Abstract
In northwestern Africa, lifestyle transitioned from foraging to food production around 7,400 years ago but what sparked that change remains unclear. Archaeological data support conflicting views: (1) that migrant European Neolithic farmers brought the new way of life to North Africa1,2,3 or (2) that local hunter-gatherers adopted technological innovations4,5. The latter view is also supported by archaeogenetic data6. Here we fill key chronological and archaeogenetic gaps for the Maghreb, from Epipalaeolithic to Middle Neolithic, by sequencing the genomes of nine individuals (to between 45.8- and 0.2-fold genome coverage). Notably, we trace 8,000 years of population continuity and isolation from the Upper Palaeolithic, via the Epipaleolithic, to some Maghrebi Neolithic farming groups. However, remains from the earliest Neolithic contexts showed mostly European Neolithic ancestry. We suggest that farming was introduced by European migrants and was then rapidly adopted by local groups. During the Middle Neolithic a new ancestry from the Levant appears in the Maghreb, coinciding with the arrival of pastoralism in the region, and all three ancestries blend together during the Late Neolithic. Our results show ancestry shifts in the Neolithization of northwestern Africa that probably mirrored a heterogeneous economic and cultural landscape, in a more multifaceted process than observed in other regions.
Similar content being viewed by others
Main
North Africa’s geographic location, centred between the vast Saharan desert, the fertile Near East and Mediterranean Europe, has resulted in a complex human history in the area7,8. The fossil record suggests long-term hominid and human presence9, although continuity over the past 100,000 years cannot be deduced due to the fragmented nature of the record. In the Late Pleistocene, 15,000 years ago, the remains of foragers excavated in Morocco show a distinct genetic make-up intermediate between contemporary Levantine foragers and sub-Saharan African populations10. Current-day North Africans are largely related to Eurasian populations, which was probably caused by ‘back-to-Africa’ migrations7.
Both archaeological records and archaeogenomic data show that Neolithic farmers (genetically distinct from European foragers) dispersed from the northern Levant and Anatolia to the Mediterranean islands, Italian peninsula and Iberia11,12,13,14,15,16,17,18. Mediterranean coastal routes have long been recognized in the archaeological record as an important part of the Neolithic expansion in Europe. In the western Mediterranean, Impressed Ware technology—and further the Cardial Horizon—spread along the European mainland coast and islands to reach the Iberian peninsula, where both phenomena are present at 7,550 calibrated years before the present (cal bp) (refs. 19,20).
Whereas some studies support a simultaneous appearance of the Neolithic in northwestern Africa (Eastern Rif, Ifri Oudadane site) and Iberia around 7,550 cal bp (ref. 21), the earliest evidence for pottery, domestic cereals and husbandry is found in northern Morocco approximately two centuries later at Kaf Taht el-Ghar (KTG) around 7,350 cal bp (refs. 2,3,22,23). Although Early Neolithic material culture and the first domestic mammals and pulses suggest a connection to Iberia1,2,3, the extent and legacy of these connections remain unclear. However, the first genomic analysis of Early Neolithic farmers from northwestern Africa (from the site Ifri n’Amr o’Moussa (IAM) in central Morocco) shows no traces of admixture with European Neolithic farmers. Instead, it shows long-term population continuity since the Upper Palaeolithic in the region6. This result aligns with the hypothesis that the Neolithic transition in northwestern Africa was initiated by local Epipalaeolithic communities adopting technological innovations4,5, such as those found at IAM: impressed Cardial-like ceramics, similar to those present throughout the western Mediterranean Neolithic Europe, and domestic cereals (for example, a grain of Hordeum vulgare dated around 7,050 cal bp)2. This pattern implies a Neolithization process that contrasts markedly with that of Europe, where it has been established that agriculture was introduced by the west- and northward demic diffusion of Anatolian early farmers11,12. The local development, or acculturation, of the North African Neolithic is further supported by signs of increasingly sedentary Epipalaeolithic groups developing strategies for resource management, such as the exploitation of wild plants and pottery1,4,24,25,26. Rapid climatic changes favoured mobile herding27 and, whereas it has been hypothesized that cattle were independently domesticated in the Sahara28, radiocarbon data suggest a gradual introduction of pastoralism in the Sahara in a southwestwards direction 7,000–6,000 cal bp, possibly from the Near East29,30.
Whereas palaeogenomic studies on the European Mediterranean Neolithic transition are abundant15,31,32,33, North Africa has been the focus of only a single study that generated human genetic data from one Early and one Late Neolithic site6, leaving substantial gaps in the chronology of events. It is evident that the site of IAM shows a Neolithic lifestyle and an absence of European Neolithic ancestry, but whether this was an independent development or the inspiration came from other groups in northwestern Africa or across the Mediterranean Sea remains unclear. Hence, the timeline and processes involved in the Neolithization of the region, the nature and dynamics of different economies in North Africa and the role they may have played in the broader European Neolithic remain understudied and controversial.
In this study we investigate a time series of human remains from four archaeological sites spanning the Epipalaeolithic to Middle Neolithic in current-day Morocco: the Epipalaeolithic site of Ifri Ouberrid (OUB), the Early Neolithic sites of IAM and KTG and the Middle Neolithic cemetery of Skhirat-Rouazi (SKH), co-analysed with previously published genetic data from that region6,10. By sequencing the genomes of nine individuals excavated from these four archaeological sites, we can demonstrate that the Neolithic transition in northwestern Africa was ignited by migration of Neolithic farmers from Mediterranean Europe.
