An application of optimal diving models to diving behaviour of Brunnich's guillemots

Although theoretical models predict that the quality of foraging patches has little effect on optimal dive time with increasing depth, many empirical studies show that dive time at a given depth may vary. We developed a model that incorporated patch quality as a parameter of energy intake as a nonlinear function of time, and applied it to the diving behaviour of Brunnich's guillemots, Uria lomvia. The model indicated that optimal dive time can vary widely depending on the parameter. It also explained the convergence of observed dive times with travel time. Assuming the birds dived optimally, this parameter can be estimated from travel time and dive time for each dive. Foraging patches with larger estimated parameter values were favoured by the birds, suggesting that the parameter indicated patch quality. We used this parameter to test an optimal patch use model in divers. The results indicate that Brunnich's guillemots adjust their diving behaviour adaptively depending on patch quality, and that the optimal diving model is valid for prediction of observed dive patterns if patch quality is incorporated appropriately. (C) 2002 The Association for the Study of Animal Behaviour. Published by Elsevier Science Ltd. All rights reserved.