How many unique DNA samples does a population need in order to survive?

[Maloney]

Say you had a group of modern humans who are all unrelated. What would be the minimal amount of people you'd need to sustain a population for 10 generations? 100? A million?

Assuming that every member was heterosexual, fertile, had no genetic problems, and was mating completely optimally*, how few people could populate an entire planet without problems?

What would be the worst-case scenario? What problems would arise if there were too few unique members?

*Also, what would this optimal mating structure be? How would it work?

[halfeye]

Say you had a group of modern humans who are all unrelated. What would be the minimal amount of people you'd need to sustain a population for 10 generations? 100? A million?

Assuming that every member was heterosexual, fertile, had no genetic problems, and was mating completely optimally*, how few people could populate an entire planet without problems?

What would be the worst-case scenario? What problems would arise if there were too few unique members?

*Also, what would this optimal mating structure be? How would it work?

It is said that a population below 100 is in trouble. We don't have that, and are unlikely to have that in the near future.

That said, there's some reason to think that if the population can grow exponentially from that point on, the minimum starting population may be a lot smaller than 100. We don't know what it is because nobodies been in that situation and survived and written about it.

The question sounds suspiciously as if it's trying to set up someone else to suggest a harem as a solution, which is an unrealistic fantasy if there are no other people about.

Basically, we don't know, because for obvious reasons no one's done the experiment on humans.

[DavidSh]

Do we know the bottleneck population size of the Bounty mutineer settlement on Pitcairn? The size of a typical Polynesian colonization fleet? I was going to ask about the size of the group of colonizers of Hawaii, except there may have been more than one wave of colonizers for that.

[Maloney]

Do we know the bottleneck population size of the Bounty mutineer settlement on Pitcairn? The size of a typical Polynesian colonization fleet? I was going to ask about the size of the group of colonizers of Hawaii, except there may have been more than one wave of colonizers for that.

The Pitcairn settlement was only two generations, going from 14 to 46.

[Radar]

It is said that a population below 100 is in trouble. We don't have that, and are unlikely to have that in the near future.

That said, there's some reason to think that if the population can grow exponentially from that point on, the minimum starting population may be a lot smaller than 100. We don't know what it is because nobodies been in that situation and survived and written about it.

The question sounds suspiciously as if it's trying to set up someone else to suggest a harem as a solution, which is an unrealistic fantasy if there are no other people about.

Basically, we don't know, because for obvious reasons no one's done the experiment on humans.

There were no experiments but with current knowledge of genetics and medicine, we can make pretty solid statistical prognosis and this is what scientists came up with.

edit: short answer is about 10 000 or more.

[Tvtyrant]

I doubt there is a single organism whose ancestral line hasn't dipped below 10,000 in the past. They based that on stable or healthy amounts, ie high enough to avoid rapid changes.

[halfeye]

There were no experiments but with current knowledge of genetics and medicine, we can make pretty solid statistical prognosis and this is what scientists came up with.

edit: short answer is about 10 000 or more.

That's for a population with absolutely no chance to grow. It sounds like a lot, but it might be right, in that particular and peculiar circumstance.

[NichG]

It looks like those simulations don't include new mutations or selection effects as inbreeding starts to kick in, and the focus is mostly on whether you preserve most of the existing human diversity rather than whether you could have a surviving population. So I'd guess the minimal number just for survival is actually quite a bit lower, and the easier it is to surviving in the environment you're transplanting the group into, the smaller you could push the starter population. But you might have a period of time where say 2 in 3 children die before reproducing due to accumulated recessive traits - which is still a growing population if everyone tries to have 6 kids for example. And if the environment is at all harsh, that could certainly make the population go under.

[Grey_Wolf_c]

Rule of thumb I’ve heard in the past is 50 for short term viability, 500 for long term. Any smaller and the genetic defects accumulate faster and more commonly than natural selection can weed them out.

The depressing study of species on the verge of extinction has plenty of hard facts on the matter, I’d imagine.

Grey Wolf

[Kato]

I think we had a similar thread a few months ago, but I don't know how much archiving gitp is still doing. Though maybe you can still find it.

