PREDICTION OF VEGETATION PATTERNS AT THE LIMITS OF PLANT LIFE - A NEWVIEW OF THE ALPINE-NIVAL ECOTONE

The distribution pattern of individual plant species, as well as of pl ant communities, at the transition between the alpine and the nival en vironment (= alpine-nival ecotone) is likely to he drastically affecte d by climate change. Currently, the best way to explore the vegetation structure and to detect possible changes is the application of spatia l modeling to predict vegetation patterns over larger areas combined w ith dynamic modeling techniques. Schrankogel in Tyrol, Austria, was se lected as a typical high alpine mountain to establish and test such mo dels. As a first step, the predictive model for the spatial pattern of species and plant communities is presented here. Direct and indirect gradient analyses (CA, CCA) were combined with GIS-techniques based on a fine-grained Digital Elevation Model (DEM; pixel size: 1 m(2)). App roximately 1000 field samples (vascular plant species and cover within 1 m(2) squares) distributed over the alpine-nival ecotone of the moun tain were taken as the vegetation data input. Topographic descriptors were derived from the DEM as habitat characters of those samples. Usin g the correlations between vegetation samples and habitat characters, single plant species as well as community distribution could be predic ted for the pixels of the whole model area (the studied ecotone area c ontained a total of 650,000 pixels) for which habitat characters were known from the DEM. Distinct distribution patterns at different spatia l resolutions appeared for individual species, species groups, and com munities in relation to the relief. Descriptors of relief curvature an d roughness explained more of the variability than ''classical'' terra in attributes, such as elevation or exposure. Nevertheless, the altitu dinal gradient was clearly reflected by the CCA ordination. As species richness of vascular plants was recorded in each sample plot, biodive rsity distribution patterns could be modeled. These patterns showed th e general trend of decline of biodiversity with altitude, but with a m aximum of species richness at the ecotone itself. Since the relief mod ifies the high mountain climate remarkably, this differentiated relief dependency of vegetation supports the view that this type of environm ents will be affected significantly by climate change.