| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
What's this? |
1 Department of Environmental and Resource Sciences/370, University of Nevada, Reno, Nevada 89557 USA; and 2 USDA Forest Service Intermountain Research Station, 920 Valley Road, Reno, Nevada 89512 USA
Received for publication August 18, 1998. Accepted for publication June 8, 1999.
ABSTRACT
Geographic patterns of genetic variation in chlorolast (cpDNA) and nuclear ribosomal (nrDNA) DNA were examined to test the hypothesis of hybridization between Juniperus osteosperma and Juniperus occidentalis in the Great Basin of western Nevada. Noncoding DNA from the trnL-trnF intergenic spacer and the trnL intron of the chloroplast genome was sequenced from seven populations of J. osteosperma and four populations of J. occidentalis sampled over a large proportion of their respective ranges. An adenine nucleotide at position 436 in the aligned sequence and within a Tru 9I restriction site was found to be present in individuals of J. osteosperma sampled from western Colorado and central Utah, but absent in sequences of J. osteosperma sampled from central and western Nevada and all sequences of J. occidentalis. Two hundred fourteen individuals from 34 populations of J. osteosperma and J. occidentalis were then screened for cpDNA haplotype by Tru 9I digestion of the trnL-trnF polymerase chain reaction (PCR) product. Two cpDNA haplotypes were evident, each consisting of restriction fragment profiles that differed solely with respect to the presence or absence of the Tru 9I site encompassing the adenine nucleotide at position 436. One of these haplotypes was monomorphic in J. occidentalis and exhibited a decreasing frequency in J. osteosperma with increasing geographic distance from J. occidentalis in west-central Nevada. Geographic patterns in nuclear ribosomal DNA (nrDNA) variation were examined by restriction fragment analysis and, although spatially more restricted, exhibited patterns of clinal variation similar to those observed in cpDNA haplotype. Genetic relationships based on DNA sequences and geographic patterns of genetic variation in chloroplast and nuclear ribosomal DNA are consistent with morphology in suggesting interspecific gene flow between J. occidentalis and J. osteosperma.
Key Words: chloroplast DNA • Cupressaceae • hybridization • Juniperus occidentalis (subsp. australis and subsp. occidentalis) • Juniperus osteosperma • nuclear ribosomal DNA • restriction fragment length polymorphism (RFLP)
Early studies of evolutionary change in chloroplast DNA (cpDNA) indicated limited variability within species (Banks and Birky, 1985
; Neale et al., 1986
; Birky, 1988
). This finding has been attributed to low rates of sequence evolution and has been maintained as justification for the lack of intraspecific sampling in studies examining relationships at higher taxonomic levels. However, documentation of intraspecific variation in cpDNA has become increasingly common and has been attributed in many cases to "chloroplast capture" following genetic exchange across species boundaries (Wagner et al., 1987
; Govindaraju, Dancik, and Wagner, 1989
; Milligan, 1991
; Soltis et al., 1991
; Bobola et al., 1995
; Mason-Gamer, Holsinger, and Jansen, 1995
; Bain and Jansen, 1996
; see Soltis, Soltis, and Milligan, 1992
; Rieseberg and Wendel, 1993
; and Rieseberg, 1995
, for reviews). Rieseberg and Wendel 1993
list 37 cases of intraspecific variation in cpDNA that are purported to result from hybridization, 24 (65%) of which they considered to be probable instances of introgression. Moreover, Rieseberg 1995
suspected that a review of the literature at that time would reveal over 100 cases of intraspecific variation in cpDNA that could be attributed to hybridization and introgression. That intraspecific variation in cpDNA is potentially indicative of hybridization is founded on the expectation that slowly evolving loci or genomes will produce greater molecular variation between than within species (Soltis, Soltis, and Milligan, 1992
). In cases where a species is polymorphic for cpDNA and at least one of the molecular variants is diagnostic for a second species, interspecific hybridization is a plausible explanation. Concordance in the geographic distribution and relationships suggested by independently evolving characters (e.g., cpDNA, nuclear ribosomal DNA, and morphology) provides additional support for introgression (Wendel and Albert, 1992
; Rieseberg, 1995
; Rieseberg, Whitton, and Linder, 1996
).
