The Fathers Who Stayed
Women’s genetic lines spread from village to village; men’s stayed home, and often died out together. Why male lineages collapsed worldwide after farming began, and how scientists model it.
Two small pieces of your DNA come down to you whole. Your mitochondrial DNA came from your mother, who had it from hers, in an unbroken line of women. If you are a man, your Y chromosome came from your father, who had it from his, in an unbroken line of men. Neither is shuffled by recombination along the way, so each is a single thread through a family tree that, a thousand years back, has more slots than there were people alive. An earlier essay here followed the rest of the genome, the part that is cut up and reshuffled every generation. This one follows the two threads.
A story about them has been going around, in articles and videos, and it goes like this. The male lines are far fewer than the female ones. The reason is marriage: for most of the last few thousand years sons stayed in their father’s village and daughters married into other villages. A village was a cluster of brothers, cousins and uncles who all carried the same Y chromosome, while the women’s lines spread across the region. When a village was wiped out, a whole male lineage went with it. And the Y chromosome has a further weakness: it collects harmful mutations, so in small groups it can be pushed out by a healthier one. That, the story goes, is what happened to the Neanderthals, whose Y chromosome was replaced by ours.
Most of this is right, and some of it is better established than the people repeating it seem to know. Part of it is a hypothesis presented as a finding. And the most interesting part, the reason it happened when it did, is barely mentioned. This essay takes the claims one at a time and checks each against the original studies. It then opens up the two computer models that offer the main explanations and runs a simplified version of them here, so that the mechanism can be seen rather than taken on trust.
The geography of a marriage rule
The first claim is the oldest and the best supported. In 1998 Mark Seielstad, Eric Minch and Luca Cavalli-Sforza compared how genetic variation is spread across human populations for different parts of the genome. For the autosomes and for mitochondrial DNA, 80 to 85 per cent of the variation is found within any one population, and only the rest separates populations from each other. For the Y chromosome only 35.5 per cent is found within populations. Male lines are much more tied to place. Their explanation was the one the story gives: “a higher female than male migration rate (via patrilocality, the tendency for a wife to move into her husband’s natal household)”.[1] Later estimates put the women’s migration rate at nearly eight times the men’s.[2]
The cleanest test came three years later, from the hill tribes of northern Thailand. Some of them are patrilocal, with the wife moving to the husband’s family, and some are matrilocal, with the husband moving to the wife’s. Hiroki Oota and colleagues sampled three of each. In the patrilocal tribes mitochondrial diversity was high and Y diversity low; in the matrilocal tribes it was the other way round. The genetic pattern follows the marriage rule, whichever way the rule runs.[3]
There are two limits on this, and the story usually skips both. The first is scale. When Jason Wilder and Michael Hammer’s group compared the two systems with properly matched samples across ten populations worldwide, the difference between continents nearly disappeared. Patrilocality, they wrote, “may be important at the local scale”, but it does not shape the pattern between continents.[2] It is a fact about villages and regions, not about humanity as a whole. The second limit is time. Today about 71 per cent of the societies in Murdock’s Ethnographic Atlas are mainly patrilocal and 11 per cent matrilocal.[4] Hunter-gatherers are different. In a survey of 32 foraging societies, Kim Hill and colleagues found that “either sex may disperse or remain in their natal group”, and Mark Dyble’s team showed that bands in which men and women have an equal say over who lives with whom end up with many unrelated people living together.[5][6] The rule of sons staying home is mostly a rule of farmers and herders.
Ancient DNA lets this be checked in the graves themselves. In the Lech valley south of Augsburg, Corina Knipper and colleagues combined DNA with strontium isotopes, which record where a person grew up, for 84 people buried between about 2500 and 1700 BC. Most of the women had grown up elsewhere and arrived as adults; the men were local. “The striking patterns of patrilocality and female exogamy prevailed over at least 800 y.”[7] A follow-up study of 104 genomes from the same valley found farmsteads built around a high-status core family related through the male line, with incoming wives and unrelated low-status people in the same households, “and the stability of this system over 700 years”.[8] In a Neolithic tomb at Hazleton North in the Cotswolds, built about 5,700 years ago, 27 of the 35 people buried belonged to a single five-generation family. All 15 links from parent to child in that family ran through men. The sons’ wives were buried there, but no adult daughters were.[9]
One in seventeen, and what it means
So far this explains why male lines are local. It does not explain why there are so few of them, and here the story is pointing at something real and recent. In 2015 Monika Karmin, Toomas Kivisild and a large team analysed high-coverage Y chromosomes from 456 men, 299 of them newly sequenced from 110 populations, and used the mutations on them, together with the mitochondrial DNA of the same men, to reconstruct how the effective number of men and of women had changed over time.[10] The method, called a Bayesian skyline plot, works backwards from how closely related the sampled chromosomes are. Lineages that all trace back to a few recent ancestors point to a small effective population at that time; lineages that stay separate deep into the past point to a large one.
