1009 December 2013, Vol. 103, No. 12 (Suppl 1) SAMJ
The Lemba have a strong oral history of non-African
origins, and a culture similar to that practiced by
Jewish/Arabic people. They live among other larger
groups of people in southern Africa, mainly the
Venda (Limpopo Province, South Africa (SA)) and
Northern Sotho or Pedi in Sekhukhuneland (Mpumalanga, SA) and
among the Shona (Kalanga) in the southern parts of Zimbabwe.
While some Lemba, particularly those from SA, claim Jewish
origins, the Jewish link is not universally accepted, and there have
been several studies suggesting Islamic connections with Arabs.[2-4]
Moreover, some of the Lemba who live in Zimbabwe, referred to
here as Remba (their name for themselves, since there is no ‘L’ sound
in their spoken language Shona), identify with Arabic ancestry, and
several clan names in use are Arabic in origin.
According to van Warmelo, the ancestors of the Lemba came from
a huge town somewhere across the seas, where they were skilled in
metalwork, pottery, textiles and ship-building – presumably the
same place Mathivha refers to as Sena or Sa’na. They came to this
country to trade their goods, especially for gold. They began leaving
some of their men behind with unsold cargo and thus established
posts. They moved further and further inland and became well
known with the locals, but did not mix with them as they deemed
A few genetic studies have attempted to elucidate the ancestry of
the Lemba. When ‘classical’ serogenetic markers (unpublished) and
mitochondrial DNA restriction fragment length polymorphisms
(RFLPs) were used, no differences between the Lemba and other
southern African populations could be detected. However,
Lemba origins revisited: Tracing the ancestry of
Y chromosomes in South African and Zimbabwean Lemba
H Soodyall, BSc (Hons), MSc, PhD
Division of Human Genetics, School of Pathology, Faculty of Health Sciences, University of the Witwatersrand and
National Health Laboratory Service, Johannesburg, South Africa
Corresponding author: H Soodyall (firstname.lastname@example.org)
My interest in the history of sub-Saharan African populations was sparked over 25 years ago when listening to the interesting and captivating
stories Professor Trefor Jenkins shared about field trips to various parts of the country, about various groups of people he encountered, and
how the use of classical genetic markers added value in studies concerning the prehistory of southern African populations. He also challenged
us ‘youngsters’ to read more widely across disciplines, to consider genetic data in conjunction with data from other fields like archaeology,
anthropology, history and palaeontology, and to use genetic information judiciously to refine and/or validate theories concerning population
history and their affinities put forward by scholars of other disciplines.
Thus, when he was told by a friend and colleague, Dr Margaret Nabaro, a musicologist at the University of the Witwatersrand, about the
parallels in music she heard played by Lemba people from Limpopo Province with that played by Jewish people, and when he learnt about
certain cultural practices that the Lemba shared with Jews, Professor Jenkins took up her offer to visit the Lemba. They attended the Annual
Cultural festival at Sweetwater in Limpopo Province in 1987 and 1988 and, following an invitation from Professor Mathiva, a member of the
Lemba Cultural Association (LCA), collected blood samples. I was fortunate to have accompanied Professor Jenkins to meet with Professor
Mathiva and members of the LCA and to conduct additional field work in 2000.
It gives me great pleasure, as a former PhD student of Professor Jenkins, to write this paper in his honour as we celebrate his contribution to
science and to present an updated story using new data on the origins of the Lemba.
Background. Previous historical, anthropological and genetic data provided overwhelming support for the Semitic origins of the Lemba, a
Bantu-speaking people in southern Africa.
Objective. To revisit the question concerning genetic affinities between the Lemba and Jews.
Methods. Y-chromosome variation was examined in two Lemba groups: one from South Africa (SA) and, for the first time, a group from
Zimbabwe (Remba), to re-evaluate the previously reported Jewish link.
