Molecular identification of sheep at Blydefontein Rock Shelter, South Africa

Abstract

The Blydefontein Rock Shelter is a Later Stone Age archaeological site in the eastern Karoo of South Africa. No remains of domesticated animals have been reported although a dung layer, interpreted as deriving from sheep, dates to approximately one thousand years ago. The published morphological analyses of the site’s fauna include many wild taxa, but also report that the majority of the bones in the assemblage were too fragmentary to identify. A recent re-examination of the assemblage identified ten specimens as examples of sheep or goats. In this paper we report on ancient DNA research on the eight specimens we were sent to study, six of which have preserved DNA. Of these, five are examples of wild animals, all of which had been previously identified as present at the site. One specimen was confirmed as a sheep, and it likely comes from a layer that dates to a period well after the initial introduction of domesticates. Direct dating of the specimen is not possible as the entire sample was consumed by the genetic testing. This study highlights the importance of ancient DNA as confirmation of taxon identification when the results of morphological identification challenge the broader culture history.

Full text

65
Southern African Humanities 27: 65–80 November 2015 KwaZulu-Natal Museum
Molecular identication of sheep at Blydefontein Rock Shelter,
South Africa
1K. Ann Horsburgh and 2J. Víctor Moreno-Mayar
1Department of Anthropology, Southern Methodist University, Dallas, TX 75275; School of
Geography, Archaeology and Environmental Studies, University of the Witwatersrand, P.O. Wits,
2050 South Africa; horsburgh@smu.edu
2Centre for GeoGenetics, Natural History Museum of Denmark, 1350 Copenhagen K, Denmark;
morenomayar@gmail.com
ABSTRACT
The Blydefontein Rock Shelter is a Later Stone Age archaeological site in the eastern Karoo of South Africa.
No remains of domesticated animals have been reported although a dung layer, interpreted as deriving
from sheep, dates to approximately one thousand years ago. The published morphological analyses of
the site’s fauna include many wild taxa, but also report that the majority of the bones in the assemblage
were too fragmentary to identify. A recent re-examination of the assemblage identied ten specimens as
examples of sheep or goats. In this paper we report on ancient DNA research on the eight specimens
we were sent to study, six of which have preserved DNA. Of these, ve are examples of wild animals, all
of which had been previously identied as present at the site. One specimen was conrmed as a sheep,
and it likely comes from a layer that dates to a period well after the initial introduction of domesticates.
Direct dating of the specimen is not possible as the entire sample was consumed by the genetic testing.
This study highlights the importance of ancient DNA as conrmation of taxon identication when the
results of morphological identication challenge the broader culture history.
KEY WORDS: Faunal analysis, ancient DNA, Blydefontein, Ovis aries, Capra hircus.
Blydefontein Rock Shelter, in South Africa’s eastern Karoo, has been excavated twice,
once in 1967 and once in 1985 (see Figure 1 for a map showing the site’s location,
and Figure 2 for a stratigraphic section of the site). The faunal remains from those
excavations were analyzed by Richard G. Klein and Kathryn Cruz-Uribe. The results
were published, and at that time no domesticated fauna were reported (Sampson 1970;
Klein 1979; Bousman 1998, 2005). Further, it was reported that the majority of the
bones in the assemblage were too fragmentary to identify (Klein 1979: 36). In 2008
Ina Plug and Karin Scott undertook a re-analysis of the faunal remains recovered from
the 1985 excavations and ten specimens were identied as domestic caprine (sheep or
goat). One specimen in particular (BFT138) proved interesting as a consequence of
a direct AMS (Accelerator Mass Spectrometry) date that seemed to provide evidence
of domestic stock in southern Africa considerably earlier than the generally accepted
date of approximately 2000 years ago (Sealy & Yates 1994; Pleurdeau et al. 2012).
Analyses of DNA from both modern (Horsburgh et al. 2013) and archaeological
domestic fauna (Horsburgh & Rhines 2010; Orton et al. 2013) have proven valuable
in reconstructing relationships among populations, and in verifying morphological
identications of skeletal elements (Loreille et al. 1997; Moss et al. 2006). There was
some doubt about whether the caprine associated with the old radiocarbon date was
from a sheep or a goat, so we were approached to attempt to recover ancient DNA
(aDNA) from eight domestic caprine specimens (Bousman pers. comm. 2013).
ISSN 2305-2791 (online); 1681-5564 (print)

66 SOUTHERN AFRICAN HUMANITIES 27: 65–80, 2015
Fig. 1. Southern Africa, showing the location of Blydefontein Rock Shelter and Die Kelders 1. Redrawn
after Bousman (2005).
Fig. 2. Stratigraphic section of Blydefontein Rock Shelter showing the location of the genetically identied
sheep specimen and relevant radiocarbon determinations. Redrawn after Bousman (2005).

HORSBURGH & MORENO-MAYAR: MOLECULAR IDENTIFICATION OF SHEEP 67
LABORATORY METHODS
All DNA extraction and Illumina library preparation before PCR amplication was
undertaken in the University of Otago’s purpose-built aDNA facility (Knapp et al.
2012). DNA was extracted from between 160 mg and 1.11 g of tooth or bone material
using a standard silica and guanidinium thiocyanate protocol (Rohland & Hofreiter
2007a, b). Two negative controls were processed alongside the eight specimens.
Barcoded Illumina sequencing libraries were constructed directly from both the
aDNA extracts and the associated negative controls using custom Illumina shotgun
adapters as described by Meyer and Kircher (2010). The amplication plateau was
estimated by quantitative PCR using SYBR Green dye (Applied Biosystems) on the
Stratagene MxPro 3000P platform, and amplied libraries were visualized with SYBR
Safe DNA Gel Stain (Invitrogen) on 2 % agarose gels. Libraries were immortalized by
PCR amplication to plateau using ABI’s AmpliTaq Gold with the following reagent
concentrations: 1× AmpliTaq PCR Buffer, 2.5 mM MgCl2, 1 mM dNTPs, 0.2 μM of
each extension primer and 3.75 units of AmpliTaq Gold. Immortalized libraries were
puried using MinElute columns (QIAGEN) following the manufacturer’s protocol
with the addition of a second PE wash and a ve-minute incubation with 0.1× TE
buffer instead of the provided elution buffer.
Libraries were enriched for the mitochondrial genome by in-solution hybridization
following Maricic et al. (2010), but employing a few modications. Each sequencing
library was enriched independently, and pooled in equimolar ratios only after a further
10 cycles of PCR amplication and quantication by qPCR as above. Additionally,
the libraries were eluted from the MyOne Streptavidin C1 Dynabeads (Invitrogen) by
heating for three minutes to 95°C instead of treatment with sodium hydroxide. The
DNA used in the manufacture of sheep mitochondrial DNA capture bait was extracted
from a lamb chop bought in a supermarket in Dunedin, New Zealand. Pooled libraries
were sequenced on the Illumina MiSeq platform with 2 × 75 base paired-end reads.
SEQUENCE PROCESSING AND MAPPING
First, we trimmed adapter-derived sequences, leading Ns and low-quality runs from
the raw sequencing reads using AdapterRemoval 1.5.4 (Lindgreen 2012). We mapped
the ltered reads to a set of reference mitochondrial genomes using the BWA-ALN
algorithm version 0.7.5a-r405 (Li & Durbin 2009). Seeding (-l parameter) was disabled
in order to prevent 5' terminal substitutions characteristic to aDNA to bias the mapping
(Schubert et al. 2012). PCR duplicates were then identied and removed from resulting
bam les using the MarkDuplicates command from the Picard tools suite (http://
broadinstitute.github.io/picard/). We then performed indel-based local realignment
on the reads using GATK (DePristo et al. 2011). Finally, the MD tag was recomputed
for each read using the SAMtools 0.1.18 (Li et al. 2009) calmd command. Mapping
statistics are reported in Table 1. The sequences are available in the supplementary
online materials (www.sahumanities.org).
SPECIES IDENTIFICATION FROM SEQUENCING READS
We took three complementary approaches for identifying the species to which the
biological remains belong.

68 SOUTHERN AFRICAN HUMANITIES 27: 65–80, 2015
TABLE 1
Mapping statistics. For each specimen the rst three rows of statistics refer to mapping to Ovis aries: GI:3445513, Capra hircus: GI:612342193 and Bos taurus:
GI:662034268, and the last row refers to mapping to the assigned species.
Specimen Total reads
Trimmed
reads
# Mapped
reads
# Mapped
reads
(w/o PCR
duplicates)
%
Endogenous
(2 d.p.)
Mean read
length
Coverage
(2 d.p.)
Depth of
coverage
BFT03 1146144 1139066 22629
24556
18374
116147
3378
3425
2616
21050
0.29
0.29
0.23
1.84
81.26
84.09
79.81
95.00
0.46
0.3
0.3
1.00
11.0
11.1
7.8
121.1
BFT07 260737 257604 7917
1597
867
7921
574
131
78
572
0.22
0.05
0.03
0.22
84.28
79.07
83.26
84.00
0.0
0.26
0.2
0.91
2.9
0.5
0.3
2.9
BFT09 370584 366533 365
256
497
NA
92
82
93
NA
0.02
0.02
0.03
NA
73.61
69.88
82.8
NA
0.16
0.12
0.25
NA
0.3
0.3
0.4
NA
BFT133 1067987 1054310 8268
8760
6478
39324
434
481
367
3901
0.04
0.05
0.03
0.37
75.73
78.58
75.41
89.00
0.17
0.16
0.14
0.96
1.5
1.6
1.2
21.1

HORSBURGH & MORENO-MAYAR: MOLECULAR IDENTIFICATION OF SHEEP 69
TABLE 1(continued)
Mapping statistics. For each specimen the rst three rows of statistics refer to mapping to Ovis aries: GI:3445513, Capra hircus: GI:612342193 and Bos taurus:
GI:662034268, and the last row refers to mapping to the assigned species.
Specimen Total reads
Trimmed
reads
# Mapped
reads
# Mapped
reads
(w/o PCR
duplicates)
%
Endogenous
(2 d.p.)
Mean read
length
Coverage
(2 d.p.)
Depth of
coverage
BFT134 660184 855415 899
896
606
4589
395
382
264
2360
0.04
0.04
0.03
0.27
76.22
75.65
74.14
86.00
0.2
0.21
0.16
0.96
1.4
1.4
0.9
12.3
BFT135 344958 341132 829
713
631
NA
104
91
86
NA
0.03
0.03
0.02
NA
62.15
64.32
65.03
NA
0.12
0.09
0.08
NA
0.3
0.3
0.3
NA
BFT137 660184 656024 2348
2564
1811
8675
100
105
75
579
0.02
0.02
0.01
0.09
67.29
63.66
62.92
78.00
0.13
0.1
0.1
0.83
0.3
0.3
0.2
2.8
BFT138 508668 504008 728
734
647
722
106
109
149
332
0.02
0.02
0.03
0.07
65.07
58.79
65.16
66.00
0.09
0.09
0.15
0.54
0.3
0.3
0.4
1.3

