| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Systematics |
Jepson Herbarium and Department of Integrative Biology, University of California, 1001 Valley Life Sciences Building # 2465, Berkeley, California 94720-2465 USA
Received for publication August 2, 2001. Accepted for publication January 29, 2002.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
Key Words: Compositae • cryptic diversity • ETS • Heliantheae • ITS • Lasthenia californica • trnK
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
and sensu Ornduff, 1993
; both = L. chrysostoma sensu Ornduff, 1966
; hereafter L. californica sensu lato [s.l.]), are often found in abundance at low elevations in the California Floristic Province and adjacent deserts from southern Oregon in the USA to northern Baja California, Mexico, including the Californian Islands (i.e., the Channel Islands and Guadalupe Island), and central Arizona (Ornduff, 1966
, 1993
). Across its broad geographic range, Lasthenia californica s.l. occupies the widest range of habitats and soil types of any species in Lasthenia. Populations have been found growing in alkali flats, chaparral, coastal bluffs, deserts, oak woodlands, open grasslands, and serpentine outcrops.
Extensive morphological variability of L. californica s.l. is reflected in its complex taxonomic history. In various taxonomic treatments before Ornduff's (1966)
study, members of L. californica s.l. were assigned to different species and subspecies spanning three genera. Ornduff (1966)
found that some of the variability in L. californica s.l. was due to phenotypic plasticity in response to climatic and edaphic conditions. Crossing and hybridization studies conducted by Ornduff (1966)
also indicated that "ecological diversity" within L. californica s.l. "is associated with the existence of numerous genetically distinct races, many of which show distinct interracial morphological differences" (p. 21). He suggested that genetic differences among races have accumulated as a result of reproductive isolation between geographically distinct populations. Subsequent biosystematic studies of L. californica s.l. based on flavonoid chemistry, isozymes, and morphology by Desrochers and Bohm (1993
, 1995
) provide strong evidence for divergence between geographically separated sets of populations. Rajakaruna and Bohm (1999)
showed that divergence among populations of L. californica extends to ecophysiological characteristics.
One component of variation in L. californica s.l. is ploidy level. Diploid (n = 8) and tetraploid (n = 16) populations occur across California. Diploid populations are most commonly found (Ornduff, 1966
). Ornduff (1966)
suggested that tetraploidy arose independently several times in L. californica s.l. and represents allopolyploidy, based on detailed observations of meiotic activity. Desrochers and Bohm (1995)
reported a hexaploid population, and a hexaploid population was also discovered during the course of our study.
Two species recognized by Ornduff (1966
, 1993
), L. leptalea and L. macrantha, are easily confused with L. californica s.l. Lasthenia macrantha (n = 16, 24) comprises all perennial members of Lasthenia and is found only in coastal habitats of Oregon and northern California. Three subspecies within L. macrantha were recognized by Ornduff (1966
, 1971
) and are distinguished by morphology, chromosome numbers, and biogeography. Lasthenia macrantha subsp. prisca (n = 16) is found only in coastal Oregon (Ornduff, 1971
). The other two subspecies, L. macrantha subsp. macrantha and L. macrantha subsp. bakeri, have chromosome numbers of n = 24, and both are endemic to coastal California. Apart from perenniality, members of L. macrantha can be exceedingly difficult to distinguish from L. californica. Members of L. leptalea (n = 8) are often misidentified as L. californica s.l. but can be consistently distinguished from L. californica s.l. by subulate anther tips, glabrous phyllaries (except at the tips), and more restricted interior distribution.
Taxonomic treatments of plants referable to L. californica s.l. prior to Ornduff's (1966)
survey of Lasthenia traditionally recognized three infraspecific taxa (see Ornduff, 1966
). The "gracilis" taxon was recognized on the basis of strigose pubescence, linear cypselae, and pappi of opaque, white, ovate-lanceolate scales, each tapering to an awn (Fig. 1B). The "chrysostoma" taxon was recognized by its hirsute pubescence, subclavate cypselae, and pappi of clear to brown subulate awns (Fig. 1C). The "hirsutula" taxon comprised the maritime races of the "chrysostoma" taxon, with similar pubescence, cypselae, and pappus morphology. Ornduff (1966)
found that although "these distinctions hold for a large number of different populations," (p. 58) a significant proportion of the plants he examined could not be unambiguously assigned to any of the traditionally recognized taxa. Worse, he found that southern coastal populations of L. californica s.l. in San Diego County, the Channel Islands, and Baja California possess the "gracilis" pappi, cypselae, and pubescence instead of the "hirsutula" characteristics traditionally associated with coastal populations. Based on these seemingly illogical and confusing findings, Ornduff (1966)
deferred recognition of any intraspecific taxa within L. californica until further studies could be undertaken.
|
survey, extensive studies have focused on the wide morphological and ecological diversity found within L. californica s.l. In a study of morphology and isozymes, Desrochers and Bohm (1995)
suggested that patterns of variation in L. californica s.l. best fit a model of geographic speciation resulting in a northern race and a southern race. Rajakaruna and Bohm (1999)
studied edaphic factors in populations of L. californica s.l. and suggested that these factors might play an important role in explaining the patterns of variation found in races of L. californica s.l.
