| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Population Biology |
Center for Population Biology, One Shields Avenue, University of California, Davis, California 95616 USA
Received for publication April 1, 2005. Accepted for publication July 27, 2005.
ABSTRACT
Polyploid speciation is an ongoing, important source of angiosperm diversity. However, the barriers to polyploid speciation and mechanisms of establishment remain poorly understood for all but a few species. Several factors likely to have influenced tetraploid establishment, including barriers to triploid formation, consequences of mixed-ploidy pollen loads, and the reproductive success of the minority cytotype, were examined in snow buttercups (Ranunculus adoneus) through a series of pollination and transplant experiments. Tetraploid snow buttercups do not have stigmatic barriers to pollen from diploid plants, nor does pollen from tetraploid plants have an advantage over pollen from diploids when on tetraploid stigmas. Tetraploid plants transplanted into a diploid population produced 50% fewer seeds than tetraploid plants in a tetraploid population. Intrinsic barriers to triploid formation were relatively weak, but few triploid seeds formed when mixed-ploidy pollen was present. Fecundity of triploid plants was very low, and no tetraploid offspring resulted. These results indicate that in snow buttercups, a triploid plant will contribute 0.8% of the tetraploid seeds of a minority tetraploid plant making it a minor contributor to the demographics of tetraploid establishment. The reproductive costs facing minority cytotype plants may explain the previously observed spatial segregation in snow buttercups.
Key Words: flow cytometry • hybridization • reproductive isolation • pollination • polyploid speciation • Ranunculus
Polyploid speciation has played an important role in the creation of plant diversity; 30 to 70% of plant species have doubled or more than doubled chromosome complements (Stebbins, 1970
; Masterson, 1994
). Doubling the chromosome number usually results in strong postzygotic barriers to hybrids between the parental population and the newly formed polyploid plants through barriers to seed formation, viability, and fertility in triploid plants. However, minority polyploid plants should face a strong reproductive penalty unless prezygotic isolating mechanisms prevent matings with the progenitor cytotype. This should result in a high barrier to the establishment of a polyploid species (Levin, 1975
).
Several factors could intervene at different stages of reproduction to increase the success of minority tetraploid plants. First, the tetraploid plants could be self-pollinating, as are many, but far from all, polyploid taxa (Mable, 2004
). Changes in the timing of flowering, pollinator preference, or pollinator constancy could result in minority tetraploid plants transferring pollen to other tetraploids despite their minority status (Lewis, 1976
; Segraves and Thompson, 1999
; Husband, 2000
). Ecological differentiation between the cytotypes could lead to their spatial isolation, decreasing the proportion of pollen from diploid plants reaching tetraploid stigmas (Ehrendorfer, 1979
; Husband and Schemske, 1998
). Limited seed and pollen dispersal could lead to clustering of each cytotytpe and local tetraploid majorities (Li et al., 2004
; Baack, 2005
). Finally, the cost of cross-ploidy pollinations could be offset if fertile triploid plants produce tetraploid offspring (Felber, 1991
; Felber and Bever, 1997
) or high levels of diploid gamete production by diploid plants lead to steady input of tetraploid seedlings (Bretagnolle and Thompson, 1995
).
Postpollination, prezygotic isolating mechanisms may also alter the outcome of cross-ploidy pollen transfer. The proportion of tetraploid seeds produced will depend on the ability of pollen grains to germinate on the stigmatic surface and grow down the style. If pollen from tetraploid plants has an advantage on tetraploid stigmas, then even in the event of mixed deposition of pollen from diploid and tetraploid plants, the tetraploid plants would produce tetraploid seeds, greatly reducing the cost of being in the minority (Williams et al., 1999
; Husband et al., 2002
). The combination of barriers to cross-ploidy pollinations and stigmatic barriers may explain why declines in tetraploid reproduction due to minority status are not always documented when diploid and tetraploid plants are found in sympatry (e.g., Husband, 2000
).
I examined the reproductive obstacles facing minority tetraploid snow buttercups (Ranunculus adoneus) through experimental pollinations, open pollination of experimental transplants, and observations of naturally occurring mixtures of diploids and tetraploids over a 3-yr period. The two cytotypes show strong spatial segregation: only one site of the 30 surveyed had both diploid and tetraploid plants, and at that site, the shift from 90% diploid to 90% tetraploid occurred over 3 m (Baack, 2004
). In experimental transplants between populations, we found no evidence of ecological differentiation between diploid and tetraploid plants during the seed or seedling stages of the life cycle (Baack and Stanton, 2005
). Similar patterns in other species have led to the hypothesis that reproductive interference between the two cytotypes could explain their allopatric distribution (van Dijk and Bakx-Schotman, 1997
). In this series of experiments, I examined the reproductive success of tetraploid plants when surrounded either by other tetraploid plants or by diploid plants. When tetraploid plants are in the minority, they should encounter a mixture of pollen from diploid and tetraploid plants. This could lead to decreases in seed production and/or the increased production of triploid seeds relative to the number of tetraploid seeds. The particular outcome will depend in part on the intrinsic barriers to triploid seed development and the relative success of pollen from diploid and tetraploid plants on tetraploid stigmas. In particular, I examined the barriers to the formation of triploid seeds on tetraploid plants in several different pollination scenarios.
MATERIALS AND METHODS
Study system
The snow buttercup (Ranunculus adoneus Gray, Ranunculaceae) is a long-lived perennial found in the alpine snow beds of the Rocky Mountains, Colorado, USA. Plants emerge at the edge of the melting snow and flower within a few days of emergence. The protogynous flowers persist for approximately 10 days; secondary flowers on some plants open a week or two after the primary flowers. Snow buttercups are self-compatible, with insect visitation required for complete pollen transfer (Stanton and Galen, 1989
); estimates of selfing derived from allozyme frequencies in one tetraploid population ranged from 30 to 70% (Stanton et al., 1997
). Flowers are heliotropic and attract a variety of pollinators, particularly small flies (Stanton and Galen, 1989
). Receptacles bear 50–150 stigmas, which mature over several days from the top of the receptacle downwards. Photosynthetic achenes develop from the fertilized ovules (Galen et al., 1993
) and disperse mainly by gravity 3–5 weeks after fertilization. Depending on the size and shape of a snow bed, flowering within a population can extend over several weeks as the edge of the snow gradually retreats from the periphery to the center.
Snow buttercups are either diploid (2n = 2x = 16) or tetraploid (2n = 4x = 32) (Wiens and Halleck, 1962
; Kapoor and Love, 1970
; Stanton et al., 1997
). The two cytotypes are very similar, occasionally distinguishable on the basis of leaf morphology, but never on floral morphology (E. Baack, personal observation). Diploid and tetraploid populations are in close geographic proximity in the Mosquito and Ten Mile ranges of Colorado (Baack, 2004
).
Two populations near Fairplay, Colorado (CO) were used in this study. The Pennsylvania Mountain (Park County, CO) population lies within the University of Colorado, Colorado Springs' Alpine and Arctic Research Station. My study used the "Slope One" population (39°09'06'' N, 106°04'19'' W), which covers a 20 x 50 m patch sloping gently east at an elevation of 3500 m. In 2000 and 2001, snowmelt occurred over 2 weeks for the entire population. Plants selected for this study emerged in the first 4 days of snowmelt. All plants examined from Pennsylvania Mountain are tetraploid (2n = 4x = 32), based on cytological examination of pollen mother cells (Stanton et al., 1997
) or flow cytometry (N = 40, see section Cytotyping of seeds)
The second population in this study occurs on the E and NE slopes of Tundra Ridge at Hoosier Pass (Summit County, CO) on land managed by the Dillon Ranger District, Arapahoe National Forest, US Forest Service (39°13'12'' N, 106°02'31'' W). The snow bed here lies at a steep angle, forming a band 30 m wide and 450 m long. Snowmelt occurred over 4 weeks in 2000 and 6 weeks in 2001. Flow cytometric studies (Baack, 2004
) revealed that the population is complex, with diploids occupying the southern half of the area and tetraploids the northern half. Diploid plants used in cross-pollination studies in 2000 and 2001 were taken from the southern edge of the population, 
200 m away from the tetraploid populations, and emerged in the first week of snowmelt, simultaneous with the emergence of plants at Pennsylvania Mountain. In 2002, I used plants emerging in the first 10 days after snowmelt from both the diploid and tetraploid zones.
Cross- and mixed-ploidy pollination experiments
I addressed two questions using hand pollinations in 2000 and 2001. First, how does cross-ploidy pollination affect seed set compared with within-ploidy pollination? Second, how does a mixture of pollen from both diploids and tetraploids affect the seed set compared to within-ploidy pollinations?
