High parasite virulence necessary for the maintenance of host outcrossing via parasite-mediated selection

Abstract

Biparental sex is widespread in nature, yet costly relative to uniparental reproduction. It is generally unclear why self-fertilizing or asexual lineages do not readily invade outcrossing populations. The Red Queen hypothesis predicts that coevolving parasites can prevent self-fertilizing or asexual lineages from invading outcrossing host populations. However, only highly virulent parasites are predicted to maintain outcrossing, which may limit the general applicability of the Red Queen hypothesis. Here, we tested whether the ability of coevolving parasites to prevent invasion of self-fertilization within outcrossing host populations was dependent on parasite virulence. We introduced wild-type Caenorhabditis elegans hermaphrodites, capable of both self-fertilization and outcrossing, into C. elegans populations fixed for a mutant allele conferring obligate outcrossing. Replicate C. elegans populations were exposed for 24 host generations to one of four strains of Serratia marcescens parasites that varied in virulence, under three treatments: a heat-killed (control, noninfectious) parasite treatment, a fixed-genotype (nonevolving) parasite treatment, and a copassaged (potentially coevolving) parasite treatment. As predicted, self-fertilization invaded C. elegans host populations in the control and fixed-parasite treatments, regardless of parasite virulence. In the copassaged treatment, selfing invaded host populations coevolving with low- to mid-virulence strains, but remained rare in hosts coevolving with highly virulent bacterial strains. Therefore, we found that only highly virulent coevolving parasites can impede the invasion of selfing.

Keywords: Red Queen hypothesis; coevolution; experimental evolution; parasite; sex; virulence.

Figure 1.

Mortality of the ancestral mixed-mating (blue circles) and obligately outcrossing (red triangles) host populations following 48-hr assays on each of the ancestral pathogen strains. Each point represents the mortality rate on a replicate assay plate (each plated with an estimated 165–189 worms). Error bars represent ±1 SEM. Mortality rates were unaffected by host mating system and were significantly different for all possible pairwise comparisons of pathogen strains.

Figure 2.

Average mortality (±1 SEM) of host populations assayed on the ancestral pathogens (passage 0) or on their contemporary coevolved pathogens (passages 12 and 22).

Figure 3.

Selfing rates over time measured in C. elegans host populations evolving on heat-killed (control; left), fixed (nonevolving; middle), or copassaged (right) S. marcescens strains. Within each passaging treatment, replicate host populations evolved on the Db11 (low virulence, black square), Sm933 (mid virulence, blue circle), SmD1 (high virulence, green triangle), or Sm2170 (high virulence, red diamond) strain of S. marcescens. Error bars represent ± 1 SEM.