This is one of many studies that have established the presence of substantial sub-Saharan ancestry in Portugal due to the slave trade.
Diversity of mtDNA lineages in Portugal: not a genetic edge of European variation.
Annals of Human Genetics 64 (6), 491-506.
[...]
These haplogroups have been reported to be characteristic of African populations, where their frequency is inversely correlated with the North-South axis: the frequency of U6 is high in North Africa and decreases in a southerly direction, being almost absent south of the equator; the L cluster has an opposite distribution (Rando et al. 1998, 1999; Watson et al. 1996; Mateu et al. 1996).
In Portugal, as well as generally in Iberia, many migration waves from both North and sub-Saharan African populations are well documented. The geographical proximity of North Africa and the Iberian Peninsula certainly afforded many opportunities for mutual population contacts. Among them, we stress the movement of Berbers and Arabs that took place during the very recent Muslim rule of Iberia (from the 8th century to the end of the 15th, in some regions). In addition, many sub-Saharan individuals entered the region during the slave trade period, from its very beginning (middle 15th century) until its total ban in the late 19th century.
As it would be interesting to find out the origin of the L and U6 sequences detected in Portugal, we have tried to compare the motifs of the sequences observed in Portugal with those described in the literature for several populations (Figures 3 and 4). However most of the matches found for the Portuguese sequences were with sequences widely distributed in Africa, and no clear pattern of geographic clustering was detected.
A striking aspect observed for the U6 haplogroup was that 5 out of 7 of the Portuguese sequences were unique to Portugal, not allowing, therefore, any accurate assignment of their geographical origin. The Canarian characteristic sub-haplogroup U6b1 (Rando et al. 1999), observed in other Iberian samples, was not detected in the present study.
Admitting that U6 sequences could have been at least partially introduced by Berber people during the Muslim rule of Iberia, it is strange to find them restricted to North Portugal. As a matter of fact, most historical sources document a deeper influence of Berber (as well as Arab) people in Central and particularly South Iberia (as judged from toponyms and general cultural aænities), compared to North Iberia where the Muslim presence is recorded to have been more ephemeral and consequently to have made less cultural and demographic impact. The data does not exclude the possibility that U6 introductions could have been additionally reinforced by later sub-Saharan inputs mediated by the African slave trade. Even if this mixed scenario is plausible, the presence of U6 sequences exclusively in North Portugal is a question that deserves further analysis. The hypothesis of an earlier introduction in the region does not seem to be favoured, neither by its presence in a restricted geographical area, nor by the high level of heterogeneity that characterizes the set of sequences that were found among this haplogroup.
With respect to the L sequences, it is widely accepted that they have a sub-Saharan origin, excepting some L3* lineages that, as analysis of Figure 4 suggests, might indeed have a non-African origin. The presence of L sequences in North African regions does not allow us to exclude the possibility that population influxes from this region, namely the above referred Berber/Arab movement, have introduced significant fraction of L sequences into Iberia. However, it seems more likely that most of the L lineages found nowadays in Portugal have been carried by African slaves, since the country was actively involved in the Transatlantic slave trade. Nine out of 17 L sequences found in this study showed matches with widespread African sequences, and with regard to the 8 remaining sequences the absence of matches can be due to the present bias in the description of sub-Saharan mtDNA variability. Broad areas corresponding to Ivory Coast, Angola and Mozambique, which represented very important sources of African slaves, remain uncharacterized.
There were more African slaves in Portugal than in any other European country: in 1550, Lisbon boasted 10000 resident slaves in a population of 100000, and Portugal as a whole probably had over 40000 (Thomas, 1998). In the mid-sixteenth century the birth of slaves' children was stimulated in Portugal for internal trade purposes. Inter-breeding between autochthonous individuals and African slaves certainly occurred and the predominant mating must have been between slave African females and autochthonous males, due to social pressures and also for legal reasons: offspring of slave females would be slaves, whereas offspring of slave males would not. Therefore, breeding between slave African males and white females, besides being socially repressed, would not bring any economic profit.
If the pattern of genetic admixture was markedly sex influenced, the signature of this recent African influence would be expected to be very different in the maternally inherited gene pool and in the paternally inherited one. In a recent study based on Y chromosome biallelic markers (Pereiraet al. 2000) we have reported the absence of typical sub-Saharan haplogroups in the Y chromosome Portuguese pool. This finding, and the detection of L sequences at 7.1% in the mitochondrial pool, both seem to support the above-mentioned pattern of admixture with African slaves.
Sharing the features of mtDNA diversity generally registered in Europeans (all European haplogroups were detected), Portugal has in addition received significant North and sub-Saharan African influences. Frequencies of haplogroups specific to these regions were higher than those reported for other European populations: 7% of North African sequences were detected (restricted to North Portugal and representing almost 3%of the total sample), and sub-Saharan African sequences were found to be spread throughout the country, with frequencies between 5% and 9.8%. Although statistically significant differences were not detected between the three sub-samples considered, the geographic distribution pattern observed for U6 and L sequences strongly suggest that different population movements were responsible for their introduction into the country, although none of them had enough demographic impact to induce regional differentiation.
The introduction of L sequences in Portugal was tentatively imputed mainly to the modern slave trade that occurred between the 15th and 19th centuries. Both the great number of slaves that entered Portugal and their very diverse African geographic origin are consistent with the data set now reported. However, we cannot exclude some North-African contribution to present-day Portuguese L lineages.
While the population movement associated with the slave trade may be responsible by some U6 inputs, we suggest that U6 sequences were predominantly introduced into Portugal during the Berber/Arab invasion of the Peninsula. However, the observation that haplogroup U6 is restricted to North Portugal is puzzling, considering the more pronounced impact of the Muslin rule in south Iberia and the widespread presence of African slaves throughout the country, and deserves further investigation.