thresholds to total dissolved free amino acids (DFAAs) were found to range from 4.4 × 10−8 M in fish fed to satiation to 1.4 × 10−11 M in fish tested after 4 days of food deprivation—the response threshold fell by a factor of 3000 due to food deprivation. These observations suggest that fish deprived of food are capable of responding to baited gears from considerably longer distances than are satiated fish (see Behavior before Stimulation in Section 5.4.1). 5.5 INTERACTIONS BETWEEN THE FISH AND THE LONGLINE GEAR 5.5.1 The Capture Process—A Multitude of Stimuli and Responses The behavior that leads to hooking has traditionally been described as a simple sequence of events, in the course of which a fish moves into the active space, swims against the current in the odor plume toward the longline, and eventually attacks a baited hook and becomes hooked. However, we now know more about the activity and search patterns of fish before odor stimulation and how these patterns influence the probability of entering the active space, the mechanisms of location of an odor source, and the behavior involved after contact with the gear. This new knowledge suggests that the interactions between the fish and the different kinds of stimulus emitted from a baited longline are fairly complex, with the outcome influenced by a great number of physical and biological variables. Some of these variables can be tuned to optimize the gear.

The increase in the number of species studied permits species comparisons, which enhance understanding of the behavior of individual target species. Longlining is primarily based on chemical stimulation involving both the olfactory and gustatory senses, but other types of stimulus are also involved in the catch process. Visual stimuli provided by the line and baits as well as by hooked and unhooked fish play an important role. Movements of baits and struggling fish may also be sensed by the lateral line organ. The relative importance of the sensory modalities differs among species. For instance, on the basis of prey preferences, the relative size of the sensory organs, brain anatomy, and the diel activity rhythm and time of day of responding to baits, visual stimuli seem to be more important in ling than in cod (Fernö et al. 2006; Kotrschal et al. 1998; Løkkeborg et al. 2000; Løkkeborg and Fernö 1999). The influence of different stimuli may also vary within a species. Presoaked baits with reduced release rate of attractants resulted in poorer catch rates for bottom-set longlines but not for pelagic longlines that targeted cod migrating to the spawning grounds (Løkkeborg and Johannessen 1992). In the latter situation, the fish seem to react more to visual than to chemical stimulation. Fish are not always exposed to different stimuli from the gear in a strictly sequential way but may simultaneously encounter olfactory and gustatory stimuli as well as chemical and visual stimuli. The fish must then integrate stimuli from different sensory channels. Different stimuli presumably have an additive effect, but in some cases the stimuli may compete, with the fish trading off their effects. In addition, the order in which an animal encounters different stimuli from the gear may determine the outcome. King crabs approaching a baited pot upstream and then presumably attracted by chemical stimuli were trapped in the odor plume, butting against the net of the pot when they encountered a side without an entrance (Stiansen 2004). In contrast, king crabs approaching the pot down- or across-current and presumably attracted by visual or auditory stimuli showed more flexible behavior. These crabs searched around, and they more often located the entrance and entered the pot. Similar observations of ling and cod have showed that fish that encountered pots upstream when the current

Fish Behavior in Relation to Longlines was perpendicular to the entrance stayed within the odor plume and did not find the entrance area (unpublished data). The response of fish to a longline may also be influenced by which stimulus they initially encounter. The sequence of events leading to hooking can be divided into the following phases: Behavior before Stimulation The spatial and temporary dynamics of fish movements influence the efficiency of longlines. Actively swimming fish will have a higher probability of encountering the bait odor plume than will resting fish, and high swimming speed results in more encounters than slow swimming. Swimming capability varies among species and sizes of fish and is affected by ambient temperature. The path of fish that are searching for food also plays a role. In an individual-based behavioral model, moving at an angle to the current resulted in a higher probability of contact with an odor plume than moving straight into the current or a “random walk” strategy (Vabø et al. 2004). Furthermore, this study showed that cod tracked in a fjord often swim in a zigzag pattern against the current. Diel activity rhythms also influence the probability of encountering the bait odor plume (see Section 5.3). Arousal to the Presence of Bait and Stimulus Categorization Arousal to the presence of food when the concentration of attractants exceeds the thresholds of the chemosensory organs is a necessary first step in the hooking process of baited longlines. However, a fish stimulated by the vicinity of a food item does not automatically respond by approaching the odor source. Defined in this way, arousal is determined by the physiological detection threshold and is not influenced by motivational state and other modifiers. To react further, the fish must classify the stimuli from a source to be of relevance to its survival or reproduction and react accordingly. To respond to a longline by approaching the odor source, the fish must categorize the perceived chemical stimuli as a potential prey. Locating the Bait Odor Source The distance from which an odor source from a baited gear attracts fish (i.e., the active space) can

115

be determined by models of odor dispersion and by tracking fish in the field. Using the dispersion model by Sainte-Marie and Hargrave (1987) and the response thresholds to bait odor, maximum dimensions of the active space within which sablefish would search for food were found to vary from 10 m to several kilometers, depending on state of food deprivation, rate of attractant release from the bait, and current velocity (Løkkeborg et al. 1995). Of these variables, food deprivation had the greatest influence on the size of the active space. After 4 days of food deprivation, the attraction distance for sablefish increased by a factor of 57 over that of fish fed to satiation. Under conditions of low prey availability and great demand for food, therefore, baited gears are likely to attract fish within much larger areas. Field studies have confirmed that fish are capable of locating food odor sources at long distances (Løkkeborg 1998; Løkkeborg and Fernö 1999). In these studies, cod tagged with acoustic transmitters were tracked in their natural environment, and their responses to baited longlines set at various distances were observed (Fig. 5.2). At as far as 700 m from the baits, fish would respond to the bait odor and locate the odor source. Interestingly, this distance is similar to the distances between parallel fleets of longlines as set in commercial fishing in Norway, which suggests that fishermen have managed to optimize an important parameter. Ling have been observed to react at shorter distances (about 100 m; Skajaa 1997). It is important to identify the mechanisms that fish use to locate baited gear to optimize the gear. Upstream movements in the direction of bait have been observed in several field studies (Fernö et al. 1986; Kaimmer 1999; Løkkeborg et al. 1989; Løkkeborg and Fernö 1999; Wilson and Smith 1984). We might initially believe that fish locate the baited gear by swimming toward higher concentrations of attractants using a gradient search, but orientation in a chemical gradient is associated with many problems. Furthermore, in a realistic fishing situation, a gradient search would not always bring the fish into contact with the baited gear. Due to the rapid initial decrease in release of feeding attractants from baits over time (Løkkeborg 1990a), a reverse gradient may well be created, with the

116

Fish Behavior near Fishing Gears during Capture Processes

Baited longline Current direction

100 m

highest concentrations at long distances from the gear (Sigler 2000). In a simulation study of bait location, countercurrent strategies performed much better than strategies based on gradient search (Vabø et al. 2004). Fish orienting by rheotaxis, however, may encounter problems at the edge of the odor plume, and in the simulations, fish that had lost contact with the plume were allowed to swim in progressively larger loops until they were on track again. In such situations, fish might to a certain extent use gradients by comparing different points in space, although the large spatial and tem-

Figure 5.2. Tracks of three cod showing chemically mediated responses to baited longline.

poral variations in concentrations would make this difficult. Hence, although we cannot exclude the possibility that fish use gradients at the edge of the odor plume (Webster and Weissburg 2001), and close to the odor source (Atema 1980), it is safe to assume that rheotaxis is the main mechanism behind upstream swimming and location of a baited longline. This has important practical consequences. If gradient search is the main mechanism, much would be gained by adjusting the release rate of attractants from the baits in such as way as to create

Fish Behavior in Relation to Longlines as strong gradients as possible. If rheotaxis is the most important mechanism, we ought to select or develop baits that initially release attractants at high concentrations to attract fish from long distances and then maintain a release rate such that the fish will encounter above-threshold concentrations all the way to the bait. Sablefish seem to have a great ability to locate baits on a longline even when only a few baits remain (Sigler 1993), and cod can search along a line and localize baits (Løkkeborg and Fernö 1999). However, it may not always be straightforward for a fish to localize an olfactory source. One cod attracted from a large distance was observed to make stops under way, apparently losing the track and searching around (Løkkeborg 1998). The fish may then have entered an area with low concentrations of attractants generated by turbulence. In such situations, the fish may perform search patterns that increase the probability of reencountering the plume (Vabø et al. 2004). Visual Object Categorization, Bait Ingestion, and Hooking Behavior Fish in close contact with a baited hook display a number of behavioral patterns of different intensities. Based on the visual, taste, and mechanical stimuli it perceives, the fish may categorize the baited hook as either edible or inedible and respond accordingly. The behavior patterns have been defined and their occurrences recorded both in the laboratory and in the field (Fernö et al. 1986; Fernö and Huse 1983; Kaimmer 1999; Løkkeborg et al. 1989). Laboratory studies are appropriate for the initial stages of describing and defining different behavior patterns and investigating how the response of individual fish develops over time, but field studies are required to obtain realistic estimates of the frequency and intensity of the reactions (Løkkeborg et al. 1993). A fish makes an Approach toward the bait when it comes into close proximity to the bait but terminates the response without physical contact. During a Taste, the fish is in contact with the bait with the mouth or barbel. Incomplete bite, in contrast to Complete bite, is when the fish only takes a part of the bait into the mouth or does not close its mouth. Jerk is a rapid sideways movement with the head, and a Jerk series is several jerks

117

in rapid succession. Pulling the snood and Chewing the bait are made with the bait in the mouth. Finally, Rush is a rapid burst of swimming with the bait in the mouth, when the fish may either become hooked or escape and swim away. Some of these responses can reflect reactions toward natural prey—for example, a Jerk series could represent the behavior when a fish tries to shake a mussel out of its shell (Brawn 1969). Other responses, such as Rush, may be an escape reaction. Responses to hooks can be described by the frequencies of occurrence and the sequence of the different patterns of behavior. A flow diagram (see Huse and Fernö 1990; Kaimmer 1999) gives an overview of the behavior. A sequence analysis reveals more detail about which behavior patterns tend to precede and follow each other (Kaimmer 1999). In cod, transitions between complete bite and pulling, chewing, and rush are overrepresented (Fig. 5.3). Rushes lead in turn to hooking or the bait being pulled out of the mouth. Incomplete bites, on the other hand, seldom lead to rush or hooking. By identifying the behavior associated with hooking and relating the probability of hooking to this behavior, we can obtain an idea of the efficiency of a particular combination of bait and hook. Rush is strongly associated with hooking, although rushes in some instances can be the result and not the cause of hooking. The probability of hooking can thus be estimated as the number of hooked fish divided by the number of rushes or, alternatively, as the number of hookings relative to the number of all strong responses. The probability based on rushes ranges from 0.08 to 0.52 using different hook-and-bait combinations in laboratory and field studies of various species (Fernö et al. 1986; Huse and Fernö 1990; Kaimmer 1999; Løkkeborg et al. 1989). Many hook types thus appear to be relatively inefficient, and observations made with a highfrequency imaging sonar showed that a very low percentage of sablefish and Pacific halibut (Hippoglossus stenolepis) attracted to longlines and pots were captured (Rose et al. 2005). In one study, only 1.2% of cod observed on video were captured (He 1996). Because a rushing fish should generate sufficient power for the hook to penetrate the mouth when the movement is suddenly stopped by the resistance of the snood, the low hooking probability

118

Fish Behavior near Fishing Gears during Capture Processes

Figure 5.3. Sequence analysis of transitions between different responses of cod towards baited hooks based on a laboratory study by Fernö and Huse (1983). Combinations of behavior patterns that occur more or less frequently than expected by chance are indicated (chisquare test, P < .05). Transitions where the observed and expected numbers of transitions are not significantly different are not marked. Free swimming means that the fish leaves the near field of the baited hook.

indicates that the point of the hook often does not come into contact with the mouth. When it visually detects the longline in the nearfield, the fish may categorize the gear as a prey (resulting in a further approach), a predator (resulting in escape), or an object of no relevance (no response). Several properties of the bait and hook influence the behavior and thereby the outcome of the interaction. Cod and haddock approaching a large bait more often turned away before physical contact than did fish approaching smaller baits (Johannessen et al. 1993), and large baits were also less effective in comparative fishing experiments (Johannessen 1983). Furthermore, large baits have been shown to catch fewer small fish than smaller baits; there is a size-selective effect of bait size (Johannessen 1983; Løkkeborg 1990b; McCracken 1963). In a conflict situation between accepting and

rejecting a baited hook as a potential prey (Fernö and Huse 1983), fish seem to more often reject large baits. The shape of the bait may also affect responsiveness to a baited hook. Lower catch rates of small cod for rectangular shaped artificial baits than when natural shrimp bait was used were explained by a restrained response toward a novel prey item (Løkkeborg 1990b). The hooking probability of a circle hook has been shown to be higher than a J-shaped hook, presumably due to a higher probability of penetrating the mouth during a rush (Huse and Fernö 1990). Circle hooks were also more effective in catching cod in comparative fishing experiments. Similar studies that compared the circle hook and the J-hook also showed the superiority of the circle hook for catching Pacific halibut and hake (Peeling and Rodgers 1985; Quinn et al. 1985) but not for catching sword-

Fish Behavior in Relation to Longlines fish (Xiphias gladius) (Watson et al. 2005). Double and treble hooks, which increase the probability of contact, have been shown to be more effective than single hooks (Fernö et al. 1986). Hooking probability is species specific. For instance, attacks on baited hooks in cod are more often based on a complete bite and are of higher intensity than in haddock, and cod are caught more than twice as often after a rush (Løkkeborg et al. 1989). The different response intensities can be explained by differences in natural foraging behavior, with haddock feeding more on small stationary benthic organisms and cod more on mobile prey, which demand higher-intensity responses. The intensive attacks of cod on prey can therefore be expected to be preceded by more careful and selective behavior, and more cod than haddock in fact terminated their responses to baited hooks without physical contact (Løkkeborg et al. 1989). Haddock more often take only a part of the bait in their mouth and can then pull the bait from the hook, and this species has a reputation among fishermen as a bait stealer. There are generally large individual differences in fish behavior (Magurran et al. 1993), and this should also be expected during interactions with fishing gear (Fernö 1993). In the laboratory, some cod made only a few strong responses toward a baited hook, whereas other individuals made several hundred low-intensity responses (Fernö and Huse 1983). A similar difference between individual fish was also seen in haddock observed in their natural environment (Løkkeborg et al. 1989). The frequency of hooking recorded during behavioral observations cannot be used to directly estimate the efficiency of a given combination of bait and hook. One reason is that a test line differs from an actual longline in many respects, such as in the tension that determines the resistance the fish experience when pulling the snood. Furthermore, fish cannot be assumed to always make only one attack on a baited hook. With a greater number of attempts, the difference between an efficient and a less-efficient hook will decrease and, in theory, both hooks will eventually capture the fish after repeated attacks. With only one attack of each fish, two hook types with hooking probabilities of 0.4 (Hook A) and 0.2 (Hook B) will result in twice as many

119

hooked fish on Hook A, resulting in a 100% difference in catch rate between longlines with Hooks A and B. If the fish make five attempts, the difference will decrease to 37% (1 − 0.65 divided by 1 − 0.85). This may be the reason that the difference between two hooks tends to be smaller in comparative fishing experiments than is suggested by the difference in hooking frequency observed in the laboratory (Huse and Fernö 1990). Nevertheless, behavioral studies of hooking frequency show which hook type is the most efficient. 5.5.2 Internal Factors Food deprivation and motivational state have been shown to affect food searching behavior and responsiveness to prey in several fish species (Atema 1980; Hart 1993; Hart and Connellan 1984; Pearson et al. 1980). Accordingly, hunger has been shown to affect food searching behavior and response intensity to baits. Sablefish responded to lower bait odor concentrations when tested under conditions of lower food rations and longer duration of food deprivation (Løkkeborg et al. 1995). The intensity of behavioral responses to bait odor (swimming speed and turning rate) and response duration were also shown to increase in sablefish with increasing food deprivation (Løkkeborg et al. 1995; Stoner and Sturm 2004). These behavioral changes will affect the distance from which sablefish will start a food search, its search pattern, the location time, and the time spent searching for the odor source. Similarly, Pacific halibut were shown to locate more bait with increasing food deprivation, and fish deprived of food located the baits more rapidly (Stoner 2003). These observations indicate that when food demand is higher, fish intensify their search for food and therefore increase the probability of locating a food odor source, whereas satiated fish will probably not display active food searching behavior. Feeding motivation has also been shown to influence hooking behavior. Whiting and cod with low liver weight, an indication of starvation and increased food demand, tended to swallow the hook and be caught in the stomach, whereas fish with higher liver weight were usually caught in the mouth (Fernö et al. 1986; Johannessen 1983). Seasonal changes in the feeding motivation relative to other motivational systems (e.g., reproduction)

120

Fish Behavior near Fishing Gears during Capture Processes

can also lead to variations in the response. The probability of a cod being hooked was higher in September/October than in December (Løkkeborg et al. 1989), and similarly, the hooking probability of whiting was higher in October than during May– July (the spawning season) (Fernö et al. 1986). Prey abundance affects the level of hunger, and evidence exists for the effects of prey density on responsiveness to baited hooks. Longline fishermen have experienced extremely low catches of cod in the Barents Sea in early spring when the cod are preying on dense schools of capelin (Engås and Løkkeborg 1994). Fishermen from the Faeroe Islands have also observed lower catch rates in years with high prey density in the sea (Petur Steingrund, Faroese Fisheries Laboratory, personal communication). Similarly, longline catch rates of tunas in the tropical Pacific were low in areas where there were large, high-density prey patches (Bertrand et al. 2002). 5.5.3 The External Environment (Light, Temperature, Current) Behavioral observations have shown that cod and other species exhibit diel rhythms in swimming and feeding activities (e.g., Løkkeborg et al. 1989; Løkkeborg and Fernö 1999; Thorpe 1978). The

swimming speed of acoustically tagged adult cod in the autumn increased at dawn, remained high during the day, then gradually decreased during the evening, remaining relatively low throughout the night (Løkkeborg and Fernö 1999). The recorded swimming activity of cod in this study appeared to be related to light level and was regarded as reflecting food search behavior. In a similar study in late spring, with higher light levels at night, a similar but less pronounced rhythm was observed (Løkkeborg 1998) (Fig. 5.4). The locomotory activity of Pacific halibut in a laboratory study was three times as high at high than at lower light levels (Stoner 2003). These observations suggest that activity decreases below a critical light level. This is supported by a study of responses of cod to baited hooks, in which the fish displayed higher activity during the day in both September and December, but the increase in activity in the morning was observed later in December in connection with the changing time of sunrise (Løkkeborg et al. 1989). Similar seasonal changes in diel rhythms of responses to bait have been demonstrated in whiting (Fernö et al. 1986). These findings indicate that the success of fishing with baited gears is influenced by the time of day when the gear is set. More cod were thus observed

Figure 5.4. Diel rhythm in the mean swimming speed of cod in May (open circles) and September (closed circles) observed in a fjord in northern Norway. Arrows indicate sunrise and sunset in September, whereas May has midnight sun at this latitude.

Fish Behavior in Relation to Longlines to locate baits during periods of high (daytime) than low (nighttime) activity, and the time that elapsed until the fish located baits was 50% shorter during the day (Løkkeborg and Fernö 1999). Fishing experiments conducted in commercial fishing showed that longlines set before dawn caught twice as much haddock as longlines set later in the day (Løkkeborg and Pina 1997). Thus, time of day is an important factor to consider when fishing with baited gears. Furthermore, there is little doubt that light level has a direct effect on locomotion in fish, independent of diel rhythms (Stoner 2004). The higher probability of locating baits during the day may also be related to increased visibility. In a laboratory study, Pacific halibut located a larger proportion of the baits offered as the light level increased, and the time taken to locate baits was shorter in the light conditions than in darkness (Stoner 2003). Several fish species have been shown to be relatively unsuccessful in locating and attacking baits in darkness, although the lower light thresholds for feeding were low (e.g., McMahon and Holanov 1995; Ryer and Olla 1999; Stoner 2003). Interestingly, night setting has been shown to be an effective mitigation measure to reduce incidental bycatches of seabirds in longlining, and more seabirds were caught on night-set hooks set in bright moonlight conditions than when there was no moonlight (see Section 5.2). Temperature is another external factor that can influence how fish interact with longlines. Up to a certain limit, increasing temperature increases the scope of activity and swimming activity (Castonguay and Cyr 1998; He 2003; Stoner and Sturm 2004), leading to more encounters with the odor plume and the gear. In addition, temperature will affect the metabolic rate and gastric evacuation (Fry 1971). In most cases, food consumption will increase in line with temperature, leading to stronger responses to baited gear and a higher probability of being hooked. Sablefish at a low temperature swam slowly and attacked and consumed fewer baits (Stoner and Sturm 2004). For a thorough review of the effect of temperature, see Stoner (2004). Current is another external factor that affects food-searching behavior both through its effect on bait odor dispersion and on fish activity. Currents will increase the active space and permit rheotactic

121

responses to the odor source, and more whiting were attracted to bait in the presence of a current (Fernö et al. 1986). However, responses in fish to bait have been shown to decrease when current velocity is high. When current speeds were less than 18 cm s−1, the number of cod and haddock in the vicinity of baits were two or three times as high as in periods with stronger currents (Løkkeborg et al. 1989). However, variations in current velocity below 18 cm s−1 did not appear to influence activity. This could be explained in terms of energy optimization. As food-searching fish swim predominantly upstream to the odor source, it would be energetically advantageous to be active during periods of moderate or low current velocity and to remain in shelter when the current is strong (see Weihs 1987). The current can also influence the probability of contact with a baited gear by influencing the vertical distribution of fish (Michalsen et al. 1996). It is clear that the motivational and physiological state of the target species as well as its sensory capabilities have a major impact on food searching behavior and the likelihood of locating baits. Environmental conditions such as light, temperature, and current also affect locomotion, the probability of bait location, and responsiveness. These internal factors and environmental variables thus have obvious implications for the outcome of fishing operations using baited gears. 5.5.4 Intraspecific and Interspecific Interactions Food-searching behavior and responses to baits are likely to be affected by fish size because larger fish have greater swimming capability and are at less risk of predation than are smaller fish. According to optimal foraging theory, larger fish use a larger foraging area due to their higher optimal swimming speeds (Hart 1993), and constraints on feeding activity are less pronounced due to the lower risk of predation (Milinski 1993), resulting in exploitative competition. Interference competition among fish attracted to baits is also related to size. Both intraspecific (cod, whiting, ling, tusk, halibut) and interspecific (codhaddock, wolffish-cod/haddock) competition for baits has been observed, with the largest individuals emerging as the most successful competitors (Allen

122

Fish Behavior near Fishing Gears during Capture Processes

1963; Bertrand 1988; Fernö et al. 1986; Godø et al. 1997; Løkkeborg and Bjordal 1992; Stoner and Ottmar 2004). Rodgveller et al. (2008) found that longline catch rates of sablefish were negatively correlated with giant grenadier (Albatrossia pectoralis) and rougheye rockfish (Sebastes aleutianus) catch rates, indicating competition for baited hooks in the area studied (Alaskan waters). No negative correlations were found for trawl catches taken in the same area, indicating that the negative correlations for longlines were not due to differing habitat preferences among these species. The presence of other fish in the vicinity of a longline does not always have a negative effect on catchability but may in fact stimulate responses in fish. Groups of fish are less at risk to predators than are single fish, with the balance between feeding and antipredator behavior thus shifted toward the former. Social facilitation and copying the behavior of other fish also stimulate responses (Ryer and Olla 1992), as does competition for food. A laboratory study by Stoner and Ottmar (2004) found that several halibut located and consumed baits more quickly than did single fish. In the field, we observed a haddock that swam slowly back and forth for a long time, nibbling on the baits on the test line (Løkkeborg et al. 1989). Eventually the haddock was hooked and fought to get free, and within a few seconds 10 to 15 large cod entered the field of observation. One cod vigorously attacked and tried to swallow the hooked haddock, while other cod attacked the neighboring hooks and became hooked. Similarly, more whiting became hooked on a test line when there were hooked fish present (Fernö et al. 1986). However, fright responses released by hooked fish have never been observed, and fish of the species studied do not seem to react negatively to trapped fish—a situation that presumably does not take place under natural situations. The stimulatory effect of competitors (social facilitation), hooked fish, and the movements of neighboring baits helps to explain the observation that fish are often found clustered along a longline (Johannessen 1983; Sigler 2000). 5.5.5 Learning The literature on learning in fish has expanded enormously during the past decade, and it is now well

documented that fish have highly developed learning skills and cognitive abilities (Brown et al. 2006). Although the ease with which fish can form an association between a stimulus and a reward or punishment is limited by conceptual learning schemes, learning processes should be expected to modify how the longline is categorized and the sequence of responses toward the gear. The behavior of any given species toward baited hooks does not seem to be fixed, indicating flexibility and learning. In the field, fish are often observed to swim away from the gear after few responses, supporting the idea that they modify their behavior over time. Before a fish encounters the gear, its feeding behavior and its responses to gear-related stimuli can be molded by its recent experience with prey. How fish search for and catch prey may influence both search patterns and which sensory channel fish primarily use. Visual stimuli thus seem to be more important than chemical stimuli for cod feeding on capelin (see Section 5.5.2). This can be used in longline fishing by tuning the type of bait to local natural prey items. Along the same lines, the lower efficiency of artificial baits for small cod than for large cod may be explained by the fact that small cod have less experience of a range of prey organisms, making them more reluctant to attack a novel prey (Løkkeborg 1990b). A clear example of learning can be found in observations of cod that were tagged in situ by allowing them to ingest acoustic transmitters wrapped in mackerel bait (personal observations). A fish that had ingested one tag followed the research vessel to the next tagging station and took another transmitter shortly after the tagging rig was put in place on the bottom. Eventually the fish had taken four valuable tags (US $280 each), intended for tracking other individuals. However, the perturbed researchers hit back, and at its fifth attempt the saboteur was surprised by the baited and wrapped transmitter being replaced by a baited hook. The end of the story is shown in Figure 5.5. This fish responded fast and before the odor plume had dispersed more than a few meters. It was presumably attracted by auditory stimuli from the vessel and learned to associate this sound with the presence of food.

Fish Behavior in Relation to Longlines

123

Figure 5.5. Learning/ conditioning: this cod was able to locate and ingest four acoustic tags wrapped in mackerel bait by associating the sound of the research vessel with the presence of food. For color detail, please see color plate section.

Physical contact with a baited hook has been shown to modify the response. Cod in the laboratory typically terminated their response sequence after a single strong response toward a hook and did not make a new attempt for some time (Fernö and Huse 1983). Over time, the fish apparently came into a conflict situation between feeding and avoiding pain, and some individuals made long series of approaches and retreats without touching the bait. When this article was first submitted for publication, one of the referees questioned whether fish can experience pain, but with our current understanding of fish learning and cognition, it is now generally accepted that fish perceive pain and modify their behavior accordingly (Chandroo et al. 2004). In contrast to cod, haddock typically take small pieces of the bait during incomplete bites without contact with the hook and are thereby rewarded, stimulating further responses. Acoustically tagged cod that encountered a bait bag that was too large to ingest also seem to experience a conflict and were often observed to leave the baited gillnet (Kallayil et al. 2003). After some time, they returned from distances of several hundred meters. This indicates that the output from the reward-evaluating mechanisms in the brain assessing an experience shows a temporal dynamics, with the effect of a negative experience fading away and the tendency to respond to food again

becoming dominant. In some cases, however, the modification can persist for a long period. Carp (Cyprinus carpio) with experience of hooks were still difficult to catch after 1 year (Beukema 1970). If what a fish experiences after terminating a strong response toward a hook is capable of modifying its behavior, there may also be an effect of experience as soon as the fish takes the baited hook into its mouth and manipulates and chews on it. In that case, the efficiency of a hook can be partly dependent on how much it reduces the strength of further responses. Circle hooks are more efficient than J-hooks (see Section 5.5.1), and this is believed to be primarily due to the higher probability of curved hooks penetrating the mouth during a rush. However, it is also possible that the point of a circle hook that is bending inward will come less often into contact with the mouth when a fish is chewing on the bait, and the escalation of the response is therefore not inhibited. 5.6 CONSERVATION CHALLENGES AND POTENTIAL SOLUTIONS 5.6.1 The Main Problems Concerns about the numbers of seabirds that are incidentally killed in various types of fisheries are growing. Most attention has been given to bycatch of seabirds in longlining, especially albatrosses

124

Fish Behavior near Fishing Gears during Capture Processes

taken in the Southern Ocean longline fisheries (Brothers 1991; Cherel et al. 1996; Weimerskirch et al. 1997). Many southern albatross populations are in decline, and longline-induced mortality is believed to be an important factor contributing to this situation (Croxall et al. 1990; Moloney et al. 1994; Poncet et al. 2006; Prince et al. 1994; Weimerskirch and Jouventin 1987). The second conservation issue related to longline fisheries concerns interactions with sea turtles. There are seven species of sea turtles living in the world’s oceans, of which six are listed as endangered (http://www.redlist.org/). The failure of sea turtle populations to recover is attributed in part to incidental capture by fishing gears, and although most concerns have been raised about bycatch of sea turtles in trawl fisheries (Magnuson et al. 1990), pelagic longlines have been implicated as a major source of anthropogenic mortality for loggerhead (Caretta caretta) and leatherback (Dermochelys coricea) sea turtles (Lewison et al. 2004). The final conservation-oriented issue covered here is bycatch of nontarget fish. This problem relates both to discards of specimens below minimum landing size (i.e., juveniles of the target species) and to nontarget species. Bycatch of juveniles can be reduced by improving longline size selectivity, and factors affecting size selection are discussed in Sections 5.1, 5.4, and 5.5 and have been reviewed by Løkkeborg and Bjordal (1992). The effect of fishing on shark stocks has become the focus of considerable international concern (Megalofonou et al. 2005), and the FAO has developed an International Plan of Action for the Conservation and Management of Sharks. Thus, with regard to incidental catch of nontarget fish, we focus on efforts to monitor and reduce shark bycatch in pelagic longlining. It is often difficult to attribute population declines to a specific factor, as the marine environment is subject to much natural variation and thus provides a noisy background for observing changes that can be directly attributed to fishing activities (Gislason 1994). Some catch statistics exist for sharks, but accurate information on the numbers of seabirds and sea turtles killed is difficult to obtain (Lewison et al. 2004; Wienecke and Robertson 2002). Estimates of annual fishing-induced mortality of

these species are poor because captures are rare and observations are few (Pradhan and Leung 2006). Regardless of the actual number of nontarget individuals caught in a fishery and the consequent population-level effects, it is not consistent with the principles of ecologically sustainable management for fisheries to take large numbers of nontarget species (Løkkeborg and Robertson 2002). Furthermore, incidental bycatches of nontarget species are a twofold problem, especially with regard to seabirds. Most seabirds that attack baited hooks manage to seize the bait without becoming hooked, and it is likely that sea turtles and sharks also scavenge baits from longline hooks. Such interactions reduce gear efficiency and profitability due to the associated loss of baits. It should therefore be in the interest of fishermen to reduced interactions with individuals of nontarget species, as reducing bait losses is likely to result in higher target catch rates. A 32% increase in target catch rates was obtained for longlines that were set using a bird-scaring line, compared with longlines set without any mitigation measure (Løkkeborg 2001). 5.6.2 Mitigation Measures Aimed at Reducing Incidental Catches of Seabirds When longlines are set, baited hooks are available to foraging seabirds because they float on the surface for a short while before they start sinking. Seabirds are killed during the line-setting operation when they seize baited hooks, become hooked in the bill or body, and are drawn underwater by the sinking longline. Some seabird species are also capable of diving to depths of several meters and may thus attack baited hooks during the first part of the sinking phase. Birds occasionally become hooked during line hauling, but with careful handling they can be released alive. Most efforts should thus be put into developing measures that will prevent seabirds from seizing baited hooks during the setting operation. Mitigation measures should not only be efficient in minimizing bird capture but also practical and easy to implement in commercial fishing, enforceable, cause no loss of target catch, and offer fishermen incentives to use them (Gilman et al. 2003, 2005). Several mitigation measures capable of reducing the likelihood of seabird bycatch have

Fish Behavior in Relation to Longlines been described (Brothers et al. 1999a; Bull 2007), and they can be divided into four main categories: 1. Avoid peak areas and periods of bird foraging (night setting, area and seasonal closures). 2. Hinder access to baited hooks (underwater setting funnel, weighted lines, thawed bait, line shooter, bait-casting machines, side-setting). 3. Deter birds from taking baited hooks (streamer [bird-scaring] lines, acoustic deterrents, water cannon). 4. Reduce the attractiveness or visibility of the baited hooks (dumping of offal, artificial baits, blue-dyed bait). The most promising and widely tested mitigation measure in demersal longlining is the streamer line (bird-scaring line, tori line, Fig. 5.6a), which is a line with streamers that is towed behind the vessel and deters seabirds from attacking baited hooks while longlines are set. This device has been shown to virtually eliminate seabird bycatch in the longline fisheries in Alaska and in the northeast Atlantic where interactions with the northern fulmar (Fulmarus glacialis) are most common. Experimental studies carried out in the former fishery showed that paired streamer lines reduced seabird bycatch by 88% to 100% compared with control lines with no mitigation measures installed (Melvin

125

et al. 2001). Similar studies conducted in the northeast Atlantic demonstrated that a single streamer line reduced the bycatch by 98% to 100% (Løkkeborg 2003). Mitigation measures such as setting funnel and weighted lines were also shown to significantly reduce seabird bycatch in these fisheries but were not as efficient as the streamer line. Streamer lines are likely to be less efficient in reducing bycatch of diving seabirds as birds may still reach baited hooks beyond the aerial portion of streamer lines. This deficit may be solved or at least significantly reduced by using weighted longlines in combination with streamer lines. In the cod fishery in the Bering Sea, paired streamer lines in combination with integrated weight lines were shown to reduce bycatch of short-tailed shearwaters by 97% compared with control lines with no mitigation measure (Dietrich et al., 2008). Setting longlines at night is the most widely tested mitigation measure in the Patagonian toothfish (Dissosticus eleginoides) fishery in the Southern Hemisphere (Fig. 5.6b). In an experiment conducted in the South Indian Ocean, mortality rates were reduced by 62% when lines were set at night compared with when lines were set during daylight hours (Cherel et al. 1996). Furthermore, for the lines set at night, the mortality rate was 75% lower when the powerful deck lights were turned off. Observer data showed that night setting reduced

Figure 5.6. The bird-scaring (streamer) line is the most efficient mitigation measure tested in demersal fisheries in the northern hemisphere (a), and night-setting when there is no moonlight has proved to be efficient in fisheries in the southern hemisphere (b).

126

Fish Behavior near Fishing Gears during Capture Processes

mortality of white-chinned petrels (Procellaria aequinoctialis) by 81%, and only 1 of a total of 78 albatrosses were caught at night (Weimerskirch et al. 2000). Several other studies have also demonstrated the efficacy of this measure (Ashford et al. 1995; Nel et al. 2002; Reid et al. 2004; Ryan and Watkins 2002). Other efficacious mitigation measures for reducing seabird catches in demersal longlining in the Southern Oceans include dumping offal during line setting (98% and 54%, respectively, in Cherel et al. 1996 and Weimerskirch et al. 2000), weighted longlines (80% in Agnew et al. 2000; 61% to 99% in Robertson et al. 2006), streamer lines (Moreno et al. 1996), and setting funnel (Ryan and Watkins 2002). Night setting has been shown to be an efficient mitigation measure also in pelagic longlining with the probability of taking bycatch reduced by up to 85% (Brothers et al. 1999b; Klaer and Polacheck 1998; Murray et al. 1993). Seabirds were also more likely to be caught on night-set hooks set in bright moonlight condition than on those set when there was no moonlight. In Hawaii, the underwater setting chute was shown to eliminate captures of albatrosses in tuna longlining (Gilman et al. 2003), and in the swordfish fishery, branch lines with added weights and bait that was dyed blue (making it less apparent) reduced the number of contacts with albatrosses by about 90%, and the use of a streamer line reduced contact by about 70% (Boggs 2001). Bait thawing, use of a bait thrower, area fished, and season are other factors that significantly affect seabird mortality in pelagic longlining. There is no single solution of the problem of incidental seabird mortality in longline fisheries, as the efficiency of any given mitigation measure will be influenced by the seabird species assemblage at the particular fishing ground as well as the type of longline gear used (Løkkeborg 2008). Where the northern fulmar is the predominant seabird captured (i.e., northern Atlantic and Pacific Oceans), streamer lines have proved to be very effective in demersal fisheries. In pelagic longlining and in the Southern Hemisphere, where albatrosses and petrels are dominant, night setting is an effective measure, although it should be used in combination with other mitigation devices (e.g., streamer lines and weighted lines) in areas inhabited by nocturnal seabirds and

in bright moonlight conditions. Seabird mortality rates in all longline fisheries can therefore be significantly reduced by appropriate and effective mitigation measures. Considerable reductions in seabird bycatch rates have thus been obtained in many fisheries (Agnew et al. 2000; Gilman et al. 2003; Murray et al. 1993; Reid et al. 2004). For example, from 1993 to 2003, the bycatch rate was reduced from 0.66 to 0.0003 bird per 1000 hooks in the toothfish fishery around South Georgia as a result of the implementation of the CCAMLR conservation measure 25-02 (Reid et al. 2004). 5.6.2 Mitigation Measures Aimed at Reducing Incidental Catches of Sea Turtles Unlike seabirds, sea turtles are seldom caught during the setting operation when the baited hooks are floating at the surface. Sea turtles are able to dive to much greater depths than most seabirds and are caught mainly on pelagic longlines after the gear has sunk to the fishing depth. Thus, different mitigation measures have to be developed to reduce sea turtle bycatch. Due to their limited diving depth, sea turtles are primarily caught by shallow-set longlines and not by deep pelagic or bottom-set longlines (Polovina et al. 2003). Sea turtles caught on pelagic longlines set close to the surface can swim up to the surface to breathe. Nearly all sea turtles caught on shallowset longlines are therefore most likely alive at gear retrieval. Observations in the Spanish surface longline fishery in the western Mediterranean (1999– 2004) showed that less than 2% of the incidental catches of loggerhead turtles were dead (46 of 3480 turtles; Caminas et al. 2006), and all loggerhead turtles (188) caught in a study carried out in the Ionian Sea (Italy) were alive (Deflorio et al. 2005). Thus, it is important to identify and develop the best practices to handle and release captured turtles in such a way as to minimize injury. There is little empirical data on rates of posthooking mortality in sea turtles released from longlines. Findings strongly suggest low rates of postrelease mortality in lightly hooked Olive Ridley sea turtles (Lepidochelys olivacea) caught by shallow-set longlines (Swimmer et al. 2006). Sasso and Epperly (2007) used pop-up archival tags to estimate survival rates, and their results also suggested

Fish Behavior in Relation to Longlines that lightly hooked loggerhead turtles did not suffer any additional mortality relative to control turtles captured by dip net. Casale et al. (2008), however, observed high mortality rates for loggerhead turtles that had swallowed the hook. Thus, mortality rates are likely to differ greatly with regard to hooking position. Measures tested to reduce sea turtle mortality include hook types and bait types that are less efficient in catching turtles, setting at depths beyond the diving depth of turtles, and night setting to prevent turtles from seeing the baited hook. Gilman et al. (2005) have reviewed published studies on the subject as well as others that are planned or in progress. The use of circle hooks to reduce the mortality of sea turtles has been reviewed by Read (2007). Because most research has begun only recently and has not yet been peer-reviewed or published, our knowledge of the efficacy of the proposed mitigation measures is still limited. The most comprehensive work on development of mitigation measures to reduce sea turtle bycatch in longlining is a study of a U.S. pelagic swordfish fishery in the northwestern Atlantic (Watson et al. 2005). Traditionally, J-hooks baited with squid were used in this fishery, until it was closed in 2001 due to interactions with loggerhead and leatherback sea turtles. Watson et al. (2005) evaluated the effectiveness of circle hooks and mackerel bait with respect to reducing interactions with sea turtles and maintaining swordfish catch rates. Circle hooks with mackerel bait reduced bycatch of loggerhead turtle by 90%. Circle hooks with squid bait and J-hooks with mackerel bait significantly reduced loggerhead bycatch (by 86% and 71%, respectively). Mackerel bait, irrespective of hook type, reduced the bycatch of leatherback turtle by about 65%, and circle hooks with squid bait reduced leatherback catch by 57%. The catch rate of swordfish rose when mackerel was used as bait (63% with J-hooks and 30% with circle hooks), whereas the swordfish catch rate was reduced by about 30% using circle hooks with squid bait. In the tuna longline fishery in Hawaii, the use of saury as bait significantly reduced sea turtle bycatch compared with the use of squid as bait (Pradhan and Leung 2006). In conclusion, sea turtle interactions associated with the western Atlantic pelagic swordfish longline

127

fishery can be significantly reduced by using circle hooks or mackerel bait, and when used in combination, a reduction in bycatch of 90% for loggerheads and 65% for leatherbacks can be obtained with an increase in target catch rate. Implementation of this fishing technique made it possible to reopen both this fishery and the Hawaii-based pelagic longline fishery in 2004. Furthermore, few loggerheads caught on circle hooks swallowed the hooks (27%), whereas the majority of the loggerheads caught on J-hooks ingested the hooks (69%) (Watson et al. 2005). Also, Brazner and McMillan (2008) observed that lower proportions of circle hooks were swallowed by loggerheads compared with J-hooks. Swimmer et al. (2006) showed that lightly hooked Olive Ridley turtles (nine turtles) all survived their encounter with shallow-set longline gear using circle hooks. Thus, postrelease mortality is likely to be lower for individuals caught on circle hooks than for those caught on J-hooks. Leatherback turtles were usually hooked externally or entangled in the lines. Read (2007) concluded that circle hooks would significantly reduce sea turtle mortality. In an experiment studying sea turtle bycatch in the fishery targeting swordfish in the western Mediterranean Sea, 93% of the loggerhead specimens were caught on longlines during daytime soak, while swordfish captures were independent of retrieval time (Baez et al. 2007). Watson et al. (2005) found that loggerhead turtle catches increased with increasing soak time. As swordfish longlines were set at sunset and retrieved in the morning, increased soak time implies an increase in daylight soak time. Although the effect of daylight soak time was inconclusive, the authors suspected that there was a confounding effect between total soak time and daylight soak time. Also, in the swordfish-targeted longline fishery in Hawaii, incidental catches of loggerhead turtles increased with increasing soak time (Pradhan and Leung 2006). There was no effect of soak time on leatherback catch rates in these studies, indicating a difference in the time of day at which these two sea turtle species are most likely to interact with baited longlines. Night setting could thus be an efficient way of reducing loggerhead bycatch. Sea surface temperature also affects sea turtle bycatch (Watson et al. 2005). The loggerhead catch

128

Fish Behavior near Fishing Gears during Capture Processes

rate increased 200% to 350% with every 2.8°C increase in surface temperature, with a somewhat lower increase for leatherback. Fishing cooler water did not seem to affect swordfish catch rates when mackerel was used as bait. Analysis of observer and landings data from the Canadian pelagic longline fishery indicated that bycatch of loggerheads was concentrated above 22°C and catch of target species (tuna and swordfish) peaked between 17° and 18°C (Brazner and McMillan 2008). Dying the bait blue has been proposed and tested as a means of mitigating interactions with sea turtles. However, Swimmer et al. (2005) found no differences in rates of interactions with Olive Ridley turtle (8.4 and 8.1 individuals per 1000 hooks) when using untreated versus blue-dyed squid baits. On the basis of these results and other studies, the authors concluded that dying bait blue, which has been shown to reduce interactions with seabirds (see Section 5.2), is not effective in reducing sea turtle bycatch. Setting pelagic longlines deeper has been shown to decrease sea turtle bycatch (Brazner and McMillan 2008; Pradhan and Leung 2006). Analysis of observer data from the Canadian pelagic longline fishery showed that no loggerheads were captured when hooks were set at depths greater than 40 m (Brazner and McMillan 2008). Observer data collected by the Secretariat of the Pacific Community indicated that deep-set longlines reduced sea turtles catches by one order of magnitude compared with shallow-set gear, and turtles caught on deep-set gear were taken on the shallowest hooks (SPREP 2001). Similarly, observer data from Hawaii showed that loggerhead turtles were caught only by shallow longlines targeting swordfish and not by longlines set deep to target bigeye tuna (Polovina et al. 2003). In pelagic tuna longlining, the mainline takes the shape of a catenary curve, which gives a wide range in depth between the shallowest hooks close to the float lines and the deepest hooks in the middle of a basket (Suzuki et al. 1977). Shiode et al. (2005) developed a new method to set all branch lines at almost the same depth, to reduce sea turtle bycatch. The mainline was kept almost horizontal by using midwater floats to lift the section of the mainline between adjacent main floats. This setting method, with only 5-m difference in depth between the shal-

lowest and deepest hooks, may prove to be effective in reducing sea turtle bycatch when the mainline is set deeper and allows a more precise adjustment of the depth of the baited hooks to the swimming depth of the target fish (Shiode et al. 2005). Recently, Shiga et al. (2008) further developed the midwater float system by using two (double) midwater floats. An alternative approach to ensure that all hooks fished at depths greater than 100 m has been developed by Beverly et al. (2009), who suspended the fishing portion of the mainline on long, weighted float lines. 5.6.4 Mitigation Measures Aimed at Reducing Incidental Catches of Sharks The slow growth, late maturity, and low fecundity of sharks make them extremely vulnerable to even modest levels of fishing. Although many regions lack a pelagic fishery that specifically targets sharks (Yokota et al. 2006), other longline fisheries may be a great threat to shark stocks because species with higher production rates, such as swordfish and tuna, continue to support the fishery (Megalofonou et al. 2005). Most studies that deal with shark stock conservation have focused on monitoring and analyzing incidental catch and discards in pelagic fisheries (e.g., Francis et al. 2001; Marin et al. 1998; Megalofonou et al. 2005), whereas to date, few efforts have been made to develop and test measures to prevent shark capture. The most important gear parameters that determine species and size selection in longline fishing are bait type and hook type, of which bait type is the more important factor affecting species selectivity (Løkkeborg and Bjordal 1992). An artificial bait using squid liver (a waste product of the industry) was developed for tuna longlining and tested off the Hawaiian Islands (Januma et al. 2003). This bait significantly reduced shark bycatch, with catch rates that on average were 67% lower than with traditional squid bait. Although catches of tunas were also somewhat lower with the artificial bait, the difference in target catch rates between the two bait types was not significant. In the study carried out in the U.S. Atlantic swordfish fishery, mackerel bait, which proved efficient in reducing sea turtles bycatch (see Section 5.3), was also shown to reduce the catch of blue sharks (30% to 40%) compared

Fish Behavior in Relation to Longlines with squid bait (Watson et al. 2005). In the Hawaiibased swordfish fishery, regulations requiring vessels to switch from using a J-shaped hook with squid bait to a wider circle-shaped hook with fish bait led to a 36% decline in shark catch rate (Gilman et al. 2007). These studies thus confirm the speciesselective effect of bait type, suggesting that there is great potential for reducing shark bycatch by using alternative bait types. Elasmobranches are the most important bycatch species in the semipelagic longline fishery for hake on the coast of the Algarve (southern Portugal), and the effects of removing the lower hooks (i.e., those near the bottom) were evaluated in terms of bycatch reductions and target catch rates (Coelho et al. 2003). Most sharks were caught on hooks near the bottom. These hooks caught very few hake, which were mainly taken by the middle range of the hooks. These results thus indicate that the removal of the lower hooks would result in a significant reduction in shark bycatch with only a small reduction in target catch. Furthermore, the lower hooks often become entangled in the bottom substrate, so that removing them will also offer benefits in terms of handling, gear loss, and bait costs (Coelho et al. 2003). Experiments conducted off northeastern Australia on commercial pelagic longline vessels targeting tuna and billfish (Istiophoridae and Xiiphidae) found lower catch rates of sharks on nylon than on wire leaders (Ward et al. 2008). Sharks may bite through the nylon leaders and escape. Higher catch rates of tuna on nylon than on wire leaders were explained by higher visibility of wire leaders. Although shark caught on pelagic longlines tend to be in good condition and predicted postrelease survival may be high (Francis et al. 2001; Moyes et al. 2006; Ward et al. 2008), this mitigation method has potential to inflict injury and unaccounted mortality in sharks. Information on survivorship is difficult to evaluate (Skomal 2007), and postrelease mortality is evident and survival enhancement in even the more resilient shark species is advocated (Mandelman et al. 2008). Soak time may affect both catch rate and survival rate. Blue shark catch rates were found to increase with increasing soak time (Ward et al. 2004), and increased soak times led to increases in the propor-

129

tion of individuals retrieved dead (Diaz and Serafy 2005). Although shortening longline soak times might be a way of reducing blue shark mortality, this measure would probably be unacceptable to the fishermen and thus difficult to implement because the catches of swordfish would also be lowered with shorter soak times (Ward et al. 2004). The implementation of mitigation measures in longline fisheries (e.g., circle hooks to reduce sea turtle bycatch in the U.S. swordfish fishery; see Section 5.3) requires examination of the effects of such measures on catching efficiency for other nontarget species. Yokota et al. (2006) conducted fishing experiments off the coast of Japan to test the effects of circle hooks on blue shark (Prionace glauca) catches in pelagic longlining and found no significant differences in catch rates or proportion of dead individuals between circle hooks and conventional tuna hooks. In the north-western Atlantic, blue shark catch rates were 8% to 9% higher using circle hooks than when using J-hooks (Watson et al. 2005). Using circle hooks therefore seems to have no effect on reducing blue shark bycatch. However, circle hooks significantly reduced gut hooking in blue sharks compared with J-hooks; presumably post-hooking survival rates also increased with circle hooks (Watson et al. 2005). In most longline fisheries, pelagic sharks are primarily nontarget species and are discarded (Yokota et al. 2006), and it is therefore important to develop practices to minimize injury and postrelease mortality. We may thus conclude that there are few comprehensive studies aimed at the development of effective measures to reduce shark bycatch. More research is needed before appropriate measures can be implemented in the management of longline fisheries where shark bycatch is a problem. Elasmobranches make up a different group of animals than the species targeted in these fisheries, and they have a different feeding behavior with regard to sensory modalities involved, search strategy, prey preference, and diel rhythm. Several species of sharks are capable of detecting magnetic fields (Kalmijn 1971), and the 2006 winner of the “Smart Gear” competition suggested taking advantage of this unique sensory modality in sharks to deter them from taking baited hooks (see http://www.smartgear.org). A first step would be to study in depth

130

Fish Behavior near Fishing Gears during Capture Processes

how these animals search for and capture their food. 5.7 CONCLUDING REMARKS The responses of fish to baited hooks, and how their behavior is influenced by physical and biological variables, affect capture efficiency and size and species selectivity in longline fishing. An overview of the interactions between fish behavior and environmental and gear-related variables is shown in Figure 5.7. Understanding these processes is a prerequisite for improving the accuracy of stock size estimates when catch data from baited gears in resource assessment work are used (Løkkeborg

et al. 1995; Stoner 2004). Also, these processes affect the outcome of commercial fishing operations, and knowledge of how fish react to baited gears is therefore also of great importance for the sustainable harvesting and management of fish stocks. Finally, such information may also help develop mitigation measures for reducing incidental catches of seabirds, sea turtles, and nontarget fish without loss of target catch. Longlining is regarded as a size-selective fishing method, and several behavioral aspects explain why baited gears are more selective than, for example, trawling. Large fish are capable of exploiting food resources more efficiently than small fish, a situa-

Light INTERNAL AND EXTERNAL VARIABLES Ch. 5.2 - 5.5

- Hunger state - Reproductive status - Diel rhytms - Previous experienses

Temperature Current Prey density

STIMULI Ch. 4

VISUAL

OLFACTORY

GUSTATORY / MECHANICAL taste size

type GEAR VARIABLES Ch. 3

Bait texture size

Bait

Attractant concentration

shape

Hook design

prey

OBJECT CATEGORIZATION Ch. 5.1

Object

Object nonedible

predator prey nonedible

prey Object nonedible Escap (compe e titi

Escap e

) on

Search (olfactory rheotaxis) Ig

Attack Ig

est

Ing

Hooked

re no

re no

w

im

Locate

ay

Detect

aw

RESPONSES Ch. 5

S

Figure 5.7. Fish behavior to baited hooks. Variables (internal and external) that affect fish behavior but are not related to the baited gear are shown above the broken line. Stimuli, gear variables, and object categorization and the corresponding behavioral responses are shown at different distances down-current of the baited hooks. Grey and color shades illustrate odor plume and visual range, respectively.

Fish Behavior in Relation to Longlines tion that leads to exploitative competition for baits on a longline (Stoner 2004; Section 5.4). Thus, before the fish come into contact with the gear, a selection process has already taken place that exposes a high proportion of large individuals to the gear (Løkkeborg and Bjordal 1992). Interestingly, seismic air-gun noise has been shown to cause greater reductions in longline catches of large than small cod (Engås et al. 1996b). The stronger response of larger fish was explained by size-dependent swimming capability. There is also competition among fish that encounter baited longlines (i.e., interference competition) (Stoner 2004). In the seismic experiment by Engås et al. (1996b), longline catches of small cod were even shown to increase during seismic shooting, indicating that small fish were more successful in taking the available baits when their larger conspecifics had left the area. Furthermore, the characteristics and visual appearance of the gear may affect small and large individuals differently (see Section 5.1). Baited hooks are a novel food item, and because smaller fish have less experience concerning prey types and more limited diet breathe, their restraint in attacking a novel prey may be more pronounced (Løkkeborg 1990b). There is also a relationship between predator and prey size (Hart 1993; Werner 1974). Last, the hooking probability may be length dependent (Kaimmer 1999). Selective capture of large over small individuals on baited hooks has been demonstrated in several fishing experiments (Bertrand 1988; Engås et al. 1996a; Hamley and Skud 1978; Hovgård and Riget 1992; Huse et al. 1999; 2000). The factors that affect size selectivity are also likely to affect species selection. Large species swim faster and are more successful competitors than small species, and Pacific halibut have been shown to be more efficient than other species in competing for available baits (Skud 1978). Interspecific competition for baits has also been observed in other field studies (Fernö et al. 1986; Godø et al. 1997; Løkkeborg and Bjordal 1992). Perhaps the most important factor that affects species selectivity is bait type (Løkkeborg and Bjordal 1992). There is a very large body of literature demonstrating species-specific preference for feeding attractants, and comparative fishing experiments with different baits have demonstrated clear

131

effects on species composition (see Section 5.3). Fishermen use saithe when targeting wolffish in the Barents Sea and have experienced negligible catches when their longlines have happened to be set on fishing grounds with a low abundance of wolfish, indicating that other species (e.g., cod and haddock) show low preferences for saithe bait. Prey-size preferences may differ among species, and smaller baits resulted in significant increases in catch rates for haddock (120%) but no increase for large cod (Johannessen 1983). In addition to the factors that affect selectivity, longline catching efficiency is affected by gear parameters such as size and type of hook, swivel (versus traditional snood attachment), and mainline and snood materials, as well as operational factors such as setting time, fishing depth (pelagic lines), season, and setting direction relative to current direction. As these factors have pronounced effects on catch rates, they can and must be standardized when using catch per unit effort (CPUE) data from longline surveys as an index of stock size. In the fishery for Pacific halibut, there has been an almost total conversion from J-hooks to circle hooks, and this required adjustments to the CPUE data used for stock assessment as the circle hook is 2.2 times as efficient (Quinn et al. 1985). However, some environmental variables affecting catchability are not easy to measure or standardize. Stoner (2004) provides an interesting review of how physical and biological conditions in the environment influence fish activity, feeding motivation, searching behavior, and bait location and concludes that temperature, light level, current velocity, and prey density have the greatest effects on longline catchability, potentially affecting variation in CPUE by a factor of 10. Target species can occupy wide ranges of these environmental conditions over time and space, and longline survey data used for stock assessments may thus simply reflect variations in fish behavior and catchability rather than trends in fish abundance (Stoner 2004). During the past few years, promising results and important technical advances have made longline fishing more conservation oriented, particularly regarding incidental catches of seabirds. Mitigation measures such as streamer lines, weighted lines, and night setting have nearly eliminated seabird

132

Fish Behavior near Fishing Gears during Capture Processes

bycatch in important demersal fisheries and greatly reduced incidental catches in most pelagic fisheries. Thus, the implementation of appropriate and efficient measures and their compliance have the potential to contribute to negligible longlineinduced mortality in most seabird populations. Measures to reduce incidental captures of sea turtles, such as the circle hook, using finfish as bait, and deep setting, have proved efficient in protecting these endangered species. However, there is no single solution to the problem of incidental catches of sea turtles, and there are great variations in gear construction and mode of operating pelagic longlines. Reported experiments have been carried out only in a few fisheries, and our knowledge of the effects of proposed mitigation measures on sea turtle bycatch and target species catch is too limited to draw firm conclusions except in the case of a few fisheries. Information on how to reduce shark bycatch is very scarce. In general, bait type is the most important gear parameter affecting longline species selectivity (Løkkeborg and Bjordal 1992), and findings suggest that this also applies to sharks (Watson et al. 2005). There are also indications that magnetic fields could be used to deter sharks from baited hooks.

5.8 FUTURE CHALLENGES Ecologically sustainable management for fisheries and conservation-oriented harvesting practices is a topic of growing interest, and consumers are becoming more concerned about how their food has been produced. Several factors that affect species and size selectivity in longlining were discussed in this chapter. Identifying the gear parameters and operational strategies with the greatest potential for selective fishing should be given priority. Combining the most effective factors should improve longline selectivity and thus make this fishing method more sustainable. Gear research carried out in the course of the past few decades has considerably improved the catching efficiency of longlines. One example is the conversion from J-hooks to circle hooks in demersal fisheries. These two hook types are very different in design, and there is probably potential for further improvement through minor changes in hook

design. Choice of bait type is mainly based on fishermen’s experiences, availability, and price, and few comparative fishing experiments have tested the catch efficiency and species selectivity of different natural bait types. Thus, experiments designed to compare different bait types are likely to reveal a potential for improved efficiency and selectivity. This is supported by findings from the study carried out in the western Atlantic swordfish fishery, where mackerel bait was tested as a mitigation measure for sea turtles. Mackerel bait produced 63% higher catch rates of swordfish compared with traditional squid bait in the experiments with J-hooks (Watson et al. 2005). More research should also be aimed at the development of artificial baits based on surplus products or waste materials, as most baits used today are made from resources that could be better used for human consumption. Attention should be paid to developing long-lasting baits, as the release rate of attractants from natural baits decreases rapidly (Løkkeborg 1990a). Furthermore, research should aim to optimize aspects of operational fishing strategy, such as time of setting in relation to diel and tidal cycles, soak time, setting direction relative to the current, and fishing depth. The main conservation issue related to longline fishing is incidental captures of threatened and endangered species such as seabirds, turtles, and sharks. Efficient mitigation measures to reduce incidental captures have been developed for many fisheries, and promising results have been obtained for others. With respect to seabirds, future research should use an experimental approach to fine-tuning the most promising mitigation measures for each specific fishery as there is no single solution to reducing incidental mortality in longline fisheries. These measures include streamer lines, night setting, weighted longlines, and side-setting. Research aimed at reducing sea turtle mortality in longlining is in its infancy, although promising results have already been obtained in the western Atlantic swordfish fishery. Studies should be carried out to test these findings (e.g., circle hooks, finfish as bait, weighted leaders, deep sets, reduced daylight soak time, avoidance of areas of high turtle densities) in other fisheries where interactions with sea turtles are a problem. New experiments are

Fish Behavior in Relation to Longlines essential because, as with seabird interactions, there are differences between fishing grounds in species assemblages, longline gear design, and operational strategies, and best fishing practices to reduce incidental captures will differ accordingly. Although studies aimed at reducing shark bycatch are few, much can be learned from our general knowledge of longline species selectivity. Modern statistical methods (e.g., generalized additive models) and commercial fisheries data can be used to examine and estimate the relative influence of various factors on catch rates, and these models may be used to identify the spatiotemporal, environmental, and operational variables that have the greatest potential for reducing shark bycatch while maintaining profitable catch rates of target species (see Bigelow et al 1999; Walsh and Kleiber 2001). A better understanding of how fish respond to baited hooks is essential for improving the selectivity, efficiency, and conservation aspects of longline fishing. There are a few fisheries, such as the Barents Sea pelagic fishery for haddock, that take high proportions of undersized fish (Huse and Soldal 2000; Løkkeborg and Bjordal 1995), and better understanding of fish behavior may aid in solving this problem. Improved stock size estimation models are perhaps the issue that could gain most from more research on fish behavior in response to baited gear. Stoner (2004) concludes that several environmental variables can have a significant impact on the feeding behavior of fish and produce serious biases in stock assessment surveys carried out using baited gear. To produce unbiased stock size estimates, it is essential to understand relationships between the environment and feeding biology and how physical and biological variables affect fish activity and responsiveness to bait. Such information can be obtained by linking continuous recordings of environmental variables with behavioral observations using acoustic and archival tags, optical and acoustic cameras, acoustic surveys, and laboratory facilities (Stoner 2004). ACKNOWLEDGMENTS We are grateful to Anne-Britt Skar Tysseland for preparing the figures. We also thank Alan W. Stoner for valuable comments that greatly improved the manuscript.

133

REFERENCES Adron JW and Mackie AM. 1978. Studies on the chemical nature of feeding stimulants for rainbow trout, Salmo gairdneri Richardson. J. Fish Biol. 12: 303–310. Agnew DJ, Black AD, Croxall JP and Parkes GB. 2000. Experimental evaluation of the effectiveness of weighting regimes in reducing seabird by-catch in the longline toothfish fishery around South Georgia. CCAMLR Sci. 7: 119–131. Allen KR. 1963. The influence of behavior on the capture of fishes with baits. In: The Selectivity of Fishing Gear. Int. Comm. Northwest Atl. Fish., Spec. Publ. 5: 5–7. Dartmouth, NS, Canada. Ashford JR, Croxall JP, Rubilar PS and Moreno CA. 1995. Seabird interactions with longlining operations for Dissosticus eleginoides around South Georgia, April to May 1994. CCAMLR Sci. 2: 111–121. Atema J. 1980. Chemical senses, chemical signals and feeding behavior in fishes. In: Bardach JE, Magnuson JJ, May RC and Reinhart JM (eds). Fish Behavior and its Use in the Capture and Culture of Fishes. pp 57–101. Manila: International Center for Living Aquatic Resources Management. Baez JC, Real R, Caminas JA and Juan A. 2007. Differential distribution within longline transects of loggerhead turtles and swordfish captured by the Spanish Mediterranean surface longline fishery. J. Mar. Biol. Assoc. U.K. 87: 801–803. Bardach JE,Todd JH and Crickner R. 1967. Orientation by taste in fish of the genus Ictalurus. Science. 155: 1276–1278. Belghaug R and Døving KB. 1977. Odor threshold determined by studies of the induced waves in the olfactory bulb of the char (Salmo alpinus L.). Comp. Biochem. Physiol. 57(A): 327–330. Bertrand J. 1988. Selectivity of hooks in the handline fishery of the Saya de Malha banks (Indian Ocean). Fish. Res. 6: 249–255. Bertrand A, Josse E, Bach P, Gros P and Dagorn L. 2002. Hydrological and trophic characteristics of tuna habitat: consequences on tuna distribution and longline catchability. Can. J. Fish. Aquat. Sci. 59: 1002–1013. Beukema JJ. 1970. Angling experiments with carp: decreased catchability through one trial learning. Neth. J. Zool. 20: 81–92. Beverly S, Curran D, Musyl M and Molony B. 2009. Effects of eliminating shallow hooks from tuna longline sets on target and non-target species in the Hawaii-based pelagic tuna fishery. Fish. Res. 96: 281–288.

134

Fish Behavior near Fishing Gears during Capture Processes

Bigelow KA, Boggs CH and He X. 1999. Environmental effects on swordfish and blue shark catch rates in the US North Pacific longline fishery. Fish. Oceanogr. 8(3): 178–198. Bjordal Å. 1983. Effect of different long-line baits (mackerel, squid) on catch rates and selectivity for tusk and ling. ICES CM/B: 31. 9 pp. Bjordal Å and Løkkeborg S. 1996. Longlining. Oxford: Fishing News Books. 156 pp. Boggs CH 2001. Deterring albatrosses from contacting baits during swordfish longline sets. In: Melvin EF and Parrish JK (eds). Seabird Bycatch: Trends, Roadblocks, and Solutions. pp 79–94. AK-SG01-01. Fairbanks, AK: University of Alaska Sea Grant. Bossert WH and Wilson EO 1963. The analysis of olfactory communication among animals. J. Theor. Biol. 5: 443–469. Brawn VM. 1969. Feeding behavior of Cod (Gadus morhua). J.Fish. Res. Bd Can. 26: 583–596. Brazner JC and McMillan J. 2008. Loggerhead turtle (Caretta caretta) bycatch in the western North Atlantic and opportunities for mitigation. Fish. Res. 91: 310–324. Brothers N. 1991. Albatross mortality and associated bait loss in Japanese longline fishery in the Southern Ocean. Biol. Conserv. 55: 255–268. Brothers NP, Cooper J and Løkkeborg S. 1999a. The incidental catch of seabirds by longline fisheries: worldwide review and technical guidelines for mitigation. FAO Fish. Circ. 937: 100 pp. Brothers N, Gales R and Reid T. 1999b. The influence of environmental variables and mitigation measures on seabird catch rates in the Japanese tuna longline fishery within the Australian Fishing Zone, 1991– 1995. Biol. Conserv. 88: 85–101. Brown C, Laland K and Krause J. 2006. Fish Cognition and Behavior. Fish and Aquatic Resources Series 11. Oxford: Blackwell Publishing. 328 pp. Bull LS. 2007. Reducing seabird bycatch in longline, trawl and gillnet fisheries. Fish & Fish. 8: 31–56. Caminas JA, Baez JC, Valeiras X and Real R. 2006. Differential loggerhead by-catch and direct mortality due to surface longlines according to boat strata and gear type. Sci. Mar. 70: 661–665. Caprio J. 1978. Olfaction and taste in the channel catfish: an electrophysiological study of the responses to amino acids and derivatives. J. Comp. Physiol. 123: 357–371. Carr WES. 1982. Chemical stimulation of feeding behavior. In: Hara TJ (ed). Chemoreception in Fishes. pp 259–273. Amsterdam: Elsevier.

Carr WES and Derby CD. 1986. Chemically stimulated feeding behavior in marine animals. J. Chem. Ecol. 12: 989–1011. Carr WES, Netherton JC III, Gleeeson RA and Derby CD. 1996. Stimulants of feeding behavior in fish: analysis of tissue of diverse marine organisms. Biol. Bull. 190: 149–160. Casale P, Freggi D and Rocco M. 2008. Mortality induced by drifting longline hooks and branchlines in loggerhead sea turtles through observation in captivity. Aquat. Conserv. Mar. Freshw. Ecosys. 18: 945–954. Castonguay M and Cyr DG. 1998. Effects on temperature on spontaneous and thyroxine-stimulated locomotor activity of Atlantic cod. J. Fish Biol. 53: 303–313. Chandroo KP, Duncan IJH and Moccia RD. 2004. Can fish suffer? Perspectives on sentience, pain, fear and stress. App. Anim. Behav. Sci. 86: 225– 250. Cherel Y, Weimerskirch H and Duhamel G. 1996. Interaction between longline vessels and seabirds in Kerguelen waters and a method to reduce seabird mortality. Biol. Conserv. 75: 63–70. Coelho R, Bentes L, Goncalves JMS, Lino PG, Ribeiro J and Erzini K. 2003. Reduction of elasmobranch by-catch in the hake semipelagic near-bottom longline fishery in the Algarve (Southern Portugal). Fish. Sci. 69: 293–299. Croxall JP, Rothery P, Pickering S and Prince PA. 1990. Reproductive performance, recruitment and survival of Wandering Albatrosses Diomedea exulans at Bird Isand, South Georgia. J. Anim. Ecol. 59: 775–796. Davis MW, Spencer ML and Ottmar ML. 2006. Behavioral responses to food odor in juvenile marine fish: acuity varies with species and fish size. J. Exp. Mar. Biol. Ecol. 328: 1–9. Deflorio M, Aprea A, Corriero A, Santamaria N and Metrio G. 2005. Incidental captures of sea turtles by swordfish and albacore longlines in the Ionian sea. Fish. Sci. 71: 1010–1018. Dempsey CH. 1978. Chemical stimuli as a factor in feeding and intraspecific behavior of herring larvae. J. Mar. Biol. Assoc. U.K. 58: 739–747. Diaz GA and Serafy JE. 2005. Longline-caught blue shark (Prionace glauca): factors affecting the numbers available for live release. Fish. Bull. 103: 720–724. Dietrich KS, Melvin EF and Conquest L. 2008. Integrated weight longlines with paired streamer lines- Best practice to prevent seabird bycatch in

Fish Behavior in Relation to Longlines demersal longline fisheries. Biol. Conserv. 141: 1793–1805. Elkinton JS, Cardé RT and Mason CJ. 1984 Evaluation of time-average dispersion models for estimating pheromone concentration in a deciduous forest. J. Chem. Ecol. 10: 1081–1108. Ellingsen OF and Døving KB 1986. Chemical fractionation of shrimp extracts inducing bottom food search behavior in cod (Gadus morhua L.). J. Chem. Ecol. 12: 155–168. Engås A and Løkkeborg S. 1994. Abundance estimation using bottom gillnet and longline—the role of fish behavior. In: Fernö A. and Olsen S (eds). Marine Fish Behavior in Capture and Abundance Estimation. pp 134–165. Oxford: Fishing News Books. Engås A, Løkkeborg S, Ona E and Soldal AV. 1996b. Effects of seismic shooting on local abundance and catch rates of cod (Gadus morhua) and haddock (Melanogrammus aeglefinus). Can. J. Fish. Aquat. Sci. 53: 2238–2249. Engås A, Løkkeborg S, Soldal AV and Ona E. 1996a. Comparative fishing trials for cod and haddock using commercial trawl and longline at two different stock levels. J. Northw. Atl. Fish. Sci. 19: 83– 90. Fernö A. 1993 Advances in understanding of basic behavior: consequences for fish capture studies. ICES Mar. Sci. Symp. 196: 5–11. Fernö A and Huse I. 1983. The effect of experience on the behavior of cod (Gadus morhua L.) toward a baited hook. Fish. Res. 2: 19–28. Fernö A, Huse G, Jakobsen PJ and Kristiansen TS. 2006. The role of fish learning skills on fisheries and aquaculture. In: Brown C, Laland K and Krause J (eds). Fish Cognition and Behavior. pp 278–310. Oxford: Blackwell Publishing. Fernö A, Solemdal P and Tilseth S. 1986. Field studies on the behavior of whiting (Gadus merlangus L.) toward baited hooks. Fisk Dir. Skr. Ser. Hav Unders. 18: 83–95. Francis MP, Griggs LH and Baird SJ. 2001. Pelagic shark bycatch in the New Zealand tuna longline fishery. Mar. Fresh. Res. 52: 165–178. Franco J, Johnstone ADF and Mackie AM. 1991. Studies of bait preference in the cod, Gadus morhua L.: characterization of feeding stimulants using an operant conditioning technique. Fish. Res. 10: 229–242. Fry FEJ. 1971. The effect of environmental factors on the physiology of fish. In: Hoar WS and Randall DJ (eds). Fish Physiology, Vol. 6. Environmental

135

Relations and Behavior. pp 1–98. San Diego, CA: Academic Press. Gilman E, Boggs C and Brothers N. 2003. Performance assessment of an underwater setting chute to mitigate seabird bycatch in the Hawaii pelagic longline tuna fishery. Ocean Coast. Manag. 46: 985–1010. Gilman E, Brothers N and Kobayashi DR. 2005. Principles and approaches to abate seabird by-catch in longline fisheries. Fish Fish. 6: 35–49. Gilman E, Kobayashi D, Swenarton T, Brothers N, Dalzell P and Kinan-Kelly I. 2007. Reducing sea turtle interactions in the Hawaii-based longline swordfish fishery. Biol. Conserv. 139: 19–28. Gislason H. 1994. Ecosystem effects of fishing activities in the North Sea. Mar. Pollut. Bull. 29: 520–527. Godø OR, Huse I and Michalsen K. 1997. Short communication: Bait defence behavior of wolffish and its impact on long-line catch rates. ICES J. Mar. Sci. 54: 272–275. Goh Y, Tamura T and Kobayashi H. 1979. Olfactory responses to amino acids in marine teleosts. Comp. Biochem. Physiol. 62(A): 863–868. Hamley JM and Skud BE. 1978. Factors affecting longline catch and effort: II. Hook spacing. Int. Pac. Halib. Comm. Sci. Rep. 64: 15–24. Hara TJ. 1972. Electrical responses of the olfactory bulb of Pacific salmon, Oncorhynchus nerka and Oncorhynchus kisutch. J. Fish. Res. Bd. Can. 29: 1351–1355. Hara TJ. 1973. Olfactory responses to amino acids in rainbow trout, Salmo gairdneri. Comp. Biochem. Physiol. 44A: 407–416. Hara TJ. 1975. Olfaction in fish. Prog. Neurobiol. 5: 271–335. Hara TJ. 1993. Role of olfaction in fish behavior. In: Pitcher TJ (ed). Behavior of Teleost Fishes. pp 171–199. London: Chapman and Hall. Hara TJ. 1992. Fish Chemoreception. London: Chapman and Hall. 373 pp. Hart PJB. 1993. Teleost foraging: facts and theories. In: Pitcher TJ (ed). Behavior of Teleost Fishes. pp 253–284. London: Chapman and Hall. Hart PJB and Connellan B. 1984. The cost of prey capture, growth rate and ration size in pike (Esox lucius) as functions of prey weight. J. Fish Biol. 25: 279–291. He P. 1996. Bait loss from bottom-set longlines as determined by underwater observations and comparative fishing trials. Fish. Res. 27: 29–36. He P. 2003. Swimming behaviour of winter flounder (Pleuronectes americanus) on natural fishing

136

Fish Behavior near Fishing Gears during Capture Processes

grounds as observed by underwater video camera. Fish. Res. 60: 507–514. Heppell SS, Heppell SA, Read AJ and Crowder LB. 2005. Effects of fishing on long-lived marine organisms. In: Norse E and Crowder L (eds). Marine Conservation Biology: The Science of Maintaining the Sea’s Biodiversity. pp 211–231. Washington, DC: Island Press. Hovgård H and Riget FF. 1992. Comparison of longline and trawl selectivity in cod surveys off West Greenland. Fish. Res. 13: 323–333. Huse I and Fernö A. 1990. Fish behavior studies as an aid to improved longline hook design. Fish. Res. 9: 287–297. Huse I and Soldal AV. 2000. An attempt to improve size selection in pelagic longline fisheries for haddock. Fish. Res. 48: 43–54. Huse I, Gundersen AC and Nedreaas KH. 1999. Relative selectivity of Greenland halibut (Reinhardtius hippoglossoides, Walbaum) by trawls, longlines and gillnets. Fish. Res. 44: 75–93. Huse I, Løkkeborg S and Soldal A. 2000. Relative selectivity in trawl, longline and gillnet fisheries for cod and haddock. ICES J. Mar. Sci. 57: 1271–1282. Imai T. 1972. Studies on the several raw fish baits in a tuna long-line. Fishing-I. Mem. Fac. Fish., Kagoshima Univ. 21: 45–50. Imai T and Shirakawa O. 1972. Studies on the several raw fish baits in tuna long-line. Fishing-II. Mem. Fac. Fish., Kagoshima Univ. 21: 51–62. Ishida Y and Kobayashi H. 1992. Stimulatory effectiveness of amino acids on the olfactory response in an algivorous marine teleost, the rabbitfish Siganus fuscescens Houttuyn. J. Fish Biol. 41: 737–748. Januma S, Miyajima K and Abe T. 2003. Development and comparative test of squid liver artificial bait for tuna longline. Fish. Sci. 69: 288–292. Johannessen T. 1983. Influence of hook and bait size on catch efficiency and length selection in longlining for cod (Gadus morhua L.) and haddock (Melanogrammus aeglefinus L.). MSc. thesis, University of Bergen (in Norwegian). Johannessen T, Fernö A and Løkkeborg S. 1993. Behavior of cod (Gadus morhua) and haddock (Melanogrammus aeglefinus) in relation to various sizes of long-line bait. ICES Mar. Sci. Symp. 196: 47–50. Johnstone ADF. 1980. The detection of dissolved amino acids by the Atlantic cod, Gadus morhua L. J. Fish Biol. 17: 219–230.

Johnstone ADF and Mackie AM. 1990. Laboratory investigations of bait acceptance by the cod, Gadus morhua L.: identification of feeding stimulants. Fish. Res. 9: 219–230. Jones KA. 1992. Food search behavior in fish and the use of chemical lures in commercial and sports fishing. In: Hara TJ (ed). Fish Chemoreception. pp 288–320. London: Chapman and Hall. Kaimmer SM. 1999. Direct observations on the hooking behavior of Pacific halibut, Hippoglossus stenolepis. Fish. Bull. 97: 873–883. Kalmijn AJ. 1971. The electric sense of sharks and rays. J. Exp. Biol. 55: 371–383. Kallayil JK, Jorgensen T, Engås A and Fernö A. 2003. Baiting gill nets—how is fish behavior affected? Fish. Res. 61: 125–133. Kasumyan AO and Døving KB. 2003. Taste preferences in fishes. Fish Fish. 4: 289–347. Klaer N and Polacheck T. 1998. The influence of environmental factors and mitigation measures on bycatch rates of seabirds by Japanese longline fishing vessels in the Australian region. Emu. 98: 305–316. Kotrschal K, van Staaden MJ and Huber R. 1998 Fish brains: evolution and environmental relationships. Rev. Fish Biol. Fish. 8: 373–408. Krebs JR. 1978. Optimal foraging: Decision rules for predators. In: Krebs JR and Davis NB (eds). Behavioral Ecology: An Evolutionary Approach. pp. 23–63. Oxford: Blackwell Scientific Publications. Laberge F and Hara TJ. 2001. Neurobiology of fish olfaction: a review. Brain Res. Rev. 36: 46–59. Lewison RL. Freeman SA and Crowder LB. 2004. Quantifying the effects of fisheries on threatened species: the impact of pelagic longlines on loggerhead and leatherback sea turtles. Ecol. Lett. 7: 221–231. Løkkeborg S. 1990a. Rate of release of potential feeding attractants from natural and artificial bait. Fish. Res. 8: 253–261. Løkkeborg S. 1990b. Reduced catch of under-sized cod (Gadus morhua) in longlining by using artificial bait. Can. J. Fish. Aquat. Sci. 47: 1112–1115. Løkkeborg S. 1991. Fishing experiments with an alternative longline bait using surplus fish products. Fish. Res. 12: 43–56. Løkkeborg S. 1994. Fish behavior and longlining. In: Fernö A and Olsen S (eds). Marine Fish Behavior in Capture and Abundance Estimation. pp 9–27. Oxford: Fishing News Books. Løkkeborg S. 1998. Feeding behavior of cod, Gadus morhua: activity rhythm and chemically mediated food search. Anim. Behav. 56: 371–378.

Fish Behavior in Relation to Longlines Løkkeborg S. 2001. Reducing seabird bycatch in longline fisheries by means of bird-scaring lines and underwater setting. In: Melvin EF and Parrish JK (eds). Seabird Bycatch: Trends, Roadblocks, and Solutions. pp 33–41. AK-SG-01-01. Fairbanks, AK: University of Alaska Sea Grant. Løkkeborg S. 2003. Review and evaluation of three mitigation measures—bird-scaring line, underwater setting and line shooter—to reduce seabird bycatch in the north Atlantic longline fishery. Fish. Res. 60: 11–16. Løkkeborg S. 2008. Review and assessment of mitigation measures to reduce incidental catch of seabirds in longline, trawl and gillnet fisheries. FAO Fish. Circ. 1040: 24 pp. Løkkeborg S and Bjordal Å. 1992. Species and size selectivity in longline fishing: a review. Fish. Res. 13: 311–322. Løkkeborg S and Bjordal Å. 1995. Size-selective effects of increasing bait size by using an inedible body on longline hooks. Fish. Res. 24: 273–279. Løkkeborg S and Fernö A. 1999. Diel activity pattern and food search behavior in cod, Gadus morhua. Environ. Biol. Fish. 54: 345–353. Løkkeborg S and Johannessen T. 1992. The importance of chemical stimuli in bait fishing— fishing trials with presoaked baits. Fish. Res. 14: 21–29. Løkkeborg S and Pina T. 1997. Effects of setting time, setting direction and soak time on longline catch rates. Fish. Res. 32: 213–222. Løkkeborg S and Robertson G. 2002. Seabird and longline interactions: effects of a bird-scaring streamer line and line shooter on the incidental capture of northern fulmars (Fulmarus glacialis). Biol. Conserv. 106: 359–364. Løkkeborg S, Bjordal Å and Fernö A. 1989. Responses of cod (Gadus morhua) and haddock (Melanogrammus aeglefinus) to baited hooks in the natural environment. Can. J. Fish. Aquat. Sci. 46: 1478–1483. Løkkeborg S, Bjordal Å and Fernö A. 1993. The reliability and value of studies of fish behavior in longline gear research. ICES Mar. Sci. Symp. 196: 41–46. Løkkeborg S, Olla BL, Pearson WH and Davis MW. 1995. Behavioral responses of sablefish, Anoplopoma fimbria, to bait odor. J. Fish Biol. 46: 142–155. Løkkeborg S, Skajaa K and Fernö A. 2000. Foodsearch strategy in ling (Molva molva L.): crepuscular activity and use of space J. Exp. Mar. Biol. Ecol. 247: 195–208.

137

Mackie AM. 1982. Identification of the gustatory feeding stimulants. In: Hara TJ (ed). Chemoreception in Fishes. pp 275–291. Amsterdam: Elsevier. Mackie AM and Adron JW. 1978. Identification of inosine and inosine-5’-monophosphate as the gustatory feeding stimulants for the turbot, Scophthalmus maximus. Comp. Biochem. Physiol. 60(A): 79–83. Magnuson JJ, Bjorndal KA, Dupaul WA, Graham GL, Owens FW, Peterson CH, Pritchard PCH, Richardson JI, Saul GE and West CW. 1990. Decline of the Sea Turtles: Causes and Prevention. Washington, DC: National Academies Press. Magurran A, Seghers BH, Carbalho GR and Shaw PW. 1993. Evolution of adaptive variation in antipredator behavior. Mar. Behav. Physiol. 23: 29–44. Mandelman JW, Cooper PW, Werner TB and Lagueux KM. 2008. Shark bycatch and depredation in the U.S. Atlantic pelagic longline fishery. Rev. Fish Biol. Fish. 18: 427–442. Marin YH, Brum F, Barea LC and Chocca JF. 1998. Incidental catch associated with swordfish longline fisheries in the south-west Atlantic Ocean. Mar. Fresh. Res. 49: 633–639. Martin WR and McCracken FD. 1954. Relative efficiency of baits for groundfish. Progr. Rep. Atl. Biol. Sta. 126: 17–20. Marui T and Caprio J 1992. Teleost gustation. In: Hara TJ (ed). Fish Chemoreception. pp 171–198. London: Chapman and Hall. McCracken FD. 1963. Seasonal movements of the winter flounder, Pseudopleuronectes americanus, (Walbaum) on the Atlantic coast. J. Fish. Res. Bd Can. 20: 551–586. McMahon TE and Holanov SH. 1995. Foraging success of largemouth bass Micropterus salmoides at different light intensities: implications for time and depth of feeding. J. Fish Biol. 46: 759–767. McQuinn IH, Gendron L and Himmelman JH. 1988. Area of attraction and effective area fished by a whelk (Buccinum undatum) trap under variable conditions. Can. J. Fish. Aquat. Sci. 44: 2054–2060. Megalofonou P, Constantinos Y, Damalas D, De Metrio G, Deflorio M, de la Serna JM and Macias D. 2005. Incidental catch and estimated discards of pelagic sharks from the swordfish and tuna fisheries in the Mediterranean Sea. Fish. Bull. 103: 620–634. Melvin EF, Parrish JK, Dietrich KS and Hamel OS. 2001. Solutions to seabird bycatch in Alaska’s demersal longline fisheries. Washington Sea Grant Program. Project A/FP-7. WSG-AS 01-01.

138

Fish Behavior near Fishing Gears during Capture Processes

Michalsen K, Godø OR and Fernö A. 1996. Diel variation in the catchability of gadoids and its influence on the reliability of abundance indices. ICES J. Mar. Sci. 53: 389–395. Milinski M. 1993. Predation risk and feeding behavior. In: Pitcher TJ (ed). Behavior of Teleost Fishes. pp 285–305. London: Chapman and Hall. Moloney CL, Cooper J, Ryan PG and Siegfried WR. 1994. Use of population model to assess the impact of longline fishing on Wandering Albatrosses Diomedea exulans populations Biol. Conserv. 70: 195–203. Moreno CA, Rubilar PS, Marschoff E and Benzaquen L. 1996. Factors affecting the incidental mortality of seabirds in the Dissostichus eleginoides fishery in the southwest Atlantic (Subarea 48.3, 1995 season). CCAMLR Sci. 3: 79–91. Moyes CD, Fragoso N, Musyl MK and Brill RW. 2006. Predictiong postrelease survival in large pelagic fish. Trans. Am. Fish. Soc. 135: 1389–1397. Murray TE, Bartel JA, Kalish R and Taylor PR. 1993. Incidental capture of seabirds by Japanese bluefin tuna longline vessels in New Zealand waters, 1988– 1992. Biol. Conserv. 3: 181–210. Nel DC, Ryan PG and Watkins BP. 2002. Seabird mortality in the Patagonian toothfish longline fishery around Prince Edward Islands. Antarct. Sci. 14(2): 151–161. Nikonov AA, Ilyin YN, Zherelova OM and Fesenko EE. 1990. Odor thresholds of the Black Sea skate (Raja clavata). Electrophysiological study. Comp. Biochem. Physiol. 95(A): 325–328. Pawson MG. 1977. Analysis of a natural chemical attractant for whiting, Merlangius merlangus L. and cod, Gadus morhua L. using a behavioral bioassay. Comp. Biochem. Physiol. 56(A): 129–135. Pearson WH and Olla BL. 1977. Chemoreception in the blue crab, Callinectes sapidus. Biol. Bull. 153: 346–354. Pearson WH, Miller SE and Olla BL. 1980. Chemoreception in the food-searching and feeding behavior of the red hake, Urophycis chuss (Walbaum). J. Exp. Mar. Biol. Ecol. 48: 139–150. Peeling D and Rodgers DM. 1985. Circle hook comparison study. Fisheries and Oceans, Scotia-Fundy region, Halifax, Canada. Project Rep. 78: 16 pp. Polovina JJ, Howell E, Parker DM and Balazs GH. 2003. Dive-depth distribution of loggerhead (Caretta caretta) and olive ridley (Lepidochelys olivacea) sea turtles in the central North Pacific: might deep longline sets catch fewer turtles? Fish. Bull. 101: 189–193.

Poncet S, Robertson G, Phillips RA, Lawton K, Phalan B, Trathan PN and Croxall JP. 2006. Status and distribution of wandering, black-browed and greyheaded albatrosses breeding at South Georgia. Polar Biol. 29: 772–781. Pradhan NC and Leung P. 2006. Incorporating sea turtle interactions in a multi-objective programming model for Hawaii’s longline fishery. Ecol. Econ. 60: 216–227. Prince PA, Rothery P, Croxall JP and Wood AG. 1994. Population dynamics of Black-browed and Greyheaded Albatrosses Diomedea melanophrys and D. chrysostoma at Bird Island, South Georgia. Ibis 136: 50–71. Quinn TJ, Deriso RB and Hoag SH. 1985. Methods of Population Assessment of Pacific Halibut. Sci. Rep. IPHC. 72: 52 pp. Read AJ. 2007. Do circle hooks reduce the mortality of sea turtles in pelagic longlines? A review of recent experiments. Biol. Conserv. 135: 155– 169. Reid TA, Sullivan BJ, Pompert J, Enticott JW and Black AD. 2004. Seabird mortality associated with Patagonian Toothfish (Dissostichus eleginoides) longliners in Falkland Islands waters. Emu 104: 317–325. Robertson G, McNeill M, Smith N, Wienecke B, Candy S and Olivier F. 2006. The effectiveness of integrated weight (fast sinking) longlines in reducing the mortality of white-chinned petrels (Procellaria aequinoctialis) and sooty shearwaters (Puffinus griseus) in demersal longline fisheries. Biol. Conserv. 132: 458–471. Rodgveller CJ, Lunsford CR and Fujioka JT. 2008. Evidence of hook competition in longline surveys. Fish. Bull. 106(6): 364–374. Rolen SH, Sorensen PW, Mattson D and Caprio J. 2003. Polyamines as olfactory stimuli in the goldfish Carassius auratus. J. Exp. Biol. 206: 1683–1696. Rose CS, Stoner AW and Matteson K. 2005. Use of high-frequency imaging sonar to observe fish behavior near baited fishing gear. Fish. Res. 76: 291–304. Ryan PG and Watkins BP. 2002. Reducing incidental mortality of seabirds with an underwater longline setting funnel. Biol. Conserv. 104: 127–131. Ryer CH and Olla BL. 1992. Social mechanisms facilitating exploitation of spatially variable ephemeral food patches in a pelagic marine fish. Anim. Behav. 44: 69–74. Ryer CH and Olla BL. 1999. Light-induced changes in the prey consumption and behavior of two juve-

Fish Behavior in Relation to Longlines nile planktivorous fish. Mar. Ecol. Prog. Ser. 181: 41–51. Sainte-Marie B and Hargrave BT. 1987. Estimation of scavenger abundance and distrances of attraction to bait. Mar. Biol. 94: 431–443. Sasso CR and Epperly SP. 2007. Survival of pelagic juvenile loggerhead turtles in the open ocean. J. Wildl. Manag. 71: 1830–1835. Shiga M, Shiode D, Hayashi S, Tokai T and Hu F. 2008. Method for estimating buoyancy of midwater float required to standardize hook depth in pelagic longline. Fish. Sci. 74: 479–487. Shimada K and Tsurudome M. 1971. On the bait for tuna long-line-II. On the saury, mackerel and mackerel scad baits for tuna fishing. Mem. Fac. Fish. Kagoshima Univ. 20: 119–130. Shiode D, Hu F, Shiga M, Yokota K and Tokai T. 2005. Midwater float system for standardizing hook depths on tuna longlines to reduce sea turtle by-catch. Fish. Sci. 71: 1182–1184. Sigler MF. 1993. Stock Assessment and Management of Sablefish, Anoplopoma fimbria, in the Gulf of Alaska. PhD thesis, University of Washington. Sigler MF. 2000. Abundance estimation and capture of sablefish (Anoplopoma fimbria) by longline gear. Can. J. Fish. Aquat. Sci. 57: 1270–1283 Silver WL, Caprio J, Blackwell JF and Tucker D. 1976. The underwater electro-olfactogram: a tool for the study of the sense of smell of marine fishes. Experientia 32: 1216–1217. Skajaa K. 1997. Basic Movement Pattern and Chemooriented Search Toward Baited Gears in Demersal Species: A Field Study of Ling and Edible Crab. MSc thesis, University of Bergen. Skomal GB. 2007. Evaluating the physiological and physical consequences of capture on post-release survivorship in large pelagic fishes. Fish. Manag. Ecol. 14: 81–89. Skud BE. 1978. Factors affecting longline catch and effort: I. General Review. Int. Pac. Halib. Comm. Sci. Rep. 64: 5–14. South Pacific Regional Environment Programme (SPREP). 2001. A review of turtle bycatch in the western and central Pacific Ocean tuna fisheries. A report prepared for the South Pacific Regional Environment Programme (SPREP) by the Oceanic Fisheries Programme, Secretariat of the Pacific Community (SPC). 26 pp. Stiansen S. 2004. Fangstforskjeller Mellom to Typer Kongekrabbeteiner: Adferdstudier i Nærfeltet og Komparative Fiskeforsøk. MSc thesis, University of Bergen (in Norwegian).

139

Stoner AW. 2003. Hunger and light level alter response to bait by Pacific halibut: laboratory analysis of detection, location and attack. J. Fish Biol. 62: 1176–1193. Stoner AW. 2004. Effects of environmental variables on fish feeding ecology: implications for the performance of baited fishing gear and stock assessment. J. Fish Biol. 65: 1445–1471. Stoner AW and Ottmar ML. 2004. Fish density and size alter Pacific halibut feeding: implications for stock assessment. J. Fish Biol. 64: 1712–1724. Stoner AW and Sturm EA. 2004. Temperature and hunger mediate sablefish (Anoplopoma fimbria) feeding motivation: implications for stock assessment. Can. J. Fish. Aquat. Sci. 61: 238–246. Sutterlin AM. 1975. Chemical attraction of some marine fish in their natural habitat. J. Fish. Res. Bd. Can. 32: 729–738. Sutterlin AM and Sutterlin N. 1971: Electrical responses of the olfactory epithelium of Atlantic salmon (Salmo salar). J. Fish. Res. Bd Can. 28: 565–572. Suzuki N and Tucker D. 1971. Amino acids as olfactory stimuli in freshwater catfish, Ictalurus catus (Linn.). Comp. Biochem. Physiol. 40(A): 399–404. Suzuki Z, Warashina Y and Kishida M. 1977. The comparison of catches by regular and deep tuna longline gears in the western and equatorial Pacific. Bull. Far Seas Fish. Res. Lab. 15: 51–73. Swimmer Y, Arauz R, Higgins B, McNaughton L, McCracken M, Ballestero J and Brill R. 2005. Food color and marine turtle feeding behavior: can blue bait reduce turtle bycatch in commercial fisheries? Mar. Ecol. Prog. Ser. 295: 273–278. Swimmer Y, Arauz R, McCracken M, McNaughton L, Ballestero J, Musyl M, Bigelow K and Brill R. 2006. Diving behavior and delayed mortality of olive ridley sea turtles Lepidochelys olivacea after their release from longline fishing gear. Mar. Ecol. Prog. Ser. 323: 253–261. Thorpe JE. 1978. Rhythmic Activity of Fishes. London: Academic Press. 312 pp. Vabø R, Huse G, Fernö A, Jørgensen T, Løkkeborg S and Skaret G. 2004. Simulating search behavior of fish toward bait. ICES J. Mar. Sci. 61: 1224–1232. Walsh WA and Kleiber P. 2001. Generalized additive model and regression tree analyses of blue shark (Prionace glauca) catch rates by the Hawaii-based commercial longline fishery. Fish. Res. 53: 115–131.

140

Fish Behavior near Fishing Gears during Capture Processes

Ward P, Myers RA and Blanchard W. 2004. Fish lost at sea: the effect of soak time on pelagic longline catches. Fish. Bull. 102: 179–195. Ward P, Lawrence E, Darbyshire R and Hindmarsh S. 2008. Large-scale experiment shows that nylon leaders reduce shark bycatch and benefit pelagic longline fishers. Fish. Res. 90: 100–108. Watson JW, Epperly SP, Shah AK and Foster DG. 2005. Fishing methods to reduce sea turtle mortality associated with pelagic longlines. Can. J. Fish. Aquat. Sci. 62: 965–981. Webster DR, Rahman S and Dasi LP. 2001. On the usefulness of bilateral comparison to tracking turbulent chemical odor plumes. Limnol. Oceanogr. 46: 1048–1053. Webster DR and Weissburg MJ. 2001. Chemosensory guidance cues in a turbulent chemical plume. Limnol. Oceanogr. 46: 1034–1047. Weihs D. 1987. Hydromechanics of fish migration in variable environments. Am. Fish. Soc. Symp. 1: 254–261 Weimerskirch H and Jouventin P. 1987. Population dynamics of the Wandering Albatross, Diomedea exulans, of the Crozet Islands: causes and consequences of the population decline. Oikos 49: 315–322. Weimerskirch H, Brothers NP and Jouventin P. 1997. Population dynamics of Wandering Albatross Diomedea exulans and the Amsterdam Albatross D. amsterdamensis in the Indian Ocean and their rela-

tionship with longline fisheries: conservation implications. Biol. Conserv. 79: 257–270. Weimerskirch H, Capdeville D and Duhamel G. 2000. Factors affecting the number and mortality of seabirds attending trawlers and long-liners in the Kerguelen area. Polar Biol. 23: 236–249. Werner EE. 1974. The fish size, prey size, handling time relation in several sunfishes and some implications. J. Fish. Res. Bd. Can. 31: 1531–1536. Wienecke B and Robertson G. 2002. Seabird and seal—fisheries interactions in the Australian Patagonian toothfish Dissostichus eleginoides trawl fishery. Fish. Res. 54: 253–265. Wilson RR and Smith KL Jr. 1984. Effect of near bottom current on detection of bait by the abyssal grenadier fishes Coryphaenoides spp, recorded in situ with a video camera on a free vehicle. Mar. Biol. 84: 83–91. Yamaguchi Y, Nonoda T, Kobayashi H, Izawa K, Jinno T, Ishikura I, Uchida M and Tonogai M. 1983. Effectiveness of artificial bait for obtaining higher hooking rate on bottom set long-line fishing. Bull. Jpn. Soc. Sci. Fish. 42: 1819–1824. Yokota K, Kiyota M and Minamia H. 2006. Shark catch in a pelagic longline fishery: comparison of circle and tuna hooks. Fish. Res. 81: 337–341. Zimmer-Faust RK and Case JF. 1983. A proposed dual role of odor in foraging by the California spiny lobster, Panulirus interruptus (Randall). Biol. Bull. 164: 341–353.

Fish Behavior in Relation to Longlines SPECIES MENTIONED IN THE TEXT albacore, Thunnus alalunga American plaice, Hippoglossoides platessoides Atlantic cod, Gadus morhua bigeye tuna, Thunnus obesus bluefin tuna, Thunnus thynnus carp, Cyprinus carpio coho salmon, Oncorhynchus kisutch conger eel, Conger sp. goldfish, Carassius auratus Greenland halibut, Reinhardtius hippoglossoides haddock, Melanogrammus aeglefinus, hake, Merluccius sp. herring, Clupea harengus ling, Molva molva Pacific cod, Gadus macrocephalus, Pacific halibut, Hippoglossus stenolepis Patagonian toothfish, Dissostichus eleginoides pigfish, Orthopristis chrysopterus

pinfish, Lagodon rhomboides plaice, Pleuronectes platessa rabbitfish, Siganus fuscescens rainbow trout, Oncorhynchus mykiss red hake, Urophycis chuss sablefish, blackcod, Anoplopoma fimbria saithe, Pollachius virens salmon, Salmo salar sea catfish, Arius felis spiny dogfish, Squalus acanthias swordfish, Xiphias gladius turbot, Scophthalmus maximus tusk, Brosme brosme walleye pollock, Theragra chalcogramma white catfish, Ictalurus catus white hake, Urophycis tenuis wolffish, Anarhichas lupus yellowfin tuna, Thunnus albacares

141

Chapter 6 Fish Pots: Fish Behavior, Capture Processes, and Conservation Issues Bjarti Thomsen, Odd-Børre Humborstad, and Dag M. Furevik

sea with low mortality. Pots are thus seen as a benign fishing method and superior to other gear from an environmental point of view (Cole et al. 2003; Dayton et al. 1995; Jennings and Kaiser 1998). This chapter offers a brief overview of the use of fish pots in various parts of the world, describes how they work in relation to fish behavior, and discusses the conservation challenges involved in their operation. Further details on various types of pots from around the world and their construction and mode of operation can be found in work by Furevik (1994), Sainsbury (1996), Slack-Smith (2001), and Gabriel et al. (2005). Discussions on the large-scale fixed fishing gear, the trap, are given in Chapter 7.

6.1 INTRODUCTION A 2000-year-old Greek document on fishing, the Halieutika, described fish traps that “work while their masters sleep” (Bekker-Nielsen 2002). These ancient words underline one of the characteristics inherent in passive fishing gears such as pots and traps in contrast to most modern gears, which are operated actively. According to the FAO fishing gear classification, pots are a subgroup of traps. Fish traps are passive fishing gears that allow fish to enter easily but then make escape difficult. Traps may be constructed as fixed weirs or fences along the shore or in rivers that guide fish into confined spaces. Fish pots, on the other hand, are transportable boxlike or basketlike enclosures designed to capture fish by attracting them to the pot and luring them inside through one or more narrow “one-way” entrances (Fig. 6.1). Fish pots may be deployed individually or in strings of several pots. Pots range in size from a few liters up to a few cubic meters. In some places, small pots are known as “creels.” Pots have several characteristics that are desirable for a modern fishing gear. They are not laborintensive and use less energy in operation than active gears. Pots only affect an area of seabed of the same order of magnitude as the footprint of the gear itself. They deliver the catch alive and with minimal physical damage, which is a prerequisite for a high-quality product. Catching the fish alive also allows unwanted bycatch to be returned to the

6.2 WORLDWIDE USE OF FISH POTS Pots are important fishing gears for crustaceans (crabs and lobsters), but for fish they are less commercially important compared with other fishing gears. On a global scale, all trap fishing (including large-scale trap and small-scale pots for fish and crustaceans) ranks only seventh in catch quantity, preceded by seine, midwater trawl, bottom trawl, gillnet, hook and line, and dredge (Watson et al. 2006). However, fish pots are important in some regions. In tropical waters, coral reefs and outcrops largely prohibit fishing with trawl and many other gears, whereas pots are often the preferred gear for

143

144

Fish Behavior near Fishing Gears during Capture Processes

10

9

1 6

11

4

3

2

5 7 8

current

Figure 6.1. General illustration of fish pots. 1, Rigid frame covered by netting. 2, Entrance, usually cone or wedge shaped and may be equipped with triggers (3) to prevent fish from escape. 4, Bait bag. 5, Escape ring to allow small animals to escape. 6, Break-away panel, which will corrode or rot away to open the pot if lost. 7, Attachment line. 8, Anchor. 9, Buoy line. 10, Surface buoy. 11, Ground line (only used if pots are deployed in a longline). (Drawing: A-B Tysseland, Norway.)

bottom-dwelling fish species. In other parts of the world, pots have found use in specific areas or fisheries due to their special characteristics, mainly with respect to their low impact on the habitat and their ability to capture fish alive. Fish pots are probably best known from the Caribbean, where the Antillean pot has been the principal fishing gear for centuries (Earle 1889; Munro et al. 1971). In many parts of the Caribbean, they still account for more than 50% of landings. The type of pots used throughout the Caribbean is bound by tradition and they are usually constructed from locally available materials. Since around 1920, pots have mainly been constructed from galvanized hexagon-shaped mesh wire on a framework of mangrove or other wooden sticks. The pot fishery in the Caribbean is a typical multispecies fishery and many different species are taken. In Bermuda, pots were also the principal fishing gear until they were banned in 1990 because of overfishing. On the eastern seaboard of the United States, pots are used for a limited number of species, with black

sea bass (Centropristis striata) being the most important. Sea bass typically concentrate in structurally complex habitats, where pots are more effective than mobile fishing gears. Pots account for an average of 45% of the total sea bass landings from this region (Shepherd et al. 2002). On the West Coast of the United States and Canada, as well as offshore Alaska, pots were introduced in the fishery for sablefish (Anoplopoma fimbria), also called black cod, in the early 1970s. For a few years, pots were the most important gear, but their popularity decreased rapidly during the 1980s. However, in the Bering Sea and Aleutian Islands, pots have gained popularity in recent years because longlining has problems due to depredation by whales (Hanselman et al. 2006). Off the coast of Alaska, pots are used for Pacific cod (Gadus macrocephalus). In this fishery, heavy steel frame pots with dimension of 1.8 × 1.8 × 0.9 m are set as single units using purpose-built fishing vessels with pot launchers to handle the heavy gear (Fig. 6.2). Similar pots that are used for Pacific cod in Alaska

Fish Pots: Fish Behavior, Capture Processes, and Conservation Issues

145

Figure 6.2. Pacific cod pot used off Alaska. These pots are operated by purpose-built vessels with pot launchers to handle the heavy gear. (Photograph: P. Munro, USA.)

