Sex-allocation trade-offs and their genetic architecture revealed by experimental evolution

The theory of sex-allocation is hailed as one of evolutionary biology's great successes. But while it has successfully predicted strategies of resource allocation to sons versus daughters under a wide range of conditions in species with separate sexes, it has been difficult to apply to hermaphroditic species because of the repeated failure to verify the fundamental assumption of a direct genetic trade-off between male and female allocations made by hermaphrodites. This assumption supposes that mutations that increase an individual's male allocation reduce its female allocation in equal measure, yet there is almost no evidence for such 'trade-off' loci in hermaphroditic populations. Here, we use experimental evolution to generate wide variation in hermaphroditic sex allocation in an annual plant and demonstrate a clear trade-off between the two sexual functions within populations as well as over time, with populations evolving both increased male and reduced female allocation. Quantitative trait locus (QTL) analysis performed on multiple crosses further reveals the segregation of allelic variation at trade-off loci. Taken together, our results demonstrate a phenotypic trade-off in sex allocation in hermaphrodites and provide compelling evidence for sex-allocation trade-off loci, verifying the fundamental assumption of sex-allocation theory and securing its general applicability to both dioecious and hermaphroditic species.