I did read the paper, but didn’t spend time sorting out their trees
thoroughly. It is a concept that I don’t think will have much
traction.
All of modern taxonomy is done on molecular genetics. Much of it done
by mitochondria DNA. Although as David said, this isn’t the best
because of introgression and hybridization involved in producing a
number of the recognized species.
The 23 categories they break it down into, are too small by the modern
convention. The modern classification currently being used is
Angiosperm Phylogeny Group III system, or APG III, for short. Kept
updated and posted at http://www.mobot.org/MOBOT/research/APweb/ .
Clades are groups from single common ancestor. Based on this premise is
how ‘Iris’ was redrawn to include turbosa and Iris domestica (formerly
Belamcanda chinensis) as they were well nested in the clade.
Tillie (2001) and Wilson (a number of publications) have already done
trees, and although there are a few problems, are the current basis of
the ‘Iris’ phylogony. Some of the larger and more widely distributed
species ( siberica, pumila, and setosa) are represented in the computer
analysis by individuals that may not be represented of the species.
That is, there should be phylogonies of these large groups, and the
genome chosen for over all analysis should be the basal or as close as
we can find, to basal species. Otherwise the rest of the tree can be
off. And certain nuclear genes should be used as well, as introgression
and interspecies hybridization may give mitochondria alone, a wrong hit.
Older phylogonies based on plant characteristics just don’t work any
more. Rodionenko's carefully drawn up tree is based on physical
characteristics. A method that served biology well , before molecular
genetics. Unfortunately there are parallel derived traits and off
plant characteristics which can mislead, based on systemics.
Another large group I hope to look at with more detail is brevicaulis .
Thisis a wide spread species. I have plants of Candian origin.
Looking at genetic code of this , compared to other brevicaulis from
other locatiobns may give us a nice tree reflecting it’ migration as a
species. And help with placing it on phylogogy tree. This is the route
I think needs to be done. Group into small groups, find basal genome,
and then relate groups.
There does seem to be some natural groupings, but crested, beardless,
and bearded are not clades. A nice paper by Wilson addressed crested
situation. She has a dedicated grad student pursuing crested iris. I
made a couple of crosses with tectorum onto MTB plants to help her out
with her studies.
Chuck Chapman
