A curious shift occurs during and right after a war: more boys tend to be born than girls. It's been documented for decades in many nations, especially during long conflicts with many troops deployed. The cause of this boy boom has long flummoxed thinkers and scientists. Ideas have veered from the theological—a divine call for new men to replace those lost in battle—to the coital—returning soldiers have lots of sex, and so will be more likely to fertilize at a time in their ladies' cycle that's ripe for making boy babies. A new study in the journal Evolutionary Biology rejects them all. Instead, it pins the "returning soldier effect" on a gene expressed by men only. It also shows how researching your family tree can help you place bets on the sex of your next kid.
"I wasn't satisfied with the explanation that it was due to couples having more sex," says Newcastle University's Corry Gellatly, who did the work as part of his Ph.D. thesis. Gellatly was curious about studies of male shrimps, marine worms, and yes—human males—that showed that their likelihood of producing male offspring seemed to mimic that of their parents. In other words, males who have more brothers than sisters would in turn produce more sons than daughters. Thinking this may be the root of the wartime peaks, Gellatly sought to investigate the trend on a large scale. He sifted through 927 North American and European family trees posted on an online database by both amateur and professional genealogists, and tallied the sex ratios of siblings for each generation.
"The family tree study showed that whether you're likely to have a boy or a girl is inherited," says Gellatly. "We now know that men are more likely to have sons if they have more brothers but are more likely to have daughters if they have more sisters." Women, however, did not have the same tendency.
Gellatly argues that a gene—which is carried by both men and women but only active in men—influences what proportion of a man's sperm carry the X chromosome and how many carry the Y. The sperm's X or Y status determines the sex of the baby upon meeting the egg, which only carries the X chromosome. More Y sperm=more XY (male) babies.
How might this gene tip the baby balance after a war? Consider the legacy of two hypothetical men—let's call them John and Rich. John has three sons, all of whom go off to fight, and one daughter, who does not. Rich has the opposite family structure: three daughters and one son. John is more likely to see multiple sons return from war alive—and with Y-leaning sperm. So they'll produce more sons of their own. Rich is likely to lose his only son, which if alive, would have fathered girls to even out the sex ratio. The mechanism, Gellatly's genetic model shows, shifts the sex ratio back to normal as the dip in male mortality recovers.
The genetic explanation of the returning soldier effect is, for now, a thought experiment—the gene responsible has not been found. In the meantime, let's ground-truth Gellatly's hypothesis further: Guy PopSci readers—have you noticed a tendency toward boys or girls in your own family trees? Do tell.
I have noticed the same effect in my family. Over the last 4 generations, the males in my family have averaged 4 male children for every female child born. I tried to explain that to our ob-gyn as he performed an ultrasound on my wife, but he insisted it was a straight 50/50 shot. Who's laughing now?
Well, he is, I suppose, since he still makes 4 times what I do.
I think it's more likely that levels to testosterone can effect the levels of y sperm and x sperm.
Men who are sent to war are embedded in a testosterone environment both off the field and on the field where aggression is key to success.
They say boys go away boys, and come back Men... which is after being behaviorally and environmentally embedded in a more aggressive testosterone laden environment for years on end.
It makes sense they would produce more testosterone in response to the environment. It makes sense from an evolutionary point of view that creatures in aggressive environments would have greater testosterone levels... but a greater risk of males dying.
The obvious survival trait is that the higher levels of Testosterone resulted in replacement males.
If what he says is true, and this is single most important mechanism for determining a child's sex then after thousands of conflicts over the history of mankind, their would be mostly men, and few females.
If his hypothesis was correct, it would affect more than one generation of children, because this supposed male producing gene would be carried by all the grand-children, and great grand children, etc. Boom another war happens and it further reduces the female favoring gene. We know this isn't the case though.
The problem with these studies is they try to isolated complex systems (in this case genetic programming) to a single variable, if you believe in evolution, you have to consider the possibility that multiple biological mechanisms or regulators developed in parallel for things like this.
So I throwback that, yeah maybe there is a correlation, but you still haven't figured it out completely.
20 some years ago I read of a study that said societies under stress start breeding sooner. I think the point was that hormones were produced earlier to replace the population sooner.
To me it helped explain why children of the poor, the disenfranchised, fatherless and hopeless, often living in violent neighborhoods, might start having sex at a very young age.
We know better now, tho. We simply blame it on TV and Democrats. (haha)
And WOW to that powerful photo. You just know that somehow a male child was forming in the nurse's belly at that very moment!
Hmm, pretty valid question. Never thought about it that way before.
My family is loaded with boys. My brother and I (and my sister) have only produced sons and no daughters. I can't wait to see what our sons produce. After they are married and self sufficient, of course.
timias, your logic assumes that most of the population of the world was at war at any one time. While the number of wars at any one time may be considered high, all wars have detractors, which would serve to balance the fighters and non-fighters. So, even if the fighters brought in more sons, based on the ideas in the article, the non-fighters would have more balanced offspring numbers. This would water down the "returning soldiers effect" and preclude the massive male domination you describe.
The article's hypothesis is simply wrong. If half the men sent to battle tend to father more boys than girls and the other half tend to father more girls than boys, then just as many men from each group will come back.
The idea that fathering more boys than girls is genetically inherited is unlikely since over time (in both peace and in war) more and more men with that trait would be born, eventually overwhelming the other type of man.
Dougl, whilst your main point is most definitely valid (removing random individuals from a population with multiple traits can't affect the overall balance of those traits within the population absent other factors), your assumption that the traits will be distributed 50/50 is almost certainly wrong - because absent other factors that's not a stable balance.
If men from families with a larger number of boys tended to father boys themselves, then (starting from an equal distribution and picking males to reproduce purely at random) after a single generation the balance would have shifted in favour of the boy-fathering trait. The balance would shift further with each successive generation, and the system would tend to a significant imbalance of males over females in the population as a whole.
As that's not the case, if we accept that a boy-fathering/girl-fathering divide exists, then it also follows that, in a stable population, girl-fathering males must normally be producing more children per head than their boy-fathering counterparts - i.e. they must normally have some compensating degree of advantage in finding mates, with a relatively larger proportion of boy-fathering men normally failing to produce any offspring at all.
Which brings us back to wars, and a more likely mechanism. One of the most notable effects of war has historically been to reduce the overall male population, often drastically. And whilst the balance of boy-fathering men versus girl-fathering men in the remaining population won't be changed to any meaningful degree, the reproductive balance of males to females most definitely will be. With the same number of women but fewer men, boy-fathering men who might otherwise not have had offspring become more likely to reproduce - and if that happens, the balance of the sexes shifts for a generation (after which the old girl-fathering advantages, whatever they are, start to reassert themselves, and things stabilise).
So whilst I entirely agree that Mr. Gellatly's statistical reasoning is flawed, I don't necessarily disagree with his overall conclusion that boy- versus girl-fathering may be the mechanism triggering a surge in the male population after a war. (And for what it's worth, I'd also point out that there are a couple of testable predictions in my own version of the scenario.)
fredd - While I agree that the smaller number of males, left after a conflict, allows for each male to produce more children (as they have to serve a greater number of females each,) it is not just the boy-fathering men, as you indicated, who will be having all of this reproductive sex. Both the boy-fathering men and girl-fathering men will be having a good time and hence there would be no change in the male-female ratio of births after a conflict.
I think instead, that the previous comment about testosterone being involved here in the boy-fathering has some basis:
One way - if boy-fathering men tend to run at higher testosterone levels, they will be more aggressive than the female-fathering men and hence more boys are being born.
A second way - That higher level of testosterone in the system, that perhaps being at war for awhile created, increases their Y carring spearm % levels (theory). However, the higher levels of testosterone would have to go back to more normal levels after awhile or the boy-fathering effect would continue and upset the balance of males to females in the long run. A reasonable expectation if action in war increased the soldiers testosterone levels over time, then it should go back to more normal levels, over time, as the stress of war was removed from their lives.
My father's family included 3 boys and 3 girls (all alternating in birth order). My family included 1 oldest sister, 2 twin girls (miscarried), and the youngest male child (myself). I haven't had any children yet, so I don't know how that will pan out. As for the article, while I think it is a really well-done analysis, I'm not sure if a fixed gender ratio for specific lineages is really likely. Having daughters vs. sons is more often the best evolutionary strategy, because females generally have better odds at producing offspring, while males are better, only if they can monopolize resources visa vis other males (which is partially dependent on parents ability to invest in the male offspring). That would predict trait(s) that were more sensitive to ecological variability, rather than being fixed, in determining gender ratio for offspring.
I've been enjoying the comments posted here, but I need to clear something up. It is commonly assumed that wars remove men from the population at random. In fact, this is not the case, I looked at conscription records from WWI and found that at the start of the war, most soldiers were 18, while the average age was under 27.
Now, imagine an extreme hypothetical example, where every 18 year old male is suddenly removed from the modern population. It is clear that the removal of males from the population would not be random, but distributed evenly between families, because each family can only lose one male (unless they have male twins).
