Integrated soil profile mapping reveals regional patterns of soil condition across Southern African agricultural lands

Résumé fourni par la source

Soil information at regional scales remains limited in Africa despite its importance for sustainable agricultural management, food security and land degradation assessment. Existing digital soil mapping approaches commonly model soil properties as independent depth layers and apply depth harmonisation after prediction, limiting their capacity to characterise changes in soil condition. Here, we present Integrated Soil Profile Mapping, a digital soil mapping framework that incorporates vertical profile information directly into prediction through an energy-minimisation approach. Integrated Soil Profile Mapping represents soil profiles as continuous three-dimensional systems and analytically derives harmonised depth-specific soil estimates. The framework was applied to seven soil properties including clay, pH, cation exchange capacity, soil organic carbon, total nitrogen, phosphorus and potassium using the Soils4Africa dataset across agricultural lands in Eswatini, Lesotho, Namibia and South Africa. Comparison with iSDAsoil revealed contrasting regional patterns in soil condition. Soil organic carbon increased by 86–130% (effect size = 1.10–1.40 SD), while subsoil total nitrogen increased by 46% (effect size = 1.41 SD). In contrast, topsoil pH declined by 7.4% (effect size = −1.96 SD), indicating substantial changes in soil biogeochemical condition. These patterns are consistent with differences in agricultural management and regional policy contexts, although attribution requires further investigation. The findings demonstrate the potential of vertically integrated soil mapping to characterise regional soil condition and provide a basis for monitoring environmental change, assessing soil resilience and supporting depth-driven sustainable land management across data-limited regions.