Locomotion, Volume 7: Locomotion (Fish Physiology)

FISH PHYSIOLOGY Volume VII Locomotion

CONTRIBUTORS F. W. H. BEAMISH QUENTIN BONE

C. C. LINDSEY

WILLIAM R. DRIEDZIC

WILLIAM H. NEILL

JOHN J. MAGNUSON

P. W. HOCHACHKA

DAVID J. RANDALL

DAVID R. JONES

E. DON STEVENS PAUL W. WEBB

FISH PHYSIOLOGY Edited by W. S. HOAR DEPARTMENT OF ZOOLOGY UNIVERSITY OF BRlTISH COLUMBIA VANCOUVER, CANADA

and

D. J. RANDALL DEPARTMENT OF ZOOLOGY UNIVERSITY OF BRlTJSH COLUMBIA VANCOUVER, CANADA

Volume VII

Locomotion

ACADEMIC PRESS New York San Francisco London 1978 A Subsidiary of Harcourt Brace Jovanovich, Publishers

COPYRIGHT 6 1978, BY ACADEMIC PRESS, INC. ALL RIGHTS RESERVED. NO PART OF THIS PUBLICATION MAY BE REPRODUCED OR TRANSMITTED IN ANY FORM OR BY ANY MEANS. ELECTRONIC OR MECHANICAL, INCLUDING PHOTOCOPY, RECORDING, O R ANY INFORMATION STORAGE AND RETRIEVAL SYSTEM, WITHOUT PERMISSION IN WRITING FROM THE PUBLISHER.

ACADEMIC PRESS, INC.

111 Fifth Avenue, New York. New York 10003

United Kingdom Edition published by ACADEMIC PRESS, INC. (LONDON) LTD.

24/28 Oval Road, London NW1

7DX

Library of Congress Cataloging in Publication Data Hoar, William Stewart, Date Fish physiology. Includes bibliographies. CONTENTS: v. 1. Excretion, ionic regulation, and metabolism.--v. 2. The endocrine system.--v. 3. Reproduction and growth, bioluminescence, pigments, and poisons. [etc.] 1. Fishes--Physiology. I. Randall, D. J., joint author. 11. Conte, Frank P., Date 111. Title. QL639.1.H6 597’.01 76-84233 ISBN 0-12-350407-4 ( V. 7 )

PRINTED IN THE UNITED STATES OF AMERICA

8182

9 8 7 6 5 4 3 2

CONTENTS ix

LIST OF CONTRIBUTORS

xi

PREFACE

TERMINOLOGY TO DESCRIBE

SWIMMING ACTIVITYIN

FISH

xiii xv

CONTENTS O F OTHER VOLUMES

1. Form, Function, and Locomotory Habits in Fish C. C . Lindsey I. Introduction 11. Early History of Studies on Fish Locomotion 111. Modes of Swimming IV. Nonswimming Locomotion V. Propulsive Anatomy VI. Locomotory Habits of Wild Fish References

1 5 8 39 46 75 88

2. Swimming Capacity F. W. H . Beamish I. Introduction 11. Field Measurements of Performance 111. Laboratory Measurements of Performance IV. Energetics of Swimming V. Application to Management Practices References

101 103 117 163 168 172

3. Hydrodynamics: Nonscombroid Fish Paul W. Webb 1. Introduction Steady Swimming Unsteady Propulsion Unsteady versus Steady Propulsion Mechanics of Median and Paired Fin Propulsion References

11. 111. IV. V.

V

190 191 22 1 228 230 232

CONTENTS

vi 4. Locomotion by Scombrid Fishes: Hydromechanics, Morphology, and Behavior John J . Magnuson 11. Observed Swimming Speeds 111. General Considerations of Swimming Equilibria

240 24 1 250

IV. height, Buoyancy, Hydrodynamic Lift, and Prediction of Sustained Speeds V. Resistance to Forward Movement VI. Thrust Production References

25 1 267 288 308

I. Introduction

5. Body Temperature Relations of Tunas, Especially Skipjack E . Don Stevens and William H. Neil1 I. Introduction 11. His tory 111. What Is the Real (i.e,,Typical) Excess Body Temperature of Tunas? IV. What Is the Heat Source Responsible for Large Excess Body Temperatures in Tunas? V. Anatomical Basis oif Warm-Bodiedness in Tunas VI . Exchange of Heat between Tunas and Their Environment VII. Do Tunas Regulate Their Body Temperature? VIII. Adaptive Values of Warm-Bodiedness and Large Thermal Inertia IX. Physiological Insights into the Natural History of Tunas X. Conclusion: A Thermocentric Overview of Tuna Evolution References

3 16 317 319 32 1 325 334 340 348 353 354 356

6. Locomotor Muscle Quentin Bone I. Introduction 11. The Organization of the Myotomes

111. IV. V. VI.

Fin Muscles Fiber Types Proprioception Fish Muscle and the Muscles of Higher Forms References

36 1 363 368 368 4 10 416 417

CONTENTS

vii

7. The Respiratory and Circulatory Systems during Exercise David R . Jones and David J . Randall I. General Introduction 11. Assessment of Exercise Performance 111. The Respiratory System during Exercise

IV. The Circulatory System during Exercise References

425 426 442 466 492

8. Metabolism in Fish during Exercise William R . Driedzic and P. W. Hochachku I. Introduction 11. Biochemical Insights from Respiratory Physiology

111. Red-White Muscle Differences I v. Metabolism of Adenylates and Related Compounds V. Carbohydrate Metabolism VI . Lipid Metabolism VII. Protein Metabolism VIII. Citric Acid Cycle References

503 504 505

507 517 525 530 533 536

AUTHOR INDEX

545

SYSTEMATICINDEX

56 1

SUBJECTINDEX

570

This Page Intentionally Left Blank

LIST OF CONTRIBUTORS Numbers in parentheses indicate the pages on which the authors’ contributions begin.

F. W. H. BEAMISH (101),Department of Zoology, University of Guelph, Guelph, Ontario N 1 G 2W1, Canada QUENTINBONE(361),The Marine Laboratory, Citadel Hill, Plymouth PLl 2PB, United Kingdom WILLIAMR. DRIEDZIC* (503),Department of Zoology, University of British Columbia, Vancouver, British Columbia V6T 1W5, Canada

P. W. HOCHACHKA (503), Department of Zoology, University of British Columbia, Vancouver, British Columbia V6T 1W5, Canada DAVIDR. JONES (425),Department of Zoology, University of British Columbia, Vancouver, British Columbia V6T 1W5, Canada C. C. LINDSEY(11, The University of Manitoba, Department of Zoology, Dufi Roblin Building, Winnipeg, Manitoba R3T 2N2, Canada J. MAGNUSON(239),Laboratory of Limnology, Department of Zoology, University of Wisconsin-Madison, Madison, Wisconsin 53706

JOHN

WILLIAMH. NEILL (315),Department of Wildlife and Fisheries Sciences, Texas A G M University, College Station, Texas 77843

DAVIDJ. RANDALL (425), Department of Zoology, University of British Columbia, Vancouver, British Columbia V6T 1W5, Canada E. DON STEVENS (315), Department of Zoology, University of Guelph, Guelph, Ontario N1 G 2W1, Canada PAULW. WEBB(189),The School of Natural Resources, The University of Michigan, Ann Arbor, Michigan 48104 * Present address: Department of Biology, Mount Allison University, Sackville, New Brunswick EOA 3C0, Canada. ix

This Page Intentionally Left Blank

PREFACE Locomotion in fish varies considerably, both quantitatively and qualitatively, between species and within species with age and environment. Only a few swimming modes in a small number of species have been examined in detail by a relatively small number of investigators. However, these studies, drawing on expertise from a variety of disciplines, have greatly increased our understanding of how fish move, and the flurry of activity is throwing some light on a previously confused and confusing field. This volume attempts to detai'1 ourpresent stage of understanding of fish locomotion. In order to clarify discussions, we have attempted to categorize some swimming activities in fish and hope that these categories will gain general acceptance. Neither the terms nor the groupings are new, but are drawn from already published work referred to in the text. The limitations of such a classification are that it tends to obscure variability and to present swimming activity as a series of discrete categories rather than a continuum. The contents of this volume reflect areas of study rather than a balanced report on all aspects of fish locomotion. For instance, the first chapter describes the variety of locomotor patterns in fish, and then swimming modes in trout and tuna are discussed in detail in two further chapters. This is not to imply that there is something very different about tuna and trout locomotion compared with that of other fishes, only that these modes have been studied in detail, whereas other modes of locomotion in fish have not. The neural control and generation of locomotor patterns are poorly understood in any vertebrate. I n fish, we are still at the stage of describing the types of nerve and muscle fibers present, how they are arranged and function to initiate and generate movements. There are only a few studies of the control of the respiratory and cardiovascular systems in fish during exercise and there are huge gaps in our understanding of these fields. The same can be said of metabolic changes associated with exercise. Each of these subjects is reviewed in an attempt to indicate areas of knowledge and regions of ignorance. A relatively small number of fish, in particular, tunas and lamnid sharks, are able to maintain muscle temperatures above ambient xi

xii

PREFACE

levels. These animals swim continuously utilizing the hot trunk muscle. This subject is reviewed as a special aspect of fish locomotion. Finally, many people advised and helped us in editing this text; in particular, the chapters were reviewed by many people other than ourselves. We are grateful for all help given. The result is a better text, more useful we hope, to those interested in fish locomotion. W. S. HOAR D. J. RANDALL

TERMINOLOGY TO DESCRIBE SWIMMING ACTIVITY IN FISH

Sustained Swimming: A spectrum of swimming activities and speeds that can be maintained for an indefinite period-in operational does not involve fatigue. terms for longer than 200 min-and Metabolism is aerobic and the activities would include foraging, station holding, schooling, cruising at preferred speeds in negatively buoyant fish, and steady swimming at low speeds, including migration. Burst Swimming: Rapid movements of short duration and high speed, maintained for less than 15 sec. Energy is made available largely through anaerobic processes. Burst activity may be subdivided into an acceleration period and a sprint, when swimming speed is high but steady. Prolonged Swimming: Covers a spectrum of speeds between burst and sustained and is often categorized by steady swimming with more vigorous efforts periodically. The swimming period lasts between 15 sec and 200 min and if maintained will end in fatigue. Energy is supplied from either or both aerobic and anaerobic processes. Fatigue: A fish is fatigued when it collapses and can no longer maintain a given swimming speed. Critical Swimming Speed, Ucrit:This is a useful operational term for comparing swimming speeds of different fish. Swimming speeds of individual fish of the same species but different swimming abilities, because of differing physiological states, can be compared b y expressing the swimming speed as %UCrit.In order to measure Ucritrfish are subjected to stepwise increases in swimming speed (usually in a water tunnel) until fatigue occurs. The critical swimming speed is computed from the maximum speed achieved.prior to fatigue. For instance, if a fish can maintain a swimming speed of 10 cm/sec for 60 min (the most usual duration for each step) but fatigues after 30 min at a swimming xiii

xiv

TERMINOLOGY TO DESCRIBE SWIMMING ACTIVITY IN FISH

speed 12 cm/sec, then the critical swimming speed is between 10 and 12 cm/sec and is computed as follows:

10

+ [(12 - 10) x

30/60] = 11 cm/sec

where 10 is the speed at the last completed step, (12 - 10) is the size of each step, and 30/60 is the time to fatigue at the last step divided b y the step duration. The duration and velocity increment of each step will influence the critical swimming speed obtained, and thus the step duration and velocity increment should be stated; for example, the 60 min, 5 cm/sec Ucritwas 20 cm/sec.

Steady uersus Unsteady Swimming: In many experiments using water tunnels or fish wheels, fish are trained to swim at a constant speed in one direction (steady swimming), whereas in their normal environment fish usually accelerate and change direction continually (unsteady swimming). Steady swimming is clearly more easily quantified and analyzed than unsteady swimming and so is used in experiments; however, caution must be used in applying conclusions based on the analysis of steady swimming to fish swimming in an unsteady manner. It is important to note, therefore, if the swimming is steady or unsteady. Thus burst, prolonged, and sustained swimming may contain either steady or unsteady components. The critical swimming speed can only b e measured for steady swimming. Fry (1971, pp. 3 and 4) has discussed the terms “standard,” routine,” and “active” levels of metabolism and the term “scope for activity” in his article “The Effect of Environmental Factors on the Physiology of Fish” in Fish Physiology, Volume VI. Standard metabolism is an approximation of the minimum metabolic rate for the intact organism. Routine metabolism, usually referring to aerobic metabolism and measured as oxygen uptake, refers to the metabolic rate when movements are restricted and the fish is protected from outside stimuli but is free to move and does so occasionally. Active metabolic rate refers to the oxygen uptake at the maximum sustained rate for a fish swimming steadily and is equivalent to oxygen uptake at the critical swimming speed. “

CONTENTS OF OTHER VOLUMES Volume I

The Body Compartments and the Distribution of Electrolytes W. N . Holmes and Edward M . Doraldson The Kidney Ckvelund P . Hickman, Jr., and Benjamin F . Trump Salt Secretion Frank P . Conte

The Effects of Salinity on the Eggs and Larvae of Teleosts F . G . T . Holliday Formation of Excretory Products Roy P . Forster and Leon Goldstein Intermediary Metabolism in Fishes P. W. Hochachka Nutrition, Digestion, and Energy Utilization Arthur M . Phillips, Jr. AUTHOR INDEX-SYSTEMATICINDEX-SUBj ~ c rINDEX Volume I1

The Pituitary Gland: Anatomy and Histophysiology J. N . Ball and Bridget I . Baker The Neurohypophysis A. M . Perks Prolactin (Fish Prolactin or Paralactin) and Growth Hormone J. N . Ball Thyroid Function and Its Control in Fishes Aubrey Gorbmun xv

CONTENTS OF OTHER VOLUMES

xvi

The Endocrine Pancreas August Epple The Adrenocortical Steroids, Adrenocorticotropin and the Corpuscles of Stannius I . Chester Jones, D . K. 0 . Chan, 1. W. Henderson, and J . N . Ball The Ultimobranchial Glands and Calcium Regulation D. Harold Copp Urophysis and Caudal Neurosecretory System Howard A. Bern AUTHOR INDEX-SYSTEMATICINDEX-SUB JECT INDEX Volume 111

Reproduction William S . Hoar Hormones and Reproductive Behavior in Fishes N. R. Liley Sex Differentiation Toki-o Y a m m o t o Development: Eggs and Larvae J. H . S . Blader Fish Cell and Tissue Culture Ken Wolf and hl. C . Quimby Chromatophores and Pigments Ryozo Fujii Bioluminescence J . A. C . Nicol Poisons and Venoms Findluy E . Russell AUTHOR INDEX-sYSTEM.4TIC

INDEX-SUBJECT

INDEX

Volume IV

Anatomy and Physiology of the Central Nervous System Jerald 1. Bernstein

CONTENTS OF OTHER VOLUMES

The Pineal Organ James Clarke Fenwick Autonomic Nervous Systems Graeme Campbell The Circulatory System D. 1. Randall Acid-Base Balance C. Albers Properties of Fish Hemoglobins Austen Riggs Gas Exchange in Fish D. J. Randull The Regulation of Breathing G . Shelton Air Breathing in Fishes Kiell Johallsen The Swim Bladder as a Hydrostatic Organ Johan B. Steen Hydrostatic Pressure Malcolm S. Gordon Immunology of Fish John E. Cushing

AUTHORINDEX-SYSTEMATIC INDEX-SUBJECT INDEX Volume V Vision: Visual Pigments F . W . Munz Vision: Electrophysiology of the Retina T . Tomita Vision: The Experimental Analysis of Visua1,Behavior David lngb Chemoreception Toshiaki J. Hara

xvii

xviii

CONTENTS OF OTHER VOLUMES

Temperature Receptors R . W. hlurray Sound Production and Detection William N . Tavolga The Labyrinth 0. Lowenstein The Lateral Organ Mechanoreceptors Ake Flock The Mauthner Cell J. Diamond Electric Organs h1. V. L. Bennett Electroreception M . V . L. Bennett

AUTHOR INDEX-SYSTEMATIC INDEX-SUB JEGT INDEX Volume VI The Effect of Environmental Factors on the Physiology of Fish F. E . J. Fry Biochemical Adaptation to the Environment P. W. Hochachka and G . N . Somero Freezing Resistance in Fishes Arthur L. DeVries Learning and Memory Henry Gleitman and Paul Rozin The Ethological Analysis of Fish Behavior Gerard P. Baerends Biological Rhythms Horst 0. Schwassmann Orientation and Fish Migration Arthur D . Hasler Special Techniques D . J . Randall and W. S. Hoar AUTHORINDEX-SYSTEMATIC INDEX-SUB JECT INDEX

FISH PHYSIOLOGY Volume VII Locomotion

This Page Intentionally Left Blank

1 FORM. FUNCTION. AND LOCOMOTORY HABITS IN FISH

.

C . C LZNDSEY I . Introduction ................................................... I1 . Early History of Studies on Fish Locomotion .................... 111. Modes of Swimming ...........................................

A . Nomenclature of Modes ..................................... B . Propulsion by Body and/or Caudal Fin ....................... C Propulsion by Undulation of Median or Pectoral Fins ......... D. Propulsion by Oscillation of Median or Pectoral Fins .........

.

IV. Nonswimming Locomotion ...................................... A . Jet Propulsion ............ ........ ............... B . Terrestrial Locomotion . . . . ............................. C . Moving on the Bottom and Burrowing ... D . Jumping, Gliding, and Flying ................................ V. Propulsive Anatomy ......... ................................ A . Trunk ......................................................

...........................................

Wild Fish ................................. A . Records of Long Distance Movements ........................ B . Short-Term Components of Long-Term Movements . . . . . . . . . . . C Activity Cycles in Wild Fish ................................. D . Schooling .................................................. E . Some Pitfalls in Locomotory Studies ......................... References .......... ............................................

.

1 5

8 8 11 26 37 39 39 40

43 46 46 62 75 75 80 83 85 87 88

.

I INTRODUCTION A fish moving through water is constrained by physical forces quite different from those affecting an animal moving on land or through the air . Some of the problems facing fish may be appreciated by examining the favorable and unfavorable features of water. in contrast to those of land or air. as a medium for locomotion . Most fish swim by pushing back against the water with undulations of their body or their fins . Water is unfavorable in that it presents a 1 FISH PHYSIOLOGY. VOL. V1I Copyright @ 1978 by Academic Press. Inc . All rights of reproduction in any fonn reserved . ISBN 0-12-350407-4

2

C . C. LINDSEY

yielding medium against which to push, and much energy may be wasted in’making profitless eddies. Water is favorable in that it offers little drag from friction, but drag of another sort, d u e to inertial forces (“pressure drag”), is high, because water is so dense. At the same time, the density of water makes it a very favorable medium in which to live because it buoys up the body of an organism. Fish do not require strong internal structures to carry their weight, in contrast to land animals which are severely limited b y their need for structural support. Also because of the buoyancy of water, work to keep from failing is minimal. Compared with terrestrial life, fish need expend little energy to move vertically. Because liquids are almost incompressible, pressure is not usually a problem to fish, except that it constrains the rapid vertical movements of any species carrying a chamber of compressible gas. In contrast to animals underwater, those inhabiting the air live in an insubstantial medium even more yielding than water. Hence most terrestrial animals perform their locomotion by pushing against the earth rather than against the air. The ground is a medium with almost no “give,” so that little energy is lost by imparting waste motion to the soil, and all goes into forward momentum of the animal. But, because frictional drag is very great, few animals can glide rapidly over (or within) the earth‘s surface. (Despite the give of water, fish can swim faster in water than snakes can crawl over the ground.) Most terrestrial animals have abandoned fishlike techniques of propulsion; instead they “walk,” taking advantage of the frictional forces and resistance of the ground to provide thrust for the limbs, and of the negligible drag of the air to allow forward progression of the body and the individual limbs. Walking is an inappropriate technique for attaining speed underwater, as is apparent to a human who tries to run while partly immersed. Many aquatic invertebrates which are heavier than water do walk slowly on the bottom, but to attain speed they must leave the bottom and swim. Few species of fish walk on the bottom underwater with their paired limbs. On the other hand, those fish which make excursions out onto land usually have to resort to walking or skipping in some fashion, since the air provides so insubstantial a medium against which their usual body undulations can act. Proof of this is the observation that a live fish on a slippery deck flaps futilely without achieving forward progression, even though the same muscular undulations performed underwater would have propelled it swiftly forward. Eels can progress overland through grass or on rough ground, but here the body

1. FORM, FUNCTION,

AND LOCOMOTORY HABITS

3

undulations are thrusting against the ground, not the air; the eel is not swimming through the air, but is crawling like a snake. Just as the fastest aquatic animals abandon the bottom and swim through the water, so the fastest terrestrial animals leave the ground and fly through the air. Roughly half the living species of animals can fly. Most are small animals (insects),in which the surface area for lift is great relative to the weight. For them the viscosity of the medium is important relative to its inertia (i-e., they operate at low Reynolds numbers, as explained in Chapter 3). Their physical constraints are therefore in some ways comparable to those of larval fish and other small animals underwater. One feature of locomotion under such conditions is that stopping and starting are no problem; when viscosity is dominant a flying insect or swimming fish larva stops as soon as it ceases to propel itself, and begins to move at full speed with negligible time-lag as soon as the propulsive movements begin (Lighthill, 1969). The larger flying species (birds and bats) on the other hand, operate with speeds and dimensions where the viscous forces are less significant; to them, as to larger fish, the inertial forces of the medium are dominant (although the absolute values are very different in water and in air). To these animals, extra energy is required to accelerate the body from rest; conversely a moving bird or large fish, if well streamlined, can glide a long way after it ceases propulsive movements. Gravity, which is of minor concern underwater, is a major factor in air. To birds supporting themselves in the air, a severe constraint is weight. Extreme structural economies (such as hollow bones and feathers) are needed to minimize the effects of gravity in a large flying animal. Underwater, some of the midwater fishes which lack gas chambers display comparable weight economies (such as reduced skeleton, and substitution of light fatty tissues wherever possible). Fish can thereby achieve neutral buoyancy in water. But in air the density of the medium is so slight that even the most lightly built bird still weighs much more than the air it displaces. No animal has attained neutral buoyancy in air, although this should be theoretically possible through development of a large bladder of hydrogen or other light gas. In water, on the other hand, “weightlessness” is relatively easy to attain, by inclusion of a small gas chamber which displaces an equivalent volume (and 800 times the weight) of water. The majority of bony fish (but not sharks) have such a gas chamber; to them, structural weight economies are not necessary. Animals flying through air are like winged aircraft; they must divert part of their locomotory effort into lift to overcome gravity. Neutrally

4

C . C . LINDSEY

buoyant fish in water are like dirigibles; they can concentrate all their effort into forward thrust. Only in sharks, rays, tunas, and other fish which may lack a gas chamber does lift become important, and so their locomotion through water has some features in common with the flight of birds or aircraft. Most of the few fish which are capable of brief aerial locomotion progress through the air by passive gliding rather than by flying (and swim underwater by conventional body undulations). They are comparable to those restricted groups of mammals, amphibians, and reptiles which can glide but which are not primarily adapted to this mode of locomotion. Only very few fish (includingthe Gasteropelecidae, and Pantodon) can fly in air by beating their wings, and this they d o ineptly. The formidable problems of simultaneous adaptations for locomotion both in midair and underwater have been overcome, among the vertebrates, only by diving birds such as loons. Among the invertebrates a few types of insects (including water beetles and some bugs) can swim underwater and fly in air in the same life history stage. Most which can move effectively in both media do so at two radically different life history stages (e.g., dragonfly nymph and adult). Another set of locomotory constraints arising from properties of water are those factors determining the oxygen available to the propulsive system. Water contains only about one-thirtieth as much oxygen as does an equal volume of air. Water is also much heavier and more viscous to move. To meet these problems, fish have a flow-through respiratory system which extracts a high proportion of the available oxygen. This is achieved without undue expansion of the respiratory surface, which in tunas is about the same as that in terrestrial vertebrates of equivalent weight, and in other fish is smaller. A very large gill area would present too great a drag, apart from allowing excessive ion exchange (a nonexistent problem in terrestrial lungs). But even with its highly efficient countercurrent design and moderate gill area, the energy expenditure for gill ventilation in an active fish may amount to about one-tenth of its total metabolic output. In contrast, an active man probably uses no more than 3% of his total oxygen consumption for breathing. Some fish species move the water over the gills with their branchial muscles. Others simply swim with their mouths open, in which case their drag is increased and the energy for gill ventilation must come from additional work by the propulsive body muscles. The proportion of the total energy output used for ventilation is probably less in small than in large fish. This is just one of many ways in which

1. FORM, FUNCTION, AND LOCOMOTORY HABITS

5

physical constraints of the environment operate differently on small and on large fish. An elegant survey of aquatic animal propulsion (including many major groups omitted above) is given by Lighthill (1969),whose view is that of an aerodynamicist. Other insights into the locomotory problems in different media are to be found throughout the works of Alexander (1967, 1968, 1971), Gray (1968 and earlier), Schmidt-Nielsen ( 1972a,b), and Tucker (1975).

11. EARLY HISTORY OF STUDIES ON FISH LOCOMOTION Few of the features of fish locomotion are evident to the naked eye. The usual method of locomotion in fish is now thought to depend on passing alternating waves of contraction backward along the body muscles. Thrust against the water either is generated by the sides of the body pushing obliquely backward (as in a swimming eel), or else has become progressively more concentrated in the tail fin (as in fast swimmers such as the tuna). The paired fins seem to contribute little in conventional forward swimming, and are reserved for maneuvering. Not only are these points difficult to discern by direct observation; they also do not follow by logical extension from the locomotory habits of terrestrial animals. Consequently, the historical development of opinions on fish locomotion has been marked by contradictions and controversies. In the fourth century B.C., Aristotle referred to fish locomotion in various passages in his three works “Parts of Animals,” “Movement of Animals,” and “Progression of Animals.” H e wrote that fish with very long bodies and no paired fins, such as the moray, move along by an undulating motion of the body; “that is, they use the water just as serpents use the ground.” He made the acute observation that eels move in the same way in water and on land but with fewer bends in the former medium (a hydrodynamic explanation for which might only now be attempted). Aristotle anticipated Newton by 20 centuries in his statement that there must always be something “immovable” outside an animal “supported upon which that which moves moves. F i r if that which supports the animal is to be always giving way . . . there will be no progress, that is, no walking unless the ground were to remain still, and flying or swimming unless the air or sea were to offer resistance.” Aristotle fell short of an understanding of undulatory locomotion in that he did not perceive that the crests of the body waves pass backward, and he tried to draw an analogy between a

6

C . C. LINDSEY

snake’s motion and the walking of quadrupeds. He correctly described a ray as swimming by means of the edges of its flattened disc. He was apparently wrong with respect to the locomotion of most fishes, as he thought that the two pairs of fins were their principal means of propulsion. He suggested that the caudal fin was primarily for steering. The ancient Hindu medical work Susruta-samhita probably reached its final form in the early years A.D., but it may contain components which predate Aristotle. According to the free translation by Hora (1935), it suggests a correlation between body form, habitat, and locomotion in some freshwater fishes; river fish are said to be bulky in the middle because they move with their head and tail; pool fish, having little space to move about, are deep-bodied; torrent fish are flattened because they crawl with their chests on the bottom. The first attempts at mechanical analysis of fish locomotion were by Borelli, a disciple of Galileo, who in 1680 published a diagram of a fish swimming b y sweeping its caudal fin and peduncle side-to-side in an arc (reproduced in Gray, 1968). Borelli thought of the tail fin as operating like an oar sculled behind a boat. He dispelled the notion of Aristotle that the paired fins when present are the main locomotory organs, stating they are held at the sides during swimming; Borelli pointed out also that the muscles of the body, which provide power for the tail strokes, are large in proportion to those of the paired fins. He outlined the role of the gas bladder in controlling the specific gravity and hence the position of the fish in the water. He departed from observation, however, when he described the tail as operating like a frog’s foot, contracted during a “preparatory” outward sweep and expanded during a powerful inward propulsive sweep. Two fundamental points he made, which display the influence of Galileo, were: (1) the fish cannot move its tail without also moving its body, and (2) the force moving the fish forward is due to resistance of the water against the surface of the moving tail. Pettigrew ( 1874) challenged Borelli’s view that forward motion could result simply by lashing the tail from side to side. He argued that moving the tail from the midline outward during the “preparatory” sweep would produce a backward movement of the whole. Breder (1926) comments “His logic would seem to be correct, but that he was in error has been positively demonstrated by the construction of a model.” (Breder built a model boat, which did swim forward simply by means of an oscillating rigid tail vane.) What Pettigrew overlooked was that the outward sweep of the tail, on a fish or on a boat, does not move the whole backward; instead, it swings the main body (ahead of the joint) slightly in a sidewise arc opposite to that of the tail. The subsequent inward stroke moves the whole forward as

1. FORM, FUNCTION,

AND LOCOMOTORY HABITS

7

well as swinging the body back into line. Pettigrew invoked complex rolling motions of the fish to overcome his supposed difficulty. Breder ( 1926) attacked Pettigrew enthusiastically for these and other “absurdities,” and complained that uncritical paraphrasing of Pettigrew’s views by Bridge (1904) in the Cambridge Natural History has given them undeserved circulation. However, Pettigrew did observe that fish throw their bodies into a double or sigmoid curve rather than into the simple arc described by Borelli (although he believed this must occur in all fish, which it does not). Pettigrew correctly observed that the tail tip of a slowly moving sturgeon described a figure-8 when viewed from above. He implied that this was an adaptation for efficient propulsion; Gray (1933a) showed it to be the inevitable result of the propagation of a wave of curvature along any inextensible body. The possibility of precise observation on locomotion dates from Marey’s (1895) use of cinematography. He produced sequential photographs of swimming fish, a technique since widely emulated. His pictures demonstrated that waves of curvature pass along the length of the body. Unfortunately, his Fig. 199 of a swimming conger eel shows one in which the body undulations are either stationary or are slowly movingforward relative to the water, so that the fish must actually have been slowing down when it was photographed. Since Marey did not attempt to analyze the forces involved in swimming, he missed the fact that this was an unhappily chosen figure to illustrate swimming. Although Gray (1933b)drew attention in a footnote to the aberration, Marey’s figure has been widely copied. It still appears in the third edition of Norman and Greenwood’s (1975) otherwise admirable text. [Nikolsky’s (1963) text does not make this mistake, but it does use as an illustration of locomotion (his Fig. 32) a sequence from Gray (1933a) which actually shows an eel swimming backward.] Notwithstanding, Marey provided a strong incentive for studies on fish locomotion, and inspired later experimental studies in France by Houssay (1912), who tried to measure the thrust and drag of fish, and by Magnan (1929, 1930). Dean (1895) reproduced one of Marey’s illustrations showing the wave form in a swimming eel. Dean explicitly stated “It is the pressure of the fish’s body against the water enclosed in these incurved places which causes the forward movement.” In an outstanding synthesis called “The Locomotion of Fishes,” Breder (1926)gave an extensive treatment of the mechanical principles, as well as a systematic description of locomotion in different fish groups. Breder concluded that “All the movements of fishes when swimming (except exhalation) are fundamentally of an undulatory muscular nature even though obscured by various specializations, and are induced

8

C. C. LINDSEY

by the serial action of metameral muscles.” H e categorized the types of movement of the body and of the fins, and coined many of the terms (e.g., Anguilliform, Carangiform) now in use. Th dominant figure in marrying precise measurement of moving animals with mathematical analysis has been Sir James Gray. The wide-ranging curiosity of Gray and his colleagues, notably H. W. Lissmann, has been brought to bear dn the locomotion of large vertebrates, and of sperm, and of most groups in between. In a series of papers starting in 1933 Gray analyzed photographs to show how undulatory swimming movements generate thrust. Gray (1936b) compared the calculated drag of a swimming dolphin with the calculated power output of its muscles, and concluded that dolphins (and some fish) are observed to swim at speeds which according to theory are impossibly fast. This famous “Gray’s Paradox” has stimulated much research and controversy ever since. In trying to balance the equation, some have sought improved measures of the muscular power output. Some have recorded much more precise observations on swimming fish [notably Bainbridge (1958 and later) using an ingenious “fish wheel”]. Some have looked for drag-reducing mechanisms. Most important, some have questioned the assumption that drag of a swimming fish could be equated to that of a rigid body or model. Some tried to measure the drag on actual fish, but their measurements seemed only to add to the growing confusion. Sir Geoffrey Taylor (1952), and Sir James Lighthill (1960), proposed mathematical models which might allow calculation of the drag of a swimming fish. Newer “bulk momentum” models concentrate attention on the kinematics of the trailing edge of the tail throughout one propulsive cycle, from which they attempt to calculate the thrust and power which must have been generated. Their significance is described in Webb’s (1975) publication, which is itself an important bridge across the communication gap between biologists and physical scientists. Recently DuBois et al. (1974) implanted pressure sensors at various points on the surface of live and of dead fish in a water tunnel. Webb (1975; see also Chapter 3) brings up to date the developments in fish hydrodynamics and energetics. One might summarize that the gap between the swimming fish and the scientists is closing, but the fish is still well ahead.

111. MODES OF SWIMMING A. Nomenclature of Modes

The different types, or modes, of propulsive movements of fish were classified by Breder (1926), whose nomenclature, somewhat ex-

1. FORM, FUNCTION, AND LOCOMOTORY HABITS

9

panded,, is followed here. Examples of fish displaying these modes are shown in Fig. 1. As Breder stated, the suffix “-form” (e.g., in anguilliform) refers to the types of movement and not to the body forms, and is therefore not strictly parallel to words such as “fusiform.” Indeed,

Fig. 1. Modes of forward swimming in fish, arranged along the vertical axis according to the propulsive contributions of body and fins (indicated by density of shading), and along the horizontal axis according to a scale running from serpentine undulation (more than one wavelength present) to oscillation (a rigid wigwag or fanlike motion). Species illustrated are: (A) Anguilla anguilla, (B) Squalus acanthias, (C) Cadus morhua, (D) Salmo gairdneri, (E) Caranx hippos, (F) Clupea harengus, ( G ) Zsurus glaucus, (H) Thunnus albacares, (I) Ostracion tuberculatum, (J)Amia calua, (K) Gymnotus carapo, (L) Balistes capriscus, (M) Lagocephalus laeuigatus, (N) Raja undulata, (0)Diodon holocanthus, and (P) Cymatogaster aggregata.

10

C. C. LINDSEY

one fish may show more than one mode of swimming, as in the surfperch Cymatogaster aggregata, which usually swims with its pectoral fins (labriform mode) but switches to caudal fin locomotion (carangiform mode) at high speeds (Webb, 197313). Breder (1926), Bainbridge (1963), and Webb (1975)have stressed that these classifications refer to average types within an essentially continuous range of swimming modes, and should not be applied too rigorously. Use of Breder's nomenclature in the following discussion and in Fig. 1 is simply for convenience. The arrangement of swimming modes implies no evolutionary or taxonomic affinities; clearly there has often been functional convergence on one swimming mode by taxonomically remote groups (e.g., locomotion is anguilliform both in lampreys and in blennies). The rationale for a classification according to propulsive mode is that similar hydrodynamic analysis may be applicable to animals which swim in the same way, regardless of diverse phyletic origins. In a swimming eel, most of the body is bent into backward-moving waves, whose amplitude is quite wide over the whole body length (Fig. 2A). In progressively shorter, thicker fish, propulsive waves tend to be increasingly concentrated in the tail region, so that only about half a wavelength is visible, and its amplitude rises rapidly in the

\'

\\

'!\\\

Fig. 2. Gradation of swimming modes from (A) anguilliform, through (B) subcarangiform, and (C) carangiform, to (D) thunniform. The black silhouette (dorsal view) is superimposed on successive positions one-half tail beat earlier (broken outline) and one-half tail beat later (stippled). (A) Anguilla anguilla, 7 em long, about 1.5 beatskec. (B) Gadus rnerlangus, 24 cm long, about 1.7 beatdsec. (C) Scornber scombrus, 40 ern long, about 2.4 beatdsec. (D)Euthynnus ajinis, length unknown, perhaps about 40 em, about 2.4 beats/sec. A, B, and C based on Gray (1933a, 1968); D based on Fierstine and Walters (1968).

1.

FORM, FUNCTION, AND LOCOMOTORY HABITS

11

region of the tail base (Fig. 2B, C). This propulsive mode reaches its climax in the very swift mackerels and tunas that appear to the naked eye to swim by moving only the caudal fin (Fig. 2D). A few species like the boxfish Ostrucion have totally inflexible bodies which cannot be thrown into a wave; they swim by oscillating the caudal fin back and forth like a fan or pendulum. This gradation from multi-waved undulation, through progressive and eventually exclusive concentration of propulsive movement in a pendulumlike oscillation, can be seen not only in the body but also in the median and paired fins. The nomenclature “anguilliformcarangiform-ostraciiform” was used by Breder (1926) as a scale to refer to flexures in the dorsal, anal, and pectoral fins as well as in the whole body, although he coined additional names for the swimming modes which involve the fins. A comparable “undulation-oscillation” scale is used in Fig. 1. The modes of swimming dealt with in this section refer to straightforward horizontal motion in still water while free from contact with any solid. They therefore omit accelerating, turning, rising, stopping, and other maneuvers which may be vitally important components of locomotion. Nonswimming movements such as burrowing, creeping, jumping, and flying are dealt with in Section IV, as is jet propulsion. A somewhat different analysis of locomotion types was presented by Kramer (1960).He grouped fish into ten categories using combinations of the swimming characteristics: use of trunk versus median or paired fin muscles; straight versus curving path; adaptation for high sustained speed versus rapid acceleration; inclination of fin axes on the body. Fish were also categorized according to ecological types: fast swimmers, roamers, swimmers between obstacles, slow and precise maneuverers, and bottom resters. Kramer proposed no nomenclature for his swimming modes.

B. Propulsion by Body and/or Caudal Fin

1. CLASSIFICATION

The classification of modes of propulsion by the body or tail has since been somewhat expanded from that designed by Breder (1926), but nomenclature has not been uniform. Table I shows the apparent equivalence between terms used by different authors. The two extremes of the spectrum “anguilliform” and “ostraciiform” have been used consistently. Breder’s intermediate term “carangiform” covers a

C. C. LINDSEY

12 Table I

Classification by Various Authors of Methods of Fish Propulsion Involving the Body and/or Caudal Fin Breder (1926)

Bainbridge (1961)

Marshall (1971)”

Anguilliform Carangiform

Anguilliform More anguilliform carangiform Carangiform More ostraciiform carangiform Ostraciiform

Anguilliform Subanguilliform

Anguilliform Subcarangiform

Anguilliform Subcarangiform

Fusiform Thunniform

Carangiform Carangiform with lunate tail Ostraciform

Carangiform Thunnifom

Carangiform Carangiform Ostraciiform

Webb (1975)

Present

Ostraciiform

Marshall did not propose a formal classification, but used these terms in his text.

wide range of swimming patterns, which are now suspected to require more than one hydrodynamic model (Lighthill, 1969). Hence three “ terms, subcarangiform,” “carangiform,” and “thunniform,” will be used, the latter as a more convenient substitute for “carangiform with large lunate tail” (Lighthill, 1969) or “carangiform mode with semilunate taiI” (Webb, 1975). Some characteristics of each of these propulsive modes are given in Table 11, and examples of taxonomic groups in which some (but not necessarily all) members use that mode.

2. ANGUILLIFORMMODE Anguilliform is a purely undulatory mode of swimming, in which most or all of the length of the body participates. The side-to-side amplitude of the wave is relatively large along the whole body, and it increases toward the tail. The body is long and thin; in eels it may be nearly cylindrical anteriorly, and somewhat laterally compressed toward the posterior. The caudal fin is often small, or even absent. Figure 2A shows three successive positions of a swimming eel Anguilla, after which the mode is named. It must not be supposed from these outlines that there are fixed pivots of nodes around which the body sections oscillate. The manner in which the waves of contraction move smoothly backward is better illustrated in the successive outlines of a swimming herring larva in Fig. 3. Each wave is generated by contractions of the body muscles in a few anterior segments on one side of the vertebral column, while those on the opposite side are relaxed and are slightly stretched. The resultant bending of the body toward the contracted side passes backward as the wave of muscle

Table I1 Comparison of Swimming Modes Involving the Body and/or the Caudal Fin Anguilliform mode

Subcarangiform mode

Wave length Body length

Short, always <1.0

Wave lengths visible on body Amplitude Body length

Always >0.5, usually > 1.0 Large along whole body. Anguilla max. 0.36, Clupea larva max. 0.46 Long thin. Anterior cylindrical, posterior compressed

Uusually 0.5, usually not more than 1.0 Undulations wide only in posterior yi or % of body, max. about 0.2 Fusiform. Peduncle fairly deep

Character

Body shape

Carangiform mode

Thunniform mode

Ostraciiform mode

Usually >LO, but 0.93 in Scomber scombrus

1-2

Pendulum motion

Up to 0.5

0.5-1.0

Pendulum motion

Undulations confined to posterior %, max. about 0.3 Mass concentrated anteriorly. Peduncle quite narrow

Undulations confined to peduncle and tail, max. >0.3c Massive rounded anterior. Surface streamlined. Extreme narrownecking of peduncle, with lateral keels

Tail pivots on caudal peduncle Torpedo nobiliana max. about 0.25d Expanded or depressed, lateral1y inflexible, often armored, poorly streamlined

(Continued)

Table II-Continued Character Span of body and median fins

Anguilliform mode

Subcarangiform mode

Taper extreme: Chimaeridae, Saccophaqmgoidei, Notacanthiformes, Regalecidae, Cepolidae, Trichiuridae Taper moderate: Petroymzontidae, Anguillidae, Chlamydoselache Span expanded posteriorly by tail:

2-3 dorsal, 1-3 anal fins, gaps filled by vortex sheets (Gadiformes). Or dorsal or anal may be long to reduce yawing (Cyprinidae). Or short dorsal followed by adipose (Salmonidae)

SqUdUS,

Siluriformes Expanded dorsal and anal fins: Osteoglossidae, Trachypteridae, Pleuronectiformes

Carangiform mode May have stiff median fins resisting yawing: Carangidae

Thunniform mode High first dorsal fin fixed (Lamnu, Zsurus, Carcharodon) or

collapsible (Thunnus, Euth ynnus, Acanthocybium). Finlets (5-11), on peduncle (Scombridae), or small second dorsal and anal fins (fast sharks)

Ostraciiform mode Span variable; median fins often small, not placed to reduce yawing

Caudal fin

Aspect-ratio low (Siluriformes) or moderate (Squalus).Often small, rounded or absent. Span alterable

Rather low aspectratio. Posterior margin almost straight or slightly scooped. Span alterable','

Other examples (not necessarily including all species in group indicated)

Myxinidae, Blennioidei, Ophidioidei, Ammodytidae, Synbranchiformes, Trichiuridae,' Pieuronectiformes (which undulate on sides). Young of most fish even if adult mode differs

Triakis, Esocoidei, Poeciliidae, Mugil

Webb (197%). (1971).

* Gray (19334.

Rather high aspectratio. Posterior margin scooped or notched. Span alterable moderately (Clupea)or not at all (Caranx, Scomber). Stiff tips lead during beats, center follows Clupeidae Characidae, Mormyridae, Pomatomus, Sarda, Sardinops

Fierstine and Walters (1968).

Roberts (19694.

Very high aspectratio. Lunate margin. Span almost fixed. Center leads during beats, tips follow

Low aspect-ratio. Often rounded or square. Quite rigid, pivots on peduncle

Some Scombroidei, some Lamnoidei. whales and dolphins, ichthyosaurs. (Thunniform mode questionable in Rhineodon, Istiophoridae, Xiphiidae, Luvaridae, Stromateidae, Bramidae, Coryphaenidae)

Loricariidae, Ostraciidae, Tetraodon tidae, Diodontidae, Lophiiformes, Lophotidae; Trichiuridae'

Bainbridae (1963). 'Webb (1975).

g

Bone

16

C. C. LINDSEY

Fig. 3. Anguilliform swimming by herring larva Clupea harengus, about 6.5 mm long, with yolk sac. Successive cinephotos (at 0.021 sec intervals) are displaced to the right. Lower broken line represents a fixed position on background. Movements of snout and tail tip indicated by dots. Position of wave crests shown by crosses and circles. (Based on Rosenthal, 1968.)

contractions moves toward the posterior. Meanwhile the anterior muscles on the side which had been contracted relax, their partners on the opposite side contract, and a bend in the reverse direction is initiated, and passed backward in turn. In Fig. 3 the crests of the propulsive waves can be seen to move backward with respect to the fixed background, producing a thrust which drives the whole fish forward. Quantification of the thrust generated by this type of swimming is discussed in Chapter 3. Anguilliform swimming is widespread among fishes. The young of most forms probably swim in the anguilliform mode even though their adult locomotion (and body form) may be very different. [In herring Clupea harengus the swimming of the larva is anguilliform (Fig. 3) but in the adult is carangiform (Webb, 1975; see.also Fig. lF).] Several phylogenetically remote groups of fish with long flexible bodies swim as adults in this way. Performance in the anguilliform mode is evidently improved by a compressed cross section; this may be achieved by laterally compressing the body itself, or by augmenting the thrust with expanded dorsal or anal fins to produce a wide “span” (Fig. 1A). Highly flexible ribbon-shaped bodies with wide span are found in many species among the groups named in Table 11. High speed is not characteristic of the pure anguilliform mode. Many anguilliform swimmers live close to the bottom. Flatfish (Pleuronectiformes) usually swim in the anguilliform mode; they swim, however, on their side, so that the undulations are vertical rather than horizontal. Avery wide body span in flatfish is achieved both by a high compressed body and by elongate dorsal and anal fins. Some flatfish augment the propulsive thrust of the body by also passing undulations

1.

FORM, FUNCTION, AND LOCOMOTORY HABITS

17

of wider amplitude along the dorsal and anal fin rays (Aleev, 1963).

The Greenland halibut Reinhardtius hippoglossoides, which is less flattened than other flatfish and has less complete asymmetry, swims like other flatfish when it is close to the bottom, but in midwater it adopts a vertical position. “It is still an open question whether the Greenland halibut is a round flatfish or a flat roundfish” (deGroot, 1970). A few other fish swim pelagically in the anguilliform mode, although perhaps not very fast; these include some cusk-eelsOphidiidae, and presumably the eel Anguilla during its ocean migrations. Anguilliform swimming is evidently least efficient (in the sense described in Chapter 3) when associated with bodies whose span tapers toward the posterior, Table I1 lists some groups which taper markedly, and some with an almost constant span along the whole body.(partly by virtue of the median fins). Toward the other extreme, in many catfishes, Siluriformes, and more particularly in many of the smaller sharks, the body moves in a distinctly anguilliform mode bbt the span is expanded posteriorly by a moderate sized although flexible caudal fin. The shape of the caudal may vary considerably as measured by its “aspect-ratio.” [Aspect-ratio is the square of the maximum fin height (or span) divided by the fin area.] The aspect-ratio of the tail may be quite low in catfish, but is higher in some sharks. The dogfish shark Squalus in Fig. 1B has the body undulations more pronounced toward the posterior than has the eel, a n d i t has a heterocercal tail with moderate aspect-ratio. Heie, the anguilliform mode merges imperceptibly into the subcarangiform mode.

3. SUBCARANGIFORM MODE Body movements in subcarangiform swimming (e.g., trout) are essentially similar to those in anguilliform, the major difference being that the side-to-side amplitude of the undulations is slight at the anterior, and expands significantly only in the posterior half or one-third of the body. The tip of the snout does not move in the straight mean path of progression, but oscillates about this with a moderate amplitude (Fig. 2B). There is no fixed node; no point on the fish’s body seems to move forward on, or parallel to, the mean path of progression. The whole body in fact generally executes a sinuous path through the water (Bainbridge, 1963). The amplitude of subcarangiform undulations does not change with swimming speed except at speeds lower than 1 or 2 body lengths

18

C. C . LINDSEY

(t)/sec.Amplitude is constant at moderate and high steady swimming speeds even in fish of different lengths. At low speeds, below 1 or 2 Clsec, Bainbridge ( 1958) believed he detected a correlation between amplitude and tail beat frequency, but Hunter and Zweifel (1971) regarded his data as inconclusive on this point. The speed is altered by varying the velocity at which the waves are passed backward, and hence also the tail beat frequency. The maximum frequency attainable decreases with increasing size of the fish. Wavelength relative to body length is generally assumed to remain constant within a species, but there is some reason to suspect that longer fish may use slightly shorter specific wavelengths. The body tends to be heavier and more rounded anteriorly compared to anguilliform swimmers. The caudal peduncle is fairly deep; the caudal fin which it bears has a rather low aspect-ratio, with its posterior margin almost straight or only moderately indented (“scooped out”) in the center (Fig. l C , D). The caudal fin in a subcarangiform swimmer tends to be flexible, and is provided with intrinsic muscles (Fig. 7D) which can slightly open or shut the “fan” of caudal fin rays so as to alter the fin area by as much as 10% at different phases of one beat (Bainbridge, 1963). In rainbow trout, the depth of the whole tail is also increased at higher swimming speeds (Webb, 1975). The movements of the caudal fin during swimming are obviously very complex; they probably involve rapid adjustments which control the thrust, but they are not understood. Surprisingly, amputation of the caudal fin has little effect on straight forward swimming in the subcarangiform mode; the welldeveloped caudal in these fish has probably evolved primarily in response to requirements for high acceleration, fast turnicg, and highspeed maneuverability (Webb, 1973a). Cods and their relatives (Gadiformes) have two or three dorsal and one or two anal fins which are separated from each other and from the caudal fin b y narrow gaps (Fig. 1C). When the fish is moving forward, these gaps become filled by “vortex sheets” which behave hydrodynamically almost like a solid surface (see Chapter 3). Such fish therefore have functionally a wide span, and their subcarangiform swimming may approach the anguilliform. In other fish which use the subcarangiform mode the median fins are less continuous; the dorsal fin may be short and followed by a small adipose fin (Fig. lD), or the dorsal or the anal fins may be relatively long and serve to reduce the side-to-side yawing of the body in response to the lateral tail beats. The most complete descriptions of subcarangiform swimming are by Bainbridge (1958, 1963), who studied rainbow trout Salmo

1. FORM, FUNCTION, AND LOCOMOTORY

HABITS

19

gairdneri, bream Abramis brama, dace Leuciscus leuciscus, and goldfish Carassius auratus, and by Webb (1973a), who studied sockeye salmon Oncorhynchus nerka.

4. CARANGIFORM MODE In carangiform swimmers only the posterior portion of the body is capable of wide flexure. Undulations are largely confined to the last third of the length, and the thrust is delivered by the rather stiff tail (Fig. 2C). Carangiform swimming is faster and probably more efficient than anguilliform swimming, since the same thrust can be delivered, provided the amplitude has risen to a large value in the region immediately ahead of the trailing edge, with less energy lost in displacing water laterally and forming vortices (Lighthill, 1969). Several morphological adaptations are necessary for efficient carangiform swimming. Because there is never a complete wavelength on the body, and because the lateral flexures are concentrated at the posterior, the side forces produced by the flexures do not cancel out their net effect as they do in the anguilliform mode. There is thus a tendency for the body to recoil in response to the tail movement by sideslipping and yawing (Fig. 2C). Recoil movements, which waste energy, are minimized in two ways. Wave amplitude increases very rapidly as it approaches the caudal peduncle. I n order to avoid generating a large sideways thrust in this region, the depth of the peduncle is greatly reduced (Fig. 1E). This local reduction of the span is called narrow necking. Narrow necking occurs to a moderate degree in some of the subcarangiform swimmers, is well developed in the carangiform, and reaches its extreme in the thunniform (Fig. 1H). Lateral oscillations grow in amplitude mainly in the region of this slim peduncle, and reach a large approximately uniform value over the caudal fin. A second adaptation which reduces recoil in carangiform fish is the concentration of mass and of body depth toward the anterior (Fig. l E , F). Stiff median fins may further increase the overall span and help resist sideways movement of the body. The caudal fin, which delivers most of the thrust, is stiff so as to allow little dorsoventral bending. The upper and lower edge are “swept back,” and the center of the fin is scooped out to various degrees, often as a V-shaped notch (Fig. l E , F). As the fin moves, the scooped out center area fills with a vortex sheet which works as effectively as the rest of the tail (Lighthill, 1969). The high span and reduction of area due to scooping of the center produce a high aspect-ratio

20

C. C. LINDSEY

compared with subcarangiform swimmers. The span is under less control by intrinsic muscles than in subcarangiform swimmers; the fan of caudal rays can be expanded only moderately (Clupea)or scarcely at all (Caranx, Scomber). In the carangiform mode, the caudal fin is not simply wagged back and forth like a stiff blade on a hinge (the ostraciiform mode). Instead, the angle of inclination is altered as it moves from side to side so that the fin always has a backward-facing component even when moving away from the midline (Lighthill, 1969).The details of how the stiff fin is manipulated in this way by the muscles of the trunk and peduncle have not been measured. Carangiform swimmers in the family Clupeidae (Clupea, Sardinops) have only moderate narrow necking of the caudal peduncle (Fig. lF), and so have some Characins. The more extreme carangiform swimmers have pronounced narrow necking, and sharply swept back tails scooped out with a deep V (Fig. 1E). These include the bluefish Pomatomus, many species in Carangidae including Caranx, and some species in Scombridae (Scomber japonicous, S . scombrus). Also within Scombridae are species, with lunate tails, verging on the thunniform (Sarda chiliensis), as well as the tunas which exemplify the thunniform mode. The unique Mormyridae, freshwater African fishes with electric sensory organs, possess paired longitudinal bones below the caudal electric organs which provide rigidity to that region of the body (Lissmann, 1961a). The tendons which operate the tail fin run over this region and have their origin on more anteriorly placed myotomes. Lissmann writes that their swimming movements appear to be “of the normal carangiform type.” The system of tendons running pulleylike past a rigid secton are suggestive of the thunniform mode, but the whole physiognomy of mormyrids, and the speed attained, is quite unlike the tunas. The mormyrid system has probably evolved to provide a stable base for the electric sensory system.

5. THUNNIFORM MODE I n thunniform swimmers the thrust is generated exclusively by a high stiff caudal fin mounted on an extremely narrow peduncle (Fig. 1H). Significant lateral movement occurs only in the peduncle and tail fin (Figs. 2D and 4).The propulsive force is delivered from the massive body muscles to the caudal fin by a system of tendons which run like pulleys past two joints in the posterior of the vertebral column. The wavelength is long, and its amplitude is wide at the trailing edge (Fig. 4). The body is heavy toward the anterior, sometimes almost

1. FORM, FUNCTION, AND LOCOMOTORY HABITS

21

i

1

Fig. 4. Thunniform swimming by kawakawa Euthynnus affinis. Length unknown, perhaps about 40 cm. Cinephoto intervals 0.06 sec. Symbols as in Fig. 3. (Based on Fierstine and Walters, 1968.)

circular in cross section, and is beautifully streamlined. The massive anterior, the extreme narrow necking at the peduncle, and the high aspect-ratio of the tail, all combine to minimize sideways recoil despite the power of the caudal thrusts. In tunas, the swiftest of all fish, “the quick and powerful strokes of [the tail fin] can be understood from the quick and high-pitched sound produced by the fish in its death-struggle on the deck of a boat” (Kishinouye, 1923). The vertebrae of the caudal peduncle form a rigid unit, strengthened by lateral keels which make the peduncle wider than it is deep. Immediately in front is the prepeduncular joint which allows the peduncle to swing in a wide arc. Immediately behind is the postpeduncular joint, with the last three vertebrae shortened and providing a hinge on which the stiff caudal fin can swing. This double-jointed system is operated by several sets of tendons. A deep-seated row of tendons lies on either side of the backbone, each tendon running backward from the anterior-facing cones of one myomere to insert on a vertebra from three to seven segments behind it. The most posterior of these, the “posterior oblique tendon” inserts on the first peduncular vertebra behind the prepeduncular joint. A pull on the posterior oblique tendon will swing the peduncle to that side. Flexure at the postpeduncular joint is accomplished by the “great lateral tendon” lying outside the posterior oblique tendon and running out along the peduncle on either side to insert on the bases of the caudal fin rays. A cross section through this tendon reveals it as a series of nested cones representing contributions from the myosepta of successive myomeres. Also included are smaller deep-lying tendons

22

C.C . LINDSEY

which originate from particular parts of the posterior myomeres and insert on particular groups of caudal fin rays. I n various scombroid fishes, other tendons and muscle bands have been described in the caudal complex (Fierstine and Walters, 1968). The wide bony keel of the peduncle serves as a pulley which increases the angle of pull of the great lateral tendons which run on either side of it to flex the tail. The skin surrounding the peduncle is reinforced by collagenous fibers; it forms a strong sleeve which keeps the tendons from bowstringing away from the vertebral column during flexure of both the peduncular joints. The keels on either side of the peduncle, as well as its reduced span, probably also serve to streamline its extremely rapid sideways beats. Sometimes well-developed keels are present as purely external structures (e.g., in Carangidae) which do not separate the tendons but which provide transverse streamlining of the peduncle. The angle of inclination of the caudal fin is altered during each phase of the beat, even more effectively than in the carangiform mode, so as to develop a maximum thrust at all times. The elaborate plan of tendons obviously allows fine control of the tail movements, but again the details are as yet undescribed. The caudal fin is high, short antero-posteriorly, and h a t e (Fig. 1H). It therefore has-a very high aspect-ratio. The tips may project so far up and down as to operate in water relatively undisturbed by passage of the body. The analogy with a high aspect-ratio wing of a bird is rather complete, and the hydromechanical models used to analyze the thunniform mode are based on lifting-wing theory for oscillating aerofoils (Lighthill, 1969; Webb, 1975). The intrinsic muscles of the caudal fin are reduced in contrast to those in subcarangiform fish, and the rays overlap the skeletal base widely. The span can probably be altered only very slightly in most scombrid fishes. In response to the powerful thrusts, even the stiff caudal fin rays show some bending. Here, unlike the subcarangiform mode, the center of the fin leads as it beats from side to side, and the tips follow. In Scombridae there are from five to eleven small separate nondepressible saillike finlets in a row running from the dorsal and anal fins out onto the peduncle (Fig. 1H). Small anal and second dorsal fins occupy much the same positions in Istiophoridae, in Xiphiidae, and in some of the fast pelagic sharks (Fig. lG). These finlets probably contribute little direct propulsive force; they probably serve to deflect water along the peduncle so as to prevent separation of the boundary

1. FORM, FUNCTION, AND LOCOMOTORY HABITS

23

layer and so reduce drag, working in the same fashion as multiple wing-tip slots. Scombroids (with the exception of Xiphiidae) also have a pair of short fleshy horizontal keels, slightly converging toward the rear, on either side of caudal fin base; these may direct a jet which reduces cross flow and boundary layer separation (Walters, 1962). Swimming in the thunniform mode, the scombrid Euthynnus uffinis continues to increase the amplitude of the tail beat so long as speed increases (Fierstine and Walters, 1968). This contrasts with the observation by Hunter and Zweifel (1971) that in the carangid Truchurus symmetricus the tail beat amplitude is a constant proportion of body length regardless of speed. Although Webb (1975) lists T . symmetricus as swimming in the thunniform mode, its shape suggests that it may in fact adopt the carangiform mode. Striking convergence toward the thunniform mode of propulsion can be seen in four unrelated vertebrate groups: among the bony fish, several families of Perciform fishes (most or all being in the suborder Scombroidei); among the sharks (suborder Lamnoidei); among the marine mammals (whales and dolphins); and in the extinct marine reptiles, the ichthyosaurs (Lighthill, 1969).The mammals have a large lunate tail, narrow necking, and a massive streamlined anterior. The fact that their tails lie in a horizontal instead of vertical plane does not alter their basic similarity to the thunniform mode in tunas. Ichthyosaurs also had narrow necking, and lunate tails which, however, lay in a vertical plane; the vertebral column extended into the ventral lobe of the tail, rather than into the dorsal lobe as it does in the heterocercal tail of sharks. The convergence between some sharks and scombroid fish shapes is remarkable (Fig. lG, H). The caudal fin of a thunniform shark, although slightly heterocercal, has a high aspect-ratio. The narrow caudal peduncle carries lateral keels. Anteriorly the body is wide, heavy, and streamlined. A high stiff first dorsal fin reduces recoil. The small second dorsal fin lies far back and opposite to the small anal fin, suggesting similar hydrodynamic function to the finlets described in teleosts. Further parallels between the sharks Lurnna and Zsurus and the tunas are that all are heavier than water, have respiration geared to continuous swimming, achieve high velocities, and have countercurrent blood systems which maintain their body temperatures well above that of the water (see Chapter 4). The whale shark Rhineodon has a shape suggestive of the thunniform swimmers (Lighthill, 1969). Possibly this and some other sharks may use the thunniform mode, but the majority of sharks have

24

C. C. LINDSEY

highly flexible bodies (with many vertebrae) and swim in the anguilliform or subcarangiform modes.

6. OSTRACIIFORM MODE Ostraciiform swimmers have a body incapable of lateral flexure. Propulsion in the ostraciiform mode is b y pendulumlike oscillation of the tail, which pivots on the caudal peduncle. The wigwag motion is induced by a nearly simultaneous contraction of all myomeres involved, on each side alternately. The body shape is variable, but poorly streamlined, and only low speeds are attained. Curiously, the ostraciiform mode does not seem to have been described in detail in any living fish, but it has been simulated repeatedly by diligent builders of working models (Breder, 1926; Oehmichen, 1958; Kramer, 1960; Smith and Stone, 1961; Hertel, 1966; Gray, 1968). Breder’s model, with a rigid oscillating tail, was used to disprove Pettigrew’s (1874) contention that such a system would not produce net forward motion (see Section 11). The closest approach to an analysis of ostraciiform mode in fish is suggested by Webb (1975) to be the description by Gray ( 1933c) of the propulsive cycle of a whiting Gadus merlangus in which the caudal fin had been amputated. Probably a perfectly rigid tail is never found in live fish. Even in the boxfishes, family Ostraciidae, while the body is encased in an inflexible bony armor (Fig. lI), there are a few peduncular myomeres which act “almost as a unit in sweeping this flexible tail from side to side” (Breder, 1926). So also in the puffers, family Tetraodontidae, and porcupinefishes, family Diodontidae, there is some bending of the peduncle and tail as it is oscillated in the ostraciiform fashion. In all these fish, alternative means of locomotion are available by swimming with the dorsal, anal, or pectoral fins. Norman and Greenwood (1975) write that in Ostracion the dorsal and anal fins normally form the chief propelling agents, but when greater speed is required the fish swings the tail vigorously from side to side. The electric ray Torpedo nobiliana cannot bend the body laterally because of its wide expansion (Fig. 5). The caudal fin is quite well developed and almost symmetrical and beats from side to side while the edges of the expanded pectorals are held so as to provide some lift (Roberts, 1969~).Although the caudal fin is somewhat flexible and bends sideways during the tail beats (Fig. 5), this provides an approximation to the ostraciiform mode. Breder (1926) points out that, as demonstrated by his working model, the more flexible the tail in ostraciiform motion, the less the nose oscillates. The anterior of Torpedo

1. FORM, FUNCTION, AND LOCOMOTORY HABITS

25

Fig. 5. Swimming by electric ray Torpedo nobiliana, approximating the ostraciiform mode. Length about 150 cm, Cinephoto intervals 0.5 sec. Symbols as in Fig. 3. (Based on Roberts, 1969c.)

can be seen in Fig. 5 to move very little from side to side, probably in part because the tail is not perfectly rigid. The scabbardfish Aphanopus carbo is an elongate midwater species of the family Trichiuridae which may also use what can be considered an ostraciiform mode of swimming when slowly approaching its prey (Bone, 1971). Its small, deeply indented caudal fin, mounted on a narrow peduncle, can be sculled from side to side using only the musculature of the caudal region. The degree of sweep-back of the tail lobes is varied occasionally. The long body is held rigid, with the dorsal and anal fins retracted. This mode is evidently used in order to produce minimum disturbance to the well-developed lateralis system which senses the prey which it is stalking, and to provide minimum warning to the prey. In the final attack, the scabbardfish switches to a rapid swimming pattern involving powerful undulation of the entire body, with the median fins held erect. Thus, its rapid swimming is anguilliform,. its slow stalking is ostraciiform. Bone (cited in Webb, 1975) suggests that two such swimming modes may also occur in the crestfish Lophotus, and in the trichiurid g$nera Diplospinus, Benthodesma, Lepidopus, and Assurger (Bone, 1971). Although the ostraciifofm mode involves only the peduncle and caudal fin, and so does the thunniform and to some extent the carangiform mode, nevertheless the hydrodynamics are probably entirely different. Species using ostraciiform swimming are poorly streamlined, and scull at very low speeds compared with carangiform or thunniform swimmers. The ensemble of features so characteristic of thunniform swimmers (fusiform body, narrow necking, lunate tail, and fixed caudal rays) are missing in whole or in part in ostraciiform swimmers. The term ostraciiform is intended for pendulumlike oscillation of an essentially rigid blade, unlike the active alteration of the angle of attack characteristic of the tails of carangiform and thunniform

26

C . C . LINDSEY

modes. Although ostraciiform waving of a somewhat elongate and flexible tail may produce undulations (comparable to flag waving), these undulations if passive are not moving faster than the water, and hence provide no thrust, unlike the anguilliform mode. It must be admitted that the ostraciiform category is a mixed bag of diverse forms, the details of whose swimming is not yet understood.

C. Propulsion by Undulation of Median or Pectoral Fins Fish can propel themselves by passing undulations along longbased fins, in a manner analogous to the anguilliform undulations of the body. Instead of myomeres contracting to bend the body, muscles inserted on the sides of the fin rays at their bases (see Section V,B,2) contract to deflect the ray in an appropriate direction. Each fin ray can be moved about independently on a universal joint; successive rays are connected by a flexible membrane. The possible amplitude of the undulation is limited (compared to body flexures) because each ray is attached at its base to the body. Wavelengths can be very short, with several complete wavelengths present on a fin at once. Frequency of fin waves may be much higher than of body waves, reaching 70 Hz in the dorsal fin of a seahorse. The net effect of propulsion by undulation of extended fins rather than of the body is to attain only low or moderate speeds, but to achieve more precise control and maneuverability. Fin undulation usually allows both forward and backward movement, rapid reversal of direction without turning, and the ability to hover and to “drift” into confined apertures with precision. Fin propulsion also allows the body axis to remain straight; this is unavoidable in fish whose body shape or armor allows no flexure, and it may be desirable as an operating base for certain electrosensory or acoustico-lateralis systems. Breder’s ( 1926) compartmentalization of swimming modes was based on the particular fins involved. As explained earlier, his nomenclature is followed here and in Fig. 1 and Table 111 as a convenience, but it is to be supposed that essentially the same hydrodynamic processes may be involved regardless of whether the undulations progress along the dorsal fin, the anal fin, or both. The following discussion may be supplemented by reference to Table I11 and to the publications cited there.

1. AMIIFORMMODE Breder (1926) applied the name “amiiform” to swimming by undulations of the long dorsal fin, as exemplified by the bowfin Amia calva

Table 111 Swimming Modes Involving the Median or Pectoral Fins Swimming mode and principal fin Primarily undulations Amiiform Long-based dorsal

Examples of fish groups

Amia Gymna rchus

Mormyrus

Trachip terus, Regalecus

Short-based dorsal

Trichiuridae: Eupleurogrammus, Trichiurus, Lepturacanthus Syngnathus, Nerophis Hippocampus

Remarks

Reference

May also use all other fins. No anal or caudal fins, ca. 200 dorsal rays. Up to 7 wa~velengthsvisible. Wave velocity 0.63 t'lsec produces body velocity 0.5 elsec. Body held straight. Electrosensory Short anal, propulsion primarily by dorsal. Body held straight. Electrosensory Very long dorsal, highly compressed. Little body flexure. At slow speeds, inclination of body axis compensates for lack of gas bladder No anal or caudal fins. Body held straight. Detect prey by lateral line (decurved to avoid disturbance from dorsal)

Breder (1926) Lissmann (1961a); Alexander (1967)

Uses dorsal and pectorals, and largely ineffectual body curvature. No caudal fin. Pectorals also used. Body axis inclined or vertical. Dorsal fin with 19 rays, 3.5 waves, frequency 70 Hz, amplitudes 0.160.24 times fin base, adjacent rays diverge by 17"

Breder (1926); Aleev (1963)

Bennett (1971a) Nishimura (1964); Nishimura and Hirosaki (1964)

Bone (1971)

Breder and Edgerton (1942); Alexander (1967)

(Continued)

Table III-Continued Swimming mode and principal fin

Examples of fish groups

Remarks ~~

Gymnotiform Long-based anal

Gymnotoidei (South American electric fishes): Gymnotus, Eigenmannia, Electrophorus

Hyperopisus (Mormyridae) Notopteridae: Xenomystus El 0)

Balistiform Long-based dorsal and anal

Monacanthus, Balistapus, Zeus

Balistes lsichthyes (Mormyridae) Cichlidae: Pterophyllum, S ymphysodon; Chaetodontidae

Reference

~

Dorsal and caudal very small or absent. Pectorals small. Several half-waves visible on dorsal. Body held straight. Electrosensory. Electrophorus can also undulate body using wavelengths 4~ fin waves Anal fin base 5 x dorsal, locomotion primarily Gymnotiform. Electrosensory Dorsal tiny or absent. Body held straight, but no special sensory system known. “Idles” with waves on anal passing in both directions

Bennett (1971a); Hertel (1966)

Body deep, compressed, inflexible. Several halfwaves visible on dorsal and anal. Fins inclined to each other. Can reverse direction of undulations. Ray bases can be bent 90”. May augment by caudal or trunk undulation Anterior rays thickened, fins diverge in angle. Can also flap fins suggesting Tetraodontiform mode Body elongate, dorsal and anal undulate in synchrony, 3 wavelengths visible. Electrosensory Deep-bodied, undulate hind edges of soft dorsal and anal, whose margins nearly vertical. Augment by pectoral sculling

Hertel ( 1966); Aleev (1963)

Bennett (1971a) Hertel(l966)

Breder (1926)

Lissmann (1961a)

Bergmann (1968); Bliim ( 1968)

Ophichthyidae, Branchiostegidae, Mastacembelidae, Muraena Pleuronectiformes

Short-based dorsal and anal

Fistulariidae, Aulostomidae Centriscidae

ta

Rajiform Wide horizontal pectorals

Rajidae, Dasyatidae

Myliobatidae, Mobulidae

Pristidae, Rhinobatidae

Elongate; swim either by body undulation or localized undulations of dorsal and anal, or both

Breder ( 1926); Magnan (1930)

Flatfish swim in midwater by body undulation plus synchronized flexures of dorsal and anal of wider amplitude Body slender, rather inflexible. Undulations in dorsal and anal. May augment by caudal Body compressed, armored, stiff. Soft dorsal, anal and caudal close. Rapid vibration of these and pectorals move fish in any direction, usually with body vertical Pectorals expanded, undulate vertically, amplitude greatest at middle rays. No anal; caudal reduced; dorsals reduced or spinelike or absent. More than one wavelength visible. Winglike disc provides lift Pectorals further expanded, width exceeds main body length. Mobile pectoral motion analogous to bird flight. Less than 1 wavelength visible. Pectoral base provides lift Undulate pectorals, but also use moderate caudal fin and body undulation. Trunk and pectoral motions under separate nervous control

Aleev ( 1963); Ling and Ling (1974); Kruuk (1963) Lighthill (1969)

Atz (1962); Klausewitz ( 1963)

Marey (1895); Magnan ( 1930)

Lighthill (1969); Klausewitz (1964)

Campbell (1951); Alexander (1967)

(Continued)

Table III-Continued Swimming mode and principal fin i*,

0

Diodontiform Moderate pectorals with nearly vertical bases

Caudal undulation

Examples of fish groups Diodontidae, Tetraodontidae, some Balistidae

Esocidae, Umbridae, Gasterosteidae, Gobiidae Serranidae, Pomacentridae

Remarks

Reference

Can vary plane of pectoral base, and angle of divergence from body. Rays diverge, partially cancelling vertical components of thrust. Fin base raised, versatile rays operated by tendons. May have 2 wavelengths at a time. May also flap fins (Labriform mode) Pectoral undulation is combined with other modes in many teleosts

Breder (1926); Herald (1961); Schneider (1964); Webb i1975)

Vertical undulations of caudal fin translated into forward thrust by convergence of rays. Or upper and lower lobes may be moved in opposite directions, producing figure-8 sculling. Tail may open and shut scissorlike, or single lobe vibrated while maneuvering. Always accessory to other modes

DuBois-Reymond (1914); Breder (1926); Emery (1973); McCutchen (1970); Arita (1971)

Primarily oscillations Tetraodontiform Short-based dorsal and anal

Tetraodontidae, Diodontidae, Ostraciidae, some Balistidae

Mola Labriform Narrow pectorals with nearly vertical bases

Some Labridae (e.g., Tautoga), Scaridae, Gas terosteidae, Umbridae Some Labridae, Embiotocidae, Pomacentridae, Centrarchidae, Chaetodontidae, Acanthuridae, Serranidae, Sciaenidae, Cyclopteridae, Characinidae, Lamprididae

Paddlelike dorsal and anal flapped side-to-side. These occasionally alternate. Anteriors of fins often stiff. “Erector” and “depressor” muscles modified. Augmented by pectoral or caudal propulsion High dorsal and anal fins flapped synchronously Pectorals flapped rapidly and synchronously, often fanlike and rounded. May also undulate (Diodontifonn mode) Row with pectorals, moving them forward horizontally and backward broadside. Base may rotate slightly. Cymatogaster achieves speed of 3.9 t?/sec, with pectoral frequency 4sec. For high speeds, augment with caudal

Schneider (1964); Breder (1926)

Lighthill (1969)

Webb (1973b); Brett and Sutherland (1965); Harris ( 1953); Emery (1969, 1973); Rosenblatt and Johnson (1976)

32

C . C. LINDSEY

(Fig. 1J).This fish, whose body is jacketed b y a rather heavy armor of scales, usually employs its dorsal fin for locomotion, although to attain rapid movement it can also undulate its body in the subcarangiform mode. It has a full complement of developed paired and median fins. The swimming of Amia has not been studied in detail. An amiiform swimmer that has been better described is the African freshwater Gymnarchus niloticus [for which Lissmann ( 1961a) suggested the term “gymnarchiform”]. The dorsal fin of Gymnarchus extends along most of the length of the body, which tapers to a posterior point. There is no anal or caudal fin. When swimming, the body axis is usually held straight. Lissmann ( 19614 shows from cinematographs that locomotor waves may pass in either direction along the dorsal fin, and may show widely varying amplitude, particularly during turning or braking. The straight body and naked posterior provide a stable base line for a highly developed electrosensory system. A related group of African freshwater electric fish, the Mormyridae, can also hold their bodies straight and propel themselves by undulations of the median fins. Most use the dorsal and anal fins together (balistiform mode) or the caudal, but in the genus Mormyrus the anal is very short compared to the dorsal (Bennett, 1971a), and locomotion is probably primarily amiiform. The ribbonfishes, Trachypteroidei, have a dorsal fin running the whole length of the elongate and highly compressed body. Three species have been observed to swim slowly in the amiiform mode, passing undulations along the dorsal fin with little or no body flexure. At the usual low speeds they swim with the head inclined upward (even with the whole body vertical), so that the undulations of the fin furnish lift to compensate for the lack of a gas bladder. At increased swimming speed the body axis rotates to approach the horizontal, and in Regalecus the body as well as the dorsal fin may contribute to intermittent undulations. Reference was made in the preceding section to certain genera or trichiurid fish which stalk their prey with a rigid body, sculling with the small caudal fin. In these the lateral line passes along the body in the usual position midway between the dorsal and ventral surfaces. In other trichiurid genera (Table 111),the caudal and anal fins are absent, and locomotion is probably in the amiiform mode, using the enlarged dorsal fin and again holding the body straight. Here the lateral line, which is apparently used to detect prey, runs close to the ventral surface and is thus remote from disturbances caused by undulations of the dorsal fin. The seahorse Hippocampus swims characteristically in an upright

1. FORM, FUNCTION,

AND LOCOMOTORY HABITS

33

position, using both dorsal and pectoral locomotion, sometimes aided by the anal fin. There is no caudal fin. Breder and Edgerton (1942) measured dorsal fin waves in Hippocampus with remarkably high frequencies. Variations in wavelength, amplitude, and to a lesser extent frequency may occur in the versatile dorsal fin all at the same time. Maximum swimming speeds in seahorses are low; highest speeds are attained by inclining the body axis from the vertical toward the horizontal.

2. GYMNOTIFORM MODE The swimming of the South American electric fish Gymnotus carupo, after which the gymnotiform mode is named, is comparable to an upside-down Gymnarchus swimming in the amiiform mode (Fig. 1K). There is no dorsal fin. The body is held straight, and forward or backward locomotion is produced by passing rapid undulations of short wavelength in either direction along the elongate anal fin. The related electric eel EZectrophorus can supplement the thrust from its anal fin oscillations by undulating the body; the waves in the body are about four times as long as those in the fin (Hertel, 1966). The South American gymnotoid fishes described above probably combine a straight body and locomotion b y fin undulation for the same reasons as do the unrelated gymnarchid and mormyrid fishes of Afrka; all possess an electrosensory system (Bennett, 1971b) which requires a stable base undisturbed by body undulation. Another tropical freshwater group, the featherbacks, family Notopteridae, have a long anal fin extending from just behind the head to the tip of the tail. The dorsal fin is tiny or absent. Propulsion is in the gymnotiform mode, with numerous waves moving along the anal fin to drive the fish either forward or backward. The body is usually held straight, although no special sensory system has been described in this group.

3. BALISTIFORM MODE The term balistiform refers to simultaneous undulation of the dorsal and anal fins. In trigger fish and file fish, family Balistidae, their fins are somewhat inclined to each other (Fig. lL), so their propulsive waves produce a net forward thrust. Fin spines which lie ahead of the rays are detached from them and do not impede flexure of the fin rays. According to Hertel (1966),BaZistapus aculeatus can swim backward by reversing the usual direction of wave generation on both fins. It can also move downward in the water by passing waves backward on the

34

C . C. LINDSEY

anal and forward on the dorsal fin, or can move upward by reversing both these wave directions. The caudal can be spread like a fan from the folded position to a span 2.5 times as wide, and is used to add propulsive strokes at maximum speeds. The John Dory, Zeus faber, also has great transverse mobility of its posterior dorsal and anal rays which are the sole locomotor organs for slow swimming. For abrupt spurts, the entire trunk is thrown into undulations, and the median fins revert to stabilizers (Aleev, 1963). Other similarly deep-bodied fish such as the cichlids, and the butterfly fish, family Chaetodontidae, undulate the hind edges of their soft dorsal and anal fins. The soft dorsal and anal fins are without anterior thickening in the file fish Monacanthus, and display conventional fore-and-aft undulations. In the trigger fish Balistes the anterior rays are thickened and elongated, and they can on occasion be flapped from side to side as a unit instead of undulating each, so as to produce a rowing motion approaching the tetraodontiform mode. Such thickening in the anterior rays of several forms tends to be associated with angular divergence of the fins away from the vertical. Several different groups ofhighly elongate fish which are capable of anguilliform undulation of the whole body can also move by localized undulations passed along the dorsal and anal fins (Table 111). In flatfish also, anguilliform flexures of the body while swimming in midwater may be supplemented by synchronous flexures of the dorsal and anal fin rays which attain a wider amplitude. The sole Solea vulgaris can also move about on the bottom or dig into the sand without flexing the body, by means of pronounced undulations passed along the dorsal and anal fins (Kruuk, 1963). The slender but rather inflexible fishes in the suborder Aulostomoidei swim by undulations of the soft dorsal and anal fins. Their swimming mode could be described as balistiform; although the fins have shorter bases than those of the previous species, their motion is typically undulation rather than oscillation. In this group the cornetfishes, family Fistulariidae, and trumpetfishes, family Aulostomidae, supplement balistiform swimming with caudal fin movement when hard pressed. The snipefishes, family Macrorhamphosidae, which are deeper-bodied, are armored and stiff, and depend largely on balistiform locomotion; they swim backward as easily as forward, usually with the body vertical and the head down (Herald, 1961). The shrimpfishes, family Centriscidae, also swim with the body vertical, either head-up or head-down. The soft dorsal, caudal, and anal fins are close together near the posterior and almost at right angles to the rigid

1. FORM, FUNCTION, AND LOCOMOTORY HABITS

35

body axis. By rapid vibrations of these three fins, and of the pectorals, shrimpfish can move in almost any direction with astonishing agility. Although the three fins can be closely coordinated, each is short-based, and the swimming shares features of the ostraciiform and tetraodontiform modes.

4.

RAJIFORM

MODE

The primary locomotor organs in most rays, order Rajiformes, are the greatly enlarged pectorals which form a wide lateral expansion of the body. Waves of undulation in a vertical plane are passed backward along the mobile fin margins (“rajiform mode,” Fig. 1N). In the majority of rays the only organs other than the pectoral fins which may contribute at all to forward locomotion are the pelvic fins. These lie at the posterior of the disc created by the snout and pectoral margins and may be used to kick back as the ray is taking off from the bottom. Two families, the eagle rays, Myliobatidae, and the mantas, Mobulidae, have the pectorals.even further expanded than in the skates, Rajidae, so that their width is greater than the length of the main part of the body. These animals are secondarily adapted for a free-swimming rather than a sedentary existence. Their tremendously mobile pectorals produce a graceful rajiform swimming motion (Fig. 6) analogous to the flight of birds (Lighthill, 1969). Longitudinal

T T

u -

I

Fig. 6. Lateral view of rajiform swimming by manta Mobula diabolis. Length unknown. Successive cinephotos (at 0.64 sec intervals), of fish swimming from left to right, are displaced upward. Vertical broken line represents a fixed position on background. (Based on Klausewitz, 1964.)

36

C. C. LINDSEY

sections through the body show it as a series of hydrofoils. The base of the pectorals assumes a sharp angle of attack during the upstroke of the wing tips (Klausewitz, 1964), providing lift to overcome the excess of weight over buoyancy. Mobula also possesses free cephalic flippers modified from the anterior margins of the pectorals; these are operated by a separate musculature to capture food, and have not been shown to take part in forward locomotion. Chimaeras (Holocephali) are said by Breder (1926) to have two types of pectoral locomotion. They may row through the water (labriform mode), or they may hold the pectorals out at right angles to the body and undulate them in the rajiform mode. Their pectorals are not so wide-based as in Rajiformes, and the pectoral locomotion described by Breder is perhaps closer to the diodontiform mode.

5. DIODONTIFORM MODE The diodontiform mode of swimming practiced by porcupinefishes involves localized anguilliform undulation of the moderately broad pectoral fins. Unlike the horizontal pectoral fins of Rajiformes, the bases of the pectorals in diodontiform swimmers are held in a variable plane which may be nearly vertical (Fig. 1 0 ) . The undulatory waves produced by movement of successive pectoral rays on their bases may therefore have a vertical component, partly cancelled out by the fanlike divergence of the rays. The two opposing pectoral fins also diverge from the body laterally at complementary angles which can be altered. Undulatory waves may also be combined with flapping of the whole fin in the labriform mode. Combinations of these variables can generate thrust in almost any direction, allowing slow but very precise maneuvering. The families Diodontidae, Tetraodontidae, and some Balistidae exemplify diodontiform swimming. The versatile pectorals of Diodon can show as many as two wavelengths at a time. Schneider (1964) describes how in Tetraodon the well-developed abductor muscles are attached to each pectoral ray by long tendons. In all these fish the base of the pectoral fin is raised, and the rays can be bent over at a sharp angle. Undulation of the pectorals may be an important locomotory component, always used in combination with other modes, in many other teleost groups (Table 111).

6. CAUDALUNDULATION A category of propulsive techniques not included in the preceding survey and not formally named as a mode involves the propagation of

1. FORM, FUNCTION, AND LOCOMOTORY

HABITS

37

vertical undulations in the caudal fin. This is always an accessory to other locomotory modes; it is not shown in Fig. 1, but could be entered in the center of the diagram. Vertical undulations of the tail fin may be translated into forward thrust because the rays are convergent rather than parallel. “Considering each ray separately, in waving from side to side, it naturally has a forward reaction of the ostraciiform type” (Breder, 1926). A slightly different type of caudal thrust, described b y DuBois-Reymond (1914), results when the upper lobe of the tail is moved to one side and the lower lobe to the other side, creating two backward diagonal thrusts. In this case the central caudal ray remains still, while in the former case all rays may pass through a complete cycle of oscillation. Diagonal thrusts of the opposing caudal lobes have been observed in the sea basses Morone (=Roccus) and Epinephalus, and may correspond to the figure-8 sculling of the caudal in young Pomacentridae (Emery, 1973). In other fish the tail is sometimes opened and shut in a scissorlike motion; one lobe may be vibrated while maneuvering; and individual rays may perhaps be capable of active bending (McCutchen, 1970; Arita, 1971).

D. Propulsion by Oscillation of Median or Pectoral Fins At the opposite extreme from the undulations of extended longitudinal fins (dorsal, anal, or pectoral), which may be compared to ’ anguilliform propulsion, lie very short-based fins whose oscillation may be compared with ostraciiform propulsion. Various intermediate fin modes may be compared roughly with carangiform or other body modes, depending on the length of the fin and the number of contained wavelengths. The analogy is not complete, because short-based fins may also be capable of rotation on the base so as to produce locomotory strokes without parallel in the body/caudal fin series of modes. Ventral segments of the pectoral fin may be rotated relative to dorsal segments, and single fin rays may be capable of active bending (Arita, 1971). These and other complexities, most of which have scarcely been investigated, suggest that this section of the Breder‘s classification of swimming modes is crude at best.

1. TETRAODONTIFORM MODE Puffers in the family Tetraodontidae use their short paddlelike dorsal and anal fins as a unit, simply flapping them from side to side (Fig. 1M). Breder, who named this the tetraodontiform mode, writes that

38

C . C. LINDSEY

these fins may be thought of as an ostraciiform tail in two parts moved slightly forward dorsally and ventrally. While Lagocephalus laevigatus usually fans its dorsal and anal in unison in this fashion, it may occasionally alternate them. Tetraodon Juwiatilis swims with its dorsal and anal fins and its pectorals; the caudal fin is used for steering. The “erector” and “depressor” muscles of the dorsal and anal are enlarged and modified in this species so as to move the fin rays laterally rather than vertically (Schneider, 1964). As mentioned under Balistiform Mode (Section III,C,3), some triggerfish, family Balistidae, which have stiffened anterior rays may flap the dorsal and anal fins from side to side instead of undulating the rays. This type of tetraodontiform swimming reaches an extreme in the ocean sunfish, family Molidae, where the body musculature, and the tail, are virtually lost, and propulsion depends on synchronized sideto-side paddling of the immensely high dorsal and anal fins. “These operate in the bird-flight mode of the Myliobatidae (eagle-rays) but, turned through 90”; apparently a unique case within the animal kingdom of a disjoint pair of high aspect-ratio wings deployed in a vertical plane” (Lighthill, 1969).

2. LABRIFORM MODE Fish which swim by oscillations of their narrow-based pectoral fins are said to swim in the. labriform mode (named from the wrasse family Labridae). The fins are often fanlike and rounded, and may be flapped synchronously. In the stickleback Gasterosteus and the mudminnow Umbra the pectorals may move in this way so rapidly as to appear as a blur. Breder ( 1926) likens this form of labriform motion to the action of a pair of anteriorly placed ostraciiform caudal fins, each producing a large forward thrust, plus a smaller lateral thrust which is cancelled out by the opposite fin. A more complex style of labriform swimming is displayed in some other wrasse species, and in surfperches, family Embiotocidae, and several other families (Fig. 1P). Here, the fish “rows” with its pectorals, bringing them far forward almost edgewise, and forcing them back broadside. Webb (1973b) analyzed the pectoral movements of the shiner surfperch Cymatogaster aggregata. The fin base is inclined down and back at 35” from the horizontal. As the stiff leading edge swings forward it also moves slightly downward; the more flexible following rays lag behind, and do not complete their forward motion until the leading edge starts its backward swing. At the end of the back swing, the leading edge tip comes to rest against the body slightly below its point of departure; the fin is then rotated slightly upward while still pressed against the body, so that the tip starts the next

1. FORM,

FUNCTION, AND LOCOMOTORY HABITS

39

stroke in its original position. The forward and backward phases of the cycle each generate thrust; they also produce components of positive and negative lift, respectively, which combined with slight negative buoyancy cancel out over a complete fin-cycle. At higher speeds the pectoral movements described above are replaced by a slightly different pattern in which the phase difference between posterior and anterior fin-rays is lessened and the fin tends to work more as an ostraciiform unit. The prolonged speed attained by a Cymatogaster using labriform propulsion compares favorably with similar activity levels for fish which use body and caudal propulsion. At this speed, pectoral beat frequencies are of the same order as caudal beat frequencies at similar speeds. The damselfishes, family Pomacentridae, rely to various degrees on rowing with the pectorals. The fins are usually moved in synchrony, but may occasionally be alternated (as also in some Labridae). Emery (1969, 1973) describes different combinations of pectoral and caudal propulsion in the many species of damselfish inhabiting a coral reef, often involving progress in the form of a series of vertical loops. The pectoral girdles are massively developed in those species using their pectorals as the main locomotor organ, and poorly developed in those scarcely depending on labriform propulsion. The pectoral fins, often augmented by other fins or by the trunk, are prominent in propulsion in several other families (Table 111). Many other fish use their pectorals for slow swimming or for holding position, by means of subtle motions of the fins which have yet to be analyzed.

IV. NONSWIMMING LOCOMOTION

The preceding section dealt with straightforward horizontal motion free from the bottom. Obviously many fish spend much of their time in activities other than swimming in a straight line. Indeed, their abilities to perform other short-term maneuvers required for feeding, reproduction, or escape may be crucial in determining their survival. Such activities do not usually lend themselves as well as does straight swimming to kinematic analysis, physiological quantification, or calculation of energy budgets. Their brief discussion here is intended as a guide to further reading. A. Jet Propulsion Breder (1924, 1926) believed that water exhaled from the gill orifices of fish is sometimes an important locomotor asset. He recorded

40

C. C . LINDSEY

that the porcupinefish Chilomycterus schoepfi and the whale shark Rhineodon ejected strong currents through the gill slits. So do banjo catfish, family Aspredinidae, and the sargassum fish, family Antennariidae (Gradwell, 1971; Gregory, 1928). A posterior view of a swimming batfish (family Ogcocephalidae) shows the large circular exhalent apertures which bear a remarkable resemblance to jet aircraft engines, Despite these observations, it is now generally believed that in most fish the small volume of water expelled during respiration makes the thrust which it could generate negligible (Bainbridge, 1961).Jet propulsion, although important in cephalopods (Packard, 1972), is most feasible only at high Reynold’s numbers when inertia dominates viscosity (Lighthill, 1969). In fish it may be important most often in providing impetus during acceleration, particularly from a standing start (e.g., when a ray takes off from the bottom). The slight forward thrust resulting from water ejected through the gill slits during normal respiration requires some locomotory compensation in a fish holding position in midwater. Stationary fish in aquaria are often seen to back-paddle gently with the pectorals or with some versatile fin tip such as the end of the soft dorsal or the caudal lobes. Exhaled water may have another effect in fish locomotion. Aleev and Ovcharov (1973) showed by cinematography that dye ejected from the gill slits of a swimming fish forms a sleeve which envelops the body. It has been suggested that water from the gill slits may keep the boundary-layer fluid surrounding the body charged with enough kinetic energy to avoid separation (Lighthill, 1969). Exhalation has been shown to be coordinated with swimming movements (Satchell, 1968). Further references are given by Webb (Chapter 3, Section 11, c ,1 7 4 . B. Terrestrial Locomotion

Elongate fish which swim in water by anguilliform motion may progress over land in the same way, although, as noticed by Aristotle, usually with undulations of wider amplitude (Breder, 1926). Eels can move overland in this way, and so can the mullet Mugil corsulu (Ganguly and Mitra, 1962). The cyprinodont Fundulus notti moves on land by a series of “flips” which are apparently directed, the body being aligned before each jump by sun-compass so as to return a stranded fish to the water (Goodyear, 1970). The air-breathing catfish Clarius moves on land by locomotory contractions which alternately affect the whole of the lateral muscula-

1.

FORM, FUNCTION, AND LOCOMOTORY HABITS

41

ture of the body halves. The anterior is raised and the body rolled so as to bring the tip of the erect pectoral spine on one side in firm contact with the ground. The body is thrown into a semicircle, and the spine serves as a lever which heaves the body forward in one of a series of steps. At the next step the body is bent in the opposite way. Boulenger (1907) believed that Clarias progressed by independent motions of its pectoral spines, but Johnels ( 1957) observed in Clarias senegalensis that both spines were held rigid. Nawar (1955) describes a large erector muscle, and a locking system, in the pectoral spine of Clarias lazera. The “climbing perch” Anabas testudineus, family Anabantidae, moves over land at up to 3 m/min by spreading its gill covers and placing their ratchetlike spiny lower edges to the ground, and then rocking forward by vigorous action of the paired fins and tail (Herald, 1861). Fishes in this family have accessory air-breathing organs allowing extended aerial excursions. So have the snakeheads, family Channidae, which can move over land using rowing movements of their pectoral fins. The most competently amphibious fish are the mudskippers Periophthalmus and their relatives in the family Gobiidae. Eggert (1929), Harris (1960), and Klausewitz ( 1967)give extensive anatomical and behavioral details. The pectoral fin consists of two functional parts, a rigid platelike proximal region, hinged to the shoulder girdle, which on land is functionally equivalent to the upper arm of a tetrapod, and a fanlike distal part equivalent to the lower arm and plantar surface (Gray, 1968). In water, the mudskipper can paddle slowly with the pectorals out forward, the bucco-pharynx air-filled, and the dorsal fin erect, or it can swim rapidly by vigorous tail movements, the air expelled, and the fins close against the body except for four to five of the pectoral rays which are flexed to form hydrofoils. It can also skim over the water in a series of bounds alternating with brief swimming. (A 14 cm fish covered 2.5 m/sec in. two bounds.) On land the progression is usually ambipedal or “crutching”; the pectorals are swung forward in unison while the pelvic fins support the body; the pectorals are then pressed down and backward so as to lift the body and draw it forward. If these actions are speeded up the whole body leaves the ground in a series of small leaps. An even more rapid skipping movement is achieved by digging in the tail and straightening it violently, while also thrusting upward with the pectorals. (A 14 cm fish can cover 30-40 cm in a single skip.) Mudskippers can also climb, using the crutching motion. A general account of mudskippers in mangrove environments is given by Macnae (1968). The amphibious clinid Mnierpes macrocephalus inhibits steep

42

C . C . LINDSEY

rocky Sores, and makes terrestrial excursions for up to 30 min (Graham, 1970, 1973; Graham and Rosenblatt, 1970).Its lower surface is covered by thick protective pads, and its pectoral, pelvic, and anal fins are heavily padded and incised. On land, it curls its tail up toward its head, extends its pectorals forward, and raises itself up on its pelvic fins. It then simultaneously extends its tail and adducts its pectorals, thrusting the body forward. The tail is alternately curled to either side of the head. One fish was seen to jump 2 m from the shore to the sea. In this species several other physiological and anatomical adaptations to terrestrial excursion have been recorded, including cutaneous respiration, and optical modifications for acute vision in both air and water. The catfish Astroblepus (=Arges)marmoratus was observed to use adhesion by its mouth and by its pelvic region to climb 6 m up a smooth slightly overhanging wall (Johnson, 1912). Suctorial discs formed by the mouth or the paired fins may also be used in combination with inching movements of the body in Petromyzontidae, Loricariidae, and Gobiidae.

C. Moving on the Bottom and Burrowing Many fish which are heavier than water habitually rest on the bottom, but move forward by conventional body and tail movements that carry them on slight excursions away from the substrate. A few move forward by “walking” in continuous contact with the bottom. In Lophius, Antennarius, and Ogcocephalus (order Lophiiformes) the pectoral base is thick and muscular, and the rays are handlike. The frogfish Histrio climbs among sargassum weeds, moving each pectoral and pelvic fin independently, and actually clasping the branches with its prehensile fingerlike dermal rays (Gregory, 1928; Herald, 1961). The goby Gobius and scorpionfish Scorpaena have less highly modified pectorals, but can move along the bottom either by drawing the whole fin in toward the body, or by passing a wave of successive contractions along the adductors of the fin rays (Aleev, 1963). In the family Synancejidae, the goblinfish Choridactylus multibarbis has the three lower rays of each pectoral fin detached from the web, protected by conical horny caps, and capable of independent movement (Samuel, 1961). The sea robins, family Triglidae, also have the lower rays of the large fanlike pectorals separate and capable of independent movement as the fish “feels” its way forward over the bottom. Various cypriniform fish with suckerlike mouths modified for algae-scraping or bottom feeding can pull themselves along the bottom by rapid motions of their jaws. The hillstream loaches, family

1.

FORM, FUNCTION, AND LOCOMOTORY HABITS

43

Homalopteridp.e, adhere to rocks with a large sucking disc formed by the pectoral and pelvic fins; they can creep using the anterior fin rays (Wickler, 1971). The lungfish Protopterus often rests on the bottom supported on the pectoral and pelvic fins. Larvae 2-3 cm long will walk on firm bottom using alternating movements of the paired fins “highly resembling the movements of a lower tetrapod” (Johnels and Svensson, 1954).The lungfish Neoceratodus, whose paired fins are much thicker, also holds both the pectorals and pelvics down below the belly, and uses them actively during restricted locomotion in an aquarium. Burrowing into soft bottom may be achieved by the same anguilliform mode used in swimming [e.g., in Myxine (Foss, 1968), Cobitidae, Ammodytidae, and Synbranchidae]. Fish that can swim backward to burrow include the Indian snake eel Ophichthys (=Pisodonophis) boro (Tilak and Kanji, 1967), the cusk eel Otophidium scrippsi (Greenfield, 1968), and the cucumber fish Carapus acus (Arnold, 1953). Nonanguilliform fish such as the mudminnow Umbra and several wrasses, family Labridae, “swim” into the soft bottom. Killifish, family Cyprinodontidae, bury themselves b y diving into sediment headfirst at 45” (Minckley and Klaassen, 1969). Many different kinds of fish, including rays and flatfish, cover themselves by throwing up sand with undulatory fin movements; some burrow by ejecting water from the buccal cavity (Trueman and Ansell,

1969).

D. Jumping, Gliding, and Flying A fish usually jumps into air by swimming rapidly upward through the surface of the water; its momentum alone carries it forward after the tail has left the water entirely. A basking shark Cetorhinus about 9 m long was seen to jump at least 2 m clear of the water (Matthews and Parker, 1951).Tarpon leap 7-8 ft, salmon 1-2 ft; a Manta, over 5 ft long and weighing over 1000 pounds, can leap over 5 ft with a noise audible a mile away (Norman and Greenwood, 1975). Thunnus leaps several feet, Mugil and Labidesthes leap 10-20 times their lengths in a low arc (Hubbs, 1933). Gray (1968) discusses the ascent by fish of waterfalls, either by swimming in the water stream or jumping over it. Aleev and Ovcharov (1969,1971)published photographs of inked fish jumping through the water surface. Swanson ( 1949) and Gunter (1953) summarize accounts of a variety of fish (including shark, herring, needlefish, halfbeak, and silversides) which have been seen to hurdle repeatedly over a floating object or a

44

C . C. LINDSEY

line in the water. No satisfactory explanation has been offered for this curious behavior, which at least in some cases cannot be attributed to removal of ectoparasites, and may fall into the category of “play.” Some fish skim or skip along the surface in a series of small leaps or a hydroplanelike precursor to gliding flight. The mullet Mugil corsula either swims rapidly along the surface with its head projecting, or skips along using its muscular caudal peduncle, tail fin, and paired fins (Ganguly and Mitra, 1962). Mugil and the halfbeak Hemiramphus have been seen to spread their pectoral fins when in the air (Hubbs, 1933).The needlefish Tylosurus rushes forward as much as 200 m, the body held rigidly at an angle of 30” or more above the surface while the strengthened lower lobe of the caudal fin remains submerged and vibrates rapidly (Breder, 1926). The halfbeak Euleptorhamphus longirostrms, with large pectoral fins, can skim in somewhat the same way and then rise entirely free from the water into a stiff wind and glide for over 50 m in the manner of a flying fish (Myers, 1950). The most proficient aerial gliders are the flying fishes, family Exocoetidae, that despite their name do not flap their wings in active flight but rather use them as sail-planes. In Cypselurus, flight is initiated when the fish accelerates underwater and then emerges partially from the water, spreads its pectoral fins, and skims along with the body obliquely in the air while the elongate lower caudal lobe beats in the water. This taxi may average 9 m in length, with tail beats reaching nearly 70/sec (Hubbs, 1933). Franzisket (1965) calculated that from an underwater speed of 28 km/hr, one fish reached 61.7 km/hr at the end of its taxi. The fish then rises into the air by spreading its pelvic fins, which elevates the posterior part of the body. The pectorals and pelvics are held out to serve as aerofoils during the ensuing glide. After a flight of up to 13 sec, as speed through the air drops, the lower caudal lobe may reenter the water, and by vigorous beats provide increased velocity for another flight, often with a change in direction. As many as twelve consecutive flights have been recorded. Distances covered by one flight may reach 400 m. Hertel ( 1966) discusses some aerodynamic aspects of this type of flight. Details of the flights and of the relative development of the pectorals and pelvics vary among species of flying fish. Some fish may burst directly out of the water without a preliminary taxi. A modification in some Asian cyprinids, not previously described, involves a curious “neck-bending” which is probably associated with escape by skittering along the surface. Chela maassi (Weber and de Beaufort) can snap its head back so that the dorsal surface is almost at right angles to the contour of the back. The action swings the large

1. FORM, F m C T I O N , AND LOCOMOTORY

HABITS

45

pectoral fins downward, so as to thrust the whole fish upward. The pivot for this action is between the first vertebra and the skull, and on either side between the upper end of the cleithrum and a large rounded expansion of the transverse processes of the first two vertebrae. A tough fibrous sheath protects the nerve cord at the site of bending. When fish in an aquarium, having body lengths of 5 cm, were alarmed, the neck-bending and pectoral thrust shot them abruptly upward in the water about 8 cm in 0.08 sec. If they were close to the top, they would break through the water surface. In their natural habitat, swampy pools in the Malaysian region, they elude capture by skittering away over the surface in what may be a rapid sequence of neckbending. Several other Asian cyprinids have a similar physiognomy, with large horizontal pectorals, depressed lateral line, and sharpedged thorax. Among them I have observed the neck-bending only in Chela maassi and C . caeruleostigmata; other genera, including Esomus, may be suspected of using similar escape patterns (Hertel, 1966) even though the neck-bending may be poorly developed. The action of Chela maassi when alarmed suggests a Mauthner reflex such as that of Gasteropelecus (see Section V,A,5), but the neural system has not been investigated. A different neck-bending type of locomotion may have evolved in some South American characoids. The large, elongate, predatory Rhaphiodon (family Cynodontidae) has a sharp-edged thorax and the ability to bend the neck upward, perhaps coupled with downward thivsts of the strong pectoral fins. The anatomy and behavior of these fish, and of the Asian speci.es, offer intriguing research topics. The South American hatchetfishes, family Gasteropelecidae, have large highly placed pectorals operated by massive muscles originating on a greatly expanded pectoral girdle which again forms a sharp leading edge or “bow.” A hatchetfish can taxi over the surface with the tail immersed and the pectorals beating; after about 12 m it may break entirely clear of the water and fly for about 3 m. Unlike the flying fish, hatchetfish apparently beat their wings in flight, emitting an audible buzz. Unlike the pectoral muscles of the Exocoetidae, which are not notably well developed, those in hatchetfish may comprise one quarter of the total weight. The anatomy (Weitzman, 1954) and innervation, (Auerbach, 1967) have been studied, but accurate analysis of the flight of hatchetfish is lacking. A quite unrelated freshwater African osteoglossoid, the butterflyfish Pantodon buchholzi, can also flap its wings. It has large fanlike pectoral fins, with powerful muscles which can move them up and down but cannot fold them against the body (Greenwood and Thom-

46

C. C. LINDSEY

son, 1960). Butterfly-fish can make leaps of up to 2 m, but whether the fin muscles are used in “flight” or whether they only assist in a taxiing take-off is uncertain. The “flying gurnards,” family Dactylopteridae, have generated heated debate concerning their flying ability. They have large winglike pectorals, but these are delicately built, while the body is heavy and clumsy, and the caudal is small. Some authors have attributed to them grasshopperlike flight, but others have expressed disbelief. Hubbs (1933) pointed out that the pectorals are “assuredly utterly insufficient to hold the fish in the air, even for a moment” and their outer portions are “positively raglike.” He concluded that their flight was a myth. On the other hand, these fish often turn up on the decks of small boats at sea. Moreover, if aquarists do not cover tanks containing flying gumards, the fish are liable to escape and land several Just how meters away in other tanks or on the ground (Bertin, 1958~). they do this is not quite clear.

V. PROPULSIVE ANATOMY A. Trunk 1.

SEGMENT

NUMBERAND FLEXIBILITY

A fish is a chain of metameres, with a head on one end and a tail on the other. More than in most other vertebrates, locomotion in fish depends on the propagation of lateral propulsive waves by the sequential action of basically similar units arranged in a series. Each unit has a set of skeletal elements, a set of muscles, a set of nerves, and a set of arteries and veins. Hydrodynamically the chain could be indefinitely long and could still swim in the anguilliform mode; earlier attempts at mathematical analysis of undulatory locomotion indeed assumed an animal with no ends (Taylor, 1952). The number of vertebrae (and hence of metameres) in living fish species ranges from 16 (Mola)to over 600 (Nemichthys).While many selective factors must converge to determine the optimum number for each species, locomotory ability is likely to be prominent. Such a striking range in segment number between very divergent forms could be expected to show some correlations with locomotor patterns. Perhaps more significantly, the considerable variation between races of the same species, and between populations, and even within populations, is likely to have locomotory and hence ecological im-

1. FORM, FUNCTION, AND LOCOMOTORY HABITS

47

portance. Few investigations so far have considered the significance of segment number with respect to locomotion. The greater the number of myotomes, the more flexible the body is likely to be. Magnan (1929) demonstrated this by determining the arc of a circle into which various types of fish could be bent. Fish (such as eels) which could form more than one circle had 5 6 1 2 0 myomeres; fish (such as tuna) in which the head could not touch the tail had only 19-30. There does not seem to be a simple correlation between speed and segment number; the very fast thunniform swimmers have rather low counts (about 31), but fast sharks may have well over 100, and so may relatively slow congrid eels. Species which do not depend on trunk or tail propulsion may have very few (Mola) or very many (Gymna rchus ) vertebrae. There is a definite inverse correlation between vertebral number and robustness of the body. Among related species, those with wider or deeper bodies tend to have fewer vertebrae. The generalization is less valid between widely differing groups of fishes; here, also, narrow bodies tend to go with high vertebral count, but there are elongate blennies with rather few vertebrae, and fat sharks with many. The thickness of the body rather than its segment number may be dominant in determining flexibility. Aleev ( 1963)measured flexibility along five equal sections of the body in a number of species. In the eel, which is fairly uniform in width (Fig. 2A), flexibility was high in all sections, and only moderately greater in the most posterior than in the most anterior section. In the cod, whose width tapers smoothly (Fig. 2B) the flexibility was zero in the most anterior region and rose evenly to a high value at the posterior. In a scombrid (Fig. 2C, D), flexibility was zero at the head and low in the next two sections (where the body is robust); it rose sharply in the region of the peduncle, and fell again in the tail region (a reflection of the double peduncular joint followed by stiff caudal rays). These differences along the length of the body do not, except in the case of the scombrid, reflect any notable difference in vertebral spacing, but they closely parallel the width of the visceral and muscle mass. There is suggestion of a phyletic decrease in vertebral number in the course of evolution. No primitive fish (cyclostomes, elasmobranchs, lower teleosts) have low counts. Strikingly low counts occur irregularly among some but not all the higher teleosts. The forms retaining the most cartilage have many vertebrae and high flexibility. Ontogenetically, the young of all species have high flexibility quite uniformly distributed along the body; rigidity of the anterior may set in with ossification coupled with increased body width. Most young

48

C . C . LINDSEY

fish swim in the anguilliform mode (Fig. 3) regardless of their adult habits. A widespread tendency for fish species with larger adult size to have more vertebrae than smaller related species may have locomotory significance. This phenomenon, termed pleomerism (Lindsey, 1975), occurs within families of differing shapes (sharks, scombrids, seahorses, and sand lances). It exists .within genera, and sometimes between races, between populations, and even between the sexes. It holds only for comparisons between fish with roughly the same form. The average increase in vertebrae is about 10%for each doubling in maximum adult length, but ranges widely between groups. The ubiquity of pleomerism is surprising. If, as is suggested, there is some functional correlation between body size and segment number, it may well be related to locomotion. The size upon which selection operates with respect to vertebral number is not likely to be the maximum size the species ever attains; it might more probably be the size when the larva starts to swim. Selection may not be operating directly on vertebral number, but possibly on some correlate such as optimal length of muscle fiber. Pleomerism seems to be so general a phenomenon that its explanation, if forthcoming, is likely to shed some light on the functional significance of segment number in fish. Numbers of vertebrae (and of fin rays and other meristic series) are subject to several other puzzling sorts of variation which may have locomotory implications. Vertebral counts tend to be higher in forms from cooler waters, or from higher latitudes, than in related forms from the tropics. This phenomenon, known as Jordan’s Rule, is not explainable through pleomerism, although it is reinforced by a tendency toward a higher proportion of large species at high latitudes (Lindsey, 1966, 1975). Aleev (1963) suggests that high vertebral counts in cool regions are functionally associated with larval adaptation to swimming under conditions of low Reynold’s numbers, but experimental data are lacking. Many authors have implied that high vertebral counts at high latitudes are due to direct phenotypic modification by low developmental temperatures. But laboratory rearing of fish of one genetic type at different controlled temperatures does not produce a consistent negative correlation between meristic counts and rearing temperature; in half of the fourteen fish species which have been studied experimentally the curve of vertebral number against rearing temperature has been V-shaped (Lindsey and Harrington, 1972). Latitudinal gradients are more probably the result of selective factors which have produced a genetic cline. Meristic variation is surprisingly high within as well as between wild populations (Fowler, 1970). Some

1. FORM,

FUNCTION, AND LOCOMOTORY HABITS

49

of this is phenotypic, some genotypic. Evidently the precise number of parts in a fish is not critical, but the range of variation which can be tolerated may depend on locomotory performance.

2. SKELETON

The vertebral column provides a series of incompressible blocks, separated b y joints. Contraction of the body muscles on one side therefore produces bending instead of telescoping, and the bending is translated into backward thrust against the water in the manner which has been described. Each vertebra is a complex of several firmly joined pieces (Bertin, 1958a), composed of cartilage in elasmobranchs, and of bone in most adult teleosts. Strength against compression is provided by a spoolshaped cylinder, the centrum, that articulates front and back with adjacent centra. I n most fishes, both the anterior and posterior faces of the centra are concave, although in a few (Lepisosteus, and the blenny Andamia) they are convex in front, and in some eels both surfaces may be flat, or even convex in front (Norman and Greenwood, 1975). The centrum may be supported internally by longitudinal struts of stronger material, in chondrichthyes (Bertin, 1958a), and in teleosts (Hubner, 1961). The relative lengths and widths of centra may vary widely between species, and also along the vertebral column of one fish. Above each centrum is the neural arch which protects the nerve cord, and which (in teleosts but not in elasmobranchs) carries a median neural spine. Below each centrum in the caudal region is the hemal arch which protects an artery and vein, and in teleosts carries a median hemal spine; in the trunk region the transverse processes corresponding to the hemal arch carry pleural ribs which extend down into the abdominal wall. Overlapping extensions (zygopophyses) from the tops and bottoms of adjacent vertebrae maintain the alignment between vertebrae; they allow lateral flexure, but minimize dorsoventral bending. The neural’and hemal spines, lying in the same sagittal plane as the vertebrae, can be rigidly connected to them, since almost no dorsoventral movement is required. I n contrast, those bony structures which project laterally (ribs, intermuscular bones) must either be hinged at their point of attachment to the vertebrae or be completely free-floating. Vertical flexibility of the axial skeleton does occur in some fish at the junction between the skull and the vertebral column. Gosline (1971) describes how the girdle is connected to the skull via a twoplane hinge system (involving the forked posttemporal and the supra-

50

C. C. LINDSEY

cleithrum) that allows some head movement without dislocating the lateral line sensory system where it passes forward onto the skull. In many teleosts the whole cranium tilts when the mouth is opened during feeding (Alexander, 1973). Some myctophids can bend their necks upward (Gosline, 1971). The clingfish Lepadichthys lineatus has a dorsal gap between the first two vertebrae and can bend its head up while attached to the substrate by its ventral sucker (Fishelson, 1968).The neck-bending of some species of cyprinids and characoids has been described in Section IV,D. The coelacanth Latimeria has a jointed cranium, the anterior part of which tilts up when the mouth is opened (Alexander, 1973). Some dorsoventral flexion of the vertebral column may be possible in sticklebacks (Nursall, 1956) and in tunas (Gosline, 1971). But these are limited abilities; the skeletal architecture of fish generally suits them for lateral, not vertical, undulation. Flexibility may vary markedly along the vertebral column, controlled in part by the degree that the zygapophyses of one vertebra overlap the next. The prepeduncular and postpeduncular joints of some thunniform swimmers were described earlier. At each of these joints the zygapophyses are reduced and the adjacent vertebral faces form a lateral hinge. Between the two joints, the zygapophyses greatly overlap so as to lock several vertebrae into a rigid unit, further strengthened by a wide bony lateral keel (Fierstine and Walters, 1968). Between adjacent vertebrae there are fluid-filled cushions, which are evidently remnants of the notochord. Gosline (1971) points out that were it not for these cushions the entire thrust when adjacent vertebrae are bent would have to be transmitted across the point where their lateral rims touched. The hydrostatic cushion, which fills the hollow between facing centra, receives the force from one rim of a vertebral pair and distributes it over the full face of the next vertebra. The cephalochordate Branchiostoma lanceolatum lacks vertebrae, and retains a liquid-filled notochord as a hydrostatic skeleton. Guthrie and Banks (1970a,b) have found that the resistance of the notochord to bending can be altered at different swimming speeds, through altering its turgidity by means of muscle fibers in the curved lamellae which divide the notochord into compartments. Although flexibility of the notochord has not been shown to be controllable in vertebrates, Guthrie and Banks (1970a) note that parallel fibers have been reported in the notochord walls of the hagfish Myxine. In addition to the pleural ribs, there are highly variable series of paired intermuscular bones which develop in the myocommata and are often separated from the vertebral bones, although they may be

1. FORM, FUNCTION,

AND LOCOMOTORY HABITS

51

connected to them by a ligament (Nursall, 1956). The bones may be straight, C-shaped, or Y-shaped. The epipleural intermuscular bones extend out into the horizontal septum; the epineural intermuscular bones extend up and back in the epaxial myocommata. In seventeen freshwater species, Lieder (1961) found the number of epipleural bones (branched plus unbranched) varied from 0 (Perca) to 25 (Aspius), and the epineurals from 7 (Acerina)to 71 (Aspius).Intermuscular bones are exceedingly numerous and complex in some flounders (Pleuronectiformes); they may occur in four rows; each bone may be brushlike at both ends (Amaoka, 1969). The nomenclature of the ribs and intermuscular bones is confusing (Emelianov, 1935; Bertin, 1958a; Harder, 1964). So is their function. Nursall (1956) concluded “These appear to have developed in reponse to stresses and strains set up by the musculature.” Jarman (1961) pointed out that inelastic intermuscular bones might actually impede the contraction of adjacent muscles if they were parallel to the muscle fibers, and also if they were at right angles (since contracting muscles must increase in cross section). Jarman calculated the angle which must exist between the bone and the muscle fibers in order that there be no restriction. Unfortunately, this angle (55”) has not been found in any living animal. Perhaps if the intermuscular bones lie exclusively in the inextensible myocommata, onto which the muscle fibers insert at a considerable angle, they do not impede contraction, but serve somewhat the same function as the tendons discussed in the next section. Intermuscular bones are also discussed in Chapter 5. Other bones unattached to the vertebral column may lie in the midline, often extending inward to interdigitate with the neural (or rarely, the hemal) processes of the vertebrae. These are usually serially continuous with the bones bearing spines or rays in the median fins, the pterygiophores, and are probably homologous with them. In addition to these rayless or spineless pterygiophores, there may be a series of detached median rods ahead of the first dorsal fin, the supraneurals, which usually correspond one-to-one with the neural vertebral processes (Goodrich, 1930; Eaton, 1945; Lindsey, 1955). All these free-floating median elements, which are very variable, are bound tightly into the median septum, and seem to lend strength without detracting from lateral flexibility.

3. MYOMERES

The great lateral bands of muscles running along either side of the body are divided transversely into successive segments, the myo-

52

C. C. LINDSEY

meres. Myomeres correspond in number with the vertebrae (over most of the vertebral column), but alternate with them, so that in the midline each myomere lies opposite the front half of one vertebra and the back half of the next. Within each myomere the muscle fibers are short and tend to run roughly parallel to the long axis of the body, although some may depart by as much as 35”(Alexander, 1969).Therefore most muscle fibers d o not attach to skeletal parts, but instead to tough sheets of connective tissue, the myocommata (or “transverse septa”), which separate adjacent myomeres. The myocommata are anchored in the median plane to the vertebral column, and to its neural and hemal spines, and to the tough median septum. Within the myocommata there may lie the segmentally arranged ribs and intermuscular bones already referred to. Embryonically, the myomeres develop as simple vertical bands, but these then fold in zigzag patterns whose complexity varies widely. The various grades of complexity are illustrated in Chapter 5 (Fig. 1). Myomeres of Brunchiostomu are simply chevron-shaped, with the apex pointing forward. In lampreys and hagfish they are W-shaped with the center of the W, which points forward, low and rounded, and each myomere sloping forward and inward so that its attachment to the median line is ahead of its exposed lateral surface. The angles of the W-flexures become more acute toward the posterior, and additional dorsal flexures may exist in the region of the median fins. Hagfish also have a specialized ventral musculature (Nishi, 1938). In elasmobranchs and teleosts the zigzag folding is complex in three dimensions (Chevrel, 1913; Shann, 1914).The salmonid shown in Fig. 7A illustrates a relatively simple arrangement of myomeres, which are W-shaped on the surface. The forwardly directed flexure of each myomere lies halfway down the flank, with backward flexures above and below it. In some forms an additional one or even two flexures may occur at the dorsal extremity of the myomeres, and occasionally one at the ventral extremity, producing up to six zigzag arms (Nishi, 1938). Beneath the surface, each flexure becomes sharper and assumes the shape of a plough. Internally the forwardly directed flexure often becomes divided into two, most pronouncedly in the more posterior myomeres (Fig. 7A). Each myomere folds forward or backward under its neighbor, so that successive myomeres nest together. The apices are sometimes applied to the vertebrae or to the median septum, forming “halfcones” (Alexander, 1969) or “pyramids” (Shann, 1914). These have traditionally been referred to as complete cones (presumably considering the left and right halves together) or as deep cones. 1eDanois

1.

FORM, FUNCTION, AND LOCOMOTORY HABITS

53

Fig. 7. Muscle systems in chinook salmon Oncorhynchus tshawytscha. (Based on Greene and Greene, 1914.)(A) Lateral view with some myomeres removed. (B) superficial head muscles after removal of the skin and part of the jaws. (C) Dorsal fin after removal of skin. (D) Superficial muscles of the caudal fin. (E) Section across anal fin in ihe plane of the pterygiophores. (F)Ventral view of pelvic region after removal of skin. (G) Oblique ventral view of pectoral fin muscles, with part of protractor ischii removed. AB, abductor ventralis superficialis; AC, adductor mandibulae, cephalic portion; A D , adductor ventralis profundus; AM, adductor mandibulae, mandibular portion; AR, anal fin ray; AS, abductor superficialis; DA, depressor analis; DT, dorsal tendon of lateralis superficialis; DO, dilator operculi; EX, extensor pectoralis; ID, inclinator dorsalis; IN, inclinator analis; LO, levator operculi; PD, protractor dorsalis; PI, protractor ischii; PT, pterygiophore; RD, retractor dorsalis; RI, retractor ischii; T, terminal tendons of lateralis profundus; VT,ventral tendon of lateralis superficialis.

(1958) likens this arrangement to a series of babouches (oriental slippers) placed side b y side with the toes pointed in alternate directions; the median skeleton and septum form a sort of common sole to all the slippers (Fig. 7A). Often the apices of the forward or backward projecting zigzag do not touch the median plane, but instead lie in the midst of the muscle mass lateral to the midline. The myotomes then form nesting cones, which have been called incomplete cones or superficial cones. The nests of these cones will appear in transverse section as concentric circles of successive myomeres (Fig. 7E). Because of their sharper

54

C. C. LINDSEY

flexures, the more posterior myomeres show greater overlap of cones. In many teleosts the muscle cones extend past 4 or 5 vertebrae; in the swift lamnid sharks and in Thunnus they may extend past as many as 11 and 19 myomeres, respectively (leDanois, 1958). If the apices of the cones lie away from the median plane, the myocommata which wrap the cones become attenuated at the apices, to form connective tissue extensions of varying thickness among different species (Nursall, 1956). In most higher teleosts the extensions are weak and disappear in the myomeres into which they project, except in the caudal region, Amia has no tendons; Anguilla and Salmo have tendons much less prominent than those in elasmobranchs (Alexander, 1969). In the shark Scyliorhinus there is a broad tendon which projects, lateral to the apex of each cone, in the direction in which the cone points. The scombroid fishes have very distinct tendons which attach to the vertebral column or to the pectoral girdle anteriorly, or to caudal structures posteriorly (Kishinouye, 1923). The origin and attachments of tendons which operate the prependuncular and postpenduncular joints in tuna have been described earlier (Section III,B,5). In nonscombroid teleosts the cones of the last few myomeres may also give rise to tendons which insert on the caudal base (Fig.

7D). I n sharks and bony fish (but not in lampreys or hagfish) the myomeres are divided into dorsal epaxial and ventral hypaxial portions by a horizontal septum of connective tissue. The horizontal septum runs inward to the centra usually slightly below the level of the forwardly directed myomere flexure. Kafuku ( 1950) concluded from examination of fifty-two species that the horizontal septum consists of two sheets of tendons which can slide over each other; an anterior oblique set runs forward and inward and attaches to the front half of each centrum, while a posterior oblique set runs backward and inward and attaches to the back half of the centra three or more vertebrae toward the posterior. In fish other than scombroids the distal (outer) end of each tendon in the septum, in the form of connective tissue, attaches to the superficial red muscle which lies as a thin strip or sheet just beneath the skin. In scombroids the comparable tendons attach to the deep-seated red muscles, which Kafuku believes are homologous with the superficial red muscles in nonscombroids. The absence in lampreys and hagfish of both a horizontal septum and a lateral band of red muscle supports the likelihood of the functional relationship between the two implied by Kafuku’s observations. The significance of the red muscle system in swimming is referred to again later. An additional ligament, parallel to the vertebral column, may be

1.

FORM, FUNCTION, AND LOCOMOTORY HABITS

55

found fused to the median septum, where successive myocommata come together (thereby excluding all muscle fibers) to form a “deep longitudinal ligament” (Nursall, 1956). I t commonly lies between the main anterior flexure and the epaxial posterior flexure. From species to species it varies in width and density. Longitudinal ligaments are particularly well developed in elasmobranchs, where they may occur in the hypaxial as well as the epaxial regions. The muscle system contains t w o functionally different groups, designated red and white because of their usual color. Red muscle in most fish is superficial, lying in a relatively thin sheet along either flank, with its greatest thickness forming a wedge along the outer margin of the horizontal septum (which divides it into dorsal and ventral bands). In salmon the red muscles (musculus lateralis superficialis of Greene and Greene, 1914) run together at the posterior to form tendons attached to the caudal fin base (DT and VT in Fig. 7D). I n some fish red muscle is scattered in a mosaic of red fibers among the more numerous white. In scombroids there is a substantial block of red muscle straddling the horizontal septum and in some species reaching in to the midline (Kafuku, 1950; Graham, 1975). The relative development of red and white muscle in different kinds of fish may be correlated roughly with their mode of life (Boddeke et al., 1959). Red muscle is usually slow, with low contractile power, and is used for prolonged activity sustained by aerobic metabolism. White muscle is faster, more powerful, and capable of burst activity which may be anaerobic. Literature on the physiological and histological differences between red and white muscle is reviewed by Patterson and Goldspink (1972) and Johnston and Goldspink (1973a) and in Chapter 6. Although individual muscle fibers extend only from one myocomma to the next, it is possible to tease out successive fibers which lie end-to-end on opposite sides of myocommata, and to follow this thread over as many as fifteen myomeres (Kashin and Smolyaninov, 1969). The paths followed by these threads are called muscle-fiber trajectories by Alexander (1969), who provides the most complete analysis so far available. The fibers of red muscle tend to run parallel to the body axis, but the fibers in white muscle are arranged in complex three-dimensional patterns. There are two basic patterns, one found in all myomeres of elasmobranchs, primitive teleosts, Anguilla and Salmo. In higher teleosts this pattern is restricted to the last few myomeres. T h e patterns differ in the direction of slope of the muscle fibers relative to the myocommata. In the chondrichthyean pattern some of the resultant muscle-fiber trajectories run between a tendon of an anterior cone and

56

C . C. LINDSEY

a tendon of a posterior cone, while others run between the tendon of a cone and either the vertebral column or the median septum. I n the teleostean pattern the trajectories form segments of helices with axes roughly parallel to the body. The helices are in coaxial bundles, with four bundles on each side in a typical teleost. They are illustrated in Chapter 6 (Fig. 3) of this volume. Mast trajectories seem to begin and end at the median plane. Probably the lengths and angles of the white fibers are disposed so that all contract to a similar extent when the fish bends (Alexander, 1969). The interposing of inextensible tendons (and perhaps intermuscular bones?) between some muscle trajectories and their ultimate attachments may serve to unify the rate of muscle contraction in different regions. In teleosts where tendons are poorly developed, myocommata running obliquely laterally from the points of attachment of small tendons to cones may act as lateral parts of the tendons. An additional reason for the transformation of muscles to tendons in the slender caudal peduncle of thunniform swimmers is that the tensile strength of the tendons, which transmit immense forces, is much greater than the maximum isometric tension of muscle. Just where the body muscles attach to the axial skeleton has been the subject of debate. Nursall (1956) observed an “arch of muscle” visible in a near-horizontal section of a whole fish, which gave the impression of being attached to the vertebral column both anteriorly and posteriorly. The contraction of the arch on one side would bend the vertebral column. Nursall was aware that the pattern might be misleading, since it depended on the precise depth and plane at which the horizontal section was cut. Willemse ( 1959) published figures showing that this is so, and that no arch appears in sections cut at some other levels. H e further dismissed the “arch of muscle” theory on the grounds that for a wave of flexure to pass backward the morphological structure would have to do likewise. Willemse also pointed out the myocommata are not designed to transmit bending forces to the vertebrae because of their angles of insertion. Willemse ( 1959) proposed instead a “bimetal theory,” whereby contractions of the muscles on one side of the body cause it to bend toward that side, after the manner of a bimetal strip with halves of different contractability. In this view, the myosepta should not be considered as tendonlike structures transmitting muscular forces to the vertebrae. H e pointed out the extra advantage that “In our theory the intricate and rather confusing form of the myosepta is not used.” Szarski ( 1964) accepted Willemse’s theory, emphasizing that if the myocommata were actually transmitting contracting forces to the vertebrae, one would expect the myosepta to be strongest near the body

1. FORM, FUNCTION, AND LOCOMOTORY HABITS

57

axis “whereassexactly the opposite is true.” Szarski went on to seek a reason for the complex folding of the myomeres. He concluded that it is to ensure the smooth passage of contractile waves down the body. If the myomeres were flat transverse plates, a smooth passage of contraction could be achieved only by a complicated pattern of nervous impulses from the successive metameric spinal nerves. But with the overlapping myomere folds, even if all fibers in a single myomere contract and relax simultaneously, a gradual increase and then decrease is achieved in the total number of contracted fibers across any cross section of the body. Hence any type of myomere folding is advantageous, and works equally well at all swimming speeds. Alexander ( 1969) argued against Willemse, who, he wrote, was “apparently forgetting that a bimetal strip will only be bent by differential expansion if its components are fastened together in such a way that they cannot slide past each other.” Lund (1967)was also critical of some of Willemse’s views, but he noted that fish could swim according to the bimetal theory if the lateral muscles were held together at the median septum. H e also noted that leptocephalus larvae swim without any skeletal or cartilaginous elements to resist telescoping. Alexander’s analysis of the lateral muscles as a series of muscle fiber trajectories has been given above. Most of these trajectories probably run from bone to bone, bone to tendon, tendon to tendon, or in a helical path out from the median septum and then back to it farther along the body. If these trajectories are the functional unit‘s, they are suggestive of a whole series of Nursall’s “muscle arches”; the problem of moving the arch no longer remains, if successive trajectories can be contracted in sequence. But since each “trajectory” is the sum of many fibers, each of which belongs to a separately innervated myomere, there remains the large question of how nervous coordination could contract each trajectory in sequence. If the connection of the .superficial strip of red muscle to the vertebral column via a double set of diagonal tendons in the horizontal septum (Kafuku, 1950) is of general occurrence, then the red muscle locomotory system seems to be mechanically independent of the white muscle system (and easier to understand). In salmon, the red muscle band inserts on to the caudal fin base by its own tendons. The red muscle system can function independently; for example, trout swimming slowly use only their red muscle, and switch on the much larger white (mosaic) muscle system only at intermediate and high speeds (Hudson, 1973). It seems that to some extent fish may have two alternate sets of engines whose wiring, fueling, and propulsive systems are separate. It will be apparent from the foregoing that myomeres, septa, mus-

58

C.

C. LINDSEY

cle fibers, tendons, and bones interlock in a three-dimensional jigsaw puzzle which is difficult to comprehend. Sir James Gray (1968) wrote “This complex arrangement is almost certainly of functional significance, but so far no completely convincing analysis seems available.” At present, the emphasis on the trajectories traced by threads of end-to-end muscle fibers (Kashin and Smolyaninov, 1969; Alexander, 1969), rather than on the geometry of the fibrous walls (myocommata) which those trajectories pierce, seems to promise an improvement in the functional analysis of the jigsaw puzzle. 4. OTHER TRUNKMUSCLES The myomeres described above usually make up the great bulk of the lateral musculature, and provide almost all the propulsive force. Other muscle systems or modifications in the trunk will be described briefly. In addition to the references given below, Takahashi (1917) describes the structures and homologies of the carinal muscles in fish, and Winterbottom ( 1974) gives an extensive compilation of literature on the striated muscles in teleosts. In the region of the visceral cavity the hypaxial trunk muscles of right and left are separated, and the myomere arms are flattened and modified where they form the abdominal wall. Below the lateral line there may be a patch of superior oblique muscles (with fibers sloping forward and up); ventral to it, a patch of inferior oblique muscles (with fibers sloping forward and down); and sometimes, on the ventral surface of the belly, a sheet of straight rectus muscles (Maurer, 1913; Nishi, 1938; leDanois, 1958). The epaxial trunk muscles are fastened anteriorly to the back of the skull and to the upper pectoral girdles (Fig. 7B). The muscles expand and anchor into large concave spaces on the posterior face of the skull; in higher teleosts they may reach forward and attach to three longitudinal crests of bone, the supraoccipital and two frontal-parietals (Gosline, 1971). Lying along the extreme dorsal margins of the lateral muscles are separate longitudinal muscles, the supracarinals. They run from head to tail, but may be interrupted by the dorsal fin(s). The anterior supracarinals (or protracter dorsalis) run from the skull and upper part of the pectoral girdle back to the pterygiophores of the first dorsal fin (PD in Fig. 7B and C). In salmon they have segmentally arranged septa which are in step with the myomeres but which have their own irregular and complex folding (Greene and Greene, 1914). Despite their name, they serve more to produce dorsal flexion of the body than erection of the dorsal fin.

1.

FORM, FUNCTION, AND LOCOMOTORY HABITS

59

Behind the dorsal fin, the posterior supracarinals (or retractor dorsalis) run from the posterior pterygiophores of the dorsal fin back to the neural elements of the caudal base (RD in Fig. 7C). In salmon the muscles are cylindrical, and are continuous past the base of the adipose fin as a pair of slender tendons. In species with two or three rayed or spinous dorsal fins the supracarinals may be interrupted. In scombraids the posterior supracarinal may insert on each finlet. In some eels with a long continuous dorsal fin the supracarinals run the whole length of the body with separate insertions on each fin ray (leDanois, 1958). Lying along the mid ventral line is another set of paired longitudinal muscles, the infracarinals, that are interrupted by the pelvic and anal fins. The anterior infracarinals (or protractor ischii) run from the branchiostegal plate or the pectoral girdle, backward on either side of the midline to the pelvic girdle (PI in Fig. 7F and G). In salmon they are segmented, comparatively simple at the anterior but spirally folded toward the posterior (Greene and Greene, 1914).Although their contraction pulls the pelvic girdle forward, their more important functions may be to flex the body ventrally, and possibly to extrude eggs from the abdominal cavity. The anterior infracarinals are best developed in soft-rayed fishes whose pelvic fins are far back on the body; they may be small or absent in advanced forms whose pelvics are thoracic or jugular in position. In eels and brotulids which lack pelvic fins there are no anterior infracarinal,s, but in the sand lance Ammodytes the infracarinals are continuous from the throat to the anal fin (leDanois, 1958). Behind the pelvic fin the retractor ischii pass backward from the pelvic girdle, running on either side of the anus to insert on an interhemal bone at the anterior of the anal fin (R1 in Fig. 7F). Their contraction may contribute to ventral flexion of the body, or to retraction of the pelvics, or to protraction of the anal fin. A pair of slender muscles, the retractor analis, run from the posterior pterygiophores of the anal fin back to connective tissue and to the hemal spines at the caudal base. These muscles produce retraction of the anal fin, but are only slightly developed (Greehe and Greene,

1914).

5. MUSCLESOF RESPIRATION The possibility of jet propulsion, and of avoidance of boundary layer separation, by the ejection of water through the gill apertures has been discussed in Section IV,A. Water is pumped into the mouth and out through the gill apertures by a coordinated expansion and contrac-

60

C. C. LINDSEY

tion of the oral chamber and the branchial chamber (Hughes and Shelton, 1962). An outline of the pump musculature and skeleton, in teleosts, elasmobranchs, and cyclostomes, is given by Shelton (1970). In Fig. 7B only one of the three muscles involved in inspiration (the dilator operculi), and one of the seven involved in expiration (the levator operculi) are labeled. Much of the muscle mass of the head functions to operate the jaws in the capture of food rather than in respiration; in Fig. 7B the cephalic and mandibular parts of the mandibular adductors are prominent and are labeled. The branchial muscles are described b y Greene and Greene (1914), Nishi (1938), Thomas (1956), leDanois (1958), and Harder (1964). The mechanics of breathing by fishes are described by Hughes and Shelton (1962), Alexander (1967), Gosline (1971), and Shelton (1970). AND COORDINATION OF TRUNKMUSCLES 6. INNERVATION

The anatomy of the central nervous system in fishes has been reviewed in an earlier volume of this series (Bernstein, 1970). Patterns of peripheral innervation of the muscles are reviewed by Barets (1961) and Bone (1964). Nishi (1938) discusses innervation of the trunk musculature. The anatomy of the spinal nerves of elasmobranchs is described by Norris and Hughes (1920), and Roberts (1969a). Innervation, proprioception, and coordination of the muscles is discussed in Chapter 6. Of significance in the generation of waves of muscular contraction is the observation that while each spinal nerve provides branches mainly to its own myomere, the sensory fibers and probably the motor fibers also send branches across several adjacent segments (Nishi, 1938; Bertin, 1958d). “Because of the extent of sensory innervation any one segment will be provided with information about the activity of neighbouring segments, whilst the overlap of motor innervation, leading to the simultaneous contraction of many muscle fibres in different segments, will ensure a smooth transition from the contracting to the relaxed zone of the musculature” (Roberts, 1969a). As would be expected from their ability to function independently, the red muscles in each segment receive motor nerves separate from those going to the white muscles. Roberts (1969a) found this in dogfish, and McMurrich (1884) observed a distinct superficial nerve plexus supplying the superficial lateral muscles in catfish. But the two muscle systems do not seem to have separate sensory nerves. Fish muscles lack the spindles which in tetrapod muscles are length detectors. Evidently the proprioceptors in fish are not deep among the

1. FORM, FUNCTION, AND LOCOMOTORY HABITS

61

muscles, but lie in the skin or in the very outer layers of the myomeres (Roberts, 1969a). Sensory discharges from these receptors in the body wall of dogfish are capable of providing information on the frequency and angle of bending of the trunk during swimming (Roberts, 1969d). Sense organs of the lateral line system (reviewed by Disler, 1960) may also provide information useful in locomotory coordination (Roberts, 1972; Roberts and Russell, 1972). Regarding locomotory coordination, there has been much experiment and discussion as to whether the rhythmic muscular activity during swimming is generated (a) by the spinal neurons (with proprioceptive feedback being capable of modifying but not of initiating the rhythm) or (b)by impulses from the proprioceptors when the muscles develop tension or are passively stretched. The subject is well reviewed b y Healey (1957). In teleosts, if the spinal cord is cut it is not possible to initiate locomotory activity by passive stretch of body muscles posterior to the cut. This suggests theory (a), in which spinal neurons in the central nervous system are essential (Gray, 1936a, 1968). Experiments on elasmobranchs give different results (Lissmann, 1946a,b). In dogfish the timing of the locomotory rhythm is dependent on proprioceptars which are excited b y swimming movements-that is, theory (b). The spinal neurons are capable of intrinsic activity, but this is normally overriden by proprioceptive input (Roberts, 1969b). The mechanism which controls the tonic posture, and difference in phase of the body waves, is different from that which controls rhythmic movements (Gray, 1968). Moreover, very young elasmobranchs show rhythmic activity which must be myogenically induced because it persists even after total destruction of the whole brain and spinal cord (Harris and Whiting, 1954). Apparently the control of rhythmic movement may have input from several sources, whose relative importance may vary in different phyletic groups and also in different developmental stages. A table showing theoretical ways in which spinal activity waves might be propagated is presented by Riss (1972). A summary of current views on the role of proprioception in muscular control is given in Chapter 6. For quick avoidance reaction when a fish is startled, the usual reflex pathways between sensory centers and motor nerves may be shortcircuited by the Mauthnerian system. This is a pair of large nerve cells in the medulla of the brain, each connected to a long axon running the whole length of the spinal cord. The Mauthner cells receive input from the optic and acoustico-lateralis centers in the brain; their axons give off branches to the motor nerves of the muscles. The small

62

C. C. LINDSEY

number of synapses, and the large diameter and heavy sheathing of the Mauthner fibers, allow for sudden and coordinated response of the swimming muscles, The result is usually a violent “startle response,” usually in a direction unpredictable (by predators), followed by rapid directed swimming as the usual motor pathways take over from the Mauthnerian system. The Mauthnerian system is present in Cyclostomes, Chimaeras, and Teleosts, but not in adult Elasmobranchs. Among Teleosts it is best developed in species relying on quick movement for escape. It is poorly developed in bottom dwellers and in some anguilliform swimmers, and is absent in scombroid fishes and in fishes lacking tails. The anatomy and physiology of the Mauthnerian system have been described earlier in this series (Diamond, 1971). Its differential development in fish with different habits is discussed by Marshall (1971).

B. Fins

1. ANATOMYOF RAYS

AND SPINES

The mechanical-plan of the fin rays and fin spines is easier to comprehend than that of the myomeres. The flexible web of each fin is a double sheet supported by stiff rods like spokes of a fan. Each rod is attached at its base through a hinge or swivel joint to internal supports buried in the body. Each rod is operated by sets of muscles which can swing it in one or more planes. The whole complex of web, rods, internal supports, and muscles which make up a fin is usually inserted into the body as a largely self-contained unit without serial correspondence to the adjacent myomeres. Teleost fin rays and spines arise embryonically as thickenings in the basement membrane of the dermal fin fold. The opposed pairs become fused along most of their length to form the shaft of the ray or spine. The halves remain separated at the base, where they are expanded to form a pair of tongs which in fin rays grasp the outer edge of their internal support (Fig. 7E). Fin rays may branch fore-and-aft several times as they approach the outer edge of the web, and their shafts are divided into segments by regularly spaced rings which probably contribute flexibility. Fin spines, in contrast, come to a point at their tips, rather than branching; the halves of the shaft are fused almost to the base to form a rigid, unsegmented spear, and the base typically forms a strong hinge on the internal support which can swing in the median plane but not laterally.

1. FORM, FUNCTION, AND LOCOMOTORY HABITS

63

In adult teleosts the spines, rays, and internal supports are ossified; the bony rays are called lepidotrichia, and are thought to be homologous with the scales. The fins of elasmobranchs are supported by ceratotrichia composed of collagen; they are more closely packed then teleost lepidotrichia but are also paired structures, each grasping a median cartilaginous support. Very fine horny filaments, the actinotrichia, also occur in the fin fold of larval teleosts, and may persist in the outer edge of the adult fin membrane. Rays in the dorsal and anal fins of most teleosts correspond one-toone with their internal supports, and can be erected, depressed, or inclined freely on a universal joint. In the other teleost fins, and in elasmobranches, the rays are more numerous than their supports and are usually more restricted in their movements. The sets of muscles producing fin movement are described in the following sections. In addition to movement about the ray base, there is increasing evidence that sometimes individual rays are capable of active bending (McCutchen, 1970; Arita, 1971). This may depend on the fact that a ray consists of two halves whose tips are attached and whose bases are operated by separate muscles on each side (Gosline, 1971); it may also be that there are fine muscle fibers in or between the rays. General discussions on the anatomy of fish fins may be found in Bridge (1896), Goodrich (1906, 1930), Schmalhausen (1912, 1913), Grenholm (1923), Eaton (1945), Lindsey (1955), Bertin (1958b), Francois (1959), Alexander (1967), and Gosline (1971). 2. DORSALAND ANAL FINS In elasmobranchs the internal support to the dorsal or to the anal fin is a flat median mosaic of cartilaginous plates called radials. The outer of the two or three rows of radials (which are joined by ligaments) projects out into the fat fin base, sometimes extending almost to the fin margin (Francois, 1959). The tightly packed ceratotrichia greatly outnumber the underlying radials, which they grasp (Roberts, 1969b). There is a single muscle on each side of each radial, originating on the outer margin of the myotomes and inserting by a broad tendon on the bases of all the adjacent ceratotrichia. Contraction of the radial muscles bends the fin to the side, along the longitudinal joints between the rows of radial elements. The fin of a shark cannot be collapsed, nor apparently can its area be altered. In the Chondrostei, the most primitive living teleosts, including Acipenser and Polypterus, the median fin rays also outnumber their internal supports, but in all other teleosts the two correspond serially except at the fin extremities. The rays or spines are more widely

64

C . C . LINDSEY

spaced, and so are their internal supports. Each of these median supports, the pterygiophores, has two or three elements: a large proximal dagger-shaped element called a basal (whose inner end may overlap the tips of the neural or hemal spines), a small mesial element which slopes backward to form a spacer” between successive pterygiophores and which during ontogeny usually becomes fused as part of the basal, and a distal radial element which may be grasped by the ray base. Each fin ray rides on a double joint. Its tonglike base can pivot in the median plane on the distal radial beneath, so the fin can be erected or depressed. Each distal radial, in turn, can rotate from side to side at its junction with the more proximal pterygiophore. Spines, however, usually cannot be inclined to the side, although they can be depressed. Their bases may be massively constructed to resist forces from the side, and may have devices to lock them in the erect position (Hoogland, 1951; Bertin, 1958~).Useful diagrams of the joints in dorsal fin rays and spines of Tilapia are given by Geerlink and Videler (1974). A fin ray is typically provided with three pairs of muscles, one of each pair inserted on each half of the ray base. The erectors and depressors originate on the basals and the median septum and are inserted on the front and back surfaces of the ray base, respectively. Contraction of one of these sets will cause the ray either to rise or to fall. The inclinator muscles originate on the inner side of the skin or from the myomeres, and run diagonally backward to insert on the sides of the ray bases (ID in Fig. 7C, I N in Fig. 7E). Their contraction will bend the ray to one side. There is much variation in the inclinator muscles of the median fins. Spines, which cannot be bent sideways, usually lack inclinators. In Tilapia the spines have small inclinators and also a fourth set of “interinclinators” alternating with these; they may give support to the pterygiophores when strong lateral pressure is exerted on the fin (Geerlink and Videler, 1974). Catfish may lack inclinators but be able to swing the dorsal fin sideways, perhaps either by contracting the erector and depressor on one side simultaneously (Alexander, 1967),or by using the epaxial trunk musculature (Mahajan, 1967). The puffer Tetraodon has no inclinators, but uses its large erectors and depressors, which reach the ray bases via tendons, to swing the dorsal or anal fin sideways (Schneider, 1964). There are many other variations in the origin and insertion of all these median fin muscles, which in various teleosts have become attached to the skull, or to the cleithrum, or to the neural or the hemal spines of the vertebrae. Both red and white muscles can occur; the “

1.

FORM, FUNCTION, AND LOCOMOTORY HABITS

65

dorsal fin of dogfish has an outer layer of red and a much thicker layer of white fibers (Roberts, 1969b). Bergman (1964) found that red fibers in the dorsal fin of a seahorse actually contracted faster than their associated white fibers. The foregoing description applies equally to dorsal and to anal fins, which are essentially alike in their anatomy. Both may be composed only of soft rays, or, in teleosts, or spines anteriorly and rays posteriorly. If spines are present either in a separate fin or confluent with a soft-rayed fin, they almost invariably lie ahead of the soft rays. If no spines are present, the leading edge of the dorsal or anal may be stiffened b y close-packed rays. This pattern of a rigid anterior and mobile posterior seems to conform to hydrodynamic requirements. The anterior spinous part can be erected rigidly for stabilization, defense, or display, and it can sometimes be folded flat. The posterior soft part provides the mobility. The evolution of amiiform, gymriotiform, balistiform, and tetraodontiforrfi modes of swimming (Fig. 1) has depended on the development of versatile propellors provided by fin rays moving on universal joints. In salmoniform, cypriniform, and a few other teleost groups there is an adipose fin lying between the dorsal rayed fin and the tail. It is often merely a small tab in the adult fish, but in some catfishes the adipose is both long and high. The adipose almost always lacks supporting structures other than fine actinotrichia in the membrane (San don, 1956). Muscles are absent, or rarely form inclinators inserted on a median membrane (Grenholm, 1923). Gosline (1971) suggests that the adipose may be significant chiefly for hydrodynamic reasbns during the juvenile stages.

3. CAUDALFIN Heterocercal caudal fins, found in most elasmobranchs and in chondrosteans (sturgeons and their relatives), have the posterior section of the vertebral column bent upward and continued almost to the tip of the fin. The hemal processes of the vertebrae are elongated, and project down and back to provide a stiff base to the expanded lower (or hypochordal) lobe of the fin. Some anterior hemal processes involved in the fin base may bear detached radial elements. The neural processes are shorter than the hemal, and carry an outer row of more numerous radials. These radials support a tightly packed row of fin rays (ceratotrichia in elasmobranchs, lepidotrichia in sturgeons) which occupy the long low ridge which is the upper (or epichordal) fin lobe. The hypochordal lobe is much deeper, and is strengthened by longer

66

C. C. LINDSEY

rays which grasp the expanded hemal processes. It is often indented on its trailing edge to form a ventral hypochordal lobe with long rays, and a longitudinal hypochordal lobe with shorter rays (nomenclature of Thomson, 1971). The posterior end of the latter lobe may be separated, by a subterminal notch, from a small flap carried beneath the tip of the vertebral column (called the inferior lobe of the epichordal lobe b y Simons, 1970). In some swift pelagic sharks the vertebral column is bent up more sharply, the ventral lobe is extended, and the resultant high aspect-ratio outline resembles the symmetrical tail of a tuna (Fig. lG). The myomeres in elasmobranchs extend out along the posterior section of the vertebral column and retain their zigzag pattern almost to its tip. They are capable of resisting bending of the column. In the hypochordal lobe there are numerous stratified muscle bundles, running diagonally down and forward from the skin which covers the ventral edge of the myotomes, and inserting on the hemal processes (Nishi, 1938). It is commonly stated that when a shark tail is swung sideways during swimming the hypochordal lobe lags behind the vertebral column and so contributes lift. This may not be so. Simons (1970) has shown that the hypochordal lobe may actually move in advance of the rest of the tail. The radial muscles in the lobe are so placed as to be able to alter the contribution it makes to the action of the whole tail (Alexander, 1965a). Simons believes that the hypochordal lobe is a device for altering the “horizontal trim” of the fish. Thomson (1971) points out that even though the hypochordal lobe used in this way opposes lift, the net effect of the whole tail is still to produce lift. Apparently muscular control of the lower lobe allows precise orientation of the direction of thrust from the tail. In higher teleosts, the fin is superficially symmetrical top and bottom (homocercal). Internally, the vertebral column turns up sharply, but does not reach the hind edge of the fin as it does in elasmobranchs. On the dorsal side of the upturned vertebrae are a few epural elements probably derived from the neural processes. None of the major caudal fin rays arises here, and only a few minor rays above and ahead of the main caudal fin. Almost all the caudal fin arises from the lower surface of the upturned column, which may retain several centra or which may be reduced to a single rod, the urostyle. The hemal processes of the last few centra are greatly expanded, and together form a fanlike plate which retains several distinguishable elements (hypurals) in more primitive fish but which may become almost wholly fused to each other and to the urostyle in higher fish. On the dorsal side of the

1.

FORM, FUNCTION, AND LOCOMOTORY HABITS

67

upturned terminal vertebrae are a few epural elements probably derived from the neural processes. Anteriorly, the caudal skeleton grades almost imperceptably into the backward sloping hemal and neural processes from the vertebrae of the caudal peduncle. The caudal fin rays of teleosts are structurally similar to the dorsal and anal fin rays. Their split bases are widened and flattened to grasp the hind edge of the hypural bones. The major rays (whose number is largely fixed and characteristic for each species) bifurcate dorsoventrally several times as they approach the trailing fin edge; the series of minor rays, which do not bifurcate, are continuous with the major rays and run forward onto the upper and lower surfaces of the peduncle (Fig. 7D). The caudal rays may all be nearly horizontal, or they may fan out very widely, particularly in caudals with high aspect ratio. The rays of the upper and lower lobes may be separated by a slight gap at their bases, continuous with a gap in the underlying hypural elements (Nursall, 1963, Fig. l), facilitating the operation of the two lobes as separate units (Gosline, 1971). In most teleosts the bases of the rays near the center of the caudal fin usually overlap the hypural supports only slightly, providing a hinge along which the fin can bend somewhat from side to side. Nursall (1963) illustrates the joint, which also allows the rays to rotate in the sagittal plane (for expanding of the “fan”). Rays toward the upper and lower edges overlap their supports more. The result is a fin stiffer at its margins than its center; typically in the subcarangiform mode of swimming, as the tail is swept sideways the tips lead and the center bellies out behind. In contrast, tunas have extremely stiff and close packed caudal rays whose bases overlap the hypural base greatly and permit almost no bending there. The rays reaching the extreme upper and lower tips of the tuna tail are more than twice as long as the center rays, and they can be bent more. Consequently, during the very powerful sideways thrusts, the center leads and the tips follow. In those teleosts which swim primarily with the pectoral, dorsal, or anal fins (Sections II1,C and II,D), the propulsive importance of the caudal fin is often reflected in the number of the caudal fin rays. These are typically 19 in soft-rayed fishes and 17 in spiny-rayed groups, but in groups relying on the dorsal, anal, or pectoral fins as principal propellers the caudal rays may be reduced to 12-15 (Zeiformes, Labridae, Scaridae), 9-12 (most Tetraodontiformes), or even none in Molidae (Marshall, 1971). The caudal musculature of teleosts ranges from a comparatively simple series of extensions from the posterior myomeres onto each fin ray base (Polypterus),to a complex of seven or more sets of muscles in

68

C. C. LINDSEY

higher fishes. Nursall (1963) provides a valuable series of drawings and descriptions of the caudal muscles of the perciform Hoplopagrus guntheri, and a table of synonyms.of muscles names by earlier authors. Other useful references on the caudal muscles are Schmalhausen ( 1913),Greene and Greene ( 1914),Grenholm (1923),and Nishi (1938). The posterior myomeres of both the red muscle [lateralis superficialis of Greene and Greene (1914)l and white muscle (lateralis profundus) may be continued backward as tendons which insert on the caudal fin bases (DT and VT in Fig. 7D). Between these, at the surface at the horizontal septum, is a prominent deltoid flexor tendon [the terminal tendon of lateralis superficialis, of Greene and Greene ( 1914)l which originates in tendons of the myosepta and inserts on the central caudal rays (T in Fig. 7D). All the foregoing seem to be involved primarily in lateral flexion of various groups of caudal fin rays. Deeper than the deltoid flexor is the hypochordal longitudinal muscle, running up and back from a prominent projection on the terminal hypural called the hypurapophysis. The muscle inserts on the upper three to five caudal rays. Marshall (1971) believes that the hypochordal pulls these rays sideways and down to produce a caudal twist which may be important in the locomotion of some groups. Three more sets of muscles lie within the fin base: the deep dorsal flexor, the deep ventral flexors, and the dorsal adductors. These originate variously on the posterior vertebrae, epurals, .or hypurals, and insert on caudal ray bases. They can adduct, abduct, or flex particular groups of rays. Finally, a thin mat of interradial muscles knits together the proximal portions of the major caudal rays. These are of three sorts, forming fans overlying each other diagonally and connecting neighboring rays at various angles. Red fibers occur in the interradials, although the rest of the caudal musculature in Hoplopagrus is white (Nursall, 1963). Interradials can probably contract the fin, and produce some lateral bending of individual rays. Some of the fine caudal fin movements observed by Bainbridge (1963), and the caudal undulation described in Section III,D,4 may be produced by the interradial muscles.

4. PECTORALFINS The base of the elasmobranch pectoral fin is usually two flat cartilage plates, the mesopterygium and metapterygium, which articulate with the pectoral girdle. In skates a third member, the propterygium, supports the anterior of the fin; in some sharks only the metapterygium persists. Continuous with the distal margin of the pterygia is a row of

1.

FORM, FUNCTION, AND LOCOMOTORY HABITS

69

triserial radial rods. The pattern of segmentations and fusions of pterygia and radials varies, but the whole forms a tightly knit mosaic, rigid proximally and somewhat flexible distally, which extends well out into the fin. As in the dorsal and anal fins, the outer web is supported by closely spaced unbranched unsegmented ceratotrichia, completely covered by skin and muscles so as to be externally invisi-

ble. The muscles of the pectoral fin in the shark Scyllium consist of a deep and a superficial abductor, and a deep and a superficial adductor, each originating on the girdle and inserting on the fin supports; a tongue from the lateral musculature also recurves ventrally to insert on the lower fin base (leDanois, 1958). The pectoral fins are less mobile in sharks than in most teleosts, but are more mobile than the dorsal or anal. The articulation with the pectoral girdle principally allows the fin to swing forward and back in its own plane. Th.e leading edge turns slightly down as it swings forward, so that the amount of lift provided by the fin alters (Alexander, 1967). The low and horizontal position of the pectorals allows them to act as hydroplanes, which counteract the negative buoyancy and the pitch produced by the heterocercal tail. Marshall (1971) makes the interesting observation that in a few mesopelagic sharks which have achieved nearly neutral buoyancy the pectorals have shifted upward and rotated, coming to resemble the paddlelike pectorals of higher teleosts. Except perhaps in these aberrant forms, sharks do not use their pectorals as brakes the way teleosts do, and sharks are incapable of making sudden stops. In skates the pectorals are highly modified and become the principal locomotor organ (Section III,C,4). In most teleosts the pectoral fin base is a row of hourglass-shaped radials (or actinosts) which articulate with the vertical posterior margin of the scapula and coracoid of the pectoral girdle. Their number is most often four, but may be higher or lower. Unlike the elasmobranch radials, they do not extend out into the fin, which is supported entirely by bony rays (lepidotrichia). These articulate along the distal margins of the radials, their split (and curved) bases sometimes encompassing'a row of small round distal cartilages borne by the radials. The ray shafts are usually jointed, and they may branch toward the tips. The anterior (dorsal) edge of the fin is usually the stiffest. The radials are progressively longer from top to bottom, and the fin therefore tends when it is extended to swivel about the anteriormost ray as an axis (Gosline, 1971). The pectorals of the more primitive teleosts are low on the body and largely horizontal, like those of sharks. In higher teleosts the pectorals have moved up on the sides,

70

C. C. LINDSEY

and their bases have changed to an essentially vertical alignment. The axis of rotation when the pectoral is extended is usually maintained, so that the outer ray swings out ahead of the lower (tending to push the head upward). A nearly vertical hinge for the fin base allows fanning with the pectoral in order to maintain a stationary position. It also allows fish which habitually rest on the bottom to clap their pectorals back against the body to accelerate from a standing start (Gosline, 1971). In most fish with highly placed pectorals which are used as paddles, the lateral line curves upward as it approaches the front, presumably to avoid water disturbances from the pectorals (Marshall, 1971). The fins can also now be used as a brake; the upward shift of the fin origin ensures that braking by extending the pectorals will impart less of a turningcouple to the head. The muscles operating the pectoral fin of the salmon were identified by Greene and Greene (1914) as follows: abductor superficialis running from the coracoid of the pectoral girdle to the ventral half of each ray base, bends fin forward and down and closes the rays (AS in Fig. 7F); abductor profundus, from the coracoid to the ray base inner margins, draws fin down; extensors, from the cleithrum to the outer surface of the first ray, spreads fin out in the horizontal position; adductor superficialis and adductor profundus, both from the coracoid-cleithrum junction to the ray bases, draw fin back against body; and interfilamentous, a network of delicate fibers running diagonally between the ray bases on their ventral surface, capable of closing up the rays. Variations on this basic scheme of pectoral fin musculature have been discussed by Grenholm (1923), Shann (1924), and Sewertzoff (1926). The anatomy of particular species, including some highly modified ones, is described by McMurrich (1884), Howell (1933), Nawar (1955), Samuel (1961), Schneider (1964), Alexander (1965b), and Tilak and Kanji (1967). Correlations between relative development of the pectoral musculature and mode of life are shown by Ganguly and Nag (1964), Keast and Webb (1966), Emery (1973), and Horn (1975). Some references to occurrence of red muscle in the pectoral fins are given by Johnston and Goldspink (1973b) and Rosenblatt and Johnson (1976).

5. PELVICFINS The contribution of the pelvic fins to locomotion is usually minimal. They seldom contribute to forward propulsion, but may serve as

1.

FORM, FUNCTION, AND LOCOMOTORY HABITS

71

stabilizers and maneuvering vanes. They are frequently modified for nonlocomotory purposes, and become copulatory organs, defensive spines, or adhesive discs. In quite a few groups the pelvic fins are absent. The pelvic girdle of elasmobranchs is generally a single transverse cartilaginous bar. It bears a backwardly directed basipterygium on either side. The lateral margin of the basipterygium carries a row of triserial radial rods, and these in turn support the closely spaced ceratotrichia. The girdle is imbedded in the ventral myotomes. In advanced sharks there are more or less distinct muscle bundles evidently derived from myotomes and running between the basipterygium and the rays. In skates there is a rectus muscle attached to the pelvic girdle and to the first caudal vertebra, and a pelvic subspinal muscle (leDanois, 1958). In male elasmobranchs there are additional pelvic muscles associated with complex axial and terminal cartilages projecting back from the basipterygium. In lower teleosts the two halves of the pelvic girdle are fused in the midline, and there are a few skeletal nodules, equivalent to the radials in the pectoral base, along the posterior margins where the rays insert. The innermost is the largest. This allows some rotation of the fin about the outermost ray as an axis. In higher teleosts the girdle halves are usually separate, there are no radials, and the fins can be extended or retracted only along an essentially single plane (Gosline, 1971). There is a concomitant reduction in the number of pelvic fin rays. In higher fish where the pectoral fins have moved upward the pelvics often move forward, and the pelvic girdle becomes strongly attached by ligaments to the cleithra of the pectoral girdle. Forms having such a firm pelvic base may develop strong pelvic spines, which may articulate with the girdle through a joint allowing movement only in one plane. In the anterior position, the locomotory role of the pelvic fins is largely to provide braking or turning, and to counteract the pitch produced when the pectoral fins are extended. The muscles operating the pelvic fin in salmon include the two infracarinal muscles already referred to (Section V,B,3), and the following: abductor superficialis (AB in Fig. 7F), bends the fin downward away from the body, and closes up the rays; abductor profundus, bends the fin outward; adductor superficialis, draws the fin inward; adductor profundus ( A D in Fig. 7F), rotates the fin inward and spreads the rays. The foregoing muscles originate at various places on the pelvic girdles and insert on appropriate parts of the fin ray bases. The anatomy of pelvic fins was reviewed by Sewertzoff (1934). Sheldon (1937)presented an extensive study of catfish pelvic girdles.

72

C. C. LINDSEY

The morphology of the pelvics is highly modified in forms with adheSome peculiar insive or other specialized pelvics (Bertin, 1958~). stances are described by Tyler (1962) and Saxena and Chandy (1966a,b).

6. INNERVATION OF THE FINS

All the median and paired fins are innervated by spinal nerves from body segments in the vicinity of the fin. Because the fin rays and their muscles are almost always more tightly packed than the adjacent body segments, there is usually an anastomosis of the nerves, and the nervous supply to a particular fin muscle cannot be allocated to a particular spinal nerve. Even in the dorsal fin of the stickleback Gasterosteus aculeatus, in which a single set of fin muscles lies opposite each vertebral segment, branches of the spinal nerves anastomose to form a longitudinal collector. It has been shown experimentally that excitation of a single spinal nerve can cause contraction in two radial muscles in the pectoral fin of a dogfish, and in six to eight in a skate (Bertin, 1958d). The pectoral muscles of the ray Rhina are served by a brachial plexus receiving contributions from the first ten spinal nerves plus three occipital-spinal and three occipital nerves. The pelvic plexus contains a large collector nerve with contributions from spinal nerves of the twenty-fifth to thirty-eighth segments (Bertin, 1958d).Similarly in the sturgeon Acipenser the pelvic fin is served by a plexus from thirteen spinal nerves. Longitudinal collector nerves connecting the spinal nerves and fin muscles have been found under the dorsal and anal fins of catfish (McMurrich, 1884) and of Polyodon (Danfortb, 1913). The path of nerves into the fin muscles is figured b y Howell (1933) and Schneider (1964). Nishi (1938)summarizes the innervation of fish fins. Nursall (1963) provides a diagram and chart showing distribution of spinal nerves from the last five vertebrae into the various caudal muscles of Hoplopagrus. For literature on the experimental study of fin innervation and coordination see Roberts ( 1969b).

