Mitochondrial DNA Genetic Diversity among Four Ethnic Groups in Sierra Leone

Bruce A. Jackson¹, Jamie Lee Wilson², Salwa Kirbah³, Sheree S. Sidney³, Joshua Rosenberger³, Larry Bassie⁴, Joe A. D. Alie⁵, David C. McLean⁶, W. Timothy Garvey^6*, and Bert Ely³.

¹Department of Biochemistry, Boston University School of Medicine, 715 Albany Street, Boston, MA 02118; ²Department of Chemistry, University of Massachusetts, Lowell, MA 01854; ³Department of Biological Sciences, University of South Carolina Columbia, SC, 29208; ⁴Department of Biological Sciences and ⁵Department of History, Fourah Bay College, University of Sierra Leone, Freetown, Sierra Leone, West Africa; ⁶Division of Endocrinology, Diabetes and Medical Genetics, Medical University of South Carolina, Clinical Sciences Bldg 816, 96 Jonathan Lucas St, Charleston, SC 29425

Running Title: Sierra Leonean Mitochondrial DNA

*Current address: Chair, Department of Nutrition Sciences, Webb Building 232, University of Alabama at Birmingham, 1675 University Boulevard, Birmingham, AL 35294-3360

Address for correspondence and reprints: Dr. Bruce A. Jackson, Department of Biochemistry, Boston University School of Medicine, 715 Albany Street, Boston, MA 02118-2394. Email: bjackon@bu.edu

ABSTRACT

Although there are numerous ethnic groups in Sierra Leone, the Mende and Temne together account for approximately sixty percent of the total population. To see if genetic differences could be observed among ethnic groups in Sierra Leone, the nucleotide sequence of the hypervariable 1 (HV-1) region of mitochondrial DNA (mtDNA) was determined from samples of the two major ethnic groups, the Mende (n=59), and Temne (n=121) and of two of the minor ethnic groups, the Loko (n=29) and Limba (n=67). Among these 276 HV-1 sequences, 164 individual haplotypes were observed. An analysis of molecular variance indicated that the distribution of these haplotypes within the Limba sample was significantly different from that of the other ethnic groups. No significant genetic variation was seen among the Mende, Temne and Loko. These results indicate that distinguishing genetic differences can be observed among ethnic groups residing in historically close proximity to one another. Furthermore, we observed some mitochondrial DNA haplotypes that are common among the Sierra Leone ethnic groups but have not been observed in other published studies of West African ethnic groups. Therefore, we may have evidence for mtDNA lineages that are unique to this region of West Africa.

Previous studies of African mtDNA have focused primarily on discerning the origins of modern humans and the worldwide expansion of human populations. These same studies have demonstrated that Africa contains the greatest level of mtDNA diversity in the world (Ingman et al., 2000). However, there has been no intensive, published study to characterize genetic differences in the mtDNA of the many ethnic groups of Africa that were the sources of slaves. Such a study would be important to African-Americans and Caribbean blacks of African descent whose heritage was lost due to slavery. To this end, the “African-American DNA Roots Project” was created to identify ethnic-specific African mtDNA and Y chromosome haplotypes and determine their presence in the African-American and Caribbean populations.

Numerous historical documents suggest that Sierra Leone was a major source of slaves for the Southern United States. Approximately six percent of slaves bought through Charleston, SC during the final years of legal slave importation (1716 – 1807) emanated from Sierra Leone (Pollitzer, 1993). These African slaves were valued because of their knowledge of rice agriculture and were largely responsible for the financial success of Georgia and South Carolina rice plantations. Historical- and cultural evidence suggests that a Sierra Leone connection exists between the “Gullah” and “Geechee” communities of these states (Pollitzer, 1993). However, the people of Sierra Leone are ethnically diverse. The Mende and Temne ethnic groups, for example, each account for about 30% of the population. The remaining 40% of the population is divided among numerous smaller ethnic groups (Alie, 1990).

