Habitat specialisation and matrix resistance predict responses of butterfly populations to landscape features in tropical grassland–forest complexes

Ravi Jambhekar, Kavita Isvaran | 2022

Abstract

A fundamental question in ecology is to understand how a species is distributed in a landscape. In terrestrial landscapes, the size and isolation of habitat patches, and matrix properties, are thought to drive population density patterns. Yet, given the same set of landscape features, why do species from a single taxon vary so widely in their density patterns? A primary hypothesis for such variation, proposed by community-level studies, is that key ecological characteristics of species influence how they respond to landscape features. However, robust tests of this hypothesis, which require measurements of populations of multiple species in an assemblage, are still scarce. We investigated the ability of ecological specialisation and the interaction of species with the matrix (matrix resistance) to predict population responses of butterfly species to patch size and connectivity in naturally heterogeneous tropical forest–grassland complexes. We surveyed 56 habitat patches in a 65 sq. km area by laying 276 transects, along which the identity and abundances of butterfly species were recorded. We also used transects that cut across two habitats to estimate matrix resistance. We find that habitat specialisation predicted the strength of area-density and isolation-density relationships. Matrix resistance also predicted variation in area-density relationships, highlighting the importance of species interactions with the matrix. Specialists showed higher matrix resistance and stronger area and isolation effects than did generalists. Our findings suggest that investigating how traits related to ecological specialisation and matrix resistance affect demographic parameters can contribute towards understanding mechanisms underlying species distributions in heterogeneous landscapes.