Additive genetic contribution in multi-trait selection of soybeans under La Niña conditions

Abstract

Abstract The objective of this study was to quantify the additive genetic contributions in soybean genotypes evaluated under La Niña conditions. The experiment was conducted during the 2024/2025 season in southern Brazil using an augmented block design with 713 soybean lines from different generations and 38 cultivars as controls. Morphological, phenological, and yield-related traits were evaluated throughout the crop cycle, and thermal sum was estimated from meteorological data. Generalized additive models accounted for spatial dependence and revealed significant genotypic effects for developmental traits, while mixed models detected significant genetic effects for plant height, grain number, and grain weight. Environmental influence exceeded 90% of phenotypic variation, resulting in low heritabilities and reduced selection accuracy under water stress. Multi-trait selection proved effective in identifying genotypes closer to the desired ideotype. La Niña conditions constrained genetic expression, demonstrating that integrated mixed, additive, and multi-trait approaches are essential to achieve genetic gains under adverse environments.