Detection probabilities of the native Eurasian otter and the invasive American mink are independent of their co-occurrence

Abstract

Detection probability is an important input when studying and managing species. It reduces the miscalculations of the estimates of occupancy, population abundances and their trends inherent to the imperfect detection during sampling. However, detection probability may vary in relation to environmental and biological factors, such as the presence of other species of similar ecological niche. We used a two-species single-season occupancy model to analyse the detection probability of the native Eurasian otter Lutra lutra and the invasive American mink Neovison vison in the northeast of the Iberian Peninsula and to test for potential differences in their detection when co-occurring and when not. Despite the known behavioural change of American mink towards a more terrestrial diet and a more diurnal activity pattern in the presence of Eurasian otter, their detectability did not vary when both species co-occurred versus when they did not. However, we observed seasonal differences in the probability of detection. Specifically, the probability of detection was higher in winter for both species. Furthermore, Eurasian otter was more often detected in sites with clay substrate, whereas American mink in sites with clay and rocks. These results indicate that the effectiveness of monitoring of Eurasian otter and American mink remains independent of the species co-occurrence. Yet, seasonal variations in the detection should be taken into account when monitoring.