GENEALOGY-DNA-L Archives

Archiver > GENEALOGY-DNA > 2011-09 > 1316278041


From: "Kenneth Nordtvedt" <>
Subject: [DNA] Wild Demographic Dice
Date: Sat, 17 Sep 2011 10:47:21 -0600


I am studying again the wild statistical flucuations of y line survival and prosperity due to PURELY the statistics of reproduction.

Here is a run I did this morning to illustrate things. A population of 1000 males undergoes 4 percent per generation growth for past 150 generations, resulting in 370,000 males today. With sufficient dna tags (STRs, SNPs, etc) one could imagine each of those 1000 original males representing a distinctive genetic y line.

What do we see today among those 370,000 descendants in the way of distinct original lines? 87 percent of those 1000 lines went extinct; they are gone from today’s population. The most populous represented original line accounts for 4.4 percent of today’s population. The 10 most populous lines represented today account for 24.1 percent of the 370,000. And the 20 most populous lines represent 40.6 percent of the 370,000.

The above results indicate the very strong creation of demographic “winners” and “losers” of y lines based solely on statistical flucuations --- before any other factors are brought into consideration. This source of demographic winners versus losers must be understood before speculations about connnections of populous y haplogroups today to migrating, militarily advanced, technically advanced, culturally advanced......... etc. tribes of the distant past can be even tried intelligently. There are some forums engaging almost exclusively in various speculations of this latter type, and I think this is all mostly fairy tales until we can better understand how much of the y haplogroup or clade clumping looking back several thousand years to its foundings is accounted for by the basic statistical flucuations of routine male reproductive results.

So I’m urging some mathematically inclined folks to get into extinction and demographic analysis, so some of us can talk about this more.

Ken

This thread: