Ancient DNA supports lineage replacement in European dog gene pool: insight into Neolithic southeast France
ABSTRACT We report palaeogenetic analysis of domesticated dog (Canis familiaris) remains excavated from three archaeological sites from southeast France and dating from Middle Neolithic. Ancient DNA analysis was attempted on teeth and bone samples taken from 11 dogs. Three 266-base-pair fragments of the mitochondrial genome Hypervariable Region I (HVR-I) could be retrieved and revealed two haplotypes belonging to HVR-I lineage C. These three sequences were compared to the sequences of Swedish and Italian Neolithic dogs and permitted to confirm that clade C was largely represented all over Western Europe during this period. One haplotype defined in Neolithic French dog was observed for the first time in Canis mtDNA, underlining the loss of mitochondrial diversity in Europe since the Neolithic. Finally, these results point out mitochondrial lineage replacement in Europe, since lineage C represents only 5% of extant European dogs. Altogether, these results support the proposition that palaeogenetic studies are essential for the reconstruction of the past demographic history and the domestication process of dogs.
- SourceAvailable from: princeton.edu[show abstract] [hide abstract]
ABSTRACT: We used molecular markers to study genetic relationships in a diverse collection of 85 domestic dog breeds. Differences among breeds accounted for approximately 30% of genetic variation. Microsatellite genotypes were used to correctly assign 99% of individual dogs to breeds. Phylogenetic analysis separated several breeds with ancient origins from the remaining breeds with modern European origins. We identified four genetic clusters, which predominantly contained breeds with similar geographic origin, morphology, or role in human activities. These results provide a genetic classification of dog breeds and will aid studies of the genetics of phenotypic breed differences.Science 06/2004; 304(5674):1160-4. · 31.20 Impact Factor
- [show abstract] [hide abstract]
ABSTRACT: The complete nucleotide sequence of the mitochondrial genome of the domestic dog, Canis familiaris, was determined. The length of the sequence was 16,728 bp; however, the length was not absolute due to the variation (heteroplasmy) caused by differing numbers of the repetitive motif, 5'-GTACACGT(A/G)C-3', in the control region. The genome organization, gene contents, and codon usage conformed to those of other mammalian mitochondrial genomes. Although its features were unknown, the "CTAGA" duplication event which followed the translational stop codon of the COII gene was not observed in other mammalian mitochondrial genomes. In order to determine the possible differences between mtDNAs in carnivores, two rRNA and 13 protein-coding genes from the cat, dog, and seal were compared. The combined molecular differences, in two rRNA genes as well as in the inferred amino acid sequences of the mitochondrial 13 protein-coding genes, suggested that there is a closer relationship between the dog and the seal than there is between either of these species and the cat. Based on the molecular differences of the mtDNA, the evolutionary divergence between the cat, the dog, and the seal was dated to approximately 50 +/- 4 million years ago. The degree of difference between carnivore mtDNAs varied according to the individual protein-coding gene applied, showing that the evolutionary relationships of distantly related species should be presented in an extended study based on ample sequence data like complete mtDNA molecules.Molecular Phylogenetics and Evolution 11/1998; 10(2):210-20. · 4.07 Impact Factor
- [show abstract] [hide abstract]
ABSTRACT: The spectacular diversity in size, conformation, and pelage that characterizes the domestic dog reflects not only the intensity of artificial selection but ultimately the genetic variability of founding populations. Here we review past molecular genetic data that are relevant to understanding the origin and phylogenetic relationships of the dog. DNA-DNA hybridization data show that the dog family Canidae diverged about 50 million years ago from other carnivore families. In contrast, the extant canids are very closely related and diverged from a common ancestor about 10 million years ago. The evidence supporting a close relationship of dogs with gray wolves is overwhelming. However, dogs are remarkably diverse in mitochondrial and nuclear genes. Mitochondrial DNA analysis suggests a more ancient origin of dogs than has been indicated by the fossil record. In addition, dogs have originated from or interbred with wolves throughout their history at different times and different places. We test the possibility of an independent domestication event in North America by analysis of mtDNA variation in the Xoloitzcuintli. This unusual breed is believed to have been kept isolated for thousands of years and may be one of the most ancient breeds in North America. Our results do not support a New World domestication of dogs nor a close association of the Xoloitzcuintli with other hair-less breeds of dogs. Despite their phenotypic uniformity, the Xoloitzcuintli has a surprisingly high level of mtDNA sequence variation. Other breeds are also genetically diverse, suggesting that dog breeds were often founded with a large number of dogs from outbred populations.Journal of Heredity 01/1999; 90(1):71-7. · 2.00 Impact Factor
Ancient DNA supports lineage replacement in European dog gene pool:
insight into Neolithic southeast France
M.F. Deguillouxa,*, J. Moquelb, M.H. Pemongea, G. Colombeaub
aUniversite ´ Bordeaux 1, UMR 5199 PACEA, Laboratoire d’Anthropologie des Populations du Passe ´, ba ˆt. B8, av. des Faculte ´s, 33405 Talence cedex, France
bUniversite ´ Bordeaux 1, UMR 5199 PACEA, Institut de Pre ´histoire et Ge ´ologie du Quaternaire, ba ˆt. B18, av. des Faculte ´s, 33405 Talence cedex, France
a r t i c l e i n f o
Received 28 July 2008
Received in revised form 1 October 2008
Accepted 2 October 2008
a b s t r a c t
We report palaeogenetic analysis of domesticated dog (Canis familiaris) remains excavated from three
archaeological sites from southeast France and dating from Middle Neolithic. Ancient DNA analysis was
attempted on teeth and bone samples taken from 11 dogs. Three 266-base-pair fragments of the
mitochondrial genome Hypervariable Region I (HVR-I) could be retrieved and revealed two haplotypes
belonging to HVR-I lineage C. These three sequences were compared to the sequences of Swedish and
Italian Neolithic dogs and permitted to confirm that clade C was largely represented all over Western
Europe during this period. One haplotype defined in Neolithic French dog was observed for the first time
in Canis mtDNA, underlining the loss of mitochondrial diversity in Europe since the Neolithic. Finally,
these results point out mitochondrial lineage replacement in Europe, since lineage C represents only 5%
of extant European dogs. Altogether, these results support the proposition that palaeogenetic studies are
essential for the reconstruction of the past demographic history and the domestication process of dogs.
