By Chelsea Harvey
The Washington Post
Published 29 Feb 2016

March 11 will mark the five-year anniversary of the Fukushima Daiichi disaster, a series of nuclear meltdowns, triggered by a devastating earthquake-induced tsunami, that released massive amounts of radioactive material and resulted in the largest nuclear disaster since Chernobyl. Since then, the plant’s damaged drainage system has continued to leak radioactive water into the environment, and one of the biggest enduring public concerns has been the safety of fish caught in the area’s surrounding waters.

Now, with the fifth anniversary approaching, new research may allow the public to breathe a little easier over the state of its seafood. A study published Monday in Proceedings of the National Academy of Sciences indicates that the overall risk of contamination in fish around Japan is quite low — with a few notable exceptions.

“Highly contaminated foods attract people’s attention,” the new paper’s lead author, Hiroshi Okamura, a researcher with the National Research Institute of Fisheries Science in Japan, said by email. “Some people cry not to eat seafoods and other people argue many foods are not dangerous.”

In fact, the authors note in their paper that conflicting reports have been released over the past few years on the contamination of fish in Japanese waters, with some arguing that the concentration of radioactive material had declined rapidly since 2011 and other arguing that it had hardly declined at all. “These contradictory statements must confuse the general public,” the authors wrote.

The confusion comes from variations in the way different studies collect data on radiation contamination, said Ken Buesseler, a senior scientist at the Woods Hole Oceanographic Institution, who was not a part of the new study. Different data collection methods have different “detection limits” when it comes to radioactive material, or the threshold above which the researchers can detect and record contamination. A high detection limit, for instance, will only detect radioactive content above a certain level, typically a level that’s relevant for human health concerns.

But having such high detection limits also means that the public is not getting the whole picture of how quickly contamination in the region is dissipating — and when different research groups publish studies using different detection limits, the results can sometimes appear contradictory.

“I don’t think the controversy is in the data, but in the interpretation of the data,” Buesseler said.

This is a problem the new research group attempted to address in their study. Okamura and his colleagues first took data from Japan’s Ministry of Health, Labor and Welfare on radiocesium in aquatic food sources. This is one major type of radioactive contaminant released by the Fukushima disaster.

The researchers intended to perform a statistical analysis on all the data to evaluate the risks for different species and locations. But they noticed that many of the measurements were labeled “N.D.,” or “not detected,” meaning contamination could have been present but was not measured.

Having so many N.D. designations in the data could have reduced the precision of their results, the authors noted — so they came up with a solution. They developed a model that can be used to help fill in the blanks when there’s missing data by using data from similar species and locations to generate information.

The researchers quantified risk by estimating the probability that a fish would contain a radiocesium content above a certain threshold. Radioactivity is measured in a unit known as “becquerels.” So for each species, the researchers calculated its risk of having a radiocesium content higher than 20 becquerels per kilogram, 50 becquerels per kilogram and 100 becquerels per kilogram.

For reference, the current designated safe level in Japan is 100 becquerels per kilogram — and this is among the lowest in the world. Prior to the Fukushima disaster, Japan’s official safe level was 500 becquerels per kilogram (it was lowered after the incident in order to reassure consumers that their food was being monitored closely). And in many other places around the world, the safe limit is set at 1,000 becquerels per kilogram or higher.

After finally conducting their analysis, the research team found that the risk of cesium contamination around Fukushima has steadily decreased since 2011. Overall, they found that in all species and locations, the median probability of contamination higher than 100 becquerels per kilogram was nearly zero.

There were some variations across species and location, though. In general, freshwater species were found to have a higher risk than marine species. This has to do with the physiology of freshwater fish, Okamura explained — they tend to discharge radiocesium from their bodies more slowly than ocean fish. In addition, bottom-dwelling fish with large bodies and long lives also generally exhibited a higher risk.

Japanese white seaperch, for instance — a large, bottom-dwelling fish — had a probability of 0.45, or 45 percent, of having radiocesium concentration above 20 becquerels per kilogram and a 9 percent chance of having greater than 50 becquerels per kilogram, although just a 0.7 percent chance of greater than 100 becquerels per kilogram. And the freshwater Japanese eel had a 62 percent probability of a concentration above 20 becquerels per kilogram. (It’s worth noting that 20 becquerels per kilogram is still well within Japan’s standards for human safety.)

By comparison, the Japanese jack mackerel — a non bottom-dwelling fish — had a probability of less than two decillionths of having a concentration above 20 becquerels per kilogram. It’s an impossibly small probability already (imagine a decimal point with 32 zeroes behind it), and many times smaller for the probability of having a concentration higher than 100 becquerels per kilogram.

One big takeaway from the research is that contamination around Fukushima seems to be declining faster than researchers have previously thought, Okamura noted. Radioactive material breaks down over time, so in the case of a nuclear disaster, a big concern is close monitoring to see how quickly the contamination is declining.

As far as the safety of seafood around Japan is concerned, Okamura said he and his colleagues “were careful not to say whether seafoods are safe or dangerous while writing our paper,” noting that they preferred to leave it to individual consumers to decide whether they feel a food is dangerous, given the results of this study.

However, the paper does conclude that the overall risk of contamination in the region’s fish is very low and in most cases has an extremely small chance of rising above Japan’s official safety limit — which is already far more cautious than most national standards throughout the rest of the world.

Buesseler noted that more fisheries are likely to continue opening up around Japan as contamination risks continue to fall, although greater restrictions will likely remain on freshwater systems. But he added that continued careful monitoring of the region is still important.

This study only looked at radiocesium, but there are other types of contamination that ought to be periodically assessed as well — a radioactive form of strontium, for instance, Buesseler pointed out. In general, concentrations of radiostrontium are much lower in the region’s marine ecosystem and are not thought to pose great risks, but it will be important to keep an eye on things in the future.

As Buesseler pointed out, water leaks at the damaged Fukushima plant have continued leaching radioactive water into the environment since 2011, so the levels of contamination in the area may still be fluctuating. “We have to be diligent to continue to monitor this,” he said.

And when it comes to monitoring, all the trouble with the “not detected” data designations has prompted the authors of the new study to recommend data collection plans with reduced detection limits in the future — a suggestion Buesseler also agrees with. As the authors note, gathering many measurements with high detection levels may not actually be as valuable as collecting fewer measurements with low detections levels, which give scientists a better of idea of the true state of the environment.

And that’s important for communicating accurate and up-to-date information to a worried public, even five years after the accident. “We’ve done a poor job communicating to the public, who are concerned, when they should be and when they don’t have to be,” Buesseler said. “And this data addresses the latter — how quickly things have gotten better, at least in the seafood.”

But, he noted, it’s also important to relay to the public that Fukushima is not out of the woods yet. Discussions surrounding the plant tend to skew very dire or very optimistic, he pointed out. The reality is that things are still “somewhere in the middle between those extremes.”