Inaugural Post

This post is to mostly document my thoughts on the settlement of the americas based on the genetics data.

Introduction

Colonization and settlement of the Americas is an ongoing area of archaeological and archeo-genomics research that with each new sequence a complex genetic history with relationships to multiple different world-wide populations is being revealed.  Patterson et al. (2012) noted a strong genetic relationship between Western European and Native American populations and they postulated the existence of an ancient common ancestor to both populations which was termed Ancient Northern Eurasians (ANE).  Raghavan et al. (2014b) analyzed the genetics of 24,000 year old (cal. 14C years BP) remains of a boy excavated in Ma’lta Siberia, demonstrated that the Ma’lta boy ( MA-1) contained genetic elements in common to both Native Americans and to some contemporary European populations.  They then postulated that MA-1 was representative of the putative ANE population (Raghavan et al. 2014b).  ADMIXTURE data clearly demonstrated that ~29% of the MA-1 genome contains alleles also present in Native Americans (Raghavan et al. 2014b; Supplemental Information (SI) p55).  Reciprocally, MixMapper analysis indicated that the Karitiana people as having 26.1% (range: 7.7-44.4%) of their ancestry coming from a lineage represented by MA-1 (Raghavan et al. 2014b; SI p76).    Further, contemporary European populations contain Native American related alleles that appear to have introgressed during the Bronze Age from mixing with Yamnaya Steppe herders whom share genetic affinity with MA-1 (Haak et al. 2015, Jones et al. 2015).  Flow of Far Eastern/Native American like alleles into Western Eurasians dating to around 4,000 years ago was previously demonstrated by Lipson et al. (2013).   Collectively, these data indicate that Western Eurasian and Native American pre-Columbian genetic relationships are mediated due to the flow of Native American-like alleles into Western Eurasians. 

It is unclear if the ancient population represented by MA-1 is ancestral to Native Americans or instead represents a mixture of a Western Eurasian Paleolithic Hunter gatherer population and a Far Eastern-related population(s) that was itself directly ancestral to Native Americans. Raghavan et al. (2014b; Figure SI17) used MixMapper analysis to demonstrate that the Karitiana people can be modeled as a mixture of Sardinia and Han populations and concluded that Native American populations contained Western Eurasian ancestry.  However, at least two studies have indicated the Sardinians have Native American-like alleles (Lipson et al. 2013, Patterson et al. 2012).  As Raghavan et al. (2014b; SI p77) indicated that estimates of Western Eurasian ancestry would be unreliable should the Sardinian population contain Native American-like alleles, the presence of Western Eurasian ancestry in the Karitiana people is questionable.  In addition, Reich et al. (2012; Table S3) estimated the ancestry in various contemporary Native American populations and demonstrated that individuals within 34 of 52 tested populations, mostly Central and South American, were not admixed and contained no Western Eurasian Ancestry.  Further, all tested individuals in multiple SA and CA populations have no Western Eurasian alleles (Reich et al., 2012, Raghavan et al., 2015).   The lack of Western Eurasian ancestry in SA populations is further supported by CHROMOPAINTER/ GLOBETROTTER analysis (Hellenthal et al. 2014; admixturemap.paintmychromosomes.com).  Overall, these data suggest that the Western Eurasian genetic components found in MA-1 were either lost during migration into the Americas or that they were never present in populations entering via Beringia. 

Over the last several years significant progress in elucidating the pre-Columbian peopling of the Americas has been achieved (Skoglund et al. 2015, Reich et al. 2012, Patterson et al. 2012, Raghavan et al. 2015, Raghavan et al. 2014b, Raghavan et al. 2014a, Malaspinas et al. 2014a, Rasmussen et al., 2015, Malaspinas et al. 2014b, Moreno-Mayar et al. 2014, Gravel et al. 2013, Moreno-Estrada et al. 2013, Jones et al. 2015).  Collectively, the data supports a model and timeline as in Table 1 of this manuscript.  This timeline is simplified and does not include all migrations so as to focus on events that are more related to Northeastern and Eastern North America.   Regardless, these data all suggest that the majority of genetic structure in contemporary Native Americans is derived from a Far-Eastern population(s) that genetically split from other Far Eastern populations by approximately 23,000 years ago (dating based on genetic calculations) (Llamas et al. 2016, Raghavan et al. 2015).  The Beringia populations remained may have remained isolated for ~7,000 years followed by an expansion southward beginning ~16,000 years ago, nearly 3,000 year prior to the opening of the ice free corridor (Heintzman et al. 2016, Llamas et al. 2016).  Subsequently, populations then diverged over time via genetic drift due to isolation and/or the introduction of additional genetics from other populations.

