Polyploidy: some things old to go with the new

Abstract

Some of these columns will focus on newer sources of data and data analyses, other columns will emphasize the contributions of new data to plant systematics, and others, like this one, will consider how modem and more traditional approaches may be combined to produce refined insights into evolutionary processes in plants. Plant evolution via polyploidy is one such topic, and the recent conference on polyploidy sponsored by the Linnean Society and the Royal Botanic Gardens, Kew, illustrates the diversity of techniques (traditional and modem) now being employed to investigate this process. Polyploidy is arguably the most pervasive process in plant evolution, with perhaps half of all vascular plants being of polyploid origin (Grant, 1981; Haufler, 1987). Various aspects of polyploidy have occupied plant systematists-evolutionists for decades (Stebbins 1971, 1980; D. E. Soltis & P. S. Soltis, 1993). For the past 20 years many studies of polyploidy have used molecular data (especially analyses of allozymes) to document the mode of polyploidization and/or to confirm the ancestry of allopolyploids (Roose & Gottlieb, 1976; Werth & al., 1985a, b; D. E. Soltis & P. S. Soltis, 1993). The utility of these data resides in the joint presence in the polylploid of markers characteristic of or unique to each diploid parent. One of the best-known examples of recent polyploid origins and evolution that has been studied using molecular markers is in the genus Tragopogon. Studies during more than five decades have documented the origins and spread of two tetraploid species in the northwestern United States resulting from hybridizations among three introduced diploid species (Ownbey, 1950; Owbney & McCullum, 1953; Roose & Gottlieb, 1976; D. E. Soltis & P. S. Soltis, 1989; P. S. Soltis & D. E. Soltis, 1991; Novak & al., 1991; Cook & al., 1998). The results from Tragopogon and several other groups are in marked contrast to the more traditional view that the origin of a given polyploid is a rare, perhaps one-time occurrence; there is now convincing evidence from molecular markers for multiple origins of polyploids in Tragopogon and other genera (Ashton & Abbott, 1992; D. E. Soltis & P. S. Soltis, 1999). Also, the spread of the polyploids beyond the ranges of the diploids is more rapid than traditional concepts may have expected. Polyploidy as a process is fascinating because, in the case of allopolyploids, two morphologically distinct species that are genetically and chromosomally quite divergent may hybridize and produce polyploids that are more successful than either parent. A relatively new but very active area of current research on polyploidy employs molecular methods to understand what occurs at the genetic and chromosomal levels when two disparate genomes are brought together into the same organism. Such studies examine genomic and genetic divergence between polyploid genomes and their diploid progenitors. Both the extent and nature of the interactions between homoeologous gene pairs (including silencing) and the tempo of changes following polyploid formation are of primary interest. A future column will consider the nature of the changes in depth. Thus, the present contribution will focus on how more traditional data are valuable in molecular studies of polyploids and how the old time data enhance the value of the molecular data in providing greater insights into the tempo and mode of evolution in polyploid genomes. Recent reviews on the evolution of polyploid genomes (Comai, 2000; P. S. Soltis & D. E. Soltis, 2000; Wendel, 2000; Osborn & al., 2003, among others) provide insightful overviews of results from the increasing number of original research papers that have appeared in the past several years (e.g., Song & al., 1995; Feldman & al., 1997; Liu & al., 1998a, b; Axelsson & al., 2000; Ozkan & al., 2001; Kashkush & al., 2002; Madlung & al., 2002). The picture emerging from these studies is that rapid genomic and non-Mendelian changes may occur in the polyploids, although there have been contradictory results for the same polyploids (cf. Song & al., 1995; Axelsson & al., 2001), and some studies have failed to detect genomic changes in newly synthesized polyploids (Liu & al., 2001). In any study of polyploid evolution, a basic problem is disentangling the initial effects of polyploid formation from the divergence occurring subsequent to polyploid formation (Stebbins, 1971, 1980), and this issue has been considered in recent reviews (Wendel, 2000; Osborn &