Skip to main content
U.S. flag

An official website of the United States government

Detection rates of geckos in visual surveys: Turning confounding variables into useful knowledge

December 1, 2015

Transect surveys without some means of estimating detection probabilities generate population size indices prone to bias because survey conditions differ in time and space. Knowing what causes such bias can help guide the collection of relevant survey covariates, correct the survey data, anticipate situations where bias might be unacceptably large, and elucidate the ecology of target species. We used negative binomial regression to evaluate confounding variables for gecko (primarily Hemidactylus frenatus and Lepidodactylus lugubris) counts on 220-m-long transects surveyed at night, primarily for snakes, on 9,475 occasions. Searchers differed in gecko detection rates by up to a factor of six. The worst and best headlamps differed by a factor of at least two. Strong winds had a negative effect potentially as large as those of searchers or headlamps. More geckos were seen during wet weather conditions, but the effect size was small. Compared with a detection nadir during waxing gibbous (nearly full) moons above the horizon, we saw 28% more geckos during waning crescent moons below the horizon. A sine function suggested that we saw 24% more geckos at the end of the wet season than at the end of the dry season. Fluctuations on a longer timescale also were verified. Disturbingly, corrected data exhibited strong short-term fluctuations that covariates apparently failed to capture. Although some biases can be addressed with measured covariates, others will be difficult to eliminate as a significant source of error in longterm monitoring programs.

Publication Year 2016
Title Detection rates of geckos in visual surveys: Turning confounding variables into useful knowledge
DOI 10.1670/14-048
Authors Bjorn Lardner, Gordon H. Rodda, Amy A. Yackel Adams, Julie A. Savidge, Robert N. Reed
Publication Type Article
Publication Subtype Journal Article
Series Title Journal of Herpetology
Index ID 70162129
Record Source USGS Publications Warehouse
USGS Organization Fort Collins Science Center