Back to Reports




Monday, March 1, 2004


SUBJECT: carcinogenic organochlorine contaminants  " including dioxins, PCBs, toxaphene, dieldrin - in fishmeal, and the risk to human health

CURRENT ISSUE: On 9th January 2004, UK media widely reported that a major comparison of levels of 14 pollutants in farmed and wild salmon had discovered much higher levels in the farmed salmon. The most polluted salmon was said to come from UK and the Faroe Islands. The feeding of fishmeal and fish oil were blamed for the higher concentrations in farmed salmon.

 “Global Assessment of Organic Contaminants in Farmed Salmon,” by R.A. Hites at Indiana U. in Bloomington, IN; J.A. Foran at U. of Michigan in Ann Arbor, MI; D.O. Carpenter at U. at Albany in Rensselaer, NY; M.C. Hamilton at AXYS Analytical Services in Sidney, BC, Canada; B.A. Knuth and S.J. Schwager at Cornell U. in Ithaca, NY. Published in the journal Science on 9th January 2004:

and there was a press release (apparently issued on behalf of its authors)
“Study Reveals Health Risks of European Farm-Raised Salmon " Study in the journal Science suggests sharp restrictions in consumption” (Institute of Health and the Environment, University at Albany, State Univ.of New York: press release embargoed until 7pm GMT January 8th 2004).

The study said:
 700 samples of farmed and wild salmon studied looking for contaminants including dioxins, total PCBs, dieldrin and toxaphene.
 Concentrations in farmed salmon from Europe and North American farmed salmon were higher than in wild salmon.
 Concentrations of PCBs, dioxin and some other contaminants highest in farmed salmon from Scotland and Faroe Islands and lowest in that from Chile and Washington State.
 Contamination most likely a function of diet.
 Concentration in feed purchased from Europe was significantly higher than that in feed purchased from North and South America.
 Individual contaminant concentrations in farmed and wild salmon DO NOT exceed US FDA action or tolerance levels for PCB and dieldrin…..but using the EPA approach…(lots more here) consumption of farmed Atlantic salmon may pose risks that detract from the beneficial effects of fish consumption

 Example from the Times of 9.1.2004- “The most likely explanation for the high levels of pollutants in farmed salmon is that the feed that they are generally given consists of a mixture of a high fat mixture of other fish, ground into fish meal and fish oil. As organochlorines build up in the fatty tissue of the fish, they become concentrated in this high fat food, and are passed on to the farmed salmon.”
 From the Mail " “These chemicals come from the feed used on fish farms”.
Fishmeal and fish oil sector response


1. The Study does NOT present new safety concerns. The Study assesses risks against an assessment process that is not recognised by international organisations responsible for food safety and public health. The health benefits of fish continue to far outweigh risk from trace contaminants.

2. Fish, fishmeal and fish oil are not the primary source of this contamination.  Efforts to reduce contamination at source are succeeding in substantially reducing human exposure. There is evidence which contradicts the Study’s key assertion that farmed salmon is more contaminated that wild fish produced in the same area.

3. Fishmeal and oil traded in the EU complies with tough maximum permitted levels of dioxins. In addition to complying with regulations the fishmeal and oil industry is investing heavily in processes which minimise levels of all such contaminants in their products.

4. Response to comments on Industrial fishing and feed conversion efficiency

5. Criticisms of the Study itself " methodology etc

EU limits on trace contaminants

1. The Study does NOT present new safety concerns. The health benefits of fish continue to far outweigh risk from trace contaminants.
(FROM FSA Q & A Jan 9 2004)

 “This study does not present any new safety concerns. The levels of dioxins found in farmed salmon are below the safety levels set by the World Health Organisation (WHO), the European Union (EU) and the United States Food and Drug Administration (FDA) 

 “This study has based its conclusions on a risk assessment process that has been under consideration in the United States since 1991 by the Environmental Protection Agency (EPA). This process is not recognised by international organisations responsible for food safety and public health that consider it scientifically flawed. The EPA is concerned with environment protection and sets levels for fish caught recreationally in the United States.

 “The World Health Organisation set safety levels for dioxins and PCBs in 2001 based exclusively on public health protection. These form the basis of safety levels set for consumers who eat fish sold in shops.

 “These levels of dioxins have previously been reported by the FSA. What is new is the comparison between different fish caught in different oceans. Since dioxins and PCBs are associated with industrial discharges into the sea, it is not surprising that fish from oceans remote from such areas have lower levels of these chemicals.

 “The known benefits of eating oily fish outweigh any possible risks. There is good evidence that eating oily fish reduces the risk of death from recurrent heart attacks and that there is a similar effect in relation to first heart attacks. Our advice is that people should consume at least two portions of fish a week " one of which should be oily.”


 If the EPA process for assessing risks from contaminants were applied to all foods (as the Study does with farmed salmon) " the result would be that most animal products and human breast milk would be considered to pose similar serious health risks.

2. Fish, fishmeal and fish oil are not the primary source of this contamination. Efforts to reduce such contamination at source are succeeding in substantially reducing human exposure. There is evidence which contradicts the Study’s assertion that farmed salmon is more contaminated that wild fish produced in the same area.

