STUDIES OF FORAGING BEHAVIOUR AND TIME BUDGETING IN GREAT TITS (Parus major) by Alejandro Kacelnik (Wolfson College) A thesis submitted for the degree of Doctor of Philosophy at the University of Oxford Michaelmas Term 1979 a Li» con amor a Federico,con bronca a Oliver,con esperanza Studies of foraging behaviour and time budgeting in Great Tits (Parus major). Alejandro Kacelnik, D.Phil.. University of Oxford, 1979. Part One is an investigation of the dawn chorus in the Great Tit. The major conclusions are as follows: Foraging and singing are partly incompatible because of the temporal and spatial separation of microhabitat conditions required for each activity. The dawn chorus occurs at a time of day when: a) foraging efficiency is limited by low light intensity and possibly by lack of mobility of the prey due to low temperature; b) acoustic comunication is more efficient than visual displays because of poor visibility and an enhanced sound transmission; c) the rate of territorial intrusion is at its daily peak. Great Tit's readiness to sing and react agonistically towards a territorial intruder is enhanced by experimentally lowering profitability of feeding, but it does not appear to be affected by time of day. Restricting access to food early in the morning leads to higher body weight in the evenings as an anticipatory compensation. Part Two is a study of sampling strategies based on the statistical decision paradigm known as the Two-Armed Bandit problem. I studied transition behaviour in foraging experiments using concurrent variable ratio schedules. The birds' foraging behaviour approximated the predictions of a dynamic programing algorithm that calculates the optimal balance between exploring and exploiting for finite time horizons. The birds' response to patch diversity and time horizon was closer to molar maximization (total rewards over a certain period) than to molecular maximization (maximising the instantaneous probability of reward). When the environment is not totally stable these two possibilities conflict, and my results supported molar maximization. Part Three investigates the trade-off between maximising foraging efficiency and efficient territorial defence. Great Tits behaved close to the predictions of the Marginal Value model when there were no territorial intrusions, but modified their feeding behaviour in the predicted way when the probability of intrusion was higher. ACKNOWLEDGEMENTS First, I would like to thank my supervisor, John Krebs, for his constant and enthusiastic encouragement and advice. The many Sundays spent discussing chapters of this thesis were only the last stage or the most fruitful professional association that I have enjoyed. I would also like to thank especially David McFarland who, unwittingly, was responsible for my approach to quantitative ethology and my coming to Oxford in 1974. Most of my companeros in the Animal Behaviour Research Group were a source of inspiration ana learning by osmosis, but the osmotic x gradient was specially steep with respect to some of them, who I feel deserve special mention: Alasdair Houston, Robin McCleery, Richard Sibly, Marian Dawkins, Shelly Cohen, Roger Tarpy, Mac Hunter, Richard Dawkins, Jon Erichsen, David Macdonald, Adeline Nunez will always mean Oxford for me, as will - very especially - Pat Searle. Alasdair Houston read and criticised the whole of this thesis, and I benefited enormously from his advice and enjoyed thoroughly all aspects of working with him. Carlos Bernstein, Robin McCleery and Richard Sibly also read some chapters and suggested very good ideas. I did my best to take them into account. John Roberts introduced me to the world of acoustics. I wish he enjoyed as much as I did all the rainy dawns we spent x^ith Mac Hunter in Wytham. % Meyer and Elisa, my parents, endured bravely^the worst times of their lives while I was away. I am most grateful to them for too many things to list, and feel deeply sad that Meyer did not live to the joy of my return home. My research was supported financially by the following sources: The British Council, The Science Research Council, Wolfson College, Su and Ed Kleinman, and, last but not least, Lidia Rapaport. Professor J. W. S. Pringle and Dr. D. J. McFarland provided facilities in the Department of Zoology and the Animal Behaviour Research Group respectively. Nick Styles gave me efficient and enthusiastic technical support and helped me to train the birds and run some experiments. Dick Cheney, Dave Palmer and Terry Barker from the different workshops of the Department built most of my gadgets. With respect to the pre-history of this thesis, I am grateful to Enrique Segura for the good neurophysiological years together and his encouragement to come to Oxford, and to Su Kleinman for teaching me English. Lidia Rapaport helped me to proof read, by making drawings, typing references, and correcting my English. Her support has been invaluable. 