Use of DArT molecular markers for QTL analysis of drought-stress responses in soybean. I. Phenotypic evaluation of traits

摘要

Physiological drought stress responses were assessed in recombinant inbred lines (RILs) from three soybean (Glycine max (L.) Merr.) crosses, in preparation for quantitative trait locus (QTL) analyses using Diversity Arrays Technology (DArT) markers. The three RIL populations were derived from pairwise crosses between three genotypes, cv. Valder, CPI 26671 and G2120, which in previous studies had differed in drought-stress response. Of particular interest was the landrace variety G2120, which in the previous reports had recovered better after severe drought. To assess drought stress response, the plants were grown in deep cylindrical pots in the glasshouse and exposed to severe water deficit followed by re-watering. Two plants to be genotyped were grown in each pot, together with one plant of G2120, which served as a reference plant against which the responses of the two other plants were assessed. Traits recorded included measures of relative water content (RWC), epidermal conductance (ge) and recovery in growth following re-watering. The responses in the reference and parental plants and the RIL populations were broadly consistent with previous studies. As plant-available water in the soil declined, both RWC and ge declined, although the relation between RWC and ge was exponential, rather than linear as in previous studies. Analysis of variance revealed large environmental effects on most of the traits, which resulted in high coefficients of variation and low estimates of broad-sense heritability. However, there were significant differences at both the population and genotype levels for all key traits, confirming the presence of genetic variation for drought-stress response. Some opportunities for enhancing the observed genetic differences and reducing the environmental noise in future studies are canvassed. Application of the observed phenotypic data reported in this paper in subsequent QTL analyses based on DArT markers is reported in the companion paper.