Sympatric sister species of Californian antirrhinum and their transiently specialized pollinators.
The plant-pollinator relationship is a topic of intense interest because of its potential importance in facilitating or preventing plant speciation. Through traits such as flower color or patterning, animal pollinators can be influenced into moving assortatively among the plants, thereby avoiding random mating and allowing for lineage splitting (Grant, 1949). The idea that plants might have intimate relationships with particular classes of pollinators led to the paradigm of floral syndromes (van der Pijl, 1961), which carried the implication that the relationship between a plant and its pollinator is specialized (Waser et al., 1996; Johnson and Steiner, 2000). Waser et al. (1996), however, argued that plant-pollinator relationships are more generalized than commonly thought. The debate over generalist versus specialist pollination systems has, however, faded in importance compared to the agreement that more data is needed on plant-pollinator interactions and especially their temporal and spatial fluidity (Waser, 1998; Johnson and Steiner, 2000; Jones, 2001; Fenster et al., 2004).
Many empirical studies have shown that animal pollination can provide a mechanism for reproductive isolation, via a variety of ethological and mechanical means. Pollinators have been shown, for example, to distinguish among taxa with divergent floral morphologies through non random foraging behavior, thus reducing pollen flow between plant taxa (e.g., Stanton, 1987; Fulton and Hodges, 1999; Schemske and Bradshaw, 1999). Extensive work in Ficus (Ronsted et al., 2005) and Yucca (Pellmyr, 2003), has also demonstrated that pollinator behavior can reinforce existing reproductive barriers. Work in Mimulus illustrates that assortative visitation is relatively easy to accomplish and results in significant premating reproductive isolation (Bradshaw and Schemske, 2003). However, the extent to which pollinator mediated assortative mating contributes to the initial divergence of flowering plant taxa remains uncertain.
An investigation into the contribution of plant-pollinator interactions to the initial divergence of plant lineages requires a pair of plant taxa that meet certain criteria. They should be at least sister taxa, lack complete reproductive isolation, and occur in sympatry or parapatry (i.e., within the foraging distance of their pollinators). Studying close relatives that are not sister taxa may reveal much about the maintenance of barriers to gene flow following lineage divergence but does not provide a proper phylogenetic context for studying processes involved in lineage splitting. If the two taxa being compared already have complete reproductive (i.e., genetic) isolation between them, then the significance of any ethological isolation observed is equivocal since they are already genetically isolated lineages. Similarly, since mating must be possible for pollinator behavior to have an effect, allopatric taxa are unsuitable. Investigations of plant-pollinator interactions within such a system are rare (e.g., Ramsey et al., 2003) but important because they help clarify how much of the potential for pollinator mediated isolation is realized in the initial divergence of plant lineages.
In this study, we investigated the pollination system of two taxonomic species of Californian snapdragon, Antirrhinum subcordatum and A. vexillo-calyculatum, which meet all the requirements laid out in the previous paragraph. Molecular phylogenetic analyses of the Californian Antirrhinum reveal that A. subcordatum and A. vexillo-calyculatum form a clade (Oyama and Baum, 2004), confirming their previously hypothesized close relationship (Thompson, 1988). These two taxa are also interfertile and hybrids have been observed in nature (R.K. Oyama, pers. obs.), suggesting either that the taxa have diverged recently without time for complete lineage sorting or that there is ongoing gene flow. The geographic ranges of these two taxa overlap (Fig. 1), but they tend to occur on different soil types, with A. subcordatum found on shale and A. vexillo-calyculatum on serpentine. Morphologically, A. subcordatum and A. vexillo-calyculatum are differentiated almost exclusively by their flower color, with white and purple petals, respectively (Thompson, 1988). Except for color, there are no observed differences in the flowers of the two species (e.g., no difference in patterning or UV reflectance) (Thompson, 1988), but some slight differences in leaf color and shape are visible.
To determine the nature of the relationship between the two plant species and their animal pollinators we gathered two types of data. First, we observed floral visitors in natural populations. These qualitative data tell us whether or not the two plant species have separate suites of pollinator species and thus provide an initial assessment of the degree to which the plant taxa have or have not specialized onto different pollinator species. Second, we investigated the preferences and flower constancy of individual pollinators by observing their behavior in experimental arrays containing both the white and purple flowered plant taxa. These quantitative data reveal whether individual pollinating insects visit the flowers randomly or assortatively. Taken together, these data contribute toward a more nuanced picture of how the plant-pollinator interaction could effect assortative mating and facilitate the early stages of lineage divergence in flowering plants.
