Modelling changes in forest attributes driven by human activities at different spatial scales in the subtropical Atlantic Forest

Abstract

Human activities are the main drivers of biotic homogenization, thus affecting ecosystem functions. In this study, we aimed to investigate the relationship among anthropogenic disturbances and forest attributes. Moreover, we sought to identify direct and indirect effects of topographic heterogeneity (TH), more specifically of standard deviation of altitude, on anthropogenic disturbances and forest attributes, respectively, through a path analysis. We used data gathered on 186 systematically distributed sample plots located in the Brazilian subtropical Evergreen Rainforest. We selected 14 predictor variables related to anthropogenic disturbances aiming to model seven forest attributes related to species diversity, composition, and structure. The aboveground biomass, rarefied species richness, proportion of standing dead trees, and proportion of individuals of pioneer species in the regeneration layer were better predicted by global linear regression models. The proportions of individuals of pioneer species in the canopy layer and of threatened species in the canopy and regeneration layers were better explained by local geographically weighted regression models. Human activities at different spatial scales may lead to disturbances (e.g., edge effects and habitat fragmentation), thus driving changes in forest attributes. The land use amidst forest remnants was related to biomass production and biotic homogenization. Logging and road networks may imperil the maintenance of threatened species. The path analysis showed that TH indirectly affects species richness via pasture area. Furthermore, topography appeared to act as a barrier for the expansion of certain human activities over the landscape. Our findings evoked the need for implementation of adaptative conservation strategies.