? 2008 Elsevier Ltd. All rights reserved.
The domesticated dog Canis familiaris L. is descended from the
grey wolf Canis lupus L. (Clutton-Brock,1995). While this theory has
been well studied and established from genetics as well as
morphological data (Clutton-Brock, 1995; Vila ` et al., 1997), the
location, time and mode of dog domestication are still much
debated. For instance, the origin of domestic dogs from a single
wolf population or from multiple populations at different times
remains unclear. The first dog attested fossils have been found in
Spain (w17,00014C YA, Altuna et al., 1985; Vigne, 2005), Germany
(w14,00014C YA, Nobis,1979,1986) and Russia (13,000–17,00014C
YA, Sablin and Khlopachev, 2002) and consisted in robust wolf-
sized dogs. At the opposite, smaller dog remains were discovered in
Middle East, in Natoufian sites (w12,00014C YA, Tchernov and
Valla, 1997). These findings would favour independent domestica-
tion events. This would also indicate that founders were recruited
from a large and varied wolf population (Clutton-Brock, 1995),
explaining why living dogs present high phenotypic and genetic
variability. This extreme phenotypic diversity of dogs could also
imply occasional interbreeding with wild wolf populations,
following founding events (Vila ` et al., 1997).
Archaeological evidence suggests an origin of the domestic dog
during the last glaciations. Recent genetic studies focusing on the
mitochondrial genome Hypervariable Region 1 (HVR-1) of extant
dog have attempted to test this hypothesis. They point towards the
existence of several dog clades (Tsuda et al., 1997; Vila ` et al., 1997;
Leonard et al., 2002; Savolainen et al., 2002), which are interpreted
as indicative of multiple and independent origins from multiple
wolf matrilines. The molecular time estimates of the dog–wolf
divergence vary from 76,000–130,000 YA for Vila ` et al. (1997) to
15,000–40,000 YA for Savolainen et al. (2002). The analysis of intra-
clade genetic diversity and the distribution of unique haplotypes
led Savolainen et al. (2002) to favour an East Asian origin for the
three main dog clades, defined as haplogroups (Hg) A, B and C, all
emerging about 15,000 YA.
Humans have significantly influenced the genetics of wolves
and dogs during the last centuries. Wolves were highly persecuted,
leading to the extinction or at least the bottlenecking of most
populations (Vila ` et al., 1999; Randi et al., 2000; Flagstad et al.,
2003). Dogs breeding and distribution were highly influenced by
human, and the elaboration of the main dog breeds over the last
century had important consequences for the make-up of the dog
gene pool. Genetic studies have suggested that most of the current
breeds may represent a recent radiation from a common stock
(Parker et al., 2004). Moreover, the combined analyses of paternally
inherited Y chromosome markers with maternally inherited
mitochondrial DNA and biparentally inherited autosomal micro-
satellite markers in domestic dogs have demonstrated a sex bias in
* Corresponding author. Tel.: þ33 (0)5 40 00 37 38; fax: þ33 (0)5 40 00 25 45.
E-mail address: email@example.com (M.F. Deguilloux).
Contents lists available at ScienceDirect
Journal of Archaeological Science
journal homepage: http://www.elsevier.com/locate/jas
0305-4403/$ – see front matter ? 2008 Elsevier Ltd. All rights reserved.
Journal of Archaeological Science 36 (2009) 513–519
the origin of breeds, with fewer males than females contributing
genetically (Sundqvist et al., 2006). Therefore, these historic human
influences must have strongly changed genetic signatures. Only
studies based on ancient DNA (aDNA) can give direct insight into
the composition of the ancient dog and wolf gene pool, before the
confounding factor of artificial selection and persecution.
This palaeogenetic approach has already been successfully
applied to canid remains. In 2005, Verginelli et al. succeeded in the
genetic analysis of five ancient Italian canid remains (ages ranging
from 15,000 to 3,000 YA) and found that mitochondrial sequences
were highly diverse. Indeed, the five ancient mitochondrial
sequences encompassed the three major clades of extant dog
sequences (i.e. clades A/B/C and I/II/IV according to Savolainen
et al., 2002 and Verginelli et al., 2005, respectively). Phylogenetic
analysis pointed out relationships between the ancient Italian
sequences and geographically widespread extant dog matrilines
and extant wolf matrilines of mainly East European origin. The
authors concluded that European wolves played a role in the origin
of the three major clades and that multiple independent domesti-
cation events took place. This conclusion contrasts with the more
simplistic scenario proposing an East Asian origin for the three
major clades (Savolainen et al., 2002). More recently, Malmstro ¨m
et al. (2005, 2008) carried out a palaeogenetic analysis on Neolithic
and Medieval dog remains from Scandinavia. Surprisingly, they
identified sequences corresponding to the Hg A and C which are
supposed to originate from Asia (Savolainen et al., 2002), instead of
lineage D, that is restricted to Europe and consequently presented
as the potential result of central or northern Europe domestication
event. This led to the conclusion that modern data are essential to
resolve the number and the antiquity of domestication events,
however, they are inadequate to uncover their geographic origin.
Altogether, both studies highlighted the need to use aDNA to
build a better understanding of past canid population genetics.
Specifically, aDNA analyses give a direct measurement of the
impact of human activities on the haplogroup frequencies among
domesticates. The present study contributes to this promising
approach, by describing novel aDNA sequences obtained from
Middle Neolithic dogs originating from a site in southeast of France
and belonging to the Chassean culture. The three French Neolithic
HVR-I sequences obtained were compared with sequences of 541
modern dogs (from Verginelli et al., 2005) and 22 ancient canids
(Verginelli et al., 2005; Malmstro ¨m et al., 2008) sequences to
provide novel insight into the ancient gene pool of European dogs.
2. Materials and methods
2.1. Archaeological material
Dog specimens were provided by the Institut National de
Recherches Arche ´ologiques Pre ´ventives (Table 1). Remains were
collected from three Chassean culture sites in the southeast of
France: Berriac (Vaquer, 1998), Villeneuve-Tolosane (Vaquer et al.,
1980) and Le Cre `s (Loison et al., 2004). For reason of biomolecule
preservation, teeth are the sample of choice for palaeogenetic
studies (thought to be the place of better DNA conservation and
protection against contaminations, Ricaud et al., 2004). As far as
possible our analyses were conducted on this type of material.
Remains from the site of Les Plots - Berriac consisted in teeth
and bones recovered from dogs deposited in two different features
(Table 1). The first structure contained the nearlycomplete remains
of two adults and one young dog and was interpreted as the
voluntary and simultaneous deposit of the three individuals. The
second structure contained one complete dog skeleton. The pres-
ence of ash and burn marks on the skull, as well as the lack of
cervical vertebrae, both tend to exclude the hypothesis of a funeral
deposit. Teeth or tibias were collected from all four individuals of
Berriac and were submitted to genetic analysis (Table 1).
The sample of ancient dog remains from Villeneuve-Tolosane
consisted in numerous teeth found in a pit 1.5 m wide and 7.5 m
deep. This feature was interpreted as a rubbish pit and was full of
sediment, ash and faunal remains. Teeth were collected from each
0.5 m level, in order to minimize the risk that they belong to the
same individual. Five teeth corresponding to five different levels
were analysed (Table 1).
The last archaeological site, Le Cre `s of Be ´zier, consisted of a large
burial site with more than 200 structures encompassing 44 pits (33
of which contain human burials). A total of 7 dog individuals was
found in four large circular pits, four of them presenting clear
Samples analysed in this study, their excavated location, age, and amplification success (7/9¼7 positive amplifications over 9 tested).
Sample nameCodeArchaeological site Individual informationSample analysedAgeAmplification success
HVR-I BHVR-I D
Berriac dog A1BerA1 Berriac Adult 1 of sepulchre AInferior canine (right)4200–4000
Berriac dog A2 BerA2 BerriacAdult 2 of sepulchre ADistal epiphysis tibia (right) 0/60/6
Berriac dog A3BerA3 Berriac Young individual of sepulchre AFirst superior premolar (right) 0/60/6
Berriac dog BBerBBerriac Individual 1 of deposit BTibia (left)0/6 0/6
Le Cre `s dog 1
LC1Le Cre `s
Tooth in pit (2.5–3 m depth) Superior canine (right)7/9 6/10
Tooth in pit (3.5–4 m depth) Second superior premolar (right) 4000–3800
Tooth in pit (4.5–5 m depth) Deciduous superior premolar
Second inferior premolar (right) 4000–3800
Tooth in pit (1.5–2 m depth)
Tooth in pit (6–6.5 m depth) Inferior canine (right)0/80/8
Adult individual associated with human
Adult in dog dual sepulchre
Inferior canine (left) 0/80/8
Le Cre `s dog 2 LC2Le Cre `sFirst superior molar (right) 0/80/8
M.F. Deguilloux et al. / Journal of Archaeological Science 36 (2009) 513–519 514
evidence of voluntary inhumation. Only two dogs were submitted
to genetic analysis: LC1 that was found associated with human
remains and LC2 recovered from a dog dual sepulchre. Teeth were
processed for DNA isolation (Table 1).
High conservation variability was observed between the three
archaeological sites (Fig. 1). Based on macroscopic observation,
bones remains from Villeneuve-Tolosane were histologically well
preserved, as they were covered in protective ash levels (Vaquer
et al., 1980), although they were also severely mechanically frag-
mented. In contrast, the remains from Le Cre `s showed very faded
osseous surfaces due to the corrosive action of the sediment and
were also fractured post-mortem. Lastly, the bones from Berriac
were also fractured and showed highly damaged surfaces, an
alteration that is due to the chemical action of plant roots.
2.2. Ancient DNA authentication
Many studies have demonstrated the risks of contamination
with external DNA when dealing with ancient material (Gilbert
et al., 2005; Malmstro ¨m et al., 2005). As a consequence, criteria to
support the authenticity of aDNA results have been proposed by
palaeogeneticists (Richards et al., 1995; Poinar, 2003; Gilbert et al.,
2005), the most important being (i) the use of a dedicated labora-
tory, isolated from the laboratory where fresh samples are pro-
cessed, (ii) the reproducibility of the results, (iii) the fact that the
results ‘‘make sense’’ genetically (e.g. phylogenetically), and (iv) the
cloning and sequencing of the amplification products to detect PCR
artefacts associated with post-mortem template modification and/
or contaminations (Poinar, 2003). All analyses performed in our
study followed those basic authenticity criteria.