7. FUNCTIONAL TOPOGRAPHY OF

FINS

Among the more than 20,000 species of fishes, there is spectacular variation in the number, shape, size, and position of the fins. Codfish

1. FORM, FUNCTION, AND LOCOMOTORY HABITS

73

can have ten paired and median fins; some eels have virtually none. Some fins are soft-rayed and some are spiny, and these perform different roles. A gymnotid anal fin may have over 500 rays, while a scombrid finlet contains only one. Many different and sometimes conflicting selective forces must affect fin development and topography, but some general trends are observable when large numbers of species of widely different habit are compared. Magnan (1929) measured the size and position of the fins, along with many other parameters, in 171 fish species. Several characters showed trends with respect to the estimated speeds of the fish. Greenway (1965) presents calculations based on Magnan’s data. The relative surface area of dorsal, anal, and pectoral fins all decrease regularly between “very slow” and “very fast” swimmers. The dorsal and anal fins also tend to shift toward the posterior. The caudal fin shows no clear trend in surface area, although its shape varies markedly (see Section 111,B); the aspect-ratio increases strikingly between slow and swift species (Aleev, 1963). Kramer (1960) shows diagramatically how typical tail shapes alter from long and low in abyssal forms to short and high in pelagic forms. The decreased area and backward shift of the dorsal and anal fins with increasing speed are probably related to increasing emphasis on the caudal propellor, which reaches its acme in thunniform locomotion. Webb ( 1975) discusses the hydrodynamic consequences of fins (or hydrofoils) having different geometric plans. Kramer ( 1960) measured the performance in a working model of cutout “fins” of various shapes. The outline shapes of different types of fins were categorized by Magnan (1929). Gregory (1928), as part of an ambitious analysis of body forms of fishes, distinguished twenty-four categories of shapes and positions of the fins, under each of which a species could be placed in one of three or four grades. Among adult fish, broad and rounded pectoral, dorsal, anal, and caudal fins are typical of slower swimmers. Narrow, falcate pectorals and median fins often accompany high aspect-ratio tails in swift pelagic species. Changes in fin shape and size also occur during ontogeny; simple rounded fins typify the early stages even in forms which as adults acquire angular fins. Aleev (1963) suggests that rounded fins are adaptive to anguilliform movement under conditions of low Reynold’s numbers, and hence characterize small fish, slow fish, and, he states, also fish adapted to colder (i,e., more viscous) water.

74

C . C . LINDSEY

The effects of fin position on swimming and maneuvering have been discussed by Harris (1953).The positions which fins occupy on the body will depend on the allocation of the functions of generating forward movement, stabilizing that movement, changing direction, and stopping. One fin can fulfill more than one of these functions (and of course, other nonlocomotory functions such as display or defense). In anguilliform swimmers the requirements are fairly well distributed along the body, and the fins may be little differentiated. But in carangiform and thunniform swimmers the propulsive thrust is concentrated in the caudal fin, and the body is stout and rigid; the fins are more specialized, and each is most effective if located in a particular region of the body. Aleev recognizes four functional regions: (1) an anterior zone of rudders and lifting surfaces, (2) a zone of keels, (3)a zone of stabilizers, and (4) a posterior zone of rudders and locomotor organs. The first zone involves the pectoral fins, and in higher fins the pelvic fins which are directly beneath the pectorals. The farther these lie ahead of the center of gravity the more effective they can be as rudders. The second zone may involve part of the dorsal fin, and of the anal fin, if they are situated far enough forward, and the pelvics if they are posteriorly placed. The third zone involves parts of the dorsal and anal fins behind the center of gravity; their effectiveness as stabilizers is greater when they are farther back, except that they should receive minimal lateral displacement from the propulsive waves. The fourth zone involves the caudal fin, and also sometimes the dorsal and anal fins if they are well back. The relative lengths of these zones differ widely. For example, zones 2 and 3 are farther forward in a tuna than in a dogfish (Fig. 1). During ontogeny the positions of the fins often shift, the anterior of the first dorsal often moving forward and the posterior of the second moving back, both in elasmobranchs and in teleosts. Aleev (1963) has made extensive calculations of the dynamic stability resulting from fins of different sizes variously placed with respect to the center of gravity. He has also examined the relative body height, the location of the greatest height along the body, and the cross-sectional shape, all of which vary widely between species and sometimes during ontogeny. His conclusion is apt for the diffuse subject of functional topography of fins: It is clear that a certain change in the position and structure of a particular fin will affect the position and structure of the other fins and other hydrodynamically sensitive structural features merely through the resultant change of the total

1.

FORM, FUNCTION, AND LOCOMOTORY HABITS

75

activity of all the fins and body parts. This example clearly illustrates that any morphological features of an organism, in both ontogeny and phylogeny, are connected with the environment not as independent parts of the structure but as parts of a whole. Here the effect is felt of the unity of all parts of the organism, in both the functional and morphological sense.

VI. LOCOMOTORY HABITS O F WILD FISH Measurements of locomotion and metabolism of fish in the laboratory are steadily improving in precision, but the relevance of such data to wild fish is largely unknown. The gap will have to be closed before energy budgets can be drawn up for natural populations, in which food intake, metabolic costs, and growth can be balanced. Most of the discussion in this volume is of necessity based on laboratory observation, but first, knowledge will be summarized concerning the locomotory habits of fish in natural surroundings. Although current information is diffuse and anecdotal, substantial improvements can be expected soon through advancing technology. Methods of recording activity of fish, in the wild or, more often, in confinement, are discussed in Chapter 2.

A. Records of Long Distance Movements The best evidence concerning long distance movenients of fish comes from the recapture of individuals which previously had been marked in a distinctive manner at a known time and place. While there are important weaknesses in this type of data, tag recaptures offer positive evidence of movement, and allow the calculation of minimum distances covered and minimum rates of travel. A few examples of long distance movements are shown in Table IV, arranged roughly in descending order of the sizes of the fish involved. See also Table I in Chapter 2 for more detailed observations. The distances covered by some fish are spectacular. Two large bluefin tuna Thunnus thynnus tagged at Cat Cay, Florida, were recaptured 7800 km away at Bergen, Norway, 118 and 119 days later (Mather, 1962). The same species in the Pacific has crossed from west to east-Japan to Baja, California (9700 km, Table IV), or east to west-San Diego to Japan (8800km, Clemens and Flittner, 1969),and Guadalupe Island to Japan (10,750 km, Orange and Fink, 1963). By tagging albacore Thunnus alalunga. Clemens ( 1961) obtained evi-

Table IV Some Records of Long-Distance Movements by Tagged Fish Species Thunnus thynnus (bluefin tuna) Zstiophorus platypterus (sailfish) Hippoglossus stenolepis (Pacific halibut) Tetrapturus albidus (white marlin) Galeorhinus zyopterus (soupfin shark) Oncorhynchus tshawytscha (chinook salmon) Squalus acanthias (spiny dogfish) Gadus morhua (Atlantic cod) Thunnus alalunga (albacore) Salmo salar (Atlantic salmon) Salmo salar (Atlantic salmon) Thunnus thynnus (bluefin tuna) Oncorhynchus nerka (sockeye salmon)

At tagginglat recapture.

Distance (km) 7800 F1a.-Bergen 4440 Windward I.-Pensacola 2370 Alaska-Cape Blanco 5000 Bahamas-off mouth Amazon R. 2040 Calif.-Q .C.I. 1610 Q.C.1.-Ore. 8700 Wash.-Honshu 2100 W. Greenland-Iceland 8534 Calif.- Japan 1100 Norway 4270 Devon-Greenland 9700 Japan-Baja Calif. 1670 N. Pac.-Bristol Bay

Time (days)

Rate (cmlsec)

Length" (cm)

WeighP (kg)

118

77

-

ca. 170

98

52

228

20.9

175

16

-

-

516

11

-1185

103

23

1401145

-

Herald and Ripley (1951)

60

31

-

-

Manzer (1946)

94-

-

Holland (1957)

2699

3.7

16/24

Author Mather (1962) Mather et al. (197413) Manzer (1946) Mather et al. (1974a)

147

17

86

-

Meyer (1965)

196

51

75376

-

Clemens (1961)

11

116

85

-

554

9

173

14

Dahl and Spmme (1936, cited in Bainbridge, 1958) Allan and Bulleid (1963)

323

35

36/68

1.U7.3

Clemens and Flittner (1969)

35

55

-

-

Neave (1964)

1.

FORM, FUNCTION, AND LOCOMOTORY HABITS

77

dence that they move in schools which probably make the two-way trip to California to the mid-Pacific or to Japan and then back to California. Other species of scombroid fishes may also migrate thousands of kilometers; Table IVgives a few examples. Diverse groups of fish, including clupeoids, salmonids, cods, flatfish, and sharks, may move long distances; a spiny dogfish tagged off the Washington coast turned up in Japan 8700 km away (Table IV), and the same species has undertaken various transatlantic movements (Holden, 1967). Many more examples could be given from tagging returns, and many species which have not been tagged may be suspected to migrate. The European eel Anguilla anguilla probably migrates over 6000 km between its freshwater feeding streams and its spawning ground south of Bermuda, but there are no tagging data. All the distances quoted are the shortest possible distance by water between points of marking and recapture. The dogfish which crossed the Pacific may have covered substantially more than the 8700 km direct distance to Japan, if it traveled at its accustomed depth north around the Continental Shelf (Holland, 1957). Oceanic currents may also have deflected some of the courses followed. Calculated rates of travel of tagged fish may be appreciably hastened or slowed b y water movements. Mather (1962) points out that the two bluefin tuna swimming from Florida to Norway may have been ,assisted by the Gulf Stream and the North Atlantic Drift. But many of the migrations cannot be attributed to water movement. Bluefin tuna and albacore cross the Pacific in both directions. So too tagged Atlantic cod have been recorded moving between Greenland and the Icelandic spawning grounds in either direction; T h i n g (1937) noted however that the swim from Iceland to Greenland (roughly 2000 km), going with the current, takes only 90 days, while the reverse journey takes at least 164 days. The fastest daily rate of cod in West Greenland (length roughly 90 cm) was 32 cm/sec, which is close to a daily rate of 28 cm/sec 'recorded earlier for cod leaving the spawning grounds at Lofoten, Norway. Atlantic salmon Salmo salar also move both ways in the North Atlantic; some tagged fish have moved from several points on thy British Isles to Greenland (Menzies and Shearer, 1957; Swain, 1963; Allan and Bulleid, 1963); others have moved to Greenland from New Brunswick (Kerswill and Keenleyside, 1961). Oceanic circulation may have modified these migrations, but it cannot be their sole cause. The question of dependence on water currents in ocean migrations by Pacific salmon Oncorhynchus is ably discussed by Neave (1964),

78

C. C. LINDSEY

who synthesized results of massive tagging experiments in the North Pacific. His conclusions are as follows. The velocity and direction of recorded currents are often entirely inadequate to account for the rate of travel. In the central and eastern parts of the Gulf of Alaska, where current velocities are commonly of the order of 1-4 miles per day, the minimum travel speeds of tagged sockeye and pink salmon over distances of several or many hundred miles has frequently exceeded 25 and sometimes 40 miles per day. Moreover, these journeys were made in a wide variety of directions relative to known currents . . . . It is concluded that salmon are able to reach distant ocean areas which offer favourable conditions for survival and growth, and also to return from these areas, by relatively rapid journeys which are not closely controlled by currents.

Adult Pacific salmon of various species, usually with body lengths of roughly 50 cm, maintain average speeds of from 0.8 to 1.5 t/sec for many days or even weeks; one pink salmon 0. gorbuscha about 44 cm long achieved an average of 90 cm/sec (over 2 t/sec) for 6 days (Hartt, 1966; Shepard et al., 1968). The same order of magnitude of travel rates (in absolute terms) has been recorded for a soupfin shark Galeorhinus zyopterus (length 142 cm) which averaged 43 cmlsec over 2 days (Herald and Ripley, 1951), and for a striped marlin Tetrapturus audax (length unknown) which averaged 75 cm/sec over 90 days swimming from Baja, California, to Hawaii (Squire, 1974). A strikingly higher rate was recorded, over shorter periods, in schools of albacore Thunnus alalunga (Clemens, 1961). One school covered 428 km in 1 day, achieving the average rate of nearly 500 cm/sec. (The rates are calculated from recoveries at different times of fish all tagged in one school at the same time; they rest on the assumption that the fish remained together as a school, which is probable but not certain.) If these fish were about 80 cm long, the highest rate was about 6 tlsec, well above the 3 e/sec usually considered as normal cruising speed in fishes. Possibly harassment by the fishing boats which made the tag recaptures produced abnormally high speeds, comparable to the high speed of a skipjack tuna (5 t/sec) observed by Yuen (1970) after he had chased it a long way from its home bank (see Section V,B,3). Daily rates of travel, if calculated for a long period, may fall far short of the swimming speed displayed during the periods of each day when the fish was actively migrating. Sockeye salmon smolts migrating out of Babine Lake spent only about 8 hr per day traveling; much time was spent on feeding (Johnson and Groot, 1963).The daily period

1.

FORM, FUNCTION, AND LOCOMOTORY HABITS

79

allocated to active travel b y Pacific salmon in the ocean is uncertain, but they are known to feed extensively during migration; daily rate of movement might represent only a third of the actual active swimming speed (Neave, 1964). On this basis, maturing pink salmon (body length perhaps 44 cm) maintain travel rates of 46-55 km/day over many days, but if this is compressed into an 8 hr working day it represents swimming speeds of 155-180 cm/sec. Migration rates of young fish are harder to obtain because tagging is usually inappropriate. Even so, rough measures can be made of the time and extent of the seaward migration of young Pacific salmon, capitalizing on their known life cycles and strong homing tendency. Many half-grown fish have been captured in the open North Pacific and tagged. Subsequent recapture of these as spawning adults in Asian or North American streams established where they themselves had originally come from as fry; the data and distance from home at their first capture coupled with their known age enabled Neave (1964) to estimate their movement rates since first entering the ocean. Thus, over periods of from 9 to 12 months, young pink salmon must have moved at daily rates of about 10.7 cmlsec, young sockeye at from 3.2 to 6.5 cm/sec, and young chum at about 8.6 cm/sec. These rates agree with Johnson and Groot’s (1963) observations that young sockeye on their migration out of Babine Lake swam at daily rates of 6.5-13 cm/sec over a distance of 50-117 km. Although the absolute speeds of young salmon referred to above are small compared with those of adult salmon, their “specific speeds” (i-e.,body lengthdsec) are in the same range. In fact small fish tend to do better than large fish in their specific cruising speeds (Webb, 1975), as measured over short periods in the laboratory. Data on long distance travel given above are too rough for such a generalization concerning wild fish, but at least it can be said that, when related to body lengths, the performances of small fish seem to be at least comparable to those of large fish. Much information has been accumulated about fish ascending rivers to spawn. In the confines of the river, tagging and recaptures are relatively easy. The ground distance covered, and also the elevation attained, are known more precisely than in marine migrations. What is not known with any precision is the length of the water column through which the fish have passed in stemming the current. Hence speed relative to the water, and work done, are imprecise. In general, the rates of movement upriver by Pacific salmon resemble the rates described earlier for their migrations in the sea (Idler and Clemens,

80

C. C . LINDSEY

1959; Osborne, 1961; see also Chapter 2). Swimming against the flow of the river must require some additional outlay of energy, but just how much is a matter of conjecture.

B. Short-Term Components of Long-Term Movements

1. UPSTREAM MIGRATION Distances of upstream fish migration have sometimes been calculated by adding to the ground distance covered a measure of the water column which has moved downstream during the time of the ascent. The calculation is illusory, since fish are cunning at seeking paths with unknown although probably low velocities. It has even been suggested (Osborne, 1961) that migratory salmon may extract energy from turbulent velocity fluctuations of the river; the possibility has not been disproven. Observations on fish moving up rivers show that progress is irregular. During daylight, sockeye and coho salmon ascending a river progressed through slow stretches (less than 1.0 m/sec) in steadily swimming schools (Ellis, 1962, 1966). In somewhat faster water (1.0-1.5 m/sec) they interspersed their progress by rest periods. To navigate the rapids (more than 1.5m/sec), they abandoned their schools and moved individually, alternating between bursts of high activity and short holding periods. Paths chosen by all fish represented only a small part of the total water space available. Migration through pools and past obstructions showed a die1 pattern. Some migration occurred at night. Ellis (1966) gives estimates of their burst and sustained speeds, but it would not be practical under such conditions to measure water velocities and elapsed time along the precise path followed by each fish. Upstream migration of sturgeons also shows complex patterns. Acipenser gueldenstaedti were found by Gayduck et al. (1971) to swim usually at a depth of 10-15 m (at 2-4 m above the bottom of the river), but they would occasionally rise and swim for 5-20 min only 1-2 m beneath the surface. Studies by Shubina (1971) on Acipenser stellatus in the Volga also demonstrated complex variation. The largest upstream-migrating fish rested for periods of 5-31 min; downstream-migrants did not rest, but moved slower than the flow of the river. Fish ascended the shallow east bank, and descended on the opposite side. The largest fish migrated fastest. Both in upstream and downstream migration, the distribution of males and females in the flow was different, and the distribution of fish of different sizes over

1.

FORM, FUNCTION, AND LOCOMOTORY HABITS

81

the stream bed was different. Such data defy simple calculations of distance covered or energy expended. Much of the literature on upstream migration of salmonids is reviewed by Banks (1969). Additional sustained cruising speeds of migrating fish, mostly measured over shorter time periods, are given in Chapter 2.

2. MIGRATIONIN

THE OCEAN

The problems of recording motion of a swimming fish with respect to the surrounding water are less in a nonturbulent ocean or lake than in a river. An eel, about 80 cm long, outbound in the North Sea on its spawning migration, was tracked for 14.25 hr, during which time it covered 25 km (Tesch, 1972).It swam in midwater so long as the depth exceeded 20 m. Its speed through the water did not vary greatly, with an average of about 51 cm/sec and a maximum for 1 hr of 77 cm/sec. Its ground speed was of course greater when moving with the tide than against it, but the effects of,tidal drift on migrant eels apparently were largely cancelled out by alternating directions of flow. Less regular locomotory habits have been observed in flatfish. A 43 cm plaice, Pleuronectes platessa, tracked for 15 hr in water 27 m deep (Greer Walker et al., 1971) spent 70% of the time off the bottom. During strongest tide flow it moved a ground distance of 7.41 km in 137 min, or 90 cm/sec (about the same speed and direction as the tidal current). At night its average height off the bottom increased.from 2 to 6.7 m. As the tide slackened, it settled on the bottom and buried itself. These direct observations are in agreement with earlier sightings of another flatfish species, the sole Solea vulgaris, often seen at the surface during night. deVeen (1967) analyzed fishermen’s reports of such sightings, usually made in March, April, and May when the spawning migration is under way. H e suggested that sole seen at the surface are those at the top of the main body of migrants which have left the bottom to use the tidal current for passive transport. Those seen at the surface are nearly always still, and drift with the tide. “It is not known how the diurnal vertical movements of the sole are synchronized with the tidal cycle so that they leave the bottom, at night, on an easterly tide” (Jones, 1968). The rise off the bottom at night, observed directly in the tagged plaice and inferred from sightings in soles, is in agreement with aquarium experiments on plaice by Verheijen and deGroot (1966),and on sole by Kruuk (1963); these flatfish were active at night and dug into the bottom during daylight. There is a concomitant die1 rhythm in trawl catches of sole and other species (Woodhead, 1966).

82

C. C. LINDSEY

From the viewpoint of calculating locomotory output for these types of migration, much of the lateral displacement is apparently passive, but this must be achieved by expending energy in vertical movement since the flatfish are heavier than water. To compute an energy budget, one would need to know how much work is required simply to remain off the bottom.

3. MIGRATIONOF SCOMBROIDS Large scombroid fishes range widely, often over water so deep that contact with the bottom is lost. A Pacific blue marlin 365 cm long, tracked for 22.5 hr, spent half the time within 10 m of the surface, one-sixth at from 10 to 30 m, and one-third at depths over 30 m, once diving to 80 m (Yuen et al., 1974). The depth selected showed no die1 pattern. During the period, speeds ranged from 31 to 228 c d s e c , with an average of 83 cm/sec. Another marlin remained at depths of 115185 m during a 5.5 hr tracking period, but was probably suffering from effects of capture. Longer tracking periods of skipjack tuna in Hawaiian waters are reported by Yuen (1970),who maintained contact with one fish (length 44 cm) for 12 hr and another for 7 days. His summary reads as follows. The following picture of the behavior of skipjack tuna that are associated with banks can be drawn from these results. They have a general daily pattern. They usually spend the day at the bank, where they swim to and fro and are away from the surface a good part of the time. This type of swimming is probably associated with searching for food and feeding. Later in the day, within a couple of hours of sunset, they leave the bank and swim with few changes in direction until approximately 2:00 AM, when they seem to adopt a more erratic swimming pattern. Although they leave the bank by different routes from day to day, they are usually back at the bank by sunrise. They are presumed to be close to the surface throughout the night. The repeated returns to the same spot by the fish at Kaula Bank after journeys of 25-106 km by various routes imply that skipjack tuna can navigate. Their consistent arrival times suggest that they have a sense of time. It is interesting to note that from 3:00 AM to 600 AM of the first morning after it was tagged, when it was unusually far from the banks [having been pursued for 28 km before tagging], the tuna at Kaula Bank averaged 8 kmlhr (223 cmlsec), seven times its average speed for that time of day, as if it were compelled to arrive at the bank by a certain time.

Tracking of a fish carrying an acoustic tag usually involves fixing the position at regular intervals, and calculating speeds in terms of time elapsed and straight-line distance between successive fixes. Calculated speeds are therefore minima, since the method is open to the same objection, on a smaller scale, as is the determination of migration speeds from tag recoveries: The fish may very well have not swum in a

1. FORM, FUNCTION, AND LOCOMOTORY HABITS

83

straight line, and hence may have swum farther and faster than shown by its net movement. There is reason to believe that swimming behavior of many wild fish includes important fine variations which may radically alter their energy expenditures. Fish heavier than water may alternate short periods of active swimming with periods of downhill “gliding.” (The technique is not available to very small fish because of their relatively high drag and small momentum.) The energy saving could in theory amount to 50% over a given horizontal distance (Weihs, 1973a). Not only scombroids, but sharks, rays, flatfish, and others may be suspected of employing two-stage swimming modes. The subject of such locomotion in negatively buoyant midwater fishes is explored in Chapter 4.

C. Activity Cycles in Wild Fish Levels of swimming activity alter rhythmically in many kinds of fish, timed to tidal, diel, or seasonal cycles. Locomotory changes may be spectacular in species which undergo long migrations. Young socke y e salmon, after a year or more of lake residence in Babine Lake, suddenly switch to a well-oriented directional swimming at speeds of 20-30 cm/sec (body length 8.3 cm), traversing 100 km of the lake and then proceeding down-river to the ocean (Johnson and Groot, 1963). The same species, homeward-bound from the high seas 2 years later, may cut across prevailing water currents for hundreds of kilometers at average speeds of 58 cm/sec (body length about 65 cm) (Neave, 1964). Increase in swimming speed are common at spawning time in many other species. They are particularly evident when reproduction involves the ascent of swift streams. Levels of locomotor activity also alter with season independent of reproduction. Swimming speed of yellow perch in Lake Mendota was much higher in summer than in winter, and was linearly related to water temperature (Hergenrader and Hasler, 1967). Temperature seems to be the dominant control of seasonal changes in activity (Andreasson, 1969; Gibson, 1969), although photoperiod may also be effective (Woodhead and Woodhead, 1955; Olla and Studholme, 1972). Tidally controlled cycles in movement of fish in the littoral zone have been reviewed by Gibson (1969). Die1 changes in swimming activity have been found in many species of fish (reviewed by Woodhead, 1966). Some species, such as the flatfishes, are largely nocturnal in their activity. Others, such as the pelagic clupeids, are much

84

C. C. LINDSEY

more active at dusk and at dawn. Activity peaks at dawn and dusk occur in freshwater (Spencer, 1939; Davis, 1964) as well as in marine species (Gibson, 1969; Stickney, 1972). Many sight-predators are active only in daylight. Diurnal-nocturnal changes are particularly striking in the variegated fish fauna of coral reefs (Hobson, 1965,1968,1973,1974; Collette and Talbot, 1972; Emery, 1973). The changeover between day-active and night-active fish groups is also clear in many freshwater communities. Lissmann (1961b) found in South American rivers that numerous gymnotid fishes, which have electrosensory systems, hid quietly in vegetation during daylight, but at night they swam actively throughout the open water, emitting a chorus of electrical signals. Temperate freshwater communities may also show exchanges between day and night occupants, with complex vertical exchanges in the open water, and onshore-offshore exchanges, between different species (Northcote et al., 1964) and sometimes between young and adults of the same species. Young brown bullheads Zctalurus melas are active by day with peaks at twilight, while adults are active at night (Darnel1 and Meierotto, 1965). The diel changes in locomotor output of individual fish may therefore be pronounced. The immediate stimulus seems usually to be light intensity, but it is difficult to allocate the causes of diel activity patterns in wild fish among pursuit of food, avoidance of predation, response to light or .other exogenous factors, and endogenous rhythm. Die1 vertical migration is widespread among diverse kinds of fish. Although most fish are reported to be deeper during daylight (Yuen, 1970; Cushing, 1973; Badcock, 1970), some show the reverse diel pattern (Northcote et al., 1964; Hasler and Villemonte, 1953). Vertical migration may be important in bringing about lateral transport from water currents. Examples have been given of flatfish rising off the bottom into a tidal flow destined to carry them toward their spawning ground. In many other species, too, diel vertical migration must expose them to changes in the actions of water currents. In fact, by appropriate vertical movements, fish in some regions might achieve most of their horizontal migratory movements by passive transport. Jones (1968),however, cautions that for no species are there data supporting the drift hypothesis which could be regarded as critical or conclusive. Vertical migrations of fish must be achieved at the cost of some energy expenditure. The daily migration of small mesopelagic fish from the deep scattering layer involves an upward swim of 500 m

1. FORM, FUNCTION, AND LOCOMOTORY HABITS

85

lasting for an hour or more (Marshall, 1971). The metabolic cost, following calculations for zooplankton, may be less than 1% of the organic matter of the body. The descent needs little energy as the fish is likely to be slightly heavier than water. Vertical migration costs in other groups will depend partly on their buoyancy. Species which can adjust their gas bladder volumes ride up and down with less cost, but adjustments have been too slow in those (shallow water) species which have been studied, and a gas bladder imposes limits on the range of rapid depth change (Jones, 1952). With respect to the adaptive advantages which must exist for vertical migrations in fish, Woodhead (1966) writes: It seems unlikely that the limited excursions from the sea bed of flatfishes, the nocturnal dispersions of single cod into midwater, the twilight migrations of large schools of many thousands of herring, both feeding and non-feeding, the extensive vertical movements of mesopelagic fish, and the deep scattering layer, can all be accommodated within a single simple theory. Vertical migration varies greatly in its nature and extent and is likely to hold different significance for different species of fish.

D. Schooling Schooling has been observed in over 4000 species of fish (Shaw, 1962). The advantages of schooling may relate to reproduction, predator avoidance, feeding, learning, or energy conservation (Breder, 1959; Radakov, 1973). Models have been developed, based on theories of search, which show that there may be an advantage in schooling for prey, and in some circumstances for predators (Cushing and Jones, 1968). Schools of fish can be very large indeed; Cushing and Jones (1968) refer to schools of spawning herring continuous in the Straits of Dover for 17 miles. Within large groups are smaller more homogenous formations which behave in closely concerted ways (Zuyev and Belyayev, 1970). Schools usually contain fish all of the same size, swimming at the same rate and changing direction more or less as a unit. The range in relative body lengths within a school is usually from 1.0 to 0.6 (Breder, 1965). There are no permanent “leaders.” The sustained swimming speed of fish in a school may be different from that of isolated individuals. Single fish have been reported as faster than those in schools by some authors (Breder and Nigrelli, 1938; Escobar et d.,1936; Ohlmer and Schwartzkopff, 1959; Schuett,

86

C. C. LINDSEY

1934) and slower by others (Hergenrader and Hasler, 1967; Kleerekoper et al., 1970).These observations were made under a variety of laboratory and field conditions, and on different species. Fish cannot school readily if an experimental chamber is too confining (Zuyev and Belyayev, 1970). Therefore, the performance of nonschooled specimens in the laboratory may give a misleading view of those species which school in the wild. Important hydromechanical advantages may arise from swimming in a school. Belyayev and Zuyev (1969) stated that the endurance of fish may be increased two to six times b y swimming in schools. Breder (1965) suggested that there may be effects of vortices in the wake of swimming fish on the fish following behind, but he did not analyze them. The Karman vortex sheet (the row of alternating eddies) may be used to lower the hydrodynamic resistance of followers; Zuyev and Belyayev ( 1970) offered experimental support of this hypothesis. Weihs ( 197313) demonstrated, by calculation, the advantages of fish swimming in a diamond-shaped configuration such as is actually observed; he calculated that there would b e reductions in relative speed of up to 30% between the best and worst positions a fish might adopt with respect to the wake of those fish ahead. Arguments for the reality of this effect are strong. The diamond pattern occurs in fish when they are swimming but disappears when they stop (Keenleyside, 1955). Dimensions of the diamonds observed in actual schools (van Olst and Hunter, 1970; Pitcher, 1973) agree with hydrodynamic calculations. From the hydrodynamic advantages of schooling which have been proposed, it follows that the lead fish gets no advantage and must work harder. Observations by Zuyev and Belyayev (1970) are in agreement; the horsemackerel swimming at the head of a grouping, although the same size as its followers, oscillated its caudal fins with up to 1.5 times the frequency. They concluded that the lead fish is in the least favorable circumstances, and it expends more energy on motion than the rest. Distance between fish (or packing density) is of interest to those calculating fish numbers from echo records of fish schools. Interfish distances in various genera, reviewed by van Olst and Hunter (1970), suggest that mean distance to the nearest neighbor for most pelagic schooling species is generally about half a body length. But Cushing (1973) notes from observations of interfish distances that larger fish move at a greater distance apart than might be expected merely from linear differences in size. H e suggests the increase is in proportion to their volumes. Perhaps an explanation may be found from hydrodynamic analysis.

1. FORM,

FUNCTION, AND LOCOMOTORY HABITS

87

E. Some Pitfalls in Locomotory Studies Physiologists can easily make fools of themselves when generalizing about wild fish on the basis of observing laboratory fish. I n most experiments the artificial constraints are so great that fish are likely to behave in a highly unnatural manner. Much of the repertoire of behavior patterns characteristic of a species in the wild may be physically impossible under laboratory conditions. The capture and handling of a fish, even with all possible gentleness, has been shown to sometimes impose extreme stress and even death, without any apparent morphological damage. Fright stress simply from movements of the observer may induce prolonged physiological changes in a fish exposed without cover in an experimental chamber. Appropriate cover may radically alter locomotory performance, but just what cover is appropriate may not be obvious to the experimenter. Jones (1956) found that minnows Phoxinus laevis in an open aquarium were active by day and inactive at night, but provision of a simple shelter completely reversed their behavior so that they remained in the shelter by day and swam about b y night. Flounders studied by Bregnballe (1961) swam both day and night in a tank, until sand was provided, whereupon they buried themselves all day and swam only in the dark. Light intensity, and also temperature, can reverse the response of young fish to running water (Northcote, 1962; Pavlov et al., 1968). Even the tone or color of the walls of the holding chamber is commonly observed to affect locomotor activities. This may be due to the light level which ought to be no more intense in the laboratory tank than that normally encountered by wild fish (Woodhead, 1966).Light intensities convenient to the physiologist may be acutely uncomfortable to the fish, for the latter, unlike the former, cannot close his eyes. Size of the chamber in which a fish is confined imposes constraints on its locomotory possibilities, particularly for large fish species. Although big fish (and aquatic mammals) can be kept alive and apparently healthy in aquaria, their locomotory repertoire is severely limited. Studies on captive tuna have been of only limited value in predicting their wild behavior (Tester, 1959).Even small species, if they habitually cruise for long distances in the wild, may exhibit aberrant locomotion and metabolism in a small aquarium. Only the smallest fish in the largest tanks can achieve top forward speed even for a moment, before they have to turn. Moreover, wild fish are likely to switch frequently from one activity level to another, perhaps including important alternation between swimming and gliding (Weihs,

88

C . C . LINDSEY

1973a). Their spectrum of activities is far wider than can be observed even in experimental tunnels or treadmills. The significant hydrodynamic advantages of swimming in a school are also denied by the size limitation of almost all experimental devices. Many of the foregoing sources of error vanish when observations are made on free rather than on captive fish. Extended underwater fish-watching is now possible with free diving gear, underwater habitats, research submarines, or underwater television. Detailed accounts can be compiled of the minuie-to-minute and daily locomotory patterns in natural fish communities. Such descriptions are essential precursors to drawing up metabolic budgets of wild fish, but they must be followed by quantification. Miniaturization now offers the possibility of transmitting many types of information from a wild swimming fish. One limitation is the restricted distance over which the underwater signals carry [currently from a few hundred meters, up to 2.3 km for large tags (Yuen, 1970)l. More important are the possible effects which the shock of catching the fish and attaching the transmitting device may have on subsequent behavior. Yuen et al. (1974) found that the trauma of capturing marlin in order to attach a tag usually resulted in aberrant behavior and even early death. The device is likely also to have an effect on locomotor activity after the fish is released. Tags fastened outside the fish will have a hydrodynamic effect. Tags placed either inside or out may affect buoyancy; Gallepp and Magnuson (1972) concluded that bluegills provided with transmitters might, if released over deep water without adequate time (300min) for buoyancy adjustment, sink and remain for a long time on the bottom. The further that technology of miniaturization develops, the greater is the temptation to load animals with measuring devices, with the ever present danger that the behavior then recorded is not natural. With due precautions, new technology is certainly going to allow great strides in understanding fish locomotion. But zoologists should never forget that they suffer from the same problem as quantum physicists; they cannot observe the course of nature without disturbing it.

REFERENCES Aleev, Y. G . (1963). “Function and Gross Morphology in Fish.” Izd. Akad. Nauk SSSR. (Transl. from Russian, Isr. Program Sci. Transl. No. 1773, Jerusalem, 1969.) Aleev, Y. G., and Ovcharov, 0. P. (1969). On the development of processes of vortex formation and the character of border layer in the movement of fishes. Zool. Zh. 48, 781-790.

1. FORM, FUNCTION, AND LOCOMOTORY HABITS

89

Aleev, Y. G., and Ovcharov, 0. P. (1971).The role of vortex formation in locomotion of fish and the influence of the boundary between two media upon the flowing pattern. Zool. Zh. 50, 228-234. Aleev, Y. G., and,Ovcharov, 0. P. (1973).The three-dimensional pattern of flow round a moving fish.]. Ichthyol. (USSR) 13,933-936. Alexander, R. M. ( 1965a).The lift produced by the heterocercal tails of Selachii. J . E x p . Biol. 43, 131-138. Alexander, R. M. (196513). Structure and function in the catfish.]. Zool. 148,88-152. Alexander, R. M. ( 1967). “Functional Design in Fishes.” Hutchinson, London. Alexander, R. M. (1968). “Animal Mechanics.” Sidgwick & Jackson, London. Alexander, R. M. (1969). The orientation of muscle fibres in the myomeres of fishes.]. Mar. Biol. Assoc. U . K . 49, 263-290. Alexander, R. M. (1971).“Size and Shape.” Arnold, London. Alexander, R. M. (1973). Jaw mechanisms of the coelacanth Latimeria. Copeia pp. 156-158. Allan, I. R. H., and Bulleid, M. J. (1963). Long-distance migration of Atlantic salmon. Nature (London)200,89. Amaoka, K. (1969). Studies on the sinistral flounders found in the waters around Japan-taxonomy, anatomy and phylogeny. J : Shimonoseki Unio. Fish. 18, 1-340. Andreasson, S . (1969). Locomotory activity patterns of Cottus poecilopus Heckel and C. gobio L. (Pisces). Oikos 20, 78-94. Arita, G. S. (1971). A re-examination of the functional morphology of the soft-rays in teleosts. Copeia pp. 691-697. Arnold, D. C. ( 1953). Observations on Campus acus (Brunnich), (Jugulares, Carapidae). Pubbl. Staz. Zool. Napoli 24, 152-166. Atz, J. W. (1962). Does the shrimpfish swim head up or head down?Anim. Kingdom 65, 175-179. Auerbach, A. A. (1967).The synchronous effector system of Casteropelecus. Diss. Abstr. B . 28, 1367. Badcock, J. (1970).The vertical distribution of mesopelagic fishes collected on the Sond cruise.]. Mar. Biol. Assoc. U.K.50, 1001-1044. Bainbridge, R. (1958). The speed of swimming of fish as related to size and to the frequency and amplitude of the tail beat.]. E x p . Biol. 35, 109-133. Bainbridge, R. (1961).Problems of fish locomotion. Symp. Zool. SOC.London 5, 13-32. Bainbridge, R. (1963). Caudal fin and body movement in the propulsion of some fish.j. E x p . Biol. 40, 23-56. Banks, J. W. (1969).A review of the literature on the upstream migration of adult salmonids.]. Fish Biol. 1, 85-136. Barets, A. (1961). Contribution a I’Btude des systemes moteurs “lent” e t “rapide” du muscle lat6ral des TelBostkeqs. Arch. Anat. Microsc. Morphol. E x p . 50, 91-187. Belyayev, V. V., and Zuyev, G. V. (1969). Hydrodynamic hypothesis of school formation in fishes. Probl. Ichthyol. 9, 578-584. Bennett, M. V. L. (1971a). Electric organs. In “Fish Physiology” (W. S. Hoar and D. J. Randall, eds.), Vol. 5, pp. 347-491. Academic Press, New York. Bennett, M. V. L. (1971b). Electroreception. In “Fish Physiology” (W. S. Hoar and D. J. Randall, eds.), Vol. 5, pp. 493-574. Academic Press, New York. Bergman, R. A. (1964). Mechanical properties of the dorsal fin musculature of the marine teleost, Hippocampus hudsonius. Bull. Johns Hopkins Hosp. 114,344-353. Bergmann, H. H. ( 1968). Eine deskriptive Verhaltensanalyse des Segelflossers (Pterophyllumscalare Cuv. u. Val., Chichlidae, Pisces).Z. Tierpsychol.25,559-587. Bernstein, J. J. (1970). Anatomy and physiology of the central nervous system. In “Fish

90

C . C . LINDSEY

Physiology” (W. S. Hoar and D. J. Randall, eds.), Vol. 4, pp. 1-90. Academic Press, New York. Bertin, L. (19584. Squelette axial. In “Traite de Zoologie” (P. P. Grass&,ed.), Vol. 13, Part 1, pp. 688-709. Masson, Paris. Bertin, L. (195%). Squelette appendiculaire. In “Traite de Zoologie” (P. P. Grass&,ed.), Vol. 13, Part 1, pp. 710-747. Masson, Paris. Modifications des nageoires. In “Trait6 de Zoologie” (P. P. Grass&, Bertin, L. (1958~). ed.), Vol. 13, Part 1, pp. 748-782. Masson, Paris. Bertin, L. (1958d). Systeme nerveux. In “Trait6 de Zoologie’.’ (P. P. Grass&,ed.), Vol. 13, Part 1, pp. 854-922. Masson, Paris. Blum, V. (1968). Das Kampfverhalten des braunen Diskusfisches, Symphysodon aequifasciata axelrodi L. P. Schultz (Teleostei, Cichlidae). Z . Tierpsychol. 25, 395-408. Boddeke, R., Slijper, E. J., and Van Der Stelt, A. (1959). Histological characteristics of the body-musculature of fishes in connection with their mode of life. Proc. K . Ned. Akad. Wet., Ser. C 62 576-588. Bone, Q. (1964). Patterns of muscular innervation in the lower chordates. Int. Reo. Neurobiol. 6, 99-147. Bone, Q. (1971). On the scabbard fish Aphanopus carbo. J. Mar. Biol. Assoc. U.K. 51, 219-225. Boulenger, G. A. (1907). “Zoology of Egypt: The Fishes of the Nile.” Published for the Egyptian Government, London. Breder, C. M. (1924). Respiration as a factor in locomotion of fishes. Am. Nut. 58, 145-155. Breder, C. M . (1926). The locomotion of fishes. Zoologica (N.Y.)4, 159-297. Breder, C. M. (1959). Studies on social groupings in fishes. Bull. Am. Mus. Nut. Hist. 117, 393-482. Breder, C. M. (1965). Vortices and fish schools.Zoologica (N.Y.)50,97-114. Breder, C. M., and Edgerton, H. E. (1942).An analysis of the loconiotion of the seahorse, Hippocampus, by means of high speed cinematography. Ann. N.Y. Acad. Sci. 43, 145- 172. Breder, C. M., and Nigrelli, R. F. (1938). The significance of differential locomotor activity as an index to the mass physiology of fishes. Zoologica (N.Y.)23, 1-29. Bregnballe, F. (1961). Plaice and flounder as consumers of the microscopic bottom fauna. Medd. Dan. Fisk.-Havunders. 3, 133-182. Brett, J. R., and Sutherland, D. B. (1965). Respiratory metabolism of pumpkinseed (fepomis gibbosus) in relation to swimming speed./. Fish. Res. Bd. Can. 22,405409. Bridge, T. W. (1896).The mesial fins of ganoids and te1eosts.J. Linn. Soc. London, Zool. 25,530-602. Bridge, T. W. (1904). Fishes (exclusive of the systematic account of Teleostei) In “The Cambridge Natural History” (S. F. Harmer and A. E. Shipley, eds.), Vol. 7, pp. 139-537. Macmillan, London. Campbell, B. (1951). The locomotor behavior of spinal elasmobranchs with an analysis of stinging in Urobatis. Copeia pp. 277-284. Chevrel, R. (1913). Essai sur la morphologie et la physiologie du muscle lateral chez les poissons osseux. Arch. 2001.Exp. Gen. 52, 473-607. Clemens, H. B. (1961). The migration, age and growth of Pacific albacore (Thunnus germo), 1951-1958. Fish. Bull., Calif. 115, 1-128. Clemens, H. B., and Flittner, G. A. (1969). Bluefin tuna migrate across the Pacific Ocean. Calif. Fish Came 55, 132-135.

1. FORM, FUNCTION, AND LOCOMOTORY HABITS

91

Collette, B. B., and Talbot, F. H. (1972). Activity patterns of coral reef fishes with emphasis on nocturnal-diurnal changeover. Nut. Hist. Mus. Los Angeles C t y . Sci. Bull. 14, 98-124. Cushing, D. H. (1973).“The Detection of Fish.” Pergamon, Oxford. Cushing, D. H., and Jones, F. R. H. (1968).Why do fish school? Nature (London)218, 918-920. Dahl, K, and Sflmrne, S. (1936).Experiments in salmon marking in Noway, 1935. Skr. Nor. Vidensk.-Akad. Oslo, 1 1935, No. 12, 1-27. Danforth, C. H. (1913).The myology of Po1yodon.J. Morphol. 24, 107-146. Darnell, R. M., and Meierotto, €3. R. (1965). Diurnal periodicity in the black bullhead, lctalurus melas (Rafinesque). Trans. A m . Fish. S O C . 94, 1-8. Davis, R. E. (1964). Daily “predawn” peak of locomotion in fish. Anim. Behao. 12, 272-283. Dean, B. (1895).“Fishes, Living and Fossil.” Columbia Univ. Press, New York. decroot, S. J. (1970).Some notes on an ambivalent behaviour of the Greenland halibut Reinhardtius hippoglossoides (Walb.) Pisces: Pleuronectiformes. J . Fish Biol. 2, 275-279. deVeen, J. F. (1967). On the phenomenon of soles (Solea solea L.) swimming at the surface.]. Cons., Cons. Perm. lnt. Explor. Mer 31,207-236. Diamond, J. (1971). The Mauthner cell. In “Fish Physiology” (W. S. Hoar and D. J. Randall, eds.), Vol. 5, pp. 265-346. Academic Press, New York. Disler, N. N. (1960). “Lateral Line Sense Organs and Their Importance in Fish Behaviour.” Izd. Akad. Nauk. SSSR. (Trans]. from Russian, Isr. Program Sci. Transl. No. 5799, Jerusalem, 1971.) DuBois, A. B., Cavagna, G . A., and Fox, R. S. (1974). Pressure distribution on the body surface of swimming fish.J. E x p . Biol. 60, 581-591. DuBois-Reymond, R. (1914). Physiologie der Bewegung. In “Handbuch der Vergleichenden Physiologie’’ (H. Winterstein, ed.), Vol. 3, Half 1, Part 1, pp. 1-248. Custav Fischer, Jena. Eaton, T. H. (1945). Skeletal supports of the median fins of fishes. J . Morphol. 76, 193-212. Eggert, B. (1929). Die Cobiidenflosse und ihre Anpassung an das Landleben. 2. Wiss. ZOOZ.133,411-439. Ellis, D. V. (1962). Preliminary studies on the visible migrations of adult sa1mon.J. Fish. Res. Board Can. 19, 137-148. Ellis, D. V. (1966). Swimming speeds of sockeye and coho salmon on spawning migration.]. Fish. Res. Board Can. 23, 181-187. Emelianov, S. W. (1935). Die Morphologie der Fischrippen. Zool. Jahrb., Abt. Anat. Ontog. Tiere 60, 133-262. Emery, A. R. (1969). Comparative ecology of damselfishes (Pisces: Pomacentridae) at Alligator Reef, Florida Keys. Diss. Ahstr. B 29, 2962. Emery, A. R. (1973).Comparative ecology and functional osteology of fourteen species of damselfish (Pisces: Pomacentridae) at Alligator Reef, Florida Keys. Bull. Mar. Sci. 23, 649-770. Escobar, R. A., Minahan, R. P., and Shaw, R. J. (1936).Motility factors in mass physiology: Locomotor activity of fishes under conditions of isolation, homotypic grouping, and heterotypic grouping. Physiol. 2001.9, 66-78. Fierstine, H. L., and Walters, V. (1968). Studies in locomotion and anatomy of scombroid fishes. Mem. South. Calif. Acad. Sci. 6, 1-31. Fishelson, L. (1968). Structure of the vertebral column in Lepadichthys lineatus, a clingfish associated with crinoids. Copeia pp. 859-861.

92

C . C . LINDSEY

Foss, G. (1968). Behaviour of Myxine glutinosa L. in natural habitat. Investigation of the mud biotope by a suction technique. Sarsia 31, 1-13. Fowler, J. A. (1970). Control of vertebral number in teleosts-an embryological problem. Q. Reu. Biol. 45, 148-167. Francois, Y. (1959). La nageoire dorsale anatomie comparee e t evolution. Annie Biol. 35, 81-113. Franzisket, L. (1965). Beobachtungen und Messungen am Flug der fliegenden Fische. Zool. Jahrb.,Abt. Allg. Zool. Physiol. Tiere 70, 235-240. Gallepp, G. W., and Magnuson, J. J. (1972). Effects of negative buoyancy on the behavior of the bluegill, Lepomis macrochirus Rafinesque. Trans. Am. Fish. SOC. 101, 507512. Ganguly, D. N., and Mitra, B. (1962). On the vertebral column of the teleostean fishes of different habits and habitats. I. Mugil corsula, Pama pama, Triacanthus breuirostris and Andamia heteroptera. Anat. Anz. 110,289-311. Ganguly, D. N., and Nag, A. C. (1964).On the organisation of the myomeric musculature of some benthonic teleostean fishes with special reference to the functional morphology. Anat. Anz. 115,418-446. Gayduck, V. V., Malinin, L. K., and Poddubnyy, A. G . (1971). Determination of the swimming depth of fish during the hours of daylight. Zchthyologica 11, 140-143. Geerlink, P. J,, and Videler, J. J. (1974). Joints and muscles of the dorsal fin of Tilapia nilotica L. (Fam. Cichlidae). Neth. J . Zool. 24, 279-290. Gibson, R. N. (1969). The biology and behaviour of littoral fish. Oceanogr. Mar. Biol. 7, 367-410. Goodrich, E. S . (1906). Notes on the development, structure and origin of the median and paired fins of fish. Q. J . Microsc. Sci. 50, 333-376. Goodrich, E. S. (1930). “Studies on the Structure and Development of Vertebrates.” Macmillan, London. Goodyear, C. P. (1970).Terrestrial and aquatic orientation in the starhead topminnow, Fundulus notti. Science 168, 603-605. Gosline, W. A. ( 1971).“Functional Morphology and Classification of Teleostean Fishes.” Univ. of Hawaii Press, Honolulu. Gradwell, N. (1971).Observations on jet propulsion in banjo catfishes. Can.J.Zool. 49, 1611-1612. Graham, J. B. (1970). Preliminary studies on the biology of the amphibious clinid Mnierpes macrocephalus. Mar. Biol., 5, 136140. Graham, J. B. (1973).Terrestrial life of the amphibious fish Mnierpes macrocephalus. Mar. Biol. 23, 83-91. Graham, J. B. (1975). Heat exchange in the yellowfin tuna, Thunnus albacares, and skipjack tuna, Katsuwonus pelamis, and the adaptive significance of elevated body temperatures in scombrid fishes. U.S. Fish Wildl. Seru., Fish. Bull. 73, 219-229. Graham, J. B., and Rosenblatt, R. H. (1970). Aerial vision: Unique adaptation in an intertidal fish. Science 168, 586588. Gray, J. (1933a). Studies in animal locomotion. I. The movement of fish with special reference to the ee1.J. E x p . Biol. 10, 88-104. Gray, J. (1933b). Studies in animal locomotion. 11. The relationship between waves of musculature contraction and the propulsive mechanism of the ee1.J. E x p . Biol. 10, 386390. Gray, J. ( 1 9 3 3 ~ )Studies . in animal locomotion. 111. The propulsive mechanism of the whiting (Gadus merlangus).J . E x p . Biol. 10, 391-400. Gray, J. (1936a). Studies in animal locomotion. IV. The neuromuscular mechanism of swimming in the ee1.J. E x p . Biol. 13, 170-180.

1. FORM, FUNCTION,

AND LOCOMOTORY HABITS

93

Gray, J. (1936b). Studies in animal locomotion. VI. The propulsive powers of the dolphin. J. Exp. Biol. 13, 192-199. Gray, J. (1968). “Animal Locomotion.” Weidenfeld & Nicolson, London. Greene, C. W., and Greene, C. H. (1914).The skeletal musculature of the king salmon. Bull. U S .Fish. Bur. 33, 21-59. Greenfield, D. W. (1968). Observations on the behavior of the basketweave cusk-eel Otophidium scrippsi Hubbs. Calif. Fish Game 54, 108-114. Greenway, P. (1965). Body form and behavioural types in fish. Erperientia 21,489-498. Greenwood, P. H., and Thomson, K. S. (1960). The pectoral anatomy ofpantodon buchholzi Peters (a freshwater flying fish) and the related Osteoglossidae. Proc. Zool. SOC. London 135, 283-301. Greer Walker, M., Mitson, R. B., and Storeton-West, T. (1971).Trials with a transponding acoustic fish tag tracked with an electronic sector scanning sonar. Nature (London) 229, 196-198. Gregory, W. K. (1928).Studies on the body-forms of fishes. Zoologica (N.Y.)8,325-421. Grenholm, A. (1923).Studien iiber die Flossenmuskulatur der Teleostier. Inaug. Diss., Uppsala Univ. Arsskr., (Matem.Naturuet. Uppsala) 2, 1-296. Gunter, G. (1953). Observations on fish turning flips over a line. Copeia pp. 188-190. Guthrie, D. M., and Banks, J. R. (1970a).Observations on the function and physiological properties of a fast paramyosin muscle-the notochord of amphioxus (Branchiostoma lanceolatum).J . E x p . Biol. 52, 125-138. Guthrie, D. M., and Banks, J. R. (1970b).Observations on the electrical and mechanical properties of the myotomes o f the lancelet (Branchiostoma Zanceolatum).J . E x p . Biol. 52, 401-417. Harder, W. (1964). Anatomie der Fische. In “Handbuch Binnenfischerei Mitteleuropas,” (R. Demoll, H. N. Maier and H. H. Wundsch, eds.) Vol. 2a, pp. 1-308, Figs. 1-96. Schweizerbart’sche Verlags., Stuttgart. Harris, J. E. (1953).Fin patterns and mode of life in fishes. I n “Essays in Marine Biology” (S. M. Marshall and A. P. Orr, eds.), pp. 17-28. Oliver & Boyd, Edinburgh. Harris, J. E., and Whiting, H. P. (1954).Structure and function in the locomotory system of the dogfish embryo. The myogenic stage o f movement. J . E x p . Biol. 31,501-524. Harris, V. A. ( 1960). On the locomotion of the mud-skipper Periophthalmus koelreuteri (Pallas): (Gobiidae). Proc. Zool. SOC.London 134, 107-135. Hartt, A. C. (1966).Migrations of salmon in the North Pacific Ocean and Bering Sea as determined by seining and tagging, 1959-1960. Int. North Pac. Fish. Comm. Bull. 19, 1-141. Hasler, A. D., and Villemonte, J. (1953). Observations on the daily movements o f fishes. Science 118, 321-322. Healey, E. G. (1957).The nervous system. In “The Physiology of Fishes” (M. E. Brown, ed.), Vol. 2, pp. 1-119. Academic Press, New York. Herald, E. S. (1961).“Living Fishes of the World.” Doubleday, Garden City, New Yoyk. Herald, E. S., and Ripley, W. E. (1951). The relative abundance o f sharks and bat stingrays in San Francisco Bay. Calif. Fish Game 37, 315-329. Hergenrader, G. L., and Hasler, A. D. (1967). Seasonal changes in swimming rates of yellow perch in Lake Mendota as measured by sonar. Trans. Am. Fish. SOC. 96, 373-382. Hertel, H. (1966).“Structure, Form and Movement.” Reinhold, New York. Hobson, E. S . (1965). Diurnal-nocturnal activity of some inshore fishes in the Gulf of California. Copeia pp. 291-302. Hobson, E. S . (1968).Predatory behavior of some shore fishes in the Gulf of California. Bur. Sport Fish. Wildl. ( U . S . ) ,Res. Rep. 73, 1-92.

94

C. C. LINDSEY

Hobson, E. S. (1973). Die1 feeding migrations in tropical reef fishes. Helgol. Wiss. Meeresunters. 24, 361-370. Hobson, E. S . (1974).Feeding relationships of teleostean fishes on coral reefs in Kona, Hawaii. U . S . Fish Wildl, Seru., Fish. Bull. 72, 915-1031. Holden, M. J. (1967).Transatlantic movement of a tagged spurdogfish.Nature (London) 214, 1140-1141. Holland, G . A. (1957). Migration and growth of the dogfish shark, Squalus acanthias (Linnaeus), of the eastern North Pacific. Wash. Dep. Fish., Fish. Res. Pap. 2(1), 43-59. Hoogland, R. D. (1951). On the fixing-mechanism in the spines of Gasterosteus aculeatus L. Proc. K. Ned. Akad. Wet. C 54, 171-180. Hora, S. L. (1935).Ancient Hindu conception of correlation between form and locomotion of fishes. J . Asiat. Soc. Bengal, Sci. 1, 1-7. Horn, M. (1975). Swim-bladder state and structure in relation to behavior and mode of life in Stromateoid fishes. U.S. Fish Wildl. S e w . , Fish. Bull. 73, 95-109. Houssay, F. (1912). ‘‘Forme, Puissance et Stabilite des Poissons,” Collection d e Morphologie Dynamique, Vol. 4. Herman, Paris. Howell, A. B. (1933).The architecture of the pectoral appendage of the codfish. Anat. Rec. 56, 151-158. Hubbs, C. L. (1933). Observations on the Hight of fishes, with a statistical study of’the Hight of the Cypselurinae and remarks on the evolution of the Hight of fishes. Pap. Mich. Acad. Sci., Arts Letts. 17, 575-611. Hudson, R. C. L. (1973).On the function of the white muscles in teleosts at intermediate swimming speeds.]. Exp. Biol. 58, 509-522. Hubner, H. (1961).Die Wirbelsiule des Karpfens (Cyprinus carpio L.). Z. Fisch. Deren Hilfswiss. 10, 429-505. Hughes, G . M., and Shelton, G. (1962). Respiratory mechanisms and their nervous control in fish. Adti. Camp. Physiol. Biochem. 1, 275-364. Hunter, J. R., and Zweifel, J. R. (1971).Swimming speed, tail beat frequency, tail beat amplitude, and size in jack mackerel, Trachurus symmetricus, and other fishes. U S . Fish Wildl. Serti., Fish. Bull. 69, 253-266. Idler, D. R., and Clemens, W. A. (1959). The energy expenditures of Fraser River sockeye salmon during the spawning migration to Chilko and Stuart Lakes. Znt. Pac. Salmon Fish. Comm. Prog. Rep. pp. 1-80. Jarman, G . M. (1961).A note on the shape of fish myotomes. Symp. Zool. Soc. London 5, 33-35. Johnels, A. C . (1957). The mode of terrestrial locomotion in Clarias. Oikos 8, 122129. Johnels, A. G., and Svensson, G . S. 0. (1954).On the biology of Protopterus annectens (Owen).Ark. Zool. 7, 131-164. Johnson, R. D. 0. (1912).Notes on the habits of a climbing catfish (Arges mannoratus) from the republic of Colombia. Ann. N.Y. Acad. Sci. 22, 327-333. Johnson, W. E., and Groot, C. (1963).Observations on the migration of young sockeye salmon (Oncorhynchus nerka) through a large, complex lake system. J. Fish. Res. Board. Can. 20,919-938. Johnston, I. A., and Goldspink, G . (1973a).A study of the swimming performance of the Crucian carp Carassius carassius (L.) in relation to the effects of exercise and recovery on biochemical changes in the myotomal muscles and 1iver.J. Fish Biol. 5, 249-260. Johnston, I. A., and Goldspink, G . (197313). Quantitative studies on muscle glycogen utilization during sustained swimming in Crucian carp (Carassius carassius L.).J . Exp. Biol. 59, 607-615.

1. FORM,

FUNCTION, AND LOCOMOTORY HABITS

95

Jones, F. R. H. (1952). The swimbladder and the vertical movements of teleostean fishes. 11. The restriction to rapid and slow movements. 1.E x p . Biol. 29, 9 4 1 0 9 . Jones, F. R. H. (1956).The behaviour of minnows in relation to light intensity.]. E x p . Biol. 33, 271-281. Jones, F. R. H. (1968). “Fish Migration.” Arnold, London. Kafuku, T. (1950). “Red muscles” in fishes. I. Comparative anatomy of the scombroid fishes of Japan. Gyoruigaku Zasshi 1, 89-100. Kashin, S . M., and Smolyaninov, V. V. (1969). Concerning the geometry of fish trunk muscles. J. lchthyol. (USSR) 9, 923-925. Keast, A., and Webb, D. (1966). Mouth and body form relative to feeding ecology in the fish fauna of a small lake, Lake Opinicon, Ontario. 1. Fish. Res. Board Can. 23, 1845- 1874. Keenleyside, M. H. A. (1955). Some aspects of the schooling behaviour of fish. Behaviour 8, 18.3-248. Kerswill, C. J., and Keenleyside, M. H. A. (1961). Canadian salmon caught off Greenland. Nature (London) 192,279. Kishinouye, K. ( 1923). Contributions to the comparative study of the so-called scombroid fishes.]. Coll. Agric. Tokyo Imp. Univ. 8, 293-475. Klausewitz, W. (1963). Wie schwimmt der Schnepfenfisch? Natur Mus. 93,69-73. Klausewitz, W. ( 1964). Der Lokomotionsmodus der Fliigelrochen (Myliobatoidei). Zool. Anz. 173, 110-120. Klausewitz, W. ( 1967). Uber einige Bewegungsweisen der Schlammspringer (Periophthalmus).Natur Mus. 97, 211-222. Kleerekoper, H., Timms, A. M., Westlake, G. F., Davy, G . F., Malar, T., and Anderson, V. M. (1970). An analysis of locomotor behaviour of goldfish (Carassius auratus). Anim. Behav. 18,317-330. Kramer, E . (1960). Zur Form und Funktion des Lokomotionsapparates der Fische. Z . Wiss. Zool. Abt. A 163, 1-36. Kruuk, H. (1963).Diurnal periodicity in the activity of the common sole, Solea vulgaris Quensel. Neth. 1.Sea Res. 2, 1-28. leDanois, Y. (1958).Systeme musculaire. In ‘‘Trait6 d e Zoologie” (P. P. Crass&,ed.), Vol. 13, Part 1, pp. 783-817. Masson, Paris. Lieder, U. (1961). Untersuchungsergebnisse iiber die Gratenzahlen bei 17 SiisswasserFischarten. Z . Fisch. Deren Hilfswiss. 10, 329-350. Lighthill, M. J. (1960). Note on the swimming of slender fish.]. Fluid Mech. 9,305-317. Lighthill, M. J. ( 1969). Hydromechanics of aquatic animal propulsion. Annu. Rev. Fluid Mech. 1,413-446. Lindsey, C. C. ( 1955). Evolution of meristic relations in the dorsal and anal fins of teleost fishes. Trans. R. SOC. Can. 49 (Ser. 3, Sect. 5), 35-49. Lindsey, C. C. (1966). Body sizes of poikilotherm vertebrates at different latitudes. Evolution 20, 456465. Lindsey, C. C. (1975). Pleomerism, the widespread tendency among related fish species for vertebral number to be correlated with maximum body length.]. Fish. Res. Board Can. 32, 2453-2469. Lindsey, C. C., and Harrington, R. W. (1972). Extreme vertebral variation induced by temperature in a homozygous clone of the self-fertilizing cyprinodontid fish Rivulus marmoratus. Can. 1.2002.50, 733-744. Ling, S. C., and Ling, T. Y.J. (1974).Anomalous drag-reducing phenomenon at a water/ fish-mucus or polymer interface.]. Fluid Mech. 65, 499-512. Lissmann, H. W. (1946a). The neurological basis of the locomotory rhythm in the spinal dogfish (Scyllium canicula, Acanthias vulgaris). I. Reflex behaviour. J . E x p . Biol. 23, 143-161.

96

C. C. LINDSEY

Lissmann, H. W. (1946b). The neurological basis of the locomotory rhythm in the spinal dogfish (Scyllium canicula, Acanthias uulgaris). 11. The effect of de-afferentiation. J . E x p . Biol. 23, 162-176. Lissmann, H. W. (19614. Zoology, locomotory adaptations and the problem of electric fish. In “The Cell and the Organism” (J. A. Ramsay and V. B. Wigglesworth, eds.), pp. 301-317. Cambridge Univ. Press, London and New York. Lissmann, H. W. ( 1961b). Ecological stuaies on gymnotids. In “ Bioelectrogenesis” ( C . Chagas and A. P. decarvalho, eds.), pp. 215-226. Elsevier, Amsterdam. Lund, €3. (1967). An analysis of the propulsive mechanisms of fishes with reference to some fossil actinopterygians. Ann. Carnegie Mus. 39, 195-218. McCutchen, C. W. (1970). The trout tail fin: A selfcambering hydrofoi1.J. Biomech. 3, 271-281. McMurrich, J. P. (1884). The myology ofAmiurus catus (L.) Gill. Proc. R. Can. Inst. 2, 311-351. Macnae, W. (1968). A general account of the fauna and flora of mangrove swamps and forests in the Indo-West-Pacific region. Ado. Mar. Biol. 6, 73-270. Magnan, A. (1929). Les caractbristiques gComCtriques et physiques des poissons. Premiere partie. Ann. Sci. Nut. Zool. 12, 1-133. Magnan, A. ( 1930). Les CaractCristiques gCom6triques et physiques des poissons. Deuxieme partie. Ann. Sci. Nat. 2001.13, 134-269. Mahajan, C. L. (1967). Sisor rabdophorus-a study in adaptation and natural relationship. 111. The vertebral column, median fins and their musculature. J . Zool. 152, 297-318. Manzer, J. I. (1946). Interesting movements as shown by the recoveries of certain species of tagged fish. Progr. Rep. Pac. Biol. Sta. 67,31. Marey, E. J. (1895). “Movement.” Heinemann, London. Marshall, N. B. (1971). “Explorations in the Life of Fishes.” Harvard Univ. Press, Cambridge, Massachusetts. Mather, F. J. (1962). Transatlantic migration of two large bluefin tuna. J . Cons. Cons. Perm. lnt. Explor. Mer 27, 325-327. Mather, F. J., Mason, J. M., and Clark, H. L. (1974a). Migrations of white marlin and blue marlin in the western North Atlantic Ocean-tagging results since May, 1970. NOAA (Natl. Oceanic Atmos. Adm.) Tech. Rep. N M F S (Natl. Mar. Fish. Sero.) SSRF (Spec. Sci. Rep.-Fish.) No. 675, pp. 211-225. Mather, F. I., Tabb, D. C., Mason, J. M., and Clark, H. L. (1974b). Results of sailfish tagging in the western North Atlantic Ocean. NOAA (Natl. Oceanic Atmos. Adm.) Tech. Rep. NMFS (Natl. Mar. Fish. Seru.)SSRF (Spec.Sci. Rep.-Fish.) No. 675, pp. 194-210. Matthews, L. H., and Parker, H. W. (1951). Basking sharks leaping. Proc. Zool. SOC. London 121,461-462. Maurer, F. (1913). Die ventrale Rumpfmuskulatur der Fische (Selachier, Ganoiden, Teleostier, Crossopterygier, Dipnoer). Jen. Z. Naturwiss. 49, 1-118. Menzies, W. J. M., and Shearer, W. M. (1957). Long distance migration of salmon. Nature (London) 179,790. Meyer, A. (1965). Results of cod tagging by the Federal Republic of Germany in the Greenland area from 1959 to 1964. Int. Comm. Northwest Atl. Fish. Redbook Part Ill pp. 148-152. Minckley, C. O., and Klaassen, H. E. (1969). Burying behavior of the plains killifish, Fundulus kansae. Copeia pp. 200-201. Myers, G. S. (1950). Flying of the halfbeak, Euleptorhamphus. Copeia p. 320. Nawar, G. (1955). On the anatomy of Clarias lazera. 11.The muscles of the head and the pectoral gird1e.J. Morphol. 97, 23-58.

1. FORM, FUNCTION, AND LOCOMOTORY HABITS

97

Neave, F. (1964). Ocean migrations of Pacific salmon. J . Fish. Res. Board Can. 21, 1227- 1244. Nikolsky, G . V. (1963).“The Ecology of Fishes.” Academic Press, New York. Nishi, S. (1938). Muskelsystem 11. Muskeln des Rumpfes. Bolks Handb. Vgl. Anat. Wirbeltiere 5, 351-446. Nishimura, S . ( 1964).Additional information on the biology of the dealfish, Trachipterus ishikawai Jordan and Snyder. Bull. Jpn. Sea Reg. Fish. Res. Lab. (Nihonkai-Ku Suisan Kenkyusho Kenkyo) 13, 127-129. Nishimura, S., and Hirosaki, Y. (1964). Observations on the swimming behavior of some taeniosomous fishes in aquaria and in nature. Publ. Set0 Mar. Biol. Lab. 12, 165171. Norman, J. R., and Greenwood, P. H. (1975). “A History of Fishes,” 3rd Ed. Benn, London. N o d s , H. W., and Hughes, S. P. (1920).The cranial, occipital, and anterior spinal nerves of the dogfish, Squalus acanthias. J . Comp. Neurol. 31,293-404. Northcote, T. G . ( 1962).Migratory behaviour ofijuvenile rainbow trout, Salmo gairdneri, in outlet and inlet streams of Loon Lake, British Columbia.J . Fish. Res. Board Can. 19,201-270. Northcote, T. G., Lorz, H. W., and MacLeod, J. C. (1964). Studies on die1 vertical movement of fishes in a ’British Columbia lake. Verh. Int. Verein Limnol. 15,940946. Nursall, J. R. (1956).The lateral musculature and the swimming of fish. Proc. Zool. SOC. London 126, 127-143. Nursall, J. R. (1963). The caudal musculature of Hoplopagrus guntheri Gill (Perciformes: Lutjanidae). Can. J . Zool. 41, 865-880. Oehmichen, E. (1958). Locomotion des poissons. In “Trait6 d e Zoologie” (P. P. Grass&, ed.), Vol: 13, Part 1, pp. 818-853. Masson, Paris. Ohlmer, W., and Schwartzkopff, J. ( 1959). Schwimmgeschwindigkeiten’vonFischen aus stehenden Binnengewassern. Naturwissenschaften 46, 362-363. Olla, B. L., and Studholme, A. L. (1972). Daily and seasonal rhythms of activity in the bluefish (Pomatomus saltatrix).In “Behavior of Marine Animals” (H. E. Winn and B. L. O h , eds.), Vol. 2, pp. 303-326. Plenum, New York. Orange, C. J., and Fink, B. D. (1963). Migration of a tagged bluefin tuna across the Pacific Ocean. Calif. Fish Game 49, 307-309. Osbome, M. F. M. (1961). The hydrodynamical performance of migratory salmon. J . E x p . Biol. 38,365-390. Packard, A. (1972). Cephalopods and fish: the limits of convergence. Biol. Rev. Cambridge Philos. SOC. 47, 241-307. Patterson, S., and Goldspink, G. (1972).The fine structure of red and white myotomal muscle fibres of the coalfish (Gadus virens). Z . Zellforsch. Mikrosk. Anat. 133, 463-474. Pavlov, D. S., Sbikin, Y. N., and Mochek, A. D. (1968). The effect of illumination in running water on the speed of fishes in relation to feature of their orientation. Probl. Ichthyol. 8, 250-255. Pettigrew, J. B. (1874).“Animal Locomotion.” Int. Sci. Ser., London and New York. Pitcher, T. J. (1973). The three-dimensional structure of schools in the minnow, Phoxinus phoxinus (L.).Anim. Behav. 21,673-686. Radakov, D. V. (1973). “Schooling in the Ecology of Fish.” (Transl. from Russian, Isr. Program Sci. Transl. No. 22076, Jerusalem, 1973.) Riss, W. (1972).Overview ofthe design of the central nervous system and the problem of the natural units of behavior. IV.Locomotion and spinal organization. Brain, Behav., Evol. 4.439-462.

98

C. C. LINDSEY

Roberts, B. L. ( 1 9 6 9 4 The spinal nerves of the dogfish (Scyliorhinus).J . Mar. Biol. ASSOC. U.K. 49,51-75. Roberts, B. L. (1969b).The co-ordination of the rhythmical fin movements of dogfish.]. Mar. Biol. Assoc. U . K . 49, 357-378. Roberts, B. L. ( 1 9 6 9 ~ The ) . buoyancy and locomotory movements of electric rays.]. Mar. Biol. Assoc. U.K.49, 621-640. Roberts, B. L. (1969d). The response of a proprioceptor to the undulatory movements of dogfish. ]. E x p . Biol. 51, 775-785. Roberts, B. L. (1972).Activity of lateral-line sense organs in swimming dogfish.]. E x p . Biol. 56, 105-118. Roberts, B. L., and Russell, I. J. (1972). The activity of lateral-line efferent neurones in stationary and swimming dogfish. J . E x p . Biol. 57, 435-448. Rosenblatt, R. H., and Johnson, G . D. (1976). Anatomical considerations of pectoral swimming in the opah, Lampris guttatus. Copeia pp. 367-370. Rosenthal, H. (1968). Schwiinmverhalten und Schwimmgeschwindigkeit bei den Laryen des Herings Clupea harengus. Helgol. Wiss. Meeresunters. 18, 453-486. Samuel, C. T. (1961). On the ambulatory mechanism in Choridactylus multibarbis Richardson. Bull. Cent. Res. Inst. Unio. Kerala, Trioandrum 8, 79-83. Sandon, H. (1956). An abnormal specimen of Synodontis membranaceus (Teleostei, Siluroidea), with a discussion on the evolutionary history of the adipose fin in fish. Proc. Zool. Soc. London 127,453-459. Satchell, G . H. (1968). A neurological basis for the co-ordination of swimming with respiration in fish. Comp. Biochem. Physiol. 27, 835-841. Saxena, S. C., and Chandy, M. (1966a). The pelvic girdle and fin in certain Indian hill qtream fishes.]. Zool. 148, 167-190. Saxena, S. C., and Chandy, M. (1966b).Adhesive apparatus in certain Indian hill stream fishes.]. Zool. 148, 315-340. Schmalhausen, J. J. (1912).Zur Morphologie der unpaaren Flossen. I. Die Entwicklung des Skelettes und der Muskulatur der unpaaren Flossen der Fische. Z. Wiss. Zool. 100, 509-587. Schmalhausen, J. J. (1913). Zur Morphologie der unpaaren Flossen. 11. Bau und Phylogenese der unpaaren Flossen und insbesonders der Schwanzflosse der Fische. Z. Wiss. Zool. 104, 1-80, Schmidt-Nielsen, K. ( 19724. “How Animals Work.” Cambridge Univ. Press, London and New York. Schmidt-Nielsen, K. (19721,). Locomotion: Energy cost of swimming, flying, and running. Science 177, 222-228. Schneider, H. (1964).Untersuchungen zur Schwimmweise der Kugelfische. I. Die Flossenmuskulatur der Flusskugelfisches (Tetraodon fluviatilis) in vergleich zu der der Schleie (Tinca tinca). Z. Morphol. Oekol. Tiere 54,414-435. Schuett, F. (1934).Studies in mass physiology: The activity of goldfishes under different conditions of aggregation. Ecology 15, 258-262. Sewertzoff, A. N. (1926).Die Morphologie der Brustflossen der Fische.]ena. Z. Naturtoiss. 62, 343-392. Sewertzoff, A. N. (1934).Evolution der Bauchflossen der Fische.Zool.]ahrb., Abt. Anat. Ontog. Tiere 58, 415-500. Shann, E. W. (1914).On the nature of the lateral muscle in Teleostei. Proc. 2001.SOC. London pp. 319-337. Shann, E. W. (1924). Further observations on the myology of the pectoral region in fishes. Proc. Zool. Soc. London pp. 195-215. Shaw, E. (1962). The schooling of fishes. Sci. Am. 206, 128-138.

1.

FORM, FUNCTION, AND LOCOMOTORY HABITS

99

Sheldon, F. F. (1937).Osteology, myology and probable evolution of the nematognath pelvic girdle. Ann. N.Y. Acad. Sci. 37, 1-96. Shelton, G. (1970).The regulation of breathing. I n “Fish Physiology” (W. S. Hoar and D. J. Randall, eds.), Vol. 4, pp. 293-359. Academic Press, New York. Shepard, M. P., Hartt, A. C., and Yonemori, T. (1968). Salmon of the North Pacific Ocean-Part 8. Chum salmon in offshore waters. Int. North Pac. Fish. Comm. Bull. 25, 1-69. Shubina, T. N. (1971). Spawning and postspawning migrations of the sevryuga (Acipenser stellatus (Pallas)) in the lower Volga-routes and speeds. J. Zchthyol. (USSR) 11,88-97. Simons, J. R. (1970).The direction of the thrust produced by the heterocercal tails of two dissimilar elasmobranchs: the Port Jackson shark, Heterodontus portusjacksoni (Meyer), and the piked dogfish, Squalus megalops (Macleay). J . E x p . B i d . 52, 95-107. Smith, E. H., and Stone, D. E. (1961). Perfect fluid forces in fish propulsion. Proc. R . SOC., Ser. A 261,316-328. Spencer, W. P. (1939). Diurnal activity rhythms in fresh-water fishes. Ohio /. Sci. 39, 119-132. Squire, J. L. (1974). Migration patterns of Istiophoridae in the Pacific Ocean as determined by cooperative tagging programs. NOAA (Natl. Oceanic Atmos. Adm.) Tech. Rep. N M F S (Natl. Mar. Fish. S e w ) S S R F (Spec. Sci. Rep.-Fish.) No. 675, pp. 226237. Stickney, A. P. (1972). The locomotor activity of juvenile herring (Clupea harengus harengus L.) in response to changes in illumination. Ecology 53,438-445. Swain, A. (1963). Long distance migration of salmon. Nature (London) 197, 923. Swanson, P. L. (1949). “Hurdling” by the needlefish. Copeia p. 219. Szarski, H. (1964).The functions of myomere folding in aquatic vertebrates. Bull. Acad. Pol. Sci., Ser. Sci. B i d . 12, 305-306. Takahashi, N. (1917). On the homology of the median longitudinal musclessupracarinalis and infracarinalis-with the fin muscles of the dorsal and anal fins, and their functi0ns.J. Coll. Agric., Tokyo Imp. Coll. 6, 199-213. T h i n g , A. V. (1937). Some features in the migration of cod. J. Cons. Cons. Perm. Znt. Explor. Mer 12,3-35. Taylor, G . (1952).Analysis of the swimming of long and narrow animals. Proc. R. Soc., Ser. A 214, 158-183. Tesch, F.-W. (1972).Versuche zur telemetrischen Verfolgung der Laichwanderung von Aden (Anguilla anguilla) in der Nordsee. Helgol. Wiss.Meeresunters. 23,165-183. Tester, A. L. (1959). Summary of experiments on the response of tuna to stimuli. In “Modern Fishing Gear of the World” (H. Kristjonsson, ed.), pp. 538-542. Fishing News (Books), London. Thomas, A. M. (1956).The cranial muscles of teleosts-1.J. Anim. Morphol. Physiol. 3, 13-24. Thomson, K. S. (1971).The adaptation and evolution of early fishes. Q. Reu. Biol. 46, 139- 166. Tilak, R., and Kanji, S . K. (1967). Studies on the morphology of the pectoral girdle of Pisoodonophis boro (Ham.) in relation to its habit. Anat. Anz. 120, 404-408. Trueman, E. R., and Ansell, A. D. (1969).The mechanisms of burrowing into soft substrata by marine animals. Oceanogr. Mar. Biol. 7, 315-366. Tucker, V. A. (1975).The energetic cost of moving about. Am. Sci. 63,413-419. Tyler, J. C. (1962).The pelvis and pelvic fin of plectognath fishes; a study in reduction. Proc. Acad. Nut. Sci. Philadelphia 114, 207-250.

100

C. C. LINDSEY

van Olst, J. C., and Hunter, J. R. (1970). Some aspects of the organization of fish schools. /. Fish. Res. Board Can. 27, 1225-1238. Verheijen, F. J., and deGroot, S. J. (1966). Diurnal activity of plaice and flounder (Pleuronectidae) in aquaria. Neth. J . Sea Res. 3, 383-390. Walters, V. (1962). Body form and swimming performance in the scombroid fishes. Am. Zool. 2, 143-149. Webb, P. W. (1973a). Effects of partial caudal-fin amputation on the kinematics and metabolic rate of underyearling sockeye salmon (Oncorhynchus nerka) at steady swimming speeds.J. E x p . Biol. 59, 565-581. Webb, P. W. (1973b). Kinematics of pectoral fin propulsion in Cymatogaster aggregata. J . E x p . Biol. 59, 697-710. Webb, P. W. (1975). Hydrodynamics and energetics of fish propulsion. Bull., Fish. Res. Board Can. 190, 1-159. Weihs, D. (1973a). Mechanically efficient swimming techniques for fish with negative buoyancy. J . Mar. Res. 31, 194-209. Weihs, D. (1973b). Hydromechanics of fish schooling. Nature (London)241,290-291. Weitzman, S. H. (1954). The osteology and the relationships of the South American characid fishes of the subfamily Gasteropelecinae. Stanford Ichthyol. Bull. 4, 21 1-263. Wickler, W. ( 1971). Verhaltensstudien an einem hochspezialisierten Grundfisch, Gastromyzon hwneensis (Cyprinoidea, Gastromyzonidae). Z. Tierpsychol. 29, 467480. Willemse, J. J. (1959).The way in which flexures of the body are caused by muscular contractions. Proc. K . Ned. Akad. Wet.,Ser. C 62, 589-593. Winterbottom, R. (1974).A descriptive synonomy of the striated muscles of the Teleostei. Proc. Acad. Nut. Sci. Philadelphia 125, 225-317. Woodhead, P. M. J. (1966).The behaviour of fish in relation to light in the sea. Oceanogr. Mar. Biol. 4, 337-403. Woodhead, P. M. J., and Woodhead, A. D. (1955). Reactions of herring larvae to light: A mechanism of vertical migration. Nature (London) 176, 349-350. Yuen, H. S. H. (1970).Behavior of skipjack tuna,Katsuwonuspelamis, as determined by tracking with ultrasonic devices.J. Fish. Res. Board Can. 27, 2071-2079. Yuen, H. S . H., Dizon, A. E., and Uchiyama, J. H. (1974).Notes on the tracking of the Pacific blue marlin, Makaira nigricans. NOAA (Natl. Oceanic Atmos. Adm.) Tech. Rep. N M F S (Natl. Mar. Fish. Seru.) SSRF (Spec. Sci. Rep.-Fish.) No. 675, pp. 265-268. Zuyev, G. V., and Belyayev, V. V. (1970). An experimental study of the swimming of fish in groups as exemplified by the horsemackerel (Trachurus mediterraneus ponticus Aleev).]. Zchthyol. ( U S S R ) 10, 545-549.

2 SWIMMING CAPACITY F. W. H . BEAMISH I. Introduction .................................................... 11. Field Measurements of Performance ............................. A. Methodology. ............................................... B. Swimming Capacity ......................................... 111. Laboratory Measurements of Performance ........................ A. Swimming Chamber Design ................................. B. Experimental Procedure ..................................... C. Biological Constraints on Performance ........................ D. Environmental Constraints on Performance ................... IV. Energetics of Swimming.. ....................................... V. Application to Management Practices ............................. References .........................................................

101 103 103 106 117 117 128 137 150 163 168 172

I. INTRODUCTION Progress of fish maneuvering in water, the density of which is approximately equal to that of the animal itself, can be classified into three major categories: sustained, prolonged, and burst swimming speeds. Each reflects not only on the constraints imposed by time, but also on the biochemical processes which supply the fuel for their application. Sustained swimming performance is applied to those speeds which can be maintained for long periods (greater than 200 min) without resulting in muscular fatigue. Included within sustained performance under the subcategory of cruising are those speeds achieved by migrating fish as well as the velocities which negatively bouyant species such as the scombroid and xiphoid fishes must achieve to maintain 101 FISH PHYSIOLOGY, VOL VII Copyright @ 1978 by Academic P r w , InL All sight, of repiodurtion in any form reserved ISBN 0-12-350407-4

102

F. W. H. BEAMISH

hydrostatic equilibrium. A second subcategory is sustained schooling which includes speeds displayed by groups of fish distributed in a regular array, examples of which include many of the clupeids and tunas. Also within sustained performance is routine activity which represents the daily movements of fish including foraging and station holding, Routine activity thus includes periods of steady and unsteady swimming. The quantitative measurement of routine activity in the field entails enormous practical difficulties which have yet to be overcome, In laboratory studies, routine activity is generally equated to the locomotion displayed by fish whose only movements are spontaneous which again is highlighted by periods of steady and unsteady swimming. Prolonged swimming speed is of shorter duration (20 sec-200 min) than sustained and ends in fatigue. In field studies it is often difficult to separate sustained and prolonged, not only because of the practical difficulties imposed when attempting to track fish for long periods but also on account of the variability in swimming speed expressed by fish even when migrating or in schools. Rarely, if ever, is it possible to assess fatigue in the field. Prolonged speeds are most accurately measured in the laboratory in swimming flumes. Critical swimming speed, a special category of prolonged, was first defined and employed by Brett (1964) to designate the maximum velocity fish could maintain for a precise time period. It is measured by interpolation for those fish that do not fatigue exactly at the beginning or end of a prescribed period. The details of its calculation are described elsewhere in this chapter. The application of critical swimming speed is confined to laboratory investigations. The highest speeds of which fish are capable are organized under the category of burst swimming. These high speeds can be maintained only for short periods (less than 20 sec) and are characterized by an initial acceleration phase of unsteady swimming followed by a steady phase hereafter termed sprint. The capacity for short-term high performance is essential to the survival of many species as it facilitates the capture of prey, avoidance of predators, or the negotiation of rapid currents as may be encountered in rivers during spawning migration. This chapter attempts to synthesize measurements of swimming performance within the categories described above in relation to environmental and biological factors. Some guidelines are provided for the measurement of performance both in the field and in the laboratory. Finally, the application of information on swimming performance to fishery management programs is discussed.

2.

SWIMMING CAPACITY

103

11. F I E L D MEASUREMENTS OF PERFORMANCE A. Methodology 1. DIRECTOBSERVATION AND CONVENTIONAL TAGS Perhaps the simplest procedure employed to measure swimming speed is to estimate by direct observation the time required for fish to swim a gauged distance (Lane, 1941) or to relate swimming velocity to an estimated rate of movement on land. For example, the burst speed of pike was obtained by estimating the speed at which a boy had to run to keep pace with the fish in still water (Stringham, 1924). Where migrating fish pass upstream through shallow channels or culverts of known length and velocity, direct estimates of burst, prolonged, or sustained cruising speed have been made (Wales, 1950; Dow, 1962; Ellis, 1966). Wantanabe (1942) measured the time required by skipjack tuna, Euthynnus pelamis, to swim a marked distance of 2 m alongside a ship but ignored currents in his calculation. Davidson (1949) observed the time required by young Atlantic salmon, Salmo salar, to cover specific distances in circular ponds in which the current pattern was well described. Conservative estimates of sustained cruising or schooling swimming speeds have been made from conventional tagging studies, assuming that the fish travel the distance between the points of capture and recapture in a straight line.

2. SPORTSFISHING GEAR Modified sports fishing gear has been applied to the measurement of swimming speed (Magnan, 1929; Lane, 1941; Gero, 1952; Walters and Fierstine, 1964). Lane (1941) attached a motorcycle speedometer to a rod and reel to measure the burst swimming speed of a tuna caught by hook. Gero (1952) developed an instrument he termed a piscatometer, for measuring swimming velocity of hooked fish. The piscatometer was equipped with conventional sports fishing tackle. After a fish was hooked, a remote camera recorded swimming activity for subsequent analyses. Tension in the line was measured indirectly by a hydraulic strut and pressure was transmitted to an instrument where it was registered on a meter. Line velocity was determined by running the fishing line over a pulley which activated a tachometer. Iron particles, evenly spaced on a fishing line, in conjunction with a

104

F. W. H. BEAMISH

magnetic tape recorder head and an oscilloscope further facilitated measurement of the speed at which a hooked fish pulled line from a reel (Walters and Fierstine, 1964). Measurements of swimming speed made in this way assume that once hooked, fish swim along a straight course. Walters and Fierstine ( 1964) suggested that reliable estimates of burst speed could be made only within the first 10-20 sec after a fish was hooked. Thereafter changes in velocity occurred which they attributed to fish altering their direction of swimming. Other factors may also reduce line velocity including fishing line drag and probable turbulance induced in the boundary layer of the fish by the line and lure (Walters and Fierstine, 1964). The procedure of attaching a line to fish has been applied also in laboratory studies on the swimming capacity of fish (Ohlmer and Schwartzkopff, 1959).

3. PHOTOGRAPHY AND TELEVISION Motion picture records have been employed to determine swimming capacity of tunas at sea or in large outdoor tanks, (Magnuson and Prescott, 1966; Yuen, 1966; Magnuson, 1967, 1970, 1973; Dixon, Chang, Byles, and Neill, personal communication). High speed movie cameras are positioned so that the direction of swimming is at right angles to the long axis of the camera lens. Swimming speed, estimated from photographs taken at sea, can be corrected for the vertical component of motion caused by rolling and pitching and for the horizontal component when the ship is moving forward, by realignment of the images in the manner described in detail by Yuen (1966). Television was used by Brawn (1960) to estimate swimming performance of Atlantic herring, Clupea harengus. The television camera was attached at one end of a towed cage in which herring were forced to swim. The camera provided a visual record of swimming behavior at depths to 30 m. A current meter suspended within the cage measured the main current velocity, and a small streamer tag attached at the center of the chamber, where it was easily seen on the television screen, gave visual indication of the direction of minor currents.

4. ACOUSTICS

The development of echo sounders, ultrasonic pulse sonar (Nishimura, 1963; Komarov, 1971), and tags containing radio or ultrasonic transmitting devices has greatly improved measurements of sustained cruising and schooling speeds, and to a lesser extent, routine velocities (Trefethen, 1956; Johnson, 1960; Novotny and Esterberg, 1962; Bass and Rascovich, 1965; Henderson et al., 1966; McCleave et

2. SWIMMING CAPACITY

105

al., 1967; Poddubny, 1967; Hasler et al., 1969; McCleave and Horrall, 1970; Yuen, 1970; Dodson et al., 1972; Madison et al., 1972; Tesch, 1974). Echo sounders, particularly those which operate in the higher range of frequencies (above 38 kHz), have been used to measure the sustained cruising and schooling speeds of fish with only moderate success. This method suffers in that, in general, it is not possible to identify fish to species without representative catches within the area under investigation. Further, it is extremely difficult to follow the swimming activity of individual fish for extended periods of time. The technique of echo sounding is described in numerous articles which should be consulted for a comprehensive understanding of the method (Cushing, 1957; Harden-Jones and McCartney, 1962; Dowd, 1967; Midttun and Nakken, 1968; Craig and Forbes, 1969; Dowd et al., 1970). Ultrasonic tags, whi'ch in water may approach neutral buoyancy, are customarily attached to the dorsal musculature (Tesch, 1974), dorsal fin (McCleaveet al., 1967; McCleave and Horrall, 1970), or placed in the stomach of fish (Henderson e t al., 1966; Hasleret al., 1969).The effective transmission range of sonic tags varies from several hundred meters to over 2 km which allows for the tracking of fish with a minimum of interference to their behavioral pattern. The life expectancy of sonic tags is dependent on battery size, which in turn is dictated, at least in part, by the size of the fish to be tagged. Thus tags applied to white bass, Morone chrysops, transmitted sound for about 20 hr (Hasler et al., 1969), considerably less than the Sweek expectancy from sonic devices attached to the larger American shad, Alosa sapidissima (Dodson et al., 1972) and European eel, An-guilla anguilla (Tesch, 1974). The pulsed sound emitted from the tags is received by a hydrophone housed in the tracking vessel. Navigational instruments facilitate measurement of the distance covered by the tracking vessel. Currents which may assist or hinder the movement of fish are difficult to determine with the precision necessary to correct for actual swimming speed. Further, with the methods presently available, it is not yet practical to clearly define distances covered by routinely active fish, such as when foraging. However, among migrating fish, sonic tags have contributed greatly to the present understanding of their speed of movement. The development of pressure sensing ultrasonic transmitters by Stasko and Rommel (1974) has facilitated the measurement of swimming depth. Gayduk et al. (1971) measured swimming depth by attaching transmitters containing high-frequency photoresistors to fish. The photoresistors provide a measure of intensity of illumination

106

F. W. H . BEAMISH

which can be translated to vertical location assuming a constant attenuation of intensity with depth. Swimming speed can be measured in the field by any of several methods. However, few allow for the identification of the category of swimming either through their failure to record a fish’s progress for sufficient time or to provide a measure of fatigue. Among the methods employed, ultrasonic tags allow for a quantitative description of sustained swimming speed of migrating and, often, schooling fish. The shortcomings of the acoustic method is that it fails to define clearly the distance moved by routinely active fish. Methods for the identification and quantitative measurement of prolonged and burst swimming are not well developed. Pressure transducers have not been applied to the measurement of locomotion but offer an optimistic alternative to those methods presently in use.

B. SWIMMINGCAPACITY Quantitative categorization of performance in free swimming fish is frustrated by environmental and biological constraints. Sustained and prolonged speeds are difficult to distinguish because fish seldom maintain a given speed for as long as 200 min. Actual swimming speed may be enhanced or hindered by currents that often are not measured. Thus migrations may be achieved largely on the strength of currents with little expenditure of energy for locomotion (Radakov and Solovyev, 1959; Zaitsev and Radakov, 1960; Harden- Jones, 1968; Royce et al., 1968; Hasler et al., 1969). Most field measurements, exclusive of burst, have been made for migrating or schooling fish. Until quantitative procedures become available, it is convenient to consider the performance of migrating and schooling fish under sustained swimming with the realization that the range of recorded velocities may well extend to prolonged. Speeds achieved and maintained by individuals or a portion of a group of fish for under 20 sec and resulting either in fatigue or greatly reduced performance are classified within burst.

1. SUSTAINEDSWIMMINGSPEED a. Cruising.Estimates of swimming performance by migrating fish based on tag and recapture studies suffer in that they assume a straight line course from the point of tagging to the place of recapture. Further, they fail to provide any information on the variation in swimming speed during the migration. The literature on tag and recapture

2. SWIMMING

CAPACITY

107

studies is large, and in the light of recent studies with ultrasonic tags will not be dealt with in this chapter. Several of the more excellent references on tag and recapture studies include Dahl(1937),Dahl and Samme (1938),Huntsman (1942),Pritchard (1944),Maar ( 1947),Dannevig (1953),Rasmussen (1959),Trout (1957),Fridriksson (1958), Liihmann and Mann (1958),Marty (1959),and Maslov (1960). Ultrasonic tagging experiments have shown that the speed of migration depends on numerous factors including the negotiation of currents, topography, shoreline, temperature, meteorological conditions, and the physiological status of the fish (Malinin, 1973).Migratory fish seldom follow a strictly linear course although directionality in their pattern of movement is generally apparent (Poddubny, 1967;Hasler et al., 1969; McCleave and Horrall, 1970; Yuen, 1970; Madison et al., 1972;Dodson et al., 1972). Swimming speed among some migratory species appears related to the proximity of the land, which may serve as a navigational aid, while in others differences are not demonstrable. Cutthroat trout, Salmo clarki, in open water swam at an average speed of 22.9 cm sec-l, appreciably less than the 36.6 cm sec-' recorded when swimming near shore (McCleave and Horrall, 1970;see also Table I). In contrast, the migratory speed of adult sockeye salmon, Oncorhynchus nerka, did not differ with distance from shore (Madison et al., 1972).The migratory speeds of sturgeon, Acipsenser nudiventris, sevryuga, A. guldenstadtii, and Atlantic salmon in freshwater rivers approximated those in the sea (Malinin, 1973). Among some anadromous species, entry into freshwater may temporarily impair progress in the upstream migration. American shad tracked from the sea to freshwater exhibited considerable wandering and passive drift within the confines of the estuary (Dodson et al., 1972).Further, mean swimming speed was reduced from 75 cm sec-I in the saline environment to about 39 cm sec-' in the estuary (Fig. 1). On entry into the estuary of the Miramichi River from the Atlantic Ocean, Stasko (1975)found Atlantic salmon tended either to drift with the tidal currents or hold station for as much as 14 days. Fish that achieved overall upstream progress did so by drifting with the flood tide and stemming ebb currents. Both Dodson et al. (1972)and Stasko (1975)considered this reduction in performance a manifestation of the physiological stress encountered in the transition from salt to freshwater. Light conditions appear also to influence the speed and direction of migrants (Hasler et al., 1958). Pronounced die1 fluctuations in mi-

108

F. W. H. BEAMISH Table Sustained Cruising Speeds ~~

Swimming speed Range Length (em)

Species Carcharhinus leucas Acipenser Acipenser Acipenser Adpenser nudiventris Acipenser nudiventris Acipenser nudiventris Acipenser guldenstadtii ' Acipenser guldenstadtii Acipenser guldenstadtii Anguilla anguilla Alosa sapidissima Alosa sapidissima Alosa sapidissimu Oncorhynchus gorbuscha Oncorhynchus gorbuscha Oncorhynchus gorbuscha Oncorhynchus keta Oncorhynchus keta Oncorhynchus kisutch Oncorhynchus kisutch Oncorhynchus nerka Oncorhvnchus nerka Oucorhyuchus tshuw!ltscho Salmo clarki Sulmo clarki Salmo gairdneri Salmo gairdueri Salmo solar Salmo salar Salmo salar Sulmo sulor Esox lucius Abramis brama Lota lota Mororie chnjsops Tautoga onitus Makaira nigricaus Pleurouectes platessa

ca. 110 ca. 110

Number

9 6

cm sec-'

Mean

e see-'

18-202 130-500

1.2-4.5

25-58

0.2-0.5

ca. 110

ca. 110

80-130

70-120 69-96

64

58

55 56 65 84

.US 72 42-59

11 886 5 401

E-173

0.2-2.0

11

23-78

0.4-1.2

1 228 834 18 37 42 7 2

30-162 40-158 33-99 4-176

0.5-2.9 0.7-2.8 0.5-1.7 0.05-1.8

0-82 0-44.7 9-44 0.226

0-1.3 0.1-0.6 <0.01-0.7

Maximum"

cm sec-'

P sec-'

165 15 33 35 25 15 23 27 33 55-72 75 39 64 55 29

1.5 0.14 0.3

54 7-8 31 52-96 53-97 4659 154-176 22-37

0.1-0.2

cm sec-'

70

e secl

0.5-0.9

300 0.3-0.5

0.60.9

173

2.2

0.5 0.9-1.7 0.9-1.7 0.7-0.9

162 248 133-170

2.9 4.4 2.0-2.6

o.ai.8 82

54 50-70 5.3-87 ca. 80 ca. 40 ca. 50 2638 47 ca. 215-360 43

15

44-69 12-25 0-17 5-13

26

225 0-0.2

5-2 1

0.1-0.3 16.7 0.2-0.7

27.5-90

0.62.1

1

1

5.5

0.07

0.3 13 60 61-95

0.4 1.3 0.20.5

104 300

86228

1.3 3.8

0.3-0.7

" Above true sustained speed. Data reported in Malinin (1973).

gratory swimming speeds of sockeye salmon were reported by Madison et al. (1972). Maximum speeds of about 70 cm sec-' (60-90 cm sec-'; Fig. 1) occurred at midday (1200 hr) with minimum speeds averaging near 35 cm sec-* (20-50 cm sec-l) during the hours of darkness (2300-0200 hr). Periods of greatest change in swimming speed coincided approximately with sunrise and sunset, which together with

2.

109

SWIMMLNG CAPACITY

I for Migrating Fish

Time 1-11 days 1 hr I day 1 day Many hr Many h r Many hr Many h r Many hr Many hr 1-19 hr Many hr Many hr Many hr M m y hr Many h r :&SO hr Many hr Mtmy hr 1-4 hr .<30sec <30 sec 2-66 hr < 1-16 hr 1113.5 hr 3.5-18.5 hr 1-0 hr 10 hr

hliiny hr ,Many hr 3-86 hr 24 hr 22-168 hr 48 hr 41-10.6 hr 2.9 hr 1-22.5 hr 2.3 hr

Temperature (“C)

20 20 20

5-6

15.4-23.3 15.4-23.3

5

0.5

21.7

Location

Reference

Lake Nicaraguia and Caribbean Sea Volga River, U.S.S.R. Volga River, U.S.S.R. Saratov HES, U.S.S.R. Volga River,U.S.S.R. Volga River estuary, U.S.S.R. Volga River, U.S.S.R. Kura River, U.S.S.R. Volgti River, Rybinsk, and Volgogradsk Reservoir. U.S.S.R Kuban River, U.S.S.R. North Sea Long Island Sound, Atlantic Ocean Connecticut River estuary, U.S.A. Connecticut River U.S.A. Amnr River, U.S.S.R. Amur River estuary, U.S.S.R. Pacific Ocean Amur River, U.S.S.R. Amur River estuary, U.S.S.R. Columbia River, U.S.A. Somass River, Canada Somass River, Canada North Pacific Ocean Columbia River, U.S.A. Yellowstone Lake, U.S.A. Yellowstone Lake, U.S.A. Culuml,ia River, U.S.A. Lake Chuzenji, Japan Pechora River, U.S.S.R. Baltic Sea N-Tulemskove Reservoir. U.S.S.R. Mirainichi River, Cnnada R y l h s k Reservoir, U.S.S.R. Withiun River, U.K. Sogozha River, U.S.S.R. Lake Mendota, U.S.A. Atlantic Ocean, U.S.A. Pacific Ocean North Sea

Thorson (1971) Malinin et al. (1971) Malinin et 01. (1971) Malinin et al. (1971) Batchykov (1963Y Pavlov ( 1969)b Malinin (1973) Derzhavin ( 1922)b Poddubny (1967) Malinin (1973) Teach (1974) Dodson et ol. (1972) Leggett and Jones (1973) Leggett and Jones (1973) Krykhtin ( 1964P Krykhtin (1964)O Stasko et al. (1973) Soldatov ( 1912Jb Krykhtin ( 1964)b Johnson (1960) Ellis (1966) Ellis (1966) Madison et nl. (1972) Johnson (1960) McCleave and Horrall (1970) McCleave and LaBar (1972) Johnson (1960) Kazihiira rt a / . (1969)” Letovaltseva ( 1967)b Carlin (1968)b Malinin (1973) Stasko (1975) Poddubny et 01. (1970) Langfnrd ( 1974) Malinin (1971) Hasler et 01. (1969) Olla et al. (1974) Yuen et or. (1974) Greer et nl. (1971)

other information on angular change prompted the hypothesis of a daily reorientation process related to light conditions. Migrating adult chinook salmon in accord with sockeye also exhibit a reduction in swimming speed during the hours of darkness (Johnson, 1960).Differences between day and night migratory behavior of the European eel were not demonstrable; however, the prevailing influence of the tides

110

F. W. H. BEAMISH

-'8

12m

u)

E

A 0. r m r k a l 0. kisutch

Lcipnser luldenstaedti

.

S. clarki

Acl inner nu1 entrls

M.&ryMpa

0

I

Aclpenser guldenstaedtl

i

I

I

1

-

1

1

I

120

80

Length, cm 1 2 0 t Oncorhynchus nerka

B

0000

1200 a00 Time of Day

1200

Salt (28%)

Estuarine (0-28960)

Fresh (Oh01

Fig. 1. Sustained cruising speeds of migrating fish. (A) Mean and range of sustained speed in relation to length (see text). (B) Die1 fluctuations (redrawn from Madison et al., 1972,J. Fish. Res. Board Can.). (C) Mean speed on entry into freshwater from the estuarine and marine environment (Dodson et al., 1970, 1971, 1972).

precluded a valid comparison of the effects of light and darkness (Tesch, 1974). Among migrating schools of skipjack tuna, variation occurred in the spatial distance covered, although not necessarily in swimming speed (Yuen, 1970). During the hours of darkness, skipjack

2.

111

SWIMMING CAPACITY

traveled from 25 to 106 km, equivalent to speeds approaching 100 cm sec-'. By day the tuna remained in a reasonably confined area where they engaged in feeding, Under conditions such as these, ultrasonic devices often fail to provide a precise description of the actual distance covered. In fact, in the case of Yuen's tuna, the calculated speeds while feeding during the daylight hours were less than those necessary to maintain hydrostatic equilibrium (Magnuson, 1970). Mean sustained cruising speeds of fish migrating in a horizontal plane lie mainly from 0.5 to 0.2 C sec-' (Fig. 1 and Table I). A low mean speed of swimming may have adaptive ecological significance by increasing that energetic efficiency of migration. Wiehs ( 19734 calculated that the maximum distance that can be accomplished for a given expenditure of energy occurs when the energy utilized in propulsion is equal to the resting metabolic rate. In rainbow trout, Salrno gairdneri, the velocity for which this is the case approximates 1 C sec-'. Considerable data are available on swimming speeds of scrombroid fishes (Tables I1 and 111), many of which are negatively buoyant (Magnan, 1929; Watanabe, 1942; Aleev, 1963) and include bonitos, mackerels, and tunas. To overcome negative buoyancy

Table I1 Sustained Cruising Speeds Necessary to Overcome Negative Buoyancy among the Scombroid Fish" Swimming speed Mean

Range NumSpecies

Acanthocybium solanderi Auxis rochei Euthynnus afinis Euthynnus pelamis Sarda chiliensis Thunnus albacares Thunnus obesus Thunnus obesus

ber

Length (cm)

cm sec-l

31 36 38-48 52 35 55 36

55-76 71-89 39-125 47-83 41-62 55-94 43-60

sec-'

cm sec-I

't sec-I

1.8-2.5 2.0-2.5 1.0-3.3 0.9-1.6 1.2-1.8 1.0-1.7 1.2-1.7

68 76 70 83 46 66 48

0.3 2.2 2.1 1.6 1.6 1.3 1.2 1.3

t?

12 12 12 12

" Data from Magnuson (1973). Observations were made in large tanks 23"-26"C over a period of 1-4 hr.

112

F. W. H. BEAMISH

scrombroids must swim continuously (sustained cruising) with pectoral fins extended which produces a lift that balances their weight in water (Aleev, 1963; Alexander, 1967, 1968; Magnuson, 1970, 1973). Observations by Magnuson (1973) on the sustained cruising speeds necessary to overcome negative bouyancy for seven scrombroid species indicate a range from 0.3 to 3.3 C sec-' (Table 11). The three species with the fastest speeds necessary to counter negative buoyancy, skipjack, kawakawa, Euthynnus afinis, and Pacific bonito, Sarda chiliensis, are without a swim bladder. Wahoo, Acanthocybium solanded, the species with the lowest predicted speed has the largest bladder and the lowest density. Among the three species without swim bladders, skipjacks are the heaviest for their length, have the smallest lift area of pectoral fins, and have the highest range of minimum speeds. Many fishes display regular diel vertical migrations which are of considerably smaller magnitude than those displayed in the horizontal direction. Patterns of vertical migration are best described for marine demersal and some pelagic species (Hjort, 1914; Saetersdal, 1956; Konstantinov, 1958; Richardson et aZ., 1959; Brunel, 1964; Parrish et aZ., 1964; Sundnes, 1963; Woodhead, 1964; Beamish, 1966a).Observations on vertical migration have been made almost entirely with echo sounders. This method generally precludes accurate information on the actual size of the migrants or on the rate of migration by individuals. Rather, it does provide information on the general rate of movement by concentrations of fish. In addition to the practical problems of obtaining quantitative measurements, the extent of the migration itself depends very much on factors such as season and depth (Venvey, 1960; deVeen, 1964; Beamish, 1966a; Hawkins et al., 1974). The vertical distance migrated b y haddock, Melanogrammus aeglefinus, at night may be as much as 1000 m (Hjort, 1914) but more frequently is around 100 m (Beamish, 1966a). Atlantic cod, Gadus morhua, display diel vertical migrations of under 70 m (Richardson et al., 1959). Redfish, Sebastes marinus, usually remain within 50 m from bottom (Beamish, 1966a). While the exact rate of ascent or descent is unknown, most appear to be completed within an hour. Assuming steady swimming over a 1-hr period, the mean rate of swimming in most cases is under 5 cm sec-'. Echo sounder records made at sea from a stationary vessel indicate that periods of steady swimming alternate with periods of unsteady activity; however, the relative proportion of each is not clear.

Tabie I11 Sustained Schooling Speeds of Fish Swiniiiiing ?peed Range Species

Number

Length (cm)

Clupea harengus harengus

ca. 25-30

Clupea harengus harengus Clupea harengus harengus Clupea harengus harengus Clupea harengus harengus Clupea harengus pallasi Clupea hurengus pallasi Engraulis encrasichnlus Engraulis encrasicholus Seriola quinqueradiata Euthynnus afinis Euthynnus afinis

ca. 2.530 ca. 5-30 ca. 2 5 4 0 Juvenile ca. 28 15

Euthynnus pelamis Euthynnus pelamis Euthynnus pelamis Euthynnus pelamis Thunnus alhacares Thunnus alhacares Thunnus alhacares

1

10 510 2

33

Data reported in Blaxter (1967). Data reported in Radakov ( 1964).

80 35-42 40 49-57 48-79 44 ca. 40

ca. 50 ca. 50 S2

cni sec-'

Mean P sec-'

cm w c - '

9-12 50-100

<25178

5W1000 3C690 20-435 6XL850

Temperature

("C)

0.5

2-5

0.h2.3

9.7-19.4 0.4-14.3 059.9 16-21

150 90

5-6 3 4-5 7 7.5

Hour-

200 113

0.8

8 hr 1-4 hr

65-80 75 60

82 200

2.C2.3

770

14.8 2.5-4.2

200

1100

100-150 160-540

t sec-'

0bsen.atinn time

3.1-10.4

18.5

2.>26 25

<3 cec 7 days 2-3 hr

27-29 24-26

Norwegian Sea

Fridriksson and Aasen ( 1952) Jones ( 1957) S c h d e (1960) Harden-Jones (1962) Hempel" Chestonoya High and Lusz ( 1966) Lebedev (1936)* Radakov (1962)" Ichihara et al. (1972) Slagnuson (1967) Diznn et 01. (personal communication) Watanabe (1942) Yuen (1966) Yuen (1970) Neill (personal communication Kishinouye (1923) Nishimura (1963) Yuen (1966)

Pacific Ocean Tank Tank Pacific Ocean Tank Tank

25

Pacific Ocean Pacific Ocean Pacific Ocean Pacific Ocean

2426

Pacific Ocean Pacific Ocean Pacific Ocean

12

Z3 < 3 sec

Reference

North Sea North Sea North Sea Tank

20min

5

Location

114

F. W. H. BEAMISH

The *swimming depth of American eel, Anguillu rostruta, was monitored in Passamaquoddy Bay, Bay of Fundy, using pressure sensing ultrasonic transmitters (Stasko and Rommel, 1974). Eels made frequent dives from the surface to bottom during the hours of daylight and darkness at speeds of 0.8-1.1 t' sec-'. The maximum rate of ascent was 0.6-0.8C sec-'.

b. Schooling. Schools of fish are recognizable by the regular array of the component individuals and the synchrony displayed in their swimming speeds (Shaw, 1960). Belyayev and Zuyev (1969) and Wiehs ( 1973b) have suggested that the organized spatial distribution of fish within a school may, in fact, reduce the hydrodynamic resistance of individuals, thus providing for increased swimming efficiency. Swimming speeds for the schooling clupeoid fish have been reported from under 1 to more than 7 t' sec-' (Lebedev, 1936, reported in Radakov, 1964; Fridriksson and Aasen, 1952; Jones, 1957; Scharfe, 1960; Radakov, 1962, reported in Blaxter, 1967; Chestnoy, reported in Radakov, 1964; High and Lusz, 1966; see also Table 111, Fig. 2). The higher velocities were recorded for herring swimming ahead of towed nets, for short periods of time, and probably represent prolonged speeds. Blaxter (1967) reported that a relative speed of 3-4 C sec-' is a valid estimate of sustained schooling performance among the clupeoid fish. Swimming speeds of schooling scombroids, including the velocity necessary to counter negative buoyancy, have been reported from 2 to 21 4 sec-' (Table 111). Yuen (1966) calculated speeds of 30-690 cm sec-' for feeding skipjack tuna from photographic records and found no demonstrable dependence on length (48-79 cm). Watanabe (1942) observed skipjack tuna as they swam by a measured distance of 2 m alongside a ship. The mean speed recorded for ten individuals (52 cm) was 770 cm sec-' with a range of 500-1000 cm sec-'. Yuen (1970) provides information from which it may be calculated that a 44 cm skipjack tuna maintained for 107 min a minimum average speed of 435 cm sec-' (9.9 t' sec-l). Such a speed is, according to Neil1 (personal communication), consistent with reports of skipjack schools "disappearing over the horizon after being pursued by fishing vessels at 650 to over 850 cm sec-I for 2 or 3 hours." Whether, in fact, tuna fatigue at these speeds is as yet unanswered. Among the other tunas for which speeds attributable to sustained schooling have been reported, Yuen ( 1966) calculated velocities of 160-540 cm sec-' for a school of yellowfin tuna (52 cm) higher than the 100-150 cm sec-' reported by Nishimura (1963) for this species.

115

2. SWIMMING CAPACITY

200

C lupea harengus

c

150

I

0

t!

5 UP)

g

i

100

cn

D

.-E E .-E s

50

0

0

10

20

30

Length, cm Fig. 2. Field and laboratory observations on the swimming capacity of herring, Clupea harengus, in relation to length. Sustained schooling and prolonged speeds are indicated by closed circles. Speeds reported as burst are recorded as open circles. [From Fridriksson and Aasen, 1952; Jones, 1957; Blaxter and Dickson, 1959; Bishai, 1960; Brawn, 1960; Schirfe, 1960; Boyar, 1961; Blaxter, 1962; Chestnoy (cited in Radakov, 1964); Blaxter and Parrish, 1966; High and Lusz, 1966.1

Kishinouye (1923) estimated the speed of yellowfin tuna at 1100 cm sec-'. Dizon, Chang, Byles, and Neil1 (personal communication) observed speeds of 200 cm sec-' for kawakawa (40 cm) feeding in large outdoor pools. The hydrodynamic and vascular achievements in tunas (Barrett and Hester, 1964; Stevens and Fry, 1971; Linthicum and Carey, 1972; Graham, 1973, 1975; Magnuson, 1973; Stevens et al., 1974) are manifested in their relative swimming performances which exceed by severalfold those recorded for any other schooling species. It is, of course,

116

F. W. H . BEAMISH

not possible to categorize tuna performance, without question, into prolonged and sustained schooling speeds. However, that relative velocities of 2-21 t sec-I can be maintained for at least several hours by skipjack tuna without fatiguing, supports the suggestion of high sustained schooling speeds. The lower speeds reported for tuna undoubtedly represent sustained routine and cruising speeds, rather than schooling performance.

2. BURSTSWIMMINGSPEED In an entertaining article, Lane (1941) described the maximum speeds which free swimming or tethered fish could achieve but based his measurements on estimated distances or movements in streams with little attention to currents. In a quantitative laboratory study on burst performances, Bainbridge (1958) supported the view that fish up to 1 m in length should be able to swim up to 10 times their own body length but only for a brief period of about 1 sec, beyond which velocity would decrease exponentially. Recent investigations suggest variability in burst swimming among species but that for at least some the relative performance maximum of l o t sec-' is a conservative measure. Speeds of more than 20 t sec-' may better represent the upper threshold for steady or sprint burst swimming (Table IV). Thus the speed at which alewives, Pomolobus pseudoharengus, migrated upstream through a fishway ranged from 14 to 208 sec-' (Dow, 1962; see also Table IV). Fierstine and Walters (1968) recorded burst acceleration velocities of 11.9 t sec-' for a 0.19 sec period by a yellowfin tuna and 18.4 t sec-' for a wahoo over the short time span of 0.05 sec. Skipjack tuna have been recorded in pursuit of food at speeds of over 20 t sec-' (Dizon et al., personal communication). Komarov (1971) obtained speeds of 18.4 and 21.6 c sec-I for golden mullet, Mugil aurutus, and mullet, Mugil saliens, respectively, but it is not clear whether the measurements were made in the field or in the laboratory. Of the species examined by Komarov (1971), carp, Cyprinus carpio, and green wrasse, Creilabrus tinca, were among the poorest swimmers but still registered relative velocities of 7-8 t sec-'. Maximum swimming speed was calculated by Wardle (1975) on the basis of contraction time of white lateral muscle and the relation between tailbeat frequency and forward motion, summarized by the equation:

where Urnax,maximum swimming speed (m sec-I); SL, stride length [0.6-0.8 times fish length (Bainbridge, 1958)l; t,fish length (m); tM, muscle contraction time (sec). Application of the formula indicates

2. SWIMMING CAPACITY

117

maximum velocities for a 10 cm fish of 8-15 sec-'. Maximum speeds forfishof5Ocm wereestimatedto liebetween 6.4and 10.6Csec-'. Actual measurement of swimming speeds were calculated from videotape recordings of fish stimulated into movement or conditioned to race between two feeding points by association of underwater flashing lights with the appearance of food. Haddock and sprats, Clupeu spruttus, both 10 cm in length, achieved velocities of 26 8 sec-' while the maximum speed recorded for salmon (25-28 cm) was 10 8 sec-'. A burst acceleration speed (0.06 sec) of 16.9e sec-I was estimated for a sprat of 8.3 cm. The capacity for burst swimming is, for many species, a prerequisite for their continued well being and even existence. The successful completion of the spawning migration by those species that ascend fast flowing waterways may depend on their burst performance capacity. Burst swimming is important also in escaping predators and in capturing food. Recently, Wiehs ( 1974) offered the intriguing suggestion that by alternating periods of fast swimming and motionless gliding, fish can reduce the expenditure of energy required to cover a given distance by over 50%.

111. LABORATORY MEASUREMENTS OF PERFORMANCE

The role of environmental and biological factors is perhaps most easily investigated in the laboratory. The advantages offered by the laboratory, however, are countered to at least a considerable extent by the difficulties experienced in the design of a suitable swimming chamber. A. Swimming Chamber Design 1. SPONTANEOUS ACTIVITY Swimming activity exhibited by fish that are neither confined nor forced to perform frequently has been measured in conjunction with oxygen consumption, often with the specific intent of estimating basal (standard or resting) and intermediate levels of metabolism. The first qualitative measures of spontaneous activity were made by Spoor (1946). He equated activity to the number of deflections of an aluminum paddle located within an enclosed metabolic chamber to movements of a fish. Corti and Weber (1948) designed an activitymeasuring apparatus which consisted of a delicately suspended experimental tank which moved with the fish's activity. Deflections of a glass plate suspended in a respiration chamber caused by water movements which were created when a fish was active were equated to activity by Fontaine (1956). Beamish and Mookherjii (1964) used a

Table Burst Swimming Velocity Species Ceruharhiiius leucus Negqirion breuirostria Alliula oulpes Anguilla oulgaris Aloss sapidissinin Clupea horengus Clupeu horengus Chrpea hurengus Clupea spruttus Poniulobus pseudoharerigus Pornulohus pseudohorengus Eiigruulis encrosicholus poiitioira Onciirhynchus kisutch Oricorhynchua kisutch Oncorhynchus kisrrtch Oncorhynchus rierka Oncorhyrlchus rierka Oticorhynchus tshawytscho Solrno fario Salmo gairdneri Salmo gairdneri Sirlrno goirdiieri Salmo gairdneri S d m o gairdneri Salrno irideus Solnio irideus

Nlim-

Length

Ixr

(cin!

1 1

1

55 1551 1551 4 21 -

152 184 <90 60 1.0-26 6.0-21.9 6.0-21.9 >75 27.1-31.2 12.1

9

44-67 51-76 36-61 5,569

5

4 15

-

-

6 1

51-97 25

10 6

61-81

4 4

Salnio salar Solmo sulur

58.67 14.3 14.3 10.3-28.0 10.b28.0 35 -

75-85

Sulrno salar Sulaio solar sulrrlo trlrtta

Soloelinus fontinalis Euthynnus pelamis Esox Esox lucius Corassius aurutus Corassius auratus Carassius auratus Carassius auratus gibelio Chalcolburnus chaleoides montoides Cyprinus carpio Cyprinus carpio Hypophtholmichthys molotrix Leuciscus leuciscus Leuciscus leuciscus Leuciscus leuciscus Rutilus rutilus Tlnco tinca Catastomus occidentalis Gadus aeglejina Gadus morhuu Gadus merlongus Gadus oirens Odontogadus merlongus auxinas

19 1

10-20

IS37 11.2 40-50

1

7 8 8

-

9.0 6.7-21.3 6.7-21.3 23.0 12.5 35.0

-

3

-

7 7

10 6 17 13 17

-

27.0 10.0-21.4 10.0-21.4 25.5 37-43 2 4 41 12-57 14-20 12-19 18.0

118

cm sec-I

e sec-’

522 244

3.4 1.3

1000

>11.1

114 S35CL402 6-150 67-131 40-104 24-36 415-485 304 162 161-216 22CL.372 287-533 155-203 268-313 543-6638 350 550-1050 186226 536-817 30-250 30-180 105-270 32-73 347 300 429-Mx) 805 137-305 93 590-1370 360-450 138 74-200

4e80 226 187 236

-

248 410 110-240 4690 455 138 305 90-180 90-240 70- I80 60-122 19.5

1.9

Time (sec)

-

2-5

5.8-6.0 6.2-9.7 5.1 3.2-4.8 13.615.7

2-5 30 30 2 5 2.66.2

13.4 2.7-4.4 4.3-6.4 6.2-9.2 2.2-3.1


-

6.9-10.7

-

1.5

10.0

2.8-3.7 7.5-13.4 2.0-17.5 2.0-12.6 7.610.2 2.63.1

10-15 1.5 0.08 0.04 1

20

-

5.8-8.4 8.2-10.5 8.3 13.5-18.6

2-5 10 0.2

15.3 9.4-11.0

3.8-6.3 9.8 15.0 8.2 5.2 9.2 11.0-11.2 4.2-4.4 7.5 7.1-8.3 3.2-4.4 4.7-7.5 5.0-9.0 5.0-6.4 10.8

2-5

1 20
20

IV Speeds of Fish Temper:itnre ("C)

10.0-15.0 5.0-18.0 1.4-11.2 1.4-11.2 < lo-> 15

10

15

Comments

References

Hook and line Hook and line Hook and line Timed over measured distance (swimming tunnel) Timed over measured distance (fishway) Timed over measured distance (swimming tunnel) Annular trough Annular trough Timed over measured distance (swimming tunnel) Timed over measured distance (fishway) Timed over measured distance (fishway) Distance covered by one tailbeat Annular trough Annular trough Timed over measured distance (fishway) Annular trough

Gero (1952) Gero ( 1952) Lane (1941) Blaxter and Dickson (1959) Weaver (1965) Blaxter and Dickson (1959) Boy= ( 1961) Boyar (1961) Blaxter and Dickson (1959) Dow ( 1962) Stringham (1924) Komarov (1971) Paulik and DeLacy (1957) Paulik and DeLacy (1957) Weaver (1963) Paulik and D e h c y (1957) Napier (1914)' Weaver ( 1963) Denil ( 1937) Lane (1941) Paulik and DeLacy (1957) Weaver (1963) Webb (personal communication) Webb (personal communication) Bainbridge (1960) Bainbridge (1960) Denil (1909) Lavollee ( 1902)' Denil ( 1937) Lane (1941) Blaxter and Dickson (1959) Peterson (1974) Neil1 et al. (personal communication) Lane (1941) Stringham (1924) Blaxter and Dickson (1959) Bainbridge (1960) Bainbridge ( 1960) Komarov (1971) Komarov (1971) Komarov (1971) Regnard (1893) Komarov (1971) Lane (1941) Bainbridge (1960) Bainbridge (1960) Lane (1941) Komarov (1971) Wales (1950) Blaxter and Dickson ( 1959) Blaxter and Dickson ( 1959) Blaxter and Dickson (1959) Blaxter and Dickson ( 1959) Komarov (1971)

Timed over measured distance (fishway)

7

Hook and line Annular trough Timed over measured distance (fishway)

14 14

Annular trough Annular trough

10-15 15 25

Timed Over measured distance (fishway) Timed over measured distance (swimming tunnel) Swimming tunnel Timed over measured distance (annular tank) Hook and line

< 10.0 14 14

14 14

11.5-17.0 9.5-12.0 9.0-17.0 9.0-12.0

Timed over measured distance (swimming tunnel) Annular trough Annular trough Distance covered by one tailbeat Distance covered by one tailbeat Distance covered by one tailbeat Annular trough Distance covered by one tailbeat Tank observation Annular trough Annular trough Tank observation Distance covered by one tailbeat Timed over measured distance (culvert) Timed over measured distance (swimming tunnel) Timed over measured distance (swimming tunnel) Timed over measured distance (swimming tunnel) Timed over measured distance (swimming tunnel) Distance covered by one tailbeat

(Continued) 119

120

F. W. H. BEAMISH Table IVVelocity Specie3

Zoarces uilii parous Gasterosteus spinarhiu Potnicrops itajora Serrunus seribo Gyintiocephalus cernua Lepomis cyanellus Leponiio cyanellus Micropterus Percu Perca fluoiatilis Pornatornus salatrir Trachura mediterraneus pontious Spicara smoris Mullus barbatus pontious Creilahrus tinca Mugil curatus Mugil saliens Sphyroena barracuda Pholis gunnelus Gobius minutus Acauthocybium solanderi Acanthocyhium solanderi Scomber scomhrus Thuuricis albacores Thunnus alhucures Thunnus thynnus Istiophorus Trigla ~ p p . Pleuronectes flesus Pleurouectes microcephalus Pleuronectes platesea Pleuronectes platesea "

Number 1

1 1

Length (cm)

6.4 10.0 -

18.5 10.5

x 8

2 1 2 3 13 I 5 1 1

2 1 14I 6

11.5 16.0 15.7 17.0 11.8 15.0 24.0 1R.5 121-130 10 6.5 92-113 113.1 33.0-38.0 53-98 98 147

IR.O 27.5 8.0 0.7-1.0 6.0-15.0

cni sec-'

la21 72 174 227 133 10-70 10-150

145 204 258 235 109 107 442 400 1220 30 27 1204-2 140

2082

P sec-'

2.u-3.3 7.2 -

2-5 2-5

-

12.3 12.7 1.3-8.8 1.3-18.8

<1
12.6 12.8 16.4 13.8 9.2 7.1 10.1 21.6 9.4-10. I 3.0 4.2 12.2-21.4 18.4


1RQ-300

-

52S2072 1165

6.0-21.1 11.9 3.4

3000 129-135 105 15 3.7-15.0 690

Time (sec)

7.2-7.5 3.8 1.9 4.9-16.0


-

2-5 2-5 10-20


2-5 2-5 2-5 1-4 2-5

Data froin Lane (1941).

thermostatic heater probe to measure activity. Heat produced by the Joule effect is lost at a fixed rate if the water temperature is constant and the water is motionless. Disturbances within the water such as might be created by fish movement increase the heat loss from the thermistor; the temperature decreases and its electrical resistance increases. Recent investigators have improved the design of this apparatus (Heusner and Enright, 1966; Mathur and Shrivastava, 1970). A manometric method was employed by Ruhland and Heusner (1959) in which swimming by fish caused waves on the water surface of the metabolic chamber. One electrode was immersed in the chamber water, and the second was suspended immediately above the surface so that the generation of waves allowed for the periodic completion of the electrical circuit. The length of time the circuit was completed was recorded and equated to spontaneous activity.

121

2. SWIMMLNG CAPACITY Continued Temperature ("C) 15.0 < 10

> 15.0 > 15.0

>I5

I&> 15 < 10.0 r15 6..%7.5 < 10

Comments Timed over measured distance (swimming tunnel) Timed over measured distance (swimming tunnel) Hook and line Distance covered by one tailbeat Distance covered by one tailbeat

Tank observiition Tank observation Distance covered by one tailbeat Distance covered by one tailbeat Distance covered by one tailbeat Distance covered by one tailbeat Distance covered by one taillieat Distance covered by one tailbeat Distance covered by one taillxxt Distance covered b y one tailbeat Hook and line Timed over me;iaured distance (swimming tunnel) Timed over measured distance (swimming tunnel) Hook and line Hook and line (ocean) Timed over measured distance (swimming tunnel) Huok and line Hook and line (ocean) Hook and line Hook and line Timed over measured distance (\wimming tunnel) Timed over measured distance (swimming tunnel) Timed over measured distance (swimming tunnel) Timed over measured distance (swimming tunnel) Timed over measured distance (swimming tunnel)

References Blaxter and Dickson (1959) Blaxter and Dickson (1959) Gero (1952) Komarov (1971) Komarov (1971) Webb (persond communication) Webb (personal communication) Lane (1941) Lime (1941) Komarov (1971) Komarov (1971) Komarov (1971) Komarov (1971) Komarov (1971) Komarov (1971) Komarov (1971) Komarov (1971) Gero ( 1952) Blaxter and Dickson (1959) Blaxter and Dickson (1959) Walters and Fierstine (1964) Fierstine and Walters (1968) Blaxter and Dickson (1959) Walters and Fierstine (1964) Fierstine and Walters (1968) Lane ( 1941) Lane (1941) Blilxter and Dickson (1959) Blaxter and Dickson (1959) Blaxter and Dickson (1959) Ryland (1963) Blarter and Dickson (1959)

Locomotory activity of the American eel, Anguilla rostrata,generated water currents, causing electrical contact between an armature wire and rods suspended in the experimental tank (Bohun and Winn, 1966). von Kausch ( 1968) recorded spontaneous activity by transformation of turbulence created by fish movement into electric impulses. Recently Spoor et al. (1971) developed a technique whereby water currents, generated by the activities of fish, by their influence on equilibrium potentials cause measurable changes in the potential difference between electrodes contained within the water bath. Subsequent application of multiple electrodes by Spoor and Drummond (1972) allowed a rather precise description of the movements of fish within a large gradient tank. These methods while often extremely sensitive provide only a relative measure of activity. Several methods have been employed in

122

F. W. H. BEAMISH

recent years to provide absolute estimates of spontaneous activity. The applidation of a permanent magnet to a fish was used by Lillelund (1967), DeGroot and Schuyf (1967), and Schuyf and DeGroot (1971), together with a Hall effect probe, to measure unrestrained movement within a tank. Smit (1965), Kleerekoper et al. (1969, 1970), Beamish (1973), and others have used photocells to record and quantify spontaneous movements. High-frequency sound has received attention as a technique to record fish movements in an annular trough (Cummings, 1963; Muir et al., 1965; Meffert, 1968). Movement by fish within the ultrasonic transmission field shifts the audio frequency of the reflected signal, thereby providing a method for estimating activity.

2. FORCEDACTIVITY The design of swimming chambers can broadly be grouped into two categories: one in which the chamber is moved (Fry and Hart, 1948) and, second, where water flows through a stationary chamber (Katz et al., 1959; Blaika et al., 1960). Within each category there are numerous variations in design, performance, and cost. An annular open trough, rectangular in cross section, mounted on a rotating turntable (Fig. 3) was first described by Fry and Hart (1948)

Channel Rotation

Plan View

U

u

Elevation View

Fig. 3. Diagram of an annular swimming chamber, modified by Hammond and Hickman (1966) after that described and employed by Fry and Hart (1948). (From Hammond and Hickman, 1966,J. Fish. Res. Bourd Can.)

2. SWIMMING CAPACITY

123

and subsequently adopted in modified form by Graham (1949), Gibson and Fry ( 1954), Bainbridge and Brown ( 1958), Brett et al. ( 1958), Boyar (1961), Smit (1965), Hammond and Hickman (1966),Kutty (1969), and Fry and Cox (1970).The trough is rotated at a speed equal to that of the fish so that the animal remains stationary relative to the observer. The swimming response is attributable to the effect ofthe short radius ofrotation (Gray, 1937) and is usually elicited by a fixed visual cue such as a lighted table lamp. The speed of the water within the trough in relation to the speed of rotation of the chamber was determined by Fry and Hart (1948) by measuring the rate at which a loose ball of absorbent cotton was carried about the chamber. Slippage measured in this way was independent of temperature. A thorough examination of the velocity profile has not been described but it is likely to be complex. For a given speed of rotation, a velocity differential is to be expected between the boundaries of the inner and outer radius. Further, vertically oriented clockwise spiral currents generated in the direction of flow alter the drag relations at all surfaces. Along the lower edges of the trough cumulative drag from both the walls and bottom would exceed that at other points along the trough surface, thereby reducing the velocity in this region. To measure active metabolic rate the trough was closed (Job, 1955; Fry, 1957; Wohlschlag, 1957; Basu, 1959). A second broad category of swimming chambers includes those in which the chamber itself remains stationary and a current of water against which fish are encouraged to swim is generated by gravity flow or a pump. Chambers of this general design often take the shape of a simple tube or a series of tubes with or without expansion, and reduction cones and other devices such as screens appropriately constructed and positioned to encourage a rectilinear velocity profile within that portion of the chamber in which one or more fish are required to swim. A current generated by gravity flow through a series of small diameter tubes was employed by Bishai (1960) to measure the swimming capacity of larval fish (Fig, 4) and subsequently adopted by Ryland (1963) and Houde (1969). A reservoir located above the swimming chamber provided the head necessary to generate a current. Water from the reservoir is conducted to the swimming portion of the chamber through a small bore tube. The rate and direction of flow are controlled by appropriately located needle valves. A description of the velocity profile was not provided. However, it can be anticipated that it would not follow a rectilinear pattern in profile as the relatively large internal boundary surface together with the low velocities against which larvae could swim would encourage the formation of layers of different velocity, resulting in a parabolic velocity profile.

Bishai Cisso)

Funnel Entrance Swimmina Chamber

~

4

1 BlaZka, Volf, and Cepala

B1---'8 1---:

+Swimming

Brett

Ci960)

Chamber

Cigsa)

Screens Chamber

Centrifugal Pump

-

Thomas, Burrows, and Chenoweth (1964)

Motor

Fig. 4. Swimming flumes. The direction of flow is indicated by arrows. (Redrawn from Bishai, 1960, Thomas et ul., 1964, Smith and Newcomb, 1970.)

2. SWIMMING CAPACITY

125

When laminar flow exists, a precise description of swimming capacity can be made only with great difficulty as it is imperative not only to continuously monitor the positions of the swimming fish, but also to calibrate the chamber sufficiently well that a velocity can be assigned to each location. A stationary oval chamber in which water is circulated by a paddle wheel was first described by Lemke and Mount (1963) and subsequently modified and employed by MacLeod and Smith (1966), MacLeod (1967), and Oseid and Smith (1972). A description of the velocity profile was not provided; however, no specific attempt was made to encourage rectilinear flow. On the basis of the chamber design which was open at the top and rectangular in cross section, variations in drag are to be expected which would contribute to uneven flow. Recently, Lett (1975) used a paddle wheel to generate a current in an annular trough which was semi-circular in cross section. Even with the addition of a series of screens which acted as flow straighteners, a detailed description of the velocity profile indicated a 10% cross-sectional variation in current speed. Swimming chambers in which the current is generated either with a pump, propeller, or impellor are of three basic types (Smith and Newcomb, 1970): small volume, low horsepower; first described by Blaika et al. (1960), (Fig. 4) and later modified and employed by Smit (1965), Beamish (1966b), Rao (1968), Kutty (1969), Smith and Newcomb (1970), Hunter and Zweifel(1971), Smit et al. (1971), Shazkina (1972a),Kutty and Saunders (1973), and Peterson ( 1974);large volume, low horsepower, (Fig. 4), independently designed and used by Thomas et al. (1964), Arnold (1969), and Griffiths and Alderdice (1972); small volume, high horsepower. In the latter category several chambers have. been independently designed (Katz et al., 1959; Mar, 1959; Brett, 1964; Farmer and Beamish, 1969). Recently, Bell and Terhune (1970), in collaboration with Brett (Brett and Glass, 1973) enlarged and further refined the earlier design. Swimming chambers in this latter category have received the greatest engineering input and the velocity profiles closely approximate rectilinear flow. Each has advantages and disadvantages. A swimming chamber with small water volume provides the greatesf sensitivity for measurement of changes in dissolved gases and consequently is recommended where the objective is not only concerned with swimming performance but also in the determination of metabolic rate. Chambers with large volumes tend to provide the least turbulent flow at lowest head; however diffusion grids greatly assist in smoothing the flow pattern. Low horsepower systems reduce both

126

F. W. H . BEAMISH

construction and operating costs as well as the input of heat into the flume. This is usually accomplished at the expense of a uniform crosssectional velocity. Blaika’s apparatus (Blaika et al., 1960) consists of two horizontal concentric cylinders in which water is driven through the inner cylinder by a propeller and returns between the outer and inner cylinders. The flow characteristics of this apparatus are regarded by Smith and Newcomb (1970)as the least desirable among the three basic types of tunnel chambers described above. In part this is attributable to the spiral movement of water encouraged by the propellor. Smith and Newcomb (1970) suggest this can be reduced by the use of a jet outboard impellor. Vanes located between the outer and inner chambers and within the inner chamber at the upstream end further reduce spiral movement. A characteristic of water flowing through a tunnel of fixed diameter is the growth of a boundary layer with length resulting ultimately in a parabolic velocity profile. Excessive turbulence may be avoided b y the addition of appropriately machined end plates (Smith and Newcomb, 1970). Chambers of large volume and low horsepower are used in swimming studies not involving the determination of metabolic rate and often in dealing with large numbers of individuals at the same time. The flume described by Thomas et al. (1964) consists of a head tank into which water is forced from an axial flow pump and from which water enters the swimming portion. Fluid acceleration between the large head tank and the entrance of the chamber, circular in cross section, encouraged a nearly uniform velocity profile throughout the length of the swimming portion of the flume. Provision is made for individual fish, unable to maintain a given speed, to collect in a trap immediately downstream from the swimming chamber. Energy which is converted to heat is added continually to the water of a swimming chamber in operation mainly as the result of fluid friction. This problem is most severe when the flume volume is low and horsepower high, but can be avoided by the addition of a heat exchange system. To assure continuous swimming it is common practice to apply an electrical stimulus at the downstream end of the chamber, using metal electrodes in freshwater (Fry and Hart, 1948) or graphite rods in saltwater (Beamish, 1966b; Griffiths and Alderdice, 1972). Intermediate between chambers of a large volume and low horsepower and those of small capacity and high horsepower is that described by Katz et al. (1959), which consists of a horizontal tube in which fish swim against a current generated by a centrifugal pump

2.

127

SWIMMING CAPACITY

and regulated by a gate valve. Wire mesh screens located near the upstream end of the swimming chamber serve as baffles for dissipating gross turbulence or reducing it to finer patterns. The flow pattern based on the motion of particles suspended in water was reported as rectilinear. However, without some mechanism to impair boundary layer growth, a parabolic velocity profile would be anticipated toward the downstream portion of the swimming chamber. The application of sophisticated engineering principles to flume design was applied by Mar (1959)and Brett (1964) and more recently by Farmer and Beamish (1969).These flumes, which were designed for metabolic studies, incorporated reduction and expansion cones to impair boundary layer growth in the swimming portion of the chamber. The cones were appropriately designed to discourage the formation of a Venus contracta, manifested as an area of “dead” water, immediately downstream from a reduction cone. Grids of various mesh sizes further facilitated the production of a rectilinear plane of uniform micro turbulence. Where the velocity profile does not approach rectilinearity, quantitative estimates of swimming performance must account for the fish’s location within the swimming chamber. This is often sufficiently difficult to support the additional effort required in the design and construction of a more refined flume. For precise measurement of swimming speed it is necessary to correct also for the effect of the fish’s body on current velocity by causing a narrowing of the available water channel resulting in an acceleration of flow over the body. This error can be corrected by the equation given in Smit et al. (1971):

U , = Us(1+Ai/Aii) where U,, corrected velocity; U s ,velocity in the absence of a fish; Aii, cross-sectional area of the swimming chamber; Ai , cross-sectional area of the fish, which is assumed to approximate an ellipse and thus equal n/3d/+w, where d and w represent the maximum body height and width, respectively. Drag experienced by fish in an enclosed flume is higher than that expected at comparable freestream velocities. Extra drag arises from horizontal buoyancy and solid blocking effects (Pope and Harper, 1966).The former effect results from the growth of the boundary layer along the chamber walls, which tends to decrease the effective crosssectional area of the tunnel through which water can flow. A pressure gradient is set up along the chamber, which tends to draw the fish toward the exit and, hence, increase the drag. The solid blocking effect arises from the increase in water velocity around the fish which

128

F. W. H. BEAMISH

results from the presence of that fish in an enclosed chamber. These corrections have been described by Webb ( 1971a). The horizontal buoyancy effect can be accounted for by a correction to the freestream velocity of about 1%. The solid blocking effect can similarly be accounted for by a correction of 7.5-15%, depending on the size of the fish. Consideration should be given also to the pressure effects on the propellor jet which results in an apparent thrust higher than expected. The correction applied to the freestream velocity is about 1% and opposite in effect to the horizontal buoyancy correction (Webb, 1971a). Thus the horizontal buoyancy and propellor corrections tend to cancel out.

B. Experimental Procedure Swimming performance may be measured for single fish (Fry and Hart, 1948; Brett, 1964, 1965, 1967; Houde, 1969; Farmer and Beamish, 1969; Beamish, 1970, 1974; Oseid and Smith, 1972; Hocutt, 1973)or for groups of fish (Katz et al., 1959; Boyar, 1961; Green, 1964; Thomas et al., 1964; MacLeod and Smith, 1966; Dahlberg et al., 1968; Griffiths and Alderdice, 1972; Johnston and Goldspink, 1973; Otto and Rice, 1974). In most cases groups consist of not more than ten individuals, although Thomas et al. (1964) ran performance tests on groups of one-hundred fish. Fish are customarily deprived of food for a sufficient period prior to testing to assure the postabsorptative state. For some species and environmental conditions this is achieved within 24 hr (Davis et d., 1963),while for others a longer interval is required (Molnir and Tolg, 1962; Farmer et al., 1975). After transfer of fish from the holding tank to the swimming chamber some investigators have followed the practice of allowing fish a sufficient period to recover from the effect of handling (Black, l955,1957a,b,c, 1958a,b; Black et al., 1961).The usual period allowed is 12-16 hr, during which time a slow flow is generated which serves to provide orientation for the fish and to avoid oxygen depletion (Brett, 1964; Dahlberg et al., 1968; Rao, 1968). The longest posthandling period followed appears to be the 3 days applied b y Smit et al. (1971). Glova and McInerney (1977) suggest this extended period may not be necessary, at least for the measurement of critical swimming speed. They were unable to demonstrate significant differences in performance between young coho salmon allowed 1 and 12 hr recovery in the swimming chamber.

2.

SWIMMING CAPACITY

129

Often a period of recovery is not allowed, rather only the few minutes normally required for fish to orientate to a low velocity (Fry and Hart, 1948; Blaxter and Dickson, 1959; Bainbridge, 1960; Beamish, 1966b; MacLeod and Smith, 1966; MacLeod, 1967; Hocutt, 1973). When a current is initiated most fish react by swimming against the flow. However, in most studies the objective is not simply to induce fish to swim, but rather to confine their activity within a limited portion of the chamber. This is facilitated by providing one or more visual cues together with a mild electrical stimulus. Alternating black and white stripes on the outer surface of the upstream end of the swimming chamber are frequently used to provide visual cues for orientation of swimming fish (Griffiths and Alderdice, 1972). At the downstream end an electric barrier often in association with a beam of light is employed to discourage fish (Fry a n d Hart, 1948; Brett, 1964). A small voltage (6-20 V, ac or dc) may be applied across the electrodes either continuously or as required (Brett, 1964; Thomas et al., 1964; Griffiths and Alderdice, 1972). Larimore and Duever (1968) assisted young smallmouth bass, Micropterus dolomieu, in maintaining station by suspending a small open cylinder within the chamber. The avoidance displayed by bass for bright light encouraged their entry into the darkened cylinder. Where the chamber consists of an open channel fish are sometimes prodded into swimming by gently tapping the caudal fin with a rod (Fry and Hart, 1948). Particularly noticeable among some fish including the salmonids (Byrne et al., 1972; Kutty and 'Saunders, 1973), centrarchids (Beamish, 1970), and scorpaenids (Beamish, 1966b) is their ability to hold station against a current using their large pectoral fins as depressors. This is usually effective only at the lower velocities and can be avoided by initiating tests at higher swimming speeds or rapidly alternating velocity until fish begin to swim. Occasionally individual fish do not perform well in swimming chambers despite the presence of visual cues, electrical stimuli, and repeated efforts. Where this occurs it is expedient to discard that individual from further tests. With experience an investigator can often eliminate within a few minutes after their introduction those individuals that are unlikely to perform. Physical conditioning prior to measuring swimming performance can b e an important experimental procedure and was early recognized by Gray (1953, 1957). Fatigue times of unconditioned hatchery rainbow trout forced to swim in a flume were much less than those for stream-conditioned trout (Reimers, 1956). Unexercised young sockeye and coho salmon tended to fatigue earlier than physically conditioned individuals (Brett et al., 1958). Further, prolonged swimming speeds

130

F. W. H. BEAMISH

were higher among the conditioned salmon. Hammond and Hickman (1966) found that conditioning rainbow trout resulted in a marked increase in time required to fatigue fish subjected to strenuous exercise. The method employed by Brett (1964) of generating a current of water in oval or circular holding tanks against which the fish must swim offers a practical and efficient means of physically conditioning fish prior to experimentation. The prolonged swimming speed of young largemouth bass, Micropterus salmoides (5.7cm), subjected to a daily conditioning program increased from 18 to 30 cm sec-' (3.1-5.3 e sec-I) after four trials (MacLeod, 1967). Similarly, the maximum prolonged speed of largemouth bass (22.5 cm) estimated by elevating velocity at 30 min intervals by increments of 10 cm sec-' (30 min, 10 cm sec-*) increased with exercise (Beamish, 1970). Prolonged speed for bass exercised once every 3 days increased from 50 to 58 cm sec-I ( 2 . 2 2 . 6?. sec-I) after three trials beyond which no further improvement in performance was demonstrable. In contrast, Bainbridge (1962) failed to demonstrate significant differences in burst speeds between rainbow trout exercised in a low current for up to 12 months and unexercised fish, although the best performances by the former were better than those by the latter group. Concordant with an elevation in prolonged swimming performance with physical conditioning is a general increase in metabolic efficiency. Hochachka (1961) found exercised rainbow trout had higher levels of blood hemoglobin and relatively larger hearts than unexercised trout. Further, physically conditioned trout developed a greater oxygen debt and utilized more of their restricted glycogen reserves than untrained fish for comparable levels of exhaustion. Hochachka (1961) suggested stamina was limited not by the amount of energy reserves but by the excessive accumulation of lactic acid during exercise together with a limited capacity of the muscles to buffer this acid. High lactate levels in both muscle and plasma of conditioned rainbow trout have been demonstrated by Hammond and Hickman (1966).The high hemoglobin levels in conditioned fish would provide both for greater buffering and oxygen carrying capacity. That the primary effect of physical conditioning may be on enzyme systems involved in the mobilization of energy supplies for strenuous activity and in the rapid recovery from fatigue was suggested by Hammond and Hickman

(1966). Adaptation within the muscle fibers of fish exposed to physical conditioning was demonstrated for coal fish, Gadus virens, and Atlantic cod, by Greer Walker (1970, 1971) and Greer Walker and Pull

2.

SWIMMING CAPACITY

131

(1973). Both red and white muscle fibers hypertrophied with physical conditioning, although the white fibers which are customarily associated with anaerobic metabolism increased in diameter only at the higher sustained swimming speeds. Hypertrophy of the muscle fibers would, on contraction, yield an increase in power output per unit effort. Production of anabolic steroids, although not experimentally demonstrated for fish, might accompany hypertrophy of the white muscle fibers and those of the red that are recruited to function in a similar manner to white fibers (George, personal communication). Anabolic steroids in mammals are known to elevate muscular activity by enhancement of the anaerobic metabolism which is in accord with the earlier observations of Hochachka (1961) and Hammond and Hickman (1966) on rainbow trout. Procedural uniformity in the rate and magnitude of velocity increment is generally lacking from performance studies, particularly at the level of sustained, prolonged, and burst swimming. Velocity may be increased in a stepwise progression or gradually until either fatigue occurs or the prescribed swimming speed is realized. Where a stepwise progression is followed, the time required of fish to swim at.each velocity plateau may vary from a few minutes (Fry and Hart, 1948; MacLeod, 1967; Larimore and Duever, 1968; Oseid and Smith, 1972) to 1 h r or more (Brett, 1964; Houde, 1969; Rao, 1968; Griffiths and Alderdice, 1972; Kutty and Saunders, 1973). Not always is the time period fish must swim at a given speed, nor the magnitude of the velocity increase held constant (Fry and Hart, 1948; MacLeod, 1967; Oseid and Smith, 1972). The variation in swimming speed attributable to the method employed has received a minimum of investigation. To this end Jones (1971) was unable to detect differences in critical speed of rainbow trout where velocity increments of one-sixth to one-ninth of the maximum performance were applied. Moreover, demonstrable differences in performance were not apparent for stepwise intervals of 20-40 min between velocity increments. Beamish (unpublished data) compared critical speed of rainbow trout in relation to intervals of 30 and 60 min between velocity increments of 5 and 10 cm sec-' after allowing fish to recover overnight from the effects of handling (Table V). For an increment of 5 cm sec-' the critical speed was 36.1 ? 2.91 (95% confidence intervals) and 39.5 ? 5.90 cm sec-' (4.1 and 4.1 e sec-l) when the interval between velocity increases was 30 and 60 min, respectively (Table V). When the increment was 10 cm sec-', the critical speed was 42.4 ? 8.81 and 39.9 f 3.87 cm sec-' (4.1 and 3.9 cm sec-') for intervals of 30 and 60 min, respectively. Thus, there is little

Table Critical Swimming Velocity increSpecies

Hiodon alosoides Coregotius autumnalis Coregonus clupeafomiis Coregonus nusus Coregonus sardinella Oncorhynchus kisutch Oncorhynchus kisutch Oncorhynchus nerko Oncorhynchus nerko Oncorhynchus nerko Oncorhynchus nerka Oncorhynchus nerko Oncwhynchtrs nerka Oncorhynchus nerka Oncorhynchus nerka Oncorhynchus nerko Oncorhynchus nerko Oncorhynchus tierko Oncorhynchus nerku Oncorhynchus nerka Prosopiurn willtunisoni Solrno gairdneri Sulmo gairdneri Salrno goirdneri Solnro gairdneri Solnio gairdneri Solnio gairdneri Snlnro gairdneri Sulmo goirdiieri Snlmo guirdneri Salrno goirdneri Solrnu goirdneri Salnio gairdneri Solnio gairdnsri Salnio guirdrieri Soloelinus alpinus Steriodus leucichthys Thymallus articus Esox lucius Notropis otherinoides Notropis spilopterus Plotygobio gracilis cotostonius cotostonrus Catostonius cotnmersoni lctalurus punctatus Percopsis oiniscotnaycus Lota lotu Pollnchiirs oieru k p o m i s gibbosus .Micrnpterus stilmoides Microptenis solmoides M i c r o p t m o solmoides Perca Jlatiescens Perca JlaL;e,scetis Percu jlave,scens Perco f7aoescens Stizostedion oitreum vitreum

NumI>er

2 4 159 33 2 1340 290 5 5 5 5 5 5 5 10

5 go 9 1,5 18

73 15 14 6 6 10 in 10 10 in 10 10 11 22 94 192 4 15 28 169 20 25 3 56 91 6 15 10 10 20 2n 30 30 54

Length (cm)

Weight (9)

22.5 42.1

651 633 29.5 7.5-9.5 9.7 16.9 15.7 16.2 17.2 16.2 17.2 18.4 9.2-16.6 16.2 7.7-53.9 5.661.4 18.5 52-6.0 30.4 10.9 12.5 29.2 11.8 12.5 30.6 32.8 9.3 10.4 10.3 8.9 9.2 9.5 9.3 35.5 &4 1 7-34 12-62 6.5 7.5-8.4 17-30 4-53 17-37 14.0-15.4 7.2 12-62 14.8-17.1 12.7 5.2-6.4 10.2 10.0 9.5 9.5 9.5 9.5 8-38

2-1500 1-500

ments (cni

sec-')

10 10 10 10 10

10 10 10 10 10

u,.,,

60

ucr,,

49.3 40.5 43.0 51.3 44.1 52.1 62.2 8.9-36.7 32.8 3.4-1438.0 I. I- 19t32.0 52.2

13.2 23.4 26.4 18.1 21.0

7.5 11.7 10.2 6.4 7.6 7.9 8.1 11-700 1.5-800 7-1800

40-300 0.5-2200 50-550

7-1800 26.9-43.9 44.9

4-500

132

Time between increments (min)

10 10 10 10 10 10 10 10 10 10 10 10 5 10 9 9 6 9 9 10 10 5 10 10 5 5 10 10 10 10 10 10 10 6 10 10

in 6 10 10 5 6 6

60 60 60 60 60 60 60 60 60 60 60 60 60 60 10 20 20 60 2n 2n 10 10 60 30 60 30 60 30 60 10 10

10 10 10 20 10 10 10 20 10

LO

10

60 60 20 in

10 5

60 15

5 5

15

5 10

15 in

15

Critical velocity

cm

bee-'

60 80 34.1-72.1 2 1.7-46.8 60 7.5-55.1 38.657.9 50.0 61.5 65.8 67.7 52.5 33.5 21.9 40.5-54.2 59.0 51.S178.0 39.7-128.0 69.4 38.7-42.6 42.5 65.9 43.4 58.1 52.3 79.3 66.6 91.0 39.5 42.2 39.9 36.1 39.2 35.2 41.9 100.2 144-490 5272 19-47 59 23.2-67.2 42.9-62.7 2S91 48-7.3 31.7-61.4 55 3 6 41 61.0-68.9 37.2 30.650.0 45.7 35.1 1.5.%210 25.233.0 33.5 15.5 38-84

e 2.7 1.9 1.4-5.7 1.4-3.6 2.0 1.0-5.8 4.0-6.0 2.9 3.9 4.1 3.9 3.2 2.0 1.2 3.b4.4 3.7 3.3-6.7 2.1-7.1 3.8 7.1-7.3 1.4 6.1 3.5 2.0 4.4 6.3 2.2 2.8 4.2 4.1 3.9 4.1 4.3 3.7 4.4 2.8 12.0-18.0 1.9-7.5 o.ai.6 9.1 3.0-8.6 2.1-2.5 1.7-5.8 2.0-2.8 2.1-4.2 7.6 0.7-3.0 3.64.2 3.0 5.28.1 4.5

3.5 1.62.2 2.7-3.5 3.5 1.6 2.2-4.7

Speeds of Fish Teniperature ("C) Acclirn;,talll

12 I2 7-19 7-19 I2 2-26 5-19 15

IS 15 15

15 15 1s 5 I0

I5 20 21

Experi. mental 12 12 7-19 7-19 12

2-26 5- I!) 5 10

15 "2..( 22.6

22.4 22.6

15

IS 1.5 IJ 15 IS 15

1s 1.5 15

IS 15

IS 15

12

12

I2

'30

15-'1S

12-19 7-19 12-19

12-19 7-19 12-lo lS0,5

.30

12 7-12

12

I0 20 :30

10

25 25 I0 21 1 10 21

I9

Sodium pentachloriipheiiate pre\ent ((L50 pph)

15

12 12-19 12-19 I2

12 12- I!) 19-19

Acute temperature exposure Acute temperature expo5 lire Acute ternperilhire erposure Acute temperature exposure

PO

14. I

I2 11.9

Acclimation and acute tumperatue expowre Acclimation to teniperature and d i n i t y (0-20%) Acute temperntiire exposure .Acute teniperature exposure

15

20 24.5 26 27.5 5 I0 15 24 15 7-12 11.9 14.1

IS

Comments

Wild atock. 1 h r recovery priiw to test Uatchery 5tock. I 1ir recovery prior to tebt Overnight recovery prior to t a t Overnight recoveiy prior ti, tert Overnight recovery p i o r to t r a t Overnight recovery prior to te\t Overnight recovery prior hi t r h t Overnight recovery prior to te*t Overnight recovery prior to teht

7-12

20 15-'3.j

2s 25 10 20 20

I0 I $1

133

Reference Jones et ul. (1974) Jones ei ol. (1974) Jones e t 01. (1974) Junes et o l . (1974) Jones e f uI. (1974) G r i s t h s and Alderdice (1972) G l o w ;idMcIoerney (1977) Brett (1967) Brett i 1967) Brett (1967) Brett (1967) Brett (1967) Brett (1967) Brett (1967) Brett and Class (1973) Brett and Glas* (1973) Brett and Gl;i\\ (1970) Brett and Glass (1973) Brett end Glass (1973) Wel)l) iirtd Brett (3973) Jane\ et n l . (1974) Joner (1971) Joiie.;(1971) W e l h (19711)) Jones (1971) Jones i 1971) June\ ei (11. ( 1974) Jme\et ul. (1974) Beainish (tinpuldished) Beamish (~inpul)lished) Bemiish (unpul~lished) B e m i s h (unpiildirhed) Beamish (nnpnl~lished) Beairiish (unpulrlislied) Beiinii\h (unpuliliahcd) ]one\ et a / . (1974) Junes ei id (1974) Joner ci 01. i 1974) June\et d. (1974) Jme\ r i id. (1974) Hocutt (1973) lone, ei nl. (1974) Joiie.;et (11. (1974) Jane, et d . (1974) Hocutt (1973) Jane\ et ( t / . (1974) Jmies ei nl. (1974) (;rerr \Valker a n d Pull (1973) Brett :ind Sutherl.incl ( 1965) Hociitt (1973) Farlinger .ind Beiuni\h ( 1977) Farlinger and Beamish ( 1977) Otto ;MIRice i 1974) Othi and Rice (1974) 0 t h ) and Rice i 1974) Otto imd Rice i 1974) Jiiiie* el (11. (1974)

134

F. W. H. BEAMISH

indication that the critical speed of rainbow trout is influenced by velocity increments between one-fourth and one-ninth of the critical speed or the time interval, within the range of at least 20-60 min. Recently, Farlinger and Beamish (1977) examined the influence of the magnitude of and interval between velocity increments on the critical swimming speed of largemouth bass (Fig. 5). Critical performance decreased curvilinearly with increasing intervals of time for a given velocity increment. When time was fixed, critical speed increased with velocity between 2.5 and 10 cm sec-' and with further increases changed little or declined. The highest critical speeds were achieved when velocity increment was 10 cm sec-I and the interval, low. Ideally the relationship between interval and increment on swimming performance should be established before an experiment is begun. If this is impractical, a velocity increment of 10 cm sec-l appears a satisfactory choice. In the absence of quantitative information, the selection of the time interval may be determined by the objectives of the study. Ultimately, if the velocity increments are continued, a speed is reached against which fish are unable to swim for the prescribed time period. When fatigue occurs in tunnel chambers, fish are forced

b

Fig. 5. Critical swimming speed of largemouth bass, Micropterus salmoides (10 cm), in relation to the interval between and magnitude of velocity increments. (Redrawn from Farlinger and Beamish, 1977.)

2.

135

SWIMMING CAPACITY

through the electric field when present, against the downstream retaining screen (Boyar, 1961; Brett, 1964) or into a recovery section of reduced water velocity (Thomas et al., 1964; Griffiths and Alderdice, 1972).Where retaining screens are present the criteria for fatigue vary but fatigue is usually accepted when a fish by repeated efforts and despite the application of electrical stimulus or prodding can no longer hold itselfoff the screen. One variation of this procedure is that followed by Smit et aZ. ( 1971).When goldfish fell against the downstream screen, the velocity was reduced. If fish continued to swim even after the “fatigue” velocity was resumed, the first failure was ignored and the experiment was continued. However, when the second failure was recorded the experiment was terminated. Where oval or annular chambers are employed and a retaining screen is not used, fatigue is usually presumed when fish begin to loose laps. Larimore and Duever (1968) presumed young smallmouth bass to be fatigued when they left a darkened cylinder suspended in the chamber. Critical swimming speed is measured by interpolation for those fish that do not fatigue exactly at the beginning or end of a prescribed swimming period. The formula described by Brett (1964) is as follows.

Ucrit (critical swimming speed)

= ui

+ (tihiitx uii)

where ui, highest velocity maintained for the prescribed period (cm sec-I); uii,velocity increment (cm sec-I); ti, time (min) fish swam at the “fatigue” velocity; ti,, prescribed period of swimming (min). Thus, a fish successfully swimming for a prescribed period of 60 min at 50 cm sec-’ but fatiguing at 60 cm sec-I after 22 min would have a critical speed of

50 +

x 10) = 53.7 cm sec-’

Critical swimming speed is usually represented by the median performance of the fish used. This is determined graphically by plotting the logarithm of critical speed against the cumulative percentage of fish fatigued on a probit scale. As a rule this transformation allows the application of a linear regression. In some cases, however, the distribution is best described by more than a single linear regression implying mixed fatigue effects or compound responses to velocity (Griffiths and Alderdice, 1972).A summary of critical swimming speeds is presented in Table V. Proper differentiation of sustained and prolonged swimming speed requires a description of the relationship between time to fatigue and velocity such as that presented by Brett (1964) for sockeye salmon (Fig. 6). In this figure there appears an obvious distinction between

136

F. W. H. BEAMISH

I!

1000 5001

I

1

Sustained (Transition)

Salmon

0

C

Q)

E

i=

Burst

0.05 0.01

0

2

4

Velocity,

6

6

10

1 sec-l

Fig. 6. Identification of sustained, prolonged, and burst speeds for rainbow trout, Salmo gairdneri, and sockey salmon Oncorhynchus nerka, on the basis of their fatigue time at different swimming velocities. Results for rainbow trout obtained from Bainbridge (1960, 1962), and for sockeye salmon from Brett (1964). (Redrawn from Brett, 1964,J. Fish. Res. Board Can.)

prolonged and sustained swimming. The logarithm of time to fatigue at prolonged speeds increased linearly as velocity was decreased. With a continued decrease in swimming speed, sockeye salmon did not fatigue. The separation between sustained and prolonged swimming is surprisingly sharp and represents only a few cm sec-'. Where it is not practical to determine the response between velocity and time to fatigue it is assumed those speeds which can be maintained for a minimum of 200 min represent sustained. Just exactly how long fish will swim at sustained velocities has not been well investigated. Johnston and Goldspink (1973)were able to force coalfish to swim in a flume for 16 days at low velocities, Largemouth bass swam continuously at slightly in excess of 1C sec-l for periods of 2 weeks (Beamish, 1975). Swimming endurance studies on redfish suggest some individuals were able to swim steadily at over 2 e sec-I for 10 days or more (Beamish, 1966b). Smit et al. (1971) forced goldfish to swim at low speeds for periods of 1 week as a general routine. Distinct alteration in the coefficient of the linear relation between velocity and time to fatigue for burst and prolonged swimming speed implies physiological differences in the availability and mobilization of the fuel for muscular activity. Again, where it is not possible to determine the point of inflection between burst and prolonged speed,

2.

137

SWIMMING CAPACITY

a period of 20 sec is presumed a reasonable approximation. The procedure generally applied in the measurement of burst speed in the laboratory is to prod, by mechanical or electrical stimuli, fish swimming steadily at a moderate velocity. This causes the fish to dart forward with an initial expression of accelerated or unsteady swimming followed by steady or sprint swimming. Photographic or electronic devices have been applied to improve precision (Bainbridge, 1960; Komarov, 1971).

C. Biological Constraints on Performance 1. SIZE a. Length. Of the constraints on performance capacity, size is among the most important. As early as 1917, Thompson argued that sustained or prolonged speeds should be proportional to the length of fish raised to the power of 0.5. This conclusion was based not on measurements of swimming speed, but on the assumption that fish volume and the proportionate amount of muscle increases as the square of length. Assuming further that power is limited not by the volume of muscle but rather the surface area of the gills, Thompson (1917) suggested that maximum or burst speed was independent of length. This conclusion was reached also by Hill (1950) on the basis of heart capacity and blood flow through the vessels whose cross-sectional area increases as the square of length. Most frequently the relationship between length of fish and performance is described by the equation

log u

=u

+ b(log e)

where u is swimming speed (cm sec-l) and 8 , length (cm) (Blaxter and Dickson, 1959; Bainbridge, 1960; Brett, 1965; Brett and Glass, 1973). The relation has been described also by a linear regression without logarithmic transformation (Glova and McInemey, 1977) or after logarithmic transformation of swimming speed but not length (Beamish, 1970). Concordant with the earlier views of Thompson (1917), Brett ( 1965) found the (60 min, 10 cm sec-l) critical speed of sockeye salmon (8-55 cm) was proportional to a fractional power of length equal to (Fig. 7A). At burst swimming speeds, the regression coefficient appears to approach unity. Blaxter and Dickson (1959) found the (2-5 sec) burst speed of Atlantic herring to increase linearly with length ( 1-26 cm) after logarithmic transformation, the coefficient of which

138

F. W. H . BEAMISH 140 -

100

-

-

50

10

5

50

l

0 0)

-

u)

E

0

300

-

100

-

B

U

tn

.-

E I

I

3 tn

0

I

1

20

I

I

60

40

4-

2 -

hrca

,z/””’ostedion

vitreum

U’ I

l

0.6

’

l

0.8

l

l

1.o

l

l

l

1.2

l

14

l

l

i

1.6

L e n g t h , cm Fig. 7. Swimming speed and length. (A) Critical swimming speed of sockeye salmon (Brett, 1964). ( B and C) Burst (sprint) swimming speeds (Blaxter and Dickson, 1959; Houde, 1969).

2.

SWIMMING CAPACITY

139

was unity (Fig. 7B). Similarly, Bainbridge (1960) found a coefficient of unity for the (1-20 sec) burst swimming speed of the dace, Leuciscus leuciscus, (10.0-21.4 cm). Houde (1969) found a coefficient of approximately 1 for (0.5-5.0 sec) burst speed for larval yellow perch, Perca jauescens, (0.9-1.4 cm), and walleye, Stizostedion vitreurn, (1.0-1.6 cm) following absorption of the yolk sac (Fig. 7C). Pavlov et al. (1968) reported a coefficient of unity for several species of minnows and yellow perch. Methods have been described to correct swimming speed for variation in length where it is not practical to determine the precise relationship. Relative performance as e sec-I sometimes allows for comparison of fish of different length. Bainbridge (1960) found the relative burst speed (4 sec-l) for dace, rainbow trout, and goldfish, Carassius auratus, was independent of length and equal to about 10 sec-'. Accordingly, relative burst speeds of larval plaice changed little with length (0.7-1.4 cm) kom 10 e sec-' (Ryland, 1963). Relative burst speeds of longer duration generally favor the smaller individuals. Bainbridge (1960) found that when burst swimming speed was extended from 1 to 20 sec, relative performance by the smaller individuals displayed an improvement over that of larger fish. Thus over the size range of 7-20 cm, relative performance of goldfish decreased from about 6.3 to 4.0 sec-' and that for dace (10-22 cm) declined from 4.8 to 4.0 e sec-'. Burst speeds (2-5 sec) determined for a number of teleosts by Blaxter and Dickson (1963) followed a similar pattern of decline in relative performance with size. Burst speed of brown trout, Salrno trutta (10-40 cm) decreased from approximately 17.5 to 1.5 C sec-I while that for Atlantic herring (1-20 cm) was reduced from 10.1 to 5.6 sec-'. Among the many species examined only coalfish displayed an increase in relative speed with size, 5 . 2 5 . 7 e sec-' for individuals of 13-20 cm, which may be at least partly attributable to their narrow length range. Relative performance at prolonged and critical swimming speeds generally favors the smaller individuals of a species. Thus, Brett and Glass (1973) found the (60 min, 10 cm sec-l) critical speed of sockeye salmon decreased form 4.5 to 2.0 e sec-I as length increased from 10 to 90 cm. Moreover, Brett and Glass (1973) demonstrated a similar pattern over the entire ecological range of temperatures for the species. Beamish (1970) found that relative prolonged performance of largemouth bass favored the smaller individuals at temperatures approximating their physiological optimum (Niimi and Beamish, 1974) but that at lower temperatures there was little evidence of this difference. More recently Glova and McInerney (1977) corroborated the

140

F. W. H. BEAMISH

dependency of the relation between length and performance on temperature for coho salmon but found it to be independent of salinity. Smit et al. (1971) converted prolonged speeds of goldfish to relative velocity by the equation u, =

-

tSU2

e

whereu,, relative velocity for the standard fish (cm sec-I); tS, length of standard fish (cm);u , measured velocity of fish (cm sec-I); t,measured length of fish (cm). The distance or length of time (endurance or stamina) fish are able to swim against a particular current is also dependent on length. Boyar (1961) found that as Atlantic herring increased in length their endurance increased, and that this relationship was best described by a linear regression after logarithmic transformation. Over the range of prolonged speeds applied Boyar found endurance to be a function of approximately the fourth power of length (Fig. 8). The distance sea lamprey, Petromyzon marinus, were able to swim at a fixed velocity and temperature increased approximately as the square root of their weight (Beamish, 1974; see also Fig. 9). The percentage of muscle in at least some species of fish tends to increase with length (Bainbridge, 1960). However, hydrodynamic drag increases also with length. Sustained and prolonged swimming in contrast to burst is limited by the rate at which muscles can be supplied with the raw materials for contraction and relieved of waste products (Bainbridge, 1958; Jones, 1971). To this end large fish are able to provide for a higher relative metabolic scope for activity (Fry, 1947; Basu, 1959), although for some species this relationship is temperature dependent (Brett and Glass, 1973).After consideration of the pertinent factors, Brett ( 1965) attributed reduced relative performance by large sockeye salmon to an increase in hydrodynamic drag which he suggested outweighed the advantage of increased metabolic scope and body musculature. Burst speed is reliant on a store of raw materials such as glycogen within the muscle cells or possibly oxygen bound in muscle hemoglobin. The relative store of glycogen appears to be independent of size for at least Atlantic cod (Beamish, 1968). Hence in burst swimming it is likely the influence of increased drag in large fish is countered by proportional elevations in muscle development and metabolic fuel.

b. Weight. Swimming performance is most often expressed on the basis of length but has been described also for weight. Fry and Cox (1970) found the (1min, 11-18 cm sec-l) prolonged speed of rainbow trout to increase with weight (4-100 g) raised to the power of 0.13.

2.

141

SWIMMING CAPACITY

/

C lupea harengus

6

15

10

20

Length, cm Fig. 8. The endurance of Atlantic herring, Clupea harengus harengus, of different lengths at three swimming speeds. (Redrawn from Boyar, 1961, Swimming speed of immature Atlantic herring with reference to the Passamaquoddy Tidal Project, Trans. Am. Fish. SOC. 90, 21-26.)

A method of swimming performance rating based on the ratio of useful work done to the muscle power available for fish of different weight but of the same species was developed by Thomas et al. (1964) and is described by the equation: Performance rating

=

C,

JY5 1.3 x 10 ui v 2 t i

where ui, the relative velocity between the fish and water; ti, time interval; u, water viscosity; M , weight of fish; C 3 is a constant.

142

F. W. H . BEAMISH

Fig. 9. Distance sea lampreys, Petromyzon marinus, swam in relation to speed, size, and temperature. (From Beamish, 1974.)

c . Condition. Swimming capacity is influenced also by the weight of fish relative to their length, most often described by a condition factor such as that computed by Fulton (1911, reported in Ricker, 1975): ~ 1 e 3

where M is weight and e, length. Bams (1967)expressed the condition of unfed sockeye migrant fry by the factor 10~1'3

c

2.

143

SWIMMING CAPACITY

Ryland (1963) equated size of larval plaice, Pleuronectes platessn, to their length times the height of their musculature midway along the animal's length in describing the relationship with burst swimming speed. The importance of condition to swimming capacity has been explored most thoroughly for the salmonids. Generally domestic stocks of trout not only grow faster than wild but are heavier for a given length. Vincent (1960) noted the chemical composition of wild and domestic stocks of brook trout, Salvelinus fontinalis, were similar except in the fat content which was higher in domestic stocks even when both groups were reared in the hatchery from the egg stage under experimentally similar conditions. Wild stock brook trout consistently out-performed domestic in stamina tests. In a comparison of stamina among three stocks of brook trout, Green (1964) first reared eggs and the young stages under similar hatchery conditions. Stamina tests were conducted at two velocities, 45 and 57 cm sec-I. The number of fish able to swim for 2 min against the prescribed velocity increased consistently with length in all three stocks (Fig. 10).However, for fish of a given length, a greater proportion of fish from the wild stocks (Long Pond Outlet and Honnedaga Lake, New York State) were able to sustain the respective velocities for 2 min than was found for 100 r 57cm sec-'

s.

.-c E

80

-

@4 L

al m

60-

c ZI) 4 0 -

.-E E .-E

3

fn

20-

.6

8

10

12

Length, cm Fig. 10. Comparison of swimming perfonnance among three stocks of brook trout, Snloelinus fontinalis. Performance is expressed as a percentage of the trout swimming after 2 min at 58 cm sec-'. (From Green, 1964.)

144

F. W. H. BEAMISH

domestic stocks. Both Vincent (1960) and Green (1964) attributed the poorer performance by domestic stocks of brook trout to their higher fat content. Additionally, hydrodynamic drag would be expected to increase with condition factor. On the assumption that gill area and the efficiency of the pumping mechanism.and gaseous exchange are similar for fish of a given species and length, the metabolic scope for activity would decline with an increase in weight. Relative stamina of unfed migrant sockeye salmon fry was examined by Bams (1967) based on ranking order of fatigue. Since salmon were not fed, their condition factor decreased with progressive absorption of the yolk. Once the yolk was absorbed stored energy reserves were catabolized so that the fry became increasingly emaciated. Bams (1967) found relative swimming performance was optimal at the stage of almost total yolk absorption. With a well-developed yolk and hence a high condition factor, relative performance declined, which Bams (1967) attributed to the high drag associated with the protruding yolk sac. Performance was poor also, when the condition factor was low, which reflected the depletion of energy reserves among unfed fry after absorption of the yolk.

2. SEX Little information is available on the influence of either sex or stage of maturity on swimming capacity. Brett (1965) compared the (60 min, 10 cm sec-l) critical speed of adult male and female sockeye salmon for which relative performance favored the slightly shorter males. The critical speed for males (41.8 cm) was 125 cm sec-I (3.0C sec-l) while for females (53.9 cm) it was 143 cm sec-1 (2.7 8 sec-I). Males were tested at their temperature optimum of 15°C (Brett, 1964), whereas females were examined at 17"C,suggesting that at comparable temperatures differences in performance would be minimal.

3. DISEASE Restriction in capacity for swimming imposed by parasitic infection has received some consideration. However, there appears to be no information on the influence of bacterial or viral infection on swimming performance. Relatively heavy infections in rainbow trout (11.0 cm) with metacercariae of the trematode, Bolbpophorus confusus, which locates in the body muscles, reduced the prolonged speed of 75 cm sec-' b y about 35%when compared with control fish (Fox, 1965; Butler and Milleman, 1971). Olson (1968, cited in Butler and Milleman, 1971)

2.

SWIMMING CAPACITY

145

was unable to demonstrate differences in the swimming capacity of rainbow trout infected with metacercariae of the trematode, Cotylurus erraticus, which lodges in the pericardial cavity of its host. Swimming endurance of rainbow trout infected with metacercariae of the intestinal fluke, Crepidostomum farionis, and the cestode, Proteocephalus sp., was not significantly different from fish containing only Proteocephalus sp. (Klein et al., 1969). However, while differences were not significantly different, the length of time C. farionis-infected fish swam, 32.5 min, was appreciably less than the 47.7 min recorded for control fish. Unfortunately the velocity at which tests were conducted was not reported nor was the diet fed control fish the same as that offered to C. farionis-infected trout. Smith and Margolis (1970) measured the (30 min, 6 cm sec-l) prolonged speed of sockeye salmon (2-4 g) free from and infected with the cestode, Eubothrium saluelini. Infected salmon, which were about 1 g lighter in weight, fatigued after swimming about 66% of the distance covered by control fish. It was estimated by Smith and Margolis (1970) that a reduction in swimming capacity of this magnitude was sufficient to reduce the success with which infected salmon are able to descend long rivers in their seaward migration, realizing that fish must procure food and escape from predation along the route. Performance of rainbow trout and coho salmon infected with cercariae of the parasitic trematode, Nanoph ytes salmincola, was measured and compared with control fish by Butler and Milleman (1971) (Table VI). Two methods were employed to evaluate swimming performance. In one case, velocity was gradually increased to the prescribed velocity of 38.1 and 32.3 cm sec-I for rainbow trout and coho salmon, respectively, at which endurance time was recorded. The second method employed a s tepwise progression of velocity increments to delineate the maximum prolonged speed. Velocity was increased at 3 cm sec-' every 20 min to 22.9 cm sec-' and thereafter, 3 cm sec-' at 10 min intervals until fish were fatigued. Immediately after exposure to 1500 cercariae, the maximum prolonged speed of infected trout was 32% below that found for control fish. With time this differential decreased so that after 15 days, infected and control fish displayed similar swimming capacity. Endurance was more markedly impaired among infected fish, being reduced by 3 6 5 4 %over the first 96 hr after infection when the parasites were migrating through the tissues or had not yet completed development as metacercariae. After 15 days when the parasites had encysted, the differential in endurance time was only 3%. The pattern among coho salmon was similar to that for trout; however, the absolute reduction in endurance was the more marked

Table Prolonged Swimming ~~~

Species Petromyzon mu. .nus Petroniyzon marinus Petrotnyzon marinus Alosufrnito Clupeu harengus CIupea hureiigus Clupeo hurengus Oncorhynchus kisutch Otrorhynchus kisutch Oncorhyrlchus kisutch Oncnrhynchus kisutch Oirorhynchtrs kisutch Oncorliynchus kisutch Otrorhynchus kisutch Oncorhynchus kisutch Otlccwhynchrrs kisutch Oncorhynchus kisutch Otrorhynchus kisutch Oricnrhynchus nerku Oncorhynchus nerku Oncorhynchus nerko Oncorhynchus nerko Oncorhynchus nerka Oncorliynchus nerku Oncorhyrichus nerku Oncorhynchus nerku Oncorhynchus nerku Oncorhynchri,s nerko Oncorhynchus nerku Oncorhynchus tshuiuytschu Oilcorhynchus tshowytscliii Oncorhynchus tshuwytscho Oncorhynchns tshawytscha Solmo goirdneri Solnicr goirdneri Srilnio guirdneri Solino goirdneri Srilnio guirdneri Sulnao irideus Snlino solar Sulmo sirlor Snlnio tnrttu Suluelinus fontinalis Soloelinus fimtinolis Sultielinus fimtinulis Sultielinus fontinulis Suloelinus /ontitidis Suluelinirs fimtinalis Soloelinus fnntinulis 'I

Number

53 53 53 30.0

22 16 10

11

10 10 5

20 26 23 [3151 11501 [ 1501 3 5 15 10

5 5

7 17 29 19 14 160 9 11 16 129 10

I28 [ 1501 [ 1501

> 100 5

6 6 6 18 6 28 6

Tohl length (cm)

Weight (g)

5- 100

14.5-39.0 14539.0 14.5-39.0 29.7 15.226.0 6.8-15.7 12.9-20.5 4.68.8 4.9-9.1 5.7-8.9 5.69.0 6.3 7.9-9.0 6.7-8.9 7.68.9 7.69.3 5.7-6.0 5.7-6.0

5-100

5- IW

0.7-7.8 1.48.1 1.7-8.1 1.9-8.3 2.5

3.5-6.9 1.41.7 1.P1.7 1.7 2.9 1.9-35.9

6.2 6.6 6.2-15.2 7.1-13.9 7.4 7.4 14.5 14.3 13.4 12.6 13.8 .3.8 8.1- 12.6 5.1-7.3 5.7-7.6

2.8-26.9 3.4 3.4

2 . S 196 20.2 7.5-24 5.7-6.0 5.7-6.0 20.0 15-20 23.4 34.0 9.9 9.6 10.2 10.3 10.4 10.1 10.9

5-100

1.4-1.7 1.41.7

110.3

Velocity increments (cm sec-')

Time between increments (min)

Gradually to prescribed velocity Gradually to prescribed velocity Gradually to prescribed velocity Gradually to prescribed velocity Gradually to prescribed velocity Gradually to prescribed velocity Gradually until laps lost Gradually until laps lost Gradually until laps lost Gradually uiitil laps lost Gradually until Lips lost Gradually to prescribed velocity Gradually to prescribed velocity Graduslly to prescrihed velocity 2.3 10 2.3-3.0 1c-20 2.3-3.0 Icr20 Gradually until laps lost Gradually until laps lost Gradually until laps lost Gradually until laps lost Gradually until laps lost Gradually until laps lost Gradually to prescribed velocity Gradually to prescribed velocity Gradually to prescribed velocity Gradually to prescribed velocity Gradually to prescribed velocity Gradually to prescribed velocity Crndudly to prescribed velocity Gradually to prescribed velocity Gradually to prescribed velocity 12-27 60 Gradually to prescribed velocity 11- I8 I 2.,>3.0 10-20 2.,%3.0 10-20 Gradually to prescribed velocity Gradually to prescribed velocity

~~

Swimming time at iiiarimum velocity (min)

10 10 10 1 1-120 1-17

5 5 5 5

5

10 10 10

5 5 5 5 5

5

>300 2300 66 21 6 10

60 1

10 10

20

34.1

3 3 3 3 3 '3 3

Tunnel swimming chamber.

', Oval or iiiinular swimming chamber. *Towed .;winlining cage.

146

30 30 30 30 60 30 15

30 30 30 30 60 30 15

VI Speeds of Fish Temperature ("C)

Maximum velocity cm sec-'

t sec-I

16.633.6 16.F34.7 24.2-41.3 75 91-143 36.6 97.5 14.9-26.5 22.0-35.4 25..>37.5 30.5-41.2 28.4 :10-49 49-55 2.3-55 28-64 22.9-34.3 19.8-29.7 14.3 22.9 26.8-43.3 02.C46.4 27.1 21.7 46.4 51.1 53.6 55.4 66.2 40 23-67 29-53 2,3-53 19-73 30-70 48-70 25.1-43.3 13.7-34.3 170 70-100 50-76

0.9-1.2

82

34.7 37.4 47.6 49.5 48.2 48.9 55.9

0.!L1.2 1.1-1.7 2.5 5.4-6.0 2.3-5.3 4.8-7.6 3.0-3.2 3.9-4.5 4.2-4.4 4.65.4 4.5

3.S7.8

Acclimation

Experimental

5 10 15

5 10 15

3.6

4.0 4.4 4.8 10.5

Endurance Endurance

5-6

5-6

10

10 15 20 24 10 15 20 20

15 20 24 10 15 20 20

1 6

1 6

10 15 20 24 15 15 15 15 15

10 15 20 24 15 15 15 15 15 [221

11.5 15 1Y.5

11.5 15 19.5 ,!&I5 15

s 1.5 1.S3.5 55-6.4

15

3.90 4.67 4.82 4.6 4.8 5.I

1-14 15

1-14 15

15 15

15 15 15 15 15 15 15

15 15 15 15

15

Endurance. 20%; O,, .%I9 mglliter 3-19 mglliter Endurance, 20%; 0,. Endurance, 20%,0%. .>I9 muliter O,, 2-26 muliter; CO,, 1-120 mglliter Control Infected; Notiophyetus siilniincolo

Endurance, 50% fatigue Endurance, 50% fatigue Endurance, 50% fatigue Endurance, 50% fatigue Endurance, 50% htigue Endurance, 50% fatigue Endurance, 20% Fatigue; O,, 2-10 mgiliter Endurance. 20% fatigue; 4, 2-10 mglliter Endurance, 20% fatigue; O,, 2-10 mglliter Salinity, 0-30% 02,2.0-2.3 inyiliter

10

10

18-19 18-19

8.5 3-4 2.1-3.2 2.7 3.50

Endurance Endurance Endurance

12

1.4-5.6 1.45.6

18-19 18-19 2.3 3.5 4.3-2.8 3.3-4.6 3.7 2.9 3.2

Comments

Control Infected; Nanophyetus sulrniricola Photography Salinity, @GO% 0,. 3.8-5.0mgiliter Exposure to 1.5mglliter fenitrothion Exposure to 0.5 mglliter fenitrothion Exposure to 0.15 mglliter fenitrothion 0 fenitrothion Metabolism Metabolism Metabolism

Reference Beamish (1974)" Beamish ( 1974)" Beamish (1974)" Magnan (1929) Brawn (1960)' Boyer (196Ub Boyer ( 1961)b Brett et al. ( 1958)' Brett et al. (1958)' Brett et a/. ( 1958)' Brett et al. (1958Ib Brett et al. ( 1958)b Davis et ul. (1963)" Davis et al. (1963)" Davis et al. ( 1963)" Dahlberg et al. (19SaP Butler and Milleman (1971)" Butler and Milleman (1971)" Brett et al. ( 1958)' Brett et al. ( 1958)' Brett et al. (1958)' Brett et al. ( 1958Ib Brett et al. (1958)' Brett et al. (1958)' Brett et al. (1958)b Brett et a/. ( 1958)b Brett et ul. (1958)b Brett et al. (1958)' Brett et 01. (1958)' Kerr (1953)" Davis et a / . (1963)" Davis et al. (1963)" Davis et al. (1963)" Rao ( 1968)" Kutty ( 1968)" Fry and Cox ( 1970)' Butler and Millemail (1971)" Butler and Milleman (1971)" Gray (1953) Byrne et al. (1972)" Kutty and Saunders (1973)" Magnan (1929) Peterson (1974) Peterson (1974) Peterson (1974) Peterson ( 1974) Peterson ( 1974)" Peterson (1974)" Peterson (1974)"

(Continued)

147

Table VI-

Species

Saloelinus fontinalis Saloe!inus fontinults Saloelinus namaycush Saloelinus namaycush Esox lucius Esox lucius Carussius auratus Carassius auratus Carassius auratus Carassius auratus Carassius auratus Carussius auratus Curassius uuratus Carassius aurutus Carassius auratus Cyprinus carpi0 Leuciscus leuciscus Pimephales promelas Scardinius erythrophthalmus Scardinius erythrophthalmus Gadus luscus Gadus merlungus Gadus morhua Gadus morhua Merluccius vulgaris Macrozoarces americanus Morone saxotilis Morone saxotilis Lepomis macrochirus

Number

6 6

Weight (B)

Velocity increments (cm sec-I)

11.0

3

11.6

3 4-12 PI2

27.7 82.8 16.5 37.8 4.4 4.4 4.4 4.4 4.4 4.4 4.4

3 3 3

3 3 3 10 7

465

34 40 12 1090 340

18.2 15-17 13.5 18.2 4.8 19.0 18.2 16.7 17.7 35.5 35-37 22.6 33.6-38.4 2-14 8.9-11.4 4.5-5.7 5.1-5.4

30 105 65 45 53 51 51

45 30

30 32 45 45 48 51 38 137 181

1.5 0.5 2 2

Swimming time at maximum velocity (min) 15 15 2 2

17

3

Lepomis macrochirus Micropterus salmoides Micropterus salmoides Micropterus salmoides Micropterus salmoides Micropterus salmoides Micropterus salmoides Micropterus salmoides Micropterus salmoides Micropterus sulmoides Micrapterus salmoides Micropterus saltnoides Micropterus salmoides Micropterus salmoides Micropterus salmoides Micropterus salmoides Micropterus salmoides Micropterus sulmoides Perca Jlaoescens Perca perca Stizostedion vitreum oitreurn Sdaena aguilu Mugil capita

Total length (cm)

Time between increments (min)

50-60

18.8 16.5 17.7 560 580-635 23.7 156237

1.9-3.7 2.9-3.4

21.3 5.7 8.0-8.5

4.a6.4 5.67.4

8.0-8.6 2.0-2.2 2.0-2.2 2.0-2.2 2.0-2.2 2.0-2.2 2.0-2.2 5.7 15-27 15-27 15-27 15-27 15-27 15-27 0.61.4 18.3 0.7-1.5 29.7 26.5

45-270 45-270 45-270 4S270 45-270 45-270

4-12 2 4-12 2 4-12 2 4-12 2 4-12 2 4-12 2 2 4-12 Gradually to prescribed velocity 110 40

Gradually until laps lost 13

Gradually to prescribed velocity Gradually to prescribed velocity 19 Gradually to prescribed velocity Gradually to prescribed velocity Gradually to prescribed velocity 2-3 2-5 2-3

2-5

2 2 2 2 2 2 2 40-300

3

>5.6240 >4.2-240 1.1-6.6 10 30 31-201 22-28

Gradually until laps lost 2.3 10 2.3 10 2.2 3 2.2 3 2.2 3 2.2 3 2.2 3 2.2 3 Gradually until laps lost 10 30 10 30 10 30 10 30 10 30 10 30 Gradually to prescribed velocity

3 3 30 30 30 30 30 30 60

Gradually to prescribed velocity

60

3 10

10 3 3 3 3 3

18.4 29.5 26.0

148

Continued Temperature ("C)

Maximum velocity cm sec-'

67.9 88.7 35-54 4a83 210 148 22.4-34.7 29. I 22.4-40.3 51 28-51 50 15.3-38.8 15-85 60-126 170 170 19.6 114 130 55 23 75-135 75-135 79 90-120 35-87 67 22.5

&'

sec-'

6.2 7.7

Acclimation

Experimental

15 15 a23 10-22

15 15 8-23 10-22

12.7 3.9

1-3.2 3.8-8.4 12.6 9.2 4.1 6.0 5.9 3.3

5 10 15 20 25 30 35 20 15-30

5-25 10

Acute exposure to temperature

5-25 20 15-35 30 20-38 20 15-30

Acute exposure to temperature

15

15

Acute exposure to temperature O,, 0.8-1.9 mgiliter Thermal acclimated, metabolism Photography Photography

1.3

21

28.0

5.s5.5

21

21

88 18.8-30.7 20-41 24-43 4.8-14.6 5.S16.8 7.2-23.9 11.1-27.0 8.5-29.2 17.7-31.2 I8.X-30.7 24-55 32-3-58 4s-€3 47-64 48-66 40-60 0.64.6

4. I 3.2-5.4 2.45.0 2.8-7.8 2.2-6.5 2.b7.8 3.blO.l 5.1-12.4 3.6-13.0 7.8-13.6 33-5.4 1.6-2.0 2.2 2.0-3.0 2.43.1 2.43.2 2.2-2.7 1.0-3.3 3.6 0.7-3.3 3.8 2.3

20 25 25

20 25 25 5-20 5-25 10-30 10-30 10-30 20-30 20

0..>5.0

Acute exposure to temperature

Photography

21

I13 61

Metabolism Metabolism

Photography

2.1-3.8 2.1-3.7 3.5 2.4-3.2 7.6-12.6 5.9-7.5 4.0-5.0

6fi

Comments

8 5 8

8 5

Endurance Endurance

8

Endurance Endurance, 50% fatigue Endurance, 50% fatigue Eudurance, O,, 6.5 mdliter; H2S, 0-0.15 mglliter Endurance, O,, 6.5 niglliter; H,S, 0-0.01 mglliter

22 22

5 10

15 20 25 30 20 10

15 20 25 .30

34 13 13

02,1-24 nig/liter 4, 1.2-8.1; CO, 3-54 mglliter Acute exposure to temperature Acute exposure to temperature Acute exposure to temperature Acute exposure to temperature Acute exposure to teniperature Acute exposure to temperature

10

15 20 25 30 34 13

Velocity, 50% fatigue

13

Velocity, 50% fntigue

Reference Peterson (1974)' Peterson (1974)' Gibson and Fry (1954)' Gibson and Fry ( 1954)' Gray (1953) Magnan (1929) Fry and Hart ( 1948)& Fry and Hart ( 1948)' Fry and Hart ( 1948)' Fry and Hart ( 1948)' Fry and Hart ( 1948)' Fry and Hart ( 1948)' Fry and Hart ( 1948)' Kutty ( 1 9 6 8 ) O Smitet al. (1971)' Gray (1953) Gray (1953) McLeod (1967)' Magnan ( 1929) Gray (1953) Magnan ( 1929) Magnan (1929) Beamish (1966h)" Beamish (1966b)' Magnan ( 1929) Beamish (1966b)" Kerr ( 1953)' Kerr ( 1953)" Oseid and Smith (1972)' Oseid and Smith (1972)' Mngnan (1929) MacLeod (1967)' Dahlberg et ul. (1968)' Dahlberg et ul. (1968)' Larimore and Duever ( 1968)b Larimore and Duever ( 1968)' Larimore and Duever (1968)' Larimore and Duever ( 1968)' Lwimore and Duever (1968)b Larimore and Duever ( 1968)' Larimore and Duever (I968)* Beamish (1970)" Beamish (1970)" Beamish (1970)" Beamish (1970)'' Beamish (1970)" Beamish (1970)" Houde (1969)" Maguan (1929) Houde (1969)" Magnan (1929) Magnan (1929)

(Continued) 149

150

F. W. H. BEAMISH

Table VI-

Species

iiumIher

Trirchurus symmetricus Truchunrs symmetricus Cobivs puuiotilus Gobius syrnisii Cohius nielonostoniu Scornber scornlnus Sehster dnctyloptenrs Sebostes morinus Sehos tes nu rinu,s Sebostes inarinus Hernilriptews uniericunus .Uyoxocephulus oclodecieispinosus Pseudul,leirrotiectes unmricunus Pseurlopleiironectes umericunus Pseudopleurmiectes omericonus Pseudopleuronectes oinericonrrs

Totnl length (cm)

15 8 13

14.6 11.2 8.8

I1 42

10.6 9.0 25.3

40 40 70 16 50 .30 30 30

20

26.8 17-19 1617 1617 18.8-22.7 19.8-21.3 19-21 19-21 19-20 22-23

Weight

(n)

2R.6 25.2 29.3

Velocity increments (cm sec-')

Time hetween increments (niin)

Gradually to prescribed Gradually to prescribed Gradually to prescribed Gradually to prescribed Gradually to prescrihed

velocity velocity velocity velocity velocity

Swimming time at inaximum velocity (inin) 3.4 4.5 1&15 R-10 ,>4

,MI 127-166 113-137 114-140 231-271 13C-162 135-193 149-190 130-171 215-219

Gradually to prescribed velocity Gradually to prescribed velocity Gradually to prescribed velocity Gradually to prescribed velocity Gradually to prescribed velocity Gradually to prescribed velocity Gr:idually to prescribed velocity Gradoally to prescribed velocity Gradually to preacribed velocity

0.7-14.2 0.R-13.2 0.612.7 0.4-0.7 0.5-6.7 1.%14.1 1.7-10.3 1.4-10.0 5.3-25.4

among the former species. When trout and salmon were exposed to 100 cercariae daily for 15 days, which more closely approximated the natural rate of infection than a single large dose, endurance was reduced by 51 and 34% in trout and salmon, respectively. Hemorrhagic areas developed in the infected fish but not in the controls, to which Butler and Milleman (1971) attributed impairment of relative swimming performance. The incidence of parasitism among fish used in swimming studies is seldom reported. The recent findings strongly suggest that closer attention to the species, stage, and abundance of parasites present in experimental fish would significantly reduce variability in measurements of performance, and better facilitate quantitative comparisons among the results of swimming speed studies.

D. Environmental Constraints on Performance

1. TEMPERATURE Swimming capacity is regulated by the metabolic capacity of fish to convert chemical energy into propulsive thrust through muscular contraction. Adenosine triphosphate (ATP) generated by the stepwise degradation of carbohydrate and lipid is an essential prerequisite for muscle contraction. In sustained swimming the formation of ATP by aerobic processes would be expected to be more important than that

2.

151

SWIMMING CAPACITY

Continued Temperature ("C) Maximum velocity c ~ nsec-' 160

139 48-50 44 34 HI 98 52- 135 52-135 52-135 105-120 6(!-120 75-13.5 75-135 7,5135 75-135

't

sec-I

10.8 12.4 5.5-5.7 4.2 3.8

Accliination

Experimental

18.5

18.5

18.5

18.5

Comments

Reference

Endurance, 50% fatigue Endurnnce, 50%ftitigue

Hunter (1971)" Hunter (1971)" Shazkiiia (197213)" Shazkina ( 197211)" Shazkina (197213)" Mngnan (1929) Magnan ( 1929) Beamish (196613)" Beamish (196611)" Beamish ( 196613)" Beninish (196613)" Beainish (196613)" Beamish (196611)" Beamish (1966h)" Beamish (196613)" Beainish (196613)"

19-21 19-21

19-21

:3.2 3.6 3.1-8.0 3.1-8.0 3.1-8.0 4.6-6.4 2.8-5.7 4.0-6.5

3.6-6.6 3.9-6.6 3.S6.1

5 8

5

8

8 11 8 8

5 8 11

5 8 11

14

14

11 8

Endurance Endurance Endurance Endurance Endurance Endurance Endurance Enduraiice Endurance

by anaerobiosis. The contribution by both aerobic and anaerobic reactions is important in prolonged swimming (Bilinski, 1974). Burst swimming relies heavily on the anaerobic mobilization of metabolites from carbohydrate sources (Drummond and Black, 1960; Black et al., 1961; Dean and Goodnight, 1964; Beamish, 1966c, 1968; Drummond, 1967, 1971; Dando, 1969).Since the physiological mechanisms associated with swimming vary with the category of locomotion, it is not surprising that they should differ in their response to temperature.

a . Sustained and Prolonged Swimming Speed. A t swimming speeds where aerobic processes contribute significantly to the production of ATP the influence of temperature can perhaps best be understood from its relationship with oxygen consumption (see Brett, 1970b; Fry, 1971, for comprehensive reviews). I n the metabolic processes oxygen serves as a final acceptor in the electron transport system rather than participating directly in the enzyme reactions involved in biological oxidation. Of the oxygen consumed, a portion serves to meet the basal metabolic requirements while the remainder provides for other activities including swimming, digestion, excretion, and growth. The amount of oxygen that can be extracted is, within limits, related to or dependent on the environmental conditions (Beamish and Dickie, 1967; Fry, 1967))each factor acting independently or interacting with others to alter the potential expression of maximum metabolic rate. The term "scope for activity" was employed by Fry (1947) to illustrate the effect of environmental identities on the oxygen available to verte-

152

F. W. H. BEAMISH

brate poikilotherms for activities excluding basal metabolism. Changes in scope for activity in relation to temperature are illustrated in Fig. 11 for brook trout (Graham, 1949), lake trout, Salvelinus namaycush (Gibson and Fry, 1954) and largemouth bass (Beamish,

1970).

Temperature, O C Fig. 11. Metabolic scope for activity of (A) brook trout, Salvelinus f o n t i n a h (Graham, 1949); (B) lake trout, Salvelinus namaycush (Gibson and Fry, 1954); (C) largemouth bass, M i c r o p t m s salmoides (Beamish, 1970), in relation to temperature.

153

2. SWIMMING CAPACITY

Within the thermal range of tolerance for a species, prolonged speeds typically increased with temperature to a maximum and thereafter decline. Examples of prolonged swimming speeds in relation to temperature of acclimation are presented in Fig. 12. Temperature compensation, customarily applied to adaptive evolution of metabolic performance (Bullock, 1955; Fry, 1958; Roberts, 1966; Brett, 1970b)

0

10

20

30

40

Temperature,O C Fig. 12. Prolonged swimming speed and temperature. (From Fry and Hart, 1948; Gibson and Fry, 1954; Brett et al. 1958; Wohlschlag, 1964; Larimore and Duever, 1968; Beamish, 1970; Brett and Glass, 1973.)

154

F. W. H. BEAMISH

applies also to prolonged swimming speeds. Among the eurythermal temperate species such as goldfish (Fry and Hart, 1948) and large and smallmouth bass (Beamish, 1970; Larimore and Duever, 1968) the temperature for maximum prolonged performance lies within the range of 25"-30°C (Table VI). Stenothermal temperate species such as the lake trout, Saluelinus namaycush (Gibson and Fry, 1954), display maximum prolonged speeds between 15"and 20°C.The prolonged performance of the antarctic stenothermal, Trematomus borchgreuinki (Wohlschlag, 1964), in contrast to the temperate species was highest at -0.8"C and fish were unable to swim at temperatures above 2°C. Prolonged performance may vary severalfold within the range of thermal tolerance for a species. The ( 3 min, 2.2 cm sec-l) prolonged speed for smallmouth bass (2.2cm) increased from 4.8 to 31.2 cm sec-' (2.2-14.2 e sec-') between 5" and 30°C (Larimore and Duever, 1968). The prolonged (60 min, 10 m sec-I) critical for sockeye salmon (20 cm) varied from just over 50-90 cm sec-l (2.5-4.5 6 sec-l) between 1" and

15°C. The response of prolonged swimming speed to thermal acclimation is presented for three species in Fig. 13. In each, maximum performance progressively shifts toward a higher exposure temperature as acclimation is increased. This, of course, suggests that the capacity for activity over much of the range of thermal tolerance for a species is greatest at environmental temperatures equal to or above those to which it is acclimated. For example, goldfish acclimated to 5" and 25°C performed best at 18" and 28"C, respectively (Fry and Hart, 1948). Similarly, the ( 3 min, 2.2 cm sec-*) prolonged speed of smallmouth bass, acclimated to lo" or 15°C occurred at 22" and 25"C, respectively (Larimore and Duever, 1968). Griffiths and Alderdice ( 1972) thoroughly investigated the influence of acute temperature exposure on the (60 min, one-eighth critical) critical speeds of young coho salmon (7.9-9.5 cm). Maximum critical speeds occurred at test temperatures above the acclimation, denoted in Fig. 13 by line A. At a given test temperature, maximum performance coincided closely with acclimation, in the figure described b y ridge B. Optimum performance occurred at a combination of acclimation and test temperatures near 20°C. Fry (1967) concluded, based on earlier evidence, that maximum swimming performance for a given temperature occurs when acclimation and exposure are identical which is generally concordant with the observations of Griffiths and Alderdice (1972).

b . Burst Swimming Speed. Temperature appears to exert little influence on burst speed although information at this level is particularly

2.

155

SWIMMING CAPACITY

B M i c r o ~ t e r u sdolomieu

,

,

,

,

,

,

,

,

l O

0

10

0

40

30

20

10

26

t

Oncorhynchus klsulch

0

2

,

1

1

8

20

90

40

Test Temperature,'C

Temperature, OC 1

1

14

1

1

1

1

C

20

26

Acclimation TemperaturePC Fig. 13. Prolonged swimming speed and temperature. Th e heavy lines in panels (A) and (B) denote prolonged for fish acclimated to the test temperatures. Light lines indicate the response between prolonged and test temperature for a particular acclimation temperature (Fry and Hart, 1948; Larimore and Duever, 1968). I n panel (C) the U critical speed isopleths (cm sec-l) for coho salmon, Oricorhynchus kisutch, are presented in relation to acclimation and test temperature (Griffiths and Alderdice, 1972).

156

F. W. H. BEAMISH

scarce. Blaxter and Dickson (1959) measured the burst speeds for a number of marine and freshwater teleosts (Table IV) and were unable to demonstrate any correlation between performances and temperature. Based on metabolic studies of sockeye salmon in relation to performance, Brett ( 1964) anticipated temperature independence for burst swimming. The swimming endurance of redfish, Atlantic cod, and winter flounder, Pseudopleuronectes americanus, at speeds approaching burst velocities did not vary appreciably over the ecological range of temperatures experienced throughout much of the year in the northwest Atlantic ( Beamish, 196613). More recently Groves (reported in Brett, 1970b) reported a temperature independence in the burst speed achieved but a dependence in terms of endurance by sockeye salmon.

2. OXYGEN In aquatic organisms which use oxygen for their respiration, the ambient oxygen consumption itself can limit swimming performance. There appears for most fish a threshold oxygen concentration below which swimming performance is reduced (Dizon, 1977).

a. Sustained and Prolonged Swimming Speed. Kutty (1968) and Kutty and Saunders (1973) introduced the term “critical oxygen concentration” to describe the concentration at which fish are unable to maintain sustained or prolonged speeds. Thus Atlantic salmon (23.4cm) sustained speeds of 50 and 70 cm sec-l (2.1and 3.0 C sec-l) for several hours until ambient oxygen was reduced to 4.0 and 4.8 mg O2 liter-’, respectively (Kutty and Saunders, 1973; see also Fig. 14). Critical oxygen concentrations of goldfish (18.5 cm) at 59.2 and 18.6 cm sec-’ (3.2 and 1.0 C sec-l) were 1.8 and 0.8 mg O2 l i t e r 1 , respectively (Kutty, 1968). Similar reductions in sustained and prolonged swimming performance in the presence of low oxygen have been demonstrated for a number of species (Graham, 1949; Katz et al., 1959; Davis et al., 1963; Whitworth and Irwin, 1964; McLeod and Smith, 1966; Dahlberg et al., 1968; see also Table VI; Fig. 14). Above air saturation, prolonged performance of coho salmon and largemouth bass was independent of oxygen (Dahlberg et al., 1968; see also Fig. 14). In contrast to the earlier observations by Prosser et al. (1957) on goldfish, Kutty (1968) found acclimation to low ambient oxygen did not alter the critical oxygen concentrations for a given sustained swimming speed. Failure to swim at low ambient oxygen was not attributed by Kutty (1968) to fatigue as fish began to swim as soon as oxygen levels were increased, but rather to an oxygen sensing mecha-

2.

157

SWIMMING CAPACITY

v I

0

Q

u)

E

0

'0 Q

Q

a v)

D

.-C

E

.-E

3

v)

407-M icropterus salmoides

20

0

1

1

1

1

1

1

1

1

1

1

,

1

,

Fig. 14. Swimming speed and ambient oxygen concentration. In the upper panel, critical oxygen concentrations at which fish are unable to maintain a sustained speed (Kutty, 1968; Kutty and Saunders, 1973). In the lower two panels, the relationship, between prolonged speed and oxygen (Dahlbery et ul., 1968).

nism such as the peripheral or central oxygen receptors reported by Saunders and Sutterlin (1971).

b. Burst Swimming Speed. The effect of dissolved oxygen on burst swimming has not been measured. However, burst speed depending as it does on anaerobic energy sources may be expected to be largely

158

F . W. H. BEAMISH

independent of ambient oxygen except that between swimming events the accumulated metabolic debt must be repaid before the next burst of swimming can realize its full potential. The mobilization of energy resources for repeated bursts and therefore the frequency of rapid swimming may well be restricted by moderate oxygen deficiency.

3. CARBONDIOXIDE Carbon dioxide has long been known to reduce the affinity of blood for oxygen (Root, 1931) and to influence the metabolic rate of fish (Basu, 1959; Beamish, 1964b). Particularly little information is available on the effect of carbon dioxide on swimming, a notable exception being the research of.Dahlberg et al. (1968). They measured the prolonged speed of largemouth bass and coho salmon in response to dissolved oxygen and free carbon dioxide. The prolonged speeds of largemouth bass did not change in response to concentrations of carbon dioxide to 48 mg liter-'. The performance of coho salmon in contrast to that observed for bass declined on exposure to concentrations of carbon dioxide between 2-61 mg liter-'. In low concentrations of oxygen the influence of carbon dioxide was less pronounced. For an oxygen concentration of 10 mg liter-l, the prolonged speed of salmon decreased from about 60 to fractionally above 50 cm sec-' when carbon dioxide increased from 2 to 61 mg liter-'. In contrast, when ambient oxygen was about 2 mg liter-I, prolonged speed did not change with increase in free carbon dioxide.

4. SALINITY Salt concentration in the blood of fish is less than that of seawater. In a marine environment water is lost at the gills and other body surfaces (Potts, 1954; Black, 1951). Conversely freshwater homeostasis is dependent on the elimination of absorbed water, the concentration of the body fluids being greater than that of the environment (Black, 1957).The mechanism by which osmoregulation is achieved may vary among species (Parry, 1958; Gordon, 1963; Threadgold and Houston, 1964) but each requires the expenditure of energy. Few measurements have been made on the relationship between salinity and swimming performance. A consistent pattern of change in the swimming speed of skipjack and yellow fin tuna did not occur in response to a salinity decrease from 34 to 2W/00 (Dizon, 1977).Critical speeds (60 min, one-eighth critical) of coho salmon in relation to salinities and temperatures between 0-20./, and 3"-23"C, respectively, were measured by Glow and McInerney (1977) (Fig. 15, Table

2.

159

SWIMMING CAPACITY

Oncorhynchus k isutc h

,

25

20

Fry, 8.9 crn

--_:--

'

40

44

i' \

Fig. 15. Critical swimming speed isopleths (cm sec-I) for coho salmon fry and smolts, Oncorhynchus kisutch, in relation to salinity and temperature (Glova and McInerney, 1977).

V). The combined effects of salinity and temperature indicated critical swimming performance of underyearling coho was predominantly a temperature-dependent response during the premigratory stages of development. Swimming performance of fry was almost independent of salinity as reflected by the flat configuration of the performance isopleths in Fig. 15. Coho smolts achieved maximum critical speeds at

160

F. W. H. BEAMISH

salinities ranging from just under 8 to about 190/,. Relative to this salinity optimum, critical, performance declined by about 6 and 4 cm sec-’ at 0 and 20”/,, respectively. Just prior to, or concurrent with smoltification, coho appear to lose their euryhaline capacity to function efficiently over a range of salinities which is concordant with observations on salinity tolerance for this species (Alderdice, 1963). Performance restrictions attributable to salinity have not been measured for other species. However, fluctuations in metabolic expenditure in association with prolonged speeds have been measured for rainbow trout (Rao, 1968) and Tilapia nilotica (Farmer and Beamish, 1969) in salinities of 0-30”/,. In both studies the energy actually expended in swimming was independent of salinity. Nevertheless, changes in metabolic rate did occur, suggesting that performance capacity would be reduced in proportion to the energy expended in osmotic regulation.

5. THEDISTURBED ENVIRONMENT The influence of those identities introduced into the environment either directly or indirectly by man or through his activities has received attention only in recent years with most of the effort expended in the determination of their lethal concentrations. Of particular concern to the swimming capacity of fish are those identities which influence the exchange of respiratory gases or the metabolic pathways involved in the mobilization of energy. From among the many factors that may contribute to the disturbed environment, only a few have been examined with respect to their influence on swimming performance. Conifer pulpwood fiber in suspension impairs the removal of oxygen from water by physically clogging the gill lamellae and interrupting the respiratory flow during gill cleaning reflexes (MacLeod and Smith, 1966). On exposure to suspensions of pulpwood fiber equivalent to 200 mg liter-’ the endurance of fathead minnows, Pimephales promelas, forced to swim at a low velocity was significantly less than that of fish in freshwater under comparable concentrations of dissolved oxygen (Fig. 16A). The influence of the pulpwood suspension was most pronounced at the higher temperatures (Fig. 16B) which is consistent with its impairment of gaseous exchange. In contrast to the influence on endurance at prolonged swimming, burst swimming speed was independent of the concentration of pulpwood fiber. Hydrogen sulfide is found not infrequently in the aquatic environment and results from the decomposition of material either naturally occurring or present through the activities of man. Endurance of bluegills (3.2 cm) forced to swim at 22.5 cm sec-’ increased from just

161

2. SWIMMING CAPACITY 80-

60

-

Q 40

-

E

Plmephales promelas

t

al 0

c

15OC

CI

rn

i 20-

0

-

40

A ,

Pulpwood Fiber 1

2

1

1

4

1

1

1

6

al

1

0

0

50 Saivelinus fontinails

u)

E

Pulpwood Filter

21°c

200

100

Fiber Concentration, mg liter"

Oxygen, mg liter''

I

0

-

t1

tI

Salmo galrdneri

t

0

Fenitrothion, mg Iihr-'

Copper, mg liter"

Fig. 16. (A) The influence of pulpwood fiber and dissolved oxygen on the distance swum by fathead minnows, Pimephales promelas (MacLeod and Smith, 1966). (B) Temperature, oxygen, and pulpwood fiber effect on swimming distance by fathead minnows (MacLeod and Smith, 1966): ( C ) Effect of fenitrothion on prolonged speed of brook trout, Saloelinus fontinalis (Peterson, 1974). (D) Effect of copper on critical speed of rainbow trout, Salmo gairdneri (Waiwood, personal communication).

over 200 min in the absence of hydrogen sulfide to 240 min at 0.4 pg liter-'H2S, and with further increases, decreased so that at 14.6 p g liter-' fish swam only for 30 min (Oseid and Smith, 1972). Coincident with the long term exposure to H2S was an increase in the rate of gill irrigation which undoubtedly lead to an appreciable reduction in the mobilization of energy through aerobic processes. Sodium pentachlorophenate (PCP), used as a defoliant or for the protection of timber from wood-boring insects and fungal infection, is considered a general metabolic poison for fish (Webb and Brett, 1973).

162

F. W. H. BEAMISH

However, the (60 min, 5 cm sec-l) critical speed of sockeye salmon (5.3-6.0 cm) did not change significantly from about 40 cm sec-I (7.3C sec-’) on exposure to concentrations of PCP between 0-50 pl-1 (Webb and Brett, 1973). Similarly, Krueuger et al. (1966) found that swimming performance of Cichlasoma bimaculatum was not reduced by pentachlorophenol until the concentration approached lethal levels. Webb and Brett (1973) proposed that a general metabolic poison such as PCP should not “preferentially” affect the gas exchange system particularly where excitement, as included in fish forced to swim, serves as a “stressor.” The influence of fenitrothion, an organophosphate insecticide used in the control of spruce budworm, on prolonged swimming speed of brook trout was determined by Peterson (1974). Prolonged swimming speed decreased from 5.0 C sec-’ for controls to 3.5 C sec-’ for trout exposed to 1.5 mg liter-’, the highest concentration applied (Fig. 16C). While the mechanism through which fenitrothion reduces swimming performance is unknown, Peterson (1974) suggested it may cause impairment of those areas of the nervous system concerned with muscle activity or alternately by causing indirect effects through “motivational” disturbances. In a comprehensive study on the influence of bleached kraft mill effluent (BKME), Howard (1975) measured the (60 min, 5 cm sec-l) critical speed of coho salmon to concentrations equivalent to 90% of the level at which 50% of the fish died within 96 hr (96 hr LC,,). Exposure for 18 hr to a concentration of 0.9 LCso resulted in a 72% reduction in swimming capacity. Further critical speed for a given concentration of BKME was independent of exposure time beyond 18 hr and returned to control levels within 6-12 hr after being placed in effluent-free water. In swimming fish, Howard (1975) suggests BKME retards gaseous exchange either by absorption of the effluent on the gill surface or through the formation of a weak chemical bond to the gill epithelium. In addition, Javaid ( 1973) observed ventilatory irregularities among sockeye salmon exposed to BKME. Effluents from mining operations discharged into waterways may also exert a pronounced influence on the capacity of fish to perform. Waiwood (personal communication) measured the influence of total copper and p H in relation to water hardness on the (60 min, 5 cm sec-l) critical swimming speed of rainbow trout (Fig. 16D). H e found that for a given hardness, critical speed was reduced by increasing concentrations of copper but that the effect diminished with time of exposure to about 10 days (Fig. 16D). Further, the influence of a given concentration of copper on performance decreased inversely with

2.

SWIMMING CAPACITY

163

water hardness. Copper is known to have a deleterious effect on the composition of blood (McKim et al., 1970) and to damage various tissues including the kidney, liver, intestine, and cephalic lateral canals (Baker, 1969; Gardner and Laroche, 1973). Tissue damage would undoubtedly cause an elevation in basal metabolism and a decline in the scope for activity. Swimming performance, depending as it does on the immediate recruitment of energy, has been recommended for use as a criterion in the determination of the sublethal effects of pollutants on fish (Brett, 1967; Sprague, 1971). However, the proper application of swimming speed, as well as the category of performance to be tested as a criterion of sublethal effect, requires a comprehensive prior knowledge of the pharmacological effects of the pollutants concerned. Impairment of the gaseous exchange of mechanism might, for example, be masked in burst swimming speeds which depend on anaerobic processes. Further, the capacity exhibited by some fish to acclimate in part or even fully to a given pollutant should be respected by the serious investigator.

IV. ENERGETICS OF SWIMMING

The expenditure of energy during swimming is reflected in gaseous exchange and should include measurements of both oxygen consumption and carbon dioxide production. However, due to limitations imposed by the techniques available, measurements of carbon dioxide production are infrequently made, the researches of Kutty (1968) being a notable exception. More generally, calculation of the energy expenditure for swimming is made from units of oxygen consumption and converted to units of energy on the basis of an oxycalorific coefficient derived for domestic homeothermic animals (Brody, 1945). This coefficient assumes not only the complete oxidation of catabolized substrates but also a normal balance of the sources such as is implied by an average respiratory quotient of 0.8. Winberg (1956) suggested that irrespective of the components oxidized, the oxycalorific coefficient will not vary more than 1.5%. It is generally agreed an oxycalorific coefficient of 3.363.44 is most applicable for teleosts (Warren and Davis, 1967; Brett, 1973; Beamish et al., 1975). In contrast, Krueger et al. (1968) reported that calorific loss based on lipid depletion in strenuously exercised salmon was substantially greater than that estimated from respiratory rates, and questioned the method of evaluation of energy production indirectly from oxygen consump-

164

F. W. H. BEAMISH

tion. Brett (1973) on the other hand, concluded an oxycalorific coefficient of 3.36 cal mg 0 2 - 1 consumed is acceptable for teleosts and that under carefully regulated experimental conditions, estimates of energy expenditure made from the oxygen consumed by sockeye salmon are not at variance with those based on the utilization of body components. In sustained swimming the mobilization of energy is achieved through aerobic processes so that the quantity of oxygen consumed is proportional to the amount of work performed. Fish swimming at prolonged speeds derive energy from both aerobic and anaerobic processes, the contribution from the latter increasing with the severity of exercise. At prolonged speeds, utilization of glycogen stores was reported by Pritchard et al. (1971) as the principal cause of swimming failure in the jack mackerel. The evolution of respiratory gases has not been measured for burst swimming because of the practical difficulties imposed by the short duration of muscular activity. It is presumed, however, that at burst swimming, fish consume some oxygen and that the remainder of the energy requirement is met through anaerobic processes. The latter results in an oxygen debt which is repaid subsequent to the termination of exercise. The allocation of aerobic and anaerobic processes in relation to the category of swimming is summarized in Fig. 17. Swimming energetics is the subject of several reviews which should be consulted (Fry, 1957; Fry and Hochachka, 1970; Brett, 1962, 1970a, 1972; Beamish and Dickie, 1967; Randall, 1970; Doudoroff and Shumway, 1970; Schmidt-Nielsen, 1972; Bilinski, 1974). Present evidence indicates that swimming may elevate the total metabolic rate by as much as 15-fold (Beamish, 1964a; Brett, 1964). The oxygen consumed at sustained and prolonged swimming speeds is presented for a number of species in Fig. 18. Subtraction of standard or basal metabolism from the total oxygen uptake has been used to approximate the expenditure of energy associated with a particular swimming speed. Where anaerobic processes do not contribute significantly this would appear a satisfactory procedure. The rate of increase in the logarithm of oxygen uptake with relative swimming speed in Fig. 18 is surprisingly similar among species despite obvious variation in methodology, size, and temperature and is reasonably well represented by a coefficient of 0.36. Thus for each increase in relative swimming speed of C sec-' there is a corresponding 2.3-fold elevation in metabolic rate. In severe prolonged and burst swimming, caution must be exercised in not accounting for energy expenditure by anaerobic processes. The anaerobic contribution is perhaps most conveniently as-

165

2. SWIMMING CAPACITY Resting Repayment

0

45

Reperfoman

90

Swimming Speed, cm sec-l

Recovery Time, h

Fig. 17. Oxygen consumption and debt for sockeye salmon, Oncohynchus nerka (18 cm), in relation to swimming speed and recovery at 15°C.(Redrawn from Brett, 1964, I. Fish. Res. Board Can.)

sessed by continued measurement of oxygen consumption on completion of swimming until it returns to preexercise levels at which time the 'oxygen debt is presumably repaid (Heath and Pritchard, 1962; Brett, 1964; Smit et al., 1971).This procedure assumes the products of anaerobic metabolism such as lactate are not excreted but subsequently oxidized during the recovery phase following exercise. Recently, Karuppannan (1972, reported in Kutty and Peer Mohamed, 1975) has shown that Tilapia mossambica excrete some lactate after strenuous exercise, corroborating the earlier observations by Blaika (1958) on the anaerobic metabolism of crucian carp, Carussius carassius. The maximum rate of oxygen consumption among fish species appears to vary at least Sfold with maximum values in excess of 2000 mg kg-l hr-I (Stevens, personal communication). The metabolic capacity of the higher vertebrates is generally one or two orders of magnitude above that demonstrated for teleosts (Bartholomew and Tucker, 1963, 1964; Bartholomew et al., 1965; Tucker, 1970; Brett, 1972). This discrepancy is compensated for, in part by a greater tolerance by teleosts to oxygen debt but from which recovery is slow. In sockeye salmon the rate of replacement of oxygen debt following fatigue was in excess of 3 hr and independent of temperature (Brett, 1964). The magnitude of the debt accumulated at the time of fatigue was influenced by temper-

mm 100

100

1.0

21)

Swimming Speed,

3)

4.0

0

1 sec-l

Fig. 18. Oxygen consumption, energy utilization, and swimming speed. Th e total oxygen consumed is presented in the upper panel. Subtraction of the basal from total metabolic rate provides a measure of the energy required for a given speed of swimming. Th e heavy line in the lower panel denotes the general rate of increase in net oxygen consumption and was fitted by eye. (From Basu, 1959; Beamish, 1964a, 1970; Brett, 1964; Brett and Sutherland, 1965; Farmer and Beamish, 1969; Kutty, 1969; Webb,

1971b.)

2.

167

SWIMMING CAPACITY

ature and increased 2-fold between 5” and 15°C. Similarly, Heath and Pritchard ( 1962) found that bluegill sunfish, after strenuous exercise, maintained a high consumption of oxygen followed after 1 hr by a gradual decline to preexercise levels 10-24 hr later. Schmidt-Nielson ( 1972) expressed the energy cost for locomotion independently of swimming speed as the caloric expenditure to transport 1 unit of body mass 1 km. A reanalysis of Brett’s data on prolonged swimming speeds of sockeye salmon b y Schmidt-Nielsen showed a logarithmic linear decrease in energy expenditure with increase in body weight over a range of three orders of magnitude (Fig. 19). The application of this expression of the energetic cost of locomotion to other species exercised at sustained and prolonged speeds under different environmental conditions shows a remarkable similarity to the relation described for salmon in Fig. 19. Closer examination of the comparative energy cost at low and high prolonged speeds based on measurements of oxygen uptake indicates a reduction as velocity is increased. Thus, the energy expenditure for mullet, Liza rnacrolepis, declined from 2.09 to 1.48 cal g-I km-’ between 10 and 22.5 cm sec-’ (Kutty, 1969),which presumably reflects a greater contribution of anaerobic metabolism at the higher speeds. Refinement in the measurement of total metabolism of swimming fish, while desirE ’Y

Swimming

1.0

Fish

-

Micropterus salrnoides

Lepornis gibbosus’

*

Coregonus

P

al C

w

0.1

I

1

I

10

I

I

I

100

1000

Weight, g Fig. 19. Energy cost of swimming relative to body size calculated by SchmidtNielsen (1972) from data collected by Brett (1964), Wohlschlag et al. (1968), Matyukhin and Stolbow (1970, reported in Schmidt-Nielsen, 1972), Rao (1971),and Smitet al. (1971). In addition, measurements made by Brett and Sutherland (1965), Farmer and Beamish (1969), Kutty (1969), Tytler (1969), and Beamish (1970) have been recalculated and added to the figure.

I

168

F. W. H. BEAMISH

able, is unlikely to alter significantly the linear relationship described by Schmidt-Nielsen. Following Schmidt-Nielsen’s interesting hypothesis, Gold (1973, 1974) and Calder (1974) expressed the energy cost of swimming one body length in terms of the mass of propulsive muscles relative to total weight and multiplied by the number of muscle contractions or tailbeats required to transport the animal one length. This assumed a constancy in the quantity of energy available per contraction per unit of muscle mass equal to 1 cal kg-l. Energy for muscular contraction is derived from the hydrolysis of adenosine triphosphate to adenosine diphosphate and inorganic phosphate. The evolution of adenosine triphosphate through the catabolism of organic compounds may occur under both aerobic and anaerobic conditions, the former being the more efficient in terms of yield but each offering distinct advantages to the swimming teleost. In sustained swimming where the amount of oxygen consumed is proportional to the work performed, the main source of energy is from long-chain fatty acids and to a lesser extent protein and glycogen (Greene, 1926; Idler and Tsuyuki, 1958; Drummond and Black, 1960). When the capacity for aerobic metabolism is exceeded as in severe prolonged or burst swimming, adenosine triphosphate is synthesized by anaerobic glycolysis of stored muscle glycogen. Lactic acid, the end product of glycolysis, diffuses from the muscle into the bloodstream (Nakatani, 1957; Black, Connor et al., 1962; Driedzic and Hochachka, 1975). Both swimming and the accumulation of lactic acid may continue until the glycogen deposits are depleted or the end product of anaerobic glycolysis exerts a detrimental effect on activity. Where exercise is extreme in its severity, death may result during the recovery period (Parker et al., 1959; Beamish, 1966c; Caillouet, 1967). The actual cause of death is uncertain but may result from interference with the acid base equilibrium, coupled with reduced affinity of hemoglobin for oxygen, and, in the presence of excess acid, lowered affinityfor carbon dioxide (Black, 1958a).When death does not follow strenuous exercise, the elevated rate of oxygen consumption serves not only to meet the routine metabolic requirements but also to replace muscle supplies of adenosine triphosphate, creatine phosphate, and glycogen (Bilinski, 1974).

V. APPLICATION TO MANAGEMENT PRACTICES Hatchery breeding programs have tended to select for qualities such as rapid growth, early maturity, high fecundity, and disease resis-

2.

SWIMMING CAPACITY

169

tance which, of course, are of obvious benefit to the culturist. However, when the objective is to stock desirable waterways with the view of generating a sustainable population, selective breeding programs may have overlooked qualities essential to the continued well being of the population. Swimming performance is an important component of viability as it relates to a fish’s capacity to maintain station against current, avoid predators, and acquire food. Bams (1967) proposed that unless severe environmental conditions impose a serious constraint, the most important component of survival is stamina. Vibert (1956) used the ability of fish to swim against a current as a test of adaptability for stocking. The importance of swimming performance is reflected by the higher survival of fish which were conditioned to a stream habitat prior to stocking (Shuck and Kingsbury, 1948; Miller,

1957). The stamina of hatchery and stream-conditioned rainbow trout was investigated b y Reimers (1956). Hatchery rainbow trout were able to swim against a current of 90 cm sec-’ for 5-10 min before fatigued, considerably less than the 30 min recorded for stream-conditioned trout. The performance of wild stocks of brook trout, even though reared under hatchery conditions, was consistently superior to that recorded for domestic stocks of the same species (Vincent, 1960; Green, 1964). The size at which fish are stocked may also influence their success. Survival of planted chinook salmon suggest a greater success among fingerlings than fry (Junge and Phinney, 1963). Thomas et al. ( 1964) attributed the greater survival of fingerlings to a number of factors including performance capacity. The role of nutrition on swimming performance and the ultimate capacity of planted fish to cope with the environment has not been examined but represents a potentially profitable area of research. Hatchery procedure in the incubation of eggs may also iegulate the ultimate capacity of the species to perform. Bams (1967) examined different methods of incubating eggs of sockeye salmon on the ultimate relative performance of fry. Naturally propagated salmon demonstrated the best relative performance followed by fry reared in gravel from the time of hatching and held, prior to the advanced alevin stage, in baskets or trays without a substrate. The poorest stamina was registered by fish reared in hatchery troughs without gravel at any stage. Investigations aimed at determining size and structure of fish stocks can be influenced b y the species’ resistance to fatigue. This may result when fish are tagged and released subsequent to capture by any method which involves severe muscular exertion on the part of the

170

F. W. H. BEAMISH

fish. Fox example, marine demersal species are frequently captured for tagging purposes by otter trawls which are towed along the seabed at speeds of 140-200 cm sec-' for 30 min or more. Many fish are unable to swim at these speeds for long and fatigue. This imposes a severe metabolic load, manifested by an oxygen debt, a depletion of glycogen reserves, and elevation in lactate, as well as a number of other physiological changes. Mortalities following capture by otter trawl have been reported and attributed to muscular fatigue. Among haddock, Melanogrammus aeglefinus, mortalities ranged between 7 and 78% of those captured b y otter trawl (Beamish, 1966~).Fatigue deaths in ocean troll-caught chinook and coho salmon were observed by Milne and Ball (1956,1958),Parker and Black ( 1959) and Parker et al. (1959). Barrett and Connor (1962) attributed some of the deaths of hook and line-caught yellowfin and skipjack tuna during recovery to fatigue. The steadily increasing demand for greater utilization of waterways has resulted in the construction of dams on rivers and the location of electrical generating plants near rivers, lakes, and oceans (Kerr, 1953).One of the problems associated with the construction of dams is that of preserving fish populations indigenous to the waters. On the west coast of North America, particular concern has been expressed for the well being of valuable stocks and anadromous trout and salmon. This entails providing safe passage for fish through waterways and over obstacles. One of the considerations in the construction and operation of a fishway is to provide flows at the entrance which will attract the desired species. Weaver (1963) conducted a series of velocity preference studies at the site of the Bonneville Dam on the Columbia River. The experiments were conducted in large dual channels and compared the frequency of fish passing through each in relation to water velocity. The results suggested the proportion of rainbow trout and chinook and coho salmon selecting the channel with the highest current speed applied, 240 cm sec-', was appreciably greater than that at any other velocity, the lowest of which was 60 cm sec-'. Information on the critical length of the passageway was provided from performance studies designed to measure the distance salmonids could swim at velocities to 500 cm sec-'. Mean maximum speed which rainbow trout (68.5 cm) could maintain for 9.14 m (1.5 sec) was 642 cm sec-' (9.5 8 sec-l) with one individual (61 cm) achieving 817 cm sec-' (13.4e sec-I). Maximum speeds for chinook (75.3cm) and coho salmon (51.0cm) were 604 and 421 cm sec-' (8.2and 8.2 8 sec-l), respectively. Earlier, Paulik and DeLacy (1957) measured the swimming speed of rainbow trout, coho, and sockeye salmon to provide information needed in the design of fishways. They found, in laboratory studies,

2.

SWIMMING CAPACITY

171

the maximum prolonged speed for rainbow trout (63.6cm) was 213 cm sec-1 (3.4e sec-I), well below that found by Weaver (1963). Similarly the maximum prolonged speed of coho (56.2 cm) 190 cm sec-l (3.5e sec-I) fell short of the subsequent measurements made by Weaver ( 1963).

Another of the basic problems in the design of fishways is the location, number, and size of resting pools. Recovery of coho salmon (65.6 cm) from an exhaustive swimming effort at 100 cm sec-' was found to be 31% complete after 1 hr rest and 67% after 3 hr. All fish recovered when allowed 18-24 hr (Paulik et al., 1957). From these data the investigators concluded the necessity for adequate resting facilities along a fishway when velocities exceeded 100 cm sec-' for more than a few minutes. Swimming performance of salmonids has been found to decline slightly as adult fish migrate upstream (Paulik and DeLacy, 1958). In Passamaquoddy Bay of the Bay of Fundy, a study was initiated to examine the possible effects of the construction of a series of dams on the fishery. Of major importance were the Atlantic herring which accounted for the vast majority of the total fish landings. Movements of herring indicated they entered the bay through narrow passages in which water velocity occasionally reached 300 cm sec-'. With the construction of the dams currents would have exceeded this speed. Laboratory measurements of swimming endurance at prescribed velocities indicated that had the dams been constructed, high currents together with the periods during which the dam gates were closed would have denied herring access to the bay for all but about 20 min every 24 hr. Observations on the swimming capacity of the western sucker, Catostornus occidentalis, as they moved upstream through a culvert prompted Wales (1950) to note the possibility of excluding undesirable species from portions of a river by regulating current speed. Swimming speed studies have been applied also in the design and assessment of fishing gear. The efficiency of otter trawls has received considerable attention as they supply much of the total annual harvest of fish from the marine environment. Reports from divers and from photographic observations (Blaxter and Parrish, 1966; Beamish, 1967) have indicated the orientation of fish swimming ahead of the trawl. With this information and the swimming capacity of the species the probability of escape can be estimated, assuming a straight line course and a fixed speed of swimming. Such estimates have been made for a number of demersal species in t h e northwest Atlantic by Beamish (1967) and the North Sea by Blaxter (1967; see also Fig. 20).

172

F. W. H. BEAMISH A

“ I

e ‘t

8.”

\\

I

I

impossible

t

12oL

B D

c U

100

-

1

f -

I

1

- 7

Speed required to escape, m sec-‘ Fig. 20. Swimming speed required to avoid capture by otter ?awl at different angles

of escape when the fish is in the center of the path of the net and reacts 3 m away (Blaxter, 1967). Arrow at 2 m sec-I indicates maximum burst speed for adult clupeoids. Panel (A) illustrates the change when nets of 12, 18, and 24 m in width are employed at speeds of 1.2 ni sec-I. Panel (B) indicates change when towing speed is varied from 1.2 to 2.4 m sec-l and net width is held at 12 m.

ACKNOWLEDGMENTS 1 am most grateful to Dr. E. D. Stevens for his comments on the manuscript and to Mrs. E. Thomas for her assistance with the figures.

REFERENCES Alderdice, D. F. (1963).Some effects of simultaneous variation in salinity, temperature and dissolved oxygen on the resistance of juvenile coho salmon (Oncorhynchus kisutch) to a toxic substance. Ph.D. Thesis, Univ. of Toronto. Aleev, Y. G. (1963). “Function and Gross Morphology in Fish,” 245 pp. Izd. Akad. Nauk SSSR, Moscow. (Transl. by Isr. Program Sci. Transl., Jeruselam, 1969; available as TT67-51391, NTIS, Springfield, Virginia.) Alexander, R. M. ( 1967). “Functional Design in Fishes,” 160 pp. Hutchinson, London. Alexander, R. M. (1968). “Animal Mechanics,” 346 pp. Univ. of Washington Press, Seattle.

2. SWIMMING

CAPACITY

173

Arnold, G . P. (1969). A flume for behaviour studies of marine fish. J . Exp. Biol. 51,

671-679. Bainbridge, R. (1958). The speed of swimming of fish as related to size and to the frequency and amplitude of the tailbeat.]. Exp. Biol. 35, 109-133. Bainbridge, R. (1960). Speed and stamina in three fish.]. E x p . Biol. 37, 129-153. Bainbridge, R. (1962). Training, speed and stamina in trout.]. Exp. B i d . 39, 537-555. Bainbridge, R., and Brown, R. H. J. (1958). An apparatus for the study of the locomotion of fish.J. Exp. Biol. 35, 134-137. Baker, J. T.P. (1969). Histological and electron microscopical observations on copper poisoning in the winter flounder Pseudopleuronectes americanus. J. Fish. Res.

Board Can. 26, 2785-2793. Bams, R. A. (1967). Differences in performance of naturally and artificially propagated sockeye salmon migrant fry, as measured with swimming and predation tests.]. Fish.

Res. Board Can. 24, 1117-1153. Barrett, I., and Connor, A. R. (1962). Blood lactate in yellow fin tuna Neothunnus macropterus and skipjack Katsuwonus pelamis following capture and tagging. Bull.

Inter-Am. Trop. Tuna Comm. 6,233-280. Barrett, I., and Hester, F. J. (1964).Body temperature of yellow fin and skipjack tunas in relation to sea surface temperature. Nature (London) 203, 96-97. Bartholomew, G . A., and Tucker, V. A. (1963).Control of changes in body temperature, metabolism, and circulation by the agamid lizard, Amphibolurus barbatus. Physiol.

Zool. 36, 199-218. Bartholomew, G . A,, and Tucker, V. A. (1964).Size, body temperature, thermal conductance, oxygen consumption, and heart rate in Australian varanid lizards. Physiol.

Zool. 37, 341-354. Bartholomew, G . A., Tucker, V. A., and Lee, A. K. (1965).Oxygen consumption, ternial conductance and heart rate in the Australian skink, Tiliqua scincoides. Copeia No. 2 , pp. 1969-1973. Bass, G . A., and Rascovich, M. (1965). A device for the sonic tracking of large fishes. Zoologica (N.Y.)50, 75-82. Basu, S. P. (1959). Active respiration of fish in relation to ambient concentrations of oxygen and carbon dioxide. J. Fish. Res. Board Can. 16, 175-212. Beamish, F. W. H. (1964a). Respiration of fishes with special emphasis on standard oxygen consumption. 111. Influence of weight and temperature on respiration of several species. Can. J . Zool. 42, 177-188. Beamish, F. W. H. (1964b). Respiration of fishes with special emphasis on standard oxygen consumption. IV. Influence of carbon dioxide and oxygen. Can. J . Zool. 42,

847-856. Beamish, F. W. H . (1966a).Vertical migration by demersal fish in the Northwest Atlantic.

J . Fish. Res. Board Can. 23, 109-139. Beamish, F. W. H. (1966b). Swimming endurance of some Northwest Atlantic fishes.J.

Fish. Res. Board Can. 23, 341-347. Beamish, F. W. H. ( 1 9 6 6 ~ )Muscular . fatigue and mortality in haddock, Melanogrammus aeglefinus, caught by otter traw1.J. Fish. Res. Board Can. 23, 1507-1521. Beamish, F. W. H. (1967).Photographic observations on reactions of fish ahead of otter trawls. FA0 Conf. Fish Behav. Relation Fish. Tech. Tactics, Bergen, Norway Exp. Pap. No. 25, pp. 1-11. Beaniish, F. W. H. (1968). Glycogen and lactic acid concentrations in Atlantic cod (Gadus morhua) in relation to exercise.]. Fish. Res. Board Can. 25, 837-851. Beamish, F. W. H. (1970). Oxygen consumption of largemouth bass, Micropterus salmoides, in relation to swimming speed and temperature. Can. J. Zool. 48, 1221-

1228.

174

F. W. H. BEAMISH

Beamish, F. W. H. (1973). Oxygen consumption of adult Petromyzon marinus in relation to body weight and temperature.]. Fish. Res. Board Can. 30, 1367-1370. Beamish, F. W. H. (1974). Swimming performance of adult sea lamprey, Petromyzon marinus, in relation to weight and temperature. Trans.A m . Fish. Soc. 103,355-358. Beamish, F. W. H. (1975). Apparent specific dynamic action of largemouth bass, Micropterus salmoides. J . Fish. Res. Board Can. 31, 1763-1769. Beamish, F. W. H., and Dickie, L. M. (1967). Metabolism and biological production in fish. In “The Biological Basis of Fresh Water Fish Production” (S. D. Gerking, ed.), pp. 215-242). Blackwell, Oxford. Beamish, F. W. H., and Mookherjii, P. S . (1964). Respiration of fishes with special emphasis on standard oxygen consumption. I. Influence of weight and temperature on respiration of goldfish, Curassius auratus L. Can. J . Zool. 42, 161-175. Beamish, F. W. H., Niimi, A. J., and Lett, P. F. K. T. (1975). Bioenergetics ofteleost fishes: Environmental influences. In “Comparative Physiology-Functional Aspects of Structural Materials” (L. Bolis, H. P. Maddrell, and K. Schmidt-Nielson, eds.), pp. 187-209. North-Holland Publ., Amsterdam. Bell, W. H., and Terhune, L. D. B. (1970). Water tunnel design for fisheries research. Fish. Res. Board Can., Tech. Rep. No. 195, pp. 1-69. Belyayev, V. V., and Zuyev, G . V. (1969). Hydrodynamic hypothesis of school formation in fishes. J . Ichthyol. (USSR) 9, 578-584. Bilinski, E. (1974). Biochemical aspects of fish swimming. In “Biochemical and Biophysical Perspectives in Marine Biology” (D. C. Malins and J. R. Sargent, eds.), pp. 239-288. Academic Press, New York. Bishai, H. M. (1960). The effect of water currents on the survival and distribution of fish larvae.]. Cons., Cons. Perm. I n t . Explor. Mer 25, 134-146. Black, E. C. (1955). Blood levels of haemoglobin and lactic acid in some freshwater fishes following exercise.]. Fish. Res. Board Can. 12, 917-929. Black, E. C. (1957a). Alterations in the blood level of lactic acid in certain salmonid fishes following muscular activity. I. Kamloops trout, Salmo gairdneri. J. Fish. Res. Board Can. 14, 117-134. Black, E. C. (1957b). Alterations in the blood level of lactic acid in certain salmonid fishes following muscular activity. 11. Lake trout, Saluelinus namaycush. J . Fish. Res. Board Can. 14, 645-649. Black, E. C. (1957~).Alterations in the blood level of lactic acid in certain salmonid fishes following muscular activity. 111. Sockeye salmon, Oncorhynchus nerka. ]. Fish. Res. Board Can. 14, 807-814. Black, E. C. (1958a). Hyperactivity as a lethal factor in fish.]. Fish. Res. Board Can. 15, 573-586. Black, E. C. (1958b). Energy stores and metabolism in relation to muscular activity in fishes. In “The Investigation of Fish Power Problems” (P. A. Larkin, ed.), H. R. hlacMillan Lectures in Fisheries, pp. 51-67. Univ. of British Columbia, Vancourver. Black, E. C., Robertson, A. C., and Parker, R. R. (1961). Some aspects of carbohydrate metabolism in fish. In “Comparative Physiology of Carbohydrate Metabolism in Heterothermic Animals” (A. W. Martin, ed.), pp. 89-124. Univ. of Washington Press, Seattle. Black, E. C., Connor, A. R., Lam, K., and Chiu, W. (1962). Changes in glycogen, pyruvate and lactate in rainbow trout (Salmo gairdneri) during the following muscular activity. J. Fish. Res. Board Can. 19, 409-436. Black, V. S. (1951). Changes in body chloride, density and water content of chum and

2.

SWIMMING CAPACITY

175

coho salmon fry when transferred from freshwater to seawater. J . Fish Res. Board Can. 8, 164-177. Black, V. S. (1957).Excretion and osmoregulation. In “The Physiology of Fishes” (M. E. Brown, ed.), Vol. 1, pp. 163-205. Academic Press, New York. Blaxter, J. H. S . (1962). Herring rearing. 4. Rearing beyond the yolk-sac stage. Mar. Res. pp. 1-18. Blaxter, J. H. S. (1967). Swimming speeds of fish. FA0 Conf. Fish Behau. Relation Fish. Tech. Tactics, Bergen, Norway Rev. Pap. No. 3, pp. 1-32. Blaxter, J. H. S., and Dickson, W. (1959). Observations on swimming speeds of fish.J. Cons., Cons. Perm. Int. Explor. Mer 24, 472-479. Blaxter, J. H. S., and Parrish, B. B. (1966).The reactions of marine fish to moving netting and other devices in tanks. Mar. Res. 1, 1-15. Blaika, P. (1958).The anaerobic metabolism of fish. Physiol 2001.31, 117-128. Blaika, P., Volf, M., and Cepela, M. (1960).A new type of respironieter of the determination of the metabolism of fish in an active state. Physiol. Bohemosloo. 9, 553-558. Bohun, S., and Winn, H. W. (1966). Locomotor activity of the american eel (Anguilla rostrata).Chesapeake Sci. 7, 137-147. Boyar, H. C. (1961).Swimming speed of immature Atlantic herring with reference to the Passamaquoddy Tidal Project. Trans. A m . Fish. SOC. 90, 21-26. Brawn, V. M. (1960). Underwater television observations of the swimming speed and behaviour of captive herring.J. Fish. Res. Board Can. 17, 689-698. Brett, J. R. (1962). Some considerations in the study of respiratory metabolism in fish, particularly sa1mon.J. Fish. Res. Board Can. 19, 1025-1038. Brett, J. R. (1964). The respiratory metabolism and swimming performance of young sockeye sa1mon.J. Fish. Res. Board Can. 21, 1183-1226. Brett, J. R. (1965). The relation of size to rate of oxygen consumption and sustained swimming speed of sockeye salmon (Oncorhynchusnerka).J.Fish. Res. Board Can. 23, 1491-1501. Brett, J. R. (1967). Swimming performance of sockeye salmon (Oncorhychus nerka) in relation to fatigue time and te1nperature.J. Fish. Res. Board Can. 24, 1731-1741. Brett, J. R. (1970a). Fish-The energy cost of living. In “Marine Agriculture” (W. J. McNeil, ed.), pp. 37-52. Oregon State Univ. Press, Corvallis. Brett, J. R. (1970b). 3. Temperature, 3.3 Animals, 3.32 Fishes. In “Marine Ecology, Vol. 1, Environmental Factors” (0. Kinne, ed.), Part 1, pp. 513-560. Wiley (Interscience), New York. Brett, J. R. (1972).The metabolic demand for oxygen in fish, particularly salmonids and a comparison with other vertebrates. Respir. Physiol. 14, 151-170. Brett, J. R. (1973).Energy expenditure of sockeye salmon, Oncorhynchus nerka, during sustained performance. J. Fish. Res. Board Can. 30, 1799-1809. Brett, J. R., and Glass, N. R. (1973). Metabolic rates and critical swimming speeds of sockeye salmon, Oncorhynchus nerka, in relation to size and temperature. J. Fish. Res. Board Can. 30,379-387. Brett, J. R., and Sutherland, D. B. (1965). Respiratory metabolism of pumpkinseed (Lepomis gibbosus) in relation to swimming speed. J. Fish. Res. Board Can. 22, 405-409. Brett, J . R., Hollands, H., and Alderdice, D. F. (1958).The effect of temperature on the cruising speed of young sockeye and coho salmon. J. Fish. Res. Board Can. 15, 587-605. Brody, S. (1945).“Bioenergetics and Growth,” 1023 pp. Reinhold, New York. Brunel, P. (1964).Food as a factor or indicator of vertical migrations of cod in the western

176

F. W. H. BEAMISH

Gulf of St. Lawrence. lCNAF Enuiron. Symp., Rome, 1964 Contrib. No. C-2, pp. 1-16. Bullock, T. H. (1955). Compensation for temperature in the metabolism and activity of poikilotherms. Biol. Reo. Cambridge Philos. SOC. 30, 311-342. Butler, J. A., and Milleman, R. E. (1971).Effect of the “salmon poisoning” trematode, Nanophyetus salmincola, on the swimming ability of juvenile salmonid fishes. J . Parasitol. 57, 860-865. Byme, J. M., Beamish, F. W. H., and Saunders, R. L. (1972). Influence of salinity, temperature, and exercise on plasma osmolality and ionic concentration in Atlantic salmon (Salmo salar).J. Fish. Res. Board Can. 29, 1217-1220. Caillouet, C. W., Jr. (1967). Hyperactivity, blood lactic acid, and mortality in channel catfish. lowa State Uniu. Sci. Tech. Res. Bull. No. 551, pp. 897-915. Calder, W. A., 111 (1974). Energy cost of animal locomotion. Science 184, 1098. Corti, U. A., and Weber, M. (1948). Die Matrix der Fische. 11. Intersuchungen iiber die Vitalitat von Fischen. Schweiz. Z. Hydrol. 11, 297-300. Craig, R. E., and Forbes, S. (1969). A sonar for fish counting. Fiskeridir. Skr., Ser. Hauunders. 15, 210-219. Cummings, W. C. (1963).Using the Doppler effect to detect movements of captive fish in behavior studies. Trans. Am. Fish. SOC. 92, 178-180. Cushing, D. H. (1957).The number of pilchards in the channel. Fish. Inuest. Minist. Agric. Fish. Food (G.B.)Ser. 11 Salmon Freshwater Fish. 21, 1-27. Dahl, K., (1937). Salmon migrations off Noway. Salmon Trout Mag. 88,229-234. Dahl, K., and Sgimme, S. (1938). Salmon markings in Norway 1937. Skr. Nor. Vidensk.Akad. Oslo 1 2, 1-45. Dahlberg, M. L., Shumway, D. L., and Doudoroff, P. (1968). Influence of dissolved oxygen and carbon dioxide on swimming performance of largemouth bass and coho sa1mon.J. Fish. Res. Board Can. 25,49-70. Dando, P. R. (1969).Lactate metabolism in fish. J . Mar. Biol. Assoc. U . K . 49, 209-223. Dannevig, G. (1953). Tagging experiments on cod, Lofoten 1947-1952: Some preliminary resu1ts.J. Cons., Cons. Perm. lnt. Explor. Mer 19, 195-203. Davidson, V. M. (1949).Salmon and eel movement in constant circular current.]. Fish. Res. Board Can. 7 , 4 3 2 4 4 8 , Davis, G. E., Foster, J., Warren, C. E., and Doudoroff, P. (1963). The influence of oxygen concentration on the swimming performance of juvenile Pacific salmon at various temperatures. Trans. Am. Fish. SOC. 92, 111-124. Dean, J. M., and Goodnight, C. J. (1964). A comparative study of carbohydrate metabolism in fish as affected by temperature and exercise. Physiol. Zool. 37,280299. DeGroot, S. J., and Schuyf, A. (1967).A new method for recording the swimming activity in flatfishes. Experientia 23,574-576. Denil, G. (1909). Les bchelles a poissons et leur application aux barrages des Meuse e t d’Ourthe. Ann. Trau. Publ. Belg. 2, 1-152. Denil, G. (1937). La m6canique du Poisson de rivi6re. Chapitre X. Les capacitbs mhcaniques d e la truite et du saumon. Ann. Trau. Publ. Belg. 38, 412-423. deVeen, J. F. (1964).On the phenomenon of soles swimming near the surface of the sea. Rapp. P.-V. Reun., Cons. Explor. Mer No. 155, p. 51. Dizon, A. E. (1977). Effect of dissolved oxygen concentration and salinity on swimming speed of two species of tunas. Fish. Bull. 75,649-653. Dodson, J. J., Leggett, W. C., and Jones, R. A. (1970). A study of the orientation and migration of American shad in Long Island Sound and the Connecticut River Estuary. U S . Bur. Commer. Fish. Prog. Rep. Contract AFC 6-1, 16 pp.

2. SWIMMING CAPACITY

177

Dodson, J. J., Lekgett, W. C., and Jones, R. A. (1971). A study of the orientation and migration of American Shad in Long Island Sound and the Connecticut River Estuary. U.S. Bur. Commer. Fish. Prog. Rep. Contract AFC &2, pp. 1-26. Dodson, J. J.. Leggett, W. C., and Jones, R. A. (1972).The behaviour of adult American shad (Alosa sapidissima) during migration from salt to freshwater as observed by ultrasonic tracking techniques. J. Fish. Res. Bourd Can. 29, 1445-1449. Doudoroff, P., and Shumway, D. L. (1970). Dissolved oxygen requirements of freshwater fishes. FA0 Fish. Tech. Pap. No. 86, 291 pp. Dow, R. L. (1962).Swimming speed of River herringPomo1obu.s pseudoharengus (Wilson). J . Cons., Cons. Perm. Int. Explor. Mer 27, 77-80. Dowd, R. G. (1967). An echo counting system for demersal fishes. FA0 ConJ Fish Behao. Relation Fish. Tech. Tactics, Bergen, Norway Exp. Pap. No. 7, pp. 1-6. Dowd, R. G., Bakken, E., and Nakken, 0. (1970). A comparison between two sonic measuring systems for demersal fish.J. Fish. Res. Board Can. 27, 737-747. Driedzic, W. R., and Hochachka, P. W. (1975). The unanswered question of high anaerobic capabilities of carp white muscle. Can. J . 2001.53, 7Of3-712. Drummond G. I. (1967). Muscle metabolism. Fortschr. 2001. 18, 360-429. Drummond, G. I. (1971). Microenvironment and enzyme function: Control of energy metabolism during muscle work. Am. Zool. 11, 83-97. Drummond, G. I., and Black, E! C. (1960). Comparative physiology: Fuel of muscle metabolism. Annu. Rev. Physiol. 22, 169-190. Ellis, D. V. (1966). Swimming speeds of sockeye and coho salmon on spawning migrati0n.J. Fish. Res. Board Can. 23, 181-187. Farlinger, S., and Beamish, F. W. H. (1977). Effects of time and velocity increments on the critical swimming speed of Iargemouth bass. Trans. Am. Fish. Soc. 106, 4 3 6 439. Farmer, G. J., and Beamish, F. W. H. (1969).Oxygen consumption ofTilapia nilotica in relation to swimming speed and sa1inity.J. Fish. Res. Board Can. 26,2807-2821. Farm‘er, G . J., Beamish, F. W. H., and Robinson, G. A. (1975).Food consumption of the adult landlocked sea lamprey, Petromyzon marinus, L. Comp. Biochem. Physiol. A 50, 753-757. Fierstine, H. L., and Walters, V. (1968). Studies in locomotion and anatomy of scombroid fishes. Mem. South. Caiif. Acad. Sci. 6, 1-31. Fontaine, M. (1956). Appareil pour la mbsure de I’activitb motrice d’yn Poisson. In “Notice sur les Travaux Scientiques de Maurice Fontaine,” p. 81. F. Paillart, Abbeville. Fox, A. C. (1965). Some effects o€ strigeid metacercariae on rainbow trout (Salmo gairdneri).Trans. Am. Microsc. Soc. 84, 153. Fridriksson, A. (1958). The tribes in the north coast herring of Iceland with special reference to the period 1948-1955. Rapp. P.-V.Reun., Cons. Explor. Mer 143,3644. Fridriksson, A., and Aasen, 0. (1952).The Norwegian-Icelandic herring tagging experiments. Report No. 2. Rit Fiskideildar 1, 1-54. Fry, F. E. J. (1947).Effects of the environment on animal activity. Unio. Toronto Stud., Biol. Ser. 55, 1-62. Fry, F. E. J. (1957).The lethal temperature as a tool in taxonomy.Ann. Biol. 33,205-218. Fry, F. E. J. (1958).Temperature compensation. Annu. Reu. Physiol. 20, 207-224. Fry, F. E. J. (1967). Responses of vertebrate poikilotherms to temperature. In “Thermobiology” (A. H. Rose, ed.), pp. 375-409. Academic Press, New York. Fry, F. E. J. (1971). The effects of environmental factors on the physiology of fish. In “Fish Physiology” (W. S. Hoar and D. J. Randall, eds.), Vol. 6, pp. 1-98. Academic Press, New York.

178

F. W. H. BEAMISH

Fry, F. E. J., and Cox, E. T. (1970).A relation of size to swimming speed in rainbow trout.]. Fish. Res. Board Can. 27,976978. Fry, F. E. J., and Hart, J. S. (1948). Cruising speed of goldfish in relation to water temperature.]. Fish. Res. Board Can. 7, 169-175. Fry, F. E. J,, and Hochachka, P. W. (1970). Fish. In “Comparative Physiology of Thermoregulation, Vol. 1, Invertebrates and Nonmallian Vertebrates” (G. C. Whittow, ed.), pp. 79-134. Academic Press, New York. Gardner, G. R., and Laroche, G. (1973). Copper-induced lesions in estuarine teleosts.]. Fish. Res. Board Can. 30, 363-368. Gayduk, V. V., Malinin, L. K., and Poddubnyy, A. G. (1971). Determination of the depths of fishes during the hours of day1ight.J. Zchthyol. 11, 140-143. Gero, D. R. (1952). The hydrodynamic aspects of fish propulsion. Am. M U S .Nooit. No. 1601, pp. 1-32. Gibson, E. S., and Fry, F. E. J. (1954). The performance of the lake trout, Saloelinus namaycush, at various levels of temperature and oxygen pressure. Can.J.Zool. 132,

252-260. Glova, G. J., and McInerney, J. E. (1977). Critical swimming speeds of coho salmon (Oncorhynchus kisutch) fry to smolt stages in relation to salinity and temperature.J. Fish. Res. Board Can. 34, 151-154. Gold, A. (1973). Energy expenditure in animal locomotion. Science 181,275-276. Gold, A. (1974). Comment on paper by Calder. Science 184, 1098. Gordon, M. S. (1963). Chloride exchange in rainbow trout (Salmo gairdneri) adapted to different salinities. Biol. Bull. 124, 45-54. Graham, J. B. (1973). Heat exchange in the black skipjack, and the blood-gas relationship of warm-bodied fishes. Proc. Nat. Acad. Sci. U.S.A. 70, 1964-1967. Graham, J. B. (1975). Heat exchanges in the yellow fin tuna, Thunnus albacares, and skipjack tuna, Katsuwonus pelamis, and the adaptive significance of elevated body temperatures in scombrid fishes. U.S. Fish. Wildl. Seru., Fish. Bull. 73,

219-229. Graham, J. M. (1949). Some effects of temperature and oxygen pressure in the metabolism and activity of the speckled trout, Saloelinus fontinalis. Can. 1.Res., Sect. D. 27, 270-288. Gray,J. (1937).Pseudorheotropism in fishes.J. E x p . Biol. 14, 95-103. Gray, J. (1953).The locomotion of fishes. Essays Mar. Biol. pp. 1-16. Gray, J. (1957). How fishes swim. Sci. Am. 192,48-54. Green, D. M. (1964).A comparison of stamina of brook trout from wild and domestic parents. Trans. Am. Fish. SOC.93, 9 6 1 0 0 . Greene, C. W. (1926). The physiology of the spawning migration. Physiol. Reu. 6, 201-

241. Greer Walker, M. (1970). Growth and development of the skeletal muscle fibres of the cod (Gadus morhua L.) J. Cons., Cons. Perm. Int. E x p l o r . Mer 33, 228-

244. Greer Walker, M. (1971).Effect of starvation and exercise on the skeletal muscle fibres of the cod (Gadus morhua L.) and the coal fish (Gadus oirens L.) respectively. J. Cons., Cons. Perm. Znt. E x p l o r . Mer 33, 421-427. Greer Walker, M., and Pull, G. (1973). Skeletal muscle function and sustained swimming speeds in the coal fish (Gadus uirens L.). C o m p . Biochem. Physiol. A 44,

495-501. Greer Walker, M., Mitson, R. B., and Storeton-West, T. (1971).Trials with a transponding acoustic tag tracked with an electronic sector scanning sonar. Nature (London)229,

196198. Griffiths, J. S., and Alderdice, D. F. (1972). Effects of acclimation and acute temperature

2.

SWIMMING CAPACITY

179

experience on the swimming speed of juvenile coho salmon. J. Fish. Res. Board Can. 29,251-264. Hammond, B. R., and Hickman, C. P., Jr. (1966).The effect of physical conditioning on the metabolism of lactate, phosphate, and glucose in rainbow trout, Salmo gairdneri. J . Fish. Res. Board Can. 23,65-83. Harden-Jones, F. R. H. (1962). Further observations on the movements of herring (Clupea harengus L.) shoals in relation to the tidal current. J. Cons., Cons. Perm. Znt. Explor. Mer 27, 52-76. Harden-Jones, F. R. H. (1968). “Fish. Migration,” 325 pp. Arnold, London. Harden-Jones, F. R. H., and McCartney, B. S. (1962). The use of electronic sectorscanning sonar for following the movements of fish shoals: Sea trials on R.R.S. Discovery 11. J. Cons., Cons. Perm. Fnt. Explor. Mer 27, 141-149. Hasler, A. D., Horrall, R. M., Wisby, W. J., and Braemer, W. (1958). Sun-orientation and homing in fishes. Limnol. Oceanogr. 3,353-361. Hasler, A. D., Cardella, E. S., Horrall, R. M., and Henderson, H. F. (1969).Open-water orientation of white bass, Roccus chrysops, as determined by ultrasonic tracking methods. J. Fish. Res. Board Can. 26, 2173-2192. Hawkins, A. D., MacLennan, D. N., Urquhart, G. G., and Robb, C. (1974). Tracking cod Gadus morhus L. in a Scottish sea. Loch J . Fish Biol. 6, 225-236. Heath, A. G., and Pritchard, A. W. (1962).Changes in the metabolic rate and blood lactic acid of bluegill sunfish, Lepomis macrochirus, Raf., following severe muscular activity. Physiol. 2001.35, 323-329. Henderson, H. F., Hasler, A. D., and Chipman, G. C. (1966).An ultrasonic transmitter for use in studies of movements of fishes. Trans. Am. Fish. SOC. 95, 350-356. Heusner, A. A., and Enright, J. T. (1966). Long-term activity recording in small aquatic animals. Science 154, 532-538. High, W. L., and Lusz, L. D. (1966). Underwater observations on fish in an off-bottom trawl. J . Fish. Res. Board Can. 23, 153-154. Hill, A. V. (1950). The dimensions of animals and their muscle dynamics. Sci. Prog. (London) 38,209-230. Hjort, J. (1914). Fluctuations in the great fisheries of Northern Europe. Rapp. P.-V. Reun., Cons. Explor. Mer 20, 1-228. Hochachka, P. W. (1961).The effect of physical training on oxygen debt and glycogen reserves in trout. Can J. Zqol. 39, 767-776. Hocutt, C. H. ( 1973). Swimming performance of three warmwater fishes exposed to a rapid temperature change. Chesapeake Sci. 14, 11-16. Houde, E. D. (1969). Sustained swimming ability of larvae of walleye (Stizostedion oitreum oitreum) and yellow perch (Percajaoescens).J. Fish. Res. Board Can. 26, 1647- 1659. Howard, T. E. (1975). Swimming performance of juvenile coho salmon (Oncorhynchus kisutch) exposed to bleached kraft pulpmill effluent. J. Fish. Res. Board Can. 32, 789-793. Hunter, J. R. (1971). Sustained speed of jack mackerel Trachurus symmetricus. U.S. Fish. Wildl. Sew., Fish. Bull. 69,267-271. Hunter, J. R., and Zweifel, J. R. (1971). Swimming speed, tail beat frequency, tail beat amplitude, and size in jack mackerel, Trachurus symmetricus, and other fishes. U.S. Fish Wildl. Serv., Fish. Bull. 69, 253-266. Huntsman, A. G. (1942). Return of a marked salmon from a distant place. Science 95, 381-382. Ichihara, T., Soma, M., Yoshida, K., and Suzuki, C. (1972). An ultrasonic device in biotelemetry and its application to tracking a yellowtail. Bull. Far Seas Fish. Res. Lab. 7, 27-49.

180

F. W. H. BEAMISH

Idler, D. R., and Tsuyuki, H. (1958). Biochemical studies on sockeye salmon during spawning migration. I. Physical measurement, plasma, cholesterol, and electrolyte levels. Can. ]. Biochem. Physiol. 36, 783-791. Javaid, M. Y. (1973). Effect of DDT on temperature selection in some salmonids. Pak. J . Sci. Znd. Res. 15, 171-176. Job, S. V. (1955). The oxygen consumption of Saloelinus fontinalis. Unio. Toronto Studies Biol. Ser. 61, pp. 1-39. Johnson, J. (1960). Sonic tracking of adult salmon at Bonneville Dam, 1957. U.S. Fish. Wildl. Sera, Fish. Bull. 176, 471-485. Johnston, I. A., and Goldspink, G . (1973). A study of glycogen and lactate in the myotomal muscles and liver of the coalfish (Cadus oirens L.) during sustained swimming.]. Mar. Biol. Assoc. U.K. 53, 17-26. Jones, D. R. (1971).The effect of hypoxia and anaemia on the swimming performance of rainbow trout (Salmo gairdneri).]. E x p . Biol. 55, 541-551. Jones, D. R., Kiceniuk, J. W., and Bamford, 0. S. (1974). Evaluation of the swimming performance of several fish species from Mackenzie River.]. Fish. Res. Board Can. 31, 1641-1647. Jones, F. R. H. (1957). Movements of herring shoals in relation to the tidal currents.]. Cons., Cons. Perm. Znt. Explor. Mer 22, 322-328. Junge, C. O., Jr., and Phinney, L. A. (1963). Factors influencing the return of fall chinook salmon (Oncorhynchus tshawytscha) to Spring Creek Hatchery. U.S. Fish Wildl. Ser., Spec. Sci. Rep.-Fish. No. 445, pp. 1-32. Katz, M., Pritchard, A., and Warren, C. E. (1959). Ability of some salmonids and a centrarchid to swim in water of reduced oxygen content. Trans. Am. Fish. SOC. 88, 88-95. Kerr, J. E. (1953). Studies of fish preservation at the Contra Costa steam plant of the Pacific Gas and Electric Company. Calif. Dep. Fish Came Fish, Bull. 92, 1-66. Kishinouye, K. (1923). Contributions to the comparative study of the so-called scombroid fishes. J . Coll. Agric., Imp. Univ. Tokyo 8, 293-475. Kleerekoper, H. A., Timms, M., Westlake, G. F., Davy, F. B., Malar, T., and Anderson, V. M. (1969). Inertial guidance system in the orientation of the goldfish (Carassius auratus). Nature (London) 223, 501-502. Kleerekoper, H. A., Timms, M., Westlake, G. F., Davy, F. B., Malar, T., and Anderson, V. M. (1970). An analysis of locomotor behaviour of goldfish (Carassius auratus). Anim. Behao. 18, 317-330. Klein, W. D., Olsen, 0. W., and Bowden, D. C. (1969). Effects of intestinal fluke, Crepidostomum farionis, on rainbow trout, Salmo gairdneri. Trans. Am. Fish. SOC. 98, 1-6. Komarov, V. T. (1971). Speeds of fish movement. Zool. Herald 4,67-71. [Transl. by Fish. Res. Board Can., Transl. Ser. No. 2030 (1972).] Konstantinov, K. G. (1958). Diurnal vertical migrations of the cod and haddock. Tr. VNlRO 36,62-82. Krueger, H. M., Liu, S . D., Chapman, G . A., and Chang, J. T. (1966). Effects of pentachlorophenol on the fish Cichlasoma bimaculatum. lnt. Pharmacol. Congr., 3rd, Sao Paulo, Abstr. No. 649. Krueger, H. M., Sadler, J. B., Chapman, G. A., Tinsley, I. J., and Lowry, R. R., (1968). Bioenergetics, exercise, and fatty acids of fish. Am. Zool. 8, 119-129. Kutty, M. N. (1968). Influence of ambient oxygen on the swimming performance of goldfish and rainbow trout. Can J . 2001.46, 647-653. Kutty, M. N. (1969). Oxygen consumption in the mullet (Liza macrolepis) with special reference to swimming velocity. Mar. Biol. 4, 239-242.

2.

SWIMMING CAPACITY

181

Kutty, M. N., and Peer Mohamed, M. (1975). Metabolic adaptations of mullet Rhinomugil corsula (Hamilton) with special reference to energy utilization. Aquaculture 5, 253-270. Kutty, M. N., and Saunders, R. L. (1973). Swimming performance of young Atlantic salmon (Salmo salar) as affected by reduced ambient oxygen concentration.J. Fish. Res. Board Can. 30,223-227. Lane, F. W. (1941). How fast do fish swim? Country Lije, London 534-535. Langford, T. E. (1974). Trials with ultrasonic tags for the study of coarse fish behaviour and movements around power stations outfalls. J . Znst. Fish. Mgrnt. 5, 61-62. Larimore, R. W., and Duever, M. J. (1968). Effects of temperature acclimation on the swimming ability of smallmouth bass fry. Trans. A m . Fish. SOC.97, 175-184. Leggett, W. C., and Jones, R. A. (1973).A study of the rate and pattern of shad migration in the Connecticut River-utilizing sonic tracking apparatus. U.S. Natl. Mar. Fish. Seru., Rep. AFG 5, pp. 1-118. Lemke, A. E., and Mount, D. I. (1963). Some effects of alkyl benzene sulfonate on the bluegill, Lepomis machrochirus. Trans. A m . Fish. SOC.92, 372-378. Lett, P. F., Farmer, G. J., and Beamish, F. W. H. (1976).The effect of copper on some aspects of the bioenergetics of rainbow trout (Salmo gairdneri). J . Fish. Res. E d . Can. 33, 1335-1342. Lilleliind, K. (1967). Versuche zur Erbriitung der Eier vom Hecht, Esox lucius L., in Abhangigkeit von Temperatur und Licht. Arch. Fischereiwiss. 17, 95-120. Linthicum, D. S., and Carey, F. G. (1972).Regulation of brain and eye temperatures by the bluefin tuna. C o m p . Biochem. Physiol. A 43, 425-433. Luhmann, M., and Mann, H. (1958). Wiederfange markierter Elbaale vor der Kuste Dinemarks. Arch. Fischereiwiss. 9, 200-202. McCleave, J. D., and Honall, R. M. (1970).Ultrasonic tracking ofhoming cutthroat trout (Salrno clarki) in Yellowstone Lake. J . Fish. Res. Board Can. 27, 715-730. McCleave, J. D, and LaBar, G. W. (1972). Further ultrasonic tracking and tagging studies of homing cutthroat trout (Salmo clarki) in Yellowstone Lake. Trans. Am. Fish. SOC. 1,44-54. McCleave, J. D., Jahn, L. A., and Brown, C. J. D. (1967). Miniature alligator clips as fish tags. Prog. Fish-Cult. 29, 60-61. McKim, J. M., Christensen, G. M., and Hunt, E. P. (1970).Changes in the blood of brook trout Saloelinus fontinalis after short-term and long-term exposure to copper. J. Fish. Res. Board Can. 27, 1883-1889. MacLeod, J. C. (1967). A new apparatus for measuring maximum swimming speeds of small fish.]. Fish. Res. Board Can. 24, 1241-1252. MacLeod, J. C., and Smith, L. L., Jr. (1966). Effect of pulpwood fiber on oxygen consumption and swimming endurance of the fathead minnow, Pimephales promelas. Trans. A m . Fish. SOC. 95, 71-84. Madison, D. M., Horrall, R. M., Stasko, A. B., and Hasler, A. D. (1972). Migrating movements of adult sockeye salmon (Oncorhynchus nerka) in coastal BritishaColumbia as revealed by ultrasonic tracking.J. Fish. Res. Board Can. 29, 1025-1033. Miir, A. (1947). Uber die Aalwanderung im Baltischen Meer auf Grund der Meerbusen Wanderaalmarkierungsversuche im Finnischen und Livischen Meerbusen in den Jahren 1937-1939. Meed. Statens Unders.- 0 . Foers. Anst. Soetoatt Fiskeridir. 27, 1-56. Magnan, A. ( 1929). Les caracteristiques geometriques et physiques des poissons. Ann. Sci. Nat., Zool. 13, 355-489. Magnuson, J. J . (1967). Swimming activity of the scombroid fish Euthynnus afinis as related to search for food. FA0 Conf. Fish Behao. Relation Fish. Tech. Tactics, Bergen, Norway Exp. Pap. No. 3, pp. 1-13.

182

F. W. H. BEAMISH

Magnuson, J. J. ( 1970). Hydrostatic equilibrium of Euthynnus afinis, a pelagic teleost without a gas bladder. Copeia pp. 56-85. Magnuson, J. J. (1973).Comparative study of adaptatioiis for continuous swimming and hydrostatic equilibrium of scombroid and xiphoid fishes. U.S. Fish Wildl. Seru., Fish. Bull. 71, 337-356. Magnuson, J. J., and Prescott, J. H. (1966).Courtship, locomotion, feeding and miscellaneous behaviour of Pacific bonito (Sarda chilieisis).Anim. Behau. 14, 54-67. Malinin, L. K. (1971). Behaviour of burbot. Piroda No. 8,77-79. (Transl. Ser. No. 2171. J. Fish. Res. Board Con., 1972, 8 pp.) Malinin, L. K. (1973). Speed of fish migration. Fish. Znd. 8, 1 6 1 7 . (Transl. Ser. No. 3146. Fish. hlar. Serv. Can., 1974.) Malinin, L. K., Poddubny, A. G., and Gaiduk, V. V. (1971). Stereotypes of the Volga sturgeon behaviour in the region of the Saratov hydro-electric power station before and after the river regulation. Zoologicheskii Zhurnal 50,847-857. (Transl. Ser. No. 2155.3. Fish. Res. Board Con., 1972, 27 pp.) Mar, J . (1959).A proposed tunnel design for a fish respirometer. Tech. Memo. 5 4 3 . Pac. Nav. Lab., D.R.B., Esquimalt, B.C. Marty, J. J. (1959). The fundamental stages of the life cycle of Atlantic-Scandinavian herring. U.S. Fish Wildl. Serv., Spec. Sci. Rep.-Fish. No. 327, pp. 5-6a. Maslov, N. A. (1960).Soviet investigations on the biology of the cod and other demersal fish in the Barents Sea.Zn “Soviet Fishery Investigations in North European Seas,” pp. 185-231. VNIRO/PINRO, Moscow. Mathur, G. B., and Shrivastava, B. D. (1970). An improved activity meter for the determination of standard metabolism in fish. Trans. A m . Fish. S O C . 99, 602-603. Meffert, D. (1968). Ultrasonic recorder for locomotor activities. Trans.Am. Fish. S O C . 97, 12-17. Midttun, L., and Nakken, 0. (1968). Counting of fish with an echo-integrator. Znt. Counc. Explor. Sea Coop. Res, Rep. Ser. B 17, 1-7. Miller, R. B. (1957). Permanence and size of home territory in stream-dwelling cutthroat tr0ut.J. Fish Res. Board Can. 14, 687-691. Milne, D. J., and Ball, E. A. R. (1956). The mortality of small salmon when caught by trolling and tagged or released untagged. Fish. Res. Board Can., Pac. Prog. Rep. No. 106, pp. 10-13. Milne, D. J., and Ball, E. A. R. (1958).The tagging of spring and coho salmon in the Strait ofGeorgia in 1956.Fish.Res. Board Can., Pac. Prog. Rep. No. I l 1 , p p . 14-18. MolnBr, G., and Tolg, I. (1962). Relation between water temperature and gastric digestion of largemouth bass (Micropterussalmoides LacBpBde).J . Fish. Res. Board Can. 19,1005-1012. Muir, B. S., Nelson, G. J., and Bridges, K. W. (1965). A method for measuring swimming speed in oxygen consumption studies on the aholehole, Kuhlia sanduicensis. Trans. Am. Fish. SOC. 94,378-382. Nakatani, R. E. (1957).Changes in the inorganic phosphate and lactate levels in blood plasma and muscle tissue of adult steelhead trout after strenuous swimming. Uniu. Wash., Sch. Fish. Tech. Rep. 30, pp. 1-14. Niimi, A. J., and Beamish, F. W. H. (1974).Bioenergetics and growth of largemouth bass (Micropterus salmoides) in relation to body weight and temperature. Can. J. Zool. 52, 447-456. Nishimura, M. (1963).Investigation of tuna behaviour by fish finder. FA0 Fish. Rep. 3, 1113-1123. Novotny, A. J., and Esterberg, G. F. (1962). A 132-kilocycle sonic fish tag. Prog. FishCult. 24, 139-141.

2.

SWIMMING CAPACITY

183

Ohlmer, W., and Schwartzkopff, J. (1959). Schwimmgeschwindigkeiten von Fischen aus stehenden Binnengewassern. Naturwissenschaften 10, 362-363. O h , B. L., Bejda, A. J., and Martin, A. D. (1974). Daily activity, movements, feeding and seasonal occurrence in the tautog, Tautoga onitis. Fish. Bull. 72, 27-35. Oseid, D., and Smith, L. L., Jr. (1972). Swimming endurance and resistance to copper and malathion of bluegills treated by long-term exposure to sublethal levels of hydrogen sulfide. Trans. Am. Fish. SOC. 101, 620-625. Otto, R. G., and Rice, J. O’H. (1974). Swimming speeds of yellow perch (Perca Flauescens) following an abrupt change in environmental temperature.]. Fish. Res. Board Can. 31, 1731-1734. Parker, R. R., and Black, E. C. (1959). Muscular fatigue and mortality in troll-caught chinook salmon (Oncorhynchus tshawytscha).].Fish. Res. Board Can. 16,95106. Parker, R. R., Black, E. C., and Larkin, P. A. (1959). Fatigue and mortality in troll-caught Pacific salmon (Oncorhynchus).]. Fish. Res. Board Can. 16,429-448. Parrish, B. B., Blaxter, J. H. S., and Hall, W. B. (1964). Diurnal variations in size and composition of trawl catches. Rapp. P.-V. Reun., Cons. Int. Explor. Me7 155,27-34. Parry, G. (1958). Size and osmoregulation in salmonid fishes. Nature (London) 181, 1218-1219. Paulik, G. J., and DeLacy, A. C. (1957). Swimming abilities of upstream migrant silver salmon, sockeye salmon and steelhead at several water velocities. Uniu. Wash., Sch. Fish., Tech. Rep. No. 44, pp. 1-40. Paulik, G. J., and DeLacy, A. C. (1958).Changes in the swimming ability of Columbia River sockeye salmon during upstream migration. Uniu. Wash., Sch. Fish., Tech. Rep. No. 46, pp. 1-67. Paulik, G. J., DeLacy, A. C., and Stacy, E. F. (1957). The effect of rest on the swimming performance of fatigued adult silver salmon. Unio. Wash., Sch. Fish., Tech. Rep. No. 31, pp. 1-24. Pavlov, D. S., Sbikin, Y. N., and Mochek, A. D. (1968). The effect of illumination in running water on the speed of fishes in relation to features of their orientation. Vopr. Ikhtiol. 8, 318-324. (Transl. in Probl. Ichthyol. Am. Fish. SOC.,Washington, D.C. Vol. 8, pp. 250-255, 1968.) Peterson, R. H. (1974). Influence of fenitrothion on swimming velocity of brook trout (Saluelinus fontinalis).J . Fish. Res. Board Can. 31, 1757-1762. Poddubny, A. G. (1967). Sonic tags and floats as a means of studying fish response to natural environmental changes and to fishing gear. FA0 Conf. Fish Behau. Relation Fish. Tech. Tactics, Bergen, Norway Exp. Pap. No. 46, pp. 1-8. Poddubny, A. G., Malinin, L. K., and Gaiduk, V. V. (1970;. Experiment in telemetric observations under ice of the behaviour of winter fish. Biologiya Unutrehnikh. Vod. 6, 65-70. (Transl. Ser. No. 1817.1. Fish. Res. Board Can., 1971.) Pope, A., and Harper, J. J. (1966).“Low-Speed Wind Tunnel Testing.” Wiley, New York. Potts, W. T. W. (1954).The energetics of osmotic regulation in brackish and freshwater animals.]. E x p . Biol. 31, 618-630. Pritchard, A. L. (1944). Return of two marked pink salmon (Oncorhynchus gorbuscha) to the natal stream from distant places in the sea. Copeia pp. 80-82. Pritchard, A. W., Hunter, J. R., and Lasker, R. (1971). The relation between exercise and biochemical changes in red and white muscle and liver in the jack mackerel, Trachurus symmetricus. Fish. Bull. 69, 379-386. Prosser, C. L., Barr, L. M., Pinc, R. D., and Lauer, C. Y. (1957).Acclimation of goldfish to low concentrations of oxygen. Physiol. Zool. 30, 137-141. Radakov, D. V. (1964). Velocities of fish swimming (in Russian). Pamphlet A.N. Seoertsou Inst. Anim. Morphol., Moscow pp. 4-28.

184

F. W. H. BEAMISH

Radakov, D. V., and Solovyev, B. S . (1959). First attempts to employ a submarine for observing the behaviour of herring. Rybn. Khoz. (Moscow) 35, 16-21. (Fish. Res. Board Can., Transl. Ser. No. 338.) Randall, D. J. (1970). Gas exhange in fish. Zn “Fish Physiology” (W. S. Hoar and D. J. Randall, eds.), Vol. 4, pp. 253-291. Academic Press. New York. Rao, C. M. M . (1968). Oxygen consumption gf rainbow trout (Salmo gairdneri) in relation to activity and salinity. Can. 1.Zool. 46, 781-786. Rao, C . M. M. (1971). Influence of activity and salinity on the weight-dependent oxygen consumption of the rainbow trout Salmo gairdneri. Mar. Biol. 8, 205-212. Rasmussen, B. (1959). On the migration pattern ofthe West Greenland stock of cod. Ann. Bfol. 4, 123-124. Regnard, M. P. (1893).Sur un dispositif qui permet d e mesurer la vitesse d e translation d’un Poisson se mouvant dans I’eau. C.R. SOC. Biol. 5, 81-83. Reimers, N . (1956).Trout stamina. Prog. Fish-Cult. 18, 112. Richardson, I . D., Cuihing, D. H., Harden-Jones, F. R., Beverton, R. J. H., and Blacker, R. W. (1959). Echo sounding experiment in the Barents Sea. Fish. Inuest. Minist. Agric., Fish. Food (G.B.)Ser. 11 Salmon Freshwater Fish. 22(9), 7-16. Ricker, W. E. (1975).Computation and interpretation ofbiological statistics of fish populations. Bull., Fish. Res. Board Can. 191, 1-382. Roberts, J. L. (1966). Systematic versus cellular acclimation to temperature by poikilotherms. Helgol. Wiss. Meeresunters. 14, 451-465. Root, R. W. (1931).The respiratory function of the blood of marine fishes. Biol. Bull. (Woods Hole, Mass.) 61, 427-456. Royce, W. F., Smith, L. S., and Hartt, A. C. (1968). Models of oceanic migrations’of Pacific salmon and comments on guidance mechanisms. U.S.Fish Wildl. Serv., Fish. Bull. 66, 441-462. Ruhland, M. L., and Heusner, A. (1959). Chambre respiratoire pour la determination simultane d e I’activitk et de la comsommation d’oxygene par une methode manom6trique chez des poissons de 3-log. C.R. SOC. Biol. 153, 161-164. Ryland, J. S. (1963). The swimming speeds of plaice larvae.]. E x p . Biol. 40, 285-299. Saetersdal, G. (1956). Fisheries research in northern waters-study items and results. Dep. Rep. No. 1. Deep Sea Research Inst., Fish. Directorate, Bergen. (Transl. by Cent. Off. Inf., London.) Saunders, R. L., and Sutterlin, A. M. (1971). Cardiac and respiratory responses to hypoxia in the sea raven, Hemitripterus americanus, and an investigation of possible control mechanisms. J . Fish. Res. Board Can. 28, 491-503. Scharfe, J. (1960). A new method for aimed one-boat trawling in midwater and on the bottom. Stud. Reu. Gen. Fish. Counc. Mediterr. 13, 1-38. Schmidt-Nielsen, K. ( 1972).Locomotion: Energy cost of swimming, flying, and running. Science 177,222-228. Schuyf, A., and DeGroot, S. J. (1971).An indictive locomotion detector for use in diurnal activity experiments in fish.]. Cons., Cons. Perm. Int. Explor. Mer 34, 126-131. Shaw, E. (1960). The development of schooling behaviour in fishes. Physiol. 2001.33, 79-86. Shazkina, E. P. (1972a). Energy metabolism and food rations of steelhead trout under conditions of the Chernaya Rachka trout industry. Tr. USES Nauchno. Issled. Znst. Morsk. Rubn. Khoz. Okeanogr. 76, 13CL134. (Transl. for Ref. Zh. Biol., No. 61203, pp. 138-144, 1972.) Shazkina, E. P. (197213).Active metabolism in Azov goby. Proc. All-Union Res. Znst. Mar. Fish. Oceanogr. 85, 138-144. (Transl. Ser. No. 3040. Fish. Mar. Serv. Can., 1974.)

2.

SWIMMING CAPACITY

185

Shuck, H. A., and Kingsbury, 0. R. (1948). Survival and growth of fingerling brown trout (Salmofario) reared under different hatchery conditions and planted in fast and slow water. Trans. Am. Fish. SOC. 75, 147-156. Smit, H. (1965). Some experiments of the oxygen consumption of goldfish (Carassius auratus L.) in relation to swimming speed. Can. J. Zool. 43,623-633. Smit, H., Amelink-Koutstaal, J. M., Vijverberg, J., and von Vaupel-Klein, J. C. (1971). Oxygen consumption and efficiency of swimming goldfish. Comp. Biochem. Physiol. A 39, 1-28. Smith, H. D., and Margolis, L. (1970). Some effects ofEubothrium saluelini (Schrank, 1790) on sockeye salmon, Oncorhynchus nerka (Walbaum), in Babine Lake, British Columbia. J. Parasitol. 56, Sect. 11, 321-322. Smith, L. S., and Newcomb, T. W. (1970). A modified version of the Blazka respirameter and exercise chamber for large fish.J. Fish. Res. Board Can. 27, 1321-1324. Spoor, W. A. (1946). A quantitative study of the relationship between the activity and oxygen consumption of the goldfish, and its application to the measurement of respiratory metabolism in fishes. Biol. Bull. (Woods Hole, Mass.) 91, 312-325. Spoor, W. A., and Drummond, R. A. (1972). An electrode for detecting movement in gradient tanks. Trans, Am. Fish. SOC.101, 714-715. Spoor, W. A., Neiheisel, T.W., and Drummond, R. A. (1971). An electrode chamber for recording respiratory and other movements of free-swimming animals. Trans. Am. Fish. SOC. 100, 22-28. Sprague, J. B. (1971). Measurement of pollutant toxicity of fish. 111. Sublethal effects and “safe” concentrations. Water Res. 5, 245-266. Stasko, A. B. (t975). Progress of migrating Atlantic salmon (Salmo salar) along an estuary observed by ultrasonic tracking. J . Fish Biol. 7, 329-338. Stasko, A. B., and Rommel, S. A,, Jr. (1974). Swimming depth of adult American eels (Anguilla rostrata) in a saltwater bay as determined by ultrasonic tracking./. Fish. Aes. Board Can. 31, 1148-1150. Stasko, A. B., Horrall, R. M., Hasler, A. D., and Stasko, D. (1973). Coastal movements of mature Fraser River pink salmon (Oncorhynchus gorbuscha) as revealed by ultrasonic tracking.J. Fish. Res. Board Can. 30, 1309-1316. Stevens, E. D., and Fry, F. E. J. (-1971).Brain and muscle temperatures in ocean caught and captive skipjack tuna. Comp. Biochem. Physiol. A38, 203-211, Stevens, E. D., Lam, H. N., and Kendall, J. (1974). Vascular anatomy of the countercurrent heat exchanger in skipjack tuna. J. Exp. Biol. 61, 145-153. Stringham, E. (1924).The maximum speed of freshwater fishes. Am. Nut. 58,156-161. Sundnes, G. (1963).Swimming speed of fish as a factor in gear research. Fiskeridir. Skr., Ser. Havunders. 13, 126132. Tesch, F. W. (1974). Speed and direction of silver and yellow eels, Anguilla anguilla, released and tracked in the open North Sea. Ber. Dtsch. Wiss.Komm. Meeresforsch. 23, 181-197. Thomas, A. E., Burrows, R. E., and Chenoweth, H. H. (1964). A device for stamina measurement of fingerling salmonides. Bur. Sport Fish. WiZdZ. ( U . S . ) ,Res. Rep. No. 67, pp. 1-15. Thompson, D. W. (1917). “On Growth and Form,” 793 pp. Cambridge Univ. Press, London. Thorson, T. B. (1971). Movement of bull sharks, Carcharhinus leucas between Caribbean Sea and Lake Nicaragua demonstrated by tagging. Copeia No. 2, pp. 336338. Threadgold, L. T., and Houston, A. H. (1964). An electron microscope study of the “chloride cell” of Salmo salar L. E x p . Cell Res. 34, 1-23.

186

F. W. H. BEAMISH

Trefethen, P. S. (1956). Sonic equipment for tracking individual fish. U.S. Fish Wildl. Seru. Spec. Sci. Rep.-Fish. 179, 1-11. Trout, G . C. (1957). The Bear Island cod: Migrations and movements. Fish. Invest. Minist. Agric. Fish. Food (G.B.)Ser. I1 Salmon Freshwater Fish. 21, 1-51. Tucker, V. A. (1970).Energetic cost of locomotion in animals. Comp. Biochem. Physiol. A 34, 841-846. Tytler, P. (1969). Relationship between oxygen consumption and swimming speed in the haddock, Melanogrammus aeglefinus. Nature (London) 221,274-275. Verwey, J. (1960). Uber die Orienteir ung wandern der Meerestiere. Helgol. Wiss. Meeresunters. 7 , 51-58. Vibert, R. (1956). Methode pour I’btude et I’amelioration de la survie des alevins de repeuplement (triuites et saumons). Ann. Sta. Cent. Hydrobiol. Appl. 6, 347-439. Vincent, R. E. (1960).Some influences of domestication upon three stocks of brook trout (Saluelinusfontinalis Mitchill). Trans. Am. Fish. Soc. 89, 35-52. von Kausch, H. (1968). Der Einfluss der Spontanaktivitat auf die Stoffwechselrate junger Karpfen (Cyprinus carpio L.) im Hunger und bei Fiitterung. Arch. Hydrobiol., Suppl. 33,263-330. Wales, J. H. (1950). Swimming speed of the western sucker Catostomus occidentalis Ayres. Calif. Fish Game 36, 433-434. Walters, V., and Fierstine, H. L. (1964). Measurements of swimming speeds of yellow fin tuna and wahoo. Nature (London) 202,208-209. Wardle, C. S. (1975).Limit of fish swimming speed. Nature (London)255, 725-727. Warren, C. E., and Davis, G. E. (1967). Laboratory studies on the feeding, bioenergetics and growth of fish. In “The Biological Basis for Freshwater Fish Production” (S. Gerking, ed.), pp. 175-214. Blackwell, Oxford. Watanabe, N. (1942). Measurements on the bodily density, body temperature, and swimming velocity “katuwu”, Euthynnus vagans (Lesson). Nippon Suisan Gakkaishi 11, 146-148. Weaver, C. R. (1963). Influence of water velocity upon orientation and performance of adult migrating salmonids. U . S . Fish Wildl. Serv., Fish. Bull. 63,97-121. Weaver, C . R. (1965). Observations on the swimming ability of adult American shad (Alosa sapidissima). Trans. Am. Fish. SOC.94, 382-385. Webb, P. W. (1971a). The swimming energetics of trout. I. Thrust and power output at cruising speeds.]. Exp. B i d . 55, 489-520. Webb, P. W. (1971b). The swimming energetics of trout. 11. Oxygen consumption and swimming efficiency. J . E x p . Biol. 55,521-540. Webb, P. W., and Brett, J. R. (1973). Effects of sublethal concentrations of sodium pentachlorophenate on growth rate, food conversion efficiency, and swimming performance in underyearling sockeye salmon (Oncorhynchus nerka).]. Fish. Res. Board Can. 30, 49S507. Whitworth, W. R., and Irwin, W. H. (1964). Oxygen requirements of fishes in relation to exercise. Trans. Am. Fish. Soc. 93. 209-212. Wiehs, D. (1973a). Optimal fish cruising speeds. Nature (London) 245,48-50. Wiehs, D. (197313). Hydromechanics of fish schooling. Nature (London)241,290-291. Wiehs, D. (1974). Energetic advantages of burst swimming of fish. J. Theor. Biol. 48, 215-229. Winberg, G. G. (1956).“Rate of Metabolism and Food Requirements of Fishes,” 251 pp. Beloiussian State Univ., Minsk (In Russian). (Fish. Res. Board Can., Trans]. Ser. No. 194.) Wohlschlag, D. E. (1957). Differences in metabolic rates of migrating and resident freshwater forms of an arctic whitefish. Ecology 38, 502-510.

2.

SWIMMING CAPACITY

187

Wohlschlag, D. E. ( 1964). Respiratory metabolism and ecological characteristics of some fishes in McMurdo Sound, Antarctica. Antarct. Res. Ser. 1, 33-62. Wohlschlag, D. E., Cameron, J. N., and Cech, J. J., Jr. (1968). Seasonal changes in the respiratory metabolism of the pinfish (Lagodon rhomboides).Contrib. Mar. Sci. 13, 89-104. Woodhead, P. M. J. (1964). Diurnal changes in trawl catches of fishes. Rupp. P.-V. Reun., Cons. Int. Erplor. Mer 155, 35-44. Yuen, H. S. H. (1966). Swimming speeds of yellowfin and skipjack tuna. Trans. Am. Fish. Soc. 95,203-209. Yuen, H. S. H. (1970). Behaviour of skipjack tuna, Katsuwonus pelamis, as determined by tracking with ultrasonic devices.J. Fish. Res. Board Can. 27, 2071-2079. Yuen, H. S. H., Dizon, A. E., and Uchiyama, J. H. (1974). Notes on the tracking of the Pacific blue marlin, Makaira nigricans. In “Proceedings of the International Billfish Symposium, Kailua-Kona, Hawaii, 9-12 August 1972” (R. S. Shomura and F. Williams, eds.), Part 2, Review and Contributed Papers, pp. 265-268 (NOAA Tech. Rep. NMFS SSRF-675). NOAA, Honolulu. Zaitsev, V. P., and Radakov, D. V. (1960). Submarines in fishery research. I n “Soviet Fisheries Investigations in North European Seas” (1. J . Marty et al., eds.), pp. 463-466. VNIRO and PINRO, Moscow.

This Page Intentionally Left Blank

3 HYDRODYNAMICS: NONSCOMBROID FISH PAUL W.WEBB

. . .. . . . . . .. . ..

I. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11. Steady Swimming . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A. Biological Approach . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . B. Hydromechanical Approach ......................... C. Drag Reducing Mechanisms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111. Unsteady Propulsion . . . . . . . , . . . . . . . . . . . . . . . . . . , , . . . . . , . . . . . . . . A. Kinematics . . . . ..,... .. . .. .............................. B. Acceleration Performance . . . . , . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . C. Mechanics ... . . .. .. .. .. . .. ......... ....... .... ... .. ......... ........................ D. Work Done in Acceleration E. Drae Reduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . IV. Unsteady versus Steady Propulsion.. . . . . . . . . . . , . . . . . . . . . . . . . . . . A. Mechanics and Structure . . , . . . .. , .. . . . .. . . .. .. ... . . ... B. Performance , . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . , , , . , . . . . . . . . V. Mechanics of Median and Paired Fin Propulsion . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

..

.

Y

.. . . .

189

. .. .

.

190 191 192 199 218 22 1 222 226 226 226 227 228 228 229 230 232

FISH PHYSIOLOGY, VOL. VII Copyright @ 1978 b y Academic Press, Inc. All rights oi reproduction in any form re5ewed. ISBN 0-12-350407-4

190

PAUL W. WEBB

Symbols

A CD Cf C f lam C f turb DF

d dT

f e

m P PK PT R L

S t

U V W W

6* 77P

K

A Y Pe

d

tailbeat amplitude drag coefficient frictional drag coefficient frictional drag coefficient for laminar boundary layer flow turbulent drag coefficient for turbulent boundary layer flow frictional drag body depth; mean of depth and width for equivalent rigid body trailing edge depth tailbeat frequency length virtual mass per unit length total rate of working rate of energy loss to wake thrust power Reynolds number wetted surface area time swimming speed velocity of propulsive wave lateral velocity of propulsive segments resultant velocity of water displaced by propulsive segments displacement thickness of boundary layer mechanical efficiency of caudal propeller angle subtended by plane normal to the trailing edge to transverse axis of motion of that edge propulsive wavelength kinematic viscosity of water density of environment phase difference

I. INTRODUCTION Modem studies of fish propulsion mechanics can be considered to have started with the use of cinematographic techniques by Marey (1895). Although many years have elapsed since Marey’s pioneer work, research is still concerned with the same problems: How do fish swim? What is the resistance to motion? How much power is expended in swimming? Significant advance toward answering these questions has largely been made between 1960 and the mid-1970’s. This advance can be attributed to a technological advance in the de-

3.

HYDRODYNAMICS: NONSCOMBROID FISH

191

sign of efficient water tunnel respirometers (Marr, 1959; Blaika et al., 1960; Brett, 1963, 1964) and theoretical advances in hydrodynamic models, particularly by Lighthill (1969, 1970, 1971) and Wu (1971a,b,c,d). Fish propulsion patterns are extremely diverse (see Chapter 1). Although significant advances have been made in the understanding of some swimming modes, the majority have not progressed beyond the descriptions given in Breder’s (1926) collation of locomotory types. Emphasis has been placed on the relatively simple mechanics of body/caudal fin propulsion. Following Gray’s ( 1933a,b) classical discussion of propulsion in the eel, numerous quantitative observations have been made relating body and caudal fin movements to swimming speed (Bainbridge, 1958, 1963; Magnuson and Prescott, 1966; Yuen, 1966; Fierstine and Walters, 1968; Pyatetskiy, 1970a; Hunter and Zweifel, 1971; Smit et al., 1971; Webb, 1971a, 1973a); such observations have been encompassed by numerous theoretical models (Webb, 1975a). In contrast, the more complex locomotor patterns involving paired and noncaudal median fins have progressed little beyond qualitative descriptions of fin movements (Harris, 1937, 1953; Breder and Edgerton, 1943; Lissman, 1961; Nursall, 1962; Webb, 1973b). The rapid advance in understanding some swimming modes, contrasting with little advance in others, sets constraints on any discussion of propulsion mechanics. Only qualitative discussion, often speculative, is possible for swimming modes other than those involving body/ caudal fin movements. The latter are amenable to quantitative analysis. In addition to these constraints, others apply as a result of research emphasis on certain activity levels. These levels are steady swimming at cruising (sustained), prolonged, and sprint (burst) activities. Unsteady activities at acceleration (burst) and routine levels have been neglected. These constraints restrict the present discussion largely to bodykaudal fin propulsion in steady swimming activity.

11. STEADY SWIMMING Biologists and hydrodynamicists have taken different approaches to the problems of how fish swim and the calculation of thrust and power required. The traditional biological approach to the mechanics of fish swimming has not led to satisfactory quantitative solutions. The hydrodynamic approach has led to numerous models amenable to quantitative solution but these have largely been neglected by

192

PAUL W. WEBB

biologists. Recent hydromechanical models have led to generalizations on the mechanical importance of variation in body and fin morphology. Concerted efforts to rectify the two approaches have recently begun (Wu et al., 1975; Pedley, 1977; Alexander and Goldspink, 1977; this volume). A. Biological Approach

1. GENERATION OF THRUST

The mechanisms by which fish generate thrust have been described in terms of arbitrarily defined “segments” of the body and/or fins, regarded by Gray (1968) to be “the fundamental unit of aquatic undulatory propulsion.” Each segment is considered in a quasi-static time frame, i.e., at any instant the forces acting on a segment are the same as those for an identical segment moving steadily at the same speed and subtending the same angle to incident flow. The forces acting on the segment can be qualitatively described in terms of the. change in momentum of water deflected by the surface and by comparing a segment with a hydrofoil (Borelli, 1680; Pettigrew, 1873; Marey, 1874, 1895; Breder, 1926; Gray, 1933a, 1968; Bainbridge, 1963). Use of such quasi-static models was very effective in establishing the kinematic requirements for the generation of thrust (Gray, 1933a,b,c). Gray showed that thrust develops only when a propulsive segment subtends a positive angle of attack to the incident water flow. This occurs when a propulsive wave travels backward over the body at a velocity greater than the speed of forward swimming. The relationship between muscle shortening activity and the formation of propulsive movements has recently been described by Blight (1977). 2. THRUST,DRAG,AND POWER The quasi-static approach can be used to calculate net thrust and power. These are obtained by integrating instantaneous forces for each segment throughout a propulsive cycle (von Holste and Kuchemann, 1942; Parry, 1949; Gero, 1952; Taylor, 1952; Gray, 195313).Most models based on these principles are somewhat unsatisfactory because they usually involve too great an oversimplification of both body movements and body form and do not take into account interactions between segments (Webb, 1975a). Drag and power requirements for swimming fish have most commonly been calculated assuming a swimming fish can be compared with an equivalent manmade rigid body (the rigid-body analogy). The

3.

HYDRODYNAMICS: NONSCOMBROID FISH

193

analogy apparently stems from classicial observations by Sir George Cayley (ca. 1809, see Gibbs-Smith, 1962) who noted the streamline profiles of trout and dolphin, two animals capable of impressive swimming performance. In using the rigid-body analogy, drag may be calculated from standard hydrodynamic equations or determined by drag measurements on dead, anesthetized, or model fish (Magnan, 1930; Kempf and Neu, 1932; Denil, 1936; Harris, 1936; Richardson, 1936; Parry, 1949; Gero, 1952; Gray, 1936, 1957; Bainbridge, 1961; Kent et al., 1961; Osborne, 1961; Brett, 1963, 1965; Sundnes, 1963; Smit, 1965; Smit et al., 1971; Brown and Muir, 1970; Pershin, 1970; Pyatetskiy, 1970a,b; Webb, 1975a).

3. THE RIGIDLBODYANALOGY Details of flow patterns and the origins of drag for rigid bodies are given in standard hydrodynamic texts (e.g., Prandtl and Tietjens, 1934a,b). The main concepts pertinent to a discussion of drag are summarized by Webb (1975a). Drag arises from frictional and inertial (pressure) forces. The former arises as a result of viscosity in areas of flow with large velocity gradients. Pressure drag results from distortion of flow around solid bodies and, since it depends on the shape of a body, is often referred to as form drag. The flow around a body is divisible into two regions (Fig. 1).At the body surface, there is no slip between the body and the water, and the velocity is equal to that of the body. The velocity increases rapidly to 1% of that of the undisturbed free stream over a very shprt distance from the body. This region of flow is defined as the boundary layer and is characterized by steep velocity gradients; consequently, frictional resistance arises in this region. Beyond the boundary layer is a region of outer flow where velocity gradients are negligible and viscosity can be neglected. The outer flow is deflected around the body, and this deflection can lead to pressure drag. The velocity of the water increases as it is deflected around the front surface of a body. A maximum velocity is attained at the shoulder, the point of maximum thickness. Velocity decreases downstream over the tapering portion of the body. The changes in velocity, resulting from distortion of the flow, are associated with pressure changes. From Bernoulli’s theorem:

where AP, pressure difference; U , velocity.

194

PAUL W. WEBB

PRESSURE adverse

GRADIENTS

OUTER FLOW

favorable

LAMINAR TRANSITIONAL TURBULENT

BOUNDARY-LAYER I AMINAR

ATTACHED

TRANSITIONAL (PARTLY LAMINAR, PARTLY TURBULENT

TURBULENT

Some causes of Transition and Turbulence I) poor body shape. 2) adverse pressure gradient. 3) turbulence in outer flow. 4)surface roughness. 5) large size and/or speed.

Separation is encouraged SEPARATED

by adverse pressure gradients.

Fig. 1. Diagrammatic representation of streamline flow about a streamline body. The body is shown at rest relative to the fluid. The flow can be divided into two regions, the outer flow and the boundary layer flow (shown stippled). The boundary layer is a thin skin ofwater surrounding the body across which the velocity of the water increases from zero at the surface to the velocity of the free stream or outer flow. Flow in either region may b e laminar, turbulent, or may be transitional, changing from laminar to turbulent flow. The boundary layer separates from a solid body near the downstream edge (trailing edge) as a result of backflow in the boundary layer. The separated boundary layer forms a wake downstream of the body. Adverse pressure gradients, with increasing pressure in the opposite direction to mean flow, occur downstream of the point of maximum thickness (shoulder) of the body. Adverse pressure gradients facilitate premature boundary layer separation. Transition is encouraged by poor body shape, adverse pressure gradients, turbulence in the outer flow, surface roughness, and high Reynolds numbers (large size and/or speed). (From Webb, 1975a, Bull. Fish. Res. Board Can. No. 190, 159 pp.)

3. HYDRODYNAMICS: NONSCOMBROID FISH

195

Consequently, the pressure in the water decreases up to the shoulder, and subsequently increases toward the trailing edge. The former pressure change results in a favorable pressure gradient increasing in the direction of mean flow. The increase in pressure downstream results in an unfavorable pressure gradient increasing in the opposite direction to mean flow. The unfavorable pressure gradient interacts with the boundary layer, and together with frictional forces reducing boundary layer velocities, eventually causes back flow in the boundary layer. As a result the boundary layer separates from the body and a wake is formed. This dissipates kinetic energy and conti-ibutes to form drag. In addition, the downstream flow remains distorted and the net pressure across the wake is lower than in the upstream undisturbed water. Therefore, there is also a net pressure force acting along the body resisting motion. The magnitude of the frictional drag force is readily calculated from the standard Newtonian equation:

1 Df = z p eSU2Cf

( 2)

where Df,frictional drag; pe, density of environment; S, wetted surface area; Cf, drag coefficient. Power expended against frictional drag will be the product DfU . Cf depends on a nondimensional number, Reynolds number, and boundary layer flow conditions. Reynolds number can be visualized as a ratio of inertial to viscous forces that can be closely approximated to

where RL, Reynolds number; e, length, the characteristic dimension related to size; v, kinematic viscosity of water. For well-engineered bodies and smooth flat plates with the long axis at zero angle of incidence to the flow, boundary layer flow tends to be laminar up to RL of approximately 5 x lo5.This Reynolds number would be achieved, for example, by a 30 cm fusiform body or flat plate at a speed of 180 cm/sec (6 t/sec). Above this critical Reynolds number transition tends to occur to turbulent boundary layer flow. Transition is encouraged by such factors as turbulence in the outer flow, adverse pressure gradients, and surface roughness as well as large size or speed that increase RL. Cf may be expressed as a function ofRL.For laminar boundary layer flow Cf lam = 1.33 R L - 0 . 5 (4)

196

PAUL W. WEBB

and for turbulent boundary layer flow

Cf turb = 0.072RL-o.2

(5)

Frictional drag calculated from Eqs. (4) and (5)is greater at a given

RL for turbulent boundary layer flow than for laminar boundary layer flow because of the additional motions of water particles in the former flow pattern. Pressure drag is more difficult to determine. It is dependent on the point at which the boundary layer separates. If the boundary layer separates prematurely, that is, relatively close to the shoulder on a streamlined body, pressure drag is greatly increased. Pressure drag is lowest on well-designed fusiform bodies when separation occurs close to the trailing edge. Then pressure drag can be calculated as a multiple of Cf so that

c, = cf [ i + i.s(d/e)1.5+ 7(d/t)31

( 6)

where C,,, total drag coefficient; d, mean of depth and width. The factors enclosed in the square brackets of Eq. (6) give values of the order of 1.2 for fish (Bainbridge, 1961). Equations (2) to (6) calculate the theoretical drag for a rigid body geometrically similar to a fish. This is the minimum possible drag for a swimming fish. Gray (1936) first applied such calculations of theoretical drag to compare thrust power required with estimated muscle power available for a porpoise and a dolphin. The animals were probably swimming at sprint speeds. Muscle power was calculated by comparison with the sustained performance of an Olympic rowing crew. Gray found that the estimates of power required exceeded the estimates of muscle power available, the result being the basis for “Gray’s Paradox.” Gray’s Paradox has been widely interpreted to imply that the drag of swimming fish and cetacea is lower than that of the best engineered manmade body. In order to attempt to improve on theoretical drag, measurements of drag have been made for dead fish, anesthetized fish, and models, collectively referred to as dead drag measurements. These are reviewed by Webb (1975a). Such measurements are of little value however because pressure drag is greatly increased as a result of fluttering of fins and the body or inadequate methods. The magnitude of dead drag measurements expressed as drag coefficients as a function of RL is included in Fig. 5. Results from drag measurements, both of theoretical and dead drag, have been inconclusive with respect to Gray’s Paradox, some apparently being supportive (Osborne, 1961;

3.

HYDRODYNAMICS : NONSCOMBROID FISH

197

Brett, 1963, 1965; Smit et d.,1971) while others conflict (Bainbridge, 1961; Walters, 1966; Smit, 1965). With the perspective of time, it is clear that Gray did not have complete data (Bainbridge, 1961).Consequently, it is unfortunate that much subsequent effort has been expended in trying to explain Gray’s Paradox rather than evaluating the underlying assumptions. Bainbridge (1961) has discussed the question of muscle power, and his higher values for muscle power loading are now most commonly used. Using more comprehensive data than were available to Gray, Bainbridge (1961) found little problem for fish and cetacea in meeting theoretical drag power requirements. 4. FLOWA N D DRAGOF A SWIMMING FISH The basic mechanical assumption of the rigid-body analogy is that the flow around a swimming flexing fish is mechanically equivalent to that around an equivalent rigid body. There is currently no evidence to support this assumption, while the scant data available tend to refute the assumption. The body of most fish is not held rigid during swimming. Exceptions are apparently fish such as Carangids and some Scombrids, and also Cetaceans (Breder, 1926; Walters, 1962; Hertel, 1966; Fierstine and Walters, 1968; Magnuson, 1970). In most fish, both the body and caudal fin execute lateral movements with propulsive segments moving across, and at some angle to the incident flow. Large pressure differences will be set up on either side of a propulsive segment. As a result cross-flows are expected around dorsal and ventral surfaces leading to boundary layer separation and formation of a more extensive wake than for an equivalent rigid body. Such cross-flows have been observed by means of threads and dyes (Houssay, 1912; Rosen, 1959; Gray, 1968; Webb, unpublished observations). Pressure recordings along the.body of a swimming fish (DuBois et al., 1974) also suggest strong cross-flow (Section II,C, 1,c). Aleyev and Ovcharov (1969, 1971) present photographs of cross-flow about several species that show separation over the posterior half to one-third of the body of fish swimming in subcarangiform modes. The formation of a relatively extensive wake as a result of swimming movements will result in increased distortion of the outer flow and increased pressure drag. Some fish approaching and including the carangiform mode with lunate tail minimize this problem by concentrating thrust at the caudal fin while holding the body fairly straight and by streamlining of the caudal peduncle (Hertel, 1966; Magnuson, 1970) (see Chapter 4).

198

PAUL W. WEBB

Frictional drag of many fish is also expected to be greater than that for an equivalent rigid body as a result of premature transition from laminar to turbulent boundary layer flow. Observations by Allan (1961), Walters (1962), and Aleyev and Ovcharov (1969) show that projections such as eyes and nares or gill effluent can cause such premature transition. Turbulence is probably induced in some fish to reduce the chance of flow separation, notably in elasmobranchs covered with projecting scales which make the body surface rough (Walters, 1962; Bone and Howarth, 1966; Ovchinnikov, 1966; Bone, 1972). In addition, locomotory movements themselves will increase local frictional drag of propulsive segments. Such segments move discontinuously and with a mean resultant velocity higher than that of the forward speed of the fish. The boundary layer thickness is reduced on the leading surface and increased on the trailing surface, with an increase in the net velocity gradient in the boundary layer. Therefore this boundary layer thinning phenomenon increases net frictional drag (Lighthill, 1971; Webb, 1973a). An observation by Stevens (1950) has been widely used to support an alternate hypothesis that a laminar boundary layer is found under conditions where turbulent boundary layer flow would be expected. Stevens observed that the only disturbance caused by a dolphin swimming in phosphorescent water was two diverging lines in the wake. In contrast, the wake downstream of a seal was highly turbulent. The observations are clearly for the wake downstream of the trailing edge and not for the flow in the boundary layer anterior to that edge. Consequently no direct information on laminar or turbulent flow conditions in the boundary layer can be deduced. In practice, Stevens’ observation implies that the boundary layer flow remained attached for the dolphin but separated prematurely for the seal. Such separation would result in a highly turbulent wake as indicated by the extensive phosphorescence in the seal’s wake. The phosphorescence in the wake of the dolphin was probably caused by tip vortices (von Mises, 1945; see also Chapter 4) about the tail flukes (Webb, 1975a). These considerations suggest that the flow about a swimming fish will not be mechanically similar to that about a rigid body. Unfortunately, observations are few, and rigorous experiments have so far proved technically difficult (see, e.g., Allan, 1961). Nevertheless, it is clear that the rigid-body analogy should not be used to calculate swimming thrust or power unless the basic assumptions can be supported. This seems unlikely for most fish, so that further use of the rigid-body analogy is difficult to justify for calculating thrust and

3.

HYDRODYNAMICS: NONSCOMBROID FISH

199

power for steady propulsion. Therefore an alternate method is desirable.

B. Hydromechanical Approach

1. MODELS Theoretical physicists, mathematicians, and engineers have usually attempted to formulate mathematical models describing the observed kinematics of fish. These lead to various hydromechanical models of how fish swim which are amenable to quantitative solution. Earliest hydrodynamic models were usually based on a quasistatic approach (Taylor, 1952; Gray, 1953b) and considered simplified undulatory motions in real fluids. Such models are generally oversimplified in terms of real fish motions and also in terms of real fish shapes (Webb, 1975a). In addition, interaction between segments is not included as Gray (1933a, 1953b) points out in connection with his quasi-static approach to propulsion models. Later models, starting with Gadd (1952) and particularly Lighthill (1960) and Wu (1961), considered more realistic fish-type motions but in an inviscid (frictionless) fluid when viscosity is neglected. Both resistive (viscous) and reactive (inertial) forces are involved in fish propulsion but the relative magnitude of these varies with Reynolds number. Reynolds numbers for fish commonly exceed lo3 when inertial forces are very much greater than viscous forces. Consequently, reaction, inertial forces are most important in thrust generation and resistive, viscous forces can initially be neglected. Lateral flattening typical of fish enhances the magnitude of inertial forces so that viscosity effects are further reduced relative to inertial effects (Lighthill, 1969, 1970). Although most models of fish propulsion have treated fishlike motions in an inviscid fluid, several have considered the effects of viscosity on model predictions. Such discussion leads to approximate solutions that further support the hypothesis that viscous forces are negligible at higher Reynolds numbers of the order observed for fish (Lighthill, 1960, 1970; Wu, 1971d). Numerous hydromechanical models have been formulated since the early models by Gadd (1952), Lighthill (1960), and Wu (1961) (see Webb, 1975a, for references). More recent models have emphasized a continuous range of propulsion patterns within which the various modes described by Breder (1926) are 'wlrvenient reference points. Variation in body/caudal fin locomotory systems can be mechanically

200

PAUL W. WEBB

separated into two types for nonscombroid fish up to the lunate tail condition and for scombroid fish, cetacea, and some extinct aquatic reptiles with a lunate tail (Lighthill, 1969). The models therefore provide a homogeneous approach to fish propulsion questions. However, the majority of models are mathematically complex and would require measurement of parameters that are difficult to obtain with reasonable or repeatable accuracy ffom live fish. Models formulated by Lighthill (1970) and Wu (1971a,b,c,d) lead to an important conclusion that for steady swimming, an acceptable estimate of the mean rate of working and mean thrust developed by the propulsive wave can be obtained from a simple treatment of trailing edge kinematics. This solution, obtained by Lighthill (1969),provides a key framework about which biological problems may be approached. Most hydromechanical models have also emphasized small amplitude harmonic lateral movements of long slender fish of constant depth. The assumption that lateral movements are harmonic permits mathematical treatment of overall propulsive movements, but is not necessary for detailed analysis (Lighthill, 1971). The assumption is not unreasonable, since the difference between observed fish motions and simple harmonic motion is small (Gray, 1933a). Furthermore, most fish are not long and slender, and neither are they of constant depth. Some of the most important recent advances in the physical approach to fish propulsion have come from models that take into account fish morphological variation ( Lighthill, 1969, 1970, 1971; Wu, 1971a,b,c,d; Wu and Newman, 1972; Newman, 1973; Newman and Wu, 1973). Current understanding of fish propulsion is based almost entirely on these recent studies that lead to new functional interpretations of morphology and swimming mode of fish. 2. LIGHTHILL’SSIMPLIFIEDBULKMOMENTUM MODEL In order to consider how fish swim, and to calculate mechanical power developed, the bulk momentum model described by Lighthill (1969) will be discussed. The detailed models of Lighthill (1960,1970) and Wu (1971a,b,c,d) consider mechanics and interaction among all propulsive segments, and lead to the conclusion that bulk momentum changes can be predicted from trailing edge kinematics. Lighthill’s bulk momentum model considers these latter movements and associated forces, and consequently provides a simple reference point about which to discuss locomotion problems. In the first instant, it is convenient to consider a long slender fish of negligible thickness compared to its length and of constant depth. A

3.

HYDRODYNAMICS: NONSCOMBROID FISH

201

propulsive wave passes backward over the fish’s body at a velocity, V, greater than the fish’s forward velocity, U . Propulsive segments execute small amplitude lateral harmonic movements, the amplitude of which increases with distance along the body to a maximum at the trailing edge. These various assumptions will be subsequently relaxed. To visualize how the net effect of interacting propulsive segments can be described in terms of trailing edge motions, consider two adjacent segments of the fish body. Each segment is of unit length. It accelerates the water in its vicinity and thrust is generated proportional to the product of the mass of water affected and the increase in velocity given to that mass. Let the incident velocity to the first segment equal U . The water is accelerated to U1 > U , in time t, and the thrust T1 developed by the first segment is proportional to

u, - u

T1am-

t

where m is the mass of water affected. The incident velocity at the second segment is U , . Because amplitude increases along the body, the second segment further increases the velocity of the water to U 2 .Thus, thrust T2 is generated by the second segment proportional to

Since the fish is assumed to be of constant depth, m will be the same for each segment. As the propulsive wave will travel along the fish at reasonably constant velocity, t will also be the same for each segment. From Eqs. (7) and (8) the total thrust from both segments is then proportional to u 2

T1+T2am-

-u t

and hence the total velocity increment can be calculated from the kinematics of the second segment. This example shows that the propulsive wave can be visualized as a series of interacting segments that progressively increase the momentum of water in the vicinity of the fish to reach maximum values at the trailing edge. The rate at which momentum is shed by the trailing edge is equal to net thrust on the fish, generated by the propulsive movements of all interacting propulsive segments.

202

PAUL W. WEBB

Ligh6ill's bulk momentum model shows how net thrust and power generated by propulsive movements may be simply obtained from the kinematics of the trailing edge. From Eqs. (7) to (9) two aspects must be taken into account to calculate momentum changes of the water in the vicinity of the fish. These are the mass of water affected and the movements of the trailing edge, the latter accelerating the water to increase its momentum and hence generating thrust. As soon as a segment begins to move, the water in the vicinity is set in motion. The mass of water affected by movements of a segment is the virtual mass. At the trailing edge, m is given by

m = k p e r dT2 4 where dT,trailing edge depth; pe, density of environment; k, constant dependent on body sectional shape. The dimensions for Eq. (10) are mass x length-'. In order to understand how propulsive movements generate thrust, momentum changes are calculated for a water slice accelerated by a segment. In the small amplitude case, a water slice is given by Eq. (10) as a cylinder of unit length, with a diameter equal to dT oriented normal to the axis of progression of the fish. Orientation of the water slice normal to the segment becomes important for large amplitude motions, and the small amplitude restriction is relaxed below (Section II,B,4,a). The effects of propulsive movements on the water slice are considered for a propulsive segment at the trailing edge (Fig. 2). Movements of the segment are followed for a short period of time, t , after which the water slice is just left behind at the trailing edge. The length of the segment is defined for mathematical convenience in relation to the distance V t covered by the propulsive wave traveling backward over the body at velocity V. During the same period of time, the segment moves forward a distance Ut at the velocity of the fish U , and the trailing edge moves laterally a distance Wt at the trailing edge lateral velocity W. Thus the segment moves from aa, to bb, in Fig. 2. The water slice just left behind by the trailing edge at t is displaced a distance w t , given a velocity w smaller than W. The velocity w given to the water slice can be calculated from the geometry of movements of the segment. From similar triangles

wt

wt= and

(V - U ) t Vt

3.

HYDRODYNAMICS: NONSCOMBROID FISH

203

B

Vt

Wt

Fig. 2. Simplified representation of body and trailing edge movements for the calculation of thrust and power generated by propulsive movements using Lighthill’s bulk momentum model. (A) Tracings of successive positions at 0.015 sec intervals of the body centerline of a trout (30cm total length) swimming at 50 cm/sec. The position of a water slice is represented that is just left behind by the trailing edge at position 4. (B) Representation of a segment at the trailing edge at two instances in time. The position of a water slice is illustrated that‘is just left behind by the trailing edge segment. This segment moves from u u , to bb, in time t. Further explanation is given in the text. (Modified after Webb, 1975a, Bull. Fish. Res. Bourd Can. No. 190, 159 pp.)

The momentum given to the water as a result of movements of the trailing edge segment is equal to mw. Water with momentum mw is shed to the wake (i.e., left behind at the trailing edge) as the fish moves forward at velocity U . Thus the rate of shedding momentum (thrust) is mwU. Power is generated by the trailing edge working against the wake momentum at the rate W. Therefore the net total power, P, expended b y propulsive movements is given by

P=mwUW

( 13)

P has the dimensions of mass x length’ x time-3 and is calculated in ergs sec-I or watts.

204

PAUL W. WEBB

P is the total rate of working of the propulsive system, but not all the power generated is available as thrust power to overcome drag. Some power is lost energy to the wake. This lost component arises from the energy required to increase the momentum of the water in the vicinity of the fish. Since each water slice is accelerated to w , the lost energy is equal to the kinetic energy of the water shed into the wake at U . The lost energy, PK,is therefore given by

The power component to overcome drag, the thrust power PT, is found by difference from Eqs. (13)and (14)

PT = P - PK

( 15)

In using a bulk momentum approach to fish propulsion, it is important to note that the average effect of all propulsive segments is obtained for a complete propulsive cycle. Thrust calculated for any shorter period of time must take into account instantaneous contributions of all segments. Parameters for use with Lighthill’s bulk momentum model are easily measured from photographs of swimming fish (Webb, 1971a, 1973a, 1975a). Measurements of the tailbeat frequency, f, amplitude, A, propulsive wavelength, A, and trailing edge depth, dT,are required at any speed, U . Then, assuming simple harmonic motion V

=fx

( 16)

By substitution for W and V in Eq. (12) we have

and m is given by Eq. (10). 3. EFFECTSOF BODYFORM O N

SWIMMING

MECHANICS

a. Body and Fin Sections. The virtual mass of a propulsive segment depends on the structure of the body and fins comprising that segments’ cross-sectional shape (Lighthill, 1970). The constant k in Eq. (10) is close to 1 for many typical fish body or fin sections. k is usually less than 1where the body section includes both body and fins (Fig. 3). Lowest values of k in this case occur when the fins represent

3.

205

HYDRODYNAMICS: NONSCOMBROID FISH

A

k = 1.0

B

C

1.0 0.83

D

0.76

Fig. 3. Diagrammatic representation of some body and fin cross section shapes and values of k for Eq. (10).(A) Elliptical body section without fins; (B) fin section as at the caudal fin trailing edge; (C) elliptical body section with equal median fins comprising 25%of total depth; (D) circular body section with equal median fins comprising 25%of total depth. Values for k based on Lighthill (1970).

approximately 25% of total depth, a not uncommon value for fish. Thus for a fish with an elliptical body section and equal dorsal and ventral fins comprising 0.25 total depth, k would be 0.83, and for a circular body section but the same fin depth, k would be 0.76. k would be closer to 1 for greater proportions of fin or body depth and also for a single dorsal or ventral fin comprising the same proportion of total depth (Lighthill, 1970; Wu, 1971~).

b. Znteraction between Median Fins. The depth of most fish is not constant, but is discontinuous because of median fins and tapering b,ody Sections. The trailing edge of any fin does work in the same way as the trailing edge of the body. A wake is formed downstream of any trailing edge carrying that ‘momentumand energy given by the trailing edge movements. In hydromechanical theory, a vortex sheet is shed at any trailing edge carrying that momentum and energy. The vortex sheet concept is important in treating flow in inviscid fluids. It permits more realistic treatment of discontinuous flow patterns to describe flow that would occur in a fluid with viscosity (Prandtl and Tietjens, 1934a,b; Schlichting, 1968). The vortex sheet shed b y an upstream median fin flows along the body at the speed of the fish. Where there is a second re-entrant me-. dian fin (that is, a second downstream fin with depth exceeding that of the vortex sheet) the vortex sheet is absorbed at the second fin’s lead: ing edge. Thus the gap between the fins is filled by a vortex sheet. The vortex sheet shed by the second fin’s trailing edge is dictated by that edge’s motion, and the upstream vorticity does not contribute to the wake. However, the vortex sheet filling the gaps between the fins functions in the same way as a continuous fin (Lighthill, 1970; Wu, 1971c,d; Wu and Newman, 1972; Newman and Wu, 1973; Newman, 1973).

206

PAUL W. WEBB

The phenomenon of vortex sheets filling gaps between fins leads to improved propulsive effectiveness for ~o mechanically different body and median fin patterns. The first pattern is typical of Gadids, where the gaps between fins are small. The second is typical of elasmobranchs where gaps are large. In both cases, the vortex sheet filling the fin gap does not reduce thrust generated nor energy loss (efficiency). However, the surface area of the body is reduced between the fins so drag will be lower (Lighthill, 1970; Wu, 1971c,d). In contrast to the case where fin gaps are small, thrust can be increased in addition to the reduction in drag when fin gaps are large (Lighthill, 1970; Wu, 1971d). This occurs when the phase difference between momentum carried by the vortex sheet and the lateral movement of the trailing edge exceeds 0 . 5 ~ The . phase difference in lateral movements between an anterior fin trailing edge at a positionx, along the body and a posterior fin leading edge at X2 is given by

However, the vortex sheet shed by the anterior fin moves downstream at velocity U while the propulsive wave travels backwards at a velocity, V, greater than U . Consequently, the phase difference, 4, between body motions incident to the vortex sheet is

Gray ( 1933c) shows figures of the elasmobranch Acanthias vulgaris (8 = 46.0 cm) with U = 17.8 cm/sec, and V = 25.6 cm/sec. The trailing edge of the first dorsal fin was about 19.4 cm from the nose. The second dorsal fin was small so the bulk of the vortex sheet from the first dorsal fin trailing edge would be absorbed at the caudal fin leading edge, approximately 37.1 cm from the nose. These data give a value of 4 of 0.52~. In many fish, the depths of the downstream fin leading edges are less than the depth of upstream trailing edges. Then, the upstream vortex sheet will not be completely absorbed at the downstream leading edge. Under these conditions upstream trailing edges contribute to both thrust and energy losses (Wu, 1971c,d). However, the net contribution of such anterior edges is likely to be relatively small. Although the virtual mass of these sections is higher than the trailing edge, lateral displacement and hence velocities and the momentum given to the water will be small in comparison with body trailing edge contributions.

3.

HYDRODYNAMICS: NONSCOMBROID FISH

207

Just as gaps between median fins can lead to reduced drag without significantly affecting thrust and efficiency, so can scooping out the caudal fin. The central portion of a swept back caudal fin, typical of many pelagic fish, will be filled by a vortex sheet. The tail is then mechanically similar to a structurally complete tail fin (Lighthill, 1970) but drag will be reduced as a result of the decreased area. Relatively large savings in drag are possible because the drag per unit area will increase to a maximum along the body as a result of increases in lateral velocity of propulsive segments with caudally increasing amplitude. Then boundary layer thinning and hence frictional drag will increase toward the caudal fin trailing edge.

c . Viscosity and Virtual Mass. The effects of viscosity in real fluids are initially neglected by hydromechanical models. The virtual mass of a segment is affected by the boundary layer and hence viscosity. The importance of this modification to virtual mass is partially dependent on the distribution of depth along the body. Weihs (personal communication) showed that the virtual mass was reduced according to m - 2p6*d (21) where 6* is the displacement thickness of the boundary layer. Precise determination of 6* is difficult because of boundary layer thinning and possible boundary layer separation upstream of a segment (D. Weihs, personal communication). Standard hydrodynamic equations suggest that correction for virtual mass would be approximately 1.5% at the trailing edge of a 10 cm fish of constant depth swimming at 50 cm/sec, decreasing to 0.9% for a geometrically similar 100 cm fish swimming at the same speed. The thickness of the boundary layer increases along the length of any body and also increases with time. Where body depth varies, as with discrete fins, the boundary layer will tend to increase in thickness along each fin. Consequently, the boundary layer may be thinner for a discrete fin compared to a similar point on the body of a fish of constant depth. Therefore, it is likely that the virtual mass correction will be smaller for fish of discontinuous depth. In general, effects of viscosity on virtual mass are small enough to neglect. d . Tapering Body Depth. The body depth of many fish tapers, usually only as far as the caudal peduncle prior to a large caudal fin. When the body is lenticular in section or when there is a tapering median fin [Wu's (1971c,d) ribbon fin-type] a vortex sheet is shed from the sharp tapering trailing edge. The vortex sheet will be absorbed as usual by a reentrant downstream fin, most commonly the caudal fin,

208

PAUL W. WEBB

with little effect on thrust or energy loss (Lighthill, 1970; Wu, 1971c,d; Wu and Newman, 1972). Some fish, swimming in the anguilliform mode, do not have reentrant downstream fins. If the body or fin taper to the trailing edge is gradual, as in Anguilla, it has little effect on thrust or efficiency. When the body tapers over an extended’length, thrust and efficiency are affected (Lighthill, 1970; Wu, 1971d). Thrust is generated as usual by all propulsive segments in proportion to the mass of water affected and the lateral velocity. However, when the body tapers, the mass of water affected decreases, In the anguilliform mode, increases in lateral velocity are usually small over the tapering portion of the body (Gray, 1933a) so that there is unlikely to be compensation for the decrease in virtual mass. Therefore, more posterior segments will contribute little additional thrust, but the kinetic energy given to the water will contribute to wasted energy. Consequently, propulsive efficiency will be reduced. e. Body Thickness. In the first instance, basic hydromechanical models consider slender bodies, where body thickness is small compared to length. Most fish are relatively short and thick. Newman and Wu (1973) and Newman (1973) have shown that, under these circumstances, vortex sheets shed upstream can interact with body thickness to affect both thrust and drag. Newman and Wu (1973) and Newman (1973) considered flow for an axisymmetrical body (circular body section) with side fins. They found that the importance of body thickness depended on the proportion of total depth represented by side fins and the depth of the caudal fin relative to maximum depth. Both thrust and drag, and the ratio between thrust and drag, increase to some maximum as the body comprises an increasing proportion of total depth. The maximum depends on the depth ofthe caudal fin. The caudal fin dominates generation of thrust and drag as its trailing edge becomes large relative to the maximum depth of the body. When the caudal fin trailing edge is equal to, or exceeds, maximum upstream depth, effects of finite body width become neglible (Newman, 1973). For most fish, the caudal fin is large enough that body thickness effects are small enough to neglect.

4. EFFECTSOF

SWIMMING

MOVEMENTS

a. Large Amplitude Mouements. Real fish propulsive movements are of large amplitude. Lighthill (1971) showed that energy loss to the wake was increased by large amplitude movements because momentum shed at the trailing edge is not aligned with the axis of

3.

HYDRODYNAMICS: NONSCOMBROID FISH

209

motion. Wake momentum is shed across a plane normal to the trailing edge and therefore energy loss to the wake increases in proportion to the magnitude of the angle subtended b y that plane to the axis of forward progression. If the angle subtended by that plane is K , then energy loss to the wake is increased by l/cos K, and thrust is reduced proportionally from Eq. (15). Lighthill (1971) corrected for large amplitude movements of Leuciscus using data from Bainbridge (1963) by calculating instantaneous values for all kinematic parameters including K. The mean value of Ucos K was 1.18.Data for Anguilla from Gray (1933a) indicate l/cos K would be 1.20. Measurements on a single trout at several cruising swimming speeds gave lower values, independent of swimming speed, the overall mean llcos K being 1.06 (Webb, unpublished observations).

b. Swimming Mode. A major evolutionary trend in bodykaudal fin propulsion of fish is rednction in the proportion of the body involved in significant lateral propulsive movements. This was recognized by Breder ( 1926), who identified various modes within a continuously changing spectrum from anguilliform propulsion (most of the body involved in several half wavelengths in large amplitude lateral displacements) to carangiform propulsion (less than one-half of the body included in. large amplitude displacements over less than half a propulsive wavelength). The trend toward decreasing the proportion of the body executing large amplitude movements is of mechanical importance because the mass of water affected by propulsive movements increases with time to exceed m as calculated from Eq. (10)(Lighthill, 1970; Wu, 1971d). When a segment moves, there is an immediate increase in the momentum of water affected that is proportional to the segment’s velocity and virtual mass. In addition there is a time-dependent increase in momentum and kinetic energy associated with the vortex force gradually shed by upstream segments (Lighthill, 1970). Lighthill (1960, 1970) showed that the time-dependent increase in vortex force momentum was poorly correlated with trailing edge motions; that is, the trailing edge works against a component of the total momentum that is out of phase with that edge’s movements. As a result, this momentum does not contribute to thrust but does increase energy losses to the wake. Crude estimates of the vortex force led to the conclusion that trends reducing the proportion of the body involved in locomotory movements would reduce the magnitude of that momentum poorly correlated with downstream lateral movements (Lighthill, 1970). For example, in the anguilliform mode, the high

210

PAUL W. WEBB

proportion of the body executing lateral movements permits time for the development of a relatively large vortelr force poorly correlated with the movements of the trailing edge. This is reduced to become insignificant in carangiform modes when the distance over which lateral movements increase in amplitude is relatively smaller with reduction in the time-dependent vortex force. The inclusion of a large part of the body in the propulsive wave also increases the time and distance over which cross-flows occur. Such cross-flow could encourage separation of the boundary layer and increase drag. With the progressive concentration of significant propulsive movements caudally, recoil forces are generated. In the anguilliform mode, lateral forces are more or less balanced b y equal and opposite lateral forces distributed along the more complete propulsive wave. Large recoil forces generated by the tail in carangiform modes tend to cause the anterior of the body to yaw. This would lead to a large energy loss unless minimized. The problem is greatest for fish swimming in the carangiform mode with lunate tail, this mode being the logical culmination in reducing the propulsive portion of the body. Recoil yawing movements are minimized by increasing body and fin depth anteriorly. The body is also frequently thicker anteriorly, concentrating body mass forward. As a result the virtual mass plus body mass is increased, increasing the inertial resistance to yawing forces. This is best illustrated in scombroid fish where the caudal fin and body are separated by a narrow caudal peduncle. The trend toward a reduced caudal peduncle is called narrow necking (Lighthill, 1969), and is a trend toward the separation of two mass centers, i.e., the tail generating thrust and yawing forces, with the latter being minimized by anterior body mass and enhanced virtual mass. Narrow necking is thus a morphological requisite for efficient propulsion in advanced carangiform modes (Lighthill, 1969, 1970). Fish swimming in the subcarangiform mode include a complete propulsive wavelength within the body length (Breder, 1926). However, recoil forces will be incompletely equalized because major increases in amplitude are restricted to the posterior half to third of the body. Such fish, for example, trout and salmon, have moderate narrow necking and anterior median fins increase virtual mass forward. Then recoil forces are partly offset by the inertia of the anterior part of the body together with some cancellation of recoil forces within the length of the propulsive wave. Narrow necking of a less abrupt form than in scombroid fish is a feature of many fish with elongate fusiform bodies, for example, most

3. HYDRODYNAMICS: NONSCOMBROID FISH

211

small elasmobranchs. Illustrations in Gray ( 1933a,c) show that recoil yawing forces are not effectively minimized in comparison with fish having more truncate bodies and swimming in carangiform modes. Elasmobranchs typically have a fairly large anterior median fin which probably contributes more to thrust as for Acanthias vulgaris (Section II,B,3) and less to stability.

c . Kinematic Variability. Treatment of mechanics of fish movements considers two dimensional motions of plane surfaces. In practice, detailed movements are more complex, particularly for the caudal fin of fusiform fish. Variability in caudal fin movements has been described in greatest detail by Bainbridge (1963). He was also able to give approximate explanations based on hydrodynamic forces acting on the caudal fin during a propulsive cycle. However, much of the detailed kinematics are apparently under active neuromuscular control by the fish (Breder, 1926; von Holste and Kuchemann, 1942; Harris, 1937; Bainbridge, 1963; Webb, 1971a). Complex movements of the caudal fin web are presumably related to continuous fine control of thrust and stability. I n addition, some long slender fish swimming in the anguilliform mode may include spiral, three-dimensional components in the propulsive wave (see Chapter 1).The significance of such components is not known. They have not been reported for shorter, more truncate fish swimming in mechanically more advanced propulsion modes. 5.

PROBLEMS IN

APPLYINGHYDROMECHANICAL MODELS

Recent hydromechanical models permit increasing accuracy in interpreting propulsion mechanics relating to variability in fish functional design at cruising swimming speeds (Lighthill, 1969, 1970, 1971; Wu, 1971a,b,c,d; Wu and Newman, 1972; Newman and Wu, 1973; Newman, 1973). However, it is recognized that no current model is complete. I n some cases, solutions to hydromechanical aspects of morphological variation are approximate. For example, quantifying the effects of vortex force poorly correlated with lateral movements (Lighthill, 1970) or body thickness effects (Newman, 1973). Nevertheless, such studies indicate the order of magnitude of such interactions is usually small (Lighthill, 1970; Wu, 1971d; Newman, 1973). All hydromechanical models neglect viscosity effects. This again probably involves small error in determining thrust for most fish with compressed bodies swimming at high Reynolds numbers (Lighthill, 1969, 1970). Viscosity becomes important for small fish when

212

PAUL W. WEBB

Reynolds number is low and when the body is not compressed. Lighthill ( 1969) clearly states the need for a composite reaction-resistance model but formulation of such a model is complex and problematic (Vlymen, 1974). Application of such models to practical biological problems is likely to be difficult. Taylor (1952) pointed out that complex models might lead to improvement in quality but when input data required become too great, simple collection of those data mitigates against applying the model in biologically comprehensive situations. Webb (1975a) suggested that a useful compromise would be the determination of modifying coefficients related to key morphological and kinematic parameters that could be applied within the framework of simpler models. Such an approach combines the advantages of detailed theoretical studies with limitations of practical observation and has recently been successfully applied to animal flight (Weis-Fogh, 1973). Comparative data for a sufficient number of fish species to test this approach are currently lacking. The difficult problem of including viscosity applies in theory mainly to smaller fish and less compressed fish swimming in anguilliform modes (Lighthill, 1970; Wu, 1971d). However, such effects probably change mechanical efficiency of larger fish swimming at lower cruising speeds in carangiform modes. The hydromechanical models discussed here predict very high efficiencies for the propulsive system (Wu, 1971c,d). For Lighthill’s bulk momentum model the mechanical efficiency of the propeller system would not be less than 0.5. This follows from Eqs. (12)-( 15)when caudal propeller efficiency, r),, is given by qP = 1 - 0.5

( v - u 7)

so that when U 4V, T , --* 1 and when U + V, qP + 0.5. In contrast, Webb (1971b) deduced that r), for rainbow trout (e = 29.2 cm) at low cruising speeds would be less than 0.5. For example, at a speed of 20 cm/sec r), would be 0.12 while the value from Eq. (22) is 0.67. The difference may be related to interactions between the upstream viscous vortex force growing with time but poorly correlated with trailing edge motions. The magnitude of this force, contributing to energy loss but not thrust, will be greater when V is large relative to U , compared to the case when V approaches U as is predicted for good propeller efficiency. Thus the relation U/V is often taken as representative of r), (Gadd, 1952; Taylor, 1952; Lighthill, 1960; Hertel, 1966; Wu, 1971d). When U / V is small, r), will be lowest from Eq. (22) and

3. HYDRODYNAMICS: NONSCOMBROID FISH

213

further reduced through poor correlation of part of the vortex force with trailing edge motions. U / V increases with swimming speed so that prediction of qp is reasonably accurate for high steady speeds (Webb, 1971b, 1975a). Possible error in calculating vp does not apparently alter predictions of thrust. vp is reduced through increased energy loss to the wake (increased P K )and the total rate of working (P) will be increased by the same amount. Then thrust from Eq. (15)is not changed. In general, therefore, the bulk momentum model is expected to give good first approximations of thrust (=drag). These values can be improved b y correcting for large amplitude motions. Effects of upstream vorticity not absorbed by reentrant downstream fins and interactions with body thickness can be taken into account by detailed analysis, but corrections appear to be small. The same applies to viscosity corrections for virtual mass. Mechani.ca1 efficiency can also be calculated for higher swimming speeds but correction is required at low speeds. This requires more detailed experiments but can be determined. Problems of viscosity for small fish are recognized but cannot as yet be satisfactorily incorporated in a first principles model. For larger fish, neglect of viscosity is not likely to be important. 6. VALUESFOR THRUSTAND POWER Good estimates of thrust and thrust power can be calculated from propulsion parametersf, A, A, and K at a given U . Few studies have measured all these parameters and few consider a range of speeds. Most complete data can be compiled for Carassius, Oncorhynchus nerka, and Salmo gairdneri at cruising speeds (Bainbridge, 1958, 1963; Smit et al., 1971; Webb, 1971a, 1973a, 1975a). Thrust power for these fish is shown as a function of swimming speed in Fig. 4. The slopes of the relationship between thrust power and speed for salmonids are approximately 2.8 and for goldfish the slope is approximately 2.5. These are values that would be expected for theoretical drag of a rigid body with turbulent and laminar boundary layer flow, respectively. However, it cannot be concluded from these data that boundary layer flow is turbulent or laminar in each case. Thrust power exceeds theoretical drag and the slope of the powedspeed relationship may reflect changes in pressure drag with speed. Figure 4 also includes a comparison between three methods of determining thrust power; Lighthill’s bulk momentum model, a detailed quasi-static analysis following Lighthill ( 1971; Webb, 1975a) and measurements from added drag loads (Webb, 1971a). Agreement

214

PAUL W. WEBB

SWIMMING SPEED

-

cmlrec

Fig. 4. Relationships between thrust power and swimming speed for three species of fish. Curussius uurutus, solid line, thrust power calculated from Lighthill’s bulk momentum model using equations for kinematic parameters in Bainbridge (1963) and Smit et 01. (1971). Oncorhynchus nerku, closed squares calculated from Lighthill’s bulk momentum model modified for large amplitude using data for each speed shown. Data from Webb (19734. Sulmo guirdneri, open squares show thrust power calculated in the same way as 0. nerku. Open triangles show calculations from a detailed quasi-static analysis and open circles show values obtained from experiments with added drag loads. (Data from Webb, 1971a, 1975a.) Stippled areas show power required exceeding that from red muscle, and solid triangles show the percentage of myotomal white muscle required to make up the red muscle deficit. Further explanation is given in the text.

among the three methods is good and provides support for use of hydromechanical models. Thrust or thrust power can also be expressed as nondimensional drag coefficients calculated from Eq. (2) and expressed as a function of the nondimensional Reynolds number (Fig. 5). Such drag coefficients are variable among species but are most commonly three to five times theoretical drag values for turbulent boundary layer flow. Metabolic studies lead to similar conclusions (Kliashtorin, 1973).Most of the data in Fig. 5 are for fish swimming in subcarangiform modes. Preliminary calculations for propulsion in advanced carangiform modes suggest drag exceeds theoretical drag by a factor similar to that for subcarangiform propulsion (see also Chapter 4).

3.

215

HYDRODYNAMICS: NONSCOMBROID FISH

I

I

to4

I

105 lo6 REYNOLDS NUMBER

10’

Fig. 5. Drag coefficients, based on thrust power, as a function of Reynolds number. Drag coefficients a;e calculated from thrust power values calculated from hydromechanical models using standard hydrodynamic equations [see Eq. (2)].The stippled area shows the range of observed dead drag coefficients, calculated in the same way from measurements on fish and models. Theoretical drag coefficients are shown for laminar (C, lam) and turbulent boundary layer flow (C, ,ulb). (Data and references are summarized in Webb, 1975a.)

7. ENERGETICS AND

VALIDATION O F

HYDROMECHANICAL MODELS

The various models for fish propulsion mechanics are based on hydromechanical theory and although predictions of thrust and thrust power appear reasonable they require validation. Ideally, validation would be based on observations of flow about swimming fish, but such observations remain problematic (Houssay, 1912; Gray, 1936; Allan, 1961; Kent et al., 1961; Rosen, 1959; Aleyev and Ovcharov, 1969, 1971). Comparison between models and measurements of thrust are tedious (Webb, 1971a). Estimates of muscle or metabolic power have most commonly been used for indirect evaluation of predictions of thrust and power (Gray, 1936; Hill, 1950; Bainbridge, 1961; Osborne, 1961; Brett, 1963; Smit, 1965; Walters, 1966; Smit et al., 1971; Webb, 1971a,b, 1973a, 1975a,b).

216

PAUL W. WEBB

Estimates of muscle power deduced from comparative studies by Bainbridge (1961) are most commonly used. Measurements of thrust power by Houssay (1912) indicate that Bainbridge's values are of the correct order of magnitude but are probably conservative (Webb, 1975a). A comparison between predictions of thrust power and maximum white muscle power is given in Fig. 6 for various fish. Half the myotomal muscle is assumed active at any instant. A caudal propeller efficiency of 0.75 was assumed, avalue that is probably conservative (Wu, 1971d). Thrust power values are within the range of expected power outputs for vertebrate white muscle systems. Red muscle is used at sustained cruising speeds and prolonged speeds. It is frequently found that insufficient red muscle power would be developed to meet requirements at higher prolonged speeds (Fig. 4). Red muscle power during swimming can be estimated assuming half the muscle works at any instant and that caudal propeller efficiency is 75%. For the species shown in Fig. 4, red muscle comprises 7.4% of the myotome for Carassius (Johnson and Goldspink,

WEIGHT

- gm

Fig. 6. A comparison between thrust power given by hydromechanical models and estimated propulsive muscle power output. Further explanation is given in the text. (Modified after Webb, 1975,Bull. Fish. Res. Bourd Can. No. 190, 159 pp.)

3.

HYDRODYNAMICS : NONSCOMBROID FISH

217

1973a) and 4% of the myotome for the salmonids (Webb, 1975~). Myotomal muscle mass is assumed to be 50% of body mass. This figure is conservative (Bainbridge, 1963; Brett, 1963). However, support of the red muscle at prolonged speeds by small amounts of white or pink muscle would generate sufficient power. Direct and indirect observations support such joint muscle function at prolonged speeds (e.g., Brett, 1964; Beamish, 1968; Pritchard et al., 1971; Smit et al., 1971; Webb, 1971b; Hudson, 1973; Johnson and Goldspink, 1973a,b,c). Oxygen consumption data are available for fish swimming at cruising and prolonged speeds. Metabolic power can be calculated from oxycalorific equivalents. These data usually indicate sufficient metabolic power to generate thrust at normal overall efficiencies for the muscles and caudal propeller (Kilashtorin, 1973; Webb, 1975a). A major exception is known for goldfish (Smit et al., 1971) at prolonged speeds. However, energy accounting is not complete at such speeds because anaerobic energy of uncertain magnitude contributes to total metabolic power. Adequate data are not available to calculate metabolic power at burst speeds (Beamish, 1968; Johnson and Goldspink, 1973a; Webb, 1975a). In general, the current understanding of muscle and metabolic systems lends support to thrust predictions from models. However, because muscle powers are estimates, and metabolic data are frequently absent or incomplete, decisive validation of models from energetic data is not currently possible. Nevertheless, since comparative studies can suggest likely performance levels for components of the propulsive system, muscle and metabolism data can be used as a rough check on models. The same estimates can be used to suggest areas requiring modification, an approach that has been applied to bird flight (Tucker, 1973). The corrections required for qp at low swimming speeds of rainbow trout (Section II,B,5) are based on such estimates (Webb, 1971a,b). 8. EFFICIENCY

The overall efficiency of converting metabolic power to thrust power increases with swimming speed (Brett, 1963; Smit, 1965; Smit et al., 1971; Webb, 1971a, 1975a). Maximum efficiencies calculated for prolonged speeds are of the order of 20%,as for other animals (Webb, 1971b, 1973a, 1975a). Efficiencies at higher burst speeds are not known. Several different theoretical and empirical studies suggest mean routine speeds of one body length per second are common for

218

PAUL W. WEBB

nonscombroid fish (Brett, 1965; Weihs, 1973c; Kerr, 1971). If these were steady swimming speeds, overall efficiencies would be low, that is, less than 10%. Efficiencies for unsteady routine swimming are not known, but probably exceed those for steady speeds in covering a given distance in a given time (Weihs, 1974).

C. Drag Reducing Mechanisms Numerous mechanisms have been proposed that might reduce the drag of swimming fish (Webb, 1975a). Most plausible mechanisms are found for fish with high speed cruising swimming behavior, particularly scombroid fish. The following summary discusses only likely drag reducing mechanisms. Theories of drag reduction involve reduction of frictional resistance, reduction of pressure drag, and reduction in wake energy losses. Minimal frictional drag in normal fluids occurs with laminar boundary layer flow and maximum frictional drag with turbulent boundary layer flow. Minimum pressure drag occurs when the boundary layer remains attached to the body but is increased whenever the boundary layer separates prematurely. Pressure drag can greatly exceed frictional drag so that it is often more important to reduce the former. A laminar boundary layer is less resistant to separation than a turbulent boundary layer. Consequently, it is often advantageous to induce more stable turbulent boundary layer flow. Frictional drag is then increased but the potential reduction in pressure drag is very much greater with a net reduction in total drag.

1.

MECHANISMS TO

MAINTAIN LAMINARBOUNDARY LAYERFLOW

a. Distributed Dynamic Dumping. Walters (1963)has described a subdermal canal system in the integument of the trachypterid Desmodema, an elongate fish. Critical Reynolds numbers would be exceeded at even low swimming speeds. The fish swims in the amiiform mode holding the body straight and Walters postulated that the subdermal canal system could delay transitions by distributing local high pressure disturbances over the whole body via the canal system. This system is mechanically feasible but possible reductions in total drag are probably small. Drag associated with propulsive movements of the long dorsal fin is likely to greatly exceed that for the body. Pore canal systems have also been described in the Istiophoridae

3. HYDRODYNAMICS: NONSCOMBROID FISH

219

(Walters, 1962) but it is doubtful if such a system could neutralize disturbances of the intensity expected at the swimming speeds of these fish. Ovchinnikov (1966) suggests these fish are covered by a completely turbulent boundary layer (Section II,C,2,a).

b. Body Shape. Well-designed body shapes can maintain a high proportion of laminar boundary layer flow by extending the distance up to the shoulder, experiencing a favorable pressure gradient (Walters, 1962; Hertel, 1966). In most nonscombroid fish the shoulder is fairly far forward (Houssay, 1912) so most of the body would in practice experience an unfavorable pressure gradient. Transition may be encouraged by propulsive movements (Section II,A,4). LaMinar boundary layer flow would be expected to cover a small area of such fish, but the major drag problem would be pressure drag. c . Properties of the Propulsive Wave. The propulsive wave propels water backward and the increase in water velocity will result in reduced pressure caudally. Such pressure reduction could partly offset the adverse pressure gradient otherwise found downstream of the shoulder (Gray, 1936; Lang, 1966). Webb (1975a) has pointed out that this effect must be small if propulsive efficiency is to be high, since the greater the increase in water velocity the greater the kinetic energy lost to the wake (Alexander, 1968). Measurements of pressure distribution along the sides of a swimming fish have shown that local pressure changes are small (DuBois et al., 1974). Pressures measured for the dorsal surface decrease in the required fashion to facilitate flow over the posterior portion of the body. However, since lateral pressures do not increase in the same way, the dorsal pressure changes probably result from cross-flow. Richardson (1936), Hertel (1966), and Wu (1971d) have suggested the harmonic nature of propulsive movements could stabilize laminar boundary layer flow. Such stabilization occurs with small amplitude harmonic oscillations (Schlichting, 1968), but the situation is not known for large amplitude fishlike movements when propulsive segments move at a small angle of attack to the incident flow. d. Ejection of Kinetic Energy into the Boundary Layer. Breder (1926), Walters (1962), and Gray (1968) have suggested that the gill effluent might inject kinetic energy into the boundary layer to maintain laminar flow, Observations on small fish have suggested gill effluents lead to transition and separation (Allan, 1961; Walters, 1962). Further observations under better controlled conditions are required (Aleyev and Ovcharov, 1971).

220

PAUL W. WEBB

2. MECHANISMS FOR THE DELAYAND PREVENTION OF SEPARATION a. Surface Roughness. Some fish have rough body surfaces. It is presumed that one of the functions of such a surface is to induce turbulent boundary layer flow, delaying separation. Fish in the family Istiophoridae have a rostrum with denticles to induce turbulent boundary layer flow. In Xiphids, the rostrum is sufficiently long that transition is expected without roughness (Ovchinnikov, 1966). Bone and Howarth (1966) have suggested the major drag problem for most elasmobranchs would be boundary layer separation and this would be minimized by turbulent boundary layer flow induced by denticles.

b. Vortex Generators. The integument of the castor-oil fish, Ruuettus pretiosus, includes special ctenoid scales, the tips of which project well beyond the body surface (Bone, 1972). Bone suggested that the scales could act as vortex generators, vorticity being shed by each scale caudally and towards the body. In this way energy derived from the free stream could assist boundary layer flow. These fish also possess a subdermal canal system that may inject fluid into the boundary layer when the fish swims. This fluid might increase turbidity of boundary layer flow or add further kinetic energy. 3. REDUCTION OF VISCOSITY It has been conclusively demonstrated that fresh fish mucus can reduce frictional drag. The literature is reviewed by Hoyt (1975) which may be referred to for further references. There is substantial variability in the effectiveness of mucus in reducing frictional resistance. This may occur because the concentration of active polymers has not been controlled, the experiments being performed on fresh homogenized mucus of unknown composition. There is an approximate relationship between swimming performance and the effectiveness of mucus in reducing frictional resistance (Rosen and Cornford,

1971). It is not certain that mucus functions to reduce drag under natural condition. Relatively large amounts of mucus are required to give major reductions in frictional resistance. For example, a 5% solution of mucus is close to maximally effective for many species. This concentration results in a 65% reduction in frictional drag of the Pacific barracuda (Rosen and Cornford, 1971). In addition, like other linear long-chain polymers, mucus only reduces friction in turbulent flow (Hoyt, 1975). Many fast swimming fish with effective friction-reducing

3. HYDRODYNAMICS: NONSCOMBROID FISH

22 1

mucus swim in advanced carangiform modes and have streamlined body shapes that are assumed to maximize laminar boundary layer flow. If this hypothesis is correct, then mucus would be of little importance for these fish. Providing mucus concentrations were large enough in the boundary layer of swimming fish, frictional reduction would be most important for fish swimming in anguilliform and lower carangiform modes when turbulent boundary layer flow is expected. 4. BEHAVIOR

Schooling fish have frequently been reported swimming in regular diamond formations (e.g., Keenleyside, 1955; Nursall, 1973). Some forin of drag reduction through interaction among members of the school has been postulated (Breder, 1926; Belyayev and Zuyev, 1969; Zuyev and Belyayev, 1970; Weihs, 1973b). Weihs (1973b) has shown that interactions can occur between the wake generated by oddnumbered rows in a school and fish in downstream even-numbered rows when fish swim in a regular diamond formation. Interaction reduces thrust required by second row fish while third row fish experience a mechanically uniform incident flow. Weihs calculated that net thrust could be reduced approximately 10% in the horizontal plane. Further reductions of up to 30% were feasible for interactions in the vertical plane in three dimensional schools. It should be noted that the evolution of schooling among fish is difficult to explain on the basis of locomotory energetic savings alone. Instead, if fish school, for whatever reason, then locomotory energetic advantages may accrue to some members of school.

111. UNSTEADY PROPULSION

Unsteady propulsion is an integral part of normal locomotor activity. Houssay (1912) described normal routine activity of fish as bursts interspersed with glides, that is, continuous unsteady propulsion. High speed acceleration and maneuver, such as in predatodprey behavior, negotiating fish ladders and water falls, and avoidance of nets also involves unsteady activity. This area of fish propulsion has largely been neglected. Some metabolic measurements have been made for routine unsteady activity (Fry, 1957, 1971); mechanics and performance have been described for high levels of unsteady activity (Gray, 1933d; Hertel, 1966; Weihs, 1972, 1973a; Webb, 1975a,c, 1976, 1977).

222

PAUL W. WEBB

A. Kinematics Acceleration behavior is variable. Acceleration at high rates is found in behavior ranging from simple fast starts (Hertel, 1966; Weihs, 1973a) to turning maneuvers (Gray, 1933d; Weihs, 1972), most movements being intermediate between the extremes (Webb, 1975c; Eaton et al., 1977).The same variation is presumed to occur at lower routine unsteady activity levels. Although behavior patterns are variable, mechanics are fairly stereotyped. Two major kinematic patterns can be distinguished, described as C- and S-starts. In the former case, the body is bent into a C- or L-shape during the fast start. In S-starts the body bends into a double flexure, with an S-shape (Webb, 1976). In both cases, mechanics can be divided into three stages as illustrated for a C-start in Fig. 7. First, the body anterior and posterior to the center of mass (about which propulsive forces act) moves laterally and the fish begins to accelerate, as measured by the motion of the center of mass (frames 1 to 3). This kinematic stage 1 is followed by a kinematic stage 2 (frames 3 to 8) when thrust is dominated by the tail moving in the opposite direction to stage 1, at a small angle of attack to the water. Acceleration behavior is completed by a variable kinematic stage 3. During this stage, fish adopt various behavior patterns, ranging from an unpowered glide through steady swimming to continued acceleration (Hertel, 1966; Weihs, 1972, 1973a; Webb, 1975~). C-starts and S-starts differ in an important respect. During a C-start, the bending of both the anterior and posterior parts of the body in opposite directions during kinematic stage 1 results in a turn being included as part of the acceleration. The angle of the turn is proportional to acceleration rate (Hertel, 1966; Weihs, 1973a). The turn results from unbalanced recoil forces generated by the tail. In contrast, the double body flexure in an S-start results in more balanced recoil forces such that a turn is not an unavoidable concomitant of acceleration. Turns of various magnitudes are part of normal fast starts in both patterns and are under behavioral control b y the fish. The duration of kinematic stages 1 and 2 is short (Table I). The duration of kinematic stage 3 will, of course, depend upon the behavior of the fish and will be of variable duration. In considering the mechanics of acceleration, net forces and hence net body movements are determined for the center of mass. Figure 8 shows time relations for distance covered, velocity, and acceleration rate for the center of mass of the trout in the sequence illustrated in Fig. 7. For this sequence, distance covered and velocity do not in-

3.

HYDRODYNAMICS: NONSCOMBROID FISH

223

Fig. 7. Tracings of the midline of a 33 cm trout accelerating from rest. Closed circles Biorlzeology 10,343-350. Copyright 1973 show the center of mass. (From Weihs, 1973~1, by Pergamon Press.)

crease linearly with time. The former increases exponentially while velocity first increases rapidly, but the rate of increase in velocity decreases with time. Acceleration rate is therefore not uniform for this sequence but decreases with time, presumably after an initial increase to a high level. The pattern of time relations for distance covered and velocity are similar among fish, but acceleration rate relations with time are variable. Acceleration rate often increases through kinematic stages 1 and 2. Distances covered during kinematic stages 1 and 2 are usually small (Table I) particularly at higher acceleration rates observed for fish (Section 111,B).Larger distances may be covered during kinematic

Table I Acceleration Performance of Fish During Fast Starts Using Various Methods

Species Salmo irideus Esox lucius Cyprinus carpio Scardinus erythrophthalamus Leuciscus leuciscus Salmo gairdneri Perca flauescens Thunnus albacares Acanthocybium solandri Salmo trutta Salmo gairdneri

Etheostoma caeruleum Cottus cognatus Esox sp Notropis cornutus Perca jlauescens Lepomis mcrochirus Lepomis cyanellus

Length (cm) 20.0 16.5-20.0 13.5

Mass (9)

-

-

-

22.0 18.5 29.0

-

-

108.9

98.0

-

113.1

-

33.0 9.6 20.4 29.6 38.7 14.3 18.4 19.5 6.2 8.2 21.7 10.7 15.5 15.3 8.0

7.2 79.0 270.8 561.9 27.3 48.41 69.90 2.25 6.94 41.80 11.18 33.48 64.50 8.18

Maximum acceleration rate (m/secP) 40 40 40 40 40 17 52 43 37 36 34 25 33 32 36 41 42 20 33 32 23 40 29

24 29 16

Mean acceleration rate (mlsecZ)

Initial velocity (cdsec)

Final velocity (cmisec)

Distance traveled (cm)

Duration (sec)

-

-

-

-

-

-

-

5 5 5

0.05 0.05 0.05

A A A

Gray ( 1953a) Gray ( 1953a) Gray (1953a)

-

-

-

-

-

5 5

0 -

260

-

346 204 1903 1305

1166 764 2082 1682

0.05 0.05 0.15 0.016 0.19 0.15 0.05 0.11

A

-

Gray ( 1953a) Gray ( 1953a) Hertel ( 1966) Gero (1952) Fierstine and Walters (1968)

-

16.0 15.1 15.7 17.6 12.9 8.0 10.6 10.3 6.1 10.4

11.0 9.3 12.3 8.4

-

0 0 0

0 0 0 0 0 0 0 0 0 0 0

-

-

-

-

153 167 181 285 121 133 158 89 77 156 114 115 131 67

3.5 7.3 9.5 16.3 5.4 8.4 8.7 2.4 3.4 8.5 3.8 5.2 5.9 2.9

-

0.071 0.078

Methodo

A

B C BC

Source

BC C C

Weihs ( 19734 Webb ( 1976)

C C C

Webb (19754 Webb ( 1977) Webb (unpublished) Webb (unpublished) Webb (unpublished) Webb (unpublished) Webb (unpublished) Webb (unpublished) Webb (unpublished) Webb (197%)

0.096 0.100 0.078 0.115 0.114 0.057 0.081 0.115 0.078 0.103 9.088 0.079

c

C C C

C C C

'(A) f = 2 . M . (B) f = ( U , - U , ) / t . (C) Acceleration rate calculated from moving point regressions from distance data obtained with high speed photography or similar. (BC)Velocity calculated as acceleration rate in (C), and acceleration calculated from the velocity data as in (B).f, acceleration rate; s, distance covered; t, time; Ut, initial velocity; U I , final velocity.

3.

225

HYDRODYNAMICS: NONSCOMBROID FISH

$

6

0.04

' 0

, , , , , , ,

0.08

,

042

2200

.:

016

1

1

I

I

I

I

I

'

0.b4

'

Oh8

'

&!2

'

80

60 -I

9 40 20

' 0

V

;

Y 0.04

0.08

042

6oo0

0.16

TIME

800-

-

046

sac

Fig. 8. Time relations for (A) distance covered, (B) velocity, and (C) acceleration rate calculated for the acceleration sequence shown in Fig. 7.

stages 1and 2 at low acceleration rates, such as that illustrated in Fig. 8, but velocities attained will be relatively lower. Kinematic stage 3 appears important in covering large distances at the high speeds attained at the end of kinematic stage 2. Kinematic stage 3 is also important for orientation control. In C-starts, the center of mass yaws in proportion to the acceleration rate so that the path of the fish after acceleration differs from the original body axis (Hertel, 1966; Weihs, 1973a). Correction must be made by fish accelerating toward a specific objective such as shelter or prey, and this correction occurs during kinematic stage 3. The problem is less important in' S-starts because recoil forces are better balanced. Acceleration activity can be considered to be behaviorally composed of two phases. A phase 1coincides with kinematic stages 1and 2 during which high velocities are acquired. Phase 2 coincides with kinematic stage 3 during which large distances are covered and reorientations are made (Webb, 1975~).

226

PAUL W. WEBB

B. Acceleration Performance Fish may attain very high but transitory speeds by the end of kinematic stage 2. Walters and Fierstine (1964)report speeds in excess of 20 Clsec for wahoo (lengths from 92 to 113 cm) and yellowfin tuna obtained maximum veloci(lengths from 56 to 98 cm). Webb (1975~) ties of the same magnitude for rainbow trout (8 = 14.3 cm) and green sunfish (C = 8.0 cm). However velocities attained by the end of stage 2 ranged from 20 Clsec to less than 2 Clsec for the same species, with mean values of 8.5 Clsec for the trout and 8.3 tlsec for the sunfish. Consequently, the very high speeds recorded are probably not representative of normal activity. This is also suggested by data for wahoo and skipjack when speeds ranged from 6 to 21 elsec, the mean being 12 Clsec (Fierstine and Walters, 1968). Maximum acceleration rates of fish vary from about 16 to 50 mIsec2 (Table I). Where sufficient observations have been made (Gero, 1952; Weihs, 1973a, Webb, 1975c, 1976, 1977) it has been found that acceleration rates are time dependent, so that the maximum values do not adequately describe overall performance. Mean values of acceleration rates for a fast start are much lower, ranging from 6 to 16 mlsec?.

C. Mechanics The mechanics of uniform acceleration for fishlike motions of small amp1itu.de have been described by Wu (19714. More appropriate, large amplitude movements have been described by Weihs (1972, 1973a) starting from Lighthill's (1971) treatment of large amplitude movements in steady swimming. A single theory is appropriate for acceleration activity ranging from fast starts to turning maneuvers (Weihs, 1972, 1973a).Weihs (1972, 1973a) has shown that large area is required for good acceleration performance. Large area is particularly important caudally, usually the caudal fin, where lateral movements are greatest. The importance of these morphological requirements for thrust have been demonstrated experimentally (Webb, 1977).

D. Work Done in Acceleration Work is done during acceleration against frictional drag and body inertia. Webb (19754 describes a simple method for obtaining first estimates of work done against frictional drag and hence for calculating total drag. The proportions of the two components making up total work done varies with acceleration rate and speed attained (Fig. 9). Inertial work is calculated as the increase in kinetic energy which will

3. HYDRODYNAMICS: NONSCOMBROID FISH

ACCELERATION RATE

-

227

cm/sec2

Fig. 9. Relationships for work done in accelerating from rest to a speed of 300 cm/sec (solid lines) and 600 cmlsec (dashed lines) at various acceleration rates for a hypothetical salmonid fish (C = 30 cm). Work done against frictional drag (EF), inertia ( E J ,and total work done (E) were calculated as described by Webb (1975a).

be constant for any acceleration rate when fish start from rest and accelerate to a given final speed. However, the work done against frictional drag in accelerating to a given final speed decreases with acceleration rate. Work done increases as final speed increases. Thus at high acceleration rates work done is mainly against body inertia. The relationships between the two drag components indicate that it is energetically advantageous to accelerate at high acceleration rates, as commonly observed.

E. Drag Reduction 1. BODY SHAPE Specialized predators, for example, Esox, with unsteady lunging behavior typically have elongated cylindrical bodies. This body shape

228

PAUL W. WEBB

minimizes body surface area and added mass while maximizing body volume for maximum muscle mass. The body shape consequently minimizes frictional and added mass drag during acceleration.

2. BEHAVIOR Houssay (1912) described the normal pattern of fish locomotion as bursts of activity followed by passive glides. Weihs (1973d, 1974) has shown that this bursvglide behavior can reduce the energy required to cover a given distance in a given time. In comparison with swimming at uniform velocity, bursdglide behavior can reduce thrust required up to 50%.

IV. UNSTEADY VERSUS STEADY PROPULSION A. Mechanics and Structure The structural requirements for high thrust in steady swimming differ from those for unsteady swimming. The former requires large trailing edge depth, concentrated primarily at a reentrant caudal fin. In contrast, high unsteady activity requires a large area concentrated caudally where lateral movements are largest. However, a large caudal area is associated with greatly increased drag in steady swimming (Webb, 1973a) resulting from boundary layer effects of propulsive movements (Lighthill, 1971). In the absence of morphological specialization, thrust increases with the trailing edge virtual mass, but drag increases with area in similar proportion to increased thrust. As a result, there is little change in steady cruising performance with an increase in tail depth and area. Thus, morphological requirements for high steady swimming must be qualified as large trailing edge depth without equivalent increase in area and related drag. These requirements are then apparently not consistent with high unsteady performance, such that morphological requirements for high unsteady activity and for high steady activity are mutually exclusive. As a result, body shapes that maximize acceleration do so at the cost of steady performance (Fig. 10). This does not present a problem for fish specialized for either high unsteady or high steady propulsion strategies. Fish specialized for lunging habits, for example Cottids or Esocids, tend to maximize area along the whole length of the body, or to maximize caudal area. However, their routine locomotory habits are commonly “hovering” in the

3.

229

HYDRODYNAMICS: NONSCOMBROID FISH

r

h

m

A

B

C

D

E

F

Fig. 10. Relative performance of various lateral body profiles typical of fish in steady swimming and fast start acceleration; Performance is normalized with respect to that of a reference case, rainbow trout. (Data from Webb, 1977.)

water using alternate median fin propulsion (e.g., Esocids) or habits may be sedentary (Cottids). Fish specialized for cruising show well scooped out (swept back) caudal fin shapes leading to the specialized lunate tail of carangids and scombroids. Such caudal fin structure achieves a large'trailing edge depth with small tail area. The problem of unsteady versus steady performance is greatest for locomotor generalists when a balance of both activities is required. Because of the difference in morphological characteristics for the two types of performance, such generalists must be capable of modifying their lateral body profiles. This, in turn, is only possible with collapsible median fins which have evolved in bony fish. Such fish characteristically expand their median fins before accelerating (Eaton et al., 1977; Webb, 1977). Fish which are unable to significantly modify fin depth and area (notably Elasmobranchs) are not locomotor specialists.

B. Performance During and following acceleration, speeds are attained that are very much higher than normally attributed to fish. On theoretical energetics grounds (Webb, 1975a) a 10 cm salmonid might obtain a maximum (transitory) speed of 400 cm/sec (40t'/sec). Maximum speeds are expected to increase with size while specific speeds decrease as for steady swimming, so that a 50 cm salmonid might obtain a

230

PAUL W. WEBB

maximum unsteady speed of 550 cm/sec (11 t/sec). Maximum unsteady specific speeds of 20 e/sec have been reported for fish (Section III,B), but mean values are usually lower (Table I). Maximum unsteady speeds can thus exceed maximum steady speeds. The reason for this difference is that energy required is relatively small for a fish to accelerate rapidly (in usually 50.1 sec) to high speeds but covering a relatively small distance. I n comparison the energy required is high to sustain high speeds, for 21 sec, when a large distance is covered. For example, assuming swimming drag is four times theoretical drag, energy expended by a 30 cm salmonid swimming at 300 cm/sec for 1 sec is expected to be of the order of lo* ergs. According to theory (Webb, 1975a) the energy expended in accelerating at 5000 cm/sec2to the same speed is of the order of 1.8x los ergs. The difference is extremely large and is too large to be accounted for by error in assumptions made by Webb’s model for work done by accelerating fish. Consequently, it is important to distinguish between steady and unsteady performance in calculating energy expenditure. If this distinction is not made, excessive energy requirements may be calculated for the very high speeds observed for fish and cetaceans resulting in typical paradox situations analogous to Gray’s Paradox.

V. MECHANICS OF MEDIAN AND PAIRED FIN PROPULSION Breder (1926) describes a large number of morphological types of paired and noncaudal median fin propulsion systems. For mechanical purposes Breder’s classification can be simplified in the same way as body/caudal fin modes by differentiating between use of propulsive waves with short or long wavelengths. Fins with short waves relative to the fin base are found in the balistiform mode (long paired dorsal and ventral fins), the rajiform mode (long paired pectoral fins), the amiiform mode (long dorsal fins), and the gymnotiform (long ventral fins). The diodontiform mode includes fish using relatively short pectoral fins but which swim by means of short wavelength movements passed along that fin. Long wavelengths are found in the labriform mode with relatively short-based pectoral fins. The tetraodontiform mode uses short paired dorsal and ventral fins, probably using long wavelength propulsive movements. The mechanics of most non-body/caudal fin modes have only been discussed qualitatively ( Breder, 1926; Harris, 1937, 1953; Breder and Edgerton, 1943; Gray, 1968; Stickney et al., 1973; Webb, 1973a, 1975b), with the exception of models for skatelike movements in the

3.

HYDRODYNAMICS: NONSCOMBROID FISH

231

rajiform mode (Kelly, 1961; Wu, 1961, 1971a; Siekmann, 1962, 1963). It is important to recognize that median and paired fins are often used primarily for maneuver rather than steady swimming. As a result movements are variable, and repeatability in experimental situations is problematic (Harris, 1937) except under unusual circumstances (Webb, 197313, 1975b).The paucity of data for these swimming modes is probably attributable to these problems of replicating experimental observations. The mechanics of median and paired fin modes can be discussed in terms of movements of propulsive segments as for bodyhaudal fin modes. Harris (1937)and Gray (1968)considered propulsive segments to generate thrust from a combination of resistive and reactive forces throughout a propulsive cycle in similar fashion to a fan. Alternatively, long fin modes may be compared to anguilliform body/caudal fin propulsion. However, long fin propulsion is undoubtedly less efficient than anguilliform propulsion because there is no increase in amplitude along the fin (Breder and Edgerton, 1943; Lissman, 1961).As a result successive propulsive waves will probably interfere because there would be no sequential increase in water momentum. Then more posterior segments may contribute relatively little to thrust but may increase energy losses. Skates and rays apparently include just over one wavelength within the length of the pectoral fins (Breder, 1926). The shape of the fin also tends to be triangular and the amplitude of the propulsive wave apparently increases to reach a maximum coinciding with the apex of the fin. Amplitude then decreases across the tapering posterior system. This system may be functionally different from most other long fins for amplitude functionally increases to the maximum width of the fin at the apex. Subsequently, amplitude decreases posteriorly but over a tapering trailing portion of the fin. Forces will be generated not only parallel to, but also normal to the fin base because lateral movements of the fin are restricted proximally by articulation with the body. Such normal forces relative to the fin base will not contribute to thrust when the fin base is parallel to the body axis but will increase energy losses. In some fish, for example, Balistes, the paired median fins are oriented backward and forces parallel and normal to the fin base will contribute to thrust. Forces normal to the fin base will generate yawing couples for pectoral fins and pitching couples for median fins. I n the majority of fin propulsion modes the occurrence of fins in pairs will cancel out these forces for symmetrical fin movements. Asymmetric movements will be important in powered maneuver. The situation is uncertain where fins are not paired, for example, Gymnarchus orAmia. These fish have not

232

PAUL W. WEBB

been reported to rise or sink in the water when swimming or gliding to rest, so buoyancy compensation for normal propulsive forces seems unlikely. Such fish do include relatively large numbers of propulsive wavelengths within the fin lengths (Breder, 1926; Breder and Edgerton, 1943; Lissman, 1961). It is possible that some movements, for example, a downward swing at the end of a cycle, might generate forces opposite to those during the remainder of the cycle. Then a large number of wavelengths would result in a net zero force component normal to the fin base. No data are available to confirm this hypothesis. Short fin propulsion mechanics are probably more similar to those of lifting surfaces comparable to the scombroid caudal fin or wings (Lighthill, 1969; Webb, 1975~).All such fins are paired and yawing and pitching forces would cancel out. Continuous steady propulsion is possible with short median fins, for example Mola, but usually the beating of short fins results in discontinuous movements. Fish recoil vertically during abduction and adduction movements and slip backward between these active stages (Webb, 1973b). These movements are associated with increased energy losses. The thrust developed and mechanical efficiency for fin modes are not well known. The maximum mechanical efficiency in generating thrust is of the order of 0.60-0.65 for Cymatogaster aggregata swimming by means of short pectoral fins (Webb, 197%). This efficiency is somewhat lower than expected for body and caudal fin propulsion at steady swimming speeds (Wu, 1971d; Webb, 1975a).The efficiency of long-based fin propulsion modes with short propulsive wavelengths is expected to be very much lower because of increased energy losses from non-thrust-enhancing segment interactions and as a result of forces normal to the fin base.

REFERENCES Alexander, R. M . (1968). “Animal Mechanics,” Biology Series, 346 pp. Sidgewick & Jackson, London. Alexander, R. McN., and Goldspink, G . (1977). “Mechanics and Energetics of Animal Locomotion,” 346 pp. Wiley, New York. Aleyev, Y. G., and Ovcharov, 0. P. (1969). Development of vortex forming processes and nature of the boundary layer with movement of fish. Zool. Zh. 48, 781-790. Aleyev, Y. G., and Ovcharov, 0. P. (1971).The role of vortex formation on locomotion of fish, and the influence of the boundary between two media on the flow pattern.Zool. Zh. 50,228-234. Allan, W. H. (1961). Underwater flow visualization techniques. US.Nau. Ord. Test Stn., Tech. Publ. No. 2759, 28 pp.

3.

HYDRODYNAMICS : NONSCOMBROID FISH

233

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. Bainbridge, R. (1961). Problems of fish locomotion. Symp. Zool. SOC. London 5,1%32. Bainbridge, R. (1963). Caudal fin and body movements in the propulsion of some fish.J. Exp. Biol. 40, 23-56. Beamish, F. W. H. (1968). Glycogen and lactic acid concentrations in Atlantic cod (Cadus morhua) in relation to exercise.J. Fish. Res. Board Can. 25, 837-851. Belyayev, V. V., and Zuyev, G. V. (1969).Hydrodynamic hypothesis of school formation in fishes.J. Ichthyol. (USSR) 9,578-584. Blaika, P., Volf, M., and Cepela, M. (1960).A new type of respirometer for the determination of metabolism of fish in an active state. Physiol. Bohemosloo. 9, 553-558. Blight, A. R. (1977). The muscular control of vertebrate swimming movements. B i d . Reu. 52, 181-218. Bone, Q. (1972). Buoyancy and hydrodynamic functions of the integument in the castor-oil fish, Reoettus pretiosus (Pisces: Gempylidae). Copeia pp. 78-87. Bone, Q., and Howarth, J. V. (1966). “Report to Council 1966-67,” p. 19. Mar. Biol. Assoc. U.K., Plymouth, England. Borelli, G. A. (1680). “De motu animalium e x principio mechanic0 statico.” Rome. Breder, C. M. (1926). The locomotion of fishes. Zoologica (N.Y.) 4, 159-256. Breder, C. M. and Edgerton, H. E. (1943).An analysis ofthe locomotion ofthe seahorse, Hippocampus, by means of high speed cinematography. Ann. N.Y. Acad. Sci. 43, 145- 172. Brett, J. R. (1963).The energy required for swimming by young sockeye salmon with a comparison of the drag force on a dead fish. Trans. R. SOC.Can. 1, Ser. IV, 441-457. Brett, J. R. (1964). The respiratory metabolism and swimming performance of young sockeye sa1mon.J. Fish. Res. Board Can. 21, 1183-1226. Brett, J. R. (1965).The swimming energetics of salmon. Sci. Am. 213, 80-85. Brown, C. E., and Muir, B. S. (1970). Analysis of ram ventilation of fish gills with application to skipjack tuna.]. Fish. Res. Board Can. 27, 1637-1652. Denil, G . (1936). “La Mbcanique du Poisson de RiviBre; Qualitb Nautique du Poisson; ses MBthodes Locomotrices; ses Capacitbs; ses Limites; Resistance du Fluide; Effet de la Vitesse, d e la Pente; Resistance du Seuil,” Ann. Trao. Publ. Belg. 36, 1-395. DuBois, A. B., Cavagna, G. A., and Fox, R. S. (1974). Pressure distribution on the body surface of a swimming fish.J. Exp. Biol. 60, 581-591. Eaton, R. C., Bombardieri, R. A., and Meyer, D. L. (1977). The Mauthner-initiated startle response in teleost fish.J. E x p . Biol. 66, 65-81. Fierstine, H . L., and Walters, V. (1968).Studies of locomotion and anatomy of scombroid fishes. Mem. South. CaZiJ Acud. Sci. 6, 1-31. Fry, F. E. J . (1957).The aquatic respiration of fish. “Physiology of Fishes” (M. E. Brown, ed.), Vol. 1, pp. 1-63. Academic Press, New York. Fry, F. E. J. (1971). The effect of environmental factors on the physiology of fish. In “Fish Physiology” (W. S. Hoar and D. J. Randall, eds.), Vol. 6, pp. 1-98. Acadeillic Press, New York. Gadd, G . E. (1952). Some hydrodynamic aspects of swimming in snakes and eels. Philos. Mag. 58, 663-760. Gero, D. R. (1952).The hydrodynamic aspects of fish propulsion. Am. Mus. Nooit. No. 1601, pp. 1-32. Gibbs-Smith, C. H. (1962). “Sir George Cayley’s Aeronautics 17961855,’’ pp. 41-42. H. M. Stationary Off. Sci. Mus., London. Gray, J. (19334. Studies in animal locomotion. I. The movement of fish with special reference to the ee1.J. E x p . Biol. 10, 88-104.

234

PAUL W. WEBB

Gray, J. (1933b). Studies in animal locomotion, 11. The relationship between the waves of muscular contraction and the propulsive mechanisms of the eel.]. Exp. Biol. 10, 386-390. Gray, J. (1933~).Studies in animal locomotion. 111. The propulsive mechanism of the whiting.]. Exp. Biol. 10, 391-400. Gray, J. (1933d). Directional control of fish movement. Proc. R. Soc., Ser. B 113, 115125. Gray, J . (1936). Studies in animal locomotion. VI. The propulsive powers of the dolphin. ]. E x p . Biol. 13, 192-199. Gray, J. (1953a). The locomotion of fishes. In “Essays in Marine Biology” (S. M. Marshal and P. Orr, eds.), Elmhirst Memorial Lectures, pp. 1-16. Oliver & Boyd, Edinburgh. Gray, J. (1953b). Undulatory propulsion.]. Microsc. Sci. 94, 551-578. Gray, J. (1957). How fish swim. Sci. Am. 197, 48-54. Gray, J. (1968). “Animal Locomotion,” World Naturalist Series, 479 pp. Weidenfeld & Nicolson, London. Harris, J. E. (1936). The role of fins in the equilibrium of swimming fish. 1. Wind-tunnel tests on a model of Mustelus canis (Mitchell).]. E x p . Biol. 13, 476-493. Harris, J. E. (1937). The mechanical significance of the position and movements of the paired fins in the teleostei. Tortugas Lab. Pap. No. 31, pp. 173-189. Harris, J. E. (1953). Fin patterns and mode of life in fishes. I n “Essays in Marine Biology” (S. M. Marshal and P. Orr, eds.), Elmhirst Memorial Lectures, pp. 17-28. Oliver & Boyd, Edinburgh. Hertel, H. (1966). “Structure, Form and Movement,” 251 pp. Reinhold, New York. Hill, A. V. (1950). The dimensions of animals and their muscular dynamics. Sci. Prog. (London) 38,209-230. Houssay, S. F. (1912). “Forme, Puissance et StabilitC des Poissons,” 372 pp. Herman, Paris. Hoyt, J. W. (1975). Hydrodynamic drag reduction due to fish slimes. In “Swimming and Flying in Nature” (T. Y. T. Wu, C. J. Brokaw, and C. Brennen, eds.), pp. 653-672. Plenum, New York. Hudson, R. C. L. (1973). On the function ofthe white muscles in teleosts at intermediate swimming speeds.]. Exp. Biol. 58, 509-522. Hunter, J. R., and Zweifel, J. R. (1971). Swimming speed, tail beat frequency, tail beat amplitude and size in jack mackerel, Trachurus symmetricus, and other fishes. U.S. Fish Wildl.Serv., Fish. Bull. 69, 253-266. Johnson, I. A., and Goldspink, G. (1973a). Quantitative studies of muscle glycogen utilization during sustained swimming in crucian carp (Carassius carassius).].E x p . Biol. 59, 607-615. Johnson, I. A., and Goldspink, G. (1973b). A study of glycogen and lactate in the myotomal muscles and liver of the coalfish (Gadus uirens) during sustained swimming.]. Mar. Biol. Assoc. U.K. 53, 17-26. Johnson, I. A., and Goldspink, G. ( 1 9 7 3 ~ )A. study of the swimming performance of the crucian carp (Carassius carassius) in relation to the effects of exercise and recovery on biochemical changes in the myotomal muscles and liver. J . Fish. Biol. 5, 249260. Keenleyside, M. H. S. (1955). Some aspects of the schooling behavior of fish. Behaviour 8, 183-248. Kelly, H. R. (1961). Fish propulsion hydrodynamics. Deu. Mech. 1, 442-450. Kempf, G., and Neu, W. (1932). Schleppversuche mit Hechten sur Messung des Wassersiderstandes. Z. Vergl. Physiol. 17, 353-364.

3. HYDRODYNAMICS: NONSCOMBROID FISH

235

Kent, J. C., DeLacy, A., Hirota, T., and Bates, B. (1961). Flow visualization and drag about a swimming fish. Tech. Rep., 23 pp. Fish. Res. Inst. College Fish., Univ. of Washington, Seattle. Kerr, S. R. (1971).A simulation model of lake trout growth.]. Fish. Res. Board Can. 28, 815-819. Kliashtorin, L. B. ( 1973).The swimming energetics and hydrodynamic characteristics of actively swimming fish (in Russ.). Express l n f . pp. 1-19. Lang, T. G. (1966). Hydrodynamic analysis of cetacean performance. In “Whales, Dolphins and Porpoises” (K. S. Norris, ed.), pp. 410-432. Univ. of California Press, Berkeley. Lighthill, M. J. (1960).Note on the swimming ofslender fish.]. Fluid Mech. 9,305-317. Lighthill, M. J. ( 1969). Hydromechanics of aquatic animal propulsion. Annu. Reu. Fluid Mech. 1,413-446. Lighthill, M. J. ( 1970). Aquatic animal propulsion of high hydromechanical efficiency.]. Fluid Mech. 44,265-301. Lighthill, M. J. (1971).Large-amplitude elongated-body theory of fish locomotion. Proc. R. Soc., S e r . B 179, 125-138. Lissman, H. W. (1961).Zoology, locomotory adaptions and the problem of electric fish. I n “The Cell and the Organism” (J. A. Ramsay and V. B. Wigglesworth, eds.), pp. 301-307. Cambridge Univ. Press, London and New York. Magnan, A. ( 1930). Les caractkristiques gCometriques et physiques des poissons. Ann. Sci. Nut., Zool. Biol. Anim. 13, 1971-1981. Magnuson, J. J. (1970). Hydrostatic equilibrium of Euthynnus afinis, a pelagic teleost without a gas bladder. Copeia pp. 56-85. Magnuson, J. J,, and Prescott, J. H. (1966). Courtship, feeding and miscellaneous behaviour of Pacific bonito (Sardn chiliensis).Anirn. Behao. 14,54-67. Marey, E. J. (1874).“Animal Mechanism,” 283 pp. Masson, London. Marey, E. J. (1895).“Movement,” 323 pp. Masson, London. Marr, J. (1959).A proposed tunnel design for a fish respirometer. Tech. Memo. No. 58-3, pp. 1-13. Pac. Nav. Lab., Esquimalt, B.C. Newnian, J. N. (1973).The force on a slender fish-like b0dy.J. Fluid Mech. 58,689-702. Newman, J. N., and Wu, T. Y. (1973). A generalized slender-body theory for fish-like fonns.]. Fluid Mech. 57, 673-693. Nursall, J. R. (1962). Swimming and the origin of paired fins. Am. Zool. 2, 127-141. Nursall, J. R. (1973). Some behavioral interactions of spottail shiners (Notropis hudsonius), yellow perch (Perca faloescens), and northern pike (Esox lucius). J . Fish. Res. Board Can. 30, 1161-1178. Osbome, M. F. M. (1961).Hydrodynamic performance of migratory sa1mon.J. Exp. Biol. 38,365-390. Ovchinnikov, V. V. (1966).Turbulence in the boundary layer as a method of reducing the resistance of certain fish on movement. Biophysics (USSR) 11, 186-188. Parry, D. A. (1949).The swimming of whales and a discussion of Gray’s Paradox.]. E x p . Biol. 26, 24-34. Pedley, T. J. (ed.) (1977). “Scale Effects in Animal Locomotion,” 545 pp. Academic Press, New York. Pershin, C. V. (1970).Kinematics of dolphin propulsion (in Russ.). Bionika pp. 31-36. Pettigrew, J. B. (1873).“Animal Locomotion,” 264 pp. King, London. Prandtl, L., and Tietjens, 0. G. (1934a). “Fundamentals of Hydro- and Aeromechanics,” 270 pp. Dover, New York. (New Ed., 1957.) Prandtl, L., and Tietjens, 0. G. (1934b). “Applied Hydro- and Aerornechanics,” 311 pp. Dover, New York. (New Ed., 1957.)

236

PAUL W. WEBB

Pritchard, A. W., Hunter, J. R., and Lasker, R. (1971). The relation between exercise and biochemical changes in red and white muscle and liver in the jack mackerel, Trachurus symmetricus. U S . Fish Wildl. Seru., Fish. Bull. 69,379-386. Pyatetskiy, V. E. (1970a). Kinematic characteristics of some fast marine fish (in Russ.). Bionika pp. 11-20. Pyatetskiy, V. E. (1970b). Hydrodynamic characteristics of swimming of some fast marine fish (in Russ.). Bionika pp. 20-27. Richardson, E. G . (1936).The physical aspects of fish locomotion.]. E x p . Biol. 13,63-74. Rosen, M. W. (1959). Water flow about a swimming fish. U . S . Nav. Ord. Test Stn., Tech. Publ. No. 2298, pp. 1-96. Rosen, M. W., and Comford, N. E. (1971). Fluid friction of fish slimes. Nature (London) 234, 49-51. Schlichting, H . (1968).“Boundary Layer Theory,” 6th Ed., 747 pp. McGraw-Hill, New York. Siekmann, J. (1962). Theoretical studies of sea animal locomotion. Part 1. Zng.-Arch. 31, 214-228. Siekmann, J. (1963).Theoretical studies of sea animal locomotion. Part 2,Zng.-Arch.32, 40-50. Smit, H. (1965). Some experiments on the oxygen consumption of goldfish (Carassius auratus L) in relation to swimming speed. Can.J . Zool. 43, 623-633. Smit, H., Amelink-Koutstaal, J. M., Vijverberg, J., and von Vaupel-Klein, J. C. (1971). Oxygen consumption and efficiency of swimming goldfish. Comp. Biochem. Physiol. A 39, 1-28. Stevens, G . A. (1950). Swimming of dolphins. Sci. Prog. (London)38, 524-525. Stickney, R. R., White, D. B., and Miller, D. ( 1973).Observations of fin use in relation to feeding and nesting behavior in flat fishes (Pleauronectiformes) Copeia pp. 154156. Sundnes, G . (1963). Energy metabolism and migration of fish. Znt. Comm. Northwest Atl. Fish., Environ. Symp., Spec. Publ. No. 6, pp. 743-746. Taylor, G. (1952).Analysis of the swimming of long narrow animals. Proc. R. Soc., Ser. A 214, 158-183. Tucker, V. A. (1973). Bird metabolism during flight: Evaluation of a theory.]. E x p . Biol. 58,689-709. Vlymen, W. J. (1974). Swimming energetics ofthe larval anchovy. U.S. Fish Wildl. Seru., Fish. Bull. 72, 885-899. von Holste, E., and Kuchemann, D. (1942). Biological and aerodynamic problems of animal flight. J . R. Aeronaut. SOC.4 6 , 4 4 5 4 . von Mises, R. (1945). “Theory of Flight,” 629 pp. Dover, New York. (New Ed., 1959.) Walters, V. (1962). Body form and swimming performance in scombroid fishes. Am. Zool. 2, 143-149. Walters, V. (1963). The trachypterid integument and an hypothesis on its hydrodynamic function. Copeia pp. 260-270. Walters, V. (1966). The “problematic” hydrodynamic performance of Gero’s great barracuda, Sphyraena barracuda (Walbaum). Nature (London)212,215-216. Walters, V., and Fierstine, H. L. (1964).Measurements of swimming speeds of yellowfin tuna and wahoo. Nature (London)202,208-209. Webb, P. W. (1971a). The swimming energetics of trout. I) Thrust and power output at cruising speeds.]. E x p . Biol. 55, 489-520. Webb, P. W. (1971b). The swimming energetics of trout. 11) Oxygen consumption and swimming efficiency.]. E x p . Biol. 55, 521-540. Webb, P. W. (1973a). Effects of partial caudal fin amputation on the kinematics and

3.

HYDRODYNAMICS: NONSCOMBROID FISH

237

metabolic rate of underyearling sockeye salmon. (Oncorhynchus nerka) at steady swimming speeds. J. E x p . Biol. 59, 565-581. Webb, P. W. (1973b). Kinematics of pectoral fin propulsion in Cymatogaster aggregata. J. Exp. Biol. 59, 697-710. Webb, P. W.(1975a). Hydrodynamics and energetics of fish propulsion. Bull. Fish. Res. Board Can. No. 190, 159pp. Webb, P. W. (1975b). Efficiency ofpectoral fin propulsion in Cymatogaster aggregata. In “Swimming and Flying in Nature” (T. Y. T. Wu, C. J. Brokaw, and C. Brennen, eds.), pp. 573-584. Plenum, New York. Webb, P. W. (1975~).Acceleration performance of rainbow trout Sulmo gairdneri (Richardson) and green sunfish Lepomis cyanellus (Rafinesque). J . E x p . B i d . 63, 451-465. Webb, P. W. (1976). The effect of size on the fast-start performance of rainbow trout, Salmo gairdneri, and a consideration of piscivorous predator-prey interactions. J. E x p . Biol. 65, 157-177. Webb, P. W. (1977). Effects of median-fin amputatiop on fast-start performance of rainbow trout (Salmo gairdneri).J.E x p . Biol. 68, 123-135. Weihs, D. (1972). A hydrodynamicd analysis of fish turning manoeuvres. Proc. R. SOC., Ser. B . 182,59-72. Weihs, D. (1973a). The mechanism of rapid starting of slender fish. Biorheology 10, 343-350. Weihs, D. (1973b). Hydromechanics of fish schooling. Nature (London)241,290-291. Weihs, D. (19734. Optimal cruising speed for migrating fish. Nature (London) 245, 48-50. Weihs, D. (1973d). Mechanically efficient swimming techniques for fish with negative buoyancy.]. Mar. Res. 31, 194-209. Weihs, D. (1974). Energetic advantages of burst swimming of fish. J. Theor. B i d . 48, 215-229. Weis-Fogh, T. (1973). Quick estimates of flight fitness in hovering animals, including novel mechanisms for life production.]. E x p . Biol. 59, 169-230. Wu, T. Y. (1961). Swimming of a waving p1ate.J. Fluid Mech. 10,321-344. Wu, T. Y. (1971a). Hydromechanics of swimming propulsion. Part 1. Swimming of a two-dimensional flexible plate at variable forward speeds in an inviscid fluid. J . Fluid Mech. 46,337-355. Wu, T. Y. (1971b). Hydromechanics of swimming propulsion. Part 2. Some optimum shape prob1ems.J. Fluid Mech. 46, 521-544. Wu, T. Y. (19714. Hydromechanics.of swimming propulsion. Part 3. Swimming and optimum movements of slender fish with side fins.J. Fluid Mech. 46,545-568. Wu, T. Y. (1971d).Hydromechanics of swimming fishes and cetaceans.Ado. Appl. Math. 11, 1-63. Wu, T. Y., and Newman, J. N. (1972). Unsteady flow around a slender fish-like body. Proc. Int. Symp. Directional Stabil. Control Bodies Mooing Water, London Inst. Mech. Eng. Publ., Pap. No. 7. pp. 33-42. Wu, T. Y. T., Brokaw, C. J., and Brennen, C. (eds.) (1975). “Swimming and Flying in Nature,” 1005 pp. Plenum, New York. Yuen, H. S. H. (1966). Swimming speeds of yellowfin and skipjack tuna. Trans. Am. Fish. SOC. 95,203-209. Zuyev, G. V., and Belyayev, V. V. (1970). An experimental study of the swimming of fish in groups as exemplified by the horse mackerel (Trachurus mediterruneus ponticus Aleev).J. Ichthyol. (USSR) 10, 545-549.

This Page Intentionally Left Blank

LOCOMOTION BY SCOMBRID FISHES: HYDROMECHANICS, MORPHOLOGY, AND BEHAVIOR J O H N J . MAGNUSON I. Introduction ................................................... 11. Observed Swimming Speeds.. . . . . . . . . . . . , . . . , . . . . . . . . . . . . . . . . . A. Sustained Swimming . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . B. Prolonged Swimming . . . . . . . . . . . . . . , . . . . . . . . . . . . . . . . . . . . . . . . C. Burst Swimming .... .. .. .. ........ . ... . , ,....... ........... . 111. General Considerations of Swimming Equilibria . . . . . . . . . . . . . . . . . IV. Weight, Buoyancy, Hydrodynamic Lift, and Prediction of Sustained Speeds.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . , . . . . . A. Mechanisms of Hydrostatic Equilibrium . . . . . . . . . . . . . . . . . B. The Lifting Surfaces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . , . . . . . . . . . C. Surface Loading of Lifting Surfaces . . . . . . . . . . . . , . . . , . . . . . . . . . . . . .. . . . . .. , . D. Predictions of Sustained Speed . . . . . . V. Resistance to Fonvard Movement.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . A. Components of Drag . . . . . . ts . . . , . . , . . . . . . . . . . . . . . B. Distribution of Drag among C. Adaptations for Drag Reduction.. . . . . . . . . . . . . . . . . . . . . . , . . . . . . . .. . . .. .. . . . . . . .. . . D. Summary . . . . . . . . . . . . . . . . . . . ... .. .. . .... .. ... .. . VI. Thrust Production . . . . . . . . . . . . . . . . . A. Anatomy . ... .. .. ... . . . .. .. .. .. . . .. . . . . . . . . . . . .. ............ B. Movements and Postures of the Caudal Fin ... . ........... C. Theory and Magnitude of Thrust . . . . . . . . . . . . . . . . . . . . . . . . . VlI. The Locomotory System . . , . . . . . . . . . . . . , . . . . . . . . . . . . . . . . . . . .. . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

.

..

.

.

..

..

.

.. .. . .

.

.

. ..

...

.

240 241 241 247 248 250 251 252 256 262 265 267 268 274 277 288 288 289 295 301 305 308

Symbols A

AR b b,

amplitude of trailing edge of caudal movement (Fig. 13) in centimeters aspect ratio of hydrofoil = b / c = b2/S span of caudal fin (Fig. 13) in centimeters span of pectoral fins (Fig. 4) in centimeters 239 FISH PHYSIOLOGY. VOL VII Copyright @ 1978 b y Academic Pre% Inc All right, of reproduction in m y form re5erved ISBN 0-12 350407-4

JOHN J. MAGNUSON

buoyant force in dynes chord length of hydrofoil in centimeters coefficient of friction plus form drag coefficient of friction drag coefficient of lift coefficient of lift for keel coefficient of lift for pectoral fins coefficient of thrust drag in dynes frequency of caudal fin beats in number per second ratio of gliding to swimming drag = ( 1 - qP)/l fork length in centimeters lift in dynes mass of fish in grams Reynolds number surface area of caudal fin, one side only (Fig. 13), in square centimeters projected surface area of keel (Fig. 4) in square centimeters total lifting area of pectoral fins (Fig. 4) in square centimeters wetted surface area of part in question in square centimeters weight of fish in dynes maximum transverse velocity of the trailing edge of caudal fin measured at axis of tail beat in centimeters per second root mean square transverse velocity or the average transverse velocity of trailing edge of caudal fin in centimeters per second swimming speed in centimeters per second maximum angle of attack of caudal fin measured when crossing axis of tail beat (Fig. 13E) in degrees or radians hydromechanical efficiency of the caudal fin or proportion of total thrust that contributes directly to forward movement proportion of thrust force lost as drag induced from caudal movement feathering angle (Fig. 13E) in degrees or radians density of environment in grams per cubic centimeter density of fish in grams per cubic centimeter angular velocity in radians per second angle of sweepback (Figs. 4,13) ofcaudal fin (Eq.28) in degrees

I. INTRODUCTION Fishes of the family Scombridae (tuna, bonito, and mackerel) swim quickly and swim continuously. These two unique features of their locomotion have stimulated two recurring questions in the literature:

4.

LOCOMOTION BY SCOMBFUD FISHES

24 1

Why do they never stop swimming, and how can they swim so fast? Partial answers to these and similar questions are to be found in considerations of hydromechanics and of the adaptations related to locomotion that seem to dominate the anatomy, behavior, and physiology of scombrids. Their continuous shimming is closely tied to hydromechanics because they are negatively buoyant pelagic fishes that oppose their weight with lift generated b y their extended pectoral fins as they swim forward (Magnuson, 1966,1970,1973).Their shape provides evidence to the idea that they are the most highly evolved teleosts in respect to minimizing resistance to forward movement (Walters, 1962; Aleev, 1963; Hertel, 1966). Their carangiform mode of locomotion with lunate tail (Lighthill, 1969, 1970, 1975; Wu, 1971; Chopra, 1975; Webb, 1975; Aleyev, 1977; Chopra and Kambe, 1977) concentrates all thrust at the caudal fin and represents the extreme of adaptations for fast and efficient swimming. While scombrids may best exemplify the fast and continuous swimmer of the pelagic ocean, they are not entirely unique. Mackerel sharks (Lamnidae) such as the shortfin mako, Zsurus oryrinchus, and white, Carcharodon carcharias, have the same general characteristics of both fast and continuous swimming as do the dolphin fishes, Coryphaenidae. In addition the mode of fast swimming with stiff body and rigid lunate tail is shared with other teleosts such as billfishes (Istiophoridae), the swoTdfish (Xiphias gladius), and with many marine mammals such as dolphins, porpoises, and whales. Many of these are not obligatory swimmers but can stop and float in the water. Similarly, many of the continuous swimmers do not have lunate tails nor swim with stiff body movements-some jacks (Carangidae), herring (Clupeidae), the bluefish (Pomatomus saZtaZtrix) to name a few, as well as most sharks. My purpose is to explore locomotion by scombrids, a group at the edges of two lines of specialization in regard to locomotionswimming quickly and swimming continuously. Swimming performance, hydromechanics, and related anatomy and behavior are treated.

11. OBSERVED SWIMMING SPEEDS A. Sustained Swimming Scombrids do not rest in the sense of not swimming (Magnuson, 1973). The minimum speed published for adults is for the wahoo (Acanthocybium solanderi) at 41 cmlsec (0.33t'lsec). Minimum speeds

242

JOHN J. MAGNUSON

observed for seven other species are usually greater than 1 t//sec and for two species minimum speeds are near 2 k'/sec (Fig. 1A; Table I). Roberts (1975) observed 2-12 cm long Atlantic mackerel (Scomber scombrus) holding stationary positions in an aquarium. On the other hand, I have observed the young, ca. 2-3 cm long, of other scombrids swimming actively in aquaria. Sccmbrids in large shoreside tanks and at sea settle into a mode of relatively steady continuous travel. Swimming is volitional and spontaneous. The mean speed of travel for eleven species (Table I; Fig, 1A) ranges from 22 to 280 cm/sec and 0.33 to 2.2 e/sec, with Reynolds numbers of from 4.2 x lo4 to 6.3 x lo6.The Reynolds number ( R L )is the product of the fish's length (8) and speed (U) divided by the kinematic viscosity of water: RL=

eu 0.01

The highest mean Reynolds number for sustained swimming (6.3 x 1@), observed for bluefin tuna over 200 cm long, is just above the range over which flow would be expected to be laminar over a rigid streamlined body (Webb, 1975). With seven exceptions, the above speeds were measured by an observer viewing a tank (7.5m in diameter and 1.1 m deep) from above through a clear acetate sheet onto which the path of the fish was traced (Magnuson, 1969). One species was observed in a water tunnel and one was observed in a 2.9 m diameter doughnut-shaped tank. Three species were observed at sea-two with acoustic transmitters attached to them and one from sequences of aerial photographs. Species do not all have the same speed of sustained swimming. In terms of fork lengths per second, the genera Acanthocybium and Thunnus are slower than A U X ~Euthynnus, S, and Sarda. The wahoo and yellowfin tuna that had the slowest sustained speeds (fork lengths per second) are the two species with the maximum burst speeds observed (Fig. 1B). The persistence of the sustained swimming of scombrids is impressive (Fig. 2). Time of day, temperature, dissolved oxygen, and food deprivation have little influence on sustained swimming. Most striking differences from other fishes are that sustained speeds or volitional activity do not change throughout the die1 period (Fig. 2A), at least when food is not present. These data for kawakawa (Euthynnus afinis) correspond with data from other species, in that day and night speeds are similar (Magnuson, 1973, Fig. 2). The persistence is not laboratory induced because two species observed at sea continued swimming all night (Fig. 1A; Table I).

4.

243

LOCOMOTION BY SCOMBFUD FISHES I

I

I

I

I

u

Q

c

E

n W W

n m n W

f a I-

In

3 In

0

1

0

25

I

I

I

50 75 100 F O R K L E N G T H (cm.)

I

,

125

2.000 0 P a c i f i c Bonito A l l a n l i c Bonilo Bullet Mackerel 0 Kowokawa 0 S k i p l a c k Tuna 0 Y e l l o w f i n Tuna

+

n

1,000 n In

t

m [L

3

m

500

0 25

50 75 L E N G T H (cm)

I00

125

Fig. 1. Swimming speeds of scombrids expressed in terms of distance per unit time, body lengthslsec (t?/sec),and Reynolds number ( R J . (A) sustained speeds and (B) burst speeds. Sources: (1) Blaxter and Dickson (1959); (2) Waiters and Fierstine (1964); (3) Magnuson and Prescott (1966); (4) Walters (1966); (5) Yuen (1966); (6) Yuen (1970); (7) Hunter and Zweifel (1971); (8) Magnuson (1973); (9) Dizon and Neil1 (unpublished); (10) Laurs et al. (1977);(11) Pyatetskiy (19704; (12) MacKay (1976). Bluefin tuna data from Table I are not shown but fish averaging 230 cm long swam at 280 cm/sec (Anonymous, 1975).

Table I Sustained Swimming Speeds of Eleven Scombrids Listed in Order of Decreasing Reynolds Number (RL) Speed R L

flsec

cmlsec

Length (cm)

7.5 x 106

1.6

350

213

6.9

1.3

300

226

6.4~ 106

1.1

270

241

6.2

1.3

290

216

5.5 x 106

0.88

220

250

5.3x 106

1.1

240

219

5.5

X

105

0.77

65

84

All night

5.5

X

105

0.76

65

85

All night

4.4 x 105

0.57

50

87

All night

Species l.a

4 4

Thunnus thynnus (bluefin tuna)

Thunnus alalunga (albacore)

X

X

106

106

Observed duration

Method

Source

Sequential aerial photographs Sequential aerial photographs Sequential aerial photographs Sequential aerial photographs Sequential aerial photographs Sequential aerial photographs Tracked fish with acoustic transmitter Tracked fish with acoustic transmitter Tracked fish with acoustic transmitter

Anonymous (1975) Anonymous (1975) Anonymous (1975) Anonymous (1975) Anonymous (1975) Anonymous (1975) Laurs et al. (1977) Laurs et al. (1977) Laurs et al. (1977)

Acanthocybium solanderi (wahoo) Sarda chiliensis (Pacific bonito) Euthynnus pelamis (skipjack tuna)

01

Thunnus obesus (bigeye tuna) Euthynnus afinis (kawakawa) Auxis rochei (bullet mackerel) Thunnus albacores (yellowfin tuna) Scomber scombrus (Atlantic mackerel) Sarda sarda (Atlantic bonito)

5 . 1 105 ~

0.33

41

125

Continuous

Cinema at sea

Magnuson (1973)

5.0 x 105

1.5

88

57

Continuous

3.5 x 105 3.3 x 105 3.3 x 105

1.5 2.2 1.7

72 84 76

48 39 44

Continuous Continuous All night

Magnuson and Prescott (1966) Magnuson (1973) Magnuson (1973) Yuen (1970)

2.2 x 105 3.3 x 105 1.7 x 105 2.7 x 105

1.6 1.1 1.3 2.1

59 60 47 76

38 55 36 36

Continuous Continuous Continuous Continuous

Time to swim 7.2 m in large tank Traced path large tank Traced path large tank Tracked fish with acoustic transmitter Traced path large tank Traced path large tank Traced path large tank Traced path large tank

2.1 x 105

2.2

68

31

Continuous

Traced path large tank

Magnuson 1973)

1.6 x 105

1.3

46

35

Continuous

Traced path large tank

Magnuson 1973)

9.0 x 104 4.2 x 104 5.6 x 104

0.88 1.2 2.2

28

32 19 16

Continuous Continuous 5-10 sec

Traced path large tank Traced path large tank Swimming" tunnel

MacKay (1976) MacKay (1976) Pyatetskiy (197Oa)

Slowest speed observed from Pyatetskiy's Fig. 2.

22 35

Magnuson Magnuson Magnuson Magnuson

1973) 1973) 1973) 1973)

246

JOHN J. MAGNUSON

Fig. 2. Persistence of sustained swimming of scombrid fish, genus Euthynnus, as influenced by (A) time of day, (B) dissolved oxygen, (C) water temperature, and (D) food deprivation. (A, D) Kawakawa in large shoreside tanks (modified from Magnuson, 1969). (B) Skipjack tuna transferred directly into each concentration from holding tanks saturated with 0,(modified from Gooding and Neill, 1978). (C) Skipjack tuna during day (modified from Dizon et al., 1977).

Sustained speeds of several species are also remarkably independent of ambient temperatures of from 14" to 34°C (Fig. 2C). Stevens and Fry (1972) found no significant change in sustained speeds of two skipjack tuna experiencing a decreasing temperature of 4"Clmin from 24" to 14°C in a small tank of 2400 liters. Dizon et al. (1977) expanded these measurements to more species and individuals, to slower rates of temperature change, and to a 20" range of temperature from lower to upper lethal temperatures. Slower rates of temperature change (l"C/ day over a range of 14"-34°C) did not greatly influence speeds sustained by kawakawa or skipjack tuna in small tanks of 2200 liters. Qlo for both species was approximately 1.1. Similar Qlo values were observed for skipjack tuna and kawakawa at intermediate rates of cooling (0.05"-0.1"C/min from 29" to 18°C) in a larger ring-shaped tank. On the other hand, three yellowfin tuna (Thunnus albacares),at intermediate rates of cooling and corrected for body temperature lags behind ambient temperature, had a Qloof approximately 2.1 (Dizon et al., 1977). Thus, speed of sustained swimming by the euthynnids was essentially temperature independent, but for the thunnid sustained speed in-

4. LOCOMOTION BY SCOMBRID FISHES

247

creased with increased temperature at a Qlonot unlike that observed for many physiological rates in nature. Sustained speeds of skipjack tuna (e = 30-50 cm) in small tanks are independent of dissolved oxygen from 4 to 7 mg/liter (Fig. 2C) (Gooding and Neill, 1978).Oxygen concentrations below 4 mg/liter produced faster swimming but these were not sustained speeds because the fishes almost always died within 150 min. Sustained swimming decreases somewhat with food deprivation (Fig. 2D) (Magnuson, 1969), but even after 5 days without food, kawakawa (C = 41 cm) continue swimming at speeds near 60 cm/sec (1.5k'/sec). These measurements were made in large shoreside tanks. In summary, scombrids of their own volition swim continuously at speeds usually in excess of 0.5 e/sec and as high as 2.2 t'/sec. For the largest species this can result in speeds with R L over 1 x 106. This speed is remarkably uniform for a given species at a given size and persists day and night, over a temperature range of 14"-34"C, over a range of 0, concentrations of 4-7 mg/liter, and in the face of food deprivation for 5 days.

B. Prolonged Swimming There are few data available on prolonged swimming to exhaustion in scombrids because few species have been forced to swim at speeds that result in fatigue. Observations on speeds of fish in feeding activities in the field and laboratory are intermediate to sustained and burst speed. Skipjack in feeding schools at sea had modal speeds of 210-290 cm/sec (3.4-4.3 e/sec) (Yuen, 1966) and feeding kawakawa in large shoreside tanks averaged 180 cm/sec (4.5 t?/sec) (Walters, 1966). Skipjack tuna schools have been followed for over an hour at 400 cm/sec (5-8 l//sec) (Commercial Fisheries Review, 1969). We do not know how long these speeds can be maintained. Several species have been forced to swim to exhaustion. Atlantic mackerel at 16"-17"C in a water tunnel swam about 300 lengths near their top speeds of 300 cm/sec (8.5 e/sec) or from about 35 sec (Table 11). In ongoing studies by Gooding and Neill (1978) skipjack tuna swam to exhaustion at low oxygen concentrations. Examples of these data are: at 1.5 mg/liter dissolved oxygen, the fish swam 10 min at 3.9 [/see before death, and at 3 mg/liter dissolved oxygen they swam 60 min at 2.1 Usec. Skipjack tuna were chased with brooms for 30 min in a large shoreside tank and did not die (Stevens and Fry, 1972),but one

248

JOHN J. MAGNUSON

chased with a dipnet in a small tank stopped swimming and died (Neill et al., 1976). C. Burst Swimming While scombrids have been long observed to be fast swimmers, Walters and Fierstine (1964)were the first to lend scientific credibility to the fact that speeds greater than 10 thee were possible. They measured the rate that yellowfin tuna and wahoo pulled a fishing line from a reel. The line was calibrated with iron filings that were counted by a magnetic pickup on the fishing reel. Speeds were as high as 2066 cm/sec (27 l/sec) for yellowfin and 2134 cm/sec (19 e/sec) for wahoo (Table 11, Fig. 1B). Since 1964 no additional values this high have been measured, but of the eight species for which data are available (Table 11, Fig. lB), measurements on five exceed 10 thee and on all but two exceed Reynolds numbers (R,) 1 x 106. This critical Reynolds number indicates the speed at which flow should shift from laminar to turbulent. Two species apparently exceed an R L of 1 x 10'. Earlier, Watanabe (1942) timed skipjack tuna over a 2 m distance alongside a fishing vessel and obtained speeds greater than 1 x 106 Reynolds number, but it is difficult to associate the size of fish with speeds from his paper. Speeds estimated by Magnuson and Prescott (1966), Walters (1966), Yuen (1966),and Dizon and Neill (unpublished) were all from frameby-frame analyses of motion picture films. Yuen made the estimate by filming feeding schools through underwater viewing ports from the R. V. Charles H. Gilbert. Other films were of fish feeding or pursuing food in large shoreside tanks. Speeds of the mackerels were measured in a swimming tunnel. Where information was given, speeds were measured over time intervals from 0.07 to 5 sec but, for the most part, the duration that speeds could be maintained cannot be extracted from the data. The highest speeds were for short periods as can be observed from two records in Fig. 2 of Walters and Fierstine ( 1964). Only data collected with the calibrated fishing line contain speeds that exceed an R L of 6 x 106 (Fig. l B , Table 11). Walters and Fierstine ( 1964) believed that their estimates, if biased, were underestimates because the fish could turn and thus not be going directly away from the boat, and in addition the line and lure would add to the drag and slow the fish. Yuen (1966)also felt that these high speeds were reasonable when compared to slower speeds he measured because the activities of the fish were quite different. Those collected with fishing

Table 11: Burst Swimming Speeds of Eight Scombrids in Order of Decreasing Reynolds Number (RL) Speed Species

RL

ekec cm/sec

Length '(cm)

Observed duration Method Calibrated fishing line Calibrated fishing line Calibrated fishing line Calibrated fishing line Calibrated fishing line Calibrated fishing line Calibrated fishing line Calibrated fishing line Cinema at sea Cinema at sea Cinema large tank

Source

106 106

7" 10 8 19

2134 1211 1201 2066 1253 993 688 522 545 636 950

113.T.L. 100 T.L. 92 T.L. 76 T.L. 76 T.L. 71 T.L. 71 T.L. 71 T.L. 52 F.L. 79 F.L. 50 T.L.

0.1-2.2 0.1-2.2 -

3.8 x 106

15

750

50 T.L.

0.5

3.6 x 106 3.3 x 106 3.4 x 106 2.1 x 106

9 14 10 6

562 688 588 770 370

64 F.L. 48 F.L. 57 F.L. 57 F.L.

0.1-2.2 0.1-2.2 0.1-2.2 0.1 -

Cinema at sea Cinema at sea Cinema at sea Direct observation at sea Cinema large tank

2.0 x 106 1.6 x 106 1.1x 106

12 10 8

500

400 300

40 F.L. 40 F.L. 36

0.1-0.2 0.1-0.2

-

Cinema large tank Cinema large tank Swimming tunnel

Walters and Fierstine (1964) Walters and Fierstine (1964) Walters and Fierstine (1964) Walters and Fierstine (1964) Walters and Fierstine (1964) Walters and Fierstine (1964) Walters and Fierstine (1964) Walters and Fierstine (1964) Yuen (1966) Yuen (1966) Dizon and Neill (unpublished) Dizon and Neill (unpublished) Yuen (1966) Yuen (1966) Yuen (1966) Watanabe (1942) Magnuson and Prescott ( 1966) Walters (1966) Walters (1966) Blaxter and Dickson (1959)

Scomberjaponicus

6.6 x 105

8

226

29

-

Swimming tunnel

Hunter and Zweifel (1971)

(Pacific mackerel) Sarda sarda (Atlantic bonito)

2.0 x 105

7.8

128

16 T.L.

Water tunnel

Pyatetskiy (197Oa)

Acanthocybium solanderi

(wahoo) Thunnus a1bacores

(yellowfin tuna)

Euthynnus pelamis

(skipjack tuna)

Sarda chiliensis (Pacific bonito) Euthynnus afinis

(kawakawa) Scomber scombrus

2.4 x 1.2 x 1.1 x 1.6 x 9.5 x 7.0 x 4.9 x 3.7 x 2.8 x 5.0 x 4.8 x

107 107 107

lo7 106 106 106 106

106

19 12 13 27" 16a 14" lo"

ca. 5

-

-

ca. 5

-

-

-

Cinema large tank

(Atlantic mackerel)

5-10

Body lengths per second were recalculated from Table 1 of Walters and Fierstine (1964) on the basis of median lengths of fish in school rather than mm2mum length of fish in school.

250

JOHN J. MAGNUSON

gear were fleeing, while Yuen's data were on fish in a feeding school. However, even if the data collected with the calibrated reel were dismissed as questionable, other measurements on three scombrids exceed an R L of 106. The skipjack tuna exceed an R L of 4 x 10s as measured b y two authors and 18 t'/sec by one author. While I think the estimates from the fishing reel may be on the high side, burst speeds between 10 and 20 t'/sec and in excess of an R L of 1 x 10s are well documented for scombrids. Burst swimming capabilities vary amofig scombrids. Ranking the species either by maximum lengths swam per second or maximum Reynolds number suggests that the genera Acanthocybium, Thunnus, and Euthynnus are faster swimmers for their length than are Sarda and Scomber. In summary, while burst speeds vary among scombrids, the fastest measured to date are near 2000 cm/sec, 20 t'/sec and R L of 2 x 107. Most have burst speeds that exceed 10t'/sec. Scombrids are among the fastest if not the fastest fishes for which good data are available.

111. GENERAL CONSIDERATIONS OF SWIMMING EQUILIBRIA

The swimming speeds of scombrids are well documented (see Section I1 and Chapter 2) and indicate that these fish are indeed continuous swimmers and that they can swim extremely fast. In both modes of swimming (sustained and burst), speeds of certain large sized or fast species exceed a Reynolds number of 1 x 106. One level of explanation for the necessity and the possibility of their particular swimming behaviors comes from an analysis of the equilibria between both the vertical and horizontal forces acting on the animal. These forces are weight, buoyancy, and hydrodynamic lift in the vertical direction and thrust and drag in the horizontal direction (Fig. 3). A swimming scombrid produces thrust and drag which oppose each other. The weight of the fish is opposed by buoyancy and hydrodynamic lift. When opposing forces are equal the fish is in equilibrium, that is, swimming at a constant speed or constant depth. If not, the fish accelerates or decelerates or rises or sinks until the forces are again in equilibrium. For scombrids the equilibrium between thrust and drag is intimately related to the equilibrium between weight, buoyancy, and lift. This is because scombrids are heavier than water and they balance the vertical forces by swimming continuously with pectoral fins extended. Hydrodynamic lift is produced on the pectoral fins. While this adds to buoyancy and balances the vertical forces, the imposed requirement of

4. LOCOMOTION BY

WElGWT

251

SCOMBFUD FISHES

:@!+gj ... .....

“t

BUOYANCY PLUS HlDRODVNbUlC LIFT

Fig. 3. Forces acting on a swimming scombrid and axis of yaw, pitch, and roll. @ indicates location of the center of gravity. (Adapted from Magnuson, 1970.)

swimming to produce hydrodynamic lift complexes the balance between horizontal forces. Drag or resistance, as expected, results from the movement of the animal through the water, but significant drag also is induced in conjunction with the lift produced on the pectorals and even, in a sense, from the thrust produced on the caudal fin. Thus, an understanding of the hydromechanics of swimming by scombrids involves not only an analysis of thrust and drag but also an analysis of the mechanisms and effects of maintaining vertical equilibrium. These relationships are treated below-vertical forces in Section IV, horizontal forces in Sections V and VI. The interdependencies are pointed out and discussed in each section but an attempt to integrate them is presented in Section VII. I n addition the hydrodynamic stability of the swimming fish can be considered in terms of pitch, roll, and yaw (Fig. 3) (Alexander, 1967). Pitch and yaw refer to longitudinal stability in the vertical and horizontal planes, respectively. Roll refers to lateral stability or rotation in a transverse plane. Forces produced by the paired and median fins oppose forces producing these instabilities, or forces produced by the paired and median fins contribute in a controlled way to pitch and yaw for vertical and horizontal steering. These aspects of swimming are not treated here.

IV. WEIGHT, BUOYANCY, HYDRODYNAMIC LIFT, AND PREDICTION OF SUSTAINED SPEEDS The sustained speeds cited above were referred to b y Magnuson (1970, 1973) as minimum speeds for hydrostatic equilibrium. Hydro-

252

JOHN J. MAGNUSON

dynamic lift on the extended pectoral fins is generated as the fish swims forward with thrust from the caudal fin. The lift on the pectorals balances weight not met by buoyancy and, as discussed in Section 111, results in an equilibrium between vertical forces. The mechanism of maintaining hydrostatic equilibrium is analyzed below to develop quantitative predictions of the speeds sustained by swimming scombrids. There are few other groups of fishes for which such an analysis can be used to accurately predict minimum levels of activity over the size range of the species. These estimates of speed and lift production also contribute to estimation of drag on a swimming scombrid.

A. Mechanisms of Hydrostatic Equilibrium 1. WEIGHT The estimation of the fish’s weight is a first step in the analysis. Weight, like all forces, is the product of a mass and an acceleration and is measured in dynes or as the force that will accelerate 1 g of mass 1 cm per second per second. Thus, a fish’s weight (WJ is equal to its mass (Mf) times the acceleration of gravity (g):

Wf = M , g where g is approximately 980 cm/sec2.

( 2)

The mass-length relationships common in fishery biology provide a good data base for weight estimation and, as expected, are approximately the function of length cubed for scombrids (Magnuson, 1973). Yet, weights of scombrids of the same length differ considerably; for example, yellowfin tuna are more than four times heavier than the wahoo of the same length. Such differences are significant in the balance between vertical forces and contribute to differences in predicted speed.

2. BUOYANCY Fortunately for a scombrid most of its weight is opposed by buoyancy that results from the large mass of water the fish displaces. If this were not the case, the hydrodynamic lift required would be too great to produce by sustained swimming as it would impose impossible speeds. The buoyant force (B is equal to the mass of water displaced ( M w ) times acceleration of gravity: f)

Bf = M w g

(3)

4.

LOCOMOTION BY SCOMBFUD FISHES

253

where the mass of water displaced (M w) is equal to the volume of the fish (Vf) times the density of seawater (pe):

Mw = PeVf ( 4) Volume of the fish can be calculated by determining the density of the fish pf with Archimedes’ principle (Alexander, 1968) and dividing:

v f =Mf Pf

The resulting equation after substitution and rearranging to solve for buoyancy is

From Eq. (6) it can be seen that the buoyant force is equal to the weight of the fish ( M f g )times the proportion countered by buoyancy, pelpf.This ratio between the density of water and the density of the fish when multiplied by 1000 (an approximation of g ) has been referred to as the sinking factor (Lowndes, 1955; Jones, 1957). The following example will demonstrate that most of a scombrid’s weight is opposed b y buoyancy. A skipjack tuna with a mass of 10 kg ( M f = 10,000) weighs 9.8 x 106 dynes from Eq. (2) ( W f = 10,000 x 980 = 9.8 x 106 dynes). The proportion of the weight opposed by buoyancy‘is 0.94 (pelpf = 1.025l1.091 = 0.94) and the buoyant force is 9.2 x 106 dynes from Eq. (6)(B, = Wf x 0.94 = 9.2 x 106 dynes). Even though the skipjack has no gas bladder, 94% of its weight is opposed by buoyancy. All scombrids measured to date (Table 111) have densities greater than the seawater in which they were caught. Consequently not all of their weight is opposed b y buQyant forces. Since scombrids are pelagic, their excess weight also is not countered by resting on the bottom. Even though buoyant forces are great, weight does increase approximately as a cubic function of length, and large fish become exceedingly heavy for their length. One adaptation that increases the proportion of weight countered by buoyancy is the gas bladder. Many scombrids have no gas bladder, but some, such as the yellowfin tuna, bigeye tuna (Thunnus obesus), albacore (Thunnus alahnga), wahoo, and Pacific mackerel (Scomber japonicus), have. Some species such as yellowfin tuna (Magnuson, 1966, 1973)and albacore (Dotson, 1977)have allometric growth of the gas bladder that results in a greater proportion of their weight being

254

JOHN J. MAGNUSON

Table I11 Densities of Twelve Scombrid Species in Order of Decreasing Density Density of scombrids (g/cm3)

Species

Euthynnus pelamis" (skipjack tuna) Euthynnus afinis" (kawakawa) Auxis rochei" (bullet mackerel) Sarda sardab (Atlantic bonito) Sarda chiliensis" (Pacific bonito) Auxis thazordb (frigate mackerel) Thunnus albacoresn (yellowfin tuna) Thunnus alalunga (a1bacore) San Diego July" San Diego Sept." Thunnus ohesusn (bigeye tuna) Scomberjaponicuso (Pacific mackerel) Scomher scombrus 4tlantic mackerel) Juneb August Septemberb Novemberb September" AI rnthocyhiurn solanderi (wahoo)

Gas bladder present

Source

1.090 1.090 1.094 1.097 1.094 1.091

No

Magnuson (1973)

1.087 1.088 1.091

No

Magnuson (1973)

1.086

No

Magnuson (1973)

No

Aleev (1963)

1.075 1.080

No

Aleev (1963)

1.07

No

Aleev (1963)

1.086 1.087 1.068 1.058 1.054

Yes

Magnuson (1973)

Yes

Dotson (1977)

Yes

Magnuson (1973)

Yes

Magnuson (1973)

No

Aleev (1963)

Yes

Aleev (1963) MacKay (1976) Aleev (1963) Aleev (1963) Lowndes (1955) Magnuson (1973)

0-2

1.08

2-4

Mass (kg) 4-6 6-8

8-10

10

1.08

1.088 1.082 1.063 1.060 1.066 1.047

1.056

1.054

1.06 1.06 1.03 1.02 1.051 1.028

Density of habitat, 1.025 &m3. 1.01 g/cm3.

* Density of habitat,

carried by buoyancy after they attain masses greater than 2 kg. By the time yellowfin tuna reach 8-10 kg, their weight not balanced by buoyancy, would be twice as great if the gas bladder had not developed. The gas bladder of a 10 kg wahoo is sufficiently large that 99.7% of the fish's weight is opposed by buoyancy (Magnuson, 1973)

4. LOCOMOTION BY SCOMBRID

FISHES

255

in contrast to the 94% calculated above for skipjack tuna. Of those species that attain masses greater than 70 kg, all have gas bladders, whereas of those with a maximum mass less than 6 kg only 25% have gas bladders (Magnuson, 1973). Apparently, larger scombrids tend to have a gas bladder, to have lower densities, and to oppose a greater proportion of their weight by buoyancy than do smaller scombrids. Available evidence indicates that this generalization applies to comparisons both within and among species. A second mechanism for increasing the weight carried by buoyancy is to have reduced density by increased fat content. In some sharks, fatty livers greatly reduce density (Bone and Roberts, 1969; Baldridge, 1970, 1972). Aleev (1963) (Table 111) noted that seasonal changes in densities of Atlantic mackerel were associated in a sensible way with seasonal changes in fat content. The lowest density (Table 111) was associated with 23.1%fat in the flesh in fall and the highest density with 2.9% fat in spring. Albacore are more dense when first arriving on the California coast in July than after 2 months of feeding in coastal waters (Dotson, 1977). Calculations on kawakawa (Magnuson, 1970) indicate that 44% fat is required if buoyancy were to equal weight. Variations in observed densities of kawakawa are explained by a range of fat content of only 5-15% fat. However, any increase in fat content would reduce density, and 10%fat content in kawakawa can reduce the weight not balanced by buoyancy by about 25% over a fat-free condition. Of the species listed in Table I11 without a gas bladder, observed densities reveal that fat content is not sufficiently high to attain neutral buoyancy. Only the Atlantic mackerel is reported to have higher fat and lower density at larger sizes (Aleev, 1963). While gas bladders and fat contents reduce somewhat for scombrids the weight not countered by buoyancy, additional upward forces are required to obtain equilibrium between vertical forces. It is these forces remaining to be opposed by hydrodynamic lift that necessitate continuous swimming. For a realistic prediction of sustained speed, special care must be taken to obtain accurate measurements of density of fish that have a gas bladder (see Methods in Magnuson, 1970, 1973; Dotson, 1977).

3. HYDRODYNAMIC LIFT Among scombrids the excess weight not balanced by buoyancy is opposed by hydrodynamic lift generated b y the extended pectoral fins as the fish swims forward (Magnuson, 1966, 1970, 1973). The same mechanism has been explained by Aleev (1963) with sturgeon and by Alexander (1965, 1968) with sharks.

JOHN J. MAGNUSON

256

The lift (L) required to obtain hydrostatic equilibrium can be calculated as the difference between weight and buoyancy of the fish:

L=WI-BI

(7)

On substitution of Eqs. (2) and (3) and rearrangement, this becomes

which is the excess density (pr- pe) times the volume of the fish (&If/ p3 times the acceleration of gravity. Excess density, like sinking factor, has been used to characterize the weight-buoyancy relationships of fishes. It is the same as the “buoyancy A” of Aleev (1963). In summary, the lift required of a fish to balance the inequity between weight and buoyancy depends on the size of the fish (M3 and the difference between the density of the fish and its habitat. Less lift is required for smaller fish and those with a density closer to that of the water in its habitat.

B. The Lifting Surfaces 1. THE PECTORAL FINS

Scombrids not only swim continuously, but do so usually with their pectoral fins extended laterally from their body and with their pelvic and spiny dorsal fins appressed (Figs. 4 and 5) (Magnuson, 1966, 1970). The pectoral fins are structured and used in such a way that they are obvious lifting hydrofoils. Lift on a pectoral fin would be produced by the lower pressure on the upper than on the lower surface. Water flows farther and thus faster over the upper surface and the resulting pressure difference produces a net lift. The fin and fish are essentially lifted upward rather than pushed upward. In addition, two other sources of hydrodynamic liA have been considered for scombrids, the caudal keel and the body itself (Fig. 4A). Lift production b y a hydrofoil, such as the extended pectoral fins, depends directly on the lifting area of the fins (S J and characteristics of the hydrofoil that influence its coefficient of lift (CL), Lift is also proportional to the speed of the fish squared (U2)and the density of water. The equation that relates the required lift to speed and the characteristics of the lifting surfaces L = Y2peCL.SpU2

4.

LOCOMOTION BY SCOMBRID FISHES

257

Fig. 4. Diagrams based on kawakawa of the lifting surfaces, especially the pectoral fins (adapted from Magnuson, 1970).(A) Dorsal view showing lifting areas of the pectoral fins, keel, and body. (B) Angle of attack. (C) Dorsal view showing method to determine sweepback, span, and hydrodynamic center of the pectoral fins. (D) Examples of variable sweepback. (E) Dihedral angle from frontal view. (F)Longitudinal locations of centers of lift from pectoral fins, keel, and body, in respect to the center of gravity.

is taken directly from aerodynamics of wings. General discussions of its basis and derivation are in von Mises (1945), Prandtl and Tietjens (1934),von Kirmin (1954), and, in respect to aquatic animals, Alexander (1968). A number of other postural characteristics influence the coefficient of lift for the pectoral fin and consequently the relation between speed and lift production. A cambered fin (Fig. 4B) at zero angle of attack relative to the horizontal or even at slight negative

258

JOHN J. MAGNUSON

Fig. 5. Dorsal photograph of ( t o p )a euthynnid and (bottom)a thunnid showing variation in lifting areas and posture of pectoral fins. (Photos from National Marine Fisheries Service, Biological Laboratory, Honolulu.)

4.

LOCOMOTION BY SCOMBRID FISHES

259

angles of attack would produce lift. Within limits the greater the camber the greater the coefficient of lift (CL).The lift coefficient (C,) also increases with increasing angles of attack (Fig. 4B) until the fin stalls. A twisted fin has smaller angles of attack nearer the tip, a feature which helps prevent the tip of the fin from stalling first. The relations between lift and speed and the characteristics of the pectoral fins are well illustrated b y the behavior and postures of a scombrid swimming at different speeds. As a scombrid swims faster, lift increases as speed squared [Eq. (9)]but lift required remains constant. To maintain a level course at faster speeds, the pectoral fins must be regulated in such a way that hydrodynamic lift does not surpass excess weight and cause the fish to rise. Scombrids use at least two mechanisms to reduce lift from pectorals at faster speeds. Kawakawa reduce the time that the pectoral fins are extended from 100%at slow speeds to 0% at moderate speeds (Magnuson, 1970). For example, a captive kawakawa, 42 em long, that had just fed extended its fins continuously at 1.6 t'lsec and appressed them continuously at speeds of 2.8 t'lsec and greater. The second mechanism of sweeping them back farther at faster speed reduces the effectiveness of the extended pectoral fins of kawakawa (Magnuson, 1970) and Atlantic mackerel (MacKay, 1976). Increasing sweepback reduced the area, the span, and the lift coefficient of an extended pectoral fin. For example, 42 cm kawakawa in steady swimming swept the fins back more at faster speeds. They swam at 1.3 t'/sec with the pectorals swept back 25" after 5 days of food deprivation, but when just stimulated with food odor they swam at 3.0 tlsec with a 52" sweepback.

2. THE KEEL AND BODY

The keel and the body (Fig. 4A) are less obvious as lifting hydrofoils and may be rather unimportant in lift production compared with the pectorals (Magnuson, 1970). The body of scombrids sometimes can be discounted as a lifting surface at sustained speeds. In contrast to the bodies of sturgeon (Aleev, 1963) and sharks (Harris, 1936), they are not cambered (Magnuson, 1970). Kawakawa at sustained speeds swim with no angle of attack to the body (Magnuson, 1970), and no lift would be expected from a symmetrical body at zero angle of attack (Munk, 1924). But Atlantic mackerel do swim with a positive angle of 9" to the body (Muir and Newcombe, 1974) and the body was calculated to provide 14% of the lift (MacKay, 1976).

260

JOHN J. MAGNUSON

The keel (Fig. 4A,F) at best is a complex lifting surface (Magnuson, 1970). Water flow alternates across it with each beat of the caudal fin (Fig. 6). Flow is directed across it by the finlets that act as flow fences (Magnuson, 1970; Aleyev, 1977).The keel is a noncambered, flattened area with lateral extensions at the peduncle. At sustained speeds the caudal region behind the second dorial is deflected downward an average of 5"-6". Since flow passes across the keel and the keel has an angle of attack it should produce some lift. Its area is about 20%of the lifting area of the pectorals and has a lift coefficient similar to the pectorals (Magnuson, 1970). Since the caudal region is deflected slightly, the caudal fin would be expected to produce a vertically directed vector along with the horizontal vector of thrust. Magnuson (1970, 1973), in an analysis, discounted this as a significant source of lift for Euthynnus, Thunnus, A U X ~ Sand , Sarda. But in the Atlantic mackerel (MacKay, 1976) the caudal fin is expected to provide significant lift because the body is at a positive angle of attack and because that species also swims with the plane of the caudal fin tilted to the side. Thus a lift force is produced as the caudal fin is swept back and forth (Aleyev, 1977).

Fig. 6. Dorsal photograph showing keel and posture of finlets during sustained swimming. Keel and caudal fin being swept toward bottom of page. (From Magnuson, 1970, Hydrostatic equilibrium of Euthynnus ufinis, a pelagic teleost without a gas bladder, Copeia pp. 5685.)

4. LOCOMOTION BY SCOMBFUD FISHES

3.

COMPUTATION O F

26 1

LIFT COEFFICIENTS

Estimates of the lift coefficients of the lifting surfaces are needed to quantitatively relate swimming speed and required lift. The lift coefficient of the pectorals and keel can be computed from Eq. (9) solved for C L. Required lift, the density of the environment, the lifting area of the pectorals, and the speed of sustained swimming can all be measured (Magnuson, 1970). In addition, required lift to counter excess weight must first be partitioned between the pectorals and the keel. This can be done by solving two simultaneous equations (Alexander, 1968; Magnuson, 1970): one, that lift from the pectorals and keel must sum to equd required lift and, two, that the lift must balance around the center of gravity to maintain longitudinal stability. The excess weight can be considered to act at the center of gravity. The greater distance from the center of gravity to the keel than to the pectoral means that a smaller lift from the keel could balance the larger lift from the pectorals (Fig. 4F) b y analogy with a first order lever arm. In the solution for kawakawa, lift from the pectorals should be x l / x l = 0.46/0.11, or four times the lift from the keel. The location of the center of buoyancy could alter the above interpretation. In scombrids the center of buoyancy may be a few millimeters behind the center of gravity (Magnan, 1929; Bone, personal communication). If so, since the buoyant force is so large, it might, even with such a short level arm, balance the lift from the pectoral fins about the center of gravity for longitudinal stability. This would not alter the idea that the keel should act as a lifting hydrofoil based on its anatomy and posture, but perhaps the keel does not carry as much as 20% of the excess weight. Consequently Magnuson ( 1970) calculated lift coefficients for two situations: (1)lift behind the center of gravity from keel only, and (2) lift behind the center of gravity from buoyancy only. Computed coefficients pf lift (C,) (Table IV) calculated from captive kawakawa range from 0.6 to 1.0 for pectoral fins based on total lifting area (Fig. 4A)and from 0 to 0.9 for the keel. The coefficients are not unreasonable for efficient cruising but are less than the maximum that can be achieved from a wing with plain flaps when C, may equal 1.9, or modern slotted wings with leading and trailing edge flaps. Since the area of the hydrofoil is so important in determining the lift produced, the large variations among scombrids in pectoral fin size (Fig. 5) are important in determining the differences in sustained swimming speeds of scombrids (Magnuson, 1973). At a length of 100 cm, yellowfin tuna have a lifting area almost four times that of the wahoo. Fish with a larger pectoral should be able to produce more lift at a given speed. Those species that have a maximum mass greater

262

JOHN J. MAGNUSON

Table IV Coefficients of Lift (C,) of the Pectoral Fins and Peduncular Keel Computed for a Kawakawa, 42 cm long" Condition of fish Assumptions

Lifting surface

5 Days without food

Just fed

Lift behind center of gravity from keel Lift behind center of gravity from buoyancy

Pectoral fins Keel Pectoral fins Keel

0.8 0.9 1.0 0

0.6 0.6 0.8 0

a

Data from Magnuson (1970).

than 70 kg tend to have pectoral fins that are relatively longer than those of smaller species. As the fish grow, the lifting area increases approximately as a function of fish length squared. Exponents range from 1.7 to 2.7 (Magnuson, 1973; Dotson, 1977).Thus, both among and within species, lifting areas are relatively greater for larger than for smaller fish. Species with longer pectoral fins such as the bigeye tuna and yellowfin tuna have greater sweepback (A = 60")during sustained swimming than those with shorter pectoral fins such as skipjack tuna and kawakawa (A = 43") (Magnuson, 1973).Those with longer fins also have a higher aspect ratio (span/mean chord) which would increase the ratio of lift to drag over that of a shorter fin.

C. Surface Loading of Lifting Surfaces A number which allows comparison of the lifting characteristics of the pectoral fins and other lifting surfaces in the animal kingdom is the loading of the fin (LIS,), or the dynes of weight carried per square centimeter of lifting surface (dynes/cm2).Loading of scombrids range from about 400 dynes/cmZ for wahoo to 6000 dynes/cm2 for skipjack tuna (Fig. 7). The skipjack tuna is the maximum for scombrids because it is the largest species with small pectoral fins and no gas bladder and for which data are also available (Magnuson, 1973). If the keel were assumed to carry 20% of the weight as estimated for kawakawa, then maximum loading of the pectorals would be slightly less or about 5000 dynes/cm*.Three species without gas bladders that are larger than the skipjack tuna might have a higher loading: the black skipjack (Euthynnus alletteratus), king mackerel (Scomberomorms cavalla), and Scomberomorus commerson.

4. LOCOMOTION BY

263

SCOMBRID FISHES

0o0 l0,OOO

\OO'

A

00"

\O'

0O0 \*

0 Wahoo 0 P a c i f i c Bonito

+ Bullet

Mackerel

0 Kawakawa I

1,000

Bigeye 'H X Lemon

In Q c

Tuna Shark

,O0

21

-0

0

n

E

Atlantic Sharpnose Shork @ Tiger Shark A Bull S h a r k

@ .

0 \ v) 0)

t

\o

-," W

z -

I-

r c3

n U

\ s

W

3

10

I

I

10

too

1,000

10,000

L I F T I N G A R E A (ern?) Fig. 7. Comparison of the loadings of lifting surfaces among scombrids, sharks, birds, and bats. Diagonal lines are surface loading isopleths. Areas for bats i n d birds are for the wings, for fish S , in Fig. 4A. Weight is less buoyancy for fish or L in Eq. (7). Sources: albacore (Dotson, 1977); other scombrids (Magnuson, 1973); sharks (Baldridge, 1970, 1972, and personal communication); birds (Hertel, 1966); and bats (Findley et al., 1972). Line fitted by eye on twenty-four species.

At smaller size, skipjack tuna not only weigh less, but have lower fin loadings (Fig. 7). The weight carried per unit area increases approximately in direct proportion to the fish length for species with a constant density. The relations are given by L cc C3and S a e2 so that LISPa e. For skipjack tuna, L a e3.36and S, 0: P2*so that LIS, a (coefficients from Magnuson, 1973, Table IV). Consequently larger skipjack have to swim faster in absolute terms to produce sufficient lift than do smaller skipjack [see Eq. (9)]. Other scombrids without a gas bladder have fin loadings similar to skipjack tuna, but those with a gas bladder have lower loadings either for a given weight or lifting area (Fig. 7). This results not only because the gas bladder greatly reduces density but also because fish with a gas bladder often have larger pectoral fins as well. Pectoral fin loading for yellowfin tuna increases more slowly than for skipjack tuna as the fish grows larger (Fig. 7). At smaller sizes the

264

JOHN J. MAGNUSON

yellowfin have lower fin loading than skipjack tuna because their pectoral fins are larger than skipjack tuna's but as they attain weights near 100 x 103 dynes, the allometric growth of the gas bladder results in a greater proportion of their weight being assumed by buoyant forces. A trend of increasing fin loading with length is countered by the development of the gas bladder. For example, a yellowfin tuna 100 cm long has a fin loading only 1.1 times that of 25 cm fish. With the same change in length, fin loading of a skipjack tuna would have increased by a factor of four. Consequently for yellowfin, larger masses do not produce as great an increase in fin loading and they should be able to maintain'slower speeds than skipjack to counter their weight. Fin loadings of the albacore even decrease at larger sizes because both the gas bladder and the pectoral fins are growing allometrically (Dotson, 1977) with exponents greater than 3 for the gas bladder and greater than 2 for the pectorals. Data on large sharks are especially informative because fin loading has not been measured to date on any scombrid with a mass greater than 12 kg, the wahoo and largest albacore. Pectoral fin loading of the small Atlantic sharpnose shark, Rhizoprionodon terraenovae, and of three large sharks-the lemon shark (Negaprion brevirostris), the tiger shark (Galeocedo cuvieri), and the bull shark, (Carcharhinus levcas)-were all intermediate between scombrids without a gas bladder and birds (Fig. 7). The large sharks ranged from 60 to 240 kg in mass and the largest had the lowest loading among the large sharks. Large fatty livers increase the proportion of the sharks' weight carried by buoyant forces at larger sizes (Baldridge, 1970, 1972). Consequently, the large negatively buoyant sharks, like the larger scombrids, maintain lower fin loadings by lowered densities at larger sizes. They should be able to swim at slower sustained speeds at larger sizes. I think that large scombrids for which we have no data mag also have low fin loadings and retain slower speeds at larger sizes (Magnuson, 1973). Data for the lemon shark, compared with the bullet mackerel (Fig. 7), demonstrate the above point. The lemon shark has a mass of approximately 100 kg. Yet, its pectoral loading is the same as that of bullet mackerel that has a mass of only 0.5 kg but no fatty liver or gas bladder. Assuming that their lifting surfaces are equally efficient in producing lift, both should have to swim at the same speed to produce sufficient lift to keep from sinking. This speed, about 60 cm/sec, is approximately 2 e/sec for the bullet mackerel, but only 0.25 thee for the lemon shark. The lowered speeds and associated energetic advantage of lower fin loading are dramatic, especially for large fish.

4.

LOCOMOTION BY SCOMBRID FISHES

265

Some scombrids have lifting areas about one order of magnitude smaller than birds or bats at the same weight (Fig. 7). Thus they have fin loadings about one order of magnitude greater than birds or bats (Fig. 7). For example, a 22 kg skipjack tuna in seawater has about the same weight as an adult cormorant or vulture in air (Hertel, 1966) but has a lifting area similar to the wing areas of medium size birds such as woodpeckers and doves. As pointed out by Magnuson (1970) this order of magnitude difference results in part from the differences in the density of water and air. Water is about 860 times the density of air. If an airborne flyer and a waterborne flyer have the same weight in their respective media, then the product of the C LSJ 2 for the aquatic animal has only to be Vssoth of the same product for an aerial animal. Tuna can provide sufficient lift with smaller lifting surfaces, slower speeds, and perhaps less efficient lifting surfaces than birds and bats. The smaller lifting surfaces account for about one order of magnitude of the approximate three orders of magnitude difference in density between air and water. The other two orders of magnitude must be accounted for by faster speeds of birds or more efficient flight mechanisms. Lifting areas of all groups considered in Fig. 7 increase with weight. The slope or rate of increase appears similar for skipjack tuna, birds, and bats. Hertel (1966) indicated that for birds plus insects the relation was S = 15 W2'3 based on the surface-volume relationships. More recent analyses on birds give similar information (Greenwalt, 1975). The relationships fit data on insects and birds quite well. In all cases, however, the weight increases more rapidly than lifting area and the larger species or larger individuals carry more weight per unit area of lifting surface. Birds approach 1000 dynes/cm2 while scombrids approach 10,000dynes/cm2.The only animals plotted that do not seem to obey this particular two-thirds rule are the yellowfin tuna and albacore. Again this is because buoyancy from the developing gas bladder carries proportionately greater weight as the fish grows, and because (Dotson, 1977). for albacore lifting area increases not as t2but as P.s7 The largest scombrids would be expected to have pectoral loadings even more similar to birds, as was observed for large sharks that had surface loadings similar to storks and pelicans.

D. Predictions of Sustained Speed The minimum speed at which a scombrid can swim and still produce sufficient hydrodynamic lift to balance their excess weight can be

266

JOHN J. MAGNUSON

calculated from the information on required lift and the properties of their lifting surfaces discussed above. Basically Eq. (9) can be rearranged to solve for speed (U) under the condition that the pectoral fins are extended 100%of the time. This equation (Magnuson, 1970, 1973), including both the pectoral and keel lifting surface, is

where Uloo, the speed (cm/sec) of fish with pectorals continuously extended; CL,, coefficient of lift for pectorals; S p, total lifting area of pectoral (cm2);CL,, coefficient of lift for the keel; Sk, lifting area of keel (cmz);L, required lift (dynes); pe, density of environment (g/cc). Minimum hydrostatic speeds of various scombrids were computed from Eq. (10) and compared with observed sustained speeds of scombrids in shoreside tanks at Kewalo Basin, Honolulu. The predictions were quite close (Fig. 8A). Correlation coefficients were +0.9 if all lift were assumed to come from the pectorals or +0.8 if the lift was partitioned between pectorals and the keel. Thus, minimum speeds required for hydrostatic equilibrium are for practical purposes the same as the observed sustained speeds of captive scombrids. The model was used (Magnuson, 1973; Dotson, 1977) to predict sustained speed over a greater range of lengths (Fig. 8B,C). The poten-

d

4

lc

,Z+0.89 0.5

1.0

OBSERVED

1.5

2.0

S P E E D (f/sec.l

FORK

L E N G T H (em)

Fig. 8. Predicted speeds of sustained swimming for scombrids. (A) Comparison of predicted and observed speeds; (B) predicted speed (cmhec) related to t;and (C) predicted speed (tkec) related to t.(Adapted from Magnuson, 1973; Dotson, 1977.)

4. LOCOMOTION BY SCOMBRID FISHES

267

tial for lowered metabolic rates from the presence of a gas bladder is apparent both within and among species. Minimum speeds for hydrostatic equilibrium of the three thunnids (yellowfin tuna, bigeye tuna, and albacore) decline at larger sizes owing to development of the gas bladder and pectoral fins. Skipjack tuna, a species without a gas bladder, at lengths of 100 cm has a minimum speed 3.5 times greater than yellowfin tuna and many times greater than bigeye (Magnuson, 1973). The influence and potential energetic savings of large pectoral fins can be seen by comparing, at a length of 50 cm, the minimum speeds of yellowfin tuna with large fins and skipjack tuna with small fins. At this length the gas bladder has not developed, and yellowfin tuna speeds are 70% of the skipjack tuna speeds. The facts that scombrids do not stop swimming (Fig. 2) and that sustained speeds are predictable (Fig. 8A) are a product of the physical system by which they maintain a balance between vertical forces of weight, buoyancy, and hydrodynamic lift. In addition, variations in pectoral fin areas, gas bladder' sizes, and sustained speeds among species exemplify alternate solutions to balancing these vertical forces. Sustained swimming has brought with it other sets of adaptations-ram gill ventilation and perhaps the lunate tail. Ram gill ventilation, characteristic of scombrids, also requires continuous forward movement (Hall, 1930; Brown and Muir, 1970; Roberts, 1975). The structures for pumping water across the gills are poorly developed or even lost in some species. If these fish stop swimming suffocation results. The mechanisms by which scombrids maintain equilibrium among vertical forces, i.e., swimming continuously and producing hydrodynamic lift on the pectorals, and the related adaptations such as ram gill ventilation, significantly alter the relative importance of the various components of drag and thrust in the equilibrium amon'g horizontal forces. These interrelations are considered next.

V. RESISTANCE TO FORWARD MOVEMENT Scrombrids are among the most highly evolved fish in regard to adaptations to reduce drag or the resistance to forward movement. This is not surprising since they have both high sustained speeds and high burst speeds, and drag is proportional to speed squared. Adaptations of the body such as streamlined shape have been noted frequently (Walters, 1962; Marshall, 1965; Hertel, 1966; Alexander, 1967, 1968; Webb, 1975; Aleyev, 1977). Gill resistance and drag produced with hydrodynamic lift on the pectoral fins were considered

268

JOHN J. MAGNUSON

by Brown and Muir (1970) and MacKay (1976) but drag induced from

thrust by the caudal fin has not been evaluated in discussions of total drag. Other adaptations to reduce the energy requirement of swimming, such as variations in locomotory behavior, have been identified as possible in theoretical papers (Weihs, 1973a,b, 1974)but have as yet received little or no confirmation from experiment or observation.

A. Components of Drag Drag results from skin friction between a fish and the boundary layer of water (friction drag), from pressures formed in pushing water aside for the fish to pass (form or pressure drag), and from energy lost in vortices formed by the pectoral fins and the caudal fin as they produce lift or thrust (induced drag). Alexander (1968), Brown and Muir (1970), and Webb (1975) provide more complete explanations of these general sources of drag.

ESTIMATION OF DRAG The various structures producing drag and the components of drag are discussed below for scombrids. Methods of estimating drag are described and the magnitude of the components of drag for major structures are calculated and presented in Table VI for a skipjack tuna swimming at sustained speeds. When these component drag forces are summed they provide an estimate of the thrust required to oppose drag forces. Total drag can also be estimated by estimating total thrust and equating it to total drag as was done by Pyatetskiy (1970b). These estimates will be discussed in the section on thrust. a. Friction and Form Drug. Webb (1975) argued that theoretical calculation of drag and dead-drag measurements on towed fish produced unsatisfactory results with the possible exception of scombridlike fishes with stiff bodies and carangiform locomotion with lunate tail. Measurements which Webb (1975) cited from Bone of dead-drag on towed Atlantic mackerel were similar to theoretical calculations for laminar flow conditions. However, another set of measurements on free falling mackerel by Richardson (1936) produced estimates of drag coefficients that differed from theoretical estimates by three orders of magnitude (Webb, 1975, Fig. 46). Thus, care must be taken in any measurements of dead drag even for scombrids. Frictional drag (Of) is proportional to the density of water ( p J , a

4.

269

LOCOMOTION BY SCOMBFUD FISHES

measure of surface area (S), speed squared (Uz),and a coefficient of frictional drag (CJ. It can be estimated from

D f = ‘/2Cfp&’U2

(11)

Mechanics of drag are discussed in Webb (1975). The coefficient of friction can be calculated for laminar, turbulent, or transitional flows as follows (Prandtl and Tietjens, 1934):

C

f

Cf laminar = 1.33 RL-l12

( 12)

cfturbulent = 0.074 RL-’’~

( 13)

transitional

= 0.074 R L-lI5

-

1700 R LP1

( 14)

At sustained speeds (Fig. 1A) R , for scombrids may be low enough for flow to be laminar over much or all of the body. At burst speeds (Fig. 1B) or sustained speeds for large bluefin tuna which are often above 1 x lOS, flow would become turbulent at some point along the body. Flows over the fins would be laminar because fins have short chords or lengths and consequently lower Reynolds numbers than the body. As mentioned earlier, drag coefficients measured by Bone (Webb, 1975) and those computed from Eq. (12) are similar (Table V). At R L from 7 x 104 to 1.4 x 105 the values were remarkably close to expectation with laminar flow. If turbulent flow were assumed, the estimates were usually higher than measured values but the fit was best at the highest and lowest RL.Laminar flow was assummed for calculations in Table VI for sustained speeds of skipjack tuna. I n spite of the prob-

Table V Coefficients of Friction Drag Measured from a 35 cm long Atlantic Mackerel by Bone in Webb (1975)Compared with Theoretical Values Computed from Eqs. (12)and (13)

Cf Speed (cm/sec)

RL

10

3.5 x lW

20

7.0 x 1W 1.0 x 1w 1.4x 1W 1.8 x 1w

30 40 50

Measured dead drag

1x 4x 4x 5x 5x

10-2 10-3 10-3 10-3 10-3

Calculated cf

laminar

7x 5x 4x 4x 3x

10-3 10-3 10-3 10-3 10-3

cf

luibulenl

9 x 10-3 8x 7 x 10-3 7~ 10-3 6x

270

JOHN J. MAGNUSON

Table M Total Drag Estimated for a 44 cm Skipjack Tuna Swimming at a Sustained Speed of 66 cmlsec" Type of drag

Part of body Body Second dorsal and anal fins Caudal fin Pectoral fins Gills Total

Friction [dynes

Form [dynes

Induced [dynes

Gills [dynes

Total [dynes

(%)I

(%)I

(%)I

(%)I

(%)I

4,720 (24)

1,180 (6)

0

-

5,900 (30)

500 (2) 1,340 (7) 870 (4)

120 (1) 1,080 (5) 700 (4)

0 2,970 (15) 2,900 (15)

-

-

-

-

7,430 (37)

3,080 (16)

5,870 (30)

3,400 (17) 3,400 (17)

620 (3) 5,390 (27) 4,470 (23) 3,400 (17) 19,780 (100)

-

Modified from Brown and Muir (1970) with recalculated values for gill resistance and fin drag and an inclusion of hydromechanical efficiency of the caudal based on Chopra (1975). Flow was assumed to be laminar over all parts and over the body. See text, Table XI, and the following tabulation (values of variables used in computation and their data source) for the calculations.

U S (cm/sec) (cm') Body* 66 Second dorsal and anal finsf 66 Caudal fin$ 73 Pectoral fins 66

840* 15 41 30 $

C

b

(cm)

(cm)

44 1.3 1.7 1.5f

12 24t

c,

RL

2.9 x 8.6 x 1.2 x 9.9 x

1W 101 104 101

0.0025 0.015 0.012 0.013

* Magnuson and Weininger (1978). f Brown and Muir (1970). $ Magnuson (1973) and Table X (this chapter). lems involved, additional measurements of dead drag and computations of theoretical drag on scombrids would be useful in future analyses of forces acting on a swimming scombrid. The coefficient of drag can be adjusted to include form drag. The thicker the body ( d ) in respect to length (8)the greater the form drag and so the greater the combined drag coefficient (C,) in respect to Cf. C , , the coefficient of frictional and form drag together, can be determined from (Hoerner, 1958, cited in Bainbridge, 1961)

c ,= C J 1 + is(d/e)+ 7(d/e)31

( 15)

Brown and Muir (1970), using test data from Hoemer (1965), judged form drag of the body to equal 25%of the body friction drag for

4.

LOCOMOTION BY SCOMBRID FISHES

271

skipjack tuna. They also estimated form drag for all extended fins (pectorals, second dorsal, anal, and caudal fins) as 50% of friction drag from the same sources. Since I am interested in the individual contribution of each fin I did not use the 50% average. Rather I used the figure of 80% suggested for the pectoral fins by Brown and Muir for pectorals and caudal fins. I used 25% for the anal and second dorsal that operate at no angle of attack. I n addition I used more recent values for the wetted areas of the exposed pectoral fins and the span and area of the caudal fin than had been used in Brown and Muir. Speed of the caudal fin was increased slightly (1.1 x U ) over swimming speed to account for its oscillatory path. The above computations do not include resistance to water flow from ram gill ventilation used by scombrids (Brown and Muir, 1970), nor do they account for induced drag from the lift produced on the pectoral fins (Brown and Muir, 1970) or the thrust from the caudal fin.

b. Gill Resistance. Scombrids ventilate their gills by swimming forward with mouth slightly agape. Water passing through the mouth, over the gills, and out the opercular gap produces internal drag and loses momentum. Brown and Muir (1970) estimated that speed of water through the respiratory path declined by about 30% from entrance at the mouth to exit at the opercular gap. As can be observed from Eq. (16), gill resistance (R.) is directly proportional to ventilation volume (VG),swimming speed ( U ) , water density (pe), and approximately to the head loss (AH):

This equation developed by Brown and Muir (1970) appears to provide a good fit to observed data (Stevens, 1972). The problem of accurate estimation is to obtain reasonable values for ventilation volume and head loss from a “difficult” experimental animal. Ventilation volume was estimated by Brown and Muir (1970) from observed rates of oxygen consumption, the fraction of oxygen removed from the water, and the oxygen concentration in water. They used VG of 56 ml/sec or 33 ml/sec . kg for a 44 cm skipjack tuna based on preliminary work. The value is probably low. Stevens (1972) was unable to keep restrained skipjack tuna alive for even 1 hr at similar rates of gill perfusion but had better luck at rates near 50 ml/sec * kg. Reasonable values for head loss seem to be about 2000 dynes/cm2 or about twice the value used by Brown and Muir (1970). Stevens (1972) directly measured head loss on restrained skipjack tuna (AH = 1885 dynes/cm2, n = 6, mass = 1.5-1.7 kg) and a restrained

272

JOHN J. MAGNUSON

kawakawa (AH = 1816 dynes/cm2, mass = 2.3 kg). Roberts (1975) has also determined for small Atlantic mackerel a minimum speed at which ram ventilation is used. This speed by his estimates would produce a head loss of 2100 dynedcm2. The best estimate of gill resistance is 3400 dynes for a 44 cm skipjack tuna. This estimate, about three times the first estimate by Brown and Muir, was obtained from Eq. (16) with pe = 1.025 g/cc, U = 70 cm/sec, AH = 2000 dynes/cm2, and VG= 85 ml/sec.

c. Induced Drag. Hydrodynamic lift from the pectoral fins and thrust from the caudal fin are not produced without some losses from tip vortices and trailing edge vortices (Fig. 9). Prandtl and Tietjens (1934),Alexander (1968),and especially Lighthill ( 1975)provide more complete discussion of the vortices. Not shown in the diagram is the lateral movement of the trailing edge vortices observed in scombrids by McCutcheon (personal communication, cited in Lighthill, 1975). Owing to greater pressure below the fin or on alternative sides of the beating caudal fin, some water flows around the tips of the pectoral fins or the caudal fin and forms a vortex behind the fish (Fig. 9). Also trailing edge vortices on the caudal fin result from the water accelerated backward by the caudal movement (Hertel, 1966; Lighthill, 1969, 1975). This induced rotating motion of water left in the wake adds to the drag. Induced drag (D,)from the pectorals is directly proportional to the required lift squared (L2).It is inversely proportional to the speed of and the span of the fins squared (b:). The magthe fish squared (U2)

@TIP VORTICES @TRAILING EDGE VORTICES

Fig. 9. Diagram of vortices induced at the tips of pectoral fins and at the tips and trailing edge of the caudal fin that are the sources of drag induced by production of lift and thrust. (Adapted from two-dimensional views-Prandtl and Tietjens, 1934, Fig. 154; Lighthill, 1975, Fig. 8.)

4. LOCOMOTION BY SCOMBRID

FISHES

273

nitude of induced drag can be estimated from an equation taken from Prandtl and Tietjens (1934):

D,=-

2LZ

mU2bP2

where E = an efficiency factor of approximately 0.85. Brown and Muir (1970) estimated induced drag from the lift on the pectorals at 2800 dynes for a 44 cm skipjack tuna with a fin span of 24 cm. Induced drag is slightly higher, 2900 dynes in my calculations because required lift I used was greater. Lift (100,000 dynes) was calculated from Eq. (8) from data in Magnuson (1973, Table V) for a fish with a stomach content equal to 5% of body weight. Preliminary estimates of drag resulting from thrust from the caudal fin can be obtained by methods developed by Wu (1961) and Lighthill (1969, 1975) and reiterated in Webb (1975). A hydromechanical efficiency factor (qp) for thrust production estimates the proportion of power exerted by the caudal fin that goes into producing thrust. One minus the efficiency factor ( 1 - qp)is used here as an estimate of losses owing to induced drag. The magnitude of the efficiency factor depends on the posture, form, and rate of movement of the caudal fin. Parameters for estimating efficiency factors were developed by Webb (1975, p. 98) and will be discussed in more detail under thrust production. The values for efficiency from Table 17 in Webb (1975) for kawakawa swimming 3-4.7 t/sec varied from 0.77 to 0.86 to give losses (1- qp)of 0.14-0.23. For present purposes I use 0.15 as the proportion of total thrust power lost to induced drag based on a reanalysis of data on kawakawa presented in the section on thrust. Total drag can be calculated, as can drag induced from thrust (0.15 x total drag) by summing all drag other than that induced by the caudal fin, and then solving for total drag b y Total drag = (0.15 total drag) drag from all sources other than induced drag from the cuadal fin (18)

+

Drag from all other sources than induced by the caudal fin was summed from Table VI.

d . Total Drag. The estimate of total drag on the 44 cm skipjack tuna swimming at sustained speed of 66 cm/sec was 19,780 dynes (Table VI). This includes frictional and form drag on the body and extended fins, induced drag on the pectoral fins and caudal fin, and gill resistance. Usually these components cannot be isolated and total drag regardless of source is sometimes related with a single drag coefficient

274

JOHN J. MAGNUSON

to the body surface area and speed (see, e.g., Pyatetskiy, 1970b; Sharp and Francis, 1976). For this skipjack the coefficient of total drag is 0.010 from 19,780/(0.5 x 1.025 x 840 x 66) in analogy with Eq. (11) for frictional drag. Total drag and coefficients of total drag were also estimated in the section on thrust for kawakawa swimming at faster speeds. Estimates were from summing components of drag and from calculations of total thrust which should equal total drag (Table XI). For R L of 5.0 X 1 8 1.3 x 108 total drag coefficients of kawakawa average 0.012 (range, 0.009-0.016) from the component of drag analysis and 0.027 (range, 0.011-0.044) from thrust analysis. The values from the two methods of estimation are within a factor of 2 of each other and are two and four times larger than the frictional drag coefficient for turbulent flow from Eq. (13). The coefficients for total drag appear to reach a maximum near R , of 7 x lo5(Table XI, column 27). Pyatetskiy (1970b) estimated coefficients of total drag based on calculations of thrust for 16 cm Atlantic bonito. At speeds between R L of 1 and 2 x 105 these values were near 0.010, but at speeds just below 1 x lo5 values near 0.04 were calculated. The best estimate of the coefficient of total drag for a swimming scombrid, genera Euthynnus and Sarda, is 0.01-0.03.

B. Distribution of Drag among Components 1.

SUSTAINED-SWIMMING

SKIPJACKTUNA

For skipjack tuna swimming at sustained speeds, the calculated magnitudes of drag components are given in Table VI. The first dorsal fin and pelvic fins were not included in the calculation because they are appressed at sustained speeds. Calculations were described above and values were taken or recomputed from Brown and Muir (1970) and Webb (1975). Values will certainly change as methods and data are refined. Similar values were obtained b y MacKay (1976) with slightly different components for Atlantic mackerel. Frictional drag and form drag on the body and fins constitute 53% of total drag b y my calculations. The remainder comes from induced drag from lift and thrust (30%)and from gill resistance (17%).Viewed in another way only 30% of the drag comes from the body while 53% comes from the various fins and again 17% &om the gill resistance. Drag resulting from the use of the pectorals to maintain hydrostatic equilibrium is 23% of total drag and somewhat greater than the 17% going to gill resistance. Together, these two functional aspects of continuous swimming account for 40% of the drag. Fish not required to

4.

LOCOMOTION BY SCOMBRID FISHES

275

meet these functions by swimming would be expected to have lower drag at the same speeds. The importance of adaptations to reduce drag from the production of thrust and lift become especially apparent when their magnitude is compared with the drag on the streamlined body. A total of 50% of total drag comes from the caudal and pectoral fins but only 30%comes from the body at sustained speeds. The most surprising conclusions from an analysis of the components of drag on a skipjack tuna at sustained speeds (Table VI) are (1) most resistance to forward movement comes from the fins, not from the body, and (2) the two functions of continuous swimming, hydrostatic equilibrium and gill ventilation, increase resistance to forward movement b y a factor of 1.7.

2.

CHANGES IN

DISTRIBUTION OF DRAG

The distribution of drag among sources (Table VI) will be expected to vary with speed of swimming, and among species and sizes of scombrids. a . Changes with Speed. Gill resistance will decrease with increasing speed because at burst speeds a scombrid closes its mouth. Closing the mouth would reduce drag b y about 17%, a considerable saving. Even at sustained speeds Pacific bonito (Sarda chiliensis) in the oceanarium at Marineland of the Pacific closed their mouth up to 42% of the time (Magnuson and Prescott, 1966) and may have thereby obtained a 7% reduction in drag (0.17 x 0.42 = 0.07). Drag produced by the pectoral fins will decrease to zero with increasing speed because at faster speeds they appress their pectorals for increasing proportions of time. For example, captive kawakawa, that had just eaten, extended the pectorals 100% of the time while swimming at 1.6 tlsec but 0% of the time at 2.8 Msec and faster (Magnuson, 1970). At burst speed, drag on pectorals would b e zero. Also, the ratio between form and frictional drag on the body should decrease at faster speeds because at Reynolds numbers near 106 the change of flow from laminar to turbulent should decrease the form drag (Alexander, 1968). Thus, at burst speeds, relative savings in drag would be expected from absence of gill resistance and absence of drag on the pectorals and from reduction in form drag on the body. However, even though gill resistance and the pectorals contribute 30%of the drag of a slowly swimming scombrid (Table VI), saving at high speeds would be small because gill resistance and induced drag on the pectorals would re-

276

JOHN J. MAGNUSON

main relatiGely constant with increase in speed. Drag on a burst swimming scombrid, while not calculated here, would be dominated by friction on the body and induced drag from the caudal fin (see Table XI).

b. Changes with Species and Size. Most forms of drag are proportional to speed squared so drag would be lower for species with slower sustained speeds. Species with larger gas bladders and larger pectoral fins have slower sustained speeds (Magnuson, 1973) and should have less drag as a consequence. Differences in distribution of drag among components in different species also will result from differences in morphometry, especially morphometry associated with the determination of the minimum speed necessary for hydrostatic equilibrium. Induced drag on the pectorals is directly proportional to weight squared [Eq. (17)] and induced drag on those species that were lighter would tend toward zero and could reduce total drag by up to about 15%. In addition induced drag on the pectorals is inversely proportional to the span squared, and species with long fins such as yellowfin tuna should have less drag from the pectorals for that reason alone. However, induced drag is also inversely proportional to speed squared and the slower sustained speed of species with long fins would cancel some of the savings. Species with larger fins should also have more frictional and form drag on their fins than those with smaller fins. Frictional drag is directly proportional to surface area and pectoral areas differ markedly among species. For example, areas of pectoral fins are four times greater for bigeye tuna than for wahoo (Magnuson, 1973), and two times greater for small albacore than for skipjack tuna (Dotson, 1977). Similar differences would occur among sizes of the same species because growth of gas bladders and pectoral fins are allometric (Magnuson, 1973; Dotson, 1977). For example, yellowfin 50 cm long have no gas bladder and have faster sustained speeds than larger yellowfin that do have a gas bladder. Consequently small yellowfin should have a greater proportion of drag as induced drag than would larger yellowfin. Species with more optimum body shapes would have reduced form drag on the body. Many scombrids, Scomber, Sarda, Euthynnus, Thunnus, and Auxis, have shapes near optimum for sustained swimming, but others such as Acanthocybium have bodies more slender than optimum (Magnuson, 1973). In summary, species and sizes with more optimum body shapes would have reduced form drag on the body, those with adaptations for slower sustained speeds would have less induced drag from the pec-

4.

LOCOMOTION BY SCOMBFUD FISHES

277

toral fins, and those with relatively smaller fins in general would have less frictional and form drag associated with the fins. Since all of these factors are interactive, the net effect of all differences would have to be considered in concert with one another.

C. Adaptations for Drag Reduction Energetic consequences of adaptations that reduce drag are selfevident. Among scombrids with their continuous swimming and their continuous production of thrust with the caudal fin, lift with the pectorals, and drag from ram gill ventilation, adaptations for drag reduction have developed to perhaps a greater extent than in any other group.

1.

SHAPE AND STRUCTURE

Most scombrids have nearly ideal longitudinal streamlining of body shape (Fig. 10) in terms of thickness ratio measured here by maximum thickness as a percentage of length. Larger estimates for thickness are obtained if only maximum height is used rather than when maximum width and height are averaged, but the latter seems more representative of shape and is used here. For bodies of the same volume, drag is minimum when the body thickness is about 22% of length and is increased by less than 10% when between 14 and 33% (Alexander, 1968; von Mises, 1945; Webb, 1975; Hertel, 1966). Three genera, Thunnus, Euthynnus, and Auxis (Table VII), have body thicknesses between 21 and 23% of length or have almost exactly the optimum shape. Sarda and Scomber are somewhat more slender with a thickness of 1618%.Acanthocybium with a thickness of 12%is the only species for which data are published that is too slender to have minimum drag per unit volume. Scomberomous has a shape similar to

Acanthocybium. A second feature of body shape, the position of maximum thickness along the length, is an index to the proportion of the body over which laminar flow can be expected at high speeds. A laminar profile has the thickest portion farther from the snout and tends to maintain laminar flow over a greater length of the body. Scombrids have laminar flow profiles with the position of maximum thickness 4&50% (Table VII) of the distance from the snout to the tail. A 70% figure given by Webb (1975) originates from a single photograph from a semidorsal view of an unidentified miniature tuna hanging from a hook (Fig. 166 in Hertel, 1966). I think this value is an overestimate. Values in Table VII,

278

JOHN J. MAGNUSON

1 1

30

B

0

20

40

I

I

60

80

0 100

P e r c e n t o f B o d y L e n g t h , E x c l u d i n g the Keel

Fig. 10. Body shape of kawakawa 32 cm fork length showing,(A)cross section shape and (B) longitudinal disbibution of average height calculated by dividing cross section area by maximum width. (From Magnuson, 1970, Hydrostatic equilibrium ofEuthynnus nfinis, a pelagic teleost without a gas bladder, Copein pp. 56-85.)

usually about 41%, are probably underestimates because the length used included the length of peduncle (Aleev, 1963) and the caudal fin (Magnuson, 1973). Exclusion of the caudal fin gives numbers near 46% and an estimate for kawakawa with length terminated just anterior to the keel is about 50% (Fig. 10). Cross sections of the body of a scombrid are a series of vertical ellipsoids along most of the length (Fig. 10). Anteriorly the body is wider above the midline. Just anterior to the peduncle it becomes first circular and, at the peduncle, well-developed lateral keels form a horizontal ellipse which in some species has sharp extended edges. Horizontal keels at the peduncle probably reduce turbulence from the rapid lateral movements that occur here as the tail beats back and forth (Walters, 1962; Aleev, 1963; Webb, 1975). Finlets along the dorsal and ventral midline direct flow across the peduncle (Magnuson, 1970; Webb, 1975).The vertically thin, horizontally reinforced peduncle, with finlets to direct flow, appears to be an adaptation that reduces form drag during lateral movements and still provides a strong, me-

4. LOCOMOTION

279

BY SCOMBFUD FISHES

Table VII Longitudinal Streamlining of the Bodies of Scombrids Distance from snout to position

Maximum maximum body thickness height (% 0 (% 0

of

Species"

Thunnus Bigeye tuna Yellowfin tuna Albacore Median Euthynnus Kawakawa Skipjack tuna Median Aunis Bullet mackerel Frigate mackerel Median Sarda Pacific bonito Atlantic bonito Median Scomber Atlantic mackerel Acanthocybiurn Wahoo

tb (cm)

Maximum body thickness, (height + width)/2 (% 0

Source

4660 28-45 68

37 41 46 41

28.6 27.8 26.6 28

23.5 22.3 23 (4.3)c

Magnuson (1973) Magnuson (1973) Aleev (1963)

37-43 37-57

40 42 41

27.2 25.9 27

22.6 21.0 22 (4.6)e

Magnuson (1973) Magnuson (1973)

32-37 39

40 46 43

24.3 23.0 24

20.8 21 (4.8)"

Magnuson (1973) Aleev (1963)

45-65 14-45

40 34-49 41

21.8 21.0 21

18.4 18 (5.6)?

Magnuson (1973) Aleev (1963)

12-25

4G50

19

16d(6.2)'

Aleev (1963)

123-130

37

15

12 (8.4)?

Magnuson (1973)

Species listed by genera in order of decreasing thickness. Aleev (1963) used length of skeleton. Fineness ratio ( t h a x i m u m thickness) is given in parentheses for each genera. Estimated by reducing maximum height by 3% to give maximum thickness.

chanically advantageous link (Fierstine and Walters, 1968) between the body musculature and the caudal fin. Internally the tendons to the fin pass over the bony lateral extensions which serve as a pulleylike bearing surfice. Other functions of the keel were discussed above under hydrostatic considerations. The body is smooth. Scales are usually absent except in a small corselet immediately behind the opercular opening. The eyes, nares, jaw, and opercular bones provide extremely smooth exterior surfaces. Pelvic fins and first dorsal fin are used only in maneuvers (Magnuson, 1970).During sustained swimming they are appressed into grooves or

280

JOHN J. MAGNUSON

depressions in the skin surface which are perfectly shaped to receive the appressed fins and retain a smooth body surface. At burst speeds the pectorals are also appressed into grooves on the side. Walters (1962) speculated that the corselet of scales and the thick skin just behind the opercular gap might act as a tripping wire to reduce form drag, but Webb’s (1975) idea that it provides the fairing for the appressed pectoral fins seems more likely. Areas between the fin rays on the caudal fin, pectoral fins, and the anal and second dorsal fins are filled with tissue and provide a smooth surface to the water. Fish slime when mixed with water can reduce friction of water in turbulent flow through a tube (Rosen and Cornford, 1971). But slimes of scombrids were the least effective of eight species tested and were considered exceptions. Slime from Pacific mackerel (Scomber japonicus) was not effective until the solutions were 50% slime. For Pacific bonito even 100% slime did little to reduce friction over ordinary seawater. Slimes as suggested by Breder (1926) might streamline the body by reducing surface irregularities, but the slippery nature of slime itself is probably unimportant, at least to scombrids. Breder (1976) argues that slimes are more effective in solitary than in schooling fishes like scombrids. Various methods of damping disturbances in the flow with unique structures of the skin have been proposed (Walters, 1962) to help maintain laminar flow. Webb (1975),in a review of evidence, doubts if they occur in scombrids and Aleyev (1977) presents interesting information that indicates no drag reduction even when they do occur in mammals, Walters (1962) suggested that a scombrid body with ram ventilation acted like a leading edge slat with water entering the mouth and being injected to the boundary layer behind the opercular gap with favorable effects. The injection of high energy water from the gills would delay boundary layer separation and decrease surface drag. Whether this applies is not clear. Brown and Muir (1970)calculate that water coming from the gills would be lower in energy and it thus might not have the desired effect. Rather it appears that gill effluent causes turbulence (Webb, 1975). Further discussion can be found in Walters (1966) and Breder (1965). The shapes of the pectoral fins (Magnuson, 1970) and caudal fins (Fierstine and Walters, 1968) are characterized by high aspect ratios. The aspect ratio (AR) is the ratio between span and mean chord. Induced drag is less per unit of lift or thrust for fins with higher aspect ratios. As indicated by Eq. (17) induced drag is inversely proportional to span squared. For pectoral fins of kawakawa AR of the exposed area

4.

LOCOMOTION BY SCOMBFUD FISHES

281

ranged from 6 to 7 (Magnuson, 1970). If the span across the body is included, AR is greater than 10. Kawakawa have relatively short fins for a scombrid, and species with long pectorals such as the thunnids have AR well over 10 by either method of computation. For the caudal fin Fierstine and Walters (1968)measured AR that ranged from 5.2 to 7.7 from nine scombrid species. These will be compared in more detail with more recent estimates in the section on thrust production. But with the exception of fast swimming istiophorids such as sailfish and marlin they are the highest known among bony fishes. Pectoral fins are high (>lo) and the caudal fin median (5-7) AR devices when compared with flying animals and airplanes. AR for planes ranges from 2 to 15 (Perkins and Hage, 1949) and from 3 to 12 for birds (Hartman, 1961; Storer, 1948; Greenwalt, 1975). The greater AR of pectorals may explain in part the observed higher lift : drag ratio of the pectorals compared to the caudal fin. For example, lift : drag and thrust :drag ratios of a 44 cm skipjack tuna can be calculated and are 22 : 1 (100,000: 4470) for pectoral fins and 4 : 1 (22,870: 5390) for the caudal fin. The 22 : 1 lift :drag ratio for the pectoral fins is somewhat above the maximum of 17 : 1 for a thin cambered airfoil (see von Mises, 1945, Fig. 109). The lift coefficient for such an airfoil operating at a maximum lift : drag ratio would be about 0.9 (von Mises, 1945, Fig. 109) and is in the range of estimates for swimming kawakawa (Table IV). The estimates of thrust : drag ratio for the caudal fin was not nearly as high as for the pectorals. While it provides no further explanation for the difference, it is interesting that the vertical forces in sustained swimming are much greater than horizontal forces. Total drag force on the 44 cm skipjack swimming at 66 c d s e c are about 20,000 dynes or one-fifth of the total lift force of 100,000 dynes required by the fish to maintain hydrostatic equilibrium. In summary, two general adaptations characterize the shape and structure of scombrids in regard to drag reduction. These are (1) smooth streamlining of the body and fins, and (2)high aspect ratios for fins that produce lift or thrust.

2. PHYSIOLOGICAL ADAPTATIONS Any increase in efficiency of oxygen uptake could contribute to drag reduction. Gill resistance in Eq. (16) was directly proportional to respiratory volume, Given a particular metabolic rate, the less water that was required to ventilate the gills, the lower the drag from ram gill ventilation. Scombrids have highly specialized gills with surface areas similar to small mammals of equal mass and almost an order of mag-

282

JOHN J. MAGNUSON

nitude greater than most other fishes (Muir and Kendall, 1968; Muir, 1969; Muir and Hughes, 1969; Muir and Brown, 1971). Oxygen removal from ventilated water (Stevens, 1972) averaged 71%for six skipjack tuna and 79%for one kawakawa. Averages for individual skipjack tuna ranged from 47 to 94%.These uptake efficiencies are adaptations of scombrids and are much higher than for other fishes. If they were half as efficient, had twice the ventilation volume, gill resistance would double its present level. Webb (1975) discounted the hypotheses by Walters (1962) that tuna could warm their boundary layer sufficiently to decrease viscosity.

3. BEHAVIORALADAPTATIONS Several exciting ideas have been developed that if true would be significant adaptations to reduce the energy required of scombrids to swim through the water. These are that properly spaced schooling fish cancel the vortices in each other’s wakes (Breder, 1965,1976; Belyayev and Zuyev, 1969; Weihs, 1973;, 1975) and that alternating swimming and gLJing reduces energy expenditures for a given distance covered (Weihs, 1973b, 1974). a. Schooling. Scombrids with rare exceptions are schooling fishes (Magnuson, 1963) of the obligatory type as defined by Breder (1967). Polarized, parallel orientation is maintained except during feeding and courtship (Magnuson and Prescott, 1966) by these continuous swimmers with an attraction for each other. Schools are composed of animals of the same species and approximate length (see references in Magnuson, 1963). Even in so-called mixed schools of yellowfin and skipjack tuna, parallel orientation is among those of the same species (Yuen, 1962). In such a formation the caudal fin of each fish sheds two rows of trailing edge vortices spinning in opposite directions in the wake (Fig. 11). These vortices degenerate and fade as the fish pulls away. Weihs (1973a) points out that one fish swimming directly and closely behind another would have to swim at a faster relative speed to counter the backward directed flow in the wake of the first fish (Fig. 11). This tandem position is avoided by the nearest neighbor as demonstrated in an aerial photograph of about 100 bluefin tuna (Thunnus thynnus) in which only four appear to be in tandem (Breder, 1965). Rather their spacing was more in a diamond or diagonal pattern more typical of schooling fishes (Breder, 1976). In the diamond pattern (Weihs, 1973a, 1975) the second row of fish in a school swims diagonally behind and midway between fish in the

4.

283

LOCOMOTION BY SCOMBRID FISHES

ROW 1

ROW 2

-ROW 3

Fig. 11. Diagram of scombrid school showing possible interaction between trailing edge vortices and swimming efficiency (modified from Weihs, 1973a, 1975). Distances and angles between fish based on Atlantic mackerel (van Olst and Hunter, 1970). See text for rationale. Not diagrammed is Weihs' (1975) suggestion that fish abeam from each other beat in opposition.

first row and obtains a swimming advantage because these fish are traveling in the direction of vortex rotation induced by the first row of fish (Fig. 11). In analyses of overhead photographs of small schools of captive kawakawa, few of the nearest neighbors swam in the abeam position (Cahn, 1972). The direction to the nearest neighbor in small schools of captive Pacific mackerel averaged 14" from the heading of the following fish (van Olst and Hunter, 1970). A fish in the optimum position (Weihs, 1975) behind the first row of fish could have its relative speed reduced by 50-60%. Since drag is proportional to speed squared, the possible savings in energy is even greater. Weihs (1973a) goes on to show that fish swimming in third or later rows in a school would be swimming through water in which the vortices from fish in the front two rows tend to cancel each other (Fig. 11) if the lateral distance to the abeam fish is twice the width of the vortex trail. By his calculations the energy savings of an entire school would be one-half the savings of the second row discussed above. Tip vortices from the pectoral fins may also produce a favorable

284

JOHN J. MAGNUSON

result in schools, Weihs (1975) points out that in the diamond pattern the tip vortices from the pectoral fins produce upwash from which following fish may receive some lift or reduced induced drag on their pectorals. In birds, a 71% reduction in induced drag is suggested (Lissaman and Schollenberger, 1970). Weihs ( 1973a, 1975) discussed still another hydromechanic advantage of schooling. Fish directly abeam and above or below may act as streamlined walls and as such should also positively influence thrust without any change in energy expenditure or caudal movements. This effect is greatest when fish are close together and is essentially absent at lateral distances of 1 e. Increases in thrust could be 10% or more. Estimates of the lateral distance between the longitudinal axis of nearest neighbors range from 0.4 to 0.6 C for adult scombrids in schools (bluefin tuna, Breder, 1965; Pacific mackerel, van Olst and Hunter, 1970; kawakawa, Cahn, 1972). Greater distances of 2.4 e were observed for subadult Pacific mackerel, 5 cm long, (van Olst and Hunter, 1970) and 0.6-22 e for bluefin tuna (Anonymous, 1975). When kawakawa were separated by transparent Plexiglas partitions they tended to swim directly abeam of the nearest neighbor at a lateral distance of 1.3-2.9 e and Cahn (1972) concluded that hydrodynamic clues were necessary for normal spacing. The above observations suggest that scombrids often space themselves appropriately in a school to take advantage of the savings described by Weihs. Data to substantiate the energetic advantages of swimming in a school do not exist for scombrids but Zuyev and Belyayev (1970) observed for the similarly shaped horse mackerel, (Truchurus mediterruneus), that fish of the same length that were farther back in the school swam at the same speed as lead fish, but with a slower rate of tailbeat. Tailbeat rate of the second, third, and fourth fish in the group averaged 88, 85, and 83%, respectively, of the tailbeat rate of the lead fish. These results suggest a possible savings of ca. 10-20% from schooling.

b. Gliding. Gliding downward with no swimming movements and then swimming to regain altitude should result in significant energy savings by negatively buoyant fish according to a theoretical analysis by Weihs (1973b). Using kawakawa about 42 cm long as an example, he estimated savings of 20% to travel the same horizontal distance in 7% more time. This comes out to about a 25% savings in energy expended per unit time. The pattern of swimming proposed by Weihs (1973b) is diagrammed in Fig. 12. The fish glides down from A to B at an angle of descent, a,and then actively swims to the surface at angle p to point C.

4. LOCOMOTION

285

BY SCOMBRID FISHES

WATER SURFACE

c B

Fig. 12. Diagram of path of a negatively buoyant scombrid alternating gliding and active swimming-a behavior that could decrease energy used per unit time (modified from Weihs, 1973b). Terms and rationale in text [see Eqs. (18)-(23)l.

Horizontal distances traveled are a , for descent and u2 for active swimming to regain altitude. The maximum and minimum depth of swimming are separated by distance h. Angles, distances, and energy savings (Table VIII) can be estimated from the following equations derived by Weihs (1973b) for the path with maximum energy savings per horizontal distance traveled:

D L

a = arc sin -

(18)

1 -a p = arc cos k,

( 19)

Table VIII Optimum Swimming Behavior and Expected Energy Savings of a Negatively Buoyant Euthynnid and Thunnid Obtained by Alternating Gliding and Active Swimming as Described in Fig. 12 and TexP Equations (18)-(23)

State variables

e Species Skipjack Albacore

L

(cm) (dynes) 45 45

D

h

(dynes) k I b I(m)

100,OW 20,OoOe 100,OOod 5,OW

1.2 100 1.2 100

a

p

a,

u2

(m)

(m)

12" 21" 470 260 0.3" 33.2" 20,000 150

Energy savings (proportion)

0.07 0.16

T

1.04 1.00

" Model from Weihs (197313). k, based on hydromechanical efficiency of qp = 0.15 from Table XI. Approximate values for fish in Table VI. Approximate values based on manuscript by Dotson (1977) that estimated minimum speed of 45 cm albacore as about 0.5 that of skipjack tuna. I assumed a drag reduction to 0.25 of drag from skipjack tuna since most drag values are proportionak speed squared.

286

JOHN J. MAGNUSON

where D, drag during gliding; L , excess weight; kl, ratio between swimming and gliding drag or about 1.2 for a scombrid (estimated from Table VI).

a , = h cot a

( 20)

p

(21)

tan a

(22) (1+Ga) sin a

a2 = h cot Energy saving = 1 -

sin

p + (tan a)(cos p)

or the proportion of energy saved compared with level swimming. 7=

+

sin a sin /3 sin(a + p)

(23)

where r , ratio of time to travel a given horizontal distance by sinking and rising versus level swimming. The euthynnid with no gas bladder and with small pectoral fins would sink rapidly in a glide (ca. 14 cm/sec) and rise steeply to obtain maximum advantage (Table VIII). The thunnid, here an albacore with gas bladder not yet developed but with large pectoral fins, sinks slowly (0.2 cm/sec) and rises steeply. Possible energy savings are 16% by the albacore compared with 7% for the skipjack tuna (Table VIII). Weihs (1973b) points out that greater savings are to be expected for fish that are capable of smaller glide angles. Also the'greater the ratio (k,)between drag during swimming and gliding, the greater the possible savings. Variations in angle of ascent (p) on the other hand have little influence over calculated energy savings. Scombrids with large well-developed gas bladders could have a problem with this mode of swimming near the surface because the rapid ascent can cause over expansion of the gas bladder. But the mode of swimming is quite reasonable over shorter vertical distances at water depths near 100 m or so (at the thermocline in the Central Pacific) where vertical displacement has less influence on gas volume than it does near the surface (Boyle's law). A number of scombrids are believed to live in the region of the thermocline (see Chapter 5 ) and conceivably they could swim and glide to their advantage in a path between lower and upper avoidance temperatures. Weihs (1973b) notes that observational data are not available to verify whether scombrids use this mode of swimming. Strasburg (1961)recorded that skipjack tuna schools in the Central Pacific often disappear from the surface and reappear every 3-38 min. A skipjack tuna carrying an acoustic transmitter in the Central Pacific was deeper than 4 m 39% of the day and 3% of the night hours (Yuen, 1970).

4.

LOCOMOTION BY SCOMBFUD FISHES

287

Skipjack tuna then do change depth, but the pattern of this movement is not known with enough precision to test Weihs’ idea. However, the ideas raised by Weihs (1973b) are attractive and seem plausible for scombrid fishes. Weihs’ (1973b) ideas also raise the possibility that scombrids can soar in vertically moving water like birds do in rising air (Cone, 1962; Lighthill, 1974) by gliding. The possible sites for soaring are two, (1) in the surface waves, and (2) in internal waves at the thermocline. In surface waves vertical velocities of water particles (Sverdrup e t al., 1942) are greater than the rates of sinking expected for gliding scombrids. The skipjack tuna and albacore in Table VIII should sink at 14 and 0.2 cm/sec, respectively. Vertical velocities in surface waves are of this magnitude as deep as 2-20 m below the surface for a wave as low as 1 m high. The major problem with soaring in a surface wave is that its horizontal velocity (624 cm/sec for a 1 m high wave) is so fast that a gliding scombrid probably would be left behind. However, the prolonged speeds recorded for some scombrids at sea are as high as 400 cm/sec (Commercial Fisheries Review, 1969). The possibility that scombrids can glide this fast should be considered especially if an occasional burst of tailbeats is used to keep up with the progress of a surface wave. Internal waves are out of consideration for euthynnids because they sink too fast, but are more promising for thunnids that sink slowly while gliding. Vertical velocities of internal waves near the thermocline in Sverdrup et al. (1942, Fig. 155) are as fast as 0.5 cm/sec and forward progress of such waves are apparently in the order of 60-200 cm/sec. The thermocline may be an especially suitable location for large thunnids to meet needs of gill ventilation, migration, and food search with little muscular effort. If this were true negative buoyancy (necessary for gliding) would be energetically advantageous compared to neutral buoyancy. Another possible technique of burst swimming and gliding may result in energy savings even for neutrally buoyant fish (Shebalov, 1969; Weihs, 1974). It is again based on the ratio between the drag on gliding and swimming fish and, according to Fig. 2 in Weihs (1974), could save as much as 20%for a fish such as a scombrid. Savings nea; 50% are projected for fishes that produce more drag during active swimming than scombrids do. The slower the average speed and the smaller the variation between burst and gliding speeds, the greater the expected savings. Speeds of 1-2 lengths per second were in the range to provide significant savings in calculations on nonscombrids. In my experience, a captive scombrid does tend to beat its tail regularly for a few seconds and then miss a few beats during steady swimming at a

288

JOHN J. MAGNUSON

constant depth, I do not know whether such behavior qualifies for the savings described above. Sustained speeds of scombrids also fall within the general magto provide the nitude of optimum speeds proposed by Weihs (1973~) greatest distance traveled per unit of energy expended.

D. Summary Scombrids are rich with adaptations to reduce drag or to move efficiently through the water with minimum energetic costs. All external structures are streamlined. Pectorals and caudal fins have high aspect ratios with a resulting reduction in drag induced from lift and thrust production. Gill resistance is reduced by a remarkably efficient uptake of oxygen from ventilated water. Behaviors such as schooling, gliding, and possibly even soaring, promise great savings based on theoretical analyses. Drag forces have apparently exerted intense selection during scombrid evolution. Yet, drag resulting from continuous swimming, induced from lift production on the pectorals and resulting from the gills by ram ventilation, imposes on scombrids sources of and a persistence of drag generation not faced by many other groups of fishes. This drag is large and must be balanced by a continuous production of large amounts of thrust, to be discussed next.

VI. THRUST PRODUCTION Scombrids have the extreme form of carangiform locomotion, name-

ly, carangiform with lunate tail (Lighthill, 1969, 1970, 1975; Wu, 1971; Webb, 1975; Aleyev, 1977) (see Chapter 1). The name comes from the carangid fishes that swim with movements confined to the posterior half of the body (Breder, 1926), and from the high aspect ratio caudal fin which has the appearance of a quarter moon. Lighthill (1975) sees three stages in the evolution of the lunate tail and Aleyev (1977) compares the differences among fishes. First is the development of the highly forked tail of many fishes such as the herrings (Clupeidae). No thrust should be lost by the loss of surface area between the lobes because the vortex-sheet in the gap functions as part of the fin, but drag should be reduced by the reduction of surface area. Second is a decrease in sweepback such as found in many carangids. This results in increased span without increased area, and increases thrust with less drag. Third is the development of the high aspect ratio caudal fin with the lunate form which Lighthill believes to be the

4.

