This preview shows page 1. Sign up to view the full content.
Unformatted text preview: PHYLOGENETIC ANALYSIS CONSTRUCTING A PHYLOGENETIC TREE Introduction Perhaps the most common phrase used to describe evolution is "descent with modification." Descent entails an ancestor and its descendants: a genealogy. Modification involves a change in a characteristic or attribute in the descendant relative to the ancestor. Genealogical relationships among taxa cannot be directly observed. However, characteristics of living (or fossilized, when available) descendants can be observed. Since not all characters change at the same rate or to the same degree, descendants will be mosaics of unmodified (ancestral) and modified (descendant or derived) character states. It is the derived character states that you will identify and use to reconstruct the ancestordescendant relationships, or phylogen y, of the vertebrates. Because we can never directly observe these relationships, the tree that you develop is a hypothesis that can be tested by the examination of more characters. As a consequence, earlier phylogenetic hypotheses may be rejected with the inclusion of more data. There are four basic steps in constructing a family tree: 1. Identify homologous characters. 2. Outgroup comparison: determine the order and polarity of the character states. 3. Code the character states and construct a matrix. 4. Group by synapomorphies: analyze the matrix to produce a phylogenetic hypothesis. Homologous Characters A character is an observabl e trait of an organism. It may be morphological, physiological, behavioral, molecular, or ecological. A character may be passed on from an ancestor to its descendant either in unmodified or in some modified character state. For instance, if the character is eye col or, then character states might be brown eyes and blue eyes. If your parents have brown eyes which are the ancestral state, and you have blue eyes, then you have the derived character state. Characters used for analysis must be homologous. Homol ogous characters in two or more species are derived from the same structure in a common ancestor. This definition presents a problem, since we would need to have some estimate of relationships in order to determine homology. If we then attempt to determine relationships among taxa using characters whose homology has been determined by reference to some estimate of phylogeny we will be confounded in an endlessly circular argument. In order to avoid this circularity, we recognize homology by developmental, structural, or positional similarity. If it looks the same and is found in the same place then we will assume it is homologous. An example of homology is the vertebrate forelimb. Recall from your study of comparative morphology, that the basic forelimb plan is retained throughout the vertebrates: there is a humerus, a radius and ulna, carpals, metacarpals, and digits comprising phalanges. Recall also that there is variation among taxa in the exact morphology; frogs have one lower arm bone since the radius and ulna are fused; birds have a carpometacarpus which consists of fused carpals and metacarpals. How do we know which bones are homologous? The answer is that studies have shown that the bones develop in the same way and in the same position in the different taxa. During early devel opment in the frog, for example, both radius and ulna bones can be distinguished and only later on do they fuse to look like a single bone. A similar pattern has been found for bird forelimb development. Types of Characters Apomorphy =a new or descendant character state. When an apomorphy (apo = awa y from [the origin of life], morphy = form; derived character sate) is found in two or more taxa it is called a synapomorphy (syn =shared) which is a shared derived character. It is the synapomorphies which are used to infer phylogenetic relationships. Phylogenetic reconstruction may be viewed as a search for synapomorphies. A character can be a synapomorphy for a group only if no other organisms outside the group under study have the same character state. For instance, the presence of feathers may be considered a synapomorphy for different species of birds. Plesiomorphy = an ancestral character state (plesio = near [the origin of life]). A symplesiomorphy is a shared ancestral character state. These characters provide no information in resolving phyl ogenetic relationships. However, what is a symplesiomorphy at one level may become a synapomorphy at a higher level. For instance, the presence of feathers will provide no information if you are trying to reconstruct relationships among species of birds, because all birds have feathers of some form. However, if you are undertaking an analysis at the level of the tetrapods, then the presence of feathers becomes synapomorphic for birds. This illustrates an important point. What is derived at one level of analysis may be ancestral at another level of analysis, and vice versa. Outgroup Comparison Once you have identified potential characters