Results. A sample of 261 males (76 Lemba, 54 Remba, 43 Venda and 88 SA Jews) was initially analysed for 16 bi-allelic and 6 short tandem
repeats (STRs) that resulted in the resolution of 102 STR haplotypes distributed across 13 haplogroups. The non-African component in the
Lemba and Remba was estimated to be 73.7% and 79.6%, respectively. In addition, a subset of 91 individuals (35 Lemba, 24 Remba, 32 SA
Jews) with haplogroupJ were resolved further using 6 additional bi-allelic markers and 12 STRs to screen for the extended Cohen modal
haplotype (CMH). Although 24 individuals (10 Lemba and 14 SA Jews) were identified as having the original CMH (six STRs), only one
SA Jew harboured the extended CMH.
Conclusions. While it was not possible to trace unequivocally the origins of the non-African Y chromosomes in the Lemba and Remba, this
study does not support the earlier claims of their Jewish genetic heritage.
S Afr Med J 2013;103(12 Suppl 1):1009-1013. DOI:10.7196/SAMJ.7297
1010 December 2013, Vol. 103, No. 12 (Suppl 1) SAMJ
Y-chromosome RFLP studies conducted by Spurdle and Jenkins[7,8]
provided the first definitive evidence that the male gene pool of SA
Lemba was derived, in part, from non-African sources. Based on the
49a/TaqI system, 53% of Y chromosomes in the Lemba were assigned
to haplotypes that were also found at considerable frequencies in
SA Indians (38%) and in Jewish populations from SA, Europe and
Yemen (>50%). Although Spurdle and Jenkins could not distinguish
between Jewish and non-Jewish Y chromosomes at this level of
resolution, they concluded that the non-African Y chromosomes in
the Lemba were of Semitic origin.
Using a combination of 6 short tandem repeat (STR) markers in
conjunction with 6 bi-allelic markers, Thomas et al. were able to
further resolve the Y chromosomes in the Lemba. They found that a
particular Y-chromosome haplotype – referred to as the Cohen modal
haplotype (CMH) – was present in the Lemba at a frequency of 8.8%
(12/136). The CMH had previously been reported at frequencies of
44.9% in Ashkenazi and 56.1% in Sephardic Cohanim (descendants of
Jewish priests) and among Ashkenazi and Sephardic Israelites at 13.2%
and 9.8%, respectively. Thomas et al. estimated that the CMH
originated approximately 2000 - 3200 years ago and suggested that
this Y-chromosome haplotype was useful in tracing Jewish ancestry.
The CMH was subsequently shown to segregate on the background
of haplogroup J-12f2a and was present in samples of Ashkenazi,
Sephardic and Kurdish Jews (without considering religious status) at
frequencies of 7.6%, 6.4% and 10.1%, respectively.
Increasing the resolution of Y-chromosome analysis that included
75 binary markers and 22 STR markers, Hammer et al. genotyped
122 Ashkenazi and 93 non-Ashkenazi Cohanim to further resolve
the paternal ancestry of Jewish priests. Haplogroup J, defined by the
presence of the 12f2a/M304 mutation, was resolved further into 16
sub-haplogroups defined by 15 binary markers. A new mutation,
P58T→C, defined the most common lineage J-P58* in the Cohanim
and it was on this haplogroup background that the original CMH
defined on the basis of the 6-STR marker system was found. The
inclusion of 6 additional STRs extended this haplotype to 12 STRs
which was subsequently referred to as the extended CMH. Of the
99 J-P58* Y chromosomes they examined, 87 carried the CMH using
the 6-STR marker system. However, when the 12-STR marker system
was used, only 43 Y chromosomes were found to have the extended
CMH. Also, whereas the original CMH was found in several Near
Eastern populations, the extended CMH, together with its two closely
related haplotypes, had a much more restricted distribution and was
only found among Cohanim (29.8%) and Israelites (1.5%), but not
among the Levites and non-Jewish populations surveyed.
Given that the higher resolution at both the haplogroup and
haplotype level helped refine the definition of the CMH, its frequency
among the Cohanim, and its distribution in other Jewish and
non-Jewish populations, the present study revisited the question
concerning genetic affinities between the Lemba and Jews. This was
done following screening for the extended CMH in a sample of SA
Lemba and Zimbabwean Remba.