Plants were caged at Pennsylvania Mountain (tetraploid) and Hoosier Pass (diploid) within 2 days of emergence, before flowers had opened. Cages were 15 cm high x 30 cm diameter cylinders of hardware cloth wrapped with two layers of bridal veil, with a wire closure at the top. The lower edges of bridal veil were sealed with mud to prevent the entry of crawling insects. At the time of flower opening, anthers were removed, and each flower was randomly assigned to one of three pollination treatments: within-ploidy, with pollen from either diploid or tetraploid plants, matching the maternal cytotype; cross-ploidy, with pollen from plants of the other ploidy from the maternal cytotype; or mixed-ploidy, with pollen from a diploid and a tetraploid plant to pollinate a flower. I used only the larger primary flowers in this study, removing all secondary flowers prior to opening. I tied colored embroidery thread to the pedicel below each flower to indicate the treatment applied. Thirty flowers were used in each population for each treatment in 2000, from 52 diploid and 58 tetraploid plants. In 2001, a late snowstorm limited the number of diploid flowers available, so 40 flowers from 20 plants were used for the within-cytotype and mixed-cytotype pollinations in the diploid populations. Forty and 36 flowers, from a total of 58 tetraploid plants, were used for the within and mixed-ploidy treatments. The cross-ploidy pollination treatment was not carried out in either population in 2001. Each pollination treatment was applied to a maximum of one flower on any given plant.
Flowers were pollinated three times over the course of 8 days. Pollen donor flowers were caged 1 day prior to use to reduce pollen loss to pollinators. On the morning of the pollination, I harvested donor flowers at each site, with the stems kept in water to maintain freshness. Due to low rates of pollen production in snow buttercups, collecting pollen in a tube and applying it with a brush was ineffective. Instead, I removed petals from the donor flower then brushed its anthers against stigmas of the pollen recipient, with each donor used once on any given day. Donor flowers were kept overnight, and those showing sufficient pollen during visual inspection on the next day were used a second time. Pollinations were carried out every third day, with diploid and tetraploid plants receiving pollen on separate days. On any given day, within-cytotype pollinations used a single donor flower of the same cytotype, cross-cytotype pollinations used a single donor flower of the alternate cytotype, while mixed-ploidy pollinations used two flowers in quick succession, one of each cytotype. I alternated the order of application of the two cytotypes used in mixed-ploidy pollinations between recipient flowers and between days to sample equally in case of priority effects. Although self-pollination is likely important in both cytotypes, the necessity of harvesting flowers to apply pollen over several days prevented the use of self-pollen in this experiment.
After recipient flowers lost their petals and were clearly no longer receptive to pollen (approximately 12 days after opening), the bridal veil top was removed from the cages to allow more solar radiation to reach the photosynthetic achenes. Seed heads were bagged with bridal veil in the week preceding seed collection to prevent the loss of experimental seeds. Achenes were sorted to exclude unfilled or under-developed seeds, then counted. All achenes from each flower were combined and weighed on a microgram balance at the University of California, Davis (Mettler-Toledo UMT2, Greifensee, Switzerland).
Because some experiments were carried out over 2 years while others were completed in a single year, I analyzed each experiment separately. Flowers producing no seeds were excluded from the individual seed mass analysis. Different data sets were used for the seed count and seed mass analyses, precluding the use of MANOVA. I analyzed the barriers to triploid formation by comparing the 2000 within-ploidy and cross-ploidy pollinations using ANOVA (PROC GLM, SAS Institute, 2000). Seed counts were square-root transformed to improve the fit to model assumptions, while the mean mass per seed was natural-log (ln) transformed. I examined the effect of maternal cytotype, pollen treatment, and their interaction on seed number and seed mass. Comparisons between pollen treatments within each maternal cytotype were made using preplanned contrasts. Mean-square errors and associated degrees of freedom from the ANOVA were used to calculate F tests for contrasts. For the test of the effects of mixed-ploidy pollen loads, I compared within-ploidy and mixed-ploidy pollinations carried out in 2000 and 2001. Plant ploidy, pollination treatment, year and all interactions were tested, with year treated as a fixed factor (PROC GLM, SAS Institute, 2000). Nonsignificant interactions (P > 0.25) were dropped from the model for simplicity and greater statistical power. Year interacted with pollination treatment in the analysis of seed mass (F1,192 = 3.19, P = 0.12), so the two years were analyzed separately for seed mass. Pairwise differences between the within-ploidy treatment and the mixed-ploidy pollination treatment within each maternal ploidy level were tested using preplanned contrasts.
