| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
2 Ecology Center, Utah State University, Logan, Utah 84322 USA; 3 Department of Biology, Utah State University, Logan, Utah 84322 USA; 4 Rocky Mountain Biological Laboratory, Crested Butte, Colorado 81224 USA; 5 Department of Ecology and Evolutionary Biology, University of California, Irvine, California 92717 USA; 6 Department of Biology, University of California, Riverside, California 92521 USA; 7 Department of Evolution, Ecology, and Organismal Biology, Ohio State University, 1735 Neil Avenue, Columbus, Ohio 43210 USA
Received for publication December 7, 1999. Accepted for publication April 18, 2000.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
Key Words: hummingbird • hybrid zone • pollination • reproductive isolation
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
). Conversely, strong reproductive barriers or poor hybrid fitness, or both, can lead to sharp contact zones between two taxa, with at most a small contingent of hybrid individuals evident (Barton and Hewitt, 1985
; Barton and Bengtsson, 1986
). The extent to which hybrids are formed can thus determine whether closely related taxa continue to diverge following secondary contact, and can affect the pattern and rate of speciation (e.g., Endler, 1977
). Successful hybridization between taxa requires success at all stages of the complex reproductive process. In animal-pollinated angiosperms, the formation of F1 hybrids requires synchronous flowering and a pollinator that visits both taxa and transfers pollen between them. Furthermore, pollen must be able to germinate and send a pollen tube down the style, a functional zygote must form, and a seed must be able to germinate and grow into a fertile plant. Breakdown at any of these stages can limit the formation of a hybrid zone.
Here we explore barriers to hybrid formation in the Ipomopsis aggregata complex (Polemoniaceae), a group of montane perennials found throughout western North America. This group is particularly well suited to the study of reproductive isolation because it includes the full spectrum from species pairs that hybridize readily to those that appear not to form hybrids in nature. Grant and Wilken (1986)
describe three species in the complex, I. tenuituba, I. aggregata, and I. arizonica. The first of these species produces white to pink flowers and has a floral morphology suggestive of hawk moth pollination, whereas the latter two species are both red-flowered and thought to be visited primarily by hummingbirds (Grant, 1992, 1993
; Grant and Wilken, 1986, 1988
). However, these "pollination syndromes" are far from absolute; hawk moths visit I. aggregata, and hummingbirds visit I. tenuituba (Campbell, Waser, and Meléndez-Ackerman, 1997
). Furthermore, the usual color preferences of pollinators can be reversed artificially by manipulating floral characters other than color (Meléndez-Ackerman, Campbell, and Waser, 1997
). Based on floral variation in nature, Grant and Wilken (1988)
concluded that I. aggregata and I. tenuituba hybridize readily and form relatively broad hybrid zones in many (but not all) areas where the two species come into contact. This is supported by results from pollination experiments that show a strong potential for interspecific pollen transfer (Campbell, Waser, and Meléndez-Ackerman, 1997
; Meléndez-Ackerman, Campbell, and Waser, 1997
; Campbell, Waser, and Wolf, 1998
). In contrast, I. arizonica, which is distributed in northern Arizona, southern Utah, and adjacent areas, does not appear to hybridize with the other species (Grant and Wilken, 1986
). On the Kaibab Plateau of northern Arizona, I. arizonica comes into very close contact with I. aggregata (Grant and Wilken, 1986
). In this area, I. arizonica is found in piñon-juniper communities at elevations below 2100 m. Above this elevation I. aggregata is abundant in Pinus ponderosa forests and clearings. At the contact zones the two species come into proximity and some populations contain both species. Broad-tailed hummingbirds (Selasphorus platycerus) have been observed alternately visiting the two species at the contact zones (P. Wolf, personal observation). Furthermore, artificial hybrids between I. aggregata and I. arizonica from various sites have been produced in cultivation (D. Wilken, personal communication). However, plants of intermediate morphology have not been observed in nature (Grant and Wilken, 1986
).
