| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Systematics, Phytogeography, and Evolution |
Department of Biology, Texas A & M University, College Station, Texas 77845 USA
Received for publication August 24, 2000. Accepted for publication February 15, 2001.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
Key Words: chloroplast DNA • edaphic endemism • genetic diversity • natural selection • ribosomal ITS • speciation
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
; Proctor and Woodell, 1975
; Brooks, 1987
; Proctor, 1999
). Serpentine occurrences are often characterized by an abrupt change in vegetation, with greatly reduced plant densities, biomass, and diversity, as well as distinctly different species compositions. Most plant species are excluded from serpentine soils; however, a small minority of taxa has evolved physiological and developmental mechanisms that allow them to survive—and even thrive. The mechanisms of adaptation to serpentine are largely unknown, although ability to grow at low calcium levels appears to be important (Walker, 1948
; Kruckeberg, 1954
). Using "paired population" experiments, Kruckeberg (1951, 1954)
demonstrated that serpentine races of Streptanthus glandulosus Hook., Phacelia californica Cham., and Salvia columbariae Benth. had much higher levels of tolerance to serpentine soils than did nonserpentine races of these taxa, indicating the existence of distinct genetic determinants of serpentine tolerance.
Serpentine-derived soils are a significant impetus for plant speciation and endemism worldwide (Brooks, 1987
; Proctor, 1999
). Because of strong selective pressures associated with adaptation to serpentine, hybrids between serpentine-tolerant plants and their nonserpentine neighbors may exhibit reduced fitness in either environment. It is not surprising that adaptation to serpentine soils is often concomitant with reproductive isolation from parapatric nonserpentine populations, as determined by hand-pollination experiments (Kruckeberg, 1954, 1957, 1986
). Serpentine soils contribute more rare and endemic plants to the rich flora of California than any other edaphic substrate (Skinner and Pavlik, 1994
). One explanation for restriction of these taxa to serpentine soils is that the rigors of the serpentine environment exclude nonadapted species, thus providing serpentine-tolerant species a refuge from competition.
The Santa Barbara Jewelflower, Caulanthus amplexicaulis var. barbarae (J. Howell) Munz (Brassicaceae), is a rare herbaceous annual plant restricted to an archipelago of serpentine exposures in the San Rafael Mountains, near the southern terminus of the outer Coast Ranges of California, USA (Howell, 1962
). Its sister taxon, C. amplexicaulis var. amplexicaulis S. Watson, has a more widespread distribution in the Transverse Ranges of southern California. Although the Transverse Ranges contain a complex and diverse mixture of geologic substrates, C. amplexicaulis var. amplexicaulis is largely restricted to granitic soils. Two small populations (near Thorne Meadow in Ventura County, California) are found on shale outcrops. At a distance of 
75 km, these are the closest var. amplexicaulis populations to the var. barbarae on serpentine. Although ecologically and geographically isolated, C. amplexicaulis var. barbarae and C. amplexicaulis var. amplexicaulis are morphologically quite similar and remain fully interfertile in artificial intervarietal crosses (Kruckeberg, 1984
; A. E. Pepper and L. E. Norwood, unpublished data), with 95% viable F1 pollen, and F1 and F2 seed viability that was statistically indistinguishable from intravarietal crosses. The existence of genetically compatible taxa with such distinct edaphic requirements presents a unique opportunity for intensive study of the genetic basis of tolerance to serpentine soils.
Caulanthus (±13 spp.) and related genera Streptanthus (±40 spp.), Streptanthella (1 sp.), and Guillenia (3 spp.) are sometimes referred to as the "Streptanthoid Complex" of genera due to their close taxonomic affinity. To provide a solid foundation for evolutionary and ecological-genetic studies of C. amplexicaulis var. barbarae, we sought to determine the genetic and the phylogenetic relationships of this taxon to the nonserpentine taxon C. amlpexicaulis var. amplexicaulis and to other serpentine and nonserpentine taxa in the Streptanthoid Complex. The overwhelming majority of these species are annual plants with a chromosome complement of n = 14, 2n = 28 (Rollins, 1993
). The ±57 species in the complex are restricted to the central and western United States and northern Mexico (Rollins, 1993
). Of the ±40 species in Streptanthus, ±16 have affinities for serpentine soils, providing the largest contribution of any genus to the serpentine floras of the California Floristic Province (CFP). Not surprisingly, Streptanthus and related species are considered to be important models for the evolution of serpentine plant taxa (Raven and Axelrod, 1978
; Kruckeberg, 1984, 1986
). Based on traditional taxonomic treatments, most of the serpentine endemic taxa of Streptanthus have been placed in the subgenus Euclesia (Morrison, 1941
). Within Euclesia, the S. glandulosus species complex has been particularly well characterized using traditional taxonomic methods (Kruckeberg, 1957, 1958
) and molecular analyses of systematic relationships and population structure (Mayer and Soltis, 1994, 1999
; Mayer, Soltis, and Soltis, 1994
).
Within Caulanthus, however, C. amplexicaulis var. barbarae is the only taxon with known affinity for serpentine—although some cogeners show tolerance to other substrates with unusual chemistries, such as gypsum, alkali, limestone, and gabbro soils (Kruckeberg, 1984
; Beauchamp, 1986
; Rollins, 1993
). Because C. amplexicaulis var. barbarae is found on one of the most southerly and isolated serpentine exposures of the CFP, it is biogeographically well suited for studies of the evolution of a serpentine-adapted endemic species. In the present study, we used nuclear ribosomal internal transcribed spacer (ITS) sequences and chloroplast trnL intronic sequences to generate hypotheses for the intraspecific genetic differentiation within C. amplexicaulis and its phylogenetic relationships to other serpentine and nonserpentine taxa of the Streptanthoid Complex. Further, we attempted to integrate a phylogenetic perspective with what is known about the comparative edaphic biology of several of the streptanthoid taxa to develop a hypothetical model for the evolution of C. amplexicaulis var. barbarae.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
. The taxa examined are listed in Table 1, and locations are shown in Fig. 1. Populations of C. amplexicaulis var. barbarae are clustered in a set of serpentine occurrences in the San Rafael Mountains of Santa Barbara County, California (inset in Fig. 1). We sampled all five known populations of this taxon (designated CAB1 through CAB5). However, some populations probably remain undocumented in this rugged mountainous region. CAB1 was obtained from a population growing on the tailings of an historic chromium mine (containing serpantinite, chromite, and other ultramafic rocks). CAB2 and CAB3 were obtained from natural serpentine outcrops. CAB4 and CAB5 were obtained from distinct populations on a recently burned alluvial slope containing soils derived from a mix of weathered serpentine and other parental rock types. Accessions of C. amplexicaulis var. amplexicaulis were obtained from a granite talus slope on Portal Ridge, Los Angeles County, California (CAA1) and from shale occurrences near Thorne Meadow, Ventura County, California (CAA2, CAA3). Samples from 11 additional specific and intraspecific taxa in the allied streptanthoid genera (Caulanthus, Streptanthus, Guillenia), including the serpentine taxa S. albidus ssp. paramoenus (E. Greene) Kruckeb., S. batrachopus Morrison, S. glandulosus ssp. pulchellus (E. Greene) Kruckeb., and S. niger E. Greene, were obtained from the habitats and locations indicated in Table 1 and Fig. 1. Voucher specimens were not taken from small populations of rare, threatened, or endangered taxa; these populations have been adequately documented (Howell, 1962
; Dieringer, 1991
; Mayer and Soltis, 1994
; Mayer, Soltis, and Soltis, 1994
; Skinner and Pavlik, 1994
). Representative voucher specimens previously collected from these populations, or from nearby populations of the same species, are listed in Table 1. Information on the voucher specimens of nonsensitive taxa is also listed in Table 1.
|
|
DNA sequence analysis
Ribosomal ITS fragments were amplified using primers "ITS4" and "ITS5" (White et al., 1990
). Primers "CP-C" and "CP-D" (Taberlet et al., 1991
) were used to amplify an 0.5-kb intron between the chloroplast trnL (UAA) 5' and 3' exons. Polymerase chain reaction (PCR) was performed using established conditions (Konieczny and Ausubel, 1993
) in 50-µL reactions containing 10–20 ng of genomic DNA. In order to remove residual dNTPs and oligonucleotide primers prior to sequencing, PCR products were purified by precipitation with 26% PEG 8000, 6.5 mmol/L MgCl2, 0.6 mol/L NaOAc pH 7.0 (Rosenthal, Coutelle, and Craxton, 1993
). An aliquot of 20–50 ng of purified double-stranded PCR product was used as template for 35 cycles of sequencing using BigDye terminator chemistry (Applied Biosystems, Foster City, California, USA). The primers used for direct sequencing were ITS4, ITS5, ITS11 (5'-ATCTCGGCTCTCGCATCGATG-3'), and ITS12 (5'-CAAAGACTCGATGGTTCACG-3'). Primers CP-C and CP-D were used to sequence the trnL (UAA) intron. Cycle sequencing products were purified by Bio-Gel P-30 size exclusion chromatography (Bio-Rad, Richmond, California, USA) and analyzed using an ABI 377 semiautomated sequencer (Applied Biosystems). Double-stranded DNA sequence contigs from each taxon were assembled and vetted using Sequencher 3.0 (Gene Codes, Ann Arbor, Michigan, USA). Finished sequences from the various taxa were aligned using Sequencer 3.0, and alignments across insertion/deletion differences (indels) were vetted manually. Nuclear ITS and chloroplast trnL sequences reported in this study were submitted to GenBank and assigned the accession numbers listed in Table 1.
In one case, an ITS sequence dimorphism within an individual accession was verified using a cleaved amplified polymorphic sequence (CAPS) strategy of Konieczny and Ausubel (1993)
. Ten microliters of the original ITS PCR reaction was digested for 8 hr using 10 units of the restriction enzyme HaeII, which only cleaves one of the two dimorphic alleles (recognition site = 5'-GGCC-3'; the dimorphic nucleotide is underlined). HaeII digestion products were then electrophoresed on a 2% agarose gel.
Data analysis
Phylogeny reconstructions were performed using PAUP* 4.0.1b3a (Swofford, 1999
). For parsimony analysis, indels of one nucleotide or longer were defined as missing data, based on arguments discussed by Wojciechoski et al. (1993); these indels were later superimposed onto trees in order to exploit their possible phylogenetic utility. Ambiguous nucleotides (e.g., divergent paralogs) were defined as polymorphisms. Branch and bound searches using unweighted (Fitch) parsimony and accelerated transformation (ACCTRAN) of character state optimization were performed to identify the most parsimonious trees. Branches of zero length were collapsed. For neighbor-joining analysis (Saitou and Nei, 1987
), Kimura two-parameter distances (Kimura, 1980
) and a minimum evolution objective function were employed, and ambiguous and "missing" data (e.g., indels and divergent paralogs) were ignored. The ITS sequences from Sinapus alba L. (Brassicaceae) (Yang et al., 1999
) and chloroplast trnL intron sequences from Draba tomentosa (Brassicaceae) haplotype DT-2 (Widmer and Baltisberger, 1999
) were used as outgroups for the appropriate phylogenetic analyses. Relative support for various clades was determined by bootstrap analysis (Felsenstein, 1985
) employing 200 replicates. MacClade 3.05 (Maddison and Maddison, 1993
) was used as an aid to visually search for phylogenetically incongruent "blocks" of sequence containing two or more variable sites within a given taxa or clade that might constitute evidence of prior hybridization events followed by recombination between divergent paralogs (Baldwin et al., 1995
). These searches were performed using the "tree window display," with the "character trace" option active. MacClade 3.05 was also used for presentation of phylogenetic trait maps. A global tree-based relative rate test was performed by the likelihood method of Felsenstein (1988, 1993)
.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
) by digestion of the original PCR product with HaeII, which resulted in partial (50%) cleavage at the affected site, while ST9 (and other positive controls) showed complete digestion (not shown). Similarly, Streptanthus glandulosus ssp. pulchellus accession SGP1 showed distinct G/A heterozygous profiles at nucleotide 139 of ITS2. However, the latter dimorphism could not be confirmed by the CAPS method due to the absence of an affected restriction enzyme recognition site. No other ambiguous nucleotides were observed.
Phylogenetic analysis based on ITS sequences
Parsimony analysis using the branch and bound method yielded ten equally parsimonious trees of 154 steps. Alternative most-parsimonious trees differed only with respect to the relative placement of Streptanthus campestris and in the arrangement of terminal (intraspecific) taxa. The most parsimonious trees had a consistency index of 0.822 (0.710 excluding uninformative characters), a retention index of 0.860, and a homoplasy index of 0.184 (0.290 excluding uninformative characters). For comparison, neighbor-joining analysis was performed using Kimura two-parameter distances. Neighbor-joining yielded a tree that differed from a 50% majority-rule consensus parsimony tree only in the topological arrangement of a few terminal, intraspecific taxa. One of ten most parsimonious trees, which had identical topology to the neighbor-joining tree, is depicted in Fig. 2. This tree, along with 50% majority rule and strict consensus parsimony trees, indicated that Streptanthus and Caulanthus were each nonmonophyletic.
|
from chloroplast restriction site and ITS sequence analysis of the Streptanthus glandulosus species complex and related taxa. A possible sister clade to this group (bootstrap value = 67) included species from Streptanthus, Caulanthus, and Guillenia. Within this postulated sister clade, several topological relationships were poorly supported, possibly indicating an episode of rapid divergence with "lineage sorting" (Wendel and Doyle, 1998
). The morphologically distinctive "desert candle" C. inflatus fell to a basal position within this postulated sister clade. Caulanthus amplexicaulis var. barbarae, C. amplexicaulis var. amplexicaulis, and S. tortuosus formed a distinct subclade supported by three synapomorphic nucleotide substitutions, indicating that the largely nonserpentine S. tortuosus species complex is a likely sister group to C. amplexicaulis. Caulanthus amplexicaulus var. barbarae and var. amplexicaulis formed a monophyletic subclade (bootstrap value = 93) supported by two synapomorphous indels, although they showed substantial divergence within this subclade. Other likely members of the postulated "non-Euclesian" clade include the diminutive "Tamalpais Jewelflower" S. batrachopus, a highly restricted serpentine endemic, and S. campestris, a nonserpentine taxon with a wide distribution in the Transverse and Peninsular Ranges of southern California. A poorly supported subclade (bootstrap value = 31), that was nonetheless present in all most-parsimonious trees, included C. heterophyllus, a fire-following annual species from southern California, S. bracteatus, a biennial limestone endemic from central Texas, and the common desert annual Guillenia lasiophyllum.
Visual examination with the aid of MacClade 3.05 was used to search for phylogenetically incongruent "blocks" (two or more variable sites) of sequence, potentially indicative of recombination between divergent paralogues; no such blocks were observed. However, within C. amplexicaulis var. barbarae there was a block containing three variable nucleotides at positions 238 and 240 of ITS1 and position 28 of ITS2, which were not shared with C. amplexicaulis var. amplexicaulis. One of these nucleotides, a T at position 238 of ITS1, was shared with representatives of the Euclesia clade. We considered this block to be potentially significant since reticulation might explain the apparent branch asymmetry within the C. amplexicaulis clade (Fig. 2). However the nucleotide characters at positions 240 of ITS1 and 28 of ITS2 were not shared by any of the streptanthoid taxa in this study or those reported by Mayer and Soltis (1999)
. Furthermore, the intervening 212 nucleotides (with 11 variable positions), as well as the immediately surrounding nucleotides (also with several variable positions), were phylogenetically consistent with the remainder of the C. amplexicaulis ITS sequence and did not share any variable characters with the Euclesia clade. It is therefore likely that all three of the nucleotide characters in question are derived within the C. amplexicaulis var. barbarae clade rather than the result of reticulation followed by recombination.
