UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

A Comparative study of test procedures and measures of behaviour in the male three-spined stickleback,… Wootton, Robert John 1968

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

Item Metadata

Download

Media
831-UBC_1969_A1 W68.pdf [ 6.56MB ]
Metadata
JSON: 831-1.0102223.json
JSON-LD: 831-1.0102223-ld.json
RDF/XML (Pretty): 831-1.0102223-rdf.xml
RDF/JSON: 831-1.0102223-rdf.json
Turtle: 831-1.0102223-turtle.txt
N-Triples: 831-1.0102223-rdf-ntriples.txt
Original Record: 831-1.0102223-source.json
Full Text
831-1.0102223-fulltext.txt
Citation
831-1.0102223.ris

Full Text

A COMPARATIVE STUDY OF TEST PROCEDURES AND MEASURES OF BEHAVIOUR IN THE MALE THREE-SPINED STICKLEBACK, (Gasterosteus  aculeatus L.) by ROBERT JOHN WOOTTON B.A., Cambridge University, 1965 A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF Doctor of Philosophy i n the Department of Zoology We accept t h i s thesis as conforming to the required standard THE UNIVERSITY OF BRITISH COLUMBIA November, 1968 In p r e s e n t i n g t h i s t h e s i s i n 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 an a d v a n c e d d e g r e e a t 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 , I a g r e e t h a t t h e 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 r e f e r e n c e a n d S t u d y . I f u r t h e r a g r e e t h a t p e r m i s s i o n f o r e x t e n s i v e c o p y i n g o f t h i s t h e s i s f o r s c h o l a r l y p u r p o s e s may b e g r a n t e d b y t h e Head o f my D e p a r t m e n t o r b y h i s r e p r e s e n t a t i v e s . I t i s u n d e r s t o o d t h a t c o p y i n g o r p u b l i c a t i o n o f t h i s t h e s , i s f o r f i n a n c i a l g a i n s h a l l n o t b e a l l o w e d w i t h o u t my w r i t t e n p e r m i s s i o n . D e p a r t m e n t o f " 3 ^C^Q L_<OC)'/ The 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 V a n c o u v e r 8 , C a n a d a D a t e 1+ ^ a w - l ^ - a e : - \«=t fe.'g i ABSTRACT This study compared three methods used to measure aggression i n the male three-spined stickleback. These methods have been used i n d i f f e r e n t studies without the c e r t a i n t y that they measure the same behavioural phenomenon. A- number of d i f f e r e n t measures of aggression must be highly c o r r e l a t e d i f aggression i n the male stickleback i s to be regarded as a unitary d r i v e . The methods consisted of recording the behaviour of an experimental male when: ( i ) a tube containing another male (or a female i n some tests) was v i s i b l e f o r 5 min; ( i i ) another male was v i s i b l e across a glass p a r t i t i o n at a l l times; ( i i i ) a fish-shaped wax model was v i s i b l e f o r 2 min. Each method showed that i n the 12 days a f t e r f e r t i l i z a t i o n of an egg-mass, there was a U-shaped trend i n frequency of b i t i n g , frequency of charging and the rate of b i t i n g per minute spent oriented towards the other f i s h or the model. In Method ( i ) , t o t a l oriented time formed about 70 % of a t e s t period and t h i s measure was not correlated with frequency of b i t i n g , nor did i t follow a U-shaped trend. In Methods ( i i ) and ( i i i ) , t o t a l oriented time formed less than 50 % of a te s t period, was correlated with frequency of b i t i n g , and did follow a U-shaped trend. In Methods ( i ) and ( i i ) , the measures, frequency of b i t i n g and b i t e s per min of oriented time were at a maximum i i i n the f i r s t 15 min a f t e r f e r t i l i z a t i o n . Method ( i i i ) did not show t h i s maximum i n biting„ A l l methods showed that frequency of b i t i n g , of charging, and b i t e s per min of oriented time were higher f o r f i s h with nests than f o r f i s h which had yet to. b u i l d nests. Gonadectomy of f i s h with nests reduced a l l measures to the le v e l s found f o r i n t a c t f i s h that did not have nests. An experiment using only Method ( i ) showed that a male with a nest attacked another male more than a non-gravid female. A male without a nest attacked both a male and a non-gravid female equally. The use of more than one measure f o r a behaviour pattern improved both the comparisons between methods and the analysis'of changes i n behaviour of the male. Examples of t h i s are given f o r b i t i n g and parental fanning. Results from Method ( i i i ) were not always consistent with those from Methods ( i ) and ( i i ) . Method ( i i i ) was exceptional i n the number of charges the experimental males made with both dorsal and ventral spines erect. A wider range of behaviour patterns were recorded more re g u l a r l y i n Method ( i i ) than i n Method ( i ) . But these two methods con s i s t e n t l y showed the same trends i n frequency of b i t i n g , charging and bit e s per min of oriented time. In s p i t e of the s i m i l a r i t i e s i n the r e s u l t s from the three methods there was not s u f f i c i e n t evidence to conclude that aggression i n the male stickleback i s a unitary d r i and that a l l p o t e n t i a l measures of aggression w i l l be equivalent. LIST OF CONTENTS Introduction A. Statement of the problem B. L i f e h i s t o r y of the male three-spined stickleback. Materials and Methods A. C o l l e c t i o n and Maintenance B. Recording C. Time of t e s t i n g D. A c t i v i t i e s recorded E„ Gonadectomy P. Testing procedures Parental Cycle Experiment I: changes i n measures i n the 12 days a f t e r f e r t i l i z a t i o n . A. Introduction B„ Experiment l a : changes over the parental period measured using tube t e s t s . C. Experiment l b : changes over the parental period measured using p a r t i t i o n t e s t s . D. Experiment l c : changes over the parental period measured using model t e s t s . E. Comparison of selected measures. F„ ^ p i n e - r a i s i n g . G„ D i v i s i o n of a c t i v i t i e s i n a t e s t period. H. Parental fanning. I. Experiment 2: changes i n measures i n the 24 hr a f t e r f e r t i l i z a t i o n . The P r e - f e r t i l i z a t i o n Period A. Introduction B. Experiment 3: a comparison of the pre-nest and empty-nest phases. C. Experiment 4: e f f e c t of gonadectomy i n the pre-nest and empty-nest phase. D. Experiment 5: e f f e c t of s i z e and sex of the t e s t f i s h . Discussion and Conclusions Bibliography V LIST OF TABLES TABLE PAGE I. F - r a t i o s and p r o b a b i l i t y values obtained 24 by Trend Analysis of selected measures over the twelve day parental period. Experimental males tested against a mature male i n a p l e x i g l a s s tube. I I . Comparison of d a i l y means f o r the parental 26 period using Duncan's New Multiple Range Test. I I I . Correlations between selected measures 27 calc u l a t e d using a l l recordings from the parental period. IV. Correlations between selected measures 31 fo r each day of the parental period. V. Correlations between selected measures 34 fo r the i n d i v i d u a l f i s h over the parental period. VI. F - r a t i o s and p r o b a b i l i t y values obtained 40 by Trend Analysis of selected measures over the twelve day parental period. Experimental males tested against a mature male i n continual v i s u a l contact. VII. Correlations between selected measures using 45. a l l recordings from the parental period. VIII. Correlations between selected measures f o r 46 each day of the parental period. IX. . Correlations between selected measures f o r 48 i n d i v i d u a l f i s h over the parental period. X. Comparison of the l e v e l of response to 61 Model S/R over the 12 day parental period by a Friedman 2-way analysis of variance. XI. Non-parametric trend analysis of selected 61 measures obtained i n tests using Model S/R. XII. A comparison of the l e v e l s of response to 64 four models during the parental period by a Friedman 2-way analysis of variance. X I I I . Comparison of d a i l y means f o r selected measures 70 obtained i n the parental period using a l i v e male f i s h i n a tube with d a i l y means obtained from the same period from the p a r t i t i o n t e s t s . v i TABLE. XIV. XV. XVI. XVII. XVIII, XIX. XX. XXI. XXII. XXIII, PAGE' 71 72 73 Comparison of d a i l y means f o r selected measures obtained i n the parental period using a l i v e male f i s h i n a tube with d a i l y means obtained from the same period using models as tes t objects. Comparison of d a i l y means for selected measures obtained i n the parental period using p a r t i t i o n t e s t s with d a i l y means obtained f o r the same period using models as te s t objects. Comparison of d a i l y means f o r selected measures obtained i n the parental period using a l i v e , gravid female as t e s t f i s h i n a tube with d a i l y means obtained from the same period using models as tes t objects. Sample sizes used i n Experiment 2 Comparison of l e v e l s of selected measures from the pre-nest and empty-nest phases. Data from the tube t e s t s . Correlations between selected measures from f i s h i n the pre-nest and empty-nest phases. Data from tube t e s t s . Comparison of l e v e l s of selected measures from the pre-nest and empty-nest phases. Data from p a r t i t i o n t e s t s . Correlations between selected measures from f i s h i n the pre-nest and empty-nest phases. Data from p a r t i t i o n t e s t s . Comparison of l e v e l s of selected measures between f i s h gonadectomised before experience of nest-building and sham-operated f i s h s t i l l i n pre-nest phase. Data from t e s t s with mature male i n a tube. Four weeks a f t e r the operation. Comparison of l e v e l s of selected measures betweenll5 f i s h gonadectomised a f t e r b u i l d i n g nests and f i s h sham-operated a f t e r b u i l d i n g nests. Data from tests with mature male i n a tube. Four weeks a f t e r the operation. 89 99 100 103 105 114 XXIV. Comparison of selected measures from gonadect-omised and sham-operated f i s h . Data from tube t e s t s . 116 V I 1 TABLE XXV. XXVI. XXVII, XXVIII. XXIX. Mean number of b i t e s per 2 min at a wax model, four weeks a f t e r operation. Mean number of zig-zags per 2 min to a wax model, four weeks a f t e r operation. Comparison of selected measures obtained from p a r t i t i o n tests on gonadectomised and sham-operated f i s h . F - r a t i o s and p r o b a b i l i t i e s from analysis of variance of selected measures from f i s h i n pre-nest phase. Tested against males and non-gravid females i n two s i z e categories. F - r a t i o s and p r o b a b i l i t i e s from analysis of variance of selected measures from f i s h i n empty-nest phase. Tested against males and non-gravid females i n two s i z e categories, PAGE 118 118 121 125 126 XXX Means f o r sex and s i z e f a c t o r s . 129 v i i i LIST OF FIGURES FIGURE. PAGE l o Example of an Esterline-Angus record. 11 2. Changes i n the l e v e l s of selected measures 20 during the reproductive cycle, tests with a male f i s h i n a tube. 3. Relationship between t o t a l oriented time, 29 number of bouts of oriented time, mean bout length of oriented time, and mean bout length of non-oriented time. Data from tube t e s t s . 4. Plot of the c o r r e l a t i o n c o e f f i c i e n t , r, 36 between b i t e s per 5 min and t o t a l oriented time f o r i n d i v i d u a l f i s h against the mean l e v e l of t o t a l oriented time f o r each f i s h over the twelve day parental period. 5. Changes i n the l e v e l of selected measures 42 during the reproductive c y c l e , tests using the p a r t i t i o n design. 6. Relationship between t o t a l oriented time 50 number of bouts of oriented time, mean bout length of oriented time and mean bout length of non-oriented time. Data from p a r t i t i o n t e s t s . 7. Changes i n the l e v e l s of selected measures 53: during the twelve day parental period shown by the f i s h with no clutch i n the p a r t i t i o n t e s t s . 8. Changes i n the responses of four f i s h to 58 Model S/R. over the 12 days following the f e r t i l i z a t i o n of a c l u t c h of eggs. 9. Relationships between selected measures i n 62 the model t e s t s . 10. Comparison of the l e v e l s of response to the 65 four models on the f i r s t , fourth, eighth and twelth day a f t e r f e r t i l i z a t i o n . 11. Comparison f o r the three t e s t methods of spine- 78 r a i s i n g i n charging. 12. Mean percentage of t e s t period spent i n time 80 at the nest; fanning 1, and oriented towards the t e s t object. i x FIGURE PAGE 13'-. a) Mean t o t a l duration of fanning (sec 84 per 15 min) f o r s o l i t a r y males and f o r males with a partner ( p a r t i t i o n t e s t s ) , b) Mean number of bouts of fanning i n a 15 min period f o r s o l i t a r y males and f o r males with a;:;partner«, 14. Frequency d i s t r i b u t i o n f o r 5 sec classes 86 of mean bout length of fanning f o r the f i r s t ten days a f t e r f e r t i l i z a t i o n of one clu t c h of eggs. 15. Changes i n selected measures i n the f i r s t 92 24 hr a f t e r f e r t i l i z a t i o n of one clu t c h of eggs. 16. Changes i n spine - r a i s i n g i n the f i r s t 24 hr 93 af t e r f e r t i l i z a t i o n of one c l u t c h . Data from tube and p a r t i t i o n t e s t s . 17. -Relative changes i n frequency of b i t i n g 119 following gonadectomy i n the empty-nest phase. Data from tube and model t e s t s . X ACKNOWLEDGEMENTS; I wish to thank my supervisor Dr. N.R. L i l e y f o r his advice and encouragement during t h i s study and for h i s c a r e f u l and valuable e d i t i n g of the manuscript. I also thank my doctoral committee D r s . W.S. Hoar, I.E. E f f o r d , J.D. McPhail and D.J. Randall. Their suggestions and reviews of the manuscript greatly improved the t h e s i s . R. Black and D. Kramer read parts of the manuscript and made many useful suggestions. S. Borden advised on s t a t i s t i c a l problems. Miss R. Wood drew a l l the i l l u s t r a t i o n s . My wife Maureen did most of the tedious work of counting Esterline-Angus charts. Her patience, commonsense and encouragement were most important contributions to t h i s study. INTRODUCTION 2 A. Statement of the problem In the causal study of behaviour i t i s often necessary to use standard methods to obtain reproducible, quantitative measures of selected behaviour patterns. Frequently workers studying the same species have used d i f f e r e n t methods to measure the same behavioural phenomenon. Whether the d i f f e r e n t methods y i e l d the same r e s u l t s f o r a given experiment has r a r e l y been considered. This study compares three standard methods that have been used to obtain measures of aggression i n the male three-spined stickleback (Gasterosteus aculeatus L . ) . In one method, a tube containing a f i s h i s placed within the t e r r i t o r y of a s o l i t a r y male stickleback. Usually the t e s t period i s three or f i v e minutes long. The frequency of b i t e s d i r e c t e d by the t e r r i t o r i a l male at the f i s h i n the tube i s then recorded. Van I e r s e l (1958), Sevenster (1961), Segaar (1961) and Sevenster-Bol (1962) have a l l used t h i s method to measure the aggressive drive of the male. Symons (1965) used i t to investigate the r e l a t i o n s h i p between aggres-sion and s p i n e - r a i s i n g . Baggerman (1966) used another method i n her study of the r o l e of hormones i n the reproductive cycle of the male stic k l e b a c k . The experimental male was separated from another male by a glass p a r t i t i o n . Thus the experimental male was i n v i s u a l contact with another male throughout the experiment and not j u s t at the time of t e s t i n g . A. record was made of the 3 behaviour of the experimental male directed towards the other f i s h . Morris (1958) used a set of v a r i o u s l y coloured models to measure changes i n the aggression of the male ten-spined stickleback (Pyqosteus punqitius L ) . Again frequency of bites per t e s t period was regarded as a measure of aggression. The t e s t period was one minute long. Models have also been used by Tinbergen (1951) and Muckensturm (1967, 1968) to study the s t i m u l i to which the male three-spined stickleback responds i n f i g h t i n g and courtship. These are the three t e s t methods considered i n t h i s study, but other methods have been used. For instance Hoar (1962) studied the r o l e of hormones i n the reproductive cycle of the male stickleback. He kept four f i s h to an aquarium and ranked the l e v e l of aggressiveness i n each tank on a scale ranging from 0 (no attacks) to IV (dominance-subordination r e l a t i o n s h i p s with one or more males strongly t e r r i t o r i a l ) . Methods, i n which more than one f i s h i s considered, are not included i n t h i s study. Why i s such a comparative study necessary? In many of the above studies, only one measure has been taken as representative of the aggressive d r i v e . This has been c r i t i s i z e d by M i l l e r (1957). He writes, "... the type of analysis required ... to make intervening v a r i a b l e s , such as 4 d r i v e s , meaningful i s one i n which i t i s possible to compare the e f f e c t s of a number of d i f f e r e n t experimental operations on a number of d i f f e r e n t measures of the v a r i a b l e . " He suggests that i f discrepant r e s u l t s are found, that i s some measures are not highly c o r r e l a t e d , the discrepancy can be explained i n two ways. E i t h e r the drive i s not a single unitary variable but a c l u s t e r of i n t e r - r e l a t e d v a r i a b l e s , or the measures are impure and are affected by factors other than the d r i v e . I t i s not c l e a r how these two a l t e r n a t i v e s can be distinguished i n p r a c t i c e . Hinde (1959) has reviewed the concept of unitary d r i v e s . He suggests that the concept hinders a causal analysis of behaviour because a unitary drive may be regarded as an unanalysable e n t i t y i n the causal network. Therefore he urges that i n the causal analysis of a complex sequence of a c t i v i t i e s unitary concepts such as drives be abandoned. This r a d i c a l suggestion eliminates the need to d i s t i n g u i s h between impure measures and c l u s t e r s of i n t e r - r e l a t e d v a r i a b l e s . Both M i l l e r and Hinde imply that more than one measure of a behaviour sequence should be taken. This has not been done i n studies on the three-spined stickleback. Therefore the purpose of t h i s study i s to compare selected measures of behaviour obtained by d i f f e r e n t t e s t procedures at d i f f e r e n t phases i n the reproductive c y c l e . I t also r e l a t e s measures not previously used i n such studies to measures such as frequency of b i t i n g that have been used. 5 The following questions were asked: ( i ) What changes take place i n recorded behaviour i n the experiments? ( i i ) What are the r e l a t i o n s h i p s between measures i n an experiment? ( i i i ) How does the t e s t method a f f e c t the answers to questions ( i ) and ( i i ) ? (iv) How do the r e s u l t s compare with those of other authors? Hagen (1967) has shown that i n south-western B r i t i s h Columbia, the non-migratory l e i u r u s form of the three-spined stickleback Gasterosteus aculeatus i s , to a large extent, reproductively i s o l a t e d from the migratory trachurus form. The r e s u l t s of an e t h o l o g i c a l study of one form may not correspond c l o s e l y to the r e s u l t s obtained with the other. Many authors, including Symons (1965) and Van den Assem (1967) have used the trachurus, the l e i u r u s and the hybrid forms within one experiment. Only l e i u r u s f i s h from a r e s t r i c t e d area of a s i n g l e stream were used i n t h i s study. B. L i f e h i s t o r y of the male three-spined stickleback Complete descriptions of the behaviour of the three-spined stickleback have been given by Tinbergen (1953, 1954) and van I e r s e l (1953). Some of these descriptions are f o r the migratory trachurus form. But the form of the behaviour patterns described c l o s e l y f i t t e d those seen i n the leiurus males . 6 The l e i u r u s form of the three-spined stickleback remains i n streams or ponds for the whole of i t s l i f e . L i t t l e i s known about the s o c i a l structure of a population during the winter. In spring the mature males are t e r r i t o r i a l . The male bu i l d s a nest within t h i s t e r r i t o r y . When the nest i s complete, the male courts gravid females and induces them to spawn i n the nest. During the courtship encounter the male, which i s normal-l y dark, blanches to a pale creamy brown. This colour change l a s t s only as long as the courtship. When the male has a number of clutches, i t stops courting and d i s t r i b u t e s i t s time between v e n t i l a t i n g the eggs, and t e r r i t o r i a l defence. In t h i s parental period the male becomes very dark, i n some cases the dorsal surface i s almost black. When the eggs hatch the male tends the f r y f o r a few days, returning them to the nest p i t i f they swim away. However, the young soon disperse and the male may r e b u i l d i t s nest. A f t e r the young disperse the males loose the very dark colour. At the end of the spawning season, there i s a high adult m ortality (Hagen, 1967). 7 PART I I : MATERIALS, and METHODS 8 A. C o l l e c t i o n and maintenance S-ticklebacks were c o l l e c t e d from January to May i n the L i t t l e Campbell River i n southwestern B r i t i s h Columbia. They came from a region of the stream i n which no trachurus forms were caught. Hagen (1967) gives d e t a i l s of the r i v e r and the c h a r a c t e r i s t i c habitats of the two forms. The f i s h were placed i n cold (c.lO° C), running water, under a r t i f i c i a l l i g h t s whose photoperiod corresponded with the natural photoperiod. As required batches of eight to ten f i s h were placed i n 45 l i t r e glass tanks. The water temperature was allowed to r i s e to room temperature (18-20° C). A photoperiod of 16 hours l i g h t and 8 hours dark was maintained then and throughout the subsequent experiments unless otherwise stated. F i s h were fed frozen brine-shrimp augmented with l i v e Daphnia, chopped earth-worm, and other l i v e food when a v a i l a b l e . Feeding was c a r r i e d out a f t e r a l l recording f o r the day had been completed. Under these conditions the males developed t h e i r c h a r a c t e r i s t i c breeding colours and could be separated from the females. With s u f f i c i e n t feeding, the females became gravid and would spawn several times i n a season. S l i g h t i n f e c t i o n s of Ichthyophthirius were treated e i t h e r by keeping the f i s h i n fresh water t e l e o s t s a l i n e f o r a week or two or by d a i l y treatment with malachite green. B. Recording In most experiments selected behaviour patterns were recorded on a twenty-channel Esterline-Angus event recorder. The chart speed was 7.4 cm per min (15 inches = 5 min). Fanning i n the parental cycle was measured by counting the number of bouts of fanning i n a 15 min period. The t o t a l duration of fanning i n the 15 min was measured to the nearest two seconds. C. Time of Testing In the analysis of data no account was taken of the time of day at which i n d i v i d u a l records were made. A preliminary experiment had suggested that the l e v e l s of the variables measured did not vary s i g n i f i c a n t l y within a given sixteen hour or eight hour period of l i g h t . Consequently recordings were usually made between 11.00 A.M. and 5.00 P.M. D. A c t i v i t i e s recorded The same a c t i v i t i e s were recorded i n each te s t method. Thus one measure could be compared across the three methods. Relationships between measures recorded i n the same t e s t procedure could also be examined. Those a c t i v i t i e s recorded i n t h i s study are defined 10 1. A c t i v i t i e s r e l a t e d to the other f i s h or model i ) Orientation This i s a l l the time the experimental male i s facing i t s opponent, plus any time spent i n the head-down threat posture. A l l a c t i v i t i e s d i r e c t e d towards the opponent are thus included within the oriented time (See F i g . 1). i i ) Charge The male swims towards i t s opponent. The aspect of the spines during the charge i s recorded. Spines are recorded as r a i s e d only when they are f u l l y erect. Charges are not defined on the basis of the speed of the approach. This i s because the distance covered i n a charge i s often too small f o r a reasonable estimation of the speed of the approach. Such a d e f i n i t i o n of charging can be c r i t i c i s e d on the grounds that i t i s i n s u f f i c i e n t l y p r e c i s e . However i n almost a l l the t e s t s , slow approaches formed only a very small proportion of the t o t a l , and i t was not considered worthwhile adding a f u r t h e r category f o r these approaches. i i i ) Bump The male h i t s the glass near i t s opponent or h i t s the model with i t s mouth closed. i v ) Bite The male opens and closes i t s mouth against the glass near i t s opponent or against the model. Figure 1 Example of an Esterline-Angus record showing that a c t i v i t i e s directed towards the t e s t f i s h or model are included i n the oriented time. FIGURE I f a n J X J T J T J T n . lead JIIIUIIIUIIIUII1 , . : z i g zag 1 ; . ( T n r 1 at nest „ I I I I undefined activity J T J ~ T _ r i n : I l i l T l _ l 1 : oriented glue Lil bump 111 bite charge, dorsal and ventral spines erect charge,dorsal spines erect J U charge, no spines erect 3" = I min 13 A b i t e or a bump i s always preceded by a charge, although a charge does not always end i n e i t h e r a b i t e or a bump. v) Head-down threat The male stands v e r t i c a l l y with i t s head down and i t s broadside towards the opponent. I t s spines are usually -r a i s e d . Often a male i n t h i s p o s i t i o n jerks i n a character-i s t i c way. vi ) Zig-zag This i s the c h a r a c t e r i s t i c approach of a courting male to a gravid female. Tinbergen (1953) describes i t thus, "... consists of a serie s of leaps during which the male f i r s t turns as i f going to swim away from the females, then abruptly turns towards them ...". In t h i s study each leap was recorded. E s p e c i a l l y when the t e s t f i s h i s i n the tube i t i s d i f f i c u l t to decide i f the leap i s towards or away from the tube because the experimental f i s h tends to c i r c l e the tube. I f a zig-zag approach ends i n a b i t e , the sequence recorded i s : zig-zag, charge, b i t e . 2. A c t i v i t i e s r e l a t e d to the nest i ) Time at Nest This combines time spent boring, pushing and glueing. I t also includes time spent over the nest with i t s head di r e c t e d towards the nest but not fanning. I t also includes 14 time spent creeping through the nest, although on the Esterline-Angus chart a separate record was made i f creeping through took place. Nelson (1965) described changes i n behaviour consequent on creeping through, most notably an increase i n 'Vacuum1 zig-zagging and a decrease i n fanning. But creeping through did not occur very often during t e s t s , so the r e s u l t s do not include many instances of sharp changes i n behaviour such as Nelson described. i i ) Fanning The male positions himself i n f r o n t of the nest entrance and at an angle to i t . He r a p i d l y beats h i s t a i l but holds a stationary p o s i t i o n by compensatory movements of the pectorals. I t i s an a c t i v i t y c h a r a c t e r i s t i c of the male i n the nest-present and parental phases of the breeding c y c l e . 3. Variables c a l c u l a t e d from raw data i ) Bites per minute of oriented time. i i ) Proportion of charges ending i n a b i t e . i i i ) Mean bout length of oriented time. i v ) Bites per bout of oriented time. E . Gonadectomy For 24 to 48 hours before the operation the f i s h were not fed. This prevented the i n t e s t i n e bulging through the i n c i s i o n s a f t e r the operation. The f i s h were anaesthetised 15 using a concentration of t r i c a i n e methane sulphonate (MS222-Sandoz) which caused immobilization i n about ten minutes (1:9,000). Small i n c i s i o n s were made on both sides of the abdomen through which the testes were removed. The f i s h were then placed i n fresh water t e l e o s t s a l i n e f o r 24 hours before being returned to t h e i r experimental tank. Sham operations followed the same procedure except the testes were not removed. This technique was found to give 80.- 90 % s u r v i v a l with f i s h both before they had b u i l t a nest and i n the period when they had a nest. F. Testing procedures Three s i t u a t i o n s were used i n which the reactions of the experimental male f i s h to other f i s h could be recorded. 1. Tube Tests Experimental males were taken from the 45 l i t r e tanks and placed one to a 23 l i t r e tank. Each of these tanks had s u f f i c i e n t vegetation f o r nest-building. External f i l t e r s were used to clean and aerate the water. A p l e x i g l a s s tube 6 cm i n diameter containing a f i s h , confined to the lower 12 cm by a plunger, was introduced into the experimental male's tank. I f the male had a nest or had had a nest, then the tube was placed 20 cm from the nest 16 entrance. Recording continued f o r f i v e minutes a f t e r the f i r s t approach of the male to the f i s h i n the tube. 2. Model Tests Fish-shaped pieces of wax were painted, and suspended from a piece of wire so that they could be moved up and down i n f r o n t of an experimental male. These males were i s o l a t e d i n 2 3 l i t r e tanks. I f the male had a nest the model was kept 20 cm from the nest. Marks on the wire ensured that the models were always moved the same distance i n the v e r t i c a l plane. Each cyc l e took about two seconds. The t e s t lasted two minutes from the f i r s t approach of the male to the model. 3. P a r t i t i o n Tests. The t h i r d s i t u a t i o n was a modification of that used by Baggerman (1966). A 45 l i t r e tank was divided i n t o two equal sections with a transparent glass p a r t i t i o n . Vegetation was placed i n both halves. Each section was f i l t e r e d and aerated by a sub-gravel f i l t e r . A. male was allowed to b u i l d a nest i n one section and then the experimental male was placed i n the other s e c t i o n . A. t e s t sequence i n t h i s s i t u a t i o n lasted f i f t e e n minutes. For the f i r s t f i v e minutes no record was made. Then the a c t i v i t i e s of one male were recorded f o r f i v e minutes, followed immediately by a f i v e minute record of the second male's a c t i v i t i e s . The order i n which the males were recorded was determined by tossing a coin. PART I I I : THE PARENTAL PHASE 18 Experiment I: Changes i n measures i n the 12 days a f t e r f e r t i l i z a t i o n Ao Introduction The parental phase i s temporally the most c l e a r l y defined period i n the male's breeding c y c l e . The times of f e r t i l i z a t i o n , hatching, and dispersion of the f r y from around the nest can be known to within a few hours:. At constant temperature, the time of hatching i n terms of days a f t e r f e r t i l i z a t i o n i s approximately the same f o r a l l f i s h . Because of t h i s synchrony, the parental phase was used f o r a d e t a i l e d comparison between methods of t e s t i n g . Thus r e s u l t s from the parental phase are presented before r e s u l t s from other phases of the reproductive cy c l e of the males. In the experimental tanks f r y were usually v i s i b l e i n the nest p i t eight days a f t e r f e r t i l i z a t i o n . They had dispersed from about the nest by day twelve. The o l d nest was p u l l e d to pieces a f t e r the f r y had hatched. The male then b u i l t another nest. Changes i n the behaviour of males as recorded i n the three t e s t methods were followed f o r twelve days a f t e r f e r t i l i z a t i o n . For selected measures the following questions were asked: i ) What changes does a given measure show over the twelve day period? i i ) How do the r e l a t i o n s h i p s between given measures change over the twelve day period? 19 i i i ) Are the changes i n a given measure the same i r r e s p e c t i v e of the te s t method i n which i t was measured? B. Experiment l a : Changes over the parental period measured using tube t e s t s . Twelve f i s h were tested against a male i n the p l e x i -glass tube. Each experimental male was tested once a day for the twelve days following the f e r t i l i z a t i o n of one cl u t c h of eggs. The male i n the plexiglass, tube was a mature f i s h , i n most cases one that had a nest i n i t s home tank. Changes i n the measures obtained over the twelve days were examined by a trend analysis (Edwards 1962). Means f o r each day were compared by a Duncan's New Multiple Range Test (Steel and T o r r i e 1960). Correlations between the measures were also c a l c u l a t e d . Results i ) Trend Analysis f o r the 12-day period Changes i n some of the measures over the parental period are shown i n F i g . 2, For completeness, r e s u l t s obtained by the same method from f i s h i n the pre-nest and empty-nest phases have been included i n the graphs, but were not included i n the a n a l y s i s . Graphing the r e s u l t s suggested the measures f e l l i n t o two groups. In one group the d a i l y mean value f o r each Figure 2 Changes i n the l e v e l s of selected measures during the reproductive c y c l e , tests with a male f i s h i n a tube. Length of open bar i s two standard e r r o r s , the v e r t i c a l l i n e gives the range. P.N., pre-nest phase; E.N., empty-nest phase; numbered days are days a f t e r the f e r t i l i z a t i o n of one clutch of eggs. FIGURE 2 200 180 160 140 c 120 E 100 in 80 « CL 60 U) tt .tr 40 CD 20 0 260 PN EN I 2 3 4 5 6 7 8 9 10 II 12 Days E m a. a> o> k_ o 280 260 240 220 200 180 160 140 120 100 80 308 I PN EN I 2 3 4 5 6 7 8 9 10 II 12 Days Mean bout length of oriented time OJ * Ol o o o - B B --RR--SB--AN—10 —m-— s -Bites per bout of oriented time O u> O Oi o Ol —F=»=q : v 8 — s :  _ s s 3-— s — —Eel— — & Total oriented time (sec per 5 min) o o o 01 o CD o o o IN) o o 01 o CD o OJ o o ro OJ Ol CD 10 - S 3 -o a _ • — G) m Bouts of oriented time •o z — ro o o OJ o O 01 o o U3 O O O O o 13 m z -BB-0) m ran 2 5 - p s ^ — — H S -— S B — 23 measure declined to a minimum around the eighth day a f t e r f e r t i l i z a t i o n . I t then increased over the next four days. Frequency of b i t i n g per f i v e minutes, frequency of charging, the proportion of charges ending i n a b i t e and the frequency of b i t i n g per minute of oriented time a l l showed t h i s pattern of change. The other group, which consisted of t o t a l oriented time, number of bouts of oriented time, mean bout length of oriented time, and number of b i t e s per bout of oriented time, showed no comparable pattern of change over the parental period. This d i v i s i o n i n t o two groups was confirmed by the trend analysis (Table I ) . For those measures that showed a s i g n i f i c a n t trend over the twelve day period the quadratic component was the most important element i n that trend. This was expected from the conspicuous U-shape i n the graphs. In addition the trends f o r both b i t e s per f i v e minutes and charges per f i v e minutes had a cubic component, that i s the curve over the twelve day period has both a minimum (around day eight) and a maximum (at the s t a r t of the period). There was a weak l i n e a r component i n the trend f o r b i t e s per f i v e minutes (0.1>P>0.05), but t h i s component was s i g n i f i c a n t i n the trends f o r charges per f i v e minutes and b i t e s per minute of oriented time. This l i n e a r component represents an o v e r a l l increase i n the l e v e l of a measure between the f i r s t days of the period and the l a s t days of the period. 24 Table I F - r a t i o s and p r o b a b i l i t y values obtained by trend analysis of selected measures over the twelve day parental period. Experimental males tested against a mature male i n the p l e x i g l a s s tube. N=12. Measure Bites per 5 min Source of sums of F P squares Overall trend 4.95 <.005 l i n e a r component 2.85 <»i quadratic component 41.8 <.005 cubic component 5.95 <.025 other trends <1 -Charges per 5 min Overall trend 3.25 <.005 l i n e a r component 4.97 <.05 quadratic component 19.5 <.005 cubic component 5.15 <>025 other trends <1 -Bites per min of oriented time Overall trend 4.20 l i n e a r component 8.70 quadratic component 32.6 cubic component 1.11 other trends <1 <;.005 <.005 <.005 ->.25 Proportion of charges ending i n a b i t e Overall trend 3.57 <;.005 l i n e a r component <ll -quadratic component 33.8 <;.005 cubic component <1 -other trends <1 -Bites per bout of oriented time Overall trend T o t a l oriented time Overall trend Number: of bouts of Overall trend oriented time <1 <1 Mean bout length of oriented time Overall trend <1 25 Thus the trends f o r these measures over the twelve day period did not follow i d e n t i c a l patterns of change, although the quadratic component was conspicuous i n a l l of them. The comparison of d a i l y means i s shown i n Table I I . Consider those measures which had a s i g n i f i c a n t F - r a t i o i n the trend a n a l y s i s . The l e v e l s on and around day eight were s i g n i f i c a n t l y lower than the lev e l s f o r the same measure at the beginning and end of the parental phase. There were no comparable s i g n i f i c a n t differences f o r the, measures that did not have a s i g n i f i c a n t F - r a t i o i n the trend a n a l y s i s . i i ) Correlations between measures A c o r r e l a t i o n matrix for selected measures was ca l c u l a t e d using a l l the recordings from the parental period. Each measure was represented by 144 values (Table I I I ) . Bites per f i v e minutes was highly correlated with b i t e s per minute of oriented time (r = 0.83), charges per f i v e minutes (r = 0.81), and the proportion of charges ending i n a b i t e (r = 0.81). There was no s i g n i f i c a n t c o r r e l a t i o n between b i t e s per f i v e minutes and t o t a l oriented time. There was a s i g n i f i c a n t c o r r e l a t i o n between bi t e s per f i v e minutes, and b i t e s per bout of oriented time (r = 0.46). Bites per minute of oriented time was negatively correlated with t o t a l oriented time (r = -0.38). However, charges per f i v e minutes was p o s i t i v e l y correlated with t o t a l oriented time (r = 0.42). 26 Table II Comparison of d a i l y means f o r the parental period using Duncan's New Multiple Range Test. Means f o r days not underscored by the same l i n e are s i g n i f i c a n t l y d i f f e r e n t at a s p e c i a l protection l e v e l of 95 %. Measure Bites per 5 min Charges per 5 min Proportion of charges ending i n a b i t e Bites per bout of oriented time T o t a l oriented time Number of bouts of oriented time Mean bout length of oriented time Days of the parental period 12 11 1 2 10 4 3 9 7 5 6 8 12 11 2 10 1 4 3 5 7 9 6 8 12 1 11 2 3 4 10 6 7 9 5 8 No differences are s i g n i f i c a n t No differences are s i g n i f i c a n t 11 9 12 5 10 2 6 3 4 1 7 8 No differences are s i g n i f i c a n t 27 Table I I I Correlations between selected measures calcul a t e d using a l l recordings from the parental period. Data from tube tests.N=144. Only c o r r e l a t i o n s s i g n i f i c a n t at the 5 % l e v e l are included. HH £ 4H O C •H o 4H to CU -P m rC o d E cu -P o CO -P •H Cn-H •o -P Cn -p e in d cu 3 CD d CD 3 CU (0 -p •H s O £ cu £ O £ LD U ,c m c E -H ,3 -H i—\ *r\ u cu •P •P •p -P 5-1 a c •H u MH -P <u CD TJ O XS <U T J a CO O 0 CD o cu O h 0) cu o> 4-> M -P A -P •P 10 o d H co d cu d d CO d CD u a -H (fl 0> OJ ,b OJ d 0) CD CU -P (0 o *o 4J -P -rH £ -rl (ti tH •P -H •H x: M d O •H JH <U 5H •rl U CD. u cw <u EH EQ O 2 O s o CQ O Bites per 5 min 1.00 .81 .81 .83 .46 Charges per 5 min 1 .00 .35 .42 .52 -.20 .30 .55 Prop, of charges 1.00 -.18 .83 .22 ending i n a b i t e T o t a l oriented time 1.00 -.38 -.80 . 67 .55 Bites per min of 1.00 .37 -v,24 oriented time Number of bouts of 1.00 -*77 .64 oriented time Mean bout length of oriented time 1.00 .83 Bites per bout of oriented time 1.00 28 Not a l l of the measures used i n the c o r r e l a t i o n matrix can be regarded as independent from each other. Thus i n the group, t o t a l oriented time, number of bouts of oriented time and the mean bout length of oriented time, the f i r s t two were derived from the raw data and the t h i r d was ca l c u l a t e d . Only two of these three can be used to give independent information (Cane 1961). But as Mertz and Barlow (1966) have shown, such derived data can give a d d i t i o n a l indices of temporal changes i n the i n t e r - r e l a t i o n s h i p s between the measures from which they are derived. An example of r e l a t i o n s h i p s between three i n t e r - r e l a t e d v ariables i s seen i n F i g . 