Figure 6.3. Traditional Turkish fish pots on the deck of a local wooden boat. The pots are made of wire and 80 cm in diameter with entrance in top middle. (Photograph: A. Lok, Turkey.)

have been introduced for Atlantic cod (Gadus morhua) in Newfoundland and New England with some success, but wider commercial use remains to be seen (Pol et al. 2005; Walsh et al. 2006). Many Asian countries and island countries in the Indian Ocean have a tradition of using pots. In the Arabian Gulf, wire basket pots (known as gargoor) that are semicircular in cross section with sizes up to 2 m in diameter are used. In Kuwait, pots account for 50% of finfish landings, and in Oman, pot fishing represents up to 19% of all gear used by the artisanal fishery (Al-Masroori et al. 2004). In the Seychelles, pots, mainly constructed of bamboo strips, account for 20% of the reef fish catch (Mahon and Hunte 2001). Since 1990 a live food-fish trade has developed throughout Southeast Asia, and in some places up to 55% of live catches come from pots. Pots are commonly used in various parts of Japan together with other fishing gears. One example is in Hokkaido, where pots were introduced in 1980 in the important fishery for arabesque greenling (Pleurogrammus azonus) (Li et al. 2006). In Australia, pots are an established fishing method in a 700-tonne (metric ton) multispecies fishery, with snappers (Pagrus auratus) accounting for more than half of the value (Stewart and Ferrell 2003). On the northwest shelf of Australia, a pot fishery developed using cylindrical pots to target

species from Nemipteridae, Lethrinidae, Lutjanidae, and Serranidae families. These pots, unique to this area, are 1.5 m in diameter and 0.9 m high (Whitelaw et al. 1991). In New Zealand, a commercial pot fishery for blue cod (Parapercis colias) takes place in depths of 30 m or more (Cole et al. 2003). Pots have a steel frame with dimensions of 1.9 × 1.4 × 0.9 m and conical entrances on the vertical sides. In Europe, the commercial use of fish pots is of minor importance compared with that of other fishing gears. One example is a traditional circular pot (Fig. 6.3) used on a small scale in Turkey and Greece. In Norway there is a limited small-scale fishery for wrasse (Labridae) that are used as cleaner fish in the salmon farming industry (Bjordal 1993). Recently, the development of a new twochamber pot (Figure 6.4) for Atlantic cod, tusk (Brosme brosme) and ling (Molva molva) has enjoyed some success and a few costal vessels have changed from longlining and gillnetting to pot fishing (Furevik and Skeide 2003). Commercial fish pots are thus in use all over the world, and are produced in various sizes and designs, ranging from small pots that can be easily handled by one person to large heavy pots that can only be operated from large purpose-built vessels with special handling equipment.

146

Fish Behavior near Fishing Gears during Capture Processes The first three phases of catching processes of baited fish pots involving chemoreception and food localization are similar to those used with other baited gear, such as baited hooks. Some related discussions on chemoreception and food search are given in Chapter 5.

Figure 6.4. Norwegian two-chamber pot. The pot is made of three frames covered by netting. The bottom frame is made from steel and has some weight. The other two frames are aluminum. Floats are attached to the upper panel to erect the pot when deployed. Two cone-shaped entrances lead fish into the bottom chamber. A wedgeshaped entrance allows fish to enter the upper chamber. Two collapsed pots are seen next to a 12-m-long GRP pot vessel. (Photograph: B. Thomsen, Faroe Islands.)

6.3 FISH BEHAVIOR IN RELATION TO POTS Pots are passive fishing gear and catching success will depend on their ability to attract fish, lure the fish inside, and keep them in captivity until the pot is retrieved (hauled). Fish behavior in relation to pots may be described according to senses of fish influenced by the pot itself as well as additional “long-range” stimulation, such as bait. In tropical coral reef fisheries, pots were traditionally deployed unbaited (Earle 1889), whereas in most other pot fisheries, bait is used to increase the area within which fish may react and be attracted to the pot. The behavior involved in the catching process of baited fish pots can be divided into several phases: • • • • • •

Behavior before stimulation Arousal to the presence of bait Localization of the food odor source Approaching the pot Ingression and entrapment Capture or possible escape

6.3.1 Attraction Attraction to unbaited pots can involve a multitude of interacting factors such as exploratory behavior, adoption of the pot as a shelter or residence, intraspecific social behavior, and predator–prey interactions. In pot fishing on coral reef fish assemblages, several authors stress the importance of the immediate proximity of the pot to the reef (High and Beardsley 1970; Sylvester and Dammann 1972). According to Munro (1983), pots were usually set among corals and should never be more than 2 m from the nearest coral cover. Luckhurst and Ward (1985) studied the composition of pot catches at 3, 20, and 50 m from a patch reef and found that for most species, more fish were caught closest to the reef. High and Beardsley (1970), Munro et al. (1971), and Luckhurst and Ward (1985) observed the clear attraction of a number of fish species to captured conspecifics, resulting in a sharp increase of ingress after the first few individuals had entered the pot. Antillean pots have also been used with bait in coral reefs habitat. Some authors observed no effect of bait (High and Ellis 1973), whereas others saw only short-term effects with an increase in ingress rate as long as the bait lasted (Luckhurst and Ward 1985; Munro 1974). According to Furevik (1994), bait seemed essential in most pot fisheries. Valdemarsen (1977) tested pots for gadoids and found that fish had very low interest in unbaited pots compared with baited pots. Whitelaw et al. (1991) reported that a large amount of suitable bait (approximately 4 kg) was required for effective potting. Wolf and Chislett (1974), however, reduced the amount of bait from 12.5 kg to 2.5 kg per pot and found no noticeable effect on catch rates. The release rate of attractants from the bait is initially high followed by a rapid decline

Fish Pots: Fish Behavior, Capture Processes, and Conservation Issues (Løkkeborg 1990), so that the period a baited pot will “fish” efficiently is fairly short, perhaps as little as a few hours. Furevik (1994) reported that most fish arrived at the pot within first 2 hours after deployment and only a few fish arrived after 1 to 2 days. Whitelaw et al. (1991) observed an increasing number of fish in the pot within the first 3 hours with the number leveled and then decreased afterward, indicating that more fish might have escaped from the pot when the bait was depleted. The type of bait can influence catch rates and proportions of different species. For instance, the use of squid instead of herring increased catch rates of Atlantic cod, whereas only a minor difference was observed for tusk (Furevik 1994). Cole et al. (2003) used pilchards (Sardinops neopilchrdus) and paua (Hailiorrtis iris) guts as bait for blue cod and found paua gut–baited pots caught more blue cod but the fish were smaller. In addition, bycatches of other species were greatly reduced using paua guts as bait. Knowledge of the area within which fish are attracted to the pot is important when pots are used as a survey gear for fish abundance studies as well as for the determination of an optimal distance between pots in a commercial fishery. Bjordal and Furevik (1988) tested different pot spacing measures and caught 2.5 fish per pot at 37-m spacing compared with 3.4 fish per pot at 74-m spacing. These differences may be related to fish density and bait competition. Conners et al. (2004) found that there were indications of competition between pots spaced at 556 m apart. 6.3.2 Fish Behavior when Approaching the Pot When fish arrive at the pot, their behavior will be influenced by “short-range” senses such as vision and lateral line stimulation. Luckhurst and Ward (1985) caught a very small number of fish in the large-mesh pots compared with the number caught in the small-mesh pots (even if the fish were large enough to be retained by the large meshes) and speculate that the large-mesh pots have a relative weak visual image and that many fish may simply not respond to their presence in the same way as they do to the small-mesh pots.

147

High and Beardsley (1970) observed interspecies differences in behavior around fish pots. Groupers were consistently found as solitary individuals and approached the pot with caution, while schooling species entered the pot as a group (squirrelfish and goatfish) or independently (parrotfish, bigeyes) and paired fish (butterflyfish and some parrotfish) followed their mates readily into the pots. In the North Atlantic, Furevik (1994) reported qualitative species-dependent behavior patterns vis-à-vis baited pots. Atlantic cod and ling exhibited search behavior and would occasionally butt against the net. Tusk and catfish (Anarhichas lupus) appeared to approach the pot more slowly. Haddock (Melanogrammus aeglefinus) also seemed to be more careful in their approach to the pot than did Atlantic cod and ling. Almost all fish approached baited pots from downstream, where the bait plume had dispersed (Furevik, 1994). Zigzag swimming behavior was often seen (B. Thomsen, personal observations), which seemed to improve the locationing of the bait. Most fish attracted to a baited pot stayed fairly close to the pot in the downstream area independent of the position of the entrance relative to the current (Furevik, 1994). Only a few fish were sufficiently motivated to enter the pot to reach the bait inside. On several occasions, Atlantic cod were seen to take ownership of the pot and to guard the entrance by chasing other fish away from the pot (B. Thomsen, personal observations). As fish arrive at baited pots from downstream, it is obvious that pot entrances should face downstream so that fish can more easily reach the entrance. Several experiments confirmed that the maximum catch efficiency was realized when the pot entrance was oriented downstream (Furevik 1994; Furevik et al. 2008; Valdemarsen et al. 1977). To pursue the idea of an entrance facing downstream, Whitelaw et al. (1991) constructed a rotational pot design as an alternative to the cylindrical pot used in Western Australia, but they did not achieve significantly higher catches. In Norway, better catch rates were obtained in pots anchored at one end and floated off the bottom, which allows the pot to turn in response to changes in current so that the entrance is always facing downstream (Bjordal and Furevik 1988; Furevik et al. 2008).

148

Fish Behavior near Fishing Gears during Capture Processes

6.3.3 Entrance Design and Ingress/Egress Behavior Several examples are cited in the literature of large numbers of fish arriving at a pot but ingress numbers remaining low. Valdemarsen et al. (1977) caught only 16 of 1033 gadoids that entered the potting area and concluded that pots catch only a small proportion of the fish that come into contact with the gear. Hirayama et al. (1999) studied the fishing mechanism of pots by observing the behavior of puffer fish (Lagocephalus wheeleri) in and around pots in the sea and in tanks and concluded that catch efficiency was 2%. Rose et al. (2005) used a high-frequency imaging acoustic camera to observe the behavior of sablefish and Pacific halibut (Hippoglossus stenolepis) around baited fish pots. The pot caught 9 and 10 sablefish in two sets, where sablefish were observed entering the acoustic camera field 2000 to 5000 times. Cole et al. (2004) made continuous videorecordings of entries and exits from blue cod pots and found that less than 8% of approaches to pot entrance during 30-min sets led to pot entries and that 34% of the blue cod that entered were able to escape the pot before it was hauled. Munro (1974) studied coral fish behavior in Antillean pots and found ingress rates to be fairly constant over time, whereas egress was found to be a fixed proportion (around 12% per day) of the number of fish inside the pot. From his observations, Munro derived a model predicting that the catch would approach an asymptote when egress equals ingress. The motivation of a fish to enter a pot may depend on the state of the fish as well on environmental variables. It has been shown that hunger and light level alter the responses of fish to bait (Stoner 2003) and activity and feeding motivation of the fish may be affected by environmental factors, such as temperature (Stoner et al. 2006). Munro et al. (1971) found fish ingress to be dependent on the phase of the moon and the corresponding tidal rhythms. They reported the highest rates of ingress during the new and full moon. Dalzell and Aini (1992), however, found a single peak during the full moon. It is also reported that daytime catches are higher than nighttime catches (High and Beardsley 1970; High and Ellis 1973).

The critical phase in pot fishing is when fish move into the entrance area (Furevik 1994). The design of the entrance is thus crucial to the fishing success of the pot. When species enter, some are more cautious than others. Atlantic cod and catfish may push aside part of a net panel at the entrance to enter the pot, whereas haddock and ling actively search, but they may turn if resistance is met (Furevik 1994). Fuwa et al. (1995) studied the behavior of puffer fish in the entrance funnel of a pot and derived a selection curve for the pot that depended on the ratio between fish length and the width of the entrance. The entrance design may therefore affect which species and what sizes of fish are caught. In Antillean pots, the conical entrance is usually made of chicken wire and terminates in a funnel opening that faces downward like a horse’s neck. If these pots are placed upside down, they catch only few fish (Sylvester and Dammann 1972). Luckhurst and Ward (1985) tested pots with both straight and horse-neck funnels. In pots with horse-neck funnels, there was a steady buildup of fish inside the pot over 7 days, whereas with a straight funnel, the numbers of fish inside the pot fell after 2 to 4 days. They found horse-neck funnels to be very effective in reducing escape among the species that were sampled in their study. Several authors have described experiments aimed at increasing pot efficiency by optimizing the entrance design. Bjordal and Furevik (1988) tested entrances made of net panels and shaped like a wedge pointing into the pot. They tested horizontal and vertical net panels and found no difference in Atlantic cod and tusk catches. Li et al. (2006) tested responses of arabesque greenling to pot entrance design in a tank experiment using different lengths and inclination angles of wedge-shaped entrance panels. They found an inclination angle of 34 degrees to be most efficient, whereas the effect of funnel length was less conclusive. When pots are equipped with a wider entrance, the ingress rate increases but so does the escape rate (Furevik 1994). Munro (1974) concludes that in Antillean pots the rate of escape is the main determinant of catch rates and the development of effective nonreturning devices fitted to the entrance funnels will substantially enhance catch efficiency.

Fish Pots: Fish Behavior, Capture Processes, and Conservation Issues However, nonreturn devices often reduce the rate of entry into the pot and thus reduce catch rates to below those obtained without such devices (Munro 1972). One simple nonreturn device is the use of triggers. These are simple fingers of metal or plastic that may be easily pushed inward when fish enters but are not able to bend outwards. High and Ellis (1973) found that fewer fish entered when triggers were fitted but the escape rate was much higher without triggers. Hughes et al. (1970) modified king crab pots for sablefish and found that triggers made the pots more efficient. Carlile et al. (1997) found an approximately 10-fold increase in Pacific cod catches when retention triggers were fitted to crab pots. In baited pots, the importance of triggers may increase after the bait is fully consumed (Salthaug 2002). Another way to improve entrance efficiency is to use two successive entrances, sometimes called a parlor pot. Usually the outer entrance is relatively large while the inner entrance is much smaller. Furevik and Løkkeborg (1994) tested several entrance designs and found that pots with two successive entrances yielded catches of Atlantic cod that were three times as high as those for singleentrance pots but found no difference for tusk. The optimum entrance design depends on the soak time (Furevik 1994). Sheaves (1995) found higher catch rates at short soak times with simple entrances, whereas no difference was found between short and long soak durations when a complex entrance design was used. In baited pots, fish ingress rates are initially high. The number of fish inside the pot may reach a peak around the time when the bait is exhausted (Whitelaw et al. 1991). In baited pots for blue cod, Cole et al. (2004) found the optimal soak time to be 30 to 40 min. Some species are more adept than others at escaping (Munro 1983), and catch composition may therefore change with soak time. 6.3.4 Fish Behavior inside the Pot Behavior of fish inside a pot is species specific, and within the species individual specimens may also behave differently. After they enter the pot, most fish appear to be calm and mill around, although a few immediately display aggressive behavior and try vigorously to find an escape route. Furevik

149

(1994) found that most fish (Atlantic cod, haddock, tusk, and ling) were more active when they first entered the pot and frequently butted the net. They become less active with time inside the pot. In baited pots, some fish show interest in the bait for a short time (Furevik 1994) but soon lose interest. Luckhurst and Ward (1985) also found that most fish ignored the bait after entering the pot. Luckhurst and Ward (1985) found that many fish display butting behavior after entering the pot, causing abrasion of the nape and laceration of pectoral and pelvic fins, leading to secondary infection by fungi and/or bacteria. The most vulnerable were Scaridae, which were dead or dying within 2 to 3 days of entering pots. Munro et al. (1971) found that almost all fish that were trapped for up to 2 weeks displayed signs of physical deterioration or abrasions with secondary fungal infections. Cooke et al. (1998) also reported fungal growth on potted fish. However, Whitelaw et al. (1991) observed no mortality of fish inside pots even after extended periods of detention. Fish inside pots may escape not only through the entrance but, if they are small enough, also through the meshes of the cover whether made of synthetic netting or wire mesh. Mesh size is a determinant of catch rates and fish size at which fish recruit to pot fisheries (Mahon and Hunte 2001). Stewart and Ferrell (2002) found that size selection in rigid welded wire mesh could be accurately predicted by the maximum body depth of the fish and the maximum mesh aperture. However, Luckhurst and Ward (1985) found escape not to be simply a function of body depth. Some fish species turn on their side to escape through meshes, whereas others do not. In Antillean fish pots, Gobert (1998) found a shift in the selection curve, depending on average catch per pot. He concluded that the squeezing behavior of fish was density dependent. Aggressive behavior has been observed with larger individuals chasing smaller individuals and frequently predating on smaller species. In pots containing conger eels (Conger verreauxi), blue cod either attempted to escape or were consumed (Cole et al. 2001). The activity of fish inside the pot may affect fishing efficiency of the pot. High and Beardsley (1970) observed a “saturation effect” as the rate of

150

Fish Behavior near Fishing Gears during Capture Processes

entry dropped sharply when the fish inside the pot reached a certain number. It appeared that the presence of a large number of fish inside the pot scared off other fish in the area. It has been observed that larger pots usually have higher catch rates for most species (Collins 1990). The chance of finding the entrance for escaping from pots is inversely proportional to the area or volume of the pot if the sizes of the entrances are equal (Munro 1974). 6.4 CONSERVATION CHALLENGES AND SOLUTIONS Pots are generally regarded as an environmentally friendly fishing gear, with few undesirable side effects when catching target fish species. That does not mean that there are no negative conservation consequences in pot fisheries. Conservation challenges include issues related to lost gear and ghost fishing, escape and discard mortality, incidental megafauna interactions, and specific issues related to habitat alteration in vulnerable habitats. The prospects of solving these and other potential problems are good, given the inclusion of conservation characteristics in gear development and appropriate management of fisheries in sensitive areas. By using pots, access to fishery resources can probably be retained while benthic species and habitats remain protected. 6.4.1 Undersized and Nontarget Species: Catch Avoidance, Escapees, and Discards Problems related to capture of undersized or nontarget species occur frequently with most fishing gear. Means to solve these challenges in pot gears include improving selectivity by preventing unwanted organisms from entering the pots in the first place, providing effective escape opportunities, or optimizing capture and handling techniques for live release of discards. The most appealing way of solving potential discard and escape mortality is to find methods of preventing unwanted species from locating and entering the pots. Furevik et al. (2008) found that by floating pots off the bottom, catches of Atlantic cod increased, whereas bycatch of red king crab (Paralithodes camtschaticus) was eliminated. Zhou and Kruse (2000) found that horizontal excluders and slick-board ramps both had the potential to

reduce tanner crab (Chionoecetes bairdi) bycatch in Pacific cod pots. Carlile et al (1997) evaluated seven different pot modifications and found that the installation of halibut excluders in standard pots increased Pacific cod catch and reduced Pacific halibut bycatch. Bait selection is another way to reduce unwanted catch, because it can influence the catch rate and the proportions of different species and size classes. The use of squid as bait instead of herring increased the catch rates of Atlantic cod but not tusk (Furevik 1994). Cole et al. (2003) obtained larger catches of blue cod and greatly reduced bycatch species by using paua guts instead of pilchards as bait. Pedersen (2000) found that catch rates of Atlantic cod were higher when pots were baited with squid compared with mackerel bait in June, whereas no differences were found in October and February. He also found that in February, squid-baited pots caught more Atlantic cod larger than 40 cm than did mackerelbaited pots. Optimizing efficiency for target species and sizes may reduce effort, and thereby bycatch, in a quota-controlled fishery. Other ways to avoid unwanted bycatch may include developing speciesspecific deterrents or shorter soak times to avoid slow-moving scavenging invertebrates or simply avoiding problem areas and time periods. By far the most frequently used method to achieve a “cleaner catch” is to mount specific escape rings, panels, or vents to allow unwanted catches to escape after they have entered the pots (e.g., Pedersen 2000; Shepherd et al. 2002; Stewart and Ferrell 2002). Selecting web mesh size (Luckurst and Ward 1985; Robichaud et al. 1999; Sary et al. 1997) in accordance with minimum landing size, for example, is another option, which offers numerous escape routes. Successful reduction of bycatch mortality by improving selectivity relies heavily on the assumption that escapees are unharmed. If they are not, increasing the opportunity for escape may result in a higher level of unaccounted fishing mortality (Chopin and Arimoto 1995). There is currently little information on the survival of escapees from pots. Signs of eagerness to escape may range from subtle pushing against the net to squeezing through meshes (Robichaud et al. 1999), a behavior that may be exaggerated in the presence of predatory fish

Fish Pots: Fish Behavior, Capture Processes, and Conservation Issues (Hartsuijker and Nicholson 1981). Laboratory studies have shown that the simple passage of a fish through a netting mesh or other selective device does not necessarily inflict fatal injury (see Chapter 11). If undersized and nontarget species cannot be avoided or do not escape, they may be discarded after being brought on board (e.g., Stewart and Ferrell 2003). Due to the benign nature of the fish capture process, it is expected that the mortality of fish discarded from pots may be low as the catch is usually alive, with low injury rates (Nøstvik and Pedersen 1999) and low capture-related stress (Pilling et al. 2001). Survival rates of discarded bycatch can be inferred indirectly from tagging studies. When cod were caught for tagging, Nøstvik and Pedersen (1999) found that fish pots had one of the lowest injury rates, the highest tagging percentage, and the highest recapture rates, compared with fyke nets, hand line, pelagic trawl, and bottom trawl. Although survival rates of discards in pot fisheries are probably among the highest, passage up through the water column and handling on deck pose several additional threats to survival, including barotrauma, thermal shock, air exposure, physical impact, and predator exposure (Davis 2002). Returning to the depth after discard may also increase vulnerability of predation. For these reasons alone, discards should be avoided. 6.4.2 Lost Gear and Ghost Fishing Ghost fishing is defined as the ability of fishing gear to continue fishing after control of the gear has been lost by the fisherman (Smolowitz 1978). Most studies on ghost fishing by lost pots have been performed in the crab and lobster fisheries. Only a few deal with fish pots, although problems and mitigation measures in the different fisheries may overlap. Moored static gears may be lost for several reasons. Strong currents may force buoys down to depths at which they collapse. End markers may be cut off by propellers. Groundlines and buoy lines may be caught on rough bottoms or otherwise break by wave action during retrieval in bad weather. Whole fleets of pots may be displaced from their original positions due to currents or gear conflicts. Pots are often rigid structures made of strong materials and are likely to maintain their configuration after they

151

are lost. They may therefore continue to catch and potentially kill captured organisms after being lost. Baited pots lose most of their catching efficiency for target species when their bait has been consumed or the release of feeding attractants ends. The potential for continued catches of target species beyond this initial catch is low. Captured species, however, are inadequate evidence to prove ghost fishing mortality. If fish are able to escape and dead bodies cannot be found, the ghost fishing mortality has been defined as zero (Matsuoka et al. 2005). If entrapped fish cannot escape they will eventually die due to starvation or injury resulted from behaviors such as bumping on the webbing inside pots (Bullimore et al. 2001; Matsuoka et al. 2005). Captured fish that die inside the pot may re-bait the pots, enhancing ghost fishing capability. Selfbaiting through the death and decay of trapped fish may attract and capture scavenging species (especially crustaceans), which also may be trapped and die. In the worst case scenario, there may be a continuous cycle of capture, decay, and attraction of scavenging species for as long as the gear remains intact (Carr et al. 1990). In the case of unbaited pots, the capture of species does not rely on olfaction, and they may continue to fish target species and nontarget species for a long period of time. Matsuoka et al. (1997) found that some finfish pots continued ghost fishing for as long as 3 years. No general conclusions can be drawn from published studies regarding ghost fishing mortality in fish pots, simply because published studies are few and differ in many aspects such as area, species, type of pots, circumstance of loss, and study methods. One of the few studies of ghost fishing in fish pots found that, depending on fish density, fish learned to escape from pot openings or bar gaps after 10 days (Ayaz et al. 2006). After 24 days, no fish were observed in the pots. The pots were collapsed after 6 months when the study was terminated. In another field study quantifying catch rates of simulated lost fish pots at five traditional fishing grounds near Muscat and Mutrah, Sultanate of Oman (Al-Masroori et al. 2004), escape rates were low even after long soak times, with an estimated ghost fishing mortality of 95%. However, the adverse effects of fish-pot ghost fishing may easily be solved. Decomposition of the

152

Fish Behavior near Fishing Gears during Capture Processes

gear itself may not be a desirable option in reducing the catching capacity of a lost gear because of the longevity of the materials used. However, escape can be enhanced by mounting degradable panels/ threads and/or galvanic time-release mechanisms (GTRs) into the pots, allowing escape after a predetermined time (Blott 1978; Breen 1990; Scarsbrook et al. 1988; Wyman 1996). Scarsbrook et al. (1988) tested the effectiveness of various escape mechanisms in conical pots used in the sablefish fishery and found that square or triangular panels were more effective than just a “slash” secured with biodegradable twine and that these measures virtually eliminated ghost fishing. Such measures are simple and can be quick and easy to service once installed (Breen 1990). These mitigation measures should be considered if ghost fishing is a problem in both baited and unbaited pots. Retrieval of lost gear might be the ultimate means to reduce ghost fishing (Stevens et al. 2000). Retrieval is important, not only as a means of reducing mortality but also because lost pots constitute a littering problem (Edyvane et al. 2004; Hess et al. 1999; Lee et al. 2006) and the risk of losing more gear will increase as previously lost gear poses new threats to snagging and further losses of gear. Due to a certain reluctance to report losses and to the inherent difficulty of performing quantitative surveys of lost fishing gear, quantitative estimates of lost gears vary considerably (Bullimore et al. 2001). 6.4.3 Megafauna Interaction Incidental interaction between pot gears and megafauna such as whales and turtles seems to be mostly related to groundlines and buoy lines. In some pot fisheries that involve endangered species, such entanglements are of utmost concern and may be a significant part of the mortality of the species. In some other cases, the use of pots instead of other static gears is encouraged to avoid predation on bait or captured target species by birds and marine mammals. In a study by Johnson et al. (2005), entanglements of endangered right whales (Eubalaena glacialis) and humpback whales (Megaptera novaeangliae) were analyzed to determine the

types and parts of gear that were involved. Of all the entanglements, 89% were attributed to pot and gill-net gears, and when the gear part could be identified, 81% involved entanglements in buoy lines and/or groundlines. Consequently, reducing the number of lines in the water column, for example, by using sinking or neutrally buoyant groundline could reduce the risk of entanglement. The risk presented by buoy and surface system lines might be mitigated by a suitable insertion of weak links with sufficiently low breaking strengths. Another solution to reduce the number of buoy lines may be to set fleets of multiple pots instead of setting them singly. The efficacy of such mitigating measures remains to be demonstrated. Pots and their associated ropes may also entangle and drown marine turtles. Dayton et al. (1995) speculated that leatherback turtles (Dermochelys coriace) may mistake marker buoys for jellyfish and become entangled in buoy lines. Sewell and Hiscock (2005) performed a thorough review of marine turtle bycatch in the United Kingdom and Ireland. The leatherback turtle is the only species of turtle that is significantly affected by fisheries in U.K. and Irish waters. Although the global significance of bycatch in this area is not fully known, measures to reduce the impact on the declining global population are encouraged. 6.4.4 Habitat Alteration The potential for habitat disturbance due to potting is low compared with that for mobile fishing gear (Jennings and Kaiser 1998). Pots are static gears, and each pot thus affects an area of seabed of the same order of magnitude as the footprint of the gear itself. Furthermore, the forces on the seabed can be mainly attributed to the weight of the gear, which is usually made as light as possible to minimize production costs and provide ease of handling. At present, the spatial distribution of pots is also limited, and the few reported disturbances of potential significance have been in areas of erect fragile coral epifauna, areas that are probably vulnerable to all bottom-contact gears (Mortensen et al. 2005). In fact, the only long-term negative effect of pot fishing on habitat is based on anecdotal information from British Columbia, where red tree corals were reported to have disappeared in an area where

Fish Pots: Fish Behavior, Capture Processes, and Conservation Issues prawn pots were set, because corals became entangled in the mesh of the pots (Risk et al. 1998). Pots are usually regarded as environmentally friendly, and their use is encouraged to retain access to fish resources while protecting benthic species and habitats (Blyth et al. 2004; Kaiser et al. 2000; Mangi and Roberts 2006; Risk et al. 1998). Such habitats indeed sometimes serve as “de facto” reference or control areas that are subjected to minimal disturbance (Kaiser et al. 2000). Stone (2006) found that the greatest potential for disturbance existed when pots are dragged along the bottom during retrieval. He studied the coral habitat in the Aleutian Islands of Alaska with regard to depth distribution, fine-scale species association, and fisheries interactions in areas fished by bottom trawls, longlines, single-set pots, and fleet-set pots by means of video transect analysis. Only one coral site was found where fleet-set crab pots were involved and no sites where single-set pots were used. Fleet-set pots with ropes connecting the adjacent pots cover larger areas and thus come in contact with more epifauna. They may therefore cause more damage than single-set pots. At the one site where disturbance was observed, the seabed was scored to the bare substrate, but no information was offered as to whether any fauna were damaged. Although this finding demonstrates that pots may cause habitat alterations, the study also showed that the frequency of damage by this gear is extremely low (a single incident in 25 video transects, or 0.9% of the total length of video transects). Further, dragging of pots and lines represents an abnormal situation that fishermen strive to avoid due to the risk of losing or damaging their gears. Fishermen normally try to minimize drag forces, because connecting ropes cannot withstand high dragging forces, due to the friction of the pots and moorings being towed along the seabed. Instead, hauling is carried out in such a way that each pot sits in its deployed position on the bottom until it is lifted vertically off the bottom. Precarious situations may arise when strong wind and current conditions force the vessel in the opposite of intended direction or off track, or produce a lifting angle facilitating bottom drag, or when fishing takes place at great depths with difficult positioning or in unpredictable areas with steep irregular bathymetry (Stone 2006). In such areas, it

153

is often necessary to use heavy moorings. This component intuitively poses a threat through direct contact with fauna on deployment (Eno et al. 2001), but the numbers and amount of biomass affected will be relatively low, even with a fishing effort far beyond current and prospective levels. Eno et al. (2001) examined the effects of fishing with crustacean pots on benthic species. They demonstrated few or no immediate effects on several species that are perceived to be sensitive to mechanical disturbance. Notably, sea pens (Penatula phosphorea, Virgularia mirabilis, and Funiculina quadrangularis) bent away due to the pressure wave exerted by the dropping pot, which effectively prevented direct contact with their tips. After smothering and even uprooting, they reestablished themselves when they regained contact with the muddy substrate. Other than damage sustained by large individual Ross corals (Pentapora foliacea), the short-term effects of pot fishing on sensitive benthic species did not appear to be detrimental. The environmental impacts of artisanal fishing gear on coral reef ecosystems were studied in the multigear fishery of southern Kenya (Mangi and Roberts 2006). No direct evidence was offered regarding physical damage by pots to corals. However, coral heads were removed to hold pots on the bottom during fishing and therefore had direct effects on the reef. However, sustainability of coral reef is more likely to be achieved by reducing the use of beach seines and spears, while maintaining the use of pots, gillnets, and handlines, which cause the least damage. Blyth et al. (2004) examined benthic communities at sites within and adjacent to the U.K. Inshore Potting Agreement (IPA) area, which had been subjected to four different commercial fishing regimens (one static gear regimen and three towed gear regimens) since 1978. Significantly greater total species richness and biomass of benthic communities were reported at sites under static gear regimens than those under towed gear regimens. The benthic community biomass under static gear regimens was also significantly greater than that under all other regimens. Similar findings were also reported from the same IPA area by Kaiser et al. (2000), who found that communities within areas closed to towed fishing gears were dominated by higher biomass

154

Fish Behavior near Fishing Gears during Capture Processes

and emergent epifauna that increased habitat complexity, whereas towed areas were dominated by smaller-bodied fauna and scavenging taxa. 6.5 CONCLUDING REMARKS Fish pots have a long history as a fishing gear and are widely used in many parts of the world with a wide variety of designs and modes of operations. However, the volume of fish captured with pots is small and catch efficiency is low in comparison with other gear. Pots possess several superior and appealing characteristics compared with other fishing gear: low labor and energy use, minimal habitat impact, and live fish delivery. Growing concern regarding the sustainability of the marine environment and growing support for responsible utilization of natural resources may be incentives to revive the use of pots as an alternative fishing method. However, fish pots need further development to increase catch efficiency. Most observations on fish behavior in relation to catching mechanisms have identified the entrance as the most important factor in improving the efficiency. Effective bait is also a major contributor to fishing success. The release rate of odor from the bait decreases rapidly, and a system that prolongs the release of bait odor would produce substantial advantages. One potential adverse effect of pot fishing is ghost fishing by lost pots, which could probably be solved by mounting timed-release systems. In comparison with most other gears, fish pots have more favorable characteristics than drawbacks. Fish pots might therefore be developed into a viable alternative to other fishing gear in many fisheries and even become the preferred fishing gear in the future. REFERENCES Al-Masroori H, Al-Oufi H, McIlwain JL and McLean E. 2004. Catches of lost fish traps (ghost fishing) from fishing grounds near Muscat, Sultanate of Oman. Fish. Res. 69: 407–414. Ayaz A, Ozekinci U, Altinagac U and Ozen O. 2006. An Investigation of Ghost Fishing of Circular Fish Traps used in Turkey. Presented at the ICES-FAO WGFTFB meeting, April 2006, Izmir, Turkey. Bekker-Nielsen T. 2002. Nets, boats and fishing in the Roman world. Classica et Mediaevalia. Vol. 53.

Bjordal Å. 1993. Capture techniques for wrasses (Labridae). ICES CM. 1993/B: 22. Bjordal Å and Furevik DM. 1988. Full scale fishing trials for tusk (Brosme brosme) and cod (Gadus morhua) with collapsible fish trap. ICES CM. 1988/B: 33. Blott AJ. 1978. A preliminary study of timed release mechanisms for lobster traps. Mar. Fish. Rev. 40: 40–49. Blyth RE, Kaiser MJ, Edwards-Jones G and Hart PJB. 2004. Implications of a zoned fishery management system for marine benthic communities. J. Appl. Ecol. 41: 951–961. Breen PA. 1990. A review of ghost fishing by traps and gillnets. Proceedings of the Second International Conference on Marine Debris, 2–7 April 1989, Honolulu, Hawaii. pp 571–599. Bullimore BA, Newman PB, Kaiser MJ, Gilbert SE and Lock KM. 2001. A study of catches in a fleet of “ghost-fishing” pots. Fish. Bull. 99: 247–253. Carlile DW, Dinnocenzo TA and Watson LJ. 1997. Evaluation of modified crab pots to increase catch of Pacific cod and decrease bycatch of Pacific halibut. N. Am. J. Fish. Manag. 17: 910–928. Carr HA, Amaral EH, Hulbert AW and Cooper R. 1990. Under-water survey of simulated lost demersal and lost commercial gill nets off New England. In: Coe JM and Rogers DB (eds). Marine Debris: Sources, Impacts and Solutions. pp 171–186. New York: Springer. Chopin FS and Arimoto T. 1995. The condition of fish escaping from fishing gears: a review. Fish. Res. 21: 315–327. Cole RG, Tindale DS and Blackwell RG. 2001. A comparison of diver and pot sampling for blue cod (Parapercis colias: Pinguipedidae). Fish. Res. 52: 191–201. Cole RG, Alcock NK, Tovey A and Handley SJ. 2004. Measuring efficiency and predicting optimal set durations of pots for blue cod Parapercis colias. Fish. Res. 67: 163–170. Cole RG, Alcock NK, Handley SJ, Grange KR, Black S, Cairney D, Day J, Ford S and Jerrett AR. 2003. Selective capture of blue cod Parapercis colias by potting: behavioral observations and effects of capture method on peri-mortem fatigue. Fish. Res. 60: 381–392. Collins MR. 1990. A comparison of three fish trap designs. Fish. Res. 9: 325–332. Conners ME, Munro P and Neidetcher S. 2004. Pacific cod pot studies 2002–2003. AFSC Processed Report 2004.

Fish Pots: Fish Behavior, Capture Processes, and Conservation Issues Cooke SJ, Bunt CM and McKinley RS. 1998. Injury and short-term mortality of benthic stream fishes: a comparison of collection techniques. Hydrobiologia. 379: 207–211. Dalzell P and Aini JW. 1992. The Performance of Antillean wire mesh fish traps set on coral reefs in Northern Papua New Guinea. Asian Fish. Sci. 5: 89–102. Davis MW. 2002. Key principles for understanding fish bycatch discard mortality. Can. J. Fish. Aquat. Sci. 59: 1834–1843. Dayton PK, Thrush SF, Agardy MT and Hofman RJ. 1995. Environmental effects of marine fishing. In: Aquatic conservation: Mar. Freshw. Ecosys. 5: 205–232. Earle EM. 1889. The fish pot of the Caribbean Sea. J. Mar. Biol. Assoc. UK. 1: 199–204. Edyvane KS, Dalgetty A, Hone PW, Higham JS and Wace NM. 2004. Long-term marine litter monitoring in the remote Great Australian Bight, South Australia. Mar. Pollut. Bull. 48: 1060–1075. Eno NC, MacDonald D, Kinnear J, Amos SC, Chapman C, Clark R, Bunker F and Munro C. 2001. Effects of crustacean traps on benthic fauna. ICES J. Mar. Sci. 58: 11–20. Furevik DM. 1994. Behavior of fish in relation to pots. In: Fernö A and Olsen S (eds). Marine Fish Behavior in Capture and Abundance Estimation. pp 28–44. Oxford: Fishing News Books. Furevik DM, Humborstad O-B, Jørgensen T and Løkkeborg S. 2008. Floated fish pot eliminates bycatch of red king crab and maintains target catch of cod. Fish. Res. 92: 23–27. Furevik DM and Løkkeborg S. 1994. Fishing trials in Norway for Torsk (Brosme brosme) and Cod (Gadus morhua) using baited commercial pots. Fish. Res. 9: 219–229. Furevik DM and Skeide RL. 2003. Fiske etter torsk (Gadus morhua), Lange (Molva molva) og brosme (Brosme brosme) med tokammerteine langs norskekysten. Bergen: Institute of Marine Research. Fuwa S, Ishizaki M, Sako K and Imai T. 1995. A catching model of fish trap for puffer. Nippon Suisan Gakkaishi. 61: 356–362. Gabriel O, Lange K, Dahm E and Wendt T. 2005. Von Brandt’s Fish Catching Methods of the World. 4th ed. Oxford: Fishing News Books and Blackwell Publishing. 523 pp. Gobert B. 1998. Density-dependent size selectivity in Antillean fish traps. Fish. Res. 38: 159–167. Hanselman DH, Lunsford CR, Fujioka T and Rodgveller CJ. 2006. Alaska Sablefish Assessment

155

for 2007. NPFMC Bering Sea, Aleutian Islands and Gulf of Alaska SAFE, Alaska Sablefish: 341–428. Hartsuijker L and Nicholson WE. 1981. Results of a potfishing survey on Pedro Bank (Jamaica): The relations between catch rates, catch composition, the size of fish and their recruitment to the fishery. FAO/TCO/JAM 8902: Potfishing survey on Pedro Bank. Hess NA, Ribic CA and Vining I. 1999. Benthic marine debris, with an emphasis on fishery-related items, surrounding Kodiak Island, Alaska, 1994– 1996. Mar. Pollut. Bull. 38: 885–890. High WL and Beardsley AJ. 1970. Fish behavior studies from an undersea habitat. Comm. Fish. Rev. 32(10): 31–37. High WL and Ellis IE. 1973. Underwater observations of fish behavior in traps. Helgolander wiss. Meeresunters. 24: 341–347. Hirayama M, Fuwa S, Ishizaki M and Imai T. 1999. Behavior of puffer Lagocephalus and the fishing mechanism of the pot trap. Nippon Suisan Gakkaishi. 65: 419–426. Hughes SE, Worlund DD and Hipkins FW. 1970. Adaption of king crab pots for capturing sablefish (Anoplopoma fimbria). J. Fish. Res. Bd. Can. 27: 1747–1755. Jennings S and Kaiser MJ. 1998. The effects of fishing on marine ecosystems. Adv. Mar. Biol. 34: 201–352. Johnson A, Salvador G, Kenney J, Robbins J, Landry S and Clapham P. 2005. Fishing gear involved in entanglements of right and humpback whales. Mar. Mam. Sci. 21: 635–645. Kaiser MJ, Spence FE and Hart PJB. 2000. Fishinggear restrictions and conservation of benthic habitat complexity. Conserv. Biol. 15: 1512–1525. Lee D-I, Cho H-S and Jeong S-B. 2006. Distribution characteristics of marine litter on the sea bed of the East China Sea and the south of Korea. Estuarine, Costal and Shelf Science 70: 187–194. Li Y, Yamamoto K, Hiraishi T, Nashimoto K and Yoshino H. 2006. Behavioral responses of arabesque greenling to trap entrance design. Fish. Sci. 72: 821–828. Løkkeborg S. 1990. Rate of release of potential feeding attractants from natural and artificial bait. Fish. Res. 8: 253–261. Luckhurst B and Ward J. 1985. Behavioral dynamics of coral reef fishes in Antillian fish traps at Bermuda. Proc. Gulf. Carib. Fish. Inst. 38: 528–546. Mahon R and Hunte W. 2001. Trap mesh selectivity and the management of reef fishes. Fish Fish. 2: 356–375.

156

Fish Behavior near Fishing Gears during Capture Processes

Mangi SC and Roberts CM. 2006. Quantifying the environmental impacts of artisanal fishing gear on Kenya’s coral reef ecosystem. Mar. Pollut. Bull. 52: 1646–1660. Matsuoka T, Nakashima T and Nagasawa N. 2005. A review of ghost fishing: scientific approaches to evaluation and solutions. Fish. Sci. 71: 691–702. Matsuoka T, Osako T and Miyagi M. 1997. Underwater observation and assessment on ghost fishing by lost fish-traps. Proc. Asian Fish. Forum 4: 179–183. Mortensen PB, Mortensen LB and Gordon DC Jr. 2005. Effects of fisheries on deepwater gorgonian corals in the Northeast Channel, Nova Scotia. Am. Fish. Soc. Symp. 41: 369–382. Munro JL. 1972. Large volume stackable fish trap for offshore fishing. Proc. Gulf Carib. Fish. Inst. 25: 212–228. Munro JL. 1974. The mode of operation of Antillean fish traps and the relationships between ingress, escapement, catch and soak. ICES J. Mar. Sci. 35: 337–350. Munro JL. 1983. The composition and magnitude of trap catches in Jamaican waters. In: Munro JL. (ed). Caribbean Coral Reef Fishery Resources. ICLARM Studies and Reviews. No. 7. Manila, the Philippines: International Center for Living Aquatic Resources Management. Munro JL, Reeson PH and Gaut VC. 1971. Dynamic factors affecting the performance of the Antillean Fish Trap. Proc. Gulf Carib. Fish. Inst. 23: 184–194. Nøstvik F and Pedersen T. 1999. Catching cod for tagging experiments. Fish. Res. 42: 57–66. Pedersen K-A. 2000. Effekter av agntype, maskevidde og settetidspunkt på fangsteffektivitet og størrelsessammensetning av torsk i fiske med teiner. University of Tromsø, Norway (in Norwegian). Pilling GM, Purves MG, Daw TM, Agnew DA and Xavier JC. 2001. The stomach contents of Patagonian toothfish around South Georgia (South Atlantic). J. Fish Biol. 59: 1370–1384. Pol M, Walsh P and Marcella R. 2005. Cod potting in Massachusetts. A report submitted to Northeast consortium. Available online: http://northeastconsortium. org/ProjectFileDownload.pm?report_id=661 &table=project_report. Risk JR, McAllister DE and Behnken L. 1998. Conservation of cold and warm water seafans: Threatened ancient gorgonian groves. Sea Wind 12(1). Ocean Voice International. Robichaud D, Hunte W and Oxenford HA. 1999. Effects of increased mesh size on catch and fishing

power of coral reef fish traps. Fish. Res. 39: 275–294. Rose CS, Stoner AW and Matteson K. 2005. Use of high-frequency imaging sonar to observe fish behavior near baited fishing gears. Fish. Res. 76: 291–304. Sainsbury J. 1996. Static gear. In: Commercial Fishing Methods: An Introduction to Vessels and Gears, 3rd ed. Oxford: Fishing News Books and Blackwell Science. Salthaug A. 2002. Do triggers in crab traps affect the probability of entry? Fish. Res. 58: 403–405. Sary Z, Oxenford HA and Woodley JD. 1997. Effects of an increase in trap mesh size on an overexploited coral reef fishery at Discovery Bay, Jamaica. Mar. Ecol. Prog. Ser. 154: 107–120. Scarsbrook JR, McFarlane A and Shaw W. 1988. Effectiveness of experimental escape mechanisms in sablefish traps. N. Am. J. Fish. Manag. 8: 158– 161 Sewell J and Hiscock K. 2005. Effects of fishing within UK European marine sites: guidance for nature conservation agencies. Report to the Countryside Council for Wales, English Nature and Scottish Natural Heritage from the Marine Biological Association, Plymouth: CCW Contract FC 73-03214A. 195 pp. Sheaves MJ. 1995. Effect of design modifications and soak time variations on Antillean Z fish trap performance in a tropical estuary. Bull. Mar. Sci. 56: 475–489. Shepherd GR, Moore CW and Seagraves RJ. 2002. The effect of escape vents on the capture of black sea bass, Centropristis striata, in fish traps. Fish. Res. 54: 195–207. Slack-Smith RJ. 2001. Fishing with traps and pots. FAO Training Series 26. Rome, FAO. Smolowitz RJ. 1978. Trap design and ghost fishing: Discussion. Mar. Fish. Rev. 40: 59–67. Stevens BG, Vining I, Byersdorfer S and Donaldson W. 2000. Ghost fishing crab (Chionoecetes bairdi) pots off Kodiak, Alaska: pot density and catch determined from sidescan sonar and pot recovery data. Fish. Bull. 98: 389–399. Stewart J and Ferrell DJ. 2002. Escape panels to reduce by-catch in the New South Wales demersal trap fishery. Mar. Freshw. Res. 53: 1179–1188. Stewart J and Ferrell DJ. 2003. Mesh selectivity in the New South Wales demersal trap fishery. Fish. Res. 59: 379–392. Stone RP. 2006. Coral habitat in the Aleutian Islands of Alaska: depth distribution, fine-scale species

Fish Pots: Fish Behavior, Capture Processes, and Conservation Issues associations, and fisheries interactions. Coral Reefs. 25: 229–238. Stoner AW. 2003. Hunger and light level alter response to bait by Pacific halibut: laboratory analysis of detection, location and attack. J. Fish Biol. 62: 1176–1193. Stoner AW, Ottmar ML and Hurst TP. 2006. Temperature affects activity and feeding motivation in Pacific halibut: Implications for bait-dependent fishing. Fish. Res. 81: 202–209. Sylvester JR and Dammann AE. 1972. Pot fishing in the Virgin Islands. Mar. Fish. Rev. 34: 33– 35. Valdemarsen JW. 1977. Analysis of pot as a bottom gear and studies of some factors influencing the catch efficiency. Dept. of Fishery Biology, University of Bergen, Bergen, Norway. Valdemarsen JW, Fernö A and Johannessen A. 1977. Studies on the behavior of some gadoid species in relation to traps. ICES CM. 1977/B: 42.

157

Walsh PJ, Hiscock W and Sullivan R. 2006. Fishing trial for cod (Gadus morhua) using experimental pots. St. John’s: Fisheries and Marine Institute of Memorial University of Newfoundland. Watson R, Revenga C and Kura Y. 2006. Fishing gear associated with global marine catches I. Database development. Fish. Res. 79: 97–102. Whitelaw AW, Sainsbury KJ, Dews GJ and Campbell RA. 1991. Catching characteristics of four fish-trap types on the North-West Shelf of Australia. Aust. J. Mar. Freshw. Res. 42: 369–382. Wolf S and Chislett G. 1974. Trap fishing. Explorations for snapper and related species in the Caribbean and adjacent waters. Mar. Fish. Rev. 37: 49–61. Wyman E. 1996. Selective groundfish pots offer solution to bycatch problems. University of Alaska Sea Grant Report 96-03. Zhou S and Kruse GH. 2000. Modifications of cod pots to reduce Tanner crab bycatch. N. Am. J. Fish. Manag. 20: 897–907.

158

Fish Behavior near Fishing Gears during Capture Processes

SPECIES MENTIONED IN THE TEXT Atlantic cod, Gadus morhua arabesque greenling, Pleurogrammus azonus black sea bass, Centropristis striata blue cod, Parapercis colias catfish, Anarhichas lupus conger eel, Conger verreauxi haddock, Melanogrammus aeglefinus humpback whales, Megaptera novaeangliae ling, Molva molva leatherback turtle, Dermochelys coriace Pacific cod, Gadus macrocephalus Pacific halibut, Hippoglossus stenolepis

paua, Hailiorrtis iris pilchard, Sardinops neopilchrdus puffer fish, Lagocephalus wheeleri red king crab, Paralithodes camtschaticus right whale, Eubalaena glacialis Ross coral, Pentapora foliacea sablefish, black cod, Anoplopoma fimbria sea pens, Penatula phosphorea, Virgularia mirabilis, and Funiculina quadrangularis snapper, Pagrus auratus tanner crab, Chionoecetes bairdi tusk, Brosme brosme wrasse, Labridae

Chapter 7 Large-Scale Fish Traps: Gear Design, Fish Behavior, and Conservation Challenges Pingguo He and Yoshihiro Inoue

servation challenges including interactions of the gear with megafauna species and mitigation measures.

7.1 INTRODUCTION In a broad sense, traps can be defined as stationary fishing gears to which fish or shellfish are misled, drift into, or are attracted to an enclosure and cannot found their way out, therefore being “trapped.” In that sense, traps include large-scale stationary netting structures such as cod traps in Newfoundland, set nets in Japan, and small pots such as lobster pots. These small fish or shellfish pots are also called traps in many regions of the world. To avoid confusion, traps and pots are distinguished in this chapter. Traps here refer to large stationary fixed structures (including trap nets, set nets, pound nets, and weirs) to lead and trap fish, whereas pots refer to small-scale baited enclosures to attract and retain fish. This chapter is concerned with traps. Fish capture by baited pots is discussed in Chapter 6. Traps are stationary fishing gears set in coastal waters for capturing various fish and shellfish species. They have been called set nets (or setnets), bag nets, pound nets, stake nets, weirs, traps, and other names. In this chapter, we refer to Japanese traps as “set nets” and other traps as “traps” or other names to coincide with literature. Here we discuss worldwide trap fisheries and fish behavior patterns and exemplify specific fish behavior near Newfoundland cod traps, Japanese set nets, and Baltic traps for salmon and whitefish. The chapter also describes species and size selectivity and con-

7.2 TRAP FISHERIES AND TRAP DESIGNS Traps are used all over the world’s marine and freshwater environment. Traps are fuel-efficient compared with many other fishing gears (Nomura 1980). Fish trapped in the net are usually alive before harvested and allow for selective harvesting and live release. Japan may be the leading nation in the use of the gear and research activities on the subject. In 1986, there were more than 1600 large-scale set nets and 15,000 medium-size and small-scale set nets in use in Japan (Inoue and Arimoto 1989). Annual landing by set nets in later 1980s was about 650,000 metric tons, about 25% of the total landing from coastal waters in Japan. Another area where traps are widely used is eastern Canada, particularly Newfoundland and Labrador. Traps in Newfoundland are primarily used for Atlantic cod (Gadus morhua) but also Atlantic mackerel (Scomber scombrus), Atlantic herring (Clupea harengus), and capelin (Mallotus villosus). In 1920s, there were reported 7365 cod traps in use in the area. In 1989, just before the cod

159

160

Fish Behavior near Fishing Gears during Capture Processes

moratorium, there were approximately 4000 cod traps around the province of Newfoundland and Labrador (He and Nemoto 1999) and accounted for 57% of the cod landings at that time. Cod traps virtually disappeared in 1992 when the Northern cod was closed to commercial fisheries. 7.2.1 Early Traps Traps are one of the oldest commercial fishing gears in the world. Stonewalls constructed to trap fish in tidal waters, estuaries, and rivers are reported to date back to Neolithic times or earlier (Nishimura 1964). Graffiti found in a grotto on an Italian island depicted trapping systems made of palm tree branches for bluefin tuna (Thunnus thynnus) some 4000 years ago (Sara 1980). Sara (1980) further described more modern tuna traps from various sources as early as Aristotle’s time. Trap designs from 2 a.d. by Oppiano very much resemble the present-day traps with “entrances, gateways and access roads.” Reef nets were used by native peoples in Vancouver Island on the Canadian west coast many centuries and probably thousands years ago. Earlier reef nets use kelps spliced into cordage made of natural fiber to form walls to guide salmon into the net between two canoes (Claxton and Elliott 1994).

The natives believed that salmon are more comfortable to enter the reef nets surrounded by kelps. Remains of wooden stakes with formation of chevron- and heart-shaped traps were recently found in Vancouver Island (Greene 2005). Isotope aging indicated that these remains are more than 1000 years old. At the beginning of the 1900s, Europeans introduced large-scale salmon traps to the same fishing grounds using the same fish migration and other behavior patterns. The large traps consist of wooden poles with the netting hung onto them. These large traps were used until1958, when more restrictive fishing regulations, increased cost of material and operation, and development of the modern seine fishery made the trap less feasible. In mid-1990s, a new floating salmon trap (Fig. 7.1) was tested on the same fishing ground where the traditional reef net and staked salmon traps were once operated (He and Walsh 1997, 1998). Fishing for salmon by traps has a long history in Alaska. The first reported use of a salmon trap was in 1885 (Hipkins 1968). The first salmon traps were pile traps and could only be used in shallow water with a good soft bottom so that the piles could be drilled. The first floating salmon trap was built in 1890 and was reported to have been imported from Norway. In 1930, there were 423 floating traps in

Figure 7.1. A new floating salmon trap tested in Vancouver Island, British Columbia. (Courtesy Gordon Curry.)

Large-Scale Fish Traps: Gear Design, Fish Behavior, and Conservation Challenges use in Alaska. This trap was abolished from Alaskan waters in 1959 due to its high fishing efficiency. By 1967, only three traps were operated by Indian Reserves. In 1994, only one salmon trap was operated by an Indian Reserve on Annette Island in Alaska. Weirs made of sticks were used to catch Atlantic salmon (Salmo salar) as early as 1606 in the Canadian Maritime provinces (Anderson and Brimer 1976). There were records of partition to erect fish weirs in New Hampshire in 1729 according to the same source. 7.2.2 Japanese Set Nets A variety of stationary fishing gears resembling the present-day set nets were used in Japan in the early years of the second millennium. A prototype of set net known as oshiki-ami (Fig. 7.2) was used during the Edo period around 1600 a.d. (Inoue et al. 2002). It had a triangle-shaped bag net with a leader net. A similar set net but with a box-shaped bag net

161

called daibo-ami (Fig. 7.2) appeared around 1900. The net was, however, liable to deformation by currents and allowed easy escape of fish. Therefore, the net and schools of fish had to be monitored by watchers constantly and the net had to be hauled soon after the entry of a fish school. Many workers and many boats were required to haul the entire net. Otoshi-ami, which has additional chambers and a bag net, was developed from the daibo-ami around 1910. Otoshi-ami is the most popular set net in Japan at the present. In comparison to the daibo-ami, the escape of the fish is effectively prevented, and the net can be hauled at a certain time of the day rather than immediately after the entry of a school of fish. Only the bag net is hauled up, so it requires less labor and less time to haul. Large set nets are used to catch large pelagic species such as tuna, yellowtail, and salmon as well as demersal species such as sea bream. These set nets are normally owned by fishery cooperatives and involve as many as 75 fishermen in the setting

Figure 7.2. Evolution of Japanese set nets.

162

Fish Behavior near Fishing Gears during Capture Processes

and hauling (Inoue 1988). These large-scale set nets can have variations in the number of entrances and bag nets. The bag net can also be located on the surface, in mid-water, or on the seabed. Mediumscale pelagic set nets are used for small schooling pelagic fish. The largest trap is reported to be 1000 m long and 500 m wide with a leader of as long as 5 km. One of the most important species groups caught by set nets is chum salmon (Oncorhychus keta). About 100,000 tons of salmon are harvested annually in Japan, of which 60% are from coastal waters. More than 95% of coastal salmon landings are from set nets. Occasionally, 10,000 chum salmon are caught in one haul (Inoue and Arimoto 1989). Figure 7.3 illustrates a larger-scale Japanese salmon set net of otoshi-ami style showing the names of the parts of the net. Traditionally set nets are set on the surface, but recently mid-water set nets have been developed and used in northern Japan. Mid-water set nets are more resistant to bad weather and sea conditions and have been reported to catch more salmon than surface set nets.

Figure 7.3.

Japanese set nets are also used for demersal fish species such as flatfishes and Gadidae. The Japanese-style cod trap in use in Newfoundland (described later) was introduced from northern Japan, where it was used to catch Pacific cod (He and Nemoto 1999). Set nets are also used to catch shellfish species such as squid in Japan. These shellfish traps are normally small in scale and operated in coastal waters of shallow depths. The set net technology has been extensively developed in Japan; however, it still possesses the fundamental characters of passive fishing gear. For example, the catches are not very stable because fishing depends on the migration of fish, which may vary with environmental conditions. In recent years, various improvements have been rendered to make this fishery more stable and profitable. 7.2.3 Newfoundland Cod Traps The cod trap is one of the most cost-efficient fishing gears and can be spectacularly successful when used under the appropriate conditions. Since its “invention” in the 1860s, the cod trap has been

An otoshi-ami–style Japanese set net with names of different parts of the net.

Large-Scale Fish Traps: Gear Design, Fish Behavior, and Conservation Challenges modified to catch squid, mackerel, herring, and capelin. In more than 120 years of development, three distinct cod trap styles have emerged—the traditional Newfoundland cod trap, the modified Newfoundland cod trap, and the Japanese-style cod trap. In 1989, there were about 4000 active cod traps in Newfoundland, of which 48.6% were the traditional traps, 37.6% were modified traps, and 13.7% were Japanese traps. The traditional Newfoundland cod trap consists of a square box and a leader. A typical cod trap is 60 fathoms on the round and 10 fathoms deep. The

163

trap is box-shaped and without winker panels (Fig. 7.4). The box has a bottom but does not have a roof. The trap is normally set on the seabed with its floatlines 1 to 2 fathoms below the water surface. The footropes of the trap are weighted with lead ropes and sand bags are added on the corners. The length of the leader is 200 to 300 m depending on the location and bottom topography. In many cases, the leader is fastened at the shore. The Japanese-style cod trap was introduced into the Newfoundland cod fishery in the early 1960s from northern Japan, where the traps were used to

Figure 7.4. Three popular styles of Newfoundland cod traps. (A) Traditional Newfoundland cod trap. (B) The modified Newfoundland cod trap. (C) The Japanese-style cod trap.

164

Fish Behavior near Fishing Gears during Capture Processes

catch Pacific cod (Fig. 7.4). The major difference between the Japanese trap and the traditional Newfoundland traps is that the former has a porch and a funnel leading to the box-type net. There is a roof over the porch and in the box. The Japanese traps are often submerged 5 to 7 m below the surface. The modified Newfoundland trap was styled from the traditional Newfoundland trap by adding two winker panels at the entrance. The modified traps are typically set in waters 3 to 5 m deeper than the depth of the net with its headlines submerged. There is no roof on the trap. 7.2.4 Baltic Fish Traps In the Baltic Sea, traps are used for Atlantic herring, Atlantic salmon, and European whitefish (Coregonus lavaretus). Whitefish traps were evolved from the hoop net in the middle of nineteenth century in the west coast of Finland and their use spread across the bay to the Swedish coast (Toivonen et al. 1992). A typical whitefish/salmon trap consists of a leader, a wing net, one or more sections of middle chambers, and a bag net where fish are raised and harvested (Fig. 7.5) (Lehtonen and Suuronen 2004; Toivonen et al. 1992). There were 200 to 300 whitefish traps along the Finnish coast in the early 1990s, targeting the spring migration of whitefish as they enter Bothnia Bay. A trap can catch as much as

10,000 kg of whitefish in one fishing season, with salmon and brown trout as bycatch. The annual landing of whitefish from traps grew from 10 tons to about 110 tons in the 1980s (Toivonen et al. 1992). Each whitefish trap is different in design. A survey by Toivonen et al. (1992) of fishermen from the Finnish coast indicated that the total trap lengths ranged from 206 to 468 m, of which 150 to 400 m were leaders. The width of the trap (between wings) was 32 to 55 m. Leaders are made of large-mesh polyethylene materials, and other parts of the net are made of nylon. Traditionally, salmon traps use mesh sizes that result in meshing. Whitefish traps use thick twines and mainly use a guiding mechanism to lead the fish into the bag net. Salmon and whitefish traps in the Baltic Sea have undergone tremendous changes in recent years due to increasing interaction between traps and seals. Seals eat fish that are captured or concentrated in the traps and can be accidentally caught by the trap. A discussion on seal-proof traps is provided later. Herring traps in use in the Baltic Sea are also called herring pound net and were introduced from North America at the end of nineteenth century (Toivonen et al. 1992). Most herring traps operate in spring during the spawning migration. The general shapes of herring traps are very similar to

Figure 7.5. A typical salmon/whitefish trap in use in the Baltic Sea. (Redrawn after Lehtonen and Suuronen 2004.)

Large-Scale Fish Traps: Gear Design, Fish Behavior, and Conservation Challenges that of whitefish trap, except that herring traps have much smaller mesh size (Tschernij et al. 1993). A typical trap can land 15 to 20 tons of herring for a season of 1 to 2 months. 7.2.5 Other Traps Scottish bag nets are commonly used for catching salmon in waters beyond defined estuaries. However, their numbers are decreasing as the abundance of wild stock of salmon decreases. These bag nets take advantage of salmon swimming along the coast to reach their spawning river. Bag nets are floated off the seabed. Juvenile bluefin tunas are captured by traps in the Mediterranean Sea and transferred to cages to grow up and fatten for the Japanese sushi market. Bluefin tuna traps are also used on the Canadian east coast. These traps are large-scale fishing gears. Bluefin tuna are large and they need to swim constantly to stay alive; therefore, the box of the trap needs to be large to hold the fish. The main part of trap measures about 44 × 41 m. Whitefish traps are also used in inland waters of Canada and the United States. Traps, or pound nets, are used in the mid-Atlantic coast of the United States, targeting a variety of pelagic species primarily used for bait in the blue crab fishery in Chesapeake Bay. Mackerel and herring traps in use in Newfoundland have very similar designs; in many cases, fishermen use the same trap for both species. A heart-shaped herring trap, or weir, is used in the coastal waters of Quebec and Nova Scotia. Capelin is a small pelagic species found in the North Atlantic, notably Norway and Newfoundland. Capelins spawn on sand beaches along Newfoundland, and some stocks spawn on offshore banks. Capelin is caught with traps and purse seines in Newfoundland during June and July. Capelin traps are very similar to cod traps in shape, except that they have much smaller meshes (around 48 mm) and they are set near the surface. Traps are used widely in other parts of the world. Various types of traps, some with designs similar to the Japanese set nets, are used in various locations along the Chinese coast (Feng et al. 1987). The use and designs of other traps in different parts of the world are described by Gabriel et al. (2005).

165

7.3 FISH BEHAVIOR IN AND AROUND TRAPS As mentioned earlier, a good understanding of fish behavior is very important for successful trap fishing. As a large-scale fishing gear, these traps are often set in a fixed location for the entire season. Failure to understand fish behavior, especially the migration and movement patterns, can result in catch failure for the season. Here we discuss fish behavior in three major trap fisheries—Newfoundland cod trap, Japanese salmon set net and Baltic salmon/whitefish trap. 7.3.1 Cod Behavior near a Newfoundland Cod Trap He (1993) reported a fish behavior study near a modified Newfoundland cod trap using a seabedmounted sonar and an underwater video camera. The trap under observation was a typical cod trap of the “modified” design. The trap was 120 m on the round and 22 m deep. The trap was set at 26-m water depth at the trap entrance, with its floatline 4 m below the surface. The following summarizes findings reported by He (1993). Cod at a distance from the trap. Whether fish at a distance from the trap swim into the trap depends on their route of migration, presence of prey or predator near the trap, or water temperature. Capelin are reported to play a significant role in the inshore migration of cod in Newfoundland. Cod near the leader. The leader of a trap is designed to intercept and guide the fish toward the trap entrance. The basic behavior of fish when blocked by a wall of netting is to swim along the leader toward the deep end (Fig. 7.6). However, if fish approached the trap from the trap box area, the fish may be guided toward the shore (Fig. 7.6). Fish were very reluctant to swim through meshes during voluntary swimming. During the experiment with the cod traps, very few cod were observed to swim through the leader with a mesh size of 203 mm during fishing conditions. The leader is visible, especially when grown with algae, after it is in water for some time. Cod at the entrances of the trap. There are two entrances separated by the leader in cod traps in Newfoundland. Cod were seen to swim into the trap from both entrances. Likewise, they swam out from

166

Fish Behavior near Fishing Gears during Capture Processes

Figure 7.6. Movements of schools of cod (Gadus morhua) near a modified Newfoundland cod trap. (A) Fish guided away from the trap entrance. (B) Fish guided toward the trap entrance. (Redrawn from He 1993.)

both entrances. There are a number of examples indicating that schools of fish swam into the trap from one entrance and a portion of it exited the other entrance. On one occasion, a large school of more than 1100 cod was observed by camera to swim into the trap in 3 min, representing an entering rate of 367 fish/min. However, not all the schools of fish swam into the entrance in a smooth manner. In many instances, a school of fish paused at the entrance while only a portion of it swam into the trap. In one occasion, a large school of cod was spotted with the use of sonar to swim into and subsequently exit the trap in a period of 3.5 min (He 1993). The swimming speed of cod entering and exiting the entrances of the trap was less than one body length per second (BL/s), well below the maximum sustained swimming speed for this species of similar size (approximately 50 cm). Startled schools of fish showed a speed of 4 BL/s. Cod of 50 cm can

swim at 2 BL/s for an indefinite duration and up to 10 BL/s during a brief burst (see Chapter 1). The depth of fish near the trap seemed to relate to the ground swell, which in turn was related to both the wind direction and strength. When there were heavy ground swells, fish stayed 10 to 15 m off the bottom (26-m water depth). On a calm day with no ground swells, fish stayed very close to bottom, often as close as 0.5 m. Cod inside the trap. Cod inside the trap were observed to swim in circles similar to that observed in a large tank. They swam in either a clockwise or a counterclockwise direction without preference. Cod remained about 1 m from the netting of the trap wall. Large schools of cod were observed to swim into and exit out of the trap at ease, indicating that the design of the entrance can be modified to reduce a large-scale exodus. Fish were also observed to swim through 92-mm mesh netting “drying twine” during hauling. This escape behavior may be related

Large-Scale Fish Traps: Gear Design, Fish Behavior, and Conservation Challenges

167

Figure 7.7. Distribution of cod (Gadus morhua) in relation to wind-induced water temperature variations and vulnerability to cod traps set at water depth of 22 m (12 fathoms) in northeastern Newfoundland. (Adapted after Lear et al. 1986.)

to stress associated with reduced space and crowding. Effect of temperature. The success or failure of cod trapping in Newfoundland has largely been related to fish abundance and environmental factors. Temperature is a very important factor affecting the route and timing of migrations, as well as relative distances from the shore and depth of water they reside (Rose 1993). Cod in the Newfoundland area prefer waters with temperature between 0.7° and 4.2°C (Lear 1984). In most of the Newfoundland coast, the peak cod trap season was June and July. Strength and direction of prevailing winds can affect the availability of cod to cod traps set near the shore, as illustrated in Figure 7.7. 7.3.2 Fish Behavior near Japanese Set Nets Fish behavior around the set net The behavior of large schools of sardine (Sardinops melanosticta) in and around the set net was analyzed from sonar image recordings made with a scanning sonar (Inoue 1988; Kim et al. 1995). When large schools of sardine moved along the outside of the set net, the shape of the school gradually changed—the front portion extended forward in the direction of movement and the rear portion concentrated in the same direction such that the school retained its original shape. When large

schools of sardine entered the main net of the set net, the school formation was loosened and pointed in various directions before forming a dense school pattern again and then moved directly to the slope net. When the size of the front portion of the fish school enlarged, the maximum recorded moving speeds were 1.76 and 2.77 m/s for schools inside and outside the set net, respectively. Fish behavior near the leader. Many schools of fish were caught due to the effect of blocking and leading by the set net leader. The behavior of the schools of fish of several species was investigated in the set net fishing grounds around the coast of Japan using a scanning sonar to determine the function of the leader (Inoue 1988). The leader was effective in blocking off the course of fish school, although its mesh size was large enough for the fish to pass through. Only 8% of the schools of fish passed through the leader (Fig. 7.8), whereas 76% of the schools of fish encountered the leader moved along it (Inoue 1988). Of those swimming along the leader, three times as many schools of fish moved offshore toward the set net as they did toward the shore. The leader influenced schools of fish as far as 60 m from the leader. The distance between the leader and fish school depends on species. Barracuda (Sphyraena pinguis) remained 6 to 7 m from the leader, while jack mackerel remained 5 to 20 m and

168

Fish Behavior near Fishing Gears during Capture Processes

Figure 7.8. Proportion of fish schools guided by leaders or penetration through leaders of Japanese set nets. (Redrawn from Inoue and Arimoto 1989.)

yellowtail (Seriola quinqueradiata) remained 10 to 15 m away (Nomura 1980). Different mesh sizes are used for the leader according to different species targeted. The ratio of the mesh size of trap leader to that of gillnet targeting the same species can range from 0.4 to as great as 18 (Nomura 1980). For species such as sardine (Sardinia melanosticta), the ratio can be 18 times because they usually stay far from the leader. The ratio for flounder (Paralitchthys oblivaceus) was only 0.4 because smaller mesh sizes used in the leader can avoid meshing of fish in the leader. Behavior of fish in the playground. The behavior of yellowtail schools in the playground of a

large-scale set net was investigated in relation to the catching function of the funnel net by use of scanning sonar (Inoue 1988; Kim and Inoue 1998). More schools of fish were observed in the playground in the morning but more schools were observed in the bag net in the afternoon. The fish remained in the playground for a long time. Yellowtail schools changed the shape when passing the funnel net. The rate of entering the bag net was 24% among the schools of fish heading toward the funnel net (Kim and Inoue 1998). The rate of exit to the playground from the bag net was 27% among those heading toward the funnel (Fig. 7.9). It seems that the funnel was not very effective in leading in

Large-Scale Fish Traps: Gear Design, Fish Behavior, and Conservation Challenges

169

Figure 7.9. Fish behavior at the slope net. (A) The proportion of schools entering into the slope net. (B) The proportion of schools exiting from the slope net. (Redrawn from Kim and Inoue 1998.)

the fish but was effective in preventing fish from escaping. Capture efficiency of set nets. By counting the number of salmon schools near the trap and the number of schools swimming into the trap, Inoue (1988) found that almost half of the salmon schools near the trap were captured by the trap. Some 46% of the schools observed with sonar within the 500-m range were captured by the trap (Fig. 7.10). 7.3.3 Salmon and Whitefish Behavior near Baltic Traps Research indicates that Atlantic salmon in the Finnish coast of Baltic Sea follow isothermal surface water between 13° and 14°C during spawning migration (Westerberg 1982). Whitefish, however, swim near the seabed, feeding on gastropods during that period (Toivonen and Hudd 1993a). Whitefish change swimming behavior later in the season and prefer to swim near the surface. These behavioral differences were used to reduce salmon bycatch in whitefish traps (see details later in this chapter). Lunneryd et al. (2002) used tags and an acoustic positioning system and tracked whitefish in relation to an experimental leader net with a net enclosure in Sweden. Three leader mesh sizes were used: 100,

300, and 800 mm. They found that none of the 28 individuals (mean weight 0.85 kg) in 478 h of observation swam through the meshes, even though the size of large meshes (800 mm) was large enough for the largest fish (1.396 kg) to do so. Using a random walking model, they projected that the individuals would have 120 net encounters if they were not avoiding the leader net (Lunneryd et al. 2002). They also found that turning distance is larger with the small-mesh leader when compared with the medium-size leader; presumably the small-mesh leader is easier to detect. Lunneryd et al. (2002) recorded the distance at which whitefish react to a netting panel. They found that there were two peak frequencies: one at 3 to 5 m and the other at 15 to 20 m. They argued that the reaction at 3 to 5 m may be visual, but the reaction at the 15- to 20-m range could not be explained. Acoustic pressure from the netting was not strong enough to elicit a response, as measured sound pressure level produced by the leader structure were not of sufficient intensity to be detectable by whitefish at greater than 4 m (Wahlberg et al. 2000). Both salmon and whitefish showed diurnal activity. They were most active between 09:00 and 11:00 h (Toivonen and Hudd 1993a). As a result, they speculated that “fishermen may have disturbed

170

Fish Behavior near Fishing Gears during Capture Processes

Figure 7.10. Proportion of fish schools guided by the leader in Japanese set nets as observed by scanning sonars. (Redrawn from Inoue 1988.)

the entering of fish into their trap by hauling between 06:00 and 09:00.” However, there was little difference of turning distance between night and day. 7.4 FISH BEHAVIOR AND TRAP DESIGNS Traps are passive fishing gears as they are strategically placed on the fishing grounds for fish to swim into them and be trapped. Therefore, knowledge of fish behavior, particularly schooling and migration patterns, is very important. The behavior of fish differs among species, but some aspects can be generalized. Ancient traps such as reef nets use kelp (dune grass) existing on the fishing ground. It was believed that “salmon entering the reef net felt safely surrounded by the dune grass in swift running tides” (Claxton and Elliott 1994). Traps are set at a fixed location awaiting fish and intercept them on their migration route. Prior

knowledge and predictable migration patterns in terms of route and timing are extremely important to successful trapping operations. Salmon are known to return to native rivers after spending some time at sea. Various early salmon traps such as reef nets were thus developed to catch returning salmon by native residents of the North American west coast (Claxton and Elliott 1994). Atlantic cod migrate to inshore waters of Newfoundland and Labrador during summer months to feed on capelin that spawn on beaches (DFO 1988). Cod traps set during these months near the shore trapped a large amount of cod. On the other hand, a cod trap introduced to northern Norway to catch spawning migrating cod that remained in relative deeper waters was not very successful. The areas that schools of fish frequently pass are referred to as the “fish route.” A successful trap operation depends on the position and angle of net in relation to the fish route. Important factors that

Large-Scale Fish Traps: Gear Design, Fish Behavior, and Conservation Challenges

171

Figure 7.11. Illustration of the leader and the trap in relation to bottom contour, headland, and typical fish swimming route.

determine the formation of a fish route include (1) characteristics of the coastal topography and isobaths, (2) presence and location of natural reefs in the area, (3) sea states, and (4) consistency of the seabed. With knowledge of these factors, it is possible to predict types of fish species, size of school, and seasonal changes that occur in a given fish route. Trap locations are called “berth.” Good trap berths are determined through generations of fishing in the area. Most fish seek deeper waters when they encounter danger. Therefore, fish normally swim toward the deeper end of a net set across depths. Traps should therefore be located in the deeper end of the leader to accept fish intercepted by the leader, as illustrated in Figure 7.11. Fish are guided into the playground and, with time, they may swim into the bag net. The history of improving trap and set net designs has therefore focused on two conflicting

goals: (1) allow fish to enter easily and (2) prevent fish from escaping once they have been trapped. Most fish swim against the current; this is called rheotropism. In a laboratory flume tank, fish can be induced to swim against current until they cannot sustain the current and duration. In the field, fish were observed to swim against the current when searching for food and migrating upstream, the latter especially in salmonid. Fish take minimum risk during their routine activities. Experiments in tanks show that fish avoid large-mesh netting panels, even though the mesh sizes are several times larger than their body (Glass et al. 1993). In the field, fish avoid netting, such as the leader of a cod trap with mesh size that is considerably larger than what would effectively gill them (He 1993). Many fish spend their lives in schools; this is particularly true of small pelagics. Fish in a school

172

Fish Behavior near Fishing Gears during Capture Processes

are more affected by their neighbors. In traps and in other fish gears, fish were observed to swim into the nets en masse due to schooling behavior. Presence of the same species of fish on the other side of the leader can cause the fish to swim through the leader, which otherwise does not usually occur. Visibility of the trap and the leader is important in fish avoidance or penetration through the leader. Whitefish (Coregonus lavaretus) show net avoidance both during the day and during the night (Lunneryd et al. 2002). The fish were more likely to detect the net and turn at 15 to 20 m and at 3 to 5 m. The 3- to 5-m detection may be due to visual response, whereas the 15- to 20-m detection may be due to detection mechanisms other than visual, as it does not relate to turbidity and light conditions. Generally speaking, the more visible the leader, the more effective it is in leading the fish along the leader. Visibility of netting underwater is determined by the relative contrast of the twine against background. Wardle (1986) demonstrated how the same twine could have different visibility and appearance depending on the direction at which the twine is being viewed. Yellow twines were reported to have the greatest visibility in clear water, but visibility reduced in turbid water (Nomura 1980). Scientists from Aberdeen (SOAFD 1992) found that glow net was best in blocking Atlantic mackerel (Scomber scombrus) against a blue background in a large tank, whereas monofilament net was the poorest (see Chapter 8). Jarvik (1985) found that the white net leader is better in guiding herring than is dark net in the Estonia coastal herring pound net fishery. 7.5 SIZE AND SPECIES SELECTIVITY AND MORTALITY OF ESCAPEES AND DISCARDS 7.5.1 Size Selectivity Mesh size is the most important factor affecting the size of fish captured in traps. In Newfoundland cod traps, the minimum “dry twine” mesh size is 89 mm. Studies indicated that if the mesh size was increased, the proportion of the small cod (less than 43 cm) would be substantially reduced. As undersized cod can be as much as 66% of the total cod trap landings when using the 89-mm mesh size netting (Brothers

and Hollett 1991). Increasing mesh size would reduce landing in the short term but would contribute to stock recovery. In addition to large mesh sizes in the “dry twine” area, grids were also tested (Brothers 2002). Square mesh panels were tested in the west coast of Newfoundland in 1997 (Brothers 2000). A 6 × 6 m square mesh panel of 117-mm mesh size (85-mm bar length) installed in the drying twine seemed to improve size selectivity and reduce undersized cod (less than 43 cm). Fujimori et al. (2000) reported that diamond mesh hung at 70% has a better passing ratio than at other hanging ratios and than square mesh of the same mesh size when used in the bag net of Japanese salmon set net. In his laboratory tank experiment, the passing ratio, which can be considered as the escape ratio in real fishing situations, increased with time passed and stabilized after approximately 60 s. 7.5.2 Species Selection and Bycatch Reduction Traps generally have good species selection characteristics because traps target schooling fish that are usually of similar species and sizes. However, on some occasions, different species mix and are caught together by the traps. On the Canadian east coast, Atlantic salmon (Salmo salar) has been closed to commercial fishing since early 1990s. However, they were caught in various fish traps, such as cod and capelin traps. In the early 1990s, various experiments were carried out to reduce salmon bycatch in these traps, and some gear modifications were quite successful. The modifications were made primarily to the leader of the cod traps: the sunken leader, the large mesh top panel, and the deflector panels (Fig. 7.12). The sunken leader design (Fig. 7.12A) was the best design and reduced salmon bycatch by 88% (Brothers 1996). Trap leaders made of large mesh sizes in the top portion were also tested in capelin traps with varying degrees of success (Brothers 1997). In the Baltic Sea off the coast of Finland, a large number of Atlantic salmon were caught in whitefish traps during spring shoreward migration. This has resulted in the ban of all stationary gears, including traps, in the state waters during spring and early summer. Research indicates that during spawning

Large-Scale Fish Traps: Gear Design, Fish Behavior, and Conservation Challenges

Figure 7.12.

173

Cod trap modification to reduce salmon bycatch. (Redrawn from Brothers 1996.)

migration in the spring, salmon follows isothermal surface water between 13° and 14° while whitefish are feeding on gastropods near the seabed (Toivonen and Hudd 1993a). Installing a 3-m-deep prohibiting net (Fig. 7.13) near the surface made of large-mesh polyethylene material in front of the trap entrance reduced salmon bycatch by 62% without reduction in whitefish catch during May and June (Toivonen and Hudd 1993a). However, the same design resulted in a large reduction (17%) of whitefish as they come near the surface in the fall. 7.5.3 Survival of Fish Discarded from Traps Very few studies were conducted on trap-caught and discarded fish. Judging from the high survival

rate of cod transferred from trap to farming cages for grow-out, the mortality rate of discards from traps may be quite low if handled properly. In fact, several studies have used trap-caught fish for tagging studies (Jokikokko 2002). Mortality of fish discarded from traps is affected by physical injury caused by contact with netting or other gear parts. This is especially serious when there is strong water currents that deform the shape of the netting. Stress due to confinement and hypoxia, especially when there is a large catch, can cause mortality, as seen in purse seines (Lockwood et al. 1983). Undersized salmon less than 60 cm in length caught in traps are required to be released in the

174

Fish Behavior near Fishing Gears during Capture Processes

Figure 7.13. Whitefish trap modifications to reduce salmon bycatch in the Baltic Sea. (Adapted from Toivonen and Hudd 1993.)

Finnish Baltic salmon trap fishery. It is thus important to know whether those salmon released from traps die due to capture and handling processes. An experiment conducted in 1993 indicated that undersized Atlantic salmon caged for 10 days had a mortality rate of 35% (Toivonen and Hudd 1993b). This mortality rate was considered high and might have been caused by caging stress, high surface water temperature, strong currents, and handling. Indeed, a subsequent experiment using tag and recapture method indicated a mortality rate of about 11% (4% to 21%). Furthermore, migration behavior of the trap-caught and released salmon was not altered due to the capture and handling process (Siira et al. 2006). 7.6 CONSERVATION ISSUES AND MITIGATION MEASURES IN TRAP FISHERIES As a type of large-scale fishing gear, fish traps are particularly susceptible to interactions with mega-

fauna species in the sea, including mammals, turtles, and birds. Most traps target schooling pelagic species, such as herring and capelin, which are often the food of the megafauna species. Some mammals take advantage of traps that concentrate fish to catch prey (Fjalling 2005; Suuronen et al. 2006). Conflict arose when humans and animals target the same species, resulting in mortality of animals and loss of or damage to fishing gears (Lien et al. 1989; Wickens 1995). 7.6.1 Interaction of Marine Mammals and Sea Birds with Newfoundland Traps Collisions and other interactions of whale or other large marine animals with cod traps have been reported in Newfoundland, especially before the cod moratoria in 1992 (Lien et al. 1989, 1992). Annual reports from Lien and his Whale Research Group of Memorial University of Newfoundland between 1979 and 1993, as analyzed by He and Howse (1994), indicated that an annual average of 24 large

Large-Scale Fish Traps: Gear Design, Fish Behavior, and Conservation Challenges whales, mostly humpbacks (Magaptera noveangliae), were killed as a result of interaction with traps, primarily cod traps in Newfoundland and Labrador. Another 54 were released live as a result of the entrapment and release effort by the Whale Research Group. Whale collisions damage or take away traps, as well as kill or wound the whale. Research to prevent this conflict include the use of whale alarms, acoustic signaling devices that warn the whale of the presence of the net (Fig. 7.14). Other megafauna and nontarget species that interact with cod traps in Newfoundland include dolphins, seals, seabirds, and sharks (Lien et al. 1993). Many marine mammals possess ability to echolocate objects, such as prey items in the water, but their capability varies with species and acoustic characteristics of the habitat (Evans 1973). These animals may also be able to perceive the existence of the net in water through echo-location techniques. Todd (1991) found that traps that use smaller mesh size netting (e.g., capelin traps) have

Figure 7.14. A prototype whale alarm developed by the Whale Research Group of Memorial University of Newfoundland.

175

a stronger acoustic signal than do cod traps that use large meshes. Consequently, more humpback whales were entrapped in cod traps than in capelin traps. These factors, as well as behavior and sensory capability of cetaceans and mechanisms of their detection and interaction with stationary fishing gears, were reviewed by Nelson and Lien (1992). One device tested by Lien et al. (1992) was an acoustic alarm that produces a 3- or 6-s sound at 4-kHz peak frequency with intensity of 135 dB (re 1 μPa at 1 m). Testing of the alarm in cod traps in Newfoundland indicated a significant decrease in the collision and entrapment rate when the alarms were used (Lien et al. 1992). The use of the alarm did not reduce target species (cod) catch during the test period. 7.6.2 Seals and Baltic Fish Traps Interaction of traps and seals in the Baltic are twofold: seals may be caught and killed by traps and traps and captured fish may be damaged by seals. Westerberg et al. (2006) reported that 462 grey seals (Halichoerus grypus) were caught by commercial fishing gears in Sweden, some of which were by traps targeting whitefish and salmon. A recent survey in the northern Baltic Sea by Kauppinen et al. (2005) found that seals, particularly grey seals, damaged at least 37% of salmon (S. salar) in Bothnian Sea and 3% to 9% in other areas in older traps. Damage to whitefish ranged from 5% to 7%. Gear damages due to seal ranged from 2% to 15% per trap hauls. More interactions between seals and traps may have resulted from increased seal population and more offshore setting of traps (Kauppinen et al. 2005). Fjalling (2005) demonstrated that there are hidden effects of seals to the trap fisheries in addition to the observed remains of the damaged fish. Some fish may be consumed whole as seals seem to prefer smaller fish to larger fish, or small parts of remaining fish may have fallen out of the net. The presence of seals in the trap area may discourage fish from swimming into the trap. Damage to the net due to seals allows more fish to escape through the holes made by the seals. Traditional estimates of seal damages based on the number of fish remains in the net could underestimate the seal effect by as much as 46% (Fjalling 2005).

176

Fish Behavior near Fishing Gears during Capture Processes

Various measures were tested or being tested to protect catch from damage from seals. Lehtonen and Suuronen (2004) tested a wire grid at the entrance of the fish bag located at the last section of the trap and a strong and larger fish bag. The grid was made of 2-mm steel cables with 175-mm spacing and was fitted to the entrance vertically (Fig. 7.15). The experimental fish bag was made of Dyneema material of 1-mm diameter and 80-mm mesh size. The commercial fish bag was made of 210/30 twisted nylon and 130- to 140-mm mesh size. The experimental fish bag was almost twice as large in volume. The result indicated that the experiment trap with a grid and a strong, large bag net was able to reduce seal-induced salmon damage in the fish bag by 70%. Catch was higher in the experimental trap, indicating that salmon were not

prevented from entering into the fish bag by the wire grid but seals were prevented. There were no damages to the strong fish bag, indicating that the strong material prevented seals from causing damage. Seals were observed to turn away from the wire grid during underwater observations. Lunneryd et al. (2003) used large meshes (400 mm) in side panels of the first and second sections of a modified trap to allow escape of salmon and trout chased by seals. While the amount of catch of salmon and trout was similar between the two traps, the standard trap with 200-mm mesh size sustained considerably more damage due to seal feeding on fish enmeshed in the side panel. Seal activity around the standard trap was 16 times greater than that around the modified trap. It was estimated that 65% of potential catch might have

Figure 7.15. Grid designs to reduce interaction of seals with salmon and whitefish traps in the Baltic Sea. (Lehtonen and Suuronen 2004.)

Large-Scale Fish Traps: Gear Design, Fish Behavior, and Conservation Challenges

177

Figure 7.16. A rope leader design to reduce turtle bycatch in the mid-Atlantic pound net fishery. (Adapted from DeAlteris and Silva 2007.)

been lost in the standard trap due to seal damages, while 52% of potential catch might have been lost due to escape through the large meshes. Using large mesh in the trap makes it less rewarding to seals to feed inside the trap and may lead to fewer seals on the trap grounds in the long term (Lunneryd et al. 2003). Suuronen et al. (2006) tested two traditional and five modified salmon traps during 2003 and 2004 on the Finnish coast of Baltic and found that a combination of large meshes in the wings together with a protected fish bag and a wire grid at the entrance to the fish bag can reduce seal damage and improve catch. More recent work includes develop-

ment of a pontoon trap to catch live seals and release them far away from the traps (Lehtonen and Suuronen 2010). 7.6.3 Turtles and the Pound Net Fishery in the Mid-Atlantic Coast of the United States Several species of sea turtles, including hawksbill (Eretmochelys imnricates), green (Chelonia mydas), leatherback (Dermochelys coriacea), Kemp’s Ridley (Lepidochelys kempii), and loggerhead (Caretta caretta), seasonally visit Chesapeake Bay in the eastern United States. They are occasionally caught in the leader and the bag net of pound nets

178

Fish Behavior near Fishing Gears during Capture Processes

targeting various fish species such as weakfish (Cynoscion regalis), croaker (Micropogonias undulates), harvest fish (Peprilus alepidotus), butterfish (Peprilus triacanthus), and threadfin shad (Dorosoma petenense). To reduce interaction with turtles and subsequent mortality, DeAlteris and Silva (2007) tested a modified leader made of a combination of vertical ropes (top two-thirds) and netting (lower one-third) (Fig. 7.16). The theory was that pelagic species such as harvestfish, butterfish, and threadfin herring would be guided by vertical lines toward the bag net located at the deep end of the leader, whereas turtles, which are usually also near the surface, would pass through the gaps between the ropes (61-cm spacing) without being caught. Comparative fishing between the modified leader and commercial leaders made of 280-mm mesh size indicated a substantial reduction in turtle interaction. Twenty-three turtles interacted with the leaders of four pound nets with a control leader and none with the same nets with an experimental leader. The use of the modified leader effectively eliminated turtle encounters during the experimental time period. As a result, pound net leader specifications similar to that tested were put into rule by the National Marine Fisheries Service (Federal Register 2006). The rule requires that all offshore pound net (greater than 3 m in depth at the land end of the leader) must use hard lay ropes of greater than 8-mm diameter in the top part of the leader (twothirds) with spacing not less than 61 cm. The mesh size of the bottom one-third netting must not be greater than 203 mm. 7.7 CONCLUDING REMARKS Trap is one of the oldest fishing gears used in commercial fisheries. Traps are used all over the world, especially Japan and Newfoundland, catching a variety of fish and shellfish species. Successful operation of the trap requires good understanding of fish behavior, especially migration pattern and reaction to fishing gears. The design of a trap is a compromise between two conflicting objectives: easy access for the fish and prevention of fish escape. Challenges to trap fisheries include interactions with mammal mammals and other charismatic species. These interactions cause morality to the animal and damage to the target species and fishing

gears. Mitigation measures are promising for some fisheries, with enhanced research efforts required in other fisheries. REFERENCES Anderson GJ and Brimer AE. 1976. Salar: The Story of the Atlantic Salmon. New York: The International Atlantic Salmon Foundation. 74 pp. Brothers G. 1996. Salmon bycatch in cod traps. CASEC Project Summary. Department of Fisheries and Oceans, St. John’s, Newfoundland, Canada. 6 pp. Brothers G. 1997. Salmon bycatch in capelin traps. CASEC Project Summary. Department of Fisheries and Oceans, St. John’s, Newfoundland, Canada. 5 pp. Brothers G. 2000. Testing square mesh panels in trap nets to reduce the catch of juvenile Atlantic cod. ICES CM/J: 15: 8 pp. Brothers G. 2002. Cod trap selectivity: an experiment to reduce the catch of small fish with the use of rigid grates. DFO Proj. Sum. EACT-25.2002.DFO (FDP 374). 5 pp. Brothers G and Hollett J. 1991. Effect of mesh size and shape on the selectivity of cod traps. Can. Tech. Rep. Fish. Aquat. Sci. 1782: 73 pp. Claxton E Jr and Elliott J Jr. 1994. Reef net technology of the saltwater people. Brentwood Bay, BC, Canada: The Saanich Indian School Board. 55 pp. DeAlteris J and Silva R. 2007. Performance in 2004 and 2005 of an alternative leader design on the bycatch of sea turtles and the catch of finfish in Chesapeake Bay pound nets, offshore Kiptopeake, VA. Final summary report submitted to National Marine Fisheries Service. DFO. 1988. The science of cod. Fo’c’sle. 8(2): 29 pp. St. John’s, Newfoundland: Canadian Department of Fisheries and Oceans. Evans WE. 1973. Ecolocation by marine dolphinids and one species of freshwater dolphin. J. Acoust. Soc. Am. 54: 191–199. Federal Register. 2006. Sea turtle conservation; modification to fishing activities. 50 CFR Parts 222 and 223. Vol. 71, No. 121. June 23, 2006. Feng S, Huang X and Ma S. 1987. China Atlas of Marine Fishing Gears. Hangzhou, Zhejiang, China: Zhejiang Sci. Technol. Pub. 386 pp. Fjalling A. 2005. The estimation of hidden sealinflicted losses in the Baltic Sea set-trap salmon fisheries. ICES J. Mar. Sci. 62: 1630–1635. Fujimori Y, Abe K, Zhimizu S and Miura T. 2000. Analysis of the escape behavior of juvenile salmon

Large-Scale Fish Traps: Gear Design, Fish Behavior, and Conservation Challenges Oncorhynchus keta from the bag-net for bycatch prevention in a set-net fishery. Fish. Sci. 66: 424–431. Gabriel O, Lange K, Dahm E and Wendt T. 2005. Von Brandt’s Fish Catching Methods of the World. 4th ed. Oxford: Blackwell. 523 pp. Glass CW, Wardle CS and Gosden S. 1993. Behavioral studies of the principles underlaying mesh penetration by fish. ICES Mar. Sci. Symp. 196: 92–97. Greene N. 2005. A new angle on northwest coast fish trap technologies: GIS total station mapping of intertidal wood-stake features at Comox Harbor, BC. Presented at the Canadian Archaeology Association 2005 Annual Conference, Nanaimo, BC. He P. 1993. The behavior of cod around a cod trap as observed by an underwater camera and a scanning sonar. ICES Mar. Sci. Symp. 196: 21–25. He P and Howse K. 1994. Groundfish Harvesting Technologies: An Annotated Bibliography. Fishing Technology Unit Report 11/94. St. John’s, Newfoundland: Fisheries and Marine Institute. 115 pp. He P and Nemoto M. 1999. The Newfoundland cod trap: its origin, development and fishery. Trap Net Fish. 95: 34–43. (in Japanese). He P and Walsh P. 1997. Behavior of salmon near the new Sooke trap. Fishing Technology Unit Report 02/97. St. John’s, Newfoundland: Fisheries and Marine Institute. He P and Walsh P. 1998. Behavior of salmon near the new Sooke trap. Fishing Technology Unit Report 02/98. St. John’s, Newfoundland: Fisheries and Marine Institute. 18 pp. Hipkins FW. 1968. Construction and operation of a floating Alaska salmon trap. US Fish Wildl. Bur. Comm. Fish. Leaflet 611: 12 pp. Inoue Y. 1988. Fish behavior in the set net fishing grounds using a sonar. Bull. Nat. Res. Inst. Fish. Eng. Japan. 9: 227–287 (in Japanese with English abstract). Inoue Y and Arimoto T. 1989. Scanning sonar survey on the capturing process of trapnet. Proc. World Symp. Fish. Gear and Fish. Vessel Design. pp 417– 421. St. John’s, Newfoundland: Marine Institute. Inoue Y, Matsuoka T and Chopin F. 2002. Technical guide for set-net fishing. International Set Net Fishing Summit in Himi, Kita-Nihon Kaiyo Center, Himi, Japan. 42 pp. Jarvik A. 1985. Possibility for rationalization of a spring spawning herring poundnet fishery. Finnish Fish. Res. 6: 118–126.

179

Jokikokko E. 2002. Migration of wild and reared Atlantic salmon (Salmo salar L.) in the river Simojoki, northern Finland. Fish. Res. 58: 15–23. Kauppinen T, Siira A and Suuronen P. 2005. Temporal and regional patterns in seal-induced catch and gear damage in the coastal trap-net fishery in the northern Baltic Sea: effect of netting material on damage. Fish. Res. 73: 99–109. Kim MK and Inoue Y. 1998. Studies on the behavior of fish schools in the main-net of a large scale setnet using scanning sonar V. The behavior of yellowtail Seriola quinqueradiata school entrapped in a large set-net and the catching function of the funnel-net. Bull. Kor. Soc. Fish. Tech. 34(1): 13–20 (in Korean with English abstract). Kim MK, Inoue Y and Park JS. 1995. Studies on the behavior of fish schools in the main-net of a large scale set-net using scanning sonar II. The behavior of large school of sardine Sadinops melanosticta in and around the Set-net. Bull. Kor. Soc. Fish. Tech. 31(1): 8–14 (in Korean with English abstract). Lear WH. 1984. The winter distribution of cod in NAFO Division 2J, 3K and 3L, based on research vessel catches during 1978–1983. NAFO SCR Doc. 84/VI/24. 799: 9 pp. Lear WH, Baird JW, Rice JC, Caescadden JE, Lilly GR and Akenhead SA. 1986. An examination of factors affecting catch in the inshore cod fishery of Labrador and Eastern Newfoundland. Can. Tech. Rep. Fish. Aquat. Sci. 1469: 71 pp. Lehtonen E and Suuronen P. 2004. Mitigation of sealinduced damage in salmon and whitefish trapnet fisheries by modification of the fish bag. ICES J. Mar. Sci. 61: 1195–1200. Lehtonen E and Suuronen P. 2010. Live-capture of grey seals in a modified salmon trap. Fish. Res. 102: 214–216. Lien J, Barney W, Todd S. et al. 1992. Effect of adding sounds to cod traps on the probability of collisions by humpback whales. In: Marine Mammal Sensory Systems. pp 701–707. Thomas JA, Kastelein RA and Supin AY (eds). New York: Plenum. Lien J, Barney W, Ledwell W. et al. 1993. Incidental entrapments of marine mammals by inshore fishing gears reported in 1992, some results of bycatch monitoring and tests of acoustic deterrents to prevent whale collisions in fishing gear. Whale Research Group, Memorial University of Newfoundland, St. John’s, Newfoundland. 28 pp. Lien J, Stenson GB and Ni IH. 1989. A review of incidental entrapment of seabirds, seals and whales

180

Fish Behavior near Fishing Gears during Capture Processes

in inshore fishing gear in Newfoundland and Labrador: a problem for fishermen and fishing gear designers. Proc. World Symp. Fish. Gear and Fish. Vessel Design. pp 67–71. St. John’s, Newfoundland: Marine Institute. Lockwood SJ, Pawson MG and Eaton DR. 1983. The effect of crowding on mackerel (Scomber scombrus L.)—Physical condition and mortality. Fish. Res. 2: 129–147. Lunneryd S-G, Fjalling A and Westerberg H. 2003. A large-mesh salmon trap: a way of mitigating seal impact on a coastal fishery. ICES J. Mar. Sci. 60: 1194–1199. Lunneryd S-G, Westerberg H and Wahlberg M. 2002. Detection of leader net by whitefish Coregonus lavaretus during varying environmental conditions. Fish. Res. 54: 355–362. Nelson D and Lien J. 1992. A review of gear and animal characteristics responsible for incidental catches of cetaceans in fishing gear, along with proposed solutions. Proceedings of the World Fisheries Congress, Athens, Greece, May 3–8, 1992. Nishimura A. 1964. Primitive fishing methods. Ryukyuan Culture and Society. pp. 67–77. Nomura M. 1980. Influence of fish behavior on use and design of setnets. In: Fish Behavior and its Use in the Capture and Culture of Fishes. ICLARM Conf. Proc. 5: 446–472. Bardack JE, Magnuson JJ, May RC and Reinhart JM (eds). Manila, Philippines: International Center for Living Aquatic Resource Management. Rose GA. 1993. Cod spawning on a migration highway in the northwest Atlantic. Nature. 366: 458–461. Sara R. 1980. Bluefin tuna trap fishing in the Mediterranean. ICCAT Col. Vol. Sci. 11: 129–144. Scottish Office Agriculture and Fisheries Department (SOAFD). 1992. Marine Laboratory Annual Review 1990–1991. Scottish Office Agriculture and Fisheries Department, Aberdeen, UK. 84 pp. Siira A, Suuronen P, Ikonen E, and Erkinaro J. 2006. Survival of Atlantic salmon captured in and released from a commercial trap-net: Potential for selective harvesting of stocked salmon. Fish. Res. 80: 280–294.