If you remove a single son from a family with two sons, this removes 50% of their sons, whilst from a family with five sons, it removes 20%. Across a population, this translates to a greater loss of males from families with less sons.
I have found by looking at family tree records, that males inherit a tendency to produce either more sons or more daughters from their parents. It is possible to say, therefore, that families with more sons were more likely to have sons still alive after the wars and those men were more likely to have sons themselves, causing an increase in male births after the wars.
I intend to do further work on exactly how the age structure of fathers changed after the wars, but I hope I have cleared up the misconception that my hypothesis cannot be right simply because wars would have removed males from the population at random.
I am a member of Mensa. I recently retired after working for 45 years, the first 11 (peacetime) years in firstly the Navy and then the Air Force, the remaining 34 years as an accountant.
During my time in the armed forces I noticed, beyond any reasonable doubt, that my colleagues collectively produced significantly more male offspring than female. During my time as an accountant the reverse was clearly obvious amongst my accountant colleagues, although not so amongst sales and engineering colleagues,
My theory about this is that timing is of the essence; accountants and their chosen life-partners being more careful and methodical, and therefore more likely to plan their mating, whereas pilots and their ilk as well as salesmen and engineers etc are less bothered about precise planning and just go for the "scatter gun" approach.
Certainly, I've read that male and female-producing sperm have inbuilt differences that produce different rates of sperm travel and sperm longevity, with a consequent favouring of one resultant gender as against the other for each of the sperm types.
Could it therefore be that wartime engenders a "now or never" environment which increases "scattergun" mating?
raverman, if your hypothesis is correct, that it's the testosterone level and not simple genetics, then that would still lead to men who have more brothers than sisters fathering more sons than daughters, because brothers tend to boost each other's manliness just by roughhousing, while an abundance of sisters tends to make guys more sensitive.
If we look at the genetics on the other hand, if both genders truly carry this gene, then you could have a situation where the father had only brothers and the mother had only sisters and vice versa, so which gene does the son end up inheriting?
I'd be curious to see what children I have, since my father had only brothers, my mother had only a sister, and I'm an only child. ;-)
Interesting Theory, but I suspect something else. In a real shooting war, if you don't get killed, you get highly stressed.
Stress is known to trigger hormones such as cortisol, norepinephrine, epinephrine (adrenaline), and human growth hormone. Some of these are nasty hormones known to be very destructive to parts of the body. Human growth hormone on the other hand has some very positive effects, plus some tendency to increase the testosterone previously mentioned. I think it is more likely that either one or more of the hormones has a negative impact on the X chromosomes or that HGH or some other hormone has a positive impact on the Y chromosomes.
I suspect the high stress environment is tilting the hormone balance in a way which is then tilting the X/Y chromosome balance. This may even be damaging the male body's ability to produce X chromosomes.
I would be curious to see a population study of mental and physical stress (or various hormone levels) versus offspring gender.
>>If you remove a single son from a family with two sons, this removes 50% of their sons, whilst from a family with five sons, it removes 20%.
A family with 5 sons will have 5 times the chance to have a son of "conscriptive" age at the time of war. So if we extend the extreme hypothetical example to 5 families of each type - results will even out.
Is there something else I am missing?
Larisa: you have got to the crux of the idea. It is true that if all men (of fatherhood age) were available to fight and their names were picked out of a hat, then families with more sons would also lose more sons. However, in Britain in WWI, for example, the most common age of soldiers was 18 at the start of war, and most soldiers were quite young. It was not as if the soldiers were composed of a random selection of men of fatherhood age.
In order to simplify the argument, imagine only 18 year old soldiers were sent to war and they all died. In this case, it is clear that all families who lost a son would have lost one son (except those with male twins, but let's leave that aside).
Imagine the following families in the population before the war, all of which have adult sons, one of which is 18 years old:
A) 100 "male-biased" families, who have 1000 children altogether, 800 of which are sons.
B) 800 "equal-bias" families, with 8000 children altogether, 4000 of which are sons.
C) 100 "female-bias" families, who have 1000 children altogether, 200 of which are sons.
You will notice that 16% of sons are from "male-biased", 80% from "equal-biased" and 4% from "female-biased" families.
Now, the war happens and removes 1 son from every family, so you get this:
A) The "male-bias" families have 700 boys left.
B) The "equal-bias" families have 3200 boys left.
C) The "female-bias" families have 100 boys left.
In the population, you now have 17.5% of sons from "male-biased", 80% from "equal-biased" and 2.5% from "female-biased" families.
The percentage of boys from "male-biased" families has increased relative to those from "female-biased" families, and because those sons are more likely to inherit the tendency to have more sons themselves, there is an increase in male births in the next generation.
