Rice yield and its determinants in Eastern Uttar Pradesh of India

Abstract

A diagnostic survey was conducted in nine districts of Eastern Uttar Pradesh (India) under Cereal Systems Initiative for South Asia (CSISA) through Krishi Vigyan Kendra (KVK) of each district. The survey data consisted of observations from farmer’s field on different variables encompassing crop yield, inputs and management practices. Information in the survey was collected from a total of 1218 farmers following two stage sampling designs from all the districts. A total of nine varieties were observed to be cultivated in the surveyed districts. The four dominant rice varieties/hybrids preferred by majority of households (HHs) were MTU7029, BPT5204, Moti and Arize6444Gold. On an average, the highest yield was observed in Maharajgang district and the lowest was in Mau and Balia districts. The highest average yield (across the districts) was observed for MTU7029 and lowest was for Moti, and there were significant variety X district interactions. For yield determinants, multiple linear regression (MLR) model with stepwise selection technique (SST) was applied, with twelve regressors covering both quantitative as well as qualitative variables. Across districts and varieties simultaneously, regressors could not explain much of the variability in yield. However, on carrying out the analysis by variety across districts (i.e. fixing genotypic as well as the climatic/environmental effects and looking for the contribution of management variables for determining yield), the value of R2 ranged from 20 to 85% for different variety-district combinations. This analysis revealed that transplanting date, number of irrigations, farm size, drainage class and amount of N and K applied were the most significant yield determinants.