Nitrogen application effects on drip-fertigated cotton yield in Xinjiang: A meta-analysis of management, soil, and climatic moderators with agronomic and economic optimization

Chronic over-application of nitrogen (N) fertilizer in Xinjiang's drip-fertigated cotton systems reflects the absence of regionally calibrated management thresholds. To inform such thresholds for this production system, we conducted an integrated synthesis combining meta-analytic subgroup analysis, path and random forest analyses, and an economic optimization based on current market prices. We synthesized 1096 observations from 46 peer-reviewed field trials to quantify N effects on soil properties, crop growth, and seed cotton yield; identify management, edaphic, and climatic moderators; characterize dominant physiological pathways; and derive agronomically and economically optimal N rates. N application increased soil nitrate-N and ammonium-N by 72.22% and 40.66%, respectively, and raised seed cotton yield by 41.35% through coordinated improvements in leaf area index (LAI), dry matter accumulation, N uptake, and net photosynthetic rate. The greatest yield response (45.47%) occurred at 240–360 kg N ha⁻¹ , with returns declining markedly above 480 kg N ha⁻¹ . Allocating 20% of total N as basal fertilizer outperformed 30% basal allocation (61.15% vs. 33.17%), and responses were amplified at planting densities above 240,000 plants ha⁻¹ and under low initial soil organic carbon and available N. Climatic aridity intensified N responsiveness, with the strongest effects under annual precipitation below 60 mm and evaporation above 2000 mm. Random forest analysis identified N application regime and rate as the two dominant yield predictors (31.28% and 27.46% IncMSE). Path analysis revealed a dominant mechanistic chain—soil nitrate-N → LAI (direct path coefficient 0.684) → seed cotton yield. The agronomically optimal N rate (AONR = 345 kg N ha⁻¹) converged with the economically optimal rate (EONR = 318 kg N ha⁻¹), both stable across ±20% price variation. Agronomic N use efficiency declined linearly with rate (R² = 0.263, P < 0.001), from 9.05 to 2.70 kg kg⁻¹ . The convergence of agronomic, mechanistic, economic, and resource-efficiency evidence on 240–360 kg N ha⁻¹ provides a robust basis for N management in Xinjiang's drip-fertigated cotton.