Levels of contaminants in fish from the Barents Sea are in general relatively low and below EU and Norway’s maximum levels set for food safety. For most substances, concentrations are stable or slightly decreasing.
IMR conducts regular monitoring of chemical contaminants in biota through two different programs. 1) A three-year monitoring program designed to monitor the level of pollution in the Barents Sea, and 2) An annual monitoring program with focus on seafood safety and pollution level in indicator species. In program 1, levels of certain organic contaminants (PCB, chlorinated pesticides and PBDEs) are analysed mainly in liver of fish. The exact species sampled varies from year to year, but some species have been sampled repeatedly in three year cycles and temporal data exist for Greenland halibut, long rough dab, haddock, capelin, polar cod, saithe (Pollachius virens), herring, cod and golden redfish. The sampling programme is designed to monitor pollution levels over time. Samples are mainly taken on the ecosystem cruise in summer/early fall. In program 2, levels of metals including As, Cd, Hg and Pb are analysed in fillet and liver of Atlantic cod, whole capelin and polar cod as well as whole and peeled boiled northern shrimp (Pandalus borealis). Levels of organic pollutants (POPs) are analysed in liver of cod, whole capelin and polar cod and whole boiled shrimp, and a few samples of cod fillet have also been analysed. The POPs include dioxins and dioxin-like PCBs, non-dioxinlike PCBs (PCB6, PCB7), organochlorine pesticides, brominated flame retardants (PBDEs, HBCD and TBBP-A), per- and polyfluoralkyl substances (PFAS) and PAHs. The monitoring program is designed to document levels of contaminants with regards to food safety, while also gaining information on pollution levels by analysing indicator organisms representing varying trophic levels and niches. Samples are mainly taken on the winter cruise in January-March. In addition to these regular monitoring programs, samples of several species have been taken in the Barents Sea and analysed. Some are sampled and analysed on a regular basis and as a part of special surveys. Contaminants in saithe and Greenland halibut are monitored annually. Species where we collected data on contaminants for special surveys include redfish species (Sebastes norvegicus and S. mentella, (Nilsen et al. 2020), tusk (Brosme brosme), haddock and wolffish (Atlantic, spotted and jelly; Anarhichas spp.) and snow crab (Frantzen and Maage 2016) as well as red king crab from coastal areas (Julshamn et al. 2015). In both program 1 and 2, where temporal data exist, samples are not taken at fixed positions or at fixed fish size, so temporal trends must be interpreted with caution. The positions sampled during 2006-2019 for program 2 are shown in figures 18.104.22.168-22.214.171.124.
Analyses are performed with accredited analytical methods according to ISO 17025.
Levels of contaminants with focus on food safety
In general, levels of contaminants in fillet of fish from the Barents Sea are very low and below EU and Norway’s maximum levels for food safety for substances where these exist (Hg, Cd, Pb, sum dioxins and dioxin-like PCBs and sum PCB6) (Figure 126.96.36.199). Mercury is the contaminant that is most often of concern for food safety, especially with regard to muscle of lean fish. The levels of mercury in muscle of fish from the Barents Sea are generally lower than in other (Norwegian) sea areas. The highest concentrations are found in Greenland halibut, tusk and Atlantic wolffish, and the lowest concentrations are found in some of the most commercially important species such as cod, saithe and haddock (Figure 188.8.131.52). Beaked redfish had somewhat higher level of mercury than golden redfish. The varying levels of mercury are probably at least partly due to the different species’ trophic level, as methylmercury is a typically biomagnifying contaminant. Other factors that can lead to between-species variation in mercury levels are for instance age, growth rate and geographical area. Arsenic levels in some of the fish species were relatively high (Figure 184.108.40.206). The arsenic present in fish muscle is in general arsenobetain, which has very low toxicity. The most toxic species of arsenic is inorganic arsenic (EFSA 2009). A large number of fish samples from Norwegian sea areas were previously analysed for total and inorganic arsenic, and even those with very high total arsenic concentrations had very low levels of inorganic arsenic (Julshamn et al. 2012). In Norway and EU, there is no maximum level for arsenic, while Russia has a maximum level of 5 mg/kg, which some of the fish species exceeded (Figure 220.127.116.11). Differences between species in arsenic level may at least in part be related to their diet, where a more benthic diet seems to lead to higher arsenic levels than a predominantly pelagic diet (Neff 1997).