Juniperus osteosperma (Torr.) Little is a dominant member of woodland communities over much of the intermountain west (West, Tausch, and Tueller, 1978
). Populations extend from the Colorado plateau northward into southwestern Wyoming and southern Idaho and westward through Nevada to the Sierra Nevada of northeastern California (Little, 1971
). Juniperus occidentalis Hook (western juniper) is confined largely to the mountains of the Sierra Nevada and Cascade ranges (Little, 1971
). Populations of J. osteosperma and J. occidentalis are parapatric over a relatively large region in extreme western Nevada and northeastern California. Sympatric populations occur within this region (e.g., the Pine Nut Mountains, Douglas County, Nevada and Bodie Hills, Mono County, California), as well as on a few mountain ranges in central Nevada (e.g., Monitor and Toquima Ranges, Nye County, Nevada; R. Tausch, personal observation). Recent phytogeographic study has documented 146 records of J. occidentalis from 54 mountain ranges in Nevada, with single individuals or small populations occurring as far east as the Snake Mountains of extreme eastern Nevada (Charlet, 1996
). In Nevada, populations of J. occidentalis typically are part of high sagebrush steppe or mountain brush communities well above the elevational limits of J. osteosperma. However, mixed stands of J. occidentalis and J. osteosperma occur occasionally along drainages at lower elevations, e.g., along San Juan Creek in the Toiyabe Range of central Nevada (R. Tausch, personal observation), or more frequently at higher elevations in some ranges (Charlet, 1996
).
Juniperus osteosperma and J. occidentalis were described from type specimens collected near the extremes of their ranges, i.e., western New Mexico and northern Oregon, respectively. The two species have been distinguished using both vegetative (bark color and foliar gland characteristics) and reproductive (number of seeds per berry and seed length) features (Adams, 1993
). Vasek (1966)
studied morphological variation in J. osteosperma and J. occidentalis collected over a wide geographic area. He concluded that, while individual traits have broad ranges of overlap, allopatric populations of J. osteosperma and J. occidentalis could be reliably distinguished using suites of morphological characters in combination with differences in ecological preference. However, he noted populations of J. occidentalis occidentalis sympatric with populations of J. osteosperma in northwestern Nevada that were morphologically intermediate between J. occidentalis occidentalis and J. osteosperma, as well as populations of J. osteosperma from the eastern base of the Sierra Nevada that were morphologically similar to J. occidentalis australis (Vasek, 1966
). These patterns of variation were attributed to introgression between sympatric or parapatric populations of each subspecies of J. occidentalis and J. osteosperma (Vasek, 1966
). The recent documentation of populations of J. occidentalis over a large portion of Nevada and their geographic overlap with J. osteosperma over much of this range suggests the potential for introgression over a much larger area than previously suspected (Charlet, 1996
).
In this study, patterns of genetic variation in cpDNA and nrDNA are used to test hypotheses of hybridization between J. osteosperma and J. occidentalis. The taxonomic limits of genetic markers are examined, and biogeographic hypotheses are proposed that address the distribution of cpDNA and nrDNA variability in J. osteosperma. Genetic relationships based on DNA sequences and geographic patterns of genetic variation in chloroplast and nuclear ribosomal DNA are consistent with morphological features in suggesting interspecific gene flow between J. occidentalis and J. osteosperma.
MATERIALS AND METHODS
Plant sampling
Samples from 214 individuals collected from 34 populations of J. occidentalis and J. osteosperma were included in this study (Table 1). Seventy-five individuals were sampled from 11 populations and both subspecies of J. occidentalis (J. occidentalis occidentalis and J. occidentalis australis) from south-central Oregon and the Cascade and Sierra Nevada ranges of northern California. One hundred thirty-nine individuals of J. osteosperma were sampled from 23 populations that ranged from extreme western Nevada, eastward into Utah, Wyoming, and Colorado (Fig. 1). Taxonomic determinations were based on the treatments of Adams (1993)
and Vasek (1966)
. Voucher specimens of all individuals included in this study are stored at the USDA-Forest Service Intermountain Research Station, Reno, Nevada, USA.
|
|
DNA sequencing and restriction site analysis of chloroplast DNA
Our experimental approach included DNA sequencing to identify nucleotide differences in cpDNA that were putatively diagnostic at the species level, i.e., were variable between but invariant within species. Individuals were then screened using restriction enzymes that targeted species-specific nucleotide differences. This approach provided an economically feasible alternative to direct sequencing of large numbers of individuals and allowed genetic variation in cpDNA to be examined from many populations sampled over a wide geographic area.
A region of the chloroplast genome containing the trnL intron as well as the trnL-trnF intergenic spacer was amplified using the polymerase chain reaction (PCR) and primers C and F of Taberlet et al. (1991)
. Double-stranded PCR products were directly sequenced using TAQ-fs polymerase and fluorescent dye terminators at the Cancer Research Center, University of Chicago, Chicago, Illinois, USA. In an initial analysis designed to examine intra- and interspecific sequence variation at the trnL-trnF locus, ~500 bp of sequence were obtained for single individuals from the Fruita, Salina, Snake Creek, and War Canyon populations of J. osteosperma and the Sonora Pass, Cisco Grove, Stony Ridge, and Juniper Mountain populations of J. occidentalis (Table 1). In addition, sequences were obtained from representatives of four additional species of Juniperus (J. californica, J. communis, J. pinchotii, and J. saltillensis) that occur in the western and southwestern United States. All sequences have been deposited in GenBank under the accession numbers AF211509–AF211526.
Sequences were aligned using ClustalW in the SeqPup program (D. Gilbert, University of Indiana). An adenine nucleotide at position 436 in the aligned sequence and within a Tru 9I recognition site was present in those sequences obtained from populations of J. osteosperma from Colorado, Utah, and extreme eastern Nevada, but absent in all other populations and species (Fig. 2). All individuals sampled from J. osteosperma and J. occidentalis were then screened for variation in the presence or absence of the Tru 9I site by digestion of the trnL-trnF PCR product and electrophoresis in 4% NuSieve GTG agarose (FMC Bioproducts, Rockland, Maine, USA). Two restriction fragment profiles were generated, one having fragments sizes of ~250, 190, 176, and 120 bp and the other having fragments of ~310, 250, and 176 bp. These profiles differed only with respect to the presence or absence of a Tru 9I site at positions 433–436 and are hereafter referred to as the Utah and general cpDNA haplotypes, respectively (i.e., general in that this haplotype is characteristic of all species included in the analysis, with the exception of some J. osteosperma). Several individuals that were putatively J. osteosperma based on morphology and geographic location but possessed the general cpDNA haplotype were also sequenced for the trnL-trnF locus. These included individuals from the Castle Valley, Confusion Range, Fruita, Ft. Bridger, Lida Summit, Monitor, and Lovell Summit populations. This allowed the nature of the genetic variation resulting in loss of the Tru 9I site to be determined and provided a reasonable test of the assumption that the absence of the Tru 9I site is homologous in different populations.
|
). The PCR generated two distinct products of ~700 and 1500 bp. These fragments were used simultaneously as hybridization probes and revealed banding patterns for the enzymes used that were essentially identical to those detected by hybridizing with a recombinant probe encompassing the 45S coding region of the nuclear ribosomal repeat of soybean (Jackson and Lark, 1982
). Probes were nonisotopically labeled using random primer extension and digoxigenin 11-dUTP and hybridized to target sequences overnight at 40°C in Dig Easy Hyb hybridization solution (Boehringer Mannheim, Indianapolis, Indiana, USA). Four posthybridization washes were conducted, two for 5 min each at room temperature in 2X SSC, 0.1% SDS, followed by two for 15 min at 60°C using 0.1X SSC, 0.1% sodium dodecyl sulfate (SDS). Hybridized blots were exposed to X-ray film for 15–20 min following treatment with the chemiluminescent substrate CSPD (Tropix Inc., Bedford, Massachusetts, USA). Probe labeling and signal detection were conducted according to the manufacturer (Genius Chemiluminescent Detection System, Kit 7, Boehringer Mannheim, Indianapolis, Indiana, USA).
Fragment sizes in restriction fragment length polymorphism (RFLP) analyses were estimated by comparison of relative electrophoretic migration distances to that of markers of known molecular mass (
-Hind III controls). At least three lanes of control markers were run for each gel of 36 lanes. Fragments were treated as homologous if their inferred molecular masses were within ±5%, and individual fragments were assumed to be independent.
Data analysis
Neighbor-joining analysis of DNA sequences was performed using PAUP* test version 4.0d61 written by David Swofford. A total aligned length of 460 bp was obtained for most accessions. Thirteen length mutations were encountered ranging from 1 to 15 bp in size, only one of which was potentially informative and included in the analysis. Analyses were conducted using the Standard Distances, Total Character Differences, and Among Site Variation options in PAUP* and were rooted with the trnL-trnF sequence of Juniperus communis (section Juniperus sensu Adams, 1993
).
Statistical differences in mean frequencies of nrDNA markers were examined for groups of populations defined by species designation and cpDNA haplotype. Comparisons were made among three groups of populations: (1) J. osteosperma that were monomorphic for the Utah cpDNA haplotype (i.e., the Snake Creek, Salina, I-Sky, and Cache Valley populations); (2) J. osteosperma populations that were either monomorphic or polymorphic for the general cpDNA haplotype [i.e., all J. osteosperma populations except those listed in (1) above]; and (3) J. occidentalis. One-way analysis of variance (ANOVA) and least significant difference comparison of means (LSD) were performed using the Statistix software analysis package (version 4.0, Analytical Software, St. Paul, Minnesota, USA) following transformation of the frequency data where 
= arcsine p1/2 and p is the population frequency (Sokal and Rohlf, 1981
). Transformed data were converted back to frequencies for reporting.
RESULTS
Eleven point mutations and nine length mutations comprising eight cpDNA haplotypes were evident among the individuals of J. osteosperma and J. occidentalis sequenced for the trnL-trnF spacers in this study. All individuals from J. occidentalis possessed different sequences, with the mean distance between sequences being 0.8%. At most, six different sequences were present in the sampled J. osteosperma, with individuals from the Snake Creek, Salina, and Fruita (number 5) populations having unique sequences of mean distance 0.4%. All other individuals of J. osteosperma shared the absence of the adenine nucleotide at position 436 in the aligned sequence and had sequences that were identical to one another as well as to the sequence obtained from the Juniper Mtn. population of J. occidentalis occidentalis, with the exception of the War Canyon sequence, which had two positions that could not be scored unequivocally, and the Monitor sequence, which had two unique mutations. Of the 20 mutations present in the trnL-trnF sequences, only one was present in more than one individual, i.e., the insertion/deletion of the adenine nucleotide at position 436 that comprises the 3'-most position of a Tru 9I restriction site (5'-TTAA-3'). Population frequencies for cpDNA haplotypes that were differentiated based on the presence (Utah cpDNA haplotype) or absence (general cpDNA haplotype) of this Tru 9I site were determined by digesting the trnL-trnF amplification product with Tru 9I for all sampled individuals (Table 2). All populations of J. occidentalis were invariant for the general cpDNA haplotype, while populations of J. osteosperma generally exhibited a decreasing frequency of the general cpDNA haplotype along a west-east transect from western Nevada to central Utah (Fig. 2). The four northwesternmost populations of J. osteosperma sampled in this study (i.e., Kumiva Peak, Lava Beds, Virginia Mountains, and War Canyon) were monomorphic for the general cpDNA haplotype, while a single population from extreme eastern Nevada (Snake Creek), two populations from southern Utah (Salina and I-Sky), as well as a single population from northern Utah (Cache Valley) were monomorphic for the Utah cpDNA haplotype. Four of the easternmost populations of J. osteosperma were polymorphic for cpDNA haplotype, with the Castle Valley, Ft. Bridger, Fruita, and Mesa populations having frequencies for the general cpDNA haplotype of 0.14, 0.14, 0.25, and 0.14, respectively (Table 2; Fig. 2). Sequences were determined for individuals having the general cpDNA haplotype from eight populations of J. osteosperma sampled over a wide geographic area ranging from west-central Nevada eastward to Utah and western Colorado and northward to southwestern Wyoming (Fig. 2). In all cases the loss of the Tru 9I site was attributable strictly to deletion of an adenine nucleotide at position 436.
|
Five nrDNA markers had mean frequencies that were significantly different when compared between those populations of J. osteosperma that were monomorphic for the Utah cpDNA haplotype and all populations of J. occidentalis (Table 3). One-way ANOVA and comparison of means for three of these markers (i.e., the 11.5, 8.2, and 7.5 kb Hind III markers) indicated that those populations of J. osteosperma having the general cpDNA haplotype had mean population frequencies that were intermediate in value to those J. osteosperma populations monomorphic for the Utah cpDNA haplotype and populations of J. occidentalis. However, mean frequencies for these markers were not significantly different between those J. osteosperma populations having at least some individuals with the general cpDNA haplotype and J. osteosperma populations monomorphic for the Utah cpDNA haplotype (Table 3). The 6.8-kb Eco RI and 16.0-kb Dra I markers had significantly different mean frequencies between species based on T tests, but were not present in any populations of J. osteosperma.
|
Neighboring-joining analysis of DNA sequences and the adenine insertion/deletion at position 436 identified four principal groups of J. osteosperma and J. occidentalis. One consisted of those J. osteosperma possessing the Utah cpDNA haplotype (Snake Creek, Salina, and Fruita number 5), which was most closely related to a group containing all J. osteosperma having the general cpDNA haplotype and Juniper Mountain of J. occidentalis occidentalis (Fig. 3). Nested at the base of the J. osteosperma-Juniper Mountain group are two groups of J. occidentalis, one consisting of a Stony Ridge-Sonora Pass clade and the other of Cisco Grove (Fig. 3).
|
Three possible explanations are proposed for the cpDNA haplotype variation in J. osteosperma detected by Tru 9I restriction of the trnL-trnF PCR product: (1) inheritance of ancestral polymorphism; (2) intraspecific polymorphism; and (3) hybridization between J. occidentalis and J. osteosperma.
Of these hypotheses, inheritance of ancestral polymorphism seems the least likely. Results from neighbor-joining (Fig. 3) and parsimony (data not shown) analyses indicate that J. occidentalis and J. osteosperma are sister groups. Moreover, mean divergence in trnF-trnL sequences between J. occidentalis and J. osteosperma was only 0.6%. Thus, J. occidentalis and J. osteosperma appear to be closely related, and if they evolved from a common ancestor that was polymorphic for Tru 9I restriction site variation, then we would expect to find haplotype variation in populations of J. occidentalis. However, despite sampling 75 individuals from 14 populations over a large geographic area, only the general cpDNA haplotype was found in J. occidentalis. Moreover, all species included in this study have the general cpDNA haplotype, suggesting that the most recent common ancestor of J. occidentalis and J. osteosperma was monomorphic for the general cpDNA haplotype. Perhaps some populations of J. occidentalis that were not sampled in this study are polymorphic for the general and Utah cpDNA haplotypes. If this is the case, that variation may be distributed over a relatively narrow geographic range, possibly near the southern or northern limits of the species' distribution. Another possibility is that differential lineage sorting has eliminated the Utah cpDNA haplotype in J. occidentalis. The apparent recent divergence of J. occidentalis and J. osteosperma, as well as the lack of cpDNA haplotype variation in J. occidentalis populations sampled over a large geographic area, suggests that if the lack of variation is attributable to differential lineage sorting, then it probably occurred early in the evolutionary history of the group before the species expanded to occupy its current range.
