ArticlePDF Available

Experimental determination of dietary carbon turnover in bovine hair and hoof


Abstract and Figures

Stable isotopes measured in keratinized tissues like hair or hoof have proven to be a useful tool for reconstructing the dietary history of animals with a weekly to daily resolution. Quantitative reconstruction of dietary preferences requires a precise estimate of tissue turnover by means of controlled feeding experiments. We determined the turnover rates of carbon in hoof and tail hair of growing steers (Bos taurus L., 1758) fed a C 3-based diet, followed by a C4-based diet, for 168 d. As with horses, turnover in steer hair was successfully described by a three-pool modelling approach, with apparent half-lives of 1.7, 7.7, and 69.1 d for each of the pools, each contributing 53%, 20%, and 28% of the total signal, respectively. Two pools only were identified in bovine hoof, which recorded the diet switch more slowly than hair with a reduction in the amplitude of short-term isotope changes. We interpreted this result as a sampling artefact and found that the hooves reflected the same pools as the hair if growth geometry is taken into account. The model parameters defined in this study allowed us to quantitatively reconstruct previous diets of steers of different breeds and individual history with a precision of ±1‰.
Content may be subject to copyright.
Experimental determination of dietary carbon
turnover in bovine hair and hoof
A. Zazzo, S.M. Harrison, B. Bahar, A.P. Moloney, F.J. Monahan, C.M. Scrimgeour,
and O. Schmidt
Abstract: Stable isotopes measured in keratinized tissues like hair or hoof have proven to be a useful tool for reconstruct-
ing the dietary history of animals with a weekly to daily resolution. Quantitative reconstruction of dietary preferences re-
quires a precise estimate of tissue turnover by means of controlled feeding experiments. We determined the turnover rates
of carbon in hoof and tail hair of growing steers (Bos taurus L., 1758) fed a C3-based diet, followed by a C4-based diet,
for 168 d. As with horses, turnover in steer hair was successfully described by a three-pool modelling approach, with appa-
rent half-lives of 1.7, 7.7, and 69.1 d for each of the pools, each contributing 53%, 20%, and 28% of the total signal, re-
spectively. Two pools only were identified in bovine hoof, which recorded the diet switch more slowly than hair with a
reduction in the amplitude of short-term isotope changes. We interpreted this result as a sampling artefact and found that
the hooves reflected the same pools as the hair if growth geometry is taken into account. The model parameters defined in
this study allowed us to quantitatively reconstruct previous diets of steers of different breeds and individual history with a
precision of ±1%.
´:La mesure de la composition en isotopes stables des phane
`res comme le poil ou le sabot est un outil ave
reconstruire l’histoire alimentaire des animaux avec une re
´solution temporelle de l’ordre de la semaine a
`la journe
´e. La re-
construction quantitative des pre
´rences alimentaires ne
´cessite d’estimer pre
´ment le taux de renouvellement du tissu
au moyen d’expe
´riences en conditions d’alimentation contro
´e. Nous avons de
´le taux de renouvellement du car-
bone dans le sabot et le poil de la queue de taurillons (Bos taurus L., 1758) nourris avec une alimentation C3puis C4
pendant 168 j. Comme chez le cheval, le taux de renouvellement du poil est bien de
´crit par un mode
`le a
`trois re
avec des demi-vies apparentes de 1,7, 7,7 et 69,1 j pour chaque re
´servoir contribuant chacun pour 53 %, 20 % et 28 %, du
signal total. Deux re
´servoirs seulement ont pu e
ˆtre identifie
´s dans le sabot, qui enregistre le changement d’alimentation
plus lentement que le poil avec une re
´duction dans l’amplitude des changements isotopiques a
`court terme. Ce re
´sultat est
´comme un artefact d’e
´chantillonnage qui disparaı
´trie de croissance du sabot est prise en compte. La
´termination des parame
`tres du mode
`le nous a permis de reconstruire quantitativement les valeurs de d13C de l’alimen-
tation de taurillons de race et d’histoire individuelle diffe
´rentes avec une pre
´cision de ±1 %.
The stable isotope composition of keratinized tissues, in-
cluding hair, hoof, claw, feather, nail, and horn, has poten-
tial as a high-resolution recorder of the individual history of
animals and humans. This tool has been applied to a wide
array of research areas, ranging from wildlife ecology
(Bearhop et al. 2003; Cerling and Viehl 2004; Cerling et al.
2004) to vegetation and climate change (Witt et al. 1998;
Schnyder et al. 2006) and archaeology (Macko et al. 1999;
Schwarcz and White 2004; Iacumin et al. 2005). The use of
keratinized tissues offers several advantages. They are com-
posed of 90% keratin, a protein which contains all the major
light elements (H, C, N, O, and S). Hair, for example, can
be sampled noninvasively, which is an advantage for wild-
life ecology, especially when working on endangered spe-
cies. Like tooth enamel, keratinized tissues such as hair,
nail, or hoof are metabolically inert once formed, can grow
Received 22 May 2007. Accepted 26 September 2007. Published on the NRC Research Press Web site at on 19 December
A. Zazzo1,2 and O. Schmidt. UCD School of Biology and Environmental Science, Agriculture and Food Science Centre, University
College Dublin, Belfield, Dublin 4, Ireland.
S. Harrison and B. Bahar. UCD School of Biology and Environmental Science, Agriculture and Food Science Centre, University
College Dublin, Belfield, Dublin 4, Ireland; UCD School of Agriculture, Food Science and Veterinary Medicine, Agriculture and Food
Science Centre, University College Dublin, Belfield, Dublin 4, Ireland.
A. Moloney. Teagasc, Grange Beef Research Centre, Dunsany, Co. Meath, Ireland.
F. Monahan. UCD School of Agriculture, Food Science and Veterinary Medicine, Agriculture and Food Science Centre, University
College Dublin, Belfield, Dublin 4, Ireland.
C. Scrimgeour. Scottish Crop Research Institute, Invergowrie, Dundee DD2 5DA, UK.
1Corresponding author (e-mail:
2Present address: Muse
´um national d’Histoire naturelle, De
´partement E
´cologie et Gestion de la Biodiversite
´, USM 303 / Centre National
de la Recherche Scientifique (CNRS) Unite
´Mixte de Recherche (UMR) 5197 « Arche
´ozoologie, histoire des socie
´s humaines et des
peuplements animaux », C.P. 56, 55 rue Buffon, F-75231 Paris CEDEX 05, France.
Can. J. Zool. 85: 1239–1248 (2007) doi:10.1139/Z07-110 #2007 NRC Canada
continuously, and thus are able to record time series up to
several years (Cerling et al. 2004; Iacumin et al. 2005).
Under favourable climatic conditions (cold and (or) dry),
keratin can be preserved physically and chemically for sev-
eral thousand years, making keratinized tissues useful for
palaeoenvironmental and palaeoecological reconstructions
(Macko et al. 1999; Iacumin et al. 2005).
A significant fraction of dietary carbon is rapidly incor-
porated into keratin (Ayliffe et al. 2004; West et al. 2004).
The rate of incorporation of the dietary signal into body tis-
sues can be determined by exposing an animal consecu-
tively to two different diets (or drinking water) of distinct
isotopic composition. Accurate estimation of turnover rate
requires that the studied tissue is initially in isotopic equili-
brium with the first diet, and then approaches equilibrium
with the second diet at the end of the experiment. This ap-
proach has been used to determine turnover rates in a num-
ber of animals (mammals, fish, birds) and for a variety of
tissues (muscle, liver, blood, hair) (Dalerum and Angerbjo
2005). Most studies have used a single exponential best fit
to describe isotope turnover in soft tissues (Fry and Arnold
1982). This approach assumes implicitly that exchange reac-
tions are described by first-order rate kinetics and that one
single metabolic pool is present. It usually describes quite
accurately turnover in muscle tissues and allows the rate
constant to be decomposed into two components if the ani-
mal is gaining mass during the time of the experiment: a
‘growth’’ component and a ‘‘pure’’ turnover component
(Hesslein et al. 1993).
However, for incremental tissues such as hair, Ayliffe et
al. (2004) and Cerling et al. (2007) demonstrated that turn-
over curves were better described with a three-pool model.
This approach, called the reaction progress approach, was
developed based on earlier work using radionucleides
(Thompson 1952a, 1952b; Thompson and Ballou 1956).
Although the exact meaning of each of the metabolic pools
identified by the model is still debated, this method has sev-
eral advantages over the single-pool modelling approach
from a descriptive point of view. These advantages include
a better restitution of initial and final stable isotope values
for both the studied tissue and the reconstructed diet, and
the possibility to consider results from different experiments
together even if the initial and final isotope compositions are
different (Cerling et al. 2007).
Limited data are available on dietary carbon turnover in
hair and none on hoof. To date, dietary carbon turnover in
hair have only been estimated for Mongolian jirds (Meriones
unguiculatus (Milne-Edwards, 1867)) using a single-pool
model (Tieszen et al. 1983) and for horses using a three-
pool model (Ayliffe et al. 2004; West et al. 2004). The
present study represents an additional effort at documenting
how the dietary signal is recorded in animal incremental
tissues. In this paper we calculate turnover rates of carbon
in the hair and hoof keratin of steer (Bos taurus L., 1758)
raised under controlled conditions. Our aims were twofold:
first, to examine if hair and hooves provide a similar re-
cord of the kinetics of carbon turnover following a diet
switch; second, to determine which model better describes
turnover of dietary carbon in bovine hair and hoof and to
test the accuracy of the model to reconstruct bovine dietary
Materials and methods
Seventy continental crossbred steers born in Ireland in
2002 were brought to Teagasc Research Centre in Grange,
County Meath, Ireland, in mid-December 2003. Between
the day of arrival and mid-February 2004, each animal
was raised on a C3-based pre-experimental diet consisting
of grass silage fed ad libitum and 1 kg barley (Hordeum
vulgare L.) / d. From mid-February to mid-March, the ani-
mals were gradually adapted to a barley-based diet over 14 d
with ad libitum access to wheat straw (genus Triticum L.).