Blood samples were collected from healthy, unrelated adult
volunteers with their informed consent and approval from the Ethics
Committee for research on human subjects at the University of the
Witwatersrand (protocol no. M980553). Two groups of Lemba were
examined in this study – one from Limpopo Province, SA (referred
to herein as Lemba) and the Remba from Zimbabwe. The Venda
from Limpopo, the people among whom the Lemba reside in this
region, and a sample of predominantly Ashkenazi Jewish males from
Johannesburg, SA, were included for comparative analysis in the
study. Altogether, 261 individuals were examined for Y-chromosome
polymorphisms (76 Lemba, 54 Remba, 43 Venda and 88 SA Jews).
Typing of Y-chromosome polymorphisms
DNA samples were extracted from peripheral blood using the
standard salting-out method. Initially, 16 bi-allelic Y-chromosome
polymorphisms were used to identify the major haplogroups
(Fig. 1A).[14,15] Haplogroup designation follows the nomenclature
proposed by the Y Chromosome Consortium. Subsequently,
haplogroupJ* defined by the M304 mutation, was resolved further
after screening for the mutations M267, M172, P58, M410, M318
and M12 using a single-base extension method.
Haplotypes were resolved using both bi-allelic and STR
polymorphisms. The original CMH was defined using 6 STR loci:
DYS19, DYS388, DYS390, DYS391, DYS392 and DYS393 (in this
order) that were amplified in two separate multiplex polymerase chain
reactions (PCR). The 12 STRs (DYS19, DYS385a, DYS385b, DYS388,
DYS389I, DYS389II, DYS390, DYS391, DYS392, DYS393, DYS426 and
DYS439 that defined the extended CMH in this order) were genotyped
in a subset of 91 individuals who were found to have haplogroup
J-12f2a or J-M172 Y chromosomes. This was accomplished by using
AmpFISTR Y-filer (Applied Biosystems) which amplified 17 Y-STRs:
DYS19, DYS389I, DYS389II, DYS390, DYS391, DYS392, DYS393,
DYS385a, DYS385b, DYS437, DYS438, DYS439, DYS456, DYS458,
DYS635 and YGATA H4 (underlined loci part of the extended CMH)
and the Multiplex II PCR amplification kit (Applied Biosystems) that
included the two STRs DYS388 and DYS426.
Y-chromosome variation was analysed at the haplogroup level
defined by bi-allelic polymorphisms only, and at the haplotype
level defined by both bi-allelic and STR markers. Haplogroup and
haplotype diversity within populations was determined as described
by Nei. Population differentiation based on haplogroup and
haplotype frequency data was performed using an exact test using
Arlequin software (version 3.5). Variation within haplogroups was
determined by calculating haplotype diversity and mean pairwise
differences between haplotypes using Arlequin.
Haplogroup variation using 16 bi-allelic and
6 STR markers
Typing for the 16 bi-allelic polymorphisms in the total sample of 261
males resulted in the derivation of 13 Y-chromosome haplogroups
(Fig.1A). The bi-allelic markers together with the 6 STR loci defined
102 haplotypes (Appendix1, available online).
Using the distribution of Y-chromosome haplogroups in global
populations,[14,15] haplogroups B-SRY10831.1, E-M2 and E-M40 were
used to trace the African ancestry of Y chromosomes in the combined
Lemba/Remba, whereas haplogroups F*, J, K*, L, P* and R were used to
trace the non-African contribution. Although haplogroups E-M40 and
E-M35 occur in both African and non-African populations, we were
able to resolve the ancestry of the E-M40 chromosomes in the Lemba
to African sources and the E-M35 chromosomes to non-African
sources using the haplotype data from comparative analysis. Thus, the
African contribution of Y-chromosome haplogroups in present-day
Lemba and Remba was estimated to be 26.3% and 20.4%, respectively.