1. CONTENTS (cid:9)Page No. CHAPTER 1 Procedure and General Methods 1 1.1. Introduction 1 1.2. The Species 1 1.3. The Subject 2 1.4. Acclimitisation to the Laboratory 2 1.5. The holding aviaries 3 1.6. Roosting boxes 3 1.7. Diet 4 1.8. Experimental Aviaries 6 1.9. The feeding discs 7 1.9.1. The perch 7 1.9.2. The logic 8 1.9.3. The discs 10 1.10. Training birds to use the discs 11 1.11. Setting up territorial defense experiments 13 1.12. Data Recording 16 1.13. Data Analysis 16 PART I - Studies on Time Budget CHAPTER 2 2.1. The Dawn Chorus 18 2.2. Hypotheses about the dawn chorus 18 2.2.1. Proximate explanations 18 2.2.1.1. The fixed threshold 19 2.2.1.2. Internal clock hypothesis 20 2.2.2. Functional explanations 23 2.2.2.1. Ecological constraints 23 2.2.2.1.1. Acoustical factors, especially wind 23 2.2.2.1.2. Non-acoustical effects of temperature 25 2.2.2.1.3. Light intensity 26 2.2.2.1.4. Predation 27 2.2.2.2. Intraspecific factors 27 2.2.2.2.1. The behaviour of receivers 27 2.2.2.2.2. The behaviour of other actors 28 2.3. The dawn chorus in the Great Tit 30 CHAPTER 3 Environmental Factors affecting the Dawn Chorus 34 3.1. Introduction 34 3.2. Light intensity 35 3.2.1. Methods 35 3.2.1.1. Subjects 35 3.2.1.2. Apparatus 35 3.2.1.3. Prey 36 3.2.1.4. Testing 37 3.2.1.5. Units 37 3.2.1.6. Training 38 cont'd 11. CHAPTER 3 continued 3.2.2. Results 39 3.3. Temperature 41 3.3.1. Method 41 3.3.2. Results 42 3.4. Wind 43 3.4.1. Methods 44 3.4.2. Results 45 3.5. Conclusions 46 CHAPTER 4 Anticipatory control of body weight 49 4.1. Introduction 49 4.2. Method 51 4.3. Results 52 4.4. Discussion 55 CHAPTER 5 On the motivational mechanisms underlying the dawn chorus 61 5.1. Introduction 61 5.2. Methods . 62 5.2.1. Birds 62 5.2.2. Aviary arrangements 62 5.2.3. Experimental design 63 5.2.4. Deprivation 64 5.2.5. Intrusion 65 5.2.6. Recording 65 5.2.7. Measures of responses 66 5.3. Results 69 5.3.1. Description of responses 69 5.3.2. Supression of feeding by intrusion 70 5.3.3. Return to baseline feeding rate 72 5.3.4. Intensity of the song response 73 5.3.5. Return to baseline singing rate 73 5.3.6. Number of attacks 74 5.3.7. Approach 75 5.3.8. Displays 75 5.4. Discussion 75 5.4.1. Foraging opportunity 75 5.4.2. Circadian clock 78 5.4.3. Intruders pressure 79 5.5. General conclusions 80 PART II - Studies of foraging behaviour: Optimal sampling CHAPTER 6 6.1. Optimal foraging and sampling 81 6.2. The Memory Window 82 6.3. Psychological studies of concurrent schedules 84 cont'd Ill. CHAPTER 6 continued 6.4. The Two Armed Bandit 88 6.5. Summary 90 CHAPTER 7 An experiment on optimal sampling 92 7.1. Introduction 92 7.2. Methods 92 7.2.1. Foraging Patches 92 7.2.2. Experimental Sessions 93 7.2.3. Decision criteria 95 7.2.4. Subjects 95 7.2.5. Pilot experiments 95 7.3. The Model 95 7.4. Results 102 CHAPTER 8 v, Rules of Tumb for the Two Armed Bandit Problem 108 \ 8.1. Introduction 108 8.2. Methods 109 8.2.1. Simulation procedures 109 8.2.2. The Rules 110 8.2.2.1. IMMAX 111 8.2.2.2. MATCH ' 112 8.2.2.3. PLWIN 112 8.2.2.4. CULLEN d: Cullen Rules 113 8.2.3. The evaluation of the rules 113 8.3. Results 115 8.3.1. Comparison with real birds 115 8.3.2. Effect of average profitability 116 8.3.3. Effect of patch diversity in a rich environment 116 8.3.4. Effect of the time horizon 117 8.4. Discussion 118 CHAPTER 9 Optimal Sampling and Time Horizon 120 9.1. Introduction 120 9.2. Methods 123 9.2.1. Subjects 123 9.2.2. Instruments 123 9.2.3. Types of run 123 9.2.3.1. Treatments 123 9.2.3.1. Neutralisations 124 9.2.4- Measures of sampling 125 9.2.4.1. Decision point 125 cont'd IV. CHAPTER 9 continued i 9.2.4.2. Commitment 126 9.2.5. Experimental sequence 126 9.3. Predictions 126 9.3.1. Molar Predictions 126 9.3.2. Molecular Predictions 129 9.4. Results 129 9.4.1. General Behaviour 129 9.4.2. Test of predictions 130 9.4.2.1. Global amount of sampling 130 9.4.2.2. Difference in sampling between treatments 131 9.4.2.2.1. Graphical analysis 131 9.4.2.2.2. Statistical analysis 133 9.4.2.3. Pattern of switching between patches 133 9.5. Discussion 134 PART III - The trade-off between foraging and territorial defence CHAPTER 10 Optimal foraging and territorial defence 136 10.1. Introduction 136 10.2. Methods 139 10.2.1. The birds 139 10.2.1. The experimental set up 140 10.2.3. Foraging schedules 140 10.2.4. Experimental sessions 141 10.3. Predictions 142 10.3.1. The behaviour of control sessions 142 10.3.1. The behaviour in intrusion sessions 143 10.4. Results 143 10.4.1. Do undisturbed birds forage optimally? 143 10.4.2. The effect of intrusions 146 10.4.3. Inverse optimality 147 10.5. Discussion 149 REFERENCES APPENDICES 1. Algol program to solve the finite horizon two armed bandit problem with equal sampling before decision. 2. Kacelnik, A. 1979 The foraging efficiency of great tits (Parus ma jor L.) in relation to light intensity. Animal Behaviour, 27, 237-241. £_ cont'd V. 3. Roberts, J. , Kacelnik, A. and Hunter, M. 1979 Sound interference patterns in animal acoustic communication. Technical memorandum 66. Polytechnic of the South Bank, London. 4. Roberts, J., Kacelnik, A. and Hunter, M. 1979 Acoustic properties of the environment as determinants of birds' song patterns. Animal Behaviour(in press). 5. Kacelnik, A. and Tarpy, R. 1979 The effect of hunger and vigilance during prey handling in the great tit (Parus major) . Manuscript. 6. Krebs, J.R.> Kacelnik, A. and Taylor, P. 1978 Test of optimal sampling by foraging great tits. Nature, 275 27-31. 7 \A/±± <Lou«. n* cell dW^ ^^ v^i"1^ A. kreks *•* w 10 u,