3. With an increase i n t o t a l oriented time, the number of bouts of oriented time f i r s t increased, then decreased. The mean bout length of oriented time increased exponentially whilst the mean bout length of non-oriented time decreased. C o r r e l a t i o n matrices f o r each day of the parental period were c a l c u l a t e d . Each measure was represented by twelve values (Table IV). Stable s i g n i f i c a n t p o s i t i v e c o r r e l a t i o n s were found between: i ) Bites per f i v e minutes and charges per f i v e minutes. i i ) Bites per f i v e minutes and the proportion of charges ending i n a b i t e . i i i ) Bites per f i v e minutes and b i t e s per minute of oriented time. Figure 3 Relationship between t o t a l oriented time, number of bouts of oriented time, mean bout length of oriented time and mean bout length of non-oriented time. Data from tube t e s t s . With an increase i n t o t a l oriented time there was an increase i n the mean bout length of oriented time ( A ) , a decrease i n the mean bout length of non-oriented time ( • ), the number of bouts of oriented time ( • ) f i r s t increased then decreased. H o —* o =1 CL 3 0) o Mean bout length of oriented time (sec ) Number of bouts of oriented time o o O O CO O > to o > ro ro (0 o OJ w to O > & to o > a CO O > a a) ci co o > a CO O > a 55 co co o > a co (0 co o > a ro ro o o co o > a to ro co o > a ro fo CO o > a ro ro Ol-W co o > a r ro CD O ro ro Ol Ol CO o ' a > ro ro co O a > ro to ^ (0 o • > ro ro oo oo CD O • ro ro co <£ co o a * ro o o o o o O 13 9 •9 • 9 9 9 9 G> m OJ ro CD Mean bout length of non oriented time (sec ) Table IV Correlations between selected measures f o r each day of the parental period N=12. Only c o r r e l a t i o n s s i g n i f i c a n t at the 5 % l e v e l are included. Days a f t e r f e r t i l i z a t i o n Measure 1 2 3 - 4 5 6 7 8 9 10 11 Bites per 5 min:charges .89 .85 .84 .82 .74 .73 .82 .76 .84 .81 per 5 min Bites per 5 min:prop. of .84 .76 .84 .86 .84 .95 .91 .94 .81 .64 charges ending i n a b i t e Bites per 5 min:bites per .75 .89 .89 .83 .88 .83 .84 .81 .87 .84 .74 min of oriented time Bites per 5 min:bites per .62 .64 .66 .75 .69 .64 bout of oriented time Bites per 5 min:total oriented time Bites per 5 min:mean bout length of oriented time Charges per 5 min:prop. of charges ending i n a b i t e Charges per 5 min:total .58 * »56 .59 oriented time Charges per 5 min:mean bout .60 length of oriented time Bites per min of oriented time:total oriented time -.75 Table IV cont Measures 1 2 3 Bites per min of oriented t i m e : t o t a l oriented time Bites per min of oriented time:mean bout length of oriented time T o t a l oriented time:bites .68 .66 .62 per bout of oriented time T o t a l oriented time:mean .82 .94 .71 bout length of oriented time Mean bout length of oriented .79 .72 .98 time:bites per bout of oriented time Days a f t e r f e r t i l i z a t i o n 4 5 6 7 8 9 10 11 12 60 .64 .70 .62 .57 .70 .71 .78 69 .72 .83 .68 .77 .92 .84 .68 .83 98 .91 .75 .94 .58 .82 .98 .90 ro 33 i v ) Total oriented time and b i t e s per bout of oriented time. v) To t a l oriented time and mean bout length of oriented time. v i ) Mean bout length of oriented time and b i t e s per bout of oriented time. Sim i l a r r e l a t i o n s h i p s were found when the c o r r e l a t i o n s between measures were calcul a t e d f o r i n d i v i d u a l f i s h (Table V). In some cases the c o r r e l a t i o n s between measures f o r i n d i v i d u a l f i s h seemed dependent on the mean l e v e l of one of the measures taken over the whole parental period. This i s most c l e a r l y i l l u s t r a t e d by the r e l a t i o n s h i p between r (bites per f i v e minutes: t o t a l oriented time) and mean t o t a l oriented time f o r i n d i v i d u a l f i s h ( F i g . 4). Fish which had a higher l e v e l of t o t a l oriented time over the parental period tended to have low and negative values for r . Other a c t i v i t i e s recorded during the tests occurred too r a r e l y to be included i n the above an a l y s i s . Zig-zagging was seen at low mean frequencies up to f i v e days a f t e r f e r t i l -i z a t i o n . Head-down threats occurred at a very low mean f r e -quency and pe r s i s t e d throughout the parental period. There were small negative c o r r e l a t i o n s between the frequency of head down threats and i ) b i t e s per f i v e minutes (r = -0.19) i i ) proportion of charges ending i n a b i t e (r = -0.31). Table V Correlations between selected measures f o r i n d i v i d u a l f i s h over the parental period.N=12.Only c o r r e l a t i o n s s i g n i f i c a n t at the 5 % l e v e l included. 3 .67 .94 .92 .98 Measures 1 2 Bites per 5 minrcharges .91 .71 per 5 min Bites per 5 minrprop. of .95 charges ending i n a b i t e Bites per 5 minrbites per .81 .95 min of oriented time Bites per 5 min:bites per bout of oriented time Bites per 5 min:total -.60 oriented time Bites per 5 min:mean bout length of oriented time Charges per 5 min:prop, of charges ending i n a b i t e Charges per 5 min:total oriented time Charges per 5 min:mean bout length of oriented time Bites per min of oriented -.81 -.56 time:total oriented time 4 .86 .83 Individu a l males. 6 7 8 5 ,68 74 .90 .85 .86 .94 .98 .76 .76 .59 .55 .91 .91 .66 ,88 64 9 .86 .97 .95 .93 .97 .97 10 .91 .98 .98 ,66 .83 -.61 11 .96 .71 .88 .56 .72 12 .97 .73 .89 .63 .89 .87 .85 -.69 Cont. Table V cont. Measures Bites per min of oriented time:bites per bout of oriented time Bites per min of oriented time:mean bout length of oriented time Total oriented time:bites per bout of oriented time T o t a l oriented time:mean bout length of oriented time Mean bout length of oriented time:bites per bout of oriented time 1 2 3 .94 .64 .72 .91 .84 Individu a l males 4 5 6 7 8 9 10 11 12 .62 .77 .93 -.56 -.66 70 .94 .91 .88 .73 .68 .73 .75 .99 .94 .90 .67 .56 .86 .66 93 .99 .90 .65 .61 .87 .95 Figure 4 Plot of the c o r r e l a t i o n c o e f f i c i e n t , r, between bi t e s per 5 min and t o t a l oriented time f o r i n d -i v i d u a l f i s h against the mean l e v e l of t o t a l oriented time f o r each f i s h over the twelve day-parental period. FIGURE 4 e 140 160 180 200 220 ^40 260 280 300 Mean total oriented time (sec ) o 38 Summary Experiment l a suggested that over the parental period the frequency of b i t i n g by the males f i r s t decreased then increased. There was no comparable change i n the t o t a l time spent oriented towards the f i s h i n the tube, except f o r those f i s h that, over the whole period, spent a less than average time oriented towards the f i s h . The important changes took place i n the rate of b i t i n g within the oriented period. This i s r e f l e c t e d i n the high c o r r e l a t i o n between b i t e s per f i v e minutes and bit e s per minute of oriented time. Charges per f i v e minutes also showed a U-shaped trend over the parental period. Even i f the proportion of charges ending i n a b i t e had remained constant t h i s would have resulted i n a U-shape trend i n b i t e s per f i v e minutes. However, the proportion of charges ending i n a b i t e also showed a decrease towards day eight of the cycle and a subsequent increase. The small sample sizes severely r e s t r i c t the value of the c o r r e l a t i o n c o e f f i c i e n t s f o r i n d i v i d u a l days and f o r i n d i v i d u a l f i s h . A- number of the r e l a t i o n s h i p s between measures, however, were stable over most of the twelve day period i n spi t e of the changes i n l e v e l s taking place. I t i s also true that i n d i v i d u a l f i s h showed s i m i l a r r e l a t i o n s h i p s with the l i m i t a t i o n that i n some cases the r e l a t i o n s h i p was dependent on the l e v e l of one or other of the measures. 39 C. Experiment lb Changes over the parental period measured using p a r t i t i o n tests'. Eleven males, each separated from another male by a transparent glass p a r t i t i o n , were allowed to f e r t i l i z e one c l u t c h of eggs each. Daily recordings of the behaviour of the parental male and i t s partner were made f o r twelve days a f t e r the day of f e r t i l i z a t i o n . The same analysis used i n experiment l a were used i n t h i s experiment. They were c a r r i e d out on untransformed data, although f o r some of the measures the variance was not homogenous. Edwards (1960) states that where the number of observations i s the same fo r various treatments, the F te s t f o r the means i n the analysis of variance i s l i t t l e influenced by heterogeneity of variance. Results Table VI gives the r e s u l t s of the trend a n a l y s i s . For those measures that showed a s i g n i f i c a n t trend, the quadratic component of the trend was dominant. For t o t a l oriented time, b i t e s per bout of oriented time and proportion of charges ending i n a b i t e , the quadratic trend was the only component to y i e l d s i g n i f i c a n t F - r a t i o s . Bites per f i v e minutes and charges per f i v e minutes had a highly s i g n i f i c a n t l i n e a r trend, whilst the l i n e a r trend i n b i t e s per minute of oriented time approached s i g n i f i c a n c e at the 5 % l e v e l . None of the s i g n i f i c a n t trends had a cubic component. The 40 Table VI F - r a t i o s and p r o b a b i l i t y values obtained by trend analysis of selected measures over the twelve day parental period. Experimental males tested against a mature male i n continual v i s u a l contact.N=ll. Measure Bites per 5 min Charges per 5 min Bites per min of oriented time Bites per bout of oriented time T o t a l oriented time Number of bouts of oriented time Source of sums of squares Overall trend l i n e a r component quadratic component other trends Overall trend l i n e a r component quadratic component other trends Overall trend l i n e a r component quadratic component other trends Overall trend l i n e a r component quadratic component other trends Overall trend l i n e a r component quadratic component other trends Overall trend l i n e a r component quadratic component other trends F P 4.39 <.005 9.42 <.005 34.7 <;.005 <:i 5.04 <.005 11.1 <.005 39.7 <;.005 -s:i 2.60 <;.01 3 . 72 <. 1 14.6 <.005 1.15 r>.25 <.01 >.l <:.005 2.68 <.01 1.75 > . l 22.1 <.005 ^ 1 3.59 <-.005 <1 31.5 ^.005 ^ 1 4.41 *c.005 4.68 <.05 36.2 sC-005 <1 Proportion of charges ending i n a b i t e Overall trend 2.51 li n e a r component 1.82 quadratic component 12.6 other trends 1.47 Mean bout length of Overall trend oriented time <1 41 U-shaped trend suggested by the strong quadratic component i n the trends i s seen i n P i g . 5. Daily means were compared using Duncan's New Multiple Range Test. This confirmed that the minimum values f o r measures which showed a quadratic trend, were s i g n i f i c a n t l y d i f f e r e n t from values f o r days at the beginning and end of the twelve day period. The c o r r e l a t i o n matrices f o r a l l recordings (Table VII), f o r i n d i v i d u a l days (Table VIII) and f o r i n d i v i d u a l f i s h (Table IX) showed that i n addition to s i g n i f i c a n t c o r r e l a t i o n s s i m i l a r to those found i n experiment l a , b i t e s per f i v e minutes was p o s i t i v e l y correlated with t o t a l oriented time and the number of bouts of oriented time. The r e l a t i o n s h i p s between t o t a l oriented time, number of bouts of oriented time, mean bout length of oriented time and mean bout length of non-oriented time were somewhat d i f f e r e n t from those found i n the tube t e s t ( c f . F i g . 6 with F i g . 3). Table VII also shows that the t o t a l oriented time of the partner f i s h was p o s i t i v e l y correlated with t o t a l oriented time, number of bouts of oriented time and mean bout length of oriented time of the experimental f i s h . Thus, not s u r p r i s -i n g l y , the behaviour of the experimental f i s h was not independ-ent of the behaviour of the partner. However, the measures f o r the partner d i d not follow the c l e a r U-shaped trend seen i n measures from the parental males. There was some tendency f o r Figure 5 Changes i n the l e v e l of selected measures during the reproductive cy c l e , tests using the p a r t i t i o n design. Length of open bar i s two standard e r r o r s , the v e r t i c a l l i n e gives the range. P.N., pre-nest phase; E.N., empty-nest phase; numbered days are days a f t e r f e r t i l i z a t i o n of one cl u t c h of eggs. o . o Bites per min of oriented time — ro ci u m s en to o o o o o o o o o o . o o •BB--B8-- B B -Proportion of charges ending in a bite O — r o w * ui cn N 00 <D O Ul CO -BB-- B B -Bites per 5 min _ — ro u * en o> s c o i f l o O O Q O O O O O O O O o o >< ro u en CO o L~cF=h 1—ft—4  -E3E3 -SB • o O <Z 3D m Charges per 5 min c n ro <fe o> O O O O 8 N A m a o o o o -4 z m z -GB-cn -F=g=)--4—•—I-EE-— B B -H3B -E3B-FIGURE 5 cont. -~ 200 o 0 I L_1_J PN EN I 2 3 4 5 6 7 8 9 10 II 12 Days 11.0 o L l - A — i  PN EN I 2 3 4 5 6 7 8 9 10 II 12 Days 40 PN EN I 2 3 4 5 6 7 8 9 10 II 12 Days 5.6 * [3 H PN EN I 2 3 4 5 6 7 8 9 10 II 12 Days 45 Table VII Correlations between selected measures calcul a t e d using a l l recordings from the parental period. Data from the p a r t i t i o n tests.N=132. Only c o r r e l a t i o n s s i g n i f i c a n t at the 5 % l e v e l are included. 0J 4) e £ c •H o MH •H •H w <u -P m Xl <u O -P a 6 0J -P 0 W -P £ •H 13 -P cn-H -P *d W e IT) <U c <u 3 0) C -P 3 OJ (D >H -P •H s O £ <U O £ •P 4H LO C £ -H £t -H rH X) £t "H C u OJ -P •P <D -P QJ S-i u a d •H SH MH +> -P U •rl QJ Q) M 0 T3 S-I c a, w 0 o a, a) 0J o a> O -P cn -p in -P -p S-I 10 a> • c rH w c OJ C SH W C rH rt) <D m <U (U ,Q CD C 0 0 QJ (0 a •P fd O T3 -P -P -H £ -H rS -P -H -P •H xt 0 •H M 3 U <U 4H •rl M O HH CO u EH CQ O S O 2 O CQ O EH O 00 .91 .66 .74 .70 .77 .77 1.00 .45 .86 .54 ..84 .64 1.00 .37 .81 .43 .70 1 .00 .26 .81 .36 .49 .36 1.00 .44 .72 1.00 .34 .40 1.00 .40 .40 Bites per 5 min Charges per 5 min Prop, of charges ending i n a b i t e T o t a l oriented time Bites per min of oriented time Number of bouts of oriented time Mean bout length of oriented time Bites per bout of 1.00 oriented time T o t a l oriented time of partner f i s h 1.00 Table VIII Correlations between selected measures f o r each day of the parental period.N=ll.Only values s i g n i f i c a n t at 5 % l e v e l are included. Days a f t e r f e r t i l i z a t i o n Measures 1 2 3 4 5 6 7 8 9 10 11 12 Bites per 5 min:charges .90 .85 .90 .75 .96 .91 .90 .97 .96 .89 .89 .83 per 5 min Bites per 5 min:prop. .88 .61 .76 .73 .83 .76 .70 .67 .58 .56 of charges ending i n a b i t e Bites per 5 min:bites .85 .54 .71 .78 .72 .76 .84 .87 .55 .81 .55 per min of oriented time Bites per 5 min:bites per .88 .83 .90 .81 .76 .75 .81 .92 .91 .76 .82 bout of oriented time Bites per 5 min:total .70 .62 .72 .74 .78 .82 .86 .79 .73 .76 oriented time Bites per 5 min:mea.n bout length of oriented time Charges per 5 min:prop. .73 .52 .72 .65 of charges ending i n a b i t e Charges per 5 min:total .84 .89 .58 .75 .80 .83 .85 .90 .87 .96 .83 .95 oriented time Charges per 5 min:mean .59 .54 bout length of oriented time Cont. Table VIII cont. Measures 1 2 Bites per min of oriented time:total oriented time Bites per min of oriented .75 .66 time:bites per bout of oriented time Bites per min of oriented time:mean bout length of oriented time Total oriented time:bites .59 per bout of oriented time Tot a l oriented tirae:mean .75 .66 bout length of oriented time Mean bout length of oriented .64 time:bites per bout of oriented time 3 Days a f t e r f e r t i l i z a t i o n 4 5 6 7 8 9 .65 .63 10 11 74 .91 .79 .89 .81 .94 .86 .81 .53 .53 .80 .85 .54 .63 77 .65 .60 .60 Table IX. Correlations between selected measures f o r i n d i v i d u a l f i s h over the parental period.N=12. Only c o r r e l a t i o n s s i g n i f i c a n t at the 5 % l e v e l are included. Individual f i s h Measures A - B C D E F G H I J K Bites per 5 rain:charges .97 .95 .88 .96 .90 .94 .99 .99 .89 .94 .93 per 5 min Bites per 5 min:prop. of .91 .67 .53 .80 .65 .58 .56 .81 .79 .56 charges ending i n a b i t e Bites per 5 min:bites per .89 .81 .66 .72 .84 .61 .54 .73 .79 .57 min of oriented time Bites per 5 minrbites per .87 .81 .70 .94 .75 .94 .89 .86 .84 .85 bout of oriented time Bites per 5 min:total .94 .87 .68 .90 .63 .78 .86 .93 .55 .90 oriented time Bites per 5 minrmean bout .72 .56 length of oriented time Charges per 5 miniprop. .84 .85 .54 .52 .63 .62 of charges ending i n a b i t e Charges per 5 minrtotal .97 .92 .79 .96 .67 .92 .89 .94 .67 .68 .92 oriented time Charges per 5 minrmean bout .84 .53 length of oriented time Bites per min of oriented .75 .54 .58 time:total oriented time Cont Table IX cont. Measures Bites per min of oriented time:bites per bout of oriented time Bites per min of oriented time:mean bout length of oriented time Tot a l oriented time:bites per bout of oriented time Tot a l oriented time:mean bout length of oriented time Mean bout length of oriented time:bites per bout of oriented time Individual f i s h B C D E P G H I J K 89 .72 .87 .97 .88 .64 .76 .97 .67 .80 -.54 .66 .58 .90 .84 .67 .84 69 .85 .70 .64 .76 .68 .79 .74 M3 Figure 6 Relationship between t o t a l oriented time, number of bouts of oriented time, mean bout length of oriented time and mean bout length of non-oriented time,, Data from p a r t i t i o n t e s t s . With an increase i n t o t a l oriented time there i s l i t t l e change i n the mean bout length of oriented time ( A ) , the mean bout length of non-oriented time ( • ) decreased and the number of bouts of oriented time ( • ) increased l i n e a r l y . Mean bout length of non oriented time (sec ) Number of bouts of oriented time o Q CD — • (X) CL 3 <x> in O o to o to o ro ro co o ID o <o o <o o en at to o to o co a> to o to <*> to o o o (0 O (0 o ro O ro O OJ Ol o 01 Ol o a 9 a a a a a 9 a 9 a m 9 9 r> >> 9 > 3 CD c z m CD ro OJ Ol Mean bout length of oriented time 52 a number of the measures to decline over the parental period ( F i g . 7 ) . The c o r r e l a t i o n matrices f o r i n d i v i d u a l days and i n d i v i d u a l f i s h have l i m i t e d value because of the small sample s i z e . They do suggest that the r e l a t i o n s h i p s between measures remained stable over most of the parental period and that the r e l a t i o n s h i p s between measures were s i m i l a r f o r i n d i v i d u a l f i s h . Some of the f i s h showed l i t t l e r e action to the partner over the parental period so that values f o r the proportion of charges ending i n a b i t e , and mean bout length of oriented time were very v a r i a b l e . In s p i t e of changes i n measures such as b i t e s per f i v e minutes there was only one change i n the distance between the nest of the parental male and the nest of the partner male over the twelve day period. On day four of the parental cycle of male MAYSUI, the nest s i t e of i t s partner APREI was moved from a distance of 18 cm from the p a r t i t i o n to only 4 cm from the p a r t i t i o n . Two days l a t e r APREI moved the p o s i t i o n of the nest to 20 cm from the p a r t i t i o n , and 21 cm from the o r i g i n a l nest s i t e . In t h i s period the b i t e s per f i v e minutes di r e c t e d by MAYSUI at APREI were 53 on day three, 44 on day four when the nest was only 4 cm from the p a r t i t i o n , and 20 on day s i x when the nest was 20 cm from the p a r t i t i o n . The decline i n the frequency of b i t i n g was accompanied by an increase i n the distance between nests. Figure 7 Changes i n the l e v e l s of selected measures during the twelve day parental period shown by the f i s h with no clutch i n the p a r t i t i o n t e s t s . Length of open bar i s two standard e r r o r s . Numbered days are the days a f t e r f e r t i l i z a t i o n of one clu t c h by the other male. FIGURE 7 so 40 30 E . w O. 20 in CD 10 6 L J I 2 3 4 5 6 7 8 9 10 II 12 Days 60 W 50 in at • E '~- 40 X > a> c O 30 o 20 10 I 2 3 4 5 6 7 8 9 10 II 12 Days I 2 3 4 5 . 6 7 8 9 10 II 12 Days I 2 3 4 5 6 7 8 9 10 II 12 Days 55 For a l l the pairs of f i s h used i n the experiment the mean distance between nest was 32.8 - 2.0 cm. The maximum possible distance would have been 45 cm. By contrast the mean distance between nests f o r a wild population of trachurus f i s h l i v i n g i n pools i n the splash zone of a rocky c o a s t l i n e , was found to be 73.0 - 2.4 cm (Black and Wootton unpublished observations). The presence of the glass p a r t i t i o n probably allows the close spacing of nests. Head down threats and zig-zagging occurred very r a r e l y i n the p a r t i t i o n t e s t s . Summary A number of measures showed a U-shaped trend over the twelve day period. These included t o t a l oriented time, number of bouts of oriented time and b i t e s per bout of oriented time. This i s i n contrast to experiment l a i n which none of these measures were found to have a s i g n i f i c a n t trend over the parental phase. In the p a r t i t i o n arrangement there was both a decline i n the rate of b i t i n g within the oriented time over the f i r s t two-thirds of the period, and also a decline i n the time spent oriented towards the partner f i s h . As i n experiment l a , both the frequency of charging and the proportion of charges ending i n a b i t e , declined towards the day of hatching, then showed a s i g n i f i c a n t increase. 