EN
Estimating animal population size is a critical task in both wildlife management and conservation biology. Precise and unbiased estimates are nonetheless mostly difficult to obtain, as estimates based on abundance over unit area are frequently inflated due to the “edge effect” bias. This may lead to the implementation of inappropriate management and conservation decisions. In an attempt to obtain an as accurate and conservative as possible picture of Eurasian otter (Lutra lutra) numbers, we combined radio tracking data from a subset of tracked individuals from an extensive project on otter ecology performed in Southern Portugal with information stemming from other data sources, including trapping, carcasses, direct observation of tagged and untagged individuals, relatedness estimates among genotyped individuals, and a minor contribution from non-invasive genetic sampling. In 158 km of water network, which covers a sampling area of 161 km2 and corresponds to the minimum convex polygon constructed around the locations of five radio-tracked females, 21 animals were estimated to exist. They included the five radio-tracked, reproducing females and six adult males. Density estimates varied from one otter per 3.71–7.80 km of river length (one adult otter per 7.09–14.36 km) to one otter per 7.67–7.93 km2 of range, depending on the method and scale of analysis. Possible biases and implications of methods used for estimating density of otters and other organisms living in linear habitats are highlighted, providing recommendations on the issue.