AN EXPERIMENTAL STUDY OF HABITAT SELECTION BY BIRDS IN A COFFEE PLANTATION

Unique components of tropical habitats, such as abundant vascular epiphytes, influence the distribution of species and can contribute to the high diversity of many animal groups in the tropics. However, the role of such features in habitat selection and demography of individual species has not been established. Understanding the mechanisms of habitat selection requires both experimental manipulation of habitat structure and detailed estimation of the behavioral and demographic response of animals, e.g., changes in movement patterns and survival probabilities. Such studies have not been conducted in natural tropical forest, perhaps because of high habitat heterogeneity, high species diversity, and low abundances of potential target species. Agroforestry systems support a less diverse flora, with greater spatial homogeneity which, in turn, harbors lower overall species diversity with greater numerical dominance of common species, than natural forests. Furthermore, agroforestry systems are already extensively managed and lend themselves easily to larger scale habitat manipulations than protected natural forest. Thus, agroforestry systems provide a good model environment for beginning to understand processes underlying habitat selection in tropical forest animals. Here, we use multistate, capture–recapture models to investigate how the experimental removal of epiphytes affected monthly movement and survival probabilities of two resident bird species (Common Bush‐Tanager [ Chlorospingus ophthalmicus ] and Golden‐crowned Warbler [ Basileuterus culicivorus ]) in a Mexican shade coffee plantation. We established two paired plots of epiphyte removal and control. We found that Bush‐Tanagers were at least five times more likely to emigrate from plots where epiphytes were removed compared to control plots. Habitat‐specific movement patterns were not detected in the warbler. However, unlike the Golden‐crowned Warbler, Common Bush‐Tanagers depend upon epiphytes for nest sites and (seasonally) for foraging. These dispersal patterns imply that active habitat selection based on the presence or absence of epiphytes occurs in C. ophthalmicus on our study area. Survival rates did not vary with habitat in either species. Interestingly, in both species, survival was higher in the nonbreeding season, when birds were in mixed‐species flocks. Movement by Common Bush‐Tanagers into areas with epiphytes occurred mostly during the breeding season, when mortality‐driven opportunity was greatest.