As of last night my latest paper has come out, coauthored with Darren Naish and Innes Cuthill. Those with access to the journal Lethaia can get it here. Believe it or not I’ve been juggling with the idea as to whether or not to blog about this for quite some time. This is, I think, the most significant paper that I’ve produced and it’s the product of literally years of work (though at least part of that was as a result of very difficult editors and referees at various times, this was started back in 2007!) and I’m really rather proud of it.
Then why not blog it? Well the short answer is that this is a long and complex paper and it ultimately deals with a huge range of difficult issues (and not in the length we’d have liked, it had to be cut down severely to fit the journal and we still incurred page charges). It touches at various times on pterosaurs, sauropod body size, various ornithischian lineages, theropod sociality, the origins of feathers among other themes. All of this means that it’s very hard to blog about and cover the salient points for a non-expert audience without writing thousands upon thousands of words and, well, I did that for the paper.
This is obviously counterintuitive for a blog that is effectively about science communication, but I can’t do everything all the time (I certainly haven’t blogged all of my papers of the last few years). Moreover, in my experience, a paper like this which rather stomps a bit over some much cherished hypotheses of people tends to attract huge number of comments along the lines of “but what about *this* contrived example!” which I can assure you gets very annoying when people won’t let it drop.
None of this means I *won’t* be blogging it at length. But I know it’s likely to be covered a bit elsewhere on the web and thus it’ll look odd that I’m not doing it right away and it seemed sensible to provide an explanation up front. What I will at least talk about it mutual sexual selection – it’s right there in the title and the abstract and is, I suspect, a concept unfamiliar to most, perhaps nearly all, readers. it is after all, something almost entirely absent from the literature on dinosaurs and pterosaurs, Darren and I could only find two other references to it ever and one of those was what we put into the Taylor et al. paper on sauropod necks and the other sprang from Portsmouth. So it’s something that’s only really just coming into the literature.
Sexual selection is probably familiar – the idea that some traits are selected for by the opposite sex and can drive the development of bight colours, crests. displays and all manner of other things. The obvious one that’s endlessly used is the train of a peacock, that makes the male look very different to the drab female. This is typically coupled with sexual dimorphism (again, like the peacock) where the male is bigger than the female and has the extra ornaments etc. and males compete for females, with the best ornaments males advertising their fitness through the size and quality of their fitness (though in some cases like jacanas, this is reversed with bigger females).
So far, so simple. Mutual sexual selection is simply an extension of this into both genders. Both males *and* females are ornamented (or rather, have sexually selected traits) and just as males are competing with other males for the best females, so too the females are competing with each other for the best males. This means that dimorphism is limited or even absent – both genders having such traits. This is in fact, well known for quite a number of bird species and the number of papers on the subject in living species is increasing in leaps and bounds.
In the paper we hypothesis that this may have been common in the ornithodirans. It explains (potentially) quite a lot and solves a couple of previous paradoxes about crests evolution and development. Critically it means that you *don’t need* dimorphism of a feature for it to be sexually selected – both genders can have a crest and it can still be a sexually selected feature. This needs testing, this paper does little more than lay out all the conceptual issues and evolutionary biology and ecology behind the hypothesis, but at the same time, I think we do have some pretty good support for our ideas.
But as ever, what really needs to happen is for you to go and read the paper! And yes, I do have a PDF if you want it.
HONE, D. W., NAISH, D. and CUTHILL, I. C. (2011), Does mutual sexual selection explain the evolution of head crests in pterosaurs and dinosaurs?. Lethaia. doi: 10.1111/j.1502-3931.2011.00300.x
Dave Hone's Archosaur Musings 
Why didn’t you submit to PLoS ONE and incur no page charges or lose any length? Surely impact factor wasn’t that much of a bugger.
We did, or rather Andy Farke took it to their chief editor, but they declared it a review and not ‘original research’ and wouldn’t accept it.
At least they didn’t declare it to “not of interest to their readers”. 🙂
Could I bother you for a PDF please? (Presuming you still have my e-mail address.) I’m out of the academic loop now.