For most of human history the two curves moved together, with the female line somewhat higher, as one expects if men have always varied more than women in how many children they have.[11] Then, in the period “around 8–4 kya”, the male curve fell sharply while the female curve kept rising. At the widest point “the female Ne is up to 17-fold higher than the male Ne”.[10] The drop appears in Africa, Europe, the Near East, South and Central Asia, East Asia, the Andes and Siberia, at different times: about 8,300 years ago in the Near East, 5,000 in Europe, 1,400 in Siberia, as Guyon and colleagues later read Karmin’s regional results.[12] Two other large studies found the same thing independently. Chiara Batini’s team found that 64 per cent of 334 European and Middle Eastern men fall into just three lineages that expanded in the last seven thousand years, with no matching event in the women’s line.[13] David Poznik and colleagues, with 1,244 Y chromosomes from the 1000 Genomes Project, dated sudden bursts of single lineages to about five thousand years ago in Africa, between 4,800 and 5,900 years ago in western Europe and 4,000 to 4,500 years ago in South Asia.[14]
Two things about the number seventeen are often misread, and I misread one of them myself when I first answered the question that led to this essay. It is not a headcount. It does not mean seventeen women for every man, or that most men had no children. “Effective population size” is the size of an idealised population that would lose genetic diversity at the observed rate. It is a measure of how many lines get passed on, not of how many people are alive. And seventeen is the peak ratio between the female and the male curves, which also includes the female curve’s own growth. The fall in the male curve itself was smaller: by Guyon’s reading of Karmin’s figures, between 2.4-fold in Siberia and 6.7-fold in Europe, about 3.5-fold on average.[12] That is still very large for a period in which the number of men alive was growing. Many men had sons; their lines simply did not last.
The dates also carry an uncertainty that the headlines leave out. They depend on how fast the Y chromosome mutates. Karmin’s team calibrated the rate on only two ancient genomes, a twelve-thousand-year-old boy from Montana and a four-thousand-year-old man from Greenland, and said so; Batini’s team used a faster rate from Icelandic families and noted that the “mutation rate remains a difficult issue”.[10][13] A different rate moves the dates by a thousand years or more. And skyline plots can produce a false decline when a structured population is sampled unevenly, a weakness that Karmin and later authors both flag.[15] None of this makes the bottleneck go away, since three independent studies see it, but the exact timing should be read as approximate.
After property, before states
The timing is the most interesting part of the story, and also the part that the version going around leaves out. Farming began about 11,000 years ago in the Near East. If farming itself had caused the bottleneck, the male line should have fallen right away. It did not. In every region the bottleneck peaks one to two thousand years after the local start of farming, and it ends, in the regions where it ends, at about the time the first states appear.[16] Something took a while to build up after farming, and something about states brought it to an end.
The obvious explanations do not fit. A change in climate or environment would have hit men and women alike. A founder effect, a few farmers spreading out, would show up at the start of farming, not later, and it is not what the ancient DNA shows. The explanation most people reach for first is that a few powerful men had many wives. That certainly happened, but the numbers do not work: a seventeen-to-one ratio “falls far outside ratios typical in ethnographic data”, and the bottleneck ended while inequality between individuals was still increasing.[16] When Guyon’s team modelled polygyny stronger than that of the Kipsigis of Kenya, one of the most polygynous societies on record, it reduced the male effective size about four times, “not sufficient on its own”.[12]
What did build up in those thousand or two thousand years was wealth that could be inherited: fields, houses, herds, later copper and bronze. Wealth passed from father to son makes the line of fathers the unit that matters, and in many societies it was organised as a clan or lineage, a group of men who traced descent from a common male ancestor, held land together, married out, and brought wives in. Anthropologists call such societies patrilineal. This is the context in which Karmin’s team placed the bottleneck, “a change in social structures that increased male variance in offspring number”, and the two models that follow make that idea precise.[10]
States brought it to an end because they changed what a lineage was. A chiefdom or kingdom draws its soldiers, officials and settlers from many lineages at once, suppresses feuds between them, and moves people around. A man’s chance of passing on his Y chromosome stops depending so much on which clan he was born into. The male curve recovers, and grows again with the population.[16]
War between clans
In 2018 Tian Chen Zeng, Alan Aw and Marcus Feldman at Stanford proposed that the bottleneck was caused by competition between patrilineal clans.[16] Their paper has two parts. The first is a pair of equations of the Lotka–Volterra kind, the equations ecologists use for predators and prey or for species competing for food. Here the two competitors are two groups of men drawing wives from a shared pool of women. The result is that when competition between the groups is stronger than competition within them, “almost always one population of males takes over the other”.
