AJB Advance Article published on October 24, 2017, as 10.3732/ajb.1700180. The latest version is at http://www.amjbot.org/cgi/doi/10.3732/ajb.1700180 RESEARCH ARTICLE A M E R I C A N J O U R N A L O F B O TA N Y Karyotypic variation and pollen stainability in resynthesized allopolyploids Tragopogon miscellus and T. mirus1 Jonathan P. Spoelhof2,3,7, Michael Chester2,3,4, Roseana Rodriguez2,3, Blake Geraci2,3, Kweon Heo5, Evgeny Mavrodiev2, Pamela S. Soltis2,6, and Douglas E. Soltis2,3,6 PREMISE OF THE STUDY: Polyploidy has extensively shaped the evolution of plants, but the early stages of polyploidy are still poorly understood. The neoallopolyploid species Tragopogon mirus and T. miscellus are both characterized by widespread karyotypic variation, including frequent aneuploidy and intergenomic translocations. Our study illuminates the origins and early impacts of this variation by addressing two questions: How quickly does karyotypic variation accumulate in Tragopogon allopolyploids following whole-genome duplication (WGD), and how does the fertility of resynthesized Tragopogon allopolyploids evolve shortly after WGD? METHODS: We used genomic in situ hybridization and lactophenol-cotton blue staining to estimate the karyotypic variation and pollen stainability, respectively, of resynthesized T. mirus and T. miscellus during the first five generations after WGD. KEY RESULTS: Widespread karyotypic variation developed quickly in synthetics and resembled that of naturally occurring T. mirus and T. miscellus by generation S4. Pollen stainability in resynthesized allopolyploids was consistently lower than that of natural T. mirus and T. miscellus, as well as their respective diploid progenitor species. Logistic regression showed that mean pollen stainability increased slightly over four generations in resynthesized T. mirus but remained at equivalent levels in T. miscellus. CONCLUSIONS: Our results clarify some of the changes that occur in T. mirus and T. miscellus immediately following their origin, most notably the rapid onset of karyotypic variation within these species immediately following WGD. KEY WORDS Aneuploidy; Asteraceae; chromosomal translocation; genome duplication; genomic in situ hybridization; neopolyploidy; pollen; Tragopogon Polyploidy is a recurrent and defining feature of eukaryotic genome evolution. Most eukaryotic lineages are characterized by ancient polyploidy events (e.g., Wendel, 2000; Kellis et al., 2004; Gordon et al., 2009; Adolfsson et al., 2010; Jose and Dufresne, 2010; Jiao et al., 2011; Moghadam et al., 2011; P. S. Soltis and D. E. Soltis, 2012; 1 Manuscript received 16 May 2017; revision accepted 8 September 2017. Florida Museum of Natural History, University of Florida, Gainesville, Florida 32611, USA; 3 Department of Biology, University of Florida, Gainesville, Florida 32611, USA; 4 Royal Botanic Gardens, Kew, Natural Capital and Plant Health Department, Richmond, Surrey TW9 3DS, UK; 5 Kangwon National University, Department of Applied Plant Sciences, Chuncheon 24341, Korea; and 6 Genetics Institute, University of Florida, Gainesville, Florida 32610, USA 7 Author for correspondence (e-mail: firstname.lastname@example.org) https://doi.org/10.3732/ajb.1700180 2 Albert et al., 2013). More recent instances of polyploidy can be found in fishes, reptiles, amphibians, insects, fungi, ferns, lycophytes, gymnosperms, and angiosperms (Bogart, 1979; Lokki and Saura, 1980; Leggatt and Iwama, 2003; Wood et al., 2009; Albertin and Marullo, 2012; Li et al., 2015). Polyploidy is often divided into two broad classes: autopolyploidy and allopolyploidy. Although defined in multiple ways, autopolyploids are characterized by the possession of three or more homologous sets of chromosomes following genome duplication within a species, whereas allopolyploids are characterized by the possession of two or more homoeologous sets of chromosomes following interspecific hybridization (Stebbins, 1947). Autopolyploidy was long considered rare in comparison to allopolyploidy, but more recent evidence suggests that autopolyploidy and allopolyploidy occur at roughly equal rates in flowering plants (Ramsey and Schemske, 1998; Barker et al., 2016; but see Doyle and Sherman-Broyles, 2016). A M E R I C A N J O U R N A L O F B OTA N Y 104(10): 1–9, 2017; http://www.amjbot.org/ © 2017 Botanical Society of America • 1 Copyright 2017 by the Botanical Society of America 2 • A M E R I C A N J O U R N A L O F B OTA N Y In plants, the formation of polyploid genomes through wholegenome duplication (WGD) often precipitates remarkable evolutionary phenomena, including rapid speciation, genetic and morphological novelty, and, in some cases, adaptation to new biotic or abiotic ecological niches (Levin, 1983; Segraves and Thompson, 1999; Comai, 2005; Sobel et al., 2010; Ramsey, 2011; P. S. Soltis et al., 2014b; P. S. Soltis and D. E. Soltis, 2016; Visger et al., 2016). WGD contributes to speciation by causing immediate postzygotic reproductive isolation between polyploid and diploid individuals, as intercytotype mating often results in sterile or low-fertility offspring (Ramsey and Schemske, 1998). Concordantly, increases in ploidy have been estimated to account for up to 15% of speciation events in angiosperms and up to 31% of speciation events in ferns and lycophytes (Wood et al., 2009). Furthermore, these estimates may be conservative because autopolyploid plant speciation is rarely recognized taxonomically (D. E. Soltis et al., 2007; Barker et al., 2016). Polyploidy may also have a long-term impact on speciation, as multiple studies have concluded that polyploidy fosters species diversification in certain clades (e.g., D. E. Soltis et al., 2009; Tank et al., 2015; Zhan et al., 2016; but see Scarpino et al., 2014). In addition to representing a major mechanism of speciation, polyploidy also contributes to the development of biological novelty (Edger and Pires, 2009; P. S. Soltis et al., 2014b; P. S. Soltis and D. E. Soltis, 2016) through processes such as subfunctionalization (partitioning of gene function among duplicate genes following WGD), neofunctionalization (evolution of novel gene functions in duplicated genes following WGD) (P. S. Soltis et al., 2015; Wendel, 2015), transcriptional and proteomic novelty (D. E. Soltis et al., 2014a; D. E. Soltis et al., 2016a), and—particularly in allopolyploids—increased phenotypic variability (Ramsey and Schemske, 2002; Comai, 2005). Polyploid evolution tends to follow a trajectory in which each WGD is followed by a long process of genome evolution—often termed “diploidization”—that erodes the polyploid nature of the genome until a functionally diploid genome (i.e., one containing two sets of homologous chromosomes) is produced (Wendel, 2015; D. E. Soltis et al., 2016b). This cycle has multiple stages: the initial formation of neopolyploids (WGD), neopolyploidy (nascent polyploid lineages), mesopolyploidy (older lineages that are still undergoing diploidization), and ultimately paleopolyploidy (lineages whose genomes have become fully diploidized). We emphasize that these stages are delineated arbitrarily; diploidization is a continuous process that may differ among polyploid species or lineages. Although the study of polyploidy, particularly in plants, has advanced quickly, in large part through the study of genetic models and crop species, naturally occurring neopolyploids remain poorly understood because relatively