MAIZE DOMESTICATION

Multiproxy evidence highlightsa complex evolutionary legacyof maize in South AmericaLogan Kistler1,2*, S. Yoshi Maezumi3,4, Jonas Gregorio de Souza3,Natalia A. S. Przelomska1,5, Flaviane Malaquias Costa6, Oliver Smith7,Hope Loiselle1,8, Jazmín Ramos-Madrigal7, Nathan Wales9, Eduardo Rivail Ribeiro1,Ryan R. Morrison2, Claudia Grimaldo10, Andre P. Prous11, Bernardo Arriaza12,M. Thomas P. Gilbert7,13, Fabio de Oliveira Freitas14*, Robin G. Allaby2*

Domesticated maize evolved from wild teosinte under human influences in Mexico beginningaround 9000 years before the present (yr B.P.), traversed Central America by ~7500 yr B.P.,and spread into South America by ~6500 yr B.P. Landrace and archaeological maize genomesfrom South America suggest that the ancestral population to South American maize wasbrought out of the domestication center in Mexico and became isolated from the wild teosintegene pool before traits of domesticated maize were fixed. Deeply structured lineages thenevolved within South America out of this partially domesticated progenitor population.Genomic, linguistic, archaeological, and paleoecological data suggest that the southwesternAmazon was a secondary improvement center for partially domesticatedmaize. Multiple wavesof human-mediated dispersal are responsible for the diversity and biogeography of modernSouth American maize.

Maize (Zea mays ssp. mays) evolvedfrom wild Balsas teosinte (Z. maysssp. parviglumis, hereafter parviglumis)inmodern-day lowlandMexico beginningaround 9000 years ago (1) and spread

to dominate food production systems through-out much of the Americas by the beginning ofEuropean colonization in the 15th century.Archaeological and genetic data from ancientDNA studies have highlighted aspects of maizenatural history, including the evolution and fixa-tion of agricultural traits and adaptation of maizeto diverse new environments (2–6). Archaeological

remains establish thatmaize was brought to thesouthwestern United States and the ColoradoPlateau by ~4000 years before the present (yr B.P.)(7), traversing Panama by ~7500 yr B.P. (8) andarriving in Coastal Peru (9), the Andes (10), andlowland Bolivian Amazon (11) between ~6500and 6300 yr B.P. (Fig. 1 and table S1). Today,maize is a staple food species, yielding over 6%of all food calories for humans, plus more inlivestock feed and processed foods (12).Maize domestication is thought to have oc-

curred once, with little subsequent gene flowfrom parviglumis (13, 14). However, archaeoge-nomic evidence reveals maize was only partiallydomesticated in Mexico by ~5300 yr B.P. (2, 3),carrying a mixture of wild-type and maize-likealleles at loci involved in the domestication syn-drome. For example, the domestic-type TGA1gene variant responsible for eliminating the toughteosinte fruitcase was already present by this timeperiod (2), whereas other loci associated withchanges to seed dispersal and starch productionduring domestication still carried wild-type var-iants (2, 3). The state of partial domesticationsets these archaeogenomes apart from modernfully domesticated maize, which carries a com-plete, stable set of domestication alleles con-ferring the domesticated phenotype. This partiallydomesticated maize was grown in Mexico wellafter maize had become established in SouthAmerica, which raises the question of how SouthAmerican maize came to possess the full com-plement of fixed domestication traits. To reconcilearchaeobotanical and genomic data concerningthe domestication and dispersal history of maizein South America, we sequenced maize genomesfrom 40 indigenous landraces and 9 archaeo-logical samples from South America (Fig. 1 and

tables S2 and S3) and analyzed them alongsidepublished modern (n = 68) and ancient (n = 2)maize and teosinte genomes (15).Model-based clustering highlights extensive

admixture and population overlap betweenmaizepopulations, but we observe several robust lin-eages (15) (Fig. 1): (i) the Andes and the Pacificcoast of South America; (ii) lowland SouthAmerica, including the Amazon and BrazilianSavanna; (iii) North America north of the do-mestication center; and (iv) highlandMexico andCentral America, previously observed to containintrogression from wild Z. mays ssp.mexicana(14, 16). We also observe a widespread “Pan-American” lineage spanning fromnorthernMexicointo lowland South America. In a previous analysisbased on multiple nuclear microsatellites, maizeformed a monophyletic subset of teosinte, withSouth American lineages as the most derivedelements in a phylogenetic tree (13). This patternhas been interpreted as evidence for a singleepisode of domestication followed by dispersalculminating in the Andes after maize becameestablished throughout the rest of the rangeof cultivation (13). However, archaeologicalevidence for persistent maize cultivation in-dicates it was established in numerous loca-tions throughout South America by ~6500 to4000 yr B.P. regionally. On the basis of thisinformation, we propose that South Americanmaizewas carried away from theMesoamericandomestication center soon after initial stagesof domestication and may have been one ofseveral partially domesticated maize lineagesthat independently fissioned from the primarygene pool after the onset of domestication inMexico (Fig. 2).Using f4 statistics (17), we observe asymmetry

in parviglumis ancestry among modern maizepopulations (Fig. 2). This reveals that maize-parviglumis gene flow was ongoing in somelineages after others became reproductivelyisolated. Whereas later gene flow from Z. maysssp.mexicana, a highland subspecies of teosinte,is well documented in somemaize (6, 14, 16), thisfinding contradicts the assumption that dis-persal and diversification throughout the Americashappened only after the severance of gene flowfrom parviglumis (13, 14). Thus, while SouthAmerican maize became reproductively isolatedfrom the wild progenitor when it was carriedaway from the domestication center, maize lin-eages remaining inMexico underwent continuedcrop-wild gene flowbefore diversifying into extantlandraces over subsequent millennia. The Pan-American lineage shows excess shared ancestrywith parviglumis relative to all othermajor groups(Fig. 2B), suggesting that this group emerged fromthe domestication center and dispersed after othermaize lineages became regionally established.Because the Pan-American lineage carries excessparviglumis ancestry relative to the strictly SouthAmerican lineages, it appears to represent asecond episode of maize dispersal from Meso-america, reinforcing two major waves of maizemovement into South America as previouslysuggested (5).

RESEARCH

Kistler et al., Science 362, 1309–1313 (2018) 14 December 2018 1 of 4

1Department of Anthropology, National Museum of NaturalHistory, Smithsonian Institution, Washington, DC 20560,USA. 2Department of Life Science, University of Warwick,Coventry CV4 7AL, UK. 3Department of Archaeology, Collegeof Humanities, University of Exeter, Laver Building, NorthPark Road, Exeter EX4 4QE, UK. 4Department of Geographyand Geology, The University of the West Indies, MonaCampus, Kingston, Jamaica. 5Center for ConservationGenomics, Smithsonian Conservation Biology Institute,National Zoo, Washington, DC 20008, USA. 6University ofSão Paulo, Escola Superior de Agricultura Luis de Queiroz,Piracicaba, SP 13418-900, Brazil. 7Centre for GeoGenetics,Natural History Museum of Denmark, University ofCopenhagen, Øster Voldgade 5-7, 1350 Copenhagen,Denmark. 8Department of Anthropology, University ofWashington, Denny Hall 314, Seattle, WA 98195, USA.9Department of Archaeology, University of York, King'sManor, York YO1 7EP, UK. 10Department of Oncology,University of Oxford, Old Road Campus Research Building,Roosevelt Drive, Oxford, OX3 7DQ, UK. 11Museu de HistoriaNatural e Jardim Botânico da Universidade Federal de MinasGerais, Belo Horizonte, MG 31270-901, Brazil. 12Instituto deAlta Investigación, Universidad de Tarapacá, Arica, Chile.13Norwegian University of Science and Technology,University Museum, 7491 Trondheim, Norway. 14EmbrapaRecursos Genéticos e Biotecnologia, Brasília, DF, CEP70770-901, Brazil.*Corresponding author. Email: kistlerl@si.edu (L.K.); fabio.freitas@embrapa.br (F.O.F.); r.g.allaby@warwick.ac.uk (R.G.A.)

