lumping, splitting, and the morphological species concept

A couple of weeks ago Marcus Wilson & I were down in Taranaki to run a Scholarship preparation day. During one of my sessions the biology students & I were discussing the concept of ‘species’. Most of you are probably familiar with the idea of a ‘biological species’, defined by its lack of ability to interbreed with members of other populations. Well, we agreed, that works just fine for living organisms, but what do you do about fossils? The answer is that we tend to rely on a ‘morphological species’ concept, identifying likely members of a species on the basis of structure & appearance (morphology).

Now, you might be aware that this is not necessarily reliable. Take our own moa species, for example. Up until quite recently, palaeontologists believed that there were around 20 species of these big flightless birds. Interestingly enough, moa species appeared to come in pairs – one big, one small, but otherwise morphologically very similar. This understanding was turned on its head when a team from the Allan Wilson Centre used ancient DNA techniques to look into the apparent relationship. Sex determination in birds is based on Z and W chromosomes, where males are ZZ and females are ZW (ie they’re ‘heterogametic’, whereas in mammals the XY males are the sex that produces two types of gametes), and the AWC team found a genetic marker on the sex chromosomes that allowed them to reliably distinguish between males & females in living birds. They then applied the same technique to ancient DNA from moa remains – & found that of the apparent pairs of species, the larger ‘species’ was in fact female, & the smaller, male. An elegant demonstration that the morphological species concept definitely has its limitations.

Now, tthat technique worked because it’s possible to obtain useable amounts of aDNA from moa remains. What about fossil species that are much older, like the dinosaurs?

Dinosaurs come in a range of sizes & body forms. One of the more extreme groups is the pachycephalosaurs – the dinos with a big domed head ornamented with bumps & spikes around the base of the dome (& also on their muzzles). Thanks to PZ I’ve found a rather nice video (below) on the possible resolution of a species puzzle in this group – basically, dino experts Jack Horner & Mark Goodwin are suggesting that three apparent species be ‘lumped’ into one (hence the title of my post. ‘Splitters’ tend to generate a new species name for each new fossil, while lumpers take the opposite approach). While on the face of it these animals do look quite dissimilar, it’s proposed that they represent a growth series, rather than reproductively isolated species (Horner & Miller, 2009 – you can read the PLoS One paper about it here.) The idea of a growth series in dinosaurs isn’t a novel hypothesis – they’ve previously presented evidence of the same thing in Triceratops.

The evidence in this case comes from an analysis of the structure of the animals’ skulls & of the bone from which they’re constructed. in the proposed adult (Pachycephalosaurus wyomingensis), the bones of the skull are fused & not particularly well vascularised, suggesting that it’s reached its adult size. At the other end of the spectrum, ‘Dracorex’ has unfused cranial plates that are well supplied with blood vessels – this animal was still actively growing. ‘Stygimoloch spinifer’ seems to sit between the two.

Craniohistological coronal sections of “Stygimoloch” (A–C) and Pachycephalosaurus (D–F) skulls, from Horner & Miller (2009), doi:10.1371/journal.pone.0007626.g008. A-C represent sections from ‘Stygimoloch’ skulls while D-F are from Pachycephalosaurus.

You can see the unfused suture between bones in ‘Stygimoloch‘ (A) – compare those with a similar section from Pachycephalosaurus in (D). Sections (B) and (E) show the differences in vascularisation between the two specimens.

And the various cranial protuberances, while different in form, are in pretty much the same places on each skull (below).

Cranial ontogenetic sequence of Pachycephalosaurus wyomingensis with morphological landmarks highlighted in color from Horner & Miller (2009), doi:10.1371/journal.pone.0007626.g003



This isn’t quite the whole story – as Brian Switek in 2007, when this hypothesis was first put forward, the evidence remains circumstantial. We don’t have a whole lot of remains from each ‘species’, and while ‘Dracorex’ is definitely juvenile, we can’t be sure that there isn’t some adult version out there somewhere in the North American fossil beds. Horner & Miller (2009) note that Pachycephalosaurus & ‘Stygimoloch‘ remains have been found in the Hell Creek Formation of Montana, so that does tell us the taxa were sympatric – if they were never found together in the same area, it would be much harder to argue that they belong to the same species. But this is how science operates – by proposing & testing hypotheses, and by incremental change as those hypotheses are accepted or rejected.

(Unfortunately, such is the way of taxonomy, if the ‘lumping’ recommendation is accepted then we’ll lose a wonderful species name: Dracorex hogwartsia – all you J.K.Rowling afficionados will know where that one came from!)

J.R.Horner & M.B.Goodwin (2009) Extreme cranial ontogeny in the Upper Cretaceous dinosaur Pachycephalosaurus. PLoS One 4(10) e7626 doi:10.1371/journal.pone.0007626



One thought on “lumping, splitting, and the morphological species concept”

  • The statement, “A species is what a competent taxonomist says it is.” still has considerable verisimillitude. We can have more information about a living species than a paleontologst can have about a fossil species, but there are still large unknowns and most species have not been studied with other than morphological methods.

Leave a Reply

Your email address will not be published. Required fields are marked *