Postcapture movement rates can inform data-censoring protocols for GPS-collared animals

Abstract

Wildlife research often requires the capture, chemical immobilization, and handling of free-ranging animals. These activities have been documented to yield detrimental effects among some study subjects and may bias study results. We developed an empirical censoring protocol that investigated the postcapture movement rates of 89 white-tailed deer (Odocoileus virginianus) outfitted with global positioning system (GPS) collars to identify data that should be censored from analyses due to postcapture recovery and acclimation. Average daily postcapture movement rates were significantly lower than average daily movement rates for a given calendar day for approximately 14 days postcapture. We tested the influence of biased data in a simple random walk simulation and found that net displacement values were significantly lower after 1,000 iterations when an uncensored data set was used than for data sets that did not include the first 2 weeks of GPS locations. Our results indicate that study animals exhibited sublethal effects of capture, handling, and instrument acclimation (reduced movement rates) for approximately 14 days postcapture. We recommend that all studies involving marked animals assess the influence of capture, handling, and instrumentation on postcapture movements to guard against data sets that are not representative of normal movement behavior. Inclusion of these data in movement analyses may bias results, especially when data sets are small (,90 days).