Governing for transboundary environmental justice: a scientific and policy analysis of fish consumption advisory programs in the Upper St Lawrence River

FCAs are guidlines to avoid or limit the consumption of certain species of fish because of contamination. They are generally issued due to concerns about the health impacts of eating fish contaminated with mercury and persistent bioaccumulative toxins (PBTs), such as polychlorinated biphenyls (PCBs), PCB-like dioxins and furans, as well as some insecticides (e.g., dichlorodiphenyltrichloroethane (DDT), Mirex, Chlordane). As Cleary et al. (2021, p. 72) explained, FCAs present a “critical dilemma” for fish eaters, especially pronounced for communities with a strong culture of fish consumption (Duhaime et al. 2004; Lee et al. 2019; Lowitt et al. 2019). On the one hand, fish is a nutritious food source high in omega-3 polyunsaturated fatty acids with benefits for cardiovascular health, brain development, and a reduced risk of some cancers (Cleary et al. 2021). However, contaminants that bioaccumulate in fish tissue can present health risks to certain segments of the population—typically denoted as “sensitive” populations and considered as children <15 years of age and women of child-bearing age (<50 years)—especially vulnerable (Cleary et al. 2021).

FCAs use a traditional risk-based approach that attempts to measure exposure and make recommendations for public health (Cleary et al. 2021). Levels of contaminants that may harm human health are estimated by collecting fish and monitoring their tissue loadings (King et al. 2021). The methods needed to achieve this require substantial investment and presume that seasonal sampling can effectively and reliably catch fish to represent the broad array of species and size ranges needed for this purpose. The understanding of the risks underlying FCAs has also been critiqued by Arquette (2004) for only capturing biological and physical data and overlooking the risks to the social, cultural, and spiritual practices closely tied to fish consumption in many Indigenous communities. Similarly, Dawson et al. (2008) point out that FCAs often ignore the social and cultural contexts in which risk is understood among fish eaters (Dawson et al. 2008). These critiques have led to some efforts to develop more holistic approaches to risk assessment and decision-making (Arquette et al. 2002; Arquette 2004; Aven and Kristensen 2005).

The first FCAs for the Upper St Lawrence River were issued in the early 1970s. They were meant to be temporary until contamination sources could be eliminated, although various FCAs have remained in place ever since. In 1985, the St Lawrence River at Cornwall, Ontario, was designated an AOC¹ under the Great Lakes Water Quality Agreement due to high levels of discharges over many decades from industrial facilities, including auto manufacturers and pulp and paper companies, located along the river in Cornwall, Ontario and Massena, New York (Neff et al. 2013). This AOC includes the jurisdictions of the Mohawk community of Akwesasne (comprised of the Mohawk Council of Akwesasne on the north side of the river and the St Regis Mohawk Tribe on the south), Canada, and the United States (Ritcey et al. 2011). The impacts of environmental contamination on the health and well-being of Akwesasne, located just downstream from these industrial sites, are well documented in the environmental justice literature (see Hoover 2018). Along the St Lawrence River, studies have associated higher concentrations of methylmercury with neurobehavioural deficits among fish eaters (Mergler 2002; Morrissette et al. 2004). Previous environmental health research has also linked development deficits, neurological problems, and disruptions in reproductive parameters to fish consumption throughout the Great Lakes watershed (Johnson et al. 1999).

Recent decades have seen progress in addressing contamination in the AOC and the Upper St Lawrence River. Overall, mercury and PCBs are trending downward, due to government and community efforts to remediate the river and a reduction in industrial discharges (Neff et al. 2013; St Lawrence Action Plan 2016, 2019). However, levels of mercury and PCBs remain higher in certain sections of the river and in older, and generally larger, specimens of fish (St Lawrence Action Plan 2016). Alongside these legacy contaminants, contaminants of emerging concern (CECs) have been detected in the river (Elliott et al. 2017; Baker et al. 2022; Ren et al. 2022). CECs are a group of chemicals, such as pharmaceuticals, personal care products, and per- and polyfluoroalkyl substances (PFAS), that are suspected of causing harm to human or ecological health, especially as they interact in aquatic ecosystems (Baker et al. 2022).

Presently, there are active multi-million dollar recreational fisheries throughout the Upper St Lawrence River, with millions of anglers each year fishing in the larger Great Lakes watershed (Ontario Ministry of Natural Resources and Forestry 2015; Responsive Management National Office 2019). Smallmouth bass (Micropterus dolomieu), walleye (Sander vitreus), northern pike (Esox lucius), and yellow perch (Perca flavescens) are among the most popular angling species along the upper part of the river (New York State Department of Environmental Conservation 2015; Responsive Management National Office 2019). Anglers often keep a portion of the fish they catch, with over 50% of respondents in a recent survey of Ontario anglers reporting keeping and/or consuming fish (Howarth et al. 2021). Studies have also indicated that consumption of recreationally caught fish is especially important for anglers that are less financially secure and among certain ethnic groups (Nordenstam and Darkwa 2010; Lauber et al. 2017; Hunt et al. in press).

In the Mohawk community of Akwesasne, a range of fish, including perch, bullhead, and sturgeon, have traditionally been eaten (Hoover 2013). Fish consumption patterns have significantly changed because of FCAs introduced since the 1970s, with Hoover (2013) describing the emergence of “generational divides” in fish consumption, with those beyond their childbearing years more likely to eat fish. Cook (2003) similarly documented the pronounced impacts of FCAs on Mohawk women, understood as the “first environment” within Mohawk cultural philosophy and whose life-sustaining roles are threatened due to contaminants. Indigenous peoples' cultures reflect the water and landscapes they emerged from and formed over thousands of years of relationship-building. In this biocultural context, the Haudenosaunee connection to creation demonstrates an interwoven knowledge system and kinship network that “inform their values, teachings, knowledge, practices, and identity” (Francis et al. 2023). Therefore, damage to the water and fisheries has cascading effects beyond just direct impacts on human health. The Ohenton Karihwatehkwen (“Words Before All Else”) articulates this relationality in which fish carry the responsibility to offer themselves as sustenance and clean the waters. Fish consumption is about reinforcing a relationship between the community and relatives, which ensures cultural integrity and continuity.

As demonstrated in this overview, the impacts of contamination and FCAs are unevenly distributed, with women of childbearing age, children, anglers who rely on recreationally caught fish for food security, and Indigenous communities disproportionately affected. Therefore, in our following analysis, we pay attention to the environmental justice implications of FCA programs in terms of the principle that everyone is entitled to equal protection of environmental and public health regulations and has the right to participate in decision-making affecting their health and local environment (Mohai et al. 2009, p. 407). Recent revisions to the Canadian Environmental Protection Act recognize Canadians’ rights to a healthy environment (Environment and Climate Change Canada 2022). Environmental justice takes on added significance in Indigenous contexts, as seen in documents such as the United Nations Declaration on the Rights of Indigenous Peoples (UNDRIP), the Missing and Murdered Indigenous Women and Girls (MMIWG) 231 Calls for Justice, and the Truth and Reconciliation Commission's (TRC) 94 Calls to Action. These documents are expressions of Indigenous voices in Canada and globally that call for addressing discrimination and upholding Indigenous rights based on traditional governance systems. Of central importance to this study, these documents underscore shared governance for Indigenous Nations as a right in the transboundary waterway of the Upper St Lawrence River.

As shown in Table 1, various fish species and contaminants are monitored in FCAs for the Upper St Lawrence River. Notable fish species with specific advice across all four, or three out of four jurisdictions, include brown bullhead (BB), smallmouth (SMB) and largemouth bass (LMB), channel catfish (ChC), common carp (CC), northern pike (NP), yellow perch (YP), walleye (WLYE), and white sucker (WS). Advice for the region varies for species and is dependent on the jurisdiction. This ranges from seven or eight fish species with specific advice to over 21 different species covered by the SRMT advisory. For example, species like sturgeon are on the SRMT advisory as connected to their inherent rights and responsibilities (i.e., sturgeon is a species of concern and illegal for non-Indigenous fishers to harvest). Additional nuance surrounding the target species was also identified. For example, Quebec's FCA provides advice for “bass” without species-level advice, while others typically provide separate advice for both largemouth and smallmouth bass.

Thresholds for establishing species-specific size thresholds also vary by jurisdiction. For example, Ontario considers increments of 5 cm within a certain species size range, while Quebec uses two thresholds to delineate which fish are considered one of three small, medium, and large sizes for that species. The New York State and SRMT FCAs use a single size threshold to note either less or greater than an established size. The central difference between New York State and SRMT is that New York State uses a species-specific size for this threshold, while SRMT uses a similar size threshold for many different species of fish. It is of note that the single threshold sizes generally align with the medium- to large-sized fish in Quebec's advice but do not relate to a similar type of advice in the Ontario FCA.

The dominant chemical contaminants triggering FCAs in the Upper St Lawrence River are mercury and PCBs (and related dioxins, furans, and dioxin-like compounds). These priority contaminants are monitored by Ontario, Quebec, and New York State and are also considered in all four of the FCAs, including the SRMT FCA, which is largely based on fish sampled from the surrounding regions in international monitoring programs and with specific coordination between New York State and the SRMT Environment Division (see Skinner et al. 2018). Advice bodies using single-size threshold triggers for FCAs are typically dominated by mercury (Ontario, Quebec) and PCBs (New York State, SRMT) as the result of historical and present-day industry in the area, but also dioxin-like PCBs, dioxins, and furans make up a large proportion of PCB-related priority contaminants in fish that have been identified at levels considered a risk to human health (Pinheiro et al. 2020; Richter and Skinner 2020). Additionally, concentrations of PFAS in the tissues of fish, mainly perfluoro-octane sulfonic acid (PFOS) and perfluorooctanoic acid (PFOA), indicate that these are a rising concern throughout the Great Lakes with respect to their ubiquity and increasing presence in the aquatic environment (Christensen et al. 2019; Valsecchi et al. 2021). As with some PCBs and dioxins, PAHs and CECs, including PFOS and PFOAs, and PBDEs, are inconsistently monitored within FCA jurisdictions and contain discrepancies between the specific analytes and congeners of these compounds (e.g., Mirex; Photomirex) across contaminant monitoring programs. A small number of certain pesticides are also monitored in some incongruency across the four FCAs. For example, Aldrin, Chlordane, DDT, and Heptachlor are predominantly monitored in New York State, while Mirex and Toxaphene are monitored in New York State, Ontario, and Quebec. The SRMT FCA provides a similar approach, but using only historical priority contaminants, mercury and PCBs, and their tissue loads in local fishes, and provides no specific advice for CECs, such as PFAS.

Within the public-facing FCAs, the presentation of this information also varies considerably by jurisdiction. FCAs in Ontario and New York both indicate the specific contaminants of concern in fish within public guidance documents. In Ontario, detailed information on contaminant levels is available in a separate datafile accessible through a web link in the FCA. No information on specific contaminant levels is provided in the New York State FCA, nor is a web link to associated datasets provided. In Quebec, levels of mercury are provided for all fish that are monitored, whether these levels are high enough or not to warrant restrictive meal advice due to health risks. Other contaminants of concern that may be contributing to local restrictions are not indicated in the FCAs in Quebec. These public-facing FCAs also provide varying levels of detail about how contaminant levels are set. For example, the New York State FCA indicates using federal fish marketing standards to set mercury guidelines in FCAs; in Ontario and Quebec, it is not clear from public-facing guidance if the mercury threshold used in FCAs is the same or different than that set federally by Health Canada for mercury in retail fish.

The differences in monitoring that we have documented can be attributed to a lack of coordinated mechanisms for developing and harmonizing FCAs, which leaves it to individual jurisdictions to put in place monitoring frameworks. King et al. (2021) previously pointed out the lack of a comprehensive approach to fish monitoring across Canada and the United States, and these differences are particularly apparent in the transboundary waterway of the Upper St Lawrence River. In Canada, there is no federal legislation that requires fish monitoring. As such, individual provinces, territories, and First Nations can decide if, when, and how often to conduct fish monitoring. In the US, states are required to monitor for mercury in fish in all waters designated for fishing, according to the federal Clean Water Act (King et al. 2021). There is no legislative basis for other contaminants (King et al. 2021). Federally recognized tribes can also set their own rules outside of state laws.

Collaboration on a shared and transparent approach to monitoring contaminant levels and fish species with requirements in the Upper St Lawrence River would be a significant improvement to the present patchwork approach that our research has documented and would help provide more consistent protections for environmental justice across the waterway. With a coordinated mechanism, differences in monitoring across jurisdictions—for example, to recognize local areas of concern, ecologically unique species, or to incorporate fish species that are locally preferred for eating—could then be accommodated. Capacity building with Indigenous Nations to implement their own monitoring frameworks and standards in unison with surrounding settler government agencies will be key to a more just and coordinated approach within the transboundary setting of the Upper St Lawrence River. In this region, a community-driven science initiative (Great River Rapport—see www.riverrapport.ca) was initiated in 2018 in response to concerns regarding the health of the river. Community engagement identified fish contamination as a priority. A collaborative effort among the Mohawk Council of Akwesasne and the River Institute, a local non-governmental agency with government funding support, has since facilitated fish contaminant sampling in 2021 and 2022 for assessment in the traditional territory of Akwesasne. The results of this project will provide much-needed information to address data gaps in the status of fish contaminant loads and provide the opportunity for capacity building within the local community.

