Land-use-dependent source apportionment around a closed Pb-Zn tailings pond: Integrating PMF, Pb isotopes and MixSIAR model.

Legacy Pb-Zn tailings ponds can remain long-term sources of heavy metal(loid) contamination even after mine closure, but distinguishing direct source inputs from secondary redistribution pathways remains challenging in small and heterogeneous rural landscapes. This study investigated topsoil contamination around a closed Pb-Zn tailings pond in Xiangtan County, Hunan Province, China, where farmland, woodland, irrigation canals, village roads and residual tailings-related materials occur in close proximity. Elemental concentrations, spatial interpolation, positive matrix factorization (PMF), Pb isotope fingerprinting and MixSIAR Bayesian mixing models were integrated to identify source factors and quantify Pb source contributions in farmland and woodland topsoils. Mean Cd and Pb concentrations were 0.43 and 56.04 mg kg-1, respectively, representing 8.6- and 5.7-fold enrichment relative to the local background. PMF resolved four factors, including a background-dominated lithogenic factor with diffuse mining/industrial overprint (14.73%), a secondary hydrological/sedimentary redistribution factor (27.49%), a localized industrial or combustion-related particulate factor (10.77%), and a mixed mining-metallurgical source (47.01%). Pb isotope diagrams showed substantial end-member overlap but indicated that mixed geogenic and anthropogenic Pb inputs. MixSIAR further indicated that atmospheric deposition (23.20%) was the largest Pb contributor to farmland soils, followed by road dust and irrigation water (18.2%), while woodland soils showed broadly comparable contributions among the potential sources (13.6-15.1%). These results demonstrate that Pb accumulation was controlled by multiple source inputs and subsequent by dust deposition, road resuspension, irrigation-related transport and land-use-dependent redistribution. The integrated framework provides a practical basis for source-oriented management and targeted remediation in legacy tailings-affected landscapes.