Context: The measurement of home range size and configuration has been a powerful and enduring method of quantifying animal-habitat relationships. Traditionally, home range estimators have been built using bivariate location data (e.g., x-y coordinates) which inherently assumes that animal movement is two-dimensional (2D). However, this is not representative of real-world systems.
Objectives: Home range estimators that make these assumptions may underestimate animal home range size where 2D+ movement is probable. Our objective is to evaluate if landscape complexity impacts the accuracy of traditional home range estimators.
Methods: We randomly sampled 50,000 animal home range extents at four different spatial scales (5, 10, 100, and 250 km2). We then quantified 2D and 2D+ home range estimates for each home range extent and quantified landscape complexity within.
Results: When landscape complexity was low (i.e., global terrain ruggedness index [TRI] < 5.23) 2D and 2D+ home ranges were not statistically (α < 0.05) different. However, when landscapes were complex (i.e., TRI ≥ 5.23), 2D home ranges significantly underestimated home range area (range 5% to 83.66%). Importantly, 18% of the world’s terrestrial surface exceeds TRI level of 5.23.
Conclusions: We found that as landscape complexity increased, so did the percent difference between 2D and 2D+ home range estimates across all spatial scales. We recommend that landscape complexity should be considered when modeling animal home ranges. By doing so, quantitative biases may be mitigated.