A Historical-Genetic Reconstruction of Human Extra-Pair Paternity
Download started
Ok
Download started
Ok
Keywords
Results and Discussion
In behavioral ecology, the occurrence and adaptive significance of extra-pair copulations (EPCs) in species with long-term pair-bonds has been the subject of intense research and debate [2, 3, 6, 7, 12–17]. Indeed, theory has shown that both partners can enhance their fitness by seeking reproductive opportunities elsewhere. EPCs provide males with the opportunity to sire extra offspring without having to pay the cost of extended paternal care [2, 14]. Females may also benefit from pursuing EPCs—for example, if this allows them to mate with males of superior fertility or genetic quality, or if it enables them to obtain additional resources from the extra-pair partner [9, 10, 15, 16]. At the same time, EPC behavior can be harmful, because it increases the risk of attracting sexually transmitted diseases, spousal aggression, divorce, and reduced paternal investment by the social partner [3]. Additionally, EPP can result from forced copulations by an extra-pair male with serious costs for the female, the occurrence of which can in turn incentivize the male partner to exhibit anti-cuckoldry tactics motivated by sexual jealousy, such as mate guarding [3, 5, 14]. Features of the social environment, such as population density and resource availability, can strongly affect to what extent individuals have opportunities to engage in EPCs and can modulate the evolutionary costs and benefits both for seeking and preventing EPCs [5, 14, 15, 18, 19]. Thus, in both the animal kingdom and in human societies, it has been suggested that strategies for pursuing and preventing EPCs by both sexes should be highly context dependent [5, 9, 10, 13–15, 18–21].
12.
Akçay, E. ∙ Roughgarden, J.
Extra-pair paternity in birds: review of the genetic benefits
Evol. Ecol. Res. 2007; 9:855-86813.
Eliassen, S. ∙ Kokko, H.
Current analyses do not resolve whether extra-pair paternity is male or female driven
Behav. Ecol. Sociobiol. 2008; 62:1795-180414.
Westneat, D.F. ∙ Stewart, I.R.K.
Extra-pair paternity in birds: causes, correlates, and conflict
Annu. Rev. Ecol. Evol. Syst. 2003; 34:365-39615.
Brouwer, L. ∙ van de Pol, M. ∙ Aranzamendi, N.H. ...
Multiple hypotheses explain variation in extra-pair paternity at different levels in a single bird family
Mol. Ecol. 2017; 26:6717-672916.
Greiling, H. ∙ Buss, D.M.
Women’s sexual strategies: the hidden dimension of extra-pair mating
Pers. Individ. Dif. 2000; 28:929-96317.
Anderson, K.G.
How well does paternity confidence match actual paternity? Evidence from worldwide nonpaternity rates
Curr. Anthropol. 2006; 47:513-52013.
Eliassen, S. ∙ Kokko, H.
Current analyses do not resolve whether extra-pair paternity is male or female driven
Behav. Ecol. Sociobiol. 2008; 62:1795-180414.
Westneat, D.F. ∙ Stewart, I.R.K.
Extra-pair paternity in birds: causes, correlates, and conflict
Annu. Rev. Ecol. Evol. Syst. 2003; 34:365-39615.
Brouwer, L. ∙ van de Pol, M. ∙ Aranzamendi, N.H. ...
Multiple hypotheses explain variation in extra-pair paternity at different levels in a single bird family
Mol. Ecol. 2017; 26:6717-672918.
Maldonado-Chaparro, A.A. ∙ Montiglio, P.O. ∙ Forstmeier, W. ...
Linking the fine-scale social environment to mating decisions: a future direction for the study of extra-pair paternity
Biol. Rev. Camb. Philos. Soc. 2018; 93:1558-157719.
Shellman-Reeve, J.S. ∙ Reeve, H.K.
Extra-pair paternity as the result of reproductive transactions between paired mates
Proc. Biol. Sci. 2000; 267:2543-254620.
Bellis, M.A. ∙ Hughes, K. ∙ Hughes, S. ...
Measuring paternal discrepancy and its public health consequences
J. Epidemiol. Community Health. 2005; 59:749-75421.
Geary, D.C.
Evolution of fatherhood
Salmon, C.A. ∙ Shackelford, T.K.
Family Relationships: An Evolutionary Perspective
Oxford Scholarschip Online, 2007
Here, we present the first large-scale study of how social context affects the incidence of EPP in humans, using a genetic genealogical approach that combines information from in-depth legal genealogies with Y chromosome profiles of patrilineally distantly related males (cf. [6, 8]). Our study covers a time period of several centuries during which there were dramatic changes in the human social environment, including the rapid urbanization that accompanied the Industrial Revolution in 19th-century Western Europe. Against this backdrop, we test the hypothesis that the emerging high-density urban centers fostered elevated rates of extra-pair paternity caused by increased opportunities for EPCs and reduced levels of social control afforded by the anonymity of the city [22–24]. In addition, we test whether EPP incidence is elevated in the lower socioeconomic classes, because of higher incentives to seek additional benefits from extra-pair mates, reduced protection against forced copulation, and/or lower incentives to prevent partner EPC [10].
22.
Laslett, P. ∙ Oosterveen, K. ∙ Smith, R.M.
Bastardy and Its Comparative History: Studies in the History of Illegitimacy and Marital Nonconformism in Britain, France, Germany, Sweden, North America, Jamaica, and Japan
Harvard University Press, 198023.
Griffin, E.
Sex, illegitimacy and social change in industrializing Britain
Soc. Hist. 2013; 38:139-16124.
Matthys, C.
Discourses versus life courses: servants’ extramarital sexual activities in Flanders during the nineteenth and early twentieth centuries
J. Urban Hist. 2016; 42:81-100To estimate historical EPP rates among married couples in a Western European population, we identified 513 pairs of contemporary adult males living in Belgium and the Netherlands who, based on genealogical evidence, shared a common paternal ancestor and hence should have the same Y chromosome in the absence of any EPP events (115 of these pairs were also included in previous studies that analyzed the averaged historical EPP rate in Belgium [8] and the Netherlands [25]). Most of the male ancestors within the 513 genealogical pairs were born before the introduction of modern contraception (median year of birth: 1840, quartiles: 1762–1896, range 1315–1974), implying that our estimated EPP rates more reliably reflect actual EPC rates than studies focusing on contemporary populations [6, 26]. Subsequently, we genotyped all of our contemporary male DNA donors at a panel of 191 Y chromosome SNP (Y-SNPs) and 38 Y chromosome STR (Y-STR) loci and used mismatches in the Y chromosomal haplotype of genealogically paternally related pairs of men as evidence for the occurrence of one or more EPP events within their genealogy [8] (see STAR Methods for details). The high quality of both the historical demographic data and the genealogical records available for our study area meant that we could not only reconstruct the mean historical EPP rate from these mismatches but also estimate EPP as a function of socio-demographic factors that are expected to influence its incidence [9, 10, 20, 25, 27]. Specifically, we examined archival records to obtain in-depth genealogical details of all 6,818 male ancestors that occurred in our genealogies, including place and date of birth and, for the males born between 1750 and 1950, the occupation of the legal father (Figure 1). We always verified in the genealogical records that the birth of each child in the patrilineages happened within wedlock and was declared officially by the father himself (for children born after the start of the civil records; after 1800) or that there was clear genealogical evidence that the father was alive when the child was baptized (for children born between 1600 and 1800, when only church records were available). We then used the occupation of the father to infer the socioeconomic status of the ancestors and linked place and date of birth to recorded (or estimated) historical population sizes and densities. Finally, we used a newly developed empirical Bayes-adjusted logistic regression model to estimate the probability of EPP as a function of year of birth, population density, and socioeconomic status (cf. [8]; Figure 1).
