Contents lists available at ScienceDirect
Journal of Archaeological Science
journal homepage: www.elsevier.com/locate/jas
Combined influence of meteoric water and protein intake on hydrogen
isotope values in archaeological human bone collagen
Christine A.M. Francea,∗, Haiping Qib, Gwénaëlle M. Kavicha
a Smithsonian Museum Conservation Institute, 4210 Silver Hill Rd., Suitland, MD, USA
bUnited States Geological Survey, 12201 Sunrise Valley Dr., Reston, VA, USA
A R T I C L E I N F O
Keywords:
Bone
Collagen
Hydroxyapatite
Hydrogen
Oxygen
Nitrogen
Isotopes
A B S T R A C T
Hydrogen isotopes in archaeological human bone collagen are poorly understood, but present an opportunity to
add new depth to our understanding of ancient populations. The competing influences of meteoric water versus
protein intake on human bone collagen hydrogen isotope values were examined through comparison with the
well-understood proxies of hydroxyapatite oxygen and collagen nitrogen isotopes, respectively. Consideration of
the data set as individual points compared to averaged pools of individuals in each of 11 archaeological sites
suggested the latter partially eliminates inherent variability due to food choice or regional movement. Collagen
hydrogen isotopes were moderately correlated with hydroxyapatite oxygen isotopes (R= 0.695, site averages)
and collagen nitrogen isotopes (R=0.562, site averages). Correlation improved with a multiple linear regres-
sion including both oxygen and nitrogen (R=0.745, site averages). Correlation between meteoric water hy-
drogen and oxygen isotope values converted from hydroxyapatite and collagen values, respectively, yielded a
slope well below the expected value of ∼8 observed directly in meteoric water (i.e. the “meteoric water line”).
Correlation between converted meteoric water hydrogen and the measured collagen non-exchangeable hydrogen
isotope values showed a slope well below the expected value of 1.0. Theoretical meteoric water hydrogen isotope
values and theoretical herbivorous collagen hydrogen isotope values were calculated based on previously es-
tablished equations in order to construct a hypothetical framework free of trophic level influences. Deviations
between actual values and these theoretical values correlated weakly with collagen nitrogen isotope values,
suggesting that direct trophic level enrichment/depletion is not controlling the disparity between expected and
measured values. The deviations are hypothetically caused by non-local food sources, and a decoupling of ex-
pected oxygen and hydrogen relationships as individuals consumed more meat and decreased in vivo non-
essential amino acid production. This work presents a new model that facilitates understanding of the complex
relationship between meteoric water and protein intake controls on hydrogen isotopes in omnivorous human
populations that can potentially inform about past meteoric water values and amounts of animal protein con-
sumption.
1. Introduction
Stable isotope analysis of bones is relatively common in archaeology
and paleontology to determine dietary components, provenance, mi-
grations, climate proxies, metabolic functioning, and social demo-
graphics. Several decades of research have established a solid under-
standing of stable carbon, nitrogen, and oxygen isotope dynamics in
archaeological bone collagen and hydroxyapatite. Hydrogen isotopes in
bone have been addressed only recently. The routing of hydrogen into
bone collagen in particular is less well-understood, but presents new
options for understanding archaeological remains.
Hydrogen isotopes have been examined more thoroughly in tissues
which are similar to the collagen protein and can serve as basic com-
parisons. Keratin (i.e. feathers, claws, nails) has been studied most
heavily, although blood, muscle, lipids, and other organ tissues have
been examined as well (Chesson et al. 2009, 2011; Hobson et al., 1999;
Tuross et al., 2008; Wolf et al., 2011). Hydrogen is routed to keratin
from both dietary food and drinking water, where the former pathway
provides trophic information and the latter indicates latitudinal pro-
venance (Bowen et al., 2005, 2009; Ehleringer et al., 2008; O'Brien and
Wooller, 2007; Sellick et al., 2009). Where some studies suggest keratin
hydrogen isotopes largely reflect drinking water isotope values (Hobson
et al., 1999, Wolf et al., 2011), others suggest secondary dietary hy-
drogen input as well (Bowen et al., 2009; Ehleringer et al., 2008;
https://doi.org/10.1016/j.jas.2018.05.011
Received 8 August 2017; Received in revised form 22 May 2018; Accepted 26 May 2018
∗ Corresponding author.
E-mail address: francec@si.edu (C.A.M. France).
Journal of Archaeological Science 96 (2018) 33–44
Available online 01 June 2018
0305-4403/ Published by Elsevier Ltd.
T
Kirsanow and Tuross, 2011; Pietsch et al., 2011). Bulk blood, muscle,
lipid, and organ hydrogen isotope values reflect largely drinking water
sources (Chesson et al., 2011; Hobson et al., 1999; Wolf et al., 2011),
although dietary input may have some influence (Commerford et al.,
1983). While this previous research provides background for under-
standing hydrogen isotopes in bone collagen, keratin and other tissues
have a more rapid turnover, a fundamentally different structure, and
potentially different hydrogen sources rendering them inadequate
proxies for collagen.
Collagen is the primary protein in animal bones and includes hy-
drogen atoms bound to carbon, or bound within carboxyl, amide, and
minimal sulfhydryl side-groups. These side-group hydrogen atoms are
labile and exchange with hydrogen from other water sources. The more
stable carbon-bound hydrogen atoms comprise a calculated fraction of
0.742–0.829 (majority 0.77–0.81) of all hydrogen atoms in collagen
(Cormie et al. 1994b, 1994c; Leyden et al., 2006; Sauer et al., 2009;
Topalov et al., 2013) and are generally non-exchangeable with external
water sources. The total (i.e. TOT) hydrogen isotope composition of
bone collagen (i.e. COLL) can be represented as
δ2HCOLL-TOT = (1-f) *δ2HCOLL-NEX + f *δ2HCOLL-EX
where f represents fraction of exchangeable hydrogen (i.e. ∼0.19-
0.23), δ2HCOLL-NEX represents the isotope value of non-exchangeable
hydrogen atoms, and δ2HCOLL-EX represents the isotope value of ex-
changeable hydrogen atoms. Isotope values are in standard delta no-
tation:
δX= [(Rsample – Rstandard)/Rstandard]
where R is the ratio (i.e. 2H/1H), values are in parts per thousand (‰),
and the standard is V-SMOW.
The δ2HCOLL-NEX represents the isotope signal incorporated via water
or dietary food and can be considered with a general conceptual fra-
mework:
δ2HCOLL-NEX = (δ2Hingested water + εa) + (δ2Hdietary amino acids + εb).
The δ2Hingested water represents δ2H of water taken into the body via
food water or direct drinking and incorporated into amino acids syn-
thesized in vivo during collagen construction (i.e. “non-essential”
amino acids). The εa represents hydrogen isotope fractionation during
this process. The δ2Hdietary amino acids represents δ2H of amino acids
synthesized ex vivo and are incorporated directly from consumed
dietary proteins (i.e. “essential” amino acids). The εb represents sub-
sequent fractionation as these amino acids are incorporated into col-
lagen, although this value is suspected to be minimal and constant
within a given species. Cormie et al. (1994a), Cormie et al. (1994c) and
Chesson et al. (2011) present a thorough review of factors contributing
to bone collagen δ2Hingested water, εa, δ2Hdietary amino acids, and εb; addi-
tional insight is gained from detailed discussions of keratin hydrogen
incorporation (Ehleringer et al., 2008; Bowen et al., 2009).
A relatively strong linear correlation between δ2Hingested water and
δ2HCOLL-NEX exists in strict herbivores obtaining all dietary fractions
(i.e. amino acids, carbohydrates, water) from plants. Since leaf and
stem δ2H values reflect local precipitation δ2H values, the herbivore
δ2HCOLL-NEX correlates with these local precipitation δ2H values (Cormie
et al. 1994a, 1994c; Pietsch et al., 2011; Reynard and Hedges, 2008).
Hydrogen isotope values in herbivore bone collagen can be considered
with the simpler representation of δ2HCOLL-NEX= δ2Hingested water + εa.
Carnivores tend to show an apparent trophic level effect where δ2HCOLL-
NEX deviates from the expected correlation with δ2Hingested water
(Birchall et al., 2005; Pietsch et al., 2011; Reynard and Hedges, 2008;
Topalov et al., 2013; Tuross et al., 2008). This is due likely to the ad-
ditional δ2Hdietary amino acids variable which can show considerable range
depending on the type and amount of animal protein consumed.
Humans present a complex case of omnivory. Limited research
examining human collagen δ2H values suggests a combination of in-
gested water and dietary input (Reynard and Hedges, 2008), which
agrees with limited data from other omnivorous mammals (Reynard
and Hedges, 2008; Tuross et al., 2008). As archaeological human re-
mains are of high interest, examining human collagen δ2H could pro-
vide another dimension by which to examine dietary input and inges-
tion of environmental water in a uniquely coupled pathway. It has the
potential to contribute additional information to the study of geo-
graphic origin, migrations, and dietary choices or available foods.
This study uses the well-known relationships of bone nitrogen and
oxygen with trophic structure and meteoric water, respectively, to ex-
plore these mechanisms’ effects on δ2HCOLL-NEX. Nitrogen in collagen
(i.e. δ15NCOLL) is represented in standard delta notation as indicated
previously where R is15N/14N and the standard is atmospheric air. The
δ15NCOLL increases approximately 3–4‰ with trophic level (Bocherens
and Drucker, 2003; DeNiro and Epstein, 1981; Schoeninger and DeNiro,
1984) providing a proxy for amount and type of dietary protein intake.