First off, if by 'healthy' you mean not only no obvious problems but actually people with a 'perfect' genome, you could work with a very small number. (but I'm not sure if such a person exists) Birth defects hardly appear spontaneously and if there's no genetic problem, it's likely that almost all kids will be healthy and healthy to interbreed.

Picking random people from the population and doing some basic checks, I still think the number isn't too high, as long as you don't expect disasters to wipe out half of them or something. Of course this is just survival and not maintaining diversity of human genome etc.

I think that link was brought up in the previous thread, too, but it does little elaborating on what exactly those simulations were and a colonizing another star system while maintaining human diversity is not quite the same as the smallest bottleneck population.
Somebody posted a study on endangered animals which, while not directed at humans I found more convincing...
Oh, wow, I actually found it again, but posting links on my phone is annoying. Look for 'How does the 50/500 rule apply to MVPs?' by ian G. Jamieson and Fred W. Allendorf and you should find it. It's not perfect but might give you a clue.

[The Jack]

How unethical are we talking here?
You could get away with a much smaller starting population if you selectively bred and tried to maximise the output of your starting populations and you enforced cycling partners in order to minimise incestual combinations. You could return to normalcy eventually.

If you're going to be humane, you need a whole lot more. If you're in limited conditions (IE you can't just have every woman give a child to every guy and expect to have everyone remain fed) you need a whole lot more.


Other than that, I'd be interested to know if there are people that don't have any bad genes and who could interbreed frequently without any issues for many generations.

[Balyano]

Native Americans are suspected to have had a starting population in the range of 50-200 or 200-300 or 250-750. Depends who you ask. And those people were related to each other.
Best to avoid anything closer than 2nd cousin on a regular basis.
I recall reading that a population can have 3rd cousin marriage every generation without a problem.

You can get away with a very small starting population if you can maintain a high birthrate. The rate of individuals reaching adulthood to breed needs to exceed the death rate. If a population remains small for too long the mutational load will build up and lead to what is known as mutational meltdown, where stillbirth and birth defects make up such a large portion of fertilizations that death exceeds births.

Based on the rate of stillbirth, birth defects, and severe mental dissability among inbred couples, including comparison of 1st cousin, uncle-niece, sibling, and parent child. The average individual contains within their genome between 4 and 7 lethal recessive alleles, and about the same number of deleterious alleles that cause a variety of birth defects. The average individual contains about 1 allele that causes severe mental handicap.

If you really want the minimum starting population you can do it with 1 person.
Create a bunch of artificial wombs. Create sperm and eggs from the persons genome.
Commit a more extreme form of incest than any in history.
You will end up with about 0.25%-2.24% "healthy" individuals afterward.
I say "healthy" because they will be homozygous for a large number of genes that are suspected to be better when heterozygous.
So if you perform 1 million fertilizations, and assuming some losses for other developemental reasons, perhaps you get between 2,000-20,000 people from that.
Repeat the process and each generation you do this you get fewer and fewer stillbirths and defects.
This is called genetic purging.
After a few generations you will have more or less removed the lethal alleles and birth defect causing alleles from the gene pool.
After that point you can stop using the artificial wombs, since the people can then breed faster than the stillbirth rate.
The population would most likely suffer from what is called ''overdominance''.
If so expect them to be a bit shorter and overall less healthy than a normal human.
This is common in purebred livestock.
High and low human intelligence has been linked to high mutational load.
It is likely that many such mutations would be purged by this process.
Expect the population to have very average intelligence, with less variance than you would see in a human family, let alone a human town or nation.
Probably fewer idiots, but also probably fewer geniuses.
Over time mutation will add in new variation, mostly negative and silent mutations, but sometimes you get a good one.
The population would be vulnerable to disease, especially viruses.

Can't really suggest this one for real life.

Could make for an interesting cultish/totalitarian scifi culture, perhaps decended from the lone survivor of a disaster isolated far from the rest of humanity. Perhaps their ''All-Father'' generations in the past set them on a century long mission to return to humanity. The orginal purpose was to interbreed and regain normal human variation, but over the generations their interpretation of his orders morphed into one of conquest? extermination? who knows. Could keep using the artificial wombs for cultural reasons. They see it as normal. Perhaps they have strong pheromonal instict that makes them repulsed by the idea of breeding with eachother? Spend all the sexual energy on work, very very industrious?

[HouseRules]

For humans?
Inbreeding Depression (2nd cousin or closer)
Inbreeding Enhancement (3rd or 4th cousin)
Outbreeding Enhancement (5th or 6th cousin)
Outbreeding Depression (7th cousin or further)

Since a population needs to have sufficient variability to have 3rd or 4th cousin, for the shortest of terms, 4 generations back is 2^4 = 16 at the absolute minimum.
That is, an absolute minimum of 8 unique and totally unrelated males, and 8 unique and totally unrelated females.

For intermediate terms, 5 generations, or 2^5.
That is, an absolute minimum of 16 unique and totally unrelated males, and 16 unique and totally unrelated females.

For long terms, 6 generations, or 2^6.
That is, an absolute minimum of 32 unique and totally unrelated males, and 32 unique and totally unrelated females.

These absolute minimums are violated when the unique and totally unrelatedness are violated.

Given that in order to have 1 unique DNA sample, you need at least 512 people.
Thus, at a minimum, you need 16384 males and 16384 females, for at least 32768 people.


*Ideal Mating Structure clearly depends on culture.
Ideally,
(Mass of Male)^2 / (Mass of Female)^2 = Females per Male
(Mass of Female)^2 / (Mass of Male)^2 = Males per Female

Polygamy - which kind?
Polygyny may increase the minimum number of unique females.
Polyandry may increase the minimum number of unique males.
Polyamory may increase the minimum number of unique individuals.
--------------------------------------------------------------------------------
Second Half Cousin vs Double Third Cousin?
They are genetically the same distance.

Isotope dating is flawed.
It assumes that the sun emits Carbon 13 and Carbon 14 at a constant proportion.
That assumption if flawed because this is assuming that a hydrogen bomb does not use its hydrogen during the fusion stage.

[Radar]

Isotope dating is flawed.
It assumes that the sun emits Carbon 13 and Carbon 14 at a constant proportion.
That assumption if flawed because this is assuming that a hydrogen bomb does not use its hydrogen during the fusion stage.

Are there any publications on those flaws? From a first glance the timescale for changing Sun's elemental proportions is many of orders of magnitude larger then that for dating human or animal remains by C14.

[Themrys]

*Also, what would this optimal mating structure be? How would it work?

Female choice, in a matrilineal family structure where only women own property, so men's wealth can't skew women's choices.

It's scientifically proven that people like the smell (and probably facial features, but I don't recall a study on that) of those who have sufficiently different immune systems. Men have too much sperm to waste, so are not choosy enough, therefore if you want to use all women's fertility optimally, you need to get rid of rape and all kinds of coercion, as all male-chosen matings will produce suboptimal offspring - men lose nothing by producing less-than-perfect offspring, women do.

DNA testing is not nearly far enough progressed to be able to predict which to humans will have the fittest offspring. Completely free female choice is your best bet. (Ideally, you'd get rid of all psychological problems that can lead women to reenact childhood trauma rather than choosing the best mate, too.)

If you want things to be a bit wacky, you could have match-making that works by women smelling men's clothes, in an effort to eliminate all other factors. (Could be that facial features and the like also play a role, though, in which case this would be counterproductive.)




Something to consider: Mutations occur more often with older parents. Since mutations presumably increase genetic variation, it might be wise to make sure a lot of women have their first child at over 30. (You want it to be over 20, perhaps even older, to reduce the rate of death in childbirth. You can't afford death in childbirth with your as-small-as-possible-already population size. Early pregnancy is unhealthy or sub-optimal for many mammals, and usually prevented by groups with an intact social structure - that's the real reason why the "alpha wolf" doesn't let other wolves in the pack mate - in a natural structure, they'd be her immature children.)

You also need to reduce child death rate. Big cities with low tech increase child mortality. You want people to live in small villages where diseases don't spread easily.