70 SOUTHERN AFRICAN HUMANITIES 27: 65–80, 2015
Mapping-based approach
We mapped the sequencing reads to three different reference mitochondrial genomes
(Ovis aries: GI:3445513, Capra hircus: GI:612342193 and Bos taurus: GI:662034268)
independently. We called majority-rule-based haploid consensus sequences using
ANGSD (Korneliussen et al. 2014) after ltering reads with mapping quality lower than
30 and nucleotides with base quality lower than 20. We then performed a BLASTN
(Altschul et al. 1997) search against the nucleotide database, using each consensus
sequence as a query. For each of the three consensus sequences from each sample, we
retrieved the best hit based on the e-value and coverage (Table 2).
De novo assembly-based approach
Mapping-based consensus calling (and further species identication) is susceptible to
being biased by the reference genome to which reads are mapped. Therefore, we also
conrmed the results from the approach above through a de novo assembly procedure.
Since only a small fraction of the ltered reads appeared as endogenous mtDNA
fragments (Table 1), we restricted this analysis to the reads that could be mapped to
any of the mitochondrial references. For each sample, we assembled the corresponding
reads using Velvet 1.2.03 (Zerbino & Birney 2008) with a k-mer length of 21 bases,
and allowing for automatic coverage ltering. We used the resulting contigs as a query
for a BLAST search against the nt database and retrieved the best hit for each contig.
In Table 2 we report the species that was recovered by the majority of the contigs as
a best hit, for each sample.
Species identication
We assigned each sample to a candidate species based on the BLAST results from
both approaches, which were highly concordant. For the mapping-based approach,
we considered the results from the reference that yielded the most mapped reads, and
for the assembly-based approach we considered the two most frequent BLAST hits.
For the cases in which both approaches coincided in the genus assignment but not the
species, the result from the assembly approach was kept.
We then mapped all the reads to a reference mitochondrial genome from the
candidate species and generated consensus sequences, as described above. We then
performed a BLAST search using such consensus sequences and assigned each sample
to the best BLAST hit (Table 2). Note that samples for which we could not map 100 or
more reads to at least two of the three reference genomes were not included at this stage.
CYTOCHROME B ALIGNMENT
Cytochrome b (cyt b) is a mitochondrial gene involved in energy production in the cell.
The DNA sequences of cyt b are relatively stable within species and variable between
species so it has been one of the genes of choice for species identication in archaeology
(Loreille et al. 1997), forensics (Parson et al. 2000; Lee et al. 2009) and conservation
(Hsieh et al. 2001). Its reliability and efcacy for the purposes of species identication
has been demonstrated across many taxa (Kocher et al. 1989; Irwin et al. 1991).
We aligned the cyt b sequences from each of the archaeological specimens with cyt
b from sheep and goat, and all the members of the genus identied by the previous
two approaches using Geneious 6.1.8 (Kearse et al. 2012). We list here the comparative

HORSBURGH & MORENO-MAYAR: MOLECULAR IDENTIFICATION OF SHEEP 71
TABLE 2
Species diagnoses based on mapping available sequences to reference genomes and de novo assembly. Note that while BFT07 most closely resembles an Ovis
orientalis genome deposited in GenBank (accession number KF938360.1), we have assigned BFT07 to Ovis aries, the domestic sheep, because Ovis orientalis (also
known as Ovis gmelinii (IUCN/SSC Caprine Specialist Group 2000)) is the likely ancestor of modern domestic sheep (Demirci et al. 2013) and is therefore
expected to be closely related to domestic sheep.
Specimen
Morphological
identication
Morphological
element
Mapping-
based species
identication
De novo assembly-
based species
identication
Cytochrome b
comparison
species
identication
Species
identication
BFT031Ovis aries Left M1 Antidorcas
marsupialis
Antidorcas
marsupialis Antidorcas marsupialis Antidorcas
marsupialis
BFT071Ovis/Capra Complete 1st
incisor Ovis orientalis Ovis orientalis Ovis aries Ovis aries
BFT091Ovis/Capra Proximal 2nd
phalanx Excluded from analyses due to low DNA quality and quantity
BFT133 Ovis aries Calcanium Antidorcas
marsupialis
Antidorcas
marsupialis
Antidorcas
marsupialis
Antidorcas
marsupialis
BFT134 Ovis aries 3rd phalanx Pelea capreolus Pelea capreolus Pelea capreolus Pelea capreolus
BFT135 Ovis/Capra Excluded from analyses due to low DNA quality and quantity
BFT137 Ovis/Capra Proximal 1st
phalanx Redunca fulvorufula Redunca fulvorufula Redunca fulvorufula Redunca fulvorufula
BFT138 Ovis/Capra Complete 3rd
phalanx
Tragelaphus
strepsiceros Tragelaphus oryx Tragelaphus oryx Tragelaphus oryx

