Optimal foraging theory predicts that the energy value of a marginally acceptable item equals the overall energy available to the animal in the environment (calories per unit time). If true, then the selective behavior of the consumer may be used as a measure of this important parameter. In many arthropod predators the time spent feeding on an individual item should vary inversely with energy availability. This hypothesis is successfully tested using two species of tiger beetles (Cicindelidae). In many vertebrate predators the minimal acceptable food size should vary directly with energy availability. This hypothesis is not tested here, although some support for it exists in the literature. If valid, the simplicity of the method described allows the assessment of global patterns of energy availability for some taxa. Preliminary information on tropical-temperate gradients for birds is given. THE AMOUNT OF ENERGY (food) available to animals in nature is one of ecology's fundamental parameters. Unfortunately, it is also one of the most difficult to measure. Are tropical birds more food limited than temperate zone birds? What are the seasonal patterns of energy availability? What makes one habitat more attractive than another? These questions are all easier to ask than to answer. However, animals themselves assess energy availability, and often modify their behavior on the basis of their 'estimates.' This paper explores the possibility of using the behavior of the consumer to deduce the energy available to it in the environment. OPTIMAL FORAGING THEORY Most theories of selective feeding assume that the organism acts on the following principle: It searches for food, and on encountering an item asks itself the question, will it gain more energy by processing the item, or would it gain more by neglecting it and continuing to search (see Schoener 1974 for a review; this concept, however, does not take into account long-term considerations, such as discussed by Katz ( 1974) ) ? In other words, every item is judged by its 'e(x)/t(x)' ratio, where: e(x) -the 'energy profit' of the item (x) (this is a close correlate of biomass, adjusted for the amount of energy it took to capture it, indigestible matter content, etc.: it is the net energy received from that item); and t(x) -the amount of time it takes to process that item (this is the time during which the feeder is unable to search for, or process, other items). If the food is not particulate, these factors can be expressed per unit biomass. Also, it is clear that uch variables as success of capture can be incorporated into these terms. For example, the time expended on unsuccessful captures per successful capture can be included in the t (x) term, and the energy expended on unsuccessful captures per successful capture can be included in the e(x) term. By this process the animal partitions its set of encountered prey into a subset which it will accept and a subset which it will ignore. The main factor determining which subset a particular items falls into is the abundance of more desirable items (desirability defined by its e ( x) /t (x) ratio); but the animal itself may be spatially confined to a territory or to a refuge in order to escape predation, and in limiting he rea it can search, this restriction will ultimately affec what it will accept. The animal may have seve al feedin modes (e.g. hawking or foliage gleaning for birds) or a choice of several feeding patches. It is assumed that the organism will operate in the mode and patch that offers the most energy and, within these, follow the same principle. The result is some conglomerate optimization of many strategies, with the animal acquiring some amount of energy, E, over the total time spent feeding, T (its daily feeding bout, for instance), or an average of E/T calories per unit time. In the most realistic sense, this figure expresses the amount of energy available to the animal. But it is so subjective and involves so many compromises that it would be very difficult for the ecologist to measure it. It is now necessary to investigate the set of items that are on the boundary line of being accepted or rejected. The question once again is whether the animal will gain more energy by processing the item, I Present address: Department of Zoology, University of Washington, Seattle, Washington 98195, U.S.A. 96 BIOTROPICA 8(2): 96-103 1976 This content downloaded from 157.55.39.111 on Sat, 17 Sep 2016 05:47:56 UTC All use subject to http://about.jstor.org/terms or whether it could more profitably spend that time searching. If the animal processes the item it will get e(x)/t(x) calories per unit time throughout the processing time. If it continues searching it will receive, on the average, E/T calories per unit time. The boundary condition for accepting an item is thus: