RESUMEN
Environmental characteristics including water quality and sediment properties alter the hazard that metals pose to aquatic systems by governing the speciation and partitioning of metals between water, sediment, and biotic ligands; however, alternate ligands are being introduced into aquatic systems through anthropogenic activity. Microplastics are a ligand on which metals interact through adsorption to the plastic surface. It remains unknown what factors determine the amount of metal bound to microplastic. Using a combination of laboratory experiments and machine learning, we tested a suite of eighteen environmental parameters (inclusive of both water and sediment) to understand how they influence association of cadmium to a representative microplastic, polyethylene. From this, we developed and tested a predictive model that outlines the characteristics that favour the association of cadmium to microplastic. Alkalinity, humification index of dissolved organic matter, and pH (all of which are water quality characteristics) were the three factors determining the proportion of cadmium adsorbed to plastics. These results align with other predictive models, such as the Biotic Ligand Model in demonstrating the governance of metal behaviour by water quality characteristics. To assess the relationship of the amount of cadmium bound to microplastic and cadmium uptake, an exposure was completed in which fathead minnows (Pimephales promelas) were acclimated to environments representing each of the potential outcomes of the model. The uptake of cadmium was not significantly different between groups, indicating that the stress of alterations to water quality may be a confounding factor in determining the exposure risk of microplastics and cadmium.
RESUMEN
Wetlands have been used to treat anthropogenic effluents for decades due to their intense biogeochemical processes that transform and uptake nutrients, organic matter, and toxins. Despite these known functions, we lack generalizable knowledge of effluent-derived dissolved organic matter (DOM) cycling in wetlands. Here, we quantify the cycling of DOM in one of Canada's more economically important wetland complexes (Frank Lake, Alberta), restored to hydrologic permanence in the 1980s using urban and agro-industrial effluents. Optical analyses and PARAFAC (parallel factor analysis) modelling showed a clear compositional change from more bioavailable and protein-like DOM at effluent input sites to more aromatic and humic-like at the wetland outflow, likely due to DOM processing and inputs from marsh plants and wetland soils. Microbial incubations showed that effluent DOM was rapidly consumed, with the half-life of DOM increasing from as low as 35 days for effluent, to 462 days at the outflow, as a function of compositional shifts toward aromatic, humic-like material. Long-term averaged dissolved organic carbon (DOC) export was low compared to many wetlands (10.3 ± 2.0 g C m-2 yr-1). Consistent with predictions based on water residence time, our mass balance showed Frank Lake was a net source of DOM across all measured years, but shifted from a source to sink among wet and drought years that respectively shortened or lengthened the water residence and DOM processing times. Overall, Frank Lake processes and transforms effluent DOM, despite being a longer-term net source of DOM to downstream environments. Supplementary Information: The online version contains supplementary material available at 10.1007/s10533-022-01002-x.