From participatory systems mapping to spatial modelling: Co-produced soil erosion reduction pathways in western Uganda

Soil erosion poses severe threats to agricultural land in Sub-Saharan Africa, yet locally validated evidence to guide conservation planning remains scarce. Most catchment-scale assessments apply the Universal Soil Loss Equation (USLE) but often omit the support practice factor (P-factor), introducing a systematic bias into erosion estimates. This study introduces a participatory systems mapping approach (CoSMoS) to co-produce spatially explicit soil and water conservation practice (SWCP) inventories for the Mpanga catchment, Western Uganda, thus translating stakeholder knowledge into current and future spatially distributed P-factor layers, integrated into a USLE-based erosion modelling framework. Current conservation efforts, represented by a catchment-wide P-factor of 0.78, show a mean erosion rate of 7.8 t·ha⁻¹·y⁻¹ and reduce mean annual soil erosion by 28% relative to the baseline without SWCPs (10.8 t·ha⁻¹·y⁻¹), However, severe erosion hotspots persist, with over 25% of the catchment exceeding erosion of 11.4 t·ha⁻¹·y⁻¹. Future conservation pathways were developed by upscaling adoption rates of stakeholder-identified priority agricultural conservation practices by 50% and 100% under both general and spatially targeted allocation strategies. Targeted allocation scenarios consistently lower erosion compared with general upscaling at the same levels of SWCP adoption upscaling. The spatially targeted conservation with a 10 t·ha⁻¹·y⁻¹ threshold, achieved a mean catchment-wide erosion reduction of 35%, which was greater than the 22% achieved under 100% upscaling. Hotspot erosion is reduced by more than 60% under the most ambitious targeted scenario (5.1 t·ha⁻¹·y⁻¹). These findings provide Catchment Management Committees with spatially explicit, locally validated evidence directly addressing Uganda's Catchment Management Plan priorities and Land Degradation Neutrality commitments. The CoSMoS approach offers a replicable and cost-effective framework for spatial distributed P-factor estimation in data-scarce catchments across Sub-Saharan Africa.