publisher colophon
Sex-Specific Genetic Admixture of Mestizos, Amerindian Kichwas, and Afro-Ecuadorans from Ecuador
Fabricio González-Andrade
Molecular Genetics Laboratory, Metropolitan Hospital, Quito, Ecuador.
Dora Sánchez
Molecular Genetics Laboratory, Metropolitan Hospital, Quito, Ecuador.
Jorge González-Solórzano
Molecular Genetics Laboratory, Metropolitan Hospital, Quito, Ecuador.
Santiago Gascón
Department of Legal Medicine, Faculty of Medicine, University of Zaragoza, Spain.
Begoña Martínez-Jarreta
Department of Legal Medicine, Faculty of Medicine, University of Zaragoza, Spain.
Abstract

Three main ethnic groups live in the South American country of Ecuador: Mestizos, Amerindian natives, and African-derived populations, or Afro-Ecuadorans. Mestizos and Afro-Ecuadorans can be considered trihybrid populations containing genes originating in the Americas, Europe, and Africa, as is the case with equivalent populations in other Latin American countries. The proportion and the dynamics of the admixture process remain unknown. However, a certain sex asymmetry of the admixture process can be expected for historical reasons. We typed 11 Y-chromosome short tandem repeats (STRs) in these three ethnic groups to provide adequate allele and haplotype frequencies for forensic genetic purposes and to quantify admixture proportions in male lineages. In addition, a data set of 15 autosomal STRs in the same samples were reanalyzed for the same purpose. Contributions to Mestizo Y chromosomes were estimated to be 70% European, 28% Amerindian, and 2% African, whereas in autosomes the contributions were 19%, 73%, and 8%, respectively, which underlines the sexual asymmetry in mating, with Europeans contributing mostly males. European Y-chromosome haplotypes in Mestizos were similar to those in Spain. Moreover, about 10% of European Y chromosomes were found in the Amerindian Kichwa. As for Afro-Ecuadorans, their contributions to the male line are 44% African, 31% European, and 15% Native American; the last value is the highest percentage reported so far for an African-derived American group. Autosomal admixture was estimated as 56% African, 16% European, and 28% Amerindian.

Keywords
Kichwas, Afro-Ecuadorans, Mestizos, Ecuador, Y-chromosome STRs, genetic admixture.

Three main ethnic groups live in Ecuador: Mestizos, Amerindian natives, and Afro-Ecuadorans. Mestizos are the most numerous group, with 8 million (or 60% of the total population); they are the Spanish-speaking descendants of Europeans (mostly Spanish) and Amerindian natives. The proportion and dynamics of the processes that caused this admixture of populations remain unknown.

A number of different native Amerindian populations retain their culture, language, and distinct identity in Ecuador. Of these the most numerous by far are [End Page 51] the Kichwa (often spelled Quichua), which number 3 million. The Kichwa language is the result of the absorption of local populations into the former Inca empire (the related but different language spoken by the descendants of the Inca in the core Inca regions of Peru and Bolivia is called Quechua). The Kichwa live mostly in the Andes highlands, but some are found in the Amazonian region (Moya 2000; Vásquez and Saltos 2003). The cultural and dialectal variation between the highlands and the plain and within the highlands itself is considerable. Finally, half a million Ecuadorans descend from African slaves and retain distinct phenotypic features as well as African cultural traits, such as music, dance, and religion. They live in rural areas in two separate provinces, in the Valle del Chota (in the Andes) and in the coastal Esmeraldas province (Moya 2000; Vásquez and Saltos 2003). The extent to which they have absorbed European and native Amerindian gene contributions remains uncharacterized.

We have typed 11 Y-chromosome short tandem repeats (STRs) in these three ethnic groups with a twofold purpose: to provide adequate allele and haplotype frequencies for forensic genetic purposes and to characterize the three groups genetically. Significant genetic differences are expected among Kichwa, Mestizos, and Afro-Ecuadorans, and, accordingly, separate population databases should be used in forensic casework. We have typed the standardized Y-chromosome STRs that are considered best and that are currently most widely used in forensic practice worldwide so that our results can be compared with data from other populations.

Mestizos and Afro-Ecuadorans can be considered trihybrid populations containing genes originating in the Americas, Europe, and Africa in various proportions, as is the case with equivalent populations in other Latin American countries. We aim to quantify these proportions and to ascertain the sex asymmetry of the admixture process by comparing Y-chromosome markers with autosomal markers. Y-chromosome genetic variation is particularly deeply partitioned among populations and among continental groups, which enables precise identification of the origin of each Y chromosome and makes Y-chromosome STRs a good tool for admixture quantification. However, Y chromosomes convey only the male side of the story, and for this reason we have reanalyzed a published data set of 15 forensic autosomal STRs in the same populations (González-Andrade et al. 2003, 2005; González-Andrade and Sánchez 2004).

Materials and Methods

Population Sample

Whole blood was obtained in vacutainer tubes containing EDTA by venipuncture from healthy unrelated Kichwas, Mestizos, and Afro-Ecuadoran populations, of both sexes, born and living in Ecuador. Samples from Kichwas and Afro-Ecuadorans were obtained directly in their communities. Samples of Mestizos were taken from the paternity test bank of our laboratory. All the samples were collected after obtaining informed written consent, and the study was approved by the Bioethics Committee of our hospital. We selected the individuals using criteria such as skin color, surnames, town of origin, and language. [End Page 52]

DNA Extraction

DNA was extracted using the Wizard Genomic DNA Purification Kit System (Promega, Foster City, California, 1998), and quantity was estimated by ultraviolet absorbance (Gene Quant Calculator, Pharmacia, Uppsala, Sweden).

PCR

Amplification was performed in a Techne Thermal Cycler, model Genius Techne, New York City), following the manufacturer's recommendations.

Typing

The 11 Y-chromosome STRs in the Power Plex Y kit were typed with an ABI Prism 310 automated sequencer. Fragment size and allele designation of different loci were determined by comparison with allelic ladders distributed with the kit. The recommendations of the DNA Commission of the International Society for Forensic Haemogenetics for analysis of STR systems were followed (Bär et al. 1997; International Society for Forensic Haemogenetics 1992). We also used the experience of our team (Bell et al. 2000; Martínez-Jarreta 1999).

Quality Control

Our laboratory participated in proficiency testing provided by the GEP-ISFG Working Group (International Society for Forensic Genetics, http://www.gep-isfg.org).

Data Analyses

Number of different haplotypes, haplotype diversity, pairwise haplotype differences, and allele size variance in Y-chromosome STRs were computed with Arlequin 2.000 (Schneider et al. 2000). Median-joining networks (Bandelt et al. 1999) were produced with Network 4.1.0.8 (available at http://www.fluxus-engineering.com). STRs were given weights that were inversely proportional to their allele size variances. Admixture proportions in autosomal STRs were computed with Admix 2.0 (Dupanloup and Bertorelle 2001).

Results and Discussion

Y-Chromosome STRs

Within-Population Diversity

We typed DYS19, DYS389I, DYS389II, DYS390, DYS391, DYS392, DYS393, DYS385, DYS437, DYS438, and DYS439 Y-chromosome STRs in 94 Afro-Ecuadorans, 102 Kichwa, and 102 Mestizos, all from Ecuador. Allele frequencies can be found in Appendixes 1–3, phenotype frequencies at the duplicated DYS385 locus in Appendix 4, and haplotype frequencies in Appendix 5. General descriptors of intrapopulation genetic diversity can be found in Table 1. Haplotype diversity is high and quite close to 1 in all three populations; it should be noted that in haploid systems such as mtDNA and the Y chromosome this parameter is numerically identical to a priori forensic information parameters, such as the power of discrimination or the power of exclusion in paternity cases. Therefore this 11-locus set has ample power to discriminate unrelated male individuals in all three populations and can be used in situations such as sex crimes, where it is most appropriate.

Mestizos and Afro-Ecuadorans show slightly (and nonsignificantly) higher diversities, as measured by the average number of loci showing different alleles in [End Page 53] a random chromosome pair and the average variance of the allele size. This trend toward higher diversity is expected of admixed populations.

