| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Systematics |
2L. H. Bailey Hortorium, Cornell University, Ithaca, New York 14853-5908 USA; 3Pringle Herbarium, Department of Botany and Agricultural Biochemistry, The University of Vermont, Burlington, Vermont 05405-0086 USA
Received for publication March 29, 2002. Accepted for publication September 26, 2002.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
Key Words: biogeography • Dryopteridaceae • phylogeny • Polystichum • rbcL
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
, 1995
; Wolf et al., 1994
; Pryer et al., 1995
). However, most of this research has focused on small genera and relationships above the genus level. Three barriers have prevented the exploration of large fern genera using these innovations. First, all large fern genera have a worldwide distribution, which has led to unbalanced representation of the species from different continents. Second, large genera are often not monophyletic, having been rendered paraphyletic by the recognition of distinctive small groups of species as segregate genera. Third, the choice of outgroup for rooting is often unclear. Nevertheless, studies of reproductive biology and speciation in the ferns have often concentrated on members of large genera, as they are likely to present localized groups of species with challenging problems. Consequently, innovative phylogenetic work on large genera is critical if we are to provide a useful context for studies of speciation and reproductive biology in these groups.
Polystichum Roth is one of the 10 largest genera of ferns, having been variously estimated to comprise between 200 and 300 species (Barrington, 1995
; Mabberley, 1997
). Like most large fern genera, Polystichum is nearly cosmopolitan in distribution. The center of diversity is in southwest China and adjacent regions, with a secondary centers of diversity in tropical America and Malesia. Polystichum is found in montane regions throughout its range, with a preference for disturbed situations such as talus slopes, stream banks, and road cuts. In tropical regions few species grow below 1000 m, so that species distributions are often highly discontinuous (Barrington, 1992
, 1993
).
Polystichum has never been subjected to a consistent, exhaustive systematic analysis and revision. As a result researchers have been hampered by unclear generic circumscription and incomplete infrageneric classification schemes. Large numbers of species have sometimes been included as varieties of single species; e.g., most twice-pinnate species were once classified as varieties of P. aculeatum (L.) Roth (see Christensen, 1906
). Morphological and cytological evidence suggests that Polystichum is paraphyletic; the segregate genera Acropelta Nakai, Cyrtomidictyum Ching, Cyrtomium C. Presl, Cyrtonogonellum Ching, Papuapteris C. Chr., Phanerophlebia C. Presl, Phanerophlebiopsis Ching, Plecosorus Fée, Ptilopteris Hance, and Sorolepidium Christ are among those suggested to belong within a monophyletic Polystichum (e.g., Tryon and Tryon, 1981
; Yatskievych et al., 1988
).
A few specialists have addressed the circumscription of sections within Polystichum (e.g., Tagawa, 1940
; Daigobo, 1972
; Fraser-Jenkins, 1997
), and additional work has been done on the delineation of specific sections (e.g., Kung, 1989
; Xiang, 1994
; Zhang, 1994
; Zhang and Kung, 1996a
, b
; Roux, 2000
). A set of 8–16 morphological characters is commonly employed to circumscribe sections. These characters include lamina and petiole scale form and color; lamina dissection, design, and texture; presence/absence of proliferous bulbils; leaf venation; and characteristics of the indusium. Each of the proposed sectional classifications is geographically restricted in scope; as a result there is no global infrageneric classification of the genus. The two most ambitious efforts (Daigobo, 1972
; Fraser-Jenkins, 1997
) do however afford a basis for further work by providing a guide to sampling diversity within the genus and by providing a synopsis of variable morphological characters (for further details of these subgeneric classifications see Appendix 1, archived at the Botanical Society of America website [http://ajbsupp.botany.org/v90/]).
