Black Carib - Wiley Online Library

AMERICAN JOURNAL OF HUMAN BIOLOGY 22:36–44 (2010)

Original Research Article

The Garı´funa (Black Carib) People of the Atlantic Coasts of Honduras: Population Dynamics, Structure, and Phylogenetic Relations Inferred from Genetic Data, Migration Matrices, and Isonymy ´ NGEL CARRACEDO3 EDWIN-FRANCISCO HERRERA-PAZ,1* MIREYA MATAMOROS,2 AND A 1 Facultad de Medicina, Universidad Cato´lica de Honduras, Campus San Pedro y San Pablo, San Pedro Sula, Honduras 2 Direccioń de Medicina Forense, Minı´sterio Pu´blice de Honduras, Tegucigalpa M.D.C., Honduras 3 Instituto de Medicina Legal, Grupo de Medicina Geno´mica, Facultad de Medicina, Universidad de Santiago de Compostela, Santiago de Compostela, Galicia, Spain

ABSTRACT The aim of this study is to assess population dynamics, structure, and phylogenetic relations of the populations that inhabit the Caribbean coasts of Honduras: the Garı´funa (or Black Carib) people, an admixture of Black Africans and Red Carib Native Amerindians. Thirteen autosomal tetranucleotide microsatellite markers of the DNA (namely short tandem repeats) were genotyped in samples from the Garifuna communities of Bajamar, in the Department of Corte´s; Corozal, in the Department of Atlańtida; and Iriona, in the Department of Gracias a Dios. Each subject in the study filled a questionnaire with the following information: complete name and surname of participant, and places of birth of the participant, his/her parents, and grandparents. We performed analyses that included determination of migration rates and residence patterns from information of places of birth, fixation indices from genetic data, and analysis of surnames of the sampled subjects (isonymy). Migration matrices showed a migration wave from east to west in the parents and grandparents of the subjects. A raise in migration rates and a shift in predominating residence pattern from neolocality to matrilocality from grandparents to parents were observed. Analysis of isonymy conjunctly with values for FIS in each community showed high endogamy in Bajamar, and recent, high immigration in Iriona. A dendrogram constructed with allele frequencies of the Garifuna and other populations from the Americas, Africa, and Europe revealed the close relationships of this ethnic group with Afro-Caribbean and African Populations. Am. J. Hum. Biol. 22:36–44, 2010. ' 2009 Wiley-Liss, Inc.

The Garifuna is a population with an out of the ordinary, complex history. Admixture in the Lesser Antilles of Arahwak and Carib Native Amerindians from South America originated the Red Island Caribs. In the 17th Century, two ships carrying slaves from West Africa wrecked near the coasts of St. Vincent, the largest island of the archipelago known as St. Vincent and the Grenadines, in the Lesser Antilles. The African slaves acquired some of the Red Carib features (such as their language and tools) and kept the music and religion of their African ancestors. The admixed population came to be the Garı´funa, whose culture is a model of syncretism in the New World (Rivas, 1993). Near the end of 1796, the British Government sent most of the Garinagu (plural for Garı´funa) to Baliceaux Island, located 18 km. south of St. Vincent, where they suffered a dramatic reduction in their number due to hunger and disease. On April 1797, the Garinagu were deported to Honduras, Central America, making their arrival to the island of Roatan, in the Honduran Caribbean. Days later, the Spaniards transported them to the city of Trujillo, which is the same place where Christopher Columbus first arrived (in his fourth trip) to the continental coasts of the Americas (see Fig. 1 for geographic locations). About 2,500 Garinagu settled near Trujillo, founding the villages of Cristales and Rıó Negro. The population expanded from this point, migrating eastward and westward, founding villages throughout the Atlantic coastline of most Central America (for history and ethnography of the Garı´funa people see: Andrade-Coelho, 1955, 2002). C 2009 V

Wiley-Liss, Inc.

Columbus wrote in his journal the word Caniba or Canima, referring to the Red Caribs of the Lesser Antilles, who liked to eat human flesh (Rouse, 1993). The sound then transformed to cannibal or canibal, which is now used in many western languages. Despite of the term Black Carib that is sometimes used in literature instead of Garı´funa, these fishers and farmers, inhabitants of the Caribbean coasts of Honduras are pacific, friendly people; even more, the word Garı´funa has a vegetarian connotation meaning ‘‘manioc eaters.’’ The tri-ethnic origin, a unique history and the enormous population expansion to an estimate of 200,000 individuals in merely two centuries, are only some of the remarkable features of the population under study. Several articles examining the genetics of Garinagu have been published to this day, presenting studies performed in Garifuna populations from the Countries of St. Vincent, Guatemala, and Belize, using classical genetic markers (Crawford et al., 1981, 1982; Crawford, 1983, 1986, 1998; Roberts, 1984) or mitochondrial markers (Monsalve and Hagelberg, 1997; Salas et al., 2005). In the present work, we analyzed the data from genetic profiles *Correspondence to: E. F. Herrera-Paz, Bufete Herrera Doninelli y Asociados. 3 ave. 15 y 16 calle S.O. #124, Barrio La Guardia, San Pedro Sula, Honduras. E-mail: [email protected] or [email protected] Received 25 December 2008; Revision received 11 February 2009; Accepted 11 February 2009 DOI 10.1002/ajhb.20922 Published online 21 April 2009 in Wiley InterScience (www.interscience. wiley.com).

THE GARIFUNA PEOPLE OF HONDURAS

37

Fig. 1. Geographic locations of Honduras, in Central America, and St. Vincent and the Grenadines, in the Lesser Antilles. Dotted arrow shows the path followed by the Garinagu deported from Baliceaux Island to Honduras.

for 13 autosomal STR markers of the DNA (widely used in forensic laboratories around the world), and from questionnaires obtained in three Garifuna communities at the Atlantic coasts of Honduras. The information was used to establish migration rates, population structure, residence patterns, differences between the studied populations, and the relationship between the Garı´funa people and other populations of the world. DNA isolation and genotyping procedures, allele frequencies, and calculations of forensic interest were previously published (Herrera-Paz et al., 2008). MATERIALS AND METHODS Subjects During the summer of the year 2002 blood samples where taken to 209 healthy, unrelated residents at three Garı´funa communities from the Atlantic coasts of Honduras: 73 residents of the community of Iriona, 67 residents of the community of Corozal, and 69 residents of the community of Bajamar. The sampled communities are typical Garifuna villages, and each one is located in a different Department (distribution of Garı´funa communities in Honduras is shown in Fig. 2). Each blood donor completed a questionnaire with the following information: Gender, age, complete name (including two surnames), and place of birth of the subject and his or her parents and grandparents (including community or city, and department). Voluntary informed consent for blood extraction and DNA analysis was obtained from every participant in the study. One of the authors (EFHP) performed explorations of the communities and neighboring cities during the fieldwork, collecting useful information from visual recognition and interviews with subjects and community leaders.

Genotyping One hundred and seventy four individuals were genotyped: 52 from Bajamar, 57 from Corozal, and 65 from Iriona. Allele frequencies were determined for the following CODIS autosomal STR loci: D13S3171 (n 5 170), vWA (n 5 169), D5S818 (n 5 167), D7S820 (n 5 162), FGA (n 5 159), D18S112 (n 5 159), CSF1PO (n 5 158), D3S1358 (n 5 157), D16S539 (n 5 157), D81179 (n 5 157), D21S112 (n 5 156), TH01 (n 5 154), and TPOX (n 5 154). Migration rates To estimate migration rates between departments (here, the highest level of subdivision) we used a procedure similar to the one described by Cavalli-Sforza and Edwards (1967), based on pairwise comparison of the places of birth of parents and their offspring. The method was extended to include both immigration and emigration rates in several geographic areas, and to establish the direction of the migration wave. Briefly, if a parent was born in village 1 localized in department A, and his (or her) offspring was born in village 2 localized in department B, the parent had to move from village 1 to village 2. Such case was scored as an emigration event for department A, and as an immigration event for department B. Then, for a given department, immigration rate would be the proportion of parent-offspring comparisons in which the individual was born in that department, and the parent was born in a different one. Similarly, emigration rate for a department is the proportion of comparisons in which the parent was born in that department, and the individual was born in another. A matrix of n x n was constructed, with n as the number of departments. The emigration rate for department i and the immigration rate for American Journal of Human Biology

38

E.-F. HERRERA-PAZ

Fig. 2. Distribution of Garifuna villages in four Departments of the continental Caribbean coasts of Honduras. The communities are from west to east: A. Department of Corte´s: Masca, Travesıá, Bajamar, and Saraguayna; B. Department of Atlańtida: Tornabe´, San Juan, La Ensenada, Triunfo de la Cruz, Corozal, Sambo Creek, and Nueva Armenia; C. Department of Coloń: Rıó Esteban, Punta Gorda, Guadalupe, San Antonio, Santa Fe, Cristales and Rıó Negro, Barranco Blanco, Santa Rosa de Aguan, Limoń, Punta de Piedra, Cusuna, Ciriboya, Iriona, San Jose´ de la Punta, Sangrelaya and Cocalito; D. Department of Gracias a Dios: Tocamacho, San Pedro de Tocamacho, Coyoles, La Fe, Buena Vista, Bataya, Pueblo Nuevo, and Plaplaya. Straight, vertical lines are schematic and intend to show the limits of each Department in the coastline, and do not represent the real interdepartmental boundaries.

department j are simply described by P the equations: ERi P P P 5 j,j=iDij/ iDij and IRj 5 i,i=jDij/ iDij, respectively, where Dij is the number of parents born in the ith department with his son or daughter born in the jth department. Separate calculations for each generation (grandparents/ parents and parents/subjects) and gender were performed. Residence patterns Crude estimations of residence patterns based on intercommunity migrations in parents and grandparents of subjects were assessed by means of comparisons of places of birth of members of the father-mother-offspring trio. If mother and son (or daughter) were born in the same village and the father in a different one, the event was scored as matrilocality. Inversely, patrilocality events would be those in which father and son were born in the same village and the mother in a different one. Finally, if both parents were born in a different village from that of their son, the event was scored as neolocality. In events in which the three members were born in the same village, the residence pattern cannot be achieved and such cases were scored as ‘‘without migration.’’ Isonymy Popularity of studies based on surname sharing within and between populations (isonymy) have grown because they offer the possibility of testing hypothesis about migration, drift and admixture. As in most of Latin America, in Honduras people use two surnames (the first one of each parent). Some authors have taken advantage of this fact to obtain reliable values of isonymy and to assess residence patterns (Pinto-Cisternas et al., 1985, 1990). We used the surnames obtained from the questionnaires to calculate isonymy values through the approach described by Crow and Mange (1965) extended to include both surnames. Briefly, random component of isonymy within comAmerican Journal of Human Biology

munities (defined as the expected proportion of marriages P isonymous by descent in case of panmixia) is I 5 iPi2, where Pi is the frequency of surname i (in the pooled database of first and second surnames) in a given community. This calculation is equivalent to the expression I 5 P iPiQi, where Pi is the frequency of surname i in the database of first surnames, and Qi is the frequency of surname i in the database of second surnames (Jorde and Morgan, 1987). Additionally, crude estimations of the inbreeding coefficient Fst within populations ([1/4] I) and Fisher’s a (1/I)) can be calculated from isonymic data (Fisher, 1943; Barrai et al., 2001; Rodriguez-Larralde et al., 1998). Because inferences of population structure through isonymy are based on several assumptions that are rarely accomplished, the information is useful only in a comparative rather than absolute manner within a studied territory (reviews in: Colantonio et al., 2003; Jobling, 2001). All isonymy calculations were performed over the surnames of the subjects (only present population). To determine residence patterns from isonymy we used a simple approach: in matrilocal communities, we should observe lower values for isonymy calculated from first surnames, whereas in patrilocal communities lower values would correspond to isonymy from second surnames. We note that, in Hispanic populations that use two surnames and have a patrilineal transmission of first surnames, differences in mobility between fathers and mothers may have a profound effect in modifying frequencies of first and second surnames in the offspring population, which will be reflected on isonymy values. For instance, high mobilization of fathers and low mobilization of mothers (in the case of matrilocality) would result in a higher interchange of male’s surnames between locations. Then, an increased diversity of first surnames within communities would be seen in the offspring population and, thus, lower values of isonymy from first surnames when compared to isonymy from second surnames (recall that mothers transmit second surnames). The opposite apply to patrilocality.


39

TABLE 1. Village (or city) and department of birth of three generations


of Garinagu (subjects, parents, and grandparents)

of Garinagu (subjects, parents, and grandparents)

Place of birth

Subjects (number)

Parents (number)

Grandparents (number)

Santa Rosa de Aguan, Coloń Iriona, Coloń Trujillo, Coloń Limoń, Coloń Santa Fe, Coloń Tornabe´, Atlańtida Sangrelaya, Coloń Triunfo de la Cruz, Atlańtida Armenia, Atlańtida San Juan, Atlańtida La Ceiba, Atlańtida Saba´, Coloń Tela, Atlańtida San Antonio, Coloń Jutiapa, Atlańtida Belize, Belizea Puerto Castilla, Coloń Total

29 23 8 2 2 2 1 1 0 1 1 1 1 0 0 0 0 72b

69 29 16 12 4 3 4 2 1 0 0 0 0 2 1 0 0 143c

137 61 32 22 8 6 10 4 0 2 0 0 0 4 0 1 1 288c

Sample from Iriona, Coloń. a Foreign country. b One subject didn’t know his place of birth. c One or more subjects didn’t know place of birth of one or both parents. d One or more subjects didn’t know place of birth of one or more grandparents.

Fixation indices From data obtained from the genotyped markers, we used the emulator of Wright’s fixation indices based on variance partition (Weir and Cockerham, 1984) to estimate differentiation of pairs of populations (Fst) and the within populations component of inbreeding (FIS). Mean values were obtained from the 13 genotyped loci. High Fst values in pairwise analyses would indicate differentiation, whereas high FIS values would be observed in communities with some degree of endogamous behavior, or strong genetic drift and high inbreeding (as seen in isolated groups) or, alternatively, admixture within the community (Wahlund effect); however, it is difficult to differentiate one from the other. For this purpose, several approaches have been used (for example the regression model of Harpending and Ward, 1982). In our study, isonymy values within communities were analyzed conjunctly with FIS values to roughly address the issue of endogamy versus admixture. Briefly, in a community with a high FIS value, high isonymy would point toward endogamy, whereas low isonymy would be an evidence of high gene flow from outside the community. Phylogenetic analysis For phylogenetic analysis, we used allelic frequencies distributions of the 13 genotyped markers of the (pooled) Garı´funa sample and of several populations of Africa, The Americas, and Europe. Pairwise genetic distance was computed using Nei’s Da Normalized Distance (Nei et al., 1983). Dendrograms were drawn using Unweighted Pair Group Method with Arithmetic Mean (UPGMA) and Neighbor Joining (NJ) as clustering methods. The following populations were analyzed: Izalco, Conchagua, Pachimalco, San Alejo (Gomez et al., 2007), Kichwas, Afro-descendant from Ecuador (Gonza´lez-Andrade et al., 2006), Choco´ and San Andre´s Island (Paredes et al., 2003), Costa Rican (Rodrı´guez et al., 2007); Trinidian, Bahamian, Jamaican, Afro-American, Caucasian, Latin-American

Place of birth

Subjects (number)

Parents (number)


Corozal, Atlańtida La Ceiba, Atlańtida Rio´ Esteban, Coloń Trujillo, Coloń Cristales, Coloń Santa Fe, Coloń San Antonio, Coloń La Mosquitia, Gracias a Dios San Pedro Sula, Corte´s Santa Rosa de Aguan, Coloń Sambo Creek, Atlańtida Guadalupe, Coloń Sangrelaya, Coloń Limoń, Coloń Rıó Negro, Coloń Triunfo de la Cruz, Atlańtida Iriona, Coloń Nueva Armenia, Atlańtida Roatan, Islas de la Bahia Buena Vista, Gracias a Dios Armenia, Atlańtida Corte´s, Corte´s Grand Caiman Islanda San Esteban, Coloń Tegucigalpa, Francisco Morazań Tornabe, Atlańtida Total

47 8 4 2 2 1 1 1 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 67

78 3 2 8 5 9 6 0 0 4 4 3 2 2 2 1 1 1 1 1 1 0 0 0 0 0 134

129 3 8 17 8 19 23 0 0 10 8 8 5 3 3 7 5 2 2 1 0 2 1 1 1 1 267b

Sample from Corozal, Atlańtida. a Foreign country. b One or more subjects didn’t know place of birth of one or more grandparents.

