Selection of Superior Tillage and Fertilizer Practices Based on Rainfall and Soil Moisture Effects on Pearl Millet Yield under Semiarid Inceptisols

Based on seven field experiments of pearl millet with nine treatment combinations of tillage and fertilizer nutrients conducted during 2000–2006 in a semiarid inceptisol at Agra, an assessment is made in this paper about sustainability of treatments using rainfall received during the crop growing period and available soil moisture at sowing, 20, 40, and $60\text{days}$ after sowing (DAS) and harvest. Three practices, each of tillage: conventional $\text{tillage}+\text{mechanical}$ weed control, low $\text{tillage}+\text{mechanical}$ weed control, and low $\text{tillage}+\text{mechanical}$ weed $\text{control}+\text{herbicide}$ ; and fertilizer application of $60\mathrm{kg}\mathrm{N}$ $\left[\text{farm}\text{yard}\text{manure}\left(\mathrm{FYM}\right)\right]+40\mathrm{kg}\mathrm{P}∕\mathrm{ha}$ , $30\mathrm{kg}\mathrm{N}$ $\left(\mathrm{FYM}\right)+30\mathrm{kg}\mathrm{N}$ $\left(\text{urea}\right)+40\mathrm{kg}\mathrm{P}∕\mathrm{ha}$ and $60\mathrm{kg}\mathrm{N}$ $\left(\text{urea}\right)+40\mathrm{kg}\mathrm{P}∕\mathrm{ha}$ were tested in the same site over seven years. The F-test indicated significant soil moisture differences on different DAS and also between different tillage treatments. Significant yield differences were found among treatments of tillage and fertilizer and their interaction in all seasons, except 2001 and 2002. Treatment-wise correlation of yield with monthly rainfall received in June–September and available soil moisture on different DAS indicated that September rainfall had a negative and significant correlation with yield attained by tillage and fertilizer treatments. The soil moisture at 20 DAS had a negative and significant correlation with yield under all treatments except conventional $\text{tillage}+\text{mechanical}$ weed control. The soil moisture at 60 DAS and harvest had a positive and significant correlation with yield attained under different tillage and fertilizer treatments. Regression models of yield were calibrated for tillage and fertilizer treatments through monthly rainfall during July–September and soil moisture on different DAS. The predictability of yield improved significantly by inclusion of both rainfall and soil moisture variables in the models compared to either of the two groups of variables. Ranks were assigned to tillage and fertilizer treatments for yield attained in individual years and mean yield, prediction error, and sustainable yield index over years. The study indicated that conventional $\text{tillage}+\text{mechanical}$ weed control among tillage together with $30\mathrm{kg}\mathrm{N}$ $\left(\mathrm{FYM}\right)+30\mathrm{kg}\mathrm{N}$ $\left(\text{urea}\right)+40\mathrm{kg}\mathrm{P}∕\mathrm{ha}$ was superior with a minimum rank sum compared to other treatments. The treatment gave a maximum sustainable yield of $\mathrm{1,683}\mathrm{kg}∕\mathrm{ha}$ with a net return of Rs $\mathrm{5,670}\mathrm{ha}$ , benefit–cost ratio of 1.16, and sustainable yield index of 47.2% under semiarid inceptisols of Agra.