Genetic Structure and Origin of Peopling in The Azores Islands (Portugal): The View from mtDNA

Authors

  • C. Santos,

    Corresponding author
    1. Unity of Anthropology, Department BABVE, Faculty of Sciences, Autonomous University of Barcelona, 08193 Bellaterra (Barcelona), Spain
    2. Department of Anthropology, University of Coimbra, 3000 Coimbra, Portugal
    Search for more papers by this author
  • M. Lima,

    1. CIRN and Department of Biology, University of the Azores, 9500 Ponta Delgada, S. Miguel, Azores, Portugal
    Search for more papers by this author
  • R. Montiel,

    1. Unity of Anthropology, Department BABVE, Faculty of Sciences, Autonomous University of Barcelona, 08193 Bellaterra (Barcelona), Spain
    2. CIRN and Department of Biology, University of the Azores, 9500 Ponta Delgada, S. Miguel, Azores, Portugal
    Search for more papers by this author
  • N. Angles,

    1. Unity of Anthropology, Department BABVE, Faculty of Sciences, Autonomous University of Barcelona, 08193 Bellaterra (Barcelona), Spain
    Search for more papers by this author
  • L. Pires,

    1. Department of Anthropology, University of Coimbra, 3000 Coimbra, Portugal
    Search for more papers by this author
  • A. Abade,

    1. Department of Anthropology, University of Coimbra, 3000 Coimbra, Portugal
    Search for more papers by this author
  • Ma. P. Aluja

    1. Unity of Anthropology, Department BABVE, Faculty of Sciences, Autonomous University of Barcelona, 08193 Bellaterra (Barcelona), Spain
    Search for more papers by this author

*Corresponding author: Cristina Santos, Unity of Anthropology, Department BABVE, Faculty of Sciences, Autonomous University of Barcelona, 08193 Bellaterra (Barcelona), Spain, Fax: +34 935811321, e-mail: Cristina.Santos@uab.es

Summary

The Azores islands (Portugal), uninhabited when discovered by Portuguese navigators in the fifteenth century, are located in the Atlantic Ocean 1500 km from the European mainland. The archipelago is formed by nine islands of volcanic origin that define three geographical groups: Eastern (S. Miguel and Sta. Maria), Central (Terceira, Faial, Pico, Graciosa and S. Jorge) and Western (Flores and Corvo). To improve the genetic characterisation of the Azorean population, and to clarify some aspects related to the history of settlement, a study of mtDNA was conducted in the population of the archipelago. The HVRI region was sequenced and specific RFLPs were screened in 146 samples obtained from unrelated individuals with Azorean ancestry (50 from the Eastern group, 60 from the Central group, and 37 from the Western group). Samples were classified into haplogroups based on the information obtained from both sequencing and RFLP analysis.

All the analyses performed support the idea that, in the whole group of islands, the majority of mtDNA lineages originated from the Iberian Peninsula, mainly from Portugal (mainland). However contributions from other European populations, especially from Northern Europe, cannot be disregarded. The values obtained for the various diversity parameters in the Azores archipelago indicate that the Azorean population, as a whole, does not exhibit the typical characteristics of an isolated population. The analysis of genetic data by groups of islands showed that the Western group exhibited particular features. The distribution of haplogroups in the Western group is very atypical, being significantly different from what is observed in the Eastern and Central groups. Furthermore, the diversity values are, in general, lower than those observed in other populations used for comparison. African haplogroups were found in all the groups of islands. Therefore the presence of Moorish and African slaves on the islands, as reported in historical sources, is supported by the mtDNA genetic data, especially in the Eastern group. The presence of Jews in the Central group is also supported by the mtDNA data. Neither historical nor genetic data (phylogeography of mtDNA) supports the idea of a differential settlement history for the Western group; however, it is represented in the phylogenies as an isolated branch. The effect of genetic drift, induced by the reduced population size since peopling occurred, has led to a very atypical distribution of haplogroups/haplotypes in this group of islands.

We cannot ignore the influence of biodemographic and genetic processes, namely founder effect, genetic drift, migration, and even recent mutational events in the mtDNA lineages of the Azorean populations. Nevertheless, a great part of the variation in the Azorean mtDNA can be explained by the settlement history.

Introduction

The Azores archipelago (Portugal), located in the Atlantic Ocean, 1500 km from the European mainland, is formed by nine islands of volcanic origin whose relative positions define three geographical groups: Eastern (S. Miguel and Sta. Maria), Central (Terceira, Faial, Pico, Graciosa and S. Jorge) and Western (Flores and Corvo). With an area of 2344 km2, the Azores presently has a total population of 237315 inhabitants, distributed in a very asymmetric way among the three groups of islands (134885 in the Eastern group, 98101 in the Central group, and only 4329 in the Western group) (INE, 2001).

Historical data indicates that the islands were uninhabited when discovered by Portuguese navigators in the fifteenth century. The peopling of the islands was a slow and difficult process, and was initiated in 1439 in the islands of Sta. Maria and S. Miguel; the small islands of Flores and Corvo were only inhabited towards the end of 16th century (Matos, 1989; Mendonça, 1996). According to historical records, the first settlers came mainly from various regions of mainland Portugal and from Madeira Island. However, people of different origin, such as Spanish, French, Italian, English, German and Flemish (whose important presence is always referred to in accounts of the peopling of the Central group), also made up part of the early settlers (Matos, 1989; Mendonça, 1996). There is clear evidence that Jews also contributed to the peopling of the archipelago. Sephardim Jews expelled from the Iberian Peninsula may have used the Azores as a refuge. The presence of African and Moorish slaves in the islands is also referred to in historical documents (Matos, 1989; Mendonça, 1996), being well-documented especially for the Western group (Gomes, 1997). However, the specific contribution of these settlers to the different islands is not clear, due to the lack of both historical and genetic data.

The complex process of peopling, the evolution of island populations, always affected by migration along with geographic isolation, has been the main stimulus for carrying out biodemographic and genetic studies on the Azorean population. To improve the genetic characterisation of this population and enhance our knowledge about several aspects related to its origin and dynamics, a research project that includes the analysis of nuclear autosomal markers, Y chromosomal and mitochondrial DNA (mtDNA), is presently being undertaken by our group. Here we present the results of the sequencing of HVRI, and specific RFLP screening of mtDNA. Our main goals were: (i) to describe the variation of mtDNA in the Azorean population; (ii) to gather data in order to contribute to the establishment of the genetic profile of the archipelago; (iii) to shed more light on some aspects related to the history and dynamics of the settlement of the Azores.

Material and Methods

Samples

One hundred and forty-six unrelated individuals, born in the Azores and with Azorean ancestry, were sampled, including 42 previously tested by Pires (1999). The birthplace of the earliest known maternal ancestor was chosen as the sample location. Fifty individuals were originally from the Eastern Group, 60 from the Central Group, and 36 from the Western Group. The samples were taken by buccal swabs and existing samples collected by Pires (1999) by venupuncture. Voluntary donors were asked for informed consent and to fill in an anonymous inquiry concerning the birthplaces of all their known ancestors. The comparative data set consisted of 1493 mtDNA HVRI sequences from the populations listed in Table 1. Populations were selected, when possible, in order to include groups that are mentioned in historical documents as sources of settlers for the Azores islands.

