Kovács I. (szerk.): A Magyar Természettudományi Múzeum évkönyve 78. (Budapest 1986)
Szujkó-Lacza, J. ; Rajczy, M.: Similarity investigations on a loess steppe fragment in a forest-steppe in Hungary
The Salvio (nutanti-nemorosae) —Festucetum rupicolae pannonicum community (ZÓLYOMI 1958) and our steppe fragment have 28 species in common, while the Salvio (nutantinemorosae) —Festucetum rupicolae tibiscense and our steppe fragment have 37 (independently of the coenosystematical rank of any species). That is, our steppe fragment is more related to the tibiscense variant of this plant community. The zonal steppes "Kazatzkaja step" and "Streletzkaja step" in Ukraine and our steppe spot have 46 species in common (DOHMAN 1. c). According to DOHMAN, 6 of these, Agropyron intermedium, Festuca rupicola, Koeleria cristata, Phleum phleoides, Phlomis tuberosa and Verbascum lychnitis are true steppe species. The rest are "steppe-forest-steppe" ("steplesostep") or forest-steppe species. An interesting phenomenon is the absence of any Stipa species, though the S-Ukrainian steppes are very rich in species of this genus. The same thing was recognized by ZÓLYOMI (1958, 1962) who found only one species —Stipa capillata —in both geographical variant of the loess steppe community (Salvio-Festucetum rupicolae) in Hungary. Analysis of the relevé series Taking into consideration only the presence-absence of the species (using the resemblance function of Jaccard) we obtained the following clusters at the 5-clusters level; 1; 2; 3-12; 13-17; 18-20 (Fig. 1 — the numbers represent the serial numbers of the relevés). As a result the first two relevés standing alone are separated from all the rest and from each other as well. If we neglect the first and the second separated relevés 3 clusters are left, the beginning, the middle and the end of the relevé series. At the 4-clusters level (1; 2; 3-17; 18-20) the main part of the relevé series is divided to two parts with a boundary between relevés 17 and 18. If the cover values of the species are taken into account as well, a more complex aspect of the matter is revealed. The cluster analysis based on the resemblance function of Hummon at 6-clusters level segregates the following clusters: 1 ; 2; 4-5; 3,6-15; 16-17; 18-20 (Fig. 2). The separation of the relevés 1 and 2 is the same as in the previous case, but there are differences afterwards, two small clusters being separated. At 4-clusters level we obtained the two main groups again, but with the boundary between relevés 15 and 16! The dendrogram produced by the cluster analysis based on the resemblance function of Horn (Fig. 3.) and that of Czekanowski (Fig. 4.) are very similar at the examined levels (5-clusters level: 1; 2; 3; 6-13; 4-5; 14-15; 16-20-4-clusters level: 1; 2; 3-15; 16-20). Studying the dendrograms first it must be stated that the relevés 1 and 2 do not belong to the examined steppe plant community. Most probably microclimate (the vicinity of the valley) and the accompanying Polygonato-Quercetum are responsible for the species composition which is quite dissimilar to those of the other relevés. The remaining part of the relevé series is divided into two parts; Cluster A, with relevés 3-15; Cluster B, with relevés 17-20. One can find in both clusters small groups slightly different from the centre of the cluster (in Cluster A: relevés 4-5 and 14-15; in Cluster B: relevés 16-17). Because of their position they are most probably somewhat intermediate groups. RISE analysis gave a clear explanation. Two species played the most important role in composing the clusters, Festuca valesiaca ssp. pseudodalmatica and Brachypodium pinnatum. One can follow it easily in Fig. 5 that Cluster A could be characterized by the dominance of Festuca valesiaca. In the centre of the cluster its cover was is 50-80% of the total, while the cover of Brachypodium pinnatum was not higher than 30% here. In the two small subgroups of Cluster A the cover of Brachypodium increases (40%) and that of Festuca decreases (20-60%). In Cluster B Brachypodium pinnatum is dominant (70-90%), the cover of Festuca is never higher than 10%, moreover it is absent in the centre of this cluster.
