| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Department of Botany and Microbiology, University of Oklahoma, Norman, Oklahoma 73019
Received for publication January 12, 1998. Accepted for publication March 8, 1999.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
Key Words: biogeography • Chelone • evolution • isozymes • phenetics • polyploidy • Scrophulariaceae
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
; Werth, Guttman, and Eshbaugh, 1985a, b
; Soltis and Soltis, 1989, 1993
; Brochmann, Soltis, and Soltis, 1992
; Wyatt, Odrzykoski, and Stoneburner, 1992
). These studies generally supported hypotheses that multiple and polytopic origins of polyploids were not uncommon and that diverse and multiple origins contributed to the genetic heterogeneity and evolutionary potential of polyploids (Soltis, Doyle, and Soltis, 1992
; Soltis and Soltis, 1991, 1993
). Previous research also demonstrated that the potential for hybridization and polyploidy was influenced greatly by ecohistorical factors, such as the pronounced geoclimatic changes and subsequent plant migrations during the Quaternary Period (Ehrendorfer, 1959
; Stebbins, 1971
). In eastern North America, the repeated expansion, contraction, fragmentation, and amalgamation of species' ranges during the Pleistocene and Holocene epochs, which began two million years ago (Cain, 1944
; Holt and Paterson, 1970
; Jacobs, Werth, and Guttman, 1984
; Delcourt and Delcourt, 1987
; Delcourt, 1991
), undoubtedly resulted in new contacts between previously isolated populations (Qiu and Parks, 1994
; Kuser et al., 1997
), increased the probability of interlineage hybridization and introgression (Parks et al., 1994
; Parker et al., 1997
), and led to extinctions of progenitor taxa (Werth, 1991
).
Chelone is a useful model for investigating the pattern and process of hybrid and polyploid evolution among eastern North American plants. The genus Chelone is a morphologically distinctive group of perennial herbs whose monophyly is defined by a galeate corolla, subspicate inflorescence, a bracteate inflorescence, and a reticulate-rugulate pollen exine surface (Nelson, 1995
). It comprises three diploid (2n = 28) species and one species complex with tetraploid and hexaploid races (Nelson, 1995
), is subdivided into a series of morphological and geographic variants (Pennell, 1935
), has several purported examples of inter- and intraspecific hybridization (Fernald, 1950
; Radford, Ahles, and Bell, 1968
; Bittman, 1980
), and is distributed both north and south of the boundary of the last glacial maximum.
Species and varietal determinations in Chelone are notoriously difficult because of inconsistency and intergradation among characters (Pennell, 1935
; Gleason, 1952
; Crosswhite, 1965
; Radford, Ahles, and Bell, 1968
). Corolla color, beard color, staminode tip color, petiole length, and leaf base have been used to delimit species in Chelone (Nelson, 1995
). Chelone glabra has some white in the corolla and beard, a green-tipped staminode, and cuneate leaf bases. Both C. cuthbertii and C. lyonii have purple corollas, yellow beards, and rounded leaf bases. Whereas C. cuthbertii has sessile leaves and a purple-tipped staminode, C. lyonii has petiolate leaves and a white to pink-tipped staminode. The polyploid C. obliqua has a distinctive combination of morphological characters, but individual characters are found in one or more of the diploid species.
Two species, C. glabra and C. obliqua, have been subdivided into seven and three varieties, respectively (Pennell and Wherry, 1928
; Pennell, 1935
) (Table 1). Chelone cuthbertii, which is disjunct between the southern Blue Ridge region of North Carolina and the coastal plain of Virginia, is exclusively diploid (2n = 28) (Pennell, 1935
; Nelson, 1995
) as is C. lyonii, a southern Blue Ridge endemic that occurs in Tennessee, North Carolina, and South Carolina. Chelone glabra, ranging from the Carolina coast to Iowa and from Mississippi to Newfoundland, is diploid with one reported tetraploid population (Pennell, 1935
; Cooperrider and McCready, 1970
; Nelson, 1995
). Populations of C. obliqua are either tetraploid endemics of the southern Blue Ridge or occur as hexaploids in the Interior Highlands and Plains region (Arkansas to Michigan to Tennessee) or the Atlantic Coastal Plain from South Carolina to Maryland (Pennell, 1935
; Cooperrider and McCready, 1970
; Nelson, 1995
). Chelone glabra is the most widespread and polymorphic species in the genus; several of its varieties are hypothesized to hybridize with other diploid and polyploid taxa of the genus (Radford, Ahles, and Bell, 1968
; Bittman, 1980
).
|
inferred an autoploid origin of C. obliqua tetraploids (var. erwiniae) from C. glabra, an alloploid origin of the hexaploids of C. obliqua (vars. obliqua and speciosa) between C. lyonii and tetraploid populations of C. obliqua, and possible interspecific hybridization between C. lyonii and C. glabra.
