Multivariate association patterns of morpho-physiological responses in soybean under short-term drought at the early vegetative stage using a drought resistance index framework

BACKGROUND: Drought stress severely limits soybean productivity, including during early vegetative growth, highlighting the importance of identifying cultivars with favorable early-stage drought responses. Because drought evaluation based on single traits may not adequately capture coordinated plant responses, integrative multivariate approaches are needed to characterize drought-response patterns among soybean cultivars. METHODS: Six soybean cultivars were evaluated under greenhouse conditions following a five-day water-withholding treatment at the V3 stage. Morphological, physiological, and biochemical traits were assessed and integrated using a standardized Drought Resistance Index (DRI) and composite D-value framework. Multivariate analyses, including hierarchical clustering, correlation analysis, principal component analysis (PCA), and partial least squares regression (PLS-R), were used to evaluate coordinated drought-response patterns among cultivars. RESULTS: Drought stress reduced most physiological and morphological traits, particularly stomatal conductance, photosynthetic performance, plant growth, and leaf water status. In contrast, proline accumulation increased under drought stress across cultivars, whereas increases in several root-related traits and transpiration responses were observed only in certain cultivars despite substantial reductions in stomatal conductance. Clear differences in drought-response profiles were observed among cultivars. Dering1 exhibited the highest D-value and, together with Dega1, was grouped within the drought-resistant response category, whereas Biosoy1 showed the lowest D-value and was classified as drought-sensitive. Correlation analysis indicated coordinated associations among plant water status, photosynthetic traits, and root-related characteristics under drought conditions. Principal component analysis revealed distinct multivariate response profiles among tolerant cultivars, with Dering1 more closely aligned with root-related and water-status-associated traits, whereas Dega1 was more strongly associated with photosynthetic and biomass-related indices. Partial least squares regression identified several traits, including net photosynthetic rate, stomatal density, instantaneous carboxylation efficiency, and root volume, as showing relatively stronger associations with D-value variation within the analytical framework. CONCLUSIONS: The integrative DRI-D-value framework, combined with multivariate analyses, provided an exploratory approach to evaluating coordinated drought-response patterns during early vegetative growth under short-term greenhouse drought conditions. The findings further suggest that early vegetative drought responses in soybean involve coordinated variation among water-status, photosynthetic, and root-related traits. Dering1 and Dega1 may therefore serve as promising candidates for further evaluation under broader developmental stages and environmental conditions.