Suuronen P, Siira A, Kauppinen T, Riikonen R, Lehtonen E and Harjunpaa H. 2006. Reduction of seal-induced catch and gear damage by modification of trap-net design: Design principles for a seal-safe trap-net. Fish. Res. 79: 129–138. Todd SK. 1991. Acoustic Properties of Fishing Gear: Possible Relationships to Baleen Whale Entrapment. MSc Thesis, Memorial University of Newfoundland. 213 pp. Toivonen A and Hudd R. 1993a. Behavioral differences of Atlantic salmon (Salmo salar) and whitefish (Coregonus lavaretus) as the basis for improving the species selectivity of whitefish trapnets. ICE Mar. Sci. Symp. 196: 51–58. Toivonen A and Hudd R. 1993b. Survival of undersized salmon after release from the trap net. ICES CM/B:10: 6 pp. Toivonen A, Hudd R and Heikkilä P. 1992. European whitefish trap net fishing gears in the southern part of the Bothnian Bay (Baltic). Pol. Arch. Hydrobiol. 39: 879–884. Tschernij V, Lehtonen E and Suuronen P. 1993. Behavior of Baltic herring in relation to a poundnet and the possibility of extending the poundnet season. ICE Mar. Sci. Symp. 196: 36–40. Wahlberg M, Lunneryd SG, Bégout-Anras ML and Westerberg H. 2000. Whitefish leader net avoidance: Possible role of auditory cues. Adv. Fish Telemetry. Fish. Aquaculture Sci. 137–147. Wardle CS. 1986. Fish behavior and fishing gear. In: Pitcher TJ (ed). The Behavior of Teleost Fishes. pp 463–495. London: Croom Helm. Westerberg H. 1982. Ultrasonic tracking of Atlantic Salmon (Salmo salar L.) – I. Movements in coastal regions. Rep. Inst. Freshw. Res. Drottoinghoim. 60: 81–101. Westerberg H, Lunneryd A-G and Fjalling A. 2006. Reconciling fisheries activities with the conservation of seals throughout the development of new fishing gear: a case study from the Baltic fishery-grey seal conflict. Am. Fish. Soc. Symp. 49: 587–597. Wickens PA. 1995. A review of operational interactions between pinnipeds and fisheries. FAO Fish. Tech. Pap. 346: 86 p.

Large-Scale Fish Traps: Gear Design, Fish Behavior, and Conservation Challenges SPECIES MENTIONED IN THE TEXT albacore, Thunnus alalunga alewife, Alosa pseudoharengus American shad, Alosa sapidissima Atlantic cod, cod, Gadus morhua Atlantic herring, herring, Clupea harengus Atlantic mackerel, mackerel, Scomber scombrus Atlantic salmon, Salmo salar barracuda, Sphyraena pinguis blueback herring, Alosa aestivalis bluefin tuna, Thunnus thynnus bluefish, Pomatomus saltatrix butterfish, Peprilus triacanthus capelin, Mallotus villosus chum salmon, Oncorhynchus keta coho salmon, Oncorhynchus kisutch croaker, Micropogonias undulates European whitefish, Coregonus lavaretus flathead, Platycephalus bassensis green turtle, Chelonia mydas grey seal, Halichoerus grypus haddock, Melanogrammus aeglefinus harvest fish, Peprilus alepidotus hawksbill turtle, Eretmochelys imnricates humback whale, Magaptera noveangliae jack mackerel, Trachurus japonicus Japanese mackerel, chub mackerel, Scomber japonicus kawakawa, Euthunnus affinis

181

Kemp’s ridley turtle, Lepidochelys kempii lake whitefish, Coregonus lavaretus leatherback turtle, Dermochelys coriacea loggerhead turtle, Caretta caretta lemon shark, Negaprion brevirostris leopard shark, Triakis semifasciata Pacific bonito, Sarda chiliensis plaice, Pleuronectes platessa redfish, Sebastes marinus sailfish, Istiophorus platypterus saithe, Pollachlus virens sardine, Sardinops melanosticta seabass, Dicentrarchus labrax skipjack tuna, Katsuwonus pelamis sockeye salmon, Oncorhynchus nerka sole, Solea solea sprat, Sprattus sprattus striped bass, Morone saxatilis striped mullet, Mugil cephalus threadfin shad, Dorosoma petenense wahoo, Acanthocybium solandrei weakfish, Cynoscion regalis white marlin, Tetrapturus albidus whiting, Gadus merlangus winter flounder, Pseudopleuronectes americanus yellowfin tuna, Thunnus albacares yellowperch, Perca flavescens yellowtail, Seriola quinqueradiata yellowtail flounders, Pleuronectes ferruginea

Chapter 8 Fish Behavior near Gillnets: Capture Processes and Influencing Factors Pingguo He and Michael Pol

8.1 INTRODUCTION Gillnets are simple and versatile gears that catch a variety of fish and shellfish. Unlike mobile gears such as trawls, gillnets do not need to be towed or moved to catch fish; unlike baited gears such as hooks and pots, gillnets do not require the addition of bait; and unlike fixed gears such as traps and weirs, gillnets do not require fixed structures and are much more easily portable. Gillnets may be one of the simplest fishing gears in design with a plain sheet of webbing salvaged to frame ropes. They are used in every region of the world and operated from small boats of a few meters in length to highly mechanized offshore vessels. On closer examination, however, greater complexity is revealed. Even small details of gillnet construction appear to affect species and size selectivity. Although gillnets are simple in design and operation, the behavior of fish during the gillnet capture process is largely undocumented and not well understood. The history of gillnetting may be as old as that of net making, and references can be traced back to 3000 years ago in Egyptian tombs. The modern commercial gillnet fishery in the Northwest Atlantic dates to the mid-1800s when natural fibers such as cotton and hemp were used to knit the netting. Gillnetting expanded in this region after migration of hauling technology from the Laurentian Great Lakes to Massachusetts in the 1930s. Synthetic materials were tested in fishing nets in the 1950s

and became very popular on both sides of the North Atlantic due to the large catch increase observed and the almost maintenance-free nature of the material (He 2006a; Pol and Carr 2000; Potter and Pawson 1991). Modern gillnet webbings are made as invisible as possible to mesh fish before they can avoid them. Fish meshing into the net are often caught behind their gills, or “gilled,” and thus the term “gillnet” is used, although other methods of capture are also common in gillnets. Gillnets and entangling nets are one of the nine basic fishing gear categories in the U.N. World Food and Agricultural Organization (FAO) classification of fishing gears (Nedelec and Prado 1990). Set gillnets, driftnets, trammel nets, fixed gillnets, and encircling gillnets are five major subtypes in this category. Some key features of these nets are listed in Table 8.1. These different types of nets or the same type of nets of different mesh sizes and rigging may be combined to form a “combination gillnet.” A typical gillnet consists of webbing and frame ropes (headrope and footrope) (Fig. 8.1). The webbing is manufactured as diamond mesh in a single piece, including a selvedge of usually double monofilament along the top and bottom. The webbing is cut to length and lashed at intervals to the headrope and the footrope with hanging line. The headrope may have internal floatation or a series of floats attached to it. Footropes are often

183

184

Fish Behavior near Fishing Gears during Capture Processes

Table 8.1. Types of Gillnets and Their Key Features Gear Type

Important Features

Set gillnets

Anchored/weighted to the bottom; relatively stationary; can be set on the bottom, in midwater or near the surface Not fixed to the bottom; drift with the current; usually near surface; either tied or not tied to the vessel. Three layers of nets; a middle net with a smaller mesh size and two outer nets with larger mesh sizes Hung onto stakes to form a wall or “fence”; usually in tidal and shallow waters or in rivers

Drift gillnets

Trammel nets

Fixed gillnets

constructed from braided line with built-in lead (“leadline”), although a series of single weights may also be used. Typically, the gear is also anchored at both ends with solid weights or Danforth anchors using bridle lines. Buoy lines with buoys and/or highflyers are used to mark both ends of the gear at the surface. Depending on the fishing area, jurisdiction, and the type of fishery, a gillnet may require attachment of tags, pingers, weak links, breakaway swivels, radar reflectors, or acoustic gear-finding transmitters. There may be specific requirements on the size, strength, and density of ropes (e.g., NOAA 2008), as well as the maximum number of nets allowed. 8.2 CAPTURE MECHANISMS, GEAR DESIGNS, AND FISHING EFFICIENCY Four basic mechanisms of fish capture by gillnets can be identified: gilling, wedging, snagging, and entangling (Hovgård and Lassen 2000), as shown in Figure 8.2. • Gilling—caught with the mesh behind the gill cover • Wedging—caught by the largest part of the body

• Snagging—caught by the mouth or teeth or other part of the head region • Entangling—caught by spine, fins, or other parts of the body as a result of struggling Fish may be caught by more than one of these mechanisms in the same gillnet. Key design features of a gillnet include netting material and color, twine diameter and number of filaments, mesh size or opening, vertical and horizontal hanging ratios, and net dimension (length and height). The mesh size of a gillnet determines to a great extent the size of fish caught in the net as proposed by Baranov (1948) in his geometric similarity theory. He predicted that the majority of fish retained by a gillnet would have their length within 20% of the optimal length (modal length). In practice, this relationship may not be that simple, considering a wide range of species, gear design features, and operational conditions (Hamley 1975). While surveying young Atlantic cod (Gadus morhua) in Greenland waters, Hovgård (1996) found that more cod were caught by gilling, and less by other mechanisms, as mesh size was increased. Comparative fishing trials using gillnets of 127and 140-mm mesh size on the south coast of Newfoundland targeting redfish resulted in 3.6 times more fish caught in the smaller mesh size nets (Brothers and Yetman 1982). However, catch rates of Greenland halibut Reinhardtius hippoglossoides increased with larger meshes (Melindy and Flight 1992). Gillnets made of 203-mm mesh size caught 38% more fish in weight than those with 140-mm mesh size (Melindy and Flight 1992). The increase in catch rates was due to the increase in the size of fish caught for the larger-mesh nets as the average weight of fish caught in the large-mesh nets was 3.5 times heavier than those in the smallmesh nets. Comparing gillnets of 180- and 220-mm mesh sizes in the Barents Sea for Greenland halibut, Nedreaas et al. (1993) found that the modal length was 55 cm for the smaller mesh size compared with 66 cm for the large-mesh size. While regulating mesh size to reduce undersized fish has been a common management measure in many fisheries, for some species, such as paddlefish (Polyodon spathula), size selectivity cannot be established,

Fish Behavior near Gillnets: Capture Processes and Influencing Factors

185

Figure 8.1. Schematic illustration of a string of gillnet while fishing. Inset: an anatomy of a gillnet with names of gear components. (He 2006a.)

perhaps due to unusual morphology (Scholten and Bettoli 2006). Hanging ratio is another factor for design consideration. Horizontal hanging ratio is the ratio of rope length (headrope or footrope) to the stretched length of the attached webbing. Vertical hanging ratio is similarly the ratio of the skirt line to the stretched length of the webbing. Both are expressed as a decimal or percentage, with lower values indicating more “slackness” and affecting the shape of the mesh (Fig. 8.3) and resistance to penetration. Takagi et al. (2007) modeled forces acting on a bottom sink gillnet and determined that discontinuities and strong localized forces developed on the net surface. These results suggest that hanging ratios create different relative forces on the meshes, which may increase or decrease the amount of force required

for mesh penetration by fish. Comparative fishing trials between gillnets of different headrope hanging ratios (0.5–0.7) indicated that the best hanging ratio for catching Atlantic cod (Gadus morhua) was 0.6 instead of the traditional 0.5 used by Norwegian fishermen (Angelsen et al. 1979). For European dab, nets with a hanging ratio of 0.2 caught twice as many fish as nets with a hanging ratio of 0.6 (Hovgård and Lassen 2000). Samarayanka et al. (1997) reported 40% more catch of tuna (mostly skipjack tuna, Katsuwonus pelamis, and yellowfin tuna, Thunnus albacares) and sharks with a hanging ratio of 0.5 versus 0.6. Slackly hung gillnets have been found to result in more fish becoming entangled than gilled, which results in poorer size selectivity (Angelsen et al. 1979; Hamley 1975; Samaryanka et al. 1997; Stewart 1987). When

186

Fish Behavior near Fishing Gears during Capture Processes

Figure 8.2. Fish capture by gillnets, illustrating four modes of capture: gilling, wedging, snagging, and entangling.

studying Tilapia, Hamley (1975) obtained a size range (90% of catch) of 18 to 23 cm in a tightly hung net but 8 to 22 cm in a slackly hung net. Sulaeman et al. (2000) attributed the improved

catch efficiency of slackly-hung gillnets to an increase in the mesh per unit area of the net panel. However, a 25% increase in catch was observed by Samarayanka et al. (1997) even when adjusted for the increased area. Webbing material, numbers of filaments in the twine, and twine size affect visibility of the netting in water and the “softness” of the netting, which in turn affects the mechanism of fish capture. Monofilament nets are less visible and generally produce larger catches than multifilament nets (Collins 1979; Larkins 1963, 1964; Pristas and Trent 1977). Larkins (1964) reported that the monofilament nets in a monofilament and multifilament string caught 1.9 to 4.1 times more Pacific salmon than did the multifilament nets in the same string. A larger percentage of fish are gilled in monofilament gillnets than are in nets made of multifilament and multimonofilament, which tend to result in tangling. Thinner twines generally catch more fish (Holst et al. 2002; Hovgård 1996; Hovgård and Lassen 2000), as they are less visible and softer, but they may have poorer size selection (larger selection range) due to elongation when a fish pushes into the mesh (Hansen 1974) and ease of entanglements (Yokota et al. 2001). Turunen (1996) reported no change in size frequency but a 190% increase in catch of pikeperch (Stizostedion lucioperca) when comparing 0.15-mm twine and 0.20-mm twine. Hovgård and Lassen (2000) reported that monofilament nets with 0.16-mm twine caught 2 to 3 times more European dab (Limanda limanda) than a net with 0.28-mm twine. Holst et al. (2002) found that gillnets made of four-strand No. 1.5 multimonofilament twine (0.28-mm diameter) caught about 1.5 times that of six-strand No. 1.5 twine (0.36-mm diameter) for Baltic cod (Gadus morhua). Hovgård (1996) found that fishing efficiency was inversely related to the ratio of twine thickness to mesh size for a number of species in Greenland waters. However, nets made of thin twines are more easily damaged, which may result in increased costs and lost fishing time. Further, they may produce increased catch of undesired species such as crustaceans. Net dimension (height and length) may also affect catch, although net length generally does not alter the species or size composition of catches—it

Fish Behavior near Gillnets: Capture Processes and Influencing Factors

187

Figure 8.3. Explanation of hanging ratio of a gillnet. (He 2006a.)

merely increases effort. Fish may lead along the nets, and length may affect the likelihood of striking the mesh. Net height, however, appears to alter species selectivity and affect fishing efficiency of some species. Height in gillnets is affected by a number of factors—the amount of buoyancy in the headrope, the presence of “tie-down” lines (lines connecting footrope and headrope that restrict the height of a gillnet), and the strength of currents and resistance of the webbing. In the Northwest Atlantic, Atlantic cod are targeted with nets with a great deal of buoyancy, so that they reach their maximum height. Flatfish are targeted using headropes with less floatation and with tie-down lines. 8.3 SIZE SELECTIVITY OF GILLNETS The selection of fish by a fishing gear is the process that causes the catch of the gear to have a different composition, either of sizes or of species, than that of the population on the fishing grounds (Wileman et al. 1996). Selectivity is the quantitative assessment of this selection process. Gillnet selectivity processes, mechanisms, and analysis methods are reviewed by Hamley (1975), Millar and Fryer 1999, Hovgård and Lassen (2000), and Fujimori and Tokai (2001). Gillnet size selectivity curves are approximated as Gaussian or bell-shaped and may have two or more peaks reflecting different mecha-

nisms of capture (discussed earlier) or multiyear class population. Bimodal curves were found to provide the best fit in several studies (Fonseca et al. 2005; Madsen et al. 1999; Moth-Poulsen 2003); in others, the normal scale curve (Revill et al. 2007) or the lognormal (M. Pol, unpublished data) provided the best fit. Efficiency of gillnets is affected by mesh size, webbing material, hanging ratio, twine size, and fish behavior as discussed earlier. However, mesh size is likely the most important factor affecting gillnet size selectivity. Experiments confirm that larger meshes result in catches of more large fish, shifting the selectivity curve to the right as shown in Figure 8.4A; these results conform Baranov’s (1948) geometry similarity rule on fish size and mesh size. If the x-axis is expressed as length divided by the mesh size, selectivity can be expressed as one master curve as seen in Figure 8.4B. Gillnets generally catch larger fish compared with other gears, if the proper mesh size and netting materials are used. Comparative fishing trials have demonstrated that gillnets caught more large fish than other fishing gears. When used simultaneously on the west coast of Greenland, gillnets caught more large Greenland halibut than did longlines (Boje 1991). Although both gears caught fish of the

188

Fish Behavior near Fishing Gears during Capture Processes

Figure 8.4. Selectivity curve of gillnets of different mesh sizes for European hake (Merluccius merluccius), and master curve using transformed length. (Redrawn from data in Fonseca et al. 2005.)

same peak length of about 70 cm and had a similar length range from 45 to 115 cm, longlines caught a larger proportion of fish between 50 and 65 cm, while gillnets caught a larger proportion of fish between 65 and 85 cm. Cod on the Flemish Cap in the Northwest Atlantic are fished by several fleet sectors that can be identified by gear (trawl/mesh size, gillnet, longline) and by country (Boje 1991). Portuguese gillnetters caught the largest cod with an average weight of 2.5 kg, whereas Spanish and Portuguese freezer trawlers caught the smallest cod

with an average weight of 0.4 kg and 0.9 kg, respectively (Boje 1991). Lowry et al. (1994) compared gillnets and trawls with the same mesh sizes ranging from 105 to 130 mm targeting Baltic cod and found that gillnets caught fish with peak lengths 7 to 16 cm longer than fish caught with trawls using the same mesh size. Nedreaas et al. (1993) and Huse et al. (1999) compared 220-mm mesh size gillnets with a 135-mm codend mesh size trawl and No. 12/0 EZbaiter hooks in a longline targeting Greenland halibut in the Barents Sea off northern Norway.

Fish Behavior near Gillnets: Capture Processes and Influencing Factors They found gillnet catches were composed of mostly mature females of large size, whereas the trawl and longline had a much lower percentage of large mature females. The average length of gillnet fish was 65.9 cm, the longline caught fish that averaged 59.6 cm, and the trawl caught fish that averaged 50.1 cm. Comparison of three gear types targeting cod and haddock showed similar results (Huse et al. 2000). The selection range of gillnets are also narrower than that for other gears (Erzini et al. 2003). Santos et al. (2002) compared gillnets to longlines in a hake (Merluccius merluccius) fishery and found gillnets had a narrower size selectivity but found the longlines yielded better-quality fish, attributed to long soak times (more than 8 h) in the gillnets. 8.4 FISH BEHAVIOR AND GILLNET FISHING Fish availability, vulnerability, and mobility are the most important factors influencing fishing efficiency of stationary gears. Horizontal and vertical migrations are well known in many fish species. Diurnal vertical migration related to light levels and semidiurnal vertical excursions related to tide can affect gillnets set on the seabed. Increases in the rate of horizontal movement increases the probability of fish encountering gillnets. The amount of horizontal movement is especially important for set gillnets that await encounter on predicted fishing routes or foraging grounds. Fishing operators therefore need local knowledge of fish availability to set nets in the right place at the right time to be successful. Temperature may be the most important factor affecting distribution, movement, and swimming capacity. Vertical and horizontal temperature distribution patterns can cause localized concentrations and dispersal of fish and make them more or less vulnerable to gillnets (Perry and Neilson 1988; Rose and Leggett 1989; Woodhead 1964). The fishing range (the size of fishing area) of a gillnet and encounter rate of a fish (Engås and Løkkeborg 1994; He 2003; McQuinn et al. 1988) may be reduced at lower temperatures, influencing fishing efficiency of gillnets (Stoner 2004). Swimming speed of fish in relation to temperature has been discussed by Wardle (1975, 1980) and He (1991,

189

1993, 2003) and in Chapter 1. In general, swimming speeds are lower at lower water temperatures (Figure 8.5). He (2003) measured the rate of movement of winter flounder (Pseudopleuronectes americanus) on fishing grounds using a video camera and found the rate of movement was reduced by 70% when bottom water temperature was reduced from 4.4° to −1.2°C. He (2003) further discussed how fishing areas or the active fishing space of a gillnet may be altered due to a change in water temperature and soaking durations (Fig. 8.6). Swimming speed is also related to fish body length, as discussed in Chapter 1. Consequently, larger fish can swim faster and have larger geographical ranges than smaller fish. Gillnets (and other stationary gear) may have an intrinsic size selection property as larger fish are likely to reach the net and become available to the nets set some distance away (Fig. 8.6). This length-related difference in encounter probability in gillnets was discussed by Rudstam et al. (1984), who also applied size-related differences in encounter probability to correct abundance estimates for some freshwater species in the Laurentian Great Lakes. Other factors affecting swimming and local movement may include satiation or hunger (Robinson and Pitcher 1989) and prey density (Asaeda et al. 2001), light level (Gjelland et al. 2004), oxygen level (Beamish 1978), and spawning condition. Robinson and Pitcher (1989) found that the swimming speed of herring (Clupea harengus) was highest when they were hungry, presumably related to active food searching behavior. When prey species were plentiful (high prey density), swimming speed may be slowed down (Asaeda et al. 2001). Angelsen (1981) reported that more male spawning cod and halibut are caught by gillnets than are females because they are more active on spawning grounds. Reports of underwater observations of fish behavior near gillnets in the field are scarce and limited to freshwater or coral reef environments where shallow water and good lighting conditions provided better opportunities for observations either by the naked eyes or by underwater video cameras. Laboratory tank observations of fish capture by gillnets (Potter and Pawson 1991) revealed that Atlantic

190

Fish Behavior near Fishing Gears during Capture Processes

Figure 8.5. Reduction of activity and swimming capacity due to lower water temperature, assuming 1 at the higher end of temperature. (From He 2003.)

salmon (Salmo salar) initially struggled powerfully for less than 30 s. This powerful struggle was followed by a long period of weak activities, similar to the behavior observed in hooked fish (Bjordal and Løkkeborg 1996). Gilled fish tended to swim forward, pulling the net with them. Smaller fish may escape by squeezing through the meshes. Eleven salmon that escaped by squeezing through meshes did so in less than 25 s. Tangled fish were more likely to wrench their head or tail and to swim backward or alongside the net. The fate of capture or escape was also largely determined during the first 25 s (Potter and Pawson 1991). Fujimori et al. (1994) classified rainbow trout (Oncorhynchus mykiss) behavior in laboratory experiments in two ways: swimming straight into the net head-on or contacting the net with abdomen or tail. Visibility (or invisibility) of the net is the most important aspect of gillnet design and operation. Fish vision and underwater visual characteristics of fishing gear components have also been discussed in Chapter 2. The visibility of the net is determined by the fish’s visual characteristics, material of the net, light level and composition, water clarity, contrast of the net, and relative position of the fish to

the net. Angelsen and Huse (1979) tested seven nets made from different materials and colors and found that monofilament nylon was least visible and multifilament nylon was the most visible at various water depths. Wardle (1989) illustrated how different shaded twines hung vertically in water had different visibilities when viewed at different angles (Fig. 8.7). White twines disappeared toward the surface, black twines disappeared near the bottom, and grey twines were least visible when viewed horizontally. Transparent monofilament lines hung vertically are almost invisible when viewed horizontally (Gabriel et al. 2005). Comparison of mesh penetration or avoidance of four types of twines of different colors by Atlantic mackerel (Scomber scombrus) indicated that the fish more readily penetrate the naturally colored (transparent) monofilament netting, whereas nets made from glow twine were more likely to be avoided (Fig. 8.8) (SOAFD, 1992). Faulkner (1994) discussed a “window-pane” gillnet with a highly visible large mesh in the top section and regular gillnet webbing in the bottom section. The highly visible netting on the top was believed to drive surface-swimming fish into the deep water, where

Fish Behavior near Gillnets: Capture Processes and Influencing Factors

191

Figure 8.6. Predicted fishing range of gillnets at different water temperatures and at different soaking durations. (From He 2003.)

they were subsequently caught in the bottom part of the net. Once the fish are in the vicinity of the net, those individuals unaware of the presence of the net may swim into it and become caught. Visibility of the net is reduced when there is low contrast between the net and its background. Smaller-diameter materials are also less visible. Nighttime hours, periods with no moon, high-latitude winter days, deep water, and turbid water (near estuaries, tidal area with muddy bottom) all contribute to lower visibility of the net. Fish are not able to detect the netting easily at lower light levels, increasing the chance of

being caught by gillnets. However, lower light level conditions also cause fish to slow down (Gjelland et al. 2004), which reduces encountering probability with gillnets. The lower visibility of synthetic nets made of thin twines may be largely responsible for the increase in catch rates compared with natural fibers (Potter and Pawson 1991). Fish may also be deterred by strong smells from preservatives used in natural fibers. Vibrations from water current passing through meshes of the netting may also make fish aware of the net (Gabriel et al. 2005). Fish seeing or otherwise detecting the net may turn and swim

192

Fish Behavior near Fishing Gears during Capture Processes

Figure 8.7. Contrast of white, grey, and black twines hung vertically in water in relation to viewing angle. (Redrawn from Wardle 1989.)

parallel to the net, similar to the behavior observed in the leader of a trap (see Chapter 7). Acosta and Appeldoorn (1995) described observations of gillnet catch in relation to soak time and found that catch rate initially decreased after setting for 6 to 10 h but increased between 10 and 20 h. It was argued that the initial decline in catch efficiency may be related to reduction of fish density soon after the net was set as well as increased visibility of the net with meshed fish. Acosta and Appeldorn (1994) observed that several fish turned away after seeing a struggling fish caught in the net. The geometry of a gillnet may also be affected by the fish caught in the net, especially if it is entangled and twisted into several meshes. Grant (2002) observed walleye (Sander vitreus) using a stereo camera system in a shallow lake. He examined three mesh sizes and counted the number of fish blocked (turned away), passing through meshes, and caught in the meshes. Many small fish passed through meshes of large-mesh nets while many large fish were blocked by the small-mesh net, as expected. Of 147 fish observed, 35 were caught, while 46 swam through the meshes, 29 escaped after becoming temporarily wedged or entangled, and 39 were

blocked (turned away) and never contacted the net. These specific numbers are important in determining catchability of specific gillnets during stock assessment surveys. The possibility of increasing catches in gillnets through the use of additional stimuli has been investigated. Properties of fish attraction using bait are discussed in Chapter 5. Baited gillnets were tested in Norway (Engås et al. 2000; Kallayil et al. 2003). The idea of using bait in gillnets came from unexpectedly high catch rates when gillnets were set for a longer duration. It was postulated that fish caught at the beginning of the soak period might have acted as bait to attract more fish to the area (Engås et al. 2000). In some fisheries, such as the deepwater Greenland halibut fishery off Labrador, fishermen may set nets for longer than 2 weeks (Melindy and Flight 1992). In that research, however, catch rates of Greenland halibut were reduced and the amount of spoiled fish increased when soaked for a longer period of time. Baited gillnets caught more Atlantic cod, saithe (Pollachius virens), ling (Molva molva), and Greenland halibut in a study by Engås et al. (2000) but not in one by Kallayil et al. (2003). Analysis of tagged and acoustically tracked Atlantic cod near baited and nonbaited gillnets indicated that fish spent more time near baited gillnets and have more encounters with the nets (Kallayil et al. 2003). Fish were observed to turn and make directional movements toward the bait as far away as 800 m. Fish swam more slowly near baited gillnets than near nonbaited gillnets. Kallayil et al. (2003) attributed a decrease in catch rates to this slower swimming behavior despite more observed encounters. Slower swimming may give fish more time to avoid gillnets. The presence of bait may have interrupted swimming and made fish more aware of the net. In their experiment, Kallayil et al. (2003) noticed that relatively more cod were tangled than gilled or wedged in baited gillnets compared with nonbaited gillnets, indicating that most of fish were not directly swimming into the baited gillnets. 8.5 MEASURES TO REDUCE BYCATCH AND DISCARDS IN GILLNETS Gillnets can catch a variety of species, with many of them being considered as bycatch and subsequently

Fish Behavior near Gillnets: Capture Processes and Influencing Factors

193

Figure 8.8. Avoidance of Atlantic mackerel (Scomber scombrus) to nets of different colors. (Redrawn from data in SOAFD 1992.)

discarded. While the rate of discard from gillnet fisheries was considered low on a global scale (0.5% by weight), discard in some specific gillnet fisheries is quite high (Kelleher 2005). Morizur et al. (1996, cited in Kelleher 2005) reported that up to 100% of fish caught in offshore gillnets soaked for more than

6 days may be discarded due to poor quality. Murawski (1993) reported that 44% of fish by weight were discarded from Gulf of Maine groundfish gillnets in 1991. In this case, the majority of discards from gillnets were due to a lack of market for the bycatch species (Alverson et al. 1994).

194

Fish Behavior near Fishing Gears during Capture Processes

Bycatch of nontarget species may be reduced by understanding and using differences in the vertical and horizontal distribution of fish. Different species may occupy different levels of the water column and some spend a considerable time on the substrate. Video camera observations in the natural environment indicated that winter flounder spent 33% to 68% of time on the substrate with a larger proportion of time at lower temperatures (He 2003). The fish were never observed to rise to more than 0.6 m from the seabed. Although some flounder species take prolonged excursions to higher levels in the water column (Cadrin and Westwood 2004; Walsh and Morgan 2004), they usually reside very close to the seabed. In commercial practice, gillnets targeting flounders often have a reduced vertical height created through reduced floatation or tiedown lines, thus avoiding or reducing catch of other species that live higher off the bottom. Pol (2006) tested the effect of reduction of gillnet height through the addition of spaced weights on the headrope and using nets with double footrope and no headrope. Flatfish catch was maintained while bycatch of Atlantic cod was reduced by 49% and 58%, respectively, compared with standard flatfish gillnets in the Gulf of Maine. He (2006b) tested two low vertical height nets in the Gulf of Maine.

The 8 meshes deep (MD) experimental gillnets caught significantly less cod than the regular 25 MD net, whereas the catch efficiency for flounders (mainly American plaice [Hippoglossoides platessoides]) was similar (He 2006b). The extended gillnets with an extra 10 meshes of webbing (35 MD) caught significantly more Atlantic cod than the standard 25 MD nets in tests in Newfoundland (Yetman 1989). Norwegians use much higher gillnets (60 MD) when targeting cod (Engas et al. 2000). Similarly, nets with tie-down lines caught more flounder and other bottom-dwelling animals (e.g., lobsters) but less cod than the standard cod net due to a reduced vertical profile and a large amount of slack netting near the seabed (He 2006b). Crabs and lobsters are strongly substrateassociated and thus are often caught in groundfish gillnets (Godøy et al. 2003; He 2005, 2006b). In some jurisdictions, retention of crustacean species is prohibited in gillnet fisheries or they are so abundant as to become a nuisance. To avoid the catch of these species that live on the seabed, the footrope of a gillnet may be raised. Norwegian researchers tested a gillnet rigged with “Norsel lines” of 0.5 m long (Fig. 8.9). Although there was some reduction in the targeted Atlantic cod, the Norsel nets significantly reduced the catch of king crabs (Paralithodes

Figure 8.9. Design of the Norsel net to avoid bycatch of king crabs (Paralithodes camtschaticus) in northern Norway. (Redrawn based on Godøy et al. 2003.)

Fish Behavior near Gillnets: Capture Processes and Influencing Factors camtschaticus) (Godøy et al. 2003). Similar experiments were carried out in Newfoundland to reduce snow crab bycatch in Greenland halibut fishery (Brothers 2002). Preliminary experiments in the Gulf of Maine to reduce cod catch when targeting haddock (Melanogrammus aeglefinus) and pollock (Pollachius virens) with Norsel nets provided encouraging but limited results (Eayrs and Salerno 2008). Bycatch species escaped, released, or discarded from gillnets may experience physiological and physical impact, and eventual mortality, related to capture and discard processes (Suuronen 2005; also see Chapter 11). Buchanan et al. (2002) estimated that traditional gillnets targeting other species caused 35% to 70% of total mortality on coho salmon (Oncorhynchus kisutch) in the Pacific coast. Shorter soak durations together with an onboard recovery procedure reduced mortality rates significantly. The mortality rate from 40-min set was 6.7% and those from 140-min set was 52% to 72% after being held in net pens for 48 h (Buchanan et al. 2002). Vander Haegen et al. (2004) found an immediate survival rate of greater than 95% for spring chinook salmon (Oncorhynchus tshawytscha) caught by several tangle nets and gillnets using tagging and recapture methods, but fish released from tangle nets recovered better than those from gillnets. The recovery rate from 114-mm mesh size tangle nets was 1.9 times that of 203-mm gillnets. They found that fish in small-mesh tangle nets were often snagged by the snout rather than gilled and argued that snagging would have reduced injury compared with gilling or wedging and also allowed the fish to continue to respire. 8.6 INTERACTION OF MARINE MAMMALS, SEABIRDS, AND SEA TURTLES WITH GILLNETS Bycatch and related mortality of charismatic animals, including marine mammals, seabirds, sea turtles, and others, has created negative images for gillnets. As a result, gillnets and driftnets are banned in some areas. While the interaction of megafauna species with fishing gears and mitigation measures are discussed in detail in Chapter 13, a brief account of issues related to gillnet fishing is provided here.

195

Seabird bycatch occurs in almost all gillnet types, especially those nets set near the surface, adjacent to bird colonies, and in shallow waters (DeGange and Day 1991; Forney et al. 2001; Lewison et al. 2004; Lien et al., 1989; Melvin et al. 1999). During the capelin spawning season in Newfoundland, intense inshore feeding by birds and peak commercial fishing activities with both bottom and surface gillnets coincided, resulting in significant seabird mortality (Lien et al. 1989). The greatest bycatch of birds by gillnets was near bird breeding colonies, with diminishing bycatch as distance from the colony was increased (Lien et al., 1989). Common murres (Uria aalge) were most often caught in monofilament groundfish gillnets, whereas Atlantic puffins (Fratercula arctica) were more often caught in surface gillnets for salmon. Descriptions of mitigation measures to reduce seabird mortality are limited (Manville 2005). Hayase and Yatsu (1993) submerged high-sea driftnets 2 m below the surface and significantly reduced seabird entanglement. However, there was a substantial reduction in targeted species. Faulkner (1994) theorized that a “window-pane” gillnet containing a thicker twine and larger mesh top panel and regular gillnets underneath it might reduce seabird bycatch. Melvin et al. (1999) used a similar principle and tested a modified salmon gillnet with the top 20 meshes of the webbing made of highly visible white multifilament twine, and they were able to significantly reduce bycatch of common murre and rhinoceros auklet (Cerorhinca monocerata) (Fig. 8.10). Melvin et al. (1999) also found that acoustic pingers (1.5-kHz frequency in 4-s bursts at 120 dB re 1 μPa) attached to the headrope of a gillnet were able to reduce seabird bycatch. Because most seabird bycatch occurred at dawn and dusk and during certain times of the year, a combined measure of gear modification, time restriction, and area closure may reduce seabird bycatch by 70% to 75% without significant reduction in target species (Melvin et al. 1999). Interactions between marine mammals and gillnets can result in animal mortality and severe damage to the fish and the fishing gear (Lewison et al. 2004; Lien 1995; Lien et al. 1989; Northridge 1991). Globally, more than 80,000 small cetaceans have been reported killed annually in coastal waters,

196

Fish Behavior near Fishing Gears during Capture Processes

Figure 8.10. Design of a gillnet with high contrast white netting on the top to reduce seabird bycatch in Puget Sound sockeye salmon driftnet fishery. C, control net; E1, experimental net with 20 mesh white netting on the top; E2, experimental net with 1.5-kHz pingers. (Redrawn from data in Melvin et al. 1999.)

with many of them killed as a result of fishing activities (Jefferson and Curry 1994). According to Lien et al. (1989), an annual average of 24 humpback whales (Megaptera novaeangliae) were entrapped in Newfoundland groundfish gillnets between 1978 and 1987. Belden et al. (2006) reported that 2292 marine mammals were caught in the U.S. Northeast groundfish gillnets in 2004, with another 231 animals in the Mid-Atlantic coastal gillnets. Carretta et al. (2005) reported marine mammal, sea turtle, and bird mortality in the California driftnet fishery for swordfishes and sharks between 1996 and 2002 with a variety of mortality rates for different species. Bycatch of marine mammals by large-scale drift nets resulted in a ban of driftnet fishing in the high seas. Harbor porpoises (Phocoena phocoena) are incidentally caught in gillnets throughout their distribution range in northern waters (Perrin et al. 1994). Acoustic pingers tested in groundfish gillnets in the Gulf of Maine and Bay of Fundy reduced mortality of harbor porpoises, and the use of the devices has become mandatory in the fishery (Kraus et al. 1997; Trippel et al. 1999). Kraus et al. (1997) demonstrated that 10-kHz pingers were able to reduce porpoise catch while maintaining target species

catch of cod and pollock. However, the mechanism by which acoustic pingers were able to reduce porpoise bycatch was not clear (Dawson et al. 1998; Kraus et al. 1997). Harbor porpoise feed on herring, and it was argued that herring can hear high-frequency sound and might have avoided gillnets with pingers. Less herring near gillnets with pingers may have resulted in less porpoise bycatch (Kraus et al. 1997; Trippel et al. 1999). Additionally, Cox et al. (2004) monitored bottlenose dolphin (Tursiop truncatus) in an inshore gillnet site and found that dolphins stayed farther away from gillnets with active pingers. Borodino et al. (2002) reduced Franciscana dolphin (Pontoporia blainvillei) bycatch in gillnets using pingers but found increased pinniped depredation on target species. Mixing of barium sulfate (BaSO4) within monofilament nylon increases reflectivity of the net (Cox and Read 2004; Mooney et al. 2004) and has been reported to reduce harbor porpoise bycatch without a reduction in target species (cod, haddock, and pollock) (Trippel et al. 2003). Trippel et al. (1996) found that the majority (96%) of porpoise bycatch was on the upper twothirds of the gillnet that had a standup height of approximately 4 to 5 m. A reduced height gillnet

Fish Behavior near Gillnets: Capture Processes and Influencing Factors may have a positive effect on reducing porpoise bycatch in the gillnet fishery. For some bottomdwelling species such as flounder, the height of the net may be reduced without affecting the catch of the target species (He, 2006b). Other measures to reduce bycatch of harbor porpoises focus on prevention of entanglement through the use of stiff, neutrally buoyant or sinking ropes or on escape once entanglement occurs, through weak ropes or weak links. These measures are required in New England gillnet and pot fisheries (NOAA 2008). 8.7 DERELICT GILLNETS: GHOST FISHING PROBLEMS AND SOLUTIONS Gillnets and other fishing gears can become lost due to adverse weather or sea conditions or by conflict with other fishing gears or vessel traffic (Matsuoka et al. 2005). Gillnets, whether lost unintentionally, abandoned, or otherwise discarded at sea, have a similar effect on animals and the environment. The derelict gillnets may continue to fish for an extended period of time, causing additional mortality to fish and other organisms. This phenomenon is called “ghost fishing.” Gear designs and modification to eliminate ghost fishing or to reduce the fishing capacity of ghost gears are called “de-ghosting” technologies. With the introduction of synthetic materials in gear construction, these derelict gillnets may continue to fish for several years before they become inactive. In the Northeast Atlantic, including the Mediterranean, as many as 25,000 gillnets were reported lost each year, with loss rates as great as 3.2% (Macfadyen et al. 2009). Canadian Fishery Consultant Ltd (CFCL 1994) estimated that around 5000 gillnets were lost annually in Atlantic Canadian waters. Gear loss in this region may be aggravated by increased use in deep waters for species and in more hostile sea conditions such as in the Greenland halibut fishery. Cooper et al. (1988) conducted video camera surveys by a remotely controlled underwater vehicle in the Gulf of Maine and estimated that there might be 2497 nets (91 m each) on a 64-nm2 area of traditional gillnet grounds on Stellwagen Bank and Jeffries Ledge, equivalent to 39 derelict nets per square nautical mile.

197

Direct observations of derelict gillnets or simulated lost gillnets confirm that these nets continue to fish (Carr et al. 1985; Cooper et al. 1988). Gillnets deliberately set over wrecks in U.K. coastal waters continued to catch and kill fish for at least 2 years (Revill and Dunlin 2003). It was estimated that lost salmon nets might fish for 2 years for fish and 6 years for crabs (High 1985). In shallower waters, derelict gillnets may become overgrown by algae, or “biofouled.” Because these algae-laden nets are more visible, their fishing capacity is correspondingly reduced. Takagi et al. (2007) estimated that the height of bottom gillnets declined rapidly after 15 days and to zero after 25 days of deployment. Fishing capacity decreased to about 15% to 20% of a typical net in the first few weeks (Carr and Cooper 1988; Revill and Dunlin 2003). However, shallow water is usually rich in marine life, so catch rates can still be considerable. Erzini et al. (1997) set “damaged” gillnets in 15 and 18 m off southern Portugal. They estimated that the lifetime of a ghost gillnet was between 15 and 20 weeks. Observations after 8 to 11 months indicated complete destruction or heavy colonization by algae resulting in incorporation into a reef. They estimated that a lost 100-m length of gillnet will catch 314 fish over a 17-week period (Erzini et al. 1997) but thought that this number was likely an underestimate due to predation and scavenging. Prevention of ghost fishing may include prevention of gear loss, derelict gear retrieval, and deghosting technologies. Inshore gillnets use floats on the headrope and leadline on the footrope to spread the net vertically. Therefore, use of degradable material that causes the lost gillnet to lose floatation could reduce the vertical profile and hence fishing capacity. Carr et al. (1992) tested degradable plastic plates for attaching floats to the headrope of gillnets (Fig. 8.11). The gear was set to simulate ghost fishing for a period of 220 days. Two types of degradable plastic plates were used. Divers made underwater observations to check net profiles and catch. Only 2 of the 20 degradable attachment panels partially degraded after 220 days—apparently the panel can be modified to increase degrading process. No significant differences in catch were observed between sections of the net rigged with degradable float attachments and those with

198

Fish Behavior near Fishing Gears during Capture Processes

Figure 8.11. Degradable plastic panel for attaching floats to the headrope of a gillnet. (He 2006a.)

regular rigging. Although there was a report of degradable fishing nets being developed in Japan (Anon. 1993), the commercial application of degradable nets has not been implemented. The final attempt to reduce ghost fishing is to retrieve the derelict gillnets and “clean up” the fishing grounds. Several countries conduct gear retrieval operations regularly (see Macfadyen et al. 2009; Matsuoka et al. 2005). Mandatory reporting of gear loss and the use of acoustic transponders can facilitate retrieval of nets if they become lost. 8.8 CONCLUDING REMARKS Gillnets are very size selective, landing only a narrow range of fish size. Size selectivity of gillnets is closely related to mesh size and changes in type of webbing material, twine size, and hanging ratio. Although gillnets are simple in design and operation, the behavior of fish near gillnets and their capture processes are not well understood. Research is needed to better understand the influence of tide and other factors on net height and on vertical distribution of target species. There are two major conservation challenges facing gillnet fisheries. Gillnets have poor species selectivity, resulting in bycatch and discards and causing mortalities to marine mammals, sea birds, and turtles. Research on mitigation measures has shown progress, includ-

ing successful implementation of pinger use in some fisheries, but more work is needed. Ghost fishing of derelict gillnets remains a challenge, except in the countries that conduct active retrieval of derelict net. REFERENCES Acosta AR and Appeldoorn RS. 1995. Catching efficiency and selectivity of gillnets and trammel nets in coral reefs from southwestern Puerto Rico. Fish. Res. 22: 175–196. Alverson D, Freeberg M, Murawski S and Pope J. 1994. A global assessment of fisheries bycatch and discards. FAO Fish. Tech. Pap. 339: 233 pp. Angelsen KK and Huse I. 1979. Qualitative contrast comparison of different colored gillnets. ICES CM. 1979/B: 20. 3 pp. Angelsen KK. 1981. Engineering and fish reaction aspects of gillnetting—a review. ICES CM. 1981/B: 34. Angelsen KK, Haugen K and Floen S. 1979. The catching efficiency of cod gillnets with different hanging ratio (E) and different floatline buoyancy. ICES CM. 1979/B: 19. Anon. 1993. Biodegradable fishing nets. World Fishing. Oct. 1993. Asaeda T, Priyadarshann T and Manatunge J. 2001. Effects of satiation on feeding and swimming behavior of planktonvores. Hydrobiologia. 443: 147–157.

Fish Behavior near Gillnets: Capture Processes and Influencing Factors Baranov FI. 1948. Theory and assessment of fishing gear. Pishchepromizdat, Moscow. (Translated from Russian by the Ontario Department of Lands and Forests). Beamish FWH. 1978. Swimming capacity. In: Fish Physiology. Vol. 7. Locomotion. pp 161–187. Hoar WS and Randall DJ (eds.). New York: Academic Press. Belden DL, Orphanides CD, Rossman MC and Palka DL. 2006. Estimates of cetacean and seal bycatch in the 2004 northeast sink gillnet and mid-Atlantic coastal gillnet fisheries. NOAA Northeast Fisheries Science Center. Ref. Doc. 06-13: 24 pp. Boje J. 1991. A comparison of selectivity in longlines and gillnets in the fishery for Greenland halibut in West of Greenland. NAFO SCR Doc. 91/39. Borodino P, Kraus S, Albareda D, Fazio A, Palmeiro A, Mendez M and Botta S. 2002. Reducing incidental mortality of Franciscana dolphin Pontoporeia blainvillei with acoustic warning devices attached to fishing nets. Mar. Mam. Sci. 18: 833–842. Bjordal Å and Løkkeborg S. 1996. Longlining. Oxford: Fishing News Books. 156 pp. Brothers G. 2002. Reducing snow crab bycatch in turbot gillnets. Project Summary. EACT-10.2002. DFO (FDP 299). Brothers G and Yetman L. 1982. A study to determine the impact of mesh size on the redfish catches. Department of Fisheries and Ocean Newfoundland Region. Industrial Development Branch. FDN1981/82 -06. 5 pp. Buchanan S, Farrell AP, Fraser J, Gallaugher P, Joy R and Routledge R. 2002. Reducing gillnet mortality of incidentally caught coho salmon. N. Am. J. Fish. Manag. 22: 1270–1275. Cadrin SX and Westwood AD. 2004. The use of electronic tags to study fish movement: a case study with yellowtail flounder off New England. ICES CM. 2004/K:81. Carr HA and Cooper RA. 1988. Manned submersible and ROV assessment of ghost gillnets in the Gulf of Maine. IEEE Conf. Proc. 1987: 622–625. Carr HA, Amaral EA, Hulbert AW and Cooper R. 1985. Underwater survey of simulated lost demersal and lost commercial gill nets off New England. NOAA Tech. Memo. NMFS. 54. Carr HA, Blott AJ and Caruso PG. 1992. A study of ghost gillnets in the inshore waters of southern New England. Mar. Technol. Soc. Conf. Proc. 1992: 361–367. Carretta JV, Price T, Peterson D and Read R. 2005. Estimates of marine mammal, sea turtle, and seabird

199

mortality in the California drift gillnet fishery for swordfish and thresher shark, 1996–2002. Mar. Fish. Rev. 66(2): 21–25. CFCL. 1994. Review of fishing gear and harvesting technology in Atlantic Canada. Ottawa, Canada: Fisheries and Oceans Canada, Fishing Industry Services Branch, Fishing Operations. Collins JJ. 1979. Relative efficiency of multifilament and monofilament nylon gillnet towards lake whitefish (Coregonus clupeaformis) in Lake Huron. J. Fish. Res. Bd Can. 36: 1180–1185. Cooper RA, Carr HA and Hulbert AH. 1988. Manned submersible and ROV assessment of ghost gillnets on Jefferies and Stellwagen Banks, Gulf of Maine. NOAA Undersea Res. Prog. Res. Rep. 88–4. Cox TM and Read AJ. 2004. Echolocation behavior of harbor porpoises Phocoena phocoena around chemically enhanced gill nets. Mar. Ecol. Prog. Ser. 279: 275–282. Cox TM, Read AJ, Swanner D, Urian K and Waples D. 2004. Behavioral responses of bottlenose dolphin, Tursiops truncatus, to gillnets and acoustic alarms. Biol. Conserv. 115: 203–212. Dawson SM, Read A and Slooten E. 1998. Pingers, porpoises and power: uncertainties with using pingers to reduce bycatch of small cetaceans. Biol. Conserv. 84: 141–146. DeGange AR and Day RH. 1991. Mortality of seabirds in the Japanese land-based gillnet fishery for salmon. The Condor. 93: 251–258. Eayrs S. and D.J. Salerno. 2008. Testing raised—webbing gillnets to reduce bycatch of cod while targeting pollock. Online: http://northeastconsortium.org. Accessed: 05/31/2009. Durham, NH: Northeast Consortium. Engås A and Løkkeborg S. 1994. Abundance estimation using bottom gillnet and longline—the role of fish behavior. In: Marine Fish Behavior in Capture and Abundance Estimation. pp 134–165. Fernö A. and Olsen S (eds). Oxford: Fishing News Books. Engås A, Jørgensen T and Angelsen KK. 2000. Effect on catch rate of baiting gillnets. Fish. Res. 45: 265–270. Erzini K, Monteiro CC, Ribeiro J, Santos MN, Gaspar M, Monteiro P and Borges TC. 1997. An experimental study of gill net and trammel net “ghost fishing” off the Algarve (southern Portugal). Mar. Ecol. Prog. Ser. 158: 257–265. Erzini K, Gonçalves JMS, Bentes L, Lino PG, Ribeiro J and Stergiou KI. 2003. Quantifying the roles of competing static gears: comparative selectivity of longlines and monofilament gill nets in a multi-

200

Fish Behavior near Fishing Gears during Capture Processes

species fishery of the Algarve (southern Portugal). Sci. Mar. 67: 341–352. Faulkner G. 1994. Getting the most out of your gillnet. National Fisherman, Sept. 1994: 32–33. Fonseca P, Martins R, Campos A and Sobral P. 2005. Gill-net selectivity off the Portuguese western coast. Fish. Res. 73: 323–339. Forney KA, Benson SR and Cameron GA. 2001. Central California gillnet effort and bycatch of sensitive species 1990–1998. In: Proc. Seabird Bycatch: Trend, Roadblocks, and Solutions. Univ. Alaska Sea Grant, AK-SG-01-01. pp. 141–160. NOAA. 2008. Guide to the Atlantic large whale take reduction plan. Online: http://www.nero.noaa. gov/whaletrp/plan/ALWTRPGuide.pdf. Accessed: 05/31/2009. US National Oceanic and Atmospheric Administration. Fujimori Y, Tokai T and Matuda K. 1994. Effect of diurnal activity of rainbow trout and light intensity on gillnet catching in water tank experiments. Nippon Suisan Gakkaishi. 60: 577–583. Fujimori Y and Tokai T. 2001. Estimation of gillnet selectivity curve by maximum likelihood method. Fish. Sci. 67: 644–654. Gabriel O, Lange K, Dahm E and Wendt T. 2005. Von Brandt’s Fish Catching Methods of the World. 4th ed. Oxford: Blackwell. 523 pp. Gjelland KØ, Bøhn T, Knudsen FR and Amundsen P-A. 2004. Influence of light on the swimming speed of coregonids in subarctic lakes. Ann. Zool. Fennici. 41: 137–146. Godøy H, Furevik D and Løkkeborg S. 2003. Reduced bycatch of red king crab (Paralithodes camtschaticus) in the gillnet fishery for cod (Gadus morhua) in northern Norway. Fish. Res. 62: 377–384. Grant GC, Radomski P and Anderson CS. 2002. Using underwater video to directly estimate gear selectivity: the retention probability for walleye (Sander vitreus) in gill nets. Can. J. Fish. Aquat. Sci. 61: 168–174. Hamley JM. 1975. Review of gillnet selectivity. J. Fish. Res. Bd Can. 32: 1943–1964. Hansen RG. 1974. Effect of different filament diameters on the selective action of monofilament gillnets. Trans. Am. Fish. Soc. 103: 386–387. Hayase S and Yatsu A. 1993. Preliminary report of a squid subsurface driftnet experiment in the North Pacific during 1991. Int. North Pac. Fish. Comm. Bull. 53: 557–576. He P. 1991. Swimming endurance of cod, Gadus morhua L. at low temperatures. Fish. Res. 12: 65–73.

He P. 1993. Swimming speeds of marine fish in relation to fishing gears. ICES Mar. Sci. Symp. 196: 183–189. He P. 2003. Swimming behavior of winter flounder (Pleuronectes americanus) on natural fishing grounds as observed by an underwater video camera. Fish. Res. 60: 507–514. He P. 2005. Characteristics of bycatch of porcupine crabs, Neolithodes grimaldii (Milne-Edwards and Bouvier, 1894) from deepwater turbot gillnets in the northwest Atlantic: Fish. Res. 74: 35–43. He P. 2006a. Gillnets: gear design, fishing performance and conservation challengers. Mar. Technol. Soc. J. 40(3): 11–18. He P. 2006b. Effect of the headline height of gillnets on species selectivity in Gulf of Maine. Fish. Res. 78: 252–256. High WL. 1985. Some consequences of lost fishing gear. Proceedings of the Workshop on the Fate and Impact of Marine Debris, 26-29 November 1984, Honolulu, Hawaii. NOAA—TM—NMFS— SWFC—54: 430–437. Holst R, Wileman D and Madsen N. 2002. The effect of twine thickness on the size selectivity and fishing power of Baltic cod gill nets. Fish. Res. 56: 303–312. Hovgård H 1996. A two-step approach to estimating selectivity and fishing power of research gill nets used in Greenland waters. Can. J. Fish. Aquat. Sci. 53: 1007–1013. Hovgård H and Lassen H. 2000. Manual on estimation of selectivity for gillnet and longline gears in abundance surveys. FAO Fish. Tech. Pap. 397: 84 pp. Huse I, Gundersen AC and Nedreaas KH. 1999. Relative selectivity of Greenland halibut (Reinhardtius hippoglossoides, Walbaum) by trawls, longlines and gillnets. Fish. Res. 44: 75–93. Huse I, Løkkeborg S and Soldal AV. 2000. Relative selectivity in trawl, longline and gillnet fisheries for cod and haddock. ICES J. Mar. Sci. 57: 1271–1282. Jefferson TA and Curry BE. 1994. A global review of porpoise (Cetacea: Phocoenidae) mortality in gillnets. Biol. Conserv. 67: 167–183. Kallayil JK, Jørgensen T, Engås A and Fernö A. 2003. Baiting gill nets—how is fish behavior affected? Fish. Res. 61: 125–133. Kelleher K. 2005. Discards in the world’s marine fisheries. An update. FAO Fish. Tech. Pap. 470: 131 pp. Kraus SD, Read AJ, Solow A, Baldwin K, Spradlin T, Anderson E and Williamson J. 1997. Acoustic alarms reduce porpoise mortality. Nature 388: 525.

Fish Behavior near Gillnets: Capture Processes and Influencing Factors Larkins HA. 1963. Comparison of salmon catches in monofilament and multifilament gill nets. Commer. Fish. Rev. 25(5): 1–11. Larkins HA. 1964. Comparison of salmon catches in monofilament and multifilament gill nets—Part II. Commer. Fish. Rev. 26(10): 1–7. Lewison RL, Crowder LB, Read AJ and Freeman SA. 2004. Understanding impacts of fisheries bycatch on marine megafauna. Trends Ecol. Evolut. 19: 598–604. Lien J. 1995. Conservation aspects of fishing gear: cetaceans and gillnets. In: Solving bycatch: Considerations for today and tomorrow. pp 219– 224. Alaska Sea Grant College Program Report 9603. University of Alaska Fairbanks. Lien J, Stenson GB and Ni IH. 1989. A review of incidental entrapment of seabirds, seals and whales in inshore fishing gear in Newfoundland and Labrador: a problem for fishermen and fishing gear designers. Proc. World Symp. Fish. Gear and Fish. Vessel Design. pp 67–71. St. John’s, Newfoundland: Marine Institute. Lowry N, Knudsen LH and Wileman DA. 1994. Mesh size experiments in the Baltic cod fishery. ICES CM. 1994/B: 13. Macfadyen G, Huntington T and Cappell R. 2009. Abandoned, lost or otherwise discarded fishing gear. FAO Fish. Aquacult. Tech. Pap. 523: 115 pp. Madsen N, Holst R, Wileman D and Moth-Poulsen T. 1999. Size selectivity of sole gill nets fished in the North Sea. Fish. Res. 44: 59–73. Manville AM, II. 2005. Seabird and waterbird bycatch in fishing gear: next steps in dealing with a problem. USDA Forest Service. Gen. Tech. Rep. PSW-GTR191: 1071–1082. Matsuoka T, Nakashima T and Nagasawa N. 2005. A review of ghost fishing: scientific approaches to evaluation and solutions. Fish. Sci. 71: 691–702. McQuinn IH, Gendron L and Himmelman JH. 1988. Area of attraction and effective area fished by a whelk (Buccinum undatum) trap under variable conditions. Can. J. Fish. Aquat. Sci. 45: 2054–2060. Millar RB and Fryer RJ. 1999. Estimating the sizeselection curves of trawls, traps, gillnets, and hooks. Rev. Fish. Biol. 9: 89–116. Melindy S and Flight J. 1992. Development of a deep water turbot fishery by inshore gillnetters. CAFID Report. 30 pp. Melvin EF, Parrish JK and Conquest LL. 1999. Novel tools to reduce seabird bycatch in coastal gillnet fisheries. Conserv. Biol. 13: 1386–1397.

201

Mooney TA, Nachtigall PE and Au WWL. 2004. Target strength of a nylon monofilament and an acoustically enhanced gillnet: predictions of biosonar detection ranges. Aquat. Mam. 30: 220–226. Morizur Y, Pouvreau N and Guénolé A. 1996. Les rejets dans la pêche artisanale française de Manche occidentale. Plozané, France: IFREMER. 123 pp. Moth-Poulsen T. 2003. Seasonal variations in selectivity of plaice trammel nets. Fish. Res. 61: 87–94. Murawski SA. 1993. Factors influencing bycatch and discard rates: analysis from multispecies/multifishery sea sampling. NAFO SCR Doc. 93/115. 17 pp. Nedelec C and Prado J. 1990. Definition and classification of fishing gear categories. FAO Fish. Tech. Pap. 222 (Rev 1): 92 pp. Nedreaas K, Soldal AV and Bjordal Å. 1993. Performance and biological implication of a multigear fishery for Greenland halibut (Reinhardtius hippoglossoides). NAFO SCR Doc. 93/118. 15 pp. Northridge SP. 1991. An updated world review of interactions between marine mammals and fisheries. FAO Fish. Tech. Pap. 251 (Suppl. 1): 219 pp. Perrin WF, Donovan G P and Barlow J. (eds.). 1994. Gillnets and Cetaceans: Report of the International Whaling Commission, Special Issue 15. Cambridge: International Whaling Commission. Perry RI and Neilson JD. 1988. Vertical distributions and trophic interactions of age-0 Atlantic cod and haddock in mixed and stratified waters of Georges Bank. Mar. Ecol. Prog. Ser. 49: 199–214. Pol M. 2006. Testing of Low-Profile, Low CodBycatch Gillnets: Phases I and II. Report to the Northeast Consortium, Durham, NH, USA. Online: http://www.northeastconsortium.org. Accessed 05/31/2009. Durham, NH: Northeast Consortium. Pol M and Carr HA. 2000. Overview of gear developments and trends in the New England commercial fishing industry. Northeast. Naturalist. 7(4): 329–336. Potter ECE and Pawson MG. 1991. Gill netting. Ministry of Agriculture, Fisheries and Food, Directorate of Fisheries Research, Lowestoft, Laboratory Leaflet, No. 69. 34 pp. Pristas PJ and Trent L. 1977. Comparisons of catches of fishes in gillnets in relation to webbing material, time of day, and water depth in St. Andrew Bay, Florida. Fish. Bull. 75(1): 103–108. Revill AS and Dunlin G. 2003. The fishing capacity of gillnets lost on wrecks and on open ground in UK coastal waters. Fish. Res. 64: 107–113. Revill A, Cotter J, Armstrong M, Ashworth J, Forster R, Caslake G and Holst R. 2007. The selectivity of

202

Fish Behavior near Fishing Gears during Capture Processes

the gill-nets used to target hake (Merluccius merluccius) in the Cornish and Irish offshore fisheries. Fish. Res. 85: 142–147. Robinson CJ and Pitcher TJ. 1989. The influence of hunger and ration level on shoal density, polarization and swimming speed of herring, Clupea harengus L. J. Fish Biol. 34: 631–633. Rose GA and Leggett WC. 1989. Interactive effects of geophysically-forced sea temperatures and prey abundance on mesoscale coastal distributions of a marine predator, Atlantic cod (Gadus morhua). Can. J. Fish. Aquat. Sci. 46: 1904–1913. Rudstam LG, Magnuson JJ and Tonn WM. 1984. Size selectivity of passive fishing gear: a correction for encounter probability applied to gillnets. Can. J. Fish. Aquat. Sci. 41: 1252–1255. Samaranayaka A, Engås A and Jørgensen T. 1997. Effects of hanging ratio and fishing depth on the catch rates of drifting tuna gillnets in Sri Lankan waters. Fish. Res. 29: 1–12. Santos MN, Gaspar MB, Monteiro CC and Vasconcelos P. 2002. Gill net and long-line comparisons in a hake fishery: the case of southern Portugal. Sci. Mar. 66: 433–441. Scholten GD and Bettoli PW. 2006. Lack of size selectivity for paddlefish captured in hobbled gillnets. Fish. Res. 83: 355–359. SOAFD. 1992. Marine Laboratory Annual Review 1990–1991. Scottish Office Agriculture and Fisheries Department, Aberdeen, UK. 84 pp. Stewart PAM. 1987. The selectivity of slackly hung cod gillnets constructed from three types of twine. J. Cons. Int. Explor. Mer. 43: 189–193. Stoner AW. 2004. Effect of environmental variables on fish feeding ecology: implications for the performance of baited fishing gear and stock assessment. J. Fish Biol. 65: 1445–1470. Sulaeman M, Matsuoka T and Kawamura G. 2000. Effect of hanging ratio on size selectivity of gillnet. Nippon Suisan Gakkaishi. 66: 439-445. (In Japanese with English abstract). Suuronen P. 2005. Mortality of fish escaping trawl gears. FAO Fish. Tech. Pap. 478: 72 pp. Takagi T, Shimizu T, and Korte H. 2007. Evaluating the impact of gillnet host fishing using a computational analysis of the geometry of fishing gear. ICES J. Mar. Sci. 64: 1517–1524. Trippel EA, Wang JY, Strong MB, Carter LS and Conway JD. 1996. Incidental mortality of harbor

porpoise (Phocoena phocoena) by the gillnet fishery in the lower Bay of Fundy. Can. J. Fish. Aquat. Sci. 53: 1294–1300. Tripple EA, Strong MB, Terhune JM and Conway JD. 1999. Mitigation of harbor porpoise (Phoceona phocoena) bycatch in the gillnet fishery in the lower Bay of Fundy. Can. J. Fish. Aquat. Sci. 56: 113–123. Trippel E, Holy N, Palka D, Shepherd T, Melvin G and Terhune J. 2003. Acoustic reflective net reduces porpoise mortality. Mar. Mam. Sci. 19: 40–43. Turunen T. 1996. The effects of twine thickness on the catchability of gillnets for pikeperch (Stizostedion lucioperca (L.)). Ann. Zool. Fennici. 33: 621–625. Vander Haegen GE, Ashbrook CE, Yi KW and Dixon JF. 2004. Survival of spring Chinook salmon captured and released in a selective commercial fishery using gill nets and tangle nets. Fish. Res. 68: 123–133. Walsh SJ and Morgan MJ. 2004. Observations of natural behavior of yellowtail flounder derived from data storage tags. ICES J. Mar. Sci. 61: 1151–1156. Wardle CS. 1975. Limit of fish swimming speed. Nature 225: 725–727. Wardle CS. 1980. Effect of temperature on the maximum swimming speed of fishes. In: Environmental Physiology of Fish. pp 519–531. Ali MA (ed). New York: Plenum. Wardle CS. 1989. Understanding fish behavior can lead to more selective fishing gears. Proc. World Symp. Fish. Gear and Fish. Vessel Design. pp 12– 18. St. John’s, Newfoundland: Marine Institute. Wileman D, Ferro RST, Fonteyne R and Millar R. 1996. Manual of methods of measuring the selectivity of towed fishing gears. ICES Coop. Res. Rep. 215: 126 pp. Woodhead PMJ. 1964. Changes in the behaviour of the sole, Solea vulgaris, during cold winters, and the relation between the winter catch and sea temperatures. Helgol. Wiss. Meeresunters. 10: 328–342. Yetman L. 1989. Comparison of extended and standard gillnets for harvesting Atlantic cod. DFO NFLD Atl. Fish. Dev. Proj. Sum. 16: 3 pp. Yokota K, Fujimori Y, Shiode D and Tokai T. 2001. Effect of thin twine size on gill net size-selectivity analyzed with the direct estimation method. Fish. Sci. 67: 851–856.

Fish Behavior near Gillnets: Capture Processes and Influencing Factors SPECIES MENTIONED IN THE TEXT American plaice, Hippoglossoides platessoides Atlantic cod, Gadus morhua Atlantic mackerel, Scomber scombrus Atlantic puffin, Fratercula arctica Atlantic salmon, Salmo salar Baltic cod, Gadus morhua bottlenose dolphin, Tursiop truncatus chinook salmon, Oncorhynchus tshawytscha coho salmon, Oncorhynchus kisutch common murre, Uria aalge European dab, Limanda limanda Franciscana dolphin, Pontoporia blainvillei Greenland halibut, Reinhardtius hippoglossoides haddock, Melanogrammus aeglefinus hake, Merluccius merluccius

203

harbor porpoise, Phocoena phocoena herring, Clupea harengus humpback whale, Megaptera novaeangliae king crab, Paralithodes camtschaticus ling, Molva molva paddlefish, Polyodon spathula pikeperch, Stizostedion lucioperca pollock, Pollachius virens rainbow trout, Oncorhynchus mykiss rhinoceros auklet, Cerorhinca monocerata saithe, Pollachius virens shearwater, Puffinus spp. skipjack tuna, Katsuwonus pelamis winter flounder, Pseudopleuronectes americanus walleye, Sander vitreus yellowfin tuna, Thunnus albacares

Chapter 9 Electric Senses of Fish and Their Application in Marine Fisheries Hans Polet 9.1 INTRODUCTION Discussion of the use of electricity in fisheries can be traced to the middle of eighteenth century. In 1765, Job Baster, in his work “Natuurkundige Uitspanning—Part 2,” speculated that electricity would have an effect on shrimp as electric shocks had similarities to the shocks produced by the electric eel, and he proposed further investigation (De Groot and Boonstra 1974). However, it was not until the middle of the nineteenth century that the first experiment with electric fields on animals in water was carried out. In 1863, a British patent was granted to Isham Baggs for electric fishing (Snyder 2003). At that time, the four basic reactions of fish to electric fields had been distinguished: fright, anodic electrotaxis, tetanus, and electrocution. Research and application of electricity in fisheries were started in the freshwater environment in the 1920s and 1930s (e.g., Scheminzky 1924). A successful and widely used application of electric fishing (electrofishing) in freshwater followed as a research and management technique. Nowadays it is mainly used in shallow rivers, brooks, and lakes for stock assessment; fish health surveys; tagging; catching broodstocks; and eliminating undesirable species. The technique has been used around the world, but there has been growing concern among fishery biologists and managers regarding its potential for harming fish and other marine organisms. Snyder (2003) pointed out that electrofishing involves a very dynamic, complex, and often misunderstood mix of physics, physiology, and behavior. Although often not externally obvious or fatal,

spinal injuries and associated hemorrhages sometimes have been documented in over 50% of fish examined internally. Other harmful effects, such as bleeding at the gills or vent and excessive physiological stress, also are of concern and electrofishing over spawning grounds can harm embryos. The research of B. M. Bary (1956) on the behavior of marine fish in electric fields was the start of a boom in the research into applications of electricity in capturing fish at sea. The possibility of increasing shrimp catches with electricity was started in the United States in the early 1950s. During the next four decades, this work was also taken up in Europe, the former USSR, India, and China (van Marlen 1997). Studies of fish behavior in electric fields in the laboratory and in the sea resulted in electrofishing techniques for marine fisheries. Commercial application was, however, rarely reported. By the end of the 1980s, all work in this field was stopped in Europe due to a ban on the method by the European Union. In contrast, the fishing method flourished in China, where, by the end of the twentieth century, more than 3000 vessels used electric fields to capture shrimps. However, mismanagement and abuse of the method led to a national ban in 2001 (Yu et al. 2007). The only countries where the literature indicates that marine electrofishing is still considered as a realistic option for future commercial application are the Netherlands (van Marlen et al. 2006) and Belgium (Polet 2003). Fifty years of research worldwide to find ways of using electricity in marine fishing and great

205

206

Fish Behavior near Fishing Gears during Capture Processes

spending of financial resources have led to a very poor outcome, not in terms of knowledge but in terms of commercial products. In going through the literature, it is stunning to see so many positive reports of field trials that never led to any longerterm commercial application. If any future new design is to be successful, it will have to do better in both prospects for economic performance and minimization of ecosystem effects. It will have to take into account the complexity of electrofishing, the vulnerability of the technology at sea, and the often-unnoticed side effects on biota.

9.2 PROPERTIES OF AN ELECTRIC FIELD IN WATER Figure 9.1 shows a variety of types of electric currents that can be used for electric fishing (Snyder, 2003). Alternating current (AC) and direct current (DC) are the two main types that are not interrupted. Those that can be interrupted are referred to as pulsed current. Interrupted AC current is rarely used for electrofishing. Interrupted DC currents will be referred to as pulsed DC or PDC. Pulsed waveforms generated by electrofishing equipment occur in a wide variety, most often square, half-sine, quarter-sine, or capacitor discharge. According to Snyder (2003), the frequencies applied in the field (mainly rivers and lakes) lie between 15 and 120 Hz but experimentally they can range from 1 to about 500 Hz. The pulse pattern can be simple or complex. In the simple pattern, the same single pulse is generated at a certain frequency, but the complex pattern consists of a high primary frequency interrupted secondarily at a much slower frequency, thus producing bursts or trains of higher frequency pulses. The main variable determining the reaction of fish is the electric field itself (i.e., the strength of the electric field). Field strength can be described by the voltage gradient (E in V/cm) and/or the current density (J in μA/cm2) and/or power density (D in W). In the case of an interrupted electric field, the pulse type, the frequency (Hz) and pulse duration or length (ms) are also determining factors. Duty cycle is sometimes used as a characteristic for the electric field, such as in studies on the effect of electrofishing on the health of fish. It depends on the frequency and the pulse duration and the time

during which the electric field is active expressed as a percentage. In natural waters, electric fields are usually heterogeneous—that is, the current and thus the field lines do not run parallel to each other, depending on the shape of sources (electrodes) and the medium between them. In such fields, the field lines radiate and spread widely around and between the electrodes (Fig. 9.2). In the laboratory, an almost homogeneous field can easily be created by using two platelike electrodes in a tank with sides having the same dimensions as the electrodes. In this case, the current flows parallel to the sides of the tank directly from one electrode to the other, providing a constant voltage gradient, current density, and power density. A homogeneous field simplifies experimental conditions and is ideal for lab experiments, but it may be difficult to extrapolate to commercial electrofishing operations, during which the electric fields will usually be heterogeneous. Depending on the conductive properties and porosity of the substrate, the electric field can extend into the bottom. Haskell (1954) and Zalewski and Cowx (1990) found that a substrate with fine particles and organic material has a better conductivity (the electric field will penetrate deeper) compared with coarse gravel and rubble. Snyder (2003) also noted that interactions between the water and an object or substrate with different conductivity distort the electric field (Fig. 9.3). At the boundary between the water and the object, the current in the water near its surface progressively concentrates (current density increases, voltage gradient declines) if the adjacent medium is more conductive than the water and vice versa. This is the case for fish and boats made of metals. It is therefore advised against using metal boats in the vicinity of an electric field. B. M. Bary (1956) states that to obtain electric control of fish in freshwater lakes and streams, high field strength is required because the conductivity of the fish is higher than that of the water. This is, however, more than offset by the reduced current needed to maintain the field strength. Thus, the overall power requirement for obtaining a particular response is comparatively low. Seawater, by contrast, has a high conductivity compared with that of the fish so that a high current is required to maintain a given field strength and therefore the overall

Sine wave, alternating current (AC)

A

Volts

Volts

+ 0

Direct current (DC), smooth, generated by a battery or DC generator

B

+

– 0 Time C

Time D

DC, rippled, generated by partially filtered, full-wave, rectified AC

Half-sine, full-wave pulsed direct current (PDC), generated by unfiltered, half-wave, rectified DC +

Volts

Volts

+

0

0 Time E

Time F

Half-sine, full-wave PDC, generated by unfiltered, full wave, rectified AC

Square PDC, generated by interrupting smooth or rippled DC +

Volts

Volts

+

0

0 Time G

Time H

Quarter-sine, half-wave PDC, generated by controlled, half-wave, rectified AC

Exponential PDC, generated by capacitor discharge +

Volts

Volts

+

0

0 Time I

Time J

Square pulse train (PDC) (Coffelt's CPS©)

Hybrid pulsed direct current (PDC/DC), PDC on top of a DC base +

Volts

Volts

+

0

0 Time

Figure 9.1.

Time

Different waveforms used in electrofishing. (Snyder 2003.)

207

+100

–100 50

–30 20 10

–20 5

0

5

–10

Closely spaced spherical electrodes Current lines

Constant voltage lines

+100 50 20 10

Isolated Spherical Electrode Note: Total current between any two current lines is the same; zone illustrations assume both polerites are equally effective AC accelerator fields. Voltage levels are for illustration only. Flectrade

Danger zone

Differential zone

Figure 9.2. Hypothetical three-dimensional diagrams of heterogeneous electric fields around and between electrodes. Contrary to the diagrams, current flows from negative to positive electrodes. (Novotny 1990.)

208

Electric Senses of Fish and Their Application in Marine Fisheries

γ1<γw

γ1<γw

Figure 9.3. Distortion of homogenous electric fields around fish in water that is less conductive (top) and more conductive (bottom) than the fish. Horizontal lines are current (flux) lines. (Snyder 2003; adapted from Sternin et al. 1976.)

power consumption necessary to induce reactions similar to those obtained in freshwater is greatly increased. It is therefore necessary, for any practical application in seawaters, that PDC is used instead of continuous current.

9.3 THE ELECTRIC FIELD 9.3.1 General Observations There is a wide range of possible applications to fish with electricity and the different variables

209

determining its effect make it a complex method. Different animals show different reactions in electric fields (e.g., fish and crustaceans) and different fish species often show different reactions in a given electric field depending on the species, the size, and orientation in the field. Fish reactions in electric fields are often described with different terminology, which can be confusing when going through the literature over several decades. In general, however, the reaction of fish to an electric field with increasing field strength can be described as follows: • Natural mobility with lack of orientation • High mobility (first reaction, occasional tremors, dashes to the side); the fish turns perpendicular to the field lines. • Low mobility (retarded swimming, tremors, disintegration of schools, partial immobility). • Forced swimming (anodic electrotaxis) • Pseudo-forced swimming (unbalanced swimming, convulsive muscle contractions, partial loss of equilibrium); directed swimming (anodic or cathodic galvanotaxis, frightening effect) • State of shock, circular motion without escape from the field. • Death Often the three basic reactions are used—fright, electrotaxis, and electronarcosis. A generalized scheme of fish reaction in an electric field is given in Figure 9.4. Snyder (2003) depicts fish reactions in electric fields indicating the outer boundaries of response zones for a spherical anode at the surface (Fig. 9.5). These zones are more-or-less hemispherical shells around the anode that represent field intensity thresholds for the associated responses. The actual and relative sizes of the zones are specimen dependent and vary with electric output, electrode size and shape, and environmental conditions. The labels in italics in Figure 9.5 represent corresponding phases of epilepsy as suggested by Sharber and Black (1999), except that the phase of contractions (quivering or pseudo-forced swimming) between petit mal and grand mal (narcosis and tetany) is treated as the initial part of grand mal (partial tetany).

210

Figure 9.4.

Fish Behavior near Fishing Gears during Capture Processes

Fish behavior in an electric field. (Adapted from Sternin et al. 1972.)

Figure 9.5. Major intensity dependent electrofishing response zones. (Snyder 2003.)

9.3.2 Factors Affecting Fish Behavior in Electric Fields The simple scheme given in Figure 9.4 is applicable to many fish species, but each case will have its own particularities and will need a separate study before

application. The reason is that many factors determine the reaction of a fish and the different stages described earlier can occur simultaneously for different animals of the same species in the same electric field. It is of basic importance to have the

Electric Senses of Fish and Their Application in Marine Fisheries

211

Figure 9.6. Different variables influencing the behavior of fish in an electric field. (Sternin et al. 1972.)

electric field characteristics correct to meet the envisaged goal. Figure 9.6 shows the main variables determining the response. This figure clearly demonstrates the complexity of the factors affecting fish behavior in an electric field. The main variables are discussed later. Except as specified, the experiments have been carried out in freshwater. In general, the freshwater reactions can also be applied to seawater conditions, but certain electric fields cannot realistically be used in seawater due to the high conductivity. The Electric Field Depending on the type of electric field, reactions of fish are different (Sternin et al. 1972): Behavior of fish in a direct current field. The first defensive reaction of fish as the intensity of the effect increases (with the body situated along the lines of the force) is a shuddering at the instant the current is switched on and slowed swimming.

If the fish body is situated transversely to the field lines, the fish turns toward the anode. The reaction is characterized by excitement and fright. Possibilities of using it alone in electric fishing are limited. When the current increases, the fish show a rapid and strictly orientated swimming motion. The reaction is characterized by an attractive effect that can be used for concentrating fish in fishing gear zones. When the current density increases further, a state of immobility sets in together with loss of equilibrium and insensibility to external stimuli. As a rule, there is no orientation of the fish at this stage. This stage of response can be used in electric fishing. Behavior of fish in a pulsating direct current field. Most of the responses for the noninterrupted DC field also apply to a direct pulse field. Because, however, there are a number of additional parameters characterizing a pulse field (pulse frequency, pulse shape, pulse duration), fish behavior is more varied. The first shuddering effect and excitement

212

Fish Behavior near Fishing Gears during Capture Processes

at low amplitude are more intensive with pulses. The orientated swimming behavior is less clear with pulses and dependent on the pulse characteristics in relation to the fish species and other variables. With increasing power of the electric field, often this stage of behavior does not occur and the state of immobility sets in immediately. Therefore, pulse field is less useful for concentrating fish in a fishing operation. With pulse fields, according to Sternin et al. (1972), it is mainly immobilization of the fish that can be used for fishing purposes. Behavior of fish in an alternating current field. AC has the strongest effect on fish. Usually there is no oriented swimming. The peculiar features of fish behavior under the effect of AC make it possible to use it in electric fishing for repelling, stunning, or disorienting the fish or for blocking migration routes. The characteristics of the electric field have an important bearing on the magnitude and the type of the response observed. Pulsed DC was found to elicit responses at lower field strengths (Blancheteau 1971). This, together with the interrupted character of the electric field, means lower energy consumption. Thus, this is considered to be the best power source for electric fishing in freshwater and the only option for seawater. However, certain behavioral reactions will not be available when using interrupted currents. Vibert (1967) states that a pulse shape with a steep increase and a slow decrease (condenser discharges) are best for fish capture. Quarter-sine waves can be used for fish repulsion. The Electrodes The characteristics of the electrodes have a determining effect on the functioning of the electric field. The main characteristics are the shape, the size and surface area, the material, and the spacing. The electrodes determine the shape and the size of the electric field and the variations in current density. More details are given in Section 4.2. Conductivity Conductivity of water, expressed as micro-Siemens (μS)/cm, is the capacity of water to conduct an electric current. It is a determining factor in the establishment of an electric field and depends on the nature and concentration of ions in water.

Conductivity in natural waters ranges from as low as 5 μS/cm in pure mountain streams to over 50,000 μS/cm in seawater (Snyder, 2003). Conductivity can vary strongly in a body of water depending on the degree of mixing of waters with different conductivity (e.g., in estuaries) and on the water temperature. For natural waters between 10° and 25°C, conductivity varies 2% to 2.3% per °C (Snyder 2003). Reynolds et al. (1988), Reynolds (1996), and Koltz et al. (1998) used the equation c2 = c1/[1.02(t1−t2)] and Sternin et al. (1972, 1976) used c2 = c1/[1 + 0.023(t1 − t2)] to describe the effect of temperature (t) on conductivity (c). When referring to conductivity, it is important to indicate whether it was recorded as ambient (actual temperature) or specific (normalized to 25°C). Cowx et al. (1990) noted some problems associated with high conductive waters, giving the example of a canal with conductivity of 2000 μS/ cm where it was impossible to use conventional electric fishing equipment. PDC was seen as the only alternative. Pulsed currents are the only possibility for electrofishing in water with high conductivity like seawater. Equipment developed for electrofishing in lakes and rivers has its limitations, and the use will become more and more problematic with increasing conductivity. The power consumption will increase sharply with increasing conductivity. The shape of the pulses may also change with conductivity (Jesien and Hocutt 1990). Metabolism and Other Physiological Status According to Vibert (1967), fish with a high metabolic rate (measured by oxygen consumption, intestinal temperature, or respiration rate), such as trout, are more prone to galvanotaxis and less prone to galvanonarcosis compared with fish with a lower metabolic rate such as carp. Fish that are exhausted, sick, or at a stage of sexual maturity do not react very well to electric current. Fish Size Larger fish show a greater potential difference from snout to tail compared with small fish for a given voltage gradient. Thus, larger fish are likely to exhibit an increased response to a given electric field. The gradient of potential difference with length, however, required to elicit a response differs

Electric Senses of Fish and Their Application in Marine Fisheries for the three types of responses. It is minimal for the minimum response and increasing for electrotaxis and electronarcosis. Consequently, the effects of the length of fish are likely to be different dependent on the type of response being induced. This has implications for changes in selectivity during electrofishing; potential gradients sufficient for inducing “minimum responses” may not affect the size selectivity of the gear as much as the higher potential gradients that induce electrotaxis or electronarcosis. Contrary to what some authors axiomatically accepted, B. M. Bary (1956) found that an increase in the length of the fish demands an increase in the potential difference between the nose and the tail to induce any of the three reactions with the three types of electric fields studied. The increase of potential is not linearly proportional to the increase in length. A longer fish requires a higher potential over its length but a lower voltage gradient than a shorter fish. Species: Fish Daniulyte and Petrauskiene (1987) classify swimming behavior of fish in three classes—swimming, crawling, and fleeing. In the case of electrofishing, most interest is in swimming as this is related to anodic electrotaxis. Forms of fish swimming have been discussed in Chapter 1 and are given by Lindsay (1978). Anodic electrotaxis is observed mainly in fish that move by using the body muscles, such as anguilliform, subcarangiform, and carangiform fish (Daniulyte and Petrauskiene 1987). According to Maksimov (1977), no anodic electrotaxis is observed for ostraciiforms, diodontiforms, tetraodontiforms, and ballistiforms. Most flatfish can be classified as ballistiform, and it has indeed been observed that flatfish seem to be less susceptible to electrotaxis than other fish. The best known work on fish behavior in electric fields are those by Bary (1956) and Lamarque (1967). Bary studied the effectiveness of uniform AC and DC electric fields of 50 Hz and PDC fields in producing three reactions—the minimum response, electrotaxis, and electronarcosis—for three marine fish species—mullet (Mugil auratus), seabass (Dicentrarchus labrax), and flounder

213

(Platichtys flesus). The variables taken into account were fish length, pulse duration, frequency of pulsing, voltage gradient, power required, and temperature. Very informative figures were produced, of which a few examples are given in Figure 9.7. For the three species studied, no or little difference was found in the voltage gradient needed to induce the minimum response or electronarcosis. The voltage needed for the minimum response when the head of the fish was pointing to the anode was twice the voltage needed when a fish is pointing to the cathode. The voltage gradients of AC or DC needed to induce electronarcosis appear not to depend on the temperature, but higher gradients are needed for a fatigued fish compared with a rested fish. Requirements for electrotaxis and electronarcosis induced with PDC conform to a similar pattern. With the pulsed electric field, however, extra variables play a role. For a given pulse duration, there is an optimum pulse frequency (Fig. 9.7, right, dashed line). At higher frequencies, the voltage gradient required does not increase, but the power increases proportionally with the frequency. As the pulse duration is lengthened, lower voltage gradients and frequencies are required to elicit a given response. Lamarque (1967) studied a mix of fish species having different forms of swimming. The fish were subjected to a homogeneous DC electric field and the responses were noted in detail. A summary of the behavioral observations is given in Table 9.1. Klima (1972) reported effective combinations of voltage gradient and pulse rate for inducing electrotaxis in a number of coastal pelagic and demersal fish species (Table 9.2). Diner and Le Men (1974) carried out behavior experiments with marine species seabass, mullet, and sardine (Sardina pilchardus) in electric pulse fields. The study demonstrated that electrotaxis and tetanus vary with the frequency and the duration of the pulse, rather independent of the species. Response curves comparable to the ones produced by Bary (1956) are given in Figure 9.8. For electrotaxis, the minimum frequency is 100 Hz and the minimum duration is 3 ms. Swimming reactions differed with the frequency of the pulse. At low frequencies, taxis are less coordinated compared with higher frequencies. The duration of the pulse, if

214

Fish Behavior near Fishing Gears during Capture Processes

30

83 110

26 et 240

8 7 6 5

et 1000

4 3

et DC mr DC+

2 1 4

mr DC– mr AC 8 12 16 20 24 28 Length of fish (cm)

250

22

320 1.3

18 550 14

1500 2000

10

DC 6 AC 2 0

4 8 12 16 20 24 28 Length of fish (cm)

Voltage gradient (V/m)

9

Potential difference along fish (V)

Potential difference along fish (V)

10

1.0

0.7

0.4 10

50 100 Frequency (Hz)

Figure 9.7. Left, Effect of the length of mullet on the potential difference between nose and tail required to induce electrotaxis (et) and the minimum response (mr). et 240, et 1000: electrotaxis at pulse duration of 240 and 1000 μs respectively; et DC: electrotaxis with DC; mr DC+, mr DC−: minimum response with DC with the head to the anode and with the head to the cathode respectively; mr AC: minimum response with AC. Middle, Effect of the length of mullet on the potential difference between nose and tail required to induce electronarcosis when using pulsed DC at pulse durations between 83 and 2000 μs and also using DC and AC. Right, Effect of frequency of pulsing on voltage gradient for various pulse durations when inducing electronarcosis in mullet with pulsed DC. The dashed line is the optimum frequency line. (Reproduced from Bary 1956.)

above 1 ms, seems to have little influence on the strength of taxis, except at low frequencies. Stewart (1977) carried out field experiments at sea with a specially designed trawl with a 50 to 80 V/m stimuli, continuous or in bursts. The roundfish observed during the trial were mainly small gadoids. On experiencing the field, they displayed either tetanus or pulsations and escaped rapidly in apparently random directions. Sandeels were forced out of the substrate. The 40-Hz stimuli induced tetanus that “froze” the sandeels for the 1-s duration of the burst. Stewart (1977) carried out an extensive study in seawater on the response of flatfish (plaice, Pleuronectes platessa, dab, Limanda limanda, and flounder) to electric stimulation. The stimulus applied was either continuous PDC or intermittent PDC with1-s-long bursts of exponential shaped DC pulses with 1-s intervals, each pulse being of 4 ms

duration. Pulse frequencies were 4, 10, 20, 30, and 40, and the field strengths were 50 or 80 V/m. The most common reaction to the intermittent field was a jump off the bottom followed by swimming. Some fish reacted to the field but failed to leave the bottom. At frequencies below the threshold for tetanus, the muscles of these fish reacted to each pulse, giving an effect similar to the normal fin flutter of a flatfish when burying itself (pulsation). At frequencies above the threshold for tetanus, the muscles were totalized and the fish arched their backs (paralysis). With the continuous field, at 10 Hz, a greater percentage of the fish that pulsated failed to leave the bottom compared with the intermittent field. At 20 Hz, especially for the higher voltage, strong paralysis was observed and the fish failed to move after the electric field had passed. The author concluded that for application in a trawl, the intermittent field was most suitable because it

Table 9.1. V/cm

a

Reactions of Fish in Homogeneous Fields of DC. c

Reactionsb

Fish facing anode

Species Ro Sk Ee Ca

First reactionsd

* –

Jerks of head Inhibition of swimming 0.10

Forced swimming

*

Galvanonarcosis

*

Protonos

–

*

*

*

*

*

*

*

*

*

–

–

–

–

*

*

*

*

*

*

*

*

*

*

–

*

–

–

*

*

* *

*

*

*

*

*

*

*

*

*

*

*

*

*

*

*

*

*

*

*

*

*

*

*

*

*

*

*

*

*

*

*

*

*

*

*

*

*

*

*

*

*

*

*

*

*

*

*

Quivering of tail, sagittal plane Pseudo-forced swimming

*

*

*

Tetanus of body, nervous origin

*

*

*

Tetanus of body, muscular origin

* *

*

–

*

Body pigmentationf

*

*

*

*

*

*

Tetanus of gill covers

Opistotonos

*

*

Bending of fins Tetanus of maxillaries

*

Gu Te Go Br Ra Se Bu Su Ti Dr Pl So

1.25 *

*

*

*

*Reaction observed; −, reactions not observed; blank, not studied (Snyder 2003, modified from Table 1 in Lamarque 1967). a Approximate variation of voltage-gradient thresholds. b Main reactions are underlined. c Species. Symbols are Ro, roussette (Scyliorhinidae); Sk, skate (Rajidae); Ee, eel (Anguilla anguilla, Anguillidae); Ca, common carp (Cyprinus carpio, Cyprinidae); Gu, gudgeon (Gobio gobio, Cyprinidae); Te, tench (Tinca tinca, Cyprinidae); Go, golden fish (Cyprinidae); Br, brown trout (Salmo trutta, Salmonidae); Ra, rainbow trout (Oncorhynchus mykiss, Salmonidae); Se, seahorse (Hippocampus sp., Syngnathidae); Bu, bullhead (Cottus gobio, Cottidae); Su, sunfish (Lepomis sp., Centrarchidae); Ti, tilapia M (Tilapia mossambica, Cichlidae); Dr, dragonet (Callionymus sp., Callionymidae); Pl, plaice (Pleuronectes platessa, Pleuronectidae); So, sole (Solea solea, Soleidae). d First reactions of fish facing anode; transient anodic curvature. These reactions occur only at closing current. They are thus more concerned with interrupted current (PDC). By contrast, the “first reactions” of fish facing the cathode take place at the same threshold, regardless of the conditions of potential input. e Forced swimming. This reaction does not occur with flatfish, which would just flatten themselves on the bottom of the tank. In the case of Callionymes and Hippocampus, this swimming is induced by pectoral or dorsal fins. f Body pigmentation, discoloration. These reactions were not thoroughly studied.

215

Table 9.1.

Continued

Fish facing cathode 0.10

Ro Sk Ee Ca d

First reaction

*

Straightening of fins Cathodic galvanotaxis

*

*

*

*

*

*

*

*

*

*

*

*

*

* *

* *

*

*

*

*

*

*

*

*

*

*

*

*

*

*

*

*

*

*

*

*

*

*

*

*

*

Maxillary spasms Opistotonos

*

Se Bu Su Ti Dr Pl So

*

Half turn towards anode Tetanus of body, nervous origin

Gu Te Go Br Ra

*

*

Tetanus of body, muscular origin

*

Discoloration of bodyf

*

*

*

*

* *

*

*

*

*

*

*

*

*

*

*

*

*

*

*

*

1.25

Fish across the field

*

Ro Sk Ee Ca

Temporary anodic curvatured

*

*

*

Gu Te Go Br Ra *

*

*

*

*

*

*

*

*

*

*

*

Se Bu Su Ti Dr Pl So *

*

*

*

*

*

*

*

0.14 Temporary cathodic curvatured Sustained anodic curvature 0.35

* *

*

*

*

*

*

*

Fin straightening on anode side, * Fin bending on cathode side

Table 9.2. Effective Combinations of Voltage Gradient and Pulse Rate for Inducing Electrotaxis in the Species Studied and the Approximate Current (Klima 1972). Species Coastal pelagic fish Scaled sardine (Harengula jaguana)

Volts (V/m)

Frequency (Hz)

Current (A)

Fish Length (mm)

15 30 15 30 30 30 15 30 15 15 15

62.5 86.5 43.3 86.5 86.5 98.3 41.3 86.5 53.8 57.0 42.3 62.5 43.3

118 118 130 130 104 96 118 118 180 173 133 146 148

43.3 43.3

120 128

Chub mackerel 9 (Scomber japonicas) Bumper (Chloroscombrus chrysurus) Rough scad (Trachurus lathami) Thread herring (Opisthonema oglinum) Round scad (Decapterus punctatus)

15

15–55 8–28 35–45 15 25–45 35–45 35 45 15 15 15–25 15 15

Bottom fish Spot (Leiostomus xanthurus) Longspine porgy (Stenotomus caprinus)

15 15

15–35 25–35

Spanish sardine (Sardinella aurita) Round herring (Etrumeus teres) Silver anchovy (Engraulis eurystole) Butterfish (Peprilus triacanthus)

216

Figure 9.8. Response curves for voltage gradient by length of fish, frequency, temperature, and salinity.

217

218

Fish Behavior near Fishing Gears during Capture Processes

had an efficient “tickler” effect. In general, a greater percentage of large fish exhibited strong reactions than small fish. Daniulyte et al. (1987) investigated the effect of electric fields in seawater on Baltic herring, cod and plaice. For herring in a continuous DC field, all normal primary reactions like electrotaxis and electronarcosis were observed. In a PDC field, electrotaxis was most easily obtained at frequencies and voltage gradients of 20 to 40 Hz and 10 to 22 V/m and electronarcosis at 80 to 100 Hz and 18 to 22 V/m. Even at a low voltage gradient of 6 V/m, a clear electrotaxis was present. Pulse duration between 2.5 and 5 ms seemed more effective for electrotaxis than 0.5 to 1 ms. Also for cod, all normal primary reactions were observed. A voltage gradient of 4 V/m was sufficient to induce the primary reactions. A frequency of 10 to 40 Hz at this field strength led to electrotaxis. Maximum electrotaxis was observed at field strength of 6 V/m, pulse duration of 5 ms, and frequency of 20 Hz. At a frequency of 20 to 80 Hz and a voltage gradient of 14 to 20 V/m, electronarcosis started after 5 s. On the other hand, behavioral reactions in an electric field are quite weak for plaice. With frequency between 2.5 and 100 Hz and pulse duration of 0.2 to 2.3 ms, no electrotaxis was observed. A pulsed electric field causes fright reactions and trembles in plaice. Species: Invertebrates Crustaceans show a different behavior compared with fish. Theirs is more like stimulation rather than taxis. Nephrops norvegicus, for example, can be induced to emerge from their burrows when stimulated. Stewart (1972) demonstrated that Nephrops emerge to avoid the electric field by slowly walking forward or by a rapid backward movement by a strong tail flick. Laboratory experiments showed that in 25% of the cases, the animals emerged in less than 6 s. A 40-ms exponential pulse appeared to be the best choice, based on effectiveness and power consumption. Pulse width only played a role if it is less than 40 ms and the optimum pulse frequency was between 1 and 5 Hz. The length of the animal did not influence the rate of emergence. However, as Stewart (1972) pointed out, Nephrops burrows can be deeper and more extensive in their

natural habitat compared with the tank, and the higher water pressure on the seabed can cause the mud to be firmer, hindering the animals while emerging from their burrows. Brown shrimp (Crangon crangon), the most abundant North Sea shrimp species, show strong contractions of the abdomen in an electric pulse field resulting in a quick backward movement. Polet (2003) tested brown shrimps in a nonhomogeneous electric pulse field at different pulse strength and different frequencies. The pulse duration was 0.5 ms. The lowest voltage gradient that invoked a response was 0.4 mV/m, in the case of large shrimps (60 mm) perpendicular to the electrodes. At 80 mV/ cm, all animals were startled. For small shrimps (40 mm), a slightly higher voltage was needed for the first reaction, that is, 60 mV/cm. At 120 mV/cm, all animals responded. For large animals parallel to the electrodes, 180 mV/cm was needed to invoke a 100% reaction; for small animals, 2.4 V/m was needed. Comparable threshold values were obtained by Kessler (1965) for Penaeus duorarum with PDC of 0.05- to 0.5-ms pulse duration. The voltage gradients for shrimps lying perpendicular to the electrodes were between 12 and 75 mV/cm. As soon as a sufficiently strong pulse field was switched on, a startle reaction was observed for all shrimps at the first pulse as observed by Polet (2003) in a tank. The body movement was a contraction of the abdomen, referred to as tail-flip. With the following pulses, the shrimps continued the startle reaction and maintained a swimming phase until they experience pulse fatigue, during which the tail-flips weakened and the animals slowly sank to the bottom. It took at least 15 s for pulse fatigue to show. After a few minutes without electric stimuli, the shrimps recovered and dug themselves into the sand. A detailed analysis of the tail-flip movement showed that the frequency of the contraction coincided with the frequency of the pulse. A low frequency (less than 3 Hz) resulted in a discontinuous tail-flip with a short rest between the contractions. A frequency between 5 and 6 Hz resulted in continuous and complete contractions of the body. Higher frequencies also gave a continuous tail-flip but the contractions were incomplete, resulting in a slower movement. The optimal pulse amplitude in the electric field lies between 80 and

Electric Senses of Fish and Their Application in Marine Fisheries 220 V/m. Higher or lower amplitude has a negative effect on the response of shrimps. The startle response for small shrimps is slightly lower compared with that of large animals and the maximum response is usually obtained within 4 s after the start of the pulses. The length effect was confirmed by Higman (1956) for pink grooved shrimp (Penaeus duorarum). Klima (1968) found that reactions of P. aztecus and P. duorarum were similar. A voltage gradient of at least 30 V/m and a frequency of 4 or 5 Hz were most effective to force the animals out of the substratum, but shrimps escaped more easily from mixtures of sand, silt, and clay compared with sand. A higher water temperature and a low light intensity result in a stronger response (Polet 2003). Higman (1956), however, found that there was no significant difference in reactions of pink grooved shrimp at temperatures between 20° and 30°C. Kessler (1965) found that higher threshold voltage values were needed at extreme temperatures compared with the mid-range and that threshold values decreased as the pulse duration increased from 50 to 500 μs. Burba and Petrauskiene (1987) found a clear electrotaxis toward the anode for brown shrimp with a voltage gradient of 40 V/m. A voltage gradient of 280 V/m caused a clear discoloration of the body. These conclusions have been confirmed by Yu et al. (2007) for four commercial shrimp species in China (e.g., Penaeus japonicus). The Chinese tests also demonstrated that the effect of pulses on shrimps largely depends on the sharpness of the rising and falling edges of the pulse. In an AC field, shrimps exhibit a tetanus reaction and are quickly narcotized. Shrimps seem to be sensitive to exponentially discharged pulses but less sensitive to DC stimuli. Tetanus and electronarcosis can be instigated by a 50-Hz AC stimulus and prolonged exposure can lead to mortality (Hong et al. 1981). Maksimov and Tamasauskas (1987) investigated the behavior of Japanese scallops (Patinopecten jessoensis) in a PDC field. The scallops were found to move toward the cathode. Their highest locomotive activity was recorded when the animals were subjected to a pulse frequency of 10 to 40 Hz and a voltage gradient of 20 to 40 V/m according to a schedule “1 s current, 1 s interval.” Stewart (1977) found that a 50- to 80-V/m stimulus of 30 Hz and

219

above induced razor shells (Ensis spp.) to emerge from the bottom. 9.4 APPLICATION IN MARINE FISHERIES Electrofishing has been widely used in freshwater, mainly for management purposes (Snyder 2003). The fishing gear is commercially available and the methodology has been well described and discussed in the literature. The way the technology is used in freshwater cannot be extrapolated to applications in seawater, because of high conductivity of the seawater and the dimension and depth of the sea. As a result, application of electricity in marine fisheries has been less successful. The literature is full of examples of promising field experiments, very close to commercial introduction, but very rarely was the next step to introduction in the commercial fishery successfully taken. This chapter gives a short overview of ways electricity could be used in marine fisheries and some examples of applications. All electric fishing is based on the three main effects of electricity on fish—that is, fright, attraction or repulsion (stimulation), and narcosis. A few examples of possible fishing methods in seawater are: Use of Repelling Electrode Systems • Directing fish migrating close to the shore into traps and gillnets • Creating commercial concentrations in front of trawls • Stimulating fish and invertebrates from the seafloor in demersal fisheries • Influencing the functioning of selective devices based on behavioral differences between fish species and/or size classes Use of Attracting Electrode Systems • Creation of commercial concentrations for subsequent catching by conventional means • Attracting fish from the seafloor in demersal fisheries • Attracting fish to passive fishing gear Other means include stunning and collecting fish with a fish pump. Most of these methods, however,

220

Fish Behavior near Fishing Gears during Capture Processes

are less desirable fishing practices because of possible damage to and mortality of the target and nontarget species and the possible detrimental effect on the marine ecosystem. Taking these into account, electric fields can undoubtedly be used in a positive way in fishing gears, but the complexity of the method, the vulnerability of the equipment, and the side effects can hinder their commercial application. A wide range of variables affect the functioning of electric fishing systems. Each situation requires a detailed study of the behavior of the different fish species in relation to environmental conditions and electric field characteristics. 9.4.1 Historic Overview and Description of Selected Cases While the idea of using electricity in marine fishing is quite old, the major research and development were carried out between 1950 and 1980. van Marlen (1997) gave a comprehensive review on the subject. Most of the work was carried out in countries around the North Sea and in the United States, China, and the former USSR with the development of fishing gear that came close to commercial application (van Marlen 1997). In the United States (Seidel and Watson 1978), India (Sreedharan et al. 1977), and China (Yu et al. 2007), commercial fishing gear was developed, but only in China did industry take up the ideas. Despite the huge effort in basic and applied research, the outcome in terms of commercial applications has been very limited. The only systems that are close to introduction in the commercial fishery are the Dutch electrobeam trawl targeting flatfish (Solea solea, Pleuronectes platessa) (van Marlen et al. 2006) and the Belgian electrobeam trawl targeting brown shrimp (Polet 2003). A very successful commercial fishery with electrobeam trawls was developed in China, but the increase in fishing effort due to the efficiency of the method, bad practice in the choice of pulse characteristics, and mismanagement led to a ban of the system in 2001 (Yu et al. 2007). Trawling for Shrimps Research on the application of electric fields in shrimp trawling started in the early 1950s in the Netherlands (De Groot and Boonstra 1970). Van

Marlen (1997) gave an overview of the Dutch experiments between 1966 and 1979, the “shrimp period.” Until 1969, some behavioral experiments were carried out in aquaria at RIVO (Netherlands Institute for Fisheries Research) and oyster basins on both shrimps and sole subjected to electric fields. The potential for using this kind of stimulus for catching the animals was clearly identified, leading to an increase in effort in the development of a commercially applicable system. From these basin trials the following specifications arose: optimum pulse width for shrimps of 0.2 ms, and for sole, 0.7 ms. During subsequent sea trials, the pulse generator was mounted directly on the beam of the trawl and used an internal power unit, but apparently its specifications in amplitude (10 V or less) were too low. Field tests were carried out in 1972 with the first pulse unit on 3-m electrobeam trawls. Results were discouraging and led to new specifications defined for a follow-up pulse unit. Reports describe an improvement in catches of shrimps ranging from 8% to 35% but with malfunctioning of the equipment in 20% of the time. Again, a new pulse generator was built with higher amplitude (65 V) and tested in 1975, but it still had many malfunctions. Research on electrified shrimp trawling stopped in 1976 after a series of experiments indicating a catch ratio of 1 : 1. During the same period, electro-trawls for brown shrimps were tested in Belgium (Vanden Broucke 1972), the United Kingdom (Baker 1973), and Germany (Horn 1976) with the most of the work pointing to good prospects for this type of fishery. The main objective of the work was to reduce fuel consumption and to increase the commercial catches with no or very little attention to bycatches. Some experiments, however, already pointed at possibilities for selective fishing with electricity (Stewart 1975). Research on electric fishing continued into the 1980s but then stopped almost simultaneously in all North Sea countries by national bans on electric fishing driven by the fear of overfishing. In the 1990s, the experimental work with electrobeam trawls was taken up again in Belgium. The Belgian electro-trawl for brown shrimp has been tested on board of commercial vessels in the late 1990s. The first results were positive, although practical problems and seasonal variations in catch efficiency

Electric Senses of Fish and Their Application in Marine Fisheries caused some concern. It was concluded that new, finely tuned equipment was desirable to achieve constant satisfactory results (Polet 2003). After a pause of 5 years, the research restarted in 2007 (Verschueren, 2008) with the construction of a new optimized impulse generator. Sea trials are still ongoing, but preliminary results already show significant improvement. Testing by direct comparison with a standard catch shrimp trawl revealed that at least as much shrimp can be caught. Furthermore, the catch efficiency seems less influenced by the time of day. This is in contrast with the traditional shrimp trawl where catch efficiency increases with lower light intensity and higher turbidity of the seawater. At the same time, it is important to state that the experimental trawl reduces (Fig. 9.9) bottom contact by 80% in comparison to the standard

Figure 9.9.

221

brown shrimp trawl. This is after all not surprising since the use of a bobbin rope, to startle the shrimp, is no longer needed. Considering bycatch, the sea trials show that substantial reductions can be accomplished. However, at this moment it is too early to quantify these reductions, because the trials are ongoing. It is distinct, though, that electrofishing can play an important role in dealing with the discard problem of the brown shrimp fishery. In other parts of the world, interest in the fishing method, especially with an application for shrimps, was maintained after the ban in western Europe. In 1987 experiments with brown shrimps in electric fields were reported in Lithuania (Burba and Petrauskiene 1987). In the United States, a selective electrified shrimp trawl was developed (Holt 1992), although commercial application was not reported.

The Belgian electrobeam trawl. (Polet 2003.)

222

Fish Behavior near Fishing Gears during Capture Processes

In India experiments were carried out with electric fishing (van Marlen 1997). Most of these electric fishing experiments were carried out with otter trawls, using the footrope, headline, and/or warps as electrodes (Namboodiri et al. 1977). Other designs filled up the net mouth with electrodes parallel (Pease 1967; Seidel 1969) or perpendicular (Ellis 1972) to the footrope. Seidel and Watson (1978) suggested a selective electrified otter trawl (Fig. 9.10), without bottom contact and a closed net

mouth to eliminate fish bycatch. The belly of the net consisted of large meshes, allowing shrimps to enter after stimulation by electric pulses. No commercial application was reported. The most widespread and commercially successful application of electric pulse trawling for shrimps was developed in China. In the early 1990s, the electrobeam trawl (Fig. 9.11) was introduced in the East China Sea off the Zhejiang province targeting a wide mix of shrimp species, including

Figure 9.10. Selective shrimp electro otter trawl without bottom contact.

Figure 9.11. The Chinese electrobeam trawl. (Yu et al. 2007.)

Electric Senses of Fish and Their Application in Marine Fisheries

223

Figure 9.11. Continued

Parapenaeopsis hardwickii, Solenocera crassicornis, and Trachypenaeus curvirostris (Yu et al. 2007). The catch rates were significantly increased using the electric stimuli in the traditional beam trawl, especially for burrowing shrimps. At its peak, more than 3000 fishing vessels applied electric stimuli in that province alone. Typical shrimp electrobeam trawl vessels in the eastern Chinese Sea have beam lengths of 24 to

36 m and a length of net between 15 to 19 m. The mesh size typically starts from 45 mm at the front of the net and gradually reduced to 25 mm at the codend. Earlier beam trawls have two codends, but the number of codends have recently been gradually increased to four, six, eight, and ten, with the majority of the trawl now having six to eight codends. A trawl with more codends seemed to have higher catch rates, better catch quality, and a

224

Fish Behavior near Fishing Gears during Capture Processes

more stable gear. An electrobeam shrimp trawl is rigged with a pulse generator and associated components installed on the trawl. Parallel electrodes are installed in front of the groundgear to form an electric field. Stimulated shrimps jump out of substrate or off the seabed and fall into the trawl mouth. A shrimp electric pulse stimulus apparatus (SEPSA) typically consists of a power source, a pulse generator, electrodes, and electric cables (Fig. 9.11). It can be divided into cabled and cableless system. At the beginning of the development in 1980s, SEPSAs were cabled. A DC transformer raised voltage of the power that was transmitted through an underwater cable to the pulse generator attached to the beam of the trawl. Output from the pulse generator was passed to the electrodes installed in front of the groundgear. The pulse generator was encased in a 20-cm-diameter, 100-cm-long steel cylinder. Because the power was from a power generator on board the vessel, electric fields were stable and were not affected by the length of the tow. Compared with transmitting actual pulsed stimuli through a large cable as practiced in 1970s, transmitting higher-voltage DC power through a small cable reduced power consumption and costs. Cableless systems used a battery pack installed in a longer steel tube (250 cm) together with the pulse generator. The use of a cable was avoided and the systems are much easier to operate during deployment and retrieval of the gear. However, the battery pack needed to be changed and recharged every tow to ensure sufficient power to generate pulses of sufficient strength. Usually, buried shrimps jumped out of the substrate and to the height above the groundgear within two or three pulses. The use of the Chinese pulse generator increased shrimp catch rates by 40%, and large shrimps, by more than 100%. However, with illegally increased pulse parameters, more small shrimps of 50 mm or less were caught. As the use of the electro-trawl became popular, some shrimp species became less abundant. In response, some manufacturers and vessels increased power output and reduced the electrode distance to increase the intensity of the electric field. Increased fishing effort put greater pressure on the resource. At the peak of its use, more than 35% of beam trawlers were using electro-trawls. In Zhejiang province alone, there were

more than 3000 vessels fishing with electricity. As a result, shrimp biomass reduced drastically. For example, biomass in 1995 was 33.5% less than in 1994, and in 1996, biomass was further reduced by 11.1%. The trend of shrimp biomass reduction continued through 1998. In addition, shrimp biomass in offshore area was also seen to fluctuate and reduce. The use of illegal equipment caused damages to the resource, but fishery management authorities were unable to control. As a result, the electro-trawl was banned by the Zhejiang Provincial Fishery Bureau from its coastal waters as of January 1, 2001. The ban included manufacture, sale, repair, and use. Trawling for Fish Kreutzer (1963) states that the main reason for using electric fields in otter trawls is the low efficiency of that type of trawl in catching a relative amount (10% to 60%) of fish present in the trawl path. Electricity would increase the efficiency. Because the fish stay in a rather small zone in front of the net, he suggested that, even with a small electric field covering only about the height and the width of the net opening, it would be possible to increase the catching efficiency considerably. With anodes attached to the net, the fish would first be attracted into the net and then stunned to preclude further escape. Even if they recover and try to swim out, through the net mouth, they would meet the “electric wall,” which acts as a valve letting the fish pass only in one direction—into the net. The method was tested and underwater observations confirmed the theory. The peak current needed for the electric field was 15,000 to 40,000 Å. The pulse duration was chosen between 1.2 and 1.5 ms. Based on 82 comparative tows, it was found that the catches increased by 100% to 500%. The increase was more pronounced for roundfish than for flatfish and the catches with use of an electro-trawl contained more large fish. Despite these apparently good results, no commercial introduction has been reported. Maksimov et al. (1987) reported several experiments with otter electro-trawls in the Barents Sea, the North Sea, the Southeast Atlantic, and the Gulf of Mexico. No details were given on the design of the trawls but an effect on length selectivity was stressed. The electro-trawls had an improved catch-

Electric Senses of Fish and Their Application in Marine Fisheries ing efficiency for large fish (cod, whiting, and several pelagics) and a decreased catching efficiency for flatfish like plaice. Most of the research on electro-beam trawls targeting flatfish was carried in the North Sea countries (1970s and 1980s). In the Netherlands (van Marlen 1997), several types of pulse generators were built by staff of RIVO based on capacitor discharge technology. Variables under investigation were frequency, voltage on electrodes, and length of electrodes. It appeared that the catches on the electrified trawl increased with rising frequency and voltage. No difference was found as a result of changes in electrode length or between various gear designs. Differences in catchability were found between day and night with higher efficiency at night. Experiments on electric fish barriers resulted in an idea to decrease the pulse width. The amplitude of the pulse was increased to 1000 V in a new design. This design also featured higher voltage of the power supply generator, safety circuits for current and voltage, sequential discharge of pairs of electrodes by an electronic ring counter, and discharge through thyristors. Trials indicated better catches at night (+50%) for the electrified net but lower during the day (−10%). The electrified gear appears to have better size selectivity, especially for small soles, which did not appear in the catches. The electrodes showed a large degree of corrosion. A new design followed with a maximum capacitor voltage of 2000 V. The new gear design appeared to catch 45% more sole during the day and 65% during the night. The optimum voltage with this net was around 700 V, at 20-Hz frequency. However, size selectivity was not improved with this design. More recently, a private company developed a prototype pulse generator with two capacitor containers that were mounted onto the shoes of the beam trawl. The design was based on requirements concerning robustness and quick interchange of components. Tests resulted in higher sole but lower plaice catches than the conventional beam trawl. The project stopped with a ban on electric fishing by the Dutch Ministry of Agriculture and Fisheries out of fear of further increasing fishing effort in the beam trawling fleet, which was under severe international criticism. In the 1990s, the experimental work with electrobeam trawls was initiated again in

225

Figure 9.12. The Dutch electrobeam trawl. (Photo: Jochen Depestele) For color detail, please see color plate section.

the Netherlands. The Dutch electro-trawl (Fig. 9.12) has been tried extensively onboard a commercial trawler in 2005 and 2006 (van Marlen et al. 2006). The results are positive, although practical problems and seasonal variations in catch efficiency remained. Work in Belgium indicated that a frequency between 5 and 10 pulses/s and a voltage between 60 and 100 V should be used for flatfish. To reduce damage to the fish, a short pulse length of 1 ms was applied. The results indicated that the heavy tickler chains rigged in traditional beam trawls could be replaced by lighter ones using the electric stimulation, without loss of catch. The pulse length seemed to play an important role in stimulating sole. During trials in Germany (van Marlen 1997), a series of the electric characteristics of the pulse generator was used to find the optimal configuration. A voltage of 110 V and a current of 1.31 A at every pair of electrodes with a pulse length of 0.51 ms at a frequency of 25 Hz proved to be most effective. An increase of catch (sole) of 114% was obtained. At the same time, bycatch of benthic organisms and sand was reduced by almost 50%. In the United Kingdom, similar experiments were carried out (van Marlen 1997). A set of laboratory experiments indicated that a pulsed electric field at 4 Hz, a pulse of 1 ms, and estimated field strength of 150 V/m proved

226

Fish Behavior near Fishing Gears during Capture Processes

adequate to elicit minimum response from sole, causing them to lift up off the seabed. These electric field characteristics were achieved in experiments carried out from 1979 on. Later experiments on beam trawls solved a number of technical problems and tested various electrode configurations and materials. By late 1980s, the system was developed to the point where long-term trials were contemplated. The main benefit was fuel saving for a similar catch per effort due to the reduced towing speeds and reduced load required for electric fishing. However, the reduction in the relative cost of fuel during the early 1980s rendered these savings less attractive than when the research project was initiated, resulting in discontinuation of the work around 1983. Pump Fishing for Pelagic Fish In 1954, a new method of fishing for Caspian kilka (Clupeonella spp.) was established using pumps in conjunction with underwater light. In the early 1960s, a growing fleet was engaged in this fishery with a total yearly catch of over 100,000 tons. The attraction of this fish species to light is so strong that the light stimulus alone was sufficient to bring the fish into the suction area of the pump. The method has been tried for other species (Nikonorov 1963) such as saury (Cololobis saira). Although dense concentrations of fish were observed in the zone of illumination, the fish easily avoided the suction intake aperture of the nozzle. To enhance the method, the pump was combined with an electric field. The cathodes consisted of two steel pipes suspended on the booms near the bow and the stern of the vessel. The suction nozzle, insulated on the outside, served as the anode. The experimental fishing proved very successful with a maximum catch of 1.5 tons of saury in 9 min. Dethloff (1963) describes the method and problems associated with electro fishing with pumps. Spherical electrotaxis with diameters of 6 to 20 m dependent on species and length of fish can be obtained with a PDC field with powers considerably below 50 kW. The electrotaxis range must be larger than the suction range of the pump. The electrotaxis range should enclose as many fish as can be sucked in by the pump after concentration around the intake. However, this is limited by the time period

during which the fish can be kept in the stage of electrotaxis before they sink down or rise up fatigued or stunned by muscular stress. In principle, this depends on the electric power output, the capacity of the pump, and the density of the school. Another condition is that the electronarcosis range must be smaller than the suction range. If the narcosis range is larger than the suction range, some fish are not able to enter the suction range but are narcotized and lost. Further, the range of low mobility must be prevented. To fine-tune these parameters, the “pulse compensation method” can be used (Dethloff 1963). Kreutzer (1963) describes a similar method for the menhaden fishery on the east coast of the United States. In this fishery, the fish are concentrated by means of a purse seine, which is then emptied with a pump. To facilitate the collection of the fish, an electric field is used. When the fish are concentrated in the bunt of the seine, the electricity is switched on. The fish are electrically attracted to the nozzle so that the pump can continuously operate at full capacity. In this way, the bunt does not have to be lifted. After a dense concentration of fish has formed around the nozzle, there is a lower current flow due to the lower conductivity of the fish compared with the water and less fish are attracted from a distance. When most of the fish in the concentration are pumped off, the electric field increases in strength again and more fish are attracted so that a balance is established between fish attracted and those pumped away. For this application of electrotaxis, a rather low current of 2000 to 3000 Å is sufficient. This type of electrified pumping was installed during 1958 and 1959 on all menhaden purse seiners operated by the Smith Company off the U.S. east coast. It led to a reduction in crew members from 22 to 10. Diner and Le Men (1971) suggested that pump fishing for sardines would be feasible and gave following pulse characteristics: for the electrotaxis of sardines, a frequency of 200 to 50 Hz and even down to 30 Hz, a pulse duration of 1 ms. and a voltage gradient of 10 to 12 V/m. An increase to 30 V/m would lead to tetanus. Despite the positive trials with pump fishing reported by the respective authors, no longer-term commercial application has been found in the literature.

Electric Senses of Fish and Their Application in Marine Fisheries An Automated Unmanned Fishing System Driven by the oil crisis in the 1970s, Seidel and Vanselous (1976) designed a fishing system that would automatically catch fish without the use of a vessel (and thus fuel). It has not been reported in the literature that this system ever worked, but originality of the ideas make it worth mentioning. The authors state that in Gulf of Mexico, there is a huge fish resource that is basically not used. These species do not form large schools and are often found in shallow water on hard rocky bottom, which hinders an efficient fishery. To harvest these fish, the authors describe an automated fishing system that offers a method to effectively harvest the coastal pelagic fish resource. The only manpower or vessel requirement would be to service and maintain fishing platforms and to offload the processed products. The proposed system (Fig. 9.13) would be based on unused oil rig platforms. The oil rig would be surrounded with fish aggregating devices that would concentrate fish based on their natural behavior. A

227

series of lights that would be switched on and off in a sequence would lead the fish to the oil rig and bring the fish into the electrotaxis range of an electrofishing device. The electric field would periodically be switched on to concentrate the fish into the suction range of a fish pump that would bring the fish to the surface. Further study of fish behavior in electric fields may identify optimum voltage gradients that would induce the desired effect on target species while repelling unwanted species. A scanner-type device, scanning the area immediately surrounding the pump nozzle, could provide biomass measurements to be used by a computer for operating decisions. 9.4.2 The Electrodes According to Snyder (2003), spherical electrodes are considered electrically superior to other shapes. Electric fields around spheres are uniform and lack the hot spots produced near the corners and edges of other shapes. According to Sharber et al. (1994),

Figure 9.13. An automated unmanned fishing system. (Reproduced from Seidel and Vanselous 1976.)

228

Fish Behavior near Fishing Gears during Capture Processes

charge is not distributed uniformly over long thin electrodes but rather concentrated at their distal ends. Novotny (1990) suggested that circular and ringlike shapes produce electric fields similar to those of spheres but are somewhat more extended in the horizontal direction (Kolz 1993). The problem of small electrodes was explained by Vibert (1967) in relation to injuries. Experiments with fish indicate that below a threshold value of field strength, they tend to avoid the electric field. At a higher value, they move toward zones of increasing intensity. At a critical value of field intensity, they become paralyzed. It follows that the field around the anode should be as wide as possible between the threshold and the critical value and that the supercritical zone should be eliminated as much as possible. The supercritical zone represents a set of conditions where an excessive current density occurs. A larger electrode surface will eliminate this zone and make the fishing zone larger. Within the supercritical zone, it is possible to produce lethal effects to fish (especially in water with low conductivity). If a fish encounters the supercritical zone close to a small electrode, it is subjected to a very rapidly increasing field, much more intense at its head than at its tail. Because the conductivity of the fish may be higher than that of the water, the current density in the body will be greatly increased. The use of larger electrodes of, for example, wire netting may avoid this problem. In seawater, though, the maximum size of the electrodes is dictated by the capacity of the generator (Snyder 2003). Here, the size of the electrodes may have to be reduced to prevent generator overload. Any form of electric fishing in seawater requires the passage of intense current from the generator to the water that contains the fish, to develop useful electric field strength throughout a significant volume. The materials used for electrodes are highly conductive metals. The passage of intense currents from metallic electrodes into a concentrated conductive solution like seawater can cause physical and chemical effects such as corrosion and polarization. These effects can be prevented by using electrodes of inert noble metals like platinum. If the electrodes are subject to wear and tear, such as in trawling, this is a very expensive option. To prevent wear and tear, electrodes can be imbedded

in synthetic carriers with openings allowing the current to flow. This is the principle used in the recently developed Dutch electro-trawl (van Stralen, personal communication), where the synthetic electrodes are fitted with six bronze electrodes. For the commercial shrimp electro-trawling in China, copper electrodes were used (Yu et al. 2007). Stewart (1973) studied different materials for electrodes and concluded that barrier layers on titanium and aluminum have great effects on the transmission of current pulses. This makes them unsuitable for use as electrodes in experimental conditions when unnecessary distortions must be excluded. Severe corrosion took place on anodes when galvanized steel and copper were tested. It was found, however, that the salt around the anodes of these metals tended not to adhere and had no effect on current flow. If used as an electrode on a trawl, abrasion during towing would remove these deposited salts more rapidly. Stainless steel, galvanized steel, copper, or brass could be used as electrode materials without undue distortion of current flow. Stainless steel and brass are the most attractive because of their greater resistance to corrosion by seawater. Stainless steel appears to be most suitable because it is cheaper, mechanically stronger, and readily available in useful forms such as trawl warps. If aquarium experiments are to be conducted, the most suitable material is mild steel, containing no elements that are poisonous. AC or current reversal can limit corrosion (Kreutzer 1963). Stewart (1973), however, found that although this can prevent the buildup of salts during operation, the removal of metallic ions continued when the electrodes were not energized. For some applications like taxis, current reversal is not possible. He also concluded that the value of current reversal for limiting corrosion would be determined by relative costs of the electrode material and the more complex pulse generator. 9.5 9.5.1

CONSERVATION ISSUES Injuries

Types of Injuries While electrofishing has become a common capture technique in fisheries research and may gain interest

Electric Senses of Fish and Their Application in Marine Fisheries for commercial fishing applications, the potential impact of this technique on the fish is not completely understood. During the first decade when electrofishing was spread over the world, possible negative effects of electrofishing were often trivialized or just neglected. The review on alternative stimulation in fisheries by van Marlen (1997) mentioned a large amount of research (mainly the 1960s through the 1980s) on possible applications of electricity to increase the fishing efficiency of sea fishery trawls but almost no research on effects on target or nontarget species. Some authors have stressed the harmlessness of electrofishing (e.g., Halsband 1967; Sternin et al. 1976) and mainly focused on the applications. Many damaging effects of electrofishing are internal, like spinal injuries or internal bleeding, and they do not kill the fish instantly (Snyder, 2003). As a consequence, when a captured fish was released and was apparently alive, it was considered unharmed. More recent literature, however, has indicated that electrofishing can be harmful in many ways. An excellent and extensive review was made by Snyder (2003). A summary of the possible effects of electrofishing is given below, but it should be noted that most of the studies have been carried out in freshwater. Mortality. Sudden death of fish by electrocution in electrofishing operations has rarely been reported in the literature. Northrop (1967) mentioned that temporary cardiac arrest might occur in narcotized fish. Several other studies (Taylor et al. 1957; Schreck et al. 1976) have shown that the electric field can have an influence on the cardiac activity but that usually the normal heart activity is restored after some time. Schreck et al. (1976), however, described a case where cardiac and respiration activity was disturbed after an electricity treatment of about 1 min with slow recovery of normal heart activity, until finally the fish died, probably due to lack of oxygen. Because respiration is strongly reduced or stopped with narcotized or tetanized fish, this is often the main cause of death. Fish that do recover after respiratory failure show increased respiratory rates (Kynard and Lonsdale 1975). Oxygen debt can take several hours to recover (Kolz and Reynolds 1990). For some fish species, the mucus formed on the gills can cause the problem (Lamarque 1990).

229

Bleeding. Electrofishing can cause internal wounds and bleeding. The main cause is the breaking of parts of the skeleton by electrofishing (see Spinal Injuries later), rupturing arteries or veins. Bleeding of the gills has also been reported (Hauck 1949) but the cause of this would be different to discern (Snyder 2003). Spinal injuries. Electrofishing can cause powerful convulsions of the musculature of the fish. These convulsions can be so intense as to lead to injuries to the skeleton, resulting in secondary injuries. Snyder (2003) summarizes injuries as compressed, broken, or misaligned vertebrae; separated or damaged ribs; damaged swim bladders; ruptured dorsal and hemal arteries; torn muscles or ligaments; and other internal hemorrhages (Figs. 9.14 and 9.15). Jerks or seizures caused by electroshocks occur simultaneously on both sides of the body, subjecting the vertebral column to opposing forces that can

Figure 9.14. Dorsal and lateral view radiographs of a rainbow trout revealing spinal misalignment and fractured vertebrae caused by electrofishing. (From Snyder 2003 with permission.)

230

Fish Behavior near Fishing Gears during Capture Processes

Figure 9.15. Necropsy filets of rainbow trout revealing hemorrhages and associated tissue and vertebrae damaged caused by electrofishing. (From Snyder 2003 with permission.) For color detail, please see color plate section.

break, crush, or dislocate the vertebrae (Lamarque 1990; Sharber et al. 1994; Sharber and Black 1999). Some authors have observed hemorrhages without corresponding damage to the skeleton (Holmes et al. 1990; Fredenberg 1992). Snyder (2003) reports that electrofishing-induced spinal injuries can occur anywhere along the spinal column, including immediately behind the head, but most have been observed near or posterior to the middle of the spine. Predominant location varies with species. Different types of electric field can cause different injuries with regard to severity or type of injury (Snyder 2003). Behavior. After being subjected to an electric field, fish will often need a recovery period to return to their normal behavior. Reduced feeding has been observed in several studies (Callahan 1996; Mesa and Schreck 1989). Callahan also found small bluegill to be more susceptible to predation after being shocked. Swimming performance can also be reduced (Mitton and McDonald 1994). The recovery period can last minutes to several hours. Stress. The stress caused by electrofishing is reflected in the blood chemistry of fish. Schreck

et al. (1976) found an increase in the blood plasma corticoid (adrenal hormones), lactic acid (byproduct of anaerobic muscular activity) and thrombocytes (white blood cells instrumental in blood clotting— indication of possible tissue trauma, bleeding) in yearling rainbow trout. Other physiological indicators of stress were also reported, including catecholamines, metabolic acid, carbon dioxide, and pH (Mitton and McDonald 1994). Many studies have been carried out to evaluate stress caused by electrofishing. In general, it can be concluded that electrofishing causes a similar stress as extreme muscular activity. Stresses can be cumulative as electrofishing stress can add to environmental stress like pollution (Wydoski 1980). Snyder (2003) concluded that increased mortality can occur directly as a result of stress and fatigue or indirectly through greater susceptibility to predators, disease, and parasites. In some cases, delayed stress-related mortality may be more significant than immediate electrofishing mortality. Injury-related stresses may persist and affect the fish’s physiology, behavior, growth, and reproduction for a long time. Heart rhythm. Kazlauskiene (1987) found that under a rising voltage gradient, the heart rhythm of Baltic cod, herring, and flounder became more intense and arrhythmia increased. The durations of changes and recovery were dependent on the current intensity. Under low DC (0.1–0.5 V/cm), the heart rhythm changes were recorded over 1 h. Under high voltages (up to 1.0 V/cm), no recovery was recorded. Factors Affecting Electrofishing Injury and Mortality Snyder (2003) provided an extensive and detailed overview of electric field parameters and the effect on fish. The author discussed different parameters of the electric field that affect injuries and mortality but stressed that comparison of results are difficult and often susceptible because of differing and often inadequately described biological, field, or experimental conditions, including electric parameters. Type of current. In general, AC is considered the most harmful to fish, DC the least harmful, and PDC as in-between, although exceptions have been reported. With sufficient field intensity and duration of exposure, any current can be lethal. Short-term mortalities for PDC are often as low as for DC, but

Electric Senses of Fish and Their Application in Marine Fisheries this is not the case for spinal and related injuries when using moderate- to high-frequency PDCs. Field intensity. In general, lethality of electrofishing fields increases with field intensity (i.e., voltage gradient, current density or power density). It is, however, the voltage differential across the fish that determines the effect, which depends on the orientation of the fish relative to the field lines. Unlike mortality, the relation between electrofishinginduced injuries and field intensity, beyond some threshold level, remains unclear. Spinal and related injuries can occur in high- as well as in low-field intensity. Duration of exposure. Increased mortality with increased exposure has been well documented, especially at field intensities sufficient to induce tetanus. But beyond a necessary minimum threshold, duration of exposure in AC and DC fields does not appear to have an important effect on spinal and related injuries, which is not necessarily the case for PDC. Waveform, pulse shape. The effect of waveform or pulse shape on AC or PDC electrofishing mortality or injury has been poorly studied and remains inconclusive. Pulse frequency. Pulse frequency appears to be a primary factor affecting PDC-caused spinal injuries and may be a significant but probably secondary factor in electrofishing mortalities. Most spinal and related injuries are caused by sudden changes in electric potential, such as the moments when currents are switched on and off. Consequently, it is not surprising that investigators have generally reported, with some exceptions, increasing incidences of such injuries with increasing PDC frequency. If field intensity and exposure time are maintained above the threshold for lethal effects, there is some evidence that mortality can also be greater when using higher-frequency PDCs. Pulse duration, duty cycle, voltage spikes, and fish size. The effects of these parameters on mortality or injury remain inconclusive. 9.6 CONCLUDING REMARKS Fishermen are generally very practical and rely on their experience for the correct design and rigging of their fishing gears and their knowledge of fish behavior to catch the fish. Hence, the idea of

231

catching fish by flipping an electric switch may be quite upsetting. Putting two electrodes in the water and watch fish swimming toward you may also not suit with the hunting spirit of the fisherman. In “A Shocking Way to Fish” (Idyll 1958), the interest in using electricity to fish was referred to as “man-from-Mars” talk. Scientists, however, were very enthusiastic and a lot of opportunities were seen. In streams and lakes, the typical behavior of fish in electric fields has been extensively exploited. The method made it possible to collect fish where it was previously impossible, such as on a rough bottom, in dense underwater vegetation, or among overhanging brushes. The reason for its success and why the method did work in freshwater probably were the small scale of the fishery. Furthermore, freshwater electrofishing has mainly been used for scientific purposes and there was no economic incentive involved. By the time scientists started thinking of applications in seawater, electric fishing in freshwater was referred to as “old hat.” Until the 1990s, electrofishing was generally considered to be harmless for the fish. During the past decade, several researchers have proved that this is not the case. Electricity can cause internal damage to the fish and concern was expressed especially for Salmonidae and endangered species. The main difference with electric fishing in seawater is the dimension of the water basin and the size of the fishing gear. Extrapolating the experience and the method of electric fishing in lakes and rivers to the sea without taking the differences into account has often led to disappointing results. Using the electrotaxis behavior of fish, which is very efficient in freshwater, in marine fishing gear, is usually not very effective due to the scale of the operation. The electrotaxis range is quite small and consequently this response of the fish can only be used in specific circumstances (i.e., where fish occur in dense aggregations or where fish are first aggregated by other means than electricity). Applying electronarcosis in the sea can usually be classified as “bad fishing practice,” mainly because of the lack of control over the narcotized fish. These fish can sink out of the range of the collecting device or can be taken by the current. The fish can be wounded or die or can be an easy prey for predators.

232

Fish Behavior near Fishing Gears during Capture Processes

The “fright” response is probably the most realistic to be used in fishing gears. The electric field can be used as an alternative to mechanical stimulation to scare the animals from their hiding position on the sea floor, into the range of the fishing gear. The environmental impact of fishing gears such as bottom trawls could thus be reduced significantly. The difference in behavior between fish species or between crustaceans and fish can be used to increase the species selectivity of a fishing gear. Animals can be guided in or out of the range of the fishing gear. It is also possible to guide different species to different compartments of a fishing gear and allow certain species to escape, thus reducing discards. The difference in response of small and large fish in electric fields can also be used to increase the length selectivity of a fishing gear. Any creative mind can think of a wide variety of possible applications of electric fields in marine fishing, but taking the step to commercial application is not obvious. Many researchers have come up with very promising results during the past six decades, but commercial fishing with electricity has been rare. The vulnerability of the machinery in a harsh environment like the sea has often discouraged the fishermen who had to bring the method into practice. The complexity of the combination of the different variables affecting the efficiency of electricity and seasonal changes in the behavior of fish does not add to the user friendliness of the method. On top of this, concern regarding the possible harmfulness of electricity on target and nontarget species and the fear for bad practice have led to legal bans of the method. Electric fields undoubtedly have potential, but it is a technique to be developed thoughtfully and to be used with care. REFERENCES Baker C. 1973. Experimentally electric fishing voyage on DAFS research vessel Clupea. White Fish Authority FR 066. Bary BM. 1956. The effect of electric fields on marine fishes. Scot. Home Dept. Mar. Res. No. 1. Blancheteau M. 1971. Choix du stimulus approprié à la pêche à l’électricité en mer. Rev. Trav. Inst. Pêches Marit. 35(1): 13–20. Burba A and Petrauskiene L. 1987. Structure of the Baltic shrimp reactions in DC and pulsed current

fields. Acta Hydrobiol. Litunica. 6: 79–92 (translated from Russian). Callahan S. 1996. Effects of electrofishing on sunfish. IL Nat. Hist. Surv. Rep. 338: 4–5. Cowx IG, Wheatley GA, Hickley P and Starkie AS. 1990. Evaluation of electric fishing equipment for stock assessment in large rivers and canals in the United Kingdom. In: Cowx IG (ed). Developments in Electric Fishing. pp 10–18. Oxford, England: Fishing News Books. Daniulyte G and Petrauskiene L. 1987. Anode electrotaxis of fishes of various ecological groups with respect to their locomotion mode and its mechanism. Acta Hydrobiologica Litunica. 6: 20–27 (translated from Russian). Daniulyte G, Naktinis J and Petrauskiene L. 1987. Dependence of anode electro-taxis efficiency of marine fishes on pulse current parameters. Acta Hydrobiol. Litunica. 6: 3–11 (translated from Russian). De Groot SJ and Boonstra GP. 1970. Report on the development of an electrified shrimp-trawl in The Netherlands. ICES CM. 1970/B:5. De Groot SJ and Boonstra GP. 1974. Notes on the further development of an electrified shrimp trawl in The Netherlands. ICES CM. 1974/B:5. Dethloff J. 1963. Problems of electro-fishing and their solutions. Second World Fishing Gear Congress, London, May 25–31, 1963. Diner N and Le Men R. 1971. Etude du champ électrique nécessaire à la taxie anodique du poisson. Rev. Trav. Inst. Pêches marit. 35(1): 21–34. Diner N and Le Men R. 1974. Seuils de taxie et de tétanie de bars, des mulets et des sardines dans un champ électrique impulsionnel uniforme. ICES CM. 1974/ B:16. Ellis JE. 1972. The use of electricity in conjunction with a 12.5-meter (headrope) Gulf-of Mexico shrimp trawl in Lake Michigan. NOAA Tech. Rep. NMFS SSRF-653. Fredenberg WA. 1992. Evaluation of electrofishinginduced spinal injuries resulting from field electrofishing surveys in Montana. Helena, MT: MT Dept Fish Wildl. Parks. Halsband E. 1967. Basic principles of electric fishing. In: Vibert R (ed). Fishing with Electricity, Its Application to Biology and Management. pp 57–64. London: Fishing News Books. Haskell DC. 1954. Reactions and motion of fish in a direct current electrical field. N.Y. Fish Game J. l(l): 47–64.

Electric Senses of Fish and Their Application in Marine Fisheries Hauck F. 1949. Some harmful effects of the electric shocker on large rainbow trout. Trans. Amer. Fish. Soc. 77: 61–64. Higman JB. 1956. The behavior of pink grooved shrimp, Penaeus duorarum Burkenroad, in a direct current electric field. Marine Laboratory, Florida. Tech. Ser. No. 16. Holmes R, McBride DN, Viavant T and Reynolds JB. 1990. Electrofishing induced mortality and injury to rainbow trout, Arctic grayling, humpback whitefish, least cisco, and northern pike. Manuscript 9, Div. Sport Fish, AK Dept Fish Game, Anchorage, AK. Hong QS, Huang YK and Chen ZX. 1981. Study on the electrical shrimp capture technology. Proc. China Fish. Soc. Fish Capture Group Meeting. pp 170–192 (in Chinese with English abstract). Holt J. 1992. Innovative (non-traditional) marine technologies which utilize animal behavior to enhance fishing selectivity and efficiency. Mar. Technol. Soc. Conf. Proc. 1992: 347–353. Horn W. 1976. Rationalisiering der Seezungenfisherei durch Einsatz elektrifizierter Baumkurren. Informationen für die Fishwirtschaft. 23(1): 20–22. Idyll CP. 1958. A shocking way to fish! Sea Front. 4(2): 88–100. Jesien RV and Hocutt CH. 1990. Method for evaluating fish response to electric fields. In: Cowx IG (ed). Developments in Electric Fishing. pp 10–18. Oxford, England: Fishing News Books. Kazlauskiene N. 1987. DC effect on the heart rhythm of fishes. Acta Hydrobiologica Litunica. 6: 54–68. Kessler DW. 1965. Electrical threshold responses of pink shrimp Penaeus duorarum, Burkenroad. Bull. Mar. Sci. 15(4): 885–895. Klima EF. 1968. Shrimp behavior studies underlying the development of the electric shrimp trawl system. Fish. Ind. Res. 4(5): 165–181. Klima EF. 1972. Voltage and pulse rates for inducing electrotaxis in twelve coastal pelagic and bottom fishes. J. Fish. Res. Bd Can. 29: 1605–1614. Kolz AL. 1993. In-water electrical measurement for evaluating electrofishing systems. USFWS Biol. Rep. 11. Kolz AL and Reynolds JB. 1990. Principles and techniques of electrofishing. U.S. Fish and Wildlife Service Fisheries Academy, Office of Technical Fisheries Training, Kearneysville, WV. Kolz AL, Reynolds J, Temple A, Boardman J and Lam D. 1998. Principles and techniques of electrofishing. U.S. Fish and Wildlife Service National Conservation Training Center.

233

Kreutzer CO. 1963. Utilization of fish reactions to electricity in commercial sea fishing. Second World Fishing Gear Congress, London, May 25–31, 1963. Kynard B and Lonsdale E. 1975. Experimental study of galvanonarcosis for rainbow trout (Salmo gairdneri) immobilization. J. Fish. Res. Bd. Can. 32: 300–302. Lamarque P. 1967. Electrophysiology of fish subject to the action of an electric field. In: Vibert R (ed). Fishing with Electricity: Its Application to Biology and Management. pp 65–92. Published by arrangement with the FAO by Fishing News (Books) Ltd. Lamarque P. 1990. Electrophysiology of fish in electric fields. In: Cowx IG and Lamarque P (eds). Fishing with Electricity, Applications in Fresh Water Fisheries Management. pp 4–33. Oxford: Fishing News Books. Lindsay CC. 1978. Form, function and locomotory habits in fish. In: Hoar WS and Randall DJ (eds). Fish Physiology, Vol. 7. pp 1–100. New York: Academic Press. Maksimov YM. 1977. Peculiarities in the manifestation of certain defensive reactions of pelagic fishes from the Gulf of Mexico in the area fished by an electrified bottom trawl. J. Ichthyol. 17(2): 305–312. Maksimov Y and Tamasauskas P. 1987. Reactions of Japanese scallop in homogeneous pulsed electric current field. Acta Hydrobiologica Litunica. 6: 75–78. Maksimov Y, Malevicius S and Yudin V. 1987. Selective effect of electrified field while fishing by electrotrawl. Acta Hydrobiologica Litunica. 6: 54– 74 (translated from Russian in Dutch). Mesa MG and Schreck CB. 1989. Electrofishing markrecapture and depletion methodologies evoke behavioral and physiological changes in cutthroat trout. Trans. Am. Fish. Soc. 118: 644–658. Mitton CJA and McDonald DG. 1994. Consequences of pulsed DC electrofishing and air exposure to rainbow trout (Oncorhynchus mykiss). Can. J. Fish. Aquat. Sci. 51: 1791–1798. Namboodiri KS, Vijayan V and Hridayanathan C. 1977. Development of electr-trawl system in marine environment. Fish. Technol. 14(1): 61–65. Nikonorov IV. 1963. Pump fishing for saury with light and electric current attraction. Second World Fishing Gear Congress, London, May 25–31, 1963. Northrop RB. 1967. Electrofishing: IEEE Trans. Biomed. Eng. 14: 191–200. Novotny DW. 1990. Electric fishing apparatus and electric fields. In: Cowx IG and Lamarque P (eds). Fishing with Electricity, Applications in Fresh

234

Fish Behavior near Fishing Gears during Capture Processes

Water Fisheries Management. pp 34–88. Oxford: Fishing News Books. Pease NS. 1967. The design and field testing of an electro-shrimp trawling system. FAO World Scientific Conference on the Biology and Culture of Shrimps and Prawns, Ciudad de Mexico, June 12– 24, 1967. Polet H. 2003. Evaluation of Bycatch in the Belgian Brown Shrimp (Crangon crangon L.) Fishery and of Technical Means to Reduce Discarding. PhD thesis, University of Ghent. Reynolds JB. 1996. Electrofishing. In: Murphy BR and Willis DW (eds). Fisheries Techniques, 2nd ed. pp 221–254. Bethesda, MD: American Fisheries Society. Reynolds JB, Kolz AL, Sharber NG and Carothers SW. 1988. Comments: Electrofishing injury to large rainbow trout. N. Am. J. Fish. Manag. 8: 516–518. Seidel WR. 1969. Design, construction, and field testing of the BCF electric shrimp-trawl system. Fish. Ind. Res. 4(6): 213–231. Seidel WR and Vanselous TM. 1976. An automated unmanned fishing system to harvest coastal pelagic fish. Mar. Fish. Rev. 38 (2): 21–26. Seidel WR and Watson JW. 1978. A trawl design: employing electricity to selectively capture shrimp. MFR Pap. 1325: 21–23. Sharber NG, Carothers SW, Sharber JD deBos, Jr. and House DA. 1994. Reducing electrofishing-induced injury of rainbow trout. N. Am. J. Fish. Manag. 14: 340–346. Sharber NG and Black JS. 1999. Epilepsy as a unifying principle in electrofishing theory. a proposal. Trans. Am. Fish. Soc. 128: 666–671. Scheminzky F. 1924. Versuche über Electrotaxis und Electronarkose. Pflügers Arch. ges. Physiol. 202: 200–216. Schreck CB, Whaley RA, Bass ML, Maughan OE and Solazzi M. 1976. Physiological responses of rainbow trout (Salmo gairdneri) to electroshock. J. Fish. Res. Bd Can. 33: 76–84. Snyder DE. 2003. Electrofishing and its harmful effects on fish. Information and Technology Report USGS/BRD/ITR-2003-0002. US Government Printing Office, Denver, CO. Sreedharan N, Vijayan V and Hridayanathan C. 1977. Development of electro-trawl system in marine environment. Fish. Technol. 14, No 1. Sternin VG, Nikonorov IV and Bumeister YK. 1972. Electrical Fishing, Theory and Practice. Pishchevaya Promyshlennost, Moskva, 1972. Translated from Russian: Israel Program for Scientific Translations, 1976.

Sternin VG, Nikonorov IV, and Bumeister YK. 1976. Electrical Fishing, Theory and Practice [English translation of Sternin et al. 1972 from Russian by E. Vilim]. Israel Program for Scientific Translations, Keter Publishing House Jerusalem Ltd, Jerusalem. Stewart PAM. 1972. An exploratory investigation into the effects of electric fields on burrowed Nephrops. Unpublished Typescript No. IR 71-34. Aberdeen, Scotland: DAFS Marine Laboratory. Stewart PAM. 1973. The selection of electrode materials for electrical fishing. ICES CM. 1973/B:11. Stewart PAM. 1975. Catch selectivity by electrical fishing system. J. Cons. int. Explor. Mer. 36(2): 106–109. Stewart, PAM. 1977. A study of the response of flatfish (Pleuronectidae) to electrical stimulation. J. Cons. Int. Explor. Mer. 37(2): 123–129. Taylor GN, Cole LS and Sigler WF. 1957. Galvanotaxis response of fish to pulsating direct current. J. Wildl. Manag. 21: 201–213. van Marlen B. 1997. Alternative stimulation in fisheries. Final report to the European Commission, Contract No. AIR3-CT94-1850. van Marlen B, Grift R, van Keeken O, Ybema MS, van Hal R. 2006. Performance of pulse trawling compared with conventional beam trawling. RIVOReport No. C014/06. Vanden Broucke G. 1972. Eerste resultaten in de electro-visserij. Mededelingen van het Rijksstation voor Zeevisserij, 68-TZ/50/1972. Verschueren B. (2008) Project “Ontwikkeling en demonstratie van een selectieve pulskor voor de visserij op grijze garnaal met het oog op een reductie van de teruggooi en de milieu-impact (PULSKOR)”: Eindrapportering, FIOV-project. Vibert R. 1967. Fishing with Electricity—Its Application to Biology and Management. Published by arrangement with the FAO by Fishing News Books. Wydoski RS. 1980. Effect of electric current on fish and invertebrates [mimeo]. USFWS National Fisheries Centre, Leetown, Kearneysville, WV. Yu C, Chen Z, Chen L and He P. 2007. The rise and fall of electrical beam trawling for shrimp in the East China Sea: technology, fishery, and conservation implications. ICES J. Mar. Sci. 64: 1592–1597. Zalewski M and Cowx IG. 1990. Factors affecting the efficiency of electric fishing. In: Cowx IG and Lamarque P (eds). Fishing with electricity, applications in freshwater fisheries management. pp 89– 111. Oxford: Fishing News Books and Blackwell Scientific Publications.

Electric Senses of Fish and Their Application in Marine Fisheries SPECIES MENTIONED IN THE TEXT brown trout, Salmo trutta bullhead, Cottus gobio bumper, Chloroscombrus chrysurus butterfish, Peprilus triacanthus Caspian kilka, Clupeonella spp. chub mackerel, Scomber japonicas common carp, Cyprinus carpio dragonet, Callionymus sp. eel, Anguilla anguilla flounder, Platichtys flesus gudgeon, Gobio gobio longspine porgy, Stenotomus caprinus mullet, Mugil auratus plaice, Pleuronectes platessa rainbow trout, Oncorhynchus mykiss

rough scad, Trachurus lathami round herring, Etrumeus teres round scad, Decapterus punctatus roussette, Scyliorhinus sp. Scyliorhinidae scaled sardine, Harengula jaguana seabass, Dicentrarchus labrax seahorse, Hippocampus sp. silver anchovy, Engraulis eurystole spanish sardine, Sardinella aurita sole, Solea solea spot, Leiostomus xanthurus sunfish, Lepomis sp. tench, Tinca tinca thread herring, Opisthonema oglinum tilapia, Tilapia mossambica

235

Part Three Contemporary Issues in Capture and Conservation in Marine Fisheries

Chapter 10 Technical Measures to Reduce Bycatch and Discards in Trawl Fisheries Norman Graham

10.1 INTRODUCTION Trawling is one of the most important fish capture techniques in use today (Fig. 10.1). It accounts for approximately 22% of the world’s fish production and is the principal technique used in the majority of demersal fish and shrimp fisheries (Kelleher 2005). However, it is also responsible for a disproportionate amount of discard; recent estimates (Kelleher 2005) suggest that over 50% of the world total discards can be contributed to this fishing method. This not only represents a waste of raw material, but it raises moral issues relating to “good” practice and can significantly reduce stock productivity and therefore available protein, which can present food security issues (Cook 2003; Harrington et al. 2005; Revill et al. 1999; Shepherd 1990). The terms “bycatch” and “discard” have been a source of confusion in many literature sources and can differ in meaning in different countries and management jurisdictions. It is important for the reader to distinguish between these terms. Both have become synonymous with negative aspects of commercial fishing and are considered by many to be interchangeable, although this is not the case. In the context of the work presented here, discard is considered to be the dumping of the unwanted portion of the catch, whereas bycatch is the part of the catch that is captured incidentally to the species toward which there is directed effort (nontarget

species) (Saila 1983) and as such may have some economic value. Figure 10.2 illustrates the relation among catch, landing, bycatch, discard, and the total fishing-related mortality. It should be noted that discard is NOT a subset of bycatch. In practice, the relative levels associated with each element are likely fishery dependent. For example, in Ghana, unwanted bycatch from the industrial shrimp fishery is not discarded but is sold to artisanal fishers, who then sell it for human consumption or for the production of agricultural feed (Nunoo et al. 2009). In the Nigerian beam trawl fishery for estuarine prawn (Nematopalaemon hastatus), 75% of the bycatch is discarded, and only bycatch species larger than 10 cm are retained and landed for human consumption (Ambrose et al. 2005). In summary, in many mixed-species trawl fisheries, not all bycatch ends up as discard and it can represent a significant source of revenue. These distinctions are important in determining fisheryspecific issues and identifying economically viable mitigation measures. This chapter briefly reviews the levels of bycatch and discard in world trawl fisheries and discusses the causes of bycatch and discard in different eases. The chapter then concentrates on technical measures to reduce bycatch and discard. Issues related to unaccounted or unobserved mortality of discards and escapee are discussed in Chapter 11.

239

240

Contemporary Issues in Capture and Conservation in Marine Fisheries

Figure 10.1. Principal components of a demersal single boat trawl. (Crown copyright, reproduced with the permission of Marine Scotland.) For color detail, please see color plate section.

10.2 BYCATCH AND DISCARD IN WORLD FISHERIES Trawl fisheries targeting shrimp, particularly those conducted in shallow tropical areas, account for almost 28% of global discards (1.6 million tons) with a weighted average discard rate of 62% (Kelleher 2005). Earlier estimates (Alverson et al. 1994) suggest that this is even higher at 37% of global discards, some 9.5 million tons. The degree of discarding varies considerably (Saila 1983), depending largely on the geographic location and whether all or part of the bycatch is used for other purposes such as animal feed. In the Gulf of California shrimp fishery, Perez-Mellado et al. (1982) estimate a mean weight of bycatch-toshrimp ratio of 9.8:1. Furnell (1982) calculated that only 5% by weight of the catch by Guyanese shrimp vessels consisted of the target shrimp species. Kelleher (2005) notes that over 30,000 and 23,000 tons are discarded in the Madagascar and Mozambique shrimp fisheries, while almost 94% of the 513,000 tons discarded in the Gulf of Mexico is attributed to shrimp trawl fisheries, accounting for almost 7.5% of the global total. The high rates observed are a consequence of several combined

factors. These areas tend to have high biodiversity and often function as nursery areas for juvenile fish and other organisms (Tonks et al. 2008). This, combined with the relatively small mesh size needed to retain the target species, results in the high discard rates observed. However, it should be noted that some shrimp fisheries have low discard rates due to improved gear selectivity, such as the North Atlantic Pandalus fishery, or full (or partial) utilization of the bycatch. Where bycatch is utilized, the impact on stocks of commercial and noncommercial fish populations will be the same as if the bycatch was discarded and may still warrant intervention to improve the exploitation pattern of the fishery. Demersal trawl (including beam trawl) finfish fisheries account for almost 19% of the global total landing (Kelleher 2005). The high use in European demersal fisheries such as the otter trawl fishery targeting fish and Nephrops (Graham and Ferro 2004), the beam trawl fishery for sole and plaice (ICES 2006a), and the mixed gadoid fishery are the main contributors to discards, estimated to be in excess of 900,000 tons annually in the North Sea alone, making it the single largest marine ecosystem contributor to global discards at 13% (Kelleher

Technical Measures to Reduce Bycatch and Discards in Trawl Fisheries

241

Animals Affected by Fishing Gear

Catch

Target species

Legal size Possession limit

Landing

Escaped/released

Nontarget species (Bycatch)

Undersized No economic value Over possession limit

Permitted Valuable Desirable

Live

Prohibited Undersized No value

Unobserved Mortality

Discard

Dead

Total Fishing Mortality

2005). It is estimated that over 330,000 tons with a discard rate of 69% is attributed to the demersal beam trawl fleet alone, although a more recent report estimated a lower rate of 55% (Anon. 2008). It should be noted, however, that since the publication by Kelleher (2005), it is likely that absolute discard levels have decreased due to significant contraction of both otter trawl and beam trawl fleets operating in the North Sea, although discard rates are likely to remain high. Other demersal trawl fisheries of note are the U.S. East Coast demersal trawl fisheries (Harrington et al. 2005), which is estimated by Kelleher (2005) to have discarded over

Dead

Live

Figure 10.2. Potential fate of animals entering a trawl and its relation to bycatch, discard, landing, and other components of fishing mortality.

130,000 tons in 2002, and the Moroccan and Argentinean industrial trawl fisheries (Kelleher 2005). Excessive discarding in the U.S. East Coast fishery has been recognized as far back as 1935 by the U.S. Department of Commerce. Herrington (1935) reported that since the introduction of trawling at the start of the nineteenth century, the impact of discarding on the stock was of growing concern. Discard rates by weight of 38% and 40% for cod and haddock, respectively, were reported and these levels were considered unsustainable at the time. The relative contribution of individual trawl fisheries in relation to global discards can be found in

242

Contemporary Issues in Capture and Conservation in Marine Fisheries

Kelleher (2005). The contribution that discarding makes to fishing mortality at a stock level is unknown in many fisheries due to a lack of analytical stock assessments, but given the examples here, the impact is likely to be considerable in many cases. However, it is important to note that while certain fisheries may have high discard rates, from a fishing mortality perspective it is the absolute discard levels (rather than rates) that are of significance. For example, small-scale trawl fisheries or those with relatively low applied effort may have very high discard rates, but their overall contribution to fishing mortality will be considerably less due to their lower catch levels.

10.3 CAUSE OF BYCATCH AND DISCARD Before considering the effectiveness of gear-related technical conservation measures (TCMs) for reducing discards, it is necessary to consider why discarding occurs in the first instance and then to determine appropriate mitigation methods. Crean and Symes (1994) note that the management framework used to regulate fisheries has a strong influence on discard rates. Those that rely extensively on output controls (e.g., quotas, catch composition limits, and minimum sizes) are often associated with excessive discard rates as fishers are often compelled to discard to remain within the regulatory constraints. Quota-induced discarding is often problematic in mixed-species fisheries in which the quota uptake is regulated by the monitoring of landings rather than catches. Where catching (quota) opportunities remain for one species but are exhausted for others, fishing activity can continue by retaining the species for which quota remains and discarding the remainder. Similarly, the practice of high grading is common where quotas are restrictive and fishers maximize their economic value by retaining only the most valuable grades or species. Discard rates are often elevated in fisheries that are overexploited, with few “old” fish in the exploitable stock. Fishers rely more on younger age classes that tend to be close to minimum market or legal limits. As gear selectivity is not knife-edged, large quantities of fish below minimum legal size (MLS) can be caught and subsequently discarded. The other primary stimulus to discard is associated

with market demands where fish of little or no commercial value are discarded due to market size preferences or absence of a market. This is particularly problematic in fishing areas with a moderate to high biodiversity such as tropical and temperate shrimp fisheries. While discarding caused by quota restrictions and high grading is problematic and is often justifiably maligned as the dumping of “good” fish, in practice, the discard of noncommercial organisms or juveniles of commercial species makes the greatest contribution to discarded biomass in the majority of trawl fisheries It is therefore the combined interplay of fishery regulations with economic considerations and the catch composition that determine the level and pattern of discarding associated with a particular fishery. The relative importance of each will be fishery specific, but understanding these drivers will help elucidate the optimal strategy to reducing unwanted catch.

10.4 TECHNICAL MEASURES TO REDUCE BYCATCH AND DISCARD In simplistic terms, trawls can be thought of as rather coarse filtering devices. Once inside the trawl, retention of fish is a function of individual morphology and some physical attribute of the gear such as mesh size, construction, and shape (Wileman et al. 1996). In practice, the process is more complex. Species- and size-specific behavioral reactions, fish condition, and environmental factors such as temperature, light turbidity levels, and fish condition and their interaction are all important in determining vulnerability to the trawl and, hence, catch composition. In addition, catching efficiency is related to other aspects of trawl design and operation. Door and sweep width, sweep angle, headline height, and ground gear construction are all known to be important design features (among others) that are important, and adjustments in these features/parameters can also offer important tools to reduce discards. The choice of mitigation method and appropriate configuration is dependent on the factors outlined here, and the specific design characteristics will be particular to the individual fishery. In general, many of the designs tested (and mandated) are based on generic concepts that have been trialed in a wide

Technical Measures to Reduce Bycatch and Discards in Trawl Fisheries range of fisheries and have tended to vary in technical aspects of their design and positioning within the trawl. Where several species are targeted and have similar morphological characteristics, manipulation of mesh size and/or shape or the inclusion of another modification to the trawl such as the inclusion of a square mesh panel (SMP) may be sufficient to reduce the level of discards. The situation is more complex when significant morphological differences exist between a range of target and nontarget species. This is typical of fisheries with a medium to high target species mix and particularly those where crustacean form an important component of the catch. The small mesh size necessary to retain the target shrimp often results in the retention of “juvenile” or unwanted fish species. Reducing discard levels in these fisheries is more complex and technically challenging. Even a relatively small increase in mesh size is likely to result in high losses of target species while having a marginal impact on the retention of unwanted species. Therefore, alternative mechanisms are needed. It is not the intention of this chapter to provide a fishery-by-fishery description that uses or has tested mechanisms to reduce discarding. It is, however, intended to give the reader an overview of some of the tools available to reduce unwanted catches. For the purposes of this chapter, three rather broad fishery definitions that attempt to encapsulate the majority of trawl fisheries are given in Table 10.1. This allows for very generic classifications of primary discard issues and identification of management goals and appropriate measures. The main interventions are those appropriate to: 1.

Reduce discards through alterations to mesh selectivity 2. Techniques to reduce bycatch with little or no economic value 3. Discard reduction in fisheries where a component of the bycatch is economically important or the fishery has a medium to high species diversity The list is by no means exhaustive and unlikely to cover all trawl fisheries, but it demonstrates the diversity of the discard issue in trawl fisheries. It also illustrates that the appropriate tools needed to

243

reduce discards must consider not only the management objectives in terms of mitigation but also what fishers are targeting in the first instance. The management objectives in Table 10.1 provide the basis of the grouping of available gear-related technical measures outlined in the remainder of this chapter. Some consideration is also given to the motivations necessary to get such technologies into practice in commercial fisheries. While the primary focus of this chapter relates to the modifications to conventional trawls, it should be recognized that gear modifications are not the only tools available to reduce discarding. Regulating spatial and temporal fishing activity is also used as an effective management tool in a wide range of fisheries. Permanent and real-time closures of areas with high discard levels can effectively reduce discard levels, as can restrictions on the time when fishing can occur (Graham et al. 2007). Often, combining spatial and/ or temporal activity in combination with gear modifications can be the best solution. 10.4.1 Reducing Target Species Discards by Controlling Size Selectivity By the early twentieth century, trawling had expanded to most shelf (less than 200 m) fisheries around the world and with associated increases in catching power, overexploitation and excessive discarding became evident in a number of fisheries, and managers attempted to control fishing mortality through the use of mesh size regulations. This of course was based on the assumption that (1) fish actually escaped and (2) they survive the escape process. Davis (1934) was the first to demonstrate that fish escaped from the codend during the trawling process. This was an important step, as the effectiveness of mesh regulations had long been argued by both scientists and fishermen. The conventional wisdom of the time considered that due to lateral tension, meshes remained closed during the trawling process and escapement only occurred during the hauling process when the mesh became slack and more open. As a result, fish would not survive the escapement process due to barotraumas. The observations of Davis (1934) were later confirmed by underwater observations of codends taken using remote vehicles and divers. These showed that as the catch accumulates, a band of

244

Contemporary Issues in Capture and Conservation in Marine Fisheries

Table 10.1. Broad Categorization of Demersal Trawl Fisheries with Generic Fishery Examples, Desirable Management Objectives, and Possible Technical Interventions to Achieve These. Single/Low Species Diversity (Finfish/Pelagic)

Fishery Type Example Fishery

Management Objective Technical Objective

Technical Solution

Medium to High Target Species Diversity with Crustacean or Other “Small” Target Species

Single Target Species with Unwanted Bycatch

North Atlantic/Pacific mixed gadoid trawl • North Sea flatfish beam trawl Reduce capture of juveniles of target species Increase length/age of first capture

•

Temperate/ tropical shrimp fisheries • Pelagic trawl

• •

Reduce capture of unwanted fish and megafauna Exclude unwanted fish and megafauna

Increase mesh size Alter mesh geometry Size sorting grids Incorporate species specific selection process

Physical or behavioral separation and exclusion with grids, coverless trawls, fisheye

Improve exploitation pattern of fish while retaining “small” target species Increase length/age of first capture of fish. Maintain exploitation pattern “small” target species Physical or behavioral separation with additional species specific selection process

•

open meshes forms in front of the catch through which fish—those that are physically able to do so—escape (Engås et al. 1989; O’Neill et al. 2003). However, survival was not actually demonstrated until the end of the twentieth century, and thus far it has only been assessed in a limited number of fisheries and for a small range of species (Suuronen 2005; also see Chapter 11). Studies have shown that the survival rate is dependent on both the species and the size of individuals (Anon 2007). In general, pelagic species appear to be more susceptible to post escape mortality than are demersal species, although the results, even for the same species, can be highly variable (Suuronen 2005). This is clearly an area that needs further research to determine if gear-related technical measures actually achieve the desired reductions in fishing mortality in practice. However, this field of gear science is technically challenging and needs robust methodologies as experience has shown that elevated mortality rates can be a consequence of poor experimental design rather than the escape process itself.

Tropical shrimp Mediterranean fish/ crustacean trawl

Codend mesh selection is not knife edged (there is no critical length at which all fish below escape and fish above are retained) but rather the probability of retention increases with fish length (Fig. 10.3). The length at which a fish has a 50% chance of retention is referred to as L50—in the case shown here, this lies at 30 cm. The steepness or sharpness of the curve is referred to as the selection range (SR) and is the numerical difference between L75 and L25. The lower the SR value (the dashed curve), the sharper is the selection. In practice, the selection profile varies between hauls about a mean value (Fryer 1991). The position of the selection curve relative to the x-axis is dependent on species, codend construction (mesh size, shape, and twine thickness), or, in the case of rigid grids, bar spacing and grid angle. It is also affected by a range of other uncontrollable factors such as catch size (O’Neill and Kynoch 1996), catch rate (Jørgensen et al. 2006), fish condition, and environmental factors (Özbilgin 1998; Özbilgin and Wardle, 2002). Many fisheries manage selectivity by regulating key

Technical Measures to Reduce Bycatch and Discards in Trawl Fisheries

245

Figure 10.3. Selection profile of a codend showing the parameters L50, the length that which a fish has a 50% probability of being retained, and the selection range (SR), a measure of the sharpness of selection. For comparison, the dashed selection curve shows an idealized selection profile with a narrow selection range.

codend design features such as minimum mesh size (MMS). Similarly, MLSs or marketing restrictions are often used to discourage the capture and marketing of juveniles. While intuitively these make sense, discarding can be encouraged if these regulatory controls are not correctly aligned or if there is economic advantage to fishing in areas with high abundance of juveniles to catch the larger individuals in the population. Figure 10.4 shows three theoretical selection profiles relative to an MLS of 30 cm (vertical line). Figure 10.4A shows a selection profile of a codend with a comparatively low L50 (approximately 25 cm), almost all (approximately 95%) fish above legal size are retained. However, as selection is not knife edged, a large proportion of fish below legal size are also retained, represented by the area beneath the selection curve (left half of panel). It may therefore be better to reduce the MLS so that it lies at a point with a low probability of retention (e.g., 20 cm) or use a gear that has a higher L50 so that all fish below MLS escape (Fig. 4B). However, to use the latter option, a considerable proportion of marketable fish will also escape, thus introducing an incentive to modify the fishing gear to make it less selective. Figure 10.4C shows the effect of having a wide selection range, resulting in consider-

able retention of fish below 30 cm and loss of fish above. To minimize loss of marketable catch and minimize retention of fish below MLS, it would be desirable to make selection as sharp as possible. In practice, discard levels are largely dependent on the population structure (age/size) of the fish entering the trawl. If the population being fished consists mainly of large fish, then discard rates with be correspondingly low, even if a small mesh size is used. Conversely, if the fished population consists of small fish, the associated discard rates will be high. This is typically seen during years of high recruitment or in overexploited fisheries. Therefore, discard rates are determined by the selection of the trawl and the population of fish on which it is deployed. In mixed-species fisheries, the relationship between mesh size and minimum legal sizes is often more complex, where a single mesh size is used to select a range of species often with differing MLSs. In practice, the choice of MLS is often a compromise to limit retention of small fish rather than one based on biological suitability. While it is generally acknowledged that increasing mesh size results in the retention of fewer “small” fish, it was perceived that such increases in mesh size would also result in high losses of marketable fish (e.g., that the selection range of

246

Contemporary Issues in Capture and Conservation in Marine Fisheries 1 (A)

Discards

0.75

Lost Marketable Landings

MLS

0.5

0.25

L50

0 1

Probability of retention

(B) 0.75

0.5

0.25 L50 0 1 (C) 0.75

0.5

0.25 SR 0 10

15

20

25

30 35 Length (cm)

conventional netting was too large). A range of mechanisms to try to “sharpen” the selection range of codends have been developed and tested. These have tended to focus on alterations to the mesh

40

45

50

Figure 10.4. Effect of modifying L50 relative to MLS (A, B) and the effect of a high SR (C) on retentions of legal and sublegal fish. The vertical line in each denotes the minimum landing size.

geometry in an attempt to make them open further, to maintain mesh opening during all stages of the trawling operation, and to increase the area of open meshes or through the addition of complementary

Technical Measures to Reduce Bycatch and Discards in Trawl Fisheries

247

Figure 10.5. Examples of mechanisms to adjust size selectivity. (A) Square mesh codend. (B) Square mesh codend panel. (C) Sort-V selection grid. (Crown copyright, reproduced with the permission of Marine Scotland.) For color detail, please see color plate section.

devices in addition to the codend itself. The perception of “poor” selectivity of diamond mesh codends is not new. Holt (1895) expressed concern about the number of small fish retained by English trawlers in the nineteenth century and constructed and tested

the first codend made from square mesh to try to alleviate this problem. Square mesh is constructed by orientating conventional diamond mesh so that it is hung on the “square” (Fig. 10.5A). The theory was to construct a codend so that all the meshes

248

Contemporary Issues in Capture and Conservation in Marine Fisheries

remained fully open during the entire trawling process. Holt reported that this design of codend “caught considerably less small fish than an ordinary codend with mesh of the same size.” Renewed research into the utility of square mesh codends began again in early 1980s (Walsh et al. 2000). They were tested in a range of fisheries across the globe, including Scotland (Robertson and Stewart 1988), Canada (Halliday et al. 1999), Norway (Isaksen and Valdemarsen 1988), Australia (Broadhurst et al. 2004), and the Mediterranean (Bahamona et al. 2006; Sala et al. 2008; Sardà et al. 2006). Research has shown that for roundfish species such as Atlantic cod (Gadus morhua) and haddock (Melanogrammus aeglefinus), for a given mesh size, the L50 increases (by approximately 10%) when switching from a diamond to a square configuration. The converse is true for flatfish species such as flounder and sole, where it has been shown that square mesh reduces L50 (Fonteyne and M’Rabet 1992; Walsh et al, 1992). The effect on SR is less clear—some authors report significant reductions in selection range of approximately 33% (Halliday and Cooper 2000; He 2007), whereas others note no significant effect (Campos et al. 2002). In general, square mesh codends offer means to regulate L50 and can also help reduce the capture of unwanted benthos and detritus (Revill and Jennings 2005). However, constructing codends from square mesh creates several problems in terms of mesh strength and distortion (Robertson and Polanski 1984) and is generally disliked as a concept by the commercial fishing industry for these and other practical considerations such as repair difficulties. Despite this, the use of full square codends is mandatory in a number of North West Atlantic demersal fisheries and in EC Mediterranean demersal trawl fisheries. Intuitively, it would be expected that increasing the number of more “open” meshes would result in improved selection over conventional diamond mesh codends. However, underwater observations (Engås et al. 1989) have shown that despite having many opportunities to escape through open square meshes, Northeast Atlantic haddock only attempt escape when in close proximity to the accumulated catch in the codend regardless of whether the codend is constructed from diamond or square mesh

codends. Behavioral studies conducted by Glass and Wardle (1995) and Madsen et al. (1998) have attempted to encourage escapement by providing a visual stimulus through the addition of areas of dark netting or canvas panels behind the area of open meshes. Earlier laboratory experiments demonstrated that Atlantic mackerel (Scomber scombrus) were less inclined to pass through netting that appeared to be impaired by visual illusion through the inclusion of an area of dark contrast behind the netting panel (Glass et al. 1995). While the later field experiments of Glass and Wardle (1995) demonstrated significant increases in escape rate, the work of Madsen et al. (1998) was not so conclusive. However, laboratory experiments clearly demonstrate that fish behavior can be modified through visual stimulus. Further investigation into how these and other stimuli could be used in promoting fish escapement is required. In the 1990s, researchers from Poland and Germany began experiments with T90 codends, where codends are constructed from conventional diamond mesh but the netting is turned through 90 degrees. By hanging a mesh in such a way, the construction of mesh knots is such that there is a physical resistance to close, resulting in a mesh that remains more open in comparison to conventionally hung netting (Hansen 2004). Twine thickness (as a proxy for bending stiffness) and the number of meshes in circumference of a codend have also been shown to affect selection (L50). To avoid negating the effects of increasing mesh size to increase L50, both factors are regulated in a number of fisheries. Reducing these, also results in a higher L50 for a given mesh size. Based on a meta-analysis of North Sea Haddock selectivity data, it was estimated that reducing the twine thickness by 1 mm, results in an increase in L50 of approximately 1.2 cm while reducing the codend circumference by 20 meshes results in an increase in L50 of approximately 1.7 cm (Anon. 2003). Herrmann et al. (2007) noted that the latter effect as the principal factor in the “improved” selectivity associated with the T90 codend rather than the alteration to mesh geometry per se. To compensate for the increased codend circumference (due to the “wider” mesh opening), the number of T90 meshes has to be lower than for a conventional codend.

Technical Measures to Reduce Bycatch and Discards in Trawl Fisheries While the above section has focused on alterations to codend construction for controlling size selection, a number of researchers have focused on the use of complementary (to the codend itself) size-selective devices. These have focused primarily on two concepts: a rigid or semirigid grid (Fig. 10.5B) and/or exit or panels and/or windows (Fig. 10.5C). Both provide areas that remain fixed for fish to escape during the entire trawling process. While the intention is to provide the primary escape zone, the addition of these devices effectively creates a dual selection process—escape through the grid or panel and the codend itself. To retain the perceived benefits of square mesh codends and avoid practical problems of strength and knot slippage, Arkley (1990) started investigations into the utility of the SMP or “window” (Fig. 10.5C) to reduce discards of bycatch species in the U.K. Nephrops fisheries. This concept was not new and was originally tested in the 1920s as the “Gelder” codend, a patented design from the Savings Trawl Net Company. While successful at lowering levels of discards in the Nephrops fishery, square mesh panels have also been extensively tested and mandated in mixed demersal finfish fisheries to improve size selection (Graham et al. 2003; Graham and Kynoch 2001). The inclusion of an SMP improves the size selection of certain species, depending on escape behavior and the positioning of the SMP in the trawl. In general, citing the SMP close to or in the codend improves its efficiency (Graham et al. 2003; Graham and Kynoch 2001; O’Neill et al. 2006). The BACOMA panel, where almost the entire upper part of the codend is constructed from square mesh, is mandated in the Baltic cod fishery. This configuration has been shown to significantly increase L50 and reduce SR in comparison to conventional diamond mesh codends in a wide range of experiments (ICES 2006b). This is a surprising observation, given the somewhat inconclusive results from experiments with full square mesh codends described earlier. However, this may simply be an artifact of the species and the mesh size used. Graham et al. (2003) conclude that where SMPs are used to adjust the size selectivity of target species, the same selective properties can be achieved by simply increasing mesh size. The same conclusion was reached in 1935 by Herrington (1935).

249

At around the same time as the investigations into square mesh codends, researchers in Norway began experimenting with the use of rigid grids as an alternative to mesh selection (Fig. 10.6A). This was motivated by the success of grids in reducing unwanted bycatch in the small mesh fishery for pink shrimp (Pandalus borealis) (see Section 10.4.2) and problems encountered with the meshing of redfish (Sebastes spp.), which rendered the use of square mesh codends inappropriate for the Barents Sea demersal trawl fishery. The need to improve selectivity and reduce the capture of fish below MLS in the Norwegian Barents demersal fishery was important not only to the administrators but also to the fishing industry. In the 1980s, Norway introduced area closures and the obligation for vessels to switch fishing ground if the catch composition has in excess of 15% of fish below minimum catch size [note the distinction between catch rather than landing size]. At the time, it was considered that the selection range of conventional diamond mesh codends was too wide to offer any reduction in the capture of fish below MLS through increasing mesh size as this would lead to unacceptably high losses of marketable fish. Extensive research was conducted into determining the selective properties of rigid grids (ICES 1996). However, much of this work focused on determining the selective properties of the grid in isolation—rather than the combined effect of grid and codend selectivity that would be used in commercial conditions. Notwithstanding, the use of grids with a bar spacing of 55 mm fitted in a trawl with a 135-mm codend reduced the retention of fish below MLS to such an extent that it resulted in access to a large number of otherwise closed areas and in wide-scale voluntary use. Recent research into grids has focused on the combined effect of grid and codend selectivity and comparing this with the selection profile of “standard” diamond mesh codends. Kvamme and Isaksen (2004) compared the selectivity of a 55-mm grid fitted into a trawl with a 135-mm codend and derived the selectivity of the individual components (codend only, grid only, and combined grid and codend). They observed a significant difference in L50 when compared with the 135-mm codend but no significant difference in selection range. This raised the question whether similar selection

250

Contemporary Issues in Capture and Conservation in Marine Fisheries

Figure 10.6. Examples of designs aimed at improving species selectivity. (A) Inclined separator panel. (B) Nordmøre shrimp grid. (Crown copyright, reproduced with the permission of Marine Scotland.) For color detail, please see color plate section.

profiles (both L50 and SR) could be achieved simply by increasing codend mesh size. Jørgensen et al. (2006) conducted a series of experiments comparing the combined selectivity of a 55-mm grid and 135-mm codend with that of a 155-mm diamond mesh codend. They found no significant difference in either L50 or SR between the two designs and conclude that the mandatory introduction of the selection grid into Norwegian fisheries could have been achieved by increasing the then mesh size (135 mm) to 155 mm and that the earlier concerns regarding high fish loss due to the “flat” selection of conventional codends were largely unfounded.

10.4.2 Reducing Unwanted Bycatch in Single-Species Fisheries The earlier sections focused on controlling the selectivity of trawls to regulate the size of fish retained as a means to prevent the capture of small individuals; in many circumstances reducing unwanted bycatch through such means would have a significant impact on target catches. It is, however, possible to manipulate differences in physical size or behavior between the target and unwanted component of the catch. Here, it is necessary to distinguish between fisheries where all the bycatch is unwanted and those where a certain component

Technical Measures to Reduce Bycatch and Discards in Trawl Fisheries of the bycatch is economically important to the fishery. In many fisheries, all of the bycatch is unwanted and subsequently discarded. This is typical of many shrimp fisheries, yet these suffer from the highest discard rates in comparison to other trawl fisheries. Shrimp fisheries generally use small mesh sizes, which are necessary to retain the target species and are often conducted in areas with high concentrations of juveniles for example in nursery areas. In the 1960s and 1970s, attention was given to using the unwanted portion of the catch for animal/aquaculture feed or through novel processing techniques to produce fish protein for human consumption. However, by the 1980s, focus shifted toward reducing bycatch to limit the impact on commercially important and other species. Broadly speaking, the choice of mitigation method depended on whether it is possible to physically exclude the bycatch through physical (size and shape) and behavioral differences between the target and nontarget species. This is largely a choice based on the relative size difference between the target and the unwanted component of the catch. Physical separation of unwanted bycatch has tended to focus on the use of netting or rigid gird (grate) barriers inserted inside the trawl that inhibit the passage of unwanted species into the codend. Due to the smaller size of the target species, these are able to pass unhindered through the barrier and into the codend. Such an approach is often used to separate and release mega fauna such as turtles, cetaceans, pinnepeds, and sharks. The Nordmøre (shrimp) grid was originally developed by Norwegian shrimp fishermen to reduce the capture of jellyfish, but its potential to reduce the capture of fish bycatch was soon recognized by researchers. The adaptation of this device has proved highly successful in reducing unwanted fish bycatch in several temperate shrimp fisheries and is also extensively used worldwide to prevent the capture of marine turtles in many tropical shrimp fisheries. The Nordmøre grid consists of a series of longitudinal bars held inside a frame, which offer a physical barrier to the passage of fish, which are guided out through the escape hole, while the shrimp, which are physically smaller, pass through the bars and continue to the codend (Fig. 10.6A). The bar spacing determines the size-

251

selective properties of the grid. Nordmøre grids, mandatory in most North Atlantic Pandalus borealis shrimp fisheries, have been shown to successfully reduce the bycatch of juvenile cod and haddock (Isaksen et al., 1992) as well as a range of other species. Richards and Hendrickson (2006) note that since the introduction of the grid into the Gulf of Maine (U.S.) fishery, the percentage of bycatch declined from nearly 50% before the Nordmøre grid was required to about 15% after the Nordmøre grate was required, although the authors note that this is about 35% less than suggested from controlled experimental studies. It is estimated that over 1000 vessels, catching over 300,000 ton of shrimp annually use grids in the Pandalus borealis fisheries, due to both mandatory legislation and voluntary usage in the North Atlantic (ICES, 2005). In recognition of the global significance of bycatch associated with tropical shrimp trawling, between 2002 and 2008, the United Nations Environment Program (UNEP), co-funded by the Global Environment Facility (GEF), FAO, and 12 participating countries, have undertake a program, “Environmental Impact from Tropical Shrimp Trawling through the Introduction of Bycatch Reduction Technologies and Change of Management.” A range of bycatch reduction devices have been tested and, although not finally reported, the mid-term evaluation of the project (Westlund 2006) and the evaluation conducted by the ICES Working Group on Fishing Technology and Fish Behavior (WGFTFB) provides preliminary information on the success of the project. In fisheries where the bycatch is of no economic importance, devices such as square mesh panels, fish eye, and the Super-Shooter (see Eayrs 2007 for a technical description) have been tested. The effectiveness of these has been variable and, in some cases, significant reductions in discard were achieved, while in others, the use of the devices resulted in high losses of shrimp with little or no impact on discard levels. From the results presented in ICES (2006b), it is concluded that fine-tuning is required to optimize the performance in maximizing reductions in discard and minimizing losses of target species. This fine-tuning is likely to be fishery and gear specific and demonstrates that technology transfer is not straightforward and can require considerable extension work.

252

Contemporary Issues in Capture and Conservation in Marine Fisheries

As well as being effective in shrimp fisheries, the Nordmøre grid has been shown to reduce unwanted bycatch in finfish fisheries. Halliday and Cooper (2000) have demonstrated reductions of 85% to 98% by weight of pollock and haddock and 48% to 70% of other species associated with the Canadian fishery for silver hake (Merluccius bilinearis). Kvalsvik et al. (2006) noted reductions in bycatch of haddock excess of 60% in the North Sea industrial fishery for Norway pout (Trisopterus esmarki). In the same fishery, Eigaard and Holst (2004) combined the grid together with an SMP. Although the reductions in bycatch were marginally lower, losses of commercial species were considerably lower than reported by Kvalsvik et al, (2006). Grids have also been tested in a range of other tropical and temperate fisheries. Hannah and Jones (2007) note that prior to the introduction of grid technology into the U.S. Pacific Coast fishery for ocean shrimp (Pandalus jordani); the fishery was generally mixed, with fish bycatch accounting for between 32% and 61% of the total catch weight. However, due to low TACs for bycatch species and rapid uptake, premature closures of the shrimp fishery were common. With the mandatory introduction of grids, the fishery was transformed to a single species and bycatch reduced to less than 8%. One of the greatest successes attributed to grids is the worldwide reduction in the unintentional bycatch of marine turtles. Tropical shrimp trawling was responsible for a significant proportion of anthropogenic turtle mortality, but this has been significantly reduced through the widespread use of the turtle excluder device (TED) in many tropical and subtropical shrimp fisheries. The TED was first developed in the 1970s as a means to reduce bycatch in the Gulf of Mexico but was later modified and subsequently patented to reduce capture of marine turtles. The TED is constructed in much the same way as the Nordmøre grid mentioned earlier with the exception that the bar spacing is much larger. Driven largely by a U.S. trade embargo on shrimp imports from countries not certified as using TEDs, this technology is now being used in over 40 countries This is likely to have a major impact on reducing sea turtle mortality globally as it is estimated that TEDs reduce turtle bycatch in excess of 95% (Brewer et al. 2006; Watson and Seidel 1980).

Nordmøre style grids are mandatory in Swedish coastal Nephrops fisheries, where single grids, used in conjunction with square mesh codends, have created a single-species target fishery by greatly reducing the capture of finfish and improving the size selection of Nephrops. This has effectively reduced the capture of sublegal fish by approximately 70% and excludes almost all fish above legal size (Graham and Ferro 2004). Due to handling issues associated with rigid grids voiced by fishermen, “soft” designs that try to mimic the effect of grids have also been developed. The inclined separator panel (Figure 10.6B) is mandated in certain areas of the Irish Sea to exclude groundfish (primarily cod) in the Nephrops fishery. Instead of using a series of bars, a gap below the lower leading edge of the panel is maintained to allow the passage of Nephrops to the codend, while the netting panel guides fish out of the escape hole. This has also been modified to maintain the marketable (above legal size) component of the catch by the attachment of a larger mesh codend to the escape opening. While these designs tended to focus on the separating species physical means, separation breaks down when there is little difference in size between the target and nontarget catch, and for this reason, grids do not exclude small fish that are able to pass through the bars of the grid. Behavioral differences have been used in fisheries to try to both exclude all bycatch or to separate and improve size selectivity through the use of a secondary device (e.g., another codend). Considerable research has focused on studying and using species-specific behavior as a separating mechanism. Typically, this relies on exploiting species specific differences in swimming ability, spatial preferences within the trawl, reactions to light contrast, and water flow. Shrimp have limited swimming ability, and this has been successfully used in a number of cases to reduce discards of juvenile fish. Red snapper (Lutjanus campechanus) is an important species in commercial and recreational fisheries in the Gulf of Mexico (Watson et al. 1988). Bycatches of juveniles in the shrimp fishery are known to inhibit stock productivity, and attempts to reduce bycatch levels through physical separation have largely been unsuccessful. Behavioral studies (e.g., Engås and Foster, 2002) have led to developments such as the fisheye, fish

Technical Measures to Reduce Bycatch and Discards in Trawl Fisheries box, and the radial escape section, which provide escape openings within the trawl that rely on active escape behavior of fish. In the case of the fishbox and the nested cylinder bycatch reduction device, a runner-up in the 2008 WWF Smartgear competition (see www.smartgear.org for further details), also provide stimulus by providing areas of minimal water flow or generate countercurrents or turbulence adjacent to the escape opening to encourage fish to exit. The nested cylinder device also uses negative phototaxis by discouraging fish, which are dark-adapted (red snapper), to move away from a brightly illuminated area and out of a darkened escape vent. This device has shown to reduce discarding in the Gulf of Mexico shrimp fishery by as much as 55%. Eayrs (2007) and Broadhurst (2000) provide excellent overviews of the available techniques to reduce unwanted bycatch in tropical shrimp fisheries. Eayrs (2007) noted that the choice of device is largely governed by the physical size of the bycatch to be excluded and provided a range of options depending on the discard problem. For the exclusion of “small” fish, devices such as the square mesh codend, square mesh window, or grid are considered to be the most appropriate. For larger fish, the insertion of an exit hole or large mesh panel that relies on swimming ability to access them are considered more appropriate, while the escapement of mega fauna is best achieved through the use of a physical barrier such as grids with large bar spacing. From a practical perspective, the effectiveness of all of these devices will be dependent on fine-tuning to attain optimal positioning—particularly those that rely on active escape behavior. The positioning of SMPs, fish eyes, and other such devices must be tailored to local conditions and will require some degree of trials to determine what these are. Brewer et al. (2006) note that suboptimal positioning of the SMP was likely to have reduced its effectiveness in excluding sea snakes (and other small bycatch) in comparison to earlier studies, and Heales et al. (2008) conclude that panel efficiency may be improved by moving it closer to the codend. To ensure optimum performance, in practice, it is necessary that fishermen, gear technologists, and regulators collaborate to maximize the reduction of unwanted bycatch. Regulations are also need to be sympathetic to local

253

conditions and to understand variability in performance between variations in gear designs. 10.4.3 Reducing Capture of Specific Species or Sizes in Mixed-Species Fisheries With increasing frequency, specific species caught in mixed species fisheries are subject to restrictive fishing opportunities (low TACs or moratorium) due to poor stock status of one or more species in the complex. This generally means that fishing opportunities for other species that are taken in the fishery are curtailed, although on their own they could support a viable fishery. As an alternative to reductions in fishing opportunities, a number of recent research programs have focused on excluding particular species in mixed-species fisheries as a means to reduce fishing mortality. The cutaway trawl (Fig. 10.7) has been developed for use in fisheries where particular species are to be avoided because they are of little or no economic value or because of restrictive fishing opportunities. The design uses behavioral differences between the target and nontarget species. Unlike the other devices described in this chapter, the cutaway trawl allows the escape of the unwanted species before they enter the trawl. The majority of demersal trawls are designed with a cover or roof, which helps retain species that tend to rise in the trawl mouth. By removing this, fish rise up and escape unhindered. The design has been found effective in mixed flatfish/cod fisheries in the United States (Mike Pol, unpublished observations) and the Faroe Islands (Thomsen 1993) and in the Nephrops fisheries in Europe (Revill et al. 2006) and Pandalus shrimp fishery in Gulf of Maine (He et al. 2007). Thomsen (1993), using underwater video observations, quantified the behavior of cod and flatfish at various positions within the trawl. The majority of cod were observed entering close to the lower panel, but as they moved farther back, cod were seen to rise upward. Thomsen (1993) then used these observations to exclude the majority of cod and haddock while maintaining catches of flatfish by removing the upper panels of the trawl. Similar results were obtained in the New England multispecies demersal fisheries with the objective of excluding cod from flatfish and successfully reduced the capture of cod by in excess of 75% compared with

254

Contemporary Issues in Capture and Conservation in Marine Fisheries

Figure 10.7. Cutaway trawl aimed at improving species selectivity. (Crown copyright, reproduced with the permission of Marine Scotland.) For color detail, please see color plate section.

a standard net (Mike Pol, personal communication). While some loss of target yellowtail flounder (Limanda ferruginea) was noted, it is considered that this could be offset as the modification allows access to a fishery that would otherwise be closed. As Nephrops tend to keep low as they enter the trawl mouth, the only reason for having an upper panel above and forward of the groundgear in a Nephrops trawl is to maintain the capture of fish that tend to rise as they fall back into the trawl. Recent trials by Arkley and Dunlin (2002) and Revill et al. (2006) have shown that by extending the headline in conjunction with a large mesh panel behind, it is possible to exclude some finfish species (haddock and whiting). Beutel et al. (2008) successfully designed a demersal trawl to capture only haddock in the mixed species fisheries of the Georges Bank. They simply increased the mesh size (up to 240 cm from approximately 15 cm) and “eliminated” the capture of all species of flounders, skates, and monkfish as well as significant reductions in cod catches. Similar results were obtained when these modifications were tested in the North Sea (Holst and Revill 2009). Other modifications have also been tested to reduce the retention of cod in demersal trawl fisher-

ies through modification to the ground gear. Species-specific escapement under trawls has been demonstrated by several authors (Engås et al. 1998; Engås and Gødo 1989; Ingolfsson and Jørgensen 2006). This has allowed the development of trawls aimed to reduce the capture of Atlantic cod by modifying or removing the ground gear to facilitate the escapement of cod. This demonstrates that significant reductions in unwanted species can be obtained by relatively straightforward modifications to trawl design. The remaining question is economic feasibility—Is it viable to change to a single-species fishery? Section 10.4.3 provided examples of how discards can be reduced in single- or mixed-species fisheries through alterations to mesh size or geometry or by the inclusion of additional selective devices for altering size selectivity. Preceding paragraphs in this section discussed some mechanisms to reduce unwanted by catch through physical exclusion or through the use of behavioral differences. There are circumstances where these two approaches can be combined to reduce the capture of sublegal bycatch while maintaining the capture of the legally sized component. Such a dual approach is necessary where a small mesh size is

Technical Measures to Reduce Bycatch and Discards in Trawl Fisheries required to retain the target species, such as shrimp or Nephrops, which is suboptimal for the wanted bycatch. To retain the target species, shrimp fisheries, the mesh sizes required is smaller than that for most finfish species. Bycatches and subsequent discarding can be considerable due to minimum landing size restrictions on the bycatch (Evans et al. 1994; Stratoudakis et al. 2001). However, the legal size fish bycatch is an important economic component of many shrimp (Ambrose et al. 2005; Eayrs et al. 2007; Nunoo et al. 2009) and Nephrops (Graham and Ferro 2004) fisheries and the exclusion of this does not pose an economically viable option to reduce discards. While the use of juvenile fish for human consumption or animal feed purposes in tropical shrimp fisheries may represent an underutilization of natural resources, it is important to recognize that food security is often a critical issue in many regions. In the Mid-Term Review of the UNEP/ GEF project “Reduction of Environmental Impact from Tropical Shrimp Trawling through the Introduction of Bycatch Reduction Technologies and Change of Management,” FAO recognizes that in countries such as the Philippines, Indonesia, and Vietnam (Eayrs et al. 2007) in Southeast Asia and in Nigeria, all or most of the bycatch carries some commercial value and this should be considered when deciding on appropriate devices that are suited to local needs. This precludes the use of devices such as Nordmøre grids that exclude this catch component. The Juvenile and Trash Excluder

255

Device (JTED) was developed in Southeast Asian Fisheries Development Centre (SEAFDEC) and tested in several Southeast Asian countries (see Chokesanguan et al. 2002 for details). The JTED permits the retention of marketable fish (typically greater than 10–15 cm) while excluding fish that would normally be converted to fish meal. In the wide and extensive trials conducted under the auspices of the UNEP/GEF project, releasing of “juvenile” and “trash” fish has been achieved through the introduction of JTED into the fleets on a either voluntary or mandatory basis. With the advent of underwater observations on towed fishing gears conducted using underwater vehicles and divers in the 1970s and 1980s, understanding of the fish capture process expanded tremendously (Priestley et al, 1985; Urquhart and Stewart 1993). Prior to this, trawl design and its effect on fish capture were largely a process of trial and error and evolved due to the ingenuity of individual fishermen. While the first underwater observations of trawls were undertaken in the 1950s, with the advent of SCUBA, Main and Sangster (1981) were the first to report distinct between-species behavioral differences in response to an approaching trawl and inside the trawl mouth. This led to the classic description of cod, haddock, and whiting behavior shown in Figure 10.8. By using differences in vertical behavioral patterns at the mouth of the trawl, the horizontal panel separator trawl (Fig. 10.9A) was developed to segregate species into specific areas within the net. A single panel of netting

Figure 10.8. Behavior of cod, haddock, and whiting in the mouth of a demersal trawl. (Crown copyright, reproduced with the permission of Marine Scotland.)

256

Contemporary Issues in Capture and Conservation in Marine Fisheries

is inserted horizontally within the trawl, dividing the trawl into upper and lower components; the separating panel may start level with the groundrope center or further aft. Generally, separator trawls are designed with two separate codends and extensions, which are tailored to join at the upper and lower body of the trawl, divided by the panel. Shrimp/Nephrops and some finfish species, particularly cod and flatfish, will tire and fall back under the separating panel, whereas haddock and whiting

tend pass over the panel and into the upper portion of the trawl (Dunlin 1998; Main and Sangster 1981, 1985; Wardle 1986). More recently, attempts have been made to also encourage cod to enter into the upper component of the trawl through the addition of guiding ropes to act as visual stimulus (Fig. 10.9B). The upper and lower codends may have different specifications (e.g., mesh sizes), providing the opportunity to manipulate codend selectivity to suit the species entering each codend. Separator

Figure 10.9. Examples of designs to segregate and select different species components. (A) Horizontal separator trawl used in mixed finfish or mixed Nephrops/finfish fisheries. (B) Horizontal separator with guiding ropes to “encourage” cod into the upper compartment. (Crown copyright, reproduced with the permission of Marine Scotland.) For color detail, please see color plate section.

Technical Measures to Reduce Bycatch and Discards in Trawl Fisheries

257

Figure 10.10. Doublegrid and codend system for improving size selection of Nephrops and finfish. (Crown copyright, reproduced with the permission of Marine Scotland.) For color detail, please see color plate section.

trawls have also been successfully used to improve size selection in demersal mixed finfish fisheries. Cotter et al. (1997) and Arkley et al. (1995) observe an average of 87% (range 74% to 96%) cod were retained in the lower codend with 20% to 24% of the haddock and all flatfish. Engås et al. (1998) conducted a series of experiments in the Barents Sea with a separator trawl. The retention of haddock and saithe in the upper codend was approximately 89% and 66%, respectively. As an alternative to the separator trawl, Graham and Fryer (2006) combined the segregation properties of grids with a two-tier codend to reduce the capture of sublegal fish while retaining the marketable component (Fig. 10.10). Most small Nephrops passed into the lower codend, with the proportion passing into the upper codend increasing to between approximately 40% and approximately 70% at carapace lengths of 50 mm. The proportion of haddock passing into the upper codend increased from approximately 60% for 10 cm haddock to approximately 85% for haddock of 20 cm or more. The proportion of dab (Limanda limanda) passing into the upper codend increased from approximately 30% for 10 cm dab to approximately 75% for dab of 20 cm or more. Clearly, improving the selection profile in multispecies fisheries is technically challenging and present operational issues associated with the practical use of such modifications. Facilitating the use of these gears in commercial practice is likely to require strong incentives to overcome these barriers.

10.5 IMPLEMENTATION OF DISCARD REDUCTION MEASURES IN TRAWL FISHERIES From the overview presented here, it is clear that there are a number of fisheries where the use of technical modifications have successfully improved selectivity and resulted in significant reductions in discards. However, despite these successes, many fisheries still have excessive discarding (Kelleher 2005). While the technology to reduce discards may exist, its utilization in commercial fisheries appears to be lacking. There are two likely explanations. First, it may be technically difficult to separate the target from the unwanted species without incurring losses of target species that render the fishery economically unviable. Second, where the technology has been tried and tested, there is a lack of incentive for fishers to use it. Short-term losses or technical implementation difficulties such as onboard handling issues, safety concerns, or simply conservative attitudes can present a significant barrier to utilization also need to be considered. However, it is clear that in many management zones, there appears to be little or no incentive at an individual vessel level to reduce discards. Where fisheries are regulated based on landings rather than catches, the most cost-effective option available is to discard unwanted catches rather than to use technical solutions to avoid initial capture. This lack of cost associated with discarding means that there is little benefit at the individual vessel level to undertake remedial action and apply

258

Contemporary Issues in Capture and Conservation in Marine Fisheries

mitigation methods. While there may be a desire to improve the exploitation pattern of one’s fishing gear, the fact is that reducing discards will often lead to reductions in fishing efficiency and/or require increased capital investment. As a consequence, unilateral action to reduce discarding will tend to result in a competitive disadvantage relative to others engaged in the fishery, providing little or no incentive. Unless costs associated with discarding are internalized at an individual level, where failure to reduce discards results in a competitive disadvantage, those who try to act in a responsible manner under the current framework will continue to be disadvantaged. Graham et al. (2007) reviewed the management framework of three demersal trawl fisheries in the North Atlantic and Pacific and noted that in fisheries where there is a penalty associated with the capture of unwanted species or juveniles, these tend to have lower discard levels. There are a number of penalties or costs that can apply, such as limiting of access to fishing areas unless catches are maintained below predefined limits or where premature closures of fisheries are triggered by the uptake of a bycatch quota. These have all provided strong incentives to fishers to improve their selectivity in the cases examined. Maximizing fishing opportunities but maintaining realistic and predefined boundaries is likely to offer the best incentive for fishers to reduce unwanted discards and bycatches in trawl fisheries. Successful implementation of mitigation measures in shrimp fisheries in developing countries is problematic due to inadequate enforcement and the importance of bycatch as a cheap food source (Eayrs 2007). Under such conditions, it is difficult to identify incentives that could result in the adoption of more selective gears without these issues being resolved. However, the introduction of the U.S. trade embargo in 1989 on countries that do not employ tools to reduce turtle bycatch offered a successful stimulus to the adaption of TED technology across a number of countries. Adopting a similar approach to shrimp fisheries with excessive discard problems may offer a potential solution. Coupling this to the international efforts on technical solutions may offer some prospect of improving exploitation in these fisheries.

10.6 DISCUSSION Discarding of juvenile fish of commercial importance can be reduced through sharpening the selection range of trawls. This is commercially advantageous as it maximizes the retention of marketable fish while minimizing retention of juveniles. While reducing selection range may be desirable from a fisher ’s perspective, MacLennan (1992) suggests that actually increasing the selection range may be more beneficial in terms of reducing year-on-year variability in fishing opportunities (e.g., TAC) and would provide a more stable fishery. Similarly, Kvamme and Frøysa (2004) note that improvements in stock health are more readily achieved through increases in the length of first capture (e.g., higher L50 values) rather than reducing selection range. Therefore, reductions in SR may be in conflict with the desire to have more stable production from the fishery over the medium term. However, if increasing L50 and limiting selection range is an objective, there appears to be little difference in the available tools (e.g., mesh size/shape/ construction or through supplementary modifications such as grids and SMPs). It is important to recognize that the use of supplementary devices can increase gear costs and introduce more complex monitoring, control, enforcement, as well as practical handling issues. Considering this, it is important to determine whether “alternative” designs offer any advantage over simply increasing mesh size alone, as often this will offer the simplest option. However, the selectivity of trawls fitted with supplementary devices may be less affected by environmental, morphological, and operational conditions, and reducing variability in selectivity may be attractive by providing a more stable exploitation pattern. There are only a few examples that directly compared the selectivity of different technical options with gear and vessel characteristics maintained over an extended time period and on a range of different populations. Excluding unwanted species from single-species fisheries is an area where gear technology has clearly helped to reduce discarding. It is likely to have already had a significant impact in reducing mortality of charismatic species such as marine turtles as well as fish previously taken as unwanted bycatch. The widespread use of the Nordmøre grid

Technical Measures to Reduce Bycatch and Discards in Trawl Fisheries in the North Atlantic shrimp fisheries as well as SMPs in tropical shrimp and temperate Nephrops fisheries have proved to have significant impacts in reducing discards. However, there are technical challenges in these and other shrimp fisheries. Unwanted bycatch of fish that are not physically excluded through the use of grids still presents problems. The capture of red snapper in the Gulf of Mexico shrimp fishery is still a major challenge, as is the retention of small cod and redfish in the North Atlantic shrimp fishery. Alternative options are needed to reduce discard to acceptable levels. If alterations to gear design and operation are to be used, then it is necessary to utilize differences in species- and/or size-specific behavior to achieve desired reductions. This requires further research into observing and quantifying behavior as well as technical means to exploit these differences. Understanding these behavioral differences is becoming increasingly more important, particularly in multispecies fisheries. In many, one or more species are subject to restrictive catching opportunities or subject to moratorium. To avoid blanket reductions in fishing effort, which would result in underutilization of “healthy” stocks, the challenge facing managers and fishermen is to apply more species-selective gears. It may even be necessary to transform these into single-species fisheries at the expense of the marketable bycatch, if this is economically feasible. Again, this may be challenging from a commercial and managerial perspective in the short term, but such issues need to be viewed in a longer-term context. Societal demands to reduce discarding and other impacts associated with trawling are growing. Consequently, pressure is increasing on policy makers, fishermen and scientists to “do something” about the “discard problem.” To identify what options are most suited and their likely impacts, fisheries should first be evaluated (audited) to identify the specific discard problems and to reference these against the available mitigation tools. In some cases it may not be feasible to reduce discards through modification to trawl design, and other measures need to be applied. Indeed, these options may offer greater levels of protection and may provide the biologically optimal solution. Readers should also be cognizant that discarding of com-

259

mercially exploited species tends to be problematic when stocks are overexploited, and improving stock health through reductions in fishing mortality will go a long way to resolving or mitigating the impacts of discarding at a stock level. Analytical tools are needed to compare and contrast the biological, social, and economic consequences of each measure if we are to define the appropriate course of action to mitigate a particular problem. Trawl modifications are likely to continue offering attractive means to reduce negative impacts over others—ultimately it is the fishing gear that catches the fish—and, once identified as a suitable measure, the challenge is to successfully apply these into commercial fisheries. Fundamental to successful implementation are basic policies that discourage discarding and offer incentives, through tangible benefits, to those that reduce unwanted catches to the point where they do not constitute a threat to the reproductive capacity of the stock and satisfy societal demands. 10.7 CONCLUDING REMARKS Gear-related technical measures have been shown to reduce negative impacts of trawling in a wide range of situations, and there are a number of success stories. To ensure that fisheries head toward sustainable development and respond to increasing societal demands, it is necessary that fishermen, scientists, and mangers continue to develop and apply trawling techniques that are more selective. Central to this is the introduction of management policies that promote and incentivize the use of more selective gears and shift away from regulating fisheries on what they land toward the one that is based on what they catch. REFERENCES Alverson D, Freeberg M, Murawski S and Pope J. 1994. A global assessment of fisheries bycatch and discards. FAO Fish. Tech. Pap. 339: 233 pp. Ambrose EE, Solarin BB, Isebor CE and Williams AB. 2005. Assessment of fish bycatch species from coastal artisanal shrimp beam trawl fisheries in Nigeria. Fish. Res. 71: 125–132. Anon. 2003. Appendix 5 of Report of Expert Meeting on Cod Assessment and Technical Measures, Brussels, April/May 2003. Brussels: DG Fish. European Commission,

260

Contemporary Issues in Capture and Conservation in Marine Fisheries

Anon. 2007. SURVIVAL: An assessment of mortality in fish escaping from trawl codends and it use in fisheries management. EU Contract No Q5RS2002-01603. Anon. 2008. Scientific, Technical and Economic Committee for Fisheries (STECF). Report of the SGMOS-08-01 Working group on the reduction of discarding practices. Graham N and Beare D (eds). Ispra, Italy. 102 pp. Arkley K. 1990. Fishing Trials to Evaluate the Use of Square Mesh Selector Panels Fitted to Nephrops Trawls: MFV ‘ “ HEATHER SPRIG ” ’ (BCK 181): November/ December 1990. Seafish Industry Authority, Hull, England, UK. Seafish Report No.383: 84 pp. Arkley K and Dunlin G. 2002. Improving the selectivity of towed fishing gears: new prawn trawl designs to avoid capture of unwanted bycatch. Seafish Industry Authority, Hull, England, UK. Seafish Report 542. Arkley K, Lart W and Swarbrick J. 1995.The Use of Separator Trawls as a Tool for Improving Selectivity for Cod. Seafish Industry Authority, Hull, England, UK. Seafish Report 441: 84 pp. Bahamona N, Sardà F and Suuronen P. 2006. Improvement of trawl selectivity in the NW Mediterranean demersal fishery by using a 40 mm square mesh codend. Fish. Res. 81: 15–25 Beutel D, Skrobe L, Castro K, Ruhle P Jr, Ruhle P Sr, O’Grady J and Knight J. 2008. Bycatch reduction in the Northeast directed haddock bottom trawl fishery. Fish. Res. 94: 190–198. Brewer D, Heales D, Milton D, Dell Q, Fry G, Venables B and Jones P. 2006. The impact of turtle excluder devices and bycatch reduction devices on diverse tropical marine communities in Australia’s northern prawn trawl fishery. Fish. Res. 81: 176–188 Brewer D, Rawlinson N, Eayrs S and Burridge C. 1998. an assessment of bycatch reduction devices in a tropical Australian prawn trawl fishery. Fish. Res. 36: 195–215 Broadhurst MK. 2000. Modifications to reduce bycatch in prawn trawls: a review and framework for development. Rev. Fish Biol. Fish. 10: 27–60. Broadhurst MK, Millar RB, Kennelly SJ, Macbeth WG, Young DJ and Gray CA. 2004. Selectivity of conventional diamond- and novel square-mesh codends in an Australian estuarine Penaeid-trawl fishery. Fish. Res. 67: 183–194 Campos A, Fonseca P and Erzini K. 2002. Size selectivity of diamond and square mesh cod ends for rose

shrimp (Parapenaus longirostris) and Norway lobster (Nephrops norvegicus) off the Portuguese south coast. Fish. Res. 58: 281–301. Chokesanguan B, Suppachai A, Somboon S, Worawit W and Nguyen L. 2002. Studies on juvenile and trash fish excluder devices (JTEDs) in Vietnam. Bangkok: Southeast Asian Fisheries Development Centre, Training Department. 20 pp. Cook R. 2003. The magnitude and impact of bycatch mortality by fishing gear. In: Responsible Fisheries in the Marine Ecosystem. pp 219–234. Sinclair M and Valdimarsson G (eds). CABI Publishing and FAO. Cotter AJR, Boon TW and Brown CG. 1997. Statistical aspects of trials of a separator trawl using a twin rig trawler. Fish. Res. 29: 25–32. Crean K and Symes D. 1994. The discards problem: towards a European solution. Mar. Pol. 18: 422– 434. Davis FM. 1934. A mesh experiment indicating that small fish escape while the trawl is being towed. J. Cons. Int. Explor. Mer. 90: 22–26, Appendix C. Dunlin G. 1998. Evaluation of bycatch reduction devices in UK Nephrops fisheries. The use of separator trawls in the Irish Sea. Seafish Industry Authority, Hull, England, UK. Seafish Report No SR522. Eayrs S. 2007. A Guide to Bycatch Reduction in Tropical Shrimp-Trawl Fisheries. Revised edition. Rome: FAO. 108 pp. Eayrs S, Hai NP and Ley J. 2007. Assessment of a juvenile and trash excluder device in a Vietnamese shrimp trawl fishery. ICES J. Mar. Sci. 64: 1598: 1602. Eigaard OR and Holst R. 2004. The effective selectivity of a composite gear for industrial fishing: a sorting grid in combination with a square mesh window. Fish. Res. 68: 99–112 Evans SM, Hunter JE, Elizal and Wahju RI. 1994. Composition and fate of the catch and bycatch in the Farne Deep (North Sea) Nephrops fishery. ICES J Mar. Sci. 51: 155–168. Engås A and Foster D. 2002. The response of red snapper (Lutjanus campechanus) and pinfish (Lagodon rhomboides) to inclined water flow. Fish. Res. 58: 315–321 Engås A and Godø OR. 1989. Escape of fish under the fishing line of a Norwegian sampling trawl and its influence on survey results. J. Cons. Int. Explor. Mer. 45: 269–276. Engås A, Isaksen B and Valdemarsen JW. 1989. Escape behavior of fish in codends of trawl. Proc.

