Advertisement
Research Article

Can Soil Seed Banks Serve as Genetic Memory? A Study of Three Species with Contrasting Life History Strategies

  • Bohumil Mandák mail,

    bohumil.mandak@ibot.cas.cz

    Affiliations: Institute of Botany, Academy of Sciences of the Czech Republic, Průhonice, Czech Republic, Faculty of Environmental Sciences, Czech University of Life Sciences Prague, Prague, Czech Republic

    X
  • Petr Zákravský,

    Affiliation: Institute of Botany, Academy of Sciences of the Czech Republic, Průhonice, Czech Republic

    X
  • Václav Mahelka,

    Affiliation: Institute of Botany, Academy of Sciences of the Czech Republic, Průhonice, Czech Republic

    X
  • Ivana Plačková

    Affiliation: Institute of Botany, Academy of Sciences of the Czech Republic, Průhonice, Czech Republic

    X
  • Published: November 21, 2012
  • DOI: 10.1371/journal.pone.0049471

Abstract

We attempted to confirm that seed banks can be viewed as an important genetic reservoir by testing the hypothesis that standing (aboveground) plants represent a nonrandom sample of the seed bank. We sampled multilocus allozyme genotypes from three species with different life history strategies: Amaranthus retroflexus, Carduus acanthoides, Pastinaca sativa. In four populations of each species we analysed the extent to which allele and genotype frequencies vary in consecutive life history stages including the summer seed bank, which has been overlooked up to now. We compared the winter seed bank (i.e., seeds collected before the spring germination peak), seedlings, rosettes, the summer seed bank (i.e., seeds collected after the spring germination peak) and fruiting plants. We found that: (1) All three species partitioned most of their genetic diversity within life history stages and less among stages within populations and among populations. (2) All genetic diversity parameters, except for allele frequencies, were similar among all life history stages across all populations in different species. (3) There were differences in allele frequencies among life history stages at all localities in Amaranthus retroflexus and at three localities in both Carduus acanthoides and Pastinaca sativa. (4) Allele frequencies did not differ between the winter and summer seed bank in most Carduus acanthoides and Pastinaca sativa populations, but there was a marked difference in Amaranthus retroflexus. In conclusion, we have shown that the summer seed bank is not genetically depleted by spring germination and that a majority of genetic diversity remains in the soil through summer. We suggest that seed banks in the species investigated play an important role by maintaining genetic diversity sufficient for recovery rather than by accumulating new genetic diversity at each locality.

Introduction

The term “seed bank” refers to dormant seeds that are present in the soil. Buried seeds of many species can retain their viability for long periods [1], [2]. New individuals are recruited from the seed bank over time, which may strongly affect vegetation composition, structure and dynamics [3]. In general, there are two types of seed banks: transient and persistent [3], [4]. Transient seed banks are composed of seeds that germinate within a year of their initial dispersal and may not be viable at the onset of the next growing season. Persistent seed banks are made up of seeds that remain viable in the soil for longer than one year. While many plants develop vegetatively, some species depend on seed reserves that have accumulated in the soil between disturbances. Thompson and Grime [5], who studied ten contrasting habitats, found that at every site there were some species which had a persistent seed bank and some whose seeds were present for only part of the year. The size of the buried seed pool reflects the type, intensity and frequency of disturbance [6] and can significantly differ among species. Over the past 20 years, however, a growing volume of literature dealing with buried seeds has mainly focused on the importance of seed banks for population recovery following disturbances of established vegetation, that is, from an ecological point of view [3], [4], [6], [7].

Besides the ecological role of seed banks, several authors have studied their importance from an evolutionary perspective. This idea was first formulated by Templeton and Levin [8]. In their view, seed banks can serve as (i) regeneration pools with reasonable genetic variability that may be an important determinant of the success of different species at a locality, buffering the effects of local extinction of genotypes in adult populations caused by selection or drift, or (ii) as an “evolutionary memory” which stores genotypes for a variable number of years [9], possibly slowing the rate of evolutionary change [8], [10].

Many authors have tried to confirm that seed banks represent important genetic reservoirs by testing the hypothesis that standing (aboveground) plants represent a nonrandom sample of their species’ seed bank [11][15]. If seed banks really function as genetic reservoirs by maintaining genetic diversity, they must be genetically more diverse than subsequent life history stages due to accumulation of different genotypes over time. The presence of diverse genotypes in soil seed banks may then serve as raw material for evolutionary processes that are behind the transformation of soil seeds into reproductive populations. Despite their effort, scientists have been unable to conclusively confirm such a role of seed banks in evolutionary dynamics of plant populations [9], [11][17]. While some authors have demonstrated differences between seed banks and aboveground populations such as higher genetic diversity of seed bank populations [9], [13], higher heterozygosity of aboveground populations [11][15] or significant differences in allele frequencies [11], [12], [14], [15], studies of others show no differences in any of the population genetic parameters investigated [16], [17]. This discrepancy stems partly from different methodological approaches to comparing genetic diversity parameters of seed banks with the parameters of different stages of aboveground populations (i.e., seedlings or adults, or combination of both). Most importantly, ignoring the seedling stage could lead to incorrect interpretations of differences between seed banks and adult populations, as such differences may be due to microselective forces acting on early aboveground populations, rather than due to storing genetic variability in the soil or germination and establishment success [15].

Three mechanisms have been proposed to explain the commonly observed pattern of higher homozygosity and inbreeding in seed banks [9], [11][16], [18], [19]. First, it could be explained by higher inbreeding in past years, as argued by Tonsor et al. [14]. Secondly, it could be attributed to a temporal Wahlund effect [14], but see [20]. Thirdly, Vitalis et al. [20] point out that the most probable mechanism responsible is natural selection, which progressively eliminates less fit homozygotes through, for example, their lower germinability [21], [22] or a strong self-thinning process between the seedling and mature plant stage, as described by Mandák et al. [15].

thumbnail

Table 1. Analysis of molecular variance in Amaranthus retroflexus, Carduus acanthoides and Pastinaca sativa.

doi:10.1371/journal.pone.0049471.t001

Only a few empirical studies have investigated the genetic relationship between above- and belowground populations on large data sets (i.e., from more than five localities of a single species). Moreover, few studies (e.g., [12], [17], [23]) have investigated the genetic relationship between seed banks and surface plants across large spatial scales. Cabin et al. [12] found differences between above- and belowground populations at four out of five localities of Lesquerella fendleri. By contrast, Lundemo et al. [23] and Falahati-Anbaran et al. [17] did not find any differences in gene diversity for Arabidopsis thaliana and Arabidopsis lyrata subsp. petraea, respectively. Such large-scale, multipopulation studies are necessary to test whether there are any broad genetic differences between soil seeds and surface plants rather than mere idiosyncratic differences between these life-history stages in particular populations. These studies will also allow to test the hypothesis that seed banks reduce the rate of population differentiation by retaining alleles preserved from different generations [13], [14].

In this study, we have introduced the term “summer seed bank”. It refers to seeds that persist in the soil after the spring germination peak and before new seeds are added to the soil by fruiting plants. It represents the actual seed bank for the next year(s) even if the aboveground population is completely destroyed. The summer seed bank thus holds very important information concerning the real amount of genetic diversity persisting in the seed bank through summer. If we compare only genetic diversity in the soil seed bank with the diversity of standing plants, we completely miss any information concerning genetic diversity that is actually present in the seed bank through summer. Despite this, the role of the summer seed bank has been overlooked until now, and studies that deal with the genetic relationship among various life history stages, including the summer seed bank on larger spatial scales, are missing. Such studies may answer the question: “What factors are responsible for nonrandom patterns of genetic variability commonly found in natural populations?” In this study, we sampled multilocus allozyme genotypes from three species with different life history strategies to test whether there are any broad genetic differences between soil seeds and surface plants. We analysed four populations of each species to ascertain the extent to which allele and genotype frequencies vary among life history stages including the summer seed bank. Specifically, we sought answers to the following questions: (1) Are there any differences in population genetic characteristics among species with different life history strategies? (2) Is the pattern of genetic variability consistent over several localities within each species? (3) Does the winter and summer seed bank differ from other life history stages and do seed banks in general thus serve as a genetic reservoir of the population? (4) Do seed banks reduce the rate of among-population differentiation?

thumbnail

Table 2. Population genetic characteristics for individual life history stages of Amaranthus retroflexus, Carduus acanthoides and Pastinaca sativa.

doi:10.1371/journal.pone.0049471.t002

Results

Hierarchical Analysis of Molecular Variance

Hierarchical analyses of molecular variance (AMOVA) computed separately for Amaranthus retroflexus, Carduus acanthoides and Pastinaca sativa revealed that most of the genetic diversity was partitioned within life history stages; the analysis accounted for 71.19, 93.44 and 95.55% of the overall variance, respectively (Table 1). In Carduus acanthoides and Pastinaca sativa, only 0.82 and 1.42% of the genetic diversity was partitioned among stages within populations, and 5.74 and 3.03% was partitioned among populations, respectively. Amaranthus retroflexus, on the other hand, showed higher differentiation among populations as well as among stages within populations (Table 1).

thumbnail

Figure 1. Population genetic characteristics of Amaranthus retroflexus.

Comparison of allelic richness (RS), observed heterozygosity (HO), gene diversity (HS) and inbreeding coefficient (FIS) for individual life history stages of Amaranthus retroflexus at four localities. * Significant differences among life history stages within a locality were identified by two-sided t-tests after 10 000 permutations (analysis performed using FSTAT [47]).

doi:10.1371/journal.pone.0049471.g001
thumbnail

Figure 2. Population genetic characteristics of Carduus acanthoides.

Comparison of allelic richness (RS), observed heterozygosity (HO), gene diversity (HS) and inbreeding coefficient (FIS) for individual life history stages of Carduus acanthoides at four localities. * Significant differences among life history stages within a locality were determined by two-sided t-tests after 10 000 permutations (analysis performed using FSTAT [47]).

doi:10.1371/journal.pone.0049471.g002
thumbnail

Figure 3. Population genetic characteristics of Pastinaca sativa.

Comparison of allelic richness (RS), observed heterozygosity (HO), gene diversity (HS) and inbreeding coefficient (FIS) for individual life history stages of Pastinaca sativa at four localities. * Significant differences among life history stages within a locality were determined by two-sided t-tests after 10 000 permutations (analysis performed using FSTAT [47]).

doi:10.1371/journal.pone.0049471.g003

Comparison of Life History Stages Across Populations

When we compared population genetic characteristics – i.e., allelic richness (RS), observed heterozygosity (HO), gene diversity (HS), inbreeding coefficient f(FIS) and levels of differentiation among populations θ(FST) – among life history stages across all populations of individual species, we did not reveal any statistically significant differences (Table 2). Moreover, inbreeding coefficients f(FIS) in Amaranthus retroflexus and Carduus acanthoides, and fixation indices θ(FST) in all species and life history stages were significantly different from zero (Table 2).

thumbnail

Table 3. Tests for differences in allele frequencies of individual loci and over all loci.

doi:10.1371/journal.pone.0049471.t003

Comparison of Life History Stages within Populations

When we tested differences among life history stages separately for individual populations of the three species, we obtained inconsistent results concerning individual parameters, populations and species (Figs. 1, 2, 3). We found significant difference only for HO and f(FIS) in Carduus acanthoides at one locality (locality Orech) (Fig. 2) and only for HS in Pastinaca sativa at one locality (Fig. 3). Statistically significant differences detected in Amaranthus retroflexus were not consistent over all sampled populations (Fig. 1). Significant differences in RS and HS were found at the locality Besno, but not at the other localities. In the same way, observed heterozygosity (HO) and inbreeding coefficient f(FIS) were significantly different among life history stages only at locality Psov.

Although allele frequencies differed among life history stages at all localities in Amaranthus retroflexus, the results for both Carduus acanthoides and Pastinaca sativa were not consistent (Table 3). Table 3 also shows the influence of particular loci on the results computed over all loci. Allele frequencies are usually significantly different only at some specific loci over all localities. For Amaranthus retroflexus, it is locus Est-1 and in some cases Dia-2. In Carduus acanthoides Shdh participates importantly with a lower contribution of other loci. The Est locus and in some cases the Pgm locus are important for differentiating among life history stages in Pastinaca sativa (Table 3).

thumbnail

Table 4. Tests for differences in allele frequencies among individual life history stages of Carduus acanthoides.

doi:10.1371/journal.pone.0049471.t004
thumbnail

Table 5. Tests for differences in allele frequencies among individual life history stages of Amaranthus retroflexus.

doi:10.1371/journal.pone.0049471.t005
thumbnail

Table 6. Tests for differences in allele frequencies among individual life history stages of Pastinaca sativa.

doi:10.1371/journal.pone.0049471.t006

To understand changes in allele frequencies further, we tested allele frequencies separately for each species and locality (Tables 4, 5, 6). Significant differences in allele frequencies were most often encountered in Amaranthus retroflexus, followed by Carduus acanthoides and Pastinaca sativa. The results are very diverse even within species. In other words, there are populations with stages completely different among each other (e.g., Besno and Pocedelice in Amaranthus, Table 5) as well as populations with similar allele frequencies over all stages (e.g., Dobre Pole and Zidneves in Pastinaca; Table 6).

However, from simple allele presence/absence counting within individual life history stages, it is evident that, in Amaranthus retroflexus, only in two cases out of 64 was an allele present in the summer seed bank and not in other stages; in no case was an allele detected only in the winter seed bank (Table S1). Carduus acanthoides did not allocate a significantly higher proportion of genetic diversity to its winter or summer seed bank than to its aboveground populations. Only in five cases out of 100 (Table S2) did we detect alleles present only in the seed bank and absent from any other stage. Pastinaca sativa, exhibited a similar pattern as Carduus acanthoides, i.e. only in two cases out of 120, the allele was present in the winter seed bank and absent from other stages (Table S3).

thumbnail

Table 7. Genetic diversity of Amaranthus retroflexus, Carduus acanthoides and Pastinaca sativa (from other studies).

doi:10.1371/journal.pone.0049471.t007

Discussion

The potential role of soil seed banks in evolutionary dynamics of plant populations has evoked many theoretical [20], [24][31] as well as empirical studies [9], [11][17], [32][35]. Some of these studies view seed banks as important genetic reservoirs that affect the evolution and dynamics of aboveground populations. If seed banks were true genetic reservoirs, they should accumulate genotypes over time and would have the effect of slowing down population differentiation by retaining alleles preserved over multiple generations. Our study, contrary to most published results, does not support the hypothesis that seed banks contain significantly larger amounts of genetic diversity than other life history stages. The patterns of partitioning of genetic diversity were similar in all three species. Amaranthus retroflexus, however, partitioned a higher amount of molecular variance among populations and among stages within populations compared to the other two species (Table 1). The annual invasive Amaranthus retroflexus is predominantly a selfing species. It is therefore not surprising that it forms genetically differentiated populations. Besides this, it has the highest reproductive output, and its seedling populations are the densest, so self-thinning can alter them substantially, leading to differentiation among stages (see also [15] for Atriplex tatarica). As pointed out by Morris et al. [9], the capacity of seed banks to retain higher levels of genetic diversity may depend on (1) seed dormancy characteristics and potential multigenerational contributions to the seed bank, and (2) the relative size of seed bank populations. Species producing large seed banks with dormant seeds would be expected to contain higher levels of genetic diversity in their seed banks in comparison to other life history stages [8], [9], [13]. Among our study species, it seems that mainly Amaranthus retroflexus has the potential to build large, persistent seed banks. This species, however, did not accumulate higher genetic diversity in its seed bank either. The winter seed bank did not accumulate genetic diversity and was not significantly different from other stages in other species such as Calluna vulgaris [16], Atriplex tatarica [15], Cardamine amara [19] and Arabidopsis lyrata subsp. petraea [17]. Honnay et al. [36], in their meta-analysis of 13 published studies comparing genetic diversity between the seed bank and other life history stages, found no evidence of high levels of genetic diversity accumulating in the soil seed bank. However, differences in allele frequencies between the seed bank and aboveground plants were detected in almost all cases [36]. They noted, like Vitalis et al. [20] or Mandák et al. [15] before, that “if genetic differences are present between the standing crop and the seed bank, they are very likely the result of local selection acting either directly or indirectly as a filter on the alleles present in the seed bank”. We also found differences in allele frequencies in all species and at all but two localities (Table 3). However, consistent differences among individual life history stages were only encountered in Amaranthus (Tables 4, 5, 6). We observed the same trend when we compared allele frequencies between the winter seed bank and the summer seed bank (Tables 4, 5, 6). The changes in allele frequencies in Amaranthus contrast with the other population genetic characteristics, in which no significant differences among individual life history stages were observed (Table 2). These results suggest that while spring germination does not genetically deplete the seed bank when most of the genetic diversity is maintained in the soil through summer, forces affecting allele frequencies such as natural selection or drift (or a combination of both) may play a more important role in shaping the populations of Amaranthus through their maturation process (see also [15] for Atriplex tatarica). In the annual and invasive Amaranthus retroflexus, the number of seeds in the seed bank must decrease dramatically due to massive spring germination. The process of self-thinning can later remove a significant portion of individuals, which might be the reason why changes in allele frequencies were observed among nearly all life history stages. In the perennial, outcrossing species, this question seems less relevant. However, the analysis of Honnay et al. [36] does not confirm this hypothesis entirely since by far the most of their investigated species were perennials. One possible explanation for this discrepancy is that different stages of aboveground populations were investigated, i.e., seedlings, small immature plants, mature plants or fruiting plants, in our and in the other studies (see also [12], [15]).