A comparison of the haplogroup distribution among Remba,
Lemba, Venda and Jews revealed that Remba and Lemba had some
haplogroups in common with the Venda and others with Jews
(Fig. 1A). Of the 11 haplogroups found in the combined Remba/
1011 December 2013, Vol. 103, No. 12 (Suppl 1) SAMJ
Lemba sample, only four (E-M2, J-12f2a,
J-M172 and L-M11) were shared between
them. Although the four haplogroups
together constituted a major component
of the Y chromosomes in both Lemba
(73.7%) and Remba (94.4%), the frequencies
of these haplogroups (with the exception
of haplogroup E-M2) differed in the two
populations (Fig. 1A). The most common
haplogroup in the Remba was J-M172, which
accounted for 42.6% of their Y chromosomes,
whereas J-12f2a was most common in the
Lemba (39.5%). Both J haplogroups were
observed in Jews at frequencies of 18.1% and
22.7%, respectively. The CMH (14-16-23-10-
11-12), which was resolved on haplogroup
J-12f2a was only found in the Lemba (13.2%)
and Jewish (15.9%) groups, but not in the
Remba (H54, Appendix1).
With the exception of L-M11, which
was only found in Lemba and Remba, all
haplogroups derived from non-African
ancestry in the combined Lemba/Remba
sample were also found in Jews.
Using the exact test of differentiation, all
population pairs, including the Remba v.
Lemba, were found to be significantly different
from each other using both haplogroup
(p<0.001) and haplotype (p<0.001) frequency
data. Only 5 (H26, H58, H65, H67, H84) of
the 42 haplotypes found in the combined
Lemba/Remba sample were shared between
the two groups (Appendix 1). Altogether,
these lineages accounted for 39.2% of the
Y chromosomes in the Lemba/Remba and
they shared predominantly higher-frequency
There was little divergence, as reflected in
the small mean pairwise differences and low
haplotype diversity values (Table 2), within
haplogroups found in the Lemba/Remba.
Several haplogroups were represented only
by a single haplotype, some of them at
high frequency (Table2). Other haplogroups
comprised one major haplotype from
which few one-step neighbours had evolved
Finer resolution of haplogroup J*
and the extended CMH
Screening for the 6 binary markers within
haplogroup J* in 35 Lemba, 24 Remba and
32 SA Jews resolved haplogroup J* into
4 sub-haplogroups (Fig. 1B). Haplogroup
J-P58 which is associated with the CMH was
found in the Lemba (57.1%) and SA Jews
(56.3%). Haplogroup J-M267* was found
in the Lemba (28.6%) and Remba (16.0%)
but not in the SA Jews. Haplogroup J-M172
was further resolved into sub-haplogroups
J-M410* in the Remba (43.5%) and SA Jews
(37.5%) and J-M12* in the Lemba (14.3%),
Remba (44.4%) and SA Jews (6.3%).
The 12-STR marker system resolved the
91 M304 Y chromosomes (Fig. 1B) into 46
haplotypes (Appendix 2, available online).
The extended CMH (14-13-15-16-13-30-23-
10-11-12-11-12) on the J-P58 haplogroup
background was only found in 1/14 Jewish
individuals who were found to have the
original CMH (6 STRs). None of the 10
Lemba who were found to have the original
CMH had the extended CMH. The most
common haplotype 14-13-19-16-13-29-23-
10-11-12-11-12 found in eight Lemba differed
from the extended CMH by 5 mutational
steps; 4 at the DYS385b locus and 1 at the
DYS389II locus. The closest haplotype in the
Lemba to the extended CMH differed by 4
mutational steps (3 at the DYS385b locus and
1 at the DYS389II locus).
At this higher level of resolution at both
haplogroup and haplotype level, the Lemba
and Remba only shared 3 haplotypes – 1
on the J-M267* background and 2 on the
J-M12 background (Appendix 2). None of
the combined Lemba/Remba haplotypes
were shared with the Jews.
The Lemba have attracted a great deal of media
attention following on Y-chromosome studies
that lent support to their claims of having
Jewish ancestry.[7-9] Given that the refinement
in the resolution of Y chromosomes at both
a haplogroup and haplotype level delineated
an extended CMH that was found at a lower
frequency than the originally defined CMH
among the Jewish priesthood (Cohanim), and
had a more restricted global distribution,
this study sought to revisit the question
concerning the origin(s) of non-African
Y chromosomes in the Lemba, in particular
whether the extended CMH was present
in the Lemba, and to examine the genetic
affinities between the Lemba and Remba.
