| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
What's this? |
Systematics and Phytogeography |
Section of Integrative Biology, The University of Texas at Austin, 1 University Station C0930, Austin, Texas 78712 USA
Received for publication January 3, 2007. Accepted for publication August 21, 2007.
ABSTRACT
The sunflower genus, Helianthus, is recognized widely for the cultivated sunflower H. annuus and scientifically as a model organism for studying diploid and polyploid hybrid speciation, introgression, and genetic architecture. A resolved phylogeny for the genus is essential for the advancement of these scientific areas. In the past, phylogenetic relationships of the perennial species and polyploid hybrids have been particularly difficult to resolve. Using the external transcribed spacer region of the nuclear 18S-26S rDNA region, we reveal for the first time a highly resolved gene tree for Helianthus. Phylogenetic analysis allowed the determination of a monophyletic annual H. sect. Helianthus, a two-lineage polyphyletic H. sect. Ciliares, and the monotypic H. sect. Agrestis, all of which were nested within a large perennial and polyphyletic H. sect. Divaricati. The distribution of perennial polyploids and known annual diploid hybrids on this phylogeny suggested multiple independent hybrid speciation events that gave rise to at least four polyploids and three diploid hybrids. Also provided by this phylogeny was evidence for homoploid hybrid speciation outside H. sect. Helianthus. Finally, previous hypotheses about the secondary chemistry in the genus were tested in a phylogenetic framework to obtain a better understanding of the evolution of these compounds in Helianthus.
Key Words: Asteraceae • ETS • Helianthus • hybrid • phylogeny • polyploidy • reticulate evolution • rDNA • secondary chemistry
Given the economic and cultural importance of the common sunflower, Helianthus annuus, there has been broad interest in the genus in which it evolved. Helianthus also has become an important genus for the study of speciation, especially hybrid speciation, thanks to the work of Rieseberg and his coworkers (Rieseberg, 1995
, 1996
, 1997
; Rieseberg et al., 1995
, 1999a
, b
, 2003
). Helianthus includes 49 species with native distributions covering much of North America. There have been many attempts at reconstructing the phylogeny of the genus using a range of different characters: morphological (Schilling and Heiser, 1981
), chemical (Spring and Schilling, 1989
, 1990
, 1991
), RFLP (Rieseberg, 1991
; Rieseberg et al., 1991
; Gentzbittel et al., 1992
; Schilling, 1997
), and DNA sequence data (Schilling et al., 1998
; Schilling, 2001
). To date, however, relationships between the species have been largely unresolved, especially within the perennials. Two main factors have contributed to this lack of resolution. One is the lack of markers that evolved fast enough to record the speciation events during the rapid radiation in the group, and the other is the difficulty in detecting and reconstructing the extensive hybrid speciation, both diploid and polyploid, which has strongly impacted the phylogenetic history of the genus.
Hybrid speciation is common in many groups of plants, but phylogenetic treatment of the origins of hybrid species has been relatively sparse in the literature. The origins and relationships of the hybrid species in Helianthus have long been of interest and still are largely unresolved in the case of the polyploids. Thirteen species have some form of polyploidy (Table 1), either tetraploidy or hexaploidy, and all are in the perennial sections. Helianthus section Ciliares has one polyploid species, whereas H. section Divaricati has 12. It is currently unknown which polyploid species are autopolyploids and which allopolyploids. All of the previously confirmed diploid hybrid species were in the annual section, H. section Helianthus (Rieseberg, 1990
, 1991
; Rieseberg et al., 1991
), although researchers have speculated about the existence of homoploid hybrid perennials (Long, 1955a
, 1963
). During the 1950s and 1960s, the literature was filled with studies on the perennial hybrid species, focusing primarily on crossing data and cytological analysis (Dewer, 1893
; Long, 1955a
, b
, 1957
, 1959
, 1960
, 1961
, 1963
; Jackson, 1956
; Jackson and Guard, 1956
, 1957
, 1958
; Smith and Guard, 1958
; Smith and Martin, 1959
; Heiser and Smith, 1960
; Smith, 1960
, 1961
; Clevenger, 1963
; Heiser and Smith, 1964
). Aside from this initial burst of research on potential origins of the polyploids, little work has focused on the perennial hybrids over the past 40 years. Instead, the annual hybrids have dominated the literature in the past 20 years and made Helianthus a model system for studying speciation. Rieseberg and his colleagues have pioneered much of this research, revealing that H. annuus and H. petiolaris have hybridized repeatedly to produce three homoploid hybrid species: H. deserticola, H. paradoxus, and H. anomalus (Rieseberg, 1990
, 1991
; Rieseberg et al., 1991
). Rieseberg also found evidence that H. bolanderi may be a more ancient homoploid hybrid formed from the ancestors of H. annuus and H. petiolaris (Rieseberg, 1991
). Subsequent molecular phylogenetic work has added support to the parentage of these annual hybrids (Schilling, 1997
, 2001
; Schilling et al., 1998
).
|
). The classification of Helianthineae is undergoing revision (Schilling and Panero, 2002
), but current data strongly suggest that Phoebanthus is sister to Helianthus (Schilling, 2001
; Schilling and Panero, 2002
) and that these two genera are sister to another clade containing Pappobolus (previously considered South American Helianthus), Simsia, Tithonia, and some Viguiera species (Schilling and Panero, 1996
, 2002
; Schilling, 2001
).
Although many botanists over the past two centuries have worked on the systematics of Helianthus (Candolle, 1836
; Torrey and Gray, 1842
; Gray, 1884
; Dewer, 1893
; Cockerell, 1919
; Watson, 1929
), the most comprehensive knowledge of the genus comes from Heiser's extensive morphological work and crossing studies (Heiser et al., 1969
). During the first half of the 20th century, many other botanists focused on smaller taxonomic studies within the genus, publishing some 50 papers characterizing Helianthus species and hybrids. Heiser et al. (1969)
summarized the extensive crossing studies between most of the 49 species and exposed the rampant ability of many Helianthus species to hybridize with one another. So common was hybridization within the genus that Heiser (1969)
remarked that "the discovery of another interspecific hybrid combination in the genus would scarcely be noteworthy." Schilling and Heiser (1981)
assigned formal names to the subgeneric classification for the 49 species using a morphological matrix of 42 characters and a PRIM network summarizing all the information from previous crossing studies.
The phylogenetic utility of flavonoids for chemotaxonomy also has been widely examined in Helianthus, although the study of these compounds as phylogenetic characters was abandoned because of the low number of synapomorphies for infrageneric clades, perhaps due to convergent evolution of the flavonoids used (Schilling and Mabry, 1983
; Rieseberg et al., 1987
). A resolved phylogeny would enable us to revisit these characterized traits, this time in the context of their evolution in Helianthus. Flavonoids have been described for about two-thirds of the species (Schilling and Mabry, 1981
, 1983
; Gao et al., 1987
; Rieseberg et al., 1987
; Schilling et al., 1987
). Some researchers proposed explanations regarding the presence/absence of flavonoids across sunflower species. Schilling and Mabry (1981)
hypothesized that flavone aglycones only occur in plants with glandular trichomes on the abaxial surface of their leaves, whereas chalcone aglycones, found in glabrous species, were sequestered in specialized epidermal cells or in the waxy cuticle as Rieseberg (1987)
found in H. annuus.
Sesquiterpene lactones (STLs), a different class of secondary compounds considered to function in plant defense, growth regulation, and allelopathy (Picman, 1986
), were characterized for three sections in Helianthus (Spring and Schilling, 1989
, 1990
, 1991
). Also scored in the Spring and Schilling studies was the presence/absence of capitate glands on the leaves, phyllaries, and anther appendages for each species. Spring and Schilling (1990)
observed differential expression of the STLs in different plant organs, e.g., they found that all Helianthus species have anther appendages that produce STLs but that species lacking leaf glands also lack STLs in their leaves. Although they noted this pattern, they reported the presence/absence of STLs over the entire plant and not for specific organs, so the direct correlation between leaf glands and STLs cannot be assessed using their published data. Yet, we can test for patterns that are consistent with this hypothesis. Because the leaf glands are a variable gland character, we can ask whether certain classes of STLs are absent when the leaf glands are absent. In other words, we can ask whether an STL is completely absent from a plant or lineage when leaf glands are also lost.
In a recent study (Linder et al., 2000
), considerable phylogenetic resolution was shown in a small sample of Helianthus species using the DNA sequence from the external transcribed spacer region (ETS) of the nuclear 18S-26S ribosomal repeat. The ETS sits upstream of the 18S subunit of the ribosomal DNA repeat and downstream of the nontranscribed spacer (NTS). Based upon those promising results, we collected complete ETS sequences for nearly all of the species in Helianthus and used those data to reconstruct a molecular phylogeny of the genus. This phylogeny was then used to address three main issues in the genus: whether the current subgeneric classification reflects natural groups, whether the evolution of polyploidy happened once or multiple independent times in the genus, and whether the evolution of secondary compounds in Helianthus correlates with glandular-trichome characters. The latter two questions are of broad interest to plant biologists. Detailed studies, like this one, that focus on polyploidy and the evolution of secondary compounds within a specific genus increase our understanding of the role that they might have played in the overall evolution of plants.