FCAs provide consumption advice based on the levels of contaminants found in representative fish sampled according to their location, species, and length. However, we observed some differences in the existing transboundary FCAs that could be in conflict and, at best, confusing for fish eaters (see Table 1). This includes some variation in how the population demographics are considered in the advice provided, the sizes of portions considered “meals”, and the level of detail behind the restrictions in frequency for consuming fish.

Notably, FCAs issued by Ontario, Quebec, and New York State provide guidance for the general population and sensitive populations, including children and women of childbearing age. Quebec provides some additional breakdown of the age of children considered, while the SRMT FCA offers broad advice for the whole family (i.e., men and women of all ages, young, and elderly). The latter is the result of rigorous community engagement with Akwesasronon, reflecting Indigenous kinship networks compared to the nuclear family or the individual eaters to whom FCAs are typically targeted. What remains is that the act of fish consumption at the community level may not reflect the Western linearity seen in the other FCAs, as one portion may not be eaten as a single meal or the portion may not be made up of a single fish or single species. Thus, to effectively estimate a safe number of meal portions that can be eaten, the Akwesasronon would need to know the species, capture location, and size of all fish contributing to a meal and consider each restriction per portion individually to estimate daily allotments of safe fish intake.

There is also variation across the FCAs with respect to the standard size of fish fillet, as a portion or meal, and its consumption frequency as a meal. The recommended meal or portion sizes range from 227 to 230 g for Ontario, Quebec, and New York State and half this amount in the SRMT FCA (114 g), reflecting whole-family advice. While a 3 g variation in portion size may have a minimal increase in risk for occasional fish eaters, it approximates an additional portion per month for those who eat fish daily, considering the SRMT advice.

Broadly, the levels of meal restriction are similar, with restrictions of 0 or “Do Not Eat” advice up to 1, 2, 4, and 8 meals per month provided across all FCAs except Ontario. Further, Quebec, New York State, and SRMT advise up to a maximal allotment of eight meals per month of any species of fish; however, it is not clearly indicated how many fish of a particular size contribute to this maximum, adding to the confusion. In FCAs employing one or two size thresholds, no specific concessions are made for the exact sizes of fish that make up the meal amount considered in the restrictive advice or the number of fish meals consumed per month (e.g., eating two medium-sized fish may range between 30 and 40 cm, while one large fish could be over 65 cm). Ontario advice is unique in that it provides an additional level of detail within the consumption advice that considers 5 cm intervals. This considers contaminant loads for those eating fish daily up to a maximum allotment of 32 meals per month (Do Not Eat (0), 1, 2, 4, 8, 12, 16, and 32 mpm), where consumption advice is functionally unrestricted. However, there are also instances in which unrestricted consumption of specific species is misaligned. For example, for yellow perch from the St. Lawrence River (zone 14) in Ontario, the general population is advised to consume up to eight to 32 meals per month (depending on the size of the fish), and children/women of childbearing age are advised to consume up to four to 16 meals. However, in New York State, across the river at Massena, the general population is advised to eat only up to four meals per month, and children/women of childbearing age are advised “Do Not Eat” due to the concern of PCBs.

Differing guidance can be confusing for the public and fishing communities to understand and can also make it difficult for individuals and communities to engage in decision-making that affects their health. For example, the Mohawk community of Akwesasne, which spans the Canada/US border, has long had to deal with varying FCAs in their traditional waters (Boehme 2020). Recognizing this, in 2020, the International Joint Commission (IJC) initiated a project to develop more consistent and culturally appropriate FCAs in this multi-jurisdictional setting (Boehme 2020). At the time of writing this article, there is no publicly available information about the status of this initiative.

Alongside these differences, we also identified some consistent gaps in consumption guidance. First, we found minimal consideration of cultural food practices across the FCAs. The SRMT guide and its whole family advice are exceptions. Previous research has pointed out that FCAs are weak in this respect, instead providing universal dietary advice (Dellinger et al. 2018, 2019). However, this universal advice overlooks the risks or benefits associated with different eating practices. For example, studies in the Great Lakes watershed have suggested that ethnic groups are more likely to keep the fish they catch for consumption (Nordenstam and Darkwa 2010; Lauber et al. 2017; Hunt et al. in press); however, race and ethnic differences in recreational angling harvesting and consumption practices remain understudied (see Hunt et al. 2007) and are generally not reflected in consumption guidance. Indigenous communities have also traditionally eaten a greater range of parts of the fish that may have a different contaminant load than the flesh. Even within the community-based SRMT FCA, beyond general guidance in the FCAs to trim fat off the flesh of fish, we found no guidance about consuming or preparing different parts of the fish (e.g., livers, gonads, eyes, etc.). Greater engagement with the cultural eating practices of communities may be a step toward the more holistic risk assessment called for by Arquette (2004).

