UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Foraging behaviour and perceived predation risk of juvenile chinook salmon (Oncorhynchus tshawytscha)… Gregory, Robert S. 1991

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-UBC_1991_A1 G74.pdf [ 7.83MB ]
Metadata
JSON: 831-1.0100583.json
JSON-LD: 831-1.0100583-ld.json
RDF/XML (Pretty): 831-1.0100583-rdf.xml
RDF/JSON: 831-1.0100583-rdf.json
Turtle: 831-1.0100583-turtle.txt
N-Triples: 831-1.0100583-rdf-ntriples.txt
Original Record: 831-1.0100583-source.json
Full Text
831-1.0100583-fulltext.txt
Citation
831-1.0100583.ris

Full Text

FORAGING BEHAVIOUR AND PERCEIVED PREDATION RISK OF JUVENILE CHINOOK SALMON (ONCORHYNCHUS TSHAWYTSCHA) IN TURBID WATERS. by ROBERT S. GREGORY BSc(Honours Biology) Acadia University 1980 MSc(Biology) Trent University 1984 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY in THE FACULTY OF GRADUATE STUDIES (Department of Zoology) We accept this thesis as conforming to the required standard THE UNIVERSITY OF BRITISH COLUMBIA February 1991 © Robert S. Gregory 1991 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Department of Zoology  The University of British Columbia Vancouver, Canada Date A p r i l 22, 1991  DE-6 (2/88) ii ABSTRACT I investigated the effect of turbidity on the foraging behaviour of juvenile Chinook salmon (Oncorhynchus tshawytscha) in the laboratory. Specifically, I examined a behavioural " t radeoff " between visual ability and "perceived" risk. I assessed visual ability by measuring the reaction distance of juvenile Chinook to planktonic Artemia prey. I found Chinook exhibited a log-linear decline in reaction distance with increasing turbidity. To determine the effects of turbidity and microhabitat on foraging rate, I conducted separate experiments for surface (Drosophila), planktonic (Artemia), and benthic (Tubifex) prey across a range of turbidity levels (0, 25, 50, 100, 200, 400, 800 mg-L-1). Foraging rates were reduced at higher turbidity conditions for all three prey. However, for surface and benthic prey, foraging rates were also low in clear water; highest rates were attained at intermediate turbidity levels (50-200 mg-L-1). The degree to which intermediate turbidities were associated with higher foraging rates was size-dependent. Smaller individuals (150-57 mm FL) exhibited relatively higher foraging rates in clear conditions than did larger individuals. However, planktonic foraging rates by juveniles were consistently high in clear water, regardless of fish size. In experiments manipulating light level independent of turbidity, I allowed salmon to forage under conditions which were either turbid, or clear but with light intensity correspondingly reduced. Foraging rates were similar between the two treatments for planktonic prey, but differed for benthic and surface prey. Generally, foraging rates exhibited by juvenile Chinook salmon could not be explained on the basis o f visual ability alone. I suggest that young salmon also exhibited iii fo rag ing behaviour cons is tent with thei r percept ion o f r i s k to p r e d a t i o n . In a r e n a exper iments , juveni le Chinook d is t r ibuted themselves randomly in t u r b i d condi t ions ; in c l e a r conditions they assoc iated with the bottom. When bird and f i s h p r e d a t o r models were introduced the f i sh a l t e r e d t h e i r spa t ia l d i s t r ib u t ion , occupying deeper regions r e g a r d l e s s o f tu rb id i t y . However, their r e s p o n s e in turbid conditions was less marked and las ted f o r a s h o r t e r time. Turbidity apparent ly mitigated the p e r c e i v e d r i s k o f predat ion in juvenile Chinook. I developed a conceptual t r a d e o f f model tha t p red ic ted the general e f f e c t o f turbidity on forag ing behaviour. Assuming d i f f e r e n c e s in e i ther p rey quality o r perceived r i s k o f predat ion in t h r e e microhabitats ( su r face , water column, bottom), the model reso lved the a p p a r e n t d iss imi lar i t ies between planktonic and o t h e r forag ing behav iours . P e r c e i v e d r i s k o f Chinook to predat ion was significantly d i f f e r e n t between s u r f a c e and water column microhabitats. When exposed t o a non -v i sua l " f i x e d - r i s k " st imulus (sound), salmon apparent ly p e r c e i v e d l e s s r i s k as turbidity increased. I conclude t h a t in t u r b i d w a t e r s juvenile salmon exhibit forag ing behaviour in a manner cons is tent with a t r a d e o f f between thei r visual abil ity and p e r c e i v e d r i s k . iv T A B L E OF C O N T E N T S Page ABSTRACT ii LIST OF TABLES . vii LIST OF FIGURES viii ACKNOWLEDGEMENTS xiii Chapter 1. G E N E R A L INTRODUCTION 1 Chapter 2. B A C K G R O U N D AND H Y P O T H E S E S 4 2.1 Turbidity and Vision 4 2.2 Turbidity and Foraging Ability 5 2.3 Turbidity and Predation Risk 7 2.4 Foraging Under Risk of Predation 8 2.5 Hypotheses and Experiments Conducted 10 - Visual Ability 10 - Alteration of Search Behaviour 11 - Enhancement of Visual Contrast 11 - Predation Risk 12 - The Ability-Risk T radeo f f 12 Chapter 3. G E N E R A L METHODS 14 3.1 The Study Animal 14 3.2 Turbidity 15 3.3 Experiments on Foraging Rate 16 - Prey 16 - Pre-Experimental Conditioning 19 - Experimental Apparatus ; . . . . 20 - Experimental Protocol 22 3.4 Spatial Distribution and Predator Models 23 - Experimental Apparatus 23 - Pre-Experimental Conditioning 26 3.5 Statistics 28 Chapter 4. V I S U A L A B I L I T Y AND F O R A G I N G BEHAVIOUR 29 4.1 Reaction Distance - Visual Ability 29 - Introduction 29 - Methods 30 experimental apparatus 30 conditioning and observation protocol 32 - Results 33 - Discussion 35 V Page 4.2 Foraging Rates f o r Sur face , Planktonic, and Benthic Prey 35 - Introduction 35 - Methods 37 the e f f e c t of turbidity on foraging rate 37 the e f f e c t s of ontogeny on foraging rate in turbid conditions 38 - Results 38 surface foraging 38 planktonic foraging 41 benthic foraging 43 the e f f e c t of ontogeny and turbidity on foraging behaviour 45 - Discussion 47 e f f e c t of turbidity on foraging rate 47 e f f e c t s of ontogeny 49 4.3 The E f f e c t of Turbidity and Light on Foraging Behaviour 53 - Introduction 53 - Methods 54 - Results 55 surface foraging 55 planktonic foraging 56 benthic foraging 56 - Discussion 60 Chapter 5. M ICROHABITAT S H I F T S AND R E S P O N S E S T O P R E D A T O R S 62 5.1 Spatial Distribution 62 - Introduction 62 - Methods 63 - Results 64 - Discussion 65 5.2 Responses to Model Predators 71 - Introduction 71 - Methods 73 the magnitude of the e f f e c t 73 the duration of the e f f e c t - recovery from predator disturbance 75 - Results 76 the magnitude of the e f f e c t 76 the duration of the e f f e c t 78 - Discussion 85 Chapter 6. A C O N C E P T U A L MODEL OF F O R A G I N G BEHAVIOUR UNDER V ISUAL C O N S T R A I N T 91 6.1 Introduction 91 6.2 The E f f e c t of Visual Ability on Foraging Rate 92 6.3 The E f f e c t of Perceived Risk of Predation on Foraging Rate 95 6.4 The T radeo f f 99 vi Page 6.5 Predictions of the Model and Re-Interpretation of Foraging Rates 103 - The E f f e c t of Prey Quality and Microhabitat 103 - The E f f e c t of Enhanced Risk 106 6.6 Synopsis 108 Chapter 7. T E S T S OF T H E VISUAL A B I L I T Y - P E R C E I V E D RISK T R A D E O F F 110 7.1 The E f f e c t of Microhabitat and Prey Quality on Foraging Decisions 110 - Introduction 110 - Methods 112 - Results 113 - Discussion 115 7.2 The E f f e c t of Enhanced Risk on Foraging Rate in Turbid Water 118 - Introduction 118 - Methods 120 - Results 122 - Discussion 127 Chapter 8. G E N E R A L DISCUSSION AND C O N C L U S I O N S 131 8.1 General Discussion 131 - Vision as a Constraint 131 - Turbidity and the Risk of Predation 131 - T r adeo f f s Between Visual Ability and Perceived Risk in Fish Foraging Behaviour 132 - Turbid Water Foraging in Fishes 133 - Salmonid Life Histories 134 - Perceived Risk in Behaviour Studies 135 8.2 Conclusions 137 LITERATURE CITED 140 APPENDICES 152 Appendix 1. Individual Feeding Trials f o r Drosophila, Listed by Experiment 152 Appendix 2. Individual Feeding Trials f o r Artemia, Listed by Experiment 157 Appendix 3. Individual Feeding Trials f o r Tubifex, Listed by Experiment 162 Appendix 4. Godin ,T .1. and R.S.Gregory, (in prep . ) -abstract 166 vii LIST OF TABLES T a b l e 3 . 1 . P h y s i c a l and v i sua l c h a r a c t e r i s t i c s o f p r e y u s e d in f o r a g i n g r a t e d e t e r m i n a t i o n s 17 T a b l e 4 . 1 . A n a l y s i s o f v a r i a n c e o f t h e e f f e c t o f t u r b i d i t y a n d y e a r o f e x p e r i m e n t o n t h e f o r a g i n g r a t e o f j u v e n i l e C h i n o o k s a l m o n t o A . S u r f a c e p r e y ( D r o s o p h i l a ) , B . P l a n k t o n i c p r e y ( A r t e m i a ) , a n d C . B e n t h i c p r e y ( T u b i f e x ) 39 T a b l e 5 . 1 . T h e e f f e c t o f t u r b i d i t y o n t h e s p a t i a l d i s t r i b u t i o n o f j u v e n i l e C h i n o o k s a l m o n b y r e g i o n 66 T a b l e 5 . 2 . T h e e f f e c t o f t u r b i d i t y o n A . T h e l a t e r a l a n d B. T h e v e r t i c a l d i s t r i b u t i o n o f j u v e n i l e Chinook s a l m o n b y p o o l e d r e g i o n 67 T a b l e 5 . 3 . T h e e f f e c t of t u r b i d i t y on p e r c e n t a g e o f j u v e n i l e Chinook s a l m o n a s s o c i a t e d with A . T h e b o t t o m a n d B. T h e s u r f a c e 69 T a b l e 5.4. A n a l y s i s o f v a r i a n c e o f t h e e f f e c t s o f t u r b i d i t y , c o v e r , a n d e x p o s u r e t o p r e d a t o r m o d e l s o n p r o p o r t i o n s o f j u v e n i l e C h i n o o k s a l m o n l o c a t e d in A . D e e p e r w a t e r ; a n d B. S h a l l o w e r w a t e r o f an e x p e r i m e n t a l a r e n a . . . 79 T a b l e 5 . 5 . A n a l y s i s o f v a r i a n c e o f t h e e f f e c t s o f t u r b i d i t y , c o v e r , a n d p r e d a t o r t y p e on r e c o v e r y t ime f r o m e x p o s u r e t o mode) p r e d a t o r s 87 viii LIST OF FIGURES Figure 3.1. Aquarium array used f o r determinations of foraging r a t e . A. Schematic representat ion. B. Photograph. 21 Figure 3.2. The relationship of turbidity concentration used in foraging rate experiments with measures of light transmission . A . Nephelometric Turbidity Units . B . Light energy. 24 Figure 3.3. Arena used during experiments on spatial distribution and predation risK manipulation . A . Schematic representation . B . Photograph 25 Figure 3.4. The e f f e c t of depth and turbidity concentration on light, within the experimental arena (Fig. 3.3). . . 27 Figure 4.1. Schematic representation of the experimental apparatus used f o r the determination of reaction distances 31 Figure 4.2. The e f f e c t of turbidity on the reaction distance of juvenile Chinook salmon f o r Artemia prey . 34 Figure 4.3. The e f f e c t s of turbidity on mean foraging rate of juvenile Chinook salmon feeding on surface prey and the percentage of salmon foraging in 70-L aquaria. A. 1987 - 5 tr ia ls . B. 1988 - 3 trials 40 Figure 4.4. The e f f e c t s of turbidity on mean foraging rate of juvenile Chinook salmon feeding on planktonic prey and the percentage of salmon foraging in 70-L aquaria. A. 1987 - 5 tr ia ls . B. 1988 - 3 trials 42 Figure 4.5. The e f f e c t s of turbidity on mean foraging rate of juvenile Chinook salmon feeding on benthic prey and the percentage of salmon foraging in 70-L aquaria. A . 1987 - 5 tr ia ls . B. 1988 - 3 trials 44 Figure 4.6. E f f e c t of size of juvenile Chinook salmon on the slope of the ascending limb (0-100 mg*L~ 1) of the relationship between foraging rate and turbidity . A. Surface prey - Drosophila. B. Planktonic prey -Artemia . C . Benthic prey - Tubif ex 46 Figure 4.7. Hypothetical e f f e c t s on foraging rate of variables differentially a f f ec ted by turbidity and ontogeny. A . Matched e f f e c t s . B . O f f s e t e f f e c t s 52 ix Figure 4.8. The e f f e c t on juvenile Chinook salmon of turbidity and light level on foraging rate fo r surface prey (Drosophila ). A. Mean foraging r a t e . 8 . Composite trial ~ T 57 Figure 4.9. The e f f e c t on juvenile Chinook salmon of turbidity and light level on foraging rate f o r planktonic prey (Artemia). A. Mean foraging r a t e . B. Composite t r ia l . 58 Figure 4.10. The e f f e c t on juvenile Chinook salmon of turbidity and light level on foraging rate fo r benthic prey (Tubifex). A . Mean foraging r a t e . B. Composite trial 7 59 Figure 5.1. The e f f e c t of turbidity on the proportion of juvenile Chinook salmon in the lower 20 cm of a 40 cm deep experimental arena 68 Figure 5.2. The immediate spatial response of juvenile Chinook salmon to exposure to predator models in clear and turbid conditions; proportion of fish in the deepest region of an experimental arena in, A . The absence and B. The presence of additional cover in two replicates 77 Figure 5.3. The immediate spatial response of juvenile Chinook salmon to exposure to predator models in clear and turbid conditions; proportion of fish near the surface in deep water of an experimental arena in, A. The absence and B . The presence of additional cover in two replicates 80 Figure 5.4. Changes in microhabitat distribution in juvenile chinook salmon, before and a f t e r exposure to a model bird predator , in clear (A. and B.) and turbid (C. and D.) water with additional cover 81 Figure 5.5. Changes in microhabitat distribution in juvenile chinook salmon, before and a f t e r exposure to a model bird predator , in clear (A . and B.) and turbid (C . and D.) water without additional cover 82 Figure 5.6. Changes in microhabitat distribution in juvenile chinook salmon, before and a f t e r exposure to a model fish predator , in clear (A. and B.) and turbid (C. and D .) water with additional cover 83 X F i g u r e 5 . 7 . C h a n g e s in m i c r o h a b i t a t d i s t r i b u t i o n in j u v e n i l e c h i n o o K s a l m o n , b e f o r e a n d a f t e r e x p o s u r e t o a m o d e l f i s h p r e d a t o r , in c l e a r ( A . a n d B . ) a n d t u r b i d ( C . a n d D .) w a t e r w i t h o u t a d d i t i o n a l c o v e r 84 F i g u r e 5 . 8 . T h e e f f e c t o f t u r b i d i t y a n d a d d i t i o n a l c o v e r o n t h e r e c o v e r y t ime o f j u v e n i l e c h i n o o k s a l m o n a f t e r e x p o s u r e t o m o d e l p r e d a t o r s 86 F i g u r e 6 . 1 . T h e e f f e c t o f t u r b i d i t y o n v i sua l ab i l i t y ( V A ) . . . . 94 F i g u r e 6 . 2 . T h e e f f e c t o f v i sua l ab i l i t y o n : A . P r o b a b i l i t y o f f o r a g i n g , P ( F V A ) I a n d §• F o r a g i n g r a t e u s i n g H o l l i n g ' s (1959) " d i s c " e q u a t i o n s o l v e d f o r s e a r c h r a t e 96 F i g u r e 6 . 3 . T h e e f f e c t o f t u r b i d i t y o n p e r c e i v e d r i s k ( P R ) . . . 98 F i g u r e 6 . 4 . T h e e f f e c t o f p e r c e i v e d r i s k o f p r e d a t i o n o n A . p r o b a b i l i t y o f f o r a g i n g P ( F p R ) , a n d B. f o r a g i n g r a t e u s i n g H o l l i n g ' s (1959) " d i s c " e q u a t i o n , s o l v e d f o r h y p o t h e t i c a l e f f e c t s o f p e r c e i v e d r i s k o n s e a r c h r a t e 100 F i g u r e 6 . 5 . A c o n c e p t u a l m o d e l o f t h e e f f e c t o f a b e h a v i o u r a l t r a d e o f f b e t w e e n v i sua l ab i l i t y a n d p e r c e i v e d r i s k o n r e l a t i v e f o r a g i n g r a t e in t u r b i d w a t e r s 102 F i g u r e 6 . 6 . P r e d i c t e d r e l a t i v e f o r a g i n g r a t e s a t t w o l e v e l s o f p e r c e i v e d r i s k 104 F i g u r e 6 . 7 . H y p o t h e t i c a l e f f e c t o f e n h a n c i n g r i s k s t imu l i o n f o r a g i n g r a t e f o r p l a n k t o n i c p r e y . A . A s s u m i n g d e c r e a s e in p e r c e i v e d r i s k a s a f u n c t i o n o f t u r b i d i t y . B. A s s u m i n g a d e c r e a s e in f i x e d - r i s k a s a f u n c t i o n o f t u r b i d i t y 107 F i g u r e 6 . 8 . H y p o t h e t i c a l e f f e c t o f e n h a n c i n g r i s k s t imul i o n f o r a g i n g r a t e f o r s u r f a c e a n d b e n t h i c p r e y . A . A s s u m i n g d e c r e a s e in p e r c e i v e d r i s k a s a f u n c t i o n o f t u r b i d i t y . B. A s s u m i n g a d e c r e a s e in f i x e d - r i s k a s a f u n c t i o n o f t u r b i d i t y 109 F i g u r e 7 . 1 . T h e e f f e c t o f t u r b i d i t y o n f o r a g i n g r a t e in j u v e n i l e c h i n o o k s a l m o n . A . S u r f a c e f o r a g i n g f o r D r o s o p h i l a a n d d r i e d A r t e m i a p r e y . B. P l a n k t o n i c f o r a g i n g A r t e m i a a n d P o l y s o r b a t e - 8 0 t r e a t e d D r o s o p h i l a . . . 114 xi Figure 7.2. The e f f e c t of turbidity and microhabitat on the coefficient of variation (/.) in foraging rate by juvenile chinook salmon in individual t r ia ls . A . Artemia . B. Drosophila 116 Figure 7.3. Apparatus modifications to accommodate "risk enhancement" experiments . A . Schematic representat ion. B. Photograph 121 Figure 7.4. The e f f e c t of turbidity and risk on mean foraging rate on surface prey by juvenile chinook salmon. . . 123 Figure 7.5. The e f f e c t of turbidity and risk on mean foraging rate on planktonic prey by juvenile chinook salmon. 125 Figure 7.6. The e f f e c t of turbidity and risk on mean foraging rate on benthic prey by juvenile chinook salmon. . . 126 Figure 7.7 The e f f e c t of turbidity on the di f ference between control and enhanced risk foraging rates by juvenile chinook salmon . A . Surface foraging - Drosophila . B. Planktonic foraging - Artemia. C . Benthic foraging - Tubifex 128 Figure A l . E f f e c t of turbidity on the foraging rate of juvenile chinook salmon f o r Drosophila prey , 1987 152 Figure A2. E f f e c t of turbidity on the foraging rate of juvenile chinook salmon f o r Drosophila p rey , 1988 153 Figure A3. E f f e c t of A. Light and B. Turbidity on the foraging rate of juvenile chinook salmon fo r Drosophila p rey . 154 Figure A4. E f f e c t of turbidity on the foraging rate of juvenile chinook salmon f o r A . Surface Drosophila and B. Planktonic Drosophila 155 Figure A5. E f f e c t of turbidity on the foraging rate of juvenile chinook salmon f o r Drosophila prey in A. "Normal" and B. "Enhanced" r i sk . 156 Figure A6. E f f e c t of turbidity on the foraging rate of juvenile chinook salmon f o r Artemia prey , 1987 157 Figure A7. E f f e c t of turbidity on the foraging ra te of juvenile chinook salmon f o r Artemia p rey , 1988 158 Figure A8. E f f e c t of A. Light and B. Turbidity on the foraging rate of juvenile chinook salmon f o r Artemia prey . 159 xii Figure A9. E f f e c t of turbidity on the foraging rate of juvenile chinook salmon fo r A. Surface Artemia and B. Planktonic Artemia 160 Figure A10. E f f e c t of turbidity on the foraging rate of juvenile chinook salmon fo r Artemia prey in A. "Normal" and B. "Enhanced" risk 161 Figure A l l . E f f e c t of turbidity on the foraging rate of juvenile chinook salmon f o r Tubifex prey , 1987 162 Figure A12. E f f e c t of turbidity on the foraging rate of juvenile chinook salmon fo r Tubif ex prey , 1988 163 Figure A13. E f f e c t of A. Light and B. Turbidity on the foraging rate of juvenile chinook salmon f o r Tubif ex prey . 164 Figure A14. E f f e c t of turbidity on the foraging rate of juvenile chinook salmon f o r Tubif ex prey in A. "Normal" and B . "Enhanced" risk 165 Figure A15. E f f e c t of A. Prey colour, B. Background turbidity, and C . Both foreground and background turbidity, on the reaction time of juvenile chinook salmon f o r Artemia prey (Godin and Gregory in prep . ) 167 xiii ACKNOWLEDGEMENTS Financial support for these studies was provided by a Natural Sciences and Engineering Council (NSERC) operating grant (#5-83454 to T.G. Northcote), and a Fisheries and Oceans Canada, Subvention Grant (#5-56574). Personal support was provided by NSERC Postgraduate Scholarships, a University Graduate Fellowship (UBC), UBC Teaching Assistantships, and the above operating grant to T.G. Northcote. I am grateful for these sources of funds. Juvenile chinook were obtained through the cooperation of the Chehalis River Fish Hatchery. I particularly wish to thank Larry Kahl, hatchery manager, for expediting requests. Additional equipment was provided by Cultus Lake Laboratory (Department of Fisheries and Oceans); I especially thank Dennis Martens. The many hours of laboratory analysis were conducted by Theresa Godin, Melanie Johnston, Ira Leroy, Regina Schiffer, Lisa Sennewald, and Tom Suzuki. Many of the potential pitfalls often encountered when maintaining fish at the South Campus Fisheries Compound were avoided with the advice and assistance of Rick Taylor. Field and other investigations added much valuable supporting information to the e f for ts I have presented. These investigations could not have been conducted without the help of Dana Atagi, Jim Berkson, AndrS Breault, Colin Daniel, Theresa Godin, Lois Hollett, Ira Leroy, Trish MacEachern, Carin Magnhagen, Lana Mah, Tom Northcote, Regina Schiffer, Tom Suzuki, and Peter Watts. My sincere gratitude is extended to all these individuals. I wish to thank my research committee: Dr.'s K.D. Hyatt, J.D. McPhail, W.E. Neill, T.G. Northcote, and D.J. Randall for their support, advice, and encouragement throughout the course of my investigations. My supervisor, Tom Northcote, is especially thanked for his support and xiv uncanny ability to find funds for me to hire summer students each year. I owe a special debt of gratitude to Bill Neill, who on numerous occasions took a great deal of time to critically review my research plans and generally kept me on track. The defense committee consisted of Dr.'s J . Gosline, M.C. Healey, T.G. Northcote, D.J. Randall, with K.M.J. McErlane performing as chairman. The external examiners were Dr. G. Power (University of Waterloo) and Dr. J.D. Hall (Oregon State Univers i ty ) . The various seminars in "the Institute" provided a comprehensive appraisal of the do's and don'ts of research. Nancy Butler, Moira Greaven, Glyph and the gang, Wes Hochachka, Dave Levy, Barbara McGregor, Debbie McLennan, and Peter Watts provided an endless supply of constructive criticism, advice, amusement or encouragement during all stages of my research. In addition to my research committee, Lana Mah and Peter Watts proofread drafts of various chapters, greatly improving the manuscript. Wes Hochachka painstakingly reviewed the manuscript in anticipation of the defense. My many thanks are extended to these ind iv idua ls . My family and friends were a constant source of support. Lana, you made the tribulations bearable. 1 CHAPTER 1 GENERAL INTRODUCTION No single topic in animal ecology receives more attention than feeding. The acquisition and assimilation of food has been variously linKed to factors such as growth, size at maturity, reproductive output, survival, resistance to disease, and social dominance, to name but a few. Placed at the same level of importance are behavioural and morphological factors associated with the avoidance of predators. It is perhaps inevitable that these factors will eventually conflict. Almost every animal is the prey of another. Often, the very act of obtaining or searching for food places a foraging animal at risk to predators. In such instances, foragers must "trade of f " the benefits of further foraging against the costs of potentially being, eaten while doing so. I experimentally examined such a tradeoff for juvenile chinook salmon (Oncorhynchus tshawytscha). A rich literature has developed over the past decade regarding behavioural tradeoffs of foraging and risk of predation (for review: Lima and Dill 1990). It has been amply demonstrated in these investigations that animals respond to increased risk of predation by altering their foraging behaviour, often at the expense of energy intake. Much of this literature has been directed at fish behaviour (for review: Dill 1983 and 1987; Werner and Gilliam 1984; Milinski 1986). My work lies within the general framework of these studies. However, the investigations I have conducted focus on an aspect of foraging ecology receiving little attention - visual ability and "perceived" risk of predation. I examined a behavioural tradeoff between visual ability and perceived risk of predation and investigated its effect on the foraging decisions of juvenile chinook salmon. Underwater images are generally poor in quality, due to the high degree of light attenuation in most aquatic systems (Duntley 1963). 2 This attenuation is attributed to the presence of light-scattering particulate material (e.g. plankton and suspended sediment), and to interference by water molecules themselves. Despite this environmental characteristic, most fish species depend on vision for much of their sensory input (Hyatt 1979; Miller 1979; Guthrie 1986). In conditions of elevated turbidity, the dissemination of light signals is especially acute. Visual range declines precipitously as a function of particle concentration (Duntley 1943; DiToro 1978) and must affect visual foraging ability and the subsequent behaviour of fish. Chinook salmon have been demonstrated to occupy turbid estuaries for a significant portion of their early life (Levy and Northcote 1982; Simenstad et al. 1982). Turbidity presents an unique opportunity to examine foraging behaviour in conditions where the environment simultaneously affects both the ability of a forager 1 to locate prey and its susceptibility to detection by its predators. The hunting techniques of most predators of salmonids predominantly involve the visual sense for the detection of their prey (Hobson 1979; Guthrie 1986). Therefore, increased turbidity may act to simultaneously reduce the visual ability and the perceived risk of foraging chinook. Despite evidence demonstrating the negative effects of turbidity on visual ability and foraging rate (Vinyard and O'Brien 1976; Confer et al. 1978; Gardner 1981), juveniles of many marine and anadromous fish species rear in estuarine nurseries having high concentrations of suspended sediment (Blaber and Blaber 1980; Levy and Northcote 1982; Simenstad et al. 1982; Cyrus and Blaber 1987a). Juveniles of several species of fish have been shown to actively prefer turbid over clear water (Cyrus and Blaber 1987b). The presence of these fish in such ^ I use the term "forager" when referencing juvenile salmon or any animal in a similar trophic position. "Prey" are food items consumed by foragers. A "predator" is an animal which consumes foragers. "Predation" occurs between predators and foragers; "foraging" occurs between foragers and prey. 3 waters has most commonly been attributed to the large prey densities in these productive habitats. To investigate the potential tradeoff between visual ability and perceived risk in turbid conditions, I examined key aspects of fish foraging behaviour related to visual constraints. These included the e f fec ts of turbidity on reaction distance, light levels, perceived risk of predation, and their subsequent effects on foraging behaviour. Juvenile chinook salmon are generalist foragers (Keast 1979) feeding on a variety of prey species from several microhabitats (Levy et al. 1979; Northcote et al. 1979; Healey 1982). I investigated the effect of turbidity on the foraging rate of chinook for prey from each of three generalized microhabitats frequented by these fish: surface, water column, and bottom. The effect of turbidity on ambient light levels is microhabitat specific. Therefore, I also conducted experiments to compare the effects of turbidity and light level on the foraging behaviour of these fish. It has been suggested that turbidity provides a form of protective cover for foraging fish, especially juvenile forms (Blaber and Blaber 1980; Simenstad et al. 1982; Bruton 1985). I tested several hypotheses concerning the perceived predation risk of juvenile chinook in turbid conditions, using both model predators and non-visual risk stimuli. Using this information and that from experiments on visual ability and foraging rates, I constructed a conceptual tradeoff model to describe the foraging behaviour of juvenile chinook salmon in turbid conditions. I also tested assumptions made by this model regarding the effects of turbidity on the risk perceived by juveniles in turbid waters. 