Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

The response of rainbow trout Salmo gairdneri to lures with special references to color preference Dooley, Robert H. A. 1989

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

Item Metadata

Download

Media
831-UBC_1974_A6_7 D66.pdf [ 4.91MB ]
Metadata
JSON: 831-1.0099872.json
JSON-LD: 831-1.0099872-ld.json
RDF/XML (Pretty): 831-1.0099872-rdf.xml
RDF/JSON: 831-1.0099872-rdf.json
Turtle: 831-1.0099872-turtle.txt
N-Triples: 831-1.0099872-rdf-ntriples.txt
Original Record: 831-1.0099872-source.json
Full Text
831-1.0099872-fulltext.txt
Citation
831-1.0099872.ris

Full Text

T H E R E S P O N S E O F R A I N B O W T R O U T (Salmo ga i rdner i ) T O L U R E S W I T H S P E C I A L R E F E R E N C E T O C O L O R P R E F E R E N C E by Rober t H . A . Dooley B . E d . , U n i v e r s i t y of B r i t i s h Co lumbia , 1971 A THESIS S U B M I T T E D IN P A R T I A L F U L F I L M E N T O F T H E R E Q U I R E M E N T S F O R T H E D E G R E E O F M A S T E R O F S C I E N C E i n the Depar tment of Zoology We accept this thesis as conforming to the r equ i red standard T H E U N I V E R S I T Y O F B R I T I S H C O L U M B I A September, 1974 In p resent ing t h i s t h e s i s in p a r t i a l f u l f i l m e n t o f the requirements fo r an advanced degree at the U n i v e r s i t y of B r i t i s h Columbia, I agree t h a t the L i b r a r y s h a l l make i t f r e e l y a v a i l a b l e f o r reference and study. I f u r t h e r agree t h a t permiss ion for e x t e n s i v e copying of t h i s t h e s i s fo r s c h o l a r l y purposes may be granted by the Head of my Department or by h i s r e p r e s e n t a t i v e s . It i s understood that copy ing o r p u b l i c a t i o n of t h i s t h e s i s f o r f i n a n c i a l gain s h a l l not be a l lowed without my w r i t t e n p e r m i s s i o n . Depa rtment The U n i v e r s i t y of B r i t i s h Columbia Vancouver 8, Canada i i A B S T R A C T The response of ra inbow trout (Salmo gai rdner i ) to l u r e s was invest igated i n t r o l l i n g exper iments at L o o n Lake , B r i t i s h C o l u m b i a . The "ac t ion" of a lure was found to be an important parameter i n de te rmin ing i ts eff ic iency: of four actions tested, the f la t f i sh caught the greatest number of f i sh . Al though the co lo r of lu re was not significant, red lu res were more efficient than yel low, green, and blue, and more efficient than var ious co lo r patterns of r e d and white. The presence of a dodger wi th lu res did not affect the i r e f f ic iency, but l a rge r f i sh were caught. No size se lec t ion o c c u r r e d wi th ei ther c o l o r s or act ions of l u r e s . In l abora tory feeding exper iments using dyed trout eggs as food, r e d was selected f i r s t or second more often than yel low, green, or blue. The co lor of background against w h i c h the f i sh were fed, and ind iv idua l differences among f i sh caused signif icant changes i n the preference shown for va r ious c o l o r s of food. Combining two co lo r s a lso affected the se lec t ion intensity, depending upon the contrast between the two c o l o r s . P re fe rences for different c o l o r s of food were not influenced by the hunger l e v e l of the f ish, measured i n t e r m s of the quantity of food i n the f i sh ' s gut. In the course of the exper iments it was inc iden ta l ly observed: (1) that ra inbow trout possess a s t r i k ing ab i l i ty to match the hue of the i r sk in (mainly i n the d o r s a l region) to that of the background in which they are kept; (Z) the co lo r of background affects the ac t iv i ty l eve l of ra inbow trout: ye l low produces the highest l e v e l of ac t iv i ty and green the lowest . i i i T A B L E OF CONTENTS Page T I T L E P A G E i A B S T R A C T « T A B L E O F CONTENTS i i i LIST OF FIGURES v LIST O F T A B L E S v i ACKNOWLEDGMENTS v i i i INTRODUCTION 1 P A R T I . T R O L LING EXPERIMENTS 5 A. Classification of Lures 5 B. General Methods 7 C. Trolling Experiment I 1° 1. Methods 10 2. Results 10 D. Trolling Experiment II 12 1. Methods 1 2 2. Results 1 3 E. Trolling Experiment III 15 1. Methods 15 2. Results I 7 F. Trolling Experiments -- Discussion 17 P A R T II. F E E D I N G EXPERIMENTS 24 A. Introduction 24 B. General Methods 24 C. Feeding Experiment I 28 1. Methods 28 2. Results 29 3. Discussion 36 D. Feeding Experiment II 43 1. Methods 43 2. Results 46 3. Discussion 46 INCIDENTAL OBSERVATIONS 54 A. Effect of Background Color on Activity Level of Trout 54 1. Methods 54 2. Results 55 3. Discussion 55 i v Page B . Adaptat ion of Sk in of Trou t to Background C o l o r 57 DISCUSSION 61 A . Background L i t e r a t u r e 61 1 . C o l o r v i s i o n i n f i sh 61 2 . T r a n s m i s s iv i ty of l ight in water 66 3 . Relevance to t r o l l i n g and feeding exper iments 68 B . Interpretat ion of Resu l t s 68 R E F E R E N C E S C I T E D 74 V L I S T O F F I G U R E S F I G U R E Page 1 Checkered and s t r iped f latf ish, and dodger f r o m t r o l l i n g exper iment III, L o o n Lake , B . C . , August 25 to September 1, 1972 1 6 2 Cutaway v iew of exper imenta l tank used i n feeding exper i -ments I and II at the U n i v e r s i t y of B . C . , M a y 22 to June 6, 1973; and Ju ly 23 to August 3, 1973 26 3 Graph of mean select ion intensi t ies of co lo r s of food in var ious co lo r s of tank. Feeding exper iment I, U n i v e r s i t y of B . C . , M a y 22 to June 6, 1973 33 4 Graph of mean select ion intensi t ies of co lo r s of food among different f i s h . Feed ing exper iment I, U n i v e r s i t y of B . C . , M a y 22 to June 6, 1973 35 5 P r e p a r a t i o n of food for feeding exper iment I I - -a r rangement of eggs i n the gelatine; plan v iew. Feeding exper iment II, U n i v e r s i t y of B . C . , Ju ly 22 to August 3, 1973 44 6 Regions of sk in for sk in co lo r observat ions . Incidental observat ions 59 v i L IST OF T A B L E S T A B L E Page I C l a s s i f i c a t i o n of lure action based on a test t r o l l of 3 5 different l u r e s i n a f i s h pond at the Uni v e r s i t y of B r i t i s h Columbia, spring of 1972 6 II Numbers of f i s h caught on l u r e s f r o m t r o l l i n g experiment I, Loon Lake, B.C., July 10-17, 1972 11 III Numbers of f i s h caught on l u r e s f r o m t r o l l i n g experiment II, Loon Lake, B.C., August 5 to 9, 1972 14 IV Numbers of f i s h caught on l u r e s f r o m t r o l l i n g experiment III, Loon Lake, B.C., August 25 to September 1, 1972 18 V Mean lengths (cm) of f i s h caught on lures f r o m t r o l l i n g experi-ment III, Loon Lake, B.C., on the basis of color pattern of lur e and on the basis of whether or not a dodger was used. August 25 to September 1, 1972 19 VI Selection intensities of colo r s of food i n various colors of tank. Feeding experiment I, U n i v e r s i t y of B.C., May 22 to June 6, 1973 31 VII Mean selection intensities and interaction effects for col o r s of food i n various colors of tank. Feeding experiment I, Un i v e r s i t y of B.C., May 22 to June 6, 1973 32 VIII Mean selection intensities and interaction effects for c o l o r s "v of food among different f i s h . Feeding experiment I, Univer-sity of B.C., May 22 to June 6, 1973 34 IX Mean selection intensities and interaction effects for c o l o r s of food among different f i s h , for each replicate. Feeding experiment I, U n i v e r s i t y of B.C.,. May 22 to June 6, 1973 38 X V a r i a b i l i t y i n selection order between replicates and between tanks. Feeding experiment I, U n i v e r s i t y of B.C., May 22 to June 6, 1973 40 X I Interaction effects for second-order interaction (food color x tank color x individual differences) -- average for two rep l i c a t e s . Feeding experiment I, U n i v e r s i t y of B.C., May 22 to June 6, 1973 42 v i i T A B L E Page XII Se lec t ion in tensi t ies of b i - c o l o r e d food in va r ious co lo r s of tank. Feeding exper iment II, U n i v e r s i t y of B . C . , Ju ly 23 to August 3, 1973 47 XIII M e a n select ion intensi t ies and in te rac t ion effects for co lo r s of food wi th va r ious co lo r s of tag. Feeding exper iment II, U n i v e r s i t y of B . C . , Ju ly 23 to August 3, 1973 48 X I V M e a n se lec t ion in tensi t ies and in terac t ion effects for co lo r s of food i n var ious co lo r s of tank. Feeding exper iment II, U n i v e r s i t y of B . C . , Ju ly 23 to August 3, 1973 50 X V Effect of wide spec t ra l separat ion of food co lo r and tank co lo r on food co lo r x tank co lo r in te rac t ion effects of feeding exper iment II, U n i v e r s i t y of B . C . , Ju ly 23 to August 3, 1973 51 X V I Selec t ion intensi t ies f r o m feeding exper iment II grouped accord ing to s i m i l a r i t y of co lo r s of tank, tag egg, and food egg. Feeding exper iment II, U n i v e r s i t y of B . C . , Ju ly 23 to August 3, 1973 53 X V I I M e a s u r e s of ac t iv i ty l eve l s of f i sh i n different co lo r ed habitats 56 X V I I I D e s c r i p t i o n of sk in co lo r of f i s h f r o m different co lo r ed tanks. Incidental observat ions 60 v i i i A CKNOWLEDGMENTS I would like to thank Dr. P. A. Larkin who proposed the problem for investigation, gave helpful advice and encouragement throughout the field work and analysis phases of this study, and provided financial assistance when it was sorely needed. Thanks are also due to Dr. N. J. Wilimovsky and Dr. T. G. Northcote, who reviewed the manuscript. I am indebted to Mr. H. Sparrow and Mr. A. Mitchell of the B r i t i s h Columbia F i s h and Wildlife Branch for their cooperation in providing the fish used in the experiments at the University of British Columbia. Also in the F i s h and Wildlife Branch, Mr. G. Halsey, Mr. D. Smith, and other members of the Loon Lake Research Group helped by providing accommodation at Loon Lake and allowing me to use their research boat and other equipment. The able assistance of Mr. J. Blackburn, Mr.S.Simms, and Mr. A. Belshaw, made it possible to carry out the trolling experiments at Loon Lake. Finally, I sincerely thank my wife, Mary, who accepted the role of a student wife for many years and who never lost faith in me throughout this time. 1 I N T R O D U C T I O N The se lec t iv i ty and eff ic iency of f ishing gear is a v i t a l conce rn of f i she rmen and f i she r ies managers thoughtout the w o r l d . The f i she rman , of course , wants to use the most eff icient l ega l gear i n o rder to r a i s e h is ca tch per unit effort. On the other hand, by regulat ing the l e v e l of gear eff ic iency, a f i she r i e s manager can m a x i m i z e harvest , thus, i n the long run, a l lowing greater numbers to share in the r e s o u r c e . Se lec t iv i ty and eff ic iency embody two separate concepts, but they can be reduced to t e rms of one of them. The ef f ic iency of a unit of f i sh ing gear r e fe r s to i t s ab i l i ty re la t ive to other gear to ca tch f i sh . It i s r e a l l y the same concept as R i c k e r ' s "fishing power", which he defines as "...the ca tch taken by the giyen apparatus, d iv ided by the ca tch of a standard apparatus f ishing at n e a r l y the same t ime and p l a c e . " (Ricker , 1958) Se lec t iv i ty r e fe r s to the c h a r a c t e r i s t i c of a unit of f i shing gear whereby it catches f i s h of a c e r t a i n weight or length i n greater amounts than would be expected on the bas i s of the abundance of f i sh of that weight or length i n the total stock. F o r instance, a seine net of a ce r t a in mesh s ize may a l l ow 50 per cent of the f i sh of l e ss than 70 c m length to pass through i t . It i s sa id to select for f i s h of greater than 70 c m length. A l a rge r mesh s ize may a l low 50 per cent of the f i sh of less than 80 c m to s l i p through. It i s sa id that the second net selects for l a r g e r f i sh than the f i r s t net. But this can a lso be explained as a change i n ef f ic iency - -the efficiency of the two nets to catch f ish of 70-80 c m (and perhaps f i s h of other s izes also) i s not the same. It is not necessa ry to l i m i t the concept of se lec t iv i ty to length and weight of f i sh . Gear can be sa id to select on the bas i s of other c h a r a c t e r i s t i c s such as species or sex. A change in eff ic iency does not n e c e s s a r i l y involve a change i n se l ec -t iv i ty , although it often does. It i s at leas t theore t i ca l ly poss ib le to have some change in f ishing gear, producing an increase or decrease in eff ic iency that is p ropor t iona l ly d is t r ibu ted over the range of the va r i ab le concerned . 2 The l i t e r a tu re on gear ef f ic iency is s izable , but it tends to cons i s t for the greater part of papers dealing wi th eff ic iency of nets . The scope of fac tors found to influence the se l ec t iv i ty and eff iciency of nets is. great; some of the more prominent factors a re : s ize of mesh (Davis , 1929). s ize d i s t r i b u -t ion of f i sh stock (Borowik, 1930), m a t e r i a l of which, the net i s made (Boerema , 1956; Pycha , 1962), s ize of ca tch ( M c C r a c k e n , 1963), and co lo r of net (Jester , 1973). - I n c o m p a r i s o n wi th the l i t e r a tu re on nets, the l i t e ra tu re on lu re s i s ra ther sparse . Whi le catch per unit effort ( C P U E ) values are frequently de termined f r o m c r e e l censuses, there i s often no attempt made to ca tegor ize these values under different types of lu re (Emig , 1971). Esp inosa ,e t a l . , (1971) contras ted success of l i v e bait f i she rmen wi th that of f i s h e r m e n using a r t i f i c i a l l u r e s , but did not attempt any subd iv i s ion of the a r t i f i c i a l l u r e s . In some studies where the type of lu re i s ca tegor ized p r e c i s e l y , l i t t l e attempt has been made to con t ro l other factors that might influence the f i sh ing success such as angler s k i l l and technique, loca t ion of f i sh ing area, and t ime of year of f i shing ( L a r k i n , 1949). In other studies that have paid more attention to the "env i ronmenta l " factors , the ca tegor ies of lu re have been few and broad .i (Beukema, 1970; Boydstun, 1972). Ce r t a in ly , there have been few e x p e r i -mental studies a imed spec i f i c a l l y at de te rmin ing the c h a r a c t e r i s t i c s of a good l u r e . Of the studies done on eff ic iency and se l ec t iv i ty of l u r e s , most have been designed along the l ines of pi lot studies for management purposes . Frequent ly they have been prompted by observat ions that a substant ia l number of unders ized f i sh were being s a c r i f i c e d as a r e su l t of t r ad i t i ona l f i shing methods ('Pitre, 1970; Boydstun, 1972), or by the des i r e to assess the effectiveness of cur ren t regula t ions (Shetter, et a l . , 1965). R e g a r d l e s s of whether or not the i r purpose was to be a tool for management, most studies have accepted their f ie ld of r e s e a r c h as cover ing the l u r e s that are c u r r e n t l y used by most f i she rmen . In contras t to the studies on net f i shing (Jester , 1973; Pycha , 1962; and Zupanovic , 1963) , -se ldom has an inves t igat ion attempted 3 to determine the effect of modif icat ions of gear on i ts se lec t iv i ty or ef f ic iency. What compar i sons have been made were usual ly made among broad categories of lure (e.g., f l ies ve rsus spoons ve r sus plugs ve r sus gang t r o l l s ) . A c c o r d i n g l y , there i s often no effort made to s tandardize techniques of f ishing or even t imes of year at which f ishing occurs ; the assumpt ion being made, perhaps i m p l i c i t l y , that this w i l l balance out i f enough samples are taken. This is a dangerous assumption to make, as i t is genera l ly accepted that f i she rmen using ce r t a in types of gear exc lus ive ly are more exper ienced and probably more sk i l fu l than f i she rmen using other types of gear (Espinosa , 1971; Shetter, 1965). A l s o , lu res that are used in a winter f i shery , or that select for l a rge r older f i sh , cannot be expected to y ie ld so high a ca tch per unit effort ( C P U E ) as those used at the height of the f ishing season, or those that select for sma l l e r f i sh . The foregoing r e m a r k s are not meant to be d i sparag ing s ince the bulk of these studies have been d i rec ted to the question, "What effect have the var ious c l a s ses of lu re had on a f i she ry? ", and it i s perhaps not necessa ry to separate out the effects of f i she rman s k i l l and other factors f r o m the inherent eff ic iency of th.^ lu re when deal ing wi th this question. But I was more concerned wi th identifying the c h a r a c t e r i s t i c s of an efficient lure , and it was therefore impera t ive that I cont ro l as many confusing factors as poss ib le . It i s for this r eason that a care fu l attempt was made to match the lu res for s ize and to paint them in toto wi th standard co lo r s of paint. Since a multitude of factors in terac t to determine the success of lu res , a f ishing technique was sought that would be most amenable to s tandardizat ion. Both f ly - f i sh ing and spin-cas t ing depend substant ial ly on the s k i l l and technique of the ind iv idua l angler; use of l i ve bait makes s tandardizat ion at the other end of the l ine di f f icul t . Therefore , the type of f ishing was r e s t r i c t e d to t r o l l i n g wi th a r t i f i c i a l lures , for which i t i s r e l a t i ve ly easy to s tandardize technique. Loca t i on and behavior of the f i sh l o o m prominent in the de terminat ion of catchabi l i ty to var ious types of lu re , yet are a lmost imposs ib l e to standardize exper imenta l ly . Never the less , some degree of con t ro l can be achieved by f ishing the same area i n the same pattern for the extent of each t r i a l , and a s t a t i s t i ca l compar i son is possible by assur ing that each lu re combinat ion i s f i shed a number of t imes throughout an exper iment . In the second half of this thesis feeding exper iments are de sc r ibed which were per formed in an attempt to c o n f i r m apparent co lo r preferences f r o m the lu re exper iments . It was assumed that trout s t r ike l u r e s because they mis take them for potential food; i f this is so, then observing their p r e -ference for food of different co lo r s under con t ro l l ed condit ions might help to c l a r i f y their responses to the co lo red lu res used i n the t r o l l i n g exper iments . F o r purposes of compar i son , there i s ve ry l i t t l e l i t e ra tu re on the preference of f i sh for different co lo r s of food. Some of the e a r l i e r work on co lo r v i s i o n in f i sh sought to prove that ce r t a in species of f i sh d i s t ingu i sh hue by demonstrat ing that an assoc ia t ion of food wi th a c o l o r e d object could be developed (White, 1919; Brown, 1937). But this assoc ia t ion was developed a r t i f i c i a l l y through a condit ioning p r o g r a m and d id not re f lec t "na tu ra l " preferences of the f i sh . A s far as ra inbow trout are concerned, the only wor of which I a m aware that purports to measure these "na tu ra l " preferences (albeit in a labora tory setting) i s that of Ginetz and L a r k i n (1973), and perhap: that of Wolf and Wales (1953). 