In a preliminary study using molecular data, Chan and Ornduff (1998)
found that intraspecific sequence variation in the 18S–26S nuclear ribosomal internal transcribed spacer (ITS) region within L. californica s.l. is high, especially by comparison with negligible intraspecific ITS variation in most other members of the genus. In addition, sequence variation within L. californica diagnosed two distinct clades. Since then, we have increased our population sample size substantially and have added sequence data from the chloroplast 3' trnK intron and the 18S–26S nuclear ribosomal external transcribed spacer (ETS).
Here we present a molecular phylogeny for L. californica s.l. and close relatives in L. sect. Amphiachaenia (correct name for L. sect. Baeria sensu Ornduff [1966
] and redelimited to include L. leptalea [Chan, 2000
, 2001
; Chan, Baldwin, and Ornduff, 2001
]) to discern whether evolutionary lineages correspond with currently accepted taxa and to assess changes that occurred in molecules, morphology, chromosome numbers, and geographical distribution during diversification of L. sect. Amphiachaenia. Our choice of ingroup taxa is based on results of a broad phylogenetic study of Lasthenia (Chan, 2000
; Chan, Baldwin, and Ornduff, 2001
).
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
], formerly treated in L. sect. Burriellia [Ornduff, 1966
]) and outgroup species in Lasthenia. When fresh leaf material was not available, leaf tissue was obtained from herbarium specimens for which care was taken to sample only leaf tissue that was still green in color in order to have a greater chance of obtaining chloroplast DNA. Genomic DNA was extracted from three or more geographically separated populations of each species sampled in L. sect. Amphiachaenia. DNA samples from a total of 61 Lasthenia populations were extracted for this study. This sampling included 50 geographically isolated populations of L. californica s.l. and representatives of each of the other six Lasthenia sections recognized by Ornduff (1966)
. DNA was also extracted (by B. L. Wessa, Jepson Herbarium) from fresh leaf tissue of Amblyopappus pusillus and Eriophyllum congdonii, which were used as the outgroup based on ITS evidence that Amblyopappus belongs to the sister-group of Lasthenia and that E. congdonii (and other woolly sunflowers) are close relatives of Lasthenia (Baldwin and Wessa, 2000
). Voucher specimens of all sampled populations were deposited at UC/JEPS. Sources of all plant materials and GenBank accession numbers for DNA sequences have been archived at the American Journal of Botany Supplementary Data web site (http://ajbsupp.botany.org/v89/). Vouchers specimens from this study were examined to attempt to discern any morphological characteristics that may serve as markers for major molecular clades found in the study group. These morphological characters were then mapped onto the most parsimonious trees to assess evolutionary patterns. This procedure was especially important for determining whether any morphological characteristics diagnose clades within L. californica s.l.
DNA amplification and sequencing
All DNA extractions were performed using a cetyltrimethyl ammonium bromide (CTAB) extraction method for fresh leaf material (Doyle and Doyle, 1987
). An additional cleaning step using a 25 : 24 : 1 mixture of phenol : chloroform : isoamyl alcohol (pH 8.0) was substituted whenever herbarium material was used. The extracted genomic DNA was quantified using a GeneQuant II DNA/RNA Calculator spectrophotometer (Pharmacia, North Peapack, New Jersey, USA) and diluted to 30 ng/µL for use in the DNA amplification reactions. DNA extracted from herbarium material was further cleaned using the Ultra Clean 15 DNA purification kit (Mo Bio Laboratories, Solana Beach, California, USA) if the initial polymerase chain reaction (PCR) amplification was unsuccessful. The kit protocol for large DNA fragments was followed to avoid shearing the DNA. A tenfold dilution of this purified product was then used for PCR amplifications.
Primers used to amplify and sequence the entire ITS region of Lasthenia were designed from fungal ribosomal DNA (rDNA) sequences by White et al. (1990)
and have been used extensively in previous angiosperm studies (e.g., Baldwin, 1992
). Primer ITS5A (Downie and Katz-Downie, 1996
), based on White et al.'s (1990)
fungal primer ITS5 and corrected at two positions for angiosperms, was substituted for ITS5 in this study. Primers ITS5A and ITS4 were used to amplify the complete ITS region including the 5.8S ribosomal gene. Primers for amplifying the 3' end of the ETS (i.e., primers ETS-Hel-1 and 18S-ETS) were designed for Heliantheae by Baldwin and Markos (1998)
. Primers for amplifying the 3' intron segment of the chloroplast trnK gene were designed by Steele and Vilgalys (1994)
, with corrected reporting of the trnK2r primer by Johnson and Soltis (1994)
. The cpDNA primers were used to amplify part of the 5' end of the matK gene as well as the 3' trnK segment. Primer sequences are given in Table 1.
|
. The purified cycle sequencing products were then resolved by electrophoresis on 4.8% polyacrylamide (Amresco, Solon, Ohio, USA) gels using a Prism 377 automated DNA sequencer (Applied Biosystems, Foster City, California, USA). Sequences from both strands of each PCR product were examined, compared, and corrected using Sequence Navigator software (Applied Biosystems, Foster City, California, USA).