Humans differ from other mammals in that birth is an enormous risk and usually only produces one child. Some of your genetically unrelated individuums may die in childbirth and not give birth to a living baby, so you lose them from the start.

500 is, I think, the absolute minimum, and I would say, you probably need 1000 in all (500 female) to account for that.


Polygyny doesn't work - every species of animal I know of produces about as many males as females, or, in case of bees, even more males than (fertile) females. Modern beekeepers think they can just kill off the "excess" males, but I think that will come back to bite them later. Species would have evolved to produce fewer males if they didn't need the genetic diversity.

If 50 women in your population of 1000 choose to mate with the same man (because, as explained above, you have made them financially independent from men, so his inability to pay for that many children doesn't matter), that doesn't mean you could just have chosen one man at random for 50 women. The 49 that may not ever get to father any offspring are needed, too - precisely so that there is a choice.

[HouseRules]

Female choice, in a matrilineal family structure where only women own property, so men's wealth can't skew women's choices.

It's scientifically proven that people like the smell (and probably facial features, but I don't recall a study on that) of those who have sufficiently different immune systems. Men have too much sperm to waste, so are not choosy enough, therefore if you want to use all women's fertility optimally, you need to get rid of rape and all kinds of coercion, as all male-chosen matings will produce suboptimal offspring - men lose nothing by producing less-than-perfect offspring, women do.

DNA testing is not nearly far enough progressed to be able to predict which to humans will have the fittest offspring. Completely free female choice is your best bet. (Ideally, you'd get rid of all psychological problems that can lead women to reenact childhood trauma rather than choosing the best mate, too.)

If you want things to be a bit wacky, you could have match-making that works by women smelling men's clothes, in an effort to eliminate all other factors. (Could be that facial features and the like also play a role, though, in which case this would be counterproductive.)




Something to consider: Mutations occur more often with older parents. Since mutations presumably increase genetic variation, it might be wise to make sure a lot of women have their first child at over 30. (You want it to be over 20, perhaps even older, to reduce the rate of death in childbirth. You can't afford death in childbirth with your as-small-as-possible-already population size. Early pregnancy is unhealthy or sub-optimal for many mammals, and usually prevented by groups with an intact social structure - that's the real reason why the "alpha wolf" doesn't let other wolves in the pack mate - in a natural structure, they'd be her immature children.)

You also need to reduce child death rate. Big cities with low tech increase child mortality. You want people to live in small villages where diseases don't spread easily.



Humans differ from other mammals in that birth is an enormous risk and usually only produces one child. Some of your genetically unrelated individuums may die in childbirth and not give birth to a living baby, so you lose them from the start.

500 is, I think, the absolute minimum, and I would say, you probably need 1000 in all (500 female) to account for that.


Polygyny doesn't work - every species of animal I know of produces about as many males as females, or, in case of bees, even more males than (fertile) females. Modern beekeepers think they can just kill off the "excess" males, but I think that will come back to bite them later. Species would have evolved to produce fewer males if they didn't need the genetic diversity.

If 50 women in your population of 1000 choose to mate with the same man (because, as explained above, you have made them financially independent from men, so his inability to pay for that many children doesn't matter), that doesn't mean you could just have chosen one man at random for 50 women. The 49 that may not ever get to father any offspring are needed, too - precisely so that there is a choice.

Genetic Decay and Genetic Mutation are not the same.
In fact, it is proven that even men over 40 have sever genetic decay in their sperm cells that having children after 40 is lower quality.
Some similar issues also occur with women, though for a different reason.
For men, they create new cells all the time, but for women, they form a dormant intermediary stage that could convert to ovum.

The later a women has her first child, the earlier her menopause.
Since the number of periods is random, each pregnancy reduces the number of periods by 9...
Yes, each complete pregnancy shifts her menopause by 9 months.

[Kato]

Since the number of periods is random, each pregnancy reduces the number of periods by 9...
Yes, each complete pregnancy shifts her menopause by 9 months.

Uhm... Huh ? Do you have a source on that? Because I never came across that claim and I have no idea how one would go about studying it. (I guess you could average over enough people but considering there are other factors..)