Technical Measures to Reduce Bycatch and Discards in Trawl Fisheries Square Mesh Workshop. pp 72–76. St. John’s, Newfoundland: Marine Institute. Engås A, Jørgensen T and West CW. 1998. A species selective trawl for demersal gadoid fisheries. ICES J. Mar. Sci. 55: 835–845. Fonteyne R and M’Rabet R. 1992. Selectivity experiments on sole with diamond and square mesh codends in the Belgian coastal beam trawl fishery. Fish. Res. 13: 221–233. Fryer RJ. 1991. A model of between haul variations in selectivity. ICES J. Mar: Sci: 48: 281–290. Furnell MJ. 1982. Bycatch from shrimp trawling in Guyanese waters. In: Fish bycatch bonus from the sea; report of a technical consultation. pp 55–57. FAO Technical Consultation on Shrimp Bycatch Utilization, Georgetown, Guyana. Glass CW and Wardle CS. 1995 Studies on the use of visual stimuli to control fish escape from codends. II. The effect of a black tunnel on the reaction behavior of fish in otter trawl codends. Fish. Res. 23:165–174. Glass CW, Wardle CS, Gosden SJ and Racey DN. 1995. Studies on the use of visual stimuli to control fishescape from codends. I. Laboratory studies on the effect of a black tunnel on mesh penetration. Fish. Res. 23: 157–164. Graham N and Ferro RST. 2004. The Nephrops fisheries of the NE Atlantic—A review and assessment of fishing gear design. ICES Coop. Res. Rep. 270: 38 pp. Graham N and Fryer RJ. 2006. Separation of fish from Nephrops norvegicus into a two-tier codend using a selection grid. Fish. Res. 82: 111–118 Graham N and Kynoch RJ. 2001. Square mesh panels in demersal trawls: some data on the effect of panel position and mesh size. Fish. Res. 49: 207–218. Graham N, Ferro RST, Karp WA and MacMullen P. 2007. Fishing practice, gear design; and the ecosystem approach—three case studies demonstrating the effect of management strategy on gear selectivity and discards. ICES J. Mar. Sci. 64: 744–750. Graham N, Kynoch RJ and Fryer RJ. 2003. Square mesh panels in demersal trawls: further data on haddock and whiting selectivity in relation to panel position. Fish. Res. 62: 361–375. Hannah RW and Jones SA. 2007. Effectiveness of bycatch reduction devices (BRDs) in the ocean shrimp (Pandalus jordani) trawl fishery. Fish. Res. 85: 217–225 Hansen UJ. 2004. Performance of a trawl codend made from 90° turned netting (T90) compared with that of traditional codends. ICES Fishing Technology

261

and Fish Behavior Working Group Meeting, Gdynia, April 20–23, 2004. Online: http://www.sintef.dk/ docs/T90-FTFB.pdf. Halliday RG and Cooper CG. 2000. Size selection of silver hake (Merluccius bilinearis) by otter trawls with square and diamond mesh codends of 55– 60 mm mesh size. Fish. Res. 49: 77–84. Halliday RG, Cooper CG, Fanning P, Hickey WM and Gagnon P. 1999. Size selection of Atlantic cod, haddock and pollock (saithe) by otter trawls with square and diamond mesh codends of 130–155 mm mesh size. Fish. Res. 41: 255–271. Harrington JM, Myers RA and Rosenberg AA. 2005. Wasted fishery resources: discarded bycatch in the USA. Fish & Fish. 6: 250–361. He P. 2007. Selectivity of large mesh trawl codends in the Gulf of Maine: I. Comparison of square and diamond mesh. Fish. Res. 83: 44–59. He P, Goethel D and Smith T. 2007. Design and test of a topless shrimp trawl to reduce pelagic fish bycatch in the Gulf of Maine pink shrimp fishery. J. Northw. Atl. Fish. Sci. 38: 13–21. Heales DS, Gregor R, Wakeford J and Wang Y–G, Yarrowe J and Milton DA. 2008. Tropical prawn trawl bycatch of fish and seasnakes reduced by Yarrow Fisheye Bycatch Reduction Device. Fish. Res. 89: 76–83 Herrington W. 1935. Modifications in gear to curtail the destruction of undersized fish in otter trawling. Bureau of Fisheries, U.S. Dept. of Commer. Invest. Rep. 24: 48 pp. Herrmann B, Priour D and Krag LA. 2007. Simulationbased study of the combined effect on codend size selection of turning meshes by 90° and reducing the number of meshes in the circumference for round fish. Fish. Res. 84: 222–232. Holst R and Revill A. 2009. A simple statistical method for catch comparison studies. Fish. Res. 95: 254–259. Holt EWL. 1895. An examination of the present state of the Grimsby Trawl Fishery, with special reference to the destruction of immature fish. The Marine Biological Association Plymouth. New Series. Vol III (v). ICES. 1996. Report of the Study Group on Grid (Grate) Sorting Systems in Trawls, Beam Trawls and Seine Nets. ICES C.M. 1996/B:1. ICES. 2005. Report of the ICES-FAO Working Group on Fishing Technology and Fish Behaviour (WGFTFB), 18–22 April 2005, Rome FTC:05, Ref. ACFM. ICES, Copenhagen, Denmark. 283 pp.

262

Contemporary Issues in Capture and Conservation in Marine Fisheries

ICES. 2006a. Report of the ICES Advisory Committee on Fishery Management, Advisory Committee on the Marine Environment and Advisory Committee on Ecosystems, 2006. ICES Advice. Books 1: 10. 6, 310 pp. ICES. 2006b. Report of the ICES-FAO Working Group on Fishing Technology and Fish Behaviour (WGFTFB). ICES C.M. 2006/FTC:06, Ref. ACFM. ICES, Copenhagen, Denmark, 180 pp. Ingólfsson OA and Jørgensen T. 2006. Escapement of gadoid fish beneath a commercial bottom trawl: Relevance to the overall trawl selectivity. Fish. Res. 79: 303–312. Isaksen B and Valdemarsen JW. 1988. Selectivity experiments with square mesh codends in bottom trawls 1985–1987. Proc. Square Mesh Workshop. St. John’s Newfoundland, Canada: Marine Institute. Isaksen B, Valdemarsen JW, Larsen RB and Karlsen L. 1992. Reduction of fish bycatch in shrimp trawl using a separator grid in the aft belly. Fish. Res. 13: 335–352. Jørgensen T, Ingólfsson OA, Graham N and Isaksen B. 2006. Size selection of cod by rigid grids—Is anything gained compared with diamond mesh codends only? Fish. Res. 79: 337–348 Kelleher K. 2005. Discards in the world’s marine fisheries: an update. FAO Fish. Tech. Pap. 470: 131 pp. Kvalsvik K, Huse I, Misund OA and Gamst K. 2006. Grid selection in the North Sea industrial trawl fishery for Norway pout: Efficient size selection reduces bycatch. Fish Res. 77: 248–263 Kvamme C and Frøysa KG. 2004. Assessing the effects on stocks of selectivity changes in a fishery. Fish. Res. 69: 283–292. Kvamme C and Isaksen B. 2004. Total selectivity of a commercial cod trawl with and without a grid mounted: grid and codend selectivity of north-east Arctic cod. Fish. Res. 68: 305–318. MacLennan DN. 1992. Fishing gear selectivity: an overview. Fish. Res. 13: 201–204. Madsen N, Moth-Poulsen T and Lowry L. 1998. Selectivity experiments with window codends fished in the Baltic Sea cod (Gadus morhua) fishery. Fish. Res. 36: 1–14 Main J and Sangster GI. 1981. A study of the fish capture process in a bottom trawl by direct observation from a towed underwater vehicle. Scot. Fish. Res. Rep. 23: 24 pp. Main J and Sangster GI. 1985. Trawling experiments with a two-level net to minimise the undersized gadoid bycatch in a Nephrops fishery. Fish. Res. 3: 131–145.

Nunoo FKE, Boateng JO, Ahulu AM, Agyekum KA and Sumaila UR. 2009. When trash fish is treasure: The case of Ghana in West Africa. Fish. Res. 96: 167–17. Özbilgin H. 1998. The Seasonal Variation of Trawl Codend Selectivity and the Role of Learning in Mesh Penetration Behavior of Fish. PhD Thesis, Department of Zoology, University of Aberdeen. Özbilgin H and Wardle CS. 2002. Effect of seasonal temperature changes on the escape behavior of haddock, Melanogrammus aeglefinus, from the codend. Fish. Res. 58: 323–331 O’Neill FG and Kynoch RJ. 1996. The effect of cover mesh size and codend catch size on codend selectivity. Fish. Res. 28: 291–303 O’Neill FG, Kynoch RJ and Fryer RJ. 2006. Square mesh panels in North Sea demersal trawls: Separate estimates of panel and codend selectivity. Fish. Res. 78: 333–341. O’Neill FG, McKay SJ, Ward JN, Strickland A, Kynoch RJ and Zuur AF. 2003. An investigation of the relationship between sea state induced vessel motion and codend selection. Fish. Res. 60: 107–130. Perez-Mellado J, Romero JM and Young RH.1982. Yields and composition of (shrimp) bycatch from the Gulf of California (Northeastern Pacific). In: Fish Bycatch Bonus from the Sea; Report of a Technical Consultation; pp. 55–57. FAO Technical Consultation on Shrimp Bycatch Utilization, Georgetown, Guyana. Priestley R, Wardle CS and Hall CD. 1985. The Marine Laboratory remote controlled fishing gear observation vehicle. ICES CM. 1985/B:10. Revill A, Dunlin G and Holst R. 2006. Selective properties of the cutaway trawl and several other commercial trawls used in the Farne deeps North Sea Nephrops fishery. Fish. Res. 81: 268–275. Revill AS and Jennings S. 2005. The capacity of benthos release panels to reduce the impacts of beam trawls on benthic communities. Fish. Res. 75:73–85 Revill AS, Pascoe S, Radcliffe C, Riemann S, Redant F, Polet H, Damm U, Neudecker T, Kristensen PS and Jensen D. 1999. The economic consequences of discarding in the Crangon fisheries. European Commission Study Contract Report No.97/SE/025. Richards A and Hendrickson L. 2006. Effectiveness of the Nordmøre grate in the Gulf of Maine Northern shrimp fishery. Fish. Res. 81: 100–106. Robertson JHB and Polanski J. 1984. Measurement of the breaking strength of square and diamond mesh netting. Scot. Fish. Work. Pap. 09/84: 3 pp.

Technical Measures to Reduce Bycatch and Discards in Trawl Fisheries Robertson JHB and Stewart PAM. 1988. A comparison of size selection of haddock and whiting by square and diamond mesh codends. J. Cons. Int. Expl. Mer. 44: 148–161. Saila S. 1983. Importance and assessment of discards in commercial fisheries. FAO Fish. Circ. 765: 62 pp. Sala A, Lucchetti A, Piccinetti C and Ferretti M. 2008. Size selection by diamond- and square-mesh codends in multi-species Mediterranean demersal trawl fisheries. Fish. Res. 93: 8–21. Sardà F, Bahamon N, Molì B and Sardà-Palomera F. 2006. The use of a square mesh codend and sorting grids to reduce catches of young fish and improve sustainability in a multispecies bottom trawl fishery in the Mediterranean. Sci. Mar. 70: 347–353. Shepherd JG. 1990. Stability and the objectives of fisheries management: the scientific background. Lab. Leafl., MAFF Direct. Fish. Res. Lowestoft 64: 16 pp. Stratoudakis Y, Fryer RJ, Cook RM, Pierce GJ and Coull KA. 2001. Fish bycatch and discarding in Nephrops trawlers in the Firth of Clyde (west of Scotland). Aquat. Liv. Resour. 14: 283–291. Suuronen P. 2005. Mortality of fish escaping trawl gears. FAO Fish. Tech. Pap. 478: 87 pp. Thomsen B. 1993. Selective flatfish trawling. ICES Mar. Sci. Symp. 196: 161–164. Tonks ML, Griffiths SP, Heales DS, Brewer DT and Dell Q. 2008. Species composition and temporal variation of prawn trawl bycatch in the Joseph Bonaparte Gulf, northwestern Australia. Fish. Res. 89: 276–293. Urquhart GG and Stewart PAM. 1993. A review of techniques for the observation of fish behavior in the sea. ICES Mar. Sci. Symp. 196: 135–139.

263

Walsh SJ, Engås A, Ferro RST, Fonteyne R and van Marlen B. 2000. Improving Fishing Technology to Catch (or Conserve) More Fish: The Evolution of the ICES Fishing Technology and Fish Behavior Working Group During the Past Century. 19 pp. Online: http://www.wgftfb.org/publications/history %20html%20files/history.pdf Walsh SJ, Millar RB, Cooper CG and Hickey WM. 1992. Codend selection in American plaice: diamond versus square mesh. Fish. Res. 13: 235–254. Wardle CS. 1986. Fish behavior in fishing gear. In: The Behavior of Teleost Fishes. Pitcher TJ (ed). pp 463–495. London and Sydney: Croom Helm. Watson JW and Seidel WR. 1980. Evaluation of techniques to decrease sea turtle mortalities in the southeastern United States shrimp fishery. ICES CM. 1980/B: 31. Watson JW and Taylor CW. 1988. Research on selective shrimp trawl designs for Penaeid shrimp in the United States: a review of selective shrimp trawl research in the United States since 1973. Proc. FAO Expert Consultation on Selective Shrimp Trawl Development, Mazatlan, Mexico, November 24–28, 1986. Rome: FAO. Westlund L. 2006. Environmental Impact from Tropical Shrimp Trawling through the Introduction of Bycatch Reduction Technologies and Change of Management. United Nations Environmental Program. Online: ftp://ftp.fao.org/FI/DOCUMENT/ rebyc/FinalMid-TermEvaluationReport.pdf Wileman DA, Ferro RST, Fonteyne R and Millar RB. 1996. Manual of methods of measuring the selectivity of towed fishing gears. ICES Coop. Res. Rep. 215: 126 pp.

264

Contemporary Issues in Capture and Conservation in Marine Fisheries

SPECIES MENTIONED IN THE TEXT Atlantic cod, cod, Gadus morhua dab, Limanda limanda estuarine prawn, Nematopalaemon hastatus haddock, Melanogrammus aeglefinus pink shrimp, Pandalus borealis

Norway pout, Trisopterus esmarki ocean shrimp, Pandalus jordani red snapper, Lutjanus campechanus silver hake, Merluccius bilinearis yellowtail flounder, Limanda ferruginea

Chapter 11 Mortality of Animals that Escape Fishing Gears or Are Discarded after Capture: Approaches to Reduce Mortality Petri Suuronen and Daniel L. Erickson

11.1 INTRODUCTION Fishing mortality can be defined as the sum of all fishing-induced mortalities occurring directly as a result of capture or indirectly as a result of contact with or avoidance of the fishing gear (Broadhurst et al. 2006; Chopin et al. 1996). Fishing mortality can be attributed to landed catch; illegal, misreported, and unreported landings; discard; escape; drop-out; ghost fishing; avoidance; and habitat degradation. Except for landed catch, these subcomponents of fishing mortality are unknown or poorly understood for most fish stocks and are often referred to as “unaccounted mortalities.” A significant amount of research would be necessary to estimate all the subcomponents of fishing mortality for any given stock, and such information is available for only a few stocks, if any. Most of the work conducted so far in this field has been limited to two particular aspects of mortality—discard and escapement. These are probably the most important components of unaccounted fishing mortality, particularly for trawl fisheries (Suuronen 2005). Discarding is a widely applied practice in commercial fisheries (Kelleher 2005). Many of the world’s discarded fish are small juveniles of commercially important species, which if left to grow

to mature size would produce significant yields. Other discarded species include relatively rare or endangered animals (e.g., sturgeons) or abundant but unmarketable animals (e.g., sharks, skates, and rays in North Pacific longline fisheries). Studies on the mortality of fish discarded from the decks of fishing vessels generally show high mortality rates, although the types of injuries and their severity are species specific (Broadhurst et al. 2006). Most fish with gas bladders that inflate after capture die because of pressure changes, but there are exceptions (e.g., Parker et al. 2006). The postrelease mortality of other fish and aquatic organisms (i.e., those without gas bladders) is variable; their mortality may be low or high, depending among others on fishing conditions and on-deck handling. Mortality is also related to the overall fragility and physical characteristics of species. Bird, mammal, and turtle mortality is typically high after capture by most commercial fishing operations (e.g., trawl, gillnet, and longline) but can be markedly reduced by various mitigation measures (Valdemarsen and Suuronen 2003; also see Chapter 13). Results of many postselection studies suggest that mortality can be substantially reduced if animals escape from fishing gears before gear

265

266

Contemporary Issues in Capture and Conservation in Marine Fisheries

retrieval and landing on the deck (Broadhurst et al. 2006). However, it is also obvious that not all fish survive this escapement process. In many cases, escape from fishing gears occurs after the animals have been subjected to a wide variety of capture stressors and possible damage through contact with fish, debris, or the gear itself (Chopin and Arimoto 1995; Suuronen 2005). In addition, even though animals may not die immediately after encountering and escaping fishing gears, they may experience impaired predator-avoidance capacity or may exhibit delayed mortality (Ryer 2002). The specific reasons why some fish ultimately die are still poorly understood. There is a substantial need to improve the understanding of this mortality, identify its most likely sources, and assess its magnitude and impact on the stocks and management of fisheries. The fate of escaping fish is becoming an increasingly important issue because of a recent strong tendency among fisheries management authorities to increase minimum mesh sizes and to use various other controls that improve size or species selection (e.g., Breen and Cook 2002; Halliday and Pinhorn 2002). If unaccounted mortality is high, then the assumed benefits of changing selectivity may be largely overestimated (Suuronen 2005). For many important fish species, there are insufficient estimates of escape survival to make an adequate assessment of its impacts on stocks and fisheries. Failure to quantify the biological impacts of this largely unknown mortality could result in biases in fisheries management decision-making processes. This chapter describes the mortality of animals associated with commercial fishing and assesses principal factors affecting stress, injury, and mortality of fish that arise from capture and discarding. Potential sources of error in the assessment of mortality are examined and identified, and appropriate methodological approaches are suggested and discussed. This chapter also examines how mortality can be decreased through modifications to fishing gear and their operations. Because major bycatch problems are associated with towed fishing gears such as bottom trawls, the focus of this chapter is largely confined to these gears. Some components of other fisheries are addressed when information is available.

11.2 MORTALITY OF DISCARDS AND ESCAPEES 11.2.1 Discard Mortality It is clear that not all of the discarded fish and other organisms survive (Broadhurst et al. 2006; Davis 2002; Hill and Wassenberg 1990; ICES 2000). Discard mortality represents a large source of uncertainty for estimates of total fishing mortality and is an important issue for fisheries management worldwide. Major biological and economic losses occur as a result of this discarded bycatch (i.e., organisms that are inadvertently caught by fishing gears while targeting other species or size-groups and are subsequently discarded). Most of the work conducted on discard mortality has concentrated on the survival of fish and other organisms caught by trawls (otter, beam, and shrimp trawls). All major fishing gear types, however, involve some degree of injury to fish through internal and external wounding, crushing, scale loss, oxygen deprivation/ debt, or hydrostatic effects, with the severity of the injury depending on the gear type and its operation (Chopin and Arimoto 1995). These other gear types include gillnet, seine, hook and line, and pots. The types and extent of injuries caused by discarding are species specific (Broadhurst et al. 2006). Some species are highly sensitive to capture and discarding, while others are capable of surviving these traumas. Fish with gas bladders and other organs that inflate during the hauling process because of pressure changes may become “trapped” near the surface after discarding and may suffer near 100% mortality (due to pressure-related injuries, predation by birds, and inability to return to depth). Nevertheless, some species with gas bladders, such as black rockfish and China rockfish, may survive the capture and discarding process (Parker et al. 2006). For fish that do not have gas bladders (e.g., sablefish, lingcod, flatfishes), mortality after release is more variable (e.g., Davis 2002). Several species of flatfish, for example, appear to have relatively good chances of survival under certain conditions (Kelle 1976; Neilson et al. 1989; Pikitch et al. 1996; Van Beek et al. 1989). There are circumstances, however, where flatfish exhibit relatively high mortalities (e.g., Lindeboom and de Groot 1998; Robinson et al. 1993). Other reasons

Mortality of Animals that Escape Fishing Gears or Are Discarded after Capture for differential mortality both within and between species include differences in scale loss and skin damage, differences in susceptibility to changes in temperature (both in the water during haul-back and air temperature on deck), and body size (smaller fish are weaker and more susceptible to injuries and exhaustion than larger fish; Kelle 1976; Pikitch et al. 1996). Discard Mortality in Trawl Fisheries. Trawlcaught fish and other organisms often experience high discard mortality (e.g., Harris and Ulmestrand 2004; Hill and Wassenberg 1990; Steven, 1990; Wileman et al. 1999). Fragile species, in particular, may experience substantially higher mortality in trawls when the catch consists of a mixture of other species with hard parts, spines, shells, and carapices than when only soft-bodied individuals are present. Depending on environmental conditions, fishing parameters, and deck-handling methods, discard mortality for fishes without gas bladders (e.g., Pacific halibut and sablefish) can be high or low (e.g., Erickson and Pikitch 1999; Pikitch et al. 1996). Discard mortality of a trawled elasmobranch, the spiny dogfish, was well below 50% (Mandelman and Farrington 2007). It is notable that flatfish may be more sensitive than round fish (e.g., sablefish and lingcod) to suffocation in trawl nets from pressure on the operculum (Davis 2002). Many invertebrates have durable shells or exoskeletons (except cephalopods) and therefore are less likely to be damaged during gear contacts and handling than fish. Hill and Wassenberg (1990) found about half of the crustaceans discarded in the prawn trawl fishery survived. All crabs survived a 12-h monitoring period. Lancaster and Frid (2002) estimated that about 80% of undersized brown shrimp survive the capture and discarding processes in the U.K. Solway shrimp fisheries. The survival of trawl-caught Patagonian scallops was also high even after being subjected to 30 min of on-deck exposure (Bremec et al. 2004). Sea urchins caught by beam trawl (Kaiser and Spencer 1995) and cephalopods caught in pelagic trawls (Hill and Wassenberg 1990) exhibited high discard mortality. Discard Mortality in Gillnet Fisheries. Generally, fish released from gillnet may suffer severe skin injury and stress (e.g., Chopin et al. 1996; Thompson et al. 1971; Thompson and Hunter 1973). As was

267

shown for discard mortality for trawl-caught species, postrelease mortality caused by gillnet injuries is species and fishery dependent. The material of the net (Vander Haegen et al. 2004), water temperature, soak duration, and fish size are factors that are likely to affect the postrelease mortality for gillnet-caught fish. Discard Mortality in Seine Fisheries. Seine-nets (surrounding nets) have a high potential for catching fish alive and uninjured, so in many cases, unwanted bycatch organisms can be released from this gear type with a good chance for surviving (e.g., Suuronen et al. 1995). Nevertheless, relatively little scientific work has been conducted on the survival of fish discarded from seine operations. Erickson et al. (1999) demonstrated in some cases near 100% survival for Alaskan pollock caught by seine and held in sea-bed cages for 14 days. Hence, seine nets not only may be useful for promoting high survival for discarded species but also may offer the possibility for a live capture that can provide a higher-quality and higher-value catch. Mortality of discarded seine-caught fish is most likely if crowding occurs during the final stages of hauling the seine. Lockwood et al. (1983) investigated the effect of crowding on mackerel during purse seining and found that mortality was directly related to fish density and duration of time that fish spent in the net. Erickson et al. (1999) also showed high mortality for Alaskan pollock when crowded in the purse seine. Discard Mortality in Hook-and-Line Fisheries. Fish released from hook-and-line gears (including longlines) may suffer a variety of injuries, stresses, and mortalities depending on species, handling methods, and gear. Generally, it appears that hook penetration depth, hooking location, and the technique used to remove fish from the hook have major impacts on subsequent survival of discarded fish (Bartholomew and Bohnsack 2005). A swallowed hook may induce substantially greater injury than a hooked mouth (e.g., through the jaw, lips, or operculum). Hook design may affect injury and survival; circle hooks were found to be far less likely than the traditional J-hooks to cause serious bleeding or become lodged in areas other than the mouth of striped marlin (Domeier et al. 2003). Fish removed from hooks by a crucifier may experience

268

Contemporary Issues in Capture and Conservation in Marine Fisheries

a significantly higher mortality than fish removed manually (Kaimmer 1994; Milliken et al. 1999). Huse and Soldal (2002) showed that the mortality of undersized haddock that were gaffed and pulled off the hook exhibited markedly higher mortality (53%) than those that were torn off by means of a crucifier (39%). They demonstrated that both release methods inflicted severe injuries to the mouth parts of the fish. Fish that were released by gaffing also suffered from punctures to the body wall and damage to the abdomen and intestines. Note that gaffs can be used to remove the hook without handling or puncturing the fish (Kaimmer 1994). Discard Mortality in Pot and Trap Fisheries. Fishing with pots and traps generally results in catches that are alive and uninjured, so in most cases unwanted bycatch organisms can be released with a good chance of survival, although factors such as decompression or thermal shock (both in water and on deck), on-deck injury, and air exposure on deck may jeopardize the survival of released organisms. There have been few studies designed to estimate the survival of trap-caught and released animals. An investigation by Siira et al. (2006) showed an average of 7% mortality among adult Atlantic salmon released from large floating trapnets moored along the northern Baltic coast. There are cases, however, where discard mortality can be high, as shown by Stevens (1990) for king crab (Paralithodes camtschaticus) and tanner crab (Chionoecetes bairdi) that were landed in subzero temperatures on frozen decks in the North Pacific Ocean. This is an example of extreme temperature shock that can severely impact the survival of pot-caught organisms. Temperature shock during warmest months of the year may also result in high mortality for pot-caught specimens. Note that temperature shock can be caused by changes in temperature during both ascents and descents through the water column (i.e., within the water) as well during transfer from water to the deck and back to the water again. Scavenger and Predation Mortality. Two sources of potential discard mortality that are rarely mentioned are “scavengers” and predators injuring animals confined in pots or on longlines or animals stressed after being returned to the sea. Sand fleas

(amphipods) may “attack” fish that are recovering after discard from fishing vessels (Pikitch et al. 1996). These scavengers may also attack seemingly healthy fish that are confined within pots or on longlines, especially gears that prevent fish from swimming off the bottom (Trumble et al. 2000). In addition, large predators such as sharks may bite fish that are caught and restricted by longlines and gill nets. Erickson and Berkeley (2008) described high numbers of shark-bitten swordfish that were caught on pelagic longlines. Benthic sharks (e.g., sleeper sharks) also attack longline-caught halibut and sablefish in the North Pacific, taking bites out of the belly of large halibut, or leaving only the heads of smaller fish (e.g., sablefish) intact on the hooks (D. Erickson, unpublished data). Ryer et al. (2004) described the increased likelihood of stressed fish encountering predators due to injuries or changes in behavior; increased predation mortality on recently discarded fish is therefore very probable. 11.2.2 Mortality of Fish that Encounter but Escape Fishing Gear Most assessments of escape mortality have been made for commercially important species escaping from towed fishing gears, mainly trawls. The most studied group of fish is the gadoids, particularly haddock, whiting, and cod. Some work has also been conducted with pelagic species, flatfishes, as well as red mullet, sand whiting, and yellowfin bream (reviewed in Broadhurst et al. 2006; ICES 2000; Suuronen 2005). It is clear from these studies that escape mortality varies markedly with robustness and ability of various species to withstand physical injury and fatigue associated with capture and escape. Generally, high escape survival has been observed in many gadoids such as cod and saithe (Ingolfsson et al. 2002; Soldal et al. 1993, Suuronen et al. 1996a), with the cod being one of the most robust of all species. The mortality of haddock and whiting has been more variable but often less than 25% (Breen et al. 2007; Sangster et al. 1996; Wileman et al. 1999). Low mortality rates (mostly less than 10%) have also been observed with flatfishes such as winter flounder, yellowtail flounder, and American plaice (DeAlteris and Reifsteck

Mortality of Animals that Escape Fishing Gears or Are Discarded after Capture 1993; Robinson et al. 1993). Broadhurst et al. (1997, 1999) showed low mortality (less than 3%) associated with sand whiting and yellowfin bream that had passed through the square mesh panels of a trawl codend. It is notable, however, that some of these species have also shown relatively high mortalities. Medium levels of mortalities (approximately 50%) have been recorded with species such as vendace and Alaskan pollock (Pikitch et al. 2002; Suuronen et al. 1995). Large escape mortalities (50% to 90%) have been observed in small Baltic herring (Suuronen et al. 1996b, 1996c), although almost opposite results have been reported with the same species (Efanov 1981; Treschev et al. 1975). In general, there is a substantial amount of uncertainty associated in the mortality estimates made with pelagic fish species because the assessment of fragile and small fish is prone to severe experimental flaws. The general observation has been that escape survival increases with increasing fish length, regardless of fish species. For instance, Wileman et al. (1999) observed that the mean length of haddock that died after escape was significantly lower than that of surviving haddock—that is, survival increased with increasing mean length. Breen et al. (2007) observed that the probability of survival of escaping haddock and whiting was lowest among fish of 15 cm in length and less. Similar lengthrelated mortality has been described in many other studies conducted on gadoids (Pikitch et al. 2002; Sangster et al. 1996), and species such as vendace (Suuronen et al. 1995) and Baltic herring (Suuronen et al. 1996b, 1996c). Relatively high mortalities (greater than 40%) were measured for mackerel that escaped through sorting grids attached to purse seines (Beltestad and Misund 1996; Misund and Beltestad 2000). These authors concluded that fish passing through the sorting grids suffered severe stress and skin injuries. Misund and Beltestad (1995) investigated the survival of Atlantic herring after simulated purse seine bursts (fish school escapes through the bursting net during the final pursing phase). Their results indicated that there may be high mortality among herring connected to the net burst, which caused severe scale loss. Likewise, Pacific herring passing through a 57-mm monofilament gillnet had severe

269

scale loss of up to 40% and suffered delayed mortality (Hay et al. 1986). In conclusion, clear gear-specific and speciesspecific differences in escape survival have been observed. Larger fish generally appear to suffer less injury and lower mortality than smaller fish, in particular among trawl fisheries. One of the main findings of these investigations is the high variability in survival, even with the same species in the same experiments. This variation has not yet been explained adequately.

11.3 ASSESSMENT OF MORTALITY Measuring the mortality of fish discarded or escaping from fishing gears under various fishing conditions is subject to high variability and methodological flaws (Breen et al. 2007; Suuronen 2005). Until recently, experimental methodology applied in the assessment of escape mortality has been in its infancy. Survival figures have been affected by inferior methods of collecting, transporting, and monitoring escapees, and no investigation has been conducted without methodological compromises. Hence, the results of survival experiments to date should be considered with great caution, and new methods should be developed that reduce or eliminate postescape handling stress (e.g., caging, transporting, and handling). In most discard-mortality studies, organisms have been caught, retrieved to a vessel, handled according to conventional practices and then monitored in tanks at the surface (or onboard fishing vessel), or in sea cages at appropriate depths. A universal problem with holding discards in tanks at surface is that the critical parameters like temperature, pressure, light, food, and stocking densities are not similar to those at the discarded organisms’ normal habitats. High-tech tagging methods are now available that, with some creativity, can and should be modified and applied to studies for evaluating survival of animals after release into their natural environment. This section focuses on potential sources of error in the assessment of escape mortality associated with trawl fishery and suggests some new methodological approaches. Laboratory methods for assessing various types of capture-induced injury

270

Contemporary Issues in Capture and Conservation in Marine Fisheries

and physiological stress are beyond the scope of this discussion. 11.3.1 Assessment of Mortality of Fish Escaping Trawl Codend Typically, mortalities of trawl codend escapees have been assessed in relatively shallow waters and under relatively short trawl tows and low catch rates. Experiments have not fully simulated commercial fishing conditions in terms of tow duration, depth, catch size, and season (Suuronen 2005). Hence, these experiments may not reflect the full range of possible sources of injury and mortality encountered by the escaping fish under commercial fishing conditions. Moreover, the methods used in past survival experiments may have seriously biased mortality estimates. For most trawl-escape mortality studies, researchers have relied on “trapping” fish that pass through trawl meshes (= escapees) inside of codend covers or cages attached to the codend covers (Fig. 11.1). These escapees are subsequently held for observation on the seabed or near the surface in the cover or cages to observe mortalities over some period of time (Fig. 11.2). Bias Caused by Collection of Escapees. The sampling of escapees in the collection cover may cause them substantial damage and stress thereby resulting in overestimates of escape mortality (Breen et al. 2002, 2007; Suuronen et al. 1996c). The high mortality seen in some investigations, particularly among the smallest escapees (Suuronen et al. 1996c), could be partly the result of cover-induced mortality. Small fish may become too weak and exhausted by the capture process to have enough energy left to swim within the cover; the water flow then forces them against the cover wall. The use of a cover, however, may also result in underestimates of mortality. Breen et al. (2002) demonstrated that the water flow around a codend with a typical cover was significantly lower than that observed around a normal uncovered codend. By reducing the flow through and around the codend, the cover may cause reduced tension on the netting of the codend (i.e., the meshes remain loose or open). As a result, the passage of a fish through the meshes of a codend may be easier and perhaps cause less injury than under normal circumstances. This passage may also be less hazardous because

water flow outside of the codend will be reduced by the presence of the cover. The presence of the cover around the trawl codend may protect escaping fish from the injurious forces that are normally experienced during passage through a trawl codend. Collection of escapees should be done without causing any extra stress and injury to fish. Until the early 1990s, the collection cover was closed at the beginning of the tow and released at the end, when the codend was at the surface. Therefore, escaping fish were sampled throughout the entire haul. To reduce the potential adverse effect of the cover on escapees, the haul duration was limited to periods that were much shorter than those used in normal commercial practice (e.g., Suuronen et al. 1996c). Clearly, a sampling method was needed in which the duration of the tow was not restricted by the method used to collect escapees. Soldal et al. (1993) used acoustic, and Suuronen et al. (1996a, 1996b), mechanical, releasers to close and release the cages from the cover remotely (at the depth of capture). To minimize injury from cover exposure and simultaneously permit realistic tow durations, later designs allowed the sampling of escaping fish at any moment of a haul (e.g., Breen et al. 2007; Erickson et al. 1999; Lehtonen et al. 1998; Pikitch et al. 2002). With the advent of these techniques, sampling could be conducted at any moment of a haul, and the sampling period (= escapee-collection period during the tow) could be controlled precisely and kept substantially shorter than in previous experiments. Hence, the sampling duration was not dependent on the tow duration. Using these “new” methods, survival can be assessed for short and long tows, and for small and large catch quantities. Cover exposure time (i.e., sampling time) can be kept short enough to avoid cover-induced injury but long enough to provide adequate numbers of escapees. Improvements to the collection method developed by Lehtonen et al. (1998) were made by Erickson et al. (1999) and Pikitch et al. (2002) by using underwater video cameras to monitor and control specimen collection in real time. By observing fish within the cover and the collection cage (Fig. 11.3), the process of specimen collection became a calculated process rather than a series of

Figure 11.1. For most trawl-escape mortality studies fish that pass through trawl meshes (i.e., escapees) are caught inside of codend covers or cages attached to the codend covers.

271

Figure 11.2. These “escaping” fish are subsequently held for observation on the seabed or near the surface in the cover or cages to observe mortalities over some period of time. Control and experimental cages are anchored at close proximity with suitable number of replicates. Divers inspect the condition of cage and fish, remove dead animals, and feed the fish if they are held for a prolonged period of time.

Figure 11.3. Improvements made by Erickson et al. (1999) to the collection method of Lehtonen et al. (1998) by using underwater video cameras to monitor and control sample collection in real time. All escapees were allowed to pass through the codend cover and the collection cage and into the open ocean until an adequate catch had accumulated in the codend. This technology allows collection of a sample of escapees at any point in time and for any length of time during a commercial tow. See text for further details.

272

Mortality of Animals that Escape Fishing Gears or Are Discarded after Capture guesses. All escapees were allowed to pass through the codend cover, through the collection cage, and into the open ocean until an adequate catch had accumulated in the codend. Collection of escapees commenced by remotely closing the rear door of the collection cage that was attached to the codend cover. When there were adequate numbers of escapees in the cage, it was remotely closed and released from the trawl. Hence, the sampling period was restricted to the time that was required to collect the sample. With such a technique, it became possible to collect a statistically adequate number of escapees in each cage and the number of “invalid” tows would be close to zero. Even with the improvements designed to better reflect escapement under “commercial” conditions, use of collection covers and cages do not simulate natural conditions for escaping fish and still may cause skin injury and other damage to escapees. If these types of methods continue, then the cover design should also ensure that the water flow inside the cover (i.e., outside the codend) closely simulates that observed during commercial tows. Water flow patterns can be affected by constructional details of the cover (e.g., Ingolfsson et al. 2002). We are uncertain, however, whether the bias of cover method can ever be fully corrected. To minimize this bias, the design, twine thickness. and the mesh size of the cover, as well as the towing speed, should be carefully considered for each experiment. In addition, codend covers used to herd or collect escaping fish may cause bias due to the visual contrast between the cover netting and the surrounding water. Generally, the visual contrast of the cover should be minimized to obtain escape behavior that is reflective of commercial conditions. It is largely unknown how long individual fish swim in trawls and when they escape. Some may swim within the gear throughout the tow duration while others may escape immediately after capture (e.g., Suuronen et al. 1995). Fish escaping after a long period of swimming may be more exhausted than those that had swum for a short period of time. Therefore, it is important to assess and understand when and where during the tow the escapees should be sampled. Technology is now available to collect a sample of escapees from any point and for any length of time during a commercial tow (Suuronen

273

2005). Because mortality may vary depending on where and when fish escape, the cover must be large enough in all dimensions to collect escapees from all parts of the codend (i.e., to cover the whole selective area). Erickson et al. (1999) and Pikitch et al. (2002) demonstrated that for pelagic trawls, Alaskan pollock either actively escaped far ahead of the catch bulge before being impinged against the netting or became impinged against the netting or within the catch bulge and subsequently made their way out of the codend. Fish that escape from the front part of the codend may show different mortality from those escaping from the aft part of the codend. Bias Caused by Transporting and Holding Escapees. Once the specimens have been collected in the collection cage, they may be transferred to a sheltered and stable environment where they can be held without great risk and monitored for some period of time to assess immediate and delayed mortalities. Transferring fish to monitoring areas should involve minimum environmental changes and stress for the escapees. Abrupt changes in water temperature, hydrostatic pressure, salinity, water flow, or ambient light level may cause stress and affect survival of fish. Transportation should take place at a low speed to avoid excessive water flow, and fish should be monitored throughout the transfer process. To overcome some common problems of underwater fish transport, the collection cage may be placed in a protective container during the towing process (Lowry et al. 1996). However, this method has several practical difficulties and can be very labor intensive. Erickson et al. (1999) attempted to minimize this potential bias by ensuring that cages remained continuously deeper than 10 m, monitoring water temperatures at both capture and holding locations using temperature loggers, and continuously observing fish behavior within cages during the transport process using video surveillance. Suuronen et al. (2005) did not transfer the cages at all; they instead released the cages at the site of monitoring to avoid bias associated with transportation. The conditions where the fish are monitored should mimic as closely as possible those experienced by that species in its normal life. Mortality has generally been shown to peak in the first 2 to 3

274

Contemporary Issues in Capture and Conservation in Marine Fisheries

days after escape, the highest mortality usually occurring during the first day (Suuronen 2005). Postescape mortality rates generally decline with time, usually reaching a minimum after 1 or 2 weeks. Hence, mortality assessments over only a few hours may not be adequate for measuring shortterm capture and escape-induced mortality. On the other hand, observation periods longer than 1 to 2 weeks may not be useful because of secondary infections and the stress connected with captivity. Interpretation of results may become more difficult for longer period experiments. Delayed deaths in monitoring cages are often correlated to the onset of various skin infections and other problems (e.g., deteriorated caudal fins, lesions, and sores). These types of secondary infections have been described for many species (Erickson et al. 1999; Main and Sangster 1990; Soldal et al. 1993; Suuronen et al. 1996c). In fact, for many fish species, cage methodology may be suitable for studying only primary mortality. Assessing long-term mortality caused by secondary infections and predation may require other techniques. It is noteworthy that under typical commercial fishing operations, escapees are free to swim into their normal environment to recover. Enclosure in a cage, often with other damaged fish and in nonfavorable environmental conditions, is likely to cause additional stress that may further contribute to mortality, possibly also through cross-infection. Hence, it might be argued that the recovery of a damaged fish is hindered by captivity (for further discussion, see Main and Sangster 1990, 1991). However, it could also be argued that fish within the cage are protected from predators and other potential hazards. Therefore, great caution is needed when interpreting results of caging and confinement studies. Control Samples. If a fish dies during the monitoring period, there is often uncertainty about the cause of death. The potential effect of captivity on escapees during the monitoring period can be assessed through the use of adequately captured and held controls. That is, a representative group of fish of the same species and size are held in captivity in similar cages and in the same area as the escapees, and their survival is assessed in the same way. If captivity has no lethal effect on the captive fish

(over a specific period), there should be no observed mortality in the control group. When there is mortality among controls, it is important to know why they died. Death may be due to captivity stress, but it may also at least partly be due to stress and injury from the capture of control fish. This must be known when interpreting the mortality for escapees held in cages. Usually, control fish have not experienced all of the aspects of the experiments, such as capture and confinement in the cover and collection cage and transfer to the cage site. Furthermore, in a survival experiment there is always a possibility of severe cumulative effects that are difficult to detect and measure, even when there is an effective arrangement of controls. The lack of adequate controls has made the results of many survival studies practically useless. Usefulness of Laboratory Experiments. Laboratory experiments are useful to help understand key stressors and cumulative effects of the capture and escape process. However, it is difficult, or perhaps impossible, to simulate all potential capture stressors in the laboratory. Fish that are held in laboratory conditions experience sensorydeprived environments, which can result in behavior and stress responses that do not mimic normal responses in the field. Hence, the results of laboratory studies cannot be used to make direct conclusions about the survival of fish in commercial fishing conditions. Nevertheless, laboratory experiments can be cost-effective in investigating stress responses and assessing injuries and for corroborating or predicting results of field investigations. They allow systematic determination of the general behavioral, biological, and physiological principles of stress response up to mortality in different species, and this is rarely possible in field conditions (e.g., Davis 2002). Researchers can focus on the mechanisms of interest and control all others, and cause-and-effect relationships can be established. So far, most laboratory studies have been directed toward assessing the potential injury to fish passing through netting meshes or other selective devices, as well as assessing exhaustion caused by forced swimming inside towed gears. The results of laboratory experiments could be used to calibrate measures of fish conditions (i.e., wounding, behavioral deficits) with mortality (Davis 2005).

Mortality of Animals that Escape Fishing Gears or Are Discarded after Capture 11.3.2 Other Methods to Predict Mortality of Unrestricted Fish Present methods for estimating postcapture mortality require extended periods of holding or monitoring fish. As shown earlier, these methods can markedly limit the scope and replication of experiments and may introduce severe bias to the results. Ultimately, methods need to be developed that will provide unbiased mortality estimates for fish that are discarded or escaped from fishing gears (unrestricted by caging methods) and returned to their natural environments. Following are current (or potential) methods that attempt to satisfy these goals. Tag and Recapture Methodology. The tag and recapture methodology has been used to estimate the survival of discarded and released fish (e.g., Pacific halibut; Trumble et al. 2000). This method may provide an indication of long-term survival, but it does not explain much about those fish that have died (when, where, and why). In addition, the mark-recapture method used by Trumble et al. (2000) required a host of assumptions that make their survival estimates suspect. Electronic Tagging. Improved tag and “recapture” methods may and should be used to study the survival of escapees. For example, with current tagging technology, it is possible to mark fish with special tags that can be registered by instruments that are attached inside the trawl codend. With the help of such technology, it is possible to register the passage of marked fish through the codend. Hence, by first tagging a certain number of fish within a fishing ground and then trawling on this ground, it is possible to estimate how often a particular fish is captured in the trawl and how often it escapes successfully. It is likely that this type of new tagging technology will be used in the near future in many survival studies and other applications. Acoustic Telemetry. Acoustic telemetry (e.g., tagging fish with acoustic transmitters and manually tracking individual fish for hours or days) has been used to estimate postrelease survival for various fish species, and in particular pelagic fishes. Postrelease survival for many acoustic tagging and manual tracking studies were summarized by Pepperell and Davis (1999). The primary drawback of these types of manual tracking studies is sample size and

275

sample duration. In most cases, only one fish can be tracked in the ocean at a time, and tracking duration is often limited by weather conditions. Hence, postrelease survival estimates based on manualtracking methods often result in small sample sizes and short tracking durations (less than 1 day). Lindley et al. (2008) improved this method and demonstrated that acoustic telemetry can be used to estimate fish survival for large sample sizes (hundreds of fish) over long periods of time (longer than 1 year). They tagged 213 green sturgeons with acoustic tags and used arrays of stationary-acoustic receivers anchored in bays and the ocean along the Pacific coast of the United States and Canada to detect movements of individual fish. The high detection rate of these sturgeons by the stationaryacoustic receivers allowed Lindley et al. (2008) to estimate minimum annual survival rates. The use of stationary-acoustic receivers and coded-acoustic transmitters, therefore, provides the opportunity of estimating long-term, postrelease survival of discarded fish in their natural and unrestricted environment. The most recent high-tech method that has been used to estimate postrelease survival of fishes has been pop-up satellite archival tags (PSATs). These tags have been effective for estimating the postrelease survival of large pelagic fishes (e.g., striped marlin) caught by hook and line. For example, Domeier et al. (2003) estimated 74% survival for striped marlin caught and released by recreational sport fisheries using PSATs. They found that all mortalities occurred within 5 days of tagging. These tags have typically been somewhat large, and therefore could be used only on relatively large fish. The sizes of PSATs have recently been reduced; this technology can now be applied to a wider size range of fish than was previously possible. Condition Indices. It might be possible and often more efficient to predict discard and escapee mortality by observing fish condition during fishing experiments. Condition indices for discards have been developed for Pacific halibut based on wounding and for sharks based on revival time after discarding (e.g., Hueter et al. 2006; Trumble et al. 2000). In Alaskan pollock, a species that is sensitive to net abrasion, wounding was related to mortality; in other less-sensitive species, it was not (Davis and

276

Contemporary Issues in Capture and Conservation in Marine Fisheries

Ottmar 2006). General use of fish condition (e.g., wounding, plasma constituents) as a predictor for delayed mortality can be limited because these measures may show inconsistent responses to different types of fishing factors (Davis 2002; Davis and Schreck 2005; Parker et al. 2003). In addition, measures for wounding are typically subjective, thereby introducing bias in mortality estimates. Reflex Impairment. Reflex impairment (Davis and Ottmar 2006) is a quantitative measure of fish condition that may be rapidly evaluated in fishing experiments. It is a measure of fish condition that is correlated with stressor intensity. Reflex impairment has been used as a predictor for mortality in sablefish that were towed in a net and exposed to air and increased temperature (Davis 2005; Davis and Parker 2004). Davis and Ottmar (2006) further tested applicability of reflex impairment to predict mortality in flatfish and roundfish that had been towed in a net in the laboratory. Reflex impairment in Alaskan pollock, sablefish, Northern rock sole, and Pacific halibut was significantly related to mortality in biphasic relationship described by sigmoid curves. Davis and Ottmar (2006) noted that the ability to measure sublethal and lethal effects of seemingly unlimited combinations of fishingrelated stressors using RAMP (reflex action mortality predictor) would greatly enhance our ability to model bycatch discard and escapee mortality in fisheries and to reduce uncertainty in estimates of fishing mortality associated with bycatch. An analysis of the general applicability for using these measures to predict mortality for a wide range of species and fishing conditions in the field, however, remains the subject for future research. Measurement of reflex impairment may become a powerful tool for the assessment of survival likelihood in the field. It might also be used for comparing bycatch mortality among various fishing practices and fisheries. In conclusion, in a survival experiment, a fish may die of causes other than capture- and escaperelated damage. Most survival studies so far have resulted in estimates of limited accuracy and possibly great bias. The key issue in assessing the survival of escaping fish is to ensure that the collection and monitoring methodology does not induce stress and injury in the fish that are subject to investigation. Experimentally induced stress and injury

should be eliminated to the extent possible. Ultimately, methods should be developed to estimate mortality of escaping fish that are not subsequently restricted to cages or covers. High-tech tagging methods such as coded sonic transmitters and PSATs offer the newest possibilities to explore the behavior and survival of unrestrained fish. Although these and other methods have been applied to several survival studies, there is substantial potential to improve techniques and standardize procedures used in survival experiments. 11.4 FACTORS CAUSING STRESS, INJURY, AND MORTALITY Improving the survival of escapees and discards by using better gear and operational solutions requires detailed knowledge of basic factors affecting the stress, injury, and mortality of fish. This section attempts to review and highlight some of the key factors. It is evident that damage and mortality incurred by fish during their capture and escape are often caused by a multitude of factors (stressors) and can only rarely be ascribed to a single cause. In trawl fisheries, capture stressors include such factors as net entrainment, crushing, wounding, sustained swimming leading to exhaustion, and changes in pressure. Towing time and speed, light level, water and air temperature, anoxia, sea conditions, time on deck, and various handling procedures may affect stress and injury. Biological attributes include species, size, age, reproductive condition, and behavior. In addition, there are factors such as seabird predation near the surface and fish and mammal and invertebrate predation in the water column and near the seabed. 11.4.1 Escape Mortality In the case of towed fishing gears, fish usually escape from the codend. Most survival studies have therefore focused on the mortality of codend escapees. However, the zone of influence of a towed gear is not limited to where the fish are retained and where they escape; it also includes those parts of the gear that herd and scare fish (Fig. 11.4). Very little is known about the fate of fish escaping from these areas but it is widely assumed that stress, injury, and mortality are substantially less for fish

Mortality of Animals that Escape Fishing Gears or Are Discarded after Capture

Pre-capture

Capture process

Escape

277

Post-escape

Temperature, light, turbidity, currents, pressure, seabed, sea state Fish species, size, age, condition, behavior, swimming exhaustion Tow duration, speed, gear operation Contacts with netting Collisions with other fish

Catch size and composition

Mesh size and shape Twine type and stiffness Selection device Predators

Figure 11.4. A schematic description of potential factors that affect fish behavior, endurance, stress, and injury during capture and escape processes. The effects of these factors may be cumulative.

escaping the trawl prior to the codend than for those escaping from the codend. Fish Size, Skin Damage, and Mortality. Escape survival from trawl gear generally increases with increasing fish length, regardless of fish species (e.g., Breen et al. 2007; Pikitch et al. 2002; Sangster et al. 1996; Suuronen et al. 1996b). This may be surprising because an increase in injury and death rates would seem more likely in larger fish whose passage through a mesh or grid is more difficult due to their physical size (i.e., in the largest escapees). Very few investigations, however, have documented a marked increase in injury and mortality among the largest escapees (but see Breen et al.

2007; Efanov 1981). Clearly, there are complex relationships between the size of fish and their injury and mortality due to capture and escape. It is likely that the higher survival of larger escapees can be explained at least partly due to their sustained swimming ability. Furthermore, mortality of escapees may be partly a function of fish age. There is some evidence that in a particular year class, the smallest haddock and whiting are more susceptible to injuries or stress than are faster-growing individuals (e.g., Sangster et al. 1996; Wileman et al. 1999). Hence, the fitness (physical condition) of a particular fish, and not only its length, may play a vital role in its ability to survive codend escape.

278

Contemporary Issues in Capture and Conservation in Marine Fisheries

Many investigations have shown an inverse relationship between skin injury and the size of the escaping fish (Breen and Sangster 1997; Ingolfsson et al. 2002; Soldal et al. 1991, 1993; Soldal and Isaksen 1993). This supports observations that the smallest escapees often suffer the highest mortalities. Apparently, smaller fish with poorer swimming ability are less able to avoid injury when swimming within the gear and during escape. They may also have less physical strength to make active escape attempts and may therefore stay longer inside the gear before escaping. The smallest fish are generally more delicate than larger individuals and are therefore more susceptible to all types of captureinduced injury. Their high vulnerability may be the result of a combination of exhaustion and injury. It is notable, however, that several studies have been inconclusive regarding the relation between skin injury and fish size (Lowry et al. 1996; Pikitch et al. 2002; Sangster et al. 1996; Suuronen et al. 1996a, 1996b, 2005). Even though scale loss is often described as the most prevalent injury for fish escaping trawls (see Suuronen 2005), it is not conclusive that this “injury” is the primary cause of mortality for escapees. Nonetheless, many studies suggest that skin injuries may result in delayed mortality. Skin injury may expose fish to secondary infections that significantly increase their longer-term mortality (Bullock and Roberts 1980; Jones 1993; Roberts 1989). Although minor injuries may heal completely and open skin lesions may be replaced by scar tissue within a few weeks after injury, there may be secondary infections from potentially pathogenic bacteria and fungi that ultimately result in death. This may be magnified in warmer waters, where such infections may exacerbate the lesions and the fish will succumb to osmotic distress or septicemia if infections becomes generalized (e.g., Mellergaard and Bagge 1998). Long-term observations are required to understand the role of skin injury in determining mortality. Mesh Size and Shape. It is generally assumed that the larger the mesh opening, the easier it is for fish to pass through, and consequently the less damage that occurs. Indeed, an inverse correlation between mortality rates and increasing codend mesh size has been reported in several investigations (e.g., Lowry

et al. 1996; Main and Sangster 1991; Sangster et al. 1996; Wileman et al. 1999). In some cases this may have been mainly because more large fish escaped through the larger meshes; that is, the average mortality decreases as mesh size is increased because larger fish exhibit lower mortality. Nevertheless, Lowry et al. (1996) demonstrated that fish escaping through a larger mesh may sustain less skin and scale damage than fish of similar size escaping through a smaller mesh. Some studies have suggested, however, that codend mesh size has less influence on the survival of escaping fish (e.g., Suuronen et al. 1996b; Wileman et al. 1999). Clearly, the positive effect of an increase in mesh size on the survival of codend escapees has not yet been demonstrated conclusively. In general, the escape (per se) is not the main cause of injury and mortality. This does not mean that mesh size does not play any role in survival; it is potentially important, but many other important factors may affect survival simultaneously. There are indications that mesh shape may play a more important role than mesh size in deciding mortality of escapees. For example, haddock and whiting escaping through a square mesh codend were reported to have lower mortality than those escaping from a traditional diamond mesh codend of the same mesh size (Main and Sangster 1990, 1991). Roundfish may escape through open square meshes with less injury than through conventional diamond meshes because the latter may become almost closed due to net tension during the tow. In some investigations, escape through a special sorting device, such as a sorting grid, has resulted in lower mortality than through meshes (e.g., Ingolfsson et al. 2002; Suuronen et al. 1996c), but this has not yet been demonstrated conclusively. Effects of a selective device, whether it is a netting mesh or a sorting grid, on the injury and mortality of escapees are complex and are poorly understood. To determine the potential injury to fish passing through meshes or other selective devices (e.g., sorting grids), some studies have attempted to simplify the observation process by conducting simulated laboratory (tank) experiments. Species studied include cod, haddock, saithe, and sand whiting (e.g., Broadhurst et al. 1997, 1999; DeAlteris and

Mortality of Animals that Escape Fishing Gears or Are Discarded after Capture Reifsteck 1993 Engås et al. 1990; Jónsson 1994; Soldal et al. 1989, 1993). In most cases, mortality was indistinguishable from that of the control fish, which supports the view that the simple passage of a fish through a netting mesh or other selective device does not necessarily inflict fatal injury. However, results obtained in laboratory conditions cannot be directly extrapolated to the more complex nature of a commercial fishing process (see, e.g., Soldal et al. 1993). Netting Material. Most netting materials are abrasive. Therefore, fish may sustain severe skin injuries during the tow, especially in the codend where individuals are exhausted and crowded together. In a number of studies, gear-induced skin injuries have been observed in fish that escaped from a trawl codend (Borisov and Efanov 1981; Main and Sangster 1990, 1991; Suuronen et al. 1996a; Wileman et al. 1999). Fish can sustain injury long before the escape. Baltic herring scraped against the trawl netting along the trawl belly and codend extension prior to entering into the codend, which resulted in scale loss and skin injury (Suuronen et al. 1996b, 1996c). Apparently, these herring, at least the smallest ones, were not able to avoid contact with the trawl netting. Davis and Ottmar (2006) showed a high delayed mortality (up to 20 days) for Alaskan pollock (12–36 cm) after towing 30 min at a speed of 1.1 m/s in a tank; at this speed, the fish were pinned against the net or each other and suffered skin injury. Pikitch et al. (2002) observed Alaskan pollock striking meshes with their caudal fin during escape, being impinged against the codend meshes just prior to escape, or squeezing through meshes during escape. Fish may sustain injury while still confined by the mesh. Catch Volume and Quality. It is generally assumed that increasing catch size increases the likelihood of injurious mechanisms within the codend, owing to increasing abrasive contacts with other fish, netting, and debris, and exhaustion due to turbulent flow patterns. In survival experiments, however, no significant relationship between codend catch size and escape mortality has been demonstrated (Pikitch et al. 2002; Suuronen et al. 2005; Wileman et al. 1999). However, owing to many methodological constraints, most survival studies have been conducted with small catches that

279

typically do not reflect commercial situations. It is also notable that the effects of catch size and catch composition may be highly confounded by such variables as the towing duration, towing speed, and environmental conditions. Several studies have suggested that the proportion of abrasive objects, such as spiny fish species, crustaceans, and broken shells, in the codend may injure escaping fish (Main and Sangster 1991; Treschev et al. 1975; Wileman et al. 1999). Scale loss was significant in several tropical fish species studied by Farmer et al. (1998). Heaviest losses were observed for species with deciduous scales, such as perforated-scale sardine, pearly finned cardinal fish, and Sunrise goatfish. It is notable that escapees from a 45-mm square mesh codend were generally less damaged than those escaping from a 38-mm square mesh codend. Towing Speed and Duration. The effect of capture-induced exhaustion on the mortality of escapees is not clear. The interactions between fatigue, towing speed, injury, and survival are likely to be very complex. Xu et al. (1993) showed that small Alaskan pollock became severely fatigued during the trawl-capture process, and Beamish (1966) suggested that muscle fatigue alone could cause mortality. Young vendace subjected to a trawl capture and escape process exhibited strong symptoms of exhaustion (Turunen et al. 1996); this may contribute markedly to the high mortality observed for these escapees (Suuronen et al. 1995). Likewise, Suuronen et al. (1996b) suggested that the dramatic reduction of liver glycogen observed for small (7–11 cm) Baltic herring escapees after forced swimming inside the trawl may have increased their susceptibility to stress, thereby contributing greatly to their high mortality. Although there are little published data describing the effects of towing speed on survival of animals escaping trawls, we suggest that towing speed has a significant effect on fishes’ swimming endurance within the trawl and escape probability from the trawl. Clearly, more work is needed to assess the importance of stress and swimming exhaustion on the survival of escaping fish. It could be assumed that towing duration has a strong effect on injury. A longer tow is likely to produce more catch, crowding, abrasion, and more

280

Contemporary Issues in Capture and Conservation in Marine Fisheries

blocked meshes. The length of time that escaping fish spend inside the trawl, however, is not generally known, and there may be substantial variation depending on the conditions, species, and sizes. For fish that escape soon after entering the codend, the length of trawl tow is not necessarily the most important factor affecting their subsequent survival. Indeed, Wileman et al. (1999) and Pikitch et al. (2002) found no clear effect of trawl towing time on the escape mortality of haddock and whiting or Alaskan pollock, and Erickson et al. (1999) observed many Alaskan pollock escaping almost immediately after entering the codend. Nevertheless, towing time may be extremely important for the survival of fish that do not escape immediately. These fish will become exhausted and accumulate in the catch. Some of these fish may escape later, particularly during haul-back when tension on the codend netting is reduced. Fish that escape during haul-back are likely to exhibit very different mortality rates than individuals that escape the trawl immediately after entering. It is noteworthy that Sasso and Epperly (2006) reported that the risk of mortality to sea turtles escaping trawls increased with increasing towing time. Tows of 10 min or less had negligible mortality while intermediate tow times (longer than 60 min) resulted in rapidly escalated mortality. Haul-back and Depth Change. Mortality of fish that escape trawls during the haul-back process is likely to increase due to depth changes during haulback, especially for those fish that escape near the surface. Fish that escape at or near the surface risk decompression problems, such as injury to their gas bladders. Fish with ruptured gas bladders often have gas in their abdominal cavities and are trapped on the water surface. After escape, these “floaters” may be subjected to high predation by birds. Moreover, various mechanical forces on a codend floating at the surface can be extremely high, particularly in rough seas. Catch may become compressed in the aft part of the codend, and fish may become crowded, resulting in oxygen depletion, abrasion, and injury. Finally, fish that are hauled through great depths are not only subjected to large changes in pressure, but also in some cases, large changes in water temperature and salinity. While there is little available information on the fate of

fish escaping from codends near the surface, these factors suggest that the probability of survival is quite low. In fact, Breen et al. (2007) demonstrated that the mortality of surface escaping haddock was significantly higher those escaping at depth during towing. Clearly, escape should take place during fishing and not during hauling to increase the odds of survival. Water Temperature. Water temperature influences physiological processes and behavior of fish (He and Wardle 1988; Özbilgin and Wardle 2002); however, almost no fieldwork has been performed to quantify the effects of water temperature on the survival of escapees under commercial fishing conditions (but see Suuronen et al. 2005). In laboratory studies, mortality rates of sablefish, lingcod and Pacific halibut increased with increasing water temperature for fish that were first towed and then exposed to increased temperature, with 100% mortality at 16°C for sablefish, 18°C for Pacific halibut, and 20°C for lingcod (Davis et al. 2001; Davis and Olla 2001, 2002). Although there were substantial species-specific differences in mortality rates, these results demonstrate the marked effect of temperature on survival of fish escaping codends. An abrupt temperature increase of several degrees induced high mortality in adult sablefish (Olla et al. 1998). Exposure to high temperatures results in increased core body temperature, with smaller fish warming more rapidly. Davis et al. (2001) argued that additional stress and mortality resulting from the interaction of capture, escape, and exposure to increased temperature may be common in fisheries during warmer seasons of the year or in areas where fish are caught in cooler deeper waters. Low water temperatures may also affect fishes’ behavior and sensitivity to various capture stresses. Fish swimming speed and endurance generally decrease at lower temperatures (He 1991). Reduced swimming speed and endurance may influence the herding effect of various gear components and the ability of fish to escape from trawls. Davis (2002) argued that deficits in swimming performance and orientation in a trawl associated with low temperatures could cause fish to be injured more frequently. In conclusion, water temperature may play a critical role in the mortality of escapees and may magnify effects of other stressors.

Mortality of Animals that Escape Fishing Gears or Are Discarded after Capture Light Condition. Olla et al. (2000) found that Alaskan pollock could not swim and orient in trawls in the dark. Lack of swimming and orienting ability in darkness would likely increase injury to escapees and discards (Suuronen et al. 1995). Fish that use visual cues to orient would likely be most impacted in trawls in darkness. It is noteworthy that survival experiments often do not control for light conditions. Vessel Movement. Fishing vessels roll, pitch, and heave in response to waves and winds. Trawl “surging” associated with increased sea state may alter the water flow within the gear and make it difficult for fish to orient toward selective panels or sorting devices. Hence, one effect of heavy seas may be increased stress and injury (relative to escapement during calmer seas). On the other hand, increased sea state may also cause increased escape during towing and haul-back (e.g., Engås et al. 1999). Groundgear Collision. Many researchers have demonstrated that fish may collide with the groundgear (e.g., footrope or head rope) during trawl herding and capture (Jørgensen et al. 2006; Walsh and Hickey 1993). Although the magnitude of such collisions is uncertain, Jørgensen et al. (2006) found that a significant number of young cod exhibited external injuries caused by collisions with groundgear. The survival of fish that collide with groundgear has not been measured. If survival for these individuals is higher than those that escape through codend meshes, then escape under modified groundgear could become an alternative to codend selectivity for some species (e.g., cod). Post-escape Predation. Marine predators are known to follow trawls during towing and consume fish that escape from codend meshes (e.g., Broadhurst 1998). Under laboratory conditions, Ryer (2002) showed that Alaskan pollock subjected to simulated capture and escape stress were more likely to encounter predators than a control group. Ryer et al. (2004) emphasized that even a relatively small change in predator detection, avoidance, schooling, and shelter seeking could have profound implications for survival. Clearly, increased vulnerability to predation may be an important, yet largely unobserved, source of mortality for fish escaping trawl gears. By not accounting for predation, many

281

survival experiments may underestimate escape mortality. Moreover, fish may be transported for significant distances in front of and inside the trawl. When they escape, they may be in a very different environment where they may be more vulnerable to predators. This may further reduce the effectiveness of innate behavioral responses to predators, make it more difficult to find shelter and food, and reduce shoaling and swimming. The influences of such effects are poorly understood and require substantially more investigation. Repeated Capture and Escape. Very little is known about the effects of repeated capture and escape of organisms from trawls (but see Broadhurst et al. 2002). On grounds where fishing activity is high, young fish may repeatedly encounter trawl gears. If stress and injury are cumulative, fish may eventually die after they have experienced several capture and escape processes within a short period, without sufficient time for adequate recovery between the events. This hypothesis clearly requires further investigation. Moreover, learning may play a role in escape. A fish that has been captured and passed through the meshes of a fishing gear may be more capable of escaping during subsequent encounters, and may therefore have a higher probability of survival than a naive fish. Özbilgin and Glass (2004) provided experimental evidence that haddock can modify their behavior based on prior experience; one mesh penetration tended to increase the probability of penetration during the next encounter. The true nature and effect of learning in codend mesh penetration during commercial fishing operations remain to be investigated. In conclusion, few studies have adequately and quantitatively explained the full range of injuries and mortalities that can occur when fish escape from fishing gears under commercial fishing conditions. A number of mechanisms may cause injury, stress, and mortality; the passage through a mesh or a selective device is not the only potentially damaging factor. Fish may escape from many parts of the gear and may therefore be affected differently. Fish escaping from fishing gears may suffer immediate as well as delayed mortalities owing to physical injury, exhaustion, disease, and predation. Moreover, changes in water temperature, pressure, and light conditions may strongly affect the fate of

282

Contemporary Issues in Capture and Conservation in Marine Fisheries

escaping fish. The robustness and ability of various species to withstand physical disruptions and fatigue associated with the process of capture and escape vary substantially. The smallest escapees often appear to be the most vulnerable. Until the effects on mortality of various critical factors and their interactions are better understood, there will be a lack of confidence in generalizing escape mortality results to a wider range of fishing conditions, gear designs and operations, and fish species. Further work is required to identify damaging mechanisms that cause injuries and mortality. 11.4.2 Factors Causing Mortality of Discarded Fish Many factors that affect discarded fish are similar to those affecting escapees. Discarded fish, however, experience additional stress and injury from lifting on to the vessel deck, on-deck handling, air exposure, and the eventual discarding process (thrown from the vessel and then sinking or swimming back to their habitats). Fish with gas bladders and other organs that inflate after capture because of pressure changes often become “trapped” near the surface after discarding and may therefore suffer near 100% mortality. Hence, it is not surprising that bottomdwelling roundfish such as cod, haddock, and saithe generally do not survive discarding processes well (reviewed in Broadhurst et al. 2006; ICES 2000). On-Deck Handling. On-deck handling time and air temperature are among the most important factors influencing the survival of discards. Many studies have demonstrated that changes in fishing practices that reduce handling time and exposure to air will reduce discard mortality (assuming that lethal stress level from other sources has not been encountered). Exposure times can range from a few minutes to several hours depending on catch amount and handling procedures. Exposure may occur over a wide range of air temperatures. Higher survival is often associated with shorter air exposure times and lower air temperature (above freezing) on deck (de Veen et al. 1975; Erickson and Pikitch 1999; Kelle 1976; Neilson et al. 1989; Pikitch et al. 1996; Thurow and Bohl 1976). Impacts of air temperature on deck may be compounded by direct sunlight (Kelle 1976; Pikitch et al. 1996). Note that freezing air temperatures (less than 0°C) may cause high

mortality for animals on deck. Exposure to bright sunlight may also temporarily or permanently impair animal vision, but this possible effect has not yet been experimentally evaluated. Thermoclines. Extreme thermoclines with high surface water temperatures may strongly affect the survival of discarded fish. Erickson et al. (1997) observed very high mortality (greater than 95%) for trawl-caught and discarded sablefish when surface water temperatures were high (18° to 20°C). Mortality was substantially lower when surface water temperatures were 12° to 15°C (Erickson and Pikitch 1999). Laboratory studies have shown similar results. Exposure of sablefish held at 5°C to a range of seawater temperatures between 12° and 16°C in the laboratory resulted in loss of feeding and increased physiological stress and mortality (Davis et al. 2001). Mortality occurred after towing fish in a net at 5°C and then exposing them to 12°C. Exposure to 16°C resulted in complete mortality for sablefish. Similarly, mortality rates increased with increasing seawater temperature for Pacific halibut and lingcod that were towed at 5° and 8°C and then exposed to higher water temperatures, with 100% mortality at 18°C in Pacific halibut and 20°C in lingcod (Davis and Olla 2001, 2002). Apparently, there are marked species-specific differences in temperature sensitivity. Towing Duration and Catch Size. Pikitch et al. (1996) estimated the mortality and physiological condition of trawl-caught and discarded Pacific halibut in the Gulf of Alaska. Tow duration (1–3 h) was among factors that significantly affected halibut mortality. Similarly, the mortality of trawl-caught and discarded sablefish increased with longer tow durations (Erickson and Pikitch 1999). These studies suggest that tow duration can be a significant factor affecting the survival of discarded fish. On the other hand, it is generally not known how long discarded fish have spent in the trawl. The duration of a tow did not have a significant effect on the stress and mortality of undersized brown trout released after the haul (Turunen et al. 1994). Catch abundance, on the other hand, was a significant factor in causing stress. The effect of tow duration might be connected to codend catch size and catch composition, as well as to towing speed. It should be noted that catch volumes are rarely very

Mortality of Animals that Escape Fishing Gears or Are Discarded after Capture large in experimental tows, so this variable is likely to come out as nonsignificant during most scientific investigations. Nevertheless, catch weight significantly affected discard mortality of the spiny dogfish (Mandelman and Farrington 2007). Both longer tow duration and larger catch quantities increased mortality of undersized discarded plaice (Kelle 1976). Clearly, the effect of towing duration and catch size on discard mortality is complex and should be explored more carefully. Catch Composition. Catch composition likely has an effect on discard mortality. Higher mortality may be assumed when the catch consists of spiny species and rubbish/trash. The amount of sand mixed in with the catch was shown to be inversely correlated with survival for Pacific halibut by Pikitch et al. (1996). Predation and Scavenging. One source of mortality for discarded animals that is difficult to estimate is postdiscard mortality caused by predators or scavengers for weakened individuals. Erickson and Pikitch (1999) used an underwater camera to observe trawl-caught and discard sablefish returned to the sea in cages and noted that individual fish often laid on their side ventilating rapidly while recovering from the stresses of capture and discard. The duration of this “recovery” period often exceeded 1 h. This postdiscard behavior was also observed for discarded Pacific halibut (D. Erickson, unpublished video observations). Depending on the location of discard, the sea bottom can be teaming with scavengers such as hagfish and sand fleas (amphipods). Although underwater observations of discarded fish suggested that hagfish only “attacked” dead individuals, it appeared as if sand fleas attached to discarded fish that were confined to the seabed, whether dead or alive. Hence, discarded fish may be extremely vulnerable to scavengers such as sand fleas during initial hours after discard while laying on the seabed recovering from the stresses of trawlcapture and discard. A discarded fish may survive the discarding process and predation over a short period of time. However, delayed mortality of discards may represent a significant source of unobserved mortality. For example, disruption of schooling behavior and attractions of predators by visual, olfactory, and mechanical cues from injured fish may significantly increase

283

mortality hours or days after discarding (Ryer 2002; Ryer et al. 2004). It is evident that damage and mortality incurred by fish due to capture and discarding are often caused by a multitude of factors (stressors) and can only rarely be ascribed to a single cause. The effects of all these stressors may be cumulative and ultimately lead to higher mortalities. 11.5 MEASURES TO IMPROVE SURVIVAL 11.5.1 Improving the Survival of Escapees Efforts to reduce or eliminate bycatch mortality in trawl fisheries have centered on gear modifications that increase the opportunities for undersized fish or unwanted species to pass through codend mesh or another selective device. Many gear modifications and bycatch reduction devices have proved effective in guiding and sorting fish (Valdemarsen and Suuronen 2003). It is generally assumed that effective sorting and better selectivity result in good survival. This assumption may be true for many cases, but it should not be an automatic presumption. Improving selectivity without reducing the damage incurred by capture and escape is not an appropriate way of protecting immature fish. Hence, when developing and improving selectivity, it is important to address the whole range of stressors caused by the capture and selection process. This section presents some basic principles that may help to reduce the injury and mortality associated with fishing and sorting processes. Reducing Time between Capture and Escape. Fish often swim within or in front of the trawl mouth for a period of time before they are overtaken by the gear (Main and Sangster 1981; Tschernij and Suuronen 2002; Wardle 1993). After dropping inside of the trawl, fish continue to undergo forced swimming and experience contacts with the netting and other fish, crowding, crushing, and barotrauma. Once fish finally enter the codend, they may be too exhausted to continue swimming and are therefore unable to escape (e.g., Breen et al. 2004). The longer the capture process, the higher is the likelihood of severe exhaustion and physical damage. Therefore, the time between entering the fishing zone of the net and escaping should be minimized by proper gear construction and operational aspects.

284

Contemporary Issues in Capture and Conservation in Marine Fisheries

Consideration should be given to gear modifications that ensure that nontarget fish are physically capable of making an active escape. Moreover, selection devices should promote the escape of fish at depth of towing because the survival of escapees is likely to be poorer during haul back. Increasing Escape Opportunities. Trawl gear should be designed to allow fish to escape before they enter into the rear part of the codend, where the risk of damage is highest. This is not only because of crowding and crushing but also because of problems caused by clogged meshes in the codend. Many investigations have demonstrated that codend meshes may become blocked by larger catches (e.g., Erickson et al. 1996; Suuronen and Millar 1992). In such a situation, escape from a codend is difficult or impossible. For many species, an escape panel or selective device installed ahead of the codend will allow undersized fish to escape, even when the codend meshes become blocked with fish. Such an approach would be useful in particular for many pelagic fisheries where blocking of codend meshes is common and the fish species in question may experience high mortality when escaping from the codend (Suuronen et al. 1997). In these cases, escape opportunity should be offered as early as possible. Special guiding devices might be used to guide fish to an area where sorting and escape can occur. Even a very small sorting device may work efficiently when located in a strategically correct position. Mechanical sorting likely causes more injury to fish than behavior-based escape. Behavior-based escape requires that fish can detect the escape route and are capable of passing through the escape device. Commercial fishing operations are often conducted at great depths or at night, in dark conditions. When light levels are inadequate for visual mesh detection, fish generally have less chance of orienting properly toward a selective device (Olla et al. 1997, 2000), and survival may therefore be poor (Suuronen et al. 1995). Moreover, the codend is often surrounded by a dense mud cloud stirred up by the groundgear. The design of trawls to enhance bycatch escape and survival must consider characteristics of fish reaction under such conditions. There should be elements that will guide fish to the escape route. When there is sufficient light at fishing

depth, contrast of critical gear components should be maximized to facilitate easier escape. Contrast patterns between netting and surrounding water has been shown to affect fish escape behavior (e.g., Glass et al. 1995). Management of water flow patterns inside the codend is an option to enhance escapement when visibility of gear component is low. Flow patterns can be managed by appropriate gear design and rigging and through the use of various flow enhancement panels and other devices (e.g., Broadhurst 2000; Engås et al. 1999). Much effort has recently been made to the development of various types of bycatch reduction devices (e.g., Broadhurst 2000; Engås et al. 1998; Main and Sangster 1982). These devices work best when there are large size differences among animals that are to be separated. Species-selective devices typically exploit behavioral differences between species. For example, in prawn trawls, species selection occurs because fish exhibit specific responses to towed trawls (e.g., fish typically swim upward when inside of a trawl), whereas prawns and other benthic invertebrates show no or very limited responses. Behavioral differences may even be used to guide fish out of the trawl without the need to use any filtering or sorting devices. Such mechanisms are likely to increase survival chances of escaping fish. Modification of Fishing Practices. In addition to gear design, fishing procedures can be modified to maximize escapement and reduce injuries to escaping animals. It is generally assumed that likelihood of fish injuries within the codend increases with catch size. Therefore, shorter but more frequent tows with smaller catch sizes would improve selectivity properties of the codend and minimize injurious mechanisms, especially in fisheries where catch rates are high or the catch consists of large amount of debris. It is also evident that trawls towed at low speeds would cause less damage than those towed at higher speeds. Knotless, smooth-surfaced netting materials should generally be favored over stiff, knotted nettings in areas where the netting can cause abrasion of escapees. Modification of Groundgear. Excluding debris and large objects (e.g., boulders) before they reach the codend will reduce physical damages to fish caught in the codend, thereby increasing survival

Mortality of Animals that Escape Fishing Gears or Are Discarded after Capture chances of those that escape. By using improved groundgear constructions, seabed objects can be overrun rather than trapped inside the trawl. In addition, unwanted objects (e.g., rocks) that enter the trawl can be removed through various types of exit devices and holes. Such devices should be robust, effective, inexpensive, and easy to manufacture, maintain, and repair. Construction of a “demersal” groundgear that effectively catches fish without impacting the fragile seabed organisms (corals, etc) also contribute to ecosystem health, which in turn ensures a sustainable fishery. For some groundfish species, the groundgear selection might be used as an alternative to codend selectivity. This would be useful in particular for species that suffer high mortality due to the escape from the trawl codend. There are many potential but untested design options for developing selective groundgears in demersal trawl fishing. There are also other ways of separating species and sizes based on behavior. For instance, King et al. (2004) and Hannah et al. (2005) describe a flatfish trawl that has a cutback headrope designed to separate rockfish prior to their entrainment inside the trawl. This type of trawl design permits nontarget species to escape before entering into the gear and has a significant potential to reduce bycatch mortality. Indeed, a flatfish trawl has been required since 2005 for trawler access into shallow grounds along the U.S. West Coast. Clearly, there is substantial scope for development in this area. Another alternative that has been little explored or exploited is to use counterherding ahead of the trawl mouth. This alternative eliminates the capture of bycatch altogether. Just as herding brings fish into the path of the trawl, it may be possible to herd unwanted species and sizes out of the path of the trawl prior to their entering the trawl mouth. Counterherding would take advantage of the different responses of species to herding stimuli placed ahead of the trawl mouth. In conclusion, developing effective gear modifications to improve the survival of escapees requires a good understanding of how fish react to gear under light and dark conditions, as well as warm and cold water conditions. Evaluation of escape mortality should be an integral part of the development of selective fishing gears. Survival should

285

ultimately be used to determine suitability and success of gear modifications. There are likely to be many technical modifications to gears and changes to operational and handling procedures that can be developed and implemented to mitigate unaccounted mortality. Fisher or manager acceptance of gear modifications will also be an important factor for increasing survival of escapees. 11.5.2 Improving the Survival of Discards Generally, stress response and injury of discarded fish vary according to gear type and fishing and handling techniques. For many species, this mortality can be reduced through improved on-deck handling procedures (e.g., how fish are picked up and tossed overboard) and other operational improvements (Broadhurst et al. 2006). Changes in fishing practices, such as reduced towing duration, catch amounts, handling time, and exposure to air, are likely to reduce mortality considerably. In some cases, very small changes in operational and handling practices can markedly improve survival of discarded animals. For example, immediate and delayed mortalities of trap-caught snow crab increased substantially with increased drop height on deck (Grant 2003). Sorting catches in onboard containers like hoppers containing aerated water at appropriate temperatures can provide immediate and large benefits, especially for those species that are resilient to physical damage and stress incurred during capture but extremely sensitive to extended air exposure and large changes in temperature (e.g., brown shrimp and school prawns). Parker et al. (2006) suggested that a quick return to depth could be used to minimize mortality of discarded rockfishes in nearshore fisheries because rapid recompression reversed barotraumas. Survival may be improved if discarded fish are avoided for the exposure on the surface of water and are out of the reach of avian predators. This could be accomplished with the help of special “release tubes.” A release tube could facilitate the sinking of discards; fish would come out of the tube at about 5- to 10-m depth, with gas bladders recompressed to release the air and not pop back to the surface. Another method that could be modified and potentially applied to trawl fisheries to is the “discard cage system,” a method used by the salmon

286

Contemporary Issues in Capture and Conservation in Marine Fisheries

troll fishery in the North America (Farrell et al. 2001). Such a system has a special towed cage where undersized fish to be released are kept. In the salmon fishery, fish are allowed to recover in the cage while trolling continues (at slow speed). The similar principle could be used in trawl fishery, with a larger cage hanging over the side of the boat during sorting. The cage could be opened and fish released once the boat has moved some distance from the haul-up position (with less predators). This type of system might be most suitable for fisheries where relatively few fish are handled and discarded. The development of live release programs for discards that take advantage of presorting and discarding-sensitive species prior to general sorting of the catch could be developed. Live release requires the identification of good candidates for survival using a system such as RAMP for real-time assessment of fish condition during presorting. Climatic conditions have been shown to have a strong impact on mortality of discarded fish and invertebrates. When possible, fishing seasons should be established to exclude extreme water (and air) temperatures. For instance, Smith and Howell (1987) assessed mortality of trawl-caught American lobsters that were exposed to subfreezing (−9.5°C) air temperatures for periods of 30 to 120 min. The mortality of lobsters reached 100% at 120-min exposure.

11.6 CONCLUDING REMARKS For some species, discard mortalities can be reduced through reducing total catch per haul and sorting time (thereby reduced exposure to air) and improved on-deck handling procedures, but in many cases a significant reduction of discard mortality is difficult to achieve. Efforts should therefore be directed toward maximizing escape of fish during fishing. Mortality associated with escape process may be relatively low for many species, in particular when compared with discard mortality. There are various options available to improve survival of escapees. Installing escape panels or other devices at strategic positions in a fishing gear can increase escape and enhance survival of juveniles and nontarget species. Ideally, escape should

occur before the fish enter into the aft part of the codend where the risk of serious injury is greatest. Furthermore, facilitating voluntary escape through various constructional and operational solutions would increase likelihood of their survival. The use of nonabrasive netting materials, the exclusion of debris and large objects from codends, and better design, operations, and rigging of nets could improve survival. Clearly, improving the survival of escapees by using better gear modifications and operational solutions requires detailed knowledge of basic factors affecting stress, injury, and mortality of escaping fish. The likelihood of survival of fish that have been released, or have escaped, from a fishing gear varies markedly among gear types. The injuries caused by a demersal gillnet are different from those caused by a floating trap. Susceptibility to injury also varies with species and size of fish. Survival ultimately depends on how well the fish are able to adapt to capture (and release) conditions. Clearly, difference in injury among gear types is an important aspect when considering appropriate and sustainable fish capture methods. Many species targeted by trawls can be caught by a variety of other more selective and more environmentally friendly gears such as seine-nets, longlines, traps, and pots. Owing to their catching mechanism, these alternative fishing methods often have lower escape and discard mortalities. It is obvious that methods for assessing escape mortality rates across a wide range of fisheries and environmental conditions are not yet adequate. It is necessary to develop appropriate methodologies, collect more accurate data, and obtain a better understanding of the main sources of injury, stress, and mortality under various conditions. It should be borne in mind, however, that owing to natural variations in environmental parameters and fish conditions, there will always be some variability in mortality estimates between experiments, tows, and years. Research on the mortality of fish escaping from fishing gears has tended to focus on the mortality of fish kept in a sheltered environment such as seabed cages, for a relatively short time period. Factors such as predation on injured fish and the ability of a fish to recover fully from its injuries or

Mortality of Animals that Escape Fishing Gears or Are Discarded after Capture stress are more difficult to monitor and are therefore poorly understood. The fate of fish after multiple encounters with fishing gears is largely unknown. Moreover, the cumulative effects of all stressors are likely to have a strong influence on the probability of long-term survival. These areas clearly require more investigation. It is evident that as the trend for using more selective gears continues, the relative importance of escape mortality on stock assessment will increase. Mortality studies have been done for only a few species and in only a few fisheries. Extending investigations to other fisheries and other species would improve the reliability of predictive modeling. It would be particularly useful to extend investigations to overfished stocks and to those stocks or species for which there is likely high mortality of escaping fish. Modifications to the gear and operational methods will not eliminate all potential problems associated with unaccounted mortality. Active avoidance of areas with high densities of juveniles and nontarget species is an alternative and preferred approach to reducing unaccounted mortality. Hence, rather than attempting to develop measures to increase the fishes’ opportunities to escape from fishing gears, immature fish and nontarget species could be protected by preventing or reducing their chance of encountering the fishing gear. Moreover, counterherding, which keeps unwanted species and fish sizes out of gear, may be an effective solution to the unaccounted mortality problem for mobile fishing gears. Finally, we suggest that a better understanding of the behavior of both targeted and discard species is necessary to apply management regulations that will maximize the economic value of the fishery while minimizing encounters with unwanted bycatch species. With the advent of modern tagging technologies and other high-tech instruments, specific behavior of animals of interest is more easily attainable. REFERENCES Bartholomew A and Bohnsack JA. 2005. A review of catch-and-release angling mortality with implications for no-take reserves. Rev. Fish Biol. Fish. 15: 129–154.

287

Beamish, FWH. 1966. Muscular fatigue and mortality in haddock, Melanogrammus aeglefinus, caught by otter trawl. J. Fish. Res. Bd. Can. 23: 1507–1519. Beltestad A and Misund OA. 1996. Survival of mackerel and saithe escaping through sorting grids in purse seines. ICES CM 1996/B: 24. Borisov VM and Efanov SF. 1981. To the problem of natural and traumatical mortality of Baltic herring. ICES CM 1981/J: 13. Breen M and Cook R. 2002. Inclusion of discard and escape mortality estimates in stock assessment models and its likely impact on fisheries management. ICES CM 2002/V: 27. Breen M and Sangster GI. 1997. The injuries sustained by haddock (Melanogrammus aeglifinus) escaping from trawl cod-ends and the implications of these to survival. ICES FTFB Working Group meeting, Hamburg, April 14–17, 1997. Breen M, Dyson J, O’Neill FG, Jones E and Haigh M. 2004. Swimming endurance of haddock (Melanogrammus aeglefinus L.) at prolonged and sustained swimming speeds, and its role in their capture by towed fishing gears. ICES J. Mar. Sci. 61: 1071–1079. Breen M, Huse I, Ingolfsson OA, Madsen N and Soldal AV. 2007. An assessment of mortality in fish escaping from trawl codends and its use in fisheries management. Project SURVIVAL: Q5RS-2002– 01603. Final Report. Brussels, EC. Breen M, Sangster G, O’Neill B, Kynoch R, Jones E and Soldal AV. 2002. Evidence of sampling induced biases in mortality estimates from experiments investigating mortality in fish escaping from towed fishing gears. ICES CM 2002/V: 25. Bremec CS, Lasta ML and Hernandez, D. 2004. Survival of Patagonian scallop (Zygochlamys patagonica, King and Broderip, 1832) after the size selection process on commercial fishing vessels. Fish. Res. 66: 49–52. Broadhurst MK. 1998. Bottlenose dolphins, Tursiops truncates, removing by-catch from prawn-trawl cod-ends during fishing in New South Wales, Australia. Mar. Fish. Rev. 60(3): 9–14. Broadhurst MK. 2000. Modifications to reduce bycatch in prawn trawls: a review and framework for development. Rev. Fish Biol. Fish. 10: 27–60. Broadhurst MK, Barker DT and Kennelly SJ. 1999. Scale-loss and survival of juvenile yellowfin bream (Acanthopagrus australis) after simulated escape from a Nordmøre-grid guiding panel and release from capture by hook and line. Bull. Mar. Sci. 64: 255–268.

288

Contemporary Issues in Capture and Conservation in Marine Fisheries

Broadhurst MK, Barker DT, Paterson BD and Kennelly SJ. 2002. Fate of juvenile school prawns, Metapenaeus macleayi after simulated capture and escape from trawls. Mar. Freshw. Res. 53: 1189–1196. Broadhurst MK, Kennelly SJ and Barker DT. 1997. Simulated escape of juvenile sand whiting (Sillago ciliata) through square-meshes: effects on scale-loss and survival. Fish. Res. 32: 51–60. Broadhurst MK, Suuronen P and Hulme A. 2006. Estimating collateral mortality from towed fishing gear. Fish & Fish. 7: 180–218. Bullock AM and Roberts RJ. 1980. Inhibition of epidermal migration in the skin of rainbow trout Salmo gairdneri Richardson in the presence of achromogenic Aeromonas salmonicida. J. Fish Dis. 3: 517–524. Chopin FSM and Arimoto T. 1995. The condition of fish escaping from fishing gears—a review. Fish. Res. 21: 315–327. Chopin FS, Inoue Y and Arimoto T. 1996. Development of a catch mortality model. Fish. Res. 21: 315– 327. Davis MW. 2002. Key principles for understanding fish bycatch discard mortality. Can. J. Fish. Aquat. Sci. 59: 1834–1843. Davis MW. 2005. Behavior impairment in captured and released sablefish: ecological consequences and possible substitute measures for delayed discard mortality. J. Fish Biol. 66: 254–256. Davis MW and Olla, BL. 2001. Stress and delayed mortality induced in Pacific halibut Hippoglossus stenolepis by exposure to hooking, net towing, elevated sea water temperature and air: implications for management of bycatch. N. Am. J. Fish. Manag. 21: 725–732. Davis MW and Olla BL. 2002. Mortality of lingcod towed in a net is related to fish length, seawater temperature and air exposure: a laboratory bycatch study. N. Am. J. Fish. Manag. 22: 1095–1104. Davis MW and Parker SJ. 2004. Fish size and exposure to air: potential effects on behavioral impairment and mortality rates in discarded sablefish. N. Am. J. Fish. Manag. 24: 518–524. Davis MW and Ottmar ML. 2006. Wounding and reflex impairment may be predictors for mortality in discarded or escaped fish. Fish. Res. 82: 1– 6. Davis MW and Schreck CB. 2005. Responses by Pacific halibut to air exposure: lack of correspondence among plasma constituents and mortality. Trans. Am. Fish. Soc. 134: 991–998.

Davis MW, Olla BL and Schreck CB. 2001. Stress induced by hooking, net towing, elevated sea water temperature and air in sablefish: lack of concordance between mortality and physiological measures of stress. J. Fish Biol. 58: 1–15. DeAlteris JT and Reifsteck DM. 1993. Escapement and survival of fish from the cod-end of a demersal trawl. ICES J. Mar. Symp. 196: 128–131. De Veen JF, Huwae PHM and Lavaleye MSS. 1975. On discarding in the sole fishery and preliminary observations on survival rates of discarded plaice and sole in 1975. ICES CM 1975/F: 28. 9 pp. Domeier ML, Dewar H and Nasby-Lucas N. 2003. Mortality rate of striped marlin (Tetrapturus audax) caught with recreational tackle. Mar. Freshw. Res. 54: 435–445. Efanov SF. 1981. Herring in the Gulf of Riga: the problem of escapement and mechanical impact of the trawl. ICES CM. 1981/J: 7. 16 pp. Engås A, Isaksen B and Soldal AV. 1990. Simulated gear injuries on cod and haddock. ICES FTFB Working Group meeting, Rostock, April 23–25, 1990. 9 pp. Engås A, Jørgensen T and West CW. 1998. A speciesselective trawl for demersal gadoid fisheries. ICES J. Mar. Sci. 55: 835–845. Engås A, Foster D, Hataway BD, Watson JW and Workman I. 1999. The behavioral response of juvenile red snapper (Lutjanus campechanus) to shrimp trawls that utilize water flow modifications to induce escapement. Mar. Technol. Soc. J. 33: 43–50. Erickson D and Berkeley S. 2008. Methods to reduce bycatch mortality in longline fisheries. In: Pikitch EA and Camhi MD (eds). Sharks of the Open Ocean: Biology, Fisheries, and Conservation. pp 462–471. Oxford: Blackwell Publishing. Erickson D and Pikitch E. 1999. Survival of trawlcaught and discarded sablefish (Anoplopoma fimbria) during spring trials—1998 cruise report. National Marine Fisheries Service, Northwest Fisheries Science Center, Newport, OR, USA. Erickson D, Perez-Comas J, Pikitch E and Wallace J. 1996. Effects of catch size and cod-end type on the escapement of walleye Pollock (Theragra chalcogramma) from pelagic trawls. Fish. Res. 28: 179–196. Erickson D, Pikitch E, Methot R, Babcock E, Builder T, Wallace J, Crone P, Brodziak J and Conser R. 1997. Survival of trawl caught and discarded sablefish (Anoplopoma fimbria) off the U.S. West Coast. ICES CM 1997/FF:06. 19 pp.

Mortality of Animals that Escape Fishing Gears or Are Discarded after Capture Erickson D, Pikitch E, Suuronen P, Lehtonen E, Bublitz C, Klinkert C and Mitchell C. 1999. The great escape—selectivity and mortality of walleye pollock escaping from the cod-end and intermediate (extension) section of a pelagic trawl. Final report submitted to National Marine Fisheries Service, Alaska Fisheries Science Center. Anchorage, AK, USA, Alaska Fisheries Development Foundation. 68 pp. Farmer MJ, Brewer DT and Blaber SJM. 1998. Damage to selected fish species escaping from prawn trawl cod-ends: a comparison between square-mesh and diamond-mesh. Fish. Res. 38: 73–81. Farrell AP, Gallaugher PE and Routledge R. 2001. Rapid recovery of exhausted adult coho salmon after commercial capture by troll fishing. Can. J. Fish. Aquat. Sci. 58: 2319–2324. Glass CW and Wardle CS. 1989. Comparison of the reactions of fish to a trawl gear, at high and low light intensities. Fish. Res. 7: 249–266. Glass CW, Wardle CS, Godsen SJ and Racey DN. 1995. Studies on the use of visual stimuli to control fish escape from cod-ends. I. Laboratory studies on the effect of black tunnel on mesh penetration. Fish. Res. 23: 281–164. Grant SM. 2003. Mortality of snow crab discarded in Newfoundland and Labrador ’s trap fishery: At-sea experiments on the effect of drop height and air exposure duration. Can. Tech. Rep. Fish. Aquat. Sci. 2481: 28 pp. Halliday RG and Pinhorn AT. 2002. A review of the scientific and technical bases for policies on the capture of small fish in North Atlantic groundfish fisheries. Fish. Res. 57: 211–222. Hannah RW, Parker SJ and Buell Y. 2005. Evaluation of a selective flatfish trawl and diel variation in rockfish catchability as bycatch reduction tools in the deepwater complex fishery off the U.S. West Coast. N. Am. J. Fish. Manag. 25: 581–593. Harris RR and Ulmestrand M. 2004. Discarding Norway lobster (Nephrops norvegicus L.) through low salinity layers—mortality and damage seen in simulation experiments. ICES J. Mar. Sci. 61: 127–139. Hay DE, Cooke KD and Gissing CV. 1986. Experimental studies of Pacific herring gillnets. Fish. Res. 4: 191–211. He P. 1991. Swimming endurance of the Atlantic cod, Gadus morhua L. at low temperatures. Fish. Res. 12: 65–73. He P and Wardle CS. 1988. Endurance at intermediate swimming speeds of Atlantic mackerel, Scomber

289

scombrus L. herring, Clupea harengus L. and saithe, Pollachius virens L. J. Fish Biol. 33: 348–360. Hill BJ and Wassenberg TJ. 1990. Fate of discards from prawn trawlers in Torres Strait. Aust. J. Mar. Freshw. Res. 41: 53–74. Hueter RE, Manire CA and Tyminsky JP. 2006. Assessing mortality of released or discarded fish using a logistic model of relative survival derived from tagging data. Trans. Am. Fish. Soc. 135: 500–508. Huse I and Soldal AV. 2002. Mortality and injuries of haddock (Melanogrammus aeglefinus) that are caught by pelagic longline. ICES CM. 2002/V:31. 10 pp. ICES. 2000. Report of the FTFB Topic Group on Unaccounted Mortality in Fisheries. ICES FTFB Working Group, Haarlem, The Netherlands, April 10–11, 2000. Ingolfsson O, Soldal AV and Huse I. 2002. Mortality and injuries of haddock, cod and saithe escaping through cod-end meshes and sorting grids. ICES CM. 2002/V: 32. 22 pp. Jones JB. 1993. Net damage injuries to New Zealand hoki, Macruronus novaezelandiae. NZ J. Mar. Freshw. Res. 27: 23–30. Jónsson E. 1994. Scale damage and survival of haddock escaping through cod-end meshes (a tank experiment). ICES CM. 1994/B:16. Jørgensen T, Ingólfsson OA, Graham N and Isaksen B. 2006. Size selection of cod by rigid grids—Is anything gained compared with diamond mesh codends only? Fish. Res. 79: 337–348. Kaimmer S. 1994. Halibut injury and mortality associated with manual and automated removal from setline hooks. Fish. Res. 20: 165–179. Kaiser MJ and Spencer BE. 1995. Survival of bycatch from a beam trawl. Mar. Ecol. Prog. Ser. 126: 31–38. Kelle VW. 1976. Survival rate of undersized flatfish in the German shrimp fishery (mortality rate of discarded flatfish). Meeresforschung. 25: 77–89. Kelleher K. 2005. Discards in the world’s marine fisheries: an update. FAO Fish. Tech. Pap. 470: 134 pp. King SE, Hannah RW, Parker SJ, Matteson KM and Berkeley SA. 2004. Protecting rockfish through gear design: development of a selective flatfish trawl for the U.S. west coast bottom trawl fishery. Can. J. Fish. Aquat. Sci. 61: 487–496. Lancaster J and Frid C. 2002. The fate of discarded juvenile brown shrimps (Crangon crangon) in the Solway Firth UK fishery. Fish. Res. 58: 95–107.

