Environmental changes and human activities influence the spatial patterns of plant diversity in Eastern Eurasian Steppe

This study predicts the spatial distribution of plant diversity in the Hulun Buir Steppe using the Boruta algorithm and Random Forest model based on both single predictor variables and combined predictor variable sets. The results show that the combined predictor variables have the best predictive performance. The key variables affecting the spatial variation of plant diversity mainly include the topographic variables elevation and slope, the climatic variables mean annual temperature, annual precipitation, and precipitation during the growing season, as well as the soil variables surface soil temperature and moisture. The overall range of spatial prediction of species richness, Shannon-Wiener index, Simpson index, and Pielou’s evenness index are 5-37, 0.91-2.84, 0.522-2.90, and 0.30-0.88, respectively. The Hulun Buir Steppe exhibits a natural environmental gradient with decreasing precipitation and temperature from east to west. Consequently, the spatial distribution of species richness aligns with this trend, being higher in the eastern meadow steppes and lower in the central and western typical steppes. The Shannon-Wiener index shows higher values at the transition zone between the meadow steppes and typical steppes where hydrothermal conditions are moderate, but the Simpson index and Pielou's evenness index exhibit relatively small spatial variation across the entire study area. The above results can serve as foundational information to assist researchers in further exploring the grassland ecosystem's responses to climate change and human activities at the landscape scale, providing significant guidance for the ecological restoration and management of degraded grasslands.