Skip to main content
. 2014 Jun 17;31(9):2516–2527. doi: 10.1093/molbev/msu192

Fig. 5.

Fig. 5.

Temporal sampling distinguishes genetic drift from population structure. (A) Constant population size model. (B) Bottleneck model. (C) Replacement model. (D) PC1 stratified by sample time under the constant population size model. (E) PC1 stratified by sample time under the bottleneck model. (F) PC1 stratified by sample time under the replacement model. Each colored circle corresponds to a single-sampled individual except for the large circles at time zero which corresponds to 20 sampled individuals in A, B, and C (in D, E, and F, the 20 individuals sampled at time zero end up on top of each other). FST between samples from before and after the bottleneck/replacement events at 5,500 years ago fails to distinguish between the models (FST = 0.0154 ± 0.0003 and 0.0153 ± 0.0003, respectively, see fig. 6).