| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
2Botanical Garden and Museum, University of Oslo, Trondheimsveien 23 B, N-0562 Oslo, Norway; and 3The University Courses on Svalbard, P.O. Box 156, N-9170 Longyearbyen, Svalbard, Norway
Received for publication January 4, 2000. Accepted for publication July 5, 2000.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
Key Words: Arctic • Festuca • morphology • Poaceae • RAPDs • Svalbard • taxonomy
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
). Inclusion of characters that vary at random across taxa will, however, necessarily tend to conceal information in taxonomically significant characters. In this study of a polyploid complex in Festuca and in another study of polyploid Potentilla (Hansen, Elven, and Brochmann, 2000
), we explore the usefulness of an alternative approach involving an initial molecular analysis to identify groups of genotypes and a subsequent morphological analysis in which characters are selected and analyzed with the specific aim to test the hypothesis that the groups of genotypes are morphologically distinguishable.
The Festuca brachyphylla Schultes & Schultes fil. polyploid complex is today usually recognized as an arctic-alpine counterpart to the more temperate-montane F. ovina L. complex, which both belong to section Festuca (Conert, 1998
). Although most authors now separate clearly between the F. brachyphylla complex and the F. ovina complex (e.g., Tzvelev, 1976
; Aiken, Consaul, and Lefkovitch, 1995
), the acceptance of the F. brachyphylla complex as a separate entity and the delimitation of taxa within the complex have been controversial. For example, some American authors included the taxa of the F. brachyphylla complex as subspecies of F. ovina (e.g., Cronquist et al., 1977
; Scoggan, 1978
). Russian authors have either included them in a widely defined F. brachyphylla (e.g., Skvorcov, 1964
) or recognized several distinct species (Tzvelev, 1976
).
Members of the F. brachyphylla complex are characterized by short (0.3–1.4 mm) anthers with long filaments, whereas members of the F. ovina complex have longer (1.8–2.5 mm) anthers but shorter filaments (Scholander, 1934
; Frederiksen, 1977
). During flowering, the long anthers in plants of the F. ovina complex force the palea and lemma apart and keep the floret open, in contrast to plants of the F. brachyphylla complex, whose florets remain contracted, exposing only the short anthers on their long filaments. In addition, members of the F. brachyphylla complex are characterized by thin, broad lemmas with bowed margin, few-flowered spikelets, and three or more separate strands of leaf sclerenchyma, whereas members of the F. ovina complex have thicker and narrower lemmas with strongly involute margin, more flowers per spikelet, and continuous leaf sclerenchyma (Scholander, 1934
; Frederiksen, 1977, 1981
; Pavlick, 1984
).
According to the most recent treatment of the Festuca brachyphylla complex, four seminiferous (seed-bearing) species are probably widely distributed in the Arctic (Aiken, Consaul, and Lefkovitch, 1995
). These are F. brachyphylla, F. baffinensis Polunin, F. hyperborea Holmen ex Frederiksen, and F. edlundiae S. Aiken, Consaul & Lefkovitch. Some additional seminiferous taxa have been described, e.g., F. brevissima Jurtz. and F. minutiflora Rydb. (see, e.g., Tzvelev, 1976
), F. groenlandica (Schol.) Frederiksen (1977)
, and F. jensenii Gjærevoll & Ryvarden (1977)
. In addition, some viviparous plants originally referred to F. vivipara (L.) H. Sm. have been interpreted as viviparous derivatives of the F. brachyphylla complex, such as F. viviparoidea Krajina ex Pavlick (1984)
.
In the arctic Norwegian archipelago of Svalbard, all seminiferous plants of this complex were referred to F. brevifolia R. Br. or F. ovina until Scholander (1934)
recognized them as F. brachyphylla. Rønning (1961, 1996)
recognized three species in Svalbard (F. baffinensis, F. brachyphylla, and F. hyperborea). After the recent description of F. edlundiae from the Canadian Arctic (Aiken, Consaul, and Lefkovitch, 1995
), Elven and Elvebakk (1996)
tentatively accepted four seminiferous species of the F. brachyphylla complex in their checklist of the Svalbard flora (F. brachyphylla, F. baffinensis, F. hyperborea, and F. edlundiae). They emphasized that thorough analysis of the entire variation in Svalbard was necessary. Because the final conclusions in the present paper were in agreement with their treatment, those names are applied in the following.
In this study, we used RAPD analysis (random amplified polymorphic DNA; Williams et al., 1990
) and morphological analysis of field-collected populations to determine the number of seminiferous taxa in the Festuca brachyphylla complex in Svalbard, to find reliable morphological characters separating them, and to evaluate their morphological and genetic distinctness. In addition, we included the types of the four taxa tentatively accepted by Elven and Elvebakk (1996)
in the morphological analysis and revised the herbarium material deposited in Norwegian herbaria to map the geographic distribution of the taxa in Svalbard. The same populations were also analyzed for isozymes and variation in chromosome number; these results are published elsewhere (Guldahl, Borgen, and Nordal, in press). The level of intrapopulational variation was expected to be low in these taxa, which probably are strongly autogamous (Levkovsky, Tikhmenev, and Levkovsky, 1981
). In a pilot study carried out to assess the level of intrapopulational genetic variation, 20 plants from one population of each of the four tentative taxa were analyzed using enzyme electrophoresis (Guldahl, Borgen, and Nordal, in press) . Little or no variation was found within populations at isozyme loci, in agreement with the pattern observed in most other taxa analyzed in Svalbard (Brochmann and Steen, 1999
). Thus, we decided to analyze many small populations rather than a few large ones of each tentative taxon.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
|
|
From each plant, leaves were dried in fine-grained silica gel for use in RAPD analysis, and shoots with leaves and panicles were pressed for morphological analysis. The remainder of each plant was potted and cultivated in a phytotrone at the University of Oslo. Type material was borrowed of F. baffinensis (BM, lectotype: Canada, Baffin Island, Pond Inlet, 12 September 1934, N. Polunin 706), F. brachyphylla (BM, lectotype: Canada, Melville Island, 1820, Edwards s.n.), F. edlundiae (CAN, isotype: Canada, Bathurst Island, Polar Bear Pass, 75°43' N, 98°12' W, 11 August 1985, S. G. Aiken 3952), and F. hyperborea (C, holotype and BM, isotype: Greenland, Heilprin Land, Brønlund Fjord, at Kedelkrogselv, 82°10' N, 31° W, 28 July 1950, K. Holmen 8078).
For RAPD analysis, we used two plants from each population of the two supposedly most common ones of the tentative taxa (F. baffinensis and F. edlundiae) and three plants from each population of the two rarest taxa (F. brachyphylla and F. hyperborea); in addition, some more plants were included to represent additional multilocus phenotypes observed in isozyme analysis of the same populations (Guldahl, Borgen, and Nordal, in press). A total of 86 plants from 40 populations were thus analyzed for RAPD variation (Table 1).
With a few exceptions, the populations selected for morphological analysis were the same as those used in the RAPD analysis, and 1–6 plants (mean 3.6) were analyzed from each population (Table 1). Thirty-three of the field-collected populations (119 plants) as well as the type material were included in the morphological analysis.
RAPD analysis
One hundred milligrams fresh leaves from cultivated plants or 30 mg silica-dried leaves from field-collected plants were used for DNA extraction following Gabrielsen et al. (1997)
, except that a purifying step with RNAse (10 µg/mL DNA extract) was added. For quantification, 5 µL of each template stock was mixed with 1.5 µL loading buffer and run on a 0.7% agarose gel stained with ethidium bromide at 75–95 V for 1–2 h. The intensity of the template bands was compared with a lambda DNA marker cut with HindIII and EcoRI. DNA isolates were diluted to 0.2 ng/µL for use in PCR (polymerase chain reaction) following Gabrielsen et al. (1997)
. One nanogram template DNA was used in each PCR reaction, determined after an initial concentration test. Five microliters loading buffer was added to each PCR product, and 20 µL of this mixture was run on a 1.5% agarose gel stained with ethidium bromide.
To simplify selection of primers for this study, we first summarized the results of all previous primer tests in our laboratory (cf. Gabrielsen et al., 1997
; Gabrielsen and Brochmann, 1998
; Steen, 1998
; Tollefsrud et al., 1998
; Hansen, Elven, and Brochmann, 2000
). Thirty-nine of the total 143 primers included in these tests had worked well for several unrelated taxa and were therefore chosen for prescreening of two plants from each of the four tentative taxa of Festuca (A kit: 01, 02, 04, 10, 14, 17, 19; C kit: 01, 02, 05, 06, 08, 12 to 15, 18, 19; D kit: 02, 05, 07, 08, 11 to 13, 20; F kit: 03, 05, 07, 13; H kit: 02, 05 to 08, 11; K kit: 14; Operon Technologies, Alameda, California, USA). Ten of these primers (A01, A02, C01, C06, C12, C13, C14, C18, D05, and D20) produced distinct, polymorphic, and reproducible bands and were selected for full analysis of all plants. The gels were scored conservatively, i.e., only the most reliable bands were scored (as 1, present, or 0, absent). To check for reproducibility within PCR runs, DNA from one plant was included two times in each PCR run. The profiles of the plants used in the primer test were also compared with those of the same plants in the full RAPD analysis to check for reproducibility among PCR runs.
Morphological analysis
We initially examined variation in a number of morphological characters that potentially could discriminate among the groups of multilocus phenotypes identified in the RAPD analysis, including the characters analyzed by Aiken, Consaul, and Lefkovitch (1995)
. A total of 46 characters, 36 primary and 10 derived, were selected for full analysis (Table 2, Fig. 2). After a preliminary analysis, one measurement was considered sufficient per character per plant, except for basal leaf characters, for which three measurements were taken and the median used in further analyses, and except for the type specimens, for which five measurements (if possible) were taken per character and the median used in further analysis. Based on the results of the morphological analysis, all herbarium specimens of the Festuca brachyphylla complex from Svalbard deposited in O, TRH, and TROM were annotated and distributions were mapped.
|
|
). Analyses of the RAPD data were performed using four different similarity or distance coefficients (Dice, Simple matching, Jaccard, and Euclidean distance). These coefficients provided similar results, and we only show the analyses based on Dice similarity, given by 2a/(2a + b + c), where a is the number of shared bands and b and c are the number of bands present in one sample but absent in the other sample. Spearman's correlation coefficients between the morphological characters were computed using SPSS for Windows (Noru
is, 1993
). PCO analyses were also performed on 29 morphological characters (Table 2; excluding qualitative characters, highly correlated characters, highly variable characters, and the one of the two primary characters used in each derived character that was most correlated with the derived character; cf. Sneath and Sokal, 1973
). The morphological data matrix was standardized by ranging (i.e., the variation in each character was scaled between 0 and 1), and a distance matrix was computed using average Manhattan distance (city block), given by 1/n
k | xki - xkj |, where n is the number of plants, k is the character, and i and j are a pair of plants (Rohlf, 1994
). Spearman's correlation coefficients were also computed between the morphological characters and the PCO axes. |
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
In the PCO analysis of the entire RAPD data set, the first axis (51.6%) clearly separated one phenotype group (I; F. baffinensis) from the remaining phenotypes (Fig. 3). Another PCO analysis, performed to clarify the relationships among the remaining phenotypes, revealed three well-separated groups corresponding to F. brachyphylla, F. edlundiae, and F. hyperborea (Fig. 4). The first axis (60.0%) separated F. edlundiae (III) from F. brachyphylla (II) and F. hyperborea (IV), and the second axis (14.5%) separated all three taxa. Two of the five plants analyzed from the populations with putative hybrids had a RAPD phenotype identical to one observed in F. edlundiae, and three plants had phenotypes that were intermediate between those of F. edlundiae and F. hyperborea (Fig. 4).
