Food Preference as a Function of Resource Abundance with Multiple Prey Types: An Experimental Analysis of Optimal Foraging Theory

Abstract

Data were collected on the response of a small population of folivores (Kerodon rupestris: Caviidae) to changes in the absolute abundance of 10 different species of leaves. When absolute abundances were tripled (relative abundances were held constant) the folivore population significantly altered its proportional consumption of food types. In general, preferred foods were consumed in greater relative proportions when abundance was increased and the relative utilization of less preferred foods declined. Otherwise, our results were problematic in relation to the assumptions and predictions of the energy maximization model of optimal foraging theory. First, not all foods of low rank decreased in consumption when we increased the abundance of high-ranked foods. The partial consumption of low-ranked foods (or partial preferences as referred to in the literature) is contrary to the tenets of energy maximization. Second, preferences did not remain consistent from phase to phase. In fact, ranks shifted considerably, especially with respect to high-ranked foods. A number of factors may have been responsible for these contradictions to optimal foraging theory. Folivores may represent an example of the special case of optimal foraging with nutrient constraints (Pulliam 1980). The partial consumption of low-ranked foods would then be expected in this situation. Also, the encounter rate for a given item or type is probably not constant in time or space; the rate of encounter is an ephemeral value, affected by an animal's degree of satiation, and dependent upon its degree of familiarity with the distribution of food items within its foraging area. This suggests that animals are optimal foragers, sensu Charnov, only in the short term. The probability of accepting an item when it is encountered will be either zero or one in the short term, but in the long term (hours, days, seasons, years) it will assume an average value between zero and one. Finally, consumers should evolve search images for types or species, when within-type variability is very small and overlap among types minimal. When within-type variability is large, and distributions of food value overlap a great deal, natural selection should favor consumers that select high value items and do not generalize on a type's average values. We present a scenario which proposes that feeding habits may influence the way a consumer optimizes its foraging behavior. Empirical tests of optimal foraging models based upon energy maximization have upheld theoretical predictions when predators were presented with limited prey choices (e.g., Goss-Custard 1977; Krebs et al. 1977). Tests which have examined more complex situations (Zach and Smith 1980; this paper) suggest that models which include considerations of nutrient constraints (Pulliam 1975) or predator learning (Hughes 1979; Stenseth and Hansson 1979; Pulliam 1980) may, in fact, be more appropriate.