The steers were then blocked for breed and initial mass.
Ten steers were slaughtered as a pre-experimental control.
All the remaining animals were offered a control diet
(81% by mass of barley-based concentrate and 19% by
mass of wheat straw). Ten animals, which served as exper-
imental controls, received this diet for the duration of the
whole experiment (168 d). The remaining 50 animals were
gradually (in 2 d) switched from the control diet to an ex-
perimental C4diet (81% by mass of maize (Zea mays L.)
based concentrate and 19% by mass of wheat straw) in
groups of 10 either 168, 112, 56, 28, or 14 d prior to
slaughter. To facilitate handling of the cattle at slaughter,
animals within treatment groups were switched to the new
diet at weekly intervals over a 3-week period. This experi-
ment was carried out with the approval of Teagasc, the
Irish Agriculture and Food Development Authority. All
procedures employed in this study were in accordance
with national regulations concerning animal care and use.
The daily feed allowance was adjusted monthly to achieve
a target mass gain of 0.9 kg/d. The control and experimental
diets were produced in a pilot-scale blending plant at the
Teagasc Moorepark Food Research Centre, County Cork, in
three, unequal batches of 10–15 t each. Monthly feed samples
were taken and analysed at the end of the experiment. Iso-
topic analysis of the control ration showed a slight increase
in d13C value, from –28.4%± 0.3%(n= 5) to –27.8%±
0.3%(n= 5) between February and September 2004. This
slight change was taken into account when calculating ap-
parent fractionation factors between hair or hoof and diet.
Isotopic analysis of the experimental ration showed no trend
in stable isotope values (d13C = –12.8%± 0.3%,n= 9) with
the exception of one outlier in August (–16.4%, see the
Taking into account the relative proportion and digestibil-
ity of the concentrate and straw components of the diets, the
control and experimental diets were estimated to have d13C
values of –28.7%± 0.3%and –14.7%± 0.3%, respec-
tively, creating an isotope spacing of 14.0%between the
two diets.
Hair samples
One tail hair from seven different individuals was selected
for isotope analysis. Sampled animals comprise one animal
from the experimental control group and six animals who re-
ceived the experimental diet for 168 d, respectively. One
heavy (>15 mg) and long (>250 mm) tail hair was chosen
per animal, cleaned using soapy water, and then defatted
following published protocols for other keratin-based tissues
(O’Connell et al. 2001). Each hair was sonicated twice for
1240 Can. J. Zool. Vol. 85, 2007
#2007 NRC Canada
30 min in a solution of methanol and chloroform (2:1, v/v),
rinsed with distilled water, and oven-dried overnight at
60 8C. Individual hairs were serially sectioned into 2–5 mm
subsamples and weighed into ultralight tin capsules for sta-
ble isotope analysis. The size of individual sections was a
compromise to maximize the resolution around the time of
diet-switching while keeping in mind current analytical lim-
its of mass spectrometers used (50 mg carbon). Hair sample
masses ranged between 100 and 200 mg. Up to 50 subsam-
ples were taken from each hair and over 300 samples were
analysed in total.
Hoof samples
The hooves from five individuals that received the exper-
imental diet for 168 d and one control animal were selected
for stable isotope analysis. Of these, three (animals 2, 6, and
one control) correspond to individuals whose hair was also
selected for stable isotope analysis. Details of the sampling
procedure are provided in Harrison et al. (2007a). Briefly,
the left front hoof was separated from the rest of the leg of
the animal. Then, a 15 mm thick slice of hoof was cut paral-
lel to the growth axis using a band saw and the hoof wall
was detached from adhering tissues. The hoof wall was so-
nicated twice for 30 min in a solution of methanol and
chloroform (2:1, v/v), rinsed in distilled water, and oven-
dried at 60 8C overnight. Up to 41 samples/hoof were col-
lected from the surface of the hoof wall using a drill
equipped with a diamond bit. Samples were taken from the
top to 1 cm from the bottom of the hoof, following the hoof
growth axis, and weighed into tin cups. Samples had mean
(SD) dimensions of 3.0 ± 0.6, 1.2 ± 0.1, and 0.8 ± 0.2 mm
(length, width, and depth, respectively; n= 188). Mean (SD)
sample mass was 0.84 ± 0.06 mg (n= 188). The distance
between two consecutive samples was <1 mm.
Stable isotope analysis
Natural abundance stable isotope ratios of carbon (13C/12C)
were measured in hair and hoof samples using an Europa
Scientific ANCA-NT 20-20 stable isotope analyser with
the ANCA-NT solid–liquid preparation module (Europa
Scientific Ltd., Crewe, UK) for the larger samples and an
Europa Scientific Roboprep-CN sample preparation module
coupled with an Europa Scientific 20-20 IRMS for smaller
samples. Isotope ratios are expressed in the dnotation as
per mil (%) according to the equation d13C(%)=[(
12Csample 13C/12Creference)/13C/12Creference]1000, where
the reference is Vienna Pee Dee Belemnite (VPDB). Work-
ing standards of IA-R042 (iso-analytical powdered bovine
liver), IA-R005 (iso-analytical beet sugar), and IAEA-CH-6
(cane sugar) were measured as quality control check sam-
ples during analysis and calibrated against IAEA standards
N1 and N2. The analytical precision achieved for the
standards was 0.1%.
Isotope enrichment factors (%) between tissue and diet
were calculated using the following equation (Craig 1954):
"TD =[(d13CT+ 1000)/(d13CD+ 1000) – 1] 1000, where
d13CTand d13CDare the isotope values of the tissue (hair or
hoof) and diet, respectively, measured at isotopic equilibrium.
Growth rate calculation
Different hairs may grow at different rates within and be-
tween individual steers. To compare the hair chronologies
established for different individuals, it was first necessary to
convert hair length measurements to time. Growth rates
(mm/d) were calculated for each hair by dividing the dis-
tance between the position where the change from control
to experimental diet was first recorded in the hair and hair
base with the time elapsed between diet switch and slaugh-
ter. The centre point of the tail-hair segment that first
showed a significant change in d13C from baseline values
was assumed to mark the commencement of the C4exper-
imental diet. This point was assigned the value of 0 d plus
a lag factor of 12 h to account for the time taken for in-
gested food to be digested and laid down as keratin in the
hair follicle (Ryder 1958; Ayliffe et al. 2004). A similar
approach was followed for calculating growth rates in
hooves (Harrison et al. 2007a). In doing so, we assumed
that growth rate was linear during the time of the experi-
ment (168 d).
To model turnover of dietary carbon as recorded in steer
hair and hooves, we followed the approach outlined in Ay-
liffe et al. (2004) for extracting up to three first-order rate
constants from the data. This approach is described in more
detail in Cerling et al. (2007). Briefly, for a system com-
posed of several pools (in this case three), with all pools ini-
tially at isotopic equilibrium with the dietary input, the
progression to the new isotopic equilibrium state after a
change in diet is given by the following equation (eq. 17 in
Cerling et al. 2007):
where Ais the tissue of interest (e.g., hoof, hair); dinit
and dt
Aare the dvalues at the time of the diet switch, at
equilibrium, and at time tduring the reaction progress ex-
periment, respectively; and and fare the rate constant and
fractional contribution of pools 1, 2, and 3, respectively. We
can solve this equation for all values of fand using a
‘curve stripping’’ approach where the longer lived compo-
nents are successively subtracted to get both the fractions
and the rate constants of all components of the mixture.
The slope and intercept of the straight-line fit through the
data in the part of the curve where the isotopes of the first
two pools have reached their equilibrium states yield esti-
mates of 3and f3. Once these are known, the contribution
of the long-lived pool to the other two pools can be calcu-
lated and subtracted from the total isotope signal, leaving
only the contribution of pools 1 and 2 to the residual isotope
signal. Treating this residual data in the same fashion as be-
fore gives 2and f2, and ultimately 1and f1. All the indivi-
duals (n= 5 and n= 6 for the hoof and the hair data sets,
respectively) were pooled as a group to increase the sam-
pling resolution during the first days following the diet
switch, as well as to take interindividual variability into ac-
count when proposing turnover estimates for carbon in hair
and hoof keratin.
Ayliffe et al. (2004) noted that there is an uncertainty in-
herent in this method regarding the exact position of the t=
0 point in each profile. This temporal uncertainty is equiva-
Zazzo et al. 1241
#2007 NRC Canada
lent to the number of days represented by the first segment
of hair or hoof that shows a significant change in carbon
isotope value. Following the first estimates of the half lives
and fractional contributions derived from the initial semi-log
plot fits, the exact position of the t= 0 point (and hence, of
the rest of the profile) was slightly modified using the itera-
tive approach described in Ayliffe et al. (2004) and West et
al. (2004) to derive improved estimates for the half-life and
fractional contributions of the three components composing
the tail hair. These final estimates are the values provided
in the Results.
We must point out a caveat here. This model assumes that
the pools do not interact with each other. This is the most
simple assumption, and we are aware that it may not be the
most realistic. If, as hypothesized by Ayliffe et al. (2004),
the short-lived pool corresponds to an exogenous source of
carbon coming directly from the diet, whereas the longer
lived pools correspond to endogenous sources of carbon
coming from the metabolic recycling of the element in the
body tissues, they could well interact via the blood stream.
A model of interacting pools would lead to a different math-
ematical manipulation and, more important, to different half-
lives. At the moment knowledge about this interaction is
missing, but the problem may be solved in the future. To
make it clear that there is still a gap in knowledge and to
avoid contradiction with future studies, we will refer to ‘‘ap-
parent half-lives’’ rather than just ‘‘half-lives’’.
Growth rates
On average, hair grew in length 3.5 times faster than hoof
(Table 1). But because of the small hair diameter, the mini-
mum sample mass required by the mass spectrometer dic-
tated that each hair sample accounted for a 5 d period of
growth or about the same as the hoof sampling resolution in
this experiment. However, because the spacing between the
mid-point of two consecutive hoof samples was 1.6 mm,
overall sampling resolution for hoof was 7.4 ± 1.0 d
(mean ± SD; n= 5) compared with 4.8 ± 0.7 d (mean ±
SD; n= 6) in hair.
Isotope enrichment factors
Isotope enrichment factors were calculated using the hair
and hoof of each animal while on the control diet (Table 2).
Because the d13C value of the control diet slightly increased
over the course of the experiment (+0.7%± 0.3%; mean ±
SD), two estimates were calculated for the control animal
after 6 and 30 weeks on the control diet, respectively. For
the other animals, we selected the last segment of hair or
hoof preceding the switch to the experimental diet. Mean
(SD) enrichment factors were 3.0%± 0.1%and 2.8%±
0.1%for the seven hair and six hooves, respectively. The
enrichment factor measured from the control animal after
30 weeks on the control diet was identical within analytical
error to that measured in the other individuals who only
spent 6–8 weeks on that diet. This result strongly suggests
that all the individuals were at, or very close to, isotopic
equilibrium with the control diet at the time of the diet
switch. Hair had, on average, d13C values that were 0.2%
higher than hoof.
Hair and hoof carbon isotope profiles
Figures 1A and 1B show the d13C profiles for the tail
hairs and the hooves; the carbon isotope profile from the
control animal is also presented for comparison. The xaxis
was transformed from length units to time units using the
growth rate estimates presented above, and t= 0 is the time
of the diet switch from control to experimental diet. The tail
hair exhibited a distinctly nonlinear response to the diet
switch, with a rapid increase in d13C values during the first
days after the diet switch, followed by a much slower in-
crease from approximately day 10 to days 72–90 (Fig. 1A).
Based on the turnover pattern in hair and assuming the same
isotope enrichment factor for the experimental C4diet and
the control C3diet, it was calculated that 25%–32% of the
total change in hair d13C values occurred within 1 d. Ac-
cordingly, the change in d13C values observed was ~63% ±
4% at day 7 after the diet switch and ~88% ± 3% at days
70–90 after the diet switch. At first glance, the record of
the diet switch in hoof looks very similar to that of the hair,
with first a rapid rise and then followed by a slower increase
in d13C values (Fig. 1B). After days 70–90 on the experi-
Table 1. Growth rates and sampling resolution in steer (Bos taurus) hair and hoof.
Growth rate (mm/d) Time represented by each
sample during diet switch (d) Time between two consecutive
samples during diet switch (d)
ID No. Hair Hoof Hair Hoof Hair Hoof
2 0.795 0.219 5.0 5.5 5.0 7.3
3 0.256 — 4.7 6.3
4 0.220 — 5.5 7.3
5 0.212 — 5.7 7.5
6 0.665 0.170 4.5 7.1 4.5 9.4
7 0.787 — 3.8 3.8
8 0.785 — 5.1 5.1
9 0.698 — 4.3 4.3
10 0.840 — 6.0 6.0
Mean 0.762 0.215 4.8 5.6 4.8 7.4
SD (1) 0.060 0.027 0.7 0.8 0.7 1.0
1242 Can. J. Zool. Vol. 85, 2007
#2007 NRC Canada
mental diet, 88% ± 3% of the expected change in hoof d13C
values had taken place, which is identical to hair. But the
rate of initial increase in d13C values was much lower in
hoof, as it took ~20 d for the hoof to record 60% of the ex-
pected change, which is three times longer than found in the
After days 70–90 on the experimental diet, all the animals
recorded a sharp decrease in their d13C values, ranging from
2.5%to 6.8%in hair and from 1.8%to 2.9%in hoof. This
episode was followed by an increase of similar magnitude a
few weeks later. Dates were calculated for these two un-
planned diet switches assuming a constant growth rate for
both hair and hoof between the time of the switch to the ex-
perimental diet and slaughter (Fig. 2). All animals received
this unknown diet for 20–40 d, starting in late June to early
July 2004, except for animal 3 whose diet was switched
later in July. Isotopic analysis of seven individual C4pellets
coming from the batch of food that was offered to the ani-
mals during this period showed that this food was incorrectly
formulated. A mean (SD) d13C value of –18.2%± 1.0%,
(n= 12) was calculated. These pellets were offered with
C3straw, so the value of the composite diet during this
time should be close to –19.5%± 1.0%, suggesting that
as much as 30%–40% barley was added to maize during
the preparation of the pellets. However, these two addi-
tional diet switches occurred at a time when the hair and
hooves were close to equilibrium with the C4diet. There-
fore, it did not prevent us from modelling the hair and
hoof responses to the first diet switch, but it provided us
with the possibility of testing the accuracy of the model
parameter estimates (see Discussion).
Multicomponent model
The modelling approach outlined in Ayliffe et al. (2004)
and Cerling et al. (2007) was applied to the hair and hoof
data sets (Table 3). Only data points prior to the unplanned
July diet switch were included. For hair, the model sug-
gested the existence of three pools with apparent half-lives
of 1.7, 7.7, and 69.1 d, respectively. These pools accounted
for 53% ± 1%, 20% ± 8% and 28% ± 4% of the total diet-
ary change of 14%. By contrast, the isotope record in hoof
could only be decomposed into two pools. The first pool had
a rate constant of 11.7 d and accounted for 52% ± 4% of the
total change, whereas the second pool had a longer apparent
half-life of 34.0 d and accounted for 45% ± 7% of the total
change. Because the shape of the isotope profile in hoof was
sigmoidal rather than exponential and because of interindi-
vidual differences in turnover rates, the model fit was not
as good as for hair. This was reflected by the pools account-
ing only for 97% of the total variation in hoof. It overesti-
mated the hoof response in the first 20 d following the diet
switch and underestimated it for the next 20 d (Fig. 3).
The isotope record in steer hoof and hair
Sequential sampling and stable isotope analysis of bovine
tail hair and hoof revealed that the two tissues can provide a
detailed and continuous record of animal dietary history.
Based on their maximum length and average growth rate
calculated over a 168 d period, this record potentially ex-
tends back to ~15 month prior to the sampling date for both
hair and hoof. Temporal resolution, which is mainly deter-
mined by the tissue growth rate and current analytical limits
(a minimum of 50 mg C is routinely required for d13C anal-
ysis), is 5 d/sample, on average, for both tissues. Hair thick-
ness decreases from base to end, and temporal resolution is
maximal close to the hair base where a precision in the
range of 1–2 d can be achieved. Similarly, the temporal res-
olution obtained in hoof, which is partly limited by the
width of each sample band, could be improved by the use
of a computer-controlled micromilling device. This techni-
que, which was developed for fish otoliths and also applied
to mammal teeth, uses the edge of the drill bit instead of its
full thickness to collect a series of contiguous (and thinner)
samples (Wurster et al. 1999; Zazzo et al. 2005).
The record of diet switch from C3to C4was rapid in hair
and hoof, with most carbon (90%) having turned over in
Table 2. Isotope enrichment factors for steer hair and hoof.
d13 (%)"(%)
ID No. Equilibration
period (weeks)aDiet Hair Hoof Hair–diet Hoof–diet
Control 6 –28.7 –25.9 –25.7 2.8 3.0
Control 30 –28.0 –25.1 –25.3 3.0 2.8
2 7 –28.7 –25.8 –25.9 2.9 2.9
3 6 –25.9 — 2.8
4 6 –26.0 — 2.7
5 7 –25.8 — 2.9
6 8 –28.7 –25.7 –26.1 3.0 2.6
7 6 –28.7 –26.0 — 2.8
8 8 –28.7 –25.7 — 3.0
9 8 –28.7 –25.6 — 3.1
10 8 –28.7 –25.7 — 3.0
Mean 3.0 2.8
SD (1) 0.1 0.1
aThe 6-, 7-, and 8-week equilibration periods reflect the 3-week interval that facilitated the staggered
assignment of animals to the C4diet.
Zazzo et al. 1243
#2007 NRC Canada
<3 months. This is faster than for other incremental tissues
like teeth. Full mineralization of bovine tooth enamel re-
quires 6–7 months; correspondingly, the time represented in
each dentine sample is about 4 months (Balasse 2002; Zazzo
et al. 2006). Although they appear similar, the dietary record
in hoof was different from that in hair in terms of the time
resolution. Figures 4A and 4B compare the hair and hoof
isotope profiles for individuals 2 and 6 for which both tis-
sues were sampled. Following the first diet switch, the rate
of increase in d13C values was more rapid for hair than for
hoof. The delay between the two records was estimated to
be between 10 and 20 d. Similarly, the record of the second
and third diet switch in July–August was delayed in the hoof
compared with in the hair. Finally, the decrease in d13C
value during this period was less marked in hoof (2/3 of the
total amplitude) than in hair from the same individual.
Mathematically, hoof turnover rate could only be decom-
posed into two pools of equivalent size and relatively slow
apparent half-lives of 11.7 and 34 d, with the rapid pool
found for hair absent from the model. It is noteworthy that
similar observations, namely a slow response to a rapid diet-
ary change, associated with a large degree of attenuation
compared with the input (dietary) signal have been de-
scribed during the formation and maturation of hypsodont
(ever-growing) tooth enamel (Balasse 2002; Passey and
Cerling 2002; Zazzo et al. 2005).
-80 -60 -40 - 20 0 20 40 60 80 1 00 12 0 140 1 60
Time after diet switch (d)
C (VPDB,‰) δ
C (VPDB,‰)
-80 -60 -40 -20 0 20 40 60 80 100 120 140 160
Fig. 1. Stable carbon isotope profiles in six steer (Bos taurus) hairs
(A) and five hooves (B) expressed as time after the switch from
control to experimental diet. Result for one control animal is also
shown for comparison. Scaling from millimetres to days was per-
formed individually by calculating the tissue’s growth rates, based
on the known time and position where the diet switch was first re-
corded in either hoof or hair.
02/06 1 6 / 06 3 0/0 6 1 4/0 7 28 /0 7 1 1/ 08 2 5/0 8 08/ 09 22 / 09
02/06 16/06 30/06 14/07 28/07 11/08 25/08 08/09 22/09
Calendar date (dd/mm)
C (VPDB, ‰) δ
C (VPDB, ‰)
Fig. 2. Carbon isotope profiles in steer hair (A) and hoof (B) de-
tailing the last 16 weeks of the experiment.
1244 Can. J. Zool. Vol. 85, 2007
#2007 NRC Canada
We suggest that the differences observed between hoof
and hair could be related to the timing and geometry of
hoof formation, which directly influences the time repre-
sented in each hoof subsample. The diameter of steer tail
hair is very small (<0.1 mm), and the time represented in
each section of hair is the time needed for this section of
hair to grow in length. In contrast, the hoof wall grows in
length and increases in thickness over the first 2–3 cm
where it reaches its full thickness (Blowey 1993; Maierl
and Mu
¨lling 2004). Like in tooth enamel, newly formed ker-
atin cells are produced along an apposition front, which
makes an angle with the surface of the hoof wall in the first
3 cm from the top of the hoof. Because the hoof apposition
front is partly oblique, sampling into the hoof thickness per-
pendicular to the surface of the hoof will crosscut younger
layers of hoof keratin. If the angle between the apposition
front and the hoof surface is large, the time represented by
each subsample will exceed the time necessary to grow in
To express this idea mathematically, we propose a simple
model that takes the growth of hoof keratin into account:
½2d13Choof ;i¼
where d13Choof,iand d13Chair,iare the carbon isotope compo-
sition of hoof and hair at time i, respectively, and tis the
time represented in each sample of hoof (d). The model as-
sumes that hoof keratin is formed at the same rate as hair
keratin and, like hair, is fully formed once deposited (no
maturation time). Mathematically, this is equivalent to ap-
plying a moving average function to the hair stable isotope
record. Significant attenuation of short-term isotope varia-
tions recorded in tail hair will occur if tis longer than the
time represented in the individual hair subsample (about
5 d in this study). For individual 2, the best correspondence
between the model and the hoof profile was found for t=
20; for individual 6, the best correspondence was found
when t= 30 (Fig. 4).
The model successfully reproduces the time lag in the
hoof profile relative to the hair after the first diet switch
and the amplitude of the decrease in d13C values during the
second diet switch. According to our model, the hoof layer
located 1.0 mm deep was formed 20–30 d earlier than the
hoof located on the surface. Values of tseem to be inversely
correlated with hoof growth rate, as growth rate in animal 6
was much lower than in animal 2 (Table 1). This observa-
tion was confirmed independently by Harrison et al.
(2007b) on a larger number of individuals. A better under-
standing of hoof growth in the three dimensions of space
will indeed be required to test the accuracy of this very sim-
ple model. For this reason, and because the temporal resolu-
tion of the dietary record was higher in hair, we will restrict
most of the following discussion to hair keratin.
Implications for dietary reconstruction
To test the accuracy of the model parameters to recon-
struct bovine dietary history, we compared, for each individ-
ual, the measured value of the diet received by the animal
with the value predicted by the model developed by Cerling
et al. (2004). Cerling et al. (2004) showed that the isotopic
composition of an animal’s diet could be reconstructed at
any time tfrom a sequence of hair samples when the turn-
over rates and fractions of the three different pools, as well
as the fractionation factor between hair and diet, are known.
Also, the initial condition for the different pools must be de-
fined. As pointed out by Cerling et al. (2004), it takes about
Table 3. Apparent half-lives and fractions for steer hair and hoof
carbon pools.
12 3
Apparent half-life (d) 1.67±0.03 7.74±0.85 69.1±3.6
Fraction 0.53±0.01 0.20±0.08 0.28±0.04
Apparent half-life (d) 11.74±1.00 34.04±2.5
Fraction 0.52±0.04 0.45±0.07
2-pool model
hoof data set
C (VPDB, ‰)
ime (d)
3-pool model
hair data set
C (VPDB, ‰)
Fig. 3. Model predictions (solid lines) plotted with measured data
for the steer hair (A) and hoof (B) data sets. Model parameters are
reported in Table 3.
Zazzo et al. 1245
#2007 NRC Canada
5 half-lives for a pool to reach isotopic equilibrium (>95%).
This means that there is little memory of the initial condi-
tions chosen for the short pool after 10 d or so, which corre-
sponds to the first two hair samples of the profile. At t=0,
we assume that the two short pools are in isotopic equili-
brium with the pre-experimental diet and have a d13C value
of –26%. Based on the known long-term dietary history of
the animals, we choose a slightly lower d13C value of –27%
for the long pool. Diets of the individual steer while on the
experimental diet were calculated, assuming a fractionation
factor of 1.0030 between hair and diet (Table 2). We mod-
elled the d13C value for the incorrectly formulated diet that
was offered to the animals during July–August based on
the lowest value measured in each hair during that time
window. Diet values, predicted for each individual, ranged
between –19.0%and –22.2%except for one individual
that had a lower value of –24.1%. The mean (SD) pre-
dicted value (–21.0%± 1.1%) is 1.5%lower than the
mean (SD) d13C value (–19.5%± 1.0%) of the composite
diet during this time. The discrepancy between the model
estimate and the measured value for the composite diet
could either originate from the fact that the pellets sampled
in August were not representative of the average food
value offered during the entire summer, or that the model
parameters are inaccurate for this period of time. For ex-
ample, an increase in the fraction of the rapid pool during
and after the second diet switch could bring the diet values
closer to the measured food value. It is not possible to de-
cipher between the two hypotheses.
The second test case is based on the data of Jones et al.
(1981). In this study, two steers were raised on Pangola
grass (Digitaria eriantha Steud. subsp. pentzii (Stent) Kok),
4plant, then switched to cowpea hay (Vigna sinensis
(L.)), a C3plant, for 123 d and switched back to their origi-
nal C4diet for another 115 d. One of the two sampling
methods used was to repeatedly clip patches of hair to skin
level with electric shears, so each sample consisted of all the
hair that had re-grown since the last sampling occasion. The
d13C values, as well as the time represented in each subsam-
ple, were estimated graphically from Fig. 1 in Jones et al.
(1981) and plotted in Fig. 5. The diet was predicted assum-
ing that the long-term pool of the steer hair was in equili-
brium with the first C4diet (–11%) and a fractionation
factor of 1.0018 between hair and diet as stated in Jones et
al. (1981). The model overestimates the first shift to C3by
1%, on average, which corresponds to 7% of the amplitude
of the diet switch. However, it successfully estimates within
analytical error the value of the C4diet after the second
shift. In conclusion, we have shown that the model parame-
ters derived from our controlled feeding study can be suc-
cessfully applied to reconstruct the unknown diet of steers
from our experiment, and from steers of different breed and
feeding regime with an associated error of about 1%.
Comparison with other mammals
Prior to this study, carbon isotope turnover rates had only
been determined for hair from two other mammals: Mongo-
lian jirds (Tieszen et al. 1983) and horses (Ayliffe et al.
2004; West et al. 2004). In the case of Mongolian jirds,
Tieszen et al. (1983) reported a turnover rate of 47.5 d using
an exponential fit model. This estimate should be treated
with caution because the experimental design was not suited
to measuring hair turnover. The Mongolian jirds were sacri-
ficed at different time intervals after the diet switch, and
since the paper does not mention whether newly grown
patches of hair were shaved over the course of the experi-
ment, each hair sample probably integrated a period of time
longer than just the time the animals were on the experimen-
tal diet. High-resolution sampling is required to apply the
multicompartmental approach successfully and this was
only done for horses with individual samples representing
as little as 1 d (Ayliffe et al. 2004; West et al. 2004). Simi-
-80 -40 0 40 80 120 160 -80 -40 0 40 80 120 160
Hoof 6
Hair 6
30 d moving
ave rag e
Hoof 2
Hair 2
20 d moving
Time from first diet switch (d)
C (VPDB, ‰)
Time from first diet switch (d)
(A) (B)
Fig. 4. Comparison between the carbon isotope profiles measured in the hoof (*) and hair (*) of steer individuals 2 (A) and 6 (B). A
moving average of 20 and 30 d was applied to the hair data set from individuals 2 and 6, respectively.
1246 Can. J. Zool. Vol. 85, 2007
#2007 NRC Canada
lar to the horse study, we were able to decompose the car-
bon isotope record of a C3–C4diet switch in hair into three
pools of different apparent half-lives. The main difference
between the cattle and the horse results was in the sizes and
half-lives of the pools. The apparent half-lives of the two
short-lived pools in steers (1.7 and 7.7 d) were two to three
times longer than the values for horses (0.5 and 4.4 d).
