How Does Selfing Affect the Pace and Process of Speciation?

Surprisingly little attention has been given to the impact of selfing on speciation, even though selfing reduces gene flow between populations and affects other key population genetics parameters. Here we review recent theoretical work and compile empirical data from crossing experiments and genomic and phylogenetic studies to assess the effect of mating systems on the speciation process. In accordance with theoretical predictions, we find that accumulation of hybrid incompatibilities seems to be accelerated in selfers, but there is so far limited empirical support for a predicted bias toward underdominant loci. Phylogenetic evidence is scarce and contradictory, including studies suggesting that selfing either promotes or hampers speciation rate. Further studies are therefore required, which in addition to measures of reproductive barrier strength and selfing rate should routinely include estimates of demographic history and genetic divergence as a proxy for divergence time.

RIDGE, a tool tailored to detect gene flow barriers across species pairs.

Characterizing the processes underlying reproductive isolation between diverging lineages is central to understanding speciation. Here, we present RIDGE-Reproductive Isolation Detection using Genomic polymorphisms-a tool tailored for quantifying gene flow barrier proportion and identifying the relevant genomic regions. RIDGE relies on an Approximate Bayesian Computation with a model-averaging approach to accommodate diverse scenarios of lineage divergence. It captures heterogeneity in effective migration rate along the genome while accounting for variation in linked selection and recombination. The barrier detection test relies on numerous summary statistics to compute a Bayes factor, offering a robust statistical framework that facilitates cross-species comparisons. Simulations revealed RIDGE's efficiency in capturing signals of ongoing migration. Model averaging proved particularly valuable in scenarios of high model uncertainty where no migration or migration homogeneity can be wrongly assumed, typically for recent divergence times <0.1 2Ne generations. Applying RIDGE to four published crow data sets, we first validated our tool by identifying a well-known large genomic region associated with mate choice patterns. Second, while we identified a significant overlap of outlier loci using RIDGE and traditional genomic scans, our results suggest that a substantial portion of previously identified outliers are likely false positives. Outlier detection relies on allele differentiation, relative measures of divergence and the count of shared polymorphisms and fixed differences. Our analyses also highlight the value of incorporating multiple summary statistics including our newly developed outlier ones that can be useful in challenging detection conditions.

Separating phases of allopolyploid evolution with resynthesized and natural Capsella bursa-pastoris.

Allopolyploidization is a frequent evolutionary transition in plants that combines whole-genome duplication (WGD) and interspecific hybridization. The genome of an allopolyploid species results from initial interactions between parental genomes and long-term evolution. Distinguishing the contributions of these two phases is essential to understanding the evolutionary trajectory of allopolyploid species. Here, we compared phenotypic and transcriptomic changes in natural and resynthesized Capsella allotetraploids with their diploid parental species. We focused on phenotypic traits associated with the selfing syndrome and on transcription-level phenomena such as expression-level dominance (ELD), transgressive expression (TRE), and homoeolog expression bias (HEB). We found that selfing syndrome, high pollen, and seed quality in natural allotetraploids likely resulted from long-term evolution. Similarly, TRE and most down-regulated ELD were only found in natural allopolyploids. Natural allotetraploids also had more ELD toward the self-fertilizing parental species than resynthesized allotetraploids, mirroring the establishment of the selfing syndrome. However, short-term changes mattered, and 40% of the cases of ELD in natural allotetraploids were already observed in resynthesized allotetraploids. Resynthesized allotetraploids showed striking variation of HEB among chromosomes and individuals. Homoeologous synapsis was its primary source and may still be a source of genetic variation in natural allotetraploids. In conclusion, both short- and long-term mechanisms contributed to transcriptomic and phenotypic changes in natural allotetraploids. However, the initial gene expression changes were largely reshaped during long-term evolution leading to further morphological changes.

Cosegregation of recombinant chromatids maintains genome-wide heterozygosity in an asexual nematode.

