prilagođeno pretraživanje po punom tekstu

ŠUMARSKI LIST 3-4/2012 str. 33     <-- 33 -->        PDF

We also recorded the sex, potential reproductive status (by females: gravidity or lactation too), approximate age and body mass. Approximate ages were determined based on body mass and overall appearance.
The monitoring of small mammals took place from June 2001 to October 2003, and lasted between two outstandingly high floods (November 2000–June 2004) in comparison with normal spring and autumn high-water periods. Due to the tall and dense undergrowth the use of line transects was appropriate and suitable instead of regular quadrates, moreover, moving away from River Drava the spatial distribution of the individuals of each population and the migration relationships arising from the rising water levels were easier to examine by this method (Fig. 1). In this research, we used the collected data of captures for the examination of the spatial distribution of populations.
Statistical methods – Statističke metode
The relative number and trapping proportions of small mammals from the trapping parameters (number of captures and number of individuals) was counted. We involved those 4 character species in the analysis which were present in the area in all the three sampling years, including Sorex araneus (Linnaeus, 1758) among the shrews. Among the rodents, the three highlighted species were the striped field mouse, Apodemus agrarius (Pallas, 1771), the yellow-necked mouse, Apodemus flavicollis (Melchior, 1834) and the bank vole, Myodes glareolus (Schreber, 1780) which appeared with the highest abundance values. For evaluating the capture data of small mammals we considered the water level data of River Drava which were obtained from the South-Transdanubian Environmental Protection and Water Management Directorate. Water levels were measured along upper Drava, near Őrtilos, west from the sampled floodplain.
We used capture data of transects for the statistical analysis separately and the separated captures of the two areas together. The distribution of the trapping data of transects were tested by chi-square (c2) test for independence. Based on the proportion of captures of the 4 examined species we compared the areas by variance analysis (ANOVA, LSD-test) (Zar 1996). We calculated the habitat preference of the species with Ivlev-index (Ivlev 1961): Px=(a–b)/(a+b), where "a" is the proportion of the caught animals in the given area, "b" is the proportion of the given area in relation to the whole. "Px" is the preference (–1 ≤ Px ≤ 1) or avoidance in each area (+1 is absolute preference, –1 is absolute avoidance). Counting only the Ivlev-indexes is not sufficient for demonstrating whether the preference values of the habitat types are significant or not. Testing significance can be done by the use of Bonferroni z-test.
Spatial association and movement distances of individuals were analyzed by the Biotas 2.0 program. Association was counted between the 4 species in pairs, for A, B and the whole sampling area. For the examination of the movements of the character species’ individuals we gave the coordinates of the transect trap points. Based on these coordinates the Biotas program computed the individuals’ movement distances and patterns, relying on recapture data. When evaluating the movement vectors we counted on the basis of three rodents only from among the 4 sampled species because we recorded only few recaptures in case of the common shrew, therefore had a limited amount of applicable movement vectors.
Spatial and temporal distribution of capture parametersProstorna i vremenska distibucija čimbenika ulova
We registered the presence of 5 shrew (Soricomorpha) and 5 rodent (Rodentia) species between 2001 and 2003 in the sampled floodplain forest of River Drava. The species composition of small mammal community of floodplain forest and the abundance values of the species differed in the three sampling years (Table 1). As seen from the water level dynamics between 2000 and 2005 there was more than 250 cm water level rising in November of 2000 and July of 2004 above the sampled flooding forest which caused the flooding of the trapped area, therefore the start and end of the three-year long capture activity were marked by the two flooding periods. We identified 949 small mammal individuals through the three years and comparing the single sampling years, there was a statistical difference in the number of small mammals. We captured the most individuals in 2002 which significantly differed from both the 2001 and the 2003 summarized abundance values (c2 = 40.23–132.57, P < 0.001). Comparing the two years with lower abundance (correlated with the results of 2002) we received significant difference as well (c2 = 29.25, P < 0.001) which can be a result of the collapse of the small mammal populations in 2003.
Contrasting the capture values with water level data of the Drava suggests that the higher capture success of 2002 may result from the lower water level in January (38.06 a.s.l.) which was the greatest water level decline between the two floods. The second highest water level developed by 2002 December (211.64 a.s.l.) therefore small mammals could reach an expansive dispersal till the trapping months of 2002 which was not affected by the drastic change in water level values. Following the high water in December 2002, water levels followed a decreasing tendency in 2003, due to drier weather. In spite of this, small mammal capture numbers also declined in the summer and autumn of 2003, suggesting that the local density of species decreased even without the negative effect of high water levels, as a result of which the expected autumn density boom, typical of small mammals, did not set in (Fig. 2).
We examined the relative abundance values of the species throughout the given years. Only the species recorded in all 3 years were analysed. Common shrew showed significant differences between the years (c2 = 8.00, P < 0.05) which was due to the large increase of the proportion of this species in 2003. For the abundance of bank vole we received significant differences as well (c2 = 13.43, P < 0.01) comparing the years, which