Integrating soil enzyme activities, organic carbon, and soybean yield to assess soil health trajectories in a long-term crop rotation under no-tillage and conventional tillage

This study evaluated soil health (SH) trajectories over six years, using a four quadrant model (4QM) in a long-term soybean ( Glycine max (L.) Merr.) crop rotation experiment conducted on a clayey Oxisol in the Brazilian Cerrado (0–10 cm). Treatments included soybean monoculture under no-tillage (NT) and conventional tillage (CT), as well as systems with cover crops. Soil assessments included arylsulfatase (ARYL) and β-glucosidase (GLU) activities, SOC, and chemical properties. These indicators were integrated into soil functions related to nutrient cycling, storage, and supply, while the 4QM framework was used to identify regenerative and degrading SH trajectories. No-tillage combined with cover crops enhanced biological indicators compared to monoculture. Chemical indicators remained above adequacy thresholds, reflecting constructed soil fertility. From 2016–2021, soybean yields under CT and NT monocultures were similar and the lowest, whereas treatments including cover crops and the soybean–maize systems showed higher and more stable yields. By 2022, two years after introducing Brachiaria ( Urochloa ruziziensis ), a deep-rooted grass, into CT, increased ARYL and GLU activities translated into higher soybean yields. Soybean yield was positively associated with nutrient cycling and storage, but not with nutrient supply, indicating that chemical indicators alone are insufficient to capture soil functional complexity in soils with constructed soil fertility. The 4QM provided an effective framework for identifying SH trajectories, enabling the distinction between regenerative and degrading systems over time. The integration of ARYL and GLU with SOC provides a robust, scalable approach for assessing SH and guiding management decisions in tropical agroecosystems.