56 D. Experiment l c Changes over the parental period measured using model tests Ten males were tested with four models f o r twelve consecutive days a f t e r each male had f e r t i l i z e d one clutch of eggs. A model was presented f o r two minutes, with a one minute pause between the presentation of d i f f e r e n t models to the f i s h . The presentation sequences were randomised. A l l four models were the same shape. They were 5.5 cm long and d i f f e r e d only i n t h e i r colour patterns. Model R was a l l red. Model S was a l l s i l v e r . Model R/S was red d o r s a l l y and s i l v e r v e n t r a l l y . Model S/R was s i l v e r d o r s a l l y and red v e n t r a l l y . Response to the models varied widely from f i s h to f i s h . For some of the measures zero values were common. For t h i s reason non-parametric s t a t i s t i c s were used i n the analysis of data, A non-parametric trend analysis described by Ferguson (1965) was employed i n addition to the Friedman two-way analysis of variance (Siegel 1956). Two non-parametric multiple comparisons were used i n conjunction with the two-way an a l y s i s . These were the rank sum t e s t f o r o u t l i e r s (Thompson and Wilkie 1963) and the Wilcoxon method of multiple com-parisons (McDonald and Thompson 1967). Because the t e s t period f o r each model was only two minutes, the number of bouts of oriented time was not obtained 57 from the records. Thus the c a l c u l a t i o n of mean bout length of oriented time was precluded. Results Examples df the wide v a r i a t i o n i n the reactions of i n d i v i d u a l f i s h to a model are shown i n F i g . 8. Here the b i t e s per two minutes, zig-zags per two minutes and t o t a l oriented time to Model S/R have been graphed f o r four f i s h . Experiments l a and b had shown that there was a s i g n i f i c a n t trend i n a number of measures over the parental period. Was there a comparable trend i n the equivalent measures obtained i n the model test s ? Only the d a i l y means from the Model S/R t e s t s had shown more than one s i g n i f i c a n t c o r r e l a t i o n with the d a i l y means from the l i v e male t e s t s (See next s e c t i o n ) . And so only the r e s u l t s f o r Model S/R t e s t s were used i n the Friedman 2-way a n a l y s i s . In s p i t e of t h i s p r i o r s e l e c t i o n , none of the measures analysed showed s i g n i f i c a n t v a r i a t i o n with days (Table X). Neither of the multiple comparison techniques showed the sum of ranks f o r any one day to be s i g n i f i c a n t l y d i f f e r e n t from those f o r any other day. However, when the same measures from Model S/R tests were analysed using the non-parametric trend analysis they were a l l found to have s i g n i f i c a n t quadratic trends, i . e . the trend was U-shaped. In addition the trend f o r t o t a l oriented Figure 8 Changes i n the response of four f i s h to Model S/R over the 12 days following the f e r t i l i z a t i o n of a c l u t c h of eggs. 9 # , b i t e s per 2 min*, A A , zig-zags per 2 min; • • , t o t a l oriented time (sec per 2 min). FIGURE 8 • a I 2 3 4 5 6 7 8 9 10 II 12 I 2 3 4 5 6 7 8 9 10 II 12 Days Days 60 time had a highly s i g n i f i c a n t l i n e a r component (Table XI). The r e l a t i o n s h i p s between measures of response to the models were examined g r a p h i c a l l y . A- mean l e v e l of the Y-axis v a r i a b l e was ca l c u l a t e d f o r a given range of the X-axis v a r i a b l e ( F i g . 9). Bites per two minutes was almost l i n e a r l y r e l a t e d to charges per two minutes. The proportion of charges ending i n a b i t e was i n most cases very high,- i r r e s p e c t i v e of the frequency of charging. In contrast to the r e s u l t s obtained with the l i v e male tube t e s t s , but l i k e the r e s u l t s from the p a r t i t i o n arrangement, there was a p o s i t i v e r e l a t i o n s h i p between frequency of b i t e s i n the tes t period and t o t a l oriented time. These r e l a t i o n s h i p s between measures were la r g e l y independent of the model used. A Friedman two-way analysis of variance was used to te s t whether the reactions to the models d i f f e r e d from model to model (Table X I I ) . Results f o r four measures were tested. The measures were, b i t e s per two minutes, charges per two minutes, t o t a l oriented time and b i t e s per minute of oriented time ( F i g . 10). Only the r e s u l t s from day 1, day 4, day 8, and day 12 were used. In no cases did t h i s analysis suggest that the responses to the four d i f f e r e n t models were d i f f e r e n t . The multiple comparison analysis did not reveal any s i g n i f i c a n t d i f f e r e n c e s . Some i n d i v i d u a l f i s h did show widely d i f f e r e n t reactions to the various models. For example MAI3 I showed a high frequency of zig-zagging to Models S and S/R over the f i r s t eight days a f t e r f e r t i l i z a t i o n . Over the same period i t 61 Table X Comparison of the l e v e l of response to Model S/R over the twelve day parental period by a Friedman Two-way Analysis of Variance. Measure Test s t a t i s t i c , X 2. Bites per 2 min 7.5 >.7 Charges per 2 min 8.5 >.5 T o t a l oriented time 14.8 > - l Bites per min of 8.9 >-»5 oriented time Table XI Non-parametric trend analysis of selected measures obtained i n t e s t s using Model S/R ( s i l v e r upper-side, red lower s i d e ) . Measure Trend z-value f o r Component trend component P Bites per 2 min Linear .66 .25 Quadratic 1.99 .02 Cubic .87 .19 Charges per 2 min Linear .71 .24 Quadratic 2.57 .005 Cubic .43 .33 Bites per min of Linear .49 .31 oriented time Quadratic 2.10 .02 Cubic .73 .23 T o t a l oriented time Linear 3.49 .0002 Quadratic 2.07 .02 Cubic .33 .37 62 Figure 9 Relationships between selected measures i n the model t e s t s . The ordinate v a r i a b l e i s the mean value f o r that v a r i a b l e over a given range of the abscissa v a r i a b l e . O, Model R t e s t s ; Model S t e s t s , A Model R/S t e s t s , • •, Model S/R t e s t s , • . Mean number of bites ro 4k O o o o to o to 5 M W vC o W U o o CO O <c o at on u> o -j - J 10 o o CD a> CD O zr Q (0 to ID O (Q 5 o CO CO o (A CO o N f j to o cCO» •o<0 •o Da a e> 1 Mean proportion of changes ending in a bite co O CO o <t> o a> a o ",° 5" «>» Q W O "9 5 6 <P u> o «/> io 6 io O • O O o «a> •coo • t> o » • DCX> o » o> a t> • o oa • • •> o • > «o Mean number of bites ro O o o o • ID o p (0 o H o o fl)' a. 3 CD 3«a • > o 6 • BO O > • p a > • o. • 0 • ao • > • • DO Table XII A comparison of the l e v e l s of response to four models during the parental period by a Friedman Two-way Analysis of Variance. Days a f t e r Measures Test s t a t i s t i c f e r t i l i z a -t i o n T 1 Bites per 2 min Charges per 2 min Tot a l oriented time Bites per min of oriented time 2.46 >.3 3.72 >.2 2.28 >.5 3.15 >.3 4 Bites per 2 min Charges per 2 min Tot a l oriented time Bites per min of oriented time .69 >.8 1.26 >.7 .72 > B 8 1.29 >.7 8 Bites per 2 min Charges per 2 min T o t a l oriented time Bites per min of oriented time 2.97 >.3 1.83 >.5 1.35 7 2.43 >.3 12 Bites per 2 min Charges per 2 min Total oriented time Bites per min of oriented time 1.11 >.7 .93 >.8 1.47 >.5 1.71 >.5 Figure 10 Comparison of the l e v e l s of response to the four models on the f i r s t , fourth, eighth and twelfth day a f t e r f e r t i l i z a t i o n . V e r t i c a l l i n e give the t o t a l range. Model R (O ); Model S ( A ) ; Model R/S ( • ); Model S/R ( • ). FIGURE 10 E CM OJ CL in m 100 80 60 40 20 o OJ in OJ E T3 $> C OJ o ,o 100 80 60 40 20 8 12 Days 4 8 12 Days E. CM OJ a. 100 80 60 40 A in OJ 20 8 12 Days 50 OJ .§ 40 OJ -•— C .OJ 30 o C 20 E OJ Q. in 10 CJ m 8 12 Days 67 showed a very low frequency of zig-zagging and a r e l a t i v e l y high frequency of b i t i n g to Models R and R/S. Head-down threats were shown to a l l models. The mean frequency of threatening to Model R and Model R/S increased over the twelve day period from 0.4 and 1.9 threats per two minutes on day 1 to 7.2 and 7.7 on day 12. Threatening to Model S and Model S/R occurred at a lower mean l e v e l but again showed an increase over the l a s t few days of the parental period reaching a mean l e v e l of 2.7 and 3.1 threats per two minutes r e s p e c t i v e l y . These l e v e l s were a l l higher than those found i n the l i v e f i s h t e s t s . There was, however, wide v a r i a t i o n from f i s h to f i s h . Summary The model tests gave much more var i a b l e r e s u l t s than e i t h e r of the methods i n which the t e s t object was a l i v e f i s h . Despite t h i s , a number of measures obtained from the t e s t with Model S/R, did show a s i g n i f i c a n t U-shaped trend over the twelve day parental period. E. Comparison of Selected Measures Daily means f o r selected measures obtained from the three methods were compared. For a given measure the d a i l y means obtained by one method were plot t e d against the d a i l y means obtained by another method. Given that: a) the d a i l y means followed the same trend over the twelve day period i r r e s p e c t i v e of the te s t method, 68 b) the absolute magnitudes of the d a i l y means were the same i r r e s p e c t i v e of the t e s t method, then the p l o t would be l i n e a r and the regression l i n e would pass through the o r i g i n . The slope would be 1.0 when the comparisons were f o r methods i n which the t e s t period f o r both methods was f i v e minutes. Where values f o r d a i l y means from the model tests were pl o t t e d against values from the other methods the expected gradient was 0.4 because the t e s t period i n the model tests was only two minutes. An exception to t h i s would occur f o r a measure such as b i t e s per minute of oriented time, which was independent of the length of the te s t period. In t h i s case i n a l l comparisons a slope of 1.0 was expected. This graphical method, i n e f f e c t , produced c a l i b r a t i o n l i n e s f o r comparing one method with another. If the c o r r e l a t i o n between d a i l y means from two experiments was high, then a gradient of more then 1.0 (or 0.4) meant that the t o t a l change i n l e v e l found by one method was greater than the t o t a l change i n l e v e l found by the other method. I f the gradient was less than 1.0 (0.4) the reverse was true. Since the tube method has most often-been used i n previous studies, t h i s was used as the X-axis variable i n a l l comparisons i n v o l v i n g r e s u l t s from t h i s method. In comparisons between the p a r t i t i o n arrangement and the model t e s t s , the p a r t i t i o n design r e s u l t s were plot t e d on the X-axis. Results 69 are presented i n Tables XIII, XIV, XV. In the text, slopes are given with t h e i r standard e r r o r . i ) Bites per t e s t period There was a p o s i t i v e s i g n i f i c a n t c o r r e l a t i o n between the d a i l y means from the tube experiment with those from the p a r t i t i o n experiment. The regression was s i g n i f i c a n t l y less than 1.0 (b = 0.70 ± 0.12). Daily means from the tube experiment were also sign-i f i c a n t l y c o r r e l a t e d with d a i l y means obtained from tests using both Model R ( a l l red) and Model S/R (red underside, s i l v e r back). Both regressions had slopes less than the expected 0.4. Daily means from t e s t s using Model R gave a slope of 0.19 + 0.07 when pl o t t e d against d a i l y means from the tube method. The equivalent p l o t f o r Model S/R tests gave a slope of 0.26 + 0.06. Daily means from the p a r t i t i o n experiment were also s i g n i f i c a n t l y c o r r e l a t e d with d a i l y means from the Model S/R t e s t s . The regression slope was not s i g n i f i c a n t l y d i f f e r e n t from 0.4 (b = 0.33 •+ 0.08). i i ) Charges per t e s t period Daily means from the p a r t i t i o n experiment were sign-i f i c a n t l y c o r r e l a t e d with those from the tube experiment. The regression slope was not s i g n i f i c a n t l y d i f f e r e n t from 1.0 (b = 1.27 ± 0.26). 70 Table XIII Comparisons of d a i l y means f o r selected measures obtained i n the parental period using a l i v e male t e s t f i s h i n a tube with d a i l y means obtained f o r the same period from the p a r t i t i o n t e s t s . The regression was of the form: Y=a+bX. with d a i l y means from the p a r t i t i o n arrangement as ordinate and d a i l y means from the tube arrangement as abscissa. Measure Co r r e l a t i o n Regression Intercept F - r a t i o c o e f f i c i e n t c o e f f i c i e n t on the f o r the Y-axis a regressxon analysis Bites per 5 min Charges per 5 min Bites per min of oriented time .89 .84 .80 70 1.27 1.38 -55.2 -183 •18.6 37.3 24.3 18.2 .0001 ,0007 ,002 T o t a l oriented time .01 .03 44.1 <1 ,92 Table XIV Comparison of d a i l y means f o r selected measures obtained i n the parental period using a l i v e male t e s t f i s h i n a tube with d a i l y means obtained f o r the same period using models as t e s t objects. The regression was of the form: Y=a+bX, with d a i l y means from the model tests as the ordinate and the d a i l y means from the l i v e male t e s t as abscissa. Measure Model Co r r e l a t i o n Regression Intercept on F- r a t i o f o r P c o e f f i c i e n t c o e f f i c i e n t the Y-axis the regression r b a analysis Bites per Model R .65 .19 .17 7.15 .02 te s t period Model S .28 .10 9.4 .85 .38 Model R/S .47 .12 9.3 2.88 .12 Model S/R .80 .26 -11.1 18.1 .002 Charges per Model R .50 .17 -7.8 3.30 .1 te s t period Model S .38 .18 -10.9 1.72 .22 Model R/S .37 .10 6.3 1.56 .24 Model S/R .76 .29 -33.3 13.3 .005 Bites per Model R .49 .38 7.4 3.10 .11 min of Model S .43 .28 7.3 3.28 .16 oriented Model R/S .23 .15 15.9 .54 .49 time Model S/R .77 .68 -7.3 14.6 .003 Tot a l Model R -.10 -.10 66.1 .10 .75 oriented Model S -.52 -.74 207 3.72 .08 time Model R/S -.20 -.27 105 .43 .53 Model S/R -.49 -.70 195 3.15 .1 Zig-zags Model R .52 .16 .58 3.77 .08 per t e s t Model S .79 1.44 2.51 16.5 .002 period Model R/S .83 .27 .05 22.8 .001 Model S/R .82 .94 2.59 20.9 .001 Table XV Comparison of d a i l y means f o r selected measures obtained i n the parental period using the p a r t i t i o n design with daily"means obtained f o r the same period using models as tes t objects. The regression was of the form: Y=a+bX., with d a i l y means from model tests as the ordinate and d a i l y means from the p a r t i t i o n arrangement as abscissa. Measure Model Corre l a t i o n Regression Intercept on F- r a t i o f o r P c o e f f i c i e n t c o e f f i c i e n t the Y-axis the regression r b a Bites per Model R .55 .20 16.9 4.40 .06 t e s t period: Model S- .49 .23 15.3 3.10 .11 Model R/S .45 .15 19.7 2.51 .14 Model S7R. .80 .33 10.5 18.1 .002 Charges Model R .57 .13 16.4 4.93 .04 per t e s t Model S .40 .13 16.1 1.96 .19 period Model R/S .48 .09 20.4 2.97 .11 Model S/R .82 .21 9.4 21.8 .0009 Bites per Model R .53 .24 13.6 3.92 .07 min of Model S .46 .18 12.0 2.70 .12 oriented Model R/S' .23 .10 18.3 .66 .43 time Model S/R .79 .40 4.5 16.5 .005 Tot a l Model R .75 .32 26.7 12.9 .005 oriented Model s; .31 .19 42.5 1.06 .32 time Model R/S .69 .39 27.8 8.96 .01 Model S/R .80 .48 25.3 17.9 .002 Table XVI Comparison of d a i l y means f o r selected measures obtained i n the parental period using a l i v e , gravid female as t e s t f i s h i n a tube with d a i l y means f o r the same period obtained using models as t e s t objects. The regression was of the form: Y=a+bX, with d a i l y means from the model t e s t s as the ordinate and d a i l y means from the l i v e female tests as the abscissa. Measure. Model Co r r e l a t i o n Regression Intercept on F- r a t i o f o r P c o e f f i c i e n t c o e f f i c i e n t the Y-axis the regression r b a Bites per Model R .32 .07 16.3 1.13 .32 te s t period Model S .63 .18 6.1 6.47 .03 Model R/S .35 .07 17.7 1.44 .25 Model S/R .64 .16 6.0 6.88 .03 Charges per Model R. .11 .02 21.1 .13 .73 tes t period Model S .74 .18 -3.2 12.4 .006 Model R/S .29 .04 19.5 .91 .37 Model S/R .56 .11 5.4 4.54 .06 Bites per Model R .34 .16 16.7 1.27 .29 rain of Model S .59 .23 11.0 5.37 .04 oriented Model R/S .20 .08 19.1 .40 .54 time Model S/R .58 .30 8.5 5.04 .05 To t a l Model R .09 .04 36.1 .08 .77 oriented Model S -.27 -.18 90.2 .81 .39 time Model R/S .14 .08 29.2 .20 .67 Model S/R .06 .04 41.7 .03 .84 Zig-zags Model R .18 .01 .82 .32 .58 per t e s t Model S .90 .20 .05 40.3 .0001 Model R/S .78 .03 -.18 15.2 .003 Model S/R .70 .10 2.1 9.7 .01 74 Only d a i l y means from Model S/R t e s t s were s i g n i f i c a n t l y c o r r e l a t e d with those from the tube experiment. The regression slope was s i g n i f i c a n t l y less than 0,4, Daily means from both Model R and Model S/R tests were s i g n i f i c a n t l y c o r r e l a t e d with d a i l y means from the p a r t i t i o n experiment. The regression slopes were both s i g n i f i c a n t l y less than 0.4, i i i ) Bites per minute of oriented time S i g n i f i c a n t c o r r e l a t i o n s were found between d a i l y means from the tube experiment and the p a r t i t i o n experiment, the tube experiment and Model S/R t e s t s , and the p a r t i t i o n experiment and Model S/R t e s t s . The regression between p a r t i t i o n experiment means and the tube experiment means had a slope not s i g n i f i c a n t l y d i f f -erent from 1.0. This was also true f o r the regression of means from Model S/R tes t s against means from the tube experiment. The slope of the regression of means from Model S/R. tests against the means from the p a r t i t i o n experiment was s i g n i f i c a n t l y less than 1.0. iv ) T o t a l oriented time There were no s i g n i f i c a n t c o r r e l a t i o n s between the means f o r t o t a l oriented time obtained i n the tube experiment, and e i t h e r the means from the p a r t i t i o n experiment or means from any of the model t e s t s . 75 In contrast there were s i g n i f i c a n t c o r r e l a t i o n s between the means obtained from the p a r t i t i o n experiment and those from the t e s t s using Model R, Model R/S, and Model S/R. The regression slopes were not s i g n i f i c a n t l y d i f f e r e n t from the expected 0.4. Only comparisons between the tube experiment and the p a r t i t i o n experiments could be c a r r i e d out on the following measures. v) Proportion of charges ending i n a b i t e Daily means from the two methods were corre l a t e d , r = 0.77. The regression slope was not s i g n i f i c a n t l y d i f f e r e n t from 1.0. v i ) Bouts of oriented time The c o r r e l a t i o n between the d a i l y means from the two methods was 0.65. A p l o t of d a i l y means from the p a r t i t i o n experiment against d a i l y means from the tube experiment had a slope of 1.52, but t h i s was not s i g n i f i c a n t l y greater than 1.0. v i i ) Bites per bout of oriented time The d a i l y means f o r t h i s measure from the two methods were not co r r e l a t e d . From these r e s u l t s i t i s c l e a r that both methods using l i v e male f i s h as the tes t objects, that i s the tube method and the p a r t i t i o n method, and also the use of Model S/R, yie l d e d comparable r e s u l t s f o r a v a r i e t y of measures. 7 6 The r e s u l t s were extended by also comparing the r e s u l t s from the model tests with r e s u l t s from t e s t s i n which the t e s t f i s h i n the tube was a gravid female. Again d a i l y means for the twelve days following f e r t i l i z a t i o n were used. A comparable study using the p a r t i t i o n arrangement could not be made because of the small supply of gravid females. As Table XVI shows, both Model S and Model S/R tests yielded d a i l y means that were s i g n i f i c a n t l y c o r r e l a t e d with d a i l y means from the tube experiment using a gravid female. Only t o t a l oriented time did not show t h i s r e l a t i o n s h i p . However, i n only one case, b i t e s per minute of oriented time f o r the comparison inv o l v i n g d a i l y means from Model S, was the regression slope not s i g n i f i c a n t l y less than the expected value. Although the regression slope was less than the expected 0.4, the c o r r e l a t i o n between zig-zagging to Model S and zig-zagging to the gravid female was very high. F. Spine Raising In one respect the behaviour of the parental males tested against the models was c l e a r l y d i f f e r e n t from that of the males tested against l i v e f i s h e i t h e r i n the tube arrange-ment or i n the p a r t i t i o n arrangement. A majority of the approaches to the models were made with both dorsal and ventral spines erect. This was i r r e s p e c t i v e of the day of the parental period. No spines were erect i n a majority of charges made at a l i v e f i s h e i t h e r i n a tube or across a glass p a r t i t i o n . Up 77 to about 20 % of charges directed at a l i v e f i s h were made with the dorsal spines erect. Charges with both dorsal and ventral spines erect were much rarer (Pig. 11). G. D i v i s i o n of a c t i v i t i e s i n a test period A comparison was made of the d i s t r i b u t i o n of three groups of a c t i v i t i e s within the tes t periods f o r the three experimental arrangements. The mean percentage of the t o t a l t e s t time spent oriented towards the tes t object, i n fanning, and i n a c t i v i t i e s at the nest excluding fanning were calcul a t e d f o r each day of the cycle ( F i g . 