ctmmweb: A Graphical User Interface for Autocorrelation-Informed Home Range Estimation

Estimating animal home ranges is a primary purpose of collecting tracking data. Many widely used home range estimators, including conventional kernel density estimators, assume independently-sampled data. In stark contrast, modern animal tracking datasets are almost always strongly autocorrelated. The incongruence between estimator assumptions and empirical reality often leads to systematically underestimated home ranges. Autocorrelated kernel density estimation (AKDE) directly models the observed autocorrelation structure of tracking data during home range estimation, and has been shown to perform accurately across a broad range of tracking datasets. However, compared to conventional estimators, AKDE requires additional modeling steps and has heretofore only been accessible via the command-line ctmm R package. Here, we introduce ctmmweb, which provides a point-and-click graphical interface to ctmm and streamlines AKDE, its prerequisite autocorrelation modeling steps, and a number of additional movement analyses. We demonstrate ctmmweb's capabilities, including AKDE home range estimation and subsequent home range overlap analysis, on a dataset of four jaguars from the Brazilian Pantanal tracked between 2013 and 2015. We intend ctmmweb to open AKDE and related autocorrelation-explicit analyses to a wider audience of wildlife and conservation professionals. (c) 2021 The Authors. Wildlife Society Bulletin published by Wiley Periodicals LLC on behalf of The Wildlife Society.

Automated summary

Estimating animal home ranges is often done using conventional kernel density estimators, which assume independently-sampled data. However, modern tracking datasets are usually strongly autocorrelated, leading to underestimated home ranges. Autocorrelated kernel density estimation (AKDE) directly models the observed autocorrelation structure of tracking data, and has shown to accurately estimate home ranges.