Perhaps mutation has produced polymorphism in cpDNA haplotype within J. osteosperma. All other species of Juniperus included in this study possess the general cpDNA haplotype. Thus, this haplotype appears to be symplesiomorphic in Juniperus and may be ancestral in J. osteosperma as well. If this is the case, the Utah cpDNA haplotype is an apomorphy that has arisen within J. osteosperma, although it is not evident what ecological, population genetic, or historical factors have produced the observed geographic distribution of this haplotype. Thus, the cpDNA data alone do not differentiate between the competing hypotheses of hybridization and intraspecific variation in explaining haplotype polymorphism in J. osteosperma.
Altogether, these data most strongly support the hypothesis of gene flow between distinct lineages. This contention is bolstered by three lines of evidence. First, the biogeographic distribution of cpDNA variation is consistent with the hypothesis of gene flow from J. occidentalis to J. osteosperma. Tru 9I digestion of the trnL-trnF PCR product identified a cpDNA haplotype (i.e., the general cpDNA haplotype) that is fixed in all J. occidentalis sampled. This haplotype is fixed in several populations of J. osteosperma sampled from extreme western Nevada and exhibits a decreasing frequency along a west to east gradient that stretches from extreme western Nevada to central Nevada (Fig. 2). This geographic pattern in genetic variation would be expected if cytoplasmic introgression is occurring between J. occidentalis and J. osteosperma, if J. occidentalis is the cytoplasm donor, and if the extent of introgression is determined largely by the geographic distance between populations. In addition, each of the 15 DNA sequences of J. occidentalis and J. osteosperma included here have at least one unique mutation, with the exception of those from seven populations of J. osteosperma (i.e., individuals from the Castle Valley, Confusion Range, Ft. Bridger, Fruita, Lida Summit, Spring Mountains, and War Canyon populations), which are putative hybrids based on the presence of the general cpDNA haplotype. These sequences are identical to one another as well as to that of an individual sampled from the Juniper Mountain population (J. occidentalis occidentalis; Fig. 3). These data suggest that cytoplasmic introgression in J. osteosperma is occurring preferentially through hybridization with J. occidentalis occidentalis. A second line of evidence that supports hybridization is the presence of nrDNA markers that are characteristic for J. occidentalis in populations of J. osteosperma. Five nrDNA markers were identified that have significantly different mean frequencies between J. occidentalis and four populations of J. osteosperma that were monomorphic for the Utah cpDNA haplotype and geographically well removed from J. occidentalis (i.e., Cache Valley, I-Sky, Salina, and Snake Creek; Table 2). Three of these markers (H11.5, H8.2, and H7.5) were present in populations of J. osteosperma from western and central Nevada (Table 2; Fig. 2). Again, frequencies were highest in J. osteosperma populations that are geographically proximate to J. occidentalis and generally decreased along a west to east transect that extends from western to central Nevada (Fig. 2). Thus a second presumably independently evolving locus exhibits a pattern of geographic variation that is concordant with that of the cpDNA haplotype in supporting hybridization. Thirdly, geographic patterns in genetic and morphological variation are corroborative in substantiating introgression between J. occidentalis and J. osteosperma. Vasek (1966)
studied a number of vegetative and reproductive features in assessing taxonomic limits in J. occidentalis and J. osteosperma. He found that populations of J. osteosperma from west-central Nevada were similar to J. occidentalis with respect to branch angle, the frequency of ternate leaves, and the presence of resin-exuding foliar glands. Additional study has documented exudating foliar glands in populations of Juniperus from as far east as the Clan Alpine mountains of west-central Nevada (Charlet, 1996
) and the Toiyabe Range of central Nevada (R. Tausch, personal observation), although it is difficult to differentiate populations of J. occidentalis, J. osteosperma, and their hybrids in this region in some instances (Charlet, 1996
). We examined the co-occurrence of genetic markers and the presence of exudating glands and found that all the J. osteosperma populations from west-central Nevada in which exudating glands were observed (i.e., Virginia Mountains, War Canyon, Wassuks, White Mountains, and San Juan Creek) had at least some individuals with both the general cpDNA haplotype and nrDNA markers that are diagnostic for J. occidentalis (data not shown). Although quantitative relationships between genetic and morphological variation have not been examined, there appears to be a relationship between the frequencies and geographic distributions of exudating glands, the general cpDNA haplotype, and nrDNA markers that are characteristic of J. occidentalis, with each having its highest frequency in populations of J. osteosperma that are geographically proximate to J. occidentalis and with frequencies generally decreasing in more eastern populations. These data suggest an interdependence in the geographic covariation of morphological and genetic features that is consistent with the hypothesis of hybridization between J. occidentalis and J. osteosperma.