Figure 18.104.22.168 Concentrations of A) Arsenic (As, mg/kg wet weight) and B) mercury (Hg, mg/kg wet weight) in muscle of cod (2018-2019, N=98), saithe (2018-2019, N = 75), haddock (2014, N=40), tusk (2014, N=160), Greenland halibut (2018, N= 52), golden redfish (2018, N = 50), beaked redfish (2018, N = 249), Atlantic wolffish (2014, N=29), spotted wolffish (2014, N=27) and jelly wolffish (2014, N=12). Mean, minimum and maximum values are shown. Red lines indicate maximum allowable levels set for food safety.
The levels in muscle of the lipid soluble persistent organic pollutants dioxins and dioxin-like PCBs and sum PCB6 were higher in Greenland halibut, redfishes and wolffishes than in other fish species (Figure 22.214.171.124). The different levels of these substances in fillet are probably related to the different fat contents of the different fish species, since Greenland halibut is a fatty fish (fat content ca. 10g/100g), redfish and wolffish species are semi-fatty, while cod, saithe and haddock are typically lean fish species with fat contents lower than 1 g/100 g. Lean fish species primarily store their fat and fat soluble organic contaminants in the liver, where the concentrations of these substances may get very high as can be seen in figure 126.96.36.199. The EU and Norway have established special maximum levels applying to dioxins and dioxin-like PCBs and sum PCB6 in liver, which are set much higher than for fillet. In some areas, these higher maximum levels are exceeded. This is normally not a great food safety issue, as fish liver is not generally consumed in large amounts. However, to protect the most vulnerable parts of the population (i.e. foetuses and young children) against dioxin and dioxin-like PCB toxicity, the Norwegian Food Safety Authority has issued a warning for children and pregnant and breastfeeding women, against eating fish liver (https://www.matportalen.no/matvaregrupper/tema/fisk_og_skalldyr/barn_gravide_og_ammende_bor_ikke_spise_rognleverpostei).
Figure 188.8.131.52 Concentrations of A) Sum of dioxins and dioxin-like (dl-) PCBs (ng TEQ/kg weight) and B) sum PCB6 sum PCB7 (µg/kg wet weight) in muscle of cod (2018-2019, N=20), saithe (2019, N=5), haddock (2014, N=7*), tusk (2014+2016; N=10*), Greenland halibut (2018-2019, N= 74), golden redfish (2018, N=49), beaked redfish (2018, N = 347), Atlantic wolffish (2014, N=6*), spotted wolffish (2014, N=4*) and jelly wolffish (2014, N=2*). Mean, minimum and maximum values are shown. Red lines indicate maximum allowable levels set for food safety in EU and Norway. For non-dioxinlike PCBs, maximum level in EU and Norway applies to the sum PCB6. * Composite samples analysed.
Concentrations of some heavy metals in muscle tissue of the crustaceans red king crab, snow crab and shrimp, are shown in Figure 184.108.40.206. Levels of mercury and cadmium were very low and well below maximum levels set for food safety in EU and Norway (Figure 220.127.116.11A, B). The level of cadmium was considerably higher in shrimp than in both crab species. Cadmium is a heavy metal that accumulates in the hepatopancreas of crustaceans, and shrimp analysed whole have relatively high cadmium concentrations which in many cases exceed the maximum level. However, since shrimps are usually peeled before being consumed this is not considered a food safety issue. Pure muscle tissue in general has very low cadmium levels. Still, cadmium levels in shrimp from the Barents Sea are higher than the levels in shrimp sampled in Norwegian sea areas further south. This corresponds well with findings from other studies of increasing cadmium levels in crustaceans from south to north (Wiech et al. 2020; Zauke et al. 1996, Zauke and Schmalenbach 2006). It likely has natural causes. However, a good explanation has so far not been found. Levels of arsenic in the snow crab and shrimp were very high (Figure 18.104.22.168C). Again, this is most likely arsenobetain, which is known as a non-toxic substance. Analysis of inorganic arsenic in red king crab has shown that inorganic arsenic contributes a very small part (<0.4%) of the total arsenic concentrations (Julshamn et al. 2015). The levels of dioxins and PCBs were very low in the crustacean muscle tissue (Figure 22.214.171.124), which is also very lean with mean fat contents in the area of 0.5-2.2 g/100 g (Seafood data). The concentrations of both sum dioxins and dioxin-like PCBs and sum PCB6 were far below the maximum levels set for food safety in EU and Norway.
