The following questions / concerns were discussed:
- Are we making ecologically relevant assessments? Are regulatory protection goals explicit and clear? Are they set in relation to environmental quality? How do prospective and retrospective approaches differ?
It was considered that the correct place to start was with the protection goals. Within the legislation, there are generally no explicit statements of what we are trying to protect. Rather broad statements, such as “the need to protect ecosystem health”, or “have no unacceptable effects” are used. In addition, there is no such thing as ‘the’ ecosystem. Therefore, it is necessary to make these protection goals more specific. The use of ecosystem services allows different protection goals to be set based on a cost/benefit basis rather than just protecting everything everywhere all of the time. There is the possibility to set protection goals based on ecosystem structure and/or function, with the assumption that by protecting ecosystem structure, function and other ecosystem services would be protected.
It was apparent that there were different protection goals between different legislative frameworks. Examples of goals included:
- preservation of ecosystem structure and function based on measured endpoints;
- no to minimal impacts accepted (Sijm et al, 2002)
- acceptability of some effects (EC, 2009)
- the need to “protect aquatic life”, through the ecosystem structure, with essentially no effects acceptable (EC, 2000).
Examples were given of protection goals, mostly based on population, community or ecosystem structure/function, acceptable/no effects, with little quantification of the scale of effect. However for the US Clean Water Act, the stated protection goal is 95% of all species and this is achieved through setting a quality standard based on an HC5 from an SSD.
There was recognition of the limitations of the ecological relevance of SSDs. They are a collection of single species laboratory tests (or extrapolations), limited to direct effects, without interspecies interactions and so do not represent communities or ecosystems. Despite the fact that SSD-predictions are limited to predicting direct effects, indirect effects are more likely to be associated with concentrations to the right of the SSD curve (HC50+), since indirect effects cannot occur when direct effects are absent – so that indirect effects are unlikely at an HC5-level that is really protective. So do these limitations matter, provided there is empirical evidence that the risk assessment procedure is effective and ecosystems are protected? Certainly, the relatively limited available evidence does tend to suggest that communities are adequately protected by thresholds derived from SSDs compiled from single species laboratory tests.
Prospective and retrospective approaches to risk assessment and the application of SSDs to achieve their stated aims are very different. The aim of using an SSD in a prospective risk assessment is to derive a threshold which is protective, whilst SSDs in retrospective analysis are used in a predictive way, to assess the likely degree of impact on local communities. For prospective risk assessment a single parameter such as an HC5 may be all that is required from an SSD, whereas other characteristics of the SSD will be important for retrospective analysis, such as the relative sensitivity of species, and the slope.
2. Are all species of equal importance, or are there keystone species that are more important than others?
Participants at the workshop were in general agreement that all species are not of equal importance, there can be differences ecologically, economically or aesthetically, for instance. This question brings the discussion back to the protection goals, and deciding what it is we are trying to protect. If the protection goal is to preserve ecosystem structure, i.e. no effects, then it does not matter where the species are on the SSD. If using an SSD as a diagnostic tool, site specific protection goals (and testing of particular species) may be appropriate.
Protection goals may impact the species tested for an SSD, such as a site specific assessment, but there are normally other considerations such as availability, ability to culture, endangered species concerns or ethical considerations (vertebrates). Furthermore, not all tests are performed in light of protection goals. All of this can create a bias in the species that are being tested.
3. Is a generic PNEC derived from an SSD overly simplistic in terms of ecological representativeness or should we develop representative assemblages/communities (archetypes) to represent different typologies? Should protection goals account for local community composition?
4. How does aquatic community sensitivity vary with species composition?
The above 2 questions were tackled together. There was much discussion over the need to consider different communities taking into account geographical and climatic factors. Limited evidence exists to support whether or not these factors can be deemed important for the majority of organic chemicals, and evidence is therefore insufficient to draw general conclusions. For metals it is known that differences in water chemistry can affect not only community sensitivity, but also the relative sensitivity of different organisms within an SSD. Similarly for freshwater and marine communities, whilst acknowledging that these communities can be very different structurally, aside from chemical considerations affecting toxicity, there is little evidence of differences between the inherent sensitivity of the communities. When using SSDs in a prospective manner with no effects as the protection goal a single threshold value derived from a generic SSD is all that is needed, provided it is sufficiently protective. Specific community composition may be important when some effects are allowed as both direct and indirect effects may occur and recovery is included. Again, when using SSDs retrospectively in a site specific assessment the need to understand particular assemblages may be important. Approaches which are able to predict community composition in the absence of pressures can provide a valuable tool in assessing whether or not especially sensitive organisms would be likely to be present.
5. How can knowledge of chemical MoA help construct SSDs for HC5 estimation?
When chemicals have a non-specific, or unknown, mode of action, which species are tested should not make a difference when using SSDs in a predictive, protective manner. In these cases including a diverse taxonomic range of relevant organisms is likely to be the most appropriate approach. However when there is a specific mode of action, this needs to be taken into consideration when selecting species for testing and combining in an SSD.
6. What are the research needs?
Further validation of thresholds derived using SSD based approaches to adequately protect communities and ecosystems from adverse responses to predicted chemical exposures.