However, the apparent half-life of the long-term pool in
steers was half (69 d) of that in horses (136 d). The two
short-term pools made up 70% of the total C in steer hair
keratin compared with only 56% in horses. These differen-
ces indicate that carbon turnover in hair is faster in steers
than in horses: 88% of carbon in steer hair was turned over
after 10–13 weeks compared with 77% after 21 weeks in
horses. Trying to find an explanation is difficult because
there are too many differences between the two experiments,
such as digestive physiology (hindgut versus foregut fermen-
tation for horses and cattle, respectively), age (young cattle
versus mature horses), and food quality (percent crude pro-
tein, digestibility), that are likely to play a role in turnover
of dietary carbon and that differ. Also, the exact nature of
the pools remains uncertain. Ayliffe et al. (2004) hypothe-
sized that the pool with the fast turnover rate corresponds to
carbon that comes directly from the food (exogenous car-
bon), whereas the other two pools correspond to an endoge-
nous source derived from the turnover of metabolically
relatively very active (liver) and less active (muscle, colla-
gen, or connective tissues) tissues within the body. The
present experiment did not allow us to test the validity of
these hypotheses. There is obviously still a need for more
controlled studies to separate the effects of metabolism, di-
gestive physiology, food quality, age, nutritional status, and
tissue growth on hair and hoof turnover rates. However, the
similarity in the number of pools identified suggests that the
turnover of dietary carbon is controlled by the same proc-
esses in horse and cattle hair, and may be representative of
other herbivores.
In conclusion, continuously growing keratinized tissues
like hair and hoof provide a long-term record of bovine diet-
ary history with a higher temporal resolution than any other
biological tissue, including teeth. Although our data set only
covers the last 7 months of the animal’s life, there is suffi-
cient material in hair and hoof of steers to obtain dietary in-
formation for up to 15 months. The record of diet switch
from C3to C4was rapid in hair and hoof and about 90% of
the carbon had turned over in <3 months. A treatment of the
data through the modelling approach developed by Ayliffe
et al. (2004) showed that the turnover of carbon in the
steers’ hair as in horses could be decomposed into three res-
ervoirs with fast, intermediate, and slow turnover rates. Two
pools could only be identified in hooves, which responded
more slowly to the diet switch. The model parameters de-
fined in this study allowed us to quantitatively reconstruct
previous diets of steers of different breed and individual his-
tory. The fact that the same model can describe hair turn-
over in two different herbivores raised on different diets
and having very different digestive physiology suggests that
the reaction progress variable approach could potentially be
applied to a large range of mammal species of unknown di-
etary history, on condition that hair and hoof growth rates
can be assessed independently of the carbon isotope values.
Because hair can be sampled repeatedly and noninvasively,
we anticipate that this approach will prove more and more
useful for the investigation of short-term wildlife move-
ments and changes in dietary preferences.
This research was funded by an Irish Research Council
for Science, Engineering, and Technology postdoctoral fel-
lowship (A.Z.), the EU 6th Framework Programme TRACE
project (S.H.), a Teagasc Walsh Fellowship (B.B.), and the
Irish Government National Development Plan (2000–2006).
The Scottish Crop Research Institute is grant-aided by the
Scottish Executive Rural Affairs Department. The authors
are much indebted to the staff of the Teagasc Grange Beef
Research Centre, in particular to V. McHugh for help with
sample collection and animal care. We thank M. Cooney
for technical assistance during the preparation of the hoof
samples. We also thank S. Brookes (Iso-Analytical Limited,
Sandbach Cheshire, UK) for taking responsibility in running
the smallest hair samples. Finally, we wish to extend our
gratitude to T.E. Cerling for stimulating discussions while
driving on the roads of Ireland and for introducing us to the
reaction progress variable approach.
Ayliffe, L.K., Cerling, T.E., Robinson, T., West, A., Sponheimer,
measured hair
modeled diet
measured diet
-20 0 20 40 6 0 80 1 0 0 120 1 40 160 180 20 0 220 2 40
Time after first diet switch (d)
C (VPDB, ‰)
Fig. 5. Comparison between measured (broken line) and predicted
(*) diet d13C values based on d13C values of steer hair (*) mea-
sured during a diet-switch experiment. Data are from Jones et al.
(1981). For each sampling date, a point reflects the mean hair d13C
value of two individuals. Location and d13C value of each point are
based on a graphic estimate owing to the absence of a table of re-
sults in Jones et al. (1981). Predicted diet values were calculated
using an isotope fractionation factor of 1.0018 as proposed in Jones
et al. (1981), the three-pool model parameters defined in this study,
and the forward model outlined in Cerling et al. (2004). Although
no information was available regarding the previous long-term diet
of the steers, we assumed that the animals were in equilibrium with
their initial C4diet at the beginning of the experiment.
Zazzo et al. 1247
#2007 NRC Canada
M., Passey, B., Hammer, J., Roeder, B., Dearing, D., and Ehler-
inger, J. 2004. Turnover of stable isotopes in tail hair and breath
CO2of horses fed an isotopically varied diet. Oecologia (Berl.),
139: 11–22. doi:10.1007/s00442-003-1479-x. PMID:14730442.
Balasse, M. 2002. Reconstructing dietary and environmental his-
tory from enamel isotopic analysis: time resolution of intra-tooth
sequential sampling. Int. J. Osteoarch. 12: 155–165. doi:10.
Bearhop, S., Furness, R.W., Hilton, G.M., Votier, S.C., and Wal-
dron, S. 2003. A forensic approach to understanding diet and habi-
tat use from stable isotope analysis of (avian) claw material.
Funct. Ecol. 17: 270–275. doi:10.1046/j.1365-2435.2003.00725.x.
Blowey, R.W. 1993. Foot structure and function and laminitis. In
Cattle lameness and hoof care: an illustrated guide. Farming
Press, Ipswich, UK. pp. 3–18.
Cerling, T.E., and Viehl, K. 2004. Seasonal diet changes of the for-
est hog (Hylochoerus meinertzhageni Thomas) based on the car-
bon isotopic composition of hair. Afr. J. Ecol. 42: 88–92. doi:10.
Cerling, T.E., Passey, B.H., Ayliffe, L.K., Cook, C.S., Ehleringer,
J.R., Harris, J.M., Dhidha, M.B., and Kasiki, S.M. 2004. Or-
phan’s tales: seasonal dietary changes in elephants from Tsavo
National Park, Kenya. Palaeogeogr. Palaeoclimatol. Palaeoecol.
206: 367–376. doi:10.1016/j.palaeo.2004.01.013.
Cerling, T.E., Ayliffe, L.K., Dearing, M.D., Ehleringer, J.R., Pas-
sey, B.H., Podlesak, D.W., Torregrossa, A.M., and West, A.G.
2007. Determining biological tissue turnover using stable iso-
topes: the reaction progress variable. Oecologia (Berl.), 151:
175–189. doi:10.1007/s00442-006-0571-4. PMID:17186277.
Craig, H. 1954. Carbon-13 in plants and the relationships between
carbon-13 and carbon-14 variations in nature. J. Geol. 62: 115–
Dalerum, F., and Angerbjo
¨rn, A. 2005. Resolving temporal variation
in vertebrate diets using naturally occurring stable isotopes. Oe-
cologia (Berl.), 144: 647–658. doi:10.1007/s00442-005-0118-0.
Fry, B., and Arnold, C. 1982. Rapid 13C/12C turnover during
growth of brown shrimp (Penaeus aztecus). Oecologia (Berl.),
54: 200–204. doi:10.1007/BF00378393.
Harrison, S.M., Zazzo, A., Bahar, B., Monahan, F.J., Moloney,
A.P., Scrimgeour, C.M., and Schmidt, O. 2007a. Using hooves
for high-resolution isotopic reconstruction of bovine dietary his-
tory. Rapid Commun. Mass Spectrom. 21: 479–486. doi:10.
1002/rcm.2861. PMID:17230438.
Harrison, S.M., Monahan, F.J., Zazzo, A., Bahar, B., Moloney, A.P.,
Scrimgeour, C.M., and Schmidt, O. 2007b. Three-dimensional
growth of bovine hoof as recorded by carbon stable isotope
ratios. Rapid Commun. Mass Spectrom. 21: 3971–3976.
Hesslein, R.H., Hallard, K.A., and Ramlal, P. 1993. Replacement
of sulfur, carbon and nitrogen in tissue of growing broad White-
fish (Coregonus nasus) in response to a change in diet traced by
d34S, d13C and d15N. Can. J. Fish. Aquat. Sci. 50: 2071–2076.
Iacumin, P., Davanzo, S., and Nikolaev, V. 2005. Short-term cli-
matic changes recorded by mammoth hair in the Arctic environ-
ment. Palaeogeogr. Palaeoclimatol. Palaeoecol. 218: 317–324.
Jones, R.J., Ludlow, M.M., Troughton, J.H., and Blunt, C.G. 1981.
Changes in the natural carbon isotope ratios of the hair from
steers fed diets of C4,C
3and C4species in sequence. Search,
12: 85–87.
Macko, S.A., Lubec, G., Teschler-Nicola, M., Andusevich, V., and
Engel, M.H. 1999. The Ice Man’s diet as reflected by the stable
nitrogen and carbon isotopic composition of his hair. FASEB J.
13: 559–562. PMID:10064623.
