Radioactive fish study retracted for “significant and extensive” corrections

This article has been retracted at the request of the Authors.

The authors have informed the Editor and the publisher of the journal that substantial parts of the article need corrections. These corrections are so significant and extensive that the article could no longer be considered to be same as the one originally submitted, reviewed and accepted for publication.

The paper has been cited twice, according to Google Scholar, and its authors are at the Institute of Marine Research in Bergen, Norway, which studies the Barents Sea, among other projects. We’ve asked the corresponding author, Hilde Elise Heldal, for details on the needed corrections, and will update with anything we learn.

In the meantime, here’s the original paper’s abstract:

Dispersal of 137Cs from Komsomolets and K-159 is simulated using realistic rates and hypothetical scenarios. Furthermore, spatiotemporal 137Cs concentrations in Northeast Arctic cod and capelin are estimated based on survey data. The results indicate that only pulse discharges from K-159 will cause concentrations of 137Cs in cod muscle exceeding the intervention level of 600 Bq/kg fresh weight. A discharge of
10% of the 137Cs-inventory will result in concentrations in muscle of cod exceeding the intervention level for approximately two years. In fact, a discharge of 10% of the 137Cs-inventory results in an overlap of 8e30% between the different size groups of cod and levels that exceed the intervention level during the first year after the discharge. For capelin, individuals less than one year old during the first year after a discharge are more likely to be severely affected by discharges comprising
50% of the
inventory.

And here’s a story on how this sort of research is used from Yale Environment 360 magazine.

Update, 7:45 a.m. Eastern, 10/17/12: Heldal tells us why the paper was retracted*:

Short time after we had published the article I received critical questions from a colleague at another Norwegian institution. The questions made us go through our model inputs, and it turned out that they were wrong, giving ¹³⁷Cs concentrations in cod and capelin wrong with a factor of 1000 (1000 times to high). We were asked by the editor of the journal to write a corrigendum. It became, however, too extensive, as we have to change both parts of the text and some of the figures. The article was therefore retracted. We are now in the process of rewriting the article and will resubmit it to the same journal within short time.

I cannot say how sorry I am for this, it shouldn’t happen. However it did. I have learned that modeling can be difficult and you should be extremely critical to what comes out of the model.

*Note: This quote (via email from Heldal) has been edited since it was originally posted, at Heldal’s request. She claimed that she did not know we would be posting her comment, but as we pointed out, Ivan identified himself as a Retraction Watch blogger and included a link to Retraction Watch in his email to her.