Association of Yield and Yield Components Among Selected Bread Wheat Genotypes Under Silicon Fertilisation and Drought Conditions

ABSTRACT Wheat ( Triticum aestivum L. (2 n = 6 x = 42; AABBDD)) is a commodity crop serving diverse value chains worldwide. However, drought stress and poor soil health are major causes of a yield gap in wheat production. Silicon (Si) fertiliser application with drought‐adapted wheat cultivars may enhance wheat productivity. This study aimed to examine the impact of Si fertilisation on agronomic performance and trait associations in wheat under drought stress to identify selection criteria for improving drought tolerance and Si response. Twenty wheat genotypes were evaluated, involving two water regimes and two Si fertiliser formulations in three environments, using a factorial experiment laid out in a randomised complete block design with three replications. Both liquid and granular Si formulations enhanced the agronomic performance of wheat compared to the untreated control, under both drought‐stressed (DS) and nonstressed (NS) conditions. Variable degrees of correlations were recorded under different water regimes and fertiliser formulations. Si fertilisation improved favourable trait correlations compared to the untreated control under DS and NS conditions. Grain yield had positive associations with hundred seed weight (HSW), under NS conditions when using liquid Si ( r = 0.74, p < 0.001) and granular Si ( r = 0.69, p < 0.001). Strong trait correlations were detected between productive tiller number (TN) and productive spike number (PS) for granular Si ( r = 0.99, p < 0.001), liquid Si ( r = 1.00, p < 0.001) and the untreated control ( r = 0.98, p < 0.001), under DS conditions. With both granular and liquid Si applications under DS conditions, there were high trait correlations between spike length (SL) and TN ( r = 0.72, p < 0.001; r = 0.63, p < 0.001, respectively), and PS ( r = 0.73, p < 0.001; r = 0.62, p < 0.001, respectively). Under DS, HSW (0.84) and PS (0.64) had higher direct effects on grain yield with granular Si application, while TN and PS (0.94) had higher indirect effects on grain yield under liquid Si application. Under NS, aboveground biomass (0.68) had the highest direct effects on grain yield with granular Si compared to the liquid Si formulation and the untreated control. The genotypes MC10, MC6, MC1, MC3 and MC11, with granular Si application, were high yielders in descending order and are recommended for breeding and variety release in South Africa. The correlation and path coefficient analyses distinguished TN, PS and HSW as principal traits contributing to enhanced grain yield under drought stress, making them vital parameters for selecting Si use‐efficient breeding lines.