Hybridization, allopolyploidy, and apogamy all confound attempts at systematic revision and phylogenetic analysis of the genus Polystichum. More than 82 different hybrid combinations between Polystichum species have been recognized (Knobloch, 1996
). As expected, most of these hybrids are totally sterile (Manton, 1950
; Nakaike, 1973
; Wagner, 1973
, 1979
; Daigobo, 1974
; Barrington, 1985
, 1990
; Barrington et al., 1989
). However, swarms of fertile hybrids between Polystichum imbricans and P. munitum (for species authorities hereafter see Appendix 2, archived at the Botanical Society of America website [http://ajbsupp.botany.org/v90/]) have been reported where the two species come into close contact in habitats that are transitional between mesic and xeric (Mayer and Mesler, 1993
; Mullenniex et al., 1998
). Of the 81 species of Polystichum that have been investigated cytologically, 36 (44%) are polyploid (data largely from Löve et al., 1977
); a number of these are documented to be derived from interspecific hybrids (e.g., Manton, 1950
; Wagner, 1973
). Apogamous reproduction has also been reported within Polystichum (Tryon and Tryon, 1982
). Consequently, any study of phylogeny in the genus and its allies must take into consideration the cytology and reproductive biology of the sampled taxa.
The exact relationship of Polystichum to other genera within the Dryopteridaceae remains unclear. The results of an intuitive phylogenetic analysis using morphological data led Pichi Sermolli (1977
; see our Fig. 1) to place Polystichum near Dryopteris, Arachniodes, and several other genera, based on similar rachis structure. He postulated more distant relationships with other genera in his Aspidiaceae (= Dryopteridaceae sensu stricto [s.s.]), including Lastreopsis, Ctenitis, and Rumohra (Fig. 1). Pichi Sermolli's Aspidiaceae were in turn hypothesized to share a recent common ancestor with Elaphoglossaceae plus Lomariopsidaceae. Consistent with Pichi Sermolli's phylogeny is the recent analysis of rbcL sequences from 99 genera of leptosporangiate ferns by Hasebe et al. (1995)
. In this work, Arachniodes Blume, Ctenitis C. Chr., Dryopteris Adans., and Polystichum consistently appear to be each other's nearest allies, with Elaphoglossum J. Smith and Rumohra Raddi more remote. The Arachniodes and Dryopteris species sampled are consistently sister taxa in the analyses. Polystichum and Ctenitis are sister taxa in the maximum-likelihood and neighbor-joining analyses, but their relationship is unresolved in the parsimony analysis (Hasebe et al., 1995
).
|
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
with the cladistic groupings of Hasebe et al. (1995)
prompted us to select representatives from Pichi Sermolli's Aspidiineae as outgroup taxa. Within Polystichum, we sampled taxa representing a diversity of form and geographic provenance to increase the likelihood of representing different evolutionary lineages. We included representatives of as many of the published sections as possible (see survey of sections, Appendix 1 and Table 3 [http://ajbsupp.botany.org/v90/]). Five taxa belonging to segregate genera were included to assess the classification of taxa with extreme morphological departures from the standard Polystichum type. Voucher information for the taxa included in the present study is archived in Appendix 2 (for the rbcL data) and 3 (for the morphological data) at the Botanical Society of America website (http://ajbsupp.botany.org/v90/).
Molecular methods: genomic DNA isolation
Genomic DNA was isolated from 42 species with a modified version of the Doyle and Doyle (1987)
CTAB protocol. Approximately 0.1 g of leaf tissue—snap frozen, fresh, air-dried, or silica-dried—was ground in 600 µL of extraction buffer (200 mmol/L Tris pH 8.0; 20 mmol/L EDTA, 1.4 mol/L sodium chloride, 2% 40 kDa polyvinylpyrrolidone [omitted in some cases to increase yield], 2% hexadecyltrimethylammonium bromide [CTAB], 5 mmol/L l-ascorbic acid, 4 mmol/L diethyldithiocarbamic acid, 1% sodium metabisulfite, 0.5% sodium dodecyl sulfate, and 5% ß-mercaptoethanol) preheated to 60°C. Homogenized tissue was incubated at 60°C, shaking at 4.2 rad/s, for 10 min. Samples were extracted with one volume of chloroform : isoamyl alcohol (24 : 1). After separation of the organic and aqueous phases by centrifugation (156 800 m/s2 for 5 min), nucleic acids were precipitated from the aqueous phase with one volume of ice-cold isopropanol. The nucleic acids were pelleted by centrifugation (135 200 m/s2 for 2 min) and resuspended in 100 µL TE (10 mmol/L Tris, 1 mmol/L EDTA, pH 8.0). Following the addition of 20 µg of ribonuclease A, samples were incubated at 37°C for 10 min. DNA was precipitated at 4°C with two volumes of ethanol and half a volume of 7.5 mol/L ammonium acetate. Precipitated DNA was pelleted by centrifugation (135 200 m/s2 for 5 min) and dried, under a vacuum, at room temperature. The DNA was then dissolved in 100 µL of 10 mmol/L tris (pH 8.0).
