Advertisement
Research Article

Population Structure of Vitis rupestris, an Important Resource for Viticulture

Dániel Pap, Allison J. Miller, Jason P. Londo, László G. Kovács
Am J Enol Vitic.  2015  66: 403-410  ; DOI: 10.5344/ajev.2015.15012

Abstract

The wild North American grapevine Vitis rupestris Scheele is an important genetic resource for viticulture, but its natural populations have been severely depleted. We collected samples from seven V. rupestris populations from the Ozark Plateau in Missouri and the Ouachita Mountains in Oklahoma and genotyped them with 14 microsatellite markers to assess allelic diversity, heterozygosity, and genetic differentiation at various levels of population structure. We found that genetic diversity in V. rupestris was similar to that measured in many V. vinifera L. ssp. sylvestris populations and in other outcrossing angiosperms. We detected significant genetic differentiation among populations (ΦPT = 0.105); there was no significant deviation from Hardy–Weinberg equilibrium in some populations, and there was moderate inbreeding in others. Pronounced differentiation between Missouri and Oklahoma populations was supported by a Bayesian clustering approach and principle coordinate analyses and was apparently a function of geographic distance. Genetic differentiation among Missouri populations was modest. We posit that population differentiation and genetic drift may be inherent characteristics of V. rupestris.

North American grapevine species (Vitis spp.) are key resources for the development of new grape cultivars and rootstocks (Reisch et al. 2012). While several grapevine species thrive in their native ranges, others may be threatened by genetic erosion as a result of human activity and environmental stochasticity (Pavek et al. 2000). One Vitis species that is particularly impacted by these phenomena is the rock grape, V. rupestris Scheele (Moore 1991, Pavek et al. 2003). Once abundant in its native range across Maryland, Pennsylvania, Virginia, Kentucky, Tennessee, Indiana, Illinois, Missouri, Arkansas, Oklahoma, and Texas (Munson 1909), V. rupestris is now severely diminished and is designated as imperiled or critically imperiled in many states (NatureServe 2014). V. rupestris populations still grow on the Ozark Plateau in Missouri and Arkansas and in the Ouachita Mountains in Oklahoma (Moore 1991, Kartesz et al. 2013), but they have recently been designated as vulnerable in these regions by the Nature Conservancy (NatureServe 2014).

Despite the importance of V. rupestris for the global grape industry, all indications are that this species is in decline primarily due to habitat destruction caused by extensive hydrologic engineering projects such as channelization of rivers and construction of dams and spillways. Agricultural practices such as fertilization and cattle ranching near rivers are also thought to have had detrimental effects on these plants (Pavek et al. 2003). As a species that colonizes nutrient-poor gravel bars and rocky banks along rivers and intermittent streams, V. rupestris is adapted to harsh environmental conditions. It has evolved robust tolerance to abiotic and biotic stress, and as a result it has become a key source of germplasm for breeding stress-tolerant and disease- and pest-resistant grapevines (Reisch et al. 2012). Most notably, V. rupestris has been used as a rootstock and as parents in breeding rootstocks, providing durable tolerance to the insect pest phylloxera (Daktulosphaira vitifoliae) (Pongrácz 1983). V. rupestris Rupestris du Lot (syn. St. George), one of the most commonly used rootstocks, is a classic chloride excluder that can maintain low chloride levels in aerial organs of plants grown in saline soil (A. Walker, personal communication, 2013). V. rupestris is also a rich source for various levels of resistance to pathogens such as black rot (Takacs et al. 2014), powdery mildew (Barba et al. 2014), and downy mildew (Di Gaspero et al. 2012). In a recent report, alleles of the downy mildew resistance locus Rpv3 in more than half of 233 hybrid grape cultivars examined were traced back to V. rupestris parentage (Di Gaspero et al. 2012). Today, the significance of such a genetic resource is heightened as grape breeders strive to develop and maintain economically valuable cultivars for a changing climate (Hannah et al. 2013).

Previous analyses by Pavek et al. (2000, 2003) of natural V. rupestris populations from Texas to Pennsylvania assessed genetic variation using four microsatellite loci and identified seven target sites for in situ conservation of the species across its native range, two of which are in Missouri. Here, we expand on Pavek’s work by conducting detailed population genetic analyses in regions with the highest density of V. rupestris populations in North America, Missouri, and Oklahoma. The purpose of the present study was to measure allelic diversity, heterozygosity, and genetic differentiation at various levels of population structure in V. rupestris in the Ozark and Ouachita Mountains. The resulting data assess the genetic variation of extant V. rupestris populations and offer guidance for conservation efforts.

Materials and Methods

V. rupestris collections.

Six V. rupestris populations from four watersheds in the Ozark Mountains in Missouri and one from the Ouachita Mountains in Oklahoma were selected for this study (Supplemental Table 1). Three populations (Little Piney Creek, Wolf Creek, and Clifty Creek) were collected in the Gasconade River watershed (Missouri) within a 35 km radius around the confluence of the Little Piney and Gasconade rivers. Two populations, South Swan Creek and North Swan Creek, were surveyed in the White River watershed (Missouri). They are located along the same stream but are separated by 12 km. The Jack’s Fork population was sampled in the Black River watershed (Missouri). The Buzzard Creek population occurs in the Ouachita Mountains in Oklahoma. It is geographically distant from the Missouri populations (278 km from the nearest Ozark sampling site) and is separated from the Ozark Plateau by the Arkansas River Valley and the Boston Mountains (Supplemental Figure 1).