|
|
|
|
In most morphological characters, the type specimens of three of the taxa (F. baffinensis, F. edlundiae, and F. hyperborea) corresponded well with one particular of three (I, III, and IV) of the four groups identified in the Svalbard material (Fig. 6; Table 3). Although the type specimen of F. baffinensis had shorter spikelets than the Svalbard material belonging to RAPD phenotype group I (Fig. 6), this Svalbard material undoubtedly belonged to F. baffinensis because of the high degree of similarity in other characters (e.g., dense, unilateral panicles, strongly hairy culms, and long flag leaf blades that were very long relative to width; Table 3). The type specimen of F. edlundiae was distinctly glaucous (when collected, according to Aiken, Consaul, and Levkovitch, 1995
) and had glabrous spikelet bases and relatively long lemmas, in agreement with the Svalbard material belonging to RAPD phenotype group III. The characters of the type specimens of F. hyperborea fell well within the variation observed in the Svalbard material belonging to RAPD group IV (Fig. 6; Table 3). The type specimen of F. brachyphylla was, however, most similar to plants belonging to RAPD group III (F. edlundiae) in some characters (e.g., character numbers 7, 16, and 24) but most similar to plants belonging to RAPD group II in other characters (e.g., character numbers 13 and 39 [see Fig. 6], 8, 15, and 32; Table 3), and the Svalbard material belonging to group II was therefore only tentatively referred to F. brachyphylla.
|
|
Several characters were strongly correlated with the first two axes in the PCO analyses of the morphological data (P < 0.01; Table 4). Lemma length was strongly correlated with the first axis in the PCO analysis of the entire data set (character 23, r = -0.80), as was panicle length (character 1) and spikelet length (character 23; r = -0.77 and -0.77, respectively). Culm color was strongly correlated with the second axis (character 34, r = 0.83). In the PCO analysis without F. baffinensis, upper glume length (character 18), lemma length (character 23) and panicle length (character 1) were strongly correlated with the first axis (r = -0.85, r = -0.77, and r = -0.76, respectively). Number of trichomes at the base of the spikelet (character 8) was strongly correlated with the second axis (r = 0.79).
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
), we have shown that a combination of detailed molecular and morphological analyses represents a powerful tool for unraveling complex low-level taxonomy. When interpreted in concert, the two types of data provide a possibility of avoiding some of the commonly encountered problems in traditional morphometric analysis, such as the definition of criteria for selection of characters and the unknown relationship between groups recognized based on morphology and genetic variation. In this study, we initially analyzed each individual plant for presence/absence of 30 polymorphic molecular markers that probably represent loci that are distributed randomly across the genome (cf. Williams et al., 1990, 1993
). Our RAPD analysis was made particularly efficient; we simplified the primer selection procedure by identifying primers that previously had worked well for several unrelated taxa (Gabrielsen et al., 1997
; Gabrielsen and Brochmann, 1998
; Steen, 1998
; Tollefsrud et al., 1998
; Hansen, Elven, and Brochmann, 2000
). Further analyses to determine whether the RAPD multilocus phenotypes could be classified into distinct groups were performed without any a priori assumptions of their taxonomic identity. In the subsequent morphological analysis, the characters were selected and analyzed with the specific aim to test the hypothesis that the four groups identified in the molecular analysis were morphologically distinguishable and should be recognized as distinct taxa.
