Abstract

Introduction

The internal transcribed spacer region (ITS) of the 18S-26S ribosomal DNA cistron has been used successfully to reconstruct phylogenies at the generic and species levels with numerous examples from the Asteraceae (in Calycadenia, Argyroxiphium, Dubautia, Wilkesia, Adenothamnus, Madia, Raillardella, and Railardiopsis, Baldwin, 1992, 1993; in Krigia, Kim and Jansen, 1994; in Ratibida, Dracopsis, and Rudbeckia, Urbatsch and Baldwin, 1993; and in the Cardueae, Susanna et al., 1995). ITS has several advantages that make it an ideal region to sequence for phylogenetic analysis of congeneric species: 1) its rate of evolution is appropriate for studies at the specific and generic levels; 2) it is phylogenetically interpretable, i.e., the sequences are relatively easy to align because there tends to be very little length variation at the generic level in flowering plants; 3) it is large enough to offer potentially enough characters for phylogenetic reconstruction; and 4) it is flanked by regions that are highly conserved within genera, making PCR amplification and sequencing straightforward.

The primary goal of this study was to reconstruct the phylogeny of all (30+) amphimictic Antennaria species based on sequence divergence in ITS-1 and ITS-2. It was hoped that many of the relationships that remained uncertain in earlier studies (Bayer, 1990a; Bayer, unpublished cpDNA RFLP data) could be resolved through sequence analysis. Also, ITS provides an independent data set for comparison with the phylogenetic hypotheses that were produced in the earlier studies.

Materials and Methods

DNA isolation and PCR amplification - All of the DNAs were isolated from plants collected from natural populations, except Leontopodium alpinum which was purchased from a commercial plant nursery and subsequently cultivated in the greenhouse. Voucher specimens are deposited at ALTA.

Total DNA was isolated from 0.7 to 1.5 grams of fresh leaf material using a modified CTAB method (Doyle and Doyle, 1987), with 1.0 % b-mercaptoethanol (instead of 0.2 %) used in the extraction buffer. In most cases, RNA in the resulting samples was digested with RNAase A (Sigma R-9009, Sigma Chemical Corp., St. Louis, MO) according to the manufacturer's instructions. DNA was reprecipitated with ice-cold 95% ethanol, washed in 70% ethanol, and resuspended in TE.

The ITS region was amplified via the polymerase chain reaction (PCR) using Replitherm® DNA polymerase (Epicentre Technologies, Madison, WI). The PCR reaction mixture consisted of 5ml of 20X reaction buffer, 6ml of 25mM magnesium chloride solution, 16ml of a 1.25 mM dNTP solution in equimolar ratio, 25 pmol of each primer, 10 - 50 ng of template DNA, and 0.5 unit of Replitherm, all in a total volume of 100ml. The PCR samples were heated to 94 C for two minutes prior to the addition of Replitherm to denature proteases and nucleases. The double-stranded PCR products were produced via 30 cycles of denaturation (94 C for 1.5 min), primer annealing (55 C for 2 min), and extension (72 C for 3 min). A 15-min final extension at 72 C followed cycle 30.

The two ITS sequences were amplified separately. ITS-1 was amplified using the primers 1407F (D. Nickrent, Southern Ill. Univ., Carbondale, IL, pers. comm.) and ITS2 (White et al., 1990) in equal proportions to produce double-stranded product (Fig. 1), whereas ITS3 (White et al., 1990) and 307R (D. Nickrent, pers. comm.) were used to amplify the ITS-2 region. Double-stranded products were then used as templates to produce single-stranded DNA using the ITS2 primer to produce single-stranded DNA of the ITS-1 region and ITS3 to produce single-stranded DNA of the ITS-2 region (Fig. 1). PCR reactions to produce single-stranded DNA were the same as for double-stranded except only one primer (25 pmol) was used in the amplifications. The single-stranded DNA was precipitated with 20 % PEG/2.5 M NaCl, washed in 70 % EtOH, washed a second time in 95 % EtOH and then resuspended in 7ml of TE (Morgan and Soltis, 1993) prior to sequencing. Often, a second set of single-stranded products was produced for use in a second sequencing reaction using manganese to increase the yield of short fragments (per U.S. Biochemical Corp., Cleveland, OH).