We initiated this investigation to quantify the amount and pattern of variation for isozymes and morphological characters within the genus, to assess the species and varietal delimitations of Pennell (1935)
, to test hypotheses of hybridization and polyploid origins proposed by Bittman (1980)
, and to examine the effect of historical biogeographic factors on the pattern and process of evolution in the genus.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
) and in as many physiographic regions (Fenneman, 1938
; Brouillet and Whetstone, 1993
) as possible (Fig. 1). Population sizes of Chelone were generally less than ten individuals, although a few populations greater than 100 individuals were observed. Sample sizes for each taxon and population also are indicated in Table 1. For isozyme studies, leaf material was collected from individual plants in the field and stored on ice until used. Populations 2, 3, 7, 22, 40, 41, 50–52, 54–57, and 65 contained some individuals grown from seed collections. Vouchers were deposited at the Robert Bebb Herbarium (OKL) at the University of Oklahoma.
|
; Pennell, 1935
), whereas characters 9, 11, 14, and 16 had not been used previously. In addition, a number of characters designated qualitatively in previous studies (Pennell and Wherry, 1928
; Pennell, 1935
) were quantified in the present study. We obtained a minimum of ten measurements for each quantitative character for populations except when material was limited by number of herbarium sheets or small population sizes. Characters 1, 2, 7, 9, 11, and 14 were scaled using ratios to eliminate size differences that might result from environmental factors. NTSYS-pc (Rohlf, 1990
) was used to standardize a data matrix of populational averages for each character (available upon request from the first author), calculate a dissimilarity matrix using the average taxonomic distance coefficient, compute a cophenetic correlation coefficient, and generate a phenogram with the unweighted pair-group method using an arithmetic average (UPGMA) (Fig. 2). A correlation matrix of characters was constructed to calculate eigenvectors used in principal components analysis (PCA). Three-dimensional models of populations (Fig. 3) were constructed and used to analyze characters involved in groupings.
|
|
|
). A Canonical Variance Analysis (CVA) (Dixon and Brown, 1979
) was done to determine which combination of characters provided maximal discrimination among species, relative to the variation within species. Both a standard classification and a jackknifed classification (Dixon and Brown, 1979
) were produced in the CVA to test group relationships. All populations were included in analyses using UPGMA clustering, PCA, CVA, and ANOVA. UPGMA cluster analyses were done with and without polyploid populations (numbers 56–66) to determine clustering patterns of diploids and polyploids.
Enzyme extraction followed established protocols (Gottlieb, 1981
; Elisens and Nelson, 1993
) except that 120 mg/mL PVP-40 was added to the extracting buffer. Leaf extracts were centrifuged in 1.5-mL tubes and the supernatant was absorbed onto wicks of Whatman 17 MM chromatography paper. Ten enzyme systems coded by 16 putative loci in diploids were resolved on 12% starch gels utilizing two buffer systems. System I was histidine citrate buffer, pH 6.5 (Soltis et al., 1983
) with a gel buffer obtained from a 1 : 3 aqueous dilution of the electrode buffer and system II was tris-EDTA buffer (Soltis et al., 1983
) with a gel buffer obtained from a 1 : 3 aqueous dilution of the electrode buffer. System I was used to resolve one locus each for [NADP] glyceraldehyde 3-phosphate dehydrogenase (G3pd), 6-phosphogluconate dehydrogenase (6Pgd), and menadione reductase (Mnr). System II was employed to resolve one locus each for alcohol dehydrogenase (Adh), fluorescent esterase (Fest), glucose dehydrogenase (Gdh), and malic enzyme (Me), two loci for triosephosphate isomerase (Tpi), three loci for phosphoglucoisomerase (Pgi), and four loci for phosphoglucomutase (Pgm). Enzyme-activity staining protocols and agarose overlays generally followed Soltis et al. (1983)
and Wendel and Weeden (1989)
, except for Mnr, which was stained using a modification of Conkle et al. (1982)
. Loci were numbered consecutively beginning with the most anodal form. Alleles at each locus were labeled alphabetically from the most anodal form.
Allele and genotype frequencies were determined for each diploid species and allele presence or fixation was noted for polyploids of C. obliqua (Table 3). Three pairs of putative duplicated loci were observed (Pgi-1 and -2, Pgm-1 and -2, and Pgm-3 and -4) but were not included in quantitative analyses. Dosage effects were apparent for duplicated loci in polyploid taxa, but polyploid genotypes were not determined. The GENESTAT2 program (Whitkus, 1988
) was used to calculate Nei's (1972)
genetic identity and distance coefficients (Table 4), and gene diversity statistics within and among diploid species and varieties (Table 5), and for physiographic regions (Table 5). The BIOSYS-1 program (Swofford and Selander, 1981
) was used to calculate Nei's (1972)
genetic distance and to generate UPGMA phenograms (Sneath and Sokal, 1973
) (Fig. 4). Values for the estimated number of migrants (Nm) were calculated from GST (Wright, 1951
), which was based on patterns among common alleles. Alleles found among polyploids that were present in only a few diploid populations were tabulated (Table 6) and mapped (Fig. 5). The clustering pattern of populations on the isozyme UPGMA (Fig. 4, clusters I and II) was mapped to illustrate geographic location of populations, values for total genetic diversity (HT), and relation to the approximate southern boundary of the Wisconsinian glacial maximum (Fig. 6).
|
|
|
|
|
|
|
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
or others. Results of ANOVA analyses showed significant variation (P < 0.05) among species for the morphological characters 1–3, 8, 11, 13, and 15–16 (Table 2). The most useful characters (and their numbers) for discriminating among species using CVA were: lower lip width-to-length ratio (character 11), sepal pubescence (8), staminode-to-filament length ratio (14), staminode color (15), beard color (13), leaf base (3), and petiole length (5). Using these characters, it was possible to separate 100% of the populations in each of the species' groups. In the jackknifed classification, 97% of the four species groups were discriminated. When 12 groups were designated (corresponding to C. cuthbertii, seven named varieties of C. glabra, C. lyonii, and three named varieties of C. obliqua), upper corolla and lip color (12), sepal pubescence (8), beard color (13), leaf base (3), and petiole length (5) discriminated 66.7% of the populations (45.5% in the jackknifed classification). Means for quantitative characters in CVA overlapped among all intraspecific taxa in C. glabra and C. obliqua.