(Budowle et al., 1999), Guine´-Bissau (Gonçalves et al., 2002), Maputo (Alves et al., 2004), Black from South Africa (Kido et al., 2007), Equatorial Guinea (Alves et al., 2005), Cabinda (Beleza et al., 2004), Spain (Camacho et al., 2007) and Honduras (Matamoros et al., 2008). RESULTS Community leaders and subjects of the three communities provided important information during interviews. In most cases, subjects knew the names, places of birth and histories of more than two generations of ancestors. The history of the communities is poorly documented in literature, but orally transmitted and preserved from one generation to the next, a fact that helped us reconstruct (although roughly) the peopling pattern of the coast. Tables 1–3 show the distribution of places of birth of the subjects, parents, and grandparents. Although the number of sampled individuals is small, the number of parents and grandparents is almost two and four times bigger respectively, allowing the performance of more accurate migration estimates using the two generations of ancestors. An admixture of individuals born in Iriona and in Santa Rosa de Aguan (a Garifuna village located at 62 km., straight-line distance, west of Iriona) mostly composes the sample from Iriona. This sample showed the highest dispersion in the number of places of birth of the subjects. The sample from Bajamar showed the lowest dispersion. Migration rates of men and women from parents and grandparents are graphically displayed in Figure 3. For both, immigration rates are high in Corte´s and Atlańtida and low in Coloń, whereas the opposite occurs with emigration rates: low values are seen in Corte´s and high valAmerican Journal of Human Biology

40

E.-F. HERRERA-PAZ


of Garinagu (subjects, parents, and grandparents) Place of birth

Subjects (number)

Parents (number)


Bajamar, Corte´s Corte´s, Corte´s San Pedro Sula, Corte´s Rıó Tinto, Atlańtida Tela, Atlańtida Triunfo de la Cruz, Atlańtida Travesıá, Corte´s San Juan, Atlańtida Buena Vista, Gracias a Dios Masca, Corte´s Santa Fe, Coloń Santa Rosa de Aguan, Coloń Limoń, Coloń Saraguaina, Corte´s Sambo Creek, Atlańtida Trujillo, Coloń Tulian, Corte´s Iriona, Coloń Olanchito, Corte´s Puerto Barrios, Guatemalaa Belize, Belizea Livingston, Guatemalaa Armenia, Atlańtida Batalla, Gracias a Dios Esparta, Atlańtida Guatemala, Guatemalaa La Ensenada, Atlańtida Patuca, Gracias a Dios Rıó Esteban, Coloń Roatań, Bay Islands Tornabe´, Atlańtida Total

55 3 3 2 1 1 1 1 1 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 69

82 3 0 13 7 5 4 2 2 1 5 4 2 1 1 1 1 1 1 1 0 0 0 0 0 0 0 0 0 0 0 137b

152 3 0 29 10 13 8 2 0 3 9 13 5 4 2 2 2 1 0 0 4 3 1 1 1 1 1 1 1 1 1 274c

Fig. 3.

Emigration and immigration rates per department.

Sample from Bajamar, Corte´s. a Foreign country. b One or more subjects didn’t know place of birth of one or both parents. c One or more subjects didn’t know place of birth of one or more grandparents.

ues in Atlańtida and Coloń. A general increment in migration rates (more markedly in men) is observed in parents with respect to grandparents. These elements together show a clear migration wave going from east to west, which grows in amplitude from grandparents to parents. The frequencies of the three possible residence patterns calculated from inter-community migration events are shown in Figure 4. In grandparents, neolocality predominates (51%), while patrilocality and matrilocality show approximately equal values (25%). In parents, matrilocality increased up to 47%, while percentage of neolocality and patrilocality decreased. In general, we observed a drastic increment in the fraction of events that presented any kind of inter-community migration: from 28% in grandparents, to 50% in parents (see Fig. 5). Table 4 shows pairwise Fst values from genetic data and geographical distances between communities. A low, non-significant p value (a 5 0.05) in pairwise Fst analysis for the Corozal-Bajamar pair, and high, significant values in the other comparisons were obtained. The correlation coefficient of Fst values plotted against values of coastal line distances (CLD) was 0.82. Table 5 shows the occurrence of the most frequent surnames. Most Garı´funa surnames are from Spanish origin, adopted from priests, important Honduran Mestizo (admixture of Amerindian and Spaniard) or Criollo (descendants of Spaniards born in the Americas) personalities, but a few are English or derived from the original African names. One hundred and twenty six surnames American Journal of Human Biology

Fig. 4. Residence patterns in parents and grandparents of the subjects inferred from inter-community migrations.

Fig. 5. Inter-community migrations in parents and grandparents of the subjects. The percentage of events in which a change in community of residence of one or both parents was registered is labeled as with migration.

were found in the pooled (first and second surnames from the three communities) sample. Interestingly, findings in the analysis of isonymy within populations, along with their respective FIS values correspond with the observed migration and residence patterns in the communities. Higher isonymy values for second

THE GARIFUNA PEOPLE OF HONDURAS TABLE 4. Distance between communities and Pairwise FST values Pairs of populations

CLD (Km)

Iriona-Corozal Bajamar-Iriona Corozal-Bajamar a b

TABLE 6. Values of isonymy (I) from first and second surnames,

and FIS

FSTb

a

196 320 125

0.0112 0.0119 0.0042

Population Iriona Corozal Bajamar

Coastline distance. Calculated from genetic data.

41

a

I (first surname)

I (second surname)

I (pooled first and second surnames)

FISa

0.025333083 0.033637781 0.032556186

0.031144614 0.054577857 0.04641882

0.022183642 0.035754065 0.031821046

0.0327 0.0004 0.0746

From genetic data.

TABLE 5. Most frequent surnames in the pooled Garı´funa sample Number of surnames (occurrence) Surname

Iriona

Corozal

Cortes

Overall

Martinez Mejıá Flores Guity ´ lvarez A Chavez Fernańdez Bernardez ´ vila A Arriola Rochez Arzu´ Garcıá David Castillo Norales Lambert Coloń Lo´pez Gutie´rrez Suazo Nuń˜ez Ballesteros Dolmo Sanchez Amaya Cordoba Ramos

6 4 9 2 4 2 1 6 5 5 4 2 2 1 6 4 0 1 4 3 2 1 1 4 2 1 1 1

9 1 4 12 3 9 10 3 1 3 3 1 4 7 1 3 1 1 1 3 3 1 2 0 2 2 4 1

6 16 2 0 4 0 0 1 4 1 2 6 2 0 0 0 6 5 1 0 1 4 3 1 1 2 0 3

21 21 15 14 11 11 11 10 10 9 9 9 8 8 7 7 7 7 6 6 6 6 6 5 5 5 5 5

surnames when compared to first surnames were observed in the three communities (Table 6), which may be a signature of the high mobility of men between locations in parents. Iriona showed a high FIS and the lowest isonymy, most probably caused by admixture. A high FIS value together with high isonymy, point to high endogamy in Bajamar. Dendrograms showing the relationships between 22 populations of the Americas, Africa and Europe, including the Garı´funa, are shown in Figure 6. Both dendrograms show clustering of populations with their respective groups (Amerindian, Caucasoid or Afro-descendant), with some few exceptions. Particularly, Native Amerindians from San Alejo and blacks from Ecuador appear to be related to Caucasians and Hispanics, which is most probably due to high gene flow from Mestizo populations in close proximity to their homelands. When UPGMA and NJ trees were compared, we found some similarities in the relationships of populations within the Afro group. For instance, Maputo and Blacks from South Africa group together in both trees. Afro-descendants from the Americas tend to group together. Garinagu appear to be close to Afro-Caribbean populations in NJ tree, whereas in UPGMA tree show a closer

relationship with Africans, primarily with inhabitants of Guine´-Bissau. DISCUSSION Important founder events (that we refer to as bottlenecks) took place at the time of the shipwrecks, and in Baliceaux Island. We expect from each one of these events a change in allelic frequencies with respect to the original populations. Particularly, reduction of population during the period of captivity in Baliceaux might have been critical in approximating frequencies to present values. We found some evidence of these bottlenecks in an analysis of private alleles. For instance, alleles 46.2 and 25.2 of the FGA locus were found in relatively high frequencies in the pooled Garı´funa sample (0.025 and 0.0503, respectively). After performing a worldwide database search of published population studies that included in their analysis the FGA locus, both alleles were found to be present (however in very low frequencies) in few populations (Fregeau et al., 1998). Additionally, the low mean expected heterocygocity (calculated over the loci) when compared to other populations (data not shown) indicates a low variability despite of the multi-ethnic origin, which is in accordance with the analysis of mitochondrial haplogroups performed by Salas et al. (2005). Lower isonymy values from first surnames in the subjects of the three Garifuna communities confirm matrilocality as the predominant residence pattern in parents, a fact also inferred from places of birth of members of family trios. Lower random isonymy for first, second and pooled surnames obtained from the sample from Iriona when compared with the other communities, a high FIS value, and the different places of birth of the subjects are all congruent with recent, high gene flow from the outside in this community. The opposite is seen in Bajamar, where high isonymy and high FIS are compatible with a genetic isolate. Interestingly, low FIS value in Corozal is close to the null value expected for a panmictic population. It is probable that most of the subjects born in Santa Rosa de Aguan in our sample from Iriona were temporal immigrants after the destruction caused by hurricane Mitch in that community in 1998 (for information on devastation of Santa Rosa de Aguan, see Mastin and Olsen, 2002). Furthermore, we noted that some individuals from Iriona showed phenotypic features that are not common in Garinagu, such as green eyes, a lighter skin tone, or an Amerindian appearance. These characteristics could denote genetic contribution from the Miskitos, Amerindians with strong Black African and English components that inhabit a wide area of tropical rain forest known as the Honduran Moskitia, in the Department of Gracias a Dios. In the 19th century, the Miskito King allowed a popAmerican Journal of Human Biology

42

E.-F. HERRERA-PAZ

Fig. 6.

