The fresh model are adopted playing with numerical simulation tracking trait wavelengths when you look at the for every single sandwich-population over time. Since a separate replication of your overall performance, the new model was also implemented as an individual-based design where anyone and their attributes is explicitly artificial, and that gave almost-similar results (discover S1 Efficiency).
Migration versus acculturation eliminates between-classification adaptation.
Without acculturation, migration always eliminates between-group variation, just as in Wright’s original island model and specified by Eq 1. Fig 1 shows time series of FST for low (m = 0.01), typical (m = 0.1) and high (m = 0.3) migration rates. In all cases, the grey a = 0 lines fall to zero. Each trait in each population converges on the mean frequency of that trait across the whole population. Because each sub-population is equally sized, and each trait starts out at unity in one of these s sub-populations, this mean frequency = 1 / s for every trait. This can occur rapidly: for m = 0.01, the perfect between-group structure completely disappears within 300 timesteps, while for m = 0.1 it disappears within 30, and for m = 0.3 within 10.
Time series showing changes in FST over time for (A) a low migration rate m = 0.01, (B) a typical migration rate m = 0.1, and (C) a high migration rate m = 0.3, at varying strengths of acculturation, a. Other parameters: s = 5, n = 5, r = 0.
Conformist acculturation is look after between-group cultural version.
Fig 1 shows that as conformist acculturation increases in strength, between-group cultural variation is maintained (FST > 0) at equilibrium.