Figure 1 Historical-Genetic Reconstruction of Human Extra-Pair Paternity
Our results confirm earlier estimates of low average historical EPP rates in the Low Countries [6, 25] (1.6% of children per generation; 95% confidence intervals 1.2%–2.1%) and also confirm that there is no significant difference in EPP rate between Flanders and the Netherlands, despite there being key differences in religious affiliation, with Flanders being predominantly Catholic and the Netherlands being mostly Protestant [25]. More importantly, however, our results show that estimated EPP rates exhibited large and statistically significant variation as a function of population density and socioeconomic status (Figures 2A–2C). For an average population density, the mean EPP rate was low among farmers (1.1%) and the middle and high socioeconomic classes (1.0%, e.g., skilled craftsmen and merchants) but much higher (4.1%, p < 0.001) for the low socioeconomic classes (e.g., laborers, weavers) (Figure 2B). Likewise, the EPP rate showed a significant increase as a function of population density (p = 0.03), with the average EPP rate increasing from around 0.6% for small sparsely populated rural villages to 2.3% for cities where densities reached 10,000 inhabitants per km2 or more (Figure 2C). Combining the effect of both variables, the estimated EPP rates for the families in our dataset varied by more than one order of magnitude, from 0.4% and 0.5% among the medium to high classes and farmers living in the most sparsely populated towns to 5.9% for the low socioeconomic classes living in the most densely populated cities (Figures 2A and 3). Adding year of birth, father’s age, mother’s age, country, or region (province) to the model did not improve its fit (Table S1). Despite the absence of an effect of year of birth on the EPP rate, we do observe a temporal pattern peaking in the late 19th century (Figure 2A; Video S1), which is fully consistent with the temporal changes in population density and the initial expansion of the low-wage-earning industrial proletariat that occurred over the course of the Industrial Revolution. Interestingly, these observations on EPP rates also closely mirror patterns in illegitimacy, i.e., of children being born out of wedlock, in this same period [22, 23]. Indeed, illegitimacy peaked at approximately 5% in the countryside and at 12% in Belgian and Dutch cities (>1,000 habitants/km2) during the mid-19th century and also showed elevated rates among the lower socioeconomic classes, rising, e.g., to 36% among servants and day laborers working in Brussels [24].
Figure 2 Context-Specific Variation in Human Extra-Pair Paternity
Figure 3 Estimated Extra-Pair Paternity Rates in the Low Countries around 1850
Video S1. Predictions of Extra-Pair Paternity Rates through Time in the Low Countries, Related to Figure 3
An animated version with predictions of the extra-pair paternity rates shown through time based on historical population density estimates in the Low Countries.
Overall, our results show that human EPP rates in these Western populations are strongly influenced by social context. This finding is in line with theoretical predictions [9, 10, 13, 15, 18] that emphasize context-specific variation in the incentives and opportunities for either seeking or preventing extra-pair mating. Our finding of a positive association between EPP rates and population density is interesting, as this correlation also received some support in the animal kingdom, where it was usually interpreted as being the result of elevated encounter rates and increased opportunities for EPCs [15]. In humans, this effect is also likely compounded by reduced social control due to the anonymity of densely populated cities, as this factor is also thought to be one of the key reasons for the high rates of illegitimacy observed in mid-19th-century European cities [22, 24]. Our finding of elevated EPP in the lower socio-economic classes matches earlier results based on blood-group typing that indicated increased EPP rates in the lower socioeconomic classes of contemporary Mexico and the United States [10, 20, 27]. This provides support for the hypothesis that adverse ecological or economic circumstances can increase the incentive for obtaining additional material or social benefits from extra-pair mates, since these benefits have a stronger impact on the fitness of the women involved (and potentially even indirectly of their social mate and other family members) when resources are scarce [10]. In fact, resource limitation is also one of the main ecological factors that are thought to drive polyandry in some traditional small-scale societies [28]. Another (non-mutually exclusive) explanation for elevated EPP rates in the lower socio-economic classes is that social fathers have less of an incentive to prevent EPP, because they do not have much wealth to be inherited by their offspring [10]. In sociology, the similarly high rates of illegitimacy observed among the lower classes in mid-19th-century Western Europe and increased sexual risk-taking observed in this segment of the population has been explained as a reflection of a greater desire for sexual emancipation or upward social mobility [22–24, 29], although other scholars have also pointed out the greater vulnerability to male sexual violence and exploitation caused by poor working and living conditions [24, 29, 30]. This dual explanation also applies to the EPP patterns we observe, but, without knowing the identity and the social class of the biological father in the case of an EPP event, it cannot be tested here. Indeed, the causes for the observed historical associations between EPP and demography will always remain obscure to some extent, since we cannot know in which cases the husbands were aware that their offspring was biologically not theirs, nor can we reconstruct the circumstances and intentions of the wives involved in EPCs.
22.
Laslett, P. ∙ Oosterveen, K. ∙ Smith, R.M.
Bastardy and Its Comparative History: Studies in the History of Illegitimacy and Marital Nonconformism in Britain, France, Germany, Sweden, North America, Jamaica, and Japan
Harvard University Press, 198023.
Griffin, E.
Sex, illegitimacy and social change in industrializing Britain
Soc. Hist. 2013; 38:139-16124.
Matthys, C.
Discourses versus life courses: servants’ extramarital sexual activities in Flanders during the nineteenth and early twentieth centuries
J. Urban Hist. 2016; 42:81-100Previously, relatively high rates of extra-pair paternity (>5%) were thought to occur in only a handful of traditional South American and African societies where an informal form of polyandry is socially accepted, and multiple male partners may contribute resources to the same woman or her offspring [9, 10, 28]. For humans, this high end of the spectrum has typically been contrasted with the low EPP rates of 1%–2% that have been reported for most other traditional and Western populations, where polyandry is not socially accepted [6, 31, 32]. Although we confirm that EPP rates in our study population are low on average, we have also shown that these rates are by no means invariable and can reach relatively high levels in some parts of society. Indeed, by zooming in on specific social strata, our study uniquely shows that there is much variation to be uncovered in the degree of extra-pair paternity within human societies.
