1
Scien i ic RepoR s | 6:38817 | DOI: 10.1038/s ep38817
www.na u e.com/scien i ic epo s
S able C & N iso opes in 2100 Yea -
B.P. human bone collagen indica e
a e die a y dominance o C4 plan s
in NE-I aly
Zi a La anchi1, An onio Delgado Hue as2, Syl ia A. Jiménez B obeil1,
A senio G anados To es2 & Jose A. Riquelme Can al3
C4 plan s (e.g. maize, mille ), pa o ou cu en die , a e only endemic o educed a eas in Sou h-
Eu ope due o hei need o wa m clima es. Since he i s es iges o ag icul u e in Eu ope emains o
C4 plan s we e eco ded bu hei o e all p opo ion in he human die emains unknown. The e o e,
iso opic (δ13C and δ15N) composi ion o bone collagen om he skele al emains (human and animals)
o a Cel ic popula ion, Cenomani Gauls, om Ve ona (3 d o 1s cen u y BC) in he NE I aly p o ide a new
pe spec i e on his ma e . The δ13C collagen alues o 90 human skele al indi iduals ange be ween
−20.2‰ and −9.7‰ (V-PDB) wi h a mean alue o −15.3‰. As p esen day C4 plan s ha e δ13C alues
a ound −11‰, which is equi alen o −9.5‰ o samples o p eindus ial age, he less nega i e δ13C
alues in hese indi iduals indica e a die domina ed by C4 plan s. This palaeodie a y s udy indica es
ha some Eu opean popula ions p edominan ly consumed cul i a ed C4 plan s 2100 yea B.P. This
is suppo ed by he paleobo anical eco ds and ancien Roman sou ces (e.g. Pliny he Elde ), which
indica e ha mille was a s aple ood in Sou h-Eu ope.
The e m C4 plan s e e s o he ype o pho osyn he ic pa hway (Ha ch-Slack o C4 cycle) used by hese plan s,
which has less iso opic disc imina ion han he C3 cycle (C3 plan s). In he C4 cycle he i s s able compound
namely oxaloace ic acid is o med by 4 ca bon a oms. This g oup includes wa m clima e g asses like maize, mil-
le s, so ghum and suga cane, among o he s1.
Paleobo anical s udies ha e shown he p esence o domes ica ed C4 plan s (in pa icula b oomco n o
Panicum Miliaceum L. and ox ail mille o Se a ia I alica L. Beau .) in Cen al and Eas e n Eu ope as ea ly as
5000-4000 BC o Panicum Miliaceum and La e B onze Age o Se a ia I alica2; while in Asia (China) and in
he Caucasian egion (Geo gia) his is documen ed as ea ly as ca. 10000 yea s B.P.3–5. Based on he bo anical
e idence, i seems ha mille was i s in oduced in o Cen al and Eas e n Eu ope om he S eppe egions
du ing he Neoli hic pe iod6–8. B oomco n mille seems o ha e a i ed in he Medi e anean zone om he
no h o he no h-eas 9; in no he n I aly, mille appea s in ea ly B onze Age se lemen s (1700-1500 BC)2. The
ea lies eco ds o ox ail mille (Se a ia I alica L. Beau .) come om Peigang and Cishan (No h China) in he
6 h millennium BC3. In Eu ope he i s ca bonized seeds o ox ail mille appea in he 2nd millennium BC, om
B onze Age se lemen s in cen al Eu ope10 and F ance11. I is also epo ed om he la e B onze age Kas anas si e
(Macedonia, G eece)12. In he Nea Eas he ea lies e idence o ox ail mille cul i a ion da es back o he I on
Age (c. 600 BC) in Tille Höyük, sou h-eas Tu key7. Howe e , hese da a only indica e he p esence o mille , bu
do no e eal hei ole in he human die .
C4 plan s a e ypical o wa m clima es and hei abundance is highly co ela ed wi h clima ic ac o s, such as
empe a u e, p ecipi a ion and he deg ee o a idi y. In gene al, hey a e no p esen in en i onmen s whe e nigh
empe a u es a e lowe han 8 °C13,14. Nowadays in Eu ope C4 biomass is e y sca ce; howe e hese plan s a e
na u ally p esen in small p opo ions especially in he SW-Eu ope (e.g. Spain, Po ugal, SW-F ance and I aly)15.
1Depa men o Legal Medicine, Toxicology and Physical An h opology, Medicine Facul y, Uni e si y o G anada (UGR),
A . de la In es igación 11, 18016, G anada, Spain. 2Biogeochemical o S able Iso opes Labo a o y, Andalusian Ins i u e
o Ea h Sciences (IACT-CSIC-UGR), A . de las Palme as 4, 18100, A milla, G anada, Spain. 3Depa men o Geog aphy
and Te i o ial Sciences, A ea o P ehis o y, Uni e si y o Co doba, Plaza del Ca denal Salaza 3, 14071, Co doba, Spain.
Co espondence and eques s o ma e ials should be add essed o Z.L. (email: [email p o ec ed])
Recei ed: 07 Sep embe 2016
Accep ed: 14 No embe 2016
Published: 09 Decembe 2016
OPEN
www.na u e.com/scien i ic epo s/
2
Scien i ic RepoR s | 6:38817 | DOI: 10.1038/s ep38817
Be o e he onse o ag icul u e, he bi o es collagen iso opic δ
13C alues om S-I aly (32.6 o 13.3 ka B.P.) sugges
ha C4 biomass was p ac ically absen (calcula ed ege al biomass is a ound − 25‰ s V-PDB)16. E en du ing he
wa me pe iods o he La e Miocene, ossil iso opic da a indica e he absence o C4 biomass in cen al and sou h-
e n Eu ope17,18. Man has in oduced ce ain species, such as mille , wi h a plan ing s a egy in he wa m season
(i.e. summe ), possibly as a esponse o Holocene a id pe iods11.
