High-resolution MaxEnt modelling of habitat suitability for maternity colonies of the barbastelle bat Barbastella barbastellus (Schreber, 1774) in Rhineland-Palatinate, Germany

摘要

The barbastelle bat Barbastella barbastellus (Schreber, 1774), probably one of the rarest of western European bat species, has suffered from substantial population declines over the last several decades. In fact, it was believed to be extinct within the federal state of Rhineland-Palatinate(western Germany) until the discovery of a maternity colony in 2004. More reproduction sites have since been found, which demonstrates a substantial knowledge gap about the actual distribution and abundance of the species in Rhineland-Palatinate. Suitable habitats for maternity colonies arecrucial for the survival of a population and knowledge of their location is critical for conservation. We modelled the suitability of habitats for use by maternity colonies in Rhineland-Palatinate based on high-resolution data of the forest structure and roosting sites of maternity colonies,using the presence-only machine learning approach MaxEnt. In addition to statistical tests of the model performance, we analysed general occurrence surveys from the last few years for evidence of barbastelle and conducted an in-situ survey on one of the sites identified as highly suitableby the model, but for which no occurrence records exist. On this site, we discovered a new maternity colony. Analysis of third-party surveys resulted in two recently discovered colonies, which shows the barbastelle's range is not restricted to the area south of the Moselle River. The resultsof our study along with the scattered pattern of potentially suitable locations for maternity colonies in the region challenge previous assumptions of the geographic distribution of barbastelle in Rhineland-Palatinate. This study demonstrates the potential of habitat suitability modellingin conservation ecology and the results may provide a basis for future preservation strategies in the region.