Aridity-driven patterns and environmental controls of leaf C:N:P stoichiometry in arid ecosystems of Northwest China

Arid ecosystems occupy approximately 41% of the Earth's land surface and play crucial roles in global carbon (C) sequestration and biodiversity maintenance. However, extreme water limitation and nutrient scarcity in these regions pose significant challenges to plant growth and ecosystem functioning. This study characterized the spatial patterns of leaf C, nitrogen (N), and phosphorus (P) concentrations and their stoichiometric ratios (C:N, C:P, and N:P), identified key climatic and edaphic controls through correlation analyses and linear mixed-effects models, and assessed differences in mediating stoichiometric responses to aridity among plant life forms along a strong aridity gradient in Xinjiang Uygur Autonomous Region, Northwest China. Leaf C, N, and P were quantified for 428 samples representing 315 species across 74 sites spanning an aridity-index (AI) range of 0.03–0.47 during June–August 2024. With increasing aridity, leaf N and P declined significantly, whereas leaf C remained relatively stable, leading to higher C:N and N:P ratios. At the driest sites (AI<0.10), mean leaf N and P concentrations were only 16.61 and 2.02 mg/g, respectively—about 46%–70% (for N) and 60%–75% (for P) lower than both the overall means and the values at mesic sites (AI>0.40). Soil organic carbon (SOC) showed significant positive correlations with leaf N and P. Herbs had higher N and P and lower C:N, C:P, and N:P than shrubs and trees. Mixed-effects models showed that species identity explained approximately 40% of the variance in leaf C, whereas environmental variables explained 13%–16% of variation in leaf N, P, and N:P. Aridity-driven nutrient limitation is governed by coupled climate–soil mechanisms and modulated by plant functional traits; leaf C:N:P stoichiometry is a robust indicator for monitoring and restoration in drylands under climate change. Our findings reveal that intensifying aridity shifts plants toward P-limitation, with important implications for predicting dryland vegetation responses to future warming and for designing effective ecological restoration strategies in arid regions.