Radiation Testing of Canned Salmon Caught in the North Pacific Ocean Before and After the 2011 Fukushima Nuclear Accident

Below are the results of radiation testing on canned salmon produced by an American seafood company, Vital Choice Wild Seafood and Organics. (Here is the Japanese article).

Note 
Becquerel (Bq) is a unit of radioactivity: 1 Bq means "one disintegration event per second."
Picocurie (pCi) is a unit of radioactivity: 1 pCi = 0.037 Bq

Highlights

• In this citizen-driven project, Pacific salmon cans from a U.S. seafood company, Vital Choice Wild Seafood and Organics, underwent radiation testing by citizen labs in Japan. Salmon cans before and after the 2011 Fukushima nuclear accident (2009, 2011, 2012, 2014) were analyzed for 4 radionuclides: Cesium-134, Cesium-137, Strontium-90 and Tritium.
• A very small amount of Cesium-137 was detected in all 4 samples, mostly due to background Cesium-137 ("legacy Cesium-137) from nuclear weapons testing.
• The 2011 sample canned in August 2011 had a slightly higher level of Cesium-137 than the other 3 samples, and the difference of about 0.02 Bq/kg might be due to the Fukushima accident if Cesium-134 coexists.
• However, no Cesium-134 was detected. Based on the Fukushima-specific Cesium-137 to Cesium-134 activity ratio of 1:1 and the 2-year half life of Cesium-134, any Cesium-134 potentially present in the 2011 sample falls below the minimum detection limit. 
• The higher Cesium-137 level in the 2011 sample is likely to be related to the Fukushima accident, but data is inconclusive.
• No Strontium-90 or tritium (both free-water and organically-bound) was detected above the detection limit.

Background

This voluntary "project" came together spontaneously in the summer of 2015 when concerned citizens wanted to verify the alarmingly high strontium level detected in sockeye salmon in Vital Choice's 2015 radiation testing. 

Since the March 2011 Fukushima nuclear accident, Vital Choice has taken an initiative to conduct--once or twice year--radiation testing of seafood used in their products. For the 2015 test for Strontium 90, the sockeye salmon sample consisting of six 6-oz frozen, boneless fillets with skin on was received by the lab, SGS, on December 11, 2014. A separate batch was sent to another lab, Eurofins, for radioactive cesium testing. 

According the the SGS report obtained from Vital Choice, an unexpectedly high level of Strontium-90 was detected in sockeye salmon at 1.76 ± 0.863 pCi/g (65.12 ± 31.93 Bq/kg) with the minimum detection limit of 1.45 pCi/g (53.65 Bq/kg). 

This Strontium-90 test result of 65.12 Bq/kg seemed extraordinarily high for several reasons:

1. The highest Strontium-90 level detected in seafood testing conducted by the Japanese government was 1.2 Bq/kg with the detection limit of around 0.03 Bq/kg in rockfish harvested off the coast of Fukushima Prefecture on December 21, 2011 (refer to row #10 in this PDF found on this page of the Japanese Fisheries Agency website. The location where the sample was taken is also plotted as "10" in the first map of this PDF).
2. In the 2015 study of fish caught within the harbor adjoining the Fukushima Daiichi nuclear power plant in 2012 and 2013, the highest level of Strontium-90 detected in whole fish without internal organs was 170 ± 1.2 Bq/kg (wet). This study also showed that the Cesium-134/137 levels were 200-330 times the Strontium-90 levels.
3. As just mentioned, strontium has been found at a much lower level than cesium in general, consistent with the amounts released from the accident. For instance, the Japanese government sample of rockfish with 1.2 Bq/kg of Strontium-90 mentioned above also contained 970 Bq/kg of cesium—390 Bq/kg of Cesium-134 and 580 Bq/kg of Cesium-137. Although the Vital Choice samples tested for cesium and strontium consisted of two different sets of six 6-oz boneless fillets with skin, the sockeye salmon sample from Vital Choice's 2015 test showed non-detectable levels of both Cesium-134 and Cesium-137 with the detection limit of 1.0 Bq/kg each. Thus it is highly unlikely that the Vital Choice sockeye salmon would contain any detectable amount of Strontium-90.
4. In addition, whereas other tests mentioned here used muscle for the cesium testing and whole fish without organs for the strontium testing, the Vital Choice sockeye salmon sample was just muscle and skin without any bones, consisting of six 6-oz boneless fillets with skin, which makes the reported Strontium-90 level of 65.12 Bq/kg highly questionable.  (The 2015 study mainly analyzed otoliths—calcium carbonate structures in the inner ear—but also mentioned the strontium level in whole fish without internal organs).
5. Also, the reported result, 65.12 ± 31.93 Bq/kg, has a high error margin of 31.93 Bq/kg. With the minimum detection limit of 53.65 Bq/kg, much higher than the detection limit around 0.03 Bq/kg employed in the Japanese government's test, the quality of testing is doubtful.

