L. Hably szerk.: Studia Botanica Hungarica 23. 1992 (Budapest, 1992)
Németh, Ferenc: Vegetation structure studies on steppe and semidesert plant communities of Outer Mongolia 1. Textural relations
In Mongolia the whole range of the two-dimensional plant communities of the temperate zone can be found from the deserts to the tall-grass steppes, from the salt marshes to the mountain tundra. I sampled zonal and subalpine chamaephyton-grass semideserts and mid-grass, forb-rich steppes (ELLENBERG and MUELLER-DOMBOIS 1967a). Special perspectives and difficulties With help of the family of information statistics models (JUHÁSZ-NAGY 1972, 1973, 1976, 1980, 1984) botanists have at their disposal a very many-sided and efficient tool to describe any aspect of the synphenetical characteristics of a plant community. However because of the universality of these models the researcher faces hard tasks to formulate the problem precisely; to select adequate methods and sampling parameters; and to explain results. I tried to compare different geographical variants of cold-continental semidesert and steppe formations supposed being in climax stage, and to exploit the ability of the information statistic models to manage the joint spatial behaviour of any species combination, hoping to find the functional units of the formation, the coalitions. In this first publication I describe the sampling sites and sampling parameters and characterize the textural relations (the frequency distributions) of the samples from some traditional viewpoints: species richness, family distribution and life form spectra (ELLENBERG and MUELLER-DOMBOIS 1967b). In contrast to the elegantly established theory the practical difficulties are more prosaic. First of all, sampling requires much manpower: a minimal programme, a linear sampling of 500-1000 adjacent microquadrates demands 1-2 days of two experienced botanists, and this often proves to be insufficient because of the unexpected, larger scale patchiness of the communities. These larger scale patches (10-100 m in diameter, mostly irregular, convex ones), especially in the steppe communities are to be expected from the small scale vegetation maps (LAVRENKO and SOKOLOV 1984, B ANNIKOVA et al 1986), but they cannot be detected simply, at a glance. It would be interesting to take lines of 10,000-100,000 units, with the help of remote sensing on growing scale to establish the scale hierarchy in the patchiness of the vegetation, and Mongolia is an optimal field for such researches with its extensive, homogeneous, plain semidesert and steppe communities. Another important problem is the scale dependence of the sampling. In general, the size of the sampling unit is adjusted to the scale of the average plant size. The main problems with it are as follows: - Even the dominant species of a community can belong to different size scales (e.g. Caragana and Säpa or Allium). The communities of this type were excluded from the research except one sample (Myangat), where I applied different sample sizes to the shrubs and to all species (not included in this paper), but it has been shown that this community differs both from the clear grass-herb and the bush formations structure, and one cannot consider it simply as their combination. 136
