Botanikai-Természetvédelmi Folyóirat

Journal of Pannonian Botany

Kitaibelia vol. 7 – no. 2. (2002) p.119-139.

Az invazív késeiperje, Cleistogenes serotina (L.) Keng. szerepe nyílt homokgyepek társulásszerveződésében
Szigetvári Csaba
Cikk letöltése: [pdf] (687 Kb)


The grass Cleistogenes serotina has been a recent fast-spreading invader in open sand grassland communities. The phytosociological relations and the community structure characteristics were investigated in different sand grassland stands invaded by Cleistogenes near Fülöpháza, Central Hungary. The research site was divided into three distinct territories ([1], [2], and [3]). Ten stands, labelled as FES1, FES3, FUM1, FUM3, SEC1, SEC3, STI2, CLE1, CLE2, ZUZ3 (the first three letter of each label refers to an priori classification, the last number refers to the territory where the stand was situated) representing different coenostates were sampled in the whole area. Stands in territory (1) and (2) are known to have been invaded by Cleistogenes since at least 1990, but there are no such data for territory (3). The percentage cover of species was estimated in three 4×4 m quadrats whereas species presence was recorded in 1030 5×5 cm adjacent micro-quadrats in a circular transect within each stand. The classical phytosociological relations were evaluated on the basis of the 4×4 m quadrats using hierarchical cluster analysis and nonmetric multidimensional scaling. Community structure characteristics—in a range spatial scales—were investigated by Juhász-Nagy's information theory methods and by pairwise species-species association analysis on the basis of the transect data.

The phytosociological evaluation showed, that the major sand grassland types (typicum [FES], fumanetosum procumbentis [FUM], and stipetosum borysthenicae [STI] subassociations of Festucetum vaginatae, and the secalietosum subassociation of Brometum tectorum [SEC]) could be distinguished in the invaded vegetation. Three of the stands, however, were not identifiable with classical coenological types. In two of them (CLE) Cleistogenes was the only dominant species, but the diversity of subordinate specialist species was the highest when compared to the other types. These two stands were similar to the FUM type, but Fumana procumbens was not predominant in them. The third atypical stand (ZUZ) was dominated by Cleistogenes and Poa bulbosa, with a dense cryptogam layer. The relation of this coenostate to the classical types is not clear.

The information theory method approach showed, that the diversity of species combinations (florula diversity, FD) and the measure of overall spatial dependence (associatum, Ass) of the stands was not related to the coenological type. However, FD and Ass was higher in all stands of territory (1) than in territory (3), which indicates that the invaded vegetation of territory (3) represents a less organized, more pioneer state. CLE1 and CLE2 stands had the highest maximum FD, and CLE1 also had the highest maximum Ass.

For the analysis of pairwise associations between species I defined the scale of the immediate neighbourhood, which was calculated for each species-pair in each stand from the size-distribution of the ramets. In the immediate neighbourhood Cleistogenes had similar association patterns as other perennial grasses. The invasive species was negatively associated to the codominant species in each site except in STI2, and usually to the dwarf-shrub Fumana procumbens, to other perennial grasses and to annuals when they were abundant. Cleistogenes had no positive associations to other species an had neutral relation to subordinate perennial dicots.

It can be concluded, that Cleistogenes serotina plays similar role in the open sand grassland as other dominants of the habitat. The invasive species probably competes with other dominants, especially with Festuca vaginata and Fumana procumbens, but does not necessarily exclude them. Coexistence of Cleistogenes with the subordinate specialists of high conservation value does not seem to be problematic. The species is able to be dominant in the more or less organized (probably the successionally more or less mature) communities of different coenological type.

The ratio of samples with stellate trichomes is low in the samples from the Great Hungarian Plain too. At the same time the variety of pilosity-types is the highest just in this region. It can be explained by the absence of Q. petraea in most of the stands, while Q. pubescens and Q. robur transitional forms appear in the populations.

The introgressive hybridization contributes to the remarkable variety of Q. pubescens.

In the Hungarian Mountain Range the rate of the transitional forms between Q. pubescens and Q. petraea is significantly higher on non-basic sites than on basic ones. We can establish that the extremely dry and acidic sites are favourable for hybridization, that is the introgression between the species can play an important role in their ecological adaptation.