Leaf and stem samples were collected from 100 V. rupestris individuals from the seven populations between June and October 2012. The sampled populations ranged from four to 30 individuals. Vines were identified as V. rupestris based on growth habit and leaf morphological characteristics; these features distinguish V. rupestris from other sympatric grape species, notably V. riparia Michx. in Missouri and V. rotundifolia Michx. in Oklahoma. Samples were collected from vines with a shrub-like growth habit bearing reniform leaves in which the petiole and major veins displayed anthocyanin accumulation; the blade was thin, glabrous, and folded inward along the main leaf axis, and the blade width-to-length ratio was >1 (Munson 1909). Because V. rupestris frequently reproduces vegetatively, only vines that were separated by several meters were sampled to reduce the likelihood of collecting clones. Fresh samples were cooled and transported to the laboratory where young leaves were harvested and stored at −80°C for DNA extraction. Wood cuttings were stored at 4°C for rooting. Genomic DNA was extracted from frozen leaf tissue using a modified cetyltrimethylammonium bromide (CTAB) protocol (Doyle and Doyle et al. 1990, Di Gaspero and Cipriani 2002).

Microsatellite loci selection.

To identify informative microsatellite markers, randomly selected loci from across the grapevine genome and previously developed in other laboratories for V. vinifera L. and V. riparia were screened in a panel of 24 accessions (Supplemental Table 2). This panel included 21 V. rupestris accessions: five from each of our collection sites at Wolf Creek, South Swan Creek, North Swan Creek, and Buzzard Creek, and the rootstock V. rupestris cv. Rupestris du Lot. The panel also included the hybrid cultivar Munson (syn. Jaeger-70, V. aestivalis var. lincecumii × V. rupestris), and the V. vinifera cultivars Cabernet Sauvignon and Chardonnay as references for allele size. PCR amplification of microsatellite alleles using M13-tailed primers was performed in a 15-μL reaction volume containing 10 to 15 ng of template DNA, as described by Li et al. (2013), in a touchdown PCR program. The program consisted of an initial cycle at 96°C for 10 min followed by 10 cycles of denaturation at 94°C for 45 sec, annealing at 62°C for 45 sec, and extension at 72°C for 45 sec where the annealing temperature was decreased by 1°C in each subsequent touchdown cycle. This was followed by 30 cycles of amplification with 45-sec segments of 94, 56, and 72°C and by a final extension at 72°C for 5 min. The fragment length of PCR products was measured using capillary gel electrophoresis in a Beckman Coulter CEQ-8000 Genetic Analysis System. Evaluation of the resulting microsatellite allele sizes was conducted using the software package GeneMarker (Softgenetics, State College, PA). The 100 microsatellite markers were evaluated based on whether they amplified with PCR and, if amplifiable, on the number of alleles detected in the test panel.

Data analysis.

Genetic variation was estimated by calculating the number of alleles per population (Na), the number of effective alleles (Ne), observed heterozygosity (Ho), and Nei’s expected heterozygosity (He) using the software package GenAlEx 6.5 (Peakall and Smouse 2006, 2012). To test for significant deviation from Hardy–Weinberg equilibrium (HWE), χ2 tests were performed on observed and expected heterozygosity estimates. Population structure was characterized using Wright’s F-statistics and analysis of molecular variance (AMOVA) in GenAlEx 6.5. Principal coordinate analysis (PCoA) was conducted using the software DARwin 5.0.158 (Perrier and Jacquemoud-Collet 2006). The number of clusters based on microsatellite data was determined using STRUCTURE (Pritchard et al. 2000, Falush et al. 2003). After the initial runs, 10 runs were completed for each K-value (1 through 10) with a burn-in period of 200,000 and 4,000,000 Markov Chain Monte Carlo replicates using the admixture model. Calculation of the most probable K-value was determined using the likelihood value and by examining the highest second-order rate of change of K (Evanno et al. 2005). To test for isolation by distance, a Mantel test was performed with GeneAlEx 6.5 to examine whether a correlation existed between geographical distance and Nei’s genetic distance (Mantel 1967, Smouse et al. 1986).

Genetic variation and inbreeding estimates were calculated for all seven sampled V. rupestris populations. Characterization of population structure and differentiation focused on the Wolf Creek, Buzzard Creek, South Swan Creek, and North Swan Creek populations. The former three were included for their relatively large sample sizes (n ≥ 17), while North Swan Creek (n = 9) was included for its proximity and close relatedness to South Swan Creek, as indicated by our preliminary analyses.

Results

V. rupestris accessions.

This study surveyed variation at 14 microsatellite loci in 100 V. rupestris accessions from seven locations in the Ozark and Ouachita Mountains. The plants were identified as V. rupestris and were distinguished from natural hybrids derived from crosses with sympatric species (common hybrid parents were V. riparia and V. cinerea Engelm.) based on growth habit and key leaf morphological features (Munson 1909). Scanned leaves and shoots from a subset of our collection can be found online at http://people.missouristate.edu/LaszloKovacs/grapescans.htm. Successfully propagated genotypes are maintained in living collections at Missouri State University and will be submitted to USDA-ARS, Geneva, for ex situ conservation and for screening for breeding purposes.

Microsatellite amplification.

Of the 100 grape microsatellite markers scanned, 59 amplified at least one DNA fragment in the test panel. Of these, 14 loci that amplified consistently displayed an appropriate degree of polymorphism for population genetic analysis and were located on different chromosomes that were selected for genotyping the entire sample set of 100 V. rupestris accessions. These 14 loci were: VVC19 (Decrooq et al. 2003), VVMD27, VVMD24, VVMD5 (Bowers et al. 1996), VChr9a, VChr19a (Cipriani et al. 2008), VVIB23 (Merdinoglu et al. 2005), VrZag21, VrZag62, VrZag67 (Sefc et al. 1999), VVS4 (Thomas and Scott 1993), VMC2g2, VMC8g9, and VMC4d9 (Vitis Microsatellite Consortium). Each of these markers is located on a different grapevine chromosome. Genomic positions and primer sequences for these markers are available in the Probe database of NCBI (http://www.ncbi.nlm.nih.gov/probe). Although most loci amplified in most individuals, some loci failed to amplify in some taxa. For each individual, an average of 10.87 loci amplified, and each locus amplified in an average of 77 individuals. The absence of PCR products occurred across loci, genotypes, and populations, with no discernible trend in the geographic or genomic pattern of missing data. All loci that failed to amplify were treated as missing data.

Genetic diversity.

Genotyping 100 V. rupestris accessions at 14 microsatellite loci yielded 190 alleles (Table 1). The total number of alleles per locus ranged from 5 to 20, with an average of 13.57. The number of effective alleles for each locus for all of the collected samples ranged from 2.11 to 9.24 (average = 5.65). Thirty-four percent of observed alleles were private alleles that occurred in only one population (Table 2). Private alleles were detected in all populations surveyed, and the highest number (23) occurred in the Wolf Creek population (Supplemental Table 3). Estimates of diversity within populations are listed in Table 2. Average expected heterozygosity per population ranged from 0.57 to 0.72, and the average number of effective alleles per population per locus ranged from 2.79 to 4.06. Based on these values, the Wolf Creek population had the highest level of genetic diversity, consistent with our observation of a high number of private alleles in this population. Diversity estimates by microsatellite locus and population are shown in Supplemental Table 4.

View this table:
Table 1

Overall genetic variation at each microsatellite locus in V. rupestris samples.a

View this table:
Table 2

Genetic diversity and inbreeding coefficient estimates in seven V. rupestris populations. Values are means (standard error) across 14 loci.

Observed and expected genotype frequencies.

Analysis of observed and expected genotype frequencies across all populations revealed a deficiency in heterozygosity at all microsatellite loci (Table 1). Observed heterozygosity values ranged from 0.30 to 0.77 with an overall mean of 0.57, whereas expected heterozygosity values ranged from 0.53 to 0.89 with a grand mean of 0.79 (Table 1). To further examine the deficiency in heterozygosity, we tested whether genotype frequencies were in HWE in each population. Goodness-of-fit (χ2) tests demonstrated significant deviation (p < 0.05) from HWE at 13 of the 14 markers in at least one population (Table 3). The number of loci that deviated from HWE varied widely among the populations. In four populations (Jack’s Fork, Little Piney River, and North and South Swan Creek), 10 or more loci were in HWE. In the Clifty Creek, Buzzard Creek, and Wolf Creek populations, on the other hand, fewer than eight of the loci were in HWE. This deviation was particularly apparent for the latter two populations where the number of heterozygotes was lower than expected at more than half of the loci at high certainty (≥99%).

View this table:
Table 3

Inbreeding coefficient and significance of deviation from Hardy–Weinberg proportions in V. rupestris populations.a

Inbreeding coefficient (F) values calculated for the Jack’s Fork, Little Piney River, and North and South Swan Creek populations indicated very mild or no inbreeding. In contrast, inbreeding coefficient values for the Buzzard Creek, Wolf Creek, and Clifty Creek populations were 0.25, 0.22, and 0.22, respectively, indicating moderate inbreeding. These values were consistent with the preponderance of loci significantly deviating from HWE in these populations (Table 2).

Population structure.

To examine how genetic diversity was partitioned among populations, we conducted an AMOVA on the Buzzard Creek, Wolf Creek, South Swan Creek, and North Swan Creek microsatellite data. We focused on the former three larger populations, and we included North Swan Creek to explore its relatedness to the nearby South Swan Creek population. AMOVA results indicated moderate population differentiation (ΦPT = 0.105, p < 0.001), with 10% of the total variance due to among-population differences (Supplemental Table 5). Pairwise FST values were calculated to estimate the degree of differentiation among these four populations (Table 4). The resulting data clearly set Buzzard Creek, the only population in the Ouachita Mountains in Oklahoma, apart from the three populations in the Missouri Ozarks, with FST values ranging from 0.089 to 0.115. Within the three populations of the Ozark Plateau, the pairwise FST values ranged from 0.063 to 0.071 (Table 4).

View this table:
Table 4

Pairwise Fst values for four V. rupestris populations.

