Demographic Expansions In An Amerindian Brazil

RAMALLO V; BISSO-MACHADO, R.; BRAVI CLAUDIO M.; COBLE, M; SALZANO FM; BORTOLINI MC; HUNEMEIER, T.

Foz de Iguaçu, PR

Congreso; 58° Congresso Brasileiro de Genética; 2012

Sociedade Brasileira de Genética

Native Americans have been analyzed on various aspects concerning their linguistic affiliation, as well as the relationship of these to related geographical and genetic data. Native Brazilian linguistic landscape present between 154-170 languages divided in approximately 20 major linguistic stocks. Despite of this enormous diversity and controversies relative to relationship and hierarquical classification of these languages two major stocks are normally recognized by allauthors: Tupi and Ge. The present investigation was performed to test if the expansion model proposed by linguistics is reproducible with genetic data for Tupi and Ge individuals. In this study a total of 693 Native Americans (Tupi, n = 524; Ge, n = 169) from 15 Tupi populations and 11 Ge populations were analyzed through their mtDNA (HVSI). The different lineages/haplotypes found were compared between groups and their hierarchical relationship drawn in network webs. Mantel tests were conducted separately for both Tupi and Ge language stocks to test for correlations between genetic, geographic, and linguistic distances between populations. The Tupi stock populations showed that mtDNA distances across three South American Tupi regions are positively and significantly correlated at the 0.05 level of probability with geographic distance but not with linguistic distances, measured by years of separation between them. On the other hand, Ge groups did not show any association between genetics and geographic or linguisticdistances. Spatial autocorrelation analyses were also undertaken so to provide a measure of the genetic similarity of individuals as a function of the geographic distance between them. The populations belonging to both linguistic stocks present distinct geographic location, occupying areas of different surfaces. In the Tupi r values plot, significant negative autocorrelation occurs as the distances covered in each class begin to increase. In the last distance class (inferiorlimit 3,180 Km), the value of r is the lower -0.343 (p=0.000). In this class a comparison between divergent groups of North, Northeast and South Tupi occurs. These results are in agreement with a model of isolation by distance. In the Ge correlogram, from distance class comprising 1,225 to 1,715 km, r values are not significant and surprisingly begin to increase to a maximum of 0.538. Here are included most comparison between sequences sampled from Central andSouth Ge tribes. Taking together the results showed above for the two main linguistic stocks found in Brazil, we clearly see remarkable differences between Tupi and Ge expansions. While Tupi presented a great and radial dispersion within South America, reaching 3,000 Km (North-South) in 5,000 YBP, Ge groups presented an intricate and non-linear way of dispersion, had more discrete geographic expansion, but an extraordinary cultural and linguistic one over 6,000 YBP.