I hope this makes sense. It is an idea that assumes the way in which soldiers were drafted into the army was such that they were distributed fairly evenly among families. I think the fact that soldiers were young can explain this, but it is still an idea and more evidence is required to support it.
To Corry Gellatly:
It seems to me that it does not work this way :)
Let's agree with your model, where the son of age 18 is sent to war and dies.
Here is your basic assumption: "Now, the war happens and removes 1 son from every family".
In a family with less sons there are proportionally less chances to get a son of conscript age at the moment of war. Thus, it is less likely that such a family will loose 1 son (again, proportionally to the total number of sons in a family).
Do I miss something in this reasoning?
Talking of alternatives, there is a theory that frequency of sexual contacts is affecting the proportion of males in offspring (via % of spermatozoids with X and Y chromosome). If a man has more sexual contacts, it becomes more likely for him to father a boy. That is a pretty easy way of regulating imbalances: if there are more women than man, as it happens often after the war, more boys are born.
Did you ever consider such an explanation?
Anyway, it's an interesting theory of yours, and it definitely deserves further exploration (in my humble opinion)
To Corry Gellatly:
let's take 10,000 families of each type (A - 80% girls, B - 80% boys in the family)
Type A, same as Type B:
Let's say 10 children in each family:
10,000*10 = 100,000 children in each group
Type A: 80,000 girls, 20,000 boys
Type B: 20,000 girls, 80,000 boys
B to A boys ratio=4:1
Let's assume that a chance for a child to be of age 18 is 1/100th (lifespan of a 100 years (ha!) ), does not matter boy or girl
Type A group, same as Type B group:
1 000 children of each group are of age 18. Of these children
Type A: 800 girls of age 18; 200 boys of age 18
Type B: 200 girls of age 18; 800 boys of age 18
If every 18 year old boy is taken away:
from the 10,000 + 10,000 families we will have 200 Type A boys + 800 Type B boys conscripted.
type A: 80,000 girls; 20,000 - 200=19,800 boys
type B: 20,000 girls; 80,000 - 800=79,200 boys
If conscription is done by age - the ratio of boy-fathering males to girl-fathering males will not change after the war. If "one per family" is the rule - then you are correct, ratio might change.
(I have not read an original paper, only the first page preview and various interpretations in popsci and on other news sites.)
Rezoner: for the purposes of the example, assume that all men are of the age to go to war, but only 18 year olds get drafted. The families who do not lose a son do not come into the equation, because their contribution to the population does not change.
Larisa: it is quite paradoxical I must admit, because your sums are correct and so are mine, but it is the paradoxes that make biology so interesting, so it's a good thing! The point about taking only 18 year old boys from the population, is that you must be taking one from each family, because apart from families with male-male twins (which are really quite rare), each family can only have one son aged 18. There will be some men with two 18 year old sons by different mothers, but this will be quite rare as well. I do find this quite difficult to understand myself, so I might be missing something. p.s. send email for paper.
>that you must be taking one from each family...each family can only have one son aged 18.
Each family will have only one boy of age 18, but some families will not have any boys of age 18! Families with less boys have better chance of not sending anyone to war - that would explain my calculations.
Looks like more information is needed, specifically if "one per family" rule was in effect at any time at all. In Russia during WWII conscription was done strictly by age, exceptions based only on the profession, education status etc. During WWI and earlier there was a rule to NOT draft the oldest son in the family, which would have an opposite effect I think. This is a research for a historian, probably.
I would say that short-term variations are still not explained by "boy-fathering gene" theory. Or at least one can still argue...
P.S. I've sent an email to ncl.ac.uk address.
Not being a scientist, a doctor, or a mensa member, I am just reading these arguments and theories out of curiosity; however, it seems to me that just about every one is misguided.
I do not understand why you would be creating scenarios based on families when the only thing that matters for this argument is the male population--individuals.
For instance, if we have a male population of 100 million males of which 20% are male-biased, 20% are female-biased, and 60% are neutral, and if a portion of that male population is lost to war, there is no reason why the population lost to war would not be approximately the same make-up (20-20-60) as the general population. Their families have no bearing on it. Thus, war would not have any effect on the ratios of production of male versus female babies.
Regardless of war deaths, though, it would seem that the male-biased segment of the population would produce a disproportionate number of male offspring which would be disproportionately also male-biased; thus yielding a general population that was increasingly male-biased.
The phenomenon of increased male baby production during and after war seems to exactly contradict what you would expect to observe (above).
My questions are: 1) why do we not see a population having an increasing ratio of males to females in ordinary, peaceful times? 2) Why do we see increased male baby production in war times?