Molecular methods: PCR
The polymerase chain reaction (PCR) was used to amplify a 1379 base-pair fragment corresponding to bases 1 through 1379 of the chloroplast gene that encodes the large subunit of ribulose-1,5-bisphosphate carboxylase/oxygenase (rbcL). The PCR amplifications took place in a 100-µL reaction mixture containing: 1x PCR Buffer II (Perkin-Elmer, Foster City, California, USA), 0.2 mmol/L of each dNTP, 0.7 µmol/mL of each amplification primer (for primer sequences see Table 1), 2.5 mmol/L MgCl2, 1 unit of AmpliTaq (Perkin-Elmer), and approximately 500 ng of genomic DNA. Reactions were incubated at 95°C for 3 min in an Amplitron II (0.2-mL well size; Thermolyne Dubuque, Iowa, USA) then cycled 35 times (95°C for 30 s, 55°C for 30 s, 72°C for 60 s), and finally incubated at 72°C for 10 min. Successful PCR reactions were purified with the QIAquick PCR Purification Kit following the manufacturer's protocol (Qiagen, Valencia, California, USA).
|
Morphological methods
Herbarium specimens, often the vouchers for the molecular work, were scored for a total of 41 qualitative characters (described in Appendix 4, archived at http://ajbsupp.botany.org/v90/). The character set is derived from previous analyses of the genus Polystichum and its allies, especially those of Daigobo (1972)
, Pichi Sermolli (1977)
, Tryon and Tryon (1982)
, Barrington (1995)
, and Fraser-Jenkins (1997)
. The set of taxa comprises all the species from the molecular analysis not known to be polyploid.
Data analysis
We first analyzed all 55 taxa (14 of which are from GenBank) using the molecular data. Complete rbcL sequences were checked for accuracy and aligned by inspection. No insertions or deletions were required to align the sequences. Uncertainties in nucleotide sequence were coded as polymorphisms using standard IUPAC nomenclature. If more than one non-identical rbcL sequence was available for a given taxon a separate analysis was conducted (as outlined below; data not shown) to ensure that the sequences were sister to each other. Since this was true in all cases, the sequences were fused using the fuse-taxa option of Winclada (Nixon, 2001
), to create a single terminal for each taxon (i.e., differences between the sequences were treated as polymorphisms). We then performed a combined molecular and morphological analysis of taxa.
Winclada was used to remove uninformative characters from the matrix (mop command). The resulting rbcL matrix had 227 parsimony-informative characters and 55 taxa (3.35% of cells had missing or polymorphic data). The resulting combined matrix had 240 parsimony-informative characters (two additive) and 38 taxa (7.81% of cells had missing or polymorphic data). Gaps at the beginning and the end of the sequences were treated as missing data. NONA (version 2.0; Goloboff, 1993
) was used to search for shortest trees: we used 1000 random addition sequences with 10 trees held during tree bisection-reconnection (TBR) swapping of each addition sequence (h/10;mult*1000;). The resulting trees were swapped to completion (max*;) holding a maximum of 10 000 trees.
Analysis of the rbcL matrix yielded 411 trees of length 602 steps; excluding uninformative characters, consistency index (CI) = 0.44 and retention index (RI) = 0.66. The combined morphological and molecular matrix yielded 33 trees of length 689 steps; excluding uninformative characters, CI = 0.40 and RI = 0.57. The strict consensus tree (with ambiguously supported nodes collapsed) was calculated by Winclada. Rumohra adiantiformis was used to root the consensus tree, because in the phylogeny of Hasebe et al. (1995)
it is most basal among the taxa we sampled.
