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

physiognomy: the plain semidesert of the lower basins (up to 1200-1300 m); the high mountain semidesert above 1700 m; and the steppes of the more humid and temperate zones. Their floral compositions do not confirm it: the high mountain semideserts of distant areas are considerably more similar to their neighbouring lowland variants than to each other. Why do they appear to be richer than they are? First of all, probably because of their more even species-, family- and life form- distribution at the scale of the sampling, then probably because of their more contrasting patchiness at larger scale which could not be detected by this study. Nevertheless, for didactical reasons I follow this division and present all three groups: the two "successfur ones and the one that failed. The locality of the two poorest samples (Bayan Tsagan, 9 species, Bayan Dalay, 14 species) belong to the floral district Gobi Altai: the first one in its northwestern part, in a smaller isolated higher elevation basin, the second one at the southeastern foot of the mountains, in an extensive lowland (which is only relatively low, being at an 1200 m average altitude a. s. 1.). Despite of differences and the 400 km distance between them, their floral distribution is very similar (Bayan Tsagan: Table 1, Figs la-c, Bayan Dalay: Table 2, Figs 2a-c, Plate 1). The frequency data are influenced by size of the sampling unit and the logarithmic data by number of the sampling unit, respectively, but in these two cases these parameters are the same, so a comparison is possible. The species poorness and prevailing dominance of the most common species, the cespitous grasses and artemisioid (wormwood- relative) dwarf shrubs (Stipa glareosa and Ajania achilleoides at Bayan Tsagan and Stipa glareosa - Cleistogenes songorica with Artemisia frigida at Bayan Dalay) is remarkable. It determines the life form spectra and family distribution as well. Without further investigation the background of the lack of moderately rare species in the Bayan Tsagan sample, the break in the frequency distributions is not clear. Such distributions are characteristic of extreme desert conditions or/and to the overuse or underuse of pastures (HARPER 1977). A special multifold problem is in all desert and semidesert communities is the occupation of space by slowly decomposable dead plants and plant colonies. They can make the environment more coarse-grained (MacARTHUR and LEVINS 1967); it is hard to estimate their underground extension; and sometimes it is hard to establish whether they are entirely and definitely dead. One may not exclude their possible allelopathic effect either. This phenomenon was extremely striking in the Bayan Dalay sample where the dead tussocks of Stipa, Cleistogenes, Iris and Allium amounted to about 10-20% of the living fraction. We registered them, but excluded from this analysis. The two other lowland semidesert samples (Khovd: Table 3, Figs 3a-c, Plate 2; Myangat: Table 4, Figs 4a-c) are situated in another floral district (Great Lakes' Basin), cca. 750 km northeast of Bayan Tsagan and 1200 km of Bayan Dalay; and 100 km from each other. There is no considerable difference in the amount and distribution of the precipitation (120-150 mm/year, maximum in June-August, according to the National Economy of the MPR - Statistical yearbook), but probably is there such a difference in the mean annual temperature and that of January, even with the reservations mentioned above. The data of Khovd