Table 1.  Population (corresponding codes in Figure 2), sample size, references and source of the information used in this study
PopulationNReferenceSource
  1. 1http://www.hvrbase.de

  2. 2http://www.stats.ox.ac.uk/~macaulay/founders2000/tableA.html

Azores – Eastern Group50This workThis study
Azores – Central Group60This work,This study,
   Pires, 1999 Provided by the author
Azores – Western Group36This workThis study
Mainland North Portugal (Pt)100Pereira et al. 2000GenBank
Mainland Central Portugal (Pt)82Pereira et al. 2000GenBank
Mainland South Portugal (Pt)59Pereira et al. 2000GenBank
Spain (Mainland) (Sp)89Côrte-Real et al. 1996;HVRbase (Handt et al. 1998)1
   Pinto et al. 1996 
Galicia (Sp)92Salas et al. 1998HVRbase (Handt et al. 1998)1
Basque Country (Sp)106Bertranpetit et al. 1995;HVRbase (Handt et al. 1998)1
   Côrte-Real et al. 1996 
France (Fr)50Rousselet & Mangin, 1998HVRbase (Handt et al. 1998)1
UK (UK)100Piercy et al. 1993HVRbase (Handt et al. 1998)1
Germany (Ger)67Hofmann et al. 1997HVRbase (Handt et al. 1998)1
Austria (Aus)101Parson et al. 1998HVRbase (Handt et al. 1998)1
Italy (It)49Francalacci et al. 1996HVRbase (Handt et al. 1998)1
Israel Arabs/Palestinians (Isr)117Di Rienzo & Wilson, 1991;Database of Richards et al. 20002
   Richards et al. 2000 
Algeria (Alg)85Côrte-Real et al. 1996HVRbase (Handt et al. 1998)1
Moroccan Jews (MJ)115Thomas et al. 2002Provided by the author
Morocco (Berber) (MBer)60Rando et al. 1998Provided by the author
Morocco (Moroccans) (Mor)32Rando et al. 1998Provided by the author
Senegal (Sen)119Graven et al. 1995HVRbase (Handt et al. 1998)1
   Rando et al. 1998Provided by the author
Mauritania (Maur)30Rando et al. 1998Provided by the author
S. Tome (ST)50Mateu et al. 1997Article
Mozambique (Moz)109Pereira et al. 2001Article

DNA Extraction

Total DNA from buccal cells was extracted using Instagene matrix (BioRad) according to the manufacturer's specifications. DNA from blood samples was previously extracted by Pires (1999) using standard protocols.

MtDNA Analysis

RFLP Analysis

Relevant restriction sites for mtDNA haplogroup determination (Table 2) were selected from those proposed by Torroni et al. (1993, 1994, 1996), Torroni & Wallace (1994), Chen et al. (1995), Macaulay et al. (1999), Quintana-Murci et al. (1999) and Saillard et al. (2000). Some of the specific primers and PCR conditions for each polymorphic site were available from the literature, but others were designed in our laboratory (Montiel, 2001; Santos, 2001).

Table 2.  Positions analysed to define each haplogroup
Position1715 DdeI3592 HpaI4216 NlaIII4577 NlaIII7025 AluI8994 HaeIII9052 HaeII10032 AluI10394 DdeI10397 Alu I10871 MnlI11719 SmaI12308 HinfI13704 Bst OI14465 AccI14766 MseI15606 AluI12705 (C/T)1
  1. 1The region that encompasses the position 12705 was sequenced using primers L12553 and H13127

PreHV  +++   ++ +  
HV or pre*V   ++    +      
H        +      
V   +    +      
U     +  + +  +  
K      +/− + +  +  
J  +     + +   +  
T  +      +    ++ 
I   + ++ +   +  
X    +   +  ++  
W       +   +  
M       ++       
L1 or L2 +      +        
L3 (?)       +/−       
N*    +  +/− + + T
R*    +  +/− + + C

Six μl of PCR product were used in restriction enzyme digestions and the enzymatic cleavage was performed according to the manufacturer's specifications. Results of restriction analyses were resolved through electrophoresis in agarose gels (2%).

HVRI Amplification and Sequencing

Sequencing was performed for all samples, except those previously sequenced by Pires (1999). A fragment of 443 bp from the HVRI, located between positions 15978 and 16420 (numbered according to Anderson et al. 1981), was amplified using primers and PCR conditions described by Pereira et al. (2000). PCR products were purified using the QIAquick PCR purification kit (Quiagen). Sequence reactions were carried out using the kit ABI PRISM BigDye Terminator version 3 and run in an ABI3100 sequencer (Perkin Elmer).

Data Analysis

Sequences were aligned and manually checked using BioEdit (Hall, 1999), and all polymorphic positions were confirmed in chromatograms. Sequences without ambiguities were obtained between positions 16050 and 16399. The classification of all the samples by assigning them to haplogroups was performed using the combined information of RFLPs and HVRI motifs, according to the nomenclature summarized in Richards et al. (2000), available at: http://www.stats.ox.ac.uk/~macaulay/founder2000/motif.html (Macaulay, 2000). Furthermore, to classify African sequences, a comparison with published data (Pereira et al., 2001) was performed. In order to maximise the number of sequences available for population comparison, all analyses based on sequences were restricted to the 302 nucleotides between positions 16069 and 16370.

The haplogroup distribution in the three groups of islands was used to perform an exact test of population differentiation, as described in Raymond & Rousset (1995). The exact test was also used to compare frequencies of each haplogroup in the three groups of islands.

For each population, using the HVRI sequence, standard and molecular diversity indices were estimated using the Arlequin 2.000 software (Schneider et al. 2000); namely, gene diversity inline image (Nei, 1987), number of different haplotypes (K), number of polymorphic sites (S), mean number of pairwise differences (θπ) (Tajima, 1983), nucleotide diversity (π) (Tajima, 1983; Nei, 1987), and the theta estimator based on the number of different haplotypes (θk) (Ewens, 1972). For the major haplogroups found in the Azores (and for comparison in mainland Portugal), inline image (Nei, 1987), θπ (Tajima, 1983), π(Tajima, 1983; Nei, 1987) and the average number of sites differing between a set of sequences and a specified common ancestor (ρ) (Forster et al. 1996) were estimated.

Phylogenetic networks (Bandelt et al. 1999) among haplotypes were constructed using the program Network 3.1 (www.fluxus-engineering.com). Positions of HVRI were weighted as proposed by Richards et al. (1998).

Genetic distance matrices, according to Reynolds et al. (1983), were computed in Arlequin 2.000 software (Schneider et al. 2000) and were used as in-files in the Neighbour program from the Phylip ver 3.57c package (Felsenstein, 1995), to construct Neighbour-Joining trees (Saitou & Nei, 1987). Neighbour-Joining trees were drawn with TreeView (Page, 1996).

The analysis of non-parametric correlations between the percentage of private lineages and θk was performed using the statistical package SPSS 9.0.0 (SPSS inc., 1989–1999).