Cytotyping of seeds produced by experimental pollinations
I used flow cytometry, which measures the DNA content per nucleus, to verify that seeds produced by cross-ploidy pollinations were triploid and to investigate the ploidy of seeds produced in the mixed-ploidy crosses. Individual seeds were minced finely with a double-edged razor blade in chopping buffer [100 mM Tris, 0.1% Triton X-100, 1 mM MgCl2, 2 µg/mL 4',6-diamidino-2-phenylindole (DAPI), pH 7.5], filtered through two layers of Miracloth (California Biochem, Pasadena, California, USA), then run on a Partec Cell Analyzer II flow cytometer (Munster, Germany). I verified published chromosome counts using root-tip squashes on plants from those populations, then used leaves from these plants as cytotype standards. During flow cytometry, I calibrated each run using these known standards and checked the calibration every 30 min.
Each seed typically produced two fluorescence peaks, one for the embryo and one for the endosperm. Seeds from diploid–diploid crosses had the predicted 2 : 3 embryo to endosperm ratio, while tetraploid seeds had a 4 : 6 embryo to endosperm ratio. Seeds produced by the cross-ploidy pollinations with a tetraploid maternal parent produced a 3 : 5 ratio, while those with a diploid maternal parent produced a 3 : 4 ratio, with the embryo DNA content halfway between the diploid and tetraploid plant values. These distinct ratios made it possible to classify the ploidy of the embryo of each seed as diploid, triploid, or tetraploid. I scored 25 seeds from cross-ploidy pollinations as well as 10 mixed-ploidy crosses from each maternal cytotype, with 3–8 seeds scored per cross.
Examination of stigmatic and stylar barriers
In a small study to explore the possibility that pollen of different cytotypes varies in its ability to germinate on tetraploid stigmas, I sampled stigmas from tetraploid flowers 24 h following pollination by diploid (N = 5), tetraploid (N = 6), or mixed-ploidy (N = 8) pollen loads. One to seven pistils were removed from each flower and fixed in 3 : 1 ethanol to acetic acid. Pistils were cleared in 1 N NaOH at 65°C for 1 h, then rinsed in distilled water. Pollen tubes were stained with aniline blue (0.1% aniline blue, 0.35% tribasic potassium phosphate; Kearns and Inouye, 1993
) and examined under UV at 100x. I counted the number of pollen tubes emerging from pollen grains at the top of the stigma as well as the number of pollen tubes observed at the bottom of the style. Direct observation of pollen tube entry into the ovule was not possible. Comparisons of the mean number of emerging pollen tubes and the fraction of emergent tubes to the number of tubes reaching the bottom of the style were made between classes of crosses using ANOVA (PROC GLM, SAS Institute, 2000). Counts of emerging pollen tubes were square-root transformed, and the proportions were square-root and arcsine transformed to improve fit to model assumptions.
Experimental assessment of reproductive costs of minority cytotype status
Minority cytotype plants might be able to reproduce successfully through self-fertilization. However, this possibility was not examined in the previously described experimental pollinations. Sufficient observations of fecundity on naturally occuring minority plants were not possible due to the limited number of minority individuals found on either side of the contact zone in the one mixed population in Colorado (Baack, 2004
). Accordingly, to assess the reproductive costs of minority cytotype status, I performed experimental transplants in the neighboring diploid and tetraploid areas at Hoosier Pass. Emerging plants with two primary flowers were selected in both diploid (N = 180) and tetraploid habitats (N = 162). Half of the selected plants were transplanted into the habitat of the other cytotype (approximately 350 m away), the other half back into their original habitat (0.2–0.8 m from its initial site). Each transplanted plant was at least 0.8 m away from the nearest transplant of the same cytotype. This resulted in experimental minority and majority cytotype plants of each cytotype.
I bagged and hand-pollinated one of the two flowers on each plant to serve as a control for plant vigor and microsite variation, while I left the other flower open to natural pollinators. Both flowers were left with intact anthers. I hand-pollinated bagged flowers each day with a new donor flower of the matching cytotype (following the within-ploidy treatment described earlier) until petals dropped and I judged the flower no longer receptive. Bags were removed at this time to maximize light reception by the photosynthetic achenes. Plants were monitored daily for achene maturity, with all achenes from a flower collected on the first day any were mature.
Achenes were counted and weighed as described. I analyzed the achene count and mass data using ANOVA (PROC GLM, SAS Institute, 2000), with seed counts square-root transformed and mean seed mass log-transformed to improve the fit to model assumptions. Flowers with no seeds were not included in the analysis of mass per seed. I tested for the effects of plant cytotype (diploid or tetraploid), destination neighborhood cytotype (diploid or tetraploid), and the plant x neighborhood interaction. Achene number and mass from bagged, hand-pollinated flowers were used as covariates in the respective analyses to account for variation due to plant vigor and microsite quality. I tested the covariate for interactions with all main effects and dropped all non-significant interactions (P > 0.3) from the statistical model.
