Comparative Dietary Ecology of Sympatric, Insectivorous Neotropical Flycatchers (Tyrannidae)

Abstract

This study compares and contrasts diets, based on 2834 prey individuals from 126 stomachs, of 16 insectivorous, aerially foraging Neotropical flycatcher species (Tyrannidae) that are seasonally sympatric in the humid Caribbean lowlands of Costa Rica. Dietary parameters examined are prey type, diversity (breadth) of prey types, heterogeneity of prey types among individuals of a species, "patchiness" of morphologically indistinguishable prey within individual stomachs, and number of prey items per stomach. An R—type factor analysis of prey taxa in flycatchers' stomachs delimited four interpretable factors (axes) based on differences in prey detectability, location, escape behavior, and substrate where caught. A Q—type cluster of flycatcher species, based on factor scores from the analysis of prey variables, tended to juxtapose congeners. Noncongeners that clustered closely included a pair of species (Todirostrum sylvia and Oncostoma cinereigulare) that tend to replace each other geographically. Some flycatchers failed to cluster with any other species. "Saturation curves" of prey—type diversity against number of stomachs sampled showed (1) adequate sampling effort for most species with 4—10 stomachs, and (2) species differences in breadth of prey types eaten. The flycatchers that did not cluster with any other species on the basis of prey taxa had the largest and smallest diet breadths of all species studied. Two species are exceptional and consistent specialists: the Ruddy—tailed Flycatcher (Terenotriccus erythrurus) ate 94% Homoptera (largely Fulgoroidea), and the Long—tailed Flycatcher (Colonia colonus) ate 67% stingless bees (Trigona species). Flycatcher species that had the most "patchy" diets (i.e., many individuals of the same prey species within a stomach) tend to hawk flying prey and/or reside in open country or forest canopy. Colonia colonus had extremely patchy stomach contents, but unlike other flycatchers that hawk flying prey, it had an exceptionally homogeneous diet. "Patch feeders," compared with other flycatchers, tended to have many more items per stomach, suggesting the selection of many, relatively small prey per unit time. Significant variation in numbers of items per stomach, even among congeners, suggested differences in feeding rate and perhaps prey size. Comparison of species with repect to these dietary parameters helps identify a "food resource" for each species. The identification of the food resource (1) helps interpretation of foraging behavior of the predators, (2) facilitates discussion of prey—type and prey—size selection, (3) delineates potential ecological and evolutionary routes of species interactions, and (4) focuses attention on the nature of the food supply available to each species. The assumption that species with patchy and heterogeneous diets feed opportunistically on relatively ephemeral prey permits the following conclusions: (1) many tropical flycatchers are not opportunistic, (2) many guilds, including tropical ones, are composed of species with a variable degree of opportunism, (3) migrants are more opportunistic while wintering (in Caribbean Costa Rica) than syntopic year—round residents, and (4) open—country and canopy flycatchers tend to be more opportunistic than forest—interior species. Variation in flycatcher diets with respect to all parameters examined necessitates multiple explanatory hypotheses, and warrants a pluralistic approach to questions of community structure in these birds.