GCG 2017 - Classification framework

From systematics

Jump to: navigation, search

This classification primarily reflects the 2016 GCG tree, with additional species by Santiago Martin-Bravo and Pedro Jimenez. For a copy of the tree, contact ahipp@mortonarb.org and request 'GCG.trees.2017-01-16.v2'.


Classification -- Dividing up core Carex

Goal: identify the clades that are unambiguous, then divvy up the clades that are left.

Small core Carex, sensu Ford and Starr

This clade is a train wreck. There is anything and everything in here morphologically. Strong geographic component. This group may need to be shelved for now. Potentially an informal group, but it may need more work before we can do anything with it. Bruce and Julian can take this.

C. cryptostachys and C. durieui

probably incertae sedis... but may belong elsewhere in other analyses. We need to look at other analyses on this. C. durieui was initially placed in Spirostachyae

Decorae / Indicae / Polystachyae / one Hymenochlaenae

This clade seems to be well supported. The big paniculately branched things are problematic... we've traditionally split Indicae / Polystachyae by OW / NW. It is probably a good informal group. The Rhomboidales / Hangzhouenses branch should be excluded based on Marcia's analyses. This group is being worked on Berit Gehrke now.


about as messy as it could be. Everything in the Himalayas get dumped into this section.

Clandenstinae of Himalayas / Radicales of SE Asia / spp of random sections

Pretty well supported clade. This could be a morphologically definable section. Almost everything in one section with additional orphans. At least for now, an informal group, though it may turn into a section. Julian: how does this fit with the Radicales? Pedro: the problem is that no one has sampled a lot of Radicales. Julian: there is more to discover from here. Eric: for now, let's leave it as an informal group. Q: does that C. radicalis make sense? Julian: that looks odd... it may not belong there. It is a very long branch, probably needs some work.


Pretty clean, with the inclusion of C. heterostachya. THis is a good section: Sebastian will take this group.

Acrocystis of N. Am / Lamprochlaenae

Pretty good group, though this is complicated by the fact that some analyses do put this clade with Acrocystis s.s., from Europe. It could be an expanded Acrocystis with more sampling, but for now keep as an informal group. Eric will take this group, and Bruce is happy to help.

Next big clade

Cluster of three clandestinae

Makes sense, but there are other Clandestinae are scattered around.

sprengelii to tricephala

this looks like a problem... these are probably artifactual

Aulocystis clade

Carex alajica was initially Aulocystis in Egorova, though here it is in Careyanae. This is probably best as an informal clade now, but it is mostly Aulocystis. Matthias will take this group.

Lamprochlaenae / Aulocystis / Hymenochlaenae / Thuringiacae

A "dog's breakfast"... but some geographic structure, maybe some misplaced things. Pedro and Santi will take this one.

Larger Hymenochlaenae group

Core Hymenochlaenae (with Carex cherokeensis)

Good section most likely -- Marcia will take this

Three groups

Mitratae group 1

Masaki and Hoshino have almost complete sampling of this group: Marcia contact Also Sangtae (student he knows)

Mitratae group 2

Masaki and Hoshino have almost complete sampling of this group: Marcia contact Also Sangtae (student he knows)

Acrocystis + Rhomboidales

This has the core of Acrocystis. It might make sense to keep the Acrocystis there as incertae sedis, then cleave off the Rhomboidales as something separate. Marcia will lead on this, talking with Jin Xiao-Feng

Carex fastigiata to C. johnstonii

Another messy clade -- we probably need more data on this before we consider it an informal group

Careyanae / Griseae / Granulares

Careyenane traditionally

more sharply trigonous perigynia. Always known to comprise two groups:

  1. (C. abscondiata -- C. digitalis) lacking red bases, having just a single infl per shoot;
  2. the rest (C. plantaginea -- C. careyana) have lateral buds producing infls as well as the terminal shoot, and red bases.

Differentiating from Granulares: the latter have rounded bases

Granulares traditional

pigmentation in the pergynia and perigynium shape, more distinct beak than in the Griseae

Griseae traditional

Griseae have sharply impressed nerves: this is the primary difference. Terete perigynia, unlike the Careyanae. Basal sheath color helps a bit: C. brysonii, C. hitchcockiana, C. asynchrona all have brown bases.

What holds them together?

all have a large proliferation of nerves (25-40 in addition to the main veins of the perigynium), which shows up again in sect. Laxiflorae. Does this character show up in any Asian group? [Pedro: probably not so strongly.] The pistillate scale is also lined up with the angle of the perigynium, which separates it from the Laxiflorae. All cespitose, more or less all forest spp. for the upper clade (Careyanae Granulares clade), all brown bases. Between the C. abscondita / C. digitalis clade and the C. crawei / C. glaucodes clade, many fewer perigynia in the former. Laxiflorae for the vein character and inflorescence shape might get mixed up with this, but there is probably a number of new characters that are faithful.

The vote and conversation following

The majority voted to recognized only one section, though some voted for 2, 3, or 4 sections. Julian notes that this subject is very well studied, there are lots of characters, the clade is always well supported, and it is well supported based on DNA data. What guides the decision? a combination of number of taxa and practical diagnosibility. If we decide this is a section, then you can write an artificial key within the section. Recognizing just one section is quite practical.

