Selection of Wheat Genotypes for Drought Adaptation and Yield Response Using Tolerance Indices and Biplot Analysis

ABSTRACT Drought‐adapted new‐generation wheat genotypes enhance production and productivity in water‐limited agroecologies, including South Africa. Therefore, drought‐tolerant wheat ideotypes need to be bred and deployed using economic traits and tolerance indices. The aim of this study was to evaluate genetically diverse wheat genotypes and select drought‐adapted lines based on yield response, tolerance indices and genetic groups using biplot analyses for breeding and production. Ninety‐eight wheat genotypes were assessed in five environments, that is, two seasons and three sites under drought‐stressed (DS) and non‐stressed (NS) growing conditions using an alpha lattice design with two replications. Grain yield (GY) was recorded, and yield‐based 10 drought‐tolerance indices were computed for genotype selection and comparison of test environments using the genotype plus genotype by environment interaction (GGE) biplot model. The mean GY values of the test genotypes in descending order across the five environments (E) were 1.59 t ha −1 (Bethlehem site in 2022, designated as BHM‐E2), 1.57 t ha −1 (Kransfontein in 2021, KRANS‐E4), 1.03 t ha −1 (Ficksburg in 2021, FICKS‐E3), 0.63 t ha −1 (Bethlehemin 2021, BHM‐E1) and 0.58 t ha −1 (Kransfontein in 2022, KRANS‐E5). The following genotypes were the best yielders: LM29, LM9 and at BHM‐E1, BHM‐E2, FICKS‐E3 and KRANS‐E4; LM59, LM63, LM66 and LM67 at BHM‐E2, FICKS‐E3 and KRANS‐E4; and LM83 at BHM‐E1, BHM‐E2 and FICKS‐E3. The genotypes LM59, LM25, LM84, LM96, LM23 and LM39 exhibited low tolerance (TOL), susceptibility index (SSI) and high mean productivity (MP), geometric mean productivity (GMP), drought index (DI), yield index (YI), yield stability index (YSI) and relative drought index (RDI) values in a desirable trend. Correlation analysis revealed a strong association ( p < 0.001) between mean GY in non‐stressed conditions ( Y p ) and drought tolerance indices such as TOL ( r = 0.87), MP ( r = 0.77), SSI ( r = 0.72), SDI ( r = 0.72), GMP ( r = 0.65) and STI ( r = 0.54). In contrast, the mean GY in stressed conditions ( Y s ) positively and significantly ( p < 0.001) correlated with YI ( r = 1.00), DI ( r = 0.97), RDI ( r = 0.73), YSI ( r = 0.73), GMP ( r = 0.51) and MP ( r = 0.41). The indices were invaluable in identifying relatively high‐yielding and drought‐tolerant wheat genotypes, and their combined use could be effective for screening drought tolerance in wheat breeding programmes. Based on GGE biplot analysis, genotypes LM6, LM9, LM30, LM64, LM83 and LM95 were discerned to be stable and high‐yielding in the test environments. Developing new breeding populations is recommended using the above complementary selections through combining ability tests and progeny selection for yield and agronomic traits performance for variety registration and release.