Cytotypes produced by experimental and naturally occurring minority plants
I randomly selected eight diploid and nine tetraploid unbagged flowers from transplanted minority cytotype plants to assess the cytotype of the seeds set. I performed flow cytometry on 10 seeds from each flower, as described.
Within the 25-m contact zone between diploid and tetraploid dominated areas at Hoosier Pass (Baack, 2004
), I collected achenes from selected naturally occurring minority diploid (N = 8), and tetraploid (N = 6) plants in August of 2002. Pollen transfer distances in Ranunculus are probably quite short due to the muscid fly pollinators. Accordingly, selected plants had one or fewer neighbors of the same cytotype within 1 m to minimize the chances for within-ploidy pollinations. Achenes produced by natural pollination were counted, weighed, prepared for flow cytometry, then scored as diploid, triploid, or tetraploid, as described earlier.
Fecundity of naturally occurring triploid plants
In 2001, I examined 20 triploid plants at Hoosier Pass and found six with complete seed heads; from these, I collected 27 seeds. Seeds were weighed and analyzed using flow cytometry. The 14 plants not setting seed were excluded from statistical analysis due to the possibility that seeds had dispersed prior to collection. Ten seeds were collected from an additional plant in 2002 for analysis by flow cytometry.
RESULTS
Barriers to triploid formation
Cross-ploidy pollinations resulted in dramatic changes in seed number and seed mass. Of the pollinations attempted, 85% resulted in some seed set. Tetraploid plants pollinated by diploid plants had 39% of the crossing combinations fail to produce seed, while the other crosses had 9–13% failure rates. Maternal plant ploidy did not have a significant effect on seed set, but cross-ploidy pollinations led to a 32% decrease in seed number in diploid maternal plants and 66% reduction for tetraploid maternal plants (Table 1A, Fig. 1A). Mean mass per seed declined 55% for tetraploid maternal parents following cross-ploidy pollinations, but nonsignificantly for diploid maternal parents (Fig. 1B, Table 1B).
|
|
|
|
). To better control for possible artifactual effects of donor number and pollen load in the mixed-ploidy experiment, I added an additional treatment "2-fold within," for 36 tetraploid plants at Pennsylvania Mountain in 2001. In this treatment, pollen from two tetraploid donor flowers was used in each round of pollination. In 2001, mixed-ploidy pollen treatments resulted in a 40% drop in seed number compared with the within-pollination treatment (F1,109 = 3.38, P = 0.07), while the two-fold within pollination treatment led to a nonsignificant 8% increase in seed number (F1,109 = 0.01, P = 0.92). This indicates that the decline in seed number in the mixed-ploidy pollination treatment was due to the presence of pollen from both cytotypes rather than from excess pollen. Seed mass was not significantly affected by the different treatments.
Cytotypes resulting from hand pollinations
Flow cytometry confirmed that the seeds produced in the cross-ploidy crosses were consistently triploid (100%, N = 25). Observed peaks for the embryo of each ploidy varied by ±10% among all seeds of a given ploidy. Approximately 10% of the seeds could not be accurately scored due to too little tissue or uninterpretable results.
Mixed-ploidy crosses on diploid maternal plants resulted in 5.5% (± 3% SD) triploid seeds, with the remainder diploid (N = 10 crosses for each maternal cytotype, 3–8 seeds per cross, mean of 4.6 seeds per cross examined). Tetraploid plants were slightly more likely to form triploids in the mixed-ploidy crosses, with 19% ± 10% (SD) triploid seeds on tetraploid maternal parents (N = 10 crosses, 4–8 seeds per cross, mean of 5.3 seeds per cross examined).
Stigmatic and stylar barriers to cross-ploidy pollinations
Cross type (diploid, tetraploid, or mixed-ploidy) did not affect the number of pollen tubes germinating on tetraploid stigmas (F2,16 = 0.43, P = 0.66). Likewise, cross type did not alter the probability of a germinating pollen tube reaching the bottom of the style (F2,16 = 0.15, P = 0.86). Tetraploid, diploid, and mixed-ploidy pollen loads resulted in similar numbers of emerging pollen tubes (tetraploid = 3.83 ± 1.17; diploid = 4.13 ± 0.98; mixed-ploidy = 2.89 ± 0.58) and a similar proportion of pollen tubes reaching the bottom of the style (tetraploid = 0.41 ± 0.06; diploid 0.42 ± 0.09; mixed-ploidy = 0.34 ± 0.11).
