The Journal of Heredity 2001:92(4)
© 2001 The American Genetic Association 92:339-345
Genetic Diversity of the Macaronesian Leafy Liverwort Porella canariensis Inferred From RAPD Markers
From the Center of Biological and Geological Sciences, Campus of Penteada, 9000 Funchal, Portugal.
Address correspondence to António Brehm at the address above or e-mail: brehm;cauma.pt.
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Plant colonization of the North Atlantic raises the intriguing question of the relationships between extant island species with their continental counterparts (European, African, and American), which may provide clues to past geographic distribution and colonization history. It has been suggested that during past glaciations, many plant species with typical Mediterranean distributions survived in the Atlantic islands that belong to what is today known as Macronesia. We used random amplified polymorphic DNA (RAPD) markers to study 12 populations of the liverwort Porella canariensis partly covering its present-day distribution (Azores, Madeira, Canary and Cape Verde Islands, and Iberian Peninsula). Unweighted pair-group (UPGMA) and principal component (PCO) analyses showed a similar geographical pattern that suggested a close relationship between Iberian populations and those from the Canaries and Cape Verde Islands. Populations from Madeira had more genetic variation than those from the Azores, a result from either a richer diversity of habitats in Madeira, which prompted more population diversification, successive colonization waves from different origins, or an older colonization of Madeira. The data show that continuous patches of liverworts are often comprised of more than one individual. Finally, RAPDs can be used to investigate intraspecific diversity within a comparatively large geographic area and, with utmost care, can be used to infer a historic context to explain the patterns observed.
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
In plant species, the amount of gene flow is a major factor determining genetic variation found throughout the species' distribution. Gene flow depends on such factors as reproductive mode, breeding system, and selective constraints (Loveless and Hamrick 1984; Mishler 1988). In diploid, outcrossing species it is common that genetic variation among populations is less than that within populations (Loveless and Hamrick 1984). On the other hand, species that are primarily selfed, but short-lived, have most of their genetic diversity among populations (Hamrick and Godt 1996). Species with limited dispersal or that reproduce by selfing present a pattern of geographic genetic variation where drift may play the primary role. Such could be the case of the Macaronesian liverwort Porella canariensis.
The Macronesia region consists of the Atlantic islands of the Azores, Madeira, Canary and Cape Verde. It harbors an extremely rich endemic and relict flora which in some places fully characterizes the main elements of the Tertiary European flora (Laurasia). Madeira harbors an extremely well-preserved Tertiary forest called Laurisilva which also exists in a much smaller extension in the Canary Islands. In the Azores, Cape Verde Islands, and Iberia, few elements of this typical forest are present. There is some evidence, however, that Laurisilva also existed on the coast of Morocco (Aubreville 1976). The intermediate position of these archipelagos between three continents makes them extremely interesting for studies of relict taxa of Gondwanalandic or Laurasian origin. Many of these taxa belong to the Macaronesian bryophyte flora, which includes more than 600 species, some of them with worldwide enigmatic distributions. Of these, 9% are exclusively endemic, some with direct relationships to other regions of bryophyte endemics as far away as Australia (Sáergio 1984).
The leafy liverwort genus Porella comprises more than 100 species with a worldwide distribution. Most of the taxa are found in southwestern Asia, where the genus is believed to have originated (Hattori 1978). In Europe, only eight species are currently recognized. The genus is known because of its wide phenotypic plasticity and morphologic variation which renders classification based on morphology problematic (Hattori 1978). Molecular techniques have proved invaluable for detecting genetic differences among taxa (Boisselier-Dubayle and Bischler 1989, 1994; Therrien et al. 1998). P. canariensis is an endemic species present on all Macronesian islands. Eggers (1982) lists it for seven Azorean and five Canarian islands, but not Cape Verde and Deserta Island in Madeira. On the Iberian Peninsula, the species is found in Sintra (Portugal) and Algeciras (south of Spain), but Jelenc (1955, in Ros et al. 1999) lists it for North Africa, a distribution that suggests a dual origin for this species: Mediterranean and African. P. canariensis occurs in small isolated colonies or patches, depending on the availability of specific ecological requirements such as atmospheric humidity and shade. Patches grow horizontally on trees or rocks to 10–50 cm in diameter and are found from sea level to 1700 m altitude.
