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?
The group did not consider this question.
2. Are all species of equal importance, or are there keystone species that are more important than others?
This depends on the ecological role of the species and whether this needs to be explicitly protected (is there functional redundancy?). Most keystone species are known and can be protected appropriately.
Some keystone species are actually tropho-species, i.e. assemblages of species with an ecological role, e.g. Krill in polar marine environments.
Keystone species should be protected and, therefore, ideally would also be included in the SSD. However, the likelihood of having test data or being able to generate it is dependent on other, more practical factors.
Charismatic species often do not have ecological importance (compared to keystone species) but can be important to human society, e.g. pandas.
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?
It seems likely that the species data available to fit an SSD for prospective estimation of an HC5 would be unrepresentative because they tend to be those species conducive to laboratory testing regimes and methods. However, this does not mean that the assumed statistical distribution of the SSD would be inappropriate. The group did not decide whether representative assemblages would make a substantial difference to the generic approach. However, the group did consider that local community composition should be considered in setting protection goals. Setting protection goals from policies informed by science is key to good management of chemicals. Once protection goals are set, scientists can design risk assessment strategies as necessary. This can benefit from use of SSDs.
The group noted that toxicity data generated from studies with laboratory cultured organisms may not be representative of toxicity in field organisms where the field organisms have developed a degree of tolerance to some chemicals, e.g. metals. This raises the question of whether laboratory tests should involve organisms pre-exposed to the test chemical. Of course, for chemicals not yet present in the environment, this would be inappropriate.
4. How does aquatic community sensitivity vary with species composition?
The work of Professor Maltby (section 3.2) suggests that differences in community structure do not relate to clear trends in sensitivity. Although there may be data gaps as in the case of chronic data for EPT (the Ephemeroptera–Plecoptera–Trichoptera taxa, considered as relatively sensitive taxa), does comparison of PNECs derived from HC5s with field biomonitoring data suggest such taxa are protected? It follows that there would be little value in developing generic community scenarios/archetypes to represent different community sensitivities.
At an organism-sensitivity level, taxonomic distance starts to be important at higher levels of organisation, i.e. order or above. This implies that it is more important to include a range of broadly different taxonomic species than to add more species that are taxonomically close to species already included in the distribution.
Differences in sensitivity between freshwater and saltwater organisms can vary more than between different freshwater communities, but this is likely to be due to differences in bioavailability/chemistry.
There may be specialised organisms with specific adaptations to local conditions that could lead to higher or lower sensitivity. For example, cold water adapted species, e.g. arctic cod, may adopt different excretion routes for some chemicals (via liver rather than kidney). If exposed to a pollutant that is metabolised in the liver, this could influence toxico-kinetics and, therefore, sensitivity compared to other fish. Of course, other factors such as slower rates of degradation will influence exposures and also must be taken in consideration when undertaking risk assessments in cold water marine environments.
5. How can knowledge of chemical MoA help construct SSDs for HC5 estimation?
The slope of an SSD, i.e. its cumulative distribution function, is indicative of a chemical’s MoA. This seems to work for distinguishing narcotics, where the slope is usually uni-modal and steeper, from chemicals with specific MoAs, where the distribution is usually flatter and multi-modal. Comparison of the slope of a chemical SSD with unknown MoA with SSDs for existing chemicals (with known MoAs) can also be useful.
For pesticides, which often have wide-ranging potencies to different taxonomic groups, the SSD should be constructed from the more sensitive taxa when this can be demonstrated or separate SSDs constructed for sensitive and less-sensitive tax.
Care is needed to ensure responses used in the SSD are comparable - responses for plants tend to be measures of growth inhibition whereas invertebrate tests measure binary endpoints such as mortality. These should not be mixed in the same SSD although they often are. If there are enough data to make a SSD for fish, invertebrates, and algae, it is useful to consider if the different taxa data overlap to decide if they should be combined to represent the ‘ecosystem’. Colour coding species in the SSD plot helps visualise sensitive groups.
US EPA may also look at the most sensitive 4 or 5 species in the SSD to estimate a conservative HC5. This clearly provides a strong conservative bias although it should be noted that the statistical endpoint, HC5, is influenced by the span of the sensitivity distribution. Using SSDs to set WQSs can be applied differently in different regulatory jurisdictions.
Knowledge of the mode or mechanism of action is useful to decide if taxa should be in one SSD for one MoA. MoA needs to be described at an appropriate level. The group thought mode of action was more practical than mechanism of action.
Knowledge of MoA could help decide which species to test, e.g. herbicides and algae/plants. However, this may be a problem when guidelines for test methods are not available for some taxonomic groups, e.g. molluscs.
6. What are the research needs?
Sensitivity of cold water communities, e.g. arctic.
Better application of toxicological data in SSDs, e.g. using more chronic data, mechanistic understanding.
More use of predictive modelling to overcome limited data sets.