290

Contemporary Issues in Capture and Conservation in Marine Fisheries

Lehtonen E, Tschernij V and Suuronen P. 1998. An improved method for studying survival of fish that escape trawl-cod-end meshes. Fish. Res. 38: 303–306. Lindeboom HJ and de Groot SJ (eds). 1998. The Effects of Different Types of Fisheries on the North Sea and Irish Sea Benthic Ecosystems. IMPACT II. NIOZ-Rapport 1998—1, RIVO-DLO Report C003/98. 404 pp. Lindley ST, Moser ML, Erickson DL, Belchik M, Rechisky E, Kelly JT, Heublein J and Klimley AP. 2008. Marine migration of North American green sturgeon. Trans. Am. Fish. Soc. 137: 182–194. Lockwood SJ, Pawson MG and Eaton DR. 1983. The effect of crowding on mackerel (Scomber scombrus L.)—physical condition and mortality. Fish. Res. 2: 129–147. Lowry N, Sangster GI and Breen M. 1996. Cod-end selectivity and fishing mortality. Final Report of the European Commission Study Contract No. 1994/005. Brussels, EC. Main J and Sangster GI. 1981. A study of the fish capture process in a bottom trawl by direct observations from a towed underwater vehicle. Scot. Fish. Res. Rep. 23: 23 pp. Main J and Sangster GI. 1982. A study of a multi-level bottom trawl for species separation using direct observation techniques. Scot. Fish. Res. Rep. 26: 16 pp. Main J and Sangster GI. 1990. An assessment of the scale damage to and survival rates of young fish escaping from the cod-end of a demersal trawl. Scot. Fish. Res. Rep. 46: 28 pp. Main J and Sangster GI. 1991. Do fish escaping from cod-ends survive? Scot. Fish. Work. Pap. 18: 15 pp. Mandelman JW and Farrington MA. 2007. The estimated short-term discard mortality of a trawled elasmobranch, the spiny dogfish (Squalus acanthias). Fish. Res. 83: 238–245. Mellergaard S and Bagge O. 1998. Fishing gearinduced skin ulcerations in Baltic cod, Gadus morhua L. J. Fish Dis. 21: 205–213. Milliken H, Farrington M, Carr HA and Lent E. 1999. Survival of Atlantic cod (Gadus morhua) in the Northwest Atlantic longline fishery. Mar. Technol. Soc. J. 33: 19–24. Misund OA and Beltestad AK. 1995. Survival of herring after simulated net bursts and conventional storage in net pens. Fish. Res. 22: 293–297. Misund OA and Beltestad AK. 2000. Survival of mackerel and saithe that escapes through sorting grids in purse seines. Fish. Res. 48: 31–41.

Neilson JD, Waiwood KG and Smith SJ. 1989. Survival of Atlantic halibut (Hippoglossus hippoglossus) caught by longline and otter trawl gear. Can. J. Fish. Aquat. Sci. 46: 887–897. Olla BL, Davis MW and Rose C. 2000. Differences in orientation and swimming of walleye pollock in a trawl under light and dark conditions: concordance between field and laboratory studies. Fish. Res. 44: 261–266. Olla BL, Davis MW and Schreck CB. 1997. Effects of simulated trawling on sablefish and walleye pollock: the role of light intensity, net velocity and towing duration. J. Fish Biol. 50: 1181–1194. Olla BL, Davis MW and Schreck CB. 1998. Temperature magnified postcapture mortality in adult sablefish after simulated trawling. J. Fish Biol. 53: 743–751. Özbilgin H and Glass CW. 2004. Role of learning in mesh penetration behaviour of haddock (Melanogrammus aeglefinus). ICES J. Mar. Sci. 61: 1190–1194. Özbilgin H and Wardle CS. 2002. Effect of seasonal temperature changes on the escape behaviour of haddock, Melanogrammus aeglefinus, from the codend. Fish. Res. 58: 323–331. Parker SJ, McElderry HI, Rankin PS and Hannah RW. 2006. Buoyancy regulation and barotraumas in two species of nearshore rockfish. Trans. Am. Fish. Soc. 135: 1213–1223. Parker SJ, Rankin PS, Hannah RW and Schreck CB. 2003. Discard mortality of trawl-caught lingcod in relation to tow duration and time on deck. N. Am. J. Fish. Manag. 23: 530–542. Pepperell JG and Davis TLO. 1999. Post-release behavior of black marlin Makaira indica caught and released using sport fishing gear off the Great Barrier Reef (Australia). Mar. Biol. 135: 369–380. Pikitch EK, Erickson D, Oddson G, Wallace J and Babcock E. 1996. Mortality of trawl caught and discarded Pacific halibut (Hippoglossus stenolepis). ICES CM. 1996/B:16. Pikitch E, Erickson D, Suuronen P, Lehtonen E, Rose C and Bublitz C. 2002. Selectivity and mortality of walleye pollock escaping from the cod-end and intermediate (extension) section of a pelagic trawl. ICES CM. 2002/V: 15. 29 pp. Roberts RJ. 1989. Fish Pathology. 2nd ed. London: Bailliere Tindall. 467 pp. Robinson WE, Carr HA and Harris J. 1993. Juvenile bycatch and cod-end escape survivability in the Northeast US groundfish fishery—second year

Mortality of Animals that Escape Fishing Gears or Are Discarded after Capture study. A report of the New England Aquarium to NOAA. NOAA Award No. NA26FD0039–01. Ryer CF. 2002. Trawl stress and escapee vulnerability to predation in juvenile walleye pollock: Is there an unobserved bycatch of behaviorally impaired escapees? Mar. Ecol. Prog. Ser. 232: 269–279. Ryer CF, Ottmar ML and Sturm EA. 2004. Behavioral impairment after escape from trawl cod-ends may not be limited to fragile fish species. Fish. Res. 66: 261–269. Sangster GI, Lehmann KM and Breen M. 1996. Commercial fishing experiments to assess the survival of haddock and whiting after escape from four sizes of diamond mesh cod-ends. Fish. Res. 25: 323–346. Sasso CR and Epperly SP. 2006. Seasonal sea turtle mortality risk from forced submergence in bottom trawls. Fish. Res. 81: 86–88. Siira A, Suuronen P, Ikonen E and Erkinaro J. 2006. Survival of Atlantic salmon captured in and released from commercial trap-net: potential for selective harvesting of stocked salmon. Fish. Res. 80: 280–294. Smith EM and Howell PT. 1987. The effects of bottom trawling on American lobsters, Homarus americanus, in Long Island Sound. Fish. Bull. 85: 737– 744. Soldal AV, Engås A and Isaksen B. 1989. Simulated net injuries on saithe. ICES Fish Capture Committee, (FTFB) Working Group Meeting, Dublin, April 24–25, 1989. 8 pp. Soldal AV and Isaksen B. 1993. Survival of cod (Gadus morhua) and haddock (Melanogrammus aeglefinus) escaping from a Danish seine at the sea surface. ICES Fish Capture Committee, FTFB Working Group Meeting, Gothenburg, April 19–20, 1993. 8 pp. Soldal AV, Isaksen B and Engås A. 1993. Survival of gadoids that escape from demersal trawl. ICES J. Mar. Symp. 196: 122–127. Soldal AV, Isaksen B, Marteinsson JE and Engås A. 1991. Scale damage and survival of cod and haddock escaping from a demersal trawl. ICES CM. 1991/B:44. 7 pp. Steven BG. 1990. Survival of king and Tanner crabs captured by commercial sole trawls. Fish. Bull. 88: 731–744. Suuronen P. 2005. Mortality of fish escaping trawl gears. FAO Fish. Tech. Pap. 478: 72 pp. Suuronen P and Millar RB. 1992. Size-selectivity of diamond and square mesh cod-ends in pelagic herring trawls: only small herring will notice the

291

difference. Can. J. Fish. Aquat. Sci. 49: 2109– 2117. Suuronen P, Erickson D and Orrensalo A. 1996b. Mortality of herring escaping from pelagic trawl cod-ends. Fish. Res. 25: 305–321. Suuronen P, Erickson D and Pikitch E. 1997. Meshsize management in pelagic trawl fisheries—potential solutions. In: Hancock DA, Smith DC, Grant A and Beumer JP (eds). Proceedings of the 2nd World Fisheries Congress: Developing and Sustaining World Fisheries Resources: The State of Science and Management 1996. pp. 563–567. Queensland, Australia: CSIRO Publishing. Suuronen P, Lehtonen E and Jounela P. 2005. Escape mortality of trawl-caught Baltic cod (Gadus morhua)—the effect of water temperature, fish size and cod-end catch. Fish. Res. 71: 151–163. Suuronen P, Lehtonen E, Tschernij V and Larsson PO. 1996a. Skin injury and mortality of Baltic cod escaping from two cod-ends equipped with exit windows. Arch. Fish. Mar. Res. 44: 165–178. Suuronen P, Perez-Comas JA, Lehtonen E and Tschernij V. 1996c. Size-related mortality of herring (Clupea harengus L.) escaping through a rigid sorting grid and trawl cod-end meshes. ICES J. Mar. Sci. 53: 691–700. Suuronen P, Turunen T, Kiviniemi M and Karjalainen J. 1995. Survival of vendace (Coregonus albula L.) escaping from a trawl cod-end. Can. J. Fish. Aquat. Sci. 52: 2527–2533. Thompson RB and Hunter CJ. 1973. Viability of adult sockeye salmon that disentangle from gillnets. Int. North Pac. Fish. Comm. Ann. Rep. 1971: 107– 109. Thompson RB, Hunter CJ and Pattern BG. 1971. Studies of live and dead salmon that unmesh from gillnets. Int. North Pac. Fish. Comm. Ann. Rep. 1969: 108–112. Thurow F and Bohl H. 1976. On survival rates of Baltic cod after capture. ICES CM. 1996/P:4. 11 pp. Treschev AI, Efanov SF, Shevtsov SE and Klavsons UA. 1975. Die verletzung und die überlebenschancen des Ostseeherings nach dem durchdringen durch die Maschen des Schleppnetzsteerts. Fischerei Forschung, 13: 55–59. Trumble RJ, Kaimmer SM and Williams GH. 2000. Estimation of discard mortality rates for Pacific halibut bycatch in groundfish longline fisheries. N. Am. J. Fish. Manag. 20: 931–939. Tschernij V and Suuronen P. 2002. Improving trawl selectivity in the Baltic—Utökning av trålselektion i Östersjön. TemaNord 2002, 512. 56 pp.

292

Contemporary Issues in Capture and Conservation in Marine Fisheries

Turunen T, Käkelä A and Hyvärinen H. 1994. Trawling stress and mortality in undersized (<40 cm) brown trout (Salmo trutta L). Fish. Res. 19: 51–64. Turunen T, Suuronen P, Hyvärinen H and Rouvinen J. 1996. Physiological status of vendace (Coregonus albula L.) escaping from a trawl cod-end. Nordic J. Freshw. Res. 72: 39–44. Valdemarsen JW and Suuronen P. 2003. Modifying fishing gear to achieve ecosystem objectives. In: Sinclair M and Valdimarsson G (eds). Responsible Fisheries in the Marine Ecosystem. pp 321–341. FAO and CABI International Publishing. 426 pp. Van Beek FA, Van Leeuwen PI and Rijnsdorp AD. 1989. On the survival of plaice and sole discards in the otter-trawl and beam-trawl fisheries in the North Sea. Netherlands J. Sea Res. 26: 151–160. Vander Haegen GE, Ashbrook CE, Yi KW and Dixon JF. 2004. Survival of spring chinook salmon cap-

tured and released in a selective commercial fishery using gill nets and tangle nets. Fish. Res. 68: 123–133. Walsh SJ and Hickey WM. 1993. Behavioural reactions of demersal fish to bottom trawls at various light conditions. ICES J. Mar. Symp. 196: 68–76. Wardle CS. 1993. Fish behavior and fishing gear. In: Pitcher TJ (ed). The Behavior of Teleost Fishes. 2nd ed. pp 606–641. London: Chapman and Hall. Wileman DA, Sangster GI, Breen M, Ulmestrand M, Soldal AV and Harris RR. 1999. Roundfish and Nephrops survival after escape from commercial fishing gear. Final report. EC Contract FAIR-CT950753. Brussels, EC. Xu G, Arimoto T and Inoue Y. 1993. The measurement of muscle fatigue in walleye pollock (Theragra chalcogramma) captured by trawl. ICES J. Mar. Symp. 196: 117–121.

Mortality of Animals that Escape Fishing Gears or Are Discarded after Capture SPECIES MENTIONED IN THE TEXT Alaskan pollock, Theragra chalcogramma American lobster, Homarus americanus American plaice, Hippoglossoides platessoides Atlantic cod, Gadus morhua Atlantic halibut, Hippoglossus hippoglossus Atlantic herring, Clupea harengus Atlantic salmon, Salmo salar black rockfish, Sebastes melanops brown shrimp, Crangon crangon brown trout, Salmo trutta Baltic cod, Gadus morhua Baltic herring, Clupea harengus China rockfish, Sebastes nebulosus chinook salmon, Oncorhynchus tshawytscha coho salmon, Oncorhynchus kisutch dab, Limanda limanda green sturgeon, Acipenser medirostris haddock, Melanogrammus aeglefinus hake, Merluccius merluccius Hauraki Gulf snapper, Pagrus auratus hoki, Macruronus novaezelandia king crab, Paralithodes camtschaticus lingcod, Ophiodon elongatus mackerel, Scomber scombrus Northeast Arctic cod, Gadus morhua Norway lobster, Nephrops norwegicus Pacific halibut, Hippoglossus stenolepis Pacific herring, Clupea pallasii

293

Patagonian scallop, Zygochlamys patagonica pearly finned cardinal-fish, Apogon poecilopterus perforated-scale sardine, white sardinella, Sardinella albella pike-perch, Sander lucioperca; Stizostedion lucioperca plaice, Pleuronectes platessa red mullet, Mullus barbatus sablefish, Anoplopoma fimbria saithe, Pollachius virens sand whiting, Sillago ciliata scallop, Pecten maximus sea bream, Pagrus major shrimp; pink shrimp, Pandalus borealis snow crab, Chionoecetes opilio spiny dogfish, Squalus acanthias sole, Solea solea spider crab, Leptomithrax gaimardi striped marlin, Tetrapturus audax sunrise goatfish, Upeneus sulphureus swordfish, Xiphias gladias tanner crab, Chionoecetes bairdi vendace, Coregonus albula whiting, Merlangius merlangus winter flounder, Pseudopleuronectes americanus yellowfin bream, Acanthopagrus australis yellowtail flounder, Plueronectes ferruginea; Limanda ferruginea

Chapter 12 Effect of Trawling on the Seabed and Mitigation Measures to Reduce Impact Pingguo He and Paul D. Winger

12.2 REVIEW OF RECENT STUDIES ON THE SEABED IMPACT OF TRAWLING Our knowledge of how trawling activities affect benthic communities is still in its infancy, although numerous studies on seabed impacts of trawling have been carried out during the last decade (Auster and Langton 1999; Collie et al. 2000; Jennings and Kaiser 1998; Løkkeborg 2005). The lack of progress is mainly due to the complexity of conducting such studies, as benthic communities show large natural temporal and spatial variability, and the potential impacts of trawling are likely to be affected by many factors such as disturbance regime, bottom type, natural disturbance, and benthic assemblage. Because of this complexity, and the methodological limitations of most impact studies, the results from individual studies should be interpreted with great caution (Løkkeborg 2005). Otter trawls and beam trawls are likely to have different physical impacts on the seabed because of their different designs and catching principles. The most noticeable physical effect of otter trawling is caused by the trawl doors, which have been shown to create furrows of up to 20 cm deep depending on their weight and hardness of the sediment (Humborstad et al. 2004; Krost et al. 1990; Schwinghamer et al. 1998; Tuck et al. 1998). Beam trawling, on the other hand, causes a flattening of the bottom topography by eliminating natural irregular features such as ripples and bioturbation

12.1 INTRODUCTION Trawls operating on the seabed (bottom trawls) create tracks and other physical changes. The possible impact of bottom trawling on the seabed and demersal fisheries resources has been recognized since the otter trawl was invented in the 1880s (Wardle 1986). Scientific investigations have continued to this day to answer the same questions: Does trawling affect demersal fisheries resources beyond the catch that trawls take, and the ecosystem that supports commercial fishing? Technological solutions to lessen possible impacts have been pursued in line with precautionary approaches to fisheries and ecosystem management. This chapter will discuss the progress that has been made in gear design and operation to reduce seabed impact in otter trawls and beam trawls. In the last decade, intense discussions and heated debates about the impact of mobile gears, including trawls, on the seabed arose after declines of major commercial groundfish stocks in the Northwest Atlantic. As a result, several major reviews, books, symposia, and comprehensive studies have been completed, attempting to clarify the impact of mobile gears on the seabed and marine habitats (Barnes and Thomas 2005; Dorsey and Pederson 1998; Glass et al. 2007; Hall 1999; He 2007; Johnson 2002; Kaiser and de Groot 2000; Linnane et al. 2000; Løkkeborg 2005; NRC 2002; Rice 2006; Sinclair and Valdimarsson 2003).

295

296

Contemporary Issues in Capture and Conservation in Marine Fisheries

mounds (Fonteyne 2000; Kaiser and Spencer 1996). Severe biological impacts of otter trawling have been demonstrated on hard bottoms dominated by large erect sessile fauna, causing considerable decreases in the abundance of attached epifauna of higher vertical profiles such as sponges and corals (Freese et al. 1999; Moran and Stephenson 2000; van Dolah et al. 1987). On the other hand, trawling on sandy bottoms could cause declines in some invertebrate species but, according to some studies, did not produce large and enduring changes in the benthic communities (Kenchington et al. 2001; Kutti et al. 2005; Prena et al. 1999). Potential impacts of trawling on soft bottoms may be masked by the more pronounced temporal variability in these habitats and therefore are difficult to demonstrate over longer terms. Thus, although impacts of trawling on soft bottoms have been indicated in some studies, consistent and unambiguous effects attributable to disturbances due to otter trawling have not been detected (Drabsch et al. 2001; Hansson et al. 2000; SparksMcConkey and Watling 2001; van Dolah et al. 1991). Clear evidence of short-term effects of beam trawling, however, have been provided by several studies demonstrating that intensive beam trawling disturbances caused considerable decreases in the abundance of several benthic species (Bergman and Hup 1992; Bergman and van Santbrink 2000; Kaiser and Spencer 1996). 12.3 DESCRIPTION OF TRAWLS AND THEIR OPERATION The otter trawl was developed from the beam trawl in the 1880s, with significant technological developments occurring during the Industrial Revolution and after World War II (Gabriel et al. 2005; Graham 2006; Wardle 1986). An otter trawl consists of a pair of warps connecting the vessel and the doors, two trawl doors, a set of cable assemblies connecting the doors and the trawl (bridles), and the trawl net (Fig. 12.1A). The doors are also called otter boards. The trawl net is typically composed of the wings, the square, the body, the extension, and the codend. Floats are attached to the headline to open the trawl vertically, while the groundgear attaches to the fishing line to allow the net to tend the bottom

and prevent damage to the netting. The type of groundgear installed varies depending on the fishing ground conditions and species targeted. They can be light wires, chains, small and large rubber discs, wheels or spherical bobbins, or a combination of them (Fig. 12.1B). Trawls that operate on rough seabed or in strong currents are usually rigged with larger discs and bobbins. In most types of groundgears, only a few rollers or bobbins near the center of the trawl can roll freely; those on the quarters and wings cannot roll as their axes are not perpendicular to the towing direction. Behind the groundgear, the body of the net tapers so that the lower belly of the net is kept away from the seabed to avoid damage. The codend usually does not contact the seabed, but large and heavy catches may bring the codend down. Therefore, chafing gear is typically used to protect the codend from abrasion with the substrate. Otter trawl designs can be divided into a few helpful categories according to their association with the seabed (Fig. 12.2). Pelagic trawls are used to target fish in the water column and usually do not contact the seabed. Bottom trawls by comparison are towed along the seabed with the doors and groundgear in constant contact with the seabed. Hybrids of these two technologies are semipelagic trawls. They are among the newest designs and usually have either the doors or the groundgear in contact with the seabed but not both. Beam trawls are used in many parts of the world for harvesting flatfish and shellfish. The beam trawl uses a beam to spread the trawl horizontally instead of trawl doors (Fig. 12.3). Beams have been made of various materials, from wood and bamboo in early days to steel and aluminum in the present day. The beams vary in length depending on the size of the vessel and the design of the trawl, usually ranging between 4 and 12 m (Valdemarsen and Suuronen 2003). A pair of shoes (also called heads) supports the beam and keeps it off the bottom to avoid damage. The wingends of the trawl connect to the shoes through a set of short bridles. Large wheels are used instead of shoes in some beam trawls to help the gear roll on the seabed (Gabriel et al. 2005). Chains or a chain matrix are commonly used in flatfish beam trawls in the North Sea to stimulate bottom-dwelling species such as soles

Figure 12.1. (A, B) Components of an otter trawl system and (C) common types of groundgear.

297

298

Contemporary Issues in Capture and Conservation in Marine Fisheries

Figure 12.2. Trawling styles with reference to their interaction with the seabed. (Redrawn from He 2007; reprinted with kind permission of Springer Science and Business Media.)

(van Marlen 2000). Electric stimulating devices were used in Chinese shrimp beam trawls in the 1990s but were banned in 2001 (Yu et al. 2007). The headline of the trawl is fastened to the beam. The net itself resembles an otter trawl net. Multiple codends are used in Chinese shrimp beam trawls to improve quality of catch and ease of operation (Yu et al. 2007). Beam trawls weigh from a few hundred kilograms to several tons. In the North Sea flatfish fisheries, beam trawls are usually towed at high

speeds of up to 7 knots (Valdemarsen and Suuronen 2003). Of all fishing systems used worldwide, trawls are probably the most complex fishing gear. They have evolved steadily over the years for many fisheries, but basic designs have changed little. Continuously progressing through an endless cycle of fishing gear development (Winger et al. 2006), significant improvements have been made in fuel efficiency, selectivity and the degree of impact on the environ-

Effect of Trawling on the Seabed and Mitigation Measures to Reduce Impact

299

Figure 12.3. Schematic illustration of a beam trawl (A) and a photograph (B) of a similar beam trawl. (Graph and photograph courtesy Jochen Depestele © ILVO.)

ment. The following sections provide a review of mitigation measures, both experimental and commercially available, that help reduce the impact of trawls on the seabed. 12.4 PELAGIC AND SEMIPELAGIC TRAWLS Pelagic trawls and bottom trawls are different in their design and purpose (Fig. 12.2). Not surprising, two recent studies ranked them near polar opposites with regard to their ecological impact (Fuller et al.

2008; Morgan and Chuenpagdee 2003). The goal for many gear technologists has been to redesign bottom trawls to fly pelagically or at least semipelagically while at the same time maintaining traditional performance targets (i.e., efficiency and selectivity). This section provides an overview of developments in this area. 12.4.1 Alaska Pollock Trawls Alaskan pollock (Theragra chalcogramma) were harvested by bottom trawls by U.S. domestic

300

Contemporary Issues in Capture and Conservation in Marine Fisheries

vessels in the 1980s and 90s, although vessels from the former USSR may have exploited the resource using pelagic and bottom trawls much earlier. In the U.S. fishery, bycatch of shellfish (mainly crabs) and other groundfish (mainly Pacific halibut) by these trawls, although small in percentage, has been a concern for the North Pacific Fisheries Management Council (NPFMC). To reduce bycatch of these species closely associated with the bottom, NPFMC implemented a measure to encourage the use of pelagic trawls by allocating a large proportion of the pollock total allowable catch (TAC) to the pelagic sector. The NPFMC set up a performance standard for pelagic trawls by regulating the maximum number of crabs caught by a trawl (Pereyra 1995). Onboard observers determine whether the trawl is operating in pelagic or nonpelagic mode by the number of crabs in the catch, and assign the catch against the appropriate quota. However, the crab performance standard does not ensure that pelagic trawls are fished without substantial bottom contact. In fact, the 2005 Essential Fish Habitat Environmental Impact Statement, based on numbers provided by the pollock industry, estimated that 44% of the area coverage of pelagic trawls in the Bering Sea was in contact with the seafloor (NOAA 2005). Because the nonpelagic trawl was allocated only a small portion of the TAC, the industry soon adopted the pelagic method to harvest pollock. Ultimately, with industry support, the NPFMC eliminated bottom trawling in the Bering Sea pollock fishery in 1999 (NRC 2002). While the original concern was bycatch of shellfish and groundfish, the resulting pelagic trawling operations for pollock may have reduced the contact of some trawl components with the seabed. 12.4.2 Semipelagic Trawling for Pink Shrimp Shrimp trawls are generally similar in design to groundfish trawls except for differences in mesh size and bycatch reduction devices. In both cases, the shearing force of the doors with the seabed produces sand clouds that herd fish into the net path (see Chapter 4). But many species of shrimp, including the pink shrimp (Pandalus borealis), have poor swimming capability and are unable to react to sand clouds and fast moving trawl compo-

nents. Hence, most of the shrimp caught are those that lie directly within the net path. For these fisheries, the sand clouds produced by the doors serve little purpose and only increase finfish bycatch. This has led to the development of semi-pelagic fishing systems that operate with the doors off the bottom (Fig. 12.2C). Projects to test the feasibility of semipelagic shrimp trawls with doors off the bottom have been carried out in the inshore waters of Newfoundland (DeLouche and Legge 2004) and the Gulf of Maine (He et al. 2006). In both studies, high-aspect trawl doors were used to spread the trawl through hydrodynamic forces alone. The primary control of the door height off the seabed was achieved through the shortening of warps and was monitored in real-time through the use of acoustic door height monitoring devices. The studies revealed that lifting the doors off the bottom had no significant effect on the size range or catch per unit effort of target shrimp compared with traditional shrimp trawls fishing on the same grounds. While the results are encouraging, the technique requires active adjustment of the warp length and real-time monitoring of the door height. For this reason, it may be better suited for vessels where the winch power is isolated from the main engine and those with acoustic height monitoring equipment already installed.

12.4.3 Semipelagic Trawling Using French-Rigging Lifting the groundgear off the seabed can be achieved by attaching the top bridles directly onto the main warps, forward of the doors. This is commonly referred to as French-rigging or fork-rigging (Fig. 12.2D). If the groundgear is sufficiently lightweight, transferring this tension forward to the warps can effectively lift the groundgear off the seabed, while at the same time maintaining stable contact between the doors and the seabed. The technique was initially developed for the purpose of targeting fish higher in the water column or for avoiding uneven bottom that frequently cause gear damage (see Garner 1978 for detailed drawings). It has been successfully used on the west coast of Newfoundland to reduce the bycatch of Atlantic cod while targeting redfish (B. Johnson, Department

Effect of Trawling on the Seabed and Mitigation Measures to Reduce Impact of Fisheries and Aquaculture, NL, Canada, personal communication) as well as by French and U.K. Channel Islands fishermen targeting sea bass, black sea bream, and John Dory (K. Arkley, SEAFISH, personal communication). Most recently, the technique has been revisited because of its seabed-friendly qualities. Lifting the groundgear off the seabed significantly reduces the footprint of the fishing system, making it a variation of the semipelagic trawl as shown in Figure 12.2D. Preliminary tests have been conducted for pink shrimp in Newfoundland (G. Brothers, unpublished data) as well as for various inshore groundfish species in southwestern England (Arkley 2006) with encouraging results. 12.5 GROUNDGEAR MODIFICATIONS Conventional groundgear provides weight, stability, and protection to a fishing system. By their design, they help the net of an otter trawl tend bottom and catch targeted species that live in close proximity to the seabed. The downward footprint is often characterized by multiple contact points that rub or roll over the seabed. Several investigations have examined various mitigation measures to minimize the area and depth of this footprint. This section provides an overview of developments in this area.

301

12.5.1 Bottom Trawling with Drop Chains and Weights Red snappers (Lutjanus malabaricus and L. erythropterus) are harvested by bottom trawls in northern Australia. An experiment was carried out with a modified trawling system to reduce high discards and seabed impact (Fig. 12.4) (Brewer et al. 1996; Ramm et al. 1993). While the doors of the system operate on the seabed, the groundgear was replaced with a series of drop chains and weights, significantly lightening the trawl and reducing its footprint on the seabed. Initial sea trials by Ramm et al. (1993) used seven drop chains (0.5 m long, 10 kg each) suspended in the bosom and a heavier weight (60 kg) on each of the wingends. This rigging lifted the fishing line 0.3 m off the bottom. Catch rates for commercial species were similar compared with traditional demersal trawls, while the catch of nontarget species was reduced. The trawling system left only nine furrows 0.1 to 0.3 m wide, totaling about 2 m in width out of a 65-m trawl path between the doors. Because the design was sensitive to changes in fishing and operating conditions, a modified design was developed and tested by Brewer et al. (1996). Comparative fishing against the traditional groundgear demonstrated no reduction in the catch of commercial species while discard species and

Figure 12.4. A bottom trawl with drop chain and weights tested in the Australia red snapper fishery. (He 2007; reprinted with kind permission of Springer Science and Business Media.)

302

Contemporary Issues in Capture and Conservation in Marine Fisheries

benthos were substantially reduced. The fishing line of this trawl had a slightly greater height off bottom (0.4–0.5 m) than its predecessor, and due to its improved performance it was subsequently recommended for use in Australia’s Northern fish trawl fishery.

12.5.2 Raised Footrope Trawls and Sweepless Trawls The trawl groundgear is called the “sweep” in the northeastern United States; a “sweepless” trawl is therefore a trawl without groundgear. The sweepless trawl is a modification of the earlier “raised footrope” trawl design that was developed for the Gulf of Maine silver hake (Merluccius bilinearis) fishery (Pol 2003). In the raised footrope trawl, a horizontal sweep chain is hung from the fishing line by a number of drop chains 1 m in length. The sweepless trawl (Fig. 12.2E), by comparison, has no horizontal chain sweep or other groundgear assembly; only the ends of the drop chains are in contact with the seabed. To increase the weight of the drop chain, the link size can be increased or two chains of the same size can be hung at each attachment point. The height of the fishing line off the seabed can be adjusted by adjusting the relative length of the upper and lower bridles and by the length and size of the drop chains. Experimental fishing of the earlier raised footrope trawl design resulted in a substantial reduction in the bycatch of flatfish and lobster when the fishing line was rigged 1 m off the seabed. The sweepless trawl represents several improvements over this design, including less chain, fewer attachments, and easier rigging. Also, the simpler groundgear means it is less likely to become entangled with debris. The sweepless trawl has less impact on the sea floor because the contact with the seabed is reduced to a limited number of points instead of the whole width between the wingends (Pol 2003). In the Gulf of Maine, the sweepless trawl is required in two whiting fisheries: on the Massachusetts coast, as an alternative to the raised footrope trawl; and on the Maine coast, it must be used in combination with a Nordmøre-style grate to exclude large groundfish species such as Atlantic cod (Gadus morhua).

12.5.3 Lighter Groundgear for the Offshore Shrimp Trawls He (2001) reported on a project to reduce the seabed impact of offshore shrimp trawls off Labrador. The project investigated whether seabed contact by the existing offshore shrimp trawl could be reduced through a reduction of the number of footgear bobbins without significantly altering the engineering and catch performance of the gear. The fishing gear tested was a three-bridle Skjervøy 3600 shrimp trawl with 31 bobbins. The full groundgear weighed 5698 kg in air and 2984 kg in water. The modified nine-bobbin footgear weighed 2187 kg in air and 1306 kg in water. The total area of seabed in contact with the trawl was calculated from the width and the number of bobbins in the groundgear. The total area affected by the bobbins was reduced by 69% when the number of bobbins was reduced from 31 to 9. Flume tank experiments tested a progressive reduction of the number of bobbins from 31 to 9 and total removal of the bobbins (drop chain only) as shown in Figure 12.5. The project also included sea trials, which were conducted off the Labrador coast in the Northwest Atlantic, targeting pink shrimp. The catch rates were similar between the control gear with 31 bobbins and the experimental gear with 9 to 19 bobbins. Under good sea and ground conditions, the 9-bobbin rigging provided sufficient weight to keep the groundgear steadily on the seabed. However, in adverse sea and ground conditions, the lightweight experimental gear resulted in poor seabed contact and occasional gear damage. The rigging with drop chains only was not tested at sea. This experiment showed that the number of bobbins on the trawl might be reduced to as few as 9 without significantly altering the geometry or stability of the trawl under favorable sea and fishing ground conditions. 12.5.4 Use of Rollers and Wheels in the Groundgear The attempt to design a groundgear that can roll over the seabed came from the need to save fuel during fishing. In the 1940s, German engineers patented a design of wheeled groundgear. In the design, all of its wheels can roll over the seabed. In 1993, a German shrimp fisherman constructed and tested

Effect of Trawling on the Seabed and Mitigation Measures to Reduce Impact

303

Figure 12.5. Bosum part of Skjervoy 3600, a commercial offshore shrimp trawl in a flume tank, with (A) 31 bobbins, (B) nine bobbins, and (C) no bobbins (drop chains only). (From He 2007; reprinted with kind permission of Springer Science and Business Media.)

a wheeled groundgear based on the design. There were no reductions in commercial catch but less catch of sand and shells (Gabriel et al. 2005). The wheel gear had probably less drag and bottom impact due to smaller quantity of substrate materials caught in the net. Similar groundgear with turning wheels is being pursued in the northeastern U.S. silver hake and shrimp fisheries to reduce seabed impact and to save fuel.

Irish researchers (Ball et al. 1999, 2003) tested a Nephrops trawl using rollerballs on its groundgear to reduce seabed impact and a dropout panel to reduce benthos catch. Fourteen large rollers (4 kg each) were used on each wing and six smaller rollers (2 kg each) at the mid-section of the groundgear. Sea trials compared this rollerball net with a commercial Nephrops net on the west coast of Ireland. Catches of commercial species were

304

Contemporary Issues in Capture and Conservation in Marine Fisheries

similar, while bycatch of invertebrates and debris were reduced by 32% and 66%, respectively. Because of reduced ground friction, the experimental net required 12% less power than the commercial net when towed at the same speed of 2.6 knots. The catch from the rollerball net was cleaner with less silt around fish gills, indicating less penetration into the seabed (Ball et al. 2003). Researchers in the Faeroe Islands tested swiveled rollers and wheels to replace rockhoppers typically used in their trawls (Zachariassen 2004). Among several configurations tested, the most promising rolling gear consisted of 0.22-m-wide rubber discs with steel axles. There was a combination of small discs and rollers between the wheels. Each wheel can rotate independently, while maintaining orientation in the direction of tow. 12.5.5 Nylon Tickler Brushes as Groundgear Tickler chains are commonly used in flatfish fisheries in the northeastern Atlantic. To replace tickler

chains, rolling tickler brushes were tested by Faeroe researchers to reduce the seabed impacts of trawling (Zachariassen 2000). Brushes were made of nylon and were cylindrical in shape. Experimental fishing indicated similar catch rates for target species. Underwater video observations showed noticeable reduction in suspended sediments, indicating less seabed disturbance by the brush gear. However, a similar groundgear device, the “street sweeper,” was assessed and subsequently banned in the northeastern U.S. groundfishery in the 1990s due to its high fishing efficiency (Pol and Carr 2000). 12.5.6 Soft Chain Brush Groundgear Sterling and Eayrs (2006) tested a “soft brush” footgear to replace the traditional tickler chains in the Australian prawn fisheries. The fisheries target king prawns (Melicertus latisulcatus) and tiger prawns (Penaeus esculentus). The groundgear is composed of drop chains without horizontal tickler chains (Fig. 12.6) as chains of this nature are known to

Figure 12.6. (A) Traditional chain. (B) “Soft brush” foot gear tested in the Australian prawn fisheries. (Modified from Sterling and Eayrs 2006.)

Effect of Trawling on the Seabed and Mitigation Measures to Reduce Impact damage epifauna in some regions (Rose et al. 2000). Field experiments indicate that the catch of target prawn species was reduced by about 12%. Evaluation of the design on the reduction in the impact of the gear on nontarget benthic species and on the increase in the target species is continuing. 12.5.7 Use of Depressors and Plates in Groundgear Waterborne kites or other flexible devices have been investigated to reduce the weight of fishing gear and/or to eliminate the use or to reduce the size of trawl doors (Goudey 1999). A band of fabric panels was installed between the fishing line and the footgear. The fabric panel generated downward forces to keep the gear on the bottom, and it could reduce or replace heavyweight rollers and chains. The idea showed promise in flume tank tests but sea trials did not continue partly due to the fear of damaging the soft materials used if they contacted the rough seabed. Danish and Norwegian researchers developed a gear with a series of rubber plates hanging under the fishing line, called “self-spreading” groundgear or “plate gear” (SINTEF 2004). Increased wingspread (15% to 20%) was measured with the new gear compared with a typical rockhopper gear in both flume tank and field tests. The increase in spreading force from the groundgear would allow a reduction in door size, which in return would reduce the seabed impact of the doors. In addition, because the individual plates can flip horizontally in reaction to rocks and other obstructions, this gear appears to be less intrusive to the bottom, although more evidence is needed to evaluate the overall impact of the gear. Preliminary sea trials have shown that the gear has increased efficiency and reduced size selectivity for certain species, suggesting it might be an effective groundgear for sampling trawls rather than a commercial gear (Langeland 2005). 12.6 TRAWL DOOR CONSIDERATIONS 12.6.1 The Use of Efficiency and High Aspect Ratio Doors Trawl doors provide horizontal spread, as well as help in the herding process of demersal fish species

305

into the net path (Main and Sangster 1981; Wardle 1986; also see Chapter 4). Early trawl doors had simple designs and relied heavily on ground shear to spread the trawl and would not function when they were off the bottom. Newer trawl doors are more complicated in design and rely primarily on hydrodynamic forces to spread the trawl. They usually have a higher aspect ratio (the ratio of height to width) than older designs and some remain stable both on and off the bottom. The use of high aspect pelagic trawl doors may also reduce seabed impact even if they are on the bottom. Efficient doors typically have a narrow width and operate at a lower angle of attack, leaving a narrower footprint compared with traditional bottom doors (Goudey and Loverich 1987). For example, a low aspect ratio door with 2-m2 area could be 2 m long × 1 m high, an aspect ratio of 0.5. A high aspect door of the same area could be 1 m long × 2 m high, an aspect ratio of 2. The low aspect ratio door usually operates at a large attack angle of about 43 degrees, while the high aspect ratio door usually operates at about 30 degrees. As seen in Figure 12.7, the width of the door track of the high aspect ratio door would only be 37% that of the low aspect ratio door. High aspect trawl doors also reduce the bottom contact of ground wires behind the door by keeping a large proportion of the wire off the seabed (Goudey and Loverich 1987). For trawls for which herding by sand clouds and bridles is not critical, such as shrimp trawls, the use of high aspect trawl doors may be feasible and could reduce seabed disturbance as well as save fuel. 12.6.2 The New “Batwing” Door Sterling and Eayrs (2006) recently designed and tested a “Batwing” door to reduce footprint and to increase spreading force of the door for Australian prawn fisheries. The new Batwing door is mounted on a base plate that is towed parallel to the towing direction. The tow track left by the new door is thus only the width of the base plate compared with the much wider regular rectangular door (Fig. 12.8). The new door was 13% lower in drag and provided 5.4% more spread for the same trawl (Sterling and Eayrs 2006). Initial field experiments indicated a 10% reduction in targeted tiger and king prawn catch despite the increase in door spread. Research

306

Figure 12.7. track.

Contemporary Issues in Capture and Conservation in Marine Fisheries

Comparison between low (A) and high (B) aspect trawl doors and their width of

on the door and how they affect the catch of targeted prawn species is continuing. 12.7 OTHER TRAWL GEAR COMPONENTS 12.7.1 Short Bridles Bridles connect the wingends to the doors (Fig. 12.1). Heavy bridles strung with rubber discs, called “cookies,” are used in flounder trawls in New

England to stir up sediment to increase herding efficiency. In trawls for which bridle herding is not important or undesirable, such as the shrimp trawl fishery, shorter and lighter bridles are used. In the Gulf of Maine pink shrimp fishery, regulations require that the wire between the wingend and the door does not exceed 27.5 m, and only bare wires are allowed for bridles. The primary intent of the regulation was to reduce herding and bycatch of finfish, especially flounder, as longer bridles usually

Effect of Trawling on the Seabed and Mitigation Measures to Reduce Impact

307

Figure 12.8. The new “Batwing door” (B) designed by Sterling and Eayrs (2006) and the track width with comparison to the regular flat door (A). (Modified from a photograph courtesy of David Sterling and Steve Eayrs.)

catch more flounder (Somerton and Munro 2001). Shorter and lighter bridles may help reduce the seabed impact of the trawl as they are less likely to be in constant contact with the seabed. 12.7.2 Devices to Keep Bridles off the Bottom Rose and his colleagues (2006) tested off-bottom bridles to reduce the effect of bridle on sessile animals in the Bering Sea flatfish trawl fisheries. The trawl gear used in these flatfish fisheries empha-

size the herding of cables (sweeps or bridles) and use long cables to achieve the objectives. It is estimated that 90% of the swept area between the doors is covered by the cable and only 10% is covered by the netting and groundgear between the wingends. Reducing the effect of the cables on the benthic animals was thus considered the most promising approach to reducing the impact of trawling in these fisheries (Rose et al. 2006). To raise the cable off the seabed and to reduce the “cutting” effect of the

308

Contemporary Issues in Capture and Conservation in Marine Fisheries

cable to sessile animals, Rose et al. (2006) tested disc clusters on the cables (Fig. 12.9). They tested 15-, 20-, and 25-cm-diameter discs and found that 20-cm discs installed 10 m apart had the best effect in reducing damage to the benthic invertebrates without a reduction in targeted flatfish catch. The 20-cm discs installed in combination ropes reduced damage to seawhips by 52% and basket stars by 12%. There was little or no significant reduction (less than 2%) in the catch of several species of sole and flounder. There was an increase in the catch of Pacific cod (Gadus macrocephalus) and Alaska pollock. These findings are in contrast to earlier beliefs that sweep cables needed to be on the seabed to herd flounders.

12.7.3 Use of a Small Warp/Depth Ratio Vincent (2001) discussed an option to reduce the weight or pressure of the trawl door on the seabed by changing warp/depth ratio. With the French survey gear 25/47 GOV trawl, it was demonstrated that a reduction in the warp/depth ratio resulted in a reduced downward force of the door. At 50-m depth, the warp ratio was reduced from the usual 5.60 to 3.30 and the downward force was reduced more than three-fold. Various reductions in warp ratios and corresponding downward force reductions were also demonstrated at other depths. However, trawl geometry may change due to a change in warp/depth ratio. Subsequent change on catch efficiency needs to be determined, especially for survey trawls.

Figure 12.9. Disc clusters tested by Rose and his colleagues (Rose et al. 2006) in the Bering Sea flatfish trawls to reduce the impact of cable to sessile invertebrates. Photo courtesy of Carwyn Hammond. (Drawing modified from Craig Rose.)

Effect of Trawling on the Seabed and Mitigation Measures to Reduce Impact 12.8 BEAM TRAWLS Beam trawls are used for harvesting flatfish and crustaceans in the North Sea and other parts of the world. Efforts to reduce the seabed impacts of beam trawling include the use of electric stimuli to replace chain mats in the Netherlands (van Marlen 2000; van Marlen et al. 2001, 2005b, 2006) and the raised footrope in shrimp beam trawls in Belgium (Polet et al. 2005a, 2005b). Dropout panels were also tested in beam trawls to reduce benthos catch in the Belgian and British flatfish beam trawls. 12.8.1 Use of Electrical Stimuli in Flatfish Beam Trawls Early investigations on the use of electricity as alternative stimuli were conducted in an effort to reduce the weight of the gear and to reduce drag and energy costs (Stewart 1975; van Marlen 1997; see also Chapter 9). Practical problems and high power prevented its use in marine fisheries. Heavy tickler chains or chain mats are thus still used in many flatfish fisheries to herd species into the mouth of the trawl. The inevitable result is biological and physical damage to the seabed community (Valdemarsen and Suuronen 2003). The Netherlands and Belgium researchers have made great efforts in evaluating and reducing the impact of flatfish beam trawls. Van Marlen and his colleagues (van Marlen et al. 1999, 2001) compared a 7-m (beam length) prototype electric beam trawl and the conventional beam trawl, and found that the catch of Dover sole (Solea solea) was about the same, but plaice catches were reduced by 50%. A reduction of 40% was found in benthos catches. A 12-m prototype electric trawl caught significantly more marketable sole than the conventional trawl of the same beam length (+22% in weight). Both catch weights of marketable plaice (Pleuronectes platessa) and undersized plaice were significantly lower in the electric trawl (by 17% and 18%, respectively) as well as less benthos (by 25% in weight) (van Marlen et al. 2005b). Survival tests and physiology measurements indicated that undersized sole and plaice caught in the electric beam trawl had better survival (van Marlen et al. 2001, 2005b). The catching and economic performance of the new technique are currently being studied in a year-

309

long trial on a commercial vessel fitted with a complete system of two 12-m electric trawls and cable winches. A comparison over 5 weeks with vessels similar in engine power and fishing with the conventional trawls revealed that catches of sole and plaice by the electric beam trawl were 60% to 70% of the commercial trawl but the electric trawl caught significantly less undersized sole, sandstar (Astropecten irregularis L.), common starfish (Asterias rubens L.), and swimming crab (Liocarcinus holsatus L.), and a great reduction in fuel consumption (up to 40%) was observed as a result of the lower towing speed required (van Marlen et al. 2006). It was concluded that the electric beam trawl can reduce adverse effects of beam trawling and has potential to reduce sole and plaice discards. The use of electric pulse stimulation in beam trawling in the North Sea is currently under review by ICES. The use of electric pulse shrimp beam trawls was very popular in the East China Sea in the 1990s but was banned in 2001 due to challenges in managing the new technology (Yu et al. 2007). 12.8.2 Use of Electric Stimuli in Brown Shrimp Beam Trawls Research on the use of electric pulses as an alternative stimulus was carried out in the brown shrimp (Crangon crangon) beam trawl fishery in Belgium to improve species selectivity and to reduce bottom contact of the groundgear (Polet et al. 2005a, 2005b; see also Chapter 9). The study found that fish and invertebrates (with the exception of dabs and sole) showed weak responses to electric pulses. Brown shrimp on the other hand showed a strong response and may jump into the water column up to 50 cm. This species-specific response indicates that species-selective fishing for brown shrimp may be possible using electric pulses. Subsequent sea trials demonstrated the potential for a species-selective and benthos-friendly electric trawl without loss of the targeted brown shrimp (Polet et al. 2005b). The raised footrope design of the electric trawl created an escape opening for most of the discard fish species and benthos commonly caught in the shrimp trawl. Parallel electrode arrangement, rather than perpendicular, may also reduce the seabed effects of the gear.

310

Contemporary Issues in Capture and Conservation in Marine Fisheries

12.8.3 Drop-out Panels to Reduce Benthos Catch Belgian, Dutch, and British researchers have been working on drop-out zones and escape panels in the belly of beam trawls for flatfish and shrimp to reduce catch of benthos and other seabed materials (Fonteyne and Polet 2002; Revill and Jennings 2005; van Marlen 2000; van Marlen et al. 2005a). The Belgian tests indicated that square mesh escape panels in the belly just ahead of the codend significantly reduced benthos caught in the flatfish beam trawls (Fonteyne and Polet 2002). For commercial species, the square mesh panels produced a mixed result, but overall there was a benefit of releasing undersized fish (Fonteyne and Polet 2002). British researchers also confirmed that square mesh drop-out panels of 140- to 150-mm full mesh fixed into the belly of a beam trawl a few meshes in front of the codend were most effective in the English Channel beam trawl fishery (Revill and Jennings 2005). Around 80% of unwanted benthic invertebrates were released from the beam trawls using drop-out panels and escapees exhibited a high survival rate. No loss of target species was observed with this simple technology. 12.9 CONCLUDING REMARKS Fishing gear, including otter trawls and beam trawls, operating on or near the seabed may alter physical structure and biological composition of the seabed through their interaction with the seabed and benthic organisms. Research is continuing on the modification of fishing gear and/or operational methods to reduce seabed impacts. Alternative gear with less seabed contact, such as pelagic or semipelagic trawls, may be used instead of traditional bottom-tending gear in some fisheries where herding of target species by sand clouds is less critical. Gear modifications that have less seabed impact include designs with less contact area or fewer contact points, lower weight of groundgear and doors, use of more efficient and high aspect ratio trawl doors, provision of dropout openings in beam trawls, adoption of “sweepless” trawls (without groundgear), and “wheeled” or “rollerball” groundgear replacing rockhoppers. Electrical stimuli have been tested in beam trawls to replace or reduce traditional heavy tickler chains in some fisheries or

to selectively stimulate target species. Drop-out panels have been tested in some beam trawl fisheries to reduce the catch of benthos and to reduce dislocation and mortality associated with capture and discard processes. ACKNOWLEDGMENTS The authors are grateful to Dr. Craig Rose of NOAA Fisheries for his comments and suggestion to the manuscript and to members of the ICES/FAO Working Group on Fishing Technology and Fish Behavior for their discussions and comments. REFERENCES Arkley K. 2006. Development of off-bottom trawl and trawling techniques: report on gear handling and engineering trials: MFV Shiralee. Report for the Cornish Fisheries Development. 18 pp. Auster PJ and Langton RW. 1999. The effects of fishing on fish habitat. Am. Fish. Soc. Symp. 22: 150–187. Ball B, Munday B and Fox G. 1999. The impact of a Nephrops otter trawl fishery on the benthos of the Irish Sea. J. Shellfish Res. 18(2): 708. Ball B, Linnane A, Munday B, Davis R and McDonnell J. 2003. The rollerball net: a new approach to environmentally friendly ottertrawl design. Arch. Fish. Mar. Res. 50: 193–203. Barnes PW and Thomas JP (eds). 2005. Benthic habitats and the effects of fishing. Am. Fish. Soc. Symp. 41: 890 pp. Bergman MJN and Hup M. 1992. Direct effects of beam trawling on macrofauna in a sandy sediment in the southern North Sea. ICES J. Mar. Sci. 49: 5–11. Bergman MJN and van Santbrink JW. 2000. Mortality in megafaunal benthic populations caused by trawl fisheries on the Dutch continental shelf in the North Sea in 1994. ICES J. Mar. Sci. 57: 1321–1331. Brewer D, Eayrs S, Mounsey R and Wang YG. 1996. Assessment of an environmentally friendly, semipelagic fish trawl. Fish. Res. 26: 225–237. Collie JS, Hall SJ, Kaiser MJ and Pointer IR. 2000. A quantitative analysis of fishing impacts on shelf sea benthos. J. Animal Ecol. 69: 785–798. DeLouche H and Legge G. 2004. Reducing seabed contact while trawling: a semi-pelagic trawl for the Newfoundland and Labrador shrimp fishery. A report submitted to Canadian Center for Fisheries Innovation. St. John’s, Newfoundland: Fisheries and Marine Institute. 13 pp.

Effect of Trawling on the Seabed and Mitigation Measures to Reduce Impact Dorsey EM and Pederson J (eds). 1998. Effects of Fishing Gear on the Sea Floor of New England. Boston: Conservation Law Foundation. 160 pp. Drabsch SL, Tanner JE and Connell SD. 2001. Limited infaunal response to experimental trawling in previously untrawled areas. ICES J. Mar. Sci. 58: 1261–1271. Fonteyne R. 2000. Physical impacts of beam trawls on seabed sediments. In: Kaiser MJ and de Groot SJ (eds). Effects of Fishing on Non-Target Species and Habitats. Biological, Conservation and Socioeconomic Issues. pp 15–36. Oxford: Blackwell Science. Fonteyne R and Polet H. 2002. Reducing the benthos bycatch in flatfish beam trawling by means of technical modifications. Fish. Res. 55: 219–230. Freese L, Auster PJ, Heifetz J and Wing BL. 1999. Effects of trawling on seafloor habitat and associated invertebrate taxa in the Gulf of Alaska. Mar. Ecol. Prog. Ser. 182: 119–126. Fuller SD, Picco C, Ford J, Tsao C-F, Morgan LE, Hangaard D and Chuenpagdee R. 2008. How we fish matters: addressing the ecological impacts of Canadian fishing gear. ISBN 978-0-9734181-7-0. 25p. Gabriel O, Lange K, Dahm E and Wendt T. 2005. Von Brandt’s Fish Catching Methods of the World. 4th edition. Oxford: Blackwell. 523 pp. Garner J. 1978. Pelagic and Semi-pelagic Trawling Gear. Surrey, England: Fishing News Books. 59 pp. Goudey C. 1999. Progress in reducing the habitat impact of trawls and dredges. MIT Sea Grant College Program. MITSG 99–8. Goudey C and Loverich G. 1987. Reducing the bottom impact of Alaskan groundfish trawls. In: Oceans 87 Proceedings: Vol. 2. Marine Engineering, Policy, Education and Technology Transfer. pp 632–637. Graham N. 2006. Trawling: historic development, current status and future challenges. Mar. Technol. Soc. J. 40: 20–24. Glass CW, Walsh SJ, van Marlen B and Amaratunga T (eds). 2007. Fishing Technology in the 21st Century: Integrating Fishing and Ecosystem Conservation. ICES J. Mar. Sci. 64: 1499–1616. Hall SJ. 1999. The Effect of Fishing on Marine Ecosystems and Communities. Osney Mead (Oxford): Blackwell. 274 pp. Hansson M, Lindegarth M, Valentinsson D and Ulmestrand M. 2000. Effects of shrimp-trawling on abundance of benthic macrofauna in Gullmarsfjorden, Sweden. Mar. Ecol. Prog. Ser. 198: 191–201.

311

He P. 2001. Reducing seabed contact of bottom trawls. Proceedings of a workshop on “Fishing Impacts: Evaluation, Solution and Policy.” pp 27–35. Japanese Society for Fisheries Science Roundtable Meeting on Fishing Technology, No. 45. Tokyo, Japan. He P. 2007. Technical measures to reduce seabed impact of mobile gears. In: Kennelly S (ed). Bycatch Reduction in World Fisheries. pp 141–179. The Netherlands: Springer. He P, Hamilton R, Littlefield G and Syphers R. 2006. Design and test of a semi-pelagic shrimp trawl to reduce seabed impact. Final report submitted to the Northeast Consortium. UNH-FISH-REP-2006-029. 24 pp. Durham, NH: University of New Hampshire. Humborstad O-B, Nøttestad L, Løkkeborg S and Rapp HT. 2004. RoxAnn bottom classification system, sidescan sonar and video-sledge: spatial resolution and their use in assessing trawling impacts. ICES J. Mar. Sci. 61: 53–63. Jennings S and Kaiser MJ. 1998. The effects of fishing on marine ecosystems. Adv. Mar. Biol. 34: 201– 352. Johnson KA. 2002. A review of national and international literature on the effects of fishing on benthic habitats. NOAA Technical Memorandum NMFS-F/ SPO-57. 72 pp. Kaiser MJ and de Groot SJ. 2000. Effect of Fishing on Non-target Species and Habitats. Osney Mead (Oxford): Blackwell. 398 pp. Kaiser MJ and Spencer BE. 1996. The effects of beamtrawl disturbance on infaunal communities in different habitats. J. Animal Ecol. 65: 348–358. Kenchington ELR, Prena J, Gilkinson KD, Gordon DC Jr, Macissac K, Bourbonnais C, Schwinghamer PJ, Rowell TW, McKeown DL and Vass WP. 2001. Effects of experimental otter trawling on the macrofauna of a sandy bottom ecosystem on the Grand Banks of Newfoundland. Can. J. Fish. Aquat. Sci. 58: 1043–1057. Krost P, Bernhard MB, Werner F and Hukriede W. 1990. Otter trawl tracks in Kiel Bay (Western Baltic) mapped by side-scan sonar. Meeresforschung. 32: 344–353. Kutti T, Høisæter T, Rapp HT, Humborstad O-B, Løkkeborg S and Nøttestad L. 2005. Immediate effects of experimental otter trawling on a sub-artic benthic assemblage inside Bear Island Fishery Protection Zone in the Barents Sea. In: Barnes PW and Thomas JP (eds). Benthic Habitats and the Effects of Fishing. Am. Fish. Soc. Symp. 41: 519–528.

312

Contemporary Issues in Capture and Conservation in Marine Fisheries

Langeland MR. 2005. Escapement of Fish under a Survey Trawl: the Effect of Ground Gear Configuration. MSc thesis, University of Bergen. 77 pp. Linnane A, Ball B, Munday B, van Marlen B, Bergman M and Fonteyne R. 2000. A review of potential techniques to reduce the environmental impact of demersal trawls. Irish Fish. Invest. 7: 39 pp. Løkkeborg S. 2005. Impacts of trawling and scallop dredging on benthic habitats and communities. FAO Fish. Tech. Pap. 472: 58 pp. Main J and Sangster GI. 1981. A study of the sand clouds produced by trawl boards and their possible effect on fish capture. Scot. Fish. Res. Rep. 20: 20 pp. Moran MJ and Stephenson PC. 2000. Effects of otter trawling on macrobenthos and management of demersal scalefish fisheries on the continental shelf of north-western Australia. ICES J. Mar. Sci., 57: 510–516. Morgan LE and Chuenpagdee R., 2003. Shifting gears: Addressing the collateral impacts of fishing methods in U.S. waters. Washington, DC: Island Press. 42 pp. NOAA. 2005. Final Environmental Impact Statement for Essential Fish Habitat Identification and Conservation in Alaska. NOAA National Marine Fisheries Service, Alaska Region. NRC. 2002. Effect of trawling and dredging on seafloor habitat. National Research Council (US). Washington, DC: National Academy Press. 136 pp. Pereyra WT. 1995. Midwater trawls and the Alaska pollock fishery: a management perspective. In: Solving Bycatch: Consideration for Today and Tomorrow. pp. 91–94. University of Alaska Sea Grant College Program Report 96–03. Pol M. 2003. Tuning gear research into effective management: a case study. Presented at Conference on Managing Our Fisheries. Washington, DC. Nov. 2003. Pol M and Carr HA. 2000. Overview of gear development and trends in the New England commercial fishing industry. Northeast Naturalist. 7: 329– 336. Polet H, Delanghe F and Verschoore R. 2005a. On electrical fishing for brown shrimp (Crangon crangon): I. Laboratory experiments. Fish. Res. 72: 1–12. Polet H, Delanghe F and Verschoore R. 2005b. On electrical fishing for brown shrimp (Crangon crangon): II. Sea trials. Fish. Res. 72: 13–27. Prena J, Schwinghamer P, Rowell TW, Gordon DC Jr, Gilkinson KD, Vass WP and McKeown DL. 1999.

Experimental otter trawling on a sandy bottom ecosystem of the Grand Banks of Newfoundland: analysis of trawl bycatch and effects on epifauna. Mar. Ecol. Prog. Ser. 181: 107–124. Ramm DC, Mounsey RP, Xiao Y and Poole SE. 1993. Use of a semi-pelagic trawl in a tropical demersal trawl fishery. Fish. Res. 15: 301–313. Revill AS and Jennings S. 2005. The capacity of benthos release panels to reduce the impacts of beam trawls on benthic communities. Fish. Res. 75: 73–85. Rice J. 2006. Impacts of mobile fishing gears on seafloor habitats, species, and communities: a review and synthesis of selected international reviews. Can. Sci. Adv. Secret. Res. Doc., 2006/057: 35 pp. Rose C, Carr A, Ferro RST, Fonteyne R and MacMullen P. 2000. Using gear technology to understand and reduce unintended effects of fishing on the seabed and associated communities: background and potential directions. pp 106–122. Annex to ICES FTFB Report 2000. Rose C, Hammond C, An H-C, Stoner A and McEntire S. 2006. Modifying trawl bridles and sweeps to reduce their effects on seafloor habitat of the Bering Sea shelf. Presented at the ICES Symposium on Fishing Technology in the 21st Century, Oct. 30– Nov. 3, 2006. Boston, MA. Schwinghamer P, Gordon DC Jr., Rowell TW, Prena J, McKeown DL, Sonnichsen G and Guigne JY. 1998. Effects of experimental otter trawling on surficial sediment properties of a sandy-bottom ecosystem on the Grand Banks of Newfoundland. Conserv. Biol. 12: 1215–1222. Sinclair M and Valdimarsson G (eds). 2003. Responsible Fisheries in the Marine Ecosystems. Rome: FAO. 426 pp. SINTEF. 2004. Spreading Ground Gear. SINTEF Fisheries and Aquaculture. Hirtshals, Denmark. Somerton DA and Munro P. 2001. Bridle efficiency of a survey trawl for flatfish. Fish. Bull. 99: 641–652. Sparks-McConkey PJ and Watling L. 2001. Effects on the ecological integrity of a soft-bottom habitat from a trawling disturbance. Hydrobiologia. 456: 73–85. Sterling D and Eayrs S. 2006. Design and assessment of two gear modifications to reduce the benthic impact and fuel intensity of prawn trawling in Australia. Presented at the ICES Symposium on Fishing Technology in the 21st Century, Oct. 30– Nov. 3, 2006. Boston, MA, USA. Stewart PAM. 1975. Catch selectivity by electric fishing systems. J. Cons. Int. Explor. Mer. 36(2): 106–109.

Effect of Trawling on the Seabed and Mitigation Measures to Reduce Impact Tuck ID, Hall SJ, Robertson MR, Armstrong E and Basford DJ. 1998. Effects of physical trawling disturbance in a previously unfished sheltered Scottish sea loch. Mar. Ecol. Prog. Ser. 162: 227–242. Valdemarsen JW and Suuronen P. 2003. Modifying fishing gear to achieve ecosystem objectives. In: Sinclair M and Valdemarsen G (eds). Responsible Fisheries in the Marine Ecosystems. pp 321–341. Rome: FAO. van Dolah RF, Wendt PH and Nicholson N. 1987. Effects of a research trawl on a hard-bottom assemblage of sponges and corals. Fish. Res. 5: 39–54. van Dolah RF, Wendt PH and Levisen MV. 1991. A study of the effects of shrimp trawling on benthic communities in two South Carolina sounds. Fish. Res. 12: 139–156. van Marlen B. 1997. Alternative stimulation in fisheries. Final Report EU-project AIR3-CT94–1850, June 1997. van Marlen B. 2000. Technical modifications to reduce the by-catches and impacts of bottom gears on nontarget species and habitats. In: Effects of Fishing on Non-target Species and Habitats: Biological, Conservation and Socio-economic Issues. pp 253–268. Kaiser MJ and de Groot SJ (eds). Oxford: Blackwell. van Marlen B, Bergman MJN, Groenewold S and Fonds M. 2001. Research on diminishing impact in demersal trawling: The experiments in the Netherlands. ICES CM. 2001/R: 09. van Marlen B, Bergman MJN, Groenewold S and Fonds M. 2005a. New approaches to the reduction

313

of non-target mortality in beam trawling. Fish. Res. 72: 333–345. van Marlen B, Grift R, Keeken O, van Ybema MS and van Hal R. 2006. Performance of pulse trawling compared with conventional beam trawling. RIVOReport No. C014/06 van Marlen B, van Lavieren H, Piet GJ and van Duijn JB. 1999. Catch comparison of a prototype 7 m electrical beam trawl and a conventional tickler chain beam trawl. RIVO Internal Report 99.006b, April 1999. van Marlen B, Ybema MS, Kraayenoord A, de Vries M and en Rink G. 2005b. Catch comparison of a 12 m pulse beam trawl with a conventional tickler chain beam trawl. RIVO-Report C043b/05 Vincent B. 2001. A way to reduce impact of trawl door. Presented at the ICES WGFTFB meeting, April 2001. Seattle, WA. Wardle CS. 1986. Fish behavior and fishing gear. In: Pitcher TJ (ed). The Behavior of Teleost Fishes. pp 463–494. London and Sydney: Croom Helm. Winger PD, DeLouche H and Legge G. 2006. Designing and testing new fishing gears: The value of a flume tank. Mar. Technol. Soc. J. 40: 44–49. Yu C, Chen Z, Chen L and He P. 2007. Raise and fall of the electrical shrimp beam trawling in China. ICES J. Mar. Sci. 64: 1592–1597. Zachariassen K. 2000. Trolbustir FRS smárit 00/7 (in Faroese). Zachariassen K. 2004. Umhvørvisvinarligur trolgrunnur FRS smárit 04/4 (in Faroese).

314

Contemporary Issues in Capture and Conservation in Marine Fisheries

SPECIES MENTIONED IN THE TEXT Alaskan pollock, Theragra chalcogramma Atlantic cod, cod, Gadus morhua brown shrimp, Crangon crangon common starfish, Asterias rubens Dover sole, sole, Solea solea king prawn, Melicertus latisulcatus Pacific cod, Gadus macrocephalus

pink shrimp, Pandalus borealis plaice, Pleuronectes platessa red snappers, Lutjanus malabaricus and L. erythropterus sandstar, Astropecten irregularis silver hake, Merluccius bilinearis swimming crab, Liocarcinus holsatus tiger prawn, Penaeus esculentus

Chapter 13 Measures to Reduce Interactions of Marine Megafauna with Fishing Operations Dominic Rihan

13.1 INTRODUCTION Marine megafauna include all cetacean species (whales, dolphins, and porpoises), pinnipeids (seals, sea lions, and sirenians), sea turtles, as well as large fish species including sharks, manta ray, and sawfish. The life histories of many of these species make them highly vulnerable to human exploitation or unintended mortality, and therefore the incidental bycatch of marine megafauna associated with commercial fishing operations, leading in most cases to mortality, is an issue of global concern. The extent, nature, and impacts of direct interactions are, however, relatively poorly understood. In the face of this knowledge deficit, fisheries managers are under increasing pressure from governments, nongovernmental organizations (NGOs), and the general public to protect these large marine mammals, fish, and turtles from fishing activities. There are a number of high-profile fisheries where, from an ethical and/or biological perspective, such bycatch is considered unacceptable. Some examples include the capture of small cetaceans in gillnets or in pelagic trawls, the entanglement of whales in marker ropes, and the bycatch of turtles in both longline and shrimp trawl fisheries. In some instances, managers have closed viable fisheries due to marine mammal issues such as the tuna driftnet fishery in the Northeast Atlantic. Some fisheries have suffered economic downturn due to consumer boycotts, for example, the Eastern Pacific tuna

purse-seine fishery that resulted in the concept of “dolphin-friendly” tuna. Interactions between marine megafauna and fisheries take several forms as described by Read and his colleagues (Read et al. 2006). Some are operational where animals are in direct contact with fishing gears, while others involve disruptions of trophic pathways. Avoiding or mitigating consequences of these interactions typically require different management approaches (Northridge and Hofman 1999), though, modifications to fishing gear and/or fishing practices can in many instances provide practical and workable alternatives for fisheries managers. As identified by Werner and his colleagues (Werner et al. 2006), marine mammal bycatch reduction is a very active area of research with numerous ongoing studies and constant development and testing of novel initiatives and mitigation devices. This research in many cases has been driven by genuine concerns among fisheries managers, researchers, and fishermen to protect endangered species, whereas some research has been motivated by the need to reduce gear damage caused by interactions with marine megafauna or reducing predation of target catch by these species. Worldwide, there are a number of successful examples in which technical conservation measures have greatly reduced the impact of fisheries, and usually these have been developed with a high level of

315

316

Contemporary Issues in Capture and Conservation in Marine Fisheries

involvement from fishermen working closely with fishery scientists. Such examples include the use of turtle excluder devices (TEDs) in many tropical shrimp fisheries that have reduced the mortality rates of several turtle species (Shiode and Tokai 2004), and the modification of hook shape which has reduced turtle bycatch associated with longline fisheries (Gilman 2006). Similarly, the development of acoustic deterrents in gillnet fisheries as described by several authors (Barlow and Cameron 2003; Gearin et al. 2000; Kraus et al. 1997) and simple operational changes in purse-seine and trawl fisheries (Perrin et al. 2002a) have also been shown to reduce nontarget bycatch in some fisheries. However, it is apparent that there is only limited knowledge on the behavior of marine mammals around fishing gear, and this has hindered the development of more acceptable solutions. In some fisheries, gear–marine mammal interaction is high but with a comparatively low bycatch rate. This suggests that certain species of marine mammals, particularly cetaceans, are aware of the presence of nets or longlines and may actively use them during foraging or feeding. It could also mean that the use of certain gear types or fishing methods may result in a higher probability of fishing associated mortality in nontarget catch. Understanding circumstances that lead to incidental capture can provide ideas for promising mitigation strategies. It is also fair to say that to date most available techniques have been directed at reducing the bycatch of cetaceans, pinnipeds, and sea turtles. Although the emphasis on these taxonomic groups is justifiable given the high level of bycatch from fishing encounters and their public appeal, other nontarget species such as large sharks and the manta ray are as or more seriously endangered by conflicts with fishing operations but to date have received limited bycatch mitigation attention (Werner et al. 2006). 13.2 SPECIES AND FISHERIES INVOLVED The recorded number of interactions between marine megafauna and fishing operations is large. Probably of most concern has been the reported large number of cetacean species, particularly charismatic species such as dolphins, that die in fisheries

around the world. This has been documented for several decades; nonetheless, progress at quantifying the scale of the problem globally has remained slow and an understanding of the causes of bycatch is still limited in many cases. While bycatch in set and drifting gillnets and purse-seines remain of principal concern, incidental mortality in trawl nets, longlines, and some artisanal fishing methods such as beach seines is also problematic in many parts of the world. Of the 80 species of cetaceans, a group that includes whales, dolphins, and porpoises, Northridge (1991) noted that “most marine mammals, with the exception of the rarer ocean beaked whales, have been recorded at some time or other caught in some type of fishing gear.” A recent review carried out by a team of scientists on behalf of the World Wildlife Fund (WWF) designated cetacean species most at risk on the basis of documented species or population threats (Reeves et al. 2005). These include the following: • Irrawaddy dolphins (Orcaella brevirostris) in the crab net/trap fishery in Malampaya Sound, Philippines • Irrawaddy dolphins in gillnets in the Mekong, Mahakam, and Ayeyarwady Rivers and in Chilka and Songkhla Lakes, Southeast Asia • Indo-Pacific humpback dolphins (Sousa chinensis) and Indo-Pacific bottlenose dolphins (Tursiops truncates) in drift and bottom-set gillnets on the south coast of Zanzibar (Tanzania) • Harbor porpoises (Phocoena phocoena) in coastal gillnets in the Black Sea • Spinner dolphins (Stenella longirostris) and Fraser ’s dolphins (Lagenodelphis hosei) in large-mesh driftnets and purse-seines in the Philippines • Atlantic humpback dolphins (Sousa teuszii) in coastal gillnets in the northern Gulf of Guinea (Ghana, Togo) • Burmeister ’s porpoises (Phocoena spinipinnis) in artisanal gillnets in Peru • Franciscana dolphins (Pontoporia blainvillei) in coastal gillnets in Argentina, Uruguay, and Brazil • Commerson’s dolphins (Cephalorhynchus commersonii)in coastal gillnets and midwater trawls in Argentina

Measures to Reduce Interactions of Marine Megafauna with Fishing Operations In the tropical waters of the Pacific Ocean offshore Mexico and Central America, large yellowfin tuna (Thunnus albacares) swim together with several species of dolphins—pantropical spotted (Stenella attenuata), spinner (S. longirostris), and common dolphins (Delphinus delphis). This ecological association of tuna and dolphins is not clearly understood, but it has had two important practical consequences: it has formed the basis of developing a successful and lucrative fishery, and it has resulted in the incidental deaths of a large number of dolphins. The bycatch of dolphins in the Eastern Tropical Pacific Ocean (ETP) purse-seine tuna fishery stands apart from marine mammal bycatch in other fisheries, not only because it has been estimated as very high but also in the unique way in which the dolphins interact with the fishery. Marine mammals interact with most fishing gear only incidentally, but in the ETP tuna fishery the dolphins are an intrinsic part of the fishing operation. Fishermen reportedly have intentionally targeted both tuna and dolphins together, and then released dolphins from the net, both dead and alive. In recent years, unlike in most other fisheries, the vast majority (more than 99%) of dolphins captured by the ETP tuna fishery have been released alive; thus, an individual dolphin may be chased, captured, and released many times during its lifetime. The number of dolphins killed since the purseseine fishery began some four decades ago is estimated by Perrin and his colleagues (2002a) to be over 6 million animals, the highest known for any fishery. For comparison, the total number of whales of all species killed during commercial whaling in the twentieth century is about two million (Perrin et al. 2002a). In recent years, the kill of dolphins in the ETP tuna fishery has declined by two orders of magnitude, but even at this level it remains the largest documented cetacean kill in world fisheries. Cetacean bycatch in trawl fisheries has been noted as problematic in the United States (Northridge 1996), South America (Crespo et al. 1994, 1997, 2000; Dans et al. 1997, 2003), New Zealand (Baird 1994, 1996), and Australia (Browne et al. 2005). In the United States, some trawl fisheries have been closed due to associated mammal bycatch that exceeded levels deemed acceptable by authorities

317

(Read 2000), whereas in New Zealand there is evidence of fisheries having dwindled and moved away from the region of conflict (Tilzey et al. 2004). In Europe, the annual number of stranded dead dolphins has been high, especially during winter months on French Atlantic and English Channel coasts since the late 1980s (Collet and Mison 1996). There were 784 stranded dolphins reported off the coast of the United Kingdom in 2004 (Anon 2006). Forensic pathology suggests that a large proportion of these animals died due to fishing operations, and pelagic trawlers have been implicated in many cases (Bennett and Jepson 2000; Kuiken et al. 1994). The pelagic trawl fisheries of this region are complex and varied, involving over 12 target species and six nations and at least three major gear types. Some of these fisheries have relatively low or nonexistent cetacean bycatch rates, whereas some others, such as the European sea bass (Dicentrarchus labrax) fishery and Albacore tuna (Thunnus alalungus) fishery, clearly have relatively high bycatch rates (BIM 2002; Couperus 1997; Morizur and Tregenza 1996; Northridge 2002). A number of species of whales, such as humpback (Megaptera novaeangliae), Northern and Southern right whales (Eubalaena glacialis and Eubalaena australis), grey whales (Eschrichtius robustus), and bowhead whales (Balaena mysticelus), are known to become entangled in buoy lines of lobster and crab pots, gillnets and longlines (Bannister 2001; Knowlton et al. 2001). Such incidents are sporadic, but given the low population numbers of such species, these events are nonetheless significant. Although entanglement usually does not lead to death, it may cause secondary problems, such as infection or inability to feed, making the animals vulnerable to further entanglements because of trailing gear or exhaustion. For instance, according to Kraus (1990), while only a few right whales have been observed entangled in fishing gear, a large number of the remaining North Atlantic population exhibit scarring consistent with previous entanglements (Kraus 1990). Other species of whales are also taken as a bycatch; for instance, at least 30 to 40 minke whales (Balaenoptera spp.) are reported annually as bycatch in South Korean (Kim 1999) and Japanese waters (Tobayama et al. 1992),

318

Contemporary Issues in Capture and Conservation in Marine Fisheries

whereas there are reported bycatch of short-finned pilot whales (Globicephalas melas) and false killer whales (Pseudorca crassidens) in trawls, pelagic longlines, and gillnets in a number of areas (Kleiber 1999; Waring et al. 2006). Information on seals, sea lions, and other marine mammals taken incidentally in fishing gear are also widely reported but less well quantified than for cetaceans. Of the 33 recognized species of pinnipeds, Northridge (1991), in his review of interactions between marine mammals and fisheries, listed numerous species for which documented bycatch have been observed. Significant interactions are noted between fisheries and the northern sea lion (Eumetopias jubatus), northern fur seal (Callorhinus ursinus), harp seal (Phoca groenlandica), grey seal (Halichoerus grypus), Caribbean manatee (Trichechus manatus), Mediterranean monk seal (Monachus monachus), dugong (Dugong dugon), harbor seal (Phoca vitulina), and New Zealand fur seal (Arctocephalus forsteri). The majority of reported interactions are in gillnet fisheries, but high bycatch rates have also been recorded in trawl and pelagic longline fisheries. Of seven recognized species of marine turtle species, six species—the green turtle (Chelonia mydas), hawksbill turtle (Eretmochelys imbricata), loggerhead turtle (Caretta caretta), olive ridley turtle (Lepidochelys olivacea), Kemp’s ridley turtle (Lepidochelys kempii), and leatherback turtle (Dermochelys coriacea)—have reported interactions with fishing gears. Only the flatback turtle (Natador depressus) is not reported as bycatch in fishing operations. Most sea turtle species are currently listed by the International Union for Conservation of Nature (IUCN) as Endangered, and three of those as Critically Endangered, while the seventh is listed as data deficient. Incidental catches have been recorded in pelagic longline and shrimp trawl fisheries globally, as well as in purse-seine, gillnet fisheries, and even scallop dredge fisheries. Other long-lived, but perhaps less charismatic species, including sharks and rays, are also reported as bycatch in fisheries in many parts of the world. There are approximately 1165 species of sharklike fishes or “sharks” as broadly defined by the U.N. World Food and Agricultural Organization (FAO) shark action plan (FAO 1998), including at least 50

species of silver sharks (chimaeras), 627 species of winged sharks (batoids), and 488 species of “nonbatoid” sharks or ordinary sharks. Most sharks are relatively small, with a maximum length between 0.3 and 1.5 m. Relatively few species are over 2 m long, which would result in them being considered as “true” marine megafauna. Of these larger shark species, the relatively few large littoral and oceanic carcharhinoid and lamnoid sharks, the “large pelagic sharks,” get the most attention, primarily due to their public perception as “man-eaters,” concern over finning, and reports of catch declines in the North Atlantic and elsewhere. In addition, there are “supersharks,” a few gigantic wide-ranging, nonbatoid and batoid species such as the white (Carcharodon carcharias), whale (Rhincodon typus), and basking sharks (Cetorhinus maximus) and the manta ray (Manta birostris) with major conservation problems due to high vulnerability to exploitation combined with high commercial demand for their fins, flesh, jaws, and other products. Shark and ray bycatch are reported in pelagic longlines, drift net fisheries, and trawl fisheries in many parts of the world. 13.3 EXTENT OF BYCATCH Adequately quantifying megafauna bycatch requires essentially a high level of on board observer coverage (typically at a level of 25% to 30% of total fishing effort) to be able to provide accurate estimates and associated confidence limits around estimates (Northridge and Thomas 2003). Levels of coverage are frequently at much lower levels than this. For instance, a recent European Union (EU) regulation seeks observer coverage at a level of only 5%. This is mainly due to costs involved in maintaining observer programs, resulting in a large amount of data from anecdotal sources to supplement the quantitative data gathered from observer programs. The lack of systematic monitoring prevents the true extent and potential impacts of marine megafauna/fishery interactions from being fully understood or documented, although observer coverage certainly indicates trends in bycatch and do highlight particular “problem” fisheries. There are only two estimates of the global magnitude of the total incidental catch of marine megafauna in fisheries. An estimate by the International

Measures to Reduce Interactions of Marine Megafauna with Fishing Operations Whaling Commission in 1991 based on data from observer programs provided a figure of between 65,000 and 85,000 marine mammals caught annually in commercial fisheries (Northridge 1991). The same analysis identified 54 species/populations/ regions (SPRs) for which abundance estimates and data on incidental mortality in passive gears were available from a variety of sources including observer programs and scientific research projects. This listing of species involved only cetaceans interacting with passive gear. Bycatch taken by active fishing gears and species from other taxonomic groups, such as seals, sea lions, manatees, and dugongs, were excluded so the bycatch numbers can be considered a gross underestimate. Recent data, using expanded U.S. bycatch data, in combination with global fleet composition and landings data provided by the FAO, produced higher estimates of global bycatch of marine mammals, reaching 653,364 annually, consisting of 345,611 pinnipeds and 307,753 cetaceans (Read et al. 2006). Incidental capture in fishing operations is also considered a major threat to the recovery and protection of marine turtle species and populations. One of the greatest problems faced with estimating actual levels of bycatch in sea turtles is that, just like whales, they migrate over vast distances, often crossing national and international boundaries. An extrapolation carried out in 2004 (based on limited fishing effort and bycatch data from 40 countries and 13 observer programs) estimated that 200,000 loggerheads and 50,000 leatherback turtles were caught globally in pelagic longline gears in 2000, with an estimated mortality of approximately 20% (Read et al. 2006). To reduce these threats, a series of measures have been instigated in fisheries in the Pacific and Atlantic Oceans and the Gulf of Mexico. Such measures have included gear modifications, changes in fishing practices, and time/area closures. Despite these measures, concerns remain about the entanglement and hooking of sea turtles in pelagic longline gears and capture of loggerhead and leatherhead turtles in shrimp trawls. A more recent and interesting interaction reported by Smolowitz (2006) has been the bycatch of loggerhead turtles in U.S. scallop fisheries. During the summer of 2000, scallop captains began to report seeing loggerhead sea turtles where they had rarely, if ever,

319

seen them before and some were brought up to the sea surface inside or on top of scallop dredges. During 2001, U.S. National Marine Fisheries Service (NMFS) observers recorded 11 encounters between sea scallop vessels and loggerhead sea turtles in the Mid-Atlantic in 5,286 observed hauls. The observed bycatch, when extrapolated to the appropriate fleet effort, provided an estimated bycatch of 95 turtles. In 2002, over 20 incidences of turtle bycatch were reported in 72 observed trips into the Hudson Canyon Closed Area located in the U.S. side of the Mid-Atlantic region. In 2003, similar bycatch rates were observed in other areas of the Mid-Atlantic as well, resulting in a take estimate for the entire scallop fishery of 749 animals (Murray 2004). The extent of bycatch of other megafauna such as sharks and manta ray is less well documented. Fisheries bycatch of these species is seldom monitored or regulated, so impacts on their populations often go unreported or unnoticed until species numbers have drastically decreased. Bycatch of sharks in high seas longline fisheries for tuna and billfish is very high because these fisheries set a cumulatively large number of hooks that catch sharks attracted either to bait or to the target species. These fisheries are common throughout much of the world, mostly in the Pacific but also in the Indian and Atlantic Oceans. With so much fishing effort, it is not surprising they are the major source of shark bycatch. Blue sharks (Prionace glauca) comprise the majority of bycatch in these fisheries. Campana and his colleagues (2006) reported two independent approximations of total blue shark catch mortality of 100,000 tons in the North Atlantic alone. There is also a significant bycatch of silky (Carcharhinus falciformis), oceanic whitetip (Carcharhinus longimanus), mako (Isurus oxyrinchus), porbeagle (Lamna nasus), hammerhead (Sphyrna mokarran), and other sharks including white (Carcharodon carcharias) and tiger sharks (Galeocerdo cuvier) in longline fisheries in tropical and coastal seas. Other well-documented incidents of high bycatch include the pelagic trawl fisheries off Mauritania, in which Zeeberg and his colleagues (2006) reported figures of manta ray bycatch in these fisheries of between 120 and 620 mature manta rays and 1,085

320

Contemporary Issues in Capture and Conservation in Marine Fisheries

hammerhead sharks annually by the fleet of 5 to 10 large pelagic freezer trawlers. NMFS also documented a sharp decline in smalltooth sawfish (Pristis pectinata), which have a propensity for entanglement in gillnets. Due to the large size of adults that potentially damage fishing gear or even pose a threat to fishermen, many incidentally captured sawfish were killed before they were removed from fishing gear, even if the fishermen had no interest in keeping them. Available data indicate that the species’ distribution has been reduced by about 90% and that the population numbers have declined dramatically, perhaps by 95% or more, leading to the establishment of a Smalltooth Sawfish Recovery Team by NMFS (NOAA 2001). The plan recommends specific steps to recover the distinct population segment (DPS), focusing on reducing fishing impacts, protecting important habitats, and educating the public. There are also directed fisheries for sharks that threaten populations of some of the larger species. In the southern Philippines, for example, a traditional whale shark fishery has been documented during the period from 1993 to 1996. A joint survey by WWF-Philippines and Silliman University reported an estimated average catch of 286 individuals per year over the period 1996 to 1997. Actual landings for 1997 were 143 individuals. Such catches are considered unsustainable given the biology of this large, slow-growing species (Alava and Yaptinchay 2000). 13.4 NATURE OF THE PROBLEM There are a variety of mitigation devices and gear modifications that have either been tested or are undergoing further experimentation. Many of them have been proved to work and some have been adopted. But because they are often sporadic in nature, marine megafauna interactions are notoriously difficult to observe and therefore remain poorly understood. Many mitigation measures are thus designed based on conjecture rather than on a detailed knowledge of the behavior of the animal in and around the fishing gear. Some are the result of technology transfer from other fisheries. For instance, the TED was based on the Nordmøre grid idea that originated in Norway for releasing fish from shrimp trawls.

Most evidence suggests that incidentally caught marine mammals are healthy when entanglement or capture happens. It is therefore safe to conclude that when death or mortal injury occurs, it is almost invariably the process of bycatch that is the cause (Zollet 2005). It is most likely that the primary motivation for marine mammals, turtles, and large fish species to interact with fishing gear is feeding, as fishing gear undoubtedly represents an easy-toaccess concentrated food source. Trawling and longlining, in particular, may provide a source of fish species that are normally hard for marine megafauna to feed on and provide an abundance of food with a high calorific value (Fertl and Leatherwood1997). In other circumstances, though, bycatch is a result of direct interactions between the target species and bycatch species—for example, in the ETP purse-seine fishery, fishermen actively use schools of dolphins to locate tuna. Or it may be indirect, such as entanglement of whales in buoy lines of traps, gillnets, and longlines. Of the observed associations between megafauna species and trawls, both opportunistic and targeted feedings of discarded fish from nets have been noted. Pace and his colleagues (2003) observed bottlenose dolphins (Tursiops truncates) associating with trawlers in four phases: following the trawls, feeding on the net, waiting for trash fish, and feeding on discarded fish. Seals and sea lions have also been observed feeding in or around fishing gear in a similar manner, either predating on fish directly or feeding on discarded fish. Incidental capture by fishing gears may not always relate to the target species. They can also occur when animals feed on associated nontarget species. They may also be in the vicinity of fishing gear due to an attraction to species that are preying on the captured fish or may also be feeding on organisms stirred up by the fishing operation. There is intense speculation as to why, when, and where marine megafauna are captured by fishing gears, and the lack of understanding can be a hindrance in the development process for mitigation measures. What little is known is based on a few, opportunistic direct observations. For instance, Northridge (2003) indicated that it is likely that cetaceans are alive when inside the trawl, but they die due to drowning during some point of the fishing

Measures to Reduce Interactions of Marine Megafauna with Fishing Operations operation, most likely during hauling. It is also thought that marine mammals may be particularly vulnerable to capture during certain phases of the fishing operation. Zollet (2005) suggests that when a net is deployed, cetaceans or seals may be captured due to the proximity to a vessel. Alternatively, they may enter the mouth of a trawl net during towing but become caught when the boat slows, turns, or hauls back the gear. Changes in speed and direction may contribute to bycatch because the size and shape of the net are altered and the space for feeding animals is changed in time, causing confusion. Other factors such as the size and condition of the animal, time of day, seasonality, or even the sex of the animal may be contributing factors to incidental capture. Similarly in gillnets, fishermen claim that bycatch usually occurs when nets are being hauled. In summary, marine megafauna bycatch in fisheries is generally associated with the motivation for the interaction. This can be either that the animal is exploiting the target species, associated species, or simply as a consequence of the two occurring in the same area. Bycatch is also known to relate to the behavior and physiology of the species, age, environmental conditions, spatial and temporal effects, depth, time of the day, and even possibly sex. Consideration of these factors, as appropriate to the specific species, is essential to understanding and developing strategies to reduce bycatch. Given the differences between species and also fishing operations and gears, it is apparent that there is no single solution that fits all species and all gear types, reinforcing the need for active involvement of fishermen in research given their knowledge and experience of fishery interactions with marine megafauna species. 13.5 REGULATORY FRAMEWORKS Globally, there are a number of regulatory frameworks designed to reduce or minimize marine megafauna bycatch. Most of these frameworks are adopted in the context of the FAO’s Code of Conduct for Responsible Fisheries (FAO 1995). Article 7 of the Code states that “States should take appropriate measures to minimize catch of nontarget species, both fish and non-fish species, and negative impacts on associated or dependent

321

species, in particular endangered species.” Many of the current regulations merely set maximum levels of bycatch based on a level deemed acceptable for sustaining the population, whereas others are more specific by prescribing the use of mitigation measures such as closed areas or seasons or, in the most extreme cases, prohibiting fishing methods in particular fisheries. 13.5.1 European Union In Europe under Article 2 of Council Regulation (EC) 2371/2002 of 20 December 2002 (Common Fisheries Policy), “the Community shall apply the Precautionary approach in taking measures designed to minimize the impact of fishing activities on marine ecosystems,” while under Council Regulation 92/42/EC (Habitats Directive) Member States are required to undertake further research or conservation measures to ensure that the incidental capture and killing of cetaceans “does not have a significant negative impact on the species concerned.” Several countries in the EU are also signatories to the Agreement on the Conservation of Small Cetaceans of the Baltic and North Seas (ASCOBANS), which was concluded in 1991 under the auspices of the Convention on Migratory Species (UNEP/CMS or Bonn Convention) and entered into force in 1994. This agreement has been signed by Belgium, Denmark, Finland, Germany, the Netherlands, Poland, Sweden, and the United Kingdom. Additional nations are considering accession. These nations, and the non-Party Range States cooperating with ASCOBANS even though they are not signatories, share a common concern that continuously high bycatch rates, habitat deterioration, and anthropogenic disturbance such as noise or chemical pollution are likely to jeopardize the existence of small cetaceans. At the third Meeting of Parties to ASCOBANS, a resolution was passed that called for competent fishery authorities to ensure that the total bycatch of small cetaceans was reduced as soon as possible to 1.7% of the best estimate of a species overall abundance. More recently, European Council Regulation No. 812/2004 laid down measures concerning incidental catches of cetaceans and also prescribed requirements for Member States to monitor cetacean bycatch in pelagic and gillnet fisheries. As part of

322

Contemporary Issues in Capture and Conservation in Marine Fisheries

this regulation, all vessels of 12 m or longer in overall length are prohibited from using any bottom-set gillnet, entangling net, or driftnet without the simultaneous use of active acoustic deterrent devices. 13.5.2 United States In the United States, all marine mammals are protected under the Marine Mammal Protection Act (MMPA). The MMPA prohibits, with certain exceptions, the take of marine mammals in U.S. waters and by U.S. citizens on the high seas and the importation of marine mammals and marine mammal products into the United States. All marine mammal stocks are reviewed annually under this plan, and known bycatch species are compared with the population numbers to calculate the potential biological removal (PBR) rate. Where bycatch rates exceed PBR, take reduction teams (TRTs), consisting of representatives from industry, academia, and NGOs, are established to devise management plans to reduce bycatch below PBR levels. A number of these TRTs have successfully developed plans that resulted in a reduction in bycatch through the use of mitigation measures. For example, in the midAtlantic and Gulf of Maine, the use of pingers in gillnet fisheries, in conjunction with large-scale time and area closures, has resulted in a reduction in Harbor porpoise bycatch. In other areas and fisheries, the U.S. government has simply closed fisheries under the MMPA to reduce bycatch. This occurred in 1996 when the North Atlantic driftnet fishery for swordfish was banned by NMFS. In addition, some 20 species of marine mammals and 6 species of turtle are protected under the Endangered Species Act of 1973 (ESA). It provides for the conservation of species that are endangered or threatened throughout all or a significant portion of their range and the conservation of the ecosystems on which they depend. 13.5.3 Southeast Asia In Southeast Asia, the Dumaguete Action Plan, which was agreed at the Second International Conference on Marine Mammals of Southeast Asia held July 22–26, 2002, in Dumaguete, the Philippines, set a strategy to protect small cetaceans (Perrin et al. 2002). This plan identified seven focal

areas to reducing cetacean bycatch in Southeastern Asian fisheries as follows: • Targeted community education and awareness programs • Improved enforcement initiatives • Monitoring and assessment of bycatch and fisheries • Gear research • Promotion of alternative livelihoods • Identification of key areas and closure of fisheries • Development of laws and regulations to reduce bycatch This plan has been in place for over 5 years, and it appears that local awareness of the plight of several species, notably the Irrawaddy dolphin, has increased greatly in many areas over this time. Since its inception, however, Reeves and his colleagues (2005) report that bycatch reduction still requires interventions involving both socioeconomic and technological changes in the area. There is still a major need for comprehensive monitoring and documentation of fishing effort and bycatch, through monitoring of high-risk fleets with onboard observers and landing-site interviews. 13.5.4 AIDCP Program The Agreement on the International Dolphin Conservation Program (AIDCP) is a legally binding instrument for dolphin conservation and ecosystem management in the ETP. The objectives of the agreement are to reduce incidental dolphin mortalities in the tuna purse-seine fishery through the setting of annual limits, the identification of alternative means of capturing large yellowfin tuna not associated with dolphins, and ensuring the longterm sustainability of tuna stocks and marine resources in the ETP. The AIDCP entered into force on February 15, 1999. To date, Costa Rica, Ecuador, El Salvador, Guatemala, Honduras, Mexico, Nicaragua, Panama, Peru, United States, Vanuatu, and Venezuela have ratified the AIDCP. Bolivia, Colombia, and the European Union are applying the agreement provisionally. Since its enactment in 1972, the provisions of the MMPA have resulted in greatly reduced annual

Measures to Reduce Interactions of Marine Megafauna with Fishing Operations dolphin bycatch in the tuna purse-seine fishery in the ETP. By the early 1980s, only a few U.S. vessels remained in the fishery as a result of MMPA prohibitions on encircling dolphins. However, foreign participation in the ETP fishery continued to increase, and for many years dolphin mortality was managed under the voluntary International Dolphin Conservation Program (IDCP) supported by the Inter-American Tropical Tuna Commission (IATTC). In 1992, the nations participating in the ETP tuna fishery convened at the annual meeting of the IATTC and signed the La Jolla Agreement (http://www.ioseaturtles.org/index.php), which placed voluntary limits on the maximum number of dolphins that could be incidentally killed annually in the fishery. It was agreed that this bycatch limit would be lowered each year over 7 years with a goal of eliminating dolphin mortality in the fishery completely. In 1995, the United States and the governments of Belize, Colombia, Costa Rica, Ecuador, France, Honduras, Mexico, Panama and Spain, whose vessels fish for tuna in the ETP, negotiated the Panama Declaration. The Panama Declaration (http://www.ioseaturtles.org/index.php) established conservative species/stock-specific annual dolphin mortality limits and represented an important step toward reducing bycatch in this fishery. The signatory nations envisioned that, as a result of their actions in reducing dolphin mortality, the United States would amend its laws so their participation in the IDCP would satisfy comparability requirements of the MMPA and result in the lifting of embargoes on yellowfin tuna and its products. 13.5.5 UN Resolution 44/225 On December 22, 1989, the General Assembly of the United Nations expressed alarm at the overexploitation of living marine resources of the high seas by driftnets and the likelihood that driftnet fishing was having an adverse impact on the marine resources of the exclusive economic zones of adjacent coastal states, with particular reference to the catch of marine mammals. It unanimously adopted Resolution 44/225 recommending that all members of the United Nations agree to a moratorium on all large-scale pelagic driftnet fishing on the high seas by June 30, 1992. Implicit in the resolution is an understanding that such a measure would not be

323

imposed in a region or, if implemented, could be lifted, in the event of effective conservation and management measures being taken. This had to be based on statistically sound analysis carried out jointly by concerned parties with an interest in the fishery resources of the region. During 1990, support for this resolution was given by the International Whaling Commission (IWC) in July; by the heads of government of the South Pacific Forum Nations in August; and at the International Conference on the Conservation and Management of the Living Resources of the High Seas, attended by representatives of fifteen States in September. Also in September 1990, the Fisheries Commission of the Organization for Economic Co-operation and Development (OECD) gave their support to the implementation of the UN resolution. In December 1990, the U.N. General Assembly passed resolution 45/197 reaffirming resolution 44/225 and called for its full implementation by all members of the international community. Later in 1992 under the U.N. moratorium, the EU agreed to reduce the length of driftnets used in the European fishery for tuna species to 2.5 km, leading to a phasing out of the fishery in the preceding years. The fishery was permanently closed from January 1, 2001. 13.5.6 Right Whale Protection In the United States, a Recovery Plan for the North Atlantic Right Whales has been established by NOAA Fisheries. The plan highlights the need to reduce right whale deaths in commercial fishing operations. The plan was prepared under the U.S. Endangered Species Act and provided a muchneeded update to the previous plan prepared in 1991. The remaining estimated 400 whales in the population decimated by historical whaling are now threatened primarily by entanglement in fishing gear and by ship strikes. Recommendations include modifying fishing gear, reducing overlap between fishing operations and the whales, and the establishment of a multistakeholder Take Reduction Team (NOAA 2004). Similarly a Canadian North Atlantic Right Whale Recovery Plan has been established by the Department of Fisheries and Oceans (DFO) to reduce the number of entanglements in fishing gear as an important strategy for right whale recovery.

