Molecular Perspectives on the Origins of Chibchan Populations from the Sierra Nevada de Santa Marta, Colombia

Abstract

Current archaeological, biological and linguistic evidence points to a lower Central American origin for Chibchan speaking populations who are thought to have continuously occupied the region for the last 105000 years. However, the biological relationship of these groups to Chibchan speakers from Northern South America remains largely unresolved. This thesis examines mitochondrial DNA (mtDNA) haplogroup and haplotype diversity in three Sierra Nevada de Santa Marta Chibchan (Kogi, Arsario, Ijka) speaking populations and one neighboring Arawakan (Wayuu) group from Northeast Colombia in order to determine: (1) the nature of the biological relationship between the four study populations, (2) whether or not a relationship between Central and Northern South American Chibchan groups exists, (3) a potential timeframe for a Chibchan diaspora, (4) test hypothetical models regarding the initial peopling of the Santa Marta region and, (5) the role of Chibchan populations in the peopling of the Americas. Amerindian mtDNA haplogroups were characterized for 190 individuals using RFLP analysis and 61 HVS-I sequences were obtained. Three (A, B, and C) of the five founding Amerindian mtDNA haplogroups (A, B, C, D, and X) were found in these populations. The Kogi and Arsario exhibited only haplogroups A and C (Kogi 65% A, 35% C, Arsario 68% A, 32% C). The Ijka primarily exhibited haplogroup A (90%) with a single B (2.5%) and two C (7.5%) individuals. The Wayuu contained haplogroups A (34%), B (24%), C (32%), and undetermined (10%). Haplogroup D was not found in any of the groups examined. R-matrix analysis demonstrates that the three Santa Marta Chibchan populations are related to each other but not to the neighboring Wayuu. Analysis of these three South America Chibchan populations at the sequence level shows that they share low mtDNA haplotype diversity, low negative or positive values for Fu's Fs and Tajima's D and a peak between zero and one unit of mutational time with linguistically related populations from lower Central America and not with other indigenous South American groups. Phylogenetic reconstruction of these populations using median-joining networks indicates that all sampled Chibchan speaking populations had undergone a bottleneck and were highly influenced by a founder effect within the last 10,000 years. Using the p-statistic of Saillard et al (2000) on two clusters of Santa Marta Chibchan haplotypes gives mtDNA coalescence dates of 8,072 (±4943) and 6,985 (±3557) both of which are consistent with other temporal estimates of Chibchan genetic history. This time depth points to a long term occupation of Chibchan populations within Northern South America suggesting an in situ development for the Santa Marta groups. This study concludes that while there are biological similarities between the Chibchan speakers from the Sierra Nevada de Santa Marta and the Panamanian isthmus they diverged in the distant past. If a Chibchan diaspora did occur it may have been geographically widespread and would have occurred early during the peopling of the Americas. This diaspora may have blocked gene flow from the north and south possibly leaving genetic drift as the primary evolutionary force on the South American continent.