2015/11/26 by Clément Lafon Placette, Mario Vallejo‐Marín, Christian Parisod +2 · 1 citation
Agricultural and Biological Sciences · Biochemistry, Genetics and Molecular Biology · #Plant and animal studies #Plant Reproductive Biology #Genetic diversity and population structure #Genetic algorithm #Reproductive isolation #Biology #Ecological speciation #Isolation (microbiology) #Evolutionary biology #Ecology #Genetics #Bioinformatics
paper · pdf · doi:10.1111/nph.13756
openalex publication_date 2015/11/26 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/02
Explaining what species are and how they arise has been at the center of biological research since the first evolutionary concepts were developed by Buffon, Lamarck, and Darwin (Tirard, 2010). With the widespread acceptance of the biological species concept, according to which ‘species are a group of interbreeding natural populations that are reproductively isolated from other such groups’ (Mayr, 1996), much recent speciation research has focused on the processes and mechanisms involved in the evolution of reproductive isolation barriers (Coyne & Orr, 2004). The molecular basis of these isolating barriers was known in only a few instances at the time of the Plant Speciation New Phytologist Symposium held in 2003 (Rieseberg & Wendel, 2004). However, the last decade has witnessed major advances to current knowledge on this and other aspects of plant speciation research as was made clear at the recent European Molecular Biology Organization (EMBO) workshop ‘Mechanisms of plant speciation’ held in 2015 at Åkersberga, Sweden. Many of the presentations at the EMBO workshop demonstrated the game-changing effects that genome sequencing and genetic analysis, together with the emergence of new promising study systems (Zuellig et al., 2014), are having on our understanding of the genetics of barriers to gene flow between species (reviewed in Baack et al., 2015). For instance, the genetic architectures of traits such as flowering time and flower color (contributing to prepollination barriers) have been resolved in several plant species (Sheehan et al., 2012) and in some cases shown to be involved in the reinforcement of barriers to prevent production of low fitness hybrids (Hopkins, 2013). Similarly, evidence is accumulating on the genetics of Dobzhansky–Muller incompatibilities (contributing to postzygotic barriers), of loci underlying adaptation to environmental heterogeneity with or without chromosomal rearrangement (Ågren et al., 2013; Oneal et al., 2014), and that tight linkage of multiple functional genes can underlie speciation, as illustrated by the supergene architecture of pollination syndromes in Petunia species (Hermann et al., 2013). In addition to gathering support for well-known speciation mechanisms, the presentations at the EMBO workshop emphasized the potential role of other mechanisms, such as epigenetics, in driving speciation. Later, we highlight some of the research that was presented at the workshop and consider its effects on existing paradigms of plant speciation. The role of sexual selection as a selective force in plant speciation is still controversial, in contrast to the situation in animals (Moore & Pannell, 2011). The processes involved in the two major components of sexual selection, that is, intrasexual competition and female choice for males, were poorly understood in plants until the last decade. This was the case even though pollen was recognized to be often far more abundant than the ovules produced by a species, thus creating the conditions (i.e. a strongly biased sex ratio towards males) for intrasexual competition of pollen for ovules. In addition, there is species-dependent variation in ovule number per flower (Burd et al., 2009) and consequently, this variation could correlate with variation in sexual selection. As in animals, a high level of sexual selection in plants should speed up the evolution of traits influencing mating success, and often as a consequence, the evolution of traits responsible for pre- and post-mating hybridization barriers between species. Loren Rieseberg (University of British Columbia, Vancouver, BC, Canada) presented results on ovule number and F1 hybrid seed set in several taxa, and showed that a lower number of ovules (higher sexual selection) correlated with higher cross incompatibility between related species (D. Ortiz-Barientos, D. Ebert, R. Andrew, C. Fitzpatrick, R. Godfrey, K. Ostevik & L. H. Rieseberg, unpublished data). These findings, therefore, strongly support the hypothesis that sexual selection is tightly linked to interspecific barriers in plants and likely to be of high importance in plant speciation. Rieseberg further emphasized that sexual selection might play a role in reinforcement during sympatric speciation. Another aspect of sexual selection concerns female choice for males. Recent research has shown that ovules attract