Skip to main content
NCBI home page
Annals of Botany logoLink to Annals of Botany
. 2018 Dec 12;123(5):845–855. doi: 10.1093/aob/mcy219

Widespread co-occurrence of multiple ploidy levels in fragile ferns (Cystopteris fragilis complex; Cystopteridaceae) probably stems from similar ecology of cytotypes, their efficient dispersal and inter-ploidy hybridization

Kristýna Hanušová 1, Martin Čertner 1,2, Tomáš Urfus 1, Petr Koutecký 3, Jiří Košnar 3, Carl J Rothfels 4, Vlasta Jarolímová 2, Jan Ptáček 1, Libor Ekrt 3,✉
PMCID: PMC6526313  PMID: 30541055

Abstract

Background and Aims

Polyploidy has played an important role in the evolution of ferns. However, the dearth of data on cytotype diversity, cytotype distribution patterns and ecology in ferns is striking in comparison with angiosperms and prevents an assessment of whether cytotype coexistence and its mechanisms show similar patterns in both plant groups. Here, an attempt to fill this gap was made using the ploidy-variable and widely distributed Cystopteris fragilis complex.

Methods

Flow cytometry was used to assess DNA ploidy level and monoploid genome size (Cx value) of 5518 C. fragilis individuals from 449 populations collected over most of the species’ global distributional range, supplemented with data from 405 individuals representing other related species from the complex. Ecological preferences of C. fragilis tetraploids and hexaploids were compared using field-recorded parameters and database-extracted climate data.

Key Results

Altogether, five different ploidy levels (2x, 4x, 5x, 6x, 8x) were detected and three species exhibited intraspecific ploidy-level variation: C. fragilis, C. alpina and C. diaphana. Two predominant C. fragilis cytotypes, tetraploids and hexaploids, co-occur over most of Europe in a diffuse, mosaic-like pattern. Within this contact zone, 40 % of populations were mixed-ploidy and most also contained pentaploid hybrids. Environmental conditions had only a limited effect on the distribution of cytotypes. Differences were found in the Cx value of tetraploids and hexaploids: between-cytotype divergence was higher in uniform-ploidy than in mixed-ploidy populations.

Conclusions

High ploidy-level diversity and widespread cytotype coexistence in the C. fragilis complex match the well-documented patterns in some angiosperms. While ploidy coexistence in C. fragilis is not driven by environmental factors, it could be facilitated by the perennial life-form of the species, its reproductive modes and efficient wind dispersal of spores. Independent origins of hexaploids and/or inter-ploidy gene flow may be expected in mixed-ploidy populations according to Cx value comparisons.

Keywords: Bladder ferns, contact zone, Cx value, Cystopteris fragilis, cytotype coexistence, ecological preferences, flow cytometry, genome size, ploidy distribution, pteridophytes

INTRODUCTION

Polyploidization (whole-genome duplication) is widely considered one of the major forces contributing to the evolutionary diversification of land plants (Soltis et al., 2016; Landis et al., 2018). This is especially so in ferns, where approx. 30 % of speciation events are presumably linked to changes in ploidy, twice the rate predicted for angiosperms (Wood et al., 2009). Specifically, by providing immediate postzygotic reproductive isolation between newly arisen polyploids and their progenitors, polyploidization is an efficient mechanism of sympatric speciation (Ramsey and Schemske, 1998; Coyne and Orr, 2004). Polyploidization was suggested as the predominant mechanism of genome size expansion in ferns, and chromosome number and genome size are tightly correlated in this plant group, unlike in angiosperms (Barker, 2013; Clark et al., 2016). High accumulation of polyploidy in some fern lineages (Clark et al., 2016; Schneider et al., 2017) makes them suitable models for studying polyploid evolution.

The prevailing mode of polyploid origin is via unreduced gametes (i.e. gametes with a somatic chromosome number), produced as a consequence of rare meiotic errors (Ramsey, 2007; Kreiner et al., 2017). The rarity of polyploid formation is compounded by demographic challenges: new polyploids generally suffer from a lack of compatible mating partners, and crosses with their progenitors, which produce sterile odd-ploidy offspring, may lead to their extirpation (‘minority cytotype exclusion’; Levin, 1975). The frequency-dependent selection driving this process may similarly affect otherwise well-established cytotypes if they meet in contact zones, which makes minority cytotype exclusion a main constraint to ploidy coexistence in general (Husband, 2000). In recent decades, several mechanisms facilitating successful polyploid establishment and/or cytotype coexistence have been proposed (reviewed by Kolář et al., 2017). For example, the minority status of one of the coexisting cytotypes may be overcome by its recurrent origin (Ramsey, 2007), efficient vegetative spread (Chrtek et al., 2017), autogamy (Petit et al., 1997), non-random mating (Husband et al., 2008) or a substantial competitive advantage (Felber, 1991). Prominent among these mechanisms is the (fine-scale) spatial segregation of cytotypes, which can increase the rate of compatible, within-ploidy mating (Baack, 2005; Kolář et al., 2017). The most frequently reported cause of such spatial segregation of cytotypes is their different ecological preferences (e.g. Levin, 2002; Laport et al., 2016).

Cytogeography, the study of cytotype diversity and its distribution patterns, is usually the first step towards understanding the mechanisms of polyploid evolution (Soltis et al., 2003). Cytotype distribution patterns may point to differences in habitat preferences among cytotypes and enable the detection of zones of cytotype contact. Such information may provide insights into the temporal stability of ploidy coexistence and the origin of cytogenetic novelty, and possibly demonstrate the potential of contact zones in promoting inter-ploidy gene flow (Kolář et al., 2017). Of importance in this respect is local coexistence of different cytotypes, arising either after an in situ polyploidization event (‘primary contact’) or through cytotype immigration into populations of another cytotype (‘secondary contact’; Petit et al., 1999). Cytotype coexistence provides the opportunity for inter-ploidy crosses involving either reduced or unreduced gametes and thus may generate cytogenetic novelty (Kolář et al., 2017).

Despite the importance of polyploidy in ferns, only a handful of detailed cytogeographical studies are available for this plant group (e.g. Moran, 1982; Nakato and Kato, 2005; Chang et al., 2013; Grusz et al., 2014; Dauphin et al., 2018). The situation is further complicated by the limited geographical extent of these studies and their relatively small sample size, which can be largely attributed to using laborious methods of ploidy-level estimation (i.e. chromosome counts or measurements of spore diameter). The dearth of data on cytotype distribution and ecology in ferns, which is especially striking in comparison with the number of studies of angiosperms (reviewed by Kolář et al., 2017), prevents us from assessing whether cytotype coexistence and the mechanisms of polyploid evolution show similar patterns in the two plant groups. Nonetheless, fast and reliable ploidy assessment of high numbers of samples is now available via flow cytometry (Ekrt et al., 2010; Shinohara et al., 2010; Clark et al., 2016). Here, we use this technique to get better insight into fern polyploid evolution. We selected the Cystopteris fragilis complex (sensuRothfels et al., 2013) as a suitable model because, firstly, considerable ploidy-level diversity has been reported from natural populations of this group (six distinct cytotypes: 2x, 3x, 4x, 5x, 6x and 8x; Manton, 1950; Vida, 1974; Haufler et al., 1985; Haufler and Windham, 1991; Kawakami et al., 2010). Secondly, the complex (and especially C. fragilis) has an extremely wide geographical distribution (Rothfels et al., 2013), but whether spatial isolation of cytotypes or their sorting along ecological gradients contributes to cytotype coexistence remains unknown. Lastly, our preliminary ploidy screening in Central Europe revealed a high frequency of mixed-ploidy populations in C. fragilis. We used a wide array of complementary approaches, consisting of extensive field sampling, flow cytometric analysis of ploidy level and relative genome size, and ecological niche comparisons, to investigate the following questions: (1) What is the cytotype diversity in the C. fragilis complex and how is it geographically distributed (with a focus on Europe and North America)? (2) Could monoploid genome size comparisons provide additional clues to the evolution of cytotypes? (3) How common and widespread is cytotype coexistence in C. fragilis? (4) Is cytotype coexistence in C. fragilis driven by the underlying environmental heterogeneity?

MATERIALS AND METHODS

Model system

Fragile ferns (Cystopteris fragilis) are a complex of species within the eupolypod II family Cystopteridaceae (PPG I, 2016). Comprising approximately ten commonly recognized species, two of which are thought to be exclusively diploid (except rare incidence of autotriploidy) and the remainder predominantly polyploid, this complex is highly polymorphic and the species limits within it are uncertain, largely due to extensive patterns of allopolyploidy (Rothfels et al., 2013, 2014, 2017). The complex has a nearly global distribution (Rothfels et al., 2013), but individual taxa may be geographically restricted: the two named diploids (C. protrusa and C. reevesiana) are restricted to the Americas, as are the Mexican endemics C. millefolia and C. membranifolia and the north-eastern North American C. tenuis; C. alpina is European; C. douglasii is limited to Hawaii; and C. tasmanica occurs in Australia and New Zealand. The remaining diversity tends to be lumped into a heterogeneous north-temperate ‘C. fragilis’ or a sub-tropical ‘C. diaphana’– both of which, as typically circumscribed, occur over wide areas of multiple continents and comprise multiple ploidy levels. An additional putative taxon, C. dickieana, distinguished based on spore characters, is recognized by some authors (e.g. Vida, 1974; Fraser-Jenkins, 2008) but we here follow the recent tendency (e.g. Haufler and Windham, 1991; Parks et al., 2000; Rothfels, 2012) in lumping dickieana into C. fragilis. Members of the C. fragilis complex occasionally hybridize with the non-complex member C. bulbifera, forming the allopolyploids C. utahensis, C. tennesseensis and C. laurentiana (Haufler et al., 1993; Rothfels et al., 2017); these taxa are not included in this study.

The C. fragilis complex is a powerful system for studying polyploid evolution because of its broad geographical and habitat range, extensive ploidy variation – four cytotypes have been documented in natural populations of C. fragilis (4x, 5x, 6x, 8x) each of which probably contains multiple independent evolutionary units (Rothfels et al., 2014, 2017) – and because polyploids in Cystopteris reproduce sexually, avoiding the potentially confounding factor of apomixis (Rothfels, 2012). However, vegetative reproduction and a high frequency of inbreeding have also occasionally been reported in Cystopteris (Haufler et al., 1993; Gämperle and Schneller, 2002).

Field sampling

Cystopteris populations were sampled during 2012–2016. The initial strategy was to cover representatively all habitats occupied by C. fragilis across Europe with particular emphasis on Central Europe, where a pilot survey documented frequent cytotype coexistence. To put our results into a broader context, we additionally sampled C. fragilis populations in western Asia, and North and South America, and also included other species from the C. fragilis complex. While the locality selection was generally random, in North America, we deliberately targeted much rarer hexaploid populations of C. fragilis. Depending on the plant’s abundance at a given locality, we sampled 5–30 randomly selected plants with a minimum distance of 10 cm between sampled individuals. Each site was characterized by geographical coordinates, elevation, type of substrate (i.e. siliceous, alkaline and neutral), and general habitat description (inclusing natural vs. anthropogenic character), and herbarium vouchers were collected (see Supplementary Data Table S1). One leaf per plant was collected for ploidy determination. The leaves were either kept fresh until flow-cytometric analysis or immediately desiccated using silica gel; the method of tissue preservation did not affect the reliability of ploidy-level estimation (data not shown).

Flow cytometry and karyology

DNA ploidy levels and relative genome size of Cystopteris individuals were assessed from fluorescence intensities of DAPI-stained nuclei using flow cytometry. Sample preparation followed Čertner et al. (2017); Vicia faba ‘Inovec’ (2C DNA = 26.90 pg; Doležel et al., 1992) was used as an internal standard. Fluorescence intensity of 3000 particles was analysed using either Partec PA II or Partec CyFlow ML flow cytometers (Partec GmbH, Münster, Germany) equipped with a mercury arc lamp or a 365-nm UV LED, respectively. Up to five individuals were processed together during the ploidy screening, as our pilot analyses proved reliable detection of minority cytotypes in such pooled samples. We used DAPI-stained analyses for both ploidy-level screening and relative genome size estimation because these provide histograms with high resolution and do not require RNase treatment (Doležel et al., 2007). We are well aware that DAPI preferentially binds AT-rich regions of DNA, which may lead to seeming genome size differences among samples differing strongly in their genomic GC content. To complement and calibrate DAPI-stained analyses, genome size was determined for 51 selected samples using propidium iodide (PI) staining. Sample preparation was identical but only fresh material was used; PI was used as a fluorochrome and RNase IIA was added to the staining solution, both at a final concentration of 50 μg mL–1. Fluorescence intensity of 5000 particles was analysed using a CyFlow SL instrument (Partec GmbH) equipped with a green solid-state laser (Cobolt Samba, 532 nm, 100 mW). In PI analyses, samples were always processed individually and a mean value of three measurements on different days was used for genome size calculation.

The relationship between sample relative fluorescence and ploidy level was calibrated using chromosome counting. Spore mother cells of a single tetraploid C. fragilis individual (locality No. 126, see Supplementary Data Table S1) were pretreated with a saturated solution of p-dichlorobenzene (3 h, room temperature), fixed in a mixture of ethanol and acetic acid (3: 1), and stained using lacto-propionic orceine. The number of chromosomes was counted using a Carl-Zeiss Jena NU microscope (total magnification 1000×).

Genome size comparisons

Reported genome sizes, unless otherwise stated, are relative genome sizes based on DAPI staining (i.e. the sample to standard fluorescence ratio). We also calculated monoploid genome sizes (Cx values), by dividing the relative genome size by the ploidy level. This trait allows for comparisons of genome size independent of ploidy level; sufficient Cx value differences may in some cases be used to reconstruct modes of cytotype origin, cytotype relationships and, possibly, to detect newly originated polyploids (Čertner et al., 2017).

All Cx value comparisons were conducted on a subset of C. fragilis data with high-quality genome size estimates [coefficient of variation (CV) of both the sample and the standard peak <3 %; see Supplementary Data Table S2]; populations from Argentina with a distinct genome size were excluded. For an overall comparison of Cx values of C. fragilis tetraploids and hexaploids, and for a test for a latitudinal gradient in Cx values, the dataset was reduced by randomly selecting one sample per ploidy per population. The Cx value differences between the cytotypes (overall and between samples from uniform-ploidy and mixed-ploidy populations) were tested using a Kruskal–Wallis rank sum test. Latitudinal gradients of Cx values were tested separately for tetraploids and hexaploids in linear regression models with latitude as an explanatory variable. To corroborate the recurrent, hybrid origin of pentaploids in mixed-ploidy populations of C. fragilis, mean Cx values of cytotypes within populations were extracted. Only mixed-ploidy populations for which high-quality genome size estimates were available for all three coexisting cytotypes (i.e. 4x, 5x and 6x) were retained, resulting in a final dataset of 23 populations. The Cx value of pentaploids was used as a response variable in a linear regression model and the mean of Cx values of co-occurring tetra- and hexaploids served as a predictor. All statistical analyses were conducted in R ver. 3.4.3 (R Core Team, 2016).

Ecology of cytotypes

Three field-recorded parameters, type of substrate, habitat origin (i.e. natural vs. anthropogenic) and elevation, were combined with database-extracted climate data and used in comparisons of abiotic and climatic niches of cytotypes for the representatively sampled C. fragilis populations (see Supplementary Data Table S2). We compared ecological niches among uniformly tetraploid, uniformly hexaploid and mixed-ploidy populations; the last were identified by either co-occurrence of tetraploids and hexaploids at a site or by the presence of pentaploid individuals (resulting from inter-ploidy crosses). While climatic data were available on the global scale, the abiotic niche comparisons of cytotypes had to be restricted to the Eurasian range of C. fragilis as we lacked information on local abiotic parameters for the American samples.

The relative incidence of tetraploid, hexaploid and mixed-ploidy populations on different substrates and at habitats of natural or anthropogenic origin were compared using chi-square tests for homogeneity. Where there were significant differences, we conducted pairwise comparisons of populations with different cytotype composition and applied the Bonferroni correction. Only localities with siliceous or alkaline substrates were retained in the dataset because neutral substrates and other, intermediate or unclear assignments constituted a minority of sites (6.8 % of the data). Similarly, localities where natural or anthropogenic habitat status could not be unambiguously determined were excluded from the relevant analyses (1.9 %). The natural vs. anthropogenic origin was intended as a proxy of habitat history (e.g. higher frequency of one of the cytotypes at anthropogenic sites may indicate its relatively recent spread). Differences in mean elevation of tetraploid, hexaploid and mixed-ploidy populations were tested using one-way ANOVA. Elevation was square-root transformed prior to the analysis to meet the model assumptions. The final datasets consisted of 421, 400 and 429 population records for the tests of habitat origin, substrate type and elevation, respectively. All statistical analyses were conducted in R ver. 3.4.3 (R Core Team, 2016).

Georeferenced occurrences of tetraploid, hexaploid and mixed-ploidy populations were used to extract 19 Bioclim climate variables from the WorldClim database (http://www.worldclim.org/bioclim;Hijmans et al., 2005) downloaded in the highest available resolution (30 arc seconds ≈ 1 km2). To account for heterogeneity in sampling intensity, we used ArcGIS 10.0 (ESRI, Redlands, CA, USA) to divide the global range of C. fragilis into a grid of 0.5° × 0.5° cells and when multiple populations of the same cytotype composition were located within a cell, one was randomly selected for the analysis. A principal component analysis (PCA) was used to visualize climatic niches of the resulting 94 tetraploid, 98 hexaploid and 79 mixed-ploidy populations. Differences in climatic niches of the cytotypes were tested using a redundancy analysis (RDA) by applying a Monte Carlo test with 999 permutations. Cytotype composition of populations (tetraploid, hexaploid, mixed-ploidy) was used as an explanatory variable in RDA. Multivariate analyses were conducted in Canoco 5 (ter Braak and Šmilauer, 2012).

RESULTS

Cytotype diversity and its distribution patterns

During our detailed examination of cytotype diversity in C. fragilis, we sampled 5518 individuals from 449 localities across four continents (Fig. 1). Four ploidy levels were detected (4x, 5x, 6x and 8x; Table 1, Supplementary Data Fig. S1) and chromosome counts confirmed n = 84 in tetraploids (Supplementary Data Fig. S2), consistent with the Cystopteris base number of x = 42. The two most common cytotypes in C. fragilis, tetraploids and hexaploids, occurred at similar frequencies (51 % and 46 % of samples, respectively). However, the relative frequency of tetraploids and hexaploids differed between Eurasia and the Americas (50 % and 47 % in Eurasia vs. 80 % and 15 % in the Americas, respectively), despite the fact that hexaploid populations were deliberately targeted in North America. Pentaploid individuals were quite scarce (3 % of samples) and only a single octoploid individual (0.02 % of samples) was sampled (it grew at a site of anthropogenic origin; see Supplementary Data Table S1).

Fig. 1.

Fig. 1.

Open in a new tab

Spatial distribution of cytotype diversity in Cystopteris fragilis and related species. Mixed-ploidy populations are depicted as pie charts showing the local frequency of cytotypes. Sampling at a world-wide scale (A), with inset displaying the sampled populations of C. alpina (B). Detail of cytotype distribution in the European Cystopteris fragilis populations (C) subjected to more intensive field sampling. For the more detailed distribution of central European populations of C. fragilis see Supplementary Data Fig. S3.

Table 1.

Overview of cytotype diversity and genome size parameters for commonly recognized members of the Cystopteris fragilis complex. Using flow cytometry, we estimated the relative genome size (the ratio of sample to standard fluorescence), monoploid genome size (relative genome size divided by ploidy level) and absolute genome size (propidium iodide staining, in pg)

Taxon Ploidy Number of sampled plants Relative GS* (mean ± s.d.) Monoploid GS* (Cx value) (mean ± s.d.) Mean CV Absolute GS* (pg) (mean ± s.d.) Mean CV
C. fragilis 4x 2775 (28)† 0.436 ± 0.009 0.109 ± 0.002 2.20 ± 0.47 14.26 ± 0.070 2.48 ± 0.31
5x 176 (7)† 0.542 ± 0.010 0.108 ± 0.002 2.18 ± 0.45 17.59 ± 0.059 2.27 ± 0.14
6x 2524 (12)† 0.638 ± 0.013 0.106 ± 0.002 2.09 ± 0.45 20.80 ± 0.424 2.64 ± 0.23
8x 1 0.863 0.108 1.02
C. f. from Argentina 4x 42 0.459 ± 0.005 0.115 ± 0.001 2.20 ± 0.37
C. alpina 4x 21 0.441 ± 0.001 0.111 ± 0.001 2.81 ± 0.22
5x 6 0.536 ± 0.004 0.107 ± 0.001 2.09 ± 0.54
6x 202 (4)† 0.633 ± 0.010 0.106 ± 0.002 2.31 ± 0.36 20.91 ± 0.130 2.75 ± 0.06
8x 2 0.850 ± 0.034 0.106 ± 0.004 1.77 ± 0.33
C. diaphana 4x 36 0.577 ± 0.006 0.144 ± 0.001 2.49 ± 0.29
6x 55 0.790 ±0.014 0.132 ± 0.002 1.99 ± 0.67
C. protrusa 2x 9 0.260 ± 0.001 0.130 ± 0.000 2.61 ± 0.54
C. reevesiana 2x 54 0.214 ± 0.005 0.107 ± 0.002 2.74 ± 0.22
C. tenuis 4x 20 0.464 ± 0.008 0.116 ± 0.002 2.54 ± 0.26
Open in a new tab

*GS = genome size; only flow-cytometric analyses with CV of sample peak <3 % were used for computing genome size statistics.

†The number in parentheses indicates for how many samples absolute genome size was estimated using PI flow cytometry.

In the European range of C. fragilis, the area with the most intensive sampling, the distribution of tetraploid and hexaploid cytotypes largely overlaps (Fig. 1C, Supplementary Data Fig. S3). The two cytotypes co-occur over most of Europe in a diffuse, mosaic-like pattern with only a few regions seemingly dominated by one cytotype: tetraploids dominate in Iceland and the Iberian Peninsula whereas hexaploids dominate in the Dinaric and French Alps (Fig. 1C). Ploidy coexistence within populations is very common and accounts for 40 % of the localities where more than ten plants were sampled, whereas uniformly tetraploid and uniformly hexaploid populations constitute 35 % and 25 % of such sites, respectively. Moreover, 26.7 % of populations include tetraploids, pentaploids and hexaploids. Pentaploid individuals were only found in ploidy mixtures, and never formed uniformly pentaploid populations.

To put our results into a broader context, we also sampled 405 individuals of other species from the C. fragilis complex (Fig. 1). In the exclusively European species C. alpina, the vast majority of analysed individuals were hexaploids (87.4 %). However, tetraploids (9.1 %), pentaploids (2.6 %) and octoploids (0.9 %) also occurred, albeit rarely (Fig. 1B, Supplementary Data Table S1). These are the first reports for ploidy levels other than hexaploid in C. alpina. The rarest cytotype, octoploid, was discovered in two mixed-ploidy (4x + 6x and 5x + 6x) populations in Macedonia. The American species C. protrusa and C. reevesiana were uniformly diploid, whereas C. tenuis was tetraploid. Populations of C. diaphana were consistently tetraploid in America but hexaploid in Europe (including Macaronesia). Interestingly, both cytotypes of C. diaphana show conspicuously larger genome sizes in comparison with other taxa of the same ploidy (see Table 1).

