Comparing soil-health- and species-based sensitivity distributions for metal risk assessment in Tianjin soils.

Soil contamination threatens ecosystem functioning, while conventional ecological risk assessment (ERA) based on single-species toxicity may not capture system-level responses. We developed a probabilistic soil health sensitivity distribution (SHSD) framework integrating microcosm-derived soil health indicators with joint probability modeling and applied it to Pb, Cu, and Cd risks in Tianjin soils, northern China. SHSD-derived risks were lower than species sensitivity distribution (SSD)-based estimates by 61.90%, 59.55%, and 84.38% for Pb, Cu, and Cd, respectively. Normalizing SSD toxicity data to microcosm soil conditions reduced part of the SSD-SHSD discrepancy, highlighting the influence of soil physicochemical heterogeneity. Observed SHSDs were also shifted rightward relative to additive, non-interactive predictions, indicating lower apparent sensitivity of the integrated microcosm. These findings support SHSD as a higher tier, function-oriented complement to conventional SSDs, while emphasizing the need for independent field validation before regulatory application.