David W. Stephens

This account of the current state of foraging theory is also a valuable description of the use of optimality theory in behavioral ecology in general. Organizing and introducing the main research themes in economic analyses of animal feeding behavior, the authors analyze the empirical evidence bearing on foraging models and answer criticisms of optimality modeling. They explain the rationale for applying optimality models to the strategies and mechanics of foraging and present the basic "average-rate maximizing" models and their extensions. The work discusses new directions in foraging research: incorporating incomplete information and risk-sensitive behavior in foraging models; analyzing trade-offs, such as nutrient requirements and the threat of being eaten while foraging; formulating dynamic models; and building constrained optimization models that assume that foragers can use only simple "rules of thumb." As an analysis of these and earlier research developments and as a contribution to debates about the role of theory in evolutionary biology. Foraging Theory will appeal to a wide range of readers, from students to research professionals, in behavioral ecology, population and community ecology, animal behavior, and animal psychology, and especially to those planning empirical tests of foraging models.

What is the minimum dimension of a niche space necessary to represent the overlaps among observed niches? This book presents a new technique for obtaining a partial answer to this elementary question about niche space. The author bases his technique on a relation between the combinatorial structure of food webs and the mathematical theory of interval graphs. Professor Cohen collects more than thirty food webs from the ecological literature and analyzes their statistical and combinatorial properties in detail. As a result, he is able to generalize: within habitats of a certain limited physical and temporal heterogeneity, the overlaps among niches, along their trophic (feeding) dimensions, can be represented in a one-dimensional niche space far more often than would be expected by chance alone and perhaps always. This compatibility has not previously been noticed. It indicates that real food webs fall in a small subset of the mathematically possible food webs. Professor Cohen discusses other apparently new features of real food webs, including the constant ratio of the number of kinds of prey to the number of kinds of predators in food webs that describe a community. In conclusion he discusses possible extensions and limitations of his results and suggests directions for future research.