5 P A R T I. T R O L L I N G E X P E R I M E N T S Reac t ion of ra inbow trout (Salmo gai rdner i ) to l u r e s was invest igated i n t r o l l i ng exper iments at L o o n Lake , B r i t i s h Co lumbia , dur ing the summer of 1972. A n attempt to c l a r i f y co lor preferences was made wi th feeding exper iments at the U n i v e r s i t y of B r i t i s h Co lumbia i n the spr ing and summer of 1973. A . C l a s s i f i c a t i o n of L u r e s T h i r t y - f i v e different l u r e s were " t r o l l e d " in a f i s h pond at the U n i v e r s i t y of B r i t i s h Co lumbia by dragging them through the water on a l ine attached to a f i sh ing rod . The action' ' of each lu re was observed and c l a s s i f i e d . There appeared to be, at the most, only a half dozen dis t inct types of act ions, wi th ind iv idua l l u r e s exhibi t ing sl ight va r i a t i ons . These are desc r ibed i n Table I . Of these actions, four types o c c u r r e d most frequently: f la tf ish, spinner, spoon and "dead". The "dead" l u r e showed no wobbling or spinning act ion. 1 A l l the lu res that were tested r e l i e d in some way on the f lu id drag of water s t r ik ing the lu re to give i t whatever "ac t ion" it had (i.e., there were no bat tery-opera ted motors i n the lu re i t s e l f to provide the action). Therefore , at some m i n i m u m speed of t r o l l i n g where the drag was low, each lu re exhibi ted no action, i .e . , it fol lowed i ts point of attachment to the l ine wi th no sideways or v e r t i c a l motion of any part of the lu re and wi th no ro ta t ion of the lu re i tself , or of any part of the l u r e . A s the speed of t r o l l i n g inc reased (and therefore the pressure of the water on the lure) , some of the l u r e s s tar ted to exhibi t a sideways or v e r t i c a l motion of the whole lu re or part of the lure , i n conjunction wi th i ts fo rward mot ion. Other lu res exhibi ted ro ta tory mot ion. These "ac t ions" began at different speeds for different lu res .and continued qua l i ta t ive ly unchanged throughout a ce r t a in upward range of t r o l l i n g speeds. A t a s t i l l greater speed this more or less o rde r ly motion of the lure was blocked and/or ec l ipsed by an e r r a t i c movement of the lure through the water, consis t ing of a se r ies of j e rks forward three or four t imes the length of the l u r e . A t this speed the lure could be said to "cavi ta te" as the flow of water over its surface changed f r o m a l amina r to turbulent f low. F requen t ly the lure would r i s e and break the surface of the water . T A B L E I. C l a s s i f i c a t i o n of lure action based on a test t r o l l of 35 different l u r e s i n a f i s h pond at the U n i v e r s i t y of B r i t i s h Columbia, spring of 1972. L u r e s used in t r o l l i n g experiment I are marked with an a s t e r i s k (*), and their dimensions are given in m i l l i m e t e r s . A c t i o n T y p i c a l Shape De scr iption F l a t f i s h F r o n t end planes downward in the water and moves rap i d l y f r o m side to side in a d i r e c t i o n p e r p e n d i c u l a r to the line. T h i s gives the r e a r end of the l u r e a s i m i l a r motion due to the point of attachment of l i n e . R e q u i r e s f a i r l y slow (=^2 ft/s) speed of tow. Spinner The body of the lure (B) follows in the path of the line with a m i n i m u m of motion. A l l the action comes f r o m the spinner (S) which i s attached to the wire going through the center of the body of the l u r e . It spins around the body and d e s c r i b e s a cone whose apex is at the point of attachment of the spinner. R e q u i r e s a f a i r l y fast tow (>1.7 ft/s) to make spinner work. Spoon 13 34 The front end of the l u r e follows the line with l i t t l e action while the r e a r of the l u r e moves sideways back and forth, or up and down quite r a p i d l y R e q u i r e s a medium speed. Dead T 33 Shape is v a r i a b l e but, r e g a r d l e s s of shape there is a v e r y m i n i m u m of action, the lure following almost d i r e c t l y in the path of the line. The dead action lure that I used was obtained by clipping the spinner off of a spinner l u r e . P r o p e l l e r action Two or three l u r e s used the f o r c e of the water striking (usually) three f o i l s to cause them to rotate around a c e n t r a l axis coincident with the l i n e . These l u r e s t r o l l e d at a wide range of speeds, and, to the human eye, had a great likeness to the spinner action. 7 B. G e n e r a l Methods T h r e e t r o l l i n g e x p e r i m e n t s were p e r f o r m e d at L o o n Lake, B r i t i s h C o l u m b i a between J u l y . 1 0 and September 1 , 1 9 7 2 . L o o n L a k e i s situated i n a n a r r o w v a l l e y at an elevation of 2 8 2 0 feet a p p r o x i m a t e l y 1 2 m i l e s east of Clinton, B r i t i s h Columbia. It i s about 9 . 5 m i l e s long, and v a r i e s between one-quarter and one-half mile, in width. The mean depth i s 9 0 feet. T r o l l i n g was done f r o m two v e s s e l s : an 1 8 foot motor launch owned by the B r i t i s h C o l u m b i a F i s h and W i l d l i f e Branch, and a s m a l l ( 1 0 foot) a l u m i n u m outboard motor boat. When both boats were used, the speed of the s m a l l e r boat was s y n c h r o n i z e d to that of the l a r g e r by running alongside and noting t h r o t t l e settings. A s the s m a l l boat n o r m a l l y followed the motor launch around a set course, approximate synchrony could be maintained quite e a s i l y by keeping the distance between the boats roughly constant. The speed chosen for t r o l l i n g was d e t e r m i n e d by taking into c o n s i d e r a -tion the actions of each type of l u r e as d e s c r i b e d i n Table I. P i c k i n g a speed that would give the proper t r o l l i n g c h a r a c t e r i s t i c s f o r each l u r e was m o d e r a t e l y c r i t i c a l because there was not a wide o v e r l a p i n the acceptable speed ranges among the l u r e s . Speed was m e a s u r e d by dropping a wooden block out of the boat and t i m i n g how long it took to f a l l behind the boat 3 2 feet, the length of a piece of s t r i n g attached to the block. One speed check r e a d i n g was the average footnote 1 continued The speed at which a l u r e f i r s t exhibits definite action i s s a i d to be the " minimum t r o l l i n g speed" for that l u r e . The f a s t e s t speed b e f o r e it starts the e r r a t i c movement is c a l l e d the "maximum t r o l l i n g speed"; and between these two i s the range for the "intended" or " p r o p e r " action of the l u r e . To the extent that the proper action of the l u r e i s a s s u m e d to be that action exis t i n g b e f o r e cavitation, it could be c h a r g e d that this is a r a t h e r sub-je c t i v e determination, and not based on any known intention of the manufacturer. Nevertheless, it was only at speeds below this c r i t i c a l speed that a c l e a r d i s t i n c t i o n was evident to the human eye between the v a r i o u s actions exhibited by the d i f f e r e n t l u r e s . 8 of three measurements. The p r e c i s i o n of this method was such that the three readings s e l d o m d i f f e r e d by more than three seconds. A s the motor launch would not go slow enough f o r t r o l l i n g at m i n i m u m throttle, l a r g e p l a s t i c drogues were dragged. With one drogue attached f r o m the bow and one f r o m the stern, the speed attained was 1.75 f t / s _+ 0.25 f t / s . T h i s speed was within the range of t r o l l i n g speed for a l l four actions tested, and was the speed used f o r experiments I and II. In experimert III a t h i r d drogue was dragged f r o m the bow, and this f u r t h e r r e d u c e d the speed to 1.5 f t / s +0.25 f t / s. T h i s speed was more to the center of the range f o r f l a t f i s h action and was the bare m i n i m u m speed n e c e s s a r y f or p r o p e r action of the dodger 2. In e x p e r i m e n t I, i n which only the motor launch was used, four r o d ho l d e r s were attached at the stern of the boat to hold the tips of the rods about four feet apart i n a line p e r p e n d i c u l a r to the path of t r a v e l . E a c h r o d was assigned a po s i t i o n number (1, 2, 3, 4) sta r t i n g with the far right-hand r o d and ending with the f a r left-hand rod, and these numbers were maintained through-out the experiment. The four l u r e s used in a given f i s h i n g p e r i o d were a s s i g n e d at random. Two men i n the boat took turns either d r i v i n g or attending the l i n e s . In experiment II and III two boats were used r a t h e r than one. A n a s s i s t a n t f i s h e d f r o m the s m a l l outboard boat with two lines, and I f i s h e d two l i n e s f r o m the launch. P o s i t i o n s 3 and 4 were a s s i g n e d to the outboard boat, and the l u r e s were again a s s i g n e d at random. The f i s h i n g li n e and leade r were 6 lb nylon monofilament. Ninety feet of line t e r m i n a t e d in a b r a s s s w i v e l c l i p , attached to which was a b r a s s swivel, four feet of leader, and the l u r e . The sinker weight was attached to the line just beside the swivel c l i p on the r o d - s i d e . D e s c r i b e d i n T r o l l i n g E x p e r i m e n t III methods. See also F i g u r e 1. 9 E a c h hour of f i s h i n g was d i v i d e d into p e r i o d s of 20 or 30 minutes, depending on the experiment, with a line check (r e e l i n g in the l i n e s and checking for fouled l u r e s , etc.) at the end of each p e r i o d . A new p e r i o d s t a r t e d when the l a s t checked line was back in f i s h i n g p o s i t i o n . To check the four l i n e s n o r m a l l y took 8 - 1 0 minutes, so that in each p e r i o d a l u r e was out of its f i s h i n g p o s ition f o r 2 - 2 . 5 minutes. If a f i s h was caught v e r y c l o s e to the time of the next line check (within about 5 minutes of it), the line on which the f i s h was caught was not checked for that p e r i o d . T o r e e l i n the line, r e move the fi s h , and r e t u r n the l u r e to i t s fi s h i n g p o s i t i o n took about 3.5 minutes at least, and up to 5 minutes. Whenever a p r o b l e m o c c u r r e d such that one line m i s s e d out on a sig n i f i c a n t amount of f i s h i n g (due to bad fouling of the line, or breakage, or for some other reason), an attempt was made to compensate for this by f i s h i n g e x t r a time. U s u a l l y this was added on to the end of a f i s h i n g period, but once or twice d u r i n g the three t r o l l i n g e x p e r i m e n t s it had to be done the next day. In experiments I and II the l u r e s were painted red, yellow, g r e e n and 3 blue. The complete l u r e was painted, c o v e r i n g both m e t a l l i c parts and p r e v i o u s l y painted p a r t s . N o r m a l l y a l u r e was given two coats of paint and le f t 48 hours to dry, but as p r e l i m i n a r y t r i a l s showed no d e t r i m e n t a l effects due to f r e s h paint, l u r e s were used that had a s h o r t e r c u r i n g time. T h i s happened o c c a s i o n a l l y when there was a high l o s s of a p a r t i c u l a r c o l o r of l u r e . The f i s h i n g p e r i o d for a l l three experiments was of three hours d u r a t i o n o c c u r r i n g between 6:00 and 10:00 a.m. i n the morning, and between 6:00 and 10:00 p.m. in the evening (all t i m e s P a c i f i c Daylight Saving). The length of each f i s h caught was measured, and its sex was determined. F i s h caught and f i s h hooked but not landed were counted i n the evaluation of a l u r e . To count as a v a l i d strike, the f i s h either had to be seen on the line or had to make its presence indisputably obvious. A single j e r k on the line was usu a l l y d i s r e g a r d e d due to the p o s s i b i l i t y of it being a r e l e a s e d snag. T e s t o r 1 s hobby enamel paint, numbers: 3 red, 14 yellow, 24 green, and 8 blue. 10 C. T r o l l i n g E x p e r i m e n t I. 1. Methods F o u r actions and four c o l o r s of l u r e were t r o l l e d behind an 18 ft motor launch at L o o n Lake, B r i t i s h C o l u m b i a at two di f f e r e n t t i m e s of day through-out the p e r i o d J u l y 10 to J u l y 17, 1972. The four actions were f l a t f i s h , spinner, spoon, and dead (see T a b l e I); and the four c o l o r s were red, yellow, green.and blue. A No. 10 n i c k e l t r e b l e hook (point to point distance of 11 mm) was substituted for the hooks that came with the l u r e s (these were used f o r a l l three experiments), and a 5.5 g r a m r u b b e r c o r sinker was attached to the lin e . A f i s h i n g route was planned that would take three hours and p a r a l l e l the shoreline around the southwest end of the lake a p p r o x i m a t e l y 100-150 feet off-shore. F r o m a s e r i e s of soundings along the route, the average depth was 55 feet. A n effort was made to keep as c l o s e as p o s s i b l e to the planned route. The unit of time f o r f i s h i n g was 20 minutes, with l i n e checks at the end of each p e r i o d . M o r n i n g and evening runs were t r e a t e d as separate experiments for the purpose of a s s i g n i n g l u r e s -- 16 combinations of actions with c o l o r s - - t o each r o d . B e f o r e the expe r i m e n t started, the or d e r of f i s h i n g these combinations was ra n d o m l y d e t e r m i n e d for both the morning and evening runs, and after the f i r s t r e p l i c a t e was completed the o r d e r of f i s h i n g was s i m i l a r l y d e t e r m i n e d for the second r e p l i c a t e . The number of s t r i k e s and landings for each l u r e was r e c o r d e d for each fi s h i n g period, and this total was used as a datum point. D a t a were analyzed in a 3-way ana l y s i s of v a r i a n c e wither, the chance of making a Type I e r r o r , =.0 5. 2. R e s u l t s In t r o l l i n g experiment I over half of the f i s h were caught with the flat-f i s h lure; only two were caught with the de a d l l u r e (see Table II). The spoon T A B L E II. Numbers of f i s h caught on l u r e s f r o m t r o l l i n g experiment I, L o o n Lake, B.C., July 10 - 17, 1972. A N O V A table f r o m the a n a l y s i s of this data. E a c h group of numbers reading f r o m top to bottom r e p r e s e n t s the num-ber of f i s h caught i n the f i r s t and second r e p l i c a t e s , and the total f o r both r e p l i c a t e s . A c t i o n Red C o l o r of L u r e Y e l l o w G r e e n Blue T o t a l f i s h caught f o r each action Dead ..Morning F i s h i n g P e r i o d Evening F i s h i n g P e r i o d Spoon 2 1 0 1 0 2 1_ 2 9 2 3 1 3 Spinner 1 2 0 0 0 0 0 3 6 1 2 0 3 F l a t f i s h 3 1 2 1 4 1 3 0 15 7 2 5 1 10 8 7 7 32 Dead 0 0 0 0 0 0 0 0 0 0 0 0 0 Spoon 3 0 1 1 0 0 2 1 8 3 0 3 2 Spinner 0 2 1 0 0 0 2 2 7 0 2 3 2 F l a t f i s h 2 3 2 2 5 1_ 2 18 7 4 3 4 10 6 9 8 33 To t a l f i s h caught for each c o l o r 20 14 16 15 . 65 A N O V A T A B L E Source df MS Prob. C o l o r 3 0.43 .50< p<.75 A c t i o n 3 10.31 p « . 0 0 1 T i me 1 0.02 p ».7 5 C o l o r x A c t i o n 9 1.56 .10<p<.25 Co l o r x Time 3 0.18 p>.75 A c t i o n x Ti m e 3 0.31 p>.75 Col o r x A c t i o n x Ti m e 9 1.09 .10<p<.25 E r r o r 32 0.92 T o t a l 63 12 was s l i g h t l y m o r e s u c c e s s f u l than the spinner, the f o r m e r having about one-half the e f f i c i e n c y of the f l a t f i s h . T h i s action effect was shown to be s i g n i f i -cant (p •< .001). No di f f e r e n c e was found between the morning and evening f i s h i n g periods, a p p r o x i m a t e l y 50 per cent of the f i s h being caught d u r i n g each pe r i o d . R e d l u r e s caught the most f i s h , but f a i l e d to show a s t a t i s t i c a l l y s i g n i f i c a n t s u p e r i o r i t y over the other c o l o r s . V e r y l i t t l e d i f f e r e n c e showed up between the yellow, green and blue l u r e s , either i n the morning or evening runs. The data suggest the p o s s i b i l i t y of an i n t e r a c t i o n between the c o l o r and action of the lure, both i n the absolute sense and r e l a t i v e to the time of day of fishin g , although the effect was not sig n i f i c a n t . No s e l e c t i o n o c c u r r e d f o r s i z e or sex of f i s h either on the b a s i s of c o l o r or action of the l u r e . The average length (fork length) of f i s h was 27 cm, with a range of f r o m 13 c m t o 34 c m (standard d e v i a t i o n = 3.75). The o v e r a l l sex r a t i o was 1:1, with no si g n i f i c a n t deviations o c c u r r i n g f o r any of the c l a s s e s of l u r e for which a sex ra t i o could be determined. No f i s h were sexed f o r either the dead l u r e s or the yellow l u r e s . A l s o , there were no sexes r e c o r d e d for the r e d spinner, because the sole f i s h hooked on that l u r e was not landed. D. T r o l l i n g E x p e r i m e n t II 1. Methods In. a furt h e r attempt to show a sig n i f i c a n t c o l o r preference, the best action f r o m experiment I (flatfish) was chosen, and four l u r e s were painted red, yellow, green and blue. E a c h l u r e was then f i s h e d three t i m e s in the morning and three t i m e s in the evening between August 5 and August 9, 1972. A f i s h i n g route was mapped out at the north east end of the lake such that one c i r c u i t could be made each hour. T h e r e f o r e , in one p e r i o d of f i s h i n g the boats made appro x i m a t e l y three c i r c u i t s . The route entailed f i s h i n g a 13 U-shaped pattern along the s h o r e l i n e and then c r o s s i n g the lake, connecting the a r m s of the U. The s h o r e l i n e f i s h i n g made up a p p r o x i m a t e l y 7 5 per cent of the fi s h i n g t i m e in an average depth of water of 21 feet (6.4 meters), while the open-water f i s h i n g took the r e m a i n i n g 25 per cent of the time and was i n v a r y i n g depths not exceeding 50 feet (15.2 mete r s ) . A s there had been l i t t l e f o u l ing of lines in t r o l l i n g e x p e r i m e n t I, the unit of f i s h i n g time was i n c r e a s e d to 30 minutes. The weight of the s i n k e r was d e c r e a s e d f r o m 5.5 g r a m s to 3.0 g r a m s as repeated depth sounder readings i n d i c a t e d the f i s h were c l o s e to the surface. L u r e s were assigned at r a n d o m to the four r o d s in a s i m i l a r manner to that for experiment I. Again, the numbers of s t r i k e s and landings f o r each l u r e was recorded, and the data were analyzed i n a two-way a n a l y s i s of v a r i a n c e with a = .0 5. 2. R e s u l t s The morning f i s h i n g p e r i o d gave a g r e a t e r y i e l d than the evening p e r i o d (see Table III). T h e r e was no evidence for the existence of a c o l o r p r e f e r e n c e . A l though no c o l o r caught more than red, yellow caught as many, and g r e e n and blue were only three f i s h behind the l e a d e r s . T h e r e was no i n t e r a c t i o n between the effects of c o l o r and time of f i s h i n g . Lengths of f i s h caught were d i s t r i b u t e d over the four c o l o r s such that there was no s e l e c t i o n by any one c o l o r . A s some of the f i s h caught were ' m e a s u r e d as to length and thrown back immediately, it is not p o s s i b l e to r e p o r t on the sex of every f i s h caught; but of those f i s h caught and l a t e r sexed, 22 out of 32 were f e m a l e s . T h e r e were no substantial deviations f r o m this r a t i o among the four c o l o r s of l u r e tested. T A B L E III. Number s of f ish caughton lures f rom troll ing experiment II, Loon Lake, B.C., August 5 to 9, 1972. ANOVA table f rom the analysis of this data. Each group of three numbers reading f rom top to bottom represents the number of f ish caught in the f i rst , second, and third repl icates. Fishing Period Red Colors of Lures Yellow Green Blue Totals Morning 3 5 2 o-5 4 4 5 3 41 Evening 1 1 4 2 4 0 17 16 16 13 13 58 ANOVA T A B L E Source df MS Prob. Color Time Color x Time E r r o r Total 3 1 3 16 23 0.50 24.00 1.89 2.54 p > .75 .005 < p < .01 .5 < p < .75 15 E . T r o l l i n g E x p e r i m e n t III 1. Methods The best c o l o r f r o m t r o l l i n g e x periment II (red) was chosen and com-bined with white i n v a r i o u s patterns: s o l i d red, c h e c k e r e d r e d and white, s t r i p e d r e d and white, and s o l i d white ( F i g u r e 1). E a c h of these patterns was fished f r o m August 25 to September 1, 1972, with and without a "dodger" -- a 5 c m long c h r o m e - p l a t e d spinner f r o m a D a v i s gang t r o l l l u r e (see F i g u r e 1). The dodger spun around a six i n c h length of w i r e attached at either end to the line and l e a d e r with b r a s s s w i v e l snaps. A t h i r d f i s h i n g route was mapped out a p p r o x i m a t e l y one-half way down the lake c o n s i s t i n g of a p p r o x i m a t e l y 55 per cent s h o r e l i n e f i s h i n g and the r e s t open water fis h i n g . The average depth for the s h o r e l i n e s e c t i o n was 53 feet (16.25 meters), while open water depths exceeded 50 meters. The route was planned so that a p p r o x i m a t e l y three c i r c u i t s c o u l d be made i n the three-hour f i s h i n g p eriod. A s the lmes became fouled f a i r l y frequently in t r o l l i n g e x p e r i m e n t II (although this was mostly caused by f i s h i n g in shallow waters), the unit of f i s h i n g time was d e c r e a s e d f r o m 30 minutes to 20 minutes. F r o m echo-sounder readings it appeared that f i s h were at greater depths in this p art of the lake, and t h e r e f o r e a 23 g r a m sinker was used i n place of the 3.0 g r a m sinke r of experiment II. Two a s s i s t a n t s helped with this e x p e r i m e n t -- the f i r s t two r e p l i c a t e s were done by one of the a s s i s t a n t s and myself. Because of a shortage of time the t h i r d r e p l i c a t e could not be completed by the second a s s i s t a n t alone, so I did half of i t . The data f r o m the t h i r d r e p l i c a t e were pooled and analyzed as if the whole r e p l i c a t e had been done by a single person. E a c h l u r e combination was assigned at r a n d o m to a l i n e and the number of s t r i k e s and landings were r e c o r d e d . The data were analyzed in a four-way F i g u r e 1. C h e c k e r e d and s t r i p e d f l a t f i s h , and dodger f r o m t r o l l i n g e xperiment III, L o o n Lake, B.C., August 25 to September 1, 1972. 17 a n a l y s i s of v a r i a n c e with main effects being c o l o r pattern, p r e s e n c e or absence of a dodger, time of day, and i n v e s t i g a t o r effect (the i n d i v i d u a l d i f f e r e n c e s due to different f i s h e r m e n ) . Chance of making a Type I e r r o r was set at a = .0 5. 2. R e s u l t s A l t h o u g h the a l l - r e d l u r e caught more f i s h than any combination of r e d with white or the all-white lure, its s u p e r i o r i t y was not s t a t i s t i c a l l y s i g n i f i c a n t (see Table IV). The same number of f i s h were caught using a dodger as when not using one, and there was no s i g n i f i c a n t effect due to the time of day. The factor that came c l o s e s t to s t a t i s t i c a l s i g n i f i c a n c e was the effect of d i f f e r e n t p ersons doing the f i s h i n g (the i n v e s t i g a t o r effect): 46 f i s h were caught by one f i s herman, 40 by another, and 28 by the t h i r d . With these r e s u l t s , the p r o -b a b i l i t y of r e j e c t i n g the n u l l hypothesis i n c o r r e c t l y i s between 0.1 and 0.25. T h e r e was one s i g n i f i c a n t i n t e r a c t i o n effect between the c o l o r pattern of the l u r e and the investigator effect. T h e r e was no s e l e c t i o n for length of f i s h on the b a s i s of c o l o r pattern of l u r e . However, l u r e s t r o l l e d with dodgers caught s i g n i f i c a n t l y l a r g e r f i s h than those without dodgers (see Table V). The length d i s t r i b u t i o n s show a d i f f e r e n c e i n means of 2.3 which i s shown to be s i g n i f i c a n t by the t-test at a-.0 5. The o v e r a l l sex r a t i o of f i s h caught was 3$ : If/, and there were no important deviations f r o m this r a t i o among the v a r i o u s l u r e s , either with or without dodgers. F. T r o l l i n g E x p e r i m e n t s — D i s c u s s i o n The data f r o m experiment I show a c l e a r effect of action of l u r e on the number of f i s h caught, the l u r e s d i f f e r i n g in both quantity and quality of action. The extent of v e r t i c a l and h o r i z o n t a l motion of the lure, the j e r k i n e s s or smoothness of its action, the degree of r e g u l a r i t y or randomness in the action, the r a t i o of r o t a t o r y motion to l i n e a r motion: a l l these are c h a r a c t e r i s t i c s which help to d e s c r i b e the qualitative action of the l u r e s . In c o n t r a d i s t i n c t i o n 18 T A B L E I V . Numbers of f i sh caught on lures f r o m tro l l ing exper iment III, L o o n L a k e , B . C . , August 2 5 to September 1 , 1 9 7 2 . A N O V A table f r o m the ana lys i s of this data. M O R N I N G F I S H I N G P E R I O D C o l o r Pat tern of L u r e F i s h caught ; Investigator A l l - r e d A l l - w h i t e Str iped C h e c k e r e d per invest igator No Dodger A 7 5 6 2 2 0 B 2 4 1 1 8 C 0 2 0 2 _ 4 - 9 1 1 7 5 3 2 With Dodger A 4 0 1 3 8 B 0 4 4 0 8 C 3 3 3 4 1 3 7 7 8 7 29 1 6 1 8 1 5 1 2 61 E V E N I N G F I S H I N G P E R I O D No Dodger A 3 0 3 2 8 B 3 2 0 2 7 C 4 1 3 2 1 0 1 0 3 6 6 2 5 With Dodger A 4 1 4 1 1 0 B 1 3 1 0 5 C 4 0 3 6 11 9 4 8 7 2 8 = = = = = » 1 9 7 1 4 1 3 5 3 A N O V A T A B L E Source df M S P r o b . Co lor pattern 3 . 1 . 8 6 . 5 < p < . 7 5 Dodger 1 0 . 0 0 p » . 7 5 T i m e 1 1 . 3 3 . 2 5 < p < . 5 0 Inve stigator 2 5 . 2 5 . 1 0 < p < . 2 5 Color pattern x Dodger 3 1 . 0 0 . 5 < p < . 7 5 C o l o r pattern x T i m e 3 3 . 2 2 . 2 5 < p < .50 C o l o r pattern x Investigator 6 6 . 5 3 . 0 2 5 < p< . 0 5 Dodger x T i m e 1 . 7 5 . 5 < p < . 7 5 Dodger x Investigator 2 7 . 7 5 . 0 5 < p < .IX) T i m e x Investigator 2 4 . 0 8 . 1 0 < p < . 2 5 E r r o r 2 3 2 . 4 2 T A B L E V. Mean lengths (cm) o f f i s h caught on l u r e s f r o m t r o l l i n g e x p eriment III, L o o n Lake, B.C., on the b a s i s of c o l o r pattern of l u r e and on the b a s i s of whether or not a dodger was used. August 25 to September 1, 1972. Y = mean length of f i s h caught; n = sample size; s = standard deviation. C o l o r P a t t e r n of L u r e A l l r e d A l l - w h i t e S t r i p e d C h e c k e r e d 23 . 1 6 26.48 25.3 14 20 15 5.7 3.33 5.14 n s 25.94 18 4.01 No Dodger With Dodger Y 24.20 26.51 n 33 34 s 5.43 3.22 20 to these concerns i s the absolute quantity of mot ion which a lu re exhibi t s . Al though i t may be very diff icul t to devise a method for a s sess ing the r e l a t ive quantities of motion over the different types of actions, d is t inc t ions are eas i ly made when the differences a re l a rge . The data f r o m exper iment I could conceivably be in te rpre ted who l ly on the degree of act ion which each lu re exhib i t s . Whi le the spinner might be sa id to exhibit a greater amount of ac t ion than the spoon (due to the speed of the r evo lv ing member) , the o v e r a l l path of the lu re i s in a s traight l ine without z i g - z a g s or wobbles . A l s o , because the spinning member i s a lways at a constant angle f r o m the body of the lu re when a constant speed of t r o l l is maintained, and because of i ts high rate of revolut ion, it tends to lose its appearance of motion and looks more l ike a semi - t ransparen t cone surrounding the body of the l u r e . Therefore , at least to the human eyes, the spinner lu re appears to exhibi t r e l a t ive ly l e ss act ion than the spoon. A s s u m i n g this appearance to be the same for fish, the f ishing success of the l u r e s i s c o r -re la ted with the quantity of ac t ion. No attempt has been made to de termine which of the quali tat ive or quantitative c h a r a c t e r i s t i c s i s the more c r i t i c a l . A n a l y s e s of var iance ( A N O V A s ) were done on the lengths of f i sh caught i n t r o l l i ng exper iment I on the bas i s of both act ion and co lo r of lu re , and, as stated i n the resu l t s section, no d i f ferent ia l se lec t ion showed up. A s a further check, the no rma l i t y of the d i s t r ibu t ion of lengths was examined using the probabi l i ty paper method (Sokal and Rohlf, p. 122-123). Except for the two ta i l s , each of which represent single individuals , the l i n e a r i t y of the graph is what one would expect f r o m random sampl ing out of a n o r m a l l y d i s t r ibu ted population, i .e . , there was no d i f fe ren t ia l se lec t ion among the l u r e s . In the genera l methods for the t ro l l i ng exper iments it was stated that i t took three to five minutes to ca tch a f i sh and r e t u r n the lu re to i ts f i sh ing pos i t ion . It might be objected, then, that the number of f i sh r ecorded for each category does not ref lect the true efficiency of each lure since the t ime of f ishing was shortened by the number of f i sh caught t imes the amount of t ime it took to catch a f i sh . The datum points used in the ana lys is of va r iance are 21 biased in favor of the poorer lures. This bias is probably not as strong as I have suggested, since line checks were not performed on those lines that had just recently been out of the water. Nevertheless, to check against the worst possible case of biasing, each datum point was re-calculated for a shorter fishing period -- the total period minus five minutes times the number of fish caught—and an analysis of variance was done on these transformed values. No significant change was observed for any category of effect. It was assumed that these corrections would have a similar effect on the data from trolling experi-ments II and III, so no further analyses were made. The distribution of fish between the morning and evening fishing periods of experiment II was highly unlikely on the basis of chance alone (p = .007), and so the data probably reflect a different set of circumstances from that of experi-ment I. It is not known what caused this uneven distribution. The color x investigator interaction effect of experiment III is difficult to explain except on the basis that an improbable event occurred. A set procedure was laid out and followed in all aspects of fishing by all investigators. Normally this type of interaction effect would occur if one fisherman always used a particular lure when fishing was good, while another fisherman chose a different one; or if one fisherman tended to be more skilful than another in the use of a particular lure; but, as the order of lure testing was laid out in advance, the former possibility was ruled out, and the experimental design was specifically chosen to limit the latter. Under the conditions of the experi-ment, the only way this effect could have been manifested was if, for instance, one of the fishermen had particular difficulties with a dodger fouling the line and took an excessive amount of time correcting this problem. However, only the color x investigator effect was significant, and the assistants were directed to report any occasions when, for one reason or another, a given line was not fishing as long as the others (for a time greater than 20 minutes), so that a suitable correction could be made. This occurred only two or three times during the course of all three of the experiments. 22 Although I have not come ac ross any study on se lec t ion by dodgers per se, L a r k i n (1949) mentions that a l l of the gang t r o l l s used on Kootenay L a k e selected for s m a l l e r f i sh . There i s a s i m i l a r i t y between a lu re preceded by a dodger and the hooked end-piece of a gang t r o l l preceded by non-hooked f l a she r s . The fact that the s i l v e r dodger used i n this exper iment was actual ly a f lasher f r o m a Davis gang t r o l l lends further c r e d i b i l i t y to the c o m p a r i s o n . F r o m this point of v iew my resu l t s d isagree wi th those of L a r k i n . However , the compar i son i s , perhaps, not as c lose as i t might be. F i r s t , i t is most common for the parts of a gang t r o l l to be rather c l o s e l y matched in s ize, shape, and m a t e r i a l . Usua l ly the hooked end-piece is made of a shiny me ta l l i c m a t e r i a l l ike the non-hooked f lashers preceding i t . In my exper iment the l u r e s were a l l cons iderab ly sma l l e r than the dodger, had a different shape (except for the spinner), and had a coat of paint, while the dodgers were of shiny meta l l i c m a t e r i a l . F u r t h e r m o r e , the coat of paint i t se l f r e su l t ed i n a different type of ref lected l ight f r o m the dodgers and the l u r e s . Whi le gang t r o l l s might have up to six members joined together, my arrangement had only two. Perhaps these differences are enough to expla in the d isagreement between the two studies. It i s in teres t ing to note the p rogres s ive change i n the sex ra t ios throughout the t r o l l i n g exper iments . F r o m an even 50:50 ra t io i n Ju ly i t changed to 2?:lo" in the ear ly part of August arid 3$: Id* i n the la t ter par t . These ra t ios most l i k e l y ref lect changes in the ava i l ab i l i t y of the f i sh and do not point to differences in the se lec t iv i ty of the groups of lu res used i n each exper iment . To be speci f ic , four co lo r s of f la t f ish lu res were used in both exper iments I and II; and for each experiment, the sex ra t ios of the f la t f i sh catch m i r r o r e d that of the o v e r a l l catch f r o m that exper iment . Therefore , the noted differences between experiment I and exper iment II, at least, are not a resu l t of se lec t ion in different groups of l u r e s . F u r t h e r m o r e , there was no d i f ferent ia l se lec t ion among lu res within an experiment , thus giving further evidence against lure se lec t iv i ty being an explanation for the sex r a t ios . One poss ib le factor leading to the inc reas ing p ropor t ion of females in the catch as the summer p rogresses has been suggested by A . Tautz of the 23 B r i t i s h Co lumbia F i s h and W i l d l i f e B r a n c h (personal communica t ion) . It appears that some of the age 2+ males i n the lake w i l l r e a c h a s ize sufficient for them to spawn just before the t ime of the spawning run, and they w i l l thus j o i n the older f i sh in the mig ra t i on out of the lake . None of the 2+ females r e a c h this c r i t i c a l s i ze . The precoc ious males , in going through the r i g o r s of spawning and i n leaving the r i c h feeding areas of the lake, do not g row as fast as their female s ibl ings that stayed behind. L a t e r on, the females , wh ich have enjoyed a longer growing season, r e a c h a s ize sufficient to enter the f i she ry (around 20 cm) . Th is inc rement throws off the ra t io of the th ree -year old and older f i sh i n favor of more females as the season p rog re s se s . The f i sh ing success of the f la t f i sh is notable i n that it contras ts s t a rk ly wi th that for L a r k i n ' s work on Kamloops trout in P a u l and Kootenay L a k e s (1949). In P a u l Lake , except for the month of May, f la t f i sh had a cons is ten t ly lower catch per unit effort than a group of misce l laneous t r o l l s . In Kootenay Lake , out of twelve categor ies of lu re , f la t f i sh ranked poorest . However , there seems to be a marked difference in the success of l u r e s on va r ious l akes . On P a u l Lake the f ly was the poorest of three types of lu re in five out of s ix months, whi le on Kootenay Lake i t was the second best of twelve ca tegor ies over the season. L a r k i n also notes a tendency for lu res which have a h igh ca tchabi l i ty to select for s m a l l f i sh . Th is holds for both lakes; but for Kootenay Lake plugs, f la tf ish, and large spoons are used i n the spr ing and f a l l for what i s essen t ia l ly a separate f i shery (i.e., la rge f i sh over 1-1/4 lb) . F l y f ishing, gang t r o l l s , wobblers and spinners are used i n m i d - s u m m e r and are highly select ive of s m a l l e r f ish , thus making them not s t r i c t l y comparable to the f i r s t group. In L o o n Lake the f la t f i sh caught the sma l l e s t f i s h on the average, and spoons caught the largest ; but the differences were not s ignif icant . Al though in none of the t r o l l i n g exper iments was co lor a s ignif icant effect, i n a l l three the so l id r ed lure caught the most f i sh - - t ied wi th ye l low i n experiment II. 24 P A R T II. F E E D I N G E X P E R I M E N T S A . Introduct ion It was reasoned that i f trout s t ruck at l u res because they mis took them for food, then perhaps more could be learned of their preference for different co lo r s of l u r e s by feeding them s imul taneously different c o l o r s of food, and observing their order of se lec t ion . If four co lo r s of food were presented at a t i m e , ' t h e f i r s t and second co lo r s chosen would indicate most c l e a r l y a p r e -ference, as a se lec t ion would be made among four or three a l ternat ives , r e spec t ive ly . The th i rd and fourth choices would involve fewer a l te rna t ives , and it i s l i k e l y they could be selected on a more r andom bas i s as the " lef t -o v e r s " . Therefore , the most powerful ana ly t i ca l test would be one that ass igned values to the f i r s t and second choices only, and ignored the t h i r d and fourth cho ices . In the two feeding exper iments de sc r ibed here, this sor t of test was per fo rmed . In feeding exper iment I food of four d i sc re te c o l o r s was presented to ind iv idua l ra inbow trout in tanks of var ious co lo r s and in feeding exper iment II food of two co lo r s as w e l l as food of a single co lor was presented to see i f b i - c o l o r e d food would be p re fe r red ; and, i f so, to see i f there was a pa r t i cu l a r second co lo r that elevated the preference l e v e l more than others . B . Gene ra l Methods Trout eggs dyed red, yellow, green and blue were fed to ra inbow trout i n two exper iments between M a y 22 and August 3, 1973. Feedings were genera l ly done every second day between the hours of 2:00 p . m . and 5:30 p .m . (Paci f ic Dayl ight Saving Time) in a l abora tory at the U n i v e r s i t y of B r i t i s h Co lumbia . The f i sh used in the exper iment were juveni le ra inbow trout f r o m the 25 brood stock of the B r i t i s h Co lumbia F i s h and Wi ld l i f e B r a n c h hatchery at Abbotsford, B r i t i s h Co lumbia . When they were brought to the U n i v e r s i t y (October, 1972), they were maintained i n indoor tanks on a 12-hr l ight, 12-hr dark photoperiod unti l the exper iments were f inished,and fed d r i e d f i sh food. A t the t ime when the exper iments were run the mean s ize of the f i sh was 18.7 cm, with 68 per cent of their lengths ly ing i n the range of 16.7 to 20.6 c m . Expe r imen t s were done i n b ipar t i te tanks (see F i g u r e 2) wi th ins ide d imensions of 117 c m length by 61 c m width by 74.5 c m depth. The length of the test chamber was 90 c m and that of the prepara t ion chamber was 26 cm, wi th a s l id ing d iv ider of about 1 c m th ickness between the two, adding up to a total length of 117 c m . This d iv ide r was painted the same co lo r as the r e s t of the tank. On the floor of each test chamber, 8 to 10 c m off the bottom, was a w i r e mesh g r i l l (square mesh) wi th 11 m m holes fo rmed by w i r e of d iameter 1.5 m m . This mesh size was chosen.to a l low food to sink beneath the g r i l l where the f i sh could not r e a c h it after a feeding t r i a l . Sheets of black plas t ic supported by an angle i r o n f r amework surrounded the tank and kept out ambient l ight . A s m a l l peephole was cut in the side w a l l for observat ion without d is turb ing the f i sh . A c r o s s the top of the f ramework was a flat wooden roof to which was attached a Sylvan ia L i f e l i n e coo l white f luorescent f ix ture . The inside of both the p las t ic sheeting and the roof were painted the same color as the tank. E a c h of the four tanks were painted ei ther 4 red, yellow, green or blue. These pa r t i cu l a r paints were chosen on the bas is of their having a high pur i ty of co lo r and a high saturat ion. The water depth i n the tank was kept at 70 c m +_ 2 c m by a f low- through sys t em provid ing 2.7 l / m i n + 0 . 6 1 / m i n of f r e sh de - ion ized water . The temperature of the water r emained wi th in the range of 9 - 12°C throughout " C o l o r Your W o r l d " h igh -g los s mar ine enamel polyurethane: 1. B r igh t Red, 2. Br igh t Blue , 3. Sun Y e l l o w ; C i l u x M a r i n e E n a m e l #6210 T r o p i c G r e e n . Rolled up front flap Peephole Feeding hole Fluorescent lamp Sliding divider Outlet pipe Wire mesh grill F i g u r e 2. Cutaway view of exper imenta l tank used in feeding exper iments I and II at the Un ive r s i t y of B . C . , May 22 to June 6, 1973; and Ju ly 23 to August 3, 1973. 27 feeding exper iment I, and wi th in 12 - 13 .5°C throughout feeding exper iment II. The water was aerated wi th a bubbler which was turned off 20 minutes before an exper iment and remained off throughout the exper iment . In feeding exper iment I the f luorescent l amps were used wi th f u l l l i ne voltage (120 volts) in each of the four test tanks. Th i s gave the fo l lowing values of i l lumina t ion , measured wi th an underwater photometer ( G M model 1 5 - M - 0 2 / 1 ) equipped wi th a Weston c e l l , for the bot tom center of each tank: r e d tank, 4 5 lux; ye l low tank, 130 lux; green tank, 53 lux; and blue tank, 44 lux. In the second exper iment it was poss ib le to adjust the l ine voltage to give a l ight reading of 72 lux + 10 lux for each tank at a depth of 32 c m beneath the surface. Th is depth represents approximate ly the halfway m a r k i n the water co lumn and is about the depth where the f i sh would be using v e r t i c a l l y ref lected l ight to see the eggs. Trout eggs f r o m the Summer land Trout Hatchery of the B r i t i s h C o l u m b i a F i s h and Wi ld l i f e B r a n c h were used as exper imenta l food. The f rozen eggs were al lowed to thaw i n a cup of water and left to soak for two days. They were then d iv ided among four c o l o r s of dye bath cons i s t ing of 1 par t by weight of fabr ic dye to about 200 parts water . They were left i n the dye for 48 hours , and shaken gently about every 12 hours to ensure even d i s t r i bu t ion of the dye over the complete surface of the eggs. The exact concentra t ion of the dye solution was not c r i t i c a l , and measurements were usual ly made by eye. (The eggs reached a s i m i l a r shade of c o l o r i n g throughout a wide range of concentra-tions of dye.) They were removed f r o m the dye bath and r i n s e d in co ld f r e sh water for at least 30 minutes before being used i n an exper iment . Dyed eggs could be s tored in f resh water in a re f r ige ra to r for two to three days without los ing an appreciable amount of c o l o r . A t least 48 hours before an experiment, four f i sh were selected at r andom f r o m a stock of about 70 in the holding tanks (except that an effort was Tintex F a b r i c Dyes : #50 E n s i g n Red, #5 B r i l l i a n t Y e l l o w , #49 C r e m e de Menthe; R i t Concentrated Tint and Dye, #27 Evening B l u e . 28 made to get f i sh of at least 17.5 c m length), and one was placed i n each of four exper imenta l tanks. C . Feeding E x p e r i m e n t I 1. Methods The f i r s t feeding exper iment was conducted between M a y 22 and June 6, 1973, i n a l abora tory at the U n i v e r s i t y of B r i t i s h C o l u m b i a . The exper imenta l food was prepared by taking one egg f r o m each co lo r of dye batch and fo rming a group of four on a wet 8 inch by 8 inch p lex ig lass plate. In this manner 25 to 30 groups of four di f ferent-colored eggs were placed on the plate (enough for 20 serv ings plus a few extras) . A group of four eggs was s l i d off the p lex ig lass through the hole i n the top of the roof using a s ta inless steel spatula. This constituted one " s e r v i n g " . The eggs f e l l through the a i r to the water surface; there they were vulnerable to the trout un t i l they dr i f ted through the w i r e mesh on the bot tom of the tank. A feeding usual ly cons is ted of twenty serv ings , each of which fol lowed the previous one by about th i r ty seconds. The c o l o r s of the f i r s t and second eggs eaten were recorded , p rov id ing the f i sh ate them before any one of the four eggs dr i f ted through the w i r e mesh g r i l l . If an egg did drift through the g r i l l f i r s t , then the resu l t s f r o m that se rv ing were rejected as a l l co lo r s of egg did not have an equal chance (length of time) to be eaten. If the f i r s t two eggs were eaten s imultaneously, they were both sa id to be f i r s t choices (and scored as such), and a second choice was not r eco rded for that se rv ing . This occas iona l ly happened, and it r e su l t ed in the total score for some serv ings being s l ight ly higher than for others . Serv ings were continued unti l twenty sets of readings had been taken. Often 21 serv ings were r equ i r ed to get twenty sets of readings, but the number of ex t ra servings se ldom went over two, and never over four . The order in which the f i sh i n 29 the four tanks were fed was randomized for each day 's feeding. A score of two was given each time a co lo r of egg was chosen f i r s t and a score of one was assigned to a second choice . Therefore , for a given feeding a total score of 40 i n one co lo r n o r m a l l y indicated 100 per cent s e l e c -tion, i .e. , that co lor was chosen f i r s t out of the four for each serv ing i n that feeding. The score for each co lo r was d iv ided by 40, conver ted to a per cent, and then t r ans fo rmed using the a r c s i n t r ans fo rmat ion . E a c h t ransformant so attained represented one data point for the analys is of va r i ance . The l e v e l of s ignif icance was set a t a = .05. A second repl ica te was done before the f i sh were t r a n s f e r r e d f r o m one tank to the next. T rans fe r then o c c u r r e d wi th in 24 hours of i ts comple t ion . The f i sh a lways had at least 24 hours to a c c l i m a t i z e to a tank before the next exper imenta l feeding. The order in which the f i sh were moved f r o m tank to tank was f r o m green to ye l low to red to blue to green . Af te r a feeding was f in ished the d iv ider between the test chamber and the p repara t ion chamber was r a i s e d and the f i sh was "encouraged" into the prepara t ion chamber . The d iv ide r was then dropped and the meta l g r i l l could be r emoved f r o m the test chamber , thus a l lowing the excess food to be c leaned out without phys i ca l ly d is turbing the f i sh . A s soon as the cleanup was c o m -pleted, the g r i l l was replaced, the d iv ider ra i sed , and the f i sh was "encouraged" back to the test chamber . 2. Resu l t s R e d - c o l o r e d food was the o v e r a l l favor i te , wi th an average se lec t ion intensi ty 50 per cent of the highest poss ible (if r ed was chosen f i r s t each t ime) . 30 This was fol lowed by blue, yel low, and green wi th se lec t ion in tens i t ies 43 per cent, 39 per cent and 36 per cent r e spec t ive ly of the highest poss ib le . The effect of food co lo r was s ignif icant at the .001 l e v e l of p robabi l i ty (see Table VI ) . Al though the color of the tank is s t a t i s t i ca l ly important i n modifying the choice of co lo r s of food, no c lea r pattern emerges (see Table VII and F i g u r e 3). The greatest range in in te rac t ion effects o c c u r r e d i n the r e d tank, but the order of se lec t ion of food was not i n agreement wi th that for the o v e r -a l l exper iment . The order of se lec t ion of va r ious co lo rs of food is also dependent upon wh ich f i sh i s doing the select ing (see Table VIII and F i g u r e 4). A l l of the f i sh have the highest se lect ion intensi ty for red, but f a i l to concur on the order for the other c o l o r s . Fu r the r , a f i sh ' s order of se lect ion changes depending on the co lo r of the tank in which it i s he ld . F o r three of the f i sh this modi f ica t ion was of sufficient strength so as to effect a r e v e r s a l in se lec t ion order between tanks, i .e . , a co lo r that was the favori te i n one tank was the leas t p r e f e r r e d in a d i f fe ren t -co lored tank. Select ion intensi ty is a measure of the degree to which a f i s h selects and eats a pa r t i cu la r i t e m of food when it has a choice of s e v e r a l . It may ref lect a h i e r a r chy of preferences wh ich the f i sh has toward the food i tems, or it may resu l t f r o m a different set of factors , such as the v i s i b i l i t y of an i t e m against a given background, thus giving no indica t ion whatsoever as to the f i sh ' s l ikes and d i s l i k e s . The t e r m "per cent se lec t ion in tens i ty" r e fe r s to the score for a co lo r d ivided by the total poss ib le score for a feeding (40), expressed as a per cent; " se lec t ion intensi ty", by itself , r e fe r s to the a r c s i n t ransformant of the per cent se lec t ion intensi ty . The t e rms "preference", "preference intensi ty", and "per cent preference in tens i ty" w i l l be used in place of the above when the situation seems to indicate that the f i sh is showing responses based on a spec t rum of l i ke s and d i s l i k e s . T A B L E VI. F i s h 1 F i s h 2 Selection intensities of c o l o r s of food in var i o u s c o l o r s of tank. F e e d i n g experiment I, U n i v e r s i t y of B.C., May 22 to June 6, 1973. A N O V A table f r o m the analysis of this data. The upper number is the se l e c t i o n i n t e n -sity f r o m the f i r s t r e p l i c a t e . The lower number is the se l e c t i o n intensity f r o m the second r e p l i c a t e . The highest possible s e l e c t i o n intensity for a feeding (if that color is chosen f i r s t every time) is 90. C o l o r of Food Color of Tank Red Ye l l o w Green Blue Red 28.3 37.8 36.3 50 .8 50.7 39.2 31.6 31 .6 Yel l o w 46.4 30 24.7 47 .9 46.4 40.7 24.7 39 .2 Green 43.6 33.2 37.8 37 .8 45.0 26.6 34.8 45 .0 Blue 37.8 36.3 47.9 30 .0 55.2 36.3 26.6 33 .2 Red 30.0 28.3 49.3 45 .0 24.7 31.6 40.7 52 .2 Yell o w 55.2 43.6 26.6 24 .7 74.1 26.6 26.6 28 .3 Green 37.8 43.6 20.7 46 .4 43.6 37.8 34.8 37 .8 Blue 60.0 26.6 39.2 24 .7 58.4 24.7 30.0 36 .3 F i s h 3 Red Ye l l o w Green Blue 39.2 30.0 45.0 52.2 45.0 37.8 47.9 40.7 30.0 22.8 33.2 33.2 42.1 40.7 24.7 37.8 40.7 47.9 36.3 28.3 28.3 40.7 33.2 31.6 42.1 50.8 37.8 36.3 39.2 34.8 46.4 45.0 F i s h 4 Red Yellow Green Blue E Y Y . 42.1 43.6 49.3 52.2 36.3 47.9 47.9 65.3 1459.6 45.613 43.6 42.1 33.2 39.2 47.9 46.4 36.3 •31.6 1127.7 3 5.241 37.8 28.3 18.4 28.3 24.7 22.8 34.8 26.6 1041.0 32.531 30.0 39.2 47.9 33.2 40.7 34.8 33.2 28.3 1230.6 38.456 4858.9 37.960 Source '• df MS P r o b a b i l i t y Food color . 3 1020 p « . 0 0 1 Tank color x Food color . 9 302 p « . 0 0 1 Indiv. diff. x Food color 9 101 .01<p<.025 Indiv. diff. x Tank colo r x Food col o r 27 89.5 .005<p<.01 E r r o r 64 41.5 32 Mean selection intensities and interaction effects for c o l o r s of food i n v a r i o u s c o l o r s of tank. Feeding E x p e r i m e n t I, U n i v e r s i t y of B.C., May 22 to June 6, 1973. C o r r e c t i o n for food co l o r effects was made by subtracting the c o r r e c -tion factor for the column (color of food) in Table VIIA f r o m each s e l e c t i o n intensity in that column. C o r r e c t e d values are given in Table VIIB. Interaction effects are calculated by subtracting the c o r r e c t e d s e l e c t i o n i n t e n s i t i e s of Table VIIB f r o m the o v e r a l l mean selection intensity for the experiment (37.956) Interaction effects a r e a measure of the effect of the c o l o r of tank on the s e l e c -tion intensity of the c o l o r of food. ^ C o r r e c t i o n factor: Number in parentheses r e p r e -sents the mean selection intensity for a colo r of food minus the o v e r a l l mean s e l e c t i o n intensity for feeding experiment I (37.956). A. U n c o r r e c t e d for F o o d C o l o r E f f e c t Food Color (+7.6)* (-2.7) (-5.4) (0.4) Tank C o l o r R e d Y e l l o w G r e e n Blue Red 36.1 34.4 39.1 42.7 Yellow 52.6 35.0 26.7 36.9 G reen 42.1 39.8 30.6 39.6 Blue 51.6 31.8 33.7 34.6 182.4 141.0 130.1 153.8 607.3 Y = C o r r e c t e d for F o o d C o l o r Effects Red 28.5 37.1 44.5 42.2 Y ellow 4 5.0 37.7 32.1 36.4 G reen 34.5 42.5 36.0 39.1 Blue 44.0 34.5 39.1 34.1 152.0 151.8 151.7 151.8 607.3 Interaction Effects SlYl Red -9.46 - .86 +6.54 +4.24 21.10 Yellow +7.04 - .26 -5.86 -1.56 14.72 G reen -3.46 +4.54 -1.96 +1.14 11.10 Blue +6.04 -3.46 +1.14 -3.86 14.50 EY +0.16 -0.04 -0.14 -0.04 ZIYI 26.00 9.12 15.50 10.80 6l.62| 33 A. U n c o r r e c t e d for Effect of F o o d C o l o r MEAN SELECTION INTENSITY 55 50 45 4CH 35H 3(H 25 Red Yellow Green COLOR OF TANK Blue B. C o r r e c t e d for Effect of F o o d C o l o r 55T MEAN SELECTION INTENSITY 50 45H 40-35 30 25 — i — Red Yellow COLOR OF Green TANK Blue F i g u r e 3. G r a p h of mean select ion intensities of co lors of food in var ious c o l o r s of tank. Feeding experiment I, Univers i ty of B . C . , M a y 22 to June 6, 1973. >^ a* red food; a - o yel low food; B- • -a green food; blue food. 34 T A B L E VIII. Mean selection intensities and interact ion effects for c o l o r s of food among different f ish. Feeding experiment I, Un ivers i ty of B . C . , M a y 22 to June 6, 1973. ^Numbers in parentheses are c o r r e c t i o n factors (see T a b l e VII for explanation of calculat ions) . A . U n c o r r e c t e d for F o o d Color Effect F i s h Number 1 2 3 4 (+7.65)* R e d 44.2 48.0 42.2 48.1 182.5 F o o d (-2.72) Ye l low C o l o r (-5.43) G r e e n 35.0 32.8 33.1 40.0 140.9 33.0 33.5 35.9 27.7 130.1 (+.49) Blue 39.4 36.9 41.6 35.9 153.8 607.3 Y = 37.956 B . C o r r e c t e d for Food Color Effect 1 2 3 4 36.6 40.4 34.6 40.5 152.1 37.7 35.5 35.8 42.7 151.7 38.4 38.9 41.3 33.1 151.7 38.9 36.4 41.1 35.4 151.8 607.3 Interaction Effects 1 2 3 4 -1.36 +2.44 -3.36 +2.54 - .26 -2.46 -2.16 +4.74 + .44 + .94 +3.34 -4.86 + .94 -1.56 +3.14 -2.56 EY EfY| + .26 9.70 + .12 9.62 .14 9.58 - .04 8.20 U n c o r r e c t e d f o r E f f e c t of F o o d C o l o r MEAN SELECTION INTENSITY 45 40 35H 30H 25 20 Fish 1 1 — Fish 2 Fish 3 Fish 4 B. C o r r e c t e d f o r E f f e c t of F o o d C o l o r MEAN SELECTION INTENSITY 45n 40 35H 3 0 H 25 20 Fish 1 Fish 2 Fish 3 Fish 4 F i g u r e 4. G r a p h of m e a n s e l e c t i o n i n t e n s i t i e s of c o l o r s of f o o d a m o n g d i f f e r e n t f i s h . F e e d i n g e x p e r i m e n t I, U n i v e r s i t y of B.C., M a y 22 to June 6, 1 973. n : & r e d food ; o a y e l l o w food; D B g r e e n food; & blue f o o d . 36 3. D i s c u s s i o n It would seem reasonable that the effect of different c o l o r s of tank on the order of se lec t ion of food would be re la ted to the degree of s i m i l a r i t y between the co lo r of food and the co lo r of tank. Two hypotheses emerge f r o m this assumption: (1) that food i n a s i m i l a r - c o l o r ed tank would be l ess v i s i b l e than food i n a con t ras t ing-co lo red tank, so that a resul tant decrease in the se lect ion intensi ty f r o m that expected on the bas i s of co lo r of food alone would be manifest; and (2) that the se lec t ion intensi t ies of con t r a s t i ng -co lo red foods would be augmented in re la t ion to their " spec t ra l d i s tance" f r o m the co lo r of tank i n which they are being fed. The f i r s t of these we w i l l c a l l the "camouflage" effect, and the second, the "complement" effect. (One is r e a l l y the obverse of the other.) F r o m Table VII A it can be seen that the se lec t ion in tensi t ies for the camouflage c e l l s (along the diagonal f r o m the top left to the bot tom right) are low with respect to those for the same co lor of food i n the other tanks . The values for red, green and blue are 21 per cent, 6 per cent, and 10 per cent lower than the mean values for their respec t ive c o l o r s of food. (There i s no difference for the ye l low tank.) It i s poss ib le , however, to look at this effect separated out f r o m the effect of the co lo r of food. Table VII B shows what the select ion intensi t ies would be i n this instance. The actual in te rac t ion effects, which s tow the effect of the co lo r of tank on the se lec t ion intensi ty of the c o l o r s of food, are given in Table VII C . Except for the r e d co lo r , this effect was quite weak, and a sum of squares test on the camouflage versus the non-camouflage means of Table VII B fa i l ed to show a signif icant different at cv=.0 5. A g a i n , f r o m Table VII B, it is poss ib le to divide the means such that one group (the "complement" group) represents combinat ions of co lo r s that are spec t ra l ly distant f r o m each other (red wi th green, r ed wi th blue, and ye l low with blue), and the other group represents a l l the other combinat ions . Al though the average in te rac t ion effect for the complementary group is pos i t ive ( +1.39), and the average in te rac t ion effect for the non-complementary group 37 i s negative (-0.84), a sum of squares test on the co r r ec t ed se lec t ion in tensi t ies fa i l s to show a signif icant difference (CY = .05). Therefore , neither the camouflage effect nor the re la ted complementary-color effect are sufficient by themselves to expla in the s ignif icance of the tank co lo r x food co lo r in te rac t ion . Whatever the p rocess i s that is p r i m a r i l y respons ib le for this interact ion, i ts mode of operat ion i s not s y m m e t r i c a l , as the in te rac t ion effect for a co lo r of food in another co lo r of tank (except for co lo r combinations involving blue) is of opposite s ign f r o m that when the c o l o r s are r e v e r s e d . One other thing to notice i s the fact that the in te rac t ion effects are strongest when a red co lo r i s i nvo lved . When the absolute values of the in te rac t ion effects are summed ac ros s the rows , the greatest to ta l occu r s for the red tank, and s i m i l a r l y , when the in terac t ion effects are summed down the co lumns . It could be objected that these are not orthogonal compar i sons , and that the la rge value for the r e d food in the r e d tank boosts the totals for both the row and the co lumn. However , even i f we let that value be equal to the mean of the other ce l l s in the row (and s i m i l a r l y for the column), the totals for interact ions involving r ed are s t i l l seen to be the highest . The mean square of se lec t ion intensi t ies among the tanks was about three t imes as great as that among the f i s h . Th i s is re f lec ted in the r e s p e c -t ive values for p for the two in teract ions (for the tank co lor x food co lo r in te rac t ion p .001, while for the ind iv idua l differences x food co lo r i n t e r -ac t ion .01 < p < .0 25). Therefore , i n making quantitative measurements of food se lect ion based on food co lor , it i s more c r i t i c a l to s tandardize the background co lo r than i t is to get uni formi ty in test subjects. The l e s se r impor tance of ind iv idua l differences in de te rmin ing order of se lect ion shows up in the f o r m of a good consis tency of se lec t ion order between replicates, , both for the f i sh as a whole, and for ind iv idua l f i sh (see Table I X A ) . The order for both rep l ica tes was red, blue, ye l low and green when the f i sh are taken as a group. Look ing at se lec t ion o rde r s of ind iv idua l 38 T A B L E IX. Mean selection intensities and interaction effects f o r c o l o r s of food among different fish, for each r e p l i c a t e . F e e d ing experi-ment I, U n i v e r s i t y of B.C., May 22 to June 6, 1973. The average size of interaction effect is equal to the sum of the absolute values of the intera c t i o n effects in a row or column divided by 8. U n c o r r e c t e d for Food Color E f f e c t F i s h Number Replicate R ed C o l o r Y e l l o w of Food G r e e n Blue 1 1 2 39.0 49.3 34.3 35.7 36.7 29.4 41.6 37.2 2 1 2 45.8 50.2 35.5 30.2 34.0 33.0 35.2 38.6 3 1 2 44.3 40.2 32.5 33.6 34.6 37.1 41.4 41.7 4 1 2 43.9 52.2 40.2 39.8 28.9 26.5 38.0 33.9 EY 364.9 281.8 260.2 307.6 C o r r e c t e d for Food Color Ef f e c t 1 1 2 31.4 41.7 37.1 38.4 42.1 34.8 41.1 36.8 2 1 2 38.1 42.6 38.2 32.9 39.4 38.4 34.7 38.2 3 1 2 36.6 32.5 35.2 36.4 40.0 . 42.6 40.9 41.2 4 1 2 36.2 44.6 43.0 42.6 34.4 31.9 37.4 33.4 EY 303.7 . 303.8 304.6 303.7 Interaction Effects 1 1 2 -6.58 3.71 - .91 .46 4.14 -3.11 3.16 -1.21 A v e r a g e s i z e of i n t e r a c t i o n effect, 2.91 .2 1 2 .14 4.59 .29 -5.06 1.42 .49 -3.26 .19 1.93 3 1 2 -1.34 -5.44 -2.74 -1.61 2.09 4.59 2.92 3.26 3.00 4 1 2 -1.71 6.64 5.02 4.59 -3.61 -6.04 - .51 -4.58 4.09 Average size of interaction effect ' |Y| 3.77 2.58 3.19 2.39 3 9 f i s h it i s seen that the rank for a g iven co lor of food is never more than one different f r o m that i n the o v e r a l l o rde r . Only two f i sh (fish #1 rep l i ca te 1 and f i s h #3 repl ica te 2) had two of these r e v e r s a l s of a co lo r w i th its neares t neighbor; a l l the others had either one r e v e r s a l or none. A l s o , compar ing the se lect ion order for an ind iv idua l f i sh between rep l ica tes , it i s seen that, again, the greatest shift i n rank i s one p lace . Only f i sh number 4 was c o n -sistent between rep l ica tes , and this i s the f i sh that had the greatest average size of in te rac t ion effect (see Table I X C ) . A l l told, there is the greatest agreement on the place of r e d i n the o rder of se lect ion, i t being chosen f i r s t s ix t imes out of eight. The average s ize of in te rac t ion effect for each f i sh i s a measure of the amount that each f i sh deviated f r o m the average se lec t ion order of the group. Compar ing this amount of deviat ion among the f ish, f i s h #4 had twice as much deviat ion o v e r a l l as f i sh #2, and about 50 per cent more than either f i sh #1 or f i s h #3. It can also be seen by examining the co lumn sums that the co lumn for r ed food had the la rges t to ta l . Th is means that, even though the f i sh were most agreed on the place of red in the se lec t ion order , it had the greatest range of "a t t rac t iveness" of any c o l o r . Green came second, and ye l low and blue t ied for t h i rd place. The in te rac t ion between the co lo r of food, co lo r of tank, and ind iv idau l f i s h means that ind iv idua l f i sh make different changes in the i r order of s e l ec -t ion when tested i n one color of tank and then another. A s imple way of "getting at" this effect is to rank the co lo r s of food for each feeding 1, 2, 3, and 4, accord ing to their se lect ion intensi ty va lue . The s u m of the differences in rank order between rep l ica tes for a pa r t i cu la r f i sh w i l l give a measure of the inconsis tency of that f i sh in its choice of food co lo r s (see Table X ) . However , this i s a measure of inconsis tency wi th in a co lor of tank. To get a measure of inconsis tency among tanks, the var iance (S s ) of the mean rank for the two rep l ica tes can be found for each co lo r of food. The average of these four values w i l l be a measure of the inconsis tency of food choice of an ind iv idua l f i sh among the four co lo r s of tank. F r o m Table X it i s seen that f i s h #4 I S the TABLE X. Variability in selection order between replicates and between tanks. Feeding experiment I, University of B.C., May 22 to June 6. 1973. Numbers represent order of selection of food (selection rank) in first and second replicates. S 2 = variance of mean selection rank. See text for explanation. *D : difference Fish <?4 1 and 2). Color of Co lor of Food Tank Red Yellow Green Blue Red 4 2 3 1 I 2 3 3 *3(2.5) + 0(2) 0(3) 2(2) Yellow 2 3 4 1 i 2 4 3 1(1.5) 1(2.5) 0(4) 2(2) E*D =17 Green 1 4 2 2 r 1 4 3 1 0(1) 0(4) 1(2.5) 1(1.5) Blue 2 3 1 4 I 2 4 3 1(1.5) 1(2.5) 3(2.5) 1(3.5) S 2 = (.396) (.750) (.500) (.750) S 2 = .599 Red 3 . 4 1 2 4 3 2 1 1(3.5) 1(3.5) 1(1.5) 1(1.5) Yellow 1 2 3 4 1 3 3 2 0(1) 1(2.5) 0(3) 2(3) £ D =14 Green 3 2 4 1 r 1 2 4 2 2(2) 0(2) 0(4) 1(1.5) Blue 1 3 2 4 1 4 3 2 0(1) 1(3.5) 1(2.5) 2(3) S 2 = (1.396) (.563) (1.083) (.75) S 2 = .948 Red 3 4 2 1 3 4 1 2 0(3) 0(4) 1(1.5) 1(1.5) Yellow 1 4 3 2 1 3 4 2 0(1) 1(3.5) 1(3.5) 0(2) t D =15 Green 1 2 4 3 • r 3 1_ 1 4 2(2) 1(1.5) 3(2.5) 1(3.5) Blue 1 4 3 2 2 3 4 I 1(1.5) 1(3.5) 1(3.5) 1(1.5) S2 = (.729) (1.229) (.917) (.896) S 2 = .943 Red 2 1 3 4 1 2 4 3 1(1.5) 1(1.5) 1(3.5) 1(3.5) Yellow 1 3 4 2 J_ 2 4 3 0(1) 1(2.5) 0(4) 1(2.5) t D =12 Green 3 1 4 2 r 1_ 2 4 3 2(2) 1(1.5) 0(4) 1(2.5) Blue 1 2 3 4 1 2 4 3 0(1) 0(2) 1(3.5) 1(3.5) S2 = (.22')) (.229) (.083) (.333) S 2 = .219 4 1 most consistent both wi th in a given co lo r of tank and between co lo r s of tank. The value of the in ter- tank incons is tency measurement when the ranks are ass igned at r andom among the four c o l o r s of food is S 2 = .625. Therefore , two of the f i sh were more consistent i n their choices than would be expected if they were made on a r andom bas is , while two of the f i sh were marked ly l e s s consistent . Table X I shows the in te rac t ion effects for the second order in te rac t ion (food color x tank co lor x ind iv idua l dif ferences) . Among the va r ious c o l o r s of food the greatest, in terac t ion effects occur for r ed ( | Y | = 4.38). There i s the greatest difference among f i sh in their change of preference for r e d food upon being t r ans fe r red f r o m tank to tank. Th i s is , perhaps, to be expected, as the more neu t ra l an o rgan i sm ' s " fee l ings" are toward an object, the l e s s environmenta l factors w i l l affect those feel ings, and v ice v e r s a . Taken i n conjunction wi th the wide range of a t t ract iveness of r ed to ind iv idua l f ish, this explains why the second-order in te rac t ion effects are greatest for the r ed co lo r of food. However , the t- test fa i l s to show a s ignif icant difference between the in te rac t ion effects for the r e d food and the mean value for those of the other three c o l o r s . In a feeding exper iment such as this, where the f i sh go f r o m having an empty gut to an a lmost ful l one, any preference shown for a g iven co lo r might occur only when the f i sh has a lmost reached a state of satiat ion; so that se lect ion for co lor i s l i m i t e d to the t ime of "dwindling appeti te". If one were doing a forced-feeding p r o g r a m on a group of f ish, it would be in te res t ing to know what color was last "given up" as i t could be used for "desser t" after a feeding of n o r m a l - c o l o r e d food. Th is could be an important factor if i t was diff icul t or cos t ly to dye food that pa r t i cu la r c o l o r . Wi th the intent of invest igat ing this poss ib i l i ty , the data f r o m the exper iment were divided into two par ts : resu l t s of se rv ings 1 - 10, and See Brownlee , K . A . , 1965. T A B L E X I . Interaction effects for second-order interaction (food color x tank color x individual differences) - - average for two rep l i ca tes . Feeding experiment I, Univers i ty of. B . C . , . ..May.22 .to June 6, 1973. F i s h #1 C o l o r of Tank R e d Yel low Green Blue R e d +4.88 -5.02 +3.58 -3.88 C o l o r Ye l low o f F o o d Green +4.48 + .52 -9.52 +4.78 -5.48 -2.62 +5.38 +3.02 Blue -2.38 +5.48 + .92 -3.98 SIYI = 65.92 F i s h #2 R e d Ye l low G r e e n Blue -11.08 +9.42 -3.82 +5.08 -1.88 +2.48 +3.48 -3.68 +5.08 -1.22 -3.68 - .12 +7.52 -9.08 +4.12 -2.08 SlYl = 73.82 F i s h #3 R e d Ye l low G r e e n Blue +1.98 - .88 +2.68 -4.02 -5.72 + .28 +3.88 +1.62 +1.98 +2.08 + .68 -4.72 + .68 -3.22 -5.68 +7.92 SlYl = 48.02 F i s h #4 R e d Y e l l o w G r e e n Blue +4.32 -4.58 -2.52 +2.38 +3.82 -3.62 +2.72 -2.58 -1.08 +1.32 -1.88 +1.78 -5.48 +5.98 + .78 -1.32 SlYl = 46.16 Z | Y | 70.12 55.06 42.12 66.62 A v e r a g e size of i n t e r -act ion effect, IYI 4.38 3.44 2.63 4.16 43 resu l t s of servings 11 - 20. A n ana lys i s of va r iance s i m i l a r to the o r i g i n a l one was c a r r i e d out, except wi th a fourth main effect: whether the data were f r o m the f i r s t half or the second half of the feeding. There were no s ignif icant effects due to the data being f r o m the f i r s t or second half of the feeding (a-.0 5). None of the in terac t ions invo lv ing this factor were s ignif icant . The only noteworthy difference between the two A N O V A s was that the ind iv idua l differences x co lo r of food in te rac t ion was signif icant at the .00 5 l e v e l of p robabi l i ty ra ther than just at the .025 l e v e l . D . Feeding Expe r imen t II 1. Methods Trout were fed b i c o l o r e d food in the p rev ious ly d e s c r i b e d tanks between Ju ly 23 and August 3, 1973. F o u r f i s h were chosen at r a n d o m f r o m a holding tank and one was placed in each of the four exper imen ta l tanks. A l l tests for a given tank were done wi th this f i sh . E x p e r i m e n t a l food was p repared as fo l lows: a hot solution of one par t gelatine to ten par ts water was poured into a flat pan to a depth of 3 - 4 m m . The pan was placed in a r e f r ige ra to r , a l lowing the gelatine to harden for two days. It was then removed, and rec tangular holes were s l i c e d i n the gelatine just l a rge enough to accommodate two eggs side by side (see F i g u r e 5). F o r any given feeding one of the four c o l o r s of egg - - red, yellow, green or blue - - was chosen to be the "tag" color , i .e . , th is co lo r of egg was included as the "tag" and was one member of each p a i r ' o f eggs that were put into a hole in the gelat ine. The other egg was one of the co lo r s red, yel low, green or blue. F o r example, for a given feeding, i f ye l low was chosen as the tag color , the fo l lowing pa i r s of eggs would be formed: ye l low with red, ye l low with yel low, ye l low with green, ye l low with blue (designated Y R , Y Y , Y G and Y B , r e spec t ive ly ) . loo) too] oo tool m m loo] m m foo] (ool foo] |oo| t^ oj m lOOl foo] fool m too] l £ O j lool tool m too] too] loo] \m [750] foo] \oo\ fool tool foo] t£o] fool t M foo] [oo] too] too] too] m loo] m |oo| m loo] m loo] tool too] loo] m m I22) tool looj m too) too] tool looj lool fool tool too] foo] m too] too] m fool fool too] t22l loo| 160] fool foo] t M tool fool fool lool E g 00 Ho] loo] |oo( F i g u r e 5. P r e p a r a t i o n of food for feeding exper iment I I - -arrangement of eggs in the gelatine; plan v iew. Feeding exper iment II, U n i v e r s i t y of B . C . , Ju ly 22 to August 3, 1973. 45 A s l ight ly less concentrated solut ion of gelatine was made up, a l lowed to cool to r o o m temperature, and then poured into the pan to f i l l i n a l l the gaps and provide a thin l aye r comple te ly cover ing the tops of the eggs. The pan was re turned to the r e f r i ge r a to r and the f r e sh gelatine was a l lowed to harden for two days. Immedia te ly before a feeding rectangular b locks just l a rge enough (about .5 c m x .7 c m x 1.0 cm) to include a pa i r of eggs and enough gelatine to support them were cut out and the blocks were placed i n a ja r of co ld water . This a l lowed them to be eas i ly managed without the gelatine d i s s o l v i n g . One block of each of the four co lor combinations was taken f r o m i ts jar to f o r m a group near the edge of a wet p lex ig lass plate. Enough groups were formed to provide a complete feeding for one f ish , plus some ex t ras . A group of four b locks was fed to a f i sh in a s i m i l a r manner as in exper iment I by s l id ing the b locks through the hole in the roof. The f i sh had t ime to s t r ike at the b locks whi le they drifted down f r o m the surface to the g r i l l in the bot tom of the tank. The gelatine binding the eggs together was i n v i s i b l e to the human eye when the b locks were in the wate r . Usua l l y a f i sh s t ruck at and swal lowed both eggs in a block at once; only occas iona l ly d id i t eat f i r s t one egg and then the other. If this d id happen, the resu l t s for that se rv ing were deleted unless the f i sh ate the second egg before s t r ik ing at another b lock. None of the f i sh could take the intended twenty servings , so each was fed as many serv ings as i t would take. Feeding was stopped when the f i sh repeatedly s t ruck at no more than one block or when it regurgi ta ted parts of p rev ious ly eaten eggs. One feeding wi th each co lor of tag was done in a l l four tanks. The order i n wh ich the b locks were eaten was recorded , the co lo r of food being designated by the non-tag egg. A value of 2 was g iven for a f i r s t choice and a value of 1 for second choice . These were then tota l led and d iv ided by 2n (n=the number of servings that a f i sh would take), and conver ted to a 46 per cent. The per cent was in turn converted using the a r c s i n t r ans fo rmat ion and an analys is of var iance was done on the t ransformants . 2. Resu l t s R e d food was the favori te , fo l lowed by yellow, green, and blue food wi th se lec t ion intensi t ies 49 per cent, 43 per cent, 42 per cent and 34 per cent r e spec t ive ly of the m a x i m u m poss ible , i .e . , if that co lo r of food were chosen f i r s t eve ry t ime (see Table XII) . Red, yel low, and green food are in the same order wi th respect to each other as in feeding exper iment I, but blue s l ipped f r o m second to fourth p lace . The in te rac t ion between the tag egg and the food egg (the effect of having b i - c o l o r e d food) was s ignif icant at the .001 probabi l i ty l e v e l . In genera l , having a tag wi th a co lo r wide ly separated in the spec t rum f r o m that of the food enhanced the se lec t ion for that co lo r of food. The effect of feeding the trout i n different co lo r s of tank was a lmost s t a t i s t i -c a l l y significant, but i t is not evident just what this effect cons is ted of. It does not seem to be centered around either the complementa ry co lo r effect or the camouflage effect. A s the second-order in te rac t ion was used as the e r r o r t e r m i n the analys is of var iance , nothing can be sa id about the effect of feeding co lo r s of food wi th c o l o r s of tag i n different co lo r ed tanks. 3. . D i s c u s s i o n The strong tag co lor x food co lor in terac t ion effects are manifest as the effect of complementary co lo r s and the effect of d i v e r s i t y of co lo r (Table XIII), both of which are shown to be signif icant with the t-test (a-.05). The effect of d i v e r s i t y of co lor i s analogous to the camouflage e f fec to f the tank co lo r x food co lo r in teract ions of exper iments I and II except that a s i m i l a r co lo red tag egg cannot r e a l l y be thought of as camouflaging the food egg. Instead, it i s sa id that the two eggs "lack d ive r s i ty of c o l o r " . In contras t to the co lo r of tank x co lor of food effect of exper iment I, the sum of the absolute values of the in te rac t ion effects was least for the red color of food and for the r e d co lo r of tag. Th is means that there was the sma l l e s t difference in se lec t ion intensi t ies 47 T A B L E XII. Selection intensities of b i - c o l o r e d food in var ious c o l o r s of tank. Feed ing experiment II, Un ivers i ty of B . C . , Ju ly 23 to August 3, 1973. A N O V A table f r o m the analys is of the data. The highest poss ible selection intensity for a feeding (if that co lor is chosen f i r s t every time) is 90. F o o d Color Tank color x F o o d co lor T a g color x F o o d co lor E r r o r 3 9 9 27 472 295 936 153 Tag C o l o r o f F o o d C o l o r R e d Y e l l o w G r e e n Blue Red Tank R e d 36.9 15.4 45.0 49.0 Y e l l o w 18.4 0.0 90.0 45.0 G r e e n 49.6 54.7 35.2 0.0 Blue 33.2 63.4 26.6 33.2 138.1 133.5 196.8 127.2 Ye l low Tank R e d 39.2 45.0 26.6 42.1 Y e l l o w 54.9 24.3 45.0 32.6 G r e e n 60.0 56.8 22.8 0.0 Blue 48.4 62.0 32.0 0.0 20 2.5 188.1 126.4 74.7 G r e e n Tank R e d 37.8 48.6 41.4 37.8 Ye l low 42.1 22.8 50.8 42.1 G r e e n 60.0 45.0 16.7 24.1 Blue 39.2 47.9 26.6 39.2 179.1 164.3 135.5 143.2 Blue Tank R e d 35.2 24.1 48.3 45.0 Y e l l o w 43.2 30.0 41.4 37.8 G r e e n 55.7 34.3 31.5 34.3 Blue 52.2 47.4 20.7 32.7 186.3 135.8 141.9 149.8 EY = 706.0 621.7 600.6 494.9 Y = 44.125 38.856 37.537 30.931 A N O V A T A B L E Source df MS P r o b a b i l i t y .025 < p < .05 .05 < p < .10 p « .001 2423.2 48 T A B L E XIII. M e a n selection intensities and interact ion effects for c o l o r s of food with various co lors of tag. Feed ing experiment II, U n i v e r s i t y of B . C . , Ju ly 23 to August 3, 1973. N u m b e r s in parentheses in T a b l e XIIIB are correc t ion factors , , as explained in Table VII. . Because the selection intensities of feeding exper iment II were based on feedings having different numbers of servings , a greater amount of r a n d o m e r r o r was introduced when calculat ing the a r c s in of the per cent selection intensity than o c c u r r e d in s i m i l a r ca lculat ions for feeding experiment I. The values in Table XIIIB have been c o r r e c t e d for this r a n d o m e r r o r as wel l as for the food co lor effect; so they are not equal to the values in Table XIIIA minus the c o r r e c t i o n fac tor . Uncorrec ted for Food C o l o r Effect F o o d C o l o r T a g C o l o r R e d Y e l l o w G r e e n Blue R e d Y e l l o w G r e e n Blue 37.28 39.65 56.32 43.25 33.28 19.28 47.70 55.18 40.32 56.80 26.55 26.48 43.48 39.38 14.60 26.28 176.50 155.44 150.15 123.74 C o r r e c t e d for Food Color Effect T a g C o l o r R e d (+6.26) . F o o d Y e l l o w (+1.0) C o l o r G r e e n (-.33) Blue (-6.93) R e d Y e l l o w G r e e n Blue 30.29 32.48 51.63 37.06 31.55 17.37 48.27 54.25 39.92 56.22 28.45 26.88 49.68 45.40 23.10 33.28 151.46 151.44 151.47 151.46 Interaction Effects T a g Color R e d F o o d Y e l l o w C o l o r G r e e n Blue R e d Y e l l o w G r e e n Blue - 7.57 - 5.38 +13.77 - .80 - 6.31 -20.49 +10.41 +16.39 + 2.06 +18.36 - 9.41 -10.98 +11.82 + 7.54 -14.76 - 4.58 EY + .02 0.00 + .03 + .02 S | Y | 27.52 53.60 40.81 38.70 27.76 51.77 48.35 32.75 160.63 due to being fed wi th different co lo r s of tag for the r e d - c o l o r e d food. The r e d egg played the greatest ro le i n setting the o v e r a l l preference for the "packet" of food. The other egg modif ied this l e v e l of preference a l i t t l e up or down. On the other hand, the preference l e v e l for the ye l low food was highly influenced by the co lo r of the other egg i n the packet. There was Little difference i n the sums of the absolute values of the in te rac t ion effects for the green eggs and the blue eggs, both of them l y i n g i n a pos i t ion in termedia te to those for the ye l low eggs and r e d eggs. These sums of absolute values of in te rac t ion effects are r e a l l y a measure of the dominance of one co lo r over others in setting the o v e r a l l preference l e v e l for a packet of food; the Lower the sum, the greater i s the dominance of that food co lo r in setting the l eve l of preference. (The in te rac t ion effects are not measures re la t ive to the strength of the food color , but are absolute measures of the amount that the tag co lo r adds or subtracts f r o m the preference l e v e l of the va r ious co lo r s of food.) It i s perhaps signif icant that the in te rac t ion effect for R R (the r e d -co lo red tag wi th the r e d - c o l o r e d food) is the most negative of any in i ts co lumn or row. Eviden t ly , the effect of g iv ing a double dose of r e d is not as "favorable 1 as making one egg r ed and the other egg another co lo r - - any other co lo r . A g a i n , d i v e r s i t y of co lo r i s shown to be impor tan t i n e levat ing the se lec t ion in tens i ty . The effect of the co lor of tank on the se lect ion of va r ious co lo r s of food i s not eas i ly ident i f ied. Al though a camouflaging effect could be in te rpre ted for the r e d and green tanks, the effect of s i m i l a r c o l o r e d food i n the ye l low and blue tanks was to increase their se lect ion in tens i t ies to the m a x i m u m . The effect of having the tank color and the food co lor wide ly separated in the spec t rum was to s l ight ly decrease the se lec t ion intensi ty rather than to inc rease it (see Table X V ) . Look ing at the data (Table X I V ) , it is diff icul t to see any-thing but a random dis t r ibu t ion of effects. Perhaps that i s how we should v iew these resu l t s since p was between .0 5 and .10. 50 T A B L E X I V . M e a n selection intensities and interact ion effects for c o l o r s of food in various co lors of tank. Feed ing exper iment II, U n i v e r s i t y of B . C . , July 23 to August 3, 1973. N u m b e r s in parentheses in Table X I V B are c o r r e c t i o n factors , as explained in Table VII. The values in Table X I V B have been c o r r e c t e d for the r a n d o m e r r o r introduced in calculating the select ion intensit ies , as explained in Table XIII. U n c o r r e c t e d for F o o d Color Effect Tank Co lor R e d F o o d C o l o r Y e l l o w G r e e n Blue R e d Y e l l o w G r e e n Blue 34.52 50.62 44.78 46.58 33.38 47.02 41.08 33.95 49.20 31.60 33.88 35.48 31.80 18.68 35.80 37.45 176.50 155.43 150.16 123.73 605.82 Y = 37.86 C o r r e c t e d for Food Color Effect Tank Co lor R e d (+6.26) F o o d Y e l l o w (+0.99) C o l o r G r e e n (-0-32) Blue (-6.93) R e d Y e l l o w G r e e n Blue 28.90 45.24 37.50 39.82 33.03 46.91 39.07 32.46 50.16 32.80 33.18 35.30 39.37 26.49 41.71 43.88 151.46 151.45 151.46 151.45 Interaction Effects Tank Co lor R e d Y e l l o w G r e e n Blue R e d - 8.96 + 7.38 - .36 + 1.96 Y e l l o w 4.83 9.05 1.21 5.40 G r e e n +12.30 - 5.06 - 4.68 - 2.56 Blue + 1.51 -11.37 + 3.85 + 6.02 ELIL 27.60 32.86 10.10 15.94 EY E|Y| + .02 18.66 + .03 20.49 0.00 24.60 + .01 22.75 86.50 T A B L E X V . Effect of wide spec t ra l separat ion of food co lo r and tank co lo r on food co lo r x tank co lo r in terac t ion effects of feeding 1 exper iment II, U n i v e r s i t y of B . C . , Ju ly 23 to August 3, 1973. Note: Wide spec t ra l separat ion is defined as combinat ions where one of the c o l o r s is r ed and the other i s green or blue (RG, R B , GR, BR) or the combinat ion of ye l low and blue ( Y B or B Y ) . These co lo r s w i l l be r e f e r r e d to as "complementary" , although they may not be true psychophys ica l complements . A l l other combina-t ions are s imply designated "non-complementa ry" . Complementary C o l o r s N o n - C o m p l e m e n t a r y C o l o r s Combinat ion Interact ion Effect Combina t ion In terac t ion Effect GR B R B Y R G R B Y B - .36 + 1.96 - 5.40 +12.30 + 1.51 -11.37 R R Y R R Y Y Y G Y Y G G G B G G B B B - 8.96 + 7.38 - 4.83 + 9.05 + 1.21 - 5.06 - 4.68 - 2.56 + 3.85 + 6.02 E Y - 1.36 + 1.42 Y .23 + .14 52 Why should there have been such a strong effect of tank co lo r i n exper iment I and supposedly none i n exper iment II? The most obvious reason i s that in experiment II the effect of feeding a given co lo r of egg i n a given co lor of tank was grea t ly di luted since this combinat ion was presented wi th four different co lo r s of tag egg. F o r each packet of food there would actual ly be two in teract ions between tank co lo r and co lo r of egg: that between the tank color and food egg, and the one between the tank co lo r and tag egg. F u r t h e r m o r e , in exper iment I each f i sh was fed twice in each co lo r of tank, making a total of eight r ep l i ca tes over which a tank co lo r in te rac t ion effect could make i t se l f evident. In exper iment II one f i sh was fed in each tank for each co lo r of tag, making a total of only four rep l i ca tes over which to p ick out the tank color in terac t ion effect. A l s o , the fo rmer exper imenta l des ign makes it poss ib le to separate out the effects of ind iv idua l differences among the f i sh f r o m the effects of the .color of tank. In exper iment II it i s quite possible that the in te rac t ion effect involv ing ind iv idua l differences could work against that involv ing tank color , and thus l e s sen the impor tance of the la t te r . Al though i t was not poss ib le to evaluate the combined effect of tank co lor and tag co lor on the se lect ion of food (because this second-order i n t e r -act ion was used as the. e r r o r t e r m in the A N O V A ) , it is poss ib le to ca tegor ize each observat ion on the bas is of the d i v e r s i t y of co lo r shown among food co lor , tank co lor , and tag c o l o r . Thus, for 24 of the observat ions the food co lo r i s different f r o m both the tank co lor and the tag co lor ; for 36 observat ions there are two co lors s i m i l a r ; and for four observat ions a l l three c o l o r s are the same. Having divided the data up in this manner (see Table X V I ) , one can see that both the coefficient of var iance (CV) and the mean se lec t ion intensi ty (Y) increase with d ive r s i t y of c o l o r . 53 T A B L E X V I . Select ion intensi t ies f r o m feeding exper iment II grouped accord ing to s i m i l a r i t y of co lo r s of tank, tag egg, and food egg. Feed ing exper iment II, U n i v e r s i t y of B . C . , Ju ly 23 to August 3, 1973. C V = coefficient of va r i ance . See text for explanation. No C o l o r s S i m i l a r Two C o l o r s S i m i l a r Three C o l o r s . S i m i l a r 54.7 18.4 63.4 49.6 36.9 90.0 33.2 26.6 15.4 24.3 45.0 45.0 0.0 49.0 16.7 60.0 0.0 48.4 35.2 32.7 26.6 33.2 32.0 45.0 42.1 56.8 0.0 62.0 42.1 54.9 39.2 45.0 48.6 32.6 47.9 39.2 37.8 22.8 42.1 0.0 43.2 41.4 55.7 50.8 24.1 26.6 34.3 60.0 48.3 45.0 41.4 24.1 37.8 22.8 39.2 45.0 37.8 34.3 52.2 47.4 20.7 35.2 30.0 31.5 EY 993.5 1319.1 110.6 n 24 36 4 Y 41.4 36.6 27.65 C V 45.34% 40.47% 32.49% 54 I N C I D E N T A L O B S E R V A T I O N S A . Effect of Background Co lo r on A c t i v i t y L e v e l of Trout In the course of c a r r y i n g out the feeding exper iments two inc identa l observat ions were made which could suggest a d i r ec t i on for further study. P r e l i m i n a r y tests on these ideas are desc r ibed below. _ A t ce r t a in t imes (e.g., when the stock tank was being c leaned or repaired) , the stock of trout was d i s t r ibu ted among four s m a l l tanks that were painted the same co lo r s (red, yel low, green and blue) as the exper imenta l tanks. Dur ing many casual observat ions, I not iced that the f i s h in the ye l low tank seemed to be much more act ive than those in the red, green or blue tanks. In an attempt to ver i fy this observat ion, the fol lowing exper iment was c a r r i e d out. 1. Methods Six f i sh chosen at r andom f r o m the stock were placed in each of the four exper imenta l tanks: red, yellow, g reen and blue. A n 8 m i l l i m e t e r super-8 movie c a m e r a was placed face down on the roof of the tank so that the lens " looked" through the hole through which the f i sh had p rev ious ly been fed. The f i e ld of v i s i o n included about f ive - s ix ths of the length of the tank and a l l of the width . Br ightness of i l l umina t i on was set to the same values as in feeding exper iment II (7 2 lux at the 32 c m depth). The f i sh were fed on the day of t ransfer and given 24 hours to sett le. They were then f i lmed in 15-second "shots" i n each of the four tanks. The order of f i lm ing was randomized for each set of four shots. A t least th i r ty minutes was left between sets of shots so that the ac t iv i ty l e v e l of the f i sh dur ing one shot would not be a " le f tover" f r o m the previous set of shots. F i v e sets of shootings were done on Ju ly 19. after which the f i sh had to be r emoved f r o m these tanks. La te r , on August 6, another s ix f i sh chosen at r andom f r o m the stock were put into each of the tanks and left for 24 hours . The f i l m i n g 55 procedure was repeated three sets of shots were taken - - and then the f i s h were t r ans fe r red f r o m one tank to the other as fo l lows: blue to green to ye l low to red to blue. They were left for another 24 hours and then two sets of shots' were taken. A measurement of ac t iv i ty l e v e l was obtained in the fo l lowing manner for each r o l l of f i l m . The projector was set up and the f i l m was shown against a w a l l . A l ine of black thread was held on the wa l l wi th pins so that it b i sec ted diagonal ly the projected p ic tu re . The number of t imes any f i sh swam " a c r o s s " the l ine was total led for each 15-second shot. A s some of the shots were s l igh t ly longer than 15 seconds, each total was adjusted to an exact t ime of twenty seconds by mul t ip ly ing the total by twenty d iv ided by the dura t ion of the shot i n seconds (measured with a stop-watch). Al together , ten shots were taken in each co lo r of tank wi th three different groups of f ish, and so the data were b locked as shown in Table X V I I , and analyzed wi th a two-way A N O V A (a = .0 5). 2. Resu l t s The co lor of tank had a marked effect on the ac t iv i ty . l eve l of the f i sh . The f i sh were most active in the ye l low tank and leas t act ive in the green tank, wi th intermediate l eve l s for the red and blue tanks. A s it was o r i g i n a l l y postulated that there would be a higher ac t iv i ty l e v e l in the ye l low tank, an a p r i o r i test of the ye l low treatment mean versus "the o thers" showed a signif icant difference ato , = .0 5. A n a p o s t e r i o r i test of the green treatment mean versus the red and blue means combined was a lso s ignif icant at the .0 5 l e v e l of probabi l i ty . Nei ther group differences nor the in terac t ion between the groups and the c o l o r s of tank were s t a t i s t i ca l ly s ignif icant . 3. D i s c u s s i o n This exper iment demonstrates that a ye l low background causes an 56 T A B L E X V I I . M e a s u r e s of activity levels of f ish in different c o l o r e d habitats . A N O V A table from, the analys i s of this data. E a c h value r e p r e s e n t s a measure of the activity leve l of a group of six f ish in a p a r t i c u l a r co lor of tank, and is approximate ly equal to the number of t imes any f i sh c r o s s e d f r o m one side of the tank to the other in a 20 second interva l . Data are divided into three groups of f i lmings ("shootings") using different f i sh for each group. Group Nunber 1 Red 10.13 16T2 5 6.52 4.94 11.76 49.60 8.54 12.82 12.35 33.71 16.25 13.16 29.41 C o l o r of Tank Y e l l o w G r e e n 19.23 8.86 9.26 11.39 13.92 62.66 8.97 11.11 23.08 43.16 15.00 10.13 25.13 2.53 5.41 9.00 0.00 10.53 27.47 8.86 7.41 9.76 26.03 10.13 1.33 11.46 Blue 6.90 13.33 12.00 8.86 6.49 47.58 13.41 6.33 10.53 30.27 13.75 17.50 31.25 EY 187.31 133.17 97.25 112.72 130.95 64.96 109.10 417.73 A N O V A T A B L E Source df MS P r o b a b i l i t y C o l o r of tanks 3 79 x < .001 Groups 2 26 .10 < x < .25 Groups x Co lor of Tanks 6 11 x >.7 5 E r r o r 28 17.25 57 increase in the l eve l of ac t iv i ty over that for r ed or blue, while a green back-ground causes a decrease . If the average speed of swimming ref lec ts the average metabolic rate of the f ish, then i t i s poss ib le that f i sh r a i s e d i n a ye l low tank are using up more energy in maintenance and channel l ing less into growth. The resu l t s of the exper iment seem to support the t rad i t iona l hatchery pol icy of painting f i sh ponds and tanks a green-blue (turquoise) c o l o r . However , Bu r rows (1969), i n his study of the influence of f inger l ing quali ty on the s u r v i v a l of adult salmon, as w e l l as noting that s ize of f i sh at t ime of re lease had a posit ive effect on the s u r v i v a l of adults, found that those f i s h that had been exe rc i s ed dur ing their r ea r ing had a higher rate of r e t u r n of adults to the hatchery than f i sh that had not been exe rc i s ed . He therefore sug-gested using ponds in which a high ve loc i ty of water could be maintained. Whether or not an adequate exe rc i se l e v e l could be maintained through adjusting the background co lor of the tank i s , perhaps, wor th inves t iga t ing . B . Adaptat ion of Skin of Trout to Background Colo r The second thing noticed was the high degree to which the trout were able to adapt the color of their sk in to that of the tank c o l o r . Th is adaptation consis ted not only in a change of the l ightness or darkness of the skin , but in a convincing matching of the background hue. It was planned to leave severa l f i sh in each color of tank for two days, then photograph them and analyze the t ransparencies according to a spectrophotometr ic technique. Th i s I started to o do, but before analyzing the photographs, I became aware of a basic flaw in this technique which rendered any further analys is of no value. Therefore , what I have to repor t w i l l be rather subjective. Al though the photograph of the f i sh ' s sk in might appear to the human eye to accura te ly por t ray the skin color , in r ea l i ty only three p r i m a r y hues are coming f r o m it; what is "seen" is a resul t of the capacity of the b r a i n to "blend" those hues to f o r m a percept ion of a hue different f r o m any one of i ts components. Of course , a spectrophotometer would r eg i s t e r only the vary ing amounts of these three p r i m a r y hues. Wi th perfect exposure one might be able to calculate the dominant wavelengths 58 Look ing at a f i sh f r o m the side, I have somewhat a r b i t r a r i l y d iv ided the surface into three areas, A , B, and C (see F i g u r e 6). The a rea that is most great ly affected by the color of tank is probably the d o r s a l s t r ip (Area A ) . Th i s i s where the greatest range of hue and br ightness i s found, the reg ion of sk in that provides the c loses t match to the hue and br ightness of the tank. F o r the s ize of f i sh that I was using, the width of the d o r s a l s t r ip was 2 - 3 c m (1.0 - 1.5 c m on either side of the do r sa l fin), and it r a n f r o m the tip of the snout to the peduncle reg ion . A r e a B cons is t s of the res t of the d o r s a l half of the f i sh to the l a t e r a l l ine , and is usual ly a zone of fading f r o m the deep co lo r of the d o r s a l s t r i p to the whi t i sh color of the be l l y reg ion . Somet imes the l a t e r a l l ine showed a co lo r dif fer ing f r o m that of the surrounding skin, and i f so, this is noted in Table X V I I I . The remainder of the side is ca l l ed a rea C, and it i s the l ightest part of the f i sh . The t e rms used to descr ibe c o l o r s of sk in and s ize of spots in Table XVII I are subjective and good only for compara t ive purposes . No measu re -ments were taken; however, compar i sons between f i sh f r o m different tanks were checked wi th the aid of the co lo r s l ides . On June 12, 1973, a l l of the f i sh in the stock tank were cold-branded on the d o r s a l s t r ip near the d o r s a l f in on the left side of the body. (A black footnote 8 continued of l ight r e f l ec ted f r o m the f i sh ' s sk in if: ( 1 ) the f i l m accura te ly reproduced the co lo r of the skin; and (2) one knew a fo rmula for ca lcula t ing the "psycho-l o g i c a l " hue resul t ing f r o m the blending of the f i l m ' s p r i m a r y c o l o r s . M y p ic tures tended to be rather underexposed, and I had no way of ascer ta in ing the sens i t iv i ty of the f i l m dyes to c o r r e c t for underexposure . In short, I was not "set up" for the complexi ty and sophis t ica t ion of technique r e q u i r e d for such an exper iment . F i g u r e 6. Regions of sk in for sk in Incidental observat ions , explanation.) color observat ions . (See text for T A B L E XVIII. D e s c r i p t i o n of skin c o l o r of f i s h f r o m different c o l o r e d tanks. Incidental observations. D e s c r i p t i o n o f S k i n C o l o r of Tank Region A Region B Region C Red Darkness of skin: medium Region of fading co l o r to l a t e r a l Silver-white with a faint g r e e n i s h Hue; brown-green l i n e . L a t e r a l line is a pinkish tinge. c o l o r . L a r g e black spots are conspicuous. L a r g e black spots. Black spots smaller, but extending to mid-ventral line. Y e l l o w Darkness of skin: light Hue: g r e e n i s h gold S m a l l black spots. Region of fading gold color to l a t e r a l S ilver-white with perhaps a slight l i n e . L a t e r a l line has a pinkish c o l o r . Black spots have almost faded out by l a t e r a l line. y e l l o w i s h tinge. Only the slightest vestige of black spots beneath the l a t e r a l line. G reen Darkness of skin: dark Hue: dark g r e e n L a r g e black spots partly camou-flaged by dark green background. Region of fading green color to l a t e r a l line. B l a c k spots d e c r e a s i n g i n size. B e l l y retains slight b l u i s h - g r e e n tint on white skin. Black spots continue to decrease in size and frequency to halfway between l a t e r a l line and mid-ventral r e g i o n where they stop--bottom half of belly without spots. Blue D a r k n e s s of skin: very dark Hue: deep green-blue L a r g e black spots d i f f i c u l t to pick out against dark green-blue back-ground. Reg i o n of fading green-blue c o l o r to l a t e r a l line. B l a c k spots d e c r e a s i n g in size. B e l l y retains slight green-blue tint on white skin. B l a c k spots continue to decrease in size and frequency to half-way between l a t e r a l line and mid-ventral region, where they stop--bottom half of belly without spots. 