I wrote this article in 1995, when I was just getting seriously interested in fossils and evolution. I kept running into strange words when reading about how fossils were classified -- "cladogram," "plesiomorphy," "apomorphy," and more. I looked around for a basic introduction to taxonomy, something that a beginner could understand, and couldn't find one. So I got a couple of references from the library and wrote one myself. It isn't 100% accurate, and looking back at it now I see several things that could have been written a lot better. But it still serves the purpose of a basic introduction to taxonomy, something that anyone should be able to understand.
In order to understand the relationships between living and extinct animals and plants, it's necessary to have some understanding of taxonomy. It's especially important when dealing with extinct groups of animals like dinosaurs, because we have no modern animals to compare them to, and a great deal of what we know about them is involved in how we classify them.
First of all, what is taxonomy? Simply stated, taxonomy is the branch of biology that deals with classifying living things. Like most names of scientific fields, the word 'taxonomy' is built from Greek word roots. For example, 'biology' was created from the Greek words 'bios', meaning 'life or living', and 'ology', meaning 'the study of'. So 'biology' is 'the study of living things.' 'Taxonomy' is also a composite word, created from 'taxis' (meaning 'arrangement or order') and 'nomia' (meaning 'law'). So taxonomy = 'the laws for organizing things'.
Linnaean or classic taxonomy
Many biologists have tried to set up usable systems of taxonomy, but none really caught on until about two hundred fifty years ago. In 1737, a Swedish botanist named Carl Linnaeus developed a system of taxonomy for plants that proved so useful it was eventually adopted for all living things. This is now known as the Linnaean system of taxonomy. The Linnaean system is binomial ('two-names'), meaning that each specific organism receives a two-word name which uniquely identifies it. That name is the organism's scientific name. The first word is the genus name. The second word is the species name. For currently living things, the normal tactic is to have both names be taken from either Latin or Greek, and the English translation of the two names is a unique and (hopefully) accurate description of the creature. For example, the domestic dog is named Canis familiaris. "Canis" is the Latin word for "dog," and "familiaris" is a Latin word that means "familiar or domestic." So "Canis familiaris" means "Dog, familiar", which is indeed a good description of Man's best friend.
When writing a scientific name, it's customary to set it apart somehow from the rest of the text in which it appears. You can use underlining (Canis familiaris), or italics (Canis familiaris), or boldface type (Canis familiaris). In this essay, I'm using italics to indicate scientific names. When writing or typing a scientific name, you also always capitalize the genus name, and put the species name in all lowercase letters.
Linnaeus never formally defined what he meant by a "species", which is unfortunate since science requires rigorous definitions. However, later taxonomists did define a species. The current definition of a species is "a population of organisms that forms a single, self-contained gene pool." Two or more populations that can interbreed naturally form one gene pool. Two populations of organisms that are sufficiently different either morphologically or behaviorally are considered different species, even if they can breed in captivity. Lion and tiger can hybridize in a zoo, but in the wild, they are separated by behavior patterns as well as by shape, size, and color (morphology), so they're different species.
Linnaeus didn't stop with just his binomial naming scheme. He went much further. He grouped his plants according to relationships he saw between them, and this too was carried over into the general system of taxonomy. He indicated that several species of plants were closely related by giving them all the same genus name and different species names.
Scientists who came after Linnaeus greatly expanded his system. They defined larger and more general taxonomic groups using similarities between smaller groups as criteria. Similar genera ('genera' is the plural of 'genus') are combined to form a Family. Similar families make up an Order. Similar orders make up a Class. Similar Classes make up a Phylum. And similar phyla ('phyla' is the plural of 'phylum') are grouped togther as a Kingdom. The Kingdom is the highest, largest, and most general taxonomic grouping. So the groupings of Linnaean taxonomy look like this, from most general to most specific:
For example, your family dog, if you have one, is classified like this:
Kingdom Animalia (animals)
Phylum Chordata (animals with a notochord)
Class Mammalia (animals that produce milk)
Order Carnivora (meat-eating mammals)
Family Canidae (dogs and doglike animals)
Genus Canis (dogs)
Species Canis familiaris (familiar dog)
Naming of taxonomic groups
If you think that the family name (Canidae) looks an awful lot like the genus name (Canis), you're right. When a new family of organisms is named, the usual practice is to pick the genus that is most typical of the family, then modify that genus name to form the family name. Dogs are most typical of the dog family, so the dog family is called the Canidae. The common house cat (Felis catus) is most typical of the family of animals called 'cats', so the formal name for the cat family is Felidae. The most typical bear is the grizzly bear (Ursus horribilis), so the bear family is the Ursidae. And so on.
