Breeding mute swan habitat selection when accounting for detectability: a plastic behaviour consistent with rapidly expanding populations

A number of native and exotic animal species show dramatic population increases in terms of both numbers and geographic range. Understanding the habitat selection processes behind such increases is crucial to implement adequate management measures. Mute swan (Cygnus olor) populations have experienced a tremendous demographic and geographic expansion in Western Europe during the twentieth century, colonizing a wide variety of aquatic habitats. We aimed at assessing how swans select nesting sites during the pre-laying and laying periods on medium to large fishponds (from 10 to 50 ha) in Eastern France, while accounting for detectability biases and testing for the effects of fishpond spatial configuration, vegetation resources, human disturbance and habitat management. Our results demonstrate that the mute swan is a non-selective species regarding its nesting habitat among such fishponds, using these independently from the parameters considered although fishpond characteristics varied. Although mute swan is one of the least cryptic Anatidae, owing to its white colour and large size, detection of breeding pairs remained imperfect for each over several sampling occasions. However, because we repeated the sampling sessions, detection of swan pairs by the end of the monitoring period was as high as 0.94. These results are consistent with previous assertions that the mute swan is a species of high ecological plasticity, which may partly explain its recent colonization rates. Given that even swan breeding events were imperfectly detected on each occasion, we highlight the fact that most studies of breeding ducks (which are more cryptic) would be considerably improved by better considering detection biases.