thumbnail

Figure 4. Location of Amaranthus retroflexus, Carduus acanthoides and Pastinaca sativa populations.

Map showing the location of the 20 populations of each of the three species under study, i.e. Amaranthus retroflexus (empty circles), Carduus acanthoides (empty squares) and Pastinaca sativa (empty triangles). Four localities of each species were selected and used in the present study (highlighted by full symbols).

doi:10.1371/journal.pone.0049471.g004
thumbnail

Table 8. Summary of sampled localities of individual species.

doi:10.1371/journal.pone.0049471.t008

Even though the presence of a seed bank acting as a long-term reservoir of genetic diversity has been first pondered by Templeton & Levin [8] and later supported by quite a large number of empirical studies, experiments covering multiple species and multiple populations of each species with the summer seed bank included in the experiments do not support this idea unequivocally. In our view, the question whether seed banks differ from standing aboveground populations should be replaced by the question whether the winter seed bank differs from the summer seed bank, which represents the actual genetic reserve for the following years. The winter seed bank is a sum of genetic diversity accumulated over years due to, for example, selective germination [11], [22]. It also contains genetic diversity that is added each year in the form of seeds falling from fruiting plants [37]. On the other hand, the summer seed bank contains seeds that did not germinate in spring, and thus may store genotypes for which the conditions of the particular environment or season are not favorable, or remain dormant. Hence, by comparing the winter seed bank with aboveground populations, we cannot draw a conclusion regarding the seed bank as a possible reservoir of genetic diversity because we do not know what really remains in it. In the present study, the overall genetic diversity stored in the winter seed bank was similar to the one that actually remained in the summer seed bank and other life history stages. From this point of view, the seed bank can serve as a genetic reservoir by maintaining, but not accumulating, genetic diversity. We therefore cannot state that seed banks play an important role in accumulating local genetic diversity. Instead, our results suggest that seed banks in the species investigated play an important role by merely maintaining genetic diversity sufficient for recovery of population genetic diversity at localities.

Conclusions

We have demonstrated for the first time the extent to which the summer soil seed bank differs from other life history stages. Knowing this difference is crucial for answering the question that concerns the soil seed bank as a genetic reservoir of a population. We have clearly shown that the summer seed bank is not completely depleted by spring germination and that most of the genetic diversity can survive through summer. Our results support the assumption that seed banks may protect plant populations from detrimental consequences of size fluctuations. We suggest that seed banks in the species investigated play an important role by maintaining genetic diversity sufficient for recovery rather than by accumulating new genetic diversity at each locality. Finally, we have shown that seed banks in the species under study did not have the effect of slowing population differentiation by retaining alleles preserved over multiple generations.

Materials and Methods

Study Species

We chose the following species for our study: Amaranthus retroflexus L., Carduus acanthoides L. and Pastinaca sativa L. subsp. sativa. The principal criteria for species selection were diploid ploidy level and production of persistent seed banks. Furthermore, we selected species with different life history strategies and habitat preferences in order to test whether there are any broad genetic differences. The species characteristics were mostly taken from [38] for Amaranthus, [39] for Pastinaca and [40] for Carduus and can be summarized as follows: Amaranthus retroflexus is an annual species of highly disturbed habitats. It is a neophyte to the Czech flora, with its first record of occurrence from 1818 [41]. Amaranthus retroflexus is a self-compatible and probably primarily self-pollinated species, able to produce huge numbers of seeds, depending on the growing conditions. Seed production varies between tens and hundreds of thousands and occasionally reaches millions of seeds per plant. Seeds with impermeable and both mechanically and chemically resistant coats form persistent seed banks. Buried seeds can remain viable for at least 6 to 10 years. This period, however, may vary depending on the depth of burial, soil temperature and moisture, frequency of soil disturbance, and dormancy. Dormancy is regulated primarily by seasonal variation in temperature. Small seeds and a structured seed coat facilitate dispersal by water, animals (mainly humans) and, to a lesser extent, by wind.