Reproductive success of experimental minority plants
Seventy percent of the transplanted plants survived to produce seeds on at least one flower; 40% produced seeds on both flowers. Bagged control flowers supplied with pollen of the same cytotype significantly accounted for some of the variation in seed number and mass per seed (see Table 3). Interactions between the covariate and source ploidy and destination ploidy were not significant for seeds per flower or seed mass (P > 0.25 for all interactions).
|
|
Seed cytotypes produced by experimental and natural minority plants
Mirroring the results of the mixed-ploidy pollination treatment, open-pollinated plants transplanted into the neighborhood dominated by the opposite cytotype usually set seeds matching the maternal cytotype. Minority diploid plants set 98% diploid seeds (±2%, N = 9 plants, 9.8 seeds per plant examined), while minority tetraploid plants set 85% tetraploid seeds (± 10%, N = 8, 7.3 seeds per cross examined).
Minority diploid plants in the contact zone set complements of entirely diploid seeds (100% ± 0%, N = 8 plants, 9.1 seeds per plant, range of 3–10), while minority tetraploid plants set 91% ± 3% tetraploid seeds (N = 6 plants, 9.8 seeds per plant, range of 3–10).
Triploid fertility
Triploid plants (N = 6) set an average of 3.3 (± 1.4) seeds per flower compared with 33.6 ± 5.9 (N = 35) for diploid plants and 22.5 ± 3.1 (N = 28) for tetraploid plants. Mean mass per seed did not vary significantly among triploid (1.03 ± 0.10 mg, N = 5), diploid (0.92 ± 0.04 mg, N = 34) or tetraploid plants (0.94± 0.05 mg, N = 27). Flow cytometry results indicated that the seeds set were closest to triploid in ploidy, with one potential tetraploid seed among the 34 examined (N = 6 maternal plants).
DISCUSSION
The minority cytotype reproductive disadvantage and barriers to triploid formation documented in this study suggest that tetraploid snow buttercups would face significant barriers to formation and establishment. The rarity of triploid seed production by minority plants suggests that most unreduced gametes that might be produced are unlikely to result in triploid offspring (Fig. 4). The absence of phenological differences between cytotypes, coupled with an apparent lack of stigmatic prezygotic barriers, leads minority tetraploid plants to produce fewer seeds than the surrounding diploid majority (see Fig. 5). Reduced seed production results despite self-pollination. In conjunction with the low observed triploid fertility, this suggests that triploids are unlikely to significantly lower the barriers to tetraploid establishment. Taken together, these results may explain the spatial segregation of cytotypes observed in snow buttercups (Baack, 2004
) and suggest that significant obstacles would have existed to tetraploid establishment and speciation.
|
|
calculated the index of triploid block as one minus the ratio of seeds set in cross-ploidy pollinations to seeds set in within-ploidy pollinations [1 – (cross-ploidy/within-ploidy)]. Diploid snow buttercups had an index value of 0.334, far below the average of 0.952 for diploid maternal parents in angiosperms, and below the lowest previously recorded value of 0.57 (Ramsey and Schemske, 1998
). In contrast, tetraploid buttercups had an index of value of 0.691, still below the mean of 0.801 for tetraploid maternal parents but within the previously reported range of 0.34–1.0 (Ramsey and Schemske, 1998
). Snow buttercups are unusual in having a lower index of triploid block for the diploid plants than for tetraploid plants, previously reported in one study of the 11 reviewed by Ramsey and Schemske (1998)
. The weak triploid block contrasts with the rarity of triploid plants observed in natural populations (Baack, 2004
), the low frequency of triploid seeds produced by minority plants, and the low rate of triploid seed production from mixed-ploidy pollinations.
Reduced fecundity of triploid plants
If triploid plants produce unreduced gametes, the triploid ovules could be pollinated by haploid pollen grains from diploid plants, resulting in tetraploid embryos. Recent models suggest that high levels of unreduced gamete production by triploid plants can play a critical role in facilitating the persistence of neotetraploid plants in mixed diploid–tetraploid populations by contributing to tetraploid demographics (Felber, 1991
; Felber and Bever, 1997
). However, this study found just one potential tetraploid embryo on triploid maternal plants (although only 15 seeds were scored). Based on estimates from experiments, a triploid plant would contribute just 0.1% of the tetraploid seeds that a minority tetraploid plant would contribute. Triploid plants are thus unlikely to contribute significantly to the demographics of tetraploid persistence in snow buttercups (Fig. 5), although they may have played key role in their origin.