All analyses of dog remains were performed at the laboratory of
Past Human Populations (UMR PACEA, University of Bordeaux 1,
Bordeaux, France) in a laboratory specifically dedicated to ancient
DNA analysis. All DNA isolation and PCR experiments were carried
out under sterile conditions in separate dedicated rooms, and the
laboratory equipment and reagents used were DNA-free. Prepara-
tion of samples and extraction of DNA were both carried out in
separate laboratories in which dog DNA had not been previously
isolated or analysed. Prior to isolation, teeth were scraped with
sterile blades, superficially washed with bleach and stored in sterile
bags for further analysis, all performed in a sterile laminar flow
hood with positive air pressure, UV irradiation and bleach cleaning.
2.3. Ancient DNA extraction, amplification and sequencing
DNA was isolated from teeth and bones of eleven dogs (Table 1),
originating from the three French archaeological sites. Each tooth
or bone was first reduced to powder. After an overnight decalcifi-
cation and protein digestion at 55?C with agitation (EDTA 0.5 M,
pH¼8.5, proteinase K 1–2 mg/ml, N-lauryl sarcosyl 0.5%), the
supernatants were extracted using phenol–chloroform–isoamyl
alcohol (25:24:1) organic extraction. Subsequently, the aqueous
phase was concentrated in 100 ml of sterile distilled water with
Centricon-30 columns (Amicon?) according to the manufacturer’s
Fig. 1. Illustration of samples conservation variability between dog remains originating from Berriac (A), Villeneuve-Tolosane (B) and Le Cre `s (C).
M.F. Deguilloux et al. / Journal of Archaeological Science 36 (2009) 513–519 515
instructions. An extraction blank was systematically co-extracted
with the ancient dog samples during each extraction session, to
trace potential cross-contamination between samples. No more
than two ancient samples were co-extracted at the same time.
PCR amplifications were performed on the Hypervariable
Region 1 (HVR-I) of the mitochondrial genome. HVR-I region of the
mtDNA control region (position 15,426–15,692, including primers)
was obtained through two overlapping fragments obtained with
a set of two primers’ pairs (Verginelli et al., 2005): fragment HVR-I
B obtained with primers L15426int and H15555int, fragment HVR-I
D with L15529int and H15692int.
The pre-PCR mix was prepared in a DNA-free room in the
ancient DNA laboratory. PCR amplifications were performed in
a total reaction volume of 25 ml containing 6.5 mM MgCl2, 0.4 mM
dNTPs, 0.66 mg/ml BSA, 1 mM of each primer, 1?GeneAmp PCR
Buffer (Perkin–Elmer), 0.25 ml pure DNA extract and 1.25 U
AmpliTaq Gold? (Applied Biosystems). PCR was run for 55 cycles at
94?C for 45 s, 55?C for 45 s, and 72?C for 45 s. Three independent
blanks were carried out for each set of PCR experiment.
Sequences of each mitochondrial fragment were confirmed by
at least two separate amplifications on each DNA extract. As the
products of ancient DNA amplification generally contain a large
number of mutations generated by induced damage and Taq
polymerase errors (Handt et al., 1996), the PCR products were
subcloned into bacterial vectors using the Topo TA cloning kit
Screening of colonies was accomplished by PCR, transferring the
colonies into a 40 ml reaction mix made of 67 mM Tris–HCl, 2 mM
MgCl2, 1 mM of each M13 forward and reverse universal primer,
0.125 mM of each dNTP and 0.75 U of Taq polymerase (Perkin–
Elmer). PCR fragments were separated by agarose gel electropho-
resis and clones with the expected insert size were selected for
sequencing (Genome Express). Sequences from multiple clones of
the same PCR product were aligned in order to resolve HVR-I
ambiguities. Artefacts caused by the amplification of damaged
templates from a low copy number were carefully and systemati-
cally eliminated. The authentic sequences were always deduced
from the consensus of at least 10 clones, corresponding to at least
two different amplification products (except for VTC3 - HVR-I B
fragment). Extractions and PCR amplification blanks were used as
negative controls, in order to detect possible contamination by
2.4. Phylogenetic analysis
All sequences were aligned using the MEGA3.1 software (Kumar
et al., 2004) and all detected sequence variations were verified by
visual assessment in the electrophoregrams. The HVR-I sequences
obtained from ancient French dogs were compared with mtDNA
sequences from modern dog, using the data compiled by Verginelli
et al. (2005). The database consists of 541 sequences of extant dogs
from all continents, including 248, 232, 31 and 30 dogs from Asia,
Europe, Africa and America, respectively (Okumura et al.,1996; Vila `
et al., 1997; Savolainen et al., 2002, 2004). Five mitochondrial
sequences from ancient Italian canids (Verginelli et al., 2005) were
included, as well as 17 sequences from ancient Scandinavian dogs
(Malmstro ¨m et al., 2008). A 230 bp region encompassing position
15,458–16,687 was chosen as it was available from all data sources.
The sequence NC_002008, corresponding to the first complete
mtDNA dog sequence published (Kim et al., 1998) was used as
a reference for the numbering of polymorphisms, as proposed by
Pereira et al. (2004).
DNA variants were identified in the 230 bp sequence using
DnaSP v.3 software (http://www.ub.es/dnasp/) and Median-
Joining (MJ) networks connecting the ancient and extant dogs
sequences were constructed using the NETWORK 4.201 software
(www.fluxus-engineering.com) and the median-joining algorithm
(using default weights).
We attempted to isolate DNA from eleven dog remains from
French archaeological sites belonging to the Neolithic Chassean
culture (from 4,200 to 3,800 BC). Two separate mtDNA segments
permitting the identification of partial HVR-I sequence (266 bp)
were successfully amplified and sequenced for three of the 11
ancient samples tested, originating from the site of Villeneuve-
Tolosane and dated from 4000 to 3800 BC (samples VTC 1, 2 and 3,
Table 1). Owing to poor preservation of biomolecules, no PCR
product was obtained from the remaining samples, despite several
attempts. With regards to the samples originating from the Ville-
neuve-Tolosane site, no correlation between depth of samples and
DNA conservation was found in our analysis, since teeth found in
both greater and lesser depth yielded no results. This reinforces the
idea that different soil microenvironments inside archaeological
site lead to variable DNA conservation.
All Villeneuve-Tolosane sequences were deduced from multiple
clone alignment, from 10 to 25 clones for each mitochondrial
fragment. Sequence ambiguities were resolved by analysis of
multiple clones and deaminations observed in several clones were
systematically eliminated. All ancient sequences retrieved belong
to dog clade C (Fig. 2). Similar sequences were obtained for VTC2
and VTC3 and present the Hg C characteristic mutations: 15508T,
15526T, 15611C, 15639G and 15650C (Savolainen et al., 2002; Per-
eira et al., 2004). This sequence clusters to the central node of the
lineage, as shown in the MJ network of clade C (Fig. 3). VTC2 and
VTC3 share their HVR-I sequence with numerous individuals from
all continents except America (i.e. 64%, 33% and 3% of Asian,
European and African dogs, respectively). This sequence is similar
to one Neolithic Italian dog aged of 4,100 BP (Casal Del Dolce, PIC4,
Verginelli et al., 2005), and with nine Neolithic dogs from Scandi-
navia (Malmstro ¨m et al., 2008). Morphological observation
suggests that VTC2 and VTC3 remains correspond to two distinct
individuals. Indeed, VTC2 premolar shows usury indices corre-
sponding to a young individual (at least 1 year old), and VTC3
deciduous premolar is cutting suggesting it belonged to a few
months old individual. Nevertheless, we cannot exclude that those
dogs were related, a fact that could introduce bias in genetic
The sequence obtained for VTC1 presents the same poly-
morphisms as VTC2 and VTC3, as well as two additional mutations
at positions 15508 and 15625. The mutation at position 15508 is
a feature shared with one ancient Italian wolf (PIC1, aged 15,000
BP) from the Late Glacial levels of the Palidoro Upper Palaeolithic
rock shelter (Verginelli et al., 2005). As sample VTC1 presents an
additional mutation at position 15625, the sequence appears novel
and occupies a derived position.
The star-shaped topology of clade C network suggests an origin
from a single haplotype (Fig. 3). Briefly, clade C encompasses dog
haplotypes from Eurasia, America and Africa, as well as numerous
ancient European dog sequences: one sequence from an ancient
Italian dog dated of 15,000 BP, and a total of 15 sequences from
European Neolithic dogs (11 from Scandinavia, 3 from France and 1
from Italy). Interestingly, Asian sequences are represented by only
two haplotypes, whereas six different haplotypes are encountered
in modern and ancient European samples.
A total of 95.4% of dog sequences compiled in our worldwide
dataset belong to haplogroups A, B and C (with 67.8% and 10%
belonging to Hg A and Hg C, respectively). Despite the low reso-
lution obtained with our short mitochondrial sequence (230 bp)
the MJ network constructed for all clades shows the same pattern
already described by Savolainen et al. (2002), with Asian
M.F. Deguilloux et al. / Journal of Archaeological Science 36 (2009) 513–519 516
haplotypes distributed throughout the entire network, while
European haplotypes are absent from parts of the network (Fig. 2).
This suggested to Savolainen et al. (2002), in addition of genetic
diversity analysis, that haplotypes of Western part of the world
originate from the introduction of a subset of Asian types. The four
major Hg A, B, C, D represent 70%, 16%, 6% and 8% of the European
modern dataset, respectively. The comparison of the sequences
obtained from Swedish, French and Italian Neolithic dogs and the
contemporary European breeds indicates a significant deviation
between the modern and Neolithic Hg frequencies (p<0.00037,
c2¼12.67, df¼1). A significant reduction of Hg A and increase in
Hg C in Neolithic sample (12% and 88% respectively) is clearly
demonstrated. Using a binomial distribution, we estimated that the
frequency of Hg A must have been lower than 33.1% during the
Neolithic to have notice it only twice in 17 samples (p<0.05)
assuming that the relative frequency was the same as in modern
European dog gene pool (Malmstro ¨m et al., 2008).
Mitochondrial HVR-I sequences were obtained from three
samples originating from the Chassean French site of Villeneuve-
Tolosane. Due to poor preservation of the French Neolithic remains
analysed, only few dog samples yielded positive PCR amplification.
However, amplification failures were expected for the remains
taken from Berriac and Le Cre `s sites because of the observed
macroscopic degradation. Indeed, remains recovered from the
Berriac site presented very important traces of root chemical action
on the enamel. This chemical degradation may have impeded
satisfying fossilisation of the enamel and may have favoured the
degradation of biomolecules. Concerning the dog fossils from Le
Cre `s, they presented traces of corrosive action of the sediment. The
impossibility to obtain DNA from remains originating from both
sites gives support to previous studies proposing a correlation
between macroscopic aspect of archaeological remains and DNA
conservation (Cipollaro et al., 1998).
Strong evidences allow us to authenticate the sequences
obtained from the Villeneuve-Tolosane site: (i) dog DNA had never
been studied in the aDNA lab before this study; (ii) the sequences
were confirmed on at least two PCR amplifications, and were
always deduced from the consensus between amplification prod-
ucts of at least 10 clones; (iii) artefacts caused by the amplification
of damaged templates, typical of ancient DNA, were detected in
some clones and eliminated; (iv) the haplotypes obtained make
sense phylogenetically. Moreover, one of the sequences (VTC1) had
never been characterised in modern dog breeds. In conjunction,
these elements lend confidence to the authenticity of the
sequences presented in this paper, despite the isolated teeth orig-
inating from the Villeneuve-Tolosane site did not permit any
isolation replication in the same or in another lab.