Sequencing the single-stranded DNA - The single-stranded DNAs were sequenced using the dideoxy chain termination method (Sanger, Nicklen, and Coulson, 1977) with the use of the Sequenase® version 2.0 kit (U.S. Biochemical, Cleveland, OH) and ³⁵S-dATP initially without the addition of manganese. The ITS1 primer (White et al., 1990) was used to sequence the ITS-1 region, and the ITS4 (White et al., 1990) primer was used to sequence the ITS-2 region (Fig. 1). Fragments were separated in 6.0 % polyacrylamide gels (0.4 mm thickness; 1X TBE buffer) at 2,000 volts/80 watts. The gels were fixed in 10 % acetic acid for 20 minutes, washed in distilled water, and allowed to air dry. They were then used to expose Kodak X-Omat AR film for 24-36 hr.

Sequence analysis and phylogenetic reconstruction - The ITS sequences were aligned visually, and the alignment of the sequences required interpretation of several small (1-2 bp) insertion/deletion (indel) events and one seven-bp indel in ITS-1 (Appendix 1). Alignment of the ITS-2 sequences required the interpretation of several single-bp indels, and one each of three-, four- and five-bp indels (Appendix 1). The proportion of nucleotide differences among pairs of species was calculated using the Kimura 2-parameter model and the MEGA program (Kumar, Tamura, and Nei, 1993). A total of 81 potentially phylogenetically informative nucleotide substitutions in ITS-1 and ITS-2 was used in the analysis of the 34 taxa. The entire sequence of Pseudognaphalium microcephalum (Appendix 1) is given as a reference sequence. The sequences for the remaining 33 taxa have been submitted to the Genome Sequence Data Base and the GSDB accession numbers are given in Table 1.

Phylogenetic reconstruction was performed using PAUP version 3.1.1 (Swofford, 1991) on unweighted characters by heuristic searches using Fitch parsimony and stepwise SIMPLE addition of data. The outgroup in the analyses included the four taxa mentioned above. Invariant sites and strictly autapomorphous base changes were also ignored in the phylogenetic reconstruction ("ignore uninformative characters" option). Indels were coded as missing data following the recommendation of Wojciechowski et al. (1993) and therefore ignored in the analysis. They were later mapped on the phylogenetic reconstructions to assess their phylogenetic utility (Fig. 2). The "Tree - Bisection - Reconnection" (TBR) branch swapping option in conjunction with saving all minimal trees (MULPARS) and accelerated transformation (ACCTRAN) were used to search for the shortest topologies. Branches of zero length were collapsed to reduce the number of equally parsimonious trees. Heuristic searches employing 100 replicates of a stepwise random (RANDOM) addition of taxa were conducted to search for other groups of trees (i.e., islands; Maddison, 1991) that are equal to in length or shorter than the most parsimonious trees. Two analyses were conducted using the options described above; in the first, the monophyly of Antennaria was tested, and its potential sister-group relationships were explored. The outgroup was rooted as a basal polytomy and because the first analysis clearly showed Antennaria to be monophyletic relative to the outgroup included, a second analysis, in which the ingroup (Antennaria) was considered monophyletic and the outgroup was considered as paraphyletic, was conducted to examine further the possible sister-group of Antennaria. A strict consensus tree was constructed from the first analysis and 50% majority-rule consensus tree (Margush and McMorris, 1981) was produced (showing all compatible groups) from the second. Consensus trees were inspected using MacClade (Maddison and Maddison, 1992). In addition, a reanalysis of morphology (data matrix modified from Bayer, 1990a) for only those species of Antennaria also analyzed for ITS sequences was performed for comparison with the current study.

The relative support for the various clades was determined by bootstrap analysis (Felsenstein, 1985) employing 100 replicates. To avert the problems of memory exhaustion and unrealistically long analyses, a maximum of 2500 trees was saved during each bootstrap replicate. Additionally, a decay analysis (Bremer, 1988; Donoghue et al., 1992) was performed following the general methods of Johnson and Soltis (1994) to assess the strength of each clade.

