170 Hybridization in Plants Polyploid Complexes Polyploids and their diploid progenitors often form complexes of related and hybridizing species A group of related diploid species may hybridize and form polyploid derivatives in a variety of combinations These polyploids may backcross to various diploids, or they may hybridize among themselves and form additional high-level polyploids (4x  2x ¼ 6x, 4x  4x ¼ 8x, etc.) These species may subdivide the habitat in such a way that new intermediate niches are occupied by the new species, which have been referred to as ‘‘fill-in taxa’’ (e.g., in Antennaria (pussytoes); Bayer et al., 1991) Complex patterns of hybridization and polyploid formation among a group of species may lead to the formation of a ‘‘compilospecies’’ – a group of interacting species that share a gene pool across ploidal levels Although an interesting concept, few examples of compilospecies have been documented, although recent molecular data for the grass Paspalum (bahia grass) in South America are consistent with the compilospecies concept Extensive gene flow among populations and between ploidal levels in a compilospecies can generate substantial genetic and phenotypic novelty that may confuse scientists but have significant ecological and evolutionary impact Polyploid species complexes may also generate asexual lineages that arise via the production of sterile hybrid offspring Asexual diploid or odd-ploid lineages may perpetuate themselves via a combination of methods, such as vegetative reproduction and apomixis, the production of seeds without sexual reproduction Rubus, which includes raspberries and blackberries, is notorious for forming asexual lineages associated with hybrids and polyploids Anywhere from a dozen to hundreds of species of Rubus have been recognized based on morphological variants, most of which are either apomictic or otherwise asexual Apomictic polyploid species complexes also tend to occur at high altitude and high latitude, in groups such as Draba (in the mustard family) and Antennaria (in the sunflower family) Polyploidy and Diversification Given the foregoing discussion, it should be obvious that hybridization and polyploidy generate extensive genetic and phenotypic novelty on which selection can act Some groups seem to undergo unreduced gamete formation – the most common process by which genomes are duplicated – much more frequently than others (Ramsey and Schemske, 1998, 2002), and polyploidy is therefore unequally distributed across even the plant branch of the tree of life Analyses of the frequency of polyploids, in the context of their diploid relatives, have yielded estimates on the frequency of polyploid plant speciation These estimates range from a low of 2–4% of angiosperm speciation events and 7% of fern speciation (Otto and Whitton, 2000) to 15% of angiosperm speciation events and 31% of fern speciation (Wood et al., 2009) – values that yield standing estimates of polyploid frequency much closer to estimates based on chromosome numbers than the estimates by Otto and Whitton (2000) These estimates, although highly speculative, relate to the origin of polyploid species However, rates of diversification must take both formation and extinction into account One way of examining diversification is by plotting species numbers across a phylogenetic tree – a depiction of the overall branching history of evolution When investigating a causal factor for increased or decreased diversification (the net of speciation and extinction), species numbers are examined relative to the origin of a specific characteristic For example, in an analysis of the possible causes of the early diversification of the angiosperms, Davies et al (2004) tested such attributes as pollination syndrome, geographic distribution, dispersal mode, habit, and chromosome number, but none of them, at least alone, is significantly associated with diversification early in angiosperm history However, although chromosome number was tested as a proxy for ploidy, ploidy itself was not investigated Because of the dramatic changes that can occur to a genome after polyploidization, leading to a chromosome number that might no longer be recognized as polyploid, perhaps chromosome number itself is not an appropriate character; alternatively, ploidy might not have been significantly associated with diversification early in angiosperm history Plotting known genome duplication events on the phylogenetic tree for angiosperms suggests that, in fact, diversification at a deep level may be linked to increases in ploidy (Soltis et al., 2009) A genome duplication event occurred before the origin of the large core eudicot clade (which contains B70% of all angiosperm species), and additional duplications are associated with the origins of some of the largest families: grasses, legumes, and the potato/tomato family (Soltis and Soltis, 2009) Although more rigorous studies are needed, along with more information on the evolutionary history of WGDs, these analyses suggest that polyploidization events deep in the evolutionary history of angiosperms may have triggered rapid diversification In contrast to these results, Mayrose et al (2011) reported that diversification rates among polyploids are lower than those among their diploid relatives, consistent with views by Stebbins, for example, that polyploid genomes provided buffering that retarded the rate of evolution and speciation Although the results of Soltis and Soltis, 2009 and Mayrose et al (2011) differ, this difference may actually be one of scale Soltis and Soltis, 2009 examined diversification rates deep in the evolutionary tree, whereas Mayrose et al (2011) focused on the tips It is very possible that polyploidization events are like other types of mutations – most are lost as detrimental, but some persist as positive changes Clearly, additional research is needed in this important area Recent Polyploidy Hybridization and polyploidy may generate new biodiversity – often recognized as species – and they can have both positive and negative impacts on existing communities Understanding the role that these processes play in generating and eroding biodiversity is crucial for establishing sound conservation practices However, till date, data on hybridization and polyploidization are often anecdotal and not predictive Furthermore, most polyploids have changed substantially since their