Lastly, we note that the FCAs issued by Ontario, Quebec, and New York State do not consider fishing management regulations, including how many fish are legally allowed to be caught by recreational anglers. This siloed approach to policy could potentially contribute to the advice to eat fish low in contaminants but whose catch is limited. Consumption guidance could be better aligned with harvesting regulations. However, doing so would require working through categories such as meal and portion sizes that are not yet appropriately and consistently defined for all fish eaters, as well as collecting more information on fish that is eaten versus caught and released by recreational anglers.

How FCAs are developed and communicated is crucial to their uptake, awareness, and potential impact on anglers and fishing communities. There have been calls for more outreach to angling populations in the Great Lakes watershed to increase adherence to consumption recommendations (Connelly et al. 2017). For example, over one-third of recreational fishers surveyed in Ontario in 2020 reported no awareness of FCAs (Hunt et al. 2022). King et al. (2021) have noted a need for more public engagement on FCAs so that technical information on fish contamination can be better communicated and blended with individual and community perceptions of risk. The US Environmental Protection Agency provides recommendations for states on how to develop FCAs, including risk communication programs, identifying partners, and developing outreach plans (Environmental Protection Agency 2022). We found no similar guidance in Canada.

Awareness of cultural practices may be an important means of developing effective transboundary FCAs. In the FCAs we analyzed, only the SRMT FCA provides images of the fish contained in their advisories. The use of photos of fish species is an example of implementing culturally-based science and risk communication by changing the presentation style to make it more engaging among Akwesasronon (Medin and Bang 2014). The Ontario, Quebec, and New York State FCAs rely heavily on numerical information presented in tables. Previous studies have suggested that the highly technical information contained in FCAs is often not readily understood among the public (King et al. 2021) and that the use of plain language writing, graphics, and qualitative information can enhance appeal and communication (Connelly and Knuth 1998). As well, engagement with cultural eating practices, as discussed in the previous section, might help in developing more culturally appropriate risk assessment communication. For example, Dellinger et al. (2018) in collaboration with the Chippewa Ottawa Resource Authority on Lake Michigan, developed a fish consumption application that incorporated the nutritional and cultural benefits of eating fish alongside the risk of exposure to PBTs. Their messaging, starting with an understanding of the positive place of fish in the Ojibwe culture, was broadly desired by the community and well received.

Another key consideration for environmental justice is access to information. The FCAs we analyzed are available only in English in Ontario and New York State, and English and French in Quebec. The SRMT FCA is likewise published in English only. This may limit outreach to socio-culturally diverse populations, including ethnic groups that may be more likely to eat recreationally caught fish. FCAs in all jurisdictions are provided online. Print copies can be ordered online in Ontario and New York. A reliance on the internet as a key form of communication raises concerns about internet access and the needed technological literacy to navigate these advisories. This may be particularly important in the context of an aging demographic of recreational anglers across the Great Lakes watershed as well as the still predominantly rural geography of anglers, where high-speed internet options may be more limited (Hunt et al. in press).

Lastly, the development and communication of FCAs in ways that uphold self-determination are of central importance for Indigenous environmental justice. Within the Haudenosaunee biocultural context, fish are kin, and thus rights and public health cannot be separated from their relationality with the fish. When FCAs are not developed by or implemented in partnership with Indigenous communities, they may lead to further social, nutritional, and economic disruptions and cultural loss (Duhaime et al. 2004; McAuley and Knopper 2011; Hoover 2013). Hoover (2013) demonstrated how FCAs, starting in the 1970s, contributed to environmental violence among Akwesasne community members, including disruption of language and knowledge systems and the inability to fulfill community roles and responsibilities tied to fishing and fish consumption.

With respect to the contemporary legal context, Gagnon et al. (2018) argued that FCAs can undermine Treaty rights to fish in traditional territories. In Canada and the US, there is a lack of guidance in legislation or policy about how FCAs interface with constitutionally protected Aboriginal and Treaty rights in Canada and Treaty rights in the United States. In Canada, the Crown has a “Duty to Consult” as part of the recognition and affirmation of Aboriginal and Treaty rights in Section 35(1) of the Constitution Act, 1982, and this should be upheld in any implementation of FCAs in Indigenous territories. Foundationally, Indigenous consent and self-governance, as reinforced in the United Nations Declaration of Rights on Indigenous Peoples (UNDRIP), need to be key principles for transboundary governance in the Upper St Lawrence River, including the development of FCAs.