In asexual animals, female meiosis is modified to produce diploid oocytes. If meiosis still involves recombination, this is expected to lead to a rapid loss of heterozygosity, with adverse effects on fitness. Many asexuals, however, have a heterozygous genome, the underlying mechanisms being most often unknown. Cytological and population genomic analyses in the nematode Mesorhabditis belari revealed another case of recombining asexual being highly heterozygous genome-wide. We demonstrated that heterozygosity is maintained despite recombination because the recombinant chromatids of each chromosome pair cosegregate during the unique meiotic division. A theoretical model confirmed that this segregation bias is necessary to account for the observed pattern and likely to evolve under a wide range of conditions. Our study uncovers an unexpected type of non-Mendelian genetic inheritance involving cosegregation of recombinant chromatids.

Joint inference of evolutionary transitions to self-fertilization and demographic history using whole-genome sequences.

The evolution from outcrossing to selfing occurred recently across the eukaryote tree of life in plants, animals, fungi, and algae. Despite short-term advantages, selfing is hypothetically an evolutionary dead-end reproductive strategy. The tippy distribution on phylogenies suggests that most selfing species are of recent origin. However, dating such transitions is challenging yet central for testing this hypothesis. We build on previous theories to disentangle the differential effect of past changes in selfing rate or from that of population size on recombination probability along the genome. This allowed us to develop two methods using full-genome polymorphisms to (1) test if a transition from outcrossing to selfing occurred and (2) infer its age. The teSMC and tsABC methods use a transition matrix summarizing the distribution of times to the most recent common ancestor along the genome to estimate changes in the ratio of population recombination and mutation rates overtime. First, we demonstrate that our methods distinguish between past changes in selfing rate and demographic history. Second, we assess the accuracy of our methods to infer transitions to selfing approximately up to 2.5Ne generations ago. Third, we demonstrate that our estimates are robust to the presence of purifying selection. Finally, as a proof of principle, we apply both methods to three Arabidopsis thaliana populations, revealing a transition to selfing approximately 600,000 years ago. Our methods pave the way for studying recent transitions to self-fertilization and better accounting for variation in mating systems in demographic inferences.

Opposing effects of plant traits on diversification.

Species diversity can vary dramatically across lineages due to differences in speciation and extinction rates. Here, we explore the effects of several plant traits on diversification, finding that most traits have opposing effects on diversification. For example, outcrossing may increase the efficacy of selection and adaptation but also decrease mate availability, two processes with contrasting effects on lineage persistence. Such opposing trait effects can manifest as differences in diversification rates that depend on ecological context, spatiotemporal scale, and associations with other traits. The complexity of pathways linking traits to diversification suggests that the mechanistic underpinnings behind their correlations may be difficult to interpret with any certainty, and context dependence means that the effects of specific traits on diversification are likely to differ across multiple lineages and timescales. This calls for taxonomically and context-controlled approaches to studies that correlate traits and diversification.

Trait-dependent diversification in angiosperms: Patterns, models and data.

Variation in species richness across the tree of life, accompanied by the incredible variety of ecological and morphological characteristics found in nature, has inspired many studies to link traits with species diversification. Angiosperms are a highly diverse group that has fundamentally shaped life on earth since the Cretaceous, and illustrate how species diversification affects ecosystem functioning. Numerous traits and processes have been linked to differences in species richness within this group, but we know little about their relative importance and how they interact. Here, we synthesised data from 152 studies that used state-dependent speciation and extinction (SSE) models on angiosperm clades. Intrinsic traits related to reproduction and morphology were often linked to diversification but a set of universal drivers did not emerge as traits did not have consistent effects across clades. Importantly, SSE model results were correlated to data set properties - trees that were larger, older or less well-sampled tended to yield trait-dependent outcomes. We compared these properties to recommendations for SSE model use and provide a set of best practices to follow when designing studies and reporting results. Finally, we argue that SSE model inferences should be considered in a larger context incorporating species' ecology, demography and genetics.