Sixteen loci, coding for ten putative enzymes, were scored: one locus each for Adh, Fest, [NADP] G3pd, Gdh, Mnr, 6Pgd, Me, two for Tpi, three for Pgi, and four for Pgm. The number of isozymes detected was similar to those reported for diploid plants (Wendel and Weeden, 1989
; Murphy et al., 1996
) except for Pgi and Pgm. These loci exhibited isozyme profiles indicating duplicated loci in both diploid and polyploid populations. The apparent sets of duplicated loci were not included in quantitative analyses but, because Pgm-1 was polymorphic, it was compared among populations (Table 6; Fig. 5).
Summary allele frequencies for ten polymorphic loci among diploid species were calculated and compiled in Table 3. Allele and genotype frequencies for each population are available from the first author. All species shared the same highest frequency allele at all but two loci, G3pd-1 and Gdh-1. Although each diploid species had unique alleles (Table 3: C. cuthbertii and C. glabra had one, C. lyonii had six), all occurred at low frequencies and were confined to one or two populations. Polyploids had no unique alleles. Several low-frequency alleles (summary frequencies <0.21) were shared between two species: one between C. cuthbertii and C. lyonii, two between C. cuthbertii and C. glabra, and four between C. glabra and C. lyonii. Patterns that appeared to be fixed heterozygosity occurred in all 11 polyploid populations at Pgi-3, in ten populations for 6Pgd-1, in five populations for Adh-1, in four populations each for Gdh-1 and Mnr-1, and in one population for Tpi-3.
Five alleles had a differential occurrence among diploid and polyploid populations (Table 6; Fig. 5). Adh-1a occurred in hexaploids of C. obliqua in the Central Lowland, Interior Low Plateau, and Ozark Plateau regions and in all three diploid species. Both 6Pgd-1a and Fest-1c occurred in populations of C. glabra and C. lyonii in the southern Blue Ridge region and were both present in C. obliqua hexaploids and in C. glabra either in the Ozark Plateau (6Pgd-1a) or Interior Low Plateau regions (Fest-1c). Although the Tpi-2b allele was found only in the southern Blue Ridge area in C. glabra, C. lyonii, and tetraploids of C. obliqua; Pgm-1a was observed in the Ozark Plateau (C. glabra) and again in Blue Ridge populations of C. glabra and tetraploids of C. obliqua.
Table 4 contains coefficients for Nei's (1972)
genetic identity (I) and distance (D) within and among diploid species and varieties of Chelone. Mean I values within species were 0.848 in C. cuthbertii, 0.846 in C. glabra, and 0.767 in C. lyonii, whereas average I values among species ranged from 0.791 to 0.799. Mean I values among convarietal populations of C. glabra ranged from 0.643 to 1.000, whereas pairwise comparisons among C. cuthbertii, C. lyonii, and varieties of C. glabra ranged from 0.725 to 0.895.
UPGMA trees did not differentiate among diploid species or varieties of Chelone but did exhibit a geographic pattern. The UPGMA tree generated from genetic distances (Fig. 4) among diploid populations consisted of two primary clusters and two outliers. The two most divergent populations were found in the southern Blue Ridge and consisted of populations of C. glabra var. chlorantha and C. lyonii. Two large groups occurred that generally corresponded to populations in unglaciated (cluster I) and glaciated regions (cluster II) (Figs. 4, 6). Cluster I contained populations of C. glabra, C. cuthbertii, and C. lyonii from the Appalachian Plateau, Atlantic Coastal Plain, Piedmont/Valley and Ridge, southern Blue Ridge, and the southern Mississippi Embayment. Cluster II contained populations of C. glabra from the formerly glaciated regions of eastern New England and Superior Uplands as well as populations from the unglaciated Ozark Plateau, Atlantic Coastal Plain, Mississippi Embayment, and Central Lowland regions, and populations of C. lyonii from the southern Blue Ridge.
Genetic diversity statistics were computed for diploid taxa at 16 loci unbiased for sample size and population number (Table 5). The proportion of polymorphic loci (P), mean number of alleles per polymorphic locus (KP), and total gene diversity (HT) were consistently lowest in C. glabra and highest in C. lyonii. The estimated number of migrants exchanged between populations (Nm) was highest in C. cuthbertii and lowest in C. glabra. The southern Blue Ridge populations of C. cuthbertii had greater levels of gene diversity than populations on the Atlantic Coastal Plain, although coastal populations had higher Nm values. For populations of C. glabra, P, KP, and HT were lowest in the New England and Maritime Province and highest in the Ozark Plateau (P and KP) and Mississippi Embayment (HT) regions.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
) identified three diploid species each characterized by unique characters. Phenetic approaches such as clustering and ordination analyses (Figs. 2, 3) similarly delineated clusters of populations corresponding to three diploid species. Diploid species were delimited morphologically based on phylogenetic (sensu Nixon and Wheeler, 1990
), as well as phenetic criteria (Stuessy, 1990
).
Unlike morphological data, isozyme profiles of Chelone provided no fixed differences among diploid species. Mean genetic identity values between species (range 0.791–0.799; Table 4) were greater than the average interspecific I values of 0.67 reported by Gottlieb (1981)
and overlapped with those among conspecific populations of Chelone (range 0.440–1.00). Genetic distance-based UPGMA clustering did not discriminate clusters of populations corresponding to three primary morphotypes, but rather grouped populations according to geographic affinity (Fig. 4; evolutionary biogeography section). This pattern of relatively minor morphological differentiation accompanied by little divergence at isozyme loci is also common among diploid species from several other genera in eastern North America: Antennaria (Bayer and Crawford, 1986
), Coreopsis (Cosner and Crawford, 1990
), Heuchera (Soltis, 1985
), Marshallia (Watson, Elisens, and Estes, 1991
), Quercus (Manos and Fairbrothers, 1987
), and Sullivantia (Soltis, 1982
). Floral characters such as beard, staminode, and corolla color in the three diploid species of Chelone may be under selective pressures that provide structure in morphological analyses but do not when examining neutral molecular markers such as isozymes.
Our morphological and genetic data did not support delimitation of infraspecific taxa in C. glabra. The varieties of C. glabra delimited by Pennell (1935)
were not defined by any constant and unique morphological or isozyme characters and did not appear as consistent groups in clustering and ordination analyses. Morphological character ranges and variation overlapped extensively in comparisons between varieties defined by Pennell (1935)
and, although there were corolla color and leaf character variants in C. glabra, no consistent pattern emerged that facilitated recognition and delimitation of infraspecific taxa.
Low levels of isozyme divergence impeded elucidation of evolutionary processes responsible for observed patterns of differentiation. Testing alternative hypotheses of primary divergence vs. secondary contact and gene flow/hybridization ideally would require at least moderate levels of differences among species (Crawford, 1990
; Rieseberg, Carter, and Zona, 1990
; Wolfe and Elisens, 1993
). Nevertheless, several aspects of the isozyme data were most concordant with a recent divergence of extant diploids from a common ancestor. First, the amount of allozyme differentiation was minimal within the genus. Second, a large number of alleles (23 of 42) and highest-frequency alleles (8 of 10) were shared among diploid species (Table 3). Finally, alleles unique to species or detected in two species were few in number (8 and 11, respectively), low in frequency (16 of 19 < 0.08), and distributed generally in one to three populations (Table 3). Time and barriers to gene flow generally would be considered necessary prerequisites for genetic differentiation to occur (Crawford, 1990
; Avise, 1994
). Considering these criteria, we hypothesize that species of Chelone have diverged recently.
Although interspecific qualitative allelic differences were few, gene flow and interspecific hybridization appeared to be of limited occurrence. Alleles with discontinuous distributions generally were not clustered in contiguous populations and did not exhibit frequency or occurrence "clines" (Arnold and Bennett, 1993
). Only Me-2a had a populational occurrence that was contiguous and crossed species boundaries in C. lyonii (population 53) and C. glabra (population 21) in the southern Blue Ridge. However, isozyme markers were not robust enough to better test purported hybridization between C. glabra and C. lyonii (Bittman, 1980
). Common measures of genetic variation (e.g., P and HT) were most similar to values observed in perennials with selfing to mixed mating systems (Hamrick, 1987
; Hamrick and Godt, 1990
), a reproductive mode that promotes localized patterns of differentiation. The estimated number of migrants (Nm) exchanged between populations (all but one taxon/region 
1.41) also indicated low levels of gene flow (Table 5; Hamrick, 1987
; Ellstrand, 1992
). Because gene flow appeared to be limited among populations and species of Chelone, we suggest that the pattern and amount of isozyme divergence among species reflected local and minor variations of a widespread morphological and isozymic groundplan.
Origin and relationships of polyploids
Cooperrider and McCready (1970)
suggested a base number of x = 14 for Chelone and an euploid series with three diploid species (C. cuthbertii, C. glabra, C. lyonii; 2n = 2x = 28) and tetraploid (4x = 56) and hexaploid (6x = 84) races of C. obliqua (Nelson, 1995
). Phylogenetic analyses of morphological (Wolfe et al., unpublished data) and molecular data (Wolfe et al., 1997
) in tribe Cheloneae indicated that Chelone was most closely related to (formed an unresolved trichotomy with) western North American genera Chionophila (x = 8) and Nothochelone (x = 15). The proposed phylogenetic relationships and base chromosome number of 14 in Chelone suggested a possible ancient polyploid event in an x = 8 lineage (x = 8
16) followed by aneuploid reduction (x = 161514) (Nelson, 1995
). Observations of three pairs of duplicated loci in diploid species are concordant with hypotheses of paleopolyploidy and subsequent gene silencing rather than alternative hypotheses of three independent gene-duplication events (Haufler, 1987
; Crawford, 1990
). According to our hypothesis, a secondary cycle of polyploidy most likely resulted in tetraploid and hexaploid races and polyploid allozyme profiles observed in C. obliqua.
No unique morphological character or unique alleles delimited the C. obliqua polyploid complex. Tetraploid and hexaploid populations intergraded and overlapped morphologically with each other and appeared most similar to populations of C. glabra and C. lyonii (Fig. 2). Isozyme patterns indicated fixed heterozygosity at six loci. Even though C. obliqua was characterized by a recombinant genotype and phenotype, populations could be discriminated at a high level (85.7%) in canonical variance analyses and the complex could be delimited and identified by a unique combination of morphological characters. A purple corolla and petiolate leaves (both present in C. lyonii) that narrow to the base (present in C. glabra) characterized all 4x and 6x populations of C. obliqua. Unlike species-level analyses, the recognition of varieties in C. obliqua was not supported by morphological or allozyme data. The 4x and disjunct 6x races had allopatric ranges, but none of Pennell's (1935)
varieties of C. obliqua were defined by unique alleles or a unique set of morphological characters, and no varieties were discriminated consistently in phenetic analyses (Figs. 2, 3).
