Monitoring frequency influences the analysis of resting behaviour in a forest carnivore

Abstract

Resting sites are key structures for many mammalian species, which can affect reproduction, survival, population density, and even species persistence in human-modified landscapes. As a consequence, an increasing number of studies has estimated patterns of resting site use by mammals, as well as the processes underlying these patterns, though the impact of sampling design on such estimates remain poorly understood. Here we address this issue empirically, based on data from 21 common genets radiotracked during 28 months in Mediterranean forest landscapes. Daily radiotracking data was thinned to simulate every other day and weekly monitoring frequencies, and then used to evaluate the impact of sampling regime on estimates of resting site use. Results showed that lower monitoring frequencies were associated with major underestimates of the average number of resting sites per animal, and of site reuse rates and sharing frequency, though no effect was detected on the percentage use of resting site types. Monitoring frequency also had a major impact on estimates of environmental effects on resting site selection, with decreasing monitoring frequencies resulting in higher model uncertainty and reduced power to identify significant explanatory variables. Our results suggest that variation in monitoring frequency may have had a strong impact on intra- and interspecific differences in resting site use patterns detected in previous studies. Given the errors and uncertainties associated with low monitoring frequencies, we recommend that daily or at least every other day monitoring should be used whenever possible in studies estimating resting site use patterns by mammals.