Results and Discussion

Sequencing and RFLP Results

Data obtained for each sample from HVRI sequencing and RFLP analysis, as well as the haplogroup determined from the RFLP and HVRI motifs, are presented in the appendix. It was possible, using the combined information provided for both systems of mtDNA analysis, to classify all samples into one of the haplogroups described in the literature. This work confirms the previously established correlation between RFLPs and HVRI data (Torroni et al. 1996; Francalacci et al. 1999; Macaulay et al. 1999), and supports the idea that a correct and complete classification of mtDNA is only possible when both methods of analysis are used simultaneously.

Haplogroup Frequencies

The haplogroup frequencies observed in the total sample of the Azores, as well as the partial frequencies observed in the Eastern, Central and Western groups, are summarised in Table 3. The 146 samples were distributed among 31 different haplogroups/sub-haplogroups. Distinct numbers of clusters were found in the Eastern (21), Central (19) and Western (11) groups. The comparison of the haplogroup distribution in the three groups of islands, using the exact test of population differentiation, revealed that there is a similarity between the Eastern and Central groups (p = 0.29955). The Western group showed significant differences when compared with the other two groups (Western/Eastern: p < 0.00001; Western/Central: p = 0.00135).

Table 3.  mtDNA haplogroup distribution (number of individuals and percentage) in the total sample of the Azores, and in Eastern, Central and Western groups
 Azores (N = 146)Eastern Group (N = 50)Central Group (N = 60)Western Group (N = 36)
HaplogroupfI%fi%fi%fi%
H4732.1915302440822.22
I32.05  35  
J85.4851035  
 J110.6812    
  J1a21.371211.67  
  J1b132.05  35  
 J210.68  11.67  
L1c10.68  11.67  
L221.3724    
L3e2b10.6812    
M142.741211.6725.56
N1b21.371211.67  
T32.05  35  
 T132.0512  25.56
 T296.163646.6725.56
 T364.111223.3338.33
U        
 U210.68    12.78
 U321.3724    
 U521.3724    
  U5a/b32.0536    
  U5a1a21.371211.67  
  U5b10.6812    
 U6a32.0536    
 U6b21.37    25.56
 K85.483646.6712.78
pre-HV10.68  11.67  
pre*V21.37    25.56
V1510.271223.331233.33
W32.051223.33  
X42.741223.3312.78
R*10.68  11.67  

A comparison of the frequencies of each haplogroup in the three groups revealed that haplogroups J, U and V contribute to significantly differentiate the three groups of islands, and largely to differentiate the Western group (significant differences: J-Western/Central/Eastern: p = 0.04084, Western/Eastern: p = 0.038, Western/ Central: p = 0.023; U-Western/Central/Eastern: p = 0.00056, Eastern/Central: p < 0.0001; V-Western/ Central/Eastern: p < 0.0001, Western/Eastern: p < 0.0001, Western/Central: p < 0.0001).

In the total sample of the Azores, all European haplogroups (Torroni et al. 1996) were found, as well as the African clusters U6, M1, L1, L2 and L3 (Rando et al. 1998; Quintana-Murci et al. 1999; Chen et al. 1995, 2000; Watson et al. 1997) and the Near Eastern N1b cluster (Richards et al. 2000). This result provides evidence for a mixed composition of mtDNAs in the Azores islands, that is supported by historical data referring to the contribution of individuals from multiple origins to the peopling of the archipelago.

The analysis by groups (Table 3) revealed that in the Central group, all the main European haplogroups (Torroni et al. 1996) were found. A very low frequency of haplogroup U was observed when compared to values observed in European populations. The frequency of J is closer to that observed in North European populations. The frequencies of the remaining haplogroups were similar to those observed in European populations. African (M1, L1) and Near Eastern (N1b) sequences were found in low frequencies in the Central group.

In the Eastern Group, and when considering European haplogroups, only I is absent. The most interesting features are the low frequency of haplogroup H (30%), with respect to European populations, and the higher number of sub-clusters of haplogroup U relative to the Western and Central groups of the archipelago. In the Eastern group, high frequencies of non-European sequences were observed, with members of the U6, M1, L2 and L3 African clusters and N1b Near Eastern cluster detected.

The Western group exhibited a very atypical haplogroup distribution, where the I, J and W haplogroups are absent. Haplogroup U (excluding K) is under-represented, and the few samples belonging to this cluster are members of the U2 and U6b sub-haplogroups, which are uncommon in European populations but more frequent in Near Eastern and North African populations, respectively (Macaulay et al. 1999). The most frequent haplogroup is not H, as it is in European populations, but V, the frequency of which was higher than 33%. In all the results published so far, frequencies of V higher than 10% were only found in Catalonia (12.5%) (Montiel et al. 2001), the Basque Country (12.4%) and the Skolt Saami (52%) (Torroni et al. 2001). Members of the East African M1 cluster were also found in this group of islands.

HVRI Diversity

Table 4 presents diversity parameters for the Azorean population and populations cited in Table 1, obtained using HVRI sequences between positions 16069–16370. In the total Azorean sample, 80 different HVRI lineages, characterised by 74 polymorphic sites, were observed. The highest values of diversity estimators were obtained for the Central and Eastern group, with similar values, and the lowest values were observed in the Western group. Compared to other populations, the values for mean numbers of pairwise differences (θπ) and nucleotide diversity (π) were higher than the majority of European populations in the Azores sample (total or by groups of islands). A median value of gene diversity was observed in the Azores. However, the value registered for the Western group appears to be one of the lowest when the three groups of islands were analysed separately. The values of diversity decreased in the three groups of islands when African sequences (L1, L2, L3 and M1) were excluded (Table 4).

Table 4.  mtDNA HVRI diversity (from nucleotide positions 16069 to 16370) in the Azorean populations including and excluding African sequences (L1, L2, L3 and M1), and in populations used for comparison
PopulationNK (%K)S (%S)transi/transv/ indelinline imageθπ± sdπ± sdθkPrivate lineages (%)
  1. N-sample size; K-number of different haplotypes; S-number of polymorphic sites; inline image-gene diversity; θπ-mean number of pairwise differences; π-nucleotide diversity; θk-theta estimator based on the number of different haplotypes; Private lineages are defined as those found only in one of the populations included in the present study

Azores14680 (54.79)74 (24.50)64/13/20.9603 ± 0.01075.4339 ± 2.63220.0179 ± 0.009671.772341 (51.25)
 Eastern Group5033 (66.00)50 (16.56)41/8/20.9486 ± 0.02365.3168 ± 2.61020.0175 ± 0.009541.142817 (51.52)
 Central Group6040 (66.67)52 (17.22)47/6/00.9565 ± 0.01905.4937 ± 2.67990.0182 ± 0.009851.280116 (40.00)
 Western Group3616 (44.44)34 (11.26)30/4/10.9159 ± 0.02765.1745 ± 2.56490.0171 ± 0.009410.47256 (37.50)
Mainland Portugal241142 (58.92)94 (31.13)89/12/00.9588 ± 0.00965.1369 ± 2.49820.0170 ± 0.0092144.203784 (59.15)
 North Portugal10067 (67.00)69 (22.85)64/7/00.9533 ± 0.01625.1406 ± 2.51200.0170 ± 0.009287.700035 (52.22)
 Central Portugal8261 (74.39)63 (20.86)59/7/00.9765 ± 0.01075.2758 ± 2.57530.0175 ± 0.0094106.318226 (42.62)
 South Portugal5941 (69.49)54 (17.88)50/4/00.9433 ± 0.02464.9621 ± 2.44880.0164 ± 0.009058.238218 (43.90)
Spain8971 (79.78)68 (22.52)65/5/00.9837 ± 0.00815.4974 ± 2.66930.0182 ± 0.0098160.671637 (52.11)
Galicia9251 (55.43)55 (18.21)53/3/10.9233 ± 0.02393.3215 ± 1.72180.0110 ± 0.006346.301121 (41.18)
Basque Country10653 (50.00)53 (17.55)49/5/00.9364 ± 0.01813.1821 ± 1.65870.0105 ± 0.006041.520523 (43.40)
France5042 (84.00)52 (17.22)47/7/00.9878 ± 0.00864.85720 ± 2.40920.0161 ± 0.0089121.053722 (52.38)
UK10068 (68.00)67 (22.19)62/5/20.9752 ± 0.00924.7785 ± 2.35490.0157 ± 0.008692.252736 (52.94)
Germany6752 (77.61)56 (18.54)52/3/10.9724 ± 0.01394.1728 ± 2.10040.0138 ± 0.0077104.841230 (57.69)
Austria10164 (63.37)72 (23.84)68/7/00.9505 ± 0.01704.9565 ± 2.43200.0164 ± 0.008974.060936 (56.25)
Italy4940 (81.63)54 (17.88)52/2/00.9685 ± 0.01895.5298 ± 2.70420.0183 ± 0.009999.385328 (70.00)
Israel117101 (86.32)93 (30.79)90/12/30.9945 ± 0.00316.6033 ± 3.14120.0219 ± 0.0115348.304166 (65.35)
Algeria8530 (35.29)37 (12.25)37/0/00.9431 ± 0.01035.2725 ± 2.57290.0175 ± 0.009416.094215 (50.00)
Moroccan Jews11545 (39.13)51 (16.89)49/3/00.9077 ± 0.02244.0584 ± 2.03980.0134 ± 0.007526.732325 (55.56)
Morocco (Berber)6038 (63.33)50 (16.56)46/5/10.9633 ± 0.01475.0957 ± 2.50640.0169 ± 0.009243.508118 (47.37)
Morocco (Moroccans)3229 (90.63)47 (15.56)44/3/00.9879 ± 0.01416.8612 ± 3.31580.0227 ± 0.0122144.674712 (41.38)
Mauritania3023 (76.67)31 (10.26)31/0/00.9747 ± 0.01726.8674 ± 3.32450.0227 ± 0.012343.30568 (34.78)
Senegal240124 (51.67)80 (26.49)77/6/10.9858 ± 0.00257.3792 ± 3.46330.0249 ± 0.0129102.4841104 (83.87)
S. Tome5032 (64.00)50 (16.56)46/6/00.9731 ± 0.01109.3350 ± 4.36060.0309 ± 0.016037.276321 (65.62)
Mozambique10950 (45.87)62 (20.53)58/7/20.9611 ± 0.00809.8708 ± 4.55180.0327 ± 0.016735.172943 (86.00)
Excluding African sequences (L1, L2, L3 and M1)
 Azores13873 (52.90)67 (22.19)59/10/20.9557 ± 0.00124.9439 ± 2.42110.0163 ± 0.008862.084433 (45.21)
 Eastern Group4629 (60.04)43 (14.24)36/5/20.9391 ± 0.02744.9383 ± 2.44830.0162 ± 0.008932.652713 (44.83)
 Central Group5838 (65.52)45 (14.90)43/3/00.9534 ± 0.02024.9690 ± 2.45240.0165 ± 0.009046.772314 (36.84)
 Western Group3415 (44.12)32 (10.60)28/4/10.9073 ± 0.03044.5447 ± 2.29070.0150 ± 0.00849.70305 (33.33)

The analysis of diversity by haplogroups in the Azorean sample compared to mainland Portugal (Table 5) showed high values of diversity for the haplogroups U and J, while the diversity of the remaining haplogroups is lower in the Azores. In the Azorean sample, H and V haplogroups present the lowest values of diversity. The African haplogroups have high values of θπ, π and ρ, followed by the U haplogroup. However, in the Azores, the highest value for gene diversity is displayed by haplogroup U.

Table 5.  Diversity within major haplogroups in the Azores (Az) and Mainland Portugal (Pt)
 NKinline imageθπ± sdπ± sdρ
HaplogroupsAzPtAzPtAzPtAzPtAzPtAzPt
  1. N-sample size; inline image-gene diversity; θπ-mean number of pairwise differences; π-nucleotide diversity. Major haplogroups are defined as those with a total frequency higher than 10%; ρ-average number of sites differing between a set of sequences and a specified common ancestor (between positions 16051–16399)

U (including K)245419390.9819 ± 0.01640.9762 ± 0.01135.5696 ± 2.77195.3857 ± 2.63700.018382 ± 0.0101980.017834 ± 0.00968727/24 = 1.12559/54 = 1.093
J151710100.9333 ± 0.04490.9265 ± 0.04173.9738 ± 2.10583.3984 ± 1,82900.013158 ± 0.0078180.011253 ± 0.00677814/15 = 0.93315/17 = 0.882
T212611160.9286 ± 0.03060.9323 ± 0.03374.3704 ± 2.24894.6208 ± 2.34240.014472 ± 0.0083120.015301 ± 0.00864112/21 = 0.57122/26 = 0.846
H479919440.6966 ± 0.07680.8023 ± 0.04311.4868 ± 0.91451.9230 ± 1,10310.004891 ± 0.0033390.006368 ± 0.00404620/47 = 0.42643/99 = 0.434
V15173100.5905 ± 0.07710.7941 ± 0.10350.6632 ± 0.53941.6799 ± 1.03530.002196 ± 0.0020030.005562 ± 0.0038362/15 = 0.13312/17 = 0.706
L1+L2+L3+M18187170.9643 ± 0.07720.9935 ± 0.02108.5883 ± 4.44459.1597 ± 4.42020.028344 ± 0.0167090.030330 ± 0.01636717/8 = 2.12538/18 = 2.110

Relating gene diversity and nucleotide diversity in the three groups of islands (Table 4) to intra-haplogroup variation (Table 5), it is possible to infer that the higher nucleotide diversity observed in the Azores, as compared to European populations, is mainly the result of the presence of African sequences in the three groups of islands. Nevertheless, the influence of the variability of some specific haplogroups, especially U, also has some influence on the increase in both nucleotide and gene diversity.

According to Helgason et al. (2000), the theta estimator based on the number of different haplotypes (θk) is likely to provide a more reliable evaluation of current and historical female effective-population size than θπ. Of all the populations studied, the Western group presents the lowest value of θk, and the Central and Eastern groups also presented relatively low values (Table 4). This indicates that the Azores, and especially the Western group, have recently had a relatively small female effective-population size.

In general, the proportion of private lineages sampled from geographical proximate populations should increase as a function of θk, as this parameter reflects the probability of new lineages arising by mutation (Helgason et al. 2001). High values of private lineages associated with low values of θk, found in some populations used in the comparison such as Mozambique, are the result of the non-inclusion of a sufficient number of geographically related populations in the comparative database to detect gene flow. Considering the analysis of European populations alone, a significant positive correlation is observed between the percentage of private lineages and the values of θk (rsp = 0.596; p = 0.009). The relationship between the percentage of private lineages and the values of θk is presented in Figure 1, and varies greatly among the three groups of islands. The Eastern group presents an excess of private lineages in contrast to the Central group, which tends to a scarcity of private lineages. The western group, and the total sample of the Azores, have a proportion of private lineages that can be expected by their θk.

Figure 1.

Scatterplot of θk values and percentage of private lineages. The least-squares regression line is shown. The curved lines show the 95% confidence region around the regression line.

The excess of private lineages observed in the Eastern group could have different causes. An intuitive explanation is that the isolation of the islands can hinder the migratory flow of new lineages to and from neighbouring populations. However, six of the private lineages found in the Eastern group belong to African clusters (U6a, M1, L2 and L3). Consequently, it is more likely that a sufficient number of African populations was not included in the comparative database.

Within the Azores, the Central group shares the highest number of lineages with the other two groups of islands. Therefore, the relative scarcity of private lineages observed in the Central group could be indicative of either a very high level of emigration (where few lineages remain private for long, because of rapid outward gene flow to neighbouring populations), or of immigration (where new lineages arriving into the population would increase θk but not the proportion of private lineages).

The values of diversity parameters, reported for the archipelago, allow us to infer that the Azores population as a whole does not present the typical characteristics of an isolated population. Thus, the values recorded are similar to those observed in the majority of European populations used in the comparative analysis, and higher than those observed in isolated populations like the Basque country and Galicia. Furthermore, neither an excess nor a scarcity of private lineages is observed, in relation to the respective θk value, suggesting that values of diversity in the Azores are adjusted to the effective population size of females.

These results are in accordance with studies of autosomal STRs (Lima, personal communication), classical markers (Amorim et al. 1979) and those reported in biodemographic studies, which indicate that the Azorean archipelago, when considered as a whole, does not exhibit levels of endogamy and consanguinity higher than those reported for some populations from mainland Portugal (Cunha, 1987; Lima, 1991; Lima & Soares, 1992; Smith et al. 1992).

When analyses were carried out by groups of islands, the Western group appears to have different features. The haplogroup distribution is very atypical, diversity values are in general lower than those observed in the other groups of islands, and the value of gene diversity is one of the lowest registered in all of the populations used for comparison. However, the relationship between the percentage of private lineages and values of θk indicate that this group of islands does not undergo a higher level of isolation than that observed in the other groups of islands. This leads to the conclusion that the observed haplogroup distribution and lower diversity are strictly related to the reduced number of females present in the island since peopling, and makes the roles of founder effect and genetic drift in small populations evident.

Phylogeography of Azores mtDNA

Figure 2 (a, b, c, d, e) represents phylogenetic networks of haplotypes found in the Azores and populations used for comparison. To simplify networks, haplotypes absent in the Azores were only represented if they were phylogenetically related to haplotypes present in the Azorean population. Examining the networks, from the 80 different haplotypes (based on HVRI information) found in the Azores, only six are present in more than one group of islands. Two haplotypes are present in the three groups, three are common to the Central and Eastern groups, and one to the Central and Western groups. The Central group is the one that shares the highest number of haplotypes with the other two. This result indicates limited migratory movements among the three groups of islands, until at least the beginning of the 19th century (since the geographical location of the sample was based on the origin of the great-grand-parents of the donor).

Figure 2.

Figure 2.

Phylogenetic Networks of (a) H, V, HV, Pre-HV, Pre*V; (b) U (including K); (c) J, T, JT; (d) N, I, W, X; (e) L1, L2, L3, M. Circle sizes are proportional to the haplotype frequency. (Red - Eastern group; Blue - Central group; Green - Western group; Gray - Mainland Portugal; White - other populations).

Figure 2.

Figure 2.

Phylogenetic Networks of (a) H, V, HV, Pre-HV, Pre*V; (b) U (including K); (c) J, T, JT; (d) N, I, W, X; (e) L1, L2, L3, M. Circle sizes are proportional to the haplotype frequency. (Red - Eastern group; Blue - Central group; Green - Western group; Gray - Mainland Portugal; White - other populations).

Figure 2.

Figure 2.

Phylogenetic Networks of (a) H, V, HV, Pre-HV, Pre*V; (b) U (including K); (c) J, T, JT; (d) N, I, W, X; (e) L1, L2, L3, M. Circle sizes are proportional to the haplotype frequency. (Red - Eastern group; Blue - Central group; Green - Western group; Gray - Mainland Portugal; White - other populations).

Figure 2.

Figure 2.

Phylogenetic Networks of (a) H, V, HV, Pre-HV, Pre*V; (b) U (including K); (c) J, T, JT; (d) N, I, W, X; (e) L1, L2, L3, M. Circle sizes are proportional to the haplotype frequency. (Red - Eastern group; Blue - Central group; Green - Western group; Gray - Mainland Portugal; White - other populations).

Figure 2.

Figure 2.

Phylogenetic Networks of (a) H, V, HV, Pre-HV, Pre*V; (b) U (including K); (c) J, T, JT; (d) N, I, W, X; (e) L1, L2, L3, M. Circle sizes are proportional to the haplotype frequency. (Red - Eastern group; Blue - Central group; Green - Western group; Gray - Mainland Portugal; White - other populations).

Forty-one haplotypes (represented in the networks as single colour circles: red to the Eastern group, blue to the Central group and green to the Western group) appear to be exclusive to the Azores, and thirty-nine are shared with other populations. By analysing the shared haplotypes, thirty-three were found to be common to the Iberian Peninsula (mainly to mainland Portugal). However, only eleven are shared exclusively with the Iberian Peninsula, the remaining twenty-two also being present in one or more of the other populations. Specifically, sixteen are found in other European populations, ten in Israel, five in Moroccan Jews, eleven in North Africa and four in West Africa. Considering that these haplotypes belong to European haplogroups (H, J, T U, K, V and X) (Torroni et al. 1996) their most probable origins are the Iberian Peninsula or other European populations. However, a possible North African origin cannot be excluded for some sequences. The work of Rando et al. (1998) attests to the fact that there is strong evidence for some degree of European genetic input into North Africa. There is, for example, evidence that haplogroups U5 (Richards et al. 1998) and V (Torroni et al. 1998) are represented in North Africa. Only one haplotype (ACR34-H) is shared exclusively with non-Iberian European populations (Fig. 2a).

Haplotypes represented by A15-pre-HV (Fig. 2a) and SM11-N1b (Fig. 2d) are shared exclusively with the Israeli population, and SM18-K (Fig. 2b) with Moroccan Jews. Taking into account the fact that the Israel Arab/Palestinian sample represents an extant population living in the region to which Jews trace their origin (Thomas et al. 2002), and that Moroccan Jews probably included descendants of the Sephardim Jews that were expelled from the Iberian Peninsula, the three haplotypes (represented by A15, SM11 and SM18) previously mentioned could indicate the presence of Sephardim Jews in the Azores. The haplotype represented by SM51 (Fig. 2b), belonging to sub-haplogroup U6a, is shared with Israeli and North African populations. This haplotype is probably originally from North Africa, since the U6 haplogroup has been described as a North African cluster (Rando et al. 1998).

U6 sequences are represented in both the Eastern and Western groups; however, in the Western group (samples F1, F12) they belong to sub-cluster U6b, which, according to Plaza et al. (2001), is restricted to the Iberian Peninsula. In the Eastern group (samples SM22, SM30 and SM51) they are assigned to U6a, found in both the Iberian Peninsula and North Africa (Plaza et al. 2001; Rando et al. 1998). Thus, only lineages found in the Eastern group could plausibly constitute evidence of the influence of Moorish slaves, which is in accordance with historical data.

The haplotype represented by A16 (Fig. 2d), a member of the European haplogroup I, is present in non-Iberian European populations. In S. Tome, the same haplotype was found (Mateu et al. 1997); however that sample probably belongs to haplogroup L3 (Calafell, Comas & Mateu, personal communication).

In summary, the analysis of shared haplotypes indicates that the majority of haplotypes is shared with Iberian Peninsula populations, principally with mainland Portugal. A small influence from other European (especially North European), Near East/Jewish and North African populations can also be confirmed.

The 41 private lineages found could have been introduced into the Azores both by European and by African individuals belonging to populations that are not included in the comparison, for example, populations from Flanders and Madeira Island, for which data are not available. Alternatively, as suggested by Helgason et al. (2000), they may arise from populations included in the analysis in which these lineages have not yet been sampled, or have been lost since the inhabitation of the Azores.

On the other hand, the hypothesis of the origin of some private lineages in recent mutational events occurring after the settlement of the islands cannot be disregarded.

The possible origin of the non-shared haplotypes, to be discussed later, was inferred using the partial networks presented in Figure 2 and two mtDNA databases, namely the database of Richards et al. (2000) and the HVRbase (Handt et al. 1998), both available online.

A worldwide search for the 41 private haplotypes found in the Azores was performed in mtDNA databases, and eight matches with samples that were not included in the comparison were found. Haplotypes represented by ACR31-H (Fig. 2a), F24-T (Fig. 2c), ACR30-W (Fig. 2d) and SM49-U5a/b (Fig. 2b) are shared with Northern European populations that were not included in the analysis; ACR8-T2 (Fig. 2c), SM12-X (Fig. 2d) and ACR7-T (Fig. 2c) match with one sequence each, found, respectively, in Rome (Italy), Romany and Armenia; SM22 (Fig. 2b), which belongs to the North African cluster U6a (Rando et al. 1998), is not shared with populations used for comparison, but was found in the Canary Islands (Rando et al. 1999) and in Moroccan Berbers (Anglés, personal communication).

Haplotypes classified as J can provide valuable information for the inference of the origin of the Azorean population. This haplogroup is present at high frequencies in Northern European populations, and some specific clusters, such as J1b1, can be assigned to a very restricted geographic area. Haplogroup J is absent in the Western group, but its frequency is close to 14% in both the Eastern and Central groups, higher than that observed in the Iberian Peninsula and similar to that observed in Northern European populations. Among the ten J haplotypes found in the Azores, six are private lineages (Fig. 2c).

The haplotype represented by ACR5 could have been derived from the haplotype 16069, 16126, 16278, 16366, found in mainland Portugal and in the Eastern group (SM40). SM19 could have been derived from the haplotype 16069, 16126, 16261, found in mainland Portugal. Haplotypes represented by SM25 and SM37 could be derived from haplotype 16069, 16126 found in the Azores (SM7, SM43 and ACR23) and at high frequencies in Europe, especially in United Kingdom. Haplotype A3 was possibly derived from the haplotype 16069, 16126, 16145, 16172, 16222, 16261, found in the United Kingdom and Spain.

Nevertheless, other origins can be postulated for the previously mentioned J haplotypes. Importantly, the influence of the population from the region of Flanders that was not included in the analysis is always mentioned in the peopling of the Central group. It is expected that this population, like the surrounding Northern European populations, would have high frequencies of haplogroup J, and some of the non shared haplotypes found in the Azores could match with those which, hypothetically, would be found if the mtDNA of Flemish populations was analysed.

East African M1 sequences (SM1, ACR11, F29 and F37) (Fig. 2e) do not match any sequence present in the mtDNA databases, probably because African populations are less studied. The most similar sequence that was found is a North African sequence with the motif 16129, 16182C, 16183C, 16189, 16223, 16249, 16311. All the M1 sequences found in the Azores have a very similar motif, and probably trace their origin to the same geographical region in East Africa. With respect to L1, L2 and L3 sequences (Fig. 2e), they are located in the network near to sequences from various sub-Saharan populations and from mainland Portugal African sequences. This indicates a common origin for both African sequences present in mainland Portugal and the Azores islands, which is consistent with the historical context of the 16th century, when Portugal was actively involved in the slave trade (Thomas, 1998).

The analysis of non-shared haplotypes indicates that the majority of the European private lineages have their probable origin in Northern European and Near Eastern populations. Evidence of influences from both North Africa and sub-Saharan Africa were found in the Azores; however, a strict geographic origin for African sequences is difficult to define.

The terminal position in the networks of several Azorean haplotypes, and their strict relation with other Azorean haplotypes, seem to indicate that some of the private lineages could have originated in the Azores.

Estimates of Admixture

To estimate the proportion of admixture, the information concerning the possible origin of both shared and non-shared haplotypes was summarised. Three major groups were defined for the origin of mtDNA in the Azores (Europe, Near East and Jews, and Africa). Whenever a haplotype is shared with a European population, it is admitted that its most probable origin is Europe. Thus, it is possible to infer that probably: a minimum of 11.25% of the haplotypes found in the Azores are originally from African populations; a minimum of 7.5% are derived from Near Eastern and Jewish populations; and a maximum of 81.25% are from European populations.

Figure 3 illustrates the observation that mtDNA gene pools of all the groups of islands have their major contribution from European populations. However, there are some differences among the three groups of islands. The Eastern group exhibits the highest contribution from non-European populations (close to 25%), mainly African (18.2%), revealing the influence of slaves in the peopling of this group of islands.

Figure 3.

Bar chart showing the pattern of haplotype contribution from parental populations to the Azores.

In the Central group, the non-European influence is approximately 15%, 10% from Near Eastern/Jewish populations, and 5% from African populations. The Western group exhibits the lowest contribution of non-European lineages (6.25%), which is exclusively African. According to historical information the contribution of slaves to the peopling of the islands, mainly to the Eastern and Western groups, was very important, being well-documented for the Western group (for review see Gomes, 1997). One of the headmen of Flores island was also leader of Santo Antão island (Cabo Verde), and a great number of the slaves who arrived at the Western group were originally from Cabo Verde. Surprisingly, no L1, L2 or L3 mtDNA lineages were found in the Western group. This indicates that factors such as differential mortality may have affected the fitness of slaves, limiting the transmission of their genes until now.