The founding and genetic composition of these colonies is a matter of discussion (Bischler and Boisselier-Dubayle 1997 and references herein). They may start with one or several individuals reproducing vegetatively or sexually. It is accepted that these colonies are apparently reproductively isolated (Dewey 1989; Szweykowski 1982; Yamazaki 1982; Zehr 1980). Like most liverworts, P. canariensis has a life cycle characterized by the presence of a dominant haploid gametophyte and a short-lived diploid sporophyte. Spore dispersal is usually limited to short distances (Jonsson and Söderström 1988; Levin 1979) and sperm dispersion by water does not go far (from 4 to 65 cm) (Bischler and Boisselier-Dubayle 1997; Wyatt 1985). It is widely accepted that liverworts have limited dispersal abilities, in part because they easily reproduce vegetatively and most of their life cycle is haploid. Vegetative propagation is perhaps more common than sexual reproduction, and isolation of populations should be very effective, even within short distances. Transport of propagules requires the existence of nearby streams, which are not always present. On the other hand, it has been shown that even low rates of sexual recruitment are sufficient to maintain some genetic diversity (Watkinson and Powell 1993). Also, allozyme studies have shown that clonal plants present a remarkable genetic variation, even with low levels of sexual recruitment (Ellstrand and Rouse 1987; Widén et al. 1994).
If propagation of P. canariensis is mainly vegetative, drift could be the main force determining intraspecific differentiation, and high levels of homogeneity would be expected within populations. Self-fertilization within isolated colonies, especially if colonies are founded by a few individuals, tends to decrease clonal diversity. There is no strong evidence that selection plays a substantial role in determining any genetic variation among colonies of P. canariensis (Bischler and Boisselier-Dubayle 1997). Traditional methods of genetic analysis (isozyme or morphology) applied to these liverworts often fail to reveal variation in most conspecific populations, which implies restricted gene flow (Boisselier-Dubayle et al. 1995; but see a contrary view for other bryophytes in Stoneburner et al. 1991). This is consistent with a pattern of dispersal and founder events (Shaw 1991; Shaw and Schneider 1995). As an alternative, Wyatt et al. (1989b) proposed that intrapopulation differentiation could be due to selection over different genotypes acting because of environmental heterogeneity. Boisselier-Dubayle and Bischler (1994), working with random amplified polymorphic DNA (RAPD) markers, reported the existence of some intraspecific genetic variation in several species of European Porella.
To bring new insights to the debate over the genetic composition and breeding systems of P. canariensis populations, we used RAPD (Williams et al. 1990) banding patterns as markers to analyze intra- and interspecific genetic variation. These markers, presumed to be selectively neutral, should indicate that any pattern of variation found between different populations would be caused by historical effects, drift, and/or migration. This way, variation within populations must be determined by the mating system of the liverwort. The genetic relationships among the populations studied, partly covering the species distribution range, will be addressed which may offer insights into the dispersion and colonization history of P. canariensis in Macaronesia.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Plant Material
Any genetic analysis of liverworts like Porella requires a clear definition of "population." We follow Bischler and Boisselier-Dubayle (1997) and define it as a continuous colony or "carpet" of liverworts.
Ten populations of P. canariensis each represented by a patch 20–30 cm in diameter and belonging to the four Macronesian archipelagos (Azores, Madeira, Canary Islands, and Cape Verde) and two from Portugal (Sintra) were used in this study (Table 1, Figure 1). The number of individuals ("shoots" or "ramets") from each population were randomly taken from each patch and are given in Table 1. Locations of the sampling sites are shown in Figure 1.
|
|
DNA Extraction
Before DNA extraction took place, all specimens were carefully washed and checked under the microscope to be sure that no possible contaminants (microalgae, fungi) were left. DNA extraction of specimens about 2–3 cm long and weighing just 10 mg required modification of several standard protocols (Doyle and Doyle 1988; Murray and Thompson 1980; Rogers and Bendich 1985). We ground 10 mg of fresh material (equivalent to one individual of Porella) to a fine powder in an Eppendorf tube with a glass rod and liquid nitrogen (tubes are kept with the tissues and filled with nitrogen for only a few s). The powder was then mixed with 500 ml CTAB buffer (100 mM Tris-HCl, pH 8.0; 1.4 M NaCl; 2% p/v CTAB; 20 mM EDTA4Na), 1% ß-mercaptoethanol, 2% proteinase K (20 mg/ml), and then incubated in a bath at 65°C for 30 min. We added 335 ml of chloroform and 14 ml of 1-octanol to this mixture, followed by centrifuging 15 min at 13,000g. The supernatant was transferred to new Eppendorf tubes, mixed with 2/3 (v/v) isopropanol, and centrifuged for 15 min at 13,000g. Supernatant was discarded and the pellet washed with a solution of 76% ethanol AcNH4 10 mM for 20 min. This solution was centrifuged again for 10 min at 13,000g, the aqueous phase discarded, and the pellet resuspended in 100 ml of ultrapure water. Ten milliliters of AcNa 3 M (pH 5.2) and 250 ml of -20°C absolute ethanol were added, followed by a gentle mix and centrifugation for 15 min at 13,000g. The aqueous phase was removed and the pellet washed again with 70% ethanol and centrifuged as before. Finally, the aqueous phase was discarded and the pellet resuspended in 100 ml of ultrapure water or Tris-EDTA. This procedure recovered at least 0.5 mg of good-quality DNA for RAPD analysis.
Primers and PCR
Ten-mer primers from OPERON Technologies (Alameda, CA; series OPA-A,F,G,H,I,J) were screened on individual representatives of the populations under study. Many of the primers produced either complex banding patterns or nonreproducible and inconsistent amplification products. Hence only 10 primers were chosen for the subsequent analysis. For each primer, only bands that could be unambiguously interpreted across all the population samples were chosen. Reproducibility of bands was assessed by replicating amplifications of samples selected at random. Polymerase chain reaction (PCR) was carried out in 20 µl volumes using 2 µl DNA (20 ng/µl), 2 µl dNTP mix (2 mM of each of four nucleotides; Promega), 2 µl 10x Taq DNA buffer (Pharmacia), 2 µl oligonucleotide primer (5 pmol/µl), 1 unit Taq DNA polymerase (Pharmacia), and the remainder of water. PCR conditions were 1 min at 92°C, 1 min at 36°C, and 2 min at 72°C, for 45 cycles, performed in a Biometra thermal cycler. Amplified products were separated in 1.2% agarose gels, stained with ethidium bromide, and visualized under ultraviolet (UV) light. The 100 BPL (Pharmacia) was used as a molecular size marker.
Data Analysis
RAPD profiles were scored for each individual as discrete characters (presence or absence of amplified products) across all individuals from all populations and for each primer used. Table 1 shows chosen parameters for genetic diversity within populations: the percentage of polymorphic sites (markers), the genetic diversity calculated as the ratio of the number of polymorphic RAPD markers found in a population over the total number of individuals sampled, and Weir's D gene diversity. The second parameter may be less influenced by the sample size, but Weir's D is a good indicator of variability especially for telling species (Weir 1996). An analysis of molecular variance (AMOVA) was performed based on Euclidean distances between all pairs of haplotypes according to Excoffier et al. (1992) and using the Arlequin program (Schneider et al. 1997). The total genetic variation could be divided into three levels of grouping using the five recognizable groups of populations: among regions (archipelagos), among populations within regions, and among individuals within populations. AMOVA components were tested for significance by nonparametric randomization tests using 6000 permutations under the null hypothesis of no population structure. A test for differences in intrapopulation molecular variances was also performed based on Bartlett's statistics (Excoffier et al. 1992; Stewart and Excoffier 1996). Additional AMOVAs were carried out to investigate the among-group variances allocated for each of the five groups of populations. In this case, only two groups were considered, the particular one being tested and the remaining populations combined. Based on nested variance components, the AMOVA gave 
statistics, thus allowing the use of ST values (an estimate of FST; Excoffier et al. 1992) as interpopulation differentiation measures and the computation of Nem [Nem = 1/4(1/ST - 1)] as an estimator of gene flow (Slatkin and Barton 1989; Wright 1951). Significance of ST values were determined using a nonparametric permutation procedure. Variation among all pairs of samples was evaluated according to the index of genetic similarity of Nei and Li (1979), which is solely based on the shared presence of characters and excludes shared absence as a criterion of similarity. Relationships among populations were evaluated via the unweighted pair-group method (UPGMA) of Sneath and Sokal (1973) using Nei's (1978) genetic distance, and principal component (PCO) analysis. All analyses were performed using NTSYS (Rohlf 1993).
