386I.387II.388III.390IV.393V.394395References395 Summary A compound hypothesis positing that self‐fertilization is an evolutionary dead end conflates two distinct claims: the transition from outcrossing to selfing is unidirectional; and the diversification rate, or the balance of the speciation and extinction rate, is negative for selfing species. Both claims have enjoyed widespread informal support for decades, but have recently come under suspicion. Sources of data that apparently contradict strongly asymmetric mating system transitions often rely on statistical phylogenetic tests plagued by profound flaws. Although recently developed models mend preceding approaches, they have been employed sparingly, and many problems remain. Theoretical investigations, genetic data and applications of new phylogenetic methods provide indirect support for an association of selfing with negative diversification rates. We lack direct tests of reversals from selfing to outcrossing, and require data concerning the genetic basis and complexity of independently evolved outcrossing adaptations. The identification of the mechanisms that limit the longevity of selfing lineages has been difficult. Limitations may include brief and variable durations of selfing lineages, as well as ongoing difficulties in relating additive genetic and nucleotide variation. Furthermore, a common line of evidence for the stability of mixed mating – based simply on its frequent occurrence – is misleading. We make specific suggestions for research programs that aim to provide a richer understanding of mating system evolution and seriously challenge Stebbins’ venerable hypothesis.

The transmission advantage and reproductive assurance ideas describe components of gene transmission that favour selfing. Future work should move beyond their dichotomous presentation and focus upon understanding whether selection through pollen, seed or both explains the spread of selfing-rate modifiers in plant populations.

656I. 657II. 658III. 660IV. 661V. 663VI. 663VII. 664VIII. 664 665 References 665 Summary Baker's law refers to the tendency for species that establish on islands by long‐distance dispersal to show an increased capacity for self‐fertilization because of the advantage of self‐compatibility when colonizing new habitat. Despite its intuitive appeal and broad empirical support, it has received substantial criticism over the years since it was proclaimed in the 1950s, not least because it seemed to be contradicted by the high frequency of dioecy on islands. Recent theoretical work has again questioned the generality and scope of Baker's law. Here, we attempt to discern where the idea is useful to apply and where it is not. We conclude that several of the perceived problems with Baker's law fall away when a narrower perspective is adopted on how it should be circumscribed. We emphasize that Baker's law should be read in terms of an enrichment of a capacity for uniparental reproduction in colonizing situations, rather than of high selfing rates. We suggest that Baker's law might be tested in four different contexts, which set the breadth of its scope: the colonization of oceanic islands, metapopulation dynamics with recurrent colonization, range expansions with recurrent colonization, and colonization through species invasions.

No abstract

Self-compatibility and adaptations to self-fertilization are often found in plant populations at the periphery of species' ranges or on islands. Self-compatibility may predominate in these environments because it provides reproductive assurance when pollinators or availability of mates limits seed production. This possibility was studied in Leavenworthia alabamica, a flowering plant endemic to the southeastern United States. Populations at the center of the species' range retain sporophytic self-incompatibility, but peripheral populations are smaller, self-compatible, and have adaptations for self-fertilization. A reciprocal-transplant experiment was designed to test whether there is pollen limitation of seed set and to examine its strength in central and peripheral populations. Self-compatible genotypes produced more fruit and 17-22% more seed than self-incompatible genotypes in all environments, suggesting that the transition to self-compatibility may be favored by natural selection in all populations inhabited by L. alabamica. Sequence analyses demonstrated that two peripheral populations have 90-100% reductions in genetic variation, consistent with the effects of small population size or historical bottlenecks. Although pollen limitation of seed set occurs in all environments, self-compatibility may evolve at the periphery in L. alabamica because the benefits of reproductive assurance are influenced by population size or bottlenecks following extinction and colonization.

The evolution of selfing from outcrossing is a common transition, yet little is known about the mutations and selective factors that promote this shift. In the mustard family, single-locus self-incompatibility (SI) enforces outcrossing. In this study, we test whether mutations causing self-compatibility (SC) are linked to the self-incompatibility locus (S-locus) in Leavenworthia alabamica, a species where two selfing races (a2 and a4) co-occur with outcrossing populations. We also infer the ecological circumstances associated with origins of selfing using molecular sequence data. Genealogical reconstruction of the Lal2 locus, the putative ortholog of the SRK locus, showed that both selfing races are fixed for one of two different S-linked Lal2 sequences, whereas outcrossing populations harbor many S-alleles. Hybrid crosses demonstrated that S-linked mutations cause SC in each selfing race. These results strongly suggest two origins of selfing in this species, a result supported by population admixture analysis of 16 microsatellite loci and by a population tree built from eight nuclear loci. One selfing race (a4) shows signs of a severe population bottleneck, suggesting that reproductive assurance might have caused the evolution of selfing in this case. In contrast, the population size of race a2 cannot be distinguished from that of outcrossing populations after correcting for differences in selfing rates. Coalescent-based analyses suggest a relatively old origin of selfing in the a4 race (∼150 ka ago), whereas selfing evolved recently in the a2 race (∼12-48 ka ago). These results imply that S-locus mutations have triggered two recent shifts to selfing in L. alabamica, but that these transitions are not always associated with a severe population bottleneck, suggesting that factors other than reproductive assurance may play a role in its evolution.

Inbreeding depression is one of the leading factors preventing the evolution of self-fertilization in plants. In populations where self-fertilization evolves, theory suggests that natural selection against partially recessive deleterious alleles will reduce inbreeding depression. The purpose of this study was to evaluate this hypothesis by comparing the magnitude of inbreeding depression in self-incompatible and self-compatible populations of Leavenworthia alabamica. Within-population crosses were conducted to compare the quantity and quality of offspring produced by outcrossing and selffertilization. These progeny were grown in a common greenhouse and inbreeding depression was measured in germination, survival, biomass, transition rate to flowering, flower number, petal length, pollen grains/anther, pollen viability, and ovule number. In comparison to outcrossing, self-fertilization led to the production of fewer and smaller seeds within self-incompatible populations. Moreover, inbreeding depression was observed in eight of 11 offspring traits within self-incompatible populations of L. alabamica. In contrast, there was significant inbreeding depression only in flower number within self-compatible populations. The results of this study are consistent with the idea that selffertilization selectively removes partially recessive deleterious alleles causing inbreeding depression in natural plant populations. However, in plant species such as L. alabamica where self-compatibility may evolve in small populations following long-distance dispersal, declines in inbreeding depression may also be facilitated by genetic drift. Heredity (2005) 94, 159-165.

Because establishing a new population often depends critically on finding mates, individuals capable of uniparental reproduction may have a colonization advantage. Accordingly, there should be an over-representation of colonizing species in which individuals can reproduce without a mate, particularly in isolated locales such as oceanic islands. Despite the intuitive appeal of this colonization filter hypothesis (known as Baker's law), more than six decades of analyses have yielded mixed findings. We assembled a dataset of island and mainland plant breeding systems, focusing on the presence or absence of self-incompatibility. Because this trait enforces outcrossing and is unlikely to re-evolve on short timescales if it is lost, breeding system is especially likely to reflect the colonization filter. We found significantly more self-compatible species on islands than mainlands across a sample of > 1500 species from three widely distributed flowering plant families (Asteraceae, Brassicaceae and Solanaceae). Overall, 66% of island species were self-compatible, compared with 41% of mainland species. Our results demonstrate that the presence or absence of self-incompatibility has strong explanatory power for plant geographical patterns. Island floras around the world thus reflect the role of a key reproductive trait in filtering potential colonizing species in these three plant families.