Genetic Distances and Phylogenetic Reconstruction

To examine the relationship between all pairs of populations used in the comparison, genetic distances according to Reynolds et al. (1983) were calculated, and used to build a phylogenetic tree (not shown). In contrast to European populations, African populations present higher genetic distances among themselves. This justifies the existence of an identifiable gradient observed in the N-J tree from sub-Saharan to North African populations, and the aggregation of the European populations in one extreme of the tree. It is impossible to establish any relation among European populations. Furthermore, the three groups of islands are located in this assemblage.

To try to clarify the position of the Azores with respect to other populations, various N-J trees were built. The position of the various populations was very similar in all the trees. In Figure 4, a tree that excludes Algerian and Western and Eastern African populations is shown, where the African populations that show less genetic distance relative to the Azores are maintained.

Figure 4.

Neighbour-Joining tree (Saitou & Nei, 1987) constructed from Reynolds et al. (1983) genetic distances.

The analysis of Figure 4 reveals that the geographically and/or culturally isolated populations, such as those of the Basque Country, Galicia and Moroccan Jews, are separated from the remaining European populations. Moroccans also appear to be separated, because this population is the most differentiated relative to European populations. The three groups of islands are located in different branches of the tree that emerge from a group of European populations. However, the Western group appears in a very profound and isolated branch, reinforcing the previous results that indicate a differentiation of this group of islands. The Central group is located in the same branch as Israel, while the Eastern group is positioned in the same branch as the Moroccan Berbers. The position of the Eastern group corroborates previous results, indicating some affinities with African populations. In a similar manner the position of the Central group indicates that there is some relationship with Near Eastern populations.

The analysis of phylogeography of Azores mtDNA, estimates of admixture, genetic distances and phylogenetic reconstruction allow us to infer a possible origin for the mtDNA gene pool of the Azores. All the analyses support the idea that, in all the groups of islands, the majority of mtDNA lineages come from the Iberian Peninsula, mainly from Portugal, but the contribution of other European populations, especially from Northern Europe, cannot be disregarded. The presence of Moorish and African slaves in the islands is also supported by mtDNA genetic data, especially in the Eastern group, as well as the presence of Jews, mainly in the Central group.

Neither historical nor genetic data supports the idea of a differential settlement for the Western group; however, it appears in the phylogenies in an isolated branch. With 4329 inhabitants (INE, 2001) today, the Western group has reduced its population size by half in less than one century. Thus, this “bottleneck”, combined with a small population size since peopling, has led to a very atypical distribution of haplogroups/haplotypes that influences all the results obtained. Future genetic and biodemographic investigations in the Western group will certainly help to clarify its position relative to the rest of the archipelago, and to other populations.

Conclusions

This work confirms that studies of large population samples in restricted geographical contexts can produce valuable insights concerning specific genetic and demographic features that would remain undetectable in broad scale surveys.

The distinct analyses allow us to infer that the Azorean population does not present the typical characteristics of an isolated population, as is usually postulated for islands. However, the Azores archipelago has a recent demographic history, was peopled by individuals of multiple origins, which naturally leads to a population with high diversity, and it is probable that the effect of isolation, if it exists, remains undetectable at this time.

Separate analyses of the three groups of islands appeared to be valuable for detecting the specific features of the Western group, which presents some differentiation with respect to the other groups of islands. This differentiation should be the result of genetic drift processes, induced by the small population size since peopling, and eventually the result of the recent reduction in population size.

The analysis of lineage sharing supports the idea that mainland Portugal and Northern European populations could have contributed the majority of the mtDNA lineages that are currently observed in the Azorean mtDNA gene pool. However, many lineages are still private to the Azores, and their origins are sometimes difficult to infer, since historically related populations such as the population of Flanders are not used for comparison, because they remain uncharacterised for mtDNA.

Contrary to studies using autosomal markers (Lima et al. personal communication), it was possible to infer the influence of African and Near East/Jewish individuals in the islands, pointing out the effectiveness of non-recombining markers to reconstruct the history of populations. It was even possible to detect differential contributions of non-European populations to the peopling of the three groups of islands, mainly the contribution of Africans to the Eastern group, and the contribution of Near East/Jewish to the Central group, both in accordance with historical data.

Part of the variation in the Azorean mtDNA can be explained by the settlement history of the archipelago. Furthermore, it has been necessary to bear in mind that the variation of mtDNA was affected during the last 500 years, by biodemographic and genetic processes like founder effect, genetic drift, migration, and even recent mutational events. Thus, the results presented here arise from the combination of the history of peopling and the evolution of populations, with all their biodemographic and genetic interactions.

Acknowledgments

This work was supported by a grant (ref. SFRH/BD/ 723/2000) from Fundação para a Ciência e a Tecnologia - Operational programs Science Technology and Innovation (POCTI) and Society of Information (POSI) financed by UE founds and national founds of MCES.

Appendix

Please click here for table.

Table Appendix.  HVRI region sequences and RFLP specific screening in the Azorean population
Sample NameHVRI (16050–16399)Hapl. Seq.MnlI 10871DdeI 10394MseI 14766HinfI 12308AluI 7025NlaIII 4216NlaIII 4577HaeII 9052HaeIII 8994DdeI 1715AluI 10032AccI 14465SmaI 11719Bst0I 13705AluI 15606AluI 10397HpaI 359212705 C/THapl. RFLPHapl. Seq/RFLP
  1. Variant positions from the CRS are shown between 16050 and 16399 in HVRI (minus 16000). Substitutions are transitions unless the base change is explicitly indicated. Insertions of one and two cytosines are shown by ‘insC(1)’ and ‘insC(2)’ respectively. The results of specific RFLP screening are also presented. + and – indicates, respectively, the presence and absence of the restriction site. Classification into haplogroups is presented using individual and combined information of sequences and RFLPs. Haplogroups presenting an ambiguous classification are assigned with a question mark (?).

