A precise evaluation of spatial patterns in desert vegetation biomass, species richness and their environmental controls is essential for a deeper comprehension of the potential carbon preservation and sustainability of grassland ecosystems. There are widespread reports suggesting robust associations among biomass, species richness and mean annual precipitation (MAP) or temperature (MAT) at different scales. However, these reports were inconsistent, and knowledge on the desert grasslands of Central Asia remains limited. In this study, we showed that spatial patterns of biomass and species richness along the zonal climate of the northern Tianshan Mountains exhibited substantial regional differences and the relationship among biomass, richness and elevation exhibited a substantial exponential decline. We discovered that functional groups of biomass, total biomass and species richness in the desert exhibited exponential growth along the MAP gradient and a quadratic relationship with MAT. Furthermore, the biomass–species richness relationships were bell-shaped in the desert zone. Accordingly, the biomass and species richness had spatial differences. At a regional scale, the spatial variation in the desert biomass and species richness was primarily dependent on climate. Our results demonstrated the specificity between the desert vegetation and climate in arid regions of Central Asia and revealed the regularity between biomass and species richness in desert areas. The research results emphasized the impact of precipitation on desert vegetation in arid regions of Central Asia and the relationship between biomass and plant species richness, which is of great significance for understanding desert ecosystems and protecting the ecological environment.

Grazing exclusion is a widely implemented restoration strategy for degraded grasslands, but its effects on soil organic carbon (SOC) remain highly variable, and the ecological pathways associated with these changes—particularly the roles of soil microorganisms—are still insufficiently understood. To investigate context-dependent SOC responses, we conducted a 12-year grazing exclusion study across three contrasting grassland sites representing an environmental gradient (temperate desert, temperate steppe, and mountain meadow) on the northern slope of the Tianshan Mountains. By integrating multidimensional vegetation–soil–microbe observations with microbial network analysis and community assembly approaches, we examined how plant, soil, and microbial characteristics were associated with SOC variation under grazing exclusion. Our results showed that grazing exclusion was associated with significant increases in surface SOC content across all three investigated sites (16–81%), but these increases corresponded to distinct site-specific ecological association patterns. At the temperate desert site, SOC variation was mainly associated with soil conditions and microbial characteristics, where soil moisture and nutrient availability, together with fungal network complexity, showed significant associations with SOC content. At the temperate steppe site, vegetation recovery represented an important ecological component associated with SOC variation, while bacterial community assembly patterns were also related to SOC differences. At the mountain meadow site, plant, soil, and microbial characteristics showed multiple associations with SOC variation. Across all sites, grazing exclusion was consistently associated with changes in microbial network structure and community assembly processes, and these microbial ecological characteristics showed significant statistical relationships with SOC content. Our findings indicate that SOC responses to grazing exclusion are context-dependent and involve different ecological association pathways across contrasting grassland ecosystems, highlighting the importance of considering local environmental conditions when evaluating grassland restoration outcomes and carbon management strategies.

In modern livestock production, a reliable supply of high-quality forage is essential for sustaining animal productivity and product quality. Although nitrogen (N) fertilization can promote forage growth, excessive N inputs often result in low agronomic nitrogen use efficiency (NAUE) and increased environmental risks. Biochar, owing to its porous structure, high specific surface area, and physicochemical stability, can improve soil physical properties, enhance water and nutrient retention, and regulate soil N availability. However, the mechanisms by which biochar combined with reduced N rate fertilization affects NAUE in artificial grasslands remain insufficiently quantified. A two-year field experiment was conducted at the Grassland Science Experimental Station of Xinjiang Agricultural University on the northern slope of the Tianshan Mountains, Xinjiang, China. Eight treatments were established using a factorial design with two biochar rates (0 and 20 t·ha⁻¹; B0 and B20) and four N application rates (0, 75, 150, and 225 kg·ha⁻¹; N0, N75, N150, and N225). Results showed that biochar application significantly decreased soil bulk density and increased soil water content and electrical conductivity. It also elevated soil total carbon, total nitrogen, total phosphorus, NH₄⁺–N, and NO₃⁻–N concentrations, with B20N150 exhibiting the highest overall nutrient status. Plant community diversity indices did not differ significantly among treatments (p > 0.05), though B20 slightly enhanced Shannon–Wiener and Simpson indices under N0 and N75. Moderate N application significantly increased hay yield, whereas the highest N rate (225 kg·ha⁻¹) did not further improve yield and reduced NAUE. Biochar combined with N75 or N150 improved NAUE, and B20N150 achieved the best balance of high hay yield and high NAUE. Structural equation modeling revealed that soil water content (path coefficient = 0.45), NH₄⁺–N (0.27), and plant community diversity (0.20) were key positive drivers of NAUE, with biochar exerting indirect effects primarily via improving soil water and available N. Collectively, applying 20 t·ha⁻¹ biochar with 150 kg·ha⁻¹ N (B20N150) is recommended as an optimal strategy for N rate reduction and NAUE enhancement in artificial grasslands of arid and semiarid regions.