Road networks affect wildlife in various ways. Animals are at risk of road mortality due to collisions with vehicles, roads act as a barrier for their seasonal and daily migrations, and cause various disturbance effects on wildlife habitat. This knowledge notwithstanding road networks continue to increase worldwide.
The dilemma was caused, among other reasons, by the quality of previous road ecology research, because: (1) the research questions addressed were not relevant for decision making. (2) The study designs used were of low quality preventing useful inferences for road construction. (3) Many studies focused on local-scale effects of a single road, neglecting land-scape-scale interactions a dilemma, because effects on population persistence can only be evaluated studying large-scale processes, and because the most pressing policy decisions are at the landscape scale.
The thesis in hand develops a research Agenda for future road ecology research. I present five research questions of direct relevance for road construction in practise. For any research question I present hypothetical study designs increasing inferential strength of the results.
Based on the Agenda, the thesis in hand contributes to the knowledge about landscape-scale road ecology research. A monitoring study in the federal state of Hesse, Germany, analyses the development of road networks from 1930 to 2002. Carried out on the level of administrative districts and natural areas the monitoring is based on index effective mesh size, and serves as a basis for periodic updates. I show that road networks affect wildlife populations. The abundance of roe deer (Capreolus capreolus), wild boar (Sus scrofa), fox (Vulpes vulpes), and badger (Meles meles) populations is small in administrative districts with a high degree of frag-mentation, while at the same time number of road-kills increase. A study on brown hare (Lepus europaeus) in the Swiss Canton Aargau shows a similar pattern: The higher the density of heavily used roads, the smaller population abundances. Against this background I develop models predicting hotspots of vehicle-wildlife collisions on the basis of road and landscape data. Wildlife accidents with roe deer and wild boar take place in dawn, especially on roads with intermediate traffic densities. Hotspots are woodland-field interfaces frequented by animals when changing in-between resting places and pastures. Models are an essential basis for bundling mitigation efforts in the context of programs aiming at the de-fragmentation of the present road network.
Verknüpfung zu Publikationen oder weiteren Datensätzen