MATERIALS AND METHODS
Samples
Forty-seven of the 49 species in Helianthus were included in this study (Table 1): only H. neglectus and H. smithii were not sampled. Helianthus verticillatus, a recently rediscovered diploid perennial sunflower, was included in our sampling. Also included were representatives from several subspecies and, where possible, two or more individuals per species. Phoebanthus grandiflorus was used as the outgroup based on previous studies (Schilling et al., 1998
) and on the recommendation of J. Panero (University of Texas at Austin, personal communication). When they were available, we used the same DNA samples or individuals from the ITS study of Helianthus (Schilling et al., 1998
); these were kindly provided by L. Rieseberg (University of British Columbia) and are noted in Appendix 1. We also augmented these samples with our own collections of Helianthus. Voucher information along with GenBank accession numbers are also summarized in Appendix 1. Total DNA was extracted from silica-dried collections or dried herbarium specimens using the DNeasy Plant Minikit (Qiagen, Valencia, California, USA) and purified using the Qiaex II gel extraction kit (Qiagen).
Amplification and sequencing
The ETS region was PCR amplified in 25-µL reactions with 20 ng of genomic DNA; 30 mM tricine, pH 8.4; 2 mM MgCl2; 50 mM KCl; 5% acetamide, 100 mM dNTP; 0.5 µL of Taq polymerase; and 30 nM each of ETS1f and 18s2l primers (Linder et al., 2000
). Reactions were run on a thermal cycler (MJ Research, Waltham, Massachusetts, USA) with the following program: 94°C for 2 min; then 30 cycles of 94°C for 30 s, 50°C for 30 s, and 72°C for 90 s; and a final extension of 72°C for 5 min. Successful amplifications were visualized on an ethidium bromide-stained agarose gel and cloned using the TOPO-TA cloning kit (Invitrogen, Carlsbad, CA, USA). Ten colonies were picked from each plate and amplified using the M13 plasmid primers provided in the cloning kit. The same PCR protocol described earlier was performed with a hot start (95°C for 10 min, 72°C for 10 min), during which the Taq was added. Reaction products were cleaned with Sephadex G-50 columns (GE Healthcare Bio-Sciences, Piscataway, New Jersey, USA).
In Helianthus, the ETS has a large region of one to five subrepeats, each subrepeat being approximately 250 bp (Fig. 1). Because of differing numbers of subrepeats, individuals often had several ETS lengths as revealed by PCR amplification. For this study, up to three fragments of differing lengths per sample were chosen to be sequenced (cf. Linder et al., 2000
). Our goal was to sequence the diversity of lengths present in each sample to ensure that the full range of ETS sequences within each individual was included in our analyses. Fifty nanograms of amplified cloned DNA were cycle-sequenced using the BigDye Terminator mix (Perkin-Elmer, Waltham, Massachusetts, USA) with six internal primers specifically designed for Helianthus (Table 2). Sequences were cleaned with Sephadex columns and run on an MJ BaseStation (MJ Research) gel-based sequencer. Resulting sequences were assembled using Sequencher 4.5 (Gene Codes, Ann Arbor, Michigan, USA).
|
|
) followed by hand alignment using MacClade 4.05 (Maddison and Maddison, 2002
). Two aligned data sets were prepared for phylogenetic analyses. For one set of phylogenetic analyses, the 5' and 3' single copy regions of the ETS were concatenated into one partition, omitting the subrepeats (single copy alignment). For the other set of analyses, some of the subrepeats were also included based upon the following analysis of subrepeat orthology (single plus subrepeat alignment).
To determine the orthology of the subrepeats for their use in phylogenetic reconstruction, we aligned each subrepeat in all of the sequences (one to five per clone) and analyzed them phylogenetically using maximum likelihood (specific details are provided later) as in Koch et al. (2005)
. A total of 308 repeats (
250 bp each) were aligned and analyzed using maximum likelihood. Orthology was determined by identifying clades containing six or more taxa of single-copy repeats that mirrored a clade in the species phylogeny (Fig. 2a, clades labeled by a letter), even if relationships within the clades were much less resolved. Clades in Fig. 2 not labeled by a letter were not included in the larger analysis because of difficulty in determining the subrepeat orthology. Where there were obvious duplications within an orthologous set of repeats, one duplication was kept at random (Fig. 2b, c). Appendix S1 (see Supplemental Data accompanying the online version of this article) provides additional details on the subrepeat location included in final data matrix. The aligned orthologous subrepeats were then concatenated onto the single-copy ETS matrix (Fig. 2d). Once the alignment was finalized, identical clones within an individual were pruned from the data set, leaving only unique ETS sequences for each ETS type present in an individual.
|
) for maximum parsimony (MP), GARLI v. 0.93 (Zwickl, 2006
) for maximum likelihood (ML) searches, and MrBayes 3.1.1 (Ronquist and Huelsenbeck, 2003
) for searches using Bayesian inference. These three analyses were performed on each of two sets of taxa: one without the known hybrid or polyploid species (the nonhybrid, diploid-only data set) and one with the hybrids and polyploids included (the full-taxon data set). Maximum parsimony (MP) heuristic searches were started with a stepwise addition tree, tree-bisection-reconnection (TBR) branch swapping, and nchuck set to 100; they were then run for 1000 random addition replicates. For assessing branch support of MP trees, nonparametric bootstrap analysis was performed with 1000 replicates and the same settings as the original MP analysis, but with only 10 random addition replicates per bootstrap replicate. For the ML analyses, MrModeltest v. 2 (Nylander, 2004
) was used to identify the best-fit model. GARLI was run using a random-starting tree and the model determined from MrModeltest with random-starting parameters. The run was terminated after 10 000 generations without an improvement in the tree topology under a 0.05 score improvement threshold. Branch support was assessed with 500 nonparametric bootstrap replicates using the identical parameters as described. For the Bayesian analyses, default priors were used along with the model determined by MrModeltest. Posterior probabilities were assessed in two runs using four MCMC chains that ran until the split frequency between the independent runs dropped under 0.01, indicating stationarity. The first 25% of the trees were omitted in the split-frequency calculation and discarded as burn-in.
Tests for recombination
To examine the possibility of recombination between parental sequences in the ETSs of the diploid annual hybrids, we screened a reduced alignment containing only the annual species. Six recombination detection algorithms, all implemented in the program RDP2 (Martin et al., 2005
), were used to look for evidence of recombination. RDP, MaxChi, Chimera, Geneconv, and MaxChi were all run with the highest acceptable P-value cutoff set to 0.05. The settings for individual methods were as follows. For RDP, the reference sequences were set to internal and external and 90% or above sequence identity; a window of 10 variable sites was used. For Geneconv, the program scanned sequence triplets, indels were treated as one polymorphism, and the G-scale was set to 0. For Bootscan, the window scan was set to 200 with step size at 60. Two hundred bootstrap reps were performed with a 95% cutoff, and the analysis was run under the F84 model. MaxChi was set to scan triplets, the number of variable sites per window was set to 10, gaps were stripped, and 50 permutations were generated. Chimera used a window of 10 variable sites with 50 permutations generated. For SiScan, the scanning window was set to 200 with a step size of 60, gaps were stripped, only the 1/2/3 variable positions were used, and 1000 permutations were generated.
Character evolution
Forty-two secondary chemistry characters were collected from the literature for Helianthus (Appendix S2 of Supplemental Data accompanying online version of this article). Glandular trichome distribution on the upper (adaxial) and lower (abaxial) leaves, the phyllaries, and anther appendages were also scored from these studies (Herz and Kumar, 1981a
, b
; Ohno et al., 1981
; Schilling and Mabry, 1981
, 1983
; Watanabe et al., 1982
; Herz and Kulanthaivel, 1984
; Melek et al., 1985
; Rieseberg et al., 1987
; Schilling et al., 1987
; Spring, 1989
; Spring and Schilling, 1990
, 1991
; Alfatafta and Mullin, 1992
). Characters were coded as presence/absence on a pruned data set of one sequence per nonhybrid diploid species (38 ingroup taxa and one outgroup). When there was more than one ETS type per species, only one type was chosen at random. Subspecies were included in this core phylogeny if they had chemical and/or trichome information in the literature. A Bayesian analysis was then run on the reduced Helianthus data set using the same input parameters as described. Species with noncoalescing ETS types were collapsed to their most recent common ancestor, then only one ETS type was kept for subsequent analyses in an attempt to accommodate any ambiguity in the species phylogeny. Finally, all polytomies were resolved with very short (0.00001 substitutions/site) branches as required for the later analysis. Once the tree was finalized, the secondary chemistry states were placed onto the branch tips for the subsequent character analyses.
