Kinship practices in the early state El Argar society from Bronze Age Iberia

1
Vol.:(0123456789)
Scien i ic Repo s | (2022) 12:22415 | h ps://doi.o g/10.1038/s41598-022-25975-9
www.na u e.com/scien i ic epo s
Kinship p ac ices in he ea ly s a e
El A ga socie y om B onze Age
Ibe ia
Vanessa Villalba‑Mouco 1,2*, Camila Olia 3, C is ina Rihue e‑He ada 3,
Adam B. Roh lach 1,4, Ma ía Inés F egei o 5, Ainash Childebaye a 1, Ha ald Ringbaue 1,
Iñigo Olalde 6,7,8, E a Celd án Bel án 3, Ca he ine Puello‑Mo a 3, Miguel Valé io 3,
Johannes K ause 1, Vicen e Lull 3, Ra ael Micó 3,9*, Robe o Risch 3,9* &
Wol gang Haak 1,9*
The Ea ly B onze Age in Eu ope is cha ac e ized by social and gene ic ans o ma ions, s a ing in he
ea ly 3 d millennium BCE. New se lemen and une a y s uc u es, a i ac s and echniques indica e
imes o change wi h inc easing economic asymme ies and poli ical hie a chiza ion. Technological
ad ances in me allu gy also played an impo an ole, acili a ing ade and exchange ne wo ks,
which became angible in highe le els o mobili y and connec edness. A cheogene ic s udies
ha e e ealed a subs an ial ans o ma ion o he gene ic ances y a ound his ime, ul ima ely
linked o he expansion o s eppe‑ and o es s eppe pas o alis s om Eas e n Eu ope. E idence o
eme ging in ec ious diseases such as Ye sinia pes is adds u he complexi y o hese umul uous
and ans o ma i e imes. The El A ga complex in sou he n Ibe ia ma ks he gene ic u no e in
sou hwes e n Eu ope ~ 2200 BCE ha accompanies p o ound changes in he socio‑economic s uc u e
o he egion. To answe he ques ion o who was bu ied in he emblema ic double bu ials o he El
A ga si e La Almoloya, we in eg a ed esul s om biological ela edness analyses and a chaeological
une a y con ex s and e ined adioca bon‑based ch onologies om 68 indi iduals. We ind ha he
El A ga socie y was i ilocally and pa ilineally o ganized and p ac iced ecip ocal emale exogamy,
suppo ed by pedig ees ha ex end up o i e gene a ions along he pa e nal line. Synch onously
da ed adul males and emales om double ombs we e ound o be un ela ed ma ing pa ne s,
whe eby he incoming emales e lec socio‑poli ical alliances among El A ga g oups. In h ee cases
hese unions had common o sp ing, while pa e nal hal ‑siblings also indica e se ial monogamy o
polygyny.
The beginning o he Eu opean B onze Age (BA) in ol ed d as ic social changes ha esul ed in s ong poli i-
cal cen aliza ion, g owing economic inequali y, and se lemen and demog aphic dis up ions. These social
changes a e s iking in pa icula egions, such as Cen al Eu ope, B i any, sou he n England and sou heas e n
Ibe ia, whe e he unequal dis ibu ion o weal h, as e lec ed in he g a e goods, becomes mo e appa en and
consis en 1–3.
Recen genomic analyses ha e sugges ed ha hese changes we e ela ed o he wes wa d expansion o "s eppe-
ela ed ances y" and he educ ion in di e si y o male lineages in mos o Eu ope, ollowing a p ocess ha began
in he ea ly 3 d millennium cal BCE in Eas e n Eu ope4–8. In his con ex , inc eased iolence and social coe cion
could ha e played a ole in new social ela ions (e.g.,9–12). Howe e , he ole and na u e o popula ion mo e-
men s ( iolen o peace ul), o expansions in his p ocess emain a ma e o deba e. Following he in oduc ion
o game-shi ing inno a ions13, o he ac o s, such as new economies, clima e change o in ec ious diseases14–17
OPEN
1Depa men o A chaeogene ics, Max Planck Ins i u e o E olu iona y An h opology, 04103 Leipzig,
Ge many. 2Ins i u o Uni e si a io de In es igación en Ciencias Ambien ales de A agón, IUCA-A agosau us,
Za agoza, Spain. 3Depa men o P ehis o y, Uni e si a Au ònoma de Ba celona, Ba celona, Spain. 4School
o Ma hema ical Sciences, Uni e si y o Adelaide, Adelaide 5005, Aus alia. 5Independen esea che , Mu cia,
Spain. 6BIOMICs Resea ch G oup, Depa men o Zoology and Animal Cell Biology, Uni e si y o he Basque
Coun y UPV/EHU, Vi o ia-Gas eiz, Spain. 7Ike basque-Basque Founda ion o Science, Bilbao, Spain. 8Depa men
o Gene ics, Ha a d Medical School, Bos on, MA, USA. 9These au ho s con ibu ed equally: Ra ael Micó, Robe o
Risch and Wol gang Haak. *email: [email p o ec ed]; [email p o ec ed]; R[email p o ec ed];
[email p o ec ed]
2
Vol:.(1234567890)
Scien i ic Repo s | (2022) 12:22415 | h ps://doi.o g/10.1038/s41598-022-25975-9
www.na u e.com/scien i ic epo s/
migh also accoun o socio-economic and gene ic changes de ec ed h oughou he ans o ma i e imes o
he 3 d millennium BCE in Eu ope. In addi ion, he magni ude o ans o ma ions a ied in each egion. Mo e
in eg a i e wo k is needed o unde s and he ela i e con ibu ion o di e en ac o s in such in ica e p ocesses
o change a local and c oss- egional le els.
The a cheological complex o El A ga in sou heas e n Ibe ia p o ides a key case s udy o deepening ou
knowledge o social-poli ical eo ganiza ion in he Eu opean Ea ly BA (EBA) (Fig.1A). El A ga is one o he
a cheological en i ies in which a socio-economic di ide18,19 and a gene ic shi a e clea ly documen ed8, and i is
a guably one o he i s highly complex socie ies in Wes e n Eu ope o each he s a us o an ea ly s a e1,19–21. El
A ga de eloped o e h ee phases om ca. 2200 o 1550cal BCE19, sp eading om i s hea land in he coas al
lowlands in o he inne highlands, and co e ing ~ 35,000 km2 a i s peak. The A ga ic a cheological eco d
includes pe manen and densely popula ed hill op se lemen s o up o 5 hec a es in size. These se lemen s we e
o ganized and managed hie a chically, wi h e idence o public buildings o poli ical decision-making22, and
s uc u es o wa e supply. Mo eo e , la ge-scale s o age and p ocessing o ce eal c ops23, specialized po e y
Figu e1. Ea ly b onze age El A ga in sou he n Ibe ia. (A) Map o Ibe ia and loca ion o he A ga ic and o he
B onze Age nea bysi es. The map was c ea ed using QGIS 3.12 (h ps:// qgis. o g/ en/ si e/) and uses Na u al Ea h
ec o map da a om (h ps:// www. na u alea hda a. com/ downl oads/). (B) A iew o he La Almoloya hill op
si e om he Eas .
3
Vol.:(0123456789)
Scien i ic Repo s | (2022) 12:22415 | h ps://doi.o g/10.1038/s41598-022-25975-9
www.na u e.com/scien i ic epo s/
and me allu gical p oduc ion, and in ensi e subsis ence sys ems combining ain ed ag icul u e, manu ing and
small-scale i iga ion a e ound a A ga ic si es18,19,24.
A ga ic si es o e a unique oppo uni y o add ess ques ions o biological ela edness and kinship, since a
subs an ial p opo ion o he popula ion was bu ied in single o double ombs, placed unde he se led a eas.
These une a y p ac ices allow us o link indi iduals, g a e goods, ombs, and a chi ec u al uni s diach oni-
cally, which is a e in la e p ehis o ic Eu ope. To da e, A ga ic a cheology has app oached ques ions o kinship
h ough he analysis o double ombs, which ep esen nea ly 20% o bu ials in ce ain si es. On a e age, mo e
han hal o hese ombs we e assigned o wo adul s, a qua e we e occupied by an adul and a child, and he
combina ion o wo child en was a less common (~ 10%)25. The main ocus o p e ious esea ch has been on
double ombs wi h wo adul s, usually a male and a emale, p o iding he basis o he hegemonic model since
he la e nine een h cen u y: speci ically, ha hese g a es we e hough o e lec he e osexual, monogamous
couples (‘ma iages’) as he basis o nuclea amilies26–29.
To es his ‘ma iage’ hypo hesis, male and emale skele ons in double ombs we e sampled o adioca bon
da ing in he la e 1980s. A s a is ical analysis o pai ed 14C da es om a sample o 23 double bu ials sugges ed a
c oss-gene a ional gap o mos pai s o indi iduals30, which led o he al e na i e hypo hesis ha he ela ion-
ships be ween adul s in double ombs we e genealogical a he han socio-poli ical. As a esul , descen /consan-
guini y, ma ilocali y and ma ilineali y we e p oposed as he main p inciples o A ga ic kinship p ac ices25,31–33.
The aim o his s udy is o use s a e-o - he-a ancien DNA me hods o de e mine he na u e o he gene ic
ela ionships be ween indi iduals om El A ga con ex s, and o use hese as new e idence o shed ligh on
kinship p ac ices and social o ganiza ion o he A ga ic socie y, including inhe i ance ules ha a e po en ially
linked o households. Mo eo e , we aim o in es iga e he p e ailing hypo heses on A ga ic kinship p ac ices,
h ough an in e disciplina y app oach ha combines gene ic, os eological, and ine-g ained con ex ual in o -
ma ion om a cheological exca a ions. The bulk o ou da a comes om he si e o La Almoloya in Mu cia22,34,
which, due o i s unique p ese a ion and ex ensi e exca a ions, ea u es a a ie y o bu ial ypes and g a e
goods, complemen ed by de ailed an h opological da a and a adioca bon and s a ig aphy-based ch onology
(Supplemen a y 1, Fig.1B). The esul s a e expec ed o imp o e ou unde s anding o one o he ea lies , highly
complex socie ies in BA Eu ope.
Da a o e iew and de ec ion o gene ic ela ionships
A La Almoloya we sampled 86 indi iduals wi h sui able mo phological p ese a ion om a o al o 101 g a es,
con aining he emains o 128 indi iduals (Da ase S1.1). We ob ained high-quali y genome-wide da a (1240k
SNP cap u e da a) o 68 indi iduals (79% success a e) passing ou quali y con ol h esholds (< 4% con amina-
ion, cha ac e is ic aDNA damage p o iles, unambiguous sex de e mina ion) (“Ma e ials and me hods”, Da ase
S1.2). The 68 indi iduals om La Almoloya co e El A ga phase 2 (n = 41) and phase 3 (n = 27) o he local
s a ig aphy (2000–1750cal BCE and 1750–1550cal BCE, espec i ely), and we obse ed no bias wi h espec
o pos -deposi ional o aphonomic ac o s (“Ma e ials and me hods”). The ela i e equencies o emales and
males, adul s and subadul s wi h gene ic esul s ma ch hose o he physical an h opological examina ion o
each o hese demog aphic g oups, indica ing ha he sample is ep esen a i e o he g oup bu ied a he si e
(Da ase S1.3).
