Genetic variation, covariation, and constraints in the evolution of sexual and clonal reproduction in a plant species
Authors: Jannice Friedman 1; Christina Steinecke 1; Matthew Rubin 2
Affiliations: 1) Queen's University; 2) Donald Danforth Plant Science Center
Keywords: Ecological & conservation genetics
Plants show incredible diversity in their reproductive strategies, and a central problem for evolutionary biologists is to understand the selective forces and genetic mechanisms that are responsible for the origins and maintenance of this diversity. Many species have the ability to asexually reproduce (for example, via stolons or corms) and sexually reproduce via flowers and seed. The question, then, is why some invest resources into both, and whether a trade-off exists when investing in traits associated with one mode or the other. Trade-offs, or constraints, can arise due to genetic correlations (through pleiotropy or linked genes); and if resources are finite through resource allocation trade-offs. Here, we use the wildflower, Mimulus guttatus (syn. Erythranthe guttata) to investigate phenotypic and genetic trade-offs between vegetative and reproductive allocation. We use a series of approaches to understand the maintenance of the variation in nature, the underlying genetic architecture, and whether trait variation and covariation constrains evolutionary trajectories. First we quantify variation in natural annual and perennial populations (n=80) from across the native range of M. guttatus (California to Alaska) and demonstrate strong correlations between vegetative and reproductive traits. To determine whether these multi-trait patterns arise from pleiotropic or independent loci, we mapped QTLs on a cross between divergent populations, and followed this with three new mapping populations using recombinant F4 individuals, to investigate fitness and identify QTL in a common garden field experiment in British Columbia. We find extensive pleiotropy for QTLs related to flowering time and stolon production, and reveal different genetic architecture among the crosses, and between field versus greenhouse experiments. Finally, we quantify standing genetic variation within a single perennial population, and conduct five generations of artificial selection on high and low stolon number in the greenhouse. We use these selection lines to examine the response to selection and identify traits that evolve or are constrained through correlated evolution. Overall, we find strong multivariate trait associations, pleiotropic QTL, and patterns of covariation among traits that may determine the trajectory of adaptive divergence.