We generated genomic sequence data from nine ancient individuals from modern-day Morocco (Table 1), ranging in genome coverage from 45.75- to 0.017-fold, including five individuals with more than onefold coverage and three with more than ninefold. Chronologically the data span more than 1,000 years, covering the Late Epipaleolithic (n = 1), Early Neolithic (n = 5) and Middle Neolithic (n = 3). Two Early Neolithic sites were studied—KTG (n = 4) and IAM—where we co-analysed the newly generated genomic data of one individual and those previously reported6 (Fig. 1a,b). DNA libraries were generated from DNA extracts obtained from bones and teeth and subsequently shotgun sequenced on an Illumina platform. All libraries presented the degradation patterns expected from ancient DNA, including short fragment sizes and cytosine deamination at read ends (Supplementary Fig. 1). Contamination estimates were generally low for both the nuclear genome and mitochondria except for individual skh003, which showed 10–16% nuclear contamination (Table 1). To assess the relationship of the ancient northwestern African individuals to other ancient and present-day West Eurasian and African populations, we co-analysed our data with relevant ancient (Supplementary Data 2) and current-day groups from Africa, the Middle East and Europe34.
a, Geographic location of investigated archaeological sites. Symbol legend given in c. The map was generated using the open source QGIS Geographic Information System, http://qgis.osgeo.org. b, Chronological representation of the investigated archaeological time periods of northwestern Africa, with each site’s radiocarbon-dated timeline indicated. c, Enlarged view of a PCA plot (Supplementary Fig. 3) with focus on the ancient individuals analysed. Each projected ancient individual is represented by a coloured symbol. W. Eur., West European; hist., historical. d, Estimated ancestry proportions for relevant African, Middle Eastern and European (Eur.) modern-day and ancient individuals (assuming five ancestry components; additional results are presented in Supplementary Fig. 4). Pre-Neolithic and Neolithic northwestern African populations/individuals are highlighted by the same symbols used in a and c.
Eight thousand years of population continuity
From the Upper Palaeolithic people of Taforalt (TAF) via the Epipalaeolithic at OUB to the Early Neolithic at IAM, we observe the persistence of the unique genetic make-up that existed in northwestern African inhabitants 15,000 years ago (Fig. 1c,d and Supplementary Fig. 5), and possibly even further back in time. The Epipalaeolithic individual oub002, dating to 7,660–7,506 cal bp, is genetically very similar to individuals from TAF (15,086–14,046 cal bp)35 and Early Neolithic individuals from IAM (7,316–6,679 cal bp; Fig. 1)6,36. The genome of Oub002 demonstrates a marked population continuity in northwest Africa with no substantial gene flow across the Mediterranean Sea for at least 7,000 years across the Epipalaeolithic (Fig. 1c,d), linking the Maghrebi genetic ancestry found in the Upper Palaeolithic to the Early Neolithic individuals at IAM.
The Maghrebi lineage shows outstandingly low genetic diversity6,10 (Fig. 2b and Supplementary Fig. 9) and long and frequent runs of homozygosity (RoH) (Fig. 2a), probably as a consequence of long-lasting isolation. By investigation of the 45.8-fold genome of oub002 we show that ancient northwestern Africans went through a severe population bottleneck. Until some 70,000–60,000 years ago the effective population size (Ne) changes of oub002 follow a pattern similar to that of Eurasian populations with a relatively small effective population size reached 50,000 years ago (Fig. 2c), which is consistent with the Maghrebi lineage being related to the populations that migrated out of Africa. Interestingly, modern-day Eurasians and North Africans, as well as Neolithic Eurasians effective population size remained at around 5,000 until about 30,000 years ago but the effective population size of the Maghrebi lineage continues to decrease and reached its lowest point (Ne ≈ 1,400) between 50,000 and 27,000 years ago during the peak of the Last Glaciation. Remarkably similar patterns are observed for the Mesolithic western hunter-gatherers (WHG) of Europe (represented by Loschbour in Fig. 2c), for which low diversity measures have been attributed to high levels of background relatedness and autozygosity due to small population size37.
a–c, Ancient individuals (including oub002, ktg004 and skh001) are compared with modern-day individuals from geographically corresponding regions. a, Runs of homozygosity. ne1, Neolithic European 1; NSEG, number of homozygote segments. b, Heterozygosity, calculated from the number of variable positions per individual divided by the number of single-nucleotide polymorphism (SNP) sites per individual. c, Effective population size over time, as inferred by pairwise sequentially Markovian coalescent, for three ancient northwestern Africans with over ninefold genome coverage, as well as for a Mesolithic European individual (Loschbour) and a Neolithic European individual (Stuttgart), and modern-day individuals for comparison.
European farmers induce Neolithization
At the site of IAM, a multitude of artefacts representing the Neolithic package have been identified. However, it has been shown that the people living at IAM show autochthonous Maghrebi ancestry6 and were the descendants of earlier (Upper Palaeolithic and Epipaleolithic) northwestern African groups (Fig. 1c,d). These two observations support the view that the first stage of the Neolithic transition in Morocco was driven by local populations adopting technological innovations based on contacts across the Mediterranean2.
The Early Neolithic site of KTG, located on the North African Mediterranean coast near the Gibraltar strait (Fig. 1a), predates and partly overlaps in time with IAM2 (Table 1). At KTG a full Neolithic assemblage is found, including a diversity of cultivated cereals, domestic mammals and cardial ceramics38,39. In contrast to the people at IAM, those at KTG are genetically similar to European Early Neolithic populations (Figs. 1c,d and 3a). Interestingly, all four KTG individuals show admixture (15.4–27.4%) with local North African groups (Fig. 1d), consistent with significantly positive values for the f4 test of admixture (KTG, Mediterranean EN; TAF, Mbuti) (Supplementary Data 7). Furthermore we identify a small proportion of WHG ancestry in KTG (Fig. 1d), consistent with the observation of Early Neolithic Europeans carrying WHG ancestry14,15,31,33,40. A population history model for the KTG people with 72 ± 4.4% Anatolian Neolithic ancestry, 10 ± 2.6% WHG ancestry and 18 ± 3.3% Maghrebi ancestry is consistent with the data (qpAdm, P = 0.193). Taken together, these results suggest a European Neolithic origin of KTG farmers whose ancestors dispersed from Anatolia throughout Europe, admixing with European hunter-gatherers on their path to southwestern Europe33,40 before crossing the Mediterranean to North Africa. The presence of European hunter-gatherer ancestry excludes the possibility that Early Neolithic migrants exclusively followed North African Mediterranean shores from Anatolia or the Levant.
a, Genetic affinity of analysed Stone Age northwestern African individuals, polarized between Neolithic Anatolia and Maghrebi ancestry, using the f4 test of the form f4(Anatolia Neolithic, TAF011; Stone Age northwestern African individuals, Mbuti). b, Genetic affinity of Stone Age northwestern African individuals, polarized between Early Neolithic Iberian and Levantine ancestry using f4 test of the form f4(Iberia Early Neolithic, Levant Chalcolithic; Neolithic northwestern African individuals, Mbuti). a,b, Each symbol represents a single individual f4 value. Error bars indicate ±2 s.e., computed with a block jack-knife approach (5 Mb blocks weighted by the number of SNPs). c, Summary of inferred population history of the Stone Age Maghreb.
Iberian Early Neolithic (both as a whole and regionally) was found to be the best source population for the European ancestry in KTG, followed by Sicily Stentinello Early Neolithic (Supplementary Data 9). This is consistent with low levels of genetic differentiation in Cardial Ware-associated groups along the European shores of the Mediterranean Sea41, confirmed by direct radiocarbon dates showing that Impressed Ware farmers expanded rapidly across the western Mediterranean3,19,42.
It has been debated whether European farmers crossed from Iberia to Morocco2,3 or whether earlier crossings of the Mediterranean would have happened, through the Sicilian–Tunisian Strait followed by a Maghrebi route of expansion4,43. Direct comparisons of Early Neolithic farmers from Sicily and Iberia as ancestors of KTG farmers provide stronger evidence for an Iberian Neolithic origin (Supplementary Data 9 and Supplementary Information 8), but we cannot exclude some contribution from Sicilian farmers. Genetic data are consistent with the most parsimonious explanation for archaeological evidence on the Neolithic transition in northwestern Africa: the crossing from southern Iberia by Iberian Neolithic farmers2,23. The close geographical proximity between southern Iberia and the Tangitana Peninsula adds strength to this observation whereas the lack of reliable archaeological evidence of early domestic elements in relevant sites along the eastern Maghreb and Tunis, including sites with pottery and obsidian from Pantelleria Island, undermines the Sicily–Tunis crossing hypothesis3. Interestingly, gene flow from North Africa was found only in Mediterranean European individuals much later, from around 4,500 years ago31,44.
Different individuals from KTG date to slightly different time periods. We find a twofold larger proportion of Maghrebi ancestry in earlier KTG individuals (roughly 25%, ktg001 and ktg005, approximately 7,429–7,267 cal bp) than in later ones (about 13%, ktg004 and ktg006, around 7,247–6,945 cal bp) (Fig. 1d). This coincides with an increase in European Neolithic ancestry, shown by the significantly negative result for f4(KTG earlier, KTG later, Iberia Early Neolithic, Mbuti; z-score = −5.01). Approximately one quarter of Maghrebi ancestry in early KTG suggests that they represent at least the second generation of interbreeding between the groups. We estimated the time of admixture using two approaches based on ancestry covariance patterns and linkage disequilibrium decay, using Iberia or Sicily Early Neolithic and TAF as admixture sources. Both methods date the contact within the last six to 13 generations (Supplementary Information 8), suggesting that mixing between groups occurred for a few hundred years, which is consistent with analysis of pottery style that points to the first contact at 7,500–7,400 cal bp (ref. 23).
Kaf Taht el-Ghar farmers had slightly lower genetic diversity levels and greater RoH than most Early Neolithic European populations (Fig. 2a,b and Supplementary Fig. 9). The Maghrebi ancestry carried by KTG people shows markedly lower diversity and more extensive RoH, and is probably the cause of the reduction in overall diversity. Archaeological evidence suggests that Early Neolithic farming was restricted to enclaves in westernmost Maghreb, possibly due to climatic constraints to the south4,22. This could have limited the potential of these groups to recover from an initial founder effect.
Overall, the genetic patterns of local interaction between different groups in northwestern Africa are comparable to those found in Europe: farmers assimilated local foragers’ ancestry in a unidirectional admixture process. Cases of hunter-gatherer communities adopting certain elements of the Neolithic have been described in Europe11,14,45. However, the northwestern Africa Neolithization process involved the notable survival of genetically unadmixed local populations (represented by IAM), despite coexisting for at least 300 years with foreign farming communities (KTG), and still adopted several elements of the Neolithic ways of living from them. Whereas the archaeological findings in IAM and KTG point to the exchange of ideas between groups and support an acculturation process of foraging communities1,4, our genetic data show that the exchange of genes was unidirectional.
Influx of Levantine ancestry
Another, distinct, ancestry was introduced to northwestern Africa during the Middle Neolithic. All individuals from SKH show large proportions of a genetic component maximized in individuals from Neolithic and Chalcolithic Levant, Ptolemaic Egypt and modern-day Near Eastern populations (Fig. 1d). The ancestry in SKH can be modelled as a two-way admixture between Levant Neolithic populations (roughly 76.4 ± 4.0%) and local northwestern Africans (represented by TAF; 23.6 ± 4.0%). If a European Neolithic (for example, from Iberia) additional source population is added, the model is rejected.
Because this Neolithic Levantine ancestry has not been observed on the European side of the Mediterranean during the Neolithic, it probably represents an independent expansion of people from the Levant into North Africa. Migrations from the Levant to eastern Africa have been identified for Neolithic pastoralist individuals around 4,000 years ago, who are presumed descendants of unsampled northeastern African populations associated with the spread of Saharan pastoralism46. Both in SKH and eastern African Neolithic pastoralists, Levantine ancestry is admixed with local ancestries (Fig. 1d, Supplementary Information 8 and Supplementary Data 12). The arrival of this Levantine ancestry coincides with the appearance of a new ceramic tradition in northern Morocco, often characterized by cord-impressed motifs (‘roulette’ or wavy line), like the grave goods at Skhirat belonging to Ashakar Ware pottery47,48. In parallel, cattle pastoralism was expanding in the current Sahara territory30,47 and Afro-Asiatic language groups spread throughout the whole of North Africa22.