12). In the tube method about 70 % of a t e s t period was spent oriented towards the t e s t f i s h . There was l i t t l e change i n t h i s over the twelve day period. Fanning and nest directed a c t i v i t i e s accounted f o r only a small proportion of a t e s t . In model t e s t s , the time spent oriented towards the models was between about 30 and 50 % of the t e s t . The curve was U-shaped as expected from the trend analysis on the raw scores. Fanning did occur i n model t e s t s , and reached a maximum value of about 12 % of the t o t a l t e s t i n g time eight days a f t e r f e r t i l i z a t i o n . T otal time at the nest excluding fanning decreased from a mean of 6 % per t e s t period one day af t e r f e r t i l i z a t i o n to 0 % per tes t period on day 9. Oriented time i n the p a r t i t i o n arrangement was only one t h i r d and less of the t o t a l t e s t time. Fanning, e s p e c i a l l y Figure 11 A comparison f o r the three t e s t methods of sp i n e - r a i s i n g i n charges, O — — O , tube tests*, •———#, p a r t i t i o n t e s t s ; , Model S/R t e s t s ; numbered days are days a f t e r f e r t i l i z a t i o n . Figure 12 Mean percentage of t e s t period spent i n time at the nest excluding fanning; fanning; and oriented towards the t e s t object. O o 7 tube t e s t s ; • # , p a r t i t i o n t e s t s ; •— • , model t e s t s ; numbered days are days a f t e r f e r t i l i z a t i o n ; the length of the v e r t i c a l l i n e i s two standard e r r o r s . FIGURE 12 Total oriented time I 2 3 4 5 6 7 8 9 10 II 12 Days I 2 3 4 5 6 7 8 9 10 II 12 Days 82 i n the middle of the parental cyc l e , was an important component i n the t e s t period. The curve f o r percentage of te s t time spent fanning was an inverted U-shape reaching a maximum on and around day 7. At t h i s time 30 to 35 % of a t e s t period was spent i n fanning. The curve f o r the percentage of a te s t period spent at the nest followed a comparable decline to that i n the model t e s t , though the absolute l e v e l s were somewhat higher. Fanning i s probably an a c t i v i t y e s s e n t i a l for the successful development of a c l u t c h of eggs. In the p a r t i t i o n arrangement the parental male was i n continual v i s u a l contact with the partner male. So f o r any given period of time the parental male had to devote some time to fanning, and care of the nest, less time was therefore avai l a b l e for a c t i v i t i e s d i r e c t e d towards the other male. H. Parental Fanning The presence of the other male probably has an e f f e c t on the fanning cyc l e of the parental male. Evidence f o r t h i s comes from measurements of fanning from two groups of parental males. One group, the males used i n Experiments l a and l b , were i s o l a t e d except f o r the short t e s t periods. Measurement of t o t a l fanning duration and t o t a l number of fanning bouts per 15 minutes were made d a i l y on t h i s group of f i s h some time a f t e r the t e s t period. Similar measurements were made on a group of f i s h used i n the p a r t i t i o n experiment. 83 For the f i r s t nine days a f t e r f e r t i l i z a t i o n the i s o l males fanned f o r a longer duration than did the p a r t i t i o n males ( F i g . 13a). In both cases the maximum duration was reached on or around the 6th day af t e r f e r t i l i z a t i o n . Over the f i n a l three days of the parental period (days 10,. 11, 12) there was no diffe r e n c e In t o t a l duration of fanning between the two groups of f i s h . The number of bouts of fanning i n the f i f t e e n minute period was approximately the same f o r both groups ( F i g . 13b). Hence the diffe r e n c e i n t o t a l duration must be because the i s o l a t e d f i s h had longer mean bout lengths for the f i r s t nine days of the period. F i g . 14 compares the percent frequency of mean bout length classes f o r the two groups. For p a r t i t i o n males the maximum frequencies occurred f o r mean bout length frequencies between 5.0 seconds and 19.9 seconds. For i s o l a t e d males the maximum frequencies occurred f o r mean bout length frequencies of between 10.0 - 29.9 seconds. Perhaps the most s a l i e n t point i s that the trend i n fanning follows the same pattern over the twelve day period i r r e s p e c t i v e of the group considered. Figure 13 a) Mean t o t a l duration of fanning (sec per 15 min) f o r s o l i t a r y males ;N=12; and fo r males with a partner ( p a r t i t i o n t e s ts) N=7. 0 — — o , s o l i t a r y males; 9 j , males with partner; numbered days are days a f t e r f e r t i l i z a t i o n . b) Mean number of bouts of fanning i n a 15 min period f o r s o l i t a r y males and for males with a partner. Symbols as f o r Figure a. FIGURE 13a Days FIGURE 13 b 2 3 4 5 6 7 8 9 10 II 12 Days Figure 14 Frequency d i s t r i b u t i o n f o r 5-sec classes of mean bout length of fanning f o r the f i r s t ten days a f t e r f e r t i l i z a t i o n of one clutch of eggs. Closed bars, males with partner; open bars, s o l i t a r y males. FIGURE 14 0 5 10 15 20 25 30 35 40 45 50 55 > 6 0 4-9 9 9 14 9 19 9 24 9 29 9 34-9 39 9 44 9 49 9 54-9 59 9 Mean bout length 88 I. Experiment 2 Changes i n measures i n the 24 hours a f t e r f e r t i l i z a t i o n In the previous experiment a l l the recordings were taken at d a i l y i n t e r v a l s i r r e s p e c t i v e of the t e s t method. The three methods were also used to study changes i n measures i n the f i r s t 24 hours a f t e r f e r t i l i z a t i o n . F i s h were allowed to f e r t i l i z e one clu t c h of eggs. Then tests, were started at 1 minute, 15 minutes, 1 hour, 3 hours, and 24 hours a f t e r f e r t i l i z a t i o n . Each f i s h was tested with only one of the three methods. Only Model S/R was used i n the model t e s t s . In some cases i t was not possible to obtain a complete set of recordings from a f i s h . This led to unequal sample si z e s (Table XVII) f o r various times a f t e r f e r t i l i z a t i o n , which made s t a t i s t i c a l analysis of the data unfeasable. Graphing the r e s u l t s d i d , however, i n d i c a t e differences and s i m i l a r i t i e s i n the r e s u l t s from the three methods. Results i ) Bites per t e s t period ( F i g . 15a) B i t i n g i n both the tube experiment and the p a r t i t i o n experiment declined from a peak value immediately a f t e r f e r t i l -i z a t i o n to a minimum 3 to 24 hours a f t e r f e r t i l i z a t i o n . B i t i n g at Model S/R showed l i t t l e o v e r a l l change i n t h i s period. 89 Table XVII Sample sizes used i n experiment 2 TEST METHOD Tube P a r t i t i o n Model S/R TIME OF RECORDING Empty-nest 1 min 15 min 1 hr 3 hr 24 hr 25 18 10 10 11 10 12 8 10 23 16 90 i i ) T o t a l oriented time ( F i g . 15b) This measure increased from 1 minute a f t e r f e r t i l -i z a t i o n to a steady l e v e l 3 - 2 4 hours a f t e r f e r t i l i z a t i o n i n both the tube and the model t e s t s . In the p a r t i t i o n recordings there was a reverse trend i n t h i s measure. Total oriented time towards the f i s h across the p a r t i t i o n was a maximum one minute a f t e r f e r t i l i z a t i o n and reached a stable l e v e l one hour a f t e r f e r t i l i z a t i o n . i i i ) Bites: per minute of oriented time (Fig. 15c) This measure showed almost an i d e n t i c a l trend of change whether obtained from the tube method or the p a r t i t i o n method. A maximum was found i n the 1 minute and 15 minute tests and there was a steady decline to l e v e l s found 3 hours to 24 hours; a f t e r f e r t i l i z a t i o n . No comparable changes were found when the b i t e s per minute of oriented time was obtained from the model t e s t s . i v ) Proportion of charges ending i n a b i t e ( Fig. 15d) A more marked decrease i n t h i s measure was seen i n the p a r t i t i o n experiment than i n the tube experiment. v) Zig-zags (Fig. 15e) A l l three methods gave a maximum f o r zig-zagging 3 hours a f t e r f e r t i l i z a t i o n . v i ) Time at Nest ( F i g . 15f) This was a maximum 1 minute a f t e r f e r t i l i z a t i o n , and 91 a minimum 3 to 24 hours a f t e r f e r t i l i z a t i o n i r r e s p e c t i v e of the method of t e s t i n g . v i i ) Spine-raising ( F i g . 16) In one respect there was a close s i m i l a r i t y i n the behaviour of experimental f i s h e i t h e r towards a f i s h i n a tube or towards a partner f i s h . I n i t i a l l y a majority of charges were made with dorsal spines erect. However, the proportion of charges with dorsal spines erect f e l l s t e a d i l y the further from the time of f e r t i l i z a t i o n the recordings were made. Charges with both dorsal and ventral spines erect were rare and did not follow a c l e a r pattern of change. Charges towards Model S/R also showed a decrease i n the proportion with dorsal spines alone erect. But the proportion of charges with both dorsal and ventral spines erect increased s t e a d i l y from the f i r s t recording. Summary The tube and p a r t i t i o n tests both yielded comparable changes i n b i t i n g , charging, time at the nest and zig-zagging i n the 24 hours a f t e r f e r t i l i z a t i o n . Model S/R tests yielded comparable changes f o r only zig-zagging and time at the nest. Figure 15 Changes i n selected measures i n the f i r s t 24 hours a f t e r the f e r t i l i z a t i o n of one clu t c h of eggs. O O, tube t e s t s ; • • p a r t i t i o n t e s t s ; • • , Model S/R t e s t s ; the time at which the te s t was started a f t e r f e r t i l i z a t i o n i s given on the abscissa. Model tests were 2 min long. Tube and p a r t i t i o n tests were 5 min long. FIGURE 15 93 Figure 16 Changes i n sp i n e - r a i s i n g i n the f i r s t 24 hours a f t e r the f e r t i l i z a t i o n of one cl u t c h of eggs. Data from tube and p a r t i t i o n t e s t s . O o, tube t e s t s ; p a r t i t i o n t e s t s ; time at which t e s t was started a f t e r f e r t i l i z a t i o n i s given on the abscissa. FIGURE 16 No spines erect T3 O 'v-OJ CL to OJ CD Q. CO OJ O> k. O sz O 220 200 180 160 140 120 100 80 60 40 EN I min I5mins I hr 3 h r 2 4 hrs Time CO OJ O» v. O X: O O O OJ O C OJ O l _ cu 0. 90 80 70 60 50 H - 40 30 20 10 EN Imin I5mins I hr 3 h r 2 4 h r Time 05 OJ O> v. O 1Z O OJ O C OJ O if 90 80 70 60 50 40 30 20 10 0 Dorsal spines erect Dorsal and ventral spines erect CO OJ D> c D sz . CJ O O OJ O> O C OJ O C_ CD 0_ 9 8 7 6 5 4 3 2 I 0 L EN Imin I5min I hr 3 hr 2 4 hr Time EN Imin I5min I hr 3 hr 2 4 hr Time PART IV: THE PRE-EERTILIZATION PERIOD 96 Ao Introduction In the previous section changes i n behaviour within the parental period were considered. Many of the main elements of behaviour seen i n the parental period appear at the end of the pre-nest phase and during the empty-nest phase. Most important of these behaviour patterns are the nest directed a c t i v i t i e s such as fanning, pushing and glueing. Courtship behaviour such as the zig-zag dance and leading the female to the nest also appears. In view of these major changes, the behaviour of the experimental males under the three t e s t conditions was recorded i n both the pre-nest and empty-nest phases. Pre-nest phase i n t h i s study was regarded as the period from the introduction of the f i s h i n t o the experimental tank to the onset of nest-building behaviour at the s i t e of the f i r s t completed nest. A period of a few days was allowed to pass a f t e r the introduction of an experimental male int o a tank before the f i r s t t e s t was made.. This was to allow some time f o r the f i s h to become accustomed to i t s surroundings. I t has been assumed that any recordings made af t e r t h i s time i n the pre-nest phase were t y p i c a l of the phase as a whole. This same assumption was also made with regard to recordings from the empty-nest phase. This period extends from the completion of the f i r s t nest to the f e r t i l i z a t i o n of a c l u t c h of eggs. 97 The r a t i o n a l e f o r these assumptions was that the length of both phases, i n p a r t i c u l a r the empty-nest phase, was unpredictable. E i t h e r recordings could be made at regular i n t e r v a l s , or one or two recordings could be regarded as t y p i c a l of the phase as a whole. The former method was time consuming because of the amount of recording to be transcribed and analysed. This method was used i n the preliminary studies and was: found to be too time consuming. These preliminary studies showed that the assumption was not unreasonable. But since the assumption was not experimentally tested, more emphasis i n t h i s thesis has been placed on r e s u l t s from the parental period, where recordings were made at regular i n t e r v a l s . This i s why the r e s u l t s are presented i n a reversed sequence with the parental period being considered before nest-building. The following questions were asked. For any one test method, what changes took place between the pre-nest and empty-nest phases? For any one t e s t method, what were the r e l a t i o n s h i p s between selected measures i n the two phases? Did a r e l a t i o n s h i p change from one phase to the next? How do the r e s u l t s of one t e s t method compare with the r e s u l t s obtained by another method? B. Experiment 3 A comparison of the pre-nest and empty- nest phases The r e s u l t s of Symons (1965) and preliminary studies suggested hypothesis that the mean l e v e l f o r variables such as frequency of b i t i n g , frequency of charging, t o t a l oriented 98 time etc. would be s i g n i f i c a n t l y higher i n the empty-nest phase than i n the pre-nest phase. This hypothesis was tested by means of a one-tailed t - tes t (a = 0.05) f o r the tube and p a r t i t i o n methods. The non-parametric Mann-Whitney U-test (Siegel 1956) was used f o r the r e s u l t s from the model t e s t s . Results from the Tube Method Bites per 5 minutes, charges per 5 minutes and b i t e s per minute of oriented time were a l l s i g n i f i c a n t l y higher i n the empty-nest phase than i n the pre-nest phase. There were no s i g n i f i c a n t differences i n e i t h e r t o t a l oriented time or i n the proportion of charges ending i n a b i t e (Table XVIII). Zig-zags occur, though at a low mean frequency, i n the empty-nest phase. Nest dir e c t e d a c t i v i t i e s did not occur during the tests i n the empty-nest phase. Correlations between measures were calcula t e d f o r both the pre-nest and empty-nest phases. The outstanding di f f e r e n c e between these c o r r e l a t i o n s and those f o r the parental period was that i n both the pre-nest and empty-nest phases, t o t a l oriented time was s i g n i f i c a n t l y c orrelated with b i t e s per 5 minutes. In only two cases were the c o r r e l a t i o n s from the two phases s i g n i f i c a n t l y d i f f e r e n t . These were the c o r r e l a t i o n between b i t e s per 5 minutes and b i t e s per minute of oriented time, and between charges per 5 minutes and the proportion of charges ending i n a b i t e (Table XIX). 99 Table XVIII Comparison of l e v e l s of selected measures from pre-nest and empty-nest phases. Data from the tube t e s t s . H o : P pre-nes empty-nest H l : > Jpre-nest < ^empty-nest Measure Mean ± S,E. fo r pre-phase N = 22 Mean - S.E. fo r empty phase. N = 25 (one-tailed test) Bites per 5 min 123 — 11,9 Charges per 5 min 150 - 12.3 Prop, of charges .79 - .03 ending i n a b i t e Bites per min of 31.7- 2.6 oriented time Bites per bout 6.0 - 1.4 of oriented time T o t a l oriented 224 i 9.6 time (sec) Number of bouts 30.7? 2.7 of oriented time Mean bout length 9.9 - 2.0 of oriented time Zig-zags per 5 0 min 155 - 9.3 -2.154 .035 203 - 10.8 -3.23 .002 76 - .02 difference not i n the d i r e c t i o n predicted 39.9 - 1.8 -2.67 5.6 - 1.0 232 - 6.9 8.2 - 1.7 5.6 i 1.8 .01 difference not i n the d i r e c t i o n predicted -.696 36.9- 2.7 -1.606 .49 7 ,11 difference not i n the d i r e c t i o n predicted not c a l c u l a t e d Table XIX Correlations between selected me empty-nest phases. Data from the tube t e s t s . Measures C o r r e l a t i o n f o r pre-nest phase N=22 Bites per 5 min : charges .95 per 5 min Bites per 5 min : prop, of .76 charges ending i n a b i t e Bites per 5 min : b i t e s per .96 min of oriented time Bites per 5 min : b i t e s per .56 bout of oriented time Bites per 5 min : t o t a l .75 oriented time Bites per 5 min : mean bout .38 length of oriented time Charges per 5 min : prop, of .57 charges ending i n a b i t e Charges per 5 min : t o t a l .79 time Charges per 5 min : mean bout .38 length of oriented time from f i s h i n the pre-nest and Correlation f o r Test empty-nest phase s t a t i s t i c N=25 z .88 1.63 .2>P>.1 .40 1.87 .1>P>.05 .86 2.24 P=.03 .63 -.37 • P>.7 .66 .51 P>.6 .45 -.25 P>.8 -.07 2.49 P<.01 .69 .68 P>.4 .40 -.08 P>.9 Cont. o o Table XIX cont Measures Co r r e l a t i o n f o r pre-nest phase N=22 Bites per min of oriented .58 time : t o t a l oriented time Bites per min of oriented .45 time : bit e s per bout of oriented time Bites per min of oriented .25 time : mean bout length of oriented time Tot a l oriented time : bit e s .52 per bout of oriented time T o t a l oriented time : mean .47 bout length of oriented time Mean bout length of oriented .96 time : b i t e s per bout of oriented time C o r r e l a t i o n f o r Test empty-nest phase s t a t i s t i c N=25 z .20 1.48 .2>P>.1 .38 .27 P>.7 .15 .33- P>. 