Regional and local patterns of genetic variation and structure in yellow‐necked mice ‐ the roles of geographic distance, population abundance, and winter severity

• Main Authors: Sylwia D. Czarnomska, Magdalena Niedziałkowska, Tomasz Borowik, Bogumiła Jędrzejewska
• Format: Article
• Language: English
• Published: Wiley 2018-08-01
• Series: Ecology and Evolution
• Subjects: Apodemus flavicollis; genetic structure; January temperature; microsatellites; northeastern Poland; population density
• Online Access: https://doi.org/10.1002/ece3.4291

Abstract The goal of this study, conducted in seven large woodlands and three areas with small woodlots in northeastern Poland in 2004–2008, was to infer genetic structure in yellow‐necked mouse Apodemus flavicollis population and to evaluate the roles of environmental and population ecology variables in shaping the spatial pattern of genetic variation using 768 samples genotyped at 13 microsatellite loci. Genetic variation was very high in all studied regions. The primal genetic subdivision was observed between the northern and the southern parts of the study area, which harbored two major clusters and the intermediate area of highly admixed individuals. The probability of assignment of individual mice to the northern cluster increased significantly with lower temperatures of January and July and declined in regions with higher proportion of deciduous and mixed forests. Despite the detected structure, genetic differentiation among regions was very low. Fine‐scale structure was shaped by the population density, whereas higher level structure was mainly shaped by geographic distance. Genetic similarity indices were highly influenced by mouse abundance (which positively correlated with the share of deciduous forests in the studied regions) and exhibited the greatest change between 0 and 1 km in the forests, 0 and 5 km in small woodlots. Isolation by distance pattern, calculated among regions, was highly significant but such relationship between genetic and geographic distance was much weaker, and held the linearity at very fine scale (~1.5 km), when analyses were conducted at individual level.