Two of three authors in Argentina of a 2002 paper purporting to show evidence of bird-like fossil footprints from the Late Triassic age have retracted it after subsequent research suggested their estimates were off.
Here’s the notice for “Bird-like fossil footprints from the Late Triassic:”
In this Letter, we considered the bird-like footprints from the former Santo Domingo Formation of northwest Argentina to be of Late Triassic age. Recent radiometric dating (1) of the sedimentary sequence containing these bird-like footprints (renamed as the Laguna Brava Formation) indicated a Late Eocene age. Further geological studies (2) suggest that the region suffered a complex deformation during the Andean orogeny, including block rotation. In consequence, our previous inferences about the possible implications of this finding for the fossil record of Aves are no longer supported. This Retraction has not been signed by J.F.G. Correspondence should be addressed to R.N.M. ([email protected]).
That first link is to a Brief Communication Arising about the paper that appeared in Nature in March and began:
Bird-like tracks from northwest Argentina have been reported as being of Late Triassic age1. They were attributed to an unknown group of theropods showing some avian characters. However, we believe that these tracks are of Late Eocene age on the basis of a new weighted mean 206Pb/238U date (isotope dilution–thermal ionization mass spectrometry method) on zircons from a tuff bed in the sedimentary succession containing the fossil tracks. In consequence, the mentioned tracks are assigned to birds and its occurrence matches the known fossil record of Aves.
Ricardo Melchor — “R.N.M.” — tells us that nothing has changed since the March Brief Communication Arising:
The original version of the retraction included the new age published in March communication arising, but Nature editorial policies indicate that no new information can be included in a retraction. So the Editor suggested putting the new age information on that note (if you read it, you will see that there we indicate that we may be wrong in 2002 but we do not state it).
In this case, what happened is that background geological information changed, and the finding must be understood in a new framework. We (several colleagues and me) conducted additional field and laboratory work aiming to confirm the age of the tracks. Unfortunately the results were different of what we expected.
But, I decided to clear the situation. It was a mistake, now corrected.
The original paper has been cited 26 times, according to Thomson Scientific’s Web of Knowledge.
We’re not sure why Jorge Genise — “J.F.G” — didn’t sign the notice, and he hasn’t responded to a request for comment.
I think I remember the original paper – but I am not in that field to judge the error. I know an author in another field who is proud of having published something wrong by a time factor of 1000, but this came out only after a couple of years, so it was somehow right the time he had published it.
You know, to some extent this just seems like science – they did more work and came to different conclusions – and maybe a retraction wasn’t really necessary?
The originality of the paper has been lost in the translation from “Late Triassic” to “Late Eocene”.
How can this be a retraction? Nature should leave it in the record as-is. There is no hint of fraud — the scientists did the best they could, they made an inference, the reviewers accepted it. That they were wrong in light of new evidence is irrelevant. This was the state of knowledge or inference in 2002. Science always moves forward with new evidence. So let them write a new paper explaining how and why they were mistaken and what their current understanding is — and note this can also change in the future. It seems an abuse of the retraction system.
The obvious conclusion from this precedent is that ALL papers will be retracted at some point, as all papers will be proven wrong, in the long run.
It’s interesting that Melchor didn’t mention he already published a reappraisal of the age of the formation in January:
http://www.sciencedirect.com/science/article/pii/S0040195112006907
The implications for the bird tracks are clearly mentioned in this paper. Wouldn’t that have been enough for Nature to retract the original publication?
Uhm…that paper *is* mentioned in the retraction notice (second link).
This is difficult, indeed.
Should a paper be retracted, if additional scientific data, that was not available at the time of publication, shows that the interpretation of the original data and the conclusion was wrong?
Provided there was no error in original data collection, I would also tend to leave the paper on record and see it as “science evolving”.
The “retraction community” needs to find answers to these questions. My personal opinion: there is no clear black and white. Outright fraud like data fabrication needs a retraction. And a 100% correct paper needs no retraction. And in between, there are all shades of grey……