324

Contemporary Issues in Capture and Conservation in Marine Fisheries

“Strategy B” within the plan recommends a bycatch reporting system, the creation of a disentanglement network, and an educational program for stakeholders along with other initiatives. In August 2003, a cooperative agreement to share research, expertise, and rescue equipment in the Bay of Fundy and Gulf of Maine was reached between the U.S.-based Center for Coastal Studies and Canada’s Department of Fisheries and Oceans. 13.5.7 Protection of Sea Turtles In the United States, all six species of sea turtles occurring in its waters are protected under the Endangered Species Act of 1973. NOAA Fisheries and the U.S. Fish and Wildlife Service (USFWS) share jurisdiction for sea turtles, with NOAA Fisheries having lead responsibility for the conservation and recovery of sea turtles in the marine environment and USFWS for turtles on nesting beaches. To reduce the incidental capture of sea turtles in commercial fisheries, NOAA Fisheries has enacted regulations to restrict certain U.S. commercial fishing gears (gillnets, longlines, pound nets and trawls) that are known to produce significant bycatch of turtles. To effectively address all threats to marine turtles, NOAA Fisheries and the USFWS have developed recovery plans to direct research and management efforts for each sea turtle species. An international program for the conservation and recovery of marine turtles was also established under the Indian Ocean—South-East Asian Marine Turtle Memorandum of Understanding (IOSEA MoU). The IOSEA MoU is a nonbinding intergovernmental agreement that aims to protect, conserve, and recover marine turtles and their habitats in the Indian Ocean and Southeast Asia region. The agreement falls under the auspices of the Convention on the Conservation of Migratory Species of Wild Animals (Article IV, para. 4) and includes all turtles occurring in the two regions. The IOSEA Marine Turtle MoU implements a framework through which regional states and other concerned states can share the responsibility of protecting, conserving, and recovering depleted marine turtle populations. The Agreement Area covers 44 states (nations) and is divided into four subregions: Southeast Asia and Australia, Northern Indian Ocean, Northwest Indian Ocean, and West Indian

Ocean. The Conservation and Management Plan focuses on reducing threats, conserving critical habitat, exchanging scientific data, increasing public awareness and participation, promoting regional cooperation, and seeking resources for implementation. Due to many other pressing development issues in the region, some countries lack the resources for successful implementation of the MoU. Therefore, through the MoU, these countries are offered support and capacity-building assistance from the FAO. 13.5.8 Protection of Other Marine Megafauna Attempts to manage or conserve sharks have been few, and usually engendered by economic concerns about declining fisheries. In 1999, The Committee on Fisheries (COFI) of the FAO adopted a Voluntary International Plan of Action for the Conservation and Management of Sharks in response to concern about the global decline in numbers of many species of sharks. International plans of action (IPOAs) are voluntary and are a product of the Food and Agriculture Organization (FAO) Code of Conduct for Responsible Fisheries. The IPOA-Sharks applies to FAO member countries that contribute to the fishing mortality of sharks, rays, skates, and chimaeras (not all of which can be classed as megafauna) that are caught as either target or nontarget species. It applies to states with jurisdiction over the waters in which sharks are caught and to those whose vessels catch sharks on the high seas. The overall aim is to develop management and conservation strategies to keep total fishing mortality for each species within sustainable levels by applying a precautionary approach. In addition, states should regularly assess the status of shark stocks subject to fishing to determine whether a new shark plan is needed. Implementation of shark plans should be reviewed at least every 4 years to identify costeffective strategies for improving their effectiveness. The EU has signed up to the IPOA-Sharks on behalf of Member States but has yet to develop and adopt a shark plan of action or shark management plan. Some efforts were made in 2000 and a draft was circulated in 2001 at COFI, but due to failure to meet the IPOA-Sharks requirements, this proposal was then withdrawn. A formal plan from the EU has yet to be proposed (Fowler et al. 2004). There are other specific examples of countries that have introduced legislation to try and protect

Measures to Reduce Interactions of Marine Megafauna with Fishing Operations sharks, as described by Walker (2000). Australia has had a shark fishery since the turn of the century and has imposed restrictions on licenses and fishing methods since 1988 (Stevens 1993). The sand tiger shark (Carcharias taurus) received protected status in the Australian state of New South Wales in 1984 (Pollard 1996). Similarly in New Zealand, shark management started in 1986 over concerns of declining catch per unit effort (CPUE). In South Africa, the great white shark has been protected since 1991 (Compagno et al. 1991). In the United States, concerns about a rapidly growing shark fishery and overfishing led to a fishery management plan for the Atlantic coast in 1993 (NMFS 1993). Protected status has been given in April 1997 to five species in the United States on the Atlantic Coast: the great white shark, the whale shark, the basking shark, the sand tiger shark, and the bigeye sand tiger shark (Odontaspis noronhai). Shark fisheries along the western coast of the United States for shortfin mako (Isurus oxyrinchus) and thresher sharks (Alopias vulpinus) have been regulated by state agencies for many years. In 1989, the states of California, Oregon, and Washington enacted an interjurisdictional fishery-monitoring plan for thresher sharks (Walker, 2000). 13.6 POTENTIAL MITIGATION MEASURES As described in the recent review by Werner and his colleagues (2006), mitigation measures can be considered and organized according to whether they represented an approach (1) intended to avert contact with a fishing operation and gear altogether, (2) intended to facilitate escape from temporary capture, or (3) that required release postcapture. This review identified 55 modifications to fishing gear or methods for reducing nontarget species bycatch. It estimated that 33 of these methods were presently being used, with the remainder at development or initial application stage. Most of the techniques currently in use adopt an avoidance approach as opposed to facilitating escape or release once an animal comes into contact with fishing gear. Taking only those approaches or techniques geared toward avoiding conflicts, most operate under the principle of physically excluding animals from fishing areas or coming into contact with gears or, in the case of longline fisheries, bait. Other methods can be

325

divided according to the type of sensory detection animals use in averting contact (i.e., auditory, visual, olfactory, gustatory, tactile, or electromagnetic) (Werner et al. 2006). This review also highlighted that most mitigation technology has been applied in longline fisheries because a large number of bycatch mitigation approaches have developed exclusively to deter sea bird bycatch (not considered in this chapter) that occurs when sea birds predate on baited longline hooks as they are being set. Werner and his colleagues (2006) found evidence of very few studies that evaluated bycatch reduction methods for other marine megafauna suspected or known to perish following encounters with fishing operations, including sirenians (manatees and dugong) or large noncommercial pelagic fishes (Goodyear 1999; Milton 2001; Read et al. 2006). Of the mitigation measures identified for reducing marine megafauna bycatch, acoustic alarms, excluder devices, and simple modifications to pelagic longline gear specifically to reduce turtle bycatch are by far the most used globally. Given their widespread testing and adoption, the evolution and development of these measures are described in the following sections to illustrate some success stories in reducing bycatch using different approaches and in different fisheries. In addition, some novel emerging technologies are described, as well as, by way of contrast, some devices that have proved to be ineffective or simply do not work. 13.6.1 Acoustic Alarms Acoustic alarms can be divided into passive acoustic devices or active acoustic devices such as acoustic deterrents or “pingers.” Passive Acoustic Devices Passive acoustic devices include modifications to fishing gears that will increase the probability of detection of the gear by an echo-locating animal. Several tests have been carried out with passive acoustic reflectors, which are small rigid plastic devices with a resonant air cavity that are attached to the mesh zone of gillnets at intervals to make the gear more acoustically detectable by marine mammals such as dolphins and whales. In simplistic terms, they are designed to act as acoustic “cats’ eyes” for marine mammal species. To work

326

Contemporary Issues in Capture and Conservation in Marine Fisheries

effectively, these devices must have a significant cross section greater than approximately three wavelengths and must be omnidirectional (Goodson 1997; Goodson et al. 1994). Practical tests have been carried out in the United Kingdom with bottlenose and common dolphins and also separately in the Albacore tuna drift net fishery, now closed under EU regulations. No significant results, however, were reported from these devices. The Natal Shark Board in South Africa also deployed passive reflectors of the type tested in the United Kingdom on beach protection shark nets and reported that they were effective for a short period in reducing small cetacean bycatch (SGFEN 2001). A more recent “passive porpoise deterrent” device, developed by the Aquatec Group in the United Kingdom and a prize winner in the WWF Smart Gear competition in 2007, combines acoustic reflectors with a small number of active acoustic devices. The reflectors are fitted into a gillnet every 5 m, and when an echo-locating porpoise emits a click, the reflectors transmit back a stronger echo, making the reflectors appear to the porpoise to be much larger objects than they actually are. This device will be further tested in 2008 but does seem to be a simple and cost-effective solution (WWF 2007). Another passive acoustic approach is the use of net materials with an increased acoustic reflectivity. In 2000, the Danish Institute for Fisheries Research (DIFRES) conducted experiments in the North Sea with modified bottom set gillnets coated with iron oxide. The experiments were terminated after a short period due to very low catches of the target species, and measurements of target strength suggested that the physical acoustics of the experimental nets were not significantly different from the conventional nets. More successful experiments were carried out in Eastern Canada in reducing the bycatch of harbor porpoise with barium sulfate gillnets (Trippel et al. 2003). However, it has not been ascertained if this was because of the nets’ acoustic reflectivity, increased stiffness, or greater visibility over conventional gillnets (Werner et al. 2006). Active Acoustic Devices Active acoustic deterrents, or pingers, are small self-contained, battery-operated devices that emit regular or randomized acoustic signals, at a range

of frequencies, and typically loud enough to alert or deter animals from the immediate vicinity of fishing gear. These devices are distinct from much higherpowered systems that are used to protect aquaculture sites from seal predation, classified as acoustic harassment devices (AHDs). AHDs emit sounds of such high intensity they cause pain or harm to animals’ hearing and are not seen as a solution to bycatch reduction, given the potential harmful effects. Active pingers were first tested in Canada, primarily as a means to reduce entrapment of baleen whales in coastal gillnets and fish traps. These first devices, operated at 2.5 kHz, were subsequently tested on gillnets in the Bay of Fundy and appeared to reduce harbor porpoise bycatch. Similar pingers were also deployed in the Makah salmon fishery off the Washington coast and in Australia on beach protection nets with reasonable results (SGFEN 2001). More complex devices were developed after experiments with gillnets in the Gulf of Maine (Kraus et al. 1997). A design operating at 10 kHz was found to be effective at reducing porpoise bycatch and ultimately formed the basis for legislation under NFMS regulations. In the regulations the specifications for porpoise pingers were defined as 300-ms pulses of 10-kHz tonal pulses repeated at 4-s intervals with a minimum source level of 132 dB re 1 μPa. A third-generation pinger was developed in the late 1990s at Loughborough University in the United Kingdom on the basis of tests with captive porpoises in Holland and Denmark. These “PICE97” devices were trialed successfully in the Danish cod fishery during the autumn of 1997, with a significant reduction in harbor porpoise bycatch observed (Larsen 1999). The new pingers emitted a variety of wide band frequency sweep type signals with randomized interpulse intervals, rather than simple single tonal pulses. Acoustic devices are now used in many gillnet fisheries; currently there are five recognized manufacturers of commercial pingers, although other “cruder” devices exist. Two of these devices are made in the United States, one in the United Kingdom, one in Italy, and one in the Netherlands. These devices have varying technical specifications as shown in Table 13.1. While their signal

Measures to Reduce Interactions of Marine Megafauna with Fishing Operations Table 13.1.

327

Specifications of available deterrent models October 2005 (Cosgrove et al., 2005).

Manufacturer

Airmar

Aquatec

Fumunda

Savewave

STM-Products

Website

www.airmar. com

www.netPinger. net

www.fumunda. com

www.savewave. net

www.stmproducts.com

Model

Gill net pinger

AQUAmark 100

FMDP-2000

Dolphin Saver— High Impact System

DDD02F

Mitigation use

bycatch

bycatch

bycatch

depredation

bycatch

Dimensions: L × dia (mm)

156 × 53

164 × 58

152 × 46

202 × 67 × 42

210 × 61

Weight in air (g)

408

410

230

400

905

Max. depth

275

200

200

200

200

Attachment details

3-way holes each end

2 holes each end

3-way holes each end

2 holes each end

1 hole at end

Spacing along nets (m) (max recommended)

100

200

100

200

200

Signal human audible

Yes

No

Yes

No

Yes

Housing material

Plastic Alloy

Urethane

Co-polymer

HIPS Styrosun

Urethane

Battery type and number

1 D-Cell Alkaline

1 D-Cell Alkaline

1 lithium

1 Sealed 9v unit

NiMh Rechargeable

Approx. battery life (months)

>12

15

<3

90 hours

Battery replaceable

Yes

No

Yes

No

Yes

Battery disposal

by operator

20% discount on replacements

by operator

20% discount on replacements

NA

Wet switch

No

Yes

Yes

Yes

Yes

Tonal/Wide band

Tonal

Wide band / tonal

Tonal

Wide band

Random

Source level (dB re 1μPa @ 1m)

132 +/− 4dB

140

132 +/− 4dB

155

160–174

Frequency (kHz)

10

20–160

10

5–160

Pulse duration (ms)

300

200–300

300

200–900

Inter-pulse period(s)

4

16–24

4–30

4

4–30

1–250 100 ms—7 s Random

328

Contemporary Issues in Capture and Conservation in Marine Fisheries

characteristics are well suited for harbor porpoises, only limited success has been achieved with other cetacean species. For species such as bottlenose dolphins, tests have shown them to be rather ineffective (Anon 2006). Fishermen in a number of countries, particularly in Europe, have raised concerns about the resilience of the current commercially available “pingers” and also the practicalities of using these devices for commercial fisheries. These concerns have been addressed in a series of trials carried out in Ireland, the United Kingdom, Sweden, Denmark, and France in 2005 and 2006 (Cosgrove et al. 2005). As a result of this work, all available models of pinger have now been extensively assessed in terms of ease of use, resistance to damage, and long-term running costs. The trials have highlighted a number of serious issues and difficulties relating mainly to the reliability of the devices. Problems with deployment were found, although some of these problems have been resolved by changes to rigging or operating practice. It is clear that more consideration of the construction, practical handling and deployment of such devices are required before they can be considered a universal solution to certain bycatch problems in gillnet fisheries. Costs associated with pingers also remain an issue for fishermen. The current commercially available devices cost in the range of b50 to 100 per device, meaning a vessel fishing with 10 km of gillnet gear, using the recommended spacing between devices of 100 to 200 m, would require 50 to 100 devices at a cost in the region of b2500 to b5000. Given the reliability problems with battery life demonstrated in the experimental work carried out in Europe, these costs are significant and currently are a hindrance to acceptance by fishermen. One other issue that has been found with the use of pingers has been the monitoring of devices on deployed gear by fishery inspectors. Control and enforcement agencies in a number of countries have indicated that current regulations are practically unenforceable given the difficulties in testing whether devices are operational or whether fishermen have them attached to gear. A Danish company (Etec) has developed a long-distance control device, in cooperation with the Federal Research Institute for Rural Areas, Forestry and Fisheries (vTI),

Institute for Baltic Sea Fisheries, Germany, to determine compliance by enabling Fishery Inspection vessels to check for pinger use on actively fishing nets (ICES 2008). This device is now commercially available and would seem a useful tool to assist in not only controlling the use of pingers but also as a way of allowing researchers to monitor pinger signals and cetacean activity in the vicinity at long distances during experimental studies in real-time. With respect to pinnipeds, studies with acoustic devices in gillnet fisheries have given conflicting results. Field experiments in a California drift net fishery for swordfish (Xiphias gladius) and sharks indicated that pingers significantly reduced the bycatch of Californian sea lions (Zalophus californianus) by up to one-third (Barlow and Cameron 2003). In the salmon bottom gillnet fishery in Northern Washington, however, Gearin and his colleagues (2000) reported that bycatch of harbor seals was not significantly reduced in gillnets fitted with pingers. In the artisanal gillnet fishery in Argentina, trials with pingers to reduce the bycatch of Franciscana dolphins suggested that the pingers had the opposite effect on sea lions common in the same area. The sea lions were observed using the “pingers” as “dinner bells.” As a result, predation of fish by the dolphins from gillnets increased (Bordino et al. 2002). Work in Australia investigating whether acoustic signals produced by pingers have a deterrent effect on dugongs so far has been inconclusive (Anon. 2003). Whereas considerable effort has been put into devising methods for minimizing cetacean bycatch in gillnet fisheries, much of the work to develop mitigation measures to minimize cetacean and pinniped bycatch in mobile gear fisheries is still at a development phase. Preliminary trials with standard gillnet pingers in the United Kingdom suggest that deploying them around the mouth of pelagic trawls was ineffective. Follow-up trials with similar pingers set farther back in the trawl were carried out in Ireland, but the results, again, were inconclusive. Tests with the Italian DDDO2F device carried out by the French Research Institute for Exploration of the Sea (IFREMER) and by the University of St Andrews (USTAN) in the United Kingdom have shown a deterrent effect on common dolphins when

Measures to Reduce Interactions of Marine Megafauna with Fishing Operations used in a pelagic trawl fishery for bass. A 40% reduction in bycatch was observed compared with trawls not fitted with these devices. These results are only preliminary and based on a small number of observed hauls, and because this device has a very high source level close to the point, which causes permanent or temporary auditory threshold shift (deafness), suitability as a deterrent device is questionable (Anon. 2006, 2007). More sophisticated acoustic devices have been developed by Bord Iascaigh Mhara (BIM) in Ireland. In conjunction with Loughborough University, they have developed a remotely triggered acoustic device specifically for use in pelagic trawls (BIM 2002). This device could be activated whenever there was deemed to be a risk that dolphins are suspected of being in imminent danger of capture. To prevent habituation to the device and save on battery power, the system was designed to be remotely operated via an acoustic link from the wheelhouse of the fishing vessel. As with previous experiments, though, due to the sporadic nature of cetacean bycatch in the test fishery (Albacore tuna), no conclusive evidence was found as to whether the device did deter dolphins. It also became clear, during the course of the sea trials, that the remote operation of the system was impractical from a commercial perspective. The positioning and emissions used were very much based on a subjective understanding of cetacean behavior in and around pelagic gear. These trials were thus deemed to provide only an indication as to whether acoustic deterrents were a technically feasible solution to the problem of bycatch in pelagic trawl fisheries and not as a quantitative assessment. Recognizing the knowledge deficit of the effect of deterrent signals on different cetacean species, DIFRES in cooperation with the University of Southern Denmark conducted a series of experiments to compare the effectiveness of different acoustic signals as a means of deterring common dolphins (Delphinus delphis) from a sound source (Anon. 2007). The experiments were conducted in Spanish waters of the Mediterranean Sea (Alboran Sea) on board the research vessel RS Toftevaag. The experiments took advantage of the bow-riding behavior of the dolphins to secure semicontrolled conditions for the experiments. The bow-riding dol-

329

phins were exposed to a number of acoustic signals (tonal, frequency sweep, noise, and a control) and their reactions were monitored using both acoustic and video recordings. Preliminary results showed that sound intensities of approximately 140 dB root-mean-square (rms) re 1 μPa at 1 m did not cause the animals to react, whereas at 150 dB rms re 1 μPa at 1 m, there were observed responses. These sound levels are comparable to those of commercially available pingers. None of the signals used (tonal, sweep, or noise) resulted in a “freak-out” response from the animals. Considering the distance between the animals and the transducer in the experiments (less than 5 m), this seriously raises questions on the deterrent effect of some of the pingers currently available on the market, particularly on common and bottlenose dolphins. DIFRES has also looked at the possibility of observing the behavior of dolphins inside pelagic trawls using a Dual frequency IDentification SONar (DIDSON). Initial trials were conducted at the basin of Fjord and Belt, Kersteminde, Denmark, in August 2005, where images of captive animals were successfully recorded. The DIDSON unit was then deployed on R/S Toftevaag in the Alboran Sea, and some interesting observations were made of pilot whales, as shown in Figure 13.1. It was concluded that this technique seems feasible for observing marine mammal behavior within pelagic trawls, but more work is required to establish whether the DIDSON unit provides good (i.e., recognizable) images of cetaceans at sufficient range (20–30 m) within a trawl net (Anon. 2007). IFREMER in France has begun to investigate the possibility of adapting existing gear monitoring equipment, such as netsounder or headline sensors commonly mounted on the headlines of pelagic trawls, to produce a deterrent signal. This work is being carried out in conjunction with the French IX Trawl Company, with the aim of developing a wireless multifrequency sensor (netsounder, 75– 200 kHz). IX Trawl has already developed an operational device called a “headline sensor” that provides images of the seabed and footrope by using a frequency compatible with the wide-band expected. This sensor could also act as a “cetacean deterrent.” A study started at the beginning of 2005

330

Contemporary Issues in Capture and Conservation in Marine Fisheries

10.0

10.0

9.0

9.0

8.0

8.0

7.0

7.0

6.0

6.0

5.0

5.0

4.0

4.0

3.0

Figure 13.2. The French IX Trawl “Cetasaver” acoustic deterrent device. (Anon 2007.)

3.0

2.0

2.0

1.0 1.0 meters

Figure 13.1. Observations of pilot whales taken with the DIDSON unit on board the R/S Toftevaag in the Alboran Sea. (Anon 2007.)

had shown that a large part of the opening of a typical commercial French pelagic trawl (80 m wide and 40 m depth) can potentially be swept by a 190 dB (re 1 μPa at 1 m) transmission, which IFREMER thought was sufficient to deter cetaceans from entering the trawl. The IX Trawl “Cetasaver” device (Fig. 13.2) produces signals over a range of frequencies from 30 to 150 kHz, with pulse duration of 200 ms at an interval of 2 to 5 s. Trials with this new device are ongoing, but initial results are encouraging, with dolphin bycatch recorded in 4 hauls (6 animals) of 121 observed with the devices fitted, compared with 10 hauls (20 animals) of 129 observed hauls without any acoustic devices (ICES 2008). The concept of interactive devices in which deterrent sounds are triggered by the sonar clicks of approaching animals has also been considered. This concept addresses frequently voiced concerns that

extensive local use of pingers could result in “noise pollution” and habituation by the animals they were intended to exclude from fishing operations. Battery life would also be increased as the device would only trigger in the presence of an echo-locating animal. The Fjord & Belt Centre (FBC) of Denmark in cooperation with Kolmarden Sweden initially carried out tests of this concept using two captive harbor porpoises (SGFEN 2001). Aquatech Subsea Ltd in the United Kingdom supplied a prototype device for the trial that was controlled interactively using a computer device. The results were encouraging and further field tests in Denmark in the waters around the island of Fyn, Denmark, showed that the concept worked in that the device consistently replied to the echo-locating animals. Under an EU-funded project, BIM in Ireland subcontracted Aquatech Subsea Ltd to develop a similar device for pelagic trawls that emits an acoustic signal only when activated by the echolocation signals of cetaceans. This new Pelagic Trawl Interactive Pinger (Fig. 13.3) emits a wide-band deterrent signal and distinguishes cetacean noises from other acoustic emissions associated with fishing gear or fish-finding equipment by recording and analyzing click interval and click length. Enabling the pinger to ignore any source of noise other than an echo-locating cetacean was fundamental to its utility. Trials of this device were

Measures to Reduce Interactions of Marine Megafauna with Fishing Operations

Figure 13.3. Aquatech Subsea interactive acoustic deterrent device. (Anon 2007.)

carried out at a dolphinarium at the Kolmarden wild animal park in Sweden with bottlenose dolphins in March 2005 and at sea on wild bottlenose dolphins and common dolphins off the Irish coast during 2006 and 2007. These initial trials have demonstrated that the interactive system worked as anticipated but the deterrent sounds produced have only shown an effect on bottlenose dolphins and not common dolphins. Further tests are planned by BIM to overcome this problem by varying the frequency, pulse duration, and source level. Other Acoustic Devices Other, technologically simpler “acoustic-based” methods have been used to deter cetaceans from fishing gears. In Japan and Tunisia, oil- or waterfilled steel tubes, manually struck at intervals with a hammer, have been used in purse-seine fisheries with limited success (SGFEN 2001). This deterrent is claimed to have an effective range of 1 km, but dolphins appear to habituate quickly to the sound so that the method ends up becoming an attractant rather than a deterrent. There are also a number of reports of waterproof fireworks being used in artisanal fisheries, particularly in the Mediterranean (SGFEN 2001). This method is used to keep cetaceans away from fishing gear, using disturbance caused by the explosions. Ethically and legally such a practice is highly questionable and should be strongly discouraged.

331

It has been suggested that marine mammal bycatch occurs in trawls during hauling. One way to avoid this might be to close the codend acoustically prior to hauling. This has been proved operationally possible (Pennec and Woerther 1993) but has yet to be tested on a commercial fishing vessel. In South Africa, an arc-discharge transducer has been tested to deter cape fur seals (Arctocephalus pusillus) in the hake trawl fishery and in purseseines. The device produced underwater compression and sound levels (132 dB rms re 1 μPa at 1 m) similar to those of a 0.303-caliber rifle bullet hitting the water. The device was reported to be moderately successful at reducing the fur seal bycatch in trawls but was not effective in deterring seals from a seine net before it was pursed (Shaughnessy et al. 1981). In New Zealand, standard fish farm seal scarers were mounted on a towable paravane designed to be deployed astern of trawlers and transmit an arc of acoustic sound of 100 degrees. As configured during the trials, however, this array was not effective at deterring fur seals and the trials were abandoned (Tilzey et al. 2004). In summary, acoustic solutions are among the most widely used devices for the reduction of marine mammal bycatch. While they have proved to be effective and practical for certain fisheries and certain species, more work is needed to adapt them for other fisheries and other species, to ensure the acoustic signals produce do deter cetaceans from fishing gears. 13.6.5 Exclusion Devices Excluder devices are usually made of rigid grids or mesh barriers installed within the extension piece of trawl nets with an opening for escape at either the top or bottom of the net. They are now widely used to reduce the bycatch of unwanted fish and sea turtles in shrimp trawl fisheries. Modifications of such devices have been tested in many other trawl fisheries to reduce marine mammal bycatch based on the same principles. In the 1970s, the capture of five species of sea turtles (Kemp’s ridley, hawksbill, leatherback, green, and loggerhead) in shrimp trawl fisheries in the southeastern United States was perceived as a problem. Initially, time or area closures were considered as a potential solution, but due to concerns

332

Contemporary Issues in Capture and Conservation in Marine Fisheries

Escape opening

Guiding panel

Grid

Codend

Figure 13.4. An example of rigid turtle excluder device (TED). (Food and Agriculture Organization of the United Nations 2002.)

about the adverse impacts on fishermens’ income, work began in the 1980s on an alternative approach using gear modifications to reduce turtle bycatch by NMFS scientists and the shrimp industry. The basic TED comprises a rigid or semirigid grid placed in the extension piece of a trawl, which has an opening in the bottom or top sheets to allow a turtle caught in the net to escape (Fig. 13.4). Shrimp and other small fish pass through the bars of the grid and are retained in the codend of the trawl. Since development began in the United States, TED design has been constantly evolving, with modifications made to account for differences between fisheries and turtle species. Clark and his colleagues (1991) and, more recently, an FAO document by Shiode and Tokai (2004), summarize different designs of TED currently in use. Since the late 1980s, TEDs became compulsory in the United States and have spread to other countries following U.S. regulations that require nations exporting shrimp to the United States to introduce TEDs in their shrimp trawl fleets. In Australia, given that U.S.-style TEDs were too large for Australian trawl gears, a “soft” TED, which lacks the metal frame used in the U.S. TED, was developed. Flexible and soft grids were subsequently introduced that retained the characteristics of the conventional TED but addressed operational and safety issues specific to the Australian fisheries. These devices have proved effective not only at releasing turtles

but also for large sharks and rays (Shiode and Tokai 2004). In addition to the United States and Australia, TEDs are now used in Southeast Asian countries including Thailand, Malaysia, and the Philippines, mainly through the initiative of the Southeast Asian Fisheries Development Centre (SEAFDEC). In addition, TEDs are being tested, and their use is encouraged in Mexico, Belize, Guatemala, Honduras, El Salvador, Nicaragua, Costa Rica, and Panama in Latin America, as well as India in Asia, and Kenya, Nigeria, and other countries in Africa. NOAA in the United States has been able to show that TEDs are effective at excluding up to 97% of sea turtles with minimal loss of shrimp catch (MTCP 2004) and the introduction of TEDs to shrimp fisheries must be considered a real success story. Future improvements and refinements will concentrate on the survival of turtles that escape through TEDs. Rigid exclusion grids similar in design to TEDs have been used successfully to eject marine mammals in trawl fisheries. There has been considerable work in Australia and New Zealand to minimize the bycatch of pinnipeds, fur seals, and sea lions, in both squid and hoki midwater trawl fisheries through the use of exclusion grids (Cawthorn and Starr, in press; Gibson and Isaksen 1998). Initial trials began in the early 1990s, in an effort to reduce sea lion bycatch in the New Zealand trawl

Measures to Reduce Interactions of Marine Megafauna with Fishing Operations squid fishery and fur seal bycatch in the hoki fishery off the west coast of the South Island of New Zealand. In 2000 through 2003, trials with seal excluder devices (SEDs) were carried out off of western Tasmania in the blue grenadier trawl fishery. The SED was designed to prevent fur seals from passing through the extension piece of the net into the codend and to facilitate the ejection of the seals through an escape panel. Initial designs resulted in excessive fish loss via the seal escape hatch, blocking of the device during fishing, and from seal entry into the net via the escape hatch. However, these problems now appear to have been overcome by using a “top escape hatch” design. In Australia and New Zealand, excluder devices appear to be an effective means of preventing fur seal mortality, particularly in large midwater trawls (Anon. 2003). The University of St. Andrews (USTAN) has conducted sea trials with several modified gears tested for the pelagic trawl fishery for bass in the English Channel and Bay of Biscay (Anon. 2006). A solid steel grid (70 kg) at the front of the extension piece of the trawl was initially tested but proved to be too heavy to use in smaller boats. It was more easily managed in larger boats where a power block was installed. Some indication of a reduction in cetacean bycatch of approximately 20% was observed with this device, although the number of hauls observed with bycatch was relatively small. A tubular steel grid proved much easier to handle in both small and large vessels and again achieved some positive results, and underwater footage of dolphins escaping from the device were obtained. Questions, though, have subsequently been raised with respect to the survival of escaping animals given the distance between the mouth of the trawl and the position of the excluder device. A flexible grid was later tested at sea in the same fishery but became easily distorted, resulting in unacceptable fish losses. Various escape hatches have also been tested in combination with the grids. To date, no conclusive evidence has been obtained to confirm that these devices reduce cetacean bycatch and underwater footage obtained during these experiments has indicated that cetacean behavior within pelagic trawls is not uniform across individuals.

333

An EU-funded project entitled CETASEL aimed to test the effects of a series of ropes hung within the pelagic trawl net to determine if such ropes would prevent the entry of dolphins farther into the net. Following extensive experiments with captive animals, sea trials were conducted with the excluder panel in April 1997 in waters off South Ireland and in the Bay of Biscay. In the experiments carried out, ropes were placed in two parallel rows, 2.5 m apart, and seven escape holes were placed in the top sheet of the trawl with openings of 25 × 8 m. The results, however, were inconclusive due to a lack of interactions with cetaceans, but the panel appeared technically feasible (De Haan et al. 1998). Other netting exclusion devices have been deployed in Mauritania, northwest Africa, by Dutch freezer trawlers and proved effective in eliminating large organisms such as hammerhead sharks and manta rays but not cetacean species (Zeeberg et al. 2006). The modification tested comprised a filter grid positioned in the aft section of a pelagic trawl connected to an escape tunnel. The filter grid slopes downward at an angle of 20 degrees, which forces larger nontarget species such as sharks and manta rays downward to a tunnel entrance. A cetacean exit window was also positioned in front of the filter grid, although this did not seem to be effective. This work has been followed up by RIVO in the Netherlands as part of an EU-funded studied entitled NECESSITY. Extensive testing of two cetacean barriers (rope and tunnel barrier) in the front part of a pelagic trawl was carried out in 2005 but was largely inconclusive, with losses of marketable fish catch being a limiting factor. Similar rope exclusion panels tested by USTAN in the bass fishery, and placed at the point in the trawl where the large meshes of the fore part of the net join into smaller meshes leading to the codend of the trawl, gave negative results and proved impossible to handle without entangling the gear. These grids also resulted in no target fish catch. A 20-cm mesh Dyneema panel at the same place was found to be more manageable but increased drag greatly and reduced fish catch by 90%. Three other excluder device modifications evaluated for their efficacy by IFREMER had discouraging results during tests undertaken in 2004 and 2005 (Anon. 2006).

334

Contemporary Issues in Capture and Conservation in Marine Fisheries

• A 300-mm (bar length) square mesh tilted panel, fitted in the baitings of the trawl developed by a French net maker and tested by IFREMER • A vertical barrier of 300 mm (bar length) square mesh placed in the body of the pelagic trawl (in the part constructed in 100-mm half mesh) • A semi-rigid oval grid fitted in the extension piece of pelagic trawl One type of escape/excluder panel that is successfully used is the medina panel in the purse-seine fishery for yellowfin tuna in the ETP. This is a panel of small-mesh netting attached to the purse-seine at the farthest distance from the boat when the net is “pursed.” The mesh is small enough that dolphins are unlikely to be entangled, and assists them to escape over the top of the net. These panels are used in conjunction with a “back down” procedure developed during which the purse-seine is towed backward, lowering the height of the float line from the surface and facilitating the escape of dolphins over the top (Werner et al. 2006). Gear modifications also offer a potential solution for mitigating seal-induced damage that is prevalent in some static gear fisheries, notably the Baltic salmon trap fishery (see further discussion in Chapter 7). In 2003 and 2004, five trap modifications and two traditional models were tested in the Bothnian Sea to investigate whether seal-induced damage could be markedly reduced by making specific modifications in net design and by using different netting materials. In traditional nonprotected traps, 30% to 40% of the total salmon catch had observed seal damage (Suuronen et al. 2005). In the modified traps, the percentage of fish that were damaged varied between 1% and 30%, depending on the model. The marked differences in seal exclusion and catching efficiency resulted from the design of the fish bag and the construction and rigging of the funnels and middle chambers. In traditional traps, a substantial part of all salmon caught were captured in the meshes of the middle chambers; these fish are extremely vulnerable to seal predation. In modified traps, the middle chambers were made of thicker and stiffer PE-twine, and the proportion of salmon caught in the chambers was negligible. Clearly, thick and stiff netting in the chambers effectively prevents the meshing of fish

and thereby reduces their vulnerability to seal predation. Trials showed that the fish bag of a trap can be modified to effectively protect the catch from seals. However, more work is needed to find practical and effective mitigation measures to prevent predation by seals in the middle chambers of the trap. A large-mesh chamber would allow salmon chased by seals to escape and avoid the predation but it may cause substantial catch losses and seal entanglement (Suuronen et al. 2005). Following identification of a turtle bycatch problem in U.S. scallop fisheries, the fishing industry in the United States has been very proactive in trying to solve this particular sea turtle interaction problem. The industry has worked with Coonamessett Farm and the Virginia Institute of Marine Science (VIMS) to develop turtle chains, which are similar to rock chains using extensively in scallop dredge fisheries. Operationally, the various kinds of turtle chains are similar with the only differences being the chain material and details of hanging. There are only limited observations of turtle encounters with dredges but Coonamessett Farm designed a new dredge frame modification based on fishermen’s observations that dredges pass either under or over the sea turtles. While turtle chains prevent sea turtles from getting trapped inside the dredge, there was a perceived need for a frame modification to prevent entrapment on top of the dredge. A new frame design since developed is a significant departure from existing designs in that the cutting bar is moved forward of the depressor plate so that instead of confronting a vertical structure, a sea turtle or large fish (e.g., skate) encounters a sloping structure (Fig. 13.5). The design then increases the width of the depressor plate and extends the struts, at 30.5cm spacing, between the depressor plate and the forward positioned cutting bar. Thus, a sea turtle cannot become trapped in this space and is gently guided over the dredge (Smolowitz 2006). In summary, the testing of many of these net panel excluder devices and grids illustrates some of the problems researchers have faced in developing mitigation devices in trawl fisheries. The results have been largely inconclusive due to the sporadic nature of bycatch in the fisheries tested. Indications from this research do suggest, though, that they seem to have limited potential for cetacean species

Measures to Reduce Interactions of Marine Megafauna with Fishing Operations

335

Figure 13.5. New excluder frame design for preventing turtle bycatch in scallop dredges. (Smolowitz 2006.)

3"

8cm

2½"

7cm 6cm

2" 1½" 1"

5cm 4cm 3cm 2cm

½" 1cm

Figure 13.6. Differences in shape and design of traditional J-shaped hooks (first four from left to right) to circle hooks.

(Anon 2006) but are more appropriate for release of seals and sea lions (Cawthorn and Starr, in press). Other technical difficulties, such as incorporating them into large and complex pelagic trawls used in the fisheries, excessive losses of target species, and handling difficulties, are serious drawbacks. On the other hand, the medina panel used in the yellowfin tuna purse-seine fishery is used successfully, whereas the gear modification developed for the Baltic salmon trap fishery to reduce seal bycatch and for the U.S. scallop fisheries to prevent turtle bycatch show very real potential. 13.6.6 Circle Hooks Techniques to avoid and minimize interactions of longlines with sea turtles have being actively investigated in recent years following international pressure from NGOs to ban pelagic longline fisheries

due to accidental capture of sea turtles (Gilman et al. 2005). Most experiments carried out to date have small sample sizes and have been conducted over only a few seasons and in a small number of fisheries. Nonetheless, modifications to fishing gear, including hook size and shape, have been tested with some success. Circle hooks, in which the point of the hook is perpendicular to the hook shank, have been shown to reduce the hooking rate of sea turtles. Figure 13.6 shows the difference between a circle and traditional J-shaped hook. Experiments in the U.S. North Atlantic swordfish fishery have shown that large circle hooks (4.9-cm width) effectively reduce sea turtle bycatch rates and reduce the proportion of turtles that swallow the hook compared with smaller traditional J-hooks. These large circle hooks were effective without compromising commercial viability for some target species, notably

336

Contemporary Issues in Capture and Conservation in Marine Fisheries

swordfish. There are indications of broad uptake by global longline fisheries where use of the large hook is viable in terms of catch rates for the target species. Watson et al. (2004) found that nonoffset 4.9-cmwide circle hooks with squid bait reduced loggerhead and leatherback turtle captures by 74% and 75%, respectively, compared with conventional J-hooks. Studies completed to date in Hawaii, however, comparing small circle hooks (5.1-cm width or smaller) with smaller (4.1-cm width or smaller) J-shaped hooks found no significant difference in turtle bycatch rates, and the circle hooks had mixed results on the capture rate of target species. Smaller circle hooks, however, did reduce the incidence of deep hooking by turtles, similar to large circle hooks. It is still not well understood how the position of hooking affects post release survival of turtles, but it is possible that deeper hooking associated with J-hooks may increase the chances of gear remaining in the animal after it is released. Several other strategies, including setting gear below depths where turtles are in relatively high densities, using fish instead of squid bait, single hooking fish versus threading the hook through the bait multiple times, reduced gear soak times during the daytime, fleet communications protocols, and area and seasonal closures may also be effective and are currently undergoing assessment in longline fisheries globally. Results of completed research and potential methods still undergoing assessment provide cautious optimism that sea turtle bycatch mortality can be reduced substantially (Gilman et al. 2006). However, one negative aspect of circle hooks has been reported by the ICES Study Group on the bycatch of protected species (SGBYC), which found that the introduction of circle hooks into longline fisheries has reportedly resulted in a bycatch of pilot whales due to the fact that these hooks are stronger than conventional J-hooks (ICES 2008).

13.6.7 Whale Entanglement The entanglement of North Atlantic right whales in lobster and gillnet buoy lines is another highprofile fishing gear interaction due to the small population of this species. Accordingly, scientists in the United States and Canada (DFO 2003) have

been working with the fishing industry to experiment on alternative gear types and rigging to reduce entanglement. Materials used for buoy ropes were examined (Werner et al. 2006). These include research into using break-away lines that are designed to break at strengths substantially lower than normal, ropes and buoy line messenger systems that send down a stronger hauling line down, with a weak line messenger rope that would attach to the bottom gear, allowing gear retrieval. The intent is for the rope to be strong enough to haul the gear but break if a whale becomes entangled. Lipid-soluble rope that dissolves once embedded in the blubber of a large whale has also been considered, although no suitable material has been found as yet to allow testing. In addition, stiff and sinking/weighted ropes that rest on the seabed where they are less likely to entangle whales in the water column have also been tested and implemented in some fisheries. More sophisticated mitigation measures involve using galvanic or acoustic release devices, which secure the buoy lines on the bottom until the release device dissolves or is acoustically triggered, freeing a buoy that brings the buoy line to the surface. Working on the same principle, a time tension line cutter, which is a link connecting the bottom gear and the buoy lines that would break under any pressure sustained longer than the time it takes to haul the gear, has also been tested. Similarly, a line-cutting device, which will detach a surface buoy from the buoy line when pressure is exerted against a plate that is attached to the buoy, has also undergone preliminary testing. All of these methods seem viable solutions but require monitoring of whale entanglements to determine which ones provide the most practical solutions. 13.6.8 Other Methods Other bycatch mitigation solutions include operational changes to fishing practices or involve time/ area closures to avoid “hotspots” of conflicts between fishing and nontarget catch. Time and area closures, however, are only effective if the spatial or temporal frame is large enough to encompass a suitably high proportion of bycatch events. It is therefore necessary to know spatial and temporal (seasonal) distribution of bycatch events and for

Measures to Reduce Interactions of Marine Megafauna with Fishing Operations bycatch to be not just a transient or random occurrence, unless the closures are in real-time. Spatial and temporal closures of fisheries have been effective in certain circumstances; for instance, the Banks Peninsula Marine Mammal Sanctuary that was created in 1988 to protect Hector ’s Dolphins from gillnets in New Zealand (Read et al. 2006). The NOAA system of real-time closures referred to as dynamic area management (DAM) zones, which temporarily restrict the use of lobster pots and anchored gillnets on an expedited basis to protect right whales, has also proved to be very successful. In this system, a DAM zone is triggered by a single reliable report from a qualified individual of three or more right whales within an area (75 nm2) such that right whale density is equal to or greater than 0.04 right whales/nm2. There are, however, examples where such closures have not worked. For example, Zollet (2005) reported that higher bycatch rates of harbor porpoise in 1994 in the New England multispecies gillnet fishery resulted from a poorly designed area closure network instigated by NMFS. It was reported that bycatch of cetaceans occurs mostly at night or at dawn and dusk by Morizur et al. (1999). While this is not the case in all fisheries, this information could be used in promulgating mitigation strategies. Experiments in the Hawaiian swordfish and tuna longline fisheries have shown that setting hooks at night results in lower seabird bycatch because under reduced light baited hooks are less visible to seabirds (Boggs 2003; McNamara et al. 1999). However, there is no evidence that night sets have been advantageous in other fisheries to reduce bycatch of mammals or sea turtles. There have been several other suggestions that haulback procedure, offal-discarding practices, deck lighting arrangements, and the use of certain sonar equipment may all contribute to increasing cetacean bycatch probability. These ideas, however, have yet to be tested rigorously. The use of emetics, noxious, or scented bait types to induce a conditional food aversion has been tested in aquaria and on fish farms. The idea is to make the bait unpalatable to “pest” species so that they occur less frequently in the area of an operation. Results from trials involving Stellar sea lions, California sea lions, and Australian fur seals

337

(Arctocephalus pusillus doriferus) have been mixed (Anon. 2003). A recent novel method uses electromagnetic fields created in the vicinity of a fishing activity to deter interaction of sharks with fishing gear, bait, or target species. This idea won a major prize in the 2006 WWF Smart Gear competition. It takes advantage of sharks’ unique biology to deter them from taking the bait of longline hooks. Sharks are able to detect magnetic fields using special organs located on their snouts, and research has revealed that some species of shark are repelled by strong magnetic fields (Werner et al. 2006). 13.7 CONCLUDING REMARKS Marine megafauna bycatch reduction is a very active area of research and there are a number of successful examples where technical measures have greatly reduced fisheries impacts. The use of TEDs in many tropical shrimp fisheries has reduced the mortality rates of several turtle species, whereas modification of hook shape has reduced bycatch associated with longline fisheries. Operational changes in purse-seine and trawl fisheries have also been reported successful in some fisheries, notably in the ETP yellowfin tuna purse-seine fishery, to reduce dolphin bycatch. The sporadic nature of bycatch for larger marine mammals, however, makes accurate assessment of the nature and extent of bycatch as well as the development of solutions problematic. A better understanding of the behavioral interactions of marine mammals with fishing gears is therefore needed, as many mitigation measures are based on a conjectural understanding of behavior. While some mitigation measures are widely tested and used, such as pingers, there is still a need to fully assess the cost implications of bycatch reduction technology before they are introduced into legislation. Modifications to buoy ropes such as weak links or sinking/weighted ropes have been effective in allowing whales to break free or avoid buoy lines. In pelagic longline fisheries, large circle hooks have proved to reduce the bycatch and mortality of sea turtles. In certain fisheries, fishing gear may act as mammal attractor and thus there is a high likelihood of interactions. This behavior needs to be understood and taken into account when designing

338

Contemporary Issues in Capture and Conservation in Marine Fisheries

mitigation measures. The use of acoustic sonar could provide large time frame observations of marine mammal behavior in the vicinity of fishing gear. To date, most available techniques have been directed at reducing the bycatch of cetaceans, pinnipeds, and sea turtles. Measures specifically designed to reduce bycatch of large sharks, particularly in pelagic longline fisheries where the reported bycatch is high, are urgently required. Measures to reduce bycatch of large fish species such as manta ray and sawfish should also be investigated in fisheries or areas were problems are identified. As the global move to an ecosystem approach to fisheries management gathers pace, research into developing mitigation measures to protect these charismatic species will be of increasing importance with an onus on fishery scientists, stakeholders, and managers to develop practical solutions to avoid widespread closures of fisheries. The success stories cited in this field suggest that this can be achieved, particularly with active participation of fishermen. REFERENCES Alava MNR and Yaptinchay A. 2000. Whale Sharks in the Philippines. Presented at Shark Conference 2000. Honolulu, Hawaii, February 21–24, 2000. Anon. 2003. Technologies to reduce seal-fisheries interactions and mortalities. Agenda Item 5.1. Southern and Eastern scalefish and shark fishery ecological advisory group. Final report to of the Special SESSFEAG Meeting: Reducing seal interactions and mortalities in the South East Trawl Fishery. 81–99. Anon. 2006. Nephrops and cetacean species selection information and technology. EU Project NECESSITY. No. 501605. Interim Report. Anon. 2007. Nephrops and Cetacean species selection information and technology. EU Project NECESSITY. No. 501605. Final Draft Report. Baird RW. 1994. A program to monitor the status of small cetaceans in British Columbia. Pages 122–130 in Proceedings of the Pacific Ecozone Workshop, February 1–3, 1994, Institute of Ocean Sciences, Sidney, B.C. Canadian Wildlife Service Technical Report Series No. 222. Baird SJ. 1996. Nonfish Species and Fisheries Interactions Working Group Report. May 1996. Wellington, NZ: Ministry of Fisheries: 27 pp + Appendices.

Bannister JL. 2001. Status of Southern Right Whales (Eubalaena australis) off southern Australia. J. Cet. Res. Manag. (Sp. Iss.) 2: 103–110. Barlow J and Cameron GA. 2003. Field experiments show that acoustic pingers reduce marine mammal bycatch in the California drift gillnet fishery. Presented to the Scientific Committee of the International Whaling Commission, Grenada, 1999. Bennett PM and Jepson PD. 2000. Cetacean strandings investigation: England and Wales and Poseiden Database. 78 pp. London: Institute of Zoology. Boggs C.H. 2003. Annual Report on the Hawaii Longline Fishing Experiments to Reduce Sea Turtle Bycatch under ESA Section 10 Permit 1303. US National Marine Fisheries Service Honolulu Laboratory, Honolulu, 42 pp. Bordino P, Kraus S, Albareda D, Fazio A, Palmeiro A, Mendez M and Botta S. 2002. Reducing incidental mortality of Franciscana dolphin (Pontoporia blainvillei) with acoustic warning devices attached to fishing nets. Mar. Mam. Sci. 18: 833–42. Browne A, Welsford D and Lyle J. 2005. Monitoring marine mammal interactions in the small pelagics fishery: stage one pilot study. Rep. Aust. Fish. Manag. Auth. 19 pp. Campana SE, Marks L, Joyce W and Kohler NE. 2006. Effects of recreational and commercial fishing on blue sharks (Prionace glauca) in Atlantic Canada, with inferences on the North Atlantic population. Can. J. Fish. Aquat. Sci. 63: 670–682. Clark J, Griffin W and Richardson J. 1991. Simulated economic impact of TED regulations on selected vessels in the Texas shrimp fishery. Mar. Fish. Rev. 53: 1–8. Collet A and Mison V. 1996. Analyse des echoutages de cetaces sur le littoral Francais. Brest, IFREMER: Appendix 6: 1–32. Compagno LJV, Ebert DA and Cowley PD. 1991. Distribution of offshore cartilaginous fish (class chondrichthyes) off the west coast of southern Africa, with notes on their systematics. South Afr. J. Mar. Sci. 11, 43–139. Cosgrove R, Browne D and Robson S. 2005. Assessment of Acoustic Deterrent Devices in Irish Gill Net and Tangle Net Fisheries. BIM Report 05MT07. 30 pp. Couperus AS. 1997. Bycatch of marine mammals and discards in pelagic fisheries (MMADIS): CFP 94/018. Report to the European Commission: 1–21. Crespo EA, Corcuera J and Lopez Cazorla A. 1994. Interactions between marine mammals and fisheries

Measures to Reduce Interactions of Marine Megafauna with Fishing Operations in some fishing areas of the coast of Argentina. Rep. Int. Whaling Comm. (Sp. Iss.) 15: 283–290. Crespo EA, Alonso MK, Dans SL, Garcia NA, Pedraza SN, Coscarella MA and Conzalez R. 2000. Incidental catch of dolphins in mid-water trawls for southern anchovy off Patagonia. J. Cet. Res. Manag. 2: 11–16. Crespo EAS, Pedraza SL, Dans SL, Alonso MK, Reyes LM, Garcia NA, Coscarella M and Schiavini ACM.1997. Direct and indirect effects of the highseas fisheries on the marine mammal population in the Northwest and Central Patagonian Coast. J. Northw. Atl. Fish. Sci. 22: 189–207. Dans SL, Crespo EA, Alonso MK and Pedraza SN. 2003. Incidental catch of dolphins in trawling fisheries off Patagonia, Argentina: are populations sustainable? Ecol. App. 13: 754–762. Dans SL, Crespo EA, Garcia NA, Reyes LM, Pedraza SN and Alonso MK. 1997. Incidental mortality of Patagonian dusky dolphins in mid-water trawling: retrospective effects from the early 1980s. Rep. Int. Whaling Comm. 47: 699–703. De Haan D, Dremiére PY, Woodward B, Kastelein RA, Amundin M and Hansen K. 1998. Prevention of the bycatch of cetaceans in pelagic trawls by technical means, CETACEL. Final Report, Contract AIR III-CT94-2423, 1994–1997. 204 pp. DFO. 2003. Improved Canada-USA Cooperation To Protect Endangered North Atlantic Right Whales. Press Release August 22, 2003. FAO. 1995. Code of Conduct for Responsible Fisheries. Rome: FAO. 41 pp. FAO. 1998. Report of the FAO Technical Working Group on the Conservation and Management of Sharks. FAO Fish. Rep. 583. 28 pp. FAO. 2002–2009. Fishing Equipment. Turtle Excluder Device (TED). Text by V. Crespi and J. Prado. In: FAO Fisheries and Aquaculture Department [online]. Rome. Updated 28 November 2008. http:// www.fao.org/fishery/equipment/ted/en. Fertl D and Leatherwood S. 1997. Cetacean interactions with trawls: a preliminary view. J. Northw. Atl. Sci. 22: 219–248. Fowler S, Mogensen CB and Blasdale T. 2004. Plan of Action for the Conservation and Management of Sharks in UK Waters. August 2004. JNCC Report No. 360. Peterborough: JNCC, 2004 ISSN 0963-8091. Gearin PJ, Gosho ME, Laake JL, Cooke L, Delong RL and Hughes KM. 2000. Experimental testing of acoustic alarms (pingers) to reduce bycatch of harbour porpoise, (Phocoena phocoena), in the

339

state of Washington. J. Cet. Res. Manag. 2: 1561–1713. Gibson D and Isaksen B. 1998. Functionality of a fullsized marine mammal excluder device. Sci. Conserv. 81: 19. Gilman EL, Watson JW, Boggs C, Epperly S, Zollett E, Beverly S, Nakano H, Swimmer Y, Davis K, Shiode D, Dalzell P and Kinan I. 2005. Review of the state of knowledge for reducing sea turtle bycatch in pelagic longline gear. Rep. W. Pac. Reg. Fish. Manag. Council. 24 pp. Gilman E, Zollet E, Beverly S, Nakano H, Davis K, Shiode D, Dalzell P and Kinan I. 2006. Reducing Sea turtle bycatch in pelagic longline fisheries. Fish Fish. 7: 2–23. Goodson AD. 1997. Developing deterrent devices designed to reduce the mortality of small cetaceans in commercial fishing nets. J. Mar. Fresh. Behav. Physiol. 29: 211–236. Goodson AD, Kilnowska M and Bloom PRS. 1994. Enhancing the acoustic detectability of fishing nets. Rep. Int. Whaling Comm. (Sp. Iss.) 15: 585–595. Goodyear CP. 1999. An analysis of the possible utility of time-area closures to minimize billfish bycatch by US pelagic longlines. Fish. Bull. 97: 243–255. ICES. 2008. Report of the Study Group for Bycatch of Protected Species (SGBYC), 29–31 January 2008, ICES, Copenhagen, Denmark. ICES CM 2008/ACOM: 48. 88 pp. Kim ZG. 1999. Bycatches of minke whales in Korean waters. J. Cet. Res. Manag. 1 (Suppl.): 98–100. Kleiber P. 1999. Estimates of Marine Mammal Takes in the Hawaiian Longline Fishery. Honolulu, HI: NMFS Southwest Fisheries Science Center. Knowlton AR and Kraus SD. 2001. Mortality and serious injury of northern right whales (Eubalaena glacialis) in the western North Atlantic Ocean. J. Cet. Res. Manag. (Sp. Iss.) 2: 193–208. Kraus SD. 1990. Rates and potential causes of mortality in North Atlantic right whales (Eubalaena glacialis). Mar. Mam. Sci. 6: 278–291. Kraus SD, Read AJ, Solow A, Baldwin K, Spradlin T, Anderson E and Williamson J. 1997. Acoustic alarms reduce porpoise mortality. Nature 388: 525–526. Kuiken T, Simpson VR, Allchin CR, Bennett PM, Codd GA, Harris EA, Howes GJ, Kennedy S, Kirkwood JK, Law RJ, Merrett NR and Phillips S. 1994. Mass mortality of Common Dolphins (Dephinus delphis) in South-west England due to incidental capture in fishing gear. Vet. Rec. 134: 81–89.

340

Contemporary Issues in Capture and Conservation in Marine Fisheries

Larsen F. 1999. The effect of acoustic alarms on the bycatch of harbour porpoises in the Danish North Sea set gill net fishery: a preliminary analysis. Paper SC/51/SM41 presented at the 51st IWC meeting, Grenada. Marine Turtle Conservation Program (MTCP). 2004. Turtle Excluder Device (TED) Chronology. Charleston, SC: SC Dept. Natural Resources. McNamara B, Torre L and Kaaialii G. 1999. Hawaii Longline Seabird Mortality Mitigation Project. Honolulu: U.S. Western Pacific Regional Fishery Management Council. 108p. Milton DA. 2001. Assessing the susceptibility to fishing of populations of rare trawl bycatch: sea snakes caught by Australia’s northern prawn fishery. Biol. Conserv. 101(3): 281–290. Morizur Y and Trengenza N. 1996. Bycatch and discarding in pelagic trawl fisheries (BIOECO). Plouzane: IFREMER: 121 pp + 7 Appendices. Morizur Y, Berrow SD, Tregenza NJC, Couperus AS and Povreau S. 1999. Incidental catches of marine mammals in pelagic trawl fisheries of the northeast Atlantic. Fish. Res. 41: 297–307. Murray K. 2004. Bycatch of Sea Turtles in the MidAtlantic Sea Scallop (Placopecten magellanicus) Dredge Fishery during 2003. NEFSC Ref. Doc 0411. 28 pp. NMFS. 1993. Fishery management plan for sharks of the Atlantic Ocean. Florida, USA, National Marine Fisheries Services, National Oceanic and Atmospheric Administration, Newport, Oregon, February 4–6, 1992. Seattle, WA: Natural Resources Consultants. NOAA. 2001. Endangered and Threatened Species; Proposed Endangered Status for a Distinct Population Segment of Smalltooth Sawfish (Pristis pectinata) in the United States 66 FR 19414. April 16, 2001. Online. HTTP://WWW.NMFS.NOAA. G O V / P R / S P E C I E S / F I S H / S M A L LT O O T HSAWFISH.HTM. (ACCESSED JANUARY 20, 2007). NOAA NATIONAL MARINE FISHERIES SERVICe. NOAA. 2004. NOAA Fisheries Releases Recovery Plan for Endangered North Atlantic Right Whales: Critically Endangered Right Whale Numbers Less than 300. Press Release, August 2004. Online. http://www.epa.gov/fedrgstr/EPA-SPECIES/2004/ June/Day-01/e12356.htm (accessed February 21, 2007). NOAA National Marine Fisheries Service. Northridge SP. 1991. An updated world review of interactions between marine mammals and fisheries. FAO Fish. Tech. Pap. 734 (Suppl. 1): 58 pp.

Northridge SP. 1996. A review of marine mammal bycatch observer schemes with recommendations for best practice. Joint Nature Conservation Committee Report No. 219. Northridge SP. 2002. Report on preliminary trials of a dolphin exclusion device in the UK bass pair trawl fishery. St. Andrews, UK: Univ. of St. Andrews. Northridge SP. 2003. Reduction of cetacean bycatch in pelagic trawls. Final Report to DEFRA & JNCC, Project MF0733. Northridge SP and Hofman RJ. 1999. Marine mammal interactions with fisheries. In: Twiss JR Jr and Reeves RR (eds). Conservation and Management of Marine Mammals. pp 99–119. Washington, DC: Smithsonian Institution Press. Northridge SP and Thomas L. 2003. Monitoring levels required in European fisheries to assess cetacean bycatch, with particular reference to the UK fisheries. Final Report to DEFRA (EWD). 37 pp. Pace DS, Pulcini M and Triossi F. 2003. Interactions with Fisheries: Modalities of opportunistic feeding for Bottlenose Dolphins at Lampedusa Island. Proc. 17th Annual Conference of the European Cetacean Society (Las Palmas de Gran Canaria, Spain, March 9–13, 2003). Pennec S and Woerther P. 1993. Fermeture telecommandre de cul de chalut. EQUINOXE 46. Perrin WF, Wursig B and Thewissen JGM. 2002a. The Tuna-Dolphin Issue. In: Encyclopedia of Marine Mammals. pp 1269–1273. San Diego, CA: Academic Press. Perrin WF, Reeves RR, Dolar MLL, Jefferson TA, Marsh H, Wang JH and Estacion J. 2002. Report of the Second Workshop on the Biology and Conservation of Small Cetaceans and Dugongs of Southeast Asia. Second draft, October 2002. Pollard DA. 1996. The biology and conservation status of the grey nurse shark (Carcharis Taurus Rafinesque 1810) in New South Wales, Australia. Aquat. Conserv. Mar. Freshw. Ecosys. 6: 1–20. Read A. 2000. Potential mitigation measures for reducing the bycatches of small cetaceans in ASCOBANS waters: A Report to ASCOBANS: 30. Read A, Drinker P and Northridge S. 2006. Bycatch of Marine Mammals in US and global fisheries. Conserv. Biol. 20: 163–169. Reeves RR, Berggren P, Crespo EA, Gales N, Northridge SP, di Sciara GN, Perrin WP, Read AJ, Rogan E, Smith BD and Van Waerebeek K. 2005. Global priorities for reduction of cetacean bycatch. Report for the World Wildlife Fund. 29 pp.

Measures to Reduce Interactions of Marine Megafauna with Fishing Operations Shaughnessy PD, Semmelink A, Cooper J and Frost PGH. 1981. Attempts to develop acoustic methods of keeping Cape fur seals (Arctocephalus pusillus) from fishing nets. Biol. Conserv. 21: 141–158. Shiode D and Tokai T. 2004. A review of development, modification and implementation of TED (turtle excluder device) to reduce sea turtle bycatch in trawl fisheries. FAO Fish. Rep. 738 (Suppl.): 238 pp. SGFEN. 2001. Incidental catches of small cetaceans. Report of the first meeting of the subgroup on fishery and environment (SGFEN) of the scientific, technical, and economic committee for fisheries (STECF), Brussels, 10–14 December 2001. Commission Staff Working Paper, Commission of the European Communities. SEC (2002) 376: 83 pp. Smolowitz R. 2006. Sea scallop harvest gear: engineering for sustainability. Mar. Technol. Soc. J. 40(3): 25–31. Stevens JD. 1993. The status of chondrichthyan resources in the southwest Pacific. Report of the southwest Pacific sub-group of the IUCN Shark Specialist Group. Unpublished report. Suuronen P, Siira A and Kauppinen T. 2005. Design principles of seal-safe salmon trap-net. Presented at ICES ASC, Aberdeen, April 2005. Tilzey R, Goldsworthy S, Cawthorn M, Calvert N, Hamer D, Kirkwood R, Russel S, Shaughnessy P, Wise B and Stewardson C. 2004. Assessment of seal-fishery interactions in the winter blue grenadier fishery off west Tasmanian and the development of fishing practices and Seal Exclusion Devices to mitigate seal bycatch by factory trawlers. Final report to the Fishery Research Development Corporation, Project 2001/008: 70 pp.

341

Tobayama T, Yanagisawa F and Kasuya T. 1992. Incidental take of minke whales in Japanese trap nets. Rep. Int. Whaling Comm. 42: 433–436. Trippel EA, Holy NL, Palka DL, Shepherd TD, Melvin GD and Terhune JM. 2003. Nylon barium sulphate gillnet reduces porpoise and seabird mortality. Mar. Mam. Sci. 19: 240–243. Walker T. 2000. Assessment and management requirements to ensure sustainability of harvested shark populations. Presented at Shark Conference, Honolulu, HI, February 21–24, 2000. Waring G T, Josephson E, Fairfield CP and Foley KM. 2006. U.S. Atlantic and Gulf of Mexico marine mammal stock assessments-2005. NOAA Tech. Memo. NMFS-NE-194: 352 pp. Watson J, Foster D, Epperly S and Shah A. 2004. Experiments in the Western Atlantic Northeast distant waters to evaluate sea turtle mitigation measures in the pelagic longline fishery. Report on experiments conducted in 2001–2003. Pascagoula, MS: National Marine Fisheries Service. Werner T, Kraus S, Read A and Zollet E. 2006. Fishing techniques to reduce the bycatch of threatened marine animals. Mar. Technol. Soc. J. 40(3): 50–68. WWF. 2007. Special UK Prize Winner: The Passive Porpoise Deterrent. World Wildlife Fund. Online: http://www.smartgear.org/smartgear_winners/ smartgear_winner_2007/smartgear_winner_2007s pecial/index.cfm (accessed March 2007). Zeeberg J, Corten AD and de Graff E. 2006. Bycatch and release of pelagic megafauna in industrial trawl fisheries off Northwest Africa. Fish. Res. 78: 186–195. Zollet EA. 2005. A review of cetacean bycatch in trawl fisheries. Literature review prepared for the Northeast Fisheries Science Centre. Woods Hole, MA: National Marine Fisheries Service.

342

Contemporary Issues in Capture and Conservation in Marine Fisheries

SPECIES MENTIONED IN THE TEXT Albacore tuna, Thunnus alalungus Atlantic humpback dolphin, Sousa teuszii Australian fur seal, Arctocephalus pusillus doriferus basking shark, Cetorhinus maximus bigeye sand tiger shark, Odontaspis noronhai blue shark, Prionace glauca bottlenose dolphin, Tursiops truncates bowhead whale, Balaena mysticelus Burmeister ’s dolphin, Phocoena spinipinnis Californian sea lion, Zalophus californianus cape fur seal, Arctocephalus pusillus Caribbean manatee, Trichechus manatus Comerson’s dolphin, Cephalorhynchus commersonii common dolphin, Delphinus delhpis dugong, Dugong dugon European sea bass, Dicentrarchus labrax false killer whale, Pseudorca crassidens flatback turtle, Natador depressus Franciscana dolphin, Pontoporia blainvillei Fraser ’s dolphin, Lagenodelphis hosei green turtle, Chelonia mydas grey seal, Halichoerus grypus grey whale, Eschrichtius robustus hammerhead shark, Sphyrna mokarran harbor porpoise, Phocoena phocoena harbor seal, Phoca vitulina harp seal, Phoca groenlandica hawksbill turtle, Eretmochelys imbricate

humpback dolphin, Sousa chinensis humpback whale, Megaptera novaeangliae Irrawaddy dolphin, Orcaella brevirostris Kemp’s ridley turtle, Lepidochelys kempii leatherback turtle, Dermochelys coriacea loggerhead turtle, Caretta caretta mako shark, Isurus oxyrinchus manta ray, Manta birostris Mediterranean monk seal, Monachus monachus minke whale, Balaenoptera spp. New Zealand fur seal, Arctocephalus forsteri northern fur seal, Callorhinus ursinus northern right whale, Eubalaena glacialis northern sea lion, Eumetopias jubatus oceanic whitetip shark, Carcharhinus longimanus olive ridley turtle, Lepidochelys olivacea pantropical spotted dolphin, Stenella attenuate porbeagle shark, Lamna nasus sand tiger shark, Carcharias taurus shortfin mako shark, Isurus oxyrinchus short-finned pilot whale, Globicephalas melas silky shark, Carcharhinus falciformis smalltooth Sawfish, Pristis pectinata southern right whale, Eubalaena australis spinner dolphin, Stenella longirostris swordfish, Xiphias gladius thresher shark, Alopias vulpinus tiger shark, Galeocerdo cuvier whale shark, Rhincodon typus white shark, Carcharodon carcharias yellowfin tuna, Thunnus albacares

Appendix Species Names Mentioned in the Text

black rockfish, Sebastes schlegeli black sea bass, Centropristis striata blue cod, Parapercis colias blue grenadier, Macruronus novaezelandiae blue marlin, Makaira mazarra blue shark, Prionace glauca blueback herring, Alosa aestivalis bluefin tuna; bluefin, Thunnus thynnus bluefish, Pomatomus saltatrix bluegill sunfish, Lepomis macrochirus bottlenose dolphin, Tursiops truncates bowhead whale, Balaena mysticelus brown shrimp, Crangon crangon brown trout, Salmo trutta bullet tuna, Auxis rochei bullhead, Cottus gobio bumper, Chloroscombrus chrysurus Burmeister ’s dolphin, Phocoena spinipinnis butterfish, Peprilus triacanthus butterfly fish, Chaetodon sp.

African electric eel, Gymnarchus niloticus Alaskan pollock; walleye pollock, Theragra chalcogramma albacore tuna; albacore, Thunnus alalungus alewife, Alosa pseudoharengus American eel, Anguilla anguilla American lobster, Homarus americanus American plaice, Hippoglossoides platessoides American shad, Alosa sapidissima anchovy, Ancoa hepstus angelfish, Pterophyllum scalare arabesque greenling, Pleurogrammus azonus Atlantic bumper, Chlorscombrus chrysurus Atlantic cod; cod, Gadus morhua Atlantic croaker, Micropogonius undulates Atlantic flying fish, Cheilopogon heterurus Atlantic halibut, Hippoglossus hippoglossus Atlantic herring; herring, Clupea harengus Atlantic humpback dolphin, Sousa teuszii Atlantic mackerel; mackerel, Scomber scombrus Atlantic puffin, Fratercula arctica Atlantic salmon, Salmo salar Australian fur seal, Arctocephalus pusillus doriferus

Californian sea lion, Zalophus californianus cape fur seal, Arctocephalus pusillus capelin, Mallotus villosus Caribbean manatee, Trichechus manatus carp; common carp, Cyprinus carpio Caspian kilka, Clupeonella spp. catfish, Anarhichas lupus channel catfish Ictalurus punctatus China rockfish, Sebastes nebulosus chinook salmon, Oncorhynchus tshawytscha chub mackerel, Scomber japonicas

Baltic cod, Gadus morhua Baltic herring, Clupea harengus barracuda, Sphyraena pinguis basking shark, Cetorhinus maximus bastard halibut, Paralichthys olivaceus bigeye sand tiger shark, Odontaspis noronhai bigeye scad, Selar crumenophthalmus bigeye tuna, Thunnus obesus

343

344

Appendix

chum salmon, Oncorhynchus keta cichlid fish, Astronotus ocellatus cod; Atlantic cod, Gadus morhua coho salmon, Oncorhynchus kisutch Comerson’s dolphin, Cephalorhynchus commersonii common carp; carp, Cyprinus carpio common dab; dab, Limanda limanda common dolphin, Delphinus delhpis common murre, Uria aalge common starfish, Asterias rubens conger eel, Conger verreauxi cornet fishes, Fistsularia sp. crimson sea bream, Evynnis japonica croaker, Micropogonias undulates dab; common dab, Limanda limanda dogfish; spiny dogfish, Squalus acanthias dolphinfish, Coryphaena hippurus Dover sole; sole, Solea solea dragonet, Callionymus sp. dugong, Dugong dugon eel; European eel, Anguilla anguilla estuarine prawn, Nematopalaemon hastatus European dab, Limanda limanda European eel; eel, Anguilla anguilla European sea bass, Dicentrarchus labrax European shad, Alosa sapidissima European whitefish, Coregonus lavaretus false killer whale, Pseudorca crassidens fathead minnow, Pimephales promelas flatback turtle, Natador depressus flathead, Platycephalus bassensis flounder, Platichtys flesus forkbeard, Phycis phycis Franciscana dolphin, Pontoporia blainvillei Fraser ’s dolphin, Lagenodelphis hosei frigate tuna, Auxis thazard goby, Padagobius martensii goldfish, Carassius auratus green sturgeon, Acipenser medirostris green turtle, Chelonia mydas Greenland halibut, Reinhardtius hippoglossoides grey mullet, Mugil cephalus grey seal, Halichoerus grypus grey whale, Eschrichtius robustus gudgeon, Gobio gobio

haddock, Melanogrammus aeglefinus hake, Merluccius merluccius halibut, Paralichthys olivaceus hammerhead shark, Sphyrna mokarran harbor porpoise, Phocoena phocoena harbor seal, Phoca vitulina harp seal, Phoca groenlandica harvest fish, Peprilus alepidotus hatchet fish, Gasteropelecus sp. Hauraki Gulf snapper, Pagrus auratus hawksbill turtle, Eretmochelys imbricate herring; Atlantic herring, Clupea harengus hoki, Macruronus novaezelandia humpback dolphin, Sousa chinensis humpback whale, Megaptera novaeangliae Irrawaddy dolphin, Orcaella brevirostris jack mackerel; Japanese jack mackerel, Trachurus japonicus; Trachurus symmetricus jacopever, Sebastes schlegeli Japanese common squid, Tadorodes pacificus Japanese eel, Anguilla japonica Japanese jack mackerel; jack mackerel, Trachurus japonicus Japanese mackerel; chub mackerel, Scomber japonicus kawakawa, Euthunnus affinis Kemp’s ridley turtle, Lepidochelys kempii king crab, Paralithodes camtschaticus king prawn, Melicertus latisulcatus lake sturgeon, Acipenser fulvescens lake whitefish, Coregonus lavaretus large scale mullet, Liza macrolepis leatherback turtle, Dermochelys coriace lemon shark, Negaprion brevirostris leopard shark, Triakis semifasciata ling, Molva molva lingcod, Ophiodon elongatus loggerhead turtle, Caretta caretta longspine porgy, Stenotomus caprinus louvar, Luvarus imperialis mackerel; Atlantic mackerel, Scomber scombrus mako shark, Isurus oxyrinchus manta ray, Manta birostris

Appendix masu salmon, Oncorhynchus masou Mediterranean monk seal, Monachus monachus milkfish, Chanos chanos minke whale, Balaenoptera spp. moray eel, Muraena helena mullet, Mugil auratus New Zealand dory, Cyttus novaezelandiae New Zealand fur seal, Arctocephalus forsteri North Sea plaice, Pleuronectes platessa North Sea pollack, pollock, Pollachius pollachius Northeast Arctic cod, Gadus morhua northern fur seal, Callorhinus ursinus northern right whale, Eubalaena glacialis northern sea lion, Eumetopias jubatus Norway lobster, Nephrops norwegicus Norway pout, Trisopterus esmarki ocean shrimp, Pandalus jordani oceanic whitetip shark, Carcharhinus longimanus olive ridley turtle, Lepidochelys olivacea opah, Lampris guttatus oyster toadfish, Opsanus tau Pacific barracuda; barracuda, Sphyraena argentea Pacific bonito, Sarda chiliensis Pacific cod, Gadus macrocephalus Pacific halibut, Hippoglossus stenolepis Pacific herring, Clupea pallasii Pacific mackerel; Japanese mackerel, Scomber japonicus Pacific saury, Cololabis saira paddlefish, Polyodon spathula pantropical spotted dolphin, Stenella attenuate Patagonian scallop, Zygochlamys patagonica Patagonian toothfish, Dissostichus eleginoides paua, Hailiorrtis iris pearly finned cardinal-fish, Apogon poecilopterus perforated-scale sardine; white sardinella, Sardinella albella pigfish, Orthopristis chrysopterus pike, Esox lucius pikeperch, Sander lucioperca; Stizostedion lucioperca pilchard, Sardinops neopilchrdus pinfish, Lagodon rhomboides pink shrimp, Pandalus borealis pistol shrimp, Alpheoidea spp.

345

plaice, Pleuronectes platessa pollock; saithe, Pollachius virens porbeagle shark, Lamna nasus puffer fish, Lagocephalus wheeleri rabbitfish, Siganus fuscescens rainbow trout, Oncorhynchus mykiss rainbow wrasse, Coris julis red hake, Urophycis chuss red king crab, Paralithodes camtschaticus red mullet, Mullus barbatus red sea bream, Pagrus major red snapper, Lutjanus campechanus red snappers, Lutjanus malabaricus and L. erythropterus redfish, Sebastes marinus rhinoceros auklet, Cerorhinca monocerata right whale, Eubalaena glacialis Risso’s dolphin, Grampus griseus rock sole, Pleuronectes bilineatus Ross coral, Pentapora foliacea rough scad, Trachurus lathami round herring, Etrumeus teres round scad, Decapterus punctatus roussette, Scyliorhinus sp. sablefish; blackcod, Anoplopoma fimbria sailfish, Istiophorus platypterus saithe, pollock, Pollachius virens salmon, Salmo salar sand eel, Ammodytes tobianus sand flathead, Platycephalus bassensis sand tiger shark, Carcharias taurus sand whiting, Sillago ciliata sandeels, Ammodytidae sp. sandstar, Astropecten irregularis sardine, Sardinops melanosticta scads, Decapterus spp. scaled sardine, Harengula jaguana scallop, Pecten maximus sea bream, Pagrus major sea catfish, Arius felis sea pens, Penatula phosphorea; Virgularia mirabilis; Funiculina quadrangularis seabass; sea bass, Dicentrarchus labrax seahorse, Hippocampus ramulosus shearwater, Puffinus spp. shiner perch, Cymatogaster aggregata

346

Appendix

shortfin mako shark, Isurus oxyrinchus short-finned pilot whale, Globicephalas melas shrimp; pink shrimp, Pandalus borealis silky shark, Carcharhinus falciformis silver anchovy, Engraulis eurystole silver hake, Merluccius bilinearis skate, Raja batis skates, Raja spp. skipjack tuna, Katsuwonus pelamis smalltooth sawfish, Pristis pectinata snapper, Pagrus auratus snow crab, Chionoecetes opilio sockeye salmon, Oncorhynchus nerka sole, Solea solea southern right whale, Eubalaena australis Spanish sardine, Sardinella anchovia; Sardinella aurita spider crab, Leptomithrax gaimardi spinner dolphin, Stenella longirostris spiny dogfish; dogfish, Squalus acanthias spot, Leiostomus xanthurus spotlined sardine, Sardinops melanostictus spotted warehou, Seriolella punctata sprat, Sprattus sprattus squid, Loligo pealeii striped bass, Morone saxatilis striped beak-perch, Oplegnathus fasicatus striped marlin, Tetrapturus audax striped mullet, Mugil cephalus sunfish, Lepomis sp. sunfish, Mola mola sunrise goatfish, Upeneus sulphureus swimming crab, Liocarcinus holsatus swordfish, Xiphias gladius tadpole-fish, Raniceps raniceps tanner crab, Chionoecetes bairdi

tench, Tinca tinca thread herring, Opisthonema oglinum threadfin shad, Dorosoma petenense threestripe tigerfish, Terapon jarbua thresher shark, Alopias vulpinus tiger flathead, Neoplatycephalus richardsoni tiger prawn, Penaeus esculentus tiger shark, Galeocerdo cuvier tilapia, Tilapia mossambica turbot, Scophthalmus maximus tusk, Brosme brosme vendace, Coregonus albula wahoo, Acanthocybium solandrei walleye pollock, Theragra chalcogramma walleye, Sander vitreus weakfish, Cynoscion regalis whale shark, Rhincodon typus white catfish, Ictalurus catus white hake, Urophycis tenuis white marlin, Tetrapturus albidus white shark, Carcharodon carcharias whiting, Merlangius merlangus; Gadus merlangus winter flounder, Pseudopleuronectes americanus wolffish, Anarhichas lupus wrasse, Labridae yellowfin bream, Acanthopagrus australis yellowfin tuna, Thunnus albacares yellow perch, Perca flavescens yellowtail flounder, Pleuronectes ferruginea; Limanda ferruginea yellowtail, Seriola quinqueradiata

Index Page numbers in italics refer to figures. Page numbers followed by a t refer to tables. A Abduction phase, beat cycle, 12 Aberdeen Gulf III plankton sampler, front view of, 40 AC. See Alternating current Acceleration, 10 Accommodation in fish eye image focusing and, 25 lens and, 26 Accumulated catch, fish density in codend and, 91 Acosta, A. R., 192 Acoustic alarms, 325–326, 328–331 active acoustic devices, 326, 328–331 passive acoustic devices, 325–326 specifications for, 327t Acoustic attraction, 53–54 Acoustic cameras, 133 Acoustic coupling conditioning, 54–55 Acoustic deterrents, 51, 316 Acoustic harassment devices, 326 Acoustic impedance, defined, 46 Acousticolateralis system of fish, structures in, 45 Acoustic pingers gillnets, seabird bycatch reduction and, 195 groundfish gillnets, harbor porpoise bycatch reduction and, 196 Acoustic signals, mesh size, cod traps and, 175 Acoustic sonar, 338 Acoustic stimuli, chemical stimuli vs., 111 Acoustic surveys, 133 Acoustic tags, 133 Acoustic telemetry, predicting mortality of unrestricted fish with, 275 Acoustic transponders, lost gear and use of, 198 Active acoustic devices, 326, 328–331 Adduction phase, beat cycle, 12–13 African electric eel (Gymnarchus niloticus), 13

Agreement on the Conservation of Small Cetaceans of the Baltic and North Seas, 321 Agreement on the International Dolphin Conservation Program, 322–323 AHDs. See Acoustic harassment devices AIDCP. See Agreement on the International Dolphin Conservation Program Air, acoustic impedance of, 46 Air-filled monofilament nylon nets, 51 Alanine, 112 Alarms, cod traps, 175 Alaska, salmon traps used in, 160, 160 Alaskan fisheries, 107 Alaskan pollock (Theragra chalcogramma) bottom trawl harvesting of, 299–300 condition indices of discards for, 275 crowding in purse seines and mortality of, 267 light conditions and escape mortality for, 281 mortality of those encountering but escaping fishing gear, 269 netting material and escape mortality for, 279 post-escape predation and, 281 reflex impairment in, 276 Alaska pollock trawls, 299–300 Albacore tuna (Thunnus alalungus) bycatch rates for, 317 pelagic longline fisheries and, 107 Alewife (Alosa pseudoharengus) aversive sound, nuclear power plant and, 60 maximum swimming speed relative to body length of, 21 Algivorous rabbitfish (Siganus fuscescens), detection thresholds in, 113 Alternating current, 206, 230 Alternating current field, fish behavior in, 212

347

Ambush predators, 10 American croaker (Micropogonias undulates) pound net fishery in mid-Atlantic coast (U.S.) and, 178 sounds produced by, 48 American eel (Anguilla anguilla) backward and forward swimming by, 11 fish skin of, 8 maximum amplitude in, 12 swimming specialization of, 9 American lobster (Homarus americanus), groundfish gillnets and capture of, 194 American plaice (Hippoglossoides platessoides) gillnet height and catch efficiency for, 194 mortality of those encountering but escaping fishing gear, 268 Northwestern Atlantic Ocean demersal fisheries and, 106 American shad (Alosa sapidissima) audiogram of hearing sensitivity in, 70 maximum sustained swimming speed and endurance at prolonged speeds in, 20t maximum swimming speed relative to body length of, 21 Amino acids cod sensitivity to, 113 feeding attractants and, 111–112 as major feeding stimulants for cod, 112 Anchors, gillnets and, 184 Anchovy (Ancoa hepstus), passage through extension and, 83 Angelsen, K. K., 190 Anglefish (Pterophyllum scalare) cruising by, 10 swimming specialization of, 9 Anguilliform fish, anodic electrotaxis and, 213

348 Anode zone, electrofishing, 210 Anodic electrotaxis, 213 Anraku, K., 54 Anthropogenic sounds, hearing abilities of fish and, 52 Antillean fish pots, 144, 146, 148 Apparent contrast, 36 Appledoorn, R. S., 192 Approach, to bait, 117 Aquatec Group (UK), 326 Aquatech Subsea interactive acoustic deterrent device, 331 Aquatech Subsea Ltd. (UK), 330 Arabesque greenling (Pleurogrammus azonus), fish pots for, 145 Arabian Gulf, fish pots used in, 145 Arch-discharge transducers, 331 Archival tags, 133 Argentinean industrial trawl fisheries, discard rate attributed to, 241 Arimoto, T., 37 Arkley, K., 249, 254, 257 Artisanal fishing gear, coral reef ecosystems and, 153 ASCOBANS. See Agreement on the Conservation of Small Cetaceans of the Baltic and North Seas Aspect ratio, defined, 10 Atlantic bumper (Chlorscombrus chrysurus), passage through extension and, 83 Atlantic cod (Gadus morhua), 302 audiogram of hearing sensitivity in, 70 baited gillnets and capture of, 192 behavior of, in response to approaching vessel and vessel plus trawl, 74 best hanging ratio for capture of, 185 black tunnel and underwater photograph of escape by, just ahead of the black tunnel, 40 contrast threshold in, 36 endurance probability curves for swimming endurance of, at speeds comparable to those experienced in trawl mouth, 81, 82 fish pots for, 145 French-rigging and reducing bycatch of, 300 gilling and, 184 hearing range frequency of, 51 L , mesh size and, 248 50 maximum sustained swimming speed and endurance at prolonged speeds in, 20t Newfoundland traps and, 159

Index Northeastern Atlantic Ocean demersal fisheries and, 106 Northwestern Atlantic Ocean demersal fisheries and, 106 Atlantic flying fish (Cheilopogon heterurus), swimming specialization of, 9 Atlantic herring (Clupea harengus) maximum sustained swimming speed and endurance at prolonged speeds in, 20t maximum swimming speed relative to body length of, 21 mortality of those encountering but escaping fishing gear and, 269 Newfoundland traps and, 159 number of vertebrae in, 5 Atlantic humpback dolphins (Sousa teuszii), fishing operations and threat to, 316 Atlantic mackerel (Scomber scombrus) gillnets and avoidance of, 193 glow net and blocking of, 172 maximum sustained swimming speed and endurance at prolonged speeds in, 20t mesh netting experiments with, 248 Newfoundland traps and, 159 twine color, gillnets and behavior of, 190 vertical location of, during passage through extension of trawl net, 83 Atlantic Ocean demersal fisheries Northeastern, 106 Northwestern, 106–107 Atlantic puffins (Fratercula arctica), gillnets and bycatch of, 195 Atlantic salmon (Salmo salar) audiogram of hearing sensitivity in, 70 gillnet fishing and behavior of, 189–190 hearing range frequency of, 51 traps and bycatch reduction for, 172 weirs for catching, 161 Atlantic stingray (Dasyatis sabina), detection thresholds for, 113 Attracting electrode systems, use of, 219 Attraction, to unbaited fish pots, 146–147 Auditory brainstem response (ABR) method, on auditory physiology of fish, 52 Auditory physiology of fish, research techniques on, 51–52 Auditory threshold curves, 50, 51

Australia, fish pots in, 145 Australian fur seals (Arctocephalus pusillus doriferus), unpalatable bait types and, 337 Australia red snapper fishery, bottom trawl with drop chain and weights tested in, 301 Automated unmanned fishing system, 227, 227 Automatism zone, electrofishing, 210 Aversive sound, reducing fish entrapment in cooling water intakes and, 60 Avoidance patterns fish behavior in pretrawl zone and, 70–72 fishing mortality and, 265 B BACOMA panel, 249 Baggs, Isham, 205 Bag nets, 159, 162 Scottish, 164 Bait arousal to presence of, and stimulus categorization, 115 artificial, 112 longline gear, 108, 108 presoaked, 114 shape of and response to, 118 size of, 118 species-selective effects of, 112 Baited gears, vulnerability to, variation in, 110 Baited gillnets, fish behavior and, 192 Baited hooks fish behavior and, 130 furthering understanding of fish response to, 133 selective capture of large over small individuals on, 131 Baited longline, tracks of three cod showing chemically mediated responses to, 116 Bait fishing, chemoreception, food search and, 110–114 Bait ingestion, fish behaviors related to, 117–118 Bait odor plume, diel activity rhythms and, 115 Bait odor source, locating, 115–117 Bait selection, fish pots, reducing bycatch and, 150 Bait type ecologically sustainable fisheries management and, 132 reducing incidental catches of sharks and, 128

Index

species selectivity and, 131 species selectivity in longlining and, 109 Ballistiforms, 213 Baltic cod (Gadus morhua), gillnet twine size and capture of, 186 Baltic fish traps, 159, 164–165 salmon and whitefish behavior near, 169–170 seals and, 175–177 Baltic herring (Clupea harengus) catch volume/quality and escape mortality for, 279 mortality of those encountering but escaping fishing gear, 269 Baltic Sea, grid designs to reduce interaction of seals with salmon and whitefish traps in, 176, 176 Banks Peninsula Marine Mammal Sanctuary, 337 Baranov, F. I., 184, 187 Barimo, J. F., 50 Barium sulfate gillnet reflectivity and, 196 nets enhanced with, 51 Barracuda (Sphyraena argentea) acceleration behavior in, 10 swimming specialization of, 9 Barracuda (Sphyraena pinguis), behavior of, near Japanese set nets, 167 Bary, B. M., 205, 206, 213 Basket, longline gear, 108 Basking shark (Cetorhinus maximus) major conservation problems related to, 318 protected status for, 325 Bastard halibut (Paralichthys olivaceus) auditory threshold curves for, 50, 51 as hearing generalist, 50, 50 Baster, Joe, 205 “Batwing” doors, reducing seabed impact and use of, 305–306 Beam, for beam trawls, 299 Beamish, F. W. H., 279 Beam trawls, 310 brown shrimp, electrical stimuli used in, 309 description of, 296, 298 flatfish, electrical stimuli used in, 309 otter trawl developed from, 296 schematic illustration and photograph of, 299 seabed impact of, 295–296 Beardsley, A. J., 146, 147, 149 Beat cycle, phases of, 12–13

Behavioral control systems, 38 Behavioral research methods, on auditory physiology of fish, 51–52 Belgian electrobeam trawl, 220, 221 Benthic invertebrates, cable discs used in reducing damage to, 308, 308 Benthic sharks, predation mortality and, 268 Benthic species entrance height into trawl net by, 83 herding patterns: between net mouth and trawl doors, 76–77 Benthos catch, reduced, drop-out panels used for, 310 Berkeley, S., 268 Berth, 171 Beutel, D., 254 Beverly, S., 128 Bigeye sand tiger shark (Odontaspis noronhai), protected status for, 325 Bigeye scad (Selar crumenophthalmus), fish aggregation devices and, 58 Bigeye tuna (Thunnus obesus) fish aggregation devices and, 58 pelagic longline fisheries and, 107 Billfishes (Istiophoridae), body features of, 11 BIM. See Bord Iascaigh Mhara Binoculor vision, in typical teleost fish, 30 Biofouling of derelict gillnets, 197 Bioluminescence, underwater natural light and, 35 Bioturbation mounds, beam trawling and elimination of, 295–296 Bird-scaring (streamer) line, 125, 125, 126 Bjordal, Å, 106, 147, 148 Black rockfish (Sebastes schlegeli), discard mortality of, 266 Black sea bass (Centropristis striata), 144 Black sea bream, French-rigging and, 301 Black Sea skate (Raja clavata), detection thresholds in, 113 Blake, R. W., 12 Blaxter, J. H. S., 35, 36 Bleeding, electrofishing and, 229 Blind zone, in typical teleost fish, 30 Blocking, light and illusion used in, 37–40 Blueback herring (Alosa aestivalis), maximum swimming speed relative to body length of, 21 Blue cod (Parapercis colias), fish pots for, 145 Blue-dyed baits, 128

349 Bluefin tuna (Thunnus thynnus) cruising speeds by, 10 early traps and, 160 swimming specialization of, 9 Bluefin tuna traps, 164 Bluegill sunfish (Lepomis macarochirus), noise exposure and, 52 Blue grenadier (Macruronus novaezelandiae) endurance values for swimming in mouth of trawl and, 80 vertical location of, during passage through extension of trawl net, 83 Blue marlin (Makaira mazarra), fish aggregation devices and, 58 Blue shark catch rates, soak time and, 129 Blue shark (Prionace glauca) bycatch of, 319 circle hooks and, 129 Blyth, R. E., 153 Bobbins, 297 beam trawls, 299 otter trawls, 296 Body length of fish maximum burst speed and, 19 maximum swimming speed of some marine species relative to, 21 swimming speed, gillnet fishing and, 189 Body mass, cost of transport and, 19 Body of net, otter trawls, 297 Body shape, of fish, 7 Body size, maximum sighting distance and, 33 “Boko” fishing method, acoustic attraction and, 53 Bonn Convention, 321 Bord Iascaigh Mhara (Ireland), 329 Borodino, P., 196 Bottlenose dolphin (Tursiops truncatus) acoustic pingers for gillnets and bycatch reduction of, 196 trawler fisheries and interactions with, 320 Bottom food search behavior, in cod, 112 Bottom-set longlines, demersal species targeted with, 109 Bottom trawl, “sweepless trawl,” interaction of, with seabed, 298 Bottom trawl fishing system, complete, 68 Bottom trawls, 296 collision and escapement of fish under footgear of, 81

350 Bottom trawls (Continued) defined, 67 design and engineering for, 68 herding behavior of flatfish and skates in response to sweep of, 77 historical development of, 67–68 interaction of, with seabed, 298 low light intensities and fish behavior in mouth of, 80 pelagic trawls vs., 299 seabed impact from, 295 traditional, interaction with seabed, 298 underwater noises and, 48 zones with, 68–69, 69 Bottom trawls and fish behavior, 67–95 inside trawl net and the codend (Zone 3), 82–89 entry and orientation, 82–84 fish behavior in response to bycatch reduction devices, 85–89 fish behavior inside codend, 84–85 near trawls, extrinsic factors and, 89–91 fish density, 90–91 light level, contrast, and color, 89–90 water temperature, 90 near trawls, factors with influence on, 89–95 near trawls, intrinsic factors, 91–95 fish size, 91–92 learning and experience, 93–95 motivational state, 92–93 physiological condition, 93 in pretrawl zone (Zone 1), 69–75 avoidance patterns, 72–75 reaction distance, 70–72 underwater radiated noise, 69–70 between trawl doors and in net mouth (Zone 2), 75–82 herding patterns: benthic species, 76–77 herding patterns: roundfish, 75–76 in trawl mouth, 77–82 trawl gear and trawl fisheries, 67–69 Bowhead whales (Balaena mysticelus), entanglement incidences with, 317 Braking, 10 Brazner, J. C., 127 BRDs. See Bycatch reduction devices Breen, M., 269, 270 Brewer, D., 253, 301 Bridles beam trawls, 299

Index demersal single boat trawl, 240 gillnet, 185 off-bottom, reducing seabed impact and use of, 307–308 short, reducing seabed impact and use of, 306–307 Broadhurst, M. K., 253 Bronze electrodes, 228 Brothers, N., 106 Brown shrimp beam trawls, electrical stimuli used in, 309 Brown shrimp (Crangon crangon) electrical pulse stimulation in beam trawls and, 309 electric pulse field and, 218 Brown trout (Salmo trutta), homogeneous fields of direct current and reactions of, 215 Buchanan, S., 195 Bullet tuna (Auxis rochei), fish aggregation devices and, 58 Bullhead (Cottus gobio), homogeneous fields of direct current and reactions of, 215 Bumper (Chloroscombrus chrysurus), electrotaxis induced in, voltage gradient combined with pulse rate and, 216t Buoy ropes gillnet, 185 modifications of, 337 Buoys, gillnets and, 184 Burmeister’s porpoises (Phocoena spinipinnis), fishing operations and threat to, 316 Burst-and-coast swimming behavior, 16 of cod, part of velocity curve during, 17 Burst speeds, 19 Burst-swimming response in codend, 84 entry into trawl mouth and, 83 entry into trawl net and, 83 Butterfish (Peprilus triacanthus) electrotaxis induced in, voltage gradient combined with pulse rate and, 216t pound net fishery in mid-Atlantic coast (U.S.) and, 178 Butterfly fish (Chaetodon sp.) cruising by, 10 swimming specialization of, 9 Butterfly pattern, avoidance patterns and, 73, 74 Bycatch. See also Conservation; Discards; Marine megafauna; Mortality; Survival measures

cause of, 242 of cetaceans, 316, 317 reducing, 51 discard vs., 239 of dolphins, 316, 317 gillnets and measures for reduction of, 192–195 of juveniles, longlines and reduction of, 124 of manta ray, 319 marine megafauna, extent of, 318–320 nature of problem, 320–321 of porpoises, 316 potential fate of animals entering a trawl and its relation to, 241 reducing, 38 in single-species fisheries, 250–253 of seals and sea lions, 318 of sharks and rays, 318 of shellfish, pelagic trawls and reduction of, 300 technical measures for reducing, 242–257 reducing capture of specific species or sizes in mixedspecies fisheries, 253–257 reducing target species discards by controlling size selectivity, 243–250 reducing unwanted bycatch in single-species fisheries, 250–253 of turtle species, 318 of whales, 316, 317–318 in world fisheries, 240–242 Bycatch reduction devices escape of fish through, 87 examples of, 86 fish behavior inside trawl net, codend and, 85–89 fish density and, 91 future of research on, 88–89 ideal water flow in and around, 88 tropical shrimp trawl fisheries and, 88 typical flow field throughout extension/codend containing, 85 C Cables, seabed impact and reducing cutting effect of, 307 Californian sea lion (Zalophus californianus) pingers and bycatch reduction of, 328 unpalatable bait types and, 337

Index Callahan, S., 230 Campana, S. E., 319 Canadian Fishery Consultant Ltd., 197 Cape fur seals (Arctocephalus pusillus), arc-discharge transducers and, 331 Capelin (Mallotus villosus), 92 Newfoundland traps and, 159 Capelin spawning season, gillnets, seabird bycatch and, 195 Capelin traps, 164, 172 Capture efficiency, gillnets, 184–187 Capture mechanisms, gillnets, 184–187 Capture of fish fish vision and, 25–40 herding and capture of fish by trawl under visual conditions, 36–37 light and illusion used in guiding and blocking, 37–40 light use in fishing, 37 by gillnets gilling, 184, 186 snagging, 184, 186 tangling, 184, 186 wedging, 184, 186 multitude of stimuli and responses in, 114–119 arousal to presence of bait and stimulus categorization, 115 behavior before stimulation, 115 locating bait odor source, 115–117 visual object categorization, bait ingestion, and hooking behavior, 117–119 potential factors affecting fish behavior, endurance, stress, and cumulative injury during, 277 Carangiform fish, anodic electrotaxis and, 213 Carbohydrates, swimming and use of, 14 Cardiac arrest, in narcotized fish, 229 Caribbean manatee (Trichechus manatus), fisheries and interactions with, 318 Caribbean region, fish pots from, 144 Carlile, D. W., 149, 150 Carp (Cyprinus carpio) as hearing specialist, 50, 50 learning/conditioning in, 123 swimming styles of, 12 Carr, H. A., 197 Carr, W. E. S., 112 Carretta, J. V., 196

Casale, P., 127 Caspian kilka (Clupeonella spp.), pump fishing with underwater light for, 226 Catch composition, discard mortality and, 283 Catch efficiency, conceptual trawl design, optomotor response and improvement in, 39 Catch-induced turbulence, water movement in codend and, 84 Catch per unit effort, 131, 325 Catch rates, fish pots, bait type and, 147 Catch size, discard mortality and, 282–283 Catch volume and quality, escape mortality and, 279 Catch weight, discard mortality and, 283 Catfish (Anarhichas lupus) extended hearing frequency range in, 50 fish pot approach by, 147 sounds produced by, 48 CCAMLR. See Commission for the Conservation on the Antarctic Marine Living Resources Cetacean bycatch, in trawl fisheries, 317 Cetaceans, 321 deficit of effect of deterrent signals on, 329 fishing operations and population threats to, 316–317 hearing abilities of, 51 marine megafauna and, 315 reducing bycatch of, 316 small cetaceans, 51 Southeastern Asian fisheries and, 322 “Cetasaver” acoustic deterrent device, 330 CFCL. See Canadian Fishery Consultant Ltd. CFF. See Critical flicker frequency Chafing gear, 296 Chain mat, beam trawl, 299 Chains (or chain matrix), 296, 297 Channel catfish (Ictalurus punctatus), sounds produced by, 48 Characoids, hearing ability of, 50 Char (Salvelinus alpinus), arginine threshold for, 113 Chemical attractants, spatial range of, 111 Chemically stimulated feeding behavior, 112

351 Chemical stimuli properties, chemosensory senses and, 110–111 Chemosensory thresholds, 113–114 Chewing bait, 117 China rockfish (Sebastes nebulosus), discard mortality of, 266 Chinese electrobeam trawls, 222, 222–224 Chinese pulse generator, use of, 224 Chinese shrimp beam trawls, 298 Chinook salmon (Oncorhynchus tshawytscha), gillnetting and survival rate of, 195 Chislett, G., 146 Chorioid, 26 Chub mackerel (Scomber japonicas), electrotaxis induced in, voltage gradient combined with pulse rate and, 216t Chum salmon (Oncorhynchus keta) motion detection capability in, 30 set nets for, 162 Cichlid fish (Astronotus ocellatus), hearing range frequency of, 50 Circadian rhythm, retinomotor response and, 31 Circle hooks, 131, 335–336, 337 blue shark and, 129 catch efficiency with, 118, 123 converting from J-hooks to, 132 sea turtle bycatch reduction and, 127 Clark, J., 332 Climate, discard mortality and, 286 Cod; Atlantic cod (Gadus morhua) amino acids as feeding stimulants for, 112 amino acid sensitivity in, 113 bait localization by, 117 chemically mediated responses to baited longline by, 116 diel rhythm in swimming speed of, in May and September, northern Norway, 120 distribution of, in relation to wind-induced temperature and vulnerability to cod traps set northeastern Newfoundland, 167 entrance into trawl net by, 83 glycine behavioral response thresholds in, 114 groundgear collisions and escape mortality for, 281 learning/conditioning by cod able to locate and ingest four acoustic tags wrapped in mackerel bait, 122–123, 123

352 Cod; Atlantic cod (Continued) maximum swimming speed relative to body length of, 21 mortality of those encountering but escaping fishing gear, 268 movements of schools of, near modified Newfoundland cod trap, 166 responses of, towards baited hooks, 118, 118 separator trawls and behavior of, 256 sound source localization in, 70 sounds produced by, 48 Cod behavior near Newfoundland cod trap, 165–167 at distance from trap, 165 at entrances of trap, 165–166 inside trap, 166–167 near leader, 165 temperature and, 167 Coded sonic transmitters, 276 Codend construction, controlling size selection and alterations of, 248–249 Codend covers, escapees caught inside, 270, 271, 272, 273 Codend mesh blockage, escape opportunities and, 284 Codends. See also Trawl codends beam trawl, 299 bottom trawl fishing system, 68 in Chinese shrimp beam trawls, 298 demersal single boat trawl, 240 fish behavior inside of, 84–85 fish behavior inside trawl net and (Zone 3), 84–85 bycatch reduction devices and, 85–89 entry and orientation, 82–84 fish escape from, during trawling process, 243 otter trawls, 296, 297 selection profile of, showing parameters L , 245 50 separator trawls, 256 shrimp electrobeam trawl vessels, 223 square mesh, 247, 247–248 Code of Conduct for Responsible Fisheries (FAO), 321 Cod trap modification, reduction in salmon bycatch and, 173 Cod traps acoustic signals, mesh size and, 175 disappearance of, 160 Newfoundland, 162–164, 163

Index Coho salmon (Oncorhynchus kisutch) detection thresholds for, 113 gillnetting and mortality rate of, 195 Cole, R. G., 147, 149 Collin, S., 32 Color, fish behavior near trawls and, 89–90 Color-blind species, 27 Color discrimination, determining, 27 Color vision, 27 Comet fish, swimming specialization of, 9 Commercially exploited species, discarding of, 259 Commerson’s dolphins (Cephalorhynchus commersonii), fishing operations and threat to, 316 Commission for the Conservation on the Antarctic Marine Living Resources, 107 Committee on Fisheries (FAO), 324 Common carp (Cyprinus carpio), homogeneous fields of direct current and reactions of, 215 Common dolphin (Delphinus delphis) acoustic signals and, 329 fishing operations and threat to, 317 Common murre (Uria aalge), gillnets and bycatch of, 195 Common starfish (Asterias rubens), electric beam trawl and, 309 Complete bite, of bait, 117 Complex pulse pattern, 206 Condition indices, predicting mortality of unrestricted fish with, 275–276 Conditioning, food and acoustic coupling and, 54–55 Conductivity of water, electric current and, 212 Cone densities, 30, 33 Cone index, 31 Cones, 40 for color discrimination, 27 light intensity and, 28 optical resolving capability by, 33 retinal, 26 Conger eel (Conger sp.) backward and forward swimming by, 11 glutamine thresholds in, 113 Connaughton, M., 50 Conners, M. E., 147 Conservation. See also Bycatch; Discards; Seabed electrofishing and, 228–231 injuries, types of, 228–230 injuries and mortality, 230–231

fish pots and, 150–154 habitat alteration, 152–154 lost gear and ghost fishing, 151–152 megafauna interaction, 152 undersized and nontarget species, 150–151 mitigation measures aimed at reducing incidental catches of seabirds, 124–126, 132 of sea turtles, 126–128, 132 of sharks, 128–130, 132 mitigation measures in trap fisheries and issues related to, 174–178 of sea turtles, 324 of sharks, 324, 325 of whales, 323–324 Conservation-oriented fish-harvesting measures, visual physiology research and, 37 Conservation programs, for Agreement on the International Dolphin Conservation Program, 322–323 Contrast, fish behavior near trawls and, 89–90 Convention on Migratory Species, 321 Convention on the Conservation of Migratory Species of Wild Animals, 324 Cooke, S. J., 149 Cookies, 297, 306 Cooling water intakes, aversive sound and reducing fish entrapment in, 60 Cooper, C. G., 252 Cooper, R. A., 197 Coral reef ecosystems, artisanal fishing gear and, 153 Coral reef fisheries, fish pots used in, 146 Cornea, 26 Cornet fishes (Fistsularia sp.), tails of, 11–12 Cost to transport body mass and, 19 dimensionless, doubly logarithmic plot of, 16 energetic cost of swimming and, 15 Cotter, A. J. R., 257 Counterherding, improving escapee survival and, 285 Cowx, I. G., 206, 212 Cox, T. M., 196 CPUE. See Catch per unit effort Crabs discard mortality in trawl fisheries and, 267

Index groundfish gillnets and capture of, 194 Cranial nerves, 110 Crawling, 213 Crean, K., 242 Creels, 143 Crimson sea bream (Evynnis japonica), acoustic attraction and, 53 Cristae ampullaris, 49 Critical flicker frequency, 28 Critical value of field intensity, fish behavior and, 228 Croakers, frequency range of, 52 Cruise-swimming behavior, in codend, 84 Cruising specialists, 10 Crustaceans discard mortality in trawl fisheries and, 267 electric field and behavior of, 218–219 fish pots for, 143 gillnets and avoidance of bycatch for, 194 CT. See Cost to transport Cui, G., 38 Current electrofishing injury and mortality and type of, 230–231 fish interaction with longlines and, 121 Cutaway trawls, improving species selectivity with, 253, 254 Cyprinoids, hearing ability of, 50 D Dab, common (Limanda limanda), 257 audiogram of hearing sensitivity in, 70 electric field experiments in seawater and, 214 hearing range frequency of, 50–51 Daibo-ami, 161, 161 DAM. See Dynamic area management Danforth anchors, gillnets and, 184 Danish Institute for Fisheries Research, 326 Daniulyte, G., 213, 218 Davis, F. M., 243 Davis, M. W., 276, 279, 280 Davis, T. L. O., 275 Dayton, P. K., 152 DC. See Direct current DDDO2F acoustic device, 328 DeAlteris, J., 178 Deep-set longlines, sea turtle bycatch reduction and, 128

