Phylogenetics Phylogenetics Phylogenetics is the estimation of the Phylogenetics is the estimation of the “tree” through “time” “tree” through “time” knowing only the “leaves”
Phylogenetics Phylogenetics However, the “leaves” are scattered over “space”. Some Additionally, many related “leaves” diverge in “form”, areas have related “leaves”, others have unrelated “leaves”. while other unrelated “leaves” converge in “form”. Thus, Thus, phylogenetics is compounded by issues of both “time” phylogenetics is compounded by issues of “time” and and “space”. “space” and “form”.
1 Phylogenetics Phylogenetics In natural and phylogenetic systems of classification, characters What characters are selected or even considered, has been very are selected a posteriori for their value in correlating with other subjective. Consider Cronquist and Dalghren with mustard oil characters to form hierarchical structure of groups families . . .
Phylogenetics Phylogenetics These first phylogenetic classifications were “phyletic” - In the 1960s, two main groups of systematists became involving a subjective selection of characters for classification dissatisfied with the phyletic approach and developed more objective methods: phenetic and cladistic
2 Phylogenetics Phenetics vs. Cladistics With the rise of molecular phylogenetics in the 1980s, additional approaches are now invoked (ML, Bayesian) - a continuum of models are now seen • Phenetics uses “overall • Cladistics uses only similarity” - all characters “phylogenetically used (“distance” approaches) informative” characters • species similarity (or • derived state is shared differences) often scaled by at least 2 but not all from 0 to 1 taxa - “shared derived character states”
Phenetics Phenetics UPGMA cluster analysis Data Matrix Data Matrix taxa • convert data matrix into characters pair-wise matrix based on sympetaly bicarpellate apocarpy epipetaly trees epigyny poll. beetle tepals heterostyly states vessels overall similarity
1. Magnolia + + - - + - + + - - 1. Magnolia + + - - + - + + - - 2. Nymphaea - + - - - - + + - - 2. Nymphaea - + - - - - + + - - 3. Rosa + + ------3. Rosa + + ------4. Primula + - + + - - - - - + 4. Primula + - + + - - - - - + 5. Gentiana + - + + - - - - + - 5. Gentiana + - + + - - - - + - 6. Aster + - + + - + - - + - 6. Aster + - + + - + - - + -
3 Phenetics Phenetics UPGMA cluster analysis UPGMA cluster analysis • convert data matrix into • reduce overall similarity pair-wise matrix based on matrix by clustering together overall similarity Gentiana and Aster and recalculate similarity values • identify most similar pair of taxa and cluster them • identify most similar pair of taxa and cluster them Magnolia Nymphaea Aster Gentiana
90% 80%
Phenetics Phenetics UPGMA cluster analysis UPGMA cluster analysis • reduce overall similarity • cluster together Gentiana, matrix by clustering together Aster, and Primula and Magnolia and Nymphaea and recalculate values recalculate similarity values • identify most similar pair of • identify most similar pair of taxa and cluster them taxa and cluster them Nymphaea Aster Magnolia Gentiana Rosa Primula
