Cladistics Teacher Notes
The information that follows was written by Joseph Skulan, paleontologist at the University of Wisconsin-Madison and can serve as a resource for the teacher. It provides a relatively brief, yet basically complete, description of cladistics. As with all classroom topics it is helpful for the teacher to have a more complete understanding of the topic than they plan to help the students develop. It does provide support for class discussions and the questions that can be raised by students. As is always the case, each teacher will make a decision on the depth to which the topic is covered.

Introduction
Humans are constantly classifying things. This is because it is impossible to think about or work with all but the smallest amount of information unless it is organized in some way. Imagine trying to use a telephone book that was not alphabetized. The way we organize information depends on how we want to use it. A system that works well in one situation may not work well in another. The best way of classifying carpet samples in a carpet store is by color and texture, but this would be a very bad way of classifying books in a library (although it has been done).

Like carpet samples, books, or names in a telephone directory, living organisms can also be classified in different ways. For example, cooks classify spices into "sweet" and "savory, while fisherman classify fish into "game fish," "pan fish," "junk fish," and other categories. In each case the classification system used is based on some characteristic of the organism that is important to a particular profession. Cooks want to know the flavor spices will give to food, while fisherman want to know if a fish tastes good or if it's fun to catch.

Classification in biology
Biology deals with all organisms and no one needs a way of classifying organisms more than biologists. But how should biologists classify living things-- what are the characteristics of living things that are most important to biologists? Over the centuries these questions have been answered in different ways. But one thing has long been recognized; organisms seem to fall into natural groups. Domestic cats are clearly closer to lions than to dogs, while dogs, domestic cats, and lions are all closer to each other than any of them are to pigeons.

For a long time biologists classified organisms into what seemed like natural groups using a system devised in the 18th century by the naturalist Linneaus. This is referred to as the Linnaean system of classification (or as Linnaean taxonomy), and is the source of such familiar categories as kingdom, phylum, and class, (though not genus and species). In the Linnaean system organisms that share certain key features are classified together into groups. These groups are in turn defined according to the features of the organisms they contain. For example, in the Linnaean system birds are placed in Class Aves. All birds, and as far as is known nothing else, have feathers, so in the Linnaean system Class Aves could be defined as all organisms with feathers, and any organism with feathers would be, by definition, a bird.

Cladistics
The Linnaean system is still used in some branches of biology. But in other branches, and particularly in vertebrate paleontology, it is rapidly being replaced by a system referred to as cladistics or phylogenetic systematics. Cladistics was invented by the German entomologist Hennig in the 1950s, but the basic methods of cladistics were devised in the 19th century by philologists attempting to reconstruct the histories and interrelationships of European languages. Most cladistic classification schemes are based on Darwin's realization that the reason organisms fall into natural groups is that they are related to each other by common descent. Domestic cats resemble lions more than dogs for the same reason brothers and sisters resemble each other more than they do their cousins; they share a more recent common ancestor. Rather than just looking for natural groups, cladistics tries to classify organisms according to the thing that makes them fall into natural groups in the first place: their common ancestry or evolutionary history. For cladistics, classifying organisms and reconstructing their evolutionary relationships are one in the same thing. With the rise of cladistics, systematics-- the study of how organisms are classified-- has been transformed from a kind of glorified filing system into an important branch of biological science.

In cladistics all natural groups of organisms, called clades, consist of an ancestor and all of that ancestors descendants, and nothing else. Linnaean categories are often, but not always, clades. One major difference is that according to cladistic convention, organisms can never evolve out of their clade. For example, most paleontologists accept that birds are descended from dinosaurs. Under the Linnaean system birds are put in Class Aves, while dinosaurs are included in Class Reptilia. This means that birds evolved out of Class Reptilia and into Class Aves. Such a thing would not be possible in a cladistic system. A common cladistic definition of dinosaurs is the last common ancestor of Tyrannosaurus and Triceratops and all of its descendants. Assuming birds are descended from this common ancestor, birds are dinosaurs. Period. If at some time in the distant future birds were to return to water and evolve into something that looked exactly like a salmon, they would still be dinosaurs, because no amount of evolution will change their ancestry, just as you can do nothing to change who your great-grandparents were. Also, nothing that is not a descendant of the last common ancestor of Tyrannosaurs and Triceratops will ever become a bird, no matter how closely it may come to resemble one. In the same way your best friend, no matter how much he or she may be like and imitate you, will never be able to become the descendant of your great-grandmother, unless he or she already is. When paleontologists classify birds as dinosaurs they are not claiming that birds and other dinosaurs resemble each other in any particular way, only that they have common ancestor that they share with no other organisms.

To put all of this in a more formal way, cladistics recognizes only monophyletic groups, that is, groups that include a single ancestor and all of its descendants. Groups that do not include the common ancestor of all of its members are called polyphyletic; groups that include the common ancestor but not all of its descendants are called paraphyletic. Shellfish and pachyderms are examples of a polyphyletic groups. Paraphyletic groups include fish and reptiles (at least as these groups are usually defined), and dinosaurs, if birds are not included. Polyphyletic groups are usually avoided even in the traditional Linnaean system. The real conflict between traditional and cladistic taxonomy is over paraphyletic groups. Linnaean taxonomy recognizes many paraphyletic groups, but cladistic taxonomy rejects them all.

Definition vs. Diagnosis
In order to understand cladistics, and how it differs from other taxonomic systems, it is helpful to make a distinction between definition and diagnosis. The definition of a group tells us what organisms the group includes; the diagnosis is how we actually decide if a particular organism is a member of that group. There are several ways of defining groups in cladistics, but they all come down to common ancestry. ÝBirds are (for example) the last common ancestor of Archaeopteryx and a chicken, and all of that last common ancestors descendants. ÝThis is a definition of the monophyletic group birds. But we cant see an organisms ancestors just by looking at it, so we need indirect ways of deciding if an organism fits the definition of bird. We know from experience that the descendants of the last common ancestor of Archaeopteryx and chickens usually have feathers, pneumatic bones, inelastic lungs, and other characteristic features. These are the features that actually let us diagnose, or decide, Ýif an organism is a bird.

Think of medicine. By definition Influenza is a disease caused by a particular group of viruses. Influenza also causes a characteristic set of symptoms, and it is by these symptoms that you, or your doctor, diagnose the flu. You probably know that you have had the flu at least once in your life, but the chances are you have never actually had your tissues analyzed to see if they are infected with the flu virus.

Diagnoses can change with time, but not definitions. A bird may lose all of the diagnostic features of a bird-- it may lose its feathers, its pneumatic bones, and all of the other features that let us diagnose it as a bird, but it would still be a bird. 450 million years ago vertebrate might have been diagnosed by the presence of heavy, bony armor. ÝThis is no longer a useful diagnosis, because most living vertebrates have no armor. But all vertebrates will always be descendants of the same ancestor from which the earliest vertebrates were descended, so the definition of vertebrate will never change. Thus even if through evolutionary modification vertebrates were to lose all trace of a backbone, they would still be vertebrates. Presence of a backbone diagnoses, but does not define, the vertebrates.

Characters and Character States
There are several formal terms that are used in connection with the diagnosis of monophyletic groups, or clades. Organisms differ from each other in particular ways. The features in which organisms differ are called characters, while the particular ways they may differ are called character states. For example, people differ from each other in the color of their eyes. Eye color is a character. Eyes may be brown, blue, green, or some other color. These different colors are character states. A organism's character state may be unchanged from that of its ancestors. Such a character state is called primitive of plesiomorphic trait. Or an organisms character state may be new and different from its ancestors. These character states are called derived or apomorphic traits. If my eyes are brown and the eyes of my descendants were blue, brown eyes would be a derived trait for me. In cladistic literature character is often used to refer to both character and character state. It is usually obvious from context in which way the term is being used.

In cladistics only derived or apomorphic traits are used to diagnose groups, because only these traits give evidence of common ancestry. For example, loss of limbs is an apomorphic trait of snakes, and can be taken as evidence that snakes are all related to each other. But within the snakes loss of limbs is a primitive, or plesiomorphic, trait, and doesn't help tell us who is more closely related to whom. If you want to know whether a garter snake is more closely related to a cobra or a rattlesnake, the fact that garter snakes don't have legs help you decide. More specifically, cladistics deals with shared derived character, or synapomorphies. These are derived character states possessed by more than one group of organisms. Derived character states that only one group posses (called autapomorphies), tell us nothing about how that group is related to other groups. If I have brown eyes but everyone in both my mothers and fathers families have blue eyes, my brown eyes wont settle any arguments about which side of the family I favor.

The same character states may evolve independently in different groups. This is called convergence or homoplasy, and characters states that have evolved more than once are called homoplastic. An example of homoplasy is the long, narrow snout has evolved independently in several groups of ant-eating mammals. Because character states may be homoplastic, we cannot always be sure that the presence of a derived character state in two organisms means that these organisms are closely related. Distinguishing convergence from true synapomorphy is one of the great problems in studying evolution.

Cladograms
A cladogram is a diagram showing how organisms might be related to each other. The illustration on the next page shows a simple cladogram. It consists of a set of taxa (or groups of organisms) connected by lines. A, B, and C are the taxa that are being classified. Together they make up the ingroup. The outgroup is an organism that we are pretty sure is not part of the ingroup. We use the outgroup to help us decide which character states are primitive and derived for the ingroup. The point at which the lines leading to any two taxa meet is called a node, and can be thought of as representing the common ancestor of these two taxa (cladistics usually tries to avoid having more than two branches at a any node). The taxa being classified are almost never placed at the nodes, because we seldom, if ever, know that a particular taxon is ancestral to another taxon. Fortunately we can reconstruct the evolutionary relationships between any organisms without any direct knowledge of their ancestors. Never assume that because an organism lived a long time ago that it is ancestral to anything.

Each line that comes off a node is a monophyletic group, as long as everything that is attached to that line is included in the group. The two monophyletic groups that come of each node are called sister taxa. This means that they are more closely related to each other than either is to anything other group. The relative size of the sister taxa is not important. One may be very large, and the other very small. In real cladograms, a single species is often the sister taxa of a very large group. For example, in many reconstructions Archaeopteryx is the sister taxon of all other birds. On this cladogram A and B are sister taxa to each other. A and B together also form a monophyletic group that is the sister taxon to C, and A, B, and C together are the sister taxon to the outgroup. Any part of the cladogram around which you can draw a circle that only cross one line is a monophyletic group. Another way of thinking about it is to imagine that the cladogram is a tree. Any part of the tree that you can cut off by cutting through only one branch is a monophyletic group.

The Character Matrix
The distribution of character states among taxa is the information that is used to build a cladogram. Before you can build a cladogram you need to organize this information by constructing a character matrix: A. Look at the types of organisms and identify as many characters as possible-- these are traits in which there is some difference among the organisms you are looking at. B. Determine two states for each character (you can also determine more than two states, but this makes things much more difficult, and usually doesn't do any good). C. For each character decide which state is primitive and which is derived. This is known as polarizing the character. The primitive state is the one that outgroup has. D. Make a chart (called a character matrix) showing the character state (primitive or derived) of each character for each types of organism.

Once you have finished the character matrix look to see which types of organisms have the most derived traits in common. Ignore primitive traits. Those that have the most derived in common probably share a more recent common ancestor with each other than they do with the others, and will likely be grouped together in the final cladogram. At this point paleontologists would simply take the data matrix, plug it into a computer, and with luck get one cladogram back.

Parsimony
Going from a character matrix to a cladogram requires one more concept: the Principle of Parsimony. This is basically Ockhams Razor, the idea that when deciding between possible explanations of the same event, the simplest explanation, the one that makes the fewest assumptions, is to be preferred. In the case of building cladograms this means that the cladogram that requires the least evolution to achieve the observed distribution of character states is the best. For example, all cats have long canine teeth. It is simpler-- more parsimonious-- to assume that long canines evolved once in the ancestor to cats than that it evolved independently in each cat species. If is possible that long teeth did evolve independently, but this explanation makes much less sense than the simpler one, because there are so many more ways that it could be wrong. But things are not that simple. A cladogram that lets cats long teeth evolve only once may require another character state to evolve many times. Characters can give conflicting information, so we need to come up with the cladogram that is most parsimonious overall, even if it requires some character states to be distributed in less than the most parsimonious way. We do this by counting the total number of character state changes that occur on a particular cladogram. Every time a character state changes is considered a step. The cladogram with the fewest number of steps is the most parsimonious. In the illustration on the next page, parsimony favors grouping the green-eyed organisms together into a single clade, because this requires that green eyes evolved only once, and is therefore more parsimonious.

A cladogram* of the types of organisms found in this activity can be found below. The mosasaur would be placed nearest to the Komodo dragon and snake although its specific location is being debated.

Characteristics
• A: Bone, teeth, teeth on roof of mouth, well ossified skull roof, hinged braincase allowing movement of skull, sculpted dermal bone, no temporal openings aquatic (primitive characters)
• B: Simplification of shoulder girdle, separation of shoulder girdle from skull, loss of hinge on braincase, terrestrial
• C: Two temporal openings on skull
• D: Loss of sculpted bone, loss of lower bar on lower opening, quadrate free
• E. Loss of sculpted bone, loss of teeth on roof of mouth, aquatic
• F. Loss of teeth on roof of mouth

*This cladogram is also based on other morphological features as well as the molecular evidence of DNA and proteins