Sex uncovered: the evolutionary biology of reproductive systems.

Abstract

Reproduction is one of the most fundamental characteristics defining living organisms, yet it occurs via a tremendous diversity of mechanisms that differ in their potential for genetic mixing and evolutionary novelty. For example, sexual reproduction may involve mechanisms that promote outcrossing (such as mating types and self-incompatibility systems), whereas other species mostly self-fertilize. Individuals can specialize to different degrees into the production of either male or female gametes, which may in turn favour specific strategies for each sexual morph. Plants in particular feature a variety of sexual systems, where hermaphrodites may coexist with females and males (Barrett, 2002). Even among apparently straightforward, asexual routes to reproduction, there is considerable variation in the path taken (Suomalainen et al., 1987). Asexual reproduction can occur via production of mitotic eggs or via different forms of budding and vegetative reproduction. Other ‘asexual’ all-female lineages employ meiotic forms of parthenogenesis that vary in their genetic consequences. What are the processes that generate and maintain such a diversity of reproductive systems? Almost four decades after the seminal works of Williams (1975) and Maynard Smith (1978) on the evolution of sex, recent developments involving new technical tools and model systems have allowed these questions to be tackled from new angles. The contributions to this special issue illustrate some of the most recent achievements and review current and future challenges for the field. These recent developments show, in particular, that variation among reproductive systems is more easily explained when considering their effect at different levels: within populations, through direct and indirect fitness effects, and at the lineage level, through effects on rates of extinction or diversification. In this introduction, we give a brief overview of these different factors. The most often cited examples of direct fitness effects generated by reproductive systems result from anisogamy. Because males typically provide fewer resources to offspring than females, asexual females benefit from a higher reproductive rate. This ‘cost of males’ is twofold under equal investment of sexuals into the male and female functions (Maynard Smith, 1971). In hermaphrodites, the automatic transmission advantage of selfers over outcrossers, first described by Fisher (1941), is also a consequence of anisogamy: when self-fertilization has a negligible impact on the quantity of male gametes available for outcrossing, selfers benefit from a 50% advantage in an outcrossing population. This is because they transmit on average three copies of their genes, two by their selfed ovules and one by their outcrossed male gametes, whereas outcrossers only transmit two. Additional direct effects on fitness can emerge under certain ecological conditions. For example, habitats with low population densities or a lack of pollinators may generate situations in which finding a mate is difficult. Under these conditions, uniparental reproduction in asexuals or selfing hermaphrodites can provide an advantage over biparental breeding systems via reproductive insurance, and by avoiding costs associated with the production of gametes that will never be fertilized (e.g. Cheptou, 2004; Porcher & Lande, 2005; Charlesworth, 2006; Schwander et al., 2010). Selection for uniparental reproduction as a consequence of mate limitation is thus expected to be prevalent at species’ range margins, during colonization events following environmental changes (Pannell et al., 2014), or in metapopulations with high rates of population turnovers (Holt & Keitt, 2000; Sagarin & Gaines, 2002). Another consequence of pollinator-dependent reproduction in plants is that pollinator foraging behaviour may induce correlations between fitness components (male fertility, female self-fertility and female outcross fertility). As highlighted in the review by Devaux et al. (2014), this could explain the maintenance of mixed mating systems (involving both selfing and outcrossing; see also Johnston et al., 2009). Trade-offs between fitness components can also stem from physiological constraints. For example, female fertility is usually expected to be higher in females than in hermaphrodites, as the latter also invest resources into the production of male gametes. In populations displaying different sexual morphs (females, males and hermaphrodites), the relative fitnesses of these morphs depends both on such allocation trade-offs and on the relative abundances of the different morphs, which determine their mating opportunities (e.g. Lloyd, 1975, 1976). Several plant species display variation in the frequencies of sexual morphs across their geographical range, which makes them particularly suitable for testing models of sexual system evolution (e.g. Barrett, 1992; Pannell, 1997; Dorken et al., 2002). The studies by Yakimowski & Barrett Correspondence: Tanja Schwander, Department of Ecology and Evolution, University of Lausanne, CH-1015 Lausanne, Switzerland. Tel.: +41216924151; fax: +41216924165; e-mail: tanja.schwander@unil.ch