Assessing Error in Locations of Conspicuous Wildlife Using Handheld GPS Units and Location Offset Methods

摘要

Collecting spatially explicit locations of individual animals often is an important part of the study of habitat use. Obtaining accurate locations without disturbing an individual can be difficult for small species and may be limited for species of conservation concern, such as piping plover (Charadrius melodus), where a close approach is undesirable because of the potential for disturbance. To reduce disturbance while estimating an animal's location, an observer can collect their location using Global Positioning System (GPS) and offset that position using a distance and azimuth. Typically, error is not considered when these locations are determined, despite the potential effects of inaccuracy on habitat association results. Therefore, our objectives were to quantify potential error using the offset method and then evaluate how that error may manifest. During the plover breeding season in 2017 (Apr through Sep), we tested the error of Trimble GPS units compared with benchmark locations to assess the accuracy of locations derived from these units. We then assessed the error associated with offset locations of a small target using Trimble GPS units, laser rangefinders, and 2 compass types. Finally, we determined the potential consequence of unaccounted error in our system by comparing the difference between point locations and land‐cover data using our estimated error as a point buffer. Average error of the GPS units at benchmark locations was 0.95m. The mean error of the offset locations increased with increasing observer distance from the decoy location (from urn:x-wiley:19385463:media:wsb1055:wsb1055-math-0001=2.9m to 7.6m at distances of 20m and 100m, respectively). Error also increased with increasing error in the distance and azimuth measurements, and was greater using a digital GPS compass as compared with a handheld magnetic compass corrected for magnetic declination. In addition, potential misclassification of land‐cover type increased with increasing potential location error. When modeling animal locations, the error of point locations using this method, especially for land‐cover at the point location, should be accounted for using appropriately sized position buffers. Using this location collection method, we can increase our knowledge and study of conspicuous species to ensure that we consider habitats used throughout the species' life stages.