Correlated paternity measures mate monopolization and scales with the magnitude of sexual selection.

Indirect measures of sexual selection have been criticized because they can overestimate the magnitude of selection. In particular, they do not account for the degree to which mating opportunities can be monopolized by individuals of the sex that compete for mates. We introduce a measure of mate monopolization (m) based on the magnitude of correlated paternity and evaluate its ability to track changes in the magnitude of sexual selection. Simulation models were used to compare how well m tracked changes in the selection differential (s) for a trait regulating mating success. We further evaluated the association between m and s using ten replicate mating arrays of Sagittaria latifolia in which plants with contrasting alleles at microsatellite loci competed for siring opportunities. The computer models and mating arrays both demonstrated a positive linear association between m and s, supporting the utility of m as an index of sexual selection. Commonly used measures of sexual selection are not easily applied to organisms, such as the flowering plants, for which mating events are difficult to observe. The measure of mate monopolization introduced here could prove to be a useful addition to studies of sexual selection in these organisms.

Regional variation in sex ratios and sex allocation in androdioecious Mercurialis annua.

In androdioecious metapopulations, where males co-occur with hermaphrodites, the absence of males from certain populations or regions may be explained by locally high selfing rates, high hermaphrodite outcross siring success (e.g. due to high pollen production by hermaphrodites), or to stochastic processes (e.g. the failure of males to invade populations or regions following colonization or range expansion by hermaphrodites). In the Iberian Peninsula and Morocco, the presence of males with hermaphrodites in the wind-pollinated androdioecious plant Mercurialis annua (Euphorbiaceae) varies both among populations within relatively small regions and among regions, with some regions lacking males from all populations. The species is known to have expanded its range into the Iberian Peninsula from a southern refugium. To account for variation in male presence in M. annua, we test the following hypotheses: (1) that males are absent in areas where plant densities are lower, because selfing rates should be correspondingly higher; (2) that males are absent in areas where hermaphrodites produce more pollen; and (3) that males are absent in areas where there is an elevated proportion of populations in which plant density and hermaphrodite pollen production disfavour their invasion. We found support for predictions two and three in Morocco (the putative Pleistocene refugium for M. annua) but no support for any hypothesis in Iberia (the expanded range). Our results are partially consistent with a hypothesis of sex-allocation equilibrium for populations in Morocco; in Iberia, the absence of males from large geographical regions is more consistent with a model of sex-ratio evolution in a metapopulation with recurrent population turnover. Our study points to the role of both frequency-dependent selection and contingencies imposed by colonization during range expansions and in metapopulations.

The maintenance of hybrid zones across a disturbance gradient.

The parapatric distribution of genetically divergent lineages in hybrid zones can be maintained by ecological differences (dispersal-independent 'ecotonal' hybrid zones), by frequency- and density-dependent interference when they intermingle and mate (dispersal-dependent 'tension' hybrid zones), or by both processes acting together. One potentially important ecological factor that has received little theoretical attention is gradients in habitat disturbance. Such gradients may be particularly important in contact zones in which the interacting lineages differ in their sexual system (e.g., self-fertile versus obligately outcrossing) because self-fertility promotes the colonization of open patches. Here we use a spatially explicit metapopulation model to examine the dynamics of a dispersal-dependent ecotonal hybrid zone across a gradient in the rate of habitat disturbance, where competing lineages differ in their sexual system. We found that self-fertility promoted the maintenance of one lineage over its outcrossing counterpart at high extinction rates, predominantly because self-fertility confers reproductive assurance. Additionally, greater seed and pollen production promoted a lineage's persistence by reducing the seed fertility of its counterpart through hybridization. Our results draw attention to the joint effects of ecological and endogenous selection in regulating the location of hybrid zones. Our study also casts new light on the maintenance of the parapatric distribution of incompatible lineages of Spanish populations of the plant Mercurialis annua. In particular, we expect the rate of movement of a contact zone in eastern Spain to increase as it moves further south, contrary to earlier predictions.

Chloroplast haplotype variation among monoecious and dioecious populations of Sagittaria latifolia (Alismataceae) in eastern North America.

Aquatic plants commonly have extensive geographical distributions, implying few restrictions to dispersal. Here we investigate the postglacial history of an aquatic plant with contrasting sexual systems (monoecy and dioecy), which are predicted to affect dispersal ability. We examined the distribution of cpDNA haplotypes using polymerase chain reaction-restriction fragment length polymorphism (PCR-RFLP) among 76 populations (32 monoecious, 38 dioecious, two mixed and four undetermined populations) of Sagittaria latifolia sampled throughout eastern North America. We also use these data to investigate the polarity of the evolutionary transition between monoecy and dioecy. Using PCR-RFLP, we identified eight cpDNA haplotypes. All haplotypes were found in unglaciated areas of the species' range, clustered primarily in the southeastern United States, providing evidence that glacial refugia probably occurred in this area. Genetic diversity (hT) was more than six times greater among monoecious compared to dioecious populations. All seven of the haplotypes for which the sexual system could be determined were represented among monoecious populations. In contrast, only four haplotypes were detected in dioecious populations and 94% of individuals from dioecious populations possessed a single haplotype. Monoecious populations possessing this widespread haplotype were restricted to the southern portion of the range, indicating that dioecy probably originated in this region and then spread northwards. The distribution of cpDNA haplotypes in dioecious populations represents a subset of the variation found in monoecious populations, a pattern expected if dioecy has evolved from monoecy in S. latifolia.