The s able iso ope composi ion (δ
13C and δ
15N) is a e y e ec i e ool o s udy he ophic web in ecosys-
ems19–23. This me hodology is c ucial o s udying pas ecosys ems and paleodie s, because di ec obse a ion
is impossible24–27. In addi ion his echnique allows quan i a i e analysis, which e en in cu en ecosys ems is
di icul o ob ain wi h he classical me hods o obse a ion. Iso opic ca bon aceabili y su e s less change in he
s eps o he ood chain han iso opic ni ogen one. The e o e, he oo p in o p ima y p oduce s (e.g. plan s) is
clea e 28.
The impo ance o ca bon iso opes as a ace o he p esence o C4 plan s in he die , is ha a mosphe ic CO2
is ixed in o o ganic ma e by he enzyme PEP ca boxylase (C4 plan s) which disc imina es much less agains
13C han he enzyme RuBisCO (C3 plan s)29. In p esen day C3 plan s biomass show ypical δ
13C alues a ound
− 26‰ (V-PDB), while C4 plan s a e a ound − 11‰ (V-PDB)30,31, − 24.5‰ and − 9.5‰ (V-PDB), espec i ely in
p eindus ial age, because δ
13C alues o CO2 we e less nega i e by 1.5‰ compa ed o he p esen alues (− 8‰
V-PDB). This 15‰ di e ence be ween C3 and C4 plan s allows an accu a e calcula ion o he pe cen age o C4
plan s in he die . The δ
13Ccollagen alues o humans and animals a e p opo ional o he iso opic signal om he
base o he ood chain (plan s), bu en iched by 0.8 and 1.3‰ o he bi o es and ca ni o es, espec i ely32. The δ
15N o he collagen is also ela ed wi h he ood chain and can be used o e alua e he ophic sys em, because in a
pa icula ecosys em he e is a signi ican en ichmen in 15N be ween 3 and 5‰ o each ophic le el25–27.On he
o he hand δ
15N canno be used o dis inguish be ween C3 and C4 plan s consump ion.
The e is li le knowledge abou he nu i ional habi s o p o ohis o ical popula ions such as P e-Roman-Cel ic
ones. The e o e, we pe o med an iso opic s udy o un eil he die a y habi s o he Cel ic popula ion Cenomani
Gauls, om he nec opolis o Semina io Vesco ile in Ve ona (I aly) da ed be ween he 3 d o 1s cen u y BC33. This
nec opolis coun s wi h a minimum o 174 skele ons in a good s a e o p ese a ion, and he majo i y o hem a e
non-adul s (see Supplemen a y In o ma ion o de ailed desc ip ions o he a chaeological con ex ). Su p isingly,
we eco ded ela i ely high, i.e. less nega i es, δ
13Ccollagen alues in he human samples, and e en in some animals
ound in he nec opolis. In ou ini ial conside a ions, we suspec ed he p esence o p o ein in he die om
ma ine sou ces, o e en anomalous esh wa e sou ces (en iched 13C in DIC-Dissol ed Ino ganic Ca bon, “dead
ca bon” om ma ine ca bona es), o al e na i ely an ex ensi e consump ion o C4 plan s. The i s hypo hesis
seemed om he beginning imp obable, aking in o accoun he geog aphic loca ion o he nec opolis, which
was a om he sea (abou 120 km). Addi ionally, he less nega i e δ
13C collagen, − 10‰ (V-PDB), is highe
han wha would be expec ed o a signal om ma ine p ima y p oduc ion22. A second al e na i e explana ion
is a eshwa e opic web wi h an anomalous en iched 13C sou ce o ca bon (DIC) om dissolu ion o he -
mic decomposi ion o ca bona es. This geochemical scena io is p esen in some a eas o I aly34 in which na u al
wa e becomes en iched in CO2. This a chaeological si e was nea o he Adige Ri e ha uns h ough Ve ona.
The e o e, wa e samples om he Adige i e and i s ibu a ies we e collec ed and analysed. In addi ion, 14C
da ing was done on some human emains o de ec he p esence o “dead ca bon”.
The main objec i e o his s udy is o demons a e ha hese less nega i e δ
13C alues a e ela ed o he con-
sump ion o C4 plan s and e u e he imp obable hypo hesis o a signi ican ma ine o anomalous esh wa e
die a y in ake (en iched in 13C). A he same ime, we se ou o quan i y he p opo ion o C4 plan s in he die
and p opose an in e p e a ion abou he alimen a y habi s o Cel ic popula ions o I aly which up o now ha e
been elusi e.
