Detection of Superior Rice Genotypes and Yield Stability under Different Nitrogen Levels Using AMMI Model and Stability Statistics

Abstract

Sustainable agriculture is a prerequisite for food and environmental security. Chemical fertilization, especially nitrogenous fertilization, is considered the most consumed for field crops. In rice crops, plants consume much less than half of the applied N-fertilizer. In the current investigation, multiple N environments were generated by applying different N doses of urea fertilizer to a permanent transplanted field for two successive summer growing seasons at the rice research and training center, Kafrelsheikh, Egypt. A set of 55 genotypes consisting of 25 Jabonica, 4 Tropical Japonica, 20 Indica, and 6 Indica/Japonica were transplanted under no N (0N), Low N (LN), medium N (MN), and High N (HN) (i.e., 0, 48, 96, and 165 Kg N ha−1, respectively). Highly significant differences were detected among the tested genotypes. AMMI analysis of variance revealed the existence of the genotype via nitrogen interaction (GNI) on yield performance. The GNI principal components (IPCA); IPCA1 and IPCA2 scores were significant and contributed values of 71.1 and 21.7%, respectively. The highest-ranked genotypes were MTU1010, IR22, SK2046, SK2058, IR66, and Yabani LuLu based on their grain yield means (30.7, 29.9, 29.5, 29.3, 28.8, and 28.5 g plant−1). These genotypes were grouped into the same subcluster (SCL) according to the stability analysis ranking matrix. Based on AMMI analysis and biplots, MTU1010 and Yabani LuLu showed yield stability across environments. Meanwhile, the which-won-where biplot showed that IR22 was superior under unfavorable N-levels and MTU1010 was stable across the different environments. These findings are considered to be of great importance to breeders for initiating low-nitrogen-input breeding programs for sustainable agriculture.