Minimum Haplotype Y-Chromosome STR Matches for Ecuadoran Populations in the Y-Chromosome Haplotype Reference Database (Release 16)
Click for larger view
Table 1
Minimum Haplotype Y-Chromosome STR Matches for Ecuadoran Populations in the Y-Chromosome Haplotype Reference Database (Release 16)

Haplotype Sharing Within Ecuador

Seven different haplotypes were shared between Kichwas and Mestizos, one between Mestizos and Afro-Ecuadorans, and one between Kichwas, Afro-Ecuadorans, and Mestizos. This last shared haplotype happens to be the most frequent haplotype in Europeans, in particular, the Spanish. The total number of different haplotypes is 271.

Minimum Haplotype Sharing with Global Populations

Minimum haplotypes (i.e., DYS19-389I-389II-390-391-392-393-385) have been defined for forensic practice, and such haplotypes from global populations are stored in the YHRD (Y-Chromosome Haplotype Reference Database; available at http://www.yhrd.org). Minimum haplotypes in Ecuadoran populations were searched for in the YHRD (release 16); this release contained minimum haplotypes for 32,196 chromosomes from 271 world populations. Perfect matches were counted; for haplotypes without matches, one-step neighbors were considered (i.e., haplotypes that were different by only one repeat at one locus). Results are displayed in Table 2. No haplotype showed matches to more than one continental group.

A match or a one-step neighbor could not be found for more than half the Kichwa haplotypes. It is also remarkable that only five matches were found with other native American populations, but 26 haplotypes had one-step neighbors. These two facts can be explained by two nonmutually exclusive phenomena: the higher interpopulation differentiation among Amerindians (Salzano 2002) and the underrepresentation of these populations in the database (6 populations, compared to 201 European populations). This increases the probability of no-match chromosomes (which, a priori, could have any population origin) being of native American origin.

Overall, the number of matches with Europe is striking. This method is more sensitive to European admixture, though, because Europe (and Spain in particular) is overrepresented in the YHRD. However, perfect or near-perfect matches in Europe were found for 14% of Kichwa Y chromosomes, 67% of Mestizo Y chromosomes, and 27% of Afro-Ecuadoran Y chromosomes. [End Page 54]

Admixture Estimates

The proportion of Y chromosomes of native American, European, and African origin in each population was roughly estimated by trying to predict the haplogroup of each chromosome, because most haplogroups are geographically restricted (Jobling and Tyler-Smith 2003). This task was performed by using data sets in which both biallelic markers and STRs had been typed (Bortolini et al. 2003; Zegura et al. 2004; Beleza et al. 2005, 2006) and takes advantage of the fact that STR variation in the Y chromosome is deeply partitioned by haplogroup background (Bosch et al. 1999). A chromosome was allocated to a haplogroup when a perfect or near-perfect match was found for a chromosome with a known haplogroup or when a diagnostic allele or subhaplotype was present (such as 14 or larger alleles at DYS392 combined with DYS19*13 for haplogroup Q, DYS19*15-DYS390*21 for haplogroup E3a, or DYS392*13-DYS385*11,14 for haplogroup R1b). Because we are interested in the broad origins of each chromosome, rather than in a fine phylogeography, and because this method is prone to error, we assigned each chromosome to one of the following categories: Q (native American), R1b (European), other European (includes haplogroups E3b, G, I, J, R1a), and E3a (African). The frequencies of each class in each population can be found in Table 3, and class assignments for each haplotype are given in Appendix 5.

General Descriptors of Intrapopulation Genetic Diversity in the Study Populations
Click for larger view
Table 2
General Descriptors of Intrapopulation Genetic Diversity in the Study Populations

It is worth noting that the Kichwa contain about 10% of Y chromosomes of putative European origin. However, this value is not exceptional in South America; European origin accounts for 11% of Y chromosomes in Guarani and Ingano groups, 14% in the Kaigang, and 26% in the Wayuu (Bortolini et al. 2003). The proportion of putative European lineages is as high as 70% in Mestizos (plus an additional 2% [End Page 55] of African origin). The significance of this figure is not immediately apparent because, to the best of our knowledge, no quantitative estimates of admixture proportions have been published, although Y-chromosome STR data sets have been published for comparable urban Mestizo populations. However, in the admixture of Kichwa, mating between Mestizos and Kichwas should be considered the most important source of European Y chromosomes.