Carduus acanthoides is an annual, winter annual or biennial species of moderately disturbed habitats. It is an archaeophyte in the Czech flora [41]. It is primarily an outcrossing species, but considerable variation in the degree of selfing among individuals was recorded. Flowers are pollinated by insects. Seed production varies with climatic conditions and ranges between hundreds and thousands of seeds per head. It reproduces strictly by seeds, which can be dispersed by wind using the attached pappus. The seeds fall within 50 m of the maternal plant, but less than l percent of seeds may be carried further than 100 m apart. Seed viability remains high over several years. Carduus acanthoides survives winter either as seeds or rosettes. The species forms dense stands at disturbed sites where competition is low.

Pastinaca sativa is a biennial or short-lived, monocarpic, perennial species of habitats with a low level of disturbance [42], [43]. It is native to the flora of the Czech Republic [41]. Mostly hermaphroditic flowers of P. sativa are protandrous, promoting outcrossing between different plants. The species is insect-pollinated and able to produce up to 2000 fruits per plant. Its seeds are efficiently dispersed by wind to a distance of several meters from the maternal plant as well as by humans along roadways and highways. The seeds of P. sativa possess morphological dormancy due to an underdeveloped embryo. Most seeds therefore germinate in the following spring [44]. Seeds of this species only form a weak, persistent seed bank; 99 percent of all seedlings typically emerge within 2 years after seed dispersal ([39] and references therein). After emergence, seedlings develop a strong taproot for nutrient storage and overwintering as rosettes.

The Species’ Genetic Diversity

Population genetic diversity of the three species, regardless of their different life history stages, has been investigated earlier. A summary of population genetic diversity across 20 populations of Amaranthus retroflexus and Carduus acanthoides has been published by Mandák et al. ([45], [46], respectively). Population genetic data for Pastinaca sativa has not been published so far, but the sampling strategy (20 localities and 15 plants regularly sampled per locality) was same as in Mandák et al. [45], [46] (see Table 7 for overview of genetic diversity of individual species). The three species strongly differed in their population genetic characteristics. The most genetically impoverished species was the annual invasive species Amaranthus retroflexus for which high inbreeding and among population differentiation was indicated (Table 7). However, Mandák et al. [45] demonstrated that A. retroflexus supports a larger amount of genetic variation than other widespread selfing species. Although both Carduus acanthoides and Pastinaca sativa exhibited low levels of inbreeding, populations of Pastinaca sativa were more structured than those of Carduus acanthoides (Table 7). The low θ(FST) value is obviously the result of the high dispersal capacity of Carduus acanthoides [46] in comparison to Pastinaca sativa, which lacks any apparatus that would enable dispersal over long distances. In the present study, we therefore compared individual life history stages of species that differ not only in their life history strategies but also in levels and patterns of genetic diversity in order to test whether there are any broad genetic differences between soil seeds and surface plants.

Study Sites

We tested the relationship among plants derived from seeds extracted from the soil seed bank, seedlings, rosettes and mature plants over a range of localities and species. We took samples at four different localities of our species’ ranges. Populations were selected on the basis of known genetic variability (see above), i.e. the most genetically variable populations were selected (Fig. 4). These differed in habitat dynamics and population size or history (see Table 8, Fig. 4). This minimized the risk that our data would be distorted by attributes of one particular locality.

Sample Collection and Allozyme Analysis

In four natural populations of our three species, we took 40 samples of each life history stage from regularly distributed permanent plots. To ensure sampling consistency, we used the same sampling strategy for all three species. To obtain samples from different phases of the life history cycle for our genetic analysis, we sampled several consecutive life history stages: (1) the winter soil seed bank (i.e. seed bank before the spring germination peak), (2) seedlings, (3) rosettes, (4) the summer soil seed bank (i.e. seed bank after the spring germination peak) and (5) fruiting plants (i.e. plants that survived self-thinning as well as the competition from the surrounding vegetation, see also [15]). The number of sampled stages differed in relation to the species and its life history cycle (see Table 2). In the case of the annual species Amaranthus retroflexus, we did not sample the rosette life history stage. In the perennial Carduus acanthoides, we sampled only the rosette life history stage instead of the seedling stage because we were unable to differentiate between large seedlings and small rosettes in the field. In Pastinaca sativa, we were able to sample all stages because they strongly differed among each other. We did not succeed in obtaining the summer seed banks from four populations: Pocedelice (Amaranthus), Orech (Carduus), and Ostopovice and Sedlec (Pastinaca).

Soil samples of the winter seed bank were collected in February, of the summer seed bank in July. Amaranthus retroflexus was sampled in 2006, Carduus acanthoides in 2007 and Pastinaca sativa in 2008. First, we created a 15 by 10 m grid with gridlines spaced at 5 m intervals. We then took soil samples of winter seed bank at the intersections of the gridlines by digging up soil blocks that were 20 cm long, 20 cm wide and 10 cm deep. We recorded the spatial coordinates of each sample. Soil samples were moved to the Experimental garden of the Institute of Botany, Academy of Sciences, Průhonice, Czech Republic (49°59′30′′N, 14°34′00′′E, ca 320 m a.s.l.). Each soil block was slightly disturbed to increase the probability of seed germination and placed on a layer of sterilized soil in a 25×25×25 cm garden pot. Emerging seedlings were harvested and transplanted into small pots. After 14 days, the soil was disturbed once again and sprayed with a gibberellic acid solution to break dormancy and speed up as well as synchronise germination of ungerminated seeds. From each population, forty seedlings were randomly harvested, grown to maturity and subjected to allozyme electrophoresis. We believe that we were able to establish plants from most seeds stored in the seed banks. Using a gibberellic acid solution, we broke dormancy of a majority of ungerminated dormant seeds because only a minimal portion of the seeds remained in the soil. We have to admit, however, that our results could be influenced by a bias in the differential establishment rates of seeds from different seed banks collected from different populations and different species. The same problem was encountered by Mandák et al. for Atriplex tatarica [22]. They pointed out that this possibility is to some extent a technical issue because seedlings must grow large enough before they can be used for allozyme analysis, and many plants experience some degree of early mortality. On the other hand, the use of small seedlings is impossible because of the low enzyme activity in this life stage as found in Atriplex tatarica.

We then collected subsequent life history stages within the grid established during the winter seed bank soil sample collection. We randomly collected 4 seedlings, 4 rosettes and 4 fruiting plants around each of ten intersections of the gridlines but no further than 5 m away. Seedlings and rosettes were transported to the Experimental garden where they were grown to maturity and subjected to allozyme electrophoresis. From each fruiting plant, a single young expanded leaf was collected in the field, kept in a cool box and transported to the laboratory for subsequent analysis. Soil samples of summer seed banks were taken and processed in the same way as in the winter seed bank.

In total, 1986 samples were analysed, i.e., 759 in Carduus acanthoides, 561 in Amaranthus retroflexus and 666 in Pastinaca sativa. All samples were analysed by gel electrophoresis using selected enzymes (for methodological details of analysing Carduus acanthoides, see [46], for Amaranthus retroflexus, see [45]). Data for Pastinaca sativa have not yet been published. We have used the following enzymatic systems: Lap, Shdh, Nadh-Dh, 6-Pgdh-1Aat-1, Aat-2, Pgm and Est, using the same methodology as in Mandák et al. [45], [46]. For a summary of the enzymatic systems used, see Table 3.

Statistical Analysis

To estimate genetic diversity and genetic structure, a locus was considered polymorphic if the frequency of the most common allele did not exceed 0.95. Genetic diversity parameters, i.e. allelic richness (RS), observed heterozygosity (HO) and gene diversity (HS) were estimated using the software FSTAT [47]. Population structure was analysed using F statistics [48] following the method of Weir and Cockerham [49] by FSTAT. Wright’s F statistics [48] include f(FIS), which represents the inbreeding index due to non-random mating within populations, and θ(FST), which represents the fixation index due to population subdivision. This method allowed us to test whether these F statistics were significantly different from zero (determined by 5000 bootstrap replicates). Probability values for differences between various life history stages are given for two-sided t-tests after 10 000 permutations. All analyses were performed using FSTAT.

To test for differences in allelic composition among stages and populations, we ran an exact test of population differentiation [50] with the programme TFPGA [51], using 1000 dememorization steps, 20 batches and 2000 permutations per batch. This test was applied to a contingency table (Fisher’s R×C test) and a Markov chain Monte Carlo approach to determine whether significant differences in allele frequencies exist between stages or populations.

Genetic variation at the level of populations and stages was investigated with a nested analysis of molecular variance (stages nested within populations; AMOVA [52]). Levels of genetic differentiation were measured by FCT, FSC, and FST, referring to the differentiation among populations, among stages within populations and within stages, respectively. Levels of significance were determined by computing 1000 random permutation replicates.

Permissions

No specific permits were required for the described field studies, which took place on localities with public right-of-way and did not involve endangered or protected species.

Supporting Information

Table S1.

Allele frequencies of Amaranthus retroflexus. Allele frequencies at six loci in four populations and four life history stages of Amaranthus retroflexus are given (.pdf).

doi:10.1371/journal.pone.0049471.s001

(PDF)

Table S2.

Allele frequencies of Carduus acanthoides. Allele frequencies at eight loci in four populations and life history stages of Carduus acanthoides are given (.pdf).

doi:10.1371/journal.pone.0049471.s002

(PDF)

Table S3.

Allele frequencies of Pastinaca sativa. Allele frequencies at eight loci in four populations and five life history stages of Pastinaca sativa are given (.pdf).

doi:10.1371/journal.pone.0049471.s003

(PDF)

Acknowledgments

The authors thank Michaela Knotková and Martina Hadincová for their excellent technical, field and laboratory assistance, Karin Sramek deKott and Adéla Macková for their significant help with the laboratory work, Petr Dostál for his help in the field, and Frederick Rooks for kindly improving our English.

Author Contributions

Conceived and designed the experiments: BM. Performed the experiments: BM PZ VM IP. Analyzed the data: BM. Contributed reagents/materials/analysis tools: IP PZ. Wrote the paper: BM VM.

References

  1. 1. Roberts HA (1981) Seed banks in soils. Adv Appl Biol 6: 1–55.
  2. 2. Telewski FW, Zeevaart JAD (2002) The 12-years period for Dr. Beal’s seed viability experiment. Am J Bot 89: 264–270. doi: 10.3732/ajb.89.8.1285
  3. 3. Thompson K, Bakker JP, Bekker RM (1997) The Soil Seed Banks of North West Europe: Methodology, Density and Longevity. Cambridge: Cambridge University Press. x +276 p.
  4. 4. Baskin CC, Baskin JM (1998) Seeds – Ecology, Biogeography, and Evolution of Dormancy and Germination. San Diego: Academic Press. ivx +266 p.
  5. 5. Thompson K, Grime JP (1979) Seasonal variation in the seed banks of herbaceous species in ten contrasting habitats. J Ecol 67: 893–921. doi: 10.2307/2259220
  6. 6. Leck AS, Parker VT, Simpson RL (1989) Ecology of Soil Seed Banks. San Diego: Academic Press. 462 p.
  7. 7. Fenner M (1985) Seed Ecology. New York: Chapman & Hall. 151 p.
  8. 8. Templeton AR, Levin DA (1979) Evolutionary consequences of seed pools. Am Nat 114: 232–249. doi: 10.1086/283471
  9. 9. Morris AB, Baucom RS, Cruzan MB (2002) Stratified analysis of the soil seed bank in the cedar glade endemic Astragalus bibullatus: evidence for historical changes in genetic structure. Am J Bot 89: 29–36. doi: 10.3732/ajb.89.1.29
  10. 10. Hairston NG, De Stasio BT (1988) Rate of evolution slowed by a dormant propagule pool. Nature 336: 239–242. doi: 10.1038/336239a0
  11. 11. Cabin RJ (1996) Genetic comparison of seed bank and seedling populations of the desert mustard Lesquerella fendleri. Evolution 50: 1830–1841. doi: 10.2307/2410740
  12. 12. Cabin RJ, Mitchell RJ, Marshall DL (1998) Do surface plant and soil seed bank populations differ genetically? A multipopulation study of the desert mustard Lesquerella fendleri (Brassicaceae). Am J Bot 85: 1098–1109. doi: 10.2307/2446343
  13. 13. McCue KA, Holtsford TP (1998) Seed bank influences on genetic diversity in the rare annual Clarkia springvillensis (Onagraceae). Am J Bot 85: 30–36. doi: 10.2307/2446551
  14. 14. Tonsor SJ, Kalisz S, Fischer J (1993) A life history based study of population structure: seed bank to adults in Plantago lanceolata. Evolution 47: 833–843. doi: 10.2307/2410187
  15. 15. Mandák B, Bímová K, Mahelka V, Plačková I (2006) How much genetic variation is stored in the seed bank? A study of Atriplex tatarica (Chenopodiaceae). Mol Ecol 15: 2653–2663. doi: 10.1111/j.1365-294x.2006.02953.x
  16. 16. Mahy G, Vekemans X, Jacquemart AL (1999) Patterns of allozymic variation within Calluna vulgaris populations at seed bank and adult stages. Heredity 82: 432–440. doi: 10.1038/sj.hdy.6884990
  17. 17. Falahati-Anbaran M, Lundemo S, Ågren J, Stenøien KH (2011) Genetic consequences of seed banks in the perennial herb Arabidopsis lyrata subsp. petraea (Brassicaceae). Am J Bot 98: 1475–1485. doi: 10.3732/ajb.1100021
  18. 18. Alvarez-Buylla ER, Chaos Á, Pińero D, Garay AA (1996) Demographic genetics of a pioneer tropical tree species: patch dynamics, seed dispersal, and seed banks. Evolution 50: 1155–1166. doi: 10.2307/2410656
  19. 19. Koch M, Huthmann M, Bernhardt KG (2003) Cardamine amara L. (Brassicaceae) in dynamic habitats: Genetic composition and diversity of seed bank and established populations. Basic Appl Ecol 4: 339–348. doi: 10.1078/1439-1791-00165
  20. 20. Vitalis R, Glémin S, Olivieri I (2004) When genes go to sleep: the population genetic consequences of seed dormancy and monocarpic perenniality. Am Nat 163: 295–311. doi: 10.1086/381041
  21. 21. Kalisz S (1989) Fitness consequences of mating system, seed weight, and emergence date in a winter annual, Collinsia verna. Evolution 43: 1263–1272. doi: 10.2307/2409361
  22. 22. Mandák B, Bímová K, Plačková I (2006) Genetic structure of experimental populations and reproductive fitness in a heterocarpic plant Atriplex tatarica (Chenopodiaceae). Am J Bot 93: 1640–1649. doi: 10.3732/ajb.93.11.1640
  23. 23. Lundemo S, Falahati-Anbaran M, Stenøien HK (2009) Seed banks cause elevated generation times and effective population sizes of Arabidopsis thaliana in northern Europe. Mol Ecol 18: 2798–2811. doi: 10.1111/j.1365-294x.2009.04236.x
  24. 24. Cohen D (1966) Optimizing reproduction in a randomly varying environment. J Theor Biol 12: 119–129. doi: 10.1016/0022-5193(66)90188-3
  25. 25. Cohen D (1967) Optimizing reproduction in a randomly varying environment when a correlation may exist between the conditions at the time a choice has to be made and the sibsequent outcome. J Theor Biol 16: 1–14. doi: 10.1016/0022-5193(67)90050-1
  26. 26. Venable DL, Brown JS (1988) The selective interactions of dispersal, dormancy and seed size as adaptations for reducing risk in variable environments. Am Nat 131: 360–384. doi: 10.1086/284795
  27. 27. Venable DL (1989) Modeling the evolutionary ecology of seed banks. In: Leck MA, Parker VT, Simpson RL, editors. Ecology of Soil Seed Banks. San Diego: Academic Press. 67–87.
  28. 28. Nunney L (2002) The effective size of annual plant populations: The interaction of seed bank with fluctuating population size in maintaining genetic variation. Am Nat 160: 195–204. doi: 10.1086/341017
  29. 29. Pacala SW (1986) Neighborhood models of plant populations dynamics. 2. Multi-species models of annuals. Theor Popul Biol 29: 262–292. doi: 10.1016/0040-5809(86)90011-0
  30. 30. Rees M (1996) Evolutionary ecology of seed dormancy and seed size. Philos Trans R Soc Lond B Biol Sci 351: 1299–1308. doi: 10.1098/rstb.1996.0113
  31. 31. Caswell H, Hastings A (1980) Fecundity, developmental time, and population growth rate: an analytical solution. Theor Popul Biol 17: 71–79. doi: 10.1016/0040-5809(80)90015-5
  32. 32. Mandák B, Pyšek P (2001) Fruit dispersal and seed banks in Atriplex sagittata: the role of heterocarpy. J Ecol 89: 159–165. doi: 10.1046/j.1365-2745.2001.00536.x
  33. 33. Mandák B, Pyšek P (2001) The effects of light quality, nitrate concentration and presence of bracteoles on germination of different fruit types in the heterocarpous Atriplex sagittata. J Ecol 89: 149–158. doi: 10.1046/j.1365-2745.2001.00537.x
  34. 34. Thrall PH, Pacala SW, Silander JA (1989) Oscillatory dynamics in populations of an annual weed species Abutilon theophrasti. J Ecol 77: 1135–1149. doi: 10.2307/2260828
  35. 35. Hock S, Szövényi P, Schneller JJ, Tóth Z, Urmi E (2008) Bryophyte diaspore bank: a genetic memory? Genetic structure and genetic diversity of surface populations and diaspore bank in the liverwort Mannia fragrans (Aytoniaceae). Am J Bot 95: 542–548. doi: 10.3732/ajb.2007283
  36. 36. Honnay O, Bossuyt B, Jacquemyn H, Shimono A, Uchiyama K (2008) Can a seed bank maintain the genetic variation in the above ground plant population? Oikos 117: 1–5. doi: 10.1111/j.2007.0030-1299.16188.x
  37. 37. Mandák B, Plačková I (2009) How does population genetic diversity change over time? An experimental seed bank study of Atriplex tatarica (Chenopodiaceae). Flora 204: 423–433. doi: 10.1016/j.flora.2008.05.005
  38. 38. Costea M, Weaver SE, Tardif FJ (2004) The biology of Canadian weeds. 130. Amaranthus retroflexus L., A. powellii S.Watson and A. hybridus L. Can J Plant Sci 84: 631–668. doi: 10.4141/p02-183
  39. 39. Cain N, Darbyshire SJ, Francis A, Nurse RE, Simard MJ (2010) The Biology of Canadian weeds. 144. Pastinaca sativa L. Can J Plant Sci 90: 217–240. doi: 10.4141/cjps09110
  40. 40. Desrochers AM, Bain JF, Warwick SI (1988) The biology of Canadian weeds. 89. Carduus nutans L. and Carduus acanthoides L. Can J Plant Sci 68: 1053–1068. doi: 10.4141/cjps88-126
  41. 41. Pyšek P, Sádlo J, Mandák B (2002) Catalogue of alien plants of the Czech Republic. Preslia 74: 97–186.
  42. 42. Hendrix SD (1984) Variation in seed weight and its effects on germination in Pastinaca sativa L. (Umbeliferae). Am J Bot 71: 795–802. doi: 10.2307/2443470
  43. 43. Hendrix SD, Trapp EJ (1992) Population demography of Pastinaca sativa (Apiaceae) – effects of seed mass on emergence, survival, and recruitmnent. Am J Bot 79: 365–375. doi: 10.2307/2445148
  44. 44. Baskin JM, Baskin CC (1979) Studies on the autecology and population biology of the weedy monocarpic perennial, Pastinaca sativa. J Ecol 67: 601–610. doi: 10.2307/2259115
  45. 45. Mandák B, Zákravský P, Dostál P, Plačková I (2011) Population genetic structure of the noxious weed Amaranthus retroflexus in Central Europe. Flora 206: 697–703. doi: 10.1016/j.flora.2011.01.010
  46. 46. Mandák B, Zákravský P, Kořínková D, Dostál P, Plačková I (2009) Low population differentiation and high genetic diversity in the invasive species Carduus acanthoides L. (Asteraceae) within its native range in the Czech Republic. Biol J Linn Soc 98: 596–607. doi: 10.1111/j.1095-8312.2009.01304.x
  47. 47. Goudet J (1995) FSTAT (version 1.2): a computer program to calculate F-statistics. J Hered 86: 485–486.
  48. 48. Wright S (1965) The interpretation of population structure by F-statistics with special regard to system mating. Evolution 19: 395–420. doi: 10.2307/2406450
  49. 49. Weir BS, Cockerham CC (1984) Estimating F-statistics for the analysis of population structure. Evolution 38: 1358–1370. doi: 10.2307/2408641
  50. 50. Raymond M, Rousset F (1995) An exact test for population differentiation. Evolution 49: 1280–1283. doi: 10.2307/2410454
  51. 51. Miller MP (1997) Tools for population genetic analyses (TFPGA) 1.3: A Windows program for the analysis of allozyme and molecular population genetic data. Available: http://www.marksgeneticsoftware.net/.
  52. 52. Excoffier L, Laval G, Schneider S (2005) Arlequin (version 3.0): An integrated software package for population genetics data analysis. Evol Bioinform Online 1: 47–50.