California Aulocystis

This clade is strongly supported, 5 spp sister to the Car / Gris / Granulares clade... probably this makes a good section, though it also probably needs a new name.

Paniceae / Laxiflorae / Bicolores

Probably oversplit, in part because of the Bicolores stigma number. THis is a straightforward group: Tony, Marcia and Eric, with Sebastian -- will bring in Rob

Next Aulocystis group, with Setigerae and Thuringiaca

This could go as one group or two... probably better to consider as two. Sebastian will take the Aulocystis group, Julian is happy to help on the Setigerae / Polystahyae clade. No lead on this latter group.

Chlorostachyae: C. capillaris and friends

This is probably a good section


Another clean group

Echinochlaenae / Spirostachyae

Marcia, Marcial, Kerry Ford

Hallerianae / Pictae / Clandenstinae

This is not very well supported, but it makes some morphological sense and is worth considering as one grouping, or it might end up being more than one. Split between Europe and North America

Carex meyeriana

This is a floater in our tree, but it goes with C. stylosa in other analyses

Carex / Vesicarieae / Paludosae / Lupulinae / Praelongae [Phacocystis]

Folks have long thought that things in the Vesicarieae and Paludosae (minus the type of Paludosae) and sect. Carex etc. are all quite close... there are many hybrids within and between the traditional sections, which is unusual in the genus. These clearly go together: how you split them up will be informed by the phylogeny, but they derfinitely belong together. This also has a global distribution, different than the Careyanae / Griseae / Granulares clade. Long beaks, often sheathless bracts. Each of the major clades in this group has representatives of all the major clades. Tony et al. started lumping them in FNA based on morphology alone, and this supports it. There are small morphological groups that make sense: Core Lupulinae together with the portions of Vesicarieae that it falls with; Praelongae.

The C. retrorsa in our tree has an odd position... specimen perhaps needs to be inspected.

Old Pseudocypereae (includes C. comosa, e.g.) falls sister to the core Lupulinae.

Why is this interesting? These sections have long been known to be closely related and to hybridize with each other. We certainly need better resolution about how these sections group together... this makes a very reasonable informal clade, because these things always fall together. We would not probably want to say these altogether form a single section, at least until we have more data, but as an informal clade it's nice. Tony: having them altogether as a section might be justifiable, though we still need to know more about the 2-stigma group (Praelongae); there is no obvious very clean 3 or 4 section solution here. Sebastian: the close relationship between core Lupulinae and former Pseudocypereae go together; why they don't go together in this tree is unclear. Eric: this is a case where individuals with better data should look in to investigate sectional boundaries further, to analyze further, add additional taxa and loci, look into morphological characters.

The other issue with this group is that rapid and young divergence means that we simply don't have the characters to resolve relationships. What the big tree can do in this case is alert us to the fact that we need to sample, for example, the Praelongae. There is no obvious alternative to one section here, pending additional data.

Provisional informal name: the bladder sedge clade: section Carex s.l. We do have to be a bit careful with this particular clade, because it has the type section (sect. Carex).

Based on the combination of analyses we have from elsewhere, including Marcia's many-gene phylogeny, the Praelongae should probably be pulled out.


Probably a clean clade. Carex vestita might go in this group per Marcia.

Molliculae / Paludosae / Porocystis / Hymenochlaenae group

Marcia with Sangtae Above there, the Phacocystis + Stylosae are unplaced for now

Gracillima group: Porocystis / Hymenochlaenae / Longicaules

Marcia's group


Pedro and Tony


Leo and Marcia. The combination with a few Phacocystis needs to be considered along wtih these.


Probably a good section -- good small clade

Squarrosae + Shortianae

Probably a good section -- good small clade Bruce and Tony take this

Broader Phacocystis group

Pedro is leading on this

Podocarpa in, core Phacocystis

Pedro and Santi

Phacocystis: C. crinita, C. gynandra, C. torta, C. mitchelliana

Leo has been working on this for awhile.

Phacocystis: C. obnupta through C. macrochaeta

Classification - Siderosticta clade

Probably two sections, two clades. Julian and Bruce will talk with Yano and Hoshino

Classification - Caricoids

Two clades that often go together, but not always. As you add cpDNA data, you get these two clades going together.

Schoenoxiphium clade

Tamara, Berit, Modesto

Remaining unispicate clade

Uncinia, Kobresia, other unispicates... should this be broken up or taken as a group? Lead on this should be Julian, Shuren, Santi, Pedro Pedro: Junciformes, etc. from S. America Julian: planning to deal with the big Uncinia clade The predominantly American unispicate clades can mostly be sections... quite clean. The challenge comes up with the Kobresias. The clade that joins all the Uncinias with Schiedeanae, Nardinae, Filifoliae, etc. makes a good informal group that will break up easily into sections. This will be led by Julian and Tony and Santi, following up with others as needed. Kobresia: Shuren, Yano et al. From the current group, Marcia will contact to work on this group. For now, this is an informal group that needs work.

Classification - Vignea

Andrew, Eric, Bruce, Marcia, Enrique, Tony, Rob and others can work on this based on the current tree as well as the Vignea nuclear phylogeny, Bruce's AFLP data We would include Vignea in the framework paper, providing informal clade names Bruce: one strategy might be to take a broad brush things in the framework paper, don't hem ourselves in with the framework paper

Personal tools