PS 28-66 - Predicting patterns of native snake occurrence along Main Park Road in Everglades National Park, Florida

Thursday, August 11, 2016
ESA Exhibit Hall, Ft Lauderdale Convention Center
Charmaine Pedrozo1, Jeanelle Brisbane1, Christina Romagosa2, Nichole Bishop3, Ray Carthy4 and Brian J. Smith5, (1)Doris Duke Conservation Scholars Program, University of Florida, Gainesville, FL, (2)Wildlife Ecology and Conservation, University of Florida, Gainesville, FL, (3)School of Natural Resources & Environment, University of Florida, Gainesville, FL, (4)FL Coop Fish and Wildlife Unit, University of Florida, Gainesville, FL, (5)Wildife Ecology & Conservation, University of Florida, Davie, FL
Background/Question/Methods

Everglades National Park (ENP) is well-known for its high diversity and abundance of snakes. Across its 1.5 million acres of land, there are 24 snake species: 22 native and 2 non-native. Much of ENP is inaccessible wilderness, so the primary method for surveying snakes is by conducting driving surveys along Main Park Road (MPR). The purpose of this study was to determine the environmental factors that affect the probability of detecting native snakes while road cruising along MPR. Identification of the important factors related to snake detection can help optimize snake surveys. Data were collected from June to July of 2015 during 14 nights of road cruising surveys. Sampling was done between 20:00 h and 2:00 h. For every live snake encountered, species, time of day, GPS coordinates, and environmental data (air temperature, relative humidity, solar radiation, precipitation, wind, water level, moon phase) were recorded. We built a logistic regression model set to test for the factors that were important in predicting hourly probability of snake occurrence. We built a Poisson regression model set to determine the factors predicting the average number of snakes/h. The best model in each set was selected using AIC scores.

Results/Conclusions

Our best model from the logistic regression set indicated that temperature and hour were the important factors in predicting probability of snake occurrence. Based on the hour of the survey, the probability of seeing snakes increased at the start of the survey until it hit the optimum, the third hour, and then declined later in the evening. Warmer temperatures increased these probabilities. Our best model from the Poisson regression model set indicated that hour was the sole predictor of average snakes/h. The model predicted a peak in snakes/h in the third hour of the survey, consistent with the prediction from the logistic regression model. Our results suggested that timing of road cruising surveys for ENP’s native snakes may be the most important factor to consider. However, environmental conditions such as temperature might also play a role in snake detection. Determining the factors that affect the probability of snake occurrence will be useful in designing a long-term monitoring program for native species.