Today, Dr Jonathan Jupke tells us about his latest publication, where he and his colleagues explored how river macroinvertebrates across Europe respond to common pollutants, such as copper and imidacloprid, and what these findings mean for environmental risk assessment.
As a researcher studying freshwater ecology, I am fascinated by how aquatic communities respond to environmental stressors across different habitats, what you might call the context-dependence of stressor-responses. When the opportunity arose to investigate spatial patterns in the sensitivity of river macroinvertebrates to common pollutants at a continental scale, I was excited to dive in. Why does it matter? First of I find it very interesting to uncover large-scale patterns in ecology. However, from a practical perspective, strong and systematic differences between sensitivities would undermine the current risk assessment schemes. Current environmental risk assessment practices typically use a one-size-fits-all approach, applying a single effect threshold across different ecosystems. Assessment factors are used to account for uncertainties, including potential variations in sensitivity among different environments.
The Study
Our study, recently published in Environmental Sciences Europe, examined whether the sensitivity of macroinvertebrate assemblages to copper and imidacloprid varied systematically among different river types across Europe. We analyzed an extensive dataset of over 13,000 macroinvertebrate samples with 2,197 unique taxa from rivers spanning the continent (Figure 1). The problem is that we lack sensitivity data for most of these taxa. To be precise, we found LC50 or EC50 for 59 and 33 taxa for copper and imidacloprid in the US EPA ECOTOXicology Knowledgebase. Note, that we already selected the chemicals to maximize the amount of available data.
To determine the difference in sensitivity between our assemblages, we needed to predict unmeasured sensitivities. We did so using a hierarchical species sensitivity distribution model. The model assumes that closely related species (like those in the same family or genus) are likely to have similar sensitivities to chemicals and creates a kind of “family tree” for all the species we’re interested in, including those we haven’t tested directly. Using the known toxicity data and the family relationships, the model makes predictions about the sensitivity of untested species. Using this predicted sensitivity data, we estimated the concentrations of copper and imidacloprid that would be detrimental for 5% of species in each community (HC5).
The Results
Interestingly, we found that while sensitivity (i.e., HC5s) did vary among river types, the differences were relatively small. The largest difference we observed was for copper sensitivity between very large rivers and highland streams. The median HC5 of highland rivers was three times the median HC5 of lowland rivers. This magnitude of variation falls within the assessment factors currently used in environmental risk assessment, suggesting that a single threshold concentration may be reasonably protective across different river types. Our findings suggest that this approach may be adequate for addressing the variation in inherent sensitivity we observed among river types. However, our study focused only on taxonomic composition. Future research examining how bioavailability and interactions with other stressors vary among river types could reveal larger differences in real-world impacts of these chemicals on aquatic communities. Such findings might necessitate a more nuanced, habitat-specific approach to risk assessment in the future.
This project was supported by the European Chemical Industry Council’s Long-Range Research Initiative.
You can read the full paper under open-access Europe-wide spatial trends in copper and imidacloprid sensitivity of macroinvertebrate assemblages