Jump to Main Content
Spatiotemporal identification of roadkill probability and systematic conservation planning
- Lin, Yu-Pin, Anthony, Johnathen, Lin, Wei-Chih, Lien, Wan-Yu, Petway, Joy R., Lin, Te-En
- Landscape ecology 2019 v.34 no.4 pp. 717-735
- cost effectiveness, geographical distribution, graphs, planning, reptiles, risk, road kills, seasonal variation, simulation models, surveys, uncertainty
- CONTEXT: Accurate spatiotemporal modeling of roadkill hotspots is essential for the assessment of high risk roadkill locations. Increasing the spatiotemporal resolution of models may facilitate greater cost-effective solutions for roadkill mitigation strategies. OBJECTIVE: This study develops a novel spatiotemporal roadkill distribution model to simulate roadkill probability. Moreover, we systematically identify top prioritized road segments by the most frequent roadkill occurrence for multiple focal species. METHODS: Based on the theory of the Poisson process, the proposed spatiotemporal roadkill distribution model with seasonal effects is validated with four focal reptilian species. The model simulates spatiotemporal roadkill patterns and addresses uncertainty by referencing ensemble species distribution models. Finally, we systematically prioritize road segments by the most frequent roadkill occurrence for multiple focal species. RESULTS: The efficacy of the proposed spatiotemporal roadkill distribution model which is validated in terms of the area under the receiver operating characteristic curve (AUC) and accurate proportions. The AUC values based independent roadkill data tests ranged from 0.73 to 0.84. Both the efficacy of the proposed model, and the increases in uncertainty are attributable to decreasing seasonal sampling size and variation. Based on the independent roadkill data, more than 70% of roadkill events occurred within the top 30% priority segments by our approaches. CONCLUSIONS: The proposed model is successfully applied in simulation of spatiotemporal roadkill probability. The seasonal effects benefit identification of high roadkill probability. Through the systematic identification and the proposed model, our approach provides useful information for the design of cost-effective surveys and appropriate conservation planning and mitigation strategies.