Fig. 1. (A) Phylogeny and distribution of Y chromosome haplogroups dened by the 16 bi-allelic markers
used in this study (black bars) in the Lemba, Remba, Venda and South African Jewish populations.
e 13 haplogroups found in the total sample are indicated with solid lines, while those haplogroups
not found in the sample are shown using dashed lines. (B) Phylogeny showing the ner resolution of
haplogroup J chromosomes using 6 additional single nucleotide polymorphisms and the number of
individuals from a subset of Lemba, Remba and Jewish samples with these sub-haplogroups.
1012 December 2013, Vol. 103, No. 12 (Suppl 1) SAMJ
The majority of Y chromosomes found in the Lemba (73.7%) and Remba
(79.6%) were traced to non-African origins (Fig.1). Haplogroups C, D and
O were not found, suggesting that p opulations from geographical regions
where haplogroups C and D (Asians) and haplogroup O (Oceanic) are
commonly found did not contribute to the Y chromosomes found in the
combined Lemba/Remba sample. Haplogroup J was the most common
haplogroup in the Lemba/Remba (51.7%) and it was on the background
of haplogroup J-12f2a that the original haplotype designation of the
CMH based on the 6-STR system was found at frequencies of 9.6% in
the Lemba and 15.9% in SA Jews, but not in the Remba (Fig.1A). When
the higher resolution of haplogroup J* in combination with the 12-STR
system was used, the extended CMH was only found in one individual of
SA Jewish descent (1.1%) and not at all in the Lemba. This finding argues
against the claims made previously about possible connections between
the Lemba and the Cohanim.
Hammer et al. showed that while haplogroup J-P58* occurred
among both Jewish (18.8%) and non-Jewish (15.5%) populations, the
extended CMH was restricted to only the Jewish groups from across
the range of the Jewish Diaspora. Furthermore, after using the 12-STR
marker system to resolve haplotypes associated with haplogroup J-P58
Y chromosomes, the extended CMH was found at frequencies of 64.6%
among Cohanim, 10.7% among Israelites and 12.2% in Ashkenazi Jews
with unknown castes, but was not found in the Leviim or in 2099 non-
Jewish individuals surveyed. Overall, the extended CMH was found
at a frequency of 5.3% among Jewish populations.
Only one of the two closely related haplotypes to the extended
CMH reported was found in a Jewish individual in this study
(haplotype 8, Appendix 2). There was no haplotype sharing between
the combined Lemba/Remba sample and SA Jews when the 12-STRs
were used to derive haplotypes (Appendix 2). Also, there were no
matches to any of the J-P58 haplotypes derived in the Lemba/Remba
to any of the J-P58 chromosomes reported by Hammer et al.
The three haplogroup J-M304 chromosome haplotypes shared
between the Lemba and Remba (Appendix 2) must have been
present in the founding non-African males who contributed to their
Y chromosomes. Since none of the haplotypes resolved with the 12
STRs within haplogroups J-M267, J-M410 or J-M12 matched those
reported, it is still very difficult to unambiguously resolve the
origins of these non-African Y chromosomes in the Lemba/Remba.
Given the finer resolution of Y chromosomes among the Lemba/
Remba, the original claims of close Jewish links are not strongly
Table 1. The 6-STR system haplotypes found at >9% frequency in any of the four populations
Haplogroup Haplotype Lemba (N=76) Remba (N=54) Venda (N=43) Jews (N=88)
E-M2 17-12-21-10-11-14 - 0.093 0.023 -
E-M2 15-12-21-10-11-13 0.079 0.037 0.419†-
J-12f2a 14-16-24-10-13-12 0.105 - - -
J-12f2a* 14-16-23-10-11-12 0.132†- - 0.159†
J-12f2a 14-15-24-10-11-12 0.118 0.093 - -
J-M172 15-15-24-10-11-12 0.039 0.093 - 0.011
L-M11 14-12-22-10-15-12 0.079 0.167†- -
STR = short tandem repeat.
