Assessment of stormwater discharge contamination and toxicity for a cold-climate urban landscape

Stormwater is water resulting from precipitation events and snowmelt running off the urban landscape, collecting in storm sewers, and typically being released into receiving water bodies through outfalls with minimal to no treatment. Despite a growing body of evidence observing its deleterious pollution impacts, stormwater management and treatment in cold climates remains limited, partly due to a lack of quality and loading data and modeling parameters. This study examines the quality of stormwater discharging during the summer season in a cold-climate, semi-arid Canadian city (Saskatoon, Saskatchewan).

Seven stormwater outfalls with mixed-land-use urban catchments > 100 km² were sampled for four summer (June-August 2019) storm events and analyzed for a suite of quality parameters, including total suspended solids (TSS), chemical oxygen demand (COD), dissolved organic carbon (DOC), metals, and targeted polyaromatic hydrocarbons (PAHs). In addition, assessment of stormwater toxicity was done using the two toxicity assays Raphidocelis subcapitata (algae) and Vibrio fischeri (bacteria). Notable single-event, single-outfall contaminant pulses included of arsenic (420 µg/L), cadmium (16.4 µg/L), zinc (924 µg/L), fluorene (4.95 µg/L), benzo[a]pyrene (0.949 µg/L), pyrene (0.934 µg/L), phenanthrene (1.39 µg/L), and anthracene (1.40 µg/L). The IC₅₀ in both R. subcapitata and V. fischeri was observed, if at all, above expected toxicity thresholds for individual contaminant species. Principal component analysis (PCA) showed no clear trends for individual sampling sites or sampling dates. In contrast, parameters were correlated with each other in groups including DOC, COD, TSS, and reduced algal toxicity; and total dissolved solids (TDS), sum of metals, and pH.

In general, stormwater characteristics were similar to those of previous studies, with a bulk of contamination carried by the first volume of runoff, influenced by a combination of rainfall depth, antecedent dry period, land use, and activity within the catchment. Roads, highways, and industrial areas contribute the bulk of estimated contaminant loadings. More intensive sampling strategies are necessary to contextualize stormwater data in the context of contaminant and runoff volume peaks.

The online version contains supplementary material available at 10.1186/s12302-022-00619-x.