Evolutionary time-series analysis reveals the signature of frequency-dependent selection on a female mating polymorphism.

A major challenge in evolutionary biology is understanding how stochastic and deterministic factors interact and influence macroevolutionary dynamics in natural populations. One classical approach is to record frequency changes of heritable and visible genetic polymorphisms over multiple generations. Here, we combined this approach with a maximum likelihood-based population-genetic model with the aim of understanding and quantifying the evolutionary processes operating on a female mating polymorphism in the blue-tailed damselfly Ischnura elegans. Previous studies on this color-polymorphic species have suggested that males form a search image for females, which leads to excessive mating harassment of common female morphs. We analyzed a large temporally and spatially replicated data set of between-generation morph frequency changes in I. elegans. Morph frequencies were more stable than expected from genetic drift alone, suggesting the presence of selection toward a stable equilibrium that prevents local loss or fixation of morphs. This can be interpreted as the signature of negative frequency-dependent selection maintaining the phenotypic stasis and genetic diversity in these populations. Our novel analytical approach allows the estimation of the strength of frequency-dependent selection from the morph frequency fluctuations around their inferred long-term equilibria. This approach can be extended and applied to other polymorphic organisms for which time-series data across multiple generations are available.