The continuous nature of the variation revealed in our multivariate analyses of the morphological data in Potentilla (Hansen, Elven, and Brochmann, 2000
) as well as Festuca (Fig. 7) reflects a commonly encountered problem when attempting to interpret so-called "overall morphological variation" (cf. Sneath and Sokal, 1973
). In spite of this continuous, "overall" variation, we identified several individual morphological characters that could be used to classify each plant correctly to the particular group that corresponded to its RAPD phenotype. This result demonstrates that many of the quantitative morphological characters show complex variation across taxa and conceal the information provided by the taxonomically more significant characters. More or less random variation in quantitative characters among closely related species is probably a common pattern, because the phenotypic expression of such characters usually are determined by many genes, and different populations may maintain similar polymorphisms at many loci for a long time.
How many and which taxa in Svalbard?
The recognition of four distinct groups of RAPD multilocus phenotypes that unambiguously could be distinguished by several individual morphological characters (with exception of three putative hybrid plants; Figs. 5 and 6) indicate that there are four seminiferous taxa in the Festuca brachyphylla complex in Svalbard. Although they probably are closely related, each of them has unique features, morphologically as well as genetically, justifying their recognition as separate taxa. Our morphological analysis of the type specimens of the four taxa (Fig. 6; Table 3) that were suggested to be present in Svalbard by Elven and Elvebakk (1996)
supports their hypothesis; the four groups identified in Svalbard can be referred to three species described from Canada (F. baffinensis, F. brachyphylla, and F. edlundiae) and one species described from Greenland (F. hyperborea; with a reservation concerning the name F. brachyphylla, see below). The results of this study are in agreement with the results of the isozyme and chromosome number analyses of the same populations in Svalbard (Guldahl, Borgen, and Nordal, in press). The four taxa are distinct also at enzyme electrophoretic loci, and their chromosome numbers in Svalbard (Guldahl, Borgen, and Nordal, in press) are the same as those in other arctic areas (Aiken, Consaul, and Lefkovitch, 1995
). Three species are tetraploid with 2n = 28 (F. baffinensis, F. edlundiae, and F. hyperborea) and one is hexaploid with 2n = 42 (F. brachyphylla).
Most of the distinguishing characters we observed in the Svalbard material are in agreement with the analysis of Canadian material by Aiken, Consaul, and Lefkovitch (1995)
. Festuca baffinensis is most distinct, genetically as well as morphologically. It is easily recognized by its dark purple, unilateral panicle and densely hairy culm (Figs. 5 and 6). In the Svalbard material, Festuca hyperborea is easily distinguished from F. brachyphylla by its shorter lemmas and glumes, and the basal leaves of F. hyperborea are as broad as those of F. brachyphylla, but much shorter; leaf blade length relative to width is therefore a good separating character. In the Svalbard material, the lack of trichomes at the base of the spikelet and the glaucous color of the culms and leaves in F. edlundiae separate it well from F. brachyphylla and from F. hyperborea. The type specimen of F. brachyphylla is most similar to the Svalbard populations of F. edlundiae in some characters, but it is most similar to the Svalbard populations of F. brachyphylla in several characters that we consider taxonomically more important (e.g., the erect growth form and the presence of trichomes at the base of the spikelet; Table 3). Our recognition of the Svalbard material belonging to RAPD group II as F. brachyphylla is, nevertheless, tentative. What is currently recognized as F. brachyphylla (e.g., by Aiken, Consaul, and Lefkovitch, 1995
) is a polymorphic species with a large, nearly circumpolar distribution. The Svalbard material of F. brachyphylla represents only a small and very homogeneous part of this variation. Recently collected material from Melville Island, where the type locality of F. brachyphylla is situated, corresponds well with the type and falls into the variation pattern of American F. brachyphylla (R. Elven, unpublished data).
Festuca baffinensis definitely deserves the rank of species due to its distinctness in all data sets analyzed. These data, as well as the unique presence of apical hairs on the ovaries in this species (Aiken, Consaul, and Lefkovitch, 1995
), may suggest that F. baffinensis should be kept separate from the F. brachyphylla complex. Festuca brachyphylla, F. edlundiae, and F. hyperborea are clearly more closely related to each other than each of them is to F. baffinensis, and it is probable that these polyploids have at least one diploid genome in common. A rank of subspecies may be suggested based on the PCO analyses of the morphological data (Fig. 7), in which these three taxa are not clearly separated. However, it is probably more appropriate to recognize also these three taxa at the species level, because each taxon differs from the other two in at least two morphological characters that obviously are genetically independent (Figs. 5 and 6). Recognition of F. brachyphylla, F. edlundiae, and F. hyperborea as species is also supported by the PCO and UPGMA analyses of the RAPD data, which show three clearly distinct groups, with the exception of the three putative hybrids (Figs. 4 and 5). In addition, each of the taxa had at least one RAPD marker that was found in all plants of that taxon but not in any of the other taxa. The recognition of F. brachyphylla at the species level is also supported by its different ploidy level. The rank of subspecies seems also inappropriate because this category (at least in Europe) most often is applied for major geographic races that only are partly isolated reproductively. The taxa do not represent geographic races; their ranges are very similar (F. brachyphylla nearly circumpolar, F. baffinensis interruptedly circumpolar, F. hyperborea interruptedly High-Arctic, and F. edlundiae at least from North East Asia through North America and Greenland to Svalbard; R. Elven, unpublished data).
Although the species sometimes grow in close spatial proximity in Svalbard, they rarely occur in mixed populations and have somewhat differing habitat preferences (Guldahl, Borgen, and Nordal, in press). The taxa are also reproductively isolated from each other because of predominant autogamy (Levkovsky, Tikhmenev, and Levkovsky, 1981
). We only observed a few putative hybrids in the two populations (numbers 7 and 18) that occurred on disturbed ground near Longyearbyen Airport. Three plants (7–1, 18–1, and 18–4) were intermediate between the accepted taxa in the analyses of the morphological data as well as the RAPD data, but it is not immediately clear which of the potential parental taxa (F. brachyphylla, F. edlundiae, and F. hyperborea) have been involved. These putative hybrids were closer to F. hyperborea than to F. brachyphylla and F. edlundiae in the multivariate analyses of the RAPD data (Figs. 4 and 5), despite F. hyperborea being absent from the area where the hybrids were found (Fig. 1).
Extremely low levels of molecular variation (one or two RAPD multilocus phenotypes) were observed in the Svalbard populations of F. baffinensis, F. brachyphylla, and F. hyperborea, whereas F. edlundiae was relatively variable (12 RAPD multilocus phenotypes; Fig. 5). This arctic archipelago was heavily glaciated at the Weichselian maximum (Landvik et al., 1998
), and most, if not all, of the species in the present flora must have immigrated postglacially (Brochmann et al., 1996
; Brochmann and Steen, 1999
). Some species have probably immigrated a single time via long-distance chance dispersal and thus show low levels of genetic variation in Svalbard, whereas other species, such as Saxifraga cernua (Brochmann et al., 1998
; Gabrielsen and Brochmann, 1998
), S. cespitosa (Tollefsrud et al., 1998
), S. oppositifolia (Gabrielsen et al., 1997
), and several Draba spp. (Brochmann, Soltis, and Soltis, 1992
), are highly variable in Svalbard and probably have immigrated several times. It is possible that Festuca edlundiae belongs to the latter group. Festuca brachyphylla is very variable genetically and morphologically in other arctic areas (Aiken et al., 1994
; Aiken, Consaul, and Lefkovitch, 1995
), suggesting that this species may have been heavily bottlenecked when arriving in Svalbard. Festuca baffinensis and F. hyperborea show, however, low levels of variation also in other arctic areas (Aiken et al., 1994
; Aiken, Consaul, and Lefkovitch, 1995
).
|
|
FOOTNOTES |
|---|
4 Author for correspondence (christian.brochmann{at}toyen.uio.no
).
