Dryland wheat yields in relation to soil organic carbon, applied nitrogen, stored water and rainfall distribution

Correlations of long-term yields with soil, climatic and management variables would increase our understanding about their contribution to yield and help balance them for optimising the same. Therefore, grain yields of dryland wheat which received variable N ranging from 0–100 kg ha^–1 collected over 13 years, were analysed statistically. Wheat responded significantly to applied N in all but two years when the yields were the lowest. Grain yields across years did not relate with applied N alone (r² = 0.00), but inclusion of seasonal water supply and organic carbon content of soil (OC) in the regression accounted for 64 per cent variation in yield. When water supply was split into stored water and growing season rain the relationship improved further (R⁴ = 0.68) and the regression exhibited a significant interaction between seasonal rain and applied N.Three splits of seasonal water supply,viz. rain during 45 days after sowing (DAS) plus available water storage at seeding, rain during 46-120 DAS and rain during 121 DAS till a week before physiological maturity; soil OC and applied N explained 80 per cent variation in grain yield. Response to increase in post seeding water supplies was larger than that to the preseeding storage. Validation of the best regression against independent published data gave excellent agreement.