PCoA was performed to dissect differentiation among populations. The PCoA results explained 8.92% of variation along the x-axis and 5.11% of the variation along the y-axis (Figure 1). In the PCoA plot, South and North Swan Creek were clustered and appeared to be a single population, while the Wolf Creek and Buzzard Creek populations clustered separately. Population structure was further investigated using STRUCTURE. Estimating the number of clusters with Evanno’s method (Evanno et al. 2005) suggested that the metapopulation was most likely composed of two differentiated clusters (K = 2); STRUCTURE assigned 77 of the 79 total samples to one or the other cluster with a probability of 0.8 or higher. Of these 77 samples, 75 were assigned to their region of origin (the Ozark or Ouachita Mountains). When estimating the number of populations using only the Ozark samples, no distinct clusters were identified. Taken together, these data showed that, while discernible by PCoA and by the identification of private alleles, genetic differentiation among the Ozark populations is moderate. The Mantel test (Smouse et al. 1986) was performed to examine whether a correlation existed between geographical distance and Nei’s genetic distance. The null hypothesis that there is no relationship between geographic and genetic distance was rejected (p < 0.001, R2 = 0.0563) for the entire dataset (Figure 2), suggesting that genetic distance among these populations is correlated with geographic distance.

Figure 1
Figure 1

Principal coordinate analysis of Buzzard Creek, Wolf Creek, and North and South Swan Creek accessions. Ellipses show the 95% confidence threshold for each population.

Figure 2
Figure 2

Mantel test on the seven-population dataset examining the correlation between geographical distance (km) and Nei’s genetic distance (p < 0.001).

Discussion

North American grapevines are economically important as genetic resources for pest and pathogen resistance and environmental stress tolerance in grapevine rootstocks and for aromatic flavors integrated via hybridization with European grape cultivars (Reisch et al. 2012). While several valuable species, including V. riparia, V. berlandieri Planch. (syn. V. cinerea var. helleri [L.H. Bailey] M.O. Moore), and V. labrusca L. grow extensively across North America, others have more limited ecogeographic ranges and may be threatened by genetic erosion (Pavek et al. 2000). To our knowledge, no other North American grapevine species is experiencing a more dramatic decline than V. rupestris (Pavek et al. 2003). It has been proposed that the rapid decline of this species is due to human impacts on its unique and fragile habitat along riverbanks and dry beds of intermittent streams (Moore 1991, Pavek et al. 2003). Decline in population numbers is a concern not only for loss of breeding resources but also for the genetic prospects and vitality of wild V. rupestris. Our study stems from concerns about putative concomitant changes in genetic variation and structure associated with the apparent ongoing reduction in the number of populations. The results presented here show genetic differentiation among extant V. rupestris populations and possible inbreeding within some populations at the center of its natural range.

Genetic Diversity.

Compared to other non-Vitis perennial plants, the genetic diversity reported here for V. rupestris is comparable to that of a collection of natural and cultivated stands of Juglans regia in Italy (Pollegioni et al. 2011), wild populations of Populus simonii in East Asia (Wei et al. 2013), and wild stands of P. nigra in Europe (Slavov and Zhelev 2010). Furthermore, overall He values of 0.64 measured in natural populations of Pyrus calleryana in China (Liu et al. 2012), 0.77 in Malus sylvestris in France (Schnitzler et al. 2014), 0.69 in Quercus ilex in central Spain (Ortego et al. 2010), and 0.72 in Fagus sylvatica in northeastern Spain (Jump and Penuelas 2006) are comparable to the results reported here. These data suggest that, as expected, V. rupestris has a population structure similar to that of other natural populations of long-lived, outcrossing species (Loveless and Hamrick 1984, Petit and Hampe 2006).

In nature, wild grapevines reproduce predominantly by outcrossing enforced by a genetically encoded dioecious mating system (Picq et al. 2014). In general, outcrossing plants have high levels of genetic diversity within populations and low levels of among-population differentiation (Loveless and Hamrick 1984, Petit and Hampe 2006). As expected, V. rupestris populations were found to be rich in diverse alleles. In the populations examined, He ranged from 0.57 to 0.72, with an average value of 0.65 (Table 2). These values may be a slight underestimation of heterozygosity because we counted all nonamplified microsatellite markers as missing data, but several of these markers may have represented null alleles.

Our assessment of overall He was higher than that reported by Pavek et al. (2003), who measured an overall He of 0.50 in this species. The discrepancy was likely due to methodological differences, such as the numbers of markers and numbers of samples used in these two studies. Pavek et al. (2003) conducted their study on smaller sample sizes from populations from a wider geographic range using four microsatellite markers, whereas we employed larger sample sizes of fewer populations from a narrower section of the species’ natural range using 14 microsatellite markers.

Relative to other Vitis species, the overall He asessed here for V. rupestris was similar to that for natural stands of V. vinifera ssp. sylvestris in Europe measured with comparable nuclear microsatellite-based approaches. For example, in Portugal, Spain, France, and Hungary, He values of 0.66, 0.65, 0.73, and 0.74 were reported for natural populations of this species by Lopes et al. (2009), De Andres et al. (2012), Di Vecchi-Staraz et al. (2009), and Bodor et al. (2010), respectively. In Anatolia, overall genetic diversity of natural populations of this species tended to be higher than that in Europe. For example, Karataş et al. (2014) reported mean He of 0.79 for V. vinifera ssp. sylvestris in southeast Turkey, and Ergül et al. (2011) reported mean He of 0.81 in western and southern Turkey. Ergül et al. (2011) speculated that the higher He in Anatolian V. vinifera ssp. sylvestris populations was a consequence of this region being at the putative center of the species’ origin. More recently, Arroyo García and Ravilla (2013) suggested that the lower He in Spain and Portugal might have been a result of severe population fragmentation on the Iberian Peninsula. In general, our estimates for V. rupestris are similar to estimates for V. vinifera ssp. sylvestris from fragmented populations. This could reflect similarities in the effects of fragmentation history on natural Vitis populations. Alternatively, our relatively low diversity estimates could reflect an innate characteristic of V. rupestris. A more detailed comparative appraisal of V. rupestris diversity will be possible when comparable data become available for other wild American Vitis species.

Observed and expected genotype frequencies.

We detected statistically significant deviations from HWE at 13 of 14 microsatellite loci and measured inbreeding coefficient values that ranged from −0.09 to 0.25 among the seven sampled populations. In an obligate outcrossing species, such as V. rupestris, we expected populations to be near HWE. While we did find several populations near HWE, we also identified others that were moderately inbreeding, in two of which, Buzzard Creek and Wolf Creek, the moderately high inbreeding coefficients were supported by a preponderance of loci with highly significant deviation from HWE (Table 3). These results suggested that there was a tendency of breeding between relatives in at least a subset of the populations. This may be a result of the clumped distribution of closely related individuals, which, in turn, may be a reflection of the ease with which this species spreads vegetatively and forms prodigious clones (Pavek et al. 2000). Such sizeable clones produce large quantities of pollen or ovules, increasing the probability of mating between related plants in the proximity. The hypothesis that the Buzzard Creek and Wolf Creek populations included clusters of related plants could be tested by sampling and genotyping individual plants in close proximity to one another and by measuring the range of pollen and seed dispersal. Unfortunately, no ecological data are available on the dispersal agents and dynamics of pollen, seed, or vegetative propagules in V. rupestris. Such data, combined with genetic sampling, would serve not only basic plant ecology but would also guide in situ conservation efforts for the species.

Population structure.

Our pairwise FST and AMOVA estimates demonstrated that the sampled populations were genetically differentiated (Table 4). This differentiation was supported by the presence of private alleles in each population. Interestingly, the number of private alleles per population was proportional to sample size, suggesting that sampling more individuals would have uncovered even more private alleles. An examination of the geographical and genetic distance matrices using the Mantel test revealed that genetic differentiation among these sampled populations was likely a function of geographic distance (Figure 2). The PCoA also supported this observation; for example, the most geographically isolated population sampled in this study (Buzzard Creek in the Ouachita Mountains, Oklahoma) was genetically differentiated from all populations in the Missouri Ozarks (Figure 1), which were located ≥278 km from the Buzzard Creek site. PCoA also separated the Swan Creek and Wolf Creek populations, which were separated by 133 km. In contrast, PCoA was unable to separate the North and South Swan Creek samples, which were only 12 km apart along the same river (Figure 1). This suggested that North Swan Creek and South Swan Creek were part of the same mating population connected by gene flow through the movement of seed, pollen, or vegetative propagules such as stem segments. The pattern of isolation by distance was also supported by STRUCTURE analysis, which identified two likely clusters: one composed mostly of Ozark samples, and the other consisting of the geographically distant Buzzard Creek samples.

Correlation between geographic and genetic distance suggested a modest but statistically significant relationship between these factors in V. rupestris populations. Many factors may contribute to this pattern, including dispersal dynamics and genetic drift. These effects may be magnified by the small size and sparse distribution of V. rupestris populations. However, we cannot exclude the possibility that the observed correlation between geographic and genetic distance may reflect signatures of adaptation to spatially autocorrelated climates and ecogeographic conditions. For example, the Buzzard Creek site in the Ouachita Mountains has a considerably warmer and more humid climate than the sites in the Ozarks; therefore, forces of natural selection may differ between these two mountain ranges. Another factor that may contribute to genetic differentiation is introgression of alleles from different species of wild grapevines. Missouri populations of V. rupestris, for example, are sympatric with V. riparia, but the Oklahoma population is outside the natural range of V. rupestris.

The questions as to whether inbreeding and genetic differentiation are innate attributes of V. rupestris populations and to what extent these characteristics make these plants vulnerable to genetic erosion are relevant to basic plant ecology, conservation of this species, and, by extension, future efforts aimed at incorporating V. rupestris germplasm into grape breeding programs. An exploration of these questions will inform conservation efforts by the US Department of Agriculture (Pavek et al. 2003) and other agencies to safeguard against loss of genetic diversity of the species. Unfortunately, data on the reproductive ecology of V. rupestris are not available, and information on its population distribution and genetics are scarce. Growing in discontinuous populations and along river beds, V. rupestris is well suited to studies of the influence of riparian and lotic environments on gene flow. For example, is gene flow between North and South Swan Creek plants a result of their close proximity or is it mediated by the movement of water along the creek? To what extent does vegetative propagation play a role in gene flow along rivers? Interestingly, of the numerous American grapevine species, only the two characteristic riparian species, V. rupestris and V. riparia, readily form adventitious roots. To what extent do birds and mammals contribute to seed dispersal? Examining these questions is warranted, as our population-level allelic diversity data suggest that natural V. rupestris populations carry unique genetic variants that represent unexplored potential for grape breeding.