(Note: Vital Choice attributes the high level of Strontium-90 to lab errors, as explained at the end of their 2015 report).

With the Internet abuzz with the news of the high strontium level detected during Vital Choice's 2015 test, an idea arose to have confirmatory tests conducted by citizen labs in Japan which have—in response to people's demand—acquired sophisticated equipment and skills to detect very small amounts of radionuclides. Granted, it would not be possible to test the exact same fish, but at least fish caught in the same year, 2014, could be tested. 

Then it occurred to me that I myself owned (and still own) quite a few of Vital Choice's canned salmon with bone and skin produced in 2009 and purchased immediately after the 2011 Fukushima accident. A local friend also had Vital Choice's canned salmon with bone and skin produced in the summer of 2011 after the Fukushima accident. Furthermore, rather than having boneless frozen fillets tested, it made sense to have canned salmon with bone and skin tested for Strontium-90, because strontium accumulated in bone due to its chemical resemblance to calcium. (Shipping frozen fillets would also be cost prohibitive). In addition, testing the same sample for cesium would also yield pre-Fukushima baseline levels of Cesium-137 in salmon. 

Vital Choice generously agreed to provide canned salmon from 2014. As a "bonus" the warehouse accidentally sent me canned salmon from 2012, so now we had pre-Fukushima (2009) and post-Fukushima (2011, 2012 and 2014) sets of canned salmon, all with bone and skin. Although it was not possible to know exactly when and where in the Pacific Ocean the fish was caught, can codes showed when the can was processed. (Salmon is caught and canned in the same year). All the can codes and their explanations are provided here.

Cans were sent to Japan in multiple U.S. Postal Services flat-rate boxes. Some individuals helped with transporting them to a citizen lab in Tokyo for the cesium testing and then to another citizen lab in Fukushima for testing for strontium and tritium. (After the cesium testing was completed in Tokyo, the remaining samples were frozen and transported to the Fukushima lab by a citizen volunteer at a later time. Fee for testing was either donated by the lab or privately covered.

Results 

Actual reports (in Japanese) can be accessed here.
Results are tabulated here and detailed information on tested samples can be found here.

Strontium-90 

The strontium testing was conducted by the only citizen-run lab in Japan capable of beta radiation analysis, Mother's Radiation Lab Fukushima. The analytical method used is a liquid scintillation counter (shown in this Japanese PDF). The use of liquid scintillation counter is explained in detail here. Strontium-90 was not detectable in any of the samples with much lower detection limits than the lab originally used by Vital Choice (0.15-0.17 Bq/kg vs. 53.65 Bq/kg).

The actual report obtained from Vital Choice indicated "ASTM D5811-95" as the analytical method, which is the "Standard Test Method for Strontium-90 in Water" as described here. It was developed to measures Strontium-90 in environmental water samples in the range of 0.037 Bq/L (1.0 pCi/L) or greater, using Beta Gas Proportional Counter (β-GPC) as described here. β-GPC is an analytical method for testing for strontium in drinking water, as shown in Table 7-2 of this PDF, a chapter of Toxicological Profile for Strontium on the website of the Agency for Toxic Substances and Disease Registry (ATSDR). 

