A statistical assessment of genotypic sensitivity of groundnut (Arachis hypogae L.) to drought in line source sprinkler experiments

Comparison of genotypes in line source based drought screening has a number of statistical problems because of the systematic nature of irrigation treatments. A method of applying the joint regression approach has been developed and applied to evaluate sensitivity of 22 groundnut genotypes grown under 11 patterns of drought which were simulated in the field using a line-source sprinkler technique. The experimental errors in neighbouring environments are assumed correlated to account for the systematic nature of the environments. The estimation of parameters of the model and comparison of genotypes for their sensitivity to drought are presented for the pod yield. Stability in performance across the line-source environment was estimated for 22 genotypes. None of the genotypes tested was insensitive to drought across all patterns. Genotypes with stability and high mean yield could be identified in early and mid-season drought patterns but not in other patterns where genotypic sensitivity was strongly correlated with yield performance