**2.4 Data analysis**

Phylogenetic analysis. The sequences of the pre-Hispanic PCR products from the HVI segment were aligned with representative Amerindian mtDNA control-region sequences (GenBank accession numbers: AY195760, EU719927, EU719811, EU719679, EU720004, EU720308, EU720078, EU719797, AY195749, EU720177, EU720123, EU719764, HQ012155, EU720242, HQ012184, HQ012164, HQ012134, AY195772, EU720073, EU720339, AY195759, EU720071, EU720336, EU720102, EU720202, HQ012188, HQ012198, EU720029, HQ012255, HQ012254, HQ012253, GQ449339, EU034320, AY195748, AF214088, DQ973581, AF478614) and two ancient sequences from a prehistoric Oneota population (Stone and Stoneking, 1998) using the Clustal W program (Thompson et al., 1994). Then, the phylogenetic tree was constructed with the Jukes-Cantor method, and the distances were obtained from a neighbour joining algorithm. Finally, the tree was optimised for maximum likelihood, using Hy-Phy software (Kosakovsky-Pond et al., 2005).

Haplotype network analysis. The median-joining (є=0) networks (Bandelt et al., 1999) of haplotypes were constructed using the Network package, v4.5.1.0 (Fluxus Engineering). Sequences used were those described for phylogenetic analysis. This method is for constructing networks from recombination-free population data.