Correlation of pairwise ST values with geographic distances separating the populations (within and among sites) was assessed using a Mantel test. The significance of Mantel's Z statistics was calculated through a normalized approximation.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The 10 primers amplified 80 RAPD bands in the 222 individuals of the 12 populations studied. The number of polymorphic sites varied considerably among the populations studied (Table 1). Total genetic diversity, estimated as the ratio of polymorphic sites within a population to the number of individuals sampled (Table 1), had the lowest value in the Canaries (1.2) and the highest in Cape Verde (3.7). Weir's (1986) D values of gene diversity varied from 0.506 in MA5 to 0.832 in CA2, but the values did not show any pattern of geographic association. No relationship was found between genetic diversity with either the type of associated vegetation (ANOVA, F = 0.01, ns) or altitude (F = 1.4, ns). Markers within populations were found to be randomly associated (data not shown). The AMOVA results (Table 2) indicate significant genetic differences between regions (P < .0002) and between populations within regions (P < .0002). Of the total genetic divergence, only 19.3% was due to differences among the five regions (four archipelagos and the Iberian populations). Most was attributed to differences among populations within an archipelago (41%) and variation within populations (40%). Genetic variation within archipelagos was almost equally divided among populations (58.6%) and within populations (41.3%). Finally, more of the variation among archipelagos was due to differences within (64.1%) rather then between (35.8%) them. Bartlett's heteroskedasticity indices showed that variance heterogeneities differed significantly (P < .0001) among populations (Bp = 11.73) and among archipelagos (Bg = 4.21). Separate AMOVAs done with all groups pooled except one (Table 3) showed that the Cape Verde archipelago contributed a lower percentage of variation over the total (0.83), and Madeira contributed the highest (14.3). Analysis of variance done for each archipelago independently, showed that variance within populations was higher than variance among populations in the Azores (68.06% and 31.94%, respectively) and Sintra (52.07 and 47.93), but not in Madeira (47.98 and 52.02) and the Canaries (32.17 and 67.83).
|
|
All
ST pairwise values were significant (P = .05) and ranged from 0.32 between populations AZ1 and AZ2 to 0.82 between MA3 and CA2 (data not shown). Consequently the associated Nem values were always less than 1, suggesting a low migration rate between populations. The overall value of ST (0.603; Table 2) suggests significant heterogeneity for Porella populations, indicating that the amount of differentiation is not due to heterogeneity among archipelagos. UPGMA and PCO analyses revealed similar structuring of P. canariensis populations, identifying the corresponding geographical regions (Figures 2 and 3). Figure 2 includes two cluster analyses, one with Nei's (1978) genetic distance between the 12 populations and the other resulting from Nei and Li's (1979) similarity index between all 222 individuals. Because of the impracticability of presenting the latter tree with so many individuals, we collapsed all individual branches from the same population to the nearest common node. This made it possible to compare both dendrograms. In both cases the five populations from Madeira appear separated from the others, in one clade. The major difference between both topologies is the position of the Azores and Cape Verde as outgroups for the closely related Canary/Sintra populations. In the PCO analysis (Figure 3), the first axis accounted for 21.3% of the total variance and clearly separated the Madeira population, and to a lesser extent Cape Verde, from the others. This corresponds well to the first clade on the UPGMA analysis (Figure 2). The second axis accounted for 14.9% of the variation and isolated both populations from the Azores. The third component (not shown) accounted for 12% of variation and separated population CA2 from the group Canary/Sintra. This component also discriminated population MA3 from the remaining Madeira populations. Both these events are clearly depicted in the UPGMA clustering. The existence of three groups of populations—Madeira, Azores, and the others—is clearly distinguished by the PCO. Populations from Sintra were closer to the Canary ones, and Cape Verde occupied an intermediate position between this last group and Madeira. Finally, pairwise values between populations did not show any significant correlation with the geographic distance separating them, as revealed by a Mantel test (r = 0.103, t test = 0.45, ns).