Conclusions

In summary, at least two major genetic clusters of V. rupestris populations exist in the heart of its range, one in Missouri and one in Oklahoma. This differentiation appears to be the result of contemporary random genetic drift in small populations but also could have been caused by geographical isolation or a past population bottleneck. We detected moderate inbreeding in some populations and no excess homozygosity in others. Thus, it is possible that genetic drift, in some cases exacerbated by inbreeding, leads to increased homozygosity in V. rupestris, a condition known to be deleterious in grapevine (Mullins et al. 1992). To determine whether these conditions make this species susceptible to genetic erosion requires more extensive ecological and population genomic analyses.

Acknowledgments

Acknowledgments: The authors thank Courtney Coleman, Justin Fay, Laura Klein, and Anthony Peccoux for their assistance in collecting wild grapevine accessions; Chin-Feng Hwang for providing primers to screen for polymorphic microsatellites; and Jianglu Wang for assistance in digital scanning of specimens. This work was supported by a Faculty Research Grant and the M.S. in Plant Science programs at Missouri State University and by funds from the NSF Grape Genomics Research Coordination Network Grant (NSF grant DBI 0741876).

Footnotes

  • Supplemental data is freely available with the online version of this article at www.ajevonline.org.

  • Received February 2015.
  • Revision received April 2015.
  • Accepted May 2015.
  • Published online October 2015

