GCG - 2017-01-18 conference call

From systematics

Jump to: navigation, search



  1. [5 minutes] How the tree was built
  2. [5 minutes] Overview of the tree and the nature of the data in it
  3. Open discussion

Discussion notes I: building the tree

could we add a label to each terminal that indicates how many tips represent each species?
AH responds
we could certainly do that... I will distribute the big tree to everyone, but I can also add a label that indicates how many individuals we have of each species in the single-tip tree.
In this one-tip per species tree, we selected one best accession per species. In the big tree, things will be quite messy. This is not the definitive tree, so it needs to be distributed with that caveat.
how many tips are represented by just one gene?
Santi and Pedro
we have a framework tree based on everything that has both ITS and ETS; the other taxa are welded onto the framework tree.
this is a really good looking tree. Pedro and Santi have put a lot of work into this, and it looks really good. We are going to have a lot of well-supported clades to talk about. We are well beyond where we were previously, when we didn't know what the well-supported clades were going to be. All the sequencing and curation, and multiple rounds of modifying the tree and alignments make a big difference.
Group discussion
this is all the same a first step, just an ITS and ETS tree.
we all have work to do in separate clades. If we are a real group, we need to share this information so that we can make a solid rearrangement. We cannot depend just on this result... for example, the 9-gene tree shows the position of Carex satsumensis, while this tree does not. When we make our rearrangements, we need to take into account things that we all know, not just what is in the current tree.

The tree itself

any work we do here will be iterative. As a general comment, the tree looks great. There is certainly a strong geographic signal. I'm not pessimistic about finding morphological characters, though I think it's difficult to break out of the concepts we are used to to look at characters we aren't used to. Even the larger groups that fall together (Laxiflorae - Paniceae - Bicolores, e.g.; Fecundae - Limosae - Phacocystis). It's a bit of a tightrope walk between a large number of small morphologically tight sections and fewer sections that don't fall together as tightly. I was intrigued by a couple of other things. Intersectional hybrids are quite predictive of a couple of relationships. A lot of the spp that gave trouble morphologically in terms of assigning to section obviously fall in odd places in this tree. We have certainly generated a lot of work for students in the future.
Tree stats
we have about 1300 spp and every section sampled.
this tree is great, a bit much to look at all at once. I'm focused on Korean spp., the Anomalae group primarily, ca. 30 spp. Sangtae can send sequences based on his 9-gene phylogeny for 17 of the species. Only 5 are present in the current tree we have. In the Korean tree, Anomalae splits into two clades. Carex jaluensis is quite separate from the others in this tree, and morphologically quite divergent. This is a group that Sangtae would like to work on more. Sangtae will bring his data with him to the meeting.
I think where we are planning to make sectional recircumscription, we should be led by individuals who are working on the group. Sangtae is in a good position to do so on this section.
to the group, are there materials that should be in this tree that are not?
one thing I would encourage everyone to think about that came out of our unpublished Vignea work is that there are characters often assumed to be very homoplastic that may nonetheless turn out to be useful. The corrugated inner sheath of the leaf blade is one of these. Oddly, C. stipata and C. laevivaginata go off on their own... why? You would certainly expect it to be in with the corrugated-sheathed carices.
thinking outside of the box is going to get us some answers.
to go back to Anomalae, they mix up with Confertiflorae and Carex acutiformis in this tree. It certainly seems to fall with this group based on many analyses. Because type species are falling outside of their normal sections in some cases, there may be some major name changes to be made.
Carex acutiformis has always been a morphological oddity in this respect.
section Secalinae has a long branch, and it moves around in every analysis.
Sylvaticae / Rhynchocystis / Ceratocystis group comes out in the cpDNA trees as well. Additionally, the Anomalae group occurs close to the core North American Hymenochlaenae as well. These are morphologically similar even if they are distant geographically. There are a lot of sections in there, but the morphological coherence of the clade that spans Carex scabriuscula to Carex boliviensis. A lot of the Mexican species fall close to the North American spp.
Tony and Marcia
the geographic signal in this tree is striking.
the placement of Carex rivulorum and C. simulans is odd, moving around. (Santi: These individuals have both ITS and ETS. We don't have specimen data for these.) Hangzhouenses was published as being a distinct section, so its placement with Rhomboidales is odd. On what grounds did this name get changed to Carex rivulorum? (Pedro: World Checklist).

Notes from Cyperus

Is there already a full taxonomic / nomenclatural list for Carex?
It's complicated, because there is not index to these names. For FNA, I did what I could on the sectional names, but may have tossed those files when we downsized our offices. I don't think, however, that the problem is that hard: we have a lot done for North America, Egorova did a lot for Asia, and we can rebuild from there without starting from scratch. I think that because sections have been used so much in Carex, there is a lot of data available. A lot of the sections, however, have not been typified. If we have choices about typification, we can be smart about it: we have some opportunities there.
This is correct: many of the names need to be typified. Many old works contain sectional names that might confuse the current nomenclatural scheme we have: we will have to consider changing names or writing proposals to conserve names.
Watch out for illegitimate names: in Cyperus, many section names were published not at the section level and are illegitimate.
I started a spreadsheet...
... we also have one, which might have been built off of yours.
How many of these names are getting into the World Checklist?
I'm in contact with him.

Parting thoughts and ideas

If anyone is interested in discussing a particular clade or section, it would be great to bring photos
if you find the placement of a species odd, we can easily find the placement of your sequence on the tree if you bring it.
Personal tools