LOCOMOTION BY SCOMBRID FISHES

289

culmination of adaptations for speed and efficiency of aquatic swimmers. The adaptation is shared by some marine mammals, the fast swimming lamnid sharks, and with the extinct reptile Ichthyosaurus. Thrust is generated in the same general way as in carangiform swimming but is concentrated, not in the posterior half or third of length, but in the enlarged caudal fin that only occupies the posterior 10% of the length of scombrids. Amputation of the caudal fin rays from two yellowfin tuna reduced the pull that each exerted on a hand-line by about 90% (Fierstine and Walters, 1968). The little pull that remained was believed to come from the bases of the rays and the hypural supports that had not been amputated.

A. Anatomy

The shapes of the body and peduncle minimize the disturbance to the water encountering the caudal fin (Lighthill, 1969).The vertically narrow peduncle, referred to as narrow necking b y Lighthill, imparts minimum motion to the water immediately anterior to the caudal fin and the streamlined body should induce little turbulence. Flow around scombrids has not been studied but pictures of flow around the Pacific white-sided dolphin (Lagenorhynchus obliquidens) (Rosen, 1961) show surprisingly little disturbance, and photographs oi flow around a 14 cm long horse mackerel (Trachurus mediterraneus) indicate that turbulence formed behind the fish would affect primarily the middle of the caudal fin in the middle of a sweep (Aleyev and Ovcharov, 1973). I n the dolphin photo the cross-flow appears to be smooth across the flattened keel rather than turbulent as would be expected from flow across a less streamlined structure. The keeled peduncle of scombrids should perform as well in horizontal movement as that of a dolphin in vertical movement. Breder (1926) observed that the body of Atlantic mackerel had little lateral movement compared to the tail. Measurements by Fierstine and Walters (1968) on kawakawa indicate that the anterior half of the body moves laterally 0.02-0.03 +?or only about 10% of the caudal fin’s lateral excursions. Minimizing these lateral movements of the body not only reduces the disturbance caused by the body in swimming movements but also maximizes the lateral movement of the caudal fin for a given muscular contraction. The lateral movement of the body or recoil that results from the lateral forces exerted by the caudal fin is reduced by longitudinal distribution of body mass and the body profile (Lighthill, 1969). The mass of the fish is concentrated well forward and the body is a series of

290

JOHN J. MAGNUSON

vertical ellipses (Fig. 10). They are not round in cross section. For seven scombrids maximum height was 1.49 (range 1.40-1.65) times maximum width (Magnuson, 1973). A greater ratio of height to width increases the virtual mass of water that would be displaced by lateral movement of bodies with equal volume. The second dorsal fin and anal fin maintain and accentuate the larger vertical dimensions posteriorly to a point about two-thirds the distance from the snout to the fork of the caudal fin. This position is only a short distance anterior to the location of maximum bending of the body in swimming movements. We built a small mechanical skipjack at the National Marine Fisheries Laboratory in Honolulu that swam with oscillating caudal movements. Removal of the artificial second dorsal fin and anal fin increased lateral movement of the body and decreased even more its feeble swimming. The importance of the median fins in this regard can be easily visualized in Gray’s mechanical spring model (Gray, 1968, Fig. 2.11) which shows that a strong lateral force develops at the position of the anal and second dorsal fins during a sweep of the tail. The caudal fin itself is a tapered hydrofoil with high aspect ratio, curved leading edge and moderate sweepback (Fig. 13D). In cross section the fin is shaped like a thin symmetrical airfoil with a rounded anterior or leading edge and a sharp posterior or trailing edge (Chopra, 1975). The rounded leading edge is essential to leading-edge suction that results from the low pressures that occur here as water flows more rapidly around the thickened section. The taper of the fin in combination with moderate sweepback (Fig. 13D) concentrates the surface area near the three-quarter chord location. The front of the fin is pulled from left to right by the lateral movements of the peduncle. Since the area of the fin is to the rear, the fin changes its angle of attack around a yawing axis that is near the rear. An axis near the three-quarter or full chord location would be optimum for efficient thrust production based on theoretical work by Lighthill (1970, 1975), Wu (1971), Chopra (1975), and Chopra and Kambe (1977). The Atlantic mackerel (Aleev, 1963) and other scombrids have two deflector keels near the base of the caudal fin that may help direct flow smoothly along the chord. Aspect ratios of caudal fins range from about 4.5 to 7,O among adult scombrids (Table IX). Skipjack tuna have the highest, 7.2, and Japanese mackerel have the lowest, 4.6. Genera with the highest AR are Euthynnus, Auxis, and some Thunnus, and with the lowest are Scomber, Acanthocybium, Sarda, Scomberomous, Allothunnus, and

4. LOCOMOTION BY SCOMBEUD FISHES

291

PATH

OF

C

FISH

L L . I I I

w-? CAUDAL FIN PATHOF

D

Fig. 13. Diagram of the shapes, postures, and movements of the,caudal fin. (A) Tracings of top view of swimming kawakawa during one-half of a complete tailbeat (adapted from Fierstine and Walters, 1968). (B) Posterior view of swimming kawakawa showing dihedral angle (adapted from Fierstine and Walters, 1968). (C) Diagram of path of an oscillating caudal fin showing amplitude (A), wave length (A), and fin postures (adapted from Gray, 1968). (D) Lateral view ofcaudal fin showing surface area (S J, span (b),sweepback angle (A), and a cross section and chord (c)at one location along the span. (E) Diagram for caudal fin of angle of attack (a)and angle of feathering (0).

some Thunnus. Those with higher AR should be the more efficient and faster swimmers (Nursall, 1958; Kramer, 1960; Aleev, 1963; Lighthill, 1969, 1975; Wu, 1971; Webb, 1975; Aleyev, 1977), all other things being equal. The higher the aspect ratio, the longer the leading edge that is important in leading-edge suction. The longer trailing edge increases the mass of water deflected posteriorly by the fin’s lateral movement. Induced drag from tip vortices is also less important rela-

292

JOHN J. MAGNUSON

Table IX Aspect Ratios (AR) of the Caudal Fins from Seventeen Scombrids Listed in Order of Decreasing AR ARa

6.8-7.2

6.3-6.7 5.8-6.2

5.3-5.7 4.8-5.2 4.3-4.7

Species

Euthynnus pelamis (skipjack tuna) Auxis rochei (bullet mackerel) Euthynnus afinis (kawakawa) Thunnus albacares (yellowfin tuna) Thunnus obesus (bigeye tuna) Euthynnus alletteratus (little tunny) Euthynnus lineatus (black skipjack) Auxis thazard (frigate mackerel) Allothunnus f a h i (slender tuna) Thunnus alalunga (albacore) Sarda chiliensis (Pacific bonito) Scomberomorus sierra (sierra) Sarda sarda (Atlantic bontio) Scomber scombrus (Atlantic mackerel) Acanthocybium solanderi (wahoo) Thunnus thynnus (bluefin tuna)

-

Scomber japonicus (Japanese mackerel)

Sourceb

3,4 4 4 1, 3, 4 4 1 1, 3 2, 3 3 2, 3 3, 4 1, 3 2 2 3 3

3, 4

Values for AR (spanVcauda1 area) differed greatly among authors owing to method of measurement. Data were adjusted for bias as follows for each source: (1) + 1.8, (2) +0.8, (3) -0.6, and (4) +0.6. These were based on the differences between species where the same species was estimated by two authors and on the assumption that the mean between sources 3 and 4 was the best estimate. Fierstine and Walters expanded the fin as much as possible for measuring span; Magnuson did not. * Sources: (1) Nursall (1958); (2) Aleev (1963); (3)Fierstine and Walters (1968); (4) Magnuson (Table X and Fig. 19, this chapter).

tive to lift production in high aspect ratio hydrofoils. With high AR more of the caudal fin extends beyond the wake of the body. Aspect ratio increases with the length of the fish (Table X, Fig. 14C) and values near 8 characterize the largest skipjack tuna. The AR of Atlantic mackerel also increases with length from about 3 when 12 cm long to 5 when 25 cm (Aleev, 1963). The report that the AR’s of yellowfin tuna and skipjack tuna do not change with length (Fierstine and Walters, 1968) appears to result from a sample with too narrow a range of lengths. An increased AR and efficiency at larger sizes would be of value, since larger animals should encounter more drag per unit length than do smaller ones. Body surface area, important in friction drag, increases as P . Also longer fish are more likely to swim at faster speeds.

4. LOCOMOTION BY

293

SCOMBRID FISHES Table X

Relation between Size and Form of the Caudal Fin and Fork Length (t)of Seven Scombrids: Sweepback (A), Aspect Ratio (AR), Caudal Span ( b J ,and Caudal Area for Thrust (S,)" Correlation coefficients significant at p 5 0.01 Sample size

Scomber japonicus

A and

e

AR and C

Area (cm?

-

-

b , = 0.54+ 0.2&

s, = 0 . 0 1 ~ ~ ~ 7

14

-

-

b , = -2.22

S,

11

-

-

+ 0.2% b , = -3.24 + 0.3M

36

-0.73

+0.70

b , = -2.12

+ 0.34t

S,

31

-0.73

+0.77

b , = - 1.94 + 0.31t

31

-

+0.62

b , = -2.27

+ 0.3%

S , = 0.013P"

21

-

+0.63

b , = -3.07

+ 0.37e

S,

(Pacific bonito)

Auris rochei

Span (cm)

25

(Pacific mackerel) Sarda chiliensis

Regressions

0.022P.8B

1

S , = 0.017t'.ffl

(bullet mackerel)

Euthynnus afinis (kawakawa)

Euthynnus pelamis (skipjack tuna)

Thunnus albacares (yellowfin tuna)

Thunnus obesus

=

0.011P~08

S, = O.OOW*~M

=

0.012Pm

(bigeye tuna) Regressions shown in Fig. 14. New data on same fish used for morphometrics in Magnuson

(1973).

Interestingly, AR increases with length more rapidly for the two euthynnids than for the two thunnids. Since the sustained speed also increases more rapidly with length for the euthynnids than the thunnids (Fig. 8B,C), the greater AR for larger euthynnids again seems a reasonable adaptation. A curved leading edge of this tapered hydrofoil reduces leadingedge suction without reducing the total thrust produced, based on a mathematical analysis by Chopra and Kambe (1977). They point out that there is an advantage because if leading-edge suction were too high, boundary layer separation could occur and significantly reduce thrust. The mean sweepback angle (Fig. 13D) of the caudal fin ranges from about 25"-50" in adults of seven species (Fig. 14D). The hydromechanical efficiency of a high thrust device such as the caudal fin declines slightly with increased sweepback (Chopra and Kambe, 1977, Fig. 5b). This feature appears to have influenced the evolution of caudal shape. For example, the sweepback was less for larger fish (Table X) and for species with faster sustained speeds (Fig. 8) than for

4 294

Y

JOHN J. MAGNUSON I B '

'

'

"

'

'

c Q 0

F

'

'

'

'

'

'

'

1

20

FORK

-

'

'

~

~

'

"

"

"

'

'

'

1"-

10

0

-

:' -

0

'

0 '

=-

LENGTH Icml

-d

5,-

/--O A.

20

10

60

111

100

lZ0

*o

0

20

.o

80

80

100

120

,lo

smaller fish or those with slower sustained speeds. The caudal fins with greater sweepback should be able to operate at greater angles of attack (Fig. 13E), a feature that should be helpful at the slower sustained speeds (Fig. 1A) that characterizes most of the species or sizes with greater sweepback in Fig. 14D. Species with greater sweepback on the caudal fin also have lower aspect ratios (Fig. 14C, D). This observation is consistent with analyses by Hopkins (1951) that a lower aspect ratio is required for longitudinal stability of a wing with greater sweepback. Other comments on sweepback are that tails with greater sweepback have the center of thrust farther back on the fin, and that greater sweepback contributes to vertical stability (von Kirmin, 1954) of the laterally moving caudal fin, i.e., it should help prevent the fin from twisting on the peduncle. Areas and spans of caudal fins vary among species (Fig. 14A, B) in ways probably important to each species. The two thunnids have the greatest spans and areas. For example, at 80 cm lengths the caudal span of a yellowfin tuna is 1.2 times that of a skipjack tuna. Its caudal area is 1.7 times that of a skipjack tuna. Since thrust is proportional to area, those with larger area should be able to produce more thrust if all

4.

LOCOMOTION BY SCOMBFUD FISHES

295

other things were equal such as angle of attack and speed of caudal movement. Apparently, they do produce more thrust because yellowfin tuna travel farther per tailbeat than do skipjack tuna (see Fig. 16). Alternatively, species with a greater caudal area should be able to produce the same thrust at lower tailbeat frequencies or lower amplitudes of caudal movement. The greater spans should help reduce drag induced at the tips of the caudal fin.

B. Movements and Postures of the Caudal Fin

The caudal fin beats back and forth as the fish moves forward and traces an oscillating path of movement (Fig. 13C) (Gray, 1968; Fierstine and Walters, 1968; Magnuson, 1970; Webb, 1975). Since it is at the end of the stiff peduncle formed from fused vertebrae (Fierstine and Walters, 1968; Nursall, 1956) it has accentuated lateral movement as if the fin were at the end of a pendulum. Bending of the fish body is concentrated at the prepeduncular joint just anterior to the peduncle (Fig. 13A).I n a comparative study by Aleev (1963) the scombrids were the most specialized in regard to concentrating the body bending in an area between 0.6 and 0.8 8. The wavelike path of the caudal fin (Fig. 13C) can be characterized by a peak to peak amplitude (A), a wavelength (A), a frequency 0, and a period. The fin has the same forward velocity as the fish (V), a lateral velocity (W) at right angles to the direction of swimming, and a speed along its oscillating path (Fig. 13C). It has an angle of attack (a)in respect to its path and a feathering angle ( 6 ) (Fig. 13E). Feathering is the angle between the path of the fin and the direction of movement of the fish. In addition, camber and a dihedral angle develop with the lateral movement of the fin. Most fishes appear to have amplitudes approximately equal to 0.21 8 that do not vary with length (Hunter and Zweifel, 1971) nor with frequencies above five tailbeats/sec (Bainbridge, 1958). Amplitude has been measured on three scombrids but does not follow the above generalities. For Pacific mackerel 30 cm long A = 0.18 8 (Hunter and Zweifel, 1971, Fig. 6). For kawakawa amplitude more than doubled from 0.16 t' to 0.34 e as frequency approximately doubled from eight to fourteen beats/sec (Fig. 15). For Atlantic bonito 16 cm long ampliiude was highest at slowest speeds (0.30e at 2 8/sec), reached a minimum of 0.22 at 5-6 tlsec, and appeared to increase slowly to about 0.24 e at 8 81sec (Pyatetskiy, 1970b, Fig. 6). Thus, the amplitude is varied by scombrids to obtain appropriate swimming performance and is not as constant as previously implied. Species such as wahoo with a long narrow body may have even greater amplitude than kawakawa. This

296

JOHN J. MAGNUSON SWl M M ING SPEED

E X T R E M E AXIS

AMPLITUDE

FREQUENCY

EXTREME

Fig. 15. Transverse caudal velocity during a tailbeat at five swimming speeds. Amplitudes and frequency of tailbeat are also given. (Adapted from Fierstine and Walters, 1968.)

could compensate, in the wahoo, for a relatively low AR for the caudal fin and a body shape that does not minimize drag. Wahoo (Fig. l B , Table 11) has one of the fastest burst speeds measured. The hydromechanical efficiency (7) of the caudal fin is expected to be fairly high over a wide range of tailbeat amplitudes (Chopra, 1976) or, in other words, the proportion of total thrust lost in wake vortices does not vary much with amplitude (A). This is apparent in Chopra’s analysis of the effects of reduced frequency. Reduced frequency [see Eq. (24)] is defined as a ratio of time to swim distance equal to the chord of the caudal fin and the time for one tailbeat (Lighthill, 1975; Chopra, 1975). This value, computed from data of Fierstine and Walters (1968) and used later to calculate thrust force (Table XI), is 0.7 based on my calculations, 0.6 from Webb (1975), and 0.8 from Chopra (1976) (the original values of Webb and Chopra converted to equilvalent units). Note that reduced frequency in this chapter is defined the same as in Webb (1975), Lighthill (1975), and Chopra and Kambe (1977), but is twice that as defined in Chopra (1975, 1976) owing to differences in the definition of the caudal fin’s chord (c), Transverse caudal velocity (Fig. 15) when crossing the axis of pro-

4.

297

LOCOMOTION BY SCOMBRID FISHES

gression of the fish ranged between 133 and 500 cm/sec or about 1.5 times the swimming speed of kawakawa (see calculation of W,,, in Table XI). Thrust from the caudal fin increases in proportion to transverse velocity squared. At speeds of less than 4 t/sec transverse caudal velocity was slowest at the beginning and end of a sweep and maximum at the axis (Fig. 15C). But at the faster speeds it was much greater in the last half of each sweep (Fig. 15E). This asymmetry may be important in producing extra thrust such as suggested from underwater movements b y man (Seireg et al., 1971, personal communication 1976).The increase is also closely tied to a corresponding decrease in angle of attack during a sweep. Fierstine and Walters (1968) suggested that the increase in transverse velocity and the concurrent decrease in angle of attack are functionally related to maintain constant thrust during all phases of a tailbeat cycle. Speed appears to increase linearly with increases in tailbeat frequency (Fig. 16) at rates between 0.5 and 0.8 t per tailbeat. A tailbeat is a full cycle back and forth; Fig. 13A is only 0.5 tailbeat. The two species that go farther per tailbeat were those with the greatest caudal span per unit length ( b ,= 0.31-0.35 t versus 0.24-0.29 t) and the highest aspect ratio (6.8-7.2 versus 4.3-6.2). Frequencies of 13 beatshec have been recorded (Yuen, 1966). In I

A

-

I

Pacific M a c k e r e l

0 Pacific Bonito 10

0

-

W v)

a -

30cm F L 57cm F L

0 Atlantic Bonita

16cm T L

0 Skipjack Tuna

57cm F L

0 Yellowfin Tuna

53cm F L

a W

W

n. u)

(35 -

5 5

r3 v)

-0

n

A 5

10

15

TAIL B E A T S I S E C .

Fig. 16. Relation between frequency (f) of tailbeats and swimming speed (t'/sec) for five scombrids. Sources: A Hunter and Zweifel (1971); [?1 Magnuson and Prescott (1966);0 Pyatetskiy (1970a); 0 and 0 Yuen (1966).

298

JOHN J. MAGNUSON

Table Calculation and Comparison of Theoretical Thrust Thrustb (2)

(1)

( 3)

(4)

(5)

(6)

( 7)

(8)

( 9)

Proportional feathering ( 1Y

A

e (cm) ~

U

r

(cmisec) 124 158 160 188 328

"mnr

(cyclesisec)

Proportion of e

cm

(cmisec)

Degrees

Radians

7.7 9.1 10.1 12.5 14.5

0.16 0.26 0.25 0.20 0.34

6.4 10.4 10.0 8.0 13.6

146 23 1 289 298 500

25 50 30 30 40

0.44 0.87 0.52 0.52 0.70

wm,,

~~

40 40 40 40 40

(19)

( 20)

(21)

0.4 0.6 0.3 0.3 0.5 0.42 (mean)

(22)

Theoretical body friction

40 40 40 40 40

124 152 160 188 328

5.0 x 6.1 x 6.4 x 7.5 x 1.3x

18

lo5 10' 18 108

c,

S

( 1)

(cm')

0.0052 0.0050 0.0050

720

0.0048 0.0043

720 720 720 720

Drag force (dynes) 3.0 x 4.3x 4.7 x 6.2 x 1.6 x

l@

I@ I@ I@ 10'

Proportion of total thrust force 0.31. 0.24 0.11 0.14 0.14 0.28 (mean)

" Following methods similar to Webb (1975); however, rather than using Webb's estimates, new values were used tor most parameters tyased oil a reanalysis of source d a b Key: (1) e most likely used by Fierstine and Walters (1968) from Walters (1966). (2) 40 cm x ekec from Fig. 15. (3) Froin Fig. 15. (4) From Fig. 15. (5) 40 cni x (column 4). (6) (ZA)(f)(measured W at axis/measured mean W of full cycle). Ratios were 1.48, 1.22, 1.43, 1.49, 1.27. Taken from Fig. 15 and Webb (1975). (7) Measured LI at axis of cycle from Walters and Fierstine (1968). (8) x 0.0174. (9) From Eq. (24). (10) From Eq. (28), c = 1.9 cm, or S,, h,. (11) From Fig. 17. (12) From Fig. 17. (13) Root mean square lateral .;peed = 0.707 W,,,. (14) Froin Tiihle X. (15) From Eq. (25). (16) From Eli. (26). (17) Same BS (2). (18) R,. = UP/O.OI. (19) From Eq. (13), i.e., turbulent. (20) 0.279 P ':I from Magnuson and Weininger (1978). (21) D,from Eq. (11). (22) (Column Pl)/(column 15). (23) (Column 21) +form drag on body + friction on second dorsal and anal +form drag on second dorsal and anal + friction on caudal + form drag on caudal. Flows over fins were assumed to be laminar and friction drag wiis computed with Eqa. ( 11) and (12) using morphometry data as follows:

comparison, hummingbirds average almost 30/sec (Hertel, 1966). For scombrids to attain speeds near 20 Usec, frequencies over 20/sec are probably required. Wavelength ranges among species from 0.2 to 1.9 e at 2 tailbeatshec, from 0.7 to 1.0 f/sec at 5 beatshec, and 0.7-0.8 G at 10 beats/sec (estimated from data in Fig. 16). With the exception of the Pacific mackerel, wavelength decreases at higher frequencies, i.e., the fish goes a shorter distance per tailbeat. The angle of attack (Fig. 13E) of the caudal fin measured from overhead photographs of swimming kawakawa varied from 25" to 50" at

299

4. LOCOMOTION BY SCOMBRID FISHES XI and Theoretical Drag Forces for a 40 cm Kawakawa Swimming at Five Speeds" (10)

(12)

(11)

Reduced frequency (1) 0.7 0.7 0.8 0.8 0.5 0.70 (mean)

c,

V"

(1)

(proportion)

0.8 0.6 0.9 0.9 0.8 0.80 (mean)

(23) (24) Same as 21-22 plns fomi drag on liody and friction and fonn drag on fins (wxond dorsal, anal, and caudal)

W,.,, (cmlsec)

S, (cnil)

I03 163 204 211 354

22 22 22 22 22

0.8 0.9 0.8 0.8 0.9 0.84(mean)

(25)

(14)

(13)

( 15)

(16)

Thrust force (dynes)

Total thrust power (ergskec)

9.6 x 1.8 x 4.2 x 4.5 x 1.1 x

lW

1.2 x 2.7 x 6.7 x 8.5 x 3.6 x

I05 10) I05 IoR

107 107

107 10' 1oR

(26)

Same aa 2.>24 plus drag iqduced from caudal fin Coefficient of totd drag

DKlg force (dynes)

4.5 x lW

1W 6.9 x 1W 9.2 x lW

6.4 x

2.3x 10)

Proportion of total thnist force

Drag force (dynes)

0.47 0 3 0. I6 0.20 0.21 0.28 (rilean)

6.4 x 8.2 x 1.5 x 1.8 x 3.4 x

I@ I@ 10% 10' 105

Proportion of total thrust force

From drag calculation

0.011 0.0094 0.016 0.014 0.0086 0.012 (mean)

0.67 0.46 0.36 0.40 0.31 0.44 (mezin)

Fina

c (cm)

S (cnil)

Anal and second dorsal C;ludal

1.3

15 44

1.7

From thrust calculation 0.016 0.011 0.044 0.034 0.028 0.027 (mean)

b (4

12

(column 15). (26) (Column Form drag w a s calculated iih i n Table VI. (24) Column 23)/(column 15). (25) (Column 23) + ( 1 = total dragtp,SU* where total 25)/(mlumn 15). (27) C , ,,,ti,,= total drag/hp,SU* where total drag = column (25); C. drag = colinnii (15). The number I indicates nondimensional.

the point where the fin crossed the axis of progression (Fierstine and Walters, 1968, Fig. 6). It averaged about 30" over a whole cycle but attained values as high as 100". Angle of attack changed systematically-starting large at the beginning of a sweep from one side to the other and ending small at the end of a sweep. Within limits, the coefficient of thrust increases directly with angle of attack, but greater hydromechanical efficiency results at lower angles of attack (Chopra, 1976). The angle of attack is maintained positively during lateral movements in each direction by rotation at the postpeduncular joint (Fig. 13A). Forces for the rotation of the fin are primarily from

300

JOHN J. MAGNUSON

water pressure rather than overt muscular activity (Fierstine and Walters, 1968). The fin is somewhat flexible and is stoutly attached at its base. Consequently, during lateral movement the tips lag behind the center and a dihedral angle develops (Fig. 13B).This angle remains positive in respect to the lateral movement of the tail because pressure alternates from one side of the fin to the other when the sweep changes direction. The dihedral angle contributes to stability and helps prevent the fin from twisting (rolling) on the end of the peduncle. The same forces deflect the trailing edge of the fin and camber develops during each sweep. However, as noted by ‘Fierstine and Walters (1968),the direction of camber is opposite to typical hydrofoils because it is convex on the high pressure side. Proportional feathering, an important concept put forth by Lighthill ( 1969), greatly influences thrust production. The parameter geometrically is a ratio of slopes between a and 8 in Fig. 13E (Wu, 1971). The parameter can be computed (Webb, 1975): Proportional feathering = amaxU/W,,,

(24)

where amax,angle of attack in radians, i.e., the slope of a; W,,,, maximum lateral velocity of the fin (cmisec); U , forward velocity of the fin or swimming speed (cmlsec).W,,, and a,,, can be measured from overhead photographs of swimming fishes at the point in lateral movement where the caudal fin passes through the axis. U can also be measured. The easiest way for me to think about proportional feathering is to note that when the angle a in Fig. 13E and 8 are equal they have the same slopes and proportional feathering = 1. In this case the fin is oriented parallel to the fish’s direction of forward progression. When a = 0, proportional feathering = 0 and the caudal fin is oriented parallel to its own path. Thrust is obtained from a caudal fin by leading-edge suction and from the backwardly inclined component of lateral forces (Lighthill, 1975; Chopra, 1975). When proportional feathering = 1, little lift is produced because no water is directed posteriorly since the fin is not inclined to the direction of the fish’s swimming. The thrust coefficient is also very low (Lighthill, 1969, 1975; Wu, 1971; Chopra, 1975). When proportional feathering = 0, again no water is directed to the rear and all thrust comes from leading-edge suction. In this case the fin has a high coefficient of thrust, but has a low hydromechanical efficiency. At intermediate values of proportional feathering the fin deflects water to the rear, and postures that optimize combinations of leading-edge suc-

4. LOCOMOTION BY SCOMBFUD FISHES

301

tion and hydromechanical efficiency occur with proportional feathering between 0.6 and 0.8 (Lighthill, 1975). Proportional feathering of the caudal fin of a kawakawa, based on data from Fierstine and Walters (1968), is between 0.3 and 0.6 (see calculations by me in Table XI)and from the same data between 0.3 and 0.9 by Webb (1975). C. Theory and Magnitude of Thrust Models to represent thrust were reviewed by Webb (1975) and are under development by Wu (1961, 1971), Lighthill (1969, 1970, 1975), Logvinovich ( 1970), Chopra (1975, 1976), Chopra and Kambe (1977), and James (1975). The most appropriate model for scombrids is developing from a two-dimensional theory of oscillatory airfoils. The two-dimensional model is expected to result in an overestimation of thrust and some attempts are being made to improve it by considering three dimensions (Chopra, 1975, 1976; Chopra and Kambe, 1977; James, 1975; Lighthill, 1975). In particular, hydrofoils of different aspect ratio, sweepback, and leading edge curvature are being considered, as are larger amplitude oscillations with large angles of attack. Thrust is derived by directing a jet of water backward over a large span as depicted in Fig. 9. Forward thrust can be considered a reaction to this “jet stream” (Wu, 1971). A vortex train is set up like a vortex sheet behind a stationary object, but with the rotation in the opposite direction to that behind a stationary object. The rotation is induced by the posteriorly directed flow. A discussion of a reaction model and useful diagrams of how a caudal fin moving laterally and forward at an angle of attack sheds vortices into the wake to accomplish useful work is presented in Lighthill (1969,1975) and Webb (1975)(see Chapter 3). Thrust is also obtained at the same time as mentioned earlier from leading-edge suction. The thrust (Tforcein dynes) generated, as taken from Lighthill (1970) and Webb (1975), is proportional to the area of the caudal fin (SJ, a measure of lateral velocity of the fin (WrmJ,water density (pJ, and the coefficient of thrust (CT):

The thrust power (TpOwer in ergslsec) is

Power is force exerted over a distance or work done per unit time. As discussed earlier, not all the power expended on the water produces

302

JOHN J. MAGNUSON

useful thrust. Some is wasted in vortices that either do not contribute to thrust or do not do so efficiently. Consequently, the actual thrust power obtained is equal to:

(27)

%TPOWBI.

where r), is hydromechanical efficiency (Lighthill, 1969, 1975). In the present analyses I have considered these losses ( 1 - r),) as a part of drag induced by the propulsive movement of the caudal fin. S, can be measured from dead fishes (Table X), and W and U can be estimated from overhead photographs of swimming fish (see Table XI for computation). CT and 77p can be interpolated from graphs of functional relations developed by Lighthill (1975) and Chopra (1975) (see Fig. 17). Both values depend on proportional feathering that can be calculated from Eq. (24) and reduced frequency (c)(Lighthill, 1970): OC

(+=-

U

where o = angular velocity for assummed harmonic motion or 27rf in radians, c = mean chord of the caudal fin (caudal area t caudal span) in cm. Proportional feathering and the frequency parameter are the keys to the estimatiou problem. The relation between them and C Tand qp will differ as the models are refined. For the present computation in Table XI, I used Chopra’s (1975) graphs for a rectangular hydrofoil with AR of 4. The mean coefficient of thrust from this model applied to Fierstine and Walters (1968) data on Kawakawa (Table XI) is C , = 0.8. This is lower than that obtained if Lighthill’s (1975) model for an

-

OO

10

wc -

U

2.0

; OO

1.0

2.0

w c -

U

Fig. 17. Graphs for estimating the coefficient of thrust (C,) and the hydromechanical efficiency (qp)of the caudal fin at different values of the frequency parameter (wclU)and proportional feathering (0). Based on a rectangular caudal fin with AR = 4 with the pitching axis at full chord. Modified from Chopra (1975). Lunate-tail swiinming propulsion. In “Swimming and Flying jn Nature” (T. Y. T. Wu, C. J. Brokaw, and C. Brennen, eds.), Vol. 2, pp. 635-650. Copyright 1975 by Plenum Press.

4.

LOCOMOTION BY SCOMBFUD FISHES

303

infinitely long wing were used, C T= 1.0, and it is also lower than if Chopra and Kambe’s (1977) model for a curved and tapered hydrofoil (their B2 model) with an AR of 8 were used, C T =0.9. None of the models specified matches the dimensions of the caudal fin of kawakawa, which has a shape similar to the tapered hydrofoil with curved anterior and posterior margins (model B2 in Chopra and Kambe, 1977) but an AR of about 6 (Fig. 14), intermediate between 4 and 8, for which graphed relationships (Chopra, 1975) are published for hydrofoils with a finite span. The choice of model does not influence significantly the value for the hydromechanical efficiency of the caudal fin ( 7 ) .With all three models, it averages 0.84 (Table XI). Thrust is computed in Table XI for kawakawa swimming at speeds of from 3 to 8 Msec. The method is taken largely from Webb (1975, Table 17) with a reanalysis of data on the swimming kawakawa (Fierstine and Walters, 1968), new data from the present paper, and the use of Chopra’s (1975) relations between feathering, reduced frequency, and C Tand y p(Fig. 17). Methods of calculation and sources of data are given in the footnotes to the table. Drag calculations are also made in Table XI following procedures outlined earlier in this paper. Four values of drag were made that included different components. The first estimate included only friction on the body assuming turbulent flow. This averaged 18% of the predicted thrust (see column 22). This percentage is similar to the percentage of total drag provided by friction on the body (Table VI) of a skipjack tuna swimming at sustained speeds. If friction on the body actually produced one-fourth of the total drag on these kawakawa swimming at these faster speeds, then the estimate of total thrust and total drag for these kawakawa would be very close. This would suggest a remarkable agreement between theoretical calculations of thrust and drag which of course should equal each other. However, at these faster speeds, friction should make up a larger portion of total drag [perhaps about 40% from (column 2l)/(column 25) in Table XI] than at the slower sustained speeds. A second estimate of drag to compare with thrust calculation included friction and form drag at the body plus friction and form drag on the permanently extended fins (columns 23 and 24). Flow over the fins was assumed to be laminar. These values equaled 28% of the calculated thrust and were still too low. A third estimate included the drag induced by the caudal fin and averaged 44% of the calculated thrust. The analysis implies that either drag is underestimated or thrust is overestimated. Yet, the agreement between calculations of thrust and drag was good. The calculations were easily within one order of mag-

304

JOHN J. MAGNUSON

nitude and closer to a factor of 2 to each other after other components of drag were added to body friction. The agreement was slightly better in my analysis than in Webb’s (1975) because I included sources of drag from the fins and thrust production and because I used Chopra’s ( 1975) calculations for C for a three-dimensional hydrofoil rather than Lighthill’s (1975) for a two-dimensional hydrofoil. Underestimates of drag could have resulted from underestimates of surface areas or frictional drag coefficients, overestimates of hydromechanical efficiency,or the exclusion of any induced drag from lift. In addition, the postures of the fins for the fish were not known. I assumed that the pelvic, pectoral, and first dorsal fins were appressed and that the mouth was closed, as they should have been at these speeds. Yet Fierstine and Walters (1968) noted that the dorsal fin was extended in some sequences and the fish may have been turning (Fierstine, personal communication). However, even if drag on the pectorals and gill resistance were computed and included in the drag estimates, the total thrust estimates would still exceed the total drag estimates. Induced drag from lift should be negligible at these high speeds even if the body were acting as a lifting surface. Since turbulent rather than laminar flow over the body was assumed, I doubt that frictional drag coefficients were too low. If laminar rather than frictional drag coefficients had been used for the body, the agreement between drag and thrust would have been poorer. Perhaps the most realistic assumption would be to assume laminar flow over the anterior half of the body and turbulent flow over the posterior half based on flow and pressure observations discussed in Aleyev (1977). Overestimates of thrust could result from overestimates of either hydromechanical efficiency or the lift coefficient of the caudal fin. Only five sequences of a swimming scombrid have been analyzed to provide data for the above analysis. In four of the five measurements the distances traveled per tailbeat were only about one-half of those observed for other scombrids in Fig. 16. Perhaps the reversed camber of the caudal fin reduces thrust. Perhaps the body interferes with thrust production even with narrow necking because much of the caudal fin still lies in the turbulent wake left by the body. The agreement between thrust and drag calculations suggests we may be reaching a useful level of prediction and that the mechanisms for thrust and drag production by scombrids are reasonably well represented in the models and coefficients. Even so, it is apparent that further refinements in the models and additional data on swimming scombrids are still needed. Also, independent collaboration of thrust and drag estimates are needed from data on metabolism.

4.

LOCOMOTION BY SCOMBFUD FISHES

305

VII. THE LOCOMOTORY SYSTEM Scombrids are astounding bundles of adaptations for efficient and rapid swimming. All features of hydromechanics, morphology, and behavior, while discussed above in pieces, are parts of an interactive and coherent locomotory system. Still other pieces of the system such as those related to maneuvering and stability, physiology and metabolism, and ecology and the acquisition of energy were not considered here. The entire biology of scombrids is tightly linked to swimming as a way of life. All life’s activities occur in motion. Feeding, courtship, spawning, and resting are all done on the move (Magnuson and Prescott, 1966; Magnuson, 1969). Parental care is restricted to leaving the fertilized eggs in a suitable water mass in pelagia. Sustained swimming keeps scombrids from sinking, ventilates their gills, brings them in contact with food, and moves them rapidly over the vast distances of the open ocean. Burst swimming serves primarily in escape from predators and the pursuit of prey. In feeding schools food is only available in presence of other competitors and the development of a fast feeding attack may have evolved in response to getting there first, rather than catching up with the prey per se. Scombrids can easily outdistance their prey. While all scombrids are continuous and rapid swimmers, each species is a unique evolutionary solution for life that exemplifies a welI-tuned locomotory system. Three genera (Euthynnus, Thunnus, and Acanthocybiurn) will be discussed here to point out the integration of the locomotory system. Euthynnids, represented by the skipjack tuna and kawakawa, are small to medium in size and have fast rates of sustained swimming. Thunnids, represented b y yellowfin tuna, bigeye tuna, and albacore, are large species with intermediate rates of sustained swimming. Acanthocybium, represented by the wahoo, is a large and elongate species with slow rates of sustained swimming and the highest burst speeds. The first thing to note is that most locomotory adaptations are more closely correlated to continuous or sustained swimming than to burst swimming (Table XII). For example the aspect ratio of the caudal fih and blood hemoglobin concentrations are greatest for euthynnids with the fastest sustained speeds but with the slowest burst speed. Second, the locomotory adaptations are more closely correlated to speed measured in lengths per sec (t/sec) than to speed in either Reynolds number (RL) (Table XII) or absolute units (cmhec). For example, at sustained speeds Acanthocybium with the highest RL has the lowest

306

JOHN J. MAGNUSON

Table Comparison of Swimming Speeds and Hydromechanic Parameters Lift and weight

Swimming speed ( 2)

( 1)

(3)

(4)

(5)

(6) Pectoral loading

Sustained Genus

Euthynnus Thunnus Acanthocybiurn

Nsec

2.3 0.97 0.33

R,

3.0 x 1oJ 3.7x 1P 5.1 X 1oJ

Burst elsec

12 14 15

R

,.

3.4 x 10s 7.3x 10s 1.6 X lo7

Lengths Per tailbeat

4.2 5.0

-

( 1000 Pe

(g/cc)

1.091 1.052 1.028

S, (cm')

72 88

-

dynes1 cm')

9.4 2.4 0.4

0 Key: (1) mean from Table I; (2) mean from Table 11; (3)from Fig. 16 at 5 beatslsec; (4)largest fish measured (Table 111); (5)calculated for 50 cm long fish from Magnuson (1973,Table 4) and from Dotson (1977);( 6 ) mean from Magnuson (1973,Table 3);(7)mean from Table IX; (8)calculated for 50 cm long fish from Table X; (9)calculated for 50 cm long fish from Table X; (10)from

sustained speed in elsec, as well as the lowest aspect ratio of the caudal fin and the lowest blood hemoglobin concentration. The euthynnids have sustained speeds (e/sec) that are 2.4 times those of thunnids and 7.0 times those ofAcanthocybium (Table XII). The differences are due to the hydromechanics of countering their weight in water which they do by generating lift on their extended pectoral fins during continuous swimming. Differences in sustained speeds are expected from differences in surface loadings of the pectoral fins which for euthynnids are 4.0 times those of thunnids and 24 times those ofAcanthocybium. Faster sustained speeds have evolved along with adaptations for efficient thrust production and drag reduction. For example, the average aspect ratio of the caudal fin of euthynnids 7.0 compares with 6.5 for thunnids and only 5.5 for Acanthocybium (Table XII). Likewise the body ofAcanthocybium is not as well streamlined for sustained swimming as the euthynnids and thunnids. Finally, the physiological resources for sustained swimming are greater for euthynnids than thunnids, and for thunnids than Acanthocybium (Table XII). Magnitudes of muscles and the oxygen transport system are the physiological resources considered here. Blood hemoglobin concentrations of euthynnids are 1.1 times those of thunnids and 1.7 times those ofAcanthocybium. Gill areas of euthynnids are 1.3 times those of thunnids. Both euthynnids and thunnids have a greater body mass at a given length than Acanthocybium and consequently should have a greater total muscle mass. In addition the

4.

307

LOCOMOTION BY SCOMBRID FISHES

XI1 of Euthynnus, Thunnus, and Acanthocybium" Thrust (7 )

(8)

Drag

(9)

b,

AR

S, (cm')

(cm)

7.0 6.5 5.5

30 38 -

14 15 -

(10)

(11)

Body Position thick- maximum ness thickness (% e) (% l')

22 23 12

41+ 41+ 37+

Physiological resources ( 12)

Induced from lift High Medium Low

(13)

(14)

(15)

Red

Blood hemoglobin

(kg)

muscle (%M,)

(g/100 ml)

2.4 2.5 1.0

8.4 7.6 1.8

17 15 10

MI

(16) Gill area (m')

4.0 3.0

-

Table VII; ( 11) from Table VII; (12) proportional to pectoral loading; (13) calculated for 50 cm long fish from Magnuson (1973, Table 4) and from Dotson (1977); (14) means from Magnuson ( 1973, Table 7); (15) means from Magnuson (1973, Table 7); (16) calculated for 2.5 kg fish from equations in Muir and Hughes (1969).

red muscle (% mass) used for sustained swimming (Rayner and Keenan, 1967) is 1.1 times larger in euthynnids than in thunnids and 4.7 times larger than in Acanthocybium. Thus, adaptations that dictate fast sustained speeds are closely related to observed speeds as are adaptations for efficient thrust production, drag reduction, oxygen transport, and development of muscular forces. Burst speeds do not differ among the three genera as much as the sustained speeds do, and those slight differences observed are in the reverse order. Acanthocybium had the fastest burst speed, and euthynnids had the slowest. The burst speed of Acanthocybium was 1.2 times those of euthynnids while sustained speed of euthynnids was 7 times those ofAcanthocybium. I would guess that the faster speeds are possible for a shorter time in Acanthocybium because the white muscles are smaller, as indicated by the mass-length relationship, and because the oxygen transport system is less developed, as indicated by low concentrations of blood hemoglobin. The faster burst speeds may be related to greater caudal areas or speed of lateral movement by the caudal fin, but data are lacking. The thunnids had faster burst speeds and swam farther per tailbeat than euthynnids (Table XII). This could b e predicted in part by the large caudal area of thunnids and perhaps by a greater amplitude of tail movements. However, comparative data on amplitudes and tailbeat frequencies are not available. It appears that variation in speed of sustained swimming has more

308

JOHN J. MAGNUSON

closely influenced the evolution of swimming adaptations among scombrids than have considerations of burst speeds. All have fast burst speeds and the adaptations for continuous swimming apparently serve well in fast swimming. With the magnitude of forces involved in continuous swimming it is not surprising that adaptations for producing large amounts of thrust and lift as well as reducing drag i r e finely tuned. What is more surprising is the necessity of continuous locomotion in the first place. The loss of or reduction of the gas bladder is the adaptation that results in continuous swimming and its attendent drag for a pelagic teleost. What selective pressures have molded evolution along this energy demanding path? The most likely idea is that it increases vertical mobility especially near the sea surface (Magnuson, 1966, 1973; Lighthill, 1969,1975). This would be important in food capture and in escape from predators. Species are not slowed by the requirement to absorb expanding gases from the gas bladder during a rapid ascent. A new hypothesis, considered above, is that the negative buoyancy makes possible gliding and soaring by scombrids in waves near the surface or in the thermocline. This would make horizontal movements in the sea effortless under certain conditions. Data to support these ideas are skimpy or absent entirely. Regardless of evolutionary history, the locomotory system that has evolved in these negatively buoyant, pelagic teleosts will continue to challenge the biologists and engineers. ACKNOWLEDGMENTS

I thank my wife, Norma, for her help with.this study. I also thank T. Y. T. Wu for occasional help in translating mathematics to English for me; R. McN. Alexander, C. D. Sharp, and P. W. Webb for their comments on an early draft; and J. L. Brooks for his encouragement. New information presented on the morphometry of scombrids was extracted from photographic data from the files of the Honolulu Laboratory of the National Marine Fisheries Service. REFERENCES Aleev, Y. C . (1963).“Funktional ’nye Osnovy Vneshnego Strecniya Ryby (Functional and Cross Morphology in Fish),” 245 pp. Izd. Akad. Nauk SSSR, Moscow. (Transl. by Isr. Program Sci. Transl., T T 67-51391,268 pp. Natl. Tech. Inf. Sew., Springfield, Virginia, 1969.) Alexander, R. McN. (1965).The lift produced by the heterocercal tails of Se1achii.J. E r p . B i d . 43, 131-138.

4.

LOCOMOTION BY SCOMBFUD FISHES

309

Alexander, R. McN. (1967). “Functional Design i n Fishes.” Hutchinson, London. Alexander, R. McN. (1968).“Animal Mechanics.” Univ. of Washington Press, Seattle. Aleyev, Y. G. (1977).“Nekton.” Dr. W. Junk b.v. Publishers, The Hague. Aleyev, Y. G., and Ovcharov, 0. P. (1973).The three-dimensional pattern of flow round a moving fish.J. Zchthyol. (USSR) 13, 933-936. Anonymous (1975).A study of the applications of remote sensing techniques for detection and enumeration of giant bluefin tuna. Southeast Fish. Cent. Contrib. No. 437 (MARMAP No. 108). Bainbridge, R. (1958). The speed of swimming of fish as related to size and to the frequency and amplitude of tail beat.J. E x p . Biol. 35, 109-133. Bainbridge, R. (1961).Problems of fish locomotion. Zool. Soc. London 5, 13-32 (Vertebrate Locomotion Symposium). Baldridge, H. D., Jr. (1970). Sinking factors and average densities of Florida sharks as functions of liver buoyancy. Copeia pp. 744-754. Baldridge, H. D., Jr. (1972). Accumulation and function of liver oil in Florida sharks. Copeia pp. 306-325. Belyayev, V. V., and Zuyev, G. V. (1969).Hydrodynamic hypothesis of school formation in fishes.J. Zchthyol. (USSR)9, 578-584. Blaxter, J. H. S., and Dickson, W. (1959).Observations on the swimming speeds of fish.J. Cons., Cons. Perm. Znt. Explor. Mer 24, 472-479. Bone, Q. (1975).Muscular And energetic aspects of fish swimming. In “Swimming and Flying in Nature” (T. Y. T. Wu, C. J. Brokaw, and C. Brennin, eds.), Vol. 2, pp. 493-528. Plenum, New York. Bone, Q., andRoberts, B. L. (1969).The density of e1asmobranchs.J. Mar. Biol. Ass. U.K. 4% 913-937. Breder, C. M., Jr. (1926). The locomotion of fishes. Zoologica (N.Y.) 4, 159-256. Breder, C. M., Jr. (1965). Vortices and fish schools. Zoologica (N.Y.)50,97-114. Breder, C. .M., Jr. (1967). On the survival value of fish schools. Zoologica (N.Y.) 52, 25-40. Breder, C. M., Jr. (1976). Fish schools as operational structures. Fish. Bull. 74,471-502. Brown, C. E., and Muir, B. S. (1970). Analysis of ram ventilation of fish gills with application to skipjack tuna (Katsuwonus pelamis). J . Fish. Res. Board Can. 27, 1637-1652. Cahn, P. H. (1972).Sensory factors in the side-to-side spacing and positional orientation of the tuna, Euthynnus aflinis, during schooling. U S . Fish Wiidl. Serv., Fish. Bull. 70, 197-204. Chopra, M. G. (1975).Lunate-tail swimming propulsion. In “Swimming and Flying in Nature” (T. Y. T. Wu, C:J. Brokaw, and C. Brennen, eds.), Vol. 2, pp. 635-650. Plenum, New York. Chopra, M. G . (1976). Large amplitude lunate-tail theory of fish locomotion. J . Fluid Mech. 74, 161-182. Chopra, M. G., and Kambe, T. (1977).Hydromechanics of lunate-tail swimming propulsion. Part 11. J. Fluid Mech. 79, 49-69. Commercial Fisheries Review (1969).Underwater tuna school tracked by sonar. Commer. Fish. Reu. 31(11), 9-10. Cone, C. D., Jr. (1962). Thermal soaring of birds. A m . Sci. 50, 180-209. Dizon, A. E., Neill, W. H., and Magnuson, J, J. (1977).Rapid temperature compensation of volitional swimming speeds and lethal temperatures in tropical tunas (Scombridae). Enu. Biol. Fish. 2, 83-92.

310

JOHN J. MAGNUSON

Dotson, R. C. (1977).Minimum swimming speed of albacore (Thunnus alalunga). Natl. Mar. Fish. Serv., Southwest Fish. Cent., La Jolla, California. Fierstine, H. L., and Walters, V. (1968). Studies in locomotion and anatomy of scombroid fishes. Mem. South. Calilf.Acad. Sci. 6, 31 pp. Findley, J. S., Studier, E. H., and Wilson, D. E. (1972).Morphological properties ofbat wing. J. Mammal. 53, 429-444. Gooding, R. M., and Neill, W. H. (1978). Routine respiration rate and resistance to low oxygen levels in skipjack tuna, Katsuwonus pelamis. Natl. Mar. Fisheries Serv., Biol. Lab., Honolulu. Gray, J. (1968). “Animal Locomotion.” Clowes, London. Greenwalt, C. H. (1975). The flight of birds. Trans. Am. Philos. SOC.65(4), 1-67. Hall, F. G. (1930).The ability of the common mackerel and certain other marine fishes to remove dissolved oxygen from sea water. Am. J. Physiol. 93, 417-421. Harris, J. E. (1936). The role of the fins in the equilibrium of the swimming fish. I. Wind-tunnel tests on a model ofMustelus canis (Mitchell).J. E x p . Biol. 15,32-47. Hartman, F. A. (1961). Some locomotor mechanisms of birds. Smithson. Misc. Collect. 143(1), 1-91. Hertel, H. (1%6). “Structure-Form-Movement,” 1st Engl. Ed. Reinhold, New York. Hoerner, S. F. (1965). “Fluid Dynamic Drag.” Hoerner, Midland Park, New Jersey. Hopkins, E. J. (1951).Lift, pitching moment, and span load characteristics of wings at low speed as affected by variations of sweep and aspect ratio. Natl. Adu. Comm. Aeronaut., Tech. Notes No. 2284. Hunter, J. H., and Zweifel, J. R. (1971). Swimming speed, tail beat frequency, tail beat amplitude, and size in jack mackerel, Trachurus symmetricus, and other fishes. U . S . Fish Wildl. Sera, Fish. Bull. 69, 253-266. James, E. C. (1975).Lifting-line theory for an unsteady wing as a singular perturbation prob1em.J. Fluid Mech. 70, 753-771. Jones, F. R. H. (1957).The swimbladder. In “The Physiology of Fishes” (M. E. Brown, ed.), Vol. 2, pp. 305-322. Academic Press, New York. Gamer, E. (1960). Zur Form und Funktion des Locomotionapparates der Fische. Z. Wiss. Zool. 163, 1-36. Laurs, R. M., Yuen, H. S. H., and Johnson, J. H. (1977). Small-scale movements of albacore, Thunnus alalunga, in relation to ocean features as indicated by ultrasonic tracking and oceanographic sampling. Fish. Bull. 75,347-355. Lighthill, M. J. (1969).Hydromechanics of aquatic animal propulsion. Annu. Reo. Fluid Mech. 1,413-466. Lighthill, M. J. ( 1970).Aquatic animal propulsion of high hydromechanical ei3ciency.J. Fluid Mech. 44, 265-301. Lighthill, M . J. (1974). Aerodynamic aspects of animal flight. BHRA Fluid Eng. pp. 1-30. Lighthill, M. J. (1975). “Mathematical biofluiddynamics.” Soc. Ind. and Appl. Math., Philadelphia, Pennsylvania. Lissaman, P. B. S., and Shollenberger, C. A. (1970). Formation flight of birds. Science 168, 103-105. Logvinovich, G. V. (1970).Hydrodynamics of a thin, flexible body (evaluation of hydrodynamics of fish). In “Hydrodynamic Problems of Bionics,” Bionika, Vol. 4. Kiev. (Transl. from Russ.), JPRS 52605, pp. 4-11. Natl. Tech. Inf. Serv., Springfield, Virginia, 1971.) Lowndes, A. G. (1955). Density of fishes. Some notes on the swimming of fish to be correlated with density, sinking factor and load carried. Ann. Mag., Nat. Hist. 12(8), 241-256.

4. LOCOMOTION BY SCOMBRID FISHES

311

MacKay, K. T. (1976). Population biology and aspects of energy use of the northern population of Atlantic mackerel, Scomber scombrus L. Ph.D. thesis, Dalhousie University, Halifax, Nova Scotia, Canada. Magnan, A. ( 1929). Les caracteristiques geometriques et physiques des poissons. Ann. Sci. Nut., Zool. 10(12), 5-133. Magnuson, J. J. (1963).Tuna behavior and physiology, a review. FA0 (FA0 U N ) Fish. Rep. 3(6), 1057-1066. Magnuson, J . J. (1966).Continuous locomotion in scombrid fishes. Am. Zool. 6(4), 5. Magnuson, J. J . (1969). Swimming activity of the scombroid fish Euthynnus afinis as related to search for food. FA0 (FA0 U N ) Fish. Rep. 62(2), 439-451. Magnuson, J. J. (1970).Hydrostatic equilibrium of Euthynnus afinis, a pelagic teleost without a gas bladder. Copeia pp. 5G85. Magnuson, J . J. (1973).Comparative study of adaptations for continuous swimming and hydrostatic equilibrium of scombroid and xiphoid fishes. U S . Fish Wildl. Seru., Fish. Bull. 71, 337-356. Magnuson, J. J., and Prescott, J . H. (1966).Courtship, locomotion, feeding, and miscellaneous behaviour of Pacific bonito (Sarda chiliensis).Anim. Behau. 14,54-67. Magnuson, J. J . , and Weininger, D. (1978). Estimation of minimum sustained speeds and associated body drag on scombrids. In “The Physiological Ecology of Tunas” (G. D. Sharp and A. E. Dizon, eds.). Academic Press, New York. Marshall, N. B. (1965).“The Life of Fishes.” Weidenfeld & Nicolson, London. Muir, B. S. (1969).Gill dimensions as a function of fish si2e.J. Fish. Res. Board Can. 26, 165-170. Muir, B. S., and Brown, C. E. (1971). Effects of blood pathway on the blood-pressure drop in fish gills, with special reference to tunas.J. Fish. Res. Board Can. 28, 947-955. Muir, B. S., and Hughes, G. M. (1969).Gill dimensions for three species of tunny.J. E x p . Biol. 51, 271-285. Muir, B. S., and Kendall, J. I. (1968).Structural modifications in gills of tunas and some other oceanic fishes. Copeia pp. 388-398. Muir, B. S., and Newcombe, C. P. (1974). Laboratory observations on filter feeding in Atlantic mackerel, Scomber scombrus. Marine Ecology Laboratory, Dartmouth, Nova Scotia (cited in MacKay, 1976). Munk, M. M. (1924). The aerodynamic forces on airship hulls. Natl. Ado. Comm. Aeronaut., Rep. No. 184. Neill, W. H., Chang, R. K., and Dizon, A. E. (1976).Magnitude and ecological implications of thermal inertia in skipjack tuna, Katsutoonus pelamis (Linnaeus).Eno. Biol. Fish. 1, 61-80. Nursall, J. R. (1956). The lateral musculature and the swimming of fish. Proc. Zool. Soc. London ( B ) 126, 127-143. Nursall, J. R. (1958).The caudal fin as a hydrofoil. Evolution 12, 116-120. Perkins, C. D., and Hage, R. E. (1949).“Airplane Performance Stability and Control.” Wiley, New York. Prandtl, L., and Tietjens, 0. G. (1934).“Applied Hydro- and Aeromechanics.” Dover, New York. (New Ed., 1957.) Pyatetskiy, V. Y. (19704. Kinematic swimming characteristics of some fast marine fish. In “Hydrodynamic Problems of Bionics,” Bionika, Vol. 4. Kiev. (Transl. from Russ., TPRS 52605, pp. 12-23. Natl. Tech. Inf. Serv., Springfield, Virginia, 1971.) Pyatetskiy, V. Y. (1970b). Hydrodynamic swimming characteristics of some fast marine fish. In “Hydrodynamic Problems of Bionics,” Bionika, Vol. 4. Kiev. (Transl. from Russ., JPRS 52605, pp. 24-31. Natl. Tech. Inf. Serv., Springfield, Virginia, 1971.)

312

JOHN J. MAGNUSON

Rayner, M. D., and Keenan, M. J. (1967). Role of red and white muscles in the swimming of skipjack tuna. Nature (London) 214, 392-393. Richardson, E. G. (1936).The physical aspects offish locomotion.]. Exp. Biol. 13,63-74. Roberts, J. L. (1975). Active branchial and ram gill ventilation in fishes. Biol. Bull. (Woods Hole, Mass.) 148, 85-105. Rosen, M. W. (1961). Experiments with swimming fish and dolphins. Am. Soc. Mech. Eng. Publ. No. NWA 203, pp. 1-11. Rosen, M. W., and Cornford, N. E. (1971).Fluid friction of fish slimes.Natur-e (London) 234, 49-51. Seireg, A., Baz, A., and Patel, D. (1971). Supportive forces on the human body during underwater activities. J . Biomech. 4, 23-30. Sharp, G. D., and Francis, R. C. (1976).An energetics model for the exploited yellowfin tuna, Thunnus albacares, population in the Eastern Pacific Ocean. U.S. Fish Wildl. Seru., Fish. Bull. 74, 3 6 5 0 . Shebalov, A. M. (1969). Some questions of the influence of non-stationarity on the “mechanism” of resistance formation. Bionika 3, 61-66 (cited in Aleyev, 1977). Stevens, E. D. (1972). Some aspects of gas exchange in tuna. J . E x p . Biol. 56,809-823. Stevens, E. D., and Fry, F. E. J . (1972).The effect of changes in ambient temperature on spontaneous activity in skipjack tuna. Comp. Biochem. Physiol. A 42,803-805. Storer, J. H. (1948). The flight of birds analyzed through slow motion photography. Cranbrook Inst. Sci. Bull. 28, 1-94. Strasburg, D. W. (1961). Diving behavior of Hawaiian skipjack tuna. J. Cons., Cons. Perm. Int. Explor. Mer 26, 223-229. Sverdrup, H. U., Johnson, M. W., and Fleming, R. H. (1942).“The Oceans, their Physics, Chemistry, and General Biology.” Prentice-Hall, New York. van Olst, J. C., and Hunter, J. R. (1970).Some aspects of the organization of fish schools. J. Fish. Res. Board Can. 27, 1225-1238. von Khrmhn, T. (1954).“Aerodynamics.” McGraw-Hill, New York. von Mises, R. (1945).“Theory of Flight.” Dover, New York. (New Ed., 1959.) Walters, V. (1962).Body form and swimming performance in the scombroid fishes. Am. 2001.2, 143-149. Walters, V. (1966).On the dynamics of filter-feeding by the wavyback skipjack (Euthynnus afinis). Bull. Mar. Sci. 16, 209-221. Walters, V., and Fierstine, H. L. (1964).Measurements of swimming speeds of yellowfin tuna and wahoo. Nature (London) 202,203-209. Watanabe, N. (1942). A determination of the bodily density, body temperature, and swimming speed of the skipjack (Katsuwonus pelamis). Nippon Suisan Gakkaishi 11, 146148. Webb, P. W. (1975). Hydrodynamics and energetics of fish propulsion. Bull. No. 190. Dep. Environ., Fish., Mar. Serv., Ottawa. Weihs, D. (1973a). Hydromechanics of fish schooling. Nature (London) 241,290-291. Weihs, D. (1973b). Mechanically efficient swimming techniques for fish with negative buoyancy. J. Mar. Res. 31, 194-209. Weihs, D. ( 1 9 7 3 ~ )Optimal . fish cruising speed. Nature (London) 245,48-50. Weihs, D. (1974). Energetic advantages of burst swimming of fish. J. Theor. Biol. 48, 215-229. Weihs, D. (1975).Some hydromechanical aspects of fish schooling. In “Swimming and Flying in Nature” (T. Y. T. Wu, C. J. Brokaw, and C. Brennin, eds.), Vol. 2, pp. 703-718. Plenum, New York. Wu, T. Y. (1961).Swimming of a waving plate. J . Fluid Mech. 10, 321-344.

4.

LOCOMOTION BY SCOMBFUD FISHES

313

Wu, T. Y. (1971). Hydromechanics of swimming of fishes and cetaceans. Adu. Appl. Mech. 11, 1-63. Yuen, H. S. H. (1962). Schooling behavior within aggregations composed of yellowfin and skipjack tuna. FA0 (FA0 U N ) Fish. Rep. 6(3), 1419-1429. Yuen, H. S. H. (1966). Swimming speeds of yellowfin and skipjack tuna. Trans. Am. Fish. SOC. 95, 203-209. Yuen, H. S. H. (1970). Behavior of skipjack tuna, Katsuwonus pelamis, as determined by tracking with ultrasonic devices.]. Fish. Res. Board Can. 27, 2071-2079. Zuyev, G . V., and Belyayev, V. V. (1970). An experimental study of the swimming of fish in groups as exemplified by the horsemackerel (Trachurus mediterraneus ponticus Aleev). J . Zchthyol. (USSR) 10, 545-549.

This Page Intentionally Left Blank

5 BODY TEMPERATURE RELATIONS OF TUNAS. ESPECIALLY SKIPJACK E . DON STEVENS and WILLIAM H . NEILL

I . Introduction .................................................. I1. History ...................................................... I11 . What Is the Real (i.e., Typical) Excess Body Temperature

ofTunas? .................................................... A . The Problem and a Definition of Excess Temperature ....... B . Effect of Captivity on Excess Temperature .................. C . Effect of Activity on Excess Temperature ................... IV. What Is the Heat Source Responsible for Large Excess Body Temperatures in Tunas? ................................. V. Anatomical Basis of Warm-Bodiedness in Tunas ................. A . Introduction .............................................. B. Role of Vascular Anatomy in Tuna Systematics .............. C . Systematic Distribution of Retia in Tunas ................... VI . Exchange of Heat between Tunas and Their Environment ....... A . The Rate of Change of Core Temperature ................... B. Relative Heat Exchange via Gills and Body Surface ......... C . Estimates of Effectiveness for Heat Exchange ............... VII . Do Tunas Regulate Their Body Temperature? .................. A . Regulation of Heat Gain ................................... B. Regulation of Heat Loss ................................... C . Long-Term Temperature Responses or Thermoregulation? . . . VIII . Adaptive Values of Warm-Bodiedness and Large Thermal Inertia .............................................. A . Increased Muscle Power at Elevated Temperatures ......... B. Decreased Time between Bouts of Burst Activity at Elevated Temperatures .................................... C . Benefits Accruing from Body Temperature Stability ......... D . Perception of Weak Horizontal Temperature Gradients ...... E . Retia Function to Retard Mass Transfer Rather than Heat Transfer ............................................. IX. Physiological Insights into the Natural History of Tunas ......... X . Conclusion: A Thermocentric Overview of Tuna Evolution . . . . . . References .........................................................

316 317 319 319 319 320 321 325 325 326 328 334 334 338 338 340 341 344 346 348 349 350 350 352 353 353 354 356

315 FISH PHYSIOLOGY. VOL . VII Copyright @ I 1978 by Academic Press. Inc . All rights of reproduction in any form reserved. ISBN 012-350407-4

316

E. DON STEVENS AND WILLIAM H. NEILL

I. INTRODUCTION There is only one existing review of tuna autecology (Magnuson,

1963); it is concerned mostly with behavior and a list of what, in 1963, was unknown about tuna physiology. In fact, papers on tuna physiology (that is, those based on experimental work with live fish) probably still number less than thirty. One might then ask, “Why devote a chapter exclusively to tunas when there are 1000 times more published pages on salmonids?” For those of us who have -had the privilege of working with these animals, the answer comes quicklytunas are unique in many interesting ways. They are the fastest swimmers, among the largest of fishes, have the warmest bodies, have the highest metabolic rates, and are exceptional in many biochemical respects (unique hemoglobins; lactate, glycogen, actomyosin concentrations highest). Indeed, tunas are such extraordinary fishes that the tuna researcher frequently finds himself dividing all fishes simply into tunas and “nontunas”-a practice into which we slip in the pages that follow. The same attributes that so perfectly attune tunas to their continuously active existence in the pelagic zone of the high seas unfortunately make for serious logistical difficulties in carrying out research on captive subjects. A tank, even of oceanarium size, is a confining and otherwise alien environment for tunas. Some of the problems inherent in working with tunas recently have been overcome at the Kewalo research facility of the National Marine Fisheries Service in Honolulu, the only site in the world where tunas are routinely kept in captivity. Thus, many of the data and ideas offered in this chapter had their origins in work at the Honolulu laboratory; however, we have also drawn freely on other work, especially that of Frank Carey and his associates at Woods Hole Oceanographic Institution. The focal point of this paper is the issue of tunas’ body temperatures and their relation to activity and metabolism. Central to the construct “tuna” is warm-bodiedness: Tunas’ core tissues, especially the propulsive musculature, may be 20°C warmer than the surrounding seawater. Although the fact that tunas are warm has been known since at least 1835, the “how” and “why” of warm-bodiedness are only now yielding to experimental attack. Thus, much of the following information is a necessarily sketchy and tentative preview of work yet to appear in print. To the authors and sponsors of that work, we are sincerely grateful.

5.

BODY TEMPERATURE RELATIONS OF TUNAS

317

11. HISTORY Eschricht and Muller (1835)provided an early account of the unusual ,circulatory anatomy of tunas and excellent descriptions of some of these fishes’ vascular devices that have come to be regarded as countercurrent heat exchangers. At about the same time Davy (1835), a British physician who voyaged in the tropics, reported that tunas had warmer bodies than nontunas caught from the same waters. H e observed that skipjack tuna (Katsuwonus pelamis) were as much as 10°C warmer than the water from which they were caught. That is, tunas are significantly warm-bodied. The next major report was a beautiful monograph on the anatomy of the scombroids by Kishinouye (1923). In it he described the location and extent of the vascular countercurrent heat exchangers and drew attention to their possible relationship with activity and warmbodiedness: “The higher temperature of the body than surrounding water, and consequently great activity of fishes of the Plecostei is undoubtedly due to the peculiar circulatory systems” (p. 367). Subsequent temperature measurements, especially by Japanese and Russian workers, failed to confirm that tunas in fact d o have high body temperatures. Most of these studies were based on a few observations. However, an extensive report of tuna body temperatures over a wide range of sea temperatures showed a maximum excess muscle temperature of only 3°C in skipjack tuna (Uda, 1941).That is, tunas are not especially warm-bodied. These workers were hampered by the usual problems of working with large, active fish at sea and by the additional difficulty of making their measurements with mercury thermometers. The first observations made with an electronic thermometer on muscle temperatures of freshly caught tunas are those reported by Barrett and Hester (1964) for skipjack and yellowfin (Thunnus albacares). In 20°C water, skipjack muscle was on the average 8°C warmer than the water, and in some fish it was as much as 11°C warmer (Fig. 1B). The excess was smaller in fish taken from warmer waters and was smaller in smaller fish. Shortly thereafter, Carey and Teal (1969b) reported large excesses achieved by giant bluefin tuna (T. thynnus) swimming in cold water (Fig. 1A). That is, tunas are very warmbodied. The observation that excess was much larger in bluefin caught from cold water than in those from warm water was taken to indicate that these fish possess the capacity to regulate their body temperature.

E. DON STEVENS AND WILLIAM H. NEILL I

I

1

I

BLUEFIN

'

. _

:.

c I

.

, /

/

,

/

// / / /

/

I 10

/

I I5

/

/

/

I I5

I

/

20

WATER TEMPERATURE ("C)

I

I .

I

1

I

SKIPJACK : ! .

.* . . /

.. . .

I

9

/

/

-

/

/

..

/

/

*

/

/

.

/

/

/

/

/

/ / /

/

/

/

/ /

/

/

/

/

I I5

/ I 20

I

I

25

30

WATER TEMPERATURE ("C)

Fig. 1. Maximum red muscle temperatures of bluefin tuna (Carey, 1976 personal communication) and skipjack tuna ( x from Barrett and Hester, 1964; 0 from Stevens and Fry, 1971).

5.

BODY TEMPERATURE RELATIONS OF TUNAS

319

Subsequent measurements of temperature in freshly caught tunas (Stevens and Fry, 1971; Graham, 1973, 1975) have confirmed that tunas are indeed several degrees warmer than the water from which they are caught. However, reflection on these data has suggested a number of interpretative problems, not the least of which is the effect of precapture activity on the magnitude of the core temperature excess.

111. WHAT IS THE REAL (i.e., TYPICAL) EXCESS BODY TEMPERATURE OF TUNAS? A. The Problem and a Definition of Excess Temperature Tunas undoubtedly have higher excess muscle temperatures than nontunas of the same weight caught from water of the same temperature. However, the typical magnitude of the excess is still in question; the problem is that many of the large values mentioned above were obtained from fish that may have been hyperactive just prior to their capture. Throughout the discussion that follows, T , refers to ambient water temperature, Tb to body temperature, and T , to excess body temperature, or Tb - T,. In steady-state conditions with respectto T,, T , is the result of a combination of factors that affect the balance between metabolic heat production and heat loss to the environment.

B. Effect of Captivity on Excess Temperature Stevens and Fry (1971) observed that T , in skipjack tuna decreased from about 9.1"C at capture to 2"-4"C for the same group of fish during captivity (Fig. 2). In a long series of experiments, Neil1 et al. (1976) determined excess tissue temperatures of captive skipjack tuna and related them to a number of variables. In these experiments, the fish were lightly sedated and restrained in a water-filled trough. Excess tissue temperatures were described by the following set of regression equations: red muscle: white muscle: brain:

log T , = 0.1608 + 0.5800 log W - 0.000% log T , = 0.1110 0.6067 log W - 0.000% log( T , + 1) = -0.1724 + 0.0105 T ,

+

where W, wet weight (kg); T,, water temperature ("C); t, time fish had

320

E. DON STEVENS AND WILLIAM H. NEILL

SKIPIACK

I

I

I

,

I

I

6

I

I

Fig. 2. The decrease in excess red muscle temperature in skipjack tuna following capture. (Adapted from Stevens and Fry, 1971.)

been restrained (min). Thus, excess temperature relations of red and white muscle were similar to each other but different from that of brain. Muscle T, depended on fish weight and restraint time, whereas brain T , depended only on ambient temperature. The dose of sedative was unrelated to T , in all three tissues. Point solutions for red muscle T , at t = 0 min are 0.85 and 2.99"C for fish weighing 0.4 and 3.5 kg, respectively. It is clear that excess temperature is lower in captive tuna. The next question is: "Is it lower because of lower activity levels in captivity?"

C. Effect of Activity on Excess Temperature Stevens and Fry (1971)exercised captive skipjack tuna strenuously for 30 min and effected a small increase in red muscle mean excess temperature-from 3.3"Cbefore chasing to 4.6"C after chasing. Neill et al. (1976) continuously monitored red muscle T , in free-swimming skipjack. Chasing the larger of two fish to exhaustion caused T , to increase from 2.3" before to 5.7"C after the exercise. Thus, neither Stevens and Fry (1971) nor Neill et al. (1976) were able to generate excesses comparable to those observed in ocean-caught fish, even through violent chasing to exhaustion. However, Dizon ( 1976personal communication) has recorded excess core temperatures greater than 8.5"Cin Euthynnus afinis that were in captivity for more than a year. He telemetered temperature data from fish in a large oceanarium. Muscle temperature increased rapidly (8.5"Cwithin a few minutes) when the fish were active during feeding. We doubt that wild skipjack tuna have excess core temperatures

5. BODY TEMPERATURE RELATIONS OF TUNAS

32 1

much different from those of equally active captive fish; but captive skipjack have not been stimulated to levels of activity as high as those attained by wild fish in feeding frenzy. Moreover, the observation that schools of skipjack tuna are highly variable in the extent to which frenzy feeding develops upon live-bait chumming (Yuen, 1959, 1966) suggests that variation in Tx's recorded in the field may be as much a function of activity and feeding motivation as of ambient temperature. Thus, we believe that the excess core temperature in wild skipjack is likely to be 1"-2"C per (kg)o,6except when extremely high metabolic rates are attained during a feeding frenzy or other violent activity. The effect of activity on core temperature appears to be different in bluefin tuna when compared to the response in skipjack tuna. Activity increases core temperature in skipjack tuna. Carey and Teal (1969b) have published evidence that exercise lowers core temperature in giant bluefin tuna. They found that core temperature of trap-caught fish (little activity because they were pursed up and shot in the head, usually in less than 5 min) was higher than in fish caught by hook and line (fought for 1 to 8 hr). Although the type of exercise produced by fighting on a hook and line is different from the exercise of chasing prey, there also may be a fundamental difference between bluefin and skipjack tunas related to size difference or to the type of retia (see Section V).

IV. WHAT IS THE HEAT SOURCE RESPONSIBLE FOR LARGE EXCESS BODY TEMPERATURES IN TUNAS? Skipjack tuna get much warmer during (or after) burst activity than they are during periods of routine activity. In typical fishes, white muscle produces much of the energy for burst activity (Rayner and Keenan, 1967), and it .is generally held that white muscle metabolizes anerobically. But anerobic metabolism in white muscle cannot produce enough heat to cause the very large temperature excesses observed in the muscle ofjust-caught tunas. If initial glycogen levels are high, say, 0.0131 g glycogen/g white muscle (mean of five highest values from Barrett and Connor, 1964), then instantaneous conversion of all the glycogen to lactate with all of the chemical energy appearing as heat would increase muscle temperature only 2.5"C! Yet in one group of skipjack w e observed white muscle excess of 8.6"C and red muscle excess of 9.1"C. So, where does the heat come from? Clearly, the major portion of the heat produced during burst swimming is aerobic. And most of aerobic heat production occurs in the red muscle. Histochemical and

E. DON STEVENS AND WILLIAM H. NEILL

322

electrophoretic studies show that the necessary enzymes are in high concentrations in red muscle, but in low cpncentrations or absent in white muscle. For example, 60% of the total cytochrome oxidase activity is present in red muscle in the eel even though red muscle constitutes only 13% of the total muscle (Malessa, 1969). Certainly, tuna white muscle possesses some ability to metabolize aerobically because Gordon (1968) was able to measure oxygen uptake from white muscle minces. But, red muscle-mince metabolism was 6 to 8 times that of white muscle. Because red muscle mass is one-sixth that of white muscle in tunas, red and white muscle make nearly equal contributions to total routine aerobic metabolism [assuming that Gordon's results are representative of routine metabolism-but see Neill et al. (1976)l. Values from Gordon for red muscle at 30°C are 1580 mm3 O2 g-' hr-1 (0.12 cal min-' g-'). For a body weight of 1 kg, that is, 1580mm3 X- 1 hr a hr 60min Y

1 cm3 x -4.7 cal 1000mm3 cm3

85.3g fish = 10.6 cal

min-' fish-'

Similarly, for white muscle (0.02 cal min-' g-'), aerobic metabolism in a 1-kg fish is 260mm3 x - 1 xx 4.7 x 574.7 = 11.7 cal min-' fish-' g hr 60 1000 The total contribution of red plus white muscle to aerobic metabolism, then, is 22.3 cal min-* fish-'. The value obtained from restrained skipjack is very similar: 25 cal min-' kg-' [mean of minimum values from six restrained skipjack (Stevens, 1972) 1. The value obtained from free-swimming skipjack is nearly identical. Gooding et al. (1976 personal communication) measured oxygen uptake from nine pairs of routinely active skipjack tuna. The mean of the nine pairs for the lowest (i.e., minimal activity) of four readings from each pair was 25.3 cal min-' kg-l. Thus oxygen uptake values determined from muscle minces, restrained fish, and free-swimming fish are similar. The values based on the muscle mince data must be slightly high because they neglect the contribution of tissues other than muscle to total metabolism. Tissues other than muscle contribute 2 5 4 5 % to the total metabolic rate for fishes swimming at speeds similar to those of routinely active skipjack tuna (Fry,1971). Neill et al. (1976) have estimated red muscle metabolism in freeswimming skipjack tuna based on measurements of red muscle temperature and the rate of cooling of red muscle. They reasoned that the rate of heat production is proportional to the product k, the coefficient

5.

BODY TEMPERATURE RELATIONS OF TUNAS

323

of temperature change, and T,, the excess muscle temperature under steady-state conditions. For a l-kg fish the value was 0.051 cal min-' g-1 . From the above estimates we conclude that red muscle metabolism during routine swimming in skipjack tuna is between 0.05 and 0.12 cal min-I g-I. Maximum values of oxygen uptake for free-swimming skipjack tuna during heightened activity (immediately after capture from a frenzy-feeding school) corresponded to a metabolic rate of 113 cal min-I kg-' (median value of fourteen observations was 74 cal min-' kg-') (Neill, unpublished observations). These values are more than twice as great as those for any other fish under any condition (cf. Brett, 1972; Fry and Brett, 1974). It is likely that much of the increase in oxygen uptake during heightened activity is due to red muscle, and that white muscle aerobic metabolism remains at 11.7 cal min-' fish-'. Using the maximum value, then, red muscle metabolism during this time is

113 - 11.7 =

min fish

x - -fish - 1.19 cal min-l g-' red muscle 85.3g

or 10-20 times the routine value. Neill et al. (1976) estimated red muscle metabolism in vigorously chased, captive skipjack to be only 0.42 cal min-' g-l, based on measurements of T , and k . If all the energy appears as heat and no heat is lost, 1.19 cal min-' g-' is sufficient to increase red muscle temperature 1.5"C in 1min, or 15°C in 10 min [assuming heat capacity of skipjack red muscle is 0.8 cal g-' OC-' (cf. Slavin, 1964; Charm and Moody, 1966)l. Obviously, some of the heat is lost: Some is lost to the environment, and some is distributed to other tissues by circulation of the blood. I n fact, the actual course of the temperature rise in red muscle can be calculated by taking heat loss into account. The rate of temperature change in the red muscle is the difference between the rate of potential temperature increase d u e to metabolism, or 1.5"C min-' in the present case, and the rate of potential temperature decrease d u e to cooling, k (Trm- T,), where k is the coefficient of temperature change in "C rnin-' "C-' and T,, is red muscle temperature:

dt

= 1.5"C min-I - k ( T , - T , )

rearranging,

5% dt

+ kT,,

=

1.5 + kT,

324

E . DON STEVENS AND WILLIAM H . NEILL

But (as later shown)

i.e., red muscle warms or cools at a rate that depends on the difference between its equilibrium (T,) and actual ( T,,) temperature (Neill and Stevens, 1974). To find out the value of T, in the present case, we combined the preceding pair of equations to get

+ kT,,

k(Te - T,,)

kT,

= =

1.5 + kT, 1.5 + kT,

or,

T,

=

15 - + T,

k

The value of k for a point in the deep red muscle of a 2 k g skipjack is about 0.033"Cmiii-' "C-' (Neill et ul., 1976). Thus, T, is 45.45"Cabove ambient temperature! Of course, our fish will not get that hot; it will simply end its bout of burst activity as muscle temperature approaches the upper critical level (perhaps about 35°C). But let's see what does happen to the red muscle of this 2-kg fish during a 10 min period of burst swimming in water at 20°C:

gives

T,,

= T, -

[T, - T,,( 0) I e-kt

Assuming the fish's red muscle is 2.0"Cwarmer than the water at the start of burst swimming [i.e., T,,(O) = 22.0°C], T,, rises to 34.21"C after 10 min as shown in the tabulation below. T,, if no heat loss ("C)

Actual T,,

2 4 7 10

22.0 23.5 25.0 28.0 32.5 37.0

22.0 23.41 24.78 27.37 30.96 34.21

(x

P

t (min)

0 1

-

-

("C)

45.45 = T,)

5.

BODY TEMPERATURE RELATIONS OF TUNAS

325

We conclude that aerobic metabolism occurring primarily in the red muscle reasonably accounts for the 10°C excess temperature measured in the muscle of skipjack tuna. This is probably also true for other tunas as well.

V. ANATOMICAL BASIS OF WARM-BODIEDNESS I N TUNAS A. Introduction Thus far, we have reviewed the evidence that tunas are warm (but not always extremely warm) and have explored the question of heat source, concluding that tunas are exceptional in their red muscle’s capacity for aerobic heat production. But, for any fish to be very warmbodied, more is required than heat production, even at the remarkedly high rates achieved by’ tunas. The difficulty, so eloquently described by Carey (1973), is that any heat produced in metabolically active tissues tends to be carried by the blood to the gills where it is lost to the water (for fish gills must be excellent exchangers of heat if they are to be effective exchangers of respiratory gases)-unless that heat is short-circuited, from the venous to arterial sides of the circulatory system, before it reaches the gills. This would appear to he exactly the function of certain retia mirabila (“wonderful nets”) found only in the tunas and fast-swimming sharks of the family Lamnidae. These retia are systems of parallel arterioles and venules in close contact; direction of blood flow in the arterioles is counter to that in the venules. The retia are so arranged in the circulatory system that their role as countercurrent heat exchangers has never beerl doubted; however, experimental verification of that function has been accomplished only recently. And the possibility that the retia exchange something other than heat remains unexamined. Before w e pursue physiological and ecological aspects of heat exchange in tunas, we first will review the special and diverse features of vascular anatomy that retard the flow of heat between the core tissues of tunas and their environment. To do so conveniently w e also must delve (although not deeply) into scombrid systematics and evolution. A discussion of the anatomy of the vascular system in different types of tunas is made difficult by the large number of names that have been employed to describe the same species. For example, Gibbs and Collette (1967) recognize only seven species within the genus Thunnus, but report that over 160 different scientific names have been

326

E. DON STEVENS AND WILLIAM H. NEILL

ascribedto the seven species. This problem has arisen because of the difficulty in maintaining and examining adequate collections ( 1000-lb tuna do not fit into the standard museum jar), and because many early workers held the erroneous opinion that the tunas were geographically limited rather than having, as most do, circumglobal distributions. Many workers have used the presence, location, and peculiarities of the vascular retia as important distinguishing characters in tuna systematics (Godsil and Holmberg, 1950). Nakamura (1965) has argued that it is more appropriate to base a taxonomic scheme on osteology. His contention is supported by the fact that cutaneous retia are also present in pelagic sharks (Bume, 1923; Carey and Teal, 1969a); such convergence indicates that vascular heat exchangers are evolutionarily quite plastic. The evolution of tunas may be characterized as a series of adaptations for very rapid swimming, both horizontally and vertically. The obvious features of tuna anatomy, physiology, and natural history certainly point in that direction. Although lamnid sharks, xiphoids, coryphaenids, and a number of carangids also can attain high burst speeds, tunas are exceptional in this regard. Scheme 1 summarizes important evolutionary developments for fast swimming in the family Scombridae. All of the Thunnini possess retia that presumably can act as countercurrent heat exchangers to reduce the rate of heat loss; none of the other tribes of Scombridae possess such devices. Still, our scheme is far from perfect. For example, loss of the gas bladder has occurred at more than one point in the scombrid line. Among the Spanish mackerels, Scomber scombrus lacks a gas bladder, whereas the chub mackerel, Scomber colias, has a well-developed gas bladder. Thunnus tonggol has only a rudimentary gas bladder.

B. Role of Vascular Anatomy in Tuna Systematics Vascular anatomy has been of importance in the systematics of tunas. Regan (1909) and Stark (1910), using the traditional classification characteristics of osteology, placed the tunas and mackerels in the single family Scombridae. Kishinouye ( 1915, 1917, 1923) recognized four separate families and placed the tunas in the order Plecostei, separate from all other higher bony fishes. His analysis was based in part on the extraordinary vascular anatomy of the tunas. His placement of tunas into a separate order has been followed by Berg (1955) and

5.

BODY TEMPERATURE RELATIONS OF TUNAS

327

(1) streamlined body of fusiform shape; dorsal, anal, and pectoral fins retract into grooves (2) sickle-shaped caudal fin with high aspect-ratio; dorsal and anal finlets to reduce turbulence of water in path of the caudal fin (3) large muscle mass relative to body length (4) white and red muscle fibers segregated; white muscle with very high concentrations of contractile proteins and substrate for anaerobic metabolism. (The above features tend to be progressively developed through the family) \ SCOMBRINI (primitive mackerels) and SCOMBEROMORINI (Spanish mackerels) ( 5 ) complete, bony peduncular keel develops [vertically stabilizes caudal fin and acts as a pulley to permit more powerful pull by tendons on the caudal fin (see Fierstine and Walters, 1968)l 1 SARDINI (bonitos)

THUNNINI (advanced tunas) ~ T H U N N U S (7) gas bladder lost (permits rapid vertical movements) (8) squamation reduced to anterior corselet of fine scales (presumably reduces drag at high speeds) (9) metabolism becomes temperature-independent (or nearly so) K A T S U W O N U S E U T H Y N N U S AUXIS

I

Scheme 1

Morrow (1957), but .has been disputed by most systematists (Takahashi, 1926; Fraser-Brunner, 1950; Gibbs and Collette, 1967; Collette and Chao, 1975). Certainly one of the main features that distinguish the tunas from other fishes are their retia and cutaneous blood vessels. The location and type of vascular rete have also been used to distinguish between the various taxa of advanced tunas. Table I is not meant to be systematically complete but provides a useful scheme for comparing extent and locations of retia (based on Gibbs and Collette, 1967; Nakamura, 1965). The very large tunas tend to have cutaneous retia but no central exchangers, whereas small tunas have central retia but cutaneous retia are either small or ab'sent. Gibbs and Collette (1967) reviewed the gross anatomy of the lateral and central vascular sheets; Carey et al. (1971) reviewed the anatomy of the lateral and

328

E. DON STEVENS AND WILLIAM H. NEILL

central retia in relation to measurements of excess muscle temperature; and Graham (1975) has related anatomy and body temperature to aspects of the natural history of the various tunas (Table I).

C. Systematic Distribution of Retia in Tunas 1. CUTANEOUSRETIAL SYSTEMS Among all of the bony fishes, only the genera Thunnus, Euthynnus, Katsuwonus, and Auxis possess cutaneous arteries and veins. This is one feature that separates the tunas from their relatives, the bonitos and mackerels. The cutaneous system is very large in the bluefin tuna group (see Table I), and in fact is the major avenue of blood supply for the muscle. The relative unimportance of central blood supply is evidenced b y a reduced dorsal aorta and the absence of the postcardinal vein. In all of the other tunas the cutaneous system is present, but the dorsal aorta and the postcardinal vein are also well developed. The anterior dorsal aorta is not different from that of other fishes. It is formed by the union of the efferent branchial arteries and continues along the dorsal body wall to the hemal canal. Partially encircling the aorta at this point is a peculiar, small, almost transparent ligament found in all the tunas (Godsil, 1954). It runs posteriorly, passes ventral to the aorta, and just posterior to the coeliacomesenteric artery it turns forward again toward the occipital region on the opposite side. Its function is not known. The unpaired coeliacomesenteric artery always arises on the right side of the dorsal aorta beneath the third or fourth vertebra to supply the liver and viscera. The small subclavian arteries arise at about the same point on the dorsal aorta. The paired cutaneous arteries arise posterior to the origin of the coeliacomesenteric artery and course laterad through the kidney and body musculature. The point of origin of the cutaneous arteries differs among the tunas (Table I). Along their length, the cutaneous arteries give off arterioles so dense that they form a continuous sheet, the cutaneous rete, that runs into and nourishes the muscle. The cutaneous arteries also give rise to segmental arteries, one per segment, and, although the cutaneous segmental artery and vein parallel one another, there is no vascular plexus.

a. Bluefin Tuna Group. In the bluefin tuna group, the dorsal aorta is much reduced in size after emergence of the cutaneous arteries. The cutaneous artery on each side forms two branches (hypaxial and epaxial) that run parallel. The branches are united caudally by a posterior commissure in all species of Thunnus except albacore and sometimes

Table I Specific Differences in the Heat Exchange Retia of Tunas: Family Scombridae, Tribe Thunnini" Cutaneous retia

Maximum

Bluefin group Thtrnnrrs thynnrrs

Common name

size (kg)

Bluefin tuna

730

Habitat

Retia

Vertebra number at which cutaneous artery originates

Number of rows of arterioles from cutaneous artery

Large ( 10 mm thick) Large

3-4

2

44

Migrates 6"-20"c Cool water

%4

1, Mesial side

Central rete

Visceral retia

Postcardinal vein

Central rete

Large retia on liver and gut Large retia on liver and gut Large retia on liver and gut

None

Thtrnnes altrlrrnga

Albacore

Thrrnnirs maccoyii

Southern bluefin tuna

225

Cool water

Large

%4

2

Bigeye tnna

198

Cooler water than yellowfin

Small (2-5 mm

6 8

2

Small retia

None

Very small (2-5 mm thick) Very small Very samll

6 8

1, Lateral

Arteriolar reticiilar network surrounds hepatic portal vein

6 8 6 8

Two central retia; total smaller than skipjack

1. Lateral Arteriolar reticular network surrounds hepatic portal vein

Largest

Intermediate Tlitcnntrs obestis

Yellowfin group Thunntrr albacores

Yellowfin tuna

200

Tropical

Blackfin hina Lnngtail tuna

16

27

Tropical Tropical

Skipjack tnna

22

Tropical

Etrthynnus afinis

Kawakawa

18

Tropical

Etrthynntrs lineattrs

Black skipjack

5

Tropical

Etrthynnus alletterotrrs

Atlantic

5

Tropical

Atrxis thaiard

Frigate mackerel

2

Tropical

Thrrnntts atlantictrs Thrtnnus tonggol Skipjack group Katsrriconis pelamis

None None

thick)

Small (1 mm thick) Small, expaxial rete only Small, expaxial rete only Small, expaxial rete only Small, expaxial rete only

5

1, Lateral

2 1 1

Large Large Large

6

Data from Gibbs and Collette (1967);Goadby (1972): Gndsil and Byers (1944); Graham (1975); Magnuson (1963); Nakamura (1965);Collette and Chao (1975).

Large

330

E. DON STEVENS AND WILLIAM H. NEILL

bigeye. In a large bluefin tuna, the capillary sheet of the retia can be as much as 1cm thick (Carey and Teal, 1969b).The vessels that make up the rete are about 0.1 mm in diameter. Vascular bands of alternating arterioles and venules as much as sixty vessels wide (1-4 mm) but only a single layer thick branch from the rete and penetrate the muscle. The vascular bands divide into triads (arte,riole-venule-arteriole) that eventually nourish the muscle. The vascular bands from the rete are the main blood supply for the red muscle. There is only one segmental branch from each cutaneous artery per body segment; the branch from the epaxial runs dorsad and that from the hypaxial runs ventrad. The segmentals give rise to smaller vascular bands, about twenty vessels wide and a single vessel layer thick. These bands also divided into triads (arteriole-venule-arteriole) that eventually provide the main blood supply to the white muscle.

b. Yellowfin Tuna Group. Thunnus albacares is unique in that it has large parallel trunks connecting posterior epibranchial and cutaneous arteries on each side. Thus blood enters the cutaneous artery from both this branch and via the dorsal aorta. The alternate avenues may be involved in the control of the distribution of blood, but this bypass is unique to this species. Other aspects of the cutaneous system are similar to those of the bluefin group except that the lateral rete is not quite as thick (does not contain as many vessels). The vessels in the rete and in the segmental vascular bands are similar in size to those in the bluefin tuna group. c. Skipjack Group. Katsuwonus, Euthynnus, and Auxis have in common the fact that the ventral branch of the cutaneous artery is short and dendritic, and does not form a vascular plexus. The cutaneous system of skipjack differs from that of all the other tunas in many respects. Only in the skipjack do the cutaneous arteries divide before reaching the ribs. The arterioles and venules are sparse and more erratic in course and origin than in other tunas. The dorsal and ventral cutaneous veins do not fuse anteriorly; they empty directly into the duct of Cuvier rather than into the postcardinal vein as in the case of all other species. In Euthynnus, the dorsal cutaneous artery runs the length of the body. It gives rise to dorsal and ventral segmental arteries that do not form vascular plexuses, but do run adherent to segmental veins. It also gives rise to a single row of arterioles that form the capillary sheet of the cutaneous rete. Euthynnus is also unique in that the dorsal cutaneous artery is always dorsal and adherent to the dorsal cutaneous vein. In Auxis, the dorsal branch of the cutaneous artery does form a

5. BODY TEMPERATURE RELATIONS OF TUNAS

33 1

rete-the artery is ventral to the vein and has one row of arterioles which form the rete; the vein has two rows of venules that pass to either side of the artery and into the rete. The dorsal segmental branches of the cutaneous artery are large, but the ventral branches are smaller and form superficial capillary strands.

2. VISCERALRETIA Carey et al. (1971) measured visceral temperature in tunas and found it varied; temperatures ranged from as high as those of the warmest muscle to slightly above that of ambient water. High visceral temperatures probably speed digestion and absorption, thereby compensating for the small size of the gut in tunas. a. Bluefin Tuna Group. The bluefin group of the genus Thunnus (Table I) is characterized by having large, complex retia on the liver and gut; these retia were first described (and beautifully illustrated) by Eschricht and Muller (1835).The branches of the coeliacomesenteric artery, upon reaching the liver, give off myriads of minute vessels that ramify in the lobes of the liver so that a transverse section of the liver of an injected specimen resembles the cross section of a rope. After passing through the liver the vessels pass into other retia dorsal to the liver in the form of large cone-shaped, vascular masses and then to the gut. The above also applies to the venous circulation. In a large bluefin, one of these liver retia may be as much as 5 cm in diameter.

b. Bigeye Tuna. There is a slight retial system associated with the liver in the bigeye tuna. c . Yellowfin and Skipjack Groups. In these tunas, the branches of the coeliacomesenteric artery pass directly through the substance of the liver without any plexus formation. The liver is nourished by a few discrete branches. Although there are no retia as in the bluefin, some branches of the coeliacomesenteric invariably become dispersed into a complex, fine network that completely envelopes that portion of the hepatic portal vein with which it is associated (Godsil, 1954; Godsil and Byers, 1944). 3. RETIA NEARTHE BRAINAND EYES Brain and eyes of all tunas yet examined are considerably warmer than ambient water but are not as warm as muscle (Table 11).Linthicum and Carey (1972) observed e y e and brain Tx’s of 16°C in two bluefin specimens taken from 7°C water. Heat is produced by the brain and eyes and is retained by the operation of the retia. The only

332

E. DON STEVENS AND WILLIAM H. NEILL

Table I1 Brain, Eye, and Muscle Temperature Excesses in Tunas"

Species

Thunnus thynnus Thunnus thynnus Thunnus alalunga Thunnus obesus Katsuwonus pelamis Euthynnus afinis Euthynnus alletteratur

Weight (kg)

180-400 6-12

2 2

Water temperaturd

Temperature excess ("C)

("C)

Brain

Eye

21.0 21.7 17.8 19.2 26.1 25.0 18.1

6.1 5.1 4.2 5.5 3.9 3.7 7.5

6.1 3.4

8.8

4.3 4.8

13.2 9.7

-

6.3

Muscle

9.2

8.8 4.8

11.7

Referenceb

1 1 1 1 2 2 1

In nontunas the excess is usually less than 0.5"C.

* 1, Linthicum and Carey (1972); 2, Stevens and Fry (1971). anatomical description of the retia supplying the brain and eyes is given by Linthicum and Carey (1972) for Thunnus, and the following is based on their account. The vascular exchangers (one on each side) in the brain of bluefin and albacore are located dorsal and anterior to the first efferent branchial arteries and are closely applied to the ventral surface of the prootic bone. The major source of blood is the carotid arteries; the opercular artery makes a smaller contribution. The external carotid gives off many small branches that contribute to the rete. The carotid retia are not as well developed in bigeye and yellowfin. The main vessels supplying the retia in the bigeye tuna are the external carotid arteries. In the yellowfin, branches of the external carotids pass through most of the retia before breaking up into small vessels; the opercular artery is more important than the carotids in supplying the retia. In all species so far examined the veins of the retia are arranged concentrically around the arteries. The arterioles in the rete are 80-120 pm in diameter with thick muscular walls; the venules are 40-150 pm in diameter. Linthicum and Carey (1972) pointed out the possibility of regulating brain temperature by regulating the proportions of blood flowing via alternate routes. They reported that the excess brain temperature of bluefin caught in cold water (T, = 7°C) was 16"C, whereas that of fish caught in warm water (T, = 23°C) was only 4.5"C. Although short-term regulation has not been demonstrated, avenues of blood flow are present that would constitute a partial bypass of the rete.

5 . BODY TEMPERATURE RELATIONS OF TUNAS

333

4. THE CENTRALRETE The central rete is best developed in species with the least developed cutaneous retia and is completely absent in the bluefin tuna group, which has the most elaborate lateral retia system.

a. Bluefin Group and Bigeye Tuna. In these tunas the central rete is completely absent. In the bluefin group the postcardinal vein is absent and the dorsal aorta becomes very small posterior to the point at which the cutaneous arteries branch. In the bigeye tuna, a small postcardinal vein is present.

b. Yellowfin Group. The central rete is small and forms as two capillary sheets about 1 mm thick, one on either side of the dorsal aorta. c. Skipjack Group. In Euthynnus lineatus the venules in the central rete do not branch into smaller vessels (Godsil, 1954), but are surrounded by a mass of arterioles. The extent of branching and size of the retial vessels in the other Euthynnus species and Auxis are unknown. The central rete is best developed in Katsuwonus (Stevens et al., 1974). Cool arterial blood passes from the dorsal aorta through the central rete and then to segmental arteries. Warm venous blood from the tissues collects in segmental veins and then passes throhgh the rete to the postcardinal vein. In a 2 k g skipjack the rete is about 7 mm wide and the exchanger vessels are 10 mm long. The arterial vessels in the rete are approximately the same diameter as arterioles and so are referred to as arterioles. These vessels are thick walled and contain smooth muscle but are not innervated. The retial arteriples unite into collecting vessels at the top of the rete, and the collecting vessels become segmental vessels. (Figs. 3, 4,and 5). At the top of the rete, segmental veins divide to become collecting veins that are thin walled and invariably contain regularly spaced protuberances which look like incomplete septa. These protuberances probably act as valves to prevent backflow due to the high resistance offered by the venules in the rete. The collecting veins divide repeatedly to form the venules of the rete. Venous blood, after passing through the rete, collects and passes around the dorsal aorta to empty into the postcardinal vein. The number of arterioles is very nearly equal to the number of venules in the rete, but the venules are much larger. Numbers and dimensions of central rete vessels in a 1.9kg skipjack tuna examined by Stevens et al. (1974) are shown in the following tabulation.

334

E. DON STEVENS AND WILLIAM H. NEILL

Number of vessels Internal diameter (pm) Cross-sectioned area (cmz) Volume (cm3) Inside surface area (em')

Arterioles

Venules

133,721 35.7 f 1.32 1.34 1.21 1350

120,652 83.82 4.47 6.65 5.99 2858

The presence of the rete reduces the velocity of blood flow in the rete arterioles to 1/80 of that in the dorsal aorta, that in the rete venules to l/so of that in the postcardinal vein. The rete vessels contain about 5% of the total blood volume. VI. EXCHANGE OF HEAT BETWEEN TUNAS AND THEIR ENVIRONMENT

A. The Rate of Change of Core Temperature The observations by Stevens and Fry (1971)showed that the large excess temperatures observed in immediately sampled, ocean-caught tuna decreased to about one-half within 20 min after capture (for dead

Fig. 3. Schematic drawing illustrating the features of the vascular heat exchanger of skipjack tuna. DA, dorsal aorta; K, kidney; M, muscle; PCV, postcardinal vein; V, vertebra. Left inset shows pattern of arterial blood flow, right inset shows pattern of venous blood flow, and small arrows indicate heat transfer from venules to arterioles in the exchanger. (From Stevens et ul., 1974.)

5.

BODY TEMPERATURE RELATIONS OF TUNAS

335

Fig. 4. (A) Cross section of a 2-kg skipjack tuna. The vascular heat exchanger is below the vertebral canal. Red muscle appears dark. (B) Tangential section ofthe vascular heat exchanger of a formalin-perfused skipjack tuna (that is, a cross section of the vessels of the exchanger). The number of thick-walled arterioles is about the same as the number of thin-walled veins. (C) Same section as (B) except that the tissue was fixed rather than perfused. Red blood cells (10 pm diameter) are evident and reveal the size of the vessels. The arterioles have thick, muscular walls, whereas the venules are thin walled. (From Stevens et ( I / . , 1974.)

336

E. DON STEVENS AND WILLIAM H. NEILL

-

Fig. 5. Cross section of skipjack hina showing - the vascular heat exchanger. ComDare with Fig. 3. Small vessels in the exchanger are about 1 cm long. (From Stevens et al., 1974.) ~~

skipjack in air) and to about one-third after a few days in captivity. Neil1 et al. (1976) have examined in detail the rate of change of core temperature as a function of changes in ambient temperature in restrairied skipjack tuna. In their experiments, tuna were subjected to

5.

BODY TEMPERATURE RELATIONS OF TUNAS

337

5°C increases or decreases in T,, and tissue temperatures were continuously monitored until they ceased changing. Core temperatures of larger fish changed more slowly than those of small fish. The coefficient of temperature change, k, when plotted on double logarithmic axes against body weight, was linear with a slope of -0.45 for red muscle, white muscle, and brain (Fig. 6A). Brain cooled and warmed about 3.3 times as rapidly as muscle of the same fish, over the fish weight range of 0.1-3.5 kg. There was no difference between the rates the tissues cooled and the rates at which they warmed. Stevens and Fry (1974) have compiled coefficients of core temperature change for a variety of fishes and aquatic reptiles. Comparing k's of animals of equal weight, skipjack tuna come into steady state with a new water temperature only about 60%as rapidly as typical teleosts and even somewhat more slowly than aquatic reptiles (Neil1 et al., 1976; see also Fig. 6B). Considering that fishes must achieve respiratory exchange with water (which has much greater heat capacity and

1

0.20

0

O.

0.10-

t

- 0.06x

0.02'

' ' ' ' "' 0 4 0.6 1.0

'

'

2.0

'

.O 4.0

Body Weiqht (hq)

; 1.-

0.10

o Turtle d

Marine lquana

\

Fig. 6 . (A) Relation between fish weight and the coefficient of temperature change (k) in the red muscle, white muscle, and brain of the skipjack tuna. (B) Relation between skipjack body weight and red muscle k compared with that for other fishes and aquatic reptiles.

338

E. DON STEVENS AND WILLIAM H. NEILL

much less oxygen per unit volume than air), the observation that skipjack tuna change temperature more slowly than aquatic lung breathers is evidence of high efficiency in the skipjack's countercurrent heat exchanger.

B. Relative Heat Exchange via Gills and Body Surface Brill et al. (1978)have measured the rate of heat flow from the body surface in restrained skipjack tuna when ambient temperature was not changing. These results show that 10-40% of heat produced (estimated from oxygen uptake) was transferred through the skin. Furthermore, heat was transferred in a relatively uniform fashion across the body surface; that is, there were no areas where heat flow was exceptionally high or low. The rate of heat flow was related to the volume of muscle mass below the surface. The only area of significant heat loss in the head was the area over the eye. The rather uniform pattern of heat loss in skipjack contrasts sharply with that in Thunnus. Carey (1976 personal communication) measured heat loss in yellowfin, bigeye, and bluefin; all showed the same pattern: Heat loss varied by a factor of 2-4 around a fish's body and was highest in the region of the pectoral fin groove where probing with thermistors showed the steepest thermal gradients. The different patterns of heat loss in Thunnus, compared to the uniform heat loss in skipjack, is clearly related to the different positions of the heat exchangers in these fish. That 10-40% of the metabolic heat in skipjack tuna is transferred through the body surface is consistent with the findings of Neill et al. (1976). From their Fig. 7 it may be estimated that heat transfer between the environment and a point in the deep red muscle is partitioned between the gills and the general body surface 5 0 : 50% in a 0.5kg skipjack and 65: 35% in a 3.0-kg fish.

C. Estimates of Effectiveness for Heat Exchange

The effectiveness of the heat exchanger may be estimated by a number of different methods. Obviously it is highly effective because the rate of heat exchange in tuna is less than in nontunas (Figs. 6B and 11). Neill et al. (1976)have computed the efficiency of the skipjack tuna exchanger as a thermal barrier based on the following estimates and assumptions.

5.

BODY TEMPERATURE RELATIONS OF TUNAS

339

1. Temperature excess of venous blood just before it reaches the gills is about 0.1 times that of red muscle (measured). 2. Arterial blood entering the heat exchanger has a negligibly small temperature excess (assumed). 3. Venous blood entering the heat exchanger has a temperature excess equal to that of red muscle (assumed). 4. Venous blood not passing through the exchanger has temperature excess equal to that of the body average excluding red muscle (=0.2 T, of red muscle, assumed). 5. Of venous blood in the bulbus arteriosus 60% has passed through the heat exchanger, and 40% has not (assumed by Stevens et al., 1974). 6. Changes in blood temperature during transit of large arteries and veins are negligibly small (assumed). Then 0.60(Txexchanger output blood) + 0.4O(Tx nonexchanger output blood) = O.l(T, red muscle), and 0.60(Tx exchanger output blood) = O.l(T, red muscle) - 0.4(0.2 x T , red muscle) = 0.O2(Tx red muscle). Therefore, T , exchanger output blood = 0.02/0.60 = 0.O3(Tx

Fig. 7.Record of a heat pulse applied to the water perfusing the gills of a restrained skipjack tuna. Note that the pulse is contained within the vascular compartment and appears in venous blood but does not change the tissue temperature. The change in blood temperature has a direct effect on the heart; its rate increases with an increase in temperature.

340

E. DON STEVENS AND WILLIAM H. NEILL

red muscle). That is, 97% of the heat is transferred from venous to arterial blood within the exchanger. If the 0.2 value in assumption (4) is, in fact, between 0.1 and 0.4, or if the 60% value in assumption (5) is, in fact, between 50 and loo%, then the efficiency of the exchanger still lies between 90 and 100%. We also have examined more directly the effectiveness of the heat exchanger as a thermal barrier by looking at what happens to a heat pulse within the vascular system of skipjack tuna (Fig. 7). A heat pulse applied to the water perfusing the gills is transferred to blood passing through the gills and appears in the dorsal aorta. The efficiency of the heat exchanger as a thermal barrier is revealed by the extent to which the temperature pulse is contained within the vascular compartment and not transferred to the tissues. Thus, if the exchanger were very efficient, the heat pulse should be transferred to venous blood in the exchanger (and not to the tissues) but will be reduced in magnitude because of dilution by blood not passing through the exchanger and of longer duration because of unequal path lengths within the circulatory system. In our experiments approximately 150 cal were put into a skipjack's dorsal aorta (assuming that the temperature of blood came into equilibrium with that of water) when perfusion-water temperature was increased 5°C for 15 sec. Muscle temperature was unaffected by the heat pulse. The heat pulse appeared in the ventral aorta in about 15 sec or about three times as fast as if it had traversed the entire vascular system. The temperature of the blood in the ventral aorta only increased 0.5"C; the rise persisted more than 15 sec due to unequal path lengths within the circulatory system. More important, 73-85% of the pulse did appear in the ventral aorta. That is, assuming that the exchanger effectiveness was loo%,73-85% of the blood flow passed through the exchanger. If exchanger effectiveness was only 95%,then 77-88% of the blood flow passed through the exchanger. There is no question that heat is exchanged very effectively in the rete and that a large fraction of the blood flows through the rete in skipjack tuna.

VII. DO TUNAS REGULATE THEIR BODY TEMPERATURE? Definitive experiments to demonstrate whether tunas do, in fact, acutely regulate body temperature have not been carried out. Thermoregulation implies regulation of either heat gain (metabolic rate), or heat loss, or both in order to achieve a specific body temperature or excess body temperature suited to a particular circumstance. It is our

341

5. BODY TEMPERATURE RELATIONS O F TUNAS

view that thennoregulation does not require that body temperature be indefinitely maintained at some absolute level. There is reasonable evidence that some tunas thermoregulate.

A. Regulation of Heat Gain 1. EFFECTOF AMBIENT TEMPERATURE ON ROUTINE OXYGENUPTAKE Tunas could, all other things being equal, elevate body temperature by increasing metabolic rate. If the object of the game is to maintain body temperature constant, then we would predict that tunas would elevate metabolic rate at lower ambient temperatures. Only one experiment of this type has been carried out with tuna (Chang et al., 1976). In this experiment pairs of skipjack tuna were placed in a large annular respirometer, and oxygen uptake was monitored during spontaneous activity at a variety of ambient temperatures. Over the range of temperatures tested, 18"-30"C, there was no significant effect of temperature on the rate of oxygen uptake during spontaneous activity (Fig. 8).We can conclude that skipjack tuna do not maintain body temperature by increasing heat production when ambient temper, ature is decreased. I n contrast with many other fishes, h o ~ e v e rthey 1

.

5

1.0-

'

~I

~,

,

,

,

,

,

,

,

I

0

.a .7-

0

0"

,

73

.9-

-0 -k

1

/

C

.6-

.sd

.4 16

'

1

18

'

1

20

I

'

22

1

'

24

1

"

26

'

28

1

I

Fig. 8. Oxygen uptake in free-swimming skipjack tuna at a variety of ambient water temperatures. Fish were acclimated to 25°C.

342

E. DON STEVENS AND WILLIAM H. NEILL

do maintain metabolic rate constant over a wide range of ambient temperatbes during routine activity. Since tunas swim continuously, the major site of metabolic demand (and, thus, the major source of heat production) is the red muscle mass used during routine swimming activity. It is not surprising, then, that swimming velocity is also independent of temperature. ON ROUTINE 2. EFFECT OF AMBIENTTEMPERATURE SWIMMING SPEED

Reports of swimming speed as a function of ambient temperature indicate that, at least in skipjack tuna, routine swimming speed is maintained constant independent of changes in ambient temperatures. Tunas swim continuously in order to maintain vertical position (they are negatively buoyant) and also to maintain an adequate water flow over the respiratory interface (they ram water over the gills rather than pump it). Standard metabolism is thus meaningless for these obligatory swimmers. So, experimental work has focused instead on routine spontaneous activity and associated metabolic demands. In a crude experiment, Stevens and Fry (1972) reported that routine swimming speed of skipjack tuna remained unchanged over a temperature range of 16"-25"C. Subsequent and more refined experiments by Dizon et al. (1976) have also shown that skipjack tuna swimming speed remains independent of temperature over a temperature range of 18"-30"C (Fig. 9A). Interestingly, Dizon et al. (1976) have shown that yellowfin tuna, Thunnus albacares, decrease their swimming velocity in response to decreases in temperature (Fig. 9B). Deductions regarding benefits derived by the skipjack through maintaining constant swimming velocity in the face of changing ambient temperatures are complicated by the fact that temperature change also causes changes in the characteristics of water. For example, when ambient temperature is decreased from 30" to 20°C there is a benefit to the tuna because oxygen solubility increases about 15%. At the same time, however, there is a 26% increase in the viscosity of seawater that causes a 12% increase in frictional drag (Brown and Muir, 1970; Webb, 1975).Because the benefit of the increase in oxygen solubility is offset by the increase in drag, the metabolic cost of respiration at any particular swimming speed remains unchanged. In addition, a change in body temperature may affect the efficiency with which chemical energy is converted to mechanical work in the muscles, and the effi-

5.

343

BODY TEMPERATURE RELATIONS OF TUNAS I

1.6

-

I

I

I

SKIPJACK TUNA 47.0 cm F.L.

1.4-

I

I

1

,

I

28 -

22

M

1

,

x)

100

I50

200

250

300

350

ELAPSED TIME (MINI

Fig. 9. Swimming speed in free-swimming skipjack and yellowfin tunas at a variety of ambient water temperatures. Fish were acclimated to 25°C.

ciency of converting mechanical work in swimming muscles to propelling power; neither of these efficiencies has been quantified for swimming tuna. In conclusion, metabolic rates at various temperatures have been determined only for skipjack tuna, in which metabolic rate during routine swimming activity is constant over a wide range of ambient temperatures. Routine swimming speed also is invariant over a wide range of ambient temperatures and probably accounts for the constancy of metabolic rate because muscular work constitutes the major energy expenditure. In another tuna, the yellowfin, routine swimming speed decreases when ambient temperature is lowered, and it is likely that there is a concomitant decrease in metabolic rate. There is no evidence that tunas increase heat production in response to a decrease in ambient temperature.

344

E. DON STEVENS AND WILLIAM H. NEILL

B. Regulation of Heat Loss Body temperature regulation could be achieved, even if there is no direct regulation of heat gain, by regulation of the rate of heat loss.

1. SKIPJACK TUNADO NOT REGULATETHE U T E O F HEATLOSS The rate of heat loss, or more exactly, the rate of cooling and warming, under a variety of experimental circumstances has been examined in detail in skipjack tuna by Neill et al. (1976; see also Section VI,A). The rate at which red and white muscle cooled was the same as the rate at which they warmed. Moreover, the rate of cooling and warming did not vary systematically with the level of ambient temperature. Therefore, skipjack tuna do not seem to regulate the rate ofheat loss in response to changes in ambient temperature.

2. THERMALINERTIA IN LARGETUNAS Carey and his associates have used telemetry techniques to make many observations of core and ambient temperatures in freeswimming giant bluefin tuna. On the basis of these observations they have argued that giant bluefin tuna encountering abrupt changes of ambient temperature are capable of rapid physiological thermoregulation in the same sense as mammals. Their evidence is that body temperature changes very little and very slowly when the fish swims into water of very different temperature. Neill and Stevens (1974) have offered a different interpretation of these telemetry data-i.e., the observed stability of body temperature in the face of changes in water temperature could reflect a constant but small k coupled with constant heat production. Neill and Stevens refer to this as thermal inertia, as opposed to physiological thermoregulation. The analysis by Neill and Stevens is based on the assumption that the process of heat exchange in large bluefin tuna can be described by the same relationships that appear to be valid for skipjack tuna. That is, the rate at which the fish warms or cools is proportional to the disequilibration temperature excess (T, - Tb):

where dTb/dt, change in body temperature ("C min-I); k, coefficient of temperature change ("Cmin-' "C-l);T,, equilibrium temperature, the body temperature that would ultimately obtain at any given ambient temperature; T , = T , T,. Neill et al. (1976) have shown that this model adequately describes the rate of cooling in skipjack tuna.

+

345

5. BODY TEMPERATURE RELATIONS OF TUNAS

Neill and Stevens used the above model to simulate body temperature responses observed by Carey and Lawson (1973). For the analysis we assumed that the fish did not thermoregulate-i.e., did not alter heat gain or heat loss. Body temperatures predicted under the assumption of no physiological thermoregulation agreed very well with the observed body temperatures (Fig. 10).Thus it was possible to explain, in large part, the telemetry data by assuming that these tuna

. .. .. ........

9.

0

.

..

..

a.

B

Fig. 10. Actual and modeled body temperatures of two bluefin tuna swimming in a heterothermal environment. Actual body ( X ) and water (0)temperatures were estimated at 20 min intervals and to 0.5"C from graphs presented by Carey and Lawson (1973).By exponentially filtering the water temperature series to which a constant temperature had been added, modeled body temperatures (-0-)were generated under the hypothesis that each fish had a constant rate, i.e., did not physiologically regulate body temperature. Lack of substantial physiological thermoregulation is suggested by how closely the model, optimized with reasonable values ofthe parameters, could b e made to fit the data. (A) Muscle temperature, bluefin tuna No. 8. (B, C) Stomach temperature, bluefin tuna No. 14. (From Neill and Stevens, 1974, Science 184, 1008-1010. Copyright 1974 b y the American Association for the Advancement of Science.)

-

7

346

E. DON STEVENS AND WILLIAM H. NEILL

did not thermoregulate physiologically. We emphasize here (as we did in 1974) that this interpretation does not preclude the possibility of physiological thermoregulation; it simply shows that we do not have definitive evidence proving or disproving rapid physiological thermoregulation in tunas. It may be that, in order to demonstrate physiological thermoregulation in tunas, the fish must not be under any nonthermal stress. The Neill-Stevens model also shows that thermal inertia is important in its own right, especially in large tunas. Large thermal inertia makes possible a kind of physical thermoregulation that can result in large excess body temperatures and protection of core tissues from large, rapid fluctuations in environmental temperature persisting for periods up to several hours.

C. Long-Term Temperature Responses or Thermoregulation? In the previous section we have argued that there is no definitive evidence to support the hypothesis of short-term physiological thermoregulation in tunas. The hypothesis when originally proposed was based on observations of body temperature in tunas caught at a variety of locations with different surface water temperatures (Carey and Teal, 1966; Stevens and Fry, 1971). Excess body temperatures of tunas caught from cold water tend to be greater than those of conspecifics caught from warmer water, i.e., regressions of observed body temperature on water temperature tend to have slopes less than unity (Fig. 1). The relation is weak for yellowfin tuna, stronger for skipjack tuna, and strongest for bluefin tuna, in which observed body temperatures are nearly independent of water temperature. These observations have suggested to their examiners (Carey and Teal, 1969b; Stevens and Fry, 1971) that bluefin tuna, skipjack tuna, and yellowfin tuna are, respectively, good, fair, and poor thermoregulators. The presumption was that the body temperatures measured in just-caught fish are unbiased estimates of typical temperatures maintained by tunas as they go about their routine activities. For skipjack tuna we now are reasonably confident that this interpretation is invalid; rather, we believe that muscle excess in unexcited, routinely active skipjack is only 2”-4”C,independent of ambient temperature. But, when the fish are feeding or engaged in other speed-demanding activity, excess increases dramatically, doubling or trebling for periods up to several hours and reaching even higher levels for briefer periods. We think it likely that the skipjack tuna “regulates” only the upper

5.

BODY TEMPERATURE RELATIONS OF TUNAS

347

limit of core temperature, by simply reducing the intensity, duration, and/or frequency of activity bouts as body temperature approaches some critically high value (about 35°C; see Neil1 et al., 1976). According to our view, then, the thing that changes as a function of ambient temperature is neither resistance to heat transfer (k)nor “normal” rate of heat production but rather the fish’s thermal scope for activity, being greatest at low ambient temperatures and approaching zero at T, = [(-35°C) - routine T, 1. Decrease in thermal scope for activity with increased T, would lead automatically to distributions of observed body temperature that ( 1)are related to ambient temperature by slopes less than 1.0and (2) have lower variance at higher Ta’s. Neither of these results is inconsistent with the actual data (Fig. 1). Some major sources of variation in observed muscle temperatures of skipjack tuna, then, are (1)the extent of muscular activity during the feeding frenzy preceding capture (which may differ for skipjack in water of different temperatures because of direct effects of temperature on the fish or because of indirect effects of temperature acting to influence the kinds, distribution, and behavior of natural prey and also behavior of the live chum); (2) the time elapsing between initiation of heightened activity and measurement of muscle temperature (which varies with fishing tactics and success, with fish size, between observers, and also for a particular observer in differing sea conditions); and, (3) the immediate thermal history of the fish (which varies with vertical distribution of the fish in conjunction with thermal structure of the ocean in the fishing area). The last error source is particularly troublesome in that the T , recorded opposite a particular body temperature observation may be very much in error; a T , measured at the sea surface (as is customary) necessarily overestimates the ambient temperature experience of any fish not swimming continuously at the surface during the hour or so preceding capture. Some of the factors that cloud the issue of physiological thermoregulation in skipjack tuna also apply in the case of the bluefin tuna. However, the weight of the evidence for thermoregulation in large bluefin is convincing. Bluefin in 7°C water have an excess of 19”C, but the excess is only 2.4”C in 30°C water (Fig. 1). The regression line describing the relation between bluefin muscle and water temperature, based on all available data (173 fish; T, = 7”-30°C; Carey, personal communication, Woods Hole Oceanographic Institution), is

Tb = (0.237

* O.028)Ta + 25.25”C

where 0.028 is the standard error of the slope. Unfortunately, the data at extreme values of T , are least reliable. The fish from warmest water

348

E. DON STEVENS AND WILLIAM H. NEILL

were taken by hook and line rather than by trapping; in Carey's opinion, the hook-caught fish were more likely to have had abnormal body temperatures than fish taken in traps. The trap-caught fish taken at lowest temperatures had been at those temperatures less than 24 hr (following a storm that brought cold water into the trapping area); therefore, these fish may not have been in thermal steady-state at the time measurements were made. But, even if these questionable data are deleted, the regression relation changes very little; the more conservative regression line, based on 162 fish from Ta's between 13"and 23"C, is

Tb = (0.206

* O.049)Ta + 25.84"C

The slope of this line is significantly different from that of perfect conformers (1.0)and, in fact, is nearer that of perfect thermoregulators (0.0). When one considers that heightened activity in bluefin tuna decreases excess body temperature (Carey, personal communication), the above regression relationship constitutes convincing evidence that these fish thermoregulate. Bluefin tuna migrate annually and may acclimate to different water temperature conditions. Giant bluefin tuna of the western Atlantic spend winter in the warm south (Gulf of Mexico and Caribbean Sea) and migrate north from May to June to spend the summer in the northwest Atlantic (Butler, 1971). They enter Conception Bay, Newfoundland, when the water temperature is 6°C but are rarely hooked until water temperature increases to 10°C.It is very likely that anumber of adaptive physiological acclimatory processes, including an alteration in the anatomy of the rete or an alteration in the pattern of circulation through the various retia, occur during the migration. It may be that acclimation is different in some tunas than from typical teleosts in that tunas may compensate for cold to some extent by producing more heat or reducing the rate at which it is lost. However, nothing empirical is known about thermal acclimatory responses of tunas.

VIII. ADAPTIVE VALUES OF WARM-BODIEDNESS AND LARGE THERMAL INERTIA Tunas are warm-bodied both because they produce metabolic heat at near mammalian rates and because their retia retard the loss of that heat to the water. Their exceptionally high metabolic rates are, in turn, associated with their capability for exceptionally fast swimming. We begin our search for adaptive values by accepting a priori that,

5.

BODY TEMPERATURE RELATIONS OF TUNAS

349

for tunas, the capability for fast swimming is adaptive; this leads to a consideration of how warm-bodiedness may promote fast swimming. Then we return to the issue of physiological thermoregulation in tunas. Finally we consider some potential benefits of large thermal inertia that are independent of warm-bodiedness and, also therefore, of heat production.

A. Increased Muscle Power at Elevated Temperatures Carey has argued (Carey et al., 1971) that the adaptive value of tunas' heat exchangers lies in elevated body temperatures which promote greater muscle power for swimming. This is a logical proposition, but has not been empirically demonstrated for tuna muscle. Carey's original argument was based on experiments with isolated frog muscle b y Hill (1951). More recently, Wardle (1975) has shown that the time taken from a stimulating pulse to the peak of contraction in isolated fish white muscle and in intact fish is very temperaturesensitive. An increase in body temperature from 14"to 20°C was associated with an increase in maximum burst swimming speed from about 4.5 m sec-' to 5.3 m sec-' for a fish with a body length of 50 cm. Neil1 and Dizon (unpublished observations) have cinephotographed captive skipjack tuna, as these fish chased food thrown across a large tank, swimming at speeds on the order of 18 body lengths (C) sec-' (Le., 10 m sec-l for 50-cm fish) at an ambient temperature of 24°C. Muscle temperature during this period was not measured, but it was likely to have been between 30" and 32°C. The observed speed is greater than that predicted [5.8 m sec-' at 24°C extrapolated from the data of Wardle (1975)l by 72%,and some portion of this greater speed is likely attributable to the higher body temperature. Extrapolating Wardle's data to a body .temperature of 32°C yields a predicted maximum burst speed of about 8 m sec-*. Thus, tunas can burst-swim at greater speeds than nontunas and tunas are warmer than nontunas. But are tunas warmer because they swim faster, or are they faster because they are warmer? One difficulty in assessing this question is that there are no good simultaneous observations on fishes' core temperatures and maximum swimming speeds during burst activity. Walters and Fierstine (1964) reported a maximum of 21 e sec-l for a tuna (yellowfin) and 19 e sec-l for a nontuna (wahoo, Acanthocybium solandri).Yet the excess core temperature of the tuna was probably twice that of the nontuna. [There are many measurements of core temperatures in tunalike nontunas (Lindsey, 1968; Carey et al., 1971); all indicate that these fishes' core temperatures scarcely exceed l"C.1

350

E. DON STEVENS AND WILLIAM H. NEILL

B. Decreased Time between Bouts of Burst Activity at Elevated Temperatures Black and his colleagues have shown that the recovery time from severe exercise is very long in fish compared to mammals (Stevens and Black, 1966).For typical fishes, 12-24 hr elapse before levels of muscle and blood lactate return to pre-exercise levels, whereas in man recovery takes only 30-60 min. Barrett and Connor (1962, 1964) showed that in yellowfin and skipjack tuna lactate levels return to pre-exercise levels in about 2 hr-much more quickIy than in nontunas. Skipjack tuna, at least, are especially warm after severe exercise, and they recover from exercise quickly. These two aspects of tuna physiology may be related causally; this may be an important value of warm-bodiedness, considering that the recovery process is likely temperature-sensitive. (The temperature-sensitive process could be a physiochemical process such as the diffusion of lactate from muscle to blood or it could be biochemical.) An adaptive value of being warm after severe exercise then is that it speeds recovery, thereby permitting more frequent feeding frenzies with concomitant increases in the rate of food capture and perhaps of growth.

C. Benefits Accruing from Body Temperature Stability Whether or not tunas benefit from being consistently warmer than their environment (and relatively warmer than other fishes), tunas may profit from a relatively stable body temperature.

1. PHYSIOLOGICAL THERMOREGULATION Physiological thermoregulation permits optimization of physiological processes over a relatively narrow range of temperature. No doubt fishes would enjoy the benefits of physiological thermoregulation to the same extent that birds and mammals do-if its costs were not so great for the water breathers. Tunas, however, have already made an evolutionary investment in physiological thermoregulation. And the evidence suggests that large bluefin tuna, at least, practice substantial thermoregulation. But, why do the tunas as a group not physiologically thermoregulate more effectively on a short-term basis, simply by passing more or less blood through their exchangers? The answer seems to be that the tunas have not yet cashed in on their investment.

5.

BODY TEMPERATURE RELATIONS OF TUNAS

351

2. THERMAL LAG-MORE TIMEBELOWTHE THERMOCLINE Even in the absence of physiological thermoregulation, tunas achieve substantial independence from fluctuations of environmental temperature through large thermal inertia. We have shown that skipjack tuna cool only one-half as quickly as a nontuna of the same weight when exposed to the same decrease in ambient temperature. Thus, if the rate of some process is controlled by temperature, then the warmer tuna have an advantage. The small difference between the muscle temperature of the tuna and nontuna in steady state becomes considerably magnified when ambient temperature is changing (Fig. 11). Even small tunas, say 3-kg skipjack, can sound for several minutes into the cool water below the thermocline with only a small depression of core temperature. If higher core temperatures are necessary for faster swimming, these minutes of thermal lag may mean the difference between life and death to a skipjack being pursued through the thermocline by a marlin.

3. INSTANTANEOUSTEMPERATURE COMPENSATION-ADISCLAIMER There is some evidence that certain enzymes in some organisms compensate instantaneously for changes in temperature by changing

n CORE EXCESS

STEADY STATE

E$''

"Nu

0 03725

EXTRA TIME TO COOL 5 C'

X EXTRA TIME

:

IS 7 I min NONTUNA'K' - TUNA'K" NONTUNA *K'

,,oo

* 38 X For 3Lp

Time Imin)

Fig. 11. Simulated changes in core temperature of a tuna compared to a nontuna when ambient temperature decreases from 26" to 16°C (for example, if both swam through the thermocline). It takes 38% longer for the tuna to cool 5°C so that it will have an advantage over the nontuna if some processes slow upon temperature decrease.

352

E. DON STEVENS AND WILLIAM H. NEILL

affinity for substrate and for modulators (Hazel and Prosser, 1974). If this is so in tunas, then constant temperature is of no advantage in enzyme catalyzed reactions. However, temperature stability may be very important in certain physiochemical processes that cannot be compensated for by changes in enzyme characteristics. For example, the rate at which muscle can contract is limited by the rate at which Ca2+can move from the sarcoplasmic reticulum into the sarcoplasm. The diffusion of Ca2+ into the sarcoplasm may be extremely temperature-sensitive; if so, maximum swimming speed is temperature-dependent, whether or not the enzymatic machinery is instantaneously temperature compensated.

D. Perception of Weak Horizontal Temperature Gradients The rete increases thermal inertia to such an extent that it may enhance or permit detection of weak horizontal gradients of temperature (Neill et al., 1976). Such a mechanism could account for the effect of weak horizontal temperature gradients on the distribution of tunas in the oceans. The tuna changes core temperature so slowly that the difference between core temperature and a slowly changing ambient temperature increases sufficiently to be detected. Neill et al. (1976) suggest that fishes perceive changes of environmental temperature as a disequilibration of core temperature (Tbf T,). If this is so, tunas have a distinct advantage over typical fishes in sensing weak temperature gradients. Not only do tunas swim faster, thereby enhancing contrasts of ambient temperature over time, but their large thermal inertia causes core temperature to lag substantially for even very gradual changes in environmental temperature. Neill et al. illustrated the second point by simulating core temperature in a 2-kg skipjack swimming into a gradient of only 1°C km-' at a speed of 1.77 e sec-l (50 m min-'). After the fish had swam only 250 m into the gradient (5 min), core temperature had diverged from its equilibrium value by 0.23"C, which is approximately twice the threshold for detection of a temperature step-change in tuna (Steffel et al., 1978). For the simulation to yield the same result with a Zkg nontuna (which has about half the thermal inertia and swims at about half the speed of the tuna), the required gradient is four times as steep-4"C km-'. To put the contrast in terms of limits, we may ask, "How steep must a lineary temprature gradient be to just permit sensory detection under the model?" The answer is given by

k threshold swimming speed

5.

353

BODY TEMPERATURE RELATIONS OF TUNAS

where "threshold" is the minimum disequilibration ( Tb - T,) that is perceptible. (The above formula follows from integration of the equation in Section VII,B,2 after substituting [T,(O) gradient speed . t ] for T,) If we insert in the formula the following tabulated values, the answers to the question raised above are ?O.O66"C km-I for the tuna and +0.264"C km-' for the nontuna.

+

-

~~

k("C min-I "C-1) Speed (m min-I) Threshold ("C)

2-kg Tuna

2-kg Nontuna

0.033 50.0 0.1

0.066 25.0 0.1

We hasten to add that Neil1 et al. labeled their model for gradient " perception speculative." The basic hypothesis remains untested, to the best of our knowledge.

E. Retia Function to Retard Mass Transfer Rather than Heat Transfer We have no evidence that the exchanger exchanges anything but heat. However, lack of evidence does not preclude this possibility. If, for example, lactic acid passes across the exchanger vessels, lactate would be contained within the muscle. This would reduce the potentially deleterious effects of lactic acid on other tissues after a bout of strenuous activity. In addition, lactate would pass directly from white muscle to red muscle without entering the general circulation. It has been suggested (Braeken, 1956, reviewed in Bilinsky, 1974) that red muscle in fish functions like mammalian liver in the metabolic sense. White muscle provides the energy for swimming and produces lactate as a by-product; red muscle also provides energy for swimming but in addition oxidizes the lactate to provide more substrate for white muscle.

IX. PHYSIOLOGICAL INSIGHTS INTO THE NATURAL HISTORY OF TUNAS Barkley et aZ. (1976) have recently extrapolated laboratory data on temperature and dissolved oxygen requirements of skipjack tuna to predict geographic distribution of this species in the tropical Pacific Ocean. The predictions are consistent with Rothchilds (1965) migration model for the eastern Pacific skipjack tuna population.

354

E. DON STEVENS AND WILLIAM H. NEILL

Barkley et al. suggest that only pre-adolescent skipjack tuna can continuously inhabit tropical surface waters. The habitat of adult skipjack in the tropics, they argue, is the upper thermocline and not the surface layer, as has generally been thought. Their physiologically based model contends that the normal habitat of skipjack tuna is limited to those waters bounded by (1) a lower threshold for incipient thermal stress which may vary with prior conditioning but which apparently lies between 15" and 18°C; (2) a lower threshold for incipient oxygen stress, at or slightly below 3.5 ml liter1 (5 ppm) of dissolved oxygen; and (3) a speculatively inferred upper threshold of incipient thermal stress, ranging from 30°C for the smallest skipjack tuna normally caught by fisherman to 20°C or less for the oldest and largest of these fish. While the lower temperature and oxygen limits of the habitat may be the same for skipjack tuna of all sizes, the upper temperature limit occurs deeper in the water column and is more restrictive for the larger fish. In broad areas of the eastern tropical Pacific, the critical oxygen isopleth comes near the surface, intersecting the isotherms that mark the upper temperature limits for skipjack larger than 4 kg. Therefore, skipjack are excluded from progressively larger fractions of the eastern tropical Pacific as they grow larger than 4 kg. For example, 9 k g skipjack in summer along 119"W long. find essentially no habitable water between 5" and 12"N lat., a distance of more than 700 km (400 nautical miles). What happens to these larger fish? They logically move westward, toward the central Pacific, where deep water is both cool and oxygen-rich. Such a migration pattern is consistent with Rothschild's ( 1965) hypothesis: Mature and maturing skipjack (40-65 cm long, 1.2-6 kg) leave the Mexican and Central American fisheries, migrate westward, then spawn in the equatorial Pacific south of Hawaii. In the central Pacific and in other areas where the upper temperature limit is subsurface, skipjack tuna probably adopt a cyclical pattern of depth distribution, alternately rising to the surface to feed and navigate, then sounding below the thermocline to cool (Neil1 et al., 1976).

X. CONCLUSION: A THERMOCENTRIC OVERVIEW OF TUNA EVOLUTION Tunas inhabit parts of the world-ocean that are devoid of solid structure, highly transparent, and nutritively dilute. Forage for 10-

5. BODY TEMPERATURE RELATIONS OF TUNAS

355

1000 kg carnivores is not only scant but also is very patchily distributed. These properties of the tropical epipelagic zone have resulted in two divergent strategies of adaptation among tuna-sized fishes. Some, the energy frugalists, survive by minimizing energy expenditures through a nearly planktonic existence. The ocean sunfish (Molidae) represents the extreme case for this group; large opah (Lamprididae), too, reasonably fit the model of energy thrift. At the other extreme are the energy speculators. These fishes are wide-ranging and capable of high-speed swimming. They operate to maximize energy gain by gambling large energy expenditures (activity) on the expectation of proportionately large energy returns. In this group are certain carangids, coryphaenids, billfishes (istiophorids and Xiphius),and, at the very extreme, the lamnid sharks and the tunas. Specialization in energy speculation has brought tunas to the ragged edge of piscine existence. By this, we mean that tunas very nearly have beaten the system that we think of as “fish.” In particular, tunas are not poikilotherms; at least one tuna, the bluefin, is in fact on the verge of homiothermy. Scombrids gave up poikilothermy when they invested in vascular devices that retard heat transfer between body core and environment. An effective countercurrent heat exchanger not only damped vacillations of core temperature but also permitted high metabolic rates to be expressed as warm-bodiedness in the advanced tunas. Warmbodiedness, by the reasoning in Section VIII, enabled even higher (near mammalian) rates of metabolism in support of activity levels surpassing those of any other teleost group. But, there was a potential flaw in the tunas’ strategy for maximizing swimming speed: Any device that gets warmer as it operates faster and operates faster as it gets warmer is susceptible to the positivefeedback loop called by electrical engineers, “thermal runaway.” Tunas have obviated the potential for thermal runaway by opening the loop in at least two different ways. The skipjack group has evolved temperature-independent metabolism; that is, they do not operate faster as they get warmer. The bluefin tuna (and probably other Thunnus species) has adopted an alternative strategy. This largest of tunas, in contrast to the skipjack, has developed circulatory and metabolic arrangements that permit cooling during periods of maximum activity. Because the bluefin does not necessarily get warmer as it operates faster, it has no need for temperature-independent metabolic machinery. And, in fact, the yellowfin tuna, a congener of the bluefin, exhibits the usual elevation of metabolism with increase in temperature.

356

E. DON STEVENS AND WILLIAM H. NEILL

On the whole, it would seem that the bluefin’s solution to the temperature-activity dilemma is superior to that of the skipjack group. Although activity-induced increases in body temperature do not feedback into increased activity in skipjack, large skipjack still risk overheating in warm waters when they must be highly active to catch prey and avoid predators-a factor that may progressively restrict both vertical and geographic distribution of skipjack as they grow. Risk of overheating and the attendant restriction of thermal niche have been compromised through evolutionary limits on maximum body size; none of the skipjack group exceeds 22 kg in maximum weight. In contrast, four Thunnus species have maximum weights near or exceeding 200 kg (Table I). Large mass and correspondingly large thermal inertia have enabled the bluefin group to exploit much cooler waters than the skipjack group; at the same time, the bluefin’s capacity to be active without overheating allows it to live also in waters as warm as those inhabited by the skipjack group. Moreover, the bluefin’s solution to the temperature-activity problem has brought fishes to the verge of homiothermy. ACKNOWLEDGMENTS We thank Drs. Jeff Graham and Frank Carey for extensive and useful comments on the manuscript. Dr. Carey was especially helpful in providing us with raw data and details of how it was collected. We also thank Dr. Andy Dizon for supplying details of unpublished work from the National Marine Fisheries Laboratory at Honolulu. REFERENCES Barkley, R. A., Neill, W. H., and Gooding, R. M. (1976). Hypothetical habitat of skipjack tuna based on temperature and oxygen requirements. Nat. Mar. Fish. Sew., Honolulu, Hawaii. Barrett, I., and Connor, A. R. (1962). Blood lactate in yellowfin tuna and skipjack tuna following capture and tagging. Inter-Am. Trop. Tuna Comm., Bull. 6,231-280. Barrett, I., and Connor, A. R. (1964). Muscle glycogen and blood lactate in yellowfin tuna and skipjack following capture and tagging. Inter-Am. Trop. Tuna Comm., Bull. 9, 219-252. Barrett, I., and Hester, F. (1964). Body temperatures of yellowfin and skipjack tuna in relation to sea surface temperatures. Nature (London)203, 9 6 9 7 . Berg, L. S. (1955). Classification of fishes and fishlike animals, both recent and fossil (2nd Ed.). Tr. 2001.Inst., Akad. Nauk S S S R 20, 286 pp. Bilinsky, E. (1974). Biochemical aspects of fish swimming. In “Biochemical and Biophysical Aspects in Marine Biology” (D. C. Malins and J. R. Sargent, eds.), Vol. 1, pp. 239-288. Academic Press, New York. Brett, J. R. (1972). The metabolic demand for oxygen in fish, particularly salmonids, and a comparison with other vertebrates. Respir. Physiol. 14, 151-170.

5. BODY TEMPERATURE RELATIONS OF TUNAS

357

Brill, R. W., Guernsey, D. L., Dizon, A. E., and Stevens, E. D. (1978). Temperature profiles and heat loss in skipjack tuna. Nat. Mar. Fish. Serv., Honolulu, Hawaii. Brown, C. E., and Muir, B. S. (1970). Analysis of ram ventilation of fish gills with application to skipjack tuna.]. Fish. Res. Board Can. 27, 1637-1652. Burne, R. H. (1923). Some peculiarities of the blood vascular system of the porbeagle shark, Lamna cornubica. Philos. Trans. R . Soc. London Ser. B 212, 209-257. Butler, M. (1971). Biological investigation on aspects of the life history of bluefin tuna. Newfoundland Labrador Tourist Dev. Off., St. John’s. Carey, F. G. (1973).Fishes with warm bodies. Sci. Am. 228(2), 3 6 4 4 . Carey, F. G., and Lawson, K. D. (1973). Temperature regulation in free-swimming bluefin tuna. Comp. Biochem. Physiol. A 44, 375-392. Carey, F. G., and Teal, J. M. (1966). Heat conservation in tuna fish muscle. Proc. Natl. Acad. Sci. U.S.A. 56, 1461-1469. Carey, F. G., and Teal, J. M. (1969a). Mako and porbeagle: Warm-bodied sharks. Comp. Biochem. Physiol. 28, 199-204. Carey, F. G., and Teal, J. M. (1969b). Regulation of body temperature by bluefin tuna. Comp. Biochem. Physiol. 28, 205-213. Carey, F. G., Teal, J. M., Kanwisher, J. W., Lawson, K. D., and Beckett, J. S. (1971). Warm-bodied fish. Am. Zool. 11, 135-143. Chang, R. K. C., Ito, B. M., and Neill, W. H. (1976). Temperature independence of metabolism and activity in skipjack tuna. Natl. Mar. Fish. Serv., Honolulu, Hawaii. Charm, S. E., and Moody, P. (1966). Bound water in haddock muscle. ASHRAE J . 8, 39-42. Collette, B. B., and Chao, L. N. (1975). Systematics and morphology of the bonitos (Sarda)and their relatives (Scombridae, Sardini). US.Fish. Wildl. Seru., Fish. Bull. 73, 516625. Davy, J. (1835). On the temperature of some fishes of the genus Thunnus. P+oc. R . SOC. London 3,327-328. Dizon, A. E., Magnuson, J. J., and Neill, W. H. (1976). Swimming speed as a function of temperature in three species of tunas. Natl. Mar. Fish. Serv., Honolulu, Hawaii. Eschricht, D. F., and Muller, J. (1835).Uber die arteriosen and venosen Wundernetz an der leber und einen merkwiiidigen bau dieses Organes beim thunfische. Abh. Dtsch. Akad. Wiss. Berlin pp. 1-30. Fierstine, H. L., and Walters, V. (1968). Studies in locomotion and anatomy of scombrid fishes. Mem. South. Caltf. Acad. Sci. 6, 1-31. Fraser-Brunner, A. (1950). The fishea of the family Scombridae. Ann. Nat. Hist. Ser. 12 3, 131-163, Fry, F. E. J. (1971). The effect of environmental factors on the physiology of fish. In “Fish Physiology” (W. S. Hoar and D. J. Randall, eds.), Vol. 6, pp. 1-98. Academic Press, New York. Fry, F. E . J., and Brett, J. R. (1974). Oxygen consumption: Fishes. In “Biological Data Book” (P. L. Altman and D. S. Dittmer, eds.), 2nd Ed., Vol. 3, pp. 1625-1630. Fed. Am. SOC.Exp. Biol., Bethesda, Maryland. Gibbs, R. H., and Collette, B. B. (1967). Comparative anatomy and systematics of the tunas, genus Thunnus. U.S. Fish. Wildl. Seru., Fish. Bull. 66, 65-130. Goadby, P. (1972).“Big Fish and Blue Water,” Holt, New York. Godsil, H. C. (1954). A descriptive study of certain tunalike fishes. CaliJ Dep. Fish Game Bull. No. 97. Godsil, H. C., and Byers, R. D. (1944).A systematic study of the Pacific tunas. Calif. Dep. Fish Game Bull. No. 60. Godsil, H. C., and Holmberg, E. K. (1950). A comparison of the bluefin tunas, genus

358

E. DON STEVENS AND WILLIAM H. NEILL

Thunnus, from New England, Australia, and California. calif. Dep. Fish Came Bull. No. 77. Gordon, M. S. (1968).Oxygen consumption of red and white muscles from tuna fishes. Science 159, 87-90. Graham, J. B. (1973).Heat exchange in the black skipjack, and the blood-gas relationship of warm-bodied fishes. Proc. Natl. Acad. Sci. U.S.A. 70, 1964-1967. Graham, J . B. (1975).Heat exchange in the yellowfin tuna and skipjack tuna, and the adaptive significance of elevated body temperatures in Scombrid fishes. U.S. Fish. Wildl. Sew., Fish. Bull. 73, 219-229. Hazel, J. R., and Prosser, C. L. (1974).Molecalar mechanisms of temperature adaptation in poikilotherms. Physid. Rev. 54, 620-677. Hill, A. V. (1951).The influence of temperature on the tension developed in an isometric twitch. Proc. R. Soc., Ser. B 138,349-354. Kishinouye, K. (1915).A study of the mackerels cybiids, and tunas. Suisan Gakkwai H O 1,l-24.(In Japanese, trans]. by W. G. Van Campen, U S . Fish Wildl. Seru., Spec. Sci. Rep.-Fish. No. 24.) Kishinouye, K. (1917).A new order of teleostomi. S U ~ S QGakkwai ~ H O 2, 1-4. (In Japanese, trans]. by W. G. Van Campen, US.Fish Wildl. Seru., Spec. Sci. Rep.Fish. No. 50,pp. 1-3.) Kishinouye, K. ( 1923).Contributions to the comparative study of the so-called scombrid fishes. J. Coll. Agric., Imp. Uniu. Tokyo 8, 293-475. Lindsey, C. C. (1968).Temperatures of red and white muscle in recently caught marlin and other large tropical fish. 1. Fish Res. Board Can. 25, 1987-1992. Linthicum, D.S.,and Carey, F. G. (1972).Regulation of brain and eye temperatures by the bluefin tuna. Comp. Biochem., Physiol. A 43, 425-433. Magnuson, J. J. (1963).Tuna behavior and physiology, a review. FA0 Fish. Rep. 6(3), 1057-1066. Malessa, P. (1969).Beitrage zur Temperaturadaptation des Aales. Mar. Biol. 3,143-158. Morrow, J. E. (1957).Shore and pelagic fishes from Peru, with new records and description of a new species of Sphoeroides. Bull. Bingham Oceanogr. Collect. 16, 5-55. Nakumura, I. (1965).Relationships of fishes referable to the subfamily Thunninae on the basis of the axial skeleton. Bull. Misaki Mar. Biol. lnst., Kyoto Uniu. 8, 7-38. Neill, W.H., and Stevens, E. D. (1974).Thermal inertia versus thermoregulation in “warm” turtles and tunas. Science 184, 1008-1010. Neill, W. H., Chang, R. K. C., and Dizon, A. E. (1976).Magnitude and ecological implications of thermal inertia in skipjack tuna. Enoiron. Biol. Fish. 1, 61-80. Rayner, M. D., and Keenan, M. J. (1967).Role of red and white muscles in the swimming of the skipjack tuna. Nature (London)214,392-393. Regan, C. T. (1909).On the anatomy and classification of the scombroid fishes. Ann. Mag. Nat. Hist. Ser. 8 3, 66-75. Rothchild, B. J. (1965).Hypothesis on the origin of exploited skipjack tuna in the eastern and central Pacific Ocean. U.S. Fish Wildl. Sew., Spec. Sci. Rep.-Fish. No. 512. Slavin, J. W.(1964).Freezing seafood-now and in the future. ASHRAE J. 6,43-48. Stark, E.C. (1910).The osteology and mutual relationships ofthe fishes belonging to the family Scombridae. J . Morphol. 21, 77-99. Steffel, S., Dizon, A. E., Magnuson, J. J., and Neill, W. H. (1978).Temperature discrimination by captive free-swimming tuna, Euthynnus afinis. Trans. Am. Fish. Soc. 105,~8-591. Stevens, E. D. (1972).Some aspects of gas exchange in tuna. J. E x p . B i d . 56,809-823. Stevens, E.D.,and Black E. C. (1966).The effect of intermittent exercise on carbohydrate metabolism in rainbow trout. J . Fish. Res. Board Can. 23,471-485.

5. BODY TEMPERATURE RELATIONS OF TUNAS

359

Stevens, E. D., and Fry, F.E. J. (1971).Brain and muscle temperatures in ocean-caught and captive skipjack tuna. Comp. Biochem. Physiol. A 38, 203-211. Stevens, E. D., and Fry, F. E. J. (1972).The effect of changes in ambient temperature on spontaneous activity in skipjack tuna. Comp. Biochem. Physiol. A 42, 803-805. Stevens, E.D.,and Fry, F. E. J. (1974).Heat transfer and body temperatures in nonthermoregulatory teleosts. Can. J. Zool. 52, 1137-1143. Stevens, E.D.,Lam, H. M., and Kendall, J. (1974). Vascular anatomy of the countercurrent heat exchanger of skipjack tuna.J. E x p . B i d . 61, 145-153. Takahashi, N.(1926).On the Plecostei, an order of the Teleostomi, established by Prof. Kishinouye.]. Coll. Agric., Imp. Unio. Tokoyo 7 , 383-398. Uda, M. (1941).The body temperature and bodily features of “Katuo” and “Sanma.” Nippon Suisan Gakkaishi 9,231-236.(In Japanese, transl. by W. G. Van Campen, U.S. Fish Wildl. Seru., Spec. Sci. Rep.-Fish. No. 51.) Walters, V., and Fierstine, H. L. (1964).Measurement of swimming speeds of yellowfin tuna and wahoo. Nature (London)202,208-209. Wardle, C . S. (1975).Limit of fish swimming speed. Nature (London) 255,725-727. Webb, P. W. (1975).Hydrodynamics and energetics of fish propulsion. Bull., Fish. Res. Board Can. 190, 1-159. Yuen, H. S. H. (1959).Variability of skipjack response to live bait. U.S. Fish. Wildl. S m . , Fish. Bull. 60, 147-160. Yuen, H. S. H. (1966).Swimming speeds of yellowfin and skipjack tuna. Trans. Am. Fish. SOC. 95, 203-209.

This Page Intentionally Left Blank

6 LOCOMOTOR MUSCLE QUENTlN BONE I. Introduction .................................................... 11. The Organization of the Myotomes .............................. 111. Fin M u s cl es . . .................................................. IV. Fiber Types .................................................... A. General Considerations ...................................... B. Histology ................................................... C. Ultrastructure and Histochemistry in Different Fish Groups ................................................ D . Ontogeny of Fiber Types .................................... E. Innervation ................................................. F. Electrical and Mechanical Properties ......................... G. Functional Role of Different Fiber Types .....................

.......................

VI. Fish Muscle and the Muscles of Higher Forms .................... References

.........................................................

36 1 363 368 368 368 373 377 390 393 397 405 410 416 417

I. INTRODUCTION Muscular tissue forms a larger part of the mass of the fish body than it does of other vertebrates: Some 40-60% of the total body mass in most fish is locomotor musculature (Table I). In part this is because economy in weight is not mandatory as it is for terrestrial and aerial forms, and in part because stringent demands are placed on the locomotor system by the density of the medium, so that a large amount of muscle is needed to generate sufficient power for rapid swimming. In addition to sheer mass (which of course gives fish their culinary and economic importance), there are Several design features of the muscular system that are not commonly found in the muscles of other classes of vertebrates. For example, in many fish, twitch fibers are multiply innervated; again, no fish muscles contain muscle spindles: Such features give fish muscle an especial comparative interest. It must be emphasized at the outset, however, that despite the efforts of a 361 FISH PHYSIOLOGY, VOL VII Copyright @ 1978 by Academic Preu, In' A l l light* of reproduction in m y form re\erved ISBN 0-12 350407 4

362

QUENTIN BONE

Table I The Amount of Locomotor Muscle as a Proportion of Total Body Weight in Different Fish Species

% Muscle

Comments.

Authority

Galeocerdo arcticus Prionace glauca Scyliorhinus canicula Centroscymnus owstoni Katsuwonus pelamis Salmo irideus Carassius auratus Leuciscus leuciscus

37 36 45 55-60 68 55-67 33-45 42-60

Incl. vertebrae

Warfel and Clague (1950) Stevens (1976) Bone and Roberts (1969) Higashi et al. (1953) Fierstine and Walters (1968)

Incl. skin

Bainbridge (1960, 1962)

number of investigators (beginning with Lorenzini, 1678, who first described different muscle fiber types), knowledge of fish muscle is as yet relatively scanty. This chapter can thus only serve as a progress report, rather than as a definitive account of the muscular systems employed in locomotion. In several important respects, fish muscle is well-suited for physiological study (despite the disadvantages inherent in the myotomal muscle preparation), and certain puzzling features of vertebrate muscle fibers are perhaps most likely to be understood by work on fish. The great majority of fishes swim using the segmental myotomal musculature, so most of this chapter will deal with the arrangement of the myotomes and of the muscle fibers within them. A few fish (e.g., some Trachyurids, many labrids, rays, and holocephali) swim by using the musculature of the paired or unpaired fins: These muscles will be considered in less detail. The nomenclature of fish muscles has been bedeviled by synonymy, but fortunately Winterbottom (1974) has provided a helpful guide to the nomenclature of the locomotor (and other) muscles. In limiting this chapter only to locomotor muscles, it is important to observe that although this may yield a more uniform view of sets of fibers designed for similar ends in different forms, interesting work on other kinds of fish muscle fibers has to be omitted. For example, fish eye muscles have provided valuable material for experimental study of selective re-innervation (e.g., Mark and Marotte, 1972); the very rapid sonic muscles of the swimbladder in some forms have proven interesting (Skoglund, 1961); the gill muscles may yield useful mechanical preparations (Levin and Wyman, 1927);the muscular systems of fish barbels offer peculiar problems of coordination and control.

6.

LOCOMOTOR MUSCLE

363

Fascinating as these muscles all are, they fall outside the scope of this chapter.

11. THE ORGANIZATION OF THE MYOTOMES

The lateral musculature in all fish groups is subdivided into serially arranged myotomes of complex shape, delimited by connective tissue myosepta into which the myotomal muscle fibers insert. Figure 1 shows something of myotomal shape in different fish groups. There is plainly a phylogenetic increase in complexity of shape, from the simple V-shape in amphioxus via the shallow W-shape of the Agnatha, to the deep W-shape of gnathostomes. This phylogenetic change is reflected in the ontogeny of the higher fish groups where the somite first forms a V-shape, before folding further to yield the adult W-shape.

Fig. 1. Simplified diagrams of myotomal shape in various fish groups. An enlarged view of a midtrunk myotome in each is shown at right; a teleostean caudal myotome at bottom. Not to scale. (Redrawn from Nursall, 1956, Proc. Zool. SOC.London 126, 127143.)

364

QUENTIN BONE

A number of workers have attempted to provide a functional basis for the complex shape of the fish myotome, and for its internal arrangement. These attempts have been more successful in explaining the orientation of muscle fibers within the myotomes than the shape of the myotomes themselves; yet both presumably reflect the properties of the contractile units of the system and the properties of their insertions. The important properties insofar as myotomal design is concerned would seem to be the following. 1. Contraction takes place without volume change. 2. The muscle fibers insert onto deformable but inextensible partitions which are attached to the flexible but incompressible notochord or vertebral column. 3. Deformation of the myosepta is often (but not always) restrained b y intermuscular bones. 4. Flexion of the body is required only in the lateral plane. 5. During flexion, the radius of curvature will be least next to the vertical column, largest just under the skin. 6. Both frequency and amplitude of flexions may vary. The end result of the operation of the myotomal units with these properties is, of course, the lateral oscillations of the body brought about by transferring the contraction forces to the central strut. In the simple myotomes of amphioxus, the muscle lamellae run nearly parallel to the long axis, and the distance between inner and outer surface of the myotome in these animals is small. A single muscle lamella may span the whole of this distance (see Fig. 8), but the difference between the radii of curvature of its inner and outer edges is insignificant. The V-folding of the myotome is probably related to the need to avoid dorsoventral flexion. The notochord lies dorsally to accommodate the viscera below it, and if the myotomes were simple inclined blocks (the greater part of which lay below the strut), contraction would lead to ventral flexions of head and tail. With the V-shaped myotome arranged so that the arms of the V are unequal in length, and that the apex lies at the level of the notochord, solely lateral flexions are possible. It is significant that with increase in scale, all other fishes have their myotomes arranged so that the muscle fibers of the greater part of the myotome do not run parallel to the long axis. Indeed, as van der Stelt (1968) and Alexander (1969) point out, the muscle fibers of the white or fast part of the myotome may make large angles with the long axis. Alexander (1969) found that some muscle fibers in the myotomes of both sharks and teleosts were oriented at nearly 40"to the long axis. These orientations are not random; Fig. 2 shows the two patterns of

6.

LOCOMOTOR. MUSCLE

I

365

(a)

Fig. 2. Thick transverse slices across (a) selachian, and (b) teleostean body showing orientation of myotomal muscle fibers as viewed from behind. ac, anterior cone; pc, posterior cone; t, tendon. (From Alexander, 1969,J. Mar. Biol. Assoc. U . K . 49,.263-290.)

orientation found b y Alexander in sharks and in higher teleosts. Alexander was able to show that these two patterns of orientation of the fast or white fibers of the myotome were a consequence of the requirement for all of these fibers to contract at about the same rate, whatever their position in the myotome. In other words, although during flexion the radius of curvature of the fish as a whole will b e greatest next to the vertical column, and least superficially, suitable orientation of the fibers in these positions in the myotomes will enable each to contract to the same extent as the body flexes. The importance of this result is that, as Hill (1950) has pointed out, the shape of the forcehelocity curve for muscle fibers means that maximum power will be produced at a particular rate of contraction. We do not know the shape of the forcehelocity curve for the myotomal muscles of any fish, but we can assume that for most fishes, the white or fast portion of the myotome will b e designed for maximum power (see Section IV). The orientation of the muscle fibers directly reflects this requirement, for if they were not able to contract at about the same rate throughout the myotome, the power extractable from the fast portion of the myotome would be much lessened. Note that this argument does not apply to the superfi-

366

QUENTIN BONE

cia1 sheet of slow (red) muscle fibers, for it is thin enough that the fibers which make it up can run more or less parallel to the long axis and yet all contract at about the same rate. It is interesting that sharks and higher teleosts adopt different fiber orientations in the rnyotomes. Alexander has shown that the helical teleost arrangement (Fig. 3) will give a faster rate of body flexion for the same rate of contraction of the muscle fibers than will the selachian arrangement (at the expense of a weaker bending moment), and suggests that this is a better compromise for the teleosts since they swim faster than sharks of similar size. It is difficult to know whether this proposition is always applicable, since there are no measurements of the speed of Isurids; it would certainly be very interesting to know if the selachian arrangement found b y Alexander in Hexanchus, Scyliorhinus, and Squalus also obtains in Carcharodon or Zsurus. In the caudal peduncles of higher teleosts, there is a transition to the selachian pattern of fiber orientation, and Alexander suggests that this is to enable the fish to cope with the larger stresses in this region while maintaining a similar rate of contraction of muscle fibers throughout the body. Alexander’s analysis was notable, for it gave the first convincing functional explanation of the fiber orientation of the myotomes. What is more, it seems likely that this fiber orientation underlies the complex

Fig. 3. Schematic dorsal and lateral views of typical teleost showing course of myotomal muscle fibers in successive myotomes along the body. The helices shown were found by taking the origin of one muscle fiber from the point at which the muscle fiber in the myotome next anterior inserts onto the common myoseptum, and so on along the fish. (From Alexander, l969,J. Mur. B i d . Assoc. U .K. 49, 263-290.)

6.

LOCOMOTOR MUSCLE

367

shape of the gnathostome myotomes. The myotomal folding is so arranged that the muscle fibers insert into the myosepta at approximately the same angle, despite their very different orientations with respect to the long axis of the body. It would evidently be interesting to examine fiber orientations during ontogeny as the myotomes gradually achieve the W-shape, to provide a test of this view. The larger scombroids have the most specialized teleost myotomes. In these fish, the inertia of the caudal peduncle and foil has been reduced as much as possible (just as in the legs of fast-running terrestrial vertebrates) and the foil is oscillated by tendons passing over a flexible joint in the vertebral column (Fierstine and Walters, 1968). These tendons are formed by a series of nested cones derived from successive myosepta (an analogous arrangement is found in the caudal region of certain rays), similar smaller tendons also run anteriorly to insert upon the pectoral girdle. The result of this arrangement is that these fish have achieved the hydrodynamically desirable separation of the power source from the caudal propellor. Body flexions are less than those of fish swimming in less efficient and less rapid ways (see Chapter 3), and since the muscles used during cruising form the inner portion of the myotome (Fig. 4), it is possible that these contain muscle fibers oriented parallel to the long axis of the body. Fiber orientations have not yet been studied in the larger scombroids, nor in the similarly adapted Isurid sharks.

4a D

Fig. 4. Diagrammatic transverse sections across the body of various teleosts and elasmobranchs to show disposition of slow (red) muscle fibers in the myotomes (shaded).A, Katsuwonus; B, Rhina; C, Squalus; D, Atherina; E, Euthynnus; F, Alosa. A, redrawn from Rayner and Keenan (1967); remainder original.

368

QUENTIN BONE

Little work has been done upon the myosepta themselves. Some of the problems of their design have been considered by Willemse (1975), who describes the organization of the connective tissue fibers in teleost myosepta. In many teleosts intermuscular bones of different forms lie in the myocommata, limiting the directions in which they may b e deformed. Jarman (1961) pointed out that these bones could be arranged so that they did not interfere with contraction of the myotomal muscle fibers, but a functional analysis of the different patterns and numbers of these elements in different groups is yet to be made. They are absent from sharks.

111. FIN MUSCLES The arrangement of the muscles of paired and unpaired fins is illustrated diagrammatically for many groups by Winterbottom (1974). In most, small and large fiber portions of the fin musculature are anatomically similar in arrangement, the small-diameter fibers lying superficially to the larger-diameter fibers. In rays, the pectoral fin muscles are divided into superficial small bundles of small-diameter fibers, which arise on the pectoral girdle, and pass along the fin rays, to insert about halfway across the “wing” of the ray. These muscle fibers are, therefore, very long. By contrast, the larger-diameter fibers of the main portion of the fin ray musculature are arranged obliquely to the axis of the fin ray, inserting on the fin ray, and upon a connective tissue sheet underlying the superficial slow fibers. In this way, these fibers are virtually pinnate in arrangement. Calow and Alexander (1973) have shown how more power can be extracted from a pinnate fiber organization. In those teleosts where the fins are used for locomotion, e.g., Cymatogaster (Webb, 1975), or Gasteropelecus, the fibers are not pinnate, so far as is known. The arrangement in Holocephali is probably similar to that of the rays (see Kryvi and Totland, 1978).

IV. FIBER TYPES

A. General Considerations The muscle fibers of the myotomes or fins of few fishes have been examined histologically, yet fewer physiologically, but the different muscle fiber types found are remarkably similar in those fish where they have been studied. Although there are great differences in gen-

6.

LOCOMOTOR MUSCLE

369

era1 morphology between, say, acraniates and dipnoi, their locomotor muscle fibers are arranged in fundamentally the same way. It is reasonable therefore to generalize from the few cases which have been examined. At first sight this may seem surprising, but it is a consequence of the density of the medium in which fish swim. The greater part of the drag opposing forward motion in most fishes is skin friction drag (see Chapter 3), approximately proportional to Vz, so that the thrust required at constant speed will also be proportional to Vz, the power required from the locomotor muscles proportional to V3. There are complications of detail in the power dependence of the velocity (owing to uncertainties about the forcelvelocity curves of fish muscle and the nature of the boundary layer, see Webb, 1975; Bone, 1975) but this is an order of magnitude argument: Increased speed in water requires a great increase in power output. It is common observation that in individual species speed ranges of at least four times, and sometimes u p to twenty times are found; such performances demand a wide range of power from the locomotor system, and have led to a highly specialized arrangement of the locomotor muscle fibers. Faced with the conflicting demands of low speed cruise economy, and short bursts of maximum speed, all fishes have devised the same solution, and have divided the locomotor musculature into two very different parts, each specialized for one of these functions. Naturally enough, as we should expect, the muscle fibers composing each of the two conttasting parts are entirely different in design, differentiated by a whole spectrum of histological and ultrastructural features, as well as by biochemical and physiological criteria. For example, the “cruising” muscle fibers have a high content of mitochondria and high oxidative enzyme activity, as compared with the fiber type used during burst locomotion. The remainder of this chapter will mainly b e concerned with considering these differences, which are summarized in Table I1 and Fig. 5. Most fish use the same locomotor organs (caudal region and caudal fin, or paired pectoral fins) whether they are cruising slowly or swim-. ming rapidly for short bursts, and the two different parts of the locomotor system are therefore adjacent to each other in the myotomes or in the fin muscles. In rays, for example, the paired fins are employed during slow and fast swimming. In Holocephali, however, and such teleosts as labrids, many carangids, notopterids, young stromateoids, and alepocephalids, the paired fins only are used during slow swimming, and the myotomal musculature only during bursts of rapid swimming so that the two main muscle fiber types are physically divorced from each other.

