The goal of ecological risk assessment of chemicals (ERA) is to quantify the risk

that a given chemical would impair the structure and function of natural

ecosystems - by assessing its environmental exposure and the expected

ecological effects_ However, the environmental realism. ecological relevance.

and methodological accuracv of the currently used exposure and effect

assessment approaches have been questioned for the last 25 years (1-4).

·-Bearing in mind the-elogical-and-environmental-complexity-inherent to natural ecosystems, risk assessors increasingly realize that ecological risk cannot be adequately assessed using the existing cookbook procedures that disregard

most, if not all, of this complexity (4). In a forthcoming opinion document from SCENIHR (Scientific Committee on Emerging and Newly Identified Health Risks), SCHER (Scientific Committee on Health and Enl(ironmental Risks), and SCCS (Scientific Committee on Consumer Safety) of the European Union (5), new

challenges related to current ERA practices and possible solutions are discussed

in detail. Briefly, these challenges include:

• Exposure to chemicals is not constant in time and is not homogeneously

distributed in space. Different regions may exhibit site-specific bioavailabilitydetermining

physical and chemical characteristics, but also within one region

patterns of chemical emission as well as local variability of environmental

conditions may create a spatial and temporal mosaic of exposure (6, 7).

• Performance of individuals is not solely determined by the presence of

individual chemicals but by multiple stressors, including (mixtures of)

chemicals (8), possibly targeting different trophic levels (9), but also other

stress factors such as changes in temperature and sedimentation rates (10).

• Ecological effects are not necessarily irreversible, as recoverv may occur

both at the individual, and the population level upon removal/diminution of the

stressor, depending on the toxicokinetics and toxicodynamics of the chemical

(11) and the species' life history characteristics and immigration of unaffected

individuals (12, 13). Realistic and protective risk assessments should

therefore include the potential for recovery.

• ERA is still largely based on individual-level endpoints, although these are not

always the assessment endpoints of its protection goals (14). For

invertebrates, individual-level effects are initiators of effects at higher levels of

biological organization, i.e. populations. communities and ecosystems.

This propagation of individual-level effects may cause tolerant species or

ecosystem functions to be affected due to ecological interactions in the

community (15, 16), but likewise functional redundancy among species may

compensate species loss and sustain functions in stressed ecosystems (17).