Ipomopsis aggregata subsp. formosissima (the subspecies that comes into contact with I. arizonica) has corollas that are 15–23 mm long with anthers and stigmas exserted 3–7 mm beyond the corolla orifice, whereas I. arizonica has corollas that are 11–15 mm long with anthers and stigmas inserted 3–5 mm below the orifice (Grant and Wilken, 1986
). Grant and Wilken (1986)
propose that floral morphology plays a major role in maintaining reproductive isolation between these species. According to these authors, both species are hummingbird-pollinated, but pollen is transferred in I. aggregata via the face and head, whereas it is transferred via the bill-tip in I. arizonica. Here we report on experiments that test this "mechanical isolation" hypothesis. Specifically we examine whether hummingbirds are as effective at interspecific pollen transfer between I. aggregata and I. arizonica as they are at intraspecific transfer. We also explore postpollination effects by examining seed production and paternity under conditions of artificial conspecific pollination, heterospecific pollination, and mixed con- and heterospecific pollination. We test whether heterospecific pollen alone is capable of fertilization, and then whether it is excluded in the presence of conspecific pollen (e.g., Rieseberg, Desrochers, and Youn, 1995
, and references therein). Results of our study suggest that reproductive isolation in this system is asymmetrical at different stages of reproduction. We detected differences between interspecific and conspecific pollen performance on I. aggregata pistils and differences in pollen transfer when I. arizonica is the pollen recipient.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
), but there is considerable individual variation (Campbell, 1989
). As flower buds formed we emasculated those that were to act as recipient flowers. This ensured that any pollen on recipient stigmas came only from the experimental donor flowers. We used captive broad-tailed hummingbirds (Selaphorus platycercus; three male and two female) to examine pollen transfer in an aviary. This hummingbird species is the primary visitor to I. aggregata and I. arizonica on the Kaibab Plateau. Our primary null hypothesis was that birds are equally effective at transferring heterospecific and conspecific pollen, i.e., there is no donor by recipient interaction. We also compared I. aggregata and I. arizonica for overall effectiveness at importing and exporting pollen via hummingbirds.
We trained individual hummingbirds to visit hand-held flowers in an aviary. Birds that did not respond to training within 3 h were released, and no bird was kept >5 h in captivity. Each successfully trained bird was used in two trials, one for each pollen-donor species. For each donor species, the bird was presented with two male-phase flowers of that species followed by three triplets of female-phase recipient flowers, one I. arizonica, one I. aggregata subsp. aggregata (from RMBL), and one I. aggregata subsp. formosissima, with the order randomized within each triplet. Flowers were hand-held and presented to the birds one at a time. Before a bird was presented with each donor pair, we cleaned its bill and face with a moist cotton swab to remove residual pollen. After each series of visits from a bird, we removed the stigmas from recipient flowers, mounted them in basic fuschin stain (Beattie, 1971
), and counted the number of pollen grains under 100x magnification.
Data analysis
Because the focus here is on the two Kaibab Plateau taxa (I. aggregata subsp. formosissima and I. arizonica), stigma loads on I. aggregata subsp. aggregata were excluded from this analysis. Hereafter, I. aggregata refers to I. aggregata subsp. formosissima unless specified otherwise. Prior to analysis we summed the pollen loads of each recipient type, across triplets, for each bird. Then we assessed the effects of donor and recipient species on pollen load using a split-plot ANOVA with birds as blocks, and trials within birds as whole plots. The donor effect was tested over the bird by donor interaction, and the recipient and donor by recipient interaction were tested over residual error. Because we are using count data, and because of the potential hit or miss nature of pollen transfer, we log transformed the raw data (ln [number of grains + 1]) prior to analysis. Normality of the transformed data was tested with Shapiro-Wilks' W. We used PROC MIXED in SAS release 6.12 (SAS, 1989–1996).
Postpollination experiments
These experiments were designed to test the performance (siring ability) of conspecific and heterospecific pollen on stigmas of I. arizonica and I. aggregata. We transplanted field-collected rosettes towards the end of the flowering season in September 1993. Potted plants were overwintered in a 4°C cold room for 2 mo and then transferred to a pollinator-free greenhouse at Utah State University, where they flowered the following spring. We genotyped over 400 plants for two electrophoretic loci, 6Pgd-2 and Pgi-2, following the protocols of Wolf and Campbell (1995)
. Our overall approach was to use homozygous mothers, pollinated by homozygous fathers with mutually exclusive alleles. Thus, the ideal design was for the mother to be MM at 6pgd-2, whereas the competing fathers would be FF and SS, respectively. Because we could not always ensure the availability of such convenient genotypes it was usually necessary to identify the progeny of one father with 6pgd-2 and progeny of the other father with Pgi-2. This procedure ensured that we could always distinguish paternity, and we could also detect progeny from selfing (see below) and unwanted contaminant pollen. We used 15 plants as females and 15 as male plants, with some used as both, for a total of 23 plants. Recipient plants were not emasculated because we wanted to examine effects of pollen source under relatively natural conditions. We tested whether plants can self-fertilize, cross-fertilize between species, and whether hybrid seed production is reduced by the effects of competing conspecific pollen. We used the following treatments:
Self-pollination
Receptive stigmas were hand-pollinated with pollen from male-phase flowers from the same plant. In previous studies, <1% of selfed ovules set seed in I. aggregata subsp. aggregata (from RMBL), because of a self-incompatibility system (Waser and Price, 1991
). However, self-sterility or fertility has not been characterized in I. arizonica or I. aggregata subsp. formosissima.
Conspecific pollination
Flowers were hand-pollinated with pollen from another individual of the same species carrying marker alleles distinct from those in the mother. The purpose of this treatment was to make comparisons to selfing and to heterospecific pollination under noncompetitive conditions.
Conspecific competition
Flowers were hand-pollinated with pollen from two different fathers of the same species as the mother, such that the paternity of progeny could be determined. Here we test whether strong pollen preferences occur within species, so we can evaluate the results of heterospecific pollen competition. This treatment also tested the efficacy of our pollination methods (for selecting pollen at an appropriate stage, pollen mixing, and pollen application, see below).
Heterospecific pollination
Pollen from a single father of the other species was applied. Here we examined the potential for heterospecific pollen to sire seeds in the absence of conspecific competition.
Interspecific competition
Flowers were hand-pollinated with pollen from two different fathers, one of each species, to examine whether conspecific pollen outcompetes heterospecific pollen in mixture.
Pollen was harvested from anthers that had just begun to dehisce. Preliminary examination of anther content showed considerable variation in numbers of pollen grains, although I. aggregata anthers tended to contain more pollen than I. arizonica anthers. To achieve approximately equal pollen supply we harvested pollen from several anthers of each father and placed it in adjacent piles on a glass microscope slide. Piles of similar size were then thoroughly mixed with a clean dissecting needle and applied to the three-lobed receptive stigma. Color-coded marks, indicating treatment, were made on pedicels of treated flowers. We then monitored fruit development and harvested fruits just before capsule dehiscence. Seeds were stored in paper envelopes for 2–8 wk prior to electrophoretic genotyping, for which we randomly selected a subset of seeds from each treatment class. For each comparison, treatment means were compared using a one-way ANOVA in a randomized block design with heterogeneous variances for each treatment. Plants were considered blocks and because a plant was not used for all treatments, many blocks were incomplete. Analyses were performed using PROC MIXED in SAS release 6.12 (SAS, 1989–1996). In this and other analyses we considered a recipient plant to provide a single independent replicate for a particular treatment. Selfing and heterospecific pollination treatments were compared to conspecific pollination using unpaired t tests, comparing seeds per flower pollinated. For the pollen competition treatments, deviation from a 0.5 proportion of seeds sired by one father was tested with a one-sample t test using a two-tailed test for conspecific competition and a one-tailed test for interspecific competition, where the alternative hypothesis is that conspecific pollen will be most successful.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
= 0.05 (donor x recipient interaction; Table 1). However, I. arizonica was a significantly better pollen donor than I. aggregata (F1,4 = 8.75, P = 0.042). This resulted in large transfers of pollen from I. arizonica to I. aggregata (mean 73.8 grains for three recipients) but very little in the reciprocal direction (mean 4.40 grains; Table 2). There was considerable variation among birds in amount of pollen transfer: the fourth bird transferred over 18 times as much pollen as the fifth bird. However, the difference between the two types of donor was consistent across birds (Table 3).
|
|
|
|
Conspecific competition
Here we pollinated with pollen from two fathers of the same species as the mother. Progeny proportions ranged from 0.5 to 0.65 (Tables 4 and 5), suggesting either variation in paternal success even within species, or failure to achieve a 50:50 pollen mixture. However, deviations from 0.5 were not statistically significant for I. aggregata (t = -2.99, df = 3, P = 0.582) or I. arizonica (t = 0.442, df = 2, P = 0.508). Thus, it appears that our attempts at pollen mixing and application were adequate.
Heterospecific pollination
We pollinated 280 I. aggregata flowers with pollen from I. arizonica and obtained only 14 seeds (mean = 0.05 seeds per pollination), a similar rate to that under selfing (Table 4). This was significantly lower than seed set under conspecific pollination (t = 2.7; df = 4; P = 0.027). All 11 of the progeny genotypes that resolved were consistent with the crosses made. Under the reciprocal cross, I. arizonica produced 1.49 seeds per pollination, more than half the rate under conspecific pollination, and there was no significant statistical difference between the heterospecific and conspecific pollination (t = 1.77; df = 3; P = 0.175).
Interspecific competition
Large numbers of seeds were harvested from both species in this treatment. Of the 237 seeds tested from four I. aggregata plants, all had genotypes consistent with the I. aggregata father and none had a genotype consistent with I. arizonica as the father. However, considerable variation was seen among the four I. arizonica plants examined (Table 5). Two plants produced almost 0.5 proportions of intraspecific progeny. For the third plant, 0.37 of seeds were sired by I. arizonica, and the fourth plant produced progeny from only the I. arizonica father. For I. arizonica as a whole there was no statistical difference from a 0.5 proportion, even using a two-tailed t test (t = -1.262, df = 3, P = 0.148). This test could not be performed for I. aggregata because there was no variance for the I. arizonica father.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
). We caution, however, that these inferences rest on small samples sizes of plants. We cannot rule out the possibility that responses vary within the Kaibab contact zone, as well as among different contact zones. Indeed, even our small sample sizes reveal variation among individual I. arizonica plants in postpollination response to mixtures of conspecific and heterspecific pollen.
Hypothesized mechanism for limited pollen transfer
Hummingbirds need nectar, for which they forage voraciously at Ipomopsis and other plants. Although hummingbirds are aerial acrobats, their visits are rapid and not perfectly repeatable in terms of approach angle and movement while probing a flower (Campbell, Waser, and Price, 1994
; Smith et al., 1996
; Temeles, 1996
). Thus, although floral morphology will impose some constraints on pollen deposition, it is unlikely that pollen will always be confined to a distinct region of the bird's body. However, under some conditions (described below), variation in pollen deposition on a bird could provide the potential for pollen transfer between flowers with different morphologies. In I. arizonica, the anthers are concealed in the corolla tube so that pollen can be deposited only towards the tip of the bill, regardless of how the bird probes (Fig. 1). This bill-tip pollen can then be transferred to a stigma as the bird enters a subsequent flower, regardless of species. In Ipomopsis, the stigma is central to the line of approach, and although an exserted stigma may be more likely than an included stigma to miss bill contact, there is nevertheless opportunity for I. aggregata flowers to receive pollen from the bill-tip if contact is made durng entry (Fig. 1). In contrast to the central position of the stigma, the anthers spread, especially so in I. aggregata subsp. formosissima, where they can be exserted up to 7 mm. It is unlikely for pollen that comes from highly exserted anthers to be deposited on the bill tip and more likely for it to attach to the feathers near the distal part of the bill or on the head and face (Fig. 1). This pollen is unlikely to reach the included stigma of I. arizonica (as described by Grant and Wilken, 1986
). Thus, interspecific pollen movement is almost entirely from I. arizonica to I. aggregata.
|
). We can think of two hypotheses to explain the poor performance of I. arizonica pollen on I. aggregata flowers. One possibility is that pollen-pistil interactions inhibit growth of heterospecific pollen tubes in I. aggregata. Alternatively, I. arizonica pollen may be adapted to sending a pollen tube down a relatively short style. The style of I. aggregata is over twice the length of that of I. arizonica, so that an I. arizonica pollen tube may fail to reach an I. aggregata ovary for energetic reasons (e.g., Carney, Hodges, and Arnold, 1996
). Differences in pollen tube growth rates between the species are probably not a factor here because I. arizonica pollen performance was poor on I. aggregata stigmas even in the absence of competition. This contrasts with hybrid zones of some other species, such as Iris (Carney, Cruzan, and Arnold, 1994
), Helianthus (Rieseberg, Desrochers, and Youn, 1995
), and Brassica (Hauser, Jorgensen, and Ostergard, 1997
), where hybrids form readily when heterospecific pollen is applied alone, but fewer hybrids are produced when there is competition from conspecific pollen.