Monoploid genome size comparisons

Significant differences in monoploid genome size (i.e. the Cx value) were detected between C. fragilis tetraploids and hexaploids (Kruskal–Wallis test; χ2 = 175.0, d.f. = 1, P < 0.001): that of tetraploids is on average 2.6 % larger than that of hexaploids (mean = 0.1091 ± 0.0001 s.e. and mean = 0.1063 ± 0.0001 s.e., respectively; Fig. 3A). The lowest Cx values for both tetraploids and hexaploids were found in populations from the arctic island of Svalbard, whereas the highest originated in the Mediterranean region (southern Europe and Turkey). However, linear regression models with latitude as predictor failed to explain much of the overall variation in Cx values of either tetraploids (F1,228 = 0.014, P = 0.905, R2 = 0.000) or hexaploids (F1,226 = 3.440, P = 0.065, R2 = 0.015). The populations from Argentina, distinct by both their isolated geographical position (Fig. 1A) and slightly higher Cx values (Table 1), were excluded from all genome-size-related analyses as these could represent a special type.

Fig. 3.

Fig. 3.

Open in a new tab

Differences in monoploid genome size (i.e. relative genome size divided by ploidy level; Cx value) between C. fragilis tetraploids and hexaploids (A). Utilization of monoploid genome size divergence to compare two competing scenarios explaining the origin of pentaploids in mixed-ploidy populations of C. fragilis (B). Using a linear regression model (solid line), we related the Cx value of pentaploids to the mean Cx value of co-occurring tetraploids and hexaploids. In the case of a local independent origin of pentaploids in each mixed-ploidy popuation, the Cx values of response and explanatory variables should be identical (as represented by the dashed line), whereas in the case of immigration of established pentaploids from a distinct population, the Cx values should differ.

Taking into account the Cx value differences between tetra- and hexaploids, we further attempted to use this trait as a clue for reconstructing evolutionary relationships among locally co-occurring cytotypes. Interestingly, while there were no differences in monoploid genome size of tetraploids from mixed-ploidy and uniform-ploidy populations (Kruskal–Wallis test; χ2 = 1.9, d.f. = 1, P = 0.163), hexaploids from mixed-ploidy populations had slightly but significantly higher Cx values than their counterparts from uniform-ploidy populations (mean = 0.1067 ± 0.0002 s.e., mean = 0.1059 ± 0.0002 s.e., respectively; Kruskal–Wallis test: χ2 = 6.6, d.f. = 1, P = 0.010). Mean Cx value of co-occurring tetra- and hexaploids is a very good predictor of Cx value of pentaploids residing in these mixed-ploidy populations (F1,21 = 172.0, P < 0.001, R2 = 0.891; Fig. 3B), consistent with the recurrent origin of pentaploids from inter-ploidy crosses between coexisting tetraploids and hexaploids.

Ecological preferences of C. fragilis cytotypes

The sampled C. fragilis populations exhibited a substantial elevational range, from sea level up to 4670 m a.s.l. in the Himalayas (Supplementary Data Fig. S4), although no difference in mean elevation was found among the cytotypes (F2,426 = 1.67, P = 0.189). Similarly, tetraploid, hexaploid and mixed-ploidy populations were not differentiated by preferential occurrence at habitats of natural (e.g. rock crevices) or anthropogenic (e.g. walls) origin (χ2 = 1.38, d.f. = 2, P = 0.501; Fig. 2A). However, significant differences were observed for substrate preference (χ2 = 16.56, d.f. = 2, P < 0.001; Fig. 2B), with hexaploid populations being more common on alkaline than siliceous substrates compared to both tetraploid (χ2 = 15.32, d.f. = 1, Padj. < 0.001) and mixed-ploidy populations (χ2 = 6.64, d.f. = 1, Padj. = 0.030). Tetraploid and mixed-ploidy populations do not differ significantly in substrate preferences (χ2 = 1.09, d.f. = 1, Padj. = 0.890).

Fig. 2.

Fig. 2.

Open in a new tab

Comparisons of climatic and abiotic niches among uniformly tetraploid, uniformly hexaploid and mixed-ploidy populations of Cystopteris fragilis. Differences in habitat origin, natural vs. anthropogenic (A), and in the type of substrate, alkaline vs. siliceous (B). Lower-case letters indicate significantly different groups in pairwise comparisons using chi-square tests for homogeneity (i.e. differences in relative incidence at particular habitat types are compared between cytotypes). A principal component analysis (C) shows climatic niches of tetraploid, hexaploid and mixed-ploidy populations reconstructed from 19 Bioclim variables. Lines connect the most divergent populations in each group.

Climatic niches of tetraploid, hexaploid and mixed-ploidy C. fragilis populations, as reconstructed using 19 Bioclim variables, showed substantial overlap in a PCA (Fig. 2C). The effect of cytotype composition of populations was not significantly different in a redundancy analysis (P = 0.611, 999 permutations) and both constrained axes explained together only 0.6 % of the variation, whereas the first unconstrained axis in the analysis explained 33.2 % of the variation.

DISCUSSION

Cytotype diversity

By taking advantage of fast and reliable ploidy estimation using flow cytometry, we examined the ploidy level of 5518 individuals from 449 populations across the geographical range of C. fragilis, with particular emphasis on distribution in Europe (Fig. 1). Undoubtedly, this is the most comprehensive cytotype screening ever conducted in ferns (and one of the widest among vascular plants) with respect to both the area covered and the number of sampled individuals. All cytotypes previously reported in natural populations of C. fragilis were discovered (i.e. 4x, 5x, 6x and 8x; Manton, 1950; Vida, 1974; Kawakami et al., 2010), and these have contrasting frequencies of occurrence. Whereas the high frequencies of tetraploids and hexaploids (51 % and 46 % of sampled individuals, respectively) suggest these are well-established cytotypes, pentaploids and octoploids are rare (3 % and 0.02 %, respectively). Our data indicate that the pentaploids originate within mixed-ploidy populations from inter-ploidy crosses between tetraploids and hexaploids. Pentaploids are only found co-occurring with their putative parental cytotypes, and their recurrent hybrid origin is supported by monoploid genome size, as the Cx value of pentaploids can be accurately predicted from the mean of Cx values of the locally co-occurring tetra- and hexaploids (linear model explaining 89 % of the variation, Fig. 3B). In addition, the pentaploids have greatly reduced spore fertility (69–100 % of spores aborted compared to 0–1 % in tetra- and hexaploids, data not shown), which is generally common in odd-level polyploids (Ramsey and Schemske, 1998; Ekrt and Koutecký, 2016). The single octoploid individual could have originated via three possible scenarios involving different combinations of reduced and unreduced gametes of the other cytotypes (i.e. 4x + 4x gametes, 6x + 2x, or 3x + 5x). However, due to only subtle differences in Cx values of the co-occurring putative parental cytotypes, the three pathways cannot be reliably distinguished based on genome size, and the presence of both tetraploids and hexaploids in nearby populations prevents any inference from the regional frequency of parental cytotypes.

Apart from C. fragilis, several other related species were included in our sampling to put our results into context. In C. alpina, a European species which has previously been considered uniformly hexaploid (Manton, 1950; Blasdell, 1963; Vida, 1974), four different ploidy levels were detected (4x, 5x, 6x and 8x). Nonetheless, hexaploids are clearly the dominant cytotype – they constitute 86 % of our C. alpina samples and were found in all but one C. alpina population investigated. The remaining three cytotypes were found only in Macedonia and usually occurred alongside hexaploids in mixed-ploidy populations. Populations of American species C. protrusa and C. reevesiana were uniformly diploid, and populations of C. tenuis were uniformly tetraploid, in line with previous studies (Haufler and Windham, 1991). Populations of C. diaphana were uniformly tetraploid in America but uniformly hexaploid in Europe (including Macaronesia). While the situation in Europe is consistent with Vida (1974), three ploidy levels (2x, 4x and 6x) were previously reported from American populations by Blasdell (1963), and our results might have been influenced by our low number of C. diaphana samples. With the exception of triploids, all cytotypes previously reported in natural populations of the C. fragilis complex were discovered (i.e. 2x, 4x, 5x, 6x and 8x; Manton, 1950; Vida, 1974; Kawakami et al., 2010; Rothfels et al., 2013), although we have not specifically targeted the diploid populations where such triploids would be expected (Haufler et al., 1985). Interestingly, our thorough sampling corroborates the complete lack of diploids in European populations of the C. fragilis complex, as suggested by Blasdell (1963).

