SEARCH

SEARCH BY CITATION

Keywords:

  • allozymes;
  • Cyrtomium falcatum ;
  • homosporous fern;
  • mating system;
  • spatial autocorrelation analysis;
  • spatial demographic structure;
  • spatial genetic structure;
  • spore dispersal

Summary

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  • Spores of homosporous ferns are small, wind-borne and thus have the potential for long-distance dispersal. This common perception has led to a prediction of near-random spatial genetic structure within fern populations. Spore dispersal and spore bank studies, however, indicate that most spores fall close to the maternal plant (< 5 m), supporting a prediction of significant fine-scale genetic structure (FSGS) within populations.
  • To determine which of these two hypotheses is more likely to occur in nature, we measured inbreeding and quantified the spatial distribution of individuals and allozyme-based genotypes using spatial autocorrelation methods within four populations of the fern Cyrtomium falcatum in southern South Korea.
  • Inbreeding levels were low, and all populations exhibited significant aggregation of individuals and strong FSGS.
  • The present results support the second hypothesis, and the substantial FSGS in Cfalcatum could reflect the unique features of most homosporous ferns (outcrossing mating systems that lead a majority of spores to occur at short distances and a very limited dispersal distance of male gametes). Although fern spores are physically analogous to orchid seeds, the intensity of FSGS exhibited in Cfalcatum is four times stronger than that in 16 terrestrial orchid species.

Introduction

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

In seed plants, fine-scale genetic structure (FSGS) occurs when the spatial distribution of genetic variation among individuals within populations is nonrandom. In an attempt to understand the factors responsible for the development and maintenance of such spatial genetic structure within populations of seed plants, > 300 empirical studies have been published during the past two decades or so. Many studies have revealed that populations of seed plants have substantial kinship structure, which is known to be influenced by a number of evolutionary and ecological processes, including limited gene dispersal (Wright, 1943; Sokal & Jacquez, 1991), adult density (Hamrick et al., 1993), thinning among cohorts (Berg & Hamrick, 1995), spatial and temporal patterns of seedling establishment (Ellstrand, 1992), colonization and disturbance history (Epperson & Chung, 2001; Chung & Nason, 2007), microenvironmental selection (Linhart et al., 1981; Kalisz et al., 2001) and successional stage (Troupin et al., 2006; Chung et al., 2007). Among these factors, probably the most widely studied determinant of FSGS is the pattern of seed dispersal (Hamrick & Nason, 1996).

In part analogous to the seeds of flowering plants, spores of ferns are compact, durable dispersal units (Dyer, 1994). In particular, the small-sized spores (c. 0.02–0.13 mm; Knobloch, 1969; Makgomol, 2006) are very similar to the tiny, dust-like seeds of orchids (c. 0.05–6.00 mm; Arditti & Ghani, 2000) with regard to size. As a result, they may have high potential for wind-assisted dispersal, as evidenced by the early arrival of ferns on oceanic islands (Ranker et al., 1994). Some studies have revealed that viable spores exist in soils at several to tens of kilometres from the nearest source (During & Ter Horst, 1983; Ramírez-Trejo et al., 2004), and others have suggested several hundred to several thousand kilometres as putatively maximum spore dispersal distances (Tryon, 1970, 1972; Conant, 1978; Perrie et al., 2010). In particular, the effective population size of the clonal, weedy fern Pteridium aquilinum is the whole British island (Wolf et al., 1991). As is the case for orchid seeds (Jersáková & Malinová, 2007), however, the majority of spores fall in the immediate vicinity of the parent sporophytes (Peck et al., 1990; Dyer, 1994), suggesting that spore dispersal might also be leptokurtic, in a similar fashion to the pattern found in many flowering plants, including orchids (Fay et al., 2009). Furthermore, as fern spermatozoids require transport in water, the dispersal distance of male gametes in ferns tends to be very limited (within a few centimetres; Peck et al., 1990). As a result, for many fern species, gene dispersal is restricted to short distances by the requirement that two spores, and the gametophytes growing from them, be sufficiently close (within a few centimetres) to permit cross-fertilization (Peck et al., 1990; D. R. Farrar, Iowa State University, Ames, pers. comm.). Many fern species have these unique suites of life history traits, in contrast with seed plants in which gene dispersal is a two-step process: pollen dispersal and seed production.

Gene dispersal could be further restricted in some homosporous ferns exhibiting intragametophytic self-fertilization, such as members of Ophioglossaceae whose individual gametophytes regularly become bisexual and sporophytes are produced through intragametophytic selfing (i.e. fertilization of the egg by sperm from the same gametophyte; Wagner et al., 1985; Goswami, 1987). This is supported by allozyme studies, which show that substantial inbreeding could occur through self-fertilization via intragametophytic selfing (McCauley et al., 1985; Soltis & Soltis, 1986; Watano & Sahashi, 1992; Farrar, 1998; Chung et al., 2010). This being the case, successful gene dispersal in these species should be limited by the distance of spore travel (i.e. one-step gene dispersal) and the probability of a viable spore landing in a suitable habitat.

Many homosporous ferns share the same life history features. These include the high potential for spore dispersal, but limited dispersal of most spores around the maternal sporophytes (Peck et al., 1990; Dyer, 1994), outcrossing mating systems that require two spores to land in the same safe site and to successfully produce two genetically different gametophytes (Haufler & Soltis, 1984), and highly restricted sperm movement (Peck et al., 1990). Accordingly, two alternative scenarios can be hypothesized for the patterns of spatial demographic structure and FSGS within populations of homosporous fern species. If spores disperse at random within populations (i.e. high potential for spore dispersal), near-random distribution of individuals and little FSGS would be expected (Jiménez et al., 2010). Conversely, if most spores fall around maternal sporophytes and male gametes move, at most, a few centimetres, we should expect significant aggregation of individuals and FSGS within fern populations (Murakami et al., 1997). To date, however, only two empirical studies are available on FSGS in homosporous ferns, which reach opposite conclusions. Thus, more studies on demographic structure and FSGS are needed to test these hypotheses.

In this study, we selected a nonclonal homosporous fern Cyrtomium falcatum (Dryopteridaceae). To test which of the two hypothesized scenarios is more likely, we first determined the levels of genetic diversity and inbreeding using allozyme markers, and then quantified the spatial aggregation of plants using the O-ring statistic of Wiegand & Moloney (2004) and genetic autocorrelation using the methods of Kalisz et al. (2001) and Vekemans & Hardy (2004). Because both methods are annulus based and nonaccumulative, they permit direct comparisons of the spatial scale and magnitude of the distribution of individuals and genetic structure within populations.

Materials and Methods

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Study species

Cyrtomium falcatum (L.f) C. Presl is an evergreen fern that usually grows on coastal rocky slopes in the warmer parts of south to northeastern Asia (India, Vietnam, eastern and southern China, Taiwan, southern Korean Peninsula and Japan; Iwatsuki, 1992). However, it has become naturalized in many parts of the world (including Hawaii, North America, Australia, western and southern Europe, Réunion Island and South Africa) after escaping from gardens (Roux, 2011). The species, 10–60 cm tall, has a short, erect rhizome, and thus it is highly likely that proximally located individuals within populations are distinct genets (because of the absence of long, horizontal connecting rhizomes). In southern South Korea, C. falcatum usually grows on crevices in steep cliffs, rocks and man-made vertically oriented stone walls near seashores, and thus populations occur discontinuously. The chromosome number of C. falcatum is 2n = 82 (diploid; Lu et al., 2006), although both diploid and tetraploid individuals occur in Japan (Iwatsuki, 1992).

Study populations and sampling

To determine spatial demographic structure and FSGS, we mapped and sampled a total of 263 individuals from four populations of C. falcatum located on Gojae Island (CF-1 and CF-2) and Oenaro Island (CF-3 and CF-4) off the southern Korean Peninsula (Fig. 1). The linear distance between Gojae Island and Oenaro Island is c. 100 km. As areas of the islands containing the sampled populations are well preserved as national parks, the choice of these populations minimizes the potential effects of human-mediated disturbance (e.g. collection) on population structure. Populations CF-1 (n = 46 in 30 × 14 m2) and CF-2 (n = 65 in 10 × 4 m2) are separated by c. 4.9 km, and population CF-3 (n = 58 in 17 × 10 m2) is located c. 4.5 km northwest of CF-4 (n = 94 in 24 × 6 m2; Fig. 2). Individuals in CF-1 and CF-3 grow in crevices in nearly vertically oriented rocks, those in CF-2 occur on small gaps in artificial stone walls that were constructed to make a paddy field, and individuals in CF-4 grow on open, vertically oriented edges of hills along road sides. When mapping the individuals, we assumed that these fine-scale natural geographical features on which C. falcatum grows are vertical (i.e. two-dimensional). To minimize damage to plants, one leaf segment (pinna) from each individual was collected for allozyme analysis of genetic diversity and FSGS (see later). All sampled leaf tissue was kept on ice until its transportation to the laboratory, where it was stored at 4°C until enzyme extraction.

image

Figure 1. Map showing locations of the four sampled populations of Cyrtomium falcatum on Geojae Island (CF-1 and CF-2) and Oenaro Island (CF-3 and CF-4) in southern South Korea.

Download figure to PowerPoint

image

Figure 2. Spatial distribution of Cyrtomium falcatum at CF-1 (n = 46), CF-2 (n = 65), CF-3 (n = 58) and CF-4 (n = 94). (See Fig. 1 for locations of the sampled populations.)

Download figure to PowerPoint

Allozyme electrophoresis

We extracted enzymes by finely cutting leaf samples, adding an extraction buffer (Mitton et al., 1979) and then crushing them with a mortar and pestle. Enzyme extracts were absorbed onto chromatography wicks and stored in microtitre plates in an ultra-cold (−70°C) freezer until analysis. We conducted electrophoresis on 13% starch gels, with three buffer systems. We used a modification (Haufler, 1985) of system 6 of Soltis et al. (1983) to resolve alcohol dehydrogenase (Adh), diaphorase (Dia-1, Dia-2), fluorescent esterase (Fe-1, Fe-2) and cathodal peroxidase (Cpx). We used system 11 of Soltis et al. (1983) to resolve glyceraldehyde-3-phosphate dehydrogenase (G-3-pdh-1, G-3-pdh-2), hexokinase (Hk-1, Hk-2), isocitrate dehydrogenase (Idh), phosphoglucoisomerase (Pgi-1, Pgi-2), phosphoglucomutase (Pgm-1, Pgm-2, Pgm-3) and shikimate dehydrogenase (Skdh). In addition, we used the morpholine-citrate buffer system (pH 6.1) of Clayton & Tretiak (1972) to resolve fructose-1,6-diphosphatase (F1,6) and malate dehydrogenase (Mdh-1, Mdh-2, Mdh-3). We followed stain recipes from Soltis et al. (1983), except for diaphorase (Cheliak & Pitel, 1984). By inferring the genetic basis of the observed enzyme banding patterns from typical subunit structure and subcellular compartmentalization (Weeden & Wendel, 1989), we designated putative loci sequentially, with the most anodally migrating isozyme designated as 1, the next 2, and so on. Different alleles within each locus were numbered sequentially, giving the most anodally migrating alleles the alphabetically lowest letter. Among the 21 allozyme loci assayed, only three loci, Fe-2, Pgm-2 and Pgm-3, were polymorphic across all four populations, which were those used for the analysis of FSGS.

Genetic diversity and structure

To estimate genetic diversity and structure, we considered a locus to be polymorphic if two or more alleles were observed, regardless of their frequencies. The following genetic diversity parameters were estimated using the programs POPGENE (Yeh et al., 1999) and FSTAT (Goudet, 1995): percentage polymorphic loci (%P), mean number of alleles per locus (A), observed heterozygosity (Ho) and Hardy–Weinberg (H–W) expected heterozygosity or Nei's (1978) gene diversity (He).

To calculate individual population-level FIS (inbreeding) and its significance level by 999 permutations under the null hypothesis of FIS = 0, we used the program SPAGeDi (Hardy & Vekemans, 2002). To measure deviations from H–W equilibrium at the multi-population level, we calculated Wright's (1965) FIS and FST over loci for the set of four populations following the method of Weir & Cockerham (1984). These fixation indices measure the average deviation from H–W equilibrium of individuals relative to their local populations (FIS, a measure of local inbreeding) and local populations relative to the total population (FST, also a measure of differentiation among local populations). Using FSTAT, we constructed 95% bootstrap confidence intervals (CIs; 999 replicates) around means of multi-population-level estimates of FIS and FST, and considered the observed FIS and FST to be significant when 95% CI did not overlap zero.

