ArticlePDF Available

Four Iron Age women from KwaZulu-Natal: biological anthropology, genetics and archaeological context


Abstract and Figures

We report further details on four partial human skeletons from KwaZulu-Natal previously selected for genetic analysis. Dating and genetic results indicate that they derived from agriculturist communities of the mid-second millennium AD. Morphological and genetic analysis shows that three individuals were female; identification of the fourth as female comes from genetic analysis only. All four were adults at death, three older adults and one younger. Genetically, all four individuals cluster strongly with Bantu-speaking populations with West African roots, a result supported by craniometric data for the one individual with a complete and well-preserved cranium. All nevertheless display some admixture with Khoe-San populations. We show that three of the women, and probably the fourth, carried genetic resistance to the Plasmodium vivax malaria parasite, while two had some protection against Trypanosoma brucei gambiense-induced sleeping sickness. The unusual rock-shelter burial locations of three of the women suggest that their deaths required ritual 'cooling'. Lightning and violence are possible causes. We argue that this multipronged approach is necessary for the development of detailed and nuanced understandings of the past and of the individuals who lived in the region centuries ago.
Content may be subject to copyright.
Southern African Humanities 32: 23–56 April 2019 KwaZulu-Natal Museum
Four Iron Age women from KwaZulu-Natal: biological
anthropology, genetics and archaeological context
1Maryna Steyn, 2Gavin Whitelaw, 1Deona Botha, 3Mário Vicente,
3,4,5Carina M. Schlebusch and 5Marlize Lombard
1Human Variation & Identication Research Unit, School of Anatomical Sciences, Faculty of Health
Sciences, University of the Witwatersrand, 7 York Road, Parktown, 2193 South Africa;
2 KwaZulu-Natal Museum, Private Bag 9070, Pietermaritzburg, 3200 South Africa & School of Social
Sciences, University of KwaZulu-Natal, Private Bag X54001, Durban, 4000 South Africa
3 Human Evolution, Department of Organismal Biology, Uppsala University, Norbyvägen 18C, SE-752
36 Uppsala, Sweden
4 SciLifeLab, Uppsala, Sweden
5 Palaeo-Research Institute, University of Johannesburg, PO Box 524, Auckland Park, 2006 South Africa
We report further details on four partial human skeletons from KwaZulu-Natal previously selected for
genetic analysis. Dating and genetic results indicate that they derived from agriculturist communities of the
mid-second millennium AD. Morphological and genetic analysis shows that three individuals were female;
identication of the fourth as female comes from genetic analysis only. All four were adults at death,
three older adults and one younger. Genetically, all four individuals cluster strongly with Bantu-speaking
populations with West African roots, a result supported by craniometric data for the one individual with a
complete and well-preserved cranium. All nevertheless display some admixture with Khoe-San populations.
We show that three of the women, and probably the fourth, carried genetic resistance to the Plasmodium
vivax malaria parasite, while two had some protection against Trypanosoma brucei gambiense-induced sleeping
sickness. The unusual rock-shelter burial locations of three of the women suggest that their deaths required
ritual ‘cooling’. Lightning and violence are possible causes. We argue that this multipronged approach is
necessary for the development of detailed and nuanced understandings of the past and of the individuals
who lived in the region centuries ago.
KEY WORDS: Physical anthropology, ancient DNA, palaeopathology, Bantu-speaker expansion, Iron Age.
South Africa has an archaeological record spanning at least two million years, broadly
subdivided into Stone Age and Iron Age ‘technocomplexes’. The Stone Age includes
hunter-gatherer communities with deep histories in the region, as well as early pastoralist
communities (Lombard et al. 2012). The Iron Age, by contrast, results from the spread
of Bantu-speaking agriculturists from the Nigeria-Cameroon borderlands into savanna
regions south of the equator, to rst settle south of the Limpopo River some 1 700
years ago (Phillipson 1977: 227–30; Huffman 2007: 331–40). The recovery of ancient
human remains, either accidently or from various archaeological sites, adds a valuable
human biological dimension to our understanding of the past, because they can provide
information on past lifeways, adaptation and resistance to disease.
In a recent study, Schlebusch et al. (2017) successfully extracted and analysed the
DNA of seven ancient individuals from KwaZulu-Natal, South Africa. These skeletons
were selected based on their geographical location and potential to yield DNA. Three
of the seven skeletons originated from the Stone Age, with radiocarbon dates of
about 2 000 years ago. Full genomic DNA of the autosomes securely associated these
individuals with genetic variation found in modern Khoe-San or Khoe-San descendants.
The closest genetic match was to ‘Karretjie People’ from Colesberg in the Northern
ISSN 2305-2791 (online); 1681-5564 (print)
Cape (Schlebusch et al. 2017; also Schlebusch et al. 2011, Schlebusch et al. 2016). These
results are based on comparative data currently available, but could be rened if, or
when, additional data become available. Ribot et al. (2010) provided some physical
anthropological data on the Stone Age individuals, and Pfeiffer et al. (2019) recently
integrated new bio-archaeological data with their DNA information.
The remaining four skeletons yielded radiocarbon ages between 590 and 370 years
ago. Their DNA associated them securely with the genetic variation recorded for
Bantu-speaking Africans, and most notably for modern South African Bantu speakers.
One individual in this group, however, also carried a mitochondrial sub-haplogroup
common in modern Khoe-San (Schlebusch et al. 2017). Here we discuss further
details associated with these Iron Age individuals. We provide complete biological-
anthropological analyses of their skeletal remains, reassess and further analyse the DNA
data obtained by Schlebusch et al. (2017), and place them within a broad archaeological
context. Morris (2017: 2) decries “the divorce between bio-archaeology and genetics”,
suggesting that genetic research tends to ignore information from the physical remains
of humans to such an extent that the individual/skeleton becomes irrelevant. Here
we show that for some teams working towards reconstructing sub-Saharan population
histories, research includes mega-approaches such as the reconstruction, analysis and
discussion of full genomes (Schlebusch et al. 2017; Lombard et al. 2018; Schlebusch
& Jakobsson 2018), as well as consideration of the details of individuals and their
contexts (Pfeiffer et al. 2019).
Biological anthropology
The four sets of human remains included in this study originate from the uThukela
basin: three from its upper margins (Eland Cave, Champagne Castle and Newcastle),
and the fourth from its middle reaches (Mfongosi) (Fig. 1). They are housed at the
KwaZulu-Natal Museum in Pietermaritzburg.
For purposes of age estimation, we use standard methods of analysis for adults
(İşcan & Steyn 2013). These include assessment of cranial suture closure (Acsádi &
Nemeskéri 1970), pubic symphyses (Brooks & Suchey 1990) and general degenerative
changes, including changes in the rst rib (Kunos et al. 1999). In addition, we note the
degree of attrition of teeth (Molnar 1971; Scott 1979).
Where possible, we assess sex using the morphological features of the pelvis (Phenice
1969; İşcan & Steyn 2013) and skull (Buikstra & Ubelaker 1994; İşcan & Steyn 2013).
In addition, measurements of the available long bones supplement the assessment of
sex. For this purpose, we use data from black South Africans (Steyn & İşcan 1999;
Asala et al. 2004), as they provide the most relevant reference standards to the study.
Similarly, we use regression formulae for black South Africans to calculate antemortem
stature (Lundy & Feldesman 1987), with soft tissue correction factors from Raxter et
al. (2006). We also consider signs of trauma and pathology.
To gain more information on the population afnity of available crania, we performed
a FORDISC 3 analysis (Ousley & Jantz 2012). FORDISC 3 is a programme that uses
discriminant-function analysis to assess sex and/or ancestry in skeletal remains. In this
case we used cranial measurements to assess possible ancestral relationships, drawing
on a database of various 19th- and 20th-century ‘populations’ from which the analyst
must rst select possible populations of origin for comparative purposes. Statistical
output includes group membership (sex or ancestry), cross-validated classication
accuracy, posterior probabilities, and typicalities. Its limited database implies that
FORDISC should ideally be part of a suite of techniques used to assess archaeological
skeletons. The population categories we selected for FORDISC comparison derive
from Howells’s 1973 database and include Teita (East Africa), Dogon (West Africa),
‘Bushman’ (Khoe-San of southern Africa), and Zulu (South Africa, typically associated
with KwaZulu-Natal).
Fig. 1. KwaZulu-Natal and adjacent areas. Site codes: 1 = Eland Cave; 2 = Kaybar’s Cave; 3 = Champagne
Castle Hostel; 4 = Sewula Gorge; 5 = Ntomdadlana; 6 = iGujwana.
Moor Park
uThukela River
Delagoa Bay
provincial borders
national borders
100 km
uMzinyathi River
iMfolozi River
DNA analyses
We reanalysed the full genome shotgun sequenced genetic data (generated on an
Illumina HiSeq 2500 and HiSeq XTen platform) retrieved from each of the four
individuals (Schlebusch et al. 2017), using an extended dataset of 50 living populations
(Schlebusch et al. 2012; Auton et al. 2015; Gurdasani et al. 2015; Montinaro et al. 2016).
We explored their genetic afnities based on autosomal data (i.e. chromosomes 1–22,
excluding the X, Y and mtDNA) with principal component analysis (PCA) (Patterson
et al. 2006), clustering analysis (Alexander et al. 2009), and formal tests of admixture
and admixture fractions (Patterson et al. 2012). These analyses were done in addition
to, and in more detail than, those reported by Schlebusch et al. (2017).
We merged the four individuals with 50 African populations and three Eurasian
populations from four datasets (Schlebusch et al. 2012; Auton et al. 2015; Gurdasani
et al. 2015; Montinaro et al. 2016). Data were quality ltered using standard quality
lters, keeping only autosomal data. A maximum of 10% missing data was allowed.
We also ltered for genotyping errors (Hardy-Weinberg equilibrium), and all relatives
were excluded. Merging and data ltering were done with PLINK (Purcell et al. 2007).
Ultimately, we obtained a dataset of 882 individuals and 367 244 markers. We used the
entire dataset to conduct ADMIXTURE analyses (Alexander et al. 2009) to investigate
relationships among individuals, with the number of clusters, K, set from 2 to 8 and
replicated 50 times. The cluster-inference and visual inspection were made with Pong
v.1.4.5 (Behr et al. 2016).
Eurasian groups were excluded for PC analyses, and PCA was performed using
EIGENSOFT under default settings (Patterson et al. 2006). We tested for the presence
of gene ow between groups (i.e. admixture) using f4-ratio tests implemented in the
ADMIXTOOLS programme (Patterson et al. 2012), using Japanese (as an unrelated
group), Yoruba (as a West African ancestry proxy), Ju|’hoansi (as a hunter-gatherer
ancestry proxy), and the ancestral state present in human, chimpanzee, gorilla and
orang-utan as outgroups. Linkage-disequilibrium decay analyses were conducted with
Alder (Loh et al. 2013), under default parameters for the modern populations, and with
broader binsize resolution of 0.5 cM for the archaeological group.
We provide a brief curatorial biography for each skeleton and present the results of the
analyses. Tables 1 and 2 (after the references) contain all cranial, postcranial and dental
measurements that could be recorded, depending on the completeness of the remains.
Eland Cave 1925/037 (previously 2530F)
Archaeological and curatorial background
In 1926 Johannes S. Lombard sold a ‘Bushman hunting kit’ he had found in a rock
shelter in the Maloti-Drakensberg range to the Natal Museum (Vinnicombe 1971). The
museum director, Ernest Warren, asked that he examine the site further for paintings
and other artifacts, and Lombard subsequently sent four human bones and a broken
ivory bangle to the museum. The human remains and bangle were originally assigned
the accession number 2530 in the Ethnology Register, but the human bones were
later separated into the Archaeology Collection as 1925/037, with the ‘ctitious year’
of 1925 indicating a site in Natal rather than the year of accession (KwaZulu-Natal
Museum records). Wells (1933a) later named the rock shelter Cave of the Eland, and
it is now called Eland Cave. Today, Eland Cave is best known for its rock paintings
and is widely regarded as one of the nest painted sites in southern Africa (e.g. Pager
1971; Swart 2004; Mazel 2009). Excavations conducted in 1931 and 1980 yielded Later
Stone Age artifacts and a few potsherds (Stein 1933; Wells 1933a; Aron Mazel pers.
comm.). An AMS determination from a tibia of the Eland Cave individual yielded
a date-range of cal. AD 1430–1460 and a δ13C value of −9.4‰ (Beta-398219)
(Schlebusch et al. 2017) (Table 5).
