One of the biggest problems in evolutionary biology is the maintenance of genetic variation in populations. Without genetic variation, populations do not change in a consistent way over time. Basically, without genetic differences, all the differences between organisms would be erased at reproduction. If natural selection is consistent over time, applying the same winnowing effect to every generation, we should end up with only the alleles that selection favors, i.e. less variation. This is called “directional selection” since it favors the evolution of traits toward a particular size. This is different from “stabilizing selection” that favors one particular size and selects against extreme phenotypes. However, both of these forms of selection yield less genetic variation over time if selection is constant; if selection goes back and forth favoring one phenotype and then another, then it’s easy to maintain genetic variation. Another simple exception is heterozygote advantage, where a particular genotype is favored, rather than particular alleles: the alleles have to occur in a particular combination that does better than either does alone.
Sexual selection, the process that leads to ridiculous looking, dead-sexy animals, is usually conceived of as directional selection. If we think of a process that produces a long tail or big horns or a sexy mind, that process will favor phenotypes on one end of the range in a population. This presents two problems. After a while
- the genetic variation will all be gone
- females will not get a better mate by being choosy
This is a particularly bad problem if being choosy is costly. Looking around for a mate might be dangerous: predators could be hiding in the bushes and do you really want to mate with the male that has mated with every other female? Females may get food or resources from males, but in many cases all they get is sexy offspring. Females in many models need to get a really substantial direct benefit for sexual selection to work in the first place. Then consider that when there is no difference in genetic sexiness of males, a female that mates with the first male to come along does just as well: her offspring are just as sexy.
There’s another problem: when we look at sexually selected traits in the wild, we actually find a huge amount of genetic variation. Either these traits are not under directional selection at all, or selection is very weak, or there is something maintaining the variation, propping it up every generation. This could be mutation, or it could be particular combinations of alleles that lead to different phenotypes. If there are a huge number of genes controlling the phenotype we’re looking at, there are a lot of ways for them to combine. All of these problems come under the name “lek paradox,” named for the dancing arenas used by males to attract females. The lek paradox is basically “Why do we see choosy females when all females get from males is sexy offspring?”
Thankfully, there’s another interesting twist: sexual traits are affected by the overall health of the whole organism. Developmental variation or the social environment can lead males to display different traits based on their health, their age or their perceived level of competition. Condition-dependence is variation in traits that depends on the overall health of the individual, and therefore genetic variation in condition is revealed in signals that depend on overall health. This type of signaling is very important in the evolution of sexual signals, as it reveals the health of potential mates to females. Also, it can’t be cheated: if you can produce a big, costly signal indicating health, you must be pretty healthy; you’ll probably make vigorous, not just sexy babies. This is one type of “handicap signal.” Jerry Seinfeld explained handicap signaling best in an early episode of Seinfeld: why give someone the finger, when it takes a lot more effort to give someone the toe?
My research deals specifically with the noise introduced by age: if the signal a male displays depends on his age and his health, or his particular genes, then males of different genotypes and ages can have similar display traits. Mutations in overall condition should maintain genetic variation in the signaling trait. Some scientists have looked at a similar idea before: if a male is old, he has proven that he can survive and therefore probably has good genes. Old males should be a attractive. Some have said that it can work, and other have said it’s unlikely. I ask a different question: can those traits actually evolve? There are several problems: males with slow-growing traits, that enable to live a long time will be rare, they may never grow old enough to be sexy. If they display their sexy trait at a young age, they will be killed for it. Another problem with old males becoming attractive is that most mutations come from old, backdated sperm: will the offspring of a sexy, old male be even as sexy as their father?
My goal is to find the conditions under which males will evolve to display age-dependent traits using mathematical models and computer simulations of evolution. Age-dependent traits should help maintain genetic variation in sexual traits because males who survive differently should display differently: males in good health, with lots of good genes should wait until they get old to display their sexiness; males in not-so-good condition should display earlier in their lives, because they won’t get old. The point is that females can’t tell the difference — or we should have to demonstrate that they can, which means the evolution of an age-dependent trait! — and mate choice will thereby maintain the variation in condition.
Thanks for reading and subscribe for commentary on new papers and more details about my research as it unfolds.