Hierarchical Structuring of Trait Data

1
Hierarchical Structuring of Trait Data

• Bot 940 Evidence for Evolution
• Eric Caldera

2
Scientific Method

1. Observation and description Organisms seem to
   have changed over time
2. Formulation of a Hypothesis Evolution by decent
   with modification / common ancestry
3. Predictions Biological traits should appear in a
   nested hierarchical structure groups within
   groups
4. Experimental test of predictions Do traits
   suspected to have arisen by decent with
   modification show a greater degree of
   hierarchical structure?

3
Nested hierarchical structure groups within
groups
4
Note that we dont see overlap across groups
Example there are no fungi with vascular
tissue, insects with four limbs or amphibians
with vascular tissueetc.
5
Why do we predict nested hierarchical structure?

• Only branching evolutionary processes are
  capable of generating nested hierarchical
  structure.
• For example, human languages, which have common
  ancestors and are derived by descent with
  modification, generally can be classified in
  objective nested hierarchies (Pei 1949 Ringe
  1999).

6

• Only certain things can be classified objectively
  in a consistent, unique nested hierarchy.
• The difference drawn here is between "subjective"
  and "objective.
• Anything can be grouped into hierarchies (for
  example, automobiles), but the importance of
  characters must be weighted subjectively.

7
Subjective grouping of automobiles
three wheels
three wheels
four wheels
A
B
four wheels
Characters
four wheels
Note that one tree is not more parsimonious over
the other. In tree A, the number of wheels is
subjectively weighted over color, and vice versa
in tree B.
three wheels
red
blue
8

• A cladistic analysis of automobiles will not
  produce a unique, consistent, well-supported tree
  that displays nested hierarchies.
• A cladistic analysis of automobiles (or any
  analysis of randomly assigned characters) will
  result in a phylogeny, but there will be a very
  large number of other phylogenies, many of them
  with very different topologies, that are as
  well-supported by the same data.
• Cladistic analysis of an actual genealogical
  process will produce one or a small amount of
  trees that are much more well-supported by the
  data than the other possible trees.

9

• The nested hierarchical organization contrasts
  with other possible biological patterns, such as
  the continuum of "the great chain of being"

10

• Mere similarity between organisms is not enough
  to support evolution.
• The nested classification pattern produced by a
  branching evolutionary process, such as common
  descent, is much more specific than simple
  similarity.

11
Is demonstrating that phylogenies show
hierarchical structure enough? How much nested
structuring is necessary to show that the
structure is non-random?
www.talkorigins.org
12
Testing for Hierarchical Structure

• Plylogenetic signal (the degree to which a
  phylogeny shows a unique well supported tree) can
  be quantified.
• We will discuss two methods the randomization
  test and the consistency index (CI). (Archie,
  1989 Klasssen et al. 1991)
• For additional tests see Faith and Cranston
  1991 Farris, 1989, Felsenstein 1985 Hillis
  1991 Hillis and Huelsenbeck 1992, Huelsenbeck et
  al. 2001

13
Archie, 1989

• Provides a test to determine whether the minimum
  length tree for a given dataset is significantly
  different from that expected from random data.

14
Archies Randomization test

• 1. randomize character data and perform the
  cladistic analysis
• 2. repeat this process to obtain a distribution
  of the minimum tree lengths for the randomized
  data
• 3. Test whether the minimum length tree generated
  from the real data is significantly smaller than
  the trees generated from randomized data

15
Vascular tissue Chloroplasts Water-tight egg Four limbs
bacteria 0 0 0 0
amphibians 0 0 0 1
humans 0 0 1 1
mammals 0 0 1 1
birds 0 0 1 1
reptiles 0 0 1 1
fishes 0 0 0 0
insects 0 0 0 0
fungi 0 0 0 0
mosses 0 1 0 0
ferns 1 1 0 0
flowering plants 1 1 0 0
Real data
Vascular tissue Chloroplasts Water-tight egg Four limbs
bacteria 1 0 0 1
amphibians 0 0 0 0
humans 0 1 1 0
mammals 0 0 0 0
birds 0 0 1 0
reptiles 0 1 0 1
fishes 0 0 1 0
insects 0 0 0 1
fungi 0 0 0 1
mosses 1 0 0 0
ferns 0 1 1 1
flowering plants 0 0 0 0
Randomized data
16
Minimum length tree from real data
Distribution of minimum length trees from
randomized dataset
17
Min tree from real data
Distribution of randomized data
Archie, 1998
18
Klassen et al. 1991

• Makes use of the consistency index (CI)
• CI represents the reciprocal of the number of
  steps per character
• The further from one, and the closer to zero
  increased homoplasy
• A problem with CI is that is seems to decrease as
  a function of number of characters and taxa

19

• Klassen et al generated CI distributions for
  random datasets of varying numbers of taxa and
  characters.
• The CI values for random data can then be
  compared to real data

20
Note that most real datasets are well above the
95 confidence interval for CI values, and in no
cases are they below the 95 limit.
Klassen et al. 1991
21
Klassen et al. 1991
22
Did GOD do it?

• How might you respond to the idea that nested
  hierarchal structure is seen because certain
  traits, by design, work better together?
• Is testing hierarchical structure against a
  random alternative sufficient. What about those
  that argue that god did not design things
  randomly

23

• Archie, J. W. (1989) "A randomization test for
  phylogenetic information in systematic data."
  Systematic Zoology 38 219-252.
• Faith, D. P., and Cranston, P. S. (1991) "Could a
  cladogram this short have arisen by chance
  alone? on permutation tests for cladistic
  structure." Cladistics 7 1-28.
• Farris, J. S. (1989) "The retention index and the
  rescaled consistency index." Cladistics
  5417-419.
• Felsenstein, J. (1985) "Confidence limits on
  phylogenies an approach using the bootstrap."
  Evolution 39 783-791.
• Hillis, D. M. (1991) "Discriminating between
  phylogenetic signal and random noise in DNA
  sequences." In Phylogenetic analysis of DNA
  sequences. pp. 278-294 M. M. Miyamoto and J.
  Cracraft, eds. New York Oxford University Press.
  
• Hillis, D. M., and Huelsenbeck, J. P. (1992)
  "Signal, noise, and reliability in molecular
  phylogenetic analyses." Journal of Heredity 83
  189- 195.
• Hillis, D. M., Moritz, C. and Mable, B. K. Eds.
  (1996) Molecular systematics. Sunderland, MA
  Sinauer Associates.
• Klassen, G. J., Mooi, R. D., and Locke, A. (1991)
  "Consistency indices and random data." Syst.
  Zool. 40446-457.
• Pei, M. (1949) The Story of Language.
  Philadelphia Lippincott.
• Ringe, D. (1999) "Language classification
  scientific and unscientific methods." in The
  Human Inheritance, ed. B. Sykes. Oxford Oxford
  University Press, pp. 45-74.