Chloroplast trnL sequence analysis
Prior to our analysis of the ITS sequence data, we had the expectation that our study would encompass several distinct genera, including Guillenia as a likely outgroup. For this reason, chloroplast trnL intronic sequences were included as a set of phylogenetic characters that would be useful for comparisons at the intergeneric level. The trnL introns were amplified from 19 accessions, yielding ±379 nucleotides of unambiguous double-stranded DNA sequence information. However, trnL intron sequences had a very limited number of variable sites within the taxa examined. There were only five nucleotide substitutions, two of which were phylogenetically informative, and one autapomorphous indel. The lack of informative sites yielded parsimony and neighbor-joining trees that were fraught with zero-branch length polytomies (not shown). However, a unique and informative trnL haplotype arising from a single nucleotide substitution was synapomorphous within the C. amplexicaulis group. Another unique haplotype was synapomorphous within the Euclesia serpentine endemics examined in this study. Given the low level of divergence seen in the trnL data set, the probability that either of these apparent synapomorphies is actually homoplastic is quite remote. Thus, trees generated by trnL sequence analysis were fully consistent with ITS trees, with unique haplotypes supporting some of the key ITS-derived topological arrangements (Fig. 2).
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
). The application of DNA-based characters will play a significant role in the clarification of taxonomic and phylogenetic relationships within this group. In our limited analysis, the genera of Caulanthus and Streptanthus appear to be nonmonophyletic (for example, a minimum of five additional steps is required to produce a tree that places Guillenia outside of the clades represented by Caulanthus and Streptanthus). Although much of the structure of our tree lacks bootstrap support, this general finding holds true even if several weakly supported branches in our tree are statistical artifacts. The present study contributes to our understanding of the phylogenetic relationships within this group, through cladistic (parsimony) and distance analyses. However, further work in this direction should be undertaken through comprehensive examination of molecular and morphological evidence from all potential members of the streptanthoid group, including Streptanthella, the remainder of the Guillenia species, and the Streptanthus species of the southwest (Arizona, New Mexico, and west Texas) and southern plains (Texas, Oklahoma, and Arkansas) regions of the United States and of northern Mexico (Rollins, 1993
).
Phylogenetic perspectives on the evolution of serpentine tolerance
Although the Brassicaceae family only constitutes 4% of the native taxa of California, it contributes nearly 13% of the serpentine taxa (Kruckeberg, 1984
; Hickman, 1993
). Well represented in the California serpentine floras are the genera of Streptanthus (±16 taxa) and Arabis (five taxa). Other brassicaceous genera of the CFP with at least one serpentine species include Cardimine, Caulanthus, Guillenia, Erysimum, and Thalspi (Hickman, 1993
; Rollins, 1993). However, tolerance to serpentine environments is not universal within the Brassicaceae in general, or within Arabis and the Streptanthoid taxa in particular. Among numerous examples, Kruckeberg (1951, 1954)
used experimental studies to determine that certain Brassica species were intolerant to serpentine soil, as were several races of S. glandulosus from nonserpentine locations. In the species complex S. tortuosus Kellogg, an apparent sister group to C. amplexicaulis (Fig. 2), the bulk of the taxa are annuals that are intolerant to serpentine (Kruckeberg, 1984
). One uncommon biennial or perennial subspecies, S. tortuosus ssp. suffrutescens (E. Greene) Jepson (in part S. tortuosus ssp. orbatus Jepson), found on serpentine in central California (Kruckeberg, 1984
; Hickman, 1993
), needs further study. In the geologically complex Trinity-Klamath-Siskiyou region of extreme northern California and southern Oregon, we observed that populations of S. tortuosus were excluded on a fine scale from patches of serpentine soil. Considered together, these results indicate that tolerance to serpentine soils is determined by genetic differentiation among various lineages of the Brassicaceae and within the streptanthoid group itself.
For phylogenetic trait analysis (Fig. 3), we somewhat simplistically designated the taxa in our study as either "serpentine" or "nonserpentine." In the absence of comprehensive side-by-side serpentine tolerance tests (e.g., Kruckeberg, 1954
), these assignments were based on the habitats in which the specimens were collected and were in all cases in agreement with the available literature (including Kruckeberg, 1984
; Hickman, 1993
; Reeves, 1993; Mayer and Soltis, 1994
; Mayer, Soltis, and Soltis, 1994
). When superimposed onto a semistrict consensus parsimony tree, serpentine habit (Fig. 3) and endemism were clearly nonmonophyletic in nature. These results support an earlier finding of nonmonophyly of serpantinism in a study of chloroplast restriction site polymorphism-based phylogenies in S. glandulosus and related taxa that included the nonserpentine taxa S. tortuosus and S. hispidus (Mayer and Soltis, 1994
). In our study, the serpentine taxa emerged in at least three distinct phylogenetic clades. Given that the serpentine substrate is relatively rare (occupying only 1% of the CFP land mass; Raven and Axelrod, 1978
; Kruckeberg, 1984
), it would be prudent to assume that the ancestral or "primitive" state of the Streptanthoid Complex is intolerance to serpentine. With this assumption, one must surmise that "serpantinism" probably arose independently on several occasions. Given that all 16 species of the subgenus Euclesia have at least one intraspecific taxa on serpentine and that most of these species are in fact restricted to serpentine (Kruckeberg, 1951
), serpentine tolerance may very well be the ancestral state of the Euclesia.
