Spatial soft sweeps: Patterns of adaptation in populations with long-range dispersal
Fig 9
Different measures of diversity within a subrange are sensitive to different characteristic scales.
(a) Probability Phard,s(2) of observing a single allele in a pair drawn from a subrange of size Ls for different dispersal kernels (colours, labeled) and mutation rates [symbols, see legend in panel (b)], for 1D simulations with L = 106, as a function of the ratio Ls/〈req〉. In all cases, the population range was chosen to be many times larger than the characteristic size χ and harbours many distinct alleles. The dashed line is the prediction if all clones are of the same size Xave, in which case geometry dictates that Phard,s(2) = {1 − x/3, x < 1; 1/x − 1/(3x2), x ≥ 1}. The inset shows data for μ = {0.6, 1.0, 3.0} on log axes. (b) Number of distinct alleles nc,s observed in a subrange of size Ls, shown as a function of the ratio Ls/〈rmax〉. Values are scaled by 〈rmax〉/〈req〉, the expected number of clones in the area occupied by the average halo. The solid line corresponds to nc,s〈req〉/〈rmax〉 = Ls/(2〈rmax〉), or equivalently nc,s = Ls/(2〈req〉).