Skeletal analysis
The remains comprise a complete left rst rib, one right rib that had been fractured
postmortem, the left rst metatarsal and the distal half of the left tibia (Fig. 2). Sex,
stature and ancestry could not be determined morphologically without the cranium,
pelvis and complete long bones. The distal epiphysis of the tibia is completely fused,
suggesting that the individual was older than 17 years (if male) or 20 years (if female)
at death. No osteophytes occur on the distal articular surface of the tibia. The surface
of the costal face of the rst rib is irregular and the margins are rugged with swollen
projections, which is consistent with an age estimate of 50+ years (Kunos et al. 1999;
Kurki 2005; DiGangi et al. 2009). It is therefore possible that the remains belonged
to a mature or older adult.
Genetic prole
This individual had the XX chromosome composition of a female, and the L3e3b1
mitochondrial haplogroup. The L3e mitochondrial DNA lineage is common in
Fig. 2. (A) The left rst rib, (B) a right-sided rib (fractured), (C) left rst metatarsal and (D) left distal tibia
of the Eland Cave 1925/037 skeleton.
West Africa today and is thought to have reached southern Africa with the Iron
Age expansion early in the rst millennium AD (Salas et al. 2002; Soares et al. 2012;
Schlebusch & Jakobsson 2018). These lineages are generally absent in people of Khoe-
San descent today (with the exception of the Khwe), but common in Bantu-speaking
people (Schlebusch et al. 2013). The woman’s autosomal ancestry falls securely within
the variation of modern Bantu speakers (see Fig. 11 with discussion). Her Khoe-
San admixture comprised 5.3% (Table 3, Fig. 3). She carried a copy of the malaria-
resistant Duffy null allele and was likely heterozygous C/T at this locus, which would
have afforded her protection against Plasmodium vivax malaria (McManus et al. 2017),
especially in combination with the ATP2B4 protective allele that was also present in
heterozygous form. In addition, she was heterozygous for the APOL1 gene SNP
rs73885319 (i.e. the G1 variant), which confers resistance to African sleeping sickness
or human African trypanosomiasis (Genovese et al. 2010), a tsetse-y-borne parasitic
disease. The individual did not carry any of the known genetic variants linked to lactose
tolerance and she was therefore likely lactose intolerant.
Champagne Castle 2009/023
Archaeological and curatorial background
The Champagne Castle skeleton was originally part of the Durban Natural Science
Museum’s collection with accession number 3814. The museum apparently acquired the
skeleton from a G.R. (or C.R.) Ledward. It transferred the skeleton to the KwaZulu-Natal
Museum in the 1990s, initially temporarily for research purposes, then permanently. A
label on the occipital bone of the skull indicates that the skeleton came from a rock
shelter near the Champagne Castle Hostel in the Maloti-Drakensberg range. It reads:
Ac. No. 3814
From Cave near Champagne Castle Hostel
M. Loskop. Natal. 20/7/1933 C.R. Ledward.
All other labelled bones read ‘G.R. Ledward’. An AMS determination from the femur
yielded a 1σ date-range of cal. AD 1510–1550, 1560–1570, 1620–1660 and a δ13C value
of −11.0‰ (Beta-398218) (Schlebusch et al. 2017) (Table 5).
Results of the f4 ratio test for fractions of Khoe-San ancestry in Iron Age ancient individuals and
in modern-day South African Bantu speakers. * = estimated fraction of Khoe-San ancestry. JPT =
Japanese; YRI = Yoruba; Ancestr = Ancestral outgroup; SEBantu = Southeastern Bantu-speakers
from South Africa.
f4 numerator f4 denominator alpha stderr Z 1-alpha*
JPT-Ancestr-ElandCave-Juhoansi JPT-Ancestr-YRI-Juhoansi 0.947 0.0301 31.455 0.053
JPT-Ancestr-ChampCastle-Juhoansi JPT-Ancestr-YRI-Juhoansi 0.904 0.0569 15.876 0.096
JPT-Ancestr-Mfongosi-Juhoansi JPT-Ancestr-YRI-Juhoansi 0.870 0.0374 23.292 0.130
JPT-Ancestra-Newcastle-Juhoansi JPT-Ancestr-YRI-Juhoansi 0.867 0.0260 33.358 0.133
JPT-Ancestr-Zulu-Juhoansi JPT-Ancestr-YRI-Juhoansi 0.827 0.0104 79.310 0.173
JPT-Ancestr-SEBantu-Juhoansi JPT-Ancestr-YRI-Juhoansi 0.787 0.0086 91.476 0.213
JPT-Ancestr-Sotho-Juhoansi JPT-Ancestr-YRI-Juhoansi 0.781 0.0098 79.905 0.219
Fig. 3. (A) San admixture proportions in modern-day southeast Bantu speakers (circle symbols) and the Iron Age individuals (other symbols). (B) Weighted LD
decay curves for southern African groups.
Skeletal analysis
The remains comprise a complete cranium and mandible (Fig. 4), one cervical vertebra,
right humerus, right ulna, right radius, left pubic symphysis, right os coxa (damaged
postmortem), right femur and a few small fractured bone pieces.
The presence of a preauricular sulcus and very wide greater sciatic notch indicate a
female. A prominent preauricular sulcus indicates that she had most probably borne at
least one child during her lifetime. Small mastoids, nuchal crest and glabella, as well as
sharp supraorbital margins, are also characteristic of a female individual. The humeral
epicondylar breadth of 57 mm and femoral head diameter of 42 mm further indicate
a female (Steyn & İşcan 1999; Asala et al. 2004).
Cranial suture closure indicates that this individual was between 15 and 40 years
old when she died. The pubic symphysis is in phase 4 (26–70 years) with a mean age
of 38.2 years. The synchondrosis sphenooccipitalis is fused. These estimates suggest
that the individual was most likely a young adult. Standardization guidelines (Falys &
Lewis 2011) put the age range at 20–39 years.
A wide interorbital breadth, prognathic facial prole and smooth zygomaxillary suture
indicate an individual of African descent. The cranium has an interparietal groove/
depression to the posterior part of the parietal bones on either side of the sagittal
suture, as well as an inferior frontal eminence (only on the right side), and bilateral mons
temporosphenoidalis. These features have been described as characteristic of Khoe-San
crania (Dart 1924; De Villiers 1968). Using FORDISC 3, we compared the individual
to African samples in Howells’s (1973) database. Ten variables, in a Forward Wilk’s
stepwise analysis for females, placed the individual in the Dogon (West African) group
with a 94.5% posterior probability. With Dogon removed from the comparative list,
the Champagne Castle individual classied as either Teita (56.7% posterior probability)
or Zulu (43.1% posterior probability) (Fig. 5).
The maximum length of the humerus indicates an antemortem stature of 153.8±
3.72 cm, which is shorter than that of modern South African Bantu-speaking females,
but similar to gures reported for Khoe-San groups (Tobias 1961, 1962; Truswell &
Hansen 1976).
Eight permanent teeth and one tooth root are present and include the maxillary
right canine, right central incisor, left lateral incisor and root of the second premolar,
as well as the mandibular right second and third molars, left canine and left second
Fig. 4. Skull of Champagne Castle 2009/023 in anterior and lateral views.
Fig. 5. Graphic representation of the FORDISC results.
Fig. 6. Caries at the cemento-enamel junction of the mandibular second and third molars on the left and
right sides of Champagne Castle 2009/023.
and third molars. Moderate to advanced wear is present on all teeth. She had advanced
dental disease: periodontal disease is evident around the mandibular rst molars and
second premolars, as well as the maxillary molars and premolars. Caries occur on all
four lower molars (second and third molars) on the buccal sides (Fig. 6), as well as the
upper right canine (distal side) and remaining root of the upper left second premolar
(occlusal surface). There is an abscess on the maxilla in the area of the left second
premolar root. The mandibular rst molars, as well as the maxillary rst, second and
third molars and second premolars were lost antemortem.
-2.0 0.0 2.0 4.0
Can 1 (61.2%)
Can 2 (27.4%)
The right parietal bone presents with a depressed fracture that appears unhealed
and possibly occurred perimortem (Fig. 7A). The lesion is oblong in shape, measures
40×17 mm, and is situated 25 mm behind the coronal suture and 24 mm to the right
of the sagittal suture. The outer surface of the fracture suggests that the bone was
wet/elastic (green-bone response) at the time the defect occurred, suggesting that it
occurred perimortem. The bone showed postmortem dehydration before varnish was
applied, trapping dirt on the fracture which made observation of the compression-
tension patterns difcult. A CT scan image of the lesion shows that the fracture
penetrates the outer table of the cranium, but the inner table remains intact (Fig. 7B).
This fracture probably reects an episode of interpersonal violence, but we cannot
determine whether this traumatic injury was the cause of death. No other signs of
trauma or pathology were observed.
Genetic prole
This individual had the XX chromosome composition of a female, supporting the
results from the skeletal analysis. She had the L0d2a1a mitochondrial haplogroup,
which is associated with Khoe-San (Schlebusch et al. 2017), but occurs at moderate
Fig. 7. Skull of Champagne Castle 2009/023. (A) Photographs of the depressed fracture on the right
parietal bone. (B left) CT scan of the depressed fracture. (B right) X-ray image showing that the
lesion penetrates the outer table of the cranium, but leaves the inner table intact yet slightly indented.
frequencies in modern Bantu speakers (12%), most probably due to admixture with
Khoe-San (Schlebusch et al. 2013). Indeed, the L0d2a haplogroup is the most frequent
Khoe-San-associated haplogroup in South African Bantu speakers today (Schlebusch
et al. 2013). The Champagne Castle woman’s autosomal results, however, indicate a
genetic ancestry largely associated with Bantu speakers (90.4%), with 9.6% genome-
wide Khoe-San admixture (Table 3, Fig. 3). Presence/absence of the Duffy null allele
(which confers malaria protection) could not be determined because of the absence of
sequence coverage in this region. This individual nevertheless had at least one copy of
the allele that indicates the Duffy FY*B allele for the for the FY*A/B locus, meaning
that she possibly harboured the Duffy null allele (since the Duffy null allele occurs
on a FY*B chromosomal background). The individual did not have enough coverage
of the genomic regions containing the genetic variants coupled to lactose tolerance
and African sleeping sickness; it is therefore not possible to comment on her lactose
tolerance or her sleeping-sickness resistance.
Mfongosi 1925/036
Archaeological and curatorial background
Amateur naturalist W.E. ‘Mamba’ Jones of iMfongosi in the lower uThukela valley found
a “fairly complete” skeleton beneath the roots of a large olive tree (Olea europaea subsp.
africana) on the northern side of the uThukela river, a little downstream of its conuence
with the iMfongosi (Davies 1957: 48). The body was buried in a exed position in a
grave on the “slope of river-gorge” (KwaZulu-Natal Museum records). Jones believed
that the skull was of Broken Hill type and in May 1932 presented the skeleton to the
Natal Museum (original acc. no. 2737, now 1925/036). In a letter dated 25 July 1934,
Jones informed the museum director, Ernst Warren, that his son had unearthed more
bones from the skeleton; by then Warren was on leave, attending to museum business
in Europe and America, which ended with his retirement in December 1934. Jones
incorrectly told Oliver Davies in 1951 that Warren had lost the skull. Material apparently
associated with the grave included red-burnished potsherds, 13 disc-beads, and two
carved bone artifacts that are possibly broken snuff spoons (identication courtesy
of Anitra Nettleton and Tim Maggs) (Fig. 8). An AMS determination from the femur
yielded a 1σ date-range of cal. AD 1500–1590, 1610–1630 and a δ13C value of −7.0‰
(Beta-398220) (Schlebusch et al. 2017) (Table 5).
Skeletal analysis
The remains include the mandible, frontal bone, left and right parietal bones and
occipital bone of the cranium, all disarticulated. The mandible (Fig. 9) is small and
gracile with a pointed chin. The glabella region is smooth and the supraorbital margins
sharp. These characteristics indicate a female individual. The left humerus, left femur
and left tibia could not be used for metric sex determination as the proximal and distal
epiphyses were damaged postmortem; their incompleteness also meant that stature
could not be estimated.
Some closure of the cranial sutures is evident, although they are not completely
obliterated. As the cranial elements are separated, closure is difcult to judge. No
age estimation could be made, but the long bones appear to be of adult size and the
few teeth that are present are permanent with moderate to advanced dental wear.
The advanced tooth loss and mandibular atrophy suggest that this was an older adult
individual. Extensive wear on the teeth make them difcult to side, but it is possible
to distinguish tooth types. There is one lower molar, one lower rst incisor, two upper
incisors (possibly rst and second), as well as one upper premolar. A carious lesion is
present (on the neck of the root of the lower molar) and there are several abscesses.
The advanced attrition did not allow for dental measurements.
Genetic prole
This individual had the XX chromosome composition of a female, supporting the
results from the skeletal analysis. Her L3e1b2 mitochondrial haplogroup is common in
Bantu speakers (Salas et al. 2002; Soares et al. 2012; Schlebusch et al. 2013, Schlebusch
Fig. 9. Mandible of the Mfongosi skeleton.
Fig. 8. Mfongosi disc-beads and bone artifacts, possibly broken snuff spoons. Scale in cm.
et al. 2017). Her autosomal DNA indicated 13% admixture from Khoe-San (Table 3).
She carried the malaria-resistant Duffy null allele, and was possibly heterozygous C/T
(genome coverage in this region is 5; 4 of the 5 alleles were C (i.e. Duffy null) and one
was T) for this genetic variant and homozygous for the ATP2B4 gene variant that
has been associated with protection against malaria. Taken together, she had a strong
genetic resistance to malaria. The individual did not carry any of the known genetic
variants linked to lactose tolerance and therefore she was likely lactose intolerant. She
was homozygous for the A allele at the APOL1 gene SNP rs73885319, which does
not confer resistance to African sleeping sickness; she therefore harboured the non-
protective allele at this locus.
Newcastle 2007/006
Archaeological and curatorial background
Employees of the Drakensville Berg Resort discovered the remains of this individual
in a disturbed grave in a rock shelter on Oliviershoek Pass in the upper uThukela
basin. The skull and most of the long bones were missing. Annie van de Venter
and Louise van Heerden excavated the remaining bones in 2002, and they are now
curated at the KwaZulu-Natal Museum (acc. no. 2007/006). The grave had been dug
into decomposing sandstone bedrock on the shelter oor. Associated archaeological
material includes a lower grindstone and a shale slab, which formed part of a lining
within the grave that suggests similarity with the graves that Wells (1933b) describes.
There are rock paintings on the shelter walls; other archaeological features close by
include stone walling, a stone cairn, and some Middle and Later Stone Age artifacts
(Van Heerden & Van de Venter 2002; KwaZulu-Natal Museum records). An AMS
determination from the petrous portion of a temporal bone fragment yielded a date-
range of cal. AD 1450–1500, 1590–1610 and a δ13C value of −11.7‰ (Beta-398221)
(Schlebusch et al. 2017) (Table 5).
Skeletal analysis
The remains comprise a fragmented right parietal bone, left temporal bone (Fig. 10),
left inferior-lateral orbital rim fragment, right superior-lateral orbital rim fragment, both
clavicles, both scapulae, left and right os coxae, both patellae, right bula, four cervical
vertebrae, four thoracic vertebrae, ve lumbars, manubrium, six right ribs, seven left
ribs, one rib fragment, 10 foot bones along with 10 metatarsals and 10 foot phalanges,
three carpals, nine metacarpals and eight hand phalanges.
Small mastoid processes and sharp supraorbital margins on the cranium, as well as
a wide greater sciatic notch, preauricular sulcus and a well-dened ventral arc suggest
a female individual. The presence of a well-developed preauricular sulcus suggests
that this woman had borne at least one child. The scapular height (119 mm) and
scapular breadth (91 mm) also indicate a female (Iordanidis 1961). All visible long bone
epiphyses are fused, and all teeth are permanent, indicating an adult individual. The
pubic symphysis morphology is consistent with phase 5 (25 to 83 years) with a mean
age of 48.1 years. Advanced dental wear also suggests that the individual was of mature
age at death, most likely between 40 and 59 years (Falys & Lewis 2011). The bula
indicates an antemortem stature of 132.1±3.17 cm. This is a short stature, according
to Tobias (1961, 1962), very tentatively suggesting a Khoe-San afnity.
The upper left lateral incisor, upper right rst molar and lower right third molar are
present. Both molars show advanced dental wear, whereas the incisor displays very
little wear. The upper right rst molar has caries on both the mesial (caries appears
to penetrate the pulpal space) and distal (no penetration of pulpal space) sides of
the crown. The lower third molar has caries on the buccal side, but the lesion does
not appear to penetrate the pulpal space. Vertebral osteophytes occur on the lumbar,
thoracic and cervical vertebrae, with the lumbars most severely affected. No other
pathology or trauma was present.
Genetic prole
This individual had the XX chromosome composition of a female, supporting the
results from the skeletal analysis. Her L3e2b1a2 mitochondrial haplogroup is commonly
associated with Bantu speakers (Salas et al. 2002; Soares et al. 2012; Schlebusch et al.
2013; Schlebusch et al. 2017). Her autosomal DNA prole indicates 13.3% Khoe-San
admixture (Table 3). She was homozygous C for the malaria resistance Duffy null
allele, homozygous FY*B for the Duffy FY*A/B locus, and had two copies of the
ATP2B4 malaria-protective gene variant, which would have provided her with very
good resistance to malaria. She was also homozygous for the G variant at the APOL1
gene SNP rs73885319 (the G1 allele according to Genovese et al. 2010), which provides
resistance to African sleeping sickness. The individual did not carry any of the known
genetic variants linked to lactose tolerance and she was therefore likely lactose intolerant.
Principal component (Fig. 11B) and clustering analyses (Fig. 11D) show that the Eland
Cave, Mfongosi, Newcastle and Champagne Castle individuals cluster genetically with
Bantu-speaking populations. The rst two principal components (Fig. 11B) differentiate
speakers of Niger-Congo (including Bantu) languages (red and pink) with a West African
origin from speakers of Afro-Asiatic and Nilo-Saharan languages (brown) with an East
African origin, on PC1, and people of Khoe-San descent in southern Africa (purple)
on PC2. The four women in our study clearly cluster with the Niger-Congo-speaking
Fig. 10. (Left) Temporal bone and (right) fragments of the posterior part of the right parietal bone of
the Newcastle 2007/006 skeleton.
group, especially with its Bantu-speaking component. In the clustering analysis (Fig.
11D), they belong securely to the Bantu-language cluster (pink) at the level where 5 and
6 clusters are allowed (K=5 and 6), with some evidence of admixture from the Khoe-
San cluster (purple). At the level where 7 clusters are allowed (K=7), South African
Bantu speakers fall into their own cluster (yellow) and the four women group with
them. This South African Bantu-speaker component occurs in smaller proportions in
Bantu-speaking populations in other parts of Africa, most notably in the East African
countries of Kenya, Uganda, Rwanda and Burundi (Figs 11C, D). These results are
consistent with a common origin for Bantu speakers of East and South Africa in the
Nigeria-Cameroon borderlands.
The APOL1 G1 allele carried by the Eland Cave and Newcastle women confers
moderate resistance to Trypanosoma brucei gambiense-induced sleeping sickness. T. b.
gambiense occurs in tropical west-central Africa and causes chronic sickness, but in
the presence of the G1 allele (rs73885319) it can remain latent for years, permitting
reproduction. The Newcastle woman’s homozygosity, however, would have predisposed
her to chronic kidney disease. South Africa lies largely south of the sleeping-sickness
zone, so that selective pressure for the APOL1 G1 allele likely eased here (the G1 allele
offers no protection against the eastern and southern African parasite T. b. rhodesiense)
(Genovese et al. 2010; Cooper et al. 2017).
The likely presence of at least one Duffy null allele in all four women (some with
low sequence coverage in the region) similarly points to a northern ancestry, as over
90% of people in east and west-central Africa are Duffy negative (Guerra et al. 2010).
Duffy negativity possibly arose around 50 000 years ago due to selection pressure
from the Plasmodium vivax malaria parasite (McManus et al. 2017), which consequently
declined to near extinction in tropical west-central Africa as resistance spread through
human populations there (Carter & Mendis 2002: 573, see correction in vol. 16: 173;
Culleton & Carter 2012). Southern African hunter-gatherers apparently did not evolve a
similar genetic protection against malaria and sleeping sickness (McManus et al. 2017).
The Champagne Castle individual carried the Khoe-San-associated mitochondrial
L0d2a1a haplogroup (Schlebusch et al. 2017). Moreover, her height and some of her
cranial features tentatively reect Khoe-San ancestry. The FORDISC analysis, however,
does not support a Khoe-San origin for the Champagne Castle woman; the autosomal
genetic analysis conrms this result.
A f4-ratio test (Patterson et al. 2012) shows that Khoe-San admixture for the four
Iron Age women ranges from ~5 to 13%, less than the ~17–22% for Bantu speakers
of South Africa today (Table 3, Figs 11A, D). This result is not necessarily signicant,
because the four women cannot encompass all possible variation that may have existed
300–600 years ago in southern Africa. Similarly, the modern comparative sample does
not represent the full range of genetic variation in living people.
To explore the demographic history of the Iron Age women further, we analysed the
decay of admixture-induced linkage disequilibrium (LD) (Loh et al. 2013). To increase
sample size we grouped the four Iron Age individuals together and tested whether
their ancestry could be represented by an admixture between a West African ancestral
source (Yoruba) and southern African hunter-gatherer ancestry (Ju|’hoansi). Their
weighted LD decay exhibits a similar trend reported for modern South African Bantu
speakers, reecting a shared admixture history (Fig. 11B). Modern samples indicate
Fig. 11. (A) Geographical location of the samples used in this study. (B) Principal component analysis showing the genetic afnities of individuals in the merged
dataset. (C) Geographic distribution of the southeast Bantu-speaker (SEB) ancestral component (K=7).
Fig. 11. (D) Cluster analysis for K=5 to K=7.
ALDER (Admixture-induced linkage disequilibrium for evolutionary relationships) dates for admixture
events between southeastern Bantu-speaking groups and local hunter-gatherers and for Karretjie
people. * = average age of males and females across all their children, and not age of female at rst
birth. ** = a mixed group of people with Sotho, Zulu and Tswana ancestry, rst described in
Schlebusch et al. 2012.
Admixture dates in
SD in
Admixture dates in years
(30 years per generation)*
SD in
Karretjie 33.85 0.29 116 38.7
Sotho 25.76 1.66 773 49.8
Zulu 24.18 1.66 725 49.8
Bantu-speakers** 25.65 1.45 770 43.5
an admixture event 23–26 generations ago (Table 4) (Fenner 2005). Unfortunately,
the decay rates of the four Iron Age individuals were inconsistent (probably due to
small sample size and possible variation in the timing of admixture events associated
with each), which did not allow for a dating of the mixture event. Nonetheless,
the similarity of the Iron Age admixture curve to the admixture curves of modern
southeast Bantu speakers from South Africa (and its contrast to the admixture curve
for ‘Karretjie People’, of Khoe-San descent) is evidence of their shared admixture
history (Table 4, Fig. 11).
The Iron Age sequence of KwaZulu-Natal suggests three major episodes of settlement
by Bantu-speaking agriculturists, all following their initial expansion from the Nigeria-
Cameroon borderlands (Fig. 12). The rst (Urewe Tradition, Kwale Branch), 1 600
years ago, derived from the southward spread of agriculturists along the east coast
from Kenya-Tanzania (Klapwijk 1974; Maggs 1980). On current evidence, the second
(Kalundu Tradition) followed shortly afterwards, from the northern Angola-southern
D.R. Congo region (Huffman 1989: 161; Evers 1988: 106–11). Anthropological,
linguistic and some ceramic data suggest a third episode of settlement (Urewe Tradition,
Blackburn Branch) in the 11th century, with the spread of Nguni-speaking agriculturists
from the Great Lakes region in East Africa (Huffman 1989: 173–8, 2004; Hammond-
Tooke 2004). Nguni languages form a subset of the Southern Bantu language cluster
that includes Swati, Zulu and Xhosa.
