Determining behaviour of maize genotypes and growing environments using AMMI statistics

Abstract

Thirty three maize genotypes including hybrids and pure lines were evaluated to assess the grain yield performance under four environments; normal nitrogen (NN 120 kg ha^-1), high nitrogen (HN 160 kg ha^-1), low nitrogen (LN 80 kg ha^-1) and excess water (EW). Environment wise analysis revealed significant variance for grain yield. Analysis of data across the environments using additive main effect and multiplicative interaction (AMMI) statistics revealed significant variance for genotypes, environment and genotype x environment (GxE) interaction. Genotype main effects had the largest contribution (55.23%) to the total sum of square for grain yield followed by GxE interaction (22.73%) and environment (22.02%). The interaction component was further divided into three interaction principal component axes (IPCAs), IPCA I, IPCA II and IPCA III which accounted for 45.24%, 29.60% and 25.16% of GxE interaction component, respectively. The AMMI 1 biplot analysis using genotype main effect and IPCA I scores indicate that genotypes had high variance compared to the variance due to environments as evidenced by the distribution as well position occupied by the 33 hybrids and 4 environments on biplot display. The AMMI 2 biplot analysis identified three hybrids SCH-10, SCH-21, SCH-19 with relatively stable performance across the environments. AMMI 2 biplot analysis also categorized hybrids specifically adapted to NN, LN, HN and EW environments. However, all the hybrids exhibited specific adaptability to LN and EW environments having lower grain yield than the average yield. The inbred lines identified to have stable performance across the environments were IL-5 and IL-10, however it is interesting to note that none of the inbred lines exhibited specific adaptability to NN and HN environments which may due to inherently poor nitrogen use efficiency of inbred lines in comparison to hybrids.

SAARC J. Agri., 13(1): 162-173 (2015)