STAR★Methods
Key Resources Table
| REAGENT or RESOURCE | SOURCE | IDENTIFIER |
|---|---|---|
| Deposited Data | ||
| Y chromosome profiles (Y-STR and Y-SNP) | https://yhrd.org/ | YA003651, YA003652, YA003653, YA003739, YA003740, YA003741, YA003742, YA004300 and YA004301 |
| Software and Algorithms | ||
| R | https://www.r-project.org/ | N/A |
Lead Contact and Materials Availability
Further information and requests for resources and reagents should be directed to and will be fulfilled by the Lead Contact, Maarten H.D. Larmuseau (maarten.larmuseau@kuleuven.be).
Experimental Model and Subject Details
Samples for this study were collected as part of two distinct campaigns:
i)
An on-going global genetic genealogical project in the Benelux was started in 2009 by the KU Leuven (University of Leuven, Belgium) and the non-profit Familiekunde Vlaanderen. Within this citizen science project, participants were selected if they could provide their DNA sample under an informed consent agreement and if their oldest reported paternal ancestor (ORPA) within their patrilineage was born before 1800 in our study area. Participants were not informed if other relatives also had participated but they were made aware that an extra-pair paternity event could be detected within their patrilineage if another relative should also happen to participate. All participants were given a certificate with their personal Y chromosome profile and an anonymous code. With this code they could then later log in to the project website and if a Y chromosome match was found they had the opportunity to contact their match for genealogical purposes. As of November 2018, 1638 samples were analyzed as part of this first project.
ii)
An additional sampling campaign, called ‘De Gen-iale Stamboom’, was realized to specifically target DNA-donors who were either patrilineal related or related with a person who had already been sampled in the global genetic genealogical project. This campaign was specifically announced to calculate Y-STR mutation rates using deep-rooting pedigrees (this has been analyzed in another paper [33]) and to verify the genetic relatedness among legal relatives (this paper). For ethical reasons, donors were always separated by at least seven meioses from each other in direct paternal line. The donors were automatically provided with their own Y chromosome profile but they had the opportunity to refuse this information if they desired. This extra sampling campaign was announced in genealogical and local heritage associations, in newspapers and on social media, national television and radio. In total, data from 717 DNA-donors were analyzed as part of this second campaign.
All DNA-donors were adult males (> 18 year).
The study has been approved by the ‘Medical Ethical Committee’, the official institutional review board (IRB) of the KU Leuven – University of Leuven (procedure numbers S55864; S59085; S54010). All the samples were collected with written consent from the donors who gave their permission for Y chromosome analysis, storage of the samples, genealogical research, and scientific publication of their anonymized DNA results.
Method Details
Genealogical analysis
The region under study was delineated by the following current administrative borders:
•
The northern part of Belgium: the provinces of West Flanders, East Flanders, Antwerp, Limburg, Flemish and Walloon Brabant, and Brussels.
•
The southern part of the Netherlands: the provinces of Zeeland, North Brabant, South Holland, North Holland, Utrecht and Gelderland.
In the rest of our analysis, we will always refer to a country and province according to the current administrative borders, even for historical time periods where these borders did not yet exist.
The legal in-depth patrilineages of all male donors were collected with the help of many volunteers within a citizen science movement [34]. The patrilineages were afterward checked individually with copies of the original parish records, the civil registry and other official documents of the State Archives in Belgium (http://www.arch.be) and the Wiewaswie (“who was who”) database for the Netherlands (http://www.wiewaswie.nl). To guarantee the correctness of the legal genealogies, each step in the patrilineage was required to have at least two different records to prove the legal relatedness between two men, e.g., birth certificate and marriage certificate, or baptism record and death certificate [35].
Genealogical resources
In Belgium, only civil records older than 100 years are accessible due to the privacy legislation (http://www.privacycommission.be). In the Netherlands, the situation is more complex, with the Civil Code of the Netherlands (Burgerlijk Wetboek) stating that birth certificates can be consulted after 100 years, marriage certificates after 75 years and death certificates after 50 years. For our scientific study we obtained permission from the CBG (Centraal Bureau voor Familiegeschiedenis, Den Haag) to receive genealogical data of persons who died more than two years ago [36].
All patrilineages were verified based on the civil registry for data going back to 1796 in Belgium and going back to 1811 in the Netherlands (with exceptions for large parts of the current provinces of Limburg, North Brabant and Zeeland for which there are certificates since the 1790s). The civil registry includes the registers of birth certificates, marriages and divorces certificates, and death certificates of all citizens. The consulted birth certificates also mention the occupation of the fathers, and these were used to infer socio-economic class (see below).
Before the civil registry, the parish records of baptisms, marriages and funerals are accessible since the 17th century for most parishes in Belgium and the Netherlands, and this for all recognized religions in the Low Countries [36]. The baptism date is not identical to the birth date but due to religious and legal constraints during the Ancien Régime in the Low Countries, newborns were baptised within a few days so that the baptism date is a good proxy for the date of birth between the years 1600 and 1800 [36–38].
Before the 17th century, only secondary documents are available to reconstruct the patrilineages, namely notarial acts, judicial instruments, tax registers, burgher lists, lease contracts, etc. For Belgium and the Netherlands, the consulted secondary documents were Staten van Goed, Poorterslijsten, Wezerijakten, Penningkohieren, Gichten, Ommelopers, etc.
Selection of genealogical pairs
Following our quality check, all in-depth patrilineages of the DNA donors were entered in the genealogical software program ALDFAER v. 4.2 (Stichting Aldfaer, 2013; http://www.aldfaer.net), after which DNA donors with a common known paternal ancestor were detected based on the legal genealogies. Related patrilineages were selected for further analysis if they met following criteria:
•
First, all the donors were at least seven meioses separated from each other in direct paternal line due to ethical issues.
•
Second, there was at most one child in the patrilineage of the donors who was born outside the study area (defined above), e.g., in a former colony of the Netherlands or Belgium, or in exile during World War I. The rationale behind this criterion is that we mainly wanted to focus on ancestors living in the study area, but that at the same time we still allowed for one single isolated ancestor to be born outside our study area per genealogical pair, as that usually implied that this individual resided only for a short while there.
•
Third, there were no illegitimate or pre-marital children in the patrilineages of the donors, whether or not they were legitimised later by the marriage of their mother. This could be determined and guaranteed by checking all the original birth and baptism registers of the men in the patrilineages (see section ‘genealogical resources’). If the child was legitimised after the wedding of the mother, this legitimisation in the civil records (> 1800) would be written retroactively on the side next to the original birth record. This was a strong criterion, and we have strived not to include any patrilineages which included illegitimate children during the sampling process.
•
Fourth, there were no reported paternity denials in the patrilineages of the donors and there was genealogical evidence for each child in the patrilineage that the father was still alive when the child was born, and that he approved his paternity. For each child born since the start of the civil register (> 1800) we made sure that the father himself had officially declared the birth of his child on the original birth record (i.e., that the child was not declared by a new partner or another family member of the mother). The father always signed the certificate, unless he was illiterate but in that case this was clearly mentioned in the text. If the death certificate of the father was available (this was the case in approximately 95% of the cases – this is the most difficult civil record to find, especially if the person did not die in the same community where he married and/or lived), we also did an extra check to make sure the father did not die before the child was born. For each child born in the period since the start of the church records (> 1600) but before the start of the civil register (< 1800), we have checked each original baptism certificate to make sure that the names of both parents were written down by the priest without any specific note that the father died or left the family. We also double-checked all original wedding and funeral records in the parish records to be sure that all the children were born while their father was alive.
