Screening of newly developed wheat genotypes for improved yield and yield components drought tolerance and water use efficiency

Abstract

Bread wheat (Triticum aestivum L., 2n = 6x = 42, AABBDD) is a vital commodity crop supporting global food security and economies. Wheat production is negatively affected by drought stress and water scarcity, exacerbated by climate change leading to extreme temperatures and unpredictable rainfall patterns. This calls for genetic management of drought through the screening and development of new wheat genotypes for improved drought tolerance and water use efficiency. To develop water-use efficient and drought-tolerant wheat genotypes, the wheat research group at the African Centre for Crop Improvement - University of KwaZulu Natal, in collaboration with the University of South Africa and the Institute of Research and Development (France) developed 90 new wheat breeding populations after crosses of eight wheat lines selected based on high biomass production and two local checks adapted for dryland wheat production in South Africa. The 90 wheat breeding populations were evaluated for biomass allocation at their F2 generations. These F2 families needed to be advanced to the F3 generation and evaluated for improved yield and yield-related components, drought tolerance and water use efficiency. Therefore, the overall objective of this study was to select newly developed wheat populations for improved grain yield, biomass production, drought tolerance and water use efficiency. The specific objectives of this study were: i to determine the accuracy of using carbon isotopes in estimating water use efficiency of selected cereal and legume crops based on a global perspective; ii to determine the agronomic performance and water use efficiency of newly developed wheat populations under different water regimes to guide selection; iii to determine the response of the selected wheat populations for grain yield and component traits and metabolites under drought stress to guide selection for direct production and breeding; iv to determine the degree of association between agronomic traits and water use efficiency, drought tolerance indices and metabolite profiles of newly developed wheat populations under different water regimes for simultaneous trait selection.