pollen by producing and secreting short peptides, and that this attraction can be selective (Takeuchi & Higashiyama, 2012). Under the concept of female choice, it is expected that species subject to different levels of sexual selection such as self-incompatible (higher sexual selection) and self-compatible (lower sexual selection) species should differ in the selectiveness of their ovules towards male gametes (pollen). Consistent with this hypothesis, Leonie Moyle (Indiana University, USA) presented work showing that ovules from self-incompatible Solanum species were more selective, attracting less heterospecific pollen, compared to ovules from self-compatible Solanum species. When crossing self-compatible and self-incompatible Solanum species, this resulted in unilateral fertilization. These data support the idea that sexual selection drives the establishment of interspecific reproductive barriers, which can lead to unidirectional gene flow between species. Pollen competition not only occurs to increase mating success but also to increase access to maternal resources, as explained by the parental conflict theory (Haig, 2004). The endosperm is the battleground for this conflict of parental interests, enforcing parent-of-origin specific gene expression (genomic imprinting, see later). Such parental conflict is expected to be greater in species experiencing stronger sexual selection, for example, in self-incompatible species vs self-compatible ones. As a consequence, in a cross between a self-incompatible and self-compatible species, the former is expected to ‘overpower’ the latter, leading to direction-dependent endosperm defects in the hybrid seed, as formulated in the weak inbreeder/strong outbreeder hypothesis (Brandvain & Haig, 2005). Claudia Köhler and colleagues (Swedish University of Agricultural Sciences, Uppsala, Sweden) tested this hypothesis by investigating the presence of nonreciprocal postzygotic barriers affecting hybrid seed survival between the outbreeder Capsella grandiflora and the inbreeder C. rubella (Rebernig et al., 2015). They reported that high hybrid seed lethality was cross direction-dependent, due to endosperm defects, and mimicked the phenotypic and transcriptomic responses of interploidy hybrid seeds, a case where the parent of higher ploidy ‘overpowers’ the one of lower ploidy. In accordance with this view, Ana Marcela Florez Rueda (ETH Zurich, Switzerland) reported that cross direction-dependent defects of hybrid seeds between Solanum species were related to the disruption of imprinting in a majority of genes. To what extent genomic imprinting and other types of genomic conflicts are involved in hybridization barriers between inbreeders and outbreeders is an exciting question that remains to be addressed. Intragenomic conflicts, whereby DNA sequences with divergent fitness interests show antagonistic interactions, likely also contribute to the build-up of reproductive isolation. Christian Parisod (University of Neuchatel, Switzerland) illustrated how selfish transposons having diverged between species show massive, asymmetrical genetic and epigenetic changes when merged in hybrids of wild wheats (Senerchia et al., 2015). Repeated DNA loss and epigenetic changes around interspersed transposons in F1 and later generation hybrids indicate that early silencing of conflicting loci is likely necessary to circumvent transposon activation that would otherwise result in hybrid seed defects through mutation and/or epigenetic repatterning affecting nearby genes (Lafon-Placette & Köhler, 2015). As discussed by Luca Comai (University of California, Davis, CA, USA), divergence in CENH3 histones between different species is associated with chromosome elimination in hybrids (Tan et al., 2015), offering additional support for incompatibilities related to epigenetics. To what extent the cascade of genetic and epigenetic changes related to early-acting genomic conflicts contribute to the earlier-discussed allele-specific expression patterns in interspecies/interploidy hybrid seeds remains an open question, but it appears that they may represent overlooked actors driving speciation. While one of the earliest described cases of hybrid necrosis is between two Crepis species (Hollingshead, 1930), the phenomenon of hybrid necrosis and its role in plant speciation had received very little attention until recently. A major advance on this topic occurred when the phenomenon was described in Arabidopsis thaliana and the genetic basis underlying this Dobzhansky–Muller-type incompatibility was revealed (Bomblies et al., 2007). Hybrid necrosis resembles an autoimmune response and is triggered by the deleterious epistasis between naturally occurring alleles at two unlinked loci, one of them being a NB-LRR disease resistance gene. Detlef Weigel (Max Planck Institute for Developmental Biology, Tübingen, Germany) presented impressive work revealing that one