Phenology and temperature are the main drivers shaping the detection probability of the common wall lizard

Abstract

Measuring the abundance of organisms is essential to provide information to ecology and biodiversity conservation. Hardly ever, the probability of detecting an animal during a survey is near one. Overlooking this observational process can lead to biased estimates of population size and vital rates. In this study, through Bayesian modeling, I evaluated the effects of temperature, precipitation, wind, humidity, and phenology in determining changes in the detection probability of the common wall lizard, for which studies on the factors determining detection probability are currently not available. Additionally, I tested for two possible interactions: date-temperature and date-humidity, in order to assess if the relationships of these variables with detection probability vary through the sampling season. Detection probability was highest earlier in the season (April) and between 24 and 28 degrees. Rainfall during the survey showed a negative effect on detection probability. In contrast, cumulative precipitation in the 24 hours before the survey showed a positive relationship, indicating that lizards are easier to detect in surveys after rainy days. Furthermore, date and temperature showed a positive interaction, indicating that the relationship between detectability and temperature changed over the sampling season. Date and humidity showed a negative interaction: late in the sampling season, detectability was higher with lower humidity, however, this relationship was not found in the early season. Future studies can consider multiple sites to evaluate the extent of variation in the drivers of detection probability and to assess the factors related to abundance.