Oxygen is found in the hydroxyapatite mineral fraction of bone in both
the phosphate (i.e. PHOS) and carbonate (i.e. CARB) sites. Phosphate
and carbonate oxygen isotopes (i.e. δ18OPHOS and δ18OCARB) are re-
presented in standard delta notation where R is 18O/16O and the stan-
dard is V-SMOW. Both δ18OPHOS and δ18OCARB correlate with drinking
water isotopes (Bryant and Froelich, 1995, Daux et al., 2008, Kohn,
1996, Longinelli, 1984, Luz and Kolodny, 1985, Luz et al., 1984) pro-
viding a proxy for geographic locality. The δ18O and δ2H values of
meteoric water (i.e. MW) are strongly correlated according to the
known meteoric water line: δ2HMW = 8 *δ18OMW +10 (Craig, 1961;
Kendall and Coplen, 2001). In the absence of dietary influence, the
δ2HCOLL-NEX is expected to correlate to δ18OPHOS and δ18OCARB with a
similar slope to that of the meteoric water line. Deviations from this end
member were compared to associated δ15NCOLL values, and multiple
linear regression models constructed to determine the combined re-
lative influence of ingested water and dietary proteins on the δ2HCOLL-
NEX values. Combinations of δ2H and δ18O values in bone collagen have
been used to examine herbivores, but this study adds to the sparser
comparisons with δ15N, omnivores, and carnivores (Cormie et al.,
1994a; Kirsanow and Tuross, 2011; Kirsanow et al., 2008; Pietsch et al.,
2011; Topalov et al., 2013; Tuross et al., 2008).
2. Materials and methods
2.1. Sample collection and preparation
Human remains were sampled from 11 North American archae-
ological sites primarily on the east coast with one southern site in-
cluding individuals from Texas (Fig. 1, Table 1, Supplementary Table
S1). These sites were selected based on availability of samples, range of
geographic localities, and range of potential protein consumption. The
sites are primarily temperate regions with similar humidity and tem-
perature conditions. The exception is Glorieta Pass wherein the in-
dividuals hailed from the warmer dryer regions of Texas (Alberts,
1984). Carbon, nitrogen, and oxygen isotope data for some samples
were published previously in France et al. (2014) and France and
Owsley (2015).
Mechanical and chemical preparation methods followed France
et al. (2014). Briefly, ∼500mg of solid bone cross section (majority
cortical with traces of trabecular) was removed for collagen analysis
using pliers or a rotary tool. This cross section yields a homogenized
average isotope value across the final ∼10–20 years of life. Approxi-
mately 50mg of powdered bone for phosphate and carbonate analysis
was obtained by crushing with an agate mortar and pestle or using a
rotary tool. Phosphates were extracted via dissolving mineral phases in
hydrofluoric acid (2M), buffering in ammonium hydroxide (20%), and
precipitating silver phosphate using a silver nitrate solution (2M).
Carbonates were isolated by eliminating organics with sodium hypo-
chlorite (2–3%) and eliminating secondary carbonates using acetic acid
C.A.M. France et al. Journal of Archaeological Science 96 (2018) 33–44
34
buffered with calcium acetate (pH ∼4.5). Collagen extraction pro-
ceeded via sonication to remove sediments and labile salts, acidification
(0.6 M HCl, 4 °C) to remove mineral phases, removal of humic and
fulvic acids with sodium hydroxide (0.125M), denaturing of the col-
lagen pseudomorph in hydrochloric acid (0.03M, 95 °C), and lyophili-
zation.
2.2. Mass spectrometry and ATR-FTIR
All isotope ratios were measured on Thermo Delta V Advantage
stable isotope mass spectrometers at the Smithsonian MCI Stable
Isotope Mass Spectrometry Laboratory. Silver phosphates weighed into
silver cups (∼500 μg) were thermally decomposed (1450 °C) on a
Thermo Temperature Conversion Elemental Analyzer (TCEA) coupled
to a Conflo IV interface and measured for δ18OPHOS values. Carbonates
(∼4mg) were acidified in 100% phosphoric acid (SG > 1.92) at 25 °C
for 24 h on a Thermo GasBench II unit and measured for δ18OCARB va-
lues. Approximately 500 μg of collagen weighed into tin cups was
combusted (1020 °C) on a Costech 4010 Elemental Analyzer coupled to
a Conflo IV interface and measured for δ15NCOLL, weight % N, and
weight % C values.
A separate portion of collagen (∼350 μg) was weighed into silver
cups for hydrogen isotope analysis. Open cups remained in ambient air
for 72 h to equilibrate exchangeable hydrogen atoms with local water
vapor. Open cups were then placed in a vacuum oven at 60 °C for 72 h
to remove secondary adhered water molecules. The vacuum oven was
vented with pure argon before cups were removed, quickly sealed, and
loaded into a Costech zero-blank autosampler then flushed with ultra-
pure helium. Exposure to atmosphere was< 10min. Samples were
thermally decomposed on the TCEA with a chromium reactor column at
1100 °C (modified from Armbruster et al., 2006, Gehre et al., 2015,
Kelly et al., 2001). Resulting H2 gas was introduced to the mass spec-
trometer via a Conflo IV interface and measured for raw δ2H values.
Attenuated total reflectance Fourier transform infrared spectroscopy
(ATR-FTIR) was performed on bone powders using a Thermo Nicolet
6700 FTIR with Golden Gate ATR (diamond crystal, single bounce, 45°)
equipped with a DTGS detector. Spectra were collected from 450 to
4000 cm−1 for 128 scans at a resolution of 4 cm−1. All baseline cor-
rections and ratio calculations were performed using an automated
program in TQAnalyst EZ version 8.
2.3. Data normalization
All data were normalized to international reference materials. The
δ18OVSMOW-SLAP values of phosphates (i.e. δ18OPHOS) were corrected
against USGS 34 (δ18OVSMOW-SLAP=−27.9‰) and USGS 35
(δ18OVSMOW-SLAP = +57.5‰) nitrates. The δ18OVPDB-LSVEC values of
carbonates (i.e. δ18OCARB) were corrected against LSVEC (δ18OVPDB-
LSVEC=−26.7‰) and NBS 19 (δ18OVPDB-LSVEC=−2.2‰) carbonates
and converted to VSMOW values for easier comparison to δ18OPHOS
values using Coplen et al. (2002). The δ15NAIR values of collagen (i.e.
δ15NCOLL) were corrected against Urea_UIN3 and an acetanilide cali-
brated to USGS 40 (δ15NAIR=−4.52‰) and USGS 41
(δ15NAIR = +47.57‰) amino acids (Schimmelmann et al., 2009).
Weight % N and weight % C values were calibrated using a homo-
genous acetanilide standard. Errors associated with δ18OPHOS values
are± 0.4‰ (1σ); δ18OCARB and δ15NCOLL are± 0.2‰ (1σ); weight % N
and weight % C are± 0.5% (1σ).
Non-exchangeable hydrogen isotope values (i.e. δ2HCOLL-NEX) were
determined using three new collagen reference materials developed in
parallel experiments with the USGS Reston Stable Isotope Laboratory.
An Alaskan moose femur (AMF), Alaskan seal femur (ASF), and
Minnesotan otter leg (MOL) were selected based on their observed raw
differences in δ2H values. Whole bone segments of these reference
materials were degreased in sequential soaks of 2:1 chlor-
oform:methanol. Collagen was extracted in bulk from the degreased
reference bones according to procedures outlined above. The δ2HCOLL-
NEX and f values of AMF, ASF, and MOL were determined via methods in
Qi and Coplen (2011). Briefly, the exchangeable hydrogen atoms in two
identical sets of samples were equilibrated with waters of known and
disparate isotope composition in separate vacuum dessicators. Samples
were transferred to a vacuum oven and dried at 60 °C to remove sec-
ondary adhered water molecules. Samples were removed from the
oven, quickly sealed into silver cups, placed in a zero-blank auto-
sampler and flushed with ultra-pure helium. Samples were calibrated
against VSMOW2 and SLAP2 water standards sealed in silver tubes on a
TCEA with a chromium reactor column (Gehre et al., 2015). The frac-
tion of exchangeable hydrogen (fAMF=0.136, fMOL= 0.145,
fASF= 0.147) was calculated by f= [δ2Htot1 - δ2Htot2]/[δ2Hw1 - δ2Hw2]
where δ2Htot1 and δ2Htot2 are the isotope values of total hydrogen in the
collagen equilibrated with the two different waters. The δ2HVSMOW–SLAP
of non-exchangeable hydrogen was calculated by isotope balance (AMF
Fig. 1. Site map showing burial locations. Note individuals buried at Glorieta Pass, New Mexico, were a military unit of enlisted Texas soldiers.
C.A.M. France et al. Journal of Archaeological Science 96 (2018) 33–44
35
δ2HCOLL-NEX=−73.4‰, MOL δ2HCOLL-NEX = +18.3‰, ASF δ2HCOLL-
NEX =+164.9‰). The δ2HCOLL-NEX in unknown samples was calculated
using a 3-point linear calibration on the δ2HCOLL-NEX for AMF, ASF, and
MOL reference materials as per Wassenaar and Hobson (2003). Errors
associated with δ2HCOLL-NEX values are± 2.0‰ (1σ).
2.4. Examination of diagenesis
Collagen preservation (i.e. post-mortem diagenesis) was examined
using established protein and elemental abundance parameters
(Table 2). Hydroxyapatite phosphate and carbonate were examined
using ATR-FTIR peak height data (Table 2). Acceptable FTIR values for
well-preserved hydroxyapatite are based on modern material in this
study, previous ATR-FTIR data, and data converted from suggested
offsets from more ubiquitous transmission FTIR methods. As phos-
phates are generally more resistant to post-mortem isotope alteration
and more strongly correlated to drinking water compared to carbonates
(Iacumin et al., 1996; Person et al., 1995, 1996; Tuross et al., 1989), the
δ18OPHOS values are used in subsequent statistical analysis and mod-
eling.
