In Response: The NORMAN perspectives on prioritization of emerging pollutants

In our highly complex chemosphere, environmental managers need a rational strategy to help them identify the substances posing the greatest threat to ecosystems and human health. Ultimately, they need to guarantee consistency between the list of priority substances addressed in their program of measures and the pressures responsible for the decline in biodiversity. The major challenge, however, is that because of existing knowledge gaps (e.g., insufficient information on the effects of a substance, inadequate performance of the analytical method for quantifying its level of occurrence in the environment, insufficient understanding of substances' effects in mixtures), many chemicals of emerging concern cannot be satisfactorily evaluated or may even be overlooked. A list of priority substances—those representing a significant risk to or via the aquatic environment at the European Union level—is established and regularly reviewed under Article 16 of the Water Framework Directive. The most recent review (2007–2011) led to the current list of 45 substances reported in Directive 2013/39/EU. Member states also select the river basin–specific pollutants, that is, contaminants of potential concern for the ecological status of a river basin. Unfortunately, in the Environment Directorate-General's last review of the list of priority substances, approximately 50% of compounds on the list of candidate substances were discarded because of a lack of data or insufficient data reliability 1. So, emerging contaminants, for which data are often lacking, have very little chance of being identified for priority actions if conventional prioritization criteria are applied. The NORMAN network, as the interface organization between science and policy in the field of emerging contaminants, stimulates coordinated action to identify the emerging contaminants needing priority attention for further actions. The development of the NORMAN prioritization scheme 2, 3 started from the view that chemical contaminants need to be dealt with by taking into account existing knowledge gaps and actions needed to fill them, such as development of more powerful analytical methods, launch of monitoring campaigns, and new ecotoxicity tests. The NORMAN 2-stage procedure (see Figure 1) uses a decision tree that first classifies chemicals into 6 categories depending on the information available. The priority within each category is then evaluated based on specific property indicators such as persistence, bioaccumulation potential, and toxicological and ecotoxicological effects and on risk-assessment indicators such as the frequency of exceedance and the extent of exceedance of the lowest predicted-no-effect concentrations. The overall process involves a periodic revision of the priority substances in each category whenever, for example, new information and more reliable data are generated or feedback from applied reduction measures is available. The scheme is integrated in the web-based NORMAN EMPODAT database 4 as a tool for the prioritization of emerging substances using systematically collected data sets at the European level (e.g., in support of the development of the European Union Watch List by the Environment Directorate-General in 2014). The same scheme has already been successfully applied at the national and river basin levels (e.g., in France). The NORMAN methodology provides a decision-support framework for updating lists of substances for which actions (reduction, monitoring, and research) are to be undertaken as a matter of priority. The present scheme, however, is still biased toward known contaminants acting as individual substances. It is increasingly recognized that the number of compounds addressed today via target chemical analysis in environmental monitoring programs or via screening of long lists of suspect chemicals does not provide an exhaustive overview of the status of European water bodies. Thousands of substances typically present in each environmental sample together with their transformation products yield complex mixtures whose composition and effects can only partially be identified by current monitoring and risk-assessment schemes 5. The application of nontarget screening techniques coupled with biological tools is a promising approach opening a new era in environmental monitoring, in which biological responses and nontarget analysis will be used to screen the thousands of chemicals occurring in waters and to detect the presence of harmful substances that potentially have been overlooked. Strategies for more systematic inclusion in prioritization exercises of compounds identified via nontarget screening 6 and effect-directed analysis studies 7 are currently being developed by NORMAN in collaboration with the SOLUTIONS FP7 project 8, 9. In a recently adopted approach, data from nontarget screening are integrated into the NORMAN prioritization scheme. This relies on the assignment of substances to 5 different categories, corresponding to 5 identification confidence levels according to the concept developed by Schymanski et al. 10, followed by ranking the substances within each category. The idea is again to group substances according to defined criteria to better tackle knowledge gaps (and proceed with actions to fill them). The ranking of the compounds within each category relies on the application of indicators already defined by NORMAN such as the frequency of exceedance and the extent of exceedance of the lowest predicted-no-effect concentrations for the substances assigned to higher identification levels, with semiquantitative assessment of their concentrations, and the frequency of appearance for substances not yet identified. Moreover, it is foreseen that the newly developed chemicals' exposure index bringing together information about production tonnages and use patterns, exact masses of all produced chemicals from the REACH registry which may enter the environment, high-throughput ecotoxicity assessment in silico tools (e.g., ChemProp), and systematic use of retention time indices will be systematically applied to help interpret nontarget screening results. In this context the role played by effect-based tools is also crucial for future improvement of prioritization schemes. Biological tools and effect-directed analysis approaches are essentially applied today for prioritization of effects and relevant toxicants at the site level. The future challenge is to develop strategies to link chemical monitoring (and modeling) information with effect information for prioritization of chemicals for which actions should be taken at a larger, river basin or European scale. In conclusion, knowledge gaps about the occurrence, fate, and effects of the chemicals likely to be present in our environment remain a serious obstacle for clear identification of emerging contaminants for which management measures should be implemented. These gaps should be tackled in a systematic way as an integral part of the prioritization process (action-based prioritization), as recommended by NORMAN. A substance-based approach can still be considered valuable for the identification of contaminants for which frequency of exceedance and extent of exceedance of target environmental quality standards justify priority attention. Nevertheless, to better take into account the complexity of our environment and the effects of co-occurring contaminants, the implementation of batteries of bioassays combined with chemical analysis, and in particular nontarget screening, appear as the most promising way to identify contaminants responsible for observed effects at the site level. The final aim (and remaining challenge) will be to use these tools to support identification of more robust priority substances at a larger scale. For that, the future improved conceptual framework for prioritization of contaminants in European waters should come from a smart integration of substance-based and monitoring-based approaches. V. Dulio and J. Slobodnik are supported by the European Union's FP 7 Project SOLUTIONS (603437).