The second-millennium dates of the four skeletons associate them rmly with the
Blackburn Branch. In KwaZulu-Natal, archaeologists divide the Blackburn Branch into at
least three phases based on changes in ceramic style (Fig. 12): Blackburn (AD 1050–1300),
Moor Park (AD 1300–c. 1680) and Nqabeni (AD c. 1680–1850). In addition, Huffman
(2007: 431, 443–4) argues that Ntsuanatsatsi (AD 1400–1650) in the northeast Free
State is also part of the Blackburn sequence. Moor Park thus provides the archaeological
context for the four women.
The Moor Park phase is distinguished by the rst expansion of agriculturists in
southern Africa into higher-altitude interior grasslands. Agriculturists had until then
favoured wooded lowland environments for settlement. In the grasslands of the
uThukela basin, stonewalled Moor Park sites occur on top of steep-sided hills or spurs
(Davies 1974; Maggs 1984). These defensive locations suggest a time of social turmoil,
possibly caused by diminished agricultural productivity associated with the onset of the
Little Ice Age around AD 1300 (Holmgren et al. 1999; Tyson et al. 2000; Vogel et al.
2001; Whitelaw 2004; cf. Hall 1976; Garstang et al. 2014 for details of a more recent
cold event). Such stress might have provided the initial prompt for movements to the
interior. Moor Park and then Ntsuanatsatsi in the Free State represent the earliest of
these movements (Huffman 2007: 444).
Three hilltop sites (Moor Park, Sewula Gorge and Ntomdadlana) date to the 14th
century (Table 5), contemporaneous with the beginning of the Little Ice Age. Dates
from a fourth, iGujwana, range from the mid-15th to mid-17th centuries. A possible
maize grindstone recovered from a hut platform at iGujwana supports this later
assessment and suggests an occupation after the 1540s, when Portuguese ivory traders
began visiting Delagoa Bay. The Moor Park dates thus point to pulses of social tension,
perhaps associated with alternating warm and cool periods within the Little Ice Age.
These circumstances shaped the lives of our four women. If they relied on amasi
(soured cattle milk) as a staple food (e.g. Bryant 1967: 270), the more negative δ13C
values for the Eland Cave, Newcastle and Champagne Castle burials (Table 5) on
the inland edge of the Moor Park facies probably reect the use of C3 pasture in the
Maloti-Drakensberg, and possibly less reliance on millet and sorghum in these cooler
conditions. The less negative value for the Mfongosi woman probably reects the C4
pasture of the lower uThukela valley (e.g. Ribot et al. 2010).
UREWE,BlackburnBranch LateIronAge
AD Phase TRADITION,Branch Period
Fig. 12. Iron Age archaeological sequence of KwaZulu-Natal (see Maggs 1989; Huffman 2007: 126–30,
154–69, 304–15). Moor Park phase in grey shading.
Radiocarbon dates of the four skeletons (upper level) and of four excavated Moor Park hilltop sites (lower level). For consistency, all dates calibrated on
Calib Rev 7.0.4 (Stuiver & Reimer 1993) with SHCal13 (Hogg et al. 2013), rounded down (start of range) and up (end of range) to the nearest 10.
number Date BP 1σ calibration AD 2σ calibration AD δ13CMaterial Reference
Eland Cave Beta-398219 480±30 1430–1460 1410–1500 −9.4 bone Schlebusch et al.
Castle Beta-398218 310±30 1510–1550, 1560–1570, 1620–1660 1500–1600, 1610–1670, 1790 −11.0 bone Schlebusch et al.
Mfongosi Beta-398220 360±30 1500–1590, 1610–1630 1480–1650 −7.0 bone Schlebusch et al. 2017
Newcastle Beta-398221 430±30 1450–1500, 1590–1610 1440–1520, 1540–1630 −11.7 bone Schlebusch et al. 2017
Moor Park
Pta-849 750±50 1260–1320, 1350–1390 1220–1330, 1340–1390 −25.5 charcoal
Davies 1974Pta-850 660±50 1300–1400 1280–1410 −25.3 charcoal
Pta-853 600±50 1320–1350, 1380–1430 1300–1450 −25.4 charcoal
Sewula Gorge Pta-8370 710±50 1280–1320, 1350–1390 1270–1400 −25.6 charcoal Whitelaw 2004
Pta-8372 660±50 1300–1400 1280–1410 −25.6 charcoal
Ntomdadlana Pta-8697 630±50 1310–1360, 1380–1410 1290–1430 −6.8 bone Whitelaw 2015
iGujwana Pta-8101 390±50 1460–1520, 1540–1630 1450–1640 −24.5 wood: hut-
pole stub Whitelaw 2004
Pta-8335 360±50 1500–1600, 1610–1640 1460–1650 −26.8 charcoal
Rock shelters across the uThukela basin contain evidence of relationships between
agriculturists and hunter-gatherers during the second millennium AD. Later Stone
Age deposits contain potsherds and iron artifacts, as well as bone objects probably
shaped with iron blades (Maggs & Ward 1980; Mazel 1986a, b, 1990). There is limited
material evidence for what hunter-gatherers offered in return. Linguistic (Herbert
1990) and other evidence suggests that agriculturist men sometimes took hunter-
gatherer women as wives. The adoption of aspects of the hunter-gatherer trance
dance into divination performances (Hammond-Tooke 1998, 2002) probably began
in Moor Park times (Whitelaw 2015: 154–6), and herbalists possibly interacted with
hunter-gatherers while seeking wild medicinal ingredients. An ostrich-eggshell necklace
cached together with strings of some 5 000 glass beads at Sibudu Cave between 1450
and 1660 (Wood et al. 2009) possibly came from such interactions: ostrich-eggshell
beads commonly occur in Later Stone Age deposits of the second millennium AD,
but rarely in Late Iron Age deposits.
The three rock-shelter burials (Eland Cave, Champagne Castle and Newcastle) form
part of a larger set found in the upper uThukela basin. Murray (1933; Wells 1933a,
b) describes the skeletons of at least eight (from six shelters), and Beaumont (1967)
the excavation of one individual. Their analyses assign all the burials to Iron Age
communities. The skeletons cover a range of ages, including adult men and women,
juveniles, and a small child. They are undated, but the stone-lined, corbelled-hut-like
grave structures possibly indicate contemporaneity with the grave of the Newcastle
woman. Also, ve possible Indian red glass beads beside the pelvis of a juvenile
perhaps came from a beaded loin belt (see Wells 1933c: 190–1; also Wood et al. 2009:
249–51 on purported earthenware beads). They suggest broad contemporaneity with
the Sibudu bead cache, which Indian reds dominate, and with Portuguese observations
of red-coloured Indian beads in the 1500s (Theal 1898, I: 225; 1898, II: 303). After
1660, traders ooded African markets with a cheap Venetian substitute, the Indian-
red-on-green (Wood 2008: 185; Wood et al. 2009: 251–2).
In Iron Age communities, all death was a source of potent supernatural pollution
that threatened the health and welfare of people, their things, and potentially the
country (e.g. Ngubane 1977: chapter 5). Its amelioration demanded correct mortuary
practice. Typically, people were buried in places with which they were associated in life.
Thus, homestead heads and their agnatic male relations were buried in the cattle pen,
married women in the courtyard behind their houses, young children in the public space
between houses and cattle pen, while infant graves were commonly closely associated
with houses (Huffman & Murimbika 2003; Boeyens et al. 2009). In a sense, the grave
was “the terrestrial home of the ancestor” (Raum 1973: 328).
Local practices and the varying causes of death resulted in departures from the basic
pattern. Some chiefs were buried in forests, source of the medicines and ferocious
animals that gave them their chiey aura (Hunter 1936: 396–7; Bryant 1967: 720;
Hammond-Tooke 1975). The Hlubi of the upper uMzinyathi buried their chiefs in rock
shelters in the hills (Webb & Wright 1979: 13, 1986: 21), as did the Swazi (Kuper 1947:
87, 178), and possibly communities around the middle White iMfolozi (Zwane n.d.).
Rainmakers too were buried in rock shelters, to protect their bodies from direct rainfall
(Muzi Msimanga pers. comm., August 2018). This last practice offers an interpretative
possibility for the uThukela shelter burials, but we doubt it applies. Rainmakers were
most commonly men and always older individuals, post-menopausal in the case of
women (Berglund 1976: 63; Murimbika 2006: 195). Neither the three female burials
we describe, nor the wider set, properly t rainmaking practice.
Many rock shelters in KwaZulu-Natal contain material deposited by agriculturists. In
three—Sibudu, Mzinyashana and iNkolimahashi—the deposits are substantial (Mazel
1997, 1999; Wood et al. 2009), so they suggest another social context for the rock-
shelter burials: the rock shelters were home, possibly for impoverished agriculturists
(e.g. Mazel 1997: 32; Whitelaw 2009: 154–5). But only two of the eight ‘burial shelters’
(including Eland Cave) that the Wells team and Beaumont investigated contained
more than sparse Iron Age habitation debris. The principal exception, Kaybar’s Cave
in the Cathkin Park-Champagne Castle area, contained an almost entirely agriculturist
accumulation, including the graves of a man, a woman and a child. The nds (mirror,
mother-of-pearl button, abundant grindstones, large European-made white bead)
indicate that the deposit post-dates Moor Park times (Murray 1933; Wells 1933a, c),
but its contemporaneity with the burials is uncertain (Wells 1933b: 198). Similarly,
whiteware sherds, a aked glass scraper, a green glass bead, and a Royal Marine Artillery
button from Eland Cave (Stein 1933: 169; KwaZulu-Natal Museum collection) suggest
a terminal deposit in the 19th through to the early 20th century, 400–500 years after
the death of the Eland Cave woman. Thus, current data offer poor support for the
home-base hypothesis.
Throughout the entire region, from eSwatini (Swaziland) to the Eastern Cape and
across Lesotho into the eastern Free State, certain deaths received special treatment.
Victims of violence were left or covered where they died, or buried at a distance from
the homestead. Relatives feared that proximity to the grave would bring the same death
to others in the homestead. The same principle applied to lightning and drowning
victims—improper mortuary arrangements encouraged further such deaths (Soga 1931:
320–1; Hunter 1936: 228, 301; Kuper 1947: 125, 180; Ashton 1952: 104; Hammond-
Tooke 1962: 230; Webb & Wright 2001: 198). Lightning victims were buried away from
their homesteads in cool places that dampened the dangerous heat that so dramatically
caused their deaths. This concern for cooling extended even to the colour of the soil
in which the grave was set: red earth made the victim hotter. Bodies left to ‘burn’ in
death could bring long-term misfortune to relatives and, worse, drive away rain and so
threaten the agricultural base upon which communities depended (Berglund 1976: 41).
Lightning poses a frequent and unpredictable threat in KwaZulu-Natal today
(Erasmus 2015). It was surely the same in the past for people living on elevated ground.
We suggest that in the socially stressed, cold and generally drier conditions of the Little
Ice Age, people adhered strictly to practices believed to ward off real and supernatural
dangers and, especially, ensure the regularity of the annual environmental cycle. A failure
to do so is likely to have resulted in community-wide repercussions if the rains failed or
lightning struck other homesteads. Further, we suggest that they founded rock-shelter
burial as a local strategy to deal with lightning and other similarly ‘hot’ deaths, such as
the violent one the Champagne Castle woman possibly suffered.
In ethnographic accounts, damp earth on river banks provided an appropriate site for
the graves of lightning victims (Ashton 1952: 104; Berglund 1976: 40–1). We suggest
that in Moor Park times, some rock shelters in the Maloti-Drakensberg foothills were
similarly appropriate. Many rainmakers worked and stored their medicines in rock
shelters because of the cool, shady environment they provided. This preference was
widespread and has considerable antiquity (Callaway 1870: 393; Aukema 1989; Berglund
1976: 60; Murimbika 2006: 194). Apart from shade, the stone of shelter walls is cold,
and the deposit on the oors grey like ash—a cooling substance. In contrast, most
Moor Park homestead sites in the grasslands are built of dolerite, a rock type that gives
rise to a boldly red earth.
Wells (1933a) and Beaumont (1967) link the shelter burials to Zizi communities whose
settlement in the upper uThukela basin predates 1800. Descendants still live there.
Zizi families most likely lived in homesteads like Mgoduyanuka near Bergville, which
dates to between 1670 and 1850 (Maggs 1982). The three shelter burials in this study
predate Mgoduyanuka. Moreover, Mgoduyanuka-type sites are organized differently
from Moor Park sites, with a centre-sides rather than a front-back arrangement
(Huffman 2007: 33, 41). Mgoduyanuka sites might result from a movement of Zizi
ancestral communities into the upper uThukela basin in the mid- to late 1600s. Their
arrival might have extinguished some local cultural practices, including the treatment
of ‘hot’ deaths with shelter burial, although the principle lives on for rainmakers. This
topic remains for future research.
We can say little about the Mfongosi grave. The artifacts found with it and the
description of its location suggest the woman was buried within her homestead. The
grave site falls into what was the Mkhize chiefdom around 1800, with the iMfongosi
River possibly marking its boundary with the neighbouring Chunu chiefdom (Webb &
Wright 1982: 6, 2014: 147). Perhaps the Mfongosi woman lived in an ancestral form
of the Mkhize chiefdom.
All four sets of Iron Age human remains are incomplete or poorly preserved, which
complicated physical analysis. Sex could be determined osteologically in only three of
the four individuals, and we assessed all three as female, which was corroborated by
their DNA; DNA analysis also conrmed the fourth individual as female. All were of
adult age. The stature estimates of the Champagne Castle and Newcastle individuals
are shorter than expected for black African groups, being more consistent with San/
Khoe-San stature estimates reported in the literature (Tobias 1961, 1962; Pfeiffer &
Sealy 2006; Pfeiffer 2012).
In general, their teeth show considerable wear. The posterior teeth seem more
worn than the anterior teeth, suggesting that the teeth were used mostly for normal
masticatory functions and not as tools. Many of the teeth show caries or abscessing.
Most of the caries occur around the cemento-enamel junction, which suggests that
some dental cleaning procedures predisposed the individuals to root-neck caries.
Only the Champagne Castle woman’s skull was complete enough to attempt
a craniometric assessment of population afnity. Assessment of ancestry using
FORDISC indicated that she grouped rst with West Africans, which is consistent with
the genetic analysis demonstrating that all four women have ancestral roots in West
Africa. It is also consistent with linguistic, archaeological and genetic interpretations
regarding the origins of South African Iron Age agriculturalists.
Unfortunately, the FORDISC database contains a limited number of comparative
samples, and it may be worthwhile to do comparisons with other more sophisticated
databases in future. Morris (2017) made the point that fragmentary skeletons are often
used for DNA analysis, in an attempt to be conservative with bone material. However,
this precludes the possibility of matching up craniometrics and genetic data. We suggest
that this issue is worthy of consideration for future research design, especially where
geneticists are able to take minuscule samples, causing minimum damage, from which
they are able to reconstruct the full genomes of sampled individuals (Schlebusch et
al. 2017; Lombard et al. 2018).
Schlebusch et al. (2017: table S25) reported on genes coding for specic interesting
traits associated with the four Iron Age women. These include gene regions relating
to resistance to malaria and African sleeping sickness, as well as lactose intolerance.
The Duffy null allele, which has a strong protective effect against malaria, was found
in the women of Eland Cave, Mfongosi and Newcastle (Schlebusch et al. 2017: table
S25), and the woman from Champagne Castle probably had some genetic protection
against this illness. Our work therefore suggests that strong malaria-protective variants
existed in the Iron Age individuals discussed here. The same could not be said about
the Stone Age hunter-gatherers from KwaZulu-Natal assessed by Schlebusch and
colleagues (2017; Lombard et al. 2018; Pfeiffer et al. 2019). Furthermore, having at
least one G allele for the APOL1 gene SNP rs73885319 (referred to as the G1 allele
by Genovese et al. 2010) confers resistance to T. b. gambiense sleeping sickness of west-
central Africa. The Eland Cave woman was heterozygous for the polymorphism that
offers resistance to African sleeping sickness, while the woman from Newcastle was
homozygous for an alternative variant (Schlebusch et al. 2017: table S25).
The regions in which Bantu-speaking agriculturists settled during the last 2 000
years, as well as their common point of origin in West Africa, are known to be areas
where both malaria-bearing mosquitos and sleeping sickness-carrying tsetse ies thrive
(Aufderheide & Rodríguez-Martín 1998; Snow et al. 2005; Krafsur 2009; Schlebusch &
Jakobsson 2018). It is therefore not surprising that our four Iron Age women, whose
genetic roots lie in West Africa, all have varying degrees of resistance to malaria, and
two of them against sleeping sickness. Local Stone Age hunter-gatherers, on the other
hand, who were not equally exposed, did not develop similar genetic protection against
these diseases.
None of the Iron Age women (with relevant data) exhibited lactase persistence;
they would therefore not have been able to consume fresh milk without experiencing
indigestion. The absence of the lactase persistence trait in modern Bantu speakers
is well known, and although they do consume animal milk products, it is usually in
fermented form that aids in its digestion (e.g. Rocha 2012; Lombard & Parsons 2015).
This is in contrast to East African herding communities who arrived in southern Africa
only a few centuries before the Iron Age groups, and who brought with them the East
African genetic variant for lactase persistence now prevalent in most Khoekhoe and
Khoekhoe-descendent people of the region (e.g. Breton et al. 2014).
From an archaeological perspective, the three shelter burials reveal unusual treatment,
consistent with a worldview in which ‘hot’ deaths required ‘cool’ burial places. This
practice is still maintained in the upper uThukela basin, but today it is reserved for
rainmakers only, who on death are buried in cool places sheltered from the elements.
The social strain resulting from the Little Ice Age climatic downturn possibly drew Iron
Age agriculturists into closer, more intimate contact with hunter-gatherers, resulting in
increased Khoe-San admixture rates into Bantu-speaking populations over time. An
increased archaeological sample size may in future help address this issue.
It is nevertheless worth noting that to date there are no records of graves containing
the skeletons of Khoe-San individuals in this region. Despite interaction and genetic
admixture with, and cultural adoptions from hunter-gatherers, and the use, even
appropriation, of hunter-gatherer places for death rituals, the upper uThukela
burials suggest that agriculturist and hunter-gatherer communities maintained their
distinctiveness, at least during the Moor Park period.
With this paper we bring together information from a variety of sources—biological
anthropology, DNA analyses and broad archaeological context—giving the remains
a kind of individual identity that imbues them with humanity. Our approach offers
exciting new opportunities for a deeper understanding of the past and the individuals
who lived in KwaZulu-Natal centuries ago. The contextual richness that we were able
to create is especially encouraging in light of the fragmentary and poorly preserved
nature of the skeletons, as well as their perceived limited archaeological contexts. We
therefore argue that this holistic, inter-disciplinary approach to archaeological human
remains research holds considerable promise for enriching knowledge about similar
small, seemingly insignicant skeletal assemblages, simultaneously contributing to
large-scale interpretations of population movement, interaction and associated socio-
economic circumstances.
Maryna Steyn’s research is funded by the National Research Foundation. Gavin Whitelaw’s research is
funded by the KwaZulu-Natal Museum and the National Research Foundation. Whitelaw acknowledges
useful discussions with Muzi Msimanga, Aron Mazel, Tim Maggs, Judith Sealy and Anitra Nettleton.
Carina Schlebusch is funded by the Swedish Research Council (no. 621-2014-5211). Marlize Lombard is
funded by the African Origins Platform of the National Research Foundation (no. 98815). Neither the
funding bodies nor the discussants necessarily support the authors’ opinions, ndings and conclusions.
Alexander, D.H., Novembre, J. & Lange, K. 2009. Fast model-based estimation of ancestry in unrelated
individuals. Genome Research 19: 1655–64.
Acsádi, G. & Nemeskéri, J. 1970. History of human life span and mortality. Budapest: Akadémia Kiadó.
Asala, S.A., Bidmos, M.A. & Dayal, M.R. 2004. Discriminant function sexing of fragmentary femur of
South African blacks. Forensic Science International 145: 25–9.
Ashton, H. 1952. The Basuto. Reprint, 1955. London: Oxford University Press for the International
African Institute.
Aufderheide, A.C. & Rodríguez-Martín, C. 1998. The Cambridge encyclopedia of human paleopatholog y. Cambridge:
Cambridge University Press.
Aukema, J. 1989. Rain-making: a thousand year-old ritual? South African Archaeological Bulletin 44: 70–2.
Auton, A., Brooks, L.D., Durbin, R.M., Garrison, E.P., Kang, H.M., Korbel, J.O., Marchini, J.L., McCarthy,
S., McVean, G.A. & Abecasis, G.R. 2015. A global reference for human genetic variation. Nature
526: 68–74.
Beaumont, P. 1967. The Brotherton Shelter. South African Archaeological Society 22: 27–30.
Behr, A.A., Liu, K.Z., Liu-Fang, G., Nakka, P. & Ramachandran, S. 2016. Pong: fast analysis and visualization
of latent clusters in population genetic data. Bioinformatics 32: 2817–23.
Berglund, A.-L. 1976. Zulu thought patterns and symbolism. Reprint, 1989. Bloomington: Indiana University
Boeyens, J., Van der Ryst, M., Coetzee, F., Steyn, M. & Loots, M. 2009. From uterus to jar: the signicance
of an infant pot burial from Melora Saddle, an early nineteenth-century African farmer site on the
Waterberg Plateau. Southern African Humanities 21: 213–38.
Breton, G., Schlebusch, C.M., Lombard, M., Sjödin, P., Soodyall, H. & Jakobsson, M. 2014. Lactase
persistence alleles reveal partial East African ancestry of southern African Khoe pastoralists. Current
Biology 24: 852–8.
Brooks, S.T. & Suchey, J.M. 1990. Skeletal age determination based on the os pubis: a comparison of the
Acsádi-Nemeskéri and Suchey-Brooks Methods. Human Evolution 5: 227–38.
Bryant, A.T. 1967 (1949). The Zulu people as they were before the white man came. 2nd edition. Pietermaritzburg:
Shuter & Shooter.
Buikstra, J.E. & Ubelaker, D.H. 1994. Standards for data collection from human skeletal remains. Arkansas:
Arkansas Archaeological Survey.
Callaway, H. 1870. The religious system of the amaZulu. Facsimile reprint, 1970. Cape Town: C. Struik.
Carter, R. & Mendis, K.N. 2002. Evolutionary and historical aspects of the burden of malaria. Clinical
Microbiology Reviews 15: 564–94.
Cooper, A., Ilboudo, H., Alibu, V.P., Ravel, S., Enyaru, J., Weir, W., Noyes, H., Capewell, P., Camara, M.,
Milet, J., Jamonneau, V., Camara, O., Matovu, E., Bucheton, B. & MacLeod, A. 2017. APOL1 renal
risk variants have contrasting resistance and susceptibility associations with African trypanosomiasis.
eLife 6: e25461.
Culleton, R. & Carter, R. 2012. African Plasmodium vivax: distribution and origins. International Journal for
Parasitology 42: 1091–7.
Dart, R.A. 1924. The Rooiberg cranium. South African Journal of Science 21: 556–68.
Davies, O. 1957. A missing skull of early type from Zululand. Man 57: 48.
Davies, O. 1974. Excavations at the walled Early Iron-Age site in Moor Park near Estcourt, Natal. Annals
of the Natal Museum 22: 289–323.
De Villiers, H. 1968. The skull of the South African Negro: a biometrical and morphological study. Johannesburg:
Witwatersrand University Press.
DiGangi, E.A., Bethard, J.D., Kimmerle, E.H. & Konigsberg, L.W. 2009. A new method for estimating
age-at-death from the rst rib. American Journal of Physical Anthropology 138: 164–76.
Erasmus, J. 2015. KZN is a lightning magnet. The Witness, 7 December, <
News/kzn-is-a-lightning-magnet-20151206>; viewed 23 August 2018.
Evers, T.M. 1988. The recognition of groups in the Iron Age of southern Africa. PhD thesis, University of the
Falys, C.G. & Lewis, M.E. 2011. Proposing a way forward: a review of standardization in the use of age
categories and aging techniques in osteological analysis. International Journal of Osteoarchaeology 21:
Fenner, J.N. 2005. Cross-cultural estimation of the human generation interval for use in genetics-based
population divergence studies. American Journal of Physical Anthropology 128: 415–23.
Garstang, M., Coleman, A.D. & Therrell, M. 2014. Climate and the mfecane. South African Journal of Science
110 (5–6): <>.
Genovese, G., Friedman, D.J., Ross, M.D., Lecordier, L., Uzureau, P., Freedman, B.I., Bowden, D.W.,
Langefeld, C.D., Oleksyk, T.K., Knob, A.U., Bernhardy, A., Hicks, P.J., Nelson, G.W., Vanhollebeke,
B., Winkler, C.A., Kopp, J.B., Pays, E. & Pollak, M.R. 2010. Association of trypanolytic APOL1
variants with kidney disease in African-Americans. Science 329: 841–5.
Guerra, C.A., Howes, R.E., Patil, A.P., Gething, P.W., Van Boeckel, T.P., Temperley, W.H., Kabaria, C.W.,
Tatem, A.J., Manh, B.H., Elyazar, I.R. & Baird, J.K. 2010. The international limits and population
at risk of Plasmodium vivax transmission in 2009. PLoS Neglected Tropical Diseases 4: e774.
Gurdasani, D., Carstensen, T., Tekola-Ayele, F., Pagani, L., Tachmazidou, I., Hatzikotoulas, K., Karthikeyan,
S., Iles, L., Pollard, M.O., Choudhury, A., Ritchie, G.R., Xue, Y., Asimit, J., Nsubuga, R.N., Young,
E.H., Pomilla, C., Kivinen, K., Rockett, K., Kamali, A., Doumatey, A.P., Asiki, G., Seeley, J., Sisay-
Joof, F., Jallow, M., Tollman, S., Mekonnen, E., Ekong, R., Oljira, T., Bradman, N., Bojang, K.,
Ramsay, M., Adeyemo, A., Bekele, E., Motala, A., Norris, S., Pirie, F., Kaleebu, P., Kwiatkowski,
D., Tyler-Smith, C., Rotimi, C., Zeggini, E. & Sandhu, M.S. 2015. The African Genome Variation
Project shapes medical genetics in Africa. Nature 517: 327–32.
Hall, M. 1976. Dendrochronology, rainfall and human adaptation in the Late Iron Age of Natal and
Zululand. Annals of the Natal Museum 22: 693–703.
Hammond-Tooke, W.D. 1962. Bhaca society: a people of the Transkeian uplands, South Africa. Cape Town:
Oxford University Press.
Hammond-Tooke, W.D. 1975. The symbolic structure of Cape Nguni cosmology. In: M.G. Whisson &
M. West, eds, Religion and social change in southern Africa: anthropological essays in honour of Monica Wilson.
Cape Town: David Phillip, pp. 15–35.
Hammond-Tooke, W.D. 1998. Selective borrowing? The possibility of San shamanistic inuence on
southern Bantu divination and healing practices. South African Archaeological Bulletin 53: 9–15.
Hammond-Tooke, W.D. 2002. The uniqueness of Nguni mediumistic divination in southern Africa.
Africa 72: 277–92.
Hammond-Tooke, W.D. 2004. Southern Bantu origins: light from kinship terminology. Southern African
Humanities 16: 71–8.
Herbert, R.K. 1990. The sociohistory of clicks in Southern Bantu. Anthropological Linguistics 32: 295–315.
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, R.H. 2013. SHCal13 southern hemisphere
calibration, 0–50,000 years cal BP. Radiocarbon 55: 1889–903.
Holmgren, K., Karlén, W., Lauritzen, S.E., Lee-Thorp, J.A., Partridge, T.C., Piketh, S., Repinski, P.,
Stevenson, C., Svanered, O. & Tyson, P.D. 1999. A 3 000-year high-resolution stalagmite-based
record of palaeoclimate for northeastern South Africa. The Holocene 9: 295–309.
Howells, W.W. 1973. Cranial variation in man: a study by multivariate analysis of patterns of difference among
recent human populations. Papers of the Peabody Museum of Archaeology and Ethnology, vol. 67.
Cambridge, Ma.: Harvard University Press.
Huffman, T.N. 1989. Ceramics, settlements and Late Iron Age migrations. African Archaeological Review
7: 155–82.
Huffman, T.N. 2007. Handbook to the Iron Age: the archaeology of pre-colonial farming communities in southern
Africa. Scottsville: University of KwaZulu-Natal Press.
Huffman, T.N. & Murimbika, M. 2003. Shona ethnography and Iron Age burials. Journal of African
Archaeology 1: 237–46.
Hunter, M. 1936. Reaction to conquest: effects of contact with Europeans on the Pondo of South Africa. London:
Oxford University Press.
Iordanidis, P. 1961. Determination du sexe par les os du squelette (atlas, axis, clavicule, omoplate, sternum).
Annales de Mdclecine Legale 41: 280–91.
İşcan, M.Y. & Steyn, M. 2013. The human skeleton in forensic medicine. 3rd edn. Springeld: Charles C.
Klapwijk, M. 1974. A preliminary report on pottery from the north-eastern Transvaal, South Africa. South
African Archaeological Bulletin 29: 19–23.
Krafsur, E.S. 2009. Tsetse ies: genetics, evolution, and role as vectors. Infection, Genetics and Evolution 9:
Kunos, C.A., Simpson, S.W., Russell, K.F. & Hershkovitz, I. 1999. First rib metamorphosis: its possible
utility for human age-at-death estimation. American Journal Physical Anthropology 110: 303–23.
Kuper, H. 1947. An African aristocracy: rank among the Swazis. London: Oxford University Press.
Kurki, H. 2005. Use of the rst rib for adult age estimation: a test of one method. International Journal of
Osteoarchaeology 15: 342–50.
Loh, P.-R., Lipson, M., Patterson, N., Moorjani, P., Pickrell, J.K., Reich, D. & Berger, B. 2013. Inferring
admixture histories of human populations using linkage disequilibrium. Genetics 193: 1233–54.
Lombard, M., Jakobsson, M. & Schlebusch, C. 2018. Ancient human DNA: How the sequencing of the
genome of a boy from Ballito Bay changed human history. South African Journal of Science 114 (1–2):
Lombard, M. & Parsons, I. 2015. Milk not meat: the role of milk amongst the Khoe peoples of southern
Africa. Journal of African Archaeology 13: 149–66.
Lombard, M., Wadley, L., Deacon, J., Wurz, S., Parsons, I., Mohapi, M., Swart, J. & Mitchell, P. 2012. South
African and Lesotho Stone Age sequence updated. South African Archaeological Bulletin 67: 123–44.
Lundy, J.K. & Feldesman, M.R. 1987. Revised equations for estimating living stature from long bones of
the South African Negro. South African Journal of Science 83: 54–5.
Maggs, T. 1980. Mzonjani and the beginning of the Iron Age in Natal. Annals of the Natal Museum 24: 111–45.
Maggs, T. 1982. Mgoduyanuka: terminal Iron Age settlement in the Natal grasslands. Annals of the Natal
Museum 25: 83–113.
Maggs, T. 1984. Iron Age settlement and subsistence patterns in the Tugela River Basin, Natal. In: M. Hall,
G. Avery, D.M. Avery, M.L. Wilson & A.J.B. Humphreys, eds, Frontiers: southern African archaeology
today. Oxford: B.A.R., pp. 194–206.
Maggs, T. 1989. The Iron Age farming communities. In: A. Duminy & B. Guest, eds, Natal and Zululand
from earliest times to 1910: a new history. Pietermaritzburg: University of Natal Press and Shuter &
Shooter, pp. 28–48.
Maggs, T. & Ward, V. 1980. Driel Shelter: rescue at a Late Stone Age site on the Tugela River. Annals of
the Natal Museum 24: 35–70.
Mazel, A.D. 1986a. Mgede Shelter: a mid- and late Holocene observation in the western Biggarsberg,
Thukela Basin, Natal, South Africa. Annals of the Natal Museum 27: 357–87.
Mazel, A.D. 1986b. Mbabane Shelter and eSinhlonhlweni Shelter: the last two thousand years of hunter-
gatherer settlement in the central Thukela Basin, Natal, South Africa. Annals of the Natal Museum
27: 389–453.
Mazel, A.D. 1990. Mhlwazini Cave: the excavation of Late Holocene deposits in the northern Natal
Drakensberg, Natal, South Africa. Natal Museum Journal of Humanities 2: 95–133.
Mazel, A.D. 1997. Mzinyashana Shelters 1 and 2: excavation of mid and late Holocene deposits in the
eastern Biggarsberg, Thukela Basin, South Africa. Natal Museum Journal of Humanities 9: 1–35.
Mazel, A.D. 1999. iNkolimahashi Shelter: the excavation of Later Stone Age rock shelter deposits in the
central Thukela Basin, KwaZulu-Natal, South Africa. Natal Museum Journal of Humanities 11: 1–21.
Mazel, A.D. 2009. Unsettled times: shaded polychrome paintings and hunter-gatherer history in the
southeastern mountains of southern Africa. Southern African Humanities 21: 85–115.
McManus, K.F., Taravella, A.M., Henn, B.M., Bustamante, C.D., Sikora, M. & Cornejo, O.E. 2017.
Population genetic analysis of the DARC locus (Duffy) reveals adaptation from standing variation
associated with malaria resistance in humans. PLoS Genetics 13: e1006560.
Molnar, S. 1971. Human tooth wear, tooth function and cultural variability. American Journal of Physical
Anthropology 34: 175–90.
Montinaro, F., Busby, G.B., Gonzalez-Santos, M., Oosthuitzen, O., Oosthuitzen, E., Anagnostou, P.,
Destro-Bisol, G., Pascali, V.L. & Capelli, C. 2016. Complex ancient genetic structure and cultural
transitions in southern African populations. Genetics 205: 303–16.
Morris, A.G. 2017. Ancient DNA comes of age, but still has some teenage problems. South African Journal
of Science 113 (910): <>.
Murimbika, M. 2006. Sacred powers and rituals of transformation: an ethnoarchaeological study of rainmaking rituals
and agricultural productivity during the evolution of the Mapungubwe state, AD 1000 to AD 1300. PhD thesis,
University of the Witwatersrand.
Murray, N.L. 1933. Skeletal remains from rock shelters in Cathkin Park, Natal. Bantu Studies 7 (2): 201–15.
Ngubane, H. 1977. Body and mind in Zulu medicine: an ethnography of health and disease in Nyuswa-Zulu thought
and practice. London: Academic Press.
Ousley, S. & Jantz, R.L. 2012. Fordisc 3 and statistical methods for estimating sex and ancestry. In: D.
Dirkmaat, ed., A companion to forensic anthropology. Malden: Blackwell Publishing Ltd, pp. 311–29.
Pager, H. 1971. The rock art of the Ndedema Gorge and neighbouring valleys, Natal Drakensberg. In: M.
Schoonraad, ed., Rock paintings of southern Africa. South African Association for the Advancement
of Science, Special Publication No. 2, pp. 27–33.
Patterson, N., Price, A.L. & Reich, D. 2006. Population structure and eigenanalysis. PLoS Genetics 2: e190.
Patterson, N., Moorjani, P., Luo, Y., Mallick, S., Rohland, N., Zhan, Y., Genschoreck, T., Webster, T. &
Reich, D. 2012. Ancient admixture in human history. Genetics 192: 1065–93.
Pfeiffer, S. 2012. Conditions for evolution of small adult body size in southern Africa. Current Anthropology
53: S383–94.
Pfeiffer, S., Harrington, L. & Lombard, M. 2019. The people behind the samples: biographical features of
past hunter-gatherers from KwaZulu-Natal who yielded aDNA. Inter national Jour nal of Paleopathology
24: 158–64.
Pfeiffer, S. & Sealy, J. 2006. Body size among Holocene foragers of the Cape ecozone, southern Africa.