2.5. Isotope value conversions and comparisons
Corrected δ2HCOLL-NEX, δ18OPHOS, δ15NCOLL were compared and
considered via various conversions and regression models, both as in-
dividual points and by site averages. The δ2HCOLL-NEX was converted to
meteoric water δ2H values using the equation of Reynard and Hedges
(2008): δ2HMW-COLL = (δ2HCOLL-NEX - 71.9)/1.069 (r= 0.957,
SE= 16.2‰). This equation is based on data from humans, non-human
herbivores, and omnivores. The δ18OPHOS values were converted to
meteoric water δ18O values using the equation of Longinelli (1984):
δ18OMW-P = (δ18OPHOS - 22.37)/0.64 (r= 0.982, SE=0.68‰). The
δ2HMW-COLL values were compared to theoretical δ2HMW values calcu-
lated from the meteoric water line using the δ18OMW-P values: δ2HMW-
THEOR = 8 *δ18OMW-P +10. Differences between δ2HMW-COLL and
δ2HMW-THEORwere compared to δ15NCOLL in an effort to discern trophic
effects that may influence deviations from the meteoric water line.
Cormie et al. (1994a) provide evidence of a well-correlated relationship
(r= 0.917) between δ2HCOLL-NEX and δ18OPHOS in strict herbivores from
the same locality, thereby essentially eliminating trophic effects on the
δ2HCOLL-NEX. This study used their equation to calculate a theoretical
δ2HCOLL-NEX based on δ18OPHOS assuming an end-member scenario of
humans as strict herbivores: δ2HHERB-THEOR = 7.8 *δ18OPHOS −160
(r= 0.917, SE=11.9‰). Differences between δ2HCOLL-NEX and
δ2HHERB-THEOR were plotted against δ15NCOLL in another effort to de-
termine a relationship between trophic effects of human omnivory on
deviations from a pure meteoric water relationship between δ2HCOLL-
NEX and δ18OPHOS. Finally, multiple linear regression of δ18OPHOS and
δ15NCOLL against δ2HCOLL-NEX examined the concurrent influences of
protein intake and meteoric water on hydrogen isotopes in bone col-
lagen. All statistical analyses were run in SigmaPlot 14.0.
In an effort to eliminate other demographic factors as controlling
influences on collagen hydrogen isotopes, the δ2HCOLL-NEX, δ18OPHOS,
and δ15NCOLL values were compared with ancestry (i.e. African
American or Caucasian), sex, estimated age, and socioeconomic status
(i.e. lower, middle, upper, or military class). Socioeconomic status was
assigned based on the context of the burial site. Sex and age were not
available or were indeterminate for some individuals (Supplementary
Table S1).
3. Results
Samples adhering to the defined parameters for well-preserved
collagen and hydroxyapatite were included in subsequent analyses.
Table 3 includes all δ2H, δ18O, and δ15N data pertinent to statistical
analyses and modeling. Table 4 includes full statistical results fromTa
bl
e
1
Si
te
in
fo
rm
at
io
n.
Si
te
Si
te
Lo
ca
ti
on
Ti
m
e
pe
ri
od
O
ri
gi
n
of
In
di
vi
du
al
s
A
nc
es
tr
y
So
ci
o-
ec
on
om
ic
St
at
us
U
rb
an
/R
ur
al
A
.P
.
H
ill
Ft
.A
.P
.H
ill
,V
A
17
80
–1
83
0
So
ut
he
rn
U
ni
te
d
St
at
es
A
fr
ic
an
A
m
er
ic
an
Sl
av
es
ru
ra
l
Fi
rs
t
A
fr
ic
an
Ba
pt
is
t
C
hu
rc
h
(F
A
BC
)
Ph
ila
de
lp
hi
a,
PA
18
24
–1
84
2
Ph
ila
de
lp
hi
a
A
fr
ic
an
A
m
er
ic
an
Fr
ee
bl
ac
k,
po
ss
ib
ili
ty
of
ru
n-
aw
ay
sl
av
es
ur
ba
n
Fo
sc
ue
Pl
an
ta
ti
on
N
or
th
C
ar
ol
in
a
18
00
–1
84
9
So
ut
he
rn
U
ni
te
d
St
at
es
C
au
ca
si
an
H
ig
h
st
at
us
-
pl
an
ta
ti
on
fa
m
ily
ru
ra
l
G
lo
ri
et
a
Pa
ss
N
ew
M
ex
ic
o
18
62
Te
xa
s,
a
fe
w
po
ss
ib
le
Eu
ro
pe
an
im
m
ig
ra
nt
s
C
au
ca
si
an
M
ili
ta
ry
-
C
on
fe
de
ra
te
,
m
os
tl
y
en
lis
te
d,
1
offi
ce
r
in
cl
ud
ed
H
ill
ea
ry
C
em
et
er
y
M
ar
yl
an
d
18
50
–1
90
0
M
ar
yl
an
d
an
d
su
rr
ou
nd
in
g
ar
ea
,p
os
si
bl
y
so
m
e
Eu
ro
pe
an
im
m
ig
ra
nt
s
C
au
ca
si
an
H
ig
h
st
at
us
,b
ur
ia
l
in
fa
m
ily
ce
m
et
er
y
ru
ra
l
Pa
rk
w
ay
G
ra
ve
l
D
el
aw
ar
e
18
50
–1
90
0
D
el
aw
ar
e
A
fr
ic
an
A
m
er
ic
an
Sl
av
es
or
fo
rm
er
sl
av
es
ru
ra
l
Pe
tt
us
V
ir
gi
ni
a
17
00
's
-
18
00
's
V
ir
gi
ni
a,
po
ss
ib
ili
ty
of
di
re
ct
A
fr
ic
an
or
ig
in
A
fr
ic
an
A
m
er
ic
an
Sl
av
es
ru
ra
l
R
ob
in
so
n
C
em
et
er
y
M
ar
yl
an
d
17
75
–1
87
5
M
ar
yl
an
d,
po
ss
ib
ili
ty
of
di
re
ct
A
fr
ic
an
or
ig
in
A
fr
ic
an
A
m
er
ic
an
Sl
av
es
ru
ra
l
Tr
in
it
y
C
at
ho
lic
C
hu
rc
h
G
eo
rg
et
ow
n,
W
as
hi
ng
to
n,
D
C
18
00
–1
85
0
W
as
hi
ng
to
n,
D
C
an
d
su
rr
ou
nd
in
g
ar
ea
s
C
au
ca
si
an
M
id
dl
e
cl
as
s
ur
ba
n
W
al
to
n
Fa
m
ily
C
em
et
er
y
C
on
ne
ct
ic
ut
17
50
–1
83
0
C
on
ne
ct
ic
ut
C
au
ca
si
an
Fa
rm
er
s
ru
ra
l
W
oo
dv
ill
e
C
em
et
er
y
D
el
aw
ar
e
17
90
–1
85
0
D
el
aw
ar
e
C
au
ca
si
an
M
id
dl
e
cl
as
s
ru
ra
l
C.A.M. France et al. Journal of Archaeological Science 96 (2018) 33–44
36
regressions. Yield data, C:N ratios, FTIR data, and calculated differences
between isotope values are presented in the Supplementary Tables S2
and S3. Although δ18OCARB is not used in any subsequent analyses or
modeling, it is included in Supplementary Table S3 to present a com-
plete oxygen data set to the research community for future compar-
isons.
As a coarse examination, δ2HCOLL-NEX, δ18OPHOS, and δ15NCOLL va-
lues showed no significant differences between males and females nor
between Caucasians and African Americans (two-tailed t-tests, all
p > 0.3). No discernible correlation was observed between δ2HCOLL-
NEX, δ18OPHOS, and δ15NCOLL values and minimum estimated age (all
R2 < 0.12). These demographic groupings included individuals from
different regions. No single site yielded more than eight individuals
with determinate age and sex, nor did any single site include both
Caucasians and African Americans, thereby precluding a rigorous sta-
tistical examination of these factors while controlling for regional
variation in meteoric water isotope values. Comparison of δ2HCOLL-NEX
between socioeconomic groups showed no significant differences (two-
tailed t-test, all p > 0.05), with the exception of the lower class group
versus the upper class group (p= 0.00045) and military group
(p=0.020).
The δ2HCOLL-NEX for individual points varied from −15.8 to
+26.2‰; site averages ranged from −2.8 to +17.8‰. The δ18OPHOS
for individual points and site averages ranged from +15.8 to +20.3‰
and +16.5 to +20.0‰, respectively. The δ15NCOLL for individual
points and site averages ranged from +8.7 to +11.7‰ and +9.7 to
+11.7‰, respectively. All regressions and correlations between
δ2HCOLL-NEX, δ18OPHOS, and δ15NCOLL pass Shapiro-Wilk normality tests
(p > 0.05). Linear regression showed a moderate correlation between
individual δ2HCOLL-NEX points and δ18OPHOS or δ15NCOLL, with slightly
stronger correlation between site averages (Fig. 2, Table 4).
Conversion to meteoric water values resulted in δ2HMW-COLL ranging
from −82.0 to −42.8‰ for individual points and −69.8 to −50.6‰
for site averages. The δ18OMW-P for individual points and site averages
ranged from −10.3 to −3.2‰ and −9.2 to −3.7‰, respectively. As
these values, and subsequent δ18OMW-P values, were direct transfor-
mations of δ2HCOLL-NEX and δ18OPHOS, the correlations were similar with
site averages showing a stronger correlation than individual points
(Table 4). Both individual points and site averages showed slopes
considerably lower than that expected from the known meteoric water
line (Fig. 3).
Conversion of δ2HCOLL-NEX to theoretical values based on the me-
teoric water line resulted in δ2HMW-THEOR values ranging from−72.2 to
−15.9‰ for individual points, and −63.6 to −19.7‰ for site
averages. The δ2HMW-COLL and δ2HMW-THEOR values were compared to
one another and showed moderate correlation (Fig. 4, Table 4). Dif-
ferences between these two values were compared to δ15NCOLL (Fig. 4).
Correlation was very poor between δ15NCOLL and the δ2HMW-COLL-
δ2HMW-THEOR difference for individual points (r= 0.0524) and site
averages (r= 0.284).
Conversion of δ2HCOLL-NEX to a theoretical value that assumes hu-
mans are strict herbivores resulted in δ2HHERB-THEOR values ranging
from−36.8 to−1.7‰ for individual points and−31.4 to−4.0‰ for
site averages. The δ2HCOLL-NEX and δ2HHERB-THEOR values were com-
pared to one another and showed moderate correlation (Fig. 5,
Table 4). Differences between these two values were compared to
δ15NCOLL (Fig. 5). Correlation was very poor between δ15NCOLL and the
δ2HCOLL-NEX-δ2HHERB-THEOR difference for individual points (r= 0.176)
and site averages (r= 0.0547).
Multiple linear regression produced better predictive ability for
both individual points and site averages, where the latter again showed
better correlation (Table 4). The highest predictive power lies in a
linear combination of both δ18OPHOS and δ15NCOLL.
4. Discussion
Hydrogen isotopes in human collagen present a complex combina-
tion of influences. Before discussing the influences of meteoric water
and protein intake on the δ2HCOLL-NEX values, it is worth considering
potential confounding factors in the oxygen and nitrogen proxy vari-
ables. These may include inter-regional movement of individuals, de-
mographic factors, individual health status, variable isotope baselines
of food, and non-local food sources.
Movement between regions and cities did occur in historic times.
While the δ2HMW and δ18OMW values should still co-vary in all loca-
tions, the available food (and presumably its inherent δ2H values)
would change. However, the sites included in this study likely represent
local populations with a common set of food resources. Several (Foscue
Plantation, Hilleary Cemetery, Walton Family Cemetery, Woodville
Cemetery) are established family plots including individuals that most
likely lived and worked in the local area for significant portions of their
lives. The variability in δ18OMW-P, an indicator of regional origin, for
each of these sites is < 1.0‰ (1σ), supporting the idea that these in-
dividuals did not spend significant portions of their lives elsewhere.
Three sites (A.P. Hill, Pettus, Robinson Cemetery) are slave populations
that were not free to move around as they pleased; these sites show
individual δ18OMW-P variability of< 1.2‰ (1σ). Three northern urban
sites (FABC, Parkway Gravel, Trinity Catholic Church) are church ce-
meteries that historically represent local communities. Although the
urban setting might suggest higher likelihood for mobility, the δ18OMW-
Table 2
Collagen and hydroxyapatite preservation criteria.
Parameter Acceptable range or peak
location
Baseline
correction
References
Collagen:
Collagen yield ∼2–20% N/A Ambrose 1990, DeNiro 1985, Jorkov et al., 2007, McNulty et al., 2002
Weight % N 10–15% N/A
C:N (atomic) 2.8–3.6 N/A
Hydroxyapatite:
Phosphate peak (ѵ4) height 565 cm−1 410-740 cm−1
Phosphate peak (ѵ4) height 605 cm−1 410-740 cm−1
Lowest height between 565 and 605 cm−1 peaks ∼590 cm−1 410-740 cm−1
Phosphate peak (ѵ1) height ∼960 cm−1 800-1200 cm−1
Phosphate peak (ѵ3) height 1035 cm−1 800-1200 cm−1
Carbonate peak (ѵ3) height 1415 cm−1 1200-1800 cm−1
Carbonate peak (ѵ3) height 1455 cm−1 1200-1800 cm−1
IRSF [(565 cm−1 + 605 cm−1)/590 cm−1] < 4.4 N/A Beasley et al., 2014, Garvie-Lok et al., 2004, Lebon et al., 2010, Lebon et al.,
2011, Snoeck et al., 2014, Thompson et al., 2009, Thompson et al., 2011,
Wright and Schwarcz 1996
C/C (1455 cm−1/1415 cm−1) ∼0.9 N/A
C/P (1415 cm−1/1035 cm−1) ∼0.3 N/A
ѵ1PO4 peak position < 962.5 cm−1 N/A
C.A.M. France et al. Journal of Archaeological Science 96 (2018) 33–44
37
Table 3
Data.
Site Sample ID δ15NCOLL δ18OPHOS aδ18OMW-P δ2HCOLL-NEX bδ2HMW-COLL cδ2HMW-THEOR dδ2HHERB-THEOR
A.P. Hill, VA 44CEAPHILL-VAOCME-1 10.3 16.3 −9.5 5.7 −62.0 −66.1 −33.0
A.P. Hill, VA 44CEAPHILL-VAOCME-2 10.4 17.5 −7.7 16.4 −51.9 −51.4 −23.8
A.P. Hill, VA 44CEAPHILL-VAOCME-3 11.2 18.1 −6.7 23.0 −45.8 −43.9 −19.2
A.P. Hill, VA 44CEAPHILL-VAOCME-4 9.7 17.4 −7.8 11.7 −56.3 −52.2 −24.3
site average 10.4 17.3 −7.9 14.2 −54.0 −53.4 −25.1
FABC, PA FABC-08-63b 10.1 16.8 −8.8 1.9 −65.5 −60.2 −29.3
FABC, PA FABC-08-107a 10.2 16.2 −9.6 4.8 −62.7 −67.0 −33.6
site average 10.2 16.5 −9.2 3.4 −64.1 −63.6 −31.4
Foscue, NC 31FOSCUE-ECU-1 11.3 19.9 −3.9 19.1 −49.4 −21.2 −5.0
Foscue, NC 31FOSCUE-ECU-2 10.6 20.3 −3.2 16.2 −52.1 −15.9 −1.7
Foscue, NC 31FOSCUE-ECU-3 10.5 20.2 −3.3 15.8 −52.5 −16.5 −2.1
Foscue, NC 31FOSCUE-ECU-4 11.7 19.6 −4.4 20.2 −48.4 −25.0 −7.4
site average 11.0 20.0 −3.7 17.8 −50.6 −19.7 −4.0
Glorieta Pass, NM GLO-099-1A 11.5 18.5 −6.0 10.0 −57.9 −38.0 −15.5
Glorieta Pass, NM GLO-099-2A 10.8 17.6 −7.5 11.4 −56.6 −50.0 −23.0
Glorieta Pass, NM GLO-099-2C 10.4 18.6 −5.9 7.9 −59.9 −36.8 −14.7
Glorieta Pass, NM GLO-099-2E 10.4 19.2 −4.9 16.8 −51.5 −29.4 −10.1
Glorieta Pass, NM GLO-099-2H 10.9 19.4 −4.6 26.2 −42.8 −26.7 −8.4
Glorieta Pass, NM GLO-099-2P 10.5 19.1 −5.1 22.9 −45.8 −31.0 −11.1
Glorieta Pass, NM GLO-099-2R 10.7 18.9 −5.4 −6.6 −73.5 −33.1 −12.4
Glorieta Pass, NM GLO-099-2V 10.3 18.9 −5.4 14.2 −53.9 −33.0 −12.4
Glorieta Pass, NM GLO-099-2X 10.1 16.9 −8.5 9.9 −58.0 −58.2 −28.1
site average 10.6 18.6 −5.9 12.5 −55.6 −37.4 −15.1
Hilleary Cemetery, MD 18PR978-HILLEARY-FEA5 10.5 17.3 −8.0 6.1 −61.6 −53.7 −25.3
Hilleary Cemetery, MD 18PR978-HILLEARY-FEA4 11.4 16.7 −8.8 22.8 −45.9 −60.6 −29.6
Hilleary Cemetery, MD 18PR978-HILLEARY-GEORGEW 11.3 16.8 −8.7 15.5 −52.8 −59.7 −29.0
Hilleary Cemetery, MD 18PR978-HILLEARY-FEA2 10.5 17.2 −8.1 6.7 −61.0 −54.5 −25.8
Hilleary Cemetery, MD 18PR978-HILLEARY-HENRY 9.8 16.6 −9.0 10.3 −57.7 −62.3 −30.6
Hilleary Cemetery, MD 18PR978-HILLEARY-REBECCA 9.0 16.5 −9.1 8.2 −59.6 −63.2 −31.2
Hilleary Cemetery, MD 18PR978-HILLEARY-FEA3 9.9 17.2 −8.0 3.5 −64.0 −54.3 −25.6
site average 10.4 16.9 −8.5 10.4 −57.5 −58.3 −28.2
Parkway Gravel, DE 7NCE176-DHCA-V01 9.3 17.9 −7.0 1.5 −65.9 −45.8 −20.3
Parkway Gravel, DE 7NCE176-DHCA-V02 10.8 17.2 −8.0 7.1 −60.6 −54.3 −25.6
Parkway Gravel, DE 7NCE176-DHCA-V03 9.7 17.8 −7.1 10.5 −57.5 −46.9 −21.0
site average 10.0 17.6 −7.4 6.3 −61.3 −49.0 −22.3
Pettus, VA 44JC33-PETTUS-270 9.8 16.2 −9.7 5.1 −62.5 −67.5 −33.9
Pettus, VA 44JC33-PETTUS-201 9.1 17.1 −8.2 21.0 −47.6 −55.7 −26.5
Pettus, VA 44JC33-PETTUS-231 10.8 17.6 −7.5 9.6 −58.3 −49.9 −22.9
Pettus, VA 44JC33-PETTUS-223 10.1 17.1 −8.2 13.7 −54.4 −55.5 −26.4
Pettus, VA 44JC33-PETTUS-205 9.8 17.4 −7.8 1.7 −65.6 −52.2 −24.3
Pettus, VA 44JC33-PETTUS-191 10.4 18.4 −6.2 8.0 −59.7 −39.8 −16.6
Pettus, VA 44JC33-PETTUS-253 10.0 17.4 −7.8 −10.2 −76.8 −52.1 −24.2
site average 10.0 17.3 −7.9 7.0 −60.7 −53.2 −25.0
Robinson Cemetery, MD 44HE950-RADFORD-5 11.1 16.7 −8.8 23.3 −45.4 −60.7 −29.6
Robinson Cemetery, MD 44HE950-RADFORD-10 10.9 17.7 −7.2 15.2 −53.0 −47.9 −21.7
Robinson Cemetery, MD 44HE950-RADFORD-16 10.7 18.0 −6.8 11.8 −56.2 −44.6 −19.6
Robinson Cemetery, MD 44HE950-RADFORD-18 10.2 17.7 −7.3 2.6 −64.8 −48.6 −22.1
site average 10.7 17.5 −7.6 13.2 −54.9 −50.4 −23.2
Trinity Church, DC TRINITY-EAST-10 11.7 17.5 −7.6 7.2 −60.6 −51.0 −23.6
site average 11.7 17.5 −7.6 7.2 −60.6 −51.0 −23.6
Walton Cemetery, CT 6CT58-5-AMM01 9.8 17.2 −8.1 1.9 −65.5 −55.1 −26.1
Walton Cemetery, CT 6CT58-5-AMM02 10.1 17.2 −8.1 0.7 −66.6 −54.9 −26.0
Walton Cemetery, CT 6CT58-5-AMM04A 10.3 16.0 −10.0 3.7 −63.8 −70.0 −35.4
Walton Cemetery, CT 6CT58-5-AMM05 9.8 17.4 −7.8 −9.0 −75.7 −52.2 −24.3
Walton Cemetery, CT 6CT58-5-AMM07 9.6 15.8 −10.3 4.3 −63.2 −72.2 −36.8
Walton Cemetery, CT 6CT58-5-AMM08A 9.4 16.2 −9.6 −7.4 −74.2 −67.0 −33.5
Walton Cemetery, CT 6CT58-5-AMM09 9.7 16.6 −9.0 −1.4 −68.6 −61.7 −30.2
Walton Cemetery, CT 6CT58-5-AMM11 9.5 16.4 −9.3 −4.0 −71.0 −64.8 −32.2
Walton Cemetery, CT 6CT58-5-AMM19 9.7 17.8 −7.1 −15.8 −82.0 −47.1 −21.2
Walton Cemetery, CT 6CT58-5-AMM20 9.6 16.2 −9.6 4.6 −63.0 −66.9 −33.5
Walton Cemetery, CT 6CT58-5-AMM21 9.9 16.6 −8.9 −14.1 −80.5 −61.6 −30.2
Walton Cemetery, CT 6CT58-5-AMM22 10.0 16.5 −9.2 6.1 −61.5 −63.6 −31.4
Walton Cemetery, CT 6CT58-5-AMM23 8.7 17.4 −7.8 −3.3 −70.3 −52.0 −24.2
Walton Cemetery, CT 6CT58-5-AMM25 9.4 16.8 −8.7 −4.3 −71.3 −59.9 −29.1
Walton Cemetery, CT 6CT58-5-AMM26 9.4 17.5 −7.6 −2.3 −69.4 −51.2 −23.7
Walton Cemetery, CT 6CT58-5-AMM27 9.6 17.1 −8.3 −3.6 −70.6 −56.2 −26.8
site average 9.7 16.8 −8.7 −2.8 −69.8 −59.8 −29.0
Woodville Cemetery, DE 7NCE98A-DHCA-08 10.7 17.9 −6.9 13.1 −55.0 −45.4 −20.1
Woodville Cemetery, DE 7NCE98A-DHCA-12 11.7 17.2 −8.1 8.8 −59.0 −54.6 −25.8
(continued on next page)
C.A.M. France et al. Journal of Archaeological Science 96 (2018) 33–44
38
P variability is < 0.6‰ (1σ) for each of these sites. The Glorieta Pass
site is constituted entirely of a military unit that historic records show
mustered out of Texas. These individuals show a δ18OMW-P variability of
1.3‰ (1σ), which is the expected range in Texas.
Comparison of δ2HCOLL-NEX, δ18OPHOS, and δ15NCOLL values with
demographic factors showed no statistically significant connections
between ancestry, sex, and estimated age. France et al. (2014) found
that ancestry was in fact most strongly predicted by collagen carbon
isotope values rather than oxygen or nitrogen. Comparison of δ2HCOLL-
NEX to social class showed significant differences between the lower
class group versus the upper class and military groups. However, the
sole military group in this test is the southernmost site, while lower
class individuals were found in northern locations. This introduces re-
gional controls on drinking water isotopes as a potential factor and
precludes the conclusion that social class is the prevailing component in
this observed isotopic difference between the lower class and military
groups. The lower and upper class group contained individuals from
similar regions which does impose some control on regional variability
in meteoric water isotope values. In some populations, social class is
apparently linked to food availability and food choice, which in turn is
a rough proxy for nutritional status (France et al., 2014; Yoder, 2012).
As protein intake, and consequently essential amino acid intake, are
linked to food availability and choice, it bears consideration that social
class may influence the δ2HCOLL-NEX values in so much as social class
influences the type of food consumed. However, social class is only
moderately predictable using a combination of carbon, nitrogen and
oxygen isotope values and does not appear to be the prevailing factor
controlling stable isotope distributions in these humans (France et al.,
2014). Therefore, the potential influence of social class on δ2HCOLL-NEX
values does not preclude examination of meteoric water and protein
intake on said values.
The δ15NCOLL values can be complicated by health issues. The
δ15NCOLL values can reflect extreme circumstances including disease,
infection, and pregnancy (Beaumont et al., 2013, 2015; D'Ortenzio
et al., 2015; Fuller et al., 2004; Scorrano et al., 2014). Physiological
examination of the remains in this study suggests no obvious presence
of these factors, with the exception of two elderly individuals
(31FOSCUE-ECU-4 and 44JC33-PETTUS-270) who exhibit osteopenia
(Barca and Owsley, 2014). Neither individual has outlying δ15NCOLL
values which suggests this particular condition did not significantly
affect their nitrogen isotope values.
The baseline δ15N value of plants can vary somewhat by region and
incorporation of nitrogen isotopes into bone collagen can be affected by
climate. Mammals in hotter drier climates will sometimes exhibit 15N
enrichment compared to counterparts in cooler wetter climates
(Ambrose, 1991; Cormie and Schwarcz, 1994; Fizet et al., 1995; Sealy
et al., 1987). However, the individuals in this study's sites do not
conform to this pattern. The larger data set of France et al. (2014),
which includes most of the sites in the current study, shows no corre-
lation between human δ15NCOLL and δ18OMW-P (r= 0.0892). The ex-
ception is Walton Cemetery in Connecticut, the most northern and
coolest site. The human remains from this site do show the most ne-
gative δ15NCOLL values, but the Walton Cemetery site average (+9.7‰)
is only 0.6‰ less than the average of all sites combined (+10.3‰).
Regional climates and health issues may still exhibit some control over
δ15NCOLL values, but it appears to be minor in the remains sampled
here. The dominant controlling factor is more likely to be dietary input
as discussed below.
Although 18th and 19th century humans consumed a much more
localized diet than modern humans, they still consumed some percen-
tage of non-local food and could exercise choice in their dietary se-
lections, unlike non-human mammals considered in previous studies.
The more common imports of the time included cocoas, coffee, tea,
rum, molasses and sugar from the West Indies and other regions; rice
from Asia; wine from Europe and other regions; fresh produce and some
meat from other United States regions. However, broad interregional
and international food trade did not become widespread until the late
nineteenth century (Bruegel, 2002; Nützenadel and Trentmann, 2008;
Table 3 (continued)
Site Sample ID δ15NCOLL δ18OPHOS aδ18OMW-P δ2HCOLL-NEX bδ2HMW-COLL cδ2HMW-THEOR dδ2HHERB-THEOR
Woodville Cemetery, DE 7NCE98A-DHCA-01 11.0 17.8 −7.1 8.8 −59.0 −46.9 −21.0
Woodville Cemetery, DE 7NCE98A-DHCA-03 10.1 17.4 −7.8 10.6 −57.3 −52.5 −24.5
Woodville Cemetery, DE 7NCE98A- WOODVILLE-04 11.0 18.2 −6.5 9.0 −58.9 −42.0 −18.0
Woodville Cemetery, DE 7NCE98A- WOODVILLE-SLOPEB 10.3 16.7 −8.8 13.2 −54.9 −60.4 −29.4
site average 10.8 17.5 −7.5 10.6 −57.3 −50.3 −23.1
All values are in ‰; δ18O and δ2H are standardized to VSMOW, δ15N is standardized to AIR.
a Calculated from Longinelli (1984) where δ18OMW-P = (δ18OPHOS - 22.37)/0.64.
b Calculated from Reynard and Hedges (2008) where δ2HMW-COLL = (δ2HCOLL-NEX - 71.9)/1.069.
c Calculated from the meteoric water line where δ2HMW-THEOR = 8 *δ18OMW-P + 10.
d Calculated from Cormie et al. (1994a) where δ2HHERB-THEOR = 7.8 *δ18OPHOS - 160.
Table 4
Regression equations and statistical information.
