Ábrahám Levente (szerk.): Válogatott tanulmányok VI. - Natura Somogyiensis 19. (Kaposvár, 2010)
HORVÁTH GY., HERCZEG R., TAMÁSI K. & SALI N.: Nestedness of small mammal assemblages and role of indicator species in isolated marshland habitats
HORVÁTH ET AL.: SMALL MAMMALS 289 when all individuals of a species are found in a single group of sites, and when the species occurs in all sites of that group, it is a symmetrical indicator (its presence contributes to the habitat specificity and its presence can be predicted in all sites of the group, IndVal >55 %). Other species must be considered as accidental ones, these are asymmetrical indicators (their presence cannot be predicted in all samples of one habitat, but contributes to the habitat specificity, IndVal <55 %). The statistical significance of the species indicator values is evaluated using a randomisation procedure by 1000 random permutations. The IndVal2 package ( DUFRÉNE & LEGENDRE 1997) was used for these computations. Results Pattern of nestedness The number of small mammal species differed through the three examined years as well as the number of sampling patches within the macro-habitats. This determined the size of given matrices that we used by the exploration of nestedness patterns in the analysis. The program estimated higher weighted-interaction nestedness (WIN) values for every year than we got from the randomisation process (d rn J ). Therefore, all of the nestedness patterns of small mammal assemblages of the three years indicated that the counted nestedness values can not be random, but they differed significantly from it. This assumption was proved by the significance of z scores. The statistical results of compared nestedness values of random and original data matrix showed that the small mammal assemblages of the isolated Kis-Balaton marshland habitat patches are ordered contrary to the random pattern, and showed structured pattern (Table 2). The value of the weighted-interaction nestedness estimator was positive in all three years, which expressed the relative position between the maximum nestedness and the random pattern. The highest value of this estimator was calculated for the year of 2006 which therefore showed the greatest nestedness, however the relative value of this arose Table 2: Summarised statistic of nestedness analysis based on Weighted-Interaction Nestedness Estimator (WINE) Nestedness / years 2005 2006 2007 WIN 0.334 0.432 0.325 d max 0.400 0.472 0.401 drnd 0.266 0.350 0.268 z-score 4.503 5.407 2.778 P - value 3.34E-06 3.19E-08 0.002 n (WINE) 0.505 0.672 0.428 WIN: weighted-interaction nestedness t] (WINE): weighted-intercation nestedness estimator dnuix- weighted-interaction nestedness of the maximal nestedness matrix. d r nj. average weighted-interaction nestedness of random replicates z-score: z score of the weighted-interaction nestedness P - value: probability of the matrix having WIN value, less than expected by chance