370

QUENTIN BONE Table I1

A Comparison of Fast and Slow Muscle Fibers in Fish Slow Smaller diameter (20-50% of fast) Well vascularized Usually abundant myoglobin, red color Abundant large mitochondria Oxidative enzyme systems Low activity Ca*+-activatedmyofibrillar ATPase Little low molecular weight protein Stored lipid and glycogen Myosatellite cells abundant Sarcotubular system usually less in volume than in fast fibers Z-lines broader than fast fibers in some cases Distributed cholinergic innervation Subjunctional folds usually absent Lower resting potentials than fast fibers N o propagated muscle action potentials, except under experimental conditions Long-lasting contractions evoked by depolarizing agents

Fast Larger diameter (may be more than 300 Pm) Poorly vascularized No myoglobin, usually white Few smaller mitochondria with fewer cristae Enzymes of anaerobic glycolysis High activity of enzyme Rich in low molecular weight protein Glycogen stored, usually little lipid Fewer myosatellite cells Relatively larger sarcotubular system Z-lines usually thinner than in slow fibers Focal or distributed cholinergic innervation Subjunctional folds usually present Higher resting potentials Propagated muscle action potentials usual; may not always occur during activity of multiply innervated fibers Brief contractions evoked by depolarizing agents

In a crude way, the two main fiber types can be recognized macroscopically, even where they are both found together, because in most fish, fiber types are segregated so that a region of fibers rich in myoglobin with a large vascular bed appears as a red or pink zone as compared with the adjacent pale zone of fibers of contrasting type. The vascular bed of teleost red and white muscle has been described by Mosse (1978) and by Bone (in press). The striking difference between the vascularization of the two main fiber types can be seen by the tenfold difference in the capillary to muscle fiber ratio in favor of the red fibers. In the myotomal musculature of many fish, a superficial red layer of fibers covers the main mass of myotomal white fibers (Fig. 4).The proportions of these two fiber types differ in different species (Table 111). There are some difficulties in comparing data given by different workers, for example, Greer-Walker and Pull (1975) examined the proportion of red and white fibers at a given level in different

6.

371

LOCOMOTOR MUSCLE

Fig. 5. Schematic comparison between slow fiber (above) and fast fiber. The fast fibers of some teleosts (see text) are multiply innervated. On the left, a transverse section showing comparative capillary and mitochondria1 density.

fish, whereas other workers have calculated the total amounts of the two types of fiber within the myotomes; moreover the proportions of the two fiber types may change with fish size (Magnuson, 1973). In 84 species examined by Greer-Walker and Pull (1975),red muscle fibers never constituted more than a quarter of the total myotomal Table 111 Relative Amounts of Slow and Fast Myotomal Muscle Fibers in Different Fish" ~~~

Species

Percentage of slow fibers in caudal region

Percentage of slow fibers overall

Scyliwhinus canicula Prionace glauca Scomber colias Gadus oirens Gadus pollachius Chimaera monstrosab Capros ape+ Raia spp!

18 22 30 11 3 0.6 0.5 0.0

8 10-11

" Data for teleosts and Chimaera

from Greer-Walker and Pull (1975). The last three fish swim slowly by means of the paired fins.

372

QUENTIN BONE

musculature, and in most less than 10%.If only from consideration of the power output required from the musculature at different speeds, this distribution would suggest that the red fibers were employed during cruising, the much larger mass of white fibers during bursts of speed (see Section V). In few fish is this procrustean division of fiber types into red and white an absolute one, and although it is convenient to begin by considering the locomotor apparatus as composed of two main fiber types, it is important to avoid the temptation of falling into what Austin (1962) called the deeply ingrained worship of tidy-looking dichotomies. In most fish there are more than two fiber types and, apparently, some overlap in function between these different types. It is for these reasons that the original simple division of myotomal muscle fibers by their color [as, for example, by Lorenzini (1678) or by Arloing and Lavocat (1875)l is probably best abandoned. Nevertheless, the proportion of visibly red muscle in the locomotor apparatus gives some kind of an indication of the habits of the fish, and various workers have commented on this relationship (Boddeke et al., 1959; Bone, 1966; Greer-Walker and Pull, 1975); color is still a useful macroscopic guide to the fiber types which are found. This is, naturally, of particular importance in biochemical or electromyographic investigations. It is difficult to decide on an acceptable alternative to color as a description of the two main fiber types that does not have unsupported biochemical or physiological implications; but on the whole, it seems most appropriate to contrast slow and fast fibers. It is admittedly unsatisfactory to categorize muscle fibers by their speed when in very few instances only has this been determined, but other suggested terms, such as anaerobidaerobic, or twitchhon-twitch are still less satisfactory. In higher vertebrates, for example, in mammals (e.g., Burke et al., 1971), or in Anura (Smith et al., 1974), good correlations have been established between muscle fiber histochemistry, morphology, and speed of contraction; the method of glycogen depletion has proven invaluable in distinguishing the scattered fibers belonging to a single motor unit. The situation is very different in fish, for although morphologically and histochemically different fiber types are arranged in discrete zones or layers so that they can be distinguished by their position, the inconvenient physiological “preparation” of the myotome has not been investigated in any detail, so that no correlations have been established between fiber types and contraction velocity. Apart from crude experiments such as those of Ranvier (1873), Barets (1961),and Bone (1966) differentiating the contractions of slow and fast muscle fibers, nothing has been done in this regard for

6.

LOCOMOTOR MUSCLE

373

locomotor muscle fibers, apart from the elegant work of Tefavainen and Rovainen (1971) on lamprey slow and fast myotomal units. The significance of the various morphologically different slow or fast fiber types is therefore not clear, and the problem is complicated by the continuous growth of fish, for this may mean that some fiber types should b e regarded simply as stages in the development of others, rather than as types of especial contraction velocity and function. Thus Barker (1968) suggests that certain elasmobranch fibers are “slowtwitch” fibers, instead of “immature” twitch fibers (Bone, 1966), by analogv with the fiber types of higher forms, but this analogy may be invalid. Davies (1972) has shown that in mammals substantial changes take place in fiber types (based on histochemical criteria) as the muscle fiber population adapts to increase in body weight during growth; it is not yet known whether similar changes take place in fish muscle fibers, for detailed developmental studies are not available for any fish group (see Section IV,D). However, there is biochemical evidence, for example, in Anguilla anguilla (Bostrom and Johansson, 1972), that there are significant changes in enzyme activity profiles during development and growth, and these are probably reflected in such morphological parameters as mitochondria1 content and vascularization. I n most fish the slow fibers form a superficial sheet (the Seitenlinie) covering the main mass of myotomal fast fibers. In rays, the slow fibers form’a superficial zone in the fin ray muscle bundles. Presumably in these peripheral positions they are at a better mechanical advantage. They are only found deep in the myotome in secondarily flattened fish such as Rhina, or in fish where the slow fibers are operated above ambient temperature (Carey and Teal, 1966) (Fig. 4). There is some evidence that superficial and “deep” slow fibers in tuna aye different in properties (see Section IV,F,5). It may be that this reflects the effects of different fiber contraction velocity on different positions within the myotome, and myotomal muscle fiber arrangement and physiology in tuna should repay further study (see Sharp and Dizon, in press).

B. Histology Myotomal muscle fibers are often very long (in large fish, several centimeters) and insert at both ends into connective tissue sheets; fingers of connective tissue push into the ends of the muscle fibers in a complex interdigitation. At these points, there are couplings between the inpocketed tubes containing collagen fibers and the sarcoplasmic reticulum. These terminal couplings (first observed in anuran and

374

QUENTIN BONE

lamprey muscle fibers b y Nakao, 1975)are also found in elasmobranch fibers, and probably in all fish fibers. Nakao suggested that the couplings represented sites of calcium transfer and were in some way related to growth in length of myofibrils, but there is no evidence for this function. It is perhaps more probable that the couplings represent the response of the sarcoplasmic reticulum to an ingrowing portion of the sarcolemma, analagous to the SR response to the T-system. The organization of the sarcoplasmic reticulum and T-system is similar in most fish groups, triads occurring at Z-line level (e.g., Patterson and Goldspink, 1972; Kryvi, 1977; Kryvi and Totland, 1978).Exceptions to this general arrangement are seen in the Agnatha and Acrania. Few quantitative studies of SR and T-systems in fish locomotor muscle have been carried out. Most of these (e.g., Hidaka and Toida, 1969; Nag, 1972; Korneliussen and Nicolaysen, 1975; Kryvi, 1977) agree with studies on other vertebrates where it is found that in slow fibers these systems concerned with activation are of lesser extent than in fast fibers. However, Patterson and Goldspink (1972) were unable to find significant differences between these systems in fast and slow myotomal fibers of Gadus virens, in agreement with Kilarski’s (1967) original observations. It is, perhaps, probable that in these cases where no obvious quantitative difference between the SR of slow and fast fibers can be discerned, the fibers may overlap in function, and the dichotomy of function may not be so clear as once assumed. For example, there is indirect evidence for Gadus virens that the “fast” fibers may be active at slow speeds (Greer-Walker and Pull, 1973). Morphological correlates of the metabolism of the different fiber types (manifested by their oxygen consumption, e.g., Gordon, 1968; Modigh and Tota, 1975) are seen in the different abundance, size, and cristal complexity of mitochondria (see Section IV,C), and also in the abundance of stored metabolites. Glycogen particles are found in all fiber types, usually as small units, but occasionally (e.g., in Scyliorhinus white fibers) in chain formations. Lipid is chiefly found in slow fibers, though in some species, such as Squalus, Cetorhinus, and Ruvettus, buoyancy lipid may be stored in fast fibers (Bone and Roberts, 1969; Bone, 197213). In “fatty” fish, such as herring or mackerel (Bone, in press), not only is lipid stored in the slow fibers, but there are fat cells among the muscle fibers (Fig. 6G). Metabolic lipids ofthis kind, as distinct from buoyancy lipid, are often found in close association with mitochondria as pointed out by Nishihara (1967) and Nag (1972). Ultrastructural investigations of fish storing wax esters in muscle fibers for buoyancy would be of some interest; it is possible that “metabolic” and “buoyancy” lipid stores may b e differently compartmented within the muscle fibers. During starvation (Greer-

Fig. 6. (A) Transverse section of the edges of a slow (S) and a fast (F) myotomal muscle fiber of amphioxus, showing the larger subsarcolemmal cisternae (representing the sarcoplasmic reticulum) in the fast fiber. (B) Multiply innervated myotomal fast fibers in Euthynnus. (C) Focal innervation of pectoral fin fast fibers in Scylioshinus. (D) Single e n d formation from multiply innervated slow fibers of pectoral fin of Scyliorhinus. (E) Multiply innervated fast fibers of pectoral fin of Periopthalrnus. (F) Rich vascular b e d of slow fibers from myotome ofEsor. (G) Lipid-filled cells adjoining capillaries in the fast myotomal musculature of Gempylus serpens; these cells probably contain wax esters to provide static lift. (H) Large-diameter axon passing to focally innervated fast fiber portion of pectoral fin muscle in Clupea. Note division of axon at arrows. ( I ) Terminal pattern of innervation i n the myotomal musculature of Alepocephalus. The myoseptum runs vertically at the right, muscle fibers pass from their insertions on the right obliquely to the left. (J) Focal innervation of fast fiber from pectoral fin ofRoia; several nerve processes pass to motor endplate. All except (A) from whole mounts of silver-impregnated material. Scale bars: 100 Frn except for (A) 0.5 pm.

376

QUENTIN BONE

Walker, 1971; Johnston and Goldspink, 1973a,b), the different fiber types are affected differently; glycogen and protein diminish in the fast fibers although lipid and protein are little affected in the slow fibers. The result of protein loss is to increase the water content of the fibers to a maximum around 90%. A fairly large group of mesopelagic teleosts belonging to different families have the myotomal musculature naturally very watery as part of a general reduction of dense components in the absence of a swimbladder (Denton and Marshall, 1958; Blaxter et al., 1971).There has been no systematic investigation of the locomotor muscles in such fish; the few that I have examined (alepocephalids, stromateoids, stomatoids) have achieved economy of the dense myofibrillar protein components by reducing muscle fiber diameter and increasing interfiber spacing, as compared with normally dense fish of the same size. Greer-Walker (1971)found that after 130 days starvation, white (fast) fibers of cod were reduced in diameter by 40%, red (slow) fibers by 15%. Ultrastructural studies by Johnston and Goldspink (1973b) on carp starved experimentally for 16 weeks showed that protein decline in the white fibers due to starvation chiefly represented loss of myofibrillar protein. In the red fibers, mitochondria degenerated and disappeared. There are then some similarities between fish where reduction of dense components is a consequence of starvation, and the mesopelagic fish where it is part of the normal life of the fish. It has been found that both slow and fast fibers increase significantly in diameter during exercise (Greer-Walker and Pull, 1973). Bainbridge (1962) found no appreciable change in total muscle mass (in trout) after 1 year of continuous swimming, but his measurements did not take account of the (small) amount of slow muscle involved in this exercise. Slow fibers are smaller in diameter than fast fibers [between 22 and 53% on a series of marine teleosts examined by Greer-Walker and Pull (1975)] and are usually rich in myoglobin. Wittenberg (1970) has given grounds for supposing that myoglobin is equally distributed throughout the slow fibers, despite earlier attempts to localize it histochemically in specific regions of the sarcomeres in formalin-fixed material. In some groups, for example, Anura (Smith and Ovalle, 1973), several distinct fiber types are recognizable by their myofibrillar patterning [recalling the crude but useful distinction made by Kruger (1950) between Feldenstruktur and Fibrillenstruktur fibers], and it is probable that closer investigations will show that this is also the case for fibers in different fish groups. For example, superficial and slow fibers in sharks are recognizable in this way (Fig. 7) but distinctions between the different slow fibers have not yet been quantified. Very rapid mus-

6. LOCOMOTOR’ MUSCLE

377

cles, such as those of the seahorse dorsal fin, are rich in sarcoplasm and have relatively few myofibrils (Bergman, 1964a). A similar paucity of myofilaments occurs in certain mesopelagic fishes where the dense myofilament proteins have been reduced for buoyancy requirements (see Section IV,E). The obverse of myofibrillar patterning is the mitochondrial distribution within the fibers. Here there is a rather wide range in size and abundance of mitochondria in different fiber types, and although quantitative studies are few, in some slow fibers mitochondria make up around 15-25% of the total cross-sectional area of the fiber (Patterson and Goldspink, 1972; Best and Bone, 1973).In Scomber, up to 45% of the slow fiber volume may be occupied by mitochondria (Bone, in press). In most fiber types, the distribution of mitochondria seen in cross section of the fiber is more or less uniform, but in some fibers there may b e a peripheral mitochondrial zone, with a central zone where mitochondria are absent or rare. In sharks, Kryvi (1977) found that subsarcolemmal mitochondria make up about 10% of the crosssectional area of the slow fibers. Accumulations of mitochondria are common under nerve terminals, particularly where these are not embedded within indentations of the sarcolemma (see Section IV,E). In some species, myelin figures derived from mitochondria are abundant in different fiber types, and occasional multilamellar bodies are also observed. The possible significance of these organelles is discussed by Kordylewski (1974) and Bone (in press). Individual fiber types in the different fish groups are considered in the next section. The reader is again warned to bear in mind that the general description of slow and fast fibers given above (summarized in Table I1 and Fig. 5) represent in most fish the different ends of a discontinuous spectrum of fiber types, so that in a given fish there may b e several morphologically distinct slow fibers, or two fast fiber types. The essential plasticity of the muscular system, well shown by the changes between silver and yellow eels (Bostrom and Johansson, 1972), naturally means that fiber types may differ more in related fishes of different habit than between, say, elasmobranchs and teleosts of the same habit, but Acrania and Agnatha have peculiar arrangements of their myotomal fibers that are not seen in other fish.

C. Ultrastructure and Histochemistry in Different Fish Groups 1. ACRANIA In amphioxus (Brunchiostomu),Flood (1966, 1968)has shown that there are three distinct fiber types, each only a few micrometers thick

Fig. 7. (A, B, and C) Histochemical differentiation of myotomal fiber types in Scylio rhinus. Stained for malate dehydrogenase activity. Sup, superficial fibers; SI and SII, type1 andtype11 slow fibers;FIandFII,typeIandtypeIIfastfibers.Alltosamescale. (D, E, and F) Low power electron micrographsofsuperficia1,typeI slow,andtypeIIfastfibers respectively; from young specimen ofScyliorhinus. Note differences inmitochondrial size

6.

LOCOMOTOR MUSCLE

379

Fig. 8. Diagrammatic view of amphioxus myotome as seen from an internal aspect, showing three types of muscle fibers, and their connection (at SC) with the motor neurons at the surface of the spinal cord. F, fast fiber; S, slow fiber; I, intermediate fiber. (Redrawn after Flood, 1968,Z. Zellforsch. Mikrosk. Anat. 84, 389-416.)

(Fig. 8). A T-system is absent, and in these exceedingly flattened fibers, the sarcoplasmic reticulum is represented only by subsarcolemma1 cisternae (Hagiwara et al., 1971; Flood, 1977) (see Fig. 6A). Similar cisternae (in addition to the more usual triad couplings) have been reported in various muscles of higher vertebrates (Spray et al., 1974); I have observed them in slow fibers of Scyliorhinus and they are probably figured in lamprey fibers by Teiavainen (1971, his Fig. 21). Superficial fibers contain abundant mitochondria and glycogen, as compared with the deeper mitochondria-poor fibers. The two fiber types send processes to the spinal cord (forming the ventral root “nerves”); each is in connection with a different region of the spinal cord. A third intermediate type of fiber was found to share this innervation region with the deep fibers. By analogy with other fish, it may be tacitly assumed that the superficial fibers (succinic dehydrogenase-positive) represent “slow” fibers, the deep fibers “fast” fibers. The unusual subsarcolemmal cisternae representing the sarcoplasmic reticulum support this identification, since these are larger in the deep than the superficial fibers (Fig. 6A). The status of the intermediate fibers is not clear. Flood suggests that they may perhaps be immature deep fibers, but the ontogeny of the system has not yet been examined at the ultrastructural level. Hagiwara and Kidokoro (1971) obtained physio-

and abundance, and in myofibrillar patterning. All to same scale. (G, H, and I) Different fiber types in the myotome oflophius, differentiated by succinic dehydrogenase staining. Note SDH-negative small diameter fibers in (G) (the superficial zone of the myotome), large SDH-negative fibers in (H) surrounded by SDH-positive fibers, and in (I), great variation in size of SDH-negative fibers from the fast zone of the myotome. All to same scale. Scale bars, 100 p m except for D, E, and F, 2 pm.

380

QUENTIN BONE

logical evidence for electrotonic coupling of adjacent fibers, but as they pointed out, no gap junctions have been observed between fibers.

2. AGNATHA In lampreys, the myotomal muscle fibers are arranged in compartments (Fig. 9) built from more or less cylindrical small-diameter slow parietal fibers, enclosing several layers of flattened fast central fibers. In different species the arrangement of the central fibers differs slightly, but in all, the central fibers closest to the parietal slow fibers are different in several respects from those in the middle of the compartment. Thus these “juxtaparietal” fibers are richer in mitochondria and oxidative enzymes than are the interior central fibers and have a richer capillary bed (Lie, 1974). On histological or histochemical grounds they might b e considered to be an intermediate fiber type. However, ultrastructural and physiological investigations by Tefavainen (1971) have shown that all central fibers within a compartment are coupled together electrically, and that all contract together, although only the most central are innervated. There are desmosomal connections between central fibers and between central and parietal fibers, but (as in amphioxus) the usual gap junctions correlated with electrical coupling were not observed by Teravainen, or by Jasper (1967). SR and T-systems triads occur at Z-line level, but the T-tubules may run longitudinally along one sarcomere (Jasper, 1967). Tefavainen’s work showed conclusively that, despite some morphological diversity, the central fast fibers all operated as a unit. This result is of importance, for it shows that so-called intermediate fiber types recognized by virtue of ultrastructural or histochemical

Fig. 9. Diagrammatic view of lamprey myotome from an internal aspect. Note absence of innervation of intermediate fibers (I); innervation ofdeep central fibers (C); and multiple innervation of slow fibers bordering muscle unit (S).

6.

LOCOMOTOR MUSCLE

38 1

properties, need not have different functional characteristics from neighboring fibers of similar innervation pattern. We should be wary of multiplying functional fiber types on purely morphological grounds. Lie (1974) noted that in ammocoetes of Lampetra Juviatilis the compartments contained only a single central fiber, flanked b y two intermediate fibers, whereas the adult compartments contained one or two central fibers. He concluded that the new central fibers had differentiated from intermediate fibers and that the latter could be considered as stages of development of central fast fibers. In hagfish, detailed ultrastructural studies by Korneliussen and Nicolaysen ( 1973, 1975) have established three morphologically and histochemically distinct fiber types, first defined histochemically by Flood and Storm-Mathisen (1962) and by Dahl and Nicolaysen (1971). As in lampreys, the Myxine myotome is divided into muscle units composed of a regular arrangement of the different fiber types (see Fig. lo), and the three fiber types are segregated in these units in a regular way. Ultrastructhdly, the three fiber types are distinguished by their content of glycogen, lipid, and mitochondria, by the pattern of the myofilament fields, and b y the organization of the Z-lines. All three fiber types possess M-lines; all three possess a T-system with triads at the A-I junction (unlike those of other fishes); in each fiber type there is a different relation between fiber volume and volume of the T-system. The familiar arrangement of the T-system as a cpllar around the muscle fiber is not found in all fiber types in Myrine, and Korneliussen and Nicolaysen (1975) ingeniously analyzed the density of triads to show that slow fiber myofilament bundles are only rarely encircled by T-tubules, whereas fast fibers are rarely devoid of them.

Fig. 10. Hagfish myotome viewed from an internal aspect. Fast fibers (F) terminally innervated, slow fibers (S) irmervated by axons passing onto fibers from both myoseptal ends. Occasional intermediate fibers (I) scattered around borders of muscle unit next to slow fibers; these are apparently terminally innervated.

382

QUENTIN BONE

Regular elegant views of the T-system and its segmental relations with the 'sarcoplasmic reticulum have been demonstrated in many teleosts (e.g., Kilarski, 1965, 1967),so that it may seem unusual to find such diversity in arrangement as demonstrated for Myxine. Nevertheless, as will be seen, a similar less ordered system is found in elasmobranch myotomal muscle, and is probably common during muscle fiber ontogeny in different fish groups. The ultrastructural differences between the three fiber types briefly considered above are similar to those of the elasmobranch, where they will be considered in more detail. It is important to notice that physiological investigation (Andersen et al., 1963) has only distinguished two functional fiber types, and the significance of the (morphologically) intermediate fibers is not yet clear, although myosin ATPase activity (Dahl and Nicolaysen, 1971) suggests that they may be intermediate in contraction speed.

3. ELASMOBRANCHS In the dogfish, Scyliorhinus, which has been studied in the most detail, four fiber types were initially recognized in the myotomes, on the basis of color, lipid content, and innervation (Bone, 1966). However, closer investigation (Bone et al., 1978b) has shown that while the distinction between fast, slow, and superficial fibers seems to be an absolute one, the fast and slow fiber types should each be subdivided further into two types. There are, then, in the dogfish myotome five types of muscle fibers recognizable upon structural and histochemical grounds [just as there are in the anuran limb (Smith and Ovalle, 1973)], but this is not to say that these all play different functional roles (see Section IV,F,4). The general arrangement of the myotome is seen diagrammatically in Fig. 11; Figure 7A-C shows the distribution of fiber types seen after malate dehydrogenase staining. Similar results are obtained when sections are incubated for other oxidative enzymes such as citrate or succinate dehydrogenase (SDH). After incubation for Ca2+-activatedmyofibrillar ATPase (Bone and Chubb, 1978), the same five fiber types are distinguishable; the ATPase activity increases from the outer surface of the myotome inward and is lowest in the superficial fibers and highest in the inner fast fibers. The outer border of the myotome consists of a single, sometimes interrupted, layer of SDH-negative superficial fibers, forming a sort of thin skin over the myotome. Such superficial fibers are not found in all sharks. I have not observed them in the spurdog SquaZus acanthias (Fig. 12), for example, and they do not occur in Galeus or Etmopterus (Kryvi, 1977). Immediately below (internal to) the superficial fibers

6.

LOCOMOTOR MUSCLE

383

Fig. 11. Scyliorhinus myotome from an internal aspect. Superficial fibers (SU) overlie slow fibers (I and 11); fast fibers are terminally innervated and mainly of type I. Note dual innervation of fast fibers.

are the typical slow fibers, or type I slow fibers, which are strongly SDH-positive. There may be a number of layers of these type I slow fibers, but they then merge into a second type of slow fiber which is less strongly SDH-positive, and usually larger in diameter than the type I fibers. There are two or three layers of these type I1 slow fibers, and an abrupt discontinuity where fibers that are nearly completely SDH-negative begin and make up the remainder of the myotome. At the zone of contact with the type I1 slow fibers, thesefast fibers are

Fig. 12. Diagrammatic transverse sections of postanal region showing fiber types in myotomal musculature. Upper row, fast fibers focally innervated; lower, fast fibers multiply innervated. A, Clupeids, probably also eels; B, Scyliorhinus; C, Squalus, probably also Dipnoi; D, Salmonids, some cyprinodonts; E, Some cyprinodonts, gadoids; F, Lophius. Slow fibers, dark hatching; intermediate fibers, light hatching; fast fibers, unhatched.

384

QUENTIN BONE

only a little larger than the slow fibers in diameter, and faintly SDHpositive. As the interior of the myotome is approached, after three or four rows of fibers, the type I fast fibers merge into larger SDHnegative type I1 fast fibers. Apart from the mosaic arrangement of salmonids, or the little known Lophius pattern, this myotomal organization is as complex as any and hence may be taken as a paradigm of the fiber types in fish locomotor musculature. The ultrastructure of these different fiber types are shown in Fig. 7D-F. It is evident that the superficial fibers are entirely distinct from the adjacent type I slow fibers, and from the type I fast fibers with which Brotchi (1969) equated them. The inconspicuous M-line, virtual absence of lipid, the paucity and simplicity of the mitochondria, and the few triads of T-system and SR set these fibers aside as a particular type rather different from either slow or fast fibers. Fortunately, their position in the myotome allows us to categorize them by a noncommittal term1 Interestingly enough, the superficial fibers at hatching are characterized by a simple myofibrillar arrangement recalling that of embryonic muscle cells, although the other fiber types are similar to their homologues in the adult fish. The slow fibers of both types are rather similar ultrastructurally, differing in details of myofibrillar arrangement and mitochondria1 density, but they are clearly separated histochemically when myotomal sections are incubated for glucose phosphoisomerase, since the type I1 slow fibers are positive, while the type I fibers are virtually negative. The fast fibers of both types differ from both types of slow fiber in the virtual absence of mitochondria, in a thinner Z-line, more extensive sarcoplasmic reticulum, and abundant triads with the T-system. It is not yet clear whether the T-system is significantly different in slow and fast fibers: Identical values of C, (membrane capacitance) were obtained b y Stanfield (1972) for both types of fibers (Table IV). In Galeus and Etrnopterus, Kryvi (1977) found similar values for T-tubule volume in slow, fast, and intermediate fibers, but the volumes of sarcoplasmic reticulum were higher in fast than in slow fibers. There are relatively more mitochondria in the type I fast fibers, and these fibers are weakly SDH-positive (they were originally termed intermediate fibers). Again, there seems to be a transition from type I to type I1 fast fibers over several fiber rows, the general impression gained is that the type I fibers are in some respects like the adjacent type I1 slow fibers, but that divorced from this propinquity, the fast fibers are uniformly of type 11, evidently highly specialized hnctionally. In Torpedo, but not in Scyliorhinus, type I fast fibers are differently innervated to type I1 fast fibers.

Table N A Comparison of Membrane Properties of Different Fish Muscle Fibers' Resting potential (mV)

Diameter (pm)

-85.2 -71.7

150 50

0.14 0.75

108 136

SIOW

-82.4 -73.1

60 63

1.01 1.4

294 389

fast slow

-87.8 - 74.5

fast

-75-85 -46 - 53

Species

Sc yliorhinus Myotomal

fast

SIOW Carassius Fin

Lamvetra Central Myxine Myotomal

fast

S~OW

Amphioxus

Input resistance

RP Rm 7 Cm (M) (Wcm) (Idl/cmz) (msec) (pF/cm2)

0.19 8.5 56-155 56.4 1-2 pm thick

10.0 0.8-1.2

40 40

1.59 5.4

15.9 47

10.3 10.2

A (mm)

2'36 2.27

Stanfield (1972)

'"

Hidaka and Toida (1969)

10.9 7.0

48'4 26.6

2.55

2.1

5.0 30.0

4.4 123

2-6 4.0

250pm 4.0

7'23

3.68 2.25

5.5 39.6 1.2-2.0

Authority

Teravainen (1971)

Nicolaysen (1976a,b) Hagiwara and Kidokoro (1971)

Most values are the means of a number of experiments. Different experimental techniques and assumptions complicate direct comparisons between different species. Ri, specific internal resistance; R,, membrane resistance; T , time constant; C,, membrane capacitance; A, space constant.

386

QUENTIN BONE

Although ray fin muscle was long ag0,investigated b y Ranvier

(1873), it has received little attention since that time. As in the myotomes of sharks, there are slow and fast fibers distinguished b y their color, diameter, and SDH activity (Bone and Chubb, 1975);superficial fibers are apparently absent, but there are both SDH-positive and negative small-diameter “slow” fibers, These fibers form a superficial strap in each fin ray muscle, unlike the “pinnate” fast fibers (see Section IV,A).

4. TELEOSTS Slow and fast fibers (red and white fibers) were first characterized morphologically in teleosts (Arloing and Lavocat, 1875), and subsequent investigations at the ultrastructural level have apparently confirmed the existence of only two fiber types in several teleosts (Buttkus, 1963; Kilarski, 1965, 1967; Nishihara, 1967; Patterson and Goldspink, 1972; Nag, 1972). As usual, the two fiber types are distinguished b y diameter, mitochondria1 and lipid content, and sometimes by Z-line thickness and b y difference in volume of the tubular systems. In some cases, e.g., in the fin muscle of goldfish, Carassius, or in the myotomal musculature of pike, Esox, no differences were observed between the sarcotubular systems of the two fiber types. As Nag (1972) remarks, it is possible that earlier workers did not appreciate the difficulties of making accurate estimates of the T-system. Since a large part of the membrane capacitance of the fiber resides in the T-system, it is desirable to know values of C, to check morphological estimates of the extent of the T-system. In Carussius, fin muscles, for example, C, for fast fibers is 7.23 pF/cm2, as compared to 2.55 pF/cm2 for slow fibers (Hidaka and Toida, 1969); although Nishihara (1967) did not note any difference in extent of the sarcotubular systems in the two fiber types, it is probable that such could only be detected by quantitative electron microscopy. It seems to be the case that the volume of the sarcoplasmic reticulum (that is, the internal tubular system excluding the T-system) may be very similar in both slow and fast fibers, even if the volume of the T-system may be different. It is possible that Patterson and Goldspink (1972) were unable to demonstrate sarcotubular differences between slow and fast fibers in Gudus virens because SR and T-system were considered together. In many teleosts (e.g., catfish, Barets, 1952; Mollienesia, Franzini-Armstrong and Porter, 1964), the muscle fibers have peculiar ribbonlike myofibrillar bundles round the edges of the fiber. In

6.

LOCOMOTOR MUSCLE

387

species with fibers of smaller diameter, these elongate bundles may occupy the entire fiber which then resembles a wheel with the myofibrillar bundles arrayed like spokes from a small central sarcoplasmic hub. This ribbonlike myofibrillar arrangement [which was utilized b y Lansimaki (1910) to categorize several muscle fiber types] is apparently unique to teleosts. However, sufficient investigation has not been made in such groups as Holostei and Dipnoi to exclude these: Certainly ribbon-myofibrils are not found in elasmobranchs, holocephali, or higher vertebrates, so far as is known. Lansimaki’s different categories have not been found valuable in distinguishing functionally different muscle fiber types, but the simple duality of teleost muscle fiber types described so far (even if true in the cases mentioned), is in many teleosts complicated by the existence of “intermediate” fibers, and b y the mosaic organization of the myotomes in different species. Barets (1952) described type I fast fibers (his aberrant fibers) in the catfish Ameiurus, and Greer-Walker (1970),using lipid staining noted in cod (G. morrhua), that there was an apparent mosaic arrangement among the type I1 fast fibers, as well as an intermediate fiber type lying between the slow and fast fibers. The situation in cod has been examined b y Korneliussen et al. (1978). Brotchi (1968) described intermediate fibers in carp (C. carassius) on the basis of SDH staining [they make up 7% of the myotome (Johnston, 1977)], and both SDH and myofibrillar ATPase activity differentiate several sorts of fiber in the herring Clupea harengus (Bone et al., 1978a), in the angler (Lophius),and in various other species (Johnston and Tota, 1974; Patterson et al., 1975; Mosse and Hudson, 1977). Figure 13 summarizes these various results. As yet, insufficient ultrastructural investigations have been made to allow comparison of these different fiber types with those of elasmobranchs. It seems evident, however, that the superficial red fibers correspond in all to type I slow fibers, but it is not clear whether “pink” or intermediate muscle fiber types should be considered as part of the slow or fast systems. In carp, Johnston et al. (1977) have shown intermediate fibers to b e active at intermediate swimming speeds (see Section IV,G). The mosaic arrangement reported in cod by Greer-Walker (1970) is similar to that observed in the deep lateral (fast) muscles of herring and Lophius where larger diameter fast fibers (equivalent to type I1 fast fibers of elasmobranchs) are surrounded by a regular array of smaller diameter fibers which are richer in mitochondria, slightly

388

QUENTIN BONE ACANTHOPTERYCII

CARACANTHOPTERYGll

Lophllformr

I

Coblnwiformrr

P8rclform darlntlrct

I

Smrprenlforma Burxholdlformer

\I

ATHERINOHORPHA

-

\ OSTEOGLOSSOMORPHA

\--7 **- .

Jurassic Protoelopold

’

Pholldophorold

1 holorceans

\

PROTACANTHOPTERYGII

I

.

. .

Mrctophoidr Alcpocephalidr

>’’

Fig. 13. Distribution of terminally innervated fast fibers (underlined) in teleost groups. All higher teleosts have multiply innervated fast fibers. Only Hiodon of the Osteoglossiformes has terminally innervated fast fibers; a few catfish of the Ostariophysi also show terminal innervation. (Modified from Bone, 1970; redrawn from Greenwood et d., 1966, B d . Am. Mus. Nat. Hist. 131, 339-456.)

SDH-positive, and contain more glycogen than the type I1 fibers. These fibers (Fig. 71) are probably to b e equated with the type I fast fibers of elasmobranchs. Korneliussen et al. (1978)have categorized no less than seven different fiber types in cod (on the basis of cryostat sections incubated for different enzymes), and three of these are found in the deep “fast” motor system. As these authors imply, it is an open question whether histochemical and morphological diversity of this kind in the teleost fast motor system (indeed in elasmobranchs and agnatha also) reflects functional diversity, or whether it is a simple concomitant of the extensive pattern of growth in fishes. In lampreys, the accident of coupling between the central and intermediate fibers (Ter‘avainen, 1971, see Section IV,C,2) taken together with Lie’s observations on growth strongly suggest that intermediate fibers should properly be considered as growth stages in the formation of type I1 fast fibers. We do not know whether this is the case in other groups of fish,

6.

LOCOMOTOR MUSCLE

389

but the results of Korneliussen et al. (1978)certainly are most simply interpreted in this way. The problem of the status of intermediate fiber types distinguished by morphology and histochemistry is particularly acute in salmonids. Here, it has long been recognized that there are small-diameter “red” fibers scattered among the deeper portion of the myotome, as well as segregated in a superficial lateral zone, as in other fish. T h e mosaic of the deep portion of the musculature (Boddeke et al., 1959) was at first interpreted as a mixture of fast fibers with slow fibers that were the same as those of the lateral zone, i.e., that salmonids possessed “extra” slow fibers (e.g., Webb, 1971). More recently, Johnston et al. (1975) have examined the histochemistry of fiber types in Salmo gairdneri, and conclude that the mosaic portion of the myotome is composed of large-diameter fast fibers, and of small-diameter fibers which differ from those of the superficial zone in SDH and myofibrillar ATPase activity. Their conclusions ?re summarized in Fig. 12D. The similarity with the pattern found in the herring or in Lophius (Fig. 12F) is striking, and it is equally striking that there is in the mosaic muscle an almost continuous distribution of fiber size (from 15 to 95 pm). Evidence from electromyography of salmonids (Hudson, 1973; Bone et al., 1978a) considered in Section IV,G, is conflicting, and as yet, insufficient to decide whether slow fibers are intermingled with fast fibers, or whether the morphologically and histochemically similar mosaic of clup‘eids and gadoids indicates that the small-diameter fibers of the salmonid myotome are growth stages in the development of the larger fibers. A special case is presented by Lophius (Bone and Chubb, 1978)for in the myotomal musculature of this relatively inactive fish, there are a great variety of fiber types. As well as the mosaic fibers-of the deep (white) portion of the myotome, already referred to, which resemble those of salmonids or clupeids, Lophius is remarkable in having among the usual SDH-positive, lipid-rich, small-diameter fibers of the superficial lateral zone (the type I slow fibers), both bundles of SDHnegative, lipid-poor small-diameter fibers, and also occasional similar fibers of large diameter. This complex situation is seen in Fig. 7G-I. These fiber types are found in Lophius. of different sizes; their significance is not understood, for although the mosaic arrangement in the deep portion of the myotome may b e viewed as a consequence of a pattern of continued growth, it is less easy to interpret the arrangement in the superficial zone in such a way. It is obvious that further investigations are needed, both of the physiology of fiber types, and of their development in the mosaic muscle.

390

QUENTIN BONE

In thefins, so far as is known, slow and fast fibers are segregated, as they are in the fin muscles of other teleost groups (e.g., Nishihara, 1967).

D. Ontogeny of Fiber Types There are two linked problems in the differentiation of muscle fibers in fish. First, there is differentiation in embryonic development giving rise to the fiber types and relative numbers found on hatching or metamorphosis; second, there are the separate problems raised b y the continuous growth in all fish and, in some, the changes occurring prior to a migratory phase (as in AnguiEla). As the fish grows, increasing in length and mass, the relative amounts of slow and fast fibers change (at least in some species), in addition to actual increase in numbers of each fiber type. In other species, such as some of the pelagic stromateoids, there is a complete change in locomotor behavior during adolescence, which is probably again reflected in changes in the relative amounts of slow and fast fibers. How are the various fiber types differentiated, and what are the mechanisms for the changes observed in the fiber populations (e.g., Magnuson, 1973) as fish grow? Although it is manifest that (as Nag and Nursall, 1972, put it) “fish muscle fibers offer an interesting problem in differentiation because of the presence of two types of fibers (white and red) which are involved in two kinds of swimming activities, which in their turn appear in different stages of development of the fish,” there are few recent studies of these problems. In teleosts, Nag and Nursall studied the histogenesis of fiber types in the myotomes of Salmo gairdneri up to the free-swimming fry stage, and Waterman (1969) examined the development of the same system in the cyprinid Brach ydanio rerio. Fortunately both accounts are in substantial agreement, strongly indicating that differentiation begins from the inner (medial) face of the myotome outward and that (in the way that Vialleton, 1902, had suggested) deep myoblasts differentiate into deep fast fibers, whereas surface myoblasts later differentiate into the superficial slow fibers. As Waterman emphasized “the two main fiber types exhibit structural differences from the time of their formation and differentiate along separate pathways leading to dissimilar adult configurations.” The somites are initially made up of more or less rounded cells with large intercellular spaces between them; as development continues these cells become more closely apposed and either develop

6.

LOCOMOTOR MUSCLE

391

specialized embryonic or focal intermediate junctions (Bruchydunio) or adjacent cell membranes show indications of fusion (Salrno).The medial cells in each somite then elongate and begin fibrillogenesis to become myoblasts, containing bundles of myofilaments and the beginnings of a patterned SR. At this stage, the superficial presumptive myoblasts of the somite are not much elongated, nor do they contain myofilaments. These develop at a later stage, by which time the deep myoblasts have developed myofilament fields and may b e considered as young fast muscle cells. During this process, the myoblasts destined to become fast muscle fibers become relatively less rich in mitochondria; indeed, the early myoblasts are, perhaps, more similar to the developed slow fibers than they are to the developed fast fibers. This is not to say that fast fibers pass through an “embryonic” slow stage, or that all myotomal fibers commence as slow fibers; development of the two seems to be quite separate. Waterman found that fibrillogenesis could occur either in mononucleate myoblasts or in the multinucleate later stages of the deep myoblasts, but he did not observe nuclear division or cytoplasmic fusion in his material. On the other hand, Nag and Nursall observed appearances indicating cytoplasmic fusion at different stages and suggest that multinucleated fibers are derived by coalescence of myoblasts and presumptive myoblasts. It is certainly possible that there are species differences and perhaps more significantly, differences between embryos of different dimensions, but further investigations are needed before a “norm” of development can b e established. I n any event, considering only the broad categories of slow and fast fibers, these are apparently “fixed” from an early stage, and it is natural to ask whether this dichotomy is regulated by the nervous system. Waterman (1969)observed the first axons in the myotomes close to the medial ends of young superficial muscle cells, and concluded that this fiber type was the first to come under nervous control. Previous large irregular movements of the embryo were held to be due to myogenic contractions of the deep fast fibers, as yet not in connection with the nervous system. In a similar way, Nag and Nursall observed initial twitching movements of advanced embryos within the eggs, and stimulation of early larvae produced only short bursts of vigorous tail beats. Slow tail movements were not observed until a later stage, as Waterman had found. The early development of swimming movements is probably best known in dogfish (Harris and Whiting, 1954), and a preliminary study of the development of fiber types in dogfish ontogeny has shown the sequence to b e similar to that in teleosts. That is, deep fast fibers develop first and seem to become

392

QUENTIN BONE

innervated later than the slow fibers which appear after the fast fibers in ontogeny. The last fibers to appear are the superficial fibers which even after hatching are notably “embryonic” in appearance. Later development of Etmopterus slow and fast fibers has been studied by Kryvi and Eide (1977), who found that postembryonic growth takes place largely by muscle fiber hypertrophy; evidently this process gives rise to the notably uniform diameter of fibers in the fast region of the adult myotome, as compared to that in such teleosts as carp or herring, whose differentiation of fibers from myosatellite cells presumably continues throughout growth. Willemse and van den Berg (1978) examined the growth of myotomal fibers in Anguilla over a %year period and found that initial growth of red fibers takes place by increasing in diameter; later, new fibers are added, as they are in the white zone of the myotome at all stages, probably from myosatellite cells. Underlying the early myogenic activity of the deep fast fibers of the myotomes in both dogfish and teleosts are longitudinal connections between the muscle cells of adjacent somites, first observed in teleosts by Waterman. These interm yotomal connections are transient in both groups, presumably disappearing as the deep fibers receive innervation from the segmental nerves. It does not seem then, that the fiber types differentiate from an uniform population of late myoblasts following outgrowth of motor axons of different motoneuron classes, but it is obvious that further work is needed before the regulation of myogenesis is understood. I n this field, fish muscle offers interesting material to the experimental embryologist, for the general agreement that fiber types are fixed from an early stage, and the segregation or zonation of fiber types at once suggests possibilities of experimental alteration to determine the effects of propinquity of one fiber type upon another, or the influence of innervation upon the different fiber types. More detailed histological studies are needed too, particularly upon species which have “intermediate” fibers of various sorts, and upon scombroids with a medial mass of slow fibers. It will not have escaped notice that Nag and Nursall’s description of the development of the salmonid myotomal fiber types up to the free-swimming stage only dealt with superficial slow fibers and deep fast fibers. Yet in the adult salmonid, the deep portion of the myotome is typically “mosaic” (Section IV,C,4), containing a mixture of fiber types. It is not yet clear how this mosaic fiber population arises, nor (in other species) how the relative amounts of slow and fast muscle fibers are altered. The continuous growth shown by fish is reflected in some

6.

LOCOMOTOR MUSCLE

393

species, for example, Lophius (Figs. 12F and 7G-I), by patterns of muscle fiber types arranged in such a way that the generation of the arrangement is evident. I n most fish which have been examined, myosatellite cells are abundant and show in young specimens all stages of development of myofilament fields. I n the shark Galeus melastomus Kryvi (1975) found that myosatellite cells were about twice as abundant on the slow fibers as on the fast fibers. As he observed, the presence of peripheral microtubules, of extensions deep into the body of the muscle fiber, as well as lysosomes and dense bodies in the satellite cell cytoplasm, all suggest that the functions of fish myosatellite cells have not yet been fully determined and the possibility of some trophic influence on adjacent cells has been suggested (Flood, 1971). No detailed study has yet been made of the organization of satellite cells in fish of different ages (see, however, Kryvi and Eide, 1977), but preliminary examination of Scyliorhinus material indicates that they become less abundant as growth proceeds, in accord with observations in teleosts by Nag and Nursall(l969). It is agreed that myosatellite cells are persistent myoblasts, and that increase in fiber number during growth is attributable to differentiation of myosatellite cells, but how this process is regulated is unknown. The view that the mosaic nature of the myotomal population in many teleosts is simply a reflection of continued differentiation of myosatellite cells has certainly the merit of simplicity. If it is true, then growth in sharks seems to occur rather differently, since their myotomes do not show a mosaic of fiber size. Nor, however, does there seem to be a special growth zone (for example, between the deepest slow fibers and the most superficial fast fibers) so that it is not known how the system increases in fiber number. Plainly, slight changes in the rates of differentiation of myosatellite cells in the slow and fast portions of the myotome during growth will result in changes in the relative amounts of slow and fast fibers in the myotomes, such as are found as fish increase in size.

E. Innervation The innervation of fish muscle fibers presents unusual features since many fish have multiply innervated twitch fibers, and the pattern of innervation is of taxonomic value. There are, further, hints in some fish motoneurons of transmitter substances other than acetylcholine, and of dual innervation of the muscle fibers. In all fish groups, myotomal superficial slow fibers are multiply

394

QUENTIN BONE

innervated by small-diameter myelinated fibers that terminate in en grappe endings (Barets, 1961; Bone, 1964, 1966, 1970; Best and Bone, 1973). Where two types of slow fiber are recognized, as in dogfish, both are innervated in a similar way. In teleosts, the nerve terminals are usually embedded in the sarcolemma, and subjunctional folds are absent (Nishihara, 1967; Barets, cited in Barker, 1968); in dogfish, there are subjunctional folds under terminals on superficial and type I and type I1 slow fibers (Bone, 1972a). The slow fibers of the hagfish myotome are innervated by two axons only, passing onto the fiber at each of its myoseptal insertions (Andersen et al., 1963) but in other groups, more than two axons probably innervate each slow fiber. I n the fins, slow fibers are also multiply innervated, although the endformations may be larger than in the myotomal slow fibers (Fig. 6D and J). The interval between terminals along slow muscle fibers has been measured in a number of preparations and where the membrane constants of the fiber have also been measured there is a normal safety factor of around 10 times. For example, Stanfield (1972) found A to be around 2.27 mm for slow fibers in dogfish myotomes; motor terminals along these fibers are some 150-200 pm apart. Acetylcholinesterase is demonstrable at the terminals on all slow fibers investigated; in teleosts Pecot-Dechavassine (1961)found that other esterases (e.g., butyrylcholinesterase) are absent. However, an interesting possibility has been raised in hagfish, b y Korneliussen’s (1973) suggestion that slow fibers may have monoaminergic innervation. He showed that dense core vesicles were especially abundant in nerve terminals on slow fibers of the hagfish myotome and craniovelar muscle, although formaldehyde-induced fluorescence proved inconclusive. Earlier work by Andersen et al. (1963) showed that acetylcholine was probably the transmitter at the neuromuscular junctions on slow and fast fibers in the myotome; Korneliussen’s observations await further investigation. To judge from pictures by Teriivainen (1971) and Nakao (1976), dense-core vesicles are not found in unusual number in lamprey slow fiber terminals. The innervation of fast fiber types is, however, of more interest than that of slow fibers, for there are large differences between different fish groups, such that the pattern of innervation may serve as a taxonomic character (Bone, 1970). In all groups except most of the teleosts, the fast fibers are focally innervated at their myoseptal ends (Fig. 61); sometimes at both ends of the fiber (according to Barets, in Ameiurus),but more usually, at one end of the fiber only (in hagfish and elasmobranchs). This terminal innervation is also found in amphibia (at both ends of the fiber) and seems to be the original innervation pattern in lower chordate myotomal fast fibers. Best and Bone

6.

LOCOMOTOR MUSCLE

395

(1973) have suggested (see also Bone, 1975) that this innervation pattern is part of the various specializations of the fast fibers to increase power output b y devoting the maximum space to myofilaments. In elasmobranchs, in urodele amphibia, and in certain teleost groups, each fast fiber is apparently innervated by two separate axons, which both contribute to the formation of a single motor endplate. In most of these forms, there is no recognizable difference between the two motor terminals, but in some elasmobranchs (Bone, 1972a) they are recognizable by their different vesicle content. One type of terminal at the endplate contains “typical” cholinergic electron-lucent vesicles 50 nm diameter, while the other contains a high proportion of much larger dense core vesicles (up to 100 nm). As in the hagfish slow fibers, formaldehyde-induced fluorescence studies have so far proved inconclusive, and only acetylcholinesterase has been found in the subsynaptic folds. The significance of this dual innervation is therefore unresolved. In the dogfish, as in most focally innervated species examined, every type of fast fiber is innervated terminally, even if (as in herring) they may be diverse in size. So both type I and type I1 fast fibers are terminally innervated in dogfish, and no difference has been recognized between the motor terminals on each fiber type. But in two cases, the “aberrant” fast fibers of Ameiurus (Barets, 1952), and the most superficial fast fibers of the myotomes in Torpedo (Bone, 1964), there is focal innervation in the midregion of the fiber. Again, the significance of this situation (rather unusual in fish groups with terminal innervation since it has not been observed in clupeids, eels, holostei, dipnoi, sharks, or other rays) is not understood. No intermediate types of innervation are observed, and it does not seem very probable that the terminal innervation pattern is derived from the midregion en plaque type. In Ameiurus (and possibly also in Torpedo),these peculiarly innervated fast fibers lie external to the connective tissue fascia separating the slow fibers from the main mass of fast fibers. It is premature therefore to equate them with the type I fast fibers in dogfish, and to suppose that their different innervation precludes the type I fast fibers of other fishes from being developmental stages in the formation of type I1 fibers. Terminal innervation is found in certain teleost families (Barets, 1961; Bone, 1964, 1970);these have an interesting taxonomic distribution (Fig. 13). It is apparent from Fig. 13 that no acanthopterygians have this pattern and that, of the families regarded as primitive on other grounds, only the salmonids lack the terminal innervation pattern. The systematic position of a number of the groups possessing terminal innervation (for example, the alepocephalids) has not been

396

QUENTIN BONE

agreed upon. The pattern of innervation of other taxonomically vagrant groups (e.g., many deep-sea families) has not been examined. The “innate conservatism” of soft parts should allow us to employ muscle innervation as a character in determining the relationships of teleosts where skeletal structures give conflicting evidence for their affinities. With the exception of the groups indicated in Fig. 13, all teleost families so far investigated (species in 55 families have been examined) have an entirely different innervation of the fast fibers. Each muscle fiber is multiply innervated (Barets, 1961; Hudson, 1969); the innervation is similar to that of the superficial slow fibers, distributed and punctate (Fig. 14). According to Barets the fast fiber innervation is more distributed than that of the superficial slow fibers. As in the slow fibers, the motor terminals are embedded in the sarcolemma and subjunctional folds are absent. The richness of the innervation of teleost multiply innervated fast fibers is astonishing (Fig. 6B and E); the nexus of axons resembles the capillary bed of slow fibers in its abundance and nerve terminals are scattered all along the muscle fibers. The contrast with focal terminal innervation as seen in other fish fast fiber systems is very great and invites functional interpretation (see Section IV,F,5). Thus far, myotomal muscle fiber types have been considered. In the fin muscle, a similar division exists between the multiply innervated fast fiber fin muscles of higher teleosts (Fig 6E), and the focally innervated fast fibers of other groups (Fig. 6C). I n the fin fast muscle fibers, however, innervation is not terminal; large en plaque endings are found in the midregion of the fibers. These end-formations may be very large, and derived from several branches of the same axon which have divided earlier along its course and come into proximity again at

Fig. 14. Higher teleost myotome viewed from an internal aspect showing multiple innervation of all fiber types. It is not known whether any overlap of innervation takes place between fast (F) and intermediate (I) fibers.

6.

LOCOMOTOR MUSCLE

397

Fig. 15. Drawing from whole mount preparation of fast fibers of pectoral fin in Torpedo showing overlap of innervation between muscle fibers; a and b are two axons supplying muscle fibers labeled according to which axons innervate them.

the motor endplate. Figure 15 illustrates the complexity of motor terminations in Torpedo, perhaps directed to prolonging the period of subjunctional permeability in these large diameter muscle fibers. The ultrastructure of focally innervated fast fibers in the fins has not been examined, but that of multiply innervated fibers has been studied in several families (e.g., Nishihara, 1967; Bergman, 1964a), and resembles that of the body muscles, viz., in the absence of subjunctional folds.

F. Electrical and Mechanical Properties Several workers have investigated the electrical properties of myotomal muscle fibers in different fish groups (e.g., Barets, 1961;

398

QUENTIN BONE

Teiavainen, 1971; Stanfield, 1972), and, in Agnatha, Teravainen and Rovainen (1971) have been able to obtain simultaneous intracellular records from myotomal muscle fibers and the motoneurons supplying them; but the mechanical properties of myotomal fibers have been little studied, probably because the myoseptal insertions of these fibers are inconvenient for attachment to recording apparatus. The recent work of Wardle (1975) shows that such investigations are likely to prove rewarding. Fin muscle fibers provide better physiological preparations, and their mechanical properties are therefore easier to study (Bergman, 1964b; Hidaka and Toida, 1969; Yamamoto, 1972). I n elasmobranchs, fin muscle fibers are often of large diameter, and thus offer suitable material for studies on membrane properties (e.g., Hagiwara and Takahashi, 1974).

1. AMPHIOXUS

The interesting work of Hagiwara and his colleagues (Hagiwara and Kidokoro, 1971; Hagiwara et al., 1971) on the myotomal muscle lamellae (Fig. 8) of amphioxus has shown that there are two independent mechanisms for permeability increase; the normal action potential is mainly the result of increase in sodium conductance, but sufficient calcium ions enter during the spike to play a significant role in excitation-contraction coupling. Hagiwara and his colleagues consider that in practice what happens is that the sodium spike is important in conducting an impulse from the central motor endplate along the thin nervelike portion of the muscle fiber to its expanded contractile region where the calcium influx is concerned with excitation of the contractile apparatus. They also present evidence indicating that the subsarcolemmal cisternae of the SR are concerned with sequestering calcium rather than releasing it during contraction (as in higher forms). These unusual properties are of course related to the unique morphology of the system, presumably also responsible for the low values for the membrane resistance and time constant (Table IV). Electrical coupling between separate muscle cells was detected in half of the tests made; the morphological basis for this has not yet been established (Section IV,C,l) and it would b e interesting to know more of the properties of the (morphologically) different fiber types. 2. AGNATHA

Lampreys and hagfish offer favorable material for determination of the electrical and mechanical properties of muscle fibers. In lampreys, Teiavainen’s (1971) analysis has shown that the slow and fast (lateral

6.

LOCOMOTOR MUSCLE

399

and central) fibers have different input resistances and time constants, as expected from their differences in dimensions (Table IV); only fast fibers showed propagated overshooting action potentials (Fig. 16). Miniature endplate potentials were observed in both slow fibers and in fast central fibers (sometimes of two different rise times in the same fiber), but were not seen in lateral fast fibers. These last are electrically coupled to the central fast fibers, and are not directly innervated themselves. Again, as in amphioxus, gap junctions have not been observed; the morphological basis for electrical coupling has not been determined (Section IV,C,2). Recording tension directly from the abdominal surface of the myotome, Ter‘avainen was able to show (Fig. 17) that nerve stimulation gave a tetanic plateau at 2O/sec, and that relaxation followed a dual time course, presumably because the slow and fast fibers relaxed at different rates. In hagfish, Andersen et al. (1963) obtained similar mechanical results from the myotomes, i.e., single stimuli to the myotomal nerve gave evidence for the existence of slow and fast components, the slow component having a much longer relaxation time. As in lampreys, fast fibers gave propagated overshooting action potentials, from slow fibers only junctional potentials were elicited. But these junction potentials are unusual in their size, up to 30 niV, which is near to zero membrane potential! They have been studied b y Alnaes et al. (1963).Jansen and his colleagues conclude from records such as that of Fig. 18 where discontinuities are found on the junction potentials, that the slow fibers in hagfish may produce abortive spikes when stimulated indirectly. Stanfield (1972) suggests from his study of slow fibers in the dogfish Scyliorhinus (see next section) that the observations in hagfish may be explained in terms of some sodium conductance in these fibers. The ionic basis of electrical activity in hagfish fibers has not been investigated fully, and would seem to be of interest in view of the more recent observations of “spikes” in slow fibers when these are

Fig. 16. Overshooting spikes recorded from fast (central) muscle fibers of lamprey. Note that A (an innervated fiber) shows a prolonged after potential, whereas B is a record from a noninnervated fiber coupled to an innervated fiber and thus does not show this endplate potential. (From TerBvBinen, 1971.J. Neurojhysiol. 34, 954-973.)

400

QUENTIN BONE

Fig. 17. Tension records from abdominal surface of myotome after nerve stimulation at 6/sec (upper) and 2O/sec. Early and late relaxation seen in lower record presumably result from activity of fast and slow fibers, respectively. (From Teravainen, 1971, J . Neurophysiol. 34, 954-973.)

immersed in experimental solutions (e.g., Hidaka and Toida, 1969). Most recently, Nicolaysen (1976a,b) has examined the spread of potential in the T-system of hagfish fast and slow fibers using sinusoidal transmembrane currents.

3. ELASMOBRANCHS Hagiwara and Takahashi (1967, 1974) and Stanfield (1972) have examined the membrane properties of the fin muscles of several

Fig. 18. Junction potentials from hagfish slow fibers. Note discontinuity on rising phase (arrow) interpreted as an abortive spike (see text). (From Andersen et nl., 1963).

6. LOCOMOTOR

40 1

MUSCLE

species of tropical stingrays, and of the myotomal muscle fibers of the dogfish, Scyliorhinus. The cable properties of the myotomal fibers investigated are seen in Table IV. Fast fibers in fins and myotomes gave overshooting propagated action potentials as expected from their focal innervation; slow fibers never gave propagated potentials, but an abortive spike was seen b y Stanfield on one occasion. Using a twoelectrode clamp, Stanfield found that eight of twenty-seven myotomal slow fibers showed sufficiently large inward sodium currents as soon as depolarized to suggest that they were capable of propagating action potentials. Six other slow fibers showed no inward sodium current on depolarization, and others showed a small inward sodium current. Figure 19 illustrates active current-voltage relations in the extreme cases. No significant difference in cable properties was found between slow fibers capable of showing a marked conductance change to sodium and those in which no change was found. These results are interesting, for they suggest that reports of abortive spikes and small spike potentials at the break of strong inward currents are to be explained in terms of some sodium conductance, as Stanfield pointed out. Since Stanfield wrote, there have been further investigations of morphology of the two fiber types, and it is known that the slow fibers can b e divided into two types according to their histochemistry and position within the slow fiber portion of the myotome. Yet Stanfield

80 70 -

60 50 -

40

-

30 -

20 -

10 -

-110

I

-J-80 -10

A

-60

-40

,

-20

,

,

!

0 10

mV

Fig. 19. Active voltage-current relations obtained from dogfish slow fibers using a two-electrode voltage clamp technique, In A, the dashed line shows the voltagemembrane current relation obtained from Cole’s theorem. Note delayed rectification and absence of inward sodium current. In B, also a slow fiber, filled points indicate currents flowing at end of 100 msec pulse, open points initial currents flowing at about 2 msec. In this fiber, delayed rectification is again found, but significantly, there are large inward sodium currents, resembling those of the fast fibers. Threshold for the sodium conductance was around 60 mV. [From Stanfield, 1972,J. Phy~iol.(London) 222, 161186.I

402

QUENTIN BONE

emphasized the absence of sequestration of fibers with and without sodium currents in different parts of the slow fiber portion, so that it is not clear whether type I and type I1 slow fibers are different in this respect. Although there exists the possibility of propagated events among the slow fibers, no evidence has yet been adduced to suggest that these are found during normal swimming; crude records suggest that the slow fibers do indeed have the mechanical properties of slow fibers, contrasting with the twitch fast fibers.

4. TELEOSTS Barets (1961) and Hudson (1969) have examined myotomal muscle fibers, but most workers have used fin muscle preparations. In a variety of freshwater and marine teleosts there is good agreement between the results obtained from fibers that are innervated in a comparable way. All authors agree that slow (red) fibers of fins or myotomes do not propagate action potentials; the properties of these slow fibers are similar to those of other fish groups; as in elasmobranchs (Stanfield) there are hints of a sodium conductance mechanism in some teleost slow fibers. Where fast fibers are focally innervated, as in the myotomes of the catfish Ameiurus (Barets, 1961), or the fins of Conger (Hagiwara and Takahashi, 1967), typical overshooting propagated action potentials are found, similar to those in other vertebrate groups. Many teleosts, however, possess multiply innervated fast fibers, and here the situation may be different. Most workers have found that muscle action potentials from fibers of this kind are characterized by failure to overshoot, or by overshoots close to zero membrane potential (Takeuchi, 1959; Barets, 1961; Hagiwara and Takahashi, 1967; Hidaka and Toida, 1969). Hudson (1968) found in the marine teleost Cottus that 20 mV overshoots (Fig. 20) were obtained provided care was taken to experiment using a Ringer solution containing appropriate values of Ca2+ and Mg2+ ions. Reduction of these ions b y 50 and 30%, respectively, gave overshoots close to zero membrane potential. It is certainly tempting to suppose that appropriate ionic adjustments to the Ringer solutions used by previous workers would have allowed them to observe overshooting spikes, as Hudson suggests, but this has yet to be demonstrated. Alternative explanations of nonovershooting spikes are possible (e.g., in some fish high internal sodium may lower the sodium equilibrium potential), and as Hagiwara and Takahashi observe, multiply innervated fibers may be able to afford a lower safety factor than focally innervated fibers since contraction is not uniquely dependent upon propagated action potentials. In this way

6.

LOCOMOTOR MUSCLE

403

Fig. 20. Overshooting spike from multiply innervated fast myotomal fiber of Cottus. The arrow indicates the second of two summated junction potentials on the rising phase. (From Hudson, 1969.)

they could effect economies in ion transfer across the sarcolemma. However this may be, Hudson’s (1969) work revealed several interesting points about teleost multiply innervated fast fibers. By simultaneous recording from the nerve which was stimulated, and a single muscle fiber (Fig. 21), Hudson showed that each muscle fiber was innervated by at least two axons in a single spinal nerve and b y a similar number of axons in each of four neighboring spinal nerves, a remarkably high degree of polyneuronal innervation. Until tension records are taken simultaneously with electrical records from multiply innervated fast fibers, it will not be known for sure if such fibers are capable of local contractions and also of twitches following propagated action potentials. Observational data (e.g., Barets, 1961; Takeuchi, 1959; Bone et ul., 197th) suggest that this is the case; the functional advantage of such an arrangement is considered in the next section. The number of muscle fibers innervated by single axons (i.e., the size of the motor unit) is not well known in either myotomal or fin muscles. Teravainen and Rovainen (1971) suggest that in lamprey myotomal muscle, about ten to twelve fast motoneurons on each side of the spinal cord innervate an equal number of muscle units on that side, each muscle unit consisting of the central fibers and accompanying electrically coupled (noninnervated)lateral central fibers. In such focally innervated systems it is of course in principle possible to count motor axons in the nerve passing to the muscle, and divide the number obtained into the number of muscle fibers (making due allowance for the presence of different muscle fiber types where necessary). On this basis, the myotomal fast motor unit in Scyliorhinus consists of some 50-100 muscle fibers. Since this fiber type is apparently used only during burst swimming, it would not be surmised that it was finely

404

QUENTIN BONE

Fig. 21. Simultaneous recordings of the compound action potential from spinal nerve (upper) and concomitant electrical activity of a single muscle fiber (lower), when stimulus intensity is varied. Zero membrane potential indicated in all except (f) by baseline of nerve record. In (f) small changes in stimulus intensity give different (superimposed) muscle responses at similar latency. The experimental situation is shown at the bottom. S, stimulating electrodes; E, earth electrode. CRO, recording electrodes to oscilloscope. (From Hudson, 1969.)

graded. The focally innervated fast fibers of the pectoral fin in herring are supplied by very few large axons which branch repeatedly (Fig. 6H): Gradation must be relatively coarse in this case. It is manifestly more complicated to unravel the possibilities of gradation in the multiply innervated fin or myotomal systems. Roberts (1969b)observed in slowly swimming spinal dogfish that there was good correlation between the duration of the muscle bursts (cf. his Fig. 21), the number of impulses in the burst, and the swimming frequency, indicating that the frequency and composition of the discharges of the motoneurons were controlled by a single mechanism. That is to say, the slow motor system is graded (as in amphibia) by variations in the amplitude and frequency of the junction potentials. Observations on the slow motor system of the unpaired fins led to the same conclusion. These are slow fibers which apparently d o not exhibit propagated muscle action potentials. Gradation in the multiply innervated fast fibers of teleosts is considered b y Hudson (1973), who suggests that

6.

LOCOMOTOR MUSCLE

405

during sustained swimming activity, motor units may be rotated in the fast motor system. The speed of fish locomotor fibers varies widely, from fusion frequencies up to 120 Hz for seahorse fin muscles (Bergman, 1964b) to 20 Hz for lamprey fast fibers (Terivainen, 1971). Isometric muscle contraction time for the seahorse fin fibers was as little as 10 msec. Wardle (1975) has given comparative isotonic muscle contraction times for fast myotomal muscle fibers in various teleosts, ranging from 20 to 45 msec, contraction time being related, as expected, to the size of the fish (increasing as fish length increases). Little is known in fish of the factors determining contraction velocity of different fiber types from a single fish. It is possible that the low molecular weight proteins studied by Hamoir and his colleagues (e.g., Syrovy e t al., 1970; Hamoir et al., 1972), which are known to be present in fish fast fibers but are much less abundant in slow fibers, may be concerned in some way in regulating muscle contraction velocity, perhaps by their effects on calcium-activated myosin ATPase, acting as calcium buffers in a situation where there are few mitochondria.

G . Functional Role of Different Fiber Types A variety of suggestions have been made for the functions of the two main fiber types in fish muscle. These are summarized in Bone (1966), but it is now generally agreed that the superficial slow fibers are utilized by the fish for sustained slow-speed swimming or cruising and the deeper fast fibers for bursts of higher speed. Several lines of evidence point to this conclusion. First, direct electromyographic recording from teleosts swimming freely or in tunnel respirometers, or from spinal sharks, has shown that electrical activity is fouud within the zone of the superficial fibers when the fish is swimming slowly and within the deeper zone of the fast fibers during rapid swimming (Bone, 1966; Rayner and Keenan, 1967; Hudson, 1973; Bone et al., 197th). Second, biochemical and metabolic studies (reviewed in Bilinski, 1974) have shown that slow fibers operate mainly by aerobic glycolysis and lipolysis, fast fibers b y anaerobic glycolysis. Examination of fish after exercise of different kinds has shown the expected utilization of metabolites by slow and fast fibers (see Chapter 8). It was perhaps natural to follow Arloing and Lavocat (as did Boddeke et al., 1959) in supposing that red and white fibers (or slow and fast fibers) were distinct and separate systems, utilized for different patterns of swimming at different speeds. The earlier electromyographic work on spinal dogfish showed that muscle action potentials

406

QUENTIN BONE

could only be recorded from fast fibers in the deep portion of the myotome when the fish was strongly stimulated and swam very vigorously for a few tailbeats (see Fig. 22). At the usual slow spinal swimming rate of some 35 strokedmin, the fast fibers are silent, as they are even if the rhythm is speeded up b y deliberately oscillating the fish to higher tailbeat frequencies. There seems to be no question that in sharks, the deep fast fibers are only utilized during fast bursts of swimming. Calculations based upon the rate of depletion of the fast fiber glycogen reserves suggest these fibers could only operate for around 2 min of sustained activity (the time actually observed). This is less disadvantageous than might appear, for it would allow (in Scyliorhinus) a distance of some 600 m to be covered if the fish swam continuously. Of course, the fast fiber system is normally used by the fish to swim for a few rapid tailbeats and then glide to rest or to slow sustained swimming using the slow fiber system. Rather few teleosts have focally innervated fast fibers such as the dogfish, the herring is the only one in which fiber function has been investigated directly (Bone et al., 1978a). By observing herring swimming in a tunnel respirometer at different water velocities it is a simple matter to show that muscle action potentials from the deep fibers are only obtained when the fish is swimming in rapid bursts, and that up to 5 Clsec, herring must utilize only the slow fibers, since action potentials are not found. It is important to realize that the mosaic arrangement in the herring fast fiber portion of the myotome does not represent a mixture of larger focally innervated fibers with small mul-

1

0 . 5 sec

Fig. 22. Records obtained from swimming spinal dogfish using concentric needle electrodes. Upper in each, electrical activity; lower, record of movement of fish. (A) Activity of slow myotomal fibers; (B) absence of activity from fast fibers during slow spinal swimming; (C) prolonged bursts of action potentials from fast fibers during movements evoked by pinching tail. Note that electrical activity in C is recorded at a lower amplification than in A. Time bar: 0.5 sec.

6. LOCOMOTOR MUSCLE

407

tiply innervated fibers. All are focally innervated; hence muscle action potentials must be a concomitant of their activity. A certain amount of evidence has accumulated from different lines of investigation that where fast fibers are multiply innervated (i.e., in higher teleosts), the simple duality of slow and fast portions of the myotome is an oversimplification. There is now rather convincing support for the idea that in these fish, fast fibers are active during continuous swimming at speeds well below that which the fish is capable of sustaining for long periods. Thus, Greer-Walker (1971) and Greer-Walker and Pull (1973) found hypertrophy of both slow fibers and fast fibers when coalfish (Gadus uirens) were swimming for long periods at 2 and 3 t'lsec. Using the same species, and also the crucian carp (C. carussius) Johnston and Goldspink (1973a,b,c) were able to demonstrate from measurements of muscle lactate, that after continuous swimming at 2 Clsec and above, the fast fiber system was active. Direct evidence for this conclusion was obtained in carp b y Johnston et al. (1977) and b y Bone et al. (1978a). A similar conclusion was drawn by Hunter (1971) in an interesting study of a fast swimming species (Trachurus s ymmetricus). Hudson (1973) placed electrodes in lateral superficial slow fibers and in deep mosaic fast fibers in trout ( S . gairdneri),and swam the fish in a tunnel respirometer at different speeds. He found that electrical activity was recorded from the superficial slow fibers at all swimming velocities, but the mosaic muscle was silent until the fish reached around 75% of the maximum sustainable swimming speed. At this swimming speed, the fast fibers showed electrical activity similar to that of the superficial slow fibers, but at burst speed, much larger electrical events were observed. Hudson's interesting results suggested that at intermediate and high cruising speeds, fibers in the fast portion of the myotome were operating without the production of muscle action potentials, whereas during bursts, the fast fibers produced action potentials. Similar conclusions were drawn from electromyography in carp (Bone et al., 1978a), except that in this species, the fast fiber portion of the myotome was active even at the lowest speeds the fish would swim in the respiroineter (Fig. 23). It is not known whether these different patterns of activity from the fast motor system during cruising and during bursts of rapid swimming represent the activity of different fast fiber types. Functionally, it would be a neat trick if the fish were able either to operate the fast motor system b y local contractions of the fibers so that the fibers contracted slowly when the fish swam slowly, or could contract them rapidly (with propagated muscle action potentials) during bursts of

408

QUENTIN BONE

Isec

Fig. 23. Similar records to those of Fig. 22 obtained from fast myotomal muscle fibers ofcarp swimming in tunnel respirometer at different water velocities. At low swimming speeds (upper two records), only low amplitude electrical activity is obtained, but at higher speeds, (lower records) large events appear, probably muscle action potentials recorded extracellularly. (From Bone et al., 1978a, Fish Bull., in press.)

speed. The same fiber could then operate at the appropriate point on different force/velocity curves in both situations for maximum efficiency. Until single fiber activity records are obtained, the question must remain open, but results such as those obtained by Barets (1961) on tench (Tinca) do not rule out the possibility of two kinds of electrical (and mechanical) activity from single muscle fibers. It is something of a paradox to find that it is apparently common in higher teleosts for muscle fibers specialized for short periods of anaerobic burst swimming to play a part in sustained long-term activity. During sustained swimming the requirement must be for efficient and economical utilization of metabolites, and anaerobic glycolysis yielding lactate is a relatively inefficient source of ATP (Bilinski, 1974; see also Chapter 8). Since little lactate is excreted during sustained swimming (Bilinski, 1974), an obvious solution to the paradox is that lactate is oxidized at various sites outside the fast motor system of the deep portion of the myotome. The fish would then be in overall oxygen balance, and no oxygen debt would have to be repaid after the period of swimming (as it has to be repaid after short bursts of swimming at high speed). Two alternative sites have been proposed as capable of complete oxidation of lactate: the gills and the superficial slow muscle fibers. The discovery b y Bilinski and Jonas (1972) that gill tissue in trout had a high capacity for lactate oxidation suggested that the energy re-

6.

LOCOMOTOR MUSCLE

409

quired for exchange processes there is at least in part supplied in this way, so that lactate produced by the operation of the fast fibers at maintained high cruising speeds is oxidized to supply energy to drive the ion pumps of the gills. It is known that trout gill oxidizes lactate, but whether this mechanism is used to keep the fish in overall oxygen balance during sustained higher speed cruising is not known. It would be interesting to compare the capacity of the gill tissue to oxidize lactate in focally innervated forms (e.g., herring) which do not utilize the fast motor system during the cruise condition, with those from fish which are known to use the fast motor system during cruising (e.g., gadoids). Various workers have suggested that lactate produced by the activity of fast fibers can be oxidized by the superficial slow fibers. Some, (e.g., Braekkan, 1956) have supposed that the superficial muscle fibers are indeed not concerned with thrust generation at all, acting solely as a sort of peripheral liver, accessory to the metabolism of the fast fibers. It is plain that this view is untenable, but some evidence suggests that the superficial (red) fibers may play a part in the glycolytic metabolism of the deep fast (white) fibers. I n a series of papers Wittenberger (earlier references in Wittenberger et al., 1975) has examined the metabolic interrelationships of the two muscle types after simple experimental procedures, and has concluded that the superficial fibers store glycogen for subsequent transfer to the fast fibers as well as oxidizing lactate derived from the fast fibers. Similarly, Smit and his colleagues (Smit et al., 1971), observing that goldfish were able to swim at sustained fast speeds (up to 8.5 flsec for over 3 hr) without incurring an oxygen debt, assumed that slow fibers oxidized the lactate produced b y the fast fibers, which must presumably have been active at these high speeds. As Bilinski (1974) emphasizes, further experimental evidence is needed before this concept of “cooperative metabolism” between the two main fiber types can be accepted. The idea is in some ways an attractive one; what is needed are not only more experimental biochemical data but also some simple physiological data about diffusion pathways and capillary exchange between the two zones of the myotome. The poor vascularization of the deep fast fibers, and their distance from the superficial slow fiber zone in most fishes, would seem to make cooperative metabolism a very long-term process except in fish such as scombroids or carangids, where the deep fibers are better vascularized. Interestingly enough, the study by Pritchard et al. (1971),taken with that by Hunter (1971),on the carangid Trachyurus, and that by Johnston et al. (1977) on carp suggested sustained fast

410

QUENTIN BONE

fiber activity over a wide speed range. If it is supposed that the mosaic portion of the myotomes in salmonids are composed of a mixture of slow and fast fibers (as, e.g., by Webb, 1971), then such an arrangement would certainly allow efficient exchange of metabolites between slow and fast fibers. Johnston (1977) has investigated the glycolytic enzyme profiles in slow and fast fibers of trout and mirror carp and has shown that these are similar in the carp, but differ in trout. As he points out, either there is something lacking in our understanding of anaerobic pathways in carp (but see Chapter 8) or there must be noncirculatory transfer of metabolites from white to red fibers which would require novel transport phenomena. The experimental myography, biochemistry, and histology considered in this section, have on the whole been at a relatively crude level, so that it has only been possible to consider the roles of slow or fast fibers. The diverse other fiber types have been conflated in one or other of these two categories, and until more detailed studies are carried out, the functions of such fibers as the superficial fibers in the dogfish are unknown. An early hint of different roles for two types of slow fibers was obtained by Rayner and Keenan’s (1967) electromyographic work on tuna where it was found that the superficial slow fibers and the deep, elevated temperature “chiai” slow fibers could operate under different conditions. Curiously enough, as Graham ( 1975)points out, the reason for maintainance of an elevated (slow) muscle temperature in larger scombroids and isurids is not entirely clear. There is no direct evidence for the function of the different fiber types in the fins of fishes, so far as the author is aware, but since there are two main fiber types in fin muscle, very similar in most respects to the slow and fast fiber types of the myotomal musculature, it is natural to suppose that they function in a similar way. Nishihara (1967) points out that the pectoral fin muscles in goldfish are mainly red, slow fibers, whereas the pelvic fin muscles are chiefly composed of white fast fibers, and relates this to the different function of the two sets of fins.

V. PROPRIOCEPTION Despite careful histological search by a number of workers, neuromuscular spindles have never been observed in the muscles of any fish. It seems extremely probable that they are indeed lacking, and that the few reports of their presence are mistaken. Either fish differ from other vertebrates in not requiring proprioceptors to regulate muscular contraction, or their proprioceptors are different to those of

6.

LOCOMOTOR MUSCLE

411

higher vertebrates. At present, except for elasmobranchs, there is insufficient evidence available to rule out the first possibility, but in some fish groups proprioceptors of different kinds are known, and on the whole it seems likely that they will eventually be found in all groups. I n two groups, the sharks and the rays, there is good histological and physiological evidence for proprioceptors associated with the locomotor musculature. In rays elongate beaded endings (Fig. 24B) among the fin ray muscles were found in the last century [they have been most recently investigated histologically by Barets (1956) and by Bone and Chubb (1975)l. Their position between muscle fibers, i.e., in parallel with the muscle fibers, naturally suggested that they were stretch receptors, and Fessard and Sand (1937) demonstrated that the static sensory discharge from the fin ray nerves was dependent upon the tension imposed upon the fin ray muscles. More recently, Ridge (1977) has investigated the dynamic properties of these endings, find-

Fig. 24. The three proprioceptive endings known from elasmobranchs. (A) Wunderer corpuscles from flank of body in Scyliorhinus. Note division of parent fiber at arrow and complex coiling within corpuscles. (From Bone and Chubb, 1976,J. Mar. Biol. Assoc. U.K. 56,925-928.) (B) Stretch receptive ending from among slow fibers of pectoral fin in ray. (C) Similar receptor from caudal myotome surface in Raia. In this case, the ending is apparently more closely associated with a muscle fiber than are the endings ofthe pectoral and pelvic fins. All from whole mounts of silver-impregnatedmaterial. Scale bar: 100pm.

4 12

QUENTIN BONE

ing that they resemble neuromuscular spindles in certain respects (Fig. 25). There are some interesting features in the morphology of these elongate endings. First, they are only found between smalldiameter multiply innervated muscle fibers which form the superficial zone of the fin ray muscle bundles. The main mass of the fin ray muscle, consisting of larger focally innervated muscle fibers, does not possess these endings. Second, the coupling between the sensory neurites and the muscle fibers themselves is relatively loose. For most of the length of the ending the neurites are only coupled to the muscle fiber b y loose strands of collagen, but at certain points, the sarcolemma is invaginated and the coupling is more direct. Because there is no capsule, and the endings are large and visible b y means of Nomarski optics in the living state, it seems likely that the ray endings may be of interest to physiologists examining general aspects of stretch receptor function. It is certainly significant that these endings lie only among small-diameter muscle fibers which probably do not propagate muscle action potentials. In this respect they are similar to the nonencapsulated stretch receptors of urodele muscle (Bone et al., 1976). On the surface of the caudal myotomes just internal to the dermal connective tissue sheet, there are beaded brushlike endings derived from large diameter nerve fibers that are similar to those found in the fin ray muscles (see Bone, 1964, for reference to earlier observations).

1

0.2sec

& 200

r

Fig. 25. Discharge frequency of stretch receptor endings from pectoral fin of Raia claoata in response to ramp stretches at different velocities. Lower records (solid lines), record of displacement ( 1 mm); upper record, discharge frequency (instantaneous frequency meter record). (A) Stretch at 1.25mm/sec; (B) stretch at 2.5,5,10, and 20 mm/sec. (From Ridge, unpublished.)

6. LOCOMOTOR MUSCLE

413

These endings lie superficial to the large-diameter terminally innervated fast muscle fibers of the caudal region, near to the insertions of the muscle fibers (Fig. 24C). They are therefore unlike the endings of the fin muscles (between slow fibers). Presumably, as the tail is flexed, these endings would be stimulated by pressure from the overlying connective tissue sheet, as well as, or alternatively, b y the tension exerted by the muscle fibers next to the ending; no physiological investigations have yet been made however. It should be pointed out that, although the elongate endings of rays are the best known, both histologically and physiologically, of fish proprioceptors, it is not yet clear what use the fish makes of the information which they provide. Since spinal rays do not show the same kind of reflex swimming as do sharks, the necessity of proprioceptive input for the swimming rhythm has not yet been demonstrated. However, having entered this caveat, it is obviously reasonable to suppose that (as Fessard and Sand suggested) the elongate receptors regulate the swimming rhythm. The function of the superficial endings of the tail is less easy to apprehend, since in most rays, the tail is relatively immobile. In sharks, these elongate receptors are either absent or very rare. Fessard and Sand observed similar responses to those given by the elongate endings of rays in certain (unspecified) muscles of Scyliorhinus; Barets (personal communication) has occasionally observed them among the fin muscles. They are not found in the myotomal musculature, so far as I have been able to observe. Instead, endings of rather different morphology are found. These endihgs lie just superficial to the myotomes at the level of the myosepta. They are derived from large-diameter axons which form coiled corpuscles embedded in the connective tissue of the outer edge of the myospetum. Occasionally they are found among the superficial muscle fibers in the partition between two muscle stacks. These endings were first carefully described by Wunderer (1908) from the bases of the fins, where they were later examined physiologically by Lowenstein ( 1956), who showed them to be slowly adopting mechanoreceptors. In morphology they resemble most the coiled corpuscular endings of higher forms (e.g., Munger, 1961) being formed of a twisted skein of neurites surrounded b y connective tissue elements. Roberts (1969a) recorded from portions of the body wall as it was flexed, and showed that receptors existed which were sensitive to the frequency and amplitude of flexure. In all probability these receptors are the coiled corpuscles shown in Fig. 24A, but this has not yet been definitely proven. If so, they function during swimming as second-

414

QUENTIN BONE

order proprioceptors, being stimulated b y the alternate compression and release of the endings as the fish oscillates its body, just as at the bases of the fins they signal the bending of the fins. Spinal sharks appropriately set up (i.e., ventilated and held free from the bottom of the tank) swim continuously with a regular slow rhythm. This rhythmic activity is dependent upon sensory input; surgical removal of input abolishes the spinal swimming rhythm (Lissmann, 1946). It is very probable that it is the alternating proprioceptive input of the corpuscular endings of the myoseptal margins which maintains this swimming rhythm. There is some controversy (Grillner, 1974) whether the spinal swimming is dependent upon central oscillators, but whatever view is taken of the central organization, it is clear that proprioceptive input is of importance. The occasional occurrence of the corpuscular endings between muscle fascia rather than superficial to the muscle fibers is of some interest, for in this position the endings (more elongate than usual) are presumably stimulated by pressure from contracting myotomal muscle fibers themselves rather than less directly; perhaps we have a hint here of the way in which proprioceptors linked to muscle fibers may have arisen from mechanoreceptors. During investigation of the corpuscular endings of sharks, Roberts (19694 found similar activity from the nerves passing to the body wall of the gurnard Trigla C U C U ~ U S(L,), but the endings responsible have not been examined histologically. On the whole it seems unlikely that these endings (or indeed, any possible proprioceptive endings of teleosts) can be morphologically very noteworthy. Many histologists have examined teleost muscles without discovery of assodiated corpuscular or beaded endings that could be proprioceptive, so that such endings are probably simple branching endings in the myosepta, as are found in the hagfish, Myxine (Bone, 1963). In the hagfish it was possible to recognize the endings as sensory, since the nerve fibers giving rise to them could be traced back to their cells of origin in the dorsal root ganglia, but this is rarely likely to be possible. Since spinal teleosts do not normally exhibit steady swimming rhythms of the shark kind, it has not been shown whether proprioceptive input is required during myotomal locomotion. A number of teleosts swim by means of the paired fins, as also d o holocephali, and I have examined both groups without observing special sensory terminations associated with the fin musculature. In Trigla, the bases of the free fin rays of the pectoral fins are innervated by branching fibers resembling the endings found in the joints of

6. LOCOMOTOR

MUSCLE

4 15

higher vertebrates. Similar endings are found between the joints of the fin rays of elasmobranchs, between the vertebrae and fin ray joints of dipnoi, (personal observations) and are probably found in most teleosts. They have not received any attention physiologically, so far as I am aware. Holmes (cited in Barker, 1974) has found complex branching endings superficial to the musculature in the mobile, barbell-like. fins of the dipnoan Protopterus; again, these endings are somewhat similar to the joint receptors of amphibia. Despite these rather scattered examples of proprioceptors associated with locomotor muscle fibers in different fishes, it seems to be correct to suppose that the innervation of any locomotor muscle in any fish is set apart from that of higher vertebrates, not only by the lack of spindles, but also b y the poverty of the sensory component. I n cat hindlimb muscles, for example (Barker, 1974), some 75% of all the axons in the nerves passing to the muscle are sensory; if fusimotor axons are included, around 81% of the axons in the nerves pass to or from receptors. This is indeed very different from the arrangement in fish. Even in the fin muscles of rays, the large diameter axons supplying the elongate sensory terminals make u p at most some 15% of the total muscular nerve supply. Partly, perhaps, this is because in fish, where the body weight is wholly or almost entirely supported by the water, postural problems are of little account (Bone, 1966), and so a rich sensory innervation giving a detailed pattern of information about muscle length and tension is not required. It is notable that the aquatic urodeles resemble fishes in that they are devoid of neuromuscular spindles in the myotomal locomotor musculature; they d o have, however, sensory endings in the limb muscles which are sensitive to stretch (Bone et al., 1976). The absence of neuromuscular spindles from fish muscle is striking to the physiologist accustomed to higher forms, but to the fish physiologist what seems remarkable is that only in rays are there proprioceptive endings which seem analogous to spindles in that they are directly associated with muscle fibers, rather than being “secondorder” proprioceptors as are all the other sensory endings assumed to be proprioceptive in function. It may be that the ray method of swimming demands very delicate control of the fin ray musculature, unobtainable b y indirect proprioception, but anyone who has watched the barbels of feeding mullet or the dorsal fins ofNotopterus or gymnotids, will be aware of remarkably precise muscular movements apparently without benefit of direct proprioception. It is probably in delicately controlled muscles such as these that morphologically specialized

416

QUENTIN BONE

proprioceptive endings will first be found in teleosts. Ono (personal communication) has found, however, only simple branched sensory endings in the barbels of Mullus. From the account above it will be at once evident that further investigations are required before the role of proprioception in the control of fish muscles is understood; only in elasmobranchs have we a reasonable understanding of the system.

VI. FISH MUSCLE AND THE MUSCLES OF HIGHER FORMS The preceding sections have indicated some of the differences between the muscles of fishes and those of terrestrial vertebrates. The differences are real, and they are differences not only of degree but also of kind. For example, neuromuscular spindles are absent, and they occur in all terrestrial vertebrates so far as is known; again, different fiber types are segregated or zoned to a much greater degree than they are in terrestrial forms. What is more, slow and fast fibers are more different from each other.than they are even in amphibia. Not only are slow fibers apparently non-twitch fibers, but the fast twitch fibers are normally highly specialized for anaerobic operation, which is to say, for maximum power at the expense of sustained operation. On the whole, these differences between the muscle fibers of fishes and of terrestrial forms can be understood in terms of the rather stringent conditions set by the density of the water in which the fish lives. Fortunately for fish, the density of the water which imposes a requirement for a large power increase for small increment of swimming speed, also provides the possibility of achieving neutral buoyancy by storing small amounts of gas or lipid, and, so, with the possibility of greatly increasing the mass of locomotor muscle. The myotomal mass of fast fibers only used occasionally during escape or predatory movements is but a light penalty for the fish to carry around, since it is buoyed up by the water; such an arrangement of a mass of muscles used only as an emergency power pack would be quite unsuitable for a terrestrial animal subject to gravity. Curiously enough, although all terrestrial vertebrates (with the exception of mammals) seem to have both multiply innervated slow fibers and focally innervated fast fibers in their locomotor muscles (see review by Barker, 1968), the function of the different fiber types is only known clearly in mammals. In amphibia, reptiles, and birds, it is possible that slow fibers are used for slow movements, perhaps additionally or alternatively, for isometric postural contractions, but

417

6. LOCOMOTOR MUSCLE

this awaits investigation. Part of the difficulty resides in the mosaic arrangement of fiber types in most muscles of terrestrial forms, as opposed to the zonal fish arrangement. It seems most appropriate to end this chapter b y reminding the reader of two points. First, the study of fish muscle is as yet in a preliminary state; so far, few fishes out of the many different groups have been examined. Much remains to be done. There are a variety of fish groups still living, from acrania to dipnoi, and comparative studies are likely to prove fruitful in interpreting the functional roles of different fiber types. Second, fish muscle is very suitable experimental material (as Bilinski has already noted) for a wide variety of problems.

REFERENCES Alexander, R. McN. (1969).The orientation of muscle fibers in the myomeres of fishes.J. Mar. Biol. Assoc. U . K . 49, 263-290. Alnaes, E., Jansen, J. K. S., and Rudjord, T. (1963).Spontaneous junctional activity offast and slow parietal muscle fibers of the hagfish. Acta Physiol. Scand. 60, 240-255. Andersen, P., Jensen, J. K. S., and Lpyning, Y.(1963).Slow and fast muscle fibers in the Atlantic hagfish (Myxine glutinosa). Acta Physiol. Scand. 57, 167-179. Arloing, S., and Lavocat, A. (1875). Recherches sur I’anatomie e t la physiologie des muscles stries piles et fonces. Mem. Acud. Sci. Belles Lett. Toulouse 7 , 177-194. Austin, J. L. (1962). “Sense and Sensibilia” (G. J. Warnock, reconstr.). Oxford Univ. Press, London and New York. Bainbridge, R. (1960). Speed and stamina in three fish. J. E x p . Biol. 37, 129-153. Bainbridge, R. (1962).Training, speed and stamina in trout. J. E x p . Biol. 39, 537-555. Barets, A. (1952). Diffbrences dans le mode d’innervation des diverses portions du muscle lateral et leur rapports avec la structure musculaire chez le Poisson-chat. (Ameiurus nebulosus Les.). Arch. Anat. Microsc. Morphol. E x p . 41, 305-331. Barets, A. (1956). Les recepteurs intra-musculaires des nageoires chez les selaciens. Arch. Anat. Microsc. Morphol. E x p . 45, 254-260. Barets, A. (1961).Contribution B I’btude des systemes moteurs lent et rapide du muscle lateral des te16ostCens.Arch. Anat. Morphol. Exp. 50, Suppl., 91-187. Barker, D. (1968). L’innervation motrice du muscle strie des vertbbres. Actual. Neurophysiol. 8, 23-71. Barker, D. (1974).The morphology of the muscle receptors. In “Handbook of Sensory Physiology” (C. C. Hunt, ed.), Vol. IIIA, pp. 1-190. Springer-Verlag, Berlin and New York. Bergman, R. A. (1964a). The structure of the dorsal fin musculature of the marink teleosts, Hippocampus hudsonius and H . zosterae. Bull. Johns Hopkins Hosp. 114, 325-343. Bergman, R. A. (1964b). Mechanical properties of the dorsal fin musculature of the marine teleost Hippocampus hudsonius. Bull. Johns Hopkins Hosp. 114,344-353. Best, A. C. G., and Bone, Q. (1973). The terminal neuromuscular junctions of lower chordates. Z. Zellforsch. Mikrosk. Anat. 143, 495-504. Bilinski, E. (1974). Biochemical aspects of fish swimming. In “Biochemical and Biophysical Perspectives in Marine Biology” (D. C. Malins and J. R. Sargent, eds.), Vol. 1, pp. 239-288. Academic Press, New York.

4 18

QUENTIN BONE

Bilinski, E., and Jonas, R. E. E. (1972).Oxidation of lactate to carbon dioxide by rainbow trout (Salmo gairdneri) tissues. J . Fish. Res. Board. Can. 29, 1467-1471. Blaxter, J. H. S., Wardle, C. S., and Roberts, B. L. (1971). Aspects of the circulatory physiology and muscle systems of deep-sea fish. J . Mar. Biol. Assoc. U . K . 51,9911006. Boddeke, R., Slijper, E. J., and van der Stelt, A. (1959). Histological characteristics ofthe body musculature of fishes in connection with their mode of life. Proc. K. Ned. Akad. Wet., Ser. C 62, 576-588. Bone, Q. (1963).Some observations upon the peripheral nervous system of the hagfish, Myxine glutinosa. J . Mar. Bio’l.Assoc. U . K . 43, 31-47. Bone, Q. (1964). Patterns of muscular innervation,in the lower chordates. Znt. Rev. Neurobiol. 6, 99-147. Bone, Q. (1966). On the function of the two types of myotomal muscle fiber in elasmobranch fish. J. Mar. Biol. Assoc. U . K . 46,321-349. Bone, Q. (1970). Muscular innervation and fish classification. Simp. Znt. Zoofil., 1st Unio. Salamanca pp. 369-377. Bone, Q. ( 1 9 7 2 ~ )The . dogfish neuromuscular junction: Dual innervation of vertebrate striated muscle fibers? J . Cell Sci. 10, 657-665. Bone, Q. (197213).Buoyancy and hydrodynamic functions of integument in the castor oil fish, Ruoettus pretiosus (Pisces : Gempylidae). Copeia No. 1, pp. 78-87. Bone, Q. (1975). Muscular and energetic aspects of fish swimming. In “Swimming and Flying in Nature” (T. Y.-T. Wu, C. J. Brokaw, and C. Brennen, eds.), Vol. 2, pp. 493-528. Plenum, New York. Bone, Q. (in press). Myotomal muscle fiber types in Scomber andKatsuwonus. In “The Physiological Ecology ofTunas” (G. Sharp and A. Dizon, eds.). Academic Press, New York. Bone, Q., and Chubb, A. D. (1975).The structure of stretch receptor endings in the fin muscles of rays. J. Mar. Biol. Assoc. U . K . 55, 939-943. Bone, Q., and Chubb, A. D. (1976). On the structure of corpuscular proprioceptive endings in sharks. J. Mar. Biol. Assoc. U . K . 56, 925-928. Bone, Q., and Chubb, A. D. (1978). The histochemical demonstration of myofibrillar ATPase in elasmobranch muscle. Histochem. J . (in press). Bone, Q., and Roberts, B. L. (1969).The density of elasmobranchs.J . Mar. Biol. Assoc. U . K . 49,913-937. Bone, Q., Ridge, R. M. A. P., and Ryan, K. P. (1976). Stretch receptors in urodele limb muscles. Cell Tissue Res. 165, 249-266. Bone, Q., Kiczniuk, J., and Jones, D. R. (1978a).On the role ofthe different fiber types in fish myotomes at intermediate swimming speeds. Fish. Bull. (in press) Bone, Q., Moore, M. H., and Ryan, K. P. (1978b). Myotomal muscle fiber types in dogfish. In preparation. Bostrom, S.-L., and Johansson, R. G. (1972).Enzyme activity patterns in white and red muscle of the eel (Anguilla anguilla) at different developmental stages. Cornp. Biochem. Physiol. 42B, 533-542. Braekkan, 0. R. (1956). Function of the red muscle in fish. Nature (London) 178,747748. Brotchi, J. (1968). Identification histoenzymoloqique des fibres lentes e t rapides dans les muscles squelettiques des v e r t k b r k Arch. Znt. Physiol. Biochim. 76, 299310. Brotchi, J. (1969). Identification histo-enzymologique des types de fibres musculaires strikes squelettiques chez la Roussette. C. R. SOC. Biol. 163, 1457-1458. Burke, R. E., Engel, W. K., Levine, D. N., Tsairis, P., and Zajac, F. E. (1971). Mammalian

6.

LOCOMOTOR MUSCLE

419

motor units: Physiological-histochemical correlation in three types in cat gastrocnemius. Science 174, 709-712. Buttkus, H. (1963).Red and white muscle of fish in relation to rigor mortis.1. Fish. Res. Board Can. 2 0 , 4 5 5 8 . Calow, L. J., and Alexander, R. Mc. N. (1973).A mechanical analysis of a hind leg of a frog (Rana temporaria).J. 2001. 171, 293-321. Carey, F. G., and Teal, J. M. (1966). Heat conservation in tuna fish muscle. Proc. Natl. Acad. Sci. U.S.A. 56, 1464-1469. Dahl, H. A., and Nicolaysen, K. (1971).Actomyosin ATPase activity in Atlantic hagfish muscles. Histochemie 28, 205-210. Davies, A. S. (1972). Postnatal changes in the histochemical fiber types of porcine skeletal musc1e.J. Anat. 113, 213-240. Denton, E. J., and Marshall, N. B. (1958).The buoyancy of bathypelagic fishes without a gas-filled swimbladder. /. Mar. Biol. Assoc. U . K . 37, 753-767. Fessard, A., and Sand, A. (1937). Stretch receptors in the muscles of fishes.J. E x p . Biol. 14, 383-404. Fierstine, H. L., and Walters, V. (1968). Studies in locomotion and anatomy of scombroid fishes. Mem. South. Calif. Acad. Sci. 6, 1-31. Flood, P. R. (1966).A peculiar mode of muscular innervation in amphioxus. Light and electron microscopic studies of the so-called ventral roots. J. Comp. Neurol. 126, 181-217. Flood, P. R. (1968). Structure of the segmental trunk muscle in amphioxus. With notes on the course and “endings” of the so-called ventral root fibers. 2. Zellfwsch. Mikrosk. Anat. 84,389-416. Flood, P. R. (1971).The three-dimensional structure and frequency of myosatellite cells in trunk muscle of the axolotl (Siredon mexicanus).J.Ultrastruct. Res. 36,523-524. Flood, P. R. (1977). The sarcoplasmic reticulum and associated plasma membrane of trunk muscle lamellae in Branchiostomu lanceotatum (Pallas).Cell Tissue Res. 181, 169-196. Flood, P. R., and Storm-Mathisen, J. (1962).A third type of muscle fibre in the parietal muscle of the Atlantic hagfish, Myxine glutinosa (L.)? 2. Zellforsch. Mikrosk. Anat. 58, 638-640. Franzini-Armstrong, C., and Porter, K. R. (1964). Sarcolemmal invaginations constituting the T system in fish muscle fibers./. Cell Biol. 22, 675-696. Gordon, M. S. (1968).Oxygen consumption of red and white muscles from tuna fishes. Science 159, 87-90. Graham, J. B. (1975). Heat exchange in the yellow tuna, Thunnus olbacares, and skipjack tuna, Katsuwonus pelamis, and the adaptive significance of elevated body temperatures in scombrid fishes. U.S. Fish Wildl. Sew., Fish. Bull. 73, 219-229. Greenwood, P. H., Rosen, D. E., Weitzman, S. H., and Myers, G. S. (1966). Phyletic studies of teleostean fishes, with a provisional classification of living forms. Bull. A m . Mus. Not. Hist. 131, 339-456. Greer-Walker, M. (1970).Growth and development of the skeletal muscle fibers of the cod (Gadus morhua L.).J. Cons., Cons. Perm. Int. Explor. Mer 33, 228-244. Greer-Walker, M. (1971).Effect of starvation and exercise on the skeletal muscle fibers of the cod (Cadus morhua L.).J. Cons., Cons. Perm. Znt. Explor. Mer 33,421-427. Greer-Walker, M., and Pull, G. A. (1973). Skeletal muscle function and sustained swimming speeds in the coalfish Gadus oirens L. Comp. Biochem. Physiol. A 44, 495-501. Greer-Walker, M., and Pull, G. A. (1975).A survey of red and white muscle in marine fish.J. Fish Biol. 7, 295-300.

420

QUENTIN BONE

Grillner, S. (1974). On the generation of locomotion in the spinal dogfish. Exp. Brain Res. 20, 459-470. Hagiwara, S., and Kidokoro, Y. (1971). Na and Ca components of action potential in amphioxus muscle cells. J . Physiol. (London) 219, 217-232. Hagiwara, S., and Takahashi, K. (1967). Resting and spike potentials of skeletal muscle fibers in salt-water elasmobranch and teleost fish. J. Physiol. (London) 190, 499518. Hagiwara, S., and Takahashi, K. (1974).Mechanism of ion permeation through the muscle fiber membrane of an elasmobranch fish, Taeniura lymma. J. Physiol. (London) 238, 109-128. Hagiwara, S., Henkart, M. P., and Kidokoro, Y. (1971). Excitation-concentration coupling in amphioxus muscle cel1s.J. Physiol. (London) 219, 233-251. Hamoir, G., Focant, B., and Distbche, M. (1972).Proteinic criteria of differentiation of white, cardiac and various red muscles in carp. Comp. Biochem. Physiol. B 41, 665-674. Harris, J. E., and Whiting, H. P. (1954).Structure and function in the locomotory system of the dogfish embryo. The myogenic stage of movement. 1. Exp. Biol. 31, 501524. Hidaka, T., and Toida, N. (1969). Biophysical and mechanical properties of red and white muscle fibers in fish./. Physiol. (London) 201, 49-59. Higashi, H., Kaneko, T., and Sugii, K. (1953). Studies on utilization of the liver oil of deep sea sharks-IV. Hydrocarbon contents in “Ynmezame,” Centroscymnus owstoni Garman. Nippon Suisun Gakkaishi 19, 836850. Hill, A. V. (1950). The dimensions of animals and their muscular dynamics. Sci. Prog. (London) 38,209-230. Hudson, R. C. L. (1968). A ringer solution for Cottus (teleost) fast muscle fibers. Comp. Biochem. Physiol. 25, 719-25. Hudson, R. C. L. (1969). Polyneuronal innervation of the fast muscles of the marine teleost Cottus scorpius L.J. E x p . Biol. 50, 47-67. Hudson, R. C. L. (1973). On the function of the white muscles in teleosts at intermediate swimming speeds.]. E x p . Biol. 58, 509-522. Hunter, J. R. (1971).Sustained speed of jack mackerel, Trachyurus symmetricus. U . S . Fish Wildl. Serv., Fish. Bull. 69, 267-271. Jarman, G. M. (1961).A note on the shape of fish myotomes. Symp. Zool. SOC. London 5, 33-35. Jasper, D. (1967). Body muscles of the lamprey. Some structural features of the T system and sarcolemma. J. Cell Biol. 32, 219-227. Johnston, I. A. (1977).A comparative study of glycolysis in red and white muscles of the trout (Salmo gairdneri) and mirror carp (Cyprinus carpio).J. Fish Biol. 11,575-588. Johnston, I. A., and Goldspink, G . (1973a). Some effects of prolonged starvation on the metabolism of the red and white myotomal muscles of the plaice Pleuronectes platessa. Mar. Biol. 19, 348-353. Johnston, I. A., and Goldspink, G. (1973b).A study of the swimming performance of the crucian carp Carassius cavassius (L.) in relation to the effects of exercise and recovery on biochemical changes in the myotomal muscles and 1iver.J. Fish Biol. 5, 249-260. Johnston, I. A., and Goldspink, G. ( 1 9 7 3 ~ )A. study of glycogen and lactate in the myotomal muscles and liver of the coalfish (Gadus uirens L.) during sustained swimming.J. Mar. Biol. Assoc. U.K. 53, 17-26. Johnston, I. A., and Tota, B. (1974). Myofibrillar ATPase in the various red and white

6.

LOCOMOTOR MUSCLE

42 1

trunk muscles of the tunny (Thunnus thynnus L.) and the tub gurnard (Trigla lucerna L.). Comp. Biochem. Physiol. A 49, 367-373. Johnston, I. A., Ward, P. S., and Goldspink, G. (1975). Studies on the swimming musculature of the rainbow trout. 1. Fiber types. J. Fish Biol. 7, 451-458. Johnston, I. A., Davison, W.,and Goldspink, G. (1977). Energy metabolism of carp swimming muscles. 1.Comp. Physiol. 114, 203-216. Kilarski, W. (1965).Organizacja siateczki sarkoplazmatycznej m i h i szkieletowych ryb. Czeii. 11. Okoli (Perca fluviatilis L.) [The organization of the sarcoplasmic reticulum in skeletal muscles of fishes. Part 11. The perch (Percafluoiatilis L.)J.Acta Biol. Cracov., Ser. Zool. 8, 51-57. Kilarski, W. (1967).The fine structure of striated muscle in teleosts. Z . Zellforsch. M i k rosk. Anat. 79, 562-580. Kordylewski, L. (1974). Some observations on mitochondria in muscle fibers of Salamandra salamandra (L.). Z . Mikrosk.-Anat. Forsch. 88, 937-947. Korneliussen, H. (1973).Dense-core vesicles in motor nerve terminals. Monoaminergic innervation of slow non-twitch muscle fibers in the Atlantic hagfish (Myxine glutinosa L.). Z . Zellforsch. Mikrosk. Anat. 140, 425-432. Korneliussen, H., and Nicolaysen, K. (1973). Ultrastructure of four types of striated muscle fibers in the Atlantic hagfish (Myxine glutinosa, L.). Z. Zellforsch. Mikrosk. Anat. 143, 273-290. Korneliussen, H., and Nicolaysen, K. (1975). Distribution and dimension of the T-system in different muscle fiber types in the Atlantic hagfish (Myxine glutinosa, L.). Cell Tissue Res. 157, 1-16. Korneliussen, H., Dahl, H. A., and Paulsen, J. E. (1978). Histochemical definition of muscle fibre types in the trunk musculature of a teleost fish (cod, Gadus morhua, L.). Histochemistry 55, 1-16. Kruger, P. (1950). Ueber das Vorkommen von zweirlei Fasem in der muskulatur von Haien. Z . Naturforsch. 56, 218-220. Kryvi, H. (1975). The structure of the myosatellite cells in axial muscles of the shark Galeus melastomus. Anat. Histol., Embryol. 147, 35-44. Kryvi, H. (1977).Ultrastructure of the different fibre types in axial muscles,of the sharks Etmopterus spinax and Galeus melastomus. Cell Tissue Res. 184, 287-300. Kryvi, H., and Eide, A. (1977). Morphometric and autoradiographic studies on the growth of red and white axial muscle fibres in the shark Etmopterus spinax. Anat. Embryol. 151, 17-28. Kryvi, H., and Totland, G . K. (19%).Fibre types in locomotory muscles of the cartilaginous fish Chimaera monstrosa. J . Fish Biol. 12, 257-265. Lansimaki, T. A. (1910). Ueber die Anordnung der Fibrillenbundel in den quergestreiften Muskeln einiger Fische. Anat. Hefte, Abt. 1 42, 251-279. Levin, A., and Wyman, J. (1927). The viscous elastic properties of muscle. Proc. R. SOC., Ser. B 101, 218-243. Lie, H. R. (1974). A quantitative identification of three muscle fiber types in the body muscles of Lampetra fluviatilis, and their relation to blood capillaries. Cell Tissue Res. 154, 109-119. Lissmann, H. W. (1946). The neurological basis of the locomotory rhythm in the spinal dogfish (Scyllium canicula, Acanthias oulgaris). 11. The effect of de-afferentation. J. E x p . Biol. 23, 162-176. Lorenzini, S. (1678).“Osservazioni intomo alle Torpedini.” Onofri, Florence. Lowenstein, 0. (1956). Pressure receptors in the fins of the dogfish Scyliwhinus canicula. J . E x p . Biol. 33, 417-421.

422

QUENTIN BONE

Magnuson, J. J. (1973).Comparative study of adaptations for continuous swimming and hydrostatic equilibrium of scombroid and xiphoid fishes. U.S. Fish Wildl. Serv., Fish. Bull. 71, 337-356. Mark, R. F., and Marotte, L. R. (1972).The mechanism of selective re-innervation of fish eye muscles. IV. Identification of repressed synapses. Brain Res. 46, 149-157. Modigh, M., and Tota, B. (1975). Mitochondria1 respiration in the ventricular myocardium and in the white and deep red myotomal muscles ofjuvenile tuna fish (Thunnus thynnus L.). Acta Physiol. Scand. 93, 289-294. Mosse, P. R. L. (1978).The distribution of capillaries in the somatic musculature of two vertebrate types with particular reference to teleost fish. Cell Tissue Res. 187,281303. Mosse, P. R. L., and Hudson, R. C. L. (1977):The functional role of different muscle fibre types identified in the myotomes of marine teleosts: A behavioural, anatomical and histochemical study. J . Fish B i d . 11, 417-430. Munger, B. L. (1961).Patterns of organization of peripheral sensory receptors. In “Principles of Receptor Physiology (W. R. Loewenstein, ed.), Handbook of Sensory Physiology, Vol. 1, pp. 523-556. Springer-Verlag, Berlin and New York. Nag, A. C. (1972).Ultrastructure and adenosine triphosphatase activity of red and white muscle fibers of the caudal region of a fish, Salmo gairdneri. J . Cell Biol. 55,42-57. Nag, A. C., and Nursall, J. R. (1972).Histogenesis ofwhite and red muscle fibers oftrunk muscles of a fish Salmo gairdneri. Cytobios 6, 227-246. Nakao, T. (1975).Fine structure of the myotendinous junction and “terminal coupling” in the skeletal muscle ofthe lamprey,Larnpetra japonica. Anat. Rec. 182,321-327. Nakao, T. (1976). An electron microscope study of the neuroniuscular junction in the myotomes of larval lamprey, Lampetra japonica. J . C o m p . Neurol. 165, 1-16. Nicolaysen, K. (19764. The spread of the action potential in the T-system in hagfish twitch muscle fibers. Acta Physiol. Scand. 96, 29-49. Nicolaysen, K. (197613).Spread of the junction potential in the T-system in hagfish slow muscle fibers. Acta Physiol. Scand. 96,50-57. Nishihara, H . (1967). Studies on the fine structure of red and white fin muscles of the fish (Carassius auratus).Arch. Histol. J p n . (Niigata,J p n . ) 28, 425-447. Nursall, J. R. (1956).The lateral musculature and the swimming of fish. Proc. Zool. S O C . London 126, 127-143. Patterson, S., and Goldspink, G , (1972).The fine structure of red and white myotomal muscle fibers of the coalfish (Gadus virens). 2. Zellforsch. Mikrosk. Anat. 133, 463-474. Patterson, S., Johnston, I. A,, and Goldspink, G. (1975). A histochemical study of the lateral muscles of five teleost species. J . Fish Biol. 7 , 159-166. Pecot-Dechavassine, M. (1961). Etude biochemique, pharmacologique e t histochemique des Cholinest6rase des muscles stribs chez les poissons, les batraciens et les mammiferes. Arch. Anat. Microsc. Morphol. E r p . 50, Suppl., 341-438. Pritchard, A. W., Hunter, J. R., and Lasker, R. (1971).The relation between exercise and biochemical changes in red and white muscle and liver in the jack mackerel, Trachurus symmetricus. U S .Fish Wildl. Serv., Fish. Bull. 69, 379-386. Ranvier, L. (1873).Propri6tes et structures differentes des muscles rouges et des muscles blancs chez les lapins e t chez les Raies. C . R . Acad. Sci. 77, 1030-1034. Rayner, M. D., and Keenan, M. J . (1967). Role of red and white muscles in the swimming of the skipjack tuna. Nature (London) 214, 392-393. Ridge, R. M. A. P. (1977).Physiological responses of stretch receptors in the pectoral fin of the ray Raia clavata. J . Mar. Biol. Assoc. U . K . 57, 535-541.

6.

LOCOMOTOR MUSCLE

423

Roberts, B. L. (1969a).T h e response of a proprioceptor to the undulating movements of dogfish.]. E x p . Biol. 51, 775-785. Roberts, B. L. (196913).The co-ordination of the rhythmical fin movements of dogfish.]. Mar. Biol. Assoc. U . K . 49, 357-425. Sharp, G.,and Dizon, A. (eds.)(in press). “The Physiological Ecology ofTunas.” Academic Press, New York. Skoglund, C. R. (1961). Functional analysis of swimbladder muscles engaged in sound production of the toadfish. J . Biophys. Biochem. Cytol. 10, 187-200. Smit, H., Amelink-Koutstall, J. M., Vijverberg, J . , and von Vaupel-Klein, J. C. (1971). Oxygen consumption and efficiency of swimming goldfish. Comp. Biochem. Physiol. A 39, 1-28. Smith, R. S., and Ovalle, W. K. (1973).Varieties of fast and slow extrafusal muscle fibers in amphibian hind limb muscles. J. Anat. 116, 1-24. Smith, R. S., Blinston, G., and Ovalle, W. K. (1974).Organization of skeletal muscle in the amphibia. In “Control of Posture and Locomotion” (R.B. Stein, K. G. Pearson, R. S. Smith, and J. B. Redford, eds.), pp. 1-13. Plenum, New York. Spray, T. L., Waugh, R. A., and Sommer, J. R. (1974). Peripheral couplings in adult vertebrate skeletal muscle. J. Cell Biol. 62, 223-227. Stanfield, P. R. (1972).Electrical properties of white and red muscle fibers of the elasmobranch fish Scyliorhinus canicula. J. Physiol. (London) 222, 161-186. Stevens, J. (1976).The ecology of the blue shark Prionuce glauca. Ph. D. thesis, Univ. of London. Syrovy, I., Gaspar-Godfroid, A,, and Hamoir, G. (1970). Comparative study of the myosins from red and white muscles of the carp. Arch. Int. Physiol. Biochim. 78, 9 19-934. Takeuchi, A. (1959). Muscular transmission of fish skeletal muscles investigated with intracellular microelectrode.]. Cell. Comp. Physiol. 54, 211-220. Teraviinen, H. (1971).Anatomical and physiological studies on muscles of lamprey. J. Neurophysiol. 34, 954-973. Teraviinen, H., and Rovainen, C. M. (1971).Fast and slow motoneurons to body muscle of the sea lamprey. J. Neurophysiol. 34, 990-998. van der Stelt, A. (1968). Spiermechanica e n myotoombouw Bij Vissen. Ph. D. Thesis, Univ. of Amsterdam. Vialleton, L. (1902). Le developpement des muscles rouges. C . R. Assoc. Anat. Part 4, pp. 47-53. Wardle, C. S. (1975). Limit of fish swimming speed. Nature (London) 255, 725-727. Warfel, H . E., and Clague, J. A. (1950). Shark fishing potentialities of the Philippine seas. Fish Wildl. Sero. (U.S.),Res. Rep. No. 15, pp. 1-19. Waterman, R. E. (1969). Development of the lateral musculature in the teleost Brachydanis serio: A fine structural study. Am. J. Anat. 125, 457-494. Webb, P. W. (1971).The swimming energetics of trout. I. Thrust and power output at cruising speeds. J. Exp. Biol. 55, 489-520. Webb, P. W. (1975). Hydrodynamics and energetics of fish propulsion. Bull., Fish. Res. Board Can. No. 190, 158 pp. Willemse, J . J . (1975).Some remarks on the structure and function of musculus lateralis in the European ee1,Anguilla anguilla (L.)(Pisces, Teleostei). 2. Morphol. Tiere 81, 195-208. Willemse, J. J., and van den Berg, P. G. (1978).Growth of striated muscle fibres in the M. lateralis of the European eel Anguillo unguillu (L.) (Pisces, Teleostei).]. Annt. 125. 447-460.

424

QUENTIN BONE

Winterbottom, R. (1974). A descriptive synonymy of the striated muscles of the Teleostei. Proc. Acad. Nut. Sci. Philadelphia 125, 225-317. Wittenberg, J. B. (1970). Myoglobin-facilitated oxygen diffusion: Role of myoglobin in oxygen entry into muscle. Physiol. Reu. 50, 559-636. Wittenberger, C., Coprean, D. C. and Morar, L. (1975). Studies on the carbohydrate metabolism of the lateral muscles in carp (influence of phloridzin, insulin and adrenaline).J . Comp. Physiol. 101, 161-172. Wunderer, H. (1908). Uber terminal korperchen der Anamnien. Arch. Mikrosc. Anat. Entwicklungsmech. 71, 504-569. Yamamoto, T. (1972).Electrical and mechanical properties of the red and white muscles in the silver carp.J. E x p . Biol. 57, 551-567.

7 THE RESPIRATORY AND CIRCULATORY SYSTEMS DURING EXERCISE DAVZD R . JONES and DAVZD J . RANDALL I. General Introduction . . . . . . . . . . . . . . . . . ................... 11. Assessment of Exercise Performance ............................. A. Introduction ..... ..................... ..... B. Th e Relation betw gen Metabolism and ................................... Swimming Speed, C. Anaerobic Contrib Exercise Metabolism . . . . . . . . . . . . . . . D. Limitations on Maximum Performance; Fatigue ............... 111. Th e Respiratory System during Exercise . . . . ........... A. Introduction ....................... ................ B. Respiratory Adjustments to Exercise .......................... IV. The Circulatory System during Exercise .......................... A. Introduction ...................... ............... B. Cardiac Adjustments to Exercise ............................. C. Arterial Blood Pressure and Total Peripheral Resistance during Exercise . . . . . . . . . . .......... D. Venous Pressure and Venous Return d References .................................... ...............

425 426 426 427 437 438 442 442 443 466 466 467 480 486 492

I. GENERAL INTRODUCTION Exercise is the stress which animals most frequently experience and may be defined as work performed on the environment b y the locomotory muscles. Exercise is accompanied b y an increase in the rate of energy conversion from the resting rate. This increase provides for the energy requirements of the locomotory muscles as well as for extra work performed by the heart and respiratory muscles in supplying oxygen demanded in exercise. In this chapter the emphasis is placed on the ability of fish to increase the rate of gas exchange at the gills and tissues and the changes which occur in the components of the respiratory and circulatory systems facilitating this increase in gas exchange. On many 425 FISH PIIYSIOLOCY, VOL VII Copyright @ 1978 by Ar.idemii Pre\<, Inc All right\ of reproduction 111 m y form re\eived ISBN 0-12-350407 4

426

DAVID R. JONES AND DAVID J. RANDALL

occasions circulatory and respiratory adjustments have been measured along with indirect estimates of metabolism during exercise. Obviously metabolism, power output, and swimming speed are related but, for virtually all fish species, the relationship is too imprecisely known to use any one in calculations of the others. Consequently, before discussing respiratory and circulatory compensations in exercise, we have reviewed the literature on the metabolic adjustments to exercise in an attempt to establish criteria which could be used in assessing the exercise performance of a particular fish. These data have been more extensively reviewed by Beamish (see Chapter 2) and Fry (1971) in this and an earlier volume of this series, respectively. Finally, a word is in order outlining the terminology used in this chapter. For metabolic rate, the terms standard, routine, and active refer to basal metabolism, metabolism associated and related to a degree of random activity, and the maximum sustained metabolic rate, respectively (Fry, 1971). Values for swimming speed are given in the Terminology to Describe Swimming Activity in Fish (p. xiii), while all physiological variables are described in the text in full.

11. ASSESSMENT O F EXERCISE PERFORMANCE A. Introduction Even the most cursory review of the literature leads to the conclusion that defining exercise performance in fishes is a difficult problem. In many experiments fish are forced to exercise up to a maximum sustained swimming speed which may seldom, if ever, be attained in nature. For instance, Brett (1972) suggests that the cardiovascular and respiratory systems in salmonids have evolved in response to the demands of upstream migration which are obviously very different from the demands placed on these systems in a fish being forced to swim as fast as it can in a water tunnel. Furthermore, disparate measures of exercise performance have been used by various investigators and it is necessary to know the relationship between them and metabolism since cardiovascular and respiratory compensations are best evaluated in terms of supply of metabolites. Therefore, if absolute or relative U has been used as a measure of performance, then the relationship between U and gas exchange for that particular species is required along with an estimate of whether anaerobic contributions to power output, and consequently U , are significant. Therefore, in this section a review of the measures of “effort” made by a fish during exercise is

7. T H E RESPIRATORY A N D CIRCULATORY SYSTEMS

427

made as background against which the respiratory and circulatory adjustments, discussed in the subsequent sections, can be evaluated.

B. The Relation between Oxygen Metabolism and Swimming Speed, U It is generally assumed that, in sustained exercise, respiratory and circulatory adjustments are adequate to meet increased energy demands aerobically. Unfortunately, there have not been many comprehensive surveys of oxygen metabolism in exercise and this lack of data has contributed to a controversy that surrounds not only the relationship between oxygen metabolism and U , but also whether oxygen metabolism is best described by absolute oxygen uptake or the difference between standard and active oxygen metabolism (metabolic scope) in exercise. The total amount of oxygen available for the locomotory muscles is generally assumed to be represented b y the difference between the active and standard rate of uptake, although it is recognized that this represents the maximum estimate of aerobic energy available, since it includes increased energy demands of the heart and respiratory muscles as well as increased costs of osmoregulation (Brett, 1963, 1972; Webb, 1971a,b). In view of this, Brett (1964) explored the relation between oxygen uptake and U and concluded that oxygen uptake increased exponentially with U in sockeye salmon (Oncorhynchus nerka) or, in other words, a linear relation was obtained when the logarithm of oxygen uptake was plotted against U , as determined in forced velocity tests using 75 min periods between velocity increments (Fig. 1). On the other hand, either a power or a linear function may describe the relationship between metabolic scope and U . Fry (1971) concluded, for a range of data, that the cost of swimming (as measured b y the scope) increased approximately as the square of U . However, the data of Brett (1965a; Fig. 2), Rao (1968; Fig. 2), and Kutty (1968a) for salmon ( 0 .nerka) and trout (Salmo gairdneri) yield the relation that cost increases as U1.35-1.8 and not U 2 as indicated b y Fry (1971).Alternatively, Smit et al. (1971) claim that since power output is related to U2.64and swimming efficiency to U1.64(approximately) then scope must be related to power output divided by efficiency or U'.O. Undoubtedly, Fry ( 1971) correctly assesses the situation when he states that performance is qualitatively different from the power which produces it and that there need not be any simple proportionality between measures taken of the two. Many teleosts do not support their weight, being neutrally buoyant, so the power output of the locomo-

428

DAVID R. JONES AND DAVID J. RANDALL

; “: :L/ J

Underyearling sockeye salmon (Brett. 1965a)

100

v

c

a,

0)

>, X

0 (m-1)

I , , , , , _ L

OO

U (cm I sec) Fig. 1. Relationship between oxygen uptake (ml/kg/min), scope for activity (ml/kg/ min) and cost to swim 1 m (ml/kg/m), and swimming speed ( U ) for underyearling sockeye salmon (0.nerka). (All data taken from Brett, 1965a.)

1000:

--

Rainbow trout

t’ 10

10

U (cmIsec)

100

Fig. 2. Relationship between oxygen uptake (ml/kg/min) and swimming speed ( U ) . For underyearling sockeye salmon and rainbow trout the slopes of the line are between 1.3 and 1.4, whereas for yearling sockeye salmon the slope is 1.8. (Data from Brett, 1965a; Rao, 1968.)

7.

429

THE RESPIRATORY AND CIRCULATORY SYSTEMS

tory muscles will be largely directed to overcoming the resistance to movement presented by the water. Since the drag on the fish will increase in proportion to U 2 [or, as Webb (1971a) suggested, U1.8for fish in tunnel respirometers] then a power relation between metabolic scope and U might be expected. This demands that the muscles, which produce the power to overcome the drag, b e equally efficient at all power outputs, which is not the case. Muscular efficiency is not constant but usually increases with increasing power output until high levels are reached, when it becomes increasingly difficult and more costly in terms of energy input to augment the work output. This process can be better appreciated with reference to Fig. 3 taken from the work of Hill (1950) which shows that power output of muscle under a maximum load depends on the shortening speed, maximum power output being reached at a shortening speed of about 35% of maximum. On the other hand, efficiency of muscle contraction reaches its peak when the shortening speed is only 20% of maximum. Consequently, as the animal increases power output it does so at the ex-

0

20

40

SHORTENING

80

60 SPEED

-

Oh

100

man

Fig. 3. The relationship between muscle power input, output, and efficiency as functions of muscle shortening speed, expressed in relation to maximum power, efficiency, and shortening speeds. [Redrawn from Hill, 1950, Sci. Prog. (London) 38,209.1

430

DAVID R. JONES AND DAVID J. RANDALL

pense of muscular efficiency, and an exponential relationship of the type obtained by Brett (1964) might be expected between oxygen uptake and U . Of course a corollary of an exponential relationship between oxygen uptake and speed is that the efficiency of swimming decreases with speed. Webb (1971b) calculated the efficiency of swimming in trout and showed that a 20 cm/sec speed increase in the middle velocity range (25-45 cm/sec) was accomplished with a doubling in efficiency while a similar increase in the higher velocity range (45-65 cm/sec) caused swimming efficiency to increase only by one-quarter. In view of the above discussion it is difficult to find any justification for expecting a linear relation between scope and U , as suggested by Smit et al. (1971), unless the data are restricted to determinations in the midrange of both scope and U . In other words, the active oxygen consumption should increase sufficiently so that the standard uptake only contributes from 20 to 30% of the total, but not to the degree that the cost of increasing the power output becomes prohibitive. In the absence of a clear relationship between U and oxygen uptake or metabolic scope for a particular species there are other measures which can be used to assess exercise performance. For instance, it seems plausible that the absolute rate of active oxygen metabolism could be a good indicator of the level of exercise stress to which a fish is subjected. However, this measure of exercise stress is not without problems. Brett (1964) found that active, like standard, oxygen uptake in sockeye salmon was temperature dependent (Table I). Qlo for both active and standard metabolism over the range of 5"-15"C lay between 1.7 and 2, while at 20" and 24"C, the active metabolic rate was not significantly different from the value at 15"C, although standard rates were significantly elevated. In other words, salmon appeared to reach a maximum oxygen uptake of about 625 ml O,/kg/hr at 15"C, under the conditions of Brett's experiments. Brett ( 1964) attributed the peaking of oxygen uptake at 15°C to a failure on the part of the respiratory system to supply enough oxygen at high temperatures and showed that, in one group of fish at 20°C, raising the oxygen concentration of the water by 50% caused active metabolism to reach over 900 ml O,/kg/hr, yielding a Qlobetween 15" and 20°C of 2. Brett (19654 and Brett and Glass (1973) also explored the relation between oxygen uptake during exercise and body size and found that the active level of oxygen uptake was virtually independent of body weight in sockeye salmon (Fig. 4), which is a considerable advantage to the investigator if active rates are used as a single measure of per-

7.

THE RESPIRATORY AND CIRCULATORY SYSTEMS

431

1000

L 5100 O0I

0 \ rN

/

0.1

[log Y = -0.63t 0.78log X I

50 100 500 1000 WEIGHT -9 Fig. 4. The relationship between metabolic rate and weight for different levels of activity expressed as fractions of the maximum 60-min sustained speed (max) in 0. nerka. Experiments performed in freshwater at 15°C. Broken lines represent possible relation of standard metabolism for immature freshwater stage and for mature fish of different sex. Whenever fish were tested singly limits of 2 2 SE are indicated. (From Brett, 1965a,J. Fish.Res. Bourd Can. 22, 1491.)

1

5

10

formance. Standard metabolism of sockeye salmon is, of course, proportional to weight to the power of 0.78, so standard metabolism, expressed on a unit weight basis, falls with weight increase (W-o.22)while active metabolism does not.

Table I

Some Measures that May Be Used to Assess Maximum Exercise Performance in Teleosts’

rp

w

M

Species

Oncorhynchus nerka

Change in Active oxygen oxygen uptake uptake (active/ (ml/kg/hr) standard)

Temperature (“C)

Weight (9)

Critical velocity (cm/sec)

5 10 15 20 24

36.7 32.9 55.2 62.6 52.2

53.7 58 77 74 68

360 439 626 596 594

12.5 10.4 12.6 7.1 4.32

0.17 0.19 0.2 0.19 0.18

Brett (1964)

15 15 15 15 15 15

3.38 3.47 19.1 55.2 746 1432

51.5 59.3 53.2 90.7 150 178

644 58 1 490 626 511 502

4 7.5 5.5 12.6 10.28 16

0.26 0.23 0.2’ 0.182 0.0853 0.073

Brett (196%) Untrained fish Speed quoted by Brett and Class (1973) Cost =-0.2 ml/kg/m from data in Brett (1965a) 3 Corrected speeds from Brett and Class (1973)

Cost to swim 1m

(ml/kg/m)

Notes and references

2

42

40

200

6.7

0.12

Brett and Glass (1973)

Lepomis gibbosus

20

45

37.2

285

9

0.19

Brett and Sutherland (1965)

Kuhlia sandoicensis

23

301

54.7' 54.32

407 458

8.8 9.3

0.18 0.21

Muir and Niimi (1972) Freshwater Saltwater Scaled to 30 g weight

Salmo sp.

15 15j 152

264 271 258

58.1 45.1 28.6

460 336 304

8.7 6.47 5.76

0.19 0.17 0.24

Webb (1971b) Loaded fish Group 1 * Loaded fish Group 3

Salmo gairdneri

15

100'

90

408

5.2

0.1

Rao (1968) Scaled to 100 g weight

Tilapia nilotica

25

801

60

320

4.4

0.11

Farmer and Beamish (1969) Freshwater I Scaled to 80 g weight

Melanogrammus aeglejinus

10

52

193

5.5

0.08

Tytler (1969); fish unfit

(rainbow trout)

rp

w

w

a

156

Active and standard rates of oxygen consumption are extrapolated.

434

DAVID R. JONES AND DAVID J . RANDALL

It is generally assumed that maximum oxygen uptake is achieved just before fatigue in an incremental velocity test which should give some consistency to this variable. However, Webb (1971a,b) has shown that if drag loads are placed on a swimming trout (S. gairdneri), then maximum oxygen uptake falls along with Ucrit(Table I). Webb (1971a,b) argued that the reason for this decline is that, although all fish are assumed to be making their maximum effort at Ucrit,the rate at which they work falls as Ucritfalls. Therefore the power output falls and hence maximum oxygen uptake. However, in terms of muscular effort expended, there would appear to be little difference between loaded and unloaded fish over the range of Ucrit’sWebb investigated. For instance, in the majority of Webb’s loaded fish both tailbeat frequency and specific amplitude (amplitude of tailbeat divided by wavelength) at Ucriteither equaled or exceeded those in the unloaded group. The calculated thrust produced b y an unloaded fish at Ucritwas equal, in a loaded fish, to the calculated thrust plus the drag of the added load at a lower Ucrit.There can be no doubt, however, that the oxygen uptake fell in loaded fish, so it must be concluded that, in loaded fish, the hydromechanics of the propulsive wave may not be adequately described by measurements of tailbeat frequency and amplitude. This effect has not been confirmed in other studies with salmonids. Kiceniuk and Jones (1977) found a poor relationship between the logarithm of oxygen uptake and absolute speed in trout (S. gairdneri) carrying various loads of instruments, whereas expressing U as a proportion of Ucrltgreatly improved the relationship. This observation suggests that fish attain their maximum oxygen uptake when making maximum effort regardless of absolute U . Furthermore, in Brett’s ( 1964) experiment, when active oxygen consumption increased markedly after environmental oxygen concentrations were raised, Ucritwas not significantly elevated, indicating a divorce between power output and oxygen uptake at maximum effort in this group of fish. Many fish are exercised under conditions that suggest there could be a restriction of power output, i.e., when swum to fatigue in a circular chamber when the muscles on one side may be held below their resting length, while on the other they may always be stretched above their resting length due to body curvature. Consequently, in view of the above conflicting data, it would be virtually impossible to predict how maximum aerobic metabolism under these conditions would compare with that obtained from fish swimming in a straight line. It is possible to obtain an immediate idea of the increased demands for oxygen and carbon dioxide transport during sustained swimming by expressing activity in terms of a ratio of active to standard

7.

435

THE RESPIRATORY AND CIRCULATORY SYSTEMS

metabolism. Brett (1964) found that this ratio was about 10-12 for 33-65 g sockeye salmon over a temperature range of 5”-15”C, although it fell markedly at higher temperatures (Table I) due to the temperature-induced increase in standard metabolism in the face of no increase in active metabolism with temperature. However, sockeye salmon at 20°C in high oxygen-saturated water showed a marked increase in maximum oxygen uptake and the ratio was restored to 10. Obviously this ratio, as an estimate of exercise performance, is sensitive to temperature. Furthermore, since standard metabolism relates to W0.78 while active metabolism is size independent, then the ratio is very sensitive to body size (Table I), increasing from 4 to 8 in small fish (3-4 g) to 16 in large fish (1.4-1.5 kg) (Fig. 4; Brett, 1965a). The outcome of sustained swimming in still water is covering a distance between two points and what is probably most important to the fish is how much energy it uses doing this. Consequently, it might be more realistic to express exercise capability in terms of the metabolic cost to traverse a unit distance. This has been done previously in an attempt to establish an “optimum” speed, that is, when metabolic cost per unit distance traveled is minimal (Fig. 5; Brett, 1965b; Weihs, 1973a,b; Webb, 1975a).The shape of the curve in Fig. 5 results from the fact that when swimming speeds are low, standard lo9

L\

6 0

I

I

10

1

1

I

I

20

30

SWIMMING

1

1

I

40 SPEED

-

I

I

I

I

1

50

60

70

1 80

cmls

Fig. 5. Relation between energetic costs (erg/km) and swimming speed of yearling sockeye salmon at three temperatures. Optimum swimming speeds occur at the lowest point of the curve. Predicted optimum speeds (Weihs, 1973a) are solid circles. (Based on data in Brett, 1964;after Brett, 1965b; by Webb, 1975a, Bull. Fish. Res. Board Can. No. 190.)

436

DAVID R. JONES AND DAVID J. RANDALL

metabolic costs are high, relative to costs of locomotion. Since speed is low then the time required to cover unit distance is long and therefore total cost per unit distance is high. Obviously cost per unit distance falls as oxygen uptake increases, but, since the latter has an exponential relation with U,cost rises again at the highest U’s. Weihs (19734 derived an expression to predict the optimum speed for yearling sockeye salmon, finding that it occurred at two times the standard metabolic rate (Fig. 5 ) . Since standard metabolism increases with temperature, then Weihs’ (19734 expression predicts that the optimum speed will increase with temperature, which fits well with Brett’s (1964) data. At optimum speed and low temperature, cost of locomotion in fish is low, requiring less than one-fifth the amount of energy that a running mammal or flying bird of the same weight would require to cover the same distance (Tucker, 1970; Schmidt-Nielsen, 1972). Assessment of exercise capability in terms of cost to traverse a unit distance assumes that anaerobic energy contributions are negligible. If there are no obvious discontinuities in the relation between oxygen uptake and sustained U , then probably only one group of muscles has been used throughout the exercise period. This appears to hold for trout (Kiceniuk and Jones, 1977) and salmon (Brett, 1964), although Brett and Sutherland (1965), Webb (1971a), and Hudson (1973) have indicated that at the lowest speeds these species may change their mode of propulsion. A change in the relation between oxygen uptake and U has been monitored in the goldfish (Carassius auratus) by Smit (1965),who found that around an imposed water velocity of 10 cm/sec goldfish seemed to switch to a “thriftier use of oxygen” (Fig. 6). Assessment of energy cost per unit distance is complicated by a change in the mode of propulsion or by a shift from red to white (or mosaic) muscle at a particular speed, for in that case, muscle recruitment would not involve the same type of muscle throughout the exercise. Finally, the ideal mode of progression suggested by Weihs (1974) of alternate bursts and glides is difficult to compare, on a basis of oxygen cost per unit distance, with sustained forward progression as is observed in swimming tests. Since there is an exponential relationship between oxygen uptake and U then, obviously, cost to cover a unit distance increases with speed; but in Table I cost is calculated using scope for activity (active minus standard metabolism) so an exponential increase in cost with U would not be expected. Furthermore, the faster the fish swims then the shorter the time it takes to cover a unit distance and any increase in the cost factor is offset somewhat. For instance, underyearling salmon (0.

7.

437

T H E RESPIRATORY AND CIRCULATORY SYSTEMS 0,cons (mg/kg/hr)

300 200

100 50

Goldfish 879

20 OC 10

‘

0

I

10 20 30 Swimming speed km/sec)

.

Fig. 6. Relationship between oxygen consumption and swimming speed of an 87-g goldfish (C. auratus) when spontaneously swimming in a nonrotating annular chamber (closed circles) and when forced to swim at various speeds in a rotating chamber (open circles). Throughout the experiments oxygen concentrations were at ambient. When forced to swim at 10 cm/sec the fish seemed to switch to a “thriftier use of oxygen.” (From Smit, 1965, Can. 1.Zool. 43, 623.)

nerka) monitored by Brett (1965a) displayed a 1.7 times inorease in cost to swim a meter (0.15 to 0.26 ml/kg/m) while speed increased 4 times (12.8 to 51.5 c d s e c ) (Fig. 1). Since the cost to swim a meter at maximum sustained swimming speed is temperature-independent it is a less variable measure of exercise capability compared with maximum or relative oxygen uptake (Table I). Unfortunately, “cost” is inversely proportional to weight so that low valu6s are obtained from larger fish (Table I; Brett, 1965a) so, as Table I confirms, “cost” is far too variable to use as a single measure of whether fish performed maximally or not before fatigue. Nevertheless, it represents an estimate which, allied to absolute or relative oxygen uptake, is useful for making an assessment of exercise performance (Table I).

C. Anaerobic Contribution to Exercise Metabolism Oxygen consumption, even in steady state conditions, may not be a good indication of the total energy conversion during exercise, since some of the energy budget may be provided by anaerobiosis. Consequently, during anaerobiosis, the relation between oxygen metabolism and U would alter markedly. At moderate swimming

438

DAVID R. JONES AND DAVID J. RANDALL

speeds depletion of muscle glycogen or build up of blood lactic acid is slight (Bla’ck et al., 1960, 1962; Connor et al., 1964; Beamish, 1968; Driedzic and Kiceniuk, 1976) and it is only during burst activity that both muscle and blood lactate levels increase rapidly and muscle glycogen falls (Black, 1958; Stevens and Black, 1966; Beamish, 1968), which is indicative of anaerobiosis. The portion of the total energy budget contributed anaerobically in sustained exercise is a matter of some dispute. Smit et al. (1971)claim that anaerobiosis contributes markedly to the energy budget during sustained swimming in goldfish (C. auratus),although Kutty’s (1968a) R.Q. determinations provide no evidence for anaerobic metabolism over a period from 3 to 12 hr after the start of sustained swimming at 20°C in goldfish (Carassius sp.) and trout (S. gairdneri), since the mean R.Q. in his fish was below unity. Both species, however, may display an initial anaerobic phase at the start of sustained swimming because mean R.Q. in the goldfish is 1.66 in the first hour of exercise (Kutty, 1968a). In this first hour aerobic metabolism is also higher than in the remaining 3-12 hr swimming period, and it is likely that the stress of handling and being forced to swim raises both aerobic and anaerobic metabolism. As swimming fish approach critical velocity, then the anaerobic energy contribution increases forcing the fish into a cumulative oxygen debt (Brett, 1964). This “debt” is repaid in the postexercise recovery period. Hence, in an increasing velocity test, even one conducted on aerobic fish such as trout, anaerobic metabolism would be expected to be present immediately following each increment, and also throughout the exercise period at speeds of around 80-100% of critical velocity (Fig. 7; Webb, 1971b; Driedzic and Kiceniuk, 1976).

D. Limitations on Maximum Performance; Fatigue Although all enforced exercise is terminated b y fatigue it is surprising that there appears to be no consensus about what precipitates swimming failure in fish. Brett (1964) suggested that “burst” swimming was terminated by exhaustion of cellular energy supplies while metabolite supply limited steady (or sustained) performance. Studies b y Black et al. (1962) and Stevens and Black (1966) indicate that in salmonids short-term exhausting exercise (being chased for up to 10 min) leads to severe depletion of muscle glycogen and marked elevation in muscle and blood lactate, whereas with moderate exercise muscle glycogen is hardly depleted at all (Black et al., 1960, 1962). Obviously “burst” swimming is largely anaerobic, independent of the

7.

T H E RESPIRATORY AND CIRCULATORY SYSTEMS

439

Fig. 7. (A) Diagram illustrating the proposed use of anaerobic energy sources in an increasing-velocity test. Four speeds are represented, corresponding to 40, 60, 80, and 100% of the critical swimming speed. The solid line represents the energy made available from aerobic energy sources. The shaded areas represent the energy made available continuously from anaerobic energy sources, and the solid shading represents the anaerobic energy requirements at a velocity increment. The total amount of anaerobic energy is distributed in proportion to the time for which the system operates and the rate at which energy is dissipated. This is proportional to U2.8 x t where t is 5 min at speeds less than 80% of UCfitand 60 min at speeds greater than this. At speeds greater than 80% U,,,,, the anaerobic energy contribution at a velocity increment is assumed negligible in comparison with the continual anaerobic energy contribution. (B) Proposed energy changes over a 5 min period after a velocity increment. Note: the scales are not the same for the aerobic and anaerobic contributions. (From Webb, 1971b,J. E x p . Biol. 55,521.)

roles of circulatory and respiratory systems, and muscle ‘lactate concentrations may be some 9-10 times higher than blood lactate levels (Stevens and Black, 1966). This suggests that the failure of muscle activity may be due to the fact that the lowest tolerable p H for anaerobic metabolism is attained rather than total depletion of cellular energy supplies. Unforttmately, even in increasing velocity tests, fish indulge in some “burst” type swimming, particularly when the imposed water velocity is increased or as Ucritis approached, which must complicate any metabolic analysis of fish swum to fatigue in this manner. Brett (1964, 1972) envisages fatigue occurring in increasing velocity tests due to energy demand exceeding the supply of oxygen at higher and higher speeds, so products of anaerobic metabolism gradually accumulate. However, changes in blood or tissue p H are unlikely to cause swimming failure in the absence of a profligate bout of burst swimming since lactate accumulation in increasing velocity tests is much less than occurs after violent chasing (Driedzic and Kiceniuk, 1976). The failure to supply sufficient oxygen to contracting muscle dur-

440

DAVID R. JONES AND DAVID J. RANDALL

ing exercise has been variously attributed to limitations in the rate of oxygen transport in the body or oxygen exchange at the tissues or gills. When intracellular oxygen tension falls below the critical level for aerobic metabolism, anaerobic energy production increases greatly, which precipitates fatigue. This is obviously a different process from the swimming failure described by Kutty (1968b) in trout and goldfish. Kutty (1968b) found that fish failed to swim when the oxygen tension in the water fell below a critical level, even if the fish could normally swim at that speed for many hours. When the oxygen tension was raised, swimming restarted and was maintained for hours until the critical oxygen level was again attained. Unfortunately, since raising the water oxygen tension took a finite time and, in view of the fact that even after fatigue at Ucritmany fish will swim again at low speed after a recovery period, these observations are not conclusive with respect to the nature of the fatigue. Nevertheless, both Smit et al. (1971)and Fry (1971) have suggested that failure to swim in hypoxic water is due to hypoxic depression of the central nervous system. Although a reduction in environmental oxygen below ambient levels reduces swimming speed or active metabolism in many fish (Basu, 1959; Davis et al., 1963; Silver et al., 1963; Dahlberg et al., 1968; Jones, 1971b), increases above ambient levels do not increase swimming speed (Brett, 1964; Dahlberg et al., 1968) and appear to have variable effects on active metabolism (Brett, 1964). Therefore Smit et al. (1971) suggest that oxygen uptake rate is not limited by oxygen extraction from the medium or transport within the fish but probably b y the aerobic capacity of the mitochondria or oxygen tension in muscle. Oxygen supply to the tissues depends on the provision of metabolic energy to power the branchial and cardiac muscles. Obviously, the costs of branchial and cardiac pumping increase as oxygen demand increases and ultimately an optimum level of oxygen uptake will be attained when supply to the tissues is maximal (Fig. 8; Jones, 1971a). Further increase in oxygen uptake requires more oxygen to be used by the heart and branchial muscles so that even less oxygen reaches the tissues, and the animal enters a “vicious circle” with regard to aerobic metabolism (Fig. 8; Jones, 1971a). Critical velocity of rainbow trout ( S . gairdneri) is reduced b y hypoxia or anemia, both of which might be regarded as restricting tissue oxygen supply b y increasing the oxygen demand of the branchial and cardiac pumps (Jones, 1971b). This is probably a somewhat simplistic interpretation of the effects of hypoxia and anemia but nevertheless these results are not at variance with the theoretical analysis (Jones, 1971a).

7. T H E RESPIRATORY AND CIRCULATORY SYSTEMS

44 1

Oxygen mode ovolloble-\ to the tissues

Resting oxygen uptake

0

200

400

600

I D,

Ventilotion volume (mllrnin)

Fig. 8. The amount of oxygen made available to the tissues in an exercising salmonid-type fish at 15°C. The maximum oxygen uptake is predicted to occur when oxygen supply to the tissues is optimal. The arrow marks the optimum oxygen uptake at 15°C and the boxes the amount of oxygen consumed by the cardiac ( 0 )a n d branchial (B) pumps. (From Jones, 1971a,J. Them. Biol. 32,341.)

There are also other metabolic costs, such as that for ionosmoregulation, which require some portion of the total aerobic energy available to the fish (Rao, 1968; Farmer and Beamish, 1969). However, whether the cost of ion-osmoregulation can be regarded in the same light as that of oxygen transport, that is, as a factor restricting tissue oxygen supply, is not clear. Certainly, the cost increases exponentially with U (Webb, 1975a), but so does the oxygen uptake and only if the cost of ion-osmoregulation increased at a substantially greater rate than the rate of oxygen uptake could this cost produce a point of optimum uptake where tissue supply was maximal. Support for this interpretation is provided b y data not only on rainbow trout (Rao, 1968), but also on Tilapia niloticu (Farmer and Beamish, 1969) and Kuhlia sandviceusis (Muir and Niimi, 1972), for in these species there is no reduction in maximum performance or active levels of oxygen uptake in fresh or salt water or iso-osmotic media.

442

DAVID R. JONES AND DAVID J. RANDALL

111. THE RESPIRATORY SYSTEM DURING EXERCISE

A. Introduction The ability to exchange gases is limited b y the surface area of the gas exchanger(s), A, the permeability of the surface to gases, k, the distance from the water to the blood, d, and the mean difference in oxygen partial pressure in the fluids on. either side of the exchange surface, A P g . Therefore, oxygen uptake

=

kA (APg ) d

The main purpose of this section is to discuss how changes in the various components of the above equation facilitate an increase in oxygen uptake during exercise. It is obvious that a fish's maximum oxygen uptake will ultimately depend upon morphometric rather than physiological limitations for diffusing capacity =

oxygen uptake kA APg - d

so some general relation between metabolism ( M ) ,A, and d is to be expected. Most commonly a relationship of this type is explored with respect to increase in body weight, W (Ultsch, 1973), rather than differences in exercise performance in animals of the same size. It is well established that M is not a linear function of W but rather a power function, W, where for standard M , b lies between 0.7 and 0.8. A similar relationship would be expected between A and W if d is weight independent. Hughes (1972) has presented data on both A and d for ten tench (Tinca tinca) varying in weight from 24.7 to 376 g and our analysis of the relation to W expressed in the allometric form is

A

=KWQ.72

d

= KWQ.14

so that diffusing capacity = KWQ.58. Consequently, as body weight increases in tench, the capacity of the gas exchanger will limit even the resting metabolic needs and will be unable to supply any extra oxygen for activity (Ultsch, 1973). However, for salmonids the relation ofA to W approaches linearity, viz.,

A

= KWQ.9'

7. THE RESPIRATORY

AND CIRCULATORY SYSTEMS

443

(derived from trout data given by Hughes, 1970; Hughes and Morgan, 1973). Hence, if d is relatively size-independent (interestingly d appears to be twice as large in trout as tench) then the gas exchanger will always be able to provide for resting metabolism regardless of the size of the fish, since in resting salmonids (Brett, 1965a) M = KWQ.78 On the other hand, this relation changes during exercise to

M

= KWQ.95

at maximum swimming speed (Brett, 1965a), which means that the gas exchanger is increasingly unable to cope with the active metabolic requirements of larger and larger fish. Hence, it can b e argued that the size limitation for tench, which are relatively inactive fish, is set by resting M , while for salmonids it is set by active M .

B. Respiratory Adjustments to Exercise 1. WATERFLOWOVER AND BLOODFLOWTHROUGH THE GILLS The amount of water flowing over the gills per unit time greatly increases during exercise (Saunders, 1962; Stevens and Randall, 1967a; Randall et al., 1967; Heath, 1973; Roberts, 1975a; Kiceniuk and Jones, 1977). In resting fish, gill ventilation is achieved by alternate contractions and expansions of a buccal force pump and an opercular suction pump (Shelton, 1970) and both the rate and amplitude of these breathing movements increase during exercise (Stevens and Randall, 1967a; Davis, 1968; Sutterlin, 1969; Webb, 1971a; Heath, 1973; Kiceniuk and Jones, 1977) (Fig. 9). In rainbow trout (S. gairdneri) breathing rate falls to 7l/min at intermediate U , from the resting rate of 83/min, although it rises to reach 146/min at maximum U (Webb, 1971a). However, Heath (1973) reported that the relative increase in breathing rate in trout is small compared to the marked increase in amplitude of both buccal and opercular breathing movements. The forward motion of the fish will augment the action of the buccal and opercular pumps in irrigating the gills. In fact, when the speed of the fish reaches the range of 50-80 cm/sec then the kinetic energy of the flowing water is equivalent to a pressure of between 1 and 3 cm H,O, which is adequate to provide a ventilation volume 10-15 times that obtained at rest. It is around this speed that many fishes cease making rhythmic breathing movements and rely on ram ventilation of the gills (Fig. 10).Among fishes which have been observed to change

,

,

5

U2-+

m

Us-. 15

m

Recoveryis

5

P

m

m

n

m.

Minuter

Fig. 9. The effects of various U ’ s (V,= 0.8tlsec; U , = 1.5t/sec; U , = 2.2t/sec) on the heart and ventilation rates of brown trout (S. trutta) at 8°C in a water tunnel. The trout were between 24 and 26 cm in length. 0 ,mean heart rate; 0, ventilation rate; dots and dashes, 2 1standarddeviation. (From Sutterlin, 1969,PhysioZ.Zool.42,36. Copyright 1969 by The University of Chicago Press.)

Fig. 10. The relationship between active gill ventilation rate and enforced U in five Atlantic mackerel (S. scombrus). When U reaches between 50 and 80 cm/sec the fish stop active gill ventilation and ram ventilate. [From Roberts, 1975a. BioZ. Bull. (WoodsHole, Mass.)148, 85.1

7.

T H E RESPIRATORY AND CIRCULATORY SYSTEMS

445

from rhythmic to ram ventilation in the 50-80 cm/sec velocity range are salmon (0.nerka) (Brett, 1964; Davis, 1968), remora (Remora remora) (Muir and Buckley, 1967), bluefish (Pomatomus saltatrix), Atlantic mackerel (Scomber scombrus), northern scup (Stenotomus crysops), and blue runner (Caranx crysos) (Roberts, 1975a) (Fig. 10). Some other fishes such as the skipjack tuna (Katsuwonus pelamis) and mackerel (S. scombrus) have largely lost the ability to pump water over the gills and are obligate ram ventilators (Hall, 1930; Magnuson, 1963; Brown and Muir, 1970). However, not all fast swimming fishes totally cease rhythmic breathing. Both rainbow trout and mullet (Mugil cephalus) often continue breathing in water tunnels at U’s which would appear to be adequate to maintain ventilation (Roberts, 1975b; Kiceniuk and Jones, 1977), yet resting rainbow trout will stop making breathing movements when force ventilated with volumes similar to those pumped by unrestrained resting fish (Jones and Sch w arzfe Id, 1974). The most graphic deponstration of the pressure differences along the body of a fish, due to its forward U , has been given by Dubois et al. (1974, 1976). In bluefish (P. saltatrix), swimming at 1.8 m/sec, there is a pressure difference of more than 20 cm H,O between the tip of mouth and the back of the operculum. However, small fish would never be able to attain the U necessary to generate sufficient upstream pressure to give adequate gill ventilation, while bottom feeders, with mouths at right angles or even directed away from the flowing water stieam, would gain no assistance in gill ventilation. Small Atlantic mackerel (2-12 cm) actively ventilate their gills (Roberts, 197%) which indicates that the capacity to ram ventilate is a consequence of the ability to swim fast, which only comes with increased size. Even in those fish that do not rely on ram ventilation, forward motion may augment gill ventilation. There is evidence that respiratory and fin movements are coordinated, although in many fish the relationship between these activities is quite labile (von Holst, 1937; Satchell, 1968). However, in Cymatogaster aggregata, a fish which swims solely by means of pectoral fin movements, ventilation and fin beat are well synchronized so that the mouth-open phase of breathing coincides with the maximum rate of forward motion (Webb, 1975b).A 1 : 1 ratio between ventilation and fin movement was always observed at rest and was unchanged during exercise although the ratio rose to 2 : 1 during hypoxia. Satchell (1968) recorded discharges from medullary reticulomotor neurons in Squalus acanthias in time with respiration and suggested that these neurons may constitute a basis for this type of coordination.

446

DAVID R. JONES AND DAVID J. RANDALL

Energetically there may be two effects of switching from rhythmic to ram ventilation. First, the drag characteristics of a swimming fish may b e improved since the opercular clefts will no longer rhythmically protrude into and consequently disturb the flow profile over the fish. Furthermore, since the gills make a significant contribution to the total drag of the body there will no longer be sudden increases and decreases in drag with mouth opening, although the effect on U may be minimized b y the coupling between locomotory movements and breathing so that maximum thrust coincides with maximum drag. Second, the work of gill ventilation will be transferred from the branchial to the swimming musculature and since the efficiency of contraction of the tail muscles is much higher than that of the branchial muscles (Webb, 1975a; Jones and Schwarzfeld, 1974), a saving in metabolic energy would b e indicated. However, the situation is complicated b y the fact that the branchial muscles are not relaxed but are tonically active during ram ventilation. The mouth of an anesthetized fish is forced closed or wide open if held in a water current, depending on the shape of the head and initial mouth gape (Roberts, 1975a),whereas mouth gape is adjusted with oxygen demand in swimming fish. Consequently, the net saving in energy for the fish is difficult to estimate. Jones and Schwarzfeld (1974) calculated that the cost of breathing in resting, restrained rainbow trout (Salmo gairdneri) was about 10% of the total oxygen uptake and that the efficiency of the process was about 1%. Hughes and Saunders (1970) estimated a similar cost of breathing for the same species at ventilation volumes of 300-500 ml/ kg/min, but at 1-2 liters/kg/min, during hypoxia, they estimated that oxygen cost of breathing could be as high as 50% of the total oxygen uptake. If 1% is taken as resting efficiency it is possible to construct a curve describing the efficiency of breathing with increases in work performed in breathing (Jones, 1971a). This relationship is shown in Fig. 11 and although it is a theoretical derivation the mean value of efficiency in trout (9.8%) given by Hughes and Saunders (1970), at what we calculate to be a 60-70 times increase in power output, fits it surprisingly well (Fig. 11). Hence, if efficiency increases to a maximum of 10% and then falls off once more as power output by the branchial muscles increases (Fig. 10) a portion of the increased cost of breathing at higher ventilation volumes will be offset by increased efficiency of breathing. It seems unlikely that power output b y the branchial muscles will ever increase b y more than 100 times in trout during sustained exercise so, due to the increase in efficiency, the oxygen cost of breathing would be the same proportion of the total oxygen uptake as it is at rest. In fact, at ventilation volumes reported

7.

T H E RESPIRATORY A N D CIRCULATORY SYSTEMS

447

0

Efficiency Branchial

of

Pumps

(percent)

10

1 Increase

in Work

100 of

1000

Breathing

Fig. 11. A theoretical derivation ofthe change in efficiency ofthe branchial pumps of a salmonid fish with the number of times work performed in breathing increases. The curve is similar in shape to that derived by Jones (1971a) although the resting value is taken as 1% and not 4%. 0 represents values for efficiency obtained by Jones and Schwarzfeld (1974),while 0 represents the mean efficiency value obtained by Hughes and Saunders (1970), at a power output of 62 times resting, during hypoxia. @I is the efficiency value and power output calculated from Table I1 of Hughes and Saunders (1970).

for exercising trout (Kiceniuk and Jones, 1977), applying Hughes and Saunders’ (1970)data on oxygen cost ofbreathing gives an oxygen cost which is always less than 10% of the total oxygen metabolism up to ucrit *

The gills of skipjack tuna (Katsuwonus pelamis) (an obligate ram ventilator) contribute about 9% of the total drag [as calculated by Webb (1975b) from Brown and Muir (1970)l and Webb (1975a) suggested that gill drag might be of the order of 12% in neutrally buoyant fish because drag associated with the pectoral fins makes a smaller contribution to the total drag of the fish. Thus, it seems reasonable to assume that as the gills contribute about 10% of the drag, about 10%of the total energy expended in forward motion will be used in ram ventilating the gills. But, as Brown and Muir (1970)point out, gill drag increases as the square of the velocity of respiratory water flow. Consequently, doubling ventilation volume in tuna, at a given U , elevates gill resistance to 27% of the total swimming resistance and cost of swimming goes up. Brown and Muir (1970) showed that mouth gape in mackerel exposed to hypoxia increases at a given swimming speed and so does oxygen consumption, reflecting the additional work being performed to overcome the extra gill drag. However, under normoxic conditions, increased ventilation volume can be achieved in step with U and since ram pressure head generation and total body drag have

448

DAVID R. JONES AND DAVID J. RANDALL

the same power relation to U then it might be expected that, as in rhythmic breathers, the oxygen cost of ram ventilation, as a proportion of total oxygen uptake, would remain close to resting values (less than 10% of total oxygen uptake). The control of breathing movements during exercise and the regulation of the switch from rhythmic breathing to ram ventilation are not understood. Breathing rate changes instantaneously with increased U (Fig. 9; Sutterlin, 1969) which indicates central nervous drive or reflex excitation of the medullary respiratory neurons by receptors detecting body movements. However, in ram ventilators there must be another set of receptors whose input inhibits the output of rhythmic medullary respiratory neurons when ram ventilating U’s are attained (Ballintijn and Roberts, 1976). Roberts (197513) has suggested that the receptors whose input inhibits breathing may be those flow-sensitive elements of the acousticolateralis system. Alternatively, mechanoreceptors on the gill arches (Sutterlin and Saunders, 1969) could monitor water flow velocity and shut down breathing at a predetermined speed. Other possible candidates for initiating apnea during ram ventilation are stretch-sensitive receptors in the skin (Roberts, 1969, 1972), whose input could presumably be integrated to indicate tailbeat frequency. In any event, there does not appear to be a “master” switch in the sense that either the animal ram ventilates or not. In some instances the transfer to ram ventilation can be a graded process, rhythmic cycles falling out for longer and longer periods when U increases slowly (Fig. 12; Roberts, 1975a). On the other hand, the latency of the “switch over” can be quite short, for when sudden velocity changes, from above to below ram ventilation speed, are imposed on swimming fish then breathing movements can be initiated within 0.2 sec. Chemoreceptors may play some role in regulating ram ventilation even though they do not appear to be involved in either the switch from rhythmic breathing to ram ventilation or in causing the initial increase in rhythmic breathing at the start of exercise. Ram ventilating fish can adjust gill water flow by changing mouth gape at a given swimming speed. Mackerel ( S . scombrus) increase mouth gape in response to increased oxygen demand after feeding and during hypoxia (Brown and Muir, 1970) but the nature and location of the receptors mediating this response are unknown. Gill ventilation volume has never been directly measured in swimming fish although measurements have been made on a variety of restrained fishes by modifications of the van Dam (1938) technique or dye dilution method (Millen et al., 1966). In exercising or unrestrained fishes, ventilation volume has been calculated by application

7.

T H E RESPIRATORY AND CIRCULATORY SYSTEMS

449

01

Fig. 12. Breathing movements, indicated by slow wave "muscle potentials" from the adductur mandibularis, and electrocardiogram of a bluefish (P.saltatrix) during acceleration to and above U sufficient to support ram ventilation. Breathing movements increase in frequency at low U's and slow again as ram ventilation ensues at higher U's, stopping at U's of 91 cm/sec. The time marks denote 1 sec intervals. [From Roberts, 1975a, Biol. Bull. (Woods Hole, Mass.) 148, 85.1

of the Fick principle to measurements of oxygen uptake and oxygen content of mixed inspired and expired water (Saunders, 1962; Stevens and Randall, 1967b; Kiceniuk and Jones, 1977). Saunders (1962) observed a decrease in oxygen extraction per unit volume of water (percentage utilization) with exercise in a number of fish species, and his calculated increases in ventilation volume for a given rise in oxygen uptake were higher than those of Stevens and Randall (1967b) and Kiceniuk and Jones (1977), who observed little change in percentqge 0, utilization in the trout (Salmo gairdneri) with exercise. However, Davis and Watters (1970) criticized the sampling method used by both Saunders (1962) and Stevens and Randall (1967b) stating that small water samples taken via a polyethylene tube inserted into the opercular cavity were not representative of mixed expired water. Kiceniuk and Jones (1977) modified the technique of Davis and Cameron

450

DAVID R. JONES AND DAVID J. RANDALL

0.4

1

200

. 'icfO0

20'00

VG (ml I Kg min)

Fig. 13. The relationship between oxygen consumption and ventilation volume in rainbow trout at rest and during exercise of increasing intensity up to U,,,.(After Kiceniuk and Jones, 1977,J:Exp. B i d . 69, 247.)

(1971), placing a rubber curtain over the operculum in order to create a collecting and mixing area for expired water, from which samples were taken as the fish swam. They observed that resting utilization of 33% was little changed in exercise and ventilation volume therefore increased linearly with oxygen uptake (Fig. 13).An increase in ventilation without a fall in percentage utilization was also recorded b y Davis and Cameron (1971) who measured gill water flow directly in resting fish. They observed no change from resting utilization of 40-45% as the nonswimming fish voluntarily increased ventilation volume from 44 to 120 ml/min. However, these changes were only one-tenth those recorded by Kiceniuk and Jones (1977). The percentage utilization of oxygen from water flowing over the gills is a measure of the amount of water actually involved in gas exchange (the respiratory volume) and an inverse measure of the size of the combined residual volume and water shunt. The latter has been characterized as a series of dead spaces (Randall, 1970b), namely:

1. Diffusion dead space, where water and blood remain in contact with the gill epithelium for too short a time for blood and water gas tensions to equilibrate 2. Distribution dead space, associated with poor matching between ventilation and blood perfusion so that too much or too little oxygen is delivered either to portions or the whole gill sieve than is required to saturate the blood 3. Anatomical dead space, where water takes a nonrespiratory path through the gill sieve, for instance, between the tips of adjacent filaments

7.