Chromosome counts
Floral buds (immature heads) were collected from populations of L. californica s.l. whenever possible and immediately immersed in fixative (three parts 95% ethanol, one part glacial acetic acid). The buds in fixative were then stored in a freezer until examined. Immature anthers were macerated and microsporocytes were prepared for meiotic chromosomes using squash technique and 2% aceto-carmine stain mixed with Hoyer's solution (for permanent slides). Counts were made from sketches of cells observed under phase contrast microscopy at 600–1000x magnification. At least three unequivocal chromosome preparations were observed for each population. All slides are deposited at JEPS.
Phylogenetic analysis of ITS, ETS, and 3' trnK intron sequences
All ITS, ETS, and cpDNA sequences were aligned visually, with insertion of gaps where necessary. Unambiguous insertions and deletions (indels) shared by two or more taxa were recoded as additional absence/presence characters following the "simple indel coding" method proposed by Simmons and Ochoterena (2000)
. A compartmentalization approach (Mishler, 1994
) was used to avoid the computational problems associated with the many nearly identical sequences. First, a "fast" stepwise-addition bootstrap with 1000 replicates was performed in PAUP* (Phylogenetic Analysis Using Parsimony) software (Swofford, 2000
) with unambiguously aligned characters to find well-supported clades. The resulting clades were then examined in MacClade software (Maddison and Maddison, 1992
) to select a terminal taxon (OTU) with a zero-length branch or a branch without homoplasious changes to represent an apical, well-supported clade in subsequent parsimony analyses using PAUP*.
Multiple populations were chosen to represent an apical, well-supported clade if that clade included members of more than one taxon recognized by Ornduff (1966
, 1971
, 1993
). Conversely, more than one representative of a taxon was selected for subsequent analyses if that taxon was not resolved as a monophyletic group. In addition, parsimony analyses using the heuristic search function in PAUP* (with 1000 replicate runs each with 20 random stepwise additions) were performed separately on each of the well-supported clades with all sequences included, i.e., full heuristic searches were conducted for all samples of each subclade in L. sect. Amphiachaenia.
Maximum parsimony analysis and parsimony bootstrap analysis (with 1000 replicate runs, each with 20 random taxon additions) of the aligned ITS, ETS, and cpDNA sequences were conducted for each data set and for the combined data set using the reduced representative-taxon set via heuristic searches in PAUP*. Decay analyses were performed with PAUP* using the reverse constraints approach as implemented by AutoDecay (Eriksson, 1998
). Congruence of signal among the ITS, ETS, and cpDNA data sets was assessed using the partition homogeneity test (Farris et al., 1994
, 1995
) as implemented in PAUP*, with 100 replicate data partitions and using heuristic searches. Likelihood ratio tests for significant differences between clock-constrained and clock-unconstrained trees (Felsenstein, 1988
; see Baldwin and Sanderson, 1998
) were performed on the minimum-length trees obtained from parsimony analyses to test for evolutionary rate constancy across lineages of the individual ITS, ETS, and cpDNA data sets and the combined data set. The likelihood ratio tests were conducted with PAUP* using Hasegawa, Kishino, and Yano's (1985)
model (HKY85) of sequence evolution and a gamma distribution of rate variation among sites (with the shape parameter estimated and with four rate categories).
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
The ITS1 + ITS2 sequences for L. sect. Amphiachaenia vary in length from 460 to 473 base pairs (bp). Pairwise distances between unambiguously aligned ITS1 + ITS2 sequences range from 0 to 10.9% of nucleotides (all distance values stated are corrected HKY85 distances). An 11-bp deletion within ITS1 was found in 27 of the 42 sequences of L. californica s.l. and in all three sequences of L. leptalea. Length of the 5.8S ribosomal gene was found to be 164 bp. The 3' ETS sequences range in length from 482 to 493 bp, with pairwise distance values ranging from 0 to 14.6% of nucleotides. A 10-bp insertion was found to occur in only 12 of the ETS sequences of L. californica s.l. The cpDNA sequences range in length from 403 to 461 bp, with pairwise divergence values ranging from 0 to 3.8% of nucleotides. All length variation in the cpDNA sequences for L. sect. Amphiachaenia occurred within the trnK intron region, mostly in the form of a length-variable string of adenine nucleotides except for L. leptalea, which has length mutations in the intron region that are unique in the genus.
Considerable intraspecific sequence variation was found in both species of L. sect. Amphiachaenia recognized by Ornduff (1966)
, i.e., L. californica s.l. and L. macrantha. Pairwise distances for sequences of L. californica s.l. range from 0 to 8.2% for ITS1 + ITS2, from 0 to 14.6% for ETS, and from 0 to 2.4% for cpDNA. For L. macrantha, pairwise distances for sequences range from 0 to 5.8% for ITS1 + ITS2, from 0 to 8.2% for ETS, and from 0 to 1.1% for cpDNA.
