Monitoring program for pharmaceuticals, illegal substances, and contaminants in farmed fish

1.1 - Background

According to EU legislation (EU 2017/625, replacing Directive 96/23/EC), all food producing animals should be monitored for certain substances and residues thereof. The following residues or substance groups are monitored in Norwegian farmed fish:

Group A Substances with anabolic effects and unauthorized substances:

A1: Stilbenes, derivatives and their salts and esters

A3: Steroids

A6: Prohibited substances

Group B Veterinary drugs and contaminants:

B1: Antibacterial agents

B2a: Anthelmintics

B2c: Carbamates and pyrethroids

B2d: Sedatives

B3a: Organochlorine compounds

B3b: Organophosphorus compounds

B3c: Chemical elements

B3d: Mycotoxins

B3e: Dyes

B3f: Others

1.2 - Group A, Substances with anabolic effects and unauthorized substances

Fish tested for illegal compounds were collected at the farm by official inspectors from the Norwegian Food Safety Authority (NSFA), without prior notification to the farmers. Samples were taken at all stages of farming in order to represent farmed fish during production. Substances monitored in Group A include growth promoters like steroids and stilbenes, and unauthorized drugs. Unauthorized drugs considered most relevant for aquaculture are chloramphenicol, nitrofurans, metronidazole and dyes. Since the use of the dyes malachite green, crystal violet and brilliant green is not allowed for food producing species (EU 37/2010), they are considered Group A substances and hence monitored in samples throughout the production chain. However, according to Regulation (EU) 2017/625, these dyes belong to the group B3e. Thus, in order to fulfill criteria for group B sampling, some of the samples assigned to analysis of dyes were also collected at the slaughterhouse.

To ensure harmonized levels for the control of unauthorized substances, the analytical methods should meet a minimum required performance limits (MRPLs) set by the European Union (EU 2003/1881, EU 2004/25, CRL 2007), and European reference laboratories (EU-RLs), (EU 2003/1881, EU 2004/25, CRL 2007). Table 1 gives an overview of MRPLs of relevant compounds.

1.3 - Group B, veterinary drugs

In order to protect public health, current EU legislation (EU 37/2010) provisions the assignment of Maximum Residue Limits (MRLs) for all legally applied pharmacologically active substances in products intended for human consumption. An MRL denotes the highest permitted residual concentration of a legally applied veterinary drug and is evaluated for each substance and each food product individually. Consumption of food with drug residues below the MRL should not pose a health risk to the consumer. For fish, the MRLs are set for muscle and skin in natural proportions. Samples examined for veterinary drugs were collected from fish at processing plants and the samples are representative of fish ready to be placed on the market for human consumption.

1.4 - Group B, contaminants

Samples examined for contaminants were collected from fish at processing plants and are representative of fish ready for human consumption. The EU (EU 1881/2006) has set a Maximum limit (ML) for some of the contaminants in fish, while for others, such as pesticides, polycyclic aromatic hydrocarbons (PAH), perfluorocarbons (PFC) and brominated flame retardants (BFR), maximum limits have not been established.

2.1 - Sampling

Samples were taken on fish farms or slaughterhouses, by official inspectors from the NFSA, in all fish-producing regions in Norway. The sampling plan was randomised according to season and region. In 2021, the monitoring program included Atlantic salmon (Salmo salar), rainbow trout (Oncorhynchus mykiss), brown trout (Salmo trutta), turbot (Scophthalmus maximus), Atlantic halibut (Hippoglossus hippoglossus), Arctic char (Salvelinus alpinus), Atlantic cod (Gadus morhua) and spotted wolffish (Anarhichas minor) .

Samples were transported to the Institute of Marine Research (IMR) in a frozen state. For most analyses, the Norwegian quality cut (NQC) was used (Johnsen, Hagen et al. 2011). However, both NQC and individual liver samples were collected for analysis of antibiotics. Samples to be used for analyses of substances with anabolic effects or unauthorized substances also included small fish from early life stages, and in these cases, the whole fish except head, tail and gut were homogenised. The samples were analysed as pooled samples comprising five fish from the same cage/farm.

2.2 - Pre-treatment

Upon arrival at IMR the sample identification was anonymised for the analysts. A back-up sample was stored for all samples. Pooled samples of muscle from five fish from the same cage/farm were homogenised before analyses. Samples of liver were excised from the fish to be screened for residues of antimicrobial agents by the microbiological inhibition zone assay. Liver samples were examined individually, if residues were detected, the back-up sample of muscle was analysed by chemical methods. The maximum residue limits for veterinary drugs are set for muscle and skin in natural proportions (EU 37/2010). Therefore, according to the analytical protocol, any detection of drug residues in the muscle or liver was followed by a re-analysis of the back-up sample, consisting of muscle and skin in natural proportions, in duplicate.

2.3 - Analytical methods

The laboratory routines and most of the analytical methods are accredited in accordance with the standard ISO 17025 (Table 1). A summary of the analytical methods and their limit of detection (LOD) or limit of quantification (LOQ) is shown in Table 1. The LOD is the lowest level at which the method is able to detect the substance, while the LOQ is the lowest level for a reliable quantitative measurement. For all methods, a sample blank and a quality control sample (QC) with a known composition and concentration of target analyte are included in each series. The methods are regularly verified by participation in inter laboratory proficiency tests, or by analysing certified reference material (CRM), where such exist. 

3.1 - Substances with anabolic effects and unauthorized substances

In 2021, a total of 1085 pooled fillet samples (5 fish/sample) were tested for residues of illegal substances, including stilbenes (164 pooled samples; Table 2), steroids (166 pooled samples; Table 3), and unauthorized veterinary drugs (754 pooled samples; Table 4). The samples were mainly taken from Atlantic salmon, but also samples from rainbow trout, Atlantic cod, brown trout and Atlantic halibut were analysed. With regards to illegal substances, the samples are monitored for and evaluated towards presence. No residues of stilbenes, steroids, chloramphenicol, nitrofurans or metronidazole were detected in any of the samples.

Dyes were measured in 167 pooled samples from fish taken at aquaculture farms. Crystal violet was detected in two of the samples (0.32 and 0.50 ng crystal violet/g ww, respectively). The analysis results were reported to the NFSA, who executed follow-up investigation of these samples. Crystal violet was not detected in other samples coming from the same batch. The absence of leucocrystal violet, the metabolite of crystal violet, indicated a contamination during sampling rather than a direct application of the dye during farming.

In addition to 168 samples analysed for residues of dyes as group A (illegal drugs) samples, dyes were measured in 88 pooled group B samples (contaminants; Group B3e, dyes). No residues of malachite green, crystal violet and brilliant green were detected in any of the group B samples.

3.2 - Veterinary drugs

Samples analysed for veterinary drugs were collected from fish at processing plants, representing fish ready for human consumption. The maximum residue limit (MRL) for veterinary drugs is defined for muscle and skin in natural proportions (EU 37/2010). Therefore, according to the analytical protocol, any detection of drug residues in a sample of muscle or liver would be followed by a re-analysis of the backup sample, consisting of muscle and skin in natural proportions, in duplicate.

Antibacterial agents were monitored through a combination of a three-plate bioassay and chemical methods. The broad groups a) quinolones, b) amphenicols and tetracyclines and c) sulphonamides were screened in livers of 1665 fish (Table 5). A total of 119 pooled fillet samples, representing 595 fish were analysed by chemical methods (Table 6). No residues were detected in any of the samples analysed.

3.2.1 - Group B2a, Anthelmintics

The residues of anthelmintics, such as anti-sea-lice agents (Table 7) and agents for treatment of endoparasites (Table 8) were monitored in a total of 388 pooled fillet samples, representing 1940 fish.

Residues of the anti-sealice agent emamectin was detected in one out of 114 analysed samples, at a concentration of 10 µg/kg. This concentration is below the MRL of 100 µg/kg (EU 37/2010). Also residues of lufenuron were found in one out of 110 analysed samples. The concentration measured was 58 µg lufenuron/kg and was below the MRL of 1350 µg/kg. Residues of other agents in this group were not detected in any of the samples. LOQs for the substances are specified in Table 1.