on July 19, 2020

http://science.sciencemag.org/

Dow

nloaded from

The genomes of two ancient maize cobs fromthe Tehuacan Valley of Mexico at ~5300 yr B.P.recently revealed a state of partial domestication,a mixture of maize- and parviglumis-like allelesat loci involved in domestication (2, 3). This ispuzzling, given the sustained use of domesti-catedmaize from ~6500 yr B.P. onward in SouthAmerica (Fig. 1 and table S1) (11, 18). However,principal components analysis and f3 statis-tics reveal considerable genomic distance be-tween these twoMesoamerican archaeogenomes(Fig. 1 and fig. S2), and f3 statistics confirm thatthe SM10 genome (3) is more maize-like, whereasthe Tehuacan162 genome (2) is more parviglumis-like (fig. S2). In total, the two genomes are fromthe same region and time period, and both arepartially domesticated, but otherwise, they appearto represent independent samples out of a diversesemidomesticated population containing an arrayof domestic and wild-type alleles.Given the state of partial domestication ob-

served in the Tehuacan and San Marcos ge-nomes (2, 3), early South American maize emergingfrom their common ancestral population wouldlikely also have been a partially domesticatedform of maize containing an assortment of wildand domestic alleles. This ancestral populationlikely harbored the building blocks for fullydomesticated maize but lacked the allelic fixa-tion and linkage of the modern domesticatedcrop. We expect that in this ancestral semi-domesticated population, domestication loci underongoing selection would have been continuallydecoupled from their chromosomal neighborhoodthrough recombination (19, 20), resulting in anenrichment of the original parviglumis genomicbackground near domestication genes relativeto its genome-wide retention. If the domestica-tion syndrome was fully established in the com-mon ancestor of all extant maize, no modernparviglumis genome should carry this enrichedaffinity to domestication loci to differing degreesin different maize lineages, because the samebackground would have become fixed in theircommon ancestor. However, if South Americanmaize became isolated while fundamental do-mestication was still ongoing, as we hypothesize,then components of the parviglumis genomicbackground are expected to differ between earlystratified maize lineages. Therefore in this case,modern parviglumis genomes would carry aspecifically South American or non–SouthAmerican affinity for the enriched wild-typebackground near domestication loci.We comparedD-statistics (21) across thewhole

genome (DWG) and within 10 kb of 186 knowndomestication loci (Ddom) to test for these asym-metricalparviglumis contributions between pairsof extantSouthAmericanandnon–SouthAmericanmaize around domestication genes (15). Wefound that parviglumis enrichment associatedwith domestication is highly patterned amongmajor ancestry groups, with several parviglumisgenomes associated exclusively with either SouthAmerican or non–South American Ddom enrich-ment and a significant association with ancestryoverall (Fig. 2C; c2 test P = 2.74 × 10−6). That is,

we observe thatparviglumis ancestry is enrichednear domestication genes in a pattern demon-strating that domestication-associated selec-tion was still ongoing after the stratificationof the major extant lineages from their semi-domesticated ancestral population. This pat-tern validates a model in which the ancestralpopulation in South America was itself onlypartially domesticated during its dispersalaway from the domestication center.In total, we find support for a model of strat-

ified domestication in maize (Fig. 2). The initialstages of maize domestication likely occurredonly once within a diverse wild Balsas River basingene pool, as previously suggested (13). However,before the domestication syndromewas fixed andstable, multiple lineages separated, and selectionpressures on domestication loci continued inde-pendently outside of the primary domesticationcenter. Some of these divergent semidomesticatedpopulations likely led to terminal lineages lack-

ing sufficient diversity and ecological context tocontinue the domestication process. Others, likeancestral South American maize, evolved intofully domesticated lineages under continuinganthropogenic pressures.The earliest evidence places maize in the

southwestern Amazon by ~6500 yr B.P. (11), aregion serving as a geographic interface of thelowland and Andean-Pacific genetic lineages(Fig. 1). We hypothesize that the southwesternAmazon may have been a secondary improve-ment center for the partially domesticated cropbefore the divergence of the two South Americangroups. When maize arrived, southwesternAmazonia was a plant domestication hotspot(22). Additionally, microfossil assemblages (11, 22)reveal the presence of polyculture (mixed crop-ping) from ~6500 yr B.P. onward, such that anew crop species could be integrated into ex-isting food production systems supporting do-mestication activities.