Results

For the most part, the indels in the ITS sequences are autapomorphous (Fig. 2) and were therefore phylogenetically uninformative. Several indels, however, were phylogenetically informative, including a five-bp insertion ("PQRST"; Fig. 3) in ITS-2 (Appendix 1) which is a synapomorphy that defines the Ewartia-Leontopodium-Antennaria clade, two single-base insertions ("I,J"; Fig. 2) in ITS-1 (Appendix 1) which are synapomorphies that define the A. rosulata-A. marginata clade, and a single-bp insertion ("I"; Fig. 2) in ITS-1 (Appendix 1) which is a synapomorphy that defines the A. densifolia-A. corymbosa clade.

Phylogenetic reconstruction of Antennaria - The 50 % majority-rule tree presented (length = 212) (Fig. 3) is identical to one of the 2118 most parsimonious trees. A strict consensus tree was also constructed to determine the relative stability of the various clades in the tree (Fig. 2) and to test the monophyly of Antennaria. "Trivial" names have been placed on the majority-rule tree (Fig. 3) for ease of referral. In addition, a 50 % majority-rule consensus of 135 shortest trees found in a reanalysis of morphology (data matrix modified from Bayer, 1990a) for only those species of Antennaria also analyzed for ITS sequencing is presented (Fig. 4).

Based on the taxa analyzed for ITS sequence variation, the strict consensus tree (Fig. 2) indicates that Antennaria is monophyletic a result supported by 11 synapomorphies. The 50 % majority-rule consensus tree (Fig. 2) indicates that the possible sister genus of Antennaria is Leontopodium, and that Antennaria is more distantly related to Ewartia, Anaphalis, and Pseudognaphalium. Using Leontopodium, Ewartia, Anaphalis, and Pseudognaphalium as the outgroup, Antennaria is monophyletic, supported by 6 synapomorphies and with a bootstrap value of 94 % (Fig. 3). The strict consensus tree also indicates that Antennaria is composed of six clades, the Geyeriae, Arcuatae, Argenteae, Dimorphae, Pulcherrimae, and Catipes (Fig. 2). In the 50 % majority-rule tree (Fig. 3), these six groups form two subclades within the genus. One is composed of the Geyeriae, Argenteae, Arcuatae, Dimorphae, and Pulcherrimae here referred to as the "Leontipes" group and is supported by 3 synapomorphies with a boostrap value of 39 % . The "Leontipes" group is sister to the large Catipes group, which is supported by 8 synapomorphies and has 99 % bootstrap support. Within the Catipes the "neglecta" group is sister to the remainder of the clade (bootstrap value of 29 %). The remaining 19 members share one synapomorphy which is plesiomorphic in A. neglecta and the rest of Antennaria (Fig. 3). Among the other 19 species of the Catipes group, two weakly supported clades are evident in the majority-rule tree (Fig. 3): the "friesiana group" and the "corymbosa group". Several smaller monophyletic lineages are evident within the Catipes group (fig. 2), but most of these are also weakly supported.

Discussion

Sequence divergence among species of Antennaria was 1 to 14 % for ITS-1 and 0 to 8 % for ITS-2. This is comparable to sequence divergence values reported in Calycadenia and Osmadenia (0 - 11.2 % in ITS-1; 0 - 8.6 % in ITS-2) (Baldwin, 1993), in the Madiinae (0.4 - 19.2 % in ITS-1; 0 - 12.9 % in ITS-2) (Baldwin, 1992), and in the Cardueae (1.2 - 15.9 % in ITS-1; 0.9 - 15.5 % in ITS-2) (Susanna et al., 1995). In all these cases ITS-1 is more variable than ITS-2, but both regions are sufficiently variable to make the ITS a useful tool for phylogenetic reconstruction at the level of tribe and below in the Asteraceae.

Circumscription of Antennaria and outgroup relationships - ITS sequences have been very useful in reconstructing the phylogeny of Antennaria and also in defining its relationship to the outgroup taxa. Previous cladistic analyses (Fig. 4) did not portray Antennaria as a monophyletic group. For example, cpDNA restriction site data (unpublished data) indicated that Antennaria was monophyletic only with the inclusion of Anaphalis. In contrast, phylogenetic analysis of morphology (Fig. 4) found that Antennaria would be monophyletic only if A. geyeri were removed from the genus. However, ITS sequence data clearly define Antennaria, including A. geyeri, as monophyletic, united by 11 synapomorphies (Fig. 2). Furthermore, Anaphalis is not a member of Antennaria based on ITS sequence data. Therefore, based on these data the recently suggested generic recircumscription of Antennaria that A. dimorpha be segregated into its own genus (Weber, 1987) may be unwarranted.