Dendrograms of Garı´funa and other 21 populations. A: UPGMA. B: Neighbor joining.

ulation of fugitive Garinagu (who were part of a defeated army that battled against the Central American Confederation) to reside in Miskito communities in western Gracias a Dios (Andrade-Coelho, 2002). Today, both ethnic groups share some locations in that department (a community named ‘‘Las Marıás,’’ for instance). Miskito communities are separated from Garifuna communities of eastern Coloń by relatively small distances, thus, a genetic contribution from Miskitos is highly probable in population of Iriona. Other ethnic groups that might have contributed to admixture in Iriona are Ladinos (a term usually used for Mestizos without a clear ethnic affiliation), and the Native Amerindian minorities known as the Pech and the Tawahkas of Gracias a Dios (Rivas, 1993). Bajamar is located in the western coast. Even though Garı´funa villages extend westward to the countries of Guatemala and Belize, migrations farther west from the Department of Corte´s are in some way limited by the Honduras-Guatemala Border. We could observe the preservation of traditional Garifuna customs in Bajamar, as revealed by the architectural style of the houses, built entirely of hatched huts. High FIS and isonymy values in Bajamar reveal a structured population with low diversity. There are two probable causes for this finding: (1) Strong genetic drift due to small population size, with low genetic contribution from the outside (genetic isolate), and/or (2) endogamic behavior. Garinagu exhibit among their traditions a socially accepted type of polygyny in which female partners do not share the same house. Under this circumstance, the existence of marriages of half siblings (who do not know they are) in the paternal line, and the presence of undetected half siblings in the sampled subjects are possibilities that cannot be discarded. In contrast, if we assume that proximity of resiAmerican Journal of Human Biology

dences is a factor that could promote marriage between a man and a woman, the extended matrifocal family structure that predominates in Garı´funa culture could contribute to marriages between close relatives in the maternal line (family structure, marriage, and migration customs of Garinagu are described in Porter, 1983). It is worth to say that both, Bajamar (together with its twin village Travesıá) and Corozal are close to important Central American harbors (the cities of Puerto Corte´s and La Ceiba, respectively). We noted that, although most of the people of Puerto Corte´s (mainly Caucasoids and Mestizos) are phenotypically different from the Garinagu residents of Bajamar, many people from La Ceiba exhibit African characteristics, denoting high gene flow from Garı´funa population. Historical migrations between La Ceiba and Corozal might have increased diversity reflected in low isonymy values. The following lines contain a rough description of the peopling of the Honduran Caribbean coast by the Garinagu, constructed from information from interviews, cited literature, and results of this study; however, some elements of this reconstruction are (inevitably) speculative: After arrival and settlement in Trujillo in 1797, a fast dispersal through the coast began to take place. At that time, migration of whole families to form new villages prevailed. Small number of families from a village used to move to a new unpopulated beach searching for good fishing zones and new agricultural land, thus founding a new village. The peopling process initiated in the founder villages near Trujillo continued throughout the 19th century, extending eastward and westward. It is possible that serial bottlenecks that might have been produced upon foundation of new villages could have contributed in driving the populations to a rapid divergence, but main factors for micro-


differentiation should have been high endogamy and genetic drift, and in some cases, admixture with Mestizos or local Native Amerindians. By the 20th century, most of the villages were already established. High mobilization of men began to take place between communities and to the cities, for the most part searching for better work opportunities (chiefly in citrus and banana plantations) in Departments of Corte´s and Atlańtida (for a description of work patterns among Garinagu, see Kerns, 1982). High mobilization must have been intensified by the construction of paved roads that joined the different villages and cities, though influenced by the itinerant nature of Garı´funa men, the extended matrifocal family structure, and the practice of polygyny with female partners living in different communities. Our data suggest that the resulting shift of predominant residence pattern from neolocality (peopling pattern) to matrilocality must have occurred mainly during the second half of the 20th century. Regardless of the high migration rates between Garı´funa villages now, pairwise Fst analysis from genetic data reveals that the magnitude of these migrations has not been enough to counteract microdifferentiation. Pairwise Fst values obtained here are in accordance with results from several distance methods, most of which confirmed a high genetic similarity between populations of Corozal and Bajamar (data not shown). Although the differences found between communities could be due to Type 1 error, the high correlation between geographical and genetic distances suggests isolation by distance throughout a continuous, one-dimensional range along the Caribbean coast (Wright, 1943). This is in accordance with demonstrations that geographical distance is a primary determinant for human genetic divergence (Handley et al., 2007; Ramachandran et al., 2005); an exceptional fact in this particular case, if one considers the relatively short time (200 years) and distances involved. Apparently, the cultural component from Indo-Americans is larger than the genetic component in Garı´funa populations of Central America. The predominant African genetic contribution (relative to Amerindian) has been revealed through various studies (Crawford et al., 1981; Roberts, 1984; Monsalve and Hagelberg, 1997; Salas et al., 2005). Garinagu of Honduras are proud of their African ancestry and unique history, and of their struggle for their land and culture. History of migration from Africa to St. Vincent Island and then to Honduras and other locations (which is often referred to as ‘‘The Garı´funa Diaspora’’) represents a strong element of social cohesion, and a conducting thread of Garı´funa life. Although many South American influences can be seen in this ethnic group, such as the dialect (the Ign˜eri), agricultural methods, and tools, Black Caribs of the Honduran Coasts consider themselves Africans in heart and soul. Cultural ‘‘mimesis’’ (i.e., their ability to adapt to different cultures while maintaining their original African roots), together with the predominantly African genetic background and similar environments in Central America and Africa, must have aid in the adaptation to the new territories, driving the population to a colossal demographic expansion. In both (UPGMA and NJ) trees, Garinagu cluster with the Afro-descendant group denoting a foremost African genetic contribution; even though, lack of consistency between the two trees regarding the relationships of Garinagu within the group, and a relatively long branch connecting with the other Afro-descendants, are

