Coupling Oxygen‐Enriched Drip Irrigation With an Optimized Nitrogen Rate Improves Yield, Water Productivity and Grain Quality of Winter Wheat: A Two‐Season Field Study

ABSTRACT Drip irrigation displaces soil air at the emitter and transiently depresses dissolved oxygen (DO) in the wetted root zone, and recovery is slow in fine‐textured, slightly saline soils. Whether aeration of the irrigation stream can be combined with nitrogen (N) management to raise water productivity while safeguarding grain quality in winter wheat remains unresolved. Over two growing seasons (2023–2024 and 2024–2025) on a slightly saline sandy loam in the southern Loess Plateau, we compared conventional water (CW; dissolved oxygen 6–8 mg L −1 ) with oxygen‐enriched water produced by a micro‐nano bubble generator (OW; dissolved oxygen 12–15 mg L −1 ) under drip fertigation at three N rates (90, 180 and 270 kg ha −1 ; N1–N3) in a randomized complete block design with three replicates, the same plots carrying the same treatments in both seasons. We measured the post‐irrigation DO supply window together with soil water content, mineral N, enzyme and microbial activity, root morphology and activity, leaf gas exchange and fluorescence, dry‐matter and N accumulation, grain yield, water productivity (WP) and grain quality, and analysed all variables in a linear mixed model with year as a repeated measure. Relative to CW, OW raised topsoil DO on day 1 after irrigation by 2.5–2.6 mg L −1 ( p < 0.001) and the elevated DO was still detectable at 7 days; mineral N was repartitioned from ‐N towards NO 3 − ‐N, with ‐N in the 0–20 cm layer at heading lower by 0.92–1.06 mg kg −1 (8%–10%) and ‐N higher by 5.8–6.9 mg kg −1 (21%–24%); and soil enzyme activity, microbial biomass, root length density, root activity and net photosynthesis all increased ( p < 0.001). Grain yield rose by 1305–1314 kg ha −1 (18.4%–19.3%) and water productivity by 2.23–2.37 kg ha −1 mm −1 . Evapotranspiration averaged 578 mm and did not differ by water type or nitrogen rate, except for two lower means in 2024–2025. The WP gain therefore tracked yield at essentially unchanged water use. Protein and wet gluten increased under OW and increased further with N rate, while grain yield and WP peaked at N2, defining a yield–protein trade‐off that oxygenation shifted upward but did not remove. TOPSIS ranked OW‐N2 first in both seasons and under all four weighting schemes and three reduced indicator sets tested, and partial least squares path modelling described a covariance structure consistent with a soil oxygen–nitrogen–root–photosynthesis–biomass sequence. Soil and root variables, but not yield, differed more strongly between water treatments in the second season; two seasons cannot separate a cumulative effect from inter‐annual variation, and we treat this observation as provisional. For water‐ and nitrogen‐saving production, OW combined with 180 kg N ha −1 is recommended, allowing fertilizer N to be reduced by about one third relative to the high rate without a yield penalty; the high rate is justified only where a protein premium outweighs the yield and WP cost.