The second part is a simulation, and it is the part people have in mind when they cite the paper. Picture a table with 100 rows and 500 columns. Each row is a group, a tribe or clan; each column is a Y-chromosome lineage; each cell holds the number of men of that lineage in that group, 10,000 to 30,000 men in all. In the patrilineal version every group starts with a single lineage, as it would if it were made of brothers and cousins. In the comparison version every group starts with an equal mix of 100 lineages. Then, each generation:
The model was run for 60 generations, about 1,500 years, the approximate length of the bottleneck, 100 times for each of 18 combinations of settings. In the patrilineal runs, “a large number of haplogroups go extinct fairly early in the simulation, while one or several haplogroups rapidly become overwhelmingly dominant”. In the comparison runs almost none do. The mechanism is in one sentence of the paper: “Casualties from intergroup competition then tend to cluster among related males, and group extinction is effectively the extinction of lineages.” When a clan wins, its one lineage grows and splits into new clans of the same lineage; when it loses, its one lineage goes. The authors call this cultural hitchhiking: the Y chromosome gains or loses nothing by itself, but it rides on the fortunes of the group it is attached to. The total number of men never falls, and yet male diversity collapses.
This is the formal version of the village wiped out in the story, and it is a good model. Its limits, which the authors state, are also important. It contains only men: there are no women in the simulation, no mitochondrial DNA, and therefore no female-to-male ratio to compare with Karmin’s. It reports its results as plots of lineages lost, not as a number for how much the effective size fell. Its mutation is a jump between lineages that, in the authors’ words, “has no counterpart in real life”. And it treats patrilineal organisation as given, without explaining where it came from.
Clans that split
In 2024 Léa Guyon, Evelyne Heyer, Raphaëlle Chaix and colleagues in Paris built a model with both sexes and asked whether the violence was necessary.[12] Their simulation, written in the population-genetics program SLiM, follows real chromosomes: each man carries a Y chromosome and each woman a mitochondrial genome, both accumulating mutations at realistic rates. After a long period of random mating, the population is divided into five villages of 300 people that grow by 1 per cent a generation, as the populations of early farmers did. Ten per cent of women move to another village each generation; men stay. In the patrilineal scenarios each village begins with three descent groups, couples are formed across groups, and children belong to their father’s group.
Three features set the model apart. The first is that groups differ in how well they do, for reasons that have nothing to do with war: land, luck, alliances. Each group has a growth advantage drawn at random, and the chance that a newborn boy belongs to a group is proportional to the number of men in it times the exponential of its advantage. The spread of these advantages was set to match the growth and extinction rates of real patrilineal groups, in studies that do not involve warfare. The second is fission. When a group grows past a threshold it splits in two, and the split can be random or lineal. A random split deals the men out like cards. A lineal split follows the family tree: the descendants of one great-grandfather go one way and those of his brother the other, the way segmentary lineages are described in the ethnographic literature. The third is that each new group inherits its parent’s advantage, with some variation, so a successful group tends to produce successful daughter groups. Violence can be switched on, using Zeng’s rule for who dies, at 15 per cent of men per generation.
The model measures the effective sizes the same two ways that the real data are analysed: from the diversity within villages, and by drawing a hundred present-day samples and running the same skyline-plot software that Karmin used. After 100 generations, with clans and random splits but no differences in success and no violence, the male effective size fell by a factor of 1.3, about as much as with no clans at all. Adding violence brought it to 2.1. Adding inherited differences in success with lineal splits, and no violence, brought it to 21. Adding violence on top of that changed little, to 25.[12] Over the full 100 generations, violence made the decline faster at first but did not make it much deeper. “Violence”, they conclude, “is neither sufficient nor necessary to explain this bottleneck.”