Ano he h ee indi iduals om La Almoloya yielded au osomal SNP da a below he h eshold o 20,000 SNPs:
ALM037 (18,226 SNPs), ALM045 (10,970 SNPs), ALM033 (356 SNPs) wi h no con amina ion es ima e (Da ase
S1.2). These h ee indi iduals we e only included in some analyses o con i m o exclude po en ial 1s -deg ee
biological ela ionships, which is also possible o low co e age da a (Supplemen a y 2). To es ima e he biological
ela edness among La Almoloya indi iduals and also among and be ween o he published BA indi iduals om
Ibe ia, we i s calcula ed he pai wise misma ch a e (PMR) (Supplemen a y 2, Fig.S1), which also p o ides a
gene al h eshold o backg ound (un-) ela edness in BA Ibe ia om andomly d awn pai s o indi iduals6,7,35.
He e, indi iduals om neighbo ing El A ga and Ibe ian BA g oups we e o pa icula in e es o he compa ison
and discussion o esul s (Da ase S1.28).
In o al, we epo 13 1s -deg ee- ela ionships and 10 2nd-deg ee- ela ionships among he 68 indi iduals a
La Almoloya, in ol ing 34 indi iduals (50%) o which genome-wide da a was gene a ed (Fig.2, Supplemen-
a y 2; Fig.S1–S4, Da ase S1.4-S1.6). O no e, 1s -deg ee ela i es can be pa en –o sp ing o ull siblings, and
2nd-deg ee ela ionships include aun -uncle/niece-nephew, g andpa en -g andchild, and hal -siblings (who
sha e only one o he biological pa en s). We we e able o econs uc se en pedig ees in ol ing all 1s -deg ee
ela ionships and o ex end some o hese pedig ees up o i e gene a ions (Fig.3, Supplemen a y 3, Fig.S5).
In addi ion, we ound pai s ha we e 2nd-deg ee ela i es. We se hese aside as ‘un esol ed pedig ees’ due o
he lack o b idging 1s -deg ee ela i es and he ac ha we we e unable o unambiguously de e mine he exac
genealogical ela ionships despi e he in eg a ion o os eological and a cheological da a (Supplemen a y 3, Da a-
se S1.4–S1.6). Wi h he help o an Iden i y-by-descen (IBD) analysis o impu ed high-quali y genome-wide da a
om selec ed pai s (> 600,000 SNPs; “Ma e ials and me hods”), we also ound e idence o mo e dis an gene ic
ela ionships up o he 6–7 h deg ee be ween indi iduals om h ee o he se en econs uc ed pedig ees, as well
as mo e dis an connec ions be ween si es (Supplemen a y 2, Fig.S4, Da ase S1.7). Recons uc ed pedig ees
om nea by BA si es a e shown in Supplemen a y 4.
We no ice a simila equency o 1s -(N = 13) and 2nd-deg ee (N = 10) ela ed indi iduals. As he numbe
o 2nd-deg ee ela i es is expec ed o double ha o he 1s -deg ee ela i es, he bu ials a La Almoloya seem
o emphasize he closes biological ies i we dis ega d he numbe o un ela ed indi iduals also bu ied he e.
While he numbe o samples is oo small o p o ide obus s a is ical suppo , he unde ep esen a ion o 2nd-
deg ee ela i es indica es ha no all biological ela i es we e bu ied in he se lemen , which sugges s in e -si e
mobili y, in ol ing esiden ial changes o a subs an ial pa o he popula ion. A concomi an inc ease in child
4
Vol:.(1234567890)
Scien i ic Repo s | (2022) 12:22415 | h ps://doi.o g/10.1038/s41598-022-25975-9
www.na u e.com/scien i ic epo s/
bu ials du ing phase 3 (a common ea u e du ing la e A ga ic imes 30), sugges s an emphasis in une a y i es
on he closes biological o sp ing, o an inc ease in child mo ali y.
In wha ollows, we desc ibe he esul s o he es ima ion o biological ela edness in e e ence o he physical
dis ances be ween he bu ials, s a ing wi h closely inhuma ed indi iduals in double bu ials ollowed by ela i es
sepa a ed by la ge dis ances a he si e.
Double bu ials o wo adul s. Double bu ials a e an iconic ea u e o he El A ga g oup and play a cen-
al ole in he discussion o A ga ic kinship. This ype o bu ial had al eady been a es ed o in ea lie imes a
sou heas Ibe ian si es such as Molinos de Papel36,37, and con inued o be p ac iced du ing phases 2 and 3 o El
A ga . A La Almoloya, 48 ou o 126 indi iduals (38%) we e bu ied in 24 double ombs. The majo i y o hese
ombs (N = 20) con ained wo adul s ( ou een om phase 2 and six om phase 3), while he es we e ei he wo
Figu e2. (A) Plan o La Almoloya phase 3 highligh ing he posi ion o bu ials ha ha e yielded su icien
aDNA and close gene ic ela edness be ween indi iduals up o he 2nd deg ee. Fo ull pedig ees see SI
Appendix, Sec .3. (B) Exempla y pi hos double bu ial and g a e goods (C) om La Almoloya bu ial AY38.
