Microplastics in 132 Iowa lakes and variability in relation to abiotic, biotic, and anthropogenic factors

The global annual plastic production rate is approaching 400 million metric tons, with substantial amounts invading aquatic environments yearly. Numerous studies have been conducted monitoring marine plastic litter. An understanding of plastic litter's magnitude in freshwater ecosystems is lagging, particularly for microplastics (MP(s) 100 nm to 5 mm in length/diameter). Their ubiquitous presence and small sizes are concerning since MP effects remain inadequately understood. Our objective was to identify MP, measure concentration in lake surface waters and investigate how abiotic, biotic, and anthropogenic elements explain MP concentration variability among lakes. We sampled MPs in 132 Iowa lakes, collected throughout the water column using a Wisconsin net (53 mu m mesh size). A fully automated custom-built Bruker LUMOS-II Fourier-transform infrared (FTIR) spectromicroscope was employed to identify MPs. Average MP concentration across all lakes was 5 particles/L, with dimensions averaging 145 mu m in length and 80 mu m in width. Predominant MP polymers collected were polyvinyl chloride (46% of all MPs), polyester (30%), and polyethylene (11%). Two variance partitioning analysis models were created to explore variability in MP concentration. The first classical model explained 7.5% of data variability based on roads, developed medium and high-intensity land cover, sewage plants, and thermocline. The second model, built according to Louvain Groups, explained 8.3% of data variability based on lake area, maximum lake depth, zooplankton tow depth, Secchi depth, lake perimeter, lake shoreline development factor, evergreen forest cover, and thermocline. Roads and developed intensity cover were positively correlated with MP concentrations, alongside household visits standardized to the lake area. Maximum lake depth, zooplankton tow depth, and Secchi depth were negatively correlated with MP concentration. MP concentration was significantly lower if the thermocline was present at the time of sample collection. In seven lakes, MP concentrations exceeded half-value of the modeled hazardous concentration, affecting 5% of aquatic species.