One goal was to determine whether transitions in secondary chemistry were associated with the presence of capitate glands on the abaxial surface of leaves (adaxial glands are less common and only present when abaxial glands are present). Because the data in the literature were collected over about 15 yr, more sophisticated extraction and isolation methods in the later studies resulted in finer-scale chemical profiles, leaving the earlier studies with missing data. We performed the correlated changes test only on characters for which we had complete data. Correlated changes tests were performed (lower-leaf [abaxial] gland vs. chemistry character) using Discrete (Pagel, 1994
) as part of the BayesTraits v1.0 software package (M. Pagel and A. Meade, University of Reading, Reading, UK, http://www.evolution.rdg.ac.uk/BayesTraits.html). For each pair of characters tested, two models of evolution were considered in the maximum likelihood framework. One assumed the traits evolved independently of each other, and the other assumed dependence (or correlated evolution). If the dependence model fit significantly better than the independence model, then correlated change was accepted after significance was assessed through a likelihood ratio test.
RESULTS
Features of ETS in Helianthus
The 5' to 3' ETS region consisted of a 351-bp, single-copy region, a region with one to five tandem 250 bp subrepeats (only some of which were used in the analysis), and an 898-bp single-copy region. These lengths were similar to those found by Linder et al. (2000)
. The total aligned region used for the phylogenetic analyses that included some of the subrepeats consisted of the 5' and 3' single-copy regions plus 1288 bp of informative subrepeats, bringing the total aligned matrix to 2545 bp. Phoebanthus grandiflorus had a 173-bp insertion between the 3' single-copy region and the subrepeat section, which was removed for the phylogenetic analyses. Five classes of subrepeats were identified as homologous and concatenated to the core ETS matrix. Each subrepeat class was assigned a letter. The 3'-most subrepeat (Z) was present and homologous in almost every individual (82), whereas the four other subrepeats were restricted to smaller sets of species. Subrepeat type Q was present in 42 individuals, type V in 15, type W in six, and type X in eight (Fig. 2). (See Appendix S1 of Supplemental Data accompanying online version of this article for more detail on the subrepeat classes). Orthology in the remaining subrepeats could not be determined, so the remaining subrepeats were omitted from subsequent analyses. Across the full-taxon alignment, 1853 bp were invariant, 692 were variable, and 309 were parsimony informative. The subrepeats contributed 66 of the parsimony-informative characters. The percentage of missing data in the matrix was 36.1%. If only the single-copy ETS matrix was included, missing data dropped to 1.6%. The GC content was 49% in the single-copy ETS regions and 52% in the subrepeat regions. The best-fit model, according to the AIC criterion in MrModeltest, was GTR+
+I for both taxon data sets (with and without hybrids and polyploids).
Phylogenetic analysis: the effect of including subrepeats
To determine the effect of inclusion of the subrepeats (and the increased amount of missing data) in the sequence data matrix, we conducted phylogenetic analyses with and without the subrepeats included for maximum parsimony, maximum likelihood, and Bayesian analyses. In all cases, the addition of the subrepeats to the single-copy ETS alignment resulted in resolution of seven polytomies in the perennial lineages without any alteration of the supported relationships resolved by the single-copy-only data set. Therefore, all analyses reported here are on the full alignment including the orthologous subrepeats.
Phylogenetic analysis: diploid, nonhybrid data sets and full taxon data sets
All phylogenetic analyses were performed on two taxon sets: the nonhybrid, diploid-only set and the full set of taxa. Bayesian, ML, and MP analyses on each set of taxa all produced similar topologies, with a few slight differences in resolution at the tips. The ML tree for each set of taxa (Figs. 3, 4) is shown with branch support values that summarize all three analyses. The single best-fit ML tree for the nonhybrid diploid phylogeny had an –ln score of 8623.16; the ML tree for the full data set had an –ln score of 11 649.807. As shown in the figures, the phylogeny was well resolved in both data sets for all three methods of reconstruction, with strong support for most of the major lineages. The Bayesian support for the deeper lineages was stronger than the bootstrap support, but the three analyses generally agreed on the support for the more derived lineages. The only supported topological difference among reconstructions occurred between different phylogenetic analyses on the full set of taxa (marked with a large asterisk in Fig. 4). This resolution is different both between the full taxon phylogeny and the nonhybrid diploid phylogeny and between the different phylogenetic analyses on the full taxon phylogeny. MP and Bayesian analyses on the full taxon phylogeny placed Helianthus bolanderi sister to H. annuus, one H. argophyllus clone, and all the H. paradoxus sequences, with the three other H. argophyllus individuals sister to the previous clade (consistent with topology in Fig. 3). The ML resolution on the full taxon set (Fig. 4) placed H. bolanderi sister to the H. argophyllus individuals, and the two H. paradoxus clones formed a grade at the base of this lineage. Despite this difference, the backbone topology for the genus was the same with and without the hybrids included. There were at least three and possibly four main lineages in Helianthus, labeled as groups 1–3 in Figs. 2 and 3. The labeling of these assemblages is purely for referencing within this article, not as a surrogate for a formal taxonomy of the genus. Group 3 was highly supported as a natural group, group 2 was also monophyletic but weakly supported, and group 1 consisted of two unresolved clades, each of which was well supported.
|
|
The current infrageneric taxonomic classification in Helianthus (Schilling and Heiser, 1981
) is mapped onto the nonhybrid diploid phylogeny (Fig. 3). Aside from the monotypic Helianthus sect. Agrestis, only one of the current infrageneric sections is supported as a natural group: H. sect. Helianthus (the annuals). The other two sections, H. sect. Ciliares and Divaricati, are polyphyletic, according to the ETS phylogeny. However, there is support for the monophyly of series within H. sect. Ciliares: H. ser. Pumilii is weakly monophyletic and H. ser. Ciliares has very strong support as a natural group. The large, perennial H. sect. Divaricati and its series appear to be more of a "catch-all" group according to the ETS phylogeny, with neither the section nor its series appearing as natural groups.
In the phylogenetic analysis of the full data set, including the hybrids, the hybrid species are scattered across groups 2 and 3 (Fig. 4). None are present in group 1. It is clear from this overall pattern of hybrid placement within the genus that hybrid speciation and/or polyploidy has independently arisen in the sunflowers multiple times. Five polyploid species form a weakly supported clade in group 2. They are southeastern perennials, with the exception of H. californicus in the West USA. The diploid hybrid annuals cluster within the annual clade in group 2, except for H. anomalus, which is unresolved with respect to the annual clade. Closer inspection of variable sites between the H. anomalus lineage and the H. petiolaris vs. H. annuus lineages revealed 12 characters that would unite H. anomalus with the H. annuus lineage and 15 that would unite H. anomalus with the H. petiolaris lineage. These characters are scattered across the ETS region and do not cluster in any large recombination blocks (Table 3). Another perennial polyploid falls in the Group 3 H. series Ciliares clade; the remaining five polyploids are scattered across two clades in Group 3, (labeled 3a and 3b in Fig. 4).
|
) found a putative recombination event in Helianthus paradoxus 448.1, which is the basal lineage in the H. annuus clade. The suggested parent is H. paradoxus 448.4 in the H. petiolaris clade. SiScan, one of the six recombination programs in RDP2, found a significant recombination event between these two sequences within the aligned positions 158–299 (P = 0.01737). This event also was weakly supported by Geneconv, Bootscan, and LARD. This possible recombination event is further supported by phylogenetic analysis. When the recombinant and nonrecombinant regions were broken apart and analyzed separately, the recombinant (bp 158–299) fell out sister to H. paradoxus 448.4, and the remaining sequence was sister to H. paradoxus 448.10, suggesting it as the other parent. These two "parents" are actually different ETS types within the same H. paradoxus individual (#448). No recombination was detected in either H. anomalus individual, which also are in an unresolved basal location.
Character evolution
First, we tested the sensitivity of our data to detecting a correlation given a hypothetical 1 : 1 relationship between the presence of lower-leaf glands and a specific chemistry. This test was highly significant (p = 0.0007), indicating our data are sensitive to detecting a correlation between these characters and validating further tests.
Only four STLs of the 42 flavonoid and STL characters collected were reported for all species in Helianthus, so we concentrated on these four characters for the correlated character analyses. The presence/absence of lower-leaf glands was used as the independent character in the tests because this character was complete and variable across the species. The results of our correlated changes analyses of STLs using Discrete did not show any significant likelihood improvement under a model of character dependence, so we cannot reject the null hypothesis that any of these characters have evolved independently of each other or of abaxial leaf trichomes on our tree (Table 3). Figure 5 shows the pruned ETS bayesian phylogeny mapped with our characters of interest.
|
; Rieseberg et al., 1987
), we included these data on Fig. 5. The flavonoids were extracted only from the leaves of plants, and here the pattern is evident for the species for which data are available: flavone aglycones are only found in leaves with lower-leaf glands. Each of the five secondary chemistry profiles has a different pattern of evolution through the phylogeny (Fig. 5). While there are unique combinations of secondary chemistry that, together, form synapomorphies for some clades, no individual character is synapomorphic for clades detected using ETS sequences. For example, the eupaserrin-type STL is absent in all the annuals but also was independently lost in H. maximilliani and H. microcephalus.
