Empirical investigation on non-linear genotype x environment interactions applied to vegetable crops

An investigation was carried out during 2001 to study the different procedures available for the analysis of non-linear genotype-environment interactions. The procedures, viz two-phase regression, refined two-phase regression and non-linear regression, were tried on live data of vegetable crops, and strength and weakness of the procedures were discussed based on the results obtained. Also a new procedure was proposed to overcome the defects in the 2-phase regression approach by interpolating the yield at the hypothetical environmental index value, viz zero using the best fitting polynomial and carry out the 2-phase regression analysis after including the interpolated yield in the 2 sets of graded environments. The new procedure was very much comparable with the Pooni and Jinks procedure.