Figure 5

Allele behaviour of sexual and clonal populations under lag load. (a) A sexual population (N = 3000; host mutation rate, μ _h = 10^-4 bits/allele/generation; parasite mutation rate, μ _p = 10^-2 bits/allele/generation; number of loci, L = 3) at steady state. The population adaptation score for each locus is an average of 0.726, and the overall population adaptation score is 0.382. Allele frequencies here, and for (b) below, are for those alleles whose maxima exceed a frequency of 600. Different colours represent different allele species. Most alleles do not cycle more than once but there are occasional instances of bimodality, indicating a temporary arrest of a parasite allele in allele space. (b) Clonal population (N = 3000; host mutation rate, μ _h = 10^-4 bits/allele/generation; parasite mutation rate, μ _p = 10^-2 bits/allele/generation; number of loci, L = 3) at steady state. The population adaptation score for each locus is an average of 0.564, and the overall population adaptation score is 0.181. Maximum allele frequencies are capped at 3000 under substantial lag load for reasons discussed in the body of the paper, making the clonal population fully heterozygotic under these conditions. High-frequency alleles, capped at 3000, show much longer lifetimes than their sexual counterparts, due to interference with selection. An example of this interference is the fluctuation in allele frequency associated with very long-lived alleles (illustrated by the red line), where partial dips in frequency followed by recovery indicate early negative selection on the genotype in question, followed by an advantageous mutation at one of the other coupled loci. This brings the genotype as a whole under positive selection, reversing the impending early loss of the allele whose frequency is shown in red. Final elimination of this allele runs beyond the end of the simulation.