5
Vol.:(0123456789)
Scien i ic Repo s | (2022) 12:22415 | h ps://doi.o g/10.1038/s41598-022-25975-9
www.na u e.com/scien i ic epo s/
Figu e3. Summa y o he gene ic esul s om he double bu ials a Almoloya. (A) Resul s o he Pai wise-
misma ch a e (PMR) analysis including all pai s o indi iduals om he Ibe ian B onze Age a ailable o
da e. The x-axis shows he numbe o o e lapping SNPs be ween each pai and he y-axis he coe icien o
ela edness. All adul double bu ials all wi hin he ange o andomly d awn pai s o indi iduals om he
Ibe ian B onze Age. (B) Recons uc ed pedig ees o h ee cases in ol ing adul double bu ials and hei
common o sp ing. Below he colo ed squa es and ci cles is he con ex ual in o ma ion o all adul double
ombs, including sex, age a dea h, mi ochond ial and Y-ch omosomal haplog oups, di ec 14C da es, as well
as s a ig aphy. Pink ou lines e lec in e changeable gene ic sexes and g ay dashed lines indica e pedig ees
econs uc ed om low co e age da a.

6
Vol:.(1234567890)
Scien i ic Repo s | (2022) 12:22415 | h ps://doi.o g/10.1038/s41598-022-25975-9
www.na u e.com/scien i ic epo s/
child en o one adul and one child. In one o he adul double bu ials (AY80), he emains o a skull o a hi d
indi idual we e placed unde a slab ou side he cis , a e he omb had al eady been sealed.
We success ully e ie ed genomic da a om bo h skele ons o en adul double bu ials (20 indi iduals). The
gene ic esul s con i med he an h opological mo phological sex de e mina ion o he indi iduals, and in all
en cases he double bu ials con ained a male and a emale, which we e ound o be gene ically un ela ed, wi h
he coe icien o ela edness be ween hem being 0.001 ( anging om − 0.024 o 0.022) (Fig.3A, Da ase S1.4,
Supplemen a y 2). Each pai o double bu ial adul s yielded PMR alues which we e close o he baseline median
o he PMR alues, which was es ablished om all un ela ed pai s o Ibe ian BA indi iduals ac oss many si es.
Signi ican o e lap o he calib a ed anges om di ec adioca bon da es o all analyzed pai s makes he coexis -
ence o bo h indi iduals possible (Fig.2, Supplemen a y 5, Fig S6, Da ase S1.2, Da ase S1.8). All analyzed adul
couples ca ied di e en mi ochond ial haplog oups, which uled ou di ec ma e nal links (Da ase S1.2). This
obse a ion can also be ex ended o o he A ga ic si es, such as Ce o del Mo ón, om whe e we also analyzed a
con empo aneous, adul double bu ial ha con ained an un ela ed emale (CMO001) and male (CMO002), and
o he BA si es om sou heas e n Ibe ian BA, such as Molinos de Papel (MPD002, emale and MPD003, male)
(Supplemen a y 2, Da ase S1.2). In ano he adul double bu ial (AY82) we ob ained good co e age da a o he
emale indi idual (ALM018), bu no o he male (ALM045), howe e he esul s o he biological ela edness
es s s ill sugges ed ha he pai was no ela ed (Fig.3A, Da ase S1.4). In addi ion, in h ee ou o en cases
om La Almoloya, males and emales om double bu ials had o sp ing oge he , and hus ep esen ed ma es
(Fig.3B). This obse a ion sugges s he exis ence o emale–male socio-poli ical alliances in li e, which we e
also symbolized in he une a y p ac ices o he El A ga socie y. So a , he con empo aneous in a-si e genomic
s udies ha e epo ed he p esence o sexual pa ne s a he si e con i med by common o sp ing bu bu ied in
di e en g a es 38. In o he examples, only one o he biological pa en s was ound bu ied a he si e 39–41.
An ou s anding example is he high-s a us pi hos double bu ial AY38 om he pala ial building (Fig.2B), in
which a man (ALM039) and a woman (ALM038) we e endowed wi h ich g a e goods (Fig.2C) and hus we e
in e p e ed as ha ing been p ominen membe s o he uling class22. The woman was bu ied wi h one o he i e
sil e diadems ound uniquely a A ga ic si es ( he o he ou a e om he eponymous El A ga si e, abou 100km
sou h o La Almoloya). This ou s anding une a y i em has been in e p e ed as a symbol o dis inc ion and powe
o some El A ga women22. F om he une a y con ex we know ha he man died i s , jus sho ly be o e he
woman, because his skele on was ound benea h he ea u ing a ema kably low deg ee o disa icula ion and
join displacemen , in pa icula along he spine and he ho ax. We ound a common daugh e in he double
g a e AY30 (ALM030) bu ied in a pi wi hou goods in a di e en a chi ec u al complex, u he away om he
pa en s (Fig.3B). ALM030 had passed away p ema u ely a an es ima ed age o 14–17mon hs.
Ano he ou s anding example o a double bu ial con aining pa ne s is AY80. The s one cis con ained a
30–35-yea -old emale (ALM015) and a 35–40-yea -old male (ALM016) a ibu ed o he high social class
(Fig.3B). The une a y con ex sugges s ha he woman had died i s and he bones we e hen collec ed and
deposi ed in a bundle on op o he male bu ial, which ook place a a la e ime. Bo h had a common son
(ALM052), who eached 45–50yea s o age and was bu ied wi hou g a e goods 11m away, oge he wi h he
adul emale ALM053, who was bu ied a e him, bo h in ano he adul double bu ial AY42.
