1. Dude, where's my cow? : using high-frequency movement data to quantify animal space use
- Author
-
Munden, Rhys, Potts, Jonathan, Borger, Luca, and Wilson, Rory
- Abstract
In this thesis I investigate the use of high frequency (≥1Hz) location data for examining animal movements. This high frequency means that an animal's movement path can often be considered pseudo-continuous, which removes the necessity for interpolating between successive recordings. On the other hand, many existing techniques used to analyse animal movement data are either too computationally expensive or not directly applicable to high frequency data. I introduce novel adaptations of two techniques, namely the residence time metric and Step Selection Analysis (SSA), which I tailor for use on high frequency data. The residence time metric is used to estimate how long an animal spends in different areas. I explain how using high frequency data enables us to accurately calculate how much time an animal spends in a clearly delimited area. I also show how, through adaptations to the existing method used to calculate residence times, we can greatly reduce its computational time. Furthermore, this adapted method enables us to identify specific areas that are extensively used by an animal, which provides further opportunities to examine why an animal chooses to use those areas so much. SSA enables a user to infer the reasons underpinning an animal's decisions to move from one location to another. I explain how it is not useful to apply SSA directly to high frequency data, since there is often little interesting information contained in movements between consecutive recordings that are a second or less apart. Therefore, I first show how high frequency data can be rarefied such that the rarefied points represent times when the animal has made key movement decisions. I then adapt the SSA method for use with such rarefied paths. In comparing the existing and adapted SSA methods I find that the previous SSA method can lead to inaccurate results.
- Published
- 2020