4 CHAPTER 2 BACKGROUND AND HYPOTHESES Chapter 2.1 TURBIDITY AND VISION The teleost eye conforms to the general vertebrate plan, having an approximately spherical chamber containing an inverted retina, with outwardly facing receptors, and a focusable lens (for review: Guthrie 1986). Acclimation to different light intensities is accomplished by the movement of the outer segments of the receptors relative to the pigment layer (Brett and Ali 1958; Guthrie 1986). The sensitivity range of fish visual pigments is largely species specific and dependent upon the ambient wavelengths of light particular to the habitat (Munz and McFarland 1977). Individual species have been shown to possess different pigments in different habitats (Munz 1958; Levine and MacNichol 1979) and individuals may modify pigments over a period of weeKs to new habitat conditions (Muntz and Wainwright 1978). However, most fish apparently exhibit their highest sensitivity to the ambient light conditions of their natal habitat (Guthrie 1986). Visual ability in fish is commonly measured behaviourally as the reaction distance - the distance at which a foraging fish reacts to a prey item. Visual ability however, is a function of both the visual acuity of the forager and the scattering properties of the medium. Visual acuity is expressed as the minimum detectable angle subtended by a target on the retina of the subject (Tamura 1957). The scattering properties of media are defined by the scattering or attenuation coefficient (a) as the proportion of a collimated light beam that is absorbed or scattered within one meter (Duntley 1963). The principal effect of scatter is to reduce the visual contrast of a target with its background. Much work has been done on the visual acuity of fish (Tamura 1957; Hairston et al. 1982; Li et al. 1985) or on emphasizing 5 its relevance to foraging (O'Brien et al. 1976; Mills et al. 1984). Little research on the effect of contrast on fish visual perception has been performed (Hester 1968). However, it is widely recognized that contrast is as influential as acuity on visual perception (Duntley 1963; Hemmings 1966 and 1975; Lythgoe 1966 and 1980; Munz and McFarland 1977; McFarland 1986), especially in turbid water (Duntley 1943; DiToro 1978). Turbidity primarily a f fec ts the resolution of visual contrast. Chapter 2.2 TURBIDITY AND FORAGING ABILITY Evolutionary adaptations of fish species living in turbid waters include: sensory barbels, cutaneous sense organs, extensive cephalic-sensory networks, and electric, olfactory, and acoustic receptors (Miller 1979; Bruton 1985). Chinook salmon exhibit no evidence of these, although the use of olfactory cues in migration has been well documented (Northcote 1984). Although juvenile sockeye salmon (Oncorhynchus nerka) have been demonstrated to respond to food extracts by initiating search behaviour (McBride et al. 1963), effective feeding could only be1" accomplished visually. Juvenile chinook salmon exhibited foraging rates 300 times lower in near-dark conditions as those in well-lit conditions (Gregory unpublished data). Data for juvenile sockeye (Brett and Groot 1963) suggest similar effects of light. Chinook salmon primarily sense prey visually. I apply this observation as a working p r i n c ip l e . Of the various components of feeding strategy (for review: Hyatt 1979), prey encounter rate is likely to be the most directly influenced by turbidity, although prey capture rate may also be affected. Reduced reaction distance in turbid conditions has been demonstrated repeatedly (Vinyard and O'Brien 1976; Confer et al. 1981; Crowl 1989). Also, work on prey avoidance responses (Drenner et al. 1978; Vinyard 1980; Eggers 6 1982) suggests that higher probabilities of prey escape may be associated with turbid water, potentially decreasing the capture rate. The effect of turbidity on these rates is likely to have a negative impact on foraging ability. In order to offset this decrease in ability, fish must alter some aspect of their foraging behaviour if they are to maintain their rate of food intake. Juvenile chinook salmon, with their generalized feeding habits (Dunford 1975; Keast 1979; Levy et al. 1979; Healey 1982; Simenstad et al. 1982; Gregory pers. obs.), may demonstrate the required behavioural flexibility to alter their foraging behaviour and feed effectively in turbid waters. Juvenile chinook exhibit three general modes of feeding: benthic, planktonic, or surface foraging. Each of these involves different rewards, risks, and required search behaviour. All may be modified in turbid conditions. Feeding at the surface will result in the taking of predominantly insect prey trapped on the surface tension. Near-surface waters will possess the highest illumination, possibly the least affected by turbidity. Feeding on plankton will be affected by depth specific light conditions. Here, the visual dynamics imposed by contrast in turbid conditions are particularly relevant. Foraging on benthic organisms will become progressively more difficult with increasing turbidity, due to the reduction in light intensity. All three general modes of foraging are dissimilarly affected by the presence of turbidity. Changes in foraging behaviour commensurate with turbidity or light level have been demonstrated for fish in several studies. Threadfin shad (Dorosoma petenense) change foraging strategy from particulate feeding to filter feeding at low light levels (Holanov and Tash 1978). Benthically feeding largemouth bass (Micropterus salmoides) foraging in clear water strike only at prey of particular shape, orientation, and movement pattern (Crowl 1989). In turbid conditions, this strategy is largely abandoned; bass may respond simply to target size. It has been 7 suggested that rainbow smelt (Osmerus mordax) in Lake Superior respond to an increase in turbidity by moving to the surface, although they may have been following zooplankton which were rising to the surface (Swenson 1978). Boehlert and Morgan (1985) demonstrated higher feeding rates in turbid conditions for Pacific herring larvae (Clupea harengus  pallasi). They suggested that turbidity enhanced visual contrast, silhouetting the rot i fer prey against a uniformly illuminated background. These studies all suggest that foraging fish may alter their feeding strategy as turbidity increases. Chapter 2.3 TURBIDITY AND PREDATION RISK For juvenile salmonids, predators are of two main types, avian and fish piscivores. The most notable of these include: diving birds, gulls, herons, squawfish, sculpins, and other salmonids (Scott and Crossman 1973; Ginetz and Larkin 1976; Simenstad et al. 1982; Mace 1983). With the exception of the ambush predators (e.g. sculpins, herons), all are "active" hunters. Primarily, most predators on fish use vision to detect and attack prey (Hobson 1979). Any environmental condition that affects the visual abilities of a foraging fish is also likely to affect those of one or more of its predators. Juvenile salmon have been shown to be sensitive to predation risk (Dill 1983; Dill and Fraser 1984; Magnhagen 1988). Fry move shorter distances from cover on foraging bouts as this risk increases (Dill and Fraser 1984) or as their hunger level declines (Magnhagen 1988). Fish susceptible to predation often occupy more structurally complex habitats (Aggus and Elliott 1975; Werner 1977; Cooper and Crowder 1979; Savino and Stein 1982; Werner et al. 1983; Gotceitas and Colgan 1989) foraging within or near cover. Fish so positioned may easily be able to flee to 8 safety should the need arise. Reduced light levels have been linked to reduced predation risk (Woodhead 1956; Girsa 1973; Cerri 1983). The behaviour of juvenile salmonids indicates there is a risk avoidance component to activity patterns in various light conditions (Hoar et al. 1957; McCrimmon and Kwain 1966). Low light levels have also been implied to act as a form of cover in explanations of vertical migration behaviour in lake resident juvenile sockeye salmon (Levy 1987). Levy (1989) also suggests that vertical migrants may be tracking food resources. Work on coral reef communities has suggested that light itself may not be a universally good indicator of risk in fishes, whose responses are species specific and highly variable (McFarland and Munz 1975; Munz and McFarland 1977). By acting as cover, elevated turbidity levels may reduce predation pressure on adult (Bruton 1979) and young fish (White 1936; Blaber and Blaber 1980; Gradall and Swenson 1982; Healey 1982; Simenstad et al. 1982; Bruton 1985). Fish in turbid water may be able to evade detection or capture by a predator more effectively than in clear water. It has been suggested that salmonids are able to assess predation risk in clear visibility (Dill 1983 and 1987; Dill and Fraser 1984; Magnhagen 1988). They may also modify their behaviour in turbid waters to reflect a reduction in their perception of risk to predators. Chapter 2.4 FORAGING UNDER RISK OF PREDATION The importance of avoiding predation while foraging cannot be over-emphasized. Fish have evolved many, often complex, mechanisms to avoid being eaten (Cooper and Crowder 1979; Hobson 1979; Larkin 1979; Dill 1983 and 1987; Werner and Gilliam 1984; Milinski 1986; Mittelbach and Chesson 1987). Two basic predation avoiding strategies exist. The 9 first is to grow as quickly as possible beyond the size ranges most susceptible to predation (Werner and Gilliam 1984; Miller et al. 1988); the second is to adopt one feeding strategy or habitat early in life and change this with ontogeny (Werner and Gilliam 1984; Gilliam and Fraser 1987). Salmonids of the genus Oncorhynchus follow both. During their downstream migration, subyearling smolt and fry remain for variable amounts of time in estuarine habitats (Levy et al. 1979; Healey 1982; Levy and Northcote 1982; Simenstad et al. 1982), while the yearling smolt generally pass through the inner estuary (Healey 1982). The duration of estuarine residency is species and stock specific, ranging from days (pink salmon - O. gorbuscha) to months or years (chinook salmon). In the turbid Fraser Estuary, chinook migrants may quadruple their weight before they leave the tidal marshes after two months (Levy et al. 1979; Levy and Northcote 1982). Juvenile chinook salmon clearly achieve impressive growth rates in the Fraser Estuary. That chinook achieve these rates in turbid conditions is of ecological significance for a species commonly associated with more pristine clear water environments (Scott and Crossman 1973). The role of predation risk in determining these growth rates has not been elucidated. The problem of predator avoidance is universal among foraging animals. Almost every animal is the potential prey of another. As the very act of foraging potentially attracts the attention of a predator (Donnelly and Dill 1984), there may exist a tradeoff between the costs and benefits of foraging (for review: Lima and Dill 1990). The costs can often be expressed as the energetic e f for t of sequestering a food item and the risk to predators while doing so. The benefits can be expressed as the reward (usually expressed in energetic terms) of obtaining the food. Clearly, it may benefit a foraging animal in some instances simply to not forage at all (at least temporarily) if the costs are too high in terms of risk of predation. For fish, these tradeoffs have been well described in the literature (Dill 1983 and 10 1987; Werner and Gilliam 1984). But a vast literature also exists for other animals of a wide number of taxa representing both vertebrates and invertebrates (for review: Lima and Dill 1990). Chapter 2.5 HYPOTHESES AND EXPERIMENTS CONDUCTED Visual Ability It has been often stated or implied that turbidity reduces the visual foraging ability of searching fish (Vinyard and O'Brien 1976; Confer et al. 1978; Gardner 1981). Furthermore, these reductions occur on all prey types regardless of microhabitat (Minello et al. 1987; Crowl 1989). I have broken the above hypothesis concerning foraging ability into two components: visual ability and foraging rate. For planktonic prey, I investigated a behaviourally manifested surrogate of visual ability - reaction distance - as a function of turbidity. I assessed foraging rate of juvenile chinook salmon on prey in three generalized microhabitats: surface, water column (plankton), and bottom. In a further series of experiments, I investigated the effect of reduced light intensity expressed concurrently with increases in turbidity level. The effect of light on foraging rates and reaction distances is well established (Harden Jones 1956; Brett and Groot 1963; Vinyard and O'Brien 1976; Confer et al. 1978). In any such investigation involving turbidity, its ef fect on light level cannot be ignored. These studies are reported in Chapter 4.3. 11 Alteration of Search Behaviour The search behaviour of fish may be altered in the presence of elevated turbidity. In turbid conditions, fish may move into . microhabitats sparsely occupied in clear conditions (Bruton 1979; Swenson and Matson 1982) or adopt different foraging patterns (Holanov and Tash 1978; Crowl 1989). Simple alterations in various foraging rates may be insufficient to explain any exhibited changes in foraging behaviour by fish in turbid waters. Changes in microhabitat use may also occur. I have not addressed this hypothesis directly. The type of experiments required to fully describe all the potential forms of altered search behaviour would involve studies well beyond the scope of this thesis. However, I have conducted observations on the vertical and horizontal spatial distribution of juvenile chinook salmon in the laboratory over a range of turbidity conditions. These experiments (Ch. 5.1) provide information on microscale distributions of juvenile salmon. Enhancement of Visual Contrast To explain increased feeding rates by Pacific herring larvae in turbid conditions, Boehlert and Morgan (1985) hypothesized that rotifer prey were contrasted against a uniformly illuminated background. This hypothesis was examined in a study by Godin and Gregory (in prep.). The reaction time of juvenile chinook salmon for planktonic prey changed in various manipulations of prey colour and turbidity induced background contrast. Boehlert and Morgan's hypothesis cannot be rejected for planktonic prey. However, contrast will not be modified by increases in turbidity for surface prey and only for benthic prey if the colour of the bottom itself is altered. Although this phenomenon may be deemed 12 important for planktonic prey, it can be effectively eliminated from consideration for surface and benthic prey. Predation Risk Little hard evidence exists that predation risk declines with increasing turbidity. Gradall and Swenson (1982) show that salmonids reduce their use of overhead cover in elevated turbidity and Bruton (1979) correlates tilapia (Oreochromis mossambicus) survival and growth with the presence of turbidity. No controlled investigation of fish response to predator exposure has been documented for turbid and clear waters. I explore two hypotheses on predation risk in turbid water. First, does turbidity affect the response of juvenile chinook to the presence of a predator? I assess this question with the use of models of two general salmonid predators: a bird and a fish. Second, does turbidity - affect the post-exposure duration of the previous response? These questions are addressed in Chapter 5. The Ability-Risk Tradeoff Documented evidence of behavioural tradeoffs between the costs and benefits of foraging in the presence of varying degrees of predation risk has been growing in recent years (for review: Lima and Dill 1990). Turbidity represents an environmental variable which simultaneously alters the costs (predation risk) and the benefits (food rewards) of foraging behaviour. The existence of a tradeoff cannot easily be dismissed. In Chapter 6, I construct a conceptual model describing a tradeoff 13 between visual ability and perceived risk. I use this model to "explain" the foraging rates observed in Chapter 4.2, in terms of visual ability and perceived risk. 1 then make predictions concerning the inherent risk of foraging in various microhabitats and the effect of increasing risk stimuli on foraging rate. By testing these predictions in Chapter 7, I attempt to support the assumptions concerning risk perception made by the conceptual tradeoff model. 14 CHAPTER 3 GENERAL METHODS Chapter 3.1 THE STUDY ANIMAL The study animals used were juveniles of chinook salmon (Oncorhynchus tshawytscha [Walbaum]) from Harrison River, a tributary of the lower Fraser River. Harrison chinook are predominantly "ocean-type", beginning their seaward migration as underyearling fry (Shepherd et al. 1986). They are suspected of spending several months rearing in tidal channels of the Fraser Estuary after leaving their natal stream. As juveniles, these fish must traverse »100 km of the lower Fraser River, where I have measured turbidity levels as high as 400 mg-L-1. -Therefore, the fish stock I used throughout my work probably has a well established prior history of exposure to turbid conditions. Individuals were obtained as 0.8 g fry from the Chehalis River Hatchery (on a tributary of the Harrison River) and transferred to holding facilities at the University of British Columbia. Chinook fry were held in one of two 1000-L holding tanks at 6.0 - 10.5°C. Initial fish densities were 2.5-L-1. Densities declined as fry were used in experiments and were within accepted guidelines for salmon culture. The light regime of 14:10 (light:dark hours) was set by automatic timer; the lights were turned on at 0800 h. The water supply to the facility was treated with 5 - 2 0 mg*L~1 sodium thiosulphate concentrations to absorb free chlorine. Fry were fed "Oregon Moist Pellets" (OMP) twice daily at 1000 and 2000 h, at a rate of 38X of the ration required to achieve maximum growth rate in hatcheries (Kahl, Chehalis River Hatchery, pers. comm.). Growth rate was maintained at approximately 10'/. (weight-week - 1). The individual fish used in experiments varied in size largely as a function of their growth throughout the duration of the experimental 15 "season". Chinook generally ranged in size from 40 to 75 mm fork length (FL). During any particular experimental run and for most whole experiments, fish size was kept reasonably similar (+3 mm SD). Sizes of juveniles in the Fraser Estuary ranged from 35 - 75 mm FL (Levy and Northcote 1982; Gregory pers. obs.). The seasonal range of fish size created some analytical difficulties during data interpretation. However, the opportunity to examine some behavioural changes with ontogeny was also presented. These have been described in Chapter 4.2. Chapter 3.2 TURBIDITY Sediment for all laboratory procedures in this study was obtained from a tidal marsh in the south arm of the Fraser River Estuary, at Ladner, British Columbia. Approximately 10 L of this sediment was sieved through a 0.40 mm sieve, to remove larger detritus, then suspended in 125 L of freshwater in a plastic bucket and allowed to settle for a period of 2 h. After this time, the supernatant (75 L) was transferred to another bucket and allowed to settle for a further 48 h. The excess water was then poured off, and the remaining sediment slurry was autoclaved for 30 minutes. I repeated this procedure until the required quantity of slurry was obtained. The individual particles were classified as subrounded, using the criterion of Muller (1975). Particle sizes ranged from <2 to 25 pm diameter (s90X <5 pm). 16 Chapter 3.3 EXPERIMENTS ON FORAGING RATE Prey Three types of prey were used in experiments investigating chinook foraging in three generalized microhabitats. These were broadly categorized as surface, planktonic, and benthic prey. A prey animal of each of the three types was selected to fit several subjective criteria: easily obtainable, easily maintained under laboratory conditions, similar in size to the other prey used, non-evasive, low intraspecific variance in size, similar in colour, and representative of the type of prey encountered by chinook in the field. The three prey chosen were: surface - adult Drosophila melanogaster, planktonic - adult Artemia  salina, and benthic - Tubif ex sp. The general characteristics of each of these prey animals may be found in Table 3.1. No changes in behaviour relative to turbidity treatment, were noted for these prey. Artemia and Tubifex were obtained from a local retail outlet, and winged, wild-strain Drosophila were easily cultured in the laboratory. Preliminary comparisons between living and frozen Drosophila indicated no preferences by f ry for either of the two forms, therefore the frozen form was used as they could be easily stored for later use. Both Drosophila and Tubifex were readily identifiable prey types of chinook fry, with the latter being a demonstrated constituent of the diet of wild individuals (Gregory pers. obs.). Drosophila have not characteristically been found in the diet of wild caught fish, but was acceptable because chinook tend to be relatively indiscriminant in their selection of surface prey. Drosophila are within the size range of surface prey taken by wild fry. Artemia were intended to represent zooplanktonic prey. Although they are much larger than the zooplankton encountered by f ry in freshwater, they are not much larger than those encountered by slightly larger conspecifics in the marine environment 17 Table 3.1 Physical and visual characteristics of prey used in foraging rate determinations. Prey Length Weight mm3 mg(wet) Other D Character i st i cs Drosoph i1 a 2.5+0.3 0.9 black & white with red eyes; winged; previously frozen Artemi a 6.7+1.1 5.3 white to pink with red/black eyes; weak swirrmers (no developed escape response) Tubifex 14.8+7.0 1.9 dark red/black; burrowing; weakly evasive (no effective escape - but longer handling time than other two prey) a mean+SD (minimum N=300) D batch weighings; mean weight/individual only 18 and were much smaller, and less evasive, than Neomysis mercedis a mysid shrimp periodically taken by juvenile chinook in the Fraser Estuary (Northcote et al. 1979; Gregory pers. obs.). I established prey densities and feeding trial durations for each prey type in a series of preliminary investigations. A "working" Artemia density and trial duration was identified by feeding fry to satiation with an overabundance of prey to minimize the effects of search time (Holling 1959). I set the trial duration to be half the satiation time (i.e. 1.0 minute). I then set the trial prey density to be double the number of prey consumed-min-1 times 10 (10 fish used in a trial) in 64 Liters. The resulting initial prey density was s 3.5 Artemia'L"1, not outside the range of planktonic prey densities these fish would encounter in the field. For Tubifex, both working density and trial duration were identified in a similar manner as for Artemia. However, maximum feeding rates for this prey species occurred at elevated turbidity levels where individual feeding incidents could not be observed. Therefore, I observed satiation in clear water to obtain subsequent trial prey densities and estimated trial duration from a number of preliminary trials of differing duration at this prey density at 100 mg*L~1 turbidity level. As a result of these preliminary trials, I set initial prey density and trial duration for Tubifex at s 5600TTT2 and 5.0 minutes, respectively. Although availability continues to be a difficult area in both field and laboratory studies dealing with benthic prey, the densities used were well within natural levels of many prey species of f ry in the field (Dunford 1975; Kistritz 1978; Gregory pers. obs.). I set the trial density and duration for Drosophila in a manner similar to that for Artemia. As for Tubifex prey, maximum feeding in preliminary trials occurred at elevated turbidity levels, but individual feeding incidents were visible because the prey were taken from the 19 surface. Consequently, setting trial prey density and duration was comparatively easy; these were set at 1300 Drosophila-m^ and 10.0 minutes, respectively. The prey density for Drosophila far exceed instantaneous prey densities likely to be encountered in the field, but they seemed reasonable given the densities of surface prey I have observed in the field (a 6870'm~2,d~1). Given the search path (Keast 1979) covered by cruising chinook (8.8t2.1 rrrmin-1 [pers. obs.) x 0.656 m [2 x reaction distance in clear water, from Fig. 4.2] x 10 min) and prey densities in the field, I felt the prey densities presented experimentally were reasonable. The above prey densities have been expressed as individuals per unit area or volume. I measured prey by weight, to expedite experimental handling. These weights were predetermined to be 1.25 g Artemia [wet weight], 2.45 g Tubifex [wet weight], and 0.25 g Drosophila for each treatment aquarium and were found to be accurate within «3'/. of the actual numbers desired. Pre-Experimental Conditioning Three days prior to any given experimental trial, I transferred appropriate numbers of chinook (usually 72) from the main holding tanks to one of eight compartments in several 200-L "conditioning" tanks. In these tanks, fish were fed 30/. of the usual amount of "OMP"; feeding was augmented with approximately equal numerical amounts of all three prey types. Test fish were fed this altered diet for three days before experimentation. Chinook were allowed to evacuate guts 18 hours prior to any given experimental trial. In preliminary trials, I exposed chinook to one of three levels of turbidity (0, 50, and 200 mg-L"1) for a n period of 7 days prior to experimentation. The effect of prior exposure to turbid conditions was not significant (ANOVA; df=2,18; p»0.25), regardless of prey type. 20 Therefore, chinook were not exposed to turbidity as part of the , conditioning procedure. Experimental Apparatus The experimental array (Fig. 3.1) consisted of 7, 70-L, glass aquaria enclosed and separated from each other by 8 mm plywood. All inwardly-facing plywood surfaces were painted flat white, and were periodically repainted to maintain optical consistency. Lighting was provided by twin 40 watt fluorescent light fixtures running the full length of the array, 45 cm above the water surface. The entire array was enclosed in opaque black plastic sheeting such that it was isolated from outside visual disturbances and light sources. Each aquarium was similarly separated from the other aquaria in the array and from extraneous disturbances by additional sheeting. Each 70-L aquarium (50 x 34 x 40 cm) was fitted with a false bottom, through which a mud slurry (as previously decribed) was visible. Given that small amounts of sediment settled out of suspension at high turbidity levels (particularly 1400 mg^L-1), the potentially confounding effects of altered background contrast (Godin and Gregory, in prep.) were avoided. For substrate, I provided a 5 - 7 mm layer of clear 2 mm diameter glass beads. This allowed the Tubifex to burrow (in appropriate experiments) and also permitted the constant background slurry in the false bottom to remain visible. A 2.5 cm clear syphon hose was clamped in place at one end of each aquarium. Each syphon hose was primed prior to each experimental run before fry were transferred to the aquaria, and could be activated from the front of the array. An aquarium would take approximately 3 minutes to drain in this manner. I positioned an airstone in the middle of each aquarium to maintain the sediment suspension. 21 Figure 3.1. Aquarium array used for determinations of foraging rate. A. Schematic representation. B. Photograph. 22 Experimental Protocol Ten fish were placed in each aquarium. I added a measured amount of sediment to each of six aquaria to create turbidity levels of 25 -800 mg-L-1 in a geometric series (25 x 2 n ; where n = 0 - 5). No sediment was added to the seventh aquarium; which functioned as the control. I randomized the order of the treatments and the control within the experimental array prior to each experimental run. Chinook were allowed to acclimate to the experimental conditions for a period 1 - 1.5 hrs. before prey were added to the aquaria. Measured amounts of prey were added to each aquarium in order of its position within the array. Only one prey type was used during an individual run. Chinook were allowed to feed for an uninterrupted period specific to the prey animal being tested. For experiments using Tubifex prey, I allowed the test fish to acclimate in 4-L meshed plastic holding chambers, within each aquarium. I added prey to the main body of the aquarium ten minutes before beginning the trial. In preliminary trials, Tubifex were observed to commence burrowing into the bead substrate within one minute of introduction to aquaria. After the appropriate experimental time period had elapsed for an individual aquarium, its syphon was activated and the tank was permitted to drain for 3 minutes. Fish were then netted and transferred to a 500 ml container with a lethal concentration of methanesulfonate (MS222) anaesthetic. Fry remained in this container for a period of approximately 10 minutes before I transferred them to a sample jar with 5/. formalin. For Artemia prey, I began netting immediately after the one minute experimental feeding period had elapsed. I used an aquarium electroshocker to facilitate retrieval of fish in this case. The latter technique caused the cessation of all feeding, but was not so severe as to cause regurgitation in the chinook. These techniques were used 23 either on each tank in turn (Artemia trials) or were staggered such that several tanks could be operated simultaneously, at various stages of completion (Drosophila and Tubifex trials). Stomach contents were dissected and enumerated for Drosophila and Artemia trials. Contents containing Tubifex were blotted dry and weighed to the nearest mg. Turbidity levels were checked by taking water samples while the tanks were being drained. The water samples were analyzed with a Fisher 400 DRT Turbidimeter, measuring nephelometric turbidity units (NTUs). Light levels were measured with a Li-Cor 185-A light meter with a quantum sensor. Illumination at the water surface was s16.5 pE'rrr^-s-1 (s850 Lux; i.e. the light intensity of a dull day). Light levels were measured at the tank bottom and are reported along with NTUs in Figure 3.2. Chapter 3.4 SPATIAL DISTRIBUTION AND PREDATOR MODELS Experimental Apparatus All experiments in Chapter 5 were performed using a 1000 L (180 cm x 100 cm x 55 cm) arena, one wall of which was fitted with a 180 cm x 55 cm) clear plexiglass observation window (Fig. 3.3). The tank was fitted with a 25° slope running the width of the tank from the tank bottom to the water surface. Therefore, water depth ranged from 0 to 40 cm. A mixture of gravel covered with a layer of Fraser Estuary mud was used as benthic substrate. A screen of 5 mm green plastic mesh was positioned 40 cm from the observation window, to restrict the lateral movement of fish in the water column. The observation window was marked in a grid of 6 regions. I established these numbered regions in a manner 24 Figure 3.2. The relationship of turbidity concentration (mg*L~1) used in foraging rate experiments with measures of light transmission. A. Nephelometric Turbidity Units (NTUs). B. Light energy (pE'm~2,s~1; measured at the tank bottom [37.5 cm depth]; 1 pE'm"2*s~1, » 51.2 Lux). 0 0.5 1 1.5 2 2.5 3 Log(Turbidity[mg/L]) B loo(LIGHT) = 1.045 - 0.0108 • (TURB) r2=> 0.999 (N=5) j | u 0 50 100 150 200 Turbidity (mg/L) 25 Figure 3.3. Arena used during experiments on spatial distribution and predation risk manipulation. A. Schematic representation (lines labeled B - B' and F - F' indicate the track of the bird and fish predator models, respectively; see text, Ch. 5.2 fo r details). B. Photograph. 26 facilitating the subsequent analyses of variously grouped census data (e.g. by proximity to the surface, bottom, etc.). The entire apparatus was surrounded by black plastic curtain, restricting disturbances. In the area in front of the observation window, an additional curtain was erected to both eliminate the entry of ambient light through the observation window and to enclose the observer. Variously positioned 3 x 10 cm slits in the inner curtain facilitated observation. These were covered when not in immediate use. Lighting was provided from above by double fixtured fluorescent lights running the length of the apparatus, 60 cm above the water surface. Light levels ranged from near 0 to 10.00 pE*m~2 ,s~1 depending on turbidity and depth (Fig. 3.4). The turbidity restr icted the visibility of fish through the observation window. Preliminary trials with inanimate objects indicated there was no appreciable change in visibility as a function of depth over the range of turbidities tested. Although fish counts were lower in turbid conditions due to their visibility to the observer, I suspected no differential bias in these data with respect to region of the arena. Pre-Experimental Conditioning Fish were fed twice each day at 1000 and 2000 hours throughout the entire course of experiments. Drosophila, Artemia, and Tubifex were again used as prey and were supplied "ad libitum". The latter two prey types were introduced to the arena in each of the three depth regions of the arena via glass tubes near the middle or bottom of the water column, respectively. I added planktonic and benthic prey to each region in proportion to the relative volume or area of that region of the tank. Surface prey were added to the surface in a similar manner. As part of the conditioning phase, I exposed fish to all subsequently used 27 Figure 3.4. The effect of depth and turbidity concentration (mg-L-1) on light level (pETn" 2*s _ 1), within the experimental arena (Fig. 3.3). 5 10 15 20 25 30 Depth (cm) 28 treatment turbidity levels for at least one day, following an initial two days of clear water. In all respects, treatment of the fish in this period was identical to that during observation days. Daily, prior to the lights coming on in the apparatus at 0800 h, the water supply to the arena was turned of f and I added the required amount of suspended sediment slurry to establish the appropriate turbidity level. The presence of three airstones provided circulation within the tank and maintained the suspension of added sediment. Water flow to the arena was restored following the last observation period. Water samples and temperature readings were taken after each observation period; I added more sediment slurry at this time if required. Chapter 3.5 S T A T I S T I C S Statistics generally consisted of well described procedures such as ANOVA, G-test, linear regression, and various summary statistics (Zar 1984). I used the procedures of SYSTAT 4.0 (Wilkinson 1988) throughout. Analysis involving foraging rates were generally restricted to the use of means of each treatment within a trial, rather than data from individual fish. In all cases, I dropped the highest and lowest data values from fish in each treatment of any given trial before calculating the mean. I have documented any changes to the above procedures in the appropriate Methods section. 29 CHAPTER 4 VISUAL ABILITY AND FORAGING BEHAVIOUR Chapter 4.1 REACTION DISTANCE - VISUAL ABILITY INTRODUCTION Visual ability has been variously measured in fish (acuity -Tamura 1957; Hairston et al. 1982; Li et al. 1985; contrast resolution -Hester 1968; behavioural measures - Vinyard and O'Brien 1976; Confer et al. 1978). In most instances, this ability is manifested behaviourally as the distance at which an animal reacts to objects in its environment; the term receiving common usage is reaction distance. Reaction distance is affected by several factors including light intensity (Vinyard and O'Brien 1976; Confer et al. 1978), prey size (Ware 1971 and 1973; Confer and Blades 1975; Vinyard and O'Brien 1976; Hairston et al. 1982), fish size (Hairston et al. 1982; Breck and Gitter 1983; Wanzenbock and Schiemer 1989), prey movement (Ware 1971 and 1973; Crowl 1989), prey and background colour or contrast (Ware 1971 and 1973; Godin and Gregory in prep.), and turbidity (for review: Bruton 1985). Turbidity deleteriously affects the visual perception of every factor in this list. Even any advantage in visual range gained by fish of larger size may be negated in turbid water. In this study, I address the effect of turbidity on reaction distance. Work on bluegill sunfish (Lepomis macrochirus - Vinyard and O'Brien 1976) and lake trout (Salvelinus namaycush - Confer et al. 1978) has demonstrated that reaction distance of foraging fish for planktonic prey is reduced by turbidity. Investigations by Crowl (1989) have shown similar reductions for largemouth bass (Micropterus  salmoides) foraging benthically. 30 Chinook salmon may occupy turbid estuaries for a significant portion of their early life (Levy et al. 1979; Levy and Northcote 1982; Simenstad et al. 1982). However, no quantitative work on the visual ability of Oncorhynchus spp. in turbid water has been published. I investigated the reaction distance of juvenile chinook salmon to planktonic prey in concentrations of suspended sediment ranging from 0 to 400 mg-L-1. These levels are common in the Fraser River estuary and other Pacific coast rivers. M E T H O D S Experimental Apparatus The experimental apparatus (Fig. 4.1) consisted of a 200x30x25 cm plexiglass tank. Test fish were further confined to the centermost 10 cm by two plexiglass partitions running the length of the aquarium and to the centermost 160 cm of the long axis by gravel partitions at either end. Water depth was maintained at 5 cm. The apparatus was placed off the floor with a mirror positioned at a 45° angle underneath to facilitate observations from a remote location. A videocamera was used to record observations. The plexiglass bottom was covered in a reflective acetate film, which allowed light to pass freely downward but not upward, and effectively acted as "one-way" glass. Therefore, fry were not disturbed by movement below the tank. Illumination of 16.2 pE'm~2 ,s~1 at the water surface was provided by doubled-fixtured 40-watt flourescent lighting running the full length of the aquarium. An airstone was positioned at each end of the tank, outside the view of the test subject. The entire apparatus was enclosed in black opaque plastic sheeting to limit outside disturbances. Water was changed once each 31 Figure 4.1. Schematic representation of the experimental apparatus used fo r the determination of reaction distances. water/turbldlty 32 day, af ter observations. Conditioning and Observational Protocol A single juvenile chinook was placed in the apparatus and over a period of approximately two weeks, conditioned to strike at single prey items. Prey (live, 7-8 mm, male Artemia) were introduced at a constant release point 40 cm from the gravel partition at one end of the tank by means of a 4.0 mm diameter glass tube. The observer was not visible to the subject during this process. Subjects were required to strike the glass tube containing the prey to facilitate its release into the aquarium, where it was promptly consumed. This technique ultimately permitted complete verification of a successful "strike" and location of prey in elevated turbidity conditions (to 400 mg-L""1, « 275 NTUs). Subsequent analysis of reaction distance required only measurement from the assessed reaction point to the fixed prey location. The former was identified as a the initiation of a distinct, rapid, tail beat followed by a constant acceleration toward the prey in the tube, ending with a strike. The prey tube was left in position between observation periods. An observation was considered invalid if: 1. the tube (with prey) was introduced after the test fish had oriented toward the prey location; 2. the tube was introduced while the subject was within 60 cm of the prey location; 3. prey had moved within the tube to a position outside 1.0+0.5 cm above the tank bottom, or; 4. if the fish "hesitated" after an initial reaction to the prey. During experimental runs, prey were introduced in this manner with true introductions being interspersed with 1 - 4 "blanks". These were introductions containing no prey. No strikes were recorded in over 1000 "blank" runs. Chinook responded only to the presence of Artemia in the prey tube. Predetermined amounts of sediment slurry were added two hours 33 before observations to provide the turbidity treatment. Turbidity levels were randomized with respect to trial day, with only one level being tested on any single day. The first and last test date for any one fish were clear water treatments to determine any changes in fish response during the duration of the experiment. I observed none. The three chinook used in this study were 64, 65, and, 70 mm FL and 2.62, 3.63, and 3.82 g, respectively. Experiments were run at ambient temperatures of 13 - 17°C with temperature never varying by more than 1.5°C for any given individual. R E S U L T S A total of 216 separate reaction distance determinations were made under the above procedure at seven turbidity levels from <1.5 to 400 mg-L-1. The effect of turbidity on the median reaction distance of juvenile chinook salmon was best described by the log-linear re la t ionship : RD = 32.85 - 11.78 x log T, where, T = turbidity (in mg-L-1) and RD = reaction distance (in cm) (Fig. 4.2). An r 2 of 0.96 among the medians supported a high degree of confidence for this relationship (r2=0.72, when all data points were taken separately). I observed no marked dissimilarities in the reaction distance determinations between individual subjects. Figure 4.2. The effect of turbidity on the reaction distance of juvenile chinook salmon for Artemia prey (regression is from 13 median values from 3 fish). 34 40 r 0 I 1 1 1 1 i i i i 0 50 100 150 200 250 300 350 400 Turbidity (mg/L) 35 D I S C U S S I O N The relationship demonstrated by the chinook was similar to that previously described by Vinyard and O'Brien (1976) for bluegill sunfish and by Confer et a I. (1978) for lake trout. However, the observed relationship did not exhibit the characteristic threshold type of response as demonstrated between light level and either foraging rate (Harden Jones 1956) or reaction distance (Vinyard and O'Brien 1976; Confer et al. 1978). This suggests that turbidity and light affect visual ability in a dissimilar manner. The probability of prey detection by fish has been shown to be proportional to the reaction distance (Ware 1973; Confer and Blades 1975; Hairston et al. 1982). Reaction distance declines log-linearly with turbidity. It is likely that the probability of prey detection will decline with increasing turbidity level. The results I have demonstrated for juvenile chinook salmon in this study were similar to results demonstrated for other species. Turbidity has a deleterious e f fec t on visual ability. Chapter 4.2 FORAGING RATES FOR SURFACE, PLANKTONIC, AND BENTHIC PREY INTRODUCTION Turbidity acts to reduce the visual ability of fish (Vinyard and O'Brien 1976; Confer et al. 1978; Chapter 4.1). Consequently, foraging rate should decline with increasing turbidity. The work of Gardner (1981), Johnston and Wildish (1982), Sigler et al. (1984), Berg and Northcote (1985), and Breitburg (1988) supports this conclusion. These 36 investigations demonstrated strong negative effects of turbidity on foraging rate in numerous fish species. In addition to these experimental studies, a large body of evidence from field investigations also exists purporting the negative effects of turbid water on visually foraging fish species (e.g. Ellis 1936; Buck 1956; Alabaster 1972; Noggle 1978; Hart 1986; Eccles 1986; McLeay et al. 1987; Simenstad 1990). However, several studies support a more benign view of turbid conditions. Boehlert and Morgan (1985) depart from the established wisdom, reporting that Pacific herring larvae (Clupea harengus pallasi) increased their foraging rates in moderately turbid conditions (500 -1000 mg-L-1) over clear water controls in laboratory experiments. Also in the laboratory, Cyrus and Blaber (1987b) demonstrated that juveniles of several estuarine species actively prefer turbid over clear water. Numerous field investigations also suggest that foraging ability may not be severely impeded by turbidity or may be offset by advantages concommitent with turbid habitats (Blaber and Blaber 1980; Levy and Northcote 1982; Stone and Daborn 1987; Cyrus and Blaber 1987a). Chinook salmon can be best described as generalist foragers (Keast 1979) possessing relatively unspecialized feeding morphologies and search habits (Scott and Crossman 1973). This species has been observed to feed on a variety of prey types from assorted microhabitats (Dunford 1975; Levy et al. 1979; Healey 1982, Simenstad et al. 1982; Gregory pers. obs.). The effect of turbidity on foraging behaviour in these microhabitats may be dissimilar. Such studies have not been conducted. The majority of experimental manipulations regarding turbid waters have examined planktonic prey. Crowl's (1989) investigation of largemouth bass (Micropterus salmoides) reaction distance to crayfish and Moore and Moore's (1976) study of prey selectivity by flounder (Platichthys flesus) represent important exceptions. Ware (1973) also made some predictions of potential consequences of turbidity on epibenthic feeding 37 rates. These have gone largely untested. In the present study, I examine the effect of turbidity on foraging rate by juvenile chinooK salmon in three generalized microhabitats. Surface, planktonic, and benthic foraging is investigated for Drosophila, Artemia, and Tubifex prey, respectively. In the performance of various manipulations described in this and other subchapters, I have used chinook salmon ranging in mean size from 47 to 69 mm FL. Although the size disparity did create some anaytical difficulties, it also presented an opportunity to examine the effects of ontogeny on foraging behaviour in turbid conditions. Therefore, in this subchapter I present exploratory analyses of the ontogenetic changes in foraging behaviour. METHODS The Effect of Turbidity on Foraging Rate Using the methods described in Chapter 3, I conducted experiments determining the foraging rate of juvenile chinook salmon for the three above mentioned prey types. These experiments were conducted over two years, 1987 and 1988. For each prey type, I performed five and three trials at all turbidity levels in each of these years, respectively. Chinook mean size by trial ranged from 59.1 to 69.6 mm FL in 1987 and 52.5 to 58.3 mm FL in 1988. Individual sizes within each trial ranged within +3 mm FL for any given trial mean. 38 The Effects of Ontogeny on Foraging Rate in Turbid Conditions To analyze the ontogenic aspects of chinook behaviour in foraging trials, the data from all foraging rate experiments were collectively examined, for each of the prey types. Provided the experimental trial in question fit the "methods" criteria used in this subchapter (i.e. it was a "control" run of the particular experiment), I deemed it suitable for the analyses described here. The continuity of the methodology used across the individual studies (Ch. 3.3) made this possible. The above criteria were met by 14, 15, and 13 trials for surface, planktonic, and benthic prey, respectively. I f irst determined the slope of the relationship between turbidity and foraging rate, for each of the trials, for the four 0 to 100 mg-L-1 treatments (i.e. the ascending limb of many of the described relationships - see Results). I used the logarithm of (turbidity + 1) in calculations to avoid disproportionately weighting the determination of slope by the data from the 0 and 100 mg-L-1 turbidity treatment levels. I then performed a least squares analysis on the resulting slopes, and the mean size (mm FL). I conducted this analysis separately for each prey type. R E S U L T S Surface Foraging For juvenile chinook foraging on Drosophila prey, turbidity significantly affected foraging rate (ANOVA df=6,42j p=0.029 - Table 4.1A). Highest foraging rates were attained in intermediate turbidity treatments (50 - 200 mg*L~1). These rates were 10.7 and 7.7 preyfish" (in 10 minutes) in 1987 and 1988, respectively (Fig. 4.3). Chinook 39 Table 4.1. Analysis of variance of the e f fec t of turbidity and year of experiment on the foraging rate of juvenile chinook salmon to A. Surface prey (Drosophila), B. Planktonic prey (Artemia), and C. Benthic prey (Tubifex). A. Surface Prey Source of Var i at i on Sums of Squares DF F P Turb i d i ty 252.6 6 2.911 X Year 101.4 1 6.984 K Turbidity x Year 86.7 6 0.999 NS Error 607.4 42 K , » » , * * * - p<0.05, 0. 01, 0.001, respectively; NS - p>0.10. B. Planktonic Prey Source of Variation Sums of Squares DF F P Turb i d i ty 932.9 6 41.888 Kfttt Year 15.8 1 4.269 -Turbidity x Year 42.2 6 1.896 NS Error 155.9 42 «,**,**« - p<0.05, 0.01, 0.001, respect ively; NS - p>0.10. C. Benthic Prey Source of Variation Sums of Squares DF F P Turb i d i ty 2590.2 5 a 15.490 **» Year 1461.0 1 43.687 KMX Turbidity x Year 1100.8 5 6.583 KKK Error 1203.9 36 x,xx,xxx - p<0.05, 0.01, 0.001, respectively; NS - p>0.10. a analyses performed without the 25 mg'L" 1 treatment, which was not performed in 1987. 40 Figure 4.3. The ef fects of turbidity on mean foraging rate of juvenile chinook salmon feeding on surface prey and the percentage of salmon foraging in 70-L aquaria (Drosophila consumed in 10 minutes; vertical bars indicate standard error of the mean for each trial; N = 8 fish for each treatment level f o r each trial). A. 1987 - 5 trials. B. 1988 - 3 trials. 41 exhibited depressed foraging rates at O and 800 mg-L-1 treatment levels, significantly so in 1987 (t-test, p=0.016 and p=0.004, respectively). In 1988, chinook foraging rates exhibited trends consistent with those in 1987, although peak foraging rates were observed at different turbidities. Also, the peak was not statistically significant in 1988. Individual variability was high; some fish consumed 40 prey or more, while many others had empty guts. The effects of turbidity level on foraging rate were similarly described by the proportion of fish actively feeding (Fig. 4.3), in both years. I have reported the details of individual trials in the Appendices (Figs. A,1 and A.2). The observed relationship represented a marked departure from the foraging rate expected on the basis of the reaction distance of juvenile chinook in turbid conditions (Ch. 4.1, Fig. 4.2). Planktonic Foraging Among chinook foraging on planktonic Artemia prey, highest foraging rates (11.0 - 15.0 preyfish"1-min-1) were observed at low treatment levels (0 - 100 mg-L"1); reduced rates were demonstrated at high treatments (400 - 800 mg-L-1). The effect of turbidity was significant (Table 4.1B). These rates approached zero in the 800 rng'L"1 treatments. Foraging rates were not proportional to the percentage of fish actively foraging (Fig. 4.4). The 1987 and 1988 trials were not significantly different (Table 4.1B). In both years, the 800 mg-L-1 foraging rates were significantly lower than those of the 0 - 200 mg-L-1 turbidity levels (Tukey tests, p<0.001). Foraging on planktonic Artemia prey by chinook juveniles was more consistent than that demonstrated for surface foraging, with generally lower variance among and within means in both years (Appendix - Figs. A.6 and A.7 for details). The "high to low" foraging rate pattern exhibited by these fish 42 Figure 4.4. The ef fects of turbidity on mean foraging rate of juvenile chinook salmon feeding on planktonic prey and the percentage of salmon foraging in 70-L aquaria (Artemia consumed in 1 minute; vertical bars indicate standard error of the mean for each run; N = 8 fish for each treatment level f o r each trial). A. 1987 - 5 trials. B. 1988 - 3 trials. c 1 I" 12 co .c c cS <D 2 0 100 Eg " 0) cb CLLL « B Turbidity (mg/L) 400 800 200 Turbidity (mg/L) 400 800 43 was similar to the expectations based on the visual ability of these fish (Ch. 4.1, Fig. 4.2). The asymptotic aspects of the low turbidity end of the relationship could possibly be due to handling time restrictions within the 1.0 minute experimental period. Search time may not have limited the foraging rate at turbidity levels below 200 mg-L-1. Benthic Foraging Benthic foraging rates on Tubifex followed the same generalized pattern demonstrated by surface foraging chinook. Turbidity had a significant effect on foraging rates (ANOVA; df=5,36; p<0.001 - Table 4.1C). Highest foraging rates were demonstrated in the intermediate turbidities (25 - 200 mg-L-1); reduced foraging rates were observed at the lowest and highest treatment levels (Fig. 4.5). Although I found a general similarity in the form of the exhibited relationships between 1987 and 1988, the magnitude of the foraging rate peak was much greater in 1987. Peak mean foraging rates in the larger 1987 fish were >30 mg p r e y f i s h - 1 (in 5.0 minutes). Corresponding rates in 1988 were generally <10 mg. For 1987 trials, multiple comparisons revealed significant differences (Tukey test, p<0.01) in benthic foraging rate of chinook between intermediate turbidities (50 - 100 mg-L"1) and all other treatments. No similar differences could be demonstrated for the smaller 1988 fish, although the same general trend was exhibited. Foraging rates were loosely mirrored by the proportion of fish actively foraging at each treatment level (Fig. 4.5). Within and between trial variance was less than that exhibited for surface foraging, but was generally large (see Appendices - Fig. A.11-12 for detailed accounts). As for surface foraging, the foraging relationship I observed for benthic prey again represented a marked departure from the foraging rate predicted from the reaction distance of juvenile chinook in turbid 44 Figure 4.5. The ef fects of turbidity on mean foraging rate of juvenile chinook salmon feeding on benthic prey and the percentage of salmon foraging in 70-L aquaria (mg Tubifex consumed in 5 minutes; vertical bars indicate standard error of the mean for each run; N = 8 fish for each treatment level for each trial). A. 1987 - 5 trials. B. 1988 - 3 trials. 45 conditions (Ch. 4.1, Fig. 4.2). The Effect of Ontogeny and Turbidity on Foraging Behaviour For chinook foraging on benthic and surface prey, the effect of turbid water on foraging rate changed with ontogeny (p<0.01, Fig. 4.6). Although foraging rate for benthic and surface prey increased generally with fish size, this increase was less in clear water than in intermediate turbidity conditions. In many instances, I observed a decrease. At smaller fish sizes, turbidity seems to effect a reduction in the foraging rate, whereas larger fish exhibit an increase in foraging rate from low to intermediate suspended sediment levels (slOO mg-L"1). The chinook I used in the 1987 experimental trials were approximately 15 - 25/. larger (length) than their 1988 counterparts. As I have described in the preceeding sections, several differences existed between trials of these two groups of fish, during benthic and surface foraging experiments. Compared with the effects of fish size demonstrated in this subchapter, more pronounced departures from the described foraging rate - turbidity relationship were observed in several other experiments (e.g. Fig. 4.8 and 4.10). Similar ontogenetic effects were not demonstrated by chinook foraging on planktonic prey (Fig. 4.6B). However, changes in feeding behaviour on Artemia prey may have been masked by handling time effects. 46 Figure 4.6. E f fect of size of juvenile chinook salmon on the slope of the ascending limb (0 - 100 mg-L-1) of the relationship between foraging rate and turbidity (each point represents a unique run). A. Surface prey - Drosophila. B. Planktonic prey - Artemia. C. Benthic prey - Tubifex (data point identified by the arrow accounted for Al'/. of the variance and was omitted from regression analysis). 47 D I SCUSS ION Effect of Turbidity on Foraging Rate The visual ability of fish declines in turbid conditions (Vinyard and O'Brien 1976, Confer et al. 1978; Crowl 1989; Ch. 4.1). The general form of this evidence has been consistent between these studies; visual ability, measured as reaction distance, declines log-linearly with increasing turbidity. Many studies have also demonstrated that suspended sediment has a deleterious effect on foraging rate in fish (for review: Bruton 1985). However, many of these same studies did not demonstrate significant effects at low levels (Heimstra et al. 1969; Johnston and Wildish 1982; Breitburg 1988). One study (Boehlert and Morgan 1985) describes an increase in planktonic foraging by larval herring in turbid conditions, while another (Moore and Moore 1976) reports a shift in the foraging pattern or prey preference in turbid conditions. An investigation by Godin and Gregory (in prep.), on turbidity-induced increases in background contrast, even brings the seemingly well established evidence of negative effects on visual ability into question. The action of turbidity on foraging behaviour has not been demonstrably consistent. My investigations have provided evidence that supports the more established view of negative impacts of turbidity on foraging behaviour; foraging rate ultimately declined at high levels (« >200 mg-L-1). However, at intermediate levels, my results support a more flexible view. Planktonic foraging by juvenile chinook salmon on Artemia prey in the present study follows the well documented pattern. I observed high foraging rates at low turbidity conditions and much reduced rates at elevated levels. The decline in foraging rate may have been more distinct, such as that demonstrated by Gardner (1981) for bluegill 48 sunfish, except that handling time constraints likely limited the expressed foraging rate in clearer water. Established theory concerning visual ability in turbid conditions does not explain the foraging rates I demonstrated for surface and benthic foraging in chinook. I propose a revision in our thinking about turbid water foraging behaviour by these visual animals. Boehlert and Morgan (1985) observed peak foraging rates in elevated turbidity conditions for herring larvae. Their tentative explanation was an enhancement of visual contrast through a more constant illuminated background, effectively silhouetting the rotifer prey of the larvae. Work on contrast perception in fish (Hester 1968) and its effect on foraging behaviour (Ware 1973) suggests this may be possible. My own work (Godin and Gregory in prep.) has also suggested such a possibility may exist for juvenile chinook. In the latter study, chinook were observed to respond faster to prey (Artemia) that were more highly contrasted against turbid backgrounds, even with the demonstrated reduction in reaction distance (Ch. 4.1). My explanation for this involves detection of prey during the encounter phase of foraging (as defined by Holling 1966). Prey may be recognized as such, earlier on approach in turbid than clear water. In this way, a fish may reduce its recognition time in a diminished reactive field. Although my own work suggests this is at least possible, I do not believe it provides an adequate resolution to the dilemma presented here. Contrast enhancement conceivably would enable planktonic prey to be more quickly recognized as such. However, the same cannot be said of benthic or surface prey. Benthic prey could only be better contrasted if turbidity acted to modify the substrate characteristics. Turbid environments tend toward a uniform grey-brown background, but this is unlikely to appreciably change in varying turbidity phases in a given system. Also, in my experiments I have controlled for this effect (see Ch. 3.3), eliminating it as an explanation. Surface prey cannot assume 49 higher contrast in turbid water. The sky provides the contrast. Another explanation is required for these results. Given the variable form of the effect of turbidity on foraging rate and the magnitude of the changes, 1 suggest that a behaviour which is flexible with respect to turbidity levels is responsible. In clear water, when foraging on benthic or surface prey, chinook salmon may choose to forage at a reduced rate, sacrificing the obvious energetic reward for some other gain. A moving, foraging animal is more likely to be detected by a predator than a stationary, vigilant one. The act of foraging has been increasingly viewed as inherently dangerous for many animals (for review: Lima and Dill 1990). In clear water, a foraging fish will be more visible to a potential predator than when in turbid water. I suggest that juvenile salmon are sensitive to this perceived risk and forage accordingly. Higher rates may be realized at less "risk-prone" turbidity conditions. I investigate these ideas more thoroughly in subsequent chapters. Effects of Ontogeny The analyses I have conducted to examine ontogenetic changes in foraging behaviour must be viewed as exploratory. The results were not expected when I began my research and their nature was only revealed after the majority of my work had been completed. Further study will be required to fully describe this phenomenon. However, an understanding of the ontogenetic effects is required to facilitate the interpretation of results of several experiments that may appear inconsistent with the findings I have described in this subchapter. The results of these experiments may well be consistent, given the size of chinook used during the particular investigation. 