Behavioral Ecology Blog

The maintenance of genetic variation in natural populations has long been troubling to scientists- particularly population geneticists. Generally speaking (and using one of the simplest models), variation in coding DNA is thought to be maintained as a result of a balance between natural selection and mutation. Here natural selection plays the role of reducing variation. It goes like this:

Imagine a haploid system at t=0, say 100 individuals, and 20 alleles at locus W.. Fitness values are perfectly correlated to their “name”. So allele number 20 is 20 times more fit than allele number 1. Now it doesn’t take Einstein to figure out that after several generations, alleles close to number 20 will be more numerous than alleles near number 1. After N generations, locus W will be fixed for allele 20. Completely logical, right?

Well the problem with all this is simple- there are many systems (some just as simple as this one) where we find a lot of allelic diversity- meaning that instead of being fixed for allele # 20, many alleles are detectable.. The question is WHY?

The most obvious answer to that question is mutation- but with very little consideration, you can rule this out as the primary explanation. Turns out, mutation is pretty rare, and mutations that produce identical alleles repeatedly are even more rare. Besides, the amount of genetic variation predicted under the mutation-selection balance is generally less that what is actually observed.

Thankfully, this problem has been recognized for several decades, and theoretical work has gone along way in helping us understand the issue. Among the most probably solutions to the problem include negative frequency dependence, balancing selection, plieotropy, and things like population structure. Unfortunately, most models work well under a restricted set of parameters, but are less generalizable than what people hope for.

Taking all this into consideration as a historical framework- in comes the UK Behavioral Ecologists, Foerster, Ben Sheldon, Tim Clutton-Brock and others… They, in the current edition of Nature, propost a novel mechanism for the maintenance of genetic variation in a paper titles “Sexually antagonistic genetic variation for fitness in red deer”

Abstract: Evolutionary theory predicts the depletion of genetic variation in natural populations as a result of the effects of selection, but genetic variation is nevertheless abundant for many traits that are under directional or stabilizing selection¹. Evolutionary geneticists commonly try to explain this paradox with mechanisms that lead to a balance between mutation and selection². However, theoretical predictions of equilibrium genetic variance under mutation–selection balance are usually lower than the observed values, and the reason for this is unknown³. The potential role of sexually antagonistic selection in maintaining genetic variation has received little attention in this debate, surprisingly given its potential ubiquity in dioecious organisms. At fitness-related loci, a given genotype may be selected in opposite directions in the two sexes. Such sexually antagonistic selection will reduce the otherwise-expected positive genetic correlation between male and female fitness⁴. Both theory^{5, 6, 7} and experimental data^{8, 9, 10, 11, 12} suggest that males and females of the same species may have divergent genetic optima, but supporting data from wild populations are still scarce^{13, 14, 15}. Here we present evidence for sexually antagonistic fitness variation in a natural population, using data from a long-term study of red deer (Cervus elaphus). We show that male red deer with relatively high fitness fathered, on average, daughters with relatively low fitness. This was due to a negative genetic correlation between estimates of fitness in males and females. In particular, we show that selection favours males that carry low breeding values for female fitness. Our results demonstrate that sexually antagonistic selection can lead to a trade-off between the optimal genotypes for males and females; this mechanism will have profound effects on the operation of selection and the maintenance of genetic variation in natural populations.

Now I will tell you that they use a weird measure of fitness- de-lifing… I have not looked at the reference (Coulson 2006, Proc. R. Soc. London), but until them (and maybe even after), I’m a bit hesitant to consider it equivalent to more traditional measures of fitness… Anyway, De-lifing is:

This measure, p_t(i), estimates an individual’s annual contribution to changes in population size through both reproduction and survival, and it approximates the expected future representation of an individual’s alleles in the population gene pool¹⁷. The de-lifing method measures performance at each potential reproductive event, rather than on the basis of a per-generation time scale. This allows the incorporation of additional information about annual environmental variation, as well as data from incomplete life histories.

I won’t go into detailed analysis of the paper, but suffice it to say that it is tight- aside from that measure of fitness, which I may be questioning because of my own egnorance and nothing else.. The findings- that males with high fitness sire daughters with low fitness is really fascinating.. Not so much the proximate how questions (which coincidentally are covered in the paper itself), but instead the why questions.

Am I incorrect or is one of the papers major implications is to suggest that “female choice” is not under pure directional selection (forgetting about other “costs” of choice here, like energetics, increased predation, missed opportunities, or other “indirect” costs of choice) like many people assume… i.e. that in a given system “most choosy” equals optimal level of choice? It’s nice to be able to say (at least in Red Deer) that in addition to those indirect costs, there are direct costs (decreased reproductive success, at least for females) that are associated with choice.

I’m now getting tired of thinking about this (at least for now), but there thinking about the connections between this paper, and the Trivers-Willard Hypothesis are especially interesting.

Any other thoughts on this paper??