I’m reckoning that Nick speaks from experience 🙂
Hey Dave, I have a slightly off-topic question.
Can a random nobody like me TRY to get a paper published? I understand the process is long and often annoying, but I just want to know if I can try it out?
I have a thought on MOR 980’s metacarpal III. I wouldn’t even attempt to try it out if I can’t get access to the data I need. It’s just an idea I’ve been toying around with.
Yes you can and people do. Anyone can publish, if it’s up to scratch.
Wow, really? Thank you! I’ll give it a try.
Absolutely! Read this post (written by *cough* me) and get to work!
http://svpow.wordpress.com/2010/11/12/tutorial-10-how-to-become-a-palaeontologist/
(Some of the other tutorials on SV-POW! will be relevant, too, even if for some lame reason you’re not writing about sauropods.)
Ah yes, meant to flag up that one, sorry Mike.
Congratulations, I look forward to reading the paper.
I haven’t read the paper (though I’d like to) but does the mutual sexual selection assume that the same traits are selected for by both sexes? One could certainly argue that in humans, females look for indicators of strength (broad shoulders, muscular development) and males look for indicators of fertility (such as large breasts and high hip-to-waist ratio.) In this case, mutual selection would increase sexual dimorphism. For that matter, male peacocks could be selecting well-camouflaged mates.
I think as traditionally thought, and certainly every example I know, then we are talking about the same basic structure in each gender. But that doesn’t mean (I don’t think) that it *has* to be, but it makes sense that most of the time it would be since it’s unlikely that multiple ornaments would originate independently in different genders.
That said, the examples you are giving are not really about sexual selection. The conceptual idea of such charactersitcs is that they are an actual penalty to existence (making you more vulnerable to predation, requiring heavy investment etc.). Even if peacocks were selecting peahens for better camoflage, it’s hard to argue that as being sexually selected since the females benefit from it (or males would benefit from having greater muscle mass).
Sexual selection can definitely align with natural selection so that the same traits are advantageous to both. Such traits are traditionally called viability indicators and yield indirect “good-genes” genetic benefits. For a good overview see http://rspb.royalsocietypublishing.org/content/269/1498/1331.short
I would love the PDF if you don’t mind. Can you see my e-mail through the comments system?
Hello, yes, comment on it’s way. And thanks for the clarification, I agree these things can be aligned with natual selection, I should have been more clear, the kinds of things we’re talking about in our paper are about these odd crests and so on that don’t seem to have any obvious function and certainly at a bare minimum *appear* ornamental and are in the mould of ‘classic’ sexually selected traits.
Fantastic, thanks.
You’re not on twitter, are you? I desperately need to follow more palaeo people, especially ones that talk about sexual selection.
Funny, you’re the first person to ask. No I’m not, though it appears someone has been impersonating me on Twitter. it’s not me.
This is exciting! I want to read it. You’re a good writer and the material is fascinating. Wow.
It makes sense intuitively. Yet this blog entry is the first mention of mutual sexual selection that I’ve seen in my life. Now I’m curious about how you tell that a feature may be sexually selected without dimorphism. Can’t wait to see what you’ve presented.
Well obviously it’s a real pain doing it in the fossil record. We do think there are some correlates that suggest it, and that in some cases sexual selection / social dominance (an interlinked and not necessarily separate point) are the only viable hypotheses, with a lack of dimorphism, mutual sexual selection is the obvious conclusion. Though again, I’d emphasise that the real issue is not so much if we’re right (though i think we might well be! 😉 ) but that this simply hasn’t been tested or indeed in some cases the lack of dimorphism has been used as an argument against sexual selection. That is a massive issue for interpreting these crests and ornaments since we feel it’s a false assumption and incorrectly ruling it out before you even start can cloud or colour anything else that starts from this premise.
Wait a minute – the ‘MaybeDaveHone’ on twitter is NOT Dave Hone?? How bizarre. For those who care, I tweet @TetZoo.