Inferred Haplogroup Frequencies in Ecuadoran Populations by Comparison of STR Haplotypes with Data Sets in Which Both STRs and Haplogroup-Defining Biallelic Polymorphisms Have Been Typed
Click for larger view
Table 3
Inferred Haplogroup Frequencies in Ecuadoran Populations by Comparison of STR Haplotypes with Data Sets in Which Both STRs and Haplogroup-Defining Biallelic Polymorphisms Have Been Typed

Afro-Ecuadorans are also highly admixed. The origins of their Y-chromosome lineages can be estimated as 44% African, 31% European, and 15% native American. In this case comparable figures do exist: In different Afro-Brazilian communities, paternal contributions range from 47% to 77% for Africans, 23% to 48% for Europeans, and 0% to 4% for native Americans (Abe-Sandes et al. 2004). Compared to Brazil, the native American contribution to Afro-Ecuadorans seems larger, probably because of the native American population in the Andes, which is historically much denser than in the Amazon and Atlantic regions. The relative proportion of haplogroup R1b versus other European haplogroups is different between Mestizos and Afro-Ecuadorans (###c###2 = 6.59, p = 0.01). The proportion in Mestizos is similar to that in Spaniards; taking into account only the putatively European-derived chromosomes, the frequency of haplogroup R1b is 65.8%, whereas it is 59.6% in Spain (Flores et al. 2004). In Afro-Ecuadorans, it is 37.9%.

Native American and African Lineages in Detail

Median-joining networks were produced for the putative Q and E3a chromosomes. No discernible structure is found for Q chromosomes (Figure 1) in the network, which would suggest the presence of sublineages such as Q-M19 (Bortolini et al. 2003), whereas in the E3a network (Figure 2), two sublineages seem apparent. By comparison to Beleza et al. (2005), the bottom half of the network seems to apply to E3a7 chromosomes, and the top part of the network may belong to the E3a* paragroup. The frequencies of these two haplogroups in Afro-Ecuadorans are estimated to be 20.2% and 24.5%, respectively. [End Page 56]

Median joining network of putative E3a chromosomes. Open circles, Kichwa chromosomes; cross hatching, Mestizos; filled circles, Afro-Ecuadorans.
Click for larger view
Figure 1
Median joining network of putative E3a chromosomes. Open circles, Kichwa chromosomes; cross hatching, Mestizos; filled circles, Afro-Ecuadorans.
Median joining network of putative Q chromosomes. Open circles, Kichwa chromosomes; cross hatching, Mestizos; filled circles, Afro-Ecuadorans.
Click for larger view
Figure 2
Median joining network of putative Q chromosomes. Open circles, Kichwa chromosomes; cross hatching, Mestizos; filled circles, Afro-Ecuadorans.
[End Page 57]
FST Genetic Distances Based on 15 Autosomal STR Loci
Click for larger view
Table 4
FST Genetic Distances Based on 15 Autosomal STR Loci

Autosomal STRs

Intrapopulation Diversity

Fifteen STRs contained in the PowerPlex 16 kit were typed in 115 Kichwa, 317 Mestizo, and 104 Afro-Ecuadoran individuals. Allele frequencies were reported by González-Andrade et al. (2003) (Mestizos), González-Andrade et al. (2005) (Afro-Ecuadorans), and González-Andrade and Sánchez (2004) (Kichwas). Mean number of alleles and genetic diversity can be found in Table 4. Besides the Ecuadoran populations, data from two possible source populations were included: allele frequencies from a metropolitan population from Barcelona (including individuals born all over Spain; Paredes et al. 2003) and those from Equatorial Guinea (Alves et al. 2005), a former Spanish colony and slave trade post in Africa. The Kichwa showed the lowest diversity, in accordance with the low variability reported for many Amerindian groups. Diversity in Mestizos and Afro-Ecuadorans is greater, also according to expectations for admixed populations. However, of all pairwise comparisons, only allele number and haplotype diversity are lower in the Kichwa than in Mestizos (Wilcoxon test, p = 0.002 and p = 0.001, respectively).