American Journal of Physical Anthropology 129: 1–11.
Phenice, T.W. 1969. A newly developed visual method of sexing the os pubis. American Journal of Physical
Anthropology 30: 297–302.
Phillipson, D.W. 1977. The later prehistory of eastern and southern Africa. London: Heinemann.
Purcell, S., Neale, B., Todd-Brown, K., Thomas, L., Ferreira, M.A., Bender, D., Maller, J., Sklar, P., De
Bakker, P.I., Daly, M.J. & Sham, P.C. 2007. PLINK: a tool set for whole-genome association and
population-based linkage analyses. American Journal of Human Genetics 81: 559–75.
Raum, O.F. 1973. The social functions of avoidances and taboos among the Zulu. Berlin: Walter de Gruyter.
Raxter, M.H., Auerbach, B.M. & Ruff, C.B. 2006. Revision of the Fully technique for estimating statures.
American Journal of Physical Anthropology 130: 374–84.
Ribot, I., Morris, A.G., Sealy, J. & Maggs, T. 2010. Population history and economic change in the last
2000 years in KwaZulu-Natal, RSA. Southern African Humanities 22: 89–112.
Rocha, J. 2012. The evolution of lactase persistence. Antropologia Portuguesa 29: 121–37.
Salas, A., Richards, M., De la Fe, T., Lareu, M.-V., Sobrino, B., Sánchez-Diz, P., Macaulay, V. & Carracedo,
A. 2002. The making of the African mtDNA landscape. American Journal of Human Genetics 71:
Schlebusch, C.M., De Jongh, M. & Soodyall, H. 2011. Different contributions of ancient mitochondrial
and Y-chromosomal lineages in ‘Karretjie people’ of the Great Karoo in South Africa. Journal of
Human Genetics 56: 623.
Schlebusch, C. & Jakobsson, M. 2018. Tales of human migration, admixture, and selection in Africa.
Annual Review of Genomics and Human Genetics 19: 405–28.
Schlebusch, C.M., Lombard, M. & Soodyall, H. 2013. MtDNA control region variation afrms diversity
and deep sub-structure in populations from southern Africa. BMC Evolutionary Biology 13: 56.
Schlebusch, C.M., Malmström, H., Günther, T., Sjödin, P., Coutinho, A., Edlund, H., Munters, A.R.,
Vicente, M., Steyn, M., Soodyall, H., Lombard, M. & Jakobsson, M. 2017. Southern African ancient
genomes estimate modern human divergence to 350,000 to 260,000 years ago. Science 6363: 652–5.
Schlebusch, C.M., Prins, F., Lombard, M., Jakobsson, M. & Soodyall, H. 2016. The disappearing San of
southeastern Africa and their genetic afnities. Human Genetics 135: 1365–73.
Schlebusch, C.M., Skoglund, P., Sjödin, P., Gattepaille, L.M., Hernandez, D., Jay, F., Li, S., De Jongh, M.,
Singleton, A., Blum, M.G., Soodyall, H. & Jakobsson, M. 2012. Genomic variation in seven Khoe-
San groups reveals adaptation and complex African history. Science 6105: 374–9.
Scott, E.C. 1979. Dental wear scoring technique. American Journal of Physical Anthropology 51: 213–7.
Snow, R.W., Guerra, C.A., Noor, A.M., Myint, H.Y. & Hay, S.I. 2005. The global distribution of clinical
episodes of Plasmodium falciparum malaria. Nature 434: 214–7.
Soares, P., Alshamali, F., Pereira, J.B., Fernandes, V., Silva, N.M., Afonso, C., Costa, M.D., Musilová, E.,
Macaulay, V., Richards, M.B., Cerny, V. & Pereira, L. 2012. The expansion of mtDNA haplogroup
L3 within and out of Africa. Molecular Biology and Evolution 29: 915–27.
Soga, J.H. 1931. The Ama-Xosa: life and customs. Lovedale: Lovedale Press.
Stein, H.B. 1933. Stone implements from the Cathkin Peak area. Bantu Studies 7 (2): 159–81.
Steyn, M. & İşcan, M.Y. 1999. Osteometric variation in the humerus: sexual dimorphism in South Africans.
Forensic Science International 106: 77–85.
Stuiver, M. & Reimer, P.J. 1993. Extended 14C data base and revised Calib 3.0 14C age calibration program.
Radiocarbon 35: 215–30.
Swart, J. 2004. Rock art sequences in uKhahlamba-Drakensberg Park, South Africa. Southern African
Humanities 16: 13–35.
Theal, G.M. 1898. Records of south-eastern Africa. Vols I, II. Facsimile reprint, 1964. Cape Town: C. Struik.
Tobias, P.V. 1961. Physique of a desert folk: genes, not habitat, shaped the Bushmen. Natural History 170
(2): 16–24.
Tobias, P.V. 1962. On the increasing stature of the Bushmen. Anthropos 57: 801–10.
Truswell, A.S. & Hansen, J.D.L. 1976. Medical research among the !Kung. In: R.B. Lee & I. DeVore,
eds, Kalahari hunter-gatherers: studies of the !Kung San and their neighbors. Cambridge, MA.: Harvard
University Press, pp. 166–94.
Tyson, P.D., Karlén, W., Holmgren, K. & Heiss, G.A. 2000. The Little Ice Age and medieval warming in
South Africa. South African Journal of Science 96: 121–6.
Van Heerden, L. & Van de Venter, A. 2002. Oliviershoek Pass Shelter, KwaZulu-Natal: report on the
excavation of human skeletal remains. For Drakensville Berg Resort. Report in KwaZulu-Natal
Museum archaeological site record le.
Vinnicombe, P. 1971. A Bushman hunting kit from the Natal Drakensberg. Annals of the Natal Museum
20: 611–25.
Vogel, J.C., Fuls, A. & Visser, E. 2001. Radiocarbon adjustments to the dendrochronology of a yellowwood
tree. South African Journal of Science 97 (3–4): 164–66.
Webb, C. de B. & Wright, J.B., eds & trans. 1979 (vol. 2), 1982 (vol. 3), 1986 (vol. 4), 2001 (vol. 5), 2014
(vol. 6). The James Stuart archive of recorded oral evidence relating to the history of the Zulu and neighbouring
peoples. Pietermaritzburg: University of Natal Press.
Wells, L.H. 1933a. The archaeology of Cathkin Park: Introductory. Bantu Studies 7 (2): 113–29.
Wells, L.H. 1933b. Old Bantu graves in the Cathkin Park area. Bantu Studies 7 (2): 195–200.
Wells, L.H. 1933c. Ancient metal working, ceramics, and beads from Cathkin Park. Bantu Studies 7 (2):
Whitelaw, G. 2004. Iron Age hilltop sites of the early to mid-second millennium AD in KwaZulu-Natal,
South Africa. In: K. Sanogo, T. Togola, D. Keïta & M. N’Daou, eds, Proceedings of the 11th congress
of the PanAfrican Association for Prehistory and Related Studies, Bamako, Februar y 2001. Bamako: Institut
des Sciences Humaines, pp. 38–51.
Whitelaw, G. 2009. ‘Their village is where they kill game’: Nguni interactions with the San. In: P. Mitchell
& B. Smith, eds, The eland’s people: new perspectives on the rock art of the Maloti-Drakensberg Bushmen (Essays
in memory of Pat Vinnicombe). Johannesburg: Wits University Press, pp. 135–59.
Whitelaw, G. 2015. Economy and cosmology in the Iron Age of KwaZulu-Natal. PhD thesis, University of the
Wood, M. 2008. Post-European contact glass beads from the southern African interior: a tentative look
at trade, consumption and identities. In: N. Swanepoel, A. Esterhuysen & P.L. Bonner, eds, Five
hundred years rediscovered: southern African precedents and prospects. Johannesburg: Wits University Press,
pp. 183–96.
Wood, M., Dussubieux, L. & Wadley, L. 2009. A cache of ~5 000 glass beads from the Sibudu Cave Iron
Age occupation. Southern African Humanities 21: 239–61.
Zwane, T. n.d. Report on the rescue excavation at Inhlanhla Game Ranch, Vryheid, KwaZulu-Natal.
Report on le at Amafa aKwaZulu-Natali, Pietermaritzburg.
Cranial and postcranial measurements (in mm). * = measured on the right side.
Cranial measurements
Max cranial length 171
Max cranial breadth 132
Bizygomatic diameter 135
Basion-bregma height 133
Cranial base length 103
Basion-prosthion length 102
Maxillo-alveolar breadth 40
Maxillo-alveolar length 50
Upper facial height 66
Min frontal breadth 92
Upper facial breadth 110
Nasal height 48
Nasal breadth 31
Orbital breadth 39
Orbital height 35
Biorbital breadth 103
Interorbital breadth 28
Frontal chord 102
Parietal chord 112
Occipital chord 95
For magnum length 33
For magnum breadth 28
Mastoid length 27 27
Chin height 31 26
Mandibular body height 27 24
Mandibular body breadth 11 99
Bigonal width 102
Bicondylar breadth 118
Min ramus breadth 34 30
Max ramus breadth 40 38
Max ramus height 55
Mandibular length 76
Mandibular angle 122°
TABLE 1 (continued)
Cranial and postcranial measurements (in mm). * = measured on the right side.
Postcranial measurements
Clavicle: max length
Clavicle: ant-post diameter 9
Clavicle: sup-inf diameter 7
Scapula: height 119*
Scapula: breadth 91
Humerus: max length 328*
Humerus: epicondylar breadth 57*
Humerus: vertical head diameter 41*
Humerus: max diameter midshaft 19* 21
Humerus: min diameter midshaft 17* 15
Radius: max length 242*
Radius: ant-post diameter 12*
Radius: med-lat diameter 15*
Ulna: max length 263*
Ulna: ant-post diameter 12*
Ulna: med-lat diameter 16*
Ulna: physiological length 236*
Ulna: min circumference 35*
Sacrum: anterior length
Sacrum: ant-sup breadth
Sacrum: max trans base diameter
Os coxa: height 170
Os coxa: iliac breadth 123
Os coxa: pubis length 81
Os coxa: ischium length 88
TABLE 1 (continued)
Cranial and postcranial measurements (in mm). * = measured on the right side.
Postcranial measurements
Femur: max length 418*
Femur: bicondylar length 409*
Femur: Epicondylar breadth 71*
Femur: max diameter head 42*
Femur: ant-post subtroch diameter 24* 25
Femur: med-lat subtroch diameter 30* 34
Femur: ant-post midshaft diameter 30* 30
Femur: med-lat midshaft diameter 28* 31
Femur: midshaft circumference 88* 91
Tibia: length
Tibia: max prox epiphyseal breadth
Tibia: max dist epiphyseal breadth 38
Tibia: max diameter nutrient for 33
Tibia: med-lat diameter nutrient for 22
Tibia: circumference nutrient for 85
Fibula: max length 306
Fibula: max diameter midshaft 13
Calcaneus: max length 64
Calcaneus: middle breadth 37
Dental measurements (in mm). * = measured on the right side. MD = mesiodistal, BL = buccolingual,
CH = crown height.
Maxilla Mandible
Castle Newcastle
Castle Newcastle
MD I1 6.3*
BL I1 6.8*
CH I1 9.9*
MD I2 5.6*10.3 *
BL I2 6.2*99.3 *
CH I2 8.9*-
MD C 7.8* 77.5
BL C 8.5* 77.7
CH C 5.4* 77.3
MD M1 10.7*
BL M1 10.7*
CH M1 -
MD M2 10.7
BL M2 10.2
CH M2 5.3
MD M3 10.2 6.3*
BL M3 79.7 7.5*
CH M3 75.2 9.6*
... In recent papers (Schlebusch et al. 2017;Steyn et al. 2019), seemingly unimportant and little-studied remains from the KwaZulu-Natal region of South Africa were described in some detail and provided much information about the history of southern Africa. The Cathkin Peak skeletons, from the same general region, originate from a series of rock shelters in the Drakensberg region (uThukela Basin) of South Africa. ...