MATERIALS AND METHODS
Study species.--The genus Antirrhinum L. (sensu Thompson, 1988) consists of 25 European species in section Antirrhinum and 20 North American species in section Saerorhinum (Oyama and Baum, 2004; Vargas et al., 2004). The latter section, found in west and southwestern North America with a center of diversity in California, displays a wide range of flower morphology, coloration and patterning. Two of these species, A. subcordatum and A. vexillo-calyculatum from northern California, were the focus of this study.
[FIGURE 1 OMITTED]
Antirrhinum subcordatum is restricted to the eastern foothills of the Coastal Range in the Sacramento valley, whereas A. vexillo-calyculatum is widely distributed in the Coastal and Siskiyou Ranges and the north and central Sierra Nevada (Fig. 1). Within A. vexillo-calyculatum, three subspecies are recognized (Thompson, 1988): A. vexillo-calyculatum subsp. vexillo-calyculatum occurs from just south of San Francisco Bay northwards to the southern part of the Coastal Range; A. vexillo-calyculatum subsp, intermedium occurs in the central and northern Sierra Nevada; and A. vexillo-calyculatum subsp, brewer/connects the previous two subspecies ranges via the northern part of the Coastal Range, the southern Siskiyou Mountains and the northern Sierra Nevada. The two species (A. subcordatum and A. vexillo-calyculatum) are sympatric in the foothills of the Coastal Range, where their distribution ranges overlap and where they can be found literally touching one another (R.K. Oyama, pers. obs.).
Whereas many of the Old World species interbreed easily (Mather, 1947), the New World species of Antirrhinum generally have strong intrinsic barriers to hybridization (Thompson, 1988), thus making the interfertility of A. subcordatum and A. vexillo-calyculatum more noteworthy. Flowers of the A. subcordatum and A. vexillo-calyculatum are substantially smaller compared to the familiar garden snapdragon A. majus. However, they still have a closed corolla tube that tends to restrict access to visitors that are large enough to depress the lower lip allowing access to the corolla tube. No noticeable patterns on the flowers have been detected under ultraviolet light, giving confidence that the color difference we observe is likely to be the only visual cue that insect pollinators could use to discriminate these flowers (Thompson, 1988).
Pollinator observations in natural populations.--Visitors to flowers of plants in natural populations were observed over the course of two summers (1999 and 2000). For the observations, we selected four populations of Antirrhinum subcordatum, ten populations of A. vexillo-calyculatum and one where the two plant taxa co-occurred (Fig. 1, inset). The observations were distributed through the hours of the day and over the flowering season. The total hours of observation are given in Table 1. Voucher specimens representing the species of insects observed visiting the flowers were collected at each site and identified by Dr. R. Thorp of the University of California, Davis and deposited at the Harvard Museum of Comparative Zoology. Otherwise, insects were visually identified to the level of genus as they visited flowers in the field.
Array experiments.--To assess the behavior of pollinators when confronted with the two flower colors in sympatry, three array experiments were conducted at sites where observations of pollinator visitors had previously been made (see Fig. 1, inset). The ratio of white- to purple-flowered plants in each array was 1:1 using mostly plants grown in the greenhouse from seed collected in the wild, although we sometimes incorporated wild, rooted plants into arrays. Individual plants within each array were first mapped ca. 10 cm apart, in order to create an even spacing within the constraints of the topography of the site. A coin toss then determined which plant taxon, white- or purple-flowered, was placed into each pre-established position. All array experiments were performed in the summer of 2000.
The locations for the three arrays were chosen to have one where the background wild population was made up of both white- and purple-flowered plants (Array 1), one where the background population was exclusively purple-flowered plants (Array 2), and one where the background population was exclusively white-flowered plants (Array 3). Array 1 was adjacent to a large sympatric population of Antirrhinum subcordatum and A. vexillo-calyculatum that were still in flower. The array consisted of 18 plants, 9 of each flower color, and was situated on a 40 degree slope of mostly shale-derived soil with some serpentine outcrops. Visitation was observed for a total of 39 h. At Array 2, the background population of purple-flowered A. vexillo-calyculatum was abundant and also in flower. The array was situated on level ground at the top of a road cut on serpentine substrate. Visitors were observed for a total of 33 h. Array 3 was set up adjacent to a locality where A. subcordatum was prevalent but just past flowering and where no other Antirrhinum species were located. It