Literature Cited

  1. ↵
    1. Arroyo García, R.A., and
    2. E. Ravilla
    . 2013. The current status of wild grapevine populations (Vitis vinifera ssp. sylvestris) in the Mediterranean Basin. In The Mediterranean Genetic Code – Grapevine and Olive. Poljuha, D., and B. Sladonja (eds.). InTech, www.intechopen.com.
  2. ↵
    1. Barba, P.,
    2. L. Cadle-Davidson,
    3. J. Harriman,
    4. J.C. Glaubitz,
    5. S. Brooks,
    6. K. Hyma, and
    7. B. Reisch
    . 2014. Grapevine powdery mildew resistance and susceptibility loci identified on a high-resolution SNP map. Theor. Appl. Genet. 127:73–84.
    OpenUrlCrossRefPubMed
  3. ↵
    1. Bodor, P.,
    2. M. Höhn,
    3. A. Pedryc,
    4. T. Deák,
    5. I. Dücső,
    6. I. Uzun,
    7. K. Cseke,
    8. É.I. Böhn, and
    9. G.D. Bisztray
    . 2010. Conservation value of the native Hungarian wild grape (Vitis sylvestris Gmel.) evaluated by microsatellite markers. Vitis 49:23–27.
    OpenUrl
  4. ↵
    1. Bowers, J.E.,
    2. G.S. Dangl,
    3. R. Vignani, and
    4. C.P. Meredith
    . 1996. Isolation and characterization of new polymorphic simple sequence repeat loci in grape (Vitis vinifera L.). Genome 39:628–633.
    OpenUrlPubMed
  5. ↵
    1. Cipriani, G.,
    2. M.T. Marrazzo,
    3. G. Di Gaspero,
    4. A. Pfeiffer,
    5. M. Morgante, and
    6. R. Testolin
    . 2008. A set of microsatellite markers with long core repeat optimized for grape (Vitis spp.) genotyping. BMC Plant Biol. 8:127.
    OpenUrlPubMed
  6. ↵
    1. De Andrés, M.T.,
    2. et al.
    2012. Genetic diversity of wild grapevine populations in Spain and their genetic relationships with cultivated grapevines. Mol. Ecol. 21:800–816.
    OpenUrlCrossRefPubMed
  7. ↵
    1. Decrooq, V.,
    2. M.G. Favé,
    3. L. Hagen,
    4. L. Bordenave, and
    5. S. Decrooq
    . 2003. Development and transferability of apricot and grape EST microsatellite marker across taxa. Theor. Appl. Genet. 106:912–922.
    OpenUrlPubMed
  8. ↵
    1. Di Gaspero, G., and
    2. G. Cipriani
    . 2002. Resistance gene analogs are candidate markers for disease-resistance genes in grape (Vitis spp.). Theor. Appl. Genet. 106:163–172.
    OpenUrlPubMed
  9. ↵
    1. Di Gaspero, G.,
    2. et al.
    2012. Selective sweep at the Rpv3 locus during grapevine breeding for downy mildew resistance. Theor. Appl. Genet. 124:277–286.
    OpenUrlPubMed
  10. ↵
    1. Di Vecchi-Staraz, M.,
    2. V. Laucou,
    3. G. Bruno,
    4. T. Lacombe,
    5. S. Gerber,
    6. T. Bourse,
    7. M. Boselli, and
    8. P. This
    . 2009. Low level of pollen-mediated gene flow from cultivated to wild grapevine: Consequences for the evolution of the endangered subspecies Vitis vinifera L. subsp. sylvestris. J. Hered. 100:66–75.
    OpenUrlAbstract/FREE Full Text
  11. ↵
    1. Doyle, J.J., and
    2. J.L. Doyle
    . 1990. Isolation of plant DNA from fresh tissue. Focus 12:13–15.
    OpenUrl
  12. ↵
    1. Ergül, A.,
    2. G. Perez-Rivera,
    3. G. Söylemezoğlu,
    4. K. Kazan, and
    5. R.A. Arroyo García
    . 2011. Genetic Diversity in Anatolian wild grapes (Vitis vinifera subsp. sylvestris) estimated by SSR markers. Plant Gen. Res. 9:375–383.
    OpenUrl
  13. ↵
    1. Evanno, G.,
    2. S. Regnaut, and
    3. J. Goudet
    . 2005. Detecting the number of clusters of individuals using the software STRUCTURE: A simulation study. Mol. Ecol. 14:2611–2620.
    OpenUrlCrossRefPubMed
  14. ↵
    1. Falush, D.,
    2. M. Stephens, and
    3. J.K. Pritchard
    . 2003. Inference of population structure using multilocus genotype data: Linked loci and correlated allele frequencies. Genetics 164:1567–1587.
    OpenUrlAbstract/FREE Full Text
  15. ↵
    1. Hannah, L.,
    2. P.R. Roehrdanz,
    3. M. Ikegami,
    4. A.V. Shepard,
    5. M.R. Shaw,
    6. G. Tabor,
    7. L. Zhi,
    8. P.A. Marquet, and
    9. R.J. Hijmans
    . 2013. Climate change, wine, and conservation. P. Natl. Acad. Sci. USA. 110:6907–6912.
    OpenUrlAbstract/FREE Full Text
  16. ↵
    1. Jump, A.S., and
    2. J. Peñuelas
    . 2006. Genetic effects of chronic habitat fragmentation in a wind-pollinated tree. P. Natl. Acad. Sci. USA. 103:8096–8100.
    OpenUrlAbstract/FREE Full Text
  17. ↵
    1. Karataş, D.D.,
    2. H. Karataş,
    3. V. Laucou,
    4. G. Sarikamis,
    5. L. Riahi,
    6. R. Bacilieri, and
    7. P. This
    . 2014. Genetic diversity of wild and cultivated grapevine accessions from southeast Turkey. Hereditas 151:73–80.
    OpenUrlPubMed
  18. ↵
    1. Kartesz, J.T.
    2013. The Biota of North America program (BONAP 2013. North American Plant Atlas, Chapel Hill, NC [maps generated from J.T. Kartesz. (2013) Floristic Synthesis of North America Version 1.0, Biota of North America Program (BONAP), (in press)]. Available at http://bonap.net/napa. Accessed 4 Feb 2015.
  19. ↵
    1. Li, C.,
    2. et al.
    2013. Selection for Run1-Ren1 dihybrid grapevines using microsatellite markers. Am. J. Enol. Vitic. 64:152–155.
    OpenUrlAbstract/FREE Full Text
  20. ↵
    1. Liu, J.,
    2. X. Zheng,
    3. D. Potter,
    4. C. Hu, and
    5. Y. Teng
    . 2012. Genetic diversity and population structure of Pyrus calleryana (Rosaceae) in Zhejiang province, China. Biochem. Syst. Ecol. 45:69–78.
    OpenUrl
  21. ↵
    1. Lopes, M.S.,
    2. D. Mendonca,
    3. M. Rodrigez dos Santos,
    4. J.E. Eiras-Dias, and
    5. A. da Câmara Machado
    . 2009. New insights on the genetic basis of Portuguese grapevine and on grapevine domestication. Genome 52:790–800.
    OpenUrlPubMed
  22. ↵
    1. Loveless, M.D., and
    2. J.L. Hamrick
    . 1984. Ecological determinants of genetic structure in plant populations. Ann. Rev. Ecol. Syst. 13:65–95.
    OpenUrl
  23. ↵
    1. Mantel, N.
    1967. The detection of disease clustering and a generalized regression approach. Cancer Res. 27:209–220.
