Global climate change is altering the amounts of ice and snow in winter, and this could be a major driver of soil microbial processes. However, it is not known how bacterial and fungal communities will respond to changes in the snow cover. We conducted a snow manipulation experiment to study the effects of snow removal on the diversity and composition of soil bacterial and fungal communities. A snow manipulation experiment was carried out on the meadow steppe in Hulunbuir, Inner Mongolia, China, during the winter period October 2019-March 2020. Soil samples were collected from the topsoil (0-10 cm) in mid-March 2020 (spring snowmelt period). Snow removal significantly reduced soil moisture and soil ammonium concentration. Lower snow cover also significantly changed the fungal community structure and beta diversity. Snow removal did not affect the bacterial community, indicating that fungal communities are more sensitive to snow exclusion than bacterial communities. The relative importance analysis (using the Lindeman-Merenda-Gold method) showed that available nitrogen (AN), soil water content (SWC), total organic carbon (TOC), microbial biomass carbon (MBC), and microbial biomass nitrogen (MBN) together explained 94.59% of the variation in soil fungal beta diversity, where AN was identified as the most important predictor. These finding provide insights into potential impacts of climate warming and associated reduced snow cover on soil microbial communities and processes.

1. The polar desert biome of the Canadian high Arctic Archipelago is currently experiencing some of the greatest mean annual air temperature increases on the planet, threatening the stability of ecosystems residing above temperature-sensitive permafrost. 2. Ice wedges are the most widespread form of ground ice, occurring in up to 25% of the world's terrestrial near-surface, and their melting (thermokarst) may catalyse a suite of biotic and ecological changes, facilitating major ecosystem shifts. 3. These unknown ecosystem shifts raise serious questions as to how permafrost stability, vegetation diversity and edaphic conditions will change with a warming high Arctic. Ecosystem and thermokarst processes tend to be examined independently, limiting our understanding of a coupled system whereby the effect of climate change on one will affect the outcome of the other. 4. Using in-depth, comprehensive field observations and a space-for-time approach, we investigate the highly structured landscape that has emerged due to the thermokarst-induced partitioning of microhabitats. We examine differences in vegetation diversity, community composition and soil conditions on the Fosheim Peninsula, Ellesmere Island, Nunavut. We hypothesize that (i) greater ice wedge subsidence results in increased vegetation cover due to elevated soil moisture, thereby decreasing the seasonal depth of thaw and restricting groundwater outflow; (ii) thermokarst processes result in altered vegetation richness, turnover and dispersion, with greater microhabitat diversity at the landscape scale; and (iii) shifts in hydrology and plant community structure alter soil chemistry. 5. We found that the disturbance caused by melting ice wedges catalysed a suite of environmental and biotic effects: topographical changes, a new hydrological balance, significant species richness and turnover changes, and distinct soil chemistries. Thermokarst areas favour a subset of species unique from the polar desert and are characterized by greater species turnover (beta-diversity) across the landscape. 6. Synthesis. Our findings suggest that projected increases of thermokarst in the polar desert will lead to the increased partitioning of microhabitats, creating a more heterogeneous high arctic landscape through diverging vegetation communities and edaphic conditions, resulting in a wetland-like biome in the high Arctic that could replace much of the ice-rich polar desert.