AUSTRALIAN FRONTIERS OF SCIENCE, 2005

Walter and Eliza Hall Institute of Medical Research, Melbourne, 12-13 April

Session 6: Discussion

Question – I was wondering whether any of the genes that are involved with, for example, male display traits or female choice have been characterised. If so, do you think there is any value in perhaps identifying those alleles and allele frequencies at the level of the DNA in those kinds of studies?

Mark Blows – We are currently QTL mapping both male attractiveness and, eventually, female preference in Drosophila melanogaster. That is the ‘lab rat’, and it has a similar communication system to the one in my Drosophila species. So we are QTL mapping there, and we are eventually going to QTL map in D. serrata as well.

The reason we want to do that is that we want neutral markers to follow these genes in the field and look at selection on particular loci. It really depends on the distributional effects as to whether looking at allelial frequencies at particular loci are going to give us a lot of information about the underlying genetic basis of the genetic variances and covariances. It is very difficult to predict whether we will ever be in a situation to say, ‘There’s the group of loci involved. These are the allelial frequencies. You do the maths, and that equals the genetic variance.’

It is likely that the distribution of allelic effects is going to be log-normal; you will have a couple of major genes and then lots of little ones which you won’t be able to detect. So it is going to be very difficult to take that kind of reductionist genetic basis, which has a lot of value, and meld it with the statistical description of what the effects are.

Rob Brooks – I think with condition-dependence you are probably looking at a moving target, as well.

Mark Blows – It would be very difficult to QTL on that condition, I would have thought.

Question – Rob, I am taken that your major finding is that there is not always a single most attractive type, and you have based that on a fitness shape that has three peaks, as I understand it – iridescence, large tail, orange-black, whatever. You have said that that shape was very unusual, and I agree. Two of those three peaks are right on the boundary, and I have always been taught that, in maximum likelihood estimations and so on, one should really be rather wary of peaks at boundaries. Is that of concern, or not?

Rob Brooks – I think it is of some concern. We can’t be certain that the multiple-peak nature of that fitness surface is what’s maintaining variation, but it seems that the distribution of genes underlying the colour patterns doesn’t allow the optimisation of all of the different things that make males attractive. I wouldn’t bet my life on whether those three peaks happen to be the ones or not.

In terms of the number of individuals supporting those particular peaks, there are a handful of individuals in each case. The peaks that are on the boundaries are really, to some extent, extrapolated. There is an increase in attractiveness, but the precise height of those peaks et cetera is not necessarily something that I would stake anything particular on.

Question – If flies have some interesting chemicals in their wax, and guppies have some interesting colours, and if these are for, say, attracting the females, what happens if they have another functional role as well? For example, the waxy layer might be to prevent the organism from being eaten by predators or being attacked by viruses. Many genes are highly polymorphic and they have, for example, increased catalyst activities, but they also have other functions as well. I was wondering how that then relates to your studies.

Mark Blows – In my own system, I am pretty sure that that set of chemicals has a number of functions. So, in terms of levels of mate recognition, for a start, the ones that are important in keeping a second species away – in not getting confused – and the ones that are under sexual selection within a population, are fairly distinct. There is also a strong possibility that those waxes are involved in things like stress resistance. So desiccation resistance would be a prime candidate.

That kind of selection is certainly going to change those allele frequencies, there is no doubt about that. And it’s going to swish those hyperdimensional genetic spaces in certain ways. How that relates to the association between the direction of sexual selection and the pattern of genetic covariance that we see, is a very good question. I don’t have any data at the moment that replicates that kind of result under different environments – for instance, under different desiccational environments – which would be the prime way, I guess, that I would approach that kind of question.

Question (continued) – You have had this Drosophila changing and some that don’t, and so at the end of the day which is going to win out – the survival of the species because of the desiccation, or being able to attract the right fly (or person or something)?

Mark Blows – I think there is always going to be the notion of a trade-off between sexual selection and natural selection in that kind of context. I guess what you are looking at in these populations is what that situation currently is. They have already struck that balance. If there is any allelic variation for increasing attractiveness, given the current natural selection context, what you should see is that kind of pattern.

Question – I was just wondering about the role of female choice. Have you ever considered constructing a population where you deliberately wait for the female to make a choice, then make sure that she only mates with the non-chosen male?

