Mustard Yield Prediction based on Soil Site Characters in Udaipur, Rajasthan

The study was conducted to evaluate production potential of mustard crop in soils of Udaipur district of Rajasthan, representing 4.2 Agro-Ecological Sub-region. The yield was monitored at farmers’ fields consecutively for three years viz., 2004, 2005 and 2006. Multivariate regression model was developed to relate mustard crop yield to a set of independent variables such as soil depth, texture, slope, AWC, irrigation and agronomic management. These variables together governed 99% variability in mustard yield. The yield was predicted for the years 2005 and 2006. The predicted mustard yield ranged from 450 to 1550 kg ha₋₁ in shallow (<50 cm) sandy loam soils on gently sloping land having low AWC under low (37.5 kg N+15 kg P₂O₅+ no weeding + low/high plant population), medium (75 kg N+30 Kg P₂O₅+ one weeding + optimum plant population) and high agronomic management (150 kg N+60 kg P₂O₅+ 2 weedings + optimum plant population). One to three irrigations were provided in all the three conditions. Regression equation predicted 1120 to 1776 kg ha⁻¹ mustard yield in moderately shallow (50 to 75 cm) sandy loam to loamy soils on very gently sloping land, having medium AWC under the similar set of agro-management. In the similar set of management, mustard yield increased to 2900–3000 kg ha⁻¹ in very deep (>100 cm) clay loami clayey soils on nearly leveled plain with high AWC. The relationship between the observed and predicted yield was highly significant (r =0.99).