[Razade]

Uhm... Huh ? Do you have a source on that? Because I never came across that claim and I have no idea how one would go about studying it. (I guess you could average over enough people but considering there are other factors..)

A quick google search says that it's not even close to correct. Pregnancy does not factor into the age a woman goes into menopause. Which makes sense. Because menopause has nothing to do with the number of eggs a woman has and everything to do about the hormones involved. Which isn't something you could even factor in based on pregnancies, let alone affected by them.

[cobaltstarfire]

*Also, what would this optimal mating structure be? How would it work?

I would look into the Kinship structures/Skin Names that show up in many of the indigenous societies of Australia. One of the big things these structures controls is who one is allowed to marry, one can't marry people in ones own group and certain others. If I'm remembering right those groups one is allowed to marry into also happen to be the most genetically different, while those in the same group, and the groups that are bared to this person are more closely related.

It's a pretty complicated system, so expect to go a little cross-eyed reading about it.

[halfeye]

Assuming sexual reproduction:

It depends how small the maximum population is.

If the population can expand fast, then the initial group can get away with being small. If the population is limited, then it needs to start at the maximum size and stay there, and then the maximum size needs to be on the order of 100 or so.

[Algeh]

The later a women has her first child, the earlier her menopause.
Since the number of periods is random, each pregnancy reduces the number of periods by 9...
Yes, each complete pregnancy shifts her menopause by 9 months.

Uhm... Huh ? Do you have a source on that? Because I never came across that claim and I have no idea how one would go about studying it. (I guess you could average over enough people but considering there are other factors..)

A quick google search says that it's not even close to correct. Pregnancy does not factor into the age a woman goes into menopause. Which makes sense. Because menopause has nothing to do with the number of eggs a woman has and everything to do about the hormones involved. Which isn't something you could even factor in based on pregnancies, let alone affected by them.

Yeah, that intuitively makes no sense, because many methods of hormonal birth control work by suppressing ovulation, and you don't hear of women who spent their typical reproductive years taking birth control pills not hitting menopause yet in their 60s and and 70s. I assume that would have been all over the news had it happened, and many boomers used HBC for at least a decade or two of their otherwise-fertile lives so we'd be seeing such effects by now if that were a thing.

Back more on topic for this thread, I would imagine tech level would come into play quite a bit. At a minimum, if a starting colony could also travel with a bunch of frozen sperm samples from people not in the colony group, they'd be able to have a generation or two of children with the "extra" genetic diversity from those samples. This depends, of course, on the initial set-up. Did they know they'd be a genetically isolated group in advance and planned everything out, or did they get separated through happenstance?

[Kato]

You know, to make this a bit more obscure, if you actually had one perfect male, i.e. with no genetic problems, you could basically have him and just a lot of females (that is of course ignoring any problems that are not genetic). If he is perfect and the women at least healthy, then the next generation should be healthy as well. In the one after you might have a chance of deficiencies creeping in, depending on how healthy the original females were, but the majority would be fine. And this way you can weed out the unhealthy, if needed. So instead of say 200 pairs to start, you can get along with 101.

Of course if we just talk about health of the population a perfect female works just as well but for obvious reasons your birth rate would be a lot smaller, and if you aim to populate a planet..

[sktarq]

Actually this number may be more complicated.
Some wolves colonized an island in lake superior and have recently died out again. It was thought to be a classic case of inbreeding but more recent thinking is pointing to the idea the was too much variation in settler population (an basically a half dozen later additions) which set up too many inbreeding options.

So "maximum genetic diversity" may not be the useful matrix we have long taken it to be.

[Balyano]

Female choice, in a matrilineal family structure where only women own property, so men's wealth can't skew women's choices.

It's scientifically proven that people like the smell (and probably facial features, but I don't recall a study on that) of those who have sufficiently different immune systems. Men have too much sperm to waste, so are not choosy enough, therefore if you want to use all women's fertility optimally, you need to get rid of rape and all kinds of coercion, as all male-chosen matings will produce suboptimal offspring - men lose nothing by producing less-than-perfect offspring, women do.

