Goals: The goals of this lecture are to introduce the basic methods of systematics, the study of phylogeny, and the kinds of data used to study phylogeny.
Related Textbook Material: Freeman and Herron (2001) Chapter 13
Lab Manual Questions over this material are in Lab Manual Chapter XI
Other Assignments The material in this section also provides the basic background necessary for conducting your second outside assignment, on phylogenetic analysis of toucans.
We're going to spend a lot of time on systematics so let's take a moment first to consider why. Systematics is currently a very active field of study in evolutionary biology. It may surprise you to learn that something as basic as how species are related to each other is very poorly known. The reasons phylogenetic relationships among most groups are poorly known include the fact that there are so many different species (with new species being discovered all the time) and the fact that we continue to develop better and better ways of obtaining data on the species we know about. Even groups for which we though we had a pretty good idea of the phylogeny are being studied and studies are indicating that we may be wrong in our old understanding of phylogeny. Just this month (February, 1999) a study has come out suggesting that turtles may be close relatives of crocodiles and alligators; previous studies had suggested that they were only distant relatives to the other living reptiles. Systematics is also an active field of study because it is turning out to have a variety of applications. In making decisions about conservation of species, we are starting to use information on how species are related to each other. The same kinds of method are being used to trace the origin and spread of disease (for example, in another recent study, this kind of information has been used to identify the particular subspecies of chimpanzees from which HIV-1 spread to humans; knowing this may allow us to study how these chimpanzees can have HIV-1 but never develop AIDS and that might allow us to help treat AIDS in humans.) Finally, people are studying systematics because it allows us to learn more about how traits have evolved within groups, and to understand the process of evolution better -- this is our main goal in this course.
The goal for this lecture is to see how we determine how different species are related to each other. That is, we want to learn what the phylogeny of life is. Note first that there is just one true phylogeny -- life has evolved just once, and in just one pattern. As with everything else in science, we can't know this one true phylogeny for sure -- we can't prove that we've found the true phylogeny. What we can do is to test hypotheses of phylogeny. Any phylogenetic tree we draw can be considered a hypothesis of what the phylogeny of a group might be. We then test it by obtaining information about the species we're studying.
There are two main ways we learn how different species are related to each other. One is by studying fossils, which, as you saw in the lecture on evidence for evolution, provide a preserved record of some of the history of life. Not all previously existing organisms are preserved, however; most are not, so the fossil record can provide only an incomplete picture of the phylogeny of life. The other line of evidence is the hierarchical pattern of homology. As you saw in the lecture on evidence for evolution, when different species share structures it provides evidence that these species have that structure because they evolved from an ancestral species that had that structure -- homology thus provides evidence for relationships (phylogeny.) The hierarchical pattern of homology shows some species, with many traits in common, as being close relatives, while other species, with fewer traits in common, are only distant relatives.
Early systematists (biologists who study systematics) typically studies phylogeny by simply looking at how many characteristics different species shared with one another. Later, systematists realized that not all shared characteristics are useful in studying relationships. This realization led to a way of studying systematics called cladistics. Cladistics is the study of phylogenetic relationships based on shared, derived characteristics.
To understand what shared, derived characteristics are, and why only they, and not other characteristics, are useful for studying phylogeny, we will consider an example. The following picture shows the relationships among three species: a lizard, an alligator, and a bird. If you simply look at their characteristics, a lizard and an alligator look very similar to each other -- that is why they are typically classified together as reptiles, while birds are classified into a different group (the birds.) Studies of phylogeny suggest, however, that alligators and birds are actually more closely related to each other than they are to lizards. This pattern of relationships is shown in the following tree, as are some of the characteristics people have used to study relationships among these species.
First let's consider some of the terms used in the above figure. Primitive traits (also called primitive character states, or primitives states) are characteristics of organisms that were present in the ancestor to the group that we're studying. In this case, if we're studying the relationships among lizards, alligators, and birds, any characteristic that was present in the common ancestor to these three groups is a primitive trait. The examples shown here are: a body covered with scales, a heart with only three chambers, the lack of a gizzard, the lack of wings, and the presence of teeth.
