select search filters
roundups & rapid reactions
factsheets & briefing notes
before the headlines
Fiona fox's blog

expert reaction to study looking at toxin accumulation in fish, and overfishing and climate change

Research, published in Nature, reports that ocean warming and changes in the diets of some fish, due to overfishing, may result in build ups of  the neurotoxin methylmercury in some fish consumed by humans.


Dr Elizabeth Lund, Independent Consultant in Nutrition and Gastrointestinal Health, said:

“Although methyl mercury in fish is a potential cause for concern current evidence points toward the benefits of fish consumption, including in pregnancy, outweighs any risk.  However pregnant and nursing women should heed current guidelines to avoid certain species of fish high up the food chain such as swordfish.  The main point of this paper seems to be to say the current regulations should not be relaxed or levels may go up.”


Dr Oliver Jones, Associate Professor of Chemistry, RMIT University in Melbourne, Australia, said:

“The main idea of this paper is that the main source of mercury for most people is though eating fish, such as cod and tuna.  The mercury in the fish gets there, at least in part, from pollution of the water in which the fish live.  Now, you might think that if the amount of mercury in the water goes down, then the mercury levels in the fish, and hence humans, would also go down.  According to this new study however, you would be wrong.  The authors show that fish get mercury from both the water they live in, and the food they eat.  So if cod eat things with a high amount of mercury in, this can pass through to humans.  Taking this further – if factors like climate change and overfishing cause the cod and tuna etc. to eat things with more mercury in them than they otherwise would (which seems to be the case), then humans eating the cod would get a correspondingly higher dose of mercury than might be expected, even if the level of the mercury in the surrounding water has gone down.  It is clever work.

“There are some limitations to the study in that it reuses published (not new) data for the Gulf of Maine, so the conclusions may not hold true for other areas.  A lot of data in the models (which are themselves very complex) are estimated or extrapolated so there is likely to be some error in the final result.  Nevertheless the work is clever and detailed, and shows the importance of fully characterising natural systems before you attempt to understand them”.


Prof Sean Strain, Emeritus Professor of Human Nutrition, University of Ulster, said:

“This paper looks like it is good quality, and the suggestion based on modelling that concentrations of methylmercury may have increased in some fish is probable.  But these are predictions from a model, not actual measures.  The modelling and calculations appear to be solid, based on good quality science, and would support the suggestion by the authors that these modelled potential increases in methylmercury in cod and other fish species would be owing to overfishing and global warming.

“The implication, however, that a modelled 23% increase in methylmercury in the tissue of Atlantic cod between the 1970s and 2000s represents an increased threat to human health from the consumption of such fish is contestable.  There is no doubt that methylmercury is a potent neurotoxin to which the developing foetus is particularly susceptible.  This potency was highlighted by the tragedies of the Minamata poisoning in Japan in the 1950s caused by the dumping of methymercury into Minamata Bay by a large petrochemical company and by the grain poisoning in Iraq in the early 1970s caused by the consumption of wheat and barley imports treated with methylmercury as a fungicide.  But the suggestion that has been previously made that concentrations of methylmercury naturally found to varying extent in all fish and seafood is associated with long-term neurocognitive deficits in children and persist into adulthood, is much more contentious.”


Dr Emeir McSorley, Senior Lecturer in Human Nutrition, Ulster University, said:

“For background, the Seychelles Child Development Study, with which we are involved, was originally set up following the Iraq poisoning to evaluate the possible public health implications of methylmercury exposure from high fish consumption during pregnancy.  Mothers in the Seychelles are exposed to methylmercury concentrations at least 10 to 100 times greater than fish consuming mothers in Western countries and yet we have consistently found no adverse associations of methylmercury with neurodevelopment in three mother-child cohorts, the first of which has been followed up to adulthood.  Indeed, the children born to mothers with the highest methylmercury exposures were undertaking some developmental tests better than those born to mothers exposed to lower methylmercury.  We interpreted these findings as indicating that the benefits of fish consumption during pregnancy outweighed any risks.