Some exceptions to the geographic patterns of genetic variation described above were evident. First, populations of J. osteosperma sampled from eastern Utah, western Colorado, and southwestern Wyoming contained some individuals having the general cpDNA haplotype. These populations are at least 850 km to the east of the putative J. occidentalis-J. osteosperma hybrid zone as defined by overlap in geographic patterns of variation in cpDNA haplotype, nrDNA markers, and morphological features (Fig. 2). Perhaps these occurrences, as well as the widespread distribution of the general cpDNA haplotype in J. osteosperma from Nevada, result in part from postglacial dispersion of hybrid populations. Paleoecological studies indicate that populations of J. occidentalis and J. osteosperma were likely sympatric in northwestern Nevada during the last glacial maximum ~15 000–20 000 yr ago (Thompson, Benson, and Hattori, 1986
; Betancourt, van Devender, and Martin, 1990
), and it appears that populations of one or both species have persisted in the lower elevation valleys of western Nevada over the last 30 000 yr (Nowak et al., 1994a, b
). If these valleys served as refugia for populations of J. occidentalis and J. osteosperma during the last glacial maximum (Van Devender and Spaulding, 1979
; Thompson and Mead, 1982
; Mead, Van Devender, and Cole, 1983
) and interspecific gene flow was frequent during this time, then patterns of genetic variation in J. osteosperma could be strongly influenced by patterns of postglacial migration of J. occidentalis-J. osteosperma hybrids. This could account for the relative high frequencies of the general cpDNA haplotype in populations of J. osteosperma from the central Great Basin, an area from which refugial populations of juniper may have been largely absent due to the relatively high elevation of the valley floors (Wells, 1983
). The presence of the general cpDNA haplotype in populations of J. osteosperma from eastern Utah, western Colorado, and southwestern Wyoming may be the result of infrequent long-distance dispersals from populations in central and eastern Nevada or perhaps from infrequent colonization of hybrids from refugia in southern Idaho, where the Snake River plain may have provided habitat for juniper that was unsuitable in the high mountain basins to the south and pluvial Lake Bonneville would have excluded juniper from a large area in northwestern Utah.
Another unexpected finding was the occurrence of the 11.5-kb Hind III marker at high frequency in the Confusion Range and Cedar Mountain populations in western Utah. The geographic pattern of genetic variation in nrDNA markers characteristic for J. occidentalis but present in J. osteosperma is characterized by frequencies that are high in those populations sampled from extreme western Nevada, but that rapidly decrease eastward to a frequency of 0.20 in the San Juan Creek and Lida Summit populations of central and southern Nevada (Fig. 2). This pattern of nrDNA variation is generally concordant with the geographic distribution of morphological introgression in J. osteosperma, particularly with respect to the phytogeographic occurrence of exudating foliar glands (Charlet, 1996
; R. Tausch, personal observation). Given these observations, the presence of the 11.5-kb Hind III marker in the Confusion Range and Cedar Mountain populations in the absence of any obvious morphological introgression is noteworthy. Perhaps the occurrence of the 11.5-kb Hind III marker in these populations is a consequence of both hybridization and convergent evolution. For example, of the eight individuals that possess the 11.5-kb nrDNA marker between these two populations, three possess the general cpDNA haplotype and five have the Utah cpDNA haplotype. Assuming that cytoplasmic inheritance is paternal in juniper, the three individuals having the general cpDNA haplotype and the 11.5-kb marker could be early generation hybrids (F1 or F2) in which J. occidentalis was the cytoplasmic donor. In contrast, long-distance dispersal of J. occidentalis seed followed by hybridization in which J. osteosperma is the paternal parent could account for the co-occurrence of the 11.5-kb marker and the Utah cpDNA haplotype in some individuals of the Confusion Range and Cedar Mountain populations (Arnold, Buckner, and Robinson, 1991
). However, we make two observations in regards to this second hypothesis. First, despite the fact that pollen transfer from J. osteosperma to J. occidentalis is expected to occur at an appreciable frequency, no instance of capture of the Utah cpDNA haplotype in populations of J. occidentalis was found in this study. This observation suggests either cytonuclear incompatibility when J. osteosperma is the cytoplasm donor (Keim et al., 1989
; Rieseberg, Whitton, and Linder, 1996
) or that the Utah cpDNA haplotype arose within J. osteosperma and that hybridization between these species is not occurring. Second, the absence of morphological introgression in the Confusion Range and Cedar Mountain populations, in light of the apparent close phytogeographic association in the co-occurrence of nrDNA markers and morphological introgression in other J. osteosperma populations, suggests that the 11.5-kb nrDNA marker could have arisen convergently in these populations. Convergence either through parallel mutation or differential homogenization and concerted evolution (Jorgensen and Cluster, 1988
) could account for the apparent lack of morphological introgression in individuals of the Confusion Range and Cedar Mountain populations having the 11.5-kb nrDNA marker, irrespective of the type of cpDNA haplotype present.