75% 70%
4 Phenetics Phenetics UPGMA cluster analysis • many different methods based on similarity or differences (including multiple components, ordination, etc.) • cluster together Magnolia, Nymphaea, and Rosa and • in lab you will be using UPGMA & Neighbor-joining recalculate values using a computer program PAUP
• cluster the two remaining larger groups at 42.5% to make final phenogram
Cladistics Cladistics How do you analyze this same data based on cladistics - “shared Issue #1- How do you determine what is derived? derived character states”? vessels apocarpy sympetaly epipetaly trees epigyny poll. beetle tepals bicarpellate heterostyly vessels apocarpy sympetaly epipetaly trees epigyny poll. beetle tepals bicarpellate heterostyly 1. Magnolia + + - - + - + + - - 1. Magnolia + + - - + - + + - - 2. Nymphaea - + - - - - + + - - 2. Nymphaea - + - - - - + + - - 3. Rosa + + ------3. Rosa + + ------4. Primula + - + + - - - - - + 4. Primula + - + + - - - - - + 5. Gentiana + - + + - - - - + - 5. Gentiana + - + + - - - - + - 6. Aster + - + + - + - - + - 6. Aster + - + + - + - - + -
5 Cladistics Cladistics Issue #1- ordering or polarizing character states (primitive or derived) Issue #1- ordering or polarizing character states (primitive or derived)
vessels (+) → no vessels (-) OR eterostyly vessels noapocarpy vesselssympetaly (epipetaly -) → trees vesselsepigyny (+)poll. beetle tepals bicarpellate 1. Magnolia + + - - + - + + - - 2. Nymphaea - + • -can be- subjective- - + + - - 3. Rosa + + • -fossil- record------4. Primula + - • +development,+ - ontogeny- - - - + 5. Gentiana + - • +look +at groups- most- closely- - related+ - 6. Aster + - to+ your+ group- of +interest- (outgroup- + ) -
Cladistics Cladistics Issue #1- ordering or polarizing character states (primitive or derived) Issue #1- ordering or polarizing character states (primitive or derived) plesiomorph - primitive state apomorph - derived state Blue flowers = synapomorph - shared derived state sister group
- use outgroups
? ? ?
6 Cladistics Cladistics Add in Amborella as sister outgroup to rest of angiosperms Convert data matrix to “0” “ ”
& 1 sympetaly bicarpellate heterostyly apocarpy epipetaly trees epigyny poll. beetle tepals vessels Amborella - + - - + - + + - - vessels apocarpy sympetaly epipetaly trees epigyny poll. beetle tepals bicarpellate heterostyly 1. Magnolia + + - - + - + + - - 1. Magnolia + + - - + - + + - - 2. Nymphaea - + - - - - + + - - 2. Nymphaea - + - - - - + + - - 3. Rosa + + ------3. Rosa + + ------4. Primula + - + + - - - - - + 4. Primula + - + + - - - - - + 5. Gentiana + - + + - - - - + - 5. Gentiana + - + + - - - - + - 6. Aster + - + + - + - - + - 6. Aster + - + + - + - - + - e.g., “-” to “0” and “+” to “1”
Cladistics Cladistics
Convert data Convert data matrix to “0” matrix to “0” “ ” “ ”