    OpenUrlAbstract/FREE Full Text
  24. ↵
    1. Merdinoglu, D.,
    2. G. Butterlin,
    3. L. Bevilacqua,
    4. V. Chiquet,
    5. A.F. Adam-Blondon, and
    6. S. Decroocq
    . 2005. Development and characterization of a large set of microsatellite markers in grapevine (Vitis vinifera L.) suitable for multiplex PCR. Mol. Breeding 15:349–366
    OpenUrl
  25. ↵
    1. Moore, M.O.
    1991. Classification and systematics of eastern North American Vitis L. (Vitaceae) north of Mexico. SIDA, Contributions to Botany 14:339–367.
    OpenUrl
  26. ↵
    1. Mullins, M.G.,
    2. A. Bouquet, and
    3. L.E. Williams
    . 1992. Biology of the Grapevine. Cambridge University Press, Cambridge, UK.
  27. ↵
    1. Munson, T.V.
    1909. Foundations of American Grape Culture. T.V. Munson and Son, Denison, TX.
  28. ↵
    NatureServe. 2014 NatureServe Explorer: An online encyclopedia of life [web application]. Version 7.1. NatureServe, Arlington, Virginia. Available at http://explorer.natureserve.org. Accessed 4 Feb 2015.
  29. ↵
    1. Ortego, J.,
    2. R. Bonal, and
    3. A. Muñoz
    . 2010. Genetic consequences of habitat fragmentation in long-lived tree species: The case of the Mediterranean Holm Oak (Quercus ilex, L.). J. Hered. 101:717–726.
    OpenUrlAbstract/FREE Full Text
  30. ↵
    1. Pavek, D.S.,
    2. W.F. Lamboy, and
    3. E.J. Garvey
    . 2000. Ecogeographic study of Vitis species: Final Report for Rock Grape (Vitis rupestris). USDA-ARS National Germplasm Resources Lab, Beltsville, MD.
  31. ↵
    1. Pavek, D.S.,
    2. W.F. Lamboy, and
    3. E.J. Garvey
    . 2003. Selecting in situ conservation sites for grape genetic resources in the USA. Genet. Resour. Crop Ev. 50:165–173.
    OpenUrl
  32. ↵
    1. Peakall, R., and
    2. P.E. Smouse
    . 2006. GENALEX 6: Genetic analysis in Excel. Population genetic software for teaching and research. Mol. Ecol. Notes 6:288–295.
    OpenUrlCrossRef
  33. ↵
    1. Peakall, R., and
    2. P. E Smouse
    . 2012. GenAlEx 6.5: Genetic analysis in Excel. Population genetic software for teaching and research - an update. Bioinformatics 28:2537–2539.
    OpenUrlAbstract/FREE Full Text
  34. ↵
    1. Perrier, X., and
    2. J.P. Jacquemoud-Collet
    . 2006. DARwin software. Available at http://darwin.cirad.fr/darwin. Accessed 4 Feb 2015.
  35. ↵
    1. Petit, R.J., and
    2. A. Hampe
    . 2006. Some evolutionary consequences of being a tree. Ann. Rev. Ecol. Syst. 37:187–214.
    OpenUrlCrossRef
  36. ↵
    1. Picq, S.,
    2. et al.
    2014. A small XY chromosomal region explains sex determination in wild dioecious V. vinifera and the reversal to hermaphroditism in domesticated grapevines. BMC Plant Biol. 14:229.
    OpenUrlPubMed
  37. ↵
    1. Pollegioni, P.,
    2. K. Woeste,
    3. I. Olimpieri,
    4. D. Marandola,
    5. F. Cannata, and
    6. M.E. Malvolti
    . 2011. Long-term human impacts on genetic structure of Italian walnut inferred by SSR markers. Tree Genet. Genomics 7:707–723.
    OpenUrl
  38. ↵
    1. Pongrácz, D.P.
    1983. Rootstocks for Grapevines. David Philip Publisher, Cape Town, South Africa.
  39. ↵
    1. Pritchard, J.K.,
    2. M. Stephens, and
    3. P. Donnelly
    . 2000. Inference of population structure using multilocus genotype data. Genetics 155:945–959.
    OpenUrlAbstract/FREE Full Text
  40. ↵
    1. Reisch, B.I.,
    2. C.L. Owens, and
    3. P.S. Cousins
    . 2012. Grape. In Fruit Breeding. Badenes, M.L., and D.H. Byrne (eds.), pp. 225–262. Springer Science+Business Media LLC, Philadelphia, PA.
  41. ↵
    1. Schnitzler, A.,
    2. C. Arnold,
    3. A. Cornille,
    4. O. Bachmann, and
    5. C. Schnitzler
    . 2014. Wild European apple (Malus sylvestris (L.) Mill.) population dynamics: Insight from genetics and ecology in the Rhine Valley. Priorities for a future conservation program. PLoS ONE 9:e96569.
    OpenUrlPubMed
  42. ↵
    1. Sefc, K.M.,
    2. F. Regner,
    3. E. Turetschek,
    4. J. Glossl, and
    5. H. Steinkellner
    . 1999. Identification of microsatellite sequences in Vitis riparia and their applicability for genotyping of different Vitis species. Genome 42:367–373.
    OpenUrlPubMed
  43. ↵
    1. Slavov, G.T., and
    2. P. Zhelev
    . 2010. Salient biological features, systematics, and genetic variation of Populus. In Genetics and Genomics of Populus. Jansson, S., et al. (eds.), pp. 15–38. Springer, New York.
  44. ↵
    1. Smouse, P.E.,
    2. J.C. Long, and
    3. R.R. Sokal
    . 1986. Multiple regression and correlation extensions of the Mantel test of matrix correspondence. Syst. Zool. 35:627–632.
    OpenUrlFREE Full Text
  45. ↵
    1. Takacs, E.M.,
    2. A.D. Isby,
    3. P.M. Appleton, and
    4. B.I. Reisch
    . 2014. Delineating the mechanism and inheritance of black rot resistance from Vitis rupestris and V. cinerea. In Proceedings of the 11th International Conference on Grapevine Breeding and Genetics, Beijing, China.
  46. ↵
    1. Thomas, M.R., and
    2. N.S. Scott
    . 1993. Microsatellite repeats in grapevine reveal DNA polymorphisms when analysed as sequenced-tagged sites (STSs). Theor. Appl. Genet. 86:985–990.
    OpenUrlPubMed
  47. ↵
    1. Wei, Z.,
    2. Q. Du,
    3. J. Zhang,
    4. B. Li, and
    5. D. Zhang
    . 2013. Genetic diversity and population structure in Chinese indigenous poplar (Populus simonii) populations using microsatellite markers. Plant Mol. Biol. Rep. 31:620–632.
    OpenUrlCrossRef
Back to top
Print
View full PDF
Email Article
Citation Tools

Population Structure of Vitis rupestris, an Important Resource for Viticulture
Dániel Pap, Allison J. Miller, Jason P. Londo, László G. Kovács
Am J Enol Vitic.  2015  66: 403-410  ; DOI: 10.5344/ajev.2015.15012
del.icio.us logo Twitter logo Facebook logo Mendeley logo
Save to my folders

Jump to section