By comparison, except for L. coronaria, pairwise distances for unambiguously aligned sequences in intraspecific comparisons within other species of Lasthenia recognized by Ornduff (1966)
are much lower than for L. californica s.l. and L. macrantha, ranging from 0 to 0.7% for all three regions (Chan, 2000
; Chan, Baldwin, and Ornduff, 2001
). Pairwise distances for intraspecific comparisons within L. coronaria (between two putative geographical races) are higher than for most species of Lasthenia outside L. sect. Amphiachaenia (0–2.8% for all three regions [Chan, 2000
; Chan, Baldwin, and Ornduff, 2001
]).
Phylogenetic resolutions
The majority-rule bootstrap consensus tree (Fig. 2) from the 1000-replicate "fast" stepwise-addition bootstrap analysis of the full data matrix (i.e., with all 63 sequences mentioned above) shows with 100% bootstrap support that L. sect. Amphiachaenia is monophyletic only with the inclusion of L. leptalea from L. sect. Burrielia, as resolved in a broad-scale phylogenetic study of Lasthenia (see Chan, 2000
; Chan, Baldwin, and Ornduff, 2001
). In the "fast" stepwise-addition bootstrap tree, the two major clades corresponding to L. sect. Amphiachaenia constitute a polytomy with L. leptalea. Lasthenia californica s.l. populations are in both major clades of L. sect. Amphiachaenia. Results from parsimony bootstrap analyses using full heuristic searches for each of the major clades in L. sect. Amphiachaenia do not differ substantially from the "fast" stepwise-addition bootstrap tree.
|
The remaining 27 populations of L. californica s.l. correspond to a second, robustly supported (96% bootstrap) major clade in L. sect. Amphiachaenia, with some finer-scale lineages resolved within the clade. All representatives of the 27 populations of L. californica s.l. that constitute the second major clade were found to have an 11-bp deletion in their ITS1 sequences. We will refer to these populations of L. californica s.l. as L. californica-White, based on pappus color. All samples of L. leptalea also share this 11-bp ITS1 deletion. The 15 populations of L. californica s.l. mentioned in the previous paragraph and L. macrantha do not possess this 11-bp deletion in ITS1. We refer to the 15 populations of L. californica s.l. as L. californica-Brown, based also on pappus color.
Upon resolving the two major clades within L. californica s.l., pairwise distances for sequences of L. californica-Brown were calculated to be 0–1.1% for ITS1 + ITS2, 0–1.7% for ETS, and 0–0.54% for cpDNA. Pairwise distances for sequences of L. californica-White were calculated to be 0–5.7% for ITS1 + ITS2, 0–10.6% for ETS, and 0–1.9% for cpDNA. ITS, ETS, and cpDNA sequences in the L. californica-Brown clade are less variable both in length and sequence than those of L. californica-White.
A 10-bp insertion was found in ETS sequences of two subclades within the L. californica-White clade. The two subclades have 100% and 67% bootstrap support.
Results from the "fast" stepwise-addition bootstrap analyses (with and without recoded indels) indicate that the addition of recoded indel characters resulted in an increase in overall bootstrap support for resolved clades and increased resolution of lineages within the two major clades of L. californica s.l. Recoded indel characters were therefore included in all subsequent phylogenetic analyses. Inclusion of ambiguously aligned nucleotides did not affect the "fast" stepwise-addition bootstrap tree topology and these sites were not included in further analyses.
Trees from separate parsimony analyses of ITS, ETS, and cpDNA data sets with reduced taxon sampling (Fig. 3) are topologically congruent with each other except for the placement of L. macrantha subsp. prisca in the ITS/cpDNA trees versus the ETS trees. Within the ITS and cpDNA trees, L. macrantha subsp. prisca is positioned sister to, or in, the clade that comprises the other two subspecies of L. macrantha and L. californica-Brown. Within the ETS tree, L. macrantha subsp. prisca is resolved as sister to L. californica-White. When all three data sets are analyzed simultaneously, the resulting position of L. macrantha subsp. prisca follows that of the ITS tree.
|
Likelihood ratio tests for evolutionary rate constancy across lineages of the minimum-length trees for ITS data and for ETS data yielded significant differences in likelihood between the clock-constrained and clock-unconstrained trees; the rate of nucleotide change in the ITS region and in the ETS apparently has not been constant during the evolutionary history of Lasthenia. Only rate constancy of nucleotide change in the cpDNA of Lasthenia cannot be rejected.
