Fatty-acid based assessment of benthic food-web responses to multiple stressors in a large river system

Rivers are often exposed to multiple stressors, such as nutrients and contaminants, whose impacts on the river food webs may not be distinguished by sole assessment of biological community structures. We examined the benthic algal assemblages and the fatty acids (FA) of benthic macroinvertebrates in the lower Athabasca River in Canada, aiming to assess the changes in algal support and nutritional quality of the benthic food web in response to cumulative exposure to natural bitumen, municipal sewage discharge (hereafter, "sewage"), and oil sands mining ("mining"). Data show that the decline in water quality (increases in nutrient concentrations and total suspended solids) was associated with decreases in benthic diatom abundance, and was driven mainly by sewage-induced nutrient enrichment. Responses in nutritional quality of benthic macroinvertebrates, indicated by their polyunsaturated FA (PUFA) concentrations, were taxon- and stressor-specific. Nutritional quality of the larval dragonfly predator, Ophiogomphus, decreased nonlinearly with decreasing benthic diatom abundance and was lowest at the sewage-affected sites, although exposure to natural bitumen also resulted in reduced Ophiogomphus PUFA concentrations. In contrast, the PUFA concentrations of mayfly grazers/collector-gatherers were not affected by natural bitumen exposure, and were higher at the sewage and sewage+mining sites. The PUFA concentrations of the shredder Pteronarcys larvae did not change with cumulative exposure to the stressors. Sediment metal and polycyclic aromatic compound concentrations were not associated with the macroinvertebrate FA changes. Overall, we provide evidence that sewage induced reduction in trophic support by PUFA-rich diatoms, and was the predominant driver of the observed changes in FA composition and nutritional quality of the benthic macroinvertebrates. Fatty-acid metrics are useful to untangle effects of concurrent stressors, but the assessment outcomes depend on the functional feeding guilds used. A food-web perspective using multiple trophic levels and feeding guilds supports a more holistic assessment of the stressor impacts.