Skip to main content
Dryad

Outer bounds: forest edges emulate vertical strata as a habitat filter for butterfly assemblages.

Data files

May 29, 2025 version files 18.67 KB

Click names to download individual files

Abstract

Understanding the patterns that arise from habitat filtering processes, whether natural or human-induced, is a crucial step in developing strategies for maintaining biodiversity. Tropical forests provide unique habitat conditions in vertical and horizontal dimensions, which consequently affect the composition of local assemblages. In this study, we examined the effect of forest fragmentation and vertical stratification on the phylogenetic and functional diversity of Atlantic Forest butterfly assemblages. We hypothesized that fragmentation positively selects butterfly lineages that are more tolerant to exposed habitats. Likewise, forest canopies would converge to a similar filtering effect since they represent a strong filter for the forest-dwelling species. We analyzed two distinct datasets separately and compared their effects through phylogenetic and functional approaches. We did not detect the effect of fragmentation on species composition, but it has affected lineages composition indirectly via functional traits. Forest vertical structure affected the butterfly assemblage composition, and we detected a phylogenetic signal at the community level, which implies a filtering of butterfly lineages. Phylogenetic diversity substantially decreased from understory to canopy. Butterfly assemblages from fragmented areas and canopies converged to lower body sizes and thorax volumes and presented lower functional diversity. Our results suggest that forest fragmentation and vertical strata have similar filtering effects on species and trait composition of assemblages. It also implies that human-induced processes represent a major threat to tropical butterfly diversity, especially for the lineages that inhabit the understory and continuous forests.

The current dataset contains: 

  • Species composition matrices from previously published studies: (Uehara-Prado et al. 2007) and (Santos et al. 2017)
  • Species functional traits matrices