The millions of species of animals, plants, bacteria and so on that comprise Life on Earth are, for the sake of convenience, placed into groups, or taxa (singular: taxon). Some of these taxa are fairly simple and obvious: 'birds' and 'flowering plants' are two good examples. More specific but equally understandable taxa include 'whales', 'snakes' and 'mosses'. In the eighteenth century a Swedish naturalist called Carolus Linnaeus took upon himself the task of organising all these groups, giving them scientific names and placing related ones together within larger taxa. For instance, he placed the Reptilia (reptiles), Aves (birds) and Mammalia (mammals) together within the larger taxon Vertebrata (animals with a backbone). And his system of classification, known as Linnaean taxonomy, is still used today.
Unfortunately, while Linnaeus's system works pretty well most of the time, it is fundamentally flawed. For a long time scientists have been building up a good understanding of how all the different groups of organisms are related to one another, and have developed 'family trees' which show the ancestry of modern species. And this leads to problems. For example: if birds are descended from reptiles, at what point did they stop belonging to the Reptilia and start belonging to the Aves? In which group do you place an animal like Archaeopteryx, which possessed as many reptile-like characteristics as bird-like ones? The more we discover about the relationships between taxa, the muddier the distinctions become between them.
What is Cladistics?
In the 1950s a German zoologist named Willi Hennig devised a new system of classification called 'cladistics', which attempted to address this problem. Cladistics is beautifully simple in principle. According to its rules, every defined taxon must contain the species ancestral to that group, plus all of that species' descendants. Thus, the taxon Dinosauria includes not only those extinct reptiles that we consider to be dinosaurs, but also any descendants of dinosaurs - and that happens to include birds. So, the group Aves falls within the Dinosauria.
A group which conforms to this rule of cladistics is called a monophyletic taxon. Not all Linnaean taxa are monophyletic: the Linnaean definition of the Reptilia, for example, does not include birds, which are descendants of members of that taxon. The Reptilia, therefore, is known as a paraphyletic taxon, which means that it does not contain all the descendants of its ancestral species. Paraphyletic taxa are not recognised by cladistics.
Another difference between Linnaean and cladistic taxonomy is the nomenclature used to describe the taxa. The Linnaean system organises organisms into successively smaller groups known as (respectively) Kingdoms, Phyla, Classes, Orders, Families, Genera and Species. Kingdoms are the largest groups, and there are five of them, including Animalia and Plantae (animals and plants). Species are the smallest groups. Cladistics abandons this convention almost entirely, since such a structured hierarchy of taxa is simply not realistic when focusing on relationships. Instead, each monophyletic taxon is called a clade, and all organisms are grouped into clades within clades within clades. The only concession to Linnaean systematics is that the terms genera (singular: genus) and species are kept for convenience (genera are small groups of closely related species).
How it works
Cladistics operates in a very simple way. It compares the acquired, or derived, characteristics of different organisms, and works out how closely related they are to each other. Let us take three well-known animals: a python, a rabbit and a cow. The relationship between these animals is, apparently, clear. The rabbit and the cow are both warm-blooded, hairy and suckle their young by means of mammary glands. They are therefore almost certainly more closely related to each other than to the python. This relationship can be represented thus:
This diagram is called a cladogram and is how relationships are generally represented in cladistics. The intersections of the lines on the cladogram represent a hypothetical 'common ancestor'. Therefore, the cow and the rabbit have a shared ancestry that excludes the python, while the common ancestor of the python and the cow was also ancestral to the rabbit.
We have assumed that this cladogram is correct, but is it? The temptation is to say "of course it is - it's obvious", but that only reflects our current preconceptions, and all too often in the history of science, preconceptions have turned out later to be misconceptions. So let us examine an alternative:
Here we are putting forward a new hypothesis: that the python is more closely related to the rabbit than to the cow. How can we justify this? Well we can speculate that, at some point since the last common ancestor of the python and the rabbit, the python's ancestors lost their hair, their mammary glands, their legs, and their warm-bloodedness. This is, in theory, possible - loss of derived traits is something that happens frequently in evolution.
Occam's Razor, or the Principle of Parsimony
To resolve an issue like this, cladists turn to the Principle of Parsimony (also known as Occam's Razor). This principle states: if two possible explanations are presented, the simplest one is most likely to be true. In our example, it is more parsimonious (ie more likely) that warm-bloodedness, hair etc was not first gained by mammals and then lost by the python, but instead gained by mammals after the line leading to the python had diverged. Thus, the first cladogram we looked at is more parsimonious and therefore more likely. The importance of looking at all possibilities (and not discounting those that seem foolish) is that we become more certain of our conclusion and do not get caught out by our preconceptions.
So far so not-very-surprising. But the application of cladistics to some creatures proved revolutionary, and very hard for the scientific establishment to take. Take the example of the salmon, the lungfish and the frog. The salmon and the lungfish both have fins, it is true ... but the frog's legs have evolved from fins, so fins cannot be considered a derived trait that is missing from the frog's ancestry. The lungfish and the salmon both have internal gills ... but on the other hand the lungfish and the frog both have lungs. Either internal gills have been lost at some stage in the frog's ancestry, or lungs have been lost at some time in the salmon's ancestry. Which is more parsimonious?
Well, to resolve that question, we shall consider a fourth animal, one that is more primitive than any of the other three. The lamprey is a very simple vertebrate - so simple that it does not even possess jaws, and it has no bony skeleton, just a vertebral column that supports its spinal cord. Yet this primitive vertebrate possesses gills, and no lungs. Gills, therefore, seem to be an ancestral characteristic for vertebrates. Lungs, on the other hand, are a derived characteristic of one particular vertebrate lineage. It therefore makes sense to ally two lunged vertebrates, such as the lungfish and the frog, within one clade and exclude a lung-less vertebrate such as the salmon. It is more parsimonious to assume that lungs were gained once, after the salmon's ancestry had diverged from that of the lungfish and the frog, than to assume that lungs were gained by a common ancestor of all three animals and thereafter lost by the salmon's ancestors.
Let us now consider the fins of the lungfish. The muscles that support those fins are not positioned within the body, as is the case with the salmon. Instead the fins themselves are fleshy and muscular, just like the legs of the frog. In fact, there are several features of the lungfish that demonstrate its close relationship to terrestrial vertebrates. Thus our cladogram looks like this:
The implications of this are quite staggering: the group we call 'fish' is entirely artificial. If, as cladistics demonstrates, the lungfish is more closely related to tetrapods (amphibians, reptiles, mammals and birds) than it is to the salmon, then why should we place salmon and lungfish in a group which excludes tetrapods? And this is where the traditional system of classification falls down. According to Linnaean taxonomy, the lungfish and the salmon both belong in the Osteichthyes (bony fish), which excludes tetrapods. Cladistics, with its insistence on monophyletic groups, does not fall into this trap.
The Cladistical View of Vertebrate Relationships
Once we have used Occam's Razor on many different animals, adding them each by turn into our cladogram, we begin to build up a picture of how all animals are related to each other. Of course, a cladogram focusing on individual animals quickly becomes huge and complex, so we begin to group animals into clades (monophyletic groups), and hence build up a broader, simpler and more wieldy picture. Here, for example, is a cladogram showing the relationships between various living vertebrate clades.
This cladogram, built up from many years of scientific research, shows a great deal of information that Linnaean taxonomy does not even hint at. We can see at a glance, for example, that of all the fish in the world, our closest relatives are the lungfish, and that crocodiles are more closely related to birds than they are to lizards or turtles.
Each of the groups featured on this cladogram is a clade, a valid monophyletic group containing animals which are more closely related to each other than to any other animal, living or dead. And each clade is defined on the basis of a derived trait. The clade described here as 'Ray-Finned Fish', for example, is defined (unsurprisingly) on the basis of the 'rayed' fins which characterise all members of this group.
But a group of clades which are more closely related to each other than to any other animals is, itself, a clade. The group comprising crocodiles and birds is a clade known as the Archosauria, which is defined on the basis of a specialised ankle joint that improved locomotive ability. (Incidentally, the extinct dinosaurs and pterosaurs are also members of the Archosauria.) The clade comprising these archosaurs plus snakes and lizards is known as the Diapsida, members of which possess two characteristic skull openings behind the eye socket.
And so it goes on. The Tetrapoda, for example, is a clade consisting of the descendants of the first animal to possess true limbs with digits as opposed to muscular fins. This includes the amphibians and all clades below the amphibians in our cladogram above. 'Tetrapoda' means, quite helpfully, 'four feet', and indeed the names of clades are very often descriptive of the traits that define those clades. The snag is that some of these defining features are later lost by members of the clade. Snakes have lost their feet, but they are still tetrapods since they are descended from tetrapods.
The Linnaean method - a comparison
The Linnaean representation of these taxa would look like this:
Class: Chondrichthyes (sharks, skates and rays)
Class: Osteichthyes (bony fish)
Subclass: Actinopterygia (ray-finned fish)
Subclass: Sarcopterygia (fleshy-finned fish)
Order: Actinistia (coelacanths)
Order: Dipnoi (lungfish)
Class: Amphibia (amphibians)
Class: Reptilia (reptiles)
Order: Chelonia (turtles)
Order: Crocodylia (crocodiles and alligators)
Order: Squamata (snakes and lizards)
Class: Aves (birds)
Class: Mammalia (mammals)
This, of course, tells us nothing about how any of the taxa are related, and it is even misleading. One might suspect, looking at the Linnaean classification system, that crocodiles were closely related to lizards, and only distantly related to birds, when in fact the reverse is true. Cladistics places birds firmly within the Diapsida, and further allies them with crocodiles within the Archosauria. This, in my opinion, makes more sense.
Cladistics is a highly useful tool for palaeontologists. It enables them to achieve a far better understanding of the relationships between living things and how life on Earth evolved from simple beginnings. The relationship between, say, a dolphin and a weasel is not immediately clear and obvious, and Linnaean taxonomy helps very little with that. But a cladogram containing both those animals will reveal not only how they are related to each other, but also, through analysis of their shared derived traits, what their common ancestor might have been like, and what has happened to the rest of that ancestor's descendants.
Unfortunately, cladistics is not accepted by all of the scientific community. There are some who insist on sticking with Linnaean taxonomy, and there are some good arguments for this. Cladistics is quite complex and cladograms are subject to frequent changes as new evidence turns up. This makes it very hard to keep textbooks up to date. In addition, many scientific disciplines are unconcerned with evolutionary relationships and the simplicity of Linnaean systematics works quite well with modern species. The battle continues, and although it would be nice if cladistics were universally adopted as the sole method for classifying organisms, this is unlikely to happen for many years to come.