Derived traits (also called derived character states, or derived states) are characteristics of organisms that have evolved within the group we're studying -- they were not present in the ancestor. In this case, where we're studying the relationships among the lizards, alligators, and birds, derived traits include: the presence of a gizzard, a four-chambered heart, feathers, the absence of teeth, and wings.
Now let's see how we use this information to study phylogeny. If we just considered what characteristics these species had in common, without considering which were primitive and which derived, we would find some characteristics suggested relationships between the alligator and bird but others suggested relationships between the alligator and lizard. For example, the alligator and bird both have a gizzard and the lizard does not, so that characteristic would suggest birds and alligators are related. However, alligators and lizards both have four legs (not wings) and birds have two legs and two wings. That would suggest that alligators and lizards are related. Which characteristics do we believe?
To answer that question, note that not having wings is a primitive trait. The ancestor to the entire group did not have wings. The ancestor from which the lizard and alligator have inherited the trait of not having wings, then, is the ancestor to the entire group -- it was also an ancestor of birds. Because it was the ancestor to the entire group, it only shows that all members of the group we're studying came from that ancestor. It can not show that any species within the group are more related to each other than to the rest of the group. In general, primitive traits can not indicate anything about the relationships of species within a group because they are inherited from the ancestor to the entire group and simply reflect the fact that ALL members of the group, at some time in the past, evolved from one ancestral species. Even the species without the primitive trait evolved from this ancestral species; they have lost the primitive trait at some later evolutionary point. For example, birds lost the primitive trait of four legs when they evolved to have wings, but they did come from an ancestor with four legs.
In contrast, having a gizzard was NOT present in the ancestor to the entire group. The gizzard is a derived trait. Since it is present in the alligator and bird, but not the lizard, it is evidence that the alligator and bird have inherited the trait of having a gizzard from an ancestor that was NOT the ancestor to the whole group. Thus it suggests that the alligator and bird have an ancestor that evolved the gizzard and that was not also the ancestor to the lizard. So it suggests that the alligator and bird are more related to each other than to the lizard. In general, derived traits that are shared among some (but not all) species within a group provide evidence for phylogenetic relationships within a group because the most likely explanation for why some species would have a trait in common, if it was NOT in the ancestor to the whole group, is that some more recent ancestral species within the group evolved the trait, and the species that currently have the derived state have evolved from that more recent ancestral species. This means that the species with the derived trait descended from an ancestral species that was not the ancestor to the species without the derived trait, so the species with the derived trait are more related to each other than to the species without the derived trait.
Read over the past two paragraphs carefully! They show why we study phylogenetic relationships using cladistics, the method of studying phylogenetic relationships based on shared, derived traits. Only shared, derived traits provide evidence for phylogenetic relationships among species. Another reason for reading over the past two paragraphs carefully is that on every exam over this material I've given, I've asked for a clear, complete written explanation for why derived traits can provide evidence for phylogeny but primitive traits can not. You can expect a question on this on your next exam!
Once you have convinced yourself that only derived traits can provide evidence for phylogeny, the next question you should have is how do we know which traits are primitive and which derived? There are two main lines of evidence for this. One comes from fossils. We can sometimes see that some traits in organisms go back much farther in time, since they are found in much older fossils, than are other traits. For example, vertebrates with four legs and no wings go back very far in the fossil record; birds, with wings, appear much more recently. So we have strong evidence that having four legs and no wings is primitive and having two legs and two wings is derived.
Often, we do not have a complete fossil record and can not use it to determine which traits are primitive and which are derived. Also, there are many traits that can provide evidence for phylogenetic relationships, but which do not fossilize. Currently, many studies of phylogeny are based on studies of molecules, especially DNA -- DNA can almost never be obtained from fossils. So we need another way to determine which traits are primitive and which are derived.
The most commonly used method of determining primitive versus derived traits is called outgroup comparison. To use outgroup comparison, we find one or more species that are relatives of the group we are studying but outside of it in that they are equally related to all members of the group we are studying. We call the group we are studying the ingroup and these species that are equally related to all ingroup members the outgroup. The rule we use to determine which characteristics of species are primitive and which derived is that characteristics found within both ingroup AND outgroup species are most likely primitive while characteristics found just within some, but not all, the ingroup species that are NOT found in the outgroup are most likely derived.
Let's look at the possible ways species in the ingroup and outgroup could share characteristics to see why this method works, and to review which characteristics are going to be useful to study phylogeny. Consider the following phylogeny of some vertebrates: a shark, a trout, a frog, and a mouse. This will be our ingroup. Also shown on the phylogeny is a close relative of vertebrates, amphioxus, which will be our outgroup. The species names are given at the top, and some of their characteristics are given below that. Our understanding of the phylogeny of vertebrates is shown by the phylogenetic tree.
First, note that the outgroup, amphioxus, shares only the ancestor to the whole group with the species that make up the ingroup; it is equally related to all species in the ingroup. Remember this must always be true of the outgroup.
Now consider the characteristics of the species. Having a cartilage skeleton is present in the outgroup (amphioxus) and in an ingroup species, the shark. Because this characteristic is found in both the outgroup AND the ingroup, it is most likely that the outgroup and ingroup species inherited it from the species that was the ancestor to both the outgroup and the ingroup (the species labeled "A" on the tree.) Thus, it is a primitive characteristic. In general, characteristics found in both the outgroup and some ingroup species are most likely to have been inherited from the ancestor to the outgroup and the ingroup; that is why we consider them most likely primitive to the ingroup.
Having a bone skeleton is NOT present in the outgroup; it is present in only three ingroup species -- the trout, frog, and mouse. The presence of a bone skeleton thus suggests that the trout, frog, and mouse evolved from an ancestral species that was NOT the ancestor to the shark (since it does not have a bone skeleton.) On the phylogeny shown, the ancestor that would have evolved the bone skeleton (the ancestor to the trout, frog, and mouse, but NOT the shark) is labeled "B". The presence of the bone skeleton in the trout, frog, and mouse provides evidence that they had this ancestor. In general, sharing a derived characteristic provides evidence for ancestry.
Because it potentially provides evidence about phylogeny, we consider the characteristic of a bone skeleton to be phylogenetically informative. Primitive characteristics like the cartilage skeleton are NOT phylogenetically informative. The other two characteristics on the tree illustrate two other kinds of characteristic that are NOT phylogenetically informative.
First, consider the presence and absence of jaws. The outgroup does not have jaws; all members of the ingroup do have jaws. Because this trait is in all members of the ingroup but not the outgroup, we can not tell which form is primitive and which is derived. The ancestor to the ingroup and outgroup (A) might have had jaws and the outgroup evolved to lose jaws. Or the ancestor to the ingroup and outgroup might not have had jaws and the ancestor to just the ingroup could have evolved jaws. Based on the information given here, either possibility is equally likely so we can't tell which trait is primitive and which derived. In general, if ALL members of the ingroup share a trait that is not found in the outgroup we can not tell what form is primitive or derived so the trait is not phylogenetically informative since we have to have evidence that the trait is derived for it to be phylogenetically informative.
Now, consider the presence and absence of fur. The mouse has fur; the other species do not. Since the outgroup and several ingroup species lack fur, we have evidence that not having fur is primitive and having fur is derived. But in this case, the derived trait of having fur occurs in only one species in our phylogeny. This suggests that fur evolved in the ancestor to just this species. It does not indicate any ancestor shared with other species on the tree. As a result, it does not give information about phylogeny. In general, if the derived character state is found in JUST ONE species in the ingroup, it is NOT phylogenetically informative.
We have one more complicating factor to consider in studying phylogeny. We have been assuming so far that if different species have a trait in common, they have inherited that from some ancestral species -- in other words, that this trait is an example of homology, having a trait in common because it has been inherited from a common ancestor. This is the most likely reason for them to have a trait in common, but it is not the only reason. Species can evolve the same traits independently. Independent evolution of the same characteristic is called convergent evolution.
The presence of wings in both birds and bats is an example of convergent evolution. Consider the following phylogeny (as with the phylogeny above, it shows species and some of their characteristics.)
In this example, the salamander is the outgroup and the mouse, bat, and bird are the ingroup. Based on outgroup comparison, the following are derived character states: fur, mammary glands, and wings. Fur and mammary glands both provide evidence that mice and bats are more closely related to each other than to birds, but wings suggest that bats and birds are more closely related to each other than to mice. Both can't be correct. In general, in a situation like this, where different derived character states indicate different relationships, we accept the phylogeny that is supported by most derived character states as the best supported hypothesis of phylogeny. The reason we can do this is that we know that mutations are rare, so that for different species to have mutations leading to similar traits, and then having those traits evolve independently in both species, should be much less likely than species staying the same and having similar characteristics because they are inherited from an ancestor.
In this case, two derived character states, fur and mammary glands, suggest that mice and bats are close relatives, and only one derived character state, wings, suggests that birds and bats are close relatives, so our best supported hypothesis of phylogeny is that bats and mice are more related to each other than they are to birds.
If bats and mice are more related to each other than to birds, then bats and birds must have evolved wings independently. The two "W"s on the phylogeny show where wings must have independently evolved -- once in the ancestor leading to just the bats and once in the ancestor leading just to birds. In contrast, fur and mammary glands each evolved just once (indicated by the letters F and M on the tree), in the ancestor to the bats and mice. These are examples of homology: bats and mice both have fur and mammary glands because they have inherited them from this ancestor in which they evolved.
In studies of phylogeny, it is important to consider how good the support is for the phylogeny we find. We may find that we are looking at characteristics that agree to a large extent on what tree they support. Such characteristics apparently show little convergent evolution. If this is the case, then we can have confidence in our phylogeny -- while we're never sure it is the true phylogeny, we can say it is very likely to be true. In contrast, we may find that we are looking at characteristics that disagree with each other and that show a lot of convergent evolution. We may still find a best supported hypothesis of phylogeny, but we would have to say that we do not have much confidence that it is really the true phylogeny because apparently we are looking at characters that evolve independently quite often and may not really show us phylogeny very well.
There are a variety of statistics that can be calculated to indicate how good the support is for a phylogeny. We will consider one very simple statistic called the consistency index. The consistency index measures how well characteristics agree with each other with regard to phylogeny. It is a ratio; its highest value is 1. A consistency index (abbreviated CI) of 1 means that the phylogenetic tree we have found is consistent with all the characteristics we've used -- none shows convergent evolution, all show homology. If there is convergent evolution, the consistency index will be lower than 1. So a low CI means low confidence that the tree is really the true phylogeny; a high (close to 1) CI means high confidence that the tree is really the true phylogeny.
The CI is calculated as: (# of derived states)/(# evolutionary changes to derived states) That is, we look at the number of derived states for all the characteristics we're studying. In the example above, with bats, birds, and mice, there are a total of three derived states: fur, mammary glands, and wings. The number of evolutionary changes in these derived states is four: fur and mammary glands each evolve once (as indicated by the letters F and M on the tree) and wings evolve twice, independently (indicated by the two "W"s on the tree.) So the CI=3/4
The number of evolutionary changes to derived states on a tree is also given a name; it is called the treelength. In general, when trying to find the best supported hypothesis of phylogeny, we try to find the tree that shows the characters evolving so that they have the shortest (smallest) treelength. This gives us the highest possible CI for our characteristics.
At this point, you should try to work problems 1-3 in Lab Manual Chapter XII; these are problems in finding the best supported hypothesis of phylogeny for groups of species and characteristics. You will also find the treelength and CI for these phylogenies.