Performance of two GPS telemetry collars under different habitat conditions

The recent removal of Selective Availability from satellite signals has allowed for more accurate Global Positioning System (GPS) locations without the use of differential correction. Previous studies of GPS collars have documented decreased success rates and increased positional error with increasing canopy closure, which may confound analysis of telemetry data. We tested 2 companies' collars (Lotek, Inc. and Advanced Telemetry Systems, Inc. [ATS]) for success rates, proportion of 3-dimensional (3D) fixes, and positional error under 6 habitat types with varying canopy characteristics. On average, collars were highly successful (99% Lotek and 93% ATS) and positional error (14 m [Lotek] and 16 m [ATS]) was within the manufacturers' specifications. Most (>88%) locations for both collars had positional errors of <25m. Success rates differed between habitat types for ATS (P = 0.014) and Lotek (P<0.001) collars. The mean proportion of 3D fixes was higher for ATS collars (90%) than for Lotek (64%). Proportion of 3D fixes differed between habitats and canopy-closure classes for both collars. Success rate was negatively related to basal area but not to canopy closure. Positional error was not strongly influenced by plot basal area. Both collars performed well and offer cost-effective telemetry for medium and large animals; however, success rates were lower in plots with dense canopies. Therefore, use of areas with dense canopy closure by telemetered animals may be underestimated.