Modeling wildlife movement, especially through human-dominated landscapes, will help us understand how urbanization patterns affect species and how to better plan incorporate wildlife pathways through our roads and developments.
The most thorough expression of successful movement and breeding across large landscapes is through the exchange of DNA from one population to another. With colleagues, I have examined numerous landscape genetics methods to determine their appropriateness for estimating landscape features that facilitate or impede movement.
We have also developed a novel method to model multi-scale movement using GPS telemetry data. This multi-scale path selection function can help us understand what scales, or effect zones, species are responding to for different landscape features. For example, a species may be responding to natural features at fine scales, but roads and the road effect zone at larger scales. We have also promoted the combination of pathway data and genetic data in estimating resistance to wildlife movement to capture large scale processes, such as dispersal and mating, as well as fine scaled movement decision, such as where to cross a road.
As part of this path selection function research, we looked at how the frequency at which GPS data points are collected affect our understanding of animal movements. We found that longer intervals between fixes, sometimes even at 1-hour intervals, can introduce a surprising amount of bias into our understanding of animal movement — especially for highly mobile species.