Geology 107: Paleobiology

TAXONOMY (Classification)

Biologists who try to recognize, describe, name, define, and classify species are taxonomists or systematists, and the practice is called taxonomy or systematics or classification.

Biologists use the Linnean system of naming species, after the Swedish biologist Carl Linne who invented it in the 18th century. A species is given a unique name (a specific name) by which we can refer to it unambiguously. For example, Linne gave the little owl the specific name noctua, because it flies at night. Species that share a large number of characters are gathered together into groups called genera (the singular is genus) and given unique generic names. Linne gave the little owl the generic name Athene. She is the Greek goddess of wisdom, and the little owl is the symbol of the city of Athens. However, taxonomic names do not have to carry a message, even though a simple and appropriate name is easier to remember. (One must be careful about names: Puffinus puffinus is not a puffin, but a shearwater, and Pinguinus is not a penguin but the extinct Great Auk!) Thus, Linnean names are only a convenience, but a very valuable one. The bird that the British call the tawny owl, the Germans the wood owl, and the Swedes the cat owl, is Strix aluco among international scientists.

Genera may be grouped together into higher categories. Many species of owls are grouped together to form the Family Strigidae, or strigids, named after one of its genera, Strix. Families may be grouped into superfamilies, and after that into orders, classes, and phyla. Many other subdivisions can be coined for convenience. Although slightly different ranks of categories are used for different kingdoms of organisms, the basic units of classification recognized by all biologists remain the species and genus.

Complex rules have grown up as more and more organisms have been described by taxonomists. Botanists and zoologists have their own rules for describing new species so that names are legally established.

Describing Evolution

Linne did not believe in evolution. But we now recognize that members of a species form a genetically based biological unit that is evolutionarily separate from the rest of the organic world. The recognition and naming of a new species, therefore, is a statement about evolution. It represents the taxonomist's hypothesis that the members of the species share the same gene pool, which is different from the gene pool of any other species because there has been evolutionary divergence over time.

Most taxonomists aim to form genera and higher categories that truly reflect evolution. In evolutionary classification, species are grouped into genera on the hypothesis that genera too share a unique set of characters that are not shared by other genera. In turn, genera are grouped into families, families into orders, orders into classes, and classes into phyla. Of course, decisions about the course of evolution are not always obvious, so taxonomic decisions are always subject to reappraisal as new information becomes available. Species are moved around between genera and higher categories as taxonomists continually refine their classifications to reflect evolutionary history more effectively. The incomplete nature of the fossil record makes classification particularly difficult for paleontologists, and it often leads to uncertainties or arguments about classification.

As organisms evolve through time, their characters change. Characters may change slowly or quickly; they may change gradually or rather suddenly. As changes accumulate, one species may evolve characters that are changed from their original state. The new set of characters may be different enough that a biologist who could examine living specimens at the ends of the series would certainly regard them as separate species. But how can one draw a line between species in time? After all, descendants have always been genetically continuous with their ancestors. There is no discontinuity between ancestor and descendant species like that seen between contemporaneous species in the living world. This is a special question facing paleontologists, and it makes the taxonomy of the fossil record rather difficult. There's nothing unusual about the situation. At one extreme one could say that all living organisms are the same species, because they all evolved, continuously, from a single ancestor that was the first living cell. For convenience, however, and to reflect the reproductive gaps that exist between the species at any given time in Earth's history, a paleontologist must draw lines somewhere between species (and genera, and families, and so on), knowing that the lines are artificial if they pretend to separate ancestors from descendants. Fortunately or unfortunately, the fossil record is spotty enough, and the pace of evolutionary change is rapid enough, that truly intermediate fossils are very rarely found. The fossil record is therefore rather more easily divided into species and higher categories than one might expect.


I use a mild form of a new approach to evolutionary taxonomy, cladistics. The aim of cladistics is to trace the phylogeny (the evolutionary pathways) by which species appeared and to use that phylogeny to organize species into clades, that is, groups of species that all descended from one ancestral species. To use an analogy, a clade is a branch on a tree that represents all life: if all life on Earth is descended from the first living cell by a series of evolutionary branching events, life as a whole is one clade (the entire tree). Just as trees may branch many times, and branches then branch, and so on, clades of organisms exist in a hierarchy of scales, with the terminal branches representing single species. Every species belongs to a larger clade, which belongs to a larger clade, and so on: every clade, no matter how large, began with a single branching event that produced the ancestor species of the clade.

Cladists operate by trying to identify groups of species that share a set of characters that were evolved as new features in a common ancestor, and then passed on to all descendant species. Newly evolved characters represent a change from a primitive or original state to a novel or derived state. For example, all living birds have feathers, but no other living organisms do. Perhaps feathers were inherited from a common ancestor of all living birds that evolved a feathered skin as a newly derived character modifying a primitive one (a skin covered by scales). If that is true, then birds are a clade. Looking more closely, one finds many other shared derived characters of living birds that strengthen the argument: all living birds have a wishbone, a breastbone, and a boxlike pelvis, and they pass air through their lungs rather than in and out.

