Studies on Selective Feeding in the Pacific Starfish Pisaster in Southern California

The asteriid starfish Pisaster giganteus and Pisaster ochraceus are important predators of mollusks. Experiments were designed to determine the extent to which predation by starfish on different kinds of prey is selective, and to learn whether selectivity varies according to 1) number and relative density of various alternative prey and 2) past history of feeding of the predators. Seven species of intertidal mollusks were used as prey. When alternatives were presented in equal abundance and in pairs, both species of starfish showed a significant preference for one alternative in all but one pairing. The hierarchy of preferences was well—defined and consistent among replicates. When seven species of prey were presented together in equal abundance, the preferences were of the same patter but were, in general, weaker than those exhibited when prey were presented in pairs. In all experiments, mussels were preferred over other mollusks; the strongest preferences were for Mytilus edulis and Mytilus califonianus over other alternatives. Where mussels (M. californianus) and snails (Tegula funebralis) were presented as alternative prey, changes in their relative densities had little effect on the preference for mussels. When the relative densities of two species of snails (T. funebralis and Acanthina spirata, neither preferred when in equal densities) was varied, a slight preference was shown for the more abundant form at one extreme density only. In this experiment individual starfish exposed to the same relative densities of alternative prey differed in the proportions of alternatives they chose. Starfish fed only on snails (T. funebralis) for 3 months showed an increased preference for these snails over chitons (Nuttalina californica) as well as a reduced preference for mussels; the latter preference reappeared strongly after only 1 week of exposure to both prey. The hypothesis that predators induce stability in prey populations is discussed. The experimental results indicate the importance of preferential feeding to this question. Models of selective feeding based on short—term changes in hunger and situation may be inappropriate when applied to cold—blooded predators. An alternative interpretation, based on learning and hunting behavior, is given for the changes in selectivity observed in these experiments.