“One other feature of the mercury cycle which is not obvious from the press release or paper is that mercury in the atmosphere, and the resulting microbial metabolite, methylmercury, in aqueous food-chains, arises from natural sources, mainly volcanic activity, as well as from human activities, mainly fossil fuel burning and incinerators.  Methylmercury, therefore, has always been a component of freshwater and marine environments and will continue to be so irrespective of human activities.  While it is certainly sensible to follow the existing global treaty of the Minamata Convention to lower this toxin in the environment, it is less sensible to infer that potential increases in methylmercury in relatively low methylmercury fish (cod) through overfishing and global warming might make such fish a risk for human consumption, especially given that some 20% of the world’s population rely on fish as their primary source of protein.”


Prof Awadhesh Jha, Professor of Ecotoxicology, University of Plymouth, said:

“This study offers interesting insights into the effects of climate change on the bioaccumulation of relevant contaminants in wild fish species.  The modelling approaches move the science forward, offering a way to evaluate the hazard and risk posed not only by methyl mercury but also by other pollutants present in the environment.  And its recognition of the importance of species’ behaviour, and predictions of the historical and future release of contaminants, are also approaches which could be mirrored within future research of this nature.

“It is known that overfishing in our oceans and seas contributes imbalances in the distribution of natural populations of economically important species.  What is important about this study is that it provides valuable information that such imbalances also have a bearing on the bioaccumulation of hazardous contaminants in a species specific manner which could impact human and environmental health via the food chain.

“The study is mainly based on modelling approaches and has used data accumulated over the last three decades pertaining to MeHg concentrations in different compartments of the ecosystem (i.e. water, sediment and the biota at different trophic levels).  It then links this to the impact of overfishing on the bioaccumulation pattern of an important contaminant.  The predictions based on such modelling approaches could have inherent limitations for real world scenarios, given that contaminants do occur as complex mixtures in the environment and could interact differently based on their bioavailability and tissue specific uptake.  However, this could be verified experimentally in years to come.

“The potential impacts of rising temperatures and other environmental factors – such as hypoxia, salinity, acidification and pollution – are of increasing global importance, both in isolation and combined.  It is essential we understand any long-term health impacts, on the biota and ultimately humans, with our reliance on aquaculture to feed a growing global population set to expand hugely in the future.”


Dr Clive Trueman, Associate Professor in Marine Ecology, University of Southampton, said:

“This paper is an interesting simulation study which is very nicely described.  However the press release could lead to misinterpretation – it could be inferred that methyl mercury concentration levels in cod and tuna HAVE increased by 23% or 56% respectively since the 1970s, when in fact these are POTENTIAL increases simulated in a model, not measured values.

“This is an interesting study: mercury is a harmful toxin which accumulates in food webs so that the higher up a food chain an animal is, the greater the concentration of mercury in its tissues.  Marine food chains are longer than those on land, and humans like to eat top predators in marine food chains (like large tuna and swordfish), potentially increasing our exposure to mercury.  Methyl mercury is a particularly toxic form of mercury and global guidelines suggesting limits on the amounts of predatory marine fish that one should consume reflect the potential exposure to methyl mercury.

“Fish obtain mercury through their diets.  Mercury is often stored in fats or lipids – so if fish change their diets towards more lipid rich, fatty prey, this could increase the amount of mercury they consume.  Changes to the amount of food that fish have to eat in order to grow could also influence the amount of mercury, including methyl mercury in their tissues.  Fishing can alter food webs, influencing the diets available to top predators, and ocean warming could lead to increased energy demands in cold blooded fishes (meaning that they need to eat more just to stay alive).  So it is reasonable to suggest that combined effects of fishing and climate change could lead to increases in (methyl) mercury in fish tissues – even if the total amount of mercury in the oceans is reducing because of stricter controls on mercury pollution.

“The study by Schtarup et al is a largely theoretical, modelling study that explores how changes in physiology and diets might influence mercury loadings.  It is a well-designed model that takes sensible conceptual ideas and uses available measurements to build simulations.  The models suggest that realistic changes in temperature and food web structures could lead to increases in mercury concentrations in fish tissues even though the amount of mercury in the oceans has decreased.