few species have been identified in the very early stages after WGD. Examples of such species include the neoallopolyploids Senecio cambrensis Rosser (Hegarty et al., 2012), Spartina anglica C. E. Hubb. (Aïnouche et al., 2004), Tragopogon mirus Ownbey, and T. miscellus Ownbey (Ownbey, 1950). Tragopogon is an excellent natural system for the study of allopolyploidy across multiple stages. It contains repeatedly formed neoallopolyploids (T. mirus and T. miscellus; formed 80–90 yr ago) and mesopolyploids (e.g., T. castellanus Levier), and there are multiple ancient WGD events in Asteraceae, including an event that corresponds with the origin of the family (~40 mya) (Barker et al., 2008). Resynthesized T. mirus and T. miscellus have also been produced (Tate et al., 2009b) to investigate features of the very early neopolyploid stage in these species. Tragopogon mirus and T. miscellus are both freely selfing, biennial herbs that occur in eastern Washington state and western Idaho, USA. Ownbey (1950) determined that these two polyploid species formed via hybridization and genome doubling between the diploids T. dubius Scop. and T. porrifolius L. (T. mirus) and T. dubius and T. pratensis L. (T. miscellus). Ownbey described the new allotetraploids T. mirus and T. miscellus, perhaps within just 20 yr (or 10 generations) after their formation following the introduction of the diploid species to the northwestern United States from their native range in Europe. The earliest botanical collections of the diploids in the northwestern United States are from 1916 (T. porrifolius and T. pratensis) and 1928 (T. dubius) (Novak et al., 1991). Recent studies of T. mirus and T. miscellus have already yielded numerous insights into the cytogenetics, gene evolution, expression, and proteomics of young allopolyploids (e.g., D. E. Soltis et al., 2004; Tate et al., 2009a, 2009b; Buggs et al., 2010; Koh et al., 2010; Chester et al., 2012, 2015; Koh et al., 2012; D. E. Soltis et al., 2016a). Among these findings is the unexpected discovery that T. mirus and T. miscellus possess widespread karyotypic variation within and between populations and, in some cases, even among siblings (Chester et al., 2012, 2015). This variation includes frequent numeric aneuploidy (possession of a noneuploid chromosome number following the loss or gain of one or more chromosomes), compensated aneuploidy (possession of a euploid chromosome number after the loss or gain of one chromosome is compensated by the gain or loss, respectively, of a different chromosome), and homoeologous (intergenomic) chromosomal translocations (Fig. 1). Aneuploidy and intergenomic translocations are often associated with meiotic instability and corresponding reductions in gametophytic viability and overall fertility (Jackson, 1976; Ramsey and Schemske, 2002; Gaeta and Pires, 2010). Ownbey (1950) observed meiotic abnormalities (e.g., multivalent pairing) in plants from populations of T. mirus and T. miscellus, but also reported surprisingly high rates of pollen stainability (91.5–92.0%), which serves as a proxy measure of fertility. Tate et al. (2009b) resynthesized allopolyploid T. mirus and T. miscellus by crossing their respective diploid parents and treating the F1 progeny with colchicine to induce WGD. They reported meiotic abnormalities in the S0 (generation of polyploidy induction) and S1 (first selfed generation after WGD) generations of these resynthesized lines (Lim et al., 2008; Tate et al., 2009b). However, karyotypic diversity and pollen stainability in resynthesized T. mirus and T. miscellus have not been investigated previously. Here, we present these data across five generations in resynthesized Tragopogon allopolyploids and compare these data to karyotypic diversity and pollen stainability in naturally occurring T. mirus, T. miscellus, and their diploid progenitor species T. dubius, T. porrifolius, and T. pratensis. In doing so, we address two questions. First, does the widespread karyotypic variation observed in T. mirus and T. miscellus arise immediately after WGD, within a few generations of formation, or over many generations? Second, is fertility in resynthesized Tragopogon allopolyploids initially low or high—when compared to natural allopolyploids and diploid progenitors—following WGD, and how does it change across generations? By addressing these questions, we clarify some of the changes that occur during the neopolyploid stage. Although some of these questions have been addressed in other resynthesized allopolyploids, particularly in crop species such as Triticum aestivum L. and Brassica napus L. (e.g., Mestiri et al., 2010; Xiong et al., 2011), T. mirus and T. miscellus offer a replicated, natural system to study genome evolution O C TO B E R 2017 , V O LU M E 104 • S P O E L H O F E T A L. — C Y TO LO G Y A N D F E R T I L I T Y O F S Y N T H E T I C T R AG O P O G O N A L LO P O LY P LO I D S • 3 the S0 generation in which WGD was induced. In this experiment, 23 and 10 distinct resynthesized lineages (defined as arising from distinct WGD events) were included from T. mirus and T. miscellus, respectively. GISH and microscopy—Multiple root tips (ap- proximately 3–6 per plant, 2 cm in length) were collected from each greenhouse-grown plant of T. miscellus and T. mirus and then treated in 1 mL of 2 mM 8-hydroxyquinoline at 4°C in the dark for ~16 h. Roots were then fixed in a solution of ice-cold 90% acetic acid for 10 min, after which they were transferred into 70% ethanol and placed in cold storage (−20°C) for later analysis. Genomic in situ hybridization (GISH) analysis was later performed on chromosome squashes from fixed root tips as described in Chester et al. (2012, 2015). FIGURE 1 Example karyotypes from resynthesized S4 Tragopogon miscellus. (A) Euploid individual For chromosome imaging, slides were (2n = 24). (B) Numeric aneuploid individual (2n = 23) with a monosomic F chromosome in the T. dubius subgenome. The monosomic chromosome is marked with an ‘m’. (C) Individual with in- viewed under a Zeiss (Oberkochen, Badenstances of compensated aneuploidy of chromosomes D (monosomic in the T. dubius subge- Württemberg, Germany) Axio Imager M2 nome and trisomic in the T. pratensis subgenome) and E (trisomic in the T. dubius subgenome fluorescence microscope, and images of highand monosomic in the T. pratensis subgenome). Monosomic chromosomes are marked with an quality spreads were captured. The brightness ‘m’, and trisomic chromosomes are marked with a ‘t’. (D) Individual bearing an intergenomic of these images was adjusted in AxioVision translocation. An arrow marks the translocation breakpoint. The hue of this image has been al- version 4.8, after which these images were transferred to Adobe (San Jose, California, tered to accommodate readers with red-green color blindness. USA) Photoshop CS3 as composite TIFF files. In Photoshop, chromosomes were cut out inin the critical neopolyploid stage, which remains poorly understood dividually, paired, and aligned based on subgenome, chromosome despite its tremendous role in the survival and establishment of new type, and centromere location. Imaging and karyotyping, as well as polyploid lineages. designations of chromosome types A to F, followed Chester et al. (2012). Chromosome number, aneuploidy, and translocation data were gathered from the finished karyotypes. Karyotypes from 21 MATERIALS AND METHODS individuals (S1: n = 3; S2: n = 3; S4: n = 13; S5: n = 2) of T. miscellus and 42 individuals (S1: n = 12; S2: n = 18; S4: n = 12) of T. mirus were Generation of resynthesized T. mirus and T. miscellus—Resynthecharacterized for these traits. Unfortunately, difficulties in obtainsized lineages of T. mirus and T. miscellus were generated at the ing chromosome spreads with sufficient chromosome separation University of Florida (Gainesville, Florida, USA) as described in and integrity prevented the successful collection of karyotypic data Tate et al. (2009b). Briefly, seeds of the diploid parents (T. dubius, from many individuals, including all individuals from the S3 genT. porrifolius, and T. pratensis) were gathered from natural populaeration. These issues were the result of either ineffective fixation or tions, germinated, and grown in an air-conditioned greenhouse collection of root tips at a time when meristematic cells were not (GPS Coordinates: 29.6436, −82.3449) with supplemental lighting. actively dividing. Crosses were performed by hand between T. dubius and T. porrifolius (the parents of T. mirus) and between T. dubius and T. pratensis Pollen stainability—Pollen was sampled from 322 individuals (the parents of T. miscellus) by removing pollen from flowers of the (S1: n = 23; S2: n = 50; S3: n = 136; S4: n = 113) of resynthesized T. miscelmaternal parent with compressed air immediately after the flower lus and 335 individuals (S1: n = 55; S2: n = 85; S3: n = 117; S4: n = 78) head opened, then transferring pollen from the paternal parent of resynthesized T. mirus at anthesis. Pollen fertility was estimated once the stigmas of the maternal parent were receptive. The resultusing the traditional method of lactophenol blue staining as in Tate ing seeds were germinated on a moist paper towel in a Petri dish et al. (2009b). Pollen was sampled from naturally occurring North until the cotyledons emerged, at which point the seedlings were American populations of T. mirus (n = 9), T. miscellus (n = 26), treated with an aqueous solution of colchicine to induce genome T. dubius (n = 20), T. porrifolius (n = 3), and T. pratensis (n = 6), doubling (for more details, see Tate et al., 2009b). The treated indiincluding the populations from which resynthesized allopolyploids viduals were grown in a greenhouse at the University of Florida for were derived (Table 1). In this case, pollen was collected from ~6 mo, and flow cytometry and PCR analysis of the nuclear locus pressed specimens within 1 wk of collection and assayed for stainTDF85, which is polymorphic between each diploid progenitor ability. These specimens are located at the University of Florida species (Tate et al., 2006), were used to identify polyploid individuHerbarium (FLAS) (in queue for processing at the time of publicaals of confirmed hybrid origin. Later generations of T. mirus and T. tion). Results are reported as the proportion of stainable pollen out miscellus (S1–S5) were propagated by self-pollinating plants from of the total number scored (range: 26–318, mean = 104.6). 4 • A M E R I C A N J O U R N A L O F B OTA N Y TABLE 1. Names and locations of the natural populations of Tragopogon from which pollen samples were collected. Population State Latitude Longitude Garfield Moscow a Oaksdale Palouse Pullman a Rosalia Spangle a Spokane a Spokane Valley Tekoa WA ID WA WA WA WA WA WA WA WA 47.01 46.73 47.13 46.91 46.73 47.24 47.43 47.65 47.66 47.23 −117.15 −117.01 −117.24 −117.08 −117.19 −117.37 −117.38 −117.44 −117.20 −117.07 a Population from which the original parents of synthetic Tragopogon mirus and T. miscellus were collected (see Tate et al., 2009b: table 1). There are numerous challenges in growing multiple generations of these synthetic plants. Although we intended to test the effect of karyotype on pollen stainability directly and across generations, many individuals that were successfully karyotyped died prior to flower production. Early senescence may have resulted from cultural causes (our greenhouse conditions allowed for successful growth and reproduction of all Tragopogon species involved, but they did not closely mimic the temperature or light intensity of their habitat in Washington and Idaho) or genetic causes, such as extreme aneuploid karyotypes. These senescent individuals were neither assayed for pollen stainability nor advanced to the following generation, which contributed to a lack of representation of some lineages in the S2 through S4 generations (discussed below). The resulting loss of certain lineages, combined with previously described difficulties in obtaining clean chromosome squashes for some plants, prevented us from analyzing the direct relationship between karyotype and pollen stainability with adequate replication. As a result, we present individual analyses of each trait. Statistical analyses—All statistical analyses were performed in R (R Core Team, 2016). Differences in the proportions of euploidy, numeric aneuploidy, compensated aneuploidy, and chromosomal translocations between resynthesized S4 and natural allopolyploid T. mirus and T. miscellus were analyzed using Fisher’s exact test (Fisher, 1956). Patterns of chromosome gains and losses were tested with binomial exact tests, and the distribution of chromosome gains and losses among specific chromosomes was analyzed with a multinomial exact test. Generation S4 was selected for analysis of karyotypic diversity because each species was relatively well sampled in this generation (T. mirus: n = 12; T. miscellus: n = 13). Logistic regression analysis was used to test the effect of generation on the proportion of stainable pollen produced in resynthesized Tragopogon allopolyploids. Models that incorporated other factors, such as lineage, were not included, because all other factors significantly decreased model parsimony (as measured by Akaike’s information criterion; Akaike, 1974). Because many lineages lacked representation in the S2 through S4 generations, pollen data from complete lineages with at least two individuals in each generation were analyzed (T. mirus: n = 214; T. miscellus: n = 303) along with the complete (unrestricted) dataset to identify potential distortion from lineages that lack representation in one or more generations. These lineage-restricted datasets represented 9 of 23 and 6 of 10 independent resynthesized lineages in T. mirus and T. miscellus, respectively. All regression models were corrected for overdispersion with the R function “glm.binomial.disp” (“dispmod” package; Scrucca, 2012), and Hosmer-Lemeshow (Hosmer and Lemeshow, 2000) tests were used to confirm correct model specification (function “hoslem.test,” “ResourceSelection” package; Lele et al., 2016). Pairwise differences in pollen stainability between resynthesized T. mirus and T. miscellus generations were also tested with two-sided, unequal-variance t-tests on logit-transformed data. Due to the nonnormal and multiform distributions of pollen stainability data among naturally occurring Tragopogon species, exact permutation tests (function “permTS,” “perm” package; Fay, 2010) were used to compare the mean pollen stainability of resynthesized Tragopogon allopolyploids and naturally occurring allopolyploids (1 million permutations were used in all tests). Bonferroni-Holm (Holm, 1979) P-value correction was used to account for multiple comparisons when testing for differences in pollen stainability. RESULTS Aneuploidy—Both numeric and compensated aneuploidy (Fig. 1B, C) appeared to be rare in the first two generations after WGD; in T. mirus, only one unique numeric aneuploidy event (i.e., not originating from an event in a prior generation) was observed in each of the S1 (n = 12) and S2 (n = 18) generations. A similar pattern was seen in T. miscellus, where only one compensated aneuploid individual was observed in the S1 (n = 3) generation, and no aneuploids were observed in the S2 (n = 3) generation, although replication was low in these generations. Between generations S2 and S4, the frequency of aneuploidy appeared to increase markedly in both species. Two numeric aneuploid and three compensated aneuploid individuals of T. mirus were observed in the S4 generation (n = 12). Similarly, two numeric aneuploid and four compensated aneuploid individuals of T. miscellus were observed in the S4 generation (n = 13). All of the Tragopogon S4 aneuploids—whether numeric or compensated—had unique karyotypes within their respective lineages. In T. miscellus, two S5 individuals were also karyotyped, each of which possessed a unique aneuploid or compensated aneuploid karyotype. Chromosome gains and losses (nullisomy, monosomy, trisomy, or tetrasomy) were not distributed randomly among chromosomes in either species (multinomial exact test, T. mirus: P = 0.006; T. miscellus: P = 0.003) and appeared to be more common among the shorter D, E, and F chromosomes (Table 2). As in naturally occurring T. mirus and T. miscellus (Chester et al., 2012, 2015), there was no observed parental bias in overall chromosome gains or losses in resynthesized T. mirus or T. miscellus (binomial exact test, P > 0.5 for all comparisons), although the low replication of these events (Table 2) precluded insightful analyses of gains or losses for individual chromosomes. TABLE 2. Summary of all chromosome gains (trisomy or tetrasomy) and losses (monosomy or nullisomy) observed in resynthesized Tragopogon mirus and T. miscellus. Chromosome (gains/losses) Species T. mirus T. miscellus Subgenome A B C D E F T. dubius T. porrifolius T. dubius T. pratensis 0/0 0/0 1/0 0/0 0/0 0/0 0/0 0/0 1/3 2/1 0/0 0/0 0/2 1/1 1/1 2/2 0/1 1/1 2/2 1/1 0/1 0/0 2/1 0/2 O C TO B E R 2017 , V O LU M E 104 • S P O E L H O F E T A L. — C Y TO LO G Y A N D F E R T I L I T Y O F S Y N T H E T I C T R AG O P O G O N A L LO P O LY P LO I D S Intergenomic translocations—Intergenomic chromosomal trans- locations (Fig. 1D) in resynthesized Tragopogon allopolyploids were observed more frequently in generations S4 and S5 than in earlier generations and were more common in T. miscellus than in T. mirus. In T. mirus, one unique translocation was observed in generation S2 (n = 18), and two were observed in generation S4 (n = 12). By contrast, four unique translocations were observed in generation S4 (n = 13) T. miscellus, and one translocation was observed in generation S5 (n = 2). All translocations were distributed among the longer A, B, and C chromosomes of each subgenome in both allopolyploid species. A similar distribution of intergenomic translocations among chromosomes was observed by Chester et al. (2012, 2015) in natural T. mirus and T. miscellus. Karyotype comparison with naturally occurring allopolyploid Tragopogon—The proportions of euploid, numeric aneuploid, and compensated aneuploid individuals were highly similar between S4 plants and plants from naturally occurring (~40 generations post-WGD) populations of T. mirus. However, differences were apparent between S4 plants and plants from natural populations of T. miscellus (Table 3). The S4 resynthesized T. miscellus population had nearly twice the proportion of euploid individuals and half the proportion of compensated aneuploid individuals when compared to results for natural populations. However, these differences were not significant when tested with Fisher’s exact test (euploidy: P = 0.20; compensated aneuploidy: P = 0.12). By contrast, the proportion of individuals bearing intergenomic translocations differed between S4 and natural populations (Table 3), with the latter possessing a higher frequency of translocations in both species (Fisher’s exact test, T. mirus: P < 0.0001; T. miscellus: P = 0.017). Pollen stainability in resynthesized allopolyploids—Logistic re- gression showed a significant increase in pollen stainability between generations S1 and S4 in resynthesized T. mirus (P = 0.022), but not in resynthesized T. miscellus (P = 0.29) (Appendix S1; see Supplemental Data with this article). Summary statistics and pairwise comparisons of mean pollen stainability from each generation are presented in Appendix S2. Inclusion of all data, including lineages lacking representation in one or more generations, slightly decreased the slope of the regression in T. mirus and slightly increased the slope of the regression in T. miscellus (Appendix S1). The inclusion of these data did not affect the significance of the effect of generation on pollen stainability in either species. However, TABLE 3. Comparison of karyotype proportions between resynthesized S4 and naturally occurring (Nat.) Tragopogon mirus and T. miscellus. T. mirus Euploid Compensated aneuploid Numeric aneuploid Translocations T. miscellus Euploid Compensated aneuploid Numeric aneuploid Translocations S4 (n = 12) Nat. (n = 37) Pa 0.58 0.25 0.17 0.17 0.62 0.24 0.14 0.89 1.00 1.00 1.00 <0.0001 S4 (n = 13) Nat. (n = 58) Pa 0.54 0.31 0.15 0.38 0.31 0.59 0.10 0.76 0.20 0.12 0.63 0.017 a P value of a Fisher’s exact test comparing natural and synthetic S4 populations. Significant values are in bold. • 5 there was not a consistent increase with each passing generation; mean pollen stainability in generation S4 was significantly lower than in generation S3 in both T. mirus and T. miscellus (Appendix S2). Pollen stainability in naturally occurring allopolyploids and diploid progenitors—Consistent with the findings of Ownbey (1950), natural populations of T. mirus, T. miscellus, T. dubius, T. porrifolius, and T. pratensis display high mean pollen stainability (range: 0.82–0.97; see Table 4). At each synthetic generation, the mean pollen stainabilities of resynthesized T. mirus and T. miscellus were significantly lower than natural T. mirus and T. miscellus, respectively, with the exception of S3 T. mirus in the lineage-restricted dataset (Fig. 2). Also, natural allopolyploids had lower pollen stainabilities than their diploid progenitors; these differences were all found to be significant, with the exception of the comparison between T. miscellus and T. pratensis (Table 4). However, the low sample sizes of natural T. porrifolius (n = 3) and T. pratensis (n = 6) cast doubt on the significance of these results. Collection of pollen samples from recently pressed herbarium specimens did not appear to influence the condition of the pollen or the staining procedure, and pollen staining of both fresh and dry specimens is common in past studies (e.g., Mosquin, 1971). Still, we cannot rule out the possibility that drying specimens from natural populations reduced their pollen stainability, so it is possible that our estimate of the difference in pollen stainability between resynthesized and naturally occurring Tragopogon allopolyploids is conservative. DISCUSSION We have presented strong evidence that widespread karyotypic variation develops within the first few generations after WGD in resynthesized T. mirus and T. miscellus. The proportions of euploid, numeric aneuploid, and