Multi-criteria decision analysis of optimal planting for enhancing phytoremediation of trace heavy metals in mining sites under interval residual contaminant concentrations.

As one of the most cost-effective and sustainable methods for contaminants' removal, sequestration and/or detoxification, phytoremediation has already captured comprehensive attention worldwide. Nevertheless, the accurate effects of various spatial pattern in enhancing phytoremediation efficiency is not yet clear, especially for the polluted mining areas. This study designed nine planting patterns (monocropping, double intercropping and triple intercropping) of three indigenous plant species (Setaria viridis (L.), Echinochloa crus-galli (L.) and Phragmites australis (Cav.) Trin. ex Steud.) to further explore the effects of plants spatial pattern on phytoremediation efficiency. Considering the uncertainties of the residual contaminants' concentration (RCC) caused by soil anisotropy, permeability and land types, the interval transformation was introduced into the plant uptake model to simulate the remediation efficiency. Then multi-criteria decision analysis (MCDA) were applied to optimal the planting patterns, with the help of criteria of (a) the amount of heavy metal absorption; (b) the concentration of residual contaminant in soil; (c) root tolerance of heavy metals; (d) the total investment cost. Results showed that (1) the highest concentrations of Zn, Cd, and Pb of the polluted area were 7320.02, 14.30, 1650.51 mg kg^-1 (2) During the 180 days simulation, the highest RMSE of residue trace metals in soil are 3.02(Zn), 2.67(Pb), 2.89(Cd), respectively. (3) The result of IMCDA shows that the planting patterns of Setaria viridis, Echinochloa crus-galli and Phragmites australis in alternative a9 (269 mg kg^-1 year^-1) had the highest absorption rate of heavy metals compared with a7 (235 mg kg^-1 year^-1) and a2 (240 mg kg^-1 year^-1). After 20 years of remediation, the simulated RCC in a9 is far below the national standard, and the root toxicity is 0.12 (EC ≤ EC₂₀). In general, the optimal alternative derived from interval residual contaminant concentration can effectively express the dynamic of contaminant distribution and then can be effectively employed to evaluate the sustainable remediation methods.