Obstacles to tetraploid establishment
Current tetraploid populations are reproductively isolated from diploid populations due to nearly complete spatial segregation (Baack, 2004
). However, newly originated tetraploid plants in sympatry with diploid populations should have faced obstacles to establishment. If we consider the sequence of events leading to successful reproduction, several steps suggest that tetraploid snow buttercups should face a minority disadvantage (see Fig. 5). Reciprocal transplant experiments have demonstrated that diploid and tetraploid snow buttercup seeds and seedlings share a fundamental niche (Baack and Stanton, 2005
). Both diploid and tetraploid flowers open within a day of the snow melting, so there is no temporal separation to reduce cross-ploidy pollinations when in sympatry. Plants and flowers of the two cytotypes share similar floral morphologies and solar tracking abilities, suggesting that the small fly pollinators may not distinguish between the two. When mixed-ploidy pollen loads do reach snow buttercups, such as when minority tetraploid plants receive both pollen from diploid plants and self-pollen, a reduction in seed set results. This reduction is partially, but not completely, offset by an increase in seed mass. Taken together, these factors suggest that minority tetraploids are likely to pay a reproductive cost when in sympatry with their diploid progenitors, placing snow buttercups with rye (Hagberg and Ellerstrom, 1959
) and orchard grass (Maceira et al., 1993
) in suffering from reproductive interference between cytotypes. In contrast, the well-studied tetraploid Chamerion angustifolium did not show any reduction in seed set when in the minority due to pollinator preferences (Husband, 2000
) and differences in pollen competitive ability (Husband et al., 2002
).
The results of this study may explain two puzzles from an earlier distribution study. First, triploid plants were very rare, even in the contact zone between diploid and tetraploid snow buttercups (Baack, 2004
), which may be due to the strong barriers to triploid seed development when mixed-ploidy pollen loads are present. Because snow buttercups are self-fertile, some within-ploidy pollen will be present even when tetraploid plants are in the minority. Second, the distributional study found strong spatial segregation between diploid and tetraploid buttercups (Baack, 2004
), yet reciprocal transplants of seeds and seedlings found no differences in the performance of the cytotypes in their native sites (Baack and Stanton, 2005
). The reproductive disadvantage facing minority plants could explain the observed spatial segregation.
FOOTNOTES
1 The author thanks M. Stanton and her lab group at UC Davis and C. Galen and her lab group at U. Missouri for intellectual support and feedback (and Candi for essential logistical support); P. Lampton, D. Baack, and G. Baack for help with pollination in the field; K. Bradford and S. Gurusinghe for access to and help with flow cytometry. K. Whitney, K. Rice, L. Galloway, and four anonymous reviewers suggested valuable improvements to this manuscript. This work was supported by grants from the University of California, Davis and by National Science Foundation grant DEB 0105116. 
2 E-mail: ebaack{at}indiana.edu
; present address: Department of Biology, Indiana University, 1001 East Third Street, Bloomington, Indiana 47405 USA
LITERATURE CITED
Baack E. J. 2004 Cytotype segregation at regional and microgeographic scales. American Journal of Botany 91: 1783-1788
Baack E. J. 2005 Succeed globally, disperse locally: a stochastic spatial model for tetraploid establishment. Heredity 94: 538-546[CrossRef][ISI][Medline]
Baack E. J. M. L. Stanton 2005 Ecological factors influencing tetraploid speciation in snow buttercups (Ranunculus adoneus): niche differentiation and tetraploid establishment. Evolution 59: 1936-1944
Bretagnolle F. J. D. Thompson 1995 Gametes with the somatic chromosome number: mechanisms of their formation and role in the evolution of autopolyploid plants. New Phytologist 129: 1-22[CrossRef][ISI]
Ehrendorfer F. 1979 Polyploidy and distribution. In W. H. Lewis [ed.], Polyploidy, 45–60. Plenum Press, New York, New York, USA
Felber F. 1991 Establishment of a tetraploid cytotype in a diploid population: effect of relative fitness of the cytotypes. Journal of Evolutionary Biology 4: 195-207
