Whether or not the spread of agriculture in Europe was accompanied by movements of people is a long-standing question in archeology and anthropology, which has been frequently addressed with the help of population genetic data. Estimates on dates of expansion and geographic origins obtained from genetic data are however sensitive to the calibration of mutation rates and to the mathematical models used to perform inference. For instance, recent data on the Y chromosome haplogroup R1b1b2 (M269) have either suggested a Neolithic origin for European paternal lineages or a more ancient Paleolithic origin depending on the calibration of Y-STR mutation rates. Here we examine the date of expansion and the geographic origin of hgR1b1b2 considering two current estimates of mutation rates in a total of fourteen realistic wave-of-advance models. We report that a range expansion dating to the Paleolithic is unlikely to explain the observed geographical distribution of microsatellite diversity, and that whether the data is informative with respect to the spread of agriculture in Europe depends on the mutation rate assumption in a critical way.
Citation: Sjödin P, François O (2011) Wave-of-Advance Models of the Diffusion of the Y Chromosome Haplogroup R1b1b2 in Europe. PLoS ONE 6(6): e21592. https://doi.org/10.1371/journal.pone.0021592
Editor: Carles Lalueza-Fox, Institut de Biologia Evolutiva - Universitat Pompeu Fabra, Spain
Received: March 3, 2011; Accepted: June 3, 2011; Published: June 24, 2011
Copyright: © 2011 Sjödin, François. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: OF received support from the Institute of Complex Systems (IXXI, www.ixxi.fr). PS was supported by an Institutional Grant for Younger Researchers from STINT (The Swedish Foundation for International Cooperation in Research and Higher Education, www.stint.se) to Mattias Jakobsson and Michael Blum. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Since the development of molecular markers, genetics has been extensively used to address the question of the diffusion of agriculture into Europe, one of the long-standing debates in archaeology and anthropology –. Though archaeologists have considered more sophisticated models of diffusion in the last decade , two scenarios are frequently contrasted in the population genetics literature. In 1971, a seminal work by Ammerman and Cavalli-Sforza used radiocarbon dates from Neolithic sites to propose a “wave-of-advance” model of the spread of agriculture in Europe. In this “demic” process, local population growth and migration produce demographic expansion following a traveling wave from the southeast to the northwest of Europe . The wave-of-advance model argues that farmers expanded into Europe from West-Asia about 10,000 years ago, and replaced resident hunter-gatherers with little or no genetic admixture , . Alternatively, several archaeologists have hypothesized a cultural model of the development of agriculture, where cultivated plants, domesticated animals and the associated agricultural techniques were adopted with only limited human movements , . According to this “cultural” model, the Neolithic farmers did not migrate. Instead the technologies were transmitted to the resident hunter-gatherers who changed their lifestyle and converted to farming. Yet the prehistory of European populations is poorly understood, and the debate between the demic and cultural diffusion models is still active today.
Inference on demographic history of European populations is commonly based on the estimation of coalescent ages of mitochondrial and Y chromosome haplogroups in modern populations –. Estimations of times since the most recent ancestor (TMRCAs) based on mtDNA have suggested a Paleolithic origin of European maternal lineages –, but see ,  for suggestions of a Neolithic contribution to the maternal gene pool. In contrast, several studies of Y chromosome haplogroups have suggested more recent origins for the paternal lineages , , , , . The results in the latter studies have been interpreted as support for the demic diffusion model, implying that distinct migration patterns took place for women and men in Europe. Offering a more direct view of the past, ancient mtDNA has been recently used to genetically characterize a farming population of the Linear Pottery Culture in Central Europe –. Whereas Haak et al  lent weight to the arguments for a Paleolithic origin of Europeans, the subsequent analyses supported little admixture with hunter-gatherer populations , genetic affinities of Neolithic farmers with West-Asian populations and significant post-Neolithic events .