The recombinant patterns observed in C. obliqua are most concordant with hypotheses of alloploid origins for polyploid races. Among extant diploids, C. cuthbertii does not appear likely as a progenitor for tetraploids or hexaploids. No plants in 4x and 6x populations exhibit diagnostic morphological characters of C. cuthbertii and alleles at three loci lacking from isozyme profiles of C. cuthbertii are present in C. glabra, C. lyonii, and polyploids. Likely progenitors of the tetraploid race were similar to C. glabra but a contribution of plants similar to C. lyonii cannot be resolved. Our data suggested at least two independent origins for the tetraploids. Although tetraploids of C. obliqua were restricted to the southern Blue Ridge, 4x populations exhibited variation at several morphological characters, were internested among populations of C. glabra in morphological analyses (Fig. 2), and "northern" (56, 57) and "southern" populations (58) were differentiated by alternate fixed alleles at the Tpi-2 locus. The Tpi-2b allele, which characterized northern populations, was found at frequencies as high as 1.00 in populations of C. glabra from the Blue Ridge region and at low frequencies in C. lyonii. Based on leaf flavonoid profiles, Bittman (1980)
suggested autoploid origins of tetraploid populations within C. glabra. Although the isozyme pattern of northern and southern tetraploids also could have resulted from differential gene silencing, our observations of morphological variation, fixed heterozygosity at six loci, and alternate fixed alleles at Tpi-2, are more consistent with hypotheses of alloploidy and polytopy for the tetraploid race.
Morphological data showed significant variation among coastal and interior hexaploid populations and suggested close relationships with C. glabra, C. lyonii, and tetraploids. Isozyme data were more robust and indicated that extant 4x populations were not recent progenitors for 6x populations of C. obliqua. Mutually exclusive alleles at five loci, including two monomorphic loci among tetraploids, characterized isozyme profiles of tetraploid and hexaploid populations. If tetraploid races were recent progenitors of hexaploids in a pillar complex (Stebbins, 1971
), genomic additivity and the presence of alleles fixed for the tetraploids at Pgm-1 and Tpi-2 (Table 6) would be predicted among hexaploid populations (Roose and Gottlieb, 1976
; Werth, 1989
). Although it was possible that tetraploid progenitor genotypes were not sampled or that differential gene silencing caused this difference, a more likely explanation was the extinction of tetraploid progenitors of extant hexaploid populations.
Our data also suggest at least two independent origins for hexaploid populations of C. obliqua. Although morphological data are equivocal, coastal hexaploids had a unique isozyme profile among polyploids (Table 5) suggesting an origin separate from interior hexaploids and without the genomic contribution of extant tetraploids. The morphological and isozyme profile of interior hexaploids, which combined morphological characters and alleles at two loci found only in C. glabra and C. lyonii, also suggested multiple origins. Unique allozyme profiles at Fest-1 and 6Pgd-1 allow discrimination among Central Lowland, Interior Low Plateau, and Ozark Plateau populations of interior hexaploids. With the forementioned proviso that inadequate sampling and differential gene silencing could produce these isozyme patterns, multiple origins of interior hexaploids from extinct tetraploid progenitors is suggested.
Evolutionary biogeographic patterns
Although the most closely related genera to Chelone (Chionophila, Nothochelone) occur in western North America, the four species of Chelone are restricted to eastern North America. Only the wide-ranging diploid C. glabra had populations found north of the last glacial maximum. Two of three ploidy levels and all four species occurred in the unglaciated southern Blue Ridge region. Hexaploid populations are known only from the Atlantic Coastal Plain and unglaciated areas of the Interior Highlands. Isozyme data also indicated greater genetic diversity in the Blue Ridge, among all species and within C. glabra and C. cuthbertii (Table 5; Fig. 6). All but one of 42 alleles detected were found in this region. Blue Ridge populations of C. cuthbertii and C. glabra had similar or higher proportions of polymorphic loci, observed heterozygosity, and total genetic diversity compared to populations outside the region (Table 5). Consequently, the Blue Ridge geographic region represents the center of morphological, cytological, and genetic (isozymic) diversity for the genus.
Based on Nei's (1987)
formulas for stepwise mutation rates calculated from genetic identity coefficients, divergence estimates among three diploid species ranged from 1.25 to 1.53 million years BP (before present). These estimates and the low levels of morphological and isozyme divergence observed in Chelone indicated differentiation during the Pleistocene and Holocene epochs beginning two million years ago. Because paleoecological reconstructions of the eastern deciduous forest biome (Davis, 1983
; Delcourt and Delcourt, 1987
) suggested multiple north-south migrations during the Quaternary Period, we propose that areas south of or in the Blue Ridge region represented a recent center of diversification of the lineage.
Analyses of isozyme data consistently grouped diploid populations by geographic region and recent glacial history both within and among species (Figs. 4, 6). For example, among all diploid species, populations from the contiguous unglaciated areas of the Appalachian Plateau, Piedmont and Valley and Ridge, and southern Blue Ridge regions occurred together in the UPGMA phenogram (Figs. 4, 6). Populations of C. glabra from glaciated areas of the New England and Maritime, Central Lowlands, and Superior Upland regions also grouped together consistently (Figs. 4, 6). Isozyme divergence among diploid populations in Chelone apparently reflect geohistorical factors and migrational history rather than lineage differentiation.
The southern Blue Ridge, Ozark Plateau, and Atlantic Coastal Plain regions have been proposed as glacial refugia and as centers for revegetation and plant migration during interglacial periods (Cain, 1944
; Davis, 1983
; Delcourt and Delcourt, 1987
; Delcourt, 1991
). The infraspecific geographically based groupings observed in diploid populations of C. glabra and in polyploid races of C. obliqua generally are concordant with these hypotheses. Each of the three allopatric cytotypes of C. obliqua were confined to one of these regions (4x populations in the Blue Ridge, 6x populations disjunct in coastal and interior regions), and distance-based trees grouped diploid populations of C. glabra from each of these regions. Furthermore, populations of C. glabra from adjoining glaciated areas were associated with populations from purported refugial areas (Fig. 6). Regions with the highest overall genetic similarities were (1) the unglaciated Ozark Plateau and glaciated areas of the Interior Plains region, and (2) the coastal plain and glaciated New England and Maritime region. The isozymic similarities between the New England and Maritime and the Atlantic Coastal Plain populations were illustrated by the presence of Tpi-3b; the only allele not found among Blue Ridge populations.
Observed levels of populational genetic variation along an Atlantic coastal transect (B-B', Fig. 6) were concordant with patterns expected from the central-marginal migrational model (Barrett and Husband, 1990
; Avise, 1994
), which predicted reduced levels of variation in marginal populations resulting from founding events, extinction, and genetic drift. Populations of C. glabra from the New England and Maritime region had lower values of polymorphic loci, observed heterozygosity, and total genetic diversity (HT, Fig. 6) then those in populations of the coastal plain or in any other geographic region. Allelic diversity in the New England and Maritime region was a highly monomorphic subset of the genetic diversity observed in the Atlantic Coastal Plain populations of C. glabra. Similar patterns of reduced genetic variation resulting from putative postglacial range expansion also were observed in Pinus (Cwynar and MacDonald, 1987
; Parker et al., 1997
) and Chamaecyparis (Kuser et al., 1997
). Based on our data, we propose that C. glabra migrated to the New England and Maritime region via an Atlantic coastal route. The geographical pattern of isozymes also suggests that populations of C. glabra in the glaciated interior migrated from an Ozark or southern Blue Ridge center.
Because the three allopatric cytotypes of C. obliqua were most similar genetically to diploid populations in their own or neighboring (unglaciated) regions, occurred in areas hypothesized to be glacial refugia, and lacked any novel alleles or morphological characters, we propose that polyploid evolution in Chelone was geologically recent and was facilitated by the vegetational changes effected during the Pleistocene and Holocene epochs. Similar chromosomal evolutionary scenarios have been postulated in angiosperms and ferns in eastern North America (Werth, Guttmann, and Eshbaugh, 1985a, b;
Loveless and Hamrick, 1988
; Haufler and Zhongren, 1991
; Watson, Elisens, and Estes, 1991
; Werth, 1991
; Case, 1993
). Plant migration and range contraction/expansion cycles during the Quaternary apparently increased the probabilities of multiple differentiation events and independent, allopatric polyploidization cycles such as those described in the European flora (Ehrendorfer, 1959, 1980
). In the fern genus Dryopteris, Werth (1991)
used isozyme data to propose a southward migration of diploid species, with subsequent sympatry and hybridization during glacial advances. We propose a similar scenario for polyploid evolution in Chelone. In Chelone, hybridization/polyploidization cycles apparently occurred in different refugial areas, resulting in the allopatric distribution of extant polyploid cytotypes and several putative examples of multiple, independent alloploid origins. Because there are few morphological and isozyme differences, plastid and additional nuclear molecular markers will be examined to further test these evolutionary and biogeographic hypotheses.
|
|
FOOTNOTES |
|---|
2 Author for correspondence: Department of Biological Sciences, Tarleton State University, Stephenville, Texas 76402.
|
LITERATURE CITED |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
Avise, J. C. 1994 Molecular markers, natural history, and evolution. Chapman and Hall, New York, NY.
Barrett, S. C. H., and B. C. Husband. 1990 The genetics of plant migration and colonization. In A. H. D. Brown, M. T. Clegg, A. L. Kahler, and B. S. Weir [eds.], Plant population genetics, breeding, and genetic resources, 254–277. Sinauer, Sunderland, MA.
Bayer, R. J., and D. J. Crawford. 1986 Allozyme divergence among five diploid species of Antennaria (Asteraceae: Inuleae) and their allopolyploid derivatives. American Journal of Botany 73: 287–296.[CrossRef][ISI]
Bittman, R. L. 1980 Flavonoid and allozyme variation in diploid and polyploid members of the genus Chelone (Scrophulariaceae). Master's thesis, Ohio State University, Columbus, OH.
Brochmann, C., P. S. Soltis, and D. E. Soltis. 1992 Recurrent formation and polyphyly of nordic polyploids in Draba (Brassicaceae). American Journal of Botany 79: 673–689.[CrossRef][ISI]
Brouillet, L., and R. D. Whetstone. 1993 Climate and physiography. In Flora of North America Editorial Committee, Flora of North America North of Mexico, vol. 1, 15–46. Oxford University Press, New York, NY.
Cain, S. A. 1944 Foundations of plant geography. Hafner, New York, NY.
Case, M. A. 1993 High levels of allozyme variation within Cypripedium calceolus (Orchidaceae) and low levels of divergence among its varieties. Systematic Botany 18: 663–677.[CrossRef][ISI]
Conkle, M. T., P. D. Hodgekiss, L. B. Nunnally, and S. C. Hunter. 1982 Starch gel electrophoresis of conifer seeds: laboratory manual. Pacific Southwest Forest and Range Experiment Station, Berkeley, CA.
Cooperrider, T. S., and G. A. McCready. 1970 Chromosome numbers in Chelone (Scrophulariaceae). Brittonia 22: 175–183.[CrossRef][ISI]
Cosner, M. B., and D. J. Crawford. 1990 Allozyme variation in Coreopsis sect. Coreopsis (Compositae). Systematic Botany 15: 256–265.[CrossRef][ISI]
Crawford, D. J. 1990 Plant molecular systematics. J. Wiley&Sons, New York, NY.
Crosswhite, F. S. 1965 Variation in Chelone glabra in Wisconsin (Scrophulariaceae). Michigan Botanist 4: 62–66.
Cwynar, L. C., and G. M. MacDonald. 1987 Geographical variation of lodgepole pine in relation to population history. American Naturalist 129: 463–469.[CrossRef][ISI]
Davis, M. B. 1983 Quaternary history of deciduous forests of eastern North America and Europe. Annals of the Missouri Botanical Garden 70: 550–563.
Delcourt, H. R. 1991 Quaternary ecology: paleoecological perspective. Chapman and Hall, New York, NY.
Delcourt, P. A., and H. R. Delcourt. 1987 Long-term forest dynamics of the temperate zone. Springer-Verlag, New York, NY.
Dixon, W. H., and M. B. Brown [eds.]. 1979 Biomedical computer programs P-series. University of California Press, Los Angeles, CA.
Ehrendorfer, F. 1959 Differentiation-hybridization cycles and polyploidy in Achillea. Cold Spring Harbor Symposium Quantitative Biology 24: 141–152.
———. 1980 Polyploidy and distribution. In W. H. Lewis [ed.], Polyploidy and distribution, 45–60. Plenum, New York, NY.
Elisens, W. J., and A. D. Nelson. 1993 Morphological and isozyme divergence in Gambelia (Scrophulariaceae): species delimitation and biogeographic relationships. Systematic Botany 18: 454–468.[CrossRef][ISI]
Ellstrand, N. C. 1992 Gene flow among seed plant populations. New Forests 6: 241–256.
Fenneman, N. M. 1938 Physiography of the Eastern United States. McGraw Hill, New York, NY.
Fernald, M. L. 1950 Gray's Manual of Botany, 8th ed., 1268–1269. D. van Nostrand Company, New York, NY.
Gleason, H. A. 1952 The new Britton and Brown illustrated flora of the Northeastern United States and adjacent Canada, vol. III, 180–181. New York Botanical Garden, Bronx, NY.
Gottlieb, L. D. 1981 Electrophoretic evidence and plant populations. Progress in Phytochemistry 7: 1–46.
Hamrick, J. L. 1987 Gene flow and distribution of genetic variation in plant populations. In K. M. Urbanska [ed.], Differentiation patterns in higher plants. Academic Press, New York, NY.
———, and M. J. W. Godt. 1990 Allozyme diversity in plant species. In A. H. D. Brown, M. T. Clegg, A. L. Kahler, and B. S. Weir [eds.], Plant population genetics, breeding, and genetic resources, 43–63. Sinauer, Sunderland, MA.
Haufler, C. H. 1987 Electrophoresis is modifying our concepts of evolution in homosporous pteridophytes. American Journal of Botany 74: 953–966.
———, and W. Zhongren. 1991 Chromosomal analyses and the origin of allopolyploid Polypodium virginianum (Polypodiaceae). American Journal of Botany 78: 624–630.[CrossRef][ISI]
Holt, P. C., and R. A. Paterson. 1970 The distributional history of the Southern Appalachians, part II: flora. Virginia Polytechnic Institute and State University, Blacksburg, VA.
Jacobs, B. F., C. W. Werth, and S. I. Guttman. 1984 Genetic relationships in Abies (fir) of eastern United States: electrophoretic study. Canadian Journal of Botany 62: 609–616.
Kuser, J. E., T. R. Meagher, D. L. Sheely, and A. White. 1997 Allozyme frequencies of New Jersey and North Carolina populations of Atlantic white cedar, Chamaecyparis thyoides (Cupressaceae). American Journal of Botany 84: 1536–1541.[Abstract]
Loveless, M. D., and J. L. Hamrick. 1988 Genetic organization and evolutionary history in two North American species of Cirsium. Evolution 42: 254–265.[CrossRef]
Manos, P. S., and D. E. Fairbrothers. 1987 Allozyme variation in populations of six northeastern American red oaks (Fagaceae: Quercus subg. Erythrobalanus). Systematic Botany 12: 365–373.
Murphy, R. W., J. W. Sites, D. G. Buth, and C. H. Haufler. 1996 Proteins: isozyme electrophoresis. In D. M. Hillis, C. Moritz, and B. K. Mable [eds.], Molecular systematics, 51–120. Sinauer, Sunderland, MA.
Nei, M. 1972 Genetic distance between populations. American Naturalist 106: 283–292.[CrossRef][ISI]
———. 1987 Molecular evolutionary genetics. Columbia University Press, New York, NY.
Nelson, A. D. 1995 Polyploid evolution in Chelone (Scrophulariaceae). Ph.D. dissertation, University of Oklahoma, Norman, OK.
Nixon, K. C., and Q. D. Wheeler. 1990 An amplification of the phylogenetic species concept. Cladistics 6: 211–223.
Parker, K. C., J. L. Hamrick, A. J. Parker, and E. A. Stacy. 1997 Allozyme diversity in Pinus virginiana (Pinaceae): intraspecific and interspecific comparisons. American Journal of Botany 84: 1372–1382.[Abstract]
Parks, C. R., J. F. Wendel, M. M. Sewell, and Y. Qiu. 1994 The significance of allozyme variation and introgression in the Liriodendron tulipifera complex (Magnoliaceae). American Journal of Botany 81: 878–889.