Only genealogical pairs of two related donors were selected for further analysis. Therefore, patrilineages with more than two individuals were reduced to genealogical pairs according to the following rules:
i)
For patrilineages with three donors: one genealogical pair was obtained for analysis with those two donors with the highest number of meioses between each other (Figure S1A). When different combinations are possible, the one with the highest difference in age between the DNA-donors was selected (Figure S1B).
ii)
For patrilineages with more than three donors, if they had an even number of donors: the highest possible number of genealogical pairs was obtained without any overlap of ancestors between the genealogical pairs (Figure S1C).
iii)
For patrilineages with more than three donors, if they had an odd number of donors: the highest possible number of genealogical pairs was obtained without any overlap of ancestors between the genealogical pairs, in addition to the standard rules applied to patrilineages with three donors.
In total, 513 genealogical pairs were selected. One part, namely 119 pairs, were derived from the global genetic genealogical project for which the donors were not sure whether or not another relative would be involved in the analysis. The other 394 pairs were derived from the sampling campaign in which donors were explicitly aware of the fact that there were other, distantly related volunteers in the analysis.
Y chromosome analysis
A buccal swab sample from each selected participant was collected for DNA extraction by using the Maxwell 16 System followed by real-time PCR quantification (Quantifiler Human DNA kit, Applied Biosystems, Foster City, CA, USA) or by using the SwabSolution™ kit (both Promega, Madison, WI, USA) without quantification.
In total 38 Y-STR loci were analyzed in the Laboratory of Forensic Genetics and Molecular Archaeology. These Y-STR multiplexes include DYS19, DYS385a/b, DYS388, DYS389-I, DYS389-II, DYS390, DYS391, DYS392, DYS393, DYS426, DYS437, DYS438, DYS439, DYS442, DYS447, DYS448, DYS449, DYS454, DYS455, DYS456, DYS458, DYS459a/b, DYS460, DYS464a/b/c/d, DYS570, DYS576, DYS607, DYS635, DYS724a/b, GATA H4.1 and YCAIIa/b [8, 39]. PCR products were purified with the BigDye XTerminator® Purification Kit (Applied Biosystems) and plated out in Hi-Di™ Formamide on a 96 well-plate for analysis on an ABI PRISM 3130 XL Genetic Analyzer with POP7 and a 50 cm capillary (Applied Biosystems) using GeneScan 500 LIZ Dye Size Standard (Applied Biosystems) as a size standard. Fragment length of the 38 Y-STRs was analyzed using GeneMapper® v3.2.1 (Applied Biosystems). For samples in which some loci did not amplify, this process was repeated with alternative primer sets, thereby eliminating the possibility of technical errors or the occurrence of mutations in the standard primer binding sites.
All haplotypes were submitted to Whit Athey’s Haplogroup Predictor [40] to obtain probabilities for the inferred haplogroups in Northwest Europe. Based on these results, the samples were assigned to specific Y-SNP assays to confirm the haplogroup and to assign the subhaplogroup according to the Y chromosome tree given in Data S1A. This phylogeny is based on Karafet et al. [41], extended by including the Y-SNPs within haplogroup G described by Sims et al. [42] and by adding extra Y-SNPs within haplogroup R-M269 as earlier proposed in Larmuseau et al. [43]. When the posterior probability for any haplogroup did not reach 95% for a specific DNA sample, two general Y-SNP multiplex kits containing all phylogenetically important Y-SNPs developed by van Oven et al. [44] were used to find the main haplogroup before selecting the specific Y-SNP assay to assign the subhaplogroup. A total of 20 multiplex systems with all 191 Y-SNPs were used in this study with SNaPshot® Multiplex System for SNP genotyping (Applied Biosystems) [33, 44, 45]. The nomenclature of the subhaplogroups is based on the terminal mutation that defines them, according to the nomenclature rules of van Oven et al. [46].
Total haplogroup frequencies
The general population sample contains 236 Flemish and 63 Dutch donors. The frequencies of the main haplogroups in this sample fit with those expected for Western Europe [47]. The most frequent haplogroups are R1b and I1 (Data S1C). Neither of our general population samples had a haplogroup frequency that was different from independent whole genome sequencing samples of Flanders (WGS) [48] and the Netherlands (GoNL) [49], nor were the haplotype frequencies in the Flemish and Dutch samples different from each other (multinomial log-linear model [R function ‘multinom’ from package ‘nnet’]: Chisq. = 26.76, df. = 33, p = 0.77).
However, although these differences were not significant, the Flemish samples had a higher frequency for R1b and a lower frequency for I1 than the Dutch, as had been previously reported [48, 50, 51]. No significant differences in FST values between both samples in this study and the independent whole genome samples of Flanders and the Netherlands were found, demonstrating that our sampling strategy did not introduce any bias in the data.
The frequencies of the subhaplogroups within and between Flanders and the Netherlands are highly similar to previous observations in the Benelux [43] (Data S1D). Moreover, the observed frequency differences of subhaplogroups R-P312 and R-M529 are part of the larger gradients present in Western Europe [52].
Y chromosome comparison
Subsequently, the historical EPP rate was estimated by using the genealogical pair method, in which the EPP rate was estimated by testing whether for each pair, the genealogical most recent common (paternal) ancestor (GMRCA) was or was not also the biological most recent common ancestor (BMRCA), based on whether their Y chromosomes matched [8].
In order to infer whether the Y chromosomes matched within a genealogical pair, we first compared the evolutionary lineages (subhaplogroups) between the two individuals at the finest resolution of the used Y chromosome phylogenetic tree. Given that human Y-SNPs have a low mutation rate, ca 0.75-0.89 x10−9 [53], they can be treated as unique evolutionary polymorphisms [54]. Male individuals who share the same Y-SNP haplogroup also must share a common male lineal ancestor since the point of the first appearance of the SNP. All Y-SNPs analyzed in this study are known to be polymorphic in the population (i.e., we did not use private or family-specific SNPs as they are also included in the minimal Y chromosome tree on http://www.phylotree.com/Y [46]), non-recurrent and not located in the gene conversion hotspots on the Y chromosome [55]. Therefore, donors with a different Y-SNP profile and thus assigned to a different subhaplogroup were considered not to be related on a genealogical and historical timescale. Within the 513 considered genealogical pairs, the donors within 93 pairs showed discordant subhaplogroups by Y-SNP genotyping (Table S2).