Among European Y chromosome lineages, haplogroup (hg) R1b1b2 (R-M269) is carried by 110 million European men, and increases in frequency from east to west . Using germline mutation rates (GMR), Balaresque et al  reported that the distribution of hgR1b1b2 microsatellite diversity is best explained by spread from a single source in the Near East during the Neolithic. Mutation rate assumptions, however, have a large impact on molecular dating. Using the evolutionary mutation rate (EMR) proposed by Zhivotovsky et al , , Morelli et al  found strong support for considerably older TMRCAs than estimated in . Here we re-investigate whether the spatial distribution of microsatellite diversity of hgR1b1b2 supports a demic or a cultural model of expansion of agriculture into Europe, or if it results from a more recent expansion as suggested in . In previous studies, TMRCAs and population growth rates were estimated using the computer program BATWING, which assumes a model of exponential population growth and divergence without gene flow . The EMR was introduced to correct for the inaccuracy of approximations made by demographic models such as those assumed in BATWING –, , and it has been observed that the use of corrected rates can increase TMRCA estimates by a 3-fold factor . Another shortcoming of the BATWING model is its failure to reproduce the characteristics of a wave-of-advance, where recurrent founder events occur during range expansion. In addition, estimating the TMRCA of a haplogroup is not necessarily relevant to the study of the expansion time of this haplogroup , .To overcome these issues and better evaluate the alternative hypotheses for expansion dates, we implemented 14 realistic wave-of-advance models using GMR or EMR estimates as proposed in previous studies –. The wave-of-advance models are designed to capture the history of hgR1b1b2 implicit in the assumptions of cultural or demic diffusion of agriculture in Europe ,  and scenarios of more recent expansions. We also discuss the implications of using GMR estimates in a demic expansion model for the hgR1b1b2 data.
Results and Discussion
We used nine microsatellite markers from 840 European Y-chromosomes typed from hgR1b1b2 (R-M269), a common haplogroup in Europe . Fourteen distinct wave-of-advance models were fitted to the microsatellite allelic richness and to the geographical distribution of microsatellite diversity using computer simulations. The models sorted into three main categories representing the alternative hypotheses of a Paleolithic (21ky ago), Neolithic (10Ky ago) or post-Neolithic (3Ky ago) expansion. Two distinct calibrations of the Y-STR mutation rates, GMR and EMR, were used for each demographic scenario (see Materials and Methods and Table 1 and Figure 1).
Distribution of allelic richness
In the original data the number of alleles observed at each locus ranged between 4 and 6. To test whether similar levels of allelic richness could be reproduced by the 14 wave-of-advance models, genetic variation was simulated at 1,000 microsatellite loci under each model (Figure 1). The levels of allelic richness observed in a large proportion of simulated data sets were compatible with those observed in the original data. However we found significant differences between the 14 simulated distributions of allelic richness (Kruskall-Wallis test P<10−6). Generally, models based on the EMR estimate provided a better fit to the data than models using GMR estimates. The fit was better regardless of whether a post-Neolithic, a Neolithic or a Paleolithic origin of hgR1b1b2 was assumed. Zhivotovsky et al  suggested that the high microsatellite mutation rates estimated from germlines are a consequence of additional genetic drift due to population bottlenecks not taken into account by the simplistic evolutionary models used to estimate mutation rates. While wave-of-advance models incorporate effects of recurrent bottlenecks, these models are still obvious simplifications of human demographic expansions. Our results show that the EMR correction proposed by Zhivotovsky et al appears to be useful in the wave-of-advance simulation framework.
Spatial distribution of microsatellite genetic diversity
To investigate if wave-of-advance models could explain the geographic distribution of microsatellite diversity, we restricted our study to simulations that reproduced the number of alleles at each of the 9 microsatellite loci exactly. Using rejection sampling, we produced 100,000 data sets from each model, and measured genetic diversity by the variance in allele size for the 21 population samples in the actual and simulated data sets. Then we evaluated the respective fit of the models by computing the sum of squared differences between the simulated and the actual genetic diversity estimates , . Significantly distinct results were produced by the models (P<10−15). Our results show that an expansion in Neolithic or Mesolithic times (350 generations ago or 10 ky) leads to a lower sum of squared errors than post-glacial re-expansion started 700 generations ago (21 ky ago), regardless of assuming a GMR or EMR model (Figure 2 and Table 2). Using GMRs, simulations of recent (100 generations ago) and rapid expansions from three distinct origins provided a better fit to the geographical distribution of microsatellite diversity than did models with expansion started 350 generations ago. Although models of recent origins using GMRs provided poorer fit than a model of Neolithic expansion using the EMR (Figure 2 and Table 2), the small observed difference makes them however difficult to discriminate (odd ratio = 1.7; Figure 3).
Best fitting simulation obtained under Model A) Recent expansion from Anatolia (GMR), B) Neolithic expansion from Anatolia (GMR), C) Neolithic expansion from Anatolia (EMR), D) Genetic diversity in the actual data. Circles indicate sample locations.
Comparison to other studies
Using the program BATWING with GMRs, Balaresque et al  argued in favor of a Neolithic expansion hypothesis for hgR1b1b2 based on estimates of TMRCAs between 5.5 and 8.0 ky BP. In contrast Morelli et al  implemented the EMR and obtained much older TMRCAs, between 14.8 and 32.6 ky BP, supporting a Paleolithic origin. On the other hand, two additional studies employing the EMR obtained TMRCA estimates suggesting that a Neolithic expansion of hgR1b1b2 is more plausible than a Paleolithic expansion ,  (see Supplementary Materials of ). Myres et al  coalescent estimate for the Y-STR R1b1b2 network tree is 10,270±1,680 years BP, close to the median TMRCAs (8.6–12.2 ky) of the M269 clade obtained by Shi et al . Cruciani et al  reported expansion time estimates for hg R1b1b2g and R1b1b2h equal to 8.3 ky BP (95% CI 5.8–10.9ky BP) and 7.4 ky BP (95% CI 5.3–10.2 ky BP) respectively. Their study employed a mutation rate intermediate between the EMR and GMR, and reported TMRCA estimates in-between the estimates of Morelli et al  and Balaresque et al . This is in line with Shi et al  where the authors investigated the effect of assuming different sets of mutation rates on the outputs of the BATWING algorithm and found that TMRCA estimates based on the EMR are generally larger than estimates based on GMRs. The observations of  also imply that if the GMR estimates are the correct rates to use in a spatially expanding population model, some of the previously cited studies would point out to dates of expansion of hgR1b1b2 much more recent than 10,000 years. Moreover, studies of Y-chromosomal haplogroup J, a major haplogroup in south-eastern Europe, have suggested that the importance of more recent expansion events may have been underestimated . Although the result needs confirmation, post-Neolithic expansions are also supported by ancient DNA . In fact the most common Y chromosome hgs in modern Europe are not observed in a population of the earliest farming culture in Central Europe (3 males, ). When we used GMR estimates, our results pointed to a similar conclusion. Wave-of-advance models with a recent expansion date received higher support than models of Neolithic expansion (Table 2, Figures 2–3).
Drawing reliable conclusions about the timing and geographic origin of expansions from genetic data requires a precise modeling of the hypothetic processes that generated the observed genetic variation. As shown previously, estimates of TMRCA are strongly sensitive to prior information on mutation rates , . We found that wave-of-advance models can reproduce the geographical distribution of the microsatellite diversity of hg R1b1b2 very accurately (Figure 3). To what extent this distribution supports the demic or cultural dispersal model of agriculture in Europe critically depends on whether the faster germ-line mutation rates or the slower evolutionary mutation rate better capture reality in these models. We found it difficult to discriminate among models assuming EMR (Neolithic expansion) and models assuming GMRs (Recent expansion). An interpretation of our results is as support for the use of the correction proposed by Zhivotovsky et al ,  in wave-of-advance models. Historical events consistent with recent expansions from the south of Europe during the Bronze age  or the Greek and Roman civilization in Europe and West Asia  cannot be excluded, but the impact of such demographic events on European genomes requires confirmation from ancient DNA studies.
Materials and Methods
We used nine microsatellite markers from 840 European Y-chromosomes typed from hgR1b1b2 (R-M269), a common haplogroup in Europe. The population samples were all included in the analyses of Balaresque et al . The data set contained 21 samples from 5 populations from France, 4 from Spain, 3 from the British Isles and Turkey, and 2 from Germany, Italy, Denmark and the Netherlands.
Simulations of “wave-of-advance” models were performed with the computer program SPLATCHE2 . The program was used to run non-equilibrium stepping-stone simulations on a lattice of demes mirroring the geography of Europe. More specifically, range expansions occurred in a 64×42 lattice of 2,688 demes covering Europe from latitude 38°N to 65°N and from longitude −10°E to 40°E . In the stepping-stone simulations, local populations sent migrants to their nearest neighbors at rate m. The establishment probabilities of incoming individuals were inversely proportional to specific friction values that accounted for geographic obstacles, such as mountain areas and seas. Within each deme, the population size grew according to a logistic model with growth rate r, and saturated at the carrying capacity, K. Based on anthropological data, we used different estimates of population density for expansions started during the late Upper Paleolithic or during the Neolithic. Population density was equal to ∼0.05 individual per km2 for Paleolithic populations, and a 10-fold higher value was chosen for farming populations during the Neolithic , , . To match these prehistoric population density values, carrying capacities were set to K = 50 (Paleolithic expansion) and K = 500 (Neolithic expansion) in each deme. After the completion of a demographic phase generating a wave-of-advance of populations in Europe, SPLATCHE2 simulates multilocus microsatellite genotypes according to a stepwise mutation model. We simulated nine microsatellite loci for each of the 840 European individuals located at the geographic sites specified in Balaresque et al .