Our analyses show that the Levantine-associated component also remains in the Maghreb during the Late Neolithic in individuals from Kehf el Baroud (KEB) and in the Guanches of the Canary Islands (around 1,000 cal bp; Fig. 1c,d)6,49. Individuals from these sites are shifted towards ancient Levantine populations on the principal component analysis (PCA) space (Fig. 1c). This highlights the complex demographic processes that took place in northwestern Africa, in contrast to the gradual increase in hunter-gatherer ancestry described in Middle and Late Neolithic Europe32,33,40.
The Late Neolithic individuals from KEB can be modelled as a mix of ancestries already present in northwestern Africa during the Early Neolithic and Middle Neolithic, suggesting that there were no waves of substantial migration into this region between the Middle Neolithic and Late Neolithic (Supplementary Information 8 and Supplementary Data 13).
Conclusion
The complex population structure in modern-day northwestern Africa has been linked to various historical events, such as the Arab expansion7,8. However, our detailed chronology and high-resolution genomic data provide a new understanding of these prehistoric processes in the Maghreb and unveil a rich and diversified genetic substrate with Neolithic origin. First, human populations in northwestern Africa show genetic continuity and isolation since the Upper Palaeolithic, from at least 15,000 to around 7,500 years ago, when this period of isolation was interrupted by the migration of European Early Neolithic groups introducing farming practices. Hence, despite a relatively small geographic distance between southern Iberia and northwestern Africa (the distance today is only 13 km across the Gibraltar straight), and the fact that both regions were populated by foragers for many millennia prior to the Neolithic, gene flow across the Mediterranean Sea was not established until the Early Neolithic. The newcomers brought new ways of life, farming practices, domestication and pottery traditions that were subsequently adopted by local populations. Our results show that the Neolithization process in northwestern Africa was ignited by migrant Neolithic Europeans, but that local groups (at least the individuals analysed at IAM) adopted some of these practices without mixing with the newcomers. Two genetically distinct groups coexisted in close proximity in the region. Interestingly, cultural and technological knowledge appear to have been transferred mainly from European Neolithic farmers to local groups (for example, at IAM) whereas genetic ancestry flowed only from local groups to the incoming farmers, such as the population of KTG. Furthermore, in the Middle Neolithic a new ancestry with an eastern origin is detected in northwestern Africa. This ancestry indicates new migrating groups, potentially associated with Sahara pastoralists, which admixed with local groups (Fig. 3c).
The various waves of migration and admixture into northwestern Africa during the Neolithic possibly resulted in a heterogeneous economic and cultural landscape in that region—a mosaic of groups that included incoming farmers from Iberia, foragers adopting farming practices and eastern pastoralists admixing with local people. Most of these groups showed reduced effective population size and lower diversity than the contemporary populations in Europe (Fig. 2), suggesting that population sizes remained modest throughout the Neolithic. These patterns were probably caused by periods of isolation, which may have contributed to the distinct genetic ancestry seen in the Maghreb today. A recent study from the Iron Age suggests that northwestern Africa remained home to a diverse set of groups throughout prehistory50, making this part of the world one of the most unique places to have been studied with the archaeogenomic toolkit.
Methods
Detailed descriptions for each section can be found in Supplementary Information.
Archaeological sampling
The ancient human remains analysed in this study derive from a scientific cooperation agreement between INSAP, La Trobe and Uppsala Universities. Complete bone and teeth elements were brought to the ancient DNA facility in Uppsala, Sweden for further cleaning and sampling.
Radiocarbon dating
All individuals investigated were directly radiocarbon dated at the Tandem Laboratory, Uppsala, except for ktg001, who was dated at the Beta Analytic Carbon dating laboratory, and iam004, who’s date was obtained from ref. 6. Radiocarbon calibration for newly reported and relevant previously published dates was performed using Oxcal v.4.4 and the IntCal20 dataset41.
Ancient DNA retrieval
Human remains were sampled in dedicated clean-room laboratories at Uppsala University, Sweden after a series of stringent procedures aimed at minimization of bone and tooth surface contamination. Thirty to sixty milligrams of bone powder or solid pieces of bone material were used for DNA extraction either following ref. 51, with adaptations as described in ref. 15, or following ref. 52, with adaptations to the binding buffer, and an initial predigestion step with 1 ml of 0.5 M EDTA pH 8.0 for 30 min at 37 °C53. Sample digestion was performed overnight with 1 ml of 0.45 M EDTA pH 8.0 and 0.2 mg ml–1 proteinase K. Double-stranded, blunt-end-repaired DNA libraries were built with ligated P5 and P7 adaptors54. After assessment of DNA authenticity, quality and quantity (by estimation of endogenous DNA content, post mortem deamination patterns and fragment size distribution), the remaining DNA extract (for samples with over 1% proportion of human DNA) was used to build four to six additional double-stranded DNA libraries; for extracts with roughly 5% endogenous human content or more, 15–20 µl of DNA extract was treated with uracil DNA glycosylase (UDG) for double-stranded library building55. Libraries were PCR amplified using a unique 7 bp indexed primer54,56 in either four reactions of 25 µl or two of 50 µl, with the application of 12–20 PCR cycles depending on previous qPCR quantification cycle indication. Two extraction negative controls, two library negative controls and one PCR negative control were included per sample batch. PCR reactions were pooled and purified with AMPure XP beads (Agencourt, Beckman Coulter). Library quality was checked by electrophoresis on Tapestation (Agilent High Sensitivity D1000 ScreenTape, Agilent) and DNA concentration was quantified using a Qubit dsDNA HS (High Sensitivity) Assay Kit (Invitrogen). Equimolar pools of amplified and purified libraries were sequenced on Illumina HiSeq X at the SNP & SEQ Technology Platform in Uppsala. To reach higher coverage, between four and ten libraries were pooled equimolarly and sequenced to depletion.