DISCUSSION
Phylogenetic relationships of nonhybrid diploids
The ETS phylogeny has significantly higher resolution and better support than any previous phylogenetic study in Helianthus, especially the perennials (Rieseberg et al., 1991
; Gentzbittel et al., 1992
; Schilling, 1997
, 2001
; Schilling et al., 1998
). Some of the features indicated in these earlier studies are now well supported, and former polytomies in the genus are beginning to yield well-supported resolutions. Nonetheless, the order of divergence of the most basal clades continues to be poorly supported for the most part, suggesting either a rapid, early radiation of the genus or an inability of ETS to resolve the deeper nodes in this phylogeny.
Group 1
The two lineages contained in this group either form a clade or they represent a grade up to the two more speciose lineages in the genus, which also was suggested in a previous analysis (Schilling et al., 1998
). Each of the five species in this group has a very distinct morphology, both within the group and relative to the rest of the species in Helianthus. In addition, each species has a restricted distribution in Florida except for H. porteri, which occurs further north in the southeastern USA. Given the morphological variety in group 1, it is not surprising that it currently is represented by a diverse taxonomy: H. sect. Agrestis and two series of H. sect. Divaricati. Helianthus agrestis, an annual, and H. carnosus are both glabrous species that grow in boggy wetlands in Florida. The other lineage contains H. porteri, H. heterophyllus, and H. radula. Helianthus porteri is an annual, and the last two are sympatric perennial species of pine savannas in northern Florida.
Group 2
This weakly supported clade comprises three morphologically distinct lineages. The first well-supported lineage, which contains Helianthus angustifolius, H. simulans, and H. floridanus, makes up part of sect. Divericati series Angustifolii. These three species are all tall perennial herbs with alternate leaves and are distributed in the southeastern USA. This close relationship was previously recognized (Schilling and Heiser, 1981
; Schilling and Mabry, 1983
; Spring and Schilling, 1991
), but our data reveal that H. simulans has two ETS types. One is sister to H. angustifolius and the other to H. floridanus. There are two possible explanations for this: either H. simulans is a previously unrecognized homoploid hybrid offspring between H. angustifolius and H. floridanus (Fig. 6), or H. simulans is sister to one of the putative parents but has some horizontal gene flow with the other. Morphologically (Fig. 3), H. simulans shares characters of both putative parents: obovate basal leaves, cauline upper leaves with a prominent mid-vein, axillary fascicles, and mostly purple disk corollas with H. angustifolius and broader lanceolate leaves and height exceeding 2 m with H. floridanus. Crosses are successful between H. floridanus and H. angustifolius, but only H. angustifolius has been successfully crossed with H. simulans (Heiser et al., 1969
). Further genetic work is required to determine whether H. simulans should be recognized as the first homoploid hybrid perennial species in the genus.
|
, which hinted at a close relationship between H. pumilis and the annual sunflowers. In addition, Schilling's (1997) chloroplast RFLP data showed this western triplet forming a grade up to the annual clade, although the relationship was weakly supported.
Substantial previous evidence supported a monophyletic annual clade (the third morphologically distinct lineage in group 2) comprising the entire Helianthus sect. Helianthus (Schilling and Heiser, 1981
; Rieseberg, 1991
; Schilling, 1997
; Schilling et al., 1998
). Our ETS data strongly support this consensus. In addition, our analyses subdivide the annual clade into two well-supported lineages (Fig. 3), and each of these lineages contains one of the parental species (H. annuus and H. petiolaris) of Rieseberg's diploid hybrid species studies (Rieseberg, 1990
, 1991
; Rieseberg et al., 1991
). While there is no support for the species distinction between H. praecox and H. debilis, there is support for a geographical split between the H. debilis subspecies: H. debilis subsp. silvestris and H. debilis subsp. cucumerifolius from Texas are strongly supported as distinct from the Floridian H. debilis subsp. tardiflorus and H. debilis subsp. vestitus. Within the H. annuus lineage, H. argophyllus has two ETS types. One type forms a lineage sister to the H. annuus + H. bolanderi clade, and the other type is sister to H. annuus. Natural hybridization with H. annuus in narrow areas of overlap (Heiser, 1951
) provides opportunity for introgression between these two species and is consistent with an H. annuus ETS type being present in one of our H. argophyllus individuals.
Group 3
Sister to group 2, group 3 is a highly supported clade that includes a wide diversity of perennial sunflowers primarily in sect. Divaricati. The split in this group produces one large and one small lineage. The small lineage contains the Ciliares series of section Ciliares: the two diploid perennials, H. laciniatus and H. arizonensis, which occur in Mexico and Arizona, respectively. One of the H. arizonensis individuals in our study had two ETS types: one that groups with the other H. arizonensis individual and one that groups weakly with H. laciniatus. These two species are morphologically similar and share a narrow zone of overlap in Arizona, providing an opportunity for gene flow, which might account for the incongruent placement of the types. The large group 3 lineage contains two clades: one with mostly basal leaves and one with cauline leaves. The basal-leaved sunflowers, although weakly supported, contain a group of species that all (with the exception of H. mollis) share a basal rosette of leaves, are less than 1 m tall, and occur across the central and southeastern regions of the USA. Previous classifications have not included H. mollis with these other taxa because the species has cauline leaves instead of a basal rosette. Nonetheless, natural hybrids between H. mollis and most taxa in this clade (Jackson and Guard, 1958
; Beatley, 1963
) suggest a close relationship. Helianthus longifolius was placed into a different series based on other characters such as the degree of phyllary appression, but H. atrorubens, H. silphoides, and H. occidentalis appear to support the monophyly of H. series Atrorubentes (Fig. 3). Note that the two individuals of H. atrorubens in our study do not form a single clade. One individual is closely related to H. silphoides, a relationship previously suggested (Heiser et al., 1969
). The more basal position of the second H. atrorubens individual could be due to hybridization, which has been suggested in earlier work (Smith and Martin, 1959
).
Sister to the previous clade is a large, moderately supported lineage of tall, widespread perennials. Most of these species have overlapping ranges and blurred morphological boundaries (Long, 1955b
; Heiser et al., 1969
) This is reflected in our data by the multiple ETS types for several species, which, taken as a whole, span the breadth of this lineage. All species in this clade are classified as part of Helianthus sect. Divaricati series Microcephali or H. sect. Divaricati ser. Corona-solis. Introgression has been characterized across hybrid zones in these three lineages (Smith and Guard, 1958
; Long, 1966
), which could explain the wide diversity of ETS types within some of the individuals. Figure 6 illustrates a scenario of introgression that could have produced the pattern we observed in the ETS data. The ETS gene tree shows three well-supported and genetically distinct lineages despite the relationships within and between these lineages being unresolved (Fig. 3). One lineage contains the monophyletic species, H. salicifolius, which is also morphologically distinct. A second lineage reflects the morphological similarity of H. series Microcephali species with the close relationship of H. glaucophyllus and H. microcephalus, although there is one ETS type (37.1) that falls outside of this lineage. Conversely, two other ETS clones from H. divaricatus (69.4) and H. giganteus (AF123555) occur in this lineage. The well-supported third lineage in clade 3 contains seven species, most of which are in H. ser. Corona-solis. Helianthus maximilliani, the only monophyletic species in this clade, is unresolved with respect to the H. divaricatus individuals and the other H. microcephalus clone. Helianthus giganteus, H. grosseserratus, and H. nuttallii, previously recognized as forming a cohesive group (Heiser et al., 1969
), form a nicely supported clade here. Broader sampling of these species across their ranges will be needed to fully understand this complex clade of tall perennials.
Hybrid species
The nuclear ribosomal repeat region is known to undergo concerted evolution (Hillis et al., 1991
), so we might expect the more ancient hybridization events in Helianthus to lose the signature from one parent. However, at least one study has found deeper coalescent times for the ETS (B. Vanden Heuvel, Colorado State University, Pueblo, CO, personal communication). Generally, more recent hybrid speciation or introgression events are more likely to reveal sequences from both parents. Figure 4 shows the placement of the homoploid and polyploid hybrid sunflowers within the context of the nonhybrid diploid phylogeny. We emphasize that Fig. 4 is the ETS gene tree for the annual hybrids and perennial polyploids. The actual evolutionary history of the hybrid species would be better represented as a network or reticulate phylogeny as in Fig. 6. The network drawn in Fig. 6 represents one hypothesis for the reticulate evolution leading to the present pattern of ETS types across this complex group. Both morphology and introgression data from previously mentioned literature were used to produce this scenario.
Although not definitive, the ETS phylogeny suggests some scenarios for the evolution of hybrid speciation within the sunflowers and indicates the presence of a previously unrecognized hybrid. The evidence here clearly suggests that hybrid speciation occurred more than once in the genus. Although group 1 has no evidence of hybrid speciation, most major clades in groups 2 and 3 of Helianthus do have evidence of hybrid speciation. Also, within the major lineages of groups 2 and 3, none of the hybrid species contains ETS types from more than one major lineage, suggesting that the major lineages of the genus may have differentiated enough to prevent successful hybrid speciation at deeper levels. However, because we do not have evidence for both parents for all of the hybrids, this observation could change with future data collection. Hybrid content of groups 2 and 3 are discussed next.
Group 2
A large set of polyploids in group 2 is sister to the diploid clade containing Helianthus angustifolius, H. simulans, and H. floridanus (Fig. 4). The grouping of four of the five polyploid species in this clade makes sense based on their distribution in the southeastern USA and their shared morphological features, but the inclusion of H. californicus in this group is an unpredicted finding. Helianthus californicus is morphologically more similar to H. nuttallii, H. giganteus, and H. grosseserratus. In addition, its distribution in the western USA is far outside the present-day range of other group 2 polyploids. It is possible that its range was more widespread in the past. Because H. californicus is a hexaploid, it carries another diploid genome, which may contribute to its morphological divergence from the rest of the clade (assuming it inherited a tetraploid genome from the southeastern polyploids).
Although it is not possible to understand the complex history of this group of polyploids with just one marker, there are two plausible scenarios for this topology: (1) a single ancient autopolyploidy event at the base of the polyploid clade followed by nonhybrid speciation or (2) one or more ancient allopolyploidy events with the ancestor of H. angustifolius + H. simulans + H. floridanus plus one or more other unknown parents. If the second scenario is correct, then concerted evolution has erased the signature of the second parent(s) (and possibly a third parent in the case of the hexaploids), or we did not sample sufficiently to find an ETS sequence from other parents. After the ancient polyploidization event(s), subsequent speciation and additional polyploidy to produce the hexaploids would provide the pattern we see with the ETS data. Most of the relationships within this polyploid clade are not well resolved, but the suggested grouping of these species has interesting implications. Heiser et al. (1969)
proposed that the hexaploid H. eggertii contained genomes from the tetraploid H. laevigatus (4x) and a diploid genome from H. decapetalus. The ETS data support a loose association with H. laevigatus, but no other ETS types were recovered in this species. Helianthus schweinitzii, H. laevigatus, and H. hirsutus all have suggested parents from group 3 (Heiser et al., 1969
) but our analysis did not recover any ETS types outside group 2. It is striking that most of the former hypotheses regarding the origins of these polyploid taxa involve multiple independent polyploidizations. More independent genetic markers are needed to determine whether any species in group 3 have contributed genomes to this polyploid group, as hypothesized by Heiser and other previously mentioned researchers.
The other major lineage within group 2 contains the three annual diploid hybrids previously shown to be independent offspring of Helianthus annuus and H. petiolaris (Rieseberg, 1990
, 1991
; Rieseberg et al., 1991
). Each homoploid hybrid species has a different pattern of ETS types. Helianthus deserticola, represented by two individuals, contains only one ETS type, which is sister to H. petiolaris. Helianthus paradoxus contains five different ETS types within our two individuals. Four of these types are present in the clade containing H. annuus, two form a basal grade, and two group with H. annuus and H. argophyllus. The fifth ETS type in H. paradoxus is sister to H. petiolaris + H. deserticola. Although our unresolved basal H. paradoxus ETS types indicate apparent nonmonophyly of H. paradoxus, once sequence recombination is accounted for, the observed pattern is concordant with Welch and Rieseberg's (2002)
work supporting a single origin of H. paradoxus. Our RDP2 results show clear evidence of recombination in the basal-most H. paradoxus ETS clone (448.1), and while this mosaic clone also could be a PCR-recombinant, it still requires the presence of two divergent ETS types in a single individual for the recombinant to form.
The third hybrid, Helianthus anomalus, represented here by two individuals, has an ETS type that is unresolved in the annuals because it does not group with either of its parental species (Fig. 4). In prior studies, southern and northern populations of H. anomalus clustered with H. petiolaris and H. annuus, respectively (Rieseberg, 1991
; Schilling et al., 1998
). Our sampling only included two H. anomalus individuals from the northern population. Although no recombination was detected in our H. anomalus ETSs using the six different programs implemented in RDP2, this is likely explained by the shortness of the probable recombined segments in our sequences (Table 3). Instead of lengthy recombination blocks as seen in the previous H. paradoxus example, the sites from each parent are fairly mixed up, as would be expected after extensive recombination. This pattern suggests there has been so much recombination in the ETS region between H. annuus and H. petiolaris types that detecting recombination on such a fine scale is impossible using current techniques. While it is also possible H. anomalus is not the hybrid offspring of H. annuus and H. petiolaris, the extensive data supporting the hybrid origin of this species (Rieseberg, 1991
, 2000
; Schwarzbach and Rieseberg, 2002
; Ludwig et al., 2004
) make this an unlikely explanation.
Group 3
Two of the three main lineages discussed previously in this group contain polyploids (Fig. 4). For the lineage containing the Helianthus sect. Ciliares series Ciliares species from the southwestern USA, earlier support exists for the hybrid origin of H. ciliaris from H. arizonensis and H. laciniatus (Jackson and Hauber, 1994
). Our analysis provides further support because one H. ciliaris individual clusters with H. laciniatus while the other clusters with H. arizonensis. Helianthus ciliaris shares red florets with H. laciniatus and glaucus leaves with H. arizonensis.
The most complex network patterns in the genus occur between clades 3a and 3b (Fig. 4). Three polyploids have ETS types that occur in both 3a and 3b: Helianthus resinosus, H. tuberosus, and H. pauciflorus. Because of low support for the relationships within each clade, it is unclear from the ETS data whether these species were derived independently or from a single polyploidization event. Helianthus tuberosus has been the focus of many studies because it is an economically important crop plant (the Jerusalem artichoke or sun choke). In early studies (Kostoff, 1939
; Heiser and Smith, 1964
), an auto-alloploid genome was hypothesized for this species, suggesting two copies of an identical genome plus one copy of a different genome (H. annuus). Our study did not find a H. annuus ETS type, although additional data might recover it.
The ETS evidence of parentage for the four other polyploid species (Helianthus ‘Newhall Ranch', H. strumosus, H. decapetalus, and H. verticillatus) points to origins within only the 3a or 3b clade boundaries. Helianthus ‘Newhall Ranch' is a recently discovered tetraploid population thought to be related to H. nuttallii and H. californicus (Porter and Fraga, 2006
). We recovered one ETS type from this plant that was sister to H. nuttallii. This is consistent with an autopolyploidization event (Fig. 6) forming the new species, although additional independent markers are needed to confirm this. Helianthus strumosus and H. decapetalus both exist in different ploidy levels (Fig. 4), and each contains ETS types that are largely unresolved within group 3b. Helianthus verticillatus is a recently rediscovered rare diploid species in Helianthus (Matthews et al., 2002
) that is believed to be of hybrid origin (Beatley, 1963
; Heiser et al., 1969
). During our study, Ellis et al. (2006)
concluded that this species did not show any evidence of being a hybrid. We found two ETS types for H. verticillatus in clade 3b: one is unresolved and the other is moderately supported as sister to H. tuberosus (Fig. 4). Because there is so little resolution within this clade, we cannot definitively comment on the hybrid status of this species based on our ETS data.
Secondary chemistry
If our ETS phylogeny approximates the evolutionary history of the sunflowers, the secondary compounds we studied appear to have evolved independently of one another and, in some cases, multiple times, possibly by regulatory switches. Our analyses suggest that factors other than phylogenetic constraints, perhaps abiotic factors or levels of herbivory, are more likely to have influenced the secondary chemistry patterns seen today in the sunflowers. Testing specific evolutionary hypotheses for these patterns will require more organ-specific chemistry profiles and a better understanding of the regulatory pathways controlling the presence/absence of these characters.
Conclusions
Our broader research goal is to understand the complete pattern of reticulate evolution in Helianthus. By providing a well-resolved gene tree along with several interesting patterns suggesting hybrid parentage and deeper reticulation events within the genus, this study is a crucial step in the right direction. With a well-resolved single-gene tree, this study provides conclusive evidence for multiple, independent hybrid speciation and/or polyploidy events. But to understand the complete pattern of reticulate evolution in Helianthus, which will involve tracing all parent/offspring relationships and incidences of introgressive hybridization, researchers will need to study many independent gene phylogenies.
