Male (above) and female (below) gharial skulls. Photo courtesy of Larry Witmer.
So, yes, new paper time and which the concept behind this one was quite simple the outcome (as is so often the case) rather spiralled out into a bunch of other, very interesting aspects. As I noted in the run up to this post, I’ve been working a lot on sexual selection and what it means for dinosaurs in particular and wanted to use gharials as the perfect model for dinosaurs but lacked a dataset on these rare animals. A chance post by Larry Witmer led me to contact him about his dataset but it turned out to be only three animals, not the dozens I’d hoped for.
It was though, enough stimulus to get me hunting and with Jordan Mallon roped in with his interest in testing these ideas we just needed to get enough data. Happily, my former undergrad student Patrick Hennessey wanted to get engaged in some research and had time on his hands, so while I e-mailed every museum curator and croc research I could asking for photos of skulls, he set off to visit every collection in the south of the UK that was accessible. Some months later and we had an incredible set of over 100 specimens. We know of more too from photos that lacked scalebars (we were unuseable) or were in museums where we couldn’t get a response from the curator, or had various bits of skin preserved which concealed key bits of data. (We also found a good few mislabelled specimens of Tomistoma while we were at it). Still, 100 is a massive dataset for this kind of work and especially for such a rare animal and this gave us an excellent platform for our analyses.
Digging into the gharial literature though we soon found other issues. Despite the fame of these animals, their rarity means the literature on them is very small and very little is known in detail or was last written about in detail decades ago. To complicate things further, the two distinctive male traits (a fossa on the snout that correlates with the ghara, and a pair of palatal bullae) have never been truly convincingly shown to be definitively male accoutrements. Happily, an analysis of the data did suggest that the fossa was clearly a male feature and the bullae most likely were too.
Moving onto the central point of the project, analysis of the dataset showed that without pre-existing evidence for a given specimen being male or female, discovering any evidence of dimorphism was very hard, even for a dataset of over 100 animals. Gharials are strongly dimorphic in body size but the overlap between larger females and smaller males across much of the data, and the unknown sex of juveniles (which shown neither fossae nor bullae) makes finding this signal impossible. This matches what Jordan and I have said in a previous paper, and suggests that short of very large datasets and / or very strong dimorphsm (even more than seen here) or very good evidence for the sex of most specimens, it will be hard to find. That means that for the average data set we have for even well-represented species of dinosaurs (well under 100 incomplete specimens, no idea of levels of dimorphism but unlikely to be well above what we see in modern species, and no data on sex) we are not going to get a signal on dimorphism even if it’s there. I’m sure dimorphism is common in dinosaurs but I’m also sure we’re not finding it.
Female (left) and male (right) gharial snouts, the latter showing the expansion of the snout and the narial fossa anterior to the opening that makes the nares. Image courtesy of Larry Witmer.
That is, of course, based on things like body size or where a feature is expressed in both sexes (as, for example, ceratopsian fills appear to be). Presence-absence dimorphism (where one sex has a feature the other does not) should still show up relatively clearly with much smaller sets of data, but we’re not aware of any species that would obviously fit this criterion. The fossil record isn’t giving up numerous horn-less Allosaurus or dome-less Pachycephalosaurus specimens and while there are things like the two Khaan specimens with different tail anatomy, it’s just those two for now rather than a nice dataset of a dozen or so. Well-known taxa like Centrosaurus and Coelophysis are distinctly lacking in obvious dimorphism.
All of this is hopefully interesting and important for understanding sexual selection in the fossil record and as a guide for future research, but this work also threw up some interesting information for the gharials themselves which is worthy of comment. First of all, we were able to show that the fossa on the snout which is the correlate for the ghara is strongly positively allometric. This is no big surprise but it’s good confirmation that this feature is under sexual selection, and conforms with the (limited) evidence that the ghara starts growing around the time that these animals become sexually mature. We also note that it likely serves as an honest signal, since it would generate tremendous drag on the tip of the snout and that’s pretty critical for an animal with a super thin and presumably hydrodynamic set of jaws used to catch fish.
Surprisingly though, the bullae don’t show this pattern. They first appear on skulls around the same time as the fossa suggesting they are also linked to reproduction, but they first appear just before the fossa. We suggest that this is because the ghara while still small, may not need a fossa to hold it onto the skull and so the ghara and bullae may start growing at the same time, but the bullae would appear on the skeleton first. The bullae are also not allometric, so while they are larger in larger males, they are not disproportionately larger. This suggest that while they are an important part of the reproductive biology (and presumably as part of the palatal sinuses, potentially in making noise) it might be there merely to indicate sexual maturity rather than be an actual attractor. Either way, these give us some hints about the reproductive biology of these animals which gives us some hypotheses to test.
One last thing we spotted is that the very largest males are quite disproportionately robust. They have unusually wide skulls (including the normally slender snout) and also have very thick teeth, with animals only 20% smaller having teeth about half as thick. To our knowledge this has not been observed before and quite what this means isn’t certain. We hypothesise that these very large individuals might either have especially strong heads and teeth for fighting each other, or perhaps because they are entering a different niche and are able to exploit much larger prey than others. Either way, this points to an important issue given how endangered gharial populations are.
Very young gharials, yet to display any external features that might indicate their sex.
With animals under strong sexual selection, a few individual males will have a disproportionate amount of the mating opportunities in a population. But those males are also likely very well adapted to the prevailing conditions. They have, essentially, a good combination of genes allowing them to grow so big and maintain such a large ghara. If they are operating in a different niche and that isn’t taken into account (they may be eating much larger fish species compared to other gharial for example) when trying to protect them and conserve their habitats, then they might be especially vulnerable. If your genetically best adapted and fittest individuals are at most risk, that’s potentially very bad news and is unlikely to be good for the long term survival and genetic health of the population. This is of course, potentially rather speculative, but it’s supported by what we understand of strong sexual selection and the observations about the largest male skulls. It’s certainly something that is worth checking out in more detail and at the bare minimum it’s an interesting observation about their ontogeny and what that might mean for our taxonomy in the fossil record.
So here ends a very long process to analyse and assess dimorphism in gharials as a model for dinosaurs. It has thrown up far more complexity and nuance, especially in the living species themselves, than I ever thought but that has been in itself most interesting. It only remains for me to thank my coauthors for their contributions on this paper, and the huge number of curators and researchers who generously checked catalogues and sent in photos for us, the paper really would not exist with them all.
Dave Hone's Archosaur Musings 
Recent Comments