43

additional evidences of a founder event and a fast process of microdifferentiation. ACKNOWLEDGMENTS The authors are thankful to the Garı´funa people at the communities for their collaboration in the fieldwork; to Dr. Rita Baldrich, at Universidad Nacional de Colombia, for her priceless collaboration in the completion of the study; and to Mr. Raphael Molanfy, at Valle de Sula Institute, and Mr. Elton Herrera, for help in English editing. EFHP is especially thankful to Pastor Roberto Marroquıń for encouraging words. LITERATURE CITED Alves C, Gusmao L, Damasceno A, Soares B, Amorim A. 2004. Contribution for an African autosomic STR database (AmpF/STR Identifiler and Powerplex 16 System) and a report on genotypic variations. Forensic Sci Int 139:201–205. Alves C, Gusmao L, Lo´pez-Parra AM, Soledad-Mesa M, Amorim A, ArroyoPardo E. 2005. STR allelic frequencies for an African population sample (Equatorial Guinea) using AmpFlSTR Identifiler and Powerplex 16 kits. Forensic Sci Int 148:239–242. Andrade-Coelho RG. 1955. The Black Carib of Honduras: a study in acculturation. PhD. dissertation in Anthropology. Illinois: Northwestern University Press. Andrade-Coelho RG. 2002. Los Negros Caribes de Honduras, 2nd ed. Tegucigalpa: Editorial Guaymuras. Barrai I, Rodriguez-Larralde A, Mamolini E, Manni F, Scapoli C. 2001. Isonymy structure of USA population. Am J Phys Anthropol 114:109– 123. Beleza S, Alves C, Reis F, Amorim A, Carracedo A, Gusmao L. 2004. 17 STR data (AmpF/STR Identifiler and Powerplex 16 System) from Cabinda (Angola). Forensic Sci Int 141:193–196. Budowle B, Moretti TR, Baumstark AL, Defenbaugh DA, Keys K. 1999. Population data on the thirteen CODIS core short tandem repeat loci in African Americans, U.S. Caucasians, Hispanics, Bahamians, Jamaicans, and Trinidians. J Forensic Sci 44:1277–1286. Camacho MV, Benito C, Figueiras AM. 2007. Allelic frequencies of the 15 STR loci included in the AmpFlSTR1 IdentifilerTM PCR Amplification Kit in an autochthonous sample from Spain. Forensic Sci Int 173:241–245. Cavalli-Sforza LL, Edwards AW. 1967. Phylogenetic analysis: models and estimation procedures. Am J Hum Genet 19:(3 Part 1)233–257. Colantonio SE, Lasker GW, Kaplan BA, Fuster V. 2003. Use of surname models in human population biology: A review of recent developments. Hum Biol 75:785–807. Crawford MH, Gonzales NL, Schanfield MS, Dykes DD, Skradski K, Polesky HF. 1981. The Black Caribs (Garı´funa) of Livingston, Guatemala: Genetic markers and admixture estimates. Hum Biol 53:87–103. Crawford MH, Dykes DD, Skradsky K, Polesky HF. 1982. Blood group, serum protein, and red cell enzyme polymorphisms, and admixture among the Black Caribs and Creoles of Central America and the Caribbean In: Crawford MH, editor. Current developments in anthropological genetics, Vol. 3: Black Caribs: a case study of biocultural adaptation. New York: Plenum Press. Crawford MH. 1983. The anthropological genetics of the Black Caribs (Garifuna) of Central America and the Caribbean. Yearbk Phys Anthropol 25:155–186. Crawford MH. 1986. Origin and maintenance of genetic variation on Black Carib populations of St. Vincent and Central America, In: Roberts DF, De Stefano G, editors. Genetic variation and its maintenance in tropical populations. Cambridge: Cambridge University Press. Crawford MH. 1998. Population structure of Native Americans. In: The origins of Native Americans. Evidence from anthropological genetics. Cambridge: Cambridge University Press. Crow JF, Mange A. 1965. Measurements of inbreeding from the frequency of marriages between persons of the same surnames. Eugen Quart 12:199–203. Fisher RA. 1943. The relation between the number of species and the number of individuals in a random sample of animal population. J Anim Ecol 12:42–58. Fregeau CJ, Tan-Siew WF, Yap KH, Carmody GR, Chow ST, Fourney RM. 1998. Population genetic characteristics of the STR Loci D21S11 and FGA in eight diverse human populations. Hum Biol 70:813–844. Gomez JL, Salas A, Carracedo A. 2007. Microsatellite autosomal genotyping data in four indigenous populations from El Salvador. Forensic Sci Int 170:86–91.