Vital Choice also conducted Strontium-90 testing in November 2013. This testing, performed by Pace Analytical Services, Inc. and reviewed by SGS, also showed "ASTM D5811-95" as the analytical method used. All three types of fish tested at the time—sockeye salmon, king salmon, and albacore tuna—had no strontium detected above the minimum detection limits of 0.0513 pCi/g, 0.0635 pCi/g and 0.0456 pCi/g, respectively. 

Although these detection limits were two-fold higher than what the analytical method would be capable of (1.0 pCi/L or 0.037 Bq/L), the minimum detection limit of 1.45 pCi/g (53.65 Bq/kg) in 2015 was even two-fold higher than 0.0513 pCi/g (1.90 Bq/kg) in 2013. This large discrepancy suggests some sort of errors, such as in converting units.

Tritium

The tritium testing was also conducted by Mother's Radiation Lab Fukushima. None of the samples showed either free-water or organically-bound Tritium.

Cesium-134/Cesium-137

Cesium testing was conducted at Shinjuku Yoyogi Citizens' Radiation Lab. Results are not decay corrected. Actual reports can be seen here: 2009, 2011, 2012, and 2014.

The 2009 result of 0.084 Bq/kg shows the pre-Fukushima, background level of legacy cesium—Cesium-137 derived from nuclear testing. The 2011 result is about 25% higher at 0.108 Bq/kg, then it goes down to 0.088 Bq/kg in 2012 and 0.080 Bq/kg in 2014, about the same levels as the pre-Fukushima level. The 2011 result of 0.108 Bq/kg is 0.024 Bq/kg more than the 2009 result, and this 0.024 Bq/kg might be from the Fukushima accident. Results are not decay corrected. Actual reports can be seen here: 2009, 2011, 2012, and 2014.

Cesium-134 was not detected in any of the four samples. Due to its half-life of 2 years, Cesium-134 derived from the nuclear testing is not expected in the 2009 sample. Any potential amount in the 2011 sample and later, by the time of testing in September 2015, would fall below the detection limit.

Discussion

Neither Strontium-90 nor tritium (free-water and organically-bound) was detected above the detection limit in any of the samples. 

Can codes indicate that the 2011 sample was canned on August 9, 2011 (see this). Can codes for the 2011 sample also include a symbol that refers to "Skeena River," meaning this particular salmon originated from the Skeena River system. (Other samples do not have a similar designation of a specific river system, and it is uncertain why the 2011 sample does). Vital Choice says, 

"Nearly all salmon—if conditions are favorable—return to the same river system from which they came, and all salmon could be distinguished according to a specific river system from which they came. It is unclear why this got specifically ‘called out’ as Skeena (and others not). Salmon are harvested in the wider areas of ocean/bay feeding into the river systems by various fishing methods."

However, no information available in regards to whether this salmon was caught in Skeena River or caught in the Pacific ocean near Skeena River. Just how the 2011 salmon might have been exposed to Fukushima-derived cesium depends on whether the salmon was returning to or just came down the river, and there is no way of knowing this. According to a paper that reviewed the oceanic release and transport of Fukushima-derived radionuclides in the first 5 years after the accident, the 2011 salmon was most likely exposed to the atmospheric fallout in the ocean or river.

Interestingly, salmon fillets caught in an Alaskan river in 2011, 2012 and 2013 were tested by the Department of Nuclear Engineering at University of California at Berkeley. It was found that the 2011 salmon had traces of Fukushima-derived Cesium-134 and Cesium-137 (see their report here). Considering the fact and timing the salmon was caught in a river, it was suggested that:

[...] salmon's exposure may not have been from its life in the Pacific Ocean, but rather from the airborne fallout collecting/concentrating (still at small levels) in river water.

For the cesium testing conducted in the Japanese citizen lab, the samples were tested in a "wet" condition, and the results reported are in "wet" weight. The UC Berkeley lab report states, "The mass used in the analysis was the wet or frozen weight of the fish — not the dry ’baked’ weight, which was considerably less." Apparently the sample weight reported in their results is the wet weight, but the sample was baked. If the reported cesium levels are indeed per wet weight, the background Cesium-137 level of 0.14 Bq/kg detected from both 2012 and 2013 samples is higher than 0.080-0.088 Bq/kg detected in the canned salmon. (An inquiry was made through the online contact form if the reported results have been converted from the dry weight to the wet weight, but there has been no reply).

FDA's radiation testing in Alaskan seafood yields results in wet weight (see this PDF for the 2016 report), but the detection limit over 1 Bq/kg does not allow comparisons with the canned salmon results.

Although a different species, the 2012 Pacific bluefin tuna study shows post-Fukushima cesium levels, except they are in dry weight and not directly comparable to the canned salmon results. However, other data available corroborate the background level of Cesium-137 seen in the canned salmon: 

1)   The Japanese government data of radiation testing on salmon since 2000, collected from the environmental radiation database and compiled here.  

2)    In a joint research project between the University of Tokyo and a citizen lab (Akita Radiation Measuring Station "Beguredenega"), the University of Tokyo conducts a high sensitivity analysis on the sample dried and concentrated by the citizen lab as described in this study. In 2014, salmon caught in Hokkaido, Japan, showed the presence of Cesium-134, while salmon caught in Hokkaido, Japan, in 2015 didn't. The results shown on the citizen lab website (2014, 2015) are converted to the wet weight. The 2015 salmon contained 0.0727 Bq/kg wet weight of Cesium-137.

Conclusion

This project revealed mostly background levels of Cesium-137 in canned salmon, which were only detected due to much lower detection limits than widely employed. No Strontium-90 or tritium was detected above the detection limit. The 2011 sample showed a small increase in Cesium-137 that might be due to the Fukushima nuclear accident.


Putting it in context

This post is not intended to give dietary advices, but it might be helpful to learn a few facts in putting the measured radiation levels in perspective and making informed, personal decisions on what to eat or not.  

1. It should be noted that the very small levels of Cesium-137 detected in canned salmon were only detected because: 

a) The testing was actually conducted.b) The testing was sensitive enough to detect a very small level of Cesium-137.

2. Many foods contain background radionuclides—natural and manmade. 

FDA's Total Diet Study which includes analysis of radionuclides shows Cesium-137 and Strontium-90 present in foodstuffs. When the most recent report from 2006-2014 (here) show all but 3 food items (TDS Food No. 74, 320 and 376) with the value "0," one might erroneously conclude there isn't any Cesium-137 in most of the food. The trick is that a specific food item is reported as "0" as long as all the multiple samples (10-12 for most foods) for the item show Cesium-137 levels below the reporting limit of 5 Bq/kg. When one or more of the multiple samples for a specific food item shows Cesium-137 over 5 Bq/kg, mean, standard deviation, and maximum levels are provided. (By the way, maximum levels shown for the three food items are: 10.8 Bq/kg for item #74, raisin bran; 93.3 Bq/kg for item #320, baby food squash; and 40.5 Bq/kg for item #376, salad dressing).

For accuracy, all the "0" values should really be denoted as "<5 Bq/kg." An even better and more accurate option is to indicate the minimum detectible level (probably 1.0 Bq/kg for Cesium-137) and denote "ND (not detected)" rather than using "0" values. 

For Strontium-90, the reporting limit is 0.1 Bq/kg. Quite a few food items contain very small amounts of Strontium-90 (see pages 22-29 of this PDF). 

3. Radiation testing of fish only shows what is contained in the particular fish tested. Other fish in the vicinity might show a similar trend, but it all depends on where the fish has been and what it has eaten. 

4. Another important fact to learn is how radiation levels are described and compared. 

Radiation levels detected in air, water and food are often misleadingly compared to "the number of CT scans or X-rays" or "hours flown on airplanes." CT scans, X-rays and airplane flights represent external exposure that lasts for a limited length of time. On the other hand, intake of radioactive material through inhaling, drinking, and eating constitutes internal exposure that lasts as long as the radionuclides remain in the body. Moreover, radionuclides incorporated into the body accumulates in and affect specific organs, and thus it is inappropriate to compare external exposure and internal exposure just by the exposure dose.