Bioinformatics data processing and authentication
Data were demultiplexed according to the indexed primer sequence and adaptors were trimmed with either MergeReadsFastQ_cc.py57 or Adapter Removal v.2.1.7 (ref. 58). Forward and reverse paired-end reads were merged when an overlap of at least 11 bp was found. Mapping against the human reference genome build 37 (hs37d5) was done using Burrows–Wheeler aligner 0.7.13 (ref. 59). For each library we merged bam files resulting from all resequencing rounds using SAMtools merge v.1.5 (ref. 60). We then separately merged data from UDG-treated and untreated libraries for each individual and used data from the former for subsequent analysis, except for individuals ktg001 (for which only non-UDG data were generated) and iam004 (for which both treated and untreated data were merged and processed as non-UDG treated for downstream analysis). We used a modified version of FilterUniqSAMCons_cc.py57 to ensure random choice of bases to collapse reads with identical start and end positions into a consensus, thereby removing PCR duplicates. Reads shorter than 35 bp and more than 10% mismatches to the human reference genome were filtered out.
Contamination, sex determination, uniparental markers and kinship analyses
Sample contamination estimates were obtained using three different methods based on the mitochondrial genome61, on the X chromosome in males62 and on nuclear data63 (Supplementary Data 3). The ratio of coverage of X and Y chromosomes relative to autosomes was used to determine the biological sex of each individual64. We generated mitochondrial consensus sequences using SAMtools 1.5 mpileup and vcfutils.pl60,65. Base (BQ) and mapping quality (MAPQ) scores were set to MAPQ > 30 and BQ > 30, and only sites with at least threefold coverage were used. Haplogroups were assigned using Haplogrep 2.1.16 (ref. 66) and PhyloTree mtDNA tree Build 17 (18 February 2016)67 (Supplementary Data 4). For Y haplogroup inference we called SNPs from the International Society of Genetic Genealogy (http://isogg.org; v.11.110, 21 April 2016)) from bam files using SAMtools mpileup with option -B. We extracted sites with mapping and base quality greater than 30. Insertions and deletions, and sites showing multiple alleles, were excluded (Supplementary Data 5).
We ran kinship analysis with READ68 within each archaeological site (minimum of three individuals; Supplementary Fig. 2). When a pair of individuals with close kinship was found, such as first-degree relationships (parent–offspring or a full sibling), we excluded the individual with fewer SNPs covered from the analyses. This resulted in the removal from the analysis of iam4 (same individual as iam5), keb8 (same individual as keb1), iam6 (first-degree relative to iam004)6 and TAF012 (first-degree relative to TAF011)10.
Population genomics analysis of pseudohaploid data
Data from over 300 ancient Eurasian, North African and Sub-Saharan African individuals, organized according to geography and chronology (Supplementary Data 2), were downloaded, mapped and processed though the same pipeline as used for newly generated data. The full ancient DNA dataset was merged with publicly accessible modern-day individuals sampled across the globe from the Simons Genome Diversity Project (SGDP) dataset34 for a 2.2 million SNP panel64. Alleles were sampled from bam files by randomly drawing one read with MAPQ > 30 and BQ > 30 per SNP site for each ancient individual (using SAMtools v.1.5.0 mpileup with option -B), and that position was treated as (pseudo)haploid. For non-UDG-treated data (ktg001) or merged UDG and non-UDG data (iam004) we trimmed off 10 bp of sequence-ends to avoid integration of miscoding C-to-T and G-to-A substitutions. For the published partial UDG-treated data (UDG-half), 2 bp were trimmed off the sequence-ends. SNPs showing more than two alleles were excluded from the data, leaving 1,379,466 SNPs for analysis.
Principal component analysis was performed using smartpca v.10210 (ref. 69). Principal components were calculated based on individuals from 18 Mediterranean Eurasian or North African modern-day populations from SGDP. Ancient individuals were projected onto the PCA space with options shrinkmode: YES and lsqproject: YES. An unsupervised model-based clustering algorithm, implemented in ADMIXTURE v,1.3.0 (ref. 70), was performed for K = 3–5 (30 runs) on a fully pseudohaploidized, linkage disequilibrium-pruned dataset of modern-day and ancient individuals from Mediterranean Eurasian or North African populations, leaving 812,092 SNPs for analysis. The results were parsed, aligned and plotted with pong71.
Popstats72 was used to calculate f-statistics73, with Mbuti set as the outgroup (Supplementary Data 6 and 7). Outgroup-f3 statistics were computed with the option –f3vanilla. Standard errors (SEs) were calculated with a weighted block jack-knife approach.
Admixture modelling was performed with qpAdm74 using ADMIXTOOLS v.5.0, through an adapted version of qpAdm_wrapper (https://github.com/pontussk/qpAdm_wrapper) that cycles through all possible subsets of the list of source populations provided (selected based on previous results), to test one-, two-, three- and four-way admixture models. SEs were computed with 5cM block jack-knife. We used a set of 11 reference populations whose power to disentangle divergent strains of ancestry present in Europe, North Africa and the Near East has previously been described and that are differently related to the sources tested10,31,75. Distantly related sources were explored and, where possible, also more proximate groups (geographically, chronologically or according to standing archaeological evidence). We tried to find the most parsimonious models consistent with the data (P > 0.05) by checking the lowest possible number of ancestry sources necessary to explain the ancestry in each test population (Supplementary Data 8–13). The Admixture event in KTG was dated using ALDER76 and DATES77 (Supplementary Information 8). We calculated conditional nucleotide diversity78 by estimation of the average number of mismatches between two individuals of the same population. SEs were estimated using a block jack-knife approach and a block size of 2,000 SNPs (Supplementary Fig. 9).
Population genomics analysis of diploid data
Diploid genotype calls for a panel of 49,791,572 SNPs were performed for northwestern African ancient individuals with at least ninefold genome coverage (oub002, ktg004 and skh001), as well as relevant, previously published ancient individuals with sequenced high-coverage genomes. Before genotype calling, base quality in read ends was reduced and indel realignment conducted with GATK 3.5.0. Diploid genotypes were called using dbSNP v.142 as known SNPs, with GATK’s UnifiedGenotyper79. We computed average sequencing depth (avg.DP) over all called positions for each individual and filtered for QUAL > 30 and a depth span from fivefold to 3× avg.DP per individual, using BCFtools view. This dataset was merged with data from modern-day individuals from the SGDP dataset.