Asexual male production by ZW recombination in Artemia parthenogenetica.

In some asexual species, parthenogenetic females occasionally produce males, which may strongly affect the evolution and maintenance of asexuality if they cross with related sexuals and transmit genes causing asexuality to their offspring ("contagious parthenogenesis"). How these males arise in the first place has remained enigmatic, especially in species with sex chromosomes. Here, we test the hypothesis that rare, asexually produced males of the crustacean Artemia parthenogenetica are produced by recombination between the Z and W sex chromosomes during non-clonal parthenogenesis, resulting in ZZ males through loss of heterozygosity at the sex determination locus. We used RAD-sequencing to compare asexual mothers with their male and female offspring. Markers on several sex-chromosome scaffolds indeed lost heterozygosity in all male but no female offspring, suggesting that they correspond to the sex-determining region. Other sex-chromosome scaffolds lost heterozygosity in only a part of the male offspring, consistent with recombination occurring at a variable location. Alternative hypotheses for the production of these males (such as partial or total hemizygosity of the Z) could be excluded. Rare males are thus produced because recombination is not entirely suppressed during parthenogenesis in A. parthenogenetica. This finding may contribute to explaining the maintenance of recombination in these asexuals.

Expression pattern of resynthesized allotetraploid Capsella is determined by hybridization, not whole-genome duplication.

Polyploidization, the process leading to the increase in chromosome sets, is a major evolutionary transition in plants. Whole-genome duplication (WGD) within the same species gives rise to autopolyploids, whereas allopolyploids result from a compound process with two distinct components: WGD and interspecific hybridization. To dissect the instant effects of WGD and hybridization on gene expression and phenotype, we created a series of synthetic hybrid and polyploid Capsella plants, including diploid hybrids, autotetraploids of both parental species, and two kinds of resynthesized allotetraploids with different orders of WGD and hybridization. Hybridization played a major role in shaping the relative expression pattern of the neo-allopolyploids, whereas WGD had almost no immediate effect on relative gene expression pattern but, nonetheless, still affected phenotypes. No transposable element-mediated genomic shock scenario was observed in either neo-hybrids or neo-polyploids. Finally, WGD and hybridization interacted and the distorting effects of WGD were less strong in hybrids. Whole-genome duplication may even improve hybrid fertility. In summary, while the initial relative gene expression pattern in neo-allotetraploids was almost entirely determined by hybridization, WGD only had trivial effects on relative expression patterns, both processes interacted and had a strong impact on physical attributes and meiotic behaviors.

Mating system and speciation I: Accumulation of genetic incompatibilities in allopatry.

Self-fertilisation is widespread among hermaphroditic species across the tree of life. Selfing has many consequences on the genetic diversity and the evolutionary dynamics of populations, which may in turn affect macroevolutionary processes such as speciation. On the one hand, because selfing increases genetic drift and reduces migration rate among populations, it may be expected to promote speciation. On the other hand, because selfing reduces the efficacy of selection, it may be expected to hamper ecological speciation. To better understand under which conditions and in which direction selfing affects the build-up of reproductive isolation, an explicit population genetics model is required. Here, we focus on the interplay between genetic drift, selection and genetic linkage by studying speciation without gene flow. We test how fast populations with different rates of selfing accumulate mutations leading to genetic incompatibilities. When speciation requires populations to pass through a fitness valley caused by underdominant and compensatory mutations, selfing reduces the depth and/or breadth of the valley, and thus overall facilitates the fixation of incompatibilities. When speciation does not require populations to pass through a fitness valley, as for Bateson-Dobzhanzky-Muller incompatibilities (BDMi), the lower effective population size and higher genetic linkage in selfing populations both facilitate the fixation of incompatibilities. Interestingly, and contrary to intuitive expectations, local adaptation does not always accelerate the fixation of incompatibilities in outcrossing relative to selfing populations. Our work helps to clarify how incompatibilities accumulate in selfing vs. outcrossing lineages, and has repercussions on the pace of speciation as well as on the genetic architecture of reproductive isolation.