61 brand m a r k develops 2 to 3 days after branding). Two months la ter some of the brands had started to fade out, and those on f i sh f r o m the blue and green tanks were espec ia l ly dif f icul t to r ead . Some of the brands were ac tual ly i l l eg ib l e and could not be read even on c lose inspect ion . However , upon t r ans fe r r ing the f i sh to the ye l low tank (and to a l e s s e r extent the red tank), the brands could again be read, although they were not as dark as i n i t i a l l y . Th is I was able to do for about another month, when a l l of the brands were faded, even those on f i sh kept i n ye l low tanks. DISCUSSION A . Background L i t e r a t u r e 1. Co lo r v i s i o n in f i sh . E a r l y invest igators r e l i e d on secondary responses to test for color v i s i o n of f i sh . Graber (1884, 1885 9) and Hess (1910, 1911, 1913, 1914 9) ob-served what sect ion of tank a f i sh would s w i m to when var ious parts of the tank were i l l umina ted wi th l ight of different wavelengths. White (1919) and Hine l ine (1927) t ra ined mudminnows, and B r o w n (1937) t ra ined bass to make a posi t ive assoc ia t ion of food wi th a pa r t i cu la r co lo r and a negative assoc ia t ion wi th other c o l o r s . They then tested the f i sh ' s abi l i ty to d i s c r i m i n a t e between the co lo r s in the complete absence of the condit ioning s t i m u l i by observ ing whether a posi t ive (swimming towards) or negative ( swimming away from) response was e l i c i t ed upon presentat ion of each c o l o r . The fo rmer type of exper iment has been ca l l ed by Warner (1931) the "preference method", and the lat ter , the " learn ing method". Both types of exper iment have been plagued by inadequate attempts to cont ro l for intensi ty differences between test s t i m u l i that are supposed to d i f f e r - i n wavelength-only. At tempts to Quoted in Brown, 1937. 62 equate the l uminos i ty of the sources by photometr ic measurement are i r r e l e -vant as the sens i t iv i ty curve for the spectrophotometer is not l i k e l y to match the curve of r e t i na l sens i t iv i ty for the f i sh . Where these exper iments have had a measure of success is when they tested wi th a wide range of hue in ten-s i t ies and then fol lowed up wi th d i s c r i m i n a t i o n tests between the co lo r s and a se r i e s of greys of va ry ing in tens i t ies . A th i rd l ine of evidence involves modif icat ions of sk in pigmentat ion of ce r ta in f ishes upon being exposed to different background c o l o r s . The s t r i k i n g ab i l i ty of flounders to m i m i c the shade, color and pattern of the i r background has been demonstrated by Mas t (1 916). He presents a number of co lo r photo-graphs which attest to his descr ip t ions of co lo r changes in the f i sh . That co lo r change was mediated through the eyes, and not as a d i r ec t effect of l ight on the chromatophores , was shown in two ways . In ce r t a in exper iments one or both eyes were exc i sed . If only one eye was removed, there would r e su l t a t empora ry interference wi th the adaptive p rocesses i n the skin; i f both eyes were removed, a permanent cessa t ion of these processes occur red , such that changes in background had no apparent effect on the appear-ance of the skin . In other exper iments the f i sh was found res t ing , or was placed at the d iv id ing l ine in a tank painted half b lack and half whi te . It was found to take whichever shade the head (eyes) of the f i sh was res t ing over . If the f i sh was placed lengthwise so that one eye was on ei ther side of the d iv id ing l ine, it adapted to an intermediate shade. S i m i l a r r e su l t s were obtained when the hue was v a r i e d . Once it was ce r ta in that these responses were mediated through the eye, they proved to be good evidence for the existence of co lo r v i s i o n because, if the f i sh was able to perceive differences in hue only as differences in b r igh t -ness, there was no explanation as to how it was able to match the hue of its sk in to that of the environment. A s the ea r ly exper imenters ref ined their techniques they con t ro l l ed the 63 brightness factor more effectively, and soon evidence was mounting for co lo r v i s i o n i n a number of species of f i sh . A t the same t ime, more d i r ec t evidence was being sought by focussing on the r e t ina . It had been known for a long t ime that the re t ina had two types of receptor ce l l s : rods and cones. In 1802, Thomas Young had suggested that co lor v i s i o n depended upon the presence of at least three receptor substances, each m a x i m a l l y sensi t ive to different regions of the spect rum. A n d in 1866 M a x Schultze had stated that the cone is the receptor for photopic v i s i o n while the r o d i s the receptor for scotopic v i s ion ; and fur thermore, that the cone alone is respons ib le for co lo r v i s ion , since i n d i m light co lo r s are not v i s ib l e - - everything is seen i n shades of g rey . But for a long t ime h i s tophys io log ica l techniques were not capable of an examinat ion of the pigments f r o m ind iv idua l cones, nor a de te rmina t ion of their nervous ac t iv i ty . In the late 1950s and ea r ly 1960s the technique of m i c r o spectrophoto -met ry al lowed var ious worke r s to determine the absorpt ion spec t ra of the pigments in ind iv idua l cone c e l l s . M a r k s (1965) de termined the spec t ra l sens i t iv i ty curves for 113 goldf ish cones; eighteen had their wavelength of m a x i m u m absorpt ion near 450 my, 66 near 530 my, 24 near 630 my, 2 near 570 mp., and 3 near 480 m|ji. The two near 570 my were not f r o m spec t ra of single cones, but f r o m the composi te spect ra of touching twin cones; the three near 480 my were probably the r e su l t of a greater tendency for va r i a t i on in wavelength m a x i m u m (X max) i n the blue due to more prominent photoproducts, or because misa l ignment caused more chromat ic motion of the beam in the b lue . Taking this into considerat ion, these data indicate there are just three types of cones in the goldfish re t ina wi th \ max near 450 my, 530 my, and 630 my. Baker and Rushton (1965b) have used the method of r e t i na l dens i tomet ry to determine the act ion spect ra for the red cone pigment (erythrolabe) of the n o r m a l human eye, and have compared it to that found for the r ed - sens i t i ve pigment in the deuteranopic eye. The i r method consis ted of bleaching the red 64 pigment wi th deep red l ight and measur ing the resul tant change i n t r a n s m i s s i v i t y at the r ed end of the spec t rum. The measurement was made by shining a thin penc i l of l ight through the center of the pupi l to the fovea of the re t ina , where i t passed through the pigmented l aye r s , ref lected f r o m the choro id and t r a v e l l e d along a s l ight ly different path back through the pupil to a photosensit ive c e l l , wh ich measured i ts intensi ty. Th i s intensi ty was compared to that coming f r o m a very deep r e d standard l ight (700 mjj.) s i m i l a r l y ref lec ted through the r e t ina l l a y e r s . The dominant wavelength of this lat ter l ight was chosen so that a l l r e t ina l pigments would be t ransparent to i t . By inse r t ing an op t ica l wedge into the path of the light, in tensi t ies f r o m the test and con t ro l beams could be balanced, and the distance the wedge was inser ted would give a measure of the change in t r a n s m i s s i v i t y at the wavelength of the test l ight . The wavelength of the test l ight was v a r i e d to determine the act ion spec t rum for that pigment. Unfortunately, the dens i tometr ic method r equ i r e s accurate al ignment of the beams of l ight being shone on the re t ina . This was attained for human sub-jects by use of a dental i m p r e s s i o n and brow rest, and by request ing them to fixate on a given spot. F o r this reason r e t i na l dens i tometry has l i t t l e potential for measur ing photopigments of species other than man, but it has supplied an opportunity to check the accuracy of other techniques which are more appropr ia te . The e lec t rore t inographic methods developed by Svae t ich in and col leagues (Svaetichin and M a c N i c h o l , 1958) do not have these drawbacks, and can be used on exc i sed ret inae. By inse r t ing a mic roe lec t rode into the r e t ina ( f rom b r e a m or perch) to the approximate depth of the cone inner segments, they were able to r e c o r d what they thought was the res t ing potential of the cones (about - 4 0 mv). This was later found to be one of the so - ca l l ed "S potent ia l s" f r o m a deeper l a y e r . When l ight was shone onto the re t ina a change in potential o c c u r r e d which las ted throughout the t ime when the l ight was on. A c t u a l l y two separate effects could be isola ted depending upon the depth at which the e lec t rodes were placed. The f i r s t effect, c a l l ed the " luminos i ty r e sponse" was observed when the electrode was posi t ioned near the layer of giant ho r i zon ta l c e l l s . It c o n -65 s is ted of an inc reased negative potential , r ega rd le s s of the wavelength of the s t imulat ing l ight . The "ch romat i c i ty response" i s perhaps the more in te res t ing of the two wi th respect to color v i s i o n theory. When the electrode was located twenty to th i r ty mic rons deeper wi th in the inner nuclear and p l e x i f o r m Layer a response o c c u r r e d cha rac t e r i zed by having two m a x i m a of opposite po l a r i t y . These m a x i m a were found either i n the ye l low and blue regions of the spec t rum or i n the r e d and green reg ions . Between the max ima there was a neu t ra l point where i l l u m i n a t i o n by l ight of that pa r t i cu l a r wavelength caused no increase or decrease in the potent ial . Gene ra l ly the ch roma t i c i t y responses were negative for those wavelengths shorter than the neut ra l point and posi t ive for those wavelengths longer. Fur the r , by compar ing r i s e t imes and la tencies for the generated potentials f r o m short wavelength s t imula t ion ve r sus long wavelength s t imula t ion it was found that the ch roma t i c i t y response had two components g iv ing potentials of opposite s ign which were somehow subtracted f r o m each other. When the elctrode was posi t ioned s t i l l deeper i n the ganglion l ayer (site of the actual neura l t r an smi s s ions to the brain) another f requency-dependent response was r eco rded . If the s t imula t ion l ight was of short wave-length the ganglion c e l l s produced a burs t of impulses when the l ight was turned on. If the s t imula t ing l ight was of long wavelength, there was a suppress ion of ac t iv i ty during i l l u m i n a t i o n and a burs t of impulses when the l ight was turned off. It seemed that the gangl ia r e c e i v e d connections f r o m groups of receptors having sens i t iv i t i es in a number of regions of the spec t rum. But what i s known of the e l e c t r i c a l ac t iv i ty of the receptors themse lves? Tomi t a ; e_t al_., (1 967) desc r ibe a technique whereby a mic roe l ec t rode is s lowly advanced through the r e t i na l l aye r s of a ca rp unti l a negative potential is detected. If the electrode is wi th in the receptor l ayer and no a rea effect is apparent the spec t ra l response curves are then obtained and these are assumed to come f r o m a single cone. Once these measurements have been taken, the mic roe lec t rode is further advanced into the re t ina unt i l an S-potent ial is r eco rded . This is taken as further evidence that what was in fact measured 66 was the receptor potential . F r o m the analysis of 142 ca rp records they found three groups of cones: 74 per cent had their peak response in the red (6ll+_23 my), 10 per cent peaked i n the green (529 +_14 my), and 16 per cent peaked in the blue (462 + 15 my). It i s not known whether the respec t ive percentages ref lected the actual population d i s t r ibu t ion for the three types of cone or whether they were an ar t i fact of the sampl ing procedure . Kobayash i and A l i (1971) used a s i m i l a r e lec t ro re t inographic technique to measure the photopic spec t ra l sens i t iv i ty for brook trout (Salvel in is fontinalis) a c lose re la t ive of ra inbow trout . They did not isola te ind iv idua l cones for measurement, but r ecorded the summated response of the re t ina to wavelengths between 400 and 700 my. Never the less , the spec t ra l sens i t iv i ty curve showed three dis t inct max ima corresponding to the wavelengths of m a x i m u m absorpt ion for each of the three types of cones . These m a x i m a o c c u r r e d at 425 my, 545 my, and 595 mj,. 2. Trans mis s iv i ty of l ight i n water . The eyes of many f i sh are capable of co lo r d i s c r i m i n a t i o n , but are there c o l o r s to d i s c r i m i n a t e where the f i sh l i ve? One of the arguments Hess used against co lor v i s i o n i n f i sh was that monochromat ic l ight, e spec i a l l y of the longer wavelengths, penetrates water to such a shal low depth that co lo r v i s i o n would be of l i t t l e value to fish, should they possess i t . He pe r fo rmed an exper iment where he lowered different co lo red papers protected in c e l l u l o i d into the water, and observed them f r o m the surface. A t a depth of 5.6 meters none of the co lo r s could be recognized . Warner poses the c r i t i c a l question as -to whether this disappearance of c o l o r s was caused by the depth of water through which the l ight passed or the quantity of water . Of course , Hess should have observed the papers wi th his eye submerged to the l e v e l of the 67 Cousteau and Dumas (1953, p. 255) descr ibe a study where they photo-graphed var ious co lo red plates at depths to 120 feet. The p ic tures show that at forty feet the r ed plate appeared v i r t u a l l y black, and by 120 feet ye l low had begun "to t u r n " to green. Behan, _et al_.}(1972) take issue wi th Cousteau 's r e su l t s , c l a i m i n g that photographic f i l m does not reproduce what the human eye is capable of d i sce rn ing . They sent six teams of two d i v e r s down to th i r ty , s ix ty and ninety feet i n the ocean wi th diagnost ic co lo r b l indness plates (for testing r ed -g reen and ye l low-b lue co lo r b l indness) . There they were to attempt to identify the numbers wr i t t en on the plates, thus test ing their ab i l i ty to see co lo r s at these depths. A ch i - square test on the observed resu l t s ve rsus those expected on the bas i s of no loss of co lor for each of the three depths showed a probabi l i ty for the observed deviat ions of > 0.6. However , neither of these studies comple te ly c l e a r s up the matter . Cousteau d id not attempt to put any of his observat ions into quantitative t e rms , and he did not mention any difference in sens i t iv i ty between the eyes of the photographer and the f i l m which he was us ing . On the other hand, Behan fa i led to state what effect the na r rowing and shifting of bandwidth of l ight wi th depth has on the re f l ec t iv i ty of the test plate c o l o r s . In "whi te" l ight f r o m the sun, the var ious parts of the test plates (the numera l ve r sus the background) may ref lec t to the eye equal in tens i t ies of l ight, but i n the green-blue l ight of deep water the r e f l ec t iv i t i e s may be quite different. Hence, a c o l o r - b l i n d subject could conceivably d i s c r i m i n a t e the numera l on a co lo r -b l i ndnes s test plate by br ightness contrast at great depth when he would be unable to do so at the surface. However , Behan desc r ibes a further test which suggests the resu l t s of his study-do indicate a true co lo r percept ion ra ther than br ightness percept ion. P i c t u r e s of the co lo r bl indness plates were taken with an underwater c a m e r a in the a i r , at a depth of three feet in a swimming pool, and i n the ocean at fifty feet. Under a l l the conditions of v iewing the photographer was able to see both the co lo r s and numera l s on a l l of the plates. The p ic tures taken in the a i r 68 reproduced a l l of the co lo r s and f igures on the plates . However , those taken i n the pool and ocean fa i l ed to r e c o r d a l l of the f igures that the photographer could see on the plates. F r o m casual questionning of S C U B A - d i v i n g fr iends, I have found genera l agreement that a fu l l range of c o l o r s can be d i s c r i m i n a t e d at fifty to s ixty feet. 3.. Relevance to t r o l l i n g and feeding exper iments . The consensus of the l i t e ra ture i s c l ea r - - f i sh have co lo r v i s i o n . A l i (1961) states that, "Teleos ts as a group, wi th the poss ib le exception of deep sea fo rms wi th pure rod ret inae, are be l ieved to be able to perce ive c o l o r s . " (See also Wal l s , 1963, p. 488). He found that cones of the A t l a n t i c sa lmon (Salmo salar) were s tar t ing to l ight adapt in white l ight of intensi ty 10 foot candles (ft-c). A s s u m i n g the same amount of un i formi ty wi th respect to color v i s i o n th reshold i n the Salmo genus as there is in the Oncorhynchus genus ( A