& 1 sympetaly bicarpellate heterostyly & 1 sympetaly bicarpellate heterostyly apocarpy epipetaly trees epigyny poll. beetle tepals apocarpy epipetaly trees epigyny poll. beetle tepals vessels vessels Amborella 0 1 0 0 1 0 1 1 0 0 Amborella - + - - + - + + - - 1. Magnolia 1 1 0 0 1 0 1 1 0 0 1. Magnolia + + - - + - + + - - 2. Nymphaea 0 1 0 0 0 0 1 1 0 0 2. Nymphaea - + - - - - + + - - 3. Rosa 1 1 0 0 0 0 0 0 0 0 3. Rosa + + ------4. Primula 1 0 1 1 0 0 0 0 0 1 4. Primula + - + + - - - - - + 5. Gentiana 1 0 1 1 0 0 0 0 1 0 5. Gentiana + - + + - - - - + - 6. Aster 1 0 1 1 0 1 0 0 1 0 6. Aster + - + + - + - - + - e.g., “-” to “0” and “+” to “1” e.g., Amborella state (either “-” or “+”) to “0”
7 Cladistics - Cladistics - - - other other ------Primitive 0 Primitive 0 - - - - Derived 1 Derived 1 Tepals Tepals peals sepals + Tepals peals sepals + No vessels vessels Polypetaly sympetaly Free stamens epipetaly Trees herbs Hypogyny epigyny beetle poll. poll. carpels Various bicarpellate Homostyly heterostyly No vessels vessels Polypetaly sympetaly Free stamens epipetaly Trees herbs Hypogyny epigyny beetle poll. poll. carpels Various bicarpellate Homostyly heterostyly Apocarpy syncarpy Apocarpy syncarpy Amborella 0 0 0 0 0 0 0 0 0 0 Amborella 0 0 0 0 0 0 0 0 0 0 1. Magnolia 1 0 0 0 0 0 0 0 0 0 1. Magnolia 1 0 0 0 0 0 0 0 0 0 2. Nymphaea 0 0 0 0 1 0 0 0 0 0 2. Nymphaea 0 0 0 0 1 0 0 0 0 0 3. Rosa 1 0 0 0 1 0 1 1 0 0 3. Rosa 1 0 0 0 1 0 1 1 0 0 4. Primula 1 1 1 1 1 0 1 1 0 1 4. Primula 1 1 1 1 1 0 1 1 0 1 5. Gentiana 1 1 1 1 1 0 1 1 1 0 5. Gentiana 1 1 1 1 1 0 1 1 1 0 6. Aster 1 1 1 1 1 1 1 1 1 0 6. Aster 1 1 1 1 1 1 1 1 1 0 e.g., Amborella state (either “-” or “+”) to “0” “shared derived” character states
Cladistics - Cladistics - other - - - - -
- Issue #2 - how do you select the “best” tree? Note: 2 - uniformative - • with 3 ingroup species and one outgroup (*), there are characters Tepals Tepals peals sepals + No vessels vessels Polypetaly sympetaly Free stamens epipetaly Trees herbs Hypogyny epigyny beetle poll. poll. carpels Various bicarpellate Homostyly heterostyly Apocarpy syncarpy 3 trees possible Amborella 0 0 0 0 0 0 0 0 0 0
1. Magnolia 1 0 0 0 0 0 0 0 0 0 * 1 2 3 * 2 1 3 * 3 1 2 2. Nymphaea 0 0 0 0 1 0 0 0 0 0 3. Rosa 1 0 0 0 1 0 1 1 0 0 4. Primula 1 1 1 1 1 0 1 1 0 1 5. Gentiana 1 1 1 1 1 0 1 1 1 0 6. Aster 1 1 1 1 1 1 1 1 1 0 not “shared derived” character states
8 Cladistics Cladistics Issue #2 - how do you select the “best” tree? Issue #2 - how do you select the “best” tree?
• estimation procedure, however, usually involves vast • estimation procedure, however, usually involves vast number of possible “trees” number of possible “trees” • this study with 7 taxa - there are 10,395 possible tree topologies • for a study with 50 taxa - there are 3 X 10 74 possible 79 • examining all trees is possible here, but with larger trees or approaching number of atoms in universe (10 )! numbers of taxa (as the 14 taxa used in lab this week – • landmark paper in 1993 for angiosperms had 499 taxa - 7.9 trillion trees!) a heuristic approach is required astronomical number of possible trees! >> 10 1000 • for a study of the Tree of Life - 10 70,000,000
Cladistics Cladistics Issue #2 - how do you select the “best” tree? Issue #2 - how do you select the “best” tree?