The sister-group relationships of Antennaria have been unexplored primarily because the tribe Gnaphalieae, to which Antennaria belongs, has been largely neglected by taxonomists. Until Anderberg's (1991) cladistic analysis of the Gnaphalieae, phylogenetic relationships among the approximately 167 genera of the tribe were uninvestigated. Stebbins (1974) suggested that the "immediate relative" of Antennaria was the cosmopolitan genus Gnaphalium. However, taxonomists have long recognized that Gnaphalium sensu lato was a highly polyphyletic assemblage of species, and the genus has recently been dismantled into smaller segregate genera, such as Pseudognaphalium, Anaphaloides, and Gamochaeta, among several others (Hilliard and Burtt, 1981; Anderberg, 1991). Anderberg's (1991) analysis showed that the primarily Asian genus Anaphalis and the Australia-New Zealand genus Ewartia were the sister genera of Antennaria. Merxmüller, Leins, and Roessler (1977) included Antennaria in the "Anaphalis group" closely related to Anaphalis and the Eurasian genus Leontopodium. All of the cladistic analyses of Antennaria, based on morphology, cpDNA, and ITS sequences (Figs. 2 - 4), have shown that Gnaphalium s. l. (Pseudognaphalium) is more distantly related to Antennaria than is Anaphalis. However, based on the genera included here, ITS sequences (Fig. 3) suggest that Leontopodium is the sister genus of Antennaria, and together they are the sister genus to Ewartia. Additionally, the five-bp insertion (Fig. 3 "P-T") that defines the Ewartia-Leontopodium-Antennaria clade supports the monophyly of this group of genera. This is in conflict with Anderberg's analysis (1991), which showed Leontopodium as being quite distantly related to the Antennaria-Anaphalis-Ewartia clade. Until phylogenetic relationships of more members of the Gnaphalieae can be investigated using molecular techniques, the relationships among genera of the tribe will remain uncertain.

Phylogenetic relationships within Antennaria - Antennaria has been informally divided into six phenetic groups (Bayer, 1990a), the Geyeriae, Argenteae, Dimorphae, Pulcherrimae, Dioicae, and Alpinae. The majority-rule consensus tree (Fig. 3) shows that Antennaria is composed of two major lineages, the "Leontipes" group, which consists of species that are restricted in their distributions to the western United States, and the Catipes group, comprising the "neglecta", "corymbosa", and "friesiana" groups, occurring throughout the northern hemisphere. The "Leontipes" group, which consists of five smaller groups, the Geyeriae, Arcuatae, Argenteae, Dimorphae, and Pulcherrimae in the strict consensus tree (Fig. 2), is composed of species that are primarily diploid (tetraploids are known only in A. dimorpha and A. pulcherrima, Bayer and Stebbins, 1987) and as far as is known always amphimictic. Most of the species of the "Leontipes" group lack horizontal stoloniferous growth (exceptions are A. flagellaris and A. arcuata). Based on morphology the "Leontipes" group is considered the likely basal group in the genus, based on a number of unspecialized morphological features, such as non-stoloniferous growth, lack of extensive polyploidy, and a general lack of well-developed sexual dimorphism. The Catipes group has amphimictic diploids and tetraploids and derived from them are all of the polyploid agamic complexes in the genus. Most of the species of the Catipes group have aggressive horizontal stolons, an effective means of asexual reproduction. Therefore, the Catipes group is considered the more morphologically specialized of the two major groups in Antennaria.

The strict consensus tree (Fig. 2) demonstrates that Antennaria is composed of six monophyletic groups of equal rank; for the most part these groups correspond to traditionally recognized groups (Bayer, 1990a). The Geyeriae group is monotypic, consisting of A. geyeri, a species characterized by a large number (30) of autapomorphous nucleotide substitutions, a four-bp deletion and a large ten-bp insertion (Fig. 2). The tendency toward polygamo-dioecioy in A. geyeri, along with its lack of basal leaves, make it more similar morphologically to Anaphalis the the remainder of Antennaria. Previous studies of morphology (Fig. 4) and cpDNA restriction sites (unpublished data) left unresolved the inclusion of A. geyeri within Antennaria. It is obvious from the current investigation, however, that A. geyeri should remain a member of the Antennaria (Figs. 2, 3).