Eastern group                      
SM1129, 183C, 189, 189insC(1), 249, 311M1?+  +          + MM1?
SM2CRSH or HV or U or R+              HH
SM3343, 390U3++++ +          UU3
SM4CRSH or HV or U or R+  +           HH
SM5CRSH or HV or U or R+              HH
SM693, 189, 189insC(1), 192, 270, 311U5a or U5b++++++++       UU5a/b
SM769, 126J+++ ++           JJ
SM8CRSH or HV or U or R+         +    HH
SM9CRSH or HV or U or R+         +    HH
SM10274H or HV or U or R+              HH
SM11145, 176G, 223, 390N1b+++         OtherN1b
SM12189, 192, 223, 278X++++ ++ +     XX
SM13126, 294, 296, 304T2++ ++        +   TT2
SM14207H or HV or U or R+              HH
SM15CRSH or HV or U or R+              HH
SM16223, 260, 278, 390, 399L2+  +           + L1 or L2L2
SM17196, 224, 311, 362K+++++           KK
SM18168, 224, 311K+++++           KK
SM1969, 111A, 126, 261J1+++ ++           JJ1
SM20343, 356, 390U3 or U4++++ +          UU3
SM21CRSH or HV or U or R+              HH
SM2292, 172, 219, 278U6a++++ +          UU6a
SM23224, 270, 311U5 or K++++ +          UU5
SM2493, 189, 354H or HV or U or R+              HH
SM25126, 218AJ?+++ ++           JJ
SM26126, 234, 248, 292, 294T3++ ++        +   TT3
SM27189, 270U5b++++ +          UU5b
SM28224, 270, 311U5 or K++++ +          UU5
SM3092, 172, 219, 278U6a++++ +          UU6a
SM31189, 189insC(2)H or HV or U or R+              HH
SM3293A, 223, 278, 294, 309, 390L2+  +           + L1 or L2L2
SM33126, 163A, 186, 189, 294T1++ ++        +   TT1
SM34196, 224, 311, 362K+++++           KK
SM36CRSH or HV or U or R+              HH
SM3769, 126, 150J+++ ++           JJ
SM38126, 294, 296, 304T2++ ++        +   TT2
SM39206C, 223, 292, 311W++++ +        WW
SM4069, 126, 278, 366J+++ ++           JJ
SM4193, 189, 189insC(1), 192, 270, 311U5a or U5b++++ +          UU5a/b
SM42CRSH or HV or U or R+              HH
SM4369, 126J+++ ++           JJ
SM44256, 270, 399U5a1a++++ +          UU5a1a
SM45126, 294, 296, 304T2++ ++        +   TT2
SM46CRSH or HV or U or R++ ++        HH
SM47CRSH or HV or U or R+              HH
SM48298Pre*V or V++  +        VV
SM49189, 192, 270, 311, 336U5a or U5b++++ +          UU5a/b
SM5069, 126, 145, 231, 261J1a+++ ++           JJ1a
SM51172, 219, 278U6a++++ +          UU6a
SM52172, 182C, 183C, 189, 223, 320, 390CL3e2b+  +           L3?L3e2b
Central group                      
A169, 126, 193, 278J2+++ ++           JJ2
A2240H or HV or U or R+              HH
A369, 126, 145, 172, 222, 261, 305TJ1b1+++ ++           JJ1b1
A4CRSH or HV or U or R+              HH
A5CRSH or HV or U or R+             CHH
A6240H or HV or U or R+              HH
A7129, 145, 176G, 223, 291, 390N1b+++        TN*N1b
A869, 126, 145, 172, 222, 261, 305TJ1b1+++ ++           JJ1b1
A9126, 234, 248, 292, 294T3++ ++        +   TT3
A10CRSH or HV or U or R+              HH
A11CRSH or HV or U or R+              HH
A12129, 242H or HV or U or R+              HH
A1369, 126, 145, 172, 222, 261, 305TJ1b1+++ ++           JJ1b1
A14126, 234, 248, 292, 294T3++ ++        +   TT3
A15126, 362pre-HV++++ +++     pre-HVpre-HV
A16129, 223, 391I++++  ++       II
A1793, 224, 290, 311K+++++           KK
P1224, 311K+++++           KK
ACR1293, 294, 304T2?++ ++        +   TT2?
ACR3CRSH or HV or U or R+              HH
ACR4129, 223, 391I++++  ++       II
ACR569, 93, 126, 278, 366J+++ ++           JJ
ACR6129, 187, 189, 223, 265C, 278, 286A, 292, 294, 311, 360L1c+              + L1 or L2L1c
ACR7126, 186, 189, 294T++ ++        +   TT
ACR8126, 129, 294, 296, 304T2++++   +    +   TT2
ACR9CRSH or HV or U or R+              HH
ACR10CRSH or HV or U or R+              HH
ACR11129, 182C, 183C, 189, 223, 230, 249, 311, 359M1+             + MM1
ACR12189H or HV or U or R+              HH
ACR14126, 186, 189T?++ ++        +   TT
ACR15129, 223, 391I++++  ++       II
ACR1669, 126, 145, 231, 261J1a+++ ++           JJ1a
ACR17162H or HV or U or R+              HH
ACR18298Pre*V or V++  +        VV
ACR19176H or HV or U or R+              HH
ACR20126, 294, 304T2++ ++        +   TT2
ACR21189, 192, 223, 278, 362X+++  ++      XX
ACR22CRSH or HV or U or R+              HH
ACR2369, 126J+++ ++           JJ
ACR24183C, 189, 223, 260, 278X+++  ++      XX
ACR25278, 311R1+++++           KK
ACR26CRSH or HV or U or R+              HH
ACR27126, 186, 189, 294T++ ++        +   TT
ACR28298, 311Pre*V or V++  +        VV
ACR29126, 294, 296, 304T2++ ++        +   TT2
ACR30223, 292, 320W+++  +        WW
ACR31189, 356H or HV or U or R+              HH
ACR32129, 249H or HV or U or R+              HH
ACR33223, 292, 320W+++  +        WW
ACR34129, 316H or HV or U or R+              HH
ACR3569, 278J?+++ ++           JJ
ACR36CRSH or HV or U or R+              HH
ACR37192, 256, 270, 362, 399U5a1a++++ +          UU5a1a
ACR38CRSH or HV or U or R+              HH
ACR3993, 269, 270H or HV or U or R+              HH
ACR40304H or HV or U or R+              HH
ACR41CRSH or HV or U or R+              HH
ACR4293, 224, 290, 311K++++           KK
ACR44124H or HV or U or R+              HH
ACR45288, 311H or HV or U or R+++  ++     CR*R*
Western group                      
F151, 172, 219, 311U6b++++ +          UU6b
F2298Pre*V or V++           VV
F3129, 242H or HV or U or R+              HH
F4298Pre*V or V++           VV
F5126, 288A, 292, 294T3++++        +   TT3
F6189, 298Pre*V or V++           VV
F7298Pre*V or V++           VV
F8189, 298Pre*V or V++           VV
F9189, 298Pre*V or V++           VV
F10126, 288A, 292, 294T3++++        +   TT3
F11189, 298Pre*V or V++           VV
F12172, 219, 261, 311U6b+++  +          UU6b
F13293, 311H or HV or U or R+              HH
F14126, 163, 186, 189T?++++        +   TT
F15201, 278H or HV or U or R+              HH
F16163, 186, 189T?++++        +   TT
F17298Pre*V or V++           VV
F1951, 129C, 182C, 183C, 189, 189insC(1), 362, 399U2++++ +          UU2
F20298Pre*V or V++           VV
F21298Pre*V or V+++           Pre*V or HVpre*V
F22CRSH or HV or U or R+              HH
F23189, 298Pre*V or V++           VV
F24126, 193, 294, 296, 304T++++        +   TT
F25189, 298Pre*V or V++           VV
F26162, 209H or HV or U or R+              HH
F27CRSH or HV or U or R+              HH
F28126, 288A, 292, 294T3++ +        +   TT3
F29129, 182C, 183C, 189, 223, 249, 311, 359M1+  +          + MM1
F30201, 278H or HV or U or R+     +       HH
F31298Pre*V or V+++           Pre*V or HVpre*V
F3293, 224, 311K+++++            KK
F3386, 104, 183C, 189,  189insC(1), 223, 271, 278X+++  ++      XX
F34126, 193, 294, 296, 304T2++++        +   TT2
F35298Pre*V or V++           VV
F36CRSH or HV or U or R+              HH
F37129, 182C, 183C, 189, 223, 249, 311, 359M1+  +          + MM1

Ancillary