Page 57 - Read online
P. 57
Diaz-Vallejo et al. Carbon Footprints 2026, 5, 12 Page 13 of 23
Figure 7. Random forest model validation. RMSE: Root-mean-square error.
Figure 8. Random forest variable importance. MAP: Mean annual precipitation; MAT: mean annual temperature, USDA Texture: US
department of agriculture texture classifications.
Variability in SOC across different land uses
The results of our study highlight the strong influence of vegetation cover and land use change on SOC
dynamics. SOC concentrations were highly variable in forested areas and wetlands. In forests, this variability
reflects a combination of factors, including land use legacies, climatic gradients, and intrinsic soil
characteristics. Historical disturbances such as deforestation, grazing, and cultivation could have left
persistent imprints on SOC in the secondary forests .
[18]
