Quantifying effects of field observation on the behavior of secretive species: case study with Timber Rattlesnakes (Crotalus horridus)
Price, Ozzy
Stubbs, Amber
Powers, John
Moench, Ziv
DeSantis, Dominic L.
Department of Biological and Environmental Sciences
Georgia College & State University
Milledgeville, Georgia USA
Observation is known to affect the normal behavior of many free-ranging animals, but the depth of this knowledge is not universal across taxa. The most commonly overlooked groups are often secretive and difficult to study in nature, such as smaller-bodied reptiles. As ambush hunting mesopredators, rattlesnake (Crotalus spp.) foraging success is dependent on remaining unseen by both prey and potential predators. However, the most common method for studying rattlesnakes is radiotelemetry, which involves the consistent relocation and direct observation of individuals at close range and across long timeframes. There is a lack of literature regarding rattlesnake responses to observation, resulting in limited information for establishing best practices to minimize bias and disturbance. We have developed an integrated observational and quasi-experimental study design featuring animal-borne radiotelemetry and accelerometry (RT-ACT) to assess the effects of varying field observation frequencies on movement behavior in Timber Rattlesnakes (Crotalus horridus). A minimum sample of 21 adult rattlesnakes (balanced across sex) from the Cedar Creek Wildlife Management Area (CCWMA) in Putnam County, Georgia implanted with RT-ACT units will be maintained across the study period, with seven individuals being randomly assigned to a cohort at the outset of field monitoring. Field data collection will occur in 2025 between April and August, the foraging season for Timber Rattlesnakes at CCWMA. The cohorts will be systematically rotated though three field data collection intensity schedules (twice/week [low], three/week [medium], four/week [high]) every two weeks. Broad spatial estimates of movement paired with continuous activity measures will enable both scale-dependent analyses of factors that affect within and among individual variation in response to observer presence. Random intercept models will be used to evaluate associations between variation in behavioral response metrics and treatment group while accounting for snake concealment and behavioral state during observation, as well as individual repeatability in responses.