Ploidy distribution patterns

The distribution of C. fragilis tetra- and hexaploids shows distinct patterns in North America and Eurasia. Firstly, whereas these two cytotypes are almost equally represented in Eurasia, a striking predominance of tetraploids is apparent in North America, where they are five times more common than their hexaploid counterparts. Also, in North America, hexaploids seem to occur more often in the north and tetraploids predominate in the south, while the distribution of the two cytotypes in Eurasia is less structured (Fig. 1). However, the low number of populations (27 in total) and non-random sampling pattern in North America precluded comprehensive comparison of the spatial structure of the tetraploid–hexaploid co-occurrence and frequency of mixed-ploidy populations between the two continents. In Europe, the tetraploid–hexaploid contact zone of C. fragilis (i.e. the area of ploidy coexistence) is extremely wide and covers most of the continent (Fig. 1C). The contact zone has a diffuse, mosaic-like structure with common incidence of mixed-ploidy populations. This strongly contrasts with substantial spatial isolation of cytotypes documented in many fern species (e.g. Nakato and Kato, 2005; Chang et al., 2013), and when local ploidy coexistence is reported – e.g. in the Asplenium trichomanes complex (Moran, 1982; Ekrt and Štech, 2008; Liu et al., 2018) or the A. ceterach complex (Trewick et al., 2002) – it is usually restricted to certain regions. Moreover, such a cytotype distribution pattern is quite rare even in well-documented mixed-ploidy angiosperms (but see, e.g. Duchoslav et al., 2010; McAllister et al., 2015; Čertner et al., 2017). The frequency of mixed-ploidy populations in C. fragilis is very high (40 % for populations with >10 sampled individuals) and 67 % of 4x + 6x populations also include pentaploids. Comparable frequencies of ploidy mixtures have been documented in only a few angiosperm species (e.g. Gymnadenia conopsea and Andropogon gerardii; Trávníček et al., 2011; McAllister et al., 2015; Kolář et al., 2017). Although multiple cytotypes were previously reported from natural populations of C. fragilis (Vida, 1974; Kawakami et al., 2010), our results demonstrate the frequency and scope of ploidy coexistence across the species’ entire distributional range. The widespread ploidy coexistence makes C. fragilis a convenient model system for studying the microevolutionary mechanisms of polyploid speciation in ferns.

Monoploid genome size divergence

Flow-cytometric analysis has revealed subtle but highly significant differences in monoploid genome size (i.e. Cx value) between C. fragilis tetra- and hexaploids. Specifically, the monoploid genome of hexaploids is on average 2.6 % smaller than that of tetraploids (Fig. 3A). This might be a consequence of ‘genome downsizing’, a process of systematic DNA loss, which is known to commonly accompany polyploidy (Leitch and Bennett, 2004; Tayalé and Parisod, 2013). Some support for this explanation may be provided by two other ploidy-heterogeneous species from the C. fragilis complex, C. alpina and C. diaphana, in which a similar trend of decreasing Cx value with increasing ploidy seems to be present (Table 1). Alternatively, the C. fragilis hexaploids could have originated from unsampled ancestors with lower Cx value. Given that our data were based on flow-cytometric analysis using AT-specific DAPI staining, the observed differences could theoretically be attributed to different base composition (GC content) and not genome size of the two cytotypes. However, when we used base-unspecific PI staining on a selected subset of C. fragilis individuals (Table 1), the Cx value differences between tetra- and hexaploids were retained (F1,38 = 11.5, P = 0.002).

The Cx value, allowing genome size comparisons independent of ploidy level, may in some cases be used to reconstruct cytotype relationships and modes of cytotype origin, and, possibly, to detect recurrent origin of polyploids (Čertner et al., 2017). In this study, we used the Cx value differences as a clue to reconstruct evolutionary relationships among locally co-occurring cytotypes in C. fragilis, although the potential utility of this approach is substantially limited by the very small between-ploidy difference in this species. Nevertheless, we detected significantly higher Cx values of hexaploids from mixed-ploidy populations (i.e. those with values more similar to the Cx value of tetraploids) when compared with their counterparts from uniformly hexaploid populations. The fact that Cx values of tetra- and hexaploids were more similar in sympatry (mixed-ploidy populations) than in allopatry (uniform-ploidy populations) could be explained by recurrent polyploidization, the presence of locally originated hexaploids in some mixed-ploidy populations (i.e. primary cytotype contact; Petit et al., 1999). Alternatively, such a pattern could result from frequent inter-ploidy hybridization and gene flow, possibly involving pentaploids as intermediates. Were this the case, the gene flow would have to be strongly asymmetric to explain significant Cx value changes only in hexaploids and not tetraploids. Interestingly, such strongly asymmetric gene flow from tetra- to hexaploids but not vice versa was previously reported, e.g. in Senecio carniolicus (Hülber et al., 2015).

We also used the Cx value differences between locally co-occurring tetra- and hexaploids to corroborate a local hybrid origin of C. fragilis pentaploids in mixed-ploidy populations (Fig. 3B). Our premise was that with prevailing immigration of pentaploids from different populations (in which they are established and spread out), we would not observe a tight correlation between the Cx value of pentaploids and mean Cx value of residing tetra- and hexaploids in 4x + 5x + 6x populations. Moreover, the Cx value divergence between tetraploid samples from Argentina and the other C. fragilis tetraploids (Table 1) may be a sign that these represent a different, cryptic species. Collectively, our results suggest that Cx value differences may provide interesting insight into microevolutionary processes in ploidy-heterogeneous species. Note, however, that intraspecific genome size variation may occasionally be an artefact caused by technical difficulties (e.g. error of measurement, presence of staining inhibitors in the material, low material quality or improper material storage; Greilhuber, 2005). Here, we are convinced that our results are sufficiently robust, as repeated measurements of selected samples provided highly comparable estimates, fresh leaf material was preferred for C. fragilis (only approx. 30 % of samples were silica-gel dried, pilot tests confirmed estimates highly comparable to analysing fresh material), and we only used high-quality analyses (CV <3 %) for statistical comparisons. Nonetheless, caution is needed when interpreting such data and it should be ideally combined with other approaches (e.g. molecular-genetic analyses) before firm conclusions are reached.

The drivers of ploidy coexistence

Due to the reduced fitness (sterility) of offspring from between-ploidy crosses, plant fitness in mixed-ploidy populations depends strongly on cytotype frequency. According to theoretical models, the less common (minority) cytotypes should spend more of their reproductive effort on ineffective between-ploidy mating, and thus decline in frequency and ultimately become excluded (Levin, 1975; Husband, 2000). However, several factors and ecological processes may facilitate ploidy coexistence in natural populations (reviewed by Kolář et al., 2017), one of the most important being spatial clustering of cytotypes (favouring within-ploidy mating; Baack, 2005), which is commonly driven by their sorting along ecological gradients (Manzaneda et al., 2012; Laport et al., 2016).