*Cohen modal haplotype.
†The modal haplotypes in each population.
Table 2. Haplogroup composition in Remba and Lemba (6-STR system)
Haplogroup Population Haplotypes, nIndividuals, n
differences* Haplotype diversity
B-SRY10831.1 Lemba 1 2 0 0
E-M40 Lemba 2 2 4.00 1.0 (±0.5)
E-M2 Lemba 8 16 2.41 0.842 (±0.075)
E-M2 Remba 6 11 3.20 0.8 (±0.114)
E-M35 Lemba 3 5 0.80 0.7 (±0.218)
FI-M213 Lemba 1 2 0 0
J-12f2a Lemba 6 30 1.92 0.758 (±0.041)
J-12f2a Remba 1 5 0 0
J-M172 Lemba 2 4 0.50 0.5 (±0.265)
J-M172 Remba 8 23 2.47 0.878 (±0.031)
K,M,N-M9 Lemba 3 9 0.89 0.722 (±0.097)
L-M11 Lemba 1 6 0 0
L-M11 Remba 3 12 0.47 0.439 (±0.158)
R-M207 Remba 1 2 0 0
R-SRY10831.2 Remba 1 1 0 1 (±0)
1013 December 2013, Vol. 103, No. 12 (Suppl 1) SAMJ
More support for the unlikely Jewish link of the Lemba with Jewish
populations was reported by Mendez et al. following the finer
resolution of haplogroups L and T. In this study, a novel mutation (P326)
was found in both haplogroups L and T. However, additional single
nucleotide polymorphisms (SNPs), PS129, PS18 (M70), PS21 and L131
resolved haplogroup T into sub-branches T*, T1*, T1a* (with 6 sub-
lineages) and T1b* (with 3 sub-lineages), respectively. Y chromosomes
from Jewish populations clustered in branch T1a*. In the sample of
34 Lemba included in this study, 6 Lemba Y chromosomes (17.6%)
were assigned to the branch T1b*. Although this level of resolution
was not genotyped in the present study, T1b* chromosomes would be
represented by haplogroup L-M11 chromosomes found in the Lemba
and Remba (Fig1A). The present-day distribution of haplogroup L-M11
is mostly confined to southern, central and western Asia; it has been
found only at low frequencies in populations of the Caucasus region and
the Middle East and is absent in Palestinian and Syrian Arab.[11,24-27]
Overall, this study has shown that Y chromosomes typically linked
with Jewish ancestry were not detected by the higher resolution
analysis conducted in the present study. It seems more likely that
Arab traders, who are known to have established long-distance trade
networks involving some thousands of kilometres along the western
rim of the Indian Ocean, from Sofala in the south to the Red Sea in
the north and beyond to the Hadramut, India and even China from
about 900AD, are more likely linked with the ancestry of the non-
African founding males of the Lemba/Remba.
Historical data show that after the establishment of a trading hub
in Zanzibar Island by Arabs, Indian merchants capitalised on the
commercial ventures and settled in East Africa to extend trading ties
across the ocean with India. By the beginning of the 15th century,
Islamic trade and settlement was evident along the East African coast,
involving Zanzibar and many other places along the coast, extending
to the south to the mouth of the Zambezi River and inland where
the borders of Botswana, Zimbabwe and SA meet. This area, in the
vicinity of Mapungubwe, was rich in elephant ivory and alluvial
gold, and the traditional leaders, already wealthy in cattle, became
incredibly wealthy through this coastal trade.[29,30] Following a vicious
drought in about 1100AD, the people of Mapungubwe moved north
of the Limpopo to found Great Zimbabwe. From here, trade with the
East for luxury goods such as textiles (silk and cotton), carpets, glazed
pottery from Persia and porcelain bowls from China for African
products that included ivory, copper and gold, flourished.
Previous Y chromosome data, particularly the presence of the CMH
in SA Lemba, led to claims of genetic links between the Jewish
priesthood (Cohanim) and the Lemba. However, higher-resolution
genotyping of haplogroup J chromosomes that harboured the CMH
led to the delineation of an extended CMH that was found at high
frequency among the Cohanim and restricted to Jewish groups across
the Jewish Diaspora. In this study, the extended CMH was only found
in a single Jewish individual, and not in the Lemba or the Remba.
This finding, together with the lack of matches of haplotypes found
in the combined Lemba/Remba group with Jewish populations,
suggests that the haplogroupJ Y chromosomes in the Lemba/Remba
are not closely associated with Jewish ancestry. Rather, this study
suggests a stronger link with Middle Eastern populations, probably
the result of trade activity in the Indian Ocean. Also, although the SA
Lemba and Zimbabwean Remba originated from the same founding
population, several differences were observed in the composition and
frequencies of their Y chromosome pools, perhaps a consequence of
drift and human contacts the groups may have had following their split.
However, it should also be stressed that studies on genetic ancestry
makes use of patterns in DNA to look for similarities or differences,
and in the present study Y chromosome haplogroups and haplotypes,
and that this type of data is in no way aligned with identity testing.
Issues of identity are complex and based on many criteria and should
not be confused with ancestry testing.
Acknowledgements. e author expresses her gratitude to all the
participants who donated blood samples that made this study possible;
Trefor Jenkins, Bruce Dangereld, Paul Stidolf and Mr Moeti for their
invaluable assistance with eldwork in the Northern Province and
Zimbabwe; Bharti Morar, Carey-Anne Eddy, Almut Nebel, Heeran
Makkan, ijessen Naidoo, Rajeshree Mahabeer and Elize van der
Westhuizen for technical assistance and insightful comments to previous
dras of this manuscript. Support for this research was provided jointly by
the SA Medical Research Council, National Health Laboratory Service, the
University of the Witwatersrand and the National Research Foundation.
1. Mathivha MER. e Basena/Vamwenya/Balemba. Toyandu: Morester, 1992.
2. Mandivenga EC. e history and ‘re-conversion’ of the Varemba of Zimbabwe. J Relig Afr 1989;19:98-124.
3. Partt T. Journey to the Vanished City. e Search for the Lost Tribe of Israel. London: Phoenix, 1997.
4. le Roux M. e Lemba. A Lost Tribe of Israel in Southern Africa? Pretoria: UNISA Press, 2003.
5. van Warmelo NJ. e Classication of Cultural Groups. In: Hammond-Tooke WD, ed. e Bantu-speaking
Peoples of Southern Africa. London: Routledge and Kegan, 1974:56-84.
6. Soodyall H. Mitochondrial DNA Polymorphisms in Southern African Populations. PhD thesis. University
of the Witwatersrand, 1993.
7. Spurdle A, Jenkins T. e search for Y chromosome polymorphism is extended to negroids. Hum Mol
Genet 1992;1(3):169-170. [http://dx.doi.org/10.1093/hmg/1.3.169]
8. Spurdle AB, Jenkins T. e origins of the Lemba ‘Black Jews’ of southern Africa: Evidence from p12F2 and
other Y-chromosome markers. Am J Hum Genet 1996;59(5):1126-1133.