for analysis, how do you det ermine which are plesiomorphic and which are synapomorphic? Several methods have been proposed to address this problem, with outgroup comparison being the system most commonly used. Outgroup comparison works on the following two assumptions. The first is that the group being studied, termed the ingroup, is monophyletic (all members of the group share the same, most recent ancestor). Second, the outgroup, used to polarize characters, is not part of the ingroup. Based on these assumptions, any homologous character state found in the outgroup and in the ingroup is considered plesiomorphic for the ingroup. States found in the ingroup and not in the outgroup are considered synapomorphic for the ingroup. This technique works readily when there are just two states of a character and one is shared with the outgroup. Funct ional Outgroup What happens if there is no state shared between the outgroup and the ingroup, or, if there is more than one derived state (i.e., character states 1,2,..) found in the ingroup? In this instance you create a tree based on the characters which can be polarized unambiguously. Once you have used these characters you will have resolved some of the relationships among the ingroup taxa. By using the character state found in the basal most members of the ingroup as the plesiomorphic condition, you may polarize the remaining characters to further resolve ingroup relationships. The technical term for this is "functional ingroup/functional outgroup analysis." After you have completed polarizing the characters, you will construct a data matrix. This is a summary of the character states found in each taxon. Typically, the ancestral state is coded "0" and derived states are coded "1". Some characters are not binary (Le. present/absent) and instead exist in more than two different states. Such multistate characters may be coded using other numbers, but it is important to understand that a code of "1", "2", etc., used to represent the states in a multi state character does not necessarily imply a sequence of change for the character, only that there is more than one apomorphic state. There are several wa ys to deal with multistate characters. In this lab, you will use one technique, and it will be performed by hand. This technique is known as Hennigian argumentation, after its originator Willi Hennig. Note that this proceeds by consideration of one character at a time. Classificat ion Once you have devel oped a hypothesis of phylogeny you can create a natural classification for the groups being analyzed. A natural classification is composed of onl y monophyletic groups and directly corresponds to the hypothesis of phylogen y upon which it is based. Such a classification is based on genealogical relationships among monophyletic taxa. This will permit the inferred phylogenetic relationships to be recovered from the classification scheme. Work Example Assume that you want to construct a hypothesis of phylogeny for a group of seven species which we will label A, B, C, D, E, F, and G for convenience. In addition, you use another species (X) which has some similarities with A to G but which is not part of that group as an outgroup to polarize the characters you have chosen for analysis. The table on the next page contains a list of the presumed homologous character states found in each taxon. Character states for seven species (AG) and an ancestral species (X) Species Legs Type of repro duction (2) eggs eggs eggs eggs live live live eggs Body covering (3) spines feather feather feather feather feather feather spines Feet webbing (4) yes yes yes yes no no no yes Tail Eyes Beak Horn or Teeth antler (8) antler antler horn horn horn horn horn no (9) Yes Yes Yes no yes yes no yes A B C D E F G X (1) 4 4 4 4 4 4 4 0 (5) no no yes yes no no no no (6) yes yes no no yes yes yes yes (7) no duck duck duck duck raptor raptor no Outgroup Comparison Now that you have arranged the original data, you can use outgroup comparison to polarize the characters. By definition, the ancestral taxa, X, is coded "0" for each character. Any character state of an ingroup taxa which is the same as the outgroup is thus coded "0", and if it is different from the outgroup it is a derived character and is coded "1" or "2". Based on the characteristics exhibited by each taxon, the following data matrix can be constructed. 