Page 51 - Read online

Diaz-Vallejo et al. Carbon Footprints 2026, 5, 12 Page 7 of 23

Table 1. Summary statistics for soil organic carbon concentrations, silt + clay concentrations and mean annual temperature, and mean
annual precipitation across a database of 586 pedons in Puerto Rico

Min Median Mean Max Sd
SOC (%) raw 0.02 2.14 3.07 46.67 3.27
SOC (%) 0-30cm averaged 0.12 2.21 2.92 23.57 2.68
Silt + Clay (%) 2.60 83.10 80.90 100.00 16.91
MAT (°C) 18.10 24.00 23.98 26.70 1.64
MAP (mm) 752.00 1,810.00 1,804.68 3,156.00 513.51
MAT: Mean annual temperature; MAP: mean annual precipitation; SOC: soil organic carbon.

Table 2. Analysis of variance results for testing differences among land use types across a database of 586 pedons in Puerto Rico
Df Sum of squares Mean of squares F value P-value
Land use 4 54.61 13.651 29.5 < 0.0001
Residuals 581 268.87 0.463

RESULTS

Data descriptive results
Within the 0-30 cm layer, SOC concentrations among individual horizons ranged from 0.01% to 46.6%
[Figure 2A and Table 1], illustrating the wide variability of surface and organic-rich layers across Puerto
Rico’s soils. When averaged by pedon to represent integrated topsoil conditions, SOC values ranged from
0.12% to 23.6% (mean = 2.9%, sd = 2.7%, Table 1). The combined fine fraction (silt + clay) ranged from 8% to
99% (mean = 80%, sd 16.91%, Table 1). Climatic gradients were also well represented in the dataset, with
MAT spanning 18-27 °C and MAP ranging from 750 mm to 3,150 mm yr . Forest and pasture pedons
-1
accounted for most observations, followed by agricultural, wetland, and rangeland soils [Supplementary
Table 1].

Land-use effects on soil carbon
Mean SOC concentrations at 0-30 cm differed among land uses (ANOVA, P < 0.001; Table 2; Figure 2C). All
pairwise comparisons were significant [Supplementary Table 2] except those involving rangelands, which did
not differ from agriculture (P = 0.126), forests (P = 0.995), or pastures (P = 0.165, Supplementary Table 2).
Wetlands exhibited the greatest SOC (7.75 ± 5.80%), followed by forests (3.70 ± 3.51%), pastures (2.51 ±
1.36%), rangelands (2.20 ± 0.68%), and agriculture (1.62 ± 0.70%).

Factors affecting soil organic carbon in agricultural land and pastures
We evaluated the effects of soil order, soil suborder, MAT, MAP, silt + clay (%), pH, and USDA texture class
on SOC (%) in Agriculture [Tables 3 and 4, Figure 3] and Pastures [Tables 3 and 4, Figure 4].

Agricultural soils
Soil order (P < 0.001, R = 0.51) and suborder (P < 0.001, R = 0.50) were strong predictors of SOC in
2
2
agricultural soils. Climate variables also contributed (MAT: P = 0.001, R = 0.10; MAP: P < 0.001, R = 0.17).
2
2
Among soil properties, silt + clay (P < 0.001, R = 0.32) and texture class (P < 0.001, R = 0.19) explained
2
2
20%-32% of variability, while pH had a negligible effect (P = 0.003, R = 0.08). A stepwise multiple regression
2
identified a best-fit model for agricultural soils including soil order, silt + clay, MAT, and pH (P < 0.001, R =
2
0.60; Table 5).