Non-invasive genetic sampling as an effective tool to reveal population fragmentation in the endangered hazel dormouse (Muscardinus avellanarius)

Abstract The hazel dormouse ( Muscardinus avellanarius ) is a locally endangered European glirid species with limited dispersal abilities. Genetic methods have been increasingly applied to measure the effects of habitat fragmentation on the genetic structure of hazel dormouse populations; however, standard collection methods require intensive sampling effort. Here, we present the results of a pilot study testing the feasibility of non-invasive hair and scat collection from nest tubes for dormouse genetic monitoring. In 2022, we collected samples at 73 sites distributed across an approximately 800 km 2 area in a region of southern Germany comprising several isolated forest patches surrounded by agricultural land. Mitochondrial haplotype sequencing and microsatellite analysis resulted in 223 confirmed hazel dormouse detections from a total of 411 collected samples, with samples genetically assigned to 107 different individuals. Statistical testing for population structure was indicative of up to four genetically-differentiated populations corresponding to the main forest patches, suggesting inhibited dispersal between population clusters. The Aisch River valley was also found to be a major barrier to dispersal due to intense agricultural use and the existence of human settlements and traffic infrastructure. Our findings of strong population structuring and lower genetic diversity in isolated habitat patches highlights the greater risk of extirpation for isolated clusters of hazel dormice in Middle Franconia. Non-invasive genetic sampling is an effective tool for studying fragmentation effects on the hazel dormouse and other small mammals, and we provide recommendations for developing optimized sampling strategies that maximize efficiency while reducing researcher effort.