T H E RESPIRATORY AND CIRCULATORY SYSTEMS

451

Increases in ventilation volume during exercise will tend to increase the magnitude of thes diffusion dead space because of the reduced transit time for water (Randall, 1970b). If oxygen utilization remains constant as ventilation increases, however, then there must be a concomitant reduction in the magnitude of either the distribution or anatomical dead space. It is unlikely that the magnitude of the anatomical dead space will decrease at high ventilation rates; if anything, one might expect an increase in the anatomical dead space due to disruption of the gill sieve. Thus any increase in diffusion dead space, resulting from a reduction in transit time, is most probably offset by a reduction in distribution dead space in the gills during exercise which could be achieved b y changes in flow characteristics of water or blood at the level of the gill filament, primary, or even secondary lamella. The small leaflike secondary lamellae are the basic unit for gas transfer between blood and water (Fig. 14). Each secondary lamella

Fig. 14. Cast of the secondary lamellar vascular gpace dissected from a complete cast of the branchial vasculature. The black arrows indicate the probable paths of the erythro-

cytes. The direction of water flow (w) is shown by the white arrows. Blood enters through the afferent lamellar arteriole (ala) and leaves through the efferent lamellar arteriole (ela). The outer marginal channel (omc) and basal channel (bc) may serve as shunts or to improve the distribution of blood across the surface of the lamella (~410). (From Smith, 1976.)

452

DAVID R. JONES AND DAVID J. RANDALL

has a small afferent arteriole which widens into twenty or thirty channels through the body of the lamella, narrowing to a small arteriole connected to the efferent circulation (Hughes and Morgan, 1973; Vogel et aZ., 1973, 1976; Laurent and Dunel, 1976; Gannon et al., 1978a,b,c).The channels are formed by pillar cells which completely surround the blood spaces and function to hold the walls of the lamellae together, the latter being composed of a layer of epithelial cells. The thickness of the blood spaces within the secondary lamellae could conceivably be altered by contraction of the pillar cells (Hughes and Grimstone, 1965; Newstead, 1967; Bettex-Galland and Hughes, 1973) which could result in changes in flow distribution through the lamellae. Rankin (1976) has concluded, from direct visual observation on eels (AnguiZZa anguilla), that most of the blood flows through the somewhat larger peripheral vessels of the secondary lamellae (Fig. 14). A more even distribution of flow within the secondary lamellae during exercise would effectively increase the total surface area involved in gas exchange. Contraction of pillar cells would increase the area of the water channel correspondingly and water velocity past the lamellae would be reduced, favoring gas exchange, although diffusion distances in water would be increased. The secondary lamellae are arranged in a ladderlike series along the gill filaments and in trout the walls of the arterioles leading to the proximal lamellae are devoid of smooth muscle, and these lamellae are probably perfused with blood at all times. The arterioles at the inlet and outlet of more distal lamellae, however, have a thick smooth muscle coat (Gannon et aZ., 1978a)and it seems likely that perfusion of these lamellae with blood could b e regulated according to requirements of gas exchange. Booth and Holeton (1977) have shown in resting trout (S. gairdneri) that only about 60% of the lamellae are perfused. Intravascular injection of acetylcholine further reduced lamellar perfusion to 40%, while after adrenaline injection at least 95%of all secondary lamellae are perfused. Dilatation of distal lamellar arteries could occur during exercise due to adrenergic nerve stimulation or increased levels of circulating catecholamines, exciting adrenergic /+dilator receptors associated with the smooth muscle coats of these arteries (Smith, 1976). If lamellar recruitment occurs during exercise then the total gill area involved in gas transfer will increase and certainly change the magnitude of the distribution dead space. In resting fish, water flows over regions of the gill filament where there is little or no perfusion, creating a large distribution dead space. Perfusion of these regions of the filament during exercise could reduce the magnitude of the distri-

7.

THE RESPIRATORY AND CIRCULATORY SYSTEMS

453

bution dead space, which will tend to offset any increase in diffusion dead space associated with the increase in gill ventilation and maintain a constant oxygen utilization from water as it passes over the gills.

2. RESPIRATORY GAS TRANSPORT I N THE BLOOD AND PERIPHERAL BLOODFLOW Hemoglobin in arterial blood remains almost fully saturated with oxygen during exercise in salmonid fish. Kiceniuk and Jones (1977) found that oxygen saturation was 97% at rest in trout (Salmogairdneri) and varied between 96 and 100% during exercise up to 92% of critical velocity, Stevens and Randall (196713) observed no change in either arterial or venous oxygen tension during exercise; however, they used a different exercise pattern from Kiceniuk and Jones (1977), who observed a marked reduction in venous oxygen tension from 33 mm Hg at rest to 16 mm Hg during exercise. This was associated with a reduction in percentage saturation of hemoglobin in mixed venous blood from 70% at rest to 15% at exercise levels approaching the critical velocity for these fish. Thus the arterial-venous oxygen difference increased by a factor of between 2 and 3 during exercise in trout. Secondat (1950) also measured a reduction in venous oxygen content in carp (Cyprinus carpio) after exercise. Blood was collected by cardiac puncture and the mean venous oxygen content was 9.33 vol% (n = 6) before and 5.84 vol% (n = 6) after exercise. The reduction in venous content recorded in both carp and trout reflects increased tissue extraction of oxygen. The reduced venous oxygen tension will also increase oxygen differences across the gills and augment oxygen uptake. Cole and Miller (1973) have suggested that, for mammals, the product of heart rate, stroke volume, and venous oxygen content is constant during submaximal work. This is also true for fish swimming at up to 90% ofUcrit,whereas at maximal work outputs this product falls from values around 120 to 75 (Table 111; Kiceniuk and Jones, 1977). It may be that the constancy of the quantity of oxygen returning to the heart reflects, as Cole and Miller (1973) point out, a peripheral regulation of oxygen delivery whereby oxygen extraction or blood flow rates are adjusted to match increases in supply with demand. However, one requisite of such a regulatory mechanism would be a means of determining venous oxygen tension, or preferably, content. In fact, Taylor et al. (1968) from a computer simulation of the cardiovascular and respiratory responses to exercise in trout (Salmo gairdneri) concluded that these responses must be controlled by a venous oxygen sensor.

454

DAVID R. JONES AND DAVID J. RANDALL

It is clear that exercise in fish results in an increase in peripheral blood flow. associated with a reduction in the resistance to flow. Although there is an increase in the metabolic cost of circulation, gill ventilation, and osmoregulation during exercise, most of the increased oxygen uptake must be utilized by the muscles involved in locomotion. In mammals there is a marked shift in blood flow away from the gut and most of the increase in cardiac output is directed toward the exercising muscles. Stevens (1968), using radio-iodinated human serum albumin as a marker, could find no change in blood volume in any organ except the spleen (blood volume reduced) during exercise in trout. The method used, however, measured volumes and not flows. There might be marked changes in flow with only small changes in volume which would be undetected by this method. Clearly this question needs further examination. The fish heart consists of an outer cortex and an inner trabeculated layer (Satchell, 1971; Cameron, 1975). The outer cortical layer has a rich coronary supply and in some species consists of about 20-30% of the ventricular mass. The inner spongy layer, like the atria, has few or no capillaries and must rely on the blood flowing through the chambers for nourishment and gas exchange (Voboril and Schiebler, 1970; Oi'tadal and Schiebler, 1971; Santer and Cobb, 1972; Cameron, 1975). The ratio of cortex to inner layer mass changes with the species. Sluggish species such as the toadfish (Opsanus tau) have a very small cortical layer whereas the skipjack tuna (Katsuwonus pelamis) has a well-developed cortical layer and coronary circulation. Changes in coronary blood flow during exercise have not been measured, but coronary flow presumably increases to meet the increased requirements of the cortex during exercise. Based on data of Kiceniuk and Jones (1977), the oxygen uptake of the trout heart is calculated to be 3.5% of the total resting oxygen uptake, assuming that cardiac efficiency is 20% (Jones, 1971a) and that cardiac output multiplied by mean ventral aortic pressure is a measure of external work done b y the heart in fish. At maximum exercise in rainbow trout, assuming a drop in cardiac efficiency to 10% (Jones, 1971a),the heart consumes 4.5%of the total oxygen uptake. That is, at maximum exercise levels oxygen uptake by the heart in trout increases by a factor of 9.6 whereas total oxygen uptake of the fish increases only 7.8 times above the resting level. Cardiac output triples and the large increase in calculated oxygen uptake by the heart is due to the assumption that cardiac efficiency is halved at maximum exercise. It is important to stress that nothing is known about efficiency of contraction in

7.

THE RESPIRATORY AND CIRCULATORY SYSTEMS

455

fish hearts and the above assumptions concerning cardiac efficiency are based on observations on mammalian hearts. Around 70% of the ventricle of trout (Salmo gairdneri) is not supplied by coronary vessels but relies on blood flowing through the heart for nourishment. If one assumes that the inner spongy portion of the ventricle consumes 70% of the ventricular oxygen uptake then, as this oxygen comes from venous blood flowing through the ventricle, about 1% of the venous oxygen content will meet the ventricular requirements at rest. Venous blood flowing through the heart is 70% saturated with oxygen in resting trout (Salmo gairdneri). During exercise, however, the venous oxygen saturation falls to about 15% in rainbow trout (Kiceniuk and Jones, 1977) and 16% of the venous oxygen content must be removed to meet the requirements of the ventricle. I n fact the above calculations indicate that at 92% critical velocity in trout more than three times as much oxygen must be removed from a unit volume of blood flowing through the heart; the blood remains within the ventricle for one-third as long and contains only about one-fifth as much oxygen as venous blood in the resting fish. It is possible that the myocardium becomes partially anaerobic under these conditions. Hearts from different species show different tolerances to anoxia (Gesser and Poupa, 1974), and it is possible that the anaerobic component of cardiac metabolism is elevated during exercise in fish. Alternatively, the ventricular myocardium may be able to extract oxygen from the venous blood even at very low blood oxygen tensions and contents. Kiceniuk and Jones (1977) sampled blood before (cardinal sinus) or after (ventral aorta) it had passed through the heart. They observed no measurable differences in oxygen content between these two sites. This is to be expected in resting fish where only about 1% difference in content is predicted. A difference, however, should be apparent during maximum exercise. Unfortunately, at maximum exercise, Kiceniuk and Jones (1977) obtained only two samples from each site, and in no case were both sites sampled in a single fish. Their reported variability in venous oxygen content is in fact larger than the expected calculated difference in oxygen content in blood entering and leaving the ventricle. Thus this problem still needs further investigation. Stevens (1968) showed a reduction in splenic volume in the trout (Salmo gairdneri) during exercise. This organ has both adrenergic and cholinergic innervation and nerve stimulation or application of a-adrenergic agonists will cause splenic contraction and the release of erythrocytes in cod (Gadus morhua) and tench (Tinca tinca) and to a

456

DAVID R. JONES AND DAVID J. RANDALL

lesser extent in dogfish (Scyliorhinus canicula and Squalus acanthias) (Nilsson et al., 1975; Holmgren and Nilsson, 1975). The release of erythrocytes will cause an increase in circulating levels of hemoglobin. Hematocrit, hemoglobin, and plasma protein levels increase during exercise in the rainbow trout (Salmo gairdneri) in freshwater (Stevens, 1968; Wood and Randall, 1973a; Kiceniuk and Jones, 1977).At maximum activity the increase in hematocrit is 9-14% above the resting level. This hemoconcentration is due not only to erythrocyte release from the spleen but also to a reduction in plasma volume caused by a marked diuresis (Wood and Randall, 1973~). There are also longterm changes in hemoglobin levels with exercise. Hochachka (1961) found that rainbow trout made to swim in a current for 6 months had higher hemoglobin levels and larger hearts than untrained fish. Both short and long-term hemoconcentration presumably augments oxygen transport and facilitates acidhase regulation during exercise. Cameron and Davis (1970)have shown in rainbow trout that variations in blood oxygen capacity are compensated for by changes in cardiac output. They argue that a given species has evolved a particular blood oxygen capacity so that the heart operates over a favorable efficiency range. In some species hemoglobin levels may have been selected so that maximum efficiency of the heart (external heart work divided by the total energy expended by the heart) occurs during prolonged exercise, as, for instance, in migrating salmon. In these fish a small increase in hemoglobin levels may result in a reduction in cost of cardiac work and be of enormous survival value. Cameron and Davis (1970) reported that anemic rainbow trout could hardly sustain any swimming activity and Jones (1971b) found that a decrease in hematocrit from 32 to 11% almost halved maximum swimming velocity in rainbow trout at 21"-22"C. Cameron and Davis (1970) also suggest that in some fish, for example, the toadfish (Opsanus tau), maximum cardiac efficiency may span much lower levels of activity. The toadfish is generally sluggish and shows only short bursts of activity and may simply tolerate a large oxygen debt during these bursts. In this case we would predict no change in hemoglobin levels during activity. Blood volume of fishes lies in the range from 2 to 12% of the body weight (Thorson, 1959, 1960; Conte et al., 1963; Smith and Bell, 1964, 1967; Smith, 1966; Henimingsen and Douglas, 1970) with that of salmonids lying in the more restricted range of 4-10%. Resting cardiac outputs have been determined or estimated for a fairly large number of fishes but only in salmonids has cardiac output been determined

7.

T H E RESPIRATORY A N D CIRCULATORY SYSTEMS

457

during sustained exercise (Stevens and Randall, 1967a,b; Kiceniuk and Jones, 1977). In trout (Salmo gairdneri) resting cardiac output lies between 15 and 20 ml/kg/min for animals in the size range from 0.25 to 1.5 kg (Stevens and Randall, 1967b; Cameron and Davis, 1970; Kiceniuk and Jones, 1977). Hence one can expect long circulation times (blood volume divided by cardiac output) in these animals, ranging from 1 to 5 min. In fact Davis (1970) obtained an estimate for mean circulation time in resting rainbow trout (approximately 200 g) of 64.1 sec b y a dye injection technique, and similar circulation times were estimated b y Itazawa (1970) for carp (Cyprinus carpio) and eel (Anguilla japonica) of about 200 g. When cardiac output increases during exercise, circulation time will be reduced, but, even so, in large fish (>1 kg) it is unlikely that mean circulation time will ever fall below a minute. Exercise results in an increased production of CO, as well as oxygen uptake. The pattern of CO, excretion in resting fish is not well understood and even less is known about CO, excretion during exercise or its effect on oxygen transport. What is clear is that CO, can be transferred across the gills as either molecular CO, or bicarbonate ion and most of the excreted CO, comes from plasma bicarbonate. The transit time for blood flow through the gills is only a small fraction of that required for the uncatalyzed bicarbonate dehydration reaction, especially at temperatures of less than 20°C. Diamox (an inhibitor of carbcmic anhydrase) reduces CO, excretion (Haswell and Randall, 1978) and increases CO, tensions in arterial blood in both normal (Hoffert and Fromm, 1966) and anemic fish (Haswell and Randall, 1978). The erythrocytic carbonic anhydrase is unavailable for plasma bicarbonate dehydration (Haswell and Randall, 1976) but there are similar levels of carbonic anhydrase in the gill epithelium. Plasma bicarbonate enters the gill epithelium where it is dehydrated to CO, before diffusing into the water. Presumably the fish is able to regulate plasma bicarbonate b y controlling its movement through the gill epithelium (Randall et al., 1976) and thus adjusts the CO, :bicarbonate ratio and plasma pH. Haswell and Randall (1978) have proposed the following model for CO, excretion in fish (Fig. 15). a. CO, entering the blood from the tissues is rapidly dehydrated to form bicarbonate within the red blood cell (RBC). This reaction is catalyzed b y RBC carbonic anhydrase and drives oxygen from hemoglobin. b. The RBC swells due to water entry and, unlike mammals, there is no exchange of bicarbonate for chloride across the RBC membrane. c. Bicarbonate is formed at the uncatalyzed rate from CO, in

458

DAVID R. JONES AND DAVID J. RANDALL

Fig. 15. A diagram depicting the pathway of carbon dioxide excretion in fish, based on the description of Haswell and Randall (1978).

plasma; this reaction is slow and so bicarbonate is formed after blood has left the tissues and while it is in the veins, d. The blood in the veins is a closed system as far as gas transfer is concerned so, as plasma bicarbonate levels increase, CO, tension levels fall and RBC bicarbonate is dehydrated as CO, diffuses from the red cell into the plasma. Thus RBC bicarbonate formed while blood is in the tissue capillaries is dehydrated to CO,, diffuses into the plasma, and forms bicarbonate while blood resides in the veins. e. Blood entering the gills therefore has a higher plasma bicarbonate and a lower RBC CO, tension than blood leaving the tissues. The increase in RBC pH, resulting from the diffusion of CO, into the plasma, leads to an increased hemoglobin oxygen affinity, oxygen is bound to hemoglobin, and venous oxygen tension falls as blood flows from the tissues to the gills, augmenting oxygen transfer across the gills by decreasing venous oxygen levels. It is not known how the pattern of CO, excretion is affected by exercise but even during exercise residence times for blood in the veins is probably still sufficient for adequate rates of plasma CO, hydration at the uncatalyzed reaction velocity. Stevens and Randall (19674 observed increased venous CO, tension in trout, and Wood et al. (1977) measured a rise in total CO, as well as tension in venous blood after exhausting exercise in flounder (Platichthys stellatus). Obviously increased metabolic activity in the tissues results in in-

7.

459

THE RESPIRATORY AND CIRCULATORY SYSTEMS

creased CO, excretion to, as well as increased 0, extraction from, blood. The elevated venous CO, tension is undoubtedly a contributing factor, for a given fall in venous oxygen tension, in causing the marked reduction in venous hemoglobin oxygen saturation. Auvergnat and Secondat (1942) observed a reduction in plasma total CO, after exercise in the carp (Cyprinus carpio); they also measured a fall in blood pH. Wood et al. (1977) also recorded a fall in p H due to lactate accumulation during recovery from exercise in the flounder (Platichthys stellatus). Venous CO, content was elevated early in the recovery period but eventually fell below resting levels during recovery (Fig. 16) when the pH was reduced b y lactate accumulation. The acid conditions result in dehydration of bicarbonate

c

H

c :'.2[F

0

I

I

I

I

I

I

I

1

I

I

I

' I CI

I

I

I

I

I

I

1 1 c1

*:

I

*.

/J

l

0

e

2 4 Time (hrs.)

l

b

6

24

Fig. 16. Carbon dioxide tension ( P o , ) , calculated from the Henderson-Hasselbach equation, total CO, concentration, pH, and lactate concentration in the venous blood of two flounders ( 0 = 800 g; 0 = 450 g) before and after 10 min of exhausting activity (bar). I = initial resting sample. Time 0 = immediately postexercise. (Modified from Wood et al., 1977,j. E x p . Biol. 69, 173.)

460

DAVID R. JONES AND DAVID J. RANDALL

and a consequent rise in blood CO, tension. This increases the CO, tension difference between blood and water, enhances the excretion of CO, across the gills, and results in a reduction in plasma CO, content. Auvergnat and Secondat (1942) presumably missed the initial elevation in total CO, in blood observed by Wood et al. (1977) and reported only the secondary reduction in plasma total CO, after exercise. Lactate is produced in large quantities in white muscle fibers during violent exercise; it enters the blood slowly, however, and the fish is usually exhausted before there is aqy appreciable rise in blood lactate. Maximum levels of blood lactate occur 2-3 hr after the end of the exercise period (Blacket al., 1962). During exercise there may be a small acidosis related to elevated CO, levels, but reductions in p H due to lactate production do not appear until after the exercise period has ended. As Wood et al. (1977) have pointed out, the separation of the acidosis due to CO, and that due to lactate reduces the magnitude of the p H drop during exhausting activity, an important fact when one bears in mind the low buffering capacity of fish blood compared with that of mammals (Albers, 1970).The acidosis during exercise may also b e ameliorated b y the retention or uptake of H+ in muscle tissue, as occurs during hypercapnia (Randall et aZ., 1976), or b y the movement of H+, HC0,-, or OH- ions across the gills (Maetz, 1973; Randall e t al.,

1976). A decrease in RBC p H results in a fall in hemoglobin oxygen capacity (Root shift) as well as hemoglobin oxygen affinity (Bohr shift). A sharp rise in lactate levels might affect arterial oxygen saturation via the Root effect, but trout (SaZmo gairdneri) red cells are relatively impermeable to lactate (Randall et al., 1978).There is a detectable loss of 14C-labeled lactate from plasma, presumably into fish RBC’s, in uitro at 11°C 8 hr after elevation of plasma lactate, but equilibrium is still not complete after 24 hr of exposure. Furthermore, trout RBC’s have some capacity to regulate intracellular p H (Randall et al., 1978) so a rise in blood lactate may have little effect on oxygen transfer.

3. VENTILATION:PERFUSIONRELATIONSHIPS Both blood flow through and water flow over the gills are pulsatile. During exercise the oscillations in water flow will be reduced as ram ventilation becomes more important at higher swimming speeds but blood flow pulsatility will increase if stroke volume increases markedly. Jones et al. (1974)have argued that pulsatile blood flow will only have a major effect on gas transfer if stroke volume is larger than gill volume. Under these circumstances the transit time for blood through

7. T H E RESPIRATORY AND CIRCULATORY SYSTEMS

46 1

the secondary Iamellae at peak flow rates will be much less than that at minimum flow rates at the end of ventricular diastole. However, if stroke volume is less than lamellar blood volume, all blood will spend at least the equivalent time of one cardiac cycle during passage through the gills. Jones et al. (1974) reviewed estimates of gill blood volume and concluded that stroke volume is generally less than gill volume. However, from examination of vascular casts of trout gills ( S . gairdneri), Gannon et al. (1978a) obtained values for total lamellar blood volume of about 0.7 ml/kg. If only 60% of these lamellae are perfused at rest (Booth and Holeton, 1977) then functional lamellar volume in resting trout is 0.42 ml, expanding to 0.7 ml when all lamellae are perfused during exercise. Kiceniuk and Jones (1977) estimated that stroke volume in a l-kg rainbow trout increased from 0.46 m1 to 1.03 ml between rest and maximum exercise, so at rest or in exercise blood will only reside in the gills for one heart beat. In the resting sea raven (Hemitripterus americanus) blood takes four to seven heart beats to pass from the ventral to dorsal aorta (Stevens and Sutterlin, 1975). The gill lamellae in trout represent one-quarter (100% of lamelIae perfused) to one-sixth (60%of lamellae perfused) of the total gill blood volume and if similar rqlationships exist in the sea raven then, as in trout, there appears to be an approximately one-to-one relationship between gill lamellar volume and stroke volume of the heart. The ventilation/perfusion (VglQ) ratio in fish is usually between 10 and 15 and reflects the oxygen content of the two media, blood and water (Rahn, 1966; Piiper and Baumgarten-Schumann, 1968; Holeton and Randall, 1967b; Cameron and Davis, 1970; Cameron et al., 1971; Jones et al., 1970; Kiceniuk and Jones, 1977). Thus, in fish, about ten times more water flows over the gills than blood perfuses them per unit time. Stevens and Randall (19674 calculated a much higher VglQ ratio for trout (Salmo gairdneri), but this has not been substantiated b y other workers and probably reflects either special experimental conditions (a water tunnel) or measurement inaccuracies. There are probably differences in VglQ ratios between species related to differences in blood hemoglobin oxygen affinity and capacity, the area of the gills, and venous oxygen levels (Jones et al., 1970). Kiceniuk and Jones (1977) found that exercise in rainbow trout resulted in an increase in the Vg/Q ratio from 12 to 32 (Table 11). This increase in the VglQ ratio was the result of a much greater increase in Vg than Q during exercise. Ventilation volume in rainbow trout increased from 211 to 1700 ml/min/kg between rest and swimming speeds of up to 92% Ucrlt,that is, by a factor of 8. This must reduce the transit time for water flow through the gills and yet oxygen utilization

462

DAVID R. JONES AND DAVID J. RANDALL

from water did not change (Table 11). At the same time cardiac output increased by a factor of over 3, decreasing residence time in the gills, and yet arterial blood remained fully saturated with oxygen. Although lamellar recruitment or a more even distribution of flow within a single secondary lamella would promote oxygen transfer, these effects will be offset by the decreased residence time for blood in the secondary lamellae. However, since cardiac output increases by a factor of 3, blood residence time in the secondary lamellae only decreases to one-half the resting value, and not one-third, due to lamellar recruitment. Consequently, since water flow increases by a factor of 8, there is a much more marked increase in the ventilation : perfusion ratio at the secondary lamellar level in terms of the flow velocities of the two media, than is apparent from analysis of the ratio of cardiac output and total gill water flow. The carbon monoxide diffusing capacity of the gills can be used to indicate the ability of the gills to exchange oxygen. Fisheret al. (1969) concluded, from experiments in which carbon monoxide diffusing capacity was measured in catfish (Ameiurus nebulosus), that oxygen transport across the gills is primarily diffusion limited at ambient water oxygen levels. They observed that CO diffusing capacity doubled during hypoxia which they attributed to changes in conditions for diffusion between water and blood rather than diffusion and/or reaction velocity limitations within water or blood. The increase in diffusing capacity during hypoxia could have been related to lamellar recruitment, changing patterns of water and blood flow affecting the gill area available for gas transfer, or a change in the diffusion distance between water and blood. The catfish were “excited and exercising” so it is possible that all lamellae were perfused at all oxygen levels, and changing conditions for diffusion within each lamellae accounted for the increased CO diffusing capacity during hypoxia. Fisher et al. (1969)reported that the gill oxygen diffusing capacity changed in the same way as carbon monoxide diffusing capacity during hypoxia, indicating that increases in oxygen diffusing capacity are also related to changes in conditions for diffusion across the gill epithelium. Another measure of the capacity of the gills to exchange oxygen is given by the transfer factor (To*)which is defined as oxygen uptake divided b y mean oxygen difference across the gills (APg) between water and blood (Randall et al., 1967). Randall et al. (1967) and Kiceniuk and Jones (1977)both observed a five- to sixfold increase in gill oxygen transfer factor during exercise in trout (Salmo gairdneri) (Table 11),presumably reflecting either an increase in area available for diffusion or a decrease in the distance for diffusion between blood

Table I1 Oxygen Transfer Factor (TO,) and Ventilation: Perfusion Ratio ( V g @ ) for the Whole Gill in Resting and Exercising Rainbow Trout Compared with Values Obtained by Assuming that Only 60% of Secondary Lamellae Are Perfused at Rest and 100% at Maximum U" Mean oxygen difference across gill Apg

Mean oxygen difference Exercise Mixed Inspired Exhaled venous water water level Arterial Oxygen [HPh + across perfused (U expressed oxygen oxygen oxygen oxygen consumpPb, tension tension tension as a proportension tion MP, - lamellae AP~arn tion of maxiP+, PG". Ph Pb, 0% P,J1 mum U ) (mm Hg) (mm Hg) (mm Hg) (mm Hg) (ml/kg/min) (mm Hg) (mm Hg) Resting

137 70-7870 Uwt 123 128 81-91% Urnit 92100% U ~ t 126 a

33 23 29 16

152 155 146 151

Data from Kiceniuk and Jones (1977). Anatomical dead space assumed to be zero.

102 102 102 102

0.56 1.9 3.12 4.34

42 53 48

-

25

56

56

-

Oxygen transfer factor (ml/min/mm Hg) TO,

T&

Whole gill

Perfused lamellae

Ventilation perfusion ratio VgiQ

Whole

~) N,,JAP~)( V ~ A P , ~ gill 0.013 0.036 0.061 0.078

0.022 -

-

0.078

12 21 30 32

Perfused IamelIae

7.2 -

32

464

DAVID R. JONES AND DAVID J. RANDALL

and water, since there was only a small increase in the mean oxygen difference across the gills (APg) (Randall et al., 1967; Kiceniuk and Jones, 1977). APg increased from 42 mm Hg at rest to 56 mm Hg during exercise (Table 11)which alone could not account for the sevenfold increase in oxygen uptake during exercise. However, it may be that resting APg is not representative of 0, differences across perfused lamellae, particularly if lamellar recruitment occurs during exercise. If only 60% of the lamellae are perfused at rest (Booth and Holeton, 1977), then assuming that (a) water is evenly distributed over the gill sieve and this distribution does not change with exercise, and (b) the anatomical dead space is constant and in the absence of any data is set at zero, then water passing over perfused lamellae will mix with water which has passed over unperfused lamellae to give a final mixed expired water 0, tension of 102 mm Hg. The water passing over unperfused lamellae (40%of total flow) will have an 0, tension the same as the inhaled 0, tension, which in the experiments of Kiceniuk and Jones (1977) was 152 mm Hg. Thus, the 0, tension in water that has passed over perfused lamellae (60%of total flow) is calculated to have been 67 mm Hg. The mean oxygen difference between water and blood across perfused lamellae (APlam)is 25 mm Hg, much lower than that determined for the whole gill in resting fish (Table 11). If we assume that a+ swimming speeds of 92%Ueritall lamellae are perfused, then APLam will equal APg, which is 56 mm Hg (Table 11). Thus, at the level of the perfused lamellae the mean oxygen difference between blood and water doubles between rest and exercise due to both a decrease in venous 0, tension and a rise in exhaled water 0, tension, and T&, calculated using AP,,, rather than APg, only increases 3.5 times between rest and exercise up to 92% of Ucrit(Table 11; Kiceniuk and Jones, 1977). Doubling both the exchange surface area and the mean oxygen difference across the gills will quadruple oxygen uptake. Other subtle changes in water and blood flow or in conditions for diffusion across the gill epithelium must occur to account for the at least eightfold increase in oxygen uptake observed during exercise. These could include a redistribution of blood within each secondary lamella or a decrease in the thickness of the boundary layer of water or the mucus coat covering the gill epithelium. With respect to the former, it is noteworthy that a portion of each secondary lamella is buried in the surface of the filament and diffusion distances are larger between water and blood flowing in these basal channels. A redistribution of blood away from these basal channels would contribute to an increased gill oxygen transfer factor during exercise. With respect to the

7. T H E RESPIRATORY

465

A N D CIRCULATORY SYSTEMS

water boundary layer, this will presumably decrease with increasing water velocity.

4. IONICA N D OSMOTIC REGULATION The permeability of gills is low compared with other portions of the vascular system and the tight coupling of epithelial cells restricts water diffusion across the gills of eels to a rate which is 3000 times less than the free diffusion coefficient of water (Steen and Stray Pedersen, 1975). Although water and ion permeability is relatively low, gill area is large and water influx rates may approach values of 1pl/min/100 g for the eel (Anguilla anguil2a) (Evans, 1969) and 50 pl/min/100 g for . is a net influx the trout (S. gairdneri) (Wood and Randall, 1 9 7 3 ~ )There of water across the gills in freshwater fish which is slightly offset by the hydrostatic effect of blood pressure. I n seawater the osmotic and hydrostatic forces act in the same direction contributing to a net water loss across the gills. During exercise it is probable that lamellar recruitment results in an increase in functional gill area which facilitates the increase in oxygen uptake. This increase in functional area will also result in increased water and ion flux (Fig. 17). Farmer and Beamish (1969) observed an increase in plasma osmolality in Tilapia nilotica after exercise in saltwater and a decrease after swimming in freshwater.

SWIMMING

-- ’

Fig. 17. Model of the temporal changes in branchial water influx and renal water efflux rates in rainbow trout during the course of swimming activity. (From Wood and Randall, 1973c.)

466

DAVID R. JONES AND DAVID J. RANDALL

Byrne et al. (1972) observed little change in, the ionic composition of the plasma of freshwater-adapted Atlantic salmon (Salrno salar) after exercise in freshwater while, after exercise in seawater, there was a marked increase in plasma Na+, K+, and C1- in postsmolt fish acclimated to seawater. Wood and Randall (1973a,b) found little change in plasma sodium in rainbow trout but a marked net sodium loss during exercise in freshwater, due to an increase in passive diffusion of sodium across the gills. Plasma sodium levels remained unchanged because the ion loss was ameliorated by reduction in blood volume due to water loss via the kidney (Fig. 17). Wood and Randall ( 1 9 7 3 ~ also ) observed a net uptake of water across the gills which was initially large and caused an increase in the weight of the fish. However after about 20 min of swimming, a large diuresis occurred, body weight was reduced below, and hemoglobin and plasma protein levels were raised above, the resting level. Gill water influx gradually fell but remained above resting levels for at least 8 hr of exercise. Kidney water loss exceeded gill water influx so blood volume was reduced (Fig. 17). Active sodium influx across the gills remained unchanged during exercise (Wood and Randall, 1973a,b). During the period after exercise loss of sodium by diffusion was below uptake levels and blood volume was expanded as sodium stores were replenished. Since the cost of osmoregulation is claimed to b e at least 15% of oxygen uptake at all levels of exercise (Farmer and Beamish, 1969),then sodium cannot be typical of all ions because only the rate of passive loss changes with exercise, the rate of active uptake remaining unchanged. iV. T H E CIRCULATORY SYSTEM DURING EXERCISE

A. Introduction The maximum oxygen consumption of most fish is an order of magnitude lower than in similarly sized mammals (Brett, 1972). The maximum uptake of an active fish (salmonid, weight 50 g, 15°C) is about 1.0 ml 0,/100 g/min, representing an increase of 12 times above the resting level. Assuming that the locomotory muscles consume all of the increase in oxygen uptake, their rate of consumption will be 1.5 ml 0,/100 g/min, since muscle makes up about 63% of the body in adult salmonids (Brett, 1965a; Webb, 1971a). But, it is likely that only a small fraction of the total body musculature, the red muscle, is used during sustained swimming. I n salmonids red muscle makes u p about

7. T H E

RESPIRATORY AND CIRCULATORY SYSTEMS

467

4% of the myotomal mass (Webb, 1970) so the oxygen uptake of this tissue could be as great as 36 m1/100 g/min at maximum activity. On the other hand, if the red muscle scattered through the myotome is as active as that located in sheets at the sides of the fish, then the total red muscle mass quadruples and its oxygen consumption per unit weight is 9 m1/100 g/min. Nevertheless, it would appear that fish red muscle is capable of surprisingly high levels of oxygen consumption (in the mammalian range), and, since it is principally an aerobic tissue, the circulatory system must be capable of supplying it with sufficient oxygen. Since the capillary density of fish red muscle is about 1800/mm2, some two to three times more than that of fish white muscle, and in the same range as that of mammalian skeletal muscle (Landis and Pappenheimer, 1963; Cameron and Cech, 1970), it follows that, at the tissue level, circulatory adjustments in fish red muscle will be far more pronounced than those at the whole organism level. Furthermore, they may even b e more pronounced than those occurring in mammalian skeletal muscle in exercise.

B. Cardiac Adjustments to Exercise Since oxygen is transported from the gas exchanger(s) to the tissues by the cardiovascular system it is convenient to discuss the exercise adjustments b y using the Fick equation for oxygen and CO, transport: Oxygen uptake = heart rate x stroke volume x arterial-venous oxygen content difference or CO, elimination

=

heart rate x stroke volume x venous-arterial carbon dioxide content difference.

To meet an increase in oxygen demand the fish can increase one or more of the right-hand factors in the above equations. Since changes in A - V,, difference have already been discussed, this section will concentrate on the cardiac adjustments.

1.

HEARTRATE

Heart rate is one of the easiest variables to monitor from exercising animals, and the relative abundance of heart rate determinations from swimming fish reflects this fact. Nevertheless, there is surprisingly little consensus about the heart rate response to exercise in fish. Bradycardia at the onset of exercise has been reported for lingcod

468

DAVID R. JONES AND DAVID J. RANDALL

Fig. 18. Blood flow in the ventral aorta of an unrestrained lingcod. The start of spontaneous activity (at arrow) was associated with pronounced bradycardia, and diastolic flow fell to zero. (Modified from Stevens et al., 1972, Comp. Biochem. Physiol. A 43, 681. Copyright 1972 by Pergamon Press, Ltd.)

(Ophiodon elongatus) (Fig. 18; Stevens et al., 1972) and rainbow trout (Priede, 1974; Kiceniuk and Jones, 1977) but a decrease in heart rate with increasing intensity of sustained exercise has only been reported for “jack” sockeye salmon (Smith et al., 1967). Tachycardia during sustained swimming has been reported for sucker (Catastomus sp.), Atlantic mackerel (S. scombrus),blue runner ( C . crysos), bluefish (P. saltatrix, Fig. 12), northern scup (S. crysops), mature sockeye salmon, pumpkinseed (Lepomis gibbosus), bullhead (Zctalurus nebulosus), brown trout (S. trutta, Fig. 9), and rainbow trout (Smith et al., 1967; Sutterlin, 1969; Priede, 1974; Roberts, 1975a; Kiceniuk and Jones, 1977), while little (10-20% increase) or no change in heart rate with exercise has been reported for mullet (Mugil cephalus) (Roberts, 1975a), rainbow trout (Stevens and Randall, 1967a), cod (Gadus morhua) (Johansen, 1962), and a number of species of elasmobranchs (Johansen et al., 1966; Hanson, 1967). In those species in which heart rate increases during sustained swimming the relation between heart rate and speed (either absolute or relative) is not simple. Kiceniuk and Jones (1977)found for rainbow trout that below 50% of Ucritheart rate was little changed from resting levels whereas from 50 to 90% of Ucritheart rate increased linearly with U (Fig. 19). Sutterlin (1969) also noted in brown trout that heart rate was little affected by low swimming velocities. In fact, Priede (1974) took advantage of this relation and presented heart rate changes in rainbow trout as an exponential function of relative (or specific) velocity (Fig. 20a and b). However, since in salmonids the maximum heart rate is usually attained before maximum U an exponential relation is at best only a convenient approximation. Specific variations in the response of the heart to exercise have never been rigorously investigated, so one cannot say whether active fish make larger or more rapid chronotropic adjustments than sluggish species. However, there are certainly some interesting indications in

7. T H E

RESPIRATORY AND CIRCULATORY SYSTEMS

0

50

469

100

"/~'Jcrit

Fig. 19. Heart rate of rainbow trout at rest and during swimming (expressed as a percentage of each individual's Ucrit).The value at 20% U,,,, is for animals which were resting. Each point is a mean o f n determination's (n being given above the points) and vertical bars denote one standard error. (Modified from Kiceniuk and Jones, 1977,J. E x p . Biol. 69, 247.)

the literature which seem worthy of more concerted investigation, although the difficulties of making an assessment of the exercise performance in different species must be borne in mind. Exercise at one body length per second may b e close to maximum effort for a bullhead while it may only be one-third or one-fifth maximum U for a trout. Sutterlin (1969) showed that exercise of one body length per second provoked an increase in heart rate in bullheads of 2660% of the reiting level over a period of 10 min, whereas in pumpkinseed heart rate increased 128% of the resting rate in about 2 min in response to the same relative U , both species being examined at 20°C. On the other hand, the active rainbow trout only shows a heart rate increase of 2 5 3 0 % at one body length per second at 15"C, being about one-third of maximum heart rate at this temperature (Priede, 1974). The effect of temperature on the cardiac chronotropic response to exercise has been investigated by Priede (1974). Raising the acclimation and exercising temperature from 6.5" to 15°C caused maximum exercise heart rate to increase from 56 to over 93 beats/min in rainbow trout. However, resting heart rate also rises from 32.5 to 45.5 so the maximum potential change in gas transport, as assessed from this term in the Fick equation, only increases from 1.7 to 2.0 times (Fig. 20a and b). Changes in heart rate in fish can b e provoked either neurally or aneurally. I n fact, the hagfish heart, despite having an abundance of catecholamine-containing granules in specific cells (Augustinsson et al., 1956; Bloom et al., 1961; Hirsch et al., 1964) and even a system of ganglion cells (Hirsch et al., 1964), appears to b e functionally aneural (Jensen, 1961, 1963, 1965). In hagfish (Eptatretus stoutii) changes in

470

DAVID R. JONES AND DAVID J. RANDALL

.. I

I

I

I

I

1

I

I

Speci6c speed (U/L)

Fig. 20. (a) The relationship between heart rate and swimming speed of rainbow trout at 6.5"C. (A) Intact fish. Open circles are data from unilaterally vagotomized fish inserted for comparison but not used in the calculation of the lines shown. (B) Bilaterally vagotomized fish. The line shown is the mean basal heart rate; no significant relationship could be derived for the heart rates during swimming. (b) The relationship between heart rate and swimming speed of rainbow trout at 15°C. (A) Intact fish. Open circles are data derived from unilaterally vagotomized fish inserted for comparison but not used in the calculation of the lines shown. (B) Bivagotomized fish. (Figure and caption from Priede, 1974, J . E x p . B i d . 60, 305.)

7. THE c

RESPIRATORY AND CIRCULATORY SYSTEMS

471

heart rate are provoked mechanically by means of a stretch-induced rate-controlling mechanism in the myocardium (Jensen, 1961), a mechanism which also operates in lamprey (Petromyzon marinus) and teleosts with innervated hearts (Jensen, 1969).Another aneural mechanism for change in heart rate is a direct effect on the heart muscle of circulating catecholamines which, it is suspected, are produced from chromaffin tissue. I n elasmobranchs the chromaffin tissue is located close to the sinus venosus. The hearts of lampreys, elasmobranchs, and teleosts also receive vagal innervation which in lampreys is cholinergic and excitatory (Augustinsson et al., 1956; Jensen, 1969),in elasmobranchs cholinergic and inhibitory, while in teleosts the vagal innervation is cholinergic and inhibitory (parasympathetic) along with a sympathetic adrenergic excitatory innervation (Gannon and Burnstock, 1969). Consequently an analysis of the cardiac chronotropic response to exercise in teleosts becomes quite complex, involving the interplay of four effector mechanisms (two aneural and two neural) with the multiple feedback systems which control the integrated cardiovascular response. The neural control of heart rate during exercise has been investigated in both elasmobranchs and teleosts. In the former, heart rate appears to b e unchanged during swimming (in a large aquarium tank) although atropine injection causes heart rate to increase b y 30%, which implies that vagal tone is not reduced in exercise (Johansen et al., 1966). However, in most teleosts atropinization or bilateral vagotomy increases resting heart rate at low temperature (6"-10°C) although not at high temperature (15OC) (Stevens and Randall, 1967a; Stevens et al., 1972; Priede, 1974) while maximum heart rates during exercise are either unaffected or reduced (Fig. 20a and b). This implies that vagal integrity is essential for the maximum expression of the heart rate response and is the best in vivo evidence for a vagal cardiac excitatory function. The fall in heart rate which accompanies sudden movements of the fish or changes in imposed flow velocity is eliminated by atropinization or vagotomy in lingcod and trout (Stevens et al., 1972; Priede, 1974), and Priede (1974) suggests that this is the type of vagal effect seen in approach reflexes (Labat, 1966). On the other hand, to what extent withdrawal of vagal parasympathetic activity contributes to the exercise heart rate adjustment is uncertain. In those fish which show no elevation in resting heart rate with vagotomy or atropinization it would appear to b e negligible (Stevens and Randall, 1967a; Priede, 1974). On the other hand, in one individual sucker at 10°C (Stevens and Randall, 1967a) and perhaps rainbow trout at 6.5"C (Priede, 1974)

472

DAVID R. JONES AND DAVID J. RANDALL

withdrawal of vagal parasympathetic activity may account for the majority of the heart rate increase in exercise (Fig. 20a). However, in some bilaterally vagotomized trout at 6.5"C resting heart rate may, on occasion, fall back to that in intact animals, so in these fish vagal parasympathetic withdrawal cannot possibly effect the exercise tachycardia. Perhaps the best evidence for vagal withdrawal during exercise is the report of Stevens and Randall (1967a) describing heart rate response of rainbow trout to exercise during hypoxia. The bradycardia of hypoxia, a vagal parasympathetic effect (Holeton and Randall, 1967a,b; Randall and Smith, 1967), is transformed to tachycardia by exercise, the heart rate increase being much larger than normal, and must in part result from parasympathetic withdrawal. The neural control of heart rate during exercise, discussed above, leaves the impression that a considerable portion of exercise tachycardia, when it occurs, may be effected either humorally or mechanically. Catecholamines may be secreted quite rapidly at the onset of exercise, for adrenaline in blood plasma increases some five times above the resting values after about 2 min of disturbance in rainbow trout (Nakamo and Tomlinson, 1967).After a period of extreme disturbance of 10-15 min, both adrenaline and noradrenaline are very high in rainbow trout, the former increases by 20-30 times the resting level and the latter by 10-12 times (Mazeaud et aZ., 1978). What kind of cardiac chronotropic response would be provoked b y these levels of circulating catecholamines? If the adrenergic receptors causing the cardiac chronotropic effect are of the p-type, as is found in plaice (Pleuronectes platessa) (Falck et al., 1966), then it is to be expected that both adrenaline and noradrenaline would cause cardiac chronotropic effects. In the isolated trout heart, Bennion (1968)found that, at a given perfusion pressure at 6"C, heart rate was higher if the concentration of adrenaline in the perfusate was increased from 0.01 to 0.1 pg/ml, but at 15°C this was not the case and the lowest heart rates accompanied the highest adrenaline levels. On the other hand, in the isolated hearts of pike (Esox Zucius) and carp (Cyprinus carpio) both adrenaline and noradrenaline at concentrations of 1 pg/ml cause marked cardiac acceleration (Cardot and Ripplinger, 1967).The cardiac effects of adrenaline in vivo are equally contradictory, only this is perhaps more to be expected since a large part of the cardiovascular response to sympathomimetic drugs is determined by compensatory reflexes. For instance, in most vertebrates adrenaline stimulates peripheral a-receptors and provokes a rise in blood pressure, the effect of which is lessened by the baroreceptor reflex causing simultaneous vagal bradycardia. Hence Randall and Stevens (1967) obtained

7. T H E RESPIRATORY

473

AND CIRCULATORY SYSTEMS

bradycardia after intraarterial injection of 0.5 p g adrenaline in coho salmon (0.kitsutch) at 15°C (a barostatic reflex?) but tachycardia after atropinization. Increasing the dose level to 5 p g caused tachycardia and similar blood pressure increases in both control and atropinized fish (Fig. 21a and b). The existence of a barostatic reflex has never been confirmed in fish although there is certainly some suggestive evidence that one exists (Lutz and Wyman, 1932; Irving et al., 1935; Mott, 1950; Satchell, 1968). However, even in Randall and Stevens' (1967) experiments, marked bradycardia only occurred some 90 sec after peak blood pressure, in fact when blood pressure had almost returned to the initial level. This suggests that these variables may not be reflexly related, a conclusion apparently confirmed by the injection of the larger doses of adrenaline. On the other hand, Stevens et al. (1972) found that increased stroke volume and bradycardia accompanied the arterial pressure rise provoked by dorsal aortic injections of

a

25

b

+

t

5.0 pq eoimphrirm

x Beton

X B e b e atropin

0 AfIer 0

-2

t

0.50 pq epinephrine

-I

0

alqine I

atropim

0 AfIer oIropim 2

Tim. min

3

4

- 2 - 1

0

I

Tim,

2

3

4

min

Fig. 21. The effect of intravenous injections of epinephrine on dorsal aortic blood pressure and heart rate in coho salmon before (a) and after (b) blockade of the efferent cholineryic fibers innervating the heart with atropine. (From Randall and Stevens, 1967, Comp. Biochem. Physiol. 21, 415.)

474

DAVID R. JONES AND DAVID J. RANDALL

10 pg of adrenaline in lingcod (0.elongatus).Atropine eliminated the bradycardia and, after adrenaline injection, stroke volume fell and cardiac output was reduced b y 15%. Using a similar dose level in carp (Cyprinus carpio) (intracardiac injection) Laffont and Labat (1966) obtained bradycardia at low temperatures (1"-8"C) and tachycardia at high temperatures (9O-2OoC). The fact that the cardiac response to adrenaline is temperature-sensitive and perhaps dose-dependent may mean that the fish heart, like that of the frog, has a- and p-adrenoceptors which are allosteric conformations of the same structure, being predominantly p at high temperatures and a at low (Kunos and Nickerson, 1976).This suggestion is not confirmed by experiments with sole (Solea vulgaris) and plaice (Pleuronectes Jesus) at 18°C since adrenaline injection caused bradycardia in both intact and vagotomized fish (Labat, 1964). It has been suggested that the pacemaker cells of the fish heart are sensitive to stretch and depolarize more rapidly when extended (Jensen, 1961; Deck, 1964) although this has not been confirmed in all studies (Golenhofen and Lippross, 1969). The importance of this mechanism is that if the pacemaker is stretched during exercise due to an increase in venous pressure, then heart rate will increase (Pathak, 1972). This mechanical effect has been clearly demonstrated for the isolated ventricles of hagfish (Jensen, 1961, 1965), lamprey (Jensen, 1969), elasmobranchs (Jensen, 1970), carp (Cyrpinus sp.) (Harris and Morton, 1968), and rainbow trout (Jensen, 1969), and also in the isolated heart of rainbow trout (Bennion, 1968). A striking feature of all the experiments on isolated ventricles is that large changes in perfusion pressure are required to provoke any significant change in heart rate, much larger than occur in vivo. For instance, in the rainbow trout ventricle at 21"C, Jensen (1969) found that rate increased from 8 to 13 beatdmin for an order of magnitude change in filling pressure. It could b e argued that the ventricular pacemaker was much less stretchsensitive than that in the sinus venosus but Bennion (1968) obtained only slightly larger increases in rate at 15°C with an even bigger pressure increase in isolated trout hearts. Although high venous pressures have been reported in trout during exercise (Stevens and Randall, 1967a) they have not been confirmed by more recent work (Kiceniuk and Jones, 1977) so whether this mechanical factor plays any role in the chronotropic response to exercise is uncertain. However, saline infusion into the veins of brown trout (Sutterlin, 1969) and catfish (Ameirus nebulosus) (Labat et al., 1961) caused elevation in heart rate but, even so, venous pressure reached 13mm Hg before any effect was seen on trout heart rate. Despite reports to the contrary

7. THE RESPIRATORY AND CIRCULATORY SYSTEMS

475

(Stevens and Randall, 1967a) the increases in heart rate in Labat et al.'s (1961) catfish (sic. bullheads) could not be blocked b y atropine. Obviously the control of the fish heart during exercise is still not understood, although it appears most likely that the tachycardia is a product of a vagal sympathetic contribution allied with a withdrawal of parasympathetic activity and a direct effect of circulating catecholamines. When occurring, exercise bradycardia would appear to be mediated by the vagal parasympathetic innervation, vagal inhibitory tone also being present in those species showing no cardiac chronotropic response to exercise. The afferent nervous link of the heart rate response has not been elucidated and, as in mammals, it might be expected to come from diverse inputs, although movement of water past the fish (Sutterlin, 1969) or passive tail beating (West and Jones, 1975) do not appear to affect heart rate. 2. STROKEVOLUME Fishes are capable of large changes in stroke volume, and maximal values reported in the literature often exceed those given for similarly sized mammals. In the rainbow trout, for instance, maximum stroke volume seen in anemic animals is 2.67 ml/kg which is almost 10 times that obtained in similarly sized resting fish at the same temperature (Cameron and Davis, 1970; Davis and Cameron, 1971). Hence, the potential for change in gas transport due to changes in stroke volume in trout is equal to the maximum change in oxygen uptake or CO, excretion from resting to fully active metabolism at 8"-1OoC. It has been stated frequently that increased cardiac outputs in fish are achieved through adjustments in stroke volume rather than heart rate (Randall, 1970a).Virtually all exercise studies tend to confirm this except those on elasmobranchs where pulse flow in the ventral aorta (measured downstream of the first set of branchial arteries) appears to be unchanged by short periods of moderate activity, although pulse flow increases during the recovery period (Johansen et al., 1966). Since the majority of the exercise periods monitored by Johansen et al. (1966) were probably less than one blood circulation time, the significance of the lack of adjustments is difficult to interpret. However, in lingcod (0. elongatus) stroke volume increases markedly along with the bradycardia at the onset of exercise (Fig. 18). An increase in pulse flow is also seen in the hemoglobin-free fish Chaenocephalus aceratas at the start of activity although heart rate is unchanged (Hemmingsen et al., 1972). Although total heart output was not recorded in Chaenocephalus, if an allowance is made for flow

476

DAVID R. JONES AND DAVID J. RANDALL

through the last two pairs of branchial arches, stroke volume must have reached 10 ml/kg during activity, a testament to the remarkable stroke volumes in this fish, being the highest, on a unit weight basis, of any vertebrate. Stroke volume determinations during sustained performance at a known work output have only been inade on two occasions and both of these were b y indirect methods. To determine stroke volume indirectly, simultaneous measures of oxygen uptake or carbon dioxide output, arterial and venous oxygen or CO, contents, and heart rate are required, preferably directly and from the same fish. Stevens and Randall (1967b) report a fivefold increase in stroke volume in rainbow trout but unfortunately none of the above conditions was fulfilled. In a recent series of experiments Kiceniuk and Jones (1977)used a 1-hr test period at each swimming speed and managed to obtain four simultaneous determinations of the required variables at speeds approaching Ucrit(compared with nine resting determinations) (Table 111).In these experiments stroke volume increased 2.24 times, while oxygen uptake increased b y 7.75 times, but there was no indication that stroke volume had reached its limit, unlike the heart rate response (Fig. 22; Kiceniuk and Jones, 1977). Changes in stroke volume occurring during exercise may be

/

lot

t

L

strc)ke volume x10 (ml/ kg) A A - V O2 Difference 0

(VOIS.

%)

-

5

/beat)

I

/+' 1 1

0.5

Heart r a t e x Cardiac output I

I

V

02

I 2.5

I

x10-l (rnl/kg/min) I

I

I 4.5

I

(ml I k g I min)

Fig. 22. The relationship between a number of cardiovascular variables and oxygen consumption in exercising rainbow trout. (All data from Kiceniuk and Jones, 1977.)

7. T H E

RESPIRATORY AND CIRCULATORY SYSTEMS

477

caused by all or some of the same factors that promote heart rate increases, but the effects of these factors are not linked since large changes in stroke volume can occur independent of rate changes and vice versa. Adrenergic compounds stimulate preceptors in the fish heart and may have positive inotropic effects. In teleosts catecholamines may be released into the circulation or directly onto the heart from the sympathetic vagal innervation whereas in elasmobranchs there is no adrenergic innervation of the auricle and ventricle. Gannon et al. (1972) have suggested that in elasmobranchs amines are released from the anterior chromaffin masses into the blood in the posterior cardinal sinus and aspirated into the heart. These chromaffin cells are innervated by naked nerve endings and this indicates a neurohumoral control of the elasmobranch heart which may be much more specific than had been supposed. In the cod (Gadus morhua) sympathetic medullated fibers innervate the chromaffin cells lining the walls of the posterior cardinal veins, and following section of these nerves in acute preparations adrenaline release in response to stress is prevented (Nilsson et al., 1976). Since release of adrenaline on nerve stimulation is abolished by mecamylamine (a ganglionic blocker) it is concluded by Nilsson et al. (1976)that chromaffin tissue is sympathetically controlled by preganglionic cholinergic fibers. There seems to be general agreement that in virtually all fishes adrenergic compounds have marked inotropic effects in contrast to the weaker chronotropic effects. Falck et al. ( 1966) clearly demonstrated the positive inotropic effects of these compounds on the isolated hearts of lamprey (Lampetra jluuiatilis) and plaice (Pleuronectes platessa) but, as Gannon (1971) pointed out, their technique did not allow them to differentiate between atrial and ventricular responses. However, in teleosts Gannon (1971) showed that both atrium and ventricle responded positively to adrenergic compounds and confirmed that the receptors were of the @-type.On the other hand, Bennion (1968) found that high perfusion pressures and temperatures were required for a marked increase in stroke volume of the isolated trout heart in response to an order of magnitude increase in adrenaline in the perfusate (0.01-0.1 pg/ml). When perfusion pressure was low (2 mm Hg) or at high perfusion pressure at low temperature (6°C) there was apparently little effect. In intact fishes adrenaline injections cause both central and peripheral cardiovascular effects, which lead to the initiation of reflexes to counteract them, and complicate an assessment of the effect of adrenaline on a single target organ such as the heart. By placing lingcod (0.elongatus) in water containing 1 pg/ml adrenaline for a protracted

478

DAVID R. JONES AND DAVID J. RANDALL

period, Stevens et al. (1972) minimized compensatory reflexes and revealed the positive inotropic effect of adrenaline, for stroke volume increased b y one-third (Fig. 23). In dogfish (Squalus acanthias) it has been clearly shown that some elements of a neuronal autonomic cardiovascular control system are present (Opdyke et d., 1972) and in the skate (Raja binoculata) injection of 5 p g of adrenaline causes an elevation in ventral aortic blood velocity as well as an increase in blood pressure which again implies a positive inotropic role for adrenaline in vivo (Johansen et al., 1966). As has already been discussed, part of the cardiac chronotropic response to exercise may result from a withdrawal of vagal tone and since acetycholine has a negative inotropic effect on some teleost hearts (Falck et al., 1966), cessation of cholinergic inhibition will tend to augment the force of contraction. However, there appears to be little or no cholinergic innervation of the teleost ventricle (Cobb and Santer, 1972) and correspondingly no inotropic effect of applied acetycholine (Belaud and Peyraud, 1970; Gannon, 1971). The negative inotropic effects of acetycholine on the teleost atrium are pronounced (Gannon, 1971) and since the single atrium feeds the single ventricle an increase in atrial stroke volume in exercise would tend to raise ventricular stroke volume by means of the passive Starling mechanism, since

a

x

c

'E

1 E

.c-

€

+

2

Time, min

Fig. 23. The effect of the addition of 1.0 pg/rnl adrenaline to the water in which the fish was situated on heart rate, stroke volume, and mean blood flow in the ventral aorta of a lingcod weighing 2.3 kg. Addition of adrenaline is indicated by the dashed line. (From Stevenset cil., 1972. Reprinted with permission fromCornp. Biochern. Physio1.A 43, 681. Copyright 1972 b y Pergamon Press, Ltd.)

7. THE

RESPIRATORY AND CIRCULATORY SYSTEMS

479

atrial contraction is the major, if not the sole, contributor to ventricular filling in teleosts (Randall, 1968, 1970a). Starling (1918) stated that “the energy of contraction (of a muscle) is a function of the length of the muscle fiber.” Consequently, if a cardiac chamber is filled to a greater extent (end-diastolic volume increased) then the next contraction is stronger and a greater stroke volume is pumped. Both the aneural hagfish heart ( E . stoutii) and the isolated trout heart appear to follow Starling’s law (Chapman et al., 1963; Bennion, 1968).However, it is not clear what happens in innervated fish hearts in response to increased filling pressure, when adjustment of stroke volume to the varying needs of the animal may be more potently influenced by nervous or hormonal systems. In the intact cod (Gadus rnorhua) experimentally increasing venous return causes stroke volume to almost double in a few beats (Johansen, 1962). However, the fact that the elevated stroke volume persisted for some time after the actual infusion, while ventricular pressures were relatively unaffected throughout, implies the involvement of neurohumoral regulation in this response, and therefore casts doubt on the role of a “pure” Starling mechanism in intact fishes, at least in teleosts. In summary, there is no doubt that stroke volume in fishes is extremely labile and there is good evidence that it increases markedly during exercise in teleosts. However, the genesis of the increased cardiac contractility is not understood although neurohumoral factors are probably more important in this respect than passive mechanisms. Oxygen uptake is a function of cardiac output, the product of heart rate and stroke volume, and the A - Vo2 content difference which, on occasion, increases b y 2.5 times (Kiceniuk and Jones, 1977) in exercise, adding markedly to the capacity for gas transport (Table 111).I n teleosts cardiac output has an approximately linear relation to oxygen uptake across the activity spectrum but since heart rate approaches its maximum value when oxygen uptake has increased by, at most, three times, then this linear relation is achieved due to a proportionately greater increase in stroke volume (Fig. 21; Kiceniuk and Jones, 1977). The product of stroke volume and A - V,, content difference is the oxygen pulse which is more conveniently obtained by dividing oxygen uptake b y heart rate. In exercising rainbow trout oxygen pulse increased by 5.67 times, from 0.0148 to 0.084 ml/kg/beat, while oxygen uptake rose b y 7.75 times, showing that heart rate changes only make a minor contribution to the increase in oxygen transport in trout during exercise (Fig. 21; Table 111; Kiceniuk and Jones, 1977). In contrast, in mammals oxygen pulse would not be expected to change by more than three or four times for even a ten times increase in oxygen uptake above resting levels.

480

DAVID R. JONES AND DAVID J. RANDALL

Table An Evaluation of the Oxygen Transport Swimming speed

(9.U,,,) Rest

41-63% 70-78%

Heart rate (heatdmin)

Arterial 0, content (vol%)

Venous 0, contentb (vol%)

37.8 f 1.5 n=9

104 f 0.5

7.1 e'0.7 n=9

42.7 f 3.18 n=3

n=9

9.8 f 0.74 n=3

Arterial 0, tension

Venous 0, tension'

(Ton)

(Ton)

137 ? 4.2 n=8

4.4 2 0.8 n=3

123.5

33.2 t 3.0 n=8

* 7.5

n = 2

Hematocrit (%) Hct.

Hct

22.6 f 1.0 n=9

24.2 f 1.8 n=8

22.7 f'1.4 n=3

24.45 C 1.&5 n=2

49.0 f 1.00 n=5

9.02f 0.5 n=5

3.4 z 0.4 n=5

123.0 2 4.2 n=4

23.5 t 2.1 n=4

20.34 f 1.4 n=5

21.85f 2.4

81-91%

51.3 f 4.6 n=3

10.2 f 1.3 n=3

2.9 e 1.47 n=3

128.0 f 5.0 n=3

29.3 f 6.2 n=3

22.5 f'1.35 n=3

25.8 t 0.9 n=3

Maximum (92%+)

51.4 f 2.48 11 = 4

9.7 f'0.7 n=4

126.0 f 5.4 n=4

16.0f 2.1 n=4

25.7 f 0.8 n=4

27.4 t 1.2 n=4

1.35 z 0.4 n-4

n=4

a Data obtained from six trnut and n = the number of determinations. All values given as means f SEM except pH. and pH,. for which the values are means + and - SEM. Mean and standard errors were calculated on hydrogen ion concentration and reconverted to pH values.

C. Arterial Blood Pressure and Total Peripheral Resistance during Exercise In mammals maximum oxygen uptake during exercise may b e limited either b y the maximum cardiac output and oxygen extraction at the tissues or by the ability of the cardiovascular system to generate pressure against the minimum total peripheral resistance (TPR).It has been suggested that, in mammals, maximum oxygen uptake may b e best expressed as the product of maximum A - Vo, content difference times the ratio of maximum blood pressure to minimum peripheral resistance (Rowell, 1974). In other words, cardiac output has been redefined as this ratio. Certainly in fish, where 5 6 6 0 % of the body may be locomotory muscle, the potential exists for large changes in total peripheral resistance which may have to b e controlled more with respect to maintenance of arterial blood pressure than total blood flow. It is generally agreed that arterial blood pressure increases during exercise in teleosts although in elasmobranchs, during very short bursts of activity, there appears to b e no pressure change (Johansen et al., 1966).In teleosts ventral aortic pressure rises markedly in the early

7. T H E RESPIRATORY

481

A N D CIRCULATORY SYSTEMS

I11 System in Trout during Exercise"

PH.

PH,

7.932 +7991 -7.879

7.959 +8.025 -7.902 n=4

n = s

7.924 +R.Wfi -7.829

n = G 7.859 +7.970 -7.770 n=2 7.610 1-7.620 -7.600 n=2

7.988 +R.081

Arterialvenous 0, content (vol%)

Oxygen consumption (ml/kg/min)

G.29f0.26 11 = 9

0.56=0.02

5.4 f 0.1 n=3

1.52 f 0.24

n=9

Inspired water 0, tension (Tom)

Cardiac output (ml/kg/niin)

Stroke volume

17.6* 1.1 n=9

0.46*0.02 n = 9

152.95 1.9

28.4 f 5.0 n = 3

0.62 f 0.8 n=3

98.75 f 1.0 n=5

fmlW

n=8

Arterial 0, saturation

f%) 9 7 . 0 2 1.3 n=9

1.9 f 0.27 n=5

34.8 f 4.8 n = 5

0.7 t 0.09 n = 5

152.0 f'2.31 n = 3 155.75 2 0.95 n=5

7.3-0.49 11 = 3

3.12?00.38 n=3

42.9-5.4 n=3

0.8620.16 n=3

146.720.67 n = 3

99.7-0.67 n=3

8 . 3 ? 0.5 n=4

4.34 f 0.17 n=4

5 2 3 f 2.2 n=4

1.03f 0.07 n=4

151.8 T 2.6 n=4

9 8 . 5 f 0.87 n=4

5.6

?

ri =

0.58 5

n=3

96.0 2 5.00 n=3

-7.911 n=3 7.883 +7.950 -7.825 n = 2

7.548 +7.fi30 -7.480 n = 2

Values obtained from the common cardinal vein and ventral aorta were similar,

part of a bout of activity and mean pressure may attain values 10-15 mm Hg above the resting level (Johansen and Waage-Johannessen, 1962; Stevens and Randall, 1967a; Kiceniuk and Jones, 1977). It is interesting that Kiceniuk and Jones (1977) found that the magnitude of the increase in mean arterial blood pressure was the same whether the animal started swimming from rest or whether a new speed was imposed on a swimming fish in an incremental velocity test. Peak mean arterial pressure occurs between 5 and 15 min after the onset of swimming and then declines so that, after 1 hr of swimming, mean pressure is only elevated by 3-4 mm Hg (Fig. 24; Kiceniuk and Jones, 1977). Hence if four or five velocity tests are imposed, then mean ventral aortic pressure will increase b y about 15-20 mm Hg (Fig. 25). The fact that the change in mean pressure early in exercise is independent of the imposed velocity implies that it may be a response to disturbance in terms of cardiac overcompensation, rather than exercise stress and further emphasizes the need to differentiate between shortterm (Stevens and Randall, 1967a,b) and steady state exercise (Kiceniuk and Jones, 1977). On the other hand, dorsal aortic pressure changes are much smaller

482

DAVID R. JONES AND DAVID J. RANDALL

V

0

1 3 6 10 15 30 45 Time from increase in speed (min)

I

60

Fig. 24. Change in systolic (S) and diastolic (D) blood pressure in the dorsal aorta (DA) and ventral aorta (VA) following an increase in swimming speed. X's are means of (pressure at time t,,) - (pressure at time t o ) ,pressure at t obeing the pressure immediately before a speed increase was imposed. Vertical bars denote one standard error and numbers above the points represent the number of determinations on five individuals. (From Kiceniuk and Jones, 1977,J. E x p . Biol. 69, 247.)

than ventral aortic pressure changes in both short-term and steady state exercise. In fact Smith et al. (1967) reported that in mature and some jack sockeye salmon dorsal aortic mean pressure only started to rise when the maximum speed was approached and was closely fol-

''0

" c r it

Fig. 25. Changes in diastolic ( x ) and systolic ( 0 )blood pressures in both the dorsal aorta (DA) and ventral aorta (VA) of trout at rest and during swimming (expressed as a percentage of each individual's critical velocity). Values at less than 25% Urn, are for animals which were resting. The points are means of individual determinations, the vertical bars denote one standard error, and the numbers above the points indicate the numbers ofdeterminations on eightanimals. (From Kiceniukand Jones, 1977,J.Exp. Biol. 69, 247.)

7. THE

RESPIRATORY AND CIRCULATORY SYSTEMS

483

lowed by fatigue of the fish, although they also report that blood pressures were usually higher in the first 30 min of an exercise period, before falling back to control levels. In the early period of swimming (10-15 min) at any velocity mean dorsal aortic pressure rises by 5 mm Hg in trout which is less than one-half the change in ventral aortic pressure (Stevens and Randall, 1967a; Kiceniuk and Jones, 1977). After 1hr of swimming at any fixed speed dorsal aortic pressure is only 1 mm Hg above the starting level so an increasing velocity test with five increments may cause an elevation in dorsal aortic pressure of only 5 mm Hg. This is about one-quarter to one-third of the observed increase in mean ventral aortic pressure (Fig. 24; Kiceniuk and Jones,

1977). As we have seen, cardiac output may increase from 3 to 6 times during strenuous activity so, in the absence of change in total peripheral resistance, mean blood pressure might be expected to increase by this amount. However, this does not occur in rainbow trout for, close to Ucrit,ventral aortic pressure is usually only 20-25 mm Hg above the resting value of 35-40 mm Hg (Fig. 25; Kiceniuk and Jones, 1977). In Kiceniuk and Jones’ (1977)experiments cardiac output only increased b y 2.4 times so total peripheral resistance fell by about one-third. The pressure drop across the gills (ventral minus dorsal aortic pressure) increased from a resting level of 8 to 24 mm Hg at 292% Ucrit,which slightly exceeded the flow changes, so gill resistance increased. On the other hand, the pressure drop through the systemic circuit (dorsal aortic minus venous pressure) only increased slightly so systemic resistance fell markedly. The differential changes in resistance for these two vascular beds drastically alter the relationship seen at rest between these resistances. At rest the resistance of the body circuit is 4 times that of the gills but it falls as maximum activity is approached and reaches a level only 1.5 times that of the gill circuit. The largest fall in total peripheral resistance occurs, under steady state conditions, between rest and activity, and TPR is relatively unaffected by further speed increases up to Ucrit(Kiceniuk and Jones, 1977). In fact, as Wood (1974a) has pointed out, further small decreases ( 5 1 0 % ) in resistance that accompany increasing U , in an incremental velocity test, could be due to passive dilatation caused by the increase in blood pressure. The question arises as to what agents provoke these changes in resistance for the two vascular beds. Obviously a rise in blood pressure will cause passive vasodilatation of the vascular beds and reduce resistance, although this effect would appear to be more pronounced on the branchial than systemic circulation (Wood, 1974a; Wood and Shelton,

484

DAVID R. JONES AND DAVID J. RANDALL

1975). Further, local auto-regulation may play a role on the systemic side but some changes in vascular resistance are known to be under neural or hormonal control. Kirby and Burnstock (1969) suggested that there was a gradual transition in the innervation of arteries from a predominance of cholinergic involvement in fish to one of adrenergic predominance in mammals. Certainly muscarinic cholinergic receptors, which cause contraction on stimulation, are present in pregill vessels of trout (Kirby and Burnstock, 1969; Klaverkamp and Dyer, 1974).Their existence in postgill vessels, however, is less conclusively proved (Holmgren and Nilsson, 1974). However, a-adrenergic excitatory receptors have been shown to exist in the major arteries of dogfish (Squalus acanthias) (Capra and Satchell, 1974), trout, and cod (Gadus morhua) (Holmgren and Nilsson, 1974), and p-adrenergic inhibitory receptors in dogfish and trout (Capra and Satchell, 1974; Klaverkamp and Dyer, 1974). In many instances only slight changes in resistance will be provoked b y contraction and relaxation of smooth muscle in walls of large arteries so that any changes occurring during exercise must be attributed to contraction or relaxation of the smaller “resistance” vessels. The wall of the coeliac artery in some teleosts is particularly well invested with smooth muscle and it is conceivable that contraction of this segment could eliminate blood supply to the gut allowing blood to be redirected to the muscles. However, vascular control of this type is unlikely to make a large direct contribution to blood flow adjustments in exercise. On the other hand, the compliance of the vessels may alter drastically on receptor stimulation, which could cause marked circulatory effects. For instance, an increase in systemic compliance alone would be expected to increase ventral aortic flow pulsatility while decreasing pressure pulsatility in both ventral and dorsal aortas if stroke volume is unaltered (Satchell, 1971; Jones et al., 1974; Fig. 26). A decrease in systemic compliance would have the reverse effects on pressure and flow pulsatility. Pressure pulsatility (pulse pressure/mean pressure) in both the dorsal and ventral aortae increases b y 70%, from the resting level, in swimming trout in a steady state close to Ucrit(Kiceniuk and Jones, 1977). Furthermore, the ratio between pressure pulsatility in the ventral and dorsal aortas changes little during exercise, from 1.66 at rest to 1.55 at 80-100% U , , , (Kiceniuk and Jones, 1977). Unfortunately, since heart stroke volume, heart rate, and gill and systemic resistances all change during exercise it is impossible to judge the relative contribution of arterial compliance changes to these effects on pulsatility . With respect to resistance vessels in the gills and systemic circula-

485

7. T H E RESPIRATORY AND CIRCULATORY SYSTEMS

9”.

P&

M + +

2 SEC

I

I

.5 SEC

Fig. 26. Pressures and flows before and after “gill resistance” in a hydraulic model of the fish circulation. Traces (from top to bottom)-first: pressure proximal to “gills” in “ventral aorta” (Pva); second: flow proximal to “gills” in “ventral aorta” (QV,) (arbitrary units); third: flow distal to “gills” in “dorsal aorta” ( Q d a ) (arbitrary units); fourth: pressure distal to “gills” in “dorsal aorta” (Pda).At arrow, a compliance was introduced into “dorsal aorta’’ distal to gills. (From Jones et al., 1974, Am. J . Physiol. 226, 90.)

tion, a large number of studies have demonstrated that catecholamines cause branchial vasodilatation (Keys and Bateman, 1932; Ostlund and Fange, 1962; Kirschner, 1969; Steen and Kruysse, 1964; Richards and Fromm, 1969) and systemic vasoconstriction (Keys and Bateman, 1932; Reite, 1969; Wood and Shelton, 1975). The actual classification of the adrenoreceptors responsible for these actions is not entirely clear although &-dilatory receptors predominate in the gills (Wood, 1974a, 1975) and a-constrictor receptors in the body circulation (Wood and Shelton, 1975). Cholinergic receptors, on the other hand, cause vasoconstriction in both gills and systemic portions of the circulation (Davies and Rankin, 1973; Reite, 1969; Wood, 1974a, 1975). During exercise gill resistance changes variably but there is a marked fall in systemic resistance. Circulating catecholamines, which

486

DAVID R. JONES AND DAVID J. RANDALL

are frequently invoked to explain cardiovascular adjustments to exercise, would cause dilation in gill and constriction in systemic vessels. Certainly the levels of circulating catecholamines which occur in intact stressed dogfish (Mazeaud, 1969) and trout (Nakano and Tomlinson, 1967) are sufficient to cause gill vasodilatation (Davies and Rankin, 1973; Wood, 1974a) and presumably systemic vasoconstriction. There is evidence that p-inhibitory receptors in the gills are much more sensitive to circulating catecholamines than the a-excitatory receptors in the body circulation (Wood, 1974a; Wood and Shelton, 1975), the body resistance being primarily under direct neural control (Wood and Shelton, 1975). However, extreme stress (chasing and air exposure) is likely to cause far larger increases in circulating catecholamines than actual swimming activity. Consequently, this could explain a lack of gill resistance changes in steady state activity (Kiceniuk and Jones, 1977).On the other hand, any drop in gill resistance during exercise (Stevens and Randall, 1967a) could be due to neurally controlled vasodilatation. A decrease in vagal cholinergic tone (parasympathetic) could cause gill vasodilatation by eliminating excitatory muscarinic activity while an increase in vagal adrenergic activity (sympathetic) could promote gill vasodilatation through the agency of adrenergic a- and p-inhibitory receptors (Wood, 1974a; Johansen, 1972; Rankin and Maetz, 1971). On the body side of the circulation one might look for the enhancement of cholinergic and adrenergic neural activities in some beds while for others, such as locomotory muscles, these neural activities may be eliminated. There is some indirect evidence, due to the presence of “Mayer” type waves in resting fish after hemorrhage, that systemic vasomotor tone is neurally controlled through the agency of a-adrenergic receptors (Wood, 1974b) but in the absence of detailed knowledge on the balance between adrenergic and cholinergic controlling mechanisms, even in resting animals, it is difficult to predict what may occur during exercise.

D. Venous Pressure and Venous Return during Exercise The veins are large vessels offering low resistance to flow and, as in the major arteries, energy gradients (kinetic and pressure energy) required for flow are small. Obviously, in the absence of pronounced venomotor activity, any increase in energy gradient between the periphery and heart will increase venous return. In exercise an increase in energy gradient may be caused by a rise in peripheral venous

7. THE RESPIRATORY AND CIRCULATORY SYSTEMS

487

pressure, on a sustained or momentary basis, due to systemic vasodilatation or activities of the muscle pump. Also, in those fishes with a rigid pericardium, intrapericardial pressure may fall well below environmental pressure during activity which will expand the great veins and reduce central venous pressure, further enhancing the energy gradient for flow. Venous pressures are higher at sites peripheral to the heart and in the caudal vein of resting elasmobranchs reach 1-3 mm Hg (Hanson, 1967; Birch et aZ., 1969; Satchell, 1965) while close to the heart negative pressures are commonly obtained (Sudak, 1965a,b; Hanson, 1967; Birch et al., 1969).A similar picture seems to hold in some teleosts for Stevens and Randall (1967a) recorded pressures of 8 mm Hg in the subintestinal vein of resting rainbow trout while Kiceniuk and Jones (1977) obtained a value of 1.4 mm Hg in the central common cardinal vein. Unfortunately no recordings of venous pressure have been made from free swimming elasmobranchs but in trout pressure in the subintestinal vein rises substantially at moderate swimming speed (Stevens and Randall, 1967a) while in the cardinal veins a small increase (0.5 mm Hg) is only observed close to Ucrit(Kiceniuk and Jones, 1977).The subintestinal vein empties into the liver before reaching the heart, and Stevens and Randall (1967a) attributed the pressure rise to hepatic venomotor activity. Also, since the subintestinal vein drains an area not directly involved in locomotion, an increase in flow resistance in the vascular bed it drains might be expected in exercise and, in fact, Stevens and Randall (1967a) found the amount of blood they could drain from this vein (central end occluded) fell during swimming, while the pressure gradient from the dorsal aorta to the subintestinal vein was virtually unchanged. This observation is probably the best evidence for redistribution of blood away from visceral regions during exercise. I n many elasmobranchs cardiac contraction creates a negative pressure within the pericardium and great veins entering the ductus cuvieri (Johansen and Hanson, 1967; Hanson, 1967; Birchet al., 1969). A structural prerequisite for negative pressure generation is a rigid pericardium, which is lacking in teleosts. Nevertheless, reports have appeared of negative pressure in the central veins and sinus venosus of teleosts (Brunings, 1899; Mott, 1951).Johansen (1965a,b) has shown that as heart rate slows the pericardial pressure approaches zero whereas tachycardia causes it to become more negative; tachycardia during exercise would therefore be expected to promote venous return. However, pronounced tachycardia during exercise has only been shown to occur in teleosts, and since there is no rigid pericardium in

488

DAVID R. JONES AND DAVID J. RANDALL

teleosts, there is consequently no good evidence that negative intrapericardial pressure plays any specific role in exercise. Obviously venous return increases during exercise but the only certain agent of its promotion lies in the auxillary pumping activities of the waves of locomotory muscle contraction acting on veins or special venous reservoirs. I n elasmobranchs these reservoirs are located in the tail and under the median fins (Birch et al., 1969).However, for a muscle pump to be effective the veins must be compressible and possess valves to ensure that flow only travels unidirectionally (usually toward the heart). Although these criteria are well met in the veins associated with the median fins and tail (Fig. 29), all the major longitudinal veins in fishes, on the other hand, only possess valves where they enter the sinus venosus or atrium. In elasmobranchs the valves which prevent back flow from the ductus cuvieri into the major longitudinal vessels open during cardiac systole and close during diastole (Birch et al., 1969). However, the long veins possess no other valves and in elasmobranchs they are incompressible tunnels surrounded by connective tissue (Satchell, 1971). Only the hepatic vein of elasmobranchs is a discrete tubular structure, similar to teleostean veins, but its opening into the sinus venosus is protected b y a muscular sphincteric-type valve. In fact, there is evidence showing that this valve opens during swimming and may aid in the mobilization of the liver blood store during exercise (Johansen and Hanson, 1967). However, the lack of valves, combined with the noncollapsible nature of veins in elasmobranchs, or the protection of veins, such as the caudal vein, within the incompressible hemal canal, suggest that, if anything, these vessels are designed to be protected from the waves of muscular contraction rather than designed to be the propulsive part of a muscle pumping system. I n fact it is the segmental veins of the postpelvic regions which form the muscle pump in both elasmobranchs and teleosts. The segmental intercostal veins, draining blood from the muscle blocks, possess valves where they enter the main longitudinal veins (Fig. 27a; Sutterlin, 1969) and, in elasmobranchs which retain their tail as the major means of propulsion, the intercostal arteries also are valved where they arise from the dorsal aorta (Fig. 27b; Birch et al., 1969). The orientation of these valves is such that back flow of blood is prevented. Consequently, as the muscular wave of contraction passes backwards along the body the capillaries, veins and perhaps even the arteries are squeezed and the blood is forced into the caudal vein. When active tailbeat flexions are caused experimentally, elevations (2-4 mm Hg) in caudal vein pressure occur in both teleosts and elasmobranchs (Satchell, 1965; Sutterlin, 1969), but Satchell (1965) found

7. THE

RESPIRATaRY AND CIRCULATORY SYSTEMS

489

b

V.Cl.V.

Fig. 27. (a) The arrangement of the blood vessels in the caudal region of a trout, showing the position of the valves in the segmental veins. The anterior end of the fish is toward the left. v, valves; nc, nerve cord; vt, vertebra; ca, caudal artery; cv, caudal vein; vsv, ventral segmental vein; dsv, dorsal segmental vein; vsa, ventral segmental artery; dsa, dorsal segmental artery. (From Sutterlin, 1969,Physiol. 2001.42,36.Copyright I969 by the UniversityofChicagoPress.)(b) (i) Diagrammaticcross sectionofthetrunknearthe base ofthe tail. T h e intercostal artery and intercostal vein have their origins halfa segment apart; they are here shown atthesanie level. (ii)Thelocationofthearterialvalves. (iiifThe location of the venous valves. d.ct.v., dorsal cutaneous vein; I.ct.v., lateral cutaneous vein; v.ct.v., ventral cutaneous vein. [From Birch et u l . , 1969,J.2001.(London) 159, 31.1

490

DAVID R. JONES AND DAVID J. RANDALL

that passive rhythmic flexion in Heterodontus portusjacksoni only increased mean caudal vein pressure by 1 mm Hg. Hence, it seems unlikely that, due to the muscle pump, venous pressure will rise to such a level that the pressure difference for flow across the systemic vascular bed is substantially reduced. In fact, using an isolated trunk preparation perfused at constant pressure, Satchell (1965)obtained an increase in perfusate flow of 46% over the resting level (Fig. 28) during a l-min period when swimming movements were provoked b y stimulation of the spinal cord. There seems little doubt that specialized auxillary pumping mechanisms will also augment venous return in exercise. In elasmobranchs valved venous reservoirs are associated with the median fins (Mayer, 1888; Marples, 1935-1936; Birch et al., 1969) and fin movements probably pump blood from cutaneous vessels into the caudal vein. Mayer (1888) working with the dogfish [Scyliorhinus (Scyllium) canicula] observed that dorsal fins that had been congested with blood at rest paled as soon as movement occurred. Also in elasmobranchs there are paired, valved, interconnecting venous sinuses on each side of the tail (Fig. 29; Mayer, 1888; Birch et al., 1969). These sinuses are filled with blood squeezed from the posterior ends of the cutaneous veins by contraction of the surrounding myotome, the valve orientation ensuring that the blood flows into the caudal vein (Fig. 29). The caudal vein itself is protected from the effects of the waves of muscular movement during swimming since it is enclosed within the hemal canal and therefore its effectiveness in conducting blood to the heart

I MIN

Fig. 28. The effect of trunk movements on venous flow in the elasmobranch Heterodontus portusjacksoni. Upper trace, movement of postpelvic trunk. Lower trace, outflow of perfusate from caudal vein in 1 min. A, at rest; B, during movement; C, immediately following B. The perfusate (dextran-Ringer solution) was fed into the aorta of an isolated trunk preparation from a constant pressure reservoir and outflow was collected from the cannulated caudal vein. (From Satchell, 1965.)

7. T H E RESPIRATORY AND CIRCULATORY SYSTEMS

491

Fig. 29. The veins in the tail of an elasmobranch ( H . portusjacksod) with the blood flow pattern indicated by the arrows. d.ct.v., I.ct.v., v.ct.v., Dorsal, lateral and ventral cutaneous veins; u.c.sin., I.c.sin., upper and lower caudal sinuses. [From Birch et ul., 1969,J. Zool. (London) 159,31.1

from these auxillary pumps, as well as the generalized muscle pump, will be unimpaired in exercise. Other external influences such as respiratory movements or the hydrodynamic pressure gradients along the body surface of swimming fish (Dubois et al., 1974, 1976) may also influence venous return. The former might be expected to be more important in elasmobranchs where the location of the anterior cardinal, hyoidean, and inferior jugular sinuses indicate that they may be compressed in expiration and expanded in inspiration. I n many teleosts the largest body crosssectional area occurs in the region of the heart and, during swimming, significant negative pressures may occur on the body surface of this region (Dubois et al., 1974, 1976). If these pressures were transmitted by the body structures to the outside of the heart then the return of blood to the cardiac chambers would be augmented. Further, in posterior regions of the body the venous transmural pressure could be more or less doubled during rapid swimming. Fish achieve their highest swimming speeds in “bursts,” which are effected anaerobically, and are therefore independent of circulatory sufficiency. Consequently these hydrodynamic surface effects are only likely to be significant in fish which can maintain high swimming speeds (approximately 100 cm/sec) for long periods of time and, even then, the effect on tissue fluid balance, by changing capillary filtration pressures, may be more important than the changes in venous transmural pressure.

492

DAVID R. JONES AND DAVID J. RANDALL ACKNOWLEDGMENTS

D. R. J. was supported by grants from the NRC of Canada. D. J. R. was supported by grants from the NRC of Canada and Environment Canada. We are extremely grateful to Drs. N. H. West, W. Burggren, and C. Milliken for carefully reading early drafts of the manuscript and suggesting changes which have improved the final version and to Mrs. N. A. Virani for her patience in preparing the bibliography. REFERENCES Albers, C. (1970). Acid base balance. In “Fish Physiology” (W. S. Hoar and D. J. Randall, eds.), Vol. 4, p. 173. Academic Press, New York. Augustinsson, K. B., Fange, R., Johnels, A. and Ostlund, E. (1956). Histological, physiological and biochemical studies on the heart of two cyclostomes, hagfish (Myxine) and lamprey (Lampetra).J. Physiol. (London) 131,257. Auvergnat, R., and Secondat, M. (1942). Retentissement plasmatique de I’exercise musculaire chez la carpe (Cyprinus carpio L.). C . R. Acad. Sci. 215,92. Ballintijn, C. M., and Roberts, J. L. (1976). Neural control and proprioceptive load matching in reflex respiratory movements of fishes. Fed. Proc., Fed. Am. Soc. E x p . Biol. 3, 1983. Basu, S. P. (1959). Active respiration of fish in relation to ambient concentrations of oxygen and carbon dioxide./. Fish. Res. Board Can. 16, 175. Beamish, F. W. H. (1968). Glycogen and lactic acid concentrations in Atlantic cod (Gadus morhua) in relation to exercise. J . Fish. Res. Board Can. 25, 837. Belaud, A., and Peyraud, C. (1970). Actions de I’acetylcholine sur le coeur perfuse de congre (Conger conger L.) modifications de e.c.g. C . R . Soc. Biol. 146, 405. Bennion, G. R. (1968). The control of the function of the heart in teleost fish. M.S. Thesis, Univ. of British Columbia, Vancouver. Bettex-Galland, M., and Hughes, G. M. (1973). Contractile filamentous material in the pillar cells of fish gil1s.J. Cell Sci. 13, 359. Birch, M. P., Carre, C. G., and Satchell, G. H. (1969). Venous return in the trunk of the Port Jackson shark, Heterodontus portus jacksoni. J. Zool. (London) 159, 31. Black, E. C. (1958). Energy stores and metabolism in relation to muscular activity. In “The Investigation of Fish-power Problems” (P. A. Larkin, ed.), p. 51. Univ. of British Columbia Press, Vancouver. Black, E. C., Robertson, A. C., Hanslip, A. R., and Chiu, W. G. (1960). Alterations in glycogen, glucose and lactate in rainbow and Kamloops trout (Salmo gairdneri) following muscular activity.J. Fish. Res. Board Can. 17, 487. Black, E. C., Connor, A. R., Lam, K. C., and Chiu, W. G. (1962). Changes in glycogen, pyruvate and lactate in rainbow trout (Salmo gairdneri) during and following muscular activity.J. Fish. Res. Board Can. 19, 409. Bloom, G . , Ostlund, E., von Euler, U. S., Lishajko, F., Ritzen, M., and Adams-Ray, J. (1961). Studies on catecholamine-containing granules of specific cells in cyclostome hearts. Acta Physiol. Scand., S u p p l . No. 185, p. 1. Booth, J., and Holeton, G. F. (1977). Unpublished data. Brett, J. R. (1963). The energy required for swimming by young sockeye salmon with a comparison ofthe drag force on a dead fish. Trans. R. Soc. Can., Sect. 1 , 2 , 3 1,441. Brett, J. R. (1964). The respiratory metabolism and swimming performance of young sockeye sa1mon.J. Fish. Res. Board Can. 21, 1183.

7. T H E

RESPIRATORY AND CIRCULATORY SYSTEMS

493

Brett, J. R. (1965a).The relation of size to the rate of oxygen consumption and sustained swimming speeds of sockeye salmon (Oncorhynchus nerka).J . Fish. Res. Board Can. 22, 1491. Brett, J. R. (1965b). The swimming energetics of salmon. Sci. Am. 213, 80. Brett, J. R. (1972). The metabolic demand for oxygen in fish, particularly salmonids, and a comparison with other vertebrates. Respirat. Physiol. 14, 151. Brett, J . R., and Glass, N. R. (1973). Metabolic rates and critical swimming speeds of sockeye salmon (Oncorhynchus nerka) in relation to size and temperature. J . Fish. Res. Board Can. 30, 379. Brett, J. R., and Sutherland, D. B. (1965). Respiratory metabolism of pumpkinseed, Lepomis gibhosus, in relation to swimming speed.J. Fish. Res. Board Can. 22,405. Brown, C. E., and Muir, B. S. (1970). Analysis of ram ventilation of fish gills with application to skipjack tuna (Katsuwonus pelamis). J. Fish. Res. Board Can. 27, 1637. Brunings, W. (1899). Zur Physiologie des Kreislaufes der Fische. Pfluegers Arch. Gesamte Physiol. Menschen Tiere 75, 599. Byrne, J. M., Beamish, F. W. H., and Saunders, R. L. (1972). Influence of salinity, temperature, and exercise on plasma osmolality and ionic concentration in Atlantic salmon (Salrno salar).J. Fish. Res. Board Can. 29, 1217. Cameron, J. N. (1975). Morphometric and flow indicator studies. of the teleost heart. Can. J. Zool. 53, 691. Cameron, J. N., and Cech, J. J. (1970).Notes on the energy cost ofgill ventilation. Comp. Biochem. Physiol. 34, 447. Cameron, J. N., and Davis, J. C. (1970). Gas exchange in rainbow trout with varying blood oxygen capacity. J. Fish. Res. Board Can. 27, 1069. Cameron, J. N., Randall, D. J., and Davis, J. C. (1971). Regulation of the ventilationperfusion ratio in the gills of Dasyatis sabina and Squalus suckleyi. Comp. Biochem. Physiol. A 39, 505. Capra, M. F., a n d Satchel], G . M. (1974).Beta-adrenergic dilatory responses in isolated, saline perfused arteries of an elasmobranch fish, Squalus acanthias. Exfierientia 30, 927. Cardot, J., and Ripplinger, J. (1967). Action d e I’adrenaline e t de la noradrenaline sur la coeur lavi. du brochet e t de la carpe. J. Physiol. (Paris) 59, 399. Chapman, C. B., Jensen, D., and Wildenthal, K. (1963).On circulatory control mechanisms in the Pacific hagfish. Circ. Res. 12, 427. Cobb, J. L. S., and Santer, R. M. (1972). Excitatory and inhibitory inlervation of the heart of plaice (Pleuronectes platessa);anatomical and electrophysiological studies. J. Physiol. (London)222, 42. Cole, T. J., and Miller, G. J. (1973). Interpretation of the parameters relating oxygen uptake to heart rate and cardiac output during submaximal exercise. J . Physiol. (London) 231, 12 P. Connor, A. R., Elling, C. H., Black, E. C., Collines, G. B., Gauley, J. R.,and TrevorSmith, E. (1964). Changes in glycogen and lactate levels in migrating salmonid fish ascending experimental “endless” fishways. J. Fish. Res. Board Can. 21,255. Conte, F., Wagner, H. H., and Harris, T. (1963). Measurements of the blood volume of the fish, Salmo gairdneri. Am. J . Physiol. 205, 533. Dahlberg, M. L., Shumway, D. L., and Doudoroff, P. (1968). Influence of dissolved oxygen and carbon dioxide on swimming performance of largemouth bass and coho salmon. J . Fish. Res. Board Can. 25, 49. Davies, D. T., and Rankin, J. C. (1973).Adrenergic receptors and vascular responses to catecholamines of perfused dogfish gills. Comp. Gen. Phanacol. 4, 139.

494

DAVID R. JONES AND DAVID J. RANDALL

Davis, C. E., Foster, J., Warren, C. E., and Doudoroff, P. (1963). The influence of oxygen concentration on the swimming performance of juvenile Pacific salmon at various temperatures. Trans. Am. Fish. SOC. 92, 111. Davis, J. C. (1968).The influence of temperature and activity on certain cardiovascular and respiratory parameters in adult sockeye salmon. M.S. Thesis, Univ. of British Columbia, Vancouver. Davis, J. C. (1970).Estimation of circulation time in rainbow trout, Salmo gairdneri. J. Fish. Res. Board Can. 27, 1860. Davis, J. C., and Cameron, J. N. (1971). Water flow and gas exchange at the gills of rainbow trout, Salmo gairdneri. J. E x p . Biol. 54, 1. Davis, J. C., and W‘atters, K. (1970). Evaluation of opercular catheterization as a method for sampling water expired by fish.J. Fish. Res. Board Can. 27, 1627. Deck, K. A. (1964).Dehnungseffekte am spontanschlagenden, isoliertem Sinusknoten. Pfluegers Arch. Gesamte Physiol. Menschen Tiere 280, 120. Driedzic, W. R., and Kiceniuk, J. W. (1976). Blood lactate levels in free-swimming rainbow trout (Salmo gairdneri) before and after strenuous exercise resulting in fatigue.j. Fish. Res. Board Can. 33, 173. Dubois, A. B., Cavagna, C. A., and Fox, R. S. (1974).Pressure distribution on the body surface of a swimming fish.J. E x p . Biol. 60,581. Dubois, A. B., Cavagna, C. A., and Fox, R. S. (1976). Locomotion of bluefish./. E x p . Zool. 195, 223. Evans, D. H. (1969).Studies on the permeability to water of selected marine, freshwater and euryhaline te1eosts.J. E x p . Biol. 50, 689. Falck, B., von Mecklenburg, C., Myhrberg, H., and Persson, H. (1966). Studies on adrenergic and cholinergic receptors in the isolated hearts of Lampetra fluoiatilis (Cyclostomata) and Pleuronectes platessa (Teleostei). Acta Physiol. Scand. 68, 64. Farmer, C. J., and Beamish, F. W. H. (1969). Oxygen consumption ofTilapia nilotica in relation to swimming speed and sa1inity.J. Fish. Res. Board Can. 26, 2807. Fisher, T., Coburn, R.,and Forster, R. (1969).Carbon monoxide diffusing capacity in the bullhead catfish.J. Appl. Physiol. 26, 161. Fry, F. E. J. (1971). The effect of environmental factors on the physiology of fish. In “Fish Physiology” (W. S. Hoar and D. J. Randall, eds.), Vol. 6, p. 1. Academic Press, New York. Cannon, B. J. (1971).A study of the dual innervation of teleost heart by a field stimulation technique. Comp. Gen. Pharmacol. 2, 175. Cannon, B. J., and Burnstock, G . (1969). Excitatory adrenergic innervation of the fish heart. Comp. Biochem. Physiol. 29, 765. Cannon, B. J., Campbell, C. D., and Satchell, C. M. (1972).Monamine storage in relation to cardiac regulation in the Port Jackson shark, Heterodontus portusjacksoni. Z . Zellforsch. Mikrosk. Anat. 131, 437. Cannon, B. J., Campbell, C., Randall, D. J., and Smith, D. C. (1978a).The vasculature of the gill of rainbow trout, Salmo gairdneri. I. The respiratory vascular bed. In preparation. Cannon, B. J., Campbell, C., Randall, D. J.. and Smith, D. C. (197813).The vasculature of the gill of the rainbow trout, Salmo gairdneri. 11. Gill structures related to blood flow regulation and gas exchange in the respiratory vascular bed. In preparation. ) . vasculature of Cannon, B. J., Campbell, C., Randall, D. J., and Smith, D. C. ( 1 9 7 8 ~ The the gills of the rainbow trout, Salmo gairdneri. 111. Nonrespiratory blood channels. In preparation. Cesser, H., and Poupa, 0. (1974).Relations between heart muscle enzyme pattern and directly measured tolerance to acute anoxia. Comp. Biochem. Physwl. A 48, 97.

7. T H E

RESPIRATORY AND CIRCULATORY SYSTEMS

495

Golenhofen, K., and Lippross, H. (1969). Mechanische Koppelungswirkungen der Atmung auf den Herzschlag. Pfluegers Arch. 309, 156. Hall, F. G. (1930).The ability of the common mackerel and certain other marine fishes to remove dissolved oxygen from sea water. Am, J. Physiol. 93,417. Hanson, D. (1967). Cardiovascular dynamics and aspects of gas exchange in chondrichthyes. Ph.D. Thesis, Univ. of Washington, Seattle. Harris, W. S., and Morton, M. J. (1968).A cardiac intrinsic mechanism that relates heart rate to filling pressure. Circulation, Suppl. 6, 95. Haswell, M . S., and Randall, D. J. (1976).Carbonic anhydrase inhibition in troutplasma. Respirat. Physiol. 28, 17. Haswell, M. S., and Randall, D. J. (1978).The pattern of carbon dioxide excretion in the rainbow trout (Salmo gairdneri).J. E x p . Biol. 72, 17. Heath, A. G. (1973).Ventilatory responses of teleost fish to exercise and thermal stress. Am. Zool. 13, 490. Hemmingsen, E. A., and Douglas, E . L. (1970).Respiratory characteristics of the hemoglobin free fish, Chaenocephalus aceratus. Comp. Biochem. Physiol. 33, 733. Hemmingsen, E. A., Douglas, E. L., Johansen, K., and Millard, R. W. (1972). Aortic blood flow and cardiac output in the hemoglobin free fish, Chaenocephalus aceratus. Comp. Biochem. Physiol. A 43, 1945. Hill, A. K. (1950). The dimensions of animals and their muscular dynamics. Sci. Prog. (London) 38, 209. Hirsch, E. F., Jellinek, M., and Cooper, T. (1964).Innervation of the systemic heart of the California hagfish. Circ. Res. 14, 212. Hochachka, P. W. (1961).The effect of physical training on oxygen debt and glycogen reserves in trout. Can. J. Zool. 39, 767. Hoffert, J. R., and Fromm, P. 0. (1966). Effect of carbonic anhydrase inhibition on aqueous humor and blood bicarbonate ion in the teleost (Salvelinus namaycush). Comp. Biochem. Physiol. 18, 333. Holeton, G . F., and Randall, D. J. (1967a). Changes in blood pressure in the rainbow trout during hypoxia. J. E x p . Biol. 46, 297. Holeton, G . F., and Randall, D. J . (196%). The effect of hypoxia upon the partial pressure of gases in the blood and water afferent and efferent to the gills of rainbow trout. J. E x p . Biol. 46, 317. Holmgren, S., and Nilsson, S. (1974). Drug effects on isolated artery strips from two teleosts, Cadus morhua andSalmo gairdneri. Acta Physiol. Scand. 90, 431. Holmgren, S., and Nilsson, S. (1975).Effects of some adrenergic and cholinergic drugs on isolated spleen strips from the cod, Cadus morhua. Eur. J . Pharmacol. 32, 163. Hudson, R. C. L. (1973).On the function of the white muscles in teleosts a t intermediate swimming speeds. J. E x p . Biol. 58, 509. Hughes, G. M. (1970). Morphological measurements on the gills of fishes in relation to their respiratory function. Folia Morphol. (Prague) 18, 78. Hughes, G . M. (1972). Morphometrics of fish gills. Respirat. Physiol. 14, 1. Hughes, G. M., and Grimstone, A. V. (1965).The fine structure ofthe secondary lamellae of the gills of Gadus pollachius. Quart. J. Microsc. Sci. 106, 343. Hughes, G. M., and Morgan, M. (1973). The structure of fish gills in relation to their respiratory function. Biol. Rev. Cambridge Philos. Soc. 48, 419. Hughes, G. M., and Saunders, R. L. (1970). Responses of the respiratory pumps to hypoxia in the rainbow trout (Salmo gairdneri)./.E x p . Biol. 53, 527. Irving, L., Solandt, D. T., and Solandt, 0. hd. (1935). Nerve impulses from branchial pressure receptors in the dogfish. J. Physiol. (London) 84, 187. Itazawa, T. (1970). Characteristics of respiration of fish considered from the arteriovenous difference of oxygen content. Nippon Suisan Gakkaishi 36, 57.

496

DAVID R. JONES AND DAVID J. RANDALL

Jensen, D. (1961). Cardioregulation in an aneural heart. Comp. Biochem. Physiol. 2,181. Jensen, D. (1963).Eptatretin: A potent cardioactive agent from the branchial heart of the Pacific hagfish Eptatretus stoutii. Comp. Biochem. Physiol. 10, 129. Jensen, D. (1965). The aneural heart of the hagfish. Ann. N.Y. Acad. Sci. 127,443. Jensen, D. (1969). Intrinsic cardiac rate regulation in the sea lamprey, Petromyzon marinus, and rainbow trout, Salmo gairdneri. Comp. Biochem. Physwl. 30, 685. Jensen, D. (1970).Intrinsic cardiac rate regulation in elasmobranch: The homed shark, Heterodontus prancisci, and thornback ray, Platyrhinoidis triseriata. Comp. Biochem. Physiol. 34, 289. Johansen, K. (1962). Cardiac output and pulsatile aortic flow in the teleost Gadus morhua. Comp. Biochem. Physiol. 7, 169. Johansen, K. (19654. Dynamics of venous return in elasmobranch fishes. Hualradets Skr. 48, 94. Johansen, K. (1965b).Cardiovascular dynamics in fishes, amphibians, and reptiles. Ann. N.Y. Acad. Sci. 127, 414. Johansen, K. (1972). Heart and circulation in gill, skin, and lung breathing. Respirat. Physwl. 14, 193. Johansen, K., and Hanson, D. (1967). Hepatic vein sphincters in elasmobranchs and their significance in controlling hepatic blood flow. J . E x p . Biol. 46, 195. Johansen, K., and Waage-Johannessen, N. (1962).Unpublished observations. Johansen, K., Franklin, D. L., and Van Citters, R. L. (1966). Aortic blood flow in freeswimming elasmobranchs. Comp. Biochem. Physiol. 19, 151. Jones, D. R. (1971a). Theoretical analysis of factors which may limit the maximum oxygen uptake of fish. The oxygen cost of the cardiac and branchial pumps. J . Theor. Biol. 32, 341. Jones, D. R. (1971b). The effect of hypoxia and anaemia on the swimming performance of rainbow trout (Salmo gairdneri).]. E x p . Biol. 55, 541. Jones, D. R., and Schwarzfeld, T. (1974).The oxygen cost to the metabolism and efficiency of breathing in trout (Salmo gairdneri). Respirat. Physiol. 21, 241. Jones, D. R., Randall, D. J., and Jarman, G. M. (1970). A graphical analysis of oxygen transfer in fish. Respirat. Physiol. 10, 285. Jones, D. R., Langille, B. L., Randall, D. J., and Shelton, G. (1974).Blood flow in dorsal and ventral aortas of the cod, Gadus morhua. Am. J . Physiol. 226,90. Keys, A., and Bateman, J. B. (1932).Branchial response to adrenaline and pitressin in the eel. Biol. Bull. (Woods Hole, Mass.) 63, 327. Kiceniuk, J. W., and Jones, D. R. (1977).The oxygen transport system in trout (Salmo gairdneri) during sustained exercise. J . E x p . Biol. 69, 247. Kirby, S., and Burnstock, G. (1969). Comparative pharmacological studies of isolated spiral strips of large arteries from lower vertebrates. Comp. Biochem. Physiol. 28, 307. Kirschner, L. B. (1969).Ventral aortic pressure and sodium fluxes in perfused eel gills. Am. J . Physiol. 217, 596. Klaverkamp, J. F., and Dyer, D. P. (1974). Autonomic receptors in isolated rainbow trout vasculature. Eur. J . Pharmacol. 28, 25. Kunos, G., and Nickerson, M. (1976). Temperature-induced interconversion of a- and 8-adrenoreceptors in the frog heart. J . Physiol. (London)256, 23. Kutty, M. N. (1968a). Respiratory quotients in goldfish and rainbow trout. J . Fish. Res. Board Can. 25, 1689. Kutty, M. N. (1968b). Influence of ambient oxygen on the swimming performance of goldfish and rainbow trout. Can.J . Zool. 46,647. Labat, R. (1964). Action de I’adrenaline sur la frequence cardiaque d e Pleuronectes vagotomises. C . R . SOC. Biol. 158, 371.

7. T H E

RESPIRATORY AND CIRCULATORY SYSTEMS

497

Labat, R.(1966).Electrocardiologie chez les Poisson tClCost6ens: Influence de quelques facteur ecologiques. Ann. Limnol. 2, 1. Labat, R., Raynaud, P., and Serfaty, A. (1961). Reactions cardiaques et variations d e masse sanguine chez les T616ost6ens. Comp. Biochem. Physwl. 4, 75. Laffont, J., and Labat, R.(1966).Action de l’adrenaline sure la frequence cardiaque de la Carpe commune. Effet de la temperature du milieu sur I’intensiti. de la reaction. J . Physiol. (Paris) 58, 351. Landis, E. M., and Pappenheimer. J. R. (1963). Exchange of substances through the capillary walls. In “Handbook of Physiology: Circulation” (W. F. Hamilton and N. F. Dow, eds.), Vol. 2, p. 961. Am. Physiol. SOC.,Washington, D.C. Laurent, P., and Dunel, S. (1976). Functional organization of the teleost gill. I. Blood pathways. Acta Zool. (Stockholm) 57, 189. Lutz, B. R.,and Wyman, L. C. (1932).The effect of adrenaline on the blood pressure of the elasmobranch, Squalus acanthias. Biol. Bull. (Woods Hole, Mass.) 62, 17. Maetz, J . (1973). Na+/NH:, Na+/H+ exchanges and NH, movements across the gill of Carassius auratus. J . E x p . Biol. 58, 255. Magnuson, J. J. (1963). Tuna behavior and physiology, a review. Proc. World Sci. Meet., Biol. Tunas Relat. Species, La Jolla, Callf., 1962 3, 1057. Marples, B. J. (1935-1936). The blood vascular system of the elasmobranch fish Squatina squattna (Linne). Trans. R . SOC. Edinburgh 58, 817. Mayer, P. ( 1888). Uber Eigentumlichkeiten in den Kreislauforganen der Selachiern. Mitt. Zool. Stat. Neapel., Leipzig 8, 307. Mazeaud, M. M. (1969).Influence de stress sur les teneurs en catecholamines du plasma et des corps axillaires chez un Selacian, la Rousette (Scyliorhinus canicula L.). C . R . Soc. Biol. 163, 2262. Mazeaud, M. M., Mazeaud, F., and Donaldson, E. M. (1978). Stress resulting from handling in fish: Primary and secondary effects. Trans. Am. Fish. SOC.(in press). Millen, J. E., Murdaugh, H. V., Heam, D. C., and Robin, E. D. (1966). Measurement of gill water flow in Squalus acanthias using the dye-dilution technique. Am. J . Physiol. 211, 11. Mott, J . C. (1950). Radiological observations on the cardiovascular system in Anguilla anguilla. J . E x p . Biol. 27, 324. Mott, J. C. (1951). Some factors affecting the blood circulation in the commop eel (Anguilla anguilla).J . Physiol. (London) 114, 387. Muir, B. S., and Buckley, R. M. (1967). Gill ventilation in Remora remora. Copeia 3, 581. Muir, B. S., and Niimi, A. J. (1972). Oxygen consumption of the euryhaline fish Aholehole (Kublia sandoicensis) with reference to salinity, swimming, and food consumption. J . Fish. Res. Board Can. 29, 67. Nakano, T., and Tomlinson, N. (1967).Catecholamine and carbohydrate concentrations in rainbow trout (Salmo gairdneri) in relation to physical disturbance. J . Fish. Res. Board Can. 24, 1701. Newstead, J. D. (1967).Fine structure of the respiratory lamellae of teleostean gills-Z. Zellforsch. Mikrosk. Anat. 79, 396. Nilsson, S., Holmgren, S., and Grove, J. D. (1975).Effects of drugs and nerve stimulation on the spleen and arteries of two species of dogfish, Scyliorhinus canicula and Squalus acanthias. Acta Physiol. Scand. 95, 219. Nilsson, S., Abrahamsson, T., and Grove, J. D. (1976).Sympathetic control of adrenaline release from the chromaffin tissue in a fish. Acta Physiol. Scand. 96, C11. Opdyke, D. F., McCreehan, J. R.,Messing, S., and Opdyke, N. E . (1972).Cardiovascular responses to spinal cord stimulation and autonomically active drugs in Squalus acanthias. Comp. Biochem. Physiol. A 42, 611.

498

DAVID R. JONES AND DAVID J. RANDALL

OShdal, B., and Schiebler, T. H. (1971).The terminal blood bed in the heart of the fish and in the heart of the turtle. 2.Anat. Entwicklungsgesch. 134, 101. Ostliind, E.,and Fange, R. (1962).Vasodilation by adrenaline and noradrenaline and the effects of some other substances on perfused fish gills. Comp. Biochem. Physiol. 5,

307. Pathak, C.L.(1972).Stretch sensitive intrinsic autoregulatory mechanisms for rhythmicity and contractility of the heart. Experientia 28, 650. Piiper, J., and Baumgarten-Schumann, D. (1968).Transport of 0, and CO, by water and blood in gas exchange of the dogfish (Scyliorhinus stellaris).Respirat. Physiol. 5,

326. Priede, I. G. (1974).The effect of swimming activity and section of the vagus nerves on heart rate in rainbow tr0ut.J. E x p . Biol. 60,305. Rahn, H. (1966).Aquatic gas exchange: Theory. Respirat. Physiol. 1, 1. Randall, D.J. (1968).Functional morphology of the heart in fishes. Am. Zool. 8, 179. Randall, D.J . (19704.The circulatory system. In “Fish Physiology” (W. S. Hoar and D. J. Randall, eds.), Vol. 4,p. 133. Academic Press, New York. Randall, D. J. (1970b).Gas exchange in fishes. In “Fish Physiology” (W. S. Hoar and D. J. Randall, eds.), Vol. 4, p. 253.Academic Press, New York. Randall, D. J., and Smith, J. C. (1967).The regulation of cardiac activity in fish in a hypoxic environment. Physwl. 2001.40, 104. Randall, D. J., and Stevens, E. D. (1967).The role of adrenergic receptors in cardiovascular changes associated with exercise in salmon. Comp. Biochem. Physiol. 21,

415. Randall, D. J., Holeton, G. F., and Stevens, E. D. (1967).The exchange of 0, and CO, across the gills of rainbow trout.]. Exp. Biol. 46, 339. Randall, D.,Heisler, N., and Drees, F. (1976).Ventilatory response to hypercapnia in the larger spotted dogfish, Scyliwhinus stellaris. Am. J . Physiol. 230, 590. Randall, D.J., Milliken, C., and Haswell, M. S. (1978).Permeability of rainbow trout red blood cells to lactate. In preparation. Rankin, J. C. (1976).Personal communication. Rankin, J. C., and Maetz, J. (1971).A perfused teleostean gill preparation: Vascular actions of neurohypophysial homones and catecholamines.]. Endocrinol. 51,621. Rao, G . M. M. (1968).Oxygen consumption of rainbow trout (Salmo gairdneri) in relation to activity and salinity. Can. J. Zool. 46, 781. Reite, 0. B. (1969).The evolution of vascular smooth muscle responses to histamine and 5hydroxytryptamine. I. Occurrence of stimulatory actions in fish. Acta Physwl. Scand. 75,221. Richards, B. D., and Fromm, P. 0.(1969).Patterns of blood flow through filaments and lamellae of isolated-perfused rainbow trout (Salmo gairdneri) gills. Comp. Biochem. Physiol. 29, 1063. Roberts, B. L. (1969).The spinal nerves of the dogfish (Scyliorhinus).J.Mar. Biol. Ass.

U.K.49, 105. Roberts, B. L.(1972).Activity of lateral-line sense organs in swimming d0gfish.j. E x p . Biol. 56, 105. Roberts, J. L. (1975a).Active branchial and ram gill ventilation in fishes. Biol. Bull. (Woods Hole, Mass.) 148, 85. Roberts, J. L. (197513).Cardio-ventilatory interactions during swimming, and during thermal and hypoxic stress. Respirat. Mar. Organisms, Proc. Mar. Sect., 1st Mar. Biomed. Sci. Symp. p. 139. Rowell, L. B. (1974).Human cardiovascular adjustments to exercise and thermal stress. Physiol. Rev. 54, 75.

7. T H E

RESPIRATORY AND CIRCULATORY SYSTEMS

499

Santer, R. M., and Cobb, J. L. S. (1972). The fine structure of the heart of the teleost, Pleuronectes platessa L. Z. Zellforsch. Mikrosk. Anat. 131, 1. Satchell, G. H. (1965). Blood flow through the caudal vein of elasmobranch fish.Aust.J. Sci. 27, 241. Satchell, G . H. (1968). A neurological basis for the coordination of swimming with respiration in fish. Comp. Biochem. Physiol. 27, 835. Satchell, G . H. (1971). “Circulation in Fishes.” Cambridge Univ. Press, London and New York. Saunders, R. L. (1962).The irrigation of gills in fishes. 11. Efficiency of oxygen uptake in relation to respiratory flow, activity, and concentrations of oxygen and carbon dioxide. Can. J. Zool. 40, 817. Schmidt-Nielsen, K. (1972).Locomotion: Energy cost of swimming, flying and running. Science 177, 227. Secondat, M. (1950).Influence de I’exercise musculaire sur la valeur de la glycbme de la carpe (Cyprinus carpio L.). C . R. Acad. Sci. 231, 796. Sbelton, G. (1970).The regulation of breathing. In “Fish Physiology” (W. S. Hoar and D. J. Randall, eds.), Vol. 4, p. 293. Academic Press, New York. Silver, S. J.. Warren, C. E., and Doudoroff, P. (1963). Dissolved oxygen requirements of developing steelhead trout and chinook salmon embryos at different water velocities. Trans. Am. Fish. Soc. 92, 327. Smit, H. (1965). Some experiments on the oxygen consumption of goldfish (Carassius auratus L.) in relation to swimming speed. Can. J . Zool. 43, 623. Smit, H., Amelink-Koutstaal, J. M., Vuverberg, J., and von Vaupel-Klein, J. C. (1971). Oxygen consumption and efficiency of swimming goldfish. Comp. Biochem. Physiol. A 39, 1. Smith, D. G . (1976).The structure and function of the respiratory organs of some lower vertebrates. Ph.D. Thesis, Vols. 1 and 2. Zool. Dep., Univ. of Melbourne, Melbourne. Smith, L. S. (1966).Blood volumes of three sa1monids.j. Fish. Res. Board Can. 23, 1439. Smith, L. S., and Bell, G . R. (1964). A technique for prolonged blood sampling in free-swimming salmon. J . Fish. Res. Board Can. 32, 711. Smith, L. S., and Bell, G. R. (1967). Anesthetic and surgical techniques for Pacific sa1mon.J. Fish. Res. Board Can. 24, 1579. Smith, L. S., Brett, J. R., and Davis, J. C. (1967). Cardiovascular dynamics in swimming adult sockeye salmon. J . Fish. Res. Board Can. 24, 1775. Starling, E. H. (1918). “The Linacre Lecture on the Law of the Heart, Given at Cambridge, 1915.” Longmans, Green, London. Steen, J . B., and Kruysse, A. (1964). The respiratory function of teleostean gills. Comp. Biochem. Physiol. 12, 127. Steen, J. B., and Stray-Pedersen, S. (1975). The permeability of fish gills with comments on the osmotic behavior of cellular membranes. Acta Physiol. Scand. 95, 6. Stevens, E. D. (1968). The effect of exercise on the distribution of blood to various organs in rainbow trout. Comp. Biochem. Physiol. 25, 615. Stevens, E. D., and Black, E. C. (1966).The effect of intermittent exercise on carbohydrate metabolism in rainbow trout, Salmo gairdneri. J. Fish. Res. Board Can. 23, 471. Stevens, E. D., and Randall, D. J. (1967a). Changes in blood pressure, heart rate, and breathing rate during moderate swimming activity in rainbow tr0ut.j. Enp. Biol. 46, 307. Stevens, E. D., and Randall, D. J. (1967b). Changes in gas concentrations in blood and water during moderate swimming activity in rainbow trout. J . E x p . Biol. 46, 329.

500

DAVID R. JONES AND DAVID J. RANDALL

Stevens, E. D., and Sutterlin, A. M. (1975). A technique for measuring heat exchange across the gills of a teleost, Hemitripterus americanus. Respirat. Mar. Organisms. Proc. Mar. Sect., 1st Mar. Biomed. Sci. Symp. p. 195. Stevens, E. D., Bennion, G. R., Randall, D. J., and Shelton, G . (1972). Factors affecting arterial pressures and blood flow from the heart in intact, unrestrained lingcod, Ophiodon elongatus. Comp. Biochem. Physiol. A 43,681. Sudak, F.N. (1965a). Intrapericardial and intracardiac pressures and events of the cardiac cycle in Mustelus canis (Mitchill). Comp. Biochem. Physiol. 14, 689. Sudak, F. N. (196513).Some factors contributing to the development of subatmospheric pressure in the heart chambers and pericardial cavity of Mustelus canis (Mitchill). Comp. Biochem. Physiol. 15, 199. Sutterlin, A. M. (1969). Effects of exercise on cardiac and ventilation frequency in three species of freshwater teleosts. Physiol. Zool. 42, 36. Sutterlin, A. M., and Saunders, R. L. (1969). Proprioceptors in the gills of teleosts. Can. J . Zool. 47, 1209. Taylor, W., Houston, A. H., and Horgan, J. D. (1968). Development of a computer model simulating some aspects of the cardiovascular-respiratory dynamics of the salmonid fish. J . E x p . Biol. 49, 477. Thorson, T. B. (1959). Partitioning of body water in sea lamprey. Science 130, 99. Thorson, T. B. (1960). The partitioning ofbody water in Osteichthyes: Phylogenetic and ecological implications in aquatic vertebrates. Biol. Bull. (Woods.Hole, Mass.) 120, 238. Tucker, V. A. (1970). Energetic cost of locomotion in animals. Comp. Biochem. Physbl. 34, 345. Tytler, P. (1969). Relationship between oxygen consumption and swimming speed in the haddock, Melanogrammus aeglejinus. Nature (London) 221, 274. Ultsch, G. R. (1973). A theoretical and experimental investigation of the relationships between metabolic rate, body size, and oxygen exchange capacity. Respirat. Physiol. 18, 143. van Dam, L. (1938). On the utilization of oxygen and regulation of breathing in some aquatic animals. Ph.D. Thesis, Univ. of Groningen, Groningen, Netherlands. Voboril, Z., and Schiebler, T. H. (1970). The blood supply of fish hearts. Z. Anat. Entwicklungsgesch. 130, 1. Vogel, W., Vogel, V., and Keemers, H. (1973). New aspects ofthe intrafilamental vascular system in gills of a euryhaline teleost Tilapia mossambica. Z. Zellforsch. Mikrosk. Anat. 144, 573. Vogel, W., Vogel, V., and Pfautsch, M. (1976). Arterio-venous anastomoses in rainbow trout gill filaments. Cell Tissue Res. 167, 373. von Holst, E. (1937). Vom Wesen der Ordnung im Zentralnervensystem. Naturwissenschaften 40, 641. Webb, P. W. (1970). Some aspects of the energetics of swimming of fish with special reference to the cruising performance of rainbow trout. Ph.D. Thesis, Univ. of Bristol, Bristol, England. Webb, P. W. (1971a). The swimming energetics of trout. I. Thrust and power output at cruising speeds. J . E x p . Biol. 55, 489. Webb, P. W. (1971b). The swimming energetics of trout. 11. Oxygen consumption and swimming efficiency. J . E x p . Biol. 55, 521. Webb, P. W. (1975a). Hydrodynamics and energetics of fish propulsion. Bull., Fish. Res. Board Can. No. 190, 158 pp. Webb, P. W. (1975b). Synchrony of locomotion and ventilation in Cymatogaster aggregata. Can. J . Zool. 53, 904.

7. THE

RESPIRATORY AND CIRCULATORY SYSTEMS

50 1

Weihs, D. (19734. Optimal cruising speed for migrating fish. Nature (London)245,48. Weihs, D. (197%). Mechanically efficient swimming techniques for fish with negative buoyancy. J . Mar. Res. 31, 194. Weihs, D. (1974). Energetic advantages of burst swimming of fish./. Theor. Biol. 4 8 , l . West, N. H., and Jones, D. R. (1975). Unpublished observations. Wood, C. M. (19744. A critical examination of the physical and adrenergic factors affecting blood flow through the gills of the rainbow tr0ut.J. Exp. Biol. 60, 241. Wood, C. M. (1974b). Mayer waves in the circulation of a teleost fish. J . Exp. Zool. 189, 267. Wood, C. M. (1975). A pharmacological analysis of the adrenergic and cholinergic mechanisms regulating branchial vascular resistance in the rainbow trout (Salmo gairdneri). Can. J . Zool. 53, 1569. Wood, C. M., and Randall, D. J. (19734. The influence of swimming activity on sodium balance in the rainbow trout (Salmo gairdneri).J.Comp. Physiol. 82, 207. Wood, C. M., and Randall, D. J. (197313). Sodium. balance in the rainbow trout (Salmo gairdneri) during extended exercise. J. Comp. Physiol. 82, 235. Wood, C. M., and Randall, D. J. ( 1973~)The . influence of swimming activity on water balance in the rainbow trout (Salmo gairdneri).J . Comp. Physiol. 82, 257. Wood, C. M., and Shelton, G . (1975). Physical and adrenergic factors affecting systemic vascular resistance in the rainbow trout: A comparison with branchial vascular resistance.J. E x p . Biol. 63, 505. Wood, C. M., McMahon, B. R., and McDonald, D. G . (1977). An analysis of changes in blood pH following exhausting activity in the starry flounder (Platichthys stellatus). J. E x p . Biol. 69, 173.

This Page Intentionally Left Blank

METABOLISM IN FISH DURING EXERCISE WlLLlAM R . DRIEDZIC and P. W. HOCHACHKA I. Introduction

. . ............. .. ... .. . .,. ,... ............ ..... .. .

.

11. Biochemical Insights from Respiratory Physiology . . . . . . . . . . . . . . 111. Red-White Muscle Differences . . . . . . . . . . . . . . . . . . . . . . . . . , . . . . . IV. Metabolism of Adenylates and Related Compounds.. . . . . . . . . . . A. Studies at the Metabolite L e v e l . . . . . . . . . . . . . . . . . . . . . . . . .. .

. .. . . B. Studies at the Enzyme Level . . . . . . . . . . . . . . . . . . . . . . , . , . . . . . . C. Control ofthe Adenylate Pool . . . . . . . . . . . . , . . , . , . , . . . , . . . . . . D. Recovery of the Adenylate Pool in White Muscle following Activity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . , , . . . . . . . . . V. Carbohydrate Metabolism . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A. Studies at the Metabolite Level.. . . . . . . . . . . . . , . . . . . . . . . . . . . . B. Studies at the Enzyme Level . . . . . . . , . . . . . . . . . . . . . . . . . . . . . . . C. Control of Carbohydrate Metabolism . . . . . . . . . . . . . . . . . . . . . . . . D. Fate of Lactate Produced in White Muscle . . . . . . . . . . . . . . . . . . ............................. VI. Lipid Metabolism . . . . . . . . . . . A. General Observations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . , . , . . . .. . B. Control of Lipid Metabolism . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . VII. Protein Metabolism . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . VIII. Citric Acid Cycle . . . . . , . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . , . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . .

503 504 505 507 507 510 513 515 517 517 519 523 524 525 525 526 530 533 536

I. INTRODUCTION

The purpose of the present essay is to describe the biochemical events involved in the transition from a resting to an actively swimming condition. More specifically, we shall discuss the mechanisms of regulation of the various fuels of metabolism. Intermediary metabolism in fishes has been reviewed by both Hochachka (1969) and Tarr (1972) and the quantitative aspect of substrate utilization by fish during activity has been summarized by Love (1970)and Bilinski (1974). It is our intent to attempt to bridge the gap between our knowledge of what fuels are being utilized and the enzymatic control over this process with respect to altered energy demands. 503 FISH PHYSIOLOGY, VOL. V l l Copyright @ 1978 by Academic Press, Inc. All rights of reproduction in any form reserved. ISBN 0-12-350407-4

504

WILLIAM R. DRIEDZIC AND P. W. HOCHACHKA

The major thrust of this paper will, by necessity, deal with metabolism in myotomal muscle but there are other organs, integralIy important for the maintenance of homeostasis, which may be extremely expensive to operate during sustained swimming. For instance, the cost of osmoregulation, in freshwater, for actively swimming rainbow trout (Salmo gairdneri), may be as high as 30% of the total metabolism of the fish (Rao, 1968). Much of this energy expenditure must, of course, be borne by the gills. The cost of maintaining the cardiac pump may also be quite substantial. In fact, at high oxygen consumption rates this may approach 15% of the total metabolism of the animal (Jones, 1971). Thus, the myotomal muscle per se may account for only about 50% of the total energy output. In light of this, it is somewhat surprising that tissues other than skeletal muscle are rarely studied by fish physiologists interested in metabolic alterations during swimming. We shall endeavor to cover what is known about the control of metabolism in these support tissues and skeletal muscle; however, our understanding of metabolism in fish tissues is at best fragmentary. Hence, where appropriate, the better studied mammalian systems will be called upon to complement and supplement the discussion.

11. BIOCHEMICAL INSIGHTS FROM RESPIRATORY PHYSIOLOGY The findings of the respiratory physiologist are of particular interest to the biochemist concerned with metabolic problems, for the relationship between inspired and expired gases sets the biochemical limit of the organism. The oxygen consumption of an actively swimming fish may increase by about l@fold over its resting level (see Chapter 2).As we shall argue below, most of this increase must be due to red muscle activity. Since the red muscle represents only about 5% of the total body weight of the fish (Stevens, 1968; Webb, 1970; Smit et al., 1971; Johnston and Goldspink, 1973a), the increase in oxygen consumption of this tissue alone must be very high. The implications of this shall be discussed under the section dealing with the citric acid cycle. The respiratory quotient (R.Q.) (the molar ratio of the amount of CO, expired to the amount of O2 consumed) is of value to the biochemist for it provides an insight into the fuel being utilized under particular conditions. The R.Q. for the catabolism of carbohydrates, fats, and proteins is 1.0,0.7, and about 0.8, respectively. It is difficult to calculate the R.Q. for proteins since it is not known which amino acids,

8. METABOLISM

IN FISH DURING EXERCISE

505

if any, are preferentially utilized by fish. Under conditions of routine metabolism the R.Q. of rainbow trout is about 0.85 (Kutty, 1968); of Tilapia mossambica and mullet (Rhinomugil corsula) about 0.9 (Kutty, 1972; Kutty and Mohamed, 1975); and of goldfish (Carassius auratus) about 1.0 (Kutty, 1968). These values indicate that in a low activity state trout, Tilapia, and mullet utilize carbohydrates along with lipids and/or proteins which have an R.Q. value lower than 1.0. Goldfish, on the other hand, appear to have a very low turnover rate of “low R.Q.” fuels during routine metabolism. Alterations in R.Q. with forced swimming have been studied in trout and goldfish (Kutty, 1968). In trout, as the animal swims faster, the R.Q. increases to a value of about 1.0 at sustained and prolonged swimming speeds, thus suggesting an increased utilization of carbohydrate with an increase in metabolic rate due to activity. The response of goldfish, however, is much different. As the goldfish swims faster, the R.Q. drops to about 0.85 at prolonged swimming velocities of 6.3 cm/sec (body length, 3.6 cm). This indicates a relative decrease in the utilization of carbohydrate as an aerobic fuel source as metabolic rate increases. Thus, for the goldfish, fats and/or proteins are becoming increasingly important at sustained and prolonged swimming speeds. There is a limited amount of information on the expiration of NH4+ with respect to oxygen consumption. However, the data which do exist are of extreme interest. Kutty (1972) noted that, with Tilapia, the amount of NH4+ excreted in relation to the amount of oxygen consumed is so high as to suggest that this animal is utilizing amino acids (proteins) as the sole carbon and energy source. Furthermore, during exercise at a constant swimming speed the rate of NH4+ excretion continued to increase. The utilization of proteins as a controlled energy source as opposed to simple protein turnover is an interesting concept and should at the present not be discarded. Evidence in support of this contention will be discussed in Section VII.

111. RED-WHITE MUSCLE DIFFERENCES The myotomal muscle of fish is largely composed of two tissue types, usually termed the red and white fibers. A third intermediate type of muscle fiber also exists, but where it has been observed it is much less abundant than the other two and for the present, we are forced to neglect it. In mammals red, white, and intermediate type fibers usually occur in complex mixed muscle masses and in fact, this is the case with some fish species such as the salmonids. In rainbow trout, for instance, 8% of the myotomal mass is red muscle and 75% of

506

WILLIAM R. DRIEDZIC AND P. W. HOCHACHKA

this red muscle is interdigitated within the white fibers (Webb, 1970). This obviously creates a problem for the biochemist attempting to sort out red-white muscle differences. For the purpose of the present essay we shall refer to the major muscle mass of salmonids as “white” muscle; however, it should be recognized that this tissue contains a small percentage of red fibers. In many fish, the white fibers, which constitute 80-95% of the swimming musculature, exist as a discrete, easily separable tissue mass. Where this occurs the red fibers are often found as a thin superficial layer below the skin forming a thicker triangle of muscle at the level of the lateral line, with the white fibers making up the mass of the underlying myotome. Finally, in some very active species an additional band of red muscle is found near the spine. Love ( 1970) comments on differences between the peripheral red muscle and the deep-seated red muscle; however, in terms of biochemical energetics, there is very little we can say about the latter tissue and hence shall not make it the subject of further comment here. One final statement regarding the operational definitions of myotomal musculature is necessary. In many publications no reference is made to the type of muscle sampled, although in many cases it can be assumed, based on the anatomy of the fish, that the tissue sample was largely white muscle. In cases such as this we shall again refer to the tissue as “white” muscle. On the basis of biochemical and histological properties it is generally accepted that red muscle has a metabolism which is predominantly aerobic, whereas white muscle functions largely anaerobically. Thus, red muscle is characterized by a higher content of mitochondria (Buttkus, 1963; George, 1962), myoglobin (Hamoir et al., 1972), hemoglobin (Hamoir et al., 1972), lipid (Bone, 1966), lipolytic enzymes (George, 1962), citric acid cycle enzymes (Bostrom and Johansson, 1972), and cytochrome oxidase (Bostrom and Johansson, 1972). These properties are reflected by a high vascular supply to the tissue (Stevens, 1968), a greater capacity to consume oxygen (Wittenberger and Diaciuc, 1965; Lin et al., 1974), and a higher rate of lipid oxidation (Bilinski, 1969). White muscle, on the other hand, is poorly vascularized (Stevens, 1968), shows a low oxygen consumption rate (Wittenberger and Diaciuc, 1965; Lin et al., 1974), and is biochemically geared for anaerobic metabolism. Consequently, this tissue has a high content of glycogenolytic enzymes (Crabtree and Newsholme, 1972) and an extremely active lactate dehydrogenase designed to channel pyruvate into lactate (Bostrom and Johansson, 1972). A comprehensive review of the biochemical differences between red and white muscle of fish may be found in Love (1970).Studies on mammalian systems are in total agreement with the above findings from fish

8.

METABOLISM IN FISH DURING EXERCISE

507

species and have been reviewed by Keul et al. (1972) and Pette and Staudte (1973). Finally, it should be noted that, in terms of energy generating characteristics, heart muscle is very similar to red muscle in both mammals (Keul et al., 1972) and fish (Gesser and Poupa, 1974; Bass et ul., 1973). The differences in biochemical properties between red and white muscle correlate well with differences in function between these two tissue types. E!ectrophysiological studies show that during slow swimming the propulsive force is derived entirely from the red musculature. At the highest swimming velocities, the white muscle becomes maximally active and together with the red muscle provides the power for locomotion (Bone, 1966; Hudson, 1973). It appears that the major function of white muscle is to act as a powerful but temporally limited energy source for vigorous movement such as required in capturing prey or avoiding predators. However, it is now evident that white muscle is utilized at swimming velocities involving less than burst activity. All of the current data indicate that at some level in the transition from low to high swimming speed there is an increasing recruitment of white fibers. This is based on (a) the accumulation of lactate in the white muscle of carp (Carassiuscarussius) (Johnston and Goldspink, 1973b), coalfish (Gadus virens) (Johnston and Goldspink, 1973c),mackerel (Trachurus symmetricus) (Pritchard et al., 1971), and rainbow trout (Black et al., 1962) exercised at intermediate velocities; (b) hypertrophy of white muscle fibers in coalfish forced to swim at moderate speeds for extended periods of time (Walker and Pull, 1973); (c) the repayment of an oxygen debt by salmon (Oncorhynchus nerka) (Brett, 1964) during recovery from swimming at all elevated speeds; and (d) an observation that in goldfish the red muscle mass alone, by virtue of its quantity, is inadequate to provide the required power to propel the fish above 73 cm/sec (body length, 15 cm), although the animal can maintain speeds greater than this for extended periods of time (Smit et al., 1971).These findings do not negate the importance of white muscle during burst activity, but they do add an extra dimension to its function.

IV. METABOLISM OF ADENYLATES AND RELATED COMPOUNDS

A. Studies at the Metabolite Level The singularly most important function of catabolic pathways during energy demand situations is the formation of ATP (adenosine

508

WILLIAM R. DEUEDZIC AND P. W. HOCHACHKA

triphosphate), for the hydrolysis of ATP to ADP (adenosine diphosphate) plus Pi (inorganic phosphate) serves as the immediate source of energy for muscular contraction (Infante and Davies, 1962). In this respect, creatine phosphate levels are important in functioning as a buffer against rapid ATP depletion. In addition, the free nucleotides are instrumental in the control of metabolic pathways through their effects on rate controlling enzymes (Atkinson, 1965). Hence, knowledge of alterations in the free adenylate pool and related compounds is a prerequisite for the understanding of the control of energy metabolism. As shall be described below, the rate of utilization of ATP for work processes is beautifully interrelated to its rate of production. The relevant data with respect to alterations in the concentration of adenylates and related compounds in white muscle due to exercise may be found in Table I. In rainbow trout exercised to exhaustion, creatine phosphate levels in "white" muscle decreased by 7.35 pmoles/g (Tomlinson et al., 1965). Not surprisingly, the free creatine level in white muscle of carp (C. carassius) increased from 29 Table I The Concentration of Free Adenylates and Related Compounds in Fish White hluscle Before and After Strenuous Activity Species Trout (Salmo gairdneri)b Carp (Carassius carassiusy Cod (Gadus morhua)"

Metabolite

Resteda

Exercised

Creatine phosphate Inorganic phosphate Creatine ATP ADP AMP

9.45 44.50 29.11 5.34 0.58 0.69 1.26 0.003 4.75 4.12 0.97 0.07 1.38 3.00

2.10 60.26 47.10 0.26 0.43 0.57 5.86

IMP GTP Cod (Gadus mmrhua)e Carp (Cyprinus carpio)'

NH,+ ATP ADP AMP IMP

NH,+

0

7.0 1.87 0.73 0.08 4.01 4.10

All values are expressed in pnoles/g fresh tissue. In the experiment with C . carassius the fish were exercised strenuously at 85 cm/sec (body length, 15 cm) for 15 min.

In all other experiments the fish were exercised to fatigue. Tomlinson et al. (1965). Johnston and Goldspink (1973b). Jones and Murray (1960). Fraser et al. (1966). Driedzic and Hochachka (1976).

8. METABOLISM

IN FISH DURING EXERCISE

509

pmoles/g in unexercised fish to 47 pmoleslg in animals vigorously exercised for 15 min at 85 cm/sec (body length, 15 cm). However, free creatine levels did not increase above the resting value in fish exercised for 200 min at intermediate speeds of 17 and 55 cm/sec. Furthermore, the concentration of free creatine in red muscle was about 15 pmoles/g and this level did not change after any of the above exercise conditions (Johnston and Goldspink, 1973b). Jones and Murray (1960) compared the free nucleotide levels in “white” muscle from rested and exhausted cod (Gadus morrhua). They observed that after activity ATP had decreased by 5.08 pmoledg, ADP and AMP (adenosine monophosphate) were about constant, and IMP (inosine monophosphate) increased b y 4.60 pmoles/g. I n an independent study, Fraser et al. (1966) found that NH4+ levels in cod white muscle increased by 2.5 pmoles/g, between resting fish and those forced to swim to exhaustion. Working with the elasmobranch, Triachis scyllium, Suyama et al. (1960) collected data of a similar nature at the qualitative level. That is, exhausted fish have a lower content of ATP and a higher content of IMP and NH4+.The picture with respect to adenylate metabolism in white muscle has recently been elucidated and integrated. Metabolite concentrations in carp (Cyprinus carpio) white muscle were determined at rest and after exercise resulting in fatigue. ATP concentrations were reduced by about 65% in the fatigued animals. Levels of ADP also decreased a small but significant amount; however, AMP concentrations remained low and unchanged. Thus, in this complex way, the total free adenylate pool decreased by 2.48 pmoledg. Concomitant with this decrease was an increase in IMP concentration of 2.63 pmoles/g and NH4+ concentration of 1.10 pmoles/g (Driedzic and Hochachka, 1976). The increase in the IMP level and the decrease in the adenylate pool were essentially in 1 : 1 stoichiometry in both the cod and the carp experiments. Finally, it is interesting to note that although NH4+ levels increase in working white muscle the change is never as large as that for IMP increase. GTP (guanosine triphosphate) levels are notably low in fish muscle. In cod “white” muscle, the concentration was 0.003 pmolelg in rested fish and was undetectable in exhausted animals (Jones and Murray, 1960). Moreover, Gras et al. (1967) were unable to detect GTP in rainbow trout muscle, although it did occur in small amounts in the liver. As shall be described below, measurements of free phosphate levels in muscle are extremely important. On this point, however, there is a paucity of information. I n rainbow trout exercised strenu-

510

WILLIAM R. DRIEDZIC AND P. W. HOCHACHKA 60 r

’ 0

I

2

5

10

15

‘

Time in Minutes

Fig. 1. Muscle and plasma phosphate concentrations in trout during 15 min of strenuous exercise. (Data are adapted from Hammond and Hickman, 1966.)

ously over a period of 15 min at 53 cm/sec (body weight, 46 g), oscillations occurred in the content of P, in “white” muscle (Fig. 1).After 2 min of exercise Pi had decreased from the resting level, but after 5 min Pi concentration was higher than the resting level. The concentration of Pi after 10 min of exercise was similar to the level noted after 2 min of activity, but after 15 min of exercise the level was once again higher than that found in the unexercised fish. There were alterations in plasma phosphate that showed an inverse correlation to the pattern noted in muscle (Hammond and Hickman, 1966).It may be that there is some type of cellular control mechanism which prevents large accumulations of Pi (see lactate levels, Section IV,A). Tomlinson et al. (1965)also noted a significant increase in trout “white” muscle Pi after exercise to exhaustion.

B. Studies at the Enzyme Level Creatine kinase (EC 2.7.3.2)which catalyzes the reaction Phosphocreatine

+ ADP + Hf + creatine + ATP

is the most abundant of the sarcoplasmic proteins found in fish muscle (Gosselin-Rey et al., 1968).The equation as it is written above is in the direction relevant to an energy demand situation. Creatine kinase exists as a dimer and may be made up of an M (muscle) subunit and a B

8.

METABOLISM IN FISH DURING EXERCISE

511

(brain) subunit (Watts, 1973). Only the MM isozyme is found in red and white carp (Cyprinus carpio) muscle; however, the MB type occurs in cardiac tissue (Hamoir et al., 1972). To our knowledge, the kinetic properties of creatine kinase from a fish source have never been studied. In mammalian muscle, the enzyme seems to be controlled largely by the concentration of the adenylates and p H (Watts, 1973). However, creatine kinase from turtle heart is activated by NADH (reduced nicotinamide adenine dinucleotide) which increases the apparent affinity (i.e., lowers the K,) of the enzyme for creatine phosphate (Storey, 1975). Storey attributes the difference in control mechanism between turtle heart and mammalian heart creatine kinases to the markedly different anaerobic capabilities of these two tissues. H e argues that, since during anaerobic metabolism in the turtle heart, ATP levels and p H are not altered, these components could not possibly function as metabolic “signals.” Clearly, a comparative study of fish muscle creatine kinases from the aerobic red and heart muscles and the anaerobic white muscle could be of extreme interest. Suffice for the present discussion to point out that fish muscle demonstrates both Mg2+-activated ATPase activity as well as Ca2+activated myofibrillar ATPase activity (see Tarr, 1972, for review of this area). It is of interest to note that white muscle is about five times more active than red muscle with respect to this latter property (Johnston et al., 1972). An equilibrium is established amongst the free adenylates due to the presence of adenylate kinase (EC 2.7.4.3)which catalyzes the conversion of 2 ADP to ATP plus AMP. The enzyme has recently been crystallized from carp muscle (Noda et al., 1975). Adenylate kinase is normally thought to be controlled by mass action effects alone; however, the mitochondria1 form of the rat liver enzyme is regulated by citrate which causes a decrease in the K , for ATP (Pradhan and Criss, 1974). The physiological significance of this and the effect on adenylate kinases from other sources is yet to be ascertained. The most interesting enzyme, in terms of regulatory properties, with respect to metabolism of the adenylates is 5‘-AMP deaminase (EC 3.5.4.6). This enzyme catalyzes the conversion of AMP to IMP and NH,+ and is found in exceptionally high titers in white muscle in relation to its activity in other tissues (Table 11).The 1:1 stoichiometric relationship between adenylate depletion and IMP accumulation observed during activity (Jones and Murray, 1960; Driedzic and Hochachka, 1976) clearly shows that the adenylate pool is reduced by this enzyme. In carp white muscle, 5’-AMP deaminase is potently activated by ADP (K, = 0.5 mM) and is extremely sensitive to GTP

512

WILLIAM R. DRIEDZIC AND P. W. HOCHACHKA

Table I1 Specific Activities of 5'-AMP Deaminase in Different Tissues of the Carp (Cyprinus carpio) Tissue

Activity"

White muscle Red muscle Heart Gill Kidney Spleen Brain Liver

12.8 0.8 0.1 1.5 0.9 1.2 0.05 0.4

" Activity is expressed in micromoles of product formed per minute per gram fresh tissue weight at 25°C (Fields, 1975).

( K , = 50 p M ) inhibition (Fields, 1975). In these respects 5'-AMP deaminase from carp white muscle is very similar to the enzyme found in rat skeletal muscle (Lowenstein, 1972).Control of the enzyme from white muscle ofRaia clavata appears to be somewhat different. In this case, the enzyme is activated by ATP (ADP effect unknown), inhibited by Pi, and unaffected by GTP (Markarewicz, 1969). The differences in control properties of the enzyme from the teleost and the elasmobranch warrant further study from a comparative point of view. For a recent summary (within its proper historical framework) of 5'-AMP deaminase from fish and other sources, the reader is referred to Watts and Watts ( 1974). There are two enzymes associated with GTP production: nucleotide diphosphokinase (EC 2.7.4.6) and succinic thiokinase (EC 6.2.1.4). Nucleotide diphosphokinases are a relatively unspecific group of enzymes that catalyze the transfer of the terminal phosphate group of 5'-triphosphate nucleotides to the 5'-diphosphate nucleotides by the following general mechanism: NlTP

+ NzDP

-

NIDP + NzTP

Enzymes of this nature are ubiquitous (Parks and Aganval, 1973) and although never demonstrated in fish muscle, their presence may be taken for granted, and it is thus probable that GTP pools in muscle are in a state of equilibrium with ATP pools. Succinic thiokinase catalyzes the reaction Succinate + CoASH + GDP + succinyl CoA + GTP

8.

513

METABOLISM IN FISH DURING EXERCISE

and is a link in the citric acid cycle. This enzyme has not been characterized from fish sources; however, as with nucleotide diphosphokinase, its presence can be safely assumed since it occurs throughout the phylogenic scale.

C. Control of the Adenylate Pool It is possible to construct a fairly comprehensive scheme of metabolic alterations in the adenylate pool, in teleost white muscle, which occur between the resting and maximally active state (Fig. 2). The energy required for work is ultimately derived from the hydrolysis of ATP. When the creatine phosphate levels cannot be maintained by the energy production pathways, the content of ADP increases, and as the ADP level increases, ATP and AMP are formed by the adenylate kinase reaction simply due to a mass action effect. As the work load on the tissue exceeds its aerobic capabilities, GTP levels drop, 5’-AMP deaminase is activated, and the adenylate pool is decreased. This is possible since AMP, the substrate of the AMP deaminase reaction, is made available by adenylate kinase. 5’-AMP deaminase, in teleost white muscle, appears to be controlled largely by the removal of GTP inhibition. GTP levels are initially low in fish muscle and as demands for high energy phosphates increase during activity, GTP levels must fall, for GTP is only formed in essentially two ways: first by transphosphorylation with ATP, and second by the citric acid cycle reaction catalyzed by succinate thiokinase. When ATP levels are reduced, the rate of GTP production by the former reaction must also be reduced. CreatineP

+ ADP

creatine

-+

ATP -+ ADP

2 ADP GTP

-+

ATP

+ ATP

+ Pi

+ AMP

+ ADP + GDP + ATP

(a)

(b) (C)

(4

Succinyl-CoA -+ succinate G DP-GTP AMP -+ IMP

+ NH,+

(0

Fig. 2. Depletion of the adenylate pool in white muscle during anaerobic work. Enzymes involved in the reactions are: (a) creatine kinase, (b) myofibrillar ATPase, (c) adenylate kinase, (d) nucleotide diphosphokinase, ( e )succinic thiokinase, (0 5’-AMP deaminase.

5 14

WILLIAM R. DEUEDZIC AND P. W. HOCHACHKA

Furthermore, as energy demands are placed on white muscle, glycolysis is activated far more than the citric acid cycle activity (Wittenberger and Diaciuc, 1965); consequently the proportion of triphosphorylated nucleotides that GTP represents must decrease. Thus, as AMP levels increase, GTP levels decrease and 5'-AMP deaminase is subsequently functional. The result is depletion of the adenylate pool with a concomitant increase in IMP and NH,+. Unfortunately, we do not know when reactions of this nature take place. Do they occur only during the final burst period, do they begin at the transition into prolonged swimming, or is the adenylate pool altered even during sustained swimming? Furthermore, what stops depletion of the adenylate pool? Is this due to inhibition of 5'-AMP deaminase by Pi? Is it related to the energy status of the adenylates, as it is in mammalian liver (Chapman and Atkinson, 1973)? Are there extrinsic factors which force the fish to stop swimming prior to depletion of the adenylate pool? Obviously, the ground has just been broken with respect to adenylate metabolism in white muscle. Regardless of what controls pool depletion the question still remains of the physiological significance of the reduced adenylate pool during high muscle work rates. One simple explanation may be that 5'-AMP deaminase functions in concert with the adenylate kinase reaction to maximize ATP production by a mass action effect. However, there may be a more important thermodynamic explanation for the observation. Thus, as ATP levels drop during muscle work the ratio change of [ADP][P,]/[ATP] could drastically reduce the free energy of ATP hydrolysis according to the following relationship:

During muscle work, control of this ratio may become increasingly difficult since not only is there a change in the ADP/ATP ratio, there also occurs an increase in Pi concentrations (Table I; Fig. 1). These considerations emphasize that in the absence of external controlling mechanisms large drops in ATP concentrations could not be tolerated because they would occur concomitantly with increasing ADP levels of comparable magnitude, a situation that is clearly prevented (Table I). An appropriate relationship between ATP and ADP could be adjusted by the concerted action of adenylate kinase and 5'-AMP deaminase, the AMP formed from the adenylate kinase reaction being removed by 5'-AMP deaminase in order to minimize ADP accumulation. A final aspect to the picture of nucleotide metabolism in white

8.

METABOLISM IN FISH DURING EXERCISE

515

muscle during activity is the role of free NH4+.The increase in NH4+ level during white muscle activity is lower than the increase in the concentration of IMP. It is possible that some NH4+is spilling into the blood and could, consequently, account for the postulated anaerobic production of NH,+ during exercise by Tilapia mossambica (Kutty, 1972). Furthermore, the anaerobic production of NH4+during hypoxic excursions b y goldfish and carp, that is known with some certainty (Dejours et al., 1968; Pequin and Serfaty, 1962),may also be related to adenylate pool reduction. It is possible, though, that the unaccounted for NH4+has a metabolic fate within the muscle. It has recently been shown that the amount of nitrogen in the free amino acid pool of carp muscle increases during hypoxia. It may be that during strenuous activity NH,+ liberated from the adenylate pool is incorporated into the free amino acid pool (Driedzic and Hochachka, 1975), perhaps for nitrogen conservation. It would then be possible for the nitrogen to be called upon for regeneration of the adenylate pool during recovery from activity. The situation with respect to adenylate metabolism in red and heart muscle remains a complete unknown. Under working conditions, when oxygen is not limiting, the adenylate pool does not decrease in either mammalian skeletal muscle (Edington et al., 1973) or mammalian heart (Neely et al., 1972a), although ATP levels may decrease with concomitant increases in ADP and AMP levels. But when these tissues are subjected to hypoxic stress there is a depletion of the adenylate pool and an increase in degradation products of the adenylates (Imai et al., 1964; Deuticke and Gerlach, 1966). It thus appears that in mammalian “red” type muscle the adenylates are depleted when the energy demands on the tissue exceed the tissue’s aerobic capacity. It seems unlikely that the free adenylate pool in fish red muscle decreases during activity due to the maintenance of free creatine levels (Johnston and Goldspink, 1973b). It would be valuable, though, to ascertain if there are any alterations in the relative concentration of the adenylates during muscular work.

D. Recovery of the Adenylate Pool in White Muscle following Activity One outstanding question that we are now in a position to speculate upon is: How is the adenylate pool regenerated during recovery from strenuous work? Lowenstein (1972)has proposed that the 5‘-AMP deaminase reaction is one step in a reaction span that is termed the purine nucleotide cycle (Fig. 3).According to Lowenstein,

516

WILLIAM R. DRIEDZIC AND P. W. HOCHACHKA

AMP fumarate

adenylosucc inate

IMP GTP

+

aspartate

Fig. 3. The purine nucleotide cycle. Enzymes involved in reactions of the cycle are: (a) 5‘-AMP deaminase, (b) adenylsuccinate synthetase, and (c) adenylosuccinase. (Reproduced from Ldwenstein, 1972, Physiol. Reu. 52, 382-414.)

IMP further reacts with GTP and aspartate to form adenylosuccinate.

The adenylosuccinate in turn is converted to AMP plus fumarate. It has been shown in homogenates of mammalian skeletal muscle that the cycle functions in concert with glycolysis (Tornheim and Lowenstein, 1974). The theory, however, predicts only a transient increase in IMP with general maintenance of the adenylate pool. Clearly, in teleost white muscle, the cycle per se does not operate during activity, or stops operating at some point, since there is an accumulation of IMP. If the cycle were to become active only during recovery from exercise then the discrepancy between the in vivo results of a depletion in the adenylate pool with the in vitro results of maintenance of the adenylate pool would be apparent rather than real. Since the enzymes of the purine nucleotide cycle evidently are present in fish white muscle (Fields, 1975), it is probable that they supply a pathway for the regeneration of the adenylate pool from IMP during recovery following anaerobic work. Thus, in white muscle the reaction pathway shown in Fig. 3 is a “cycle” only in a theoretical sense, because the two arms of the cycle are functionally separated in time. One arm, catalyzed by AMP deaminase, is formally a catabolic pathway leading to AMP hydrolysis; the other arm is formally an anabolic pathway leading to AMP formation during recovery. The control properties of AMP deaminase as well as adenylosuccinate synthetase, which catalyzes the formation of adenylosuccinate, are entirely consistent with this model. Thus, during white muscle work, 5‘-AMP deaminase would be deinhibited due to dropping concentrations of GTP, and at the same time, GDP concentrations are presumably increased. GDP is a potent inhibitor of adenylosuccinate synthetase (Muirhead and Bishop, 1974) and this effect, coupled with reduced availability of one of its substrates (GTP), readily explains how this arm of the purine nucleotide cycle in white muscle is held at a re-

8.

METABOLISM IN FISH DURING EXERCISE

517

duced rate at the same time as AMP deaminase is being strongly deinhibited.

V. CARBOHYDRATE METABOLISM A. Studies at the Metabolite Level In fish tissues, the primary energy yielding pathway for carbohydrates is via the classical Emden-Myerhof-Parnas scheme of glycolysis. The majority of the glycolytic intermediates have been detected in fish muscle as well as the prerequisite enzymes (Tarr, 1972; Hochachka, 1969). Under aerobic conditions glucose or glycogenderived glucose l-phosphate is catabolized to pyruvate and subsequently oxidized to CO, and H 2 0 via the citric acid cycle. Under anaerobic conditions pyruvate is converted to lactate by the reaction catalyzed by lactate dehydrogenase which serves to maintain balance of the pyridine nucleotides (Fig. 4). Glycogen is the major, if not the only fuel of anaerobic metabolism. For each mole of glycogen-derived glucose l-phosphate converted to lactate three moles of ATP are formed. As described above, fish white muscle is particularly well designed for anaerobic metabolism during “burst” activity. Clearly, glycogenolysis in this tissue is of profound significance .to the overall energetics of fish. However, other tissues have the capacity to utilize carbohydrates and this fact should not be underestimated in importance. For instance, chloride cells from the gill of Fundulus heteroclitus are densely packed with mitochondria and glycogen granules (Philpott and Copeland, 1963). It is possible that this glycogen reserve plays an important role in the costly energetics of ion osmoregulation. There are numerous studies on the quantitative relationship between glycogen depletion and lactate accumulation, in fish muscle, as a function of exercise. This area of literature has been more than adequately reviewed by Love ( 1970), Wendt and Saunders (1973), and Bilinski (1974). Suffice for the present to reiterate the common finding. That is, the greater the energetic demands placed on muscle-either b y swimming longer at a given speed (above a certain level) or by swimming faster for the same length of time-the greater the glycogen depletion and lactate accumulation. The mobilization of glycogen may be extremely rapid, such that trout muscle may utilize the equivalent of about 40 pmoles of glycogen-derived glucoselglsec and in 15 sec deplete about one-half of the glycogen stores (Stevens and Black,

518

WILLIAM R. DFUEDZIC AND P. W. HOCHACHKA Hormones

Inactive protein

, ~ ~ - ‘ \ , k ~ n se y

(epinephrine, I glucogon) s

ATP-VCI~C

Adenyl cyclase

I

AMP,,&‘

+ PT

4

,-.Active protein I kinare

I

I

Fig. 4. Glycolytic activation in skeletal muscle of vertebrate animals. Regulatory metabolites are connected with the enzyme steps they activate by dashed lines; activation is indicated with a dark arrow. Additional abbreviations used: GlP, glucose 1-pho’sphate; G6P, glucose &phosphate; F6P, fructose &phosphate; FDP, fructose l,&diphosphate; TP, triose phosphate; l,SDPG, 1,Sdiphosphoglycerate; SPGA, 3-phosphoglycerate; 2-PGA, 2-phosphoglycerate; PEP, phosphoenolpyruvate; PP,, pyrophosphate. [From Hochachka and Storey, 1975, Science 187, 613-621. Copyright (1975) by the American Association for the Advancement of Science.]

1966). It should be noted that on a weight basis, the rate of utilization of glycogen, under severe energy demands by red muscle, may be %fold higher than white muscle (Johnston and Goldspink, 1973a). This is no doubt due to the greater aerobic capacity of the former tissue. Although the highest rate of glycogen utilization by white muscle occurs during “burst” swimming, there is now much evidence which shows the accumulation of lactate in this tissue at less than “burst speeds” (see Section 111). It is possible that the point of inflection

8.

METABOLISM IN FISH DURING EXERCISE

519

between sustained and prolonged swimming (see Chapter 2) is marked by glycogen utilization in white muscle. Clearly, if a fish is using white muscle glycogen at less than burst speeds, the level of lactate in that tissue cannot be allowed to build up indefinitely. Not surprisingly, there appears to be an upper limit to which lactate is allowed to accumulate in viuo. Hence, in trout exercised strenuously for 2 min, the level of muscle lactate increased from 7 to 25 pmoles/g; yet after 3, 7, and 13 min of further exercise, the concentration only reached 28, 32, and 32 pmoles/g, respectively (Black et al., 1962). Similar data were obtained with trout by Hammond and Hickman (1966)and Stevens and Black (1966).The level of lactate in carp (Cyprinus carpio) white muscle after maximal activity or severe hypoxia was 12 pmoles/g (Driedzic and Hochachka, 1975, 1976), but when carp white muscle was electrically stimulated until the muscle per se was exhausted, the concentration of lactate reached 33 pmoles/g (Wittenberger and Diaciuc, 1965). The fate of lactate produced in the white muscle will be treated separately in Section IV,D. There are very few measurements of alterations in glycolytic intermediates during exercise. It is important to know what changes occur between glycogen and lactate for the intermediates of the glycolytic pathway are, as we shall describe below, essential in regulating their own rate of production. In carp white muscle, sampled at rest and after maximal activity, glucose &phosphate increased from 0.67 to 1.33, fructose &phosphate increased from 0.11 to 0.18, and fructose l,&diphosphate increased from 0.85 to 1.28 pmoleslg (Driedzic and Hochachka, 1976). No change was recorded in the concentration of Sphosphoglycerate or dihydroxyacetone in cod muscle sampled at rest and after exercise to exhaustion (Burt and Stroud, 1966). Obviously, measurements of glycolytic intermediates are greatly in need. It would be extremely valuable to have time course measurements of these metabolites in both red and white muscle.

B. Studies at the Enzyme Level The utilization of glycogen is regulated at three control points. The enzymes involved at these loci are glycogen phosphorylase, phosphofructokinase, and pyruvate kinase. These enzymes share an important characteristic: Under physiological conditions they catalyze reactions that are, in effect, irreversible, for either thermodynamic or kinetic reasons, or both. Regulation at these points is extremely complex and involves hormonal signals, ionic effectors, and feedback control by free adenylates and glycolytic intermediates. The reader is referred to

520

WILLIAM R. DFUEDZIC AND P. W. HOCHACHKA

Drummond ( 1967) and Hochachka and Storey (1975) for overviews of the regulation of carbohydrate metabolism. In some tissues where blood-borne glucose is an alternative energy source, hexokinase appears as a regulatory enzyme. This enzyme, however, is in such low activities in fish muscle relative to muscle from other sources and relative to other glycolytic enzymes in fish muscle that it is unlikely that free glucose serves as an important source of energy for fish muscle (Crabtree and Newsholme, 1972). The mechanisms by which glycogen is broken down in mammalian muscle have been vigorously investigated and a brief outline of the phenomenon at this point is warranted. The enzymology of glycogenolysis and the factors which regulate the process are summarized in Fig. 4. Phosphorylase, the enzyme which degrades glycogen to glucose l-phosphate, exists in two forms, phosphorylase a and phosphorylase b (EC 2.4.1.1). Phosphorylase b differs from phosphorylase a in that it has half the molecular weight and is catalytically relatively inactive, except in the presence of relatively high concentrations of 5’-AMP. Phosphorylase a is active in the absence of 5’-AMP and can be considered to be the physiologically active form. The two proteins are interconvertible; the conversion of phosphorylase b to phosphorylase a is catalyzed by the enzyme phosphorylase b kinase (EC 2.7.1.38), the reverse reaction by a phosphatase. When muscle contracts there is an increase in the amount of phosphorylase a present and a concomitant decrease in the amount of phosphorylase b. This is due to the activation of phosphorylase kinase. Phosphorylase kinase, however, itself exists in an activated and a nonactivated form. There appear to be two mechanisms for the regulation of phosphorylase kinase activity. According to one scheme, hormonal controls such as epinephrine induce the formation of cyclic 3’,5’-AMP via action on the enzyme adenylate cyclase. Cyclic 3’,5’-AMP in turn activates a protein kinase, which sequentially catalyzes the conversion of nonactive phosphorylase kinase to activated phosphorylase kinase. Alternatively, phosphorylase kinase may be activated by Ca2+ alone. Calcium released from storage sites or transported across the plasma membrane of the cell serves to couple the electrical and mechanical events of muscular contraction (Drummond, 1967, 1971). Studies to date indicate that the regulation of glycogenolysis in fish muscle is similar to that in higher organisms. Nakatano and Tomlinson (1967)have shown that in rainbow trout, after 2 min of vigorous exercise by chasing, muscle phosphorylase a levels and tissue cyclic AMP levels increased significantly. However, epinephrine and norepinephrine levels did not change. Yamamoto (1968) provided evidence for

8.

METABOLISM IN FISH DURING EXERCISE

52 1

the presence of phosphorylase b kinase and, furthermore, found that rabbit muscle phosphorylase kinase can convert trout muscle phosphorylase b to phosphorylase a. The same phenomenon was observed with dogfish muscle phosphorylases (Cohen et al., 1971). Pocinwong et al. (1974) are of the opinion that dogfish “white” muscle phosphorylase kinase is regulated only by Ca2+ ions. Clearly, there are differences to be resolved on this question that may be related to the red-white muscle dichotomy. It is possible that hormonal controls play a role in glycogenolysis in red muscle, where the tissue is highly perfused, but that Ca2+ activation is more important in white muscle due to the lack of bloodborne signals. Following the phosphorylase system, the next major site in the control of glycolysis is the reaction catalyzed by phosphofructokinase (EC 2.7.1.11). The reaction involves the conversion of fructose &phosphate plus ATP to fructose 1,Sdiphosphate plus ADP (Fig. 4). The phosphofructokinase reaction is the first unique step in glycolysis, hence, the enzyme is precisely regulated by various metabolites in a manner that controls the rate of glycolysis in accord with the cells’ need for energy or glycolytic intermediates. In a variety of tissues, phosphofructokinase is controlled b y a number of factors: activators of the enzyme include fructose &phosphate, fructose l,&diphosphate, AMP, Pi, and NH4+,whereas ATP, citrate, and creatine phosphate are inhibitors (Mansour, 1972). For this enzyme the two cosubstrates, fructose 6-phosphate and ATP, obviously serve both as substrates per se and as metabolite modulators. One substrate, fructose &phosphate, behaves as a typical positive modulator, while the cosubstrate, ATP, behaves as an important negative modulator. All other regulatory metabolites exert their effects either by modifying enzyme-fructose 6-phosphate affinity, or by modifying enzyme sensitivity to ATP, or more usually by modifying both fructose &phosphate and ATP saturation kinetics. Phosphofructokinase from goldfish white muscle appears to be similar in general nature to the enzyme from other sources. Fructose &phosphate and fructose 1,6-diphosphate are known activators of the enzyme; ATP and citrate are inhibitors (Freed, 1971). Studies (Tsai et al., 1975) indicate that the enzyme occurs in tissuespecific isozymic form. Comparison of the regulatory properties of the enzyme from muscle, liver, erythrocytes, and brain suggests that the muscle enzyme is predominantly regulated by creatine phosphate, citrate, and inorganic phosphate, the first two being inhibitors, the latter, an activator. To complete the activation of muscle glycolysis, the activity of phosphofructokinase is integrated with the next major control site in

522

WILLIAM R. DRIEDZIC AND P. W. HOCHACHKA

the pathway: pyruvate kinase (EC 2.7.1.40). Pyruvate kinase catalyzes the conversion of phosphoenolpyruvate plus ADP to pyruvate plus ATP. Integration of the activities of phosphofructokinase and pyruvate kinase involves adenylate coupling; that is, ADP, the product of the phosphofructokinase reaction, is a substrate for pyruvate kinase, and this in itself serves to automatically coordinate the activities of these two enzymes. Moreover, further control is exerted over the enzyme by ATP and creatine phosphate inhibition (Kayne, 1973; Kemp, 1973). In fish, however, pyruvate kinase is regulated by one further factor, that is, feedforward activation by fructose l,&diphosphate, a product of the phosphofructokinase reaction. In these animals, the mechanism of fructose l,&diphosphate activation appears to be by lowering the K , of the enzyme for phosphoenolpyruvate. This phenomenon has been observed in carp (C. carassius) red muscle (Johnston, 1975), Coryphaenoides white .muscle (Mustafa et d.,1971), trout “white” muscle (Somero and Hochachka, 1968), and sturgeon (Acipenserfuluescens) muscle (Randall and Anderson, 1975). The metabolic fate of pyruvate under aerobic conditions is oxidative decarboxylation to acetyl-CoA, the fate of which shall be described in the section dealing with the citric acid cycle. During anaerobic conditions pyruvate is converted to lactate by the following reaction: Pyruvate

+ NADH + H+ + lactate + NAD+

This reaction catalyzed by lactate dehydrogenase (EC 1.1.1.27) is essential to glycolysis for it functionally maintains balance of the pyridine nucleotides by oxidizing NADH which is formed at the glyceraldehyde-%phosphatedehydrogenase step. When all tissues are considered, the number of lactate dehydrogenase isozymes reported may range from a few to over twenty in any one species (Tsuyuki, 1974). The details of the isozymic properties of this enzyme are beyond the scope of the present essay; sufficeto point out that, in terms of electrophoretic mobility, lactate dehydrogenase from heart and white muscle are markedly different and that red muscle enzyme demonstrates intermediate characteristic with respect to the other two muscle types (Love, 1970; Hamoir et al., 1972; Lim et al., 1975). On the basis of the isozyme ratio, that is, the ratio between the activity of lactate dehydrogenase in the presence of low pyruvate and the activity in the presence of high pyruvate concentrations, it appears that the enzyme from white muscle is designed to function in the direction of lactate production and that the enzyme from heart and red muscle is better adapted to function in the direction of pyruvate formation (Ges-

8.