50 For chinook foraging on benthic and surface prey, the effect of turbid water on foraging rate changed with ontogeny. However, similar ontogenetic effects were not demonstrated by juveniles foraging on planktonic prey. Ontogenetic changes in foraging behaviour have received much recent attention (for review of basic ideas: Werner and Gilliam 1984). In these investigations it is hypothesized that these changes involve "decisions" based on the costs and benefits of various foraging behaviours. One of the operating principles to emerge from this work has been the "minimize u/g" (mortality rate: growth rate ratio [Werner and Gilliam 1984]) or "minimize u/f" (mortality rate: feeding rate ratio [Gilliam and Fraser 1987]) rule. Under these rules, a foraging animal should choose a habitat, microhabitat, time, or environmental condition in which to forage which minimizes one or both of these ratios, when given a choice (assuming one exists). As a fish grows, its risk to predators will change as some function of its size (usually risk will decrease). Therefore, the above ratios will also change as a function of size. I believe that the ontogenetic changes in chinook foraging behaviour in turbid water were consistent with the above hypotheses. The predation risk may be reduced with both size (Werner and Gilliam 1984; Miller et al. 1988) and availability of cover (Savino and Stein 1982; Magnhagen 1988). Turbidity may act as a form of cover (Bruton 1985; Ch. 5). Therefore, as turbidity increases, the relative cover to young fish may be interpreted as increasing, and perceived predation risk decreasing (see Ch. 6 for full explanation). Foraging rates may be relatively low in clear water but high in turbid water (if the visual constraints are not limiting - again see Ch. 6 for details). As its size increases, the advantages of turbid water to the forager should be reduced. The reduction in predation risk commensurate with increased forager size may negate the advantages of cover offered by the turbid 51 conditions. This would seem to suggest that at larger sizes, fish should forage at higher rates in clear compared with turbid conditions. The "minimize p/g (or p/f)" rule would then be satisfied by both smaller and larger fish sizes. The "minimize p/g (or p/f)" rule seems to be contradicted by the results 1 have demonstrated. In clear water, lower foraging rates should have been expected at smaller, more vulnerable sizes of chinook. I observed the reverse. Larger, presumably less vulnerable, chinook juveniles demonstrated relatively low feeding rates at low turbidity and higher rates in intermediate suspended sediment concentrations. This result seems to suggest that larger juveniles are more susceptible to predation in clear water than are smaller conspecifics. This conclusion is inconsistent with existing theory. However, it may be possible that larger fish are more "sensitive" to their present risk to predators than are smaller individuals. Fish are more vulnerable to predation early in their life. Further, the probability of successful capture by a predator of any given size declines rapidly over a small range of forager size (Miller et al. 1988). Although the act of foraging puts a fish at risk, quick growth may be the most effective way of reducing this risk (Werner and Gilliam 1984; Miller et al. 1988). It follows that an individual fish, once having attained a larger size but still under some risk to predation, may be less likely to subject itself to that risk. As a result, a fish may adjust its foraging behaviour in a manner which at f i rs t may seem countei—intuitive. In Figure 4.7, I conceptualize this problem. At small sizes (Fig. 4.7A), visual ability and gut capacity (closed arrows) constrain foraging rate while the fish's perceived risk to predators effects a reduction in the maximum foraging rate (open arrows). The relative magnitude of the visual constraints and perceived risk are "matched", canceling each other, except at high turbidity. At larger fish sizes 52 Figure 4.7. Hypothetical ef fects on foraging rate of variables differentially af fected by turbidity and ontogeny. A. Matched effects (smaller fish). 6. Offset effects (larger fish), (see text fo r interpretation). Turbidity 53 (Fig. 4.7B), visual ability continues to constrain foraging rate, but at a higher limit because of larger gut capacities (higher satiation point). However at larger sizes, the forager's "sensitivity" to risk also increases but is "offset" from the physical constraints, affecting an increase in foraging rate at higher relative to lower turbidity condit ions. This section contains a great deal of speculation, which remains to be tested. I explore other ideas on the perceived risk of predation in subsequent chapters. Chapter 4.3 THE EFFECT OF TURBIDITY AND LIGHT ON FORAGING BEHAVIOUR INTRODUCTION Many studies examining foraging in turbid waters inevitably discuss the predominant role of light attenuation in turbid conditions (Swenson and Matson 1976; Gardner 1981; Sigler et al. 1984; Crowl 1989). Although increasing turbidity has a pronounced effect on the attenuation of light (Duntley 1943 and 1963; Munz 1958; DiToro 1978; Lythgoe 1980), any direct comparison of turbidity with light level can be misleading. Various fish species have been demonstrated to display thresholds in light sensitivity with regard to feeding success (Vinyard and O'Brien 1976; Blaxter 1966 and 1968; Brett and Groot 1963; Confer et al. 1978; McFarland 1986). Quite predictably, these thresholds have been shown to be species, habitat, and size specific even within taxonomically related groups (Henderson and Northcote 1985; Confer et al. 1978; Munz 1958). However, it is not certain whether feeding rates in turbid waters can be attributed to light level alone. If light level is not responsible for 54 the foraging relationships described in Chapter 4.3, it would be implied that turbidity modifies the visual environment of chinook fry in ways not readily explained by the existing literature. I conducted this study to compare the effects of light and turbidity on foraging rates for surface, planktonic, and benthic prey. I wished to determine if the relationships exhibited in Chapter 4.2, could be explained by the light levels commensurate with turbidity treatment. M E T H O D S The relationship between light level and turbidity could be described by the following equation: log(Light) = 1.01 - 0.0129 (Turbidity), where light was measured in pETrr 2 ,s~ 1 and turbidity was in nephelometric turbidity units (NTUs)(see also Ch. 3.3). Light levels for 0 - 200 mg*L"1 turbidity were measured directly, those for the remaining two levels were estimated from the regression. Reduction in the light levels required during the course of this experiment was effected using single sheets of onion skin paper layered on a plexiglass shelf under the flourescent lights to simulate the light level conditions as measured in turbid water. The number of sheets of onion skin was estimated from the following regression: No. sheets = 9.95 - 11.71 log(Light). This technique was successful in that the corresponding light level of each of the turbidity treatment conditions could be closely 55 approximated with f r o m 1 - 100 s h e e t s o f onion skin paper , rep resent ing 25 - 8 0 0 m g - L - 1 , respect ive ly . Three t y p e s o f t r ia l were conducted during this experiment, f o r each o f the t h r e e prey t ypes . Two s tandard forag ing r a t e t r ia l s were c o n d u c t e d as " c o n t r o l s " ; along with t h r e e light " t rea tment " t r ia ls , and one "compos i te" t r i a l . The cont ro l t r ia l s were identical in every r e s p e c t to those descr ibed previously (Ch. 3 and Ch. 4.2). The t rea tment t r ia l s were similar to the cont ro l s , except tha t t r e a t m e n t s cons is ted o f the above descr ibed light reduct ions instead o f the tu rb id i t y manipulations o f the s t a n d a r d p ro toco l . The composite t r ia l s c o n s i s t e d o f the fol lowing combination o f manipulations: one overal l c o n t r o l (0 m g - L - 1 tu rb id i t y ; 11.5 p E - r r r ^ s - 1 l ight), t h r e e tu rb id i t y r e p l i c a t e s (100 m g - L - 1 tu rb id i t y ; 1.15 p E * m - 2 - s " 1 l ight), and t h r e e light r e p l i c a t e s (0 m g - L - 1 tu rb id i t y ; 1.20 p E ' m - 2 , s ~ 1 l ight [9 sheets ] ) . The ent i re experimental array was " l i g h t - p r o o f e d " with the black opaque plast ic curta in surrounding each aquarium and the ent i re experimental a r r a y (Ch. 3). Light shining between aquaria was f u r t h e r reduced by using black tape on the plywood edges between the aquaria. Following th is p r o c e d u r e , l ight leve ls in " d a r k " conditions were less than the sens i t i v i t y limit o f the light m e t e r (0.01 p E - m ~ 2 . s - 1 ) . The juvenile chinook salmon used in this study ranged in mean size f r o m 47.9 to 51.7 mm FL (±2.5mm sd , within t r ia ls) . R E S U L T S S u r f a c e Foraging I observed a significant d i f f e r e n c e in the t r e n d exhibited by the tu rb id i t y " c o n t r o l " t r ia l s when compared to the light " t rea tment " t r ia ls 56 for chinook foraging on Drosophila prey (ANOVA df=1,18; p=0.011). Foraging rates in control trials exhibited a decline in foraging rate with increasing turbidity (Fig. 4.8A). The same result was not demonstrated by chinook foraging rate in corresponding light levels. Foraging rates were similar across all treatments within each trial. The composite trial (Fig. 4.8B) indicated no significant difference between the turbid and light treatments at the levels examined. Planktonic Foraging In general, the foraging rates on Artemia prey exhibited by chinook in both the turbid controls and the light treatments were in close agreement (Fig. 4.9A). The curves were not significantly different in an ANOVA. The composite trial also demonstrated close agreement between the feeding rates in turbid conditions and those of similarly reduced light levels (Fig. 4.9B). However, the difference between the turbid and light treatments was significant (ANOVA df=1,4; P=0.014), but not large. Similarly, a difference was also observed between corresponding levels in the control and treatment trials (Fig. 4.9A). Benthic Foraging In intermediate treatment conditions (50 - 200 mg'L-1; 3.22 -0.077 pE-m - 2 ^ - 1 ; Fig. 4.10A), juvenile chinook exhibited lower foraging rates for Tubifex prey in light treatments than in turbid controls. These differences were obscured by large variance and were not significantly different. However, the results from the composite trial demonstrated a large and significant (ANOVA df=1,4; p=0.001) difference 57 Figure 4.8. The effect on juvenile chinook salmon of turbidity and light level on foraging rate for surface prey (Drosophila). A. Mean foraging rate (preyfish~ 1 ,lO min - 1; vertical bars indicate standard error; N= 3 and 2 runs of light and turbidity treatments, respectively; solid line - light trials, dashed line - turbidity trials). B. Composite trial with three replicate treatments each of 100 mg'L - 1 turbidity (1.15 pE'trr^s" 1) and "reduced" light (1.20 pE'trT^'s""1) and one control treatment (<1 mg-L-1, 11.5 pE*m~2's"1; * on horizontal axes of A indicates treatments compared in B). Ql I II I I I I I—,t.,.i Control Light Turbid Treatment 58 Figure 4.9. The effect on juvenile chinook salmon of turbidity and light level on foraging rate for planktonic prey (Artemia). A. Mean foraging rate (preyfish"1-1.0 min"1; solid line -light trials, dashed line - turbidity trials). B. Composite trial, (see caption Fig. 4.8 fo r details) 59 Figure 4.10. The effect on juvenile chinook salmon of turbidity and light level on foraging rate for benthic prey (Tubifex). A. Mean foraging rate (preyfish~ 1 ,5.0 min - 1; solid line -light trials, dashed line - turbidity trials). B. Composite trial, (see caption Fig. 4.8 fo r details) Log(microEinsteins/m2/s) 26 60 100 • 2 0 0 4 0 0 8 0 0 Turbidity (mg/L) B 26 £ 2 0 E to 16 I | DC 10 JL Control Light Turbid Treatment 60 between the turbid and light treatments. The foraging rates in the turbid treatment of the composite trial were approximately 6 - 7 times higher than the light treatments. The levels of the latter were similar to the control. I have reported the details of each trial in the Appendices (Fig. A.3, 8, and 13, for Drosophila, Artemia, and Tubifex, respect ive ly ) . D I S C U S S I O N Support for the hypothesis - the effect of turbidity on foraging rate is similar to that of corresponding light intensity - was weak. The results of this experiment indicate that light alone, while possibly able to explain the foraging rates of chinook salmon on Artemia prey, did not predict surface and benthic foraging rates. Duntley (1943) argued that the dominant feature of turbid water optics is the interference of the light signal and subsequent distortion of any visual image. In high light conditions, a prey object might not be seen at close range if the water is even slightly turbid. The attenuation of light itself is a secondary feature of absorption and scattering by suspended particles. Indeed, several investigations have demonstrated that foraging rates (Harden Jones 1956; Brett and Groot 1963) and reaction distances (Vinyard and O'Brien 1976; Confer et al. 1978) exhibit threshold light levels, above which they are maximal and below which they decline sharply. Confer et al. (1978) demonstrated that reaction distances of both lake and brook trout (Salvelinus namaycush and S. fontinalis, respectively), and pumpkinseed sunfish (Lepomis gibbosus) exhibited thresholds of 50 - 100 lux and 20 - 50 lux, respectively. Above these light levels, reaction distance to prey of a given size was maximized. 61 Vinyard and O'Brien (1976) demonstrated a strong relationship between reaction distance and light level, but also reported little change in reaction distance above 10.8 lux in pumpkinseed sunfish. This value was similar to that reported by Confer et al. (1978). In their study of sockeye salmon fry, Brett and Groot (1963) observed substantial reductions in foraging rate only at light levels below 0.1 Lux. Above this level feeding rates were relatively constant. Species in closely related taxonomic groups often possess dissimilar sensitivity thresholds (Henderson and Northcote 1985; Confer et al. 1978). At the lower turbidity levels tested in my experiments, light levels were higher than the thresholds reported for various salmonids (Brett and Groot 1963; Confer et al. 1978). Reductions in chinook foraging rates should not have been expected at these light levels (i.e. 0 - 100 mg-L-1). Foraging rates in the two types of treatments should have been similar if "light e f fects" were predominant over "particle effects" in turbid water. This seems to have been the case for planktonic foraging by chinook. However, this hypothesis is not supported by the benthic and surface foraging rate results. Given that the turbid treatments demonstrated higher feeding rates than the light treatment runs at intermediate turbidity/light levels, a more dynamic process is suggested. These results do not statistically reject the hypothesis that light level alone is responsible for foraging rates exhibited by chinook salmon in turbid water. However, they of fer no support for the hypothesis either. On balance, results suggest rejection, but were inconsistent. Definitive statements cannot be justified. 62 CHAPTER 5 MICROHABITAT SHIFTS AND RESPONSES TO PREDATORS Chapter 5.1 SPATIAL DISTRIBUTION INTRODUCTION Often, relatively small differences in the distribution of animals within habitats can have a pronounced effect on foraging opportunities or potential risk to predators. For a fish foraging in turbid waters, these effects may be further accentuated. In laboratory experiments, Swenson and Matson (1976) provided evidence of vertical movements of larval lake herring (Coregonus artedii) in turbid water. In the field, Swenson (1978) and Bruton (1979) demonstrated changes in the spatial distribution of smelt (Osmerus mordax) and tilapia (Oreochromis  mossambicus), respectively, in response to elevated turbidity levels. Bruton (1979 and 1985) concluded that tilapia were not simply tracking food concentrations, but were actively modifying their behaviour. He observed that tilapia moved into food-rich (and turbid) shallows in response to a relaxation of predation risk. Due to the pronounced effect of turbidity on vision (Ch. 4.1), encounter rates by foragers with both predators and prey are likely to be affected by turbidity, even at relatively low levels. High densities of juvenile salmonids are often found in shallow waters of turbid estuaries (Simenstad et al. 1982; Murphy et al. 1989; Northcote and Larkin 1989; Gregory pers. obs.). However, this may also be the case in clear water estuaries (Healey 1982; Simenstad et al. 1982) or in clear water tributaries of turbid systems (Gregory pers. obs.). Therefore, the role of turbidity in determining fine scale spatial distribution in salmonids remains unclear. 63 Changes in salmomd spatial distr ibut ions in response to a l te red ambient conditions have been well documented. These changes have been o b s e r v e d mainly in connection with the ver t ica l migration behaviour o f sockeye salmon f r y (Oncorhynchus n e r k a - Levy 1987) o r microhabitat use in s t r e a m s (O. k isutch and 0_. tshawytscha - Tay lor 1989). Although the impact o f tu rb id i t y on spat ia l d i s t r ibu t ion o f juvenile chinook salmon has largely been ignored in the l i te ra tu re , work has been conducted on downstream displacement o f coho salmon and rainbow t r o u t (O. mykiss=Salmo gairdneri - Sigler e t a). 1984; Noggle 1978; Berg and N o r t h c o t e 1985). These invest igat ions primarily dealt with pulses o f suspended sediment o r levels not normally encountered in natural habi tats o f these f i sh . Also, none o f the studies r e p o r t subtle microhabi tat changes in spatial distr ibution, except f o r tha t o f B e r g and N o r t h c o t e (1985). Their work demonst rated that juvenile coho assoc iated more closely with the bottom during elevated turbidity pulses. The p r e s e n t subchapter desc r ibes an experiment investigating changes in spat ia l d i s t r ibu t ion in e l e v a t e d tu rb id i t y levels under contro l led labora to ry conditions. The experiment was designed t o a d d r e s s whether any such changes occur , and if so, to a s s e s s the qual i tat ive e f f e c t s e x h i b i t e d . M E T H O D S The chinook f r y used in this experiment were 78.0 t 1.6 mm (FLtSD) and 4.64 t 0.36 g in size when introduced into the arena. Twenty - f i ve f i s h were used. The experimental arena, conditioning p ro toco l , and r e l a t e d techniques have been previously descr ibed in Chapter 3.4. Fish were exposed t o only one turbidity t rea tment level on a given date, these t r e a t m e n t s (0, 25, 50, 100, and 200 mg*L~1) being randomized 64 with respect to date. Each treatment was performed on two dates. One additional date of 0 mg-L-1 treatment was performed at both the beginning and end of the observational phase, for a total of twelve experimental days. Observations were made daily from 0900 to 2100 hours, inclusive, at two hour intervals. Observations were made by simple census. The number of fish present in each region of the grid (Fig. 3.3) was counted. The order of the census was randomized within individual observation times, with all regions being counted. I made three counts within each observation time, 5 minutes apart, these were pooled for analysis. In instances of doubt, fish could be assigned to a region by position of the eye. Data analysis was conducted by G-test of contingency tables. R E S U L T S In clear water controls, the distribution of fish was significantly affected by the time at which the observations were taken (G-test, p=0.024, df=6). Analysis of residuals revealed that much of the significance could be attributed to the observation time periods immediately before and after feeding. Consequently, observations at 0900, 1100, 1900, and 2100 hours were dropped from further analysis. The middle three observation periods (1300-1700 hours) were not significantly different (G-test, p=0.117, df=2). All subsequent analyses were conducted using these observations. G-tests for heterogeneity among trials revealed no significant differences for each of the turbidity treatment levels. Pooling of observations by run and time period was therefore justified. My primary hypothesis - turbidity affects spatial distribution of chinook fry - could not be rejected. Analysis by G-test revealed a significant e f fec t of turbidity on distribution of test fish in the 65 exper imenta l a r e n a (Table 5.1, p<0.001, df=20). In addition t o the above hypothesis , I conducted severa l a pos te r io r i t e s t s to more fully d e s c r i b e the e f f e c t o f tu rb id i t y level on d i s t r i b u t i o n s . Significant d i f f e r e n c e s were demonst rated f o r the la te ra l and ver t i ca l distr ibut ions o f t e s t f i s h with r e s p e c t to turbidity level (Table 5.2.). These r e s u l t s suggested that f i s h were simultaneously changing thei r d i s t r ibu t ion in b o t h o f t h e s e dimensions. In c l e a r w a t e r , s ignif icant ly more f i sh (55.3X) were at g r e a t e r than 20 cm in depth than expected f r o m the volume of water in this s t r a t u m (38.5/). However, the number o f f i sh at g r e a t e r than 20 cm depth in turbid water was not significantly d i f f e r e n t f r o m that expected by chance (Fig. 5.1). F u r t h e r , the distr ibut ion f r o m only one turbidity t rea tment (100 mg-L - 1 ) was not s ign i f i cant l y d i f f e r e n t f r o m that in c lear water . These r e s u l t s also suggested the ex istence o f a positive associat ion with the bottom in c lear water (or a negative associat ion with the su r face ) , which was not readily apparent among the turbidity t r e a t m e n t s . D i f f e r e n c e s between the turbid t r e a t m e n t s (i.e. o t h e r than c lear) were e i ther not significant o r were cons is tent with this h y p o t h e s i s (Table 5.3). D I S C U S S I O N Changes in v e r t i c a l d i s t r ibu t ion re la t i ve to turbidity have been demonst ra ted b e f o r e . Swenson (1978) believed that zooplankton c o n c e n t r a t e d above a turbidity wedge, causing smelt t o move into the shallow s u r f a c e w a t e r s to f e e d on the zooplankton, accounting f o r the o b s e r v e d changes in smelt d ist r ibut ion. F o r tilapia, observed inshore movements were believed to have been in response to both higher f o o d availability and a re laxat ion o f predat ion p r e s s u r e f r o m f i s h eagles (Bruton 1979). The p r e s e n t experiment provided r e s u l t s cons is tent with 66 Table 5.1. The effect of turbidity concentration on the spatial distribution (as a percentage) of juvenile chinook salmon by region, within an experimental arena (observations pooled by observation period; Region is indicated on the inset schematic of Fig. 3.3; Chi-square is from G-test). Region Percent of Turbidity (mg/L"1) Number Arena Volume 0 25 50 100 200 1 26.41 11.20 42.17 28.38 33.98 31.91 2 26.41 35.82 19.57 26. 13 26.21 21.28 3 12. 12 8.44 10.00 4.95 8.74 6.38 4 12. 12 19.50 14.78 10.36 21.36 12.77 5 12. 12 16.04 4.78 16.22 7.77 10.64 6 10.82 8.99 8.70 13.96 1.94 17.02 N EE: 1325 723 230 222 103 47 X* = 168.60 p<0.001; df=20. Schematic of Figure 3.3 indicating regions. 67 Table 5.2A. The effect of turbidity on the lateral distribution of juvenile chinook salmon by pooled region (observations pooled by observation period and expressed as percentages; see Table 5.1 for Region; Chi-square is from G-test). Region Percent of Turbidity (mg-L"1) Number Arena Volume 0 25 50 100 200 1&2 52.81 47.03 61.74 54.50 60.19 53. 19 3,445 36.37 43.98 29.57 31.53 37.86 29.79 6 10.82 8.99 8.70 13.96 1.94 17.02 N EE= 1325 723 230 222 103 47 X* =39.54 ; p<0.001; df=8. Table 5.2B. The effect of turbidity on the vertical distribution of juvenile chinook salmon by pooled region (observations pooled by observation period and expressed as percentages; see Table 5.1 for Region; Chi-square is from G-test). Percent of Turbidity (mg-L"1) Number3 Arena Volume 0 25 50 100 200 1&3 5O.0 26.2 60.3 47.7 47.3 52.9 244 50.0 73.8 39.7 52.3 52.7 47.1 N EE=1023 542 199 155 93 34 a regions 5 * 6 omitted from analysis X* =86.35; p<0.001; df=4. 68 Figure 5.1. The ef fect of turbidity on the proportion of juvenile chinook salmon in the lower 20 cm of a 40 cm deep experimental arena (vertical bars are 95ZCL [after Bailey 1980]; horizontal line represents the proportion expected by chance). 0 50 100 150 200 Turbidity (mg/L) Table 5.3A. The effect of turbidity on percentage of juvenile chinook salmon associated with the bottom (observations pooled by observation period; see Table 5.1 for Region; Chi-square is from G-test). Region Percent of Turbidity (mg-L"1) Number Arena Volume 0 25 50 100 200 U3 38.53 19.64 52.17 33.33 42.72 38.30 3,4,546 61.47 80.36 47.83 66.67 57.28 61.70 N EE=1325 723 230 222 103 47 X*=99.14; p<0.001; df=4. X? =16.87; p=0.001; df=3 - for turbid treatments alone. Table 5.3B. The effect of turbidity on percentage of juvenile chinook salmon associated with the surface (observations pooled by observation period; see Table 5.1 for Region; Chi-square is from G-test). Region Percent of Turbidity (mg-L"1) Number Arena Volume 0 25 50 100 200 1,3,546 61.47* , 44.67 65.65 63.51 52.43 65.96 244 38.53 55.33 34.35 36.49 47.57 34.04 N EE=1325 723 230 222 103 47 X* =48.43; p<0.001; df=4. X? =5.66; p=0.129 NS; df=3 - for turbid treatments alone. 