| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Systematics |
2University of Utah, Utah Museum of Natural History, 1390 E. President's Circle, Salt Lake City, Utah 84112 USA; 3Missouri Botanical Garden, P.O. Box 299, St. Louis, Missouri 63166 USA
Received for publication April 8, 2003. Accepted for publication July 24, 2003.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONLITERATURE CITED |
|---|
Key Words: Argyrochosma • Cheilanthes • Cheilanthoideae • Cheiloplecton • chromosomes • cytotaxonomy • Mexico • Notholaena • Pteridaceae • western United States
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONLITERATURE CITED |
|---|
; Pinkava et al., 1992
; Baldwin et al., 2002
). Realizing the value of chromosome data for evolutionary studies, systematists have pursued the long-term goal of determining chromosome numbers for as many species and populations as possible (Carr et al., 1999
; Strother and Panero, 2001
). Ubiquitous or economically important families such as Asteraceae and Poaceae have benefited greatly from this attention, but many other groups remain undersampled.
Among the undersampled taxa, the xeric-adapted cheilanthoid ferns (Pteridaceae subfamily Cheilanthoideae) present a special challenge to cytogeneticists. As is typical of homosporous ferns (see Haufler and Soltis, 1986
), the chromosome numbers of cheilanthoids are high, with the lowest record being x = 27. Many species are confined to remote desert habitats, where they tend to grow from rock cracks or specialized substrates that make transplantation difficult. Finally, meiotic studies in the group have been limited by the fact that sporangia are scattered (not aggregated into discrete sori) and usually are protected by a covering of hairs, scales, and/or reflexed marginal indusia. All these factors have contributed to the perception that cheilanthoid ferns are chromosomally intractable, with the result that fewer than 20% of the species have been counted to date.
The limited number and questionable accuracy of counts available for many cheilanthoid ferns have hindered attempts to determine generic base numbers and resolve phylogenetic relationships (Reeves, 1979
). The problems are especially acute in Cheilanthes, the largest and most diverse genus of subfamily Cheilanthoideae. Published reports for Cheilanthes suggest that closely related species (or populations of a single species) often have different base numbers. Although infraspecific chromosomal variability is common in angiosperms, it is rarely encountered among the ferns and fern allies, in which the chromosome base numbers of most genera are extraordinarily stable (Britton, 1974
). In genera with a variable base number, such as Lycopodium (Wilce, 1972
) and Thelypteris (Smith, 1971
), each subgenus or section usually has a stable number. Examples of seemingly random variation in chromosome base number are rare, and most of these seem to occur in groups (such as the cheilanthoid ferns) in which little cytogenetic work has been done. The systematic value of chromosome data for these taxa is limited because it is unclear whether reported variations in base number represent real biological phenomena or erroneous counts. The situation can be resolved only by more intensive sampling of the taxa in question. The current study was undertaken for this purpose.
The goals of this study are twofold: (1) to expand the taxonomic and geographic sample of chromosome counts available to researchers interested in the systematics of cheilanthoid ferns and (2) to use this information in conjunction with published reports to assess the stability of generic and infrageneric base numbers in the group. This report focuses on taxa found in the western United States and Mexico, the region in which cheilanthoid ferns achieve their greatest diversity (Tryon and Tryon, 1973
). Species from other regions will be addressed in future papers.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONLITERATURE CITED |
|---|
). Meiotic chromosome counts were derived from sporocytes at late diplotene or diakinesis. A few mitotic chromosome counts were obtained by germinating spores from herbarium specimens following procedures outlined by Windham et al. (1986)
. The resultant gametophytes were prepared using the protocol of Windham and Haufler (1986)
. Meiotic and mitotic cells providing interpretable counts were photographed using a Zeiss (Zeiss, Oberkochen, Germany) phase contrast microscope and Kodak Technical Pan 2415 film (Kodak, Rochester, New York, USA). Photographs are provided for those taxa for which our determinations differ from previously published counts.
To guide the discussion, we compiled a list of previously published chromosome counts for all taxa studied. This list was assembled by checking all accepted names and commonly used synonyms through the Cytotaxonomical Atlas of the Pteridophyta (Löve et al., 1977
) and a complete set of the Index to Plant Chromosome Numbers spanning the period 1975–1997 (Goldblatt, 1981
, 1984
, 1985
, 1988
; Goldblatt and Johnson, 1990
, 1991
, 1994
, 1996
, 1998
, 2000
). The primary literature was consulted to obtain voucher information and check photographic documentation for each count identified by this search.
A significant number of cheilanthoid ferns reproduce apogamously (Gastony and Windham, 1989
), and an overview of the apogamous life cycle is necessary to understand some of the methods and results of this study. In the standard Döpp-Manton type of apogamy, sporocytes experience an endomitotic event just prior to meiosis that doubles the number of chromosomes in the cells. Thus, for example, chromosomally unbalanced triploids momentarily behave as hexaploids. Through pairing of duplicate chromosomes, meiosis proceeds normally. The resultant spores are genetically identical to the triploid sporophytes from which they were derived (Gastony and Windham, 1989
), and they germinate to produce triploid gametophytes. Because the sporophytes and gametophytes exist at the same ploidy level (i.e., n = 2n), completion of the apogamous life cycle does not involve fertilization. Instead, the rapidly developing gametophyte produces a sporophyte by simple budding, usually without the formation of sexual organs (gametangia).
Due to the frequency of apogamy among cheilanthoid ferns and its bearing on perceived ploidy levels, gametophyte cultures were established (following Windham et al., 1986
) for all taxa whose life cycle had not been ascertained by previous studies. Free water necessary for fertilization was withheld from these cultures to determine whether the gametophytes were capable of producing sporophytes apogamously. If sporophytes developed under these conditions (especially in the absence of gametangia), the taxon was considered apogamous. However, if gametangia developed normally and sporophyte production did not occur without fertilization, the taxon was classified as sexually reproducing.
|
RESULTS AND DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONLITERATURE CITED |
|---|
|
separated the two groups because of consistent differences in rhizome scales, leaf architecture, sporangial distribution, spore and gametophyte morphology, chemical composition of the farinose indument, and chromosome number. Subsequent chromosome work on this group yielded first reports for four taxa (Table 1). Both the widespread A. incana and the Mexican endemic A. pallens were sexual diploids with n = 27. Argyrochosma formosa and A. limitanea subsp. mexicana proved to be apogamous triploids with n = 2n = 81. An earlier tetraploid count for A. jonesii (as Notholaena jonesii in MacNeill et al., 1978
) was corroborated by another collection from California, but two samples from Arizona revealed that the species also encompasses a sexual diploid cytotype exhibiting n = 27. Published reports of n = 27 for A. delicatula (Windham, 1987
) and A. fendleri (as Notholaena fendleri in Brooks, 1986
) are based on our work and are repeated here because the original citations lacked voucher information. The three remaining taxa of Argyrochosma appearing in Table 1 are subject to conflicting reports that require additional discussion.
Wagner (1963)
published a count of n = 27 for A. dealbata (as Notholaena dealbata). This report was misquoted as 2n = 58 by Löve et al. (1977)
, contributing to the mistaken assumption (e.g., Tryon and Tryon, 1982
) that the base number of Argyrochosma was x = 29. Additional collections from three populations spanning much of the range of A. dealbata confirm that the chromosome number of the species is, indeed, n = 27 (Fig. 1).
|
listed a count of 2n = c. 84 from Colossal Cave, Arizona, while Benham and Schaack (1988)
reported n = 81 from essentially the same population. Samples from four additional localities in southeastern Arizona demonstrate that this taxon is an apogamous triploid with n = 2n = 81 (Fig. 2).
Knobloch et al. (1973)
reported chromosome counts of n = 29 and 2n = c. 116 for A. microphylla (as Notholaena parvifolia). This is the only report in the primary literature that supports assertions by Tryon and Tryon (1982)
that the base number of Argyrochosma is x = 29. However, the photograph documenting this report is just as easily interpreted as n = 27, leaving the true chromosome number of this species in question. Material gathered during this study near the northern and southern limits of the species range clearly show that the diploid chromosome number of A. microphylla is n = 27 (Fig. 3).
Nine of the 16 species attributed to Argyrochosma by Windham (1987)
have now been studied cytogenetically. All taxa showed chromosome counts based on x = 27, a number unique among cheilanthoid ferns. Recent molecular work by Gastony and Rollo (1998)
strongly supports the monophyly of this group and indicates that Argyrochosma is more closely related to Pellaea (with x = 29) than it is to Notholaena (with x = 30).
Aspidotis
Although Aspidotis is sometimes included in Cheilanthes (e.g., Tryon and Tryon, 1982
), most recent authors follow Lellinger (1968)
in treating this small group as a distinct genus. Our study mirrors previous counts by Smith (1975)
for A. californica, A. carlotta-halliae, and A. densa. All but one recognized taxon (the African species A. schimperi) have now been studied chromosomally, and they consistently have a base number of x = 30.
Astrolepis
Until recently, the star-scaled cloak ferns were assigned to either Notholaena (Tryon, 1956
; Lellinger, 1985
) or Cheilanthes (Mickel, 1979
; Tryon and Tryon, 1982
). However, members of this group have a unique combination of morphological features that led Benham and Windham (1992)
to treat them as a distinct genus. The data presented in Table 1 augment reports by Benham and Windham (1992)
and Benham (1992)
for A. cochisensis (subspp. arizonica and chihuahuensis), A. sinuata subsp. mexicana, and A. windhamii. They also corroborate earlier approximate counts for A. cochisensis subsp. cochisensis (as Notholaena cochisensis in Knobloch and Tai, 1978
) and A. integerrima (as N. integerrima in Knobloch et al., 1973
). All but one recognized taxon (the Mexican species A. beitelii) have now been counted, suggesting that the base number of the group is uniformly x = 29. This character distinguishes members of the segregate genus Astrolepis from all species of Notholaena sensu stricto (s.s.) (x = 30; discussed later) and the majority of those assigned to Cheilanthes (also x = 30). The base number of x = 29 is shared with Pellaea, some elements of which appear as the sister group of Astrolepis in the molecular analyses of Gastony and Rollo (1998)
.
Bommeria
The only previous report for B. elegans (as Hemionitis elegans in Haufler and Soltis, 1986
) is based on our work and is repeated here because the original citation lacks voucher information. Studies on this group also supported earlier counts for B. hispida (Smith, 1974
; Gastony and Haufler, 1976
) and B. pedata (Gastony and Haufler, 1976
). All species assigned to this genus have now been studied chromosomally, and they consistently have a base number of x = 30.
Cheilanthes
With more than 150 species worldwide (Tryon and Tryon, 1982
), Cheilanthes is, by far, the largest genus of cheilanthoid ferns. Because the group is very diverse morphologically, studies emphasizing different characters have produced very different classification schemes. For purposes of this discussion, we have chosen the infrageneric classification proposed by Tryon and Tryon (1982)
, which specifically addresses the diversity of forms found in the Americas. This classification divides the New World species of Cheilanthes among 10 informal species groups, five of which are represented in our sample. A number of morphologically isolated species could not be accommodated into this scheme, and these were treated separately as taxa of uncertain affinity. Each of these groups are next discussed individually.
Cheilanthes ("brandegei group")
Our study yielded a single chromosome count for this species group. It was a first report for the Mexican endemic C. aurea, a sexual diploid with n = 30. Our survey of published records indicates that none of the other taxa listed by Tryon and Tryon (1982)
as belonging to this group have been counted.
Cheilanthes ("fraseri group")
New counts for this group include a first report for the Mexican endemic C. longipila. This species appears closely related to the more northern taxa C. parryi and C. lanosa, and all three share a diploid chromosome number of n = 30. The data presented in Table 1 also corroborate earlier published counts for C. bonariensis (as Notholaena aurea in Knobloch et al., 1973
) and C. newberryi (as N. newberryi in Knobloch, 1967
). There are conflicting reports for C. feei, the only other member of the C. fraseri group included in this study. Although Knobloch (1967)
reported a chromosome number of 2n = 87 for this species, collections from two separate populations in Arizona clearly show n = 2n = 90 (Fig. 4). Given the lack of photographic documentation for Knobloch's count, the sole report of x = 29 in this group must be considered suspect. The data at hand suggest that the base number of the C. fraseri group is uniformly x = 30. Nevertheless, a broader sample of this large, morphologically diverse group will be necessary to confirm this hypothesis.
Cheilanthes ("marginata group")
A first report for C. cuneata indicates that it is an apogamous triploid with n = 2n = 90. The only previous record for this species group was a count of 2n = 87–90 on a plant identified as C. pyramidalis by Knobloch et al. (1975)
. Reexamination of the voucher specimen deposited at MSC revealed that this collection represents the taxon now known as C. arizonica. Three different populations of this species in Arizona (including the type locality in Ramsey Canyon) yielded counts of n = 2n = 90 (Fig. 5). Although most of the species assigned to this group by Tryon and Tryon (1982)
have not been studied, the available data support a base number of x = 30.
Cheilanthes ("microphylla group")
Our study yielded the first chromosome counts for C. horridula, which encompasses a sexual diploid cytotype showing n = 29 and a sexual tetraploid with n = 58. Although this species was assigned to the C. myriophylla group by Tryon and Tryon (1982)
, Reeves (1979)
argued that its morphological affinities lay with C. alabamensis, a core member of the C. microphylla alliance. Our counts of n = 29 and n = 58 for C. horridula appear to support Reeves' classification of this taxon, given that the C. microphylla group is the only assemblage of New World Cheilanthes species with a consistent base number of x = 29. The current study also confirmed previous reports based on x = 29 for C. alabamensis (Knobloch, 1967
; Knobloch et al., 1975
) and C. aemula (Knobloch et al., 1975
).
There is one conflicting report concerning the chromosome number of C. aemula, which was given as n = 30 by Knobloch (1967)
. However, there is good reason to doubt the validity of this count because the report of 2n = 58 by Knobloch et al. (1975)
is derived from the same collection. The two populations of C. aemula included in this study derive from the same area of northeastern Mexico and clearly show n = 29 (Fig. 11).
|
have now been counted, and all show a base number of x = 29. This is in sharp contrast to the majority of Cheilanthes species, which consistently exhibit x = 30. These chromosomal differences are accompanied by morphological distinctions that suggest a close relationship between the C. microphylla group and the genus Pellaea (Cranfill, 1980
). Further work will be necessary to ascertain whether this species group should remain in Cheilanthes, be transferred to Pellaea, or be recognized as a distinct genus.
Cheilanthes ("myriophylla group")
A first report for C. lindheimeri indicates that this taxon is an apogamous triploid with n = 2n = 90. The data presented in Table 1 also corroborate earlier published counts for C. covillei (Smith, 1974
; Schaack et al., 1984
), C. fendleri (Windham, 1983
; Windham and Schaack, 1983
), and C. wootonii (Windham, 1983
). All remaining species of the C. myriophylla group included in this study are subject to conflicting reports in the literature and are next discussed individually.
Cheilanthes tomentosa has been counted repeatedly, and the majority of published reports indicate that this species is an apogamous triploid with n = 2n = 90 (Knobloch, 1966
, 1967
; Whittier, 1970
; Knobloch et al., 1975
). The conflict in this case involves counts of 2n = 87 and n = c. 89 published by Wagner et al. (1970)
and Lloyd (1966)
, respectively. The first reference clearly states that the chromosome numbers listed are based upon the literature and are not new counts. Because there are no other records of 2n = 87 for C. tomentosa, this report is considered erroneous. Aside from being an approximate count that could be reinterpreted as n = 90, Lloyd's report is also suspect because this species is not known to occur in the northern Trans-Pecos region of Texas (Correll, 1956
). Attempts to locate the voucher specimen for Lloyd's count at UC were unsuccessful (A. R. Smith, University of California, personal communication), and the report of n = c. 89 for C. tomentosa should therefore be set aside because the identification cannot be confirmed. With these two reports excluded, all documented chromosome counts for C. tomentosa are n = 2n = 90, a number confirmed by the three populations included in this study (Fig. 6).