9. omas MG, Partt T, Weiss DA, et al. Y chromosomes traveling south: e Cohen modal haplotype and
the origins of the Lemba – the ‘Black Jews of Southern Africa’. Am J Hum Genet 2000;66(2):674-686. [http://
10. omas MG, Skorecki K, Ben-Ami H, Partt T, Bradman N, Goldstein DB. Origins of Old Testament
priests. Nature 1998;394(6689):138-140. [http://dx.doi.org/10.1038/28083]
11. Nebel A, Filon D, Brinkmann B, Majumder PP, Faerman M, Oppenheim A. e Y chromosome pool of
Jews as part of the genetic landscape of the Middle East. Am J Hum Genet 2001;69(5):1095-1112. [http://
12. Hammer MF, Behar DM, Karafet TM, et al. Extended Y chromosome haplotypes resolve multiple and
unique lineages of the Jewish priesthood. Hum Genet 2009;126(5):707-717. [http://dx.doi.org/10.1007/
13. Miller SA, Dykes DD, Polesky HF. A simple salting out procedure for extracting DNA from human
nucleated cells. Nucleic Acids Res 1988;16(3):1215. [http://dx.doi.org/10.1093/nar/16.3.1215]
14. Underhill PA, Passarino G, Lin AA, et al. e phylogeography of Y chromosome binary haplotypes and the
origins of modern human populations. Ann Hum Genet 2001;65(Pt 1):43-62. [http://dx.doi.org/10.1046/
15. Underhill PA, Shen P, Lin AA, et al. Y chromosome sequence variation and the history of human
populations. Nat Genet 2000;26(3):358-361. [http://dx.doi.org/10.1038/81685]
16. Y Chromosome Consortium. A nomenclature system for the tree of human Y-chromosomal binary
haplogroups. Genome Res 2002;12(2):339-348. [http://dx.doi.org/10.1101/gr.217602]
17. Naidoo T, Schlebusch CM, Makkan H, et al. Development of a single base extension method to resolve Y
chromosome haplogroups in sub-Saharan African populations. Investig Genet 2010;1(1):6. [http://dx.doi.
18. Lane AB, Soodyall H, Arndt S, et al. Genetic substructure in South African Bantu-speakers: Evidence from
autosomal DNA and Y-chromosome studies. Am J Phys Anthropol 2002;119(2):175-185. [http://dx.doi.
19. Nei M. Molecular Evolutionary Genetics. New York: Columbia University Press, 1987.
20. Raymond M, Rousset F. An exact test for population dierentiation. Evolution 1995;49(6):1280-1283.
21. Excoer L, Lischer HE. Arlequin suite ver 3.5: A new series of programs to perform population genetics
analyses under Linux and Windows. Mol Ecol Resour 2010;10(3):564-567. [http://dx.doi.org/10.1111/
22. Mendez FL, Karafet TM, Krahn T, Ostrer H, Soodyall H, Hammer MF. Increased resolution of Y
chromosome haplogroup T denes relationships among populations of the Near East, Europe, and Africa.
Hum Biol 2011;83(1):39-53. [http://dx.doi.org/10.3378/027.083.0103]
23. Sengupta S, Zhivotovsky LA, King R, et al. Polarity and temporality of high-resolution Y-chromosome
distributions in India identify both indigenous and exogenous expansions and reveal minor genetic
inuence of Central Asian pastoralists. Am J Hum Genet 2006 Feb;78(2):202-221. [http://dx.doi.
24. Semino O, Passarino G, Oefner PJ, et al. e genetic legacy of Paleolithic Homo sapiens sapiens in extant
Europeans: A Y chromosome perspective. Science 2000;290(5494):1155-1159. [http://dx.doi.org/10.1126/
25. Wells RS, Yuldasheva N, Ruzibakiev R, et al. e Eurasian heartland: A continental perspective on
Y-chromosome diversity. Proc Natl Acad Sci 2001;98(18):10244-10249. [http://dx.doi.org/10.1073/
26. Wilson JF, Weiss DA, Richards M, omas MG, Bradman N, Goldstein DB. Genetic evidence for dierent
male and female roles during cultural transitions in the British Isles. Proc Natl Acad Sci 2001;98(9):5078-
27. Weale ME, Yepiskoposyan L, Jager RF, et al. Armenian Y chromosome haplotypes reveal strong regional
structure within a single ethno-national group. Hum Genet 2001;109(6):659-674. [http://dx.doi.
28. Oliver R, Atmore A. Africa Since 1800. Cambridge: Cambridge University Press, 1981.
29. Human TN. Symbols in Stone. Unravelling the Mystery of Great Zimbabwe. Johannesburg: Witwatersrand
University Press, 1987.
30. Human TN. Snakes and Crocodiles. Power and Symbolism in Ancient Zimbabwe. Johannesburg:
Witwatersrand University Press, 1996.
Accepted 22 August 2013.