Genotype-by-environment interaction effect on yield and its physiological bases in short-duration pigeonpea

Abstract

Short-duration pigeonpea [Cajanus cajan (L.) Millsp.] is being targeted for commercial cultivation in more diverse environments than traditional cultivars used in subsistence agriculture. As it is a relatively new crop, information on performance of recently evolved lines and hybrids across a range of environments is lacking. Thirty lines were compared for grain yield in 20 environments representing major areas of cultivation [Patancheru (17°N and 72°E), Gwalior (26°N and 78°E) and Hisar (29°N and 75°E)], soil types and sowing times. Grain yield across environments varied widely from 0.36 to 2.09 t ha−1. Average yield was highest, 1.53 t ha−1, for hybrid ICPH 8. The genotype-by-environment interaction (GEI) effect analyzed using the additive main effects and multiplicative interaction (AMMI) statistical model was highly significant and was three times more important than the line effects. Line ICPL 83006 was highest yielding in one, ICPL 87101 in two, hybrids ICPH 9 in four and ICPH 8 in 13 environments. Based on the range of adaptation of the highest yielding lines, the 20 environments were grouped into four fairly homogeneous crop growing environments in which the same genotypes performed best (mega-environments). The average yield with the highest-yielding lines planted in the respective mega-environment was 12.5% higher than when ICPH 8 was planted across all the environments. Significant correlations of crop growth rate, duration of reproductive phase, partitioning and growth habits of lines with GEI patterns suggest that variable performance of the lines across the 20 environments could be due to variation in the physiological components of yield. The results suggest the presence of significant GEI and that its systematic exploitation through selection of the appropriate physiological components for each mega-environment can result in higher average yield across growing environments.