Deflector panels, cod trap modification to reduce salmon bycatch and, 173 De-ghosting technologies, 197 Degradable plastic plates, for lost grillnets, 197–198, 198 Demeral trawl fisheries, broad categorization of, generic fishery examples, desirable management objectives, and possible technical interventions, 244t Demersal beam trawl fleet, discard rate attributed to, 241 Demersal fish, bottom-dwelling, endurance and swimming speeds in, 19 Demersal fisheries Northeastern Atlantic Ocean, 106 Northeastern Pacific Ocean, 107 Northwestern Atlantic Ocean, 106–107 Northwestern Pacific Ocean, 107 Southern Ocean, 107 Demersal longline fishing, 106 worldwide, 106t Demersal longlines, setting, 109 Demersal single boat trawl, principal components of, 240 Demersal species, reaction distance and, 72 Demersal trawl, mouth of, behavior of cod, haddock, and whiting in, 255 Demersal trawl finfish fisheries, global total landing with, 240 Demersal trawl fisheries, reducing retention of cod in, 254 Density draining phenomenon, avoidance patterns and, 72 Dents, on scales, 8 Depressors, in groundgear, reducing seabed impact with, 305 Depth change, escape mortality and, 280 Derby, C. D., 112 Derelict gillnets, ghost fishing problems and solutions, 197–198 Dethloff, J., 226 DFAAs. See Dissolved free amino acids Diamond mesh codends, 249 DIDSON unit. See Dual frequency IDentification SONar unit Diel activity rhythms bait odor plume and, 115 in mean swimming speed of cod in May and September, 120 DIFRES. See Danish Institute for Fisheries Research

353 Dill, L. M., economic model of reaction distance for fish under predator’s threat, 71, 71, 95 Diner, N., 213, 226 “Dinner-bell” effect, pings and, 51 Diodontiforms, 213 Dipole sound source, 46 Direct current, 206 electrofishing and, 230 homogeneous fields of, reactions of fish in, 215–216t Direct current field, fish behavior in, 211 Direct current waveform, in electrofishing, 207 Discard cage system method, 285–286 Discard mortality, 266–268 biological and economic losses tied to, 266 differential, within and between species, 266–267 factors causing, 282–283 catch composition, 283 on-deck handling, 282 predation and scavenging, 283 thermoclines, 282 towing duration and catch size, 282–283 in gillnet fisheries, 267 in hook-and-line fisheries, 267–268 in pot and trap fisheries, 268 scavenger and predation mortality, 268 in seine fisheries, 267 in trawl fisheries, 267 Discard reduction measures, in trawl fisheries, implementation of, 257–258 Discards. See also Bycatch; Conservation technical measures for reducing reducing capture of specific species or sizes in mixedspecies fisheries, 253–257 reducing target species discards by controlling size selectivity, 243–250 Discards. See also Bycatch; Conservation; Mortality bycatch vs., 239 cause of, 242 commercial fisheries and, 265 gillnets and measures for reduction of, 192–195 improving survival measures for, 285–286 potential fate of animals entering a trawl and its relation to, 241

354 Discards (Continued) rates vs. levels of, 242 reducing, 38 societal demands for, 259 technical measures for reducing reducing unwanted bycatch in single-species fisheries, 250–253 trawling and percentage of, 239 in world fisheries, 240–242 Disc clusters, reducing impact of cable on sessile invertebrates with, 307–308, 308 Discs, otter trawls, 296 Dispersion model, locating bait odor source and, 115 Dissolved free amino acids, 114 Distinct population segment, 320 Diurnal vertical migrations, 189 Dogfish (Squalus acanthias), endurance values for swimming in mouth of trawl and, 80 Dolphinfish (Coryphaena hippurus), fish aggregation devices and, 58 “Dolphin-friendly” tuna, 315 Dolphins, 315, 316 acoustic devices and, 328–329 acoustic signals and, 329 conservation programs for, 322–323 deterrent sounds and, 331 ecological association of tuna and, 317 worldwide mortality of, in fisheries, 316 Domeier, M. L., 275 Donburi fishing method, 60 acoustic attraction and, 53, 54, 54 sound wave and sound spectrograph recorded before and after casting of, to sea, 55 Doors, bottom trawl fishing system, 68 Dorsal tip vortices, for steadily swimming saithe, 14, 14 Double-grid and codend system, for improving size selection of Nehrops and finfish, 257 Double line system, longlining, 109 Dover sole (Solea solea), flatfish beam trawls and, 309 Downward excluding Super Shooter TED, 86 DPS. See Distinct population segment Drag, overcoming, energy for swimming and, 15 Drag forces, fish pots and, 153 Dragonet (Callionymus sp.), homogeneous fields of direct current and reactions of, 215

Index Dredge frame modifications, 334 Drift gillnets, 183, 184t Drift net fisheries, shark and ray bycatch in, 318 Driftnets, banning of, in some areas, 195 Drop chains, bottom trawling with, 301–302 Drop-out, fishing mortality and, 265 Drop-out panels, reducing benthos catch with, 310 Drumming sounds, stridulation sounds vs., 48 Dual frequency IDentification SONar unit, 329, 330 Dugong (Dugong dugon), fisheries and interactions with, 318 Dumaguete Action Plan (Southeast Asia), 322 Dunlin, G., 254 Dunning, D. J., 60 Dutch electrobeam trawls, 225, 225 228 Dutch Ministry of Agriculture and Fisheries, 225 Duty cycle, 206 Dynamic area management, 337 E East Coast of U.S. fishery, excessive discarding in, 241 Eastern Tropical Pacific Ocean (ETP) purse-seine tuna fishery, dolphin bycatch in, 317 Eayrs, S., 253, 304, 305 Ecologically sustainable management for fisheries, future challenges related to, 132–133 Eel (Anguilla anguilla), homogeneous fields of direct current and reactions of, 215 Efficiency doors, reducing seabed impact with, 305 Egress behavior, fish pots and, 148–149 Eigaard, O. R., 252 Elasmobranch fins, 6–7 Electric control of fish, obtaining, in freshwater lakes and streams, 206, 209 Electric field, 209–219 factors related to fish behavior in, 210–214, 218–219 conductivity, 212 electric field, 210, 211–212 electrodes, 212 fish size, 212–213

metabolism and other physiological status, 212 species: fish, 213–214, 218 species: invertebrates, 218–219 fish reactions in, 211–212 general observations about, 209 heterogeneous, around and between electrodes, 208 homogeneous, distortion of around fish in water that is less conductive and more conductive, 209 strength of, 206 in water, properties of, 206, 209 Electricity in fisheries, historic discussions about use of, 205–106 Electric pulse stimulation, in beam trawling, flatfish and brown shrimp, 309, 310 Electric senses of fish application in marine fisheries, 205–232 attracting electrode systems, 219 automated unmanned fishing system, 227 historic overview and description of selected cases, 220–227 pump fishing for pelagic fish, 226 repelling electrode systems, 219 trawling for fish, 224–226 trawling for shrimps, 220–224 Electrobeam trawls Belgian, 220, 221 Chinese, 222, 222–224 Dutch, 225, 225 selective electrified otter trawls, 222, 222 Electrocution, in electrofishing operations, 229 Electrodes bronze, 228 characteristics of, 212 heterogeneous electric fields around and between, 208 platinum, 228 small, injuries and, 228 spherical, 227–228 Electrofishing, 205 complexity of, 206 conservation issues factors affecting injury and mortality, 230–231 types of injuries caused by, 228–230 different waveforms used in, 207 necropsy filets of rainbow trout revealing hemorrhages and vertebrae damage caused by, 230

Index radiographs of rainbow trout showing spinal misallignment and fractured vertebrae caused by, 229 Electro-fishing response zones, major intensity dependent, 210 Electronarcosis, 213, 231 requirements for, induced with PDC, 213 Electronic tagging, predicting mortality of unrestricted fish with, 275 Electrophysiological research methods, on auditory physiology of fish, 52 Electrophysiological studies, on chemosensory thresholds in fish, 113 Electroretinogram, 27 Electroretinogram amplitude, in dark-adapted eyes of walleye pollock, 28 Electroretinographic amplitude, relative, in light-adapted eyes and two dark-adapted eyes of different time, 29 Electroretinography, 29 Electrotaxis, 213 inducing, pulse rate and voltage gradient for inducing of in species studied, 216t mullet length and, 214 range, 231 for pump fishing, 226 requirements for, induced with PDC, 213 Elesmobranches flicker fusion frequency in, 29 lens movement in, 26 Ellis, I. E., 149 Emetics, 337 Encircling gillnets, 183 Endangered species discards and, 265 sea turtles, 318 Endangered Species Act of 1973, 322, 323, 324 Endurance capture and escape processes and, 277 codend construction, amount of accumulated catch and, 84 at prolonged speeds of various marine fish species, 20t swimming in mouth of trawl and, 79 swimming speeds and, 18–19 temperature and, 19 Endurance probability curves, for swimming endurance of Atlantic cod at speeds comparable to those

experienced in trawl mouth, 81, 82 Energetic cost of swimming, expression in watts, 15 Energy, needed for fish transport, doubly logarithmic plot of, 17 Energy consumption, by swimming fish, 14–18 Engås, A., 67, 72, 76, 86, 192, 257 Eno, N. C., 153 Entanglement incidences, whale bycatch and, 317–318 Entangling, 184 Entrance design, fish pots, fish behavior and, 148–149 Epperly, S. P., 126, 280 ERG. See Electroretinogram Erickson, D., 267, 268, 280, 282, 283 trawl escapees and improvements for collection method developed by, 270, 272 Escape behaviors fishing mortality and, 265 visual stimuli and manipulation of, 39–40 Escape distances, modeling of, 37 Escapees improving survival measures for, 281–285 increasing escape opportunities, 284 modification of fishing practices, 284 modification of groundgear, 284–285, 287 reducing time between capture and escape, 283–284 Escape/excluder panels, 334 Escapement, mesh size and encouragement of, 248 Escape mortality, of fish encountering but escaping fishing gear, 268–269 Escape panels, reducing discard mortality and, 286 Escape processes, potential factors affecting fish behavior, endurance, stress, and cumulative injury during, 277 Escape rings, on fish pots, 150 Essential Fish Habitat Environmental Impact Statement (2005), 300 Estuarine prawn (Nematopalaemon hastatus), 239 Europe fish pots in, 145 stranded dead dolphins in, 317

355 European dab (Limanda limanda) best hanging ratio for capture of, 185 gillnet twine size and capture of, 186 European eel (Anguilla anguilla) hearing range frequency of, 50 number of vertebrae in, 5 European sea bass (Dicentrarchus labrax), bycatch rates for, 317 European Union, regulatory frameworks and, 321–322 European whitefish (Coregonus lavaretus), Baltic sea traps for, 164 Evolutionary fitness, performance and, 5 Exclusion devices, 331–335 Exhaustion codend entrance and, 84 entry into trawl net and, 83 swimming in trawl mouth and, 82 Exit windows, fish escape rates and, 83 Exponential PDC waveform, in electrofishing, 207 Exposure time discarded fish mortality and, 282 Extended funnel bypass reduction device, 88 Extension (or intermediate section) funnel section of netting, 83 otter trawls, 297 Eye temperatures, 30 EZ-baiter hooks, 109 F Facial cranial nerve, 110 FADs. See Fish aggregation devices False killer whales (Pseudorca crassidens), bycatch of, 318 FAO. See Food and Agricultural Organization Farmer, M. J., 279 Fathead minnow (Pimephales promelas), noise exposure and, 52 Fatigue, 83 mortality and, 279 swimming in trawl mouth and, 82 Fats, swimming and use of, 14 Faulkner, G., 190, 195 Favorite, F., 67 Fay, R. R., 70 FBC. See Fjord & Belt Centre Federal Research Institute for Rural Areas, Forestry, and Fisheries, 328 Feeding attractants, 111–113, 131 Feeding rate, field trials in open seas and, 56

356 Fernö, A., 67, 95 Ferrell, D. J., 149 FFF. See Flicker fusion frequency FGDs. See Fish guidance devices Field intensity, electrofishing injury and mortality and, 231 Field trials, in open seas, 56–57 Filter grids, 333 Fine, M. L., 50 Finfish double-grid and codend system for improving size selection of, 257 separator trawls and behavior of, 256, 256 Finfish fisheries, Nordmøre grid and bycatch reduction in, 252 Fin rays, 6 Fins, 5 structure of, 6–7 Fish body shape of, 7 hearing abilities of, 49–51 interactions between water and: fish wakes, 13–14 light attraction and, 37 social learning in, 94 sounds produced by, 48 trawling for, 224–226 Fish aggregation devices, 57–59, 58, 60 Fish behavior in and around traps, 165–170 Baltic traps, 169–170 Japanese set nets, 167–169 Newfoundland cod traps, 165–167 capture and escape processes, factors affecting, 277 in electric field, 210, 211, 211–212 conductivity and, 212 different variables with influence, 211 electrodes and, 212 fish size, 212–213 metabolism and other physiological status, 212 species: fish, 213–214, 218 species: invertebrates, 218–219 electric field and, 230 gillnet fishing and, 189–192 in relation to pots, 146–150 approach to pots, 147 attraction, 146–147 entrance design and ingress/egress behavior, 148–149 inside of pots, 149–150 phases of, 146

Index species-specific research, separating mechanisms to reduce bycatch and, 252 trap designs and, 170–172 Fishbox, 253 Fishbox bypass reduction design, 88 Fish density, fish behavior near trawls and, 90–91 Fish entrapment, in cooling water intakes, aversive sound and reduction in, 60 Fisheries, ecologically sustainable management for, future challenges, 132–133 Fishery inspection vessels, 328 Fish escapement bycatch reduction devices and, 85, 86 overcoming optomotor reflex and, 88 Fish eye accommodation, 25 lens, 26 optics: collection and formation of image, 25 retina, 26–27 structure of, 25–27, 26 Fisheye bypass reduction device, 88, 251 Fish guidance devices, 55, 56 changes in numbers of fish following, with and without use of food pellet, 57 sound used in, 54–57 conditioning, 54–55 field trials in the open sea, 56–57 guiding device, 55 Fishing, light use in, 37 Fishing gear animals affected by, 241 escapement from, mortality and, 265–266 marine mammal mortality issues and, 316 marine mammals taken incidentally in, 318 mortality of fish that encounter but escape, 268–269 movement detection and fish reaction to, 30 sawfish interactions with, 320 turtle interactions with, 318 visual conditions and observations of fish in, 38 Fishing gear underwater, 35–36 color and appearance of spectral properties of seawater, 35 visual contrast of gear, 35–36 Fishing mortality, defined, 265

Fishing practices, modification of, improving escapee survival measures and, 284 Fishing range of gillnets, predicted, at different water temperatures and different soaking durations, 191 Fishing vessels, sounds produced by, 48 Fish length, mortality related to fishing gear escapements and, 269 Fish pots, 143–154 catch efficiency, 154 commercial, 145 conservation challenges and solutions, 150–154 habitat alteration, 152–154 lost gear and ghost fishing, 151–152 megafauna interaction, 152 undersized and nontarget species, 150–151 defined, 143 fish behavior in relation to, 146–150 in approaching pot, 147 attraction, 146–147 entrance design and ingress/egress behavior, 148–149 inside the pot, 149–150 general illustration of, 144 improving efficiency of, 154 for Pacific cod, 144, 145 Turkish, 145 two-chamber, 145, 146 worldwide use of, 143–145 Fish routes formation of, 171 trap operations and, 170–171 Fish schools, 171, 171 Fish size electric fields and, 212–213 escape mortality and, 277–278 fish behavior near trawls and, 91–92 inclined separator grids and, 87 response curves for voltage gradient by frequency, temperature and, 217 responses to bait and, 121–122 swimming speed, gillnet fishing and, 189 Fish skin of eels and sharks, 8 mechanically important part of, 7 Fish traps, defined, 143 Fish vertebrae, swimming and, 5 Fish vision capture of fish and application of, 36–40

Index herding and capture of fish by trawl under visual conditions, 36–37 light and illusion used in guiding and blocking, 37–40 light use in fishing, 37 fish capture and role of, 25–40 Fish wakes, 13–14 Fixed gillnets, 183, 184t Fjalling, A., 175 Fjord & Belt Centre (Denmark), 330 Flatback turtle (Natador depressus), 318 Flatfish beam trawls, electrical stimuli used in, 309 Flatfish (Pleuronectes platessa), 213, 220 beam trawls and harvesting of, 296 bodily shape of, 11 electric field experiments in seawater and, 214 electro-beam trawls and, 225 entrance height into trawl net by, 83 entrance into trawl net by, 83 extension and behavior of, 83 herding behavior of, in response to sweep of bottom trawl, 77 herding patterns and, 76 mortality of those encountering but escaping fishing gear, 268 separator trawls and behavior of, 256 swimming trajectory selection by, relative to advancing sweeps, 78 Flathead (platycephalus bassensis), maximum swimming speed relative to body length of, 21 Fleeing, 213 Fleet-set fish pots, habitat alteration and, 153 Flexible grids, 333 Flicker fusion, comparative studies on, 29 Flicker fusion frequency, 28, 29, 30 Floats demersal single boat trawl, 240 gillnet, 185 otter trawl, 296, 297 Flounder (Platichtys flexus) electric field and behavior of, 213 electric field experiments in seawater and, 214 ratio of mesh size of trap leader and, 168 Flow behind steadily swimming saithe, 14 fish mucus and, 8 Flow patterns, complexity of, 14 Flume tanks, 88

Flying fish, pectoral fins of, 11 Foil design, fish behavior variations and, 88 Food and Agricultural Organization, 183, 319 Code of Conduct for Responsible Fisheries, 321, 324 shark action plan, 318 Food conditioning, 54–55 Food deprivation, food searching behavior and, 119 Food security, bycatch issues and, 255 Footgear, 67 accidental collisions with, 82 of bottom trawl, collision and escapement of fish under, 81 in bottom trawl fishing system, 68 Footropes, gillnets, 183, 185 Foraging fish behavior, bait fishing and, 110 Forkbeards (Phycis phycis) braking by, 10 swimming specialization, 9 Fork-rigging, 300 Form vision, 30–31 Foster, D. G., 88 Fountain maneuver, of roundfish, 75, 75, 76 Frame ropes, gillnets, 183 Franciscana dolphin (Pontoporia blainvillei) acoustic pingers for gillnets and bycatch reduction of, 196 fishing operations and threat to, 316 Fraser’s dolphins (Lagenodelphis hosei), fishing operations and threat to, 316 French Research Institute for Exploration of the Sea, 328, 329, 330, 333, 334 French-rigging, semipelagic trawling with, 300–301 French XI Trawl “Cetasaver” acoustic deterrent device, 330 Fréon, P., 94 Frequency electric field strength and, 206 response curves for voltage gradient by length of fish, temperature and, 217 Friction drag, 7 Frid, C., 267 Frigate tuna (Ausix thazard), fish aggregation devices and, 58 Fright response, electric field and, 232 Frøysa, K. G., 258 Fryer, R. J., 257

357 Fujimori, Y., 172, 190 Furevik, D. M., 143, 146, 147, 148, 149, 150 Furnell, M. J., 240 Fuwa, S., 148 G Gabriel, O., 143, 164 Gadoids entrance height into trawl net by, 83 mortality of those encountering but escaping fishing gear, 268 Gaits, fish size and, 92 Galvanic time-release mechanisms, in fish pots, 152 Galvanonarcosis, 212 Galvanotaxis, 212 Ganglion cell density, visual acuity and, 32 Ganglion cell layer, of retina, 27 Gap detection, visual acuity and, 31 Gargoor, 145 Gas bladders, 49, 50 mortality of discarded fish with, 265, 266, 282 ruptured, haul-back, depth change and escape mortality tied to, 280 Gear longline, parts of, 108, 108 lost, fish pots and, 151–152 Gear design, gillnets, 184–187 Gearin, P. J., 328 Gear modifications, 310 mitigating seal-induced damage and, 334 reducing small cetacean bycatch with, 51 sea turtle bycatch reductions and, 319 Gear parameters, longline catching efficiency and, 131 Gear retrieval operations, 198 GEF. See Global Environment Facility “Gelder” codend, 249 Geometric similarity theory, 184, 187 Ghost fishing defined, 151 derelict gillnets and, 197–198 fishing mortality and, 265 fish pots and, 151–152 lost pots and, 154 Giant grenadier (Albatrossia pectoralis), longline catch rates of sablefish negatively correlated with, 122 Gibson, D., 332 Gilling, 184, 186

358 Gillnet fisheries acoustic deterrents used in, 316 acoustic devices used in, 326 discard mortality in, 267 marine mammal bycatch rates in, 318 sea turtle bycatch and, 318 Gillnet panels, fish behavior reactions to, 38 Gillnets anatomy of, with names of gear components, 185 avoidance of Atlantic mackerel in, 193 baited, 192 banning of, in some areas, 195 capture mechanisms, gear designs, and fishing efficiency with, 184–187 defined, 183 degradable plastic panel for attaching floats to headrope of, 197–198, 198 derelict: ghost fishing problems and solutions, 197–198 efficiency of, 187 fish behavior and fishing with, 189–192 fish behavior near, 183–198 fish capture and longlines vs., 187–188 fish capture by, modes of, 186 hanging ratio of, 185–186, 187 interaction of marine mammals, seabirds, and sea turtles with, 195–197 measures to reduce bycatch and discards in, 192–195 size selectivity of, 187–189, 198 structural components of, 183–184 types and key features of, 184t Gillnet size selectivity curves, 187, 188 Gillnetting, history of, 183 Gilman, E., 127 Glass, C. W., 38, 39, 67, 78, 94, 248, 281 Global Environment Facility, 251 Glossopharyngeal cranial nerve, 110 Glutamine, 113 Glycine, 112, 113, 114 Gobert, B., 149 Godø, O. R., 67, 72, 76, 91 Golden fish (Cyprinidae), homogeneous fields of direct current and reactions of, 215 Goldfish (Carassius auratus) audiogram of hearing sensitivity in, 70

Index extended hearing frequency range in, 50 Golenchenko, A. P., 93 Graham, N., 67, 249, 257, 258 Grand mal zone, electro-fishing, 210 Grant, G. C., 192 Great white shark, protected status for, 325 Greenland halibut (Reinhardtius hippoglossoides) baited gillnets and capture of, 192 endurance values for swimming in mouth of trawl and, 79 gillnet mesh size and capture of, 184 Northeastern Atlantic Ocean demersal fisheries and, 106 Northwestern Atlantic Ocean demersal fisheries and, 106 Green turtle (Chelonia mydas) exclusion devices and, 331 fishing gear interactions and, 318 pound net fishery in mid-Atlantic coast (U.S.) and, 177 Grey mullet (Mugil cephalus), attraction of, to thermal plume at nuclear power plant, 60 Grey seal (Halichoerus grypus), fisheries and interactions with, 318 Grey whales (Eschrichtius robustus), entanglement incidences with, 317 Grid designs, reducing interaction of seals with salmon and whitefish traps in Baltic Sea and, 176, 176 Grid orientation importance of, 87 shrimp fishery and, 87 Grimaldo, E., 87 Groundgear common types of, 297 lifting off seabed, 300, 301 modifications, 301–305 bottom trawling with drop chains and weights, 301–302 depressors and plates used as groundgear, 304 improving escapee survival measures and, 284–285 lighter groundgear for the offshore shrimp trawls, 302 nylon tickler brushes as groundgear, 304 raised footrope trawls and sweepless trawls, 302 rollers and wheels used in groundgear, 302–304 soft chain brush groundgear, 304–305 otter trawls, 296

Groundgear collision, escape mortality and, 281 Groundline, longline gear, 108, 108 Groupers, fish pot approach by, 147 Grunt family, stridulatory sounds produced by, 48 GTRs. See Galvanic time-release mechanisms Gudgeon (Gobio gobio), homogeneous fields of direct current and reactions of, 215 Guiding, light and illusion used in, 37–40 Gustation, fish detection of chemical stimuli through, 110 Gustatory stimuli, fish behavior relative to baited hooks and, 130 Gustatory system, final sensory evaluation of food items and, 111 H Habitat alteration, fish pots and, 152–154 Habitat degradation, fishing mortality and, 265 Haddock (Melanogrammus aeglefinus) entrance into trawl net by, 83 escape through square-mesh panel, 87 fish pot approach by, 147 gillnetting, bycatch reduction and, 195 L , mesh size and, 248 50 learned behavior modification in, 94 maximum sustained swimming speed and endurance at prolonged speeds in, 20t mortality of those encountering but escaping fishing gear, 268 Northeastern Atlantic Ocean demersal fisheries and, 106 Northwestern Atlantic Ocean demersal fisheries and, 106 responses of, towards baited hooks, 118 separator trawls and behavior of, 256 sound source localization in, 70 sounds produced by, 48 swimming speed and endurance for, 19 Hagfish, 283 Hake (Merluccius merluccius) bottom-set longlines and, 109 gillnets vs. longlines and capture of, 189

Index Half-sine waveform, in electrofishing, 207 Halibut. See Greenland halibut; Pacific halibut Halibut excluders, in fish pots, 150 Halieutika, 143 Halliday, R. G., 252 Halsband, E., 229 Hamley, J. M., 186 Hammerhead shark (Sphyrna mokarran), bycatch of, 319 Handegard, N. O., 74 Handlining, fishing by, 105 Hanging ratio gillnet design, fish capture and, 185–186, 187 gillnet efficiency and, 187 Hanging twine, gillnet, 185 Hannah, R. W., 252, 285 Hannah, W. R., 87 Harbor porpoise (Phocoena phocoena) fishing operations and threat to, 316 gillnets and bycatch of, 196 Harbor seal (Phoca vitulina), fisheries and interactions with, 318 Hargrave, B. T., 115 Harp seal (Phoca groenlandica), fisheries and interactions with, 318 Harris, G. G., 46 Harvest fish (Peprilus alepidotus), pound net fishery in mid-Atlantic coast (U.S.) and, 178 Hashimoto, T., 53 Haskell, D. C., 206 Hatakeyama, Y., 46 Hatchet fish (Gasteropelecus sp.) pectoral fins of, 11 swimming specialization of, 9 Haul-back, escape mortality and, 280 Hawksbill turtle (Eretmochelys imbricate) exclusion devices and, 331 pound net fishery in mid-Atlantic coast (U.S.) and, 177 turtle interactions with, 318 He, P., 83, 165, 174, 189, 302 Head, beam trawls, 299 Headline beam trawl, 299 bottom trawl fishing system, 68 otter trawl, 297 Headline sensor, 329 Headropes, gillnets, 183, 185 Heads, beam trawls, 296 Heales, D. S., 253

Hearing generalist specialists, noise exposure and, 52 Hearing generalist species, 46, 49, 60, 69 Hearing in fish, 49–51 noise exposure and effects on, 52–53 underwater radiated noise, pretrawl zone and, 69–70 Hearing specialist species, 46, 50, 60, 69 noise exposure and, 52 Heart rhythm, electrofishing and impact on, 230 Hector’s dolphins, protecting from gillnets, 337 Height of gillnet, bycatch reduction and, 194 Hemal spines, 5 Hendrickson, L., 251 Herding reactions, fishing gears and, 30 Herring (Clupea harengus) audiogram of hearing sensitivity in, 70 glycine behavioral response thresholds in, 114 swimming speed, gillnet fishing and, 189 Herrington, W., 241, 249 Herring traps, in Baltic Sea, 164 Herrmann, B., 248 Heterogeneous electric fields, 206 Hickey, W. M., 79, 281 High, W. L., 146, 147, 149 High aspect ratio doors, 310 low aspect ratio door vs., 306 reducing seabed impact and use of, 305 High-aspect-ratio tails, 10 Highflyer, gillnet, 185 High grading, discard rates and, 242 Higman, J. B., 219 Hill, B. J., 267 Hirayama, M., 148 Hiscock, K., 152 Holst, R., 186, 252 Holt, E. W. L., 247, 248 Homogeneous electric fields, 206 Hook, longline gear, 108, 108 Hook and line, fishing by, 105 Hook-and-line fisheries, discard mortality in, 267 Hooking, frequency of, 119 Hooking behavior, 114, 117–119 feeding motivation and, 119–120 Hooking position, mortality rates and, 127

359 Hooking probability, species-specific, response intensities and, 119 Hooks behavioral responses to, 117 materials used for, 105 types of, in longline fisheries, 109 Hook shape modifications, reducing turtle bycatch and, 316 Hook types ecologically sustainable fisheries management and, 132 reducing incidental shark catches and, 128–129 Hoop nets, 164 Horizontal avoidance pattern, 72, 74 Horizontal hanging ratio, 185 Horizontal migrations, 189 Horizontal patterning in netting, visual stimulus in fish and, 38 Horizontal separator trawl, 256 Horizontal separator with guiding ropes, to “encourage” cod into upper compartment, 256, 256 Hovering, neutrally buoyant fish species and, 11 Hovgård, H., 184, 186 Howell, P. T., 286 Howse, K., 174 Hughes, S. E., 149 Humpback whale (Megaptera novaeangliae) collisions with traps, 175 entanglement incidences with, 317 entrapment of, in Newfoundland groundfish gillnets, 196 fish pots and interactions with, 152 Hunger food searching behavior and, 119 prey abundance and levels of, 120 swimming, gillnet fishing and, 189 Huse, I., 67, 95, 190 Hybrid pulsed direct current waveform, in electrofishing, 207 Hydrophone, 48 I IATTC. See Inter-American Tropical Tuna Commission Ictalurus, gustatory receptors in, 113 IDCP. See International Dolphin Conservation Program IFREMER. See French Research Institute for Exploration of the Sea Illegal landings, fishing mortality and, 265 Illumination levels, flicker detection and, 28

360 Illusions fish behavior near trawls and, 90 guiding and blocking and use of, 37–40 Image formation, optics and, 25 Incidental catches of seabirds, mitigation measures aimed at reducing incidental catches of, 124–126, 132 of sea turtles, mitigation measures aimed at reducing incidental catches of, 126–128, 132 of sharks, mitigation measures aimed at reducing incidental catches of, 128–130, 132 Inclined separator panel, 250, 252 Indian Ocean, fish pots used in, 145 Indian Ocean–South-East Asian Marine Turtle Memorandum of Understanding, 324 Indo-Pacific bottlenose dolphins (Tursiops truncates), fishing operations and threat to, 316 Indo-Pacific humpback dolphins (Sousa chinensis), fishing operations and threat to, 316 Industrial Power Amplifier, 60 Industrial Revolution, 296 Ingress behavior, fish pots and, 148–149 Inhibited or undirected motion zone, electrofishing, 210 Injuries capture and escape processes and, 277 discarding and types and extent of, 266 mechanical sorting and, 284 on-deck handling and, 283 types of, electrofishing and, 228–230 Inner ear general morphology and functions of ancillary structures and, 48–53 effects of noise exposure on hearing abilities of fish, 52–53 hearing abilities of fish, 49–51 research techniques on auditory physiology of fish, 51–52 role of, 45 Inshore Potting Agreement (IPA) area (UK), 153 Institute for Baltic Sea Fisheries, 328 Inter-American Tropical Tuna Commission, 323 Internal bleeding, electrofishing and, 229

Index International Conference on the Conservation and Management of the Living Resources of the High Seas, 323 International Dolphin Conservation Program, 323 International Plan of Action for the Conservation and Management of Sharks, 124 International plans of action, 324 International Union for Conservation of Nature, 318 International Whaling Commission, 318–319, 323 Intrinsic fin muscles, 6 IPA. See Industrial Power Amplifier IPOAs. See International plans of action Iris, 26 Iron oxide-enhanced nets, 51 Irrawaddy dolphin (Orcaella brevirostris), fishing operations and threat to, 316 Isaksen, B., 91, 249, 332 IUCN. See International Union for Conservation of Nature IWC. See International Whaling Commission IX Trawl Company, 329 J Jack mackerel (Trachurus japonicus; Trachurus symmetricus) endurance values for swimming in mouth of trawl and, 79 maximum sustained swimming speed and endurance at prolonged speeds in, 20t maximum swimming speed relative to body length of, 21 motion detection capability in, 30 visual acuity expressed by minimum separable angle relative to body length in, 34 Jacopever (Sebastes schlegeli), auditory threshold curves for, 50, 51 Japan sea ranching operations in, 54 trap fisheries in, 159 Japanese common squid (Tadorodes pacificus), contrast threshold in, 36 Japanese eel (Anguilla japonica), motion detection capability in, 30 Japanese fisherman, fish attraction device used by, 53–54, 54 Japanese jack mackerel (Trachurus japonicus), auditory threshold curves for, 50, 51

Japanese mackerel (Scomber japonicus), maximum sustained swimming speed and endurance at prolonged speeds in, 20t Japanese sashimi market, tuna prices, 107 Japanese scallops (Patinopecten jessoensis), electric field and locomotive activity of, 219 Japanese set nets, 159, 161–162 evolution of, 161 fish behavior near around nets, 167 capture efficiency and, 169 leader, 167–168 playground, 168–169 proportion of fish schools guided by leader in, 170 proportion of fish schools guided by leaders or penetration through leaders of, 168 Japanese style cod trap, 163, 163 Japanese trap, traditional Newfoundland trap vs., 164 Japanese tuna longlining, speciesspecific effects of bait type and, 112 Jarbua tarpon (Terapon jarbua), attraction of, to thermal plume at nuclear power plant, 60 Jerk, 117 Jerk series, 117 Jerlov, N. G., 35 J-hooks, 118, 123, 131 blue shark catch rates and, 129 conversion from, to circle hooks, 132 discard mortality and, 267 Jigging, fishing by, 105 Johnson, A., 152 Jones, B. G., 91 Jones, S.A., 252 Jones/Davis bypass reduction device, 88 Jørgensen, T., 250, 281 J-shaped hooks, 109, 336 differences in shape and design of, 335 Juvenile and Trash Excluder (JTED), 255 Juvenile fish, discard reduction and, 258 K Kaimmer, S. M., 117 Kaiser, M. J., 153 Kallayil, J. K., 192 Kasumyan, A. O., 94

Index Kauppinen, T., 175 Kawakawa (Euthunnus affinis) fish aggregation devices and, 58 maximum swimming speed of, 21 Kazlauskiene, N., 230 Kelleher, K., 240, 241, 242 Kemp’s ridley turtle (Lepidochelys kempii) exclusion devices and, 331 pound net fishery in mid-Atlantic coast (U.S.) and, 177 turtle interactions with, 318 Kessler, D. W., 218, 219 Kick-and-glide swimming behavior, 16 Killer whale watching, noise exposure and, 48 Kim, Y. H., 35, 37 King, S. E., 285 King crab (Paralithodes camtschaticus) baiting pots and, 114–115 Norsel net design and avoiding bycatch of, 194, 194–195 pot and trap fishing and discard mortality of, 268 King prawn (Melicertus latisulcatus), soft brush footgear and, 304 King salmon, myotomes and myosepts on left side of, 6 Kiselev, O. N., 93 Klima, E. F., 213, 219 Knots, monofilament gillnets, spectral quality of water and, 38 Kolmarden wild animal park (Sweden), 330, 331 Kraus, S. D., 317 Kreutzer, C. O., 224, 226 Kruse, G. H., 150 Kvalsvik, K., 252 Kvamme, C., 91, 249, 258 L Labrids (Labridae), pectoral fin movements of, 12 Laevastu, T., 67 Lagena, 49, 49 La Jolla Commission, 323 Lakes, electric fishing in, 231 Lake sturgeon (Acipenser fulvescens) acoustic telemetry tagging of, 275 discards of, 265 Lake whitefish (Coregonus lavaretus), net avoidance by, 172 Lamarque, P., 213 Lamnoid sharks, 318 Lampreys lens movement in, 26 maximum amplitude in, 12

Lancaster, J., 267 Landings fishing mortality and, 265 potential fate of animals entering a trawl and its relation to, 241 quota-induced discarding and, 242 Landolt C Mark, 33, 33 Large-mesh fish pots, fish behavior on approach to, 147 Large-mesh top panel design, cod trap modification to reduce salmon bycatch and, 173 Large scale mullet (Liza macrolepis), attraction of, to thermal plume at nuclear power plant, 60 Larkins, H. A., 186 Lassen, H., 186 Lateral line, role of, 45 Lateral muscles, 5–6, 12 Lazy deck, beam trawl, 299 Leader mesh sizes, Baltic traps, 169 Leaders Japanese set nets, fish behavior near, 167–168 pound net fisheries rule, midAtlantic coast (U.S.), 177–178 in relation to bottom contour, headland, and typical fish swimming route, 171 Leadline, gillnets, 184 Learning in fish behavior near trawls and, 93–95 gear-related stimuli and, 122–123 Leatherback turtle (Dermochelys coriace) exclusion devices and, 331 fish pot buoy lines and entanglement of, 152 pelagic longline fisheries and bycatch of, 319 pelagic longlines and anthropogenic mortality for, 124 pound net fishery in mid-Atlantic coast (U.S.) and, 177 turtle interactions with, 318 LEDs. See Light emitting diodes Lehtonen, E., 176 trawl escapees and improvements to collection method developed by, 270, 272 Le Men, R., 213, 226 Lens, 26, 26 Lens muscle, 26 L , fish retention and, 244, 245 50 Li, Y., 148 Lien, J., 175, 196 Lift nets, fishing with lights and, 37

361 Light fish interaction with longlines and, 120–121 guiding and blocking and use of, 37–40 Light conditions, escape mortality and, 281 Light distribution, in seawater, 35 Light emitting diodes, 37 Light fishing, 37 Light intensity, 28 low, fish behavior in mouth of bottom trawl and, 80 trawl mouth fish behavior and, 78–79, 79 Light level, fish behavior near trawls and, 89–90 Light vision, 28 Limited endurance, measuring, 18 Lindley, S. T., 275 Lindsay, C. C., 213 Line-cutting device, 336 Lingcod (Ophiodon elongatus) thermoclines and discard mortality for, 282 water temperature and escape mortality for, 280 Ling (Molva molva) baited gillnets and capture of, 192 bait odor source location by, 115 fish pots for, 145 Northeastern Atlantic Ocean demersal fisheries and, 106 sound source localization in, 70 Littoral sharks, 318 Live release programs, for discards, 286 Lobster pots, 159 Lockwood, S. J., 267 Locomotion cruising specialists, 10 energy required for swimming, 14–18 swimming apparatus, 5–8 swimming-related adaptations, 8, 10–12 swimming speeds and endurance, 18–19 swimming styles, 12–13 Loggerhead turtle (Caretta caretta) exclusion devices and, 331 pelagic longline fisheries and bycatch of, 319 pelagic longlines and anthropogenic mortality for, 124 pound net fishery in mid-Atlantic coast (U.S.) and, 177 scallop fisheries and bycatch of, 319 turtle interactions with, 318

362 Løkkeborg, S., 106, 149 “Loner behavior,” fish densities and, 91 Longline efficiency, fish movements and, 115 Longline fisheries marine mammal bycatch rates and, 318 sea turtle bycatch and, 318 shark and ray bycatch and, 318 worldwide, regions and main target species for, 106t Longline gear capture process, 114–119 arousal to presence of bait and stimulus categorization, 115 behavior before stimulation, 115 locating bait odor source, 115–117 visual object categorization, bait ingestion, and hooking behavior, 117–119 chemoreception and food search, 110–114 chemosensory senses and chemical stimuli properties, 110–111 chemosensory thresholds, 113–114 feeding attractants, 111–113 mechanisms to locate an olfactory source, 111 conservation challenges/potential solutions, 123–130 main problems, 123–124 mitigation measures aimed at reducing incidental catches of seabirds, 124–126 mitigation measures aimed at reducing incidental catches of sea turtles, 126–128 mitigation measures aimed at reducing incidental catches of sharks, 128–130 defined, 105 fish behavior in relation to, 105–133 gear description, 108–110 interactions between fish and, 114–123 capture process, 114–119 external environment, 120–121 internal factors, 119–120 intraspecific and iinterspecific, 121–122 learning, 122–123 methods for setting of, 109 parts of, 108, 108 world wide longline fisheries, 106–108

Index northeastern Atlantic Ocean demersal fisheries, 106 northeastern Pacific Ocean demersal fisheries, 107 northwestern Atlantic Ocean demersal fisheries, 106–107 northwestern Pacific Ocean demersal fisheries, 107 pelagic longline fisheries, 107–108 Southern Ocean demersal fisheries, 107 Longlines, fish capture and gillnets vs., 187–188 Longlining historic use of, 105 as size-selective fishing method, 130 versatility of, 105, 108 Longspine porgy (Stenotomus caprinus), electrotaxis induced in, voltage gradient combined with pulse rate and, 216t Lost gear, fish pots and, 151–152 Louvar (Luvarus imperialis) body size of, 10 swimming specialization of, 9 Low aspect ratio door, high aspect ratio vs., 306 Lowry, N., 188, 278 Luckhurst, B., 146, 147, 149 Lunneryd, S. -G., 169, 176 M Mackerel bait, swordfish catch rates and, 128, 132 Mackerel (Scomber scombrus) mortality of those encountering but escaping fishing gear and, 269 swimming styles of, 12 Mackerel traps, 164 Mackerel-type fish, wavelength and swimming style of, 12 MacLennan, D. N., 258 Madsen, N., 248 Magnetic fields, sharks and detection of, 129 Main, J., 255 Main line longline gear, 108, 108 materials used for, 109 Mako shark (Isurus oxyrinchus), bycatch of, 319 Maksimov, Y., 213, 219, 224 Maniwa, Y., 53 Manta ray (Manta birostris), 315, 316 documenting extent of bycatch for, 319 fishing operations and bycatch of, 319

major conservation problems related to, 318 MAR. See Minimum angle of resolution Marine commercial fisheries, fish behavioral control systems and, 38 Marine mammal bycatch reduction, research related to, 315–316 Marine Mammal Protection Act, 322 Marine mammals gillnets and, 195–197 Newfoundland traps and, 174–175 Marine megafauna extent of bycatch, 318–320 nature of problem, 320–321 potential mitigation measures, 325–326, 328–337 acoustic alarms, 325–326, 328–331 circle hooks, 335–336 exclusion devices, 331–335 other bycatch solutions, 336–337 whale entanglement, 336 reducing interactions of, with fishing operations, 315–338 regulatory frameworks, 321–325 AIDCP program, 322–323 European Union, 321–322 protection of other marine megafauna, 324–325 Right Whale Protection, 323–324 sea turtles protection, 324 Southeast Asia, 322 United States, 322 UN Resolution 44/225, 323 species and fisheries involvement with, 316–318 species encompassed by, 315 Marine turtles, grids and bycatch reduction with, 252 Mark-recapture method, 275 Masu salmon (Oncorhynchus masou), auditory threshold curves for, 50, 51 Maximum amplitude, defined, 12 Maximum burst speed, body length of fish and, 19 Maximum sighting distance escape distances and, 37 estimating from minimum separable angle, 33 of walleye pollock (Theragra chalcogramma), 34 Maximum sustained swimming speed, at prolonged speeds of various marine fish species, 20t McMillan, J., 127 Mechanical sorting, injuries from, 284

Index Median fins, fish locomotion and, 12 Mediterranean monk seal (Monachus monachus), 318 Megafauna interaction, fish pots and, 152 Melvin, E. F., 195 Menhaden purse seiners, electrified pumping and, 226 Mesh configurations, visual stimulus in fish and, 38–39 Mesh size and shape acoustic signals, cod traps and, 175 discards reduction and, 243 escape mortality and, 278 fish capture in traps and, 172 fish pots and, 149 gillnet efficiency and, 187 gillnets and, 184 mortality issues and, 266 Metabolic exhaustion, 83 Metabolism, electric fields and, 212 Micro-Siemens, conductivity of water expressed in, 212 Midwater floats, sea turtle bycatch reduction and, 128 Mid-water set nets, 162 Migration patterns, successful trap operation and understanding of, 164, 169, 178 Migrations, horizontal and vertical, 189 Milkfish (Chanos chanos), attraction of, to thermal plume at nuclear power plant, 60 Minimum angle of resolution gap detection, target recognition and, 31 visual acuity expressed in, 31t Minimum legal size, 242 effect of modifying L relative to, 50 245, 246 Minimum mesh size, 245 Minimum response, electric field and, 213 Minimum separable angle gap detection, target recognition and, 31 Lindolt C Mark, cone density and, 32 maximum sighting distance estimated from, 33 visual acuity expressed in, 31t Minke whale (Balaenoptera spp.), bycatch of, 317 Misreported landings, fishing mortality and, 265 Mixed-species fisheries, reducing capture of specific species or sizes in, 253–257

MLS. See Minimum legal size MMPA. See Marine Mammal Protection Act MMS. See Minimum mesh size Modified Newfoundland cod trap, 163, 163 movements of schools of cod near, 166 Monkfish, herding patterns and, 76 Monocular vision, in typical teleost fish, 30 Monofilament gillnets, 186 spectral quality of water and, 38 Monofilament lines, 109 Monopole sound source, 46 Mooney, T. A., 196 Moon phases, ingress behavior in fish pots and, 148 Moray eel (Muraena helena) backward and forward swimming by, 11 swimming specialization, 9 Morizur, Y., 337 Moroccan industrial trawl fisheries, discard rate attributed to, 241 Mortality assessment of, 269–270, 273–276 fish escaping trawl codend, 270, 273–274 other methods to predict mortality of unrestricted fish, 275–276 defined, 265 of discarded fish, factors causing, 282–283 catch composition, 283 on-deck handling, 282 predation and scavenging, 283 thermoclines, 282 towing duration and catch size, 282–283 electrofishing and factors affecting, 229, 230–231 duration of exposure, 231 field intensity, 231 pulse duration, duty cycle, voltage, spikes, and fish size, 231 pulse frequency, 231 type of current, 230–231 waveform, pulse shape, 231 escape mortality, factors causing, 276–282 catch volume and quality, 279 fish size, skin damage, and mortality, 277–278 groundgear collision, 281 haul-back and depth change, 280 light condition, 281

363 mesh size and shape, 278–279 netting material, 279 post-escape predation, 281 repeated capture and escape, 281–282 towing speed and duration, 279–280 vessel movement, 281 water temperature, 280 factors causing stress, injury, and, 276–283 measuring, variability and flaws related to, 269 reducing, 265–266 total fishing, 241 unrestricted fish, predicting acoustic telemetry, 275 condition indices, 275–276 electronic tagging, 275 reflex impairment, 276 tag and recapture methodology, 275 Mortality of discards and escapees, 266–269 discard mortality, 266–268 in gillnet fisheries, 267 in hook-and-line fisheries, 267–268 in pot and trap fisheries, 268 scavenger and predation mortality, 268 in seine fisheries, 267 in trawl fisheries, 267 mortality of fish encountering but escaping fishing gear, 268–269 Mortality rates, hooking position and, 127 Motion vision, 28–30 Motivational state fish behavior near trawls and, 92–93 food searching behavior and, 119 MSA. See Minimum separable angle Mucus, on fish scales, 8 Mullet (Mugil auratus) continuously swimming, wake of, 15 electric behavior experiments with, 213 electric field and behavior of, 213 length of, and effect on potential difference between nose and tail required to induce electrotaxis, 214 Multifilament gillnets, 186 Multifilament lines, 109 Multifilament nets, 51 Munro, J. L., 146, 148, 149 Murawski, S. A., 193 Muscle fatigue, mortality and, 279

364 Myosepts, on left side of king salmon, 6 Myotomes, on left side of king salmon, 6 N Narcosis zone, electrofishing, 210 Narcotized fish, cardiac arrest in, 229 Nasal cavity, 110 National Marine Fisheries Service, 178, 319, 320 Natural selection, 5 NECESSITY study, 333 Nedreaas, K., 184, 188 Nehrops double-grid and codend system for improving size selection of, 257 separator trawls and behavior of, 256, 256 Nephrops fisheries discard levels lowered in, 249 Nordmøre grid used in, 259 Nephrops norvegicus, electric pulse field and, 218 Nested cylinder bycatch reduction device, 253 Net alterations, types of, 51 Net color, gillnet, avoidance of Atlantic mackerel and, 193 Net density, increasing, 51 Net dimension, gillnets, fish capture and, 186–187 Net height, gillnets, 187 Net mouth fish behavior between trawl doors and (Zone 2), 75–82 herding patterns: benthic species, 76–77 herding patterns: roundfish, 75–76 Net panel excluder devices, 334 Net path, defined, 75 Nets with acoustic reflectivity, 326 beam trawl, 299 gillnets, 184 Netting color, visual senses of fish and, 35–36 Netting material, escape mortality and, 279 Neural spines, 5 Newfoundland cod traps, 159, 162–164 cod at a distance from, 165 cod at entrances of, 165–166 cod inside, 166–167 cod near leader, 165 temperature and, 167 three popular styles of, 163

Index Newfoundland traps, interaction of marine mammals and sea birds with, 174–175 New Zealand, fish pots in, 145 New Zealand dory (Cyttus novazelandiae), entrance into trawl net by, 83 New Zealand fur seal (Arctocephalus forsteri), fisheries and interactions with, 318 NGOs. See nongovernmental organizations Night setting, 131, 132 as mitigation measure in pelagic longlining, seabirds and, 126 NMFS. See National Marine Fisheries Service NOAA Fisheries, Recovery Plan for the North Atlantic Right Whales established by, 323 Noise exposure, effects of, on hearing ability of fish, 52–53, 60 Noise signature, in pretrawl zone, 69 Nonbatoid sharks, 318 Nongovernmental organizations, 315 Nonotophysan fish, hearing thresholds in, 46 Nonreturn devices, fish pots, 149 Nontarget species, fish pots and: catch avoidance escapees, and discards, 150–151 Nordmøre grid, 255, 320 bycatch reduction in finfish fisheries and, 252 widespread use of, 258–259 Nordmøre shrimp grid, 250 structure of, 251 Norsel net design, to avoid bycatch of king crabs, 194, 194–195 Northern fulmar (Fulmarus glacialis), mitigation measures to reduce incidental catches of, 125 Northern fur seal (Callorhinus ursinus), fisheries and interactions with, 318 Northern right whale (Eubalaena glacialis), entanglement incidences with, 317 Northern rock sole, reflex impairment in, 276 Northern sea lion (Eumetopias jubatus), fisheries and interactions with, 318 North Pacific Fisheries Management Council, 300 Northridge, S. P., 316, 318, 320 Northrop, R. B., 229

North Sea haddock, temporal variation in codend size selection of, 93 North Sea plaice (Pleuronectes platessa), audiogram of hearing sensitivity in, 70 North Sea pollock (Pollachius pollachius), audiogram of hearing sensitivity in, 70 Norway pout (Trisopterus esmarki), reductions in bycatch of, 252 Norwegian Barents demersal fishery, reducing capture of fish below MLS in, 249 Nøstvik, F., 151 Novotny, D. W., 228 Noxious bait types, 337 NPFMC. See North Pacific Fisheries Management Council Nuclear layer, of retina, 27 Nuclear power plants, aversive sound and reducing fish entrapment in cooling water intakes from, 60 Nunnalle, E. P., 72 Nylon leaders, shark catch rates and, 129 Nylon tickler brushes, as groundgear, 304 O Oceanic carcharhinoid sharks, 318 Oceanic whitetip shark (Carcharhinus longimanus), bycatch of, 319 Ocean shrimp (Pandalus jordani), grid technology and, 252 Odor source, chemically stimulated rheotaxis and, 111 OECD. See Organization for Economic Cooperation and Development Off-bottom bridles, reducing seabed impact and use of, 307–308 Olfaction, fish detection of chemical stimuli through, 110 Olfactory bulb, 110 Olfactory cranial nerve, 110 Olfactory organ, in fish, 110 Olfactory source, locating, mechanisms for, 111 Olfactory stimuli fish behavior to baited hooks and, 130 Olfactory system, initiating feeding process and, 111 Olive ridley turtle (Lepidochelys olivacea), 126, 127, 318 Olsen, K., 92 Ona, E., 72 On-deck handling, discard mortality and, 282

Index Opah (Lampris guttatus) body size of, 10 swimming specialization of, 9 Optical cameras, 133 Optical resolution, 25 Optic nerve, 26 Optics, 25 Optimal foraging theory, 110 Optomotor reactions, fishing gears and, 30 Optomotor response (or reaction), 29, 31–32, 83 conceptual trawl design, improved catch efficiency and, 39 fishing gear evolution and, 38 overcoming, to facilitate fish escapement, 88 Organization for Economic Cooperation and Development, Fisheries Commission of, 323 Oshiki-ami, 161, 161 Ostariophysi superorder, hearing ability of fish in, 50 Ostraciiforms, 213 Otic gas bladder, 49, 50 Otolith, 49 Otolithic end organs, 49 Otophysan fish, hearing thresholds in, 46 Otoshi-ami, 161, 161, 162, 162 Otter boards, 296 demersal single boat trawl, 240 herding by, 37 Otter trawls, 310 designs, 296 development of, 296 electric fields use in, main reason for, 224 seabed impact from, 295, 296 Otter trawl system, components of, 296, 297 Ottmar, M. L., 122, 276, 279 Output controls, discard rates and, 242 Oxygen level, swimming, gillnet fishing and, 189 Oxygen use endurance in fish swimming and, 19 by swimming fish, 14 Oyster toadfish (Opsanus tau), sounds produced by, 48 Özbilgin, H., 87, 93, 94, 281 P Pace, D.S., 320 Pacific barracuda (Sphyraena argentea), 8 Pacific bonito (Sarda chiliensis), 8

Pacific cod (Gadus macrocephalus) fish pots for, 144, 145 herding pattern for, 76 Northeastern Pacific Ocean demersal fisheries and, 107 Pacific halibut (Hippoglossus stenolepis) catch composition and discard mortality for, 283 condition indices of discards for, 275 endurance values for swimming in mouth of trawl and, 80 fish pots and behavior of, 148 food deprivation in, and bait location by, 119 hooking behavior and, 117 light levels and bait locating by, 121 Northeastern Pacific Ocean demersal fisheries and, 107 postdiscard behavior in, 283 reflex impairment in, 276 thermoclines and discard mortality for, 282 water temperature and escape mortality for, 280 Pacific herring (Clupea palasii), mortality of those encountering but escaping fishing gear and, 269 Pacific mackerel (Scomber japonicus), schooling, tail beat frequency of, 18 Pacific Ocean demersal fisheries Northeastern, 107 Northwestern, 107 Pacific saury (Cololabis saira) retinal structure of, 27 visual acuity expressed by minimum separable angle relative to body length in, 34 Paddlefish (Polyodon spathula), gillnets, mesh size and, 184 Paired fins, fish locomotion and, 12 Panama Declaration, 323 Pantropical spotted dolphin (Stenella attenuata), fishing operations and threat to, 317 Parapenaeopsis hardwickii, electrobeam trawl vessels and, 223 Parrish, B. B., 36 Parsons, G. R., 88 Passive acoustic devices, 325–326 Passive porpoise deterrent device, 326 Patagonian scallop (Zygochlamys patagonica), trawl-caught, survival of, 267 Patagonian toothfish (Dissostichus eleginoides), Northwestern Pacific

365 Ocean demersal fisheries and, 107 Paua (Hailiorritis iris), fish pots and, 147 Pavlov, D. S., 32 Payao power spectrum of sound generated by, 59 sound sources and, 57, 58, 58–59 PBR rate. See Potential biological removal rate Pearly finned cardinal fish (Apogon poecilopterus), catch volume/ quality and escape mortality for, 279 Pectoral fins, 11 Pectorals, braking force and, 10 Pedersen, T., 151 Pelagic fish, pump fishing for, 226–227 Pelagic longline fisheries, 107–108 marine mammal bycatch rates and, 318 sea turtle bycatch and, 318, 319 shark and ray bycatch and, 318 Pelagic longlines, setting, 109 Pelagic longlining, 106 monofilament lines and, 109 worldwide, 106t Pelagics, small, entrance height into trawl net by, 83 Pelagic species fishing with lights and, 37 flicker fusion frequency in, 29 Pelagic swimmers, common form for, 7 Pelagic trawl, drop chain, interaction of, with seabed, 298 Pelagic trawl, interaction of, with seabed, 298 Pelagic trawl fisheries, dolphin mortality and, 317 Pelagic Trawl Interactive Pinger, 330 Pelagic trawls, 296, 299–301, 310 bottom trawls vs., 299 Penaeus duorarum, electric pulse field and, 218 Penalties, discard reduction measures and, 258 Pepperell, J. G., 275 Perez-Mellado, J., 240 Perforated-scale sardine (Sardinella albella), catch volume/quality and escape mortality for, 279 Performance, evolutionary fitness and, 5 Perrin, W. F., 317 Petit mal zone, electrofishing, 210 Petrauskiene, L., 213 Pettigrew, J. D., 32

366 Photopic vision, 27, 36, 40 Photosensitivity, 28 Physiological condition, fish behavior near trawls and, 93 Piasente, M., 83 PICE-97 devices, 326 Pigfish (Orthopristis chrysopteras), stimulatory capacity of extracts and, 112 Pigment index, 31 Pike (Esox lucius) acceleration behavior in, 10 swimming specialization of, 9 Pikeperch (Stizostedion lucioperca), gillnet twine size and capture of, 186 Pikitch, E., 280, 282 trawl escapees and improvements for collection method developed by, 270 Pilchards (Sardinops neopilchrdus), fish pots and, 147 Pinfish (Lagodon rhomboides), stimulatory capacity of extracts and, 112 Pingers, 325, 337 active, 326 costs related to, 328 drawbacks with, 51 noise pollution concerns and, 330 Pink grooved shrimp (Penaeus duorarum), 219 Pink muscle fibers, 6 Pink shrimp (Pandalus borealis) grids and bycatch reduction with, 249 lighter groundgear trials and, 302 semipelagic trawling for, 300 Pinnipeds, 315, 316 fishing gear incidences with, 318 Pipefishes, camouflage protection for, 11 Pistol shrimp (Alpheoidea spp.), snapping sounds of, 46 Pitcher, T. J., 189 Plaice (Pleuronectes platessa) electric field experiments in seawater and, 214 flatfish beam trawls and, 309 homogeneous fields of direct current and reactions of, 215 stimulants of feeding behavior for, 112 Plate gear, 305 Plates, in groundgear, reducing seabed impact with, 305 Platinum electrodes, 228

Index Playground of Japanese set net, fish behavior in, 168–169 Plexiform layer, of retina, 27 Pol, M., 194 Polet, H., 218 Pollock industry, trawling and bycatch issues in, 300 Pollock (pollachius virens) gillnetting, bycatch reduction and, 195 sound source localization in, 70 sounds produced by, 48 Pop-up satellite archival tags, 275, 276 Porbeagle shark (Lamna nasus) bycatch of, 319 swimming specialization of, 9 Porpoises, 315, 316 Postrelease mortality, 265 Pot and trap fisheries, discard mortality in, 267 Potential biological removal rate, 322 Pots, traps vs., 159 Pound nets, 159, 164 Power density, electric field strength and, 206 Predation discard mortality and, 283 escapee mortality and, 286 post-escape, mortality and, 281 Predation mortality, 268 Predators motion perception and detection of, 29 Ydenberg and Dill’s economic model of reaction distance for fish under threat of, 71 Predatory species, high temporal resolution motion vision in, 30 Presoaked baits, 114 Pressure drag, 7 Pretrawl zone (Zone 1) fish behavior in, 69–75 avoidance patterns, 72–75 reaction distance, 70–72 underwater radiated noise, 69–70 Prey abundance, hunger levels and, 120 Prey density, swimming, gillnet fishing and, 189 Prey detection, olfaction and, 111 Prolonged speeds, 19 Protasov, V. R., 29 Proteins, swimming and use of, 14 PSATs. See Pop-up satellite archival tags Psychoacoustical research methods, on auditory physiology of fish, 51–52 Puffer fish (Lagocephalus wheeleri), fish pots and behavior of, 148

Pulling snood, baited hook and, 117 Pulsating direct current field, fish behavior in, 211–212 Pulse compensation method, 226 Pulsed current, 206 Pulse duration electric field strength and, 206 swimming reactions and, 213–214 Pulse frequency, electrofishing injury and mortality and, 230 Pulse rate, voltage gradient and, for inducing electrotaxis in species studies, 216t Pulse shape, electrofishing injury and mortality and, 230 Pulse shape, fish capture and, 212 Pump fishing, for pelagic fish, 226–227 Pupil, 25 Purkinje phenomenon, 27 Purkinje shift, 29 Purse-seine fisheries acoustic device usage and, 331 mackerel mortality and fish density in, 267 operational changes and reducing nontarget bycatch by, 316 Purse-seines fishing with lights and, 37 sea turtle bycatch and, 318 Pursuit effect, 73 Pyanov, A. I., 94 Q Quantitative flow visualization techniques, 13 Quarter-sine waveform, in electrofishing, 207 Quarter-sine waves, fish repulsion and, 212 Quota-induced discarding, 242 R Rabbitfish (Siganus fuscescens), 113 Radial Escape Section, 86 Radiated noise, underwater, fish behavior in pretrawl zone and, 69–70 Rainbow trout (Oncorhynchus mykiss) gillnet fishing and behavior of, 190 homogeneous fields of direct current and reactions of, 215 mucus experiments with, 8 necropsy filets of, revealing hemorrhages and vertebrae damage caused by electrofishing, 230 radiographs revealing spinal misalignment and fractured

Index vertebrae caused by electrofishing, 229 stimulants of feeding behavior for, 112 Rainbow wrasse (Coris julis) body features of, 11 swimming specialization of, 9 Raised footrope trawls, 302 Ramm, D. C., 301 RAMP (reflex action mortality predictor), 276, 286 Rate of work, as function of swimming speed, 16 Rays (Rajiformes) bodily shape of, 11 discards of, 265 fins of, 6–7 fishery bycatch of, 318 Reaction distances, 92 of acoustically tagged cod, in response to approaching vessel and trawl, 72, 73 benthic species and, 76 fish behavior in pretrawl zone and, 70–72 from predator’s threat, 71 Reactive detection zone, electrofishing, 210 Reactive oxygen species, vitamin E and, 52–53 Read, A. J., 127, 315 Recovery Plan for the North Atlantic Right Whales, 323 Redfish (Sebastes marinus) French-rigging and targeting of, 300 grids and bycatch reduction with, 249 maximum sustained swimming speed and endurance at prolonged speeds in, 20t Red hake (Urophycis chuss), detection threshold in, 113 Red king crab (Paralithodes camtschaticus), fish pots and bycatch of, 150 Red lateral muscles, pelagic fish and use of, 19 Red mullet (Mullus barbatus), mortality of those encountering but escaping fishing gear, 268 Red muscle fibers, 6, 6, 8 Red sea bream (Pagrus major) acoustic attraction and, 53 contrast threshold in, 36 glutamine thresholds in, 113 as hearing generalist, 50, 50 visual acuity expressed by minimum separable angle relative to body length in, 34

Red snapper capture of, in Gulf of Mexico shrimp fishery, 259 Red snapper (Lutjanus campechanus) bycatch reduction for, 252–253 passage through extension and, 83 Red snappers (Lutjanus malabaricus and L. erythropterus), bottom trawls and harvesting of, 301 Reef nets, 160 Reflex impairment, predicting mortality of unrestricted fish with, 276 Refractive index gradient, within lens, 26 Regulatory frameworks for minimizing bycatch, 321–325 AIDCP program, 322–323 European Union, 321–322 protection of other marine megafauna, 324–325 Right Whale Protection, 323–324 sea turtles protection, 324 Southeast Asia, 322 United States, 322 UN Resolution 44/225, 323 Reid, D. G., 76 Release tubes, improving survival of discards and, 285 Repeated capture and escape, mortality and, 281 Repelling electrode systems, use of, 219 RES. See Radial Escape Section Research Institute of Oita Prefectural Government (Japan), 54 Research vessels, sounds produced by, 48 Respiration, swimming and, 14 Resting metabolic rate, 16 Retention triggers, fish pots, 149 Retina, 26, 26–27, 40 changes in position of layers of, during dark adaptation, transitional stage, and light adaptation, 32 neural layers of, 27 Retinomotor response, 31 Revill, A. S., 254 Rheotaxis, bait odor source location and fish orientating by, 116–117 Rheotropism, 171 Rhinoceros auklet (Cerorhinca monocerata), gillnetting and reducing bycatch of, 195 Richards, A., 251 Right Whale Recovery Plan, 323–324 Right whales (Eubalaena glacialis), fish pots and interaction by, 152

367 Rigid exclusion grids, 249, 332, 332–333 Ripples, beam trawling and elimination of, 295 Risso’s dolphin (Grampus griseus), acoustic attraction and, 53 Rivers, electric fishing in, 231 RIVO (Netherlands Institute for Fisheries Research), 220, 225 Robinson, C. J., 189 Rockhoppers, 297, 304, 310 Rock sole (Pleuronectes bilineatus), codend passage and behavior of, 84 Rod and reel fishing, 105 Rods, 40 light intensity and, 28 retinal, 26 Rollerballs, on groundgear, reducing seabed impact and, 303 Rollers, 297 on groundgear, seabed impact and, 302–304 ROS. See Reactive oxygen species Rose, C., 307, 308 Ross, Q. E., 60 Ross corals (Pentapora foliacea), 153 Rougheye rockfish (Sebastes aleutianus), longline catch rates of sablefish negatively correlated with, 122 Roughness, on scales, 8 Rough scad (Trachurus lathami), electrotaxis induced in, voltage gradient combined with pulse rate and, 216t Roundfish fountain maneuver of, in response to trawl doors, 75, 75, 76 herding patterns: between net mouth and trawl doors, 75–76 Round herring (Etrumeus teres), electrotaxis induced in, voltage gradient combined with pulse rate and, 216t Round scad (Decapterus punctatus), electrotaxis induced in, voltage gradient combined with pulse rate and, 216t Roussette (Scyliorhinidae), homogeneous fields of direct current and reactions of, 215 RS Toftevaag research vessel deterrent signal research and, 329 observations of pilot whales taken with onboard DIDSON unit, 330 Rudstam, L. G., 189

368 Rumpon, 58 Rush, hooking and, 117 Ryer, C. F., 268, 281 S Sablefish (Anoplopoma fimbria) bait location on longlines by, 117 fish pots and behavior of, 148 fish pots for, 144 food deprivation in, and response to bait odor concentrations, 119 hooking behavior and, 117 longline catch rates of, negative correlations with fish size and, 122 Northeastern Pacific Ocean demersal fisheries, 107 reflex impairment in, 276 thermoclines and discard mortality for, 282 water temperature and escape mortality for, 280 Saccule, 49, 49, 50 Sailfish (Istiophorus platypterus), swimming specialization, 9 Sailfish (Istioporus inducus), fish aggregation devices and, 58 Sainsbury, J., 143 Sainte-Marie, B., 115 Saithe (Pollachius virens) baited gillnets and capture of, 192 deceleration rate for, 10 endurance values for swimming in mouth of trawl and, 80 entrance into trawl net by, 83 maximum sustained swimming speed and endurance at prolonged speeds in, 20t maximum swimming speed relative to body length of, 21 Northwestern Atlantic Ocean demersal fisheries and, 106 steadily swimming, flow behind, 14 swimming specialization of, 9 swimming styles of, 12 Salmon bycatch cod trap modification and reduction in, 173 whitefish trap modifications and reduction in, 174 Salmon (Salmo salar), electrophysiological studies of, 113 Salmon traps, 170 early, 160 new floating, 160, 160

Index Salmon/whitefish traps, grid designs to reduce interaction of seals with, in Baltic Sea, 176, 176 Samarayanka, A., 186 Sand clouds, 67 bottom trawl fishing system, 68 herding by, 37 herding patterns and, 76 Sand eels (Ammodytes tobianus), 11 entrance into trawl net by, 83 swimming specialization of, 9 Sand flathead (Platycephalus bassensis) endurance values for swimming in mouth of trawl and, 80 extension and behavior of, 83 Sand fleas (amphipods), 283 Sandstar (Astropecten irregularis), electric beam trawl and, 309 Sand tiger shark (Carcharias taurus), protected status for, 325 Sand whiting (Sillago ciliata), mortality of those encountering but escaping fishing gear, 268 Sandy bottoms, experimental trawling and impact on, 296 Sangster, G. L., 255 Santos, M. N., 189 Sara, R., 160 Sardine (Sardinops melanosticta) behavior of, around Japanese set nets, 167 electric behavior experiments with, 213 pump fishing for, 226 ratio of mesh size of trap leader and, 168 Sashimi market, bigeye tuna prices and, 107 Sasso, C. R., 126, 280 Satiation, swimming, gillnet fishing and, 189 Saury (Cololobis saira), pump fishing with underwater light for, 226 Savings Trawl Net Company, 249 Sawfish, 315 Scads (Decapterus spp.), fish aggregation devices and, 58 Scaled sardine (Harengula jaguana), electrotaxis induced in, voltage gradient combined with pulse rate and, 216t Scales, 8 Scallop dredge fisheries, sea turtle bycatch and, 318 Scallop dredges, excluder frame design for preventing turtle bycatch in, 335

Scallop fisheries, loggerhead turtle bycatch and, 319 Scarsbrook, J. R., 152 Scavenger mortality, 268 Scavenging, discard mortality and, 283 Scented bait types, 337 Scholik, A. R., 52, 53 “Schooling behavior” energy-saving effects with, 17–18 fish densities and, 91 Schooling species, fish pot approach by, 147 Schreck, C. B., 229, 230 Sciaenids, 48 Sclera, 26 Scotopic vision, 27, 36, 40 Scottish bag nets, 164 Scup, swimming styles of, 12 Sea bass (Dicentrarachus labrax) electric behavior experiments with, 213 electric field and behavior of, 213 French-rigging and, 301 maximum swimming speed relative to body length of, 21 Sea bass landings, fish pots and, 144 Seabed beam trawls and impact on, 309–310 drop-out panels for reducing benthos catch, 310 electric stimuli used in brown shrimp beam trawls, 309 electric stimuli used in flatfish beam trawls, 309 groundgear modifications and reducing impact on, 302–305 bottom trawling with drop chains and weights, 301–302 depressors and plates as groundgear, 305 lighter groundgear for offshore shrimp trawls, 302 nylon tickler brushes as groundgear, 304 raised footrope trawls and sweepless trawls, 302 rollers and wheels used in groundgear, 302–304 soft chain brush groundgear, 304–305 other trawl gear components and impact on, 306–308 devices to keep bridles off the bottom, 307–308 short bridles, 306–307 use of small warp/depth ratio, 308 pelagic and semipelagic trawls and reducing impact on, 299–301

Index Alaska pollock trawls, 299–300 semipelagic trawling for shrimp, 300 semipelagic trawling using French-rigging, 300–301 trawl door considerations and impact on, 305–306 new “batwing” door, 305–306 use of efficiency and high aspect ratio doors, 305 trawling impact on, review of recent studies, 295–296 trawling styles with reference to their interaction with, 298 Seabirds concerns about incidental catches of, 123–124 interactions of, with gillnets, 195–197 longlining and reducing incidental catches of, 105 mitigation measures to reduce incidental capture of, 124–126, 131–132 Newfoundland traps and, 174–175 night setting, and reducing incidental bycatches of in longlining, 121 Sea catfish (Arius felis), detection thresholds for, 113 SEAFDEC. See Southeast Asian Fisheries Development Centre Seahorse (Hippocampus ramulosus) camouflage protection for, 11 homogeneous fields of direct current and reactions of, 215 swimming specialization of, 9 Seal excluder devices, 333 Sea lions, 315 fishing gear and interactions with, 320 fishing gear incidences with, 318 Seals, 315, 321 Baltic fish traps and, 175–177 fishing gear and interactions with, 320 fishing gear incidences with, 318 grid designs to reduce interaction of, with salmon and whitefish traps in Baltic Sea, 176, 176 Seal scarers, 331 Sea pens (Penatula phosphorea; Virgularia mirabilis; Funiculina quadrangularis) fish pots and, 153 Seasonality feeding motivation, hooking behavior and, 119–120

physiological condition of fish and, 93 Sea turtles, 315, 316 endangered species among, 318 estimating actual bycatch levels for, 319 exclusion devices and, 331–332 interactions of, with gillnets, 195–197 longlining and reducing incidental catches of, 105 mitigation measures to reduce incidental capture of, 126–128, 132 mortality, 265 pound net fishery in mid-Atlantic coast of U.S. and, 177–178 protection of, 324 towing time and escape mortality for, 280 turtle chains and, 334 Seawater electric fishing in, 228, 231 obtaining electric control of fish in, 206 spectral properties of, 35 SEDs. See Seal excluder devices Seidel, W. R., 222, 227 Seine fisheries, discard mortality in, 267 Seismic air-gun noise, reductions in longline catches of small cod and, 131 Selection range, 244 Selective electrified otter trawls, 222, 222 Selective gears, incentivizing use of, 259 Selectivity, gillnet, 187–189 Self-spreading groundgear, 305 Semicircular canals, 48, 49 Semidiurnal vertical migrations, 189 Semipelagic longlining monofilament lines and, 109 setting, 110 Semipelagic trawls, 296, 310 with French rigging, 300–301 interaction of, with seabed, 298 interaction of, with seabed, 298 for pink shrimp, 300 Sensory hair cells, in fish ears, 49 Separator grids, inclined hydrodynamic performance of, 86 orientation and size of fish and, 87 Separator trawls Barents Sea experiments with, 257 design of, 256

369 SEPSA. See Shrimp electric pulse stimulus apparatus Set gillnets, 183, 184t Set nets (or setnets), 159 capture efficiency of, 169 Japanese, 161–162 Sewell, J., 152 Sharber, N. G., 227 Sharks, 315, 316 best hanging ratio for capture of, 185 conservation of, 324 discards of, 265 documenting extent of bycatch for, 319 fins of, 6–7 fishery bycatch of, 318 fishery operations and bycatch of, 318 fish skin of, 8 large, reducing bycatch of, 338 longlining and reducing incidental catches of, 106 magnetic fields and, 337 mitigation measures to reduce incidental capture of, 128–130, 132 pingers and bycatch reduction of, 328 protection programs for, 325 Sharpnose tigerfish (Rhyncopelates oxyrhynchus), acoustic attraction and, 53 Sheaves, M. J., 149 Shellfish, beam trawls and harvesting of, 296 Shellfish pots, 159 Shiga, M., 128 Shiner perches (Cymatogaster aggregata), pectoral fin movements of, 12 Shiode, D., 128, 332 Shoes, beam trawls, 296, 299 Short bridles, reducing seabed impact and use of, 306–307 Shortfin mako shark (Isurus oxyrhinchus), regulatory protection for, 325 Short-finned pilot whale (Globicephalas melas), bycatch of, 318 Shrimp, electric fields and trawling for, 220–224 Shrimp, pink shrimp (Pandalus borealis), separator trawls and behavior of, 256 Shrimp biomass reduction, in China, 224

370 Shrimp electric pulse stimulus apparatus, 224 Shrimp fisheries bycatch, discard and, 240 successful implementation of mitigation measures in, 258 unwanted discards and, 251 Shrimp trawl fisheries, sea turtle bycatch and, 318 Shrimp trawls offshore, lighter groundgear for, 302 selective, water flow alteration and, 88–89 Side-opening grids, 87 Side-setting, 132 Siira, A., 268 Silky shark (Carcharhinus falciformis), bycatch of, 319 Siluroids, hearing ability of, 50 Silva, R., 178 Silver anchovy (Engraulis eurystole), electrotaxis induced in, voltage gradient combined with pulse rate and, 216t Silver hake (Merluccius bilinearis) raised footrope trawl and, 302 reducing bycatch of, 252 Silver sharks (chimaeras), 318 Simple pulse pattern, 206 Sine waveform, in electrofishing, 207 Single-species fisheries excluding unwanted species from, 258–259 reducing unwanted bycatch in, 250–253 Sirenians, 315 Sivak, J. G., 26 Size of bait, 118 Size of fish electric fields and, 212–213 escape mortality and, 277–278 fish behavior near trawls and, 91–92 inclined separator grids and, 87 response curves for voltage gradient by frequency, temperature and, 217 responses to bait and, 121–122 swimming speed, gillnet fishing and, 189 Size selectivity adjusting, examples of mechanisms for, 247 gillnets and, 184–185, 187–189, 198 reducing target species discards by controlling, 243–250 traps and, 172

Index Skate (Raja batis) homogeneous fields of direct current and reactions of, 215 swimming specialization of, 9 Skates (Raja spp.) discards of, 265 endurance values for swimming in mouth of trawl and, 79 entrance height into trawl net by, 83 herding behavior of, in response to sweep of bottom trawl, 77 herding patterns and, 76 Skin damage, escape mortality and, 277–278 Skipjack tuna (Katsuwonus pelamis) best hanging ratio for capture of, 185–186 fish aggregation devices and, 58 maximum swimming speed relative to body length of, 21 Skirt line, gillnet, 185 Skjervoy 3600, Bosum part of, 303 Slack-Smith, R. J., 143 Slope net, fish behavior at, 169 Small-mesh fish pots, fish behavior on approach to, 147 Smalltooth sawfish (Pristis pectinata), fishing gear interactions and, 320 Smalltooth Sawfish Recovery Team (NMFS), 320 Smith, E. M., 286 Smolowitz, R., 319 SMPs. See Square mesh panels SMR. See Standard metabolic rate Snagging, 184, 186 Snappers (Pagrus auratus), fish pots for, 145 Snapping shrimps, sound production by, 52 Snell notation, visual acuity expressed in, 31t Snood length of, 109 longline gear, 108, 108 materials used for, 109 Snow crab (Chionoecetes opilio), gillnetting and reducing bycatch of, 195 Snyder, D. E., 206, 209, 227, 229, 230 Soaking durations, predicted fishing range of gillnets relative to, 191 Soak time, 132 fish pots and, 149 gillnet catch relative to, 192 sea turtle bycatch in longlining and, 127 shark catch rates and, 129 Social learning, defined, 94

Sockeye salmon (Oncorhynchus nerka), metabolic rate and endurance of, as functions of swimming speed, 18, 18 Soft bottoms, trawling and impact on, 296 Soft brush chain arrangement, 304 Soft chain brush groundgear, 304–305 Solar radiation, underwater natural light and, 35 Soldal, A. V., 270 Solenocera crassicornis, electrobeam trawl vessels and, 223 Sole (Solea solea), homogeneous fields of direct current and reactions of, 215 Solid steel grids, 333 Soria, M., 94 Sort-V selection grid, 247 Sound fish guidance devices and use of, 54–57 responses of fish to, and application in fisheries, 53–60 acoustic attraction, 53–54 aversive sound to reduce fish entrapment, 60 fish aggregation devices and, 57–59 fish guidance devices and, 54–57 underwater radiated noise, pretrawl zone and, 69–70 Sound field, 45–46 Sound pressure levels calculating, 46 underwater, variations in, 46–47 of various types of sound sources, 47 Sound sources, 45–46 types of, 46 underwater, characteristics fish produced, 48 natural ambient sounds, 47–48 produced by fishing, research, and whale-watching vessels, 48 underwater, sound pressure levels of types of, 47 Southeast Asia, marine mammal protections in, 322 Southeast Asian Fisheries Development Centre, 255, 332 Southern right whales (Eubalaena australis), entanglement incidences with, 317 South Korean tuna fleet, 108

Index Spanish sardine (Sardinella anchovia; Sardinella aurita) electrotaxis induced in, voltage gradient combined with pulse rate and, 216t endurance values for swimming in mouth of trawl and, 79 Spanish system, longline setting, 109 Spawning aggregations, motivational state and harvesting of, 92 Spawning condition, swimming, gillnet fishing and, 189 Species/populations/regions, marine mammal bycatch by, 319 Species selectivity bait type and, 131 improved, designs aimed at, 250 Spectral absorption of light, in open oceans, 36 Spectral properties of seawater, 35 Spherical body, friction for given volume and, 7 Spherical electrodes, 227–228 Spinal injuries, electrofishing and, 229–230 Spinal misalignment, electrofishing and, 229 Spines, on scales, 8 Spinner dolphin (Stenella longirostris), fishing operations and threat to, 316, 317 Spiny dogfish (Squalus acanthias), Northwestern Atlantic Ocean demersal fisheries and, 106 SPLs. See Sound pressure levels Spot (Leiostomus xanthurus), electrotaxis induced in, voltage gradient combined with pulse rate and, 216t Spotlined sardine (Sardinops melanostictus), auditory threshold curves for, 50, 51 Spotted warehou (Seriolella punctata), codend and behavior of, 84 Sprat (Sprattus sprattus), maximum swimming speed relative to body length of, 21 SPRs. See Species/populations/regions Square, otter trawl, 297 Square mesh codend panel, 247 Square mesh codends, 84, 86, 247, 249 Square mesh panels, 251 discards reduction and use of, 243 fish escape rates and, 83 mesh size, size selectivity of target species and, 249 Square-mesh window, 86

Square PDC waveform, in electrofishing, 207 Square pulse train waveform, in electrofishing, 207 Squid bait, swordfish catch rates and, 128, 132 Squid (Loligo pealeii) endurance values for swimming in mouth of trawl and, 80 entrance height into trawl net by, 83 fishing with lights, 37 SR. See Selection range Stainless steel electrodes, 228 Stake nets, 159 Standard metabolic rate, 16 Stationary-acoustic receivers, 275 Station keeping, fish density and, 91 Stellar sea lions, unpalatable bait types and, 337 Sterling, D., 304, 305 Sternin, V. G., 212, 229 Stevens, B. G., 268 Stewart, J., 149 Stewart, P. A. M., 67, 214, 218, 219 Stock health, reducing fish mortality and, 259 Stock size estimates, unbiased, producing, 131, 133 Stone, A. W., 153 Stoner, A. W., 122, 131, 133 Strange, E. S., 76 Stratum compactum, 7, 8 Streamer lines, 131, 132 bird-scaring, 125, 125, 126 Streams, electric fishing in, 231 “Street sweeper,” 304 Stress capture and escape processes and, 277 electrofishing and, 230 repeated capture and escape and, 281 Stride length, 12 Stridulation sounds, drumming sounds vs., 48 Striped bass (Morone saxatilis) maximum sustained swimming speed and endurance at prolonged speeds in, 20t maximum swimming speed relative to body length of, 21 Striped beak-perch (Oplegnathus fasicatus), contrast threshold in, 36 Subcarangiform fish, anodic electrotaxis and, 213 Sulaeman, M., 186 Sunfish (Lepomis sp.), homogeneous fields of direct current and reactions of, 215

371 Sunfish (Mola mola) locomotion behavior of, 10 swimming specialization of, 9 Sunken leader design, cod trap modification to reduce salmon bycatch and, 173 Sunken leader trap, design of, 172 Sunrise goatfish (Upeneus sulphureus), catch volume/quality and escape mortality for, 279 Supercritical zone, electrodes and, 228 Supersharks, 318 Super-Shooter device, 251 Surface set nets, 162 Surfperches (Embiotocidae), pectoral fin movements of, 12 Survival measures improving for discards, 285–286 improving for escapees, 281–285 increasing escape opportunities, 284 modification of fishing practices, 284 modification of groundgear, 284–285, 287 reducing time between capture and escape, 283–284 Sustained speeds, 19 Suuronen, P., 176, 177, 270, 273, 279 Swedish coastal Nephrops fisheries, Nordmøre style grids and, 252 Sweep bottom trawl fishing system, 68 Sweep angle, herding patterns and, 76 Sweep length, increasing, catch rates and, 76 Sweepless trawls, 302, 310 Sweeps, 67 Swimbladder, 50 Swimming fish behavior and classes of, 213 differences in, across the species, 5 herding patterns and variability in, 76 in marine fish, 5–21 concluding remarks, 10 energy required for swimming, 14–18 fish wakes, 13–14 introduction, 5 styles of swimming, 12–13 swimming apparatus, 5–8 swimming-related adaptations, 8, 10–12 swimming speeds and endurance, 18–19

372 Swimming crab (Liocarcinus holsatus), electric beam trawl and, 309 Swimming flumes, 81 Swimming specializations, representation of, 8, 9 Swimming speeds classification of, 19 endurance and, 18–19 gillnet fishing and, 189 metabolic rate and endurance of sockeye salmon, as functions of, 18, 18 rate of work as function of, 16 Swimming trajectory advancing sweeps and selection of, 78 herding patterns and, 76 Swivels, catching efficiency and, 109 Swordfish catch rates, mackerel bait vs. squid bait, 128, 132 Swordfish (Xiphias gladius) acoustic devices in gillnet fisheries and, 328 body features of, 11 pelagic longline fisheries and, 107 swimming specialization of, 9 temporal resolution motion vision in, 30 Symes, D., 242 T Tadpole-fish (Raniceps raniceps), sounds produced by, 48 Tag and recapture methodology, predicting mortality of unrestricted fish with, 275 Tagging, electronic, predicting mortality of unrestricted fish with, 275 Tagging methods, discard-mortality studies and, 269 Tail beat frequency, maximum burst speed and, 19 Tail fin, 5 Tails high-aspect-ratio, 10 maximum amplitude in, 12 Takagi, T., 185, 197 Take reduction teams, 322 Tamasauskas, P., 219 Tangling, 186 Tanner crab (Chionoecetes bairdi) fish pots and bycatch of, 150 pot and trap fishing and discard mortality of, 268 Target detection, visual acuity and, 31 Target recognition, visual acuity and, 31

Index Target strength, 51 Taste, of bait, 117 Taste buds, 110 Taxis zone, electrofishing, 210 Technical conservation measures (TCMs), for reducing discards, 242 TEDs. See Turtle excluder devices Teleost fin rays kinds of, 7, 7 structure of, 7 Teleosts binocular vision, monocular vision, and blind zone in, 30 flicker fusion frequency in, 29 retina of, 26 Temperature endurance, swimming speed and, 19 eye, 30 fish interaction with longlines and, 121 gillnet fishing, reduction of activity and swimming capacity relative to, 190 gillnet fishing and, 189 Newfoundland cod trap and, 167 predicted fishing range of gillnets relative to, 191 response curves for voltage gradient by length of fish, frequency and, 217 sea turtle bycatch and, 127–128 water, escape mortality and, 280 water conductivity and, 212 Temperature shock, discard mortality and, 268 Temporal summation, motion vision and, 28 Tench (Tinca tinca), homogeneous fields of direct current and reactions of, 215 Tetanus, electric field experiments and, 213 Tetanus zone, partial to full, electrofishing, 210 Tetraodontiforms, 213 Thermal regulation systems, fish vision and, 30 Thermoclines, discard mortality and, 282 Thomsen, B., 253 Threadfin shad (Dorosoma petenense), pound net fishery in mid-Atlantic coast (U.S.) and, 178 Thread herring (Opisthonema oglinum), electrotaxis induced in, voltage gradient combined with pulse rate and, 216t

Threestripe tigerfish (Terapon jarbua), acoustic attraction and, 53 Thresher sharks (Alopias vulpinus), regulatory protection for, 325 Threshold value of field strength, fish behavior and, 228 Tickler chain, traditional, 304 Tie-down lines, gillnets, 187 Tiger flathead (Neoplatycephalus richardsoni) codend passage and behavior of, 83–84 endurance values for swimming in mouth of trawl and, 80 Tiger prawn (Penaeus esculentus), soft brush footgear and, 304 Tiger sharks (Galeocerdo cuvier), bycatch of, 319 Tilapia (Tilapia mossambica) homogeneous fields of direct current and reactions of, 215 slackly-hung gillnets and capture of, 186 Time tension line cutters, 336 Tjøstheim, D., 74 T90 codends, experiments with, 248 Todd, S. K., 175 Toivonen, A., 164 Tokai, T., 332 “Toms effect” fish, mucus and, 8 Total catch per haul reductions, discard mortality and, 286 Towing duration, discard mortality and, 282–283 Towing speed and duration, escape mortality and, 279 Trachypenaeus curvirostris, electrobeam trawl vessels and, 223 Traditional Newfoundland cod trap, 163, 163 Traditional Newfoundland trap, Japanese trap vs., 164 Trammel nets, 183, 184t Trap berths, 171 Trap design, 178 fish behavior and, 170–172 satisfying conflicting goals with, 171, 178 sunken leader, 172 trap fisheries and, 159–165 Trap fisheries conservation issues and mitigation measures in, 174–178 trap designs and, 159–165 Traps alarms, 175 Baltic fish traps, 164–165

Index bluefin tuna, 164 capelin, 164, 172 cod, 159–160 defined, 159 early, 160–161 fish behavior in and around, 165–170 herring, 164 large-scale, 159–178 mackerel and herring, 164 Newfoundland cod traps, 162–164, 163, 165–167 pots vs., 159 in relation to bottom contour, headland, and typical fish swimming route, 171 salmon, 170 seals interacting with, 175–177 size selectivity and, 172 species selection and bycatch reduction and, 172–173 successful operation of, 178 survival of fish discarded from, 173–174 whale collisions with, 175 whitefish, 164 whitefish/salmon, 164 Trawl codends assessing mortality of fish escaping from, 270, 273–274 bias caused by collection of escapees, 270, 271, 272, 273 bias caused by transporting and holding escapees, 273–274 control samples, 274 usefulness of laboratory experiments, 274 Trawl design, improving escapee survival and, 285 Trawl doors fish behavior between net mouth and (Zone 2), 75–82 herding patterns: benthic species, 76–77 herding patterns: roundfish, 75–76 Trawl-escape mortality studies, escapees caught inside codend covers and, 270, 271, 272, 273 Trawl fisheries cetacean bycatch and, 317 discard mortality in, 267 implementation of discard reduction measures in, 257–258 marine mammal bycatch rates and, 318 operational changes and reducing nontarget bycatch by, 316 shark and ray bycatch and, 318

Trawl fishing historical development of, 67–68 zones in, 68–69, 69 Trawling importance of, 239 seabed impact and, 295–310 beam trawls, 309–310 descriptions of trawls and their operation, 296, 298–299 groundgear modifications, 301–305 pelagic and semipelagic trawls, 299–301 review of recent studies, 295–296 trawl door considerations, 305–306 trawl gear components, 306–308 for shrimp, electric fields and, 220–224 worldwide expansion of, 243 Trawling styles, interaction of, with seabed, 298 Trawl modifications, ongoing discard/ bycatch reductions and, 259 Trawl mouth, fish behaviors in, 77–82 Trawl net, 67 demersal single boat trawl, 240 fish behavior inside codend and, 82–89 bycatch reduction devices and, 85–89 entry and orientation, 82–84 Trawls complexity of, 298–299 descriptions of operations, 296, 298–299 fish vision and capture process of, 36–37 Trawl “surging,” escape mortality and, 281 Trawl warps, herding by, 37 Triggers, fish pots, 149 Trippel, E., 196 Trolling, fishing by, 105 Tropical shrimp fisheries bypass reduction devices and, 88 reduction of turtle species mortality and, 316 Tropical shrimp trawling, anthropogenic turtle mortality and, 252 TRTs. See Take reduction teams Trumble, R. J., 275 TS. See Target strength Tubercles, on scales, 8 Tubular steel grids, 333

373 Tuna, ecological association of dolphins and, 317 Tuna longline fisheries, 107 Turbot (Scophthalmus maximus) stimulants of feeding behavior for, 112 swimming specialization of, 9 Turbulence attractants diluted by, 111 catch-induced, water movement in codend and, 84 Turkish fish pots, 145 Turtle bycatch new excluder frame design for preventing in scallop dredges, 335 rope leader design for reduction of, in mid-Atlantic pound net fishery, 177 Turtle chains, 334 Turtle excluder devices, 86, 252, 258, 316, 337 basic, 332 rigid, example of, 332 Turtles. See also Sea turtles fishing gear interactions and, 318 mortality, 265 Turunen, T., 186 Tusk (Brosme brosme) fish pots for, 145 Northwestern Atlantic Ocean demersal fisheries and, 106 Twine color, gillnet fishing and, 190, 192 Twine filament number, gillnets, 186 Twine size, gillnets, 186, 187 Two-chamber fish pots, 145, 146 U Unaccounted mortalities, 265 Undersized species, fish pots and: catch avoidance escapees, and discards, 150–151 Underwater ambient sounds, characteristics of, 47–48 Underwater sound and vibration properties, 45–47 sound pressure level, 46 sound source and sound field, 45–46 traveling speed of, 53 variations of underwater sound pressure levels, 46–47 Underwater visual environment, fishing gear and, 35–36 Undulatory swimmers, 12 United Nations Environment Program (UNEP), 251

374 United States, marine mammal protections in, 322 University of St. Andrews (UK), 328, 333 Unreported landings, fishing mortality and, 265 UN Resolution 44/225, 323 Upstream swimming, olfactory arousal and, 111 Upward excluding grids, 87 Urick, R. J., 47, 52 Urquhart, G.G., 67 U.S. Fish and Wildlife Services (USFWS), 324 Utricle, 49, 49 V Vagus cranial nerve, 110 Valdemarsen, J. W., 146, 148 van Bergeijk, W. A., 46 Vancouver Island, reef nets used in, 160 Vander Haegen, G. E., 195 van Marlen, B., 220, 229, 309 Vanselous, T. M., 227 Vendace (Coregonus albula) catch volume/quality and escape mortality for, 279 mortality of those encountering but escaping fishing gear, 269 Ventral tip vortices, for steadily swimming saithe, 14, 14 Vents, on fish pots, 150 Vertebral column, in fish, 5 Vertebral fractures, electrofishing and, 229, 230 Vertical avoidance pattern, 74 Vertical hanging ratio, 185 Vertical migrations, 189 Vertical patterning in netting, visual stimulus in fish and, 38 Vertical stop-start vortices, for steadily swimming saithe, 14, 14 Vessel movement, escape mortality and, 281 Vibert, R., 212, 228 Vincent, B., 308 Viscosity, fish mucus and, 8 Visibility of net, gillnet fishing and, 190–191 Vision. See Fish vision Visual acuity, 40 aspects of, 30–31 expressed in minimum separable angle, minimum angle of resolution, Snellen notation, and decimal unit, 31t optical factors related to, 33

Index relative to fish body length in yellowtail, red sea bream, jack mackerel, walleye pollock, and Pacific saury, 34 Visual axis, cone density and, 30 Visual capability of fish, predicting, 35 Visual capacity, 32–35 equation, 32–33 visual acuity, separable angle, and maximum sighting distance, 32–35 Visual cell layer thickness, 31 Visual contrast, of fishing gear, 35–36 Visual function, 27–32 color vision, 27 form vision, 30–31 light vision, 28 motion vision, 28–30 optomotor response, 31–32 retinomotor response, 31 Visual object categorization, 117 Visual stimuli behavior modification and manipulation of, 38 chemical stimuli vs., 111 fish behavior to baited hooks and, 130 Vitamin E, reactive oxygen species and, 52–53 Vitreous body, 26 Voltage, electric field strength and, 206 Voltage gradient pulse rate and, for inducing electrotaxis in species studied, 216t response curves for, by length of fish, frequency, and temperature, 217 Vortex rings chain formation and, 13, 13 energy-saving effects with schooling behavior and, 17 structures, 13 Vortices fish wakes and, 13 saithe tail tip, 14 wake of continuously swimming mullet, 15 W Wahoo (Acanthocybium solandrei), maximum swimming speed relative to body length of, 21 Wakeford, J., 84, 88 Walker, T., 325 Walleye pollock (Theragra chalcogramma) auditory threshold curves for, 50, 51

electroretinogram amplitude in dark-adapted eyes of, 27, 28 maximum sighting distance of, 34 Northwestern Pacific Ocean demersal fisheries and, 107 sound source localization in, 70 visual acuity expressed by minimum separable angle relative to body length in, 34 Walleye (Sander vitreus), geometry of grillnet and, 192 Walsh, S. J., 79, 91, 281 Ward, J., 146, 147, 149 Wardle, C. S., 35, 37, 38, 67, 78, 87, 172, 189, 190, 248 Warp bottom trawl fishing system, 68 demersal single boat trawl, 240 Warp/depth ratio, small, reducing seabed impact with use of, 308 Wassenberg, T. J., 267 Water, acoustic impedance of, 46 Waterborne kites, reducing seabed impact with, 305 Water-borne sound and vibration, 45 Water flow alteration, improved BRD designs and, 88 Water speed variations, fish behavior in extension/codend and, 85 Water temperature escape mortality and, 280 fish behavior near trawls and, 90 Watson, J. W., 88, 127, 222, 336 Waveform, electrofishing injury and mortality and, 230 Wavelength, swimming styles and, 12 Weakfish (Cynoscion regalis), pound net fishery in mid-Atlantic coast (U.S.) and, 178 Webbing, gillnets, 183, 186, 187 Weberian ossicles, 50 Wedging, 184, 186 Weight, gillnet, 185 Weighted filaments, woven into nets, 51 Weighted longlines, 131, 132 Weights, bottom trawling with, 301–302 Weirs, 159, 161 Werner, T., 315, 325 Westenberg, J., 57, 58, 58 Westerberg, H., 175 WGFTFB. See Working Group on Fishing Technology and Fish Behavior Whale alarm, prototoype, 175 Whale bycatch, entanglement incidences and, 317–318

Index Whale entanglement, mitigation measures for, 336 Whale mortality, during 20th century whaling time period, 317 Whale Research Group (Memorial University, Newfoundland), prototoype whale alarm developed by, 175 Whales, 315, 316 collisions, with Newfoundland traps, 175 conservation programs for, 323–324 Whale shark (Rhincodon typus) major conservation problems related to, 318 protected status for, 325 Whale-watching vessels, sounds produced by, 48 Wheels, in groundgear, seabed impact and, 302–304 Whiptails, in codend, behavior of, 84 White catfish (Ictalurus catus), 113 Whitefish behavior, near Baltic traps, 169–170 Whitefish traps, 164 modifications, to reduce salmon bycatch in Baltic Sea, 174 White hake (Urophycis tenuis), Northwestern Atlantic Ocean demersal fisheries and, 106 White lateral muscles, pelagic fish and use of, 19 Whitelaw, A. W., 146, 147, 149 White muscle fibers, 6 White shark (Carcharodon carcharias) bycatch of, 319 major conservation problems related to, 318 Whiting (Merlangius merlangus; Gadus merlangus) maximum swimming speed relative to body length of, 21 Whiting (Merlangius merlangus; Gadus merlangus)

glycine behavioral response thresholds in, 114 mortality of those encountering but escaping fishing gear, 268 separator trawls and behavior of, 256 vertical location of, during passage through extension of trawl net, 83 Wileman, D. A., 269, 280 Window-pane gillnet, 190 seabird bycatch reduction and, 195 Wing, otter trawl, 297 Winged sharks (batoids), 318 Wingends beam trawls, 296 otter trawls, 297 Winger, P. D., 48, 72, 90 Winter flounder (Pseudopleuronectes americanus) mortality of those encountering but escaping fishing gear, 268 temperature, gillnet fishing and, 189 Wire leaders, shark catch rates and, 129 Wires, 297 Wolf, S., 146 Wolffish (Anarchichas lupus), Northeastern Atlantic Ocean demersal fisheries and, 106 Working Group on Fishing Technology and Fish Behavior, 251 World fisheries, bycatch and discard in, 240–242 World War II, 296 World Wildlife Fund, 316 Smart Gear competition, 337 Wrasse (Labridae), fish pots for, 145 WWF. See World Wildlife Fund X Xu, G., 279 Y Yan, H. Y., 50, 52, 53

375 Yanase, K., 84 Ydenberg, R. C., economic model of reaction distance for fish under predator’s threat, 71, 71, 95 Yellowfin bream (Acanthopagrus australis), mortality of those encountering but escaping fishing gear, 268 Yellowfin tuna (Thunnus albacares) best hanging ratio for capture of, 185 fish aggregation devices and, 58 fishing operations and threat to, 317 maximum swimming speed relative to body length of, 21 pelagic longline fisheries and, 107 Yellowtail flounder (Pleuronectes ferruginea; Limanda ferruginea), 254 extension and behavior of, 83 mortality of those encountering but escaping fishing gear, 268 Yellowtail (Seriola quinqueradiata) acoustic attraction and, 53 form vision in, 30 Japanese set net leader and behavior of, 168 visual acuity expressed by minimum separable angle relative to body length in, 34 Yokota, K., 129 Young’s modulus, 8 Z Zalewski, M., 206 Zeeberg, J., 319 Zhang, X. M., 37 Zhou, S., 150 Zollet, E. A., 321, 337 Zones for fish in electric fields, 209, 210 in trawl fishing, 68–69, 69

Figure 2.2. Retinal structure of Pacific saury (Cololabis saira) indicating layers within the retina. (Hajar 2007.)

Figure 2.4. Relative Electroretinographic amplitude in lightadapted eyes (open circle) and two dark-adapted eyes of different time, showing the phenomenon of Pukinje shift. (Zhang 1992.)

Figure 2.6. Changes in the position of layers in the retina during dark adaptation, transitional stage, and light adaptation. p, shifting distance of pigment layer; c, cone position; and A, thickness of visual cell layer. (Zhang 1992.)

Figure 2.10. Spectral absorption of light in the open oceans where light in the green/blue part of the spectrum transmits deeper into the water column than other wavelengths (reprinted from http:// ultramaxincorp. com/?p2=/modules/ ultramax/catalog. jsp&id=23).

Figure 4.14. Typical flow field observed throughout a standard extension/codend containing a bycatch reduction device. (Data from Wakeford 2004.) Variation in water speed through the trawl is known to modify fish behavior. Under certain conditions, turbulent flow and eddies are expected, creating areas of flow reversal as shown here near the top of the codend.

Figure 5.5. Learning/ conditioning: this cod was able to locate and ingest four acoustic tags wrapped in mackerel bait by associating the sound of the research vessel with the presence of food.

Figure 9.12. The Dutch electrobeam trawl. (Photo: Jochen Depestele.) Figure 9.15. Necropsy filets of rainbow trout revealing hemorrhages and associated tissue and vertebrae damage caused by electrofishing. (From Snyder 2003 with permission.)

Figure 10.1. Principal components of a demersal single boat trawl. (Crown copyright, reproduced with the permission of Marine Scotland.)

Figure 10.5. Examples of mechanisms to adjust size selectivity. (A) Square mesh codend. (B) Square mesh codend panel. (C) Sort-V selection grid. (Crown copyright, reproduced with the permission of Marine Scotland.)

Figure 10.6. Examples of designs aimed at improving species selectivity. (A) Inclined separator panel. (B) Nordmøre shrimp grid. (Crown copyright, reproduced with the permission of Marine Scotland.)

Figure 10.7. Cutaway trawl aimed at improving species selectivity. (Crown copyright, reproduced with the permission of Marine Scotland.)

Figure 10.9. Examples of designs to segregate and select different species components. (A) Horizontal separator trawl used in mixed finfish or mixed Nephrops/finfish fisheries. (B) Horizontal separator with guiding ropes to ‘encourage’ cod into the upper compartment. (Crown copyright, reproduced with the permission of Marine Scotland.)

Figure 10.10. Doublegrid and codend system for improving size selection of Nephrops and finfish. (Crown copyright, reproduced with the permission of Marine Scotland.)