This repository contains the analysis for the 1st chapter of my PhD. You can find the code used for the analyses within the RMarkdown files within the RMarkdown folder.
The Data folder contains the .csv file (or .rds) for the dataframe used for the full analysess. The data can also be used to go through the material in Appendix 1 to practice home range analysis using continuous-time movement modelling and the autocorrelated kernel density estimator.
The Intermediate folder contains .rds files for the statistical models, movement models, variograms, and home range estimates. These can be read into R in case one would like to skip running the lengthy calculations and speed up computation time during the analyses.
The goal of this chapter is to assess how irregular sampling resulted from collecting data via handheld GPS influences home range estimation.
We did this by thinning high quality subsets of the data to create fake sampling regimes that are aimed to reflect possible sampling designs in the field. Then we measure the performance of the home range estimates by calculating the total data from the high quality samples that are represented within the 95% level UD home range estimates.
We then see how performance varies depending on sampling characteristics such as the number of unique weeks and the number of recorded locations in the sampling regime. Finally, we assess the influence of sampling on estimates of the effective sample size, which is approximately the number of home range crossings observed in the data. This enables us to better understand how sampling and underlying movement behaviors may influence the vulnerability of datasets to bias in home range estimation.