Sometimes problems arise because similar derived characters are found in species outside a clade; those characters have evolved more than once by parallel evolution. For example, bats and birds both have wings, and in each group the wing is a derived character that has been modified from some other structure. But bats and birds share very few other derived characters, and the weight of evidence suggests that birds are a clade, bats are probably a clade, but [bats + birds] is not a clade.

Once a clade is established, we then search for characters that are novel in species within the clade. Subclades can be established based on the distribution of such derived characters within the group, until a single best hypothesis emerges about the total evolutionary history of the group of species. The hypothesis can be tested as further characters are examined or existing ones are reassessed, and as new species are discovered and fitted into the evolutionary framework.

As an example, three living species, A, B, and C, could be related along three possible evolutionary pathways. Which is correct? Which two of the three species are most closely linked? Two species may look very similar because they share similar characters, but if those are shared primitive characters that were also present in a common ancestor, they cannot tell us anything about evolution within the group, because they have not changed within that history. The useful character for solving the problem is the novelty-the derived character-which defines the group that has changed the most since the three species all shared the characters of their common ancestor.

Cladograms such as Figure 3.8 display the distribution of characters in a visual form, and the cladogram that requires the simplest and fewest evolutionary changes is assumed to represent best the phylogenetic history of the species. A cladogram therefore expresses a hypothesis about the phylogeny of a group. Two of the species are most closely linked, and form sister groups in a clade, while the third species becomes their sister group in a larger clade.

Once the preferred cladogram is drawn to portray the best hypothesis, one can make decisions about the best way to classify the species and to describe its evolutionary history. A cladogram in itself does neither of these things.

One could introduce formal names for each clade on a cladogram. It's obvious, however, that this would lead to a great number of names, not all of which might be necessary for everyday discussion around the breakfast table. It conveys more information simply to print a cladogram and to use a minimum of hierarchical names.

A cladogram is always drawn with all the species under study along one edge (Figure 3.9a). No species in a cladogram is shown as evolving into another. Some cladists claim that one can never know true ancestor-descendant relationships, and in a strict sense this is correct because we don't have time machines. But sometimes a fossil is known that could well be an ancestor of a later fossil or of a living organism. At present, for example, it seems more reasonable (to me) to suggest that Homo erectus is the ancestor of Homo sapiens than to suggest that Homo sapiens is descended from some species of human that we haven't found yet. Hypotheses like this are expressed on phylograms or phylogenetic trees that include time information: we can show a suggested ancestor within the diagram (Figure 3.9b). Like cladograms, phylograms are not statements of fact but hypotheses subject to continuous testing.

It's sometimes useful to recognize a stem group, which includes all the early members of a clade that subsequently becomes very large but which excludes the later, advanced, derived members. (A stem group is not a clade, therefore.) For example, all living mammals may be descended from a small group of mammals that lived in Jurassic times but are not very well known. Some day we may be able to trace ancestries well enough to split that small group of early mammals into several clades, even though they were closely related to one another and not very different in structure or ecology. But until we have more information, it is more convenient to lump most Jurassic mammals into a stem group, recognizing that it contains several separate clades but also recognizing that we can't yet sort them out properly.

Counterintuitive patterns sometimes emerge in cladistics. We are all used to thinking about fishes, amphibians, reptiles, birds, and mammals as classes of vertebrates, in some way equal in rank to one another. But this is not a cladistic classification. Tetrapods are one clade within fishes, derived from them by acquiring some novel characters, and amphibians are a clade within tetrapods. Reptiles are in turn a derived subgroup of amphibians. Mammals and birds are equally clades of derived reptiles.

There's nothing intimidating about this: it simply takes some time to get used to it. The important thing is not to try to force the older, stable taxonomic units into a cladistic framework, but to combine a simple and convenient classification with an explanatory cladogram or phylogram. I have tried to use cladograms and phylograms in this way.

If we classify all living reptiles into one group and draw a cladogram, we display the well-known fact that living reptiles and birds are more alike than either is to mammals. The cladogram also carries other information. It shows that warm blood, a derived character that living birds and mammals share, must have evolved independently at least twice, unless living reptiles have secondarily lost warm blood.

As we consider smaller subgroups of living and fossil reptiles, we find that the neat picture of reptile classification breaks down, and that we must revise our interpretations. "Living reptiles" is not a clade. We can define a clade called "reptiles," but we have to include mammals and birds in it. This means that turtles, mammals, birds, crocodiles, and lizards are all reptilian clades that have diverged from an ancestral, primitive reptile, although some are more derived than others in the sense that they have evolved more novel characters that their common ancestor did not have. In the same way, humans are derived fishes, derived amphibians, and derived reptiles, all at the same time. Once one becomes used to this unusual line of thinking, evolution becomes much more real, and we can see, for example, that humans, tapeworms, and bacteria in green pond slime are all cyanobacteria: though some have accumulated more visible derived characters than others.

Return to 107 main menu