Figure 126.96.36.199 Concentrations of A) mercury (Hg, mg/kg wet weight), B) cadmium (Cd, mg/kg wet weight) and C) arsenic (As, mg/kg wet weight) in muscle of red king crab (Paralithodes camtchaticus, 2012), snow crab (Chionoecetes opilio, 2014 + 2016) and shrimp (Pandalus borealis, boiled, 2015-2019). Mean, minimum and maximum values are shown. Red lines indicate maximum allowable levels set for food safety. *Composite samples.
Figure 188.8.131.52 Concentrations of A) sum of dioxins and dioxin-like (dl-) PCBs (ng TEQ/kg weight) and B) sum PCB6 (µg/kg wet weight) in composite muscle samples of red king crab (2012, N = 29), snow crab (2014, N = 9*) and shrimp (2015-2019, N=15* cooked and peeled). Mean, minimum and maximum values are shown. *Composite samples analysed.
Levels of contaminants with focus on pollution level and environmental effects
Figures 184.108.40.206 and 220.127.116.11 show concentrations of selected lipid soluble organic contaminants (Dioxins+dioxin-like PCBs, PCB7) in liver of different fish species from the Barents Sea. Tusk liver had among the highest levels of both dioxins and dioxin-like PCBs and PCB7. Haddock had the highest mean concentration of dioxins and dioxin-like PCBs, while PCB7 was not very high. The results shown for haddock for dioxins and dioxinlike PCBs and for PCB7 are from different projects, and the differences could be due to differences in sampling area. The species where we have the most data from other areas for comparison, is cod. The mean levels of dioxins and dioxinlike PCBs and PCB7 in liver of cod from the Barents Sea were much lower than in liver of cod from the North Sea, which in 2016 were 20.6 ng TEQ/kg and 162 µg/kg, respectively (miljostatus.no). It has also been shown that the levels of contaminants in tusk and saithe from the Barents Sea in general are lower than in the same species from the Norwegian Sea and the North Sea (Frantzen and Maage 2016; Nilsen et al. 2013a, Nilsen et al. 2013b). It must be kept in mind that the different species in Figure 18.104.22.168 and Figure 22.214.171.124 were taken in different years (see figure legends), and this may also affect the observed differences between the species. Herring liver was only analysed for PCB7 in program 1, and herring had the lowest mean concentration of PCB7 of all the different species, followed by Greenland halibut. Since herring and Greenland halibut are fatty fish species, they store much of their lipids and lipid soluble pollutants in their fillet, explaining the low concentrations in their liver. In contrast, the species with the highest concentrations of these substances in liver are lean fish species, which store all surplus lipids in their liver. Brominated flame retardants (PBDEs) show almost same pattern of between-species variation as dioxins and dioxin-like PCBs and non-dioxin like PCBs (Figure 126.96.36.199). This is probably because the same factors control the between-species variations for the lipid soluble brominated compounds as the dioxins and PCBs. Also for PBDEs, concentrations in the Barents Sea are lower than in the North Sea for species where we have comparable data from the North Sea. Concentrations of chlorinated pesticides such as DDTs and HCB in fish liver vary between species in much the same way, and results are not shown here.
Figure 188.8.131.52 Concentration of sum of dioxins and dioxin-like (dl-) PCBs (ng TEQ/kg weight) in liver of cod (2018-2019, N=92), saithe (2018-2019, N=75), haddock (2014, N=8*), tusk (2014; N=9*), golden redfish (2018, N=2*), beaked redfish (2018, N = 14*), Atlantic wolffish (2014, N=6*), spotted wolffish (2014, N=4*) and jelly wolffish (2014, N=2*). Mean, minimum and maximum values are shown. Red lines indicate maximum allowable levels set for food safety in EU and Norway. For non-dioxinlike PCBs, maximum level in EU and Norway applies to the sum PCB6. * Composite samples analysed.
