CORRELATES OF THE DENSITY OF FORAGING OYSTERCATCHERS HAEMATOPUS-OSTRALEGIS AT DIFFERENT POPULATION SIZES

1. The distribution of a population of oystercatchers Haematopus ostralegus L., that varied temporally in size, over beds of mussels Mytilus edulis L., that varied spatially in quality, was investigated. It was expected that, as bird numbers on the estuary increased causing competition for preferred beds to intensify and the birds to spread out over their food gradient, the characteristics of the mussel beds with which bird densities were correlated would change. 2. Bird numbers on each bed, and thus their densities, were censused approximately monthly. The mussel bed characteristics considered likely to correlate with bird density were: mussels (densities of several length-classes and shell thickness on the dorsal and ventral sides) and beds (area, flatness, exposure time, proportion covered by algae, softness of the substrate and distance from the birds' high water roost). 3. An initial multiple regression analysis relating bird density to these bed characteristics, using data collected from 21-23 beds during the first 3 years, enabled three variables to be eliminated from further consideration: exposure time, algal cover and flatness. A subsequent similar analysis of data from the 12 'priority' beds, studied over 8 years, allowed more variables to be eliminated: the densities of small and large mussels, shell thickness on the ventral side, and bed size. The four remaining variables - substrate softness, distance to roost, density of mussels 40-50 mm long and shell thickness on the dorsal side - together accounted for most of the explainable variation in bird density between the 12 priority beds. 4. Multiple regressions of the logarithms of bird density against these four variables were then calculated separately for each of the 83 censuses. The standardized partial regression coefficients were plotted against total population size. In the presence of the other three variables, the effect of the density of 40-50 mm mussels was small at low population sizes but increased as bird numbers rose; at high population sizes, bird densities were higher where mussels were dense. At low population sizes, bird densities increased with shell thickness whereas, at high population sizes, they decreased with shell thickness. Bird densities were generally higher on beds far from the roost, particularly at the higher population sizes. The effect of substrate softness remained consistently large at all population sizes, and may even have increased; bird densities were markedly higher on beds with firm substrates on all except a few low population censuses. 5. The causal basis of these changes are discussed in relation to ideal free distribution models that assume that predators feed where their energy intake rate is maximized. Although the variables that defined the food gradient in this field study were largely different to those used in such models, the conceptual framework these models provide is nonetheless a useful way of viewing how the distribution of oystercathers over their food gradient changes with population size.