How broad is the selfing syndrome? Insights from convergent evolution of gene expression across species and tissues in the Capsella genus.

The shift from outcrossing to selfing is one of the main evolutionary transitions in plants. It is accompanied by profound effects on reproductive traits, the so-called selfing syndrome. Because the transition to selfing also implies deep genomic and ecological changes, one also expects to observe a genomic selfing syndrome. We took advantage of the three independent transitions from outcrossing to selfing in the Capsella genus to characterize the overall impact of mating system change on RNA expression, in flowers but also in leaves and roots. We quantified the extent of both selfing and genomic syndromes, and tested whether changes in expression corresponded to adaptation to selfing or to relaxed selection on traits that were constrained in outcrossers. Mating system change affected gene expression in all three tissues but more so in flowers than in roots and leaves. Gene expression in selfing species tended to converge in flowers but diverged in the two other tissues. Hence, convergent adaptation to selfing dominates in flowers, whereas genetic drift plays a more important role in leaves and roots. The effect of mating system transition is not limited to reproductive tissues and corresponds to both adaptation to selfing and relaxed selection on previously constrained traits.

Diversity and determinants of recombination landscapes in flowering plants.

During meiosis, crossover rates are not randomly distributed along the chromosome and their location may have a strong impact on the functioning and evolution of the genome. To date, the broad diversity of recombination landscapes among plants has rarely been investigated and a formal comparative genomic approach is still needed to characterize and assess the determinants of recombination landscapes among species and chromosomes. We gathered genetic maps and genomes for 57 flowering plant species, corresponding to 665 chromosomes, for which we estimated large-scale recombination landscapes. We found that the number of crossover per chromosome spans a limited range (between one to five/six) whatever the genome size, and that there is no single relationship across species between genetic map length and chromosome size. Instead, we found a general relationship between the relative size of chromosomes and recombination rate, while the absolute length constrains the basal recombination rate for each species. At the chromosome level, we identified two main patterns (with a few exceptions) and we proposed a conceptual model explaining the broad-scale distribution of crossovers where both telomeres and centromeres play a role. These patterns correspond globally to the underlying gene distribution, which affects how efficiently genes are shuffled at meiosis. These results raised new questions not only on the evolution of recombination rates but also on their distribution along chromosomes.

The relative role of plasticity and demographic history in Capsella bursa-pastoris: a common garden experiment in Asia and Europe.

The colonization success of a species depends on the interplay between its phenotypic plasticity, adaptive potential and demographic history. Assessing their relative contributions during the different phases of a species range expansion is challenging, and requires large-scale experiments. Here, we investigated the relative contributions of plasticity, performance and demographic history to the worldwide expansion of the shepherd's purse, Capsella bursa-pastoris. We installed two large common gardens of the shepherd's purse, a young, self-fertilizing, allopolyploid weed with a worldwide distribution. One common garden was located in Europe, the other in Asia. We used accessions from three distinct genetic clusters (Middle East, Europe and Asia) that reflect the demographic history of the species. Several life-history traits were measured. To explain the phenotypic variation between and within genetic clusters, we analysed the effects of (i) the genetic clusters, (ii) the phenotypic plasticity and its association to fitness and (iii) the distance in terms of bioclimatic variables between the sampling site of an accession and the common garden, i.e. the environmental distance. Our experiment showed that (i) the performance of C. bursa-pastoris is closely related to its high phenotypic plasticity; (ii) within a common garden, genetic cluster was a main determinant of phenotypic differences; and (iii) at the scale of the experiment, the effect of environmental distance to the common garden could not be distinguished from that of genetic clusters. Phenotypic plasticity and demographic history both play important role at different stages of range expansion. The success of the worldwide expansion of C. bursa-pastoris was undoubtedly influenced by its strong phenotypic plasticity.