Also the radiation levels are frequently compared to the limits permitted in food established by FDA called "Derived Interventional Levels (DILs)"—guidance levels for radionuclide activity concentration in imported and domestic food—as described here. DIL is 1200 Bq/kg for the total cesium (Cesium-134 and Cesium-137) and 160 Bq/kg for Strontium-90. (See this for details on how DILs were derived). A domestic coalition is calling for lowering of DIL from 1200 Bq/kg to 5 Bq/kg for total cesium, while two international organizations jointly recommend lowering the EU limit of 370 Bq/kg for baby food and 600 Bq/kg for all other foods to 8 and 16 Bq/kg, respectively (more on this later).

DILs are based on Protective Action Guides (PAGs). As described on the FDA website, "PAGs are radiation dose levels to an individual at which protective action should be considered to limit the radiation dose to that individual." In 1998, an annual committed effective dose of 5 mSv was adopted as the PAG, which essentially accepts increased cancer deaths of 2 in 10,000 from radiation exposure. (The FDA document on this subject is quite complicated: the April 14, 2011 article in Forbes explains it in an easier term). 

In 1998, the current DIL of 1200 Bq/kg for Cesium-134/137 replaced the Levels of Concern (LOCs) established in 1986, which was 370 Bq/kg. The increase is due to the fact that LOCs assumed 100% of of food was contaminated, whereas DILs assumed 30%, but essentially the levels mean the same: It's just that you can eat more of less contaminated food before reaching the PAG of 5 mSv.

For record, DIL of 1200 Bq/kg for cesium is twice the EU limit of 600 Bq/kg for import foods. It is ten times the Japanese limit of 100 Bq/kg with the maximum permissible dose of 1 mSv per year. It's more than twice the previous, emergency limit of 500 Bq/kg with the maximum permissible dose of 5 mSv per year imposed by the Japanese government immediately after the Fukushima accident. (See this document for details on Japan's old and new permissible levels of radioactivity in food. Japan's most current limit, 100 Bq/kg, assumes 50% of marketed foods are contaminated). 

As mentioned earlier, two different groups—Beyond Nuclear and coalition in the United States and foodwatch and German IPPNW internationally—have come up with recommendations to lower the permissible level of radioactive cesium in food as follows: 

a) From 1200 Bq/kg to 5 Bq/kg for DIL (Beyond Nuclear and coalition)

b) From 370 Bq/kg to 8 Bq/kg for baby food and milk, and from 600 Bq/kg to 16 Bq/kg for all other foodstuffs for the EU limit for the import (foodwatch/German IPPNW)

Beyond Nuclear and coalition of other groups have based their recommendation--5 Bq/kg both cesium-134 and cesium-137 combined--on the work of Yuri Bandazhevsky, a Belarusian pathologist who have done research on health effects of Cesium-137. 

Recommendation by foodwatch and German IPPNW--based on a study by German Society for Radiation Protection--is derived from the maximum annual exposure limit of 0.3 mSv for an individual from “discharge of radioactive substances through air or water” in
normally operating nuclear facilities. An excerpt from their report has some important points:

"Setting official maximum levels of radionuclides to be tolerated in food is supposed to protect the population from danger. But, in contrast to chemical toxins, there is no threshold below which radioactivity is harmless. Thus there is also no dose of radiation, no matter how small, that is harmless, benign or unobjectionable. The authority (government or international organization) that recommends or sets standards, or maximum permissible value limits, basically decides on how many fatalities or cases of illness will be acceptable in a given situation.(...)  

A significant reduction in current limits is needed to reduce the risk of health problems. To derive limits that can be used as a standard to achieve this reduction, our calculations are based on a person being exposed to a maximum annual effective radiation dose of 0.3 millisieverts (mSv). This is the maximum exposure limit set out in Germany’s radiation protection legislation for normal operations in nuclear power plants; the figure applies to the exposure pathways of air and water. 

This means that current EU value limits must be reduced to 8 becquerels per kilogram of total cesium for baby food and 16 becquerels per kilogram of total cesium for all other foods."

Acknowledgements
Appreciation is extended to Vital Choice and BR for providing the canned salmon samples, Shinjuku Yoyogi Citizens' Radiation Lab and Mothers' Radiation Lab Fukushima for offering their analytical services, and NK and AT for transporting the canned salmon after arrival in Japan.