Spatial assessment of soil fertility in high-Andean potato systems using a multivariate index in the Pasco region, Peru

Potato cultivation in the central Andes of Peru occurs under highly heterogeneous environmental conditions, where soil fertility is influenced by interactions among lithology, geomorphology, climate, and vegetation. This study evaluated the spatial variability of soil fertility in high-Andean potato systems across three river basins of the Pasco region, Peru, through the development of a multivariate Soil Fertility Index (SFI). A total of 573 soil samples were collected and analyzed for soil, climatic, and environmental variables. Principal Component Analysis was applied to construct a minimum dataset and assign indicator weights according to their contribution to the explained variance. The final SFI integrated 11 indicators, with pH, exchangeable Mg, precipitation, and cation exchange capacity identified as the most influential variables. The SFI was used to compare soil fertility among three river basins and across lithological, geomorphological, and life zone classes. Significant differences in SFI were observed among lithological, geomorphological, and life zones (p < 0.001). Clastic sedimentary and igneous lithologies exhibited the highest fertility median values (0.50), whereas metamorphic and carbonate sedimentary formations showed lower median values (0.42–0.46). Among geomorphological units, hills exhibited the highest SFI median values (0.51), while depositional slopes and terraces showed the lowest (0.45). Andean shrublands exhibited the highest fertility median values (0.53), whereas tropical dry forests showed the lowest (0.44). Districts with higher SFI values generally exhibited greater potato yields, although this relationship was influenced by additional environmental and management factors. Spatial interpolation using RK and GWRK enabled the characterization of soil fertility patterns across the three basins and the identification of areas with contrasting fertility conditions. Although model performance was moderate (R² = 0.10–0.33), the resulting maps provide a useful framework for soil fertility assessment, site-specific management, and sustainable agricultural planning in Andean potato production systems.