Dinâmica espacial do carbono e atributos do solo em sistemas de uso da terra no Cerrado

Abstract

Soil carbon stock is a crucial indicator of soil health and an important component of climate‑change mitigation, being directly influenced by land‑use and management practices. Low‑carbon agriculture seeks to reconcile food production with environmental sustainability, aligning with commitments to reduce greenhouse‑gas emissions. This study primarily aimed to evaluate the spatial variability of soil carbon stock (SCS) in the 0–60 cm layer among Integrated Livestock–Forest (IPF), Cropland, and Pasture systems, and to analyze the relationship between total carbon and soil physicochemical variables. Soil samples were collected at depths of 0–20, 20–40, and 40–60 cm, and analyses were performed to determine a range of physicochemical attributes. Principal Component Analysis (PCA) was used to identify the attributes that most contributed to variability in soil properties among management systems. Subsequently, Spearman’s correlation analysis was applied to test the association between total carbon and soil attributes within each system. Means of soil carbon stock were statistically compared using the nonparametric Mann–Whitney test (p < 0.05). PCA results indicated that attributes such as aluminum saturation, exchangeable aluminum, organic matter (OM), pH, and base cations (calcium, magnesium, potassium), in addition to textural fractions and sulfur, were the main determinants of variability in soil attributes across the evaluated systems. The correlation analysis showed that OM had the strongest positive and significant correlation with total carbon in all managements (IPF, Cropland, and Pasture). In contrast, aluminum and aluminum saturation consistently displayed negative correlations with carbon. Correlations for other attributes varied among systems: sand correlated negatively in Cropland and Pasture but not in IPF, and potassium showed no significant correlation in Pasture. For soil carbon stocks (0–60 cm), Cropland had the highest mean (117.22 t ha⁻¹), followed by IPF (107.5 t ha⁻¹) and Pasture (93.17 t ha⁻¹), revealing that SCS differs statistically among systems, depending on management and each area’s physicochemical characteristics. Cropland and IPF systems demonstrated greater potential to maintain or increase soil carbon compared with Pasture in the study area. Organic matter is the primary factor influencing soil carbon stock across all managements, reinforcing the importance of agricultural practices aimed at building OM for the sustainability of agroecosystems.