Maximum parsimony analysis of the combined ITS, ETS, and cpDNA data set (including one representative sequence from each taxon) via a full heuristic search in PAUP* generated two minimum length trees of 1079 steps each (Fig. 4). The two trees differ only in whether a minor clade is resolved within the clade containing L. californica-Brown and two of the subspecies of L. macrantha. The topology of the strict consensus tree differs from that of the "fast" stepwise-addition bootstrap tree only in (weakly) resolving the base of L. sect. Amphiachaenia, by placing L. californica-White as the most basal divergent clade.
|
|
|
As noted above, epappose and pappose plants were found among the herbarium collections for both the L. californica-White and L. californica-Brown clades. However, "chrysostoma" and "gracilis" pappus-type plants were never found together in any single herbarium collection. About 10–15% of the herbarium collections examined were of only epappose plants and were not scored. Epappose plants could not be assigned to a clade with absolute certainty; cypsela shape and/or pubescence type could not be identified unambiguously or cypselae were absent. At present, without molecular data, identifying epappose plants occurring in the northern portion of the range of L. californica s.l. as members of either L. californica-White or L. californica-Brown would be very difficult if not impossible. Lack of a pappus appeared to be more common in plants within the L. californica-Brown clade, especially in coastal populations, which also may tend towards morphology approximating that of the perennial L. macrantha, with more succulent herbage, wider leathery leaves, and larger heads (Ornduff, 1966
).
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
). Lasthenia sect. Amphiachaenia and L. sect. Burrielia traditionally have been considered to be closely related, with L. californica s.l. and L. leptalea showing particularly high similarity (Ornduff, 1966
). In Lasthenia, only L. sect. Amphiachaenia and L. sect. Burrielia possess anthochlor pigments. The flavonoid profiles of L. sect. Amphiachaenia and L. sect. Burrielia are also very similar to each other and differ from those of the rest of the genus (Bohm, Saleh, and Ornduff, 1974
; Ornduff, Bohm, and Saleh, 1974
). Plants of L. californica s.l. and L. leptalea can be easily mistaken for each other and both have a chromosome number of n = 8. In the biosystematic survey of Lasthenia conducted by Ornduff (1966)
, the only successful intersectional hybridization made was between L. californica s.l. and L. leptalea. The two hybrids obtained had pollen stainabilities of 30% and 51%, levels higher than those measured for some hybrids made between diploid populations of L. californica s.l. The flavonoid profile of L. leptalea is less similar to the other two members of L. sect. Burrielia than to L. californica s.l. (Bohm, Saleh, and Ornduff, 1974
; Ornduff, Bohm, and Saleh, 1974
). Attempts by Ornduff (1966)
to cross L. leptalea with the other two members of L. sect. Burrielia resulted only in a sterile, depauperate hybrid. Fine-scale relationships of L. leptalea within L. sect. Amphiachaenia remain uncertain. Lasthenia leptalea possesses the 11-bp ITS1 deletion found elsewhere in Lasthenia only in L. californica-White, but possesses pappus elements that are more similar to the clear to brown subulate awns ("chrysostoma" type) in pappose plants of L. californica-Brown and L. macrantha. If L. californica-White and L. leptalea do not constitute a clade, then the 11-bp deletion either occurred independently in the two groups or occurred in the common ancestor of L. sect. Amphiachaenia, with a subsequent reversal (reinsertion of lost nucleotides) at the base of the L. californica-Brown/L. macrantha clade. The possibility of an 11-bp reinsertion is tenable based on the slightly different nucleotide composition of the 11-bp region for L. californica-Brown and the Californian L. macrantha subspecies (i.e., L. macrantha subsp. bakeri and L. macrantha subsp. macrantha) compared to that of L. macrantha subsp. prisca.
Nonmonophyly and cryptic diversity of L. californica s.l.
Evidence presented here for paraphyly or polyphyly of L. californica s.l. and for only cryptic differences between epappose plants of both major clades of L. californica s.l. helps to resolve persistent taxonomic problems surrounding the group. Ornduff's (1966)
contention that the traditional recognition of three intraspecific taxa in L. californica s.l. was "confusing and unworkable" can be validated. He observed that coastal and island populations of the maritime "hirsutula" taxon located in the southern portion of the range were composed of plants possessing the "gracilis" morphology. This conclusion is supported by the observed distribution of the L. californica-White clade, which is the only clade of L. sect. Amphiachaenia found in southern California (including the Channel Islands) and Baja California (including Guadalupe Island).
Our results also help to extend evolutionary insights from previous biosystematic studies of L. californica s.l. Desrochers and Bohm (1995)
noted four pappus states (including the epappose condition) in L. californica s.l. and found a nonrandom geographical distribution for two of the pappus types. The subulate and lanceolate pappus types reported by Desrochers and Bohm (1995)
probably represent the opaque, white, ovate-lanceolate "gracilis" pappus type of L. californica-White. Although we could not correlate our DNA data with the isozyme data of Desrochers and Bohm (1995)
, their conclusion of northern and southern races of L. californica s.l. may be mirrored by our data supporting a northern distribution for L. californica-Brown and a wider, in part southern, distribution for L. californica-White. Rajakaruna and Bohm (1999)
concluded that the two geographical races of L. californica s.l. correspond to two edaphically distinct groups. Rajakaruna and Bohm's (1999)
"Race A" appears to correspond in geographical distribution to L. californica-White and their "Race C" appears to correspond in distribution to L. californica-Brown. Uncertainty remains about whether the groups L. californica-White and L. californica-Brown actually equate to Rajakaruna and Bohm's (1999)
races; members of the two clades resolved here span habitat types to which the edaphic races are restricted (N. Rajakaruna, University of British Columbia, personal communication).