|
LITERATURE CITED |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
———, ———, A. Spidle, and B. May. 1994 Allozyme and morphological observations on Festuca hyperborea, compared with F. baffinensis and F. brachyphylla (Poaceae) from the Canadian Arctic. Nordic Journal of Botany 14: 137–143[ISI]
Brochmann, C., T. M. Gabrielsen, A. Hagen, and M. M. Tollefsrud. 1996 Seed dispersal and molecular phylogeography: glacial survival, tabula rasa, or does it really matter? Det Norske Videnskaps-Akademi. I. Matematisk Naturvitenskapelig Klasse, Skrifter, Ny Serie 18: 54–68
———, D. E. Soltis, and P. S. Soltis. 1992 Electrophoretic relationships and phylogeny of Nordic polyploids in Draba (Brassicaceae). Plant Systematics and Evolution 182: 35–70[CrossRef][ISI]
———, and S. W. Steen. 1999 Sex and genes in the flora of Svalbard—implications for conservation biology and climate change. Det Norske Videnskaps-Akademi. I. Matematisk Naturvitenskapelig Klasse, Skrifter, Ny Serie 38: 33–72
———, Q.-Y. Xiang, S. J. Brunsfeld, D. E. Soltis, and P. S. Soltis. 1998 Molecular evidence for polyploid origins in Saxifraga (Saxifragaceae): the narrow arctic endemic S. svalbardensis and its widespread allies. American Journal of Botany 85: 135–143[Abstract]
Conert, H. J. 1998 Poaceae (Echte Gräser oder Süssgräser). In G. Hegi [ed.], Illustrierte Flora von Mitteleuropa. 3. Aufl. Band I, Teil 3. Parey Buchverlag, Berlin, Germany (in German)
Cronquist, A., A. H. Holmgren, N. H. Holmgren, J. L. Reveal, and P. K. Holmgren [eds.]. 1977 Intermountain flora. Columbia University Press, New York, New York, USA
Elven, R., and A. Elvebakk. 1996 Part 1: Vascular plants. In A. Elvebakk and P. Prestrud [eds.], A catalogue of Svalbard plants, fungi, algae and cyanobacteria. Norsk Polarinstitutt Skrifter 198: 9–55. Norsk Polarinstitutt, Oslo, Norway
Frederiksen, S. 1977 The Festuca brachyphylla group in Greenland. Botaniska Notiser 130: 269–277[ISI]
———. 1981 Festuca vivipara (Poaceae) in the North Atlantic area. Nordic Journal of Botany 1: 277–292
Gabrielsen, T. M., K. Bachmann, K. S. Jakobsen, and C. Brochmann. 1997 Glacial survival does not matter: RAPD phylogeography of Nordic Saxifraga oppositifolia. Molecular Ecology 6: 831–842
———, and C. Brochmann. 1998 Sex after all: high levels of diversity detected in the arctic clonal plant Saxifraga cernua using RAPD markers. Molecular Ecology 7: 1701–1708[CrossRef]
GjÆrevoll, O., and L. Ryvarden. 1977 Botanical investigations on J. A. D. Jensens Nunatakker in Greenland. Det Kongelige Norske Videnskabers Selskab, Skrifter 4: 1–40
Guldahl, A. S., L. Borgen, and I. Nordal. In press. Variation in the Festuca brachyphylla (Poaceae) complex in Svalbard elucidated by chromosome numbers and isozymes. Botanical Journal of the Linnean Society
Hansen, K. T., R. Elven, and C. Brochmann. 2000 Molecules and morphology in concert: tests of some hypotheses in arctic Potentilla (Rosaceae). American Journal of Botany 87: 1466–1479
Landvik, J. Y., S. Bondevik, A. Elverhøy, W. Fjeldskaar, J. Mangerud, M. J. Siegert, O. Salvigsen, J. I. Svendsen, and T. O. Vorren. 1998 The last glacial maximum of Svalbard and the Barents sea area: ice sheet extent and configuration. Quaternary Science Reviews 17: 43–75[CrossRef][ISI]