compensated aneuploid individuals at generation S4 were not statistically different from those of natural populations (Chester et al., 2012, 2015) in either polyploid species. Therefore, we infer that the karyotypic diversity observed in natural populations likely developed rapidly after allopolyploidization. This karyotypic diversity probably stems from meiotic abnormalities, including multivalent or homoeologous bivalent pairing resulting from residual homology between homoeologous chromosomes, and from segregation errors such as laggards and anaphase bridges that result from chromosomal rearrangements (e.g., inversions, acentric/dicentric chromosomes) or parental differences in centromeric histone structure or cell-cycle regulation (De Storme and Mason, 2014). Each of these abnormalities has been observed in resynthesized S0 and S1 T. mirus and T. miscellus (Lim et al., 2008; Tate et al., 2009b). While both naturally occurring and resynthesized Tragopogon allopolyploids are karyotypically diverse, their sporophytic chromosome numbers vary consistently between 2n = 23 and 2n = 25, with the exception of one resynthesized individual of T. mirus (2n = 22). This suggests that more extreme aneuploid karyotypes are typically nonviable. The occurrence of aneuploidy in resynthesized Brassica napus is similar to that of Tragopogon, with little aneuploidy in the S0 generation and a sharp increase in the S2 generation, followed by a leveling-off between generations S2 and S5 (Xiong et al., 2011). A study of resynthesized allohexaploid wheat (Triticum aestivum) lines (Mestiri et al., 2010) also showed an increasing occurrence of 6 • A M E R I C A N J O U R N A L O F B OTA N Y TABLE 4. Comparison of mean pollen stainability in Tragopogon between naturally occurring allopolyploids and their respective diploid progenitor species. Species T. mirus T. miscellus T. dubius T. porrifolius T. pratensis a FIGURE 2 Violin plots of proportion stainable pollen in resynthesized (A) Tragopogon mirus and (B) T. miscellus (only data from the restricted dataset are shown). Asterisks denote the significance of permutation tests (T. mirus—S1: P = 0.0020; S2: P = 0.00048; S3: P = 0.063; S4: P = 0.00020; T. miscellus—S1: P = 0.0011; S2: P = 0.00029; S3: P = 0.0021; S4: P = 8.0·10−6) comparing the mean pollen stainabilities of each generation of resynthesized T. mirus and T. miscellus to the mean of naturally occurring T. mirus and T. miscellus, respectively (ns = not significant; * P < 0.05, ** P < 0.01, *** P < 0.001). aneuploidy through generation S2. Aneuploidy in resynthesized wheat has been shown to persist through generation S20, although intergenomic translocations may be largely suppressed by the stabilizing presence of the Ph1 gene in this species (Zhang et al., 2013). When compared to resynthesized allopolyploid lines, naturally formed Brassica napus and wheat display little karyotypic deviation from parental additivity (Riley and Kimber, 1961; Xiong et al., 2011). Data from the older allopolyploid Tragopogon castellanus Mean SD n T. mirus P a T. miscellus P a 0.82 0.87 0.95 0.97 0.90 0.072 0.11 0.045 0.017 0.11 9 26 20 3 6 – – <0.001 0.018 – – – <0.001 – 0.67 P value of a two-sided exact permutation t-test. Significant values are in bold. reveal a similar level of karyotypic stability: in a survey of 99 plants from 59 populations of T. castellanus, Mavrodiev et al. (2015) found that all individuals had a euploid chromosome number (2n = 24) and also identified three independent origins of the species, two of which were euploid and one of which contained a fixed intergenomic translocation. While the stable genomes of natural wheat, B. napus, and T. castellanus all suggest a transient presence of widespread karyotypic variation after WGD, the presence of fixed intergenomic translocations in each of these species (Udall et al., 2005; Berkman et al., 2012; Mavrodiev et al., 2015) shows that a small amount of this variation can persist well past the neopolyploid stage. Differences in genomic composition (including intergenomic translocations) between independent formations of an allopolyploid species may drive differentiation, reproductive isolation, and potential speciation between the descendants of those formations (P. S. Soltis, 2013). Although extensive karyotypic variation in neoallopolyploids appears to be common, karyotypic stability in neoallopolyploids is frequently observed as well. Synthetic allopolyploids of Arabidopsis suecica (Fr.) Norrl. Ex O. E. Schulz display few meiotic aberrations and achieve stable bivalent pairing quickly after WGD, resulting in very few aneuploid progeny (Comai et al., 2003; Henry et al., 2014). Resynthesized lines of the allohexaploid species Senecio cambrensis are also meiotically stable (Weir and Ingram, 1980; Ingram and Noltie, 1986). The differences in observed aneuploidy between these species and Tragopogon allopolyploids are likely based on multiple factors. In S. cambrensis, it is possible that meiotic stability evolved in the tetraploid parent, S. vulgaris L., leading to increased meiotic stability in the resulting allohexaploid. Different species may also possess specific genes or alleles that stabilize meiosis, such as the aforementioned Ph1 gene in wheat or the BOY NAMED SUE locus in A. suecica (Henry et al., 2014). A marked difference was observed between naturally occurring and resynthesized Tragopogon allopolyploids in the proportion of plants bearing intergenomic translocations (Table 3), but confounding factors make this difference difficult to interpret. Natural allopolyploids possessed intergenomic translocations at a significantly higher frequency, but the number of unique translocations in natural populations is small (Chester et al., 2012, 2015), which is consistent with the assertion of Gaeta and Pires (2010) that allopolyploid evolution may permit the retention of a limited number of neutral or beneficial translocations, while selecting against the vast majority due to their deleterious effects on meiotic stability. Therefore, the frequency of translocations among natural allopolyploids may not reflect a higher rate of formation so much as common ancestry. A common feature of neopolyploids is reduced fertility, the causes of which are complex and include meiotic and developmental abnormalities, allelic composition, and epigenetic effects O C TO B E R 2017 , V O LU M E 104 • S P O E L H O F E T A L. — C Y TO LO G Y A N D F E R T I L I T Y O F S Y N T H E T I C T R AG O P O G O N A L LO P O LY P LO I D S (Jackson, 1976; Ramsey and Schemske, 2002; Comai, 2005). While reviewing research on the fertility of neopolyploids, Ramsey and Schemske (2002) found that their mean pollen viability was 71.5% (based on 200 different species), which was significantly lower than the mean of 91.4% in their diploid progenitors. Within this set of studies, analysis of the 20 allopolyploid species with measures of pollen viability in both the neopolyploid and progenitor species revealed similar results to their overall finding; the neoallopolyploids had an average pollen viability of 67.5%, which was significantly less than the mean of 90.7% in their respective progenitor species (exact two-sample permutation test, P < 0.001). Ramsey and Schemske (2002) also noted that pollen viability in neopolyploids responds quickly to selection; studies from neopolyploids in five species showed a mean pollen viability increase of 14% per