L. Forró szerk.: Miscellanea Zoologica Hungarica 12. 1998 (Budapest, 1998)

Horváth, Gy.: Population dynamics and trappability of four rodent species in a forest habitat

In the 1997 trapping period Jolly trappability had higher values than the maximum trappability. This is absolutely in contradiction with the former period and our expectations, and it was caused by the fact that the D-model underestimated the number of marked individuals and population sizes in this period, in comparison with MNA. For none of the species was a significant correlation between population size and either of the trappability indices found. Values of Jolly trappability for the entire period are as folows: A. flavicollis 69 %, A. agrárius 48 %, A. sylvaticus 50.18 %, C. glareolus 65.24 %, while maximal trappability values are: A. flavicollis 37 %, A. agrárius 30 %, A. sylvaticus 37 %, C. glareolus 38 %. Discussion The demographic changes of the forest rodent populations in the sampling area during the 15 four-night trapping periods in 1995-1996 and the 10 five-night periods in 1997 were described and presented. The sampling method was not identical in the two periods (difference in trap numbers and the duration of trap operation), thus it is not possible to combine their evaluation. The dominance of A. agrárius at the end of 1995 was shown; in autumn it was trapped in high numbers throughout the sampling area, and its abundance declined early next year. The analysis of population dynamics of A. agrárius during this period is reported by Horváth et al. (1997), who emphasized the rapid population growth. Earlier studies have also touched upon this characteristic demographic fluctuation of the species (Andrzejewski & Wroclawek 1961, Babiska-Werka et al. 1981). Since numbers of other rodent species in the area remained low in comparison with A. agrárius, its peaking population in 1995 in the area can be interpreted as one existing in allopatry with the other rodent species. Löfgren (1995) reported on C. glareolus allopatry with its dominance in a boreal habitat, in a study showing the niche-expansion of this species in the absence of competitors. Similarly, in our case of A. agrárius dominance, competitive relations probably have played an important role. Earlier studies have described the strong dispersal and adaptive capacities of A. agrárius, and this species has also been reported as a competitor of C. glareolus (Gliwicz 1981). The marked decrease of populations in 1996 resulted in a distortion of certain calculated parameters. Estimates by the D model of the JOLLY programme were higher in 1996 than the MNA values, while in 1997 they fell below the MNA figures. Nichols & Pollock (1983) pointed out that a condition of the acceptability of MNA method as an estimation is to achieve 100% capture, and the Jolly-Seber estimator usually gives higher values. This was found to be in accordance with our 1995-1996 data, but the opposite situation was experienced in 1997 which fact probably roots in the relation of our 1997 data to the JOLLY programme. The same problem was encountered when calculating the trappability index. Since the JOLLY programme (which considers deaths and accidentally released individuals too) underestimated the number of marked individuals, the value of the Jolly-trappability index had to be smaller than maximal trappability. According to our results trappability did not depend on the material of the traps: no significant difference was found between the capturing capacity of wooden and plastic traps. The test with different species yielded similar results in this respect. Only few animals were captured on the first two days of the trapping sessions in the 19951996 15 month sampling period. The number of captures rose on the following days of the trappings, with the peak occurring on the 4th day, although the difference between the days