|
). Finally, we searched for phylogenetically incongruous blocks of divergent ITS sequence, indicative of past recombination between divergent paralogs. No compelling evidence of recombination between divergent ITS paralogs was observed. Therefore, reticulation appears to have played a limited role in the evolution of C. amplexicaulis and closely related taxa.
Given that reticulation does not adequately explain the pattern of serpentine tolerance in the streptanthoid taxa, then the trait was apparently independently gained (or lost) multiple times. It is tempting to hypothesize the existence of an evolutionary "predisposition" within the Brassicaceae in general, and in Arabis and the streptanthoid group in particular, that facilitates adaptation to serpentine environments when and where these are encountered. Various models for the evolution of tolerance to heavy metals (Antonovics, Bradshaw, and Turner, 1971
) and serpentine soils (Kruckeberg, 1986) stipulate the necessity of preexisting genetic variation as a requisite for microevolutionary adaptation through strong selection processes (this is variously termed "preadaptation" or "genetic potential" by Kruckeberg [1986] and others). A predisposition to the evolution of serpentine tolerance might simply be the presence of the requisite genetic variation in a given population. However, the mechanisms by which such genetic variation is maintained in specific lineages over many generations on nonserpentine soils—without apparent selection—remain to be elucidated (for example, C. amplexicaulis var. barbarae does not appear to have a close phlyogenetic relationship with any of the other serpentine taxa).
Additional possible predisposing factors need only provide a slight survival advantage on transitional soils at the margins of serpentine outcrops, as suggested by Kruckeberg (1954)
. Such contact zones could provide a temporary "stage" upon which microevolutionary forces could act to exert changes in allele frequencies that would eventually confer greatly enhanced tolerance to serpentine soils. Because this model incorporates a priori selective forces and biological processes, it is more satisfactory than drastic "catastrophic selection" models (Lewis, 1962
; Raven, 1964
; Antonovics, Bradshaw, and Turner, 1971
; Kruckeburg, 1986) for explaining that the acquisition of serpentine tolerance is largely confined to a limited and recurring set of plant families (Kruckeberg, 1984). As an example, the Brassicaceae and Caryophyllaceae families, which are both conspicuous among serpentine floras (Proctor, 1999
), are predominantly nonmycorrhizal, and would therefore not suffer a fitness cost with any loss of mycorrhizal associations due to geochemical limitations on microbial growth.
Furthermore, serpentine habitats are often open, rocky, and strongly exposed to evaporation and light. An evolutionary predisposition to evolving serpentine tolerance might take the form of adaptation to similar physical habitats that might arise on any number of geochemical substrates. In Hickman (1993)
, fully 50% of the specific and intraspecific brassicaceous taxa of California have the words "barren," "cliff," "crevice," "gravel(ly)," "outcrop," "rock(y)," or "talus" in their habitat description; the percentage is highest in Arabis (84%) and Streptanthus (64%). Successful adaptation to the rocky, exposed environments in geologically active and topographically complex western North America may have placed some elements of the Brassicaceae in an excellent position to take advantage of the refugia from competition afforded by serpentine habitats.
The uptake of nickel and its sequestration in plant tissues has also been postulated to enhance tolerance to serpentine soils in certain plant taxa (Brooks, 1987
). About half of the known nickel-accumulating species occur in the Brassicaceae. Reeves, Brooks, and MacFarlane (1981)
found that some herbarium specimens of C. amplexicaulis var. amplexicaulis, S. tortuosus, and C. inflatus from nonserpentine habitats had nickel levels in their tissues (up to 95, 14, and 47 µg/g dry tissue, respectively) that were comparable to the levels observed in serpentine-tolerant Streptanthus species sampled from serpentine sites (where underlying soil nickel concentrations were 10- to 100-fold higher). These data suggest that nonserpentine taxa C. amplexicaulis var. amplexicaulis, S. tortuosus, and C. inflatus may have some genetic capacity to sequester nickel in their tissues. If this capacity were characteristic of members of the postulated clade containing these species (Fig. 2), it may have facilitated the evolution of serpentine tolerance in C. amplexicaulis var. barbarae, S. batrachopus, and possibly S. tortuosus ssp. suffrutescens (to our knowledge, none of these latter taxa have yet been tested for nickel accumulation). Further, Mayer and Soltis (1994)
provided strong evidence based on chloroplast restriction endonuclease fragment profiles that S. polygaloides Gray, the only known nickel hyperaccumulator in the streptanthoid group (Reeves, Brooks, and MacFarlane, 1981), has a very close phylogenetic affinity to S. tortuosus, and is therefore likely to be another member of this hypothetical non-Euclesian clade.
