Systematics and Phylogenetic Inference
 Home
 Course Info
 Course
Material
Picture

© 1997
David H.A. Fitch
all rights reserved

Click on the topic you would like to review:

Rule

Lecture notes

Systematics

I.  Introduction
A.  Organisms share a unique history because of heredity
1.  That is, all life arose from a single ancestor
2.  Life on this planet is unique:  if life also originated elsewhere, it would be uniquely different from life on earth (e.g., more different from us than we are from prokaryotes)

B.  So far, we have been concerned largely with mechanisms
1.  At the phenotypic level (e.g., adaptations as a result of the natural selection "algorithm")
2.  At the genetic level (evolution as changing allele frequencies with regard to drift and selection)

C.  But the other task is to understand the actual history of evolution (i.e., the task of systematics, which is fundamental to nearly all evolutionary analysis)

II.  Definitions

A.  CHANGE in evolutionary lineages
1.  ANAGENESIS:  directional, evolutionary change within a single lineage
2.  CLADOGENESIS:  branching (divergence) by speciation (may be confused by "reticulate evolution"; e.g., by hybridization or other "horizontal transfer")
3.  PHYLOGENY:  a tree-like representation of evolutionary history (cladogenetic and often anagenetic)

B.  CHARACTERS and their changes:  features of organisms
1.  HOMOLOGOUS:  the "same" character in different organisms (shared because of common ancestry); the important characters for inferring the history of cladogenesis!
Note:  The definition (or inference) of homology is not to be confused with criteria for suggesting a homology hypothesis:
a.  Positional equivalence (e.g., in a series of connected parts)
b.  "Special quality" (i.e., unique composition)
c.  Transitional series (e.g., existence of developmental transitions or fossil intermediates)
2.  ANALOGOUS:  similar characters in different organisms shared by convergence (not because of common ancestry)

C.  CHARACTER STATES:  forms of those features
1.  PLESIOMORPHY:  ancestral ("primitive") state (note that no extant species is primitive, although it may have primitive characters)
2.  APOMORPHY:  a shared ancestral change to a DERIVED state
3.  HOMOPLASY:  similar changes in different lineages that do not reflect common ancestral origin, but rather PARALLEL, REVERSAL, or CONVERGENT change

D.  TAXA and their organization
1.  TAXON:  set of organisms that fill a category (species, genus, family, order, class, phylum, kingdom, etc.); OTU ("operational taxonomic unit")
2.  BINOMEN (not binomial!):  designation for a species-level taxon that includes a genus-level name
3.  SYSTEMATICS:  organizing TAXA according to relationships and similarities; now generally refers to the inference or estimation of phylogeny
4.  TAXONOMY:  Defining TAXA by descriptions of new species and the application of systematics
5.  CLADE or MONOPHYLETIC taxon:  entire group is descended from a single ancestral species (taxon)
6.  PARAPHYLETIC taxa:  does not include all the descendents of an ancestor (but usually belong to similar GRADE); 2 reasons for this kind of classification:
a.  Similar because of parallel evolution in closely related lineages
b.  Some members have diverged so much as to define a new group
7.  POLYPHYLETIC taxa:  composed of descendents of unrelated ancestors that evolved similar features by CONVERGENCE (generally avoided in modern classifications wherever possible)

(Return to top of page.)

III.  Schools of taxonomy (general tenet:  classification should somehow reflect evolutionary history)
A.  PHENETICS (Diagram of classification = phenogram)
1.  Classifications based only on overall similarity among species (e.g., GRADES)
2.  Does not necessarily reflect cladogenetic history if...
a.  The same character arises independently in two lineages (HOMOPLASY), or
b.  Different lineages show very different rates of evolution

B.  CLADISTICS (Willi Hennig, 1979) (Diagram of classification = cladogram)
1.  Each taxon should be monophyletic to reflect common ancestry (i.e., cladogenetic history)
2.  Only APOMORPHIC changes are informative about the history of cladogenesis and thus for classification
3.  Does not necessarily reflect similarity between closely related taxa or degree of divergence of some groups with novel features

C.  TRADITIONAL ("evolutionary") classification system (advocate:  Ernst Mayr):  monophyletic and paraphyletic taxa allowed to reflect both cladogenesis and the marked differences (e.g., novel features) of some groups (e.g., Aves)

D.  Ambiguities for all systems
1.  Absolute rank of taxa is rather arbitrary (genera, families, etc.) (sound like Darwin?)
2.  Extinct species are often placed in the same taxa with extant species (i.e., rankings do not usually reflect geological times)

(Return to top of page.)