Resul s
Iso opic alues o he human and animal sample om Ve ona. The human δ
13C alues ange
om − 20.2‰ o − 9.7‰ (V-PDB) wi h a mean alue o − 15.3‰ (± 2.2‰) (V-PDB) while δ
15N alues a e
be ween + 6.9‰ and + 12.9‰ (AIR), showing a mean alue o + 9.4‰ (± 1.3‰) (AIR) (Fig.1; Table1;
Supplemen a yTableS2). The a omic C/N a io o he samples alls inside he ange o 2.9–3.6, which co e-
sponds o ypical alues o well-p ese ed samples, as sugges ed by DeNi o35. Iso opic da a om animal bones
(Tables1 and 2) sugges s ha domes ica ed he bi o es had a die mainly based on C3 plan s, al hough one o hese
(VRAR-1 wi h − 17.2‰ V-PDB) showed a less nega i e δ
13C alue ha could sugges ha his animal pa ly ed
on C4 plan s. Dogs (n = 2) usually ea men’s le o e s, showing ela i ely high δ
13C alues (≈ − 13‰ V-PDB). One
o hem also has a ela i ely low δ
15N alue o an omni o e. This implies a die poo in animal p o eins, suppo -
ing he hypo hesis o an impo an in luence in hei die o human ood scabs, ich in (C4) plan s.
Iso opic alues o DIC o Adige’s i e and Radioca bon da a. The DIC δ
13C da a (Table3 and
Supplemen a yFigu eS1b) o he Adige i e anges be ween − 4.5‰ and − 5.0‰ (V-PDB); his, a e iso opic
ac iona ion p ocesses o pho osyn hesis in algae, would esul in alues anging om − 23.5‰ and − 24‰
(V-PDB) and consequen ly i would imply δ
13C alues in ish issues e y close o hese alues (≈ − 22.5‰ o
− 23‰ V-PDB). These da a would esul in a alue o human collagen esul ing om an exclusi ely ish die (an
ex eme ha would be ela i ely a e), be ween − 20‰ o − 22‰ (V-PDB), which is no compa ible wi h he leas
nega i e alues ound in his wo k. The es o δ
13C alues o DIC o o he ibu a ies, wells, e c. a e e en mo e
nega i e so hey also canno jus i y he highe δ
13C alues ound in collagen o human emains36.
On he o he hand, he age ob ained om 14C da ing o bone collagen om ou indi iduals ( hose wi h less
nega i e δ
13C alues) is consis en wi h ha ex apola ed (Table4) om he p elimina y a chaeological analysis o
he g a e goods36. This indica es he absence o “dead” ca bon (e.g. dissolu ion o ma ine ca bona e ocks), which
would lead o appa en ly much olde ages. Such dissol ed ino ganic ca bon (DIC) ha e δ
13C alues close o + 0‰
www.na u e.com/scien i ic epo s/
3
Scien i ic RepoR s | 6:38817 | DOI: 10.1038/s ep38817
o posi i e alues, which would also lead o an aqua ic ophic chain wi h less nega i e alues o δ
13C, his being
he only al e na i e o he consump ion o C4 plan s. The absence o “dead” ca bon in he humans collagen ules
ou his hypo hesis, and hus consequen ly suppo s he p esence o C4 plan s in he die o hese indi iduals.
Discussion
The iso opic esul s show he impo an ole ha C4 plan s had in he die o his Cel ic popula ion om Ve ona
(Fig.1). Gene ally, δ
13C alues o − 20‰ (V-PDB) o less nega i es would indica e some C4 plan con ibu ions.
Thus he as majo i y o s udied indi iduals (o e 90%) seem o include C4 plan s in a di ec way (o indi ec ly
h ough he consump ion o he bi o es ha ed on hem) in hei die . Addi ionally, adul emales (mean alue
o 48.8 ± 15.1%) show signi ican ly (Mann-Whi ney U es = 198.0; p = 0.008) highe consump ion o C4 plan s
han adul males (mean o 37.4 ± 16.8%) (Fig.2). We also ound a signi ican s a is ical di e ence in δ
15N alues
be ween sex in he adul g oup ( = 3.15 p ≤ 0.01): men display highe δ
15N alues compa ed o women (Fig.3).
Thus, he e is a clea di e en ia ion in he die acco ding o sex wi h a highe in ake o animal p o ein (mea and
mea p oduc s) o men, while he die o women included mo e ce eals and ege able p o eins. I is impo an o
Figu e 1. Sca e plo o he da a o δ15N and δ13C analyses o human and animal samples om Ve ona.
The δ
13C da a o human bone collagen ange be ween − 20.2 and − 9.7‰ (V-PDB), and e eals he signi icance
o C4 plan in he die , wi h less nega i e alues indica ing an almos exclusi e die o C4 plan s and he
mo e nega i e alues ep esen ing a die lacking C4 plan s. The ange o δ
15N alues ( om 6.9 o 12.9‰
AIR) compa ed wi h hose o animals sugges s a mixed e es ial die (mea o dai y p oduc s and ege al
oods u s).
δ15N‰ (AIR) δ13C‰(V-PDB)
N Min Max MSD N Min Max MSD
Humans 90 6.9 12.9 9.4 1.3 89 − 20.2 − 9.7 − 15.3 2.2
He bi o es 5 3.8 5.9 4.7 0.9 5 − 20.8 − 17.2 − 19.3 1.4
Omni o es 2 4.9 7.1 6.0 1.6 2 − 13.7 -13.2 − 13.4 0.3
Table 1. Summa y able wi h Numbe o samples (N), Minimum (Min), Maximum (Max), Mean (M) and
s anda d de ia ion (SD) o he iso opic alues (‰) o human and animal emains.
Sample Species δ15N ‰AIR δ13C ‰V-PDB Bone C/N %PP- C4
VRSV-92 Equid 5.9 − 19.4 long bone 2.9 9
VRSV-93 Canis 4.9 − 13.2 c anium 2.9 59
VRSV-94 Canis 7.1 − 13.7 c anium 3.3 55
VRSV-95 Equid 3.9 − 20.8 long bone 3.0 0
VRSV-96 O icap id 4.2 − 20.3 ibia 3.2 1
VRAR-1 He bi o e 5.6 − 17.2 long bone 3.2 26
VRAR-2 He bi o e 3.8 − 18.9 e eb a 3.2 12
Table 2. Iso opic alues (δ15N and δ13C) o he animal samples wi h he es ima ed pe cen age o C4 plan s
in hei die . %PP-C4: minimum pe cen ages o die based on C4 p ima y p oduc ion.
www.na u e.com/scien i ic epo s/
4
Scien i ic RepoR s | 6:38817 | DOI: 10.1038/s ep38817
no e, ha some adul s ela ed wi h high in akes o animal p o ein (highes alues in δ
15N) ha e less nega i e alues
in δ
13C (≈ − 12‰ s V-PDB, see o example indi iduals VRSV-12 and VRSV-24 in he Supplemen a yTableS2),
indica ing ha domes ic he bi o ous ha e also consumed la ge amoun o C4 plan s, al hough he ew he bi -
o ous s udied show a consump ion o C3 plan s. Howe e , wo dogs show δ
13C alues which a e e y simila o
hose o humans (Fig.1). Especially he younge dog (VRSV-94) p esen s a ypical alue o he middle-low animal
p o ein in ake o omni o es: δ
15N = + 7.1‰ (AIR). While he δ
13C alue o − 13.7‰ (V-PDB) is e y close o
hose o he humans wi h a C4 die (Table2). As men ioned be o e, a possible explana ion is ha dogs we e ed
wi h he impo e ished-p o ein was e om he meals o hei owne s.
DIC alues (Table3) om local me eo ic wa e show ela i ely nega i e δ
13C alues ( om − 4.5 o − 13.3‰ s
V-PDB; i.e. an aqua ic die should be app oxima ely be ween − 23.5 and − 32.3‰ s V-PDB) and 14C alues o he
human samples wi h less nega i e δ
13Ccollagen alues does no indica e he p esence o “dead ca bon”. Consequen ly,
his excludes a eshwa e die o explain he measu ed less nega i e iso opic alues o his pa icula Cel ic popu-
la ion and hus suppo s he hypo hesis o a la ge consump ion o C4 plan s in hei die . Fu he mo e, δ
15N alues
in humans (+ 9.4‰ AIR) plo in abou a one ophic le el abo e e es ial he bi o ous (+ 4.7‰ AIR)36, ule ou
a ma ine die , in ag eemen wi h he ela i e la ge dis ance om he coas .
Conside ing ha he e a e no many known cul i a ed C4 plan species in his ime pe iod (3 d-1s BC), he
hypo hesis o mille consump ion seems easonable based on he a chaeological eco d and ancien w i en
Sample ID wa e Localiza ion δ13CDIC ‰ (V-PDB)
1Alpone c eek Ves enano a (VR) − 9.7
2P ogno c eek Giazza (VR) − 4.6
3Tana delle sponde ca e Velo Ve onese (VR) − 4.9
4Bassanella sou ce Soa e (VR) − 13.3
5Well’s wa e (108 m) Gambella a (VI) − 10
6Adige i e Bolzano (BZ) − 5
7A esa c eek A esa (VR) − 9.2
8Adige i e Chie o dam (VR) − 4.5
9Fon ana del Fe o sou ce Ve ona − 11.7
Table 3. DIC (dissol ed ino ganic ca bon) iso opic composi ion o he wa e o he zone.
Sample δ13C ‰ V-PDB YEARS B.P. YEARS B.C. (2σ)LAB ID
VRSV-12 − 12.7 2080 ± 32 193–37 CAN-2880
VRSV-16 − 11.9 2164 ± 33 261–110 CNA-2881
VRSV-17 − 16 2113 ± 32 204–47 CNA-2882
VRSV-24 − 12.2 2149 ± 32 234–89 CNA-2883
Table 4. 14C da ing o he less nega i e δ13C human collagen samples.
Figu e 2. C4 plan s minimum pe cen ages (%) in he die o he indi iduals om Ve ona. Some indi iduals
show C4 plan s p opo ion abo e 60% in hei alimen a ion, and especially emale adul s each signi ican ly
highe pe cen ages o C4 plan s consump ion compa ed o adul males.
www.na u e.com/scien i ic epo s/
5
Scien i ic RepoR s | 6:38817 | DOI: 10.1038/s ep38817
sou ce. In ac , he in es iga ed nec opolis is loca ed in an a ea which is s ill oday cha ac e ized by a e ile plane
(called Pianu a Padana o Po alley): i o e s ideal clima ic and en i onmen al condi ions o he cul i a ion
o C4 plan s du ing he wa m seasons (sp ing-summe ). Pliny he Younge (Epis ulae, IV, 6,1) p aised he Po
Valley egion and w o e “in egione T anspadana summa abundan ia”37. Since he Middle and La e B onze age
(1550-1170 yea s BC) his plane was cha ac e ized by in ensi e ag icul u e, pas o alism and he demand o la ge
amoun s o wood o building si es. This lead o a hea y de o es a ion o he whole Po plain and con e ing i in o
an a i icial s eppe de o ed o ce eal c ops38. Fu he mo e, paleoclima e eco ds o lake sedimen s and speleo-
hems show ends owa ds a d ie clima e, o al e na ing we and d y clima es, be ween 2500 and 2000 BP39,40.