Levkovsky, V. P., E. A. Tikhmenev, and E. P. Levkovsky. 1981 Cleistogamy in arctic grasses. Botanicheskii Zhurnal SSSR 66: 116–120 (in Russian)
Noruis, M. J. 1993 SPSS for Windows, release 6.0. SPSS, Chicago, Illinois, USA
Pavlick, L. E. 1984 Studies on the Festuca ovina complex in the Canadian Cordillera. Canadian Journal of Botany 62: 2448–2462
Rohlf, F. J. 1994 NTSYS-pc. Numerical taxonomy and multivariate analysis system, version 1.80. Exeter Software, New York, New York, USA
Rønning, O. I. 1961 Some new contributions to the flora of Svalbard. Norsk Polarinstitutt Skrifter 124: 1–20
———. 1996 The flora of Svalbard. Norsk Polarinstitutt, Oslo, Norway
Scholander, P. F. 1934 Vascular plants from northern Svalbard. Skrifter om Svalbard og Ishavet 62: 1–152
Scoggan, H. J. 1978 The Flora of Canada. 2. Pteridophyta, Gymnospermae, Monocotyledoneae. National Museums of Canada, Ottawa, Ontario, Canada
Skvorcov, A. K. 1964 Festuca L. In A. I. Tolmachev [ed.], Arkticheskaya Flora SSSR 2. Nauka Publishers, Moscow-Leningrad, Russia (in Russian)
Sneath, P. H. A., and R. R. Sokal. 1963 Principles of numerical taxonomy. Freeman, San Francisco, California, USA
———, and ———. 1973 Numerical taxonomy: the principles and practice of numerical classification. Freeman, San Francisco, California, USA
Steen, S. W. 1998 Population structure and RAPD phylogeography of Papaver dahlianum in Svalbard. Master's thesis, University of Oslo, Oslo, Norway
Tollefsrud, M. M., K. Bachmann, K. S. Jakobsen, and C. Brochmann. 1998 Glacial survival does not matter—II: RAPD phylogeography of Nordic Saxifraga cespitosa. Molecular Ecology 7: 1217–1232
Tzvelev, N. N. 1976 Zlaki SSSR. Nauka Publishers, Leningrad, Russia (in Russian)
Williams, J. K. G., M. K. Hanafey, J. A. Rafalski, and S. V. Tingey. 1993 Genetic analysis using random amplified polymorphic DNA markers. Methods in Enzymology 218: 704–740[ISI][Medline]
———, A. R. Kubelik, K. J. Livak, J. A. Rafalski, and S. V. Tingey. 1990 DNA polymorphisms amplified by arbitrary primers are useful as genetic markers. Nucleic Acids Research 18: 6531–6535
This article has been cited by other articles:
|
|
 |
|
  M.-S. Tiebre, J.-P. Bizoux, O. J. Hardy, J. P. Bailey, and G. Mahy Hybridization and morphogenetic variation in the invasive alien Fallopia (Polygonaceae) complex in Belgium Am. J. Botany, November 1, 2007; 94(11): 1900 - 1910. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. L. Jorgensen, I. Stehlik, C. Brochmann, and E. Conti Implications of ITS sequences and RAPD markers for the taxonomy and biogeography of the Oxytropis campestris and O. arctica (Fabaceae) complexes in Alaska Am. J. Botany, October 1, 2003; 90(10): 1470 - 1480. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. T. Tercek, D. P. Hauber, and S. P. Darwin Genetic and historical relationships among geothermally adapted Agrostis (bentgrass) of North America and Kamchatka: evidence for a previously unrecognized, thermally adapted taxon Am. J. Botany, September 1, 2003; 90(9): 1306 - 1312. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
