Multi-Stage Growth Parameter Correlation, Yield Prediction Models, and Sustainability Analysis for Soybean and Pigeon Pea

Enhancing crop productivity and sustainability is a critical priority in modern agriculture, particularly for leguminous crops such as soybean (Glycine max L.) and pigeon pea (Cajanus cajan L.), which play a vital role in global food security and soil fertility improvement. Accurate yield prediction models integrating multi-stage growth parameters can enable informed decision-making for crop management and resource optimisation. However, limited studies have simultaneously examined stage-wise parameter correlations, regression-based yield estimation, and sustainability metrics for these crops under varied treatments. This study aims to develop and evaluate multi-stage correlation matrices, regression models, and sustainability indices for soybean and pigeon pea yields using a comprehensive set of morphological and physiological parameters observed across key developmental stages. The parameters included seed yield (SY), dry matter (DM), plant height (PH), number of functional leaves (NFL), number of branches (NOB), leaf area (LA), number of pods (NOP), leaf area duration (LAD), pod weight (PW), number of seeds (NOS), seed weight (SW), test weight (TW), and harvest index (HI), with model performance evaluated using the coefficient of determination (R²) and prediction error (PE). Results revealed strong positive correlations between LA, NOP, PW, and SY in soybean, and between DM, NOP, and SY in pigeon pea at later growth stages, with regression models achieving R² values exceeding 0.90 for the best-performing treatments. Sustainability yield index (SYI) analysis indicated that specific treatments consistently maintained high yield stability across seasons. The findings conclude that integrating multi-stage growth parameters into regression models significantly improves yield prediction accuracy and aids in identifying treatments that enhance sustainability. The developed models can serve as decision-support tools for optimising crop management practices, and future research may extend this approach to other legume species and integrate remote sensing data for real-time yield forecasting.