Genetic Distances

FST genetic distances were computed among Ecuadoran and external populations (Table 5). FST distances were used rather than any STR-specific distance measure, given that 7 of the 15 loci showed imperfect repeats that cannot be accommodated by the simple stepwise mutation model on which distances such as RST are based (Slatkin 1995). Genetic distances in general are short, probably because of frequent stepwise mutations that tend to homogenize allele-frequency distributions. This is a general trend for STRs, and more so for forensic STRs, in which interpopulation homogeneity is a desirable property. Mestizos display a short distance from the Kichwa, but their distance from the Spanish is clearly shorter than the distance between Kichwa and Spanish. This is also the case for their respective distance to Guineans. This is consistent with a triple genetic origin [End Page 58] for Mestizos: Amerindian, European, and African, as demonstrated by Y-chromosome STRs.

Average Intrapopulation Diversity Parameters for 15 Autosomal STR Loci
Click for larger view
Table 5
Average Intrapopulation Diversity Parameters for 15 Autosomal STR Loci
Dendrogram and genetic distances between Mestizos, Kichwas and Afro-Ecuadorans calculated with minimal Y-chromosome haplotypes.
Click for larger view
Figure 3
Dendrogram and genetic distances between Mestizos, Kichwas and Afro-Ecuadorans calculated with minimal Y-chromosome haplotypes.

Afro-Ecuadorans are closest to Guineans but are closer to Kichwa and Spanish than Guineans are. With different admixture proportions the triple-source model proposed for Mestizos also applies to Afro-Ecuadorans. See Figures 3, 4, and 5.

Genetic Admixture

Genetic admixture was quantified as suggested by Dupanloup and Bertorelle (2001). These investigators derived a linear model that can accommodate any number of parental populations, as well as mutation rate, molecular distance among alleles, and time elapsed since admixture. Admixture proportions [End Page 59]

In the neighbor-joining dendrogram obtained from the same distance matrix, populations are more separate. India and Guinea Bissau are clearly separated, and India lies between Asia and Africa. Mestizos from Ecuador and Afro-Ecuadorans are between native populations and the rest of the world. Like other Amerindian groups, native populations such as Ecuadoran Kichwas show a lower genetic derivative effect or a greater mixture with Europeans.
Click for larger view
Figure 4
In the neighbor-joining dendrogram obtained from the same distance matrix, populations are more separate. India and Guinea Bissau are clearly separated, and India lies between Asia and Africa. Mestizos from Ecuador and Afro-Ecuadorans are between native populations and the rest of the world. Like other Amerindian groups, native populations such as Ecuadoran Kichwas show a lower genetic derivative effect or a greater mixture with Europeans.

and their standard deviations were estimated from 100,000 bootstrap iterations. By using Kichwa, Spanish, and Guineans as source populations, admixture proportions in Mestizos were 0.730 ± 0.243 Amerindian, 0.193 ± 0.280 European, and 0.078 ± 0.077 African. Large standard deviations are a reflection of short genetic distances among the source populations. These results contrast sharply with those obtained for Y-chromosome STRs but can be reconciled by postulating considerable sex asymmetry in matings, with most mixed matings involving European men and Amerindian women. However, such extreme differences could not arise if Mestizos [End Page 60] were created in a single generation by the mating of Amerindian women and European men. That is, a mixture of genes contributed by 70% European men and 30% Amerindian men and only Amerindian women would result in the proportions seen for the Y chromosome, but in autosomes admixture proportions would be 35% European and 65% Amerindian. Subsequent asymmetry in matings between Mestizos and Amerindians, again with mostly men contributing to the first group and women to the second, needs to be referred to in order to explain the results. This significant sex asymmetry has also been reported for several other Latin American populations (Seielstad 2000).