l i , 1959). we can expect the cones of Salmo g a i r d n e r i to begin l ight -2 adaptation at no greater than 10 ft-c i l l u m i n a t i o n . Th i s is a lmos t the same intensi ty of l ight r equ i red for human co lo r v i s i o n - - .003 ft-c (Geldard, 1972). In l ight of Behan 's work (1972), this v i r t u a l l y ensures that the f i sh were seeing photopical ly at a l l t imes during the t r o l l i n g and feeding exper imen t s . A rough check on the depths at which the lures were running indicated a m a x i m u m depth of fifteen feet for t r o l l i n g exper iments I and II, and no greater than fifty feet for t r o l l i n g exper iment III. In the tanks where the feeding exper iments were done the greatest decay in l ight l e v e l wi th depth o c c u r r e d in the blue tank, and even here the darkest part of the tank measured 38 lux (at 3.5 f t -c ) . This is more than enough l ight to ini t ia te cone v i s i o n . B . Interpretat ion of Resu l t s The s ignif icance of co lo r i n the feeding exper iments i s attested to else-69 where in the l i t e ra tu re . Ginetz and L a r k i n (1973) descr ibe a feeding e x p e r i -ment in which they fed seven co lo r s of dyed eggs to five ra inbow trout in 7.3 meter long troughs painted a pale greenish-b lue . T h i r t y - f i v e eggs of each of two co lo r s were placed in the inlet of the trough, and those that were not eaten as they passed down the trough were co l lec ted in a t rap at the outlet and counted. F o u r rep l ica tes were done (one trough per day) for each poss ib le combinat ion of two co lo r s of egg. The average number of eggs eaten per t r i a l over the entire exper iment were as fol lows (a = .0 5): blue, 29.54; red, 27.0 5; black, 26.78; orange, 23.84; brown, 23.46; yel low, 22.54; and green, 20.36. In a second exper iment Ginetz and L a r k i n matched the background co lo r of the tank wi th one of the two co lo r s of eggs presented to the f i sh . Th i s was done for a l l poss ib le combinat ions of red, yellow, blue and b lack only. Th is t ime the r ed came out ahead of the blue wi th the fol lowing average numbers of eggs eaten per t r i a l : red, 21.94; yel low, 21.66; blue, 20.38; and black, 15.69. Al though preference was measured in a different manner in Gine tz ' s exper iments than in mine (total number of eggs eaten ra ther than order of selection), the order of preference for red, ye l low and green eggs i n r e l a t i on to each other, in h is f i r s t exper iment is the same as in my feeding exper iments I and II. Since, among four exper iments , the rank of blue i n the preference h i e r a r c h y assumed f i r s t , second, t h i r d and fourth posi t ions , it must be con -cluded that the l e v e l of preference for blue i s highly var iab le , perhaps grea t ly dependent upon the conditions under which it is being presented to the f i sh . A unique reac t ion of f i sh to red has been mentioned by other authors. In his r ev i ew of the color v i s i on l i t e ra ture , Warner (1931) ci tes a study where f i sh reacted negatively and v io len t ly when placed i n a r eg ion of a tank i l l umina ted wi th r ed l ight . He also c i tes another study where f i sh gathered in a reg ion of r ed i l l umina t ion i n preference to darkness . B r o w n (1937) found that bass were able to d i s t inguish shades of r ed f r o m a ser ies of shades of grey wi th a greater accuracy than for other c o l o r s . They were also p a r t i c u l a r l y r a p i d in lea rn ing to associate a r e w a r d or punishment wi th r ed . Wolf and Wales (19 53) 70 "fed" painted corks to ra inbow trout and brook trout. "When green, blue, yel low, brown, b lack or white c o r k s were tossed onto the surface of the ponds, the trout demonstrated a nea r ly un i fo rm lack of in teres t . On the other hand, when a red cork was thrown among those of other co lo r s , the f i sh l i t e r a l l y "bo i l ed" around it, s t r ik ing i t so v igo rous ly that i t would be knocked out of the water and soon d r iven to shore ." It is apparent that, whether the reac t ion toward i t i s posi t ive or negative, the attention paid to the co lo r r ed by f i sh i s ve ry high. If ra inbow trout show such a definite h i e r a r c h y of co lo r preferences as indicated in the feeding exper iments , the obvious question is why these p r e -ferences were not evident in the t r o l l i n g exper iments . A number of factors may help expla in th is . F i r s t , it i s poss ib le that the f i sh "shift their at tention" f r o m one set of cues to another in a manner s i m i l a r to that proposed by Dawkins (1969). While they may at one t ime be attending to the co lo r s of objects, they may at another t ime attend to their shape or pos i t ion . It i s part of Dawkins ' thesis that only one sys tem of cues can be attended to at a t ime . Wi th this in mind, one may consider the resu l t s of exper iments pe r fo rmed by H o r i o in 1 9 3 8 . ^ H o r i o was in teres ted in the re la t ive strength of f o r m and co lor s t i m u l i to f i sh . To this purpose he t ra ined ca rp posi t ive to a r ed disc and negative to a blue one. They were able to l e a r n this assoc ia t ion more eas i ly than a s i m i l a r assoc ia t ion using a white t r iangle (positive) and a white square (negative). He then found a pair of co lo rs as dif f icul t to t e l l apart as the white t r iangle and white square: blue and violet . "T ra ined posi t ive to a v io le t d isc ve r sus a blue one and to a white t r iangle versus a white square, then offered a v iole t square ve r sus a blue t r iangle , the f i sh went to the posi t ive co lo r ra ther than to the posi t ive f o r m . " (from Wal l s , 1963). C o l o r thus seems to be more important D e s c r i b e d in Wal l s , 1963. 71 for ca rp than f o r m regard ing i ts attention value. F o r trout this may be r e v e r s e d . E v e n if not, i t i s l i k e l y that the actions exhibited by the l u r e s commanded a greater amount of attention than the static shapes used by H o r i o . If so, the f i sh in the t r o l l i n g exper iment may have been p r i m a r i l y attending to the actions of the lu res to the neglect of their c o l o r s . Since in the feeding exper iments the "ac t ions" of the eggs dr i f t ing down through the water co lumn were at least s i m i l a r , the f i sh may have shifted their attention to the sys t em of color cues, and made thei r choices on that ba s i s . This is poss ib ly why co lor was signif icant in the one exper iment and not i n the other. Fu r the r , the t ighter con t ro l of extraneous factors poss ib le wi th a l abora to ry exper iment over a f i e ld study probably played a ro le i n i l l umina t ing what may be a weaker effect. A l s o , the two-pronged effect of using a domest ic stock of trout in a labora tory experiment , whi le using w i l d t rout in a f i e ld study, may have been respons ib le for some of the differences i n resu l t s (for example, their previous feeding h i s to r i e s were different). The imp l i ca t ions of the r e su l t s f r o m the t r o l l i n g and feeding e x p e r i -ments wi th respect to the design of an effective lu re probably only apply to ra inbow trout, and may even apply only to the stocks of f i sh studied. It was found that: 1. A m o n g t r o l l i n g lures of approximate ly equal s ize, the act ion of the lu re was found to be more important than the co lo r in de te rmin ing i ts a t t ract iveness to ra inbow trout, as measured by the number of s t r ikes made on a pa r t i cu la r lu re . It is not known whether the c r i t i c a l aspects of the act ion are quantitative or qual i ta t ive; both appear to co r re l a t e wi th the measured at t ract iveness of the l u r e s . 2. Al though the color of lu re was not s t a t i s t i ca l ly s ignif icant in the t r o l l i n g exper iments , r ed lu res caught the greatest number of f i sh in a l l three exper iments . The co lor pattern was found not to cause s ignif icant dif ferences . No se lect ion for s ize or sex of f i s h o c c u r r e d on the bas i s of ei ther co lo r or action of l u r e . L u r e s t r o l l e d wi th dodgers se lected for l a rge r f i sh than those t r o l l e d without dodgers. There was no difference in the sex ra t ios of f i s h caught by lu res t r o l l e d wi th dodgers f r o m those t r o l l e d without dodgers . The t ime of day of f ishing was found to be signif icant i n one exper iment and not signif icant in the two others . The reason for this difference i s not known. Co lo r of food was found to play a significant r o l e in the order of s e l e c -t ion of dyed trout eggs fed to ra inbow trout . The highest preference was for r ed followed by yel low, then green . The preference for blue eggs was var iab le ; i t was probably more dependent on the condit ions of feeding than the preferences for the other c o l o r s . Al though a l l the f i s h p r e f e r r ed r e d eggs to other c o l o r s of eggs, the greatest range of preference intensi ty among the f i sh was found for r ed eggs. The preferences exhibited by the f i sh were not dependent on their l e v e l of hunger, as indicated by the amount of food in the i r s tomachs. Both the co lo r of background and the co lo r of the tag egg were found to affect the order of se lec t ion of dyed trout eggs. D i v e r s i t y of co lo r between the tag egg and the food egg inc reased the preference intensi ty for a packet of food. It i s not ce r t a in how the effect of background co lo r exerts its influence, but there is evidence that the preference intensity is higher if one, and espec ia l ly i f both, the tag egg and the food egg contrast with the co lo r of background. There were signif icant differences among the f i sh in their order of se lec t ion of different co lo red food. These differences were not as 73 important as those caused by changing the background co lo r of the tank. Indiv idual f i s h modify their order of se lec t ion of co lo red food in different ways in different co lo red tanks. This study has made only a start on the complex question of what makes a good l u r e . The act ion of the lu re was found to be of p r i m e impor tance . The t r o l l i n g exper iments , i n l ight of the resu l t s of the feeding exper iments , i n d i -cate the importance of color , at least to the extent that r ed i s highly p r e f e r r ed . Only when s i m i l a r exper iments are done on other populations of ra inbow trout and on other species of f i sh can we begin to assess the un ive r sa l i t y of these genera l iza t ions . P r o p e r l y designed exper iments are more l i k e l y to y i e l d re levant resuLts than are analyses of c r e e l census data. Th i s i s e spec ia l ly so if the exper iments are designed to elucidate the qualit ies ' of a good lu re rather than act as pi lot studies for management p roposa l s . Wi th the great weal th of factors that can affect the re la t ive success of a lure , utmost ca re must be taken to standardize techniques and to cont ro l , as best one can, va r iab les other than the test v a r i a b l e . However, knowledge gained f r o m this type of exper iment i s ce r t a in to benefit both the f i she r i e s manager and the f i she rman by making poss ib le f iner cont ro l on the efficiency and se lec t iv i ty of t r o l l i n g gear, which, i n the long run, should resu l t i n higher sustained y ie lds . 74 REFERENCES CITED A l i , M.A. 1959. The ocular structure, retinomotor and photobehavio^al responses of j u v e n i l e P a c i f i c salmon. Can. J . of Zool., 37:965-996. A l i , M.A. 1961. H i s t o p h y s i o l o g i c a l studies on the j u v e n i l e A t l a n t i c salmon (Salmo salar) r e t i n a . I I . Responses to l i g h t i n t e n s i t i e s , wavelengths, temperature and continuous l i g h t and dark. Can. J . Zool., 39:511-526. Baker, H.D., and W.A.H. Rushton. 1965. The r e d - s e n s i t i v e pigment i n normal cones. J . Physiol. 176:56-72. Behan-, F.L., R.A. Behan, and H.W. Wendhausen. 1972. Color perception underwater. Human Factors, 14(l):41-44. Beukema, J . J . 1970. Acquired hook-avoidance i n the pike Esox l u c i u s L. fished with a r t i f i c i a l and natural b a i t s . J . F i s h . B i o l . , 2_: 155-160. Boerema, L'.K. 1956. Some experiments on factors i n f l u e n c i n g mesh s e l e c t i o n i n trawls. J . Cons, i n t . Explor. Mer., 21:175-191. Borowik, J . 1930. On what does the catch of undersized f i s h depend? J . Cons, i n t . Explor. Mer., 5(2):195-215. Boydstun, L.B. 1972 Plugs (lures) i n the management of the C a l i f o r n i a t r o l l salmon f i s h e r y . P a c i f i c Marine F i s h e r i e s Commission, B u l l e t i n 8:5-13. Brown, Frank A. (Jr.) 1937. Responses of the large-mouth black bass to col o r s . I l l i n o i s Nat. H i s t . Surv. B u l l . , 21:33-55. Brownlee, K.A. 1965. S t a t i s t i c a l Theory and Methodology i n Science and Engineering. New York, John Wiley & Sons, 590 pp. Burrows, Roger E. 1969. The influence of f i n g e r l i n g q u a l i t y on adult salmon s u r v i v a l s . Trans. Amer. F i s h . S o c , 98 (4) : 777-784. Cousteau, J.Y.,and F. Dumas. 1953. The S i l e n t World. New York, Harper and Row. Davis, F.M. 1929. Preliminary note on experimental trawling with cod-end meshes of d i f f e r e n t s i z e s . J . Cons, i n t . Explor. Mer., 4^(3): 287-299. Dawkins, R. 1969. The at t e n t i o n threshold model. Anim. Behav., 17: 134-141. Emig, John W. 1971. C a l i f o r n i a inland angling survey for 1969, with cor-rections f o r the 1964 survey. C a l i f . F i s h & Game, 57(2):99-106. Espinosa, F.A. ( J r . ) , and James E. Deacon. 1971. Comparative angler success at Lake Mead (Arizona-Nevada). Progr. F i s h - C u l t . , 33(3): 170-174. ~ 75 G e l d a r d , F.A. 1972. The Human S e n s e s , 2nd ed. New Y o r k , J o h n W i l e y and Sons, xi+584 pp. G i n e t z , R.M., and P.A. L a r k i n . 1973. C h o i c e of c o l o r s o f f o o d i t e m s by r a i n b o w t r o u t (Salmo g a i r d n e r i ) . J . F i s h . Res. Bd. Can., 3 0 ( 2 ) : 229-234. G r a b e r , V. 1884. G r u n d l i n i e n z u r E r f o r s c h u n g des H e l l i g k e i t s - und F a r b e n s i n n e s d e r T i e r e . P r a g , Tempsky, 322 p. ( c i t e d f r o m Warner, 1931). . 1885. Ueber d i e H e l l i g k e i t s - und F a r b e n e m p f i n d l i c h k e i t e i n i g e r M e e r t i e r e . S i t z - B e r . Akad. W i s s . Wien., math-naturw. K l . , A b t . I., £1:129-150. ( c i t e d f r o m Warner, 1931). n Hess, C. 1910. Uber den a n g e b l i c h e n Nachweis v o n F a r b e n s i n n b e i F i s c h e n . A r c h i v f u r d i e gesammte P h y s i o l o g i e , 1 3 4(1,2):1-14. Aug. 15. ( c i t e d f r o m Brown, 1937). . 1911. E x p e r i m e n t e l l e U n t e r s u c h u n g e n z u r v e r g l e i c h e n d e n P h y s i o l o g i e des G e s i c h t s s i n n e s . A r c h i v f u r d i e gesammte P h y s i o l o g i e , 142 (9,10,11,12):405-46. ( c i t e d f r o m Brown, 1937). . 1913. Untersuchungen z i i r F r a g e n a c h dem Vorkommen v o n F a r b e n s i n n b e i F i s c h e n . Z o o l o g i s c h e J a h r b u c h e r ; A b t e i l u n g f u r a l l g e m e i n e Z o o l o g i e und P h y s i o l o g i e d e r T i e r e , .31(4):629-46. ( c i t e d f r o m Brown, 1937). . 1914. U n t e r s u c h u n g e n z u r P h y s i o l o g i e des G e s i c h t s s i n n e s d e r F i s c h e . Z e i t s c h r i f t f u r B i o l o g i e , 63[45 n.s.]:245-274. ( c i t e d f r o m Brown, 1937). H i n e l i n e , G e r t r u d e M. ( W h i t e ) . 1927. C o l o r v i s i o n i n t h e mudminnow. J . Exp. Z o o l . , 4 7 ( l ) : 8 5 - 9 4 . H o r i o , G. 1938. D i e F a r b - und F o r m d r e s s u r an K a r p f e n . J a p . J o u r . Med. S c i . , P a r t I I I ( B i o p h y s i c s ) , 4:395-402. J e s t e r , D.B. 1973. V a r i a t i o n s i n c a t c h a b i l i t y of f i s h e s w i t h c o l o r o f g i l l n e t s . T r a n s Amer. F i s h . S o c , 102(1) : 109-115. K o b a y a s h i , H., and M.A. A l i . 1971. E l e c t r o r e t i n o g r a p h i c d e t e r m i n a t i o n o f s p e c t r a l s e n s i t i v i t y i n a l b i n o and pigmented b r o o k t r o u t ( S a l v e l i n u s  f o n t i n a l i s , M i t c h i l l ) . Can. J . of P h y s i o l o g y and P h a r m a c o l o g y , 49(12):1030-1037. L a r k i n , P.A. 1949. R e p o r t on r e l a t i v e e f f i c i e n c i e s o f v a r i o u s t y p e s o f f i s h i n g gear w i t h s p e c i a l r e f e r e n c e t o Kamloops t r o u t . R e p o r t t o t h e 1949 P r o v i n c i a l [ B r i t i s h Columbia] Game Commission. McCracken, F.D. 19 63. S e l e c t i o n by codend meshes and hooks on c o d , haddock, f l a t f i s h and r e d f i s h , pp. 131-155. I n : The S e l e c t i v i t y o f F i s h i n g Gear. S p e c i a l p u b l . No. 5; b e i n g volume 2 o f P r o c e e d i n g s o f J o i n t ICNAF/ICES/FA0 S p e c i a l S c i e n t i f i c M e e t i n g , L i s b o n , 1957. 76 Marks, W.B. 1965. Visual pigments of single goldfish cones. J. Physiol., 178:14-32. Mast, Samuel 0. 1916. Changes in shade, color, and pattern in fishes, and their bearing on the problems of adaption and behavior, with especial reference to the flounders Paralichthys and Ancyclopsetta. Bull. Bur. Fish. (U.S.), 34:175-238. Pitre, K.R. 1970. Summary of "shaker" investigations in the west coast of Vancouver Island t r o l l fishery in 1968 and 1969. Dept. of Fish, and For., Pac. Reg. Tech. Rept., 1970-71, 10 pp. Pycha, R.L. 1962. The relative efficiency of nylon and cotton gillnets for taking lake trout in Lake Superior. J. Fish. Res. Bd. Can., 19(6):1085-1094. Ricker, W.E. 1958. Handbook of computations for biological s t a t i s t i c s of f i s h populations. Bull. Fish. Res. Bd. Canada, No. 119. 300 pp. Shetter, D.S., and G.R. Alexander. 1965. Results of angling under special and normal trout fishing regulations in a Michigan trout stream. Trans. Amer. Fish. Soc, 94 (3) : 219-226. Sokal, R., and F. Rohlf. 1969. Biometry. San Francisco, W.H. Freeman, 776 pp. Svaetichin, Gunnar, and E.F. MacNichol, Jr. 1958. Retinal mechanisms for chromatic and achromatic vision. Ann. N.Y. Acad. Sci., 74:385-404. Tomita, T., A. Kaneko, M. Murakami, and E.L. Pautler. 1967. Spectral response curves of single cones in the carp. Vis. Res., _7:519-531. Walls, G.L. 1963. The Vertebrate Eye and i t s Adaptive Radiation. New York, Hafner Publishing Co., v i i + 785 pp. Warner, Lucien H. 1931. The problem of color vision in fishes. Quart. Rev. Biol, 6(3):329-348. White, Gertrude M. 1919. Association and color discrimination in mudminnows and sticklebacks. J. Exp. Zool., 27(4):443-498. Wolf, H., and J. H. Wales. 1953. Color perception i n trout. Copeia, 1953 (4):234-236. Zupanovic, S. 1963. Experiments on the fishing effectiveness of trawls using wire cable bridle and wire cable bridle with manila, pp. 222-225. In: The Selectivity of Fishing Gear. Special publ. No. 5; being volume 2 of Proceedings of Joint ICNAF/ICES/FAO Special Scientific Meeting, Lisbon, 1957. 

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.831.1-0099872/manifest

Comment

Related Items