The skull o an adul male (ALM017) om ou side he cis in which he couple om AY80 had been bu -
ied, was ound o be ela ed in he 3 d deg ee o bo h indi iduals in AY80. Addi ional IBD analysis o he pai
ALM015 and ALM017 con i med he 3 d-deg ee ela ionship in di ec gene a ional succession, who oge he
wi h ALM016 hus ep esen g ea -g andpa en s and g ea -g andson (Fig.3B; Supplemen a y 2, Da ase S1.7).
In addi ion, we ound ha he g ea -g andson, ALM017, was 1s -deg ee ela ed o ALM060, his 18–20-mon h-
old daugh e who was bu ied in a pi hos in he same a ea. This cons i u es he la ges pedig ee econs uc ed
in La Almoloya, which connec s ela i es o e i e gene a ions spanning phases 2 and 3, om he g ea -g ea -
g andpa en s (ALM015 and ALM016) o he g ea -g ea -g anddaugh e (ALM060), in a ela i ely close a ea,
e en hough gene a ions a e missing, emained un yped, o migh ha e been bu ied elsewhe e (Fig.3B, Sup-
plemen a y 2, Fig.S4). While he small numbe o cases and limi a ions o 14C da e anges do no allow he
gene alized assump ion o genealogical links ac oss phases, his example a gues agains he possibili y ha he
d ama ic changes in se lemen layou in phase 3 we e unde aken by an en i ely new g oup o dynas y. Gene ic
con inui y a he popula ion le el be ween he wo phases, as desc ibed in8, also ules ou scena ios o d as ic
demog aphic changes (Supplemen a y 5, Fig.S7).
G a e AY22 ep esen s ano he adul double bu ial wi h common o sp ing (Fig.3B). The bu ial con ained
a 35–45-yea -old emale (ALM048) and an adul male (ALM049), whose s a e o p ese a ion did no pe mi a
mo e p ecise age es ima ion. We iden i ied hei common adul son ALM034 in a single cis bu ial (AY16) less
han 3m away om AY22.
Taken oge he , males and emales om adul double ombs we e no ound o be gene ically ela ed o each
o he . I espec i e o his inding, hese indi iduals we e in ol ed in all 1s - and 2nd-deg ee ela ionships among
adul s a he si e, which sugges s ha double bu ials displayed a cen al social ole (Fig.3B, Supplemen a y 2,
Supplemen a y 3).
Double bu ials wi h an adul and a child. A La Almoloya, we documen ed h ee double bu ials con-
aining an adul and a subadul indi idual, o which only one could be analyzed gene ically. Cis bu ial AY21
con ained a 30–35-yea -old woman (ALM073), who was bu ied holding a newbo n emale baby (ALM062)
agains he igh side o he ches . As he skele al supe posi ion indica es, he synch onic inhuma ion o he wo
bodies and he gene ic analysis sugges s a mo he /daugh e ela ionship (Fig. S5A), whe e complica ions in he
pe iod a e childbi h could be conside ed he likely cause o dea h. I is no ewo hy ha he child ca ied he
7
Vol.:(0123456789)
Scien i ic Repo s | (2022) 12:22415 | h ps://doi.o g/10.1038/s41598-022-25975-9
www.na u e.com/scien i ic epo s/
aneuploid XXX-synd ome 8, bu his condi ion is unlikely o ha e caused he un imely dea h o bo h he baby
and he mo he 42.
The combina ion o adul /subadul indi iduals in double bu ials (25%) is ela i ely uncommon a El A ga
si es, and he example om omb AY21 migh hin a excep ional ci cums ances in ol ing a close gene ic ela ion-
ship. This migh be he case o omb AY85, in which ano he emale neona e, whose sex has been de e mined
gene ically (ALM079), was ound in he a ms o an adul emale (ALM066) and in a posi ion ema kably simila
o omb AY21. Un o una ely, he sample o he adul emale did no p oduce su icien gene ic da a and he e o e
we could no asce ain a pu a i e mo he /daugh e ela ionship.
Con e sely, omb BA6 a La Bas ida shows ha o he scena ios a e possible. He e, a 25–30-yea -old male
(BAS002) was bu ied alongside a newbo n boy (BAS026), bu he wo we e no gene ically ela ed (Da ase S1.2).
I is impo an o no e ha his bu ial canno be conside ed simul aneous as in he o he wo cases, since pa
o he pi hos im b oke when he omb was eopened and he agmen s ended up on op o he adul pel is and
below he neona e skele on, hus p o iding e idence o successi e inhuma ions. Howe e , adioca bon da es
we e no s a is ically di e en a he 95% con idence le el (Supplemen a y 5). I is possible ha his bu ial was
mean o ep esen a a he /son ela ionship, bu whe he he adul male bu ied in BA6 would be awa e o he
eal biological a he hood o no (as s ep a he and he e o e social kin) emains an open ques ion.