Irrespective of ecophysiological differences associated with edaphic conditions, differences in geographic distribution of the L. californica-White and L. californica-Brown clades may reflect differences in their ecological tolerances. Plants of the L. californica-White clade are more widely distributed and may be more tolerant of extreme climatic conditions, e.g., in desert habitats. In contrast, plants of the L. californica-Brown clade are found only in the northern portion of the reported range of L. californica s.l., in which the climate is wetter and cooler on average than in southern California and Baja California. Differences in geographic distribution that may reflect ecophysiological differences between cryptically distinct, nonsister lineages also have been reported in Downingia yina (Schultheis, 1999
) and in the Madiinae (Baldwin, 2000
).
Our finding that diploid and tetraploid populations occur in both the L. californica-White and L. californica-Brown clades is consistent with Ornduff's (1966)
hypothesis (based on lines of evidence apart from cytogenetics) that polyploids referable to L. californica s.l. do not constitute a monophyletic group. The hexaploid L. californica s.l. population we discovered belongs to the L. californica-Brown clade. Congruence of cpDNA and rDNA trees and resolution in the rDNA trees alone are inconsistent with the possibility that bidirectional concerted evolution may account for apparent polyphyly of polyploids and that polyploids may constitute a monophyletic group (see Wendel, Schnabel, and Seelanan, 1995a
). Polyploidy evidently has evolved independently in the L. californica-White and L. californica-Brown clades.
Some internal clades within L. californica-White are robustly supported by rDNA data; we did not find any chromosomal or morphological characteristics that diagnose these clades. Some of the internal clades within L. californica-White are narrowly distributed geographically. Additional study (i.e., with another, unlinked set of molecular markers) is warranted to ascertain whether the robust rDNA clades within L. californica-White represent cryptic evolutionary lineages or, alternatively, manifestations of lineage sorting or hybridization.
Nonmonophyly of L. macrantha
Lasthenia macrantha is not monophyletic; the clade comprising L. macrantha also includes members of L. californica s.l. These results support previous interpretations of a close relationship between L. californica s.l. and L. macrantha, first by Gray (1857)
, who considered the two taxa to be varieties of the same species, and later by Ornduff (1966)
, who discussed the relationship in detail. Ornduff (1971)
stated that L. californica s.l. "may be viewed as an annual version of L. macrantha" (p. 98). Hybridization studies by Ornduff (1966)
show very high (60–100%) pollen stainability in hybrids between L. macrantha and tetraploid L. californica s.l., which led him to conclude that the two taxa are closely related.
Although all members of L. macrantha are perennials of coastal habitats, Ornduff (1966)
reported that California populations of L. macrantha; i.e., L. macrantha subsp. bakeri and L. macrantha subsp. macrantha, may flower the first year and behave as annuals under prolonged drought conditions, whereas Oregon populations, i.e., L. macrantha subsp. prisca, are long-lived perennials. Ornduff (1971)
also suggested that the tetraploid L. macrantha subsp. prisca may be ancestral to the two hexaploid subspecies of L. macrantha, i.e., L. macrantha subsp. bakeri and L. macrantha subsp. macrantha. The results presented here for a basally divergent position of L. macrantha subsp. prisca relative to the clade comprising L. californica-Brown and the Californian populations of L. macrantha indicate that if L. macrantha subsp. prisca is ancestral to the hexaploids, then the other genome in the hexaploids (and the rDNA fixed in the hexaploids) probably was inherited from a diploid of common genomic constitution with the diploid members of L. californica-Brown.
An amphidiploid origin of the tetraploid condition in L. macrantha subsp. prisca involving ancestors of L. californica-White and the clade comprising L. californica-Brown and the rest of L. macrantha could explain the only strong phylogenetic incongruence found between the data sets in this study, i.e., between the ITS and the ETS sequences of L. macrantha subsp. prisca. Under this scenario, the cpDNA and ITS sequences of L. macrantha subsp. prisca reflect a genomic contribution by an ancestor basal to the clade comprising L. californica-Brown, L. macrantha subsp. bakeri, and L. macrantha subsp. macrantha, and the ETS sequences of L. macrantha subsp. prisca reflect a genomic contribution by an ancestor basal to L. californica-White. Recombination between the parental rDNAs and subsequent fixation of recombined sequences via concerted evolution in the amphidiploid could lead to the observed phylogenetic patterns for L. macrantha subsp. prisca (see Wendel, Schnabel, and Seelanan, 1995b
). On the basis of our findings, recognition of L. macrantha subsp. prisca as a distinct species is warranted and represents another good example of cryptic diversity within L. sect. Amphiachaenia.
