Testing for sexual selection

I had a new paper out a few weeks ago but it was at the very height of my busy start to teaching and so barely even got a tweet out about it and completely failed to do anything on here. That’s a shame as this is a paper that has some serious and major implications for trying to detect sexually selected structures in extinct animals (and indeed looking at some odd structures in living ones too). I’ve written a huge amount about dinosaur dimorphism and sexual selection and with numerous papers covering different aspects of the evolution and behviour of dinosaurs (and pterosaus) when it comes to signals and sexually selected things like crests, spines and horns.

The short version is that these are of course hard to look at becuase we can’t directly observe behaviour in extinct animals and coupled with small sample sizes, taxonomic uncertainty of specimens and then issues like extended growth periods and cryptic dimorphism and this is a frustratingly tricky subject to tackle. One standard, if imperfect, measure has been to look at the growth trajectory of the anatomical feature in question and to see if it grows more rapidly than the rest of the naimal, especially iof this happens relatively late in ontogeny. In short, animals don’t need sexaul display structures when they are not sexually mature but when they are this is important so things like horns tend to be small for a long time and then grow very quickly.

This paper led by Devin O’Brien and featuring a host of sexaul selection theroists and biologists posits that things may be more complex still. Features that directly rate to body size will be postively allometric (this can include things like horns and crests in dinosaurs) but those that are not (like say a moths’ antenna), will not. The former are accurate representations of the animals they are attached to and so act as a proxy for their size and quality, but other traits that can still be variable and under sexual selection are not acting in this way and so wouldn’t follow this pattern. There may even be some allometry in these latter traits (non-reproducing animals will not likely invest in such features until the can mate) but the allometry will be much greater, and the correlation with body size present in visual signals.

To help resolve this, we also reccommend in the paper that allometry be tested not jsut again body size but also some other reference trait that is likely to (or been shown to) grow close to isometry. So for example, don’t just measure your dinosaur horn as it related to overall skull size, but also compre it to something like tooth size or humerus lenght. That will help keep things clear when there are other traits around that can grow rapidly or are large but that don’t function as signals. One wonderful example of this we inlcude is a comparisons of the horns on the head of a chameleon with the lenght of the tongue. We used foot size as a reference trait andf show that while both tongues and horns do show allometry, the tongue is little more than isometric but the horns (used in combat and an obvious visual siganl to reflect that) have a much greater allometric slope and show greater variability which is likely to reflect differing quality.

We include a whole raft of such measures of various animals from insects up to mammals and covering both signal and non-signal traits. Two extinct animals were included based on dataset I’ve been working on for a while and may be of interest. One was the frills of Protoceratops which I and colleagues did some time ago but now updated with some extra specimens that we did now have before. These produced a simialr result to our analysis which is no big surprise but nice to see the previous results verified. The second one though was to look at the growth of the tail vane in Rhamphorhynchus.

The standard interpretation of basal pterosaur tail vanes has been that these functioned in steering in flight and acted as something of rudder. That works out quite well since many of the shapes adopted are surprisingly close to the rudders actually made for various aircraft and putting a small vane at the end of the tail would make mechanical sense to increase the effects. However, it is notable that the vanes for Rhamphorhynchus (the only pterosaur where we have a decent sample size) seem to change quite dramatically in shape as they grow and this is rather at odds with the idea that this is purely mechanical. Similarly, there is some serious variation between various basal pterosaurs in vane shape which suggests that the tail is unlikely o be (purely) mechanical in function and the fact that the pterodactloids gfot rid of theirs implies it is hardly critical for flight. Some people have suggested that these vanes were therefor acting as some form of signal and our analysis bears this out. The height of the vane grown very considerably and shows strong positive allometry as the vane changes from a narrow leaf shape in juveniles to a triangle in adults. The vane could of course be multi-functional and it could well be that it has been co-opted from something initially mechanical to function in signaling.

The fundamentals of the methods and theory described here have been around for some time, but the nuance is important to try and distinctuish between traits that are sexually selected and those which are also likely used in some form of display and even combat. It should make for a more reliable way of assessing these kinds of traits and that should be of real benefit to palaeontologists who have an interest in these things. I hope it is not long until more animals are formally assessed for their growth trajectories and what that might mean for understanding their behaviour.

The paper is open access and is freely avaiable here:

O’Brien, D.M., Allen, C.E., Van Kleeck, M.J., Hone, D.W.E., Knell, R.J., Knapp, A., Christiansen, S., & Emlen, D.J. 2018. On the evolution of extreme structures: static scaling and the function of sexually selected signals. Animal Behaviour.


@Dave_Hone on Twitter

Enter your email address to follow this blog and receive notifications of new posts by email.

Join 501 other followers


%d bloggers like this: