The Origins of the British: The New Prehistory of Britain (73 page)

39
. Harding et al. (2000), Rees (2000).

40
. Tambets et al. (2003).

41
. Present study (mtDNA database), and Tambets et al. (2003).

42
. This problem is acknowledged by Richards in an excellent review and update of the Richards et al. (2000) founder paper, ‘The Neolithic invasion of Europe’: ‘Finally, the results are, at best, estimating the proportion of lineages in the present-day population that can be attributed to each founder event from the Near East (or to bottlenecks within Europe), rather than from the immediate source region’ (Richards 2003, p. 152). On the expansion signalled initially by Cardial Ware, he notes that ‘There are fewer Neolithic-derived lineages along the Mediterranean and the Atlantic west, at around 10%, again mainly from haplogroup J’ (p. 153).

43
. ‘The Basque region, which was an outlier in the PC analyses of both mtDNA and classical markers, has the lowest Neolithic component, at around 7%. The Basque outlier status may therefore be partly the result of reduced Neolithic penetration, as well as considerable genetic drift due to isolation and small population size. They are little more of a Mesolithic relict than any other European population’ (Richards 2003, p. 153).

44
. Represented e.g. by mtDNA haplogroup H5a, which dates to 8,000 years ago (
n
= 32, SE ±4,000), see Pereira et al. (2005).

45
. McEvoy et al. (2004).

46
. Richards (2003), p. 153.

47
. Percentage of Ivan in English: present study 19.5% (172/883). See also Rootsi et al. (2004), estimated on same dataset but including several larger islands
around England: 18.4% (174/945). For the role of I in the post-glacial and Neolithic expansions from the Balkans, see also Passarino et al. (2002).

48
. 8.3% (168/2,082), present study.

49
. ‘Gaelic Modal Haplotype’ = Ht. 199 in present study. Gaelic-named male data from Hill et al. (2000).

50
. The original publication on Rory’s ‘Gaelic connection’ (Hill et al. 2000) suggested a Neolithic age for the whole cluster.

51
. Rory re-expansions: Mesolithic: R1b-14b, distribution Wales and Western Scotland > Ireland, age: 8,440 years (
n
= 40, SD +/4060). Neolithic: R1b-14c, distribution Ireland > Scotland, age 6,541 years (
n
= 53, SD ±4,614). Irish Neolithic: R1b-14a, mainly Ireland, age 5,450 years (
n
= 83, SD ±2,090).

52
. Age of R1b-11 to 13: 9,540 years (
n
= 225, SD ±4,740); R1b-13, 7,790 years (
n
= 79, SD ±4,090); R1b-11, 4,560 years (
n
= 121, SD ±3,370); and R1b-12, 4,620 years (
n
= 25, SD ±4,140). Rate of R1b11 plus R1b-12: 7.0% (146/2,082).

53
. FMH is the modal haplotype (Ht. 151 in the present study and closely related to the AMH) of cluster R1b-8. The name is not intended to identify Frisia as the source, since it is also present in the Iberian refuge region.

54
. One-step derivative: Ht. 150.

55
. R1b-8 (Ht. 150–153 and 17; frequency as percentage of R1b in Durness 36.6% (15/41), as a percentage of total Durness 29.4% (15/51). R1b-8 founding age in British Isles: 4,674 years (rooted on Ht. 151,
n
= 96, SD ±4,674). Total size of R1b-8 in UK:
n
= 273. R1b-8 age in Frisia (no uniquely derived lineages): 2,024 years (
n
= 19, SD ±1,510).

56
. R1b-7 (Ht. 145–8) (= small cluster). Founding age in UK 3,940 years (
n
= 13, SD ±3,119).

57
. Neolithic contributions by re-expansion of existing clusters in the British Isles:
n
= 136 (R1b-14) + 146 (R1b-11 and 12) + 109 (R1b-7 and 8) = 391; 391/2,082 = 18.8% of male lines (or 25.8% of R1b in British Isles).

58
. Rate of group I in British Isles: 16.1% (336/2,082); age of haplogroup I in Europe: see Rootsi et al. (2004) and Semino et al. (2000).

59
. Age of I in entire present dataset, rooted on Ht. 311: 54,100 years (
n
= 627, SD ±16,240).