IV.  Phylogenetic inference (i.e., the estimation of history through the proposition and testing of phylogenetic hypotheses)
A.  Cladistic methods:  Parsimony
1.  Assumption (yes, parsimony does make assumptions!):  characters states will be shared more often because of shared ancestry than because of independent changes (homoplasy)
2.  Justifications:
a.  Changes are fairly rare (heredity works)
b.  Independent changes that are the same are expected to be rarer by the square of the probability of a single change
3.  Principle (criterion) of parsimony is then used to evaluate many different phylogenetic hypotheses (all if possible); the hypothesis that best explains the data is the one (or one of the set) that requires the least number of changes from common ancestral sequences

B.  Phenetic methods
1.  Assumptions:
a.  Evolutionary changes are additive over time (i.e., accumulate; a model for this accumulation is generally proposed)
b.  Rates of change in different lineages do not accumulate so much change that closest relatives are more different from each other than each is to another taxon.
2.  Measures of divergence (generally corrected for additive accumulation according to an evolutionary model) are used to estimate relationships according to an algorithm or used as criteria to evaluate many (if not all) possible phylogenetic hypotheses

(Return to top of page.)

V.  Difficulties in phylogenetic inference:  variability of evolutionary rate, mistaken homology, and homoplasy
A.  Variation in evolutionary rate (both for a particular character and between different characters) may result in incorrect phylogenetic inference (esp. if using phenetic methods)
1.  Rate varies for single character
2.  "Mosaic evolution" = different characters evolve at different rates (e.g., human phenotype vs. genome)
a.  DIVERGENT characters (rapidly diverging either in response to selection or because of genetic drift in the absence of selection)
b.  CONSERVATIVE characters (static either because they serve in a great variety of environments or because of DEVELOPMENTAL CANALIZATION (e.g., resulting from redundancy of developmental mechanism), or developmental or genomic "CONSTRAINT" (appropriate variation is either not generated or is selected against)

B.  Mistaken homology will likely result in incorrect phylogenetic inference; one way of circumventing this problem is to use suites of independent characters to define relationships, not single characters (assuming most of the characters defined are likely to be truly homologous)

C.  HOMOPLASY (refers to changes in character states):  incorrect assumptions that shared changes are due to shared ancestry (instead of homoplasy) might result in an incorrect phylogenetic inference); dealt with by:
1.  Using suites of characters (with the assumption that most changes will not be homoplastic; e.g., the main assumption of parsimony)
2.  Using characters with low probability of homoplasy (but what are these?  Not too conservative and restricted but not too divergent with superimposed transformations?  Complex and not simple, and thus likely to follow DOLLO'S "LAW"?)

(Return to top of page.)

VI.  Inferring direction of evolutionary change:  requires at least a cladogram and an ancestral state

A.  OUTGROUP often helps to define primitive state (does not necessarily mean that the primitive state is the one in the outgroup or that the primitive state is the one that is possessed by most of the taxa)

B.  Fossil record is often not helpful, or may even lead you astray

C.  Embryology
1.  Haeckel's (1866) Biogenetic "Law":  successive embryonic stages represent adult stages of ancestral forms (not correct; e.g., butterflies not derived from pupal ancestors)
2.  Von Baer's (1828) "Law":  early developmental stages tend to be more similar among related species than later stages; i.e., characteristics that differentiate taxa are additions onto a fundamentally similar developmental plan (later interpretation:  derived characters are those that show greater difference from the embryonic condition); violations occur (e.g., paedomorphosis)

D.  Thus, inferences about the direction of change are drawn largely from the pattern of distribution of features among extant taxa, and has nothing to do with adaptive significance

(Return to top of page.)

VII.  Molecular Systematics

A. Using molecules should be advantageous for 2 reasons:
1.  Closer to the actual level of heredity (especially if you use DNA sequences)
2.  The number of independently varying characters is huge

B.  Disadvantages:
1.  Homoplasy (esp. reversals) are likely to occur at a higher rate in nucleotide sequences than in morphological characters (fewer states possible, no order of change, observed difference often smaller than actual accumulation of changes)
2.  Homology among characters (esp. nucleotides) is sometimes not easily assessed (neighboring structures are simply more sequences)

C.  Kinds of data
1.  Character data (analyzed by cladistic methods)
a.  Sequences (nucleotides, amino acids)
b.  Restriction enzyme maps
2. Distance data (analyzed by phenetic methods)
a.  DNA-DNA hybridization
b.  Dissimilarity of allozyme frequencies ("genetic distance")
c.  Immunological Distance (relative affinities to antibodies)
d.  Pairwise differences between sequences

D.  Molecular clock:  opportunity to test the prediction of the neutral theory
1.  Data should provide an idea of the number of changes that have been fixed in particular lineages after they diverged from a common ancestor
2.  If paleontological data is known for that divergence date, rates can be assigned to particular lineages
3.  Applying this rate to other lineages, are the dates of divergence plausible?  (Relative rate test)

E.  Molecular phylogenies generally concur with morphological ones, but sometimes there are interesting conflicts

(Return to top of page.)

Rule

Exercises

  1. Shown below is a data set (i.e., matrix of character states) for 5 taxa (A-E) and 10 nucleotide characters (1-10).
 

1

2

3

4

5

6

7

8

9

10

A

A

C

C

C

G

G

G

T

T

A

B

A

A

C

C

G

G

G

T

T

A

C

A

A

A

C

G

G

G

T

T

T

D

A

G

A

A

A

A

G

T

T

T

E

T

T

A

A

A

A

A

A

A

A

  1. Calculate the matrix of pairwise dissimilarities (leave these values uncorrected for accumulated or "superimposed" changes).  Assuming that the least dissimilar taxa share the most recent common ancestor, assemble the phylogenetic tree that represents the best estimate of the evolutionary history of these taxa.
  2. What other assumptions have you made?  Why are (or why aren't) these assumptions justifiable?
  3. Draw all of the possible unrooted trees for 5 taxa.
  4. What is the least number of changes required for each of these phylogenetic hypotheses?  Which of the hypotheses is (are) most consistent with the parsimony criterion?
  5. What assumptions have you made in your parsimony analysis?  Why do you (or why don't you) think these are justifiable assumptions?
  6. On the most parsimonious tree(s), which of the characters show apomorphic changes?  Which of the characters show homoplastic changes?  Which of the characters show parallel changes?  reversals?  convergences?
  7. Which characters are informative about phylogenetic relationships?  Why?  Which are not?  Why not?
  8. Assume that taxon E is an outgroup to the rest of the taxa.  Draw a rooted version of a most parsimonious tree.  What are the directions of the evolutionary changes on this tree?  Can you determine with certainty the directions of all changes?  If not, why not?
  9. Now assume taxon A is an outgroup to the rest of the taxa.  Draw a rooted version of a most parsimonious tree.  What are the directions of the evolutionary changes on this tree now?

(Return to top of page.)

Rule

Simulations and software

Sorry.  None yet.

(Return to top of page.)

 Darwin  Adaptation  Evolutionary
Genetics  Speciation  Systematics  History of
Diversity  Biogeography  Origins of
Novelty  Molecular
Evolution  Coevolution  Human
Issues
[Darwin] [Adaptation] [Evolutionary Genetics] [Speciation] [Systematics] [History of Diversity] [Biogeography] [Origins of Novelty] [Molecular Evolution] [Coevolution] [Human Issues]
[Home] [Course Info] [Course Material]