Comparisons to other studies in Ipomopsis
Within I. aggregata subsp. aggregata, pollen export correlates positively with corolla width (Campbell, 1989
; Campbell et al., 1991
). Part of this is because of a phenotypic correlation between corolla width and pollen production, but the direct effect of width is still evident (Campbell, Waser, and Price, 1996
). Here we detected greater pollen export by I. arizonica, which generally produces less pollen per anther and has a narrower corolla tube than I. aggregata subsp. formosissima. There are several explanations for this incongruence between patterns of pollen export. First, mechanisms that operate over the range of variation within a species need not produce differences between species. This idea is supported by both the current study and a previous study of a Colorado hybrid zone between I. aggregata subsp. aggregata and I. tenuituba (Campbell, Waser, and Wolf, 1998
). In I. aggregata subsp. aggregata, corolla widths varied from 
2.9 to 4.3 mm, a range similar to that of I. arizonica, whereas I. aggregata subsp. formosissima has corollas that are almost 20% wider (Table 6). Thus, in the current study we are now dealing with widths above those at which the correlations of pollen export within I. aggregata were measured. Similarly, the narrow-flowered species I. tenuituba (Table 6) is not a worse pollen exporter than the wider flowered I. aggregata subsp. aggregata (Campbell, Waser, and Wolf, 1998
).
|
At the postpollination stage the results of the current study contrast markedly with the I. aggregata and I. tenuituba hybrid zone at Poverty Gulch in Colorado. Whereas I. aggregata and I. arizonica pollen behave differently when on stigmas of I. aggregata, there appear to be no differences in siring ability between I. aggregata aggregata and I. tenuituba pollen on stigmas of I. aggregata aggregata (Alárcon and Campbell, 2000).
Implications for dynamics of hybrid zones
We have examined two stages of the hybridization process: pollen transfer and fertilization potential of heterospecific pollen. However, we only examined pollen transfer by one hummingbird species known to visit Ipomopsis on the Kaibab Plateau. We do not know whether other occasional visitors, such as hawk moths, might exhibit different patterns. Based on the patterns of pollen transfer we measured here, however, we predict that if hybridization is going to be successful occasionally, it will be with I. arizonica as the maternal plant. We hypothesize that it is more likely for the "rules" to be broken by pollinators than pollen tubes. This can be tested if we can find species-specific markers that are inherited maternally, such as chloroplast DNA (Wolf, Murray, and Sipes, 1997
). We predict that maternal markers for I. arizonica will be detected in natural F1 hybrids, if such plants can be found. Fitness, fertility, and mating patterns of F1 hybrids would determine the direction of marker transfer in subsequent generations. But if pollen tube growth is constraining the direction of gene flow we would expect to see capture of I. arizonica chloroplast DNA by I. aggregata, and this can be used as evidence of historical hybridization and subsequent introgression at contact zones. These conclusions suggest several avenues for future research on this system.
In summary, we found partial support for Grant and Wilken's (1986)
mechanical isolation hypothesis for Ipomopsis. However, we detected a stronger effect from poor performance of I. arizonica pollen on I. aggregata flowers. The sharp contact zone between I. aggregata and I. arizonica, and the apparently strong barriers to hybridization, especially at the postpollination stage, are in stark contrast to the extensive hybridization between I. aggregata and I. tenuituba (Campbell, Waser, and Meléndez-Ackerman, 1997
). This makes the I. aggregata species complex excellent for studying speciation and the dynamics of hybridization in animal-pollinated angiosperms.
|
FOOTNOTES |
|---|
8 Author for correspondence (wolf{at}biology.usu.edu
).
|
LITERATURE CITED |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
Barton, N., and B. O. Bengtsson. 1986 The barrier to genetic exchange between hybridising populations. Heredity 57: 357–376
———, and G. M. Hewitt. 1985 Analysis of hybrid zones. Annual Review of Ecology and Systematics 16: 113–148[CrossRef][ISI]
Beattie, A. J. 1971 A technique for the study if insect-borne pollen. Pan-Pacific Entomologist 47: 82
Campbell, D. R. 1989 Measurements of selection in a hermaphroditic plant: variation in male and female pollination success. Evolution 43: 318–334[CrossRef][ISI]