Individual heterozygosity was calculated from the number of variable positions divided by that of sequenced SNPs, using the –het command in PLINK 1.9 (ref. 80). We estimated the length and number of runs of homozygosity after filtering with the command PLINK –geno 0. MSMC81 input files were generated from VCF files. Filters for MAPQ > 30, minimum genotype quality of 50 and sequencing depth were used. Sites not passing these filters were masked out per individual. MSMC 0.1.0 was then run for each individual.
Ethics and inclusion statement
The sampling for this study emerged from archaeology projects that involved local universities and researchers, including Y.B., whose involvement in research design included the selection of archaeological material for analyses as well as sampling supervision. The local relevance of this research is tied to the region’s history, and it is locally relevant in regard to describing the human past in northwestern Africa. The study was undertaken with the highest standards of archaeogenomic research, and relevant research by local scholars was cited.
Reporting summary
Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.
Data availability
The sequence data generated for this study are available from the European Nucleotide Archive under accession no. PRJEB59008.
References
Linstädter, J., Medved, I., Solich, M. & Weniger, G.-C. Neolithisation process within the Alboran territory: models and possible African impact. Quat. Int. 274, 219–232 (2012).
Martínez-Sánchez, R. M., Vera-Rodríguez, J. C., Pérez-Jordà, G., Peña-Chocarro, L. & Bokbot, Y. The beginning of the Neolithic in northwestern Morocco. Quat. Int. 470, 485–496 (2018).
Zilhão, J. Early prehistoric navigation in the Western Mediterranean: implications for the Neolithic transition in Iberia and the Maghreb. Isl. Archaeol. Orig. Seafaring East. Mediterr. 11, 185–200 (2014).
Mulazzani, S. et al. The emergence of the Neolithic in North Africa: a new model for the Eastern Maghreb. Quat. Int. 410, 123–143 (2016).
Linstädter, J. The Epipalaeolithic-Neolithic-Transition in the Mediterranean region of Northwest Africa. Quartär. International Yearbook for Ice Age and Stone Age Research 55, 41–62 (2008).
Fregel, R. et al. Ancient genomes from North Africa evidence prehistoric migrations to the Maghreb from both the Levant and Europe. Proc. Natl Acad. Sci. USA 115, 6774–6779 (2018).
Henn, B. M. et al. Genomic ancestry of North Africans supports back-to-Africa migrations. PLoS Genet. 8, e1002397 (2012).
Arauna, L. R. et al. Recent historical migrations have shaped the gene pool of Arabs and Berbers in North Africa. Mol. Biol. Evol. 34, 318–329 (2017).
Hublin, J.-J. et al. New fossils from Jebel Irhoud, Morocco and the pan-African origin of Homo sapiens. Nature 546, 289–292 (2017).
van de Loosdrecht, M. et al. Pleistocene North African genomes link Near Eastern and sub-Saharan African human populations. Science 360, 548–552 (2018).
Skoglund, P. et al. Origins and genetic legacy of Neolithic farmers and hunter-gatherers in Europe. Science 336, 466–469 (2012).
Omrak, A. et al. Genomic evidence establishes Anatolia as the source of the European Neolithic gene pool. Curr. Biol. 26, 270–275 (2016).
Antonio, M. L. et al. Ancient Rome: a genetic crossroads of Europe and the Mediterranean. Science 366, 708–714 (2019).
Yu, H. et al. Genomic and dietary discontinuities during the Mesolithic and Neolithic in Sicily. iScience 25, 104244 (2022).
Günther, T. et al. Ancient genomes link early farmers from Atapuerca in Spain to modern-day Basques. Proc. Natl Acad. Sci. USA 112, 11917–11922 (2015).
Barnett, W. K. Cardial pottery and the agricultural transition inMediterranean Europe. In Europe’s First Farmers 93–116 (Cambridge Univ. Press, 2000).
Manen, C. et al. The Neolithic transition in the Western Mediterranean: a complex and non-linear diffusion process—the radiocarbon record revisited. Radiocarbon 61, 531–571 (2019).
Natali, E. & Forgia, V. The beginning of the Neolithic in Southern Italy and Sicily. Quat. Int. 470, 253–269 (2018).
Guilaine, J. A personal view of the neolithisation of the Western Mediterranean. Quat. Int. 470, 211–225 (2018).
Bernabeu Auban, J. & Pardo-Gordó, S. La impressa en la península Ibérica: ¿Espejismo o realidad? Una reflexión a partir del binomio radiocarbono-cerámica (2020)
Linstädter, J., Broich, M. & Weninger, B. Defining the Early Neolithic of the Eastern Rif, Morocco–spatial distribution, chronological framework and impact of environmental changes. Quat. Int. 472, 272–282 (2018).
Broodbank, C. & Lucarini, G. The dynamics of Mediterranean Africa, ca. 9600–1000 bc: an interpretative synthesis of knowns and unknowns. J. Mediterr. Archaeol. https://doi.org/10.17863/CAM.49028 (2020).
Martínez-Sánchez, R. M. et al. Reflections on the other side. A Southern Iberia origin for the first pottery production of Northern Morocco? Open Archaeol. 7, 1054–1065 (2021).
Huysecom, E. et al. The emergence of pottery in Africa during the tenth millennium cal BC: new evidence from Ounjougou (Mali). Antiquity 83, 905–917 (2009).
Garcea, E. A. A. Semi-permanent foragers in semi-arid environments of North Africa. World Archaeol. 38, 197–219 (2006).
Dunne, J., Mercuri, A. M., Evershed, R. P., Bruni, S. & di Lernia, S. Earliest direct evidence of plant processing in prehistoric Saharan pottery. Nature Plants 3, 16194 (2016).
Marshall, F. & Hildebrand, E. Cattle before crops: the beginnings of food production in Africa. J. World Prehist. 16, 99–143 (2002).
Gautier, A. In Droughts, Food and Culture: Ecological Change and Food Security in Africa’s Later Prehistory (ed. Hassan, F. A.) 195–207 (Springer US, 2002); https://doi.org/10.1007/0-306-47547-2_12.
David, B. & McNiven, I. J. The Oxford Handbook of the Archaeology and Anthropology of Rock Art (Oxford Univ. Press, 2018).
Smith, A. B. Origins and spread of pastoralism in Africa. Annu. Rev. Anthropol. 21, 125–141 (1992).
Olalde, I. et al. The genomic history of the Iberian Peninsula over the past 8000 years. Science 363, 1230–1234 (2019).
Valdiosera, C. et al. Four millennia of Iberian biomolecular prehistory illustrate the impact of prehistoric migrations at the far end of Eurasia. Proc. Natl Acad. Sci. USA 115, 3428–3433 (2018).
Villalba-Mouco, V. et al. Survival of Late Pleistocene hunter-gatherer ancestry in the Iberian Peninsula. Curr. Biol. 29, 1169–1177 (2019).
Mallick, S. et al. The Simons Genome Diversity Project: 300 genomes from 142 diverse populations. Nature 538, 201–206 (2016).
Humphrey, L. T. et al. Earliest evidence for caries and exploitation of starchy plant foods in Pleistocene hunter-gatherers from Morocco. Proc. Natl Acad. Sci. USA 111, 954–959 (2014).
Turek, J. & Vintr, J. Neolit Maghrebu ve světle nových radiokarbonových dat. Živá Archeol. 18, 10–15 (2016).
Ringbauer, H., Novembre, J. & Steinrücken, M. Parental relatedness through time revealed by runs of homozygosity in ancient DNA. Nat. Commun. 12, 5425 (2021).
Morales, J. et al. The introduction of South-Western Asian domesticated plants in North-Western Africa: an archaeobotanical contribution from Neolithic Morocco. Quat. Int. 412, 96–109 (2016).
Martínez Sánchez, R. M. et al. Revisiting the Epipalaeolithic-Neolithic Transition in the extreme NW of Africa: the latest results of the chronological sequence of the Cave of Kaf Taht el-Ghar (Tétouan, Morocco). Afr. Archaeol. Rev. 38, 251–274 (2021).
Lipson, M. et al. Parallel palaeogenomic transects reveal complex genetic history of early European farmers. Nature 551, 368–372 (2017).
Reimer, P. J. et al. The IntCal20 Northern Hemisphere radiocarbon age calibration curve (0–55 cal kBP). Radiocarbon 62, 725–757 (2020).
Martins, H. et al. Radiocarbon dating the beginning of the Neolithic in Iberia: new results, new problems. J. Mediterr. Archaeol. 28, 105–131 (2015).
García Borja, P., Aura Tortosa, J. E., Bernabeu Aubán, J. & Jordá Pardo, J. F. Nuevas perspectivas sobre la neolitización en la cueva de Nerja (Málaga-España): la cerámica de la sala del vestíbulo. Zephyrvs https://revistas.usal.es/uno/index.php/0514-7336/article/view/7979/8431 (2010).
Marcus, J. H. et al. Genetic history from the Middle Neolithic to present on the Mediterranean island of Sardinia. Nat. Commun. 11, 939 (2020).
Mathieson, I. et al. The genomic history of southeastern Europe. Nature 555, 197–203 (2018).
Prendergast, M. E. et al. Ancient DNA reveals a multistep spread of the first herders into sub-Saharan Africa. Science 365, eaaw6275 (2019).
Martínez Sánchez, R. M. et al. The Middle Neolithic of Morocco’s North-Western Atlantic Strip: new evidence from the El-Khil Caves (Tangier). Afr. Archaeol. Rev. 35, 417–442 (2018).
Gilman, A. The Later Prehistory of Tangier, Morocco (American School of Prehistoric Research, Peabody Museum of Archaeology and Ethnology, Harvard Univ., 1975).
Rodríguez-Varela, R. et al. Genomic analyses of pre-European conquest human remains from the Canary Islands reveal close affinity to modern North Africans. Curr. Biol. 27, 3396–3402 (2017).
Moots, H. M. et al. A genetic history of continuity and mobility in the Iron Age Central Mediterranean. Preprint at bioRxiv https://doi.org/10.1101/2022.03.13.483276 (2022).
Yang, D. Y., Eng, B., Waye, J. S., Dudar, J. C. & Saunders, S. R. Improved DNA extraction from ancient bones using silica-based spin columns. Am. J. Phys. Anthropol. 105, 539–543 (1998).
Dabney, J. et al. Complete mitochondrial genome sequence of a Middle Pleistocene cave bear reconstructed from ultrashort DNA fragments. Proc. Natl Acad. Sci. USA 110, 15758–15763 (2013).
Svensson, E. et al. Genome of Peştera Muierii skull shows high diversity and low mutational load in pre-glacial Europe. Curr. Biol. 31, 2973–2983 (2021).
Meyer, M. & Kircher, M. Illumina sequencing library preparation for highly multiplexed target capture and sequencing. Cold Spring Harb. Protoc. 2010, pdb-prot5448 (2010).
Günther, T. et al. Population genomics of Mesolithic Scandinavia: investigating early postglacial migration routes and high-latitude adaptation. PLoS Biol. 16, e2003703 (2018).
Gansauge, M.-T. & Meyer, M. Single-stranded DNA library preparation for the sequencing of ancient or damaged DNA. Nat. Protoc. 8, 737–748 (2013).
Kircher, M. Analysis of high-throughput ancient DNA sequencing data. Methods Mol. Biol. 840, 197–228.
Schubert, M., Lindgreen, S. & Orlando, L. AdapterRemoval v2: rapid adapter trimming, identification, and read merging. BMC Res. Notes 9, 88 (2016).
Li, H. & Durbin, R. Fast and accurate short read alignment with Burrows–Wheeler transform. Bioinformatics 25, 1754–1760 (2009).