METABOLISM IN FISH DURING EXERCISE

523

ser and Poupa, 1974; Bostrom and Johansson, 1972). One final comment on this enzyme is warranted. Glycolysis is considered to take place in the cytosol, yet lactate dehydrogenase activity has been detected in mitochondria from eel (Anguilla anguilla) muscle (Mattisson et al., 1972), the function of which is totally unknown. C. Control of Carbohydrate Metabolism

The situation with respect to glycogenolysis in fish white muscle seems fairly clear. Free calcium ions released from the sarcoplasmic reticulum during muscle contraction could stimulate glycogen phosphorylase by action upon protein kinase. The subsequent increase in fructose &phosphate and fructose l,&diphosphate would lead to substrate and product activation of phosphofructokinase, respectively. Following this, pyruvate kinase activation would occur due to the availability of ADP and the positive feedforward effect of fructose 1,Gdiphosphate. Moreover, deinhibition of phosphofructokinase and pyruvate kinase would be expected as a consequence of falling levels of ATP and creatine phosphate. The pyruvate produced at the pyruvate kinase reaction is, of course, subsequently converted to either lactate or acetyl-CoA. Aside from the possibility that glycogen phosphorylase may be activated by Ca2+alone, the adenylates are involved in one other interesting facet to the control of glycolysis in fish white muscle. There is no evidence whatever that AMP constitutes an uniquely important metabolic signal to glycolysis in this tissue, as is generally accepted to occur in mammalian heart (Newsholme, 1972) and mammalian skeletal muscle (Hochachka and Storey, 1975), because its concentration is similar at widely differing metabolic rates. In contrast, if there is a single adenylate signal that is important to a sustained high level of glycolysis it presumably is ATP, since its overall concentration change is the greatest. As argued above (Section IV,C), however, in order to take advantage of this metabolic signal the organism must tolerate an overall reduction in the adenylate pool. As opposed to white muscle, glycolytic control mechanisms in fish red muscle are still totally unknown. Is phosphorylase hormonally controlled in this tissue? This is possible since the blood level of epinephrine and norepinephrine, in trout, increases during activity (Nakatano and Tomlinson, 1967). How is phosphofructokinase controlled? Clearly it is not under creatine phosphate regulation since creatine levels do not increase even during strenuous exercise (see Section IV,A). But what about AMP? Is there a reorganization of the adenylate pool in this tissue which leads to an AMP increase? Cer-

524

WILLIAM R. DRIEDZIC AND P. W. HOCHACHKA

tainly there are questions of fundamental importance that we must address ourselves to if we hope to gain an understanding of energy metabolism in fish.

D. Fate of Lactate Produced in White Muscle The fate of lactate produced in the white muscle of fish remains an open question. In mammalian systems which have been better studied it is known that 80-90% of the blood lactate is oxidized to COz and water (Drury and Wick, 1956), the site of oxidation being heart (Keul et al., 1972), skeletal muscle (Jorfeldt, 1970), liver (Rowell et al., 1966), and kidney (Levy, 1962). In resting conditions approximately 15% of the lactate is converted to glucose (Reichard et al., 1963)in the liver (Rowell et al., 1966); however, during activity this process (Le., the Cori cycle) may be quantitatively more important (Keul et aZ., 1972). Black, on the basis of his many studies, was forced to conclude that in fish the Cori cycle is of little importance (Black et al., 1966),for, after anaerobic stress glycogen does not return to prestress levels even after 24 hr (Black et aZ., 1962; Heath and Pritchard, 1965);therefore it is probable that in fish, as with mammals, most of the blood lactate is oxidized to COz. Bilinski and Jonas (1972) have shown that the capacity to channel lactate through the citric acid cycle per unit weight of tissue decreases in the following order: gill, kidney, red muscle, liver, heart, white muscle. An attempt was made to quantitate the in vivo significance of gill as a site of lactate utilization by sampling blood from rainbow trout before and after its passage through the branchial circulation. In most individuals sampled there was a negative arterial-venous difference during the recovery period following exercise to fatigue (Driedzic and Kiceniuk, 1976).In light of the high oxidative potential of gill and its apparent high glycogen content, the role of this tissue in the regulation of blood lactate must be further examined by more sensitive techniques. A further important site of lactate deposition may be red muscle. When the myotomal muscle of Cyprinus carpio was electrically stimulated to a moderate degree, white muscle glycogen content decreased while red muscle glycogen concentration remained the same, yet changes in lactate levels were exactly opposite this. Thus, at moderate levels of activity white muscle lactate level remained constant and red muscle lactate concentration increased (Wittenberger and Diaciuc, 1965).The authors suggest that lactate being produced in the white muscle is further oxidized in red muscle. Such a model would be consistent with observations by Issekutz et al. (1976)that working mammalian muscle of mixed fiber

8.

METABOLISM IN FISH DURING EXERCISE

525

types both produces and utilizes lactate at the same time. It has further been shown that red muscle has a high neoglycogenic capacity from added glucose (Wittenberger, 1972). Whether fish muscle can synthesize glycogen from lactate as can occur in the frog sartorius (Bendall and Taylor, 1970) is yet to be tested. It is well recognized that the mammalian heart utilizes blood lactate as a source of energy, although under aerobic working conditions free fatty acids are the preferred fuel of metabolism (Keul et al., 1972). When blood arterial lactate levels increase the rate of utilization of fats by the myocardium decreases (Spitzer, 1974). The aerobic energy source of the fish heart has never been investigated, thus it would be of interest to ascertain if metabolism is organized in a fashion similar to the mammalian tissue. If so, the problem of a switch in fuel sources may be particularly accentuated, since during the recovery period from strenuous activity blood lactate levels may increase by fivefold within 1 min (Driedzic and Kiceniuk, 1976). Clearly, metabolic regulation under these conditions is an area that should be investigated.

VI. LIPID METABOLISM A. General Observations It is well recognized that free fatty acids (FFA) derived from triglycerides serve as a major aerobic fuel source for energy metabolism in mammalian muscle (Drummond, 1967; Hochachka et al., 1977). Data on many fronts indicate that this is also true for fish muscle. During sustained swimming, for many hours, the total body lipids of coho salmon were noted to decrease in direct proportion tq the distance travelled (Kruegeret al., 1968). In light of the role of red muscle as the tissue responsible for generating the propulsive force during activity of this nature, it is probable that it is deriving much of its energy from fat catabolism. In keeping with the greater aerobic capacity of red and heart muscle as opposed to white muscle, it has been noted that the former tissues, from both salmonids (Bilinski, 1963; Jonas and Bilinski, 1964) and cyprinidonts (George and Bokdawala, 1964), can oxidize FFA at a rate at least 10 times higher than the latter tissue. A major difference exists between fish and mammals with respect to the site of storage of triglycerides. In mammals, fats are mainly deposited in the adipose tissue; however, in fish fat storage is more dispersed. Some fish do possess an adipose-like tissue in the viscera (Farkas, 1967a) but lipids are also extensively stored in the liver and

526

WILLIAM R. DRIEDZIC AND P. W. HOCHACHKA

muscle (Bilinski, 1969). Regardless of where the major site of lipid deposition occurs, the fat content of red muscle is usually about twice as high as in white muscle (Bone, 1966; Krishnamoorthy and Narasimhan, 1972; Lin et al., 1974). Moreover, fat is stored in the red muscle both intra- and extracellularly (George, 1962) and the intracellular lipids may be totally surrounded by mitochondria (Fig. 5) (Lin et al., 1974). Despite the presence of high “background noise” due to structural lipids, Bone (1966) observed decreases in the fat content of both red and white muscle of elasmobranchs after extended periods of electrically-induced sustained muscular activity. It should be kept in mind that white muscle has the capacity to utilize lipids at a low rate (Bilinski, 1969), although during intense activity it is most certainly a carbohydrate burner.

B. Control of Lipid Metabolism In light of the importance of lipid catabolism to the energetics of swimming, the paucity of information on the control of the phenomenon is astonishing! Lipid metabolism in mammalian tissues is at best poorly understood (Neely and Morgan, 1974; Hochachka et al., 1977), and our knowledge of control in fish muscle is even more obscure. We shall attempt to summarize briefly the state of the field in mammalian muscle with reference to aspects that have been studied in fish systems. The first step in the utilization of fats is the breakdown of triglycerides to FFA. The phenomenon generally referred to as lipolysis involves the sequential hydrolysis of triglycerides to diglycerides, monoglycerides, and eventually FFA and glycerol. Lipolysis in mammalian adipose tissue has been an active area of research. The process is stimulated b y a variety of hormones, including epinephrine, norepinephrine, ACTH, and glucagon, and involves activation via cyclic AMP. In this sense, the regulation appears analogous to the activation of phosphorylase (Drummond, 1971). Similar control mechanisms, however, do not take place in fish adipose tissue. Farkas (1967a,b, 1969) has shown that lipolysis in adipose tissue, from a variety of fish including carp (Cyprinuscarpio), bream (Abramis brama), and perch (Perca perca), is not stimulated in vivo or in vitro by any of the above active mammalian agents. Glucose is known to inhibit FFA release but mechanisms of activation are totally unknown. FFA derived from adipose tissue are transported to the site of utilization as albumin-bound complexes but triglycerides mobilized from other sources such as liver may be carried as lipoprotein complexes.

8.

METABOLISM IN FISH DURING EXERCISE

527

Fig. 5. Electron micrograph of pectoral red muscle of Trematomus borchgreoinki showing lipid droplets surrounded by mitochondria. Panel 1 shows a region of striated fibers, whilepanel 2 shows aperipheralregion devoid ofcontractilematerial butfilledwith mitochondria and lipid droplets. (From Lin et al., 1974,J. E x p . Zool. 189, 379-385.)

528

WILLIAM R. DRIEDZIC AND P. W. HOCHACHKA

Hydrolysis of protein-bound triglycerides occurs at the capillary endothelium and is catalyzed by a lipoprotein lipase, which in the mammalian heart appears to be activated by epinephrine (Mallov and Alousi, 1969). This, however, is the extent of our knowledge of the control properties of this enzyme. Clearly, on the basis of the sites of lipid storage and the form in which FFA reach the tissue of utilization, there is a need for a wide spectrum of muscular lipase activity. This criterion is certainly met in fish muscle. Foremost, lipase activity has been demonstrated histologically, both intracellularly and extracellularly in Labeo rohita red muscle (George and Bokdawala, 1964). A short-chain triglyceride lipase (EC 3.1.1.3)was partially purified from lingcod (Ophiodon elongatus) white muscle (Wood, 1959) and lipolytic activity specific for long-chain triglycerides Bas been demonstrated in trout red muscle (Bilinski and Lau, 1969). It is interesting to note that neither epinephrine nor cyclic 3’,5’-AMPwas stimulatory in the latter preparation. Finally, Bilinski et al. (1971) detected a lipase from trout red muscle, maximally active toward triplamitin at pH 4-4.5, displaying the sedimentation properties of a lysomal enzyme. The physiological function of this protein is totally unknown. Obviously, lipolysis in muscle is an area bountiful in problems for the metabolic biochemist. Once inside the cell FFA are esterified by the following reaction ATP

+ FFA + CoASH 5 acyl-CoA + AMP + PP1

which occurs in the cytosol. The reaction is catalyzed by a broad spectrum of acyl-CoA synthetases (EC 6.2.1.2), specific for short-, medium-, and long-chain fatty acids (Fig. 6). The rate of oxidation of FFA by trout muscle mitochondria is proportional to the amount of CoASH (reduced coenzyme A), at least up to 0.1 mM, and has a requirement for Mg2+and ATP (Bilinski and Jonas, 1964); thus, it appears that the same enzyme is present in fish systems as in the more commonly studied mammalian tissues. Palmityl-CoA synthetase, from rat heart, is inhibited by the products of the reaction (palmityl-CoA, AMP, and PPl) but under physiological working conditions is probably limited by cytosolic CoASH levels (Oram et al., 1975). Further oxidation of acyl-CoA takes place in the mitochondria but acyl-CoA derivatives cannot penetrate the inner mitochondria1 membrane. The transfer of acyl-CoA is facilitated by a carnitine-dependent translocation process (Fig. 6). As would be expected, FFA oxidation by trout muscle mitochondria is highly activated by the addition of carnitine (Bilinski and Jonas, 1970). Similar to the acyl-CoA synthetases, there are carnitine-acyl transferases (EC 2.3.1.7) specific for varying

8.

529

METABOLISM IN FISH DURING EXERCISE

-

FFA-albumsn

trlplyceride protein Cell

CoASH

rnernbcane

-+z. PP,

acyl-CoA

I

,.----camtine wter mltOchOnOrial

acyi-carnitine

membrane

L

carnitine

acyl-CoA

Fig. 6. Lipolysis and activation of free fatty acids.

chain lengths. Palmityl camitine transferase has been detected in cod white muscle and heart (Norum and Bremer, 1966). The transfer of acyl derivatives into the mitochondria may be a limiting step to FFA oxidation at high muscular work rates. When the work load on the rat heart, perfused with palmitate, was increased, the tissue levels of acyl carnitine increased in association with decreased levels of acyl-CoA and acetyl-CoA. This suggests that the oxidation of acyl carnitine was limiting the rate of utilization of palmitate. Moreover, under the same experimental conditions employing octanoate as the fuel source (which freely crosses the mitochondria1 membrane), high levels of acetyl-CoA were maintained. It thus appears that acyl translocation and not p-oxidation is the limiting step in palmitate degradation (Oram et al., 1973). Acyl-CoA derivatives are broken down within the mitochondria by a reaction scheme known as p-oxidation (Fig. 7). Each turn through the Acyl-CoA

+ FAD+ acyl-CoA dehydrogeoa\e a-p-unsaturated

a-p-Unsaturated acyl-CoA

p-Hydroxyacyl-CoA p-Ketoacyl-CoA

+ NAD+

+ H,O

-

+ FADH,

enol-CoA hydr,i\e

p-hydroxyacyl-CoA

p-hydroxyacyl-CoA dehydrogenaw

-

acyl-CoA

p-ketoacyl-CoA

+ NADH + H'

+ CoASH thio1.i'e acyl-CoA ( - 2 carbons) + acetyl-CoA

Fig. 7. Stepwise degradation of fatty acyl-CoA by the p-oxidation system.

530

WILLIAM R. DRIEDZIC AND P. W. HOCHACHKA

degradation sequence results in the production of a fatty acyl-CoA derivative two carbon atoms shorter and one molecule of acetyl CoA, which, in turn, is channeled into the citric acid cycle. p-Oxidation appears to be limited by the dehydrogenase reactions, largely b y the availability of oxidizing power. Hence, the process is tightly coupled to both the citric acid cycle and oxidative phosphorylation, which produce and utilize reduced equivalents, respectively. Unfortunately, @oxidation has never been studied in fish tissues; thus, we can only speculate that control of the phenomenon in fish is similar to that in other organisms. The integration of lipid utilization with Krebs cycle activity has been outlined elsewhere (Hochachka et al., 1977).

VII. PROTEIN METABOLISM It is well recognized that during severe food deprivation all animals utilize their muscle proteins as an energy source. However, it is now evident that amino acids, if not proteins per se, play an integral role in mammalian muscle metabolism at all times. In this tissue the degradation of amino acids may account for 20% of the CO, produced in the postabsorptive state (Beatty et al., 1974; Odessey et al., 1974). Fish are particularly interesting with respect to protein catabolism, since during induced starvation in the laboratory, the protein stores and the free amino acid pool are often called upon before glycogen and/or lipids are utilized (see Creach and Serfaty, 1974, for references). Under these conditions muscle proteins appear to be preferentially mobilized, with white muscle proteins being affected before those in red muscle (Johnston and Goldspink, 1973d). Prolonged starvation often occurs as a normal course of events with fish and as such these animals have probably evolved efficient mechanisms for dealing with the problem. But the extent to which proteins are utilized under conditions other than during a fast is still an open question. Kutty (1972) has noted that in exercising Tilapia mossambica starved for only 36 hr before the experiment, the level of NH4+excretion relative to oxygen consumption is high enough to indicate his fish were generating all of their aerobic energies from protein catabolism. This is a truly remarkable observation and warrants further investigation. It should be kept in mind that aquatic species can directly expire NH4+ without prior conversion to other energetically costly compounds such as urea or uric acid. Furthermore, the cost of making a polypeptide, given the prerequisite amino acids from dietary sources, is less than the conversion of amino acids to either glycogen or triglycerides.

8.

METABOLISM IN FISH DURING EXERCISE

531

Hence, we feel that at the present time there is no reason to rule out the possibility that fish have evolved unique storage protein(s) to be utilized in times of energy demand. Regardless of why or when proteins are utilized the process must be under stringent metabolic control. Proteins, as with all other cellular constituents, are in a state of continuous turnover. Goldberg and Dice (1974) have pointed out that this phenomenon may significantly enhance the organism’s ability to readily adapt to changes in its environment. Protein turnover is, of course, a function of the rate of synthesis and the rate of degradation, both of which are under separate control (Goldberg and Odessey, 1974). During exercise, the rate of protein synthesis in the muscle of Fundulus heteroclitus is not decreased (Jackim and La Roche, 1973); thus, control may be exerted over protein degradation. Unfortunately, very little is known about the control of proteolysis in any system (Goldberg and Odessey, 1974), and all that can be said regarding the process in fish muscle is that this tissue has an active and wide range of proteolytic enzymes (Tarr, 1972). The metabolism of amino acids is, however, slightly better understood and we shall now address ourselves to this area. In fish muscle the two major free amino acids are glycine and histidine. In numerous fish species, including a variety of carps and salmonids, glycine and histidine either individually or in combination make up about 50% of the free amino acid pool. Moreover, both are utilized extensively during starvation (Creach, 1966; Siddiqui et al., 1973; Wood et al., 1960; Fontaine and Marchelidon, 1971).The utilization of muscle glycine may involve prior conversion to glucose in the liver, since Demael-Suard et al. (1974) found that labeled glycine injected into tench (Tinca vulgaris) is rapidly incorporated into hepatic glycogen. The catabolism of histidine, on the other hand, appears to occur in muscle itself. The first step in the breakdown of histidine, involving the formation of free NH4+ and urocanate, is catalyzed by histidase (EC 4.3.1.3) which is present in mackerel (Scomber japonicus) muscle but is thought not to occur in mammalian muscle (Sakaguchi et al., 1970). Moreover, histidase from muscle has aK, for histidine which is 25 times lower than the liver analogue. In addition, measurements of free histidine in all of the tissues of migrating sockeye salmon are consistent with the contention that oxidation occurs in the muscle (Wood et al., 1960). It is probable that not only histidine but other amino acids are utilized for energetic purposes directly in the muscle. Fish muscle contains a large number of transaminating enzymes with glutamate-pyruvate transaminase (EC 2.6.1.2) and glutamate-oxaloacetate transaminase (EC 2.6.1.1) occurring in par-

532

WILLIAM R. DRIEDZIC AND P. W. HOCHACHKA

ticularly high titers (Siebert et al., 1965), as well as high activities of 5’-AMP deaminase (Markarewicz, 1969; Fields, 1975), and detectable levels of glutamate dehydrogenase (EC 1.4.1.2)(Siebert et al., 1965; McBean et al., 1966).Thus, although the site of oxidation is not known with certainty, Cyprinus carpio injected with labeled glucose, alanine, or glutamate oxidize the two amino acids to CO, much more readily than glucose (Nagai and Ikeda, 1971,1972,1973).This again is consistent with the hypothesis that prior conversion of amino acids to glucose is not necessary in order that the carbon skeletons be utilized for energetic purposes. Preparations of mammalian muscle can readily oxidize singularly alanine, glutamate, aspartate, isoleucine, leucine, and valine but apparently no other amino acids (Goldberg and Odessey, 1972; Beatty et al., 1974). However, when any of the branch chain amino acids are catabolized, provided glucose is available, alanine and glutamate accumulate and under these conditions function as nitrogen acceptors. As previously pointed out, nitrogen excretion does not create a problem for aquatic species burning amino acids. In a recent experiment with Cyprinus carpio, NH4+levels in white muscle increased from about 16 pmoles/g at rest to about 25 pmoles/g after a short period of activity at a sustainable velocity (Driedzic, 1975). The high resting levels of NH4+ relative to other experiments (Table I) suggest that proteins are being utilized as an energy source even under routine metabolic conditions and the large increase in NH4+ concentration after swimming shows an increased mobilization of amino acids during the increased energy demand situation. It is not known why amino acids were being called upon in this particular experiment. The fish were fed daily; however, the diet may have been inadequate. Other possibilities such as a seasonal switch in metabolic fuel at the present time cannot be ruled out. Clearly, the entire concept of protein and amino acid metabolism in fish tissues must be reexamined. Aside from the direct value of their carbon skeletons amino acids are important energetically for other reasons. During burst activity by white muscle, the adenylate pool is decreased with a concomitant increase in NH4+and IMP. If the NH4+liberated from the adenylate pool cannot be fixed back into the mainstream of catabolism then the tissue has essentially contracted a nitrogen debt. The free amino acid pool may be essential in repaying this debt via the purine nucleotide cycle (see Section IV,D). A further important aspect of amino acid catabolism is in providing four carbon units for augmentation of the citric acid cycle pool. This process shall be discussed in the next section.

8.

METABOLISM IN FISH DURING EXERCISE

533

VIII. CITRIC ACID CYCLE The citric acid cycle is often and appropriately termed the hub of cellular metabolism. Through a series of reactions that are initiated by oxaloacetate and which produce oxaloacetate, the cycle supplies a means for the complete oxidation of acetyl-CoA derived from either fats or carbohydrates. The citric acid cycle in fish is similar to that in other more frequently studied organisms in that (a) the intermediates have been identified, shown to be labeled in the appropriate order, upon the addition of substrate to mitochondrial preparations, and (b) the majority of the prerequisite enzymes have been detected (Hochachka, 1969; Tarr, 1972). However, very little information is available on the control of this cycle in fish tissues. Thus, it will be useful to review briefly what is current19 known about its regulation in other species. The main function of the citric acid cycle is to produce reduced flavin and pyridine nucleotides which are subsequently oxidized via the electron transport chain with the concurrent production of ATP. Hence, not unexpectedly, carbon flux through the cycle is regulated largely by ATP : ADP and NADH : NAD+ ratios. Studies at the tissue, mitochondrial, and enzyme level indicate a minimum of three control sites: citrate synthetase, isocitrate dehydrogenase, and a-ketoglutarate dehydrogenase. Control points within the cycle may differ with the physiological demand upon the tissue and during increased energy demand situations may even differ with the fuel source utilized. The reaction controlled by citrate synthetase (EC 4.1.3.7) represents the point of entry of acetyl-CoA into the Krebs cycle. There is widespread agreement h a t the major control on this enzyme is through the availability of its substrates, acetyl-CoA and oxaloacetate (La Noue et al., 1972; Randle et al., 1970). The purified enzyme, however, is inhibited by ATP from a variety of sources (Srere, 1974) including trout liver (Hochachka and Lewis, 1970), but mitochondrial studies indicate that the foremost inhibitor of citrate synthetase in rat heart is succinyl-CoA (Williamson et al., 1972). Regardless, the importance of these inhibitors during increased energy demand situations is questionable since ATP remains relatively constant, at least in highly aerobic muscles (see Section IV,A) and succinyl-CoA levels are elevated during increased activity in the rat heart (Neely et al., 1972b). In the transition from a resting to a highly active state there is an increased demand for oxaloacetate to keep pace with the production of acetyl-CoA. In some cases this demand is met not only by the intermediates within the cycle but also by supplementation from exo-

534

WILLIAM R. DFUEDZIC AND P. W. HOCHACHKA

genously derived sources. This phenomenon is particularly well documented in the perfused rat heart burning glucose, where augmentation of oxaloacetate plus other Krebs cycle intermediates is fully accounted for by aspartate depletion (Fig. 8). Under these conditions aspartate is transaminated with a-ketoglutarate to form glutamate and oxaloacetate. The latter product enters the pool of citric acid cycle intermediates, whereas the glutamate is further transaminated with glycolytically derived pyruvate to form alanine and regenerate a-ketoglutarate. Thus, by this reaction scheme. there is a 1 : l stoichiometric relationship between aspartate depletion and alanine accumulation to maintain nitrogen balance (Safer and Williamson, 1973). During activity by carp (Cyprinus carpio) white muscle, there is a significant, albeit small, decrease in aspartate levels which are associated with an increase in alanine (Driedzic and Hochachka, 1976). This observation is probably related to priming the Krebs cycle in this tissue, since white muscle does have the capacity to increase its oxygen consumption to a limited extent (Wittenberger and Diaciuc, 1965). In muscle burning fats, though, the situation is far more complex in that pyruvate ceases to be available for transamination with glutamate. In fact, with the transition to higher work loads, aspartate

glucose

1

aspartate

CYCLE

Fig. 8. Mechanism of augmenting oxaloacetate reserves during carbohydrate catabolism.

8.

METABOLISM IN FISH DURING EXERCISE

535

levels in the perfused rat heart burning fats increases not decreases; moreover, alanine levels remain constant (Neely et al., 1972a,b). The second potentially important site of citric acid cycle control, with elevated work loads, is at the reaction catalyzed by isocitrate dehydrogenase (EC 1.1.1.4). This enzyme catalyzes the decarboxylation of isocitrate to a-ketoglutarate with the concomitant reduction of either NAD+ or NADP+. Zn vitro studies show that the enzyme is generally inhibited by reduced pyridine nucleotides and ATP, whereas activity is facilitated by ADP (Plaut, 1970). Metabolic regulation at the NAD+-linked site has been demonstrated in insect flight muscle (Johnson and Hansford, 1975) and rat heart (Neely et al., 1972b), both of which, upon the initiation of activity, show decreased levels of citrate and isocitrate with associated increased contents of a-ketoglutarate. These are interesting observations for they show that one mechanism by which muscle can meet the increased demands for oxaloacetate is simply to repartition the carbon within the cycle itself. Thus, at the expense of citrate and isocitrate, levels of oxaloacetate and other citric acid cycle intermediates increase. In the insect flight muscle, activity of the NAD+-linked enzyme appears to be due to decreasing ATP:ADP ratios (Johnson and Hansford, 1975) but in rat heart mitochondria the NADH : NAD+ ratio seems to be more important (La Noue et al., 1970). Regulation of isocitrate catabolism in fish tissues may be more complex than either the insect or the mammal. In trout muscle the activity of the NADP+-linked isocitrate dehydrogenase is 400 times higher than the NAD+-linked enzyme (Newsholme and Start, 1973),and in trout liver the latter enzyme is undetectable (Moon and Hochachka, 1971). Complexity arises since the NADP+-linked enzyme is cytosolic and the reactions of the Krebs cycle are of course mitochondrial. Clearly, this is a topic which must be further elucidated. The third well-established, but b y no means final, site of citric acid cycle control is at the reaction catalyzed by a-ketoglutarate dehydrogenase (EC 1.2.4.2).Isotope studies indicate this site is rate limiting in the perfused rat heart burning acetate (Randle et al., 1970).In rat heart mitochondria a-ketoglutarate dehydrogenase is inhibited by succinyl-CoA and NADH, the latter potentiating the effect of the former (La Noue et al., 1970, 1972). However, as has been previously pointed out, succinyl-CoA levels increase during activity (Neely et al., 1972b), thus rendering this a questionable control mechanism during increased levels of exercise. It is obvious that there is much room for studies of the citric acid cycle not only in fish but in all organisms. In fish the separation of the

536

WILLIAM R. DRIEDZIC AND P. W. HOCHACHKA

red and white musculatures with their preferred fuel sources provides an exceptional experimental system in which to sort out differences in control phenomenon. We hope in the future that this will be an active site of research. ACKNOWLEDGMENTS During the writing of this chapter W. R. D. was the recipient of a Canadian National Research Council Graduate Scholarship.

REFERENCES Atkinson, D. E. (1965). Biological feedback control at the molecular level. Science 150, 851-857. Bass, A., O&idal, B., Pelouch, V., and Vitek, V. (1973). Differences in weight parameters, myosin-ATPase activity and the enzyme pattern of energy supplying metabolism between the compact and spongious cardiac musculature of carp (Cyprinus carpio) and turtle (Testudo horsfieldi).Pjdgers Arch. 343, 65-77. Beatty, C. H., Curtis, S., Young, M. K., and Bocek, R. M. (1974). Oxidation of amino acids by red and white muscle fiber groups. Am. J . Physiol. 227, 268-272. Bendall, J. R., and Taylor, A. A. (1970). The Meyerhof quotient and the synthesis of glycogen from lactate in frog and rabbit muscle. A reinvestigation. Biochem. J . 118, 887-893. Bilinski, E. (1963). Utilization of lipids by fish I. Fatty acid oxidation by tissue slices from dark and white muscle of rainbow trout (Salmo gairdneri).Can. J . Biochem. Physiol. 41, 107-112. Bilinski, E. (1969). Lipid catabolism in fish muscle. In “Fish in Research” (0.W. Neuhaus and J. E. Halver, eds.), pp. 135-151. Academic Press, New York. Bilinski, E. (1974). Biochemical aspects of fish swimming. In “Biochemical and Biophysical Perspectives in Marine Biology” (D. C. Malins and J. R. Sargent, eds.), Vol. 1, pp. 247-288. Academic Press, New York. Bilinski, E., and Jonas, R. E. E. (1964). Utilization of lipids by fish. 11. Fatty acid oxidation by a particulate fraction from lateral line muscle. Can. J . Biochem. 42, 345-352. Bilinski, E., and Jonas, R. E. E. (1970).Effects of coenzyme A and carnitine on fatty acid oxidation by rainbow trout mitochondria. J. Fish. Res. Board Can. 27, 857-864. Bilinski, E., and Jonas, R. E. E. (1972).Oxidation of lactate to carbon dioxide by rainbow trout (Salmo gairdneri) tissues. J . Fish. Res. Board Can. 29, 1467-1471. Bilinski, E., and Lau, Y. C. (1969). Lipolytic activity toward long-chain triglycerides in lateral line muscle of rainbow trout (Salmogairdneri).J.Fish. Res. Board Can. 26, 1857- 1866. Bilinski, E., Jonas, R. E. E., and Lau, Y. C. (1971).Lysosomal triglyceride lipase from the lateral line tissue of rainbow trout (Salmo gairdneri).J . Fish. Res. Board Can. 28, 1015-1018. Black, E. C., Robertson Conner, A., Lam, K. C., and Chiu, W. G. (1962). Changes in glycogen, pyruvate, and lactate in rainbow trout (Salmo gairdneri) during and following muscular activity.]. Fish. Res. Board Can. 19,409-436. Black, E. C., Manning, G. T., and Hayashi, K. (1966). Changes in levels of hemoglobin,

8.

METABOLISM IN FISH DURING EXERCISE

537

oxygen, carbon dioxide, pyruvate, and lactate in venous blood of rainbow trout (Salmo gairdneri) during and following severe muscular activity. J . Fish. Res. Board Can. 23,783-795. Bone, Q. (1966).On the function of the two types of myotomal muscle fibre in elasmobranch fish. J. Mar. Biol. Ass. U . K . 46, 321-349. Bostrom, S. L., and Johansson, R. G. (1972). Enzyme activity patterns in white and red muscle of the eel (Anguilla anguilla) at different developmental stages. Comp. Biochem. Physiol. B 42, 533-542. Brett, J. R. (1964).The respiratory metabolism and swimming performances of young sockeye salmon. J . Fish. Res. Board Can. 21, 1183-1226. Burt, J. R.,and Stroud, G. D. (1966).The metabolism of sugar phosphates in cod muscle. Nippon Suisan Gakkaishi 32, 204-212. Buttkus, H. (1963).Red and white muscle of fish in relation to rigor mortis. J. Fish. Res. Board Can. 20, 1607-1612. Chapman, A. G., and Atkinson, D. E. (1973). Stabilization of adenylate energy charge by the adenylate deaminase reaction. J . Biol. Chem. 248, 8309-8312. Cohen, P., Duewer, T., and Fischer, E. H. (1971).Phosphorylase from dogfish skeletal muscle. Purification and a comparison of its physical properties to those of rabbit muscle phosphorylase. Biochemistry 10, 2683-2694. Crabtree, B., and Newsholme, E. A. ( 1972).The activities of phosphorylase, hexokinase, phosphofructokinase, lactate dehydrogenase and the glycerol Sphosphate dehydrogenases in muscles from vertebrates and invertebrates. Biochem. J . 126,49-58. Creach, Y. (1966). Thiols protbines et acid& amines libres des tissus chez la carpe (Cyprinus carpio L.) au cours du jeune prolonge. Arch. Sci. Physiol. 20, 115-121. Creach, Y., and Serfaty, A. (1974). Le j e t n e e t la realimentation chez la carpe (Cyprinus carpio L.). J . Physiol. (Paris)68, 245-260. Dejours, P., Armand, J,, and Verriest, G. (1968). Carbon dioxide dissociation curves of water and gas exchange of water-breathers. Respirat. Physiol. 5, 23-33. Demael-Suard, A,, Garin, D., Brichon, G., Mure, M., and Peres, G. (1974). Neoglycogenese a partir de la glycine ‘*C chez la tanche (Tinca oulgaris L.) au cows de I’asphyxie. Comp. Biochem. Physiol. A 47, 1023-1033. Deuticke, B., and Gerlach, E. (1966). Abbau freier Nucleotide in Herz, Skelett-muskel, Gehirn und Leber der Ratte bei Sauerstoffmangel. Pfluegers Arch. Cesamte Physiol. Menschen Tiere 292,239-254. Driedzic, W. R. (1975).Energy metabolism in carp white muscle. Ph.D. Thesis, Univ. of British Columbia, Vancouver. Driedzic, W. R., and Hochachka, P. W. (1975). The unanswered question of high anaerobic capabilities of carp white muscle. Can.J . Zool. 53, 706712. Driedzic, W. R.,and Hochachka, P. W. (1976). Control of energy metabolism in carp white muscle. Am. J . Physiol. 230, 579-582. Driedzic, W. R., and Kiceniuk, J. W. (1976).Blood lactate levels in free swimming trout (Salmo gairdneri) before and after exercise resulting in fatigue. J . Fish. Res. Board Can. 33, 173-176. Drummond, G. I. (1967). Muscle metabolism. Fortschr. Zool. 18, 359-429. Drummond, G. I. (1971). Microenvironment and enzyme function: Control of energy metabolism during muscle work. Am. Zool. 11, 83-87. Drury, D. R., and Wick, A. N. (1956). Metabolism of lactic acid in intact rabbit. Am. J . Physiol. 184, 304-308. Edington, D. W., Ward, G. R., and Saville, W. A. (1973).Energy metabolism of working muscle: Concentration profiles of selected metabolites. Am. J . Physiol. 224, 13751380.

538

WILLIAM R. DRIEDZIC AND P. W. HOCHACHKA

Farkas, T. (1967a). The effect of catecholamines and adrenocorticotrophic hormone on blood and adipose tissue FFA levels in the fish Cyprinus carpio L. Prog. Biochem. Pharmacol. 3, 314-319. Farkas, T. (1967b). Examinations of the fat metabolism in freshwater fishes-the sympathetic nervous system and the mobilization of fatty acids. Ann. Biol. (Tihany)34, 129-138. Farkas, T. (1969).Studies 011 the mobilization offats in lower vertebrates. Actu Biochim. Biophys. Acud. Sci. Hung. 4, 237-249. Fields, J. H. A. (1975). Personal communication. Fontaine, M., and Marchelidon, J. (1971). Amino acid contents of the brain and the muscle of young salmon (Salmo salar L.) at parr and smolt stages. Comp. Biochem. Physiol. A 40, 127-134. Fraser, D. I., Dyer, W. J,, Weinstein, H. M., Dingle, J. R., and Hines, J. A. (1966). Glycolytic metabolites and their distribution at death in the white and red muscle of cod following various degress of antemortem muscular activity. Can.J . Biochem. 44, 1015-1033. Freed, J. M. (1971). Properties of muscle phosphofructokinase of cold- and warmacclimated Carassius uuratus. Comp. Biochem. Physiol. B 39, 747-764. George, J. C. (1962). A histophysiological study of the red and white muscle of the mackerel. Am. Midl. Not. 68, 487-494. George, J. C., and Bokdawala, F. D. (1964). Cellular organization and fat utilization in fish muscle.]. Anim. Morphol. Physiol. 11, 124-132. Gesser, H., and Poupa, 0. (1974). Relations between heart and muscle enzyme pattern and directly measured tolerance to acute anoxia. Comp. Biochem. Physiol. A 48, 97-103. Goldberg, A. L., and Dice, J. F. (1974). Intracellular protein degradation in mammalian and bacterial cells. Annu. Reo. Biochem. 43, 835-869. Coldberg, A. L., and Odessey, R. (1972).Oxidation of amino acids by diaphragms from fed and fasted rats. Am. J . Physiol. 223, 1384-1391. Goldberg, A. L., and Odessey, R. (1974).Regulation of protein and amino acid degradation in skeletal muscle. Excerpta Med. Int. Congr. S e r . No. 333, pp. 187-199. Gosselin-Rey, C., Hamoir, G., and Scopes, R. K. (1968).Localization of creatine kinase in the starch-gel and moving-boundary electrophoretic patterns of fish muscle. J . Fish. Res. Board Can. 25, 2711-2714. Gras, J., Gautheron, D., and Henry, J. C. (1967). Fractions phosphorylees et nucleotides libres d e divers organes de la Truite Arc-en-Ciel (Salmo gairdnerii Rich.). C . R. Acad. Sci., Ser. D 265, 1547-1549. Hammond, B. R.,and Hickman, C. P., Jr. (1966).The effect of physical conditioning on the metabolism of lactate, phosphate, and glucose in rainbow trout, Salmo gairdneri. J . Fish. Res. Board Can. 23, 65-83. Hamoir, G., Focant, B., and Distkche, M. (1972). Proteinic criteria of differentiation of white, cardiac and various red muscles in carp. Comp. Biochem. Physiol. B 41, 665-674. Heath, A. G., and Pritchard, A. W. (1965). Effects of severe hypoxia on carbohydrate energy stores and metabolism in two species of freshwater fish. Physiol. Zool. 38, 325-334. Hochachka, P. W. (1969). Intermediary metabolism in fishes. In “Fish Physiology” (W. S. Hoar and D. J. Randall, eds.), Vol. 1, pp. 351-390. Academic Press, New York. Hochachka, P. W., and Lewis, J. K. (1970). Enzyme variants in thermal acclimation. Trout liver citrate synthases. J . Biol. Chem. 245, 6567-6573.

8.

METABOLISM IN FISH DURING EXERCISE

539

Hochachka, P. W., Neely, J. R., and Driedzic, W. R. (1977). Integration of lipid utilization with Krebs cycle activity in muscle. Fed. Proc. 36, 2009-2014. Hochachka, P. W., and Storey, K. B. (1975). Metabolic consequences of diving in animals and man. Science 187,613-621. Hudson, R. C. L. (1973). On the function of the white muscles in teleosts at intermediate swimming speeds. J . E x p . Biol. 58,509-522. Imai, S., Riley, A. L., and Berne, R. M. (1964).Effect of ischemia on adenine nucleotides in cardiac and skeletal muscle. Circ. Res. 15, 443-450. Infante, A. A., and Davies, R. E. (1962). Adenosine triphosphate breakdown during a single isotonic twitch of frog sartorius muscle. Biochem. Biophys. Res. Commun. 9, 4 10-4 15. Issekutz, B., Jr., Shaw, W. A. S., and lssekutz, A. C. (1976).].Appl. Physiol. 40,312-319. Jackini, E., and La Roche, G. (1973). Protein synthesis in Fundulus heteroclitus muscle. Comp. Biochem. Physiol. A 44, 851-866. Johnson, R. N., and Hansford, R. G . (1975). The control of tricarboxylate-cycle oxidations in blowfly flight muscle. The steady-state concentrations of citrate, isocitrate, 2-oxoglutarate and nialate in flight muscle and isolated mitochondria. Biochem. ]. 146, 527-535. Johnston, I. A. (1975). Pyruvate kinase from the red skeletal musculature of the carp (Carassiuscarassius L.). Biochem. Biophys. Res. Commun. 63, 115-122. Johnston, I. A., and Goldspink, G . (19734. Quantitative studies of muscle glycogen utilization during sustained swimming i n crucian carp (Carassius carassius L.). J. E x p . Biol. 59, 607-615. Johnston, I. A., and Goldspink, G. (1973b).A study ofthe swimming performance of the crucian carp Carassius carassius (L.) in relation to the effects of exercise and recovery on biochemical changes in the myotomal muscles and liver. J . Fish. Biol. 5, 249-260. Johnston, I. A., and Goldspink, G. ( 1 9 7 3 ~ )A. study of glycogen and lactate in the myotomal muscles and liver of the coalfish (Gadus uirens L.) during sustained swimming.]. Mar. Biol. Assn. U . K . 53, 17-26. Johnston, I. A. and Goldspink, G . (1973d). Some effects of prolonged starvation on the metabolism of the red and white myotomal muscles of the plaice (Pleuronectes platessa). Mar. Biol. 19, 348-353. Johnston, I. A., Frearson, N., and Goldspink, G. (1972). Myofibrillar ATPase activities of red and white myotomal muscle of marine fish. Experientia 28, 713-714. Jonas, R. E. E., and Bilinski, E. (1964). Utilization of lipids by fish. 111. Fatty acid oxidation by various tissues from sockeye salmon (Oncorhynchus nerka). J. Fish. Res. Board Can. 21,653-656. Jones, D. R. (1971). Theoretical analysis of factors which may limit the maximum oxygen uptake of fish: The oxygen cost of the cardiac and branchial pumps.]. Theor. Biol. 32,341-349. Jones, N. R., and Murray, J. (1960). The acid-soluble nucleotides of codling (Gadus callarias) muscle. Biochem. J . 77, 567-575. Jorfeldt, L. (1970). Metabolism of L( +) lactate in human skeletal muscle during exercise. Acta Physiol. Scand., S u p p l . No. 338, pp. 1-67. Kayne, F. J. (1973). Pyruvate kinase. In “The Enzymes” (P. D. Boyer, ed.), 3rd Ed., Vol. 8, pp. 353-383. Academic Press, New York. Kemp, R. G. (1973). Inhibition of muscle pyruvate kinase by creatine ph0sphate.J. Biol. Chem. 248,3963-3967. Keul, J., Doll, E., and Keppler, D. (1972). “Energy Metabolism of Human Muscle,” pp. 104-125. Univ. Park Press, Baltimore, Maryland.

540

WILLIAM R. DFUEDZIC AND P. W. HOCHACHKA

Krishnamoorthy, R. V., and Narasimhan, T. (1972). Ascorbic acid and fat content in the red and white muscle ofcarp (Catla catla).Comp. Biochem. Physiol. B 43,991-997. Krueger, H. M., Saddler, J . B., Chapman, G. A., Tinsley, I. J., and Lowry, R. R. (1968). Bioenergetics, exercise and fatty acids of fish. Am. Zool. 8, 119-129. Kutty, M. N. (1968). Respiratory quotients in goldfish and rainbow trout.]. Fish. Res. Board Can. 25, 1689-1728. Kutty, M. N. (1972). Respiratory quotient and ammonia excretion in Tilapia mossambica. Mar. Biol. 16, 126-133. Kutty, M. N., and Mohamed, M. P. (1975).Personal communication. La Noue, K. F., Nicklas, W. J , , and Williamson, J, R. (1970).Control of citric acid cycle activity in rat heart mitochoi1dria.J. Biol. Chem. 245, 102-111. La Noue, K. F., Bryla, J., and Williamson, J . R. (1972). Feedback interactions in the control of citric acid cycle activity in rat heart mitochondria. J . Biol. Chem. 247, 667-679. Levy, M. N. (1962).Uptake of lactate and pyruvate by intact kidney of the dog. Am. J . Physiol. 202,302-308. Lim, S. T., Kay, R. M., and Bailey, G. S. (1975). Lactate dehydrogenase isozymes of salmonid fish: Evidence for unique and rapid functional divergence of duplicated H, lactate dehydrogenases. J . Biol. Chem. 250, 1790-1800. Lin, Y., Dobbs, G. H., and De Vries, A. L. (1974).Oxygen consumption and lipid content in red and white muscle of Antarctic fishes.J. Exp. Zool. 189, 379-385. Love, R. M. (1970)."The Chemical Biology of Fishes." Academic Press, New York. Lowenstein, J. M. (1972).Ammonia production in muscle and other tissues: the purine nucleotide cycle. Physiol. Rev. 52, 382-414. McBean, R. L., Neppel, M. J., and Goldstein, L. (1966). Glutamate dehydrogenase and ammonia production in the eel (Anguilla rostruta). Comp. Biochem. Physiol. 18, 909-920. Mallov, S., and Alousi, A. A. (1969). In uitro effect of epinephrine on lipase activity of heart. Am. J . Physiol. 216, 794-799. Mansour, T. E. (1972). Phosphofructokinase. C u n . Top. Cell. Regul. 5, 1-46. Markarewicz, W. ( 1969). AMP-aminohydrolase in muscle of elasmobranch fish. Purification procedure and properties of the purified enzyme. Comp. Biochem. Physiol. 29, 1-26. Mattisson, A. G. M., Johansson, R. C., and Bostrom, S. L. (1972). The cellular localization of lactate dehydrogenase in skeletal muscle of eel (Anguilla anguilla). Comp. Biochem. Physiol. B 41, 475-482. Moon, T. W., and Hochachka, P. W. (1971). Temperature and enzyme activity in poikilotherms. Isocitrate dehydrogenase in rainbow trout liver. Biochem. J . 123, 695-705. Muirhead, K. M., and Bishop, S. H. (1974).Purification of adenylosuccinate synthetase from rabbit skeletal musc1e.J. Biol. Chem. 249, 459-464. Mustafa, T., Moon, T. W., and Hochachka, P. W. (1971). Effects of pressure and temperature on the catalytic and regulatory properties of muscle pyruvate kinase from an off shore benthic fish. Am. Zool. 11, 451-466. Nagai, M., and Ikeda, S. (1971). Carbohydrate metabolism in fish. 11. Effect of dietary composition on metabolism of glucose-6J4C in carp. Nippon Suisan Gakkafshi 37, 410-414. Nagai, M., and Ikeda, S. (1972). Carbohydrate metabolism in fish. 111. Effect of dietary composition on metabolism of glucose-U-14Cand glutamate-U-"C in carp. Nippon Suisan Gakkaishi 38, 137-143. Nagai, M., and Ikeda, S. (1973). Carbohydrate metabolism in fish. IV. Effect of dietary

8.

METABOLISM IN FISH DURING EXERCISE

541

composition on metabolism of acetate-U-I4C and L-alanine-U-’4C in carp. Nippon Suisan Cukkaishi 39, 633-643. Nakatano, T., and Tomlinson, N. ( 1967). Catecholamine and carbohydrate concenbations in rainbow trout (Salmo gairdneri) in relation to physical disturbance. J . Fish. Res. Board Can. 24, 1701-1715. Neely, J. R., and Morgan, H. E. (1974). Relationship between carbohydrate and lipid metabolism and the energy balance of heart muscle. Annu. Rev. Physiol. 36,413459. Neely, J . R., Denton, R. M., England, P. J., and Randle, P. J. (1972a). The effects of increased heart work on the tricarboxylate cycle and its interactions with glycolysis in the perfused rat heart. Biochem. J. 128, 147-159. Neely, J. R., Rovetto, M. J., and Oram, J. F. (197213). Myocardial utilization of carbohydrate and lipids. Prog. Cardiooasc. Dis. 15, 289-329. Newsholme, E. A. (1972). The regulation of phosphofructokinase in muscle. Cardiology 56, 22-34. Newsholnie, E. A,, and Start, C. (1973).“Regulation in Metabolism,” p. 134. Wiley, New York. Noda, L., Schulz, G. E., and Von Zabern, I. (1975).Crystalline adenylate kinase from carp muscle. Eur. J . Biochem. 51, 229-235. Norum, K. R., and Bremer, J. (1966).The distribution of palmityl-CoA:carnitine palmityltransferase in the animal kingdom. Comp. Biochem. Physiol. 19, 483-487. Odessey, R., Khairallah, E. A., and Goldberg, A. L. (1974). Origin and possible significance of alanine production by skeletal muscle. J . Biol. Chem. 249, 76237629. Oram, J. F., Bennetch, S. L., and Neely, J. R. (1973).Regulation of fatty acid utilization in isolated perfused rat hearts.]. Biol. Chem. 248, 5299-5309. Oram, J. F., Wenger, J. I., and Neely, J. R. (1975). Regulation of long chain fatty acid activation in heart musc1e.J. Biol. Chern. 250, 73-78. Parks, R. E., Jr., and Aganval, R. P. (1973). Nucleoside diphosphokinases. In “The Enzymes” (P. D. Boyer, ed.), 3rd Ed., Vol. 8, pp. 307-334. Academic Press, New York. Pequin, L., and Serfaty, A. (1962).L’ammoniemie chez les Teleosteens dulciocoles lors de I’emersion et de I’anoxie brutale. C . R. Soc. Biol. 156, 1167-1171. Pette, D., and Staudte, H. W. (1973). Differences between red and white muscles. I n “Limiting Factors of Physical Performance” (J. Keul, ed.), pb. 23-31. Thieme, Stuttgart. Philpott, C. W., and Copeland, D. E. (1963). Fine structure of chloride cells from three species of Fundulus. J . Cell Biol. 18, 389-404. Plaut, G . W. E. (1970). DPN-linked isocitrate dehydrogenase of animal tissues. Cum. Top. Cell. Regul. 2, 1-27. Pocinwong, S., Blum, H . E., and Fischer, E. H. (1974). Purification and properties of dogfish skeletal muscle phosphorylase kinase. Fed. Proc., Fed. Am. Soc. E x p . Biol. 33(2), Abstr. No. 277. Pradhan, T. K., and Criss, W. E. (1974).Modulation of mitochondria1 adenylate kinase by citric acid cycle intermediates. Eur. J . Biochem. 43, 541-547. Pritchard, A. W., Hunter, J. R., and Lasker, R. (1971).The relation between exercise and biochemical changes in red and white muscle and liver in the jack mackerel, Trachurus symmetricus. U S . Fish Wildl. Serv., Fish. Bull. 69, 379-386. Randall, R. F., and Anderson, P. J. (1975). Purification and properties of the pyruvate kinase of sturgeon muscle. Biochem. J. 145, 569-573. Randle, P. J., England, P. J., and Denton, R. M. (1970). Control of the tricarboxylate

542

WILLIAM R. DFUEDZIC AND P. W. HOCHACHKA

cycle and its interactions with glycolysis during acetate utilization in rat heart.

Biochem. J. 117,677-695. Rao, G. M. M. (1968). Oxygen consumption of rainbow trout (Salmo gairdneri) in relation to activity and salinity. Can. J. 2001.47, 131-134. Reichard, G. A., Jr., Moury, N. R., Jr., Hochella, N. J., Patterson, A. L., and Weinhouse, S. (1963). Quantitative estimation of the Cori cycle in the human.J. Biol. Chem. 238, 495-50 1. Rowell, L. B., Kraning, K. K., 11, Evans, T. O., Kennedy, J. W., Blackmon, J. R., and Kusumi, F. (1966). Splanchnic removal of lactate and pyruvate during prolonged exercise in man.]. Appl. Physiol. 21, 1773-1783. Safer, B., and Williamson, J. R. ( 1973). Mitochondrial-cytosolic interactions in perfused rat heart: Role of coupled transaminations in repletion of citric acid cycle intermediates. J . Biol. Chem. 248, 2570-2579. Sakaguchi, M., Sugiyama, M., Sugiyama, T., and Kawai, A. (1970).Histidine metabolism in fish. IV. Comparative study on histidine deaminases from muscle and liver of mackerel and from bacteria. Nippon Suisan Gakkaishi 36, 200-206. Siddiqui, A. Q., Siddiqui, A. H., and Ahmad, K. (1973). Free amino acid contents of the skeletal muscle of carp at juvenile and adult stages. Comp. Biochem. Physiol. B . 44,

725-728. Siebert, G., Schmitt, A., and Bottke, I. (1965).Enzyme des Aminosaurestofhechsels in der Kabeljaumuskulatur. Arch. Fischereiwiss. 15, 233-244. Smit, H., Amelink-Koutataal, J. M., Vijverberg, J., and Von Vaupel-Klein, J. C. (1971). Oxygen consumption and efficiency of swimming goldfish. Comp. Biochem.

Physiol. A 39, 1-28. Somero, G. N., and Hochachka, P. W. (1968). The effect of temperature on catalytic and regulatory functions of pyruvate kinases of rainbow trout and the Antarctic fish

Trematomus bernacchii. Biochem. J. 110, 395-400. Spitzer, J. J. (1974). Effect of lactate infusion on canine myocardial free fatty acid metabolism in oioo. Am. J . Physiol. 226, 213-217. Srere, P. A. (1974).Controls of citrate synthase activity. Life Sci. 15, 1695-1710. Stevens, E. D. (1968).The effect of exercise on the distribution of blood to various organs in the rainbow trout. Comp. Biochem. Physiol. 25, 615-625. Stevens, E. D., and Black, E. C. (1966).The effect of intermittent exercise on carbohydrate metabolism in rainbow trout, Salmo gairdneri. J. Fish. Res. Board Can. 23,

471-485. Storey, K. B. (1975).Purification and properties of turtle heart creatine kinase: Role for the enzyme in glycolytic control. Int. J. Biochem. 6, 53-59. Suyama, M., Koike, J., and Suzuki, K. (1960).Studies on the glycolysis and the formation of ammonia in the muscle and blood of elasmobranchs. J . Tokyo Uniu. Fish. 46,

51-65. Tarr, H. L. A. (1972). Enzymes and systems of intermediary metabolism. In “Fish Nutrition” (J. E. Halver, ed.), pp. 255-326. Academic Press, New York. Tomlinson, N., Geiger, S. E., and Kay, W. W. (1965). Sodium, potassium and magnesium concentration and weight changes in fish stored in refrigerated sea water in relation to biochemical changes associated with rigor m0rtis.J. Food Sci. 30, 126134. Tornheini, K., and Lowenstein, J. M. (1974). The purine nucleotide cycle. IV. Interactions with oscillations of the glycolytic pathway in muscle extracts.]. Biol. Chem.

249,3241-3247. Tsai, M. Y., Gonzalez, F., and Kemp, R. G. (1975). Physiological significance of phosphofructokinase isozymes. In “Isozymes” (C. L. Markert, ed.), Vol. 2, pp. 819-835. Academic Press, New York.

8.

METABOLISM IN FISH DURING EXERCISE

543

Tsuyulti, H. (1974).Muscle proteins of fishes. In “Chemical Zoology” (M. Florkin and B. T. Scheer, eds.), Vol. 8, pp. 287-306. Academic Press, New York. Walker, M. G., and Pull, G. (1973). Skeletal muscle function and sustained swimming speeds in the coalfish Gadus uirens. L. Comp. Biochem. Physiol. A 44, 495-501. Watts, D. C. ( 1973). Creatine kinase (adenosine 5’-triphosphate-~reatinephosphotransferase). In “The Enzymes” (P. D. Boyer, ed.), 3rd Ed., Vol. 8, pp. 384-456. Academic Press, New York. Watts, R. L., and Watts, D. C. (1974). Nitrogen metabolism in fishes. In “Chemical Zoology” (M. Florkin and B. T. Scheer, eds.), Vol. 8, pp. 369-446. Academic Press, New York. Webb, P. W. (1970). Some aspects of the energetics of swimming of fish with special reference to the cruising performance of rainbow trout. Ph.D. Thesis, Univ. of Bristol, Bristol, England. Wendt, C. A. G., and Saunders, R. L. (1973). Changes in carbohydrate metabolism in young Atlantic salmon in response to various forms of stress. Znt. Atl. Salmon Symposium, 1973, Znt. Atl. Salmon Found., New YorklSt. Andrews pp. 55-82. Williamson, J. R., Smith, C. M., La Noue, K. F., and Bryla, J. (1972).Feedback control of the citric acid cycle. In “Energy Metabolism and the Regulation of Metabolic Processes in Mitochondria” (M. M. Mehlman and R. w. Hanson, eds.), pp. 185-210. Academic Press, New York. Wittenberger, C. (1972). The glycogen turnover in mackerel muscles. Mar. Biol. 16, 279-280. Wittenberger, C., and Diaciuc, I. V. (1965). Effort metabolism of lateral muscles in carp. ]. Fish. Res. Board Can. 22, 1397-1406. Wood, J . D. (1959). Lipase activity in lingcod muscle preparations. Can. J . Biochem. Physiol. 37, 937-943. Wood, J. D., Duncan, D. W., and Jackson, M. (1960). Biochemical studies on sockeye salmon during spawning migration. XI. The free histidine content of the tissues.]. Fish. Res. Board Can. 17,347-351. Yamamoto, M. (1968). Fish muscle glycogen phosphorylase. Can. J. Biochem. 46, 423432.

This Page Intentionally Left Blank

AUTHOR INDEX Numbers in italics refer to the pages on which the complete references are listed. A Aasen, O., 113, 114, 115,177 Abrahamsson, T., 477,497 Adams-Ray, J., 469,492 Agarwal, R. P., 512,541 Ahmad, K., 531,542 Albers, C., 460,492 Alderdice, D. F., 123, 125, 126, 128: 129, 131, 133, 135, 147, 153, 154, 155, 160, 172,175,178 Aleev, Y. G., 17, 27, 28, 29, 34, 40, 42, 43, 47, 48, 73, 74, 88, 89, 111, 112, 172, 241, 254, 255, 256, 259, 278, 279, 290, 291,292,295,308 Alexander, R. McN., 5, 27, 29, 50, 52, 54, 55,56, 57, 58,60,63,64,66,69,70,89, 112,172, 192,219,232,251,253,255, 257, 261, 267, 268, 272, 275,277,308, 309, 364,365, 366, 368,417,419 Aleyev, Y. G., 197, 198,215,219,232, 241, 260,267,280,288,289,291,304,309 Allan, I. R. H., 76, 77,89 Allan, W. H., 198,215,219,232 Alnaes, E., 399,417 Alousi, A. A., 528,540 Amaoka, K., 51,89 Amelink-Koutstaal, J. M., 125, 127, 128, 135, 136, 140, 149, 165, 167,185, 193, 197, 213, 214,215, 217,236, 409,423, 427,430,438,440,499, 504,507,542 Andersen, P., 382, 394, 399, 400,417 Anderson, P. J., 522,541 Anderson, V. M., 86,95, 122,180 Andreasson, S., 83,89 Ansell, A. D., 43,99 Arita, G. S., 30, 37, 63,89 Arloing, S., 372, 386, 405,417 Armand, J.. 515,537 Arnold, D. C., 43,89 Arnold, G. P., 125,173 Atkinson, D. E., 508, 514,536, 537 Atz, J. W., 29,89

Auerbach, A. A., 45,89 Augustinsson, K. B., 469, 471,492 Austin, J. L., 372,417 Auvergnat, R., 459, 460,492

B Badcock, J., 84,89 Bailey, G. S., 522,540 Bainbridge, R., 8, 10, 12, 15, 17, 18,40,68, 76, 89, 116, 117, 119, 123, 129, 130. 136, 137, 139, 140,173, 191, 192, 193, 196, 197,209, 211,213, 214, 215,216, 217,233,270,295,309,362,376,417 Baker, J. T. P., 163,173 Bakken, E., 105,177 Baldridge, H. D., Jr., 255, 263, 264,309 Ball, E. A. R., 170,182 Ballintijn, C. M., 448,492 Bamford, 0. S., 133,180 Barns, R. A., 142, 144, 169,173 Banks, J. R., 50, 93 Banks, J. W., 81,89 Barets, A., 60,89, 372,386, 387, 394, 395, 396,397,402,403,408,411,413,417. Barker, D., 373, 394, 415, 416,417 Barkley, R. A., 353, 354,356 Barr, L. M., 156,183 Barrett, I., 115, 170, 173, 317, 318, 321, 350,356 Bartholomew, G . A., 165,173 Bass, A., 507,536 Bass, G. A., 104,173 Basu, S. P., 123, 140, 158, 166, 173, 440, 492 Bateman, J. B., 485,496 Bates, B., 193, 215,235 Baumgarten-Schumann, D., 461,498 Baz, A., 297,312 Beamish, F. W. H., 112, 117, 122, 125, 126, 127, 128, 129, 130, 131, 133, 134, 136, 137, 139, 140, 142, 147, 149, 151, 152, 153, 154, 156, 158, 160, 163, 164, 166,

545

546 167, 168, 170, 171,173,174,176,177, 181,182, 217,233, 426,433, 438,441, 465, 466,493,494,496 Beatty, C. H., 530,532,536 Beckett, J. S., 327, 331, 349,357 Bejda, A. J., 109, 183 Belaud, A., 478,492 Bell, G. R., 456,499 Bell, W. H., 125, 174 Belyayev, V. V., 85, 86,89,100, 114,174, 221,233,237, 282,284,309,313 Bendall, J. R., 525,536 Bennetch, S. L., 529,541 Bennett, M . V. L., 27, 28, 32, 33,89 Bennion, G. R., 468, 471, 472, 473, 474, 477,478,479,492,500 Berg, L. S., 326,356 Bergman, R. A., 65,89,377,397,398,405, 417 Bergmann, H. H., 28,89 Berne, R. M., 515,539 Bernstein, J. J., 60,89, 90 Bertin, L., 46, 49, 51, 60, 63, 64, 72,90 Best, A. C. G., 377,417 Bettex-Galland, M., 452,492 Beverton, R. J. H., 112,184 Bilinski, E., 151, 164, 168,174, 353,356 405, 408,409,417,417,418, 503, 506, 517, 524, 525, 526, 528,536,539 Birch, M. P., 487, 488, 489,490, 491,492 Bishai, H. M., 115, 123, 124,174 Bishop, S . H., 516,540 Black, E. C., 128, 151, 168, 170,174,177, 183, 350,358, 438,439,460,492,493, 499, 507,517,518,519,524,536,537, ti42 Black, V. S., 158,174, 175 Blacker, R. W., 112,184 Blackmon, J. R., 524,542 Blaxter, J . H. S., 112, 113, 114, 115, 119, 129, 137, 138, 139, 156, 171, 172,175, 183, 243,249,309, 376,418 Blaika, P., 122, 125, 126, 165, 175, 191, 233 Blight, A. R., 192,233 Blinston, G . , 372,423 Bloom, G., 469,492 Bliim, V., 28,90 Blum, H. E., 521,541 Bocek, R. M., 530, 532,536 Boddeke, R., 55,90, 372, 389, 405,418

AUTHOR INDEX Bohun, S., 121,175 Bokdawala, F. D., 525, 528,538 Bombardieri, R. A., 222, 229,233 Bone, Q., 15,25,27, 60,90, 198,220,233, 255,261,268,269,309, 362,369,370, 372, 373,374, 377,382, 386,387,388, 389, 394, 395,403,405,406,407, 408, 411,412,414,415,417,418, 506,507, 526,537 Booth, J., 452, 461, 464,492 Borelli, G . A., 192,233 Bostrom, S.-L., 373, 377, 418, 506, 523, 537,540 Bottke, I., 532,542 Boulenger, G. A., 41,90 Bowden, D. C., 145,180 Boyar, H. C., 115, 119, 123, 128, 135, 140, 141, 147,175 Braekkan, 0. R., 409,418 Braemer, W., 107,179 Brawn, V. M., 104, 115, 147,175 Breder, C. M., Jr., 6, 7, 8, 9, 10, 11, 12, 24, 26,27,28,29,30,31,33, 36, 37, 38, 39, 44, 85, 86,90, 191, 192, 197, 199, 209, 210, 211, 219, 221, 230,231, 232,233, 280, 282,284, 288,289,309 Bregnballe, F., 87,90 Bremer, J., 529,541 Brennen, C., 192,237 Brett, J. R., 31,90, 102, 123, 125, 127, 128, 129, 130, 131, 133, 135, 136, 137, 138, 139, 140, 144, 147, 151, 153, 156, 161, 162,163,164,165,166,167,175,186, 191, 193, 197,215,217, 218,233, 323, 356,357, 426,427,428,430,431,432, 433,434,435,436,437,438,439,440, 443,445,466,468,482,492,493,499, 507,537 Brichon, G . , 531,537 Bridge, T. W., 7, 63,90 Bridges, K. W., 122,182 Brill, R. W., 338,357 Brody, S., 163,175 Brokaw, C. J.. 192,237 Brotchi, J., 384, 387,418 Brown, C. E., 193,233,267,268,270,271, 272,273,274,280,282,309,311, 342, 357, 445,447,448,493 Brown, C. J. D., 104, 105,181 Brown, R. H. J., 123,173 Brunel, P., 112,175, 176

AUTHOR INDEX

Brunings, W., 487,493 Bryla, J., 533, 535,540, ,543 Buckley, R. M., 445,497 Bulleid, M. J., 76, 77,89 Bullock, T. H., 153,176 Burke, R. E., 372,418,419 Burne, R. H., 326,357 Burnstock, G . , 471, 484,494,496 Burrows, R. E., 124, 125, 126, 128, 129, 135, 141, 169,185 Burt, J . R., 519,537 Butler, J. A., 144, 145, 147, 150,176 Butler, M., 348,357 Buttkus, H., 386,419, 506,537 Byers, R. D., 329,331,357 Byrne, J. M., 129, 147,176, 466,493

547

Chevrel, R., 52,90 Chipman, G. C., 104, 105,179 Chiu, W.-G., 168,174, 438, 460,492, 507, 519, 524,536 Chopra, M. G., 241, 270, 290, 293, 296, 299, 300,301,302,303,304,309 Christensen, G. M., 163,181 Chubb, A. D., 382, 386, 389, 411,418 Clague, J. A., 362,423 Clark, H. L., 76,96 Clemens, H. B., 75, 76, 78,90 Clemens, W. A., 79,94 Cobb, J. L. S., 454, 478,493,499 Coburn, R., 462,494 Cohen, P., 521,537 Cole, T.J., 453,493 Collette, B. B., 84,91, 325, 327, 329,357 C Collines, G . B., 438,493 Cone, C. D., Jr., 287,309 Cahn, P. H., 283,284,309 Connor, A. R., 168,170,173,321,350,356, Caillouet, C. W., Jr., 168,176 438,460,492,493 Conte, F., 456,493 Calder, W. A., 111, 168,176 Calow, L. J., 368,419 Cooper, T., 469,495 Cameron, J. N., 167, 187, 449, 450, 454, Copeland, D. E., 517,541 456,457,461,467,475,493,494 Coprean, D. C., 409,424 Campbell, B., 29,90 Cornford, N. E., 220,236, 280,312 Campbell, G., 452, 461,494 Corti, U. A., 117, 176 Cox, E. T., 123, 140, 147,178 Campbell, G . D., 477,494 Crabtree, B., 506,520,537 Capra, M. F., 484,493 Cardot, J., 472,493 Craig, R. E., 105,176 Carey, F.G., 115,181, 317, 318, 321, 325, Creach, Y.,530, 531,537 326, 327,330,331,332,338,344,345, Criss, W. E., 511, 541 346,348,349,356,357,358,373,419 Cummings, W. C., 122,176 Carre, C. G., 487, 488, 489, 490, 491,492 Cushing, D. H., 84, 85, 86, 91, 105, 112, Cavagna, G . A., 8,91, 197,219,233, 445, 176,184 Curtis, S., 530, 532,536 491,494 Cech, J. J.. Jr., 167,187, 467,493 D Cepela, M., 122, 125, 126,175, 191,233 Chandy, M., 72,98 Dahl, H. A., 381, 382, 387, 388, 389,419, Chang, J. T., 104, 162,180 421 Chang, R. K., 248,311 Chang, R. K. C., 319, 320, 322, 323, 324, Dahl, K., 76,91, 107,176 336, 337, 338, 341, 344,347, 352, 353, Dahlberg, M. L., 128, 147, 149, 156, 157, 354,357,358 158,176, 440,493 Dando, P. R., 151,176 Chao, L. N., 327,329,357 Danforth, C. H., 72,91 Chapman, A. G., 514,537 Dannevig, G . , 107,176 Chapman, C. B., 479,493 Darnell, R. M., 84,91 Chapman, G. A., 162, 163,180, 525,540 Davidson, V. M., 103,176 Charm, S. E., 323,357 Chenoweth, H. H., 124, 125, 126, 128, Davies, A. S., 373,419 Davies, D. T., 485,486,493 129, 135, 141, 169,185

548 Davies, R. E., 508,539 Davis, G. E., 128, 147, 156, 163,176, 186, 440,494 Davis, J. C., 443, 445, 449, 450, 456, 457, 461,468, 475,482,493,494,499 Davis, R. E., 84,91 Davison, W., 387,407, 409,421 Davy, F. B., 122,180 Davy, G. F., 86,95 Davy, J., 317,357 Dean, B., 7,91 Dean, J. M., 151,176 Deck, K. A., 474,494 DeGroot, S. J., 17, 81, 91, 100, 122, 176, 184 Dejours, P., 515, 537 DeLacy, A. C., 119, 170, 171, 183, 193, 215,235 Demael-Suard, A., 531,537 Denil, G., 119,176, 193,233 Denton, E. J., 376,419 Denton, R. M., 515, 533, 535,541, 542 Deuticke, B., 515,537 deVeen, J. F., 81,91, 112,176 De Vries, A. L., 506, 526, 527,540 Diaciuc, I. V., 506,514,519, 524,534,543 Diamond, J., 62,91 Dice, J. F., 531,538 Dickie, L. M., 151, 164,174 Dickson, W., 115, 119, 129, 137, 138, 139, 156,175, 243,249,309 Dingle, J. R., 508, 509,538 Disler, N. N., 61, 91 Disteche, M., 405,420, 506, 511, 522,538 Dizon, A. E., 82, 88, 100, 109, 113, 116, 117,187, 243, 246,248,249,311, 319, 320, 322, 323, 324, 336, 337, 338, 342, 344, 347, 352, 353, 354, 356,357,358, 423 Dobbs, G. H., 506, 526, 527,540 Dodson, J. J., 105, 107, 109, 110,176,177 Doll, E., 507, 524, 525,539 Donaldson, E. M., 472,497 Dotson, R. C., 253,254,255,262,263,264, 265,266,276,285,306,307,310 Doudoroff, P., 128, 147, 149, 156, 157, 158, 164,176,177, 440,493,494,499 Douglas, E. L., 456, 475,495 Dow, R. L., 103, 116, 119,177 Dowd, R. G., 105,177 Drees, F., 457, 460,498

AUTHOR INDEX

Driedzic, W. R., 168, 177, 438, 439, 494, 508,509,511,515,519,524, 525,526, 530,532,534,537,539 Drummond, G. I., 151, 168,177, 520,525, 526,537 Drummond, R. A., 121,185 Drury, D. R., 524,537 DuBois, A. B., 8, 91, 197, 219, 233, 445, 491,494 DuBois-Reymond, R., 30, 37,91 Duever, M. J., 129, 131, 135, 149, 153, 154, 155,181 Duewer, T., 521,537 Duncan, D. W., 531,543 Dunel, S., 452,497 Dyer, D. P., 484,496 Dyer, W. J., 508, 509,538 E

Eaton, R. C., 222, 229,233 Eaton, T.H., 51, 63,91 Edgerton, H. E., 27,33,90, 191,230,231, 232,233 Edington, D. W., 515,537 Eggert, B., 41,91 Eide, A., 392, 393,421 Elling, C. H., 438,493 Ellis, D. V., 80,91, 103, 109,177 Emelianov, S. W., 51,91 Emery, A. R., 30, 31, 37, 39, 70, 84,91 Engel, W. K., 372,418,419 England, P. J., 515, 533, 535,541, 542 Enright, J. T., 120, 179 Eschricht, D. F., 317, 331,357 Escobar, R. A,, 85,91 Esterberg, G. F., 104,182 Evans, D. H., 465,494 Evans, T. O., 524,542

F Fange, R., 469, 471, 485,492,498 Falck, B., 472, 478,494 Farkas, T.,525, 526,538 Farlinger, S., 133, 134,177 Farmer, G . J., 125, 127, 128, 160, 166, 167, 177,181, 433,441,465,466,494 Fessard, A., 411,413,419 Fields, J. H. A., 512, 516, 532,538

AUTHOR INDEX

549

Fierstine, H. L., 10, 15, 21, 22, 23, 50,91, Gardella, E. S., 105, 106, 107, 109,179 103, 104, 116, 121,177,186, 191, 197, Gardner, G. R., 163,178 224,226,233, 243, 248,249, 279,280, Garin, D., 531,537 281,289, 291,292, 295,296, 297, 298, Gaspar-Godfroid, A., 405,423 299, 300, 301,302, 303, 304,310,312, Gauley, J. R., 438,493 Gautheron, D., 509,538 327, 349,357,359, 362, 367,419 Findley, J. S., 263,310 Gayduck, V. V., 80,92,178 Fink, B. D., 75,97 Geerlink, P. J., 64,92 Fischer, E. H., 521,541 Geiger, S. E., 508, 510,542 Fishelson, L., 50,91 George, J. C., 506, 525, 526, 528, 538 Fisher, E. H., 521,537 Gerlach, E., 515,537 Fisher, T., 462,494 Gero, D. R., 103, 119, 121,178, 192, 193, Fleming, R. H., 287,312 224,226,233 Flittner, G. A., 75, 76,90 Gesser, H., 455,494, 507, 522, 523,538 Flood, P. R., 377, 379, 381, 393,419 Gibbs, K H., 325,327, $29,357 Focant, B., 405,420, 506, 511, 522,538 Gibbs-Smith, C. H., 193,233 Fontaine, M., 117,177, 531,538 Gibson, E. S., 123, 149, 152, 153, 154,178 Forbes, S., 105,176 Gibson, R. N., 83, 84,92 Forster, R., 462,494 Glass, N. R., 125, 133, 137, 139, 140, 153, Foss, G., 43, 92 175, 430, 432, 433,493 Foster, J., 128,. 147, 156,176, 440,494 Glova, G. J., 128, 133, 137, 139, 158, 159, Fowler, J. A., 48,92 178 Fox, A. C., 144,177 Goadby, P., 329,357 Fox, R. S., 8, 91, 197, 219,233, 445, 491, Godsil, H. C., 326,328,329,331,333,357, 494 358 Francis, R. C., 274,312 Gold, A., 168,178 Francois, Y.,63,92 Goldberg, A. L., 530, 531, 532,538, 541 Franklin, D. L., 468, 471, 475, 478, 480, Goldspink, G., 55, 70, 94, 97, 128, 136, 496 180, 192,216,217,232,234,374,376, Franzini-Armstrong, C., 386,419 377, 386,387,389,407,409,420,421, Franzisket, L., 44,92 422, 504,507,508, 509,511,515, 518, Fraser, D. I., 508, 509, 538 530,539 Fraser-Brunner, A., 327,357 Goldstein, L., 532,540 Frearson, N., 511,539 Golenhofen, K., 474,495 Freed, J. M., 521,538 Gonzalez, F., 521,542 Fridriksson, A., 107, 113, 114, 115,177 Gooding, R. M., 246,310, 322, 353, 354, Fromm, P. O., 457,485,495,498 356 Fry, F. E. J., 115, 122, 123, 126, 128, 129, Goodnight, C. J., 151,176 131, 140, 147, 149, 151, 152, 153, 154, Goodrich, E. S., 51, 63,92 155, 164,177,178,185, 221,233, 246, Goodyear, C. P., 40,92 247,312, 318,319, 320,322, 323, 332, Gordon, M. S., 158,178,322,358,374,419 334,337,342, 346,357,359, 426,427, Gosline, W. A., 49, 50, 58, 60, 63, 65, 67, 69, 70, 71, 92 440,494 Gosselin-Rey, C., 510,538 G Gradwell, N., 40,92 Graham, J. B., 42, 55, 92, 115, 178, 319, 328,329,356,358, 410,419 Gadd, G. E., 199,212,233 Gaiduk, V. V., 109,182, 183 Graham, J. M., 123, 152, 156,178 Gallepp, G. W., 88,92 Gras, J., 509,538 Ganguly, D. N., 40, 44,70,92 Gray, J., 5,6,7, 8, 10, 15,24,41,43,58,61, Cannon, B. J., 452,461,471,477,478,494 92, 93, 123, 129, 147, 149, 178, 191,

550

AUTHOR INDEX

192, 193, 196, 197, 199, ZOO, 206,208, 209, 211, 215, 219,221,222, 224,230, 231,233,234, 290, 291,295,310 Green, D. M., 128, 143, 144, 169,178 Greene, C. H., 53,55,58,59,60,68,70,93 Greene, C. W., 53, 55, 58, 59, 60,68, 70, 93, 168,178 Greenfield, D. W., 43,93 Greenwalt, C. H., 265, 281,310 Greenway, P., 73,93 Greenwood, P. H., 7,24,43,45,46,49,93, 97, 388,419 Greer Walker, M., 81, 93, 109, 130, 131, 133,178, 370,371,372,374,376,387, 407,419 Gregory, W. K., 40, 42, 73,93 Grenholm, A., 63, 65, 68, 70,93 Griffiths, J. S., 125, 126, 128,129,131, 133, 135, 150,. 155,178 Grillner, S., 414,420 Grimstone, A. V., 452,495 Groot, C., 78, 79, 83,94 Grove, J. D., 456, 477,497 Guernsey, D. L., 338,357 Gunter, G., 43, 93 Guthrie, D. M., 50,93

H Hage, R. E., 281,311 Hagiwara, S., 379,385,398,400,402,420 Hall, F. G., 267,310, 445,495 Hall, W. B., 112,183 Hammond, B. R., 122, 123, 130, 131,179, 510,519,538 Hamoir, G., 405,420, 423, 506, 510, 511, 522,538 Hansford, R. G., 535,539 Hanslip, A. R., 438,492 Hanson, D., 468, 487, 488,495,496 Harden Jones, F. R., 105, 106, 112, 113, 179,184 Harder, W., 51, 60,93 Harper, J. J., 127, 183 Harrington, R. W., 48,95 Harris, J. E., 31, 61, 74,93, 191, 193, 211, 230,231,234, 259,310,391,420 Harris, T., 456,493 Harris, V. A., 41, 63 Harris, W. S., 474,495

Hart, J. S., 122, 123, 126, 128, 129, 131, 149, 153, 154, 155,178 Hartman, F. A., 281,310 Hartt, A. C., 78,93,99, 106,184 Hasler, A. D., 83,84,86,93, 105, 106, 107, 108, 109, 110,179,181, 185 Haswell, M. S., 457, 458, 460,495,498 Hawkins, A. D., 112,179 Hayashi, K.,524,536,537 Hazel, J. R., 352,358 Healey, E. G., 61,93 Hearn, D. C., 448,497 Heath, A. G., 165, 167,179, 443,495, 524, 538 Heisler, N., 457, 460,498 Hemmingsen, E. A,, 456, 475,495 Henderson, H. F., 104, 105,106,107, 109, 179 Henkart, M. P., 379,398,420 Henry, J. C., 509,538 Herald, E. S., 30, 34, 41, 42, 76, 78.93 Hergenrader, G. L., 83, 86,93 Hertel, H., 24,28, 33,44, 45,93, 197,212, 219,221,222,224,225,234, 241,263, 265,267,272,277,298,310 Hester, F., 317, 318,356 Hester, F. J., 115, 173 Heusner, A. A,, 120,179,184 Hickman, C. P., Jr., 122, 123, 130, 131, 179, 510, 519,538 Hidaka, T., 374, 385, 386, 398, 400, 402, 420 Higashi, H., 362,420 High, W. L., 113, 114, 115,179 Hill, A. K., 429,495 Hill, A. V., 137, 179, 215, 234, 349, 358, 365,420 Hines, J. A., 508, 509,538 Hirosaki, Y.,27, 97 Hirota, T., 193, 215,235 Hirsch, E. F., 469,495 Hjort, J., 112, 179 Hobson, E. S., 84,94 Hochachka, P. W., 130,131,164,168,177, 178,179, 456,495, 503,508, 509, 511, 515,517,518, 519,520,522, 523, 525, 526,530,533,534,535,538,539,540, 542 Hochella, N. J., 524,542 Hocutt, C. H., 128, 129, 133,179 Hoerner, S. F.,270,310

AUTHOR INDEX

551

Hoffert, J. R., 457,495 Holden, M. J., 77.94 Holeton, G. F., 443, 452, 461, 462, 464, 472,492,495,498 Holland, G. A., 76, 77,94 Hollands, H., 123, 129, 147, 153,175 Holmberg, E. K., 326,357,358 Holmgren, S., 456, 484,495,497 Hoogland, R. D., 64,94 Hopkins, E. J.. 294,310 Hora, S. L., 6,94 Horgan, J. D., 453,500 Horn, M., 70,94 Horrall, R. M., 105, 106, 107, 108, 109, 110,179,181,185 Houde, E. D., 123,128,131,138,139,149, 179 Houssay, S. F., 7, 94, 197, 215, 216, 219, 221, 228,234 Houston, A. H., 158,185, 453,500 Howard, T.E., 162,179 Howarth, J. V., 198, 220,233 Howell, A. B., 70, 72,94 Hoyt, J. W., 220,234 Hubbs, C. L., 43,44, 46,94 Hiibner, H., 49,94 Hudson, R. C. L., 57, 94, 217, 234, 387, 389,396,402,403,404,405,407,420, 422,436,495, 507,539 Hughes, G. M., 60,94,282,307,31I, 442, 443, 446,447,452,492,495 Hughes, S. P., 60,97 Hunt, E. P., 163,181 Hunter, J. R., 18,23,86,94,100, 125, 151, 164,179,183, 191,217,234,236,243, 249,283, 284,295,297,310,312, 407, 409,420,422, 507,541 Huntsman, A. G., 107,179 I Ichihara, T.,113, 179 Idler, D. R., 79,94, 168,180 Ikeda, S., 532,540, 541 Imai, S., 515,539 Infante, A. A., 508,539 Irving, L., 473,495 Irwin, W. H., 186 Issekutz, A. C., 524,539 Issekutz, B., Jr., 524,539

Itazawa, T., 457,495 Ito, B. M., 341,357

J Jackim, E., 531,539 Jackson, M., 531,543 Jahn, L. A., 104, 105,181 James, E. C., 301,310 Jansen, J. K. S., 399,417 Jarman, G. M., 51,94, 368,420, 461,496 Jasper, D., 380,420 Javaid, M. Y., 162,180 Jellinek, M., 469,495 Jensen, D., 469, 471, 474, 479,493,496 Jensen, J. K. S., 382, 394, 399, 400,417 Job, S. V., 123,180 Johansen, K., 468,471,475,478,479,480, 481,486,487,488,495,496 Johansson, R. G., 373,377,418, 506,523, 537, 540 Johnels, A., 469, 471,492 Johnels, A. G., 41, 43,94 Johnson, G. D., 31, 70,98 Johnson, I. A., 216, 217,234 Johnson, J. H., lO4,109,1fi0,243,244,310 Johnson, M. W., 287,312 Johnson, R. D. O., 42,94 Johnson, R. N., 535,539 Johnson, W. E., 78, 79, 83,94 Johnston, I. A., 55, 70,94, 128, 136,180, 376,387,389,407,409,410,420,421, 422, 504,507,508,509,511,515,518, 522,530,539 Jonas, R. E. E., 408, 418, 524, 525, 528, 536,539 Jones, D. R., 131, 133,140,180, 387,389, 405,406,407,408,418,434,436,440, 441,443,445,446,447,449,450,453, 454,455,456,457,460,461,462,463, 464,468,469,474,475,476,479,481, 482,483,484,485,486,487,496,501, 504,539 Jones, F. R. H., 81, 84,85, 87,91, 95, 113, 114, 115,180, 253,310 Jones, N. R., 508,509,511,539 Jones, R. A., 105, 107, 109, 110,176, 177, 181 Jorfeldt, L., 524,539 Junge, C. O., Jr., 169,180

AUTHOR INDEX

552

Kliashtorin, L. B., 214, 217,235 Koike, J., 509,542 Komarov, V. T., 104, 116, 119, 121, 137, Kafuku, T., 54, 55, 57,95 180 Kambe, T., 241, 290, 293, 296, 301, 303, Konstantinov, K. G., 112,180 309 Kordylewski, L., 377,421 Kaneko, T., 362,420 Korneliussen, H., 374, 381, 387, 388, 389, Kanji, S. K., 43, 70,99 394,421 Kanwisher, J. W., 327,331,349,357 Kramer, E., 11, 24, 73,95, 291,310 Kashin, S. M., 55, 58,95 Kraning, K. K., 11, 524,542 Katz, M., 122, 125, 126, 128, 156,180 Krishnamoorthy, R. V., 526,540 Kawai, A., 531,542 Krueger, H. M., 162, 163,180, 525,540 Kay, R. M., 522,540 Kruger, P., 376,421 Kay, W. W., 508, 510,542 Kruuk, H., 29, 34, 81,95 Kayne, F. J., 522,539 Kruysse, A., 485,499 Keast, A., 70,95 Kryvi, H., 368, 374, 377, 382, 384, 392, Keemers, H., 452,500 393,421 Keenan, M. J., 307,312,321,358,367,405, Kuchemann, D., 192,211,236 410,422 Keenleyside, M. H. A., 77,86,95,221,234 Kunos, G., 474,496 Kusumi, F., 524,542 Kelly, H. R.,231,234 Kutty, M. N., 123,125,129,131,147,149, Kemp, R. G., 521,522,539,542 156, 157,163, 165,166,167,180,181, Kempf, G., 193,234 Kendall, J. I., 115,185, 282,311, 333, 339, 427,438,440,496,505,515,530,540 359 1 Kennedy, J. W., 524,542 Kent, J. C., 193, 215,235 LaBar, G. W., 109, 125,181 Keppler, D., 507, 524, 525,539 Labat, R., 471, 474, 475,496,497 Kerr, J. E., 147, 149, 170,180 Laffont, J., 474,497 Kerr, S. R., 218,235 Lam, H. M., 115,185, 338, 339,359 Kerswill, C. J., 77,95 Lam, K.-C., 168, 174, 438, 460,492, 507, Keul, J., 507, 524, 525,539 Keys, A., 485,496 519, 524,536 Landis, E. M., 467,497 Khairallah, E. A., 530,541 Kiceniuk, J. W., 133, 180, 387, 389, 405, Lane, F. W., 103, 116, 119, 120, 121,181 406,407, 408,415,418, 434,436,438, Lang, T. G., 219,235 439,443,445,447,449,450,453,454, Langford, T. E., 109,181 455,456,457,461,462,463,464,468, Langille, B. L., 460, 461, 484, 485,496 469,474,476,479, 481,482,483,484, La Noue, K. F., 533, 535,540,543 Lansimaki, T. A., 387,421 486,487,494,496, 524,525,537 Kidokoro, Y., 379, 385, 398,420 Larimore, R. W., 129, 131, 135, 149, 153, Kilarski, W., 374, 382, 386,421 154, 155,181 Larkin, P. A., 168, 170,183 Kingsbury, 0. R., 169,185 Kirby, S., 484,496 La Roche, G., 163,178, 531,539 Kirschner, L. B., 485,496 Lasker, R., 164, 183, 217, 236, 409,422, Kishinouye, K., 21, 54, 95, 113, 115,180, 507,541 Lau, Y. C., 528,536 317,326,358 Lauer, C. Y., 156,183 Klaassen, H. E., 43,96 Laurent, P., 452,497 Klausewitz, W., 29, 35, 41,95 Laurs, R. M., 243, 244,310 Klaverkamp, J. F.,484,496 Lavocat, A,, 372, 386,405,417 Kleerekoper, H. A,, 86,95, 122,180 Klein, W.-D., 145,180 Lawson, K. D., 327,331,345,349,357 K

553

AUTHOR INDEX IeDanois, Y.,52, 53, 54, 58, 59, 60, 69, 71, 95 Lee, A. K., 165,173 Leggett, W. C., 105, 107, 109, 110, 176, 177,181 Lemke, A. E., 125,181 Lett, P. F.T., 125, 163, 174, 181 Levin, A., 362,421 Levine, D. N., 372,418,419 Levy, M. N., 524,540 Lewis, J. K., 533,538 Lie, H. R., 380, 381,421 Lieder, U., 51, 95 Lighthill, M. J., 3, 5, 8, 12, 19, 20, 22, 23, 29, 31, 35, 36, 38, 40, 95, 191, 198, 199,200,202,204,205,206,207,208, 209,210,211,212,213,214,226,228, 232,235, 241, 272,273,287,288, 289, 290, 291, 296, 300, 301,302, 304, 308, 310 Lillelund, K., 122,181 Lim, S. T., 522,540 Lin, Y.,506, 526, 527,540 Lindsey, C. C., 48, 51, 63,95, 349,358 Ling, S. C., 29,95 Ling, T. Y. J., 29, 95 Linthicum, D. S., 115,181, 331, 332,358 Lippross, H., 474,495 Lishajko, F., 469,492 Lissaman, P. B. S., 284,310 Lissmann, H. W., 20,27, 28, 32, 61, 84,96, 191, 231, 232,235, 414,421 Liu, S. D., 162, 180 Logvinovich, G. V., 301,310 Lorenzini, S., 362, 372,421 Lorz, H. W., 84,97 Love, R. M., 503,506, 517,522,540 Lowenstein, J. M., 512,515,516,540,542 Lowenstein, O., 413,421 Lowndes, A. C.,253,254,310 Lowry, R. R., 163,180, 525,540 Lfiyning, Y.,382, 394, 399, 400,417 Liihmann, M., 107,181 Lund, R., 57,96 Lusz, L. D., 113, 114, 115,179 Lutz, B. R., 473,497 M

McBean, R. L., 532,540 McCartney, B. S., 105,179 McCleave, J. D., 105, 107, 109, 181

McCutchen, C. W., 30, 37,63,96 McDonald, D. G., 458,459, 460,501 McGreehan, J. R., 478,497 McInerney, J. E., 128, 133, 137, 139, 158, 159,178 Mackay, K. T., 243, 245, 254, 259, 260, 268,274,311 McKim, J. M., 163,181 MacLennan, D. N., 112,179 MacLeod, J. C., 84,97, 125,128,129,130, 131, 149, 156, 160, 161,181 McMahon, B. R., 458,459, 460,501 McMurrich, J. P., 60, 70, 72, 96 Macnae, W., 41,96 Mair, A., 107,181 Madison, D. M., 105, 107, 108, 109, 110, 181 Maetz, J., 460,486,497,498 Magnan, A,, 7, 29, 47, 73, 96, 103, 111, 147, 149,151,181, 193,235, 261,311 Magnuson, J. J., 88,92, 104, 111, 112, 113, 115,181,182, 191,197,235,241,243, 245, 246,247, 248, 249, 251, 252,253, 254, 255, 256, 257, 259, 260, 261, 262, 263, 264,265,266,267,270, 273, 275, 276, 278,279,280,281,282,290, 292, 293, 294,295, 297,298, 305, 306,307, 308,311, 316,329,342,352,357,358, 371,390,422, 445,497 Mahajan, C. L., 64,96 Malar, T., 86,95, 122,180 Malessa, P., 322,358 Malinin, L. K., 80,92, 107, 108, 109,178, 182,183 Mallov, S., 528,540 Mann, H., 107,181 Manning, G. T., 524,536, 537 Mansour, T. E., 521,540 Manzer, J. I., 76,96 Mar, J., 125, 1277 182 Marchelidon, J., 531,538 Marey, E. J., 7, 29,96, 190, 192,235 Margolis, L., 145,185 Mark, R. F., 362,422 Markarewicz, W., 512, 532,540 Marotte, L. R., 362,422 Marples, B. J., 490,497 Marr, J., 191,235 Marshall, N. B., 12, 62, 67, 68, 69, 70, 85, 96, 267,311, 376,419 Martin, A. D., 109,183

554 Marty, J. I., 107,182 Maslov, N. A,, 107,182 Mason, J. M., 76,96 Mather, F. J., 75, 76, 77, 96 Mathur, G. B., 120,182 Matthews, L. H., 43,96 Mattisson, A. G. M., 523,540 Maurer, F., 58,96 Mayer, P., 490,497 Mazeaud, F., 472,497 Mazeaud, M. M., 472, 486,497 Meffert, D., 122,182 Meierotto, R. R., 84,91 Menzies, W. J. M., 77,96 Messing, S., 478,497 Meyer, A., 76,96 Meyer, D. L., 222,229,233 Midttun, L., 105,182 Millard, R. W., 475,495 Milleman, R. E., 144, 145, 147, 150,176 Millen, J. E., 448,497 Miller, D., 230,236 Miller, G. J,, 453,493 Miller, R. B., 169,182 Milliken, C., 460, 492,498 Milne, D. J., 170,182 Minahan, R. P., 85,91 Minckley, C. O., 43,96 Mitra, B., 40, 44, 92 Mitson, R. B., 81,93, 109,178 Mochek, A. D., 87,97, 139,183 Modigh, M., 374,422 Mohamed, M. P., 505,540 Molnar, G., 128,182 Moody, P., 323,357 Mookherjii, P. S., 117,174 Moon, T. W., 522,535,540 Moore, M. H., 382, 403,418 Morar, L., 409,424 Morgan, H. E., 526,541 Morgan, M., 443, 452,495 Morrow, J. E., 327,358 Morton, M. J., 474,495 Mosse, P. R. L., 370, 387,422 Mott, J. C., 473, 487,497 Mount, D. I., 125,181 Moury, N. R., Jr., 524,542 Muir, B. S., 122, 182, 193,233, 259, 267, 268,270,271,272,273,274,280,282, 307,309,311, 342,357, 433,441,445, 447,448,493,497

AUTHOR INDEX

Muirhead, K. M., 516,540 Miiller, J., 317, 331,357 Munger, B. L., 413,422 Munk, M. M., 259,311 Murdaugh, H. V., 448,497 Mure, M., 531,537 Murray, J., 508, 509, 511,539 Mustafa, T., 522, 540 Myers, G . S., 44,96, 388,419 Myhrherg, H., 472, 477,478,494

N Nag, A. C., 70,92, 374, 386, 390, 391, 392, 393,422 Nagai, M., 532,540, 541 Nakano, T., 472, 486,497 Nakao, T., 374, 394,422 Nakatani, R. E., 168,182 Nakatano, T.,520, 523,541 Nakken, O., 105,177,182 Nakumura, I., 326, 327, 329,358 Narasimhan, T., 526,540 Nawar, G., 41, 70,96 Neave, F., 76, 77, 79, 83,97 Neely, J. R., 515, 525, 526, 528, 529, 530, 533, 535,539,541 Neiheisel, T.W., 121,185 Neill, W. H., 243, 246, 248, 249,310,311, 319, 320, 322, 323, 324, 336, 337, 338, 341, 342, 344, 345,346, 347, 352,353, 354,356,357,358 Nelson, G. J.. 122,182 Neppel, M. J., 532,540 Neu, W., 193,234 Newcornb, T. W., 124, 125, 126,185 Newcornbe, C. P., 259,311 Newman, J . N., 200, 205, 208, 211, 235, 237 Newsholme, E. A., 506,520,523,535,537, 541 Newstead, J. D., 452,497 Nickerson, M., 474,496 Nicklas, W. J., 535, 540 Nicolaysen, K., 374, 381, 382, 385, 400, 419,421,422 Nigrelli, R. F., 85, 90 Niirni, A. J., 139, 163,174, 182, 433, 441, 497 Nikolsky, G. V., 7,97

555

AUTHOR INDEX Nilsson, S., 456, 477, 484,495, 497 Nishi, S., 52, 58, 60, 66, 68, 72,97 Nishihara, H., 374,386,390,394,397,410, 422 Nishimura, M., 104, 113, 115,182 Nishimura, S., 27,97 Noda, L., 511, 541 Norman, J. R., 7, 24, 43, 49, 97 Norris, H. W., 60,97 Northcote, T. G . , 84, 87,97 Norum, K. R., 529,541 Novotny, A. J., 104,182 Nursall, J. R., 50, 51, 54, 5 5 5 6 , 57,67,68, 72, 97, 191, 221, 235, 291, 292, 295, 311, 363, 390, 391, 392, 393,422

0 Odessey, R., 530,531, 532,538, 541 Oehmichen, E., 24,97 Ostlund, E., 469, 471, 485,492,498 Ohlmer, W., 8 5 9 7 , 104,183 O h , B. L., 83,97, 109,183 Olsen, 0.W., 145,180 Opdyke, D. F., 478,497 Opdyke, N. E., 478,497 Oram, J . F., 528, 529, 533, 535,541 Orange, C. J., 75,97 Osborne, M. F. M., 80,97, 193, 196, 215, 235 Oseid, D., 125, 128, 131, 149, 161,183 OSiidal, B., 454,498, 507,536 Otto, R. G., 128, 133,183 Ovalle, W. K., 372, 376, 382,423 Ovcharov, 0. P., 40,43,89, 197, 198,215, 219,232, 289,309 Ovchinnikov, V. V., 198, 219, 220,235 P

Packard, A., 40,97 Pappenheimer, J. R., 467,497 Parker, H. W., 43,96 Parker, R. R., 128, 151, 168, 170,174,183 Parks, R. E., Jr., 512,541 Parrish, B. B., 112, 115, 171,175, 183 Parry, D. A,, 192, 193,235 Parry, G., 158, 183

Patel, D., 297,312 Pathak, C. L., 474,498 Patterson, A. L., 524,542 Patterson, S., 55, 97, 374, 377, 386, 387, 422 Paulik, G . J., 119, 170, 171, 183 Paulsen, J. E., 387, 388, 389,421 Pavlov, D. S.,Y37,97, 139,183 Pecot-Dechavassine, M., 394,422 Pedley, T. J., 192,235 Peer Mohamed, M., 165,181 Pelouch, V., 507,536 Pequin, L., 515,541 Peres, G . , 531,537 Perkins, C. D., 281,311 Pershin, C. V., 193,235 Persson, H., 472, 477, 478,494 Peterson, R. H., 119, 125, 147, 149, 161, 162,183 Pette, D., 507,541 Pettigrew, J. B., 6, 7, 24,97, 192,235 Peyraud, C., 478,492 Pfautsch, M., 452,500 Philpott, C. W., 517,541 Phinney, L. A., 169,180 Piiper, J., 461,498 Pinc, R. D., 156,183 Pitcher, T. J., 86,97 Plaut, G. W. E., 535,541 Pocinwong, S., 521.541 Poddubnyi, A. G., 80, 92, 105, 107, 109 178,182, 183 Pope, A., 127,183 Porter, K. R., 386,419 Potts, W. T. W., 158, 183 Poupa, O., 455,494, 507, 523,538 Pradhan, T. K., 511,541 Prandtl, L., 193, 205, 235, 257, 269, 272, 273,311 Prescott, J. H., 104, 182, 191, 235, 243, 245,248,249,275,282,297,305,311 Priede, I. G., 468, 469, 470, 471,498 Pritchard, A. L., 107,183 Pritchard, A. W., 122, 125, 126, 128, 156, 164, 165, 167,179,180,183, 217,236, 409,422, 507, 524,538,541 Prosser, C. L., 156,183, 352,358 Pull, G. A., 130, 131, 133, 178, 370, 371, 372, 374, 376, 407,419, 507,543 Pyatetskiy, V. Ye., 191, 193,236, 243, 245, 249,268,274,295, 297,311

556

AUTHOR INDEX

R

Root, R. W., 158,184 Rosen, D. E., 388,419 Radakov, D. V., 85,97, 106, 113, 114, 115, Rosen, M. W., 197,215,220,236, 280,289, 312 183,184,187 Rahn, H., 461,498 Rosenblatt, R. H., 31, 42, 70,92, 98 Randall, D. J.. 164,184,443,449,450,451, Rosenthal, H., 16,98 452,453, 456,457,458,460, 461,462, Rothchild, B. J., 353, 354,358 464, 465,466, 468,471,472,473,474, Rovainen, C. M., 373,398,403,423 475,476,478,479,481,483,484,485, Rovetto, M. J., 533, 535,541 486,487,494,495,496,498,499,500, Rowell, L. B., 480,498, 524,542 501 Royce, W. F., 106,184 Randall, R. F., 522,541 Rudjord, T., 399,417 Randle, P. J., 515, 533, 535,541, 542 Ruhland, M. L., 120,184 Rankin, J. C., 452, 485, 486,493,498 Russell, I. J., 61, 98 Ranvier, L., 372, 386,422 Ryan, K. P., 382,403,412,415,418 Rao, G . M. M., 125, 128, 131, 147, 160, Ryland, J. S., 121, 123, 139, 143,184 167,184, 427,428,433,441,498, 504, 542 s Rascovich, M., 104,173 Saddler, J. B., 163,180, 525,540 Rasmussen, B., 107,184 Saetersdal, G., 112,184 Raynaud, P., 474,475,497 Rayner, M. D., 307, 312, 321, 358, 367, Safer, B., 534,542 Sakaguchi, M., 531,542 405, 410,422 Samuel, C. T., 42, 70,98 Regan, C. T., 326,358 Sand, A., 411, 413,419 Regnard, M. P., 119,184 Sandon, H.,’65,98 Reichard, G . A., Jr,, 524,542 Reimers, N., 129, 169,184 Santer, R. M., 454, 478,493,499 Satchel], G. H., 40,98, 445,454,473,477, Reite, 0. B., 485,498 484,487,488,489,490,491,492,493, Rice, J . O’H., 128, 133,183 494,499 Richards, B. D., 485,498 Richardson, E. G . , 193,219,236, 268,312 Saunders, R. L., 129, 131, 147, 156, 157, Richardson, I. D., 112,184 176,181,184, 443,446,447,448,449, Ricker, W. E., 142,184 466,493,495,499,500, 517,543 Saville, W. A., 515,537 Ridge, R. M. A. P., 411,412,418,422 Riley, A. L., 515,539 Saxena, S. C., 72,98 Ripley, W. E., 76, 78,93 Sbikin, Y. N., 87,97, 139,183 Ripplinger, J., 472,493 S c h ~ f eJ., , 113, 114, 115,184 Riss, W., 61,97 Schiebler, T. H., 454,498,500 Ritzen, M., 469,492 Schlichting, H., 205, 219,236 Robb, C., 112,179 Schmalhausen, J. J., 63, 68,98 Roberts, B. L., 15,24,25,60,61,63,65,72, Schmidt-Nielsen, K., 5,98, 164, 167, 168, 98, 255,309, 362, 374, 376, 404, 413, 184, 436,499 414,418,423, 448,498 Schmitt, A., 532,542 Roberts, J. L., 153,184,242,267,272,312, Schneider, H., 30,31,36,38,64,70,72,98 Schuett, F., 85, 86,98 443,444,445,446,448,449,468,492, Schulz, G. E., 511,541 498 Robertson, A. C., 128, 151,174, 438,492 Schuyf, A., 122,176,184 Robertson Connor, A., 507, 519, 524,536 Schwartzkopff, J., 85,97, 104,183 Robin, E. D., 448,497 Schwarzfeld, T., 445, 446,447,496 Robinson, G . A., 128,177 Scopes, R. K., 510,538 Rommel, S. A., Jr., 105, 114,185 Secondat, M., 453,459,460,492,499

AUTHOR INDEX Seireg, A,, 297,312 Serfaty, A., 474, 475,497, 515, 530, 537, 541 Sewertzoff, A. N., 70, 71, 72,98 Shann, E. W., 52, 70,98 Sharp, G. D., 274, 308,312,423 Shaw, E., 85,98, 114,184 Shaw, R. J., 85,91 Shaw, W. A. S., 524,539 Shazkina, E. P., 125, 151, 184 Shearer, W. M., 77,96 Shebalov, A. M., 287,312 Sheldon, F. F., 71, 72.99 Shelton, G., 60,94,99, 360,443,461,468, 471,473,478,483,484,485,486,498, 499,500,501 Shepard, M. P., 78,99 Shollenberger, C. A., 284,310 Shrivastava, B. D., 120,182 Shubina, T. N., 80,99 Shuck, H. A., 169,185 Shumway, D. L., 128, 147, 149, 156, 157, 158, 164,176,177, 440,493 Siddiqui, A. H., 531,542 Siddiqui, A. Q., 531,542 Siebert, G., 532, 542 Siekmann, J., 231,236 Silver, S. J., 440,499 Simons; J. R., 66,99 Skoglund, C. R., 362,423 Slavin, J. W., 323,358 Slijper, E. J., 55,90, 372, 389, 405,418 Smit, H., 122, 123, 125, 127, 128, 135,136, 140, 149, 165, 167,185, 191, 193, 197, 213,214,215,217,236, 409,423, 427, 430,436,437,438,440,499, 504,507, 542 Smith, C. M., 533,543 Smith, D. G., 451,452,461,494,499 Smith, E. H., 24,99 Smith, H. D., 145,185 Smith, J. C., 472,498 Smith, L. L., Jr.. 125, 128, 129, 131, 149, 156, 160, 161,181,183 Smith, L. S., 106, 124, 125, 126,184, 185, 456,468,482,499 Smith, R. S., 372, 376, 382,423 Smolyaninov, V. V., 55, 58.95 Solandt, D. T., 473,495 Solandt, 0. M., 473,495 Solovyev, B. S., 106,184

557 Soma, M., 113,179 Somero, G. N., 522,542 S@mme,S., 76,91, 107,176 Sommer, J. R., 379,423 Spencer, W. P., 84,99 Spitzer, J. J., 525,542 Spoor, W. A., 117, 121,185 Sprague, J. B., 163,185 Spray, T. L., 379,423 Squire, J. L., 78,99 Srere, P. A., 533,542 Stacy, E. E., 171,183 Stanfield, P. R., 384, 385, 394, 398, 399, 400,401,402,423 Stark, E. C., 326,358 Starling,. E. H., 479,499 Start, C., 535,541 Stasko, A. B., 105, 107,108, 109, 110, 114, 181, 185 Stasko, D., 109,185 Staudte, H. W., 507,541 Steen, J. B., 465, 485,499 Steffel, S.. 352,358 Stevens, E. D., 115, 165, 172, 185, 246, 247, 271,282,312, 318, 319,320, 322, 324, 332, 333, 334, 335,336, 337, 338, 339, 342, 344, 345,346,350,357,358, 359, 438,439,443, 449,453,'454,455, 456, 457,458, 461, 462, 464, 468, 471, 472, 473,474,475,476, 478, 481,483, 486,487,498,499,500, 504, 506, 517, 518, 519,542 Stevens, G. A., 198,236 Stevens, J., 362,423 Stickney, A. P., 84,99 Stickney, R. R., 230,236 Stone, D. E., 24,99 Storer, J. H., 281,312 Storeton-West, T., 81,93, 109,178 Storey, K. B., 511, 518,520, 523,539,542 Storm-Mathisen, J., 381,419 Strasburg, D. W., 286,312 Stray-Pedersen, S., 465,499 Stringham, E., 103, 119,185 Stroud, G. D., 519,537 Studholme, A. L., 83,97 Studier, E.H., 263,310 Sudak, F. N., 487,500 Sugii, K., 362,420 Sugiyama, M., 531,542 Sugiyama, T., 531,542

AUTHOR INDEX Sundnes, G., 112,185, 193,236 Sutherland, D. B., 31, 90, 133, 166, 167, 175, 433,436,493 Sutterlin, A. M., 157, 184, 443, 444, 448, 461,468,469,474,475,488,489,500 Suyama, M., 509,542 Suzuki, C., 113,179 Suzuki, K., 509,542 Svensson, G. S. O., 43,94 Sverdrup, H. U., 287,312 Swain, A., 77,99 Swanson, P. L., 43,99 Syrovy, I., 405,423 Szarski, H., 56, 57,99 T

Tota, B., 374, 387,420,421, 422 Totland, G. K., 368, 374,421 Trefethen, P. S., 104,186 Trevor-Smith, E., 438,493 Trout, G. C., 107,186 Trueman, E. R., 43,99 Tsai, M. Y.,521,542 Tsairis, P., 372,418, 419 Tsuyuki, H., 168,180, 522,543 Tucker, V. A., 5, 99, 165, 173, 186, 217, 236, 436,500 Tyler, J . C., 72,99 Tytler, P., 167,186, 433,500

U

Uchiyama, J . H., 82, 88,100, 109,187 Tabb, D. C., 76,96 Uda, M., 317,359 Takahashi, K., 398,400,402,420 Ultsch, G. R., 442,500 Takahashi, N., 58,99, 327,359 Urquhart, G. G., 112,179 Takeuchi, A., 402,403,423 Talbot, F. H., 84,91 V Thing, A. V., 77,99 Tarr, H. L. A., 503,511,517,531,533,542 Taylor, A. A., 525,536 Van Citters, R. L., 468,471,475,478, 480, Taylor, G., 8, 46,99, 192, 199, 212,236 496 Taylor, W., 453,500 Van Dam, L., 448,500 Teal, J. M., 317, 321, 326, 327, 330, 331, Van den Berg, P. G., 392,423 346,349,357, 373,419 Van der Stelt, A., 55, 90, 364, 372, 389, Teravainen, H., 373, 379, 380, 385, 388, 405,418,423 394,398, 399, 400,403,405,423 Van Olst, J . C., 86, 100, 283, 284,312 Terhune, L. D. B., 125,174 Verheijen, F. J., 81,100 Tesch, F.-W., 81,99, 105, 109, 110,185 Verriest, G., 515,537 Tester, A. L., 87,99 Verwey, J., 112,186 Thomas, A. E., 124, 125, 126, 128, 129, Vialleton, L., 390,423 135, 141, 169,185 Vibert, R., 169,186 Thomas, A. M., 60,99 Videler, J. J., 64,92 Thompson, D. W., 137,185 Vijverberg, J., 125, 127, 128, 135, 136,140, Thomson, K. S., 45, 46, 66,93, 99 149, 165, 167,185, 191, 193, 197,213, Thorson, T. B., 109,185, 456,500 214,215,217,236,409,423, 504,507, Threadgold, L. T., 158,185 542 Tietjens, 0. G., 193, 205, 235, 257, 269, Villemonte, J., 84,93 272,273,311 Vincent, R. E., 143, 144, 169,186 Tilak, R., 43, 70,99 Vitek, V., 507, 536 Timms, M., 86,95, 122,180 Vlymen, W. J., 212,236 Tinsley, I. J., 163,180, 525,540 Voboril, Z., 454,500 Tolg, I., 128, 182 Vogel, V., 452, 500 Toida, N., 374,385,386,398,400,402,420 Vogel, W., 452,500 Tomlinson, N., 472, 486, 497, 508, 510, Volf, M., 122, 125, 126,175, 191,233 520,523,541,542 von Euler, U. S., 469,492 Tornheim, K., 516,542 von Holst, E., 192, 211, 236, 445, 500

AUTHOR INDEX

559

von Khrmin, T., 257, 294,312 359, 368,369,389,410,423, 427,429, von Kausch, H., 121,186 430, 433,434, 435,436,438,439, 441, von Mecklenburg, C., 472,477, 478,494 443,445,446,447,466,467,500, 504, von Mises, R., 198,236,257,277,281,312 506,543 von Vaupel-Klein, J. C., 125,127,128,135, Weber, M., 117,176 136, 140, 149, 165, 167,185, 191, 193, Weihs, D., 83, 86, 88,100, 207, 218, 221, 197,213,214,215,217,236,409,423, 222,223,224,225,226,228,237, 268, 427,430,438,440,499,504,507,542 282,283,284,285,286,287,288,312, Von Zabern, I., 511,541 435, 436,501 Vuverberg, J., 427, 430, 438, 440,499 Weinhouse, S., 524,542 Weininger, D., 270, 298,311 W Weinstein, H. M., 508, 509,538 Weis-Fogh, T., 212,237 Waage-Johannessen, N., 481,496 Weitzman, S. H., 100, 388,419 Wagner, H. H., 456,493 Wendt, C A. G . , 517,543 Wales, J. H., 103, 119, 171,186 Wenger, J. I., 528,541 West, N. H., 475, 492,501 Walker, M. G., 507,543 Walters, V., 10, 15, 21, 22, 23, 50,91, 100, Westlake, G. F., 86,95, 122, 180 103, 104, 116, 121,177,186, 191, 197, White, D. B., 230,236 198,215, 218,219,224,226,233,236, Whiting, H. P., 61,93, 391,420 241,243,247,248,249,267,278,279, Whitworth, W. R., 186 280, 281,282, 289,291,292, 295,296, Wick, A. N., 524,537 297,298,299,300,301,302,303,304, Wickler, W., 43,100 310,312, 327,349,357,359, 362, 367, Wiehs, D., 111, 114, 117,186 419 Wildenthal, K., 479,493 Ward, G. R., 515,537 Willemse, J. J., 56, 57, loo, 368, 392,423 Ward, P. S., 389,421 Williamson, J. R., 533,534,535,540,542, Wardle, C. S., 116,186,349,359, 376,398, 543 Wilson, D. E., 263,310 405,418,423 Warfel, H. E., 362,423 Winberg, G. G., 163,186 Warren, C. E., 122, 125, 126, 128, 147,156, Winn, H. W., 121,175 163,176,180,186, 440,494,499 Winterbottom, R., 58,100, 362, 368,424 Watanabe, N., 103, 111, 113, 114,186,248, Wisby, W. J., 107,179 249,312 Wittenberg, J. B., 376,424 Waterman, R. E., 390, 391, 392,423 Wittenberger, C., 409,424, 506, 514, 519, Watters, K., 449,494 524, 525, 534,543 Watts, D. C., 511, 512,543 Wohlschlag, D. E., 123,153,154, 167,186, Watts, R. L., 512,543 187 Waugh, R. A., 379,423 Wood, C. M., 456,458,459,460,465,466, Weaver, C. R., 119, 170, 171,186 483,484,485,486,501 Webb, D., 70,95 Wood, J. D., 528, 531,543 Webb, P. W., 8, 10, 12, 15, 16, 18, 19, 22, Woodhead, A. D., 83,100 23, 24, 25, 30, 31, 38, 40, 73, 79,100, Woodhead, P. M. J., 81, 83, 85, 87, 100, 119, 121, 128, 133, 161, 162, 166,186, 112,187 191, 192, 193, 194, 196, 197, 198, 199, Wu, T. Y., 191, 192, 199, 200, 205, 206, 203,204,209,211,212,213,214,215, 207,208,209,211,212,216,219,226, 216,217,218,219,221,222, 224,225, 231,232,235,237, 241,273,288,290, 226,227,228, 229,230,231,232,236, 291,300,301,312,313 237, 241, 242,267,268,269,273,274, Wunderer, H., 411, 413,424 277,278,280,282,288,291,295,296, Wyman, J., 362,421 298,300, 301, 303,304,308,312, 342, Wyman, L. C., 473,497

560

AUTHOR INDEX

Y

Yamamoto, M., 520,543 Yamamoto, T., 398,424 Yonemori, T., 78,99 Yoshida, K., 113,179 Young, M. K., 530, 532,536 Yuen, H. S. H., 78, 82, 84, 88, 100, 104, 105, 107, 109, 110, 111, 113, 114,187, 191,237, 243, 244, 245, 247, 248, 249,

250, 282, 286, 297, 310, 313, 321, 359

Z Zaitsev, V. P., 106,187 Zajac, F. E., 372,418,419 Zuyev, G. V., 85, 86, 89, 100, 114, 174, 221,233,237, 282,284,309,313 Zweifel, J. R., 18, 23, 94, 125, 179, 191, 234, 243, 249,295,297,310

SYSTEMATIC INDEX Note: Names listed are those used by the authors of the various chapters. No attempt has been made to provide the current nomenclature where taxonomic changes have occurred. A

A . anguilla, 9-13, 17, 77, 108, 110, 373, 452, 465, 523 Abramis brama, 19, 108, 526 A. japonica, 457 Acanthias uulgaris, 206 A . rostrata, 114, 121 Acanthocybium, 14, 242, 250,276277, A. uulgaris, 118 Anguillidae, 14, 388 -305-307 A. solanderi, 111, 112, 120, 224, 241, Antennariidae, 40 245, 249,254, 292, 349 Antennarius, 42 Acanthopterygians, 388, 395 Anura, 372, 376 Acanthuridae, 31 Aphanopus carho, 25 Acerina, 51 Arges marmoratus, 42 Acipenser, 63, 72, 108 Aspius, 51 A. fuluescens, 522 Aspredinidae, 40 A . gueldenstoedti, 80, 107-108, 110 Assurger, 25 A. nudioentris, 107-108, 110 Astroblepus mamioratus, 42 A. stellatus, 80 Atherina, 367 Atherinoinorpha, 388 Acrania, 374, 377, 380, 417 Agnatha, 363, 374, 377, 380-382, 398-400 Aulostomidae, 29, 34 Albacore, see Thunnus alalunga Auxis, 242, 260, 276277, 279, 290, 328, Albula oulpes, 118 330, 3.30 Alepocephalids, 369, 375, 388, 395 A . rochei, 111, 245, 254, 292-293 Alewife, see Pomolobus pseudoharengus A. thazard, 254, 292, 329 Allothunnus, 290 A. falai, 292 6 Alosa, 367 Barracuda, 220, see also Sphyraenu A.finita, 146 A. sapidissimu, 105, 107-108, 110, 118 ba rrucu do Bass Ameiurus, 387, 394-395, 402 A . nebulosus, 462, 474 largemouth, see Micropterus saltnoides sinallmouth, see Micropterus dolomieu Amia, 27, 32, 54, 2.31 A. caluu, 9, 26 white, see Morone chrysops Aiiiniodytidae, 15, 43 Batfish, 40 Ammodytes, 59, 216 Batlaclioidiformes, 388 Amphioxus, 363, 379, 385, 398, see n l s o Benthodesma, 25 Branchiostoma Beryciformes, 388 Anabantidae, 41 Billfish, 241, 355 Anabas testudineus, 41 Blennioidei, 15 Andamiu, 49 Blenny, 10, 49 Angler, see Lophius Bluefish, see Pomatomtis Anguilla, 54-55, 208-209, 216, 290, 392 Bluegill, 160, see also Lepomis 561

562

SYSTEMATIC INDEX

Cetorhinus, 43, 374 Chaenocephalus aceratas, 475 Chaetodontidae, 29, 31, 34 Chulcalburnus chaleoides montoides, 118 Channidae, 41 Characidae, 15, 45, 50 Characinidae, 20, 31 Chela C. caeruleostigmata, 45 C. niuassi, 4 4 4 5 Chilomycterus schoepji, 40 Chimaera, 36, 62 C. monstrosa, 371 Chimeridae, 14 Chlumydoselache, 14 C Chondrichthyes, 49 Capros uper, 371 Chondrostei, 63, 65 Carangidae, 14, 20, 22, 197, 241, 355,269, Churiductylus multihurhis, 42 Cichlasoma bimaculatum, 162 409 Carunx, 15, 20 Cichlidae, 28, 34 C . crysos, 445, 468 Clarias, 40-41 C. lazera, 41 C . hippos, 9 C. senegdensis, 41 Carapus acus, 43 Clingfish, see Lepadichthys lineatus Carassius, 213, 215-216, 385-386 C. auratus, 19, 118, 139, 148, 15,3-157, Clinid, see Mnierpes macrocephulus Clupea, 15, 20, 375 167, 214, 217, 362,386, 410, 4 3 6 C. harengus, 9, 13, 16,43, 85, 104, 114438,440, 505, 507, 521 115, 118, 138-141, 146, 171, 241, C. auratus gihelio, 118 288, 374, 387, 389, 392, 395, 404, C . carassius, 165, 387, 407, 507-508, 522, see also Carp 406, 409 Carcharhinus leucas, 241 C. horengus harengus, 113, 141 Carcharodon, 14, 366 C. harengus pallasi, 113 C. sprattus, 117-118 C. carcharias, 241 Carp, 117, 387, 392, 408, 410, 453, 457, Clupeidae, 15, 20,83, 101, 288, 383, 389, 395 459,472,474,507-508,511-512,519, Clupeomorpha, 388 522, 526, 534, see also Carassius Crucian, see Carassius carassius Coalfish, see Gadus virens Cobitidae, 43 Japanese, see Cyprinus carpio Cod, 18, 47, 73, 7 6 7 7 , 85, 455, 468, 477, Swedish, see Cyprinus curpio 479, 484, 508, see also Gadus Catfish, 17, 40, 42, 60, 64-65, 71-72, 386. Atlantic, see Gadus morhua 388, 462, 474-475, see also lctalurus ling, see Ophiodon elongatus Catostomus, 468 Coelacanth, 50 C. catostomus, 132 Conger, 402 C. commersoni, 132 Coregonus, 167 C. occidentalis, 118, 171 C. autumnalis, 132 Cenimrchidae, 31, 129 C. clupeaformis, 132 Centriscidae, 29, 34 C. nasus, 132 Centroscymnus owstoni, 362 C. sardinella, 132 Cephalopods, 40 Coryphaenidae, 15, 241,326, 355, 522 Cepolidae, 14 Cottidae, 228-229 Cetacea, 196197, 200 Bonito, see Sorda chiliensis Bowfin, see Amiu calva Boxfish, 11, 24, see also Ostracion Brachydanio rerio, 390-391 Bramidae, 15 Branchiostegidae, 29 Brunchiostomu, 52, 377, see also Aniphioxus B. lunceolatuni, 50 Bream, see Ahramis hrumu Brotulidae, 59 Bullhead, 469, 475, see also Ictalurus Butterfly fish, see Pantodon huchholzi

563

SYSTEMATIC INDEX

Cottus, 403 C . cognatus, 228 Creilabrus tinca, 116, 120 Crestfish, see Lophotus Ctenothrissiformes, 388 Cucumberfish, see Carapus acus Cyclopteridae, 31 Cyclostoines, 47, 60, 62 Cymatogaster, 31, 39, 368 C. aggreguta, 9-10, 232, 445 Cynodontidae, 45 Cyprinidae, 14, 42, 4445, 50, 65 Cyprinodontidae, 40, 43, 383, 525 Cyprinus carpio, 116, 118, 148, 166,224, 453,457,459,472,474,508,511-512, 519, 524, 526, 532, 534 Cyselurus, 44 D

Dace, see Leuciscus leuciscus Dactylopteridae, 46 Damselfish, 39 Delphinus, 216 Desmodema, 218 Diodon holocanthus, 9 Diodontidae, 15, 24, 30, 36 Diplospinus, 25 Dipnoi, 383, 387, 395, 415, 417 Dogfish, 60, 72, 74, 391, 394395, 404, 406, 486, see also Acanthias; Scylliorhinus; S cylliu m; S quulus Dolphin, 8, 15, 23, 123, 198, 241, 289, see also Lagenorhynchus

E Eel, 2, 5, 7, 10, 12, 17, 40, 43, 47, 49, 59, 77, 81, 109, 114, 122, 191, 377, 395, 452, 457,465, 523 Conger, 7 Cusk, see Otophidium scrippsi Eigenmannia, 28 Elasmobranchs, 47, 49, 52, 5455, 60-63, 65-66, 68, 71, 74, 363, 373-374, 382, 386-387, 394-395, 400-402,411, 415-416,468,471,474-475,477,480, 488,490-491 Electric fish, 28 Electrophorus, 28, 33 Elopiformes, 388

Elopomorpha, 388 Embiotocidae, 31 Engraulis E. encrasicholus, 113 E. encrasicholus pontious, 118 Epinephalus, 37 Eptatretus stoutii, 469, 479 Esocidae, 15, 30, 228-229 Esomus, 45 Esox, 118, 224, 227, 375 E. iucius, 108, 118, 132, 148, 224, 472 Etmopterus, 382, 392 Etheostomu caeruleum, 224 Euleptorhampus longirostrus, 44 Eupleurogrammus, 27 Euthynnus, 14, 215, 242, 246,250, 260, 274,276277, 279, 286, 290, 293, 305-307, 328, 330, 333, 367,375 E. afinis, 10,21-23, 111, 112-113, 116, 242-243, 245, 247, 249,254255, 258-259, 261-263,266,272-275, 278-281, 284, 289,292-293,295, 297-299, 303, 305, 320,329, 332 E. alletteratus, 262, 292, 329, 332 E . lineatus, 292, 329, 333 E . pelamis, 103, 111, 113, 118, 245,249, 254,292-293 Exocoetidae, 45 F

Featherback, 33 Filefish, see Monacanthus Fistulariidae, 29 Flatfish, 16-17, 34,43, 77, 81-85,see also Solea; Microstomus; Pleuronectes platessa Flounder, see Platichthys; Pleuronectes; Pseudopleuronectes Flying fish, 4, 44, 45 Frogfish, see Histrio Fu ndulus F. heteroclitus, 517, 531 F. notti. 40 G

Gadiformes, 14, 18,206,383,388-389,409 Gudus, 215-216 G. aeglefina, 118 G. luscus, 148

SYSTEMATIC INDEX

564

Hemitripterus americantcs, 150, 461 G. merlangus, 10, 24, 118, 148 G . morhuu, 9,76-77,112,118,130,138, Herring, see Clupea harengus Heterodontus portusjacksoni, 490-491 140, 148, 156, 387, 455, 468, 477 Hexanchus, 366 479,484, 508-509 Hiodon, 388 G. pollachius, 371 H . alosoides, 132 G. virens, 118, 130, 136, 139, 371, 374, Hippocampus, 27, 32-33, 48, 65, 405 386, 407, 507 Hippoglossus stenolepis, 76 Galeocerdo G. arcticus, 362 Histrio, 42 Holocephali, 36, 362, 36S-369, 387, 414 G. cuvieri, 264 Holostei, 387, 395 Galeorhinus zyopterus, 76, 78 Homalopteridae, 43 Galeus, 382, 384 G. melastomus, 393 Hoplopagrus, 68, 72 H . guntheri, 68 Gasteropelecidae, 4, 45 Hyperopisus, 28 Gasteropelecus, 45, 368 Hypophthalmichthys molotrix, 118 Gasterosteidae, 30-31 Casterosteus, 38 G. aculeatiis, 72 I G. spinuchia, 120 Geinpylus serpens, 375 Ichthyosaur, 15, 23, 289 Gobiesociformes, 388 Ictcilurus Gobiidae, 30, 40-41 1. melas, 84 Gohius, 42 I . nebulosus, 468 G. fluoiatilus, 150 I . punctatus, 132 G. melanostoma, 150 Isichthyes, 28 G. minutus, 120 Istiophoridae, 15, 22, 218, 220, 241, 281, G. syrmnn, 150 355 Goblinfish, see Chwidactylus multibarhis Istiophorus, 120 Goby, see Gohius Isuridae, 367, 410 Goldfish, see Carassius nurutus lsurus Gonorynchiformes, 388 1. glaucus, 9 Gurnard, 46, 414 I . oxyrinchus, 241 Gymnarchids, 33 Gymnurchus, 27, 32-33, 47, 231 J G. niloticus, 32 Gymnoceplaalus cernua, 120 Jack, 241 Gymnotidae, 33, 73, 84, 415 John Dory, 33 Gymnotus, 28 K G. curapo, 9, 33 H

Haddock, see Melanogrummus aeglejinus Hagfish, 50, 52, 54, 381, 394, 398-400, 414, 469, 474, 479 Halfbeak, 43 Halibut Greenland, 17 Pacific, 76 Hatchetfish, 45 Hemimmphus, 44

Katsuwonus, 328, 330, 333, 367 K. pelamis, 78, 82, 103, 110-111, 112, 114-117, 226,245-247, 249-250, 254, 26.3-267, 270-271, 27.3-275, 279, 282, 286287,292-295, 297 305, 317319,325, 329,331-334, 337-338, 342-344, 346.347, 350351, 353-356, 362, 445, 447, 454 Kawakawa, see Euthynnus afinis Killifish, 43, see also Fundulus Kuhlia sandvicensis, 432, 441

SYSTEMATIC INDEX

565

Macrorhamphosidae, 34 Macrozoarces umericanus, 148 Laheo rohita, 528 Makaira nigricans, 108 Luhidesthes, 43 Mantc1, 35, 43 Labridae, 3,38,43,67,362,369 M a r l i n , 82, 281 Lagenorhynchus, 215-216 striped, see Tetrapturus uudax L. ohliquidens, 289 white, see Tetrapturus albidus Lagocephnlus laevigutus, 9, 38 Mastacembelidae, 29 Lugodon rhomhoides, 167 Melanogrammus aeglefinus, 112, 117, Lumna, 14, 23 166167, 170,433 Lamnidae, 15, 23, 241 Merluccius vulgaris, 148 Lampetra, 385 Micropterus, 120 L. juuiatilis, 381, 477 M . dolomieu, 129, 153-155 Lamprey, 10, 52, 54, 374, 380, 388, 394, M . salmoides, 130, 132, 134, 139, ,398-399, 404, 471, 474, 477, see u ~ s o 148,152154, 156, 158, 166-167 Petromyzon; Lumpetra Minnows, see Phoxinus laeuis Lamprididae, 31, 355 Mnierpes macrocephalus, 41 Latimeriu, 50 Mobula, 36 Lepadichthys lineatus, 50 M . diabolis, 35 Lepidopu.s, 25 Mola, 31, 4 6 4 7 , 232 Lepisosteus, 49 Molidae, 67, 355 Lepomis Mollienesia, 386 L. cyanellus, 120, 224 Moiiacanthus, 28, 3,3-34 L. gfhhosus, 132, 166167,433,468-469 Mormyridae, 15, 20, 28, 32-33, 388 L. macrochirus, 148, 224 Morniyrus, 27, 32 Leuciscus, 215 Morone, 37 L. leuciscus, 19 118, 139, 148, 209, 224, M . chrysops, 105, 108, 110 362 M . saxatilis, 148 Liza mqrolepis, 1 6 6 167 Mudminnow, see Umbra Loach, 43 Mudskipper, see Periophthalmus Lophius, 42, 379, 38.3-384, 387, 389, 393 Mugil, 15, 4.3-44 Lophotidae, 15, 42, 388 M . auratus, 116 Lophotus, 25 M . capito, 148 Loricariidae, 15, 42 M. cephalus, 445, 468 Lota lota, 108, 132 M . corsula, 40, 43 Lungfish, see Neoceratodus; Protopterus M . curatus, 120 Luvaridae, 15 M . saliens, 116, 120 Mullet, 415, 505, see also Mugil; Mullus M Mullus, 416 M . harbatus pontious, 120 Mackerel, 11, 111,164,240,327-328,374, Muraena, 29 445,448, 507, 531 Myctophoids, 50, 388 Atlantic, see Scomber scombrus Bullet, 243, 263-264, 266, 279, see also Myliobatidae, 35, 38 Myxine, 43, 50, 381-382, 385, 414 Auxis rochei Myxinidae, 15 Chub, see Scomher colias Myoxocephalus octodecimspinosus, 150 Frigate, see Auxis thazard Horse, see Trachurus mediterraneus N Japanese, see Scomber japonicus King, see Scomberomorus caualla Needlefish, see Tylosurus Pacific, see Scoinber japonicus Negaprion hreuirostrus, 118, 264 Spanish, see Scomber scombrus

1

566 Nemichthys, $6 Neoceratodus, 43 Nerophis, 27 Notacanthiformes, 14, 388 Notopteridae, 28, 33, 369 Notopterus, 415 No tropis N . atherinoides, 132 N . cornutus, 224 N . spilopterus, 132

SYSTEMATIC INDEX

P. Juuiatilis, 120 P. perca, 148, 526

Perch climbing, see Anabas testudineus yellow, see Perca jlauescens Perciform, 23, 68, 388 Percopsiformes, 388 Percopsis omiscomaycus, 132 Periophthalmus, 41, 375 Petromyzontidae, 14,. 42 Petromyzon marinus, 140, 142, 146,471 0 Phocoenoides, 216 Pholidophoroid, 388 Odontogadus merlangus auxinas, 118 Pholis, 216 Ogcocephalidae, 40 P. gunnelus, 120 Ogcocephalus, 42 Phoxinus laevis, 87, 139, 160-161 Oncorhynchus, 77, 79, 215 Pike, 103, see also Esox lucius 0. gorbuscha, 78, 108 Pimephales promelas, 148, 160-161 0. keta, 79, 108 Pisodonophis boro, 43 0. kisutch, 80, 108, 110, 118, 128-129, Plaice, see Pleuronectes platessa 132, 140, 146, 153, 159, 162, 170Platichthys stellatus, 458-459 171,473, 525 Platygobio gracilis, 132 0.nerka, 13,19,76,78,80,83,107-110, Plecostei, 326 118, 129, 132, 135-140, 143, 145Pleuronectes 146, 153-154, 156, 162, 164-167, P.Jesus, 474 21.3-214, 427, 429, 430-431, 433, P. microcephalus, 120 435,439, 445, 468,482,504, 531 P. platessa, 81, 108, 120, 143, 472, 4 74, 0. tshawytscha, 53, 76, 108-109, 118, 477 146, 169-170 Pleuronectiformes, 14-16, 29 opah, 355 Poeciliidae, 15 Ophichthyidae, 29 Pollachius uiens, 132 Ophichthys boro, 43 Polyodon, 72 Ophidiidae, 15, 17 Polypterus, 6.3, 67 Ophiodon elongatus, 467-468, 471, 474Pomacentridae, 30, 37, 39 475, 477, 528 Pomatomus, 15, 20 Opsanus tau, 454, 456 P. saltaltrix, 120, 241, 445, 449, 468 Ostariophysi, 388 Pomicrops itajara. 120 Osteoglossidae, 14, 45, ,388 Pomolbus pseudoharengus, 116, 118 Ostracidae, 15, 24 Porcupine fish, see Chilomycterus Ostracion, 11 schoepji 0. tuberculatum, 9 Porpoises, 241 Otophidium scrippsi, 17, 43 Prionace gluucu, 362, 371 Prosopium williamsoni, 132 P Protacanthopterygii, 388 Protoelopoid, 388 Pantodon, 4 Protopterus, 43, 415 P. buchholzi, 34, 45-46 Pseudopleuronectes americanus, 150, I.56 Paracanthopterygii, 388 Pterophyllum, 28 Perca, 51, 120 P.pauescens, 83, 132, 138-139, 148,224, Puffers, 24 Pumpkinseed, see Lepomis gihhosus 526

SYSTEMATIC INDEX

R

Salmon, 43,55,57-59, 70-71,79-80, 117, 130, 164, 210, 436, 445, 507 Raia, 371, 375, 411 Atlantic, see Salmo salar R . binoculata, 478 chinook, see Oncorhynchus R. clavata, 412, 512 tshaw ytscha R . undulata, 9 chum, see, Oncorhynchus keta Raiidae, 29, 35 coho, see Oncorhynchus kisutch Raja, see Raia Pacific, see Oncorhynchus Raven, sea,see Hemitripterus americanus sockeye, see Oncorhynchus nerka ,Rays, 4,6,35,40,43,72,83,231,362,395, Salmonidae, 14, 65, 77, 81, 129, 171, 217, 411,415 229,383-384,388-389, 395,410,426, eagle, 35, 38 442-443, 466, 468, 525 electric, 24 Salvelinus Redfish, see Sebastes marinus S. alpinus, 132 Regalecidae, 14 S. fontinalis, 118, 143-144, 146, 148, Regalecus, 27, 32 152, 161-162 Reinhurdtius hippoglossoides, 17 S. nanaycush, 148, 152-153 Remora remora, 445 Sand lance, 48, 59 Rhaphiodon, 45 Sarda, 15, 242, 250, 260, 274, 276277, Rhina, 72, 367, 373 279,290 Rhineodon, 15, 23, 40 S . chiliensis, 20, 111- 112,243-245,249, Rhinomugil corsula, 505 254, 263, 266, 275, 279, 292, 297, Rhizoprionodon terraenovae, 263-264 327-328 Ribbonfish, 32 S. sarda, 245, 254, 292 Roccus, 37 Sardini, 327 Runner, blue, see Caranx crysos Sardinoides, 388 Rutilus rutilus, 118 Sardinops, 15, 20 Ruvettus, 374 Sargassum fish, 40 R . pretiosus, 220 Scabbardfish, see Aphanopus carbo Scardinius erythrophthalmus, 148, 224 Scaridae, 31, 67 Sciaena aguila, 148 Saccopharyngoidei, 14 Sciaenidae, 31 Sailfish, 76, 281 Scomber, 15, 20, 250, 276277, 279, 290 Salmo, 54-55, 215-216, 391 S. colias, 326, 371 S . clarkii, 107-108, 11C S. japonicus, 20,243,249,253-254,266, S.fario, 118 283-284, 292, 295,297, 531 S. gairdneri, 9, 18, 108, 111, 118, S. scombrus, 10, 13, 20, 120, 150, 242129-132, 134, 136, 139, 144-146, 243, 245, 247, 249-250, 254-255, 157-158, 161, 166167, 170-171, 258-259, 272, 279, 289, 292, 326213-214, 217, 224, 226, 389-390, 327,444-445, 448, 468 407,427-428,433, 438,440,443, Sconiberomorini, 327 445-446, 449, 452-453,455-457, Scomberomus, 277, 290 460-463,465,468, 474-476,479, S. cavalla, 262 504-505,507-509, 520 S. commmson, 262 S. irideus, 118, 146, 224, 362 S . sierra, 292 S. salar, 7 6 7 7 , 103, 107-108, 110, 118, Scombridae, 14-15, 20, 22-23, 47-48, 146, 156157, 466 54-55, 59, 62, 77, 82-83, 101, 114, S. trutta, 138-139, 146, 224, 444, 468, 197, 240, 260, 326-327, 329, 409-410 474 Scorpaena, 42

s

568 Scorpaenids, 129, 388 Scorpion fish, see Scorpaena Scup, northern, see Stenotomus crysops Scyliorhinus, 54, 366, 374-375, 378-379, 382-385, 399,401,403,411, 413 S. canicula, 362, 371, 393, 456, 490 Scyllium, 69 S. canicula, 362, 371, 393, 456, 490 Seabass, 37 Seahorse, see Hippocainpus Sea robins, 42

Sebastes S . dactylopterus, 150 S . marinus, 112, 150, 156 Seriolu quinqueradiata, 113 Serranidae, 30-31 Serranus seriba, 120 Shad, see Alosa sapidissima Shark, 4, 17, 22-23, 43, 48, 54, 66, 68-69, 71, 83, 255, 259, 264, 326, 376, 382, 393, 395, 406, 411, 414, see also Heteroclontus; Scyliwhinus; Scyllium; Squulus Atlantic sharpnose, see Rhizoprionodon terraenooae basking, see Cetorhinus bull, see Carcarhinus leucas laninid, 54, 241, 289, 326, 355 lemon, see Negaprion brevirostrus shortfin mako, see Isurus oxyrinchus soupfin, see Galeorhinus zyopterus tiger, see Galeocerdo cuvieri whale, see Rhineodon white, see Carcharodon carcharius Shrimpfish, 34 Sierra, see Scoinberomorus sierra Silverside, 43 Siluriformes, 14, 17 Skates, 35, 68-69, 71-72, 231, 478 Snakehead, 41 Snipefish, 34 Sole, see Solea uulguris Solea vulgaris, 34, 81,474 Sphyruena I~urracudu,120 Spicara smuris, 120 Sprat, see Clupea sprattus Squulus, 14, 366367, 374, 383 S . ucanthias, 9, 17, 7 6 7 7 , 382,445,456, 478, 484 Stenodus leucichthys, 132 Stenotomus crysops, 445, 468

SYSTEMATIC INDEX Stickleback, 38, 50, 72 Stizostedion S . oitreum, 138-139 S . vitreum vitreum, 132, 148 Stromateidae, 15, 369, 391 Sturgeons, 65,255,259,see also Acipenser Sucker, see Catostomus Sunfish, ocean, 38, 355, see also Molidae Surfperch, 10, 38, see also Cymatogaster aggregata Swordfish, see Xiphias gladius Synancejidae, 42 Synbranchidae, 43 T

Tautoga, 31 T. onitus, 108 Teleosts, 23, 47-49, 52, 54-56, 58, 60-62, 6 6 6 7 , 69, 71, 74, 363, 386390, 394396,402-407,414-416,427,471,477, 479-480, 487-488 Tench, see Tincu Tetraodon, 36 T. Juoiatilis, 38 Tetraodontidae, 15, 24, 30, 3 6 3 7 , 67 Tetrapturus T. albidus, 76 T. audax, 78 Thunnus, 14, 43, 54, 242, 250, 260, 2 7 6 277,279, 286,290-291,293,305-307, 325,327-328, 331, 338, 355-356 T. ulalunga, 75-78, 243-244, 25.3-255, 263-267, 276, 279, 286-287, 292, 305, 329, 332 T. ulhacares, 9, 111, 113, 116, 120, 224, 226, 242-243, 245, 249, 253-254, 261, 263-264, 266267, 276, 279, 289, 292-295, 297, 305, 317, 329333, 338, 342-343, 346, 350 T. atlanticus, 329 T. maccoyii, 329 T. obesus, 111,243,245, 253-254, 262263, 266267, 276, 279, 292-293, 305, 329-33.3, 338 T. thynnus, 75-77, 120, 242-244,269, 282, 292, 317-318,328433,338, 344-348, 355-356 T. tonggol, 326, 329 Thymallus, 167 T. articus, 132

569

SYSTEMATIC INDEX

Tilapia, 64, 157 T. mossambica, 165, 505, 515, 530 T. nilotica, 160, 166167, 433, 441, 465 Tinca T. tinca, 118,406, 442-443,455, 531 T. vulgaris, 531 Toadfish, see Opsanus tau Torpedo, 384, 395, 397 T. nobiliana, 13, 24 Trachipterus, 27 Trachurus T. mediterraneus, 284, 289 T. meditterraneus pontious, 120 T. symmetricus, 23, 150, 407, 507 Trachypteridae, 32, 218 Trachyuridae, 362 Trachyurus, 409 Trematomus borchgreuinki, 153, 527 Triakis, 15 Trichiuridae, 14, 15, 25, 27 Trichiuris, 27 Triggerfish, 33-34, 38 Trigla, 120, 414 T. cuculus, 414 Triglidae, 42 Trout, 17-18,57, 145, 193,209-210,222223,408, 410, 436, 438, 443, 469, 472, 481, 484, 489, 508 brook, see Salvelinus fontinalis brown, see Salmo trutta cutthroat, see Salmo clarkii lake, see Saluelinus namaycush rainbow, see Salmo gairdneri Tuna, 4, 11, 20-21, 50, 54, 6 6 6 7 , 74, 87, 101, 104, 111, 116, 240, 265, 315, 326327, 355, 410 bigeye, see Thunnus obesus blackfin, see Thunnus atlanticus bluefin, see Thunnus thynnus

longtail, see Thunnus tonggol skipjack, see Katsuwonus pelamis skipjack, black, see Euthynnus allet-

teratus slender, see Allothunnus falai yellowfin, see Thunnus albacares Tunny, little, see Euthynnus alletteratus

Tylosurus, 43-44 U

Umbra, 38, 43 Umbridae, 30-31

W Wahoo, 116, 226, 242-244, 253, 263-265,

276, 295-296, see also Acanthocyhium solanderi Walleye, see Stizostedion uitreum Whale, 15, 23, 241 Whiting, see Gadus merlangus Winged fish, 45 Wrasse, see Labridae green, see Crenilabrus tinca X

Xiphius, 355 X . gladius, 241 Xiphiidae, 15, 22-23, 101, 220, 326 2 Zeiformes, 67 Zeus, 28 Z. faber, 34 Zoarces viui parous, 120

SUBJECT INDEX A Acceleration, 11, 18, 40, 44, 70, 102, 1 1 6 117,221-227, see also Swimming speed, burst mechanics, 226 work done, 2 2 6 2 2 7 Acidhase regulation, 456 Acidosis, 460 Adhesion, see Suction Air breathing fish, 40-41 Amiiform, see Swimming modes Anal fixation, see Caudal complex Anguilliform, see Swimming modes Aspect-ratio, see Tail fin

B Balistiform, see Swimming modes Baroreceptors, 472-473 Blood circulation, metabolic cost, 454 flow redistribution during exercise, 487 pressure arterial, 480-483 levels during exercise, 480-483 pulsatility, 484 .venous, 4 8 6 4 9 1 volume, 456 Body density, 254 depth, 190, 207 length, 13, 18, 78-79, 81-83, 85, 117, 137-140 shape, 13, 73,202, 204, 219, 228 size, 79-81, 85, 137, 276 surface area, 293 tapering, 207-208 Boundary layer displacement thickness, 190, 207 separation, 22-23, 40, 59, 104, 193-194, 207, 220 turbulent flow at, 196, 215, 220 570

Breathing cost, 446447 efficiency, 446 rates during exercise, 443 rhythmic, 443, 445 Buoyancy, 2-3, 36, 69, 127-128, 250, 252-255 negative, 39, 69, 8.3, 88, 101, 111-114, 24 1 sinking factor, 253 Burrowing, 11, 34, 42-43, 81, 87 Burst swimming, see Swimming speed, burst

C Cardiac, see Heart Carangiform, see Swimming modes Caudal complex, 22 Caudal fin, see Tail fin Chemoreceptors, 448 Circulation, see Blood, circulation Condition factor, 142, 144 Critical swimming speed, see Swimming speed, critical

D Density, see Body, density Digging, 43 Diodontiform, see Swimming modes Drag, 2, 4, 7-8, 23, 83, 127, 140, 144, 192-193, 197-198,218,251,268,270, 273,369,446447, see also Boundary layer coefficient, 190, 195,215, 268-269 components, 274-275,447 frictional, 190, 195, 198, 268, 276, 369 coefficient, 190, induced, 272-273, 276 due to pectoral fin, 256 pressure, 196 reduction, 114, 218-221, 227, 275, 288 aspect-ratio, 281

571

SUBJECT INDEX and hydrodynaniic resistance, 114 and schooling, 86, 221, 282-284 in slime, 280 in streamlining, 277

E Electric sense organs, 20,2627,32-33,84 Excretion ammonia, 505, 509, 515, 530, 532 carbon dioxide, 457-460 nitrogen, 530 Exercise acidhase regulation, 456 acidosis, 460 definition, 425 death due to excess, 168 hematocrit levels, 456 hemoglobin levels, 456 lactate accumulation, 460 parameters, 432-433 physical conditioning, 129-130 plasma protein levels, 456 recovery from, 350, 515-517 respiration during, 44,3453 salinity, effect, 158, 465-466

F Feathering, 295, 300 Fibers, see Muscle, fibers Fins, 62-64 absence, 32-33 adipose, 18, 65 anal, 9, 11, 17-18, 22-26, 28, 32-34, 42, 53, 59, 63-65, 67, 73-74 angle of inclination, 20, 22, 36, 38-39 base, 63 coordination, 33-35, 205-206 depression of, 64 development, 73 dorsal, 9, 11, 18, 22-25, 27-28, 32-34, 40, 58-59, 63-65, 67, 73-74 erection, 64 innervation, 72 internal supports, 62-64 margin, 63 median, 11, 17-19, 2 6 2 7 , 51, 64, 73, 205-206 mobile margin, 35 movement while stationary, 40, 70, 102

muscles, 368 pectoral, 9-11, 24, 26-28, 33, 35-38, 40-46, 67-70, 72-74, 257-258 pelvic, 35, 41-44, 59, 70-72 position, 73-74 propulsion, 26 rays, 62-67, 71-72 rod, 62 shape, 73 size, 73 spines, 33, 62-65, 71 stabilizing motion, 34, 65, 74 surface area, 73 topography, 73-75 web, 62 Finlets, 22-23, 59, 73, 260 Fish wheel, 8 Flying, 4, 11, 43-46 distance covered, 44-45 Foraging, 102, 105

G Gill area, 4, 271-272, 443, 460-461 diffusing capacity, 442, 462, 466 resistance, 271-272 secondary lamellae, 451-452 slits, 40 ventilation, see Ventilation Gliding, 4, 43-44, 83, 88, 117, 284-288 Gray’s Paradox, 8, 196197, 230 Gymnotiform, see Swimming modes

H Heart acceleration, 472 bradycardia, 467-468, 472-473, 475 chronotropic response, 469-472, 478 efficiency, 454 innervation, 471 inotropic response, 477-478 output, 457,462, 475,479-480, 483 pacemaker cells, 474 rate, 467-475 and swimming speed, 470 and temperature, 469, 471 Starling mechanism, 478-479 stroke volume, 461,473,475-480 tachycardia, 468, 472-473, 475, 487

572

SUBJECT INDEX

Heat excess temperature body, 319, 321 brain, 319, 331-332, 337 source, 321-325 exchange, 338-340 gain, 341-343 loss, 344-346 Hematocrit levels, 456 Hemoglobin, 130, 305-306, 456, 506 absence, 475 Bohr shift, 460 Root shift, 460 Hydrodynamics, 86, 88, 189-190 lift, 255 models, 199-204, 211-21.3, 215-217 Hydrofoils, 36, 41, 73 Hydromechanics, models, 199-204, 21 1213, 215-217, 307 Hydrostatic cushion, 50 equilibrium, 252-25.3 Hypercapnia, 460 Hypoxia, 440, 472

I Ion regulation, 4, 441, 465-466

J Jet propulsion, 11, 39, 46, 59 Jumping, 11, 4.3-44, 46, 79

K Keel, 21-23, 50, 74, 254

L Labriform, see Swimming modes Lactate, 130, 165, 168, 170, 350, 407-409, 459-460, 507,519, 522,524-525 dehydrogenase, 506, 517, 522 Lamellae, perfusion, 452 Lamellar recruitment, 452-453, 462, 464-465 Laminar boundary layer flow, 194-195, 215,218 maintenance, 195, 218-219, 221 Land speed, 41 Larval forms, 16, 43, 48, 57, 63, 123

Lateralis system, 25,27,32,45,50,58,61, 70 Learning, 85 Lepidotrichia, 63, 65, 69 Lift, 4, 24, 66, 255-256, 258-260 coefficient, 257-259, 261-262 fin angle alteration, 36 Light response, 84, 87, 107-110, 114 Lipid, 374, 506, 525, 530 Lipolysis, 529 Lipoprotein lipase, 528 Locomotion, 5, 75, 88 braking, 32, 69-71 bulk momentum, 8, 201, 203 crutching, 41 distance covered, 75-79 diurnal activity, 83-84, 87 gliding, 4, 43-44, 83, 88, 117, 284-288 nonswimming, 39 oscillatory, 11, 13, 24, 33, 37, 86 rigid-body, analogy to, 192-193, 197198 scale, 9 terrestrial, 40-42 turning, 11, 18, 32, 71, 74 undulatory, 5, 7-8, 13, 16-17, 1 9 , 2 5 2 6 , 28,32-34,36,38,46 Loconiotory behavior, 291, 228 contractions, 40 waves, 32

M Maneuverability, 17-18, 26 M auth n er cells, 61 fibers, 62 reflex, 45, 61-62 Mechanoreceptors, 448 Metabolism, 75, 87-88, 442-443, see also Oxygen, consumption aerobic, 55, 405-410, 427-437, 440 anaerobic, 55, 405-410, 437-439 cooperative, 409 output, 4, 8, 88, 343 oxygen debt, 438 maximum consumption, 434 protein, 505, 530-532 and weight, 442

SUBJECT INDEX Migration, 17, 79, 81-82, 84, 101-103, 106-109, 111, 116 hydroelectricity, effect, 170 downstream, 80 ocean currents, effects, 77-79, 81, 84, 103, 105-107, 171 upstream, 80-81, 103, 107, 426 Muscle arches, 5 6 5 7 attachment, 56, 62 bundles, 71 and burst speed, 369-371 contraction, 5657, 60, 6%64 bimetal theory, 5 6 5 7 fibers chiai slow, 410 classification, 372 development, 390-39.3 differentiation, 390-393 electrical properties, 397-405 fast, 368-390, 405-410 electrical properties, 397-405 innervation, 39,3497 mechanical properties, 397-405 Feldenstruktur, 376 Fibrillinstruktur, 376 innervation, 39.3-397 mechanical prbperties, 397-405 membrane properties, 385 slow, 368-390, 405-410 electrical properties, 397-405 innervation, 39.3-397 mechanical properties, 397-405 types, 383-384 mosaic system, 55, 57, 389, 392, 407 orientation, 364-367 pattern, 55-56 pink, 217, 387 power, effect of temperature, 349 red, 54-55, 60, 64-65,68, 70, 131, 2 1 6 217, 307, 353, 369, 370,466467, 504, see also Muscle, fibers, slow biochemistry, 505-507 creatine levels, 509 excess temperature, 318-319, 322323 resting potential, 370 sustained speed, 369-371 types abductor, 36, 68-70 profundus, 70-71

573 superficialis, 53, 70-71 adductor, 42, 53, 60, 68-70 mandibulae, 53 mandibularis, 60, 449 profundus, 53, 70 superficialis, 53, 70 I,ranchial, 60 carinal, 58 depressor, 31, 38, 53 analis, see Caudal complex dilator operculi, 53, 60 epaxial trunk, 64 extensor pectoralis, 53 flexor deep dorsal, 68 ' ventral, 68 deltoid, 68 inchator, 53, 64-65 analis, see Caudal complex dorsalis, 53 inferior oblique, 58 juxtaparietal, 380 lateralis profundus, 53, 68 superficialis, 53, 68 levator operculi, 53, 60 pectoral, 45, 72 protractor dorsalis, 53, 58 ischii, 53, 59 radial, 66, 72 rectus, 71 respiratory, 59-60 retractor analis, 59, see also Caudal complex dorsalis, 53, 59 ischii, 53, 59 superficial lateral, 60 superior oblique, 58 white, 55-56, 60, 64-65, 68, 131, 216217, 353, 369, see also Muscle, fibers, fast adenylate metabolism, 508-509 biochemistry, 505-507 creatine levels, 508-509 excess temperature, 319, 322-323, 370 Myocommata, 50-52, 54-56, 58 Myofibrillar ATPase, 513 Myogenic rhythm, 61

574

SUBJECT INDEX

Myoglobin, 370, 376, 506 Myotomes, 20, 47, 63, 71 design, properties of, 364 shape, 363

N Nerves, motor action potentials, 398-406 sodium conductance, 398-406 Nocturnal activity, 81, 83-87

0 Orientation, sun-compass, 40 Osmoregulation, 441, 465-466, 504 Ostraciiform, see Swimming modes Oxygen consumption, 151, 160, 163-166, 217, 341-342, 354,408,430, 434, 439, 442, 447-449, 464,467,479, 504 gas exchange, surface area involved, 442, 452 and swimming speed, 427-428, 430, 432-437, 440-441 debt, 438 delivery to tissues, 453, 460, 467 pulse, 479 P Paddlelike motion, see Swimming modes Pectoral fin, see Fins, pectoral flapping, 31, see cclso Swimming modes Labri form undulation, 30-31, 36, see ulso Swimming modes, Diodontiform Penduncle, 18-22, 47, 67 Power, 192-193, 203, 213, 216, 301 efficiency, 217 Predation, 84 Predators, avoidance, 39, 84-85, 117, 172, 351 Pressure receptors, see Baroreceptors Proprioception, 60-61, 410-416 Propulsion, wave generation, 33 length, 13,1 8 , 2 0 , 2 6 2 9 , 3 3 , 3 6 3 7 , 190, 201, 210,231 propagation, 46, 190, 210, 219 velocity, 190

R Rajiform, see Swimming modes Red muscle, see Muscle, red Reflex, baroreceptor, 472-473 Respiratory quotient, 504-505 Retia mirabila, 325, 328-334, 353 Reynold’s numbers, 3, 40, 48, 73, 190, 194-195, 211, 215, 242

s Schooling, 85-86, 88, 101-106, 109 drag reduction, 86, 221, 282-284 energy saving, 282-284 packing density, 86 swimming speed, 113-116 Sculling, see Swimming modes Shedding momentum, see Thrust Sodium pentachlorophenate, 161-162 Span, 14-20, 22, 33-34 Spawning, 59, 79, 81, 83, 85, 117 Species distribution, 353 Startle response, 62 Streamlining, 22, 24-25, 194 Stretch receptors, 411-412, 448 Suction, 42-43, 71-72 Swim bladder, 2-4, 6, 27, 32, 85, 112, 253-254, 263 Swimming activity cycles, 83-85, 87-88 biological approach to, 192-199 chamber, 117-128 depth, 80-82, 84-85, 105, 112, 114 effect of body form, 204-208 of mode, 209 of tail amplitude, 208-209 efficiency, 17, 217, 296, 430 endurance, 1.36, 140-141, 150 energetics, 163-168 experimental studies, 128-137 fatigue, 169-170, 438-441 hydromechanical approach to, 199-218 models, 199-200 Swimming capacity, 106 effect of carbon dioxide, 158 of length, 137-140 of oxygen, 156158, 160, 166 of parasites, 144-145, 150

SUBJECT INDEX of pollution, 160-163 of salinity, 158-160, 465-466 of temperature, 150-156, 160, 343 of water hardness, 162 of weight, 140-144 endurance, 136, 140-141, 150 energetics, 163-168 experiments, 128-131, 134-137, 145 influence of maturity, 144 of sex, 144 scope for activity, 151-152 temperature, 154 Swimming mechanics biological approach to, 192-199 hydromechanical approach to, 199-218 Swimming modes, 9-1 1 , 2 0 9 4 11,230 accesory movement, 36-37 Amiiform, 9, 26-27, 32-33, 65, 218, 230 Anguilliform, 9, 11-18, 24-25, 34-36, 40, 43, 46-47, 62, 7,3-74, 208-210, 212,221 back paddling, 40 backward swimming, 33 Balistiform, 9, 28, 32-34, 38, 65, 230 Carangiform, 9, 11-16, 19-20, 23, 25, 37,74, 197, 209-211,221,241,268, 288 classification, 11 comparison, 13-15 creeping, 11 Diodontiform, 9, 30-31, 36, 230 evolution, 65 Gymnotiform, 9, 28, 33, 65, 2.30 intermediate, 36-39 Labriform, 9-10, 30, 36, 38, 230 Ostraciiform, 9, 11-15, 20, 24-25, 35-36, 38 paddlelike, 31, 37-38, 69-70, see also Swimming modes, Tetraodontiforni pectoral flapping, 31, see also Swimming modes, Labriform Rajiform, 9, 35-36, 230-231 rowing, 31, 41 sculling, 25, 28, 30, 32, 37 skimming, 44 skipping, 44 Subanguilliform, 12 Subcarangiform, 9, 12-13, 17-20,22,24, 32,67, 197, 210, 214 taxiing, 44-46

575 Tetraodontiform, 9, 28, 35, 37-38, 65, 230 Thunniforni, 9, 12-13, 15, 19-23,25,47, 50, 56, 7,3-74 two-stage, 83 walking under water, 2, 42 Swimming performance effect of body size, 137-140 of body weight, 141 influence of disease, 144 of sex, 144 Swimming speed, 8, 11, 16-18, 23-24, 27, 32-33, 41, 44, 50, 56, 73, 78-87, 101117, 166, 427 burst, 101-106, 115-117, 130, 136, 138-140, 217, 248-250,307, 439, 491, 507 effect of oxygen, 157-158 of pollution, 160 of temperature, 154-156 critical, 102, 128, 131, 134-135, 138139, 158,434,438, 440, 453 effect of anadromous change, 107, 110 of pollution, 162 of salinity, 158-159 estimation using photography, 104 using television, 104 and heart rate, 470 maximum, 116,229 optimum, 435-436 prolonged, 101-104,115-116, 130,136, 140-141, 247-248, 519 effect of oxygen, 156-158, 247 of pollution, 160 of temperature, 151-154 slow, 34 specific amplitude, 434 specific speed, 79, 468 sustained, 101-107, 112, 114-116, 135-136, 140, 241-247, 266, 519 die1 patterns, 242, 246 effect of food deprivation, 242, 2 4 6 2 4 7 of oxygen, 156-157, 242, 246-247 of shoreline, 107

576

SUBJECT INDEX

of temperature, 151-154, 242, 246247 unsteady propulsion, 221-228

T

power, 190, 204, 213-215, 302 production, 192, 288-289 supplemental, 33 unsteady versus steady motion, 191,228 Thunniform, see Swimming modes Trailing edge depth, 190, 282 kinematics, 19, 200-202, 272 Transfer fktor, 462 Turning, 11, 18, 32, 71, 74

Tagging, 75-79, 81-82, 88, 103-107 Tail fin, 6, 9-11, 15, 18-24, 28, 34, 37-38, 40, 44, 46, 57, 59, 65-68, 7,3-74, 86, 207-208, 295-301 aspect-ratio, 17-20, 22-23, 38, 6 6 6 7 , 73,239,262,280-281, 290-292 V beat Ventilation, 4 amplitude, 10, 13, 16-20, 23, 33, 190, dead spaces, 450 204,208-209,295 exercise, effect, 443, 460-461 frequency, 18, 23, 33, 44, 117, 190, perhsion ratio, 461 204, 284,297, 434 ram, 267,443-445 efficiency, 190, 212, 217 change from rhythmic, 443-444, 446, expansion, 34, 67 448 heterocercal, 17, 65, 69 obligate, 267, 445 h a t e , 12, 22-23, 25 oxygen, cost, 448 rays, 20-22, 25, 66-68 volume, 450 as rudder, 74 Viscosity, 3, 40, 74, 207, 211-213, 220 variability, 211 kinematic, 190, 195 Temperature, ,see ulso Heat reduction, 220 change, coefficient, 337 Vortex lag, 351 generators, 220 regulation, 341 sheets, 18-19, 86, 205 Tetraodontiform, see Swimming modes Thermal W inertia, 344 lag, 351 White muscle, see Muscle, white runaway, 355 Thrust Y coefficient, 301 diagonal, 37 Yawing, 18-19, 210 forward, 4-5, 8, 16, 19-20, 22, 30, 33, 3 6 3 7 , 40, 45, 50, 66, 74, 192, 203, Z 213-214,301 lateral, 38, 67 Zygopophyses, 49-50