 The best way to reduce dietary exposure to dioxins and PCBs is to reduce emissions to the environment,” - Food Standards Agency statement, January 3, 2001.

 Fishmeal and fish oil are NOT the source of this group of contaminants. Fish are victims of this contamination, carried from sources on land through waterways, to the sea. Fisheries and fishmeal interests support efforts to reduce contamination at source.

 Dioxins are often formed as unwanted by-products in a variety of industrial processes and some domestic ones like household fires. They are found throughout the environment and fish accumulate them from what they eat including other fish and fish feed. PCBs have been used since the early 1930's mainly in electrical equipment.

 There are strict controls in place on the release of dioxins. For example, the manufacture and general use of PCBs stopped in the 1970's and is no longer permitted in the UK.

 Such efforts are succeeding dramatically. For example, the latest data released by the UK Food Standards Agency (FSA from its 2001 Total Diet Survey) showed that the average intake of dioxins and dioxin-like PCBs by adults in the UK halved between 1997 and 2001 and the percentage of adults exceeding the new UK safety limits from their diets fell from 35% to 1.1% in the same period. The percentage of children as a whole who exceeded the safety limit fell from 62% in 1997 to 10% in 2001. Strict controls on industrial pollutants came into effect in 1992, which has resulted in a 70% reduction in the amount of dioxins and PCBs released into the environment over the past ten years. The concentrations of dioxins found in individual food groups in the 2001 study were all below EU regulatory limits.

 One of the major salmon feed companies has carried out private studies which show no difference between levels of trace contaminants in farmed and wild Atlantic salmon. (This data may be made available shortly.) In addition UK industry monitoring of dioxins and PCBs in fish used to produce fishmeal show similar levels in farmed salmon and wild fish. In addition monitoring has shown a significant decrease in contamination over the last two years (again the data could be made available).

3. Fishmeal and oil traded in the EU complies with tough maximum permitted levels of dioxins. In addition to complying with regulations the fishmeal and oil industry is investing in processes which minimise levels of all such contaminants in their products.

 To protect human health the EU introduced Maximum (Permitted) Levels of dioxins in fishmeal and fish oil, as well as in all other foods and feeds in 2002. These limits are in line with the (public health) safety levels set in 2001 by the World Health Organisation.

 Fishmeal and oil supplied in the EU complies with these limits. Similar EU maximum permitted levels are planned or proposed for further trace contaminants including PCBs. There are also EU Action Levels which trigger a warning to the authorities. The purpose of the maximum and action levels, together with action to reduce pollution at source, is to bring exposure of the whole population to safe levels. (See Appendix 1 for details of legislation).

 In order to further protect consumers, the fishmeal and oil sector has already introduced commercial techniques to remove (typically 90% of ) the dioxins in fish oil, and is developing techniques to remove them from fishmeal.

 Increased quality control and selection by producers could provide fish oil preparations with extremely low pollution levels for use as infant/toddler supplements. This would immediately reduce any perceived risks to these more susceptible groups even further.

4. Resonse to comments on ‘Industrial’ fishing and the efficiency of food conversion

The media has taken the opportunity to repeat other criticisms of the fishmeal and oil industry " notably that fishing for feed fish is damaging fish stocks and that feed fishing to produce fishmeal and oil feeds for aquaculture are an inefficient way to produce protein for the human diet. The suggestion is that global fish resources are being squandered at a time when a large proportion of the world population has an inadequate diet.

These points have been rebutted or answered on many occasions by the fishmeal industry. To summarise key points here:

 World aquaculture produces nearly two tonnes of seafood for every tonne of wild fish it uses in feed.

 Fishmeal and oil produced from caught wild fish are produced mainly from small, bony pelagic species " for which there is little demand for human consumption.

 In the EU about 40% of fishmeal is produced, NOT from wild fish caught for this purpose, but from trimmings from food fish processing. In six out of seven producing Member Countries, the majority of fishmeal is produced from trimmings. For this proportion, this means that, in fish stock conservation terms, no wild fish are caught. In addition use of trimmings recycles food fish waste saving the financial and environmental costs of otherwise disposing of it.

 The majority of wild fish stocks used to produce fishmeal and fish oil are within safe biological limits. Stocks are monitored according to scientific advice and subject to national and international management regimes which are recognised and supported by the fishmeal industry.

 It is be easy to make an initial assumption that it must take more than a tonne of fishmeal to produce a tonne of farmed fish " as the process cannot be 100% efficient. In fact, farmed fish diets are by no means 100% fishmeal and oil. For example salmon feed is currently typically 30% fishmeal and 28% oil. Worldwide for all the main farmed species the proportion of fishmeal in feed can be as low as 8% and as high as 50%, while for fish oil the low is 1% and the high is 28%.  Farmed fish diets also include, for example vegetable meals and flours, and binders. In addition many fish are very efficient converters of feed into weight gain. In fact for every tonne of weight gained by salmon they consume only 1.2 tonnes of feed. This in turn includes 0.36 tonnes of fishmeal and 0.34 tonnes of fish oil. To produce a tonne of farmed salmon therefore requires typically 0.36 tonnes of fishmeal and 0.33 tonnes of fish oil " or 0.69 tonnes in all.

