Landscape-Scale Forest Loss Promotes the Taxonomic Homogenization of Bird Assemblages in a Human-Modified Mountain Region

The ongoing deforestation process across the globe is reducing the extent of suitable habitat for forest-specialist species. The cross-habitat spillover hypothesis posits that in such a scenario, some species may be compelled to use supplementary resources from the adjacent anthropogenic matrix. Consequently, the compositional differentiation (beta diversity) between forest remnants and the matrix should decrease (i.e. biotic homogenization) in more deforested landscapes. We tested this prediction by assessing bird assemblages in a mountain region of Guerrero, Mexico. We surveyed birds in nine landscapes with different forest cover. Within each landscape, we measured bird beta diversity (D β ) between forest fragments and the surrounding anthropogenic matrix and then assessed the relationship between D β and landscape forest cover. We separately assessed the complete bird assemblage, and forest-specialist and habitat-generalist birds, because the cross-habitat spillover hypothesis posits that the loss of D β in more deforested landscapes should be particularly evident when assessing forest-dependent birds. The generalized linear models indicated that, as expected, D β decreased in landscapes with lower forest cover. Such a decrease was significant when assessing the complete bird assemblage and forest-specialist birds, but not when assessing habitat-generalist species. Our findings support the cross-habitat spillover hypothesis and indicate that forest loss contributes to the homogenization of bird assemblages in human-modified landscapes. Such homogenization process could also be related to an alternative but non-exclusive mechanism: the extirpation of rare, non-ubiquitous forest-specialist species in more deforested landscapes. To conserve bird assemblages in human-modified landscapes, we should prevent forest loss and promote adequate management strategies (e.g. leaving standing native trees, avoiding hunting, and removing feral predators) to prevent threats to forest-specialist species when they use the matrix.