Spatial pattern of plant species diversity and the influencing factors in a Gobi Desert within the Heihe River Basin,Northwest China

Understanding the spatial pattern of plant species diversity and the influencing factors has important implications for the conservation and management of ecosystem biodiversity. The transitional zone between biomes in desert ecosystems, however, has received little attention in that regard. In this study, we conducted a quantitative field survey(including 187 sampling plots) in a 40-km~2 study area to determine the spatial pattern of plant species diversity and analyze the influencing factors in a Gobi Desert within the Heihe River Basin, Northwest China. A total of 42 plant species belonging to 16 families and 39 genera were recorded. Shrub and semi-shrub species generally represented the major part of the plant communities(covering 90% of the land surface), while annual and perennial herbaceous species occupied a large proportion of the total recorded species(71%). Patrick richness index(R), Shannon-Wiener diversity index(H'), Simpson's dominance index(D), and Pielou's evenness index(J) were all moderately spatially variable, and the variability increased with increasing sampling area. The semivariograms for R and H' were best fitted with Gaussian models while the semivariograms for D and J were best fitted with exponential models. Nugget-to-still ratios indicated a moderate spatial autocorrelation for R, H', and D while a strong spatial autocorrelation was observed for J. The spatial patterns of R and H' were closely related to the geographic location within the study area, with lower values near the oasis and higher values near the mountains. However, there was an opposite trend for D. R, H', and D were significantly correlated with elevation, soil texture, bulk density, saturated hydraulic conductivity, and total porosity(P0.05). Generally speaking, locations at higher elevations tended to have higher species richness and diversity and the higher elevations were characterized by higher values in sand and gravel contents, bulk density, and saturated hydraulic conductivity and also by lower values in total porosity. Furthermore, spatial variability of plant species diversity was dependent on the sampling area.