 Compared with other realistic ways in which small bony pelagic fish and fish trimmings could be or are turned into food, the feed for aquaculture route is very efficient in terms of feed conversion. Where the fishmeal and oil are produced from food fish trimmings, waste is being converted into useable feed.

 The amount of fishmeal and oil used to produce a tonne of salmon, or other fish, is falling continuously. First, the total amount of feed required to grow salmon has fallen by 44% since 1972, according to feed manufacturer EWOS, due to technical progress. Second, significant technical progress is being made in substituting fishmeal and fish oil with other ingredients " mainly of plant origin. Fish have no requirement for any particular feed ingredient (e.g. fishmeal or fish oil), but they do have a requirement for dietary energy and a number of essential nutrients (e.g. 10 amino acids and certain fatty acids, minerals and vitamins) " which can be supplied by a range of ingredients in combination. Estimates of the potential for substitution vary, but most are focused around the view that the technology is in sight to substitute 50% of the meal and 75% of the oil in farmed fish feed.

5. Criticisms of the Study

 That the Study appears to have been funded by or influenced by NGOs with a strong anti salmon farming agenda - e.g. that author Foran is cited as from Citizens for a Better Environment, Milwaukee, and there is funding from the Pew Charitable Trust.

 That the study has been used to develop, not for the first time, a major orchestrated hit on the farmed salmon industry.

 That the study compared contaminant levels in farmed Atlantic salmon with wild Alaskan Pacific salmon (which just happens to be the least contaminated) on the grounds of: first, difficulty in obtaining supplies of Atlantic wild salmon and, second, that Alaskan Pacific salmon are not farmed on any great scale. Whatever the reason this meant the study compared farmed salmon with the least contaminated of wild salmon.

 Many or all of the salmon and feed samples may have been collected before the EU legislation of 2002 which imposed maximum permitted levels of dioxins in both fish and fishmeal and fish oil came into force. The proposed and actual introduction of these maximum levels, and proposed or planned limits on other pollutants such as PCBs, have resulted in considerable food chain resources being devoted to minimising contamination with carcinogenic organochlorines.

 Fishmeal interests are also seeking further information on the feed samples used in this study " their source, age etc " as these may not be representative, in terms of contaminant levels of contemporary (post EU maximum level. One source suggested one or more samples could have been four years old.

 The EPA yardstick of risk assessment is not recognized by international authorities (confirmed by FSA).

EU maximum permitted levels

 The Council Directive COM/2001/0493 amending Council Directive 1999/29/EC on undesirable substances and products in animal nutrition (dioxin limits in feed) was adopted on 27 November 2002. (The equivalent proposal on dioxin limits in food was adopted on 29 November). The legislation became effective on 1 July 2002. A new Commission Directive 2003/57/EC was published on 17 June 2003. This came into effect on 1 August 2003. This effectively repeals and replaces Directive 1999/29/EC

 The feed proposal stipulates maximum levels for fish products as:
o Maximum level for fish oil is 6.0 ng/kg product
o Maximum level for fish meal is 1.25 ng/kg product
o Maximum level for compounded fish feed is 2.25 ng/kg product

 The maximum levels for fish:
o Maximum level for fish is 4 pg WHO-PCDD/F-TEQ/g fresh weight
o Maximum level for fish oil is 2 pg WHO-PCDD/F-TEQ/g fat)

To sum up the toxicity of the different dioxins the concept of toxic equivalency factors (TEFs) has been introduced to facilitate risk assessment and regulatory controls. This translates to Tolerable Daily Intakes (TDI) and Tolerable Weekly Intakes (TWI) which relate to the amount of contaminant that can be ingested daily or weekly over a lifetime without appreciable risk

 The EU SCF has established a TWI of 14 picogram toxic equivalent equivalents (TEQ) per kilogram body weight for dioxins and dioxin-like PCBs. This TWI is in line with the provisional Tolerable Monthly Intake (TMI) of 70 pg/kg/bodyweight/month established by the Joint FAO/WHO Expert Committee on Food Additives (JECFA - June 2001). It also concurs with the lower end of the range TDI of 1-4 pg WHO-TEQ/kg body eight, established by the World Health Organisation (WHO) consultation in 1998.

 Like for like this means:
EU SCF (May 2001) Maximum intake 2 pg/kg bodyweight/day
JECFA (June 2001) Maximum intake 2.3 pg/kg bodyweight/day
WHO (May 1998) Range of 1-4 pg/kg bodyweight/day

[Source: IFFO]


Aquaculture Jobs

Check out the latest industry job openings in our Job Board.

Follow Us

We are social. Connect with us on:

Sign Up For Our Publications

Select a newsletter/magazine and submit your e-mail to subscribe.

Aquafeed Publications
HATCHERY Feed & Management Publications