No abstract

Surveys to determine the presence of a species may be used as the basis for monitoring programs, wildlife‐habitat models, or management decisions. Errors in survey data can result from nondetections, where the species is present at the site but not detected in the survey. We designed an experiment to estimate the probability of detection of red‐backed salamanders (Plethodon cinereus) using different sampling methods and salamander densities. Red‐backed salamanders were introduced into enclosures (n = 124) that were randomly assigned to a sampling treatment (daytime cover‐object searches or nighttime visual searches) and a density treatment (low or high densities). Probability of detection was significantly higher using daytime searches and in enclosures with higher salamander densities. Weather conditions influenced the probability of detection. Increases in temperature correlated with decreases in detection during daytime sampling. Low humidity limited detection during nighttime surveys. Recent rainfall was positively related to detection during nighttime surveys. Probabilities of detection estimates are important for future design and analysis of survey data. Because of the high mean probability of detection, daytime surveys of cover objects are recommended as the survey method of choice if only a small number of samples can be obtained. However, because of their lower variability, nighttime visual surveys are the recommended sampling method when multiple samples can be obtained. Daytime surveys may be most effective at determining presence or absence at a site, but nighttime surveys may be more useful for indices of population trends. Nondetections can be reduced by ensuring large sampling areas for daytime searches and multiple samples at the same location for nighttime searches.

No abstract

No abstract