MODELING THE ONION YIELD RESPONSE TO DEFICIT IRRIGATION AND MULCH USING MULTIPLE REGRESSION MODEL IN SEMI-ARID NIGERIA

Abstract

Effect of deficit irrigation, mulch practices, crop growth and water use parameters on the yield of Onion was modeled using a multiple linear regression model. The crop evapotranspiration, number of leaves, leaf height and canopy cover of the Onion were used as the independent variables. Onion bulb yield was the dependent variable under four mulching materials (rice straw, RM; wood shaving, WM; white synthetic plastic, SM and no mulch, NM) in the semi-arid region, of Nigeria. The regression analysis revealed that the independent variables in the model predicted the Onion bulb yield significantly (p < 0.05) under mulching conditions, while no mulch plots yielded no significance as indicated by the ANOVA statistic. The overall model degree of determination (r²) of the dependent variable of 0.97, 0.97, 0.98, and 0.81 were obtained under SM, WM, RM, and NM respectively indicating that the multiple regression model predicted the dependent variable satisfactorily. The co-efficient values show that the highest coefficient was obtained at the number of leaves (0.56) followed by crop evapotranspiration (0.33). It was observed that among the four multiple regression models developed, the model obtained with white synthetic plastic mulch produce a better yield. Thus, white synthetic plastic mulch conserved soil moisture thereby improving Onion bulb yield. When the models were tested, a slight overestimation of the Onion bulb yield at both mulched and no mulched regression models was observed as compared with the yield from the field. Therefore, the model obtained with white synthetic plastic mulch produces a better yield.