The apparent sister-group relationship between J. occidentalis and J. osteosperma and low levels of trnL-trnF sequence divergence between these taxa imply that the patterns of genetic variation documented here may be the result of primary contact (i.e., gene flow between divergent populations as opposed to distinct species). This contention is consistent with the rather poor morphological differentiation between these species over a relatively wide geographic area and a paucity of characters in establishing taxonomic circumscriptions (Vasek, 1966
; Adams, 1993
; Charlet, 1996
). These findings underscore the need for additional study of hybridization using multiple independently evolving loci and stress the importance of reassessment of existing taxonomic limits (Charlet, 1996
).
FOOTNOTES
1 This research was funded in part by a grant from the U.S. Department of Energy Program for Ecosystem Research (DE-FG03-93ER 61668) and by the Intermountain Research Station, Forest Service, U.S. Department of Agriculture. Financial support does not constitute an endorsement of the views expressed in the article by the DOE or the USFS. Additional support to RSN was provided by the Nevada Agricultural Research Station and by a sabbatical leave from the University of Nevada Reno. The authors thank Dr. Robert P. Adams, Baylor University, for graciously providing the DNA samples for J. communis, J. saltillensis, J. pinchotii, and J. californica, Dr. James A. Young, Agricultural Research Service, for collecting samples for the Lava Beds population, and Craig Biggart for assistance in population sampling. 
2 Author for correspondence. Current address: Division of Biological Sciences, University of Montana, Missoula, Montana 59812 USA.
LITERATURE CITED
Adams, R. P. 1993 Juniperus (Cupressaceae). In Flora of North America Editorial Committee [eds.], The Flora of North America, 412–420. Oxford University Press, New York, New York, USA.
Arnold, M. L., C. M. Buckner, and J. J. Robinson. 1991 Pollen-mediated introgression and hybrid speciation in Louisiana irises. Proceedings of the National Academy of Sciences, USA 88: 1398–1402.
Bain, J. F., and R. K. Jansen. 1996 Numerous chloroplast DNA polymorphisms are shared among different populations and species in the aureoid Senecio (Packera) complex. Canadian Journal of Botany 74: 1719–1728.
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]
Banks, J. A., and C. W. Birky. 1985 Chloroplast DNA diversity is low in a wild plant, Lupinus texensis. Proceedings of the National Academy of Sciences, USA 82: 6950–6954.
Betancourt, J. L., T. R. van Devender, and P. S. Martin. 1990 Packrat middens—the last 40,000 years of biotic change. University of Arizona Press, Tucson, Arizona, USA.
Birky, C. W. 1988 Evolution and variation in plant chloroplast and mitochondrial genomes. In L. D. Gottlieb and S. K. Jain [eds.], Plant evolutionary biology, 23–53. Chapman and Hall, New York, New York, USA.
Bobola, M. S., R. T. Eckert, A. S. Klein, K. Staplefeldt, K. A. Hillenberg, and S. B. Gendreau. 1995 Hybridization between Picea rubens and Picea mariana: differences observed between montane and coastal island populations. Canadian Journal of Forest Research 26: 444–452. [CrossRef]
Charlet, D. A. 1996 Atlas of Nevada conifers. University of Nevada Press, Reno, Nevada, USA.
Govindaraju, D. R., B. P. Dancik, and D. B. Wagner. 1989 Novel chloroplast DNA polymorphisms in a sympatric region of two pines. Journal of Evolutionary Biology 2: 49–59.
Jackson, P. J., and K. G. Lark. 1982 Ribosomal RNA synthesis in soybean suspension cultures growing in different media. Plant Physiology 69: 234–239.
Jorgensen, R. A., and P. D. Cluster. 1988 Modes and tempos in the evolution of nuclear ribosomal DNA: new characters for evolutionary studies and new markers for genetic and population studies. Annals of the Missouri Botanical Garden 75: 1238–1247. [CrossRef][Web of Science]
Keim, P., K. N. Paige, T. G. Whitham, and K. G. Lark. 1989 Genetic analysis of an interspecific hybrid swarm of Populus: occurrence of unidirectional introgression. Genetics 123: 557–565.
Little, E. L. 1971 Atlas of the United States trees. United States Government Printing Office, Washington, D.C., USA.
