Beech masting modifies the response of rodents to forest management

We investigated the impact of forest harvest and tree masting on the abundance and habitat selection of yellow-necked mice (Apodemus flavicollis) and bank voles (Myodes glareolus). Our study was conducted in managed European beech (Fagus sylvatica) forest, in four closed-canopy and four shelterwood stands, during two mast and two non-mast years. We predicted that (1) forest harvest will positively affect vegetation cover, resulting in increased abundance of yellow-necked mice; (2) abundance of bank voles will not be affected by forest harvest; (3) the increased abundance of yellow-necked mice in shelterwood vs. closed-canopy stands will be more pronounced after non-mast (low rodent abundance) than after mast years (high rodent abundance), and (4) both species of rodents will select microhabitats with protective cover (provided by coarse woody debris or dense vegetation), but this preference will be stronger after non-mast than after mast years. In agreement with the first prediction, the abundance of yellow-necked mice tended to be higher in shelterwood than in closed-canopy stands, and was positively associated with average vegetation cover (generally denser in shelterwood than in closed-canopy stands). The second prediction was partially supported: while there was no clear effect of forest harvest on bank vole abundance, it was negatively affected by soil scarification conducted at two shelterwood stands. In concurrence with predictions (3) and (4), habitat associations of both rodent species were considerably weaker after mast years at both stand and microhabitat scales. Unexpectedly, after mast years, there was a negative association between bank vole abundance and vegetation cover, perhaps resulting from interference competition with yellow-necked mice. Our findings demonstrate that masting-related fluctuations in rodent abundance strongly influence their patterns of habitat selection.