Rob Brooks –  Certainly that experiment has been done a few times. We have tried it, interestingly, with the guppies, allowing females to mate with their chosen males or mating them randomly. We didn’t get much of a response. It is conceivable that we didn’t do it for long enough, or we didn’t select hard enough, but it is also quite possible that the reason is that there’s not a hell of a lot of genetic variation in that direction.

But longer term experiments, and probably bigger-scale experiments than are possible in the guppies, have been done in various fly systems, with varying effects. I think in some cases that has had enormous ecological effects. When you allow female choice, the population tends to adapt more quickly to various environmental factors in the lab and tends to grow more quickly, et cetera, which suggests that female choice is a very benign in evolution. Other experiments have shown that males and females, if you circumvent female choice in the interaction and just mate them randomly, tend to do quite nicely – but if you allow them to do their business, males and females are at war with each other and this has all sorts of nasty effects ecologically.

So I think the answer is that it is going to be different, just on the starting conditions in the system that you are working with. It certainly impacts on a lot of other aspects of adaptation and ecology.

Question – I have a question about choosing the traits. How serious is it that you are missing something that you don’t see as being a trait, when it is one? That must come up. How is that going to skew things? How do you think about that in your research?

Mark Blows – I can answer that empirically, in that you can actually estimate the amount of genetic variance in the set of traits that you have chosen – such as the cuticular hydrocarbons – that explains the genetic variance in the holistic trait, which is mating success. When we do that with the hydrocarbons we can explain anywhere up to 80 or 90 per cent of the genetic variance in that life history trait, mating success, by those mechanisms.

At the same time, during the QTL mapping procedures we are finding that, although single QTLs don’t map to the same locus they control both traits. So there are not single pleiotrophic loci doing that; it is the combination of QTL for the CHCs that explains mating success QTL. So at both the QTL level and the quantitative genetic level we are starting to be able to quantify whether we have missed important aspects of attractiveness. And I am sure that we have. There are certainly other aspects, such as body size, which would play a role.

Chair – I guess in terms of trying to predict the dynamics in a natural population you have also got the issue of the presence of other species that is also relevant in respect of [inaudible]. I know you have also done work on that, but that’s an example of where you have something in the system that, I guess, in a similar way you have conserved to a great extent. I know you have done experiments on it.

Mark Blows – We are trying to look at the effect of the presence of the second species. We know that reinforces the way that the animals make decisions within their own population. It changes the way that they do that; it displaces their communication system away from the second species.

What we haven’t done yet is to determine whether that displacement disrupts the way that they normally go about their business within their own population. I think it is a critical question whether that level of mate recognition – getting the species right – interacts with the level of mate recognition in terms of who is most attractive within that species. We haven’t got an answer to that question yet.

Question – Rob, in relation to the calling song of the cricket, have you looked at the frequency structure of the call song, the actual spectral components located within the sound, and whether that changes – whether it is not just duration but some content of the call may change?

Rob Brooks – There is. We have measured that; it is quite nice and easy to measure various spectral properties. What we tend to see is consistently that the most attractive type is an intermediate and is pretty much where the population mean is. So it seems that what is happening is that the auditory system of the females is tuned to a particular frequency and the neuropsychological mechanism for processing the information in the song is tuned to particular aspects of the timing of the song, in such a way that there is a single most attractive way, or a single most effective way, of stimulating a female and males have very successfully managed to converge on that. That is the fitness surface with the nice single smooth peak.

There are some aspects of timing within the song, such as the end of one call and the start of the next one, that are somewhere between stimulating females with the appropriate timing and how much effort you are putting in. So that particular aspect is related, I guess, both to call structure, which is under that strong stabilising selection, and to call effort, how much effort you are making, which is very directional. You want males that put more in, essentially.

Question – In these studies, is it important to consider the degree of female genetic variation and how it influences the choice that females make – what traits they choose? It is something I have never seen discussed.

Rob Brooks – There is a tremendous amount of genetic variation in some species. In the guppies we have variation in how willing females are to choose. We have got a lot of individual variation in what they like, but that doesn’t seem to be genetic; that seems to have a lot more to do with the female’s past experience. So there are differences in what different females like. There are some that consistently go for males that are, on average, unattractive. But most of the genetic variation is in just how motivated they are to choose at all: some females don’t choose but just mate randomly, and some tend to put in quite a lot of scrutinising and choosing among the options.