Studies of niche partitioning based on diet have a long history in ecology. However, many of these studies use dietary proxies, such as foraging behavior, and fail to ensure that the proxies correspond with actual resource differences. Furthermore, actual diet studies stop at quantifying diet overlap without testing for phenotypic differences (e.g. behavior, morphology, etc.) that could account for the diet differences. Here we combine available prey, foraging behavior, and diet in a niche specialization framework, using four coexisting wood warblers (Parulidae) wintering in Jamaica. We compared foraging height, tactics, and substrate (branch tips/leaves, bark, airspace) and quantified abundances of arthropod taxa using these substrates. We combined these two data sets to generate expected diets for each species, and used an emetic to obtain observed diets based on stomach contents. We predicted that (1) all four species would differ in foraging behavior, facilitating coexistence, (2) expected diets would predict observed diets better than several null and alternative models, and (3) differences in expected diets between species would correspond to differences in observed diet.
Results/Conclusions
We found distinct differences as predicted in foraging behavior among all four warbler species. Dietary overlap was high between all species pairs, largely driven by high consumption of ants. Despite these high overlaps, distinct differences were apparent in diet between all warbler species pairs. Expected diets predicted observed diets, and performed better than several other potential models. Finally, differences in expected arthropod consumption among bird species based on foraging (substrate use) generally predicted differences in observed arthropods consumed. The one exception to this was two pairwise comparisons involving black-throated-blue warblers, which consumed unpredicted high numbers of nematoceran flies. Coupled with demonstrable winter food limitation within some of these warbler species, our finding of high dietary overlap indicates that interspecific competition is likely important in this system, and the impacts of such competition may be exacerbated by the high dependence on an abundantly available, but probably low-profit prey (small ants). We also conclude that evolved differences in foraging behavior directly generate differences in resource use, likely contributing to the ecological coexistence of these closely related, but food-limited migratory birds.