Unfortunately, nothing in science is ever simple, and so it is with taxonomy. Linnaeus thought that his seven levels of classification were sufficient. But before long, his successors began recognizing sub-groups that were larger than the next 'official' subgroup, but smaller than the parent group. For example, all mammals, birds, reptiles, amphibians, and fish are vertebrates, or animals with a backbone. However, the distinguishing feature of the vertebrates isn't just the backbone, but the large, specialized nerve cord that the backbone encloses. There was a time when the vertebrates were considered a phylum, Vertebrata. But then biologists discovered animals that had the specialized nerve cord, but did not have a backbone. Clearly these belonged in the same phylum with vertebrates, but they weren't vertebrates because they didn't have backbones. So a new phylum was created, Chordata, including all vertebrates and all the nonvertebrate animals that had the special nerve cord. The group Vertebrata is still around, but it's been changed to a "subphylum" of the phylum Chordata.
The same thing happened in other major groups. The largest animal phylum is Arthropoda, the arthropods. There are four major groups of arthropods, the trilobites, chelicerates, crustaceans, and insects, but they're all so different from one another that they are considered almost separate phyla. So, they're all subphyla of Phylum Arthropoda.
Some taxonomists recognized that all land-dwelling vertebrates were more closely related to each other than any of them were to living fishes. So a group was defined between the Class level and the Subphylum level to indicate this. The 'superclass' Tetrapoda includes all land-dwelling vertebrates: amphibians, reptiles, archosaurs, birds, and mammals.
The same thing happened again and again as more organisms were classified. Taxonomists kept adding sub-groups to indicate more and more detailed relationships between organisms. However, the more they did this, the more complicated the taxonomic system became. Three or four genera might be assembled into a group that was called a family. Then another genus or two would turn up that was similar enough to the first group to be in the same family, but different in some significant ways too. So the first group became a subfamily, part of a new, larger family which also included the later finds.
After a while, the urge to add subgroupings reached absurd levels. By 1975, the typical classification scheme for the Class Mammalia included nineteen levels between Class and Species. Most taxonomists also use subspecies, making twenty levels just within a Class. Since there can be as many as four levels above Class (Kingdom, Phylum, Subphylum, Superclass), the total number of levels in a classification can reach 25 -- over three times as many as what Linnaeus originally used. Taxonomy was getting more and more complex and difficult to remember, but still there were holes and gaps in it.
And even with all this, taxonomy was fallible. Traditional taxonomy depends on a system called "phenetic" classification. Phenetics is a system of classification that uses overall similarities and differences to group organisms. Phenetics does not distinguish between primitive traits, specialized traits, and false 'similarities' produced by convergent evolution. For example, under an old phenetic classification scheme, the elephant, rhinoceros, and hippopotamus were placed together in the Order Pachydermata, because it was thought the fact that all three have very thick, heavy skin indicated they were related. Only much later was it shown that the three animals are more properly placed in three different Orders.
It was clear that taxonomists needed a method of classification that was more difficult to fool than the traditional phenetic method.
In the 1970s, a new method of classification began to be used more and more. This method became known as cladistics. Cladistics was developed as a more powerful and systematic way of placing organisms into the Linnaean taxonomy, but as it grew it became less of a method and more of an alternative to traditional classification.
The power of the cladistic method stems from the fact that it focuses on specific traits and uses comparisons of those traits to group organisms. It's a more powerful method than phenetics, because phenetics can be fooled by superficial similarities. Cladistic analyses can't be fooled that easily. And a properly developed cladistic analysis is self-correcting. If a classification is incorrect, the cladistic analysis almost always makes that clear and enables the taxonomist to look for a better classification.
The basic rules of cladistics are:
- Some traits are defined as primitive, while others are defined as derived. Primitive traits are traits shared by a large number of species. Derived traits are traits shared by only a few species within the larger group. The formal term for a primitive trait is a plesiomorphy. The formal term for a derived trait is an apomorphy.
- Primitive traits cannot be used to distinguish sub-groups of species within the large group. Only shared derived traits, or synapomorphies, can be used to distinguish sub-groups of species.
- Two species that share a number of synapomorphies form a sub-group relative to a third species that does not share all of those advanced traits. The third species is then called an outgroup relative to the first two.
- Any species will have one or more traits unique to itself. These traits are called unique derived traits, or autapomorphies. They can be used to define species, but not to define higher taxonomic groups like genera or families.
- The taxonomist must be careful to distinguish between true shared traits (homologies), and traits produced by convergence (analogies). Convergent traits can produce an false relation.
For example, consider these six animals: the house cat (scientific name Felis catus), the cougar (Felis concolor), the lion (Panthera leo), the cheetah (Acinonyx jubatus), the wolf (Canis lupus), and the whitetail deer (Odocoileus virginianus). How can we classify them using phenetic and cladistic methods?
First, what are the distinguishing characteristics of these animals?
- House cat: A small predatory mammal, 4-8kg, 30-45cm long overall. 30 teeth adapted for biting, cutting, and shearing, not for chewing. Purely carnivorous diet. Wide, rounded, flattened skull with a short, wide muzzle and large eyes placed in the front of the face. Walks on the first four toes of its foot; fifth toe has a developed claw but doesn't touch the ground. All claws on all feet are fully retractable. Natural habitat: thin forest, edge, fields. Can purr, but cannot roar.
- Cougar: A large predatory mammal. Up to 90kg and 2 meters long overall. Teeth similar to the house cat's in number and development. Purely carnivorous diet. Large, wide, rounded, flattened skull with a wide muzzle. Feet and claws are similar to the house cat's. Generally solitary. Natural habitat: ranges widely from forest to prairie to mountains. Can purr, but cannot roar.
- Lion: A very large, heavily built predatory mammal, up to 225 kg and 3 meters long overall. Similar to cougar and house cat in dentition and diet. Shows sexual dimorphism: male is larger and usually has a mane. Long, flattened skull with a wide muzzle, eyes in front of the skull, but smaller and more widely separated than in the house cat. Highly social, living in groups of up to thirty with five to ten (plus cubs) typical. Feet and claws are similar to the house cat's. Natural habitat: mainly savannah; also found in mountains, thin forest, edge. Can both purr and roar
- Cheetah: A large predatory mammal. 30-50 kg and 2-2.3 meters long overall. Similar to cougar and house cat in dentition and diet. Purely carnivorous diet. Slender build with heavily muscled hindquarters. Distinctive spotted coat with black lines from eyes down the sides of the muzzle. Somewhat rounded skull with triangular muzzle, eyes placed well back on the sides of the skull. Four walking toes on each foot; fifth toe is well developed but nonfunctional. Partially retractable claws. Highly specialized for speed, the swiftest land animal over short distances, capable of speeds up to 100kph. Solitary. Natural habitat: savannah and edge.
- Wolf: A medium-size predatory mammal, 25-45 kg and up to 2 meters long overall. Has 42 teeth adapted for biting and chewing meat. Diet is mainly carnivorous, but will eat berries and other vegetable matter. Large skull with long, tapered muzzle and smaller eyes placed in the front of the skull, producing stereoscopic vision. Four walking toes on each foot, with nonretractile claws. Front feet have a fifth degenerate toe high on the inside of the leg, the dewclaw, which is not attached to the skeleton and can be removed with no permanent debility. Highly social; usually lives in packs of 8-20 individuals. Natural habitat: forest, fields, tundra, mountains, thin forest, edge, fields.
- Whitetail deer: A large herbivorous mammal, about 1.2 meters tall, 1.5-2 meters long overall, and up to 250 kg. Has 32 teeth adapted for chewing vegetable matter. Herbivorous diet. Long, narrow, deep skull with eyes set wide apart and looking to the sides, not straight ahead. Male is larger and has antlers which are shed and regrown yearly. Has a complex four-chambered stomach adapted for digesting leaves and twigs. There are two toes on each foot that form a cloven hoof, plus two more, degenerate toes on the back of the foot. Habitat: field, deciduous forest, edge.
Now, how do we classify these six animals?
Under the phenetic system, we would do it something like this:
- The house cat and the cougar are almost identical in most significant ways: diet, dentition, skull and skeleton, shape, claws, hunting style, voice. The differences in size, coloration, and habitat are enough to make them different species, but not enough to justify different genera. So the cougar and the house cat are placed in the same genus.
- The lion is similar to the cat and the cougar in diet, dentition, claws, general build, and hunting style. So it's also a cat. However, it shows several significant differences: its ability to roar, its sexual dimorphism, its social lifestyle. These differences are enough to justify placing the lion in a separate genus.
- The cheetah is similar to the first three in dentition and basic physiology. It's recognizably a cat, but it has several characteristics that sharply differ from the other three cats we've looked at. Its claws are only partly retractable. It cannot climb or use its claws as weapons. It's extremely specialized for speed. These are enough to place it in a genus separate from the others, but it remains in the cat family, Felidae.
- The wolf differs from all the cats in dentition, skull shape, diet (omnivorous as opposed to carnivorous), and many other aspects of its physiology. Its claws are not even slightly retractable. Its behavior is radically different. All told, the wolf is so different from the cats that it must be placed in a different taxonomic family, Canidae. However, it's still much closer to the cats than it is to the deer, and this is indicated by placing it and the cats in the same Order, Carnivora.
- The whitetail deer is radically different from all the others in almost every detail. It shares only the basic mammalian traits with them. It's a mammal, so it belongs in the Class Mammalia, but it must go as far from the others as it can, into a separate Order.
