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 respect of the mean number of morning captures, no significant difference was revealed by ANOVA (H = 4.97, NS) between the four successive morning samplings, based on the 1995-1996 15-month data series. Comparisons with data of 1997, when each trapping period extended over 5 nights, revealed that the 4th and 5th mornings had a significantly higher value of mean number of morning captures than the 1st and the 2nd (H = 20.22, p < 0.05). The number of morning recaptures was analysed on species basis in both periods. In the 4-night period none of the species exhibited significant variation with respect to the number of morning recaptures recorded during the four successive occasions (H = 3.32-6.64, NS). No significant difference was revealed by ANOVA between data of the five mornings in 1997 in A. sylvaticus and C. glareolus (H = 4.45-11.74, NS). In A. flavicollis the proportion of recaptures on the 5th morning was significantly higher than on the 1st and the 2nd (H = 16.6, p < 0.05), while in A. agrárius it was mornings 4 and 5 that were represented with significantly higher recapture values then days 1 and 2 (H = 18.04, p < 0.05). y = -9.4022Ln(x) + 87.132 R J = 0.3611 O p < 0.05 O 100 •£80 £60 100 T 3" 80 -^** S 40 a. eu S 20 O o y = -6.23Ln(x) + 98.242 R 2 = 0.3861 ° p < 0.1 A. agrárius 50 100^, mw . , 150 H /D Model Fig. 5. Jolly-trappability as a function of estimated population size (D-model) in 1995-1996 Values of Jolly trappability in the entire 1995-1996 trapping period were as follows: A. flavicollis: 57.4 %, A agrárius: 55.75 %,A. sylvaticus: 25.92 %, C. glareolus: 45.57 %. Its monthly values were plotted against the population sizes estimated by the D model (Fig. 5) in the case of two species. Negative logarithmic correlation was found at the error level 5% in A. flavicollis, and at 10% in A. agrárius, which suggests that if the whole period is considered, trappability decreases with increasing population size. However, the early summer interval with low abundance values breaks the trapping period into two terms. If the relationship between trappability and estimated population size is analysed separately for the period starting at the first trapping and lasting until the summer minimum, and for the one lasting from the second half of the summer until the last trapping occassion, it is found that smaller population size values are accompanied by lower trappability. In the autumn period of population buildup trappability values increased with increasing population size, but beyond a certain limit further increase in abundance did not bring about a notable increase in trappability. As data were fewer in the two separate regression analyses of the two shorter periods, negative and positive correlations found for these tendencies were statistically not significant. Values of maximum trappability in the 1995-1996 period were higher than Jolly trappability, which is in accordance with expectations, since values of the former are influenced by individuals with a single capture and no recapture. For the same reason maximal trappability is given only for the entire period: A. flavicollis 80.65 %, A. agrárius 76.25 %, A. sylvaticus 64.04 %, C. glareolus 67.49 %. Due to the lack of sufficient number of recapture data, the correlation between estimated population size and maximum trappability was not analysed.