Given the possibility of long-term survival in marginal serpentine zones, there would be sufficient time for microevolutionary processes and sexual recombination to produce genotypes with allelic combinations well adapted to serpentine environments. Indeed, there may even be sufficient time for new mutations to occur, thus resolving the need for preexisting genetic variation. Given the apparent frequency with which serpentine tolerance may be gained (or lost) within some evolutionary clades (Fig. 3), it seems plausible that allelic changes in a limited number of genes may be sufficient for adaptation to serpentine. Alternatively, the streptanthoid group may have evolved in a very dynamic selective context—with strong edaphic selective pressures. It should be noted that models that evoke the need for preexisting genetic variation, including catastrophic selection models, also favor the involvement of a limited number of loci.
Intraspecific ITS sequence divergence in C. amplexicaulis and related taxa
G. lasiophyllum, C. heterophyllus, C. amplexicaulis, and S. tortuosus have similar life histories, floral morphology, and seed dispersal mechanisms. The pollination biology of these species needs further study, but all were self-fertile when hand-pollinated (Preston, 1991
; A. E. Pepper and L. E. Norwood, unpublished data). Species within this group are thus well suited for comparative analyses, including comparisons of intraspecific DNA sequence variation. As indicated in Table 2, C. amplexicaulis var. barbarae and C. amplexicaulis var. amplexicaulis populations sampled 75–105 km apart were divergent at six to eight nucleotide positions in the ITS1 and ITS2 regions (1.2–1.8%). In contrast, four S. tortuosus accessions sampled over a transect nearly 300 km long, from the Sierra Nevada foothills near Sacramento, California, to the Siskiyou Mountains on the Oregon border, showed at most two nucleotides substitutions (including the dimorphic accession ST2). Further, the morphologically distinctive subspecies C. heterophyllus ssp. heterophyllus and C. heterophyllus ssp. pseudosimulans, obtained from locations 127 km apart, had only two divergent nucleotides. Finally, G. lasiophyllum populations sampled from Sonoran Desert (Yaqui Well) and Mojave Desert (Blackwater Well) stations separated by 260 km and two major (3000 m) mountain ranges had identical ITS sequences.
|
). As a third alternative, recurrent genetic bottlenecks—as might be expected in a group of small isolated populations—could lead to an increased frequency of the fixation of rare ITS alleles. Further studies are needed to ascertain the relative contributions of sampling, time, and isolation, and population dynamics to the apparent genetic divergence of C. amplexicaulis var. barbarae from C. amplexicaulis var. amplexicaulis. C. amplexicaulis var. barbarae has an exceedingly limited distribution and under typical circumstances has very small population numbers (N < 50). Several of the populations are potentially threatened by road building and maintenance, off-road vehicle activity, mining, and other pressures. Given the possibility of substantial genetic differentiation from its sister taxon C. amplexicaulis var. amplexicaulis—as evidenced by the data presented here—it should be granted a conservation status appropriate to an ecologically and phylogenetically distinct evolutionary unit.
A model for the evolution of C. amplexicaulis var. barbarae
Using a global tree-based maximum likelihood test (Felsenstein, 1988, 1993
), a clock-like model for ITS divergence within the streptanthoid taxa could not be excluded [–2(logLclock – logLno clock) = –34.1, P = 0.10]. However, local branch-length asymmetry along with a paucity of fossil records and biogeographic events for calibrating molecular clocks conspire to make any estimates of the time of divergence of C. amplexicaulis var. barbarae and C. amplexicaulis var. amplexicaulis highly speculative. However, if we presume that the actual ITS divergence rate within the streptanthoid group is within the range previously reported for herbaceous annual plants—from 1.45 x 10–9 to 3.62 x 10–9 substitutions·site–1·yr–1 in the Cucurbitaceae (Jobst, King, and Hembleben, 1998
) up to 7.83 x 10–9 substitutions·site–1·yr–1 in the genus Robinsonia (Sang, Crawford, and Stuessy, 1995
)—we can estimate that C. amplexicaulis var. barbarae and var. amplexicaulis diverged at or after the time that areas of serpentine in the San Rafael Mountains were exposed, during the middle Pliocene to early Quaternary (1–3.5 mya) (Raven and Axelrod, 1978
). A lack of reproductive barriers between the two extant taxa suggests that the duration of any parapatric coexistence of serpentine and nonserpentine populations must have been brief.