Both these phenomena seem contrary to logic to me if you assume that the bias toward producing male babies is carried in males and is inherited by their sons.
As for the adrenalin or other hormonal theories, I think that if you factor in the small percentage of the total male population that actually will be involved in combat in modern warfare, those soldiers will have to have a very, very good time in order to affect the general population statistics.
Also, male sports activity would create similar stresses, albeit to a lesser degree, in wartime or peacetime.
I will try to check for responses. Maybe one of you can show me what I am missing.
Corry. Your theory based on non-random selection of males to be lost in wars seems to make sense at first. It is true that families with more males are more likely to have surviving males after the war. I think that that advantage is offset by the disadvantage that they are also more likely to have sons of fighting age and would be more likely to lose a son in war.
That would mean that the men lost in war will be weighted in favor of more male-biased men than neutral- or female-biased. That would seem to work against producing more male babies--not for it.
You could successfully argue that a family that is male-biased will have more male children and thus be more likely to have surviving male children who will reproduce. That much may be true; but that only speaks about that family, not the general population. The offsetting phenomenon is that a family that is not male-biased is less likely to lose a son at all and thus does not lose any reproduction capability while the male-biased family did lose some of its reproduction capability.
Again, that would mean that overall, war losses should result in less male baby production, not more.
The hypothesis is flawed from the start. The selective transmission of a gene should work outside the war too, so in time the all the male population will have that gene.
(Let's assume one family and 4 kids) - 3 males with this gene vs 1 male without - next generation 9 males vs 1 male and so on....
raverman - You are assuming that higher levels of testosterone will result in more Y chromosomes - more boy babies. This may or may not be the case; you haven't said anything to support this assumption.
Isis - You mention that you are a member of Mensa, but you don't say why this is pertinent to the discussion.
Corry Gellatly - Your reasoning makes a certain amount of sense. Is there any evidence that there is, in fact, a gene which favors Y chromosomes over X chromosomes in a given male? Is there evidence that there is a greater number of X or Y chromosomes in some males and the converse in other males? It will be interesting to see what further research reveals - and I *love* that picture!
"As that's not the case, if we accept that a boy-fathering/girl-fathering divide exists, then it also follows that, in a stable population, girl-fathering males must normally be producing more children per head than their boy-fathering counterparts - i.e. they must normally have some compensating degree of advantage in finding mates, with a relatively larger proportion of boy-fathering men normally failing to produce any offspring at all."
It seems like there could be both biological and social reasons. Perhaps boy-fathering males have a similar proportion of viable y spermatids but have a higher rate of non-viable X spermatids. This would result in lower fertility for boy-fathering males when there is more competition for mates/reproductive encounters. Perhaps the genetic difference is something like a CCG repeat that becomes amplified over successive generations and ultimately results in complete Y spermatid infertility such that families of boy-fathering men ultimately give rise to female-fathering men. From a social perspective, it was common for social roles to be based on birth order with first sons getting the business, second sons going to the military, and third sons going to the clergy and becoming infertile as a social construct. Also, risk taking behavior is a common cause of early male mortality and hence loss of reproductive potential. Single male children seem to be more closely protected and less likely to engage in these behaviors (momma's boys). Males with female siblings are often also more socially comfortable with women have more chance of being introduced to females of similar age, and probably have a reproductive advantage with better courting skills. All these factors would seem to work hand in hand to maintain equilibrium once the male population was replenished.
The somewhat more uncomfortable question is whether periodic war and killing off of males is advantageous to mankind. There seems to be a predictable cycle of postwar prosperity followed by economic decline, increased unemployment and crime, and then war. These things seem to parallel increases in male population. We now fight proxy wars with low casualty rates and seem to be stuck in a state of perpetual war with prolonged economic decline and unemployment.
I think I have it.
1) Large number of women have "sweethearts". Some of them are the kind of men who will have sons, some the kind of men who will have daughters.
2) It's hard to meet and socialise before the war, and more so after the war with Europe in ruins.
3) The women who have sweethearts who have lots of brothers, after the sweetheart is killed, have ready access to his brothers who will also produce lots of sons. They meet at the memorial service, support one another in the coming months, get married and have children.
4) The women who have sweethearts who would have had daughters, but were shot, meet his sisters at the memorial service and become the generation of women who did not have kids.
If the original article were true then we would have 2 boys not 2 girls I don't agree at all with his studies as I've done our family tree to the minute detail and going based off of what was presented above I should have been a boy so 3 out of 3 time in personal experience if the researcher hasn't gotten one right pretty good chance there's no concrete basis