¹ MRL established for Salmonidae only (muscle and skin in natural proportions).

3.2.2 - Group B2c, Carbamates and pyrethroids

In addition to measurement of cypermethrin and deltamethrin with a method using GC-MS, in 2021, several carbamates and pyrethroid substances, including cypermethrin and deltamethrin were measured using GC-MS/MS (Table 9).

A total of 120 pooled fillet samples of Atlantic salmon, rainbow trout and turbot were analysed for cypermethrin and deltamethrin using GC-MS. All samples were below the LOQ of 5 µg/kg for cypermethrin and 10 µg/kg for deltamethrin.

With an LOQ of 0.99-1 µg/kg for all of the measured substances on the GC-MS/MS, cypermethrin was detected in 24 out of 92 pooled fillet samples of Atlantic salmon, and one of 6 samples of rainbow trout. Cypermethrin is a pyrethroid insecticide, used both as anti-sealice agent in aquaculture, but also in commercial agricultural applications. Therefore, residues of cypermethrin in fish may also originate from transfer via plant-based ingredients in fish feed. The maximum level measured in salmon was 3.0 µg/kg, and 1.2 µg/kg in rainbow trout, which is below the MRL of 50 µg/kg (EU 37/2010). None of the other substances included in the monitoring, were detected in any of the samples.

¹ MRL established for Salmonidae only (muscle and skin in natural proportions).      

3.2.3 - Group B2d, Sedatives

No residues of isoeugenol or eugenol were found in any of the 39 samples analysed for these sedatives (Table 10).

3.3 - Contaminants

Samples analysed for contaminants were collected from fish at processing plants and are representative of fish ready for human consumption.

3.3.1 - Group B3a, Organochlorine compounds

The levels of organochlorine compounds were determined in 186 pooled samples. The results are summarised in Tables 11 to 13.

3.3.1.1 - Organochlorine pesticides

For several of the pesticides, the amount present is calculated as a sum including metabolites or transformation products (EU GD SANTE 2017). The results for these groups of pesticides are presented in Table 11. To calculate the sum of the components, conversion factors (Table A1, Appendix) are used to adjust for different molecular weights (EU GD SANTE 2017). The sums in Table 11 were calculated according to the upper bound (UB) formula. When using UB calculations, the numerical value of LOQ is substituted for analytes with levels below LOQ. UB represents a “worst case scenario”. As an example, all measurements of endosulfan are below LOQ, however, a sum is generated based on the LOQ-values. There are currently no MRLs established in fish fillet for any of the listed pesticides (EU 2014).

¹ DDT (sum of p,p-DDT, o,p-DDT, p,p-DDD, o,p-DDD, p,p-DDE,and o,p-DDE expressed as DDT). ² Legal residue definition according to Reg. (EC) No 149/2008: DDT (sum of p,p´-DDT, o,p´-DDT, p-p´-DDE and p,p´-TDE (DDD) expressed as DDT). ³ Chlordane (sum of cis- and trans-isomers and oxychlordane expressed as chlordane). ⁴ Legal residue definition according to Reg. (EC) No 149/2008: Chlordane (sum of cis- and trans-chlordane).

The levels of pesticides calculated from a sum of several components were comparable to the previous years. The highest values of organochlorine pesticides in Atlantic salmon fillet were 9.8 µg/kg w.w. of DDT, and 5 µg/kg w.w. Toxaphene. The highest values for organochlorine pesticides in rainbow trout were DDT and endosulfan, with 5 and 2.2 µg/kg w.w, respectively.

The results for the other organochlorine pesticides are summarised in Table 12.

3.3.1.2 - Dioxin, dl-PCBs and PCB-6

The levels of dioxins (PCDD+PCDF), dl-PCBs and PCB-6 in farmed fish are shown in Table 13. The data is mainly represented by Atlantic salmon, but also samples from rainbow trout were examined. The sums of dioxins, dioxins + dl-PCBs and PCB-6 are calculated as upper bound (EU 1259/2011). Accordingly, the numerical LOQ values were used for congeners with levels below LOQ.

