Evaluation of micro-GPS receivers for tracking small-bodied mammals

Abstract

GPS telemetry markedly enhances the temporal and spatial resolution of animal location data, and recent advances in micro-GPS receivers permit their deployment on small mammals. One such technological advance, snapshot technology, allows for improved battery life by reducing the time to first fix via postponing recovery of satellite ephemeris (satellite location) data and processing of locations. However, no previous work has employed snapshot technology for small, terrestrial mammals. We evaluated performance of two types of micro-GPS (< 20 g) receivers (traditional and snapshot) on a small, semi-fossorial lagomorph, the pygmy rabbit (Brachylagus idahoensis), to understand how GPS errors might influence fine-scale assessments of space use and habitat selection. During stationary tests, microtopography (i.e., burrows) and satellite geometry had the largest influence on GPS fix success rate (FSR) and location error (LE). There was no difference between FSR while animals wore the GPS collars above ground (determined via light sensors) and FSR generated during stationary, above-ground trials, suggesting that animal behavior other than burrowing did not markedly influence micro-GPS errors. In our study, traditional micro-GPS receivers demonstrated similar FSR and LE to snapshot receivers, however, snapshot receivers operated inconsistently due to battery and software failures. In contrast, the initial traditional receivers deployed on animals experienced some breakages, but a modified collar design consistently functioned as expected. If such problems were resolved, snapshot technology could reduce the tradeoff between fix interval and battery life that occurs with traditional micro-GPS receivers. Our results suggest that micro-GPS receivers are capable of addressing questions about space use and resource selection by small mammals, but that additional techniques might be needed to identify use of habitat structures (e.g., burrows, tree cavities, rock crevices) that could affect micro-GPS performance and bias study results.

Fig 1. GPS data collection by GPS receivers at a stationary test location at Cedar Gulch.

GPS data collection by snapshot (blue circles) and traditional (yellow circles) at one stationary test location. Stationary test locations (red Xs) were distributed throughout the Cedar Gulch study site. Darker regions in the aerial imagery acquired from the National Agriculture Imagery Program (NAIP) display areas of denser shrub cover.

Fig 2. Micro-GPS receivers deployed in this study.

(A) G10 UltraLITE GPS Logger (snapshot GPS receivers) weighed 11 g. (B) FLR V GM502030 GPS Logger (traditional GPS receivers) weighed 14–16 g.

Fig 3. Relationship between duration (hr) from the start of the stationary tests and micro-GPS accuracy (log of location error [LE]).

Traditional micro-GPS receivers (left) displayed consistent performance across the entire 0range of duration time, whereas snapshot micro-GPS receivers (right) generated large LE values at the start of each stationary test before leveling to a more consistent performance.

Fig 4. Performance of micro-GPS receiver during stationary tests for locations in burrows and above ground.

Mean values (±SE) of (A) fix success rate (%) and (B) location error (m) for traditional and snapshot micro-GPS receivers placed on the ground surface and in burrows at a depth of 25 cm.

Fig 5. Mean (±SE) light values (lux) relative to micro-GPS receiver performance.

When collars with both micro-GPS receivers and light loggers were deployed on free-ranging pygmy rabbits, successfully acquired locations (Success) had greater light levels than missed fixes (Fail).

Fig 6. Fix success rates (±SE) of micro-GPS receivers deployed on free-ranging pygmy rabbits (field) deemed to be above ground and stationary tests (stationary).

“Above ground” fix attempts were classified using a predetermined light threshold.