Of course it’s a staple of postapocalyptic fiction - a handful of people (or sometimes just two, which raises all sorts of uncomfortable questions about incest) must repopulate the Earth, or humanity is doomed. But how many would it take, at a minimum, for mankind to make a new start without all sorts of inbreeding problems down the road?
4 pair. Without inbreeding there would need to be some mixing and matching of originals, but 4 pair may work…
We just did this exact question last week I think. Does anyone have it? It doesn’t take many people and inbreeding isn’t much of a problem past 1st cousins and closer than that isn’t necessarily a problem in the short-term.
Didn’t we pull this off with just one pair about 6,000 years ago?
::: d & r :::
Damned if i can find it by search, but yes, we did this very recently.
The general consensus was that one pair was actually sufficient. Or possibly even just one pregnant female, but one pair is icky enough i think so stick with that.
Of course there is no way to determine the actual number needed in reality as even with a large number of people there is no gaurantee that illness, accident and general bad luck wouldn’t wipe us out anyway. But the question (previously and now) focussed on inbreeding and that in itself doesn’t seem too much of a problem apparently.
Ah, here it is…
http://boards.straightdope.com/sdmb/showthread.php?t=410728&highlight=inbreeding
‘Inbreeding’ was the required search term! Found it pretty quickly under that.
While inbreeding does raise the odds for defects, couldn’t this be considered “accelerated mutation”?
I seem to recall (possibly from a PBS “Nature” episode?) that the cheetah has little to no genetic diversity. It was speculated that some fairly recent (in genetic terms) event killed off all but a single pair, or pregnant female.
So I googled before submitting, and it turns out that while cheetahs are genetically close, there is some dispute as to how close. cite
Since a few good answers to the question have come in, what if we replace genetic viability with some form of cultural rules about cosanguinity as the criterion?
Let’s say that retained cultural laws keep any pair as closely related as full second cousins (grandparents are full siblings) can’t have children. How big a starting population would you need to keep the comunity going on into perpetuity?
(Of course, if the founders are smart, they probably won’t have too many full siblings… mixing up the available pairs and having a lot of half-siblings makes more sense in the circumstance… though there are cultural barriers to that as well that we won’t go into now. :D)
Replying to my own question. (Tried to edit, but the time ran out.)
Sounds like 16 people would be enough for this, and 15 might be possible. With 16 people, then everyone in the fourth-generation would all be third-cousins of everyone else many many times over, but there would be plenty of people who aren’t half-second cousins, and the same thing on the third generation… you’d have 8 of the founders as your great-grandparents, and you’d need to find someone with the other 8 great-grandparents to completely avoid second cousinhood.
Thanks - good, useful answers! My concern for the future of humanity is at least somewhat assuaged. 
I think we can assume complete freedom from genetic defect in this case, no?
The lundehund dog breed was repopulated recently from only 6 individuals. I don’t know if they used insemination techniques, or if it was old-school nookie though.
Me and 10 women. I get to pick them.
That depends on your definition of consanguinuity. For instance, are double cousins (common ancestors on both sides) considered to be any more closely consanguinous than single cousins? If so, then I think you’re guaranteed to hit a wall eventually, no matter how many individuals you start with. And once you start getting any inbreeding at all (even with only nth cousins), it starts to become difficult to unambiguously define consanguinuity at all (at the least, the notion of consanguinuity becomes non-commutative).
The OP posits “…mankind to make a new start without all sorts of inbreeding problems down the road?”
I submit that the question is not answerable in light of present knowledge. Consanguinuity does not necessarily confer deleterious results. Actual results are totally unppredictable unless we know beforehand about some particular genetic abnormality. Thus one pair or one pregnant female might be sufficient from this perspective. This seems to be the only aspect of inbreeding that is being discussed.
But the OP doesn’t define “down the road” so we must also consider the loss of genetic diversity given limited breeding stock as an “inbreeding problem”. This is the point on which our knowledge is insufficient to answer the question.
What if the breeders, by random chance, lack resistance to, say, malaria? If any succeeding generation is exposed to the disease, we might lose the entire bunch regardless of number present in that generation.
Population biologists often see a phenomenon referred to as critical population minimum (or density). The real question being asked here, in population dynamics terms, is “what is the critical population minimum for humans?”. Note that millions of passenger pigeons were “above” this population minimum . But thousands of passenger pigeons were “below” this critical level-- once the population had been diminished to this extent, its extinction was inevitable.
The most diverse population you could have with just one living human being would go the other way: one woman, pregnant with fraternal twins, each having a different father.
A population can survive a very narrow bottleneck, but it can’t maintain itself forever at very low numbers; it has to reexpand and do so as rapidly as possible, so that if there are some harmful genes floating around, enough people will by pure chance not inherit them to keep the population healthy. Studies of south Pacific islanders have shown a disturbingly high incidence of inherited defects in isolated populations of as much as thirty thousand people. Fifty to one hundred thousand seems to be necessary for long-term genetic health.
I guess I’ll be the guy who brings up the fact that the vast majority of all golden hamsters are descendants from a litter discovered in 1930 and they seem to be holding up OK, to the best of my knowledge anyway.
As an aside, thanks for the link Kevbo. I was wondering about the factoid that cheetahs are so closely related that they don’t reject skin grafts from each other…it seems like one of those oft repeated things that could end up being BS, but it looks like it’s legit. That seems quite surprising to me.
Not at all. The longest truly isolated humanpopulation was found in Tasmania: ~4, 000 people isolated entirely from the rest of the species for 10, 000 years. Quite clearly you only need 4000 humans for long term genetic health.
The point you need to relaise about Pacific island populations is that they are neither long term populations nor usually isolated. Many pacific islands, especially the smaller landmasses but also larger ones like New Zealand, were settled within the last 1, 000 years. That’s less than 50 generations. Moreover few if any of the Pacific islands were isolated. The Polynesians were skilled enough Mariners ot make repeated contacts between those islands to maintain trading and genetic links.
As such the genetic defects found in Pacific populations don’t tell us anything about the size of a vaiable population. What we can say from the Tasmanaina example is that the long term viable population is less than 5, 000 individuals. The idea that 50, 000 is required simply doesn’t agree with the simple fact that a stable population of 4, 000 people survived quite happily in isolation for about 5% of the entire existence time of our species.
Are you planning something we should know about?