Affiliation:
1. Universidade Federal de São Paulo Diadema Brazil
2. Universidade Federal do ABC Santo André Brazil
Abstract
AbstractThe increasing development of highways severely threatens several vertebrates due to habitat fragmentation, edge effects and roadkills. Understanding how the landscape influences the occurrence of roadkills on highways can provide insights into the preferred routes of fauna dispersion and define the priority stretches to implement mitigation actions. This study evaluated the association between landscape characteristics and fauna roadkills on the BR‐158 highway (State of Mato Grosso do Sul, Brazil), where the adjacent vegetation encompasses phytophysiognomies of the Cerrado, a severely threatened biome. The spatial distribution of recorded road‐killed vertebrates (secondary data) was analysed using the statistical tool 2D Hotspot Identification, which defined the stretches with the highest mortality (hotspots). The surrounding landscape within a 1 km buffer zone around the highway was characterized based on area and distance information considering the land use and cover. Generalized linear models were applied to test the influence of locations on roadkill, considering the landscape characteristics as independent variables and the presence/absence of roadkill as the dependent variable. Our results indicated two hotspot stretches along the BR‐158 highway and a higher probability of roadkill in stretches associated with riparian forests and more distant from urban areas. The identification of landscape characteristics associated with the occurrence of vertebrate roadkills can help guide the planning and monitoring of highways. Furthermore, in dry ecosystems such as the Cerrado, we must provide information to implement priority actions in stretches of riparian forest, aiming to mitigate the adverse effects of highways on local biodiversity conservation.
Subject
Ecology,Ecology, Evolution, Behavior and Systematics
Cited by
1 articles.
订阅此论文施引文献
订阅此论文施引文献,注册后可以免费订阅5篇论文的施引文献,订阅后可以查看论文全部施引文献