Winclada was used to generate jackknife procedure files for NONA. The matrix was sampled without replacement 1000 times; the probability of character selection was approximately 0.63. Each jackknife replicate was searched using the following strategy: five random addition sequences; 10 trees held during TBR swapping of each addition sequence; and the strict consensus of each replicate was calculated (h/10;mult*5; inter;). The frequency of occurrence of each group present in the strict consensus tree was calculated with Winclada.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
The Polystichum species sampled fall into two monophyletic groups (Fig. 2): (1) an earlier diverging clade of three species that we will call the Cyrtomidictyum clade because it contains C. lepidocaulon (the type species of Cyrtomidictyum, a genus often included within Polystichum) and (2) a more recent clade including the remaining 31 species, which we will call Polystichum sensu stricto (s.s.). The two are separated by a clade comprising the three Cyrtomium species in the study set, rendering Polystichum paraphyletic unless Cyrtomium is included in Polystichum.
|
; Mullenniex et al., 1998
). Considering the more basal part of the tree, the Polystichum species sampled (including Cyrtomidictyum) plus Cyrtomium (which we will refer to as Polystichum sensu lato [s.l.]) together constitute a clade whose sister group, the Phanerophlebia clade, includes all three samples of Phanerophlebia (an entirely Neotropical genus) plus Polystichopsis chaerophylloides, which is from Puerto Rico in the Greater Antilles. The two Asian Ctenitis species included are sister to this clade, and Dryopteris and Arachniodes, both monophyletic, are in turn sister to the two Asian Ctenitis species plus all the close allies of Polystichum. Finally, all of the aforementioned taxa constitute a monophyletic group sister to a clade including Elaphoglossum and the Neotropical Lastreopsis effusa. The Costa Rican Megalastrum atrogriseum is sister to all of these taxa.
The analysis of combined rbcL and morphological characters (strict consensus, Fig. 3; for morphological data see Appendix 5, archived at http://ajbsupp.botany.org/v90/) yielded results largely similar to the rbcL analysis alone. The relationships of the outgroups to each other and to the sampled polystichums remained unchanged with two exceptions. First, Polystichopsis chaerophylloides did not group with the phanerophlebias, but was sister to a clade including the phanerophlebias and Polystichum s.l. Second, Ctenitis, the Dryopteris-Arachniodes clade, and the clade including Polystichum s.l. plus the phanerophlebias and Polystichopsis formed an unresolved trichotomy.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
parsimony analysis of 107 rbcL sequences from pteridophytes. Our much denser sampling of taxa closely related to Polystichum has resolved sister relationships between Polystichum s.l. and allied genera. The genus Phanerophlebia is prominent in reaching this goal, in that the three taxa we sampled lie in a clade sister to Polystichum s.l. (Figs. 2 and 3). A fourth sequence, from the West Indian Polystichopsis chaerophylloides, is included basally in the Phanerophlebia clade in the analysis using rbcL alone, although with less than 50% jackknife support. (We have chosen to follow Morton's [1960]
circumscription in recognizing the genus Polystichopsis [J. Smith] Holttum as distinct from Arachniodes because it is not part of the monophyletic group comprising Arachniodes aristata and A. denticulat in our analyses.) Together, these taxa constitute our best current estimate of the sister group to the genus Polystichum s.l.
Outside the clade including Phanerophlebia and Polystichum s.l., the Old World Ctenitis species we sampled are the next most closely allied when rbcL is analyzed alone. The Arachniodes-Dryopteris clade is sister to all of these taxa, with the Elaphoglossum-Lastreopsis clade still more remote. The results of our analyses depart from the hypothesis of Pichi Sermolli (1977
; our Fig. 1) in that Polystichum and Phanerophlebia are more closely related to Ctenitis than to Dryopteris and Arachniodes. In addition the placement of Polystichopsis is different. In Pichi Sermolli (1977
; our Fig. 1) it is allied with Dryopteris and Arachnoides, while our analyses place it with Phanerophlebia and Polystichum s.l.
Using chloroplast DNA restriction fragment length variation Yatskievych et al. (1988)
addressed the differences between Polystichum (five species, all North American), Cyrtomium (four species, all Asian), and Phanerophlebia (seven species, all American). Comparison of their work to ours is difficult, as their analyses yielded parsimony networks, not rooted phylogenies. If their network is rooted on the basis of the results of our phylogenetic analysis as represented in Fig. 2, some comparisons can be made. The two analyses agree in two prominent respects: (1) relationships among the Phanerophlebia species are the same; and (2) Cyrtomium is more closely related to Polystichum than to Phanerophlebia.
