Validating a high-throughput tracking system: ATLAS as a regional-scale alternative to GPS

Abstract

Fine-scale tracking of animal movement is important to understand the proximate mechanisms of animal behaviour. While GPS tracking is an excellent tool for measuring animal movement, it often causes a trade-off between tag weight, cost and lifespan that limits its application to relatively large individuals, a small number of individuals or short tracking durations, respectively. The reverse-GPS system, ATLAS, uses lighter, cheaper tags that last long periods of time at the similar sampling frequencies to GPS. ATLAS provides a regional-scale alternative to global GPS-based tracking with which hundreds of relatively small-bodied species can be tracked simultaneously for long periods of time. Here, we test the accuracy and precision of an ATLAS system, using concurrent GPS measurements as a reference. To measure accuracy, we calculated the distance between ATLAS and GPS location estimates for a route (mobile test) and 16 fixed locations (stationary test) on the Griend mudflat, the Netherlands. This location is the focus of a large-scale ATLAS system in the Wadden Sea consisting of 26 receivers and covering 1326 km2 of intertidal region, with almost no physical obstacles for radio signals. ATLAS-derived location estimates differed on average 4.2 m from GPS-estimated stationary test sites and 5.7 m from GPS tracks taken whilst moving between them. Signals that were collected by more receiver stations were more accurate, although even 3-receiver localisations were comparable with GPS localisations (~10 m difference). Higher receiver stations detected the tag at further distances. In conclusion, ATLAS provides an alternative to GPS for regional-scale animal movement studies by providing movement data of comparable spatial accuracy. The growing use of the ATLAS systems motivates further refinement of best-practice application, such as considering the height of receivers, their spatial arrangement, density and the movement mode of the study species (e.g., ground-dwelling or flying).

Spatial cognitive ability is associated with transitory movement speed but not straightness during the early stages of exploration

“…Since pheasants move relatively slowly (less than 0.5 m/s, unpublished data), we felt this was an acceptable reduction in temporal resolution to improve overall accuracy. ATLAS has been reported to have a median accuracy of 5-10 m [36].…”

“…We assayed the spatial cognitive ability of a cohort of pheasant chicks in early life using a five-arm radial maze with a single baited arm. Later, we released these birds into the wild with a soft-release protocol [33,35], and monitored their movements using the reverse-GPS system, ATLAS [36,37]. At our study site, these birds are limited in access to informed individuals who they could follow, since adults are caught for breeding and released elsewhere.…”

A Guide to Pre-Processing High-Throughput Animal Tracking Data

“…The package is based on , a fast implementation of data frames; thus it is compatible with a number of data structures from popular packages including , , and objects, which can be converted to data frames (Boone et al, 2020; Calenge et al, 2009; Kranstauber et al, 2011). These pre-processing techniques and package were designed with ATLAS systems in mind, motivated to meet the rapid growth of studies using this high-throughput system worldwide: in Israel (Corl et al, 2020; Toledo et al, 2014, 2016, 2020; Vilk et al, 2021), the UK (Beardsworth et al, 2021b,c), and the Netherlands (Beardsworth et al, 2021a). However, the principles and functions presented here are ready for use with other massive high-resolution data collected by GPS, reverse-GPS or any other high-throughput tracking system.…”

“…3.a; Listing 1). A bounding box filter does not require a geospatial representation such as a shapefile, and can help remove unreliable data from a tracking system that is less accurate beyond a certain range (Beardsworth et al, 2021a). In some special cases, users may wish to remove positions inside a bounding box, either because movement behaviour within an area is not the focus of a study, or because positions recorded within an area are known to be erroneous.…”

“…7). We collected the calibration data (n = 50,816) on foot and by boat, with a hand-held WATLAS tag (sampling interval = 1s) around the island of Griend (53.25°N, 5.25°E) in August 2020 (WATLAS: Wadden Sea ATLAS system Beardsworth et al, 2021a, Bijleveld et al in prep. ).…”