<![CDATA[Agricultural systems are increasingly challenged by climate variability, which disrupts crop productivity and threatens long-term sustainability. Existing approaches often separate physiological understanding from predictive modeling, limiting their ability to capture the complexity of crop responses to environmental stress. This study aims to develop an integrative framework that couples crop physiological processes with predictive statistical models to improve the accuracy and interpretability of agricultural sustainability assessments. A mixed-methods design was employed, combining field-based physiological measurements with advanced statistical and machine learning modeling. Data were collected across multiple agricultural sites, including climatic variables, soil conditions, and key physiological indicators such as photosynthetic rate, stomatal conductance, and water-use efficiency. Predictive models were developed and evaluated using regression analysis and machine learning techniques with cross-validation procedures. Results indicate that models incorporating physiological variables significantly outperform those based solely on climatic data in predicting crop yield. Physiological indicators function as critical mediators between environmental stress and productivity, enhancing both predictive accuracy and explanatory depth. Nonlinear modeling approaches further improve performance by capturing complex interactions among variables. Findings demonstrate that integrating crop physiology with predictive modeling provides a robust framework for understanding and managing agricultural systems under climate variability. This approach supports more adaptive and sustainable agricultural strategies.]]>
Copyrights © 2026
