Today in a local library I found the first paper in a short series that
deals with one of the controversial results of Charles Sibley & Jon Alhquist's
DNA hybridization work. I'll warn you in advance that I consider the paper to
be the printed version of a flame, but that I also believe the new results to
be correct and that I'm glad to see them out at last.

One of the surprising results of S&A's DNA study was a report that the
American Finfoot (Heliornis) was genetically very close to the Limpkin
(Aramus), and [less surprisingly] that both were fairly close relatives of the
cranes (Gruidae). Probably because the result was so odd, and because they
were unable to obtain DNA from either the Asian or African finfoot (I'll get
the genera wrong if I do this from memory), they treated Aramus & Heliornis as
"tribes" of a subfamily Aramidae in the Gruidae. I've seen most of the
original data, and that result is strongly indicated in the data IF all of the
DNA samples were correctly identifed.

Peter Houde has been studying the early evolution of Gruiform birds, and
Sibley heard indirectly that he & Fred Sheldon repeated the DNA measurements,
and got different answers. They have not been willing to share their results
with Sibley, but given the controversy, Sibley has acknowledged the
possibility of a lab error (in print, twice) and encouraged a study that
confirms (or not) his result. Houde & co-workers now are producing a series of
paper that seems to meet that end.

Houde's first paper finally appeared in Cladistics, issue #1 for 1994. He
examined morphological traits of finfoots, limpkins and cranes, and summarizes
the results of new DNA comparisons. Anatomically, finfoots are rail-like, and
distinct from cranes and limpkins, which are rather similar to each other.
Houde, Sheldon and coworkers devised a new technique for DNA hybridzation that
works well (they did extensive comparisons between the new technique and the
"traditional" technique, which will be reported in a later paper) with museum
specimens, and were able to incorporate all three species of finfoot into
their study. Their new DNA comparisons reproduce (as had others) the difference
between Aramus & Gruidae, but finfoots are very different and are about as
distinct as are sunbitterns (Eurypyga), which unfortunately weren't compared
to finfoots as best I could tell from a cursory reading (rails are similarly
or even more distinct). There has been speculation that the three genera of
finfoots might not be closely related, but the new comparisons seem to dispell
that notion. So the short of it is that Sibley's original data are erroneous
(probably based on two samples of Aramus, although that can't be checked), and
we now know more about the place of finfoots in the avian scheme of things.

Houde issued a general caution that one has to be careful about similar
problems elsewhere in S&A's results. I'd extend the caution to any data in any
paper, and the problem is especially severe when a new interpretation is
radically different from old ones and when the concept being changed has only
a single representative (so that internal consistency can't be checked). There
are, at most, only a few additional such cases in S&A's data, and Sibley & I
are busily trying to root them out. But it's nice to see this part being
clarified, and I (and I'm sure Sibley) await the publication of the other
papers in the series in 1995. The new DNA hybrization paper looks especially
interesting.

However, it's also unfortunate that such an adversorial approach was taken.
A frequent side effect of such play is to discredit the parts of a study or
technique that ARE correct and to discourage the use of valuable new
technologies. That's what's happened to DNA hybridization in birds, at least,
and I'm hearing elsewhere in systematic zoology. The new technique currently
being championed is gene sequencing, but its early results suggest that the
incidence of laboratory errors is as large or worse than in DNA hybridization,
and it's hard to say how that will play out. But in the current climate of
concern for biodiversity, we need to see effort expended in developing
techniques that quickly and accurately estimate genetic diversity to
complement data on faunistics in order to propose more rational programs to
preserve biodiversity. But science like all other endeavors human mirrors the
culture in which it is practiced and too often has a nasty habit of being
adversorial when it could be practiced more cooperatively.

Paul DeBenedictis
SUNY Health Science Center at Syracuse
BENEDICT at VAX.CS.HSCSYR.EDU