Soybean grain production and nutritional quality responses under elevated CO2, high temperature, and drought.

Soybean (Glycine max L. Merr.) is an important global crop that supplies protein and oil for the food and feed industries. However, its yield and nutritional quality are increasingly affected by climate change. This study combines experimental data with predictive modeling to assess how elevated CO2 (eCO2), high temperature, and drought (either alone or combined) influence soybean grain production and composition. Plants were grown under controlled conditions simulating future climate scenarios, and biochemical traits, including carbohydrates, proteins, lipids, and amino acids, were analyzed. Generalized linear models (GLMs) and machine learning algorithms (XGBoost, CatBoost) were used to predict yield and quality responses based on early-stage biomass data. Elevated CO2 increased grain production up to 142%, while high temperature and drought reduced yield by 91% and 60%, respectively. The combined "Triple Effect" (CO2 + high temperature + drought) was evaluated through predictive modeling from experimentally validated dual-stress datasets (eCO2 + temperature and eCO2 + drought), as this specific three-factor combination was not validated experimentally. Model projections a potential 50% increase in grain production, 35% soluble sugars, and a 175% rise in amino acid content, accompanied by reductions in starch (-20%) and protein (-6%). Elevated CO2 may partially offset stress, while inducing metabolic shifts that could increase productivity, but alter grain nutritional quality. The integrated experimental with modeling framework highlights the importance of early physiological indicators and predictive tools to anticipating yield-quality trade-offs and support development of soybean cultivars resilient to multifactorial climate stress under future environmental conditions.