Previous reports for Cheilanthes eatonii (here circumscribed to include C. castanea following Reeves, 1979
) include counts of 2n = 87 (Lloyd, 1966
; Knobloch et al., 1975
) and n = 2n = 90 (Knobloch et al., 1975
; Benham and Schaack, 1988
). As was the case with C. tomentosa, the count by Lloyd (1966)
is approximate and the identification of the collection cannot be confirmed because there is no voucher available at UC. As such, this report does not provide strong evidence for a base number of x = 29 in C. eatonii. The same is true of the 2n = 87 report by Knobloch et al., who cast doubt on their own count by stating "...we cannot say with certainty whether our first count of 2n = 87 was wrong or whether there are two cytotypes in C. castanea" (Knobloch et al., 1975
, p. 651). The three collections analyzed during this study (which included both eatonii and castanea morphs) provided an abundance of clear preparations showing n = 2n = 90 (Fig. 7).
Knobloch et al. (1975)
reported a count of 2n = 109 ± 3 for Cheilanthes lendigera and suggested that this species might be a tetraploid based on x = 29. Our studies confirm that C. lendigera is indeed tetraploid, but the photos clearly indicate that the chromosome number is n = 60 (Fig. 8), not 2n = 116 as hypothesized by previous authors. A similar situation was encountered in C. villosa, for which Knobloch (1967)
reported a count of 2n = 87. Collections from two populations of C. villosa from southern Arizona indicate that the chromosome number of this species is n = 90 (Fig. 9).
The final disputed count in the Cheilanthes myriophylla group involves a report by Knobloch (1967)
of 2n = 116 in a collection identified as C. wootonii. This species subsequently proved to be an apogamous triploid with n = 2n = 90 (Windham, 1983
; Table 1). Reexamination of the voucher specimen for Knobloch's tetraploid count (deposited at MSC) revealed that this collection represents C. yavapensis, a species initially segregated from C. wootonii by Reeves (1979)
. Isozyme studies indicate that C. yavapensis is a distinct allopolyploid species that probably originated through hybridization between C. lindheimeri and C. covillei (Gastony and Windham, 1989
). With both parental taxa having chromosome counts based on x = 30 (Table 1), it is highly unlikely that C. yavapensis deviates from that base number. In agreement with three earlier counts of C. yavapensis by Windham (1993a)
, the two populations surveyed for this study were composed of apogamous tetraploids showing n = 2n = 120 (Fig. 10).
Reeves (1979)
reviewed the available chromosome data for the C. myriophylla group (= subg. Physapteris) but was unable to draw any conclusions concerning the base number because of the many conflicting reports. With the addition of the counts listed in Table 1, a clearer picture is emerging. Nearly 60% of the species included in the C. myriophylla group have now been studied chromosomally, all of which have yielded counts based on x = 30. Chromosome counts based on x = 29 have been attributed to nearly half of these species by previous authors, but these earlier reports appear to be erroneous. Samples of all but two of the taxa previously reported as x = 29 were acquired for this study, and the resultant counts were, without exception, based on x = 30. The only species missing from our sample were C. chipinquensis and C. myriophylla. Knobloch and Lellinger (1969)
attributed a chromosome number of 2n = 58 to the former, and Lloyd (1966)
reported a count of n = 2n = 87 for the latter. However, Knobloch et al. (1975)
provided micrographs for these species that showed 2n = 60 and 2n = c. 90, respectively. Considering that all recent, photographically documented counts are based on x = 30, we conclude that this is the true base number of the C. myriophylla group.
Cheilanthes ("incertae sedis")
This group represents a "dumping ground" for species whose relationships to other members of the genus are unclear. Included are first reports for C. brachypus, C. lozanii var. seemanii, and C. pringlei, all of which proved to be sexual diploids with n = 30. Published counts for C. gryphus (Mickel, 1987
) and C. subcordata (as Hemionitis subcordata in Haufler and Soltis, 1986
) are based on our work and are repeated here because the original citations lack voucher information. An earlier report of n = 30 for C. leucopoda (Knobloch, 1967
) was corroborated by our sampling. Although Knobloch (1966)
reported a count of 2n = 58 for C. wrightii, samples from two populations close to his collection locality in southern Arizona clearly showed n = 30 (Fig. 12). Given the morphological diversity represented in this group, it is unlikely that these species are closely related. However, all seven had a base number of x = 30, a number shared with the majority of species assigned to Cheilanthes.
Cheiloplecton
Although included in Pellaea by many earlier authors, Cheiloplecton is recognized as a distinct genus by Smith (1981)
and Mickel and Beitel (1988)
. It is one of the last cheilanthoid genera to be studied chromosomally, and the count presented here for C. rigidum var. lanceolatum is apparently the first report for the genus. This taxon, which is endemic to southern Mexico, proved to be an apogamous triploid with n = 90. This isolated chromosome count seems to support assertions by Smith (1981)
and Mickel and Beitel (1988)
that Cheiloplecton is more closely related to Cheilanthes (with x = 30) than it is to Pellaea (with x = 29).
Hemionitis
A single count for H. palmata from Jamaica supports earlier determinations by Wagner (1963)
, Walker (1966
, 1985
), Smith and Mickel (1977)
, and others. With counts available for the majority of species, it appears that the chromosome base number of Hemionitis (including Gymnopteris following Ranker, 1989
) is uniformly x = 30.
Mildella
The apogamous triploid count of n = 90 presented here for M. intramarginalis var. serratifolia is the first report for the genus in the Americas and suggests a base number of x = 30. However, published chromosome counts for the Asian species M. henryi (Tsai and Shieh, 1983
) and M. nitidula (as Pellaea nitidula in Verma and Goloknath, 1967
) are consistently based on x = 29. Species from the two regions differed substantially in spore ornamentation, and the chromosomal data seem to support the hypothesis (Tryon and Tryon, 1982
) that Asian and American species of Mildella represent separate evolutionary lines.
