DERIVING POPULATION PARAMETERS FROM INDIVIDUAL VARIATIONS IN FORAGING BEHAVIOR .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 Haematopus ostralegus, in which individual birds vary in their competitive ability and foraging efficiency, becomes distributed over its spatially variable mussel Mytilus edulis food supply. Model predictions on the distribution of the birds over the main mussel beds of the Exe estuary, on where particular cohorts of birds feed and on their intake rates and local dominance ranks were tested against already published data. 2. The model predicted quite well the preference ranks of the 12 beds; the spreading out of birds over more mussel beds as oystercatcher numbers increased; the decrease in immature oystercatcher numbers on the most preferred beds as adult numbers increased; the higher proportion of immatures on the lower ranked beds; the high rate of movement between beds of birds with low dominance scores; and, on the most preferred beds, the higher intake rates of dominant individuals and the similarity between the intake rates of adults and immatures with the same dominance. 3. Although, with a few exceptions, the qualitative trends were predicted correctly, their magnitudes were sometimes under-predicted, 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' behaviour, such as the effect on dominance of familiarity with a feeding area, and prey depletion. 4. By assuming that birds require a certain minimum 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 population sizes, it was demonstrated how such a model can be used to predict (i) the carrying capacity for oystercatchers of beds of different preference rank, and of all the mussel beds combined, and (ii) the parameter values of the overwinter density-dependent mortality function. These findings, in turn, can be used to predict the local and global population consequences of winter habitat loss.