70 the observations made in these studies. The results detailed in this experiment supported the hypothesis that turbidity a f fects the spatial distribution of juvenile chinook salmon. Fish were found higher in the water column and in shallower water at elevated turbidity levels. The vertical and horizontal scales of this experiment were restricted in comparison to those that would be encountered in the field; however, the implications of these results are broad. First, given the significant movement of f ry to the upper portions of the water columnin turbid conditions, encounter rates with surface prey may increase relative to planktonic and benthic prey (assuming availability of the prey has not also been affected by the turbid conditions). Similar chinook movements may also have occurred during my foraging rate experiments (e.g. Chapter 4.2). In turbid conditions, feeding rates on surface prey should have increased relative to those of planktonic and benthic prey. I observed this increase for surface foraging on Drosophila. Second, although chinook moved up in the water column in turbid water, this movement suggested a threshold type of response. Beyond 25 mg-L""1, the number of test fish in the deeper regions did not decrease as a function of the turbidity level as would be expected if the fish were tracking light levels. However, scale may have been a factor. And third, this experiment demonstrated that turbidity may cause chinook to alter their spatial distribution, but this change was not sufficient to explain the feeding relationships encountered in the previous chapter (Ch. 4.2). For example, the foraging rates on benthic prey also rose in turbid water. The observed change in spatial distribution may partly explain increases in foraging rate on surface prey, but it does little to clarify the observed effect of turbidity on the foraging rate of chinook on benthic Tubifex. These results did not suggest that juvenile chinook were maintaining visual contact with the bottom by moving closer to it as demonstrated by juvenile coho (Berg and Northcote 1985). Instead, fish 71 became dispersed throughout the arena in turbid waters as compared to the clear water controls. The results were consistent with the hypothesis that turbidity may act as a form of cover against predation (Bruton 1979 and 1985; Gradall and Swenson 1982) although, the hypothesis is not directly supported. Chapter 5.2 RESPONSES TO MODEL PREDATORS INTRODUCTION A beneficial e f fec t of turbidity for foraging fish, specifically their reduced susceptibility to predators, has been postulated many times (for reviews: Miller 1976; Simenstad et al. 1982; Bruton 1985; Guthrie 1986). However, hard evidence has been lacking. The suggestion that turbidity acts as a form of cover seems to have been more a reasonable assumption than a supported principle. There have been several notable observations consistent with this hypothesis. Bruton (1979) demonstrated that tilapia grew better in turbid water shallows; he found that fish of this species avoided similar areas in clear water lakes. Blaber and Blaber (1980) and Cyrus and Blaber (1987a) observed more juvenile fish in moderately turbid portions of estuaries. Cyrus and Blaber (1987b) demonstrated experimentally that many juvenile estuarine fishes actively seek out moderate turbidity levels. Gradall and Swenson (1982) found that creek chub (Semotilus atromaculatus) and brook trout (Salvelinus fontinalis) relied less on overhead cover and became more active in turbid conditions. White (1936) observed that foraging rates by piscivorous birds were reduced after rains had caused a rise in the turbidity of salmon streams. Ginetz and Larkin (1976) correlated turbidity with decreased predation pressure during downstream 72 migrations of juvenile salmonids. Anecdotal and correlative support for the "turbidity as cover" hypothesis has been considerable. The effects of manipulations of predation pressure and turbidity on forager behaviour have not been described in the literature. There exists a myriad of investigations concerning predator mediated decision-making processes of foragers (reviews: Werner and Gilliam 1984; Dill 1983 and 1987; Milinski 1986; Lima and Dill 1990). This literature is especially rich concerning fish. Much of this work investigates the response of foragers to combinations of predation risk, food reward, and accessability of cover. None of these investigators manipulated the risk of predation simultaneously with the forager's ability to detect either predation risk or its own food. A turbid medium imposes such a conflict. It acts to restrict the visual ability of the forager, as well as those of its principal avian and aquatic predators. Whether it is primarily the forager or the predator which benefits from these visual constraints has been the subject of debate. Statements such as: "...[turbidity may] provide cover from predators and for predators stalking their prey." (Bruton 1985). acknowledge this problem. I believe he has summarized the problem well. I , conducted the experiments described in this subchapter, to qualitatively describe the response of juvenile chinook salmon to predation risk under combinations of turbidity level and additional cover. Model bird and fish predators were used throughout. The response of the chinook was assessed by using a relative index - change in spatial distribution - to describe the risk perception state of loose aggregations of 22 test fish in the laboratory. The magnitude and duration of a risk-induced response were determined in separate experiments. 73 M E T H O D S Two predator models were used in these experiments: a fish and a bird piscivore. My use of these models in this experiment was intended to be representative of the two general classes of salmon piscivores. I did not intend to conduct a detailed investigation of the intrinsic differences of their effects. The fish model was that of a spiny dogfish (Sgualus sp.), 32 cm in length and made of rubber. The bird predator was a stuffed museum specimen of an immature glaucous-winged gull (Larus glaucescens), 62 cm in length. Both of these species were known predators of juvenile salmon (Mace 1983; Hargreaves 1990 and pers. comm.). Model predators were positioned in the water column and above the experimental arena. Models were pulled with clear monofilament line at a speed of approximately 1 m-s-* along fixed guide wires (F-F' and B-B', for the fish and bird models, respectively - Figure 3.3). The guide wire for the bird model traversed diagonally above the tank at a height of 40 cm above the water surface. The fish model was drawn from near the middle of the water column in the deepest portion of the arena to the surface at the shallow end. When not in use, models were positioned in a "ready" position behind a curtain outside the view of the test fish. The Magnitude of the Effect Twenty-two juvenile chinook were used in each of two sequential replicates. These fish were 83.412.7 mm (mean+sd FL) and 78.8+3.3 mm before, and 87.5+2.9 mm and 83.6+5.1 mm after, in each of the replicates, respectively. Each replicate consisted of twelve treatment days preceeded by 7 days of preconditioning (see Ch. 3). The experiment 74 was structured in a crossed, three-factor design, with manipulations of: turbidity level (present [25 mg*L_1 only] or absent), additional cover (present or absent), and predator exposure (bird model, fish model, or absent). Each of the twelve experimental days consisted of a unique combination of treatment factors. These combinations were randomly ordered within each replicate. Additional cover was provided by positioning plastic drinking straws vertically, from above the surface to within 2 cm of the bottom, uniformly throughout the arena. These were intended to represent stems of vegetation. A density of 10C-m"2 was used and seemed reasonable from past work (Savino and Stein 1982). A trial consisted of exposing the chinook to one or neither of the models and observing the subsequent spatial distribution. Model exposure was facilitated by pulling the appropriate model from its "ready" position into the view of the test fish to a "holding" position at the opposite end of the guide wire. The model was returned to its "ready" position following observations. In "no model" trials, a sham treatment was provided by performing identical operations as for "model" trials, except that the model was not pulled into the view of the fish. Observations were conducted a standard 15 seconds after exposure. Three of the above trials were made on each experimental day at 1300, 1500, and 1700 hours. Observations were made by visual census. Regions 1 and 2 were censused separately while Regions 3, 4, 5, and 6 were counted as a unit. This method allowed the completion of all observations within 20 seconds. These observations provided an "instantaneous" assessment of spatial distribution immediately following disturbance by a predator. 75 The Duration of the Effect - Recovery from Predator Disturbance I measured the duration of the post-disturbance behaviour elicited by the predator, by videotape analysis. Additional trials of the previously outlined interactions were conducted over a period of four days following the completion of the second replicate; the fish used were also from this replicate. Using the previously described procedure, I videotaped the spatial distribution of the fish in the arena for 5 minutes prior to and for 10 minutes following exposure to the predator models. The number of fish in or entering each of the arena's regions was counted throughout sequential increments of 30 seconds each. The four experimental days were set with the following conditions: Day 1 - additional cover, clear; Day 2 - additional cover, turbid; Day 3 - no additional cover, turbid; and Day 4 - no additional cover, clear. The four treatment runs consisted of two runs each with the bird and fish model trials conducted at two hour intervals on each of the four experimental days, beginning at 1200 hours and ending at 1800 hours. Data were analyzed by chi-square. The mean frequency of observations in each region of the arena during the pre-disturbance period was used as an expected value for calculating a X* statistic. The X8 was then subsequently determined for each 30-second time interval in both the pre- and post-disturbance periods. The 95X confidence limit was used to approximate the time taken for test fish to return to a pre-disturbance spatial distribution within the arena. I have referred to this time as recovery time. I arbitrarily selected the first occurrence of three consecutive time intervals in the post-disturbance period which fell below the critical value of the statistic (X*(cc=.05;df=4) = 9.488) as evidence that the fish had recovered from the effects of the disturbance. I took the time elapsed before this occurrence as the recovery time. In instances where the above criterion could not be 76 s a t i s f i e d within t he du ra t i on o f t he exper iment , I r e c o r d e d the r e c o v e r y time as 10.0 minutes. As I have u s e d it, t he r e c o v e r y time may no t r e p r e s e n t a comple te a b s e n c e o f s t r e s s in chinook induced by t he models. However , a s a r e l a t i ve index on which t o c o m p a r e tu rb id i t y and c o v e r t r e a t m e n t s , it was e f f e c t i v e . R E S U L T S T h e Magnitude o f t he E f f e c t E x p o s u r e t o e i t he r o f t he p r e d a t o r models e l ic i ted a gene r a l s t a r t l e t y p e o f r e s p o n s e . All o r many o f t he t e s t f i s h immediately and rapid ly moved t o t he d e e p e s t r eg ion o f t he a r e n a . T h e o b s e r v a t i o n s p r e s e n t e d h e r e r e p r e s e n t e d the time pe r i od within 15 s e c o n d s o f th is r e s p o n s e . In "no p r e d a t o r " c o n t r o l s , t he p r o p o r t i o n o f f i s h in t he d e e p e s t r eg ion o f the a r e n a (Region #2) was h igher in c l e a r wate r , b o t h in t he p r e s e n c e o r a b s e n c e o f addit ional c o v e r (Fig. 5.2). In only one c o m p a r i s o n (Fig. 5.2B, R e p l i c a t e 1), was th i s d i f f e r e n c e no t s ign i f i cant . T h e s e r e s u l t s s u p p o r t e d t h o s e r e p o r t e d in t he p rev ious s u b c h a p t e r (Ch. 5.1). E x p o s u r e t o p r e d a t o r models e f f e c t e d a f u r t h e r i n c r e a s e in the p r o p o r t i o n o f f i s h in Region #2 o v e r t he "no p r e d a t o r " c o n t r o l s in c l e a r w a t e r (Fig. 5.2). T h e e f f e c t was more p r o n o u n c e d f o r t h e b i rd model than it was f o r t he f i s h model. A s t he e f f e c t o f st imulus s ize was no t c o n t r o l l e d , I have viewed this r e s u l t conse r va t i v e l y . Be tween about 3 0 / and 120/ more chinook were p r e s e n t in deep w a t e r (Region #2) in t he p r e s e n c e o f one o f t he models than when no model was u sed . T h e s e e f f e c t s were r e d u c e d in t u rb id wa te r (Fig 5.2). Resu l t s in t h e p r e s e n c e 77 Figure 5.2. The immediate spatial response of juvenile chinook salmon to exposure to predator models in clear and turbid (25 mg*L~1) conditions; proportion of fish in the deepest region (# 2) of an experimental arena in, A. The absence and B. The presence of additional cover in two replicates (open and shaded bars represent clear and turbid, respectively; vertical bars represent standard error; horizontal line represents the proportion of fish expected by chance). A Replicate 1 Replicate 2 I 1 1 i II = 11 11 Bird Fish None Bird Fish None Model Type B Replicate 1 Replicate 2 1001 i Bird Fish None Bird Fish None Model Type 78 and absence of additional cover were not significantly different (Fig. 5.2 and Table 5.4A). The main effects of turbidity and presence of a predator model were both significant (ANOVA p<0.001; Table 5.4A). The interaction between the effects of turbidity and the presence or absence of the models, was also significant (p<0.001). The above results were essentially reciprocal to those observed at the surface (Region #1 - Fig. 5.3). The presence of either of the two predators and turbidity was significant (ANOVA, p<0.001; Table 5.4B). Lower percentages of chinook were observed in surface waters in the presence of either of the models. During some trials, none were present in the case of the bird. Fish did not occupy the surface water region in representative proportion in clear water treatments, regardless of the predator treatment. The presence of turbidity effected an increase in the relative frequency of chinook occupying the surface region. Analysis of variance (Table 5.4B) demonstrated that the differences were significant, although the interaction between turbidity and predator model e f fects was not significant. The Duration of the Effect Many of the results for the duration of the post-disturbance effect were similar to the above observations. Exposure to model predators caused a significant change in the distribution of chinook in the arena. In all cases involving the bird model in clear water (Fig. 5.4 and 5.5), lengthy departures from the pre-disturbance spatial distribution were observed, often continuing until after the cessation of the videotaping (10 minutes post-disturbance). Similar departures were observed for the fish model (Fig. 5.6 and 5.7), but they were neither as strongly expressed nor as long-lasting. In one case (Fig. 5.7B), the post-disturbance was not different from the pre-disturbance 79 Table 5.4. Analysis of variance of the effects of turbidity, cover, and exposure to predator models on proportions of juvenile chinook salmon located in A. Deeper water (Region #2); and B. Shallower water (Region #1) of an experimental arena (proportional data was arcsine transformed). A. Deeper Water (Region #2) Source of Var i at i on Sums of Squares DF F P Turbidity(T) 1.689 1 69.011 K K K Cover(C) 0.015 1 0.614 NS Predator(P) 1.073 2 21.931 K K K T x C 0.025 1 1.012 NS T x P 0.460 2 9.391 X * K C x P 0.001 2 0.029 NS T x C x P 0.012 2 0.006 NS Error 1.468 60 K , K K , K K X - p<0.05, 0.01, 0.001, respectively; NS - p>0.10. B. Sha11ower Water (Region #1) Source of Var i at i on Sums of Squares DF F P Turbidity(T) 0.342 1 26.020 K K K Cover(C) 0.010 1 0.749 NS Predator(P) 0.278 2 10.558 K K K T x C 0.000 1 1.010 NS T x P 0.015 2 0.561 NS C x P 0.004 2 0.145 NS T x C x P 0.004 2 0. 158 NS Error 0.789 60 K , K K , K K K - p<0.05, 0.01, 0.001, respectively; NS - p>0.10. 80 Figure 5.3. The immediate spatial response of juvenile chinook salmon to exposure to predator models in clear and turbid (25 mg-L-1) conditions; proportion of fish near the surface in deep water (Region # 1) of an experimental arena in, A. The absence and B. The presence of additional cover in two replicates (see caption on Fig 5.2 for details). 80 C o & £ 6 0 C 40 OJ o OJ Q. Replicate 1 Replicate 2 Bird Fish None Bird Fish None Model Type B 100 80 c g c S 4 0 20 Replicate 1 Replicate 2 Bird Fish None Bird Fish None Model Type 81 Figure 5.4. Changes in microhabitat distribution in juvenile chinook salmon, before and after exposure to a model bird predator, in clear (A. and B.) and turbid (C. and D.) water with additional cover. (The significance of the change in spatial distribution, based on the pre-exposure spatial distribution, was assessed by the Pearson Chi-square statistic; the arrow and vertical line delineate the time of the exposure; to the left of this line is the pre-exposure time, to the right is post-exposure; the horizontal line represents the critical X*a=o.o5 value of 9.488). Time (min) 82 Figure 5.5. Changes in microhabitat distribution in juvenile chinook salmon, before and after exposure to a model bird predator, in clear (A. and B.) and turbid (C. and D.) water without additional cover (see caption of Fig. 5.4. fo r details). Time (min) 83 Figure 5.6. Changes in microhabitat distribution in juvenile chinook salmon, before and after exposure to a model fish predator, in clear (A. and B.) and turbid (C. and D.) water with additional cover (see caption of Fig. 5^ 4. fo r details). Time (min) 84 Figure 5.7. Changes in microhabitat distribution in juvenile chinook salmon, before and after exposure to a model fish predator, in clear (A. and B.) and turbid (C. and D.) water without additional cover Tsee caption of Fig. 5.4. fo r details). Time (min) 85 spatial distribution. Turbidity effected a reduction in both the magnitude and the duration of the post-disturbance response in all cases but one (Fig. 5.5D). I attribute this exception to random noise, but the possibility of subtle lags in post-disturbance effects cannot be discounted. In only three of the eight turbid treatments, were the post-disturbance periods statistically different from the pre-disturbance spatial d i s t r ibu t ion . The presence of additional cover appeared to have an unexpected effect on the post-disturbance spatial distribution for the bird model. Whereas juvenile chinook exhibited recovery times of approximately three minutes in clear water without cover (Fig. 5.5), times with cover exceeded the duration of the experiment (Fig. 5.4). This disparity was not observed for the fish model. Recovery time in clear water was approximately 8 times slower than in turbid water (Fig. 5.8), when averaged across all treatments. Analysis of variance (Table 5.5) demonstrated the main effects of turbidity and predator accounted for 67.3X of the variance in these data, while cover was not significant. D I S C U S S I O N Other than broad generalizations, no significance can be attributed to the differences in the responses of the test fish toward the two predator models. Exposure to both models elicited similar responses by the test fish; a distinct, rapid movement into deeper water. For the gull model, this movement would appear to be an appropriate evasive manoeuvre. However, for the model fish piscivore, this response seems counter-intuitive. A movement to the bottom would expose juvenile chinook to predation by sculpins (Leptocottus armatus), 86 Figure 5.8. The effect of turbidity and additional cover on the recovery time of juvenile chinook salmon after exposure to model predators (open - clear, shaded - turbid; each bar represents the mean of two measures). 1.2 E 1.0 c E E 0.8 >» 0.6 I O 0.4 CC O ) 0.2 O 0.0 Bird/Cover Bird/NoCover Fish/Cover Fish/NoCover 87 Table 5.5. Analysis of variance of the effects of turbidity, cover, and predator type on recovery time from exposure to model predators (times were log-transformed). Source of Var i at i on Sums of Squares DF Turbidity(T) 0.763 1 14.108 ** Cover(C) 0.009 1 0.166 NS Predator(P) 0.593 1 10.965 « Er ror 3 0.649 12 »,»*,*»» - p<0.05, 0.01, 0.001, respectively; NS - p>0.10. 3 error term includes the interaction term. 88 well known for their predatory behaviour on juvenile salmonids (Oncorhynchus keta - Mace 1983; Jones 1986). The physiology of fish o f f e r s one potential insight into this problem. Startle responses, such as those exhibited by the chinook in this experiment have been attributed to the action of giant Mauthner cells (Eaton and Bombardieri 1978) in reponse to exposure to large moving objects. The response is described as general, any object within a certain size range, will elicit a response. This type of visually mediated startle response occurs in many species (Guthrie 1986). The type of behaviour exhibited by the chinook here may represent some reflex response, adaptive for most predator situations, including most fish piscivores; albeit not for sculpins. The response exhibited by the fish in relation to cover is more curious. In the presence of an avian predator, a quicker recovery time in the absence of cover may be the appropriate response. "Cover" in this case, may represent a trap, impeding the safe retreat of the forager. Given the risk the fish may perceive in this event, a quicker recovery time during the cover treatments perhaps should not be the expected result. Also, avoidance responses and the use of cover for chinook may be predator specific. For the fish model, no differences were observed in chinook behaviour with respect to cover. Such effects were expected, given past work (Savino and Stein 1982). It is possible that the stem densities I used in the cover treatments were inappropriate to investigate what has recently been revealed to be a threshold response (Gotceitas and Colgan 1989). Turbidity disrupts the inherent visual abilities of fish. The results I observed may have resulted from the inability of the chinook to detect the presence of a particular model. I maintain this was not the case. The results suggest that juvenile chinook modify their spatial distribution in a manner sensitive to the risk and to the turbidity level (i.e. the ANOVA interaction term was significant). 89 Still, I cannot completely discount the possibility that the response resulted from a simple failure to sense the threat. The manifestation of risk cannot easily be separated from the visual recognition of risk stimuli in this experiment. A forager may be at less risk of detection by a visual predator in turbid water, but once detected may be at higher risk to capture. Conversely, the forager may be at less risk in the above circumstances because its escape responses may be more effective. "Cover" may be only a body length away for a foraging chinook fry in many turbid systems. Also, turbidity probably eliminates any advantage in visual acuity that a larger fish predator may possess (Tamura 1957; Hairston et al. 1982; Li et al.. 1985). Which of these scenarios is the most likely depends largely on whether or not the predator sees the forager first. In the presence of avian predators, the value of turbidity as a potential form of cover is easily appreciated, humans see water from a similar visual perspective. Turbidity obscures the vision of bird piscivores in most cases (for an exception: Haney and Stone 1988). White (1936) observed reduced success in two piscivorous birds, a kingfisher and a merganser, hunting juvenile salmonids in turbid conditions. Bruton (1979) reported that tilapia (Oreochromis  mossambicus) found in the food-rich shallows of a turbid lake grew to be six times larger than individuals in a similar but clear lake. He suggested differential predation pressure by visually hunting fish eagles was responsible. The results of the present study suggest that juvenile chinook salmon respond to the presence of a predator threat in both turbid and clear water. Chinook both responded less intensely and for shorter durations of time in the turbid conditions. I suggest that fish sensed the presence of the predator model rather than simply failed to detect it. Test fish exhibited responses to the presence of models in both clear and turbid water that significantly departed from non-predator 90 situations. These results were consistent with the "turbidity as cover" hypothesis (Bruton 1979 and 1985; Blaber and Blaber 1980; Gradall and Swenson 1982; Simenstad et al. 1982). 91 CHAPTER 6 A CONCEPTUAL HODEL OF FORAGING BEHAVIOUR UNDER VISUAL CONSTRAINT Chapter 6.1 INTRODUCTION The results demonstrated in Chapters 4 and 5 suggest that a tradeoff may exist between the conflicting demands of foraging ability and risk of predation in turbid waters. Reaction distance (Ch. 4.1) declines log-linearly with increasing turbidity. Some types of predation risk also decline with increasing turbidity (Ch. 5.2), although the exact nature of the relationship cannot be determined from my data. The relationships exhibited by the foraging rates of juvenile chinook salmon for surface, planktonic, and benthic prey under various turbid conditions (Ch. 4.2) also suggest such a behavioural tradeoff. I will demonstrate theoretically that an animal may make foraging decisions sensitive to both visual restrictions and perceived predation risk. Throughout the present and preceding chapters, I have referred to the potential risk of predation as perceived risk. This terminology reflects a more accurate perspective than the conventional view in that - foraging behaviour is often affected by the individual forager's perception of risk, not by any measure of absolute risk. In population studies (e.g. Werner et al. 1983; Gilliam and Fraser 1987; Werner and Hall 1984; Johannes et al. 1989), the risk of predation has often been measured as the probability of an animal being killed by a predator. Recent pond studies (Mittelbach 1986 and 1988; Mittelbach and Chesson 1987) have demonstrated changes in the foraging behaviour and habitat utilization of fishes in the presence of potential predators, but in the absence of predator-derived mortality over extended periods. Actual predation is not necessary to effect changes in foraging behaviour. The 92 perception of risk may be a more important determinant of foraging behaviour than the more classic view of predation risk. The term perceived risk has recently begun to receive more use in the behaviour literature (Lima and Dill 1990). Perceived risk must be viewed as especially important within the context of this thesis. First, no actual predators of juvenile chinook were used during any of the investigations documented herein, although model predators were used in Chapter 5.2. Second, the observed relationships of turbidity on foraging rate of chinook for surface, planktonic, and benthic prey are attributed to perception of risk decreasing as a function of turbidity. And third, the differences exhibited in these latter relationships by planktonic foraging when compared with surface and benthic foraging, may be explained in terms of perceived risk. This chapter will describe the manner in which perceived risk and visual ability may be behaviourally "traded of f " in turbid waters to produce the relationships documented in Chapter 4.2. Chapter 6.2 THE EFFECT OF VISUAL ABILITY ON FORAGING RATE Visual ability has been variously defined in the literature. The acuity of the eye is the minimum angle subtended by an object on the retina which a subject is capable of discriminating (Tamura 1957; Li et al. 1985). The contrast threshold is defined as the minimum difference in illumination between a target and background that an individual can distinguish (Hester 1968). Whereas both of these features of vision are important components of the visual ability of fishes in clear waters, the latter assumes a more prominent role as turbidity levels increase (Duntley 1943 and 1963; DiToro 1978). Because of increasing veiling brightness in the foreground, suspended particles in turbid water act to obscure objects that would be visible in clear water at equivalent light 93 conditions. The net result of these properties is a reduction in the visual ability of fish in turbid conditions. In Chapter 4.1, I demonstrated the effects of turbidity on the visual ability (reaction distance) of juvenile chinook salmon to Artemia prey. Turbidity effects a reduction in the reaction distance of foraging fish for their prey. Reaction distance declines log-linearly with increasing suspended sediment concentration (Vinyard and O'Brien 1976; Confer et al. 1981; Crowl 1989; Ch. 4.1, Fig. 4.2). The general effect of turbidity on reaction distance (Fig. 6.1) can therefore be summarized by: RD a f(TURB), where, RD = reaction distance, and TURB = turbidity. The probability of encounter will also decline with turbidity, because encounter rate is a function of the reaction distance: P(prey encounter) a RD, and also, P(prey encounter) a 1 - f(TURB). Conclusions concerning the effect of reaction distance on foraging rate have been varied. The generally held view has been that foraging rate will increase as some function of the volume or area of effective search, depending on the type of forager and prey. In the basic theoretical construction of his "disc" equation, Holling (1959) indicated the rate of successful search is highly dependant upon the distance at which a foraging animal responds to the presence of its prey (i.e. the reaction distance). Other important components of search, including any recognition or decision time required prior to an attack may also be incorporated into this rate. In general, foraging rate is Figure 6.1. The e f fec t of turbidity on visual ability (VA). 94 Turbidity 9 5 likely to increase as some function of the reaction distance (Ware 1 9 7 1 and 1 9 7 3 ) (Fig. 6 . 