Evolving Thoughts


This is the first in an irregular series of basic concepts in science, that I suggested to the Seed Bloggers we might do from time to time. If anyone wants to suggest a revision, because I got it wrong or am unclear, make a comment – this will be revised to make sure it is OK.

: this term of art is a new one in biological systematics, or the science of classification, or taxonomy. The word has given its name to a new science of classification: cladistics, which is properly known as phylogenetic systematics.

A clade is, simply expressed, any branch (Greek: klados) of the evolutionary tree which is separated from the rest of the tree by a single cut. Any branch, however large or small, that is cut off this way is monophyletic, or of a single origin. Monophyly is, under cladistic terminology, the property of a single species and all of its descendant species.

In older evolutionary taxonomy, groups of taxa (singular taxon, from the Greek taxis, meaning “order”) were often defined in terms of some shared trait or habitat. Hence, evolutionists like George Simpson would refer to the “bird grade” or “flight niche” and define birds as organisms with feathers that could fly, or whose ancestors could fly. This type of classification rests on what is now called a grade. Gradistic classification is problematic because it depends on the choice of trait, and because evolution doesn’t always retain the same traits in descendants. In the end, gradist classifications became a matter of duelling authorities.

In cladistics, the only natural grouping is based on monophyly – that is, the only natural kind of group of organisms is a clade. Cladists generate cladograms, which look like trees, but they are not exactly the same as the evolutionary tree of life. For a start, they are made for restricted groups – one way to generate them is to pick a group of probably related species and then pick an outgroup taxon, which gives the cladogram a “root” or base of the tree, and see how many traits (called “characters”) each share. The end result looks like this:


Here, species B and C are more closely related to each other than either is to A. B and C form a clade, relative to the outgroup species A. So, is this an evolutionary tree? Not yet. B and C might be later stages of A in a process called anagenesis (which means “origins from”), or A might have evolved from its common ancestor with B and C, and so on. In fact there are exponentially (more accurately, factorially) many possible historical trees for a set of taxa in a cladogram. So a cladogram is a way of testing historical scenarios, but isn’t, without other information like fossil position in the strata, or theories about evolutionary dynamics, an evolutionary tree itself.

How are cladograms formed? A systematist (one who does systematics) will try to select the most informative characters for analysis. This means getting rid of characters that might have evolved separately, such as flight (or else flying squirrels, bats and flying insects will be more closely related to birds than their probable ancestors, therapod dinosaurs). These homoplasies, as they are called, form grades, not clades.

Cladistics relies on the fact that evolving taxa tend to retain many traits no matter how much specific evolution they undergo. So we mammals are all warmblooded, have backbones, and feed our young milk, no matter if we fly, swim, or climb trees. Such traits come in two kinds, and here is where cladistics gets going: one is the ancestral or underived character (like having a vertebral column), and another is having a special or derived character which has undergone evolution. A group of organisms that has, say colour vision when its nearest outgroup relatives have bichromatic (two colour) vision is likely to have evolved it from the ancestral two-colour form. This is prima facie evidence that they are all descended from a species that evolved colour vision.

Of course, one character is not enough, because organisms evolved in a mosaic manner, with different organs and traits evolving differently at different rates, so cladistics relies on many characters, and the end result is what is called a “consensus tree”, in which lots of lines of evidence lead to a general set of cladograms, which may differ in detail but overall give you the “phylogenetic signal”.

Cladists use a specialised set of terms that were invented, for clarity if not ease of comprehension, by the founder of the school, Willi Hennig, back in 1950 (translated into English in 1966). An underived character is a plesiomorphy, from the Greek for “primitive form”. A derived character is an apomorphy, for “from-form”. A unique apomorphy is an autapomorphy, which only members of that taxon have (including a single species), while a shared apomorphy is a synapomorphy, and it unites a clade of more than one species.

Cladism is a relative classification. What is an underived trait in one part of the cladogram can be a derived trait in a wider cladogram, because all traits evolve from some previous trait. Moreover, what is an autapomorphy for one group (say, monkeys) can be a synapomorphy for a wider group (say, primates). Cladism lacks absolute ranks and states, unlike the older Linnaean classification which forces taxa into a set number of taxonomic levels.*

It helps, I find, to think of clades as being Venn diagrams, and to think of those as being cross-sections, as it were, through the evolutionary tree. A clade can be a set of organisms, but that set can also be a proper subset of another set, and can contain sets inside it. Only a proper set – one which does not intersect with any other but wholly contains or does not contain at all other sets of taxa – is thought to be a “natural” group. Two kinds of unnatural groups in cladism are polyphyletic (many origins, that is to say, many cuts of the branches) groups, or paraphyletic (not containing all the descendants of the original cut, like taking humans out of the ape clade) groups. They look like this:


Here it is as a Venn diagram:


“Invertebrata” (blue area of the Venn diagram; animals without backbones) are paraphyletic, because not all descendants of their root taxon are included. So too are “Pisces” (green are; fishes), and “Reptilia” (orange area; reptiles).

“Homothermia” (grey area; animals that have warm blood) is a polyphyletic group, because internal temperature regulation evolved more than once. The same is for the “Crossopterygians” (dark green; extinct fishes), which have living ancestors not included in their group, and which cover the root of more than one group.

[The little crosses indicate taxa that are extinct. A cladogram can be made on extinct and extant species, and in a simple cladogram, no distinction is made - they are all at the top of the cladogram. But you often get "mixed" cladograms like this one.]

A simple rule in cladistics is that once a member of a group, always a member of that group. So we are, as the cladist will tell you, bony fishes. Likewise, if some group is formed that has, say, four limbs (Tetrapoda), the fact that some member, like snakes or whales, has secondarily lost some or all limbs doesn’t mean they are no longer part of that group. They simply have a “derived form of being four-legged” (i.e., their apomorphies are, respectively, limbs = 0 and limbs = 2). It sounds unintuitive, but if you think it through, it makes sense as an evolutionary perspective. If you look at the Venn diagram and ignore the colours, then each group (except “Crossopterygians” and “Homothermians”) is a proper subset of its enclosing group, and so we are “invertebrates” and “fishes” because one of our ancestors or more was an invertebrate and a fish. This is why these excluding classifications are not “natural” in systematics. Invertebrates is just what is left over from Animalia once the vertebrates have been taken out.

So now, when someone talks about cladistics, you will know what they are talking about (or get a feeling whether they know what they’re talking about)…

* Although Linnaean classification has been forced to add sub- and super- ranks to include the number of levels in each group that taxonomists want to talk about, and there is no reason not to make a Linnaean taxon monophyletic, which is what has been happening over the past 30 years or so.


  1. #1 Torbjörn Larsson
    January 13, 2007

    Biological classifications tickle me pink, first because of the childish urge to be able to name everything I point at, second because I can easily visit Linnaeus gardens in Uppsala where he had a chair of medicine and later botany.

    But cladistics seems especially fun being both systematic (with fun methods!) and historical.

    “In fact there are exponentially many possible historical trees for a set of taxa in a cladogram. So a cladogram is a way of testing historical scenarios, but isn’t, without other information like fossil position in the strata, or theories about evolutionary dynamics, an evolutionary tree itself.”

    Ah! That cleared up a basic confusion of mine.

    Btw, good idea proposing SB blogging basic concepts. If it is not done overly often it is a great addition to the menu.

    PS. If you revise, you may consider making the effort of correcting “that were invented, [...] but the founder of the school,”. Since the rest was so nicely done. DS

  2. #2 John Wilkins
    January 13, 2007

    Thanks, and done.

  3. #3 John Wilkins
    January 13, 2007

    Incidentally, the number of trees is


    so if there are 20 taxa, there are 3.36154E+53 trees.

  4. #4 Torbjörn Larsson
    January 13, 2007

    In that case I will nitpick, since the factorial increases faster than exponential. Checking: “As n grows, the factorial n! becomes larger than all polynomials and exponential functions in n.” ( http://en.wikipedia.org/wiki/Factorial ) Eyeballing your expression the numerator will make it so, if I’m not mistaken.

    That is a forest of trees!

  5. #5 Matt
    January 13, 2007

    Very nice post- every time I thought you had missed a point- there it was in the next paragraph or section.

    Nice review for the evolutionary biologists, but I fear that some the jargon will lose others..

  6. #6 Kele
    January 13, 2007

    This was very helpful in clearing up some confusion of mine. I hope there will be more! Thanks.

  7. #7 RyanG
    January 13, 2007

    I agree with Matt that this might lose laymen. Assuming I understand clades right, you could probably throw in the following example:

    Chimpanzees and Bonobos form a clade
    Chimpanzees, Bonobos, and Humans form a clade
    Chimpanzees, Bonobos, Humans, and Gorillas form a clade (African apes)
    Chimpanzees, Bonobos, and Gorillas do not form a clade, because humans are excluded.

    To simplify things even further, you could use a family analogy. Me and my siblings form a clade, me and my cousins do not form a clade unless my siblings are included. Assuming I understood the concept right.

  8. #8 Torbjörn Larsson
    January 13, 2007

    Perhaps it is just me, but I’m a layman (perhaps prepared, though) and I believe this post illustrated the concepts very well.

    What I will have problems with is remembering the terms, even with Venn diagrams. (Though Wilkins roots them well in ordinary language.) Specific examples illustrating them is then a good thing. But I believe a reader can come up with such.

    Btw, I forgot to tease Wilkins on his agnosticism. Crosses instead of X’s for eXtinct? In that case, the FSM demands S’s for Spaghettified!

  9. #9 AndyS
    January 13, 2007

    Excellent. Love to see more like this from all the SciBlings. I like the “Basic Concepts” category you have here on Evolving Thoughts. Now, if we could just have a channel on ScienceBlogs that would serve to collect these sorts of posts….

    BTW, does the Australian version of Monophyly have a Boardwalk or is that renamed to Sydney Opera House or some such?

  10. #10 John Wilkins
    January 13, 2007

    Monophyly here has the usual properties: Fossil Bone Railway Station, Megapoda Street, Aves Avenue, and so on… There is a Get Out of Eurasia Free card, though, and a Wallace’s Line Railway Station, which you don’t find in the Old World or New World editions.

  11. #11 Jim Harrison
    January 14, 2007

    It will take some time to get use to the cladistic way of doing things in taxonomy, although the new approach isn’t really that new. How do you match up old taxa such as genera, families, and orders with the proliferation of names for clades? In the system recommended by Lecointre and Le Guyader in their book The Tree of Life, we show up in Deuterostoma/Pharyngotremata/Chordata/Myomerozoa/Craniata/
    Boreoeutheria/Euarchontoglires/Primates (sensu latu)/Euarchonta/Primates(sensu stricto)/ Haplorrhini/Similiformes/Catarrhini/Hominoidea/
    Hominoidae/Homininae/Hominini/Homo. The authors entertain the that absent human vanity, a case could be made for calling our species Pan sapiens or, in the alternative, we could refer to the chimps as Homo troglodytes and Homo paniscus.

  12. #12 John Wilkins
    January 14, 2007

    Technically, Homo has priority. Pan was introduced by Blumenbach because no right thinking Christian would ever want chimps and other apes in Homo, the way Linnaeus placed them.

    But, since the two chimps form a clade, they can be a separate genus from Homo (which now includes neanderthalensis, ergaster, erectus, and so forth, so it’s not a singleton genus).

  13. #13 W. Kiernan
    January 14, 2007

    Never even heard the word, sir. Not often anyone gives me a gift like this!

    It reminds me of Ouspensky and his “cast-off Is”.

  14. #14 Jonathan Vos Post
    January 15, 2007

    John Wilkins:

    Can you clarify your enumeration of trees?

    Labeled? Unlabeled? Rooted?

    Weisstein, Eric W. “Tree.” From MathWorld–A Wolfram Web Resource. http://mathworld.wolfram.com/Tree.html

    Also, since the term was used:

    Weisstein, Eric W. “Forest.” From MathWorld–A Wolfram Web Resource. http://mathworld.wolfram.com/Forest.html

  15. #15 rdb
    January 16, 2007

    The wikipedia article on Clade is a stub, if you wish to fill in more.

    In cladistics, a clade is a group of organisms consisting of a single common ancestor and all the descendants of that ancestor. Any such group is also considered a monophyletic group of organisms, and can be modelled by a cladogram, a diagram of the organisms in the form of a tree.

    A clade is a scientific hypothesis of evolutionary relationships among the organisms included in the analysis. A particular clade may therefore be supported or disproved by a subsequent analysis using a different set of data.

    Is the 3000 species Tree of Life a cladogram or something else?

  16. #16 John Wilkins
    January 16, 2007

    Jonathon, how mathematicians use a term, and how taxonomists use it, is related by loose rather than close technical agreement. The math of cladistics is standard, but a (cladistic) tree in taxonomy is labeled only at the terminal nodes. A phylogenetic tree, or evolutionary tree, is labeled at the internal nodes as well. In mathspeak, this might be the difference between cladograms and phylogenetic trees.

    “Forest” here has only vernacular meaning. The critical thing about cladograms is their topology, not their density. A cladogram can have three -> n terminal nodes. If the branching order is the same, then that is the same tree irrespective of the way it is displayed.

    The tree of life, rdb, is a cladogram, or ought to be. However, it actually gives ancestral names (in the sense that the more general taxa are thought to be species). However, if we found a fossil species that was close to the base of the clade in which it falls, we cannot say for sure if it is the actual ancestral species or not. So it appears in a cladogram as a “sister taxon” not as an ancestor (to put it in terms Jonathon might think appropriate, it would be a terminal not internal node of the tree).

  17. #17 saikou ba jabbie
    March 3, 2008

    please can tell me the animal that can fly can swim and can olimb trees