APPENDIX
Taxon—Individual #, Voucher specimen, HERBARIUM, Clone number (cl. #): GenBank accession.
Helianthus agrestis Pollard—#20___, ___, cl. 12: DQ486530; #90, RET 90, TEX, cl. 45: DQ486531. H. angustifolius L.—#50, RET 50, TEX, cl. 63: DQ486532. H. annuus L.— #__, CRL 2AA, TEX, cl. 1: DQ486533. H. anomalus Blake—#1244, LR 1244, IU, cl. 5: DQ486534; #1260, LR 1260, IU, cl. 7: DQ486535. H. argophyllus T. & G.—#__, CRL 7, TEX, cl. 14: DQ486536, cl. 16: DQ486537; #33, RET 33, TEX, cl. 3: DQ486538, cl. 9: DQ486539. H. arizonensis R. C. Jackson—#383, ES OS-280–4*, TENN, cl. 14: DQ486540, cl. 30: DQ486541; #80, RET 80, TEX, cl. 9: DQ486542. H. atrorubens L.—#149, ES 149*, TENN, cl. 9: DQ486543; #40, RET 40, TEX, cl. 2: DQ486544. H. bolanderi Gray—#60, RET 60, TEX, cl. 7: DQ486545. H. californicus DC—#65, RET 65, TEX, cl. 23: DQ486546. H. carnosus Small—#48, ES 48*, TENN, cl. 17: DQ486547; #89, Gary E. Schultz s.n., FLAS, cl. 38: DQ486548. H. ciliaris DC—#280, R. Jackson 121, IU, cl. 11: DQ486549; #81, RET 80, TEX, cl. 1: DQ486550. H. cusickii A. Gray—#61, RET 61, TEX, cl. 6: DQ486551. H. debilis Nutt. subsp. cucumerifolius (T. & G.) Heiser—#99, RET 99, TEX, cl. 5: DQ486552, cl. 7: DQ486553; subsp. silvestris Heiser—#5, CRL 5, TEX, cl. 13: DQ486554; subsp. tardiflorus Heiser—#__, G. Seiler 1564*, RSA, cl. 19: DQ486555; subsp. vestitus (E. Watson) Heiser—#93, RET 93, TEX, cl. 7: DQ486556. H. decapetalus L.—#240, SS 117*, TENN, cl. 17: DQ486557, cl. 20: DQ486558; #44, RET 44, TEX, cl. 51: DQ486559, cl. 64: DQ486560. H. deserticola Heiser—#__, G. Seiler 1476, RSA; cl. 3: DQ486561; #4255, Matt Lavin 4255, TEX, cl. 7: DQ486562. H. divaricatus L.—#253, SS 227, TENN, cl. 29: DQ486563; #69, RET 69, TEX, cl. 4: DQ486564, cl. 5: DQ486565; #71, RET 71, TEX, cl. 5: DQ486566. H. eggertii Small—#210, SS 90–16, TENN, cl. 14: DQ486567, cl. 15: DQ486568; #57, RET 57, TEX, cl. 10: DQ486569, cl. 6: DQ486570. H. floridanus Gray ex Chapm.—#88, RET 88, TEX, cl. 54: DQ486571. H. giganteus L.—#204, SS 171, TEX, cl. 13: DQ486572, cl. 18: DQ486573, cl. 2: DQ486574, cl. 3: DQ486575. H. glaucophyllus Smith—#138, ES 138*, TENN, cl. 13: DQ486576; #42, RET 42, TEX, cl. 10: DQ486577. H. gracilentus A. Gray—#59, RET 59, TEX, cl. 3: DQ486578. H. grosseserratus M. Martens—#101, RET 101, TEX, cl. 11: DQ486579, cl. 4: DQ486580; #98, RET 98, TEX, cl. 26: DQ486581. H. heterophyllus Nutt.—#35, CRL 35, TEX, cl. 21: DQ486582. H. hirsutus Raf.—#__, CRL 11, TEX, cl. 17: DQ486583; #207, SS 90–17*, TENN, cl. 10: DQ486584. H. laciniatus A. Gray—#59, ES 59*, TENN, cl. 4: DQ486585; #75, RET 75, TEX, cl. 24: DQ486586. H. laevigatus T. & G.—#235, ES 124*, TEX, cl. 1: DQ486587, cl. 5: DQ486588. H. longifolius Pursh—#254, SS 227*, TENN, cl. 21: DQ486589. H. maximiliani Schrad.—#2219, PI 586900, GRIN, cl. 3: DQ486590; #78, RET 78, TEX, cl. 4: DQ486591. H. microcephalus T. & G.—#18, J. Gershenzon and R. Pfeil 108, TEX, cl. 9: DQ486592; #37, RET 37, TEX, cl. 1: DQ486593, cl. 6: DQ486594. H. mollis Lamb.—#38, AMES 22796, GRIN; cl. 6: DQ486595. H. nuttallii T. & G. subsp. nuttallii T. & G.—#46, PI 420182, GRIN, cl. 15: DQ486598, cl. 6: DQ486599; subsp. parishii (Gray) Heiser—#__, Mary Meyers s.n., RSA; cl. 18: DQ486597. H. niveus (Benth.) Brandegee subsp. tephrodes (Gray) Heiser—D. Gomez 71, TEX, cl. 2: DQ486596. H. occidentalis Riddell subsp. occidentalis Riddell—#213, ES 90–23*, TENN, cl. 9: DQ486601; subsp. plantagineus (T. & G.) Heiser—#100, RET 100, TEX, cl. 8: DQ486600. H. paradoxus Heiser—#442, LR 1300, IU, cl. 3: DQ486602, cl. 8: DQ486603; #448, LR 1302, IU, cl. 1: DQ486604, cl. 10: DQ486605, cl. 4: DQ486606. H. pauciflorus Nutt.—#54, GRIN, AMES 17981, cl. 1: DQ486607, cl. 10: DQ486608, cl. 7: DQ486609. H. petiolaris Nutt. subsp. petiolaris Nutt.— #__, B. B. Simpson 22-V-00–2, TEX, cl. 11: DQ486610; #63, RET 63, TEX, cl. 1: DQ486611. H. porteri (A. Gray) Heiser—#__, ES 90–31*, TENN, cl. 1: DQ486612. H. praecox Engleman and Gray subsp. runyonii Heiser—#__, H. Nikols 2, TEX, cl. 6: DQ486613. H. pumilus Nutt.—#64, RET 64, TEX, cl. 4: DQ486614. H. radula (Pursh) T. & G.—#257, SS 192*, TENN, cl. 4: DQ486615, #83, RET 83, TEX, cl. 8: DQ486616. H. resinosus Small—#30, CRL 30, cl. 21: DQ486617, cl. 5: DQ486618. H. salicifolius Deitr.— #__,__,__, cl. 18: DQ486619; #97, RET 97, TEX, cl. 5: DQ486620. H. schweinitzii T. & G.—#264, SS 186c*, TENN, cl. 4: DQ486621, cl. 8: DQ486622. H. silphioides Nutt.—#122, ES 122*, TENN, cl. 16: DQ486623. H. simulans E. Watson—#45, ES 54*, TENN, cl. 3: DQ486624, cl. 8: DQ486625. H. strumosus L.—#41, RET 41, TEX, cl. 58: DQ486626, cl. 60: DQ486627. H. tuberosus L.—#2, CRL 16, TEX, cl. 23: DQ486628, cl. 4: DQ486629, cl. 6: DQ486630. H. verticillatus Small—#97, G. Seiler 2391, IU, cl. 1: DQ486631, cl. 9: DQ486632.
Phoebanthus grandiflorus (T. & G.) Blake—#__, Steve L. Orzell 14500, TEX, cl. 8: DQ486633.
|
1 Our broader research goal is to understand the complete pattern of reticulate evolution in Helianthus. By providing a well-resolved gene tree along with several interesting patterns suggesting hybrid parentage and deeper reticulation events within the genus, this study is a crucial step in the right direction. With a well-resolved single-gene tree, this study provides conclusive evidence for multiple, independent hybrid speciation and/or polyploidy events. But to understand the complete pattern of reticulate evolution in Helianthus, which will involve tracing all parent/offspring relationships and incidences of introgressive hybridization, researchers will need to study many independent gene phylogenies. 
2 Author for correspondence (e-mail: retimme{at}umd.edu
)
LITERATURE CITED
Alfatafta A. A. Mullin C. A.. 1992. Epicuticular terpenoids and an aurone from flowers of Helianthus annuus. Phytochemistry 31: 4109-4113..[CrossRef][Web of Science]
Atlagic J. Skoric D.. 1999. Cytogenetic study of Helianthus laevigatus and its F-1 and BC1F1 hybrids with cultivated sunflower, Helianthus annuus. Plant Breeding 118: 555-559..[CrossRef][Web of Science]
Beatley J. C.. 1963. The sunflowers (genus Helianthus) in Tennessee. Journal of the Tennessee Academy of Science 38: 135-154..
de Candolle A. P.. 1836. Helianthus. In A. P. de Candolle [ed.], Prodromus systematis naturalis regni vegetabilis, vol. II, pt. V, 585-591. Treuttel and Würtz, Paris, France..
Clevenger S.. 1963. Helianthus laetiflorus and Helianthus rigidus, hybrids or species?. Rhodora 65: 121-133..
Cockerell T. D. A.. 1919. Hybrid perennial sunflowers. Botanical Gazette 67: 264-266..
Dewer D.. 1893. Perennial sunflowers. Journal of the Royal Horticultural Society 15: 26-39..
Ellis J. R. Pashley C. H. Burke J. McCauley D. E.. 2006. High diversity in a rare and endangered sunflower as compared to a common congener. Molecular Ecology 15: 2345-2355..[CrossRef][Medline]
Gao F. Wang H. P. Mabry T. J.. 1987. Sesquiterpene lactones and flavonoids from Helianthus species. Journal of Natural Products 50: 23-29..[CrossRef][Medline]
Gentzbittel L. Perrault A. Nicolas P.. 1992. Molecular phylogeny of the Helianthus genus, based on nuclear restriction-fragment length polymorphism (RFLP). Molecular Biology and Evolution 9: 872-892..[Web of Science]
Gray A.. 1884. Synoptical flora of North America, vol. I, pt. II. Smithsonian Institution, Washington, D.C., USA..
Heiser C. B. Jr.. 1951. Hybridization in the annual sunflowers: Helianthus annuus x H. argophyllus. American Naturalist 85: 64-72..
Heiser C. B. Jr. 1965. Species crosses in Helianthus: III. Delimitation of "sections.". Annals of the Missouri Botanical Garden 52: 364-370..[CrossRef]
Heiser C. B. Jr. Smith D. M.. 1960. The origin of Helianthus multiflorus. American Journal of Botany 47: 860-865..[CrossRef][Web of Science]
Heiser C. B. Jr Smith D. M.. 1964. Species crosses in Helianthus: II. Polyploid species. Rhodora 66: 344-358..
Heiser C. B. Jr. Smith D. M. Clevenger S. B. Martin W. C.. 1969. The North American sunflowers (Helianthus). Memoirs of the Torrey Botanical Club 22: 1-218..
Herz W. Kulanthaivel P.. 1984. Ent-pimaranes, ent-kauranes, heliangolides and other constituents of 3 Helianthus species. Phytochemistry 23: 1453-1459..[CrossRef][Web of Science]
Herz W. Kumar N.. 1981a. Minor sesquiterpene lactones of Helianthus pumilis. Phytochemistry 20: 1339-1341..[CrossRef][Web of Science]
Herz W. Kumar N.. 1981b. Sesquiterpene lactones from Helianthus grosseserratus. Phytochemistry 20: 99-104..[CrossRef][Web of Science]
Hillis D. M. Moritz C. Porter C. A. Baker R. J.. 1991. Evidence for biased gene conversion in concerted evolution of ribosomal DNA. Science 251: 308-310..
Jackson R. C.. 1956. The hybrid origin of Helianthus doronicoides. Rhodora 58: 97-100..
Jackson R. C. Guard A. T.. 1956. Hybridization of perennial sunflowers in Indiana. Proceedings of the Indiana Academy of Science 65: 212-217..
Jackson R. C. Guard A. T.. 1957. Analysis of some natural and artificial interspecific hybrids in Helianthus. Proceedings of the Indiana Academy of Science 66: 306-317..
Jackson R. C. Guard A. T.. 1958. Natural and artificial hybridization between Helianthus mollis and H. occidentalis. American Midland Naturalist 58: 422-433..[Medline]
Jackson R. C. Hauber D. P.. 1994. Quantitative cytogenetic analyses of autoploid and alloploid taxa in the Helianthus ciliaris group (Compositae). American Journal of Botany 81: 1063-1069..[CrossRef][Web of Science]
Koch M. A. Dobes C. Matschinger M. Bleeker W. Vogel J. Kiefer M. Mitchell-Olds T.. 2005. Evolution of the trnF(GAA) gene in Arabidopsis relatives and the Brassicaceae family: monophyletic origin and subsequent diversification of a plastidic pseudogene. Molecular Biology and Evolution 22: 1032-1043..
Kostoff D.. 1939. Autosyndesis and structural hybridity in F1-hybrid Helianthus tuberosus L. x Helianthus annuus L. and their sequences. Genetica 21: 285-300..[CrossRef]
Linder C. R. Goertzen L. R. Heuvel B. V. Francisco-Ortega J. Jansen R.. 2000. The complete external transcribed spacer of 18S-26S rDNA: amplification and phylogenetic utility at low taxonomic levels in Asteraceae and closely allied families. Molecular Phylogenetics and Evolution 14: 285-303..[CrossRef][Web of Science][Medline]
Long R. W.. 1955a. Hybridization between the perennial sunflowers Helianthus salicifolius A. Dietr. and H. grosseserratus Martens. American Midland Naturalist 54: 61-64..[CrossRef]
Long R. W.. 1955b. Hybridization in perennial sunflowers. American Journal of Botany 42: 769-777..[CrossRef][Web of Science]
Long R. W.. 1957. A cytological analysis of a naturally occurring Helianthus hybrid. Ohio Journal of Science 57: 65-69..
Long R. W.. 1959. Natural and artificial hybrids of Helianthus maximilliani x H. grosseserratus. American Journal of Botany 46: 687-692..[CrossRef][Web of Science]
Long R. W.. 1960. Biosystematics of two perennial species of Helianthus (Compositae). I. Crossing relationships and transplant studies. American Journal of Botany 47: 729-735..[CrossRef][Web of Science]
Long R. W.. 1961. Biosystematics of two perennial species of Helianthus (Compositae), II. Natural populations and taxonomy. Brittonia 13: 129-141..[CrossRef]
Long R. W.. 1963. Cytogenetic investigations of artificial Helianthus giganteus x H. salicifolius hybrids (Compositae). Ohio Journal of Science 63: 273-281..
Long R. W.. 1966. Biosystematics of the Helianthus nuttallii complex (Compositae). Brittonia 18: 64-79..[CrossRef][Web of Science]
Ludwig F. Rosenthal D. M. Johnston J. A. Kane N. Gross B. L. Lexer C. Dudley S. A. Rieseberg L. H. Donovan L. A.. 2004. Selection on leaf ecophysiological traits in a desert hybrid Helianthus species and early-generation hybrids. Evolution; International Journal Organic Evolution 58: 2682-2692..
Maddison D. R. Maddison W. P.. 2002. MacClade: analysis of phylogeny and character evolution. Sinauer, Sunderland, Massachusetts, USA..
Martin D. P. Williamson C. Posada D.. 2005. RDP2: recombination detection and analysis from sequence alignments. Bioinformatics 21: 260-262..
Matthews J. F. Allison J. R. Ware R. Nordman C.. 2002. Helianthus verticillatus Small (Asteraceae) rediscovered and redescribed. Castanea 67: 13-24..
Matthews J. F. Barden L. S. Matthews C. R.. 1997. Corrections of the chromosome number, distribution and misidentifications of the federally endangered sunflower, Helianthus schweinitzii T&G. Journal of the Torrey Botanical Society 124: 198-209..
Melek F. R. Ahmed A. A. Mabry T. J.. 1985. 6-Methoxylated flavones from Helianthus hirsutus and H. gracilentus. Revista Latinoamericana de Ingenieria Química y Química Aplicada 16: 27..
Nylander J. A. A.. 2004. MrModeltest. Computer program distributed by the author, Florida State University, Tallahassee, FL, USA. Website http://people.scs.fsu.edu/nylander/..
Ohno N. Gershenzon J. Neuman P. Mabry T. J.. 1981. Diterpene carboxylic-acids and a heliangolide from Helianthus angustifolius. Phytochemistry 20: 2393-2396..[CrossRef][Web of Science]
Pagel M.. 1994. Detecting correlated evolution on phylogenies: a general method for the comparative analysis of discrete characters. Proceedings of the Royal Society, B, Biological Sciences 255: 37-45..
Panero J. L. Funk V. A.. 2002. Toward a phylogenetic subfamilial classification for the Compositae (Asteraceae). Proceedings of the Biological Society of Washington 115: 909-922..[Web of Science]
Picman A. K.. 1986. Biological activities of sesquiterpene lactones. Biochemical Systematics and Ecology 14: 255-281..[CrossRef][Web of Science]
Porter J. M. Fraga N.. 2006. A quantitative analysis of pollen variation in two southern California perennial Helianthus (Heliantheae: Asteraceae). Crossosoma 31: 1-12..
Rieseberg L. H.. 1991. Homoploid reticulate evolution in Helianthus (Asteraceae): evidence from ribosomal genes. American Journal of Botany 78: 1218-1237..[CrossRef][Web of Science]
Rieseberg L. H.. 1995. The role of hybridization in evolution: old wine in new skins. American Journal of Botany 82: 944-953..[CrossRef][Web of Science]
Rieseberg L. H.. 1997. Hybrid origins of plant species. Annual Review of Ecology and Systematics 28: 359-389..[CrossRef][Web of Science]
Rieseberg L. H.. 2000. Crossing relationships among ancient and experimental sunflower hybrid lineages. Evolution 54: 859-865..[CrossRef][Web of Science][Medline]
Rieseberg L. H. Archer M. A. Wayne R. K.. 1999a. Transgressive segregation, adaptation and speciation. Heredity 83: 363-372..[CrossRef][Web of Science][Medline]
Rieseberg L. H. Beckstrom-Sternberg S. M. Liston A. Arias D.. 1991. Phylogenetic and systematic inferences from chloroplast DNA and isozyme variation in Helianthus sect. Helianthus (Asteraceae). Systematic Botany 16: 50-76..[CrossRef][Web of Science]
Rieseberg L. H. Carter R. Zona S. 1990. Molecular tests of the hypothesized hybrid origin of two diploid Helianthus species (Asteraceae). Evolution 44: 1498-1511..[CrossRef][Web of Science]
Rieseberg L. H. Raymond O. Rosenthal D. M. Lai Z. Livingstone K. Nakazato T. Durphy J. L. Schwarzbach A. E. Donovan L. A. Lexer C.. 2003. Major ecological transitions in wild sunflowers facilitated by hybridization. Science 301: 1211-1216..
Rieseberg L. H. Sinervo B. Linder C. R. Ungerer M. C. Arias D. M.. 1996. Role of gene interactions in hybrid speciation: evidence from ancient and experimental hybrids. Science 272: 741-745..[Abstract]
Rieseberg L. H. Soltis D. E. Arnold D.. 1987. Variation and localization of flavonoid aglycones in Helianthus annuus (Compositae). American Journal of Botany 74: 224-233..[CrossRef][Web of Science]
Rieseberg L. H. van Fossen C. Desrochers A.. 1995. Hybrid speciation accompanied by genomic reorganization in wild sunflowers. Nature 375: 313-316..[CrossRef][Web of Science]
Rieseberg L. H. Whitton J. Gardner K.. 1999b. Hybrid zones and the genetic architecture of a barrier to gene flow between two sunflower species. Genetics 152: 713-727..
Ronquist F. Huelsenbeck J. P.. 2003. MrBayes 3: Bayesian phylogenetic inference under mixed models. Bioinformatics 19: 1572-1574..
Schilling E. E.. 1997. Phylogenetic analysis of Helianthus (Asteraceae) based on chloroplast DNA restriction site data. Theoretical Applied Genetics 94: 925-933..[CrossRef]
Schilling E. E.. 2001. Phylogeny of Helianthus and related genera. Oleagineux, Corps Gras, Lipides 8: 22-25..
Schilling E. E. Heiser C. B.. 1981. Infrageneric classification of Helianthus (Asteraceae). Taxon 30: 393-403..[CrossRef][Web of Science]
Schilling E. E. Linder C. R. Noyes R. D. Rieseberg L. H.. 1998. Phylogenetic relationships in Helianthus (Asteraceae) based on nuclear ribosomal DNA internal transcribed spacer region sequence data. Systematic Botany 23: 177-187..[CrossRef][Web of Science]
Schilling E. E. Mabry T. J.. 1981. Flavonoids of Helianthus series Corona-solis. Biochemical Systematics and Ecology 9: 161-163..[Medline]
Schilling E. E. Mabry T. J.. 1983. Flavonoids of Helianthus series Angustifolii. Biochemical Systematics and Ecology 11: 341-344..[CrossRef][Web of Science]
Schilling E. E. Panero J. L.. 1996. Phylogenetic reticulation in subtribe Helianthinae. American Journal of Botany 83: 939-948..[CrossRef][Web of Science]
Schilling E. E. Panero J. L.. 2002. A revised classification of subtribe Helianthinae (Asteraceae: Heliantheae). I. Basal lineages. Botanical Journal of the Linnean Society 140: 65-76..[CrossRef][Web of Science]
Schilling E. E. Panero J. L. Storbeck T. A.. 1987. Flavonoids of Helianthus series Microcephali. Biochemical Systematics and Ecology 15: 671-672..[CrossRef][Web of Science]
Schwarzbach A. E. Rieseberg L. H.. 2002. Likely multiple origins of a diploid hybrid sunflower species. Molecular Ecology 11: 1703-1715..[CrossRef][Medline]
Smith D. M.. 1960. The chromosome number of Helianthus decapetalus. Transactions of the Kentucky Academy of Science 21: 17-19..
Smith D. M.. 1961. Variation in the tetraploid sunflowers Helianthus decapetalus, H. hirsutus, and H. strumosus. Recent Advances in Botany 1: 878-881..
Smith D. M. Guard A. T.. 1958. Hybridization between Helianthus divaricatus and H. microcephalus. Brittonia 10: 137-145..[CrossRef]
Smith D. M. Martin W. C.. 1959. Natural hybridization of Helianthus longifolius with H. atrorubens and H. occidentalis. Rhodora 61: 140-147..
Sossey-Alaoui K. Serieys H. Tersac M. Lambert P. Schilling E. Griveau Y. Kaan F. Bervillé A.. 1998. Evidence for several genomes in Helianthus. Theoretical and Applied Genetics 97: 422-430..[CrossRef][Web of Science]
Spring O.. 1989. Microsampling: an alternative approach using sesquiterpene lactones for systematics. Biochemical Systematics and Ecology 17: 509-517..[CrossRef][Web of Science]
Spring O. Schilling E. E.. 1989. Chemosystematic investigation of the annual species of Helianthus (Asteraceae). Biochemical Systematics and Ecology 17: 519-528..[CrossRef][Web of Science]
Spring O. Schilling E. E.. 1990. The sesquiterpene lactone chemistry of Helianthus sect. Ciliares (Asteraceae, Heliantheae). Biochemical Systematics and Ecology 18: 139-143..[CrossRef][Web of Science]
Spring O. Schilling E. E.. 1991. The sesquiterpene lactone chemistry of Helianthus sect. Atrorubentes (Asteraceae, Heliantheae). Biochemical Systematics and Ecology 19: 59-79..[CrossRef][Web of Science]
Swofford D. L.. 2003. PAUP*: phylogenetic analysis using parsimony (*and other methods). Sinauer, Sunderland, Massachusetts, USA..
Thompson J. D. Gibson T. J. Plewniak F. Jeanmougin F. Higgins D. G.. 1997. The CLUSTAL_X windows interface: flexible strategies for multiple sequence alignment aided by quality analysis tools. Nucleic Acids Research 24: 4876-4882..
Torrey J. Gray A.. 1842. A flora of North America, vol. II. Wiley and Putnam, New York, New York, USA..
Watanabe K. Ohno N. Yoshioka H. Gershenzon J. Mabry T. J.. 1982. Sesquiterpene lactones and diterpenoids from Helianthus argophyllus. Phytochemistry 21: 709-713..[CrossRef][Web of Science]
Watson E. E.. 1929. Contributions to a monograph of the genus Helianthus. Papers of the Michigan Academy of Science, Arts, and Letters 9: 305-475..
Welch M. E. Rieseberg L. H.. 2002. Patterns of genetic variation suggest a single, ancient origin for the diploid hybrid species Helianthus paradoxus. Evolution 56: 2126-2137..[CrossRef][Web of Science][Medline]
Zwickl D. J.. 2006. Genetic algorithm approaches for the phylogenetic analysis of large biological sequence data sets under the maximum likelihood criterion. Ph.D. dissertation, The University of Texas at Austin, Austin, Texas, USA..
CiteULike Complore Connotea Del.icio.us Digg Facebook Reddit Technorati Twitter What's this?
This article has been cited by other articles:
|
|
 |
|
  I. Mayrose, M. S. Barker, and S. P. Otto Probabilistic Models of Chromosome Number Evolution and the Inference of Polyploidy Syst Biol, January 6, 2010; (2010) syp083v2. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 N. Ishikawa, J. Yokoyama, and H. Tsukaya Molecular evidence of reticulate evolution in the subgenus Plantago (Plantaginaceae) Am. J. Botany, September 1, 2009; 96(9): 1627 - 1635. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. A. Church and E. J. Spaulding Gene Expression in a Wild Autopolyploid Sunflower Series J. Hered., July 1, 2009; 100(4): 491 - 495. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Dunbar-Co, A. M. Wieczorek, and C. W. Morden Molecular phylogeny and adaptive radiation of the endemic Hawaiian Plantago species (Plantaginaceae) Am. J. Botany, September 1, 2008; 95(9): 1177 - 1188. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