2 A ) ; however, this relationship may be asymptotic in many instances where the Holling "disc" equation is applicable (Fig. 6 . 2 B ) or is not truncated to the ascending limb of the latter relationship (e.g. at reduced reaction distances). The upper limit would be defined by the density of prey and the handling time. Success rate could be expressed as a probability rather than an absolute value in predictions derived by the "disc" equation or simpler linear or log-linear relationships. The relationship of probability of feeding (F V A) with reaction distance can be approximated from the following well-established principles: P(Fy^) a P(prey encounter), and P(prey encounter) a RD, therefore, P ( F V A ) A R D > or alternately, P ( F V A ) A 1 " F ( T U R B ) « where, P(F V A) is the probability of feeding, given visual ability ( V A ) . This relationship is simplistic. However, in this form the burden of assumptions concerning exact rates and various prey densities are effectively removed. Chapter 6 . 3 THE EFFECT OF PERCEIVED RISK OF PREDATION ON FORAGING RATE The results of Chapter 5 . 2 and several authors (White 1 9 3 6 ; Ginetz and Larkin 1 9 7 5 ; Gradall and Swenson 1 9 8 2 ; Simenstad et al. 1 9 8 2 ; 96 Figure 6.2. The e f fec t of visual ability on: A. Probability of foraging, P(FyA), and B. Foraging rate using Holling's (1959) "disc" equation solved for search rate (y. is foraging rate; x is prey density; b is the handling time; T is the time available for foraging; a is the search rate). Search Rate (a) 97 Gregory pers. obs.) suggest that turbidity effects a reduction in the risK or the perceived risk (PR) of predation in juvenile chinook salmon: P(predator encounter) a 1 - f(TURB), and, PR cc P(predator encounter), therefore, PR a 1 - f(TURB). All of these observations have been anecdotal or were implied as reasonable conclusions, except those of the present study (Ch. 5.2). Even my experimental manipulations assume that perceived risk declined as a function of turbidity (Fig. 6.3). This assumption will be further discussed in the Synopsis of this chapter and will be tested in Chapter 7. The effect of predation risk on foraging rate has been much studied in the last decade, both directly and indirectly. There have been many reviews on this topic (e.g. Dill 1983 and 1987; Werner and Gilliam 1984; Milinski 1986; Lima and Dill 1990). A synopsis of this work can be summarized by the simple statement: in the presence of increased predation risk, the foraging rate of an animal declines. The nature of the decline however, is not entirely clear. Sih (1986) observed that movement of one species of mosquito larvae (Culex pipiens) decreases with increasing number of predators (Notonecta undulata) in a log-linear fashion, while that of another mosquito species (Aedes  aegypti) decreased linearly. The connection between these results and foraging rates by mosquito larvae were not investigated. Two innovative investigations (Abrahams and Dill 1989; Nonacs and Dill 1990) provide approximations of the cost of avoiding predation in energetic terms in guppies (Poecilia reticulata) and ants (Lasius pallitarsis), Turbidity 99 respectively. Unfortunately, in both studies only two risk levels were tested. Therefore, the functional nature of the relationship cannot be deduced from these studies. Gilliam and Fraser (1987) introduced a "minimize p/f" (deaths per unit energy) model which may ultimately prove useful in this regard. As predation risk increases, the probability of foraging will decline. Or perhaps more correctly, the probability of attempted foraging will decline. The nature of- this decline is assumed here to be linear or log-linear (Fig. 6.4A): P(F P R ) a 1 - f(PR), where, P(Fpp) is the probability of foraging given the perceived risk. Unlike the logic presented for the similar conclusion regarding foraging rate and visual ability, the present assumption cannot be empirically supported. It represents a theoretical "best guess", given available information. An asymptotic or reversed sigmoid relationship will result if the juvenile chinook respond in a manner sensitive to the most likely functional response of theoretical predators (Fig. 6.4B). These relationships simply state that the probability of successful capture of prey is inversely proportional to the perceived risk, which in turn is proportional to the visual ability of the predator. Chapter 6.4 THE TRADEOFF The relationships described in the two preceding sections provide intuitive outcomes of well-established working principles. The only relationship for which I have provided detailed evidence is that for the e f fec t of turbidity level on the reaction distance. However, this is where the apparent simplicity ends. Visual ability and perceived risk 100 Figure 6.4. The effect of perceived risk of predation on A. Probability of foraging P(FpR), and B. Foraging rate using Holling's (1959) "disc" equation, solved for hypothetical ef fects of perceived risk on search rate. Perceived Risk 101 both directly affect foraging rate, the former as a sensory constraint and the latter as a behaviour-modifying variable. Their effects conflict. At low turbidity levels, visual ability is high; correspondingly high foraging rates should be expected in the absence of a conflicting variable. However, perceived risk is also potentially high at low turbidities; therefore foraging rates should be reduced. The reverse of both of these will be manifested at high turbidity levels. The result of this tradeoff (Fig. 6.5) will be high foraging rates at conditions intermediate between two turbidity extremes. Clearly, at high turbidity levels (1800 mg-L-1 - Chapter 4.2) the visual ability may be reduced to an extent were perceived risk is no longer relevant. A behavioural decision by an animal which perceives the risk of foraging to be too high may manifest itself either as the reduction or cessation of foraging activity. The act of foraging itself can be an inherently dangerous activity for salmonids (Donnelly and Dill 1984; Dill 1987). Low foraging rates may be observed in visually superior conditions if these conditions also promote higher perceptions of risk. In such a situation, an animal may forage more actively in visually inferior conditions. The predictions of the model depicted in Figure 6.5 were generated by incorporating the previously described assumptions into a multiplicative probability model. It predicts foraging rate (FR) as a relative probability, given the probabilities of foraging with the current visual ability and perception of risk: FR a P(FVA)-P(FPR). This model clearly represents a simplification of a complicated behavioural process. However, in the form I have presented here, hypotheses concerning the implications of changing perceived risk in turbid and other visually restricted media may be constructed and 102 Figure 6.5. A conceptual model of the effect of a behavioural tradeoff between visual ability and perceived risk on relative foraging rate in turbid waters. FR a P ( F V A ) - P ( F p n ) 103 experiments designed to test them. Two of these hypotheses are outlined in the subsequent section. Chapter 6.5 PREDICT IONS O F T H E MODEL AND R E - I N T E R P R E T A T I O N O F FORAGING R A T E S The Effects of Prey Quality and Microhabitat The foraging rates of juvenile chinook salmon demonstrated in Chapter 4.2 were consistent with the predictions of the above outlined tradeoff model. In predicting the relationship between turbidity and the foraging rate of salmon for benthic and surface prey, the model forces us to make an assumption concerning perceived risk. Perceived risk must be relatively high while foraging on these prey (at least in clear water) relative to planktonic prey. Observed foraging rates on planktonic prey may be realized only if perceived risk is relatively low when foraging in clear water. If the assumptions of the model are accepted, that visual ability operates as a constraint on foraging behaviour while perceived risk functionally acts as a flexible parameter, then these apparently confusing results may be explained. The model makes either prey- or microhabitat-specific predictions about foraging behaviour (e.g.. planktonic versus benthic foraging). Juvenile chinook may perceive less risk while foraging on planktonic prey relative to either of the other two prey types. The model predicts relatively higher foraging rates at low turbidity levels for planktonic prey than for either benthic or surface prey (Fig. 6.6). A re -examination of the results of Chapter 4.2 indicates a strong similarity to these predicted relationships. Juvenile chinook salmon are opportunistic foragers (Dunford 1975; Figure 6.6. Predicted relative foraging rates at two levels of perceived risk. 104 Turbidity 105 Levy et al. 1979; Healey 1982; Simenstad et al. 1982; Gregory pers. obs.). Individuals are morphologically equipped to forage on a wide variety of prey. However, various locations in the water column will be occupied at different degrees of risk to predators. The surface of the water column may readily be visualized as being inherently "dangerous" when compared to planktonic and benthic microhabitats. Occupation of near- surface waters subjects a foraging fish to both exposure to predatory birds and also to the possibility of attack from below by piscivorous fish. The occupation of near-bottom microhabitat places a juvenile at more risk to sculpins (ambush predators), especially from behind while foragers are actively feeding. Planktonic foraging may represent the least inherently "dangerous" portion of the water column. Alternately, the fish may have responded to a quality difference in the prey. Evidence demonstrating the tendency for foragers to accept higher risk to gain greater rewards (usually measured as energy intake) have been numerous (for reviews see: Dill 1983 and 1987; Werner and Gilliam 1984; Milinski 1986; Sih 1987; Lima 1988; Werner and Hall 1988; Lima and Dill 1990). This body of literature would seem to suggest that the relationship exhibited by juvenile chinook foraging on planktonic prey (Ch. 4.2) could be explained on the basis of a prey quality difference between Artemia and the two other prey species (Drosophila and Tubifex). Foraging juvenile chinook salmon may exhibit higher foraging rates on Artemia prey because a lower cost:benefit ratio exists for this prey as compared with Drosophila or Tubifex prey. An alternate hypothesis must also be considered. The location or microhabitat occupied by the prey may affect a forager's perceived risk while feeding on this prey. Microhabitat determines of the risk a forager is exposed to while foraging, usually in relation to distance from cover (e.g. Lima 1988, Magnhagen 1988, Dill and Fraser 1984). These two conflicting hypotheses (quality versus microhabitat) may be tested by switching the microhabitats in which the different prey types are found. 106 Under the prey quality hypothesis, the foraging behaviour of juvenile chinooK salmon will be similar toward prey of equivalent quality irrespective of microhabitat. In contrast, the microhabitat hypothesis predicts that equivalent quality prey will elicit behaviour specific to the microhabitat. The results of an experiment to discriminate between these two hypotheses is presented in Chapter 7.1. The Effect of Enhanced Risk A major assumption of the tradeoff model is that foraging animals perceive (i.e. assess) risk, and that this perception of risk declines with reduced visibility. However, perceived risk may not reflect the actual risk of mortality. Enhancing the risk stimulus should not affect foraging rate to as great an extent at high as at low turbidity levels. This assumption may be tested by increasing the actual risk stimulus to the forager. Given this assumption, any non-visual stimulus effecting changes in foraging behaviour in clear conditions must elicit a change in foraging rate which is mitigated by the presence of turbidity. This must occur because perceived risk will decline as the turbidity level increases (Fig. 6.7). An experiment to test this assumption is presented in Chapter 7.2, involving an audible fright-stimulus. The predicted results of such an experiment are that foraging rate will be reduced at low turbidities, but progressively less reduced at higher turbidity levels (i.e. the enhanced risk and "control" curves will converge - Fig. 6.7A). Should the assumption prove false, there are two possible outcomes, reflecting some fixed cost of risk avoidance - either a constant or a proportional reduction of foraging rate across all turbidity levels (Fig. 6.7B). Although measuring the energetic cost of predator avoidance behaviour has been attempted with some degree of success (Godin and Smith 1988; Abrahams and Dill 1989; Nonacs and Dill 107 F i g u r e 6.7. H y p o t h e t i c a l e f f e c t o f enhancing r i s k stimuli on f o r a g i n g r a t e f o r p lank ton i c p r e y ( see C h a p t e r 4.2). A. Assuming d e c r e a s e in p e r c e i v e d r i s k a s a f u n c t i o n o f tu rb id i t y . B. Assuming a d e c r e a s e in f i x e d - r i s k a s a f u n c t i o n o f t u r b i d i t y . A 'Normal R i sk ' Turbidity B 'Normal R i sk ' Turbidity 108 1990), these f ixed c o s t s were not est imated as a funct ion o f r i sk . This work has yet to be p e r f o r m e d but would be invaluable in assess ing the c o s t s o f forag ing in d i f f e r e n t environments. The relationship between turbidity and foraging r a t e exhibited by chinook f o r planktonic p rey is ideal f o r tes t ing assumptions about perceived r i sk . C lear predict ions o f re lat ive forag ing r a t e between turbidity levels can be establ ished (Fig. 6.7). This may not be the case f o r s u r f a c e and benthic forag ing (Fig. 6.8). The re lat ive r a t e s exhibited by chinook foraging on these p rey may be too d e p r e s s e d at the turbidity levels where the g r e a t e s t e f f e c t s would be expected (i.e. 0 mg-L - 1 ) . C h a p t e r 6.6 S Y N O P S I S This chapte r p roposes a probabil ity model to descr ibe a t r a d e o f f between visual abil ity and r i s k p e r c e p t i o n in predict ing forag ing behaviour . Al though spec i f i ca l l y c r e a t e d t o explain forag ing behaviour by juvenile chinook salmon in turbid waters , the model is general . This model is potent ia l ly usefu l in making predict ions in o t h e r t r a d e o f f s where an independent var iable (e.g. tu rb id i t y , light) has the opposite e f f e c t on two o r more variables (e.g. sensory ability, perceived r isk) , which in t u r n have opposing e f f e c t s on some behaviour (e.g. foraging, mating behav iour ) . 109 Figure 6.8. Hypothetical ef fect of enhancing risk stimuli on foraging rate for surface and benthic prey (see Chapter 4.2). A. Assuming decrease in perceived risk as a function of turbidity. B. Assuming a decrease in fixed-risk as a function of turbidity. A Turbidity B Turbidity 110 CHAPTER 7 TESTS OF THE VISUAL ABILITY-PERCEIVED RISK TRADEOFF Chapter 7.1 THE EFFECT OF MICROHABITAT AND PREY QUALITY ON FORAGING DECISIONS INTRODUCTION In Chapter 4, I demonstrated that the relationship between turbidity and the foraging rate of the juvenile chinook salmon for planktonic prey was different from that exhibited for benthic and surface prey. The foraging rate of chinook was observed to decline at high turbidities for all three prey types. However, chinook foraging rates for planktonic prey were consistently high in low turbidity conditions (<50 mg-L-1), while they were often reduced at these levels for benthic and surface prey. There are two potential explanations for this disparity provided by the tradeoff model (Ch. 6). Dissimilarities could be explained by, 1. the quality of the different prey or, 2. the perceived risk associated with the microhabitat where the prey were found. In this study, I test the prey quality and the microhabitat hypotheses. Prey quality has received considerable attention by researchers. Foragers have been shown to move farther from protective cover, endure higher risk of predation, or exert more energy to obtain higher quality prey (Dill and Fraser 1984; Milinski 1986; Real and Caraco 1986; Magnhagen 1988; Lima and Dill 1990; Nonacs and Dill 1990). Also, foragers under varying risk demonstrate different prey preferences (Lima 1988). Dill and Fraser (1984) and Magnhagen (1988) describe studies showing this type of behaviour in several salmonid species. 111 The perceived risk of a foraging fish to predators has been demonstrated to change over relatively short distances (Werner et al. 1983), often with no actual predatory act occurring (Mittelbach and Chesson 1987). The perceived risk of salmon foraging on prey either at the surface or bottom could be different from that while foraging in a pelagic microhabitats. Risk sensitive feeding mode changes have been observed (Werner and Gilliam 1984; Milinski 1986). Juvenile chinook may feed more readily in the water column than at the surface or bottom because of a lower perception of risk while foraging planktonically. In this study, I describe an experiment in which I "switched" planktonic Artemia prey with surface Drosophila prey such that the former was artificially made to float and the latter to sink. In a reciprocal treatment design, the effect of food quality was effectively isolated from that of prey location. Given, the conceptual tradeoff model described in Chapter 6, the prey quality and microhabitat hypotheses make conflicting predictions concerning chinook foraging rates on these manipulated prey. The prey quality hypothesis predicts similar or identical feeding responses regardless of the position of the prey in the water column (i.e. whether it is a surface or a planktonic prey). Artemia and Drosophila prey should elicit species-specific foraging responses. However, the microhabitat hypothesis predicts similar feeding responses regardless of the prey species. Chinook foraging rates on floating Artemia prey should be similar to those exhibited toward "normal" Drosophila prey. Also, planktonic Drosophila are expected to elicit a similar feeding response in foraging chinook as do "normal" Artemia prey. The prey quality and microhabitat hypotheses make clear and conflicting predictions. 112 M E T H O D S The juvenile chinook used throughout this experiment were 52.7 -54.9 mm FL in mean size. Within trial individuals ranged +4 mm FL. The basic experimental procedure was the same as that described previously (Ch. 3). I have mentioned only the necessary changes here. Both Artemia and Drosophila prey were used in this experiment but were prepared in such a way as to make the former float and the latter sink. This was facilitated in Artemia by air drying individual prey items over a period of several hours. The resulting prey item would float for up to an hour, before again sinking. Sinking Drosophila prey were pretreated by placing an appropriate number of prey in a jar of 10 mg-L-1 Polysorbate-80 overnight. Polysorbate-80 is an emulsifying agent used commercially as a food additive. In the present context, it was used to reduce the surface tension of water, allowing the prey to absorb water and sink. In preliminary trials, no effect of Polysorbate-treated food (Oregon Moist Pellets) could be demonstrated on either growth or feeding preferences in juvenile chinook. Preliminary conditioning of the test fish to the experimental prey was adjusted from that previously described (Ch. 3), reflecting the two additional prey categories. Test fish also continued to be fed Tubifex during the pre-experimental conditioning. Artemia were "heat-killed"; no live prey were used in this experiment. The experimental design consisted of 12 trials at each of the seven turbidity levels 0 - 800 mg-L-1. A total of 320 individual prey items were used in each treatment of each trial. Two trials each, using planktonic Artemia and surface Drosophila prey, were conducted as "control" trials. Three trials each, using "switched" prey (i.e. surface Artemia and planktonic Drosophila) were performed as "treatment" trials. Trials on surface prey and planktonic prey were of 10.0 and 1.0 minutes duration respectively. This procedure was appropriate to the 113 treatment of these prey categories in other experiments (see Ch. 3). In addition, one further trial for each of Artemia and Drosophila was carried out at 10.0 and 1.0 minutes experimental feeding times, respectively. These additional trials allowed description of the effect of the different feeding times on the various prey categories. Statistical analyses in addition to those described in Chapter 3, included coefficient of variation and the F-statistic for the significance of dissimilarity of pooled variances (Zar 1984). Foraging rates of individual fish were log-transformed prior to the calculation of variances by treatment and trial, in order to satisfy the assumptions concerning normality made by these two statistics. R E S U L T S Chinook salmon foraging at the surface (Fig. 7.1A) or in the water column (Fig. 7.1B) exhibited similar foraging rates at each turbidity level, regardless of the prey species. In all cases, I observed foraging rates consistent with those exhibited toward Artemia prey in previous experiments (Ch. 4.2). Foraging rates consistent with those observed in Chapter 4.2 for Drosophila (i.e. highest rates at intermediate turbidities), were not observed here. Given the size range of chinook used here, this result was not unexpected (see Ch. 4.2 -Effects of Ontogeny). Experimental duration consistent with the methods outlined for each prey type (i.e. 10.0 and 1.0 minutes for surface and planktonic prey, respectively), continued to be the prime determinant of the number of prey consumed in each experiment. Chinook exposed to surface prey for one minute consumed fewer prey than in either the 10.0-minute trial (as expected) or in trials with planktonic prey (Fig. 7.1). Similarly, fish exposed to planktonic prey for 10.0 minutes consumed more prey than either planktonic prey at a lesser amount of time or 114 Figure 7.1. The effect of turbidity on foraging rate in juvenile chinook salmon (vertical bars represent standard error of mean of trial means). A. Surface foraging (prey in 10.0 minutes) f o r Drosophila (thin solid line) and dried Artemia prey (broken line), and for Drosophila (prey in 1.0 minutes - thick solid line). B. Planktonic foraging (prey in 1.0 minutes f o r Artemia (thin solid line) and Polysorbate-80 treated Drosophila (broken line), and for Artemia (prey in 10.0 minutes - thick solid line). A Surface Prey 24 r d) 20 0 25 60 100 200 4 0 0 8 0 0 Turbidity (mg/L) B Planktonic Prey 241 0 25 50 100 200 4 0 0 800 Turbidity (mg/L) 115 surface prey at the same amount of time. Chinook consumed planktonic prey at a faster rate than surface prey, regardless of the prey species. Although the prey consumed was quantitatively different with respect to treatment, the qualitative relationship with turbidity among the various prey categories generally exhibited similar overall form (Fig 7.1). However, coefficients of variation within the individual trial treatments were generally higher for surface prey than for planktonic prey (usually above and below 20/, respectively - Fig. 7.2). Again, I found this to be true regardless of the prey species (Fig. 7.2A Artemia; Fig. 7.2B Drosophila). Between prey species, differences in the pooled variances within each treatment were generally small, and usually insignificant. Chinook salmon exhibited patterns of variance unique to the microhabitat of foraging but not unique to species. The results presented in this subchapter are based on the means of individual trials, except when dealing with coefficient of variation. Details of these trials may be found in the appendices (Fig. A.4 and Fig. A.9). D I S C U S S I O N The microhabitat hypothesis - microhabitat effects changes in the foraging rate of juvenile chinook salmon, independent of prey species -cannot be rejected. Surface prey were fed upon at lower rates and in a more variable manner than planktonic prey irrespective of the species being preyed upon. These results were inconsistent with the predictions of the prey quality hypothesis. Foraging rates were not prey specific. Perceived predation risk due to the location, proximity, or accessibility of prey affects foraging rate in fish (Dill and Fraser 1984; Werner and Gilliam 1984; Magnhagen 1988; Mittelbach 1988). Various studies by Milinski (for review: Milinski 1986) demonstrate that 116 Figure 7.2. The effect of turbidity and microhabitat on the coefficient of variation {'/.) in foraging rate by juvenile chinook salmon in individual trials (square - surface foraging; circle - planktonic foraging; filled symbols represent "usual" position of prey, open symbols represent manipulated prey; NS, *,*»,*** - variance not significantly different [p>0.10], different at p<0.05, p<0.01, p<0.001, respectively). A. Artemia. B. Drosophila. A Artemia 100 c -B 8 0 I o .1 40 D o 0 D surface 0) 8*° • * g 1 » 5 planktonic • • • 1 1 1 1 25 50 100 200 Turbidity (mg/L) B Drosophila 120 C O100 I •§» 80 «*-o ^ 60 c 40 «•— 0) o O ao NS o surface o O 8 planktonic © o _ o 25 60 100 200 Turbidity (mg/L) 117 in clear water sticklebacks (Gasterosteous aculeatus) lower their feeding rates in the presence of high Daphnia concentrations. He attributed the change in behaviour to the increased risk associated with feeding on the denser plankton and the relaxation of vigilance required to forage effectively in such a patch. For generalist foragers, the present study suggests this principle also applies to foraging within fine scale microhabitats. The distances between forager and prey type were similar in this experiment; thus, distance to prey is unlikely to be a large source of variability. For the purpose of the present investigation, any residual cost of handling or accessibility was incorporated into the concept of prey quality. The most likely explanation of the results is that behaviour is influenced by the microhabitat occupied by the prey. Given the reciprocal manipulation, other explanations seem unlikely. Avoidance of high-risk areas by foragers has been demonstrated for a great number of species (for review: Lima and Dill 1990). Behaviour consistent with predator avoidance has also been demonstrated in cases where a predator was not actually detected (Mittelbach and Chesson 1987; Mittelbach 1988). Foraging animals may be expected to avoid inherently dangerous areas. Particular foragers may have a limited ability to escape some classes of predators. Such foragers may avoid, or limit their exposure, in areas accessible to that predator class. For juvenile chinook salmon, the surface of the water may represent an unsafe microhabitat. Not only are predatory birds numerous, but many piscivorous fish strike from below, using the background illumination to silhouette their targets (Munz and McFarland 1977; Guthrie 1986). Thus, the water surface potentially is a dangerous location for small foraging fish. The foraging rates, and associated large variances, of surface foraging chinook are consistent with the argument that f ry are sensitive to these dangers, regardless of whether they are real or perceived. Studies examining the decisions of foragers with respect to the 118 air-water interface are lacking. The present study breaks new ground in this area of fish foraging ecology. One of the expected results of this experiment had been a difference in the qualitative nature of the relationship between foraging rate and turbidity among the two prey or microhabitats. I did not observe this result. Given the size of the chinook used at the time of my experiments, this should not have been surprising (see Ch. 4.2 - Effects of Ontogeny). The observed qualitative similarity of planktonic and surface foraging rates should have been expected at this size. An experiment over a broad range of chinook sizes may prove useful in elucidating the effects of ontogeny on the use of microhabitat on finer scales than those previously identified by Werner and Gilliam (1984) and others. Chapter 7.2 THE EFFECT OF ENHANCED RISK ON FORAGING RATE IN TURBID WATER INTRODUCTION An established working principle has been that turbidity acts to reduce the predation risk to small fish from visually hunting predators (White 1936; Bruton 1979; Blaber and Blaber 1980; Simenstad et al. 1982; Cyrus and Blaber 1987a). Foragers, including fish, respond to increased risk of predation by altering their foraging behaviour, often at the expense of reduced energy intake (Dill 1983 and 1987; Werner and Gilliam 1984; Milinski 1986; Lima 1988; Lima and Dill 1990). Long-term field studies demonstrate that fish may dramatically alter their foraging behaviour in the presence of potential predators, even without actual predation events occurring (Mittelbach and Chesson 1987; Mittelbach 1988). That foraging fish change their behaviour in a manner that 119 reduces their energy intake suggests a sensitivity to the perception of risk. In the present investigation, I measure the foraging rate of juvenile chinook salmon at two levels of risk to determine the effect of enhancing the risk stimulus on foraging behaviour in turbid conditions. I test the assumption, made in Chapter 6, that turbidity reduces the risk perceived by foraging juvenile chinook. The prediction I made in Chapter 6 was that increasing a risk stimulus uniformly across a range of visibility conditions has proportionally less effect on foraging rate as turbidity increases. Foraging rates under enhanced, and baseline, risk conditions should converge with increasing turbidity. In this subchapter, I describe an experiment to test this prediction using surface, planktonic, and benthic prey. A risk stimulus was used to enhance the threat to foraging juvenile chinook salmon. The most common risk stimuli used in fish investigations are visual (for review: Dill 1987), including those of Chapter 5.2. However, since turbidity acts to reduce visual ability (Vinyard and O'Brien 1976; Confer et al. 1978; Chapter 4.1), this technique could not be used in the present investigation. I wanted to measure perceived risk using a fixed level of threat stimulus under varying turbidity conditions. Therefore, the stimulus had to be non-visual. From a number of potential mechanical, chemical, electrical, and auditory stimuli, I chose sound. Low frequency sound is part of the sensory repertoire of salmon (Hawkins 1986). Many potential predators on salmonids in the marine environment, particularly gadoids (e.g. cod, haddock, pollock), use sound in courtship and intraspecific threat displays (for review: Hawkins 1986). The sound signals of potential predators may be detected by salmon, influencing subsequent behaviour. In the present experiment, sound was an ideal stimulus because its effect on fish behaviour would not be altered by visual considerations, only by the perception of the 120 magnitude of the risk. M E T H O D S Tubifex, Artemia, and Drosophila were used as prey in separate experiments. Each of these experiments consisted of a two-way factorial design with two levels of risk ("control" and "enhanced risk" trials) and the previously described 7 levels of turbidity (0 -800 mg-L-1; Ch. 3). Three replicates were performed for each level of risk and turbidity for all prey except Drosophila, where there were two. Trials were run by risk level, with all levels of turbidity run simultaneously at each level of risk. The separate risk trials were run sequentially and were blocked into control-risk pairs. The risk stimulus was a 10-watt, 30-Hz, square-wave sound signal (i.e. low frequency clicks) originating from two sound transducers in each aquarium (Fig. 7.3). Each transducer consisted of a 7.5 cm, 8-ohm speaker in a watertight whirlpak bag. The pair of transducers in each aquarium was arranged in a parallel electrical circuit and were activated from a remote location during trials. They remained in place during all treatments, but were only activated during "enhanced risk" trials. Preliminary trials established that sound elicited a response in the chinook similar to that exhibited in the presence of predator models (Ch. 5.2). Test fish moved rapidly to the bottom upon exposure to sound " signals originating within the aquarium. During preliminary trials, test fish responded to the continuous presence of the stimulus for approximately 35 seconds and then resumed normal activity. The stimulus was presented in six randomly timed 5-second increments over the course of an experiment. This effectively lengthened the duration of the response. 