No, it’s not me. And it’s all a bit weird. My name, my photo, links to here, chatting to friends and colleagues about dinosaurs and pterosaurs. But not me.
Not to blow my own trumpet, but I pretty much twigged straight away. Much as Dave’s got a decent sense of humour, the Twitter impostor claimed to be from ‘Pterodactyle Town’ 😉 And some of the stuff he posted…oh dear…
A cool paper, as far as I can make out from the blog posts (no access from home). Let me throw the cassowaries-use-crests-as-vegetation-splitters thing into the ring. When they run, cassowaries lower their heads and the crest takes the impact of branches – something you can see even in zoo animals by the scratched and often even dented appearance of the crests. Clearly, that may be the reason why they are so unusual in having crests AND feathers.
Hmmm, that comment looks familiar…. 🙂 In the paper we note that there’s a bunch of things that cassowary casques are used for, not just a deflector. So again, multifunctionality rears it’s head.
Yep, just saw (thanks for emailing the paper) – but even one single main function outside sex alone would suffice to “explain your problem away” 😉
You mean for oviraptorosaurs?
for any taxon that doesn#t behave as predicted by mutual secual selection – I took from the LitToCh post that you think cassowaries odd because they have crests DESPITE having feathers, which would do just fine for display. If crests in them have one other important function (thus, are not a display primarily), then that takes them out of the equation, at least partly.
So if we have a fossil taxon that sticks out as a sore (feathers+crests), cassowaries show how that may not falsify mutual sexual selection – it just don’t apply in this case, y’know?
as far as my ill-informed and drunk opinion goes, that is! 🙂
Right, I getcha. Sorry, hadn’t quite followed that line of argument, but yes, basically that’s what we’re saying. it doesn’t falsify it, they just might be doing something else / extra.
f-ing pre-adaptation and conversion at wok 😉
Hello, would be too much trouble if you email me the paper? Thank you!
So I’m curious as to what you make of the ‘social signaling’ hypothesis proposed by Padian, Horner, and Goodwin (and mentioned in the Padian and Horner papers). This is not the same as ‘species recognition’, and I do not recall seeing it discussed in your review.
In ceratopsians, we see ornamental structures changing significantly through ontogeny, often becoming reduced or resorbed (nose & brow horns, frill epiossifications). You see similar things in pachycephalosaurs (even if you ignore the pachy-stygi debate) and tyrannosaurs. If the structures display social status, then this may explain their significant (non-unidirectional) ontogenetic changes. Would mutual sexually selected traits only get bigger ontogenetically?
Does this not rely on the general hypothesis of massive ontogenetic change being correct? By that I mean if Trorosaurs really is Torosaurs and not a juvenile Triceratops as advocated, then this problem doesn’t exist (the idea that something might shrink towards adulthood) and ornaments are then apparently stable or only grow during ontogeny? So as a hypothesis that aspect of social signaling is dependent on the other hypothesis being correct and I’d argue that this is still pretty contentious.
On a not unrelated note, while it didn’t get the space we wanted to devote to it, certainly yes social dominance signals can be important and could promote the kinds of ornaments we are discussing here, though actually quite of then the two are linked (as in examples we give in the paper like black swans – those with the curliest feathers are dominant within sexes [social dominance and by extension a signal of dominance] but these are also selected for by the other gender [sexual selection , and here mutual sexual selection]).
Regardless of that, I have actually got a number of ideas and thoughts about all this which I’m in the process of writing up. But quite simply there wasn’t time or space to get it into this paper and I don’t want to just shove it all out here on my blog when I’m hoping it’ll be published at some point. So I’m afraid I’m basically going to duck that one right now.
It doesn’t matter if you agree with the evidence that Triceratops and Torosaurus are synonymous. I don’t think that you can reasonably argue that all the different growth stages that we have of Triceratops are all in fact different species (from the same beds of the same formation), especially as we have intermediate specimens and specimens that show combinations of intermediate character states.
Even if you disregard Trike-Toro synonomy, the changes through the solid-frilled morphs are more significant than from triceratops to torosaurus. Epiossifications are pointed and more prominent in juveniles and become broad and blunt in adults; brow horns are relatively at their longest at subadult stage (again, ignoring trike-toro).
Uncontroversially, everyone agrees that various centrosaurines resorb brow horns and nasal horns through ontogeny (especially well illustrated for Centrosaurus / Pachyrhinosaur bonebeds), as does Chasmosaurus (may be more controversial, although I think not). In ceratopsians, ornamentation structures are often more prominent in juveniles than the larger (presumably adult) morphs. Resorption is actually the norm rather than the exception for many structures.
Similarly in pachycephalosaurs, small specimens (presumably juvies) have more prominent spiky cranial nodes, and larger forms from the same formations (presuambly more adult) have less spiked nodes: all the way round the skull. Again, whether or not you agree with Stygi-Dracorex-Pachy synonomy doesn’t matter here.
I would agree that some dinosaurs seem to show more linear growth trajectories when it comes to cranial ornamentation (most hadrosaurids for example). Maybe this means that among different clades, social signaling vs mutual sexual selection are more/less dominant.
Well I personally didn’t know about the Centrosaurus example, so that’s another issue. And with regards to your last point, that’s certain possible and yes, we are of course not arguing that mutual sexual selection was universal among the ornithodirans. In creating a review paper like this we can’t deal with every example, or even every major clade (there’s a lot more that can be said about hadrosaurs for example, and the rhamphorhynchoids don’t get much of a look in) and we do point to some structures (like Triceratops horns or pachy. domes) that were very probably nothing to down with sexual selection (or it was not likely the driving force). But I’d maintain that the central point (Mutual SS has yet to be considered at all, and is at the bare minimum plausible, if not even probable) causes a problem for at least some of the hypotheses or assumptions for hypotheses out there (e.g that dimorphism is a prerequisite for sexual selection). While clearly we are advocating MSS as a putative driving force, regardless of whether it turns out to be correct or not, it does, i think, need serious consideration. This paper is the starting point of that, and individual cases (be it ceratopsians as a whole, or just a single species) need to be evaluated on their own merits.
Quick response to Denver… If cranial ornaments in Mesozoic dinosaurs have a role in sexual display (of any sort), we should expect them to become elaborate at sexual maturity – and this is what we do see (remember that sexual maturity and skeletal maturity are not the same thing, hell no). I’m not sure we can be confident that old adults become ‘less’ elaborate with age (which pachycephalosaurs do you have in mind? All the claims regarding ontogenetic morphing are, shall we say, open to debate), but if we could, this is ok, since old individuals in populations are not necessarily the most highly ornamented, there is variation within populations, and you don’t know the ontogenetic history of all individuals.
Centrosaurine brow horns change from horn nubbins to craters, but the craters are the derived condition, and how do you know they’re not a peculiar form of ornamentation? We have a section in the paper on this: repeated LOSS of sexual display ornaments is common in lineages that evolve those structures.
I’ll be covering this stuff on Tet Zoo at some stage – glad the paper has prompted so much interest here on the Honesque Musings 🙂
>If cranial ornaments in Mesozoic dinosaurs have a role in sexual display (of any sort), we should expect them to become elaborate at sexual maturity – and this is what we do see (remember that sexual maturity and skeletal maturity are not the same thing, hell no)
This would be something to test directly by conducting histology on limb bones associated with skulls and seeing if the kink in the growth curve (which people associate with sexual maturity) is associated with the onset of display features.
An immediate problem with this is that cranial display features seem to occur at very small body sizes (ie. before sexual maturity), and change continually through ontogeny. I don’t see any point in Triceratops ontogeny where there is a “fixed” morphology. They keep changing till they die. All the Torosaurus synonomy does is give us a final morphology that is fenestrated. If you don’t agree you can just stop at the final solid-frill stage.