Genetic distances using autosomal STRs. European populations: bas, Basque Country; ext, Estremadura; pol, Poland; por, Portugal; ita, Italy; swit, Switzerland. North African populations: mor, Morocco; tuni, Tunisia; egyp, Egypt. Sub-Saharan African population: gubi, Guinea Bissau. Middle Eastern populations: turk, Turkey; syri, Syria; uae, United Arab Emirates; indi, India. Asian populations: kore, Korea; japn, Japan; viet, Vietnam; mala, Malaysia. American populations: kichwa, Kichwas; wao, Waoranis. Hybrid populations: mestizo, Mestizos; afro, Afro-Ecuadorans. We considered the following STRs: D3S1358, D5S818, D7S820, D8S1179, D13S317, D18S51, D21S11, FGA (FIBRA), VWA31, CSF1PO, D16S539, TH01, and TPOX. We used a distance coefficient. In multidimensional scaling: stress = 0.06869, variance explained = 0.99352.
Click for larger view
Figure 5
Genetic distances using autosomal STRs. European populations: bas, Basque Country; ext, Estremadura; pol, Poland; por, Portugal; ita, Italy; swit, Switzerland. North African populations: mor, Morocco; tuni, Tunisia; egyp, Egypt. Sub-Saharan African population: gubi, Guinea Bissau. Middle Eastern populations: turk, Turkey; syri, Syria; uae, United Arab Emirates; indi, India. Asian populations: kore, Korea; japn, Japan; viet, Vietnam; mala, Malaysia. American populations: kichwa, Kichwas; wao, Waoranis. Hybrid populations: mestizo, Mestizos; afro, Afro-Ecuadorans. We considered the following STRs: D3S1358, D5S818, D7S820, D8S1179, D13S317, D18S51, D21S11, FGA (FIBRA), VWA31, CSF1PO, D16S539, TH01, and TPOX. We used a distance coefficient. In multidimensional scaling: stress = 0.06869, variance explained = 0.99352.

Other urban Mestizo populations have been studied for other autosomal markers; a survey of the literature is presented in Table 6. It can be seen that Ecuadoran Mestizos represent one of the largest Amerindian contributions among the populations studied, although, given the various types of markers and levels of resolution used in the different publications, such a comparison should be made with caution. Admixture proportions for Afro-Ecuadorans were 0.564 ± 0.107 African, 0.279 ± 0.328 Amerindian, and 0.158 ± 0.367 European. As also seen from the Y chromosome, the Amerindian contribution to Afro-Ecuadorans is remarkable. Asymmetry [End Page 61]

Admixture Proportions for Various Urban Admixed American Popuations Based on Autosomal Loci
Click for larger view
Table 6
Admixture Proportions for Various Urban Admixed American Popuations Based on Autosomal Loci
[End Page 62]

is lower than in Mestizos, but again European men seem to have contributed disproportionately to admixed matings (probably mostly between them and Amerindians). Although it is well established that Hispanics generally have mixed native American, African, and European ancestry (Lisker and Babinsky 1986; Lisker et al. 1990; Mesa et al. 2000; Bedoya et al. 2006), the dynamics of the admixture of these and other different populations within South America is heterogeneous and poorly documented.

Conclusions

We have confirmed and quantified that Mestizos and Afro-Ecuadorans are trihybrid populations with various proportions of Amerindian, European, and African contributions. As seen from the Y chromosome, the male European contribution (clearly Spanish in the case of Mestizos) was much larger than when admixture estimates were computed from autosomal STRs. Other types of markers would enable the present results to be refined. For instance, Y-chromosome biallelic markers would enable the geographic attribution of Y chromosomes to be confirmed, and, in particular, would no doubt shed some light on the origin of European Y chromosomes in Afro-Ecuadorans. Autosomal ancestry informative markers (AIMs; Shriver et al. 2003) would yield much more precise autosomal admixture estimates. And, last but not least, mtDNA sequences and haplogroups would provide the female side of history. However, our study provides a reasonably detailed sketch of the composition of the main ethnic groups of Ecuador and contributes to the understanding of their diverse heritage.

Additional Information

ISSN
1534-6617
Print ISSN
0018-7143
Launched on MUSE
2007-08-27
Open Access
No
Back To Top

This website uses cookies to ensure you get the best experience on our website. Without cookies your experience may not be seamless.