Li, H. et al. The sequence alignment/map format and SAMtools. Bioinformatics 25, 2078–2079 (2009).
Fu, Q. et al. A revised timescale for human evolution based on ancient mitochondrial genomes. Curr. Biol. 23, 553–559 (2013).
Rasmussen, M. et al. An Aboriginal Australian genome reveals separate human dispersals into Asia. Science 334, 94–98 (2011).
Jun, G. et al. Detecting and estimating contamination of human DNA samples in sequencing and array-based genotype data. Am. J. Hum. Genet. 91, 839–848 (2012).
Fu, Q. et al. The genetic history of ice age Europe. Nature 534, 200–205 (2016).
Danecek, P. et al. The variant call format and VCFtools. Bioinformatics 27, 2156–2158 (2011).
Weissensteiner, H. et al. HaploGrep 2: mitochondrial haplogroup classification in the era of high-throughput sequencing. Nucleic Acids Res. 44, W58–W63 (2016).
Van Oven, M. PhyloTree Build 17: growing the human mitochondrial DNA tree. Forensic Sci. Int. Genet. Suppl. Ser. 5, e392–e394 (2015).
Monroy Kuhn, J. M., Jakobsson, M. & Günther, T. Estimating genetic kin relationships in prehistoric populations. PLoS ONE 13, e0195491 (2018).
Patterson, N., Price, A. L. & Reich, D. Population structure and eigenanalysis. PLoS Genet. 2, e190 (2006).
Alexander, D. H., Novembre, J. & Lange, K. Fast model-based estimation of ancestry in unrelated individuals. Genome Res. 19, 1655–1664 (2009).
Behr, A. A., Liu, K. Z., Liu-Fang, G., Nakka, P. & Ramachandran, S. Pong: fast analysis and visualization of latent clusters in population genetic data. Bioinformatics 32, 2817–2823 (2016).
Skoglund, P. et al. Genetic evidence for two founding populations of the Americas. Nature 525, 104–108 (2015).
Patterson, N. et al. Ancient admixture in human history. Genetics 192, 1065–1093 (2012).
Haak, W. et al. Massive migration from the steppe was a source for Indo-European languages in Europe. Nature 522, 207–211 (2015).
Harney, É. et al. Ancient DNA from Chalcolithic Israel reveals the role of population mixture in cultural transformation. Nat. Commun. 9, 3336 (2018).
Loh, P.-R. et al. Inferring admixture histories of human populations using linkage disequilibrium. Genetics 193, 1233–1254 (2013).
Narasimhan, V. M. et al. The formation of human populations in South and Central Asia. Science 365, eaat7487 (2019).
Skoglund, P. et al. Genomic diversity and admixture differs for Stone-Age Scandinavian foragers and farmers. Science 344, 747–750 (2014).
DePristo, M. A. et al. A framework for variation discovery and genotyping using next-generation DNA sequencing data. Nat. Genet. 43, 491–498 (2011).
Purcell, S. et al. PLINK: a tool set for whole-genome association and population-based linkage analyses. Am. J. Hum. Genet. 81, 559–575 (2007).
Schiffels, S. & Durbin, R. Inferring human population size and separation history from multiple genome sequences. Nat. Genet. 46, 919–925 (2014).
Acknowledgements
We thank A. R. Munters for bioinformatics support and R. P. Stjerna and F. Sánchez-Quinto for helpful discussions on data analysis. Sequencing was performed at The National Genomics Infrastructure, Uppsala. Computations and data handling were enabled by resources provided by the Swedish National Infrastructure for Computing at the Uppsala Multidisciplinary Center for Advanced Computational Science, partially funded by the Swedish Research Council through grant agreement no. 2018-05973. This project was supported by the Knut and Alice Wallenberg Foundation (to M.J.), Vetenskapsrådet (grant nos. 2018-05537 and 2022-04642 to M.J. and 2017-05267 to T.G.) and Ramón y Cajal (grant no. RYC2018-025223-I to C.V.). The Spanish–Moroccan archaeological team was supported by the European Research Council (no. ERC AdG 230561).
Funding
Open access funding provided by Uppsala University.
Author information
Authors and Affiliations
Contributions
C.V., J.C.V.-R. and M.J. conceived the study. R.M.-S., J.C.V.-R., E.I., R.R.-V., Y.B. and C.V. selected and sampled archaeological material. L.G.S. performed DNA laboratory work. R.M.-S., J.C.V.-R. and C.V. provided archaeological interpretations. L.G.S., T.G. and M.J. analysed genetic data. L.G.S., T.G., R.M.-S., J.C.V.-R., C.V. and M.J. wrote the paper, with input from all authors.
Corresponding authors
Ethics declarations
Competing interests
The authors declare no competing interests.
Peer review
Peer review information
Nature thanks the anonymous reviewers for their contribution to the peer review of this work.
Additional information
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Supplementary information
Supplementary Information
A description of the northwestern African archaeological sites studied and a detailed description of the methods used for analyses, supplementary results and supplementary discussions for these results.
Supplementary Data
Supplementary Data 1–15.
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.
About this article
Cite this article
Simões, L.G., Günther, T., Martínez-Sánchez, R.M. et al. Northwest African Neolithic initiated by migrants from Iberia and Levant. Nature 618, 550–556 (2023). https://doi.org/10.1038/s41586-023-06166-6
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1038/s41586-023-06166-6
- Springer Nature Limited
This article is cited by
-
Modelling the demographic history of human North African genomes points to a recent soft split divergence between populations
Genome Biology (2024)
-
Understanding the genomic heterogeneity of North African Imazighen: from broad to microgeographical perspectives
Scientific Reports (2024)
-
Improving data archiving practices in ancient genomics
Scientific Data (2024)
-
Isotopic evidence of high reliance on plant food among Later Stone Age hunter-gatherers at Taforalt, Morocco
Nature Ecology & Evolution (2024)
-
Human genetic structure in Northwest France provides new insights into West European historical demography
Nature Communications (2024)