Page 47 - Read online

Diaz-Vallejo et al. Carbon Footprints 2026, 5, 12 Page 3 of 23

poses challenges in regions where deforestation has been extensive and land-use histories are poorly
documented, making it difficult to define reliable baseline SOC conditions. In tropical regions, many forests
are successional and reflect past human use rather than long-term undisturbed conditions, complicating
their use as reference systems for SOC benchmarking [17,18] . In the absence of unconverted forests, successional
or secondary forests can serve as reference proxies due to their capacity for aboveground biomass
recovery [19,20] , yet SOC recovery in these forests is often inconsistent [18,21,22] . As a result, using secondary forests
in the Soil Health Gap benchmark may lead to biased assessments where no primary forests remain,
overestimating “recovery” and underestimating losses when available reference soils already have reduced
SOC stocks.

An alternative approach is to develop score-based benchmarks that use the distribution of SOC values within
specific environmental conditions to assess soil health. These models generate standardized scores that allow
users to evaluate whether SOC levels are low or high relative to regional variability. Examples include the
Comprehensive Assessment of Soil Health (CASH), which employs cumulative distributions of regional data
from the northern United States , and the Soil Management Assessment Framework (SMAF), which
[22]
integrates biological, chemical, and physical indicators using normalized 0-1 scores . A more recent tool,
[23]
the Soil Health Assessment Protocol and Evaluation (SHAPE), builds on both approaches by grouping soils
according to climatic and edaphic similarities . These frameworks demonstrate the value of using
[24]
region-specific benchmarks to interpret SOC variability, but their development for tropical ecosystems
remains limited. Creating such benchmarks requires a detailed understanding of the climatic and
physicochemical factors that regulate SOC dynamics.

Climate and soil properties are key drivers of SOC at both global and regional scales [25,26] . Temperature and
precipitation regulate organic matter inputs and decomposition rates, while soil texture and mineralogy
affect SOC stabilization and loss [27-29] . Although clay content is often considered a strong predictor of
SOC [30-32] , its explanatory power can vary across environmental conditions, with other soil variables, such as
pH, fine silt, clay, and mineralogy becoming more important . In an earlier study in the U.S. Caribbean,
[6]
Vaughan et al. reported that clay alone was a poor predictor of SOC and suggest combining clay + silt for
[7]
better results. Geologic substrates can mediate the effects of climate and land use on SOC at regional
scales [7,33,34] . These studies highlight the need to better understand interactions between land use, soil
properties, and climate to improve quantification of SOC inventories and modeling efforts of SOC response
to environmental change.

In this study, we (1) evaluated how agricultural land use and vegetation cover influence SOC variability
across a region with a diversity of soil and climate environments representative of tropical ecosystems; (2)
identified the primary factors controlling SOC within each land-use type; and (3) developed a
regionally-derived Scores Benchmark using the statistical distribution of SOC values to represent variability
across environmental gradients, and compared it with the conceptual Soil Health Gap framework to evaluate
their applicability for tropical soil health assessment. We hypothesized that agricultural lands would exhibit
smaller average SOC stocks and reduced variability compared to forests, reflecting the effects of intensive
management and residue removal. We further hypothesized that SOC variability at the regional scale would
be modulated by climate, while within-land-use variability depends on soil physicochemical characteristics.
Lastly, we expected that the Soil Health Gap model would be less effective for tropical regions lacking
primary forest reference sites, and that a Score-based benchmark capturing geographic variability provides a
more practical and scalable approach. We use Puerto Rico as a case study due to its well-documented
land-use history, diversity of soils and climates, and extensive SOC datasets. By synthesizing 586 pedons
representing nine US Department of Agriculture (USDA) soil orders, our work contributes a regionally
grounded framework for improving SOC prediction and benchmarking in tropical environments. Together,