Mason-Gamer, R. J., K. E. Holsinger, and R. K. Jansen. 1995 Chloroplast DNA haplotype variation within and among populations of Coreopsis grandiflora (Asteraceae). Molecular Biology and Evolution 12: 371–381. [Web of Science]
Mead, J. I., T. R. Van Devender, and K. L. Cole. 1983 Late Quaternary small mammals from Sonoran Desert packrat middens, Arizona and California. Journal of Mammalogy 64: 173–180. [CrossRef]
Milligan, B. G. 1991 Chloroplast DNA diversity within and among populations of Trifolium pratense. Current Genetics 19: 411–416
Neale, D. B., M. A. Saghai-Maroof, R. W. Allard, Q. Zhang, and R. A. Jorgensen. 1986 Chloroplast DNA diversity in populations of wild and cultivated barley. Genetics 120: 1105–1110.
Nowak, C. L., R. S. Nowak, R. J. Tausch, and P. E. Wigand. 1994a Tree and shrub dynamics in northwestern Great Basin woodland and shrub steppe during the Late-Pleistocene and Holocene. American Journal of Botany 81: 265–277. [CrossRef][Web of Science]
———, ———, ———, and ———. 1994b A 30,000 year record of vegetation dynamics at a semi-arid locale in the Great Basin. Journal of Vegetation Science 5: 579–590. [CrossRef][Web of Science]
Rieseberg, L. H. 1995 The role of hybridization in evolution: old wine in new skins. American Journal of Botany 82: 944–953. [CrossRef][Web of Science]
———, and J. F. Wendel. 1993 Introgression and its consequences in plants. In R. G. Harrison [ed.], Hybrid zones and the evolutionary process, 70–109. Oxford University Press, New York, New York, USA.
———, J. Whitton, and C. R. Linder. 1996 Molecular marker incongruence in plant hybrid zones and phylogenetic trees. Acta Botanica Neerlandica 45: 243–262. [Web of Science]
Sokal, R. R., and F. J. Rohlf. 1981 Biometry: the principles and practice of statistics in biological research. W. H. Freeman, San Francisco, California, USA.
Soltis, D. E, P. S. Soltis, T. G. Collier, and M. L. Edgerton. 1991 Chloroplast DNA variation within and among genera of the Heuchera group: evidence for extensive chloroplast capture and the paraphyly of Heuchera and Mitella. American Journal of Botany 78: 1091–1112.
———, ———, and B. G. Milligan. 1992 Intraspecific chloroplast DNA variation: systematic and phylogenetic implications. In P. S. Soltis, D. E. Soltis, and J. J. Doyle [eds.], Molecular systematics of plants, 441–457. Chapman and Hall, New York, New York, USA.
Taberlet, P., L. Gielly, G. Pautou, and J. Bouvet. 1991 Universal primers for amplification of three non-coding regions of chloroplast DNA. Plant Molecular Biology 17: 1105–1109. [CrossRef][Web of Science][Medline]
Thompson, R. S., L. Benson, and E. M. Hattori. 1986 A revised chronology of the last Pleistocene lake cycle in the central Lahontan Basin. Quaternary Research 37: 1–15.
———, and J. I. Mead. 1982 Late Quaternary environments and biogeography in the Great Basin. Quaternary Research 17: 39–55.
Van Devender, T. R., and W. G. Spaulding. 1979 Development of vegetation and climate in the southwestern United States. Science 204: 701–710.
Vasek, F. C. 1966 The distribution and taxonomy of three western junipers. Brittonia 18: 350–372. [CrossRef][Web of Science]
Wagner, D. B., G. R. Furnier, M. A. Saghai-Maroof, S. M. Williams, B. P. Dancik, and R. W. Allard. 1987 Chloroplast DNA polymorphisms in lodgepole and jack pines and their hybrids. Proceedings of the National Academy of Sciences, USA 84: 2097–2100.
Wells, P. V. 1983 Paleobiogeography of montane islands in the Great Basin since the last glaciopluvial. Ecological Monographs 53: 341–382.
Wendel, J. F., and V. A. Albert. 1992 Phylogenetics of the cotton genus (Gossypium L.): character-state weighted parsimony analysis of chloroplast DNA restriction site data and its systematic and phylogenetic implications. Systematic Botany 17: 115–143. [CrossRef][Web of Science]
West, N. E., R. J. Tausch, and P. T. Tueller. 1978 Phytogeographical variation within the juniper-pinyon woodlands of the Great Basin. Great Basin Naturalist Memoirs 2: 119–136.
CiteULike Complore Connotea Del.icio.us Digg Facebook Reddit Technorati Twitter What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
) and J. occidentalis (= 
) . Asterisks and arrows indicate populations for which individual sequences having the general-type and Utah-type cpDNA haplotypes were determined, respectively. Combined frequencies for the 11.5, 8.2, and 7.5 Hind III nuclear ribosomal markers are given in parentheses