Page 62 - Read online

Page 18 of 23 Diaz-Vallejo et al. Carbon Footprints 2026, 5, 12

diverse tropical landscapes.

The Soil Health Gap Benchmark, in contrast, is more effective at site-level comparisons between agricultural
or pasture soils and their forest counterparts. By using forest soil as reference conditions, this model can
reveal depletion or enrichment trends associated with management. Its performance, however, depends
heavily on the quality of the reference system. In Puerto Rico, where most forests are secondary, SOC values
may not represent undisturbed , leading to inconsistent gap estimates. Furthermore, applying the Gap
[18]
Benchmark at large scales introduces uncertainty because climatic gradients strongly affect SOC differences
between managed and natural lands [9,26,67] . Temperature, precipitation, and ecological zone interactions can
exaggerate or mask true management effects [27,28,77] . The predominance of secondary forests adds additional
variability, as studies show that SOC responses during tropical forest succession can range from gains to
losses or no net change [11,68,78] . These challenges limit the Gap Benchmark’s scalability beyond localized
studies.

Both approaches contribute valuable but different insights. The Scores Benchmark provides a generalized,
continuous scale for evaluating soils relative to regional expectations, while the Gap Benchmark offers
targeted insights into site-specific carbon loss or recovery. For the tropics, where environmental
heterogeneity and land-use history legacies influence soil conditions, score-based models are likely more
reliable and scalable. Nevertheless, integrating both frameworks could yield the most robust assessments of
SOC to inform soil health characterization: Scores Benchmarks for regional diagnostics and Gap
Benchmarks for localized monitoring and management evaluation. As the U.S. Department of Agriculture’s
Natural Resources Conservation Service recognizes SOC as a key dynamic property underpinning soil
resilience, developing benchmarks that are both scientifically rigorous and accessible to policymakers and
farmers remains essential. Our findings indicate that the Scores Benchmark, by contextualizing SOC relative
to climatic and edaphic factors, offers the most practical tool for tropical regions, enabling users to determine
whether their soils perform above or below the regional carbon potential.

Conclusion
This study evaluated how agriculture, pasture, and forest land uses influence SOC variability across the
environmentally heterogeneous landscape of Puerto Rico and used this regional synthesis to evaluate SOC
benchmarking approaches for tropical systems. SOC differed among land uses, with soil suborder, land use
type, USDA texture classification, mean annual temperature, and soil pH jointly explaining a substantial
proportion of observed variability. The magnitude and direction of SOC differences among forests, pastures,
and agricultural lands varied across soil orders and climatic conditions, demonstrating that land-use effects
on SOC in tropical systems are strongly conditioned by environmental context.

These results are particularly relevant given that much of the existing SOC literature in the tropics has
emphasized highly weathered soils and humid environments, despite the broader range of soils, climates, and
land-use histories that characterize tropical regions. Our findings show that SOC responses to land use
cannot be reliably interpreted using a single reference condition or narrow environmental subset,
underscoring the need for benchmarking approaches that accommodate this uneven coverage of tropical
conditions.

Comparison of benchmark frameworks highlighted important differences in their applicability. The Soil
Health Gap approach was highly sensitive to soil order and climate, reflecting its reliance on reference
assumptions that are difficult to satisfy in tropical landscapes shaped by long and complex land-use histories.
In contrast, the Scores Benchmark captured SOC variability across diverse soils and land uses by situating
observed values within empirically derived distributions, allowing SOC to be evaluated relative to regional