On the other hand, there are several distinctive differences between the results of Yatskievych et al. (1988)
and our own. The three genera as they sampled them are monophyletic with our rooting, whereas in our analysis Polystichum is not. This discrepancy is due to our inclusion of P. deltodon, P. tripteron, and Cyrtomidictyum lepidocaulon, which were not sampled by them. Secondly, although our study did not produce much resolution within Polystichum s.s., the species that were sampled in both studies are resolved differently. The apparent contradiction between the present study and that of Yatskievych et al. (1988)
is unexpected, given the argument of Doyle (1992)
that the chloroplast behaves as a single multistate character. Following from Doyle, cpDNA restriction data should produce a tree congruent with one produced from the rbcL sequences, which are chloroplastic. The only well-supported differences that can be be expected relate to the level of resolution of the two trees.
A monophyletic genus Polystichum
Our second goal was to decide how to divide Polystichum and its allies into a set of monophyletic genera given the results of our analysis. The options, judging from the current sample, are to (1) recognize a very large genus Polystichum including Cyrtomium, Phanerophlebia, Polystichopsis, and Cyrtomidictyum; (2) recognize Polystichopsis, Phanerophlebia, and Cyrtomium, and segregate P. lepidocaulon and its allies P. tripteron and P. deltodon as Cyrtomidictyum, leaving a more narrowly circumscribed Polystichum, which includes only our Polystichum s.s. clade; or (3) circumscribe Polystichum in the broad sense including Cyrtomium in Polystichum s.l. and recognize only Phanerophlebia among Polystichum segregates. We have chosen the second option in order to retain generic recognition for the horticulturally popular Cyrtomium. Consequently, all of the segregate genera we investigated except two should be recognized as distinct monophyletic genera. The exceptions are Plecosorus, the single species of which (Pl. speciosissimus [A. Braun ex Kze.] Moore) is almost certainly a Polystichum, and Acropelta (represented by Arachniodes aristata in our sample), which seems best left as a synonym of Arachniodes.
Biogeogaphic trends in Polystichum
Our third goal was to investigate the phylogeny of the genus Polystichum. Our results suggest that, at least in the tropics, recent evolution has been confined to single continents. The monophyly of the tropical American species provides the basis for a more intensive investigation of the genus in this region, which is at the center of our interest (see, for example, Barrington [1990]
). The close integration of the African species, whose morphological diversity and predominantly high-polyploid chromosome numbers suggest a long evolutionary history, leads us to conclude that the continent harbors the remnants of a complex clade characterized by extensive hybridization and extinction. In contrast, the phylogenetically remote positions of the morphologically and ecologically similar species found in Asia (P. craspedosorum) and the West Indies (P. ekmanii and P. underwoodii) suggest that they are convergent lineages, not members of a single lineage with a West Indian-East Asian disjunct distribution.
Morphological transformations
Optimization of morphological transformations onto the combined phylogeny provides some insights into structural evolution in Polystichum and its allies. The auriculate ultimate segment typical of Polystichum (our character 10; CI = 0.25) appears to be a synapomorphy for Polystichum and Cyrtomidictyum. This character is evidently a morphological synapomorphy for the genus in the broad sense, as Cyrtomium (excluded from the morphological analysis as it is apogamous and not diploid) also has auriculate ultimate segments. Similarly, peltate indusia (our character 21; CI = 0.50–1.00), a traditional diagnostic feature of Polystichum, may have been derived from reniform indusia in the common ancestor of these three genera, but other histories are as likely. A series of three characters is useful in combination for circumscribing the clade including Polystichum s.l. plus Phanerophlebia even though each has a low CI. These comprise ciliate petiole-base scales (our character 26; CI = 0.14), catadromous venation (our character 9; CI = 0.16, also in Polystichopsis), and sporangial receptacle terminal on the vein (our character 19; CI = 0.12). As most of the extra steps contributing to the low CIs are reversals, not parallelisms, it seems likely that these characters define ancestral character states of this clade. Our optimization allows us to characterize the common ancestor of all of these genera (the lineages immediately relevant to the history of Polystichum). Presumably this ancestor would have had ciliate petiole-base scales, once-pinnate fronds, ultimate segments with scarious tips, peltate indusia, and microscales.
