Simulating animal movements to predict wildlife-vehicle collisions: illustrating an application of the novel R package SiMRiv

• Published in: European journal of wildlife research Vol. 65; no. 6; pp. 1 - 12
• Main Authors: Quaglietta, Lorenzo, Porto, Miguel, Ford, Adam T.
• Format: Journal Article
• Language: English
• Published: Berlin/Heidelberg Springer Berlin Heidelberg 01.12.2019; Springer Nature B.V
• Subjects: accidents; Animal models; Biomedical and Life Sciences; Computer simulation; computer software; Dependence; Ecology; Fish & Wildlife Biology & Management; Heterogeneity; humans; Infrastructure; Landscape; landscapes; Life Sciences; Lutra lutra; Mathematical models; mortality; Movement ecology; Original Article; Otters; prediction; Predictions; Risk; Road Ecology; road kills; roads; Roads & highways; simulation models; Wildlife; Wildlife management; Zoology; Resistance; Individual-based mechanistic movement simulation models; Movement ecology; Landscape connectivity; Road-kill hotspots; Road ecology
• ISSN: 1612-4642; 1439-0574
• DOI: 10.1007/s10344-019-1333-z

In conservation, there is a high demand for methods to predict how animals respond to human infrastructure, such as estimating the location of road mortalities and evaluating the effectiveness of mitigation measures. Computer-based simulation models have emerged as an important tool in understanding wildlife-infrastructure interactions. Such models, however, often assume animal omniscience of the landscape yielding unrealistic movements, focus more on genetic connectivity than actual movement paths, or are case-specific and mathematically/computationally challenging to apply. Here, we illustrate the potential of SiMRiv , a novel R package for simulating spatially explicit, individual multistate (Markovian) movements incorporating landscape heterogeneity, in the subject of road ecology. In particular, we used SiMRiv to reproduce wildlife movement patterns and predict high-risk areas for road-kill, using Eurasian otters ( Lutra lutra ) as a model species. We compared the number of road crossings in real otter movements and null models (simulated, multistate Markovian movements) incorporating the effect of the landscape structure (here, water dependence). The number of road crossings in real and simulated movements was remarkably similar, and available limited real road-kill data supported SiMRiv ’s road-kill risk predictions. Further, other emergent movement properties were also very similar in real and simulated movements. Overall, results show that SiMRiv has potential for reconstructing real wildlife movement patterns, as well as for predicting road-kill risk areas. SiMRiv constitutes a flexible, powerful, and intuitive tool to help biologists and managers to test mechanistic hypotheses on wildlife movement ecology, including those related to wildlife-vehicle interactions.