72 SOUTHERN AFRICAN HUMANITIES 27: 65–80, 2015
species we used in each case, with the appropriate GenBank accession numbers. BFT137
was aligned with mountain reedbuck (Redunca fulvorufula, NC_020742.1), southern
reedbuck (Redunca arundinium, NC_020794.1) and Bohar reedbuck (Redunca redunca,
AF096626.1). BFT138 was aligned with eland (Tragelaphus oryx, NC_020750.1), nyala
(Tragelaphus angasii, NC_020748.1), mountain nyala (Tragelaphus buxtoni, AY667216.1),
bongo (Tragelaphus eurycerus, NC_020749.1), lesser kudu (Tragelaphus imberbis,
NC_020619.1), bushbuck (Tragelaphus scriptus, JN632706.1), sitatunga (Tragelaphus spekii,
NC_020620.1), and greater kudu (Tragelaphus strepsiceros, NC_020752.1). This sequence
alignment allowed us to resolve the conict between the species determinations derived
from the mapping (kudu) and de novo assembly (eland) approaches and assign BFT138
to eland.
BFT03 and BFT133 were aligned with springbok (Antidorcas marsupialis, JN632596).
Because the springbok is the only living member of its genus, we also aligned these
sequences with those of the springbok’s closest living relative (Bibi 2013), the gerenuk
(Litocranius walleri, NC_020716.1). BFT134 was aligned with the vaalribbok (Pelea
capreolus, JN632684), which likewise is alone in its genus. We therefore aligned it with
all the Redunca species listed above as well as all the members of the genus Kobus,
specically waterbuck (Kobus ellipsiprymnus, AF096625.1), kob (Kobus kob, AF052939.1),
lechwe (Kobus leche, NC_018603.1), Nile lechwe (Kobus megaceros, AF096620.1) and puku
(Kobus vardonii, AF096619.1). In the interests of space we have not reproduced these
sequence alignments, but they are available on request.
SPECIES IDENTIFICATION BY SEQUENCE DATA AND MORPHOLOGICAL DATA
Morphological analyses of archaeological fauna are integral to reconstructions of
the ancient environments and the behaviour of the people living in them. They are,
however, inherently subjective (Driver 1992; Wolverton 2013), and frequently much
of the bone recovered from archaeological sites is unidentiable to skeletal element
or zoological taxon (Klein & Cruz-Uribe 1984). One of us has argued elsewhere that
genetic data need to be interpreted carefully and integrated with other available lines of
evidence, and that we should furthermore guard against the temptations to go beyond
our data (Horsburgh 2015). We will, nonetheless, show here that the employment of
genetic data to identify the species of fragmentary bones is considerably less fraught
with potential misinterpretation.
The species at issue here are sheep (Ovis aries), goat (Capra hircus), springbok
(Antidorcas marsupialis), eland (Tragelaphus oryx), mountain reedbuck (Redunca fulvorufula)
and a vaalribbok (Pelea capreolus). In the interests of space we have not reproduced the
complete ~16 500 base pairs of the mitochondrial genomes of each of the relevant
species, although they are all available in the National Center for Biotechnology
Information database, GenBank, maintained by the US National Institutes of Health
(www.ncbi.nlm.nih.gov/GenBank). Figure S1 of the supplementary materials shows a
phylogenetic tree constructed using a neighbour-joining method with a Jukes-Cantor
model of molecular evolution executed in Geneious (Kearse et al. 2012). Genbank
accession numbers of the specimens used are listed in the gure legend. The phylogeny
visually represents the extent of evolutionary divergence among each of these species
as reected in the amount of DNA sequence divergence between them. Figure 3A
shows DNA sequences from a small portion of cyt b from each of the relevant species

HORSBURGH & MORENO-MAYAR: MOLECULAR IDENTIFICATION OF SHEEP 73
(base position numbers 14 159 to 14 392). Cyt b was chosen for this illustration because
it has been widely used for the purposes of species identication elsewhere (Parson
et al. 2000; Hsieh et al. 2001). Each of the DNA sequences presented in Figure 3A is
identied by the species name, and each of the associated GenBank accession numbers
is listed in the gure legend. In the presented example, the DNA sequences are aligned
with the sheep mitochondrial reference genome. In the diagram the dots (.) represent
positions in which a given sequence is identical to the sequence to which it has been
aligned, in this case sheep. Where nucleotide bases are represented by letters in the
sequence (A, C, T or G), the specied sequence differs from the sheep sequence to
which it has been aligned. In this way the DNA variability between species can be
observed. The degree of sequence difference is a reection of the length of time to
the most recent common ancestor, with mutations accumulating in the genome in a
more or less clock-like fashion (Zuckerkandl & Pauling 1965).
Figure 3A shows the efcacy of using mitochondrial DNA sequences to identify
species and to distinguish domesticated sheep and goats from wild bovid forms.
Hence, we can have condence in the validity of the genetic identication of the
morphologically identied archaeological specimens.
Specimen BFT07 was morphologically assigned to sheep/goat. In this, the molecular
results are consistent with the morphological diagnosis, and further, we are able to use
the DNA sequence to distinguish between sheep and goat, showing that the specimen
is from a sheep. Figure 3B shows the sheep and goat reference genomes aligned with
BFT07. It can be seen that over the displayed bases, the goat reference sequence differs
from the sheep reference sequence at 26 positions. In contrast, BFT07 does not differ
at all from the sheep reference genome. When the entire mitochondrial genome of
BFT07 is examined, it does differ from the sheep reference genome, as is to be expected;
every species shows some degree of intraspecic genetic variation (Hsieh et al. 2001).
By aligning 518bp of the mitochondrial d-loop of BFT07 with the d-loop sequences
deriving from haplogroups A, B and C sequences reported by Guo et al. (2005), we
were able to determine that BFT07 is a member of haplogroup B.
Figure 3C shows the two specimens we have identied as springbok (BFT03 and
BFT133). To make the visual comparison straightforward we have aligned BFT03,
BFT133, the sheep reference genome and the goat reference genome to the springbok
genome. The dots in this case, then, indicate where the sequences are identical to the
springbok mitochondrial genome. It can be seen that the two archaeological specimens
are very much more similar to the springbok genome than they are to either the sheep
or goat genomes. Figures 3D and 3E show the same types of alignment for specimens
BFT134, genetically identied as vaalribbok (Pelea capreolus), and BFT137, identied as
mountain reedbuck (Redunca fulvorufula).
The most critical specimen in these analyses is BFT138 because it is associated
with a direct AMS date which, if it were a domestic specimen, would demonstrate a
surprisingly early arrival of domesticated fauna in southern Africa. Figure 1F shows
the recovered mitochondrial DNA sequence from BFT138 aligned with the eland
(Tragelaphus oryx) mitochondrial genome above the homologous sequences from both
sheep and goat. BFT138 is clearly an eland, not a sheep or a goat. Specimens BFT09
and BFT135 were poorly preserved and did not yield DNA of sufcient quality or
quantity to allow determination of species.

