Geospatial analysis of soil quality index in relation to land use/land cover and spectral indices in Wukari, Nigeria

Soil quality is a key determinant of ecosystem sustainability in tropical agroecosystems, yet its spatial response to land use and land cover (LULC) change remains poorly quantified. This study addresses this gap by combining satellite-based LULC mapping with field-derived soil quality assessment in a representative tropical landscape. LULC mapping was conducted using Sentinel-2 imagery and a Random Forest classifier, which delineated six LULC classes with high accuracy (overall accuracy = 0.99; κ = 0.99). Soil sampling across the mapped LULCs produced 87 geo-referenced topsoil samples that were analyzed for 14 physical, chemical, and biological properties. Principal component analysis and correlation screening reduced these variables to a minimum data set comprising microbial biomass carbon, clay content, pH, and bulk density, explaining 87.9% of the total variance. These indicators were integrated into a weighted Soil Quality Index (SQI), which varied significantly among LULC types (p < 0.05), which was highest under dense vegetation (0.69 ± 0.08) and lowest under built-up areas (0.30 ± 0.06). Ordinary kriging indicated very strong spatial dependence of SQI (nugget-to-sill ratio = 0.1%), corroborating regression results that showed a positive influence of Normalized Difference Vegetation Index (NDVI) on SQI (β = 0.66, p < 0.001) and a negative effect of Normalized Difference Built-up Index (NDBI) (β = −0.53, p = 0.027). A geographically weighted regression improved explanatory power (adjusted R 2 = 0.73), highlighting spatially variable soil–landscape interactions. Overall, the results demonstrate that vegetation dominated LULCs strongly enhance soil quality in the Wukari landscape and that spectral indices provide effective proxies for monitoring soil degradation and recovery. • Soil quality was assessed using SQI, spectral indices, and geostatistics. • Vegetation cover was identified as the main driver of soil quality patterns. • PCA identified microbial biomass as a dominant soil quality indicator. • NDVI and NDBI served as proxies for soil quality and degradation gradients. • GWR outperformed OLS, revealing non-stationary soil–environment relationships.