Validation of the QUEFTS Model for Predicting Fertilizer Requirements in Maize and Common Beans

Inherently low soil fertility continues to constrain crop productivity among smallholder farmers in Zambia. To remedy this, fertilizer application is often guided by blanket recommendations that fail to account for differences in soil nutrient supply and crop demand. The Quantitative Evaluation of the Fertility of Tropical Soils (QUEFTS) model can estimate balanced nitrogen, phosphorus, and potassium requirements, but its performance when calibrated using rapid soil testing kit data remains limited. This study validated QUEFTS-based recommendations for maize and common beans using soil data obtained from a rapid soil testing kit. On-farm trials were conducted at 12 farmer-managed sites and one researcher-managed site during the 2024/2025 season. Fertilizer treatments included QUEFTS-generated rates, a blanket rate (current practice), and control (no fertilizer) treatments. Significant treatment effects were observed for maize grain yields and biomass (p < 0.001) with the QUEFTS treatment producing the highest mean grain yield (5.6 t ha-1), followed by the blanket rate (4.8 t ha-1) and the control (3.2 t ha-1). For common beans, grain yield under QUEFTS was significantly higher than the control but not different from the blanket treatment, while aboveground biomass did not vary among treatments. Model evaluation indicated moderate predictive performance for maize (RMSE = 1.91 and d = 0.41) but overestimated the bean yield. These findings indicate that integrating rapid soil testing with the QUEFTS model can improve site-specific fertilizer management in maize although further refinement is required for common beans under variable field conditions.