The levels of dioxins and dl-PCBs are reported as ng toxic equivalents 2005 (TEQ05)/kg and represent the sum of 17 different dioxins and 12 dl-PCBs where each congener was multiplied by a Toxic Equivalency Factor (TEF). TEF values are determined by the World Health Organization (WHO), and the toxicity of each congener is expressed relative to the most toxic form of dioxin, which has a TEF value of 1 (EU 1259/2011, Van den Berg, Birnbaum et al. 2006).

In 2021, dioxin levels found in fish fillet were somewhat lower than in the previous year. For salmon, the median of the sum of dioxins was 0.16 ng TEQ/kg w.w. The maximum value found in salmon (0.45 ng TEQ/kg w.w.) was below the EU maximum level of 3.5 ng TEQ/kg w.w.

The median of the sum of all 29 dioxins and dl-PCBs was 0.36 ng TEQ/kg w.w for salmon (0.1 ng TEQ/kg w.w. lower than in 2020) and 0.37 ng TEQ/kg w.w for rainbow trout (0.41 ng TEQ/kg w.w in 2020). The highest result for sum dioxin and dl-like PCBs was 0.90 ng TEQ/kg w.w., measured in salmon. All measured values were below the EU maximum level of 6.5 ng TEQ/kg w.w.

Compared to 2020, in 2021 decreased fillet levels were also noted for PCB-6 measurements. The median of PCB-6 for salmon was 2.9 μg/kg w.w and 2.7 in rainbow trout, with maximum concentrations of 7.7 and 4.1 μg/kg w.w, respectively. For PCB-6, a maximum level is set at 75 μg/kg w.w. in the EU.

3.3.2 - Group B3b, Organophosphorous compounds

Organophosphorous pesticide residues, chlorpyriphos, chlorpyriphos-methyl and pirimiphos-methyl, were determined in 99 pooled fillet samples (Table 14). No residues of chlorpyriphos or chlorpyriphos-methyl were detected in any of the samples. Pirimiphos-methyl was detected in 8 of 92 samples of Atlantic salmon. The maximum concentration was 0.39 µg pirimiphos-methyl/kg w.w.. There is currently no MRL established for pirimiphos-methyl in fish fillet (EU 2014).

3.3.3 - Group B3c, Chemical elements

In 2021, monitoring of the levels of chemical elements, such as arsenic (and inorganic arsenic), total mercury in addition to methylmercury, cadmium, lead included 73 samples of Atlantic salmon, 7 samples of rainbow trout, and one sample of Atlantic cod (Table 15). Mono-, di- and tributyltin were analysed in 52 samples of Atlantic salmon, 7 samples of rainbow trout and one sample of Atlantic cod.

The concentrations of total mercury were found below the EU maximum level, which is set at 0.50 mg/kg w.w. for these species. The highest concentrations of total mercury were 0.057 mg/kg w.w. in salmon, and 0.081 mg/kg w.w. in Atlantic cod, mainly present as methylmercury (Table 15).

Cadmium in all Atlantic salmon samples, Atlantic cod and 6 of 7 rainbow trout samples was below the LOQ. Only one sample of rainbow trout had cadmium at a level of 0.002 mg/kg w.w. in the fillet which is well below the EUs maximum level of 0.05 mg/kg w.w. (EU 1881/2006).

Arsenic is determined as “total arsenic”, comprising the sum of all arsenic species. In addition, inorganic arsenic was determined in 20 of the samples. The median level of total arsenic in Atlantic salmon was 0.58 mg/kg w.w., and, same as in the previous year, the highest concentration measured was 1.4 mg/kg w.w. (Table 15). The median and maximum concentration of Arsenic in rainbow trout samples were 0.8 and 1 mg/kg w.w respectively. The concentrations of inorganic arsenic were below the LOQ in all samples measured (Table 15), indicating that arsenic in fish is present mainly as organo-arsenic compounds of low toxicity (Shiomi 1994). There is currently no EU upper limit for neither total arsenic nor inorganic arsenic in fish fillets.

In all 81 samples of Atlantic salmon, rainbow trout and Atlantic cod, the concentration of lead was below LOQ and well below the EU maximum level, which is currently set at 0.30 mg/kg w.w. in muscle meat of fish (EU 1881/2006).

