The Utility of Nuclear gapCp in Resolving Polyploid Fern Origins

Although polyploidy is rampant in ferns and plays a major role in shaping their diversity, the evolutionary history of many polyploid species remains poorly understood. Nuclear DNA sequences can provide valuable information for identifying polyploid origins; however, remarkably few nuclear markers have been developed specifically for ferns, and previously published primer sets do not work well in many fern lineages. In this study, we present new primer sequences for the amplification of a portion of the nuclear gapCp gene (encoding a glyceraldehyde-3-phosphate dehydrogenase). Through a broad survey across ferns, we demonstrate that these primers are nearly universal for this clade. With a case study in cheilanthoids, we show that this rapidly evolving marker is a powerful tool for discriminating between autopolyploids and allopolyploids. Our results indicate that gapCp holds considerable potential for addressing species-level questions across the fern tree of life. Keywords—allopolyploidy, autopolyploidy, ferns, gapCp, GAPDH, molecular systematics, nuclear marker. Polyploidy—the multiplication of entire chromosome sets—has been documented in nearly all major eukaryotic lineages (Otto and Whitton 2000; Gregory and Mable 2005; Tate et al. 2005). The process is especially prevalent in ferns, which exhibit both the highest known gametic chromosome numbers (Abraham and Ninan 1954) and some of the highest incidences of polyploidy (Walker 1966; Manton and Vida 1968; Löve et al. 1977; Walker 1984). Nearly 50% of fern species that have been studied cytogenetically are polyploids of recent origin (neopolyploids, following Ramsey and Schemske 2002), exhibiting chromosome numbers that are multiples of those documented in closely related species (Vida 1976; Walker 1984). In addition, it is estimated that at least 95% of fern species have undergone polyploidization at some point in their evolutionary history (Grant 1981; Haufler 1987). Polyploidy has left an indelible mark on fern evolution and continues to serve as a dynamic source of genetic variability, ecological innovation, and species diversity (e.g. Klekowski and Baker 1966; Klekowski 1972; Walker 1984; Werth et al. 1985; Soltis and Soltis 1987; Werth and Windham 1991; Haufler et al. 1995). Evolutionary biologists typically recognize two major categories of neopolyploids (see Soltis et al. 2007 and references therein). Organisms containing multiple genomes from a single diploid species are generally called autopolyploids; those that incorporate genomes derived from two or more diploid species are called allopolyploids. Although these categories are widely used, the distinction can be difficult to operationalize because it requires detailed knowledge of diploid progenitor populations (which may be unknown or extinct) and is dependent on the species concept applied (Soltis et al. 2007). Nonetheless, it is important to be able to assign individuals or taxa to these categories, because autopolyploids and allopolyploids exhibit fundamental differences in their genetics, ecology, and evolutionary potential (Levin 1983; Thompson and Lumaret 1992; Soltis and Soltis 2000; Wendel 2000; Ramsey and Schemske 2002; Osborn et al. 2003). Through the years, evolutionary biologists have used a variety of techniques to discriminate between autopolyploids and allopolyploids. Morphology, chromosome pairing behavior, and patterns of genetic segregation (based primarily on allozyme data) all provide important clues regarding polyploid origins (Grant 1981; Jackson 1982; Soltis and Rieseberg 1986). However, these approaches are inherently phenetic, relying exclusively on the genetic similarity of the genomes involved. As we move toward more integrated species concepts, it becomes increasingly important to place the genomes found in polyploid organisms in a phylogenetic context. Thus, DNA sequencing—specifically of nuclear and organellar markers in combination—is emerging as a powerful tool for revealing polyploid origins (Ge et al. 1999; Sang and Zhang 1999; Hoot and Taylor 2001; Popp and Oxelman 2001; Popp and Oxelman 2007). Biparentally-inherited nuclear markers provide sequences unique to individual diploid species, which can be isolated from polyploids through cloning and analyzed in a phylogenetic context. This allows for a more objective assessment of whether constituent genomes came from a single diploid species (autopolyploid) or more than one diploid species (allopolyploid). Maternallyinherited (Sears 1980; Gastony and Yatskievych 1992) organellar markers, in turn, can distinguish the maternal from pa-