The Trans-Atlantic Hypothesis proposes that pre-Clovis and Clovis technology arose in the Americas due to a transfer of Solutrean technology via transport across the Atlantic during the LGM (Bradley and Stanford, 2004).  Multiple models for the way this could have taken place could be proposed.  The cultural diffusion model would suggest that a very small population made it across the Atlantic and taught the techniques to a population already present, with minimal genetic admixture.  Timing of the Beringia migrations into the Americas suggests that this model is unlikely as no population should have been present in Northeastern North America during the required time period (Llamas et al., 2016, Raghavan et al. 2015, Fagundes et al. 2008).  The demic diffusion model would involve successful colonization with a founding population from across the Atlantic.  This population would likely have been isolated for a significant period with cultural and genetic drift from its founding population.  This PPS population then developed the Clovis techno-culture as a derivative of Solutrean technology.  Based on genetics of contemporary populations, the later model would require subsequent and substantial admixture of the PPS population with population(s) that entered the Americas from Beringia. 

To date, no genome-wide study has been conducted that thoroughly tests Northeastern or Eastern populations for a Paleolithic source of Western Eurasian genetics.  Despite this lack of evidence Skoglund and Reich (2016) state: ‘This child [Aznick-1] was consistent with deriving all of his ancestry from the same founding populations as Central and South Americans, contradicting the ‘Solutrean’ hypothesis of trans-Atlantic migration...’  It should be noted that genomic sequences for the entire eastern half of North America are limited to ten individuals from the Ojibwa/Algonquin tribes (Reich et al. 2012) and one sample from 400 year old remains associated with the Mi’kmaq tribe (Raghavan et al. 2015).  The majority of the studies only use single nucleotide polymorphisms (SNP) for genome wide analysis.  SNP analysis is a good method for establishing relationships but is a proxy for whole genome sequencing and could miss relationships requiring higher resolution methods.  Lastly, in these studies relationships to Western Eurasians is deliberately suppressed (Malaspinas et al. 2014b, Gravel et al. 2013, Raghavan et al. 2015, Reich et al. 2012).  Reich et al. (2012; SI p14) removed Western Eurasian genetic components from analysis (mask) in a manner that was deliberately restrictive to confidently eliminate modern admixtures events or they selected individuals/populations with no apparent Western Eurasian or African genetics.  They claim that the methodology gives an unbiased masking of the data (Reich, et al. 2012, SI p15).  However, the Reich masking method can be shown to eliminate more genetic elements from analysis than would be predicted by ADMIXTURE analysis (Reich et al. 2012) and analysis of the individual masking data demonstrates a non-random and disproportionate over-masking  bias towards Northern North American populations (S. Oppenheimer, personal communication).  In Raghavan et al. (2015) Western Eurasian alleles in both whole genome sequences and SNP genomic data was masked and they state:  “The uneven distribution of the western Eurasian genetic components across individuals from Native American populations suggests that the admixture is relatively recent.  For this reason, admixed samples were masked for the western Eurasian (European) genetic component…”  It is a given that most Native American populations do have recent admixture with other populations.  However, no direct study has examined the Western Eurasian alleles present in NNA to the resolution sufficient to determine if any alleles have a pre-Columbian origin.   Therefore, claims that the currently available genetic data demonstrates that ancient Western Eurasian genetic components present in North American Natives did not arrive in the Americas from a PPS population via a Trans-Atlantic route are not satisfactorily demonstrated. 

It should be noted that the veracity of the Trans-Atlantic Hypothesis does not require transfer of Western Eurasian alleles.  However, should the demic model of Trans-Atlantic transfer of Solutrean technology to North America be correct then the Ojibwa and other North American populations would likely be modern genetic descendants of the PPS population.  However, the predominant interpretation of Eurasian and Native American genetic data currently dismisses the Trans-Atlantic hypothesis (Skoglund and Reich 2016).  This dismissal is done despite inadequate whole genome data, and the clear presence of Western Eurasian associated maternal and paternal lineages found at high frequency in Northeastern North American (NENA) populations.  The purpose of this work is to re-examine the genetic  data accumulated over the last several years that is relevant to the Trans-Atlantic hypothesis