Antennaria arcuata, the only member of the newly recognized Arcuatae, also has accumulated a relatively large number (10) of autapomorphous changes (Fig. 2). Previously, A. arcuata had been included in the Argenteae with A. luzuloides and A. argentea (Bayer, 1990a), but that relationship was always weakly supported. In fact, the Argenteae group is paraphyletic in cladistic analyses of both morphology (Fig. 4) and ITS sequences (Figs. 2, 3) if A. arcuata is included in the group. It seems best to recognize that A. arcuata belongs to a distinct lineage.

The Argenteae clade is composed of three taxa, A. argentea, A. luzuloides, and A. stenophylla (Fig. 2) and is supported by six synapomorphies in the strict consensus tree. It is the sister group to the A. arcuata - A. geyeri clade in the majority-rule tree (Fig. 3). Morphology (Fig. 4) indicates that A. argentea and A. luzuloides are sister taxa, whereas ITS sequences indicate that A. stenophylla is the sister taxon to A. luzuloides (Fig. 3). Moreover, morphology indicates that A. stenophylla should be a member of the Dimorphae clade (Fig. 4). The close relationship suggested between A. stenophylla and A. luzuloides based on ITS sequences is readily acceptable based on gross morphology in that they both have narrow, linear leaves and small flowering heads.

ITS sequence data provide support for a Pulcherrimae clade consisting of A. pulcherrima, A. anaphaloides, and A. lanata (Fig. 2). However, one surprising feature of the ITS-based tree (Fig. 2) is the inclusion of A. carpatica in the Catipes clade as the sister taxon to A. dioica. By contrast, morphology (Fig. 4) strongly suggests that A. carpatica is a member of the Pulcherrimae. The clade containing A. carpatica, the "nordhageniana" group (Fig. 3), is entirely European in distribution. It is also noteworthy that A. carpatica and A. dioica, also of this "nordhageniana" clade, hybridize in the Alps (Urbanska-Worytkiewicz, 1968), the area which served as the source of our material of A. carpatica (Table 1). It could be that past introgressive hybridization between A. carpatica and A. dioica has effectively transferred the ITS region of A. dioica into A. carpatica (at least in the material that we used in this study). The Pulcherrimae is a well-defined morphological group, and the suggestion based on sequence data that A. carpatica is a member of Catipes rather than the Pulcherrimae is not easily accepted. This result is problematical, and the sequencing of additional material of A. carpatica is needed.

The Catipes is a very well-supported group in both the strict (Fig. 2) and majority-rule (Fig. 3) consensus trees, although support for subclades within Catipes is weak. Traditionally, members of Catipes were split into two groups, the Alpinae, distributed in tundra, with black or olivaceous colored phyllaries, and the Dioicae taxa with phyllaries of lighter colors other than black or dark green. Based on ITS sequence data (Fig. 2 and 3), as well as morphology (Fig. 4), it is obvious that these two groups are unnatural, polyphyletic groups that should be abandoned. Species with dark phyllaries that grow in arctic and/or alpine tundra are probably the result of convergent evolution under similar environmental conditions (e.g., A. pulchella of the Sierra Nevada of the "corymbosa" group and A. aromatica of the "friesiana" group). Likewise, taxa with light-colored phyllaries, the Dioicae, are also a paraphyletic group. Lastly, although some have suggested that A. marginata and A. dioica are conspecific (Jepson, 1925), sequence data suggest that they are only distantly related (Figs. 2, 3).

