DERIVING POPULATION PARAMETERS FROM INDIVIDUAL VARIATIONS IN FORAGINGBEHAVIOR .2. MODEL TESTS AND POPULATION PARAMETERS

1. The predictions of an empirical model were tested, based on a game theory approach that describes how a population of oystercatchers Haem atopus ostralegus, in which individual birds vary in their competitive ability and foraging efficiency, becomes distributed over its spatial ly variable mussel Mytilus edulis food supply. Model predictions on th e distribution of the birds over the main mussel beds of the Exe estua ry, on where particular cohorts of birds feed and on their intake rate s and local dominance ranks were tested against already published data . 2. The model predicted quite well the preference ranks of the 12 bed s; the spreading out of birds over more mussel beds as oystercatcher n umbers increased; the decrease in immature oystercatcher numbers on th e most preferred beds as adult numbers increased; the higher proportio n of immatures on the lower ranked beds; the high rate of movement bet ween beds of birds with low dominance scores; and, on the most preferr ed beds, the higher intake rates of dominant individuals and the simil arity between the intake rates of adults and immatures with the same d ominance. 3. Although, with a few exceptions, the qualitative trends w ere predicted correctly, their magnitudes were sometimes under-predict ed, suggesting that some parameter values need to be refined. None the less, the model tests were considered encouraging. Future work should aim to incorporate into the model other aspects of the birds' behavio ur, such as the effect on dominance of familiarity with a feeding area , and prey depletion. 4. By assuming that birds require a certain mini mum rate of food intake, either to remain on the estuary or to survive the winter, and by running simulations over a wide range of initial p opulation sizes, it was demonstrated how such a model can be used to p redict (i) the carrying capacity for oystercatchers of beds of differe nt preference rank, and of all the mussel beds combined, and (ii) the parameter values of the overwinter density-dependent mortality functio n. These findings, in turn, can be used to predict the local and globa l population consequences of winter habitat loss.