Roads and highways represent one of the most important anthropogenic impacts on natural areas and contribute to habitat fragmentation, because they are linear features that can inhibit animal movement, thereby causing barrier effects by subdividing the populations adjacent to the roads. The study presented here aims to determine, to which extent roads act as a barrier, subdividing populations of three species of small forest mammals: bank vole, yellow-necked mouse and common shrew, and what is the relative importance of road width and traffic intensity on the barrier effect. The study was carried out at four 25 m long segments of roads, close to the city of České Budějovice. All segments crossed a forest. The capture-recapture method was applied to determine the crossing rates of animals. The traps were checked three times each day during four consecutive nights, in summer and in autumn. We found that: (1) roads strongly prevent crossing movements in all three studied species, (2) there are interspecific differences in road crossing rates, (3) species cross more often narrow than wide roads, (4) traffic intensity does not affect the crossing rates.
Between 1996 and 1998, microhabitat selection by three small mammal species was studied in oak-elm forest using the catch-mark-release (CMR) method. Microhabitat selection by these species was assessed by Discriminant Function Analysis (DFA) and Canonical Correspondence Analysis (CCA). Apodemus flavicollis was shown to prefer dense young forest and shrubs, but the presence of dead woody material was also important. In Microtus arvalis the data confirmed a negative relationship with stands with higher tree and shrub cover and a preference for open meadows. Clethrionomys glareolus was found to prefer non-fragmented tree microhabitats, preferring sites with developed undergrowth and ample hiding places amongst fallen logs and branches. Although C. glareolus did not change its microhabitat requirements during the year (spring to autumn), within forest stands it was connected with undergrowth regardless of species composition; this highlights the importance of undergrowth for C. glareolus. Seasonal changes in microhabitat selection could not be confirmed unambiguously.
The vertical activity of small mammals was studied in Central Bohemia, the Czech Republic, from 2003 to 2005. In total, 424 individuals were captured by the CMR technique, in live traps placed at ground level or in trees at heights of 1 and 2 m. The most commonly captured was the yellow-necked mouse, Apodemus flavicollis (96.7 %). Other species captured in the trees were: the common dormouse, Muscardinus avellanarius and the pygmy shrew, Sorex minutus</i. Although mice captures on the ground (considered as 100 %) prevailed significantly over those in trees (84.2 % and 77.9 % at 1 and 2 m, respectively), vertical activity was still quite extensive. In the peak abundance year, mice visited trees more frequently than in the year with lower abundance. There was no clear seasonal variation in vertical activity during the May–October period. No significant difference between male and female vertical activity was found. There was a slight but non-significant positive relation between the weight of an individual and the frequency of its arboreal captures. Any preference for climbing a particular tree species was not found. Our results clearly demonstrate that tree climbing by the yellow-necked mouse represents an important component of its movement activities and this fact should be considered in future studies of its ecology.