However, our results imply a very limited effect of environmental heterogeneity on cytotype distribution patterns in C. fragilis. Tetraploids and hexaploids occupy the same climatic niches and elevational ranges, and show similar preferences for habitat origin (rocks vs. man-made walls). The only sign of ecological differentiation was found in substrate preference, as the relative incidence of uniformly hexaploid populations at alkaline sites was significantly greater than observed in both uniformly tetraploid and mixed-ploidy populations (Fig. 2B). Given that hexaploids from mixed-ploidy populations showed the same substrate preferences as tetraploids, it seems unlikely these differences were caused by inherently distinct ecophysiology of the two cytotypes. Alternatively, the observed association could stem from an interplay between the founder effect, allowing hexaploids to dominate in some regions with alkaline bedrock (e.g. Dinaric Alps, French Alps), and a sampling bias (e.g. unintentionally high sampling intensity in such regions). When statistical comparisons of substrate preferences were restricted to the Central European C. fragilis populations to exclude all tetraploid/hexaploidy-dominated regions, the overall effect of cytotype composition was only marginally significant (χ2 = 6.38, d.f. = 2, P = 0.041) and none of the pairwise comparisons differed significantly after applying the Bonferroni correction. While we cannot rule out some effect of substrate quality on cytotype distribution patterns in C. fragilis, our data do not suggest that it has an important role in maintaining tetraploid–hexaploid coexistence.

We are well aware that our survey of environmental parameters may not have been comprehensive enough to reveal other signs of cytotype ecological differentiation. For example, local environmental conditions in the sampled populations could be substantially better described if we employed chemical soil analysis and direct measurements of microclimatic conditions using temperature and moisture probes. Unfortunately, the applied methodology was largely a trade-off caused by the scope of our study (nearly global-scale sampling) and the very high number of sampled populations. Nonetheless, the lack of cytotype ecological differentiation is also indirectly supported by the ploidy distribution patterns (e.g. range-wide ploidy coexistence, high incidence of mixed-ploidy populations). While we cannot rule out a fine-scale ecological segregation of cytotypes within mixed-ploidy populations, during our field sampling, we did not notice any signs of greater microhabitat diversity in mixed-ploidy compared to uniform-ploidy populations. No ecological differences among cytotypes were also documented in other ploidy-variable plant species (Buggs and Pannell, 2007; Glennon et al., 2014; Hanzl et al., 2014), suggesting that ecological segregation of cytotypes may be a common scenario rather than the rule.

The widespread and frequent ploidy coexistence, however, can be explained by other mechanisms. In general, plant longevity and the ability to reproduce clonally may partially mitigate the effect of minority cytotype exclusion (Yamauchi et al., 2004). In a recent review of cytotype diversity among angiosperms, asexual reproduction resulted in a nearly two-fold increase in the frequency of mixed-ploidy populations and also contributed to the abundance of odd-ploidy cytotypes (Kolář et al., 2017). As with most ferns, C. fragilis is a long-lived rhizomatous perennial, and even though the clonal spread of the species is limited and probably locally restricted (Hovenkamp, 1990), these traits could facilitate ploidy coexistence within populations and also favour the persistence of largely sterile pentaploid hybrids suffering from meiotic irregularities. Additionally, several other mechanisms specific to the reproductive biology of ferns could promote local cytotype coexistence. First, the movement of spermatozoids among gametophytes is more spatially restricted than pollen transfer in angiosperms, especially in Cystopteris, where suitable microhabitats (e.g. crevices in walls and rock faces) are often patchily distributed. Secondly, gametophytic selfing documented in the genus Cystopteris is likely to occur and may serve as an important mechanism of reproductive assurance for minority cytotypes (Sessa et al., 2016). Lastly, and probably most importantly, efficient wind dispersal of spores might often bring the two cytotypes together and facilitate ploidy coexistence, either by founding new mixed-ploidy populations or by supplying immigrants to the existing ones.

CONCLUSIONS

Our study has provided the first detailed insight into ploidy distribution patterns in populations of a fern species. Contrary to the substantial spatial isolation of cytotypes documented in most ferns, mosaic-like structure of the 4x–6x contact zone in C. fragilis favours ploidy coexistence even within populations and makes it a common and widespread phenomenon across the entire distributional range of the species. Both ploidy-level diversity and the frequency of mixed-ploidy populations observed here suggest that in this respect ferns can match the well-documented patterns in angiosperms (Kolář et al., 2017).

We also focused on possible evolutionary drivers of common coexistence of cytotypes in this species. Because no ecological constraints to ploidy coexistence were detected, the local co-occurrence of tetra- and hexaploids seems to be possible across the entire range of environmental conditions suitable for C. fragilis. Persistence of local ploidy mixtures could be facilitated by the perennial life-form of C. fragilis, its reproductive modes (occasional clonal spread and gametophytic selfing) and efficient wind dispersal of spores (founding new mixed-ploidy populations or supplying immigrants to existing ones). Moreover, independent origins of hexaploids and/or inter-ploidy gene flow may be expected in mixed-ploidy populations of C. fragilis as suggested by Cx value comparisons.

SUPPLEMENTARY DATA

Supplementary data are available online at https://academic.oup.com/aob and consist of the following. Table S1: Complete list of sampled localities. Table S2: Datasets used for ecological niche and genome size comparisons. Fig. S1: Flow-cytometric histogram of the relative DNA content in simultaneous analysis of DAPI-stained nuclei isolated from tetraploid and hexaploid C. fragilis plants, with the internal standard Vicia faba ‘Inovec’. Fig. S2: Microphotograph of chromosomes at meiosis of tetraploid C. fragilis. Fig. S3: Detail of cytotype distribution patterns in Central European populations of C. fragilis. Fig. S4: Overlapping elevational ranges of uniformly tetraploid, uniformly hexaploid and mixed-ploidy populations of C. fragilis.

mcy219_Suppl_Supplementary-Data-Table_S1
Click here for additional data file. (204KB, xls)
mcy219_Suppl_Supplementary-Data-Table_S2
Click here for additional data file. (444KB, xlsx)
mcy219_Suppl_Supplementary-Data-Figure_S1
Click here for additional data file. (12.3KB, png)
mcy219_Suppl_Supplementary-Data-Figure_S2
Click here for additional data file. (1.1MB, jpeg)
mcy219_Suppl_Supplementary-Data-Figure_S3
Click here for additional data file. (1.4MB, jpeg)
mcy219_Suppl_Supplementary-Data-Figure_S4
Click here for additional data file. (155.5KB, jpeg)

ACKNOWLEDGEMENTS

We thank Scottish Natural Heritage organization in Inverness for a permit (licence no. 36412) to collect Cystopteris in Scotland. We are also indebted to numerous colleagues who helped with field material collection: R. Bartošová, C. Björk, G. Blanca, S. Březina, J. Chrtek, Z. Chumová, M. Fialová, T. Hájek, J. Hadinec, M. Hanzlíková, I. Hejlová, A. Hilpold, F. Holič, B. Jonášová, K. Kabátová, A. Knotek, J. Kocková, F. Kolář, T.-T. Kuo, P. Kúr, M. Lučanová, V. Mach, H. Mašterová, H. Přívozníková, P. Robovská, J. Robovský, P. Schönswetter, T. Skalický, S. F. Smith, T. Staughton, V. N. Suaréz-Santiago, J. Suda, M. Sundue, K. Šemberová, M. Štech, M. Štefánek, P. Tájek, W. Testo, P. Trávníček, P. Vít, B. Weiss. This study was financially supported by the Grant Agency of Charles University (project GAUK 1556214; K.H., T.U.), Charles University Research Centre program No. 204069 (T.U.) and the Czech Science Foundation project No. 14-36079G, Centre of Excellence PLADIAS (L.E., P.K.). Additional support was provided by the Czech Academy of Sciences (long-term research development project no. RVO 67985939; M.C., V.J.).

LITERATURE CITED

  1. Baack EJ. 2005. To succeed globally, disperse locally: effects of local pollen and seed dispersal on tetraploid establishment. Heredity 94: 538–546. [DOI] [PubMed] [Google Scholar]
  2. Barker M. 2013. Karyotype and genome evolution in pteridophytes. In: Greilhuber J, Doležel J, Wendel JF eds. Plant genome diversity, vol. 2 Vienna: Springer-Verlag, 245–353. [Google Scholar]
  3. Blasdell RF. 1963. A monographic study of the fern genus Cystopteris. Memoirs of the Torrey Botanical Club 21: 1–102. [Google Scholar]
  4. Buggs RJA, Pannell JR. 2007. Ecological differentiation and diploid superiority across a moving ploidy contact zone. Evolution 61: 125–140. [DOI] [PubMed] [Google Scholar]
  5. Čertner M, Fenclová E, Kúr P, et al. 2017. Evolutionary dynamics of mixed-ploidy populations in an annual herb: dispersal, local persistence and recurrent origins of polyploids. Annals of Botany 120: 303–315. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Chang Y, Li J, Lu S, Schneider H. 2013. Species diversity and reticulate evolution in the Asplenium normale complex (Aspleniaceae) in China and adjacent areas. Taxon 62: 673–687. [Google Scholar]
  7. Chrtek J, Herben T, Rosenbaumová R, et al. 2017. Cytotype coexistence in the field cannot be explained by inter-cytotype hybridization alone: linking experiments and computer simulations in the sexual species Pilosella echioides (Asteraceae). BMC Evolutionary Biology 17: 87. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Clark J, Hidalgo O, Pellicer J, et al. 2016. Genome evolution of ferns: evidence for relative stasis of genome size across the fern phylogeny. New Phytologist 210: 1072–1082. [DOI] [PubMed] [Google Scholar]
  9. Coyne JA, Orr HA. 2004. Speciation. Sunderland, MA: Sinauer Associates, Inc. [Google Scholar]
  10. Dauphin B, Grant JR, Farrar DR, Rothfels CJ. 2018. Rapid allopolyploid radiation of moonwort ferns (Botrychium; Ophioglossaceae) revealed by PacBio sequencing of homologous and homeologous nuclear regions. Molecular Phylogenetics and Evolution 120: 342–353. [DOI] [PubMed] [Google Scholar]
  11. Doležel J, Sgorbati S, Lucretti S. 1992. Comparison of three DNA fluorochromes for flow cytometric estimation of nuclear DNA content in plants. Physiologia Plantarum 85: 625–631. [Google Scholar]
  12. Doležel J, Geilhuber J, Suda J. 2007. Estimation of nuclear DNA content in plants using flow cytometry. Nature Protocols 2: 2233–2244. [DOI] [PubMed] [Google Scholar]
  13. Duchoslav M, Šafářová L, Krahulec F. 2010. Complex distribution patterns, ecology and coexistence of ploidy levels of Allium oleraceum (Alliaceae) in the Czech Republic. Annals of Botany 105: 719–735. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Ekrt L, Koutecký P. 2016. Between sexual and apomictic: unexpectedly variable sporogenesis and production of viable polyhaploids in the pentaploid fern of the Dryopteris affinis agg. (Dryopteridaceae). Annals of Botany 117: 97–106. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Ekrt L, Štech M. 2008. A morphometric study and revision of the Asplenium trichomanes group in the Czech Republic. Preslia 80: 325–347. [Google Scholar]
  16. Ekrt L, Holubová R, Trávníček P, Suda J. 2010. Species boundaries and frequency of hybridization in the Dryopteris carthusiana (Dryopteridaceae) complex: a taxonomic puzzle resolved using genome size data. American Journal of Botany 97: 1208–1219. [DOI] [PubMed] [Google Scholar]
  17. Felber F. 1991. Establishment of a tetraploid cytotype in a diploid population: effect of relative fitness of the cytotypes. Journal of Evolutionary Biology 4: 195–207. [Google Scholar]
  18. Fraser-Jenkins CR. 2008. Taxonomic revision of three hundred Indian subcontinental pteridophytes: with a revised census list; a new picture of fern-taxonomy and nomenclature in the Indian subcontinent. Bishen Singh Mahendra Pal Singh. [Google Scholar]
  19. Gämperle E, Schneller JJ. 2002. Phenotypic and isozyme variation in Cystopteris fragilis (Pteridophyta) along an altitudinal gradient in Switzerland. Flora 197: 203–213. [Google Scholar]
  20. Glennon KL, Ritchie ME, Segraves KA. 2014. Evidence for shared broad‐scale climatic niches of diploid and polyploid plants. Ecology Letters 17: 574–582. [DOI] [PubMed] [Google Scholar]
  21. Greilhuber J. 2005. Intraspecific variation in genome size in angiosperms: identifying its existence. Annals of Botany 95: 91–98. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Grusz AL, Windham MD, Yatskievych G, Huiet L, Gastony GJ, Pryer KM. 2014. Patterns of diversification in the xeric-adapted fern genus Myriopteris (Pteridaceae). Systematic Botany 39: 698–714. [Google Scholar]
  23. Hanzl M, Kolář F, Nováková D, Suda J. 2014. Nonadaptive processes governing early stages of polyploid evolution: insights from a primary contact zone of relict serpentine Knautia arvensis (Caprifoliaceae). American Journal of Botany 101: 935–945. [DOI] [PubMed] [Google Scholar]
  24. Haufler CH, Windham MD. 1991. New species of North American Cystopteris and Polypodium, with comments on their reticulate relationships. American Fern Journal 81: 7–23. [Google Scholar]
  25. Haufler CH, Windham MD, Britton DM, Robinson SJ. 1985. Triploidy and its evolutionary significance in Cystopteris protrusa. Canadian Journal of Botany 63: 1855–1863. [Google Scholar]
  26. Haufler CH, Moran RC, Windham MD. 1993. Cystopteris. Flora of North America, Vol. 2 New York: Oxford University Press, 263–270. [Google Scholar]
  27. Hijmans RJ, Cameron SE, Parra JL, Jones PG, Jarvis A. 2005. Very high resolution interpolated climate surfaces for global land areas. International Journal of Climatology 25: 1965–1978. [Google Scholar]
  28. Hovenkamp P. 1990. The significance of rhizome morphology in the systematics of the Polypodiaceous ferns (sensu stricto). American Fern Journal 80: 33–43. [Google Scholar]
  29. Hülber K, Sonnleitner M, Suda J, et al. 2015. Ecological differentiation, lack of hybrids involving diploids, and asymmetric gene flow between polyploids in narrow contact zones of Senecio carniolicus (syn. Jacobaea carniolica, Asteraceae). Ecology and Evolution 5: 1224–1234. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Husband BC. 2000. Constraints on polyploid evolution: a test of the minority cytotype exclusion principle. Proceedings of the Royal Society B: Biological Sciences 267: 217–223. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Husband BC, Ozimec B, Martin SL, Pollock L. 2008. Mating consequences of polyploid evolution in flowering plants: current trends and insights from synthetic polyploids. International Journal of Plant Sciences 169: 195–206. [Google Scholar]
  32. Kawakami SM, Kawakami S, Kato J, Kondo K, Smirnov SV, Damdinsuren O. 2010. Cytological study of a fern Cystopteris fragilis in Mongolian Altai. Chromosome Botany 5: 1–3. [Google Scholar]
  33. Kolář F, Čertner M, Suda J, Schönswetter P, Husband BC. 2017. Mixed-ploidy species: progress and opportunities in polyploid research. Trends in Plant Science 22: 1041–1055. [DOI] [PubMed] [Google Scholar]
  34. Kreiner JM, Kron P, Husband BC. 2017. Evolutionary dynamics of unreduced gametes. Trends in Genetics 33: 583–593. [DOI] [PubMed] [Google Scholar]
  35. Landis JB, Soltis DE, Li Z, Marx HE, Barker MS, Tank DC, Soltis PS. 2018. Impact of whole-genome duplication events on diversification rates in angiosperms. American Journal of Botany 105: 348–363. [DOI] [PubMed] [Google Scholar]
  36. Laport RG, Minckley RL, Ramsey J. 2016. Ecological distributions, phenological isolation, and genetic structure in sympatric and parapatric populations of the Larrea tridentata polyploid complex. American Journal of Botany 103: 1358–1374. [DOI] [PubMed] [Google Scholar]
  37. Leitch IJ, Bennett MD. 2004. Genome downsizing in polyploid plants. Biological Journal of the Linnean Society 82: 651–663. [Google Scholar]
  38. Levin DA. 1975. Minority cytotype exclusion in local plant populations. Taxon 24: 35–43. [Google Scholar]
  39. Levin DA. 2002. The role of chromosomal change in plant evolution. Oxford: Oxford University Press. [Google Scholar]
  40. Liu HM, Russell SR, Vogel J, Schneider H. 2018. Inferring the potential of plastid DNA-based identification of derived ferns: a case study on the Asplenium trichomanes aggregate in Europe. Plant Systematics and Evolution 304: 1009–1022. [Google Scholar]
  41. Manton I. 1950. Problems of cytology and evolution in the Pteridophyta. Cambridge: Cambridge University Press. [Google Scholar]
  42. Manzaneda AJ, Rey PJ, Bastida JM, Weiss‐Lehman C, Raskin E, Mitchell‐Olds T. 2012. Environmental aridity is associated with cytotype segregation and polyploidy occurrence in Brachypodium distachyon (Poaceae). New Phytologist 193: 797–805. [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. McAllister C, Blaine R, Kron P, et al. 2015. Environmental correlates of cytotype distribution in Andropogon gerardii (Poaceae). American Journal of Botany 102: 92–102. [DOI] [PubMed] [Google Scholar]
  44. Moran RC. 1982. The Asplenium trichomanes complex in the United States and adjacent Canada. American Fern Journal 72: 5–11. [Google Scholar]
  45. Nakato N, Kato M. 2005. Cytogeography of the Adiantum pedatum complex (Pteridaceae, Subfamily Adiantoideae). Acta Phytotaxonomica et Geobotanica 56: 85–96. [Google Scholar]
  46. Parks JC, Dyer AF, Lindsay S. 2000. Allozyme, spore and frond variation in some Scottish populations of the ferns Cystopteris dickieana and Cystopteris fragilis. Edinburgh Journal of Botany 57: 83–105. [Google Scholar]
  47. Petit C, Lesbros P, Ge X, Thompson JD. 1997. Variation in flowering phenology and selfing rate across a contact zone between diploid and tetraploid Arrhenatherum elatius (Poaceae). Heredity 79: 31–40. [Google Scholar]
  48. Petit C, Bretagnolle F, Felber F. 1999. Evolutionary consequences of diploid–polyploid hybrid zones in wild species. Trends in Ecology & Evolution 14: 306–311. [DOI] [PubMed] [Google Scholar]
  49. PPG I 2016. A community-derived classification for extant lycopods and ferns. Journal of Systematics and Evolution 54: 563–603. [Google Scholar]
  50. R Core Team 2016. R: A language and environment for statistical computing. Vienna: R Foundation for Statistical Computing. [Google Scholar]
  51. Ramsey J. 2007. Unreduced gametes and neopolyploids in natural populations of Achillea borealis (Asteraceae). Heredity 98: 143–150. [DOI] [PubMed] [Google Scholar]
  52. Ramsey J, Schemske DW. 1998. Pathways, mechanisms, and rates of polyploid formation in flowering plants. Annual Review of Ecology and Systematics 29: 467–501. [Google Scholar]
  53. Rothfels CJE. 2012. Phylogenetics of Cystopteridaceae: reticulation and divergence in a cosmopolitan fern family. PhD Thesis, Duke University, Durham. [Google Scholar]
  54. Rothfels CJ, Windham MD, Pryer KM. 2013. A plastid phylogeny of the cosmopolitan fern family Cystopteridaceae (Polypodiopsida). Systematic Botany 38: 295–306. [Google Scholar]
  55. Rothfels CJ, Johnson AK, Windham MD, Pryer KM. 2014. Low-copy nuclear data confirm rampant allopolyploidy in the Cystopteridaceae (Polypodiales). Taxon 63: 1026–1036. [Google Scholar]
  56. Rothfels CJ, Pryer KM, Li F-W. 2017. Next-generation polyploid phylogenetics: rapid resolution of hybrid polyploid complexes using PacBio single-molecule sequencing. New Phytologist 213: 413–429. [DOI] [PubMed] [Google Scholar]
  57. Schneider H, Liu HM, Chang YF, et al. 2017. Neo‐ and paleopolyploidy contribute to the species diversity of Asplenium–the most species‐rich genus of ferns. Journal of Systematics and Evolution 55: 353–364. [Google Scholar]
  58. Sessa EB, Testo WL, Watkins JE. 2016. On the widespread capacity for, and functional significance of, extreme inbreeding in ferns. New Phytologist 211: 1108–1119. [DOI] [PubMed] [Google Scholar]
  59. Shinohara W, Ushio Y, Seo A, et al. 2010. Evidence for hybrid origin and segmental allopolyploidy in eutetraploid and aneutetraploid Lepisorus thunbergianus (Polypodiaceae). Systematic Botany 35: 20–29. [Google Scholar]
  60. Soltis DE, Soltis PS, Tate JA. 2003. Advances in the study of polyploidy since plant speciation: research review. New Phytologist 161: 173–191. [Google Scholar]
  61. Soltis DE, Visger CJ, Marchant DB, Soltis PS. 2016. Polyploidy: pitfalls and paths to a paradigm. American Journal of Botany 103: 1223–1235. [DOI] [PubMed] [Google Scholar]
  62. Tayalé A, Parisod C. 2013. Natural pathways to polyploidy in plants and consequences for genome reorganization. Cytogenetic and Genome Resources 140: 79–96. [DOI] [PubMed] [Google Scholar]
  63. ter Braak C, Šmilauer P. 2012. Canoco reference manual and user’s guide: software for ordination, version 5.0. Ithaca: Microcomputer Power. [Google Scholar]
  64. Trávníček P, Kubátová B, Čurn V, et al. 2011. Remarkable coexistence of multiple cytotypes of the Gymnadenia conopsea aggregate (the fragrant orchid): evidence from flow cytometry. Annals of Botany 107: 77–87. [DOI] [PMC free article] [PubMed] [Google Scholar]
  65. Trewick SA, Morgan-Richards M, Russell SJ, et al. 2002. Polyploidy, phylogeography and Pleistocene refugia of the rockfern Asplenium ceterach: evidence from chloroplast DNA. Molecular Ecology 11: 2003–2012. [DOI] [PubMed] [Google Scholar]
  66. Vida G. 1974. Genome analysis of the European Cystopteris fragilis complex: 1. Tetraploid taxa. Acta Botanica Academiae Scientiarum Hungaricae 20: 181–192. [Google Scholar]
  67. Wood TE, Takebayashi N, Barker MS, Mayrose I, Greenspoon PB, Rieseberg LH. 2009. The frequency of polyploid speciation in vascular plants. Proceedings of the National Academy of Sciences USA 106: 13875–13879. [DOI] [PMC free article] [PubMed] [Google Scholar]
  68. Yamauchi A, Hosokawa A, Nagata H, Shimoda M. 2004. Triploid bridge and role of parthenogenesis in the evolution of autopolyploidy. The American Naturalist 164: 101–112. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

mcy219_Suppl_Supplementary-Data-Table_S1
Click here for additional data file. (204KB, xls)
mcy219_Suppl_Supplementary-Data-Table_S2
Click here for additional data file. (444KB, xlsx)
mcy219_Suppl_Supplementary-Data-Figure_S1
Click here for additional data file. (12.3KB, png)
mcy219_Suppl_Supplementary-Data-Figure_S2
Click here for additional data file. (1.1MB, jpeg)
mcy219_Suppl_Supplementary-Data-Figure_S3
Click here for additional data file. (1.4MB, jpeg)
mcy219_Suppl_Supplementary-Data-Figure_S4
Click here for additional data file. (155.5KB, jpeg)

Articles from Annals of Botany are provided here courtesy of Oxford University Press

RESOURCES