Taxacom: the CODA challenge

[John Grehan]

In reference to some earlier comments about panbigoeography and
creationism, that there was some kind of similarity of argument. Below is a
demonstration that any such comparison is misplaced as the issues here are
very empirical and reach to the heart of what constitutes valid evidence.
The following is a text I mailed out en masse to about 150 people
involved either in biogeography in general or lepidopterists (mostly
butterflies) who have utilized the popular Center of origin, chance
dispersal, and adaptation (CODA) approach to biogeography. I have
endeavoured to be direct and specific while still respectful (I guess, but
who knows if some find even being questioned is offensive). No response
from any of the CODA people - no surprise. Nothing ever changes in that
respect.

"Dear colleagues,

This email is directed to individuals I understand either have an interest
in the issues mentioned below, or participate in methods involved. Attached
for your information is a recent summary article about the age and origin
of the Coenonymphina that was recently published in Cladistics, and copies
of which were widely circulated. The summary article hones in on the
principle global argument based on tectonic correlation evidence. In this
paper I make a rather blunt statement about the veracity of the prevailing
CODA method (center of origin, chance dispersal, and adaptation) which I
and others have shown to lack empirical foundation for the
transmogrification of minimum age fossil calibrations to actual or maximal
molecular ages, ancestral area analysis that generates sequential chance
dispersal where vicariance is equally applicable, and the use of area units
that have no empirical existence.

A colleague of mine recently said that to describe the CODA method as
'rotten to the core' (in the attached article) is not very useful. And in
isolation, perhaps so. But then, there are already decades of
circumspective critique, and Heads published on the foundational flaws of
CODA nearly 20 years ago. But to no effect (at least for butterfly
biogeography), and in that respect the circumspection has been 'not very
useful'. Over the last several decades, most CODA theorists either ignored
the critical problems (but some molecular theorists have agreed), or when
acknowledged, just carry on anyway. So I hope you understand that even if
it makes me seem like a rather obnoxious personality (and who knows,
perhaps in this context I am :), it is time to be blunt. Equally well, it
is time for CODA theorists to be equally blunt and direct and come forward
to provide the empirical evidence in support of the CODA methodology that
is more than 'this is what we believe this to be true'. And 'empirical' is
the emphasis. In a recent podcast, a butterfly authority asserted (without
empirical evidence) that New Zealand lycaenid butterflies arrived over the
sea from Australia, the one place where their closest relatives are absent.
Refutation of vicariance needs to be empirically grounded. A science that
is not empirically defended is not science – at least that is what I was
taught as a student.

On the other side, CODA theorists should look critically at
panbiogeographic methodology. Some have argued that tectonic correlation is
circular, but in reality it is a direct inference of a tectonic correlation
to a tectonically associated biogeography. The existence of spatial
tectonic-biogeographic correlations is an empirical reality that allows the
data of distribution to be critically evaluated. I admit that the recipe
driven CODA method makes for an attractive option, as one only has to plug
in the 'data' and get the answer, whereas panbiogeography is more nuanced
and requires a knowledge of the details of distribution as well as
geography and tectonics, not to mention a familiarity with biogeographic
patterns in general (for example, the overall global pattern for
Coenonymphina fits pretty well with patterns in plants that were recognized
as early as 1952). And tectonic correlations are not always straightforward
and have to be looked at critically, as noted by Heads (in litt.). So
perhaps we have made some mistake or other with respect to the
Coenonymphina correlations made in the Cladistics article, in which case an
empirically based falsification would be of great interest.

So I hope that someone in the butterfly field will publish a counter
critique that directly addresses the CODA issues that I and colleagues have
identified in the Coenonymphina article, and demonstrate how the CODA
method is actually empirically correct while the panbiogeographic method is
empirically wrong. If CODA is really empirically sound, surely it should
not be hard to show that. The Coenonymphina provides an ideal test subject
in this respect, having been analyzed in detail through both CODA and
panbiogeographic methodology. Far better for the science of butterfly
biogeography than ignoring competing research programs (sensu Lakatos), as
was unfortunately the case in a recent *Papilio *study. Recognition of
competing research programs is an attribute of a dynamic and progressive
science. I look forward to future published discussion and debate but I am
always happy to discuss these issues further with anyone, should they so
wish."

-- 
https://nam10.safelinks.protection.outlook.com/?url=https%3A%2F%2Fhepialidsoftheworld.com.au%2F&data=05%7C01%7Ctaxacom%40lists.ku.edu%7C9da1656d49c44da20bfe08dbceaf3e54%7C3c176536afe643f5b96636feabbe3c1a%7C0%7C0%7C638331024304946417%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C3000%7C%7C%7C&sdata=nXvywLjzZGHIsqcqCTgVIQ39UU4HELjlfHN3D413XwM%3D&reserved=0 (use the 'visit archived web site'
link, then the 'Ghost Moth Research page' link.