The 14 models
Fourteen distinct wave-of-advance models summarized in Table 1 and Figure 1 were compared to the data using computer simulations. The models sorted into three main categories representing the two alternative hypotheses of a demic or a cultural diffusion of agriculture and a third hypothesis of a more recent expansion scenario , . For the demic diffusion model, we assumed that range expansion started around 10ky ago, and the origin of the spread was in Anatolia (39°N, 32°E), southeast to Europe. For the cultural diffusion model, we assumed that range expansion started around 21ky ago, and three distinct origins were considered. Geographic origins in Anatolia, Iberian Peninsula (Spain 40°N, 3°E, southwestern Europe) and Italy (41°N, 13°E, southern Europe) were chosen to mirror the locations of glacial refugia in southern Europe. Recent expansions were started 3Ky ago, and we used the same 3 geographic origins as in the previous models. The SPLATCHE parameters that reproduce these demographic expansion scenarios are given in Table 1. Additionally two Y-STR mutation rate calibrations were included in the models: the comparatively high microsatellite germline mutation rate (GMR) values ranging between 6×10−4 and 3×10−3 mutation per generation  and the lower “evolutionary” mutation rate (EMR, ). In the Zhivotovsky method, the ages of haplogroups in populations are estimated using an evolutionary effective mutation rate of Y-STR of 6.96×10−4 per generation. In preliminary runs, we also investigated expansions corresponding to the initial colonization of Europe by modern humans around 40,000 years ago (1,500 generations ago) but we did not retain these models due to their poor fit to the observed data.
Conceived and designed the experiments: PS OF. Performed the experiments: OF. Analyzed the data: PS OF. Wrote the paper: PS OF.
- 1. Menozzi P, Piazza A, Cavalli-Sforza LL (1978) Synthetic maps of human gene frequencies in Europeans. Science 201: 786–792.
- 2. Cavalli-Sforza LL, Menozzi P, Piazza A (1994) The History and Geography of Human Genes. Princeton (New Jersey): Princeton University Press.
- 3. Jobling MA, Hurles M, Tyler-Smith C (2004) Human Evolutionary Genetics: Origins, Peoples and Disease. Garland Science.
- 4. Whittle A, Cummings VGoing Over: the Mesolithic-Neolithic Transition in Northwest Europe. Proceedings of the British Academy 144, Oxford University Press, 2007.
- 5. Ammerman AJ, Cavalli-Sforza LL (1971) Measuring the rate of spread of early farming in Europe. Man, New Series 6: 674–688.
- 6. Ammerman AJ, Cavalli-Sforza LL (1984) The Neolithic Transition and the Genetics of Populations in Europe. Princeton University Press, Princeton.
- 7. Diamond J, Bellwood P (2003) Farmers and their languages: the first expansions. Science 300: 597–603.
- 8. Barker G (2006) The Agricultural Revolution in Prehistory (Oxford: Oxford University Press).
- 9. Richards M, Côrte-Real H, Forster P, Macaulay V, Wilkinson-Herbots H, et al. (1996) Paleolithic and Neolithic lineages in the European mitochondrial gene pool. Am J Hum Genet 59: 185–203.
- 10. Richards M, Macaulay V, Hickey E, Vega E, Sykes B, et al. (2000) Tracing European founder lineages in the near eastern mtDNApool. Am J Hum Genet 67: 1251–1276.
- 11. Torroni A, Bandelt HJ, Macaulay V, Richards M, Cruciani F, et al. (2001) A signal, from human mtDNA, of postglacial recolonization in Europe. Am J Hum Genet 69: 844–852.
- 12. Achilli A, Rengo C, Magri C, Battaglia V, Olivieri A, et al. (2004) The molecular dissection of mtDNA haplogroup H confirms that the Franco-Cantabrian glacial refuge was a major source for the European gene pool. Am J Hum Genet 75: 910–918.
- 13. Semino O, Passarino G, Oefner PJ, Lin AA, Arbuzova S, et al. (2000) The genetic legacy of Paleolithic Homo sapiens sapiens in extant Europeans: a Y chromosome perspective. Science 290: 1155–1159.
- 14. Chikhi L, Nichols RA, Barbujani G, Beaumont MA (2002) Y genetic data support the Neolithic demic diffusion model. Proc Natl Acad Sci USA 99: 11008–11013.
- 15. Balaresque P, Bowden GR, Adams SM, Leung HY, King TE, et al. (2010) A predominantly neolithic origin for European paternal lineages. PLoS Biol 8: e1000285.
- 16. Morelli L, Contu D, Santoni F, Whalen MB, Francalacci P, et al. (2010) A comparison of Y-chromosome variation in Sardinia and Anatolia is more consistent with cultural rather than demic diffusion of Agriculture. PLoS ONE 5(4): e10419.
