Fig. 6.
Genetic differentiation between temporal sample groups. (A) FST computed on aggregated sample groups is unable to differentiate the bottleneck and replacement models. “Moderns”: 20 samples from time 0. “Young”: 10 samples from time 0–5,000 years ago. “Old”: 10 samples from 5,500 to 10,000 years ago. (B) FST between individuals adjacent in time is able to detect a sudden increase in FST between the pair of individuals that flank the demographic event (both bottleneck and replacement), but we are unable to separate the replacement and bottleneck scenarios. Standard errors are not shown but ranged between 0.002 and 0.003. (C) FST between 20 modern individuals and each ancient individual. Standard errors are not shown but ranged between 0.0010 and 0.0014.