Regression Correlation coefficient (r) Standard error (SE) F-value p-value
δ2HCOLL-NEX = 3.6 *δ18OPHOS – 55.5 Individual points 0.400 8.6 11.593 0.001
δ2HCOLL-NEX = 4.1 *δ18OPHOS – 63.0 Site averages 0.695 4.3 8.409 0.018
δ2HCOLL-NEX = 7.3 *δ15NCOLL – 67.2 Individual points 0.539 8.0 24.971 <0.001
δ2HCOLL-NEX = 5.7 *δ15NCOLL – 50.3 Site averages 0.562 5.0 4.150 0.072
δ2HMW-COLL = 2.1 *δ18OMW-P – 44.0 Individual points 0.400 8.1 11.593 0.001
δ2HMW-COLL = 2.5 *δ18OMW-P – 40.4 Site averages 0.695 4.0 8.417 0.018
δ2HMW-COLL = 0.27 *δ2HMW-THEOR – 46.7 Individual points 0.400 8.1 11.593 0.001
δ2HMW-COLL = 0.31 *δ2HMW-THEOR – 43.5 Site averages 0.695 4.0 8.415 0.018
δ2HCOLL-NEX = 0.46 *δ2HHERB-THEOR + 18.2 Individual points 0.400 8.6 11.593 0.001
δ2HCOLL-NEX = 0.53 *δ2HHERB-THEOR + 21.0 Site averages 0.695 4.3 8.417 0.018
δ2HCOLL-NEX = 1.9 *δ18OPHOS + 6.1 *δ15NCOLL – 88.6 Individual points 0.573 7.8 14.660 <0.001
δ2HCOLL-NEX = 3.3 *δ18OPHOS + 3.1 *δ15NCOLL – 80.5 Site averages 0.745 4.2 5.001 0.039
C.A.M. France et al. Journal of Archaeological Science 96 (2018) 33–44
39
Perren, 2006).
While some of the factors discussed above may have a minor in-
fluence on the isotope values, none appear to be a dominant or sig-
nificant factor in this study. However, it is worth bearing such factors in
mind in future examinations of human δ2HCOLL-NEX since the body of
published data is currently small and the understanding of how de-
mographics and health affect hydrogen routing in humans is poorly
understood. Considering data as individual points will include wider
variation while averaging data by site eliminates some of the variability
due to food choice, food availability, and lifetime movement. The δ18O
and δ15N values are likely to be reliable proxies for meteoric water
influence and protein intake, respectively. The δ18OPHOS and δ15NCOLL
values showed moderate correlation to δ2HCOLL-NEX values. Better cor-
relation through multiple linear regression suggests both are coupled to
δ2HCOLL-NEX with ∼56% of the variation in δ2HCOLL-NEX related to a
combination of δ18OPHOS and δ15NCOLL (from r=0.745). A larger data
set of pooled individuals (i.e. site averages) might facilitate a model
whereby a reasonable approximation of any one variable can be
determined from a combination of the other two. Likewise, calculations
of δ2HMW-COLL and δ18OMW-P can be performed once δ2HCOLL-NEX or
δ18OPHOS are known. This holds potential for reconstructing historic
meteoric water isotope values from archaeological remains when faunal
remains are unavailable, and subsequently examining geographic ori-
gins, migration, and movement.
Deviation of the δ2HMW-COLL and δ18OMW-P relationship from the
expected meteoric water line suggests minor input of non-local food
sources and metabolic decoupling of the δ2H-δ18O relationship. The
observed slope of this relationship (i.e. 2.1 for individual points, 2.5 for
site averages) was well below the expected value of 8.0. Regression of
δ2HMW-COLL against δ2HMW-THEOR produces a slope well below 1.0. Since
δ2HMW-THEOR represents a theoretically expected value for δ2HMW based
on the meteoric water line, a slope< 1.0 indicates the δ2H incorporated
into bone collagen is more negative than it should be if controlled
strictly by regional meteoric water. This may indicate that dietary
amino acids entering the body via protein consumption had an in-
herently depleted δ2HCOLL-NEX value. Local herbivorous protein sources
Fig. 2. (A) Phosphate oxygen isotope values versus collagen non-exchangeable hydrogen isotope values. (B) Nitrogen isotope values from collagen versus non-
exchangeable hydrogen isotope values from collagen. Enlarged symbols indicate site averages.
Fig. 3. Collagen hydrogen isotope values converted to meteoric water values versus phosphate oxygen isotope values converted to meteoric water values. Enlarged
symbols indicate site averages.
C.A.M. France et al. Journal of Archaeological Science 96 (2018) 33–44
40
(i.e. deer, cow, sheep, etc.) foddered on local vegetation should con-
form to the meteoric water line where the slope of δ2HCOLL-NEX versus
δ18OPHOS is approximately 8.0, and the slope of δ2HCOLL-NEX versus
δ2HMW is approximately 1.0 (Cormie et al. 1994a, 1994c; Pietsch et al.,
2011; Reynard and Hedges, 2008; Topalov et al., 2013). In this study,
the deviation from these expected slopes may be due to inclusion of
protein sources imported from northern latitudes, or from local animals
foddered on imported northern grains. This would result in a depleted
δ2Hdietary amino acid input into the human body, while maintaining the
local δ18OMW-P signature, thereby decreasing the δ2HCOLL-NEX versus
δ2HMW slope to<1.0. Bowen et al. (2009) and Kirsanow and Tuross
(2011) also noted that amount of local versus non-local dietary input
can affect expected correlations between δ2H and δ18O in human ker-
atin and collagen. While the influence of imported and non-local foods
on isotope values cannot be completely discounted, it is likely to be a
relatively minor factor, as discussed above.
Differences between δ2HMW-COLL and δ2HMW-THEOR did not correlate
well with δ15NCOLL, nor did differences between δ2HCOLL-NEX and
δ2HHERB-THEOR. The δ2HMW-THEOR and δ2HHERB-THEOR represent expected
values for meteoric water-controlled hydrogen isotope values without
trophic level fractionations. Nitrogen isotopes show a very systematic
enrichment during increased meat consumption where 14N is pre-
ferentially excreted with urea thereby leaving the body enriched in 15N
(Sutoh et al., 1987). If a similar mechanism fractionated hydrogen
isotopes during incorporation into collagen, one would expect δ2HMW-
COLL-δ2HMW-THEOR differences and δ2HCOLL-NEX-δ2HHERB-THEOR differ-
ences to correlate with δ15NCOLL. The noticeable lack of correlation
suggests some other metabolic mechanism at work.
Rather than a systematic enrichment or depletion of hydrogen iso-
topes during collagen formation, this data may indicate a threshold of
meat consumption exists beyond which the expected δ2H-δ18O re-
lationships decouple. Pietsch et al. (2011) demonstrated that pure
Fig. 4. (A) Collagen non-exchangeable hydrogen isotope values converted to meteoric water values versus theoretical meteoric water hydrogen isotope values based
on the meteoric water line. (B) Difference between converted values and theoretical values versus collagen nitrogen isotope values. Enlarged symbols indicate site
averages.
C.A.M. France et al. Journal of Archaeological Science 96 (2018) 33–44
41
carnivores drinking very little water show virtually no correlation be-
tween δ2H and δ18O in hair keratin, nor do keratin and meteoric water
δ2H and δ18O values correlate. Cormie et al. (1994a), Cormie et al.
(1994c), Reynard and Hedges (2008), and Pietsch et al. (2011) de-
monstrated that strict herbivores show the opposite with strong corre-
lations between bone δ2H and δ18O values and strong correlations be-
tween meteoric water δ2H and δ18O values. Data from omnivorous
humans in this study fall somewhere in between these two end mem-
bers. This supports the conclusions of Pietsch et al. (2011) that a
threshold of meat consumption may exist above which animals obtain
the majority of their amino acids from consumed meat and have
minimal need to produce amino acids in vivo. Without in vivo amino
acid production, meteoric water hydrogen will not be incorporated into
δ2HCOLL-NEX values either through direct water consumption or via
plant tissues. The primary factor contributing to δ2HCOLL-NEX will then
become the baseline δ2H value of the meat itself which has already
cycled through several metabolic processes and associated fractiona-
tions. The more meat consumed, the more the δ2H-δ18O relationships
deviate from the expected meteoric water line. Thus the overall
δ2HCOLL-NEX will show some correlation with a proxy for meat amount
consumed (i.e. δ15NCOLL), while relationships between δ2HCOLL-NEX and
δ18OPHOS (and calculated values based on these factors) will decouple.
This opens the question as to where the threshold lies. From the
previously introduced general conceptual model, one may consider the
δ2HCOLL-NEX for mammals as follows:
Herbivores: δ2HCOLL-NEX= δ2Hingested water + εa
Omnivores: δ2HCOLL-NEX = (δ2Hingested water + εa) + (δ2Hdietary amino
acids + εb)
Carnivores: δ2HCOLL-NEX= δ2Hdietary amino acids + εb
Herbivores and carnivores are presented here as extreme end
Fig. 5. (A) Measured collagen non-exchangeable hydrogen isotope values versus theoretical hydrogen isotope values based on pure herbivory. (B) Difference between
measured values and theoretical values versus collagen nitrogen isotope values. Enlarged symbols indicate site averages.
C.A.M. France et al. Journal of Archaeological Science 96 (2018) 33–44
42
members, which is oversimplified. Herbivores do obtain some essential
amino acids from plants, but these amino acids generally have similar
δ2H values to local meteoric water. Most carnivores will ingest some
meteoric water, albeit in very limited quantity, or it may enter the body
already incorporated into food (i.e. meat). Nonetheless, the point at
which one end member grades into another is unknown, but potentially
useful information. Future work in feeding studies with gradations of
consumed meat amount may be able to use deviations from the ex-
pected meteoric water line δ2H-δ18O relationships to determine per-
centage of meat in a person's diet. The slope between δ2H and δ18O in
keratin or collagen appears to decrease from ∼8.0 to ∼0.0 as animals
move from pure localized herbivory to pure carnivory; likewise the
slope between δ2H in collagen or keratin and meteoric or drinking
water δ2H appears to decrease from ∼1.0 to ∼0.0 between herbivores
and carnivores (Bowen et al., 2009; Cormie et al. 1994a, 1994c;
Ehleringer et al., 2008; Pietsch et al., 2011; Reynard and Hedges, 2008;
Topalov et al., 2013). More data is needed for bone collagen in parti-
cular. The ability to estimate percentage meat consumption would add
new depth to understanding ancient lifestyles, food choice, and food
availability in historic and archaeological populations.
5. Conclusions
Hydrogen isotopes in human bone collagen represent a combination
of drinking water and dietary input. Using phosphate oxygen isotopes
and collagen nitrogen isotopes as meteoric water and protein intake
proxies, respectively, one can estimate the collagen hydrogen isotope
values. This is contingent upon pooling multiple individuals from a
given site to eliminate inherent variability from food choice, food
availability, and movement between locations. Relationships between
meteoric water oxygen isotope values and bone collagen hydrogen
isotope values show deviation from the expected meteoric water line.
This is hypothetically due to the minor inclusion of non-local food
sources and a decoupling of oxygen/hydrogen relationships as in-
dividuals reduce in vivo amino acid production due to increased meat
consumption. Further examination of this decoupling could determine
percentage of dietary meat, thereby adding valuable new information
into archaeological dietary studies. Currently nitrogen isotope data can
provide only relative amounts of meat consumption between in-
dividuals or populations. Incorporating collagen hydrogen isotopes may
facilitate absolute determinations of meat consumption, as well as
adding depth to our understanding of locality and movement within
ancient populations.
Acknowledgements
Collection access through Smithsonian National Museum of Natural
History Skeletal Biology Program. K. Barca, K. Bruwelheide, C. Doney†,
D. Dunn†, A. Lowe†, S. McGuire, S. Mills, W. Miller, J. Ososky, B.
Pobiner, C. Potter, A. Warmack assisted with procurement/preparation
of remains. Special thanks to D. Owsley (NMNH), and T. Coplen (USGS)
for sample procurement and support. Thanks to T. Cleland, N.
Sugiyama, and five anonymous reviewers for constructive comments.
†Supported by NSF REU Site Grant SMA-1156360.
Appendix A. Supplementary data
Supplementary data related to this article can be found at http://dx.
doi.org/10.1016/j.jas.2018.05.011.
References
Alberts, D.E., 1984. Rebels on the Rio Grande. University of New Mexico Press,
Albuquerque.
Ambrose, S.H., 1990. Preparation and characterization of bone and tooth collagen for
isotopic analysis. J. Archaeol. Sci. 17, 431–451.
Ambrose, S.H., 1991. Effects of diet, climate, and physiology on nitrogen isotope abun-
dances in terrestrial foodwebs. J. Archaeol. Sci. 18, 292–317.
Armbruster, W., Lehnert, K., Vetter, W., 2006. Establishing a chromium-reactor design for
measuring δ2H values of solid polyhalogenated compounds using direct elemental
analysis and stable isotope ratio mass spectrometry. Anal. Bioanal. Chem. 384,
237–243.
[dataset]Barca, K.G., Owsley, D., 2014. The Human Skeleton Database, the Smithsonian
Skeletal Biology Program. National Museum of Natural History.
Beasley, M.M., Bartelink, E.J., Taylor, L., Miller, R.M., 2014. Comparison of transmission
FTIR, ATR, and DRIFT spectra: implications for assessment of bone bioapatite diag-
enesis. J. Archaeol. Sci. 46, 16–22.
Beaumont, J., Gledhill, A., Lee-Thorp, J., Montgomery, J., 2013. Childhood diet: a closer
examination of the evidence from dental tissues using stable isotope analysis of in-
cremental human dentine. Archaeometry 55, 277–295.
Beaumont, J., Montgomery, J., Buckberry, J., Jay, M., 2015. Infant mortality and isotopic
complexity: new approaches to stress, maternal health, and weaning. Am. J. Phys.
Anthropol. 157, 441–457.
Birchall, J., O'Connell, T.C., Heaton, T.H.E., Hedges, R.E.M., 2005. Hydrogen isotope
ratios in animal body proteins reflect trophic level. J. Anim. Ecol. 74, 877–881.
Bocherens, H., Drucker, D., 2003. Trophic level isotopic enrichment of carbon and ni-
trogen in bone collagen: case studies from recent and ancient terrestrial ecosystems.
Int. J. Osteoarchaeol. 13, 46–53.
Bowen, G.J., Ehleringer, J.R., Chesson, L.A., Thompson, A.H., Podlesak, D.W., Cerling,
T.E., 2009. Dietary and physiological controls on the hydrogen and oxygen isotope
ratios of hair from in-20th century indigenous populations. Am. J. Phys. Anthropol.
139, 494–504.
Bowen, G.J., Wassenaar, L.I., Hobson, K.A., 2005. Global application of stable hydrogen
and oxygen isotopes to wildlife forensics. Oecologia 143, 337–348.
Bruegel, M., 2002. Farm, Shop, Landing – the Rise of a Market Economy in the Hudson
Valley, 1780–1860. Duke University Press, Durham.
Bryant, J.D., Froelich, P.N., 1995. A model of oxygen isotope fractionation in body water
of large mammals. Geochem. Cosmochim. Acta 59, 4523–4537.
Chesson, L.A., Podlesak, D.W., Cerling, T.E., Ehleringer, J.R., 2009. Evaluating un-
certainty in the calculations of non-exchangeable hydrogen fractions within organic
materials. Rapid Commun. Mass Spectrom. 23, 1275–1280.
Chesson, L.A., Valenzuela, L.O., Bowen, G.J., Cerling, T.E., Ehleringer, J.R., 2011.
Consistent predictable patterns in the hydrogen and oxygen stable isotope ratios of
animal proteins consumed by modern humans in the USA. Rapid Commun. Mass
Spectrom. 25, 3713–3722.
Coplen, T.B., Hopple, J.A., Böhlke, J.K., Peiser, H.S., Rieder, S.E., Krouse, H.R., Rosman,
K.J.R., Ding, T., Vocke, R.D., Revesz, K., Lamberty, A., Taylor, P.D.P., De Bièvre, P.,
2002. Compilation of Minimum and Maximum Isotope Ratios of Selected Elements in
Maturally Occurring Terrestrial Materials and Reagants. U.S. Geological Survey
Water-Resources Investigations Report 01–4222.
Commerford, S.L., Carsten, A.L., Cronkite, E.P., 1983. The distribution of tritium among
the amino acids of proteins obtained from mice exposed to tritiated water. Radiat.
Res. 94, 151–155.
Cormie, A.B., Schwarcz, H.P., 1994. Stable isotopes of nitrogen and carbon of North
American white-tailed deer and implications for paleodietary andother food web
studies. Palaeogeogr. Palaeoclimatol. Palaeoecol. 107, 227–241.
Cormie, A.B., Luz, B., Schwarz, H.P., 1994a. Relationship between the hydrogen and
oxygen isotopes of deer bone and their use in the estimation of relative humidity.
Geochem. Cosmochim. Acta 58, 3439–3449.
Cormie, A.B., Schwarz, H.P., Gray, J., 1994b. Determination of the hydrogen isotopic
composition of bone collagen and correction for hydrogen exchange. Geochem.
Cosmochim. Acta 58, 365–375.
Cormie, A.B., Schwarz, H.P., Gray, J., 1994c. Relation between hydrogen isotopic ratios
of bone collagen and rain. Geochem. Cosmochim. Acta 58, 377–391.
Craig, H., 1961. Isotopic variation in meteoric waters. Science 133, 1702–1703.
Daux, V., Lécuyer, C., Héron, M., Amiot, R., Simon, L., Fourel, F., Martineau, F.,
Lynnerup, N., Reychler, H., Escarguel, G., 2008. Oxygen isotope fractionation be-
tween human phosphate and water revisited. J. Hum. Evol. 55, 1138–1147.
DeNiro, M.J., 1985. Postmortem preservation and alteration of in vivo bone collagen
isotope ratios in relation to palaeodietary reconstruction. Nat 317, 806–809.
DeNiro, M.J., Epstein, S., 1981. Influence of diet on the distribution of nitrogen isotopes
in animals. Geochem. Cosmochim. Acta 45, 341–351.
D'Ortenzio, L., Brickley, M., Schwarcz, H., Prowse, T., 2015. You are not what you eat
during physiological stress: isotopic evaluation of human hair. Am. J. Phys.
Anthropol. 157, 374–388.
Ehleringer, J.R., Bowen, G.J., Chesson, L.A., West, A.G., Podlesak, D.W., Cerling, T.E.,
2008. Hydrogen and oxygen isotope ratios in human hair are related to geography. P
Natl Acad Sci USA 105, 2788–2793.
Fizet, M., Mariotti, A., Bocherens, H., Lange-Badré, B., Vandermeersch, B., Borel, J.P.,
Bellon, G., 1995. Effect of diet, physiology and climate on carbon and nitrogen stable
isotopes of collagen in a Late Pleistocene anthropic palaeoecosystem: Marillac,
Charente, France. J. Archaeol. Sci. 22, 67–79.
France, C.A.M., Owsley, D., 2015. Stable carbon and oxygen isotope spacing between
bone and tooth collagen and hydroxyapatite in human archaeological remains. Int. J.
Osteoarchaeol. 25, 299–312.
France, C.A.M., Owsley, D.W., Hayek, L.C., 2014. Stable isotope indicators of provenance
and demographics in 18th and 19th century North Americans. J. Archaeol. Sci. 42,
356–366.
Fuller, B.T., Fuller, J.L., Sage, N.E., Harris, D.A., O'Connell, T.C., Hedges, R.E.M., 2004.
Nitrogen balance and δ15N: why you’re not what you eat during pregnancy. Rapid
Commun. Mass Spectrom. 18, 2889–2896.
Garvie-Lok, S.J., Varney, T.L., Katzenberg, M.A., 2004. Preparation of bone carbonate for
C.A.M. France et al. Journal of Archaeological Science 96 (2018) 33–44
43
stable isotope analysis: the effects of treatment time and acid concentration. J.
Archaeol. Sci. 31, 763–776.
Gehre, M., Renpenning, J., Gilevska, T., Qi, H., Coplen, T.B., Meijer, H.A.J., Brand, W.A.,
Schimmelmann, A., 2015. On-line hydrogen-isotope measurements of organic sam-
ples using elemental chromium─an extension for high temperature elemental ana-
lyzer techniques. Anal. Chem. 87, 5198–5205.
