Spatial pattern of a plant community in a wetland ecosystem in a semi-arid region in northwestern China

摘要

The spatial distribution characteristics of plant communities in a wetland-dry grassland ecosystem in arid and semi-arid regions in northwestern China and the main factors affecting the distribution pattern were studied. The Siertan wetland in the Haba Lake Nature Reserve was the study area. Four transect lines, each about 1 km long, were set up in the growing season along the biotope gradient in four directions: east, northeast, west and northwest. Attributes measured include frequency of occurrence, height, density, coverage, biomass and envir-onmental soil factors. The data were analyzed using a two-way indicative species analysis (TWINSPAN), detrended correspondence analysis (DCA) and Spear-man correlation coefficients. On this basis, the vegetation in the Siertan wetland was classified into three types and 14 associations. The first vegetation type was halophytic marsh vegetation, mainly distributed in the wetland zone. This type largely contains halophyte and hygric plants and its representative association is Phragmites communis + Kalidiumfoliatum. The second type was meadow vegeta-tion, distributed in the ecotone. The meadow vegetation species are mainly mesophytic and their representative association is Nitraria tangutorum+ Pennisetum flacci-dum. The third vegetation type is grassland vegetation, which mainly consists of mesophytic and xerophytes, dis-tributed in the arid grassland zone. The representative association of the third vegetation type is the association of Anaeurolepidium secalium + Saussurea runcinata. Correlation analysis between the axes of DCA and envir-onmental soil factors shows that soil moisture content, organic matter, soil salt content and total nitrogen are the main environmental factors affecting the pattern of vegetation distribution along the biotope gradients from wetland to arid grassland. Other factors affecting the vegetation are microtopography and grazing pressure. Additionally, we have reported modifications and improvements to the importance value methodology.