generation under selection. However, the only allotetraploid included in this figure was Nicotiana glauca Graham × langsdorffii Weinm. (from Kostoff, 1938), which showed a mean pollen viability increase of 17% per generation over four generations of selection. Ownbey (1950) documented rates of pollen stainability >90% in natural Tragopogon allopolyploids and their diploid progenitor species. We found similar results by sampling naturally occurring Tragopogon populations in Washington and Idaho, although our estimates of pollen stainability in T. mirus and T. miscellus were somewhat lower than Ownbey’s (Table 4). By comparing these measures with those of resynthesized Tragopogon allopolyploids across multiple generations, we found that mean pollen stainability was reduced in resynthesized T. mirus and T. miscellus, and that mean pollen stainability increased significantly in T. mirus while it remained stable in T. miscellus (Fig. 2 and Appendix S1). Unlike the pollen viability studies reviewed by Ramsey and Schemske (2002), our experiment did not apply any overt selection, so the increase in pollen stainability likely came from the loss of individuals with highly deleterious karyotypes that did not survive or flower in the next generation. The more pronounced increase of pollen stainability in T. mirus could indicate that the deleterious effects of aneuploidy are stronger in this species, which is consistent with the lower rate of aneuploid individuals among natural T. mirus than among natural T. miscellus (Chester et al., 2012, 2015). Additionally, increases in pollen stainability did not appear to be lineage specific. Regression of pollen stainability against generation did not differ greatly between lineage-restricted and unrestricted datasets (Appendix S1). Loss of low-fitness lineages, and not just low-fitness individuals within lineages, would increase the slope of regression in analyses of unrestricted data. Finally, we did not expose plants to competition, or mimic the seasonal conditions of the Pacific Northwest, so the potential rate of increase in pollen viability under natural selection may be different in a natural context. Mating system may have a significant impact on the evolution of newly formed allopolyploids (Grant, 1956; P. S. Soltis and D. E. Soltis, 2000; Husband et al., 2008). In our study, polyploid lineages were propagated solely through selfing. This simulates the initial formation of an allopolyploid species with no potential conspecific reproductive partners. In this scenario, inbreeding depression may affect the fertility of the new polyploid, although this would be limited by persistent heterozygosity among homoeologous alleles, and karyotypic changes will only arise through meiotic or mitotic events within that lineage. In some allopolyploids, such as Brassica napus, chromosomal rearrangements can destabilize meiosis, leading to a positive feedback loop whereby karyotypic diversity increases in each selfed generation (Gaeta and Pires, 2010). Tragopogon mirus and T. miscellus both self-pollinate freely, and a study of T. mirus suggests that • 7 they are largely, although not exclusively, selfing in nature (Cook and Soltis, 1999). Outcrossing could occur if a new allopolyploid individual arose among other, preexisting allopolyploid lineages, which has happened repeatedly in T. mirus and T. miscellus (P. S. Soltis et al., 1995; D. E. Soltis et al., 2004). In a scenario in which a new allopolyploid arises sympatrically with preexisting lineages, outcross progeny are more likely to inherit beneficial alleles that stabilize meiosis and increase fertility, particularly if the lineages already possess such alleles and have survived long enough for natural selection to significantly increase their frequency in the population. This may limit the proliferation of karyotypic diversity in nonprimary formations of an allopolyploid species. The karyotype and pollen stainability data presented here indicate that the frequency of aneuploidy and compensated aneuploidy is similar between resynthesized S4 and natural Tragopogon allopolyploids (Table 3), while their rates of pollen stainability differ significantly. Furthermore, flowering plants of naturally occurring T. mirus and T. miscellus set seed normally, regardless of karyotype (P. S. Soltis, personal observation). This suggests that the effects of nonadditive karyotypes on fertility in neopolyploid Tragopogon are either not gravely deleterious per se or may be obscured by other factors. Polyploids are frequently tolerant of aneuploidy, at least when compared to diploids, because increased genomic redundancy buffers against losses of individual chromosomes (Leitch and Leitch, 2008), and because genetic factors can increase meiotic stability significantly in polyploids (e.g., Henry et al., 2014). Although our data show stable or increasing rates of mean pollen stainability and an increasing frequency of aneuploidy over generations, we were unable to test the relationship between karyotype and pollen stainability on an individual basis. Data from other resynthesized allopolyploid systems have demonstrated deleterious effects of aneuploidy. For example, aneuploid karyotypes have been associated with reduced pollen viability in resynthesized wheat (Zhang et al., 2013), and Xiong et al. (2011) showed a negative relationship between pollen stainability and the number of chromosomal losses or gains in resynthesized B. napus. Xiong et al. (2011) also reported that the most fertile individuals were chromosomally additive of the two parental species, B. oleracea L. and B. rapa L., which is consistent with selection acting to maintain near parental additivity in natural B. napus and other older allopolyploids (discussed above). Similarly, if the nearly additive genome of the established, repeatedly formed allopolyploid T. castellanus is representative of the late stages of allopolyploid evolution in Tragopogon, then one might expect to find a similar, if subtle, inverse correlation between aneuploidy and fertility in T. mirus and T. miscellus. However, aneuploidy has myriad phenotypic effects that are frequently more prominent than those caused by polyploidy alone (Birchler and Veitia, 2007), so other deleterious effects of aneuploidy may also guide allopolyploids toward nearparental additivity in the long term. Finally, the effects of aneuploidy are likely heterogeneous and depend on the specific karyotypes and alleles present in a given population. ACKNOWLEDGEMENTS The authors thank S. Shan and X. Liu for collecting and preparing naturally occurring Tragopogon for pollen analysis; M. Prindle and R. Riley for greenhouse and laboratory assistance; and two anonymous reviewers for comments on the manuscript. This research was supported by funding from the National Science Foundation (DEB-1146065 and DEB-0922003). 8 • A M E R I C A N J O U R N A L O F B OTA N Y LITERATURE CITED Adolfsson, S., Y. Michalakis, D. Paczesniak, S. N. Bode, R. K. Butlin, D. K. Lamatsch, M. J. Martins, et al. 2010. Evaluation of elevated ploidy and asexual reproduction as alternative explanations for geographic parthenogenesis in Eucypris virens ostracods. Evolution 64: 986–997. Aïnouche, M. L., A. Baumel, A. Salmon, and G. Yannic. 