Hobson, K.A., Atwell, L., Wassenaar, L.I., 1999. Influences of drinking water and diet on
the stable-hydrogen isotope ratios of animal tissues. P Natl Acad Sci USA 96,
8003–8006.
Iacumin, P., Bocherens, H., Mariotti, A., Longinelli, A., 1996. Oxygen isotope analyses of
co-existing carbonate and phosphate in biogenic apatite: a way to monitor diagenetic
alteration of bone phosphate? Earth Planet Sci. Lett. 142, 1–6.
Jorkov, M.L.S., Heinemeier, J., Lynnerup, N., 2007. Evaluating bone collagen extraction
methods for stable isotope analysis in dietary studies. J. Archaeol. Sci. 34,
1824–1829.
Kelly, S.D., Heaton, K.D., Brereton, P., 2001. Deuterium/hydrogen isotope ratio mea-
surement of water and organic samples by continuous-flow isotope ratios mass
spectrometry using chromium as the reducing agent in and elemental analyser. Rapid
Commun. Mass Spectrom. 15, 1283–1286.
Kendall, C., Coplen, T.B., 2001. Distribution of oxygen-18 and deuterium in river waters
across the United States. Hydrol. Process. 15, 1363–1393.
Kirsanow, K., Tuross, N., 2011. Oxygen and hydrogen isotopes in rodent tissues: impact of
diet, water, and ontogeny. Palaeogeogr. Palaeoclimatol. Palaeoecol. 310, 9–16.
Kirsanow, K., Makarewicz, C., Tuross, N., 2008. Stable oxygen (δ18O) and hydrogen (δD)
isotopes in ovicaprid dentinal collagen record seasonal variation. J. Archaeol. Sci. 35,
3159–3167.
Kohn, M.J., 1996. Predicting animal δ18O: accounting for diet and physiological adap-
tation. Geochem. Cosmochim. Acta 60, 4811–4829.
Lebon, M., Müller, K., Bahain, J.J., Fröhlich, F., Falguères, C., Bertrand, L., Sandt, C.,
Reiche, I., 2011. Imaging fossil bone alterations at the microscale by SR-FTIR mi-
crospectroscopy. J Annal At Spectrom 26, 922–929.
Lebon, M., Reiche, I., Bahain, J.J., Chadefaux, C., Moigne, A.M., Fröhlich, F., Sémah, F.,
Schwarcz, H.P., Falguères, C., 2010. New parameters for the characterization of di-
agenetic alterations and heat-induced changes of fossil bone mineral using Fourier
transform infrared spectrometry. J. Archaeol. Sci. 37, 2265–2276.
Leyden, J.J., Wassenaar, L.I., Hobson, K.A., Walker, E.G., 2006. Stable hydrogen isotopes
of bison bone collagen as a proxy for Holocene climate on the Northern Great Plains.
Palaeogeogr. Palaeoclimatol. Palaeoecol. 239, 87–99.
Longinelli, A., 1984. Oxygen isotopes in mammal bone phosphate: a new tool for pa-
leohydrological and paleoclimatological research? Geochem. Cosmochim. Acta 48,
385–390.
Luz, B., Kolodny, Y., 1985. Oxygen isotope variations in phosphate of biogenic apatites,
IV. Mammal teeth and bones. Earth Planet Sci. Lett. 75, 29–36.
Luz, B., Kolodny, Y., Horowitz, M., 1984. Fractionation of oxygen isotopes between
mammalian bone-phosphate and environmental drinking water. Geochem.
Cosmochim. Acta 48, 1689–1693.
McNulty, T., Calkins, A., Ostrom, P., Ganghi, H., Gottfried, M., Martin, L., Gage, D., 2002.
Stable isotope values of bone organic matter: artificial diagenesis experiments and
paleoecology of Natural Trap Cave, Wyoming. Palaios 17, 36–49.
Nützenadel, A., Trentmann, F., 2008. Food and Globalization: Consumption, Markets, and
Politics in the Modern World. Berg, New York.
O'Brien, D.M., Wooller, M.J., 2007. Tracking human travel using stable oxygen and hy-
drogen isotope analyses of hair and urine. Rapid Commun. Mass Spectrom. 21,
2422–2430.
Perren, R., 2006. Taste, Trade and Technology – the Development of the International
Meat Industry since 1840. Ashgate Publishing Ltd., Aldershot.
Person, A., Bocherens, H., Mariotti, A., Renard, M., 1996. Diagenetic evolution and ex-
perimental heating of bone phosphate. Palaeogeogr. Palaeoclimatol. Palaeoecol. 126,
135–149.
Person, A., Bocherens, H., Saliege, J., Paris, F., Zeitoun, V., Gerard, M., 1995. Early di-
agenetic evolution of bone phosphate: an X-ray diffractometry analysis. J. Archaeol.
Sci. 22, 211–221.
Pietsch, S.J., Hobson, K.A., Wassenaar, L.J., Tütken, T., 2011. Tracking cats: problems
with placing feline carnivores on δ18O, δD isoscapes. PLoS One 6, 1–11.
Qi, H., Coplen, T.B., 2011. Investigation of preparation techniques for δ2H analysis of
keratin materials and a proposed analytical protocol. Rapid Commun. Mass
Spectrom. 25, 2209–2222.
Reynard, L.M., Hedges, R.E.M., 2008. Stable hydrogen isotopes of bone collagen in pa-
laeodietary and palaeoenvironmental reconstruction. J. Archaeol. Sci. 35,
1934–1942.
Sauer, P.E., Schimmelmann, A., Sessions, A.L., Topalov, K., 2009. Simplified batch
equilibration for D/H determination of non-exchangeable hydrogen in solid organic
material. Rapid Commun. Mass Spectrom. 23, 949–956.
Schimmelmann, A., Albertino, A., Sauer, P.E., Qi, H., Molinie, R., Mesnard, F., 2009.
Nicotine, acetanilide and urea multi-level 2H, 13C, and 15N-abundance reference
materials for continuous flow isotope ratio mass spectrometry. Rapid Commun. Mass
Spectrom. 23, 3513–3521.
Schoeninger, M.J., DeNiro, M.J., 1984. Nitrogen and carbon isotopic composition of bone
collagen from marine and terrestrial animals. Geochem. Cosmochim. Acta 48,
625–639.
Scorrano, G., Brilli, M., Martínez-Labarga, C., Giustini, F., Pacciani, E., Chilleri, F.,
Scaldaferri, F., Gasbarrini, A., Gasbarrini, G., Rickards, O., 2014. Palaeodiet re-
construction in a woman with probable celiac disease: a stable isotope analysis of
bone remains from the archaeological site of Cosa (Italy). Am. J. Phys. Anthropol.
154, 349–356.
Sealy, J.C., van der Merwe, N.J., Lee-Thorp, J.A., Lanham, J.L., 1987. Nitrogen isotopic
ecology in southern Africa: implications for environmental and dietary tracing.
Geochem. Cosmochim. Acta 51, 2707–2717.
Sellick, M.J., Kyser, T.K., Wunder, M.B., Chipley, D., Norris, D.R., 2009. Geographic
variation of strontium and hydrogen isotopes in avian tissue: implications for
tracking migration and dispersal. PLoS One 4, 1–9.
Snoeck, C., Lee-Thorp, J.A., Schulting, R.J., 2014. From bone to ash: compositional and
structural changes in burned modern and archaeological bone. Palaeogeogr.
Palaeoclimatol. Palaeoecol. 416, 55–68.
Sutoh, M., Koyama, T., Yoneyama, T., 1987. Variations of natural 15N abundances in the
tissues and digesta of domestic animals. Radioisotopes 36, 74–77.
Thompson, T.J.U., Gauthier, M., Islam, M., 2009. The application of a new method of
Fourier Transform Infrared spectroscopy to the analysis of burned bone. J. Archaeol.
Sci. 36, 910–914.
Thompson, T.J.U., Islam, M., Piduru, K., Marcel, A., 2011. An investigation into the in-
ternal and external variables acting on crystallinity index using Fourier Transform
Infrared spectroscopy on unaltered and burned bone. Palaeogeogr. Palaeoclimatol.
Palaeoecol. 299, 168–174.
Topalov, K., Schimmelmann, A., Polly, P.D., Sauer, P.E., Lowry, M., 2013. Environmental,
trophic, and ecological factors influencing bone collagen δ2H. Geochem. Cosmochim.
Acta 111, 88–104.
Tuross, N., Behrensmeyer, A.K., Eanes, E.D., Fisher, L.W., Hare, P.E., 1989. Molecular
preservation and crystallographic alterations in a weathering sequence of wildebeest
bones. Appl. Geochem. 4, 261–270.
Tuross, N., Warinner, C., Kirsanow, K., Kester, C., 2008. Organic oxygen and hydrogen
isotopes in a porcine controlled dietary study. Rapid Commun. Mass Spectrom. 22,
1741–1745.
Wassenaar, L.I., Hobson, K.A., 2003. Comparative equilibration and online technique for
determination of non-exchangeable hydrogen of keratins for use in animal migration
studies. Isot. Environ. Health Stud. 39, 211–217.
Wolf, N., Bowen, G.J., Martinez del Rio, C., 2011. The influence of drinking water on the
δD and δ18O values of house sparrow plasma, blood, and feathers. J. Exp. Biol. 214,
98–103.
Wright, L.E., Schwarcz, H.P., 1996. Infrared and isotopic evidence for diagenesis of bone
apatite at Dos Pilas, Guatemala: palaeodietary implications. J. Archaeol. Sci. 23,
933–944.
Yoder, C., 2012. Let them eat cake? Status-based differences in diet in medieval Denmark.
J. Archaeol. Sci. 39, 1183–1193.
C.A.M. France et al. Journal of Archaeological Science 96 (2018) 33–44
44