Tuesday, May 16, 2017

Prehistoric migrations into South Asia

A new paper here, its abstract, and commentary from Rudradev on BRF:
A diffusion based study of population dynamics: Prehistoric migrations into South Asia
Mayank N. Vahia, Nisha Yadav, Uma Ladiwala, Deepak Mathur
Published: May 11, 2017

A diffusion equation has been used to study migration of early humans into the South Asian subcontinent. The diffusion equation is tempered by a set of parameters that account for geographical features like proximity to water resources, altitude, and flatness of land. The ensuing diffusion of populations is followed in time-dependent computer simulations carried out over a period of 10,000 YBP. The geographical parameters are determined from readily-available satellite data. The results of our computer simulations are compared to recent genetic data so as to better correlate the migratory patterns of various populations; they suggest that the initial populations started to coalesce around 4,000 YBP before the commencement of a period of relative geographical isolation of each population group. The period during which coalescence of populations occurred appears consistent with the established timeline associated with the Harappan civilization and also, with genetic admixing that recent genetic mapping data reveal. Our results may contribute to providing a timeline for the movement of prehistoric people. Most significantly, our results appear to suggest that the Ancestral Austro-Asiatic population entered the subcontinent through an easterly direction, potentially resolving a hitherto-contentious issue.
Rudradev's commentary:

The paper uses a mathematical model based on Fisher's equation (a reaction diffusion equation with positive constant parameters) to estimate patterns of population migration, and admixture, in the Indian subcontinent starting from 10kya.

It begins with the assumption that at least three of the five ancestral groups identified by Moorjani et al (https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3769933/) and Basu et al ( http://www.pnas.org/content/113/6/1594.full.pdf) using genome-wide analysis: Ancestral North Indian (ANI), Ancestral South Indian (ASI), and Ancestral Austro-Asiatic (AAA) began diffusing throughout the subcontinent from putative locations of origin around 10kya. Some would say this is actually a low estimate; according to the authors they may have arrived as many as 60 kya, but 10kya is the late-bound hypothesis they go with.

Note: The fourth group identified by the Moorjani and Basu studies is Ancestral Tibeto Burman (ATB), which probably only entered the subcontinent much more recently than the first three; and the fifth is almost entirely restricted to the Andaman Islands. However, virtually every Indian likely harbors genetic markers from both the ANI and ASI groups.

Using a computer simulation of preferential migratory patterns in which various parameters of Fisher's equation are modeled using satellite data: altitude, slope of land, proximity to water etc, the authors show that by 4000 years ago, ASI, ANI, and AAA had moved ALL over the subcontinent and admixture of populations had already begun. In effect, their data support the hypothesis that, even if ASI/ANI/AAA had begun diffusing from singular places of origin only 10kya, they would have spread all over the subcontinent, and commenced interbreeding with each other, as of 4 kya. Importantly, the authors note that 4 kya is roughly contemporaneous with the IVC.

Meaning: By the time IVC (and putatively, Saraswati civilization) were in existence, the "Ancestral North Indian" population (with many markers in common with Central Asian and European populations) and "Ancestral South Indian" population had already intermingled. The genetic admixture we see of ANI and ASI markers throughout India, had already happened as of IVC. In fact, in one of the simulations run by the authors, it is already occurring by 6 kya.

Therefore: the ANI population migrated into India well before dates of "Aryan Invasion Theory" posited by race-mongering Western Indologists (1500 BCE or thereabouts). After all if ANI and ASI had already intermingled by 4000 years ago, then both groups must have been present in India very long before that.

The "relative geographical isolation" refers to the Fisher-equation based model hitting a steady state at about 4 kya (in a simulation that begins 10 kya) where the overall level of migration becomes less extensive and more discontinuous, probably because tribes had staked out their pieces of territory separated by natural barriers like mountains or rivers. With less migration comes less admixture.

This is distinct from the deliberate practice of endogamy, also indicated by genetic sampling of caste and tribe groups, to have begun around 70 generations (~1600 years) ago during the Gupta era. That was a normative preference rather than a product of geographical isolation.

None of this has any bearing on the validity of an "Out of India" hypothesis... indeed, the paper builds on a body of evidence arguing that the ANI group clearly migrated from somewhere else into India at some later date than the ASI and AAA groups.