Amphimixis, apomixis (agamospermy), and the very high levels of polyploidy (up to dodecaploid; Bayer and Minish, 1993) are prevalent among polyploid derivatives of the Catipes clade. The Catipes clade consists of diploids in which sexual dimorphism is highly evolved and in which an effective means of asexual vegetative reproduction, stoloniferous growth, is well developed (Bayer, 1990a). Additionally, many of the Catipes are specialized as edaphic endemics, such as A. virginica on Devonian-age shale barrens (Bayer and Stebbins, 1987, 1993), A. suffrutescens on serpentine (Bayer and Stebbins, 1993), and A. aromatica and A. densifolia on limestone talus (Bayer, 1989). Five polyploid agamic complexes, A. alpina (L.) Gaertn., A. howellii E. L. Greene, A. parlinii Fern., A. parvifolia Nutt., and A. rosea E. L. Greene, have evolved via multiple hybridization among members of the Catipes group (Bayer, 1987). The great success of the Catipes group seems to be correlated with their ability to grow in a diversity of habitats throughout their range from Great Britain across Eurasia and North America to Tierra del Fuego and to their acquisition of characters such as strong sexual dimorphism, aggressive vegetative reproduction (stolons), polyploidy, and agamospermy. Additional investigations, perhaps with a more rapidly-evolving nuclear DNA sequence, will be needed, however, before the phylogenetic relationships among members of the Catipes group can be resolved with greater certainty.

Phylogeography - There is some correspondence between geographic patterns and the phylogenetic relationships in Antennaria (Fig. 3). The "Leontipes" group, evident in the majority-rule tree, consists of the Geyeriae, Arcuatae, Argenteae, Dimorphae, and Pulcherrimae groups and is restricted in distribution to the western United States and portions of adjacent Canada (Fig. 3). The "neglecta" group is the sister group to the remainder of the Catipes group in the majority-rule tree and is perhaps the most widespread of the North American Catipes taxa (Fig. 3). Two clades form the "plantaginifolia" group, which are restricted to the eastern United States. The "nordhageniana" clade consists of three Eurasian taxa, although, as discussed above, the inclusion of A. carpatica in this group is anomalous (Fig. 3). Within the "friesiana" group are four clades, two of which have some geographic pattern to the relationships. The "marginata" group is composed of two species that have their centers of distribution in the southern Rockies and are largely restricted to Arizona and New Mexico. The four taxa of the "racemosa" group are from the northern Rockies.

Based on this study, the apparent sister genus of Antennaria is Leontopodium (Fig. 3), a genus most diverse in Chinese, Burmese, and European mountains (Anderberg, 1991). It is likely that the common ancestor that gave rise to both genera was also found in the Northern Hemisphere, perhaps in Asia and western North America. The most logical phytogeographic hypothesis concerning the origin and subsequent divergence of the genus Antennaria would maintain that it arose in the southern part of western North America, where the "Leontipes" clade of the genus is still found today (Fig. 3). At a later time, the most specialized clade of Antennaria, the Catipes group (Fig. 3), successfully spread into a diverse variety of habitats throughout the Northern Hemisphere. Recent speciation in Catipes appears to have occurred in rather limited geographic regions through adaptive radiation into very different niches. For example, the "racemosa" group contains A. aromatica, A. microphylla, A. racemosa, and A. umbrinella; all occur in the northern Rockies (Fig. 3), yet each species occupies its own distinct niche within that range (Bayer, Purdy, and Lebedyk, 1991). Similarly, the sister species A. marginata and A. rosulata (Figs. 2, 3) both occur exclusively in the southern Rockies, but each occurs in its own distinct habitat (Bayer, Purdy, and Lebedyk, 1991).

Conclusions - ITS sequences have been useful in resolving phylogenetic questions in the Cassiniinae. Considering the outgroup taxa that were included in the analysis, Leontopodium is the sister genus to Antennaria, and Anaphalis and Pseudognaphalium may be more distantly related to Antennaria than previously believed. Antennaria is a monophyletic group relative to the outgroups cited, and A. geyeri is clearly included within the genus. Within Antennaria are six clades, the Geyeriae, Arcuatae, Argenteae, Dimorphae, Pulcherrimae, and Catipes. The first five of these are geographically restricted to western North America and form the "Leontipes" group, in which polyploidy is rare and morphology is less specialized than in Catipes. The geographically widespread Catipes group is morphologically most specialized, and hybridization among these sexual species and subsequent acquisition of polyploidy and agamospermy by the hybrid entities has led to the complex pattern of reticulate evolution that is confined to the Catipes clade.

