Heavy metal contamination around the Daye mining district in Hubei, China, cannot be pinned on one element or one source, according to a new study that mapped and apportioned contamination across 394 topsoil samples. The research, published in Environmental Geochemistry and Health, tested two assumptions that often go unexamined in mining-district assessments: that ecological risk, the areal extent of contamination and human health risk are all driven by the same metal, and that contamination sources can be inferred simply from distance to the mine. Both assumptions failed.
The team measured seven metals in 394 topsoil samples collected from nine blocks across the district, comparing concentrations against the Hubei A-horizon background values. Median concentrations of five of the seven metals exceeded the regional background, with copper showing the strongest enrichment at 3.69 times background and cadmium at 3.00 times. The samples were mapped using inverse distance weighting and the Getis–Ord Gi* statistic to identify statistically significant hot spots, while sources were apportioned with positive matrix factorization and risk was evaluated with both the Hakanson ecological risk model and United States Environmental Protection Agency health risk models.
The most striking result is the divergence among the three risk endpoints. Cadmium supplied 64.3 percent of the mean ecological risk index, which had a median value of 200.7, placing much of the district in a high ecological risk category. Arsenic, by contrast, accounted for 82.1 percent of the carcinogenic risk in the health assessment. Copper, meanwhile, defined the areal extent of contamination, covering the largest mapped footprint. No single element, the authors conclude, can represent the district as a whole, and remediation or monitoring programs that focus on only one metal will miss the others.
Positive matrix factorization resolved four distinct source factors. One was copper-dominated, consistent with the district’s long history of copper mining and smelting. A second carried a cadmium–lead–zinc signature, a third was geogenic and marked by nickel, and a fourth reflected volatile elements, arsenic and mercury, likely associated with high-temperature processes such as smelting. The contributions of these factors followed ore type rather than distance to the mine pits, overturning the common practice of treating proximity to mining operations as a proxy for contamination source. Notably, the study measured no source-specific tracer, so the factor labels remain interpretive rather than confirmed by direct fingerprinting.
The health risk assessment produced a median children’s hazard index of 1.196, with 59.9 percent of sampled sites exceeding the threshold value of 1. The authors are careful to characterize this figure as an upper bound, derived from total metal concentrations rather than bioavailable fractions. When localized Chinese exposure parameters were applied instead of the default United States values, the median hazard index fell to 0.935, below the concern threshold. This sensitivity to parameter choice underscores how much health risk estimates for mining districts depend on the exposure assumptions built into the models.
Beyond the metals themselves, the study examined the pH–salinity state of the soils, a variable that covaries with remediation history across the district. This combined state tracked both of the anthropogenic source factors but not the geogenic nickel-marked factor. The association persisted even after block means were removed from the data, indicating it is not an artifact of spatial aggregation. The authors position pH–salinity as a monitoring covariate, a practical indicator that can flag anthropogenic influence in routine surveys, but they explicitly caution that it is not evidence of an effect of remediation on soil chemistry.
The hot-spot mapping also carries a methodological caveat. Classification of areas as statistically significant hot spots proved sensitive to the number of neighbors specified in the Getis–Ord Gi* analysis, meaning that mapped remediation priorities should be treated as exploratory rather than definitive. This sensitivity analysis adds to a growing recognition in spatial environmental studies that mapping outputs depend on analyst choices as much as on underlying contamination patterns.
The study’s design, sampling nine blocks and applying multiple risk models to the same dataset, allowed the authors to compare endpoints directly rather than assume they align.
The practical implications are straightforward. Ecological risk assessment, which in this case centers on cadmium, points to different priority areas than carcinogenic risk assessment, which centers on arsenic, and both differ from the copper-defined contamination footprint. A remediation plan built around any one of these endpoints would leave the others unaddressed. Similarly, distance-based source inference would misattribute the volatile arsenic–mercury factor, which follows ore processing rather than pit location.
For environmental managers in mining regions worldwide, the Daye results argue for multi-endpoint assessment and measured source tracers before committing remediation resources. The study also demonstrates that health risk figures reported from total-concentration models should be accompanied by sensitivity analysis using locally relevant exposure parameters, since the same dataset yielded a hazard index above and below the concern threshold depending on parameter choice. The pH–salinity covariate offers a low-cost monitoring tool, but one that the authors frame strictly as an indicator of anthropogenic influence, not a measure of remediation success.
Subject of Research: Source apportionment and multi-endpoint risk assessment of heavy metals in Daye mining-district soils
Article Title: Heavy metals in Daye mining-district soils: source apportionment, multi-endpoint risk and pH–salinity association
Article References: Yang, M., Chen, Z., Ye, M., Zhang, C., Xu, R., & Pan, X. (2026). Heavy metals in Daye mining-district soils: source apportionment, multi-endpoint risk and pH–salinity association. Environmental Geochemistry and Health, 48(17), Article 677. https://doi.org/10.1007/s10653-026-03584-z
Image Credits: AI Generated
DOI: 10.1007/s10653-026-03584-z
Keywords: heavy metals, Daye, mining district, soil contamination, source apportionment, positive matrix factorization, ecological risk, health risk, cadmium, arsenic, copper, hot-spot mapping
Tags: arseniccadmiumcopperDayeecological riskhealth riskheavy metalshot-spot mappingmining districtPositive Matrix Factorizationsoil contaminationsource apportionment
