Microplastics in rivers along an urban-rural gradient in an urban agglomeration: Correlation with land use, potential sources and pathways

Microplastics are ubiquitous and affect all environments, including rivers. In recent years the number of studies about microplastics in rivers has strongly increased. But still many questions exist regarding sources, pathways, and the role of land use patterns. In this study the relationship between microplastics abundance and anthropogenic factors (population density, urbanization, land use types), as well as the potential role of storm sewers as pathways in tributaries of the Wu River in Taichung, central Taiwan, were studied. Two river catchments of the Dali River were studied in greater detail to investigate the influence of land use on microplastics abundance along an urban-rural gradient, and to observe the change of microplastics abundance in the transition from rural to urban areas. Samples were taken from 41 different locations in urban and rural areas using a manta net with a mesh size of 0.3 mm. Results show abundances ranging from 0 pcs/m³ in unpopulated rural areas up to 230 pcs/m³ in densely populated urban centers, and are positively correlated with population density. Remarkably, a sharp increase in microplastics abundance was observed at the transition from rural to urban areas, which coincides with the appearance of storm sewers. Land use analysis revealed that microplastics abundance positively correlates with the size of industrial, residential and traffic areas in the catchment areas, and negatively correlates with the size of forest areas. Source areas for microplastics in the studied rivers are likely residential and commercial areas. Furthermore, the results of this study show that correlations between microplastics abundances and population density or land use patterns along urban-rural gradients are not trivial. Strength of correlations can depend on local factors or how well urban-rural gradients are developed. Absence of correlations need to be considered carefully, as existing correlations might be masked by the above-mentioned factors.