Literature Cited

------. 1991. Taxonomy and phylogeny of the tribe Gnaphalieae (Asteraceae). Opera Botanica 104: 1-195.

BALDWIN, B.G. 1992. Phylogenetic utility of the internal transcribed spacers of nuclear ribosomal DNA in plants: An example from the Compositae. Molecular Phylogenetics and Evolution 1: 3-16.

------. 1993. Molecular Phylogenetics of Calycadenia (Compositae) Based on Its Sequences of Nuclear Ribosomal DNA - Chromosomal and Morphological Evolution Reexamined. American Journal of Botany 80: 222-238.

 BAYER, R. J. 1985a. Investigations into the evolutionary history of the polyploid complexes in Antennaria (Asteraceae: Inuleae). II. The A. parlinii complex. Rhodora 87: 321-339.

------. 1985b. Investigations into the evolutionary history of the polyploid complexes in Antennaria (Asteraceae: Inuleae). I. The A. neodioica complex. Plant Systematics and Evolution 150: 143-163.

------. 1987. Evolution and phylogenetic relationships of the Antennaria (Asteraceae: Inuleae) polyploid agamic complexes. Biologisches Zentralblatt 106: 683-698.

------. 1989. A systematic and phytogeographic study of Antennaria aromatica and A. densifolia (Asteraceae: Inuleae) in the western North American cordillera. Madroño 36: 248-259.

------. 1990a. A phylogenetic reconstruction of Antennaria Gaertner (Asteraceae: Inuleae). Canadian Journal of Botany 68: 1389-1397.

------. 1990b. Investigations into the evolutionary history of the Antennaria rosea (Asteraceae: Inuleae) polyploid complex. Plant Systematics and Evolution 169: 97-110.

------, and D. J. CRAWFORD. 1986. Allozyme divergence among five diploid species of Antennaria (Asteraceae: Inuleae) and their allopolyploid derivatives. American Journal of Botany 73: 287-296.

------, and T. M. MINISH. 1993. Isozyme variation, ecology, and phytogeography of Antennaria soliceps (Asteraceae: Inuleae), an alpine apomict from the Spring Mountains, Nevada. Madroño 40: 75-89.

------, B. G. PURDY, and D. G. LEBEDYK. 1991. Niche differentiation among eight sexual species of Antennaria Gaertner (Asteraceae: Inuleae) and A. rosea, their allopolyploid derivative. Evolutionary Trends in Plants 5: 109-123.

------, and G. L. STEBBINS. 1987. Chromosome numbers, patterns of distribution, and apomixis in Antennaria (Asteraceae: Inuleae). Systematic Botany 12: 305-319.

------. 1993. A synopsis with keys for the genus Antennaria (Asteraceae: Inuleae: Gnaphaliinae) for North America. Canadian Journal Botany 71: 1589-1604.

BREMER, K. 1987. Tribal relationships of the Asteraceae. Cladistics 3: 210-253.

------. 1988. The limits of amino acid sequence data in angiosperm phylogenetic reconstruction. Evolution 42: 795-803.

------ , R. K. JANSEN, P. O. KARIS, M. KÄLLERSJÖ, S. C. KEELEY, K-. J. KIM, H. J. MICHAELS, J. D. PALMER, and R. S. WALLACE. 1992. A review of the phylogeny and classification of the Asteraceae. Nordic Journal of Botany 12: 141-148.

DONOGHUE, M. J., R. G. OLMSTEAD, J. F. SMITH, and J. D. PALMER. 1992. Phylogenetic relationships of Dipsacales based on rbcL sequences. Annals of the Missouri Botanical Garden 79: 333-345.

DOYLE, J. J. and J. L. DOYLE. 1987. A rapid DNA isolation procedure for small quantities of fresh leaf tissue. Phytochemical Bulletin 19: 11-15.

FELSENSTEIN, J. 1985. Confidence limits on phylogenies: An approach using the bootstrap. Evolution 39: 783-791.

HILLIARD, O. M. and B. L. BURTT. 1981. Some generic concepts in Compositae - Gnaphaliinae. Botanical Journal of the Linnaean Society 82: 181-232.

JEPSON, W. L. 1925. A manual of the flowering plants of California. Associated Students Store, Univ. of California, Berkeley, CA.

