Deep Time Project, A Comprehensive Phylogenetic Tree of Living and Fossil Angiosperms


Comments and questions: Dr. Doug Soltis

Copyright © FLMNH. This site is maintained by the FLMNH.

Last modified: 05/10/2002


New Mexico Meeting




A key theme of our initial discussions was the importance of coming up with something solid, or concrete at the end of 5 years, rather than simply a series of meetings. A major topic of discussion at this initial meeting was the construction of an interactive website and construction of a comprehensive morphological data base for angiosperm fossils; this data base could be expanded to include living taxa. A format for construction of such a web-accessed data base was developed; this concept is discussed below. Although a comprehensive data base would be beyond the scope of the Deep Time project itself, an initial effort would focus on a list of "priority" fossils.

Opportunities for funding were also discussed that would enable us to expand the data base for fossils into a more comprehensive format. Also discussed was the importance of using a small, well-defined clade of angiosperms as "test case" for experimentation in construction of a data base, as well as the issues involved in the integration of fossils into a data set of extant taxa. More details from the meeting are provided below.

Web site - general

  • It should be interactive (readers will be able to offer comments on the taxa and characters entered)
  • General editors and subeditors will be required for each group
  • There are several potential models (IPNI was mentioned; we may want to have Deep Time participants attend the next meeting of this NSF funded venture).
  • Deep Time should take a leading role to ensure that there is one, central data basing project for angiosperms.
  • There are also several paleo-data basing projects and we will need to interact with these groups (Richard Lupia may be able to take the lead on contacting these groups)

    Data base project

  • Should be able to search by taxon or character; perhaps also by age and distribution
  • Each entry contains fossil illustration, information about the fossil that is not readily available or apparent in the literature, a character data base, and a phylogenetic tree indicating presumed or approximate placement of the taxon Fossil Priority list

The need for developing and placing on the web site (using the format noted above) a group of priority fossils was discussed. What is meant by a "priority" fossil was discussed. Priority fossils may involve more than one organ, but only one organ may be suitable in some instances (if it displays a particularly critical or diagnostic character). Important fossils of the Lower Cretaceous were noted. Upper Cretaceous fossils of interest to group members may also be included. Juglandaceae, Fagales, particularly if these groups are to be used as initial models. Ultimately, all good fossils would be considered. A lower priority list could also be established --

  • Name of fossil
  • Age range (stratigraphy/years)
  • Quality ∑ Where it occurs now, as well as paleo latitude and longitude
  • key words, descriptions
  • Images (that can be measured, as in the example of the Digital Dragonfly)
  • Probable relationships
  • Annotation—other interpretations? (Interactive)
  • Other information Morphological matrix for extant taxa

    This topic was discussed further by a group of participants over dinner following our initial meeting. Walter Judd has compiled a summary of this meeting, and this summary is provided below. Construction of a Morphological Data Matrix (in connection with "Deep Time" project; based on notes taken on August 16, 2001, in meeting of subgroup interested in this issue):


  1. Get group of taxonomic experts together (including those familiar with particular taxonomic groups—extinct and extant—and pattern of variation within particular characters).
  2. Decide on potential characters to be assessed (based on a tentative list).At this point detailed procedures for recording/measuring each character would be determined, including the use of drawings/digital images. The recording of morphological s.l. information would constitute the metadata upon which an eventual data matrix (or matrices) would be based. This data should be as "transparent" as possible, as it will serve as the basis of subsequent decisions regarding state delimitation, which may be various, depending upon taxonomic judgement and the scope of later analyses. [These issues were much discussed, with some participants in our meeting suggesting that character states could be determined at this point, based on the experience of the taxonomic experts, while others believed that such decisions should only be made after the pattern of morphological variation was more thoroughly assessed.]
  3. At this point, individuals would be assigned various taxa and/or characters to observe and record, on an on-line database (which should have the ability to include images/drawings). Actual recording of data from specimens would take place at the individuals home institutions; some loans of herbarium specimens likely would be required.
  4. Taxonomic experts meet again to assess pattern of variation (i.e., review electronic database or printed "data forms") and determine characters (and character states) to be used in phylogenetic analyses. Criteria used in state delimitation determined at this point. Coding decisions made at this point. Evaluation of various characters, and some included in matrix, others excluded.
  5. Construction of matrix (see also issues relating to taxa).

    The matrix could be used in various ways, e.g., morphological phylogenetic analysis, combined (molecular + morphological) analysis, mapping morphological characters onto a molecular-based tree, etc.


