Spatially Explicit NPK Fertilization Gaps Under Water‑Limited Conditions: An Integrated Modelling Approach for Sustainable Nutrient Management in Sub‑Saharan Africa

Soil fertility depletion, spatially variable nutrient limitations, and increasing climate variability are major constraints to crop productivity and resource conservation in Sub‑Saharan Africa (SSA). Sustainable nutrient management strategies that simultaneously improve crop yields, enhance soil health, and reduce nutrient losses are urgently needed to strengthen agricultural resilience. This study presents an integrated modelling framework for quantifying gaps in nitrogen (N), phosphorus (P), and potassium (K) fertilization under water‑limited conditions. The aim is to support site‑specific nutrient recommendations that improve fertilizer use efficiency while contributing to soil and water conservation.
The framework combines three complementary models. INITIATOR simulates soil nutrient stocks and flows, providing a spatially explicit assessment of nutrient availability and potential depletion risks. WOFOST estimates water‑limited maize yield potential, capturing the influence of rainfall variability and soil moisture dynamics on attainable production. QUEFTS then models nutrient uptake and yield responses, enabling calculation of nutrient requirements that match crop demand without excess application. Together, these components enable a holistic evaluation of nutrient management in soil–crop–water systems on different management levels.
The approach was applied to Uganda, a country characterized by strong agro‑ecological diversity, high interannual rainfall variability, and widespread soil degradation. Using 1 km resolution SoilGrids data and climate records from 2001–2024, spatially explicit water‑limited yield potentials were simulated for maize across the country. Fertilizer requirements were calculated based on median attainable yields for the period . Results reveal substantial geographic variation in nutrient availability, nutrient use efficiency, and fertilizer needs. In many areas, low soil nutrient stocks combined with high potential yield estimates lead to large fertilization gaps, while in other regions the gap is smaller due to relatively higher soil fertility or lower potential yields due to  moisture constraints.
These findings demonstrate that uniform fertilizer recommendations are unlikely to be effective under the diverse soil and climatic conditions of SSA. The proposed modelling framework provides a robust decision‑support tool for developing precision nutrient management strategies that align nutrient inputs with both crop demand and water availability. Such targeted approaches can help reduce nutrient losses to the environment, maintain soil fertility, and support sustainable intensification efforts across smallholder farming systems. This work highlights the potential of integrated soil–crop–climate modelling to guide context‑appropriate nutrient management solutions that enhance productivity while promoting soil and water conservation.