Road-Based Density Estimation


Abundance is perhaps the most fundamental piece of data needed to effectively manage or conserve wildlife populations. However, some snake species are so secretive (i.e., low individual capture probability) that traditional abundance estimation methods, such as capture-mark-recapture (CMR) are unlikely to ever yield meaningful abundance information. Not surprisingly, these species are often also those that are perceived to be rare or declining, and thus are of high conservation concern. Understanding abundance of these species requires development of innovative field and analytical techniques that overcome low recapture probability. We recently developed a novel abundance estimation method that takes advantage of the fact that many rare and secretive snakes are most frequently encountered crossing roads (Willson et al. 2018).


​Our technique uses an understanding of snake spatial movement patterns and road crossing behavior to estimate snake density from frequency of snake observations during systematic road surveys. In short, our method is based on the idea that we can reasonably assume a snake will be detected during a road survey if its road crossing location and time coincide with the survey vehicle. Thus, to translate encounter frequencies into densities, we need to know 1) the probability that a snake will cross the road during a survey and 2) our probability of detecting a snake that does cross during the survey time. To estimate the probability that a snake will cross the road during a survey, we created individual-based spatial movement simulation models and parameterized those models with information derived from species-specific radiotelemetry data (i.e., movement distance, frequency, orientation towards home range center and road). Next, we use data on snake road crossing speed and average vehicle speed to calculate the probability that a snake will be detected if it crossed the survey route during a survey. Taken together, these pieces of information allow us to generate a relationship between encounter frequency of snakes on roads and density, which can be compared to empirical road data for the species to generate a density estimate and explore sensitivity of that estimate to assumptions and variation in model parameters. We developed this method using existing data for imperiled Southern Hognose Snakes (Heterodon simus) in the sandhills of North Carolina. With support from the US DoD (ESTCP and Legacy Resource Management Programs), we are currently applying our method to estimate abundance of other species of secretive upland snakes in the Southeast, as well as invasive Burmese pythons in the Florida Everglades.


Research Areas:

  • Developing novel methods for estimating abundance of secretive snakes from road survey data

  • Using road-based density estimation to estimate abundance of rare upland snakes and invasive Burmese pythons

  • Validating road-based density estimation in snakes and other model systems


Willson, J.D., S. Pittman, J. Beane, and T. Tuberville. 2018. A novel approach to estimating densities of secretive species from road-survey and spatial-movement data. Wildlife Research 45:446-456.

Primary Collaborators and Students:

Shannon Pittman (Athens State University), Jen Mortensen, Brett DeGregorio, Tracey Tuberville (SREL), Jeff Beane (NC State Museum), Chris Petersen (DoD), Jeff Lovich (DoD)

Popular Coverage:

Savannah morning news - Natural Georgia

UA Newswire

Partners & Funding Sources:

  • Department of Defense (ESTCP and Legacy Resource Management Programs)

  • Savannah River Ecology Lab

  • USGS

  • University of Florida