60
. Present study (entire database): I1a age 14,940 years (rooted on Ht. 393,
n
= 458, SD ±6,428); I1c, 20,830 years (rooted on Ht. 346,
n
= 109, SD ±4,550);
and I1b, 21,000 years (rooted on Ht. 299,
n
= 41, SD ±6,970). See also Rootsi et al. (2004), table 3: times since sub-clade divergence: I1a, 15,900 (±5,200) years; I1c, 14,600 (±3,800) years. See also Semino et al. (2000), Eu7 and Eu8 inferred estimate c.22,000 years; and Inos in Oppenheimer (2003), pp. 146, 151 (inferred 33,000 years by analogy with age of mtDNA group HV).

61
. One of the problems in working out the routes the Ivan subgroups took after the end of the Ice Age is in knowing where they had each already spread to
before
the LGM. While the main source of the Ivan group information (Semino et al. 2000, Rootsi et al. 2004) suggests that although Ivan may have already spread in a limited way with the Gravettian technology before the LGM, the main expansions of its subgroups occurred after the LGM. This is borne out in the divergence ages of each subgroup, which by their estimates are all post-glacial. Furthermore, if we look at which part of the Ivan geographical distribution holds all his ancestral as well as the filial diversity (i.e. including I* and I1a to c), we find that they are all in the same distribution as I1b*. In other words, southeast Europe, where Ivan represents up to 45% of male lines, has every appearance of a homeland (see
Figures 3.7
,
4.11a
and
4.11b
in this book). I should make it clear that this Balkan homeland of all the main branches of Ivan is not what is proposed by Rootsi et al. (2004), but the movement of Ivan (and R1a1) into Germany and Norway during the Neolithic is supported to a certain extent by Passarino et al. (2002).

62
. Present study. Age of I1b*: 23,000 years (rooted on Ht. 300,
n
= 24, SD ±7,870); I1b2, 14,330 years (rooted on Ht. 289,
n
= 17, SD ±7,500). See also Rootsi et al. (2004), who give ages of, for I1b*, 10,700 (±4,800) years and, for I1b2, 9,300 (±7,600) years.

63
. Another possibility is that the Balkan refuge extended a little farther to the east, around the north coast of the Black Sea, which during the LGM would have been a shrunken freshwater lake.

64
. Piggott (1965).

65
. British ages of the three I1c clusters: I1c-1, 12,800 years (
n
= 12, SD ±7,400); I1c-2, 14,200 years (
n
= 19, SD ±3,880); and I1c-3, 11,600 years (
n
= 25, SD ±5,470). See also similar dates for I1c in Western Europe in Rootsi et al. (2004), table 3.

66
. Ages: I1a-4, 7,700 years (rooted on Ht. 401,
n
= 20, SD ±4,440); Ib-2, 8,600 years (
n
= 6 unique descendants, SD ±4,270), dated in British expansion; total I1b2 in British including types shared with Iberia = 14 in this study. Numbers
used for founder calculations include only uniquely British derived haplotypes – i.e. less than total immigrants for each cluster (see Appendix C). For comparison, ages estimated by Rootsi et al. (2004) for I1a: time since sub-clade divergence (i.e. whole branch) 15,900 (±5,200) years, age of STR variation 8,800 (±3,200) years, time since population divergence 6,800 (±1,900) years.

67
. For distribution and expansion ages for I1a, I1b* and I1b2 in Europe, see Rootsi et al. (2004), figure 1 and table 3.

68
. Gronenborn (2003).

69
. Age of I1b*: see Rootsi et al. (2004), table 3. STR variation 7,600 (±2,700) years; time since population divergence 7,100 (±2,500) years. For the geographical distribution of I1b* and I1b2, see figure 1 in Rootsi et al. (2004). From table 3 in Rootsi et al. (2004): I1b2 age of STR variation, 8,000 (±4,000) years; time since population divergence, 7,900 (±3,600) years.

70
. European rates of I1b2: see Rootsi et al. (2004), table 1. Rates in British Isles: present study. Age of I1b2 in British Isles in present study 8,600 years (
n
= 6, SD ±4,270).

71
. Figure 1c in Rootsi et al. (2004).

72
. Gronenborn (2003).

73
. Percentages of I1a for Germany, the Netherlands, Switzerland and Normandy from Rootsi et al. (2004), table 1; for Belgium inferred from Hg-2 in Rosser et al. (2000). In spite of the view of Rootsi et al. (2004) that ‘the I1a data in Scandinavia are consistent with a post-LGM recolonization of north-western Europe from Franco-Cantabria’, their own figures for the presence of I1a show low rates in southern France (5.3%) and Lyon/Poitiers (2%). Only Normandy, in the north, shows appreciable rates (11.9%).

74
. I1a-2, 3 and 4: total 173/233 = 74% of British I1a. I1a-3: 30.4% of British I1a (71/233), British date 1,200 years (rooted on Ht. 398,
n
= 68, SD ±400). ‘Bronze Age’ of I1a-3 in southern Scandinavia 3,608 years (rooted on Ht. 398;
n
= 164, SD ±1,390).

75
. I1a-3 is also found both farther north, in Norway and Sweden (I1a in southern Sweden 35.7% (
n
= 168), in Norway 38.9% (
n
= 72), Rootsi et al. (2004); in Oslo 33% (9/27), present study), and south across the Baltic in Denmark. It is less common, however, in northern Germany (I1a in Germany 25%, Rootsi et al. (2004); in northern Germany 7.5% (14/185), present study). Unlike all the other I1a branches, however, I1a-3 has a significantly lower and more focused
representation in Britain than might be expected from its high frequency in Norway and the distribution of other I1a clusters there (I1a-3 rate in Norway: 11.6% (33/284), and 3.4% in the British Isles (71/2,082); odds ratio 0.29, i.e. I1a-3 is only about one-third of its expected frequency in the British Isles). This low ratio is consistent with the limited nature of the Viking invasions compared with the Neolithic Scandinavian entrants.

76
. I1a-5 is 5.2% of British I1a (12/233). Absolute rate in Western Isles, 5.7% (5/88); age, 5,700 years (
n
= 9, SD ±2,010).

77
. I1a-2, 27.0% of British I1a (63/233).

78
. I1a-3 is 31.8% of British I1a (74/233); I1a-3 constitutes 9% of Danish I1a (9/100), 14% (12/85) of Schleswig-Holstein/northern Germany and 6.3% of Frisia (6/94) – all from present study.

79
. I1a-4 rates: Oslo 9.1% (6/66), Bergen 7.8% (10/128), Denmark 4% (4/100), Frisia 4.3% (4/94); England overall rate 1.3% (28/2,082), Fakenham 7.5% (4/53); British age 7,700 years (rooted on Ht. 401,
n
= 20, SD ±4,440).

80
. Nine unique derived British haplotypes; four others shared with the Continent.

81
. I1a-7 rate among British I1a types 9.4% (22/233). While this is the British cluster nearest to Frisia in identity, it does not show evidence of a recent invasion: one sub-cluster, which shows a clear founding event in East Anglia, has a date suggestive of the early Bronze Age. Overall rate I1a-7 rate in British Isles 1% (22/2,082), age of I1a-7b cluster 3,940 years (rooted on Ht. 461,
n
= 13, SD ±2,420).

82
. I1a-3 represents 30.4% of British I1a types (71/233) and 3.4% in Britain overall (71/2,082); frequency in Bergen 15.6%, in Oslo 13.6%, Denmark 12%, eastern England 8–11% (all figures present study). However, although it is the only dated late founder, I1a-3 is probably not the only I1a cluster to enter Britain from the Anglo-Saxon region during the Dark Ages, with I1a-2, I1a-4, I1a-6 and I1a-7 (in that order of importance) contributing overall a further 3– 5% in eastern England.

83
. For example, I1a forms a large part of haplogroup 2 in the Principal Components Analyses by Weale et al. (2002) and Hill et al. (2000).

84. E3b is most common in the southern Balkans and southern Italy (25%), and in southern Spain (10%). E3b is rare to absent in the Basque Country and the southern French Ice Age refuge regions, although it is common at 6–14% in other parts of southern Spain and the Spanish coast, both Atlantic and
Mediterranean (Semino et al. 2004, Cruciani et al. 2004, and present study), whereas E3b is found in Galicia (11.3%) and Valencia
(6.5
%
).
Age of E3b in the British Isles: main cluster 4,500 years (rooted on Ht.
66, n
= 20, SD 2,480). Three smaller E3b clusters give dates of the same order – total British
n
= 47. Highest rate at Abergele 33.3% (6/18 four haplotypes belonging to three clusters, i.e.
not
an extreme founder event); Southwell 5.7%. For European/African distribution and ages of E3b and sub-groups see Semino et al. (2004) and Cruciani et al. (2004); see also, Richards (2003). For distribution of the E-M78-α sub-cluster of E3b as described in text and shown in Figure
5.8
[new], see Cruciani et al. (2004). For distribution of the E-M78-α sub-cluster of E3b as described in text and shown in figure
5.8,
see Cruciani et al. (2004).