Furthermore, C. amplexicaulis and related taxa have only minimally winged seeds (<1 mm), and no mechanisms for long-distance dispersal. The nearest extant population of C. amplexicaulis to the San Rafael Mountains serpentine sites is at Thorne Meadows—some 75 km distant. Considered together, these findings suggest that a "biotype depletion" scenario for the origin of endemic plant species (Stebbins, 1942
) may be applicable. According to this model, Pliocene populations of C. amplexicaulis would have had a more continuous distribution than today. As serpentine became exposed, some combination of "genetic potential" (sensu Kruckeberg, 1987) and other predisposing factors present in C. amplexicaulis facilitated the evolution of adapted ecotypes on serpentine sites. An emerging Mediterranean climate, with decreasing summer rainfall (Raven and Axelrod, 1978
), would have subsequently led to the confinement of C. amplexicaulis to a few "more local environments" (Kruckeberg, 1984
) where competition was limited. Indeed, present-day C. amplexicaulis var. amplexicaulis populations at lower elevations of the Transverse Ranges and on the fringes of the Mojave Desert are found on loose, nearly barren shale or granite "scree" slopes—environments that could also be considered refugia from competition. At higher elevations (>2000 m) where moisture is more abundant, such as Frazier Mountain and Mount Piños in Ventura County, or the San Bernardino and San Gabriel Mountains, C. amplexicaulis var. amplexicaulis is often found in environments more generally suitable for plant growth, such as the "yellow pine" forest floor. With geographic isolation resulting from biotype depletion, as is suggested by our intraspecific ITS sequence divergence data, reproductive isolation would not have been favored by selection. As a result, C. amplexicaulis var. barbarae and C. amplexicaulis var. amplexicaulis remained interfertile—as appears to be the "default" situation among several widely divergent taxa within the streptanthoid group, including C. amplexicaulis, C. inflatus, C. heterophyllus, and G. lasiophyllus, which are interfertile in hand-pollination experiments (A. E. Pepper and L. E. Norwood, unpublished data).
Further experiments, using defined growth conditions, will determine the phenotypic and genetic components of serpentine tolerance in the two C. amplexicaulis varieties. Since the obvious morphological differences between the two taxa are minimal (i.e., anthocyanin content of the perianth), adaptive differences are likely to be physiological or biochemical in nature or may involve root development. Determination of these adaptive genetic differences will permit the further development of a comprehensive model for the evolution of these two divergent taxa. Caulanthus amplexicaulis var. barbarae and other members of the streptanthoid group provide fascinating and experimentally tractable subjects for the study of serpentine tolerance and endemism. While the exact sequence of events in the evolution of C. amplexicaulis var. barbarae—or any other taxon in nature—will never be precisely known, the development of testable ecological and evolutionary hypotheses, combined with the application of powerful new tools for the study of extant genetic variation, will almost certainly lead to an enhanced understanding of the processes of natural selection and the formation of new plant species.
|
|
FOOTNOTES |
|---|
2 Author for reprint requests (Fax: 979-862-4790; apepper{at}bio.tamu.edu
).
|
LITERATURE CITED |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
Baldwin B. G. M. J. Sanderson J. M. Porter M. Wojciechowski C. Campbell M. Donoghue 1995 The ITS region of nuclear ribosomal DNA: a valuable source of evidence on angiosperm phylogeny. Annals of the Missouri Botanical Garden 82: 247-277[CrossRef][ISI]
Beauchamp M. 1986 A flora of San Diego County, California. Sweetwater River Press, National City, California, USA
Brooks R. R. 1987 Serpentine and its vegetation, a multidisciplinary approach. Dioscorides, Portland, Oregon, USA
Dieringer G. 1991 Pollination ecology of Streptanthus bracteatus (Brassicaceae): a rare central Texas endemic. Southwestern Naturalist 36: 341-343[CrossRef][ISI]
Felsenstein J. 1985 Confidence limits on phylogenies: an approach using the bootstrap. Evolution 39: 783-791[CrossRef][ISI]
———. 1988 Phylogenies from molecular sequences: inference and reliability. Annual Review of Genetics 22: 521-565[CrossRef][ISI][Medline]
———. 1993 PHYLIP: phylogenetic inference package. University of Washington, Seattle, Washington, USA
Hickman J. C. 1993 The Jepson manual: higher plants of California. University of California Press, Berkeley, California, USA
Howell J. 1962 New western plants IV. Leaflets of Western Botany 9: 223-224
Jobst J. K. King V. Hemleben 1998 Molecular evolution of the internal transcribed spacers (ITS1 and ITS2) and phylogenetic relationships among species of the family Cucurbitaceae. Molecular Phylogenetics and Evolution 9: 204-219[CrossRef][ISI][Medline]
Kimura M. 1980 A simple method for estimating evolutionary rate of base substitutions through comparative studies of nucleotide sequences. Journal of Molecular Evolution 16: 111-120[CrossRef][ISI][Medline]
Konieczny A. F. M. Ausubel 1993 A procedure for mapping Arabidopsis mutations using co-dominant ecotype specific PCR-markers. Plant Journal 4: 403-410[CrossRef][ISI][Medline]
