Accounting for bioaccumulation in PFAS ecological risk assessment
The relationships between bioaccumulation of PFAS, food chain transfer and ecological effects are still not well understood.
The build-up of PFAS in the body of animals over time is termed bioaccumulation. Long-chain PFAS can build up because PFAS is poorly metabolised in many animals and is only slowly eliminated from the body. Long-chain PFAS can be found in the blood, organs and muscle tissue but tend not to adhere to fat, unlike other bioaccumulative chemicals, such as dioxins and DDT.
The exposure of organisms higher up the food chain to long-chain PFAS can be greater than for those at lower levels of the food chain. Increasing concentrations of chemicals up the food chain is termed biomagnification. For PFAS, biomagnification has been observed in aquatic and terrestrial mammals and birds.
PFAS can also undergo maternal transfer during foetal growth and/or lactation. This means that PFAS exposure may be increased for later generations.
Quantifying these effects using data from natural systems is uncertain and challenging, given the high degree of variability in natural systems. As such, a precautionary approach is adopted in environmental site assessments, discussed further below.
Why does the level of species protection need to be increased to account for bioaccumulation in aquatic environments?
Chemicals that bioaccumulate and biomagnify can exert effects on organisms via several means, including:
- effects from direct exposure to higher concentrations over shorter time scales
- effects from build-up in the body due to direct exposure over longer time scales
- effects on predatory species due to indirect exposure through the food chain.
Most of the published aquatic ecotoxicological research on chemicals (including bioaccumulative chemicals) is limited, and focuses on single-species data that only account for the direct effects of exposure under controlled conditions. Usually, the studies involve only one generation of animals, with fewer second generation or multi-generation studies.
The Australian water-quality guidelines for aquatic ecosystems (ANZECC and ARMCANZ, 2000) state that there is currently no formal and specific international guidance for quantifying the potential effects of bioaccumulation in water-quality guidelines. For these reasons, the guidelines recommend an increased level of species protection for bioaccumulative chemicals, where local data are lacking on bioaccumulation effects, or where it is considered that the standard protection level may not protect all potentially exposed species.
Increasing the level of species protection results in more conservative water-quality guidelines. This is a qualitative approach to addressing the potential for bioaccumulation, rather than being based on a defined relationship between the level of protection and the level of effect of bioaccumulative chemicals. Having conservative water-quality guidelines also means that more aquatic species are likely to be protected, which may lessen the potential impact of exposure on an aquatic ecosystem-wide basis. Finally, adopting conservative water-quality guidelines means that the potential for the chemical to bioaccumulate at high concentrations is reduced.
The Heads of EPA PFAS National Environmental Management Plan (HEPA PFAS NEMP, 2018) provides aquatic ecosystem water-quality guidelines for PFOS and PFOA for various levels of species protection. The PFOS and PFOA guidelines are not based on the bioaccumulative effects of these chemicals, or on effects through the food chain. Therefore, the guidelines that would normally apply at a site for a given level of species protection may not protect against the effects of bioaccumulation.
For sites that are “slightly-to-moderately disturbed”, water-quality guidelines that correspond to 95 percent species protection would normally apply for non-bioaccumulative chemicals. However, for PFOS and PFOA at these same sites, the 99 percent species protection level would apply, to allow for potential bioaccumulative effects. This includes bioaccumulation within aquatic organisms as well as within animals higher up the food chain. The water-quality guideline for PFOS that corresponds to the 99 percent species protection level is currently lower than the typically achievable analytical limit of reporting (LOR).
Why is the 99 percent aquatic ecosystems guideline for PFOS in fresh water so conservative?
In Australia, the current approach for deriving water-quality guidelines protecting aquatic ecosystems uses statistical methods to interpolate and extrapolate the results of individual ecotoxicological studies to estimate ecosystem-wide effect thresholds in the form of a Species Sensitivity Distribution (SSD).
In relation to the SSD for PFOS:
- The current dataset of ecotoxicological studies consists of concentrations from PFOS studies that were associated with no observed effect on organisms, and span over five orders of magnitude.
- A statistical approach is used to extrapolate the available data from published studies in order to estimate a concentration in water that would theoretically protect 99 percent of all aquatic species.
- Because there is a wide spread of data for PFOS, this introduces uncertainty in determining the theoretical highest (effects on all organisms) and lowest (effects on no organisms) points of the SSD. As a result, the statistical approach estimates a very low 99 percent species protection concentration — much lower than the lowest measured toxicity endpoint from published studies.
It is important to note that there are still uncertainties in our understanding of the mode of action and potential ecological effects of PFOS. Future revisions to include new research data and refined statistical approaches may result in changes to the current criteria.
I have results exceeding the 99 percent screening criteria based on the increased level of protection when bioaccumulation is accounted for — what do I do next?
It is important to note that the water-quality guideline values are screening values that adopt a conservative approach to managing uncertainty. They do not necessarily represent a threshold for unacceptable risk, or mean that a management action is definitively required.
Field investigations of PFAS levels in appropriate organisms can provide additional evidence for whether or not bioaccumulation is an issue at the site under study (ANZECC and ARMCANZ, 2000). The HEPA PFAS NEMP, 2018 recommends aquatic biota sampling where there is a pathway for PFAS migration from a source to an aquatic receptor, regardless of whether PFAS is detected in water at levels above the LOR.
In AECOM’s experience, biota sampling is most useful for ecological risk assessment when measured concentrations of PFAS in water are between the current PFAS guidelines for 99 percent and 95 percent protection. This is because higher PFAS concentrations in water are typically associated with concentrations in aquatic biota that also exceed wildlife dietary guidelines in the HEPA PFAS NEMP, 2018. In these instances, biota sampling confirms what might otherwise have been inferred from water sampling of PFAS alone.
Where PFAS concentrations in biota are above wildlife dietary guidelines, food-chain modelling can be used to better understand potential ecological risks on a more site-specific basis, and to identify the pathways of exposure that drive risk.
AECOM has completed food-chain modelling on a number of sites with the benefit of identifying key areas of concern, and data gaps for further assessment and/or management. AECOM understands where food-chain modelling improves the understanding of key exposure pathways and risks to ecology associated with a site. Importantly, AECOM also understands where soil, sediment and water data will be sufficient to the understanding and prediction of risks at a site, so that the next steps may be progressed to management and/or remediation. AECOM is developing an ecological risk assessment decision tree to assist with identifying the best approach to assessing ecological risks of PFAS on contaminated sites.