Double bu ials o child en. We success ully eco e ed DNA om he only double bu ial a La Almoloya
(AY30) wi h wo child en. The g a e consis ed o a small pi , in which a emale oddle (ALM030) o 14–17mon hs
was i s bu ied, ollowed la e by an 8- o 9-yea -old gi l (ALM031). The e we e no g a e goods associa ed wi h
ALM031, and om he une a y con ex we in e ha she was bu ied a e ALM030, as he skele on was ound
ully a icula ed and pa ly on op o he jumbled pos -c anium o he younge gi l ALM030. We de ec ed a hal -
sibling ela ionship be ween hese wo gi ls on hei a he ’s side (Fig.3, Supplemen a y 3), which indica ed ha
he adul male ALM039 om he weal hies g a e AY38 o La Almoloya, loca ed in he nea by pala ial building,
was he a he o bo h. I is no ewo hy ha he adul emale ALM038, bu ied alongside his man, was he mo he
o only one o he gi ls (ALM030), and ha we ha e no iden i ied he biological mo he o ALM031 among he
success ully yped indi iduals. The a chaeological con ex does no p o ide clues as o whe he he wo mo he s
li ed a he same ime o no , no whe he his case ep esen s an example o se ial monogamy, o , al e na i ely,
polygamy. Howe e , he ac ha he hal -sis e s we e en ombed oge he e lec s awa eness (on he pa o he
people who bu ied hem) o he kin ela ionship be ween he wo child en, i espec i e o hei di e en biologi-
cal mo he s, and e y likely also he acknowledgemen o a he hood on behal o ALM039. Howe e , i is also
possible ha hese unions we e empo al and dissol able.
The ins ances in which siblings we e iden i ied gene ically also ep esen a iable a chaeological con ex s and
si ua ions open o in e p e a ion. We also de ec ed wo siblings bu ied oge he in omb BA23 om La Bas ida,
which, like AY30, also da es o he la e A ga ic phase (Supplemen a y 4). A 9- o-11-mon h-old gi l BAS017
was bu ied in a ce amic essel, ollowed by he b o he BAS018, who died sho ly a e a oughly he same age.
Con e sely, in ano he case o siblings om La Almoloya, wo boys (ALM080 and ALM081) we e bu ied e y
close o each o he bu in sepa a e g a es (AY88 and AY89, espec i ely) (Fig. S5).
As a esul , we obse e ha double ombs con aining a leas one child ea u e close biological ela ionships,
excep o he case o he adul and newbo n male ound a La Bas ida (BA6), which de ia es om he double
bu ials o adul s.
Gene ic ela ionships beyond double ombs. We also obse ed se e al biological ela ionships
be ween indi iduals bu ied in single ombs and he e o e explo ed he deg ees o ela edness in he ligh o
ch onological and spa ial dis ances. Phase 3 a La Almoloya is cha ac e ized by a ne wo k o housing complexes
ha we e buil a ound 1750cal BCE, and which o m a la ge pa o he s uc u es isible oday (Fig.2A). How-
e e , he layou o he p eceding phase 2 was la gely disman led by he u baniza ion e o s o phase 3. Thus,
building complexes canno be used as a gene al backg ound, and i is ad isable o ope a e wi h aw dis ances
only.
To analyze he spa ial dis ibu ion o biologically ela ed indi iduals, we plo ed he physical dis ances be ween
all 1s - and 2nd-deg ee ela ed pai s o indi iduals (Fig.2A). We ind ha mos o he indi iduals wi h close
pa en al links we e bu ied less han 5m apa om each o he . In e es ingly, hal -siblings (biologically 2nd-deg ee
ela i es) a e spa ially as close as ull siblings (1s -deg ee ela i es), while all o he 2nd-deg ee ela ed indi iduals
(uncle-aun /nephew-niece and g andpa en -g andchild ela ionships) a e sepa a ed by 19–32m, and a e o en
bu ied unde he loo s o di e en buildings o phase 3. We documen ed a case o ull siblings (ALM080 and
ALM081) loca ed in sepa a e bu close pi hoi g a es wi hou g a e goods. Bo h male indi iduals died p ema u ely,
one a he age o 14–20mon hs (ALM080) and he o he a he age o 18–24mon hs (ALM081), and while he
ime elapsed be ween hei dea hs emains unknown, his si ua ion sugges s he in en ional placing o he wo
b o he s in a nea by space (Fig. S5). As men ioned abo e, o he ull siblings om La Bas ida (gene ically male
and emale) we e ound bu ied oge he in omb BA23 (Supplemen a y 4). These wo examples may e lec social
ies among subadul siblings ega dless o hei biological sex.
We de ec ed a possible hal -sibling ela ionship on he a he ’s side be ween ALM068 (a 14–16-mon h-old
emale baby om omb AY8) and ALM078 (a 14–16-mon h-old male om omb AY23 in e ed some ime
la e ), bu ied in sepa a e g a es bu wi hin he same housing complex. The mo he o ALM068 was bu ied in a
sepa a e omb (AY17-ALM077), close o he daugh e bu no o he boy (ALM078), o whom she is biologically
un ela ed. Ano he possible, bu pe haps less plausible in e p e a ion o he 2nd-deg ee ela ionship be ween
he wo child en is ha he boy (ALM078) was he uncle o he gi l (ALM068) on he a he ’s side (albei one
gene a ion younge ), as he is no ela ed o he mo he , ALM077 (Fig. S5).
8
Vol:.(1234567890)
Scien i ic Repo s | (2022) 12:22415 | h ps://doi.o g/10.1038/s41598-022-25975-9
www.na u e.com/scien i ic epo s/
Ano he hal -sibling ela ionship on he a he ’s side can be sugges ed o ALM046 (a 14–18-mon h-old
boy), bu ied in g a e AY13, and ALM047 (a 6–7-mon h-old male in an ) bu ied in g a e AY14, none o which
had g a e goods. In his case, we ound nei he he a he no he mo he among he success ully geno yped
indi iduals, and we canno comple ely ule ou he possibili y o an uncle/nephew ela ionship. Howe e , hese
indi iduals we e bu ied close o one ano he , as was also he case wi h o he pai s o siblings, which s ongly
hin s a speci ic social and amilial ies (Supplemen a y 3).
