Landscape and traffic factors affecting animal road mortality

Abstract

Roadkill is a significant indicator in realizing the impacts of roads on adjacent ecosystems. This study undertakes a comprehensive survey of roadkills in Kinmen (Taiwan) and analyzes their causes. Two models, Traffic Flow Model and Geometric Model, combined with animal road-crossing behaviors, are used to derive survival probability. Survey results and model predictions yield similar results for moderate traffic flow and agree in bird and small mammal roadkill frequency prediction. It is found that traffic volume, adjacent landscape and road condition are the major contributing factors related to roadkills. Higher traffic volume near habitats always augments the probability of roadkill; however, roadside trees, adjacent landscapes, and road longitudinal slope also affect the probability of successful crossing by small animals, especially birds. This study also proposes measures, to be applied to future road planning and design, aimed to lower roadkill probability.

Keyword : roadkill, ecological conservation, traffic flow theory, Kinmen, geometric model