Eleven additional chemical elements were analyzed in addition to the above-mentioned elements. There is currently no EU-limit established for any of these elements. Copper, iron, manganese, selenium and zinc were found at levels above LOQ in all samples analyzed (Table 15), with median values similar to the year before. The maximum concentrations among all 81 samples were 0.64 mg copper/kg, 4.5 mg iron/kg, 0.13 mg manganese/kg, 0.28 mg selenium/kg and 5.1 mg zinc/kg, respectively. Cobalt and nickel were not detected in any of the analysed samples. Chromium and vanadium were detected in 22 and 17 out of 81 samples, respectively. The highest concentrations were 0.035 mg chromium/kg and 0.006 mg vanadium/kg (both salmon) in 2021.

Mono-, di- and tributyltin were monitored in a total of 60 pooled fillet samples of Atlantic salmon, rainbow trout and Atlantic cod. There is currently no EU upper limit for tin in fish fillet. Monobutyltin was found at levels above LOQ in 7 samples, with the maximum concentrations of 1 µg/kg w.w. and 0.7 µg/kg w.w. in salmon and rainbow trout, respectively. Concentration of dibutyltin was bellow LOQ (0.2 µg/kg w.w.) in all samples. A total of 16 samples contained tributyltin above the LOQ, with the highest measured level of 1.8 µg/kg w.w. found in rainbow trout (median 0.2 µg/kg w.w.).

Calculated as upper bound, using the numerical LOQ values for measurements below LOQ. Median for elements with more than 50% of values below LOQ is not presented.

3.3.4 - Group B3d, Mycotoxins

The mycotoxins enniatin A, enniatin A1, enniatin B, enniatin B1 and beauvericin were measured in 95 pooled samples of Atlantic salmon and four 4 pooled samples of rainbow trout. No residues of these mycotoxins were detected in any of the samples (Table 16).

3.3.5 - Group B3f, others

The group B3f, others is a group not required for finfish products by Regulation (EU) 2017/625, but are deemed relevant for analyses in Norwegian aquaculture by the NFSA and IMR, because these undesirable compounds are present in the environment and may affect food safety. This group currently consists of brominated flame retardants (BFR), perfluorinated compounds (PFC), polyaromatic hydrocarbons (PAHs), and since 2018 also the technological feed additive ethoxyquin (EQ) and its main transformation product ethoxyquin dimer (EQDM).

3.3.5.1 - Brominated flame retardants

In addition to the PBDEs included in PBDE-7 (PBDE 28, 47, 99, 100, 153, 154, 183), 10 other tri-hepta PBDEs (PBDE 35, 49, 66, 71, 75, 77, 85, 118, 119, 138) and five octa-deca PBDEs (PBDE 196, 197, 206, 207 and 209) were measured in 97 pooled fillet samples from Atlantic salmon and 3 pooled fillet samples of rainbow trout (Table 17) . Median values of PBDE-7 were 0.27 μ g/kg w.w. and 0.28 μ g/kg w.w for salmon and rainbow trout, respectively.

*For PBDE 196, 197, 206, 207 and 209, the number of samples analysed were 93.

TBBPA was not found in any of the 55 pooled fillet samples from Atlantic salmon samples, three pooled rainbow trout samples and one Atlantic cod sample (Table 18). Alpha-HBCD was found in 41 of the 55 salmon fillet samples and two of three rainbow trout fillet samples, while beta- and gamma- HBCD were only found in salmon samples (Table 18). The median HBCD concentration (Sum alpha-, beta-, gamma-HBCD; UB) for salmon was 0.055 μ g/kg w.w., with a maximum concentration of 0.30 μ g/kg w.w.. There is currently no EU maximum limit for BFRs in food.

3.3.5.2 - Perfluorinated compounds

The results for the analysis of perflourinated compounds are presented in Table 19.

3.3.5.3 - Polycyclic aromatic hydrocarbons

The results for PAH are summarised in Table 20. There is no maximum limit for PAH in fresh fish (EU 835/2011).