Thanks to herr doktor bimler & the University's science librarian, I now have my hands on two copies of the paper I mentioned a couple of posts ago: Positive allometry & the prehistory of sexual selection (Tomkins et al., 2010). The term 'allometry' refers to the relationship between the size of an organism & the size of various parts of that organism. When scientists study allometry, they might do this for various stages in the growth of an individual, or they might compare different organisms of the same species, or individuals from separate species. 'Positive' allometry means that as body size increases, so does the size of whatever other feature's being examined. (The Panda's Thumb has an example of this, for body size: brain size ratios in primates.) The paper by Tomkins et al. looks at the spiny 'sails' on the backs of pelycosaurs & the crests on pterosaurs' heads, and suggests that differences in size between male & female individuals is related to sexual selection.
When I first saw the newspaper report on this article, I wondered: just how did the authors identify male & female pelycosaurs & pterosaurs? With modern reptiles size can be a cue, as can colours & crests & inflatable throat pouches. (Sexual dimorphism is widespread in reptiles, but unfortunately it's not consistent: in some groups the males are larger, but in others it's the females.) And of course, hypotheses about which sex you might be looking at can be confirmed by watching to see who's on top during mating. And also by dissection: the 'who's on top' method wouldn't work for some whiptail lizard species, where all individuals are female & reproduction is by parthenogenesis. In these species, individuals take on the 'male' or 'female' role during mating depending on the levels of oestrogen & testosterone in their blood. So I was interested to know how the authors determined whether they were looking at a male or female in dealing with any particular indivdiual.
The 'standard' explanation for pteranodon crests & pelycosaur sails is that they were related to thermoregulation. Pelycosaurs were unlikely to be endotherms, generating heat internally as a side-effect of a high metabolic rate, & so like modern reptiles would have had to bask in the sun to warm their bodies before dashing around chasing prey (or avoiding being prey). As Tomkins & his colleagues point out, a 'sail' of tissue supported by vertebral spines, with blood vessels running through it, could have warmed the animal's blood more rapidly - a bit like a solar panel used to heat water. It could equally have acted like a car radiator & shed excess heat. (I need to add that not all pelycosaurs had these sails.) Something similar's been proposed for pteranosaur crests, although here there are other hypotheses, including acting as rudders in flight, or in courtship displays - in which case there could have been sexual selection operating. Certainly sexual selection can generate some quite extreme traits - the peacock's tail is just one example. Tomkins et al. comment that "[p]ositive interspecific allometry occurs in the sexually selected traits of a range of [living species]", & hypothesise that significant allometry in these two extinct taxa may also be explained by sexual selection.
For their examination of allometry in pterosaurs they used 9 skulls from Pteranodon longiceps - none of these skulls were associated with other skeletal bits & pieces from which to obtain a measure of body size. So the assumption here must be that the bigger heads/crests came from bigger-bodied individuals. Some support for this assumption came from an examination of the size of the eye socket: as you might expect this increased in size as the skulls got bigger - but in living reptiles the diameter of the eye socket is directly correlated with an increase in body size as well. However, as herr doktor pointed out in my earlier post's comments, the individual pterosaurs are described as 'putative' males (N = 6) & females (N = 3) i.e. they've been tentatively classified as male & female on an unspecified basis. If the classification was based on size, then this is something of an a priori assumption that could colour the results. In addition, the smallest 'putatively male' skull is the same size as the 3 'putatively female' skulls, with the same sized crest. It's hard to see strong evidence of sexual selection in these data. I wonder if an alternate possibility could be an age series? Reptiles do increase in size as they age, within species-specific limits, so smaller skulls & crests could simply be those of younger indivdiuals.
In their analysis of Dimetrodon, the team didn't have access to enough individuals of one species to look at intraspecific allometry. Instead, they used data from 7 Dimetrodon species, & found that the sail size did increase with an increase in body size from one species to another. But sexual selection - again, I'd like to see an explanation of just how the authors determined the sex of the individual animals concerned, & how they ruled out the possibility that they were looking at age-related size differences within particular species. Having said that, the authors do note that the sails of smaller dimetrodonts would not have had much positive effect on thermoregulation & might actually have been a thermoregulatory liability, radiating heat so fast in cooler conditions that the animals would have chilled very rapidly.
Basically, the sexual selection hypothesis as an explanation for the adornments of Dimetrodon & Pteranodon is an interesting one, but we need to see data from many more specimens, & a clear method of sexing the remains, to test it further.
Tomkins JL, Lebas NR, Witton MP, Martill DM, & Humphries S (2010). Positive Allometry and the Prehistory of Sexual Selection. The American naturalist PMID: 20565262 doi:10.1086/653001
And I see that Brian Switek has beaten me to it...