The Mende belong to a West African linguistic family called Mande. They were originally located in the Liberian hinterland but are believed to have begun arriving in Sierra Leone in the 18^th century. The early Mende immigrants established groups of settlements based on agriculture and hunting. These settlements were usually founded by one or two families, together with their slaves and/or dependants. Often, a thick strip of forest, in essence a barrier that reduced genetic exchange, was left between one community and the next, partly to avoid friction and partly for protection in war. Initially, the village headman, or head of the dominant family, was probably the ultimate authority. Over time, warlords became absolute rulers because of their ability to conquer weaker villages as well as their ability to offer protection to such villages. In the 19^th century, the Mende began to use warfare to extend their polities and influence. They dislodged some groups associated with the Temne, such as the Banta, and pushed them southward toward the coast. Other Mende leaders later transformed their original farming settlements into fortresses. In many instances, these chiefs made agreements with the original inhabitants, offering protection in return for allegiance. Sometimes they killed the leaders and enslaved their followers. By this method, the Mende rulers were able to create large political entities. Over the years, the Mende absorbed the culture and linguistics of large numbers of smaller neighbouring ethnic groups, and today virtually the entire Southern and Eastern provinces are Mende-speaking. Descent and inheritance are patrilineal, although among the Mende, a man stands in special relationship with his mother’s brother (keinya: Mende for Uncle), whose blessing is considered more important than that of his own father. The Mende, like other ethnic groups, encourage polygamy but marriages with close relatives are prohibited.

The Temne belong to a West Atlantic group that probably migrated into Sierra Leone from Futa Jallon in present-day Republic of Guinea. These migrations began before the 15^th century. By the time Portuguese explorers visited Sierra Leone’s shores from the 1440s, some Temne were already established in the Sierra Leone peninsula. The Temne claim descent from about 25 eponymous ancestors, and they bear the clan names of these ancestors. The clan names include: Bangura, Kargbo, Kamara, Koroma, Fonah, and Thula. Each clan is associated with a symbol or totem, often an animal, bird or plant that members of the clan are forbidden to see, touch, kill or eat. This form of prohibition differs and varies with each clan, and the penalty for non-observance or breach of the prohibition also varies. The clan name is transmitted from father to children and a woman belongs to the clan of her father. The social significance of the clan rested on the bond it created between all men possessing it. Originally, marriages were allowed only outside one’s clan, but later marriages within the clan became frequent. Today, the Temne occupy parts of the Sierra Leone peninsula and large sections of the northern interior. Like the Mende, the Temne have also influenced many ethnic groups in the north, culturally and linguistically (Foray, 1977).

The Limba are the third largest and one of the oldest groups in Sierra Leone. They claim no tradition of origin and maintain that they have always lived around the Wara Wara mountains in the northern interior. Some linguists place the Limba among the West Atlantic group, others put them into a separate category. The Limba language bears little resemblance to any of the other Sierra Leonean languages. The Limba, like the Temne, also have clans. The names, common to the Temne as well, include Kamara, Kargbo, and Conteh. Clan membership is also acquired through the father and no one is allowed to marry someone of the same clan.

The Loko, like the Mende, belong to the Mande linguistic group and are closely related to the Mende and Gbande (of Liberia). Most Loko occupy chiefdoms in Port Loko and Bombali Districts in northern Sierra Leone, while significant numbers are found in Bumpe and Ribbi chiefdoms in Moyamba District and in the Mountain villages overlooking Freetown, Sierra Leone’s capital city and main port. The ancestors of the Loko came to Sierra Leone in the 1550s, probably as Mane warriors from the Cape Mount area in present-day Liberia. Over time, the Loko settled peacefully in many parts of the country. While the Loko are linguistically related to the Mende, they appear to have been heavily influenced by the Temne culture. The Mende and Loko call each other Njagbe (nephew), a mutually unacceptable term. The Mende and Loko languages are intelligible to speakers of both languages.

In this study, we analysed samples from each of the aforementioned four ethnic groups to determine if significant differences could be detected among ethnic groups that have been living in close proximity to one another for several centuries. As a first comparison of these ethnic groups, we determined the nucleotide sequence of the mtDNA HV-1 region to generate a profile of the haplotypic diversity of each group.

METHODS

Sampling strategy

Cheek swabs were obtained from 166 unrelated individuals from the two major ethnic groups Mende (n=55), and Temne (n=28), and two minor ethnic groups Loko (n=29) and Limba (n=54) ethnic groups in Sierra Leone. Samples were collected at pre-arranged gatherings organized by the collaborating Chiefs of each ethnic group. Primarily male subjects were chosen because both matrilineal (mtDNA) and patrilineal (Y-DNA) studies can be conducted on DNA from a single individual. Prior to sampling institutionally (Boston University and University of South Carolina) approved informed consent forms were obtained from all individuals participating in the study. In addition, the maternal and paternal ethnic lineages over three generations as well as relevant regional information were recorded for each individual. Additional samples of Mende (n=22), Limba (n=11), and Temne (n=96) were obtained from Dr. Sahr Gevao at the University of Sierra Leone, Freetown and sent to the Medical University of South Carolina during the period 1996-1998. For this study, all subjects were classified by the maternal ancestor’s ethnic group.

DNA extraction

Total DNA was extracted using the BuccalAmp DNA Extraction kit (Epicentre, Madison, WI) according to the manufacturer’s specifications.

HV-1 amplification

The mtDNA from each subject was PCR amplified using the Platinum Taq Polymerase kit (Invitrogen, Carlsbad, CA). The forward primer: L15926 (5’-TACACCAGTCTTGTAAACC-3’) and the reverse primer H16498 (5’-CCTGAAGTAGGAACCAGATG-3’) were used in both PCR amplification and DNA sequencing. PCR reactions were performed as described by Alves-Silva et al. (2000). All amplified PCR fragments were purified using the Qiaquick PCR Purification Kit (Qiagen, Valencia, CA) and visualized by electrophoresis on a 0.8 % agarose gel with ethidium bromide.

Nucleotide sequence determination

The mtDNA haplotypes of all subjects were determined by analyzing the nucleotide sequence of the hypervariable region I (HV-1). Cycle sequencing was performed using BigDye Terminator V3 chemistry (Applied Biosystems, Foster City, CA) in an ABI PRISM 377 DNA sequencer and with the L primer. Additional sequencing reactions were performed with the H primer if any ambiguity was present in the sequence information generated by the L primer. HV-1 sequences were aligned with the Cambridge Reference Sequence (CRS, Anderson et al. 1981, GenBank accession number J01415) and edited from positions 16001-16480 to ensure the identification of unambiguous polymorphisms. From the observed polymorphisms, haplogroup assignments for each sample were made based on Salas et al. (2002). Approximately 70% of the haplotype assignments were confirmed using PCR-RFLP assays to detect the presence of mutations elsewhere on the mtDNA molecule that are specific for the haplotype as described by Salas et al. (2002). Statistical analyses were performed using the Arlequin software package (Schneider et al., 2000).

RESULTS

Haplotypic diversity

A total of 480 bases of HV-1 nucleotide sequence was determined for DNA samples from 276 individuals resulting in the identification of 164 unique haplotypes. Haplotypic diversities for the four samples ranged from 0.983 to 0.994 (Table 1). Thus, in all of our ethnic group samples most haplotypes were observed only once. However, there was considerable haplotype sharing among the four ethnic groups.

Haplogroup distribution

The three major African haplogroups, L1, L2, and L3A, were present in all four Sierra Leone samples (Table 2). When the major haplogroups were subdivided according to the scheme described by Salas et al. (2002), the combined L2a and L2a1 haplogroups were present at frequencies of 17% to 28% in all four ethnic groups. In contrast, the combined L1b and L1b1 haplogroups were only 4% of the Limba sample, but they were present at frequencies of 12% to 27% among the other samples. Similarly, the L3e4 haplogroup was present in the Limba sample at a frequency of 7% and was absent from the other three samples. When an analysis of molecular variance was performed on haplotypes pooled into the haplogroup classifications described in Table 2, 0.76 % of the genetic diversity could be attributed to among group variation and the remainder to within group variation (Table 3). Despite the low value of among group variation, this result indicated significant population subdivision among the samples. An exact test of sample differentiation based on the haplogroup frequencies in Table 2 indicated that the Limba sample was significantly different from the Mende and Temne samples (Table 4).

Comparison to other West African ethnic groups

Similar mitochondrial haplotype data are available for a number of West African ethnic groups (Graven et al., 1995; Watson et al., 1997; Rando et al., 1999). Sample sizes for three of these ethnic groups, the Mandinka, Fulbe, and Wolof, were larger than 40. Therefore, we compared these three to the three Sierra Leone samples that also were larger than 40 (Table 5). To be consistent with our data, the haplogroup assignments for the published data were redefined based on the scheme of Salas et al. (2002). When an exact test of sample differentiation was performed, we found that the Mandinka sample had haplogroup frequencies that were significantly different from those of the other ethnic groups (Table 6). All other pairwise comparisons were significantly different as well, except for comparisons between the Mende and the Temne, Fulbe or Wolof samples. Haplotype sharing was common. Exceptions include 4 instances of an L3b1 haplotype (93, 223, 278, 362) that was found uniquely in the Fulbe sample and 6 instances of an L2a haplotype (148, 150, 223, 278, 294, 355, 390) that was found uniquely in the Mandinka sample (Watson et al., 1997). Also, an L2c2 haplotype (93, 126, 223, 264, 274, 278, 311, 390) and two L2d2 haplotypes (111, 145, 184, 189, 223, 239, 278, 292, 311, 355, 360, 368, 390, 399, 400 and 111A, 145, 184, 189, 223, 239, 278, 292, 355, 390, 399, 400) that are common among the Sierra Leone ethnic groups have not been observed previously (Table 2). Thus, there are significant differences in mtDNA haplotype distribution among most populations.

DISCUSSION

Haplogroup assignments were determined for a total of 276 mtDNAs obtained from four ethnic groups in Sierra Leone. Two approaches were used to determine the haplogroup assignment of a particular mtDNA, determining the nucleotide sequence of the HV1 region or PCR-RFLP analyses of the remainder of the molecule. In both cases, the results were compared to published information about African mtDNA haplogroups to assign a particular mtDNA to a particular haplogroup (Salas et al., 2002). In almost every case, the same haplogroup assignment was obtained using either approach. In the few cases (several Temne L1b1 haplotypes in Table 2), where there was disagreement between the haplogroup assignment based on the HV1 nucleotide sequence and the PCR-RFLP data, the assignment based on the HV1 nucleotide sequence usually proved to be correct. In these cases, the PCR-RFLP assignment had been based on the presence or absence of a single restriction site at position 10806 that is considered diagnostic for the L1b1 haplogroup. Further analyses of these cases demonstrated the presence of the mutation at position 8655 that is also characteristic of an L1b haplogroup assignment (Herrnstadt et al., 2002). Thus, the haplogroup assignment based on the HV1 nucleotide sequence was correct, and an additional mutation had occurred at the 10806 HinfI site that was being used to identify a particular haplogroup. Therefore, determination of the HV1 nucleotide sequence appears to be the most efficient and reliable approach for the initial assignment of West African mtDNAs to particular haplogroups.

When the four ethnic groups sampled in Sierra Leone were analyzed, we found that the distribution of mtDNA haplogroups was similar to that observed in previous studies of West African ethnic groups. For instance, in all four ethnic groups, 30 to 46% of the individuals had haplogroup L2 mtDNA as observed previously for other West African ethnic groups (Watson et al., 1997).

To provide better resolution of possible differences among ethnic groups, we subdivided the major haplogroups according to the scheme described by Salas et al. (2002). When the distribution of these subdivided haplogroups was examined, the distribution in the Limba sample was significantly different from those of two of the other three Sierra Leone ethnic groups (Table 4). For instance, the 5 members of haplogroup L3e4 in the Limba sample were the only ones found among the Sierra Leone samples. Also, haplogroup L1b was present at a lower frequency in the Limba sample than in any of the other samples. These findings demonstrate that mtDNA haplotypes are not randomly distributed among ethnic groups in Sierra Leone even though the various ethnic groups have been living in close proximity to one another for several centuries. Although it is not possible to determine what factors contributed to the genetic differentiation of the Limba from the other Sierra Leonean ethnic groups, the fact the Limba have lived in the same isolated mountainous terrain for centuries may be important. Also, their unique language may have further reduced genetic exchange. In contrast, the other Sierra Leone ethnic groups in this study have a more mobile history that documents the assimilation of people from other ethnic groups, a process that would promote intragroup diversity and intergroup homogeneity. In Senegal, the Wolof and Mandinka samples haplogroup distributions were significantly different as well (Table 6). In contrast, the Mende were indistinguishable from the Wolof indicating that similar distributions of mtDNA haplogroups can occur between geographically distant ethnic groups while colocalized groups can exhibit different haplotype distributions.

When individual haplotypes were compared within each haplogroup, we found that many of the frequently occurring haplotypes were shared across ethnic groups. However, a few haplotypes that were observed repeatedly in a single ethnic group have not been observed in any other published sequences. Thus, we may have identified some ethnic group specific mtDNA haplotypes. Similarly, the specific L2c2 and L2d2 haplotypes found in the Sierra Leone samples (Table 2) may be unique to this region of West Africa since they have not been observed among published sequences from other parts of West Africa or elsewhere on the continent. If our ongoing analyses of samples from additional West African ethnic groups confirm these results, then these markers may prove valuable in the several research efforts to identify the ethnic origins of American and Caribbean descendents of African slaves (Brehm et al., 2002; Pereira et al., 2001; Bandelt et al., 2001). In contrast, most mtDNA haplotypes in our study were observed only once. Hence, much larger sample sizes will be needed to establish the distribution of these haplotypes among West African ethnic groups. Other mtDNA haplotypes occur commonly among many sub-Saharan ethnic groups. Thus, they will provide little information about the ethnic origins of individuals who possess them.

The differences observed among ethnic groups when mtDNA lineages are compared are likely to be more pronounced when Y chromosome lineages are compared, since most ethnic West African groups are defined by the male lineage. In fact, all four of the Sierra Leone ethnic groups in this study share a patriarchal culture where ethnicity is defined through males. In addition, the clan structure in these ethnic groups would serve to maintain clusters of distinct Y lineages while the requirement to marry a woman from another clan would tend to homogenize mtDNA lineages among clans and perhaps among ethnic groups. Furthermore, the practice of polygamy would give rise to limited male lineages with diverse mtDNAs. Our current studies of the non-recombining region of the Y chromosome will demonstrate how these cultural practices influence the distribution of Y chromosome lineages.

ACKNOWLEDGMENTS

We gratefully acknowledge the people of Sierra Leone who volunteered to participate in this study. In addition, we thank Eric Thelen for his expert technical assistance and Elizabeth Newton for her invaluable feedback with regard to statistical data. We also wish to thank Tom Parsons of the Armed Forces DNA Identification Laboratories for the generous sharing of his Sierra Leonean mtDNA database. This work was supported in part by National Science Foundation CAREER Grant MCB-9707644 to BAJ and REU grant DBI0097667 to BE.

LITERATURE CITED

Alie JAD. 1990. A new history of Sierra Leone. London: Macmillan.

Alves-Silva J, Santos M, Guimarães P, Ferreira AC, Bandelt H-J, Pena SD, Prado VF. 2000. The ancestry of Brazilian mtDNA lineages. Am J Hum Genet 67:444-461

Anderson S, Bankier AT, Barrell BG, de Bruijn MH, Coulson AR, Drouin J, Eperon IC, Nierlich DP, Roe BA, Sanger F, Schreier PH, Smith AJ, Staden R, Young IG. 1981. Sequence and organisation of the human mitochondrial genome. Nature 290:457-465.

Bandelt H-J, Alves-Silva J, Guimarães P, Santos M, Brehm M, Pereira L, Coppa A, Larruga JM, Rengo C, Scozzari R, Torroni A, Prata MJ, Amorim A, Prado VF, Pena SDJ. 2001. Phylogeography of the human mitochondrial haplogroup L3e: a snapshot of African prehistory and Atlantic slave trade. Ann Hum Genet 65:549-563.

Brehm A, Pereira L, Bandelt H-J, Prata MJ, Amorim A. 2002. Mitochondrial portrait of the Cabo Verde archipelago: the Senegambian outpost of Atlantic slave trade. Ann Hum Genet 66:49-60.

Foray CP. 1977. Historical dictionary of Sierra Leone. Metuchen, NJ: The Scarecrow Press, pp210-212.

Graven L, Passarino G, Semino O, Boursot P, Santachiara-Benerecetti S, Langaney A, Excoffier L. 1995. Evolutionary correlation between control region sequence and restriction polymorphisms in the mitochondrial genome of a large Senegalese Mandenka sample. Mol Biol Evol 12:334-345.

Herrnstadt C, Elson JL, Fahy E, Preston G, Turnbull DM, Anderson C, Ghosh SS, Olefsky JM, Beal MF, Davis RE, Howell N. 2002. Reduced-median-network analysis of complete mitochondrial DNA coding-region sequences for the major African, Asian, and European haplogroups. Am J Hum Genet 70:1152-1171.

Ingman M, Kaessmann H, Pääbo S, Gyllensten U. 2000. Mitochondrial genome variation and the origin of modern humans. Nature 408:708-713.

Pereira L, Macaulay V, Torroni A, Scozzari R, Prata MJ, Amorin A. 2001. Prehistoric and historic traces in the mtDNA of Mozambique: insights into the Bantu expansions and the slave trade. Ann Hum Genet 65:439-458.

Pollitzer WS. 1993. The relationship of the Gullah-speaking people of coastal South Carolina and Georgia to their African ancestors. Hist Methods 25:53-67.

Rando JC, Cabrera VM, Larruga JM, Hernández M, González AM, Pinto F, Bandelt H-J. 1999. Phylogeographic patterns of mtDNA reflecting the colonization of the Canary Islands. Ann Hum Genet 63:413-428.

Salas A, Richards M, De la Fe T, Lareu M, Sobrino B, Sanchez-Diz P, Macaulay V, Carracedo A. 2002. The making of the African mtDNA landscape. Am J Hum Genet 71:1082-1111.

Schneider, S., Roessli, D., and Excoffier, L. 2000. Arlequin: A software for population genetics data analysis. Ver 2.000. Genetics and Biometry Lab, Dept. of Anthropology, University of Geneva.

Watson E, Forster P, Richards M, Bandelt H-J. 1997. Mitochondrial footprints of human expansions in Africa. Am J Hum Genet 61:691-704.

Table 1

Haplotypic Diversity of Sierra Leone Samples

Sample	N	Number Haplotypes	Haplotypic Diversity
Mende	59	54	.994
Loko	29	26	.994
Limba	67	48	.983
Temne	121	86	.992

Haplotypic diversity is defined as h = (1 - Σx²_i) n/(n-1), where x is the frequency of a haplotype and n is the sample size.

Table 2. HV-1 nucleotide sequence and coding region RFLP assays in 276 individuals from Sierra Leone.

Haplo-groups	HV-1 Haplotypes	Mende	Loko	Limba	Temne	RFLP Assay
A2	111,187,223,290,319,362	1
H	278				1
L1a1	129,148,168,172,187,188G,189,223,230,301,311,320				1
L1a1	129,148,168,172,187,188G,189,223,230,311,320,325,362,390				2	L1
L1a1	129,148,168,172,187,188G,189,223,230,311,320,355				2
L1a1	129,148,168,172,187,188G,189,223,230,311,320		1	1	1	L1a
L1a1	86,129,148,168,172,187,188G,189,223,230,311,320	1
L1b	126,187,189,223,239,258,264,270,278,311	1				L1b
L1b	126,187,189,223,239,264,270,278,311	2	1	1		L1b
L1b	126,187,189,223,264,270,278,311	1	2	1	2	L1b
L1b	187,189,223,264,270,278,311,362			1		L1b
L1b1	111,126,187,189,223,239,270,278,293,311	2			1	L1b
L1b1	114G,126,187,189,223,264,270,278,293,311				2	L1b
L1b1	114G,126,187,189,264,270,278,293,311	1				L1b
L1b1	126,139,187,189,223,256,264,270,278,293,311		1			L1b
L1b1	126,154,187,189,223,264,270,278,293,311	1			1	L1b
L1b1	126,180,187,189,223,256,264,270,278,293,311	1
L1b1	126,183C,189,223,256,264,270,278,293,311				1	L1b
L1b1	126,183C,189,223,264,270,278,293,311				1	L1b
L1b1	126,183C,189,223,264,270,293,311				1	L1b
L1b1	126,187,189,223,239,264,270,278,292,293,311		1			L1b
L1b1	126,187,189,223,243,264,270,278,293,311	1
L1b1	126,187,189,223,256,264,270,278,287,293,311	1
L1b1	126,187,189,223,264,270,274,278,293,311	1				L1b
L1b1	126,187,189,223,264,270,278,293,311	1		1	1	L1
L1b1	126,187,189,223,270,278,293,311,362,400	2
L1b1	126,187,189,223,270,278,293,311				1	L1
L1b1	126,187,189,264,270,278,293,311	1	1			L1b
L1b1	126,189,223,264,270,278,293,311				2
L1b1	187,189,223,264,270,278,293,311,362				1	L1b
L1c	129,172,184,187,189,223,261,278,311,359,360,390		1
L1c	129,187,189,223,261,278,311,360		1
L1c	86,129,153,172,184,187,189,223,261,278,311,360				2	L1
L1c	86,172,173,184,187,189,223,261,278,311,360				1	L1
L1c1	129,163,187,189,212,223,278,293,294,311,325,360,368			1
L1c1	129,166,187,189,223,278,293,294,311,360				1	L1c
L1c1	129,187,189,223,278,293,294,311,360		1			L1c
L1c1	17,129,163,187,189,223,278,293,294,311,360			1
L1c1	38,86,129,187,189,192,278,293,294,360				1
L1c1	38,86,129,187,189,223,278,291,293,294,360		1
L1c1	86,129,169,172,183C,189,223,261,278,293,311,360	1				L1
L1c1	86,129,169,172,189,223,261,278,293,311,360				1	L1c
L1c1	86,92,129,169,172,183C,189,192,223,261,278,293,311,360				1
L1c1	93,129,169,172,187,189,223,239,261,278,293,311,360			1
L1c1	93,129,187,189,223,256,266,278,293,311,360				1	L1c
L2a	145,189,192,223,278,294,362,390		2	1		L2a
L2a	168,183C,189,223,278,294,355,362,390,460			1		L2a
L2a	183C,185,189,192,223,241,278,292,293,294,390				2	L2a
L2a	183C,185,189,192,223,241,278,292,294,390	1		1		L2a
L2a	183C,189,223,278,294,362,390			1	2
L2a	189,192,223,278,294,311,362,390,430,460			1
L2a	189,192,223,278,294,311,362,390				2	L2a
L2a	189,192,223,278,294,362,390			2	1	L2a
L2a	189,192,223,278,294,380,390	1
L2a	189,223,278,294,362,390,467		1			L2a
L2a	189,223,278,294,362,390		2		2	L2a
L2a	189,223,278,294,390				1	L2a
L2a	193,213,223,239,278,294,390	1				L2a
L2a	193,213,223,239,265,278,294,390		1
L2a1	104,189,223,278,294,309,390			1		L2a
L2a1	124,189,223,278,294,309,390				1	L2a
L2a1	183C,189,223,278,294,309,390				1	L2a
L2a1	189,192,223,269,278,294,309,390				1	L2a
L2a1	189,192,223,278,294,309,390	1			3	L2a
L2a1	189,223,269,278,294,309,390,460			1
L2a1	189,223,278,294,309,390	1	1	1	1	L2a
L2a1	189,223,278,294,362,390				1	L2a
L2a1	223,233,278,294,309,368,390	2				L2a
L2a1	223,278,286,294,309,390	1				L2a
L2a1	223,278,294,309,311,368,390			1
L2a1	223,278,294,309,368,390	1
L2a1	223,278,294,309,368,390			1	1
L2a1	223,278,294,309,390,399			1		L2a
L2a1	223,278,294,309,390	1		1		L2a
L2a1	31,223,256,278,294,309,368,390	1				L2a
L2a1	75,189,213,278,294,309,375,381,390				1
L2a1	75,189,223,278,294,309,390				1
L2a1	93,223,278,294,309,368,390		1			L2a
L2a1	93,223,278,294,309,390			1		L2a
L2b	114A,129,182C,183C,189,213,223,257,278,390		1			L2b
L2b	114A,129,183C,189,213,223,278,390			1		L2b
L2b	93,114A,129,213,223,278,390	1				L2b
L2b1	114A,129,169,213,223,278,355,362,390			1		L2b
L2b1	114A,129,213,223,278,355,359,362,390			1		L2b
L2b1	86,114A,129,213,223,278,342,355,362,390,465				1
L2c	111,223,278,294,390				1	L2c
L2c	168,189,213,223,278				1	L2c
L2c	168,193,223,278,390			1	1	L2c
L2c	176,223,278,390	1
L2c	188,223,278,390				1
L2c	189,223,278,362,390				1
L2c	223,278,302,309	1				L2c
L2c	223,278,311,390				3	L2c
L2c	223,278,390	1		6	4	L2c
L2c	93,223,278,362,390				1
L2c	93,223,278,390			1		L2c
L2c1	223,261,278,318,390				1
L2c1	223,278,318,390				1
L2c2	189,223,264,278,390				1
L2c2	93,126,223,264,274,278,311,390	1	1	4		L2c
L2c2	93,223,264,278,362,390				1
L2d1	129,182C,183C,189,278,300,354,390,399				1
L2d1	129,183C,189,278,300,354,390,399				2
L2d2	111A,145,184,189,223,239,278,292,311,355,360,368,390,399,400				1
L2d2	111A,145,184,189,223,239,278,292,355,362,390,399,400		1	1
L2d2	111A,145,184,189,223,239,278,292,355,390,399,400	2			1
L2d2	111A,145,184,223,239,278,292,311,355,390,399,400				2
L2d2	111A,145,184,239,278,292,311,355,390,399,400	1
L3	184,213,223,278	1
L3	209,223,278	1
L3b	124,145,189,223,278,362,460			1
L3b	124,145,189,223,278,362			1		L3b
L3b	124,148,182C,183C,189,223,278,362	1			1
L3b	124,172,189,223,278,399			1		L3b
L3b	124,223,234,278,362				1
L3b	124,223,243,278,362	1
L3b	124,223,261,278,360,362		1	2		L3b
L3b	124,223,266,278,355,362	1
L3b	124,223,278,355,362	2
L3b	124,223,278,362		1	3	6	L3b
L3b	124,223,278				1	L3b
L3b	48,124,223,278,362		1		1	L3b
L3b	93,124,223,278,355,362				1	L3b
L3b1	214,223,278,362			1
L3b1	223,278,362	1		2
L3b1	51,223,278,318,362				1
L3d	111,124,223	1		2	2
L3d	124,166,223				1
L3d	124,223,399	1	1	1	4	L3d
L3d	124,223				1
L3d	93,124,126,223	1
L3d2	86,124,223,256,368	1	1	1	2	L3d
L3d2	124,172,184,223,256				1
L3d2	124,223,256,291,368		1			L3d
L3d2	124,223,256,292				2	L3d
L3d2	124,223,256,399		1			L3d
L3e1	172,223,311,327,399			2		L3e
L3e2	223,242,320			1	1
L3e2b	172,183C,189,193,223,320	1
L3e2b	172,183C,189,223,259,291,320,467	1				L3e
L3e2b	172,183C,189,223,259,291,320				2	L3e
L3e2b	172,183C,189,223,261,320				1	L3e
L3e2b	172,183C,189,223,320				4
L3e2b	172,189,223,239,295,320				1	L3e
L3e2b	172,189,223,320	1				L3e
L3e2b	189,223,261,320,390				1
L3e4	51,189,223,264,460			1		L3e
L3e4	51,223,264			4		L3e
L3f1	129,209,223,292,295,311	1
L3f1	167,209,223,292,311	1
L3f1	176,209,223,292,295,311				2
L3f1	209,223,224,292,295,311			1
L3f1	209,223,266,274,292,311			2
L3f1	209,223,292,295,311	1			5
L3g	148,223,293,311,362	1
Other	169,187,223,266			1
Other	179,189,192,215,223,256A,284,311				1
Other	179,189,215,223,256A,284,311			1
Other	223,274,399			1
Other	37,223,399				1
U6	172,183C,189,219,278	2			2