Next, for the donors assigned to the same subhaplogroup, we compared the 38 Y-STR haplotypes of both individuals with each other. Based on the number of observed different alleles within a genealogical pair, the time of most recent common ancestor was calculated using the formulae by Walsh [56] and the mean mutation rate for the 38 analyzed Y-STRs. This mean mutation rate (5.91 × 10−3 mutations per generation) was calculated using the individual mutation rates measured by Ballantyne et al. [57]. Subsequently, for each genealogical pair we verified whether the number of separated generations based on the legal genealogy fits within the calculated range (95% credibility) based on the Y-STR haplotypes. For the remaining 420 out of 513 genealogical pairs for which both individuals belonged to the same subhaplogroup, the 38 Y-STR haplotypes of only three pairs were significantly too divergent between the donors in relation to their legal genealogical relatedness. These three pairs had 14, 15 and 19 divergent alleles meaning that the biological ancestors of those pairs should live, respectively, 18-42 generations ago, 19-43 generations ago and 24-51 generations ago (95% credibility interval). Although, they should be separated, respectively, by only 9, 15 and 7 meioses based on their legal genealogy. All these individuals belonged to subhaplogroup R-L48, the most frequent lineage in Flanders and the Netherlands with a frequency of ∼10% according to our used Y-SNP set [43]. Therefore, a type I error by declaring a mismatch wrongly is most likely ruled out in our study.
The type II error rate in our study (i.e, declaring no mismatch wrongly within a genealogical pair) was estimated by calculating haplotype differences between independent haplotypes within an unbiased sample of the general population. This so-called ‘general population sample’ is a set of Y chromosome profiles of donors who do not have any possible genealogical connection with each other and for which no recent EPP event that could connect them was expected in their patrilineage. Therefore, when samples of other related men in direct paternal line or men with the same surname (or spelling variants) were present in the dataset, only one individual for each patrilineage or surname was selected at random to avoid any familial bias in the analysis. This ‘general population sample’ contained the Y chromosome profiles of 299 donors. Within this sample, pairs of haplotypes with the same haplogroup had six or more different alleles out of 38 Y-STRs (Table S3). Thus, we may assume that in our sample of 513 genealogical pairs only those with the same subhaplogroup and with six or seven different alleles (Table S2) might be declared to have no mismatch wrongly. Pairs of 38 Y-STR haplotypes with six or seven different alleles between donors belonging to the same subhaplogroup were rare in the ‘general population sample’ after all, namely six and ten out of 44,551 combinations respectively (Table S3). These donors always belonged to subhaplogroups within the R1b (R-M269) radiation for which high Y-STR haplotype resemblance is already observed in Western Europe due to homoplasy [58]. With the R command line pbinom (0:10, size = 513, prob = 16/44,551) we calculated that the chance that we have at most two type II errors in our study is 99.9%. This analysis proved that also type II errors, which are conservative mistakes in our analysis on differences in EPP rates, are limited in our specific study sample and would not affect the final results of our global analysis.
Mismatch verification
For each genealogical pair with a deduced Y chromosome mismatch (N = 93), other reasons than EPP were excluded by:
•
Comparing the discordant Y chromosome profiles with all others in the study to investigate the possibility that samples might have been switched during sampling or genotyping.
•
Redoing the sampling and genotyping of these genealogical pairs to assure that samples were not switched during sampling or genotyping of the Y-SNPs and Y-STRs, or that the DNA quality of the first sample was not appropriate.
•
Regenerating the genealogies of the donors by three external genealogists to assure the legal relatedness between the DNA donors of divergent genealogical pairs. These three independent genealogists, which were not aware of the study aim, received only the recent (< 100 year) genealogical data of the donors (required due to the privacy laws). Genealogical mistakes were excluded based on these independent efforts.
Only when errors in sampling, genotyping and genealogical data were excluded, a mismatch was assumed to be the result of at least one EPP event in the genealogy between two genealogical relatives with a discordant Y chromosome profile.
Measurement of socio-demographic variables
Auxiliary socio-demographic factors that were measured and which could be related to the incidence of extra-pair paternity were year of birth, the father’s age at birth, the mother’s age at birth, the difference in age between the two, socio-economic status of the father at birth and the historical population density at the place of birth.
The socio-economic class of the paternal ancestors was determined on the basis of information about the occupation of the fathers in the consulted birth certificates (see earlier in section ‘genealogical analysis’). These historical occupational titles are relatively easy to classify in terms of social class by an abridged version of a social class scheme proposed by van Leeuwen and Maas [59], known as HISCLASS. This is a common, cross-national, language-sensitive coding scheme that can accommodate historical occupational titles of the kind found in historical documents from state censuses to parish records [60]. We employ the HISCLASS codes of the open access dataset provided by Mandemakers et al. [61] to categorise each occupation in the following straightforward categories: a) medium or high income, b) low income and c) farmers. Medium and high income occupations were combined, as there was only a small number of people with high incomes in our study population (2% of our recorded paternal ancestors in 19th century). In some areas within the region of interest, the occupation of the father was not mentioned in the birth certificate. For the fathers of those areas the socio-economic class was categorised as ‘unknown’ (18% of all recorded paternal ancestors in 19th century).
The population density at place of birth (or place of baptism for the data from before 1800) during the year of birth (or baptism) was determined for all male members in the patrilineages of the DNA donors. This was realized by estimating the number of inhabitants relative to the surface area of the community at that specific year. The surface areas of each community in the year of birth (or baptism) of these males were obtained from the land registers or so-called cadastres present in the State Archives of Belgium (http://www.arch.be) and the National Archives of the Netherlands (https://www.nationaalarchief.nl). An exception has been made for the communities in the current agglomerations of Brussels and Antwerp. Those communities were merged together in both agglomerations for the whole analysis as they could be considered urbanised already since the beginning of the 19th century. For the births before the start of the civil register (before ca 1800) only the parish of baptism was known. If there were more adjacent parishes in one community, for example in large cities such as Bruges, Ghent or Leuven, we considered the surface of the community based on the cadastres and not of the parish, which had only a church-administrative function.
Although we obtained many records of population size this way, it was not possible to obtain recorded census counts or estimates of the population of villages and towns for each specific date of birth in our dataset. For small communities, we simply used the available population size estimate for the year that was closest to the actual year of birth. For large communities, a more detailed approach was used, and the logarithm of population size was interpolated for each specific year of birth using a Hermite spline regression, fitted using the Fritsch and Carlson method [62], and implemented in R’s splinefun function. These spline interpolations are monotone if the underlying raw data were monotone, and therefore offer a good quality interpolation of population growth.
Archival sources for population density
The number of inhabitants in each place of birth during the year of birth was estimated from several data sources:
- After 1980:
•
Belgium: Rijksregister or Registre national Belge. Source = STATBEL or Statistics Belgium (http://statbel.fgov.be).
•
the Netherlands: Gemeentelijke Basisadministratie Personen. Source = Centraal Bureau voor de Statistiek (http://www.cbs.nl).
- 1800-1970: Official census counts in Belgium and the Netherlands
•
Belgium: 1846, 1856, 1866, 1876, 1880, 1890, 1900, 1910, 1920, 1930, 1947, 1961, 1970 and 1981. Source = STATBEL or Statistics Belgium (http://statbel.fgov.be).