———, N. M. Waser, and E. J. Meléndez-Ackerman. 1997 Analyzing pollinator-mediated selection in a plant hybrid zone: hummingbird visitation patterns on three spatial scales. American Naturalist 149: 295–315[CrossRef][ISI]
———, ———, and M. V. Price. 1994 Indirect selection of stigma position in Ipomopsis aggregata via a genetically correlated trait. Evolution 48: 55–68
———, ———, and ———. 1996 Mechanisms of hummingbird-mediated selection for flower width in Ipomopsis aggregata. Ecology 77: 1463–1472[CrossRef][ISI]
———, ———, ———, E. A. Lynch, and R. J. Mitchell. 1991 Components of phenotypic selection: pollen export and flower corolla width in Ipomopsis aggregata. Evolution 45: 1458–1467[CrossRef][ISI]
———, ———, and P. G. Wolf. 1998 Pollen transfer by natural hybrids and parental species in an Ipomopsis hybrid zone. Evolution 52: 1602–1611[CrossRef][ISI]
Carney, S. E., M. B. Cruzan, and M. L. Arnold. 1994 Reproductive interactions between hybridizing irises: analyses of pollen-tube growth and fertilization success. American Journal of Botany 81: 1169–1175[CrossRef][ISI]
———, S. A. Hodges, and M. L. Arnold. 1996 Effects of differential pollen-tube growth on hybridization in the Louisiana irises. Evolution 50: 1871–1878[CrossRef][ISI]
Endler, J. A. 1977 Geographic variation, speciation, and clines. Princeton University Press, Princeton, New Jersey, USA
Grant, V. 1963 The origin of adaptations. Columbia University Press, New York, New York, USA
———. 1992 Floral isolation between ornithophilous and sphingophilous species of Ipomopsis and Aquilegia. Proceedings of the National Acadamy of Sciences, USA 89: 11828–11831
———. 1993 Effects of hybridization and selection of floral isolation. Proceedings of the National Acadamy of Sciences, USA 90: 990–993
———, and D. H. Wilken. 1986 Taxonomy of the Ipomopsis aggregata group (Polemoniaceae). Botanical Gazette 147: 359–371[CrossRef][ISI]
———, and ———. 1988 Natural hybridization between Ipomopsis aggregata and I. tenuituba (Polemoniaceae). Botanical Gazette 149: 213–221[CrossRef]
Hauser, T. P., R. B. Jorgensen, and D. Ostergard. 1997 Preferential exclusion of hybrids in mixed pollinations between oilseed rape (Brassica napus) and weedy B. campestris (Brassicaceae). American Journal of Botany 84: 756–762[Abstract]
Meléndez-Ackerman, E., D. R. Campbell, and N. M. Waser. 1997 Hummingbird behavior and mechanisms of selection on flower color in Ipomopsis. Ecology 78: 2532–2541[ISI]
Rieseberg, L. H., A. M. Desrochers, and S. J. Youn. 1995 Interspecific pollen competition as a reproductive barrier between sympatric species of Helianthus (Asteraceae). American Journal of Botany 82: 515–519[CrossRef][ISI]
SAS. 1989–1996. The SAS system for Windows, Release 6.12. SAS Institute, Cary, North Carolina, USA
Smith, C. E., J. T. Stevens, E. J. Temeles, P. W. Ewald, R. J. Hebert, and R. L. Bonkovsky. 1996 Effect of floral orifice width and shape on hummingbird-flower interactions. Oecologia 106: 482–492
Temeles, E. J. 1996 A new dimension to hummingbird–flower relationships. Oecologia 105: 517–523
Waser, N. M., and M. V. Price. 1991 Reproductive costs of self-pollination in Ipomopsis aggregata (Polemoniaceae): are ovules usurped? American Journal of Botany 78: 1036–1043[CrossRef][ISI]
Wolf, P. G., and D. R. Campbell. 1995 Hierarchical analysis of allozymic and morphometric variation in a montane herb, Ipomopsis aggregata (Polemoniaceae). Journal of Heredity 86: 386–394
———, R. A. Murray, and S. D. Sipes. 1997 Species-independent, geographical structuring of chloroplast DNA haplotypes in a montane herb Ipomopsis (Polemoniaceae). Molecular Ecology 6: 283–291[CrossRef][ISI]
This article has been cited by other articles:
|
|
 |
|
  D. A. Tierney and G. M. Wardle The relative fitness of parental and hybrid Kunzea (Myrtaceae): The interaction of reproductive traits and ecological selection Am. J. Botany, February 1, 2008; 95(2): 146 - 155. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. B. Pascarella Mechanisms of prezygotic reproductive isolation between two sympatric species, Gelsemium rankinii and G. sempervirens (Gelsemiaceae), in the southeastern United States Am. J. Botany, March 1, 2007; 94(3): 468 - 476. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. Aldridge and D. R. Campbell Asymmetrical pollen success in Ipomopsis (Polemoniaceae) contact sites Am. J. Botany, June 1, 2006; 93(6): 903 - 909. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. A. Chapman, D. G. Forbes, and R. J. Abbott Pollen competition among two species of Senecio (Asteraceae) that form a hybrid zone on Mt. Etna, Sicily Am. J. Botany, April 1, 2005; 92(4): 730 - 735. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
