Early-acting inbreeding depression can evolve as an inbreeding avoidance mechanism

Despite the potential for mechanical, developmental and/or chemical mechanisms to prevent self-fertilization, incidental self-fertilization is inevitable in many predominantly outcrossing species. In such cases, inbreeding can compromise individual fitness. Unquestionably, much of this inbreeding depression is maladaptive. However, we show that when reproductive compensation allows for the replacement of inviable embryos lost early in development, selection can favour deleterious recessive variants that induce 'self-sacrificial' death of inbred embryos. Our theoretical results provide numerous testable predictions which could challenge the assumption that inbreeding depression is always maladaptive. Our work is applicable any species that cannot fully avoid inbreeding, exhibits substantial inbreeding depression, and has the potential to compensate embryos lost early in development. In addition to its general applicability, our theory suggests that self-sacrificial variants might be responsible for the remarkably low realized selfing rates of gymnosperms with high primary selfing rates, as gymnosperms exhibit strong inbreeding depression, have effective reproductive compensation mechanisms, and cannot evolve chemical self-incompatibility.

The impact of Fisher's reproductive compensation on raising equilibrium frequencies of semidominant, nonlethal mutations under mutation/selection balance

Fisher's reproductive compensation (fRC) occurs when a species' demography means the death of an individual results in increased survival probability of his/her relatives, usually assumed to be full sibs. This likely occurs in many species, including humans. Several important recessive human genetic diseases cause early foetal/infant death allowing fRC to act on these mutations. The impact of fRC on these genetic conditions has been previously calculated and shown to be substantial as quantified by ω, the fold increase in equilibrium frequencies of the mutation under fRC compared with its absence, i.e. ω = 1.22 and ω = 1.33 for autosomal and sex-linked loci, respectively. However, the impact of fRC on the frequency of the much larger class of semidominant, nonlethal mutations is unknown. This is calculated here as ω = 2 - h*s for autosomal loci and ω up to 2 for sex-linked loci where h is dominance (varied between 0.05 and 0.95) and s is selection coefficient (varied between 0.05 and 0.9). These results show that the actions of fRC can almost double the equilibrium frequency of deleterious mutations with low values of h and/or s (noting that "low" is s∼0.05 to 0.1). It is noted that fRC may act differentially across the genome with genes expressed early in life being fully exposed to fRC while those expressed later in life may be unaffected; this could lead to systematic differences in deleterious allele frequency across the genome.

Variation in pollen sterility and gender specialization: an investigation with distylous species of Erythroxylum (Erythroxylaceae)

Differences in pollen sterility between morphs in distylous populations may represent the first step in the evolution of gender specialization. Theoretically, for partially or completely male sterile individuals to persist they must have some fitness advantage that compensates for the loss of male function. Gender specialization is considered a widespread process in Erythroxylum, but male sterility and the resource reallocation to female function have been investigated in few species and populations. In 18 populations of four distylous species of Erythroxylum, we quantified the levels of male sterility, estimated through pollen sterility, in short- and long-styled flowers to test if sterility is morph-biased. In one population per species, we also described the frequency of floral visitors, the production of flowers and fruits, and the quality of fruits and seeds of short- and long-styled plants to evaluate the expression of trade-offs in allocation to male and female function. In some populations of E. campestre and E. deciduum, short-styled flowers possessed higher levels of pollen sterility than long-styled flowers. Although most flowers of E. suberosum and E. tortuosum also expressed pollen sterility, the frequency of sterility was similar between morphs in all populations. Differences in reproductive output between morphs occurred only in populations of species with morph-biased sterility, but none of the variations reflect resource allocation to female fitness of short-styled plants. Differences in the level of sterility between morphs indicates the potential for gender specialization in populations of E. campestre and E. deciduum, despite the apparent lack of a trade-off in allocation.

Differential allocation of parental investment and the trade-off between size and number of offspring

When parents decide how much to invest in current versus future offspring and how many offspring to divide their current investments between, the optimal decision can be affected by the quality of their partner. This differential allocation (DA) is highly dependent on exactly how partner quality affects reproductive costs and offspring benefits. We present a stochastic dynamic model of DA in which females care for a series of clutches when mated with males of different quality. In each reproductive event, females choose the size and number of offspring. We find that if partner quality affects reproductive costs, then DA in total reproductive investment occurs only via changes in the number of offspring. DA in the optimal size of the offspring occurs only if partner quality affects the offspring benefit function. This is mostly in the form of greater female investment per offspring as male quality decreases. Simultaneously, we find that adaptive DA increases the number of offspring, and thus the amount of total investment, as male quality increases. Only certain model scenarios produce the positive DA in offspring size seen in empirical studies, providing a predictive framework for DA and how partner quality affects reproductive costs and offspring benefits.