74 SOUTHERN AFRICAN HUMANITIES 27: 65–80, 2015
Fig. 3A. Sequence alignment of 234bp of the mitochondrial gene cytochrome b. In this gure, and subsequence gures, dots (.) represent nucleotide positions
where the sequence for the given specimen is identical to the sequence at the top of the alignment. For example, in the rst displayed position
(nt 14,159), the sheep mitochondrial genome has an adenine residue (A). The other ve shown species likewise have an A. All six species have the same
nucleotides for the rst four of the displayed positions (ATGA). At the fth position, however, the sheep mitochondrial genome has a thymine (T),
as does the springbok, but the other four species all have cytosines (C). It can be seen here that there are enough differences in the DNA sequences
to condently distinguish them from each other. The Genbank accession numbers for each of the DNA sequences displayed here are as follows:
sheep (NC_001941), goat (NC_005044), mountain reedbuck (JN632695), springbok (JN632596), vaalribbok (JN632684) and eland (NC_020750.1).
Fig. 3B. Sequence alignment of 234bp of the mitochondrial gene cytochrome b of sheep, goat and specimen BFT07. Over this stretch of the gene, the specimen
BFT07 does not differ at all from the sheep reference genome, but differs at 26 positions from the goat reference genome.

HORSBURGH & MORENO-MAYAR: MOLECULAR IDENTIFICATION OF SHEEP 75
Fig. 3C. Sequence alignment of 234bp of the mitochondrial gene cytochrome b of sheep, goat and specimens BFT03 and BFT133 aligned to the springbok
mitochondrial genome. The two archaeological specimens do not differ from the springbok reference genome at all, but differ from the sheep and
goat genomes by 33bp and 40bp respectively.
Fig. 3D. Sequence alignment of 234bp of the mitochondrial gene cytochrome b of sheep, goat and specimen BFT134 aligned to the vaalribbok mitochondrial
genome. The archaeological specimen differs at only one position from the vaalribbok sequence, but differs from the sheep and goat sequences by
20bp and 18bp respectively.

76 SOUTHERN AFRICAN HUMANITIES 27: 65–80, 2015
Fig. 3E. Sequence alignment of 234bp of the mitochondrial gene cytochrome b of sheep, goat and specimen BFT137 aligned to the mountain reedbuck
mitochondrial genome. The archaeological specimen differs at only two positions from the mountain reedbuck sequence, but differs from the sheep
and goat sequences by 24bp and 36bp respectively. Note that the string of 7 Ns in the BFT137 sequence indicates positions in the mitochondrial
genome of the archaeological specimen that we were unable to recover.
Fig. 3F. Sequence alignment of 234bp of the mitochondrial gene cytochrome b of sheep, goat and specimen BFT138 aligned to the eland mitochondrial
genome. The archaeological specimen differs at one position from the eland sequence, but differs from the sheep and goat sequences by 30bp and
23bp respectively.

HORSBURGH & MORENO-MAYAR: MOLECULAR IDENTIFICATION OF SHEEP 77
ANCIENT DNA DAMAGE PATTERNS
Based on the nal sequence alignments to the assigned species, we investigated whether
our sequences bore misincorporation patterns characteristic of aDNA, using bamdamage
(Malaspinas et al. 2014). We observed an excess of C to T substitutions towards the 5'
ends of the reads, which supports the authenticity of the extracted aDNA (Briggs et al.
2007; Krause et al. 2010; Skoglund et al. 2014). Note that in some cases, the expected
complementary G to A substitutions towards the 3' ends are increased to a lesser
extent. This is probably the consequence of reads being shorter than the actual DNA
fragments that were built into sequencing libraries. The heavy tail present in read length
distributions further supports this scenario (see Fig. S2 in the supplementary materials
for plots of the base substitutions). We also observed an overall increased substitution
rate, which could be attributed to the specicity limitations of our species assignment,
which is ultimately limited by the species representation in the nucleotide database.
NEGATIVE CONTROLS
We also produced sequencing data from two negative extraction controls, which we
analyzed with the above mapping pipeline. As expected, these experiments yielded
one and zero mapped reads (to any of the three reference mitochondrial genomes),
respectively. This result, together with the observed misincorporation patterns towards
the reads’ termini, supports the authenticity of the genetic data produced in this study.
DATING THE ARRIVAL OF DOMESTIC FAUNA
The specimen we were able to securely identify as sheep (BFT07) was excavated from
Unit B8, Level 1. Before the specimens were made available for aDNA analysis, they
had been subjected to AMS dating and stable isotope analysis. The remaining portion of
specimen BFT07 was only 0.36 g of material, so it was therefore completely consumed
by the DNA extraction protocols. There are no radiocarbon dates published from Level
1 of Blydefontein. There are, however, published dates from deeper in the sequence
(Bousman 2005). We have taken those published, corrected, dates, and calibrated them
using the SHCal13 calibration curve (Hogg et al. 2013). Level 2, dating to cal. AD
1020–1395 (SMU-1902) provides a lower boundary for the age of Level 1, as well as
a date of cal. AD 640–1025 (SMU-1925) from a dung deposit (CPB) within Level 2,
while a date of cal. AD 678–861 (SMU-1850) from just below the dung feature shows
that the upper layers are well within the last 2000 years. The dung deposit has been
previously interpreted as evidence for the presence of domestic stock (Bousman 1998,
2005), a nding consistent with our identication of a sheep bone in Level 1.
DISCUSSION
A morphological re-analysis of the 1985 Blydefontein fauna identied ten specimens
as either sheep or domestic caprine. This result differs from the original morphological
analyses, which identied no domestic species. We had access to eight of the specimens
identied as domestic fauna, and of those, six of the specimens’ DNA preservation
was sufcient to allow us to diagnose species condently. Crucially, only one of those
six specimens (BFT07) is a sheep. The remainder are wild species, all of which were
identied by Klein and Cruz-Uribe in the original analysis (Bousman 1998).

78 SOUTHERN AFRICAN HUMANITIES 27: 65–80, 2015
All of Africa’s domestic sheep originate outside the continent (Epstein 1971; Clutton-
Brock 1993), most likely from several semi-independent populations in southwestern
Asia (Meadows et al. 2011) where a dramatic decrease in the body size of caprines has
been observed across the Aceramic Neolithic (Meadow 1984, 1993). There are few
comparative data available for the mitochondrial genome of African sheep. A small
fragment of the mitochondrial control region of twenty sheep specimens from the
Later Stone Age deposits at Die Kelders 1 in South Africa’s Western Cape Province
has nevertheless been reported (Horsburgh & Rhines 2010). Like those from Die
Kelders 1, the Blydefontein sheep is a member of haplogroup B, a clade with a now
almost worldwide distribution (Hiendleder, Lewalski et al. 1998; Hiendleder, Mainz et
al. 1998; Hiendleder et al. 2002; Demirci et al. 2013).
It is now evident from the mtDNA that the remaining ve specimens represent
wild species known to inhabit the region, including the specimen that yielded a
surprisingly old direct AMS date, and previously argued to be a domestic species. The
wild forms include an eland (Tragelaphus oryx), two springbok (Antidorcas marsupialis), a
mountain reedbuck (Redunca fulvorufula) and a vaalribbok (Pelea capreolus). The original
zooarchaeological analyses of the fauna recovered from the 1967 and the 1985
excavations at Blydefontein (Sampson 1970; Klein 1979; Bousman 1998, 2005) did not
identify domesticates. Based on Klein’s (1979) comment that the majority of the bones
in the assemblage were too fragmentary to identify, we suggest it likely that the original
analysts regarded the eight specimens under consideration here to be nonidentiable
on morphological grounds.
We do not suggest that aDNA analyses should be considered a replacement for
traditional morphological analyses of archaeological fauna. Morphological analyses
can be undertaken on much larger sample sizes than are practical for genetic analyses,
which are tremendously costly in terms of both time and money. Furthermore,
morphological studies provide information simply inaccessible to genetic analyses. In
this instance, and others like it, aDNA analyses can provide a critical test of hypotheses
of archaeological signicance.
ACKNOWLEDGEMENTS
Horsburgh gratefully acknowledges salary support from a University of Otago Research Grant that also
funded the laboratory costs. Moreno-Mayar was supported by the Lundbeck Foundation, the Danish
National Research Foundation, and the Consejo Nacional de Ciencia y Tecnología (National Council of
Science and Technology) in Mexico. We thank Britt Bousman for providing the specimens for analysis; Lisa
Matisoo-Smith for the use of her laboratory; Zev Kronenberg for preliminary data processing; Richard
G. Klein, David J. Meltzer and Jayson Orton for productive discussions and suggestions; and Mark D.
McCoy for assistance in calibrating radiocarbon determinations and production of gures.
REFERENCES
Altschul, S.F., Madden, T.L., Schaffer, A.A., Zhang, J.H., Zhang, Z., Miller, W. & Lipman, D.J. 1997. Gapped
BLAST and PSI-BLAST: a new generation of protein database search programs. Nucleic Acids
Research 25 (17): 3389–402.
Bibi, F. 2013. A multi-calibrated mitochondrial phylogeny of extant Bovidae (Artiodactyla, Ruminantia)
and the importance of the fossil record to systematics. BMC Evol Biol 13: 166.
Bousman, C.B. 1998. The chronological evidence for the introduction of domestic stock into southern
Africa. African Archaeological Review 15: 133–50.
Bousman, C.B. 2005. Coping with risk: Later Stone Age technological strategies at Blydefontein Rock
Shelter, South Africa. Journal of Anthropological Archaeology 24 (3): 193–226.

HORSBURGH & MORENO-MAYAR: MOLECULAR IDENTIFICATION OF SHEEP 79
Briggs, A.W., Stenzel, U., Johnson, P.L., Green, R.E., Kelso, J., Prufer, K., Meyer, M., Krause, J., Ronan,
M.T., Lachmann, M. & Pääbo, S. 2007. Patterns of damage in genomic DNA sequences from
a Neandertal. Proceedings of the National Academy of Sciences of the United States of America 104
(37): 14616–21.
Clutton-Brock, J. 1993. The spread of domestic animals in Africa. In: T. Shaw, P. Sinclair, B. Andah & A.
Okpoko, eds, The archaeology of Africa: food, metals and towns. London & New York: Routledge,
pp. 61–70.
Demirci, S., Bastanlar, E.K., Dagtas, N.D., Piskin, E., Engin, A., Ozer, F., Yuncu, E., Dogan, S.A. & Togan,
I. 2013. Mitochondrial DNA diversity of modern, ancient and wild sheep (Ovis gmelinii anatolica)
from Turkey: new insights on the evolutionary history of sheep. PLoS One 8 (12): e81952.
DePristo, M.A., Banks, E., Poplin, R., Garimella, K.V., Maguire, J.R., Hartl, C., Philippakis, A.A., del Angel,
G., Rivas, M.A., Hanna, M., McKenna, A., Fennell, T.J., Kernytsky, A.M., Sivachenko, A.Y.,
Cibulskis, K., Gabriel, S.B., Altshuler, D. & Daly, M.J. 2011. A framework for variation discovery
and genotyping using next-generation DNA sequencing data. Nature Genetics 43 (5): 491–8.
Driver, J.C. 1992. Identication, classication and zooarchaeology. Circaea 9: 35–47.
Epstein, H. 1971. The origin of domestic animals of Africa. Vol. 1. New York: Africana Publishing Corporation.
Guo, J., Du, L.X., Ma, Y.H., Guan, W.J., Li, H.B., Zhao, Q.J., Li, X. & Rao, S.Q. 2005. A novel maternal
lineage revealed in sheep (Ovis aries). Animal Genetics 36 (4): 331–6.
Hiendleder, S., Lewalski, H., Wassmuth, R. & Janke, A. 1998. The complete mitochondrial DNA sequence
of the domestic sheep (Ovis aries) and comparison with the other major ovine haplotype. Journal
of Molecular Evolution 47 (4): 441–8.
Hiendleder, S., Mainz, K., Plante, Y. & Lewalski, H. 1998. Analysis of mitochondrial DNA indicates that
domestic sheep are derived from two different ancestral maternal sources: no evidence for
contributions from urial and argali sheep. Journal of Heredity 89 (2): 113–20.
Hiendleder, S., Kaupe, B., Wassmuth, R. & Janke, A. 2002. Molecular analysis of wild and domestic sheep
questions current nomenclature and provides evidence for domestication from two different
subspecies. Proceedings of the Royal Society B 269 (1494): 893–904.
Hogg, A.G., Hua, Q., Blackwell, P.G., Niu, M., Buck, C.E., Guilderson, T.P., Heaton, T.J., Palmer, J.G.,
Reimer, P.J., Reimer, R.W., Turney, C.S.M & Zimmerman, S.R.H. 2013. SHCal13 Southern
Hemisphere calibration, 0–50 000 years cal BP. Radiocarbon 55 (4): 1889–903.
Horsburgh, K.A. 2015. Molecular anthropology: the judicial use of genetic data in archaeology. Journal
of Archaeological Science 56: 141–5.
Horsburgh, K.A. & Rhines, A. 2010. Genetic characterization of an archaeological sheep assemblage from
South Africa’s Western Cape. Journal of Archaeological Science 37: 2906–10.
Horsburgh, K.A., Prost, S., Gosling, A., Stanton, J.A., Rand, C. & Matisoo-Smith, E.A. 2013. The genetic
diversity of the Nguni breed of African cattle (Bos spp.): complete mitochondrial genomes of
haplogroup T1. PLoS One 8 (8): e71956.
Hsieh, H.M., Chiang, H.L., Tsai, L.C., Lai, S.Y., Huang, N.E., Linacre, A. & Lee, J.C. 2001. Cytochrome b
gene for species identication of the conservation animals. Forensic Science International 122 (1): 7–18.
Irwin, D.M., Kocher, T.D. & Wilson, A.C. 1991. Evolution of the cytochrome b gene of mammals. Journal
of Molecular Evolution 32 (2): 128–44.
IUCN/SSC Caprine Specialist Group. 2000. Workshop on Caprinae taxonomy, Ankara, Turkey, May 8–10.
Kearse, M., Moir, R., Wilson, A., Stones-Havas, S., Cheung, M., Sturrock, S., Buxton, S., Cooper, A.,
Markowitz, S., Duran, C., Thierer, T., Ashton, B., Mentjies, P. & Drummond, A. 2012. Geneious
Basic: an integrated and extendable desktop software platform for the organization and analysis
of sequence data. Bioinformatics 28 (12): 1647–9.
Klein, R.G. 1979. Paleoenvironmental and cultural implications of late Holocene archaeological faunas
from the Orange Free State and north-central Cape Province, South Africa. South African
Archaeological Bulletin 34: 34–49.
Klein, R.G. & Cruz-Uribe, K. 1984. The analysis of animal bones from archaeological sites. Chicago: The University
of Chicago Press.
Knapp, M., Clarke, A.C., Horsburgh, K.A. & Matisoo-Smith, E.A. 2012. Setting the stage: building and
working in an ancient DNA laboratory. Annals of Anatomy–Anatomischer Anzeiger 194 (1): 3–6.
Kocher, T.D., Thomas, W.K., Meyer, A., Edwards, S.V., Pääbo, S., Villablanca, F.X. & Wilson, A.C. 1989.
Dynamics of mitochondrial-DNA evolution in animals: amplication and sequencing with
conserved primers. Proceedings of the National Academy of Sciences of the United States of America
86 (16): 6196–200.
Korneliussen, T.S., Albrechtsen, A. & Nielsen, R. 2014. ANGSD: analysis of next generation sequencing
data. BMC Bioinformatics 15: 356.

