Differential mapping of ringed bird distributions from live resightings versus dead recoveries: an illustration using Eurasian Teal Anas crecca

Abstract

Ringing and visual marking have long been used to delineate bird distribution ranges, but neither method is free from biases linked with differential re-encounter probabilities over the landscape, which may also differ among data collection methods. Modern statistical techniques enable two-dimensional smoothing of the capture/re-encounter information compared with earlier raw ring recovery location maps, improving geographical presentation of areas of greater bird encounter density. However, such technological tools do not solve the issue of the potential biases associated with different data collection methods. We used ring recovery and nasal saddle resighting data from over 14,000 Eurasian Teal Anas crecca ringed in France since 2002 to produce flyway-scale kernel distribution maps to highlight differences between distributions dependent on whether they were derived from recoveries or resighting information, despite abundant data in both cases. Results show that neither large datasets nor modern analytical tools are sufficient to overcome re-encounter probability biases, which require serious evaluation and future dedicated research.