This dataset has been in GenBank for quite some time (since 2009), and has long intrigued me, as it included sequences of an apparent
Gavia arctica, that are identical to other available sequences of
Gavia pacifica (as was the case in the data of
Wink et al. 2002, but to my knowledge has never been the case anywhere else--all other available mitochondrial sequences of these two taxa are clearly distinct).
Here, the ID of this bird was apparently changed/corrected at a late stage: in Tables 1 and 2 of the paper, it is now called a
Gavia pacifica from Mexico. But it is still labelled
Gavia arctica in the trees (figures 1 and 2), where it appears to have sequences identical to those of
another Gavia pacifica from GenBank. (And of course the sequences are still called
G. arctica in Genbank.)
To make things still muddier,
the cox1 sequence of this bird was also picked up in GenBank by BOLD and included in the barcode database of the BOLD systems. There, quite amazingly, it recently had a
nice picture of an obvious Gavia a. arctica joined to it, which claims to be an image of the specimen... (But that actually appears to come from
here and, if so, presumably represents an infringement to copyright laws.)
In the same dataset, the "
Phoenicopterus ruber" seems to be a
Ph. minor.
(And, not in the same dataset, but included in the analysis, the
complete mitochondrial genome of "Pseudopodoces humilis" is problematic too.)