Pollutants in capelin

Method

The sampling is mostly performed by one of the vessels from the Institute of Marine Research.  

Aggregated samples of minimum 25 individuals or 1-2 kg capelin are analysed from each position. Homogenised samples are freeze dried prior to analysis.

To determine dioxins (PCDD), furans (PCDF), dioxin-like PCB (DL-PCB; non-ortho and mono-ortho PCB), non-dioxin-like PCB (NDL-PCB) and polybrominated diphenyl ether (PBDE), freeze-dried samples were mixed with Hydromatrix and added to the internal standard for PCDD/F, PCB and PBDE. The samples were extracted using hexane with Accelerated solvent Extractor-300 (ASE®). The fat was broken down on-line using sulphuric acid-impregnated silica gel in the cells. The extract was further cleansed chromatographically on columns packed with silver nitrate-silica gel, sulphuric acid-silica gel, carbon and alumina respectively using a GO-HT system. Two fractions were collected.

• Fraction 1 contained mono-ortho PCB, NDL-PCB and PBDE.
• Fraction 2 contained dioxins, furanes and non-ortho PCBs.

PCDD, PCDF and non-ortho PCB are analysed using HRGC/HRMS. Mono-ortho PCB, NDL-PCB and PBDE were analysed using GC-MSMS. All compounds are quantified using the isotope dilution/internal standard method. 

NDL-PCB comprised PCB 28, 52, 101, 138, 153 and 180, and PCB7 is the sum of these, plus PCB 118. Sum 7 PBDE (PBDE7) is the sum of PBDE 28, 47, 99, 100, 153, 154 and 183. For each of the 29 congeners of PCDD/F and DL-PCB, a toxic equivalent value (TE) was calculated by multiplying the concentration by the congener’s toxic equivalent factor (TEF) (2005-TEF, Van den Berg et al., 2006). The upper bound of the sum of the TE values was then calculated, with concentrations below the limit of quantification (LOQ) being set as equal to LOQ.

To determine per- and polyfluorinated alkyl substances (PFAS), the following method was used: The weighted sample quantity was added to the mass-marked internal standard and methanol, and extracted in an ultrasonic bath. Following centrifuging, the supernatant was filtered through a 0.45 μm nylon filter, before cleansing on ASPEC®. The extract from ASPEC was further cleansed by filtering through a 3K ultrafilter. Finally, the samples were analysed using LC-MS/MS and quantified using the internal standard. This method is accredited for various matrices such as fat and lean fish muscle and liver, but not for whole fish.

A single method was used which determines a number of chlorinated pesticides, and the analyses were carried out at Eurofins GfA Lab Service GmbH using their method no. GFB53. The analysis was performed using GC-HRMS based on an accredited method.

Hexabromocyclododecane (α, β, γ-HBCD) and tetrabromobisphenol A (TBBP-A) were analysed by Eurofins GfA Lab Service GmbH using method nos. GFB71-2 and GFB86-2. The determinations were carried out using LC-MS/MS based on a accredited method.

A series of metals/elements (e.g. arsenic (As), cadmium (Cd), mercury (Hg), lead (Pb), copper (Cu), zinc (Zn), selenium (Se)) were determined by weighing the freeze-dried sample, adding nitric acid, and heating in a microwave oven in order to completely break down organic matter. The determination of the elements was carried out using an inductively coupled mass spectrometer (ICPMS), and element concentrations were calculated using an external standard curve. Rhodium, germanium, indium or thulium were used as internal standards, and gold was added to stabilise the mercury ions.

To determine polyaromatic hydrocarbons (PAH), the sample material was mixed with Hydromatrix and added to the internal standard, before being extracted using dichloromethane:cyclohexane (1:3) with ASE®. Fat was partially removed using silica gel in the ASE cell, and the extracts evaporated in Turbovap® and further cleansed in ASPEC®. The solvent was switched to isooctane and the samples were concentrated before being added to the recycling standard and analysed using GC-MSMS. A calibration curve was used for quantification.   

Quality

Analyse av samleprøver gir manglende informasjon om variasjon mellom fisk fanget ved hver posisjon og begrenser muligheten for statistisk behandling.

• For practical reasons, the samples were taken from capelin in different locations in the Barents Sea, and sometimes at different times of the year. The results are therefore representative of the whole of the Barents Sea, but seasonal variations in particular can cause substantial variations in pollutant concentrations. This means that it will be difficult to detect any changes over time, and that differences between years cannot be interpreted as time series for changes in levels.
• Nedre kvantifiseringsgrense (Level of Quantification – LOQ) og måleusikkerhet for enkeltparametere i analysemetodene er gitt i tabellen under. Når det er flere kongenere i gruppen gjelder kvantifiseringsgrense og måleusikkerhet for enkeltparametere.
• For pesticidene (DDT, klordaner, dieldrin, HCH, HCB og toksafen) har det vært brukt ulike metoder i perioden overvåkningen har foregått. Nedre kvantifiseringsgrense har endret seg mye fra år til år. Grensene som er oppgitt her er de som benyttes i dag (per november 2020).

¹Benz(a)anthracen, benzo(a)pyren, benzo(b)fluoranthen, benzo(c)fluoren, benzo(g,h,i)perylen, benzo(j)fluoranthen, benzo(k)fluoranthen, chrysen, cyclopenta(c,d)pyren, dibenz(ah)anthracen, dibenzo(a,e)pyren, dibenzo(a,h)pyren, dibenzo(a,i)pyren, dibenzo(a,l)pyren, indeno(1,2,3,-c,d)pyren, 5-methylchrysen.

²A higher LOQ and greater measurement uncertainty for dibenzo(a,e)pyrene, dibenzo(a,h)pyrene, dibenzo(a,i)pyre and dibenzo(a,l)pyrene. These analytes are not accredited.

³Depending on the concentration range, the measurement uncertainty is greater at concentrations close to the LOQ (<10 x LOQ) than at higher concentrations.

Reference level and action level

Reference values for the unaffected state of substances not found in the natural environment, like persistent organic pollutants, will be zero (really the detection limit). Since these are in practice widespread, it will be natural to compare the levels in capelin with those measured in other comparable species.

Action levels

Increase in the level of pollutants over a certain number of years, or a sudden increase from one sample to the next in the same area.

Should capelin be used directly for human consumption, it will be relevant to compare with the maximum levels applying to products intended for human consumption in the EU and Norway (EU 2006 and FOR 2002-09-27 no. 1028, see the reference list). There are three persistent organic pollutants on this list, sum of dioxins, sum of dioxins and dioxin-like PCBs and sum of six non-dioxin-like PCBs.

The maximum levels set for food and feedstuffs for fish, however, cannot be applied directly to the levels in the fish. This is because the fish undergoes an industrial process en route to fish meal and fish oil, and thence to feed, which can change the content of contaminants from the raw material. It is only when the Norwegian Food Safety Authority takes samples from ingredients like fish meal and fish oil or the finished feed product that the maximum levels are applicable and measures can be taken if the concentrations are too high.