cluster of NLR disease resistance genes was responsible for more than half of the nearly 150 inter-accession hybrid incompatibility cases observed (Chae et al., 2014). This disease resistance gene cluster is highly variable and shows signs of diversifying selection, leading the authors to propose that hybrid incompatibility observed between Arabidopsis accessions could be shaped by the selective pressure related to host–pathogen coevolution. As Weigel pointed out, deleterious interactions of disease resistance genes could limit the possible combinations of favorable allele combinations present in plant genomes. A similar hybrid incompatibility, involving the interaction between two loci containing disease resistance genes, was reported by Michael Lenhard (University of Potsdam, Germany) to occur between the sister species C. rubella and C. grandiflora (Sicard et al., 2015). Interestingly, both compatible and incompatible alleles for one of the loci are maintained at high frequency by balancing selection in C. grandiflora. Similarly, data presented by Daniel Koenig (Max Planck Institute for Developmental Biology) showed that in the inbreeding species C. rubella and C. orientalis, which have extremely low genetic diversity, ancestral polymorphisms of disease resistance genes have been maintained by means of balancing selection (D. Koenig, S. I. Wright, B. Neuffer, T. Slotte, R. Li, J. Hagmann & D. Weigel, unpublished data). Balancing or diversifying selection on disease resistance genes as a result of the selective pressure related to the host–pathogen conflict could drive the evolution of reproductive barriers through lineage sorting of alternative alleles. Nevertheless, if maintenance of the ancestral polymorphism is important to cope with pathogen selection pressure, the question of how a lineage with fixed alleles responds to this pressure remains to be addressed. An extension of this research to species outside the Brassicaceae family should offer exciting insights of substantial impact. Polyploidization through whole genome duplication is a major feature of the evolutionary history of flowering plants (Soltis et al., 2014). It may result from genome duplication within a species (autopolyploidy) or of a hybrid formed between species (allopolyploidy), usually as a consequence of unreduced gamete formation (i.e. gametes with the same chromosome number as somatic cells). Both genetic and environmental effects can influence unreduced gamete formation, and Nico de Storme (Ghent University, Belgium) pointed out that during plant evolution there may have been a causal connection between marked changes in temperature and polyploidization events. He reported that in A. thaliana acute exposure to either high or low temperatures increased formation of unreduced pollen grains (De Storme & Geelen, 2014). Genetic and environmental factors also affect the stability of meiosis in polyploids and their diploid progenitors (Bomblies et al., 2015). Using genome scans, Yant et al. (2013) have identified multiple genes involved in the control of crossing over in autotetraploid Arabidopsis arenosa that were likely targets of selection for the stabilization of meiosis following whole genome duplication. Interestingly, the appropriate alleles for some of these genes occur at low frequency in diploid populations of the same species and, consequently, an absence of such standing variation in diploid ancestral populations might be a factor preventing autopolyploids establishing in some species. From further research, Kirsten Bomblies (John Innes Centre, Norwich, UK) reported differences in the stability of synapsis among diploid strains of A. arenosa from different habitats with two core meiotic proteins (axis/cohesion proteins) showing evidence of having undergone selection. She and her colleagues found that different alleles for the same proteins are subject to selection after polyploidy. Their current hypothesis is that both polyploidy and environmental changes (mainly temperature) affect core meiotic functions, causing alleles that favor a stable meiosis to be selected. In some plant groups polyploidy is associated with asexual reproduction through apomixis, which removes the need for a stable meiosis. However, a complicating factor can be pseudogamy, that is, the requirement of pollination for seed set to occur. Simon Hiscock (University of Oxford, UK) and colleagues found this to be the case in Sorbus (whitebeams). Working with material from the Avon Gorge in the UK, a major hotspot of Sorbus diversity, they established that triploid taxa are facultatively apomictic and pseudogamous, and surprisingly also self-incompatible. Seeds are produced only after pollination by other Sorbus taxa, and because apomixis is facultative this can lead to the generation of new hybrids and further genetic diversification. Studies of very recently formed polyploid species provide insights into