Relationships between mixsmart parameters and bread wheat quality characteristics in South African dry land cultivars

Abstract

Ten hard red wheat genotypes from the National Cultivar Trials conducted in South Africa, were planted at three environments over two seasons in the eastern part of the summer rainfall area of South Africa. Wheat quality analysis, including hectolitre mass, falling number, grain and flour protein content, flour yield, corrected flour colour, farinograph analysis, alveograph analysis, mixograph analysis, loaf volume and corrected loaf volume were performed on these samples. The mixograph analysis was subjected to Mixsmart analysis and six midline Mixsmart parameters were selected based on their repeatability, coefficient of variation, coefficient of determination and their genotype contribution to variation. The selected parameters were midline left slope (MLS), midline peak time (MPT), midline peak value (MPV), midline right slope (MRS), midline right integral (MRI) and midline tail width (MTW). These parameters showed highly significant correlations with the fixed, non-negotiable primary quality criteria used during the cultivar classification process in South Africa. Therefore, they could be effective in assisting breeders during early selections for potential breeding lines that would exhibit acceptable primary criteria during the final evaluation phases for cultivar classification. Stability for the parameters were investigated using AMMI and GGE-biplots. Genotypes varied in their stability. For only two of the Mixsmart parameters, MPT and MTW, the GGE-biplot indicated the most stable genotype also as the ideal genotype. Over years scientists differed in total mixing time preferred for performing mixograph analysis (6, 8 or 10 min). This study investigated if differences in the mixing time would alter the relationships between the Mixsmart parameters and primary criteria. TimeX values were therefore set for 6, 8 and 10 min. ANOVA indicated envelope timeX slope (ETXS), envelope timeX width (ETXW), envelope timeX integral (ETXI), envelope timeX value (ETXV), midline timeX slope (MTXS), midline timeX width (MTXW), midline timeX integral (MTXI) and midline timeX value (MTXV) to show variation (causing differences at the different time intervals), but when these parameters were eliminated from linear discriminant analyses (LDA), there were no variations in mixing time interval values for the three timeX times. Correlations did not differ significantly between the primary quality criteria and Mixsmart parameters at the different time intervals, indicating that relationships between Mixsmart parameters and primary quality criteria remained constant at the three timeX times. Wheat quality analysis required for classification purposes of potential breeding lines, are conducted by the South African Grain Laboratory (SAGL) and breeders have to select the best representative samples, over a period of three years, of each breeding line over five environments for submission to SAGL. Breeders utilise white flour or whole meal when conducting pre-selection and it is of utmost importance that relationships with the primary criteria do not differ when different flour types are used. LDA confirmed that no differences existed between the white flour and whole meal used for mixograph analysis as indicators of acceptable quality.