The Utility of Phenetics

I talked about phenetic methodology in this post. I was relatively charitable towards phenetics, which has led to some readers thinking that I endorse phenetics, when I’m a dye-in-the-wool strict cladist, because cladistics is the only method that makes sense of biodiversity, as explained a few paragraphs down.

The previous post was a mere methodological guide, written because, like it or not, phenetic methods are still common and useful for practical reasons – what are morphospecies other than phenetic-based species identifications? As I said in the post, molecular phylogenetics is riddled with phenetic concepts (see Mallet (1995)). If a researcher chooses to separate taxonomy (classification of organisms) from systematics of organisms (classification with respect to evolutionary history), then phenetics is a valid way to go about things – a classification doesn’t necessarily have to reflect evolutionary history. Botanists (Rieseberg et al., 2006) and microbiologists (Rosselló-Mora & Amann, 2001) frequently use phenetic concepts; let’s not even get started with the zoological literature.

This separation of systematics and taxonomy isn’t my own invention, it’s actually a deep-seated divide that traces its roots back to the advent of evolutionary theory. Modern taxonomy emerged with Linnaeus and the core principles have remained essentially unchanged since then, even though the greatest revolution in biology occurred in the meantime that should have rendered Linnaeus’s classification of nature into types created by God senseless in a scientific sense. But it nonetheless remained.

It wasn’t until the aftermath of the Modern Synthesis that the insights of evolution were brought to bear on species classification. It’s out of that scrutiny that the modern debate came out between cladists and pheneticists. The cladists are those who think classifications ought to reflect evolutionary history; pheneticists think it doesn’t have to. A middle-of-the-road view combining both of them, evolutionary taxonomy, also emerged. As mentioned, I side with the cladists, because phenetics is in no way objective and leads to screw-ups in the long-run – the continued use of Reptilia not including the birds is an example. The fact that phenetics is subjective also leads to a difficulty in justifying higher taxa. In cladistics, a hierarchy emerges naturally through the nesting of monophyletic taxa; in phenetics, the hierarchy is artificial.

So basically, my previous post was in no way an endorsement of phenetic philosophy as a guide to phylogeny (the elucidation of the evolutionary history of the taxa studied). Like the majority of biologists, I see such a viewpoint as deeply misguided (Ridley, 1986), as do philosophers of biology (Sober, 1993). For example, Matthen (2009) simply does away with phenetics using just two encapsulating sentences: “Pheneticism is a mistake. It originates in the harmless but imprecise idea that conspecific organisms resemble each other.” My not getting bothered too much by phenetics is merely an acceptance of the practicality of clustering similar objects together; after all, I can tell you as a fact that it’s very, very rare for taxonomists to actually check whether their new species interbreeds with their purported sister species, as a cladist would have you do (if following the biological species concept).

It must also be said that forcing a phylogenetic view on taxonomy isn’t only impractical but also miguided, for the only reason that the history of a taxon cannot always be inferred. Without that, there is no way to do cladistics. Admittedly, this is now not such a big worry with molecular phylogenetics, but even ignoring all the problems with molecular phylogenetics, one simply cannot expect every single species to get sequenced even partially – not with the measly funding taxonomy gets nowadays.

I will go one step further and say that someone attacking modern pheneticism as flawed for phylogenetic purposes is attacking a strawman; no pheneticist (or pheneticist-like, since phenetics is now a dirty word) I’ve read from the past couple of decades has advocated such a use. If I’m wrong, please link me to a paper/book. As far as I know, pheneticists only state the practical case – practical for pure taxonomy, or for other types of classification (by toxicity, by virulence, whatever – think of how bacteria are grouped as Gram-positive or Gram-negative, this has no phylogenetic basis at all*), and while in the earlier days of numerical taxonomy, an objective phylogeny was thought to have been the goal (Sokal & Sneath, 1963), this is now rightly left to the phylogeneticists. In fact, from a technical viewpoint, it’s just about impossible for phenetic methods to uncover historical relationships, since they can’t even deliver branch lengths of the produced trees (Penny, 1982) – this is akin to having a line of best fit without the data points.

*: Commenter points out that this may not be true. There is modern work supporting Gram-negative bacteria as a monophyletic group. See Gupta (2011) and especially the references to Gupta’s other papers in there. I have not read up on this stuff in too much detail to judge, so it’s only fair I present all sides of the story. Apologies to anyone that got misled.

References:

Mallet J. 1995. A species definition for the modern synthesis. TrEE 10, 294-299.

Matthen M. 2009. Chicken, Eggs, and Speciation. Noûs 43, 94-115.

Penny D. 1982. Towards a basis for classification: the incompleteness of distance measures, incompatibility analysis and phenetic classification. Journal of Theoretical Biology 96, 129-142.

Ridley M. 1986. Evolution and classification: the reformation of cladism.

Rieseberg LH, Wood TE & Baack EJ. 2006. The nature of plant species. Nature 440, 524-527.

Rosselló-Mora R & Amann R. 2001. The species concept for prokaryotes. FEMS Microbiology Reviews 25, 39-67.

Sober E. 1993. Philosophy of biology.

Sokal R & Sneath P. 1963. Numerical taxonomy: the principles and practice of numerical classification.

Leave a Reply