... During this expedition, excavations were also undertaken at Elands Cave, although no mention is made of skeletons having been found there. The Elands Cave human remains mentioned in Steyn et al. (2019) were discovered in the late 1920s and were not excavated during the 1931/1932 expeditions. According to Wells (1933a), of the sites discussed here, only Kaybar's Cave was used for habitation, while Murray's Shelter, Mason's Shelter, Yellow Tree Shelter and Robinson's Shelters 1 and 2 were used as burial sites only. ...
... Rainmakers were often buried in rock shelters and were for the most part older males (Berglund 1976;Murimbika 2006;Steyn et al. 2019), which to some extent fits the demographics of the individuals described here. Rock shelters were often used for rainmaking as they provided working environments that were secluded, cool and had a strong connection with spirits (Whitelaw 2017). ...
... For this analysis cranial measurements, as outlined by Howells (1973), were recorded. The analysis was run similarly to that reported for an archaeological case from KwaZulu-Natal (Steyn et al. 2019), but it should be kept in mind that the results can only be as good as the background data (Jantz and Ousley 2005) and therefore more sophisticated analyses should be considered in future. ...
... As is the case with many other such collections, the Dart collection houses many smaller collections of which very little are known. In a recent paper by Steyn et al. (2019) it was shown that these seemingly unimportant remains can provide significant information and that we need to do all that we can to provide context and information, especially in cases such as this where there are unpublished documents that may be lost if they are not properly looked after. Future research avenues to be considered could include aDNA analysis for the two more complete skeletons in order to elucidate their population relationships better. ...
Full-text available
The Nyanga complex is situated in the Penhalonga district of Manicaland province in northern Zimbabwe. According to various archival and modern sources, six skeletons were discovered in this region in the 1930s and are supposedly curated in the Raymond A. Dart Collection of Archaeological Humans Remains at the University of the Witwatersrand, Johannesburg. In an attempt to locate these skeletons and associate them with the sites, archival records, skeletal and faunal analyses and radiocarbon dating were used to gain more information on the bioarchaeology of the region. Only three of the skeletons could be located in the Dart Collection, two of which could be reliably radiocarbon dated, one from the Hill of Paintings to before the beginning of the Nyanga complex, the other, from Mkondwe, to most probably contemporary with it. The latter shows evidence of dental modification similar to that seen in individuals recovered from the Monk’s Kop site, situated to the north of Zimbabwe. This study forms part of a larger attempt to bring context to skeletons housed in archaeological collections because of their value as sources of information on the past.
... This group of East African origin was an already admixed group with both East African and Eurasian genetic components (69% East African and 31% Eurasian ancestry), comparable to the present-day Amhara and Oromo groups from Ethiopia [2,6]. The East African migration into southern Africa was shortly followed by an independent and separate agro-pastoral migration into the region, the Bantu expansion, which introduced a West African genetic component into southern Africa [1,7,8]. Bantu speakers across sub-Saharan Africa have a clearly distinguishable West African genetic ancestry, irrespective of their present-day location [9][10][11][12]. ...
Full-text available
Background Hunter-gatherer lifestyles dominated the southern African landscape up to ~ 2000 years ago, when herding and farming groups started to arrive in the area. First, herding and livestock, likely of East African origin, appeared in southern Africa, preceding the arrival of the large-scale Bantu-speaking agro-pastoralist expansion that introduced West African-related genetic ancestry into the area. Present-day Khoekhoe-speaking Namaqua (or Nama in short) pastoralists show high proportions of East African admixture, linking the East African ancestry with Khoekhoe herders. Most other historical Khoekhoe populations have, however, disappeared over the last few centuries and their contribution to the genetic structure of present-day populations is not well understood. In our study, we analyzed genome-wide autosomal and full mitochondrial data from a population who trace their ancestry to the Khoekhoe-speaking Hessequa herders from the southern Cape region of what is now South Africa. Results We generated genome-wide data from 162 individuals and mitochondrial DNA data of a subset of 87 individuals, sampled in the Western Cape Province, South Africa, where the Hessequa population once lived. Using available comparative data from Khoe-speaking and related groups, we aligned genetic date estimates and admixture proportions to the archaeological proposed dates and routes for the arrival of the East African pastoralists in southern Africa. We identified several Afro-Asiatic-speaking pastoralist groups from Ethiopia and Tanzania who share high affinities with the East African ancestry present in southern Africa. We also found that the East African pastoralist expansion was heavily male-biased, akin to a pastoralist migration previously observed on the genetic level in ancient Europe, by which Pontic-Caspian Steppe pastoralist groups represented by the Yamnaya culture spread across the Eurasian continent during the late Neolithic/Bronze Age. Conclusion We propose that pastoralism in southern Africa arrived through male-biased migration of an East African Afro-Asiatic-related group(s) who introduced new subsistence and livestock practices to local southern African hunter-gatherers. Our results add to the understanding of historical human migration and mobility in Africa, connected to the spread of food-producing and livestock practices.
... Current-day Nama individuals display small amounts of West African ancestry, likely from recent admixtureSchlebusch et al., 2012). Bantu speakers started to arrive in the northern parts of today's South Africa 1800 years ago and gradually migrated southwards(Huffman, 1989;Steyn et al., 2019), admixing with Khoe-San populations along the way(Schlebusch et al., 2017). By the 1600s, Bantu speakers occupied much of the eastern part of southern Africa up to the Great Fish River in the Eastern Cape. ...
Full-text available
Previous studies show that the indigenous people of the southern Cape of South Africa were dramatically impacted by the arrival of European colonists starting ~400 years ago and their descendants are today mixed with Europeans and Asians. To gain insight on the occupants of the Vaalkrans Shelter located at the southernmost tip of Africa, we investigated the genetic make-up of an individual who lived there about 200 years ago. We further contextualize the genetic ancestry of this individual among prehistoric and current groups. From a hair sample excavated at the shelter, which was indirectly dated to about 200 years old, we sequenced the genome (1.01 times coverage) of a Later Stone Age individual. We analyzed the Vaalkrans genome together with genetic data from 10 ancient (pre-colonial) individuals from southern Africa spanning the last 2000 years. We show that the individual from Vaalkrans was a man who traced ~80% of his ancestry to local southern San hunter-gatherers and ~20% to a mixed East African-Eurasian source. This genetic make-up is similar to modern-day Khoekhoe individuals from the Northern Cape Province (South Africa) and Namibia, but in the southern Cape, the Vaalkrans man's descendants have likely been assimilated into mixed-ancestry "Coloured" groups. The Vaalkrans man's genome reveals that Khoekhoe pastoralist groups/individuals lived in the southern Cape as late as 200 years ago, without mixing with non-African colonists or Bantu-speaking farmers. Our findings are also consistent with the model of a Holocene pastoralist migration, originating in Eastern Africa, shaping the genomic landscape of historic and current southern African populations.
... Zimbabwe repre- sented a major ancient civilization that flourished in southern Africa between the 12th and 18th centuries AD.17Archeologists highlighted the important need of collaborative research to better understand the demographic history and migrations of populations with different sub- sistence practices and technological innovations.18,193 | NEW INSIGHTS TO STUDY PAST AND RECENT HISTORY IN AFRICAThe Bantu expansion is the most important linguistic, cultural, and demographic process in Late Holocene Africa.20It ...
Full-text available
The Keimoes 3 desert kite site, South Africa: an aerial lidar and micro-topographic exploration - Marlize Lombard, Matthew V. Caruana, Jaco van der Walt, Anders Högberg
Full-text available
Chronological record of completed research for the period 2012-2019, including a concise description of the work done, a summary of the results achieved an explanation of the significance of the work, and citation statistics.
Full-text available
Abstract Purpose Skeletons sampled for ancient human DNA analysis are sometimes complete enough to provide information about the lives of the people they represent. We focus on three Later Stone Age skeletons, ca. 2000 B.P., from coastal KwaZulu-Natal, South Africa, whose ancient genomes have been sequenced (Schlebusch et al., 2017). Methods Bioarchaeological approaches are integrated with aDNA information. Results All skeletons are male. Dental development shows that the boy, with prominent cribra orbitalia, died at age 6–7 years. Two men show cranial and spinal trauma, extensive tooth wear, plus mild cribra orbitalia in one. Conclusions Dental wear and trauma of the adults are consistent with hunter-gatherer lives. Even partial aDNA evidence contributes to sex determination. Parasitic infection such as schistosomiasis is the best-fit cause for the child’s anemia in this case. Contribution to knowledge The convergence of genomic and bioarchaeological approaches expands our knowledge of the past lives of a boy and two men whose lives as hunter-gatherers included episodes of trauma and disease. Limitations The skeletons are incomplete, in variable condition, and from poorly characterized local cultural contexts. Suggestions for further research Thorough osteobiographic analysis should accompany paleogenomic investigations. Such disciplinary collaboration enriches our understanding of the human past.
Full-text available
Given the importance of Africa to studies of human origins and disease susceptibility, detailed characterization of African genetic diversity is needed. The African Genome Variation Project provides a resource with which to design, implement and interpret genomic studies in sub-Saharan Africa and worldwide. The African Genome Variation Project represents dense genotypes from 1,481 individuals and whole-genome sequences from 320 individuals across sub-Saharan Africa. Using this resource, we find novel evidence of complex, regionally distinct hunter-gatherer and Eurasian admixture across sub-Saharan Africa. We identify new loci under selection, including loci related to malaria susceptibility and hypertension. We show that modern imputation panels (sets of reference genotypes from which unobserved or missing genotypes in study sets can be inferred) can identify association signals at highly differentiated loci across populations in sub-Saharan Africa. Using whole-genome sequencing, we demonstrate further improvements in imputation accuracy, strengthening the case for large-scale sequencing efforts of diverse African haplotypes. Finally, we present an efficient genotype array design capturing common genetic variation in Africa
Full-text available
Southern Africa is consistently placed as a potential region for the evolution of Homo sapiens. We present genome sequences, up to 13x coverage, from seven ancient individuals from KwaZulu-Natal, South Africa. Three Stone Age hunter-gatherers (~2,000 years old), were genetically similar to current-day southern San groups, while four Iron Age farmers (300-500 years old) were genetically similar to present-day Bantu-speakers. We estimate that all modern-day Khoe-San groups have been influenced by 9-30% genetic admixture from East Africans/Eurasians. Using traditional and new approaches, we estimate the first modern human population divergence time to between 350,000 and 260,000 years ago. This estimate increases the deepest divergence amongst modern humans, coinciding with anatomical developments of archaic humans into modern humans as represented in the local fossil record.
Full-text available
Reduced susceptibility to infectious disease can increase the frequency of otherwise deleterious alleles. In populations of African ancestry, two apolipoprotein-L1 (APOL1) variants with a recessive kidney disease risk, named G1 and G2, occur at high frequency. APOL1 is a trypanolytic protein that confers innate resistance to most African trypanosomes, but not Trypanosoma brucei rhodesiense or T.b. gambiense, which cause human African trypanosomiasis. In this case-control study, we test the prevailing hypothesis that these APOL1 variants reduce trypanosomiasis susceptibility, resulting in their positive selection in sub-Saharan Africa. We demonstrate a five-fold dominant protective association for G2 against T.b. rhodesiense infection. Furthermore, we report unpredicted strong opposing associations with T.b. gambiense disease outcome. G2 associates with faster progression of T.b. gambiense trypanosomiasis, while G1 associates with asymptomatic carriage and undetectable parasitemia. These results implicate both forms of human African trypanosomiasis in the selection and persistence of otherwise detrimental APOL1 kidney disease variants.
In the last three decades, genetic studies have played an increasingly important role in exploring human history. They have helped to conclusively establish that anatomically modern humans first appeared in Africa roughly 250,000-350,000 years before present and subsequently migrated to other parts of the world. The history of humans in Africa is complex and includes demographic events that influenced patterns of genetic variation across the continent. Through genetic studies, it has become evident that deep African population history is captured by relationships among African hunter-gatherers, as the world's deepest population divergences occur among these groups, and that the deepest population divergence dates to 300,000 years before present. However, the spread of pastoralism and agriculture in the last few thousand years has shaped the geographic distribution of present-day Africans and their genetic diversity. With today's sequencing technologies, we can obtain full genome sequences from diverse sets of extant and prehistoric Africans. The coming years will contribute exciting new insights toward deciphering human evolutionary history in Africa.