7 .68 -.84 P>.4 .68 -.97 P>.3 102 Thus, as i n the parental cy c l e , the change i n the mean frequency of b i t i n g (bites per 5 minutes) was a consequence of a change i n the mean rate of b i t i n g within the oriented period (bites per minute of oriented time) rather than a change i n the mean time spent oriented towards the t e s t fish„ Unlike the parental c y c l e , the proportion of charges ending i n a b i t e , remained constant whilst the frequency of b i t i n g increased. Results from the P a r t i t i o n Method With the exceptions of t o t a l oriented time and mean bout length of oriented time, a l l the measures had a higher mean l e v e l i n the empty-nest phase than i n the pre-nest phase (Table XX). But the differences were not s i g n i f i c a n t . Only the d i f f e r e n c e f o r the proportion of charges ending i n a b i t e c l o s e l y approached s i g n i f i c a n c e CP = 0.051). Unfortunately the proportion of charges ending i n a b i t e i s not a very r e l i a b l e measure i n the p a r t i t i o n experiment because i n a 5 minute period the frequency of charging i s often low. The most noticeable change was i n nest-directed a c t i v i t i e s . These were absent from the recordings made of f i s h i n the pre-nest phase. In the empty-nest phase time at the nest and, to a les s e r extent, fanning became prominent components of the recordings. Most of the c o r r e l a t i o n s between measures from the empty nest phase were not s i g n i f i c a n t l y d i f f e r e n t from the 103 Table XX Comparison of the l e v e l s of selected measures from the pre-nest and empty-nest phases. Data from p a r t i t i o n tests, H o * /^pre-nest Pempty-nest H 1 * /"'pre-nest ^empty-nest Measure Mean - S.E. for pre-nest phase N=13 Bites per 5 rain 23.9 - 6.6 Charges per 5 min Prop, of charges ending i n a b i t e 52.6 - 10.9 ,40 i .4 Bites per min of 21.3 - 3.6 oriented time Mean - S.E. f o r empty nest phase N=18 28.9 i 5.2 55.7 - 7.1 .49 i .03 25.9 ± 2.9 t P (one-tailed test) -.502 -.246 .31 .39 Bites per bout of oriented time T o t a l oriented time (sec) Number of bouts of oriented time Mean bout length of oriented time Zig-zags per 5 min Time at nest Time fanning 1.3 - 1.4 ± 14 .3 66.7 t l i . l 65.1 - 6.9 16.0 i 2.00 20.4 - 2.4 -1.324 1.676 .051 1.035 .15 -.297 .38 difference not i n d i r e c t i o n predicted 10 4.0 i .5 0 0 0 3.4 -.9 ± 1.0 diff e r e n c e not i n d i r e c t i o n predicted .45 not calcul a t e d 45.8 - 12.8 11.4 - 4.9 not cal c u l a t e d not c a l c u l a t e d 104 equivalent c o r r e l a t i o n s from the pre-nest phase. This may be a r e f l e c t i o n of an inadequate sample s i z e which meant that the di f f e r e n c e between two c o r r e l a t i o n s had to be very large to be judged s i g n i f i c a n t (Table XXI). There were differences i n the c o r r e l a t i o n s between t o t a l oriented time and both mean bout length of oriented time and b i t e s per bout of oriented time. This i s probably because nest directed a c t i v i t i e s become an important component of the non-oriented time. This may a f f e c t the p r o b a b i l i t y that a f i s h o r i e n t s towards an intruder or breaks of an encounter with an intruder. The p a r t i t i o n experiment did.not o f f e r very strong evidence i n support of the hypothesis that approach var i a b l e s increase s i g n i f i c a n t l y when a nest i s b u i l t . Results from the Model Tests The means f o r the four measures considered, b i t e s per 2 minutes, charges per 2 minutes, b i t e s per minute of oriented time, and t o t a l oriented time, were a l l higher i n the empty-nest phase than i n the pre-nest phase. This was i r r e s p e c t i v e of the model used i n the t e s t . However, i f data from a l l the f i s h , including those that did not charge the model, were included i n the ana l y s i s , no means from the empty-nest phase could be declared s i g n i f i c a n t l y higher than the. equivalent means from the pre-nest Table XXI Correlations between selected measures from f i s h i n the pre-nest and empty-nest phases. Data from the p a r t i t i o n t e s t s . Measure. Co r r e l a t i o n f o r Cor r e l a t i o n f o r Test pre-nest phase empty-nest phase s t a t i s t i c N=13 N=18 z Bites per 5 min : charges per 5 min Bites per 5 min : prop-, of charges ending i n a bite. Bites per 5 min : bit e s per min of oriented time Bites per 5 min : b i t e s per bout of oriented time Bites per 5 min : t o t a l oriented time Bites per 5 min : mean bout length of oriented time Charges per 5 min : prop, of charges ending i n a b i t e Charges per 5 min : t o t a l oriented time Charges per 5 min : mean bout length of oriented time .96 .82 .71 .92 .77 .45 .70 .77 .44 .92 .68 .78 .83 .65 -.24 .41 .84 .11 .88 .80 -.39 .98 .60 1.79 1.06 -.49 .89 > . 3 >.4 > . 6 >.3 >. 5 >.07 .. 6 Gont, o Table XXI cont. Measure C o r r e l a t i o n f o r pre-nest phase N=13 Bites per min of oriented .14 time ; t o t a l oriented time Bites per min of oriented .56 time : b i t e s per bout of oriented time Bites per min of oriented -.22 time : mean bout length of oriented time Tot a l oriented time : b i t e s .77 per bout of oriented time T o t a l oriented time : mean .84 bout length of oriented time Bites per bout of oriented .62 time : mean bout length of oriented time Corr e l a t i o n f o r Test P empty-nest phase s t a t i s t i c N=18 z .16 -.05 >.9 .77 -.95 >.3 -.31 -.68 >.4 .10 2.26 >.023 .12 2.70 >.007 .30 1.02 >.3 i—• o CT) 107 phase. The data were re-analysed with a l l zero values discarded. Discarding the zero value was an attempt to use only those f i s h that reacted to the model as though the model was an intruder. Then the following s i g n i f i c a n t differences were found. In Model S/R te s t s the mean from the empty-nest phase was higher than than from the pre-nest phase f o r bit e s per 2 minutes (P = 0.025) and charges per 2 minutes (0.05>P>0.025). In addition, the t e s t s t a t i s t i c U f o r b i t e s per minute of oriented time and t o t a l oriented time c l o s e l y approached the 5 % s i g n i f i c a n c e value. In Model S tests the means f o r b i t e s per 2 minutes and t o t a l oriented time v/ere s i g n i f i c a n t l y higher i n the d i r e c t i o n predicted, i n both cases 0.05>P>0.025. To judge whether the reactions to the four models d i f f e r e d i n ei t h e r the pre-nest or nest-empty phases a Friedman 2-way analysis of variance was c a r r i e d out on data from the two phases separately. In neither case were there any s i g n i f i c a n t differences detectable. Nor did the non-parametric multiple comparison tests show any s i g n i f i c a n t differences i n the response given to any given p a i r of models. ' The r e l a t i o n s h i p s between the four measures were s i m i l a r to those found i n the parental period ( c f . F i g . 8) In summary, the model tests suggested that there were some differences i n the l e v e l s of the approach measures con-108 sidered between the pre-nest and empty-nest phases. But any dif f e r e n c e was l a r g e l y obscured by the wide v a r i a t i o n i n the response from f i s h to f i s h . Comparison of r e s u l t s from the three methods There was. much less: agreement i n the r e s u l t s from the three methods f o r t h i s experimental than f o r experiment 1. The p a r t i t i o n method suggested that there were only minor, differences between pre-nest and empty-nest phase f i s h i n t h e i r behaviour towards the partner male. Nest directed a c t i v i t i e s became an important element i n the behaviour of the empty-nest phase male. Whereas nest dir e c t e d a c t i v i t i e s were a n e g l i g i b l e feature of recordings from e i t h e r tube or model tests performed on empty-nest f i s h . I f only means were considered, the responses to Model S, and Model S/R. suggested large quantitative differences i n behaviour between f i s h i n the pre-nest phase and f i s h i n the empty-nest phase. For - example, there was a mean l e v e l of 4.6 (7.1 i f zero values were discarded) b i t e s per 2 minutes f o r pre-nest f i s h tested with Model S/R. Equivalent tests on empty-nest f i s h yielded a mean of 15.3 (28.4 i f zero values were discarded) b i t e s per 2 minutes. This was approximately a f o u r - f o l d increase i n frequency of b i t i n g . The mean t o t a l oriented time towards Model S/R doubled from 21.9 seconds f o r pre-nest f i s h to 48.5 seconds f o r empty-nest f i s h . 109 ,. Results from the tube t e s t s indicated a d i f f e r e n c e i n the frequency of b i t i n g , but not of t h i s magnitude. There was no d i f f e r e n c e i n t o t a l oriented time i n the tube t e s t s . As i n the parental period', t o t a l oriented time i n the tube t e s t formed a much higher proportion of a t e s t period than i t did i n e i t h e r the p a r t i t i o n or model tests;. Perhaps s u r p r i s i n g l y , t o t a l oriented time i n the p a r t i t i o n t e s t s did not show a marked decrease when nest-directed a c t i v i t i e s became important. In the parental period i t had seemed that an increase i n fanning had- cut i n t o the time spent oriented towards the partner. Did the increase i n nest-directed a c t i v i t y i n the empty-nest f i s h prevent a r i s e i n t o t a l oriented time comparable to that seen i n Model S and Model S/R tests:? There was no increase i n the frequency of head-down threats i n any of the t e s t methods. Zig-zagging appeared i n records from the empty-nest phase, but i n a l l these methods the mean frequency was low. As i n the parental period most charges towards the models were, made with dorsal or dorsal and v e n t r a l spines e r e c t . This was true f o r both pre-nest and empty-nest phase f i s h . No spines were erect i n 90 % of the charges made by experimental males i n the p a r t i t i o n arrangement. Again there was no d i f f e r e n c e between f i s h i n the two phases. An i n d i v i d -ual exception to t h i s was shown by one f i s h that had a r e l a t -i v e l y high proportion of charges made with both dorsal and v e n t r a l spines erect. This f i s h had a nest and showed some 110 bouts of zig-zagging during the t e s t . The erection of the dorsal and ven t r a l spines occurred i n the zig-zag dance and they were held erect i n charges following a bout of zig-zagging. Charges made with dorsal or dorsal and ven t r a l spines erect were r e l a t i v e l y more common i n the tube t e s t s . I l l C. Experiment 4 E f f e c t of gonadectomy i n the pre-nest and empty-nest phases Introduction Some recent work on the three-spined stickleback has been concerned with the r o l e of gonadal hormones i n the reproductive cycle of the male (Hoar 1962a, 1962b, Baggerman 1966). I t i s i n such causal studies that a knowledge of the e f f e c t of the t e s t conditions on the r e s u l t s i s important. However, i t is; i n ju s t such studies that one t e s t condition has often been used without reference to other studies or to the methods of other authors. The experiment to be described i s a comparison of the r e s u l t s from the three methods of t e s t i n g when they are used to study the e f f e c t of gonadectomy on male sti c k l e b a c k s . Only f i s h i n e i t h e r the pre-nest or empty-nest phases were used. Unfortunately lack of time meant that t h i s experiment can only be regarded as a p i l o t study. I t must be c l e a r from the preceeding r e s u l t s that to obtain s a t i s f a c t o r y r e s u l t s , reasonably large sample sizes are required. This p r e - r e q u i s i t e was not achieved. Method For each method of t e s t i n g , four groups were set up. The f i r s t two groups: were pre-nest phase f i s h . Of these, one group consisted of f i s h gonadectomised before being placed i n 112 the experimental tanks, and the other group were sham operated before placement i n the experimental tanks. Other f i s h were allowed to b u i l d nests i n t h e i r tanks and a' week or more a f t e r the f i r s t nest had been completed such f i s h were eithe r gonadectomised or sham operated and replaced i n t h e i r home tank. Tube tests:, using a male 50 mm long, were c a r r i e d out 4 weeks a f t e r the operation. Some f i s h were also tested at weekly i n t e r v a l s with both males (35 mm and 50 mm long) and non-gravid females (35 mm and 50 mm long). Four models were used. Two were painted red on the ventral surface and two were painted s i l v e r on the ventral surface. The dorsal surfaces were l e f t unpainted. For each colour there were two model s i z e s , one 40 mm long, the other 55 mm long. The use of t e s t f i s h and models of d i f f e r e n t sex (colour) and s i z e was a consequence of the r e s u l t s presented i n experiment 5. In the p a r t i t i o n tests; the partner f i s h was always a male. At the end of the experiment the colour of the experimental f i s h was recorded. The body ca v i t y was examined f o r remains of t e s t i c u l a r t i s s u e . Results No s i g n i f i c a n t differences were found between the pre-nest gonadectomised group and the pre-nest sham operated group i n t h e i r responses to a male f i s h i n the pl e x i g l a s s tube (Table XXII). At the time of t e s t i n g no f i s h i n eithe r of these groups had nests. The f i s h i n the gonadectomised group 113 did not b u i l d nests i n the four month period following the operation. Nor did they develop the t y p i c a l red throat and blue i r i s e s of the mature male. The empty-nest phas:e sham-operated f i s h charged and b i t at the t e s t male i n the p l e x i g l a s s tube more often than did the empty-nest phase gonadectomised group. However, the differences were not s i g n i f i c a n t at the 5 % l e v e l using a one-t a i l e d t t e s t (Table XXIII). None of the gonadectomised f i s h zig-zagged to the t e s t male i n te s t s c a r r i e d out a f t e r the operation. Though p r i o r to the operation a l l f i s h i n these two groups had had nests f o r at l e a s t a week, four weeks a f t e r the operation none of the gonadectomised f i s h had nests. Nine of the ten sham-operated f i s h had nests. Often the nest was maintained at the same s i t e as before the operation. In appearance and behaviour the two gonadectomised groups and the pre-nest sham-operated group were very s i m i l a r (Table XXIV). This s i t u a t i o n also held f o r the behaviour of the experimental f i s h towards the wax models. Frequency of b i t i n g and frequency of zig-zagging to the models were s i m i l a r f o r the two gonadectomised groups and the pre-nest sham-operated group. In a l l cases the frequency of b i t i n g at the s i l v e r models was, higher than at the red models. The differences were not s i g n i f i c a n t . The empty-nest sham-operated group b i t 114 Table XXII Comparison of l e v e l s of selected measures between f i s h gonadectomised before experience of nest-building and sham-operated f i s h s t i l l i n pre-nest phase. Data from tests with mature male i n a tube four weeks a f t e r the operation. H o * Psham Pq gonad. H - . U , > U 1 f sham / gonad. Measure Sham-operated group N = 19 Gonadectomised group N = 14 t value P (one-t a i l e d t e st) Bites per 123 5 min Charges per 149 5 min Prop, of charges .79 ending i n a b i t e 133 160 .84 diffe r e n c e not i n d i r e c t i o n predicted d i f f e r e n c e not i n d i r e c t i o n predicted d i f f e r e n c e not i n d i r e c t i o n predicted T o t a l oriented 221 time (sec) Bites per min 32.1 of oriented time 211 36.5 0.549 n„ s, difference not i n d i r e c t i o n predicted Zig-zags per 5 min Nests present 0 0 0 0 115 Table XXIII Comparison of l e v e l s of selected measures between f i s h gonadectomised a f t e r b u i l d i n g nests and f i s h sham-operated a f t e r b u i l d i n g nests„ Data from tests with mature male i n a tube, four weeks a f t e r the operation. H o * Psham /^gonad H 1 * ^sham > /^gonad Measure Sham group Gonadectomised t value P value group (one-N 10 N 11 t a i l e d test) 77 Bites per 167 5 min Charges, per 207 5 min Prop, charges ending i n a b i t e T o t a l oriented 226 time (sec) Bites per min 41.7 of oriented time Zig-zags per 14.0 5 min Nests present 9 126 153 .75 210 34.9 0 1.21 1.42 0.19 0.54 1.23 not ca l c u l a t e d not calcul a t e d 0.15>P>0.1 0.1>P?0.05 >0.5 0.15>P>0.1 Table XXIV Comparison of selected measures obtained from tube tests on gonadectomised and sham-operated f i s h . Measure Bites per 5 min Charges per 5 min Prop, charges ending i n a b i t e Total oriented time (sec) Bites per min oriented time Zig-zags per 5 min Number of f i s h with nests Mean value f o r group gonadect-omised i n pre-nest phase N = 14 133 160 .84 211 36.5 0 0 Mean value f o r Mean value for group gonadect- group sham-omised i n empty- operated i n nest phase pre-nest phase N = 11 126 158 .76 210 34.9 0 0 N = 19 123 149 .79 22:1 32.1 0 0 Mean value f o r group sham-operated i n empty-nest phase N = 10 167 207 .77 226 41.7 14.0 117 and zig-zagged more frequently to the models than did the other three groups (Tables XXV and XXVI). Relative changes i n the frequency of b i t i n g to models and l i v e f i s h were followed f o r four weeks following gonadectomy i n the empty-nest phase. The rate of decline i n the frequency of b i t i n g was s i m i l a r f o r both s i l v e r and red models. The rate of decline i n frequency of b i t i n g i n i t i a l l y was more rapid f o r t e s t s against l i v e males than against l i v e females. B i t i n g to models declined to between 13 and 31 percent of the l e v e l p r i o r to the operation i n the four weeks following the operation. In a comparable time period b i t i n g to l i v e f i s h declined to between 48 and 74 percent of the l e v e l p r i o r to the operation. The changes were much more regular i n the model tests than i n the l i v e f i s h t e s t s ( F i g . 17). Frequency of zig-zagging had declined to near zero by a week a f t e r the operation i n both the model and l i v e f i s h t e s t s . Results from the p a r t i t i o n experiment were of very l i m i t e d value. I t was intended that there would be f i v e f i s h i n each group but a lack of healthy males r e s t r i c t e d the numbers even f u r t h e r . Such r e s u l t s as were obtained suggested that f i s h gonadectomised before they had b u i l t a nest, sham-operated f i s h i n the pre-nest phase and f i s h gonadectomised a f t e r they had b u i l t a nest were a l l v i r t u a l l y a l i k e i n t h e i r behaviour towards the male across the p a r t i t i o n (Table XXVII). Table XXV Mean number of b i t e s per 2 min to a wax model, four weeks a f t e r operation. Model s i z e and colour Group gonadect-" .Group gonadect- Group sham- Group sham-omised i n pre- omised i n empty- operated i n operated i n nest phase nest phase pre-nest empty-nest phase phase N = 7 N = 6 N = 8 N = 6 Medium s i l v e r 2.3 5.5 3.1 20.5 Large s i l v e r 3.2 9.8 6.4 23.5 Medium red 1.7 3.0 0.5 8.7 Large red 0.3 2.5 1.0 11.1 Table XXVI Mean number of zig-zags per 2 min to a wax model, four weeks afte r operation, Model s i z e and colour Group gonadect- Group gonadect- Group sham- Group sham-omised i n pre- omised i n empty operated i n operated i n nest phase nest phase pre-nest empty-nest phase phase N = 7 N = 6 M = 8 N = 6 Medium s i l v e r 0 0 0 14.1 Large s i l v e r 0 0 0 6.0 Medium red 0 0 0 3.0 Large red 0 0 0 3.0 Figure 17 The r e l a t i v e change i n the mean frequency of b i t i n g following gonadectomy i n the empty-nest phase. Le f t hand side, model t e s t ; Q O , s i l v e r model; • • , red model. Right hand side, tube t e s t s ; O — — O , female t e s t f i s h ; • • ,male t e s t f i s h . Ordinate, mean frequency of b i t i n g a f t e r gonadectomy as a percentage of the mean frequency of b i t i n g p r i o r to gonadectomy; abscissa, time i n weeks a f t e r gonadectomy. F IGURE 17 LARGE MODELS (55mm) 0 1  1 2 3 4 Weeks MEDIUM MODELS (40mm) 0 L 1 2 - 3 4 Weeks LARGE TEST FISH (55mm) 1 2 3 4 Weeks MEDIUM TEST FISH (40mm) 10 1 1 2 3 4 Week s 121 Table XXVII Comparison of selected measures obtained from p a r t i t i o n t e s t s on gonadectomised and sham-operated f i s h . Measure Mean value f o r Mean value f o r Mean value f o r group gonadect- group gonadect- group sham-omised i n pre- omised i n empty- operated i n nest phase nest phase pre-nest phase N = 5 N = 3 N = 4 Bites per 3.2 5 min Charges per 8.4 5 min Prop, of charges .44 ending i n a b i t e T o t a l oriented 18.8 time (sec) Bites per min 12.7 of oriented time 1.7 8.7 .15 18.0 5.4 5.8 11.5 .34 19.8 14.4 N.B. No f i s h sham-operated i n the empty-nest phase survived f o r more than two weeks. 122 Fi s h that were gonadectomised a f t e r they had b u i l t a nest showed no nest directed a c t i v i t i e s i n the t e s t . One male was unsuccessfully gonadectomised and was showing nest-b u i l d i n g a c t i v i t y three to four weeks a f t e r the operation. Summary These r e s u l t s confirmed that gonadectomy of the lei u r u s male prevented the expression of courtship and nest-dir e c t e d a c t i v i t i e s . The breeding colours of the male, the red throat and blue i r i s , are also dependent on the presence of t e s t i s . A l l three methods indicated that the l e v e l of b i t i n g measured i n gonadectomised f i s h was comparable to that measured i n sham-operated, pre-nest f i s h . This was i r r e s p e c t i v e of the phase the f i s h were i n when they were operated on. Results from the tube and model tests suggested that the l e v e l of b i t i n g of empty-nest males at a male i n the tube or: at a red model was more dependent on i n t a c t t e s t i s than the l e v e l of b i t i n g at a non-gravid female or s i l v e r model. There was also an o v e r a l l d i f f e r e n c e i n the response of i n t a c t empty-nest males and i n t a c t pre-nest males to a l l the models. 