A MODEL FOR THE SPACE-TIME DEPENDENCE OF FEEDING FOR PELAGIC FISH POPULATIONS

The time-varying spatial distribution of pelagic fish is an important factor in the development of feeding rate estimates for input into bio energetics growth models. Key issues are the degree to which fish and prey distributions overlap, the frequency and duration of such overlap s, and the effectiveness of foraging during overlaps. A model for the space and time dependences of feeding by the pelagic planktivorous ale wife Alosa pseudoharengus in the Great Lakes is presented. The model c ombines two submodels: (1) a diffusion-taxis model of predator and pre y dispersal and (2) a foraging rate model. Both models depend on envir onmental variables. The diffusion-taxis model is based on directional motion and random motion on the part of the fish. Directional motion i s determined from a taxis function that relies on local cues for feedi ng rate and encounter rate with predators. The foraging model expresse s the feeding rate of a predator as a function of predator and prey si zes, prey density, light level, and temperature. The diel movements, d istribution, and feeding rates of alewives were simulated with these m odels. The model simulations successfully represented the time-varying spatial distribution of alewives and predicted peaks in their feeding rate at dawn and dusk. The model simulations suggested a nighttime fe eding rate that was less than what is known from the literature. These results suggest that many of the known features of alewife distributi on can be reproduced with this class of model. Given the typically pat chy distribution of predator and prey, the nonhomogeneous character of environmental conditions, and the dynamic nature of both the biologic al and physical components of an ecosystem, a spatiotemporal approach such as this may be necessary for modeling the feeding, growth, and pr oduction dynamics of pelagic fishes.