the early stages of speciation. Mario Vallejo-Marín (University of Stirling, UK) reported on the young allopolyploid species Mimulus peregrinus which evolved from the triploid sterile hybrid between diploid (M. guttatus) and tetraploid (M. luteus) taxa that were introduced into the British Isles in the nineteenth century (Vallejo-Marín et al., 2015). Despite its recent origin, M. peregrinus has formed at least twice, and its genome seems to be undergoing changes including the potential loss of small genomic regions (Vallejo-Marín et al., 2015), as reported in other recent polyploid systems (Hegarty et al., 2006; Buggs et al., 2012). The discovery of a young allopolyploid species in Mimulus is particularly exciting, as this genus combines a long history of evolutionary and speciation research, with rapidly developing genomic tools, including the whole genome sequence of the parental taxon M. guttatus (Wu et al., 2007). Mimulus thus adds to the short list of powerful systems for the study of polyploid speciation. A major question in polyploid speciation is the extent to which genome duplication facilitates the evolution of ecological and phenotypic novelty (Abbott et al., 2013). Pamela Soltis (University of Florida, Gainesville, FL, USA) discussed the role of allopolyploidization in generating such novelty in the recently formed allotetraploid Tragopogon (goatsbeard) species, T. mirus and T. miscellus. These neopolyploids show extensive evidence of nonadditive gene expression and genome restructuring (Buggs et al., 2012; Chester et al., 2012) as well as additive and transgressive phenotypic variation relative to their diploid parents. The extent to which novel phenotypes enable allotetraploids to colonize new niches is uncertain, but Soltis and colleagues used ecological niche modeling to report that the niche of Tragopogon polyploids does not entirely overlap those of their parents. Brian Husband (Guelph University, ON, Canada) further showed from a large-scale analysis of diploid, natural tetraploids, and artificially synthesized neotetraploid cytotypes of A. thaliana that the effect of genome duplication on phenotype, including fertility, is dependent on genetic background (i.e. the particular 2x accession used to generate the neotetraploid), which complicates the assessment of effects of polyploidy on phenotypic divergence. The consequences of polyploidization at the macroevolutionary level in plants remains controversial and a better appraisal of speciation vs extinction in polyploids depends on the availability of accurate phylogenetic information and large-scale data bases of phenotypic and ecological information. Itay Mayrose (University of Tel Aviv, Israel) reported on his group's efforts to create a community resource that combines chromosome counts and phylogenetic information across plants (Rice et al., 2015). This ongoing effort will be of great value for elucidating the effect of polyploidization on diversification. The last decade has not seen dramatic changes in paradigms of plant speciation, although it is becoming clear that the role of hybridization between genetically differentiated taxa as a promoter rather than an inhibitor of reproductive isolation deserves further attention (Abbott et al., 2013). Of particular importance is that recent technological advances are allowing identification of the genetic targets of selection, natural or sexual, that drive the establishment of reproductive barriers between species. It is now realized that the architecture of speciation loci may be more variable than previously anticipated, as exemplified by cases of interspersed or supergene-like incompatible loci (Fig. 1). It is also apparent that higher level and more complex speciation patterns occur, with similar drivers and genetic mechanisms involved across taxa, such as hybrid necrosis. In addition, it is evident that an understanding of epigenetic processes provides a mechanistic basis for old concepts such as genome shock, parental conflict, or chromosome deletion. Although the origin of reproductive barriers may often remain elusive, the ongoing sequencing revolution will likely foster the accumulation of new insights at a fast pace. Thus, we can envision that the synthesis of complementary studies emphasizing processes underlying early speciation with the comparative genomics of ‘good species’ (i.e. the magnifying glass vs spyglass approaches of Via, 2009) will generate a deeper understanding of the biology of plant speciation over the coming years. The authors thank all participants of the workshop for valuable discussions and contributions. The authors apologize for studies not discussed here in detail due to space constraints. The workshop was financially supported by EMBO, the Swedish Research Councils (VR and Formas), New Phytologist Trust, the Scandinavian Society of Plant Physiology, and the Swedish University of Agricultural Sciences.