Table 3

Analysis of Molecular Variance Among Sierra Leone Samples

Source of Variation	d.f.	Sum of Squares	Variance Components	Percentage of Variation
Among Populations	3	2.089	0.00359 Va	0.76
Within Populations	272	127.208	0.46768 Vb	99.24
Totals	275	129.297	0.47127

Fixation Index FST : 0.00762

Va and FST : P(random value >= observed value) = 0.0147+/-0.0034

Table 4

Exact Test of Sample Differentiation Based on Haplogroup Frequencies

	Non-differentiation Exact P Values
Sample	Mende	Loko	Limba
Loko	0.2134+-0.0073
Limba	0.0044+-0.0028	0.0526+-0.0100
Temne	0.4853+-0.0354	0.2849+-0.0194	0.0128+-0.0071

Table 5

Haplogroup Summaries for Six West African Ethnic Groups

Group	Mende	Limba	Temne	Mandinka	Fulbe	Wolof
L1				2
L1a1	1	1	6	2
L1b	4	2	2	3	1	3
L1b1	12	1	12	19	10	8
L1c			3			1
L1c1	1	3	5
L2a	3	7	10	6	2	3
L2a1	8	8	10	7	7	7
L2b	1	1				3
L2b1		2	1	3		4
L2c	3	8	13	30	2	2
L2c1			2	7		1
L2c2	1	4	2	6
L2d1			3
L2d2	3	1	4			3
L3b	5	8	11	3	8	2
L3b1	1	3	1	3	4	1
L3b2				2
L3d	3	3	8	6	3	2
L3d1					3
L3d2	1	1	5	4
L3d3				1
L3e1		2				1
L3e2		1	1	1	3	1
L3e2b	3		9		4	1
L3e4		5		4		1
L3f					1
L3f1	3	3	7		3
L3g	1
U5					1	1
U6	2		2		2	2
A2	1
H			1
L3A	2		1		1	1
Other		3	2	1	5
Totals	59	67	121	110	60	48

Table 6

Exact Test of Sample Differentiation Based on Haplogroup Frequencies

	Non-differentiation Exact P Values
Sample	Mende	Limba	Temne	Mandinka	Fulbe
Limba	0.0023+-0.0018
Temne	0.5074+-0.0390	0.0093+-0.0030
Mandinka	0.0000+-0.0000	0.0000+-0.0000	0.0000+-0.0000
Fulbe	0.2300+-0.0348	0.0000+-0.0000	0.0065+-0.0042	0.0000+-0.0000
Wolof	0.7264+-0.0302	0.0007+-0.0004	0.0170+-0.0013	0.0000+-0.0000	0.0025+-0.0083

FIGURE LEGENDS

Figure 1. A map of western Africa. The inset shows a map of Sierra Leone and the traditional regions occupied by various ethnic groups. The ethnic groups sampled for this study are marked.