So, this analysis shows that cougar and house cat form a genus; they, lion, and cheetah are separate genera within the Family Felidae; the wolf is in a different Family within the Order Carnivora; and the deer is in a separate Order within the Class Mammalia. Useful information, certainly, but can we do better? Let's see what a cladistic analysis of these animals shows.
- First, we can say that all six animals share the traits that define the Class Mammalia. They're all vertebrates. They have hair, upright gait, are warmblooded, bear live young, and nurse them using milk. So, they're all mammals.
- Next we observe that our six animals have two basically different approaches to getting food: the deer is an herbivore, while the others are all carnivores. More, the deer has a number of traits that specialize it for its herbivorous lifestyle. All of these traits are derived ones, so they are useful for grouping the deer relative to the others. So the deer is in a different subgroup of the Mammalia from the others in our test group. This is the first big division within the group: the carnivorous animals form a different sub-group, or clade, from the herbivorous one.
- Next, we look at differences between our five carnivores. We observe that the wolf has 42 teeth in a certain pattern, while the four cats all have 30 teeth in a different pattern. Teeth patterns are considered a derived trait, so here is evidence that the wolf is substantially different from the cats. More evidence can be found in the feet and the general design of the skull and body. In all of these except one, the four cats are all closer to each other than any of them are to the wolf. So we can conclude that the wolf is in a different clade from the cats. They're all carnivores, but within the carnivores there is another division, with the wolf on one side and all the cats on the other.
- Now we're down to the four cats. We look at them, and what do we see? We see that one of them, the cheetah, is noticeably different from the others in several ways: its claws, its active chasing of prey, its specializations for speed. If retractile claws are a derived trait for cats, then the cheetah does not have all the traits that make a cat a cat. It's still close enough to the other cats to be called a cat, but they and the cheetah lie on opposite sides of a division within the cats.
- Three left: lion, cougar, and house cat. What differences are there among these animals? Well, the lion can roar; the other two can't. The lion is a social animal; the other two aren't. The lion shows sexual dimorphism; the other two don't. It appears, then, that we have yet another division, with the lion on one side and the cougar and house cat on the other.
- What's left? Very little. Besides size and some details of the skull and teeth, the differences between the cougar and the house cat are slight and subtle. They can't interbreed, so they're separate species, but they are still extremely close to each other.
(Please note that this example is neither rigid nor wholly accurate. It is intended only as an illustration.)
Now, let's assemble our cladistic analysis into a diagram, a cladogram. If we follow the divisions we noted in the analysis, the cladogram looks something like this:
When we look at this cladogram, we can observe a few things right away. First, all the branchings can be easily matched to groupings within Linnaean taxonomy. The first branch is between two Orders, the Carnivora and another Order that contains the deer. The second branch is between two Families, one that contains the wolf and one that contains the cats. The third branch, between the cheetah and the other cats, appears to be a subfamily division. The fourth branch is between two genera, and the fifth is between two species. So, by analyzing our six animals cladistically, we've arrived at a better understanding of the ways in which they are similar to one another.
Now, this example is by no means complete. It's impossible to be rigorous in such a small space. I could write a couple of pages on what makes each of the divisions I've drawn a legitimate way of dividing animals into clades. Also, this analysis is based only on the physical traits of the living animals. If we add fossil animals, we might get a different result -- for example, fossils suggest the cougar is more closely related to the cheetah than to the house cat. If we add genetic analyses, we might get a different result. If we add other species, we might get a different result -- for example, the leopard and tiger are so similar to the lion that they belong in the same genus, and the lion's sexual dimorphism and social lifestyle are unique derived traits (autapomorphies), not a good basis for a genus-level division. Still, this example catches the flavor of how cladistics work and some of what it can do.
But that's not all cladistics can do. A cladistic analysis can go far beyond simply linking living organisms. When you add fossil organisms to the cladogram, it grows into a representation of the evolutionary history of the organisms being classified. You need to know a lot about a fossil organism to put it in a cladogram, but in the cases where we do know that much, we can analyze fossils by cladistics as easily as we can analyze currently living organisms. And adding fossils often validates the cladogram by providing species at or very near the branch points, confirming that the relationships in the cladogram are real and not just our imagination.
So that's a brief look at the science of taxonomy. Actually, looking back I realize it's not all that brief -- but less wouldn't have let me do the subject justice. Whole books have been written on taxonomy, and even those don't cover all of it. There's a great deal more to modern taxonomy: monophyletic and paraphyletic groupings, and how to tell them apart; methods of cladistic analysis; how to tell a valid clade from an invalid one; how to determine what traits to use in classifying an organism; and still more.