B. Papp szerk.: Studia Botanica Hungarica 32. 2001 (Budapest, 2001)

Dobolyi, Konstantin: Phytosociological and ecological evaluation of selected habitats of Telekia speciosa (Schreb.) Baumg. (Asteraceae)

Table 1 (continued) No Site Date Altitude Exposure Angle Parent material (m s. m.) of slope (°) 69 Bodoki Mts 16.09.1981 850 NE 30 andésite 70 Bodoki Mts 16.09.1981 800 E 25 andésite tuff 71 Bodoki Mts 17.09.1981 750 W 5 andésite 72 Vitosa Mts 27.06.1981 800 NE 5 granit 73 Bükk Mts 04.08.1981 300 0 limestone 74 Bükk Mts 20.06.1981 300 0 limestone 75 Bükk Mts 31.07.1981 400 0 limestone 76 Bükk Mts 31.07.1981 450 0 clay shale 77 Bükk Mts 31.07.1981 450 0 clay shale 78 Bükk Mts 31.07.1981 450 0 clay shale The relevés were analysed by the following multivariate numerical methods with SYN-TAX 5.1 and SYN-TAX 2000 program packages (PODANI 1993, 2001). 1) hierarchical classifications: — distance optimizing procedures: single link, complete link, group average (UPGMA), — global optimization, — ordinal clustering, 2) ordinations: — centered principal component analysis (PCA), — principal coordinate analysis (metric multidimensional scaling) (PCoA), — correspondence analysis (COA), — symmetric weighting, 3) block clustering (of raw data matrices), sum of squares. The applied distance coefficients (used as dissimilarity): Euclidean distance, Jaccard, S0rensen, Bray-Curtis, similarity ratio, Goodman-Kruskal's gamma, Kendall's tau, Podani's discordance (PODANI 1997). All mentioned multivariate analyses were performed on two data matrices: "Matrix A": relevé/species matrix: 78 objects/235 variables. The original cover values according to Braun-Blanquet have been converted to a numerical scale of 0 to 6 in the following way: absence^), +->1, 1^2, 2-»3, 3->4, 4->5, 5->6. "Matrix B": relevé/phytosociological character matrix: 78 objects/32 variables. The values were calculated in the following way: cover values of species having the same phytosociological character were summed pro relevés. Based on multivariate analyses (classifications, ordinations and block clustering) four groups of relevés were separated. (These groups are not identical with those in DOBOLYI (1983).) The groups of relevés were analysed statistically on the basis of the following characteristics of their species: life form, phytogeographical area (floral element) (HORVÁTH et al. 1995), phytosociological character, social behaviour type and ecological indicator values (BORHIDI 1993). (BORHIDI has applied ELLENBERG'S relative ecological indicator value system to the Hungarian