•
the Netherlands: 1795, 1830, 1840, 1849, 1859, 1869, 1879, 1889, 1899, 1909, 1919, 1920, 1930, 1947, 1956, 1960 and 1971. Source = Centraal Bureau voor de Statistiek (http://www.volkstellingen.nl).
- Before 1800:
•
For urban areas (roughly containing > 5,000 citizens in 1850) and for the period from 1500 to 1800 we used population counts estimated from historical data, interpolated at 50 year intervals. Source = CLIO INFRA database (https://clio-infra.eu/Indicators/TotalUrbanPopulation.html).
•
For small communities, estimates were based on parish records and household counts. Official sources for Belgium: State Archives of Belgium (http://www.arch.be); for the Netherlands: National Archives of the Netherlands (https://www.nationaalarchief.nl).
Quantification and Statistical Analysis
Estimation of extra-pair paternity rates as a function of socio-demographic variables
We constructed a logistic regression model to estimate the probability of an extra-pair paternity event as a function of individual-level socio-demographic variables (social class and population density) based on the observed incidence of Y chromosome genotype mismatches. Our data was partly censored, i.e., if there was a genotype mismatch for a given genealogical pair then any of the potential ancestors could have been the result of an extra-pair paternity event (although observations were complete for each of the ancestors if there was no genotype mismatch, as this then provided certainty that no offspring were derived from extra-pair copulations in any of the ancestors). This censoring meant that we could not run a ‘regular’ logistic regression in which all ancestors of a pair with a Y chromosome mismatch would be coded as resulting from an EPP event, because this would lead to pseudo-replication in the data. Instead, we took this censoring into account in the outcome variable
We should note that even in cases where we had no information from extra DNA donors about the exact place in the genealogy where the EPP event(s) might have occurred we were still able to relate the incidence of EPP events to social class and population density due to the fact that these covariates were highly correlated within genealogical pairs. This can be seen from the fact that social mobility was very low (76%, 79% and 78% of all farmers and individuals from low and medium to high socioeconomic backgrounds stayed within the same social class as their parents) and that patrilocality was very high (87% of the individuals were born within 15 km of where their father was born; congruent to the historical averages for the Western world [63]).
After fitting an initial logistic regression model, we applied an empirical Bayes correction to account for the possibility that there could have been more than one EPP event within the same genealogical pair in case a mismatch occurred. Similar to Larmuseau et al. [8], this was done by iteratively refitting the logistic regression model to make an upward adjustment for the cases where a mismatch occurred based on the initially estimated EPP probabilities
The most parsimonious model was determined among a number of plausible models (see Table S1) on the basis of the Akaike Information Criterion AIC (see Table S4 for all coefficients and p-values). The minimal model included only social class as a predictor variable, and other predictors that were considered included population density in the natal city at birth (untransformed or log2 transformed), year of birth, the father’s age at birth, the mother’s age at birth, country or province (which correlate with predominant religion, i.e., Catholic for Belgium and Protestant for most Dutch provinces) and the sampling campaign (see above) from which the data were derived. This resulted in the selection of a best model (AIC = 225.57) in which social class and the log2 transformed population density at birth were included as terms. Adding any of the other variables increased the AIC values and therefore resulted in less parsimonious models (see Table S1 for an overview of all models run). However, it should be mentioned that a model that included social class and non-transformed population density at birth (rather than log2-transformed) had an evidence ratio of 1.1, meaning it is only 1.1 times less likely than the original model. This indicates that although it has a significant effect on EPP probability, the exact shape of the relationship between population density and EPP probability is difficult to determine from our data. In addition, we do not have the data to zoom in on the effects of density within towns or cities, only between them (the different classes may live in different sub-densities within the same city). As a robustness check on the model described above, we also ran a model on the level of the genealogical pair (rather than treating every ancestor as a data point). The outcome of this model is qualitatively the same as the model presented in the main text (results not shown). Finally, we also investigated whether a model in which population density is dichotomised between ‘urban’ (> 5,000 inhabitants in 1850) and ‘rural’ population sizes would be better able to explain the data, but this was not the case (see Table S1).
For the final interpretation of the best model we plotted overall per-case model predictions (Figure 2A) and drew effect plots of the partial effects of social class and density (Figures 2B and 2C), which were calculated using the emmeans R package. In addition, we also plotted the predicted extra-pair paternity rates in function as in historical urban and rural population densities and social class. Per-city urban densities (cities roughly containing > 5,000 citizens in 1850) were available for the entire historical range of our data, and as mentioned above were interpolated per year using Hermite spline regression, fitted using the Fritsch and Carlson method [62], and implemented in R’s splinefun function. From these values, we then calculated the aggregate per-province urban densities. Using official province-level census counts from Belgium and the Netherlands (available for the period between 1846 and 1947) (cf. above) we then calculated per-province average population densities, which were then again interpolated per year using Hermite spline regression. In addition, we extrapolated per-province average population densities for the period between 1750 and 1845 using urban densities in interaction with province as predictors in a linear heterogeneity-of-slopes model (both densities were first log transformed in this model). From these per-province average and urban population densities we finally calculated per-province rural population densities for the period between 1750 and 1950 (Figure S2). Finally, these estimated historical urban and rural population densities were used to obtain predicted EPP rates for this same period for all social classes (Figure 3).
Data and Code Availability
For privacy reasons it is required not to provide any potential link/association between surnames and genealogies with Y chromosome profiles or haplogroups [64]. The 299 Y chromosome profiles of the general population sample were submitted to YHRD, the Y chromosome STR haplotype reference database (accession numbers: YA003651, YA003652, YA003653, YA003739, YA003740, YA003741, YA003742, YA004300 and YA004301) and are given in Data S1B. If a specific surname or patrilineage will ever be associated with a profile of the general population sample in the future, this would only reveal participation of the family to the study and the information that no historical extra-pair paternity event was observed between the donor of the given profile and his paternal relative (see the defining criteria of the general population sample). The anonymised data together with the R script used for all analyses has been posted on Dryad (https://doi.org/10.5061/dryad.w6m905qk6).
Acknowledgments
The authors would like to thank all DNA donors and volunteers; Familiekunde Vlaanderen vzw, Jan Geypen, and Marie Boz for assistance in sampling; Michael Van Hecke for help in the genealogical analysis; three anonymous referees; and Jan Van Bavel, Erik Buyst, and Hendrik Larmuseau for comments on a previous version of our manuscript. This study is part of the “Gen-iaal Erfgoed” program of Histories vzw. Funding was provided by KU Leuven (BOF-C1 grant C12/15/013) and the Fund for Scientific Research – Flanders (Research grant number 1503216N and postdoc grants of M.H.D.L. and P.v.d.B.). F.C. was supported by Agencia Estatal de Investigación and Fondo Europeo de Desarollo Regional (FEDER) (grant CGL2016-75389-P), Agència de Gestió d’Ajuts Universitaris i de la Recerca (Generalitat de Catalunya) grant 2017 SGR00702, and “Unidad de Excelencia María de Maeztu,” funded by the MINECO (ref: MDM-2014-0370).
