A multi-scale approach reveals differential responses of birds to vegetation structure in dry forest on livestock ranches

Abstract

Dry forests are among the most threatened ecosystems globally, facing severe degradation from long-term human use. This degradation alters vegetation structure, composition, and spatial configuration, impacting biodiversity, ecosystem functions, and local livelihoods. However, its influence on birds at multiple scales remains poorly understood, limiting the effectiveness of conservation strategies. In this study, we examined how changes in vertical (local) and horizontal (landscape) vegetation structure influence bird communities and foraging height guilds (ground-, understory-, and canopy-foraging birds) in the threatened Espinal forest. We collected bird and local forest structure data from 30 livestock ranches along a gradient of grazing intensity (6 sites per ranch). We also estimated NDVI texture indices to characterize horizontal vegetation structure. Our results show that bird communities responded positively to increases in vertical and horizontal vegetation structure. Canopy cover, grass cover, and the spatial distribution of plant biomass emerged as key factors explaining bird responses. Foraging height guilds, however, responded differently from bird communities. Canopy- and ground-foraging birds were more sensitive to vertical vegetation structure, showing a positive relationship with canopy cover and a non-linear relationship with DBH, respectively. Understory-foraging birds benefited from shrub-dominated forests and were strongly related to the horizontal vegetation structure. Moreover, the responses of the bird community, canopy- and ground-foraging birds to local vegetation were modulated by biomass distribution at the landscape scale. The results highlight the importance of maintaining vegetation heterogeneity across spatial scales. Preserving old-growth trees, high grass cover, and dense vegetation patches can support bird communities in livestock-dominated dry forests.