123 P.. Experiment 5 E f f e c t of s i z e and sex of t e s t f i s h  Introduction At the s t a r t of t h i s study i t was necessary to examine the e f f e c t of the s i z e and the sex of the t e s t f i s h i n the tube on the response of the experimental f i s h . No single t e s t f i s h survived a whole season of t e s t i n g and so the te s t f i s h had to be changed. The form of t h i s experiment was not repeated with the p a r t i t i o n design, so that the comparisons which form the main theme of t h i s thesis could not be made. However the r e s u l t s were of s u f f i c i e n t i n t e r e s t to be included at t h i s point, e s p e c i a l l y i n r e l a t i o n to the e f f e c t s of gonadectomy described i n the previous experiment. Method The same general method was used i n each year. In 1967, two groups of f i s h i n pre-nest and one group of f i s h i n the empty-nest phase were used. Of the pre-nest phase groups, one had been kept on a photoperiod of 16 hours of l i g h t and 8 hours of dark (16L 8D. ), the other group had been kept on a photoperiod of 8 hours l i g h t and 16 hours of dark (8L 16D). In 1968, one group of pre-nest phase f i s h (16L 8D) and one group of empty-nest phase f i s h were used. The t e s t f i s h consisted of a male and a non-gravid female each 35 - 40 mm long, and a male and a non-gravid female each 50 - 55 mm long. 124 Every f i s h i n each group was tested once with each of the t e s t f i s h . Tests were spaced at one day i n t e r v a l s and the order i n which the t e s t f i s h were presented to any one experimental f i s h was randomised. In the analysis the i n d i v i d u a l f i s h i n a group were treated as blocks, and analysis of variance of s i z e against sex was c a r r i e d out on a number of measures. Results The r e s u l t s of the analysis of variance are shown.in Tables XXVII, XXIX. In the three pre-nest groups, the sex.factor did not y i e l d a s i g n i f i c a n t F - r a t i o f o r any of the measures considered with the exception of proportion of charges ending i n a b i t e (1967 group 16L 8D). The s i z e f a c t o r did y i e l d a s i g n i f i c a n t F - r a t i o f o r some measures. Thes.e were, proportion of charges ending i n a b i t e (1967 group 16L 8D), bites' per minute of oriented time (1967 group 16L 8D), and t o t a l oriented time (1968 group). By contrast i n the empty-nest phase groups, the sex f a c t o r yielded a s i g n i f i c a n t F - r a t i o f o r b i t e s per 5 minutes and b i t e s per minute of oriented time. I t was also a s i g n i f -i c a n t f a c t o r f o r charges per 5 minutes i n the 1968 empty-nest group, and approached s i g n i f i c a n c e for that f a c t o r i n 1967 (P = 0.07). The major differences between the 1967 and 1968 r e s u l t s were that i n 1968 the sex f a c t o r F - r a t i o f o r t o t a l oriented time was s i g n i f i c a n t , and the sex f a c t o r F - r a t i o 125 Table XXVIII F - r a t i o and p r o b a b i l i t i e s from analyses of variance of selected measures from f i s h i n pre-nest phase. Test against males and non-gravid females i n two s i z e categories, Spring 1967 i ) 16L 8D (N=6) Measure Factor F - r a t i o P Bites per 5 min Sex <1.0 Size 3.4 0.082 Charges per 5 min Sex <1.0 Size 1.09 0.313 Proportion of charges Sex 4.56 0.048 ending i n a b i t e Size 11.26 0.004 Tota l oriented time Sex <1.0 Size <C1.0 Bites per min of Sex 2.43 0.137 oriented time Size 6.03 0.026 Spring 1967 i i ) 8L 16D (N=8) Bites per 5 min Sex <1.0 Size <1.0 Charges per 5 min Sex <1.0 Size <1.0 Proportion of charges Sex <1.0 ending i n a b i t e Size <1.0 To t a l oriented time Sex 1.58 0.221 Size 3.07 0.091 Bites per min of Sex <1.0 oriented time Size 1.16 0.29 Spring 1968 16L 8D (N=8) Bites per 5 min Sex <1.0 Size 2.35 0.107 Charges per 5 min Sex <C1.0 Size 2.34 0.137 Proportion of charges Sex <1.0 ending i n a b i t e Size 2.34 0.137 To t a l oriented time Sex 1.47 0.234 Size 5.19 0.032 Bites per min of Sex <1.0 oriented time Size <T1.0 126 Table XXIX. F - r a t i o s and p r o b a b i l i t i e s from analyses of variance of selected measures from f i s h i n the empty-nest phase. Tested against males and non-gravid females i n two s i z e categories. Spring 1967 (N=7) Measure Factor F - r a t i o P Bites per 5 min Sex 7.35 0.014 Size 4.15 0.054 Charges per 5 min Sex 3.65 0.070 Size 5.00 0.036 Proportion of charges Sex 11.05 0.004 ending i n a b i t e Size <1.0 0.51 Tota l oriented time Sex <1.0 0.72 Size 5.20 0.03 Bites per min of Sex 13.38 0.002 oriented time Size <1.0 0.425 Spring 1968 (N=10) Bites per 5 min Sex 17.17 0.004 Size <1.0 0.425 Charges per 5 min Sex 21.61 0.0001 Size 1.91 0.175 Proportion of charges Sex <1.0 0.475 ending i n a b i t e Size 1.38 0.249 To t a l oriented time Sex 9.20 0.005 Size <1.0 0.486 Bites per min of oriented time Sex Size 13.09 <1.0 0.005 127 f o r t h e p r o p o r t i o n o f c h a r g e s e n d i n g i n a b i t e was s i g n i f i c a n t i n 1967. Where t h e r e was a s i g n i f i c a n t F - r a t i o f o r t h e sex f a c t o r , t h e mean l e v e l o f t h e measure i n q u e s t i o n , was always h i g h e r f o r t h e male t e s t s t h a n f o r t h e f e m a l e t e s t s ( T a b l e XXX). I n 1967 t h e s i z e f a c t o r i n t h e empty-nest phase a n a l y s i s was s i g n i f i c a n t f o r t h e measures, c h a r g e s p e r 5 mi n u t e s and t o t a l o r i e n t e d t i m e . No s i g n i f i c a n t F - r a t i o s f o r t h e s i z e f a c t o r were o b t a i n e d i n t h e a n a l y s i s f o r t h e 1968 d a t a . A p a r a l l e l s t u d y s u b s t i t u t i n g 'male 1 and 'female' models f o r t h e t e s t f i s h , d i d n o t i n d i c a t e what cues were u s e d f o r t h e n o n - d i s c r i m i n a t i o n between males and n o n - g r a v i d f e m a l e s . The 'male' models, one 40 mm l o n g , t h e o t h e r 55 mm l o n g , were p a i n t e d r e d on t h e v e n t r a l s u r f a c e . E q u i v a l e n t • f e m a l e ' models were p a i n t e d s i l v e r on t h e v e n t r a l s u r f a c e . A l l f o u r models were l e f t u n p a i n t e d on t h e d o r s a l s u r f a c e . No s i g n i f i c a n t d i f f e r e n c e s i n t h e r e s p o n s e s t o t h e f o u r models c o u l d be d e t e c t e d i n e i t h e r t h e p r e - n e s t o r t h e empty-nest phase f i s h . The t e n t a t i v e c o n c l u s i o n i s t h a t t h e r e was a change i n t h e r e s p o n s e o f a male s t i c k l e b a c k towards males and non-g r a v i d f e m a l e s c o n s p e c i f i c s as t h e male b u i l d s and c o m p l e t e s a n e s t . Because o f t h e outcome o f t h e 1967 e x p e r i m e n t , c a r e was t a k e n t o use o n l y t e s t males o f a p p r o x i m a t e l y t h e same s i z e i n t h e t e s t s u s i n g t h e tub e method. A n o t h e r i m p o r t a n t 128 c h a r a c t e r i s t i c of the t e s t f i s h i s the amount of movement i n the tube. Although not experimentally v e r i f i e d , i t was assumed that the t e s t f i s h should be selected to show roughly the same amount of movement i n the tube. I t may well be that the movement of males and non-gravid females i n the tube i s s u f f i c i e n t l y d i f f e r e n t to account for the r e s u l t s of t h i s experiment. F i s h i n the empty-nest phase may be more s e n s i t i v e to any movement within t h e i r t e r r i t o r y i r r e s p e c t i v e of i t s o r i g i n . 129 Table XXX Means f o r the sex and s i z e f a c t o r s . Only i f the fa c t o r y i e l d e d a s i g n i f i c a n t F - r a t i o are the means given. i ) empty-nest phase f i s h (1967) Measure Bites per 5 min Charges per 5 min Prop, of charges ending i n a b i t e T o t a l oriented time (sec) Bites per min of oriented time Sex f a c t o r male female 116 146 . 7 9 37.8 81.8 116 .71 27.1 Size f a c t o r medium large 114 173 148 203 i i ) pre-nest phase f i s h (1967) Bites per 5 min Charges per 5 min Prop, of charges ending i n a b i t e T o t a l oriented time (sec) Bites per min of oriented time .62 25.8 ,45 19.5 PART V: DISCUSSION AND CONCLUSIONS 131 The purpose of t h i s thesis has been to compare selected measures of behaviour of the male three-spined stickleback obtained by three t e s t procedures at d i f f e r e n t phases i n the reproductive c y c l e . Only t h i s comparative approach can i d e n t i f y those elements of behaviour that are c l o s e l y r e l a t e d to each other, that i s those elements that are often regarded as the consequence of some unitary intervening v a r i a b l e such as a dr i v e . I t i s also the only approach that can assess- the influence of te s t procedures on the r e s u l t s obtained. Thus t h i s thesis has been concerned with four questions. ( i ) What changes take place i n behaviour over d i f f e r e n t stages of the reproductive cycle? ( i i ) What are the r e l a t i o n s h i p s between selected measures of behaviour towards a te s t object? ( i i i ) How do the r e s u l t s from the three t e s t methods compare? (iv) How do the r e s u l t s i n t h i s study compare with other studies? The f i r s t three questions provide the framework f o r the subsequent discussion. At appropriate points within t h i s framework the comparisons implied i n the fourth question can be made. ( i ) Changes i n recorded behaviour 132 Changes i n the Parental Period The dominant change was- the U-shaped (quadratic) trend shown by a number of measures i n a l l three t e s t methods. For some measures t h i s trend took place whether fanning was an important component within the te s t period or not. But i n the tube t e s t s , i n which fanning was a n e g l i g i b l e f a c t o r , t o t a l oriented time did not show the quadratic trend. Changes took place i n the rate of b i t i n g within the oriented period rather than changes i n the length of the oriented period. This was i n contrast to the r e s u l t s from both the model and p a r t i t i o n t e s t s . In these both t o t a l oriented, time and the rate of b i t i n g within the oriented time (bites per minute of oriented time) showed the quadratic trend. In both of these t e s t methods t o t a l oriented time was less than 50 % of the t e s t period and fanning was a component of the records. Frequency of charging showed the quadratic trend i r r e s p e c t i v e of the te s t method. In both the p a r t i t i o n and tube t e s t s , the proportion of charges ending i n a b i t e showed a U-shaped trend over the twelve day parental period. This was not true of the model t e s t s . In these, the experimental f i s h could make contact with the t e s t object, and the proportion of charges ending i n a b i t e was very high and did not change during an experiment. In published studies only the tube tests have been used to follow changes i n frequency of b i t i n g i n the parental 133 period. Symons (1965) presented r e s u l t s from the f i r s t s i x days of the period and Segaar (1961) tested f i s h over the f i r s t nine days. Segaar's r e s u l t s showed a c l e a r U-shaped trend i n b i t e s at a test male i n a tube 23 cm from the ex-perimental male's nest. Symons placed h i s te s t male at 10 cm, 100 cm, and 250 cm from the nest. Only f o r the two larger distances was there a c l e a r decline i n frequency of b i t i n g shown by the experimental male. Symons used a mixed population of experimental f i s h i n cluding trachurus, l e i u r u s , and hybrid forms. Segaar does not state which form he used. Probably the trend of b i t i n g during the parental cycle i s comparable f o r both trachurus and leiurus forms. A l l three methods showed that frequency of zig-zagging declined from the day a f t e r f e r t i l i z a t i o n to a very low value towards the day of hatching. In some f i s h , creeping-through, usually regarded as a behaviour c h a r a c t e r i s t i c of f i s h w i l l i n g to court a female, was seen as l a t e as eight days a f t e r f e r t i l i z a t i o n . In both the p a r t i t i o n and model t e s t s , fanning, i n terms of t o t a l duration or percentage of t e s t period spent i n fanning, showed an inverted U-pattern of change over the twelve day period. This same trend was seen i n records of fanning from s o l i t a r y f i s h with clutches (see also van l e r s e l 1953). The duration of fanning i n the i s o l a t e d males was higher than f o r the p a r t i t i o n males, though both groups 134 i n i t i a t e d fanning bouts at the same frequency. The mean bout length of fanning i n the p a r t i t i o n males was lower than f o r the i s o l a t e d males. Van den Assem (1967) found the same phenomena. He was using mixed populations of trachurus, l e i u r u s , and semiarmatus forms. Since the trend f o r fanning duration was a mirror image of the trends for a number of the approach variables there i s presumably some d i r e c t or i n d i r e c t r e l a t i o n s h i p between fanning and the measures considered here. In the tube experiments there was; a decline i n some measures as fanning increased although fanning was not i t s e l f a component of the a c t i v i t y during the t e s t . In the p a r t i t i o n t e s t s , and to a less e r extent i n the model t e s t s , fanning was a component of the t e s t period. As fanning increased, oriented time decreased. The partner f i s h i n the p a r t i t i o n experiment did not show the marked changes i n fanning c h a r a c t e r i s t i c of the parental male. Nor was there a marked U-shaped trend i n the approach measures i n the partner males over the twelve day period. Stevenster (1961) found a weak negative c o r r e l a t i o n between the parental drive (measured by t o t a l duration of fanning i n seconds per 30 minutes) and the aggressive drive (measured by b i t e s per 5 minutes using a t e s t male i n a tube). There were no co n t r o l groups of f i s h i n the tube and model t e s t s . That i s , no group of f i s h with nests but with-out clutches were tested on twelve consecutive days. The absence of such controls, while unimportant to the comparisons 135 between methods, means that causal factors i n the changes recorded cannot be i d e n t i f i e d . Results presented i n t h i s study can only be used to formulate hypotheses about these factors but cannot i d e n t i f y the f a c t o r s . In a population of mature males chasing and b i t i n g r e s u l t s i n the formation and defence of t e r r i t o r i e s . Are there changes i n t e r r i t o r i e s that can be correlated with the changes i n behaviour i n the parental period as recorded by the three methods? Morris (1958) recorded an increase i n the si z e of the t e r r i t o r y held by a male three-spined stickleback a f t e r i t s eggs had hatched, that i s , i n the period that b i t i n g and charging were increasing i n the test conditions used here. Unpublished work by R. Black (personal communication) has shown that the s i z e of the t e r r i t o r y held by a parental male decreased over the f i r s t h a l f of the parental period. Again t h i s change p a r a l l e l s a change, a decrease, i n the measures used i n t h i s study. I t i s not c l e a r which elements i n the approach and attack r e p e r t o i r e of the male change i n the t e r r i t o r i a l s i t u a t i o n . The tube experiments would suggest that the tendency to attack decreases. The p a r t i t i o n experiment suggests that not only does the tendency to attack decrease but that more time i s spent at the nest instead of p a t r o l l i n g the t e r r i t o r y . A parental male may also be reluctant to leave the stimulus of a nest f u l l of eggs i n order to approach an intruder. 136 Changes i n the 24 hours a f t e r f e r t i l i z a t i o n From a high l e v e l soon a f t e r f e r t i l i z a t i o n the frequency of b i t i n g and of charging measured by ei t h e r the p a r t i t i o n or tube tests declined to the l e v e l c h a r a c t e r i s t i c of the day af t e r f e r t i l i z a t i o n . The dominant fac t o r was the change i n bit e s per minute of oriented time. In a l l three methods zig-zagging was minimal soon a f t e r f e r t i l i z a t i o n and was a maximum three hours a f t e r f e r t i l i z a t i o n . These r e s u l t s f o r b i t i n g and zig-zagging c l o s e l y agree with those of Sevenster-Bol (1962). Also i n a l l three methods time at nest was a maximum i n the f i r s t t e s t s a f t e r f e r t i l i z a t i o n . One might expect that t o t a l oriented time would be a minimum when time at nest was a maximum. This was true i n the tube and model t e s t s , but i n the p a r t i t i o n tests maximum t o t a l oriented time and maximum time at the nest both occur i n the tes t s immediately a f t e r f e r t i l i z a t i o n . The r e s u l t s from the model tests f o r b i t i n g and charging did not agree with the r e s u l t s from the p a r t i t i o n and tube tests -. This lack of agreement was f o r tests 1 minute and. 15 minutes a f t e r f e r t i l i z a t i o n . In the tests using l i v e f i s h , the frequencies of b i t i n g and charging were at a maximum at these times. The frequencies were at a minimum i n the model t e s t s . The experimental f i s h had j u s t had contact with another l i v e f i s h , the gravid female that deposited the eggs. I t i s not known what the e f f e c t of repeated exposures to either the t e s t o b j e c t or to a free-swimming intruder i s on the l e v e l of response i n subsequent t e s t s . Exposure to a l i v e c onspecific may have a d r a s t i c e f f e c t on the response to an inanimate object presented shortly afterwards. Segaar (1961) and Sevenster (1961) have both claimed that the a f t e r - e f f e c t s of a tube te s t are short l i v e d when the major response to the f i s h i n the tube consists of charging and b i t i n g . They do not give the data on which t h i s claim i s based. In the p a r t i t i o n t e s t s , the experimental male and the t e s t male are i n continual v i s u a l contact. Thus while over a long period of time there may be a waning i n the response of one f i s h to another, i t seems u n l i k e l y that the rapid changes i n b i t i n g and charging seen a f t e r f e r t i l i z a t i o n are a part of t h i s waning process. The low frequencies of b i t i n g and charging seen i n the model tests a f t e r f e r t i l i z a t i o n may be a r e s u l t of exposure to a l i v e f i s h and then the model within a short period of time. So i f models are to be used r e g u l a r l y i n an experiment i t would be wise to study the waning of response on repeated presentations of a model, and the e f f e c t of exposure to a l i v e f i s h on subsequent responses to a model. I t i s a serious ommission of t h i s comparative study that the e f f e c t of habituation i n the three te s t procedures has not been studied. I t may be that the anomalous r e s u l t s from the model tes t s a r i s e because the t e s t period was only two minutes long. 138 The pre-nest and empty-nest phases Presumably the pre-nest phase i n t h i s study i s equiv-alent to the period i n which t e r r i t o r i e s are set up. A p r i o r i there i s no reason to believe that s e t t i n g up a t e r r i t o r y requires a lower attack tendency than does the subsequent re t e n t i o n of that t e r r i t o r y . The r e s u l t s from the p a r t i t i o n experiment suggested that there was only a s l i g h t difference i n the frequency of charging and b i t i n g between pre-nest f i s h and empty-nest f i s h . The major diffe r e n c e between the two groups was the nest directed behaviour of the empty-nest f i s h . In contrast both the tube and model tests indicated that there was a higher l e v e l of b i t i n g and charging i n the empty-nest f i s h . I t seemed, however, that the increase i n these measures was p a r t i a l l y dependent on the sex of the t e s t f i s h . Pre-nest f i s h treated both males and non-gravid females a l i k e . But empty-nest f i s h showed a higher l e v e l of b i t i n g and charging towards males than towards non-gravid females. There was no great d i f f e r e n c e i n the frequency of zig-zagging to males or non-gravid females that would help to account f o r the d i f f e r e n c e i n frequency of charging and b i t i n g shown to them. The gonadectomy experiments also suggested that there was t h i s d i f f e r e n c e i n the response of pre-nest and empty-nest phase f i s h . I t was found that, i r r e s p e c t i v e of the t e s t method, the frequency of b i t i n g of gonadectomised f i s h was c h a r a c t e r i s t i c of i n t a c t sham-operated controls i n the pre-nest 139 phase. This was true whether the f i s h were gonadectomised p r i o r to nest-building or while they had a nest. Hoar (1962b) had suggested that aggression, as measured by h i s ranking method (see introduction) was dependent on the l e v e l of gonadotrophins. This suggestion was supported by Carew (1968). She treated male sticklebacks with methallibure, which i s believed to block the formation of gonadotrophins (Hoar et a l 1967). Treated sticklebacks showed a marked reduction i n the l e v e l of pre-spawning (pre-nest) aggression. Empty-nest f i s h , however, showed a drop i n b i t i n g at a male i n a tube a f t e r gonadectomy. The drop i n b i t i n g at t e s t f i s h a f t e r gonadectomy seemed to be s p e c i f i c to male t e s t f i s h ( c f . F i g . 17). Baggerman (1966) found an even more marked drop i n the frequency of b i t i n g and charging a f t e r gonadectomy of empty-nest f i s h . She hypothesised that there was a change i n the hormonal control of aggression with gonadal hormones becoming important a f t e r nest b u i l d i n g . Unfortunately i n both Baggerman's experiments and the experiment described i n t h i s study, an important co n t r o l has been omitted. The main behavioural change i n gonadectom-is e d f i s h i s the loss of nest-directed a c t i v i t i e s . Thus a co n t r o l group of sham-operated f i s h whose nest has been removed and r e - b u i l d i n g i s prevented, should also be used. The drop i n b i t i n g a f t e r gonadectomy may be a consequence of the loss of the f o c a l point of t e r r i t o r y , and not because of d i r e c t hormonal c o n t r o l . Symons (1965) measured frequency of b i t i n g 140 i n f i s h whose nests had been removed. He found a marked lowering compared to f i s h with nests i n place, A modification of Baggerman's hypothesis can be proposed. There i s a basic l e v e l of b i t i n g (and charging e t c ) which i s probably dependent on the l e v e l of gonadotrophic hormones. Super-imposed on t h i s l e v e l there i s a response shown by empty-nest phase f i s h e s p e c i a l l y towards cons p e c i f i c males. This increased response to males i s d i r e c t l y or i n d i r e c t l y dependent on the l e v e l of gonadal hormones. I f empty-nest males are gonadectomised Baggerman's hypothesis would pr e d i c t that the l e v e l of b i t i n g to both males and non-gravid females would decline. The alternate hypothesis would predict that only the l e v e l of b i t i n g to males would de c l i n e . The data presented here would support the second hypothesis. Spine-raising and head-down threats Morris (1958) has pointed out that there may be more than one function f o r the spines. They can act as s t a b i l i s e r s , as anti-predator devices, and as i n t r a - s p e c i f i c s i g n a l l i n g devices. Lea (1968) found that l e i u r u s males from the L i t t l e Campbell River had r e l a t i v e l y shorter spines than trachurus males from the same r i v e r . In addition, the red colouring of the ventral spines was present i n the majority of the trachurus males and absent i n the majority of the l e i u r u s males. These differences may well make the spines 141 of the l e i u r u s males less u s e f u l i n both t h e i r anti-predator and i n t r a - s p e c i f i c s i g n a l l i n g functions. Symons (1965) found that the r a i s i n g of dorsal spines was p o s i t i v e l y associated with zig-zagging and, when zig-zags were present i n the approach, negatively with b i t i n g . Raising the dorsal and v e n t r a l spines together showed a p o s i t i v e a s s o c i a t i o n with both b i t i n g and zig-zags. As i n t h i s study, he found a decrease i n the percentage of charges made with dorsal spines erect between the f i s h i n the sexual (empty-nest) phase and f i s h i n the f i r s t s i x days of the parental period. Other occurrences of spine r a i s i n g were noted i n t h i s study. For example i n the f i r s t tests a f t e r f e r t i l i z a t i o n the highest frequency of b i t i n g was combined with a high percentage of charges with the dorsal spines erect. The dorsal and v e n t r a l spines were erect i n most approaches to the models. Is the function of spine r a i s i n g i n the approach to models an anti-predator device or an i n t r a - s p e c i f i c signal? Can the 'strangeness 1 of a t e s t s i t u a t i o n or object be gauged by the degree of spine r a i s i n g ? Symons found that at the s t a r t of a tube t e s t a higher proportion of charges were made with spines erect than towards the end of the t e s t . In the tube t e s t experiments, the proportion of charges with dorsal or dorsal and v e n t r a l spine erect was highest i n t e s t s made at the beginning of a s e r i e s . For example i n the pre-nest and empty-nest phase, the r e s u l t s used i n the analysis came from 142 among the f i r s t t e s ts given to the f i s h * The t e s t s i t u a t i o n was more novel to f i s h i n the pre-nest phase than to f i s h i n the middle of the parental phase. By contrast i n the p a r t i t i o n design there were no major changes i n s p i n e - r a i s i n g i n any of the experiments except immediately a f t e r f e r t i l i z a t i o n . In the p a r t i t i o n design the t e s t s i t u a t i o n i s continuous: from the s t a r t to the end. Another aspect of the s p i n e - r a i s i n g behaviour of leiurus. males was the s c a r c i t y of head-down threats. In observation of a- wild population of trachurus f i s h i t was found that the head-down threat was frequently seen i n encounters between t e r r i t o r i a l males. In these encounters the erect, reddened ventral spines could c l e a r l y be seen. In a p a r t i t i o n arrangement, van den Assem (1967) found that i n i t i a l l y , b i t i n g against the glass, occurred most, but l a t e r on threatening became predominant. In Baggerman's (1966) r e s u l t s threatening was an important component of the agonistic behaviour of the males. Both i n t h i s study and i n a study of R, Black (personal communication) i n which there was free contact between t e r r i t o r i a l males, threatening was r e l a t i v e l y r a r e . This may be r e l a t e d to the less prominent spines and colouring of the l e i u r u s males. ( i i ) Relationships between selected measures Only a few recognizable a c t i v i t i e s occurred often enough i n the tests to be s u c c e s s f u l l y analysed. These were 143 o r i e n t i n g , charging, and b i t i n g . The raw data included frequency of b i t i n g , frequency of charging, t o t a l oriented time, and the number of bouts of oriented time. From these, a number of measures were derived, f o r example, bi t e s per minute of oriented time, and the proportion of charges ending i n a b i t e . Inevitably t h i s r a i s e d the problem discussed by Cane (1961)^of s t a t i s t i c a l dependence between some of the measures. Can the use of a number of measures which may not be a l l independent be j u s t i f i e d ? As Cane has shown i n a discussion of van I e r s e l ' s (195 3) measures of fanning i n the stickleback, one measure may be better than another f o r a p a r t i c u l a r purpose, even though they are s t a t i s t i c a l l y dependent. Van I e r s e l found that the mean bout length of fanning cor r e l a t e d more c l o s e l y with the date of hatching than did t o t a l duration. Mertz and Barlow (1966) used derived data as a d d i t i o n a l indices of temporal changes i n the i n t e r - r e l a t i o n s h i p s between measures taken d i r e c t l y from the raw data. F i n a l l y , Nelson (1965) has pointed out that a measure such as frequency per unit time i s a complex measure which may give misleading r e s u l t s , and i t may be important to break i t down int o components. An example of a derived measure which was useful i n t h i s study i s b i t e s per minute of oriented time. This was highly correlated with frequency of b i t i n g i n a l l three of the t e s t methods. However, i n the comparison between the te s t methods, the r e s u l t s f o r the derived measure showed much clo s e r 144 agreement between the three t e s t procedures than did the r e s u l t s f o r the raw measure, frequency of b i t e s . In the 24 hour period a f t e r f e r t i l i z a t i o n , the measure b i t e s per minute of oriented time shoived more obvious changes than did the frequency of b i t i n g . The change i n l e v e l s of b i t e s per minute of oriented time was v i r t u a l l y i d e n t i c a l i n both tube and p a r t i t i o n t e s t s . This derived measure also provided a good index of changes i n the tendency to b i t e over the parental period. One of the l i m i t a t i o n s of t h i s study was that only a few behaviour patterns occurred often enough i n the tests to be used i n the analyses. I t was not possible to obtain a s u f f i c i e n t number of behaviour patterns to attempt an intensive motivational study such as Blurton-Jones (1968) analysis of displays i n the Great T i t . The behavioural r e p e r t o i r e of the Great T i t was s u f f i c i e n t l y large that the use of s t a t i s t i c a l l y dependent measures could be avoided. In a l l the experiments the c o r r e l a t i o n s between the various measures of b i t i n g and charging were s i m i l a r . But there may be advantages i n using one measure rather than another. For example the frequency of b i t i n g was highly c o r r e l a t e d with frequency of charging. This would be ex-pected from the d e f i n i t i o n s . Both measures showed s i m i l a r changes during an experiment. I t may be preferable to use charging, as defined i n t h i s t h e s i s , rather than b i t i n g . This would avoid the d i f f i c u l t y of d i s t i n g u i s h i n g b i t i n g from 145 bumping. A-lthough they were defined d i f f e r e n t l y , i t was often d i f f i c u l t to d i s t i n g u i s h them when the frequency of charging was very high. Bites per bout of oriented time was correlated with frequency of b i t i n g . That i s , f i s h with high l e v e l s of b i t e s per bout tended to have high l e v e l s of frequency of b i t i n g . In s p i t e of t h i s , b i t e s per bout did not show the changes i n le v e l s seen i n frequency of b i t i n g over the parental period when the tube tests were used. The proportion of charges ending i n a b i t e may not be a useful measure i n any causal analysis using the methods described i n t h i s study. Where the experimental male could reach the t e s t object, as i n the model t e s t s , a very high proportion of a l l charges ended i n a b i t e . Where the experi-mental and t e s t f i s h were separated by glass, the proportion of charges ending i n a b i t e was highly c o r r e l a t e d with frequency of b i t i n g . Symons (1965) also found that the per-centage of approaches with b i t e s was co r r e l a t e d with t o t a l b i t e s per 5 minutes (r = + 0.716). However, changes i n the proportion of charges ending i n a b i t e did not always c l o s e l y follow changes i n the frequency of b i t i n g . In the parental stage the former measure showed only a quadratic trend. In the tube t e s t s , i t showed only a small decline i n te s t s follow-ing f e r t i l i z a t i o n , and i t was not s i g n i f i c a n t l y larger f o r f i s h i n the empty-nest phase than f o r f i s h i n the pre-nest phase. 147 length of oriented time. Cane's method of analysis i s obviously useful i n analysing the causal r e l a t i o n s h i p s i n behaviour. In t h i s section I have t r i e d to show that the use of more than one measure of an a c t i v i t y , even though the measures are s t a t i s t i c a l l y dependant, can add to the knowledge of changes i n that a c t i v i t y . Indeed the use of only one measure may obscure changes i n an a c t i v i t y . ( i i i ) Comparison of the t e s t methods There were large differences i n the frequency of b i t i n g , frequency of charging and i n t o t a l oriented time obtained by the three methods. A l l methods, however, gave comparable values f o r the measure, b i t e s per minute of oriented time. This measure does not depend on the length of the t e s t period or on the t o t a l oriented time. I t i s differences i n these that account f o r most of the differences i n frequency of b i t i n g and frequency of charging between methods.. This study has omitted another aspect of the comparison of the l e v e l of a measure across t e s t methods. Does an i n d i v i d u a l f i s h that shows high rates of b i t i n g i n one te s t method also show high rates of b i t i n g i n another t e s t s i t u a t i o n ? Would t h i s f i s h also hold a larger or a better s i t e d t e r r i t o r y i f placed amongst other f i s h ? R. Black, working at the Unive r s i t y of B r i t i s h Columbia, i s making a preliminary study of t h i s question. U n t i l the question i s answered, i t would be misleading to say that one i n d i v i d u a l i s more aggressive than another on the basis of one t e s t method. 148 This t h e s i s , however, compared changes i n the l e v e l s of measures obtained by three t e s t methods. In the twelve day parental period the frequency of b i t i n g , frequency of charging, and b i t e s per minute of oriented time a l l showed s i m i l a r changes i r r e s p e c t i v e of the t e s t method. The work of other authors, Morris (1958) and Black (personal communication), has shown that t e r r i t o r y s i z e i n the parental period changes i n a way that p a r a l l e l s the changes found i n the measures described above. This would suggest that i n the parental period these a c t i v i t i e s are sharing a common causal f a c t o r . A s i m i l a r conclusion can be drawn from the r e s u l t s of the gonadectomy experiment. While the i n t e r n a l causation of b i t i n g and charging at a male-like model (Model S/R) was not c l e a r l y separable from that f o r b i t i n g and charging at male c o n s p e c i f i c s , there were some i n d i c a t i o n s that the responses to models were not i d e n t i c a l to the responses to l i v e males. In the parental period Model S/R tests yielded a r e l a t i v e l y smaller change i n the l e v e l s of b i t i n g and charging than did the tests that used l i v e f i s h . The Model S/R tests did not reveal marked changes i n frequency of b i t i n g or charging i n the 24 hours a f t e r f e r t i l i z a t i o n . In addition, empty-nest phase f i s h were f a r more responsive to models than were pre-nest phase f i s h . Model t e s t s were also exceptional because of the high proportion of charges made by the experimental f i s h with dorsal and ventral spines erect. This was true at a l l stages of the reproductive cycle studied. 149 In view of these d i f f e r e n c e s , measures of aggression obtained using models cannot be regarded as p r e c i s e l y equivalent to measures of aggression obtained by tests i n which a l i v e male f i s h i s used. The tube and the p a r t i t i o n t e s t s c o n s i s t e n t l y gave comparable r e s u l t s f o r changes i n b i t i n g or charging i n an experiment. For a given experiment there may be p r a c t i c a l reasons f o r using one rather than the other. The tube t e s t s y i e l d e d high values f o r frequency of b i t i n g , frequency of charging, and oriented time, but few bouts of nest-directed a c t i v i t y . In the p a r t i t i o n tests a f a r broader range of behaviour i s seen. But to obtain a comparable number of charges or b i t e s would require a f a r longer t e s t period than i n a tube t e s t . The p a r t i t i o n method also requires twice as many f i s h with nests to obtain a sample s i z e equal with the tube method. Id e a l l y the behaviour of f i s h on both sides of the p a r t i t i o n should be recorded simultaneously so that records of the experimental f i s h could be corrected f o r differences i n the a c t i v i t y of the partner f i s h . There was not time to develop other methods of measuring the aggression of males. These other measures might include the rate at which a male v / i l l perform an operant response which leads to an encounter with another male, or the strength of an aversive stimulus that the male w i l l overcome i n order to make an attack on another male. These are more i n d i r e c t measures of aggression than frequencies of 150 b i t i n g or charging, measures c h a r a c t e r i s t i c of the methods used i n t h i s study. I f the i n d i r e c t measures were highly correlated with the more d i r e c t measures, then i t would be parsimonious to postulate a unitary drive to account f o r the r e l a t i o n s h i p s between such diverse measures ( M i l l e r 1957). In t h i s study, discussion of the concept of unitary drives has been l i m i t e d because only three methods were used to measure aggression. I t i s misleading to use the concept of a unitary aggressive drive i n the causal analysis of the male s t i c k l e -back behaviour i n the absence of a comparative study of measures of aggression. This study has shown that two methods of measuring aggression gave comparable r e s u l t s . But a t h i r d method was not always consistent with the other two. 151 VI. BIBLIOGRAPHY Assem, J . van den (1967). T e r r i t o r y i n the three-spined stickleback Gasterosteus aculeatus L. An experi-mental study i n i n t r a - s p e c i f i c competition. Behaviour, Suppl. 16. 164p. Baggerman, B. (1966). On the endocrine c o n t r o l of reproductive behaviour i n the male three-spined stickleback (Gasterosteus aculeatus L . ) . Symp. Soc. Exptl. B i o l . 20: 427 - 456. Cane, V.R. (1961). Some ways of describing behaviour. In .Current Problems i n Animal Behaviour (ed. by W.H. Thorpe and O.L. Zangwill), pp. 361 - 388. Cambridge University Press. Carew, B.A.M. (1968). Some e f f e c t s of methallibure (I.C.I. 33,828) on the stickleback, Gasterosteus aculeatus L. M.Sc. Thesis. Uni v e r s i t y of B r i t i s h Columbia. 64p. Edwards, A.L. (1960). Experimental Design i n Psychological Research. Holt, Rinehart and Winston. New York. 398p. Ferguson, G.A. (1965). Non-parametric Trend A n a l y s i s . McGill University Press. Montreal. 61p. Hagen, D.W. (1967). I s o l a t i n g mechanisms.in three-spined sticklebacks. (Gasterosteus). J . F i s h . Res. Bd. Canada. 24: 1637 - 1692. Hinde, R.A. (1959). Unitary d r i v e s . Anim. Behav. 7: 130 - 141. Hoar, W.S. (1962a). Hormones and the reproductive behaviour of the male three-spined stickleback (Gasterosteus  aculeatus L . ) . Anim. Behav. 10: 247 - 266. Hoar, W.S. (1962b). Reproductive behaviour of f i s h . Gen. Comp. Endocrinol., Suppl. 1: 206 - 216. Hoar, W.S., J . Wiebe, and E.H. W'ai. (1967). I n h i b i t i o n of the p i t u i t a r y gonadotropic a c t i v i t y of fishes by a dithiocarbamoylhydrazine d e r i v a t i v e . (I.C.I. 33,828). Gen. Comp. Endocrinol. 8: 101 - 109. I e r s e l , J.J.A. van (1953). An analysis of the parental behaviour of the male three-spined stickleback (Gasterosteus  aculeatus L.) Behaviour, Suppl. 3. 159p. I e r s e l , J.J.A. van (1958). Some aspects of t e r r i t o r i a l behaviour of the male three-spined stickleback. Arch, n e e r l . Zool. 13, Suppl. 1: 381 - 400. 152 Jones, N.G. Blurton (1968). Observations and experiments on the causation of threat displays i n the great t i t (Parus major). Anim. Behav. Monogr. 1: 74 - 158. Lea, G.R. (1968). L a t e r a l plates i n the three-spined s t i c k l e -back (Gasterosteus aculeatus. L. ) as a defence against predators. B.S'c. Thesis. University of B r i t i s h Columbia. 34p. McDonald, B.J. and W.A. Thompson, (1967). Rank sum multiple comparisons i n one and two way c l a s s i f i c a t i o n s , Biometrika. 54: 487 - 497. Mertz, J.C. and G.W. Barlow. (1966). On the reproductive behaviour of Jordanella f l o r i d a e with s p e c i a l reference to a quantitative analysis of parental fanning. Z„ Tierpsychol. 23: 537 - 554. M i l l e r , N.E. (1957). Experiments on motivation. Science. 126: 1271 - 1278. Morris, D. (1958). The reproductive behaviour of the ten-spined stickleback (Pygosteus pungitius L . ) . Behaviour, Suppl. 6. 154p. Muckensturm, B. (1967). L'Epinoche et les leurres, complexite" de l a react i o n . C.R. Acad. Sc. P a r i s . 264: 745 - 748. Muckensturm, B. (1968). La re"action des Epinoches aux leurres ne provient pas d'une confusion avec un congenere. C.R. Acad. Sc. P a r i s . 266: 2114 - 2116. Nelson, K. (1965). A f t e r e f f e c t s of courtship i n the male three-spined stickleback. Z. v e r g l . P h y s i o l . 50: 569 - 597. Segaar, J . (1961). Telencephalon and behaviour i n Gasterosteus aculeatus males. Behaviour. 18: 256 -287. Sevenster, P. (1961). A causal analysis of a displacement a c t i v i t y (fanning i n Gasterosteus aculeatus). Behaviour, Suppl. 9. 170p. Sevenster-Bol, A.C.A. (1962). On the causation of drive reduction af t e r a consummatory act ( i n Gasterosteus  aculeatus). Arch, ne'er 1. Zool. 15: 175 - 236. S i e g e l , S. (1956). Nonparametric S t a t i s t i c s f o r the Behavioural Sciences. McGraw-Hill, New York. 312p. S t e e l , B.G.D. and J.H. T o r r i e . (1960). P r i n c i p l e s and Procedures of S t a t i s t i c s . McGraw-Hill, New York. 481p. 153 Symons, P.E.K. (1965). Analysis of s p i n e - r a i s i n g i n the male three-spined stickleback. Behaviour 26: 1- 74. Thompson, W.A. and T.A. Wilkie. (1963). On an extreme rank sum t e s t f o r o u t l i e r s . Biometrika. 50: 375 - 383. Tinbergen, N. (1951). The Study of I n s t i n c t . Clarendon Press. Oxford. 228p. Tinbergen, N. (1953). S o c i a l Behaviour i n Animals. Methuen. London. 150p. Tinbergen, N. (1954). The Tale of John S t i c k l e . Methuen. London. 48p. 

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.831.1-0102223/manifest

Comment

Related Items