Despite uniform occurrence of the annual habit throughout the rest of Lasthenia, life-form shifts between annual and perennial habits in L. sect. Amphiachaenia appear to have occurred repeatedly based on the molecular trees. Lasthenia californica-Brown could be viewed as a "racially complex, slightly more specialized derivative" (p. 29) of L. macrantha, as suggested by Ornduff (1966)
for L. californica s.l. in general. As noted above, Californian populations of L. macrantha can behave as annuals (Ornduff, 1966
), so an evolutionary shift back to the annual habit is not difficult to envision. Another possibility is that L. macrantha represents a polyphyletic group of lineages that independently acquired the perennial habit (i.e., the annual habit of L. californica-Brown is plesiomorphic). This less parsimonious hypothesis is easier to envision ecologically and may explain evident nonmonophyly of L. macrantha subsp. bakeri and L. macrantha subsp. macrantha. Some coastal populations of annual L. californica-Brown do exhibit L. macrantha-like morphology and germinate earlier and live longer than other annuals in the group (Ornduff, 1966
). Carlquist (1965)
noted that a perennial habit should be favored in a moderate, equable climate as found on the California coast and on oceanic islands. Numerous examples of groups that have evolved a perennial or woody habit on oceanic islands have been documented (see Baldwin et al., 1998
).
Taxonomic implications
Based on the results of this molecular phylogenetic study, L. californica s.l. should be treated as a minimum of two distinct taxa and other (rDNA) clades within one of the two groups (i.e., L. californica-White) warrant further study. A revised taxonomy (Chan, 2001
) of L. sect. Amphiachaenia that better reflects phylogenetic relationships is outlined in Table 2. The two major, nonsister lineages in L. californica s.l. are unequivocally distinguished by subtle but distinct differences in pappus morphology and, to some extent, by geographic distribution. Epappose populations occur in both groups and may be distinguished from each other by geography where their ranges do not overlap. At present, distinguishing between epappose populations of the two clades in areas where their ranges overlap would be difficult if not impossible without access to molecular data. Although recognition of the two clades as distinct species may result in a taxonomy that is less convenient for purposes of identification than continued adherence to the artificial species Lasthenia californica s.l., we believe that the need to recognize only natural groups must supercede practical considerations in systematics (see Baldwin, 2000
). The examples provided here constitute strong evidence for cryptic evolutionary lineages in angiosperms and serve to illustrate how molecular data can provide a refined means of assessing plant biodiversity in widespread, variable, and evolutionarily active groups, as has proven possible in numerous animal lineages (see Avise, 2000
).
|
|
FOOTNOTES |
|---|
2 Author for reprint requests (present address: Department of Botany, University of Hawaii at Manoa, 3190 Maile Way, Room 101, Honolulu, Hawaii 96822-2279 USA; raymund{at}socrates.berkeley.edu
)
|
LITERATURE CITED |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
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[CrossRef][Medline]
Baldwin B. G. 2000 Roles for modern plant systematics in the discovery and conservation of fine-scale biodiversity. Madroño 47: 219-229
Baldwin B. G. D. J. Crawford J. Francisco-Ortega S. C. Kim T. Sang T. F. Stuessy 1998 Molecular phylogenetic insights on the origin and evolution of oceanic island plants. In D. E. Soltis, P. S. Soltis, and J. J. Doyle [eds.], Molecular systematics of plants II, 410–441. Kluwer Academic, Boston, Massachusetts, USA
Baldwin B. G. S. Markos 1998 Phylogenetic utility of the external transcribed spacer (ETS) of 18S–26S rDNA: congruence of ETS and ITS trees of Calycadenia (Compositae). Molecular Phylogenetics and Evolution 10: 449-463[CrossRef][ISI][Medline]
Baldwin B. G. M. J. Sanderson 1998 Age and rate of diversification of the Hawaiian silversword alliance (Compositae). Proceedings of the National Academy of Sciences of the United States of America 95: 9402-9406
Baldwin B. G. B. L. Wessa 2000 Origin and relationships of the tarweed-silversword lineage (Compositae: Madiinae). American Journal of Botany 87: 1890-1908
Bohm B. A. N. A. M. Saleh R. Ornduff 1974 The flavonoids of Lasthenia (Compositae). American Journal of Botany 61: 551-561[CrossRef][ISI]
Carlquist S. 1965 Island life: a natural history of the islands of the world. Natural History Press, Garden City, New York, USA
Chan R. 2001 Taxonomic changes and a new species in Lasthenia sect. Amphiachaenia (Compositae: Heliantheae sensu lato). Madroño 48: 205-210
