Mean Annual Temperature, Soil Organic Matter and Phyllospheric Bacterial Diversity Shape Biomass of Dominant Species Along a Degradation Gradient in Alpine Steppes: A Case Study from the Qinghai-Tibet Plateau.

The structure and function of alpine steppes are maintained largely by dominant species, which in turn determine the productivity and stability of plant communities. Nutrient acquisition and stress regulation may, to some extent, be mediated by phyllospheric microbiota at the interface of plants with the atmosphere, and phyllospheric microbes are capable of amplifying and transmitting vegetation responses to degradation. Previous research has mainly addressed climate, soil, vegetation and soil microbiota or has assessed phyllosphere communities as a whole, thereby overlooking the specific responses of phyllospheric bacteria associated with the vegetation-dominant species Stipa purpurea along gradients of vegetation degradation in alpine steppes. In this study, we characterised vegetation degradation at the community level (from non-degraded to severely degraded grasslands) and quantified associated changes in the dominant species Stipa purpurea (cover, height and aboveground biomass) and its phyllospheric bacterial communities, in order to elucidate response patterns within the coupled system of host plants, phyllosphere microbiota, climate (mean annual temperature and precipitation) and soil physicochemical properties. Compared with non-degraded (ND) grasslands, degraded sites had a 22.6% lower mean annual temperature (MAT) and reductions in total nitrogen, nitrate nitrogen, organic matter (OM) and soil quality index (SQI) of 49.4%, 55.6%, 46.8% and 47.6%, respectively. Plant community cover and the aboveground biomass of dominant species declined significantly with increasing degradation. Along the vegetation-degradation gradient from non-degraded to severely degraded alpine steppes, microbial source-tracking analysis of the phyllosphere of the dominant species Stipa purpurea revealed a sharp decline in the contribution of phyllospheric bacterial sources. Estimated contributions from non-degraded sites to lightly, moderately and severely degraded sites were 95.68%, 62.21% and 6.89%, respectively, whereas contributions from lightly to moderately degraded and from moderately to severely degraded sites were 34.89% and 16.47%, respectively. Bacterial richness increased significantly, and β diversity diverged under severe degradation (PERMANOVA, F = 5.48, p < 0.01). From light to moderate degradation, biomass and relative cover of the dominant species decreased significantly, while the phyllosphere bacterial community appeared more strongly influenced by the host than by environmental deterioration; the community microbial turnover index (CMTB) and microbial resistance potential increased slightly but non-significantly (p > 0.05). Under severe degradation, worsening soil conditions and hydrothermal regimes exerted a stronger influence than the host, and CMTB and microbial resistance potential decreased by 6.5% and 34.1%, respectively (p < 0.05). Random-forest analysis indicated that climate, soil, phyllosphere diversity and microbial resistance jointly accounted for 42.1% of the variation in constructive-species biomass (R2 = 0.42, p < 0.01), with the remaining variation likely driven by unmeasured biotic and abiotic factors. Soil contributed the most (21.73%), followed by phyllosphere diversity (9.87%) and climate (8.62%), whereas microbial resistance had a minor effect (1.86%). Specifically, soil organic matter (OM) was positively correlated with biomass, whereas richness, beta diversity and MAT were negatively correlated (p < 0.05). Taken together, our results suggest that under ongoing warming on the Qinghai-Tibet Plateau, management of alpine steppes should prioritise grasslands in the early stages of degradation. In these systems, higher soil organic matter is associated with greater phyllospheric microbial resistance potential and increased biomass of Stipa purpurea, which may help stabilise this dominant species and slow further vegetation degradation.