Multivariate analysis of the factors influencing soil organic carbon stock (0–30 cm) based on land use and its transitions in the Atlantic Forest Biome
¹Fluminense Federal University, Department of Agricultural and Environmental Engineering, Niterói, Rio de Janeiro, BR24210-240, Brazil
²Fluminense Federal University, Graduate Program in Ocean and Earth Dynamics, Niterói, Rio de Janeiro, BR24210-240, Brazil
*Correspondence: annaamandipdb@gmail.com
Abstract:
Soil organic carbon (SOC) stocks co‑vary with land use, land‑cover dynamics and their temporal transitions in human‑modified biomes such as the Atlantic Forest. This study explores how land‑use classes and their annual transitions help to understand the dynamics of SOC variability in the 0–30 cm layer using MapBiomas Soil data and Principal Component Analysis (PCA). This study analyzed total SOC stocks, the areal extent of major land uses (forest, pasture, agriculture, non‑vegetated areas, forestry plantations, hydrography and herbaceous vegetation), and their dominant transitions. Correlation analysis indicated positive associations between SOC and pasture, herbaceous vegetation and native forest, whereas agriculture, forestry plantations and non‑vegetated areas were negatively related to SOC stocks, in line with recent findings for Brazilian landscapes. In the PCA, the first component (PC1) captured most of the variance and was strongly associated with SOC and transitions between open and vegetated areas, representing a gradient linked to the intensity of landscape conversion. PC2 was mainly associated with transitions involving gains in forestry plantations, corresponding to a gradient of plantation expansion or contraction, while PC3 was associated with transitions affecting water bodies and wetlands, indicating a secondary axis of SOC variation. Overall, the results suggest that both current land use and the direction and intensity of land‑cover change are key factors associated with SOC stocks. These patterns are consistent with the importance of conserving vegetation cover and promoting stable, sustainable land management to maintain and enhance soil carbon in agricultural and natural environments across the Atlantic Forest biome.
Key words:
land use, MapBiomas, principal component analysis, Remote sensing