Competitive ability depends on mating system and ploidy level across Capsella species.

BACKGROUND AND AIMS: Self-fertilization is often associated with ecological traits corresponding to the ruderal strategy, and selfers are expected to be less competitive than outcrossers, either because of a colonization/competition trade-off or because of the deleterious genetic effects of selfing. Range expansion could reduce further competitive ability while polyploidy could mitigate the effects of selfing. If pollinators are not limited, individual fitness is thus expected to be higher in outcrossers than in selfers and, within selfers, in polyploids than in diploids. Although often proposed in the botanical literature and also suggested by meta-analyses, these predictions have not been directly tested yet. METHODS: In order to compare fitness and the competitive ability of four Capsella species with a different mating system and ploidy level, we combined two complementary experiments. First, we carried out an experiment outdoors in north-west Greece, i.e. within the range of the obligate outcrossing species, C. grandiflora, where several life history traits were measured under two different disturbance treatments, weeded plots vs. unweeded plots. To better control competition and to remove potential effects of local adaptation of the outcrosser, we also performed a similar competition experiment but under growth chamber conditions. KEY RESULTS: In the outdoor experiment, disturbance of the environment did not affect the phenotype in any of the four species. For most traits, the obligate outcrossing species performed better than all selfing species. In contrast, polyploids did not survive or reproduce better than diploids. Under controlled conditions, as in the field experiment, the outcrosser had a higher fitness than selfing species and was less affected by competition. Finally, contrary to the outdoor experiment where the two behaved identically, polyploid selfers were less affected by competition than diploid selfes. CONCLUSIONS: In the Capsella genus, selfing induces lower fitness than outcrossing and can also reduce competitive ability. The effect of polyploidy is, however, unclear. These results highlight the possible roles of ecological context in the evolution of selfing species.

What does the distribution of fitness effects of new mutations reflect? Insights from plants.

The distribution of fitness effects (DFE) of new mutations plays a central role in molecular evolution. It is therefore crucial to be able to estimate it accurately from genomic data and to understand the factors that shape it. After a rapid overview of available methods to characterize the fitness effects of mutations, we review what is known on the factors affecting them in plants. Available data indicate that life history traits (e.g. mating system and longevity) have a major effect on the DFE. By contrast, the impact of demography within species appears to be more limited. These results remain to be confirmed, and methods to estimate the joint evolution of demography, life history traits, and the DFE need to be developed.

Hunting for Beneficial Mutations: Conditioning on SIFT Scores When Estimating the Distribution of Fitness Effect of New Mutations.

The distribution of fitness effects (DFE) of new mutations is a key parameter of molecular evolution. The DFE can in principle be estimated by comparing the site frequency spectra (SFS) of putatively neutral and functional polymorphisms. Unfortunately, the DFE is intrinsically hard to estimate, especially for beneficial mutations because these tend to be exceedingly rare. There is therefore a strong incentive to find out whether conditioning on properties of mutations that are independent of the SFS could provide additional information. In the present study, we developed a new measure based on SIFT scores. SIFT scores are assigned to nucleotide sites based on their level of conservation across a multispecies alignment: the more conserved a site, the more likely mutations occurring at this site are deleterious, and the lower the SIFT score. If one knows the ancestral state at a given site, one can assign a value to new mutations occurring at the site based on the change of SIFT score associated with the mutation. We called this new measure δ. We show that properties of the DFE as well as the flux of beneficial mutations across classes covary with δ and, hence, that SIFT scores are informative when estimating the fitness effect of new mutations. In particular, conditioning on SIFT scores can help to characterize beneficial mutations.

Pulled Diversification Rates, Lineages-Through-Time Plots, and Modern Macroevolutionary Modeling.

Estimating time-dependent rates of speciation and extinction from dated phylogenetic trees of extant species (timetrees), and determining how and why they vary, is key to understanding how ecological and evolutionary processes shape biodiversity. Due to an increasing availability of phylogenetic trees, a growing number of process-based methods relying on the birth-death model have been developed in the last decade to address a variety of questions in macroevolution. However, this methodological progress has regularly been criticized such that one may wonder how reliable the estimations of speciation and extinction rates are. In particular, using lineages-through-time (LTT) plots, a recent study has shown that there are an infinite number of equally likely diversification scenarios that can generate any timetree. This has led to questioning whether or not diversification rates should be estimated at all. Here, we summarize, clarify, and highlight technical considerations on recent findings regarding the capacity of models to disentangle diversification histories. Using simulations, we illustrate the characteristics of newly proposed "pulled rates" and their utility. We recognize that the recent findings are a step forward in understanding the behavior of macroevolutionary modeling, but they in no way suggest we should abandon diversification modeling altogether. On the contrary, the study of macroevolution using phylogenetic trees has never been more exciting and promising than today. We still face important limitations in regard to data availability and methods, but by acknowledging them we can better target our joint efforts as a scientific community. [Birth-death models; extinction; phylogenetics; speciation.].

Establishment of local adaptation in partly self-fertilizing populations.

Populations often inhabit multiple ecological patches and thus experience divergent selection, which can lead to local adaptation if migration is not strong enough to swamp locally adapted alleles. Conditions for the establishment of a locally advantageous allele have been studied in randomly mating populations. However, many species reproduce, at least partially, through self-fertilization, and how selfing affects local adaptation remains unclear and debated. Using a two-patch branching process formalism, we obtained a closed-form approximation under weak selection for the probability of establishment of a locally advantageous allele (P) for arbitrary selfing rate and dominance level, where selection is allowed to act on viability or fecundity, and migration can occur via seed or pollen dispersal. This solution is compared to diffusion approximation and used to investigate the consequences of a shift in a mating system on P, and the establishment of protected polymorphism. We find that selfing can either increase or decrease P, depending on the patterns of dominance in the two patches, and has conflicting effects on local adaptation. Globally, selfing favors local adaptation when locally advantageous alleles are (partially) recessive, when selection between patches is asymmetrical and when migration occurs through pollen rather than seed dispersal. These results establish a rigorous theoretical background to study heterogeneous selection and local adaptation in partially selfing species.

The Road to Sorghum Domestication: Evidence From Nucleotide Diversity and Gene Expression Patterns.

Native African cereals (sorghum, millets) ensure food security to millions of low-income people from low fertility and drought-prone regions of Africa and Asia. In spite of their agronomic importance, the genetic bases of their phenotype and adaptations are still not well-understood. Here we focus on Sorghum bicolor, which is the fifth cereal worldwide for grain production and constitutes the staple food for around 500 million people. We leverage transcriptomic resources to address the adaptive consequences of the domestication process. Gene expression and nucleotide variability were analyzed in 11 domesticated and nine wild accessions. We documented a downregulation of expression and a reduction of diversity both in nucleotide polymorphism (30%) and gene expression levels (18%) in domesticated sorghum. These findings at the genome-wide level support the occurrence of a global reduction of diversity during the domestication process, although several genes also showed patterns consistent with the action of selection. Nine hundred and forty-nine genes were significantly differentially expressed between wild and domesticated gene pools. Their functional annotation points to metabolic pathways most likely contributing to the sorghum domestication syndrome, such as photosynthesis and auxin metabolism. Coexpression network analyzes revealed 21 clusters of genes sharing similar expression patterns. Four clusters (totaling 2,449 genes) were significantly enriched in differentially expressed genes between the wild and domesticated pools and two were also enriched in domestication and improvement genes previously identified in sorghum. These findings reinforce the evidence that the combined and intricated effects of the domestication and improvement processes do not only affect the behaviors of a few genes but led to a large rewiring of the transcriptome. Overall, these analyzes pave the way toward the identification of key domestication genes valuable for genetic resources characterization and breeding purposes.