In summa y, an aun -uncle/niece-nephew ela ionship canno be excluded in cases in which wo child en
a e ela ed in he 2nd deg ee. I is wo hwhile no ing ha hal -siblings, which would in e e y case be linked ia
he pa e nal line, make polygamy (polygyny) and se ial monogamy plausible p ac ices o conside . Howe e ,
we also ound o he 2nd-deg ee ela ionships be ween an adul and a subadul o which we canno disca d an
a uncula ela ion, such as ALM004 (adul emale) and ALM075 (in an emale), o ALM019 (adul emale) and
ALM069 (in an male) (Supplemen a y 3).
Las ly, we we e able o in e mo e dis an ela i es by calcula ing he numbe and ac ion o segmen s in he
1240k SNP da a ha a e iden ical-by-descen (IBD) in indi iduals wi h > 600,000 SNPs a e impu a ion (“Ma e-
ials and me hods”, Supplemen a y 2). This me hod no only allowed us o con i m and dis inguish be ween
close 1s - and 2nd-deg ee ela ionships bu also o es ima e biological ela edness up o he 6 h–7 h deg ee, by
which we we e able o connec h ee o he se en main pedig ees econs uc ed a La Almoloya (Da ase S1.7,
Supplemen a y 3).
In e ing kinship p ac ices. Following he indi idual-based e alua ion o biological ela edness a La
Almoloya, we compa ed he s uc u e o he econs uc ed pedig ees om all ela ed indi iduals and looked
o links ac oss gene a ions ha would signal lineali y o inhe i ance along p e e en ial pa en al lines. He e,
we obse ed ha all econs uc ed pedig ees a e linked h ough he pa e nal side and, in one case, he male
lineage can be aced o e a leas i e gene a ions (Fig.3, Supplemen a y 3). In addi ion, in all o he cases o
2nd- and 3 d-deg ee ela ionships (six 2nd-deg ee pai s), o which he pedig ee could no be ully es ablished
o ex ended due o missing 1s -deg ee ela ed indi iduals, we obse e ha he possible al e na i e pedig ees can
also only be explained h ough he pa e nal line (Supplemen a y 3). All 1s -deg ee ela ionships among adul s
in ol ed a leas one adul male (3 ou o 19 adul males analyzed) (Fig.4). No adul male has an adul daugh e ,
sis e , b o he , o adul hal -siblings also bu ied a he si e. The ew 2nd-deg ee ela ionships in ol ing adul s
we e all also be ween males (4 ou o 19 adul males analyzed) (Fig.4).
Con e sely, we ound no 1s - o 2nd-deg ee ela ionships be ween adul women (0 ou o 30 adul women
analyzed). Women ela ed in he 1s deg ee a he si e we e mo he s o ei he gi ls (AY21/2-ALM062, AY30/1-
ALM0030, AY8-ALM068) o adul males (AY16-ALM034, AY27/1-ALM058 and AY42/1-ALM052), bu no
boys. These women we e no ela ed o any o he adul women (Fig.4) and had no pa en s a he si e. The same
holds ue o 2nd-deg ee ela ed emales, hen conside ed in eg a ed lineage emales, who we e ound o be
aun s and/o g andmo he s (AY58-ALM004, AY87-ALM019, AY26/2-ALM086) o bo h gi ls and boys, bu ,
again, ne e ela ed o adul women (Fig.4).
These esul s ob ained by econs uc ing he pedig ees a e in line wi h esul s ha ha e al eady been epo ed
a he me a-le el in8, whe e i was desc ibed ha males o La Almoloya had mo e close ela i es a he si e han
emales. This obse a ion was made based on signi ican ly highe obse ed 3-s a is ics o 3 (male, male; Mbu i)
han 3 ( emale, male; Mbu i) han 3 ( emale, emale; Mbu i) (a e excluding all 1s - and 2nd-deg ee ela ed
pai s). Al hough he 3-ou g oup s a is ic is used a he popula ion le el o measu e sha ed gene ic d i be ween
wo popula ions a e he spli om a common ou g oup, i has been also use ul o iden i y 1s -deg ee ela i es
as hey sha e hal o hei genomes and hus, will epo highe 3- alues4,8.
Examining he pa e nal lineages, we obse e a highe esiden ial s abili y, no only a he si e bu also in
some speci ic housing complexes (i.e., close p oximi y) (Fig.2A). A p ime example is ALM034/AY16, an adul
male bu ied close o his pa en s, ALM048/AY22-1 and ALM049/AY22-2. A simila scena io could be in e ed
o emale ALM086/AY26-2, bu ied in an adul double bu ial close o he adul son ALM058/AY27-1. Un o u-
na ely, we did no ob ain su icien aDNA om he pa ne bu ied wi h ALM086/AY26-1 o p o e he biological
a he hood be ween him and ALM058/AY27-1 (Supplemen a y 3, Fig.S5). Finally, ano he example o c oss-
gene a ional pa e nal lineages bu ied close in space a e ALM015/AY80-1 and ALM016/AY80-2, an adul double
bu ial, which also con ained he skull o hei adul g ea -g andson ALM017/AY80-0 (Fig.3, Supplemen a y 3).
O he sou ces o in o ma ion o in e lineali y and locali y can be gleaned om unipa en ally inhe i ed ma k-
e s, such as mi ochond ial DNA (m DNA) and Y-ch omosomal haplog oups. I espec i e o he a ying gene ic
esolu ion, he inding o only one single Y-ch omosome lineage in La Almoloya (R1b-P312 > Z195), which is
also he p edominan lineage ac oss Ibe ia, is ema kable, bu canno be used o esol e ela ionship pa e ns
a an in a-si e le el and hus sugges s a much mo e common p ac ice a a b oade scale o a small/non-di e se
sou ce popula ion o Y ch omosome di e si y. The di e si y o he m DNA is also simila o he di e si y obse ed
in Ibe ia in he p eceding pe iods 43, wi h he excep ion o one emale indi idual ca ying m DNA haplog oup
R0a, which has no been epo ed o da e om Ibe ia (Da ase S1.2).
Aside om close ela ionships wi hin La Almoloya, we also ound a close ela ionship be ween ALM034/
AY16 and a 40–50-yea -old emale om Lo ca, bu ied in Mad es Me ceda ias Tomb 4_1 (MMI003), ano he
A ga ic si e abou 50km away. Due o he low co e age o MMI003, he biological ela edness es s e u ned
an in e media e alue be ween 1s - and 2nd-deg ee and hus need o be in e p e ed wi h cau ion. In he case o
a 1s -deg ee ela ionship, MMI003 would be a di ec example o emale exogamy. In he case o a 2nd-deg ee
ela ionship, he ype o mobili y would be unspeci ic as he pa en lea ing La Almoloya could ha e been ei he
he a he o he mo he o he emale om Lo ca (Fig.3, Supplemen a y 3).
9
Vol.:(0123456789)
Scien i ic Repo s | (2022) 12:22415 | h ps://doi.o g/10.1038/s41598-022-25975-9
www.na u e.com/scien i ic epo s/
Using IBD analyses, we also ound h ee 6 h–7 h-deg ee in e -si e ela ionships in ol ing pai s o males and
emale and male indi iduals om La Almoloya and La Bas ida (Da ase S1.7), which connec he ne wo ks o
biological ela i es ac oss bo h El A ga si es. In addi ion, we ound be ween-si e connec ions be ween indi-
iduals om phase 3 o La Almoloya and La Ho na (LHO), a Valencian BA si e 44,45. This inding highligh s he
powe o de ec long-dis ance ela ionships h ough in ensi e sampling schemes and emphasizes he ex ended
ne wo ks wi hin El A ga , bu also sheds ligh on he poli ical and economic ela ions wi h neighbo ing socie ies46
(Da ase S1.7, Fig.S4).
The combined iew o he esul s leans owa ds he p ac ice o emale exogamy and pa ilocali y, in which
young emales mo ed o a di e en esidence o build new ela ionships. The adul emales bu ied in double
g a es p o ide suppo o hese p ac ices as hey ha e no pa en s bu ied a he si e and, apa om hei o -
sp ing, also no o he adul ela i es, which sugges s ha hey came om ou side he communi y and we e in e-
g a ed h ough hei union wi h local males, and can hus be conside ed ma es o lineage males. Impo an ly, he
ac ha we do no ind 1s - o 2nd-deg ee ela ionships be ween adul women a La Almoloya sugges s ha his
p ac ice was ecip ocal and ha young emales om La Almoloya also mo ed o o he si es. Pa ilocali y does
no necessa ily imply he absence o mobili y o adul males. In ac , ou esul s also suppo subs an ial mobili y
o bo h sexes as shown by he p esence o ewe 2nd-deg ee han 1s -deg ee ela i es a he si e. Howe e , he
abili y o ace male lineages h ough gene a ions by he p esence o adul male o sp ing, bu no emale adul
o sp ing (Fig.4B), suppo s pa ilocali y despi e male and emale mobili y.
The in eg a ion o gene ic, demog aphic, and o he con ex ual da a sheds u he ligh on he social o gani-
za ion and kinship p ac ices o he A ga ic communi y. Resul s o he gene ic and an h opological sex de e -
mina ion indica e a sligh excess o males (11 males s. 8 emales) among subadul s. By con as , he numbe
o adul women (N = 53) exceeds he numbe o adul men (N = 32), esul ing in a sex a io o 1.65 in a o o
Figu e4. O e iew o gene ic sex and age a dea h de e mina ions and o al numbe s o obse ed ela ionships
pe ca ego y and long-dis ance ela i es ou side La Almoloya. (A) Simpli ied age classes o he indi iduals
analyzed om La Almoloya: subadul (including in an (0–3yea s) and child (3–12yea s)) and adul (including
young adul (20–35yea s), middle adul (35–50yea s), and old adul (50+ yea s)56) emales and males. O no e,
adolescen indi iduals (12–20yea s) we e no ound a he si e; (B) o al numbe o 1s - (pink) and 2nd-deg ee
(pu ple) ela i es be ween age and sex classes as summa ized in (A). Numbe s e lec he numbe o links in he
pedig ees be ween age/sex classes and numbe s inside symbols e lec he numbe o links wi hin each class. The
g aph illus a es he absence o ei he 1s - o 2nd-deg ee ela ionships among adul emales; (C) Long-dis ance
ela i es (6–7 h deg ee) as indica ed by sha ed IBD-blocks be ween indi iduals om La Almoloya and o he
BA a chaeological si es. The map was c ea ed using QGIS 3.12 (h ps:// qgis. o g/ en/ si e/) and uses Na u al Ea h
ec o map da a om (h ps:// www. na u alea hda a. com/ downl oads/).