Abstract
The foraging paradigm of trace fossil theory has historically accorded random behavior to non-food-limited deposit-feeders and non-random behavior to food-limited feeders. A series of randomness measures derived from empirical modeling, simulation modeling, stochastic modeling and probability theory applied to foraging patterns observed in deep-sea bottom photographs from the Arctic and Antarctic yielded a behavioral continuum of increasing non-randomness. A linear regression of trace positions along the continuum to bathymetric data did not substantiate the optimal foraging efficiency-depth dependence model of trace fossil theory, except that all traces exhibited a greater optimization than that of simulated random foraging. It is hypothesized that optimization as evidenced by non-random foraging strategies represents maximization of the cost/benefit ratio of resource exploitation to risk of predation and that individual foraging patterns reflect an exploration response to the morphometry of a patchily distributed food resource. Differential predation and competition may account for the co-occurrence of random and non-random strategies within the same bathymetric zone.