With hadrosaurs, there does seem to be similarity among juveniles, then the crest starts to develop (still at a pretty small size). This would be more consistent with mutual sexual selection.
>Centrosaurine brow horns change from horn nubbins to craters, but the craters are the derived condition…
The point is that juvenile centrosaurines develpo brow horns, which are then lost. The ornamentation increases to a point, then decreases. You’re not just seeing increased elaboration: ornamentation growth is not always linear.
Padian and Horner (2010b, p. 23) note that sexual selection, by definition (Darwin), requires dimorphism. ‘Mutual Sexual Selection’ is therefore a different process if it does not have dimorphism.
Thanks for thoughts, Denver. I’m not sure I see your point though: why are you assuming that ornamentation growth should be linear? It isn’t in many modern animals that employ cranial (and other) ornaments in sociosexual display. Appearance of ornamentation “at very small body size”: are you referring to the juvenile Triceratops? If so, I don’t see that the ornamentation they have is inconsistent with sociosexual display – ornamentation has to start growing some time, and what’s seen in those specimens is incipient of the sexually mature condition. cf, short ossicones in juvenile giraffes, tusks in juvenile elephants etc. Clearly, the species recognition hypothesis is not inconsistent with sexual selection in such cases.
Yes, Padian and Horner insist that sexual selection demands dimorphism, but this is debatable. Have you read the Hone et al. paper? We address this: sexual dimorphism is often an expected product of sexual selection, but it isn’t inevitable. There are species where mate choice is inspired by the same ornaments, present in both sexes. Herein the phenomenon of mutual sexual selection.
Many years agoI heard a spectacularly good lecture n speciation in land snails. Apparently sister-species (cryptic to the casual human observer) are often distinguishable if you look at bits of anatomy relevant to a snail looking for a mate, in particular the “darts” that are apparently used in snail fore-play. Evolution of features of this sort, I would think, is almost inevitably subject to sexual selection… but snails are hermaphrodites So would this be a good example to show the mutual sexual selection is a real phenomenon and that sexual selection doesn’t HAVE to involve sexual dimorphism?
I read your paper, but I haven’t read any of the papers that you reference that discuss extant examples. Mainly I am interested in seeing how applicable mutual sexual selection may be to ceratopsians, and whether or not it is any different from “social signaling”. We are studying long-term changes in ceratopsian ornamentation, so I am curious as to whether we can test for mutual sexual selection as a driving mechanism.
I’m not saying ornamentation growth is (always) linear; I’m giving examples of where it isn’t (also examples where a structure increases in size then resorbs). Some recent papers have suggested that ornamentation only ever increases in size, not me.
Darwin’s definition of “sexual selection” produces dimorphism (this is what Padian and Horner were arguing). Strictly speaking, “mutual sexual selection” is not Darwin’s “sexual selection”; it is something else. So again, I am trying to figure out how different it is from “social signaling”, which is basically how ornamentation structures confer social status.
One last response before I get offline for Christmas 🙂 Re: distinguishing between ‘social signalling’ (aka status signalling) and MSS – they are different phenonema but, yeah, it may be difficult or even impossible to say that one was at play and not the other, especially when both are expected to be in operation at the same time. They are in the respective extant birds. In fact some research (on mutually ornamented sparrows) suggests that MSS evolves in group-living species where there is unusually strong competition for resources and hence a pressure for ‘status signals’ to evolve. Was status signalling present in Mesozoic dinosaurs? This is an interesting hypothesis that I really think plausible, though one complication is that status signalling in extant animals is often not linked to age (whereas the variation we see in the ornaments of Mesozoic dinosaurs seems to be ontogenetically controlled). Anyway… the raison d’etre of our paper was pretty much to bring MSS to the attention of palaeontologists (and we were obviously interested in sexual signalling, hence the brevity of the section on status signalling [p. 10 of the preprint]). It’s an addition to the dialogue, and very clearly not the end of it.
Hi Dave – please send a copy of the pdf when you get a spare moment – sounds very interesting. Best regards, Mark