“Ecosystems are complex, and we have a rather limited understanding of how animals respond to change – models are always simplifications of reality, making assumptions that scientists hope are reasonable.  This study makes quite a lot of assumptions about how animals are likely to respond to change.  While the assumptions are clearly described in the paper, the results are model simulations – not real measurements.  They describe what might happen IF the assumptions and simplifications made are reasonable.  The outputs from the models are compared to measurements of mercury in tuna and other fishes – but there are not very many direct measurements of methyl mercury in fish tissues, and very few measurements over the timescales considered in the study.

“On balance I think this is an interesting conceptual study.  One of the nicest parts of this study is that it mixes different types of science.  Many scientists look at toxicology (that is the amount of dangerous chemicals in animal tissues).  Other ecologists study how food webs and metabolism are influenced by things like fishing and climate change.  Relatively few studies combine the two and to understand toxicology by understanding how the ecology of the animals might change with human impacts.  This is a really nice aspect of the study, and it shows how important it is to consider the whole ecosystem (including humans) when we think about the varied consequences of human actions.

“I am less enthusiastic about the press release.  The press release does not make it clear that the study is essentially reporting results from a model – something that is very clear from the paper itself.  It would be easy to read the press release and think that the authors had actually measured increasing concentrations of methyl mercury in fish tissues.  The press release quotes figures of 23% or 56% increases in methyl mercury in cod or tuna tissues, respectively.  It is very important to note that these figures are NOT measured (i.e. real) values.  They are outputs of models suggesting that possible changes to fish diets and physiology inferred to happen because of climate and fishing could possibly increase mercury concentrations IF the models are reliable.  That is a long way from saying that methyl mercury levels HAVE increased by 23 or 56% respectively.”


Prof Andy Smith, Senior Scientist, MRC Toxicology Unit, said:

“With global regulation anthropogenic sources of mercury have declined over recent decades leading to lower or plateauing levels of methyl mercury (MeHg) in the marine environment.  MeHg in top fish predators such as cod and bluefin tuna accumulated from fish lower in the food chain, is considered a major source for human consumption and has previously been suggested to be associated with possible neurocognitive deficits in human development.

“Based on computer modelling using data from various sources of fish and seawater collected from the Northwest Atlantic over 30 years, Schartup and colleagues show that assumptions of lower MeHg levels in such fish eaten by humans may be too simple.  The authors argue that depletion of other fish stocks by intense fishing, such as herring, change the feeding behaviour of cod and tuna with associated increased intake of MeHg.  Changes in sea water temperature perhaps linked to climate change also impact on the availability of many species that are food for cod and bluefin tuna, potentially leading to increased levels of MeHg in their tissue.

“The conclusions are based on modelling with extrapolations of incomplete data sets and restricted to the gulf of Maine.  Although it is not possible to see immediate implications for human consumers, it underlines the importance of continuing pressure to decrease global exposure to environmental mercury.”


‘Climate change and overfishing increase neurotoxicant in marine predators’ by Amina T. Schartup et al. was published in Nature at 18:00 UK time on Wednesday 7 August 2019. 

DOI: 10.1038/s41586-019-1468-9


Declared interests

Dr Elizabeth Lund: “As a research leader at The Institute of Food Research (now Quadram) I ran an EU funded research project on fish consumption and gastrointestinal health as part of the SEAFOODplus project between 2004-2009.  The salmon and cod used in this study was donated by industry.”

Dr Oliver Jones: “I have no conflicts of interest to declare.”

Prof Sean Strain and Dr Emeir McSorley: “Membership of the Seychelles Child Development Study team; we have been funded by the US National Institutes of Health and the European Commission and our partners include University of Rochester, NY and the Department of Health, Republic of Seychelles.”

Prof Awadhesh Jha: “I have no conflict of interest.”

Dr Clive Trueman: “I have no competing interests.”

Prof Andy Smith: “No conflicts.”

in this section

filter RoundUps by year

search by tag