2004. Hybridization, polyploidy and speciation in Spartina (Poaceae). New Phytologist 161: 165–172. Akaike, H. 1974. A new look at the statistical model identification. IEEE Transactions on Automatic Control 19: 716–723. Albert, V. A., W. B. Barbazuk, J. P. Der, J. Leebens-Mack, H. Ma, J. D. Palmer, S. Rounsley, et al. 2013. The Amborella genome and the evolution of flowering plants. Science 342: 1241089. Albertin, W., and P. Marullo. 2012. Polyploidy in fungi: Evolution after wholegenome duplication. Proceedings. Biological Sciences 279: 2497–2509. Barker, M. S., N. Arrigo, A. E. Baniaga, Z. Li, and D. A. Levin. 2016. On the relative abundance of autopolyploids and allopolyploids. New Phytologist 210: 391–398. Barker, M. S., N. C. Kane, M. Matvienko, A. Kozik, R. W. Michelmore, S. J. Knapp, and L. H. Rieseberg. 2008. Multiple paleopolyploidizations during the evolution of the Compositae reveal parallel patterns of duplicate gene retention after millions of years. Molecular Biology and Evolution 25: 2445–2455. Berkman, P. J., A. Karshewski, S. Manoli, M. T. Lorenc, J. Stiller, L. Smits, K. Lai, et al. 2012. Sequencing wheat chromosome arm 7BS delimits the 7BS/4AL translocation and reveals homoeologous gene conservation. Theoretical and Applied Genetics 124: 423–432. Birchler, J. A., and R. A. Veitia. 2007. The gene balance hypothesis: From classical genetics to modern genomics. Plant Cell 19: 395–402. Bogart, J. P. 1979. Evolutionary implications of polyploidy in amphibians and reptiles. Basic Life Sciences 13: 341–378. Buggs, R. J. A., N. M. Elliott, L. Zhang, J. Koh, L. F. Viccini, D. E. Soltis, and P. S. Soltis. 2010. Tissue-specific silencing of homoeologs in natural populations of the recent allopolyploid Tragopogon mirus. New Phytologist 186: 175–183. Chester, M., J. P. Gallagher, V. V. Symonds, A. V. C. da Silva, E. V. Mavrodiev, A. R. Leitch, P. S. Soltis, and D. E. Soltis. 2012. Extensive chromosomal variation in a recently formed natural allopolyploid species, Tragopogon miscellus (Asteraceae). Proceedings of the National Academy of Sciences, USA 109: 1176–1181. Chester, M., R. K. Riley, P. S. Soltis, and D. E. Soltis. 2015. Patterns of chromosomal variation in natural populations of the neoallotetraploid Tragopogon mirus (Asteraceae). Heredity 114: 309–317. Comai, L. 2005. The advantages and disadvantages of being polyploid. Nature Reviews. Genetics 6: 836–846. Comai, L., A. P. Tyagi, and M. A. Lysak. 2003. Fish analysis of meiosis in Arabidopsis allopolyploids. Chromosome Research 11: 217–226. Cook, L. M., and P. S. Soltis. 1999. Mating systems of diploid and allotetraploid populations of Tragopogon (Asteraceae). I. Natural populations. Heredity 82: 237–244. De Storme, N., and A. Mason. 2014. Plant speciation through chromosome instability and ploidy change: Cellular mechanisms, molecular factors and evolutionary relevance. Current Plant Biology 1: 10–33. Doyle, J. J., and S. Sherman-Broyles. 2016. Double trouble: Taxonomy and definitions of polyploidy. New Phytologist 213: 487–493. Edger, P. P., and J. C. Pires. 2009. Gene and genome duplications: The impact of dosage-sensitivity on the fate of nuclear genes. Chromosome Research 17: 699–717. Fay, M. 2010. Perm: Exact or asymptotic permutation tests. R package version 1.0-0.0. Website: https://cran.r-project.org/web/packages/perm. Fisher, R. A. 1956. Statistical methods for research workers. Quarterly Journal of the Royal Meteorological Society 82: 119. Gaeta, R. T., and J. Pires. 2010. Homoeologous recombination in allopolyploids: The polyploid ratchet. The New Phytologist 186: 18–28. Gordon, J. L., K. P. Byrne, and K. H. Wolfe. 2009. Additions, losses, and rearrangements on the evolutionary route from a reconstructed ancestor to the modern Saccharomyces cerevisiae genome. PLOS Genetics 5: e1000485. Grant, V. 1956. The influence of breeding habit on the outcome of natural hybridization in plants. American Naturalist 90: 319–322. Hegarty, M. J., R. J. Abbott, and S. J. Hiscock. 2012. Allopolyploid speciation in action: Origin and evolution of Senecio cambrensis. In P. S. Soltis and D. S. Soltis [eds.], Polyploidy and genome evolution, 245–270. Springer-Verlag, Berlin, Germany. Henry, I. M., B. P. Dilkes, A. Tyagi, J. Gao, B. Christensen, and L. Comai. 2014. The BOY NAMED SUE quantitative trait locus confers increased meiotic stability to an adapted natural allopolyploid of Arabidopsis. The Plant Cell Online 26: 181–194. Holm, S. 1979. A simple sequentially rejective multiple test procedure. Scandinavian Journal of Statistics 6: 65–70. Hosmer, D. W., and S. Lemeshow. 2000. Applied logistic regression. Wiley, New York, New York, USA. Husband, B. C., B. Ozimec, S. L. Martin, and L. Pollock. 2008. Mating consequences of polyploid evolution in flowering plants: Current trends and insights from synthetic polyploids. International Journal of Plant Sciences 169: 195–206. Ingram, R., and H. J. Noltie. 1986. Early adjustment of patterns of metaphase association in the evolution of a polyploid species. Genetica 78: 21–24. Jackson, R. C. 1976. Evolution and systematic significance of polyploidy. Annual Review of Ecology and Systematics 7: 209–234. Jiao, Y., N. J. Wickett, S. Ayyampalayam, A. S. Chanderbali, L. Landherr, P. E. Ralph, L. P. Tomsho, et al. 2011. Ancestral polyploidy in seed plants and angiosperms. Nature 473: 97–100. Jose, C., and F. Dufresne. 2010. Differential survival among genotypes of Daphnia pulex differing in reproductive mode, ploidy level, and geographic origin. Evolutionary Ecology 24: 413–421. Kellis, M., B. W. Birren, and E. S. Lander. 2004. Proof and evolutionary analysis of ancient genome duplication in the yeast Saccharomyces cerevisiae. Nature 428: 617–624. Koh, J., S. Chen, N. Zhu, F. Yu, P. S. Soltis, and D. E. Soltis. 2012. Comparative proteomics of the recently and recurrently formed natural allopolyploid Tragopogon mirus (Asteraceae) and its parents. The New Phytologist 196: 292–305. Koh, J., P. S. Soltis, and D. E. Soltis. 2010. Homeolog loss and expression changes in natural populations of the recently and repeatedly formed allotetraploid Tragopogon mirus (Asteraceae). BMC Genomics 11: 97. Kostoff, D. 1938. Studies on polyploid plants. XXI. Cytogenetic behaviour of the allopolyploid hybrids Nicotiana glauca Grah. × Nicotiana langsdorffii Weinm. and their evolutionary significance. Journal of Genetics 37: 129–209. Leggatt, R. A., and G. K. Iwama. 2003. Occurrence of polyploidy in the fishes. Reviews in Fish Biology and Fisheries 13: 237–246. Leitch, A. R., and I. J. Leitch. 2008. Genomic plasticity and the diversity of polyploid plants. Science 320: 481–483. Lele, S. R., J. L. Keim, and P. Solymos. 2016. ResourceSelection: Resource selection (probability) functions for use-availability data. R package version 0.2-6. Website: https://CRAN.R-project.org/package=ResourceSelection. Levin, D. A. 1983. Polyploidy and novelty in flowering plants. American Naturalist 122: 1–25. Li, Z., A. E. Baniaga, E. B. Sessa, M. Scascitelli, S. W. Graham, L. H. Rieseberg, and M. S. Barker. 2015. Early genome duplications in conifers and other seed plants. Science Advances 1: e1501084. Lim, K. Y., D. E. Soltis, P. S. Soltis, J. A. Tate, R. Matyasek, H. Srubarova, A. Kovarik, et al. 2008. Rapid chromosome evolution in recently formed polyploids in Tragopogon (Asteraceae). PLoS One 3: e3353. Lokki, J., and A. Saura. 