Science

Acids, bases and pH

Atomic structure

Biochemical nomenclature

Biotransformation

Carbohydrates

Carnivorous plants

Cell communication

Cell cycle

Chromatography and electrophoresis

Cladistics

Core metabolism

DNA replication

Electron microscopy

Enzymes

Enzyme mechanisms

Essays

Extracellular matrix

Genetics

Genomes

Kinetics

Lipids and membranes

Molecular biology methods

Moles

Natural selection

Nervous system

Nitrogen metabolism

Nuclear structure

Nucleic acids

Oxidative phosphorylation

Phloem

Photomorphogenesis

Photorespiration

Photosynthesis

Plant action potential

Plant circadian rhythms

Plant growth

Plant mineral nutrition

Plant secondary metabolites

Prokaryotes and eukaryotes

Proteins

Protein targeting

Reaction mechanisms

Seeds

Spectroscopy

Statistics

Thermodynamics

Transcription

Translation

Water potential

You are what you eat

Reference

Domains

Genetic code

Geological timeline

Model organisms

Molecules

Plants

Vegetable empire

Steve's Place

Search

Guestbook

Me

Perl

Science

Rants

Cladistics

A monkey's uncle

Every living thing on this planet is related to every other, and the differences many of you may hold dear, like those between mammals and reptiles, or even between plants and animals, really aren't very large. Although there's a fair bit of difference between you and a coliform bacterium from your faeces, you both have a lot in common: you both use protein enzymes, you both have DNA and RNA, you both have cells bounded by layers of fat, and you both have the same dictionary for translating DNA messages into proteins. The chances of Escherichia or Enterobacter and people having all these features in common by chance are so microscopically small, you may as well forget them. You are a cousin to your own bowel bacteria, like it or not.

The science of classifying organisms into species, and into higher groups like families and kingdoms, is called taxonomy, and it used to be thought of as a rather musty dull subject studied by tweed-jacketed old men in museums who smelt of formaldehyde. However, taxonomy is the way to understanding of our origins, and of the interrelationships between living things on earth.

Cladistics measures how closely two or more species are cousins. We know that everything has to be a cousin of everything else, for the reasons above, so it's just a matter of putting an exact figure on how closely related you are. This has led to a few surprises. For example: birds are crocodiles with feathers, all land vertebrates are fish, and sharks aren't fish at all.

In the common - but misguided - view that evolution must be progressing towards the ultimate goal of producing human beings, it has become common-place to consider life as some sort of hierarchy, with neat pigeon-holes labelled 'fish' and 'reptiles' arranged on some grand one dimensional ladder of life. Life is a huge, two dimensional, twiggy, branching tree, with no 'top' and certainly no ladder with bacteria at the bottom, and people in the topmost branches lording it over the amoebae, carrots and mice, arranged one on top of the other, at his feet.

Putting people in their place

• Mammals are a rather ancient group (read 'primitive', if you so desire), which nearly went extinct at the end of the Permian in competition with dinosaurs. Mammals still retain several ancestral features that the more 'advanced' dinosaurs (birds) have improved on, like the number of holes in their skulls.
• Most species of animals (in fact most species of everything) are insects. The numbers are so large, we are not even close to classifying them all. The vertebrates are an withered appendage to the arthropod corpus of the animals.
• This so called 'Age of the Mammals' is grotesquely misnamed. The most cursory glance at a field full of insects, rock pool swarming with molluscs, river full of bony fish or human gut infested with bacteria will prove beyond doubt that mammals are not the dominant animal group. There are more bird species and more species of fish than there are of mammals, and most mammals are mice or rats.
• People are just some weird bunch of hairless primates, a look also sported by those sartorial luminaries, the naked mole rats.

It is noteworthy that the list above also proves our parochial obsession with animals: in total biomass, plants outweigh animals by hundreds to one, and even their success is predicated on their possession of two sorts of symbiotic bacteria, within their own cells: eukaryotic life as we know it couldn't exist but for a number of fortuitous accidents some billion years ago, where our ancestors began to keep herds of bacteria in their cells to produce energy for them, in the form of mitochondria and chloroplasts. Humans are just another species, and any coherent classificatory structure must have knit into its very core the fact that our 'specialness' is something we only recognise in hindsight.

Adventures in flatland

In the beginning, there was no classification of living things, and there was much wailing and gnashing of teeth. Then some bright spark noticed that animals and plants were quite different: one lot moved, and the other lot was green. And so was born the first great classification of life: green things and moving things. Later, Carolus Linnaeus realised that the names of animals and plants would be a lot simpler if they were given Latin soubriquets, so that everyone (well, everyone but the illiterate plebs) would be able to describe them without fear of confusion. And so was born the Latin binomial system: Homo sapiens, Escherichia coli and Tequila sunrise.

In doing this, it was noted that animals and plants seemed to be divisible into groups of similar sort of organism: like birds, mammals, apes, pitcherplants and oak trees, some of which were more inclusive than others: chaffinches were birds, which, with the mammals and others, were all vertebrates, which themselves were a certain sort of animal. And so the idea of a hierarchy of taxonomic levels came to pass. And so classification entered flatland…

Two steps forward, one step back

After the revelation that life fitted an ordered pattern, someone, seeking desperately, high and low, for a bad analogy, thought, "Aha, 'tis like a ladder, with the worms and filth at the bottom, with us and thr angels at the very pinnacle". And so was born the Great Ladder of Life (which got its own Latin soubriquet, the Scala Naturae).

In hindsight this was something of a mistake. If the said someone hadn't been so high on his own self-importance, he might have noticed that the tree seemed to fit the analogy rather better than the ladder. Nevertheless, 300 years of this idea was peddled, biology concerning itself with the 'higher' organisms such as Man, and the 'lower' organisms such as pondweed, worms, and Worse. It was also expedient in ensuring the lesser people of the world knew their place at the feet of Europeans.

And this continued even unto the present day, despite it being not so long after Mr Linnaeus that a certain naturalist came to the conclusion that life had evolved under the influence of natural selection.

Darwin noticed that you could draw a tree of life, in much the same way you can draw a family tree, linking up different groups according to how closely they are related to each other. This 2D picture ('flatland', a definite improvement over the 1D 'lineland' of the Scala Naturae), persists to this day, and is the basis of most modern classifications based on relatedness. Such 'family-trees' (and I use the term very loosely), are called phylograms or cladograms, and the study of the relatedness of organisms is called phylogenetics.

Numerical and evolutionary taxonomy

Back to classifying things. The first thought that might come to mind is classifying Life as 'thin things' like snakes and millipedes, versus everything else, or something similar. This is obviously absurd, but if you measure more and more characters of an organism, you should be able to pigeon-hole it in a less and less arbitrary way. This technique is called 'numerical taxonomy'. It's very useful for the initial classification of nondescript things like bacteria, but it doesn't really shed any light on how things are related. This is in many ways the point of numerical taxonomy; indeed, it can work for anything: I could produce a numerical taxonomy of belly button fluff, or of books if I were feeling more useful. There's nothing 'wrong' with numerical taxonomy, but there's a more satisfying way of classifying Life, based on how living things are related. There are two ways of doing this, and the first brings us back to the dreadful Scala Naturae.

We could classify organisms based on the great ladder of life with us at the top and bugs and worms at our feet. Such classification has long since been abandoned, but vestiges still persist in the technique of 'evolutionary taxonomy'. In this system, if an organism has fins and a backbone, it is a fish. A fish is therefore a fish. This accords with common sense, but it leads to nightmarish brain tangles over fossils and how to classify them. Given that amphibians evolved from a fish-like ancestor (one with a backbone and fins used for swimming), how on earth do you classify Ichthyostega, a back-boned creature with legs it used for swimming? Such 'transitional forms' are a problem for evolutionary taxonomists, as the system relies literally on 'grading' organisms: is it 'good' enough to be an amphibian yet, or does it need to be using its legs for walking before we exalt it to the status of first amphibian, and ancestor of us all? For more discussion, see Henry Gee's excellent book, Deep Time. The problem of ancestry and descent plagues the technique. Although organisms must have ancestors, we are in no position to point our finger at a fossil ape and say, "That is the most recent common ancestor of humans and chimps", despite the media's obsession with always calling any important new fossil find a 'missing link' and 'a direct ancestor' of some organism or other. We cannot possibly know this, as fossils aren't found with convenient genealogies on bits of papyrus next to them.

Everything on Earth is a cousin of everything else, which cladistics takes at face value, whilst evolutionary taxonomy tries to see into the abyssal depth of geological time and pick out ancestors and descendants. I couldn't tell my great grandfather's skull just by looking, let alone the fossilised remain of my Ur-grandfather-amphibian from the depths of a Devonian swamp. Similarly, humans beings are not just 'descended from apes'. We are apes. Chimps may (or may not) look a little more like our most recent common ancestor than we do, but we are not qualitatively different. The things we descend from were apes, so we must be too. This is another area where cladistics clashes with traditional taxonomy: the latter puts birds on the same footing as mammals and reptiles, even though birds are reptiles (dinosaurs to be precise). In traditional evolutionary taxonomies, if they're green and scaly, they're reptiles. In cladistics, reptiles are all the species in a particular branch of the evolutionary tree, both the green scaly ones, and the feathery ones that comprise part of the reptilian clade.

The true meaning of 'primitive'

There is a perfectly decent, scientific way of using this word, but many people use it with a sort of dismissive snarl, usually followed by the use of words like 'Lower Diptera' (that's mosquitos to you) and similar:

…in plants, the production of spores is a primitive adaptation. The production of seeds is derived from this state…

There's nothing wrong with that, as long as you realise primitive means nothing more than 'the same version of a character as your ancestors' (i.e. spores, not seeds for reproduction). 'Ancestral' is the preferred term these days, however. Mosses produce spores, which is a primitive method of reproduction in plants. However, this is what you will see far too often in biology text books:

…plants possessing spores are the primitive, lower plants. Seed bearing plants are the more advanced, higher plants…

In humans, melanin in the skin is probably an adaptation to withstanding intense sunlight near the equator. Humans originated near the equator, and it is likely that dark skin pigmentation is the primitive character state, i.e. the ancestors of all humans probably had dark skin. Just change the word 'plants' to 'humans', 'spores' to 'dark skin' and 'seeds' to 'pale skin', in the second statement and you'll see one of the interesting problems with applying the word 'primitive' to groups or species rather than to character states. Dismissing mosses as primitive (rather than just recognising that the ancestor of land plants probably reproduced with spores, rather than seeds), does not make you a bigot, but it's worth drawing this exact analogy to show you just what you are saying.

Mosses branched off from the rest of the land plants a very long time ago, and have evolved on their separate ways since. A few sorts of mosses developed their own 'derived' ('advanced') characters, such as vascular systems, which move water about inside them. Calling mosses primitive as a group, rather than just noting that their sexual habits are similar to those of the first land plants, is really only saying that you think mosses are more similar to the ancestors of land plants than other plants are. In itself this doesn't sound too bad, but who are you to judge? Largely, calling a group primitive seems to be little more than asking 'is it similar to humans or their crops?'. To point out that mosses have spores and generally lack vascular tissue, and then to draw the conclusion that they are therefore more similar to ancestral land plants than a daisy is, is to judge that there is something special about vascular tissue and seeds. Unless you think there really is something evolutionarily special about humans and their crops, then calling mosses as a group 'primitive' is largely meaningless. The tree of life is full of pairs of branches, one labelled 'primitive', the other labelled 'derived'. To pick out some of these pairs and to label everything (including all species possessing later derived characters: mosses didn't just sit there doing nothing for the last 300 million years), arising from one of the two twigs as 'primitive' is dubious at best.

It is possible that (some) mosses really are objectively more similar to the ancestors of land plants than daisies are, and maybe there is some way you could justify calling them 'primitive'. However, 'lower plants' and its ilk, are terms that should die. You really ought to get it out of your heads that we are somehow the highest form of life, and that anything different to us is more lowly to some degree or other.

Cladistics

So what's so great about cladistics? It claims that 'reptile' and 'fish' have no meaning, but that such inelegant groupings as 'land-vertebrates-coelacanths-and-lungfish' do. Even worse, land-vertebrates can't be classified intuitively as 'mammals, birds, amphibians or reptiles' but could be classified as 'mammals, tortoises, crocodiles-and-birds, snakes-and-lizards or amphibians'. What on earth could such a system have going for it? Cladistics performed on just living organisms actually draws the evolutionary tree as it occurred (at least in theory). Classifying cladistically also means no tangles over ancestry, 'transitional forms', or trying to impose an order based on hindsight. It means no problems with arbitrary grading of organisms, and fewer problems making workable classifications that include extinct and fossil organisms.

Sharks and other cartilaginous, jawed fish.	Lionfish and other ray-finned bony fish	Coelacanths and other lobe-finned fishes.	Cats and dinosaurs, and other limbed, land-dwelling fish.
Sharks are an 'outgroup', the group most distantly related to the others.	'Proper (bony) fish' are related to those two → but not more closely to the coelacanth just because they both look fishy.	These two belong together in the same 'clade', because they share a number of novel features (like muscular limbs).
Still the outgroup.	The proper fish and the {coelacanths, cats} clade belong together, but more distantly.
And finally, the shark gets a look in {sharks, {proper-fish, {coelacanths, cats}} .

Groups that can be paired off because they are closely related, like the coelacanth and the molerats, form a 'clade', which is the name for the entirety of a branch (of any size) on the tree of life. In cladistics, only clades can have names, and these names apply to everything in that clade. I assume everyone is happy calling the 'proper (bony) fish' fish. Third from left is a coelacanth. It is more closely related to the molerats than it is to the lionfish, as evidence by the prescence of derived features not found in the bony fish, such as lobe fins and funny breathing apparatus. Now, you have two choices.

Either the coelacanth is a fish, in which case, you have to accept that molerats are just very odd fish too, because this means you want the word fish to apply to the entire clade {proper-fish, {coelacanths, molerats}}.

Alternatively, the coelacanth is not a fish, in which case sharks cannot be fish either, because you are forced to make the word fish to apply just to the clade {proper-fish}. You cannot define a clade by exclusion. Fish cannot mean 'the whole clade {sharks, {proper-fish, {coelacanths, molerats}}}, except molerats. The same situation applies to reptiles: birds are more closely related to crocodiles than crocodiles are to snakes. Hence either the reptile clade includes the birds, or it is a meaningless term. The choice is yours.

Another useful thing is that cladistics tries to discover how things are related without making assumptions about what organisms did with their bits. All it's interested in is which features are recently acquired, not which features are in some way 'important' to the evolutionary history of the group. The reason this is important is because if you assume the 'story' of human beings to be one about becoming upright walkers, then this will cloud your judgement as you try to classify fossils that seem to be related to us. Because your classification contains assumptions about what legs were adapted for, it makes it intellectually indefensible to use your classification to investigate how many times the upright gait evolved, and so on. It's much more sensible to make your classification without any preconceptions, and then it may be possible to use it to find out new things. For example, take a human, a dog, and a wolf. I bet if you were to draw a cladogram ('family tree', but don't let cladists hear you call it that!) of our relationships, you'd come up with something like this:

So how do we choose between these two models? Both contain just two assumptions, that is they are equally 'parsimonious'. This is where cladistics gets into a bit of trouble with the mathematically purist brigade, because cladistics 'just' chooses a species it thinks is less closely related to all the other species in the cladogram than any of them are to each other. This species is called an 'outgroup'. The problem is that there is no way to know for certain that your outgroup really is less closely related. However, all other taxonomic schools have troubles of arbitrariness inbuilt into them, and at least you can have debate and self-correction in cladistic analysis.

For my outgroup, I chose a frog. and the two trees now look like this:

All these explanations are less parsimonious than the first, so we'd accept the model you all instinctively thought was right in the first place, as our working hypothesis. It seems a bit long-winded, but we now have a classification that seems to show doggy features only evolved once, without having to assume that dog features were something important in the history of dog evolution. You could now start laying emphasis on the importance of dogginess to the ecology and success of dogs and foxes and wolves. If you discounted the second classification without some thought, you might well construct an entire theory of dog ecology based on the importance of the waggly tail and wet nose, only to have someone point out that the common ancestor of all mammals looked like a dog, and that the loss of the waggly tail in humans was the interesting evolutionary event. Such problems have actually appeared in real life, in the arguments over the origins of birds, flowering plants and eukaryotes.

The web of life

There has been an underlying assumption - a grubby secret if you will - throughout this essay. So far we've only been looking at species like dogs and wolves and people, which are conveniently unable to have sex with other species, which makes the lines on our cladograms nice and thick and clean.

Unfortunately, there are rather a lot of edge-cases and other problems:

• Bacterial plasmids.
• Mitochondria and chloroplasts (endosymbionts).
• Viruses.
• Bacteria in general

These problems all boil down to one thing: species and individuals and groups are not the units of natural selection. It is not they who evolve, but genes, and really, we should be looking at the evolution of genes, not species. Indeed, this is often what we usually do when we reconstruct cladograms: we sequence genes, and use these to construct our trees.

Genes are badly behaved. Any gene that can get itself replicated will, and it doesn't care about how easy or otherwise that makes our job reconstructing the family tree of life. Now for the most part, especially in big, sexually reproducing species like ours, most genes are quite well behaved, and are 'happy' to be inherited from parents to offspring. However, the mitochondria I mentioned earlier are the remains of bacteria that have been associated with our cells for a couple of billion years. The 'human' genome is not just a pile of 'our' genes, it is also a small pile of bacterial genes too, from a completely different branch of the tree of life. And that's not all. In sequencing genomes, we have found that some genes in two widely separated species are very similar, while the rest of their genes are completely different, giving evidence that genes have transferred from one species to another without using the usual sperm & egg route. These crossovers come under the term 'reticulate evolution' (reticulate means 'net like').

For example, our genomes are riddled with the dead remains of viruses, and it's more than likely that viruses are in fact escaped genes. You never see viruses on a tree of life, because usually we don't really know where they come from, and they don't fit nicely into our scheme anyway. Likewise, bacteria from quite different groups on the tree of life are capable of transferring genes between one another, hence the spread of antibiotic resistance from species to species, and maybe even the evolution of photosynthesis. These genes are usually carried on plasmids, which are not really part of the bacterial genome, nor capable of existing on their own. GM technologies further add to this 'cobwebbing' of the tree of life.

These intracellular parasites, and genes, and symbionts have as much right to be considered part of an organism's genome as any other gene, but their origins and their everyday lives may involve some transfer of genes 'sideways', not along the usual parent to offspring route.

If you consider the lines on the cladograms to be thick bundles of genes that are associated with each other into an organism's genome, you can consider the plasmids and viruses infecting several different species to be frayed, cobwebby bits around the edges. The mitochondria and chloroplasts are thin ropes connecting far distant parts of the tree, burrowing ever deeper into the core of the bundles as their genes are moved from inside the endosymbiont into the nucleus of the host cell over vast tracts of time.

I hope this foray into the musty world of taxonomy hasn't bored you. If it has, just go and get angry at something more controversial instead. If it hasn't, the actual Tree of Life (creationists need not apply) is at the University of Arizona: it's a distributed web project, with hyperlinks taking you to places hosting different bits of the messy bush we call Life.