Week 7 Laboratory
Phylogenetic Analysis of Asterid Eudicots
<Lab Schedule> <Next Lab>
Families sampled Sampling characters Character list Cladistic analysis Phenetic analysis Report
This laboratory exercise involves a phylogenetic analysis of the Asterid eudicots. The asterids are a fairly natural group of eudicots best identified by the fusion of petals to form a corolla tube. The aims of this exercise are (1) to familiarize yourself with the procedures of character selection and data collection; i.e., find and score as many variable characters as possible among the representative asterid families; (2) to allow you to use computer programs that can analyze such data with both phenetic and cladistic methods to determine relationships among the asterid families; and (3) to give you the opportunity to make taxonomic decisions based on your results. See lecture notes and handouts for review.
A. Operational Taxonomic Units (OTUs or families investigated)
Thirteen genera representing 13 different families of Asterids, and one genus representing the Rosids (to be used as the outgroups) are provided at each table or as demonstrations at the front desk (see Plant Systematics, 2nd ed. [T], Zomlefer[Z] and Cronquist [C] for information on each family/genus). Many of the genera are tropical or subtropical so you will not find them in Gleason & Cronquist. For these (and others), consult Cronquist's An Integrated System of Classification of Flowering Plants [C]; and Heywood's Flowering Plants of the World at the front. Although not providing information about these genera, these references will provide considerable detail concerning the families. For additional information on these families, go to the family links in the University of Wisconsin Plant Systematics Collection homepage. The recent ordinal molecular phylogeny of Asterids is largely reflected in the APG system of classification shown below (see also the asterid family molecular phylogeny).
Asterid Eudicots Order Gentianales Gentianaceae 1. Exacum 397 [T] 244 [Z] 871 [C] Apocynaceae 2. Catharanthus 394 [T] 239 [Z] 876 [C] Rubiaceae 3. Pentas 397 [T] 236 [Z] 995 [C] Order Solanales Solanaceae 4. Nicotiana 416 [T] 213 [Z] 892 [C] Convolvulaceae 5. Evolvulus 412 [T] 215 [Z] 895 [C] Order Lamiales Bignoniaceae 6. Tecomaria 402 [T] 248 [Z] 968 [C] Lamiaceae 7. Salvia 402 [T] 265 [Z] 924 [C] Verbenaceae 8. Lantana 412 [T] 262 [Z] 920 [C] 9. Antirrhinum 410 [T] 250 [Z*] 951 [C*] Acanthaceae 10. Justicia 400 [T] 259 [Z] 963 [C] Order Asterales Campanulaceae/(Lobeliaceae) 11.Lobelia 433 [T] 211 [Z] 983 [C] Goodeniaceae 12. Scaevola 433 [T] --- 989 [C] Order Ericales Ericaceae 13. Rhododendron 378 [T] 77 [Z] 479 [C] Order Myrtales --- Rosid [for outgroup purposes] Onagraceae 14. Fuchsia 353 [T] 229 [Z] 645 [C]
B. Characters and Character States
Your group will examine the living plants of each family and look for vegetative, floral, and fruit characters that vary among the families. You have been provided already a list of characters; the character states that the fifteen families possess can be obtained by (1) working in groups; (2) examining the plants macroscopically; (3) examining the flowers/fruits microscopically; (4) looking at descriptions of the families and/or genera in your Text, Zomlefer, Cronquist, and other texts provided at the front of the lab; and (5) being persistent and inquisitive. Additional characters, if necessary, can be obtained from various literature sources (Text, Zomlefer, Cronquist, Heywood).
Only qualitative characters (presence/absence; number of sepals; color; shape; etc.) should be scored on the provided blank data matrix sheets. Please do not use quantitative characters that vary continuously (length/width of leaves; number of flowers; etc.). We will avoid these latter quantitative characters as they are difficult to use in a cladistic fashion, although they can be handled in phenetic analyses.
Make certain you are carefully distinguishing between characters and character states. Character states that can not be determined for certain families (lacking information totally e.g., do not have fruits or can not find the relevant information on fruits for the family; or inapplicable to that taxa e.g., shape of corolla tube in family without fusion of petals) should be treated as "?" in your matrix. Otherwise, give each character state an integer value (0 and 1 for binary character states; 0, 1, 2 for characters with 3 states; etc.).
C. Data Matrix
Once your group has completed the data matrix, enter the data into a file using the program MacClade on the computer in the lab. Open the "Phylogenetics Lab" folder in the center of the desktop, and drag the file "Asterids.template" on the desktop into the icon "MacClade". [If a message about fonts is encountered, just press "return".] When the file opens up, you can enter "0", "1", "2" etc. into the appropriate cell (columns are the characters, rows are the taxa). Use the arrow keys at bottom left to speed up the data entry procedure. You can make new columns for additional characters you want to include or delete present characters by going to the "Edit" menu at top (see TA for further instructions). After you have entered some of the data, go to "FILE" menu at top left and "SAVE FILE AS" your file as "Asterids.[your last name]" making sure that it is saved within the folder "Phylogenetics Lab". Save again when you are completely done entering the data. This completed data set will be used for both the phenetic and cladistic analyses.
D. Phenetic Analysis
Phenetic analysis will be done with both Neighbor-Joining (N-J) and UPGMA in the program "Phylogenetic Analysis Using Parsimony" (PAUP) on the laboratory computer. These are analyses that will derive a phenogram (tree) based on overall similarity among the families sampled. N-J and UPGMA use different algorithms to construct a phenogram and will each generate a single tree, likely different.
Open the “Phylogenetics Lab” folder and drag your file “Asterids.[your last name]” into the icon Paup in the left dock of the desktop. Then press the keys “apple” + “R” at the same time to run the program with your data set. N-J is performed by going to the “Analysis” menu at top, and clicking “Neighbor Joining/UPGMA”. All the default conditions are in place, so simply click “N-J” in the upper left “Algorithm” box, and then press the “OK” button. When the search is done, the single best tree will scroll through your screen. Repeat these steps with UPGMA selected rather than N-J.
To print the tree generated from each of these two analyses, go to the “Trees” menu at top, indicate “Print/View N-J Tree” (or UPGMA Tree), click on the “Print Trees” button. After printing, exit the “Print Trees” mode by clicking “Close”. Save the tree by going to the “Trees” menu at top and indicating “Save Trees to File” and name your treefile “Asterids[your group name]NJ.tre” (or UPGMA) and saving it to the “Phylogenetics Lab” folder.
E. Cladistic Analysis [see lecture notes for discussion of cladistics]
The PAUP computer program will be used to produce a cladogram of the Asterids based on shared derived character states using the assumption of maximum parsimony (MP). Cladistic analysis will probably generate more than one tree.
Open the “Phylogenetics Lab” folder and drag your file “Asterids[your last name].nex” into the icon Paup in the left dock of the desktop. Then press the keys “apple” + “R” at the same time to run the program with your data set. The outgroup Fuchsia is already indicated in the file. The basic cladistic run is performed by going to the “Analysis” menu at top, and indicating “Heuristic Search” - make sure that "Parsimony" is checked at the top before searching. All the default conditions are in place, so simply click the “Search” button. Alternatively, one can perform “Branch and Bound Search” under the “Analysis” menu. When the search is done, one or more trees will be found. Save the tree or trees by going to the “Trees” menu at top and indicating “Save Trees to File” and name your treefile “Asterids[your group name]MP.tre” and saving it to the “Phylogenetics Lab” folder.
If a single tree is found, go to the “Trees” menu, indicate “Print/View Trees”, click on the “Print Trees” button, and then exit the “Print Trees” mode by clicking “Close”. Prior to printing you may want to check the "Include title" button at top left and name the tree (e.g., MP Tree) using the Set Text option. If more than one tree was found, you can make a consensus tree by going to the “Trees” menu and indicating “Compute Consensus”. Follow the default conditions and click “OK”. To print out the consensus tree, go to the “Trees” menu and indicate “Print/View Consensus Tree(s)”. Additional analyses may be suggested by your lab instructor — specifically, successive weighting of characters in order to reduce the number of equally parsimonious trees.
F. Examination of Morphological Evolution using a Molecular Tree
The last step in the analysis will be to examine how well the morphological characters that you have scored fit (overlay) onto the molecular cladogram that forms the basis of the APGIII classification system. This tree is actually saved in the template file you used. This procedure will be demonstrated by your instructor and involves the MacClade program. Once your data set is opened in MacClade, you can select the molecular tree to visualize on the screen. Using the “trace” function in MacClade, you will be able to see how often each morphological character has evolved in asterids (some once, some many times, or with subsequent reversals).
Instructions for mapping your data set onto the APGIII or molecular trees
1. Open your datafile (Asterids[your group name].nex) in MacClade6. OPTIONAL: If you wish to save an image of the tree with a specific character "traced" onto the tree, go to "File" menu on top left, and select "Save Graphics File". You can then save the image as a PICT file using an appropriate name
2. Open the Tree Window (apple + T).
3. In Tree Window, go to “Tree” menu, and a tree named “APG” will open. This is a tree with the Asterid genera arranged according to DNA results.
4. Go to the “Trace” menu, choose “Trace Character.” The tree should now be colored, a little box at the lower right hand corner of the screen should show which character is currently being traced, with different colors showing the character states at different branches. Note: zebra striping means that the character is “equivocal” on that particular branch segment. This means that several character states are equally probable at this point.
5. To answer the question posed in the lab write up, you need to determine which of your characters show less homoplasy (on the APGIII tree) — i.e., are derived fewer times, and which show considerable homoplasy — i.e., are derived independently a number of times, or show repeated reversals back to the original character state.
Each student will work in a small group to obtain the original character state matrix and to do the cladistic and phenetic analyses. Each student, however, should work independently on the report and will hand in his/her own report in two weeks (due during lab of Oct. 28 – 30. This report will be part of your laboratory (20 pts.) and should include (1) the data matrix, (2) the N-J and UPGMA phenograms, (3) the cladogram or consensus cladogram (or additional tree derived from successive weighting - optional), and (4) at least 2 pages (typed, single spaced) on "Systematic Analysis".
Introduction: what were you attempting to answer or do in the lab
Data collection & analytical methods: briefly mention what these were
Data table with characters & states
Figures of NJ, UPGMA, Parsimony (optional Weighted Parsimony) trees
Compare trees to each other and to DNA tree
Do families group in their orders; are relationships among Asterids same as in DNA tree
What characters are good vs. bad (mapping on DNA tree in MacClade)
Would you propose different Asterid classification?
What limitations of this analysis do you see?