Spatial distribution of individuals

To assess the spatial distribution of mapped plants in the four C. falcatum populations (Fig. 2), we calculated the univariate O-ring statistic O(r) (Wiegand & Moloney, 2004) from the mean number of individuals in an annulus of radius (r) around each plant, and plotted this against the spatial scale r at the starting ring width 1 m with a 1-m lag. Because it is annulus based, the O-ring statistic provides a measure of physical clustering that is on the same spatial scale as the analysis of FSGS. As the use of ring widths greater than one-half of the shortest plot side introduces bias as a result of edge effects, the maximal ring width was set at one-half of the shortest plot width. To test the significance of O(r) for each r, we used the common null model of complete spatial randomness (CSR). For reference to the point pattern expected under CSR, we calculated the first-order intensity λ. For each study population, 95% CI about CSR (i.e. λ) for a given r was constructed from the 25th and 975th highest of the ordered O(r) from 1000 spatial randomizations (999 replicates by Monte Carlo simulation). An observed value of O(r) outside of this envelope was judged to be a significant departure from CSR, with an observed value above, within or below the envelope indicating spatial clustering, spatial randomness or spatial repulsion (hyperdispersion), respectively, at radius r. We conducted all calculations and simulations using the program PROGRAMITA (Wiegand, 2003).

Fine-scale genetic structure

To characterize FSGS within populations, we conducted spatial autocorrelation procedures by calculating the pairwise kinship coefficient between individuals i and j located distance interval r apart (Fij; Loiselle et al., 1995; Kalisz et al., 2001). Like the O-ring statistic, these procedures are annulus based and noncumulative, and thus permit the isolation of the strength of genetic structure for specific distance classes (Chung & Nason, 2007; Chung et al., 2011). To visualize FSGS, we calculated the mean Fij for each distance interval and plotted this against the physical distance. To strike an acceptable balance between spatial resolution and statistical power, we calculated mean Fij estimates at r = 1 m at the shortest distance, then at r = 3 or 5 m, and at various distance intervals up to 14 m (CF-2) and 40 m (CF-1). For each distance interval, 95% CIs were constructed about the null hypothesis of no genetic structure (Fij = 0) using 999 random permutations of the spatial data (Kalisz et al., 2001). An observed value of Fij(r) outside of this CI signals a significantly positive or negative FSGS at distance r. To test for differences between populations, we also constructed 95% CIs about estimates of Fij at the smallest interval, Fij(1 m), as 1.96 × SE, with the SE (standard error) obtained by jackknifing, assuming Fij(1 m) to be normally distributed (Chung et al., 2011).

To test the overall pattern of genetic structure in each population, we estimated the regression slope (bF) of Fij on the loge of r. A 95% CI about the null hypothesis of no genetic structure (bF = 0) was obtained from 999 random permutations of individuals among spatial positions. Value bF was considered to be significantly different from zero if it did not lie within this 95% CI. To test for differences between populations, we assumed Fij(r) to be normally distributed and approximated 95% CIs around bF as 1.96 × SE with SE obtained by jackknifing. Slopes were considered to be significantly different among populations if their 95% CIs did not overlap. We also calculated the Sp statistic (i.e. the strength of FSGS; Vekemans & Hardy, 2004) for each locus as –bF /(1–Fij(2 m)), where Fij(2 m) is the mean Fij at the first distance class (hence 2 m). For each population, a mean Sp was obtained by averaging over loci. All spatial genetic autocorrelation statistics were obtained using the program SPAGeDi.

Results

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Genetic diversity and structure

Allozyme variation within populations was low across the four studied populations: the mean percentage of polymorphic loci within populations (%P) was 13.1, mean number of alleles per locus (A) was 1.13 and mean genetic diversity (He) was 0.042 (Table 1). Population-level FIS estimates calculated over two (CF-3) and three (CF-1, CF-2 and CF-4) polymorphic loci were positive and similar (Table 1). However, all positive FIS values, ranging from 0.104 to 0.133, were not significantly different from zero (two-tailed test; Table 1). These results, as well as the nonsignificant multi-population-level FIS (FIS = 0.111, 95% CI = −0.052 to 0.332), indicate that these populations were not significantly different from H–W equilibrium. By contrast, deviations from H–W expectations caused by allele frequency differences between populations were significantly higher than zero (FST = 0.306, 95% CI = 0.267 to 0.319).

Table 1. Summary of genetic diversity measures and mean fixation (FIS) values observed in four populations of Cyrtomium falcatum
Population N %P A Ho (SE)He (SE) F IS P value
  1. N, sample size; %P, percentage of polymorphic loci; A, mean number of alleles per locus; Ho, observed heterozygosity; He, Hardy–Weinberg expected heterozygosity or genetic diversity; SE, standard error; FIS, fixation index within populations; P, probability to accept the null hypothesis of FIS = 0.

  2. a

    Nonsignificant Weir & Cockerham (1984) estimate of FIS over populations.

CF-14614.31.140.036 (0.026)0.042 (0.026)0.1330.177
CF-26514.31.140.040 (0.026)0.045 (0.029)0.1050.232
CF-3589.51.100.041 (0.031)0.046 (0.032)0.1040.296
CF-49414.31.140.031 (0.020)0.035 (0.022)0.1120.102
Average6613.11.130.037 (0.003)0.042 (0.002)0.111a 

Spatial distribution of individuals

Except for two distance intervals at r = 7 m in CF-1 and at r = 5 m in CF-3, O-ring analyses of all four populations exhibited significant spatial aggregation of individuals up to r = 6 m (CF-1), r = 2 m (CF-2), r = 4 m (CF-3) and r = 3 m (CF-4) (Fig. 3).

image

Figure 3. Spatial structure of total Cyrtomium falcatum individuals within CF-1, CF-2, CF-3 and CF-4 as measured using the O-ring statistic (O(r)). Closed circles, mean O(r) for an annulus of radius r with 1-m lags; dotted lines, 95% confidence envelopes about the null hypothesis of random spatial structure. The first-order intensities λ of the point pattern within populations are 0.003 in CF-1, 0.018 in CF-2, 0.010 in CF-3 and 0.042 in CF-4.

Download figure to PowerPoint

Fine-scale genetic structure

Vekemans & Hardy (2004) advise that, for each distance class, the minimum number of pairs (# pairs) and the minimum percentage of individuals participating at least once (% partic) should be > 100 and 50, respectively. These recommendations were met in most cases with the exception of two cases at ≤ 1 m distance in CF-1 (# pairs = 55, % partic = 80.4) and CF-3 (# pairs = 74, % partic = 87.9).

Except for Fij(14 m) at 7–14 m in CF-2, significant FSGS was detected in all four populations (Table 2; Fig. 4). In each population, the Fij estimate at the smallest distance interval, Fij(1 m), was significant and positive, ranging from 0.121 in CF-3 to 0.280 in CF-2 (with a mean of 0.187; Table 2). However, estimates at 20–40 m (CF-1), 5–7 m (CF-2), 10–19 m (CF-3) and 5–25 m (CF-4) were significantly less than zero (Fig. 4). Accordingly, the slope (bF) of kinship on ln(r) was significantly below zero in all populations, ranging from −0.0919 (CF-4) to −0.1255 (CF-2), with a mean of −0.1081 (Table 2), but did not differ among populations as 95% CIs broadly overlapped. Consistent with this, the Sp statistics among populations were similar, ranging from 0.1117 (CF-4) to 0.1335 (CF-2), with a mean of 0.1228 (Table 2).

Table 2. Comparison of fine-scale genetic structure in four populations of Cyrtomium falcatum
Population# pairs F ij(1 m) b F Sp
EstimateP value95% CIEstimateP value95% CI
  1. # pairs, numbers of pairs at ≤ 1 m; Fij(1 m), the mean kinship coefficient calculated at ≤ 1 m; bF, slope of the pairwise kinship coefficients on distance (loge); Sp, a statistic by averaging over loci calculated for each locus as −bF/(1 − Fij(1 m)), where Fij(1 m) is the mean Fij at the first distance class (hence 1 m); P, probability to accept the null hypothesis of Fij(1 m) = 0 or under the null hypothesis of bF = 0; 95% CI, 95% confidence intervals about the estimates of Fij(1 m) and bF. Statistically significant values are indicated in bold text.

CF-1550.159 0.017 −0.028, 0.346−0.1214 0.000 −0.1794, −0.06340.1289
CF-22280.280 0.000 0.090, 0.470−0.1255 0.000 −0.1708, −0.08020.1335
CF-3740.121 0.035 −0.139, 0.381−0.0934 0.000 −0.2455, 0.05870.1170
CF-44210.187 0.000 0.002, 0.372−0.0919 0.000 −0.1248, −0.05900.1117
Average1940.187  −0.1081  0.1228
image

Figure 4. Correlograms of estimated kinship coefficients (Fij) of total Cyrtomium falcatum individuals over distance intervals within CF-1, CF-2, CF-3 and CF-4. Closed circles, mean values of Fij for corresponding distance intervals; dotted lines, upper and lower bounds of 95% confidence envelopes constructed about the null hypothesis of Fij = 0. Note that the distance scales vary among sites.

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Genetic diversity and structure

Levels of within-population genetic diversity are extremely low in C. falcatum (population-level means: %P = 13.1, A = 1.13, He = 0.042). A previous study on eight populations throughout southern South Korea also revealed similarly low within-population variation (population-level means: %P = 11.9, A = 1.12, He = 0.034; Chung et al., 2012).

Consistent with the previous study of eight southern South Korean populations (multi-population level FIS = 0.030; Chung et al., 2012), the four island populations of C. falcatum are approximately at H–W equilibrium (with statistically nonsignificant multi-population level FIS = 0.111). Many diploid homosporous ferns exhibit a high level of outcrossing (as inferred from their inbreeding coefficients; Soltis & Soltis, 1989, 1992; Ranker & Geiger, 2008). If closely related individuals of C. falcatum randomly outcrossed with others within populations, we would expect that the FIS estimate would approximately equal Fij(1 m). This is the case in C. falcatum, as we found that all but CF-2 Fij(1 m) estimates were similar to FIS: 0.159 vs 0.133 in CF-1; 0.280 vs 0.105 in CF-2; 0.121 vs 0.104 in CF-3; 0.187 vs 0.112 in CF-4 (Tables 1, 2). Although FIS < Fij suggests that inbreeding depression might be occurring (Vekemans & Hardy, 2004), this may not always be the scenario in homosporous ferns. If spore dispersal and male gamete movement are limited, individuals of close neighbourhoods are likely to be more genetically related to each other than to distantly located individuals within the same population. This neighbourhood relatedness would result in a higher mean kinship among individuals located within a short distance than the average inbreeding coefficient of the population. Therefore, we would expect a larger Fij(1 m) estimate than FIS within each population even if mating actually occurs at a larger distance than 1 m.

At the species level, we found significantly high degrees of genetic differentiation among the four populations (FST = 0.306), similar to the population structure found in the study carried out with eight populations from southern South Korea (FST = 0.543; Chung et al., 2012). Primarily owing to the patchiness of suitable habitat (which, in turn, causes restricted gene flow among populations), a high degree of among-population genetic divergence has been found in many homosporous ferns (Soltis et al., 1989; Holderegger & Schneller, 1994; Ranker et al., 1996; Schneller & Holderegger, 1996; Vogel et al., 1999; Suter et al., 2000; Pryor et al., 2001; Chung et al., 2010).

Spatial distribution of individuals within populations

Consistent with the second hypothesis, we found that populations of C. falcatum exhibited significant aggregation of individual plants (Fig. 3). In all of the four populations studied, this aggregation was primarily at a scale of < 2 m and strongest at the 0–1-m distance interval (Fig. 3). We assumed that this structure would arise because of the formation of localized spore shadows around colonizing sporophytes (i.e. good habitats occur in small patches, and so the plants are also patchily distributed). It has commonly been perceived that fern spores are capable of long-distance dispersal by wind (Tryon, 1970, 1972, 1986; Conant, 1978). Studies of several homosporous fern species, however, have demonstrated that most spores fall within a short distance of the source plant. Peck et al. (1990) showed that > 90% of spores released by Botrychium virginianum are deposited within 5 m of the source plant. Further, Dyer (1994) reported that the largest spore banks occurred in samples taken beneath sporing fronds of Pteridium aquilinum on an open hillside near Edinburgh, UK, and found that the size of the spore bank was noticeably smaller as distance increased from the source sporophyte. For comparison, one might consider the case of pollen dispersal by wind in pines. Most of the pollen disperses only short distances, whereas a few grains move long distances. For example, an isolated population of Pinus sylvestris received c. 5% pollen flow over a distance of 30 km (Robledo-Arnuncio & Gil, 2005). Despite the tendency for most spores to fall near the parent plant, spore germination and the establishment and survival of gametophytes are dependent on the distribution of adult sporophytes and environmental factors (e.g. physical features and soil conditions or properties; Richard et al., 2000). In C. falcatum, the localized structure revealed by the O-ring analyses is thus probably a consequence of limited spore dispersal and the spore preferences for suitable microhabitats (i.e. microsite mosaics).

It would be interesting to compare the present results with orchids, as ferns and orchids share similar features (e.g. small-sized propagules, high potentials for dispersal and the fact that the vast majority of propagules fall near the parent plant, following a leptokurtic distribution in most cases; Jersáková & Malinová, 2007; Fay et al., 2009). Analyses of the spatial distributions of individuals of several terrestrial orchids have revealed significant aggregation of individuals at short spatial scales in Cymbidium goeringii (Chung et al., 2011), Liparis kumokiri and L. makinoana (Chung et al., 2005a), Orchis cyclochila (Chung et al., 2005b), Orchis mascula (Jacquemyn et al., 2009), Orchis purpurea (Jacquemyn et al., 2007) and Pogonia minor (Chung & Chung, 2008). These authors attributed the significant spatial clustering of individuals to restricted seed dispersal around maternal plants. As all orchids have an obligate relationship with mycorrhizal symbionts, it is highly likely that the spatial structuring of symbionts also influences the spatial structuring of individual orchid plants (McCormick et al., 2012; and references therein).

Fine-scale genetic structure

In agreement with the second hypothesis and consistent with the significant aggregation of individuals detected within populations, all four populations exhibited significant evidence of FSGS, with mean kinship (Fij) at near-neighbour distances (≤ 1 m) being significantly positive and the slope (bF) of the regression of kinship on distance being significantly negative (Table 2). Similar to our interpretations of spatial demographic structure, we attribute the significant FSGS in these populations primarily to the presence of localized spore shadows and the establishment of new recruits around maternal plants.

To date, there have been only two other studies of FSGS in homosporous ferns, one in Pteris multifida and the other in Dryopteris aemula. Using one or two polymorphic allozyme loci and Moran's I relatedness coefficient, Murakami et al. (1997) found significant FSGS for the weedy fern Pteris multifida within three populations occurring on artificial stone walls in Japan (mean bF = −1.328; range, −0.831 to −1.545; P < 0.05; recalculated from figs. 6–8 in Murakami et al., 1997). The negative slope found in P. multifida is substantially stronger than that exhibited in C. falcatum. One possible reason is the fact that P. multifida is asexually propagated through a creeping rhizome, and the authors included clonal ramets for the calculation of Moran's I, which should inflate the strength of FSGS. Hence, we cannot confidently determine the spatial distribution of individual genotypes established through sexual reproduction and spore dispersal in P. multifida. By contrast, C. falcatum produces a short, erect rhizome. Even if its individuals are closely located in space, they remain as distinct genets. By contrast, the study by Jiménez et al. (2010) revealed no significant spatial autocorrelation at any distance class in a population of Dryopteris aemula in Spain based on four polymorphic microsatellite loci and Moran's I coefficient. The authors attributed the lack of FSGS to effective mixing of selfed genotypes by spore dispersal in the deciduous forest understory. The lack of FSGS in Dryopteris aemula should be further verified with allozyme markers, as microsatellite loci often include null alleles which may cause inaccuracies in the estimates of population genetic structure (O. J. Hardy, pers. comm.).

It is also of interest to note that both bF (ranging from −0.0919 to −0.1255, with a mean of −0.1081) and Sp (ranging from 0.1117 to 0.1335, with a mean of 0.1228) values found in C. falcatum populations are substantially higher than those reported for 16 terrestrial orchids (mean bF = −0.026 and mean Sp = 0.0297; table 5 in Chung et al., 2011). Furthermore, the strength of FSGS exhibited by C. falcatum places it in the fourth highest position when compared with the Sp values of 47 flowering plants compiled by Vekemans & Hardy (2004). Mean kinship (Fij = 0.187) at near-neighbour distances and the Sp statistic found in C. falcatum are very similar to values for the two herbaceous flowering plants Plantago major (FIS = 0.863, Fij at near-neighbour distances = 0.148, Sp = 0.1231; Van Dijk, 1987) and Androcymbium gramineum (FIS = 0.249, Fij at near-neighbour distances = 0.181, Sp = 0.1111; Caujapé-Castells & Pedrola-Monfort, 1997). Plantago major is a predominant selfer, and its pollen and seeds are dispersed by wind and gravity, respectively, whereas the mating system of A. gramineum is mixed, and its pollen and seeds are dispersed by animals and gravity, respectively.

Why do populations of C. falcatum exhibit such strong FSGS? There could be several factors involved. Ferns have certain life history traits that are strikingly different from those of seed plants. First, dispersal in ferns occurs via haploid spores. Second, the gametophytic generation in their life cycle is independent from the maternal sporophytes. Third, their spermatozoids require water for movement and the dispersal distance of the male gametes tends to be limited to a few centimetres (Peck et al., 1990). In addition, Soare & Andrei (2006/2007) reported the in vitro formation of the apogamous sporophyte (i.e. sporophyte borne out of gametophytic cells without fertilization) from a 2-month-old prothallus of C. falcatum. Although information on the distance of sperm movement and occurrence of apogamy in our study system is not available, it is likely that these factors could all have contributed to the observed strong FSGS within populations of C. falcatum.

In the absence of genetic load, in some homosporous fern species, a single spore can produce a sporophyte via intragametophytic selfing (self-fertilization of a haploid gametophyte), enabling the successful colonization of new sites (Lloyd, 1974; Flinn, 2006; Edgington, 2007; Wubs et al., 2010). Intragametophytic selfing results in completely homozygous sporophytes in a single generation (Klekowski, 1972; Vogel et al., 1999), a situation without analogues in seed plants, which, in turn, has several important genetic consequences. The spores produced by the homozygous sporophyte will be genetically identical to the original gametophyte, and the new gametophyte formed from the spore will likewise be of the same genotype, and so on, as long as intragametophytic selfing occurs. With no means of generating genetic variability, sexual reproduction in some homosporous ferns (e.g. Botychium species in Ophioglossaceae), through intragametophytic self-fertilization, becomes equivalent genetically to vegetative reproduction (D. R. Farrar, pers. comm.), which would enhance the magnitude of FSGS. However, a significant amount of inbreeding was not found within the studied populations of C. falcatum, indicating that selfing, either inter- or intragametophytic, is probably not important in the study system.

Concluding remarks

To our knowledge, this study is the first to investigate spatial demographic structure and FSGS in a homosporous fern in great detail. All four populations of C. falcatum exhibited significant aggregation of individuals and FSGS within populations. Although fern spores are analogous to the tiny, dust-like seeds of orchids, the strength of FSGS found in C. falcatum is four times stronger than that in the 16 terrestrial orchid species examined to date. There was no significant amount of selfing in the four studied populations of C. falcatum (nonsignificant FIS values), suggesting that intragametophytic selfing is probably not happening in these populations. The substantial FSGS in C. falcatum could reflect the exceptional life cycle of homosporous ferns in general, that is, outcrossing mating systems that require two spores to establish new sporophytes lead to the majority of spore dispersal occurring at short distances and extremely limited male gamete dispersal. As only two previous studies on FSGS of ferns are available and their results are in conflict, more studies are needed to determine the patterns of demographic structure and FSGS – as well as their underlying factors – in homosporous ferns.

Acknowledgements

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

The authors thank N. M. Tuyen, N. Lu, B. J. Shim, E. J. Im, M. S. Park and C. H. Chung for field and laboratory assistance. Special thanks are due to James L. Hamrick (University of Georgia, USA), Carl Rothfels (Duke University, USA), Eric Myers (South Suburban College, USA), Mei Sun (University of Hong Kong, China) and Jordi López-Pujol (Universitat de Barcelona, Spain) for improving previous versions of the manuscript. In particular, we are grateful to three anonymous referees and Lynda Delph for their comments on a previous version of the manuscript. This work was supported by the Basic Science Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Education, Science and Technology (NRF-2009-0066840 to M.Y.C. and NRF-2011-017236 to M.G.C.).

References

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  • Arditti J, Ghani AKA. 2000. Tansley review no. 110. Numerical and physical properties of orchid seeds and their biological implications. New Phytologist 145: 367421.
  • Berg EE, Hamrick JL. 1995. Fine-scale genetic structure of a turkey oak forest. Evolution 49: 110120.
  • Caujapé-Castells J, Pedrola-Monfort J. 1997. Space–time patterns of genetic structure within a stand of Androcymbium gramineum (Cav.) McBride (Colchicaceae). Heredity 79: 341349.
  • Cheliak WM, Pitel JP. 1984. Technique for starch gel electrophoresis of enzyme from forest tree species. In: Information Report PI-X-42, 1–49. Chalk River: Petawawa National Forestry Institute, Agriculture Canada, Canadian Forestry Service, Canada.
  • Chung MY, Chung MG. 2008. Conservation genetics of the endangered terrestrial orchid Pogonia minor in South Korea. Annales Botanici Fennici 45: 455464.
  • Chung MY, López-Pujol J, Chung JM, Kim K-J, Chung MG. 2012. Low within population genetic variation and high among population differentiation in Cyrtomium falcatum (Dryopteridaceae) in southern Korea: inference of population-establishment history. American Fern Journal 102: 256272.
  • Chung MY, Nason JD. 2007. Spatial demographic and genetic consequences of harvesting within populations of the terrestrial orchid Cymbidium goeringii. Biological Conservation 137: 125137.
  • Chung MY, Nason JD, Chung MG. 2005a. Patterns of hybridization and population genetic structure in the terrestrial orchids Liparis kumokiri and Liparis makinoana (Orchidaceae) in sympatric populations. Molecular Ecology 14: 43894402.
  • Chung MY, Nason JD, Chung MG. 2005b. Spatial genetic structure in populations of terrestrial orchid Orchis cyclochila (Orchidaceae). Plant Systematics and Evolution 254: 209219.
  • Chung MY, Nason JD, Chung MG. 2007. Effects of population succession on demographic and genetic processes: predictions and tests in the daylily Hemerocallis thunbergii (Liliaceae). Molecular Ecology 16: 28162829.
  • Chung MY, Nason JD, Chung MG. 2011. Significant demographic and fine-scale genetic structure in expanding and senescing populations of the terrestrial orchid Cymbidium goeringii (Orchidaceae). American Journal of Botany 98: 20272039.
  • Chung MY, Nason JD, Sun B-Y, Moon M-O, Chung JM, Park C-W, Chung MG. 2010. Extremely low levels of genetic variation in the critically endangered monotypic fern genus Mankyua chejuense (Ophioglossaceae) from Korea: implications for conservation. Biochemical Systematics and Ecology 38: 888896.
  • Clayton JW, Tretiak DN. 1972. Amine citrate buffers for pH control in starch gel electrophoresis. Journal of the Fisheries Research Board of Canada 29: 11691172.
  • Conant DS. 1978. A radioisotope technique to measure spore dispersal of the tree fern Cyathea arborea Sm. Pollen et Spores 20: 583593.
  • During HJ, Ter Horst B. 1983. The diaspore bank of bryophytes and ferns in chalk grassland. Lindbergia 9: 5764.
  • Dyer AF. 1994. Natural soil spore banks – can they be used to retrieve lost ferns? Biodiversity and Conservation 3: 160175.
  • Edgington JA. 2007. Dynamics of long-distance dispersal: the spread of Asplenium adiantum-nigrum and Asplenium trichomanes (Aspleniaceae; Pteridophyta) on London walls. Fern Gazette 18: 3138.
  • Ellstrand NC. 1992. Gene flow among seed plant populations. New Forests 6: 241256.
  • Epperson BK, Chung MG. 2001. Spatial genetic structure of allozyme polymorphisms within populations of Pinus strobus (Pinaceae). American Journal of Botany 88: 10061010.
  • Farrar DR. 1998. Population genetics of moonwort Botrychium. In: Berlin N, Miller P, Borovansky J, Seal US, Byers O, eds. Population and habitat viability assessment for the goblin fern (Botrychium mormo), final report. Apple Valley, MN, USA: The Conservation Breeding Specialist Group, 109113.
  • Fay MF, Bone R, Cook P, Kahandawala I, Greensmith J, Harris S, Pedersen HÆ, Ingrouille MJ, Lexer C. 2009. Genetic diversity in Cypripedium calceous (Orchidaceae) with a focus on north-western Europe, as revealed by plastid DNA length polymorphisms. Annals of Botany 104: 517525.
  • Flinn KM. 2006. Reproductive biology of three fern species may contribute to differential colonization success in post-agricultural forests. American Journal of Botany 93: 12891294.
  • Goswami HK. 1987. Ophioglossales I: an overview. Bionature 7: 4789.
  • Goudet J. 1995. FSTAT: a computer program to calculate F statistics, version 1.2. Journal of Heredity 86: 485486.
  • Hamrick JL, Murawski DA, Nason JD. 1993. The influence of seed dispersal mechanisms on the genetic structure of tropical tree populations. Vegetatio 107/108: 281297.
  • Hamrick JL, Nason JD. 1996. Consequences of dispersal in plants. In: Rhodess OE Jr, Chesser RK, Smith MH, eds. Population dynamics in ecological space and time. Chicago, IL, USA: University of Chicago Press, 203236.
  • Hardy OJ, Vekemans X. 2002. SPAGeDi: a versatile computer program to analyze spatial genetic structure at the individual or population levels. Molecular Ecology Notes 2: 618620.
  • Haufler CH. 1985. Enzyme variability and modes of evolution in Bommeria (Pteridaceae). Systematic Botany 10: 92104.
  • Haufler CH, Soltis DE. 1984. Obligate outcrossing in a homosporous fern: field confirmation of a laboratory prediction. American Journal of Botany 71: 878881.
  • Holderegger R, Schneller JJ. 1994. Are small isolated populations of Asplenium septentrionale variable? Biological Journal of the Linnean Society 51: 377385.
  • Iwatsuki K. 1992. Ferns and fern allies of Japan. Tokyo, Japan: Heibonsha Ltd.
  • Jacquemyn H, Brys R, Vandepitte K, Honnary O, Roldán-Ruiz I, Wiegand T. 2007. A spatially explicit analysis of seedling recruitment in the terrestrial orchid Orchis purpurea. New Phytologist 176: 448459.
  • Jacquemyn H, Wiegand T, Vandepitte K, Brys R, Roldán-Ruiz I, Honnary O. 2009. Multigenerational analysis of spatial structure in the terrestrial, food-deceptive orchid Orchis mascula. Journal of Ecology 97: 206216.
  • Jersáková J, Malinová T. 2007. Spatial aspects of seed dispersal and seedling recruitment in orchids. New Phytologist 176: 237241.
  • Jiménez A, Holderegger R, Csencsics D, Quintanilla LG. 2010. Microsatellites reveal substantial among-population genetic differentiation and strong inbreeding in the relict fern Dryopteris aemula. Annals of Botany 106: 149155.
  • Kalisz S, Nason JD, Hanzawa FM, Tonsor SJ. 2001. Spatial population genetic structure in Trillium grandiflorum: the roles of dispersal, mating, history and selection. Evolution 55: 15601568.
  • Klekowski EJ. 1972. Genetical features of ferns as contrasted to seed plants. Annals of the Missouri Botanical Garden 59: 138151.
  • Knobloch IW. 1969. The spore pattern in some species of Cheilanthes. American Journal of Botany 56: 646653.
  • Linhart YB, Mitton JB, Sturgeon KB, Davis ML. 1981. Genetic variation in space and time in a population of ponderosa pine. Heredity 46: 407426.
  • Lloyd RM. 1974. Reproductive biology and evolution in the Pteridophyta. Annals of the Missouri Botanical Garden 61: 318331.
  • Loiselle BA, Sork VL, Nason JD, Graham C. 1995. Spatial genetic structure of a tropical understory shrub, Psychotria officinalis (Rubiaceae). American Journal of Botany 82: 14201425.
  • Lu J-M, Cheng X, Wu D, Li D-Z. 2006. Chromosome study of the fern genus Cyrtomium (Dryopteridaceae). Botanical Journal of the Linnean Society 150: 221228.
  • Makgomol K. 2006. Morphology of fern spores from Phu Phan National Park. Kasetsart Journal (Natural Science) 40: 116122.
  • McCauley DE, Whittier DP, Reilly LM. 1985. Inbreeding and the rate of self-fertilization in a grape fern, Botrychium dissectum. American Journal of Botany 72: 19781981.
  • McCormick MK, Taylor DL, Juhaszova K, Burnett RK Jr, Whigham DF, O'Neill JP. 2012. Limitations on orchid recruitment: not a simple picture. Molecular Ecology 21: 15111523.
  • Mitton JB, Linhart YB, Sturgeon KB, Hamrick JL. 1979. Allozyme polymorphisms detected in mature needle tissue of ponderosa pine. Journal of Heredity 70: 8689.
  • Murakami N, Nishiyama T, Satoh H, Suzuki T. 1997. Marked spatial genetic structure in three populations of a weedy fern, Pteris multifida Poir., and reestimation of its selfing rate. Plant Species Biology 12: 97106.
  • Nei M. 1978. Estimation of average heterozygosity and genetic distance from a small number of individuals. Genetics 89: 583590.
  • Peck JH, Peck CJ, Farrar DR. 1990. Influences of life history attributes on formation of local and distant fern populations. American Fern Journal 80: 126142.
  • Perrie LR, Ohlsen DJ, Shepherd LD, Grrreff M, Brownsey PJ, Bayly MJ. 2010. Tasmanian and Victorian populations of the fern Asplenium hookerianum result from independent dispersal from New Zealand. Australian Systematic Botany 23: 387392.
  • Pryor KV, Young JE, Rumsey FJ, Edwards KJ, Bruford MW, Rogers HJ. 2001. Diversity, genetic structure and evidence of outcrossing in British populations of the rock fern Adiantum capillus-veneris using microsatellites. Molecular Ecology 10: 18811894.
  • Ramírez-Trejo MDR, Pérez-García B, Orozco-Segovia A. 2004. Analysis of fern spore banks from the soil of three vegetation types in the central region of Mexico. American Journal of Botany 91: 682688.
  • Ranker TA, Floyd SK, Trapp PG. 1994. Multiple colonizations of Asplenium adiantum-nigrum on to the Hawaiian Archipelago. Evolution 48: 13641370.
  • Ranker TA, Geiger JMO. 2008. Population genetics. In: Ranker RA, Haufler CH, eds. Biology and evolution of ferns and lycophytes. New York, NY, USA: Cambridge University Press, 175198.
  • Ranker TA, Gemmill CEC, Trapp PG, Hambelton A, Ha K. 1996. Population genetics and reproductive biology of lava-flow colonising species of Hawaian Sadleria (Blechnaceae). In: Camus JM, Gibby M, Johns RJ, eds. Pteridophytes in perspective. Kew, UK: Royal Botanic Gardens, 581598.
  • Richard M, Bernhardt T, Bell G. 2000. Environmental heterogeneity and the spatial structure of fern species diversity in one hectare of old-growth forest. Ecography 23: 231234.
  • Robledo-Arnuncio JJ, Gil L. 2005. Patterns of pollen dispersal in a small population of Pinus sylvestris L. revealed by total-exclusion paternity analysis. Heredity 94: 1322.
  • Roux JP. 2011. The genus Cyrtomium (Pteridophyta: Dryopteridaceae) in Africa and Madagascar. Botanical Journal of the Linnean Society 167: 449465.
  • Schneller JJ, Holderegger R. 1996. Colonisation events and genetic variability within populations of Aspleniumn ruta-muraria L. In: Camus JM, Gibby M, Johns RJ, eds. Pteridophytes in perspective. Kew, UK: Royal Botanic Gardens, 571580.
  • Soare LC, Andrei M. 2006/2007. Morphogenesis of apogamous sporophyte in Cyrtomium falcatum (L. Fil.) C. Presl (Dryopteridaceae). Romanian Journal of Biology–Plant Biology 51–52: 4954.
  • Sokal RR, Jacquez GM. 1991. Testing inferences about microevolutionary processes by means of spatial autocorrelation analysis. Evolution 45: 152168.
  • Soltis DE, Haufler CH, Darrow DC, Gastony GJ. 1983. Starch gel electrophoresis of ferns: a compilation of grinding buffers, gel and electrode buffers, and staining schedules. American Fern Journal 73: 927.
  • Soltis DE, Soltis PS. 1986. Electrophoretic evidence for inbreeding in the fern Botrychium virginianum (Ophioglossaceae). American Journal of Botany 73: 588592.
  • Soltis DE, Soltis PS. 1989. Polyploidy, breeding systems, and genetic differentiation in homosporous Pteridophytes. In: Soltis DE, Soltis PS, eds. Isozymes in plant biology. Portland, OR, USA: Dioscorides Press, 241258.
  • Soltis DE, Soltis PS. 1992. The distribution of selfing rates in homosporous ferns. American Journal of Botany 79: 97100.
  • Soltis PS, Soltis DE, Ness BD. 1989. Population genetic structure in Cheilanthes gracillima. American Journal of Botany 76: 11141118.
  • Suter M, Schneller JJ, Vogel JC. 2000. Investigations into the genetic variation, population structure and breeding system of the fern Asplenium trichomanes subsp. quadrivalens. International Journal of Plant Science 161: 233244.
  • Troupin D, Nathan R, Vendramin GG. 2006. Analysis of spatial genetic structure in an expanding Pinus halepensis population reveals development of fine-scale genetic structuring over time. Molecular Ecology 15: 36173630.
  • Tryon R. 1970. Development and evolution of fern floras of oceanic islands. Biotropica 2: 7684.
  • Tryon R. 1972. Endemic areas and geographic speciation in tropical American ferns. Biotropica 4: 121131.
  • Tryon R. 1986. The biogeography of species, with special reference to ferns. Botanical Review (Lancaster) 52: 117156.
  • Van Dijk H. 1987. A method for the estimation of gene flow parameters from a population structure caused by restricted gene flow and genetic drift. Theoretical and Applied Genetics 73: 724736.
  • Vekemans X, Hardy OJ. 2004. New insights from fine-scale spatial genetic structure analyses in plant populations. Molecular Ecology 13: 921935.
  • Vogel JC, Rumsey FJ, Russell SJ, Cox CJ, Holmes JS, Bujnoch W, Stark C, Barrett JA, Gibby M. 1999. Genetic structure, reproductive biology and ecology of isolated populations of Asplenium csikii (Aspleniaceae, Pteridophyta). Heredity 83: 604612.
  • Wagner WH, Wagner FS, Beital JM. 1985. Evidence for interspecific hybridization in pteridophytes with subterranean mycoparasitic gametophytes. Proceedings of the Royal Society of Edinburgh. Section B. Biological Sciences 86: 273281.
  • Watano Y, Sahashi N. 1992. Predominant inbreeding and its genetic consequences in a homosporous fern genus, Sceptridium (Ophioglossaceae). Systematic Botany 17: 486502.
  • Weeden NF, Wendel JF. 1989. Genetics of plant isozymes. In: Soltis DE, Soltis PS, eds. Isozymes in plant biology. Portland, OR, USA: Dioscorides Press, 4672.
  • Weir BS, Cockerham CC. 1984. Estimating F-statistics for the analysis of population structure. Evolution 38: 13581370.
  • Wiegand T. 2003. PROGRAMITA, a software to perform point pattern analysis with Ripley's L and O-ring statistic. [WWW document] URL http://www.thorsten-wiegand.de/towi_download_manual.html [accessed on 13 March 2012].
  • Wiegand T, Moloney KA. 2004. Rings, circles, and null-models for point pattern analysis in ecology. Oikos 104: 209229.
  • Wolf PG, Sheffield E, Haufler CH. 1991. Estimates of gene flow, genetic substructure and population heterogeneity in bracken (Pteridium aquilinum). Biological Journal of the Linnean Society 42: 407423.
  • Wright S. 1943. Isolation by distance. Genetics 28: 114138.
  • Wright S. 1965. The interpretation of population structure by F-statistics with special regard to systems of mating. Evolution 19: 395420.
  • Wubs ER, de Groot GA, During HJ, Vogel JC, Grundmann M, Bremer P, Schneider H. 2010. Mixed mating system in the fern Asplenium scolopendrium: implications for colonization potential. Annals of Botany 106: 583590.
  • Yeh FC, Yang RC, Boyle TBJ. 1999. POPGENE. version 1.31. Microsoft windows-based free ware for population genetic analysis. [WWW document] URL http://www.ualberta.ca/~fyeh/index/htm [accessed on 12 March 2012].