  6. Species used as terminals
  7. Whenever possible, species that have served as the source of DNA data should be used as the source of morphological s.l. data. [However, if a species is considered by experts to be inappropriate—due to numerous morphological specializations—a species better representing a higher taxon could be suggested, and new sequences, etc. acquired.]
  8. Morphological data from cited voucher specimens.
  9. Appropriate taxon selection. Consult experts; higher taxa considered to be potentially paraphyletic or polyphyletic would be more highly sampled than those likely monophyletic.
  10. Fossils to be include would be those most completely known, or with diagnostic synapomorphies.
  11. Construct morphological data matrix for selected species (see procedures under characters).

Goals of Project:

Goals need to be clearly articulated. Some were suggested, as follows:

  1. Incorporate fossils into phylogenies based on morphology, DNA, or combined data.
  2. Study character change in connection with phylogenetic hypotheses.
  3. Construct hypotheses of phylogenetic relationships based on morphological and molecular data.
  4. Increase confidence in taxonomic groupings (through assessment of congruence of molecular and molecular phylogenies).
  5. Better assess homology.
  6. Determine morphological synapomorphies (of clades seen in various molecular trees)
It was noted that the goals will (in part) determine what kind of data are needed. [Some discussion was given to the question of what would be an appropriate "test case" on which to apply these methods. Various taxa were discussed, including Fagales, Ranunculales, Ericales, Saxifragales, Cornales, and Proteales. Criteria for taxon selection were also discussed.] Other items available on the website
  • The most current, comprehensive angiosperm summary tree
  • The Deep Green angiosperm "click on" tree may be useful
  • Summary information for molecular biologists and others that provides general information regarding the best estimate of the age of various major groups of angiosperms, as well as a general discussion of the difficulties inherent in using a molecular clock to provide such estimates.

    Molecular biologists as well as scientists in other disciplines (as well as teachers) need access to such information. With so many possible dates with highly diverse ages presented in the literature, it is important for Deep Time to take the lead on this

    Homework assignments

  • Contact appropriate individuals involved in the several paleo-data basing projects so that we can better interact with these groups (Richard Lupia) ∑ Develop fossil "priority" list (Jim Doyle, David Taylor, Pat Herendeen, David Dilcher)
  • Find out more about the IPNI next meeting (in Greece?) and make plans for a group from Deep Time to attend (Peter Stevens) ∑ Discuss possible focus groups (Fagales? Juglandaceae? Ericales?)
  • Invite additional appropriate individuals to Feb meeting (e.g., Paul Manos) ∑ Initial development of website (Doug and Pam Soltis) ∑ Estimated time of divergence of major angiosperm clades (with help of S. Magallon)
  • List of fossil priority taxa (with help of Jim Doyle, David Taylor, Pat Herendeen, David Dilcher)
  • Discussion of approach to be used in construction of morphological data base for extant taxa (Judd, Stevens, Kron, et al.) ∑ Approach NSF about funding possibilties for a morphological data base project (Judd, Soltis)
  • Determine possible sources of funding to develop support for a large fossil data base project (Herendeen, Dilcher, others?)
  • K-12 outreach needed for BSA meetings in Madison, August, 2002 (P. Soltis, D. Soltis, Dilcher, Herendeen)
  • Advertise phyhlogenetics workshops, short courses for students (Deep Time will support several students per year) (P. Soltis, D. Soltis, Dilcher, Herendeen)
  • Ideas for developing a possible paleobotany short course (Dilcher, Herendeen)


Deep Time: Revised Agenda for First Meeting


8:30 Introductions of PIs and other participants in the project

8:45 Have others attending introduce themselves

9:00 Background on Deep Time and the RCN program; mention Deep Gene

9:15 Goals of the project; discuss four major research areas

9:30 Development of Deep Time website; Virtual Fossil Project

9:45 Announce support available, date for future meetings (particularly the proposed February meeting in Florida, Feb. 22, 23, 2002)

10:00 Describe Groups and announce group leaders (we will initially break up into groups specified by goals) and allow people to decide group(s) they want to participate in during the morning break

10:15 Break; Morning coffee

10:30 Groups begin to meet with a group leader for each group

12:00 Box Lunch

1:00 Discuss integration of fossils into the angiosperm tree and other aspects of the Deep Time project as a single group

3:00 Break; Afternoon coffee, coke, snacks

3:30 Meet with Deep Gene

5:00 End Discussions

7:00 Dinner with Deep Gene

Deep Time Home | Projects | Meetings | News and Info | Goals | Links | Site Map | FLMNH