Kistler et al., Science 362, 1309–1313 (2018) 14 December 2018 2 of 4

San MarcosTehuacan

Z2Z6

Z64

Z65

Z66

Z67

Arica4Arica5

Z61

4383 B.P.

4030 B.P.

4363 B.P.

8750 B.P.

6589 B.P.

6655 B.P.

7746 B.P.

6000 B.P.

6700 B.P.

4030 B.P.

3770 B.P.

3730 B.P.

6500 B.P.

parviglumis range

North AmericanPan-AmericanMex. and C. America HighlandsLowland South AmericaAndean-PacificZea mays ssp. parviglumisArchaeological Genome

k = 5 ancestry groups

−0.20 −0.15 −0.10 −0.05 0.00 0.05 0.10

−0.2

−0.1

0.0

0.1

PC1

PC

2

San Marcos

Tehuacan

Z2Z6

Z64Z65

Z66

Z67Arica4

Arica5Z61

Fig. 1. Distribution and ancestry proportions of maize genomes and principal components analysis(PCA) of maize and parviglumis genomes. Pie colors reflect ancestral proportions estimated bymeans of model-based clustering (k = 5) of modern maize genomes (15). Archaeologicalgenomes were projected onto the PCA to mitigate degradation biases (15). Dates reflect earlyregional maize archaeobotanical remains (table S1 and fig. S1). C., Central; Mex., Mexico; PC1,First principal component; PC2, second principal component.

RESEARCH | REPORTon July 19, 2020

http://science.sciencemag.org/

Dow

nloaded from

Kistler et al., Science 362, 1309–1313 (2018) 14 December 2018 3 of 4

Fig. 2. A stratified domestication modelfor maize. (A) Schematic comparingthe conventional domestication modelunder which maize became fullydomesticated and then dispersedthroughout the Americas, versus astratified domestication model in whichpartially domesticated subpopulationsbecame reproductively isolated beforethe fixation of the domestication syndrome.(B) f4 statistics demonstrating excessallele sharing between the Pan-Americanlineage and wild parviglumis comparedwith other maize, revealing nonuniformcrop-wild gene flow after initialdomestication. Bars are three standarderrors under a block jackknife (15).(C) Bar plot of enriched parviglumiscontributions to ancestry near domesticationgenes, in which each bar is a parviglumisgenome contributing to South Americanmaize (blue) or other maize (red)Ddom enrichment. Geographic segregationin Ddom enrichment among parviglumisgenomes suggests that the domesticationsyndrome was not yet fixed in a commondomesticated ancestor of modern maize.

Domestication center

Secondary improvement center

parviglumis semidomesticated ancestral domestic maize

diverse landraces

diverse landraces

parviglumis semidomesticatedsemidomesticated

subgroups

H0 – Simple domestication

H1 – Stratified domestication

A

B

−0.003 0.000 0.003f4(((p1,p2),parviglumis ),Tripsacum)

p1 p2

Pan-American

North American

North American

North American

Lowland South American

Lowland S. American

Andean/Pacific

Andean/Pacific

Andean/PacificLowland South American

Pan-American

Pan-American

p1 p2 parviglumis

Tripsacum

p1 p2 parv.

Trip.

p1 p2 parv.

Trip.

C

TIL12TIL09

TIL03TIL06

TIL01TIL15

TIL07TIL10

TIL02TIL04

TIL11

0

2

4

6

8

Num

ber

of s

igni

fican

t Ddo

m e

nric

hmen

t cas

es

10

2 test: p=2.74 x 10-6

Enriched in South American maizeEnriched in other lineages

Fig. 3. Genomic relatedness overlappinglinguistic and archaeological patterns in low-land South America. Maize genomes with≥50% Andean-Pacific ancestry and ≥99% SouthAmerican ancestry are connected by lines withthe two other genomes with which they sharethe highest outgroup-f3 value. Geometricenclosures and mound ring villages of southernAmazonia broadly coincide with the expansionof Arawak languages, whereas the Uru and Araturing villages coincide with the distribution ofMacro-Jê languages (15) (figs. S3 and S4).Only the earliest regional dates for eacharchaeological tradition are shown (see tableS4). Macro-Jê languages borrowing an Arawakloanword for “maize” are based on (24).Arawak homeland is shown approximatelyin the modern location of Apurinã, in accordancewith (29).40°W50°W60°W70°W80°W