1 0 1 1 1 1 1 1 1 2 0 0 0 0 0 1 1 1 3 0 0 1 1 1 1 1 1 4 0 0 0 0 0 1 1 1 5 0 0 0 1 1 0 0 0 6 0 0 0 1 1 0 0 0 7 0 0 1 1 1 1 2 2 8 0 1 1 2 2 2 2 2 9 0 0 0 0 1 0 0 1 X A B C D E F G BUILDING THE TREE Initially, there are no relationships known among the ingroup and outgroup taxa. Thus, if you were to draw a tree representing what you know of their relationships it would look like Tree #1. We build the tree character by character, by successivel y finding groups which share new characters (group by synapom orphies). Remember that a primary requirement of phylogenetic analysis is that the ingroup be monophyletic and that the outgroup is not part of the ingroup. Therefore you must have a synapomorphy that is found in all members of the ingroup and not in the outgroup. Character 1 is such a character. (Note: For the purposes of this example, we made Character 1 a synapomorphy for the whole ingroup. In your own analysis, it may be anyone of the characters). All of the ingroup taxa have legs while the outgroup does not. Therefore, having legs may be hypothesized to be a synapomorphy that defines the ingroup. Adding this character to Tree #1 produces Tree #2. By adding this character you have separated the outgroup from the ingroup and provided a basis for your decision that the ingroup forms a monophyletic group. However, there are still no relationships resolved among the ingroup taxa. Adding Character 2 (type of reproduction) to Tree #2 produces Tree#3. This still does not provide much in the wa y of resolution of ingroup relationships. Therefore proceed to add Character 3 (body covering) to the preceding tree to result in Tree #4. Based on these three characters you can now make the following observation: Taxon A is the sister group to a group consisting of Taxa B, C,D, E, F, and G. Now include Character 4 (feet webbing) in the analysis to produce Tree#5. Note that adding Character 4 to the previous arrangement did not bring any further resolution to the devel oping hypothesis of relationships. It did, however, strengthen the hypothesis that Taxa E, F, and G share a most recent common ancestor. Continue, by adding Character 5 (tail) to produceTree#6. Adding Character 5 produces the hypothesis that Taxa C and D are sister taxa. This is strengthened when Character 6 (eyes) is added, producing Tree#7. Character 7 (beak type) is a challenge because it has three states, one of which is shared with the outgroup. The condition found in the outgroup is the plesiomorphic condition. The question is, How did the character change from the plesiomorphic condition? Was it 0 → 1 → 2; 0→ 2 → 1; or 1 ← 0 → 2? This is where functional outgroups may be used to determine the order and polarity of the transformation series. Taxa EFG share a most recent common ancestor with each other that is not shared with any other of the ingroup members. Therefore, the rest of the ingroup (ABCD) may be considered to be the outgroup to EFG. By so doing, EFG functions as an ingroup (functional ingroup) and ABCD functions as an outgroup (functional outgroup). As a consequence, the state found in the functional outgroup can be considered to plesiomorphic to the state found in the functional ingroup. This also illustrates that what may be apomorphic at one level can be plesiomorphic at another. By applying functional outgroup analysis to Character 7, Tree #8 is generated. Character 8 (horns and antlers) also presents a challenge since there is no state shared with the outgroup, so it appears initially that this character cannot be used since there is no means to order or polarize its transformation. Using a similar argument to that used in the analysis of Character 7,Tree #9 can be generated. In this instance, the state found in Taxa AB is plesiomorphic and the state found in Taxa CDEFG is derived. Note that in order to use functional outgroups you must first have some resolution of ingroup relationships and that this is dependant on binary characters. The final character for analysis (9, teeth) occurs in Taxa D and G. This character is incongruent with the hypothesis of relationships depicted in Tree #9 and so we hypothesize that this is a homoplasious character. This demonstrates the distinction bet ween homoplasies and homologous characters. Homologi es are assumed before you begin your tree. Homoplasies are identified after you have completed your tree. There is no change in relationships with Character 9 included, as shown by Tree #10. However, there is not complete consistency bet ween the depicted hypothesis of relationships and the characters used in the analysis. Therefore the characters that are inconsistent with the hypothesis are indicated with an asterisk or some other means of recognition. It is important to realize that the resulting phylogenetic tree depicts relationships among taxa. There are many wa ys in which these relationships are represented. It is important that you be able to look at a tree and recognize the relationships that are being indicated. Questions and activities Complete the following exercise in building a phylogenetic tree using the cladistics method. Exercise 1 You have the following information about four plant species. Plant X is the outgroup. Recode the characters as ancestral or derived and then build a phylogen y. There may be more than one possible tree. Plant A B C D X Reproductive body Seeds Seeds Seeds Seeds Spores Type of leaves compound compound compound simple simple Type of stem smooth smooth hairy hairy hairy Rhizomes absent present present present present Plant height short tall tall short short Recoded characters Plant A B C D X Draw your tree here. 1 2 3 4 5 ...
View Full Document
- Spring '09