Y-DNA-HAPLOGROUP-I-L Archives

Archiver > Y-DNA-HAPLOGROUP-I > 2009-03 > 1238441598


From: "Ken Nordtvedt" <>
Subject: [yDNAhgI] Haplogroup Takeovers
Date: Mon, 30 Mar 2009 13:33:18 -0600


I did a rough glacial maximum simulation of how just a few ydna lines take over a population in tough times. I let a population go for 9000 years at demographic no-growth to simulate a span of time from beginning to end of the LGM. Probability of male having zero male children equaled probability of male having two male children = .33; probability of male having one male child = .34

I let the PC run through the descendant population of a male 400,000 times

Probability that any male line would go extinct = 99.01 percent

"Average" number of males in the final generation = 1.007

So those 1 percent of males whose lines did not go extinct had to have quite a few offspring --- about 100 on average. I binned the final offspring populations for those few males whose lines did not go extinct and found this number of cases out of 400,000 runs and also indicated is the percent of the surviving population which such cases would represent.

0 - 15 offspring --- 517 cases ---- but producing only 1.2 percent of final surviving population
15 - 30 offspring --- 449 cases --- representing 3.1 percent of final surviving population
......
105-120 offspring --- 217 cases ---- 7.6 percent (peak bin)
......
225-240 offspring --- 59 cases ----- 4.3 percent
.....
285-300 offspring --- 37 cases --- 3.4 percent (high-offspring tail continues .....)

Suppose a particular region of Europe had 1000 males at the beginning. Only 10 of those male lines would be expected to survive. And of those ten surviving lines just a few of them would double digit populations, and the remaining survivors would be a collection of weaker and weaker surviving lines.

Each of those surviving lines had about 600 generations to drift away from the others, even if the initial 1000 males in that valley represented one haplotype bush from an earlier haplogroup. We'd end up with a decent number of very distinct clades

Ken


This thread: