Human societies tend to be at least a little polygynous. This finding, recently reported in PLoS genetics, does not surprise us but is nonetheless important. This important in two ways: 1) This study uncovers numerical details of human genetic variation that are necessary to understand change across populations and over time; and 2) the variation across populations are interesting and, in fact, seem to conform to expectations (in a “we don’t’ really care about statistical significance” sort of way, for now) regarding human social organization.
Before examining the paper, we should mention four terms/concepts: Effective population size, Polygyny, and Operational Sex Ratio (OSR), and Variation in Reproductive Success (RS). The first two are used in the article, and the third I’m adding because it is a good one to know.
Polygyny is a term about mating systems, and unfortunately, mating system terms are usually used or understood incorrectly. Also, terms about mating systems should not be used as a substitute for terms about actual mating or population genetics. A mating system is a characteristic of a population, group, species, or some other collection of individuals, not a behavior of an individual or a pattern of genetic flow, though of course they are all related.
Briefly, “Monogamy” is the mating system where a male and female pair off, either serially (chaning mates over time) or life-long; “Polygamy” is a term we should never use in science, but technically it includes the two most common non-monogamous systems. Most of the time when you hear the word “Polygamy” used it is by lawyers or reporters and they are really talking about “Polygyny” which is a system in which one or more females may be mated with a single male. Like race horses, big horn sheep, mormons, and so on. “Polyandry” is where more than one male is mated to one female, as is seen in phalaropes (a kind of bird) and in a small number of human societies.
If there is a roughly fifty-fifty sex ratio at birth, then in Polygyny, one expects that some number of males are not mating at all, or at least not very often, with females, while some other number of different males are mating more often. But in fact, mating is not enough … what really matters is the difference among males in paternity of the various offspring borne by the females.
Since this is the number that really matters, we like the term “Operational Sex Ratio” or “OSR.” There are two ways to get a smaller number of males (relative to females) in the mating equation: One is if some of the males simply disappaer following this fifty-fifty birth ratio, so among adults, there are fewer males. The other is that these males are around perhaps wanting to mate, but they don’t for some reason, and we simply ingore them. We ignore them numerically by defining the OSR which is the ratio between actual mating/fathering/paternitizing males (the ‘operational’ males) and the actual mating/mothering/materitizing females.
Of course we are also ignoring some females, but in most natural populations of mammals or birds, far more males than are ignored by the OSR.
“Variation/variance in RS” is one of the most difficult terms for people to become totally comfortable with, in part because there is a way to understand this that makes you think you’ve “got it” but you don’t. Then you try to apply it and the world goes topsy-turvy on you and you have to start again.
The variation that is being measured here is simply the numerical variation among members of ONE SEX (totally ignoring the other sex) in reproductive output. So we look at, say females, totally ingoring the males, and we ask of a group of females “what was the average number of offspring these females had over a lifetime” and then we ask “Oh, and what was the variation in that number.” So if every female has one offspring and that is all she wrote, then the variation is zero. If some females had three or four offspring, a small percentage had zero, but most had two, then the variation is greater than zero, but not to much.
Then, we look at the other sex. “How many offspring, on average, did the MALES in this same population have?” Of course, this will be the same number as we had calculated for the females in this population, right? Then we ask “What is the variation among the males?” In a typical mammal (and humans are not typical mammals, by the way) we will find that a smallish number of males have lots of offspring each, a moderate to largish number have zero, and the remaining males have something in between. In other words, there will often be a LOT OF VARIATION among males. More importantly, there will be more variation among males than in females in most mammals … a lot more in some mammals. For birds, as a rule, there is more variation among males than in females, but not such a large difference.
This is important because any allele that affects mating will have an effect that is stronger where variation is greater. This is why we see exaggerated sexually selected traits (including large body size) in male mammals, but hardly ever in female mammals, while in birds we see sexually selected traits quite often in both sexes.
The final term to examine is “effective population size.” This is one of the more important aspects of the present study, because they were looking at the difference between males vs. females in relation to population genetic change over time. In population genetics, size is everything (size of population, that is). So, if there is a skewed OSR then the effective population size of males will be different than for females.
Here is what the study found:
… the mating system of humans is considered to be moderately polygynous (i.e., males exhibit a higher variance in reproductive success than females). As a consequence, males are expected to have a lower effective population size (Ne) than females, and the proportion of neutral genetic variation on the X chromosome (relative to the autosomes) should be higher than expected under the assumption of strict neutrality and an equal breeding sex ratio. We test for the effects of polygyny [note: here I would prefer the term “OSR”] by measuring levels of neutral polymorphism at 40 independent loci on the X chromosome and autosomes in six human populations. To correct for mutation rate heterogeneity among loci, we divide our diversity estimates within human populations by divergence with orangutan at each locus. Consistent with expectations under a model of polygyny, we find elevated levels of X-linked versus autosomal diversity. While it is possible that multiple demographic processes may contribute to the observed patterns of genomic diversity (i.e., background selection, changes in population size, and sex-specific migration), we conclude that an historical excess of breeding females over the number of breeding males can by itself explain most of the observed increase in effective population size of the X chromosome.
Here is a chart showing the results:
Behavioral biological theory would make the following two predictions regarding these data:
1) The degree of body size dimorphism (with males being a bit larger than females) should be greater in the populations with the highest OSR (the ones higher on the graph), assuming that there is some effect having to do with male-male competition (warning, in the absence of behavioral evidence of males fighting over females, this may be a weak hypothesis. This will depend on what you think about males fighting over females in each population. Do they? Do you know?); and
2) The populations with the lowest OSR (those lower on the graph) should be foragers or egalitarian societies, while those highest on the graph should be societies where variation in personal wealth is more likely to occur.
I quickly add that the 95% confidence intervals overlap all of these populations, so interpretations related to these predictions using these data would be mere arm waving. And, if arm waving is an adaptive trait, say, for males, in a certain society, then we should see longer arms in males in that society. Which brings us back to the orangutans.
Michael F. Hammer, Fernando L. Mendez, Murray P. Cox, August E. Woerner, Jeffrey D. Wall, Dmitri A. Petrov (2008). Sex-Biased Evolutionary Forces Shape Genomic Patterns of Human Diversity PLoS Genetics, 4 (9) DOI: 10.1371/journal.pgen.1000202