Notholaena
The typification and circumscription of this genus have been contentious topics in systematic pteridology (Yatskievych and Smith, in press
). Studies by Tryon and Tryon (1982)
and Windham (1993b)
revealed that the American taxa traditionally assigned to Notholaena represent at least four distinct evolutionary lines. With the transfer of several species groups to Cheilanthes (Tryon and Tryon, 1982
) and the recognition of Argyrochosma (Windham, 1987
) and Astrolepis (Benham and Windham, 1992
) as separate genera, the remainder of the American species form a coherent, monophyletic group. However, the correct name for this group is in dispute because Notholaena has been lectotypified by several authors citing three different type species. Following Yatskievych and Smith (in press)
, we accept the first lectotypification based on N. trichomanoides, a member of the monophyletic group mentioned earlier. The second and third lectotypifications (the last favored by some European workers, e.g., Pichi-Sermolli, 1983
, 1989
) are based on Old World taxa unrelated to the species herein called Notholaena.
This group has received very little attention from cytotaxonomists, and our study yielded first reports for 10 taxa. Notholaena bryopoda, N. candida, N. lemmonii, N. rosei, N. schaffneri, and N. sulphurea all proved to be sexual diploids with n = 30. Both N. aschenborniana and N. grayi subsp. grayi yielded counts of n = 2n = 90, indicating that they are apogamous triploids. An even higher ploidy level was encountered in N. californica subsp. californica, which proved to be an apogamous pentaploid with n = 2n = 150. First counts for N. neglecta revealed that it encompasses two ploidy levels; a sexual diploid cytotype with n = 30 and an apogamous triploid with n = 2n = 90. Our study also corroborated earlier counts for N. copelandii and N. greggii (Benham and Schaack, 1988
), N. galeottii and N. rigida (Knobloch et al., 1973
), and N. trichomanoides var. subnuda (as forma subnuda in Walker, 1973
). Chromosome counts are now available for 16 of the c. 25 species included in Notholaena s.s., all of which exhibit a base number of x = 30.
Pellaea
Tryon and Tryon (1982)
divided this genus into four "sections," two of which are restricted to the Old World, with a third confined to tropical South America. The biogeographical, morphological, and chromosomal differences among these "sections" suggest that most are worthy of generic recognition, a conclusion supported by the molecular data of Gastony and Rollo (1998)
. All of the taxa included in this report represent section Pellaea, which has been very well studied chromosomally. As such, the only new count represents a previously unknown apogamous tetraploid cytotype of P. intermedia showing n = 2n = 116.
Within the group commonly known as the light-stipe cliff brakes, previous reports of n = 29 for Pellaea andromedifolia (Tryon, 1968
; Smith, 1974
) and P. notabilis (Tryon, 1972
) were confirmed. Among the species with dark stipes and concolorous rhizome scales, this study verified earlier counts for P. glabella subsp. simplex (as var. simplex in Tryon and Britton, 1958
) and P. breweri (Tryon and Britton, 1958
). In the case of the latter species, a photograph of the Nevada count was published previously by Windham and Haufler (1986)
, but the collection is included here because the original report lacked voucher information.
Pentagramma
Studies by Yatskievych et al. (1990)
revealed that the genus Pityrogramma sensu lato (s.l.) comprised two distinct evolutionary lines differing in several important features (including chromosome base number) that suggested they belonged to different tribes within the family Pteridaceae. Subsequent molecular analyses by Gastony and Rollo (1998)
have shown that the tropical members of Pityrogramma (including the type species) do not belong to subfamily Cheilanthoideae. Although Pityrogramma is thus excluded from consideration here, it is appropriate to discuss the temperate taxa that Yatskievych et al. (1990)
transferred to the genus Pentagramma. This group was strongly supported as a member of Cheilanthoideae in the rbcL analyses of Gastony and Rollo (1998)
.
As currently defined, Pentagramma comprises two species, one of which (P. pallida) is endemic to north-central California, while the other (P. triangularis) ranges over much of western North America and is divisible into several subspecies. Although California populations of P. triangularis have been well studied chromosomally, the Sonoran Desert taxon (subsp. maxonii) has received little attention. Thus, it is not surprising that our survey revealed an unreported sexual tetraploid cytotype in subsp. maxonii showing n = 60. The only previous report for this taxon was a diploid count of n = 30 published by Yatskievych et al. (1990)
.
Previous reports of a diploid cytotype of Pentagramma triangularis subsp. semipallida (as Pityrogramma triangularis var. semipallida in Smith, 1980
) from California are corroborated by a new record from Tehama County. A series of counts for Pentagramma triangularis subsp. triangularis from California and Oregon supports earlier records (as Pityrogramma triangularis in Alt and Grant, 1960
; Smith et al., 1971
; Smith, 1974
) of a widespread diploid cytotype with n = 30. Two additional determinations for subsp. viscosa from San Diego County, California, agree with previous reports (as Pityrogramma viscosa in Alt and Grant, 1960
) from the same region. With chromosome counts available for all currently recognized taxa, it is clear that the base number of Pentagramma is consistently x = 30.
Stability and evolution of chromosome base numbers in cheilanthoid ferns
Our data set, comprising the largest single block of cheilanthoid chromosome counts published to date, provides important new insights into the stability and evolution of base numbers in the group. Whereas earlier attempts at synopsis (e.g., Knobloch et al., 1975
; Reeves, 1979
) were confounded by seemingly random variations in base number, our study reveals remarkable stability comparable to that observed in other fern groups (Britton, 1974
). We present a total of 131 counts representing 75 taxa of cheilanthoid ferns from the western United States and Mexico. The 16 species groups represented here do not have a single case of within-group variation in base number. The genus Cheilanthes s.l. has some variability, with x = 29 in the C. microphylla group and x = 30 in all other species sampled. However, Cheilanthes is highly polyphyletic (Gastony and Rollo, 1998
), and the distinctive nature of the C. microphylla group suggests that it may be worthy of generic recognition.