JOHNSON, L. A. and D. E. SOLTIS. 1994. matK DNA sequences and phylogenetic reconstruction in Saxifragaceae s. str. Systematic Botany 19: 143-156.

KIM, K. -J. and R. K. JANSEN. 1994. Comparisons of phylogenetic hypotheses among different data sets in dwarf dandelions (Krigia, Asteraceae): additional information from internal transcribed spacer sequences of nuclear ribosomal DNA. Plant Systematics and Evolution 190: 157-185.

KUMAR, S., K. TAMURA, and M. NEI. 1993. MEGA: Molecular Evolutionary Genetic Analysis, Version 1.0. The Pennsylvania State University, University Park, PA.

MADDISON, D. R. 1991. The discovery and importance of multiple islands of most parsimonious trees. Systematic Zoology 40: 315-328.

MADDISON, W.P. and D. R. MADDISON. 1992. MacClade, Analysis of phylogeny and character evolution, version 3. Sinauer Associates, Inc., Sunderland, MA.

MARGUSH, T. and F. R. McMORRIS. 1981. Consensus n-trees. Bulletin of Mathematical Biology 43: 239-244.

MERXMÜLLER, H., P. LEINS, and H. ROESSLER. 1977. Inuleae--systematic review. In V. Heywood, J. Harbourne, and B. L. Turner [Eds.], The biology and chemistry of the Compositae, 577-601. Academic Press, London.

MORGAN, D. R. and D. E. SOLTIS. 1993. Phylogenetic relationships among members of Saxifragaceae sensu lato based on rbcL sequence data. Annals of the Missouri Botanical Garden 80: 631-660.

SANGER, F., S. NICKLEN, and A. R. COULSON. 1977. DNA sequencing with chain-terminating inhibitors. Proceedings of the National Academy of Sciences of the United States of America 74: 5463-5467.

STEBBINS, G. L. 1974. Flowering plants: evolution above the species level. The Belknap Press of Harvard University Press, Cambridge, MA.

SUSANNA, A., N. G. JACAS, D. E. SOLTIS, and P. S. SOLTIS. 1995. Phylogenetic relationships in tribe Cardueae (Asteraceae) based on ITS sequences. American Journal of Botany (in press).

SWOFFORD, D. L. 1991. PAUP: Phylogenetic analysis using parsimony, version 3.3.1. Illinois Natural History Survey, Champaign, IL.

URBANSKA, K. M. l983a. Antennaria carpatica (Wahlb.) Bl. et Fing. s.l. in North America. I. Chromosome numbers, geographical distribution and ecology. Berichte des Geobotanischen Instituts der Eidgenössischen Technischen Hochschule Stiftung Rübel 50: 33-66.

-------. l983b. Cyto-geographical differentiation in Antennaria carpatica s.l. Botanica Helvetica 93: l23-l3l.

URBANSKA-WORYTKIEWICZ, K. M. l968. A new hybrid of the alpine flora: Antennaria carpatica x A. dioica. Berichte des Geobotanischen Instituts der Eidgenössischen Technischen Hochschule Stiftung Rübel 38: 20-27.

URBATSCH, L. E. and B. G. BALDWIN. 1993. ITS DNA sequence data and chloroplast restriction site data and the phylogenetics of coneflowers (Asteraceae). American Journal of Botany 80: 186-7.

WEBER, W. A. 1987. Colorado flora: western slope. Colorado Associated Univ. Press, Boulder, CO.

WHITE, T. J., T. BRUNS, S. LEE, and J. TAYLOR. 1990. Amplification and direct sequencing of fungal ribosomal RNA genes for phylogenetics. In M. Innis, D. Gelfand, J. Sninsky, and T. White [Eds.], PCR protocols: A guide to methods and applications, 315-322. Academic Press, San Diego, CA.

WOJCIECHOWSKI, M. F., M. J. SANDERSON, B. G. BALDWIN, and M. J. DONOGHUE. 1993. Monophyly of Aneuploid Astragalus (Fabaceae) - Evidence from Nuclear Ribosomal DNA Internal Transcribed Spacer Sequences. American Journal of Botany 80: 711-722.

Published in: American Journal of Botany 83: 516-527 (1996).