Taxacom: On describing new taxa without using phylogenetics: some suggestions

[Douglas Yanega]

At the risk of poking the proverbial bear, I find cause to dispute two 
comments that seem to be demonstrably oversimplified:
> It should also be remembered that all taxonomically relevant morphological information is ultimately encoded in DNA
> Although DNA also contains homoplasy, it tends to be random, whereas morphological homoplasy is often systematic and directional.

You speak of DNA as if it is a single thing, with uniform properties. 
Mitochondrial or chloroplast DNA is, however - and I think we can all 
agree on this - demonstrably different in terms of how it is inherited, 
and what it codes for (and, therefore, what selective forces drive its 
evolution), from nuclear DNA. Which DNA contains "taxonomically relevant 
morphological information"? Is that the only DNA that people are 
sequencing? The pattern of relationships that one constructs between 
putative taxa using DNA that is only inherited maternally, and evolves 
at a different rate from nuclear DNA, introduces its own significant set 
of complications. There are plenty of known cases where two biologically 
definitive species have identical barcodes, or where one biologically 
definitive species has more than one barcode (if either counts as 
homoplasy, it certainly isn't random), and it is only by using other 
evidence that the pitfalls of putting faith in DNA as the ultimate 
arbiter of taxon delineation become evident.

I admit that I was trained by fairly "old guard" morphological 
systematists, back when tools like PAUP had literally just been 
invented, and that they felt strongly that if you had a good reason to 
suspect that a particular set of characters (e.g., male genitalia, or 
hindwing venation, or whatever) was *full of noise*, that you should 
exclude it from your analyses a priori. Some of you might find that 
appalling, to reject potential evidence *before* running an analysis, 
but the funny thing is that, at least for the groups I'm familiar with, 
a good number of those "biased" phylogenies proposed in that time window 
(after PAUP and MacClade and such became available, but before DNA was 
part of the data pool) have held up very well against the later 
phylogenies based on more objective and inclusive datasets. I'm not 
convinced that all sources of data are equal, and not convinced that 
more data are always better. I think you *can* drown out the signal with 
too much noise, and that the best way to test phylogenetic hypotheses is 
by sampling more *taxa*, not by adding more noise.

Peace,

-- 
Doug Yanega      Dept. of Entomology       Entomology Research Museum
Univ. of California, Riverside, CA 92521-0314  voicemail:951-827-8704
FaceBook: Doug Yanega (disclaimer: opinions are mine, not UCR's)
              https://nam10.safelinks.protection.outlook.com/?url=https%3A%2F%2Ffaculty.ucr.edu%2F~heraty%2Fyanega.html&data=05%7C02%7Ctaxacom%40lists.ku.edu%7Cb4611381a60645efb3e408de06acf722%7C3c176536afe643f5b96636feabbe3c1a%7C0%7C0%7C638955537126789891%7CUnknown%7CTWFpbGZsb3d8eyJFbXB0eU1hcGkiOnRydWUsIlYiOiIwLjAuMDAwMCIsIlAiOiJXaW4zMiIsIkFOIjoiTWFpbCIsIldUIjoyfQ%3D%3D%7C0%7C%7C%7C&sdata=rI%2BjvLZ4MCfjtThz02sEYVfcF%2BD4aD4SkysLW8xddTY%3D&reserved=0
   "There are some enterprises in which a careful disorderliness
         is the true method" - Herman Melville, Moby Dick, Chap. 82