A plant scientist at the Colorado State University has retracted a fifth paper.
Here’s the notice for “Influence of ATP-binding cassette transporters in root exudation of phytoalexins, signals, and in disease resistance, a paper originally published in July 2012:
The Journal, Chief Editor and the Authors wish to retract the Original Research article cited above in its entirety. Based on information reported after publication, this article was found to have images that were inappropriately manipulated (Figure 1B: actin panel; Figure 6A: PR1, PR5; Figure 6B: AtATH6, AtATH10). The authors and the journal regret the errors and regret any inconvenience to the readers of Frontiers in Plant Science.
The last author of the now-retracted paper, Jorge Vivanco, has had four previous retractions, including one in Nature. He tells us:
This unfortunate situation is independent of the previous retractions.
It was derived from mislabeling and duplication of images that slipped by me, the co-authors, the editor and the reviewers of the journal. The data in the notebooks coincides with the results of the paper. Additionally, we repeated the studies and they coincide with the published data. The policy of the journal is very strict when it comes to these issues and that is why the paper was retracted even though the overall story of the paper is correct.
We are reviewing other publications using forensic image analysis to ascertain if additional problems exist.
Hat tip: Rolf Degen
What a load of malarkey. Blaming the editor and the reviewers for this having “slipped by them” is utterly outrageous, particularly when framed in the context of multiple retractions from the same senior author. In my opinion, the review process is based on the assumption that the authors are honest but stupid. We trust that the data they show are the honest to goodness results of the experiments that are described… we have to, because we simply have no way of knowing otherwise. What we don’t take on faith is whether the approach is appropriate for the problem under consideration, whether the results are interpreted correctly, or whether the story is as exciting and profound as the investigator insists. Delivering honest data to the reviewers is the sole responsibility of the senior author. And this one seems to have a lot of problems with this responsibility.
Jorge Vivanco is a “legend” of sorts in plant science. So, to see this is actually quite shocking since he and his work represented one of those “role models” I admired when I was an undergrad. Especially in research related to roots and the rhizosphere. He holds some key literature and if he already boasts 5 retractions, this is a very bad omen. I have just whisked past his web-site, and looking at the publications page (http://lamar.colostate.edu/~jvivanco/publications.htm) was an extreme disappointment. There are no papers listed between 2009 and 2013. If a “legend” does not bother to update their web-site with their work, thenw hy even bother to show an outdated and fragmentary CV? Is there something there that he is hiding? Apparetly yes, there are now 5 things that are being masked by his misleading web-site. I fully agree wth Peer007 that at minimum, this individual is acting extremely irresponsibly. To skirt responsibility to the publisher, instead of holding himself accountable, to his students and to the US tax payers, is actually rather irritating to see. However, from experience, Dr. Vivanco represents the typical, arrogant status quo attitude that abounds in the plant science community (no names mentioned here). Things that definately need to be addressed publically by Dr. Vivanco are:
a) Are there any more “slips” that he and his students are aware of but that he is not informing us of?
b) What consequences will he and his students face? How does CSU, which appears to be a public university, handle this case and how is the ORI viewing this? Will he suffer any consequences, or will all this just be another stream passing under the bridge? Will he have to pay a fine, return funds, or re-imburse travel or symposium fees that had been given to “advertise” these findings?
c) All of his papers now need to be examined in detail.
d) Surprisingly, until recently, one of his other retracted papers remained in full view (the original paper) without being removed, even though there was a retraction notice. When I contacted the publisher, Springer, about if they thought it was morally correct to be making profit off a paper that was supposed to have been retracted for at leas one year, I got a really stiff and arrogant claim that profit was not being made. As if a 30-40 US$/PDF download does not fall into the “profit” category.
One query to Frontiers. How was image manipulation not detected during the peer review?
Finally, for Peer007, can one equate honest stupidity with stupid honesty?
why not ask the editor/reviewers of the paper for comment on why it might have been missed?
also, I would prefer that retracted papers be left online, with a big retracted watermark and the retraction notice at the top… (with a similar marking for expressions of concern)… without a paywall:
1) this allows others to spot plagiarism/republication of the article in a less than ethical manner by others who have a copy
2) it allows others to identify if the editor /reviews should have caught the mistake and place blame on them… and for others to learn to identify bad science
3) it allows those reading papers that cited this paper to know what assumptions the authors might have based their work on and better understand if the prior work was essential to this or simply light under which to interpret results.
Yes. Vivanco says that it slipped past “me, the co-authors, the editor and the reviewers of the journal.” but of those four, the first two bear far more responsibility than the others.
What is “forensic image analysis”? What specific tools are used for this method?
“Based on information reported after publication, this article was found to have images that were inappropriately manipulated”
And what is then appropriate manipulation? Anyone? Or is the above rather an oxymoron?
For forensic image analysis, there is specific software that can detect fine structure changes in the pixels that would be apparent if Photoshop tools were used to alter an image. I don’t have any experience with these myself, and I doubt they are really necessary. Most manipulations are done by crude copy and pasting.
As for ‘appropriate manipulation’ that is when you work on an image to make it more clear. All photographs that appear in the scientific literature are manipulated. This means, at the very least, to adjust the contrast and brightness until the image looks as close to the original as possible. You will often have to cut up the images to assemble the parts you are showing and discard the rest. You may also be bringing in images from different experiments into the same figure. Prior to Photoshop these actions were done during the photography and printmaking process.
The goal in image manipulations is to allow the eventual read of your article to see the same things in the data that you see when you examine the original. (And one thing that is often forgotten is that colorimetric western blots can be very difficult to photograph for the non-expert.) This might require a significant amount of fiddling.
Inappropriate manipulation is when you use these tools to create an image that looks different from the original. There seems to be an almost unbearable drive to improve on images simply because the tools are now easily available.
I am not a forensic scientist, but I found Dan’s explanation to be intriguing. I tried to pull his comment into a real-case context in the plant sciences, where one often sees photos of plants growing in vitro or a greenhouse, often cropped to show only the desired structure. Therefore, if an original photo would show, let’s say a root and a shoot growing from let’s say, a leaf, but the authors were to “crop” the photo to only reveal the shoot OR the root, masking the other structure, in order to support claims of ONLY shoot OR root formation, then this would I believe fall into the category of “inappropriate manipulation”. If the authors crop the shoot, but declare in the text that the shoot was cropped in the photo, then would that make the manipulation any less manipulative or would it make the manipulation “appropriate manipulaton”? One could also say that in many studies in the plant sciences, scientists select the best and the worst looking plants to highlight extremes between treatments. In other words, biased selectivity and manipulation occurs even before the photo is taken. I think this aspect, at least in plant science, is worth a fresh debate. The tool of the imagination is only given shape once technological tools exist to support it, possibly explaining why we are having such an explosion of fraud and also of fraud detection in the past 2 or 3 years.
Yes, indeed, engaging in misleading editing is fraud — but this is completely unhelpful for determining whether it is appropriate to crop images in non-fraudulent ways. (The answer, incidentally, is “yes of course this is acceptable.”)
I would say you are describing editing. Manipulation, unless it refers to skillful handling of physical objects, carries a negative connotation.
http://www.thefreedictionary.com/manipulate
“Jargon,” in a scientific sense, includes both specialized vocabulary that does not exist in common language and also the use of common words with technical, specialized meanings. Obviously the use of “manipulation” is being used in the a sense very close to the first definition in your link, “To move, arrange, or control … in a skillful manner.” Any pejorative sense is a misreading.
The web-site link to Vivanco’s page at CSU is not linking:
http://lamarforwarder.colostate.edu/~jvivanco/