Freshwater ecosystems are increasingly threatened by multiple anthropogenic stressors. Release of treated sewage effluent and pollution from agricultural or urban sources can independently reduce water quality with implications for ecological communities. However, our knowledge of the combined effects of these stressors is limited. We performed a field study to quantify the combined effect of treated sewage discharge and land use on nutrient concentrations, sewage fungus presence, and communities of macroinvertebrates and benthic algae. Over three seasons in four rivers we found that a model which included an interaction between sewage pollution and time of the year (i.e., months) was the best predictor of nutrient concentrations and the abundance of algae and sewage fungus. Both macroinvertebrate and algae communities shifted downstream of sewage input. Specifically, more tolerant groups, such as cyanobacteria and Oligochaetes, were more abundant. The EPT (Ephemeroptera, Plecoptera, and Tricoptera) water quality score was best explained by an interaction between month and agriculture in the surrounding landscape. Overall, our results show that sewage discharge has a significant impact on water quality and benthic riverine communities, regardless of the surrounding land uses. Agricultural inputs, however, could be more important than treated sewage discharge in reducing the abundance of sensitive invertebrate taxa. We need both improvements to wastewater treatment processes and reductions in agricultural pollution to reduce threats to vulnerable freshwater communities.

The study was conducted during three sampling campaigns in 2021 at four rivers in the UK. Two areas were sampled at each river: 5-10 meters upstream (controls) and 5-10 meters downstream (impacts) of sewage discharge points. Nutrient analysis and sewage fungus estimation were performed on collected water samples. Benthic algae biomass was quantified using BenthoTorch. Macroinvertebrates were collected using Surber samples and identified in the lab. Land cover classification used Sentinel-2 satellite data through Google Earth Engine, distinguishing "agriculture" and "urbanisation" types.

Water samples from upstream and downstream areas were collected on each occasion to collect sewage fungus. Nutrient concentrations (nitrate, phosphate, chloride, sulfate) were analysed, and sewage fungus abundance was enumerated using preserved samples using a FlowCam machine.

Benthic algae biomass was measured using BenthoTorch on rocks collected from streams. Macroinvertebrates were collected using Surber samples, preserved, identified, and counted in the lab. The EPT score indicating sensitive macroinvertebrate groups was calculated.

Google Earth Engine and Sentinel-2 data were used to classify land cover. A region encompassing upstream areas was selected, and a supervised classification was conducted based on NDVI bands, defining "agriculture" and "urbanisation" categories. Percentages of these categories were determined for each area.

R studio was used to perform all the statistical analysis and to create the graphs.