121 122 Within the confines of the aquaria, the sound environment during enhanced risK treatments was probably complex. I did not attempt to describe these sound patterns other than to observe their effects on fish behaviour. The turbidity level was unlikely to have had a measurable effect on the sound stimulus for several reasons. First, the sound signal was at a longer wavelength than the size of the sediment particles. Second, the sound absorption coefficients of brick and concrete (similar to sediment in molecular structure) are low (0.02 -0.03). Third, the density of the medium at 800 mg*L~1 was only 0.02/. higher than clear water. All these considerations indicated that negligible changes in sound quality probably occurred over the turbidity ranges tested (Giancoli 1980). The mean size of juvenile chinook used in these experiments increased over the testing period. Fish sizes were 59.5+4.0, 64.4+3.7, and 66.6+4.9 mm FL (+ s.d.) for Tubifex, Artemia, and Drosophila trials, respectively. No significant "within prey type" differences in fish size were found among the risk or turbidity treatments. R E S U L T S The effect of the enhanced risk stimulus on the foraging of juvenile chinook on surface prey (Drosophila) was unclear. The general form of the relationship between foraging rate and turbidity was similar to that documented in Chapter 4.2. Foraging rates were highest at intermediate turbidity conditions. This result was also expected from the size range of the fish used throughout these experiments (see Ch. 4.2 - Effects of Ontogeny). Generally, the foraging rates in the enhanced risk treatments were lower than the controls at most turbidity levels and for most individual trials (Fig. 7.4). The expected convergence of foraging rates was observed at higher turbidities (100 -123 Figure 7.4. The effect of turbidity and risk on mean foraging rate on surface prey (number of Drosophila in 10.0 minutes) by juvenile chinook salmon (control risk - solid line; enhanced risk - broken line; vertical bars indicate standard error of mean of trial means). 0 25 50 100 200 400 800 Turbidity (mg/L) 124 200 mg-L-1). However, chinooK foraging rates at 0 mg-L-1 were not as expected. Foraging rates were higher in the enhanced risk trials than in controls. However, this observation was not inexplicable. The effect of enhanced risk on foraging rate may be expected to be dampened where the foraging rates were lowest (i.e. low and high turbidity levels in this prey type). Inconsistencies in the results were more likely at these levels. Even random variation may appear to affect inconsistent results. The foraging rates of salmon on planktonic Artemia supported the perceived risk hypothesis. Regardless of the risk treatment, the general form of the relationship between foraging rate and turbidity was similar to my other results for this prey type (Ch. 4.2, 4.3, and 7.1). Foraging rates were highest in low turbidity treatments (1200 mg*L~1). At most turbidity levels, I found the foraging rates of the enhanced risk treatments were lower than those of the control treatments (Fig. 7.5). This difference was significant at 0 mg-L-1 (one-tailed t-test, p<0.05). At higher turbidity levels, foraging rates in the enhanced risk and control treatments converged. Foraging rates of juveniles on benthic Tubifex prey also supported the perceived risk hypothesis. The relationship between foraging rate and turbidity again was of the form observed in Chapter 4.2. Rates were highest in intermediate turbidity conditions (25 - 100 mg-L-1), regardless of the risk treatment. At all turbidity levels, I observed lower foraging rates in the elevated risk treatments (Fig. 7.6). The largest differences between the elevated risk and control treatments were observed at lower turbidities (<100 mg-L-1). Again, as expected from the perceived risk hypothesis, I observed a convergence of the rates of these treatments at higher turbidity levels. The perceived risk hypothesis (Ch. 6.5) predicts that differences in foraging rate between enhanced risk and control treatments will decline with increasing turbidity. For chinook foraging on planktonic 1 2 5 Figure 7.5. The effect of turbidity and risk on mean foraging rate on planktonic prey (number of Artemia in 1.0 minutes) by juvenile chinook salmon (control risk - solid line; enhanced risk - broken line; vertical bars indicate standard error of mean of trial means). 126 Figure 7.6. The effect of turbidity and risk on mean foraging rate on benthic prey (wet weight of Tubifex in 5.0 minutes) by juvenile chinook salmon (control risk - solid line; enhanced risk - broken line; vertical bars indicate standard error of mean of trial means). 14 r C E 12 0 25 50 100 200 400 800 Turbidity (mg/L) 127 (Artemia) prey, a regression of these differences was significant (ANOVA; df=1,13; p=0.023; r2=0.34). The negative slope demonstrated by this relationship (Fig. 7.7B), indicated a convergence between the enhanced risk and control treatments (i.e. approaching a difference of zero). Similar results were not demonstrated by chinook foraging on surface (Drosophila) or benthic (Tubifex) prey (Fig. 7.7A and 7.7C, respectively). The perceived risk hypothesis was supported by chinook foraging on planktonic prey. Here, I have reported results based on the means of treatments within trials. The details of each individual trial are recorded in the appendices (Fig. A.5, A.10, and A.14). D I S C U S S I O N Fish generally respond to only low frequency sound (<300 Hz). By comparison, mammals and birds exhibit a much broader sensitivity range (e.g. to 15 kHz in humans; >300 kHz in bats and some marine mammals). However, within their relatively restr icted detectability range, fish are quite sensitive to sound (for review: Hawkins 1986). Atlantic salmon (Salmo salar - Hawkins and Johnstone 1978) are sensitive to sound pressure in the "near-field" (<30 to 300 Hz frequencies) but not to "far-field" sound. The exact biological significance of near- versus far-field sensitivity is not clear. I suggest that detection of proximal movement may be more important to these fish (and perhaps other salmonids) given their more confined freshwater habitats. It may also represent a type of sensitivity useful for predator detection. In the present study, juvenile chinook rapidly moved to the bottom of the tank in response to the activation of the 30-Hz, sound stimulus. This response was reminiscent of the response exhibited by chinook to the presence of predator models (Ch. 5.2). Although direct links with 128 Figure 7.7 The e f f ec t of turbidity (logrmg-L"1]) on the difference between control and enhanced risk foraging rates by juvenile chinook salmon (data points indicate [control minus enhanced risk] for the means of each pair of trials). A. Surface foraging - Drosophila. B. Planktonic foraging -Artemia (regression significant p=0.023). C. Benthic foraging - Tubifex. -18 I I I I 0 0.5 1 1.5 2 2.6 Log(Turbidity[mg/Ll) 129 predator avoidance and sound detection in fish have not been demonstrated to my knowledge, such a response may be adaptive. Many potential fish predators, such as gadoids, produce sound detectable to salmon (Hawkins 1986). Avoiding areas where such sound originates would potentially reduce predation risk. I recommend this as an area for future research. At increasing turbidity levels, foraging rates by juvenile chinook for planktonic prey became progressively more similar in control and enhanced risk conditions. I suggest, from this evidence, that foraging salmon were responding to a reduction in their perception of risk in the more turbid conditions. This may have resulted in their feeding at higher rates on planktonic Artemia prey. Effects of enhanced risk on surface and benthic foraging by chinook were inconclusive. However, reduced foraging rates were expected at low turbidity, regardless of the risk treatment (Ch. 6.5). At low turbidity, the overall effect of enhancing the risk stimulus was probably masked. Therefore benthic and surface foraging rates exhibited by chinook were at least consistent with the perceived risk hypothesis. I suggest that turbid conditions manifest a relaxation in the vigilance requirements of foraging chinook salmon which would normally required in clear water. Chinook may be more motivated to feed in turbid conditions. If prey density is high (as may often be the case in estuarine habitats), this combination of food supply and motivation may result in higher daily foraging rates and growth rates in turbid conditions. Levy and Northcote (1982) demonstrate high growth rates in high prey density, turbid conditions in the Fraser Estuary. Furthermore, my own work in the Fraser River system suggests that prey densities need not be higher under turbid conditions to effect an increase in foraging rate when compared to clear water environments (Gregory pers. obs.). I have also observed that juvenile salmonids in turbid estuarine conditions feed throughout the daylight hours. This is 130 generally not the case in most clear water systems, where feeding behaviour exhibits marked diurnal patterns (Adams et al. 1987; Levy 1987; Angradi and Griffith 1990). Reduced perceived risk in turbid waters may result in higher chinook foraging rates. Perceived risk has an impact on the foraging behaviour and habitat choices of fish (Mittelbach and Chesson 1987; Mittelbach 1988). Furthermore, these changes in foraging behaviour were demonstrated to effect a reduction in the energy intake and subsequent growth of bluegill sunfish (Lepomis macrochirus). From the results of these and my own investigations, I suggest foraging behaviour may be affected by perceived risk, not necessarily by the probability of predation itself. This suggestion may be particularly relevant in the cases involving ambush predators, which may not be seen prior to a strike. Perceptions of risk related to microhabitat may be all that forearm a foraging juvenile chinook salmon against possible ingestion by aerial or benthic predators. In turbid conditions, these predators may have their potential attack range effectively reduced. As it may be difficult to avoid capture by these species once an attack is initiated, the avoidance of detection attains high importance. Turbid conditions may be likely to reduce such detection and subsequently reduce encounters with predators. Foraging behaviour may be expected to change to reflect this reduction in perceived risk. These ideas must be explored further. 131 CHAPTER 8 GENERAL DISCUSSION AND CONCLUSIONS Chapter 8.1 GENERAL DISCUSSION Vision as a Constraint The visual ability of fish, measured as reaction distance, decreases as a function of turbidity (Vinyard and O'Brien 1976; Confer et al. 1981; Ch. 4.1). Visual ability likely manifests itself as a constraint on foraging rate by directly affecting the rate of prey encounter (Ware 1971 and 1973; Confer and Blades 1975; Luecke and O'Brien 1981). In this manner, the negative impact of suspended sediment on fish vision is easy to appreciate and has been abundantly demonstrated in the literature. Turbidity and the Risk of Predation The impact of turbidity on the predation risk of juvenile fish has not been convincingly demonstrated in past work, although anecdotal and correlative information exists from many field studies (White 1936; Ginetz and Larkin 1975; Bruton 1979 and 1985; Blaber and Blaber 1981; Simenstad et al. 1982; Cyrus and Blaber 1987a) and one laboratory investigation (Gradall and Swenson 1982). The studies I have presented examine the perception of risk in turbid and clear water in controlled laboratory experiments. The results of Chapters 5 and 7 suggest that juvenile chinook salmon perceive less risk in turbid water environments. My work supports the conclusions of the above investigations and speculations. 132 Tradeoffs Between Visual Ability and Perceived Risk in Fish Foraging Behaviour Visual constraints and reduced perception of risk may be behaviourally "traded off". The results of my work (Ch 4.2), and that of Boehlert and Morgan (1985) and Neverman and Wurtsbaugh (in prep.), demonstrate that young fish exhibit foraging rates which suggest t radeoffs relating to visual ability and some other variable. I suggest that in all three cases, perceived risk was the additional variable. Boehlert and Morgan (1985) suggested increased feeding rates by herring larvae (Clupea harengus pallasi) in turbid conditions might have been due to increased prey-background contrast within their limited reactive field to prey. I cannot entirely dismiss this possibility. Juvenile chinook salmon have demonstrated decreased reactive times to prey silhouetted against turbid backgrounds (Godin and Gregory in prep.). However, Giguere and Northcote (1987) have demonstrated that full guts in otherwise transparent Chaoborus larvae increase the likelihood of predation by fish. Larval herring are also transparent. The cost of increased foraging rates in terms of predation risk in herring would be reduced in turbid water. The visual contrast of a full gut would be reduced. Therefore, increased foraging rates in turbid water would not subject the herring larvae to the higher risk of predation which its full gut may suggest. Neverman and Wurtsbaugh (in prep.) observed peak foraging rates by young-of-the-year sculpins were highest at intermediate light levels, while reduced at higher and lower light levels. I suspect a similar tradeoff in these results also. However, my own work suggests that light levels are a poor predictor of foraging rate, except on a coarse scale. Low light levels may reduce the perceived risk in some species (sockeye salmon - Clark and Levy 1988; sculpin - Neverman and Wurtsbaugh in prep.), whereas turbidity may perform this function in others (tilapia - Bruton 1979; estuarine young-133 of-the-year fish - Blaber and Blaber 1981, Cyrus and Blaber 1987a; juvenile chinook salmon - Ch. 5 and Ch. 7). Turbid Water Foraging in Fishes Turbidity has often been viewed as detrimental to the foraging activities of fishes (Ellis 1936; Alabaster 1972; Bruton 1985; Gardner 1981) especially salmonids (Sigler et al. 1984; Berg and Northcote 1985; Confer et al. 1978). This concern has been justified. High levels of suspended sediment are deleterious to salmonid feeding rates, growth, and survival (Noggle 1978; Sigler et al. 1984; Servizi and Martens 1987). However, physiological responses to chronic exposure in fishes occur only at excessively high levels of turbidity, orders of magnitude beyond levels experienced naturally (Wallen 1951). In the face of this evidence, other studies (Blaber and Blaber 1980; Cyrus and Blaber 1987a; Bruton 1979) demonstrate that fishes, especially larval and juvenile forms, may actively seek out turbid waters, where they may exhibit higher survival rates. Further, the active preference of turbid over clear conditions has been demonstrated in the laboratory (Cyrus and Blaber 1987b; Gradall and Swenson 1982). These observations create an apparent paradox, which may be resolved by considering that fish may behaviourally trade o f f their risk of predation and visual ability in all of these situations. In an examination of the above field studies, I discovered a clear dichotomy among conclusions reached. All of the investigations which concerned human-induced turbidity (7 studies) reported negative effects; work on naturally turbid systems (4 studies) reported positive effects. While I have not conducted a comprehensive review of this topic, the dicotomy is suggestive. 134 Salmonid Life Histories Chinook salmon exhibit the most interpopulation variablity in life history strategy among all anadromous members of the genus Oncorhynchus (Taylor 1990). Basically, two chinook life history strategies have been described. The first, "stream-type", spend one year in or near the natal stream and migrate to sea as 1+ year old smolt. The second, "ocean-type" (normally from fall adult spawners), migrate to sea as underyearling fry but rear for a variable amount of time in the estuary. The Harrison River stock, from which the fish of this study originated, is "ocean-type". Although the duration of freshwater existence has been shown to be influenced by the growth potential in natal streams (Taylor 1990), it is not clear how turbidity or estuarine productivity influences life history strategy. Seaward migrating juvenile salmon demonstrate species- or population-specific estuarine residency periods. Although Healey (1982) and Simenstad et al. (1982) both categorize chinook salmon as the most estuarine dependant of the Pacific salmon species, both also described populations of this species which were transient estuarine residents. Estuarine residency in this species has been demonstrated to range from days to weeks, in 1+ year old smolt, and up to months in young-of-the-year individuals (Simenstad et al. 1982). Residency duration has been correlated with salmon size (Healey 1982). Many river systems supporting large salmon populations have turbid mainstem components (Squamish River - Levy 1977; Taku River - Murphy et al. 1989; Fraser River - Northcote and Larkin 1989). In the turbid Fraser Estuary individual juvenile chinook salmon have been resident for periods of up to two months (Levy and Northcote 1982) with residency within the estuary spanning March to July in any one year for juveniles of all salmonid species (Levy et a). 1979; Levy and Northcote 1982; Gregory pers. obs.). This residency period also corresponds to that of 135 peak discharge and highest turbidity level in the Fraser River. Prey availability may influence the residency time of salmonid juveniles within the estuary. Prey abundance in the Fraser River system has been observed to be several orders of magnitude higher in the estuary than at either downstream or upstream locations (Northcote and Larkin 1989). High foraging rates may be maintained in this system because of high food density, mitigating any negative impact of reduced visual ability. However, it has also been observed in the same river system that juvenile chinook found in clear water habitats with high prey densities (Harrison River) exhibit lower gut fullness when compared to nearby turbid areas with lower prey densities (Gregory pers. obs.). In the same study, predation on juvenile salmonids was observed in 10/ of older conspecifics in clear water. No evidence was found of this in turbid water. I suggest the presence of turbidity may have a more pronounced, and positive, impact on foraging rates than prey availability in this system. Prey density may only provide a partial explanation for the duration of residency of salmonids in the estuary. Perceived Risk in Behaviour Studies My results have obvious implications to the study of ecology and behaviour of all fish species found in turbid waters at various stages of their life history. Turbidity also a f fec ts ambient light conditions. My work (Ch. 4.3) implies that the effect of turbidity on chinook foraging behaviour may be correlated to ambient light conditions only on a coarse scale. This is unlikely to be universally true. The effect of light level has been especially well documented on coral reef communities (for review. Munz and McFarland 1977). Even the light related movements of other salmonids have been related to risk of predation (Levy 1987; Clark and Levy 1988). Also, similar relationships 136 to those I present in Chapter 4.2, have recently been demonstrated for juvenile sculpins over a range of light levels (Neverman and Wurtsbaugh in prep.). Perceived risk of predation may be a possible explanation for the relatively lower foraging rate of sculpins in higher light condit ions. Light effects are not confined to the aquatic environment. The foraging behaviour of many nocturnal animals may be reduced during periods of bright moonlight (for review: Lima and Dill 1990). Artificial increases in light level have also been demonstrated to reduce foraging rates by nocturnal animals (Kotler 1984; Brown et al. 1988). In all of these cases, exposure to predators in highly illuminated conditions has been suggested to reduce foraging activity. Terrestrial animals subject to predation may commonly assess their potential risk by some surrogate, such as the distance to cover (Schneider 1984; Elgar 1986). In some cases, the proximity to cover cannot be used as a measure of perceived risk. Cover can be a source of predation risk for some animals, such as marmots (Carey 1985) and African antelope (Underwood 1982), or a source of risk as well as an escape from it in others, such as some finches (Lima et al. 1987). Clearly, research must ultimately attempt to measure the perceived risk of foragers when investigating related behaviour. This will not be an easy task; often a given experimental animal may alter its behaviour in subtle ways when perceiving elevated risk. These subtleties may elude an investigator. The question of subtle effects may be relevant in the case of ambush predators. For many terrestrial and aquatic animals (including juvenile chinook salmon), ambush predators present an interesting dilemma. By the very nature of the hunting strategy of these predators, avoidance of predation by a forager would be best achieved by avoiding the encounter. Since these predators remain essentially hidden from the foraging animal until the attack distance is relatively short, it would 137 benefit the forager to measure risk in a manner that did not rely on the sighting of the predator itself. Of more importance to the forager may be its perception of risk, possibly genetically inherited or through learned experience of times, locations, or environmental conditions of inherently higher risk. For juvenile chinook salmon, ambush predators of two general categories - avian and aquatic - make such perceptions vitally important. Chapter 8.2 C O N C L U S I O N S The visual ability of juvenile chinook salmon, measured as reaction distance, declined with increasing turbidity, similar to that found previous investigations on other species (Vinyard and O'Brien 1976; Confer et al. 1978). None of these investigations, including my own, indicate a threshold type of decline in reaction distance for turbidity as has been demonstrated with decreasing light (Harden Jones 1956; Vinyard and O'Brien 1976; Confer et al. 1978). These results and those of my experiments comparing the effects of light and turbidity level on foraging behaviour (Ch. 4.3) suggest that turbidity cannot be viewed simply by examining its effect on ambient light conditions. However, for planktonic prey this may indeed be the case. Foraging rates by chinook on surface and benthic prey were dissimilar in turbid and equivalent light conditions. Visual ability was a poor predictor of foraging rate on surface and benthic prey by juvenile chinook. While it could be argued that planktonic foraging rates by chinook may have been loosely predicted by visual ability, this explanation is untenable for surface and benthic prey. An hypothesis based solely on visual effects is unlikely to predict increased foraging rates in intermediate turbidity as compared to clear water conditions. Visual ability alone cannot fully explain 138 the foraging behaviour of juvenile chinook. The positive effects of turbidity on perceived risk by chinook were suggested in experiments involving model predators. Avian and fish piscivore models elicited similar responses in juvenile chinook. In clear water, salmon moved into deeper waters closer to the bottom following exposure to the models. In turbid conditions, both the magnitude of the response and the recovery time of salmon following exposure were reduced. Experiments in Chapter 7, isolating the effects of the visual component of risk perception, suggest that a reduced response in turbid waters was not necessarily due to a failure to detect the presence of a predator. It appears likely that turbidity mitigates perceived risk in juvenile chinook. The more established view purporting detrimental effects of turbidity on visual foraging (for review: Bruton 1985) was re-examined and found to be inadequate. I propose a view of behavioural flexibility. The e f fec ts of turbidity on both visual ability and perceived risk must be appreciated to describe foraging behaviour. My findings regarding changes in foraging rate with ontogeny add support to a more behaviourally flexible response to turbid water conditions. Although their visual ability may be deleteriously affected by turbid conditions, juvenile chinook may be more motivated to feed because of their reduced perception of risk. Theoretical maximum feeding rates may not be attained on short time scales (minutes) for some prey due to visual constraints. However, foraging rates over longer time scales (diurnal) may be positively affected by turbidity. Fish perceiving less risk may be more willing to feed at high rates, and subsequently exhibit higher growth rates. Salmonids in clear waters tend to exhibit highest foraging rates at distinct times of the day (Adams et al. 1987; Levy 1987; Angradi and Griffith 1990). Chinook salmon in turbid estuarine waters forage actively throughout the day (Gregory pers. obs.) and exhibit elevated growth rates (Levy and 139 Northcote 1982). I suggest foraging animals may "trade of f " the conflicting goals of obtaining food and avoiding predators in conditions of limited visibility. In many cases, including turbid water, reduced visibility negatively affects the sensory abilities of predators as well as foragers (Hobson 1986; Lima and Dill 1990). I suggest foraging behaviour changes in varying visibility conditions in a manner sensitive to both of these positive and negative effects. Reduced visibility cannot be viewed simply as detrimental to a visually foraging animal. 