Ascertaining yield and grain protein content stability in wheat genotypes having the Gpc-B1 gene using univariate, multivariate, and correlation analysis

Abstract

The high performance and stability of wheat genotypes for yield, grain protein content (GPC), and other desirable traits are critical for varietal development and food and nutritional security. Likewise, the genotype by environment (G × E) interaction (GEI) should be thoroughly investigated and favorably utilized whenever genotype selection decisions are made. The present study was planned with the following two major objectives: 1) determination of GEI for some advanced wheat genotypes across four locations (Ludhiana, Ballowal, Patiala, and Bathinda) of Punjab, India; and 2) selection of the best genotypes with high GPC and yield in various environments. Different univariate [Eberhart and Ruessll's models; Perkins and Jinks' models; Wrike's Ecovalence; and Francis and Kannenberg's models], multivariate (AMMI and GGE biplot), and correlation analyses were used to interpret the data from the multi-environmental trial (MET). Consequently, both the univariate and multivariate analyses provided almost similar results regarding the top-performing and stable genotypes. The analysis of variance revealed that variation due to environment, genotype, and GEI was highly significant at the 0.01 and 0.001 levels of significance for all studied traits. The days to flowering, plant height, spikelets per spike, grain per spike, days to maturity, and 1000-grain weight were specifically affected by the environment, whereas yield was mainly affected by the environment and GEI. Genotypes, on the other hand, had a greater impact on the GPC than environmental conditions. As a result, a multi-environmental investigation was necessary to identify the GEI for wheat genotype selection because the GEI was very significant for all of the evaluated traits. Yield, 1000-grain weight, spikelet per spike, and days to maturity were observed to have positive correlations, implying the feasibility of their simultaneous selection for yield enhancement. However, GPC was observed to have a negative correlation with yield. Patiala was found to be the most discriminating environment for both yield and GPC and also the most effective representative environment for GPC, whereas Ludhiana was found to be the most effective representative environment for yield. Eventually, two NILs (BWL7508, and BWL7511) were selected as the top across all environments for both yield and GPC.

Keywords: G × E interaction; grain protein content; multivariate analysis; stability analysis; univariate analysis; wheat.

FIGURE 1

Pattern of correlation and level of significance observed among different traits across all the environments in 13 bread wheat genotypes. DTF, Number of days to flowering; PH, Plant height; SPS, Number of spikelets per spike; GPS, Number of grain per spike; DTM, Number of days to maturity; TGW, 1000-grain weight; GPC, Grain protein content.

FIGURE 2

The “AMMI1” graphs displays the main effect and IPC1 effect values describing relationship among examined genotype and environment of 13 bread wheat genotypes across four locations (Ludhiana, Ballowal, Patiala, and Bathinda) of Punjab in two consecutive years (2019–20, and 2020–21) for (A) yield (kg/plot) and (B) GPC (%). (ENV1, and ENV5 = Ludhiana; ENV2, and ENV6 = Ballowal; ENV3, and ENV7 = Patiala; ENV4, and ENV8 = Bathinda).

FIGURE 3

The “AMMI2” graphs displays both the axes of interaction (IPCA1 and IPCA2) values for genotype effect and genotype by environment interaction effect of 13 bread wheat genotypes across four locations (Ludhiana, Ballowal, Patiala, and Bathinda) of Punjab in two consecutive years (2019–20 and 2020–21) for (A) yield (kg/plot) and (B) GPC (%). (ENV1, and ENV5 = Ludhiana; ENV2, and ENV6 = Ballowal; ENV3, and ENV7 = Patiala; ENV4, and ENV8 = Bathinda).

FIGURE 4

The polygon view of “Which-won-where” model of GGE biplot representing the performance of 13 bread wheat genotypes and their interactions with environment across four locations (Ludhiana, Ballowal, Patiala, and Bathinda) of Punjab in two consecutive years (2019–20, and 2020–21) based on (A) yield (kg/plot and (B) GPC (%). (ENV1, and ENV5 = Ludhiana; ENV2, and ENV6 = Ballowal; ENV3, and ENV7 = Patiala; ENV4, and ENV8 = Bathinda).

FIGURE 5

The “mean versus stability” model describing the interaction effect of 13 bread wheat genotypes evaluated across four locations (Ludhiana, Ballowal, Patiala, and Bathinda) of Punjab in two consecutive years (2019–20, and 2020–21) for (A) yield (kg/plot) and (B) GPC (%). (ENV1, and ENV5 = Ludhiana; ENV2, and ENV6 = Ballowal; ENV3, and ENV7 = Patiala; ENV4, and ENV8 = Bathinda).

FIGURE 6

The “ranking genotypes” model of biplot assess other genotypes against the ideal genotype conferring genotype interaction and GEI for 13 bread wheat genotypes evaluated across four locations (Ludhiana, Ballowal, Patiala, and Bathinda) of Punjab in two consecutive years (2019–20, and 2020–21) for (A) yield (kg/plot), and (B) GPC (%). (ENV1, and ENV5 = Ludhiana; ENV2, and ENV6 = Ballowal; ENV3, and ENV7 = Patiala; ENV4, and ENV8 = Bathinda).

FIGURE 7

The “Discrimitiveness vs. Representativeness” model of biplot evaluate the genotypes anianst the ideal genotypes conferring genotype interaction and GEI for 13 bread wheat genotypes across four locations (Ludhiana, Ballowal, Patiala, and Bathinda) of Punjab in two consecutive years (2019–20, and 2020–21) for (A) yield (kg/plot) and (B) GPC (%). (ENV1, and ENV5 = Ludhiana; ENV2, and ENV6 = Ballowal; ENV3, and ENV7 = Patiala; ENV4, and ENV8 = Bathinda).