Author Contributions
Concept & Supervision, M.H.D.L.; Sampling & Y Chromosome Genotyping, M.H.D.L., S.C., L.G., M.V., K.N., and R.D.; Y Chromosome Comparison, M.H.D.L., F.C., and A.B.; Genealogical Data, M.H.D.L.; Data Analysis & Modeling, P.v.d.B. and T.W.; Writing, P.v.d.B., T.W., and M.H.D.L.
Declaration of Interests
The authors declare no competing interests.
Supplemental Information (3)
Document S1. Figures S1 and S2 and Tables S1–S4
Data S1. Y Chromosome Data, Related to STAR Methods
A) The used Y chromosome phylogeny. B) All 299 Y chromosome profiles of the general population sample. C) Distributions of the main Y chromosome haplogroups in the general population sample in comparison with those of previously published reference samples of the autochthonous populations of Flanders, the WGS sample [S1], and the Netherlands, the GoNL sample [S2]. For this comparison the general population sample was separated in a ‘Flemish subset’ and a ‘Dutch subset’. N, sample size; Freq, frequency. D) Distributions of all observed Y chromosome subhaplogroups in the total ‘general population sample’ and in the ‘Flemish subset’ and ‘Dutch subset’. N, sample size; Freq, frequency.
Document S2. Article plus Supplemental Information
References
Kempenaers, B. ∙ Verheyen, G.R. ∙ Van den Broeck, M. ...
Extra-pair paternity results from female preference for high-quality males in the blue tit
Nature. 1992; 357:494-496Gray, P.B. ∙ Anderson, K.G.
Fatherhood: Evolution and Human Paternal Behavior
Harvard University Press, 2010Kempenaers, B. ∙ Schlicht, E.
Extra-pair behaviour
Kappeler, P.
Animal Behaviour: Evolution and Mechanisms
Springer, 2010; 359-412
Griffith, S.C. ∙ Owens, I.P.F. ∙ Thuman, K.A.
Extra pair paternity in birds: a review of interspecific variation and adaptive function
Mol. Ecol. 2002; 11:2195-2212Platek, S.M. ∙ Shackelford, T.K.
Female Infidelity and Paternal Uncertainty: Evolutionary Perspectives on Male Anti-Cuckoldry Tactics
Cambridge University Press, 2006Larmuseau, M.H.D. ∙ Matthijs, K. ∙ Wenseleers, T.
Cuckolded fathers rare in human populations
Trends Ecol. Evol. 2016; 31:327-329Larmuseau, M.H.D. ∙ Matthijs, K. ∙ Wenseleers, T.
Long-term trends in human extra-pair paternity: increased infidelity or adaptive strategy? A reply to Harris (2016)
Trends Ecol. Evol. 2016; 31:663-665Larmuseau, M.H. ∙ Vanoverbeke, J. ∙ Van Geystelen, A. ...
Low historical rates of cuckoldry in a Western European human population traced by Y-chromosome and genealogical data
Proc. Biol. Sci. 2013; 280:20132400Scelza, B.A.
Choosy but not chaste: multiple mating in human females
Evol. Anthropol. 2013; 22:259-269Hrdy, S.B.
The optimal number of fathers. Evolution, demography, and history in the shaping of female mate preferences
Ann. N Y Acad. Sci. 2000; 907:75-96Gangestad, S.W. ∙ Simpson, J.A.
The evolution of human mating: trade-offs and strategic pluralism
Behav. Brain Sci. 2000; 23:573-587, discussion 587–644Akçay, E. ∙ Roughgarden, J.
Extra-pair paternity in birds: review of the genetic benefits
Evol. Ecol. Res. 2007; 9:855-868Eliassen, S. ∙ Kokko, H.
Current analyses do not resolve whether extra-pair paternity is male or female driven
Behav. Ecol. Sociobiol. 2008; 62:1795-1804Westneat, D.F. ∙ Stewart, I.R.K.
Extra-pair paternity in birds: causes, correlates, and conflict
Annu. Rev. Ecol. Evol. Syst. 2003; 34:365-396Brouwer, L. ∙ van de Pol, M. ∙ Aranzamendi, N.H. ...
Multiple hypotheses explain variation in extra-pair paternity at different levels in a single bird family
Mol. Ecol. 2017; 26:6717-6729Greiling, H. ∙ Buss, D.M.
Women’s sexual strategies: the hidden dimension of extra-pair mating
Pers. Individ. Dif. 2000; 28:929-963Anderson, K.G.
How well does paternity confidence match actual paternity? Evidence from worldwide nonpaternity rates
Curr. Anthropol. 2006; 47:513-520Maldonado-Chaparro, A.A. ∙ Montiglio, P.O. ∙ Forstmeier, W. ...
Linking the fine-scale social environment to mating decisions: a future direction for the study of extra-pair paternity
Biol. Rev. Camb. Philos. Soc. 2018; 93:1558-1577Shellman-Reeve, J.S. ∙ Reeve, H.K.
Extra-pair paternity as the result of reproductive transactions between paired mates
Proc. Biol. Sci. 2000; 267:2543-2546Bellis, M.A. ∙ Hughes, K. ∙ Hughes, S. ...
Measuring paternal discrepancy and its public health consequences
J. Epidemiol. Community Health. 2005; 59:749-754Geary, D.C.
Evolution of fatherhood
Salmon, C.A. ∙ Shackelford, T.K.
Family Relationships: An Evolutionary Perspective
Oxford Scholarschip Online, 2007
Laslett, P. ∙ Oosterveen, K. ∙ Smith, R.M.
Bastardy and Its Comparative History: Studies in the History of Illegitimacy and Marital Nonconformism in Britain, France, Germany, Sweden, North America, Jamaica, and Japan
Harvard University Press, 1980Griffin, E.
Sex, illegitimacy and social change in industrializing Britain
Soc. Hist. 2013; 38:139-161Matthys, C.
Discourses versus life courses: servants’ extramarital sexual activities in Flanders during the nineteenth and early twentieth centuries
J. Urban Hist. 2016; 42:81-100Larmuseau, M.H.D. ∙ Claerhout, S. ∙ Gruyters, L. ...
Genetic-genealogy approach reveals low rate of extrapair paternity in historical Dutch populations
Am. J. Hum. Biol. 2017; 29:e23046Voracek, M. ∙ Haubner, T. ∙ Fisher, M.L.
Recent decline in nonpaternity rates: a cross-temporal meta-analysis
Psychol. Rep. 2008; 103:799-811Cerda-Flores, R.M. ∙ Barton, S.A. ∙ Marty-Gonzalez, L.F. ...
Estimation of nonpaternity in the Mexican population of Nuevo Leon: a validation study with blood group markers
Am. J. Phys. Anthropol. 1999; 109:281-293Kok, J. ∙ Bras, H. ∙ Rotering, P.
Courtship and bridal pregnancy in The Netherlands, 1870-1950
Ann. Demogr. Hist. (Paris). 2016; 132:165-191Tilly, L.A. ∙ Scott, J.W. ∙ Cohen, M.
