Abstracts (first author)
The effect of linkage on sojourn times and patterns of genomic divergencePDF
The combination of spatially heterogeneous selection and gene flow can lead to clustering of adaptive mutations and concentrated genetic architectures. This might contribute to observed genome-wide variation in differentiation between populations or species, sometimes reflected in 'islands of divergence'. A mechanistic explanation is that tight linkage to previously established mutations reduces the effective migration rate experienced by novel mutations, which increases their potential of establishment. Deterministic and stochastic studies of invasion properties suggest that this benefit of linkage is constrained to a small genomic neighborhood. Other mechanisms such as so-called 'genomic hitch-hiking' or translocations might be needed to explain observed patterns. However, we argue that invasion properties alone yield an incomplete picture. With one-way migration and genetic drift, invasion of adaptive mutations is ephemeral. Therefore, it is essential to study the time a beneficial mutation spends in the population before loss. This sojourn time limits the potential for a further mutation to occur and profit from linkage. Using diffusion theory, we show that even under the strong assumption of quasi-linkage equilibrium, the presence of a previously established adaptive mutation may considerably elevate the sojourn time of a weakly beneficial mutation. Depending on the rate at which beneficial mutations occur, even loose linkage might therefore enhance their chance of invasion. In view of empirical studies, accurate estimates of both migration and recombination rates are necessary to judge if patterns of divergence and concentrated architectures can be explained by direct benefits of linkage to previously established polymorphisms.