This migh ha e induced a change in c opping s a egies and he consequen in oduc ion o C4 plan cul i a ion
(p obably mille ).
Un o una ely, du ing he a chaeological exca a ion o Ve ona, a chaeologis s ha e no ound seeds (o cha -
coal) and he ce amic g a e goods ( essels and o he s) a e s ill unde s udy and ini ially seem no o indica e
mac oscopic emains o o ganic was e. Some p e ious a chaeobo anical s udies o Neoli hic, B onze and I on
Age si es o he NE o I aly, and pa icula ly o he a ea o Ve ona, seem o con i m he p esence o Panicum
Miliaceum and o Se a ia i alica among he ca pological emains41,42. Ano he compa a i e paleobo anical s udy
ela ing o some I on Age si es loca ed be ween he No he n and Sou he n Alps (Eas e n Swi ze land, Aus ia
and No he n I aly), also desc ibes he p esence o mille (Panicum Miliaceum L.) and ox ail mille (Se a ia
I alica L. Beau .) emains43. This is also con i med by he esul s o he s udy abou ca pological emains ound
a he si e o Oppeano (Ve ona), da ed o he second I on Age (abou 6 h o 3 d cen u y BC) and loca ed in he
same geog aphical con ex o he Semina io Vesco ile nec opolis. The majo i y o he de e mined ce eal emains
co espond o mille (Panicum miliaceum L.) and ox ail mille (Echinochloa c us-galli L. Beau . and Se a ia i alica
L. Beau )44. Thus, he I on Age seems o be cha ac e ized by a g ea e c ops specializa ion ha i s wi h he
assump ion o he pu sui o a ce eal ype ha bes i s di e en clima es (biomass inc ease wi h mo e e iciency
in wa e use). In his case o he sandy Pianu a Padana (Ve ona), he c ops o mille we e p e e ed, because i is a
plan ha adap s e y well o poo subs a es cha ac e ized by wa e sca ci y du ing summe 44.
Finally, o he iso opic esul s om bone collagen analysis o some indi iduals p oceeding om he B onze
Age nec opolis o Olmo di Noga a, Ve ona (1600-1200 cal. BC) and o A ano di Cello e, Ve ona (2040-1890 cal.
BC) suppo he hypo hesis o C4 plan s consump ion, and in pa icula o mille in some a eas o he Ve ona
p o ince45,46(Fig.4). In conclusion, o hese a chaeological da a we add he es imonies o some ancien au ho s
such as Pliny he Elde (Na u alis His o iae XVIII, 83-84) and Columella (De Re Rus ica, 2, 9, 14–16) who epo
he use o mille lou o he p oduc ion o b ead and a so o po idge cooked in wa e and sal and o en
accompanied wi h ege ables and cheese and e y a ely wi h mea . Pliny (XVIII.XXIV) exac ly no ed: “mille is
used o p epa e a e y whi e puls (i.e. simila o p esen -day polen a). Panic, when g ound and eed om b an, and
mille as well, makes a po idge which, especially wi h milk, is no o be despised e en in ime o plen y”. Columella
(2.9.19), ag eed wi h Pliny and w o e: “b ead is made o mille , and i may be ea en wi hou dis as e be o e i cools”.
The iso opic da a ob ained in his s udy e eal ha he p opo ion o C4 plan s was subs an ial (abo e ≈ 40%)
in he daily die o his Cel ic popula ion. T aces o C4 plan consump ion a e epo ed in o he s udies45,46 om
mo e ancien nec opolis o he zone. While Olmo de Noga a (ODN) da ed o Middle B onze Age sha e e y sim-
ila iso opic alues wi h Ve ona (Fig.4), con i ming a p eponde an C4 plan s based die , A ano di Cello e (AC),
Figu e 3. δ15N alues o adul emales and males. Adul males ha e highe δ15N alues compa ed o adul
emales.

www.na u e.com/scien i ic epo s/
6
Scien i ic RepoR s | 6:38817 | DOI: 10.1038/s ep38817
da ed o Ea ly B onze Age, shows a die based on he consump ion o C3 ce eal- ype plan s. Hence, he da a om
AC and ODN i mly place he shi in C4 c op use in he egion a a ansi ion pe iod be ween he la e phases o
he Ea ly B onze Age and he beginning o he Middle B onze Age46, consequen ly ou da a show ha succes-
si ely, in p e-Roman imes, he die o mos habi an s o he zone was almos exclusi ely based on C4 plan s. This,
indi ec ly indica es ha his Cel ic popula ion had mas e ed ag icul u e echniques o such an ex en ha hey
could li e o he ha es o hei cul i a ed c ops he yea ound.
Me hods
A sample o 90 human ibs, consis ing o bo h sexes and di e en ages (Supplemen a yIn o ma ionTableS1),
and 7 animal bones, was selec ed o analysis o bone collagen. The animal bone sample is composed o 5 he bi -
o ous species (2 ho ses, 1 goa /sheep and 2 cows) and 2 omni o ous species (2 dogs). The ex ac ion o collagen is
pe o med using he p o ocol desc ibed by Boche ens e al.47,48 and ollowing he ou ine p ocedu es o he S able
Iso ope Biogeochemis y Labo a o y o he Andalusian Ins i u e o Ea h Sciences (CSIC, G anada, Spain). We
also de e mined he cu en DIC alues o he Adige’s i e (and o some o i s ibu a ies) and da ed wi h 14C he
collagen samples o he less nega i e human alues (a sou ce o “dead” ino ganic ca bon would show abno mally
high ages).
The δ
13C mean alues o p esen day C3 and C4 plan s a e espec i ely − 26‰ and − 11‰ (V-PDB)49,50, bu
be o e he indus ial e olu ion he iso opic composi ion o a mosphe ic CO2 was 1.5‰ less nega i e51. Hence,
o make a mo e p ecise es ima e we conside ed ha hese wo ca bon sou ces would espec i ely co espond o
alues o −24.5‰ and −9.5‰ (V-PDB) (see Supplemen a y In o ma ion o de ailed desc ip ions abou he
di e en me hodologies applied).
Re e ences
1. Van de Me we, N. J. Ca bon iso opes, pho osyn hesis and a chaeology: di e en pa hways o pho osyn hesis cause cha ac e is ic
changes in ca bon iso ope a ios ha make possible he s udy o p ehis o ic human die s. Am. Sci. 70, 596–606 (1982).
2. Zoha y, D. & Hop , M. Domes ica ion o plan s in he Old wo ld. (Cla endon P ess, Ox o d, 1993).
3. An, Z. P ehis o ic ag icul u e in China. in Fo aging and a ming: he e olu ion o plan exploi a ion (eds Ha is, D. R. & Hillman, G. C.)
643–649 (Unwin & Hyman, London, New Je sey, 1989).
4. Lisi sina, G. N. The Caucasus- A cen e o ancien a ming in Eu asia. in Plan s and ancien man (eds Van Zeis , W. & Caspa ie, W. A.)
293–305 (Balkema, Ro e dam, 1984).
5. Cheng-Bang, A. e al. Va iabili y o he s able ca bon iso ope a io in mode n and a chaeological mille s: e idence om No he n
China. J. A chaeol. Sci. 53, 316–322 (2015).
6. Hjelmq i, H. Spel and Mille om B onze Age o Sou he n Sweden and some obse a ions on i s la e his o y in Sweden. Bo .
No ise . 122, 260–283 (1969).
7. Nesbi , M. & Summe s, G. D. Some ecen disco e ies o mille s (Panicum miliaceum and Se a ia i alica) a exca a ions in Tu key
and I an. Ana olian S ud. 38, 85–97 (1988).
8. Mu phy, E. M. e al. I on Age pas o al nomadism and ag icul u e in he eas e n Eu asian s eppe: implica ions om den al
palaeopa hology and s able ca bon and ni ogen iso opes. J. A chaeol. Sci. 40, 2547–2560 (2013).
9. Van Zeis , W. Ape çu su la di usion des égé aux cul i és dans la égion Médi e anéene. in Colloque su la mise en place, l’é olu ion
e la ca ac é isa ion de la lo e e la égé a ion ci cummédi e anéenne (Funda ion L. Embe ge ), 129–145 (Na u alia Monspelensia,
Mon pellie , 1980).
10. Ne oli zky, F. Die Hi se aud al en Funden. Si zungsbe . K. Akad. Wiss. Wien. 73, 1–35 (1914).
Figu e 4. S able ca bon and ni ogen iso ope da a o human bone collagen om Ve ona (VRSV) compa ed
wi h he indi idual’s alues om Olmo di Noga a (ODN)45 and A ano di Cello e (AC)46. We ha e also
plo ed he aunal anges om he h ee nec opolises. While Ve ona and Olmo di Noga a (Middle B onze Age)
ha e e y simila iso opic alues, con i ming a la gely C4 plan s based die , A ano di Cello e, da ed o Ea ly
B onze Age shows a die based on C3 ce eal- ype plan s.
www.na u e.com/scien i ic epo s/
7
Scien i ic RepoR s | 6:38817 | DOI: 10.1038/s ep38817
11. Goude, G., Rey, L., Toulemonde, F., Ce el, M. & Ro ie , S. Die a y changes and mille consump ion in no he n F ance a he end
o P ehis o y: e idence om a chaeobo anical and s able iso ope da a. En i on. A chaeol. 21, doi: 10.1080/14614103.2016.1215799
(2016).
12. K oll, H. J. Die P lanzen unde. in Kas anas. Ausg abungen in einem Siedlungshügel de B onze- und Eisenzei Makedoniens 1975-
1979. P ähis o ische A chäologie in Südos eu opa Vol. 2 (Spiess, Be lin, 1983).
13. Tee i, J. A. & S owe, L. G. Clima ic pa e ns and he dis ibu ion o C4 g asses in No h Ame ica. Oecol. 23, 1–12 (1976).
14. Ehle inge , J. R., Ce ling, T. E. & Hellike , B. R. C4 pho osyn hesis, a mosphe ic CO2, and clima e. Oecol. 112, 285–299 (1997).
15. Pyanko , V. I., Ziegle , H., Akhani, H., Deigele, C. & Lu ge, U. Eu opean plan s wi h C 4 phosyn hesis: geog aphical and axonomic
dis ibu ion and ela ions o clima e pa ame e s. Bo . J. Linn. Soc. 163, 283–304 (2010).
16. Iacumin, P., Boche ens, H. & Hue as, A. A s able iso ope s udy o ossil mammal emains om he Paglicci ca e, Sou he n I aly. N
and C as palaeoen i onmen al indica o s. Ea h. Plane Sci. Le . 148, 349–357(1997).
17. Domingo, L. e al. La e neogene and ea ly qua e na y paleoen i onmen al and paleoclima ic condi ions in sou hwes e n Eu ope:
iso opic analyses on mammalian axa. PLoS ONE 8(5), e63739 (2013).
18. Tü ken, T., Kaise , T. M., Vennemann, T. & Me ce on, G. Oppo unis ic eeding s a egy o he ea lies old wo ld hypsodon equids:
e idence om s able iso ope and den al wea p oxies. PLoS ONE 8(9), e74463 (2013).
19. Ande son, W. B. & Polis, G. A. Nu ien luxes om wa e o land: seabi ds a ec plan nu ien s a us on Gul o Cali o nia islands.
Oecol. 118, 324–332 (1999).
20. Bal e , V., Simon, L., Fouille , H. & Lecuye , C. Box-modeling o N-15/N-14 in mammals. Oecol 147, 212–222 (2006).
21. Cabana, G. & Rasmussen, J. B. Modelling ood chain s uc u e and con aminan bioaccumula ion using s able ni ogen iso opes.
Na u e 372, 255–257 (1994).
22. Chisholm, B. S., Nelson, D. E. & Schwa cz, H. P. S able-ca bon iso ope a ios as a measu e o ma ine e sus e es ial p o ein in
ancien die s. Science 216, 1131–1132 (1982).
23. Vande Zanden, M. J., Casselman, J. M. & Rasmussen, J. B. S able iso ope e idence o he ood web consequences o species
in asions in lakes. Na u e 401, 464–467 (1999).
24. Boche ens, H., D ucke , D. G. & Madelaine, S. E idence o a 15N posi i e excu sion in e es ial oodwebs a he Middle o Uppe
Palaeoli hic ansi ion in sou h-wes e n F ance: implica ions o ea ly mode n human palaeodie and palaeoen i onmen . J. Hum.
E ol. 69, 31–43 (2014).
25. Amb ose, S. H. & DeNi o, M. J. Bone ni ogen iso ope composi ion and clima e. Na u e 325, 201–201 (1987).
26. DeNi o, M. J. Pos mo em p ese a ion and al e a ion o in i o bone-collagen iso ope a ios in ela ion o paleodie a y
econs uc ion. Na u e 317, 806–809 (1985).
27. Hedges, R. E. M. & Reyna d, L. M. Ni ogen iso opes and he ophic le el o humans in a chaeology. J. A chaeol. Sci. 34, 1240–1251
(2007).
28. Pos , D. M. Using s able iso opes o es ima e ophic posi ion: Models, me hods, and assump ions. Ecology 83, 703–718 (2002).
29. Bende , M. M. Va ia ions in C-13/C-12 a ios o plan s in ela ion o pa hway o pho osyn he ic ca bon dioxide ixa ion.
Phy ochemis y 10, 1239–1244 (1971).
30. Ce ling, T. E. The s able iso opic composi ion o mode n soil ca bona e and i s ela ionship o clima e. Ea h Plane . Sci. Le . 71,
229–240 (1984).
31. Pe e son, B. J. & F y, B. S able iso opes in ecosys em s udies. Annu Re Ecol Sys em 18, 293–320 (1987).
32. Boche ens, H. & D ucke , D. T ophic le el iso opic en ichmen o ca bon and ni ogen in bone collagen: case s udies om ecen
and ancien e es ialecosys ems. In . J. Os eoa chaeol. 13, 46–53 (2003).
33. La anchi, Z., Ma ín Fló ez, J. S., Jiménez B obeil, S. A. & Cas ellani, V. 2015. Foo polydac yly and bipa i e medial cunei o m: a
case o co-occu ence in a Cel ic skele on om Ve ona (I aly). Homo. 66, 216–228 (2015).
34. Di Benede o, F. e al. Bio ic and ino ganic con ol on a e ine deposi ion a Bullicame 3 sp ing (Vi e bo, I aly): A mul idisciplina y
app oach. Geochim Cosmochim Ac 75, 4441–4455 (2011).
35. DeNi o, M. J. Pos mo em p ese a ion and al e a ion o in i o bone collagen iso ope a ios in ela ion o palaeodie a y
econs uc ion. Na u e 317, 806–809 (1985).
36. La anchi, Z. An opología de la población P e-Romana de Ve ona (I alia). PhD Thesis. (Uni e sidad de G anada, G anada, 2015).
37. Mu phy, C. Finding mille in he Roman wo ld. A chaeol. An h opol. Sci. 8, 65–78 (2016).
38. C emaschi, M., Me cu i, A. M., To i, P., Flo enzano, A. & Pizzi, C. Clima e change e sus land managemen in he Po Plain
(No he n I aly) du ing he B onze Age: New insigh s om he VP/VG sequence o he Te ama a San a Rosa di Po iglio. Qua . Sci.
Re . 136, 153–172 (2016).
39. McDe mo , F. e al. Holocene clima e a iabili y in Eu ope: E idence om del a O-18, ex u al and ex ension- a e a ia ions in
h ee speleo hems. Qua . Sci. Re . 18, 1021–1038 (1999).
40. Magny, M. e al. No h-sou h palaeohyd ological con as s in he cen al Medi e anean du ing he Holocene: en a i e syn hesis
and wo king hypo heses. Clim Pas . 9, 2043–2071 (2013).
41. Pisoni, L. Aspe i e p oblemi dell’occupazione del e i o io, dell’o ganizzazione sociale e dell’economia ag o-pas o ale nell’e à del
e o a esina: un’in oduzione allo s udio dei sis emi alimen a i. P eis o ia Alpina 44, 225–243 (2009).
42. Ro oli, M. C op di e si y be ween Cen al Eu ope and he Medi e anean: aspec s o no he n I alian ag icul u e in Plan s and
people: choices and di e si y h ough ime (eds. Che alie , A. e al.) 75–81 (Oxbox book, Ox o d, 2014).
43. Schmidl, A., Jacome , S. & Oeggl, K. Dis ibu ion pa e ns o cul i a ed plan s in he Eas e n Alps (Cen al Eu ope) du ing I on Age.
J. A chaeol. Sci. 34, 243–254 (2007).
44. Ca a, M. L. Pe una s o ia della ce ealicol u a in I alia se en ionale dal Neoli ico all’e à del Fe o: s a egie ada i e e condizionamen i
ambien ali. PhD Thesis. (Alma Ma e S udio um, Bologna, 2012).
45. Ta u i, M. A., C aig, O. E. & Canci, A. S able iso ope e idence o consump ion o mille and o he s plan s in B onze Age I aly. Am. J.
Phys. An h opol. 139, 146–153 (2009).
46. Va alli, A. e al. Die a y con inui y s. discon inui y in B onze Age I aly. The iso opic e idence om A ano di Cello e (Illasi, Ve ona,
I aly). J. A chaeol. Sci: Rep. 7, 104–113 (2016).
47. Boche ens, H. e al. Iso opic biogeochemis y (13C, 15N) o ossil e eb a e collagen: applica ion o he s udy o a pas ood web
including Neande hal man. J. Hum. E ol. 20, 481–492 (1991).
48. Boche ens, H. e al. Paleobiological implica ions o he iso opic signa u es (13C, 15N) o ossil mammal collagen in Scladina ca e
(Sclayn, Belgium). Qua . Res. 48, 370–380 (1997).
49. Bende , M. M. Mass spec ome y s udies o ca bon-13 a ia ions in co n and o he g asses. Radioca bon 10, 468–472 (1968).
50. O’Lea y, M. H. Ca bon iso opes in pho osyn hesis. BioScience 38, 328–336 (1988).
51. F iedli, H., Lö sche , H., Oeschge , H., Siegen hale , U. & S au e , B. Ice co e eco d o he 13C/12C a io o a mosphe ic CO2 o he
pas wo cen u ies. Na u e 324, 2 (1986).
Acknowledgemen s
The au ho s would like o hank D . Giuliana Ca alie i Manasse o he Sop in endenza A cheologica o Ve ona
o p o ide he oppo uni y o s udy his skele al collec ion and he a chaeologis s Simon Thompson and
Ma zia Be sani o he use o hei a chaeological da a. The iso opic analysis we e unded by he P ojec s RNM-
www.na u e.com/scien i ic epo s/
8
Scien i ic RepoR s | 6:38817 | DOI: 10.1038/s ep38817
8011(Jun a de Andalucía, Spain), HAR 2016-75788 (Minis e io de Economía y Compe i i idad, Spain) and he
esea ch g oup RNM-309 (Jun a de Andalucía, Spain). We also wish o exp ess ou g a i ude o Lau a Bigno o
(independen esea che ) o he assis ence wi h he wa e sampling o Ve ona a ea and Alexande E.S. Van
D iessche (ISTe e, G enoble, F ance) o p oo eading he English e sion o he manusc ip . The commen s
and sugges ions om he Edi o Jens O mö and wo anonymous e iewe s g ea ly imp o ed he pape .
Au ho Con ibu ions
Z.L. and A.D.H. concei ed he objec i es and designed he p ojec . Z.L. pe o med he an h opological s udy
while S.J.B. supe ised he an h opological esul s. Z.L. and J.A.R.C. pe o med he ex ac ion o collagen om
he bone samples. A.G.T. and A.D.H. pe o med he s able iso opes analysis (bone collagen, DIC alues, 14C
da a). Z.L., A.D.H. and S.J.B. we e esponsible o in e p e ing he da a. A.G.T. and J.A.R.C. con ibu ed o he
discussion. Z.L. and A.D.H. w o e he manusc ip . All au ho s e iewed he manusc ip .
Addi ional In o ma ion
Supplemen a y in o ma ion accompanies his pape a h p://www.na u e.com/s ep
Compe ing inancial in e es s: The au ho s decla e no compe ing inancial in e es s.
How o ci e his a icle: La anchi, Z. e al. S able C & N iso opes in 2100 Yea -B.P. human bone collagen
indica e a e die a y dominance o C4 plan s in NE-I aly. Sci. Rep. 6, 38817; doi: 10.1038/s ep38817 (2016).
Publishe 's no e: Sp inge Na u e emains neu al wi h ega d o ju isdic ional claims in published maps and
ins i u ional a ilia ions.
This wo k is licensed unde a C ea i e Commons A ibu ion 4.0 In e na ional License. The images
o o he hi d pa y ma e ial in his a icle a e included in he a icle’s C ea i e Commons license,
unless indica ed o he wise in he c edi line; i he ma e ial is no included unde he C ea i e Commons license,
use s will need o ob ain pe mission om he license holde o ep oduce he ma e ial. To iew a copy o his
license, isi h p://c ea i ecommons.o g/licenses/by/4.0/
© The Au ho (s) 2016