Physiological effects of temperature do not explain prevalence of females in populations of gynodioecious Lobelia siphilitica growing in warmer climates

PREMISE OF THE STUDY

Gynodioecy is a sexual polymorphism whereby female and hermaphroditic plants co-occur within populations. In many gynodioecious species, stressful abiotic environments are associated with higher frequencies of females. This association suggests that abiotic stress affects the relative fitness of females and hermaphrodites and, thus, the maintenance of gynodioecy.

METHODS

To test whether abiotic stress affects the fitness of females and hermaphrodites, we grew open-pollinated Lobelia siphilitica families in temperature regimes characteristic of the southern portion of the species' range (where females are common) and the northern portion of the range (where females are rare). We measured physiological and phenological traits that are indicative of heat stress, and fitness components of females and hermaphrodites that could affect the maintenance of gynodioecy.

KEY RESULTS

Contrary to expectations if growth at high temperatures is stressful, we found that the hot treatment increased leaf chlorophyll content, decreased the percentage of plants that delayed flowering initiation, and did not affect the quantum efficiency of photosystem II. Growth at high temperatures did not affect the magnitude of the difference in rosette size (a correlate of flower number) between females and hermaphrodites, or the variance in pollen viability among hermaphrodites.

CONCLUSIONS

We found that growing-season temperatures typical of high female L. siphilitica populations were not stressful and did not affect either the fitness of females compared to hermaphrodites or variation in fitness among hermaphrodites. Consequently, further research is necessary to explain correlations between abiotic environmental factors and the frequency of females in this and other gynodioecious species.

Females allocate differentially to offspring size and number in response to male effects on female and offspring fitness

Female investment in offspring size and number has been observed to vary with the phenotype of their mate across diverse taxa. Recent theory motivated by these intriguing empirical patterns predicted both positive (differential allocation) and negative (reproductive compensation) effects of mating with a preferred male on female investment. These predictions, however, focused on total reproductive effort and did not distinguish between a response in offspring size and clutch size. Here, we model how specific paternal effects on fitness affect maternal allocation to offspring size and number. The specific mechanism by which males affect the fitness of females or their offspring determines whether and how females allocated differentially. Offspring size is predicted to increase when males benefit offspring survival, but decrease when males increase offspring growth rate. Clutch size is predicted to increase when males contribute to female resources (e.g. with a nuptial gift) and when males increase offspring growth rate. The predicted direction and magnitude of female responses vary with female age, but only when per-offspring paternal benefits decline with clutch size. We conclude that considering specific paternal effects on fitness in the context of maternal life-history trade-offs can help explain mixed empirical patterns of differential allocation and reproductive compensation.

The paradox of clonality and the evolution of self-incompatibility

In the January issue of New Phytologist Vallejo-Marín and O'Brien1 documented that in the genus Solanum (Solanaceae) clonality and self-incompatibility, a common genetic mechanism enforcing cross-fertilization, co-occur more often than expected by chance. Using a phylogenetic approach the authors showed that the statistical association between clonality and self-incompatibility persists even after taking into account phylogenetic relationships among species, uncertainty in the phylogenetic reconstruction, and associations between clonality and life history (annual/perennial). Vallejo-Marín and O'Brien1 suggest that clonality and self-incompatibility tend to co-occur because clonality, by allowing the persistence and propagation of a genotype in environments with limited pollinator or mate availability, reduces the selective pressure favoring the breakdown of self-incompatibility. In addition to promoting the maintenance of self-incompatibility, when clonality results in the spatial aggregation of genetically identical individuals, clonality may promote its breakdown by restricting pollen transfer between different genotypes. Here I call attention to these contradictory predictions of the effects of clonality on the evolution of self-incompatibility, and suggest that the outcome of this paradox depend on both the extent to which clonal propagation compensates for limited seed production, and on the extent to which clonality reduces pollen transfer between genotypes.