Chan R. B. G. Baldwin R. Ornduff 2001 Goldfields revisited: a molecular phylogenetic perspective on the evolution of Lasthenia (Compositae: Heliantheae sensu lato). International Journal of Plant Sciences 162: 1347-1360[CrossRef]
Chan R. R. Ornduff 1998 Goldfields revisited: phylogenetic analysis of the genus Lasthenia (Asteraceae) using morphology and DNA sequence data from the nuclear ribosomal ITS and chloroplast trnK intron regions. American Journal of Botany 85: 119 (Abstract)
Chan R. K.-G. 2000 Molecular systematics of the goldfield genus Lasthenia (Compositae: Heliantheae sensu lato). Ph.D. dissertation, University of California, Berkeley, California, USA
Desrochers A. B. A. Bohm 1993 Flavonoid variation in Lasthenia californica (Asteraceae). Biochemical Systematics and Ecology 21: 449-453[CrossRef]
Desrochers A. B. A. Bohm 1995 Biosystematic study of Lasthenia californica (Asteraceae). Systematic Botany 20: 65-84
Downie S. R. D. S. Katz-Downie 1996 A molecular phylogeny of Apiaceae subfamily Apioideae: evidence from nuclear ribosomal DNA internal transcribed spacer sequences. American Journal of Botany 83: 234-251[CrossRef][ISI]
Doyle J. J. J. L. Doyle 1987 A rapid DNA isolation procedure for small quantities of fresh leaf tissue. Phytochemical Bulletin 19: 11-15
Eriksson T. 1998 AutoDecay, version 4.0.2 (program distributed by the author). Bergius Foundation, Royal Swedish Academy of Sciences, Stockholm, Sweden
Farris J. S. M. Källersjö A. G. Kluge C. Bult 1994 Testing significance of incongruence. Cladistics 10: 315-319[CrossRef][ISI]
Farris J. S. M. Källersjö A. G. Kluge C. Bult 1995 Constructing a significance test for incongruence. Systematic Botany 44: 570-572
Felsenstein J. 1988 Phylogenies from molecular sequences: inference and reliability. Annual Review of Genetics 22: 521-566[CrossRef][ISI][Medline]
Gray A. 1857 Compositae. In J. Torrey [ed.], Description of the general botanical collections. Reports of explorations and surveys, to ascertain the most practical and economical route for a railroad from the Mississippi River to the Pacific Ocean 4: 61-182
Hasegawa M. H. Kishino T. Yano 1985 Dating the human-ape split by a molecular clock of mitochondrial DNA. Journal of Molecular Evolution 22: 160-174[CrossRef][ISI][Medline]
Johnson D. E. R. Ornduff 1978 Lasthenia californica (Compositae), another name for a common goldfield. Madroño 25: 227.
Johnson L. A. D. E. Soltis 1994 matK DNA sequences and phylogenetic reconstruction in Saxifragaceae sensu stricto. Systematic Botany 19: 143-156
Maddison W. P. D. R. Maddison 1992 MacClade version 3: analysis of phylogeny and character evolution. Version 3.08. Sinauer, Sunderland, Massachusetts, USA
Mishler B. D. 1994 Cladistic analysis of molecular and morphological data. American Journal of Physical Anthropology 94: 143-156[CrossRef][ISI][Medline]
Ornduff R. 1966 A biosystematic survey of the goldfield genus Lasthenia (Compositae: Helenieae). University of California Publications in Botany 40: 1-92
Ornduff R. 1971 A new tetraploid subspecies of Lasthenia (Compositae) from Oregon. Madroño 21: 96-98
Ornduff R. 1993 Lasthenia. In J. C. Hickman [ed.], The Jepson manual: higher plants of California, 298–300. University of California Press, Berkeley, California, USA
Ornduff R. B. A. Bohm N. A. M. Saleh 1974 Flavonoid races in Lasthenia (Compositae). Brittonia 26: 411-420[CrossRef][ISI]
Rajakaruna N. B. A. Bohm 1999 The edaphic factor and patterns of variation in Lasthenia californica (Asteraceae). American Journal of Botany 86: 1576-1596
Schultheis L. M. 1999 Systematics of Downingia (Campanulaceae) and population structure of D. bicornuta. Ph.D. dissertation, University of California, Berkeley, California, USA
Simmons M. P. H. Ochoterena 2000 Gaps as characters in sequence-based phylogenetic analyses. Systematic Biology 49: 369-381[CrossRef][ISI][Medline]
Steele K. P. R. Vilgalys 1994 Phylogenetic analyses of Polemoniaceae using nucleotide sequences of the plastid gene matK. Systematic Botany 19: 126-142
Swofford D. L. 2000 PAUP*: phylogenetic analysis using parsimony (*and other methods), version 4.0b3. Sinauer, Sunderland, Massachusetts, USA
Wendel J. F. A. Schnabel T. Seelanan 1995a Bidirectional interlocus concerted evolution following allopolyploid speciation in cotton (Gossypium). Proceedings of the National Academy of Sciences, USA 92: 280-284
Wendel J. F. A. Schnabel T. Seelanan 1995b An unusual ribosomal DNA sequence from Gossypium gossypioides reveals ancient, cryptic, intergenomic introgression. Molecular Phylogenetics and Evolution 4: 298-313[CrossRef][ISI][Medline]
White T. J. T. Bruns S. Lee 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, California, USA
This article has been cited by other articles:
|
|
 |
|
  P. M. Beardsley, S. E. Schoenig, J. B. Whittall, and R. G. Olmstead Patterns of evolution in western North American Mimulus (Phrymaceae) Am. J. Botany, March 1, 2004; 91(3): 474 - 489. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||