Figure 184.108.40.206 Concentrations of the sum PCB7 (µg/kg wet weight) in liver of cod (2018-2019, N=92), saithe (2018-2019, N=75), haddock (2015, N = 49), tusk (2014; N=9*), Greenland halibut (2015, N = 50), long rough dab (2015, N = 43), golden redfish (2018, N=2*), beaked redfish (2018, N = 14*), Atlantic wolffish (2014, N=6*), spotted wolffish (2014, N=4*), jelly wolffish (2014, N=2*), herring (2015, N = 10*), blue whiting (2015, N = 10*) and polar cod (2015, N = 5*). Mean, minimum and maximum values are shown. Red line indicates maximum level applying to fish liver in EU and Norway. * Composite samples analysed.
Figure 220.127.116.11 Left: Concentrations in 2015 of sum PBDE in liver of cod (2015, N=50), saithe (2015, N=25), haddock (2015, N = 49), Greenland halibut (2015, N = 50), long rough dab (2015, N = 26), beaked redfish (2015, N = 43), herring (2015, N = 10*), blue whiting (2015, N = 10*) and polar cod (2015, N = 5*). Right: Concentrations of PBDE 47 (µg/kg wet weight) in liver of cod (2018-2019, N=92), saithe (2018-2019, N=75), haddock (2014, N=8*), tusk (2014; N=9*), golden redfish (2018, N=2*), beaked redfish (2018, N = 14*), Atlantic wolffish (2014, N=6*), spotted wolffish (2014, N=4*) and jelly wolffish (2014, N=2*). Mean, minimum and maximum values are shown. * Composite samples analysed.
Levels of contaminants in forage fish and shrimp analysed whole are particularly interesting with regard to transfer of contaminants to higher trophic levels and ultimately to the top predators. Through the water framework directive, the EU has given a set of environmental quality standards (EQS) for harmful substances in the environment, including biota (Directive 2008/105/EC). In addition, Norway has established additional regional EQS values for substances not included by the EU (Miljødirektoratet 2016). The EQS are meant to protect the most sensitive organisms in the ecosystem. This means that levels of contaminants exceeding the EQS may potentially have a harmful effect on animals at the highest trophic levels and animals feeding exclusively on fish. Therefore, the EQS have in general been set much lower than the maximum levels for food safety, where these exist. Levels of selected compounds (PCB7, DDT, HCB, PBDE) in small organisms analysed whole, i.e. shrimp, capelin, polar cod and sandeel, are given in the graphs below (Figure 18.104.22.168). While mean levels of DDT, PCB7 and PBDE were higher in capelin than in polar cod, the mean level of HCB was higher in polar cod than in capelin. Sandeel had lower concentrations of both DDT, PCB and HCB than both capelin and polar cod. Shrimp (whole, boiled) had a mean PCB7 concentration similar to capelin, while levels of DDT and HCB were lower in shrimp than both capelin, polar cod and sandeel. Concentrations of mercury, HCB and DDT in capelin and polar cod from the Barents Sea are below the EQS values (Figure 22.214.171.124; Figure 126.96.36.199). However, the levels of PCB7 and PBDE were mostly above the EQS. The concentrations of these substances are, however, very low in these species from the Barents Sea. Since capelin and polar cod are mainly found in the Barents Sea, there is no data to compare with from other Norwegian sea areas. For sandeel, there is data from the North Sea, where concentrations of Hg, DDT and PBDEs in 2016 were much higher than in the Barents Sea (miljostatus.no). On the other hand, the level of HCB was much higher in sandeel from the Barents Sea than in sandeel from the North Sea. Concentrations of contaminants are found in fish and crustaceans at significant and measurable levels. The levels of the analysed substances are, however, relatively low in the Barents Sea compared to areas further south, except for the levels of cadmium and HCB, which seem to be higher in organisms from the Barents Sea. However, PCBs and PBDEs are above EQS values, which may indicate potentially harmful effects for animals at high trophic level such as for instance polar bears. It is not known, however, whether the levels of these substances in top predators and fish eaters such as marine mammals and seabirds in the Barents Sea actually have harmful effects on individual or population levels.