Felber F. J. D. Bever 1997 Effect of triploid fitness on the coexistence of diploids and tetraploids. Biological Journal of the Linnean Society 60: 95-106
Galen C. T. E. Dawson M. L. Stanton 1993 Carpels as leaves— meeting the carbon cost of reproduction in an alpine buttercup. Oecologia 95: 187-193[CrossRef][ISI]
Hagberg A. S. Ellerstrom 1959 The competition between diploid, tetraploid, and aneuploid rye. Hereditas 45: 369-415[ISI]
Husband B. C. 2000 Constraints on polyploid evolution: a test of the minority cytotype exclusion principle. Proceedings of the Royal Society, B, Biological Sciences 267: 217-223[CrossRef]
Husband B. C. D. W. Schemske 1998 Cytotype distribution at a diploid–tetraploid contact zone in Chamerion (Epilobium) angustifolium (Onagraceae). American Journal of Botany 85: 1688-1694
Husband B. C. D. W. Schemske T. L. Burton C. Goodwillie 2002 Pollen competition as a unilateral reproductive barrier between sympatric Chamerion angustifolium. Proceedings of the Royal Society of London, B, Biological Sciences 269: 2565-2571[CrossRef]
Kapoor B. M. A. Love 1970 Chromosomes of Rocky Mountain Ranunculus. Caryologia 23: 575-594
Kearns C. A. D. W. Inouye 1993 Techniques for pollination biologists. University Press of Colorado, Niwot, Colorado, USA
Levin D. A. 1975 Minority cytotype exclusion in local plant populations. Taxon 24: 35-43[CrossRef]
Lewis W. H. 1976 Temporal adaptation correlated with ploidy in Claytonia virginica. Systematic Botany 1: 340-347[CrossRef]
Li B. H. X. M. Xu M. S. Ridout 2004 Modelling the establishment and spread of autotetraploid plants in a spatially heterogeneous environment. Journal of Evolutionary Biology 17: 562-573[CrossRef][ISI][Medline]
Mable B. K. 2004 Polyploidy and self-compatibility: is there an association?. New Phytologist 162: 803-811[CrossRef][ISI]
Maceira N. O. P. Jacquard R. Lumaret 1993 Competition between diploid and derivative autotetraploid Dactylis glomerata L. from Galicia: implications for the establishment of novel polyploid populations. New Phytologist 124: 321-328[CrossRef][ISI]
Masterson J. 1994 Stomatal size in fossil plants: evidence for polyploidy in majority of angiosperms. Science 264: 421-424
Ramsey J. D. W. Schemske 1998 Pathways, mechanisms, and rates of polyploid formation in flowering plants. Annual Review of Ecology and Systematics 29: 467-501
Segraves K. A. J. N. Thompson 1999 Plant polyploidy and pollination: floral traits and insect visits to diploid and tetraploid Heuchera grossulariifolia. Evolution 53: 1114-1127[CrossRef][ISI]
Stanton M. L. C. Galen 1989 Consequences of flower heliotropism for reproduction in an alpine buttercup (Ranunculus adoneus). Oecologia 78: 477-485[CrossRef][ISI]
Stanton M. L. C. Galen J. Shore 1997 Population structure along a steep environmental gradient: consequences of flowering time and habitat variation in the snow buttercup, Ranunculus adoneus. Evolution 51: 79-94[CrossRef][ISI]
Stebbins G. L. 1970 Chromosomal evolution in higher plants. Edwin Arnold, London, UK
van Dijk P. T. Bakx-Schotman 1997 Chloroplast DNA phylogeography and cytotype geography in autopolyploid Plantago media. Molecular Ecology 6: 345-352[CrossRef]
Wiens D. D. K. Halleck 1962 Chromosome numbers in Rocky Mountain plants I. Botaniska Notiser 115: 455-464
Williams J. H. W. E. Freidman M. L. Arnold 1999 Developmental selection within the angiosperm style: using gamete DNA to visualize interspecific pollen competition. Proceedings of the National Academy of Sciences, USA 96: 9201-9206
Young H. J. T. P. Young 1992 Alternative outcomes of natural and experimental high pollen loads. Ecology 73: 639-647[CrossRef][ISI]
This article has been cited by other articles:
|
|
 |
|
  K. Halverson, S. B. Heard, J. D. Nason, and J. O. Stireman III Origins, distribution, and local co-occurrence of polyploid cytotypes in Solidago altissima (Asteraceae) Am. J. Botany, January 1, 2008; 95(1): 50 - 58. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. H. Kao Origins and Widespread Distribution of Co-existing Polyploids in Arnica cordifolia (Asteraceae) Ann. Bot., January 1, 2008; 101(1): 145 - 152. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Suda, H. Weiss-Schneeweiss, A. Tribsch, G. M. Schneeweiss, P. Travnicek, and P. Schonswetter Complex distribution patterns of di-, tetra-, and hexaploid cytotypes in the European high mountain plant Senecio carniolicus (Asteraceae) Am. J. Botany, August 1, 2007; 94(8): 1391 - 1401. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