- 17. Myres NM, Rootsi S, Lin AA, Järve M, King RJ, et al. (2011) major Y-chromosome haplogroup R1b Holocene era founder effect in Central and Western Europe. Eur J Hum Genet 19(1): 95–101.
- 18. Shi W, Ayub Q, Vermeulen M, Shao RG, Zuniga S, et al. (2010) A worldwide survey of male demographic history based on Y-SNP and Y-STR data from the HGDP-CEPH populations. Mol Biol Evol 27: 385–393.
- 19. Soares P, Achilli A, Semino O, Davies W, Macaulay V, et al. (2010) The archaeogenetics of Europe. Curr Biol 20: R174–83.
- 20. Barbujani G (1998) Evidence for paleolithic and Neolithic gene flow in Europe. Am J Hum Genet 62: 488–491.
- 21. Simoni L, Calafell F, Pettener D, Bertranpetit J, Barbujani G (2000) Geographic patterns of mtDNA diversity in Europe. Am J Hum Genet 66: 262–278.
- 22. Chikhi L, Destro-Bisol G, Bertorelle G, Pascali V, Barbujani G (1998) Clines of nuclear DNA markers suggest a largely neolithic ancestry of the European gene pool. Proc Natl Acad Sci USA 95: 9053–9058.
- 23. Dupanloup I, Bertorelle G, Chikhi L, Barbujani G (2004) Estimating the impact of prehistoric admixture on the genome of Europeans. Mol Biol Evol 21: 1361–1372.
- 24. Haak W, Forster P, Bramanti B, Matsumura S, Brandt G, et al. (2005) Ancient DNA from the first European farmers in 7500-year-old Neolithic sites. Science 310: 1016–1018.
- 25. Bramanti B, Thomas MG, Haak W, Unterlaender M, Jores P, et al. (2009) Genetic discontinuity between local hunter-gatherers and central Europe's first farmers. Science 326: 137–140.
- 26. Haak W, Balanovsky O, Sanchez JJ, Koshel S, Zaporozhchenko V, et al. (2010) Ancient DNA from European early neolithic farmers reveals their near eastern affinities. PLoS Biol 8(11): e1000536.
- 27. Chiaroni J, Underhill P, Cavalli-Sforza LL (2009) Y chromosome diversity, human expansion, drift and cultural evolution. Proc Natl Acad Sci USA 106: 20174–20179.
- 28. Zhivotovsky LA, Underhill PA, Cinnioglu C, Kayser M, Morar B, et al. (2004) The effective mutation rate at Y chromosome short tandem repeats, with application to human population-divergence time. Am J Hum Genet 74: 50–61.
- 29. Zhivotovsky LA, Underhill PA, Feldman MW (2006) Difference between evolutionarily effective and germ line mutation rate due to stochastically varying haplogroup size. Mol Biol Evol 23: 2268–2270.
- 30. Wilson IJ, Weale ME, Balding DJ (2003) Inferences from DNA data: population histories, evolutionary processes and forensic match probabilities. Journal of the Royal Statistical Society 166: 155–188.
- 31. Liu H, Prugnolle F, Manica A, Balloux F (2006) A geographically explicit genetic model of worldwide human-settlement history. Am J Hum Genet 79(2): 230–237.
- 32. François O, Blum MGB, Jakobsson M, Rosenberg NA (2008) Demographic history of European populations of Arabidopsis thaliana. PLoS Genet 4(5): e1000075.
- 33. Cruciani F, Trombetta B, Antonelli C, Pascone R, Valesini G, et al. (2011) Strong intra- and inter-continental differentiation revealed by Y chromosome SNPs M269, U106 and U152. Forensic Sci Int Genet in press. https://doi.org/10.1016/j.fsigen.2010.07.006
- 34. Di Giacomo F, Luca F, Popa LO, Akar N, Anagnou J, et al. (2004) Y chromosomal haplogroup J as a signature of the post-neolithic colonization of Europe. Hum Genet 115: 357–371.
- 35. Martin TR (1992) Ancient Greece: From Prehistoric to Hellenistic Times. Yale University Press.
- 36. Ray N, M Currat, M Foll, L Excoffier (2010) SPLATCHE2: a spatially explicit simulation framework for complex demography, genetic admixture and recombination. Bioinformatics 26: 2993–2994.
- 37. Francois O, Currat M, Ray N, Han E, Excoffier L, et al. (2010) Principal Component Analysis under population genetic models of range expansion and admixture. Mol Biol Evol 27: 1257–1268.
- 38. Currat M, Excoffier L (2005) The effect of the Neolithic expansion on European molecular diversity. Proc Biol Sci 272: 679–688.