Figure 188.8.131.52 Left: Concentrations of HCB, DDT (sum of p,p’-DDD, p,p’-DDE and p,p’-DDT) and PCB7 in composite samples of whole boiled shrimp, capelin, sandeel and polar cod. Right: Concentration of PBDE 47 (the most dominating PBDE congener) in composite samples of whole capelin, polar cod, boiled and raw shrimp. Yellow lines indicate EQS values for PCB7 (left panel) and PBDE (right panel).
Temporal trends of contaminant levels
In order to evaluate time trends for levels of different contaminants in the Barents Sea biota, data from cod, capelin and polar cod have been included (Figure 184.108.40.206-220.127.116.11). Mercury levels appear to be very stable in the Barents Sea, based on concentrations measured in cod fillet as well as in capelin and polar cod (Figure 18.104.22.168, Figure 22.214.171.124). The lower limit of quantification (LOQ) for mercury prior to 2010 was higher than later on, which is why it may look like a decrease for polar cod after 2009.
Levels of dioxins and dioxin-like PCBs and PCB6 in cod liver analysed in program 2 seem to have decreased slightly since 2007 (Figure 126.96.36.199). Because sampling area and fish size may affect the results, results for samples taken south and north of 73°N have been shown separately, and only fish between 50 and 70 cm length were included.
Figure 188.8.131.52 Annual concentrations of sum of dioxins and dioxin-like PCBs in liver of cod from 2007 to 2019, shown separately for the areas south of 73°N (73 N; right). Only individuals between 50 and 70 cm are included. For each year, mean ± 95% confidence interval is shown.
The levels of PCB6/PCB7 in capelin and polar cod also seem to have decreased somewhat since 2006, and more so in polar cod than in capelin (Figure 184.108.40.206). It cannot be completely ruled out that a more northern distribution resulting in a more northern sampling in later years, may have resulted in a more decreasing trend than if sampling had been done in the same areas year after year.
For DDT, there has been a clear decrease in cod liver since 1992, while HCB levels have remained very stable (Figure 220.127.116.11). For polar cod, the level of HCB seems to be almost increasing in later years (Figure 18.104.22.168). Levels of PBDE show clearly decreasing trends since 2006/2007, both in cod liver, capelin and polar cod (Figure 22.214.171.124 and 126.96.36.199). There are thus indications that the levels of persistent organic contaminants such as dioxins, PCBs and DDTs have been decreasing and are still slowly decreasing in the Barents Sea. This pattern seems to be the most evident for PBDEs, which were banned around 2005, while HCB seems not to decrease at all. The latter may be because HCB can be breakdown product from other pesticides still in use. The level of mercury in cod fillet has remained very stable since 1994.
Figure 188.8.131.52 Annual concentrations of sum DDT and hexachlorobenzene (HCB) in composite samples of whole capelin and polar cod from 2006/2007 to 2019. Results from program 1 (P-1) and program 2 (P-2) are given separately. For each year, mean, minimum and maximum values are shown.
Figure 184.108.40.206 Temporal trend of concentrations of sum of 6 PBDEs in liver of cod from 2007 to 2019, shown separately for areas south of 73°N (73 N). Only individuals between 50 and 70 cm are included. For each year, mean ± 95% confidence interval is shown.
Data accumulated over the past 10-15 years by the Joint Norwegian-Russian monitoring program has provided a reliable overview of the levels and trends of radioactive contamination in the Barents Sea marine environment. The data confirms that levels of the anthropogenic radionuclides Cs-137, Sr-90 and Pu-239,240 in seawater, sediments, fish and seaweed are currently low and generally decreasing. In recent decades, there has been a slow decrease in the levels of most anthropogenic radionuclides in the Barents Sea. Working document (WD2) present full report of the methodology and monitoring results.
The monitoring data reported by Norway and Russia has been demonstrated to be robust and comparable. Further effort to ensure data comparability is an important part of the ongoing work on future cooperation. To support this aim, a series of bilateral workshops on sampling, analysis and quality control have been launched (starting in 2016) to better understand and harmonise methodologies employed on each side of the border.