Function-based and PCA-derived soil health assessments reveal convergent land-use responses and functional trade-offs in Brazilian drylands

Drylands are highly vulnerable to soil degradation, yet soil-health assessments remain limited for agroecosystems in the Brazilian semi-arid region, and the consistency between diagnoses produced by different assessment frameworks remains poorly understood. This study compared two soil-health assessment methods and evaluated land-use effects on soil health across Brazilian dryland agroecosystems. Soil health was assessed using (i) a soil-function framework integrating chemical, physical, and biological indicators into five functional domains and a multifunctionality metric and (ii) a variance-weighted soil health index (SHI) derived from principal component analysis (PCA). Measurements were obtained from the 0-30 cm layer across three semi-arid sites of Brazil (Sobral-CE, Betânia do Piauí-PI, and Petrolina-PE). Across sites, several diversified agroecosystems showed soil-health scores comparable to native vegetation at the time of sampling and, in some cases, higher scores for specific soil functions. The two frameworks captured similar land-use patterns but differed in indicator contributions: the PCA-derived SHI emphasized chemical variation, whereas the soil-function framework highlighted function-specific strengths and limitations. Soil health interaction functional trends showed that management effects were often expressed as functional trade-offs rather than uniform gains across all functions. Together, the results indicate that the two frameworks provide convergent overall diagnoses but complementary information, with the function-based assessment offering additional resolution for interpreting functional strengths, limitations, and trade-offs. These findings provide a consistent basis for improving soil-health assessment strategies and supporting sustainable land-use management across environmentally vulnerable dryland agroecosystems.