Apogamy in Polystichum and Cyrtomium
We found that all of the apogamous polystichums sampled belong to the same clade. As the accessions of the apogamous genus Cyrtomium lie together in a second clade remote from the apogamous polystichums, it appears that apogamy has originated just twice in the study group as sampled. Two of our apogamous Polystichum taxa, the African P. luctuosum and the Asian P. tsus-simense, have recently been combined into a single species because they have similar leaf design and share petiole scales with long cilia both on the margin and faces, dark narrow scales on the petiole and rachis, and large, entire, persistent indusia (Roux, 2000
). We sampled these plants from Asia and Africa: since as the two are separated by only three unambiguous differences (plus 14 ambiguous) they would be at the lower end of the variation that we observed between accessions of the same species. Consequently we agree that P. tsus-simense ought to be included in P. luctuosum. Although P. luctuosum and P. neolobatum are sister taxa with the same chromosome number (x = 123; Löve et al., 1977
), their differences in morphology suggest that they have different origins. Our results suggest that apogamy is a prominent but not widespread evolutionary process in Polystichum and its allies.
Conclusions
Our analysis of rbcL sequence and morphological data for Polystichum and allied taxa yields four basic insights into the phylogeny of the group. First, Phanerophlebia and Polystichopsis together are the sister group to Polystichum s.l. including Cyrtomium. Second, Polystichum as commonly circumscribed is paraphyletic. We choose to recognize the clade of earliest diverging Polystichum species as a distinct genus (Cyrtomidictyum) and to continue to recognize Cyrtomium as a separate genus, leaving a monophyletic Polystichum s.s. Third, evolution of Polystichum in the tropics appears largely confined to single continents. For instance, we resolved tropical American and African clades within Polystichum. Similarly, once-pinnate, bulb-bearing calciphilic species from Asia and the West Indies appear to have evolved independently. Fourth, members of the Polystichum alliance, including Phanerophlebia, Cyrtomium, Polystichopsis, and Cyrtomidictyum, share a common ancestor that likely had ciliate petiole-base scales, once-pinnate fronds, ultimate segments with scarious tips, peltate indusia, and microscales.
|
FOOTNOTES |
|---|
4 Author for reprint requests (dbarring{at}zoo.uvm.edu
)
|
LITERATURE CITED |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
Barrington D. S. 1990 Hybridization and allopolyploidy in Central American Polystichum: cytological and isozyme documentation. Annals of the Missouri Botanical Garden 77: 297-305[CrossRef][ISI]
Barrington D. S. 1992 Climate and the disjunct distribution of Polystichum alfarii (Christ) Barr. in Mesoamerica. Rhodora 94: 327-339[ISI]
Barrington D. S. 1993 Ecological and historical factors in fern biogeography. Journal of Biogeography 20: 275-280[CrossRef][ISI]
Barrington D. S. 1995 Polystichum. In G. Davidse, M. Sousa S., and S. Knapp [eds.], Flora Mesoamericana, vol. 1, 218–225. Universidad Nacional Autónoma de México, Mexico City, Mexico
Barrington D. S. C. H. Haufler C. R. Werth 1989 Hybridization, reticulation, and species concepts in the ferns. American Fern Journal 79: 55-64[CrossRef]
Christensen C. 1906 Index filicum. H. Hagerup, Copenhagen, Denmark
Daigobo S. 1972 Taxonomical studies on the fern genus Polystichum in Japan, Ryukyu, and Taiwan. Science Report Tokyo Kyoiku Daigaku 15: 57-80
Daigobo S. 1974 Chromosome numbers of the fern genus Polystichum. Journal of Japanese Botany 49: 371-378
Doyle J. J. 1992 Gene trees and species trees: molecular systematics as one-character taxonomy. Systematic Botany 17: 144-163[CrossRef][ISI]
Doyle J. J. J. L. Doyle 1987 A rapid DNA isolation procedure for small quantities of fresh leaf tissue. Phytochemical Bulletin 19: 11-15
Fraser-Jenkins C. R. 1997 Himalayan ferns: a guide to Polystichum. International Book Distributors, Dehra Dun, India
Goloboff P. A. 1993 NONA Version 2.0. Published by the author. Available at www.cladistics.com
Hasebe M. T. Omori M. Nakazawa T. Sano M. Kato K. Iwatsuki 1994 rbcL gene sequences provide evidence for the evolutionary lineages of leptosporangiate ferns. Proceedings of the National Academy of Sciences USA 91: 5730-5734
Hasebe M. et al 1995 Fern phylogeny based on rbcL nucleotide sequences. American Fern Journal 85: 134-181[CrossRef]
Knobloch I. 1996 Pteridophyte hybrids and their derivatives. Publication of the Michigan State University Department of Botany and Plant Pathology, East Lansing, Michigan, USA
Kung H. S. 1989 A taxonomical study of sect. Scleropolystichum Daigobo (Polystichum) from China. Acta botanica Boreali Occidentalia Sinica 9: 267-278
Löve Å. D. Löve R. E. G. Pichi Sermolli 1977 Cytotaxonomical atlas of the Pteridophyta. J. Cramer, Vaduz, Germany
Mabberley D. J. 1997 The plant book, 2nd ed. Cambridge University Press, New York, New York, USA
Manton I. 1950 Problems of cytology and evolution in the Pteridophyta. Cambridge University Press, Cambridge, UK
Mayer M. S. M. R. Mesler 1993 Morphometric evidence of hybrid swarms in mixed populations of Polystichum munitum and P. imbricans (Dryopteridaceae). Systematic Botany 18: 248-260[CrossRef][ISI]
Morton C. V. 1960 Observations on cultivated ferns VI. The ferns currently known as Rumohra. American Fern Journal 50: 145-155[CrossRef]
Mullenniex A. T. M. Hardig M. Mesler 1998 Molecular confirmation of hybrid swarms in Polystichum. Systematic Botany 23: 421-426[CrossRef][ISI]
Nakaike T. 1973 Studies in the fern genus Polystichum, I. Observations on the section Metapolystichum at Gobô-sawa, Pref. Chiba. Bulletin of Natural Science Museum Tokyo 16: 437-457
Nixon K. C. 2001 Winclada. Published by the author. Available at www.cladistics.com
Pichi Sermolli R. E. G. 1977 Tentamen Pteridophytorum genera in taxonomicum ordinem redigendi. Webbia 31: 313-512
Pryer K. M. A. R. Smith J. E. Skog 1995 Phylogenetic relationships of extant ferns based on evidence from morphology and rbcL sequences. American Fern Journal 85: 205-282[CrossRef][ISI]
Roux J. P. 2000 The genus Polystichum (Dryopteridaceae) in Africa. Bulletin of The Natural History Museum (Botany series) 30: 33-79
Tagawa M. 1940 Polystichum of Japan, Korea and Formosa I. Acta phytotaxonomica et geobotanica 9: 119-138
Tryon R. M. A. F. Tryon 1981 Taxonomic and nomenclatural notes on ferns. Rhodora 83: 133-138
Tryon R. M. A. F. Tryon 1982 Ferns and allied plants: with special reference to tropical America. Springer-Verlag, New York, New York, USA
Wagner D. H. 1979 Systematics of Polystichum in western North America north of Mexico. Pteridologia 1: 1-64
Wagner W. H. 1973 Reticulation of holly ferns (Polystichum) in western United States and adjacent Canada. American Fern Journal 63: 99-115
Wolf P. G. P. S. Soltis D. E. Soltis 1994 Phylogenetic relationships of dennstaedtioid ferns: evidence from rbcL sequences. Molecular Phylogenetics and Evolution 3: 383-392[CrossRef][Medline]
Xiang L. L. 1994 A taxonomical study of the genus Polystichum Roth section Haplopolystichum Tagawa from Yunnan. Acta Phytotaxonomica Sinica 32: 258-267
Yatskievych G. D. B. Stein G. J. Gastony 1988 Chloroplast DNA evolution and systematics of Phanerophlebia (Dryopteridaceae) and related genera. Proceedings of the National Academy of Sciences USA 85: 2589-2593
Zhang L. B. 1994 A taxonomic study on the fern genus Polystichum Roth Sect. Metapolystichum Tagawa from Sichuan, China. Acta Botanica Yunnanica 16: 131-138
Zhang L. B. H. S. Kung 1996a A taxonomic study of the fern genus Polystichum sect. Metapolystichum Tagawa from China (II). Acta Phytotaxonomica Sinica 34: 68-76
Zhang L. B. H. S. Kung 1996b A taxonomic study of the fern genus Polystichum sect. Metapolystichum Tagawa from China (III). Acta Phytotaxonomica Sinica 34: 194-203
This article has been cited by other articles:
|
|
 |
|
  H. E. Driscoll and D. S. Barrington Origin of Hawaiian Polystichum (Dryopteridaceae) in the context of a world phylogeny Am. J. Botany, August 1, 2007; 94(8): 1413 - 1424. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