Kruckeberg A. R. 1951 Intraspecific variability in the response of certain native plant species to serpentine soil. American Journal of Botany 38: 408-419[CrossRef][ISI]
———. 1954 The ecology of serpentine soils III. Plant species in relation to serpentine soils. Ecology 35: 267-274[ISI]
———. 1957 Variation in fertility of hybrids between isolated populations of serpentine species, Streptanthus glandulosus Hook. Evolution 11: 185-211[CrossRef][ISI]
———. 1958 Taxonomy of the species complex Streptanthus glandulosus Hook. Madroño 14: 217-248
———. 1984 California serpentines: flora, gology and management problems. Univerisity of California Press, Berkeley, California, USA
———. 1986 An essay: the stimulus of unusual geologies for plant speciation. Systematic Botany 11: 455-463[CrossRef][ISI]
Lewis H. 1962 Catastrophic selection as a factor in speciation. Evolution 16: 257-271[CrossRef][ISI]
Maddison W. P. D. R. Maddison 1993 MacClade: analysis of phylogeny and character evolution. Sinauer, Sunderland, Massachusetts, USA
Mayer M. S. P. S. Soltis 1994 The evolution of serpentine endemics: a cpDNA phylogeny of Streptanthus glandulosus complex (Cruciferae). Systematic Botany 19: 557-574[CrossRef][ISI]
———, and ———. 1999 Intraspecific phylogeny analysis using ITS sequences: insights from studies of the Streptanthus glandulosus complex (Cruciferae). Systematic Botany 24: 47-61[CrossRef][ISI]
———, ———, andD. E. Soltis 1994 The evolution of the Streptanthus glandulosus complex (Cruciferace): genetic divergence and gene flow in serpentine endemics. American Journal of Botany 81: 1288-1299[CrossRef][ISI]
Morrison J. L. 1941 A monograph of the section Euclesia Nutt., of Streptanthus Nutt. Ph.D. dissertation, University of California, Berkeley, California, USA
Preston R. E. 1991 The intrafloral phenology of Streptanthus tortuosus (Brassicaceae). American Journal of Botany 78: 1044-1053[CrossRef][ISI]
Proctor J. 1999 Toxins, nutrient shortages and droughts: the serpentine challenge. Trends in Ecology and Evolution 14: 334-335[CrossRef]
———, andS. R. J. Woodell 1975 The ecology of serpentine soils. Advances in Ecological Research 9: 255-365
Raven P. H. 1964 Catastrophic selection and edaphic endemism. Evolution 18: 336-338[CrossRef][ISI]
———, andD. I. Axelrod 1978 Origins and relationships of the California flora. University of California Publications in Botany, vol. 72. University of California Press, Berkeley, California, USA
Reeves D. R. R. R. Brooks R. M. MacFarlane 1981 Nickel uptake by Californian Streptanthus and Caulanthus with particular reference to the hyperaccumulator S. polygaloides Gray (Brassicaceae). American Journal of Botany 68: 708-712[CrossRef][ISI]
Rollins R. C. 1993 The Cruciferae of continental North America: systematics of the mustard family from the Arctic to Panama. Stanford University Press, Stanford, California, USA
Rosenthal A. O. Coutelle M. Craxton 1993 Large-scale production of DNA sequencing templates by microtitre format PCR. Nucleic Acids Research 21: 173-174
Saitou M. N. Nei 1987 The neighbor joining method: a new method for reconstructing phylogenetic trees. Molecular Biology and Evolution 4: 406-425[Abstract]
Sang T. D. J. Crawford T. F. Stuessy 1995 ITS sequences and phylogeny of the genus Robinsonia (Asteraceae). Systematic Botany 20: 55-64
Skinner M. W. B. M. Pavlik 1994 Inventory of rare and endangered vascular plants of California. California Native Plant Society Press, Sacramento, California, USA
Stebbins G. L. 1942 The genetic approach to problems of rare and endemic species. Madroño 6: 241-272
Swofford D. L. 1999 PAUP*: phylogenetic analysis using parsimony (*and other methods), version 4. Sinauer, Sunderland, Massachusetts, USA
Taberlet P. L. Gielly G. Pautou J. Bouvet 1991 Universal primers for amplification of three non-coding regions of chloroplast DNA. Plant Molecular Biology 17: 1105-1109[CrossRef][ISI][Medline]
Walker R. B. 1948 Factors affecting plant growth on serpentine soils. Ecology 35: 259-266[ISI]
Wendel J. F. J. J. Doyle 1998 Phylogenetic incongruence: window into genome history and molecular evolution. In D. E. Soltis, P. S. Soltis, and J. J. Doyle [eds.], Molecular systematics of plants II, DNA sequencing, 265–296. Kluwer Academic, Boston, Massachusetts, USA
White T. J. T. Bruns S. Lee J. Taylor 1990 Amplification and direct sequencing of fungal ribosomal RNA genes for phylogenetics. In M. Innis, J. J. Gelfand, J. J. Sninsky, and T. J. White [eds.], PCR protocols, 315–322. Academic Press, New York, New York, USA
Widmer A. M. Baltisberger 1999 Molecular evidence for allopolyploid speciation and a single origin of the narrow endemic Draba ladina (Brassicaceae). American Journal of Botany 86: 1282-1289
Wojciechowski M. F. M. J. Sanderson B. G. Baldwin M. J. Donoghue 1993 Monophyly of Aneuploid Astragalus (Fabaceae)—evidence from nuclear ribosomal internal transcribed spacer sequences. American Journal of Botany 80: 711-722[CrossRef][ISI]
Yang Y.-W. K.-N. Lai P.-Y. Tai D.-P. Ma W.-H. Li 1999 Molecular phylogenetic studies of Brassica, Rorippa, Arabidopsis, and allied genera based on the internal transcribed spacer region of 18S–25S rDNA. Molecular Phylogenetics and Evolution 13: 455-462[CrossRef][ISI][Medline]
This article has been cited by other articles:
|
|
 |
|
  J. S. JOHNSTON, A. E. PEPPER, A. E. HALL, Z. J. CHEN, G. HODNETT, J. DRABEK, R. LOPEZ, and H. J. PRICE Evolution of Genome Size in Brassicaceae Ann. Bot., January 1, 2005; 95(1): 229 - 235. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
