Principal coordinates analysis of genotype × environment interaction in grain yield of lentil genotypes

Exploration the genotype × environment (GE) interaction pattern is an important issue in most plant improvement programs. A significant GE interaction for grain yield can limit attempts in selecting the most favorable genotypes for both new cultivar recommendation and improved genotype development. Conventional statistical models of stability analysis provide little or no insight into the pattern of the GE interaction while the principal coordinates analysis can account more effectively for the underlying GE interaction patterns. To characterize 18 lentil genotypes from performance trials, data were obtained from 12 environments (four locations in 3 years). Combined analysis of variance indicated that effects of genotype (G), environment (E) and GE interaction were highly significant. The genotypes accounted for 6.8% of the sum of squares of G+E+GE, with environment responsible for 70.4% and interaction for 22.7%. According to grand means of test environments and total mean yield (1230 kg ha-1), environments are classified to two main groups as H (seven high mean yield) and L (five low mean yield). The identified most stable and high mean yield genotypes based on the minimum spanning tree plots and centroid distances were G2 with 1366 kg ha-1 and G11 with 1374 kg ha-1, and therefore could be recommended for unfavorable or poor conditions. Also, genotypes G5 (1324 kg ha-1), G9 (1329 kg ha-1) and G14 (1402 kg ha-1) were located four, five and six times in the vertex positions of seven high cycles, respectively according to the principal coordinates analysis and so these genotypes were the most stable ones with high mean yielding properties. The principal coordinates analysis provided useful and interesting ways of investigating GE interaction of lentil genotypes.