The three-taxon statement and monophyly: Consider 3 taxa, A, B, and C. If evolution has occurred, then two of the tax must be more closely related to each other than either is to the third. That is, they must have shared a more recent common ancestor. This is a 3-taxon statement. A three-taxon statement can be drawn as a branching diagram, or cladogram, as shown below, with the two most closely related taxa joined into one clade, which is then joined with the clade representing the third taxon. The two most closely related taxa in a three-taxon statement are often called sister groups. In the diagram on page 5-11, taxa B and C would be considered sister groups.
The three-taxon statement can be applied at any level of the taxonomic hierarchy. For example, in the cladogram of the tetrapods, a three-taxon statement could be applied to the lizards +snakes, crocodilians, and birds, to display the internal relationships among these three related taxa, but these taxa could be considered one group in a three-taxon statement that considered the relationships among the turtles, mammals, and lizards-crocs-birds. An even larger three-taxon statement could be made about the turtles-mammals-lizards-crocs-birds as one group, and the amphibians and fishes as the other two groups.
Phylogenetic systematics seeks to identify monophyletic groups. That is, we want to recognize three taxon-statements based on recent ancestry, and do not want to link groups that are not closely related, nor exclude related groups. Two criteria that must be satisfied for monophyly:
1) All members of a monophyletic group must share a recent common ancestor, and
2) A monophyletic group must contain all derivatives of the most recent common ancestor as well as the ancestor.
Non-monophyletic groups include paraphyletic groups (whose members share a recent common ancestor but that do not include all members) and polyphyletic groups (whose members do not share a recent common ancestor). Note that it is not so important whether a group is para- or polyphyletic; it is much more important whether a group is monophyletic or not. Study the reptile cladogram on page 3 of the appendix to familiarize yourself with these principles.
Exercise: Referring back to the classification of the vertebrates, are the traditional classes Amphibia, Reptilia, Aves, and Mammalia monophyletic, especially if they are classified at the same rank? If they are not monophyletic, in what ways are they not? Are the groups that you named in your revised classification monophyletic? Why, or why not?
Characters and character states: How do we determine which two of the three taxa are most closely related? This is not something we know before we start, and it is not something we will know for sure when we finish. We use data on characters of the taxa to make an inference, or hypothesis, about the relationships among the three taxa. We define characters as the kinds of things we might look at in an organism (like the structure of an organ, the type of outer covering it has, etc.), and character states as the different forms that a character can take. For example, if the character was the type of outer covering in tetrapods, the character states might be moist skin with no other covering (as in amphibians), scales (as in reptiles), feathers (as in birds), and hair (as in mammals).
Exercise: In small groups, pick a group of organisms with which you are familiar (like fish, mammals, tetrapods, reptiles, birds, etc.), then decide on two or three characters you might use in classifying them, and what the character states might be.
Using characters to estimate relationships: During the course of evolution, characters undergo character state transformations from one character state to another. In any character state transformation, one state would be ancestral (or plesiomorphic in the vocabulary of phylogenetic systematics), and the other(s) would be derived (or apomorphic). If an apomorphic character state occurs in two of the three taxa in a three-taxon comparison, one hypothesis is that one character-state transformation occurred, in the ancestor of the two taxa sharing that character state, and as a result we might infer that these two taxa are the most closely related. Such a uniquely shared, derived character state is known as a synapomorphy, and synapomorphies are the character states used to define monophyletic groups. The following diagram illustrates this point:
In this diagram, taxa B and C each have the derived character state X’. Now, it is possible that B and C each acquired the derived character state independently, and that in actuality either B or C is most closely related to A. It is also possible that X’ evolved in the ancestor of A (and either B or C) and was lost in A. This is a character reversal, in which case the appearance of the formerly ancestral state X becomes a new derived state. However, the simplest hypothesis we can develop for the relationships among A, B, and C, given the information at hand, is that B and C share a common ancestor, which had acquired character state X’. This is the simplest because it required only one evolutionary event, or one “step.” Another way of saying this is that the most parsimonious hypothesis for the relationships among A, B, and C is that B and C are most closely related. One way of proceeding in science and logic is that when presented with more than one explanation for a phenomenon, you should proceed with the simplest, or most parsimonious, explanation, and this is one of the criteria used in phylogenetic analysis.
Exercise: With the help of your instructor, try making other arrangements of the tree shown above, in which A and B or A and C are actually sister groups (rather that B and C). Account for the occurrence of X’ in B and C in different ways (like independent acquisition in B and C, or acquisition of X’ in an ancestor of A and subsequent reversal). Count the number of steps, or characters state transformations, required in each of these trees. Does any alternate tree have as few steps as the one shown above?
Here is a more difficult, but more realistic exercise: Suppose we had information on two more characters, Y and Z. We almost always base phylogenetic analyses on more than one character. The distribution of character states is as follows:
A: W X Y’ Z’,
B: W’ X’ Y’ Z’
C: W X’ Y Z
Now try drawing different cladograms and determining the number of steps required in each one. What classification gives the smallest number of steps? In the simplest classification, one character disagrees with the other two. How would you account for its evolution? Another character occurs in its derived state in only one taxon. Does this character help you determine the relationships among the three taxa? Does the presence of the ancestral state in two taxa help you to determine relationships among the three taxa?
Some additional vocabulary:
A derived character state that occurs in only one of the three taxa being compared is called an autapomorphy. Autapomorphies help you to define a taxon, but do not help in determining relationships. For example, if you were comparing birds, crocodiles, and lizards+snakes, the presence of feathers in birds would be considered an autapomorphy. However, if one were, for example, comparing two groups of birds with crocodilians, the presence of feathers would be considered a synapomorphy for the two groups of birds.
Ancestral character states can be shared by two taxa in a three-taxon statement. These are called symplesiomorphies. Symplesiomorphies do not help in determining relationships, because distantly-related taxa can happen to share ancestral character states. For example, lizards+snakes and crocodiles both have scales as body coverings, while birds have feathers. If we assume that scales were the body covering for the ancestor of all of these groups (and we have good evidence for that), then the presence of scales in snakes+lizards and in crocodilians leads us astray. Other derived character states help us to determine that birds and crocodilians really are sister groups. However, the presence of scales in may actually be a derived character state when the reptiles are compared with amphibians, which lack any body covering. So the definition of synapomorphies and symplesiomorphies often depends on the groups being compared.
Sometimes derived character states evolve more than once, independently in different branches of a lineage. Such a character state is called a homoplasy, or convergence. It is often difficult to determine what shared derived character states are homoplasies and what are synapomorphies. Often (as in the exercise you conducted above), this will depend on the most parsimonious explanation for the relationships among the three taxa being compared.