1 4 0 R E F E R E N C E S Abrahams.M.V. and L.M.Dill. 1989. A determination of the energetic equivalence of the risk of predation. Ecology 70:999-1007. Adams,N.J., D.R.Barton, R.A.Cunjak, G.Power, and S.C.Riley. 1987. Diel patterns of activity and substrate preference in young Arctic char from the Koroc River, northern Quebec. Can. J. Zool. 66:2500-2502. Aggus,L.R. and G.V.Elliott. 1975. Effects of cover and food on year-class strength of largemouth bass. pp. 317-322. IN Stroud.R.H. and H. Clepper(Eds.). Black bass biology and management. Sport Fishing Institute, Washington. Alabaster.J.S. 1972. Suspended solids and fisheries. Proc. Roy. Soc. Lond. B. 180:395-406. Angradi.T.R. and J.S.Griffith. 1990. Diel feeding chronology and diet selection of rainbow trout (Oncorhynchus mykiss) in the Henry's Fork of the Snake River, Idaho. Can. J. Fish. Aquat. Sci. 47:199-209. Bailey,B.J.R. 1980. Large sample simultaneous confidence intervals for the multinomial probabilities based on transformations of the cell frequencies. Technometrics 22:583-589. Berg.L. and T.G.Northcote. 1985. Changes in territorial, gill-flaring, and feeding behaviour in juvenile coho salmon (Oncorhynchus  kisutch) following short-term pulses of suspended sediment. Can. J . Fish. Aquat. Sci. 42:1410-1417. Blaber,S.J.M. and T.G.Blaber. 1980. Factors affecting the distribution of juvenile estuarine and inshore fish. J . Fish Biol. 17:143-162. Blaxter.J.H.S. 1966. The ef fect of light intensity on the feeding ecology of herring, pp. 392-409. IN Bainbridge,R., G.C.Evans, and O.Rackham(Eds.). Light as an ecological factor. Blackwell Scientific Publications, Oxford. Blaxter.J.H.S. 1968. Visual thresholds and spectral sensitivity of herring larvae. J . Exp. Biol. 48:39-53. Boehlert,G.W. and J.B.Morgan. 1985. Turbidity enhances feeding abilities of larval Pacific herring, Clupea harengus pallasi. Hydrobiol. 123:161-170. Breck.J.E. and M.J.Gitter. 1983. Effect of fish size on the reactive distance of bluegill (Lepomis macrochirus) sunfish. Can. J . Fish. Aquat. Sci. 40: 162-167. 141 Breitburg.D.L. 1988. Effects of turbidity on prey consumption by striped bass larvae. Trans. Am. Fish. Soc. 117:72-77. Brett,J.R. and C.Groot. 1963. Some aspects of olfactory and visual responses in Pacific salmon. J. Fish. Res. Bd. Can. 20:287-303. Brett,J.R. and M.A.AIi. 1958. Some observations on the structure and photomechanical responses of the Pacific salmon retina. J. Fish. Res. Bd. Can. 15:815-829. Brown,J.S., B.P.Kotler, R.J.Smith, W.O.Wirtz. 1988. The e f fec ts of owl predation on the foraging behavior of heteromyid rodents. Oecologia 76:408-415. Bruton,M.N. 1979. The fishes of Lake Sibaya. pp. 162-245. IN: Allanson,B.R.(Ed.). Lake Sibaya. W.Junk, The Hague. Bruton,M.N. 1985. The effects of suspensoids on fish. Hydrobiol. 125:221-241. Buck,D.H. 1956. E f fects of turbidity on fish and fishing. Trans. N. Am. Wildl. Conf. 21:249-261. Carey,H.V. 1985. The use of foraging areas by yellow-bellied marmots. Oikos 44:273-279. Cerri.R.D. 1983. The effect of light intensity on predator and prey behaviour in cyprinid fish: factors that influence prey risk. Anim. Behav. 31:736- 742. Clark,C.W. and D.A.Levy. 1988. Diel vertical migrations by pelagic planktivorous fishes and the antipredator window. Am. Nat. 131:271-290. Confer,J.L. and P.I.Blades. 1975. Reaction distance to zooplankton by Lepomis gibbosus. Verh. Internat. Verein. Limnol. 19:2493-2497. Confer,J.L., G.L.Howick, M.H.Corzette, S.L.Kramer, S.Fitzgibbon, and R. Landesberg. 1978. Visual predation by planktivores. Oikos 31:27-37. Cooper,W.E. and L.B.Crowder. 1979. Patterns of predation in simple and complex environments, pp. 257-267. IN Stroud.R.H. and H.CIepper(Eds.). Predator-prey systems in fisheries management. Sport Fish. Inst., Washington. Crowl,T.A. 1989. Effects of crayfish size, orientation, and movement on the reactive distance of largemouth bass foraging in clear and turbid water. Hydrobiol. 183:133-140. 142 Cyrus,D.P. and S.J.M.Blaber. 1987a. The influence of turbidity on juvenile marine fishes in estuaries. Part 1. Field studies at Lake St. Lucia on the southeastern coast of Africa. J. Exp. Mar. Biol. Ecol. 109:53-70. Cyrus.D.P. and S.J.M.Blaber. 1987b. The influence of turbidity on juvenile marine fishes in estuaries. Part 2. Laboratory studies, comparisons with field data and conclusions. J . Exp. Mar. Biol. Ecol. 109:71-91. DiToro.D.M. 1978. Optics of turbid estuarine waters: approximations and applications. Water Res. 12:1059-1068. Dill.L.M. 1983. Adaptive flexibility in the foraging behaviour of fishes. Can. J . Fish. Aquat. Sci. 40:398-408. Dill,L.M. 1987. Animal decision making and its ecological consequences: the future of aquatic ecology and behaviour. Can. J. Zoo). 65:803-811. Dill.L.M. and A.H.G.Fraser. 1984. Risk of predation and the feeding behavior of juvenile coho salmon (Oncorhynchus kisutch). Behav. Ecol. Sociobiol. 16:65-71. Donnelly,W.A. and L.M.Dill. 1984. Evidence for crypsis in coho salmon, Oncorhynchus kisutch (Walbaum), parr: substrate colour preference and achromatic reflectance. J . Fish Biol. 25:183-195. Drenner.R.W., J.R.Strickler, and W.J.O'Brien. 1978. Capture probability: the role of zooplankter escape in the selective feeding of planktivorous fish. J . Fish. Res. Bd. Can. 35:1370-1373. Dunford.W.E. 1975. Space and food utilization by salmonids in marsh habitats of the Fraser River Estuary. M.Sc. Thesis, University of British Columbia, Vancouver. Duntley,S.Q. 1943. The mathematics of turbid media. J . Opt. Soc. Am. 33:252-257. Duntley,S.Q. 1963. Light in the sea. J . Opt. Soc. Am. 53:214-233. Eaton,R.C. and R.A.Bombardieri. 1978. Behavioural functions of the Mauthner cell. pp. 221-244. IN: Faber.D. and H.Korn(Eds.). Neurobiology of the Mauthner cell. Raven Press, New York. Eccles.D.H. 1986. Diet of the cyprinid fish Barbus aeneus (Burchell) in the P.K.le Roux Dam, South Africa, with special reference to the e f fec t of turbidity on zooplanktivory. S. Afr. J . Zool. 21:257-263. 143 Eggers.D.M. 1982. Planktivore preference by prey size. Ecology 63:381-390. Elgar.M.A. 1986. The establishment of foraging flocks in house sparrows: risk of predation and daily temperature. Anim. Behav. 34:1892-1894. Ellis.M.M. 1936. Erosion silt as a factor in aquatic environments. Ecology 17: 29-42. Gardner,M.B. 1981. E f fec ts of turbidity on feeding rates and selectivity of bluegills. Trans. Am. Fish. Soc. 110:446-450. Giancoli,D.C. 1980. Physics: principles with applications. Prentice-Hall, Englewood Cliffs, New Jersey. Giguere.L.A. and T.G.Northcote. 1987. Ingested prey increase risks of visual predation in transparent Chaoborus larvae. Oecologia 73:48-52. Gilliam,J.F. and D.F.Fraser. 1987. Habitat selection under predation hazard: test of a model with foraging minnows. Ecology 68:1856-1862. Ginetz.R.M. and P.A.Larkin. 1976. Factors affecting rainbow trout (Salmo  gairdneri) predation on migrant f ry of sockeye salmon (Oncorhynchus nerka). J . Fish. Res. Bd. Can. 33:19-24. Girsa.l.l. 1973. Alteration in the behavior and vertical distribution of certain juvenile cyprinids in relation to illumination intensity and the presence of a predator. J. Ichthyol. 13:449-454. Godin,J.-G.J. and S.A.Smith. 1988. A fitness cost of foraging in the guppy. Nature 333:69-71. Godin,T.I. and R.S.Gregory, (in prep.). Reaction times of chinook salmon f ry (Oncorhynchus tshawytscha) to prey in turbid water, see Appendix 4. Gotceitas.V. and P.Colgan. 1989. Predator foraging success and habitat complexity: quantitative test of the threshold hypothesis. Oecologia 80: 158-166. Gradall.K.S. and W.A.Swenson. 1982. Responses of brook trout and creek chubs to turbidity. Trans. Am. Fish. Soc. 111:392-395. Guthrie,D.M. 1986. Role of vision in fish behaviour, pp. 75-113. IN Pitcher,T.J. (Ed.). The behaviour of teleost fishes. Croom Helm, London. Hairston,N.G.,Jr., K.T.Li, and S .S .Easter, Jr . 1982. Fish vision and the detection of planktonic prey. Science 218:1240-1242. 144 Haney.J.C. and A.E.Stone. 1988. Seabird foraging tactics and water clarity: are plunge divers really in the clear? Mar. Ecol. Prog. Ser. 49:1-9. Harden Jones.F.R. 1956. The behaviour of minnows in relation to light intensity. J . Exp. Biol. 33:271-281. Hargreaves.N.B. 1990. Effects of predation and oceanographic factors on survival of juvenile salmon in B.C. coastal waters. Can. Conf. Fish. Res. (CCFFR) - Abstract, Ottawa. Hart.P.J.B. 1986. Foraging in teleost fishes, pp. 211-235. IN Pitcher,T.J.(Ed.). The behaviour of teleost fishes. Croom Helm, London. Hawkins,A.D. 1986. Underwater sound and fish behaviour, pp. 114-151. IN Pitcher,T.J.(Ed.). The behaviour of teleost fishes. Croom Helm, London. Hawkins,A.D. and A.D.F.Johnstone. 1978. The hearing of the Atlantic salmon, Salmo salar. J . Fish Biol. 13:655-673. Healey,M.C. 1982. Juvenile Pacific salmon in estuaries: the life support system, pp.315-341. IN Kennedy,V.S.(Ed.) Estuarine comparisons. Academic Press, Toronto. Heimstra.N.W., D.K.Damkot, and N.G.Benson. 1969. Some effects of silt turbidity on behavior of juvenile largemouth bass and green sunfish. U.S. Bur. Sport Fish. Wildl. Tech. Pap. No. 20:9p. Hemmings.C.C. 1966. Factors influencing the visibility of objects underwater. pp. 359-374. IN Bainbridge,R., G.C.Evans, and O.Rackham(Eds.). Light as an ecological factor. Blackwell Scientific Publications, Oxford. Hemmings.C.C. 1975. The visibility of objects underwater, pp. 543-545. IN Evans,G.C., R.Bainbridge, and O.Rackham(Eds.). Light as an ecological factor. II. Blackwell Scientific Publications, Oxford. Henderson.M.A. and T.G.Northcote. 1985. Visual prey detection and foraging in sympatric cuttroat trout (Salmo clarki clarki) and Dolly Varden (Salvelinus malma). Can. J . Fish. Aquat. Sci. 42:785-790. Hester,F.J. 1968. Visual contrast thresholds of the goldfish (Carassius  auratus) . Vision Res. 8:1315-1335. Hoar.W.S., M.H.A.Keenleyside, and R.G.Goodall. 1957. Reactions of juvenile Pacific salmon to light. J . Fish. Res. Bd. Can. 14:815-830. 145 Hobson.E.S. 1979. Interactions between piscivorous fishes and their prey. pp. 231-242. IN: Stroud,R.H. and H.Clepper(Eds.). Predator-prey systems in fisheries management. Sport Fish. Inst., Washington. Holanov.S.H. and J.C.Tash. 1978. Particulate and filter feeding in threadfin shad, Dorosoma petenense, at different light intensities. J . Fish. Biol. 13:619-625. Holling.C.S. 1959. Some characteristics of simple types of predation and parasitism. Can. Entomol. 91:385-398. Holling.C.S. 1966. The functional response of invertebrate predators to prey density. Mem. Ent. Soc. Can. 48:1-86. Hyatt,K.D. 1979. Feeding strategy, pp. 71-119. IN Hoar.W.S., D.J.Randall, and J. R.Brett.(Eds.). Fish physiology. Vol. 8. Bioenergetics and growth. Academic Press, New York. Johannes.M.R.S., D.J.McQueen, T.J.Stewart, and J.R.Post. 1989. Golden shiner (Notemigonus crysoleucas) population abundance: correlations with food and predators. Can. J. Fish. Aquat. Sci. 46:810-817. Johnston,D.D. and D.J.Wildish. 1982. Effect of suspended sediment on feeding by larval herring (Clupea harengus harengus L.). Bull. Environm. Contam. Toxicol. 29:261-267. Jones,M.L. 1986. The influence of natural predation on the population dynamics of Pacific salmon. Ph.D. thesis, University of British Columbia, Vancouver, British Columbia. Keast.A. 1979. Patterns of predation in generalist feeders, pp. 243-255. IN Stroud,R.H. and H.Clepper(Eds.). Predator-prey systems in fisheries management. Sport Fish. Inst., Washington. Kistritz.R.U. 1978. An ecological evaluation of Fraser Estuary tidal marshes: the role of detritus and the cycling of elements. Univ. British Columbia, Westwater Res. Centre Tech. Rep. 15. Kotler.B.P. 1984. Effects of illumination on the rate of resource harvesting in a community of desert rodents. Am. Midi. Nat. 111:383-389. Larkin,P.A. 1979. Predator-prey relations in fishes an overview of the theory, pp. 13-22. IN Stroud,R.H. and H.CIepper(Eds.). Predator-prey systems in fisheries management. Sport Fish. Inst., Washington. Levine.J.S. and E.F.MacNichol.Jr. 1979. Visual pigments in teleost fishes: e f fec ts of habitat, microhabitat, and behavior on visual system evolution. Sens. Process. 3:95-131. 146 Levy.D.A. 1977. The effects of experience on the acquisition of food by juvenile chum salmon, Oncorhynchus Keta, in a tidal creek of the Squamish River estuary, B.C. M.Sc. Thesis. University of British Columbia, Vancouver. Levy,D.A. 1987. Review of the ecological significance of diel vertical migrations by juvenile sockeye salmon (Oncorhynchus nerka). p.44-52. IN Smith, H.D., L.Margolis, and C.C.Wood(eds.). Sockeye salmon (Oncorhynchus nerka) population biology and future management. Can. Spec. Publ. Fish. Aquat. Sci. 96. Levy.D.A. 1989. The selective advantage of diel vertical migration behavior in juvenile sockeye salmon and kokanee (Oncorhynchus  nerka). Ph.D Thesis. University of British Columbia, Vancouver. Levy.D.A., T.G.Northcote, and G.J.Birch. 1979. Juvenile salmon utilization of tidal channels in the Fraser River Estuary. University of British Columbia, Westwater Research Center Technical Report 23. Levy,D.A. and T.G.Northcote. 1982. Juvenile salmon residency in a marsh area of the Fraser River Estuary. Can. J. Fish. Aquat. Sci. 39:270-276. Li.K.T., J.K.Wetterer, and N.G.Hairston,Jr. 1985. Fish size, visual resolution, and prey selectivity. Ecology 66:1729-1735. Lima,S.L. 1988. Vigilance and diet selection: the classical diet model reconsidered. J . Theor. Biol. 132:127-143. Lima.S.L., K.L.Wiebe, and L.M.Dill. 1987. Protective cover and the use of space by finches: is closer better? Oikos 50:225-230. Lima.S.L. and L.M.Dili. 1990. Behavioural decisions made under the risk of predation: a review and prospectus. Can. J. Zool. 68:619-640. Luecke,C. and W.J.O'Brien. 1981. Prey location volume of a planktivorous fish: a new measure of prey vulnerability. Can. J. Fish. Aquat. Sci. 38:1264- 1270. Lythgoe.J.N. 1966. Visual pigments and underwater vision, pp. 375-391. IN Bainbridge.R., G.C.Evans, and O.Rackham(Eds.). Light as an ecological factor. Blackwell Scientific Publications, Oxford. Mace.P.M. 1983. Predator-prey functional responses and predation by staghorn sculpins (Leptocottus armatus) on chum salmon fry (Oncorhynchus keta). Ph.D. Thesis. University of British Columbia, Vancouver. 147 Magnhagen,C. 1988. Predation risk and foraging in juvenile pink (Oncorhynchus gorbuscha) and chum salmon (O. keta). Can. J. Fish. Aquat. Sci. 45:592-596. McBride.J.R., D.R.Idler, R.E.E.Jonas, and N.Tomlinson. 1962. Olfactory perception in juvenile salmon. I. Observations on response of juvenile sockeye to extracts of foods. J. Fish. Res. Bd. Can. 19:327-334. McCrimmon.H. and W.H.Kwain. 1966. Use of overhead cover by rainbow trout exposed to a series of light intensities. J. Fish. Res. Bd. Can. 23:983-990. McFarland,W.N. 1986. Light in the sea - correlations with behaviors of fishes and invertebrates. Amer. Zool. 26:389-401. McFarland,W.N. and F.W.Munz. 1975. The visible spectrum during twilight and its implications, pp. 249-270. IN Evans,G.C., R.Bainbridge, and O.Rackham(Eds.). Light as an ecological factor. II. Blackwell Scientific Publications, Oxford. McLeay.D.J., I.K.Birtwell, G.F.Hartman, and G.L.Ennis. 1987. Responses of arctic grayling (Thymallus arcticus) to acute and prolonged exposure to Yukon placer mining sediment. Can. J. Fish. Aquat. Sci. 44:658-673. Milinski,M. 1986. Constraints placed by predators on feeding behaviour, pp. 236-252. IN Pitcher,T.J.(Ed.). The behaviour of teleost fishes. Croom Helm, London. Miller.R.J. 1979. Relationships between habitat and feeding mechanisms in fishes, pp. 269-280. IN Stroud,R.H. and H.Clepper(Eds.). Predator-prey systems in fisheries management. Sport Fish. Inst., Washington. Miller,T.J., L.B.Crowder, J.A.Rice, and E.A.Marschall. 1988. Larval size and recruitment mechanisms in fishes: toward a conceptual framework. Can. J . Fish. Aquat. Sci. 45:1657-1670. Mills,E.L., J.L.Confer, and R.C.Ready. 1984. Prey selection by young yellow perch: the influence of capture success, visual acuity, and prey choice. Trans. Am. Fish. Soc. 113:579-587. Minello,T.J., R.J.Zimmerman, and E.X.Martinez. 1987. Fish predation on juvenile brown shrimp, Penaeus aztecus Ives: effects of turbidity and substratum on predation rates. Fish. Bull. 85:59-70. Mittelbach,G.G. 1986. Predator-mediated habitat use: some consequences for species interactions. Env. Biol. Fish. 16:159-169. 148 Mittelbach,G.G. 1988. Competition among refuging sunfishes and effects of fish density on littoral zone invertebrates. Ecology 69:614-623. Mittelbach,G.G. and P.L.Chesson. 1987. Predation risk: indirect e f fec ts on fish populations. IN: Kerfoot.W.C. and A.Sih.(Eds.). Predation: direct and indirect impacts on aquatic communities. University Press of New England, Hanover. Moore, J.W. and l.A.Moore. 1976. The basis of food selection in flounders, Platichthys flesus (L.), in the Severn Estuary. J. Fish Biol. 9:139-156.. Muller.G. 1967. Methods in sedimentary petrology. Hafner, New York. Muntz,W.R.A. and A.W.Wainwright. 1978. Annual cycles in the light environments and visual mechanisms of fishes, pp.105-129. IN Thorpe,J.E.(Ed.) Rhythmic activity of fishes. Academic Press, New York. 312 pp. Munz,F.W. 1958. The photosensitive retinal pigments of fishes from relatively turbid coastal waters. J . Gen. Physiol. 42:445-459. Munz,F.W. and W.N.McFarland. 1977. Evolutionary adaptations of fishes to the photic environment, pp.193-274. IN Crescitelli,F.(Ed.) Handbook of sensory physiology, Vol. VII Part 5. The visual system in vertebrates. Springer- Verlag, Berlin. Murphy,M.L., J.Heifetz, J.F.Thedinga, S.W.Johnson, and K.V.Koski. 1989. Habitat utilization by juvenile Pacific salmon (Oncorhynchus) in the glacial Taku River, southeast Alaska. Can. J . Fish. Aquat. Sci. 46:1677-1685. Neverman,D. and W.A.Wurtsbaugh. (in prep.). Visual feeding in juvenile Bear Lake sculpin. Noggle.C.C. 1978. Behavioral, physiological and lethal e f fec ts of suspended sediment on juvenile salmonids. MSc thesis, University of Washington, Seattle. Nonacs.P. and L.M.Dill. 1990. Mortality risk vs. food quality trade-offs in a common currency: ant patch preferences. Ecology 71:1886-1892. Northcote,T.G. 1984. Mechanisms of fish migration in rivers. IN McCleave.J.D., G.P.Arnold, J.J.Dodson, and W.H.Neill. Mechanisms of migration in fishes. Plenum Publishing. Northcote,T.G., N.T.Johnston, and K.Tsumura. 1979. Feeding relationships and food web structure of lower Fraser River fishes. Univ. British Columbia, Westwater Res. Centre Tech. Rep. 16. 149 Northcote,T.G. and P.A.Larkin. 1989. The Fraser River: a major salmonine production system. Can. Spec. Publ. Fish. Aquat. Sci. 106:172-204. O'Brien,W.J., N.A.Slade, and G.L.Vinyard. 1976. Apparent size as the determinant of prey selection by bluegill sunfish (Lepomis  macrochirus). Ecology 57:1304-1310. Real.L. and T.Caraco. 1986. Risk and foraging in stochastic environments. Ann. Rev. Ecol. Syst. 17:371-390. Savino.J.F. and R.A.Stein. 1982. Predator-prey interaction between largemouth bass and bluegills as influenced by simulated, submersed vegetation. Trans. Am. Fish. Soc. 111:255-266. Schneider,K.J. 1984. Dominance, predation, and optimal foraging in white-throated sparrow flocks. Ecology 65:1820-1827. Scott,W.B. and E.J.Crossman. 1973. Freshwater fishes of Canada. Fish. Res. Bd. Can. Bull. 184. Shepherd,B.G., J.E.Hillaby, and R.J.Hutton. 1986. Studies on Pacific salmon (Oncorhynchus spp.) in Phase I of the salmonid enhancement program. Can. Tech. Rep. Fish. Aquat. Sci. 1482. Sigler,J.W., T.C.Bjornn, and F.H.Everest. 1984. E f fec ts of chronic turbidity on density and growth of steelheads and coho salmon. Trans. Am. Fish. Soc. 113:142-150. Sih.A. 1986. Antipredator responses and the perception of danger by mosquito larvae. Ecology 67:434-441. Simenstad,C.A.(Ed.). 1990. E f fects of dredging on anadromous Pacific coast fishes: workshop proceedings September 8-9, 1988. University of Washington, Seattle. Simenstad,C.A., K.L.Fresh, and E.O.Salo. 1982. The role of Puget Sound and Washington coastal estuaries in the life history of Pacific salmon: an unappreciated function, pp.343-364. IN Kennedy,V.S.(Ed.) Estuarine comparisons. Academic Press, Toronto. Stone,H.H. and G.R.Daborn. 1987. Diet of alewives, Alosa pseudoharengus and blueback herring, A. aestivalis (Pisces: Clupeidae) in Minas Basin, Nova Scotia, a turbid, macrotidal estuary. Environ. Biol. Fishes 19:55-67. Swenson,W.A. 1978. Influence of turbidity on fish abundance in western Lake Superior. Envir. Res. Lab., Duluth. Environ. Protect. Agency 600/3-78-067. 150 Swenson,W.A. and M.L.Matson. 1976. Influence of turbidity on survival, growth, and distribution of larval lake herring (Coregonus artedii). Trans. Am. Fish. Soc. 105:541-545. Tamura,T. 1957. A study of visual perception in fish, especially on resolving power and accomodation. Bull. Jap. Soc. Sci. Fish. 22:536-557. Taylor,E.B. 1989. Adaptive diversification of juvenile life histories in the chinook salmon, Oncorhynchus tshawytscha (Walbaum). Ph.D. Thesis, University of British Columbia, Vancouver. Taylor,E.B. 1990. Environmental correlates of life-history variation in juvenile chinook salmon, Oncorhynchus tshawytscha (Walbaum). J. Fish Biol. 37:1-17. Underwood,R. 1982. Vigilance behaviour in grazing African antelopes. Behaviour 79:81-107. Vinyard,G.L. 1980. Differential prey vulnerability and predator selectivity: e f fec ts of evasive prey on bluegill (Lepomis  macrochirus) and pumpkinseed (L. gibbosus) predation. Can. J. Fish. Aquat. Sci. 37:2294-2299. Vinyard,G.L. and W.J.O'Brien. 1976. E f fec ts of light and turbidity on the reactive distance of bluegill (Lepomis macrochirus). J . Fish. Res. Bd. Can. 33:2845-2849. Wallen.I.E. 1951. The direct e f fect of turbidity on fishes. Oklahoma Agr. Mech. Coll. Bull. 48:1-27. Wanzenbock.J. and F.Schiemer. 1989. Prey detection in cyprinids during early development. Can. J. Fish. Aquat. Sci. 46:995-1001. Ware.D.M. 1971. Predation by rainbow trout (Salmo gairdneri): the effect of experience. J . Fish. Res. Bd. Can. 28:1847-1852. Ware.D.M. 1972. Predation by rainbow trout (Salmo gairdneri): the influence of hunger, prey density, and prey size. J. Fish. Res. Bd. Can. 29:1193-1201. Ware,D.M. 1973. Risk of epibenthic prey to predation by rainbow trout (Salmo gairdneri). J . Fish. Res. Bd. Can. 30:787-797. Werner,E.E. 1977. Species packing and niche complementarity in three sunfishes. Am. Nat. 111:553-578. Werner,E.E. and D.J.Hall. 1984. Ontogenetic habitat shifts in bluegill: the foraging rate-predation risk trade-off. Ecology 69:1352-1366. 151 Werner.E.E. and J.F.Gilliam. 1984. The ontogenetic niche and species interactions in size-structured populations. Ann. Rev. Ecol. Syst. 15:393-425. Werner.E.E., J.F.Gilliam, D.J.Hall, and. G.G.Mittelbach. 1983. An experimental test of the effects of predation risk on habitat use in fish. Ecology 64: 1540-1548. White,H.C. 1936. The food of kingfishers and mergansers on the Margaree River, Nova Scotia. J . Biol. Bd. Can. 2:299-309. Wilkinson,L. 1988. SYSTAT: the system for statistics. Evanston, Illinois. Woodhead.P.M.J. 1956. The behaviour of minnows (Phoxinus phoxinus L.) in a light gradient. J . Exp. Biol. 33:257-270. Zar.J.H. 1984. Biostatistical analysis. Prentice-Hall, Englewood Cl i f fs , New Jersey. APPENDIX 1. Feeding Trials for Drosophila, by Experiment 152 Figure A1. E f f ec t of turbidity on the foraging rate (prey in 10.0 minutes) of juvenile chinook salmon for Drosophila prey, 1987 (vertical bars - standard error; n = 8 fish per mean). APPENDIX 1. Feeding Trials for Drosophila, by Experiment 153 Figure A2. Ef fect of turbidity on the foraging rate (prey in 10.0 minutes) of juvenile chinook salmon for Drosophila prey, 1988 (vertical bars - standard error; n = 8 fish per mean). 200 Turbidity (mg/L) 400 APPENDIX 1. Feeding Trials for Drosophila, by Experiment 154 Figure A3. Ef fect of A. Light and B. Turbidity on the foraging rate (prey in 10.0 minutes) of juvenile chinook salmon for Drosophila prey (vertical bars - standard error; n = 8 fish per mean). A B 0 28 BO TOO 200 400 800 Turbidity (mg/L) _i i 1 1- 1 <— wi o.n o* i 0.08 -1.1 ~ -aa Log(microEir»tein8/irr/8) APPENDIX 1. Feeding Trials for Drosophila, by Experiment 155 Figure A4. Ef fect of turbidity on the foraging rate of juvenile chinook salmon for A. Surface Drosophila (prey in 10.0 minutes) and B. Planktonic Drosophila (prey in 1.0 minutes) (vertical bars - standard error; n = 8 fish per mean). Turbidity (mg/L) w f 64.9*3.2 mm FL Turbidity (mg/L) APPENDIX 1. Feeding Trials for Drosophila, by Experiment 156 Figure A5. Effect of turbidity on the foraging rate (prey in 10.0 minutes) of juvenile chinook salmon for Drosophila prey in A. "Normal" and B. "Enhanced" risk (vertical bars -standard error; n = 8 fish per mean). APPENDIX 2. Feeding Trials for Artemia, by Experiment 157 Figure A6. Ef fect of turbidity on the foraging rate (prey in 1.0 minutes) of juvenile chinook salmon for Artemia prey, 1987 (vertical bars - standard error; n = 8 fish per mean). 67.6*4.3 mm FL 29 SO 100 200 400 800 Turbidity (mg/L) APPENDIX 2. Feeding Trials for Artemia, by Experiment 158 Figure A7. Ef fect of turbidity on the foraging rate (prey in 1.0 minutes) of juvenile chinook salmon for Artemia prey, 1988 (vertical bars - standard error; n = 8 fish per mean). n 1 ' • • • 1 1 63.8*3.1 mm FL Turbidity (mg/L) APPENDIX 2. Feeding Trials for Artemia, by Experiment 159 Figure A8. Ef fect of A. Light and B. Turbidity on the foraging rate (prey in 1.0 minutes) of juvenile chinook salmon for Artemia prey (vertical bars - standard error; n = 8 fish per mean). APPENDIX 2. Feeding Trials for Artemia, by Experiment 160 Figure A9. Ef fect of turbidity on the foraging rate of juvenile chinook salmon for A. Surface Artemia (prey in 10.0 minutes) and B. Planktonic Artemia (prey in 1.0 minutes) (vertical bars - standard error; n = 8 fish per mean). APPENDIX 2. Feeding Trials for Artemia, by Experiment 1 6 1 Figure A10. Ef fect of turbidity on the foraging rate (prey in 1.0 minutes) of juvenile chinook salmon for Artemia prey in A. "Normal" and B. "Enhanced" risk (vertical bars - standard error; n = 8 fish per mean). pi I I 1 I I I I ol I I I I I I I 0 28 BO TOO 200 400 800 0 28 60 100 200 400 800 Turbidity (mg/L) Turbidity (mg/L) APPENDIX 3. Feeding Trials for Tubifex, by Experiment 162 Figure All. E f fect of turbidity on the foraging rate (mg prey in 5.0 minutes) of juvenile chinook salmon for Tubifex prey, 1987 (vertical bars - standard error; n = 8 fish per mean). 0 28 SO 100 200 400 800 Turbidity (mg/L) 66.0*3.0 mm FL 800 Turbidity (mg/L) APPENDIX 3. Feeding Trials for Tubifex, by Experiment 163 Figure A12. Ef fect of turbidity on the foraging rate (mg prey in 5.0 minutes) of juvenile chinook salmon for Tubifex prey, 1988 (vertical bars - standard error; n = 8 fish per mean). 1 1 1 1 64.3*2.4 mm FL Turbidity (mg/L) APPENDIX 3. Feeding Trials for Tubifex, by Experiment 164 Figure A13. Ef fect of A. Light and B. Turbidity on the foraging rate (mg prey in 5.0 minutes) of juvenile chinook salmon for Tubifex prey (vertical bars - standard error; n = 8 fish per mean). APPENDIX 3. Feeding Trials for Tubifex, by Experiment 165 Figure A14. Ef fect of turbidity on the foraging rate (mg prey in 5.0 minutes) of juvenile chinook salmon for Tubifex prey in A. "Normal" and B. "Enhanced" risk (vertical bars - standard error; n = 8 fish per mean). APPENDIX 4. Godin.T.I. and R.S.Gregory (in prep.) 166 Godin.T.I. and R.S.Gregory, (in prep.). Reaction times of chinook salmon f ry (Oncorhynchus tshawytscha) to prey in turbid water. Abstract Suspended sediment lowers the visibility of prey to foraging fish by scattering light signals in the foreground. Turbidity can also act to increase the visibility of prey. Particulates in the water column may also scatter light in a manner which creates uniform background illumination, increasing prey contrast. A multichambered experimental arena allowing independent manipulation of background and foreground turbidities (0-400 mg*L_1) was used to determine the reaction times of chinook salmon f ry (Oncorhynchus tshawytscha) to variously contrasted Artemia prey over a fixed distance. Six conditioned subjects were used in the experiments. Reaction times were 20-40/ faster with darkened prey as compared with light prey and 25-37/. faster for prey against turbid as compared to non-turbid backgrounds. With uniform foreground and background suspended sediment levels, no significant differences in reaction times were found among conditions ranging from clear to moderately turbid (<100 mg'L-1). Given the reduced visual range in turbid waters, the latter observation suggests the possibility of faster reaction times within the reduced visual field. APPENDIX 4. Godin,T.I. and R.S.Gregory (in prep.) 167 Figure A15. Ef fect of A. Prey colour, B. Background turbidity, and C. Both foreground and background turbidity, on the reaction time of juvenile chinook salmon for Artemia prey (means of individual fish medians; vertical bars are standard deviations; data from Godin and Gregory in prep.). 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
https://iiif.library.ubc.ca/presentation/dsp.831.1-0100583/manifest

Comment

Related Items