|
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
A substantial amount of variation in RAPD profiles of P. canariensis was found within populations (39.6%) and among populations within each archipelago (41%). Much less variation (19%) was attributed to differences among regions (archipelagos), in spite of the fact that big geographic distances separate them. Similar results were found by Martin et al. (1997) and Sale et al. (1996), but Liao and Hsiao (1998) and Hsiao and Lee (1999) found that variation among regions contributed more than variation among populations within regions to the total variation. Buso et al. (1998) found 77% of variation between regions versus 12% due to variation within regions. Gurgeli et al. (1999) reported 95% of variation within populations and only 1% among regions in the outcrossing and monoclonal Saxifraga oppositifolia. No congruence was found by these authors between genetic and geographic distances. In our analysis of P. canariensis, variation within populations in RAPD haplotypes (which made every ramet in the patches sampled a different genotype) contradicts the homogeneity found in morphology, even between populations that are localized in different environments. We have sequenced several individuals from different populations of P. canariensis for the ITS region of nuclear rDNA and found no polymorphism (unpublished results).
The high variation within populations revealed by the AMOVAs and the fact that no identical haplotypes existed within any colony indicates that the differences observed might correspond to different individuals sampled from the same patch. An alternative explanation could be that Porella, which reproduces vegetatively during most of its long haploid life cycle, accumulates mutations, producing an extremely rich clonal diversity. Individuals sampled in the adult haploid phase would make mutations easily detectable, explaining the high variability found. Waycott (1998), in a study of uniclonal populations of sea grasses, proposed that long-lived individuals characterized by vegetative growth may have accumulated genotypic differences via the accumulation of somatic mutations and recombination, a fact also predicted in theoretical models by Soane and Watkinson (1979). Gabrielsen et al. (1998) also reported high levels of genetic diversity in the arctic clonal Saxifraga cernua, and Cronberg et al. (1997) explained the high genetic variation found in a clonal bryophyte as due to a perennial stayer life strategy associated with wide wind dispersal of pollen. In contrast, Bauert et al. (1998) found three isolated, relict alpine populations of the same species presented no variation at all. This was explained by postglacial events resulting in bottlenecks or founder events. Also Steinger et al. (1996) reported that patches of slow and clonally growing Carex curvula are genetically uniform. In spite of the fact that the breeding system of P. canariensis is poorly understood, it is known that the species reproduces well vegetatively (Sá-Fontinha S, personal observation) and that each individual can live and grow for a long time. In this case, recombination is possible among individuals of different generations during the short period of their diploid and sexual cycle, and vegetative growth preserves these acquired genotypes.
At least in the Madeira archipelago, mature plants in full reproductive phase have been observed in some populations (patches), but no evidence exists that sexual reproduction is the major factor contributing through recombination to the high level of genetic variability. Observations in other populations, like Porto Santo, only rarely show reproductively mature plants (Sá-Fontinha S, personal observation). The fact that RAPD markers were found to be randomly associated is evidence that some sexual reproduction exists within each patch. The widely accepted lack of substantial spore dispersal in P. canariensis is supported by 41% of the variation being attributed to differences between populations within archipelagos. Such a pattern suggests that P. canariensis is a typical outcrossing species according to Loveless and Hamrick (1984) and Hamrick and Godt (1996). In this case, however, high levels of variation still exist among populations within regions, especially in the case of Madeira, with five populations sampled, which indicates that some reduced gene flow exists. The fact that the species is not widespread, but rather is confined to particular habitat conditions where the colonies can grow, increases variation among populations belonging to the same group. The AMOVA also shows that differences among archipelagos are maximum between the Madeira populations and the others (14%) and minimum between the population of Cape Verde and the others (0.8%). These values may be inflated, however, because the samples from the Canaries are from the same island, and only two islands were screened for the Azores.
RAPD profiles revealed a structuring of populations completely congruent with their geographic location. Variation found within Madeira is high if we compare it to both populations from Tenerife and even to the two Azorean samples on different islands. Both cluster analyses and PCO showed similar grouping of populations and clearly separate the Madeira cluster from the others. In the Madeira group of populations, the two more closely related samples are MA1 on Madeira Island and MA4 on the Desertas Islands. Both populations appear in coastal vegetation which is not the case of MA2, which occurs in Laurisilva forest. In the other group of populations, the two trees slightly differ in the clustering of internal nodes. Nevertheless, both agree in placing together the populations from Sintra and Tenerife (Canaries). The populations from the Azores always cluster together and are the two most similar ones. Their place as an outgroup for the Sintra/Tenerife populations may change with Cape Verde according to the clustering method used. On a strict simple-match basis of similarity, it is clear that Cape Verde joins with both the Tenerife and Sintra populations, and only then joins the two Azores. This has implications for the general pattern of dispersion and colonization history of P. canariensis.
Mediterranean in origin, the species may have occupied a much broader area than it does today. In fact, the data allow us to speculate that P. canariensis may have existed from Iberia to the coast of Cape Verde when the Sahara was green, much of its range then was probably associated with the former Tertiary Laurisilva rain forest. This type of forest is today completely absent from Morocco, but some evidence suggests that it had previously existed (Aubreville 1976). In Cape Verde the last few colonies of P. canariensis are quickly disappearing with desertification. We have only found it on São Vicente Island on a mountain where very few Laurasian elements still exist. Starting from the mainland, colonization of the Cape Verde and Canary Islands would be due to the proximity to the coast, probably using other plant species as carriers.
The results from the AMOVA analyses, together with the lack of correlation between pairwise FST values and geographic distances among the 12 samples studied, indicate that habitat differentiation is more important than large-scale geographic differentiation (isolation by distance) to account for variation in these populations of Porella. High levels of genetic variation found in bryophytes by Wyatt et al. (1989a) have been explained as genetic adaptations to microhabitats. In fact, the variance among regions is smaller than that among populations within a single region. This result could also be due to the fact that more samples were used in Madeira than in the Canaries or the Azores, but similar results were obtained when we performed the analysis with just the three Madeira Island samples. Large distances between the archipelagos and Sintra (> 1400 km between Sintra and Canary, >2600 km between Sintra and Cape Verde, >1000 km from Sintra to either Madeira or Azores), and the fact that many extant populations are rare and confined to specific areas, make present-day gene flow among regions virtually impossible. The relatedness of populations from Sintra, Canary, and Cape Verde as revealed in the dendrograms indicates that part of their RAPD markers are shared and date back to a single large population. Particular habitat conditions on Tenerife probably prevent the two populations analyzed from diverging further. Populations from the Azores may be considered a separate founding event from continental populations. The Azores are the youngest islands and have low habitat diversity. This could explain why the two populations belonging to different islands are separated by a small genetic distance.
Taking into consideration variation within populations of P. canariensis, it is clear that colonies are comprised of more than one individual (i.e., multiclonal). Moreover, besides vegetative reproduction, sexual recombination appears to be more common than is usually admitted. The high variation among populations within regions confirms that spore dispersal, even over short distances, is probably only effective within colonies. Ecological constraints may be a strong factor limiting the possibility for colonies to diverge. Finally, the good agreement obtained between the two dendrograms suggests a continuous, large population of P. canariensis, at least from Iberia to the coast of Cape Verde in mainland Africa, from which the plant colonized the two southern archipelagos. The close relationship between P. canariensis populations from these Atlantic archipelagos and the samples from Sintra corroborates the hypothesis that some of the P. canariensis polymorphism is not recent, but rather dates back to a preglacial widespread population that found refuge in these habitats that were essentially untouched by climatic changes.
|
Acknowledgments |
|---|
The authors thank to S. Sá-Fontinha from Jardim Botânico do Funchal and J. Paiva for providing specimens. J. Jesus, S. Sá-Fontinha, and C. Sérgio provided helpful discussions on the manuscript. This research was supported by FCT contract PBICT/P/BIA/2052/95 (to A.B.). The Centre of Biological and Geological Sciences is supported by Fundação da Ciência e Tecnologia PRAXIS XXI.
|
Footnotes |
|---|
Corresponding Editor: James L. Hamrick
Received January 28, 2000
Accepted January 15, 2001
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
-
Aubreville A, 1976. Centre terciaire d'origine, radiation et migration des flores angiospermiques tropicales. Adansonia 3:297–354.
Bauert MR, Kálin M, Baltisberger M, and Edwards PJ, 1998. No genetic variation detected within isolated relict populations of Saxifraga cernua in the Alps using RAPD markers. Mol Ecol 7:1519–1627.
Bischler H and Boisselier-Dubayle MC, 1997. Population genetics and variation in liverworts. Adv Bryol 6:1–34.
Boisselier-Dubayle MC and Bischler H, 1989. Electrophoretic studies in Marchantia polymorpha L. J Hattori Bot Lab 67:297–311.