American Journal of Human Biology

44

E.-F. HERRERA-PAZ

Gonçalves R, Jesus J, Fernandes AT, Brehm A. 2002. Genetic profile of a multi-ethnic population from Guine´-Bissau (West African Coast) using the new PowerPlex1 16 System kit. Forensic Sci Int 129:78–80. Gonza´lez-Andrade F, Sańchez-Q D, Martıńez-Jarreta B. 2006. Genetic analysis of the Amerindian Kichwas and Afroamerican descendants populations from Ecuador characterized by 15 STR-PCR polymorphisms. Forensic Sci Int 160:231–235. Handley LJ, Manica LJ, Goudet AJ, Balloux F. 2007. Going the distance: Human population genetics in a clinal world. Trends Genet 23:432– 439. Harpending HC, Ward R. 1982. Chemical systematics and human evolution. In: Nitecki M, editor. Biochemical aspects of evolutionary biology. Chicago, IL: University of Chicago Press. p. 213–256. Herrera-Paz EF, Garcıá LF, Aragoń-Nieto I, Paredes M. 2008. Allele frequencies distributions for 13 autosomal STR loci in 3 Black Carib (Garı´funa) populations of the Honduran Caribbean coasts. Forensic Sci Int Genet 3:e5–e10. Jobling MA. 2001. In the name of the father: surnames and genetics. Trends Genet 17:353–357. Jorde LB, Morgan K. 1987 Genetic structure of the Utah Mormons: isonymy analysis. Am J Phys Anthropol 72:403–412. Kerns V. 1982. Structural continuity in the division of men’s and women’s work among the Black Carib (Garı´funa). In: Loveland CA, Loveland FO, editors. Sex roles and social change in Native Lower Central American societies. Illinois: University of Illinois Press. Kido A, Dobashi Y, Fujitani N, Hara M, Susukida R, Kimura H, Oya M. 2007. Population data on the AmpFlSTR Identifiler loci in Africans and Europeans from South Africa. Forensic Sci Int 168:232–235. Mastin MC, Olsen TD. 2002. Fifty-year storm-tide flood-inundation maps for Santa Rosa de Aguań, Honduras. U.S. Geological Survey Open-File Report 02-258. U.S. Department of Interior, U.S. Geological Survey, prepared in cooperation with the U.S Agency for International Development. Available on line at: http://pubs.usgs.gov/of/2002/ ofr02258/. Matamoros M, Pinto Y, Inda FJ, Garcıá O. 2008. Population genetic data for 15 STR loci (Identifiler trademark kit) in Honduras. Leg Med (Tokyo, Japan). 10:281–283. Monsalve JV, Hagelberg E. 1997. Mitochondrial DNA polymorphisms in Carib people of Belize. Proc R Soc Lond B Biol Sci 264:1217–1224.

American Journal of Human Biology

Nei M, Tajima F, Tateno Y. 1983. Accuracy of estimated phylogenetic trees from molecular data. II. Gene frequency data. J Mol Evol 19:153–170. Paredes M, Galindo A, Bernal M, Avila S, Andrade D, Vergara C, Rincoń M, Romero RE, Navarrete M, Ca´denas M, Ortega J, Suarez D, Cifuentes A, Salas A, Carracedo A. 2003. Analysis of the CODIS autosomal STR loci in four main Colombian regions. Forensic Sci Int 137:67–73. Pinto-Cisternas J, Pineda L, Barrai I. 1985. Estimation of inbreeding by isonymy in Iberoamerican populations: an extension of the method of Crow and Mange. Am J Hum Genet 37:373–385. Pinto-Cisternas J, Rodriguez-Larralde A, Castro de Guerra D. 1990. Comparison of two Venezuelan populations using the coefficient of relationship by isonymy. Hum Biol 62:413–419. Porter RW. 1983. El estilo migratorio de vida en la biografıá Garı´funa. YAXKIN. Instituto Hondurenõ de Antropolo´gia e Historia 6:56–65. Ramachandran S, Deshpande O, Roseman CC, Rosenberg NA, Feldman MW, Cavalli-Sforza LL. 2005. Support from the relationship of genetic and geographic distance in human populations for a serial founder effect originating in Africa. Proc Natl Acad Sci U S A. 102:15942–15947. Rivas RD. 1993. Pueblos Indı´genas y Garı´funa de Honduras (una caracterizacioń). Tegucigalpa: Editorial Guaymuras. pp 257–261. Roberts DF. 1984. Anthropogenetics in a Hibrid Population. The Black Caribs studies. In: Crawford MH, editor. Current developments in anthropological genetics, Vol 3. A Case Study in Biocultural Adaptation. New York. Plenum Press. Rodrı´guez A, Arrieta G, Sanoú I, Vargas MC, Garcıá O, Yurrebaso I, Pe´rez JA, Villalta M, Espinoza M. 2007. Population genetic data for 18 STR loci in Costa Rica. Forensic Sci Int 168:85–88. Rodriguez-Larralde A, Scapoli C, Beretta M, Nesti C, Mamolini E, Barrai I. 1998. Isonymy and the genetic structure of Switzerland. II. Isolation by distance. Ann Hum Biol 25:533–540. Rouse I. 1993. Rise and decline of the people who greeted Colombus. New Haven: Yale University Press. Salas A, Richards M, Lareu MV, Sobrino B, Silva S, Matamoros M, Macaulay V, Carracedo A. 2005. Shipwrecks and founder effects: divergent demographic histories reflected in Caribbean mtDNA. Am J Phys Anthropol 128:855–860. Weir BS, Cockerham CC. 1984. Estimating F-statistics for the analysis of population structure. Evolution 38:1358–1370. Wright S. 1943. Isolation by distance. Genetics 28:114–138.