Pennell, F. W. 1935 The Scrophulariaceae of eastern temperate North America. Academy of Natural Sciences of Philadelphia Monograph, number 1.
———, and E. T. Wherry. 1928 The genus Chelone of eastern North America. Bartonia 10: 12–23.
Qiu, Y., and C. L. Parks. 1994 Disparity of allozyme variation levels in three Magnolia (Magnoliaceae) species from the southeastern United States. American Journal of Botany 81: 1300–1308.[CrossRef][ISI]
Radford, A. E., H. E. Ahles, and C. R. Bell. 1968 Manual of the vascular flora of the Carolinas. University of North Carolina Press, Chapel Hill, NC.
Rieseberg, L. H., R. Carter, and S. Zona. 1990 Molecular tests of the hypothesized hybrid origin of two diploid Helianthus species (Asteraceae). Evolution 44: 1498–1511.[CrossRef][ISI]
Rohlf, F. J. 1990 Numerical taxonomy and multivariate analysis system, version 1.80. Exeter Software, New York, NY.
Roose, M. L., and L. D. Gottlieb. 1976 Genetic and biochemical consequences of polyploidy in Tragopogon. Evolution 30: 818–830.
SAS. 1990 SAS/STAT user's guide, ver. 6, ed. 4, vol. 2. SAS Institute, Cary, NC.
Sneath, P. H., and R. R. Sokal. 1973 Numerical taxonomy. W. H. Freeman, San Francisco, CA.
Soltis, D. E. 1982 Allozymic variability in Sullivantia (Saxifragaceae). Systematic Botany 7: 26–34.
———. 1985 Allozyme differentiation among Heuchera americana, H. parviflora, H. pubescens, and H. villosa (Saxifragaceae). Systematic Botany 10: 193–198.[CrossRef][ISI]
———, C. H. Haufler, D. C. Darrow, and G. J. Gastony. 1983 Starch gel electrophoresis of ferns: compilation of grinding buffers, gel and electrode buffers, and staining schedules. American Fern Journal 73: 10–27.
———, and P. S. Soltis. 1989 Allopolyploid speciation in Tragopogon: insights from chloroplast DNA. American Journal of Botany 76: 1119–1124.[CrossRef][ISI]
———, and ———. 1991 Multiple origins of the allotetraploid Tragopogon mirus (Compositae): rDNA evidence. Systematic Botany 16: 407–414.
———, and ———. 1993 Molecular data and the dynamic nature of polyploidy. Critical Reviews in Plant Sciences 12: 243–273.
Soltis, P. S., J. J. Doyle, and D. E. Soltis. 1992 Molecular data and polyploid evolution in plants. In D. E. Soltis, P. S. Soltis, and J. J. Doyle [eds.], Molecular systematics of plants, 177–201. Chapman and Hall, New York, NY.
Stebbins, G. L. 1971 Chromosomal evolution in higher plants. Edward Arnold Limited, London.
Stuessy, T. F. 1990 Plant taxonomy: systematic evaluation of comparative data. Columbia University Press, New York, NY.
Swofford, D. L., and R. B. Selander. 1981 Biosys–1: a FORTRAN program for the comprehensive analysis of electrophoretic data in population genetics and systematics. Journal of Heredity 72: 281–283.
Watson, L. E., W. J. Elisens, and J. R. Estes. 1991 Electrophoretic and cytogenetic evidence for allopolyploid origin of Marshallia mohrii (Asteraceae). American Journal of Botany 78: 408–417.[CrossRef][ISI]
Wendel, J. F., and N. F. Weeden. 1989 Visualization and interpretation of plant isozymes. In D. E. Soltis, and P. S. Soltis [eds.], Isozymes in plant biology, 5–45. Dioscorides Press, Portland, OR.
Werth, C. R. 1989 The use of isozyme data for inferring ancestry of polyploid species of pteridophytes. Biochemical Systematics and Ecology 17: 117–130.[CrossRef][ISI]
———. 1991 Isozyme studies on the Dryopteris "spinulosa" complex I: origin of the log fern Dryopteris celsa. Systematic Botany 16: 446–461.
———, S. I. Guttman, and W. H. Eshbaugh. 1985a Electrophoretic evidence of reticulate evolution in the Appalachian Asplenium complex. Systematic Botany 10: 184–192.
———,———, and ———. 1985b Recurring origins of allopolyploid species in Asplenium. Science 228: 731–733.
Whitkus, R. 1988 Modified version of GENESTAT: program for computing genetic statistics from allele frequency. Plant Genetics Newsletter 4: 10.
Wolfe, A. D., and W. J. Elisens. 1993 Diploid hybrid speciation in Penstemon revisited. American Journal of Botany 80: 1082–1094.
———, W. J. Elisens, L. E. Watson, and C. W. dePamphilis. 1997 Using restriction-site variation of PCR-amplified cpDNA genes for phylogenetic analysis of Tribe Cheloneae (Scrophulariaceae). American Journal of Botany 84: 555–564.[Abstract]
Wright, S. 1951 The genetical structure of populations. Annals of Eugenics 15: 323–354.
Wyatt, R., I. J. Odrzykoski, and A. Stoneburner. 1992 Isozyme evidence of reticulate evolution: Plagiomnium medium is an allopolyploid of P. ellipticum x P. insigne. Systematic Botany 17: 532–551.
| ||||||||||||||