The skyline plots give smaller reductions than the diversity measure, and the paper is candid about that: in the run closest to the real data, a period of bilateral descent followed by patrilineal clans, the female-to-male ratio from the skyline plots reaches between 18 and 20, close to Karmin’s seventeen, while the reduction in male effective size is about three, close to Karmin’s regional values.[12] The authors do not claim to have settled the question. The answer, they write, “may therefore be found in between” their scenarios and Zeng’s, and the model leaves out long migrations such as the spread of herders from the steppe.
A model small enough to read
Both papers publish their code, but the mechanism is easier to see in something smaller. The script for this essay is a cut-down version of Guyon’s model: five villages of 150 men and 150 women, one Y and one mitochondrial marker, three descent groups per village, a 10 per cent female migration rate, 1 per cent growth, random or lineal splits, inherited success and Zeng’s violence rule as options. Before anything starts, the villages spend 1,500 generations mixing freely, so that both markers reach a steady level of diversity. Effective size is read from the diversity within villages, as in the main results of both papers. Each scenario was run 40 times; the method note gives the details.
Figure 1 shows the whole story in one run of the model. While men and women move freely, the two effective sizes sit together. When patrilineal clans appear, the female line keeps rising with the population, and the male line falls, within a few dozen generations, to a low of 37 at generation 70, against 873 for women, a gap of about 25 to one. When the clans give way and men start moving between villages again, the male line climbs back within fifty generations and follows the female line upward. This is the shape Karmin found, and nothing in the model knows about war.
Figure 2 takes the ingredients apart, and the small model agrees with the large one in every comparison that matters. Patrilocality alone, with no clans, makes the male line 1.9 times smaller within villages, simply because men stay where their fathers were. Clans with random splits add nothing to that (1.7). Violence doubles it (3.5). Lineal splits with inherited success make it 13.5 times smaller, and violence on top changes nothing (13.4). The model’s reductions for the strongest scenarios are smaller than Guyon’s, which is to be expected of five small villages and a single marker, but the order is the same, and so is the main finding.
Why does the lineal split matter so much? Because it keeps a group genetically uniform. When a clan splits at random, each half carries a sample of the clan’s Y lineages, and whatever happens to the halves afterwards, the lineages are spread across them. When it splits along the family tree, each half is one branch of the family, and in a few rounds of splitting every group is one male line. Then any difference in how groups fare becomes a difference in how lineages fare. A group that does well grows, splits, and fills the village with copies of one Y chromosome; a group that does badly fades out and takes its lineage with it. Violence is one way for groups to fare differently. Land, luck and alliances are others, and in the model it makes little difference which.
Figure 3 counts lines, which is what the story is really about. It shows something the story leaves out: lines die out all the time, in every scenario. Of 750 female lines present at the start, about 23 still have a carrier after 100 generations; most families end with a generation that has only sons, or no children at all, and that is ordinary chance. Without clans, the male lines die out at the same rate (25 survive). Under clans only 5 do. The difference is not that men die out and women do not. It is that under patrilineal clans men’s lines die out in whole groups.
What the graves can and cannot show
The story’s image of a village wiped out has real examples behind it. At Schöneck-Kilianstädten near Frankfurt, a mass grave from around 5000 BC holds at least 26 people, killed and thrown into a ditch, many with their lower legs deliberately broken. Its excavators argue that “massacres of entire communities were not isolated occurrences but rather were frequent features of the last phases” of Europe’s first farming culture, and they note that young women are missing from the grave, a pattern that “has been suggested for other sites previously” and could mean they were taken.[17] At Talheim, a grave from the same period, women were among the 34 dead, so the pattern is not universal.[18]
In Iberia, which includes Portugal, the DNA shows the turnover directly. Between about 2500 and 2000 BC, newcomers with ancestry from the steppe arrived, and “by ~2000 BCE” they had brought about “the replacement of 40% of Iberia’s ancestry and nearly 100% of its Y-chromosomes”.[19] The men of the earlier population left almost no Y chromosomes behind, while their wives and daughters left plenty of everything else. The authors are careful to add that ancient DNA can show that the mixture was sex-biased but not what caused it. Among later examples, a single Y lineage that arose in Mongolia about a thousand years ago is carried by about 8 per cent of men across a large part of Asia, about 0.5 per cent of all men alive; its discoverers proposed that it descends from Genghis Khan and his close male relatives.[20]
So violence happened, and it left marks in the Y chromosome. What the models dispute is whether it was the engine of the worldwide bottleneck or one of several ways in which some clans did better than others. Zeng’s model needs it. Guyon’s does not, and shows that ordinary differences in success between lineages, passed on when groups split along family lines, are enough. On the evidence now available both are possible, and they are not exclusive. The story is right that a whole male lineage could disappear with one village. It is wrong only if it implies that this is the only way lineages disappeared.
A chromosome that cannot clean itself
The last claim is about mutation, and the biology behind it is standard. Most chromosomes come in pairs that exchange pieces in every generation, which lets natural selection separate a harmful mutation from the good genes around it. The Y chromosome has no partner for most of its length and so cannot do this. A harmful mutation on it is stuck with everything else on that chromosome. In a finite population the copies with the fewest harmful mutations are lost from time to time by chance and cannot be rebuilt, a process called Muller’s ratchet: “the irreversible accumulation of deleterious mutations in a finite non-recombining population”.[21] The smaller the population, the faster the ratchet turns. The same review also deflates a popular worry: the decay of gene-rich Y chromosomes is fast at first but “slows down over time and ultimately comes to a complete halt”. The human Y is not about to disappear.
The Neanderthal part is a real finding with a cautious explanation. In 2016 Fernando Mendez and colleagues sequenced part of the Y chromosome of a Neanderthal man from El Sidrón, in northern Spain, who lived about 49,000 years ago. It was very different from every modern human Y, and it carried changes in several genes, three of them genes whose proteins the mother’s immune system can react to when she carries a male fetus. The authors suggested, as a possibility, that “incompatibilities at one or more of these genes played a role in the reproductive isolation of the two groups”.[22] That would explain why no Neanderthal Y survives in anyone alive today, although Neanderthal DNA survives in everyone outside Africa. Their estimate of when the two Y chromosomes split, 588,000 years ago, has since been revised.
In 2020 Martin Petr, Janet Kelso and colleagues sequenced the Y chromosomes of three late Neanderthals and two Denisovans.[23] The Denisovan Y split from ours about 700,000 years ago, as expected from the rest of their genome. The late Neanderthal Y did not: it split from ours only about 370,000 years ago, which means that late Neanderthals carried a Y chromosome related to modern humans. Some time between about 370,000 and 100,000 years ago, men from an early population related to our own ancestors had children with Neanderthals, and their Y chromosome eventually replaced the original Neanderthal one. The authors’ explanation is the one the story repeats, but in their words it is a possibility: “This replacement is plausible if the low effective population size of Neanderthals resulted in an increased genetic load in Neanderthals relative to modern humans.” Small groups carry more harmful mutations, so an incoming Y with fewer of them would have had an advantage. Nobody has measured the fitness of a Neanderthal Y chromosome.
One fact weakens the explanation that rests on the Y alone. Neanderthal mitochondrial DNA was replaced too, by an African lineage that arrived through the same kind of early contact, between roughly 400,000 and 270,000 years ago, and Cosimo Posth’s team showed that a complete replacement is possible “even with minimal hominin introgression”.[24] Mitochondria have their own non-recombining weakness, so this does not rule out the explanation from harmful mutations. But it shows that both of the Neanderthals’ uniparental threads were replaced, which makes small population size and chance at least as important as anything peculiar to the Y.
What the two threads record
Put together, the evidence turns the story around in one respect. It is usually told as a story about men: their violence, their mutations, their failure to leave descendants. The Y chromosome, read carefully, is a record of how societies organised descent. For tens of thousands of years men’s and women’s lines moved together. In the few thousand years after farming, when inherited land and herds made the line of fathers the unit of property, male lines became tied to place and to the fortunes of groups, and most of them were lost, by war in some places and by the ordinary rise and fall of families in others. When states made lineages matter less, the male line recovered.
The mitochondrial thread records something else: the movement of women between villages, generation after generation, carrying their mothers’ lines across the map. The Y chromosome stayed where its owners were born and shows where power was held. Neither thread is the history of a sex. Each is the trace of a rule about who moves and who inherits, and the gap between them is the measure of how much that rule mattered.
Where this could go
Ancient DNA as a time series. The bottleneck was inferred from present-day men. Ancient genomes now exist in the thousands, so male diversity could be measured directly in each period and region, and the timing checked without assuming a mutation rate.
Iberia, closely. Portugal and Spain have one of the best-sampled ancient-DNA records in the world, with the almost complete Y turnover around 2000 BC in the middle of it. A regional version of the models here, with the arrival of the steppe-related newcomers, would test how much of that turnover needs violence.
The rest of the genome. If clans and fission explain the bottleneck, the X chromosome and the autosomes should show smaller, predictable effects. Karmin’s team noted that current methods are too weak to see them; tree-sequence methods of the kind described in The Ancestors Who Left Nothing may no longer be.
This essay was researched and written with Claude Opus 5.5 (Anthropic): human specification, editorial direction and critical review; machine research, drafting, computation and figure generation. The question came from a popular article Luis had read; each of its claims was checked against the original papers. Zeng et al. (2018), Guyon et al. (2024), Karmin et al. (2015), Batini et al. (2015), Poznik et al. (2016), Fowler et al. (2022), Knipper et al. (2017), Olalde et al. (2019), Mendez et al. (2016), Posth et al. (2017), Hammer et al. (2008) and Bachtrog (2013) were read in their open-access full text; for Seielstad et al., Oota et al., Wilder et al., Hill et al., Dyble et al., Mittnik et al., Petr et al., Zerjal et al. and Meyer et al. the figures and quotations come from the abstracts, and for Heller et al. from a summary. Karmin’s regional dates and reductions are taken from Guyon’s reading of Karmin’s supplementary figure, which could not be downloaded. The Talheim figures come from an encyclopedia entry.
The figures are computed, not traced. The script scripts/ylineages.py is a simplified version of Guyon’s model: five villages of 150 men and 150 women, mixing freely (5% of each sex moving between villages per generation) for 1,500 generations; then, for 100 generations, 10% of women move and men stay, villages grow 1% a generation, and in the patrilineal scenarios each village starts with three descent groups. A boy’s group is drawn with probability proportional to its number of men times er, where r is the group’s growth advantage, drawn from a normal distribution with variance 0.1; a group with more than 75 men that has not split for three generations splits in two, at random or along paternal lines (the most distant paternal ancestor, within eight generations, that divides the group), and each half draws a new advantage around its parent’s. Violence follows Zeng’s rule, at 15% of men per generation. One Y and one mitochondrial marker mutate at 0.005 per generation under the infinite-alleles model, and effective size is read from the mean gene diversity H within villages as H/(2u(1−H)). Each scenario was run 40 times; the numbers are in docs/ylineages-results.json. The model has one locus per sex, no polygyny and no long-distance migration, and its absolute values should not be compared with real populations; it is there to show the mechanism. The cover is drawn from one run of the same model.
Authored by: Luis Matos Ferreira — Physicist, Developer, Writer
- The Ancestors Who Left Nothing — the rest of the genome: why most of your ancestors left no DNA in you.
- The Complete Parts List — what biology can enumerate and what it can predict.
- The Frozen Accident — how much of life is found and how much is the tape happening to run that way.
- M. T. Seielstad, E. Minch & L. L. Cavalli-Sforza, “Genetic evidence for a higher female migration rate in humans”, Nature Genetics 20, 278–280 (1998); figures and quotation from the abstract, doi.org.
- J. A. Wilder, S. B. Kingan, Z. Mobasher, M. M. Pilkington & M. F. Hammer, “Global patterns of human mitochondrial DNA and Y-chromosome structure are not influenced by higher migration rates of females versus males”, Nature Genetics 36, 1122–1125 (2004), doi.org.
- H. Oota, W. Settheetham-Ishida, D. Tiwawech, T. Ishida & M. Stoneking, “Human mtDNA and Y-chromosome variation is correlated with matrilocal versus patrilocal residence”, Nature Genetics 29, 20–21 (2001), doi.org.
- V. Hrnčíř, P. Duda, G. Šaffa, P. Květina & J. Zrzavý, “Identifying post-marital residence patterns in prehistory: a phylogenetic comparative analysis of dwelling size”, PLoS ONE 15, e0229363 (2020), for the shares in G. P. Murdock’s Ethnographic Atlas (1967), doi.org.
- K. R. Hill, R. S. Walker, M. Božičević et al., “Co-residence patterns in hunter-gatherer societies show unique human social structure”, Science 331, 1286–1289 (2011), doi.org.
- M. Dyble, G. D. Salali, N. Chaudhary et al., “Sex equality can explain the unique social structure of hunter-gatherer bands”, Science 348, 796–798 (2015), doi.org.
- C. Knipper, A. Mittnik, K. Massy et al., “Female exogamy and gene pool diversification at the transition from the Final Neolithic to the Early Bronze Age in central Europe”, PNAS 114, 10083–10088 (2017), doi.org.
- A. Mittnik, K. Massy, C. Knipper et al., “Kinship-based social inequality in Bronze Age Europe”, Science 366, 731–734 (2019), doi.org.
- C. Fowler, I. Olalde, V. Cummings et al., “A high-resolution picture of kinship practices in an Early Neolithic tomb”, Nature 601, 584–587 (2022), doi.org.
- M. Karmin, L. Saag, M. Vicente et al., “A recent bottleneck of Y chromosome diversity coincides with a global change in culture”, Genome Research 25, 459–466 (2015); the 17-fold figure is in the Results and Supplemental Fig. S5, doi.org.
- M. F. Hammer, F. L. Mendez, M. P. Cox, A. E. Woerner & J. D. Wall, “Sex-biased evolutionary forces shape genomic patterns of human diversity”, PLoS Genetics 4, e1000202 (2008), doi.org.
- L. Guyon, J. Guez, B. Toupance, E. Heyer & R. Chaix, “Patrilineal segmentary systems provide a peaceful explanation for the post-Neolithic Y-chromosome bottleneck”, Nature Communications 15, 3243 (2024); results from Table 2 and the Discussion; the regional dates and reductions are their reading of Karmin et al., Supplemental Fig. S4B; code at github.com/lea-guyon/PatriSim, doi.org.
- C. Batini, P. Hallast, D. Zadik et al., “Large-scale recent expansion of European patrilineages shown by population resequencing”, Nature Communications 6, 7152 (2015), doi.org.
- G. D. Poznik, Y. Xue, F. L. Mendez et al., “Punctuated bursts in human male demography inferred from 1,244 worldwide Y-chromosome sequences”, Nature Genetics 48, 593–599 (2016), doi.org.
- R. Heller, L. Chikhi & H. R. Siegismund, “The confounding effect of population structure on Bayesian skyline plot inferences of demographic history”, PLoS ONE 8, e62992 (2013), doi.org.
- T. C. Zeng, A. J. Aw & M. W. Feldman, “Cultural hitchhiking and competition between patrilineal kin groups explain the post-Neolithic Y-chromosome bottleneck”, Nature Communications 9, 2077 (2018); model in Methods and Supplementary Notes 1–6; code at github.com/alanaw1/CulturalHitchhiking, doi.org.
- C. Meyer, C. Lohr, D. Gronenborn & K. W. Alt, “The massacre mass grave of Schöneck-Kilianstädten reveals new insights into collective violence in Early Neolithic Central Europe”, PNAS 112, 11217–11222 (2015), doi.org.
- “Talheim Death Pit”, Wikipedia, for the count of the dead by age and sex (the primary report, J. Wahl & H. G. König, Fundberichte aus Baden-Württemberg 12, 1987, was not consulted), en.wikipedia.org.
- I. Olalde, S. Mallick, N. Patterson et al., “The genomic history of the Iberian Peninsula over the past 8000 years”, Science 363, 1230–1234 (2019), doi.org.
- T. Zerjal, Y. Xue, G. Bertorelle et al., “The genetic legacy of the Mongols”, American Journal of Human Genetics 72, 717–721 (2003), doi.org.
- D. Bachtrog, “Y-chromosome evolution: emerging insights into processes of Y-chromosome degeneration”, Nature Reviews Genetics 14, 113–124 (2013), Box 4 and main text, doi.org.
- F. L. Mendez, G. D. Poznik, S. Castellano & C. D. Bustamante, “The divergence of Neandertal and modern human Y chromosomes”, American Journal of Human Genetics 98, 728–734 (2016), doi.org.
- M. Petr, M. Hajdinjak, Q. Fu et al., “The evolutionary history of Neanderthal and Denisovan Y chromosomes”, Science 369, 1653–1656 (2020); the 370,000–100,000-year window is from the preprint, bioRxiv 10.1101/2020.03.09.983445, doi.org.
- C. Posth, C. Wißing, K. Kitagawa et al., “Deeply divergent archaic mitochondrial genome provides lower time boundary for African gene flow into Neanderthals”, Nature Communications 8, 16046 (2017), doi.org.
Comentários
Enviar um comentário