Maierl, J., and Mu
¨lling, C. 2004. Funktionelle anatomie. In Erkran-
kungen der Klauen und Zehen des Rindes. Edited by A. Fiedler,
J. Maierl, and K. Nuss. Schattauer GmbH, Stuttgart, Germany.
pp. 1–28.
O’Connell, T.C., Hedges, R.E.M., Healy, M.A., and Simpson,
A.H.R.W. 2001. Isotopic comparison of hair, nail and bone:
modern analyses. J. Archaeol. Sci. 28: 1247–1255. doi:10.1006/
Passey, B.H., and Cerling, T.E. 2002. Tooth enamel mineralization
in ungulates: implications for recovering a primary isotopic time
series. Geochim. Cosmochim. Acta, 66: 3225–3234. doi:10.
Ryder, M.L. 1958. Nutritional factors influencing hair and wool
growth. In The biology of hair growth. Edited by W. Montagna
and R.A. Ellis. Academic Press, New York. pp. 305–335.
Schnyder, H., Schwertl, M., Auerswald, K., and Scha
¨ufele, R.
2006. Hair of grazing cattle provides an integrated measure of
the effects of site conditions and interannual weather variability
on d13C of temperate humid grassland. Glob. Change Biol. 12:
1315–1329. doi:10.1111/j.1365-2486.2006.01169.x.
Schwarcz, H.P., and White, C.D. 2004. The grasshopper or the ant?
Cultigen-use strategies in ancient Nubia from C-13 analyses of
human hair. J. Archaeol. Sci. 31: 753–762. doi:10.1016/j.jas.
Thompson, R.C. 1952a. Studies of metabolic turnover with tritium
as a tracer. I. Gross studies on the mouse. J. Biol. Chem. 197:
81–87. PMID:12981036.
Thompson, R.C. 1952b. Studies of metabolic turnover with tritium
as a tracer. II. Gross studies on the rat. J. Biol. Chem. 200: 731–
Thompson, R.C., and Ballou, J.E. 1956. Studies of metabolic turn-
over with tritium as a tracer. V. The predominantly non-dynamic
state of body constituents in the rat. J. Biol. Chem. 214: 795–809.
Tieszen, L.L., Boutton, T.W., Tesdahl, K.G., and Slade, N.A. 1983.
Fractionation and turnover of stable carbon isotopes in animal
tissue: implications for d13C analysis of diet. Oecologia (Berl.),
57: 32–37. doi:10.1007/BF00379558.
West, A.G., Ayliffe, L.K., Cerling, T.E., Robinson, T.F., Karren, B.,
Dearing, M.D., and Ehleringer, J.R. 2004. Short-term diet changes
revealed using stable carbon isotopes in horse tail-hair. Funct.
Ecol. 18: 616–624. doi:10.1111/j.0269-8463.2004.00862.x.
Witt, G.B., Moll, E.J., Beeton, R.J.S., and Murray, P.J. 1998.
Isotopes, wool, and rangeland monitoring: let the sheep do
the sampling. Environ. Manag. 22: 145–152. doi:10.1007/
s002679900091. PMID:9419292.
Wurster, C.M., Patterson, W.P., and Cheatham, M.M. 1999. Ad-
vances in micromilling techniques: a new apparatus for ac-
quiring high-resolution oxygen and carbon stable isotope
values and major/minor elemental ratios from accretionary car-
bonate. Comput. Geosci. 25: 1159–1166. doi:10.1016/S0098-
Zazzo, A., Balasse, M., and Patterson, W.P. 2005. High-resolution
d13C intra-tooth profiles in bovine enamel: implications for
mineralization pattern and isotopic attenuation. Geochim. Cos-
mochim. Acta, 69: 3631–3642. doi:10.1016/j.gca.2005.02.031.
Zazzo, A., Balasse, M., and Patterson, W.P. 2006. The reconstruc-
tion of mammal individual history: refining high-resolution iso-
tope record in bovine tooth dentine. J. Archaeol. Sci. 33: 1177–
1187. doi:10.1016/j.jas.2005.12.006.
1248 Can. J. Zool. Vol. 85, 2007
#2007 NRC Canada
... Stable isotope analyses of keratinized tissues, such as hair, have been useful for dietary reconstruction of ruminant diets 31 , as well as other species 32 . Zazzo, et al. 31 also detected dietary changes through analysis of hair in beef steers consuming C 3 or C 4 forage-based diets. ...
... Stable isotope analyses of keratinized tissues, such as hair, have been useful for dietary reconstruction of ruminant diets 31 , as well as other species 32 . Zazzo, et al. 31 also detected dietary changes through analysis of hair in beef steers consuming C 3 or C 4 forage-based diets. The use of keratinized tissues offers several advantages since keratin generally contains all the major light elements (C,H, N, O, and S) and can be sampled non-invasively 31 . ...
... Zazzo, et al. 31 also detected dietary changes through analysis of hair in beef steers consuming C 3 or C 4 forage-based diets. The use of keratinized tissues offers several advantages since keratin generally contains all the major light elements (C,H, N, O, and S) and can be sampled non-invasively 31 . Hair growth rates and length will pose an inherent challenge to the accurate representation of the animal's diet, especially since these vary between breed and individuals 31 . ...
Full-text available
Stable isotopes are useful for estimating livestock diet selection. The objective was to compare δ¹³C and δ¹⁵N to estimate diet proportion of C3–C4 forages when steers (Bos spp.) were fed quantities of rhizoma peanut (Arachisglabrata; RP; C3) and bahiagrass (Paspalumnotatum; C4).Treatments were proportions of RP with bahiagrass hay: 100% bahiagrass (0%RP); 25% RP + 75% bahiagrass (25%RP); 50% RP + 50% bahiagrass (50%RP); 75% RP + 25% bahiagrass (75%RP); and 100% RP (100% RP). Feces, plasma, red blood cell (RBC), and hair were collected at 8-days intervals, for 32 days. Two-pool mixing model was utilized to back-calculate the proportion of RP based on the sample and forage δ¹³C or δ¹⁵N. Feces showed changes using δ¹³C by 8 days, and adj. R² between predicted and observed RP proportion was 0.81 by 8 days. Plasma, hair, and RBC required beyond 32-days to reach equilibrium, therefore were not useful predictors of diet composition during the study. Diets were best represented using fecal δ¹³C at both 8-days and 32-days. By 32-days, fecal δ¹⁵N showed promise (R² = 0.71) for predicting diet composition in C3–C4 diets. Further studies are warranted to further corroborate fecal δ¹⁵N as a predictor of diet composition in cattle.
... ex. Ambrose, 2002;Ayliffe et al., 2004;Sponheimer et al., 2003bSponheimer et al., , 2003aZazzo et al., 2007). Dans le registre archéologique, les tissus les plus couramment analysés sont les os, l'émail et la dentine et plus rarement la kératine (poil et corne). ...
... ex. Ayliffe et al., 2004;Bendrey et al., 2015;Cerling and Harris, 1999;Flockhart et al., 2015;Green et al., 2018b;Lewis et al., 2017;Podlesak et al., 2008;Zazzo et al., 2010Zazzo et al., , 2007 ainsi que l'influence des stratégies d'élevage sur les compositions isotopiques (p. ex. ...
... En effet, des études d'alimentation contrôlée sur des animaux domestiques ont montré que la composition isotopique du carbone d'un nouveau régime alimentaire est rapidement enregistrée dans le poil et permet de détecter dans l'alimentation des changements dans l'alimentation de court terme, de l'ordre de quelques jours. Cette rapidité d'enregistrement offre une plus haute résolution temporelle que celle d'autres tissus biologiques West et al., 2004;Zazzo et al., 2008Zazzo et al., , 2007. ...
Les sociétés pastorales laissent des marques discrètes dans le registre archéologique. De ce fait, la description et la caractérisation fines des interactions homme-animal-environnement dans le passé au sein de ces sociétés est difficile. L’analyse isotopique des tissus biologiques des animaux d’élevage est une clé d’entrée pour comprendre ces interactions car elle livre des informations sur l’histoire alimentaire, la mobilité et l’environnement de ces animaux. Cette approche nécessite néanmoins la mise en place de référentiels actuels de qualité afin d’interpréter correctement les résultats obtenus en contexte archéologique. Le présent travail a pour objectif de valider des marqueurs isotopiques (C, N, O, Sr) de la mobilité pastorale et de la date de mort du bétail. Nous avons pour cela équipé de colliers GPS le bétail (cheval, mouton, chèvre) de plusieurs troupeaux appartenant à des éleveurs nomades de l’Altaï mongol et exploité les données issues de l’analyse isotopique séquentielle des tissus biologiques à croissance continue ou prolongée (poil, dent) de ces animaux.Nos résultats montrent que les déplacements fréquents des animaux compliquent l’interprétation des variations isotopiques enregistrées par les crins des chevaux. En revanche, la valeur moyenne de δ13C du poil et de l’émail ainsi que l’anticovariation entre les valeurs de δ13C et δ18O mesurées dans l’émail peuvent être exploitées pour inférer la mobilité altitudinale des animaux et le taux d’occupation des étages alpins. La cartographie isotopique en strontium de la zone d’étude nous a permis de discuter l’utilisation des variations isotopiques en Sr de l’émail pour retracer l’origine géographique et les déplacements des animaux. Enfin, la modélisation des séquences saisonnières enregistrées par le δ18O de l’émail donne accès à la date de mort des caprinés avec une précision de l’ordre du mois. L’analyse isotopique de l’émail des chevaux archéologiques trouvés en contexte rituel et funéraire dans la zone d’étude conduit à proposer une origine locale couplée à une utilisation des pâtures d’altitude dès l’âge du Bronze. Elle révèle également l’existence d’un calendrier de subsistance favorisant un abattage hivernal des chevaux.
... Stable isotope analysis of metabolically inert consumer tissue (e.g., hair, nail) can provide a dietary and nutritional record or chronology that is integrated for the tissue growth period. This approach allows examination of seasonal changes in consumer diet [48][49][50][51][52][53][54][55][56]. The isotope composition of Alaskan moose hooves, for example, oscillated with distance from the hairline and was interpreted as a pattern that indicated seasonal diet shifts. ...
... Since trophic discrimination factors have not been determined for captively reared moose, we used the mean published diet-keratin (e.g., horn, hair) trophic discrimination factors for other large mammalian herbivores (cattle, African ungulates, bighorn sheep (Ovis canadensis), alpaca (Lama guanicoe), llama (L. glama), goat, horse; Δ 13 C = 3.0‰, Δ 15 N = 2.7‰) [56,[76][77][78][79]. However, several of the published estimates for Δ 13 C are based on a high-protein diet (19% crude protein; [76]). ...
Full-text available
Moose (Alces alces) are generalist herbivores, but are important aquatic-terrestrial ecotone specialists. Aquatic macrophytes are a high-quality food source for moose during summer, but the importance of aquatic food sources to the moose diet is difficult to study. We used stable isotope analysis of carbon and nitrogen from moose hooves and forage (terrestrial plants, aquatic macrophytes, and arboreal lichen) to assess the diet of moose at Isle Royale National Park, Michigan, USA, using Bayesian mixing models. We also evaluated the isotopic variability along chronologies of serially sampled hooves. Overall, our mixing models indicate that 13%-27% of the summer moose diet was aquatic in origin. Among moose that died during winter, body condition was impaired and hoof δ 15 N was higher where aquatic habitats were sparse. Although isotope chronologies preserved in hooves could significantly enhance our understanding of ungulate foraging ecology, interpretation of such chronologies is presently limited by our lack of knowledge pertaining to hoof growth rate and seasonal growth variability related to age and health. Distinct isotopic values among terrestrial plants, aquatic macrophytes, and arboreal lichens indicate that continued methodological advances in stable isotope ecology will lead to more precise estimates of the contribution of aquatic feeding to moose population dynamics and other ungulates.
... Stable isotope analysis of metabolically inert consumer tissue (e.g., hair, nail) can provide a dietary and nutritional record or chronology that is integrated for the tissue growth period. This approach allows examination of changes in consumer diet seasonally [48][49][50][51][52][53][54][55][56]. The isotope composition of Alaskan moose hooves, for example, oscillated with distance from the hairline and was interpreted as a pattern that indicated seasonal diet shifts. ...
... Since trophic discrimination factors have not been determined for captively reared moose, we used the mean published diet-keratin (e.g., horn, hair) trophic discrimination factors for other large mammalian herbivores (cattle, African ungulates, bighorn sheep [Ovis canadensis], alpaca [Lama guanicoe], llama [L. glama], goat, horse; Δ 13 C = 3.0‰, Δ 15 N = 2.7‰) [56,[74][75][76][77]. However, several of the published estimates for Δ 13 C are based on a high protein diet (19% crude protein; [74]). ...
Full-text available
Moose (Alces alces) are generalist herbivores but are important aquatic-terrestrial ecotone specialists. Aquatic macrophytes are a high-quality food source for moose during the summer, however their relative importance to moose diet is difficult to study. We used stable isotope analysis of carbon and nitrogen from moose hooves and forage (terrestrial plants, aquatic macrophytes, and arboreal lichen) to estimate the diet of moose at Isle Royale National Park, Michigan, USA, and to evaluate the isotopic variability along chronologies of serially sampled hooves. We hypothesized that aquatic macrophyte consumption and winter body condition (as measured by bone marrow fat content) would be greater at the eastern end of the island where aquatic habitats were most abundant. We were unable to evaluate spatial differences in aquatic macrophyte consumption, but overall, our mixing model results suggest that between 13% and 27% of summer moose diet was from aquatic sources. Among moose that died during winter, body condition was impaired and hoof δ15N (measured at the hairline) was higher at the western end of the island, where aquatic habitats are sparse. Although isotope chronologies preserved in hooves could significantly enhance our understanding of ungulate foraging ecology, interpretation of such chronologies is presently limited by our lack of knowledge pertaining to hoof growth rate and seasonal dynamics in relation to age and health. Significant isotope distinction among terrestrial plants, aquatic macrophytes, and arboreal lichens indicate that continued methodological advances in stable isotope ecology will lead to more precise estimates of the contribution of aquatic feeding to moose population dynamics.
... Several studies have previously estimated the growth rates of tail hairs for domestic cattle (Bos taurus Linnaeus, 1758) and horses (Equus ferus caballus Linnaeus, 1758) (Schwertl et al. 2003;Ayliffe et al. 2004;West et al. 2004;Zazzo et al. 2007;Auerswald et al. 2011). Lacking the resources and ability to replicate their efforts and considering the consistency of approximate growth rates across species, the growth rate of bison hairs was estimated to be 0.76 ± 0.03 mm/day (mean ± SD) from the previously published studies (Supplementary Table S1). 1 This was then extrapolated to days and months from death to determine the approximate timing of events in the isotope records from the bison hairs. ...
Full-text available
Assessing the challenges faced by wildlife populations is key to providing effective management but is problematic when dealing with populations in remote locations. Analyses of the stable carbon and nitrogen isotope composition (expressed as δ13C and δ15N values) of sequentially grown tissues, such as hairs, can be used to track changes in the eco-physiology of organisms. We generated δ13C and δ15N values from sequentially sampled (n = 465) hairs taken from wood bison (n = 27, Bison bison athabascae (Rhoads 1897)). Samples were taken from individuals prior to and after their release from captivity as part of a larger herd into the lower Innoko-Yukon river area of Alaska in 2015. Twenty months after release individuals had a distinct seasonal pattern in δ13C values. Hairs from individuals that experienced food scarcity or long-distance movement were sampled as case studies. Nutritional stress in these cases lead to a rise in δ15N values and a decrease in δ13C values. Applications of δ13C and δ15N analyses of bison tail hairs could provide wildlife managers a valuable and minimally invasive tool to better understand bison seasonal metabolic status and determine the historical health and behaviour of living and dead individuals.
... Improving the digestibility of keratin has been achieved through various methods (Falaye, 1982;Omitoyin, 1995;Coward-Kelly et al., 2006;AFRIS, 2012). Hoofs are soft keratin (Tomlinson et al., 2004) with growth rate 0.215mm/d (Zazzo et al., 2007) and are now been processed for human consumption in England (Walsh, 2014) and in Rwanda (Asaba, 2015). Hooves are part of inedible animal by-product discarded (Kiyanjui & Noor, 2013) while its exploitation in value addition is less considered (Kakkah et al., 2014;Alao et al., 2017). ...
Full-text available
Waste generation at slaughter from ruminant has led to environmental concerns. Processing slaughter house waste will reduce the problem of disposal and possible utilisation in livestock feed. Subjecting Cattle hoof meal to different processing methods can help in enhancing its nutritive value. Cattle hoof were obtained from the slaughter house; raw hoof was subjected to processing methods by boiling; chemical treatment with 10 % soda ash + boiling; fermentation treatment in water + boiling; 10 % wood ash treatment in water + boiling; autoclave treatment and samples analysed for proximate composition, amino acid profile and mineral content analysis according to standard methods. The hoof proximal analysis ranged 9.30 ± 0.06 % – 12.39 ± 0.01 % moisture content; 0.34 ± 0.01 % – 2.50 ± 0.12 % ash; 0.31 ± 0.01 % – 1.47 ± 0.02 % crude fat; 0.19 ± 0.02 % – 12.71 ± 0.15 % crude fibre and 85.27 ± 0.20 % – 90.74 ± 0.26 % crude protein in all samples. The amino acids profile of the hoof showed significant difference among treated samples. Tryptophan an essential amino acid was below detectable limit in all processed samples and raw hoof. The essential and non-essential minerals content showed significant difference (P < 0.05) among treatments with highest (Mg, Fe, K) in wood ash; (Cu) in raw hoof; (Ca, Na, P, Mn, Zn) in autoclaved samples. This study, suggest that Cattle hoof has the potential of being exploited as a source of animal protein for feed formulation in animal nutrition. This research concludes that the different processing methods affect the nutritive profiles of treated samples hence supplementation of limiting amino acids envisaged.
... Controlled feeding experiments have shown that it is possible to detect short-term (daily) changes in diet. 8,[10][11][12] Therefore, temporally resolved hair isotope records have been used to explore fine-scale spatial ecology of mammals. 7,13,14 Under specific conditions, keratin, the protein constitutive of hair, horn and hooves, can be preserved over several thousands of years and be used for palaeoenvironmental and palaeocological reconstructions. ...
Rationale Carbon and nitrogen stable isotope time series performed in continuously growing tissues (hair, tooth enamel) are commonly used to reconstruct the dietary history of modern and ancient animals. Predicting the effects of altitudinal mobility on animal δ¹³C and δ¹⁵N values remains difficult as several variables such as temperature, water availability or soil type can contribute to the isotope composition. Modern references adapted to the region of interest are therefore essential. Methods Between June 2015 and July 2018, six free‐ranging domestic horses living in the Mongolian Altaï were fitted with GPS collars. Tail hairs were sampled each year, prepared for sequential C and N isotope analysis using EA‐IRMS. Isotopic variations were compared with altitudinal mobility and Generalized Additive Mixed (GAMMs) models were used to model the effect of geographic and environmental factors on δ¹³C and δ¹⁵N values. Results Less than half of the pasture changes were linked with a significant isotopic shift while numerous isotopic shifts did not correspond to any altitudinal mobility. Similar patterns of δ¹³C and δ¹⁵N variations were observed between the different horses, despite differences in mobility patterns. We propose that water availability as well as seasonal availability of N2 fixing type plants primarily controlled horse hair δ¹³C and δ¹⁵N values, overprinting the influence of altitude. Conclusions Our study shows that altitudinal mobility is not the main factor that drives the variations in horse tail hair δ¹³C and δ¹⁵N values and that seasonal change in the animal dietary preference also plays an important role. It is therefore risky to interpret variations in δ¹³C and δ¹⁵N values of animal tissues in terms of altitudinal mobility alone, at least in C3–dominated environments.
... These isotopes provide semi-quantified information on the diet of the individuals from which the tissues are sampled (Schwarcz & Schoeninger, 1991), and can be used, for example, to differentiate between photosynthetic pathways in plant-based diets (DeNiro & Epstein, 1978), or used as indicators of trophic level (for example as between plant-and animal-based diets: Minagawa & Wada, 1984). As a result, these isotopes have served as the basis upon which studies of dietary composition, change, and niche partitioning have been undertaken in a wide range of taxa (e.g., Balasse, Bocherens, Mariotti, & Ambrose, 2001;Caraveo- Patino, Hobson, & Soto, 2007;Codron, Lee-Thorp, Sponheimer, de Ruiter, & Codron, 2006;Codron, Lee-Thorp, Sponheimer, De Ruiter, & Codron, 2008;Harrison et al., 2007;Koch et al., 1995;Lee, Schell, McDonald, & Richardson, 2005;Zazzo et al., 2007). Iwaniec, & Nash, 1998). ...
Objectives Food scarcity is proposed to be a limitation to chimpanzees at the limits of their range; however, such a constraint has never been investigated in this context. We investigated patterns of δ¹³C and δ¹⁵N variation along a latitudinal gradient at the northwestern West African chimpanzee (Pan troglodytes verus) range limit with the expectation that isotope ratios of chimpanzees at the range limit will indicate different dietary strategies or higher physiological constraints than chimpanzees further from the edge. Materials and methods We measured δ¹³C and δ¹⁵N values in hair (n = 81) and plant food (n = 342) samples from five chimpanzee communities located along a latitudinal gradient in Southeastern Senegal. Results We found clear grouping patterns in hair δ¹³C and δ¹⁵N in the four southern sites compared to the northernmost site. Environmental baseline samples collected from these sites revealed overall higher plant δ¹⁵N values at the northernmost site, but similar δ¹³C values across sites. By accounting for environmental baseline, Δ¹³C and Δ¹⁵N values were clustered for all five sites relative to total Pan variation, but indicated a ¹³C‐enriched diet at the range limit. Discussion Clustering in Δ¹³C and Δ¹⁵N values supports that strategic shifting between preferred and fallback foods is a likely ubiquitous but necessary strategy employed by these chimpanzees to cope with their environment, potentially allowing chimpanzees at their limits to avoid periods of starvation. These results also underline the necessity of accounting for local isotopic baseline differences during inter‐site comparison.
... Bone and hair samples were selected in order to compare different moments in the life of the individuals. Bone collagen should represent the diet consumed during the last years of life of the individual (Hedges et al. 2007), while hair should reflect the diet consumed during the last months before death, assuming hair grows at a rate of about 1-2 cm per month (Jones et al. 1981;O'Connell et al. 2001;Zazzo et al. 2007). ...
Full-text available
Pastoralism and camelid management have been essential to all aspects of pre-Hispanic Andean societies. Here, we present zooarchaeological and isotopic data on domestic camelid remains from Huaca Cao Viejo (El Brujo archaeological complex) on the northern coast of Peru, and dated to the Lambayeque/Sicán period—to characterise their biological age, diet, life history, possible geographic origin and ritual use. Domestic camelids, representing a wide range of biological ages and a high rate of polydactyly, were found as burial offerings in direct association with human funerary bundles (fardos). Direct AMS dates indicated that camelids were buried over a short period of time (AD 1022–1176) confirming the Lambayeque presence in the Chicama Valley during the first half of the Late Intermediate Period. Stable isotopic analyses were carried out on both bone collagen and hair keratin, including incremental analysis. A considerable variability in δ13C values at both the intra-individual and the intra-group level and a large contribution of C4 resources to diet are shown. This clearly supports local management and camelids originating from various herds. Zooarchaeological and isotopic evidences suggest diversity in herding practices and suggest the importance of the herds in fulfilling the transportation demands for trade in goods.
Full-text available
Stable isotopes are powerful tools for elucidating ecological trends in extant vertebrate communities, though their application to Mesozoic ecosystems is complicated by a lack of extant isotope data from comparable environments/ecosystems (e.g. coastal floodplain forest environments, lacking significant C4 plant components). We sampled 20 taxa across a broad phylogenetic, body size, and physiological scope from the Atchafalaya River Basin of Louisiana as an environmental analogue to the Late Cretaceous coastal floodplains of North America. Samples were analysed for stable carbon, oxygen and nitrogen isotope compositions from bioapatite and keratin tissues to test the degree of ecological resolution that can be determined in a system with similar environmental conditions, and using similar constraints, as those in many Mesozoic assemblages. Isotopic results suggest a broad overlap in resource use among taxa and considerable terrestrial–aquatic interchange, highlighting the challenges of ecological interpretation in C3 systems, particularly when lacking observational data for comparison. We also propose a modified oxygen isotope-temperature equation that uses mean endotherm and mean ectotherm isotope data to more precisely predict temperature when compared with measured Atchafalaya River water data. These results provide a critical isotopic baseline for coastal floodplain forests, and act as a framework for future studies of Mesozoic palaeoecology.
Full-text available
We monitored the change in the isotope composition of sulfur, carbon, and nitrogen in broad whitefish (Coregonus nasus) tissues in response to a change in the isotope composition of their food. One of two batches of 2.5-yr-old fish raised in the laboratory were given a new food source with different delta(34)S, delta(13)C, and delta(15)N, which were monitored in muscle and liver tissue for 1 yr. A model including change due to tissue accumulation (growth) and metabolic replacement was developed. For all three isotopes, most of the change could be attributed to growth. Metabolic replacement expressed as a turnover rate was only 0.1-0.2%.d(-1) and was similar for the three isotopes. Although liver tissue was -4.4 and -4.1 parts per thousand, respectively, for delta(34)S and delta(13)C relative to muscle tissue, the response over time to the new food was the same as for muscle. We expect that the complete change in the isotope composition of fish tissue in response to a change in food could take years in slow-growing wild populations. The stable isotope composition would represent a long-term average of the food. In fast-growing fish the rate of change would directly reflect the growth rate.
Food products, such as meat, from grassland production systems have ‘added value’ for both food producers and consumers because of the perceived nutritional value of such products and the environmental acceptability of the production system itself. To differentiate these products from competing products produced without, or with minimal, grass inputs, robust authentication methods are required. This review reports on the application of stable isotope, α-tocopherol stereoisomer and metabolite analyses to authenticate grass-fed beef. The potential for metabolically inert animal tissues to provide an archival record of changes to diet and of urine to yield information, non-invasively, about grass-fed beef production systems is discussed.
A pilot study was conducted to evaluate the suitability of stable isotope analysis for inferring the feeding histories of cattle fed known feeds. Stable isotope ratios of carbon and nitrogen (δ13C and δ15N) were measured in meat and hair from cattle and in their feeds at five farms in different regions of Japan, and the correlations of the isotope ratios between meat and hair were analysed. The results showed that δ13C values in feed depend on the photosynthesis type: C3 or C4. The values of δ15N in feeds varied widely, indicating divergent feeds made from plant materials that have different nitrogen origins, such as soil, chemical fertilizer, manure and air. In both cattle meat and hair, the farms differed significantly in the values of δ13C and δ15N. Both δ13C and δ15N were significantly higher in hair than in meat, and high correlations between meat and hair in both δ13C and δ15N were found. The results suggested that stable carbon and nitrogen isotope analysis for cattle meat and hair could be used to trace the feeding histories of cattle in Japan, and that hair samples would be used as an alternative to meat.
This study uses the carbon isotope values (δ13C) to determine how environmental deterioration is expressed in the δ13C values of vegetation and gazelles in the southern Levant. The ultimate goal is to use these modern data to predict the climatic impact of the Younger Dryas (YD). Climatic deterioration associated with the YD has been cited as the trigger for the transition to agriculture in the southern Levant. However, the evidence for the local severity of this climatic event is equivocal. There is disagreement over whether Mediterranean forest was succeeded by arid adapted steppic plant communities in what has been termed the Natufian ‘core area’. The modern data show a moderately negative regression slope between aridity and the δ13C values of both modern C3 plants and gazelle horn keratin within the Mediterranean phytogeographic belt. This pattern is expressed in both seasonal and annual datasets. The incorporation of a C4 plant component into gazelle diets is evident in the arid Mediterranean region, and is more pronounced in the dry season. The latter is apparent even despite interference caused by gazelle foraging on cultivated land. Based on the present day data, it is predicted that the succession of Mediterranean forest by open steppic vegetation would cause a positive shift of > 2‰ in the δ13C values of C3 plants and gazelles. The argument is based on the response of C3 vegetation to growth under increasing water stress conditions and the current distribution of C3 and C4 vegetation in relation to rainfall. This study presents a new tool with the potential to assess the climatic severity of the YD and its effect on Natufian foraging strategies.