The major present-day dog populations have been presented
having a common origin from a single gene pool containing clades
A, B and C (Savolainen et al., 2002). Data obtained on ancient
European canids confirm this point (Verginelli et al., 2005; Malm-
stro ¨m et al., 2008 and present study), as the diversity found in
ancient European canids proves that those three lineages were
formerly present in Western Europe, since at least 15,000 BP for
clade C and 10,000 BP for clades A and B (Verginelli et al., 2005).
PIC (Verginelli et al. 2006)
Medieval (Malmstrom et al. 2008)
Neolithic (Malmstrom et al. 2008)
VTC (our study)
Fig. 2. Median-Joining network of 566 extant and ancient dog mitochondrial haplotypes, based on 230 bp sequence of HVR-I region (nps15,458–15,687). Circle areas are
proportional to haplotypes frequencies.
M.F. Deguilloux et al. / Journal of Archaeological Science 36 (2009) 513–519 517
However, all studies dealing with ancient European canids (Vergi-
nelli et al., 2005; Malmstro ¨m et al., 2008), including ours, clearly
demonstrate the loss of genetic diversity since Late Pleistocene and
Early Holocene. Caution is then necessary when making historical
inferences from modern population data, which has already been
noted for other Upper Pleistocene carnivores like brown bears and
hyenas (Leonard et al., 2000; Hofreiter et al., 2004).
More precisely, the data obtained on French Neolithic dogs,
combined with the analysis of Swedish (Malmstro ¨m et al., 2008)
and one Italian (Verginelli et al., 2005) Neolithic dogs, confirm that
Hg C was well represented all overthewesternpartof the European
continent during this period. As Hg A encompasses the majority of
extant European dogs (70% for Hg A, 6% for Hg C), we speculate that
Hg C has been replaced through time by Hg A in Europe. Malm-
stro ¨m et al. (2008) also clearly identified a lineage replacement in
Scandinavia (with a clear emergence of Hg D in modern dog gene
pool). As already mentioned by the authors, numerous factors can
induce Hg frequencies change, such as selection, migration and
demographic changes. Our results tend to support that such lineage
replacement has occurred over the entire continent, as all European
dogs appear to have experienced such alteration since Neolithic
period. More recent massive importation of Hg A in Europe and Hg
D in Scandinavia could be responsible for such a change. More
recently, the stringent selection of European breeds, following the
formalization of the breed concept in the 19th century, with all dog
breeds deriving from shared stock, could have contributed to this
lineage replacement. Together, those results contradict the current
hypothesis that phylogeographic patterns among domesticates
were established during the initial domestication period. Clearly, in
the case of European dogs, human activities have drasticallyaltered
the Hg frequencies since the Neolithic.
The topology of the MJ networks for clades A, B and C suggested
that haplotypes of Western part of the world originate from the
introduction of a subset of Asian types (Savolainen et al., 2002).
Because of lower haplotype diversity, this Asian origin was less
clear concerning Hg C. However, within this clade, West haplotypes
had only types shared with the East, whereas the East presented
unique haplotypes. This point led the authors to suggest an East
Asian origin for this lineage (Savolainen et al., 2002). With the
addition of ancient Italian, Swedish and French canids to the
network of clade C, the topology obtained tends to diminish this
hypothesis, since four European ancient canids occupy derived
positions and present unique haplotypes. Similar pattern of genetic
diversity have been used to argue for local domestication in other
species (Fernandez et al., 2007; Beja-Pereira et al., 2006), and could
be used in the case of dogs to propose a European origin for Hg C.
However, the combination of ancient European sequences with
modern Asian sequences introduces bias in haplotypic diversity,
comparing incomparable data. This striking point currently
remains unresolved, but the continuous enlargement of ancient
dog mitochondrial database holds promise for a solution. The
analysis of ancient Asian dogs is a necessary follow up of the
current study for example. In any case, the newly described ancient
genetic diversity in western European dogs highlights that caution
must be taken when making inferences about population history
from modern data only. This point is in total accordance with the
conclusion of Malmstro ¨m et al. (2008) who point out that modern
sequences of domestic animals could be unsuitable to resolve the
geographic origin of domestication events.
The ancient sequences indicate that western European dogs,
from Italy to France and Scandinavia, may have belonged to the
same genetic pool during Neolithic. Although this ancient sample is
PIC (Verginelli et al. 2006)
Medieval (Malmstrom et al. 2008)
Neolithic (Malmstrom et al. 2008)
VTC (our study)
Fig. 3. Median-Joining network of dog mitochondrial clade C, based on 230 bp sequence of HVR-I region (nps15,458–15,687). Circle areas are proportional to haplotypes
frequencies. The HVR-I mutations at given nucleotide positions compared with the reference sequence of Kim et al. (1998) are shown.
M.F. Deguilloux et al. / Journal of Archaeological Science 36 (2009) 513–519518
not representative of ancient European population, it is remarkable
to encounter the same haplogroup, and even the same sequences,
during the Neolithic period, in these distant regions. This could be
accountedby distant domestic
Neolithic human populations and/or shared breeding practices
since Neolithic times. Numerous French sites from the Chassean
culture show strong evidence of special attention paid to dog
burials (Arbogast et al., 1989) and the voluntary burial of dogs
appears like a common practice of those populations. This care is
often interpreted as special relationship between Neolithic human
populations and domesticated dogs. Since this human–dog rela-
tionship appears important during the Neolithic, more compre-
hensive palaeogenetic analysis of ancient dog remains belonging to
this period could demonstrate, on the one hand, the extensive
connection between dog populations and, on the other, the extant
links and exchanges between Neolithic human populations.
In conclusion, the sequences obtained on ancient European dogs
highlight the relevance of palaeogenetic data to: (i) shade light on
the loss of genetic diversity during the domestication process, (ii)
reveal lineage(s) replacement(s) through time that can be caused
by various factors such as selection, migration and extinction, and
(iii) resolve more accurately the origin of the domestication events.
The accumulation of genetic data on ancient European dog remains
begins to bring new elements on dog domestication. The obser-
vation of shared Hg C sequences between Italian, French and
Swedish Neolithic dogs permits to speculate on a mitochondrial
lineage replacement since Neolithic times, and suggests shared
breeding practices since this period. We believe that further
palaeogenetic analysis of European and Asian dog remains will give
critical insight into the domestication process of Canis.
The sequences reported in this paper have been deposited into
GenBank (accession numbers EU287460, EU287461, and EU287462).
The authors would like to acknowledge H. Duday (UMR PACEA)
for providing contacts with archaeologists and zooarchaeologists
and for proofreading the manuscript. We are gratefultoJ. Vaquer, G.
Loison and V. Forrest who supplied dog samples from Berriac,
Villeneuve-Tolosane and Le Cre `s-Be ´ziers and for informations on
archaeological sites, and to D. Chagne ´ forhis help in the preparation
of the manuscript. Funding for this study was provided by the
University of Bordeaux 1 (BQR).
Altuna, J., Baldeon, A., Mariezkurrena, K., 1985. Cazadores magdalenienses en la
Cueva de Errala (Cestona, Pais Vasco). Munibe. Antropologia-Arkeologia 37.
Arbogast, R.M., Blouet, V., Desloges, J., Guillaume, C., 1989. Le cerf et le chien dans
les pratiques fune ´raires de la seconde moitie ´ du Ne ´olithique du Nord de la
France. Anthropozoologica 3, 37–42.
Beja-Pereira, A., Caramelli, D., Laluza-Fox, C., Vernesi, C., Ferrand, N., Casoli, A.,
Goyache, F., Royo, L.J., Conti, S., Lari, M., Martini, A., Ouragh, L., Magid, A.,
Atash, A., Zsolnai, A., Boscato, P., Triantaphylidis, C., Ploumi, K., Sineo, L.,
Mallegni, F., Taberlet, P., Erhardt, G., Sampietro, L., Bertranpetit, J., Barbujani, G.,
Luikart, G., Bertorelle, G., 2006. The origin of European cattle: evidence from
modern and ancient DNA. Proceedings of the National Academy of Sciences 103,
Cipollaro, N., Di Bernardo, G., Galano, G., Galderisi, U., Guarino, F., Angelini, F.,
Cascino, A., 1998. Ancient DNA in human bone remains from Pompeii archaeo-
Clutton-Brock, J., 1995. Origins of the dog: domestication and early history. In:
Serpell, J. (Ed.), The Domestic Dog: Its Evolution, Behaviour and Interaction with
People. Cambridge University Press, Cambridge, pp. 8–20.
Fernandez, H., Hughes, S., Vigne, J.D., Helmer, D., Hodgins, G., Miquel, C., Ha ¨nni, C.,
Luikart, G., Taberlet, P., 2007. Divergent mtDNA lineages of goats in an early
Neolithic site, far from the initial domestication areas. Proceedings of the
National Academy of Sciences 103, 15375–15379.
Flagstad, Ø., Walker, C.W., Vila `, C., Sundqvist, A.K., Fernholm, B., Hufthammer, A.K.,
Wiig, Ø., Koyola, I., Ellegren, H., 2003. Two centuries of the Scandinavian wolf
population: patterns of genetic variability and migration during an era of
dramatic decline. Molecular Ecology 12, 869–880.
Gilbert, M.T.P., Bandelt, H.J., Hofreiter, M., Barnes, I., 2005. Assessing ancient DNA
studies. Trends in Ecology and Evolution 20, 541–544.
Handt, O., Krings, M., Ward, R.H., Pa ¨a ¨bo, S., 1996. The retrieval of ancient human
DNA sequences. American Journal of Human Genetics 59, 368–376.
Hofreiter, M., Serre, D., Rohland, N., Rabeder, G., Nagel, D., Conard, N., Mu ¨nzel, S.,
Pa ¨a ¨bo, S., 2004. Lack of phylogeography in European mammals before the last
glaciation. Proceedings of the National Academy of Sciences USA 101, 12963–
Kim, K.S., Lee, S.E., Jeong, H.W., Ha, J.H., 1998. The complete nucleotide sequence of
the domestic dog (Canis familiaris) mitochondrial genome. Molecular Phylo-
genetics and Evolution 10 (2), 210–220.
Kumar, S., Tamura, K., Nei, M., 2004. MEGA3: integrated software for molecular
evolutionary genetics analysis and sequence alignment. Briefings in bio-
informatics 5, 150–163.
Leonard, J.A., Wayne, R.K., Cooper, A., 2000. Population genetics of Ice Age brown
bears. Proceedings of the National Academy of Sciences USA 97 (4), 1651–1654.
Leonard, J.A., Wayne, R.K., Wheeler, J., Valadez, R., Guille ´n, S., Vila `, C., 2002. Ancient
DNA evidence for old world origin of new world dogs. Science 298, 1613–1616.
Loison, G., Fabre, V., Villemeur, I., 2004. Le Cre `s. Habitats pre ´historiques en bordure
de l’Orb. Structures domestiques et se ´pultures du Chasse ´en ancien. Rapport
final d’ope ´ration. INRAP, Montpellier, p. 304.
Malmstro ¨m, H., Stora, J., Dale ´n, L., Holmlund, G., Go ¨therstro ¨m, A., 2005. Extensive
human DNA contamination in extracts from ancient dog bones and teeth.
Molecular Biology and Evolution 22 (10), 2040–2047.
Malmstro ¨m, H., Vila `, C., Gilbert, M.T.P., Stora, J., Willerslev, E., Holmlund, G.,
Go ¨therstro ¨m, A., 2008. Barking up the wrong tree: modern northern European
dogs fail to explain their origin. BMC Evolutionary Biology 8 (71).
Nobis, G., 1979. Der a ¨lteste Haushund lebte vor 14,000 Jahren. Umschau 19, 610.
Nobis, G.,1986. Die Wildsa ¨ugetiere in der Umwelt des Menschen von Oberkassel bei
Bonn und das Domestikationsproblem bei Wo ¨lfen im Jungpala ¨olithikum. Bon-
ner Jahrbu ¨cher 186, 367–376.
Okumura, N., Ishiguro, N., Nakano, M., Matsui, A., Sahara, M., 1996. Intra- and
interbreed genetic variations of mitochondrial DNA major non-coding regions in
Japanese native dog breeds (Canis familiaris). Animal Genetics 27 (6), 397–405.
Parker, H.G., Kim, L.V., Sutter, N.B., Carlson, S., Lorentzen, T.D., 2004. Genetic
structure of the purebred domestic dog. Science 304, 1160–1166.
Pereira, L., Van Asch, B., Amorim, A., 2004. Standardisation of nomenclature for dog
mtDNA D-loop: a prerequisite for launching a Canis familiaris database. Forensic
Science International 141, 99–108.
Poinar, H.N., 2003. The top 10 list: criteria of authenticity for DNA from ancient and
forensic samples. International Congress Series 1239, 575–579.
Randi, E.V., Lucchini, V., Christensen, M.F., Mucci, N., Funk, S.M., Dolf, G.,
Loeschcke, V., 2000. Mitochondrial DNA analysis in Italian and East European
wolves: detecting the conservation of small population size and hybridization.
Conservation Biology 14, 464–473.
medieval tooth and bone samples. Forensic Science International 151 (1), 31–35.
Richards, M.B., Sykes, B.C., Hedges, R.E.M.,1995. Authenticating DNA extracted from
ancient skeletal remains. Journal of Archaeological Science 22, 291–299.
Sablin, M.V., Khlopachev, G.A., 2002. The earliest ice age dogs: evidence from Eli-
seevichi 1. Current Anthropology 43, 795–799.
Savolainen, P., Zhang, Y., Luo, J., Lundeberg, J., Leitner, T., 2002. Genetic evidence for
an east Asian origin of domestic dogs. Science 298, 1610–1613.
Savolainen, P., Leitner, T., Wilton, A.N., Matisoo-Smith, E., Lundeberg, J., 2004. A
detailed picture of the origin of the Australian dingo, obtained from the study of
mitochondrial DNA. Proceedings of the National Academy of Sciences USA 101
Sundqvist, A.K., Bjo ¨rnerfeldt, S., Leonard, J.A., Hailer, F., Hedhammar, A., Ellegren, H.,
Vila `, C., 2006. Unequal contribution of sexes in the origin of dog breeds.
Genetics 172, 1121–1128.
Tchernov, E., Valla, F.R., 1997. Two new dogs, and other Natufian dogs, from the
southern Levant. Journal of Archaeological Science 24, 65–95.
Tsuda, K., Kikkiwa, Y., Yonekawa, H., Tanabe, Y., 1997. Extensive interbreeding
occurred among multiple matriarchal ancestors during the domestication of
dogs: evidence from inter- and intra-species polymorphisms in the D-loop
region of mitochondrial DNA between dogs and wolves. Genes and Genetic
Systems 72, 229–238.
Vaquer, J., Clottes, J., Coulardou, J., Gireau, J.P., Rouzaud, F.1980. Villeneuve-Tolosane,
Haute Garonne. Rapport de fouilles.
Vaquer, J., 1998. Les se ´pultures du Ne ´olithique moyen en France me ´diterrane ´enne.
In: Guilaine, J. (Ed.), Se ´pultures d’Occident et gene `se des me ´galithismes.
Editions Errance, Paris, pp. 167–186.
Verginelli, F., Capelli, C., Coia, V., Musiani, M., Falchetti, M., Ottini, L., Palmirotta, R.,
Tagliacozzo, A., De Grossi Mazzorin, I., Mariani-Costantini, R., 2005. Mito-
chondrial DNA from Prehistoric canids highlights relationships between dogs
and South-East European wolves. Molecular Biology and Evolution 22 (12),
Vigne, J.D., 2005. L’hume ´rus de chien magdale ´nien de Errala (Gipuzkoa, Espagne) et
la domestication tardiglaciaire du loup en Europe. Munibe. Antropologia-
Vila `, C., Savolainen, P., Maldonado, J.E., Amorin, I.R., Rice, J.E., Honeycutt, R.L.,
Crandall, K.A., Lundeberg, J., Wayne, R.K., 1997. Multiple and ancient origins of
the domestic dog. Science 276, 1687–1689.
Vila `, C., Maldonado, J.E., Wayne, R.K., 1999. Phylogenetic relationships, evolution,
and genetic diversity of the domestic dog. The American Genetic Association
M.F. Deguilloux et al. / Journal of Archaeological Science 36 (2009) 513–519 519