80 SOUTHERN AFRICAN HUMANITIES 27: 65–80, 2015
Krause, J., Fu, Q.M., Good, J.M., Viola, B., Shunkov, M.V., Derevianko, A.P. & Pääbo, S. 2010. The
complete mitochondrial DNA genome of an unknown hominin from southern Siberia. Nature
464 (7290): 894–7.
Lee, J.C.I., Tsai, L.C., Liao, S.P., Linacre, A. & Hsieh, H.M. 2009. Species identication using the cytochrome
b gene of commercial turtle shells. Forensic Science International: Genetics 3 (2): 67–73.
Li, H. & Durbin, R. 2009. Fast and accurate short read alignment with Burrows-Wheeler transform.
Bioinformatics 25 (14): 1754–60.
Li, H., Handsaker, B., Wysoker, A., Fennell, T., Ruan, J., Homer, N., Marth, G., Abecasis, G., Durbin, R. &
1000 Genome Project Data Proc. 2009. The Sequence Alignment/Map format and SAMtools.
Bioinformatics 25 (16): 2078–9.
Lindgreen, S. 2012. AdapterRemoval: easy cleaning of next-generation sequencing reads. BMC Res Notes 5: 337.
Loreille, O., Vigne, J.D., Hardy, C., Callou, C., Treinen-Claustre, F., Dennebouy, N. & Monnerot, M. 1997.
First distinction of sheep and goat archaeological bones by the means of their fossil mtDNA.
Journal of Archaeological Science 24 (1): 33–7.
Malaspinas, A.S., Tange, O., Moreno-Mayar, J.V., Rasmussen, M., DeGiorgio, M., Wang, Y., Valdiosera, C.E.,
Politis, G., Willerslev, E. & Nielsen, R. 2014. Bammds: a tool for assessing the ancestry of low-
depth whole-genome data using multidimensional scaling (MDS). Bioinformatics 30 (20): 2962–4.
Maricic, T., Whitten, M. & Pääbo, S. 2010. Multiplexed DNA sequence capture of mitochondrial genomes
using PCR products. PLoS One 5 (11): e14004.
Meadow, R.H. 1984. Animal domestication in the Middle East: a view from the Eastern margin. In: J.
Clutton-Brock & C. Grigson, eds, Animals and archaeology: 3. Early herders and their ocks. Oxford:
Oxford University Press, pp. 309–37.
Meadow, R.H. 1993. Animal domestication in the Middle East: a revised view from the Eastern Margin.
In: G. Possehl, ed., Harappan Civilisation. New Delhi: Oxford & IBH, pp. 295–320.
Meadows, J.R.S., Hiendleder, S. & Kijas, J.W. 2011. Haplogroup relationships between domestic and wild
sheep resolved using a mitogenome panel. Heredity 106 (4): 700–6.
Meyer, M. & Kircher, M. 2010. Illumina sequencing library preparation for highly multiplexed target
capture and sequencing. Cold Spring Harbor Protocols 2010 (6): pdb prot5448.
Moss, M.L., Yang, D.Y., Newsome, S.D., Speller, C.F., McKechnie, I., McMillan, A.D., Losey, R.J. & Koch, P.L.
2006. Historical ecology and biogeography of North Pacic pinnipeds: isotopes and ancient DNA
from three archaeological assemblages. The Journal of Island and Coastal Archaeology 1 (2): 165–90.
Orton, J., Mitchell, P., Klein, R., Steele, T. & Horsburgh, K.A. 2013. An early date for cattle from Namaqualand,
South Africa: implications for the origins of herding in southern Africa. Antiquity 87: 108–20.
Parson, W., Pegoraro, K., Niederstatter, H., Foger, M. & Steinlechner, M. 2000. Species identication by
means of the cytochrome b gene. International Journal of Legal Medicine 114 (1–2): 23–8.
Pleurdeau, D., Imalwa, E., Detroit, F., Lesur, J., Veldman, A., Bahain, J.J. & Marais, E. 2012. ‘Of sheep
and men’: earliest direct evidence of caprine domestication in southern Africa at Leopard Cave
(Erongo, Namibia). PLoS One 7 (7): e40340.
Rohland, N. & Hofreiter, M. 2007a. Ancient DNA extraction from bones and teeth. Nature Protocols 2
(7): 1756–62.
Rohland, N. & Hofreiter, M. 2007b. Comparison and optimization of ancient DNA extraction. Biotechniques
42 (3): 343–52.
Sampson, C.G. 1970. The Smitheld Industrial Complex: further eld results. Bloemfontein: National Museum.
Schubert, M., Ginolhac, A., Lindgreen, S., Thompson, J.F., Al-Rasheid, K.A., Willerslev, E., Krogh, A. &
Orlando, L. 2012. Improving ancient DNA read mapping against modern reference genomes.
BMC Genomics 13: 178.
Sealy, J. & Yates, R. 1994. The chronology of the introduction of pastoralism to the Cape, South Africa.
Antiquity 68 (258): 58–67.
Skoglund, P., Northoff, B.H., Shunkov, M.V., Derevianko, A.P., Pääbo, S., Krause, J. & Jakobsson, M. 2014.
Separating endogenous ancient DNA from modern day contamination in a Siberian Neandertal.
Proceedings of the National Academy of Sciences of the United States of America 111 (6): 2229–34.
Wolverton, S. 2013. Data quality in zooarchaeological faunal identication. Journal of Archaeological Method
and Theory 20 (3): 381–96.
Zerbino, D.R. & Birney, E. 2008. Velvet: algorithms for de novo short read assembly using de Bruijn graphs.
Genome Research 18 (5): 821–9.
Zuckerkandl, E. & Pauling, L. 1965. Molecules as documents of evolutionary history. Jour nal of Theoretical
Biology 8: 357–66.