The evolution of sex in spatially and temporally changing environments – SexChange

The theoretical part of the project will consist in the develoment of quantitative genetics models for the evolution of sex, in order to derive qualitative and quantitative results under more realistic scenarios than previous theoretical models: large numbers of interacting loci, distributions of selection coefficients and epistasis emerging from an adaptive landscape model. We will focus in particular on the following questions: - what are the relative strength of deterministic and stochastic effects acting on sex under stabilizing or directional selection ? - how is selection for sex affected by the shape of the fitness function and distribution of mutational effects? - how does the ploidy of individuals affect selection for sex ? - how do patterns of spatial or temporal changes in environment affect selection for sex ? - can we quantify the strength of selection for sex under different scenarios in terms of measurable variables? The experimental part of the project will consist in evolution experiments on laboratory populations of the monogont rotifer Brachionus plicatilis, exploring in particular: - how does the effect of sex on the mean and variance in fitness among offspring vary with population size ? - what are the relative contributions of recombination and segregation in the genetic effects of sex ? - how does sex affects the mean and variance in fitness in spatially / temporally changing environments ? - how do these results depend on the complexity of environmental change ?

The results obtained to date mainly concern the theoretical part of the project. We have explored a haploid population model, in which the fitness of an individual depends on an arbitrary number of polygenic phenotypic traits. The rate of sex of each individual (proportion of sexually produced offspring) is also a polygenic trait. The change in the mean rate of sex per generation can be decomposed into 3 terms, representing the effects of direct selection acting on sex (cost of sex), indirect selection due to the effect of sex on genetic variances and covariances among offspring, and indirect selection generated by the effect of sex on mean trait values among offspring. Under stabilizing selection, the effect of sex on genetic variances/covariances disfavours sex, while the effect on mean trait values may favour sex if the population is not at the optimum. Interestingly, indirect selection terms can be expressed in terms of quantities that could in principle be measured in an experimental population (effect of sex on the mean fitness of offspring, and on the response to selection among offspring). These mathematical predictions are new being tested using multilocus simulation programs.

The theoretical part will be pursued by the analysis of the diploid case. We will then consider different forms of environmental change (in space and time). In parallel, we are now setting up the experimental system which will allow us to test theoretical predictions using populations of monogonont rotifers.

An article presenting theoretical results is under completion.

The evolution and maintenance of sexual reproduction has been one of the major questions in evolutionary biology for the last decades: although biparental sex entails many costs, pure asexuality is rare in the eukaryotic kingdom, and a substantial number of organisms are obligate sexuals. Understanding why this costly mode of transmission of genetic material (as opposed to clonal propagation) is so widespread still represents of one the greatest challenge for evolutionary biology.

On the theoretical side, important progress has been achieved over recent years: in particular, several plausible scenarios generating indirect selection for genetic mixing have been proposed, such as temporal fluctuations of the environment (generated for example by host-parasite interactions) or stochastic effects due to finite population size, generating strong interference between selected loci in non-recombining populations. However, it is still not clear to what extent the different theories proposed can explain the evolution of high rates of sex in the presence of strong direct costs. Furthermore, most of these models make simplistic assumptions about the genetic architecture of fitness: epistatic interactions between genes are either neglected, or assumed to be the same between all pairs of loci. Finally, experimental tests of theories on the possible benefits of sex remain scarce. Although experimentation in natural populations is technically difficult, experimental evolution on laboratory populations emerged as a promising approach to test theoretical predictions: in particular, experiments on different model species showed that sex accelerates adaptation to new conditions. However, in most of these experiments the environment stayed constant in time and space.

The SexChange project proposes to explore selective forces acting on sex in different types of environments (stable vs. changing in time or space) using a combination of theoretical and experimental approaches. The theoretical part will consist in using adaptive landscape models representing selection acting on a number of quantitative phenotypic traits. Interestingly, these models capture different aspects of the complexity of interactions between genes, such as distributions of epistatic effects and possible compensatory effects among mutations ("sign epistasis"). They have been increasingly used over recent years to explore the dynamics and adaptation and generate predictions for the distribution of fitness effects of mutations (which have been validated by experimental data), but have rarely been used to study the evolution of reproductive systems. We will use a combination of analytical and simulation methods to explore selection for sex under (i) stable environmental conditions, or (ii) temporally or (iii) spatially varying environments. These models will allow us to investigate the effect of the genetic architecture of adaptation on selection for sex, and quantify the relative importance of stochastic and deterministic forces.

The second objective of the project is to test theoretical predictions by evolution experiments on facultatively sexual rotifers (Brachionus plicatilis) in different types of environments. Mongonont rotifers represent a particularly interesting system to explore selective forces acting on sex: they are easy to maintain in the lab, have short generation times and become sexual in response to an environmental stimulus. We will use these organisms to explore the consequences of sexual reproduction on the mean and variance in fitness among offspring in different sets of conditions: stable environment with different population sizes, temporally or spatially changing environment (considering different degrees of complexity of environmental change). This combination of experimental and theoretical approaches is expected to yield new insights on one of the most challenging questions in evolutionary biology (why sex).