1980. Polyploidy in insect evolution. Basic Life Sciences 13: 277–312. Mavrodiev, E. V., M. Chester, V. N. Suárez-Santiago, C. J. Visger, R. Rodriguez, A. Susanna, R. M. Baldini, et al. 2015. Multiple origins and chromosomal novelty in the allotetraploid Tragopogon castellanus (Asteraceae). New Phytologist 206: 1172–1183. Mestiri, I., V. Chagué, A. M. Tanguy, C. Huneau, V. Huteau, H. Belcram, O. Coriton, et al. 2010. Newly synthesized wheat allohexaploids display progenitor-dependent meiotic stability and aneuploidy but structural genomic additivity. New Phytologist 186: 86–101. O C TO B E R 2017 , V O LU M E 104 • S P O E L H O F E T A L. — C Y TO LO G Y A N D F E R T I L I T Y O F S Y N T H E T I C T R AG O P O G O N A L LO P O LY P LO I D S Moghadam, H. K., M. M. Ferguson, and R. G. Danzmann. 2011. Whole genome duplication: Challenges and considerations associated with sequence orthology assignment in Salmoninae. Journal of Fish Biology 79: 561–574. Mosquin, T. 1971. Biosystematic studies in the North American species of Linum, section Adenolinum (Linaceae). Canadian Journal of Botany 49: 1379–1388. Novak, S. J., D. E. Soltis, and P. S. Soltis. 1991. Ownbey’s tragopogons: 40 years later. American Journal of Botany 78: 1586–1600. Ownbey, M. 1950. Natural hybridization and amphiploidy in the genus Tragopogon. American Journal of Botany 37: 487–499. Ramsey, J. 2011. Polyploidy and ecological adaptation in wild yarrow. Proceedings of the National Academy of Sciences, USA 108: 7096–7101. Ramsey, J., and D. W. Schemske. 1998. Pathways, mechanisms, and rates of polyploid formation in flowering plants. Annual Review of Ecology and Systematics 29: 467–501. Ramsey, J., and D. W. Schemske. 2002. Neopolyploidy in flowering plants. Annual Review of Ecology and Systematics 33: 589–639. R Core Team. 2016. R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. Available at https://www.R-project.org/. Riley, R., and G. Kimber. 1961. Aneuploids and the cytogenetic structure of wheat varietal populations. Heredity 16: 275–290. Scarpino, S. V., D. A. Levin, and L. A. Meyers. 2014. Polyploid formation shapes flowering plant diversity. American Naturalist 184: 456–465. Scrucca, L. 2012. dispmod: Dispersion models. R package version 1.1. Website: https://CRAN.R-project.org/package=dispmod. Segraves, K. A., and J. N. Thompson. 1999. Plant polyploidy and pollination: Floral traits and insect visits to diploid and tetraploid Heuchera grossulariifolia. Evolution 53: 1114–1127. Sobel, J. M., G. F. Chen, L. R. Watt, and D. W. Schemske. 2010. The biology of speciation. Evolution 64: 295–315. Soltis, D. E., V. A. Albert, J. Leebens-Mack, C. D. Bell, A. H. Paterson, C. Zheng, D. Sankoff, et al. 2009. Polyploidy and angiosperm diversification. American Journal of Botany 96: 336–348. Soltis, D. E., B. B. Misra, S. Shan, S. Chen, and P. S. Soltis. 2016a. Polyploidy and the proteome. Biochimica et Biophysica Acta. Proteins and Proteomics 1864: 896–907. Soltis, D. E., P. S. Soltis, J. C. Pires, A. Kovarik, J. A. Tate, and E. Mavrodiev. 2004. Recent and recurrent polyploidy in Tragopogon (Asteraceae): Cytogenetic, genomic and genetic comparisons. Biological Journal of the Linnean Society. Linnean Society of London 82: 485–501. Soltis, D. E., P. S. Soltis, D. W. Schemske, J. F. Hancock, J. N. Thompson, B. C. Husband, and W. S. Judd. 2007. Autopolyploidy in angiosperms: Have we grossly underestimated the number of species? Taxon 56: 13–30. Soltis, D. E., C. J. Visger, D. B. Marchant, and P. S. Soltis. 2016b. Polyploidy: Pitfalls and paths to a paradigm. American Journal of Botany 103: 1146–1166. Soltis, D. E., C. J. Visger, and P. S. Soltis. 2014a. The polyploidy revolution then… and now: Stebbins revisited. American Journal of Botany 101: 1057–1078. Soltis, P. S. 2013. Hybridization, speciation and novelty. Journal of Evolutionary Biology 26: 291–293. Soltis, P. S., X. Liu, D. B. Marchant, C. J. Visger, and D. E. Soltis. 2014b. Polyploidy and novelty: Gottlieb’s legacy. Philosophical Transactions of the Royal Society of London. Series B, Biological Sciences 369: 20130351. Soltis, P. S., D. B. Marchant, Y. Van de Peer, and D. E. Soltis. 2015. Polyploidy and genome evolution in plants. Current Opinion in Genetics & Development 35: 119–125. • 9 Soltis, P. S., G. M. Plunkett, S. J. Novak, and D. E. Soltis. 1995. Genetic variation in Tragopogon species: additional origins of the allotetraploids T. mirus and T. miscellus (Compositae). American Journal of Botany 82: 1329–1341. Soltis, P. S., and D. E. Soltis. 2000. The role of genetic and genomic attributes in the success of polyploids. Proceedings of the National Academy of Sciences, USA 97: 7051–7057. Soltis, P. S., and D. E. Soltis, Eds. 2012. Polyploidy and genome evolution. Springer, Berlin, Germany. Soltis, P. S., and D. E. Soltis. 2016. Ancient WGD events as drivers of key innovations in angiosperms. Current Opinion in Plant Biology 30: 159–165. Stebbins, G. L. 1947. Types of polyploids: Their classiφication and significance. Advances in Genetics 1: 403–429. Tank, D. C., J. M. Eastman, M. W. Pennell, P. S. Soltis, D. E. Soltis, C. E. Hinchliff, J. W. Brown, et al. 2015. Nested radiations and the pulse of angiosperm diversification: Increased diversification rates often follow whole genome duplications. New Phytologist 207: 454–467. Tate, J. A., P. Joshi, K. A. Soltis, P. S. Soltis, and D. E. Soltis. 2009a. On the road to diploidization? Homoeolog loss in independently formed populations of the allopolyploid Tragopogon miscellus (Asteraceae). BMC Plant Biology 9: 80. Tate, J. A., Z. Ni, A. C. Scheen, J. Koh, C. A. Gilbert, D. Lefkowitz, Z. J. Chen, et al. 2006. Evolution and expression of homeologous loci in Tragopogon miscellus (Asteraceae), a recent and reciprocally formed allopolyploid. Genetics 173: 1599–1611. Tate, J. A., V. V. Symonds, A. N. Doust, R. J. Buggs, E. Mavrodiev, L. C. Majure, P. S. Soltis, and D. E. Soltis. 2009b. Synthetic polyploids of Tragopogon miscellus and T. mirus (Asteraceae): 60 years after Ownbey’s discovery. American Journal of Botany 96: 979–988. Udall, J. A., P. A. Quijada, and T. C. Osborn. 2005. Detection of chromosomal rearrangements derived from homeologous recombination in four mapping populations of Brassica napus L. Genetics 169: 967–979. Visger, C. J., C. C. Germain-Aubrey, M. Patel, E. B. Sessa, P. S. Soltis, and D. E. Soltis. 2016. Niche divergence between diploid and autotetraploid Tolmiea. American Journal of Botany 103: 1396–1406. Weir, J., and R. Ingram. 1980. Ray morphology and cytological investigations of Senecio cambrensis Rosser. New Phytologist 86: 237–241. Wendel, J. F. 2000. Genome evolution in polyploids. Plant Molecular Biology 42: 225–249. Wendel, J. F. 2015. The wondrous cycles of polyploidy in plants. American Journal of Botany 102: 1753–1756. Wood, T. E., N. Takebayashi, M. S. Barker, I. Mayrose, P. B. Greenspoon, and L. H. Rieseberg. 2009. The frequency of polyploid speciation in vascular plants. Proceedings of the National Academy of Sciences, USA 106: 13875–13879. Xiong, Z., R. T. Gaeta, and J. C. Pires. 2011. Homoeologous shuffling and chromosome compensation maintain genome balance in resynthesized allopolyploid Brassica napus. Proceedings of the National Academy of Sciences, USA 108: 7908–7913. Zhan, S. H., M. Drori, E. E. Goldberg, S. P. Otto, and I. Mayrose. 2016. Phylogenetic evidence for cladogenetic polyploidization in land plants. American Journal of Botany 103: 1252–1258. Zhang, H., Y. Bian, X. Gou, B. Zhu, C. Xu, B. Qi, N. Li, et al. 2013. Persistent whole-chromosome aneuploidy is generally associated with young allohexaploid wheat. Proceedings of the National Academy of Sciences, USA 110: 3447–3452.