Women’s work and European fertility patterns
J. Interdiscip. Hist. 1976; 6:447-476Strassmann, B.I. ∙ Kurapati, N.T. ∙ Hug, B.F. ...
Religion as a means to assure paternity
Proc. Natl. Acad. Sci. USA. 2012; 109:9781-9785Claerhout, S. ∙ Vandenbosch, M. ∙ Nivelle, K. ...
Determining Y-STR mutation rates in deep-routing genealogies: Identification of haplogroup differences
Forensic Sci. Int. Genet. 2018; 34:1-10Bloothooft, G. ∙ Schraagen, M.P.
Learning name variants from inexact high-confidence matches
Bloothooft, G. ∙ Christen, P. ∙ Mandemakers, K. ...
Population Reconstruction
Springer, 2015; 61-83
van Drie, R. ∙ Needs, S.
Dutch Roots. Finding Your Ancestors in the Netherlands
Centraal Bureau voor Genealogie, 2012Larmuseau, M.H.D. ∙ Vanoverbeke, J. ∙ Gielis, G. ...
In the name of the migrant father--analysis of surname origins identifies genetic admixture events undetectable from genealogical records
Heredity. 2012; 109:90-95Athey, W.T.
Haplogroup prediction from Y-STR values using a Bayesian-allele-frequency approach
J. Genet. Geneal. 2006; 2:34-39Karafet, T.M. ∙ Mendez, F.L. ∙ Meilerman, M.B. ...
New binary polymorphisms reshape and increase resolution of the human Y chromosomal haplogroup tree
Genome Res. 2008; 18:830-838Sims, L.M. ∙ Garvey, D. ∙ Ballantyne, J.
Improved resolution haplogroup G phylogeny in the Y chromosome, revealed by a set of newly characterized SNPs
PLoS ONE. 2009; 4:e5792Larmuseau, M.H.D. ∙ Vanderheyden, N. ∙ Van Geystelen, A. ...
Increasing phylogenetic resolution still informative for Y chromosomal studies on West-European populations
Forensic Sci. Int. Genet. 2014; 9:179-185van Oven, M. ∙ Ralf, A. ∙ Kayser, M.
An efficient multiplex genotyping approach for detecting the major worldwide human Y-chromosome haplogroups
Int. J. Legal Med. 2011; 125:879-885Caratti, S. ∙ Gino, S. ∙ Torre, C. ...
Subtyping of Y-chromosomal haplogroup E-M78 (E1b1b1a) by SNP assay and its forensic application
Int. J. Legal Med. 2009; 123:357-360van Oven, M. ∙ Van Geystelen, A. ∙ Kayser, M. ...
Seeing the wood for the trees: a minimal reference phylogeny for the human Y chromosome
Hum. Mutat. 2014; 35:187-191Chiaroni, J. ∙ Underhill, P.A. ∙ Cavalli-Sforza, L.L.
Y chromosome diversity, human expansion, drift, and cultural evolution
Proc. Natl. Acad. Sci. USA. 2009; 106:20174-20179Larmuseau, M.H.D. ∙ Otten, G.P.P.L. ∙ Decorte, R. ...
Defining Y-SNP variation among the Flemish population (Western Europe) by full genome sequencing
Forensic Sci. Int. Genet. 2017; 31:e12-e16Francioli, L. ∙ Menelaou, A. ∙ Pulit, S. ..., Genome of the Netherlands Consortium
Whole-genome sequence variation, population structure and demographic history of the Dutch population
Nat. Genet. 2014; 46:818-825Larmuseau, M.H.D. ∙ Vanderheyden, N. ∙ Jacobs, M. ...
Micro-geographic distribution of Y-chromosomal variation in the central-western European region Brabant
Forensic Sci. Int. Genet. 2011; 5:95-99Larmuseau, M.H.D. ∙ Boon, N. ∙ Vanderheyden, N. ...
High Y-chromosomal diversity and low relatedness between paternal lineages on a communal scale in the Western European Low Countries during the surname establishment
Heredity. 2015; 115:3-12Busby, G.B.J. ∙ Brisighelli, F. ∙ Sánchez-Diz, P. ...
The peopling of Europe and the cautionary tale of Y chromosome lineage R-M269
Proc. Biol. Sci. 2012; 279:884-892Balanovsky, O.
Toward a consensus on SNP and STR mutation rates on the human Y-chromosome
Hum. Genet. 2017; 136:575-590Bird, S.C.
Towards improvements in the estimation of the coalescent: implications for the most effective use of Y chromosome short tandem repeat mutation rates
PLoS ONE. 2012; 7:e48638Van Geystelen, A. ∙ Decorte, R. ∙ Larmuseau, M.H.D.
Updating the Y-chromosomal phylogenetic tree for forensic applications based on whole genome SNPs
Forensic Sci. Int. Genet. 2013; 7:573-580Walsh, B.
Estimating the time to the most recent common ancestor for the Y chromosome or mitochondrial DNA for a pair of individuals
Genetics. 2001; 158:897-912Ballantyne, K.N. ∙ Goedbloed, M. ∙ Fang, R. ...
Mutability of Y-chromosomal microsatellites: rates, characteristics, molecular bases, and forensic implications
Am. J. Hum. Genet. 2010; 87:341-353Larmuseau, M.H.D. ∙ Vanderheyden, N. ∙ Van Geystelen, A. ...
Recent radiation within Y-chromosomal haplogroup R-M269 resulted in high Y-STR haplotype resemblance
Ann. Hum. Genet. 2014; 78:92-103van Leeuwen, M. ∙ Maas, I.
HISCLASS - A Historical International Social Class Scheme
Leuven University Press, 2011van Leeuwen, M. ∙ Maas, I. ∙ Miles, A.
HISCO - Historical International Standard Classification of Occupations
Leuven University Press, 2002Mandemakers, K. ∙ Muurling, S. ∙ Maas, I. ...
HSN Standardized, HISCO-Coded and Classified Occupational titles, release 2013.01
IISG, 2013Fritsch, F.N. ∙ Carlson, R.E.
Monotone piecewise cubic interpolation
SIAM J. Numer. Anal. 1980; 17:238-246Kaplanis, J. ∙ Gordon, A. ∙ Shor, T. ...
Quantitative analysis of population-scale family trees with millions of relatives
Science. 2018; 360:171-175Larmuseau, M.H.D. ∙ Bekaert, B. ∙ Baumers, M. ...
Biohistorical materials and contemporary privacy concerns-the forensic case of King Albert I
Forensic Sci. Int. Genet. 2016; 24:202-210Figures (3)
Article metrics
-
-
26Citations75Captures27Mentions
-
Supplemental information (3)
PDF (1.07 MB)
Document S1. Figures S1 and S2 and Tables S1–S4
Spreadsheet (103.32 KB)
Data S1. Y Chromosome Data, Related to STAR Methods
PDF (2.81 MB)
Document S2. Article plus Supplemental Information
