Precision estimation of lime requirements in acidic soils using varied soil acidity indices under contrasting application methods in northwest Ethiopia

Abstract Soil acidity is a major constraint to crop productivity in intensively cultivated humid and sub-humid regions worldwide. Accurate estimation of lime requirement (LR) remains challenging due to spatial variability in soil buffering capacity and the use of different acidity indices. This study aimed to develop and compare lime recommendation models derived from multiple soil acidity indices under broadcast and drilling-along-the-row application methods. Soil samples were analyzed for pH–H2O, pH–CaCl2, pH–KCl, buffer pH, exchangeable acidity, and exchangeable aluminum. Pearson correlation, linear regression, and curvilinear modeling were employed to quantify relationships among soil acidity indices and LR. Strong positive correlations were observed among pH-based indices (r = 0.96–0.98), while exchangeable acidity showed strong negative relationships with pH indices and a near-perfect positive association with exchangeable aluminum (r = 0.98). Lime requirement exhibited a curvilinear response to all soil acidity indices, with higher LR under broadcast application than drilling-along-the-row application. The developed equations achieved very high predictive performance (R2 = 0.999) and revealed systematic differences among soil acidity indices and lime application methods. Compared with the broadcast method, drilling-along-the-row application reduced LR by 40–55%. The locally derived models tended to estimate lower LR, compared with generic Shoemaker McLean Pratt (SMP) buffer method-based recommendations, highlighting the importance of site-specific calibration. This study's findings demonstrate that integrating multiple soil acidity indices with lime application methods significantly improves the precision and agronomic relevance of lime recommendations, thereby contributing to sustainable soil acidity management and improved crop productivity.