Changes in soil organic matter fractions induced by cropland and pasture expansion in Brazil's new agricultural frontier

Recently, the so-called Matopiba region (northeastern Brazil) has experienced an accelerated agricultural expansion, with large areas of native vegetation (NV: Cerrado biome) converted to extensive pasture (PA) and cropland under no-tillage (CL). However, because soils from this region contain a large amount of sand (> 60%) and a high susceptibility to erosion, concerns have been raised about the impacts of land-use change (LUC) on soil organic matter (SOM) dynamics. Therefore, we assessed the magnitude that agricultural expansion affects SOM fractions in the most common LUC scenarios observed in the Matopiba region. Soil samples were collected to a depth of 1-m in three adjacent sites (i.e., NV, PA, and CL) located in a representative agricultural area of the region. Soil organic matter was physically fractioned into particulate organic matter (POM) and mineral-associated organic matter (MAOM); and carbon (C) and nitrogen (N) were determined in SOM fractions, in addition to in bulk soil samples. Additionally, the C management index (CMI) was used as a proxy to infer soil health changes. The conversion from NV to PA decreased C (by ~28 Mg ha−1 at 0–100 cm) and N (by ~0.5 Mg ha−1 at 0–50 cm) stocks at an annual loss rate of ~1.3 and 0.02 Mg ha−1, respectively. Carbon loss occurred in both POM and MAOM fractions, which were lower in ~36% and 34%, respectively, compared with NV (0–100 cm). This C depletion resulted in the smallest CMI value (75%) in the 0–30 cm soil layer in PA. The conversion from NV to CL had no effect on total C and N stocks (0–100 cm), but impacted C dynamics in SOM fractions. A lower value was observed for MAOM-C (~22%) in the 30–70 cm layer compared to NV, but the C accrual in the POM fraction (~24%) in the uppermost soil layers (0–30 cm) offset MAOM-C loss. This increase in the POM stocks enhanced the CMI, which was 37% higher than NV for the 0–30 cm layer. Our results revealed that PA expansion in the Matopiba region, besides declining C and N levels, reduced soil health as indicated by the low CMI values. On the other hand, land conversion from NV to CL did not impact C and N stocks and increased CMI. However, we advocated that agriculture expansion over NV areas is not a sensible option to promote the region's sustainable development. It should be prioritized on degraded PA areas and be based on best management practices as a strategy to recarbonize PA soils and restore soil health. This is paramount to mitigate climate change and i