Sunday, February 17, 2013

Why are botanists more opposed to cladism than zoologists?

One interesting thing a colleague once said when we were discussing continuing opposition to phylogenetic systematics was that that opposition is stronger among botanists than among zoologists - if there is any left in zoology at all. Now obviously I have much more contact with other botanists on a daily basis than with zoologists, but that observation rings true. But why would it be so?

One might suggest that Willi Hennig himself having been a zoologist could have something to do with it, but that seems a bit spurious. I favor a different explanation, and it has to do with the way ranks are treated in the two fields.

As you may know, systematics traditionally classifies biological diversity into nested taxa at different Linnean ranks. The basic unit of systematics is the species. (Although there are various ranks below species, like variety or subspecies, there does not seem to exist a satisfactory objective criterion for their delimitation comparable to monophyly for supraspecific taxa and the various empirical species concepts for species themselves.) Above species and in ascending order, the principal ranks are genus, family, order, class, division/phylum, kingdom, domain. For ourselves, for example, the classification looks like this:

Domain Eukarya (eukaryotes)
    Kingdom Animalia (animals)
        Phylum Chordata
            Class Mammalia (mammals)
                Order Primates
                    Family Hominidae (great apes)
                        Genus Homo
                            Species Homo sapiens (humans)

In addition, there can be many intermediate ranks such as superorder, subfamily, tribe or section, and due to the massive number of species on this planet, which Linnaeus was of course unaware of when he came up with the first five principal ranks, they are badly needed. Indeed even with them there are possibly not enough ranks for a comprehensive classification of all life.

Now one of the differences between zoological and botanical nomenclature is that in botany many taxon names have defined endings indicating their rank. Orders always have names ending in "-ales", families end in "-aceae", subfamilies in "-oideae", tribes in "-eae" and subtribes in "-iinae". That is not the case in zoology.

So when our understanding of the relationships in a group of plants changes, and we need to sink one family into another, it means that the ending of the group name also has to change. For example, the former Epacridaceae, a mostly Australian group of heaths, are nested deeply inside the heath family Ericaceae and so the two had to be united. The Epacrids themselves are still a natural group, and nothing whatsoever has changed about them. What has changed are only the Ericaceae, which have become slightly bigger. But you do not get to call the Epacrids Epacridaceae any more, and so it feels as if something has changed about them. If botany were like zoology, nobody would care and you could still use the old name, and my suspicion is that that makes it easier for a zoologist to accept such a new classification.

Another issue is that, at least from what I can tell, botanists seem to take the ranks more seriously, and this may also simply result from those defined name endings. Floras and field guides are always carefully organized by plant family, and many botanists think in terms of families as the principal boxes by which they organize their knowledge of the plant kingdom. When I went through university, we had to learn the diagnostic characters of, very specifically, several plant families as opposed to the diagnostic characters of several important plant groups of varying rank. And that means that when family circumscriptions change, it feels like a very big deal to a botanist.

It isn't. All ranks above species are completely arbitrary and have no scientific meaning. A taxon has to be monophyletic to be valid, but whether a monophyletic group is ranked as a family or as a subfamily is entirely up to personal taste (as long as subfamilies are ranked below families, of course). There is nothing whatsoever in nature that tells us what to prefer, no objective and reproducible method to decide, no empirically discoverable reality behind "familyness" or "genusness". The rank is arbitrary.

But if you have grown up with these endings it feels like a significant difference whether something is Fabaceae or "merely" Faboideae, and if you have been trained using a flora that was organized into separate sections for Epacridaceae and Ericaceae it feels like a significant difference when the former suddenly disappears. It is the usual human problem of making a fetish of the concept while tuning out the reality it was meant to describe.

I increasingly feel that something is to be said for the PhyloCode approach of simply scrapping ranks altogether and thinking only in terms of clades nested inside of other clades. Having floras or field guides organized strictly after nested clades would be a good start. But the Linnean ranks are so entrenched in our thinking and in centuries of legacy literature that it will be hard to make that transition. In particular, it would be very disruptive to do without genera, not least because the genus is the first half of every species name.


  1. Just a hobby botanist at most. I have the idea that a fair number of flowering plant species are of hybrid origin. I've seen different percentage numbers, up to, as I recall, some 25%. I have thought this one reason that botanists have been slow to accept cladistics. Cladistics, as I have understood it, does not deal well with hybrid origin groups, ie, non monophyletic.

    Ichthyologists, at least, are fairly uniform in using an -iformes ending for orders, and -idae for families, of course. In the groups that I follow, I think there has been an increase in families and orders, so we have not seen the consoldiation problem so much.

  2. The question is whether the hybridogenic species are so many, so persistent (as opposed to high-level polyploidy usually being a dead end) and derived from parents so distant as to make species relationships too net-like for phylogenetic systematics to work. Looking at the reproductive boundaries across the plant kingdom, I strongly doubt that. Especially if hybridogenic speciation mostly happens between very closely related species, you just have to "zoom" out a bit and you have a tree-like structure again. And most of the controversy does not focus on classifying the species of, say, madly hybridizing Mentha section Mentha but on higher level relationships where hybridization is not an issue.

    25% also seems very high at least based on what I have seen so far. Is it possible that this number was advanced by people unaware that they would have to rule out incomplete lineage sorting as an alternative explanation for incongruent gene trees?

    Thanks for the info on ichtyology; I was completely unaware of that. Must admit that vertebrates never interested me as much as arthropods and molluscs.

  3. When I first got seriously interested in plants, first cacti and various succulents, then amaryllis,I expected to readily grasp botanical taxonomy. It turns out that plants are not just green animals which do not run away. Rather they pose a number of situations which I have never encountered in fish. I now think there are strong differences between plants and animals, and a corresponding difference in how botanists and zoologists think about their subjects.

    I'm posting as anonymous as I don't see exactly how to post by name. Jim Thomerson

  4. I am not saying there are no differences, but the only group that has plausibly been argued to potentially lack phylogenetic structure are bacteria because they steal each others genes so often - the whole hyperbolic "Darwin was wrong" and "bush of life instead of tree of life" discussion from a short while back. (Not sure whether the same applies to Archaea...)

    It is also curious how this argument is supposed to proceed: according to certain colleagues, we should accept paraphyletic taxa because of hybridogenic speciation. But if there was, in plants, a net of life instead of a tree of life then there could be no paraphyletic taxa either. They cannot have it both ways. If there is a phylogeny, then whole branches of it are the only natural taxa. If there isn't a phylogeny, then there are no branches, so one cannot formally recognize partial branches either. In that case, all that remains is phenetics, no?

    Because commenters have so far been few and civilized, I have not made any restrictions on commenting. I assumed the Name/URL option under "comment as" would allow one to enter a name.