0°

10°S

20°S

940

940

700

1070 980

810

850

530

700

2300

3770

43005500

4310

36306500

Early maize sitesArawak homeland

Archaeological traditions

Language contact

Earliest dates Earliest dates

Modern

0.191 - 0.195

0.196 - 0.201

0.202 - 0.207

Archaeological

0.182 - 0.186

0.187 - 0.190

0.191 - 0.195

Maize genomes

outgroup-f3 allele sharingPaleoecology

‘maize’ loanword fromArawak into Macro-Jê

Aratu and Ururing villages

Geometric enclosuresand mound villages

RESEARCH | REPORTon July 19, 2020

http://science.sciencemag.org/

Dow

nloaded from

Pollen and phytolith data demonstrate a west-to-east pattern of maize expansion across theAmazon and show that maize was consistentlypresent from ~4300 yr B.P. onward in the easternAmazon (18). Initially, maize in the eastern Amazonwas part of a polyculture agroforestry systemcombining annual crop cultivation with wildresource use and low-level management throughburning (18). Maize cultivation proceeded along-side the progressive enrichment of edible forestspecies and subsequent waves of new crop ar-rivals, including sweet potato (~3200 yr B.P.),manioc (~2250 yr B.P.), and squash (~600 yr B.P.).The development of anthropogenically enrichedAmazonian Dark Earth soils ~2000 yr B.P. (23)enabled the expansion and intensification ofmaize cultivation, likely increasing carrying ca-pacity to sustain growing populations in theeastern Amazon (18). The extant endemic maizelineage in lowland SouthAmerica likely originatedwith this long-term process involving millennia ofevolving land-use practices.Several landraces and two archaeogenomes

(~700 yr B.P.) in eastern Brazil also show stronggenetic links to Andean maize near the south-westernAmazon (Fig. 3). This pattern closelymirrorslinguistic patterns linking Andean, Amazonian,and eastern Brazilianmaize cultivation and sug-gests a second major west-to-east cultural ex-pansion of maize traditions. A loanword formaize with possible Andean origins was trans-mitted fromAmazonianArawak languages—mostlikely originating in southwest Amazonia (24)—into Macro-Jê stock languages in the Braziliansavanna and Atlantic coast (24) (fig. S3). Archae-ological evidence suggests this expansion occurred~1200 to 1000 yr B.P. with the spread of a culturalhorizon of geometric enclosures and mound ringvillages throughout southern Amazonia and ringvillages in the central Brazilian savannas and theAtlantic coast (Fig. 3 and fig. S4) (25–27). Thisprocess is roughly contemporaneous with archae-ological Andean-admixed genomes in the area.Thus, Arawak speakers likely brought nonlocalAndean-Pacific maize lineages into a landscapewhere maize was an established component oflong-term land management and food produc-tion strategies.Finally, we quantified the mutation load in

maize genomes—the accumulation of potentiallydeleterious alleles due to drift and selection(16)—using a phylogenetic framework to estimateevolutionary constraint (15). We observe thatSouth American lineages carry a higher muta-tion load than other maize lineages. Mutationload increases linearly with distance from thedomestication center and is linkedwith ancestry,and the Andean-Pacific group carries the highestburden of potentially deleterious variants (Fig. 4)(15). The mutation load in the Andes has beenattributed to selection for high-altitude adapta-tions (16), but the elevated mutation load inlowland maize also suggests a history of sharedselection and drift effects prior to highlandadaptation. These processes would likely haveincluded a founder episode as maize was car-ried into South America, persistent selection

pressures for regional adaptation, and the lat-ter stages of domestication after isolation fromthe founding gene pool. We also find thatAndean and Pacific maize from ~1000 yr B.P.to the early colonial period has a low mutationload compared with its modern Andean-Pacificcounterparts (Wilcoxon P = 0.002477) (15) (Fig. 4);although still elevated compared with non–SouthAmerican lineages. It is possible that Andeanmaize experienced a wave of deleterious alleleaccumulation as human and crop populationswere disrupted by changes caused by the arrivalof Europeans (28). Alternatively, the increasingmutation load in modern crops could repre-sent the ongoing effects of burdensome alleleaccumulation over nine millennia of humanintervention.

REFERENCES AND NOTES

1. D. R. Piperno, A. J. Ranere, I. Holst, J. Iriarte, R. Dickau,Proc. Natl. Acad. Sci. U.S.A. 106, 5019–5024 (2009).

2. J. Ramos-Madrigal et al., Curr. Biol. 26, 3195–3201(2016).

3. M. Vallebueno-Estrada et al., Proc. Natl. Acad. Sci. U.S.A. 113,14151–14156 (2016).

4. V. Jaenicke-Després et al., Science 302, 1206–1208 (2003).

5. F. O. Freitas, G. Bendel, R. G. Allaby, T. A. Brown,J. Archaeol. Sci. 30, 901–908 (2003).

6. R. R. da Fonseca et al., Nat. Plants 1, 1–5 (2015).7. W. L. Merrill et al., Proc. Natl. Acad. Sci. U.S.A. 106,

21019–21026 (2009).8. D. R. Piperno, K. H. Clary, R. G. Cooke, A. J. Ranere, D. Weiland,

Am. Anthropol. 87, 871–878 (1985).9. A. Grobman et al., Proc. Natl. Acad. Sci. U.S.A. 109, 1755–1759

(2012).10. M. B. Bush et al., Quat. Sci. Rev. 141, 52–64 (2016).11. S. O. Brugger et al., Quat. Sci. Rev. 132, 114–128 (2016).12. F. A. O. of the United Nations, FAOSTAT statistics database

(2018); www.fao.org/faostat/.13. Y. Matsuoka et al., Proc. Natl. Acad. Sci. U.S.A. 99, 6080–6084

(2002).14. J. van Heerwaarden et al., Proc. Natl. Acad. Sci. U.S.A. 108,

1088–1092 (2011).15. Supplementary materials are available online.16. L. Wang et al., Genome Biol. 18, 215 (2017).17. N. Patterson et al., Genetics 192, 1065–1093 (2012).18. S. Y. Maezumi et al., Nat. Plants 4, 540–547 (2018).19. W. G. Hill, A. Robertson, Genet. Res. 8, 269–294 (1966).20. M. W. Feldman, S. P. Otto, F. B. Christiansen, Annu. Rev. Genet.

30, 261–295 (1996).21. R. E. Green et al., Science 328, 710–722 (2010).22. J. Watling et al., PLOS ONE 13, e0199868 (2018).23. B. Glaser, Philos. Trans. R. Soc. Lond. B Biol. Sci. 362, 187–196

(2007).24. E. R. Ribeiro, Liames 9, 61–76 (2009).25. J. G. de Souza et al., Nat. Commun. 9, 1125 (2018).26. M. J. Heckenberger et al., Science 321, 1214–1217 (2008).27. I. Wüst, C. Barreto, Lat. Am. Antiq. 10, 3–23 (1999).28. C. R. Clement, Econ. Bot. 53, 203–216 (1999).29. R. S. Walker, L. A. Ribeiro, Proc. Biol. Sci. 278, 2562–2567

(2011).30. L. Kistler et al., Dryad Digital Repository (2018); doi:10.5061/

dryad.70t85k2.

ACKNOWLEDGMENTS

We thank Admera Health for assistance with sequence datacollection and D. Piperno for comments on the manuscript.Funding: Work was supported by Natural Environment ResearchCouncil Independent Research Fellowship NE/L012030/1 to L.K.,and a sub-award from Science and Technology Facilities Councilgrant ST/K001760/1 (PI Thomas Meagher, co-I Peter Kille) toL.K. and R.G.A. Author contributions: Study conceptualizationand design: L.K., F.O.F, and R.G.A.; Sample acquisition: F.O.F., A.P.P.,C.G., B.A., and M.T.P.G.; Genomic data collection: L.K., F.O.F., O.S.,N.W., and R.R.M.; Genomic data analysis: L.K. and N.A.S.P.;Archaeology and linguistic background and interpretation: J.G.S.,S.Y.M., F.O.F., F.M.C., and E.R.R.; Interpretation and integration ofresults: L.K., S.Y.M., J.G.S., F.M.C., J.R.-M., N.W., F.O.F., and R.G.A.;Visualization: L.K., S.Y.M., J.G.S., H.L., and N.A.S.P.; Manuscriptdrafting: L.K., S.Y.M., and J.G.S., with input from N.A.S.P., F.O.F., andR.G.A. All authors reviewed and contributed to the final manuscript.Competing interests: We declare no competing interests. Data andmaterials availability: Raw sequence data, NCBI Sequence ReadArchive accession SRP152500. In-house scripts for data handling andanalysis (allele frequency estimation, f and D statistic calculation,genome alignment conformation for mutation load analysis, andexclusion amplification duplicate removal), genome-wide GERPscoring details, genomic mappability bed file, SNP calls, andmapDamage results are available in (30). Germplasm for newlysequenced maize landraces is curated at the Embrapa gene bank inBrasilia, Brazil, and Programa Cooperativo de Investigaciones enMaíz in Peru, which provided sample material for this study to F.O.F.and C.G. Archaeological samples from Santa, Chorrillos, Ica, and Jujuywere originally obtained from the PSUM Archaeological Project,Paurarku Archaeological Project and Samaca Archaeological Project,facilitated by archaeologists V. Pimentel, K. Lane, D. Beresford-Jones,and H. Yacobaccio.