DNA testing is not nearly far enough progressed to be able to predict which to humans will have the fittest offspring. Completely free female choice is your best bet. (Ideally, you'd get rid of all psychological problems that can lead women to reenact childhood trauma rather than choosing the best mate, too.)

If you want things to be a bit wacky, you could have match-making that works by women smelling men's clothes, in an effort to eliminate all other factors. (Could be that facial features and the like also play a role, though, in which case this would be counterproductive.)




Something to consider: Mutations occur more often with older parents. Since mutations presumably increase genetic variation, it might be wise to make sure a lot of women have their first child at over 30. (You want it to be over 20, perhaps even older, to reduce the rate of death in childbirth. You can't afford death in childbirth with your as-small-as-possible-already population size. Early pregnancy is unhealthy or sub-optimal for many mammals, and usually prevented by groups with an intact social structure - that's the real reason why the "alpha wolf" doesn't let other wolves in the pack mate - in a natural structure, they'd be her immature children.)

You also need to reduce child death rate. Big cities with low tech increase child mortality. You want people to live in small villages where diseases don't spread easily.



Humans differ from other mammals in that birth is an enormous risk and usually only produces one child. Some of your genetically unrelated individuums may die in childbirth and not give birth to a living baby, so you lose them from the start.

500 is, I think, the absolute minimum, and I would say, you probably need 1000 in all (500 female) to account for that.


Polygyny doesn't work - every species of animal I know of produces about as many males as females, or, in case of bees, even more males than (fertile) females. Modern beekeepers think they can just kill off the "excess" males, but I think that will come back to bite them later. Species would have evolved to produce fewer males if they didn't need the genetic diversity.

If 50 women in your population of 1000 choose to mate with the same man (because, as explained above, you have made them financially independent from men, so his inability to pay for that many children doesn't matter), that doesn't mean you could just have chosen one man at random for 50 women. The 49 that may not ever get to father any offspring are needed, too - precisely so that there is a choice.

First having women start mating after 30 is a terrible idea. Not only does it vastly reduce population growth, it increases the chance of birth defects, stillbirths, and high mutational load. You state that more mutations is good, it is not. Nearly all mutations are either silent or negative. It's usually estimated that about 70% of mutations are harmful, and most of the remaining 30% are silent.

Also if you want to improve the survival of the population as a whole reducing the death rate of children may be a bad idea. High mutational load tends to create multiple co-morbid symptoms. A deleterious mutation has a high chance of causing physical abnormalities AND mental abnormalities AND immune system abnormalities. Those children that died young in the middle ages were more likely to have weaker immune systems and more mutated DNA than the ones in the same environment that endured the same diseases and hardships and survived.

Harsher darwinian selection keeps the population healthier long term.

You also said that men are not choosy enough. This applies some to one night stands but not to pair bonding. The vast majority of males are attracted to physical features that denote health, and therefor healthy genetics ie: low mutational load. If you are in such a situation and a man or woman does not have healthy genetics it is more evolutionary advantageous for their genes to NOT reproduce but instead invest their resources and nurturing instincts into raising a siblings children. Effectively promoting a healthier version of their own genome rather than investing in offspring that will most likely fail to pass on their genes. ie: go extinct.

[shawnhcorey]

Assuming that every member was heterosexual, fertile, had no genetic problems, and was mating completely optimally*, how few people could populate an entire planet without problems?

If there are no genetic problems, then the minimum population is two, one male and one female.

[thorr-kan]

Actually this number may be more complicated.
Some wolves colonized an island in lake superior and have recently died out again. It was thought to be a classic case of inbreeding but more recent thinking is pointing to the idea the was too much variation in settler population (an basically a half dozen later additions) which set up too many inbreeding options.

So "maximum genetic diversity" may not be the useful matrix we have long taken it to be.

Isle Royale National Park, Michigan.

Google results for Isle Royale wolf genetic study
https://www.google.com/search?rlz=1C...l8&ved=&uact=5

[Aaron Underhand]

If there are no genetic problems, then the minimum population is two, one male and one female.

Various studies of interstellar colonization I ve seen suggestions from 40,000 to 160.

Personally I would send an entirely female crew with a very wide diversity of frozen sperm...