Marshall University
Factors affecting detection of a cryptic ambush predator using camera traps, with implications for snake monitoring
Abstract
dc:description.abstractCamera traps are becoming increasingly popular in wildlife research because they can be deployed for long time periods with little maintenance and are much less invasive than traditional trapping methods. However, they are primarily used to study large mammals and have not been extensively tested for snake species, many of which are cryptic and understudied. The Adapted-Hunt Drift Fence Technique (AHDriFT) arrays were designed to improve detection of herpetofauna by attaching camera traps to drift fences. In the summer of 2022, we installed 20 AHDriFT arrays at the Marine Corps Recruit Depot Parris Island in coastal South Carolina where we are concurrently conducting a long-term mark-recapture and radio telemetry study on the imperiled eastern diamondback rattlesnake (Crotalus adamanteus, EDB). Our goal was to assess the effectiveness of the AHDriFT for surveying for EDBs by determining which factors influence detection probability of EDBs on camera traps. We used logistic regression and AICC model selection to assess covariates of detection probability at weekly and monthly time scales. Covariates included weather variables (temperature, precipitation, cloud cover, humidity), change in day length, camera functionality, EDB behavioral seasons, and the distance of telemetered EDBs, number of EDBs captured, and amount of time EDBs were within 150 meters of an array. Overall, the number of EDBs detected by camera traps was low compared to the number of EDBs captured by research personnel during the same time period. At the weekly scale, change in day length was the most important predictor of detection probability, with detection probability increasing in the fall as photoperiod decreased. At the monthly scale, EDB behavioral season and the amount of time spent near an array were the most important predictors of detection probability, with detection probability decreasing in the dormant season relative to the foraging and breeding seasons, and increasing with the amount of time EDBs were within 150 m of arrays. Thus, detection probability appears to be most strongly influenced by predictable seasonal patterns of movement that in turn are driven by the change in photoperiod, rather than weather conditions. Because EDBs are ambush predators which is a largely stationary foraging strategy, the amount of time EDBs spend near an array was predictive of detection probability while EDB proximity was not. Due to low detection probability at both the monthly and weekly scale, we do not recommend this method for surveys of this species, unless used in conjunction with another long-term monitoring method or if used for additional, more actively-foraging species.
Degree
thesis:*- Name thesis:degree_name
- Biological Sciences
- Level thesis:degree_level
- Thesis
- Discipline thesis:degree_discipline
- College of Science
Author and committee
dc:creator, dc:contributor.*- Author dc:creator
-
- Ebert, Sarah Elizabeth
- Contributors dc:contributor
-
- Dr. Jayme Waldron
- Dr. Shane Welch
- Dr. Kyle Palmquist
Subjects
dc:subject × 9Identifiers
dc:identifier.*- Repository record dc:identifier
- https://mds.marshall.edu/etd/2119