• the “best” tree is dependent on assumption of an • in the context of evolution, maximum parsimony = optimality criterion: e.g., likelihood, parsimony choosing the tree that requires the fewest number of evolutionary changes (apomorphies) • cladistics (morphology) often uses parsimony - based on “Ockham’s Razor” • choose the tree with the least amount of homoplasy - convergences or reversals or character conflict William of Ockham – Entia non sunt multiplicanda praeter necessitatem or • choose the shortest, simplest, most “Entities should not be multiplied efficient tree unnecessarily”
9 Cladistics Cladistics - - other - - - - -
Issue #2 - how do you select the “best” tree? - - - Tepals Tepals peals sepals + No vessels vessels Polypetaly sympetaly Free stamens epipetaly Trees herbs Hypogyny epigyny beetle poll. poll. carpels Various bicarpellate Homostyly heterostyly Apocarpy syncarpy Amborella 1. Magnolia 2. Nymphaea • use maximum parsimony to find http://evolution.genetics.washington.edu/phylip/software.html the “best” of 10,395 possible trees 3. Rosa • 36 of the around 370 phylogenetic software programs available! 4. Primula • many can be used on about 50 free web servers (including 5. Gentiana supercomputers or tera-grids) 6. Aster • in lab we will use two programs
Cladistics - Cladistics - other - - - -
- Consistency Index = 0.91
- • two trees are equally parsimonious 2 characters - # changes minimally expected
- • with character conflict, each is 11 conflict! steps long and not the expected 10 # changes occurred on tree Tepals Tepals peals sepals + No vessels vessels Polypetaly sympetaly Free stamens epipetaly Trees herbs Hypogyny epigyny beetle poll. poll. carpels Various bicarpellate Homostyly heterostyly Apocarpy syncarpy Amborella 0 0 0 0 0 0 0 0 0 0 1. Magnolia 1 0 0 0 0 0 0 0 0 0 2. Nymphaea 0 0 0 0 1 0 0 0 0 0 3. Rosa 1 0 0 0 1 0 1 1 0 0 4. Primula 1 1 1 1 1 0 1 1 0 1 5. Gentiana 1 1 1 1 1 0 1 1 1 0 6. Aster 1 1 1 1 1 1 1 1 1 0
10 Cladistics Cladistics
• tree 1 has vessels as • tree 2 has herbs as synapomorphy for all taxa synapomorphy for all taxa except outgroup + except outgroup + Magnolia Nymphaea • habit shows homoplasy • vessels shows homoplasy Þ trees Þ no vessels (“messy”) with an origin to (“messy”) with an origin to herb and then reversal back vessels and then reversal to tree back to vessel-less Þ vessels Þ herbs Tree 1 Þ herbs Tree 2 Þ vessels
Cladistics Phylogenetic Analysis of Asterids • a consensus tree depicts the maximum information possible from all most parsimonious trees (note: not equal to phenogram) 1. Data set for 13 asterids and one rosid outgroup report should include data set (characters/states)
11 Phylogenetic Analysis of Asterids Phylogenetic Analysis of Asterids
1. Data set for 13 asterids and one rosid outgroup 4. Mapping of characters onto DNA tree in MacClade report should include data set (characters/states) report should include discussion of good vs. bad characters (homoplasy) 2. Distance (phenetic) approach in PAUP – two different ones
report should include UPGMA and NJ trees 5. Phylogenetics and classification
3. Parsimony (cladistic) approach in PAUP report should include discussion of how asterids are or should be classified based on YOUR data report should include strict consensus tree (# trees) report optionally include strict consensus tree after weighting characters
Cladistics Cladistics Issue #3 - how do you construct a classification? Issue #3 - how do you construct a classification? • most cladists advocate monophyletic groupings only - stressing primacy of descent Milkweeds are a highly • a different group, evolutionary taxonomists, allow for derived lineage from paraphyletic grouping - stressing both descent and modification within the dogbane family - Apocynaceae
• recognizing Asclepiadaceae makes the Apocynaceae paraphyletic
• some agree since Asclepiadaceae are so divergent
12 Cladistics Cladistics Issue #3 - how do you construct a classification? Paraphyletic species - 3 options: 1. Recognize both the derived (apo) species and the Recognize paraphyletic paraphyletic ancestral (plesio) species - 2 species species?
• island or peripheral geographic speciation is a common model in plants
• ancestral species becomes paraphyletic, new Lisianthius in central peripheral isolates: new species monophyletic Panamanian cloud species forms at edge of forests retained ancestral species
Cladistics Cladistics Paraphyletic species - 3 options: Paraphyletic species - 3 options: 2. Recognize the derived (apo) species and monophyletic 3. Recognize only one monophyletic species units from the ancestral (plesio) species
13