Me too! I'm a bony fish.

This week the Blogosphere, Twitterspace, and at least one website have been giving some attention to taxonomic principles.

It all began with an article by Vasko Kohlmayer in the Washington Times that asked "Is Richard Dawkins an ape?".  The correct answer, according to that article, is "No", despite Dawkins's statement that he is indeed an ape.  Kohlmayer went on to say Dawkins can't be an ape because he has all the usual human attributes, and we humans have these because we were created by a god.

That ought to have been the end of it.  The usual scientists' response to such bronze age superstition is to point and laugh and move on.  But biblical creationists aren't the only people who don't like to think of humans as apes.  Quite a few biologists as well are unhappy with the idea that species or groups that are very different from their ancestors should nevertheless be classified with them.  For instance, humans are different from apes, birds from dinosaurs, and termites from cockroaches.  Technically speaking, when we separate out part of a natural group because it looks very different (humans, birds, termites), the left over remnants of the natural group (apes, dinosaurs, cockroaches) are said to be paraphyletic.

So scientists responded to Kohlmayer in different ways.

First, anthropologist John Hawks argued that we use folk taxonomy terms like ape in everyday parlance to mean something other than the formal Latinized names that taxonomists use, and so it's OK for these groups to be paraphyletic because that's what ordinary people understand.  I can see where he's coming from, and not long ago I sort of agreed.

That's because I was aware that Kew botanist Dick Brummitt (2006) had made an argument for recognising paraphyletic groups in an article called "Am I a bony fish?".  Brummitt argued that paraphyletic groups, like fish, have biological meaning and that including mammals and other tetrapods among the fish in order to make fish monophyletic takes that meaning away.

I disagree absolutely with Brummitt that we ought to accept paraphyletic groups in biological taxonomy.  Rather, I think his "I am not a fish" argument is a clever rhetorical device to use in an argument against cladistic taxonomy.  It sounds so self-evident, so intuitively correct, and it can win the day with people who (1) aren't familiar with phylogenetics, and (2) aren't used to stepping outside their comfort zones (something scientists ought to do).  I do agree with Hawks that "fish", like "plant", "fruit", "bug", and even "animal", has a non-scientific meaning that can be different from its scientific meaning. But biologically, the common ancestor of all fish was also my ancestor, and that makes me a fish.  That may sound ridiculous to a non-biologist, and I was inclined to agree with Hawks because it's harder to deny the logic if we simply change the words and ask, "am I a bony vertebrate" instead of fish (the correct term for the bony vertebrates is Osteichthyes).   

But although in part we're arguing about the meaning of words, we're also arguing about whether or not evolution is important.  Creationists say it didn't happen, but even some scientists who accept evolution still think we ought to classify ourselves and other living things according to what we look like, giving less weight to what we're related to.

Evolutionary biologist Jerry Coyne summarised the controversy very well in two posts on his website (here and here). Then another biologist, Brian Switek, weighed in with "I am an ape, and I'm also a fish".  Now I'm ready to put my hand (pectoral fin) up and say, "me too".

So why do I think I am a fish?  Bony vertebrates are defined by their ossified bones and traditionally include fish and tetrapods (amphibians, reptiles, mammals, and birds). The bony fish, as people historically have defined them, have an ossified internal skeleton of bones, but also scales, fins, streamlined bodies, and gills.  If we divide the bony vertebrates into two groups, bony fish and tetrapods, we can make some decidedly odd-sounding statements about evolution, e.g., a coelacanth (fish), is more closely related to a human (tetrapod) than it is to a salmon (fish).  How can a fish be more closely related to a non-fish than it is to another fish?  In a natural group, every member is related more closely to every other member than it is to any non-member.  So if fish doesn't include tetrapods, clearly it isn't a natural group.  And in dividing the bony vertebrates up, we can use natural groups like the ray-finned fish, the lobe-finned fish, and the tetrapods.

What's more, if you look at every fishy feature, like scales, gill arches, and fins, you can find those structures modified in your own body (teeth, jaw & inner ear bones, arms and legs) and so our fishiness is still present in a modified form.  Even the fish's repeated body segments, so obvious when we eat a trout, are present but rather hidden in our own bodies (when I got shingles, it affected just one segment of my body because the chickenpox virus attacks from a single segmental nerve). Yet if you were to take an unrelated group, say insects or echinoderms, you'd not find in yourself such modifications of their characteristics (exoskeletons & spiracles; tube feet).  So it makes sense to think of ourselves as modified fish, modified tetrapods, modified mammals, and modified apes, but full members of all those groups nevertheless.  Neil Shubin's "Your inner fish" covers it well and this is why Dobzhansky's statement "Nothing in Biology Makes Sense Except in the Light of Evolution" is held so dear by biologists.  (However I was quite surprised to read about how that statement came to be written, something biologists who quote it probably aren't aware of).

It's not only about trying to build a classification that reflects evolution.  Another reason is that science should be objective.  If we propose a classification, we can test it by asking how well it aligns with natural groupings.  Natural groupings are those that arose by a natural process, the branching pattern of evolution.  Cladistic taxonomists reject groups that don't pass that test, or more accurately they redraw the boundaries until the groups do pass the test (e.g., by adding humans to apes or tetrapods to fish).

I'm a botanist: why do I care about fish?  Well, first I should care because I'm a biologist, but secondly also because exactly the same arguments occur in botany.  The classification of plants has been going through quite a bit of turmoil lately for the same reasons.  Family membership and names have changed to reflect the desire to classify only natural groups, so hebes and foxgloves have been taken out of the Scrophulariaceae and put in Plantaginaceae with their close relatives the plantains (Plantago), which look very different because their flowers are adapted for wind pollination.  The same thing has happened at the rank of genus: the genus Rhododendron now includes the azaleas, so the names of azaleas now begin with Rhododendron.  That's upsetting for people who have to learn new names, and also for people who are tuned in to the differences between rhododendrons and azaleas, rather than the relationships.

At heart, I also think some taxonomists are uncomfortable with the notion of testing their ideas about classifications.  They are quite coy about this in print, but in lectures, emails, and conversations I often hear "obviously ...", or "it's totally different", or "only a fool would ...", phrases that try to pre-empt a challenge or to beg the question.  I hear pleas for taxonomic judgement to over-ride objectivity, and for intuition to trump tested hypothesis.

My own part in this has been transferring New Zealand hebes and their relatives back into the genus Veronica.  Our hebes aren't a separately evolved (and certainly not created) group that has nothing to do with the northern veronicas.  Rather, they evolved from ancestors that were veronicas, and we can understand how they evolved when we begin to see them as veronicas that have adapted to life in New Zealand.  Most of them are shrubs and have tubular flowers and capsules that are flattened parallel to the partition, but some are still quite like northern veronicas.  To reflect the fact of their origin, and it is a fact, I've provided new names (or resurrected old ones) for all New Zealand's hebes and their relatives; these names all start with Veronica.

Veronica lilliputiana, a New Zealand species that has non-woody stems, blue flowers, and short corolla tubes.

Veronica parviflora, a more typical New Zealand hebe.

I'll give the last word to Professor David Mabberley.  I've just bought his wonderful and authoritative book and was delighted to see this issue discussed in the introduction, including mention of Veronica.  Mabberley is wonderfully uncompromising in his support for classifying only natural groupings.

Writing about modern DNA-based phylogenetics, he says, "I believe that somewhat condescending, tut-tutting scientists and others are seriously underestimating the intellect of the rest of society's building on these bases to appreciate and assimilate the results from modern work. As Darwin predicted, 'Our classifications will come to be, so far as they can be made, genealogies'."

And: "To maintain these and similar examples as separate genera or families, effectively picking holes in the (monophyletic) generic fabric and denying us the framework within which we can not only begin to understand ecological-evolutionary shifts but also marvel at the workings of evolution itself, is to maintain the holey relic as paraphyletic."

And yes, I think the spelling "holey" was deliberate.