SUPPLEMENTARY MATERIALS

www.sciencemag.org/content/362/6420/1309/suppl/DC1Materials and MethodsFigs. S1 to S4Tables S1 to S4References (31–94)

8 August 2018; accepted 22 October 201810.1126/science.aav0207

Kistler et al., Science 362, 1309–1313 (2018) 14 December 2018 4 of 4

0.165

0.170

0.175

North American

Genome-wide mutation load

Pan-American

Mex. and C. America Highlands

Lowland South American

Andean-PacificArchaeological And./Pac.

0.174

0.162

0.166

0.170

0 10

Distance from the Balsas Valley, Mexico

20 30 40 50 60 70

Genome-wide mutation load

Pearson’s r 2 = 0.51; p = 4.9 x 10-12

ANOVA r 2 = 0.74; p = 1.0 x 10-12

Fig. 4. Genome-wide mutation load acrossancestry groups (non-admixed samples onlyin top panel) and load compared with dis-tance to the domestication center. Mutationload is calculated as a proportion of thetheoretical maximum load over observed single-nucleotide polymorphisms, and ancient loadscores are rescaled for missingness using aProcrustes transformation (15). Euclidean dis-tance in degrees to the Balsas River valley isshown. And./Pac., Andean-Pacific.

RESEARCH | REPORTon July 19, 2020

http://science.sciencemag.org/

Dow

nloaded from

Multiproxy evidence highlights a complex evolutionary legacy of maize in South America

Andre P. Prous, Bernardo Arriaza, M. Thomas P. Gilbert, Fabio de Oliveira Freitas and Robin G. AllabySmith, Hope Loiselle, Jazmín Ramos-Madrigal, Nathan Wales, Eduardo Rivail Ribeiro, Ryan R. Morrison, Claudia Grimaldo, Logan Kistler, S. Yoshi Maezumi, Jonas Gregorio de Souza, Natalia A. S. Przelomska, Flaviane Malaquias Costa, Oliver

DOI: 10.1126/science.aav0207 (6420), 1309-1313.362Science 

, this issue p. 1309; see also p. 1246Scienceoccurred among multiple South American populations, including those in southwestern Amazonia.maize cultivars likely involved a ''semidomesticated'' lineage that moved out of Mexico. Later improvements theninvestigate the genetic changes that accompanied domestication (see the Perspective by Zeder). The origin of modern

applied genomic analysis to ancient and extant South American maize lineages toet al.European contact. Kistler Maize originated in what is now central Mexico about 9000 years ago and spread throughout the Americas before

The complexity of maize domestication

ARTICLE TOOLS http://science.sciencemag.org/content/362/6420/1309

MATERIALSSUPPLEMENTARY http://science.sciencemag.org/content/suppl/2018/12/12/362.6420.1309.DC1

CONTENTRELATED http://science.sciencemag.org/content/sci/362/6420/1246.full

REFERENCES

http://science.sciencemag.org/content/362/6420/1309#BIBLThis article cites 86 articles, 18 of which you can access for free

PERMISSIONS http://www.sciencemag.org/help/reprints-and-permissions

Terms of ServiceUse of this article is subject to the

is a registered trademark of AAAS.ScienceScience, 1200 New York Avenue NW, Washington, DC 20005. The title (print ISSN 0036-8075; online ISSN 1095-9203) is published by the American Association for the Advancement ofScience

Science. No claim to original U.S. Government WorksCopyright © 2018 The Authors, some rights reserved; exclusive licensee American Association for the Advancement of

on July 19, 2020

http://science.sciencemag.org/

Dow

nloaded from