Boisselier-Dubayle MC and Bischler H, 1994. A combination of molecular and morphological characters for delimitation of taxa in European Porella. J Bryol 18:1–11.
Boisselier-Dubayle MC, Jubier MF, Lejeune B, and Bischler H, 1995. Genetic variability in the three subspecies of Marchantia polymorpha: isozymes, RFLP and RAPD markers. Taxon 44:363–376.
Buso GS, Rangel PH, and Ferreira ME, 1998. Analysis of genetic variability of South America wild rice populations (Oryza glumaepatula) with isozymes and RAPD markers. Mol Ecol 7:107–117.
Cronberg N, Molau U, and Sonesson M, 1997. Genetic variation in the clonal bryophyte Hylocomium splendens at hierarchical geographical scales in Scandinavia. Heredity 78:293–301.
Dewey RM, 1989. Genetic variation in the liverwort Riccia dictyospora (Ricciaceae, Hepaticopsida). Syst Bot 14:155–167.
Doyle J and Doyle J, 1988. Isolation of plant DNA from fresh tissue. Am J Bot 75:1238.
Eggers J, 1982. Artenliste der moose makaronesiens. Cryptogamie 3:283–335.
Ellstrand NC and Roose ML, 1987. Patterns of genotypic diversity in clonal plant species. Am J Bot 74:123–131.[Web of Science]
Excoffier L, Smouse PE, and Quattro JM, 1992. Analysis of molecular variance inferred from metric distances among DNA haplotypes: application to human mitochondrial DNA restriction data. Genetics 131:479–491.[Abstract]
Gabrielsen TM and Brochmann C, 1998. Sex after all: high levels of diversity detected in the arctic clonal plant Saxifraga cernua using RAPD markers. Mol Ecol 7:1701–1708.
Gurgeli F, Eichenberger K, and Scneller JJ, 1999. Promiscuity in populations of the cushion plant Saxifraga oppositifolia in the Swiss Alps as inferred from random amplified polymorphic DNA (RAPD). Mol Ecol 8:453–461.
Hamrick JL and Godt MW, 1996. Effects of life history traits on genetic diversity in plant species. Philos Trans R Soc Lond B 351:1291–1298.
Hattori S, 1978. Studies on the Asiatic species of the genus Porella (hepaticae). VII. A synopsis of Asiatic Porellaceae. J Hattori Bot Lab 44:91–120.
Hsiao JY and Lee M, 1999. Genetic diversity and microgeographic differentiation of Yushan cane (Yushania niitakayamensis; Poaceae) in Taiwan. Mol Ecol 8:263–270.
Jonsson BG and Söderström L, 1988. Growth and reproduction in the leafy hepatic Ptilidium pulcherrimum (G. Web.) Vainio during a 4-year period. J Bryol 15:315–325.
Levin DA, 1979. The nature of plant species. Science 204:381–384.
Liao LC and Hsiao JY, 1998. Relationship between population genetic structure and riparian habitat as revealed by RAPD analysis of the reophyte Acorus gramineus Soland. (Araceae) in Taiwan. Mol Ecol 7:1275–1281.
Loveless MD and Hamrick JL, 1984. Ecological determinants of genetic structure in plant populations. Annu Rev Ecol Syst 15:65–95.[Web of Science]
Martin C, Gonzalez-Benito ME, and Iriondo JM, 1997. Genetic diversity within and among populations of a threatened species: Erodium paularense Fern. Gonz. & Izco. Mol Ecol 6:813–820.
Mishler BD, 1988. Reproductive ecology of bryophytes. In: Plant reproductive biology. Patterns and strategies (Lovett Doust J and Lovett Doust L, eds). Oxford: Oxford University Press; 285–306.
Murray M and Thompson W, 1980. Rapid isolation of high molecular weight plant DNA. Nucleic Acids Res 8:4321–4325.
Nei M, 1978. Estimation of average heterozygosity and the genetic distance from small number of individuals. Genetics 89:583–590.
Nei M and Li WH, 1979. Mathematical model for studying genetic variation in terms of restriction endonucleases. Proc Natl Acad Sci USA 76:5269–5273.
Rogers S and Bendich A, 1985. Extraction of DNA from plant tissues. Plant Mol Biol 5:69–76.
Rohlf F, 1993. NTSYS-pc: numerical taxonomy and multivariate analysis system. New York: Applied Biostatistics Inc.
Ros R, Cano M, and Guerra J, 1999. Bryological monograph. Bryophyte checklist of northern Africa. J Bryol 21:207–244.
Sale MM, Potts BM, and West AK, 1996. Molecular differentiation within and between Eucalyptus risdonii, E. amygdalina and their hybrid using RAPD markers. Aust J Bot 44:559–569.
Schneider S, Kueffer J-M, Roessli D, and Excoffier L, 1997. Arlequin ver 1.1: a software for population genetic data analysis. Geneva, Switzerland: University of Geneva.
Sérgio C, 1984. The distribution and origin of Macronesian bryophyte flora. J Hattori Bot Lab 56:7–13.
Shaw AJ, 1991. Ecological genetics, evolutionary constraints and the systematics of bryophytes. Adv Bryol 4:29–74.
Shaw AJ and Schneider RE, 1995. Genetic biogeography of the rare "copper moss," Mielichhoferia elongata (Bryaceae). Am J Bot 82:8–17.
Slatkin M and Barton N, 1989. A comparison of three indirect methods for estimating average levels of gene flow. Evolution 43:1349–1368.
Sneath PHA and Sokal RR, 1973. Numerical taxonomy. San Francisco: W.H. Freeman.
Soane ID and Watkinson AR, 1979. Clonal variation in populations of Ranunculus repens. New Phytol 82:557–573.
Stoneburner A, Wyatt R, and Odrzykosky IJ, 1991. Applications of enzyme electrophoresis to bryophyte systematics and population biology. Adv Bryol 4:1–27.
Steinger T, Korner C, Schmid B, 1996. Long-term persistence in a changing climate: DNA analysis suggests very old ages of clones of alpine Carex curvula. Oecologia 105:94–99.[Web of Science]
Stewart CNJ and Excoffier L, 1996. Assessing population genetic structure and variability with RAPD data: application to Vaccinium macrocarpon (American Cranberry). J Evol Biol 9:153–171.[Web of Science]
Szweykowski J, 1982. Genetic differentiation of liverwort populations and its significance for bryotaxonomy and biogeography. J Hattori Bot Lab 56:7–13.
Therrien JP, Crandall-Stotler B, and Stotler RE, 1998. Morphological and genetic variation in Porella platyphylla and P. platyphylloidea and their systematic implications. Bryologist 101:1–19.
Watkinson AR and Powell JC, 1993. Seedling recruitment and maintenance of clonal diversity in plant populations. Computer simulation of Ranunculus repens. J Ecol 81:707–717.
Waycott M, 1998. Genetic variation, its assessment and implications to the conservation of seagrasses. Mol Ecol 7:793–800.
Weir B, 1996. Genetic data analysis II. Sunderland, MA: Sinauer.
Widén B, Cronberg N, and Widén M, 1994. Genotypic diversity, molecular markers and special distribution in clonal plants, a literature survey. Folia Geobot Phytotax Praha 29:245–263.
Williams JGK, Kubelik AR, Livak KJ, Antoni RJ, and Tingey SV, 1990. DNA polymorphism amplified by arbitrary primers are useful as genetic markers. Nucleic Acids Res 15:6531–6535.
Wright S, 1951. The genetical structure of populations. Ann Eugen 15:323–354.
Wyatt R, 1985. Species concepts in bryophytes: input from population biology. Bryologist 88:82–189.
Wyatt R, Odrzykoski IJ, and Stoneburner A, 1989a. High levels of genetic variability in the haploid moss Plagiomnium ciliare. Evolution 43:1085–1096.
Wyatt R, Stoneburner A, and Odrzykoski IJ, 1989b. Bryophyte isozymes: systematic and evolutionary implications. In: Isozymes in plant biology (Soltis DE and Soltis PS, eds). Portland, OR: Dioscorides Press; 221–240.
Yamazaki T, 1982. Genic variability in natural populations of the haploid plant, Conocephalum conicum. In: Molecular evolution, protein polymorphism and the neutral theory (Kimura M, ed). Tokyo: Japan Scientific Societies Press; 123–134
Zehr DR, 1980. An assessment of variation in Scapania nemorosa and selected related species (Hepatophyta). Bryophyt Biblio 15:1–140.
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||