Long-term effect of rainfall and fertilizer nutrients on sustainability of soybean yield, rainwater use efficiency, monetary returns and soil fertility under semi-arid Vertisols

A long-term experiment was conducted to assess the effect of rainfall and plant nutrients on soybean yield and Rainwater Use Efficiency (RWUE) by superimposingT\: Control;T₂: 20kgN(urea)+ 13 kgPha"¹;T₃: 30kgN(urea)+ 20 kg Pha"¹;T₄: 40kgN(urea) + 26kg Pha"¹; T₅: 60kgN (urea) + 35 kg Pha"¹; T₆: 20kg N(urea) + 13kgP + FYM@6tha"¹;T₇: 20kgN(urea) + 13kgP + FYM@5tlia"¹; T_g: FYM @ 61 ha"¹; and T,: Crop residue @ 51 ha"¹ to same plots during 1992–2009 in Vertisols at Indore. The yield variability was assessed using regression and Principal Component (PC) models through rainfall and plant nutrients. Based on regression analysis, the model of T_g gave maximum coefficient of determination (R²) of 0.78 compared to Tj with minimum of 0.38. Based on PC analysis, 3 PCs explained maximum variability of 74.1% in T₂ compared to minimum of 66.6% in T₆. The 1^st PC (PJ explained variance of 33.3 to 42.1%, compared to 18.1 to 21.6% by 2^nd PC (P₂) and 11.3 to 16.5% by 3^Id PC (P₃). The regression model of yield through PC scores indicated that T_ghad maximum R²of 0.46, while T₂had minimum of 0.15. Application of 20 kgN (urea) +13 kg P + FYM @ 6 t ha"¹ was superior with significantly higher mean yield of 2182 kg ha"¹, RWUE of 2.80 kg ha"¹ mm"¹, gross returns of ? 34814 ha"¹, net returns of? 24346 ha"¹, BC ratio of 3.33 and SYI of 43.0 and 44.2 based on regression and PC models respectively. The same treatment maintained maximum soil N of 269 kg ha"¹, soil P of 23.9 kg ha"¹, soil K of 757 kg ha"¹ and soil S of 18.0 kg ha"¹ at the end of study. This treatment could be adopted by farmers for attaining maximum soybean yield, RWUE and monetary returns and maintain soil fertility under semi-arid Vertisols.