A comprehensive analysis involving 518 observations was performed to determine the accuracy of using carbon isotope in estimating water use efficiency in selected cereal and legume crops. The data for environmental factors (rainfall and temperature), soil properties, crop types and crop name were collected. The results showed that the mean observed water use efficiency (WUEobs) amongst all experiments was 3.4 g l-1 and the mean absolute error (MAE) was 0.5 g l-1 or 14.7% of WUEobs, corresponding to accurate predictions at p < 0.05. However, the percentage mean absolute error of observed water use efficiency (%MAE) estimated from grains was 3.6 ± 11.5%, which was lower than the %MAE from aboveground biomass collected at harvest (3 ± 22.8%). The use of carbon isotope discrimination (CID) to estimate water use efficiency (WUE) is more accurate for soybean (1.1 ± 5.1%), maize (1.6 ± 7.2%) and rice (1.2 ± 8.6%). The use of carbon isotopes to estimate water use efficiency (WUE) is limiting in oats because they appeared to have a high %MAE (6.3 ± 39.3%). The type of climate has an impact on the accuracy of using carbon isotopes to estimate WUE, with a high level of inaccuracy under Mediterranean climates. The CID-based WUE is underestimated in crops grown on loam soils. In the first experimental study, 100 wheat genotypes (10 parental lines and 90 F3 families) were evaluated under two field experiments using a 5 x 20 alpha-lattice design under drought-stressed (DS) and non-stressed (NS) conditions. The recorded agronomic traits include grain yield (GY), shoot biomass (SB), root biomass (RB), and total plant biomass (PB). The water use efficiency for grain yield (WUEgy), shoot biomass (WUEsb), root biomass (WUErb) and total plant biomass (WUEpb) were calculated. Drought tolerance indices, including the geometric mean productivity (GMP), mean productivity (MP), harmonic mean (HM), stress tolerance index (STI), yield index (YI), tolerance index (TOL) and stress susceptibility index (SSI) were computed based on grain yield under DS and NS conditions. Significant (p < 0.05) genetic variations were recorded for agronomic traits and WUE variables. The mean grain yield value of the F3 families was higher by 29.42% and WUE by 25.00% than the parental lines under DS conditions. Among the F3 wheat populations, the WUEgy ranged from 0.05 g mm-1 (LM47 X LM70) to 0.21 g mm-1 (BW141 X LM71) under DS conditions, whilst the WUEgy for the parental lines ranged from 0.08 (BW162) to 0.18 (LM48) under DS. Twenty-one percent of the wheat populations had the best GMP, MP, HM, STI, YI, TOL, and SSI compared to parental lines. Ten wheat families, BW141 X LM71, LM71 X BW162, BW140 X LM70, BW162 X BW140, BW141 X LM26, BW162 X LM71, BW152 X LM71, LM70 X BW141, LM75 X LM47 and LM70 X BW140 were selected for their high grain yield production and high WUEgy under DS conditions. These genotypes were recommended for further selection and deployment as new cultivars in South Africa. In the second experiment, ten high-yielding genotypes out of 100 wheat genotypes evaluated for water use efficiency in the study above were selected based on grain yield and were validated for agronomic traits and water use efficiency, and grain samples were assayed to profile their key metabolites under drought stress. Significant differences existed (p < 0.05) among the tested wheat genotypes for yield and yield components, WUE, drought tolerance and major metabolites to discern trait associations. The grain yield (GY) of the 10 genotypes ranged from 590.00 g m-2 (genotype LM70 X BW140) to 800.00 g m-2 (BW141 X LM71) under drought stress, whilst under non-stressed conditions it ranged from 760.06 g m -2 (LM70 X BW140) to 908.33 g m-2 (LM71 X BW162). Grain yield-based water use efficiency of the assessed genotypes was higher under non-stressed (0.18 g mm-1) than drought-stress (0.17 g mm-1) conditions. The highest drought tolerance index (211.67) and stress susceptibility index (0.77) were recorded for BW162 X LM71, whilst the lowest tolerance index (23.33) and stress susceptibility index (0.09) were recorded in BW141 X LM71. Grain metabolites, including the apigenin-8-C-glucoside (log2Fold = 3.00) and malate (log2Fold = 3.60) were present in higher proportions in the high-yielding genotypes (BW141 X LM71 and LM71 X BW162) under drought stress, whilst fructose (log2Fold = - 0.50), cellulose (log2Fold = - 3.90) showed marked decline in the two genotypes. Based on phenotypic and metabolite profile analyses, genotypes BW141 X LM71 and LM71 X BW162 were selected for being drought-tolerant, water-use efficient, and recommended for crop production or future breeding. The third experimental study evaluated the degree and trend of associations between agronomic traits and major metabolites to identify influential traits and metabolites optimized by wheat genotypes for improved grain yield production, water use efficiency and drought tolerance. The relationship between grain yield and drought indices was further evaluated to select high-yielding genotypes under DS conditions. Grain yield water use efficiency (WUEgy) showed significant (p < 0.05) correlations with PB (r = 0.68 for DS; r = 0.46 for NS), RB (r = 0.27 for DS; r = 0.24 for NS) and SB (r = 0.45 for DS; r = 0.24 for NS). The GY was positively associated with GMP (r = 0.88 for DS; r = 0.92 for NS), MP (r = 0.89 for DS; r = 0.92 for NS), STI (r = 0.92 for DS; r = 0.86 for NS), HM (r = 0.96 for DS; r = 0.83 for NS) and YI (r = 0.99 for DS; r = 0.63). Furthermore, citric acid was strongly correlated with GY (r = 0.30) and WUErb (r = 0.30) than all the selected metabolites under DS conditions. The SB had high positive direct effects on GY under DS and NS conditions, while PB had high and positive direct effects on WUEgy under DS conditions. The SB, SW and RB had the greatest positive indirect effects on WUEgy through PB. Therefore, the study demonstrated that selection based on SB, RB, PB, GMP, MP, STI, HM, YI and citric acid will be more effective when selecting wheat ideotypes for drought tolerance and water use efficiency. Overall, the present study identified promising families including BW141 X LM71 and LM71 X BW162, that have high grain yield, water use efficiency and improved drought tolerance. These families can be advanced using the single seed descent selection method for further characterisation of end-use quality traits and comparison with local checks or commercial cultivars.