In addition to being a stable attribute of species groups among the cheilanthoid ferns we studied, chromosome base number appears to be a relatively conservative trait in the evolution of the subfamily. Figure 13 shows Gastony and Rollo's (1998)
rbcL phylogeny for the group, pruned to show only the genera and species groups included in our study. From this, it is evident that x = 30 is the plesiomorphic character state for subfamily Cheilanthoideae. Although restricted to the higher branches of the tree, taxa with x = 29 occur in two separate clades and do not appear to form a monophyletic group. A broader survey of the Pteridaceae as a whole indicates that x = 29 has several independent origins in the family (Gastony and Yatskievych, 2001
).
|
. This hypothesis requires a reversal to x = 30 along the branch (bootstrap 52) leading to the C. myriophylla alliance and its sister clade. Alternatively, x = 29 may have originated twice, once along the branch leading to the C. microphylla group and again near the base of the Pellaea/Astrolepis clade. This second origin could have occurred either on the branch (bootstrap 88) immediately subtending Pellaea and Astrolepis or on the branch (bootstrap 85) below Argyrochosma. In the first case, the unique base number of Argyrochosma is derived directly from x = 30; in the second, x = 29 represents an intermediate step in a process of descending aneuploidy ultimately leading to x = 27.
The fact that we can even propose hypotheses for the origin of x = 29 in subfamily Cheilanthoideae means that our understanding of the group is improving. However, much more remains to be done. Nearly half of the taxa assigned to Argyrochosma and Notholaena s.s. remain uncounted, and the sampling of Cheilanthes is woefully inadequate. The type species of the genus (C. micropteris) has not been subject to chromosomal or molecular analyses, so we are unable to determine which branch of Gastony and Rollo's (1998)
phylogenetic tree properly bears the name Cheilanthes. Additional potentially erroneous counts remain ensconced in the literature, and the chromosome numbers of most South American taxa are unknown.
As we work to fill these gaps in our knowledge of cheilanthoid ferns, the accuracy of our observations must be paramount. Inaccurate counts are the bane of the cytogenetic literature because it is nearly impossible to disprove an erroneous report. Photographic documentation is essential, especially among the homosporous ferns with high base numbers. No one has described the situation more clearly than Irene Manton, whose work provides the foundation and inspiration for modern fern cytogenetics. As Manton (1950, p. ix) said, "In a group like the Pteridophyta, where the technical difficulties are so great that it has been my unfortunate lot to have to correct errors in the work of almost every previous investigator, the attainment of accuracy has been a primary task without which no valid general conclusions could have been drawn. For this reason the use of photography has assumed a special importance...what cannot be photographed cannot be used as evidence."
|
|
|
|
FOOTNOTES |
|---|
4 E-mail: windham{at}umnh.utah.edu
|
LITERATURE CITED |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONLITERATURE CITED |
|---|
Baldwin B. G. B. L. Wessa J. L. Panero 2002 Nuclear rDNA evidence for major lineages of helenioid Heliantheae. Systematic Botany 27: 161-198
Benham D. M. 1992 Additional taxa in Astrolepis. American Fern Journal 82: 59-62[CrossRef]
Benham D. M. C. G. Schaack 1988 Miscellaneous chromosome counts. In A. Löve [ed.], Chromosome number reports XCIX Taxon 37: 399.
Benham D. M. M. D. Windham 1992 Generic affinities of the star-scaled cloak ferns. American Fern Journal 82: 47-58[CrossRef]
Britton D. M. 1974 The significance of chromosome numbers in ferns. Annals of the Missouri Botanical Garden 61: 310-317[CrossRef][ISI]
Brooks R. E. 1986 Polypodiaceae. In T. Barkley [ed.], Flora of the Great Plains, 49–69. University Press of Kansas, Lawrence, Kansas, USA
Carr G. D. R. M. King A. M. Powell H. Robinson 1999 Chromosome numbers in Compositae. XVIII. American Journal of Botany 86: 1003-1013
Correll D. S. 1956 Ferns and fern allies of Texas. Texas Research Foundation, Renner, Texas, USA
Cranfill R. 1980 Ferns and fern allies of Kentucky. Kentucky Nature Preserves Commission Scientific and Technical Series No. 1, Frankfort, Kentucky, USA
Gastony G. J. C. H. Haufler 1976 Chromosome numbers and apomixis in the fern genus Bommeria (Gymnogrammaceae). Biotropica 8: 1-11[CrossRef]
Gastony G. J. D. R. Rollo 1998 Cheilanthoid ferns (Pteridaceae: Cheilanthoideae) in the southwestern United States and adjacent Mexico—a molecular phylogenetic reassessment of generic lines. Aliso 17: 131-144
Gastony G. J. M. D. Windham 1989 Species concepts in pteridophytes: the treatment and definition of agamosporous species. American Fern Journal 79: 65-77[CrossRef]
Gastony G. J. G. A. Yatskievych 2001 Morphological reassessment of molecular phylogenetic inferences in the xerically adapted cheilanthoid ferns (Pteridaceae: Cheilanthoideae). Botany 2001 Abstracts: 82–83
Goldblatt P. [ed.] 1981 Index to plant chromosome numbers 1975–1978. Monographs in Systematic Botany from the Missouri Botanical Garden 5: 1-553
Goldblatt P. [ed.] 1984 Index to plant chromosome numbers 1979–1981. Monographs in Systematic Botany from the Missouri Botanical Garden 8: 1-427
Goldblatt P. [ed.] 1985 Index to plant chromosome numbers 1982–1983. Monographs in Systematic Botany from the Missouri Botanical Garden 13: 1-224
Goldblatt P. [ed.] 1988 Index to plant chromosome numbers 1984–1985. Monographs in Systematic Botany from the Missouri Botanical Garden 23: 1-264
Goldblatt P. D. E. Johnson [eds.] 1990 Index to plant chromosome numbers 1986–1987. Monographs in Systematic Botany from the Missouri Botanical Garden 30: 1-243
Goldblatt P. D. E. Johnson [eds.] 1991 Index to plant chromosome numbers 1988–1989. Monographs in Systematic Botany from the Missouri Botanical Garden 40: 1-238
Goldblatt P. D. E. Johnson [eds.] 1994 Index to plant chromosome numbers 1990–1991. Monographs in Systematic Botany from the Missouri Botanical Garden 51: 1-267
Goldblatt P. D. E. Johnson [eds.] 1996 Index to plant chromosome numbers 1992–1993. Monographs in Systematic Botany from the Missouri Botanical Garden 58: 1-276
Goldblatt P. D. E. Johnson [eds.] 1998 Index to plant chromosome numbers 1994–1995. Monographs in Systematic Botany from the Missouri Botanical Garden 69: 1-208
Goldblatt P. D. E. Johnson [eds.] 2000 Index to plant chromosome numbers 1996–1997. Monographs in Systematic Botany from the Missouri Botanical Garden 81: 1-188
Haufler C. H. D. E. Soltis 1986 Genetic evidence suggests that homosporous ferns with high chromosome numbers are diploid. Proceedings of the National Academy of Sciences, USA 83: 4389-4393
Holmgren P. K. N. H. Holmgren L. C. Barnett [eds.] 1990 Index herbariorum. Part I: The herbaria of the world, 8th ed. New York Botanical Garden, Bronx, New York, USA
Knobloch I. W. 1966 Chromosome numbers in Cheilanthes and Polypodium. American Journal of Botany 53: 288-291[CrossRef][ISI]
Knobloch I. W. 1967 Chromosome numbers in Cheilanthes, Notholaena, Llavea and Polypodium. American Journal of Botany 54: 461-464[CrossRef][ISI]
Knobloch I. W. D. B. Lellinger 1969 A new species of Cheilanthes from Mexico. American Fern Journal 59: 8-10[CrossRef]
Knobloch I. W. W. Tai 1978 The chromosome number of Notholaena cochisensis. American Fern Journal 68: 63.
Knobloch I. W. W. Tai T. Adangappuram 1975 Chromosome counts in Cheilanthes and Aspidotis with a conspectus of the cytology of the Sinopteridaceae. American Journal of Botany 62: 649-654[CrossRef][ISI]
Knobloch I. W. W. Tai T. Ninan 1973 The cytology of some species of the genus Notholaena. American Journal of Botany 60: 92-95[CrossRef][ISI]
Lellinger D. B. 1968 A note on Aspidotis. American Fern Journal 58: 140-141[CrossRef]
Lellinger D. B. 1985 A field manual of the ferns and fern allies of the United States and Canada. Smithsonian Institute Press, Washington, D.C., USA
Lloyd R. M. 1966 Chromosome counts for Pteridaceae. In A. Löve [ed.], IOPB chromosome number reports VIII Taxon 15: 282-283
Löve A. D. Löve R. E. G. Pichi-Sermolli 1977 Cytotaxonomical atlas of the Pteridophyta. J. Cramer, Vaduz, Liechtenstein
MacNeill C. D. W. Brophy A. R. Smith 1978 Ferns of the New York mountains, California, with biogeographic comments. Madroño 25: 54-56
Manton I. 1950 Problems of cytology and evolution in the Pteridophyta. Cambridge University Press, Cambridge, UK
Mickel J. T. 1979 The fern genus Cheilanthes in the continental United States. Phytologia 41: 431-437
Mickel J. T. 1987 A new fern from western Mexico and its bearing on the taxonomy of the cheilanthoid ferns. American Fern Journal 77: 109-114[CrossRef]
Mickel J. T. J. M. Beitel 1988 Pteridophyte flora of Oaxaca, Mexico. Memoirs of the New York Botanical Garden 46: 1-568
Pichi-Sermolli R. E. G. 1983 Fragmenta pteridologiae. VIII. Webbia 37: 111-140
Pichi-Sermolli R. E. G. 1989 Again on the typification of the generic name Notholaena R. Brown. Webbia 43: 301-310
Pinkava D. J. B. D. Parfitt M. A. Baker R. D. Worthington 1992 Chromosome numbers in some cacti of western North America—VI, with nomenclatural changes. Madroño 39: 98-113
Ranker T. A. 1989 Spore morphology and generic delimitation of New World Hemionitis, Gymnopteris, and Bommeria. American Journal of Botany 76: 297-306[CrossRef][ISI]
Reeves T. 1979 A monograph of the fern genus Cheilanthes subgenus Physapteris (Adiantaceae). Ph.D dissertation, Arizona State University, Tempe, Arizona, USA
Schaack C. R. Hevly M. Rusche 1982 Miscellaneous chromosome counts. In A. Löve [ed.], IOPB chromosome number reports LXXV Taxon 31: 366-367
Schaack C. J. Morefield M. D. Windham 1984 Miscellaneous chromosome counts. In A. Löve [ed.], Chromosome number reports LXXXIII Taxon 33: 353-354
Semple J. C. J. G. Chmielewski M. A. Lane 1989 Chromosome number determinations in fam. Compositae, tribe Astereae. III. Additional counts and comments on generic limits and ancestral base numbers. Rhodora 91: 296-314[ISI]
Smith A. R. 1971 Chromosome numbers of some New World species of Thelypteris. Brittonia 23: 354-360[CrossRef][ISI]
Smith A. R. 1974 Taxonomic and cytological notes on ferns from California and Arizona. Madroño 22: 376-378
Smith A. R. 1975 The California species of Aspidotis. Madroño 23: 15-24
Smith A. R. 1981 Pteridophytes. In D. Breedlove [ed.], Flora of Chiapas, Part 2, 1–370. California Academy of Sciences, San Francisco, California, USA
Smith A. R. J. T. Mickel 1977 Chromosome counts for Mexican ferns. Brittonia 29: 391-398[CrossRef][ISI]
Smith D. M. 1980 Flavonoid analysis of the Pityrogramma triangularis complex. Bulletin of the Torrey Botanical Club 107: 134-145[CrossRef][ISI]
Smith D. M. S. P. Craig J. Santarosa 1971 Cytological and chemical variation in Pityrogramma triangularis. American Journal of Botany 58: 292-299[CrossRef][ISI]
Strother J. L. J. L. Panero 2001 Chromosome studies: Mexican Compositae. American Journal of Botany 88: 499-502
Tryon A. F. 1968 Comparisons of the sexual and apogamous races in the fern genus Pellaea. Rhodora 70: 1-24
Tryon A. F. 1972 Spores, chromosomes and relations of the fern Pellaea atropurpurea. Rhodora 74: 220-241
Tryon A. F. D. M. Britton 1958 Cytotaxonomic studies on the fern genus Pellaea. Evolution 12: 137-145[CrossRef][ISI]
Tryon R. M. 1956 A revision of the American species of Notholaena. Contributions to the Gray Herbarium 179: 1-106
Tryon R. M. A. F. Tryon 1973 Geography, spores and evolutionary relations in the cheilanthoid ferns. In A. C. Jermy et al. [eds.], The phylogeny and classification of the ferns, 145–153. Botanical Journal of the Linnean Society 67, Supplement 1
Tryon R. M. A. F. Tryon 1982 Ferns and allied plants, with special reference to tropical America. Springer-Verlag, New York, New York, USA
Tsai J. L. W. C. Shieh 1983 A cytotaxonomic survey of the pteridophytes in Taiwan. Journal of Science and Engineering (Taiwan) 20: 137-158
Verma S. C. M. Goloknath 1967 Miscellaneous chromosome counts. In A. Löve [ed.], IOPB chromosome number reports X Taxon 16: 146-148
Wagner W. H., Jr. 1963 A biosystematic survey of United States ferns: preliminary abstract. American Fern Journal 53: 1-16
Wagner W. H., Jr. D. Farrar D. McAlpin 1970 Pteridology of the Highlands Biological Station area, southern Appalachians. Journal of the Elisha Mitchell Scientific Society 86: 1-27
Walker T. G. 1966 A cytotaxonomic survey of the pteridophytes of Jamaica. Transactions of the Royal Philosophical Society of Edinburgh 66: 169-237
Walker T. G. 1973 Additional cytotaxonomic notes on the pteridophytes of Jamaica. Transactions of the Royal Philosophical Society of Edinburgh 69: 109-135
Walker T. G. 1985 Cytotaxonomic studies of the ferns of Trinidad 2. The cytology and taxonomic implications. Bulletin of the British Museum of Natural History (Botany) 13: 149-249
Whittier D. P. 1970 The rate of gametophyte maturation in sexual and apogamous species of ferns. Phytomorphology 20: 30-35[ISI]
Wilce J. H. 1972 Lycopod spores, I. General spore patterns and the generic segregates of Lycopodium. American Fern Journal 62: 65-79[CrossRef]
Windham M. D. 1983 The ferns of Elden Mountain, Arizona. American Fern Journal 73: 85-93
Windham M. D. 1987 Argyrochosma, a new genus of cheilanthoid ferns. American Fern Journal 77: 37-41[CrossRef]
Windham M. D. 1993a New taxa and nomenclatural changes in the North American fern flora. Contributions from the University of Michigan Herbarium 19: 31-61
Windham M. D. 1993b Notholaena. In Flora of North America Editorial Committee [eds.], Flora of North America North of Mexico, vol. 2, 143–149. Oxford University Press, New York, New York, USA
Windham M. D. C. H. Haufler 1986 Biosystematic uses of fern gametophytes derived from herbarium specimens. American Fern Journal 76: 114-128[CrossRef]
Windham M. D. C. G. Schaack 1983 Miscellaneous chromosome counts. In A. Löve [ed.], IOPB chromosome number reports LXXXI Taxon 32: 664-665
Windham M. D. P. G. Wolf T. A. Ranker 1986 Factors affecting prolonged spore viability in herbarium collections of three species of Pellaea. American Fern Journal 76: 141-148[CrossRef]
Yatskievych G. A. A. R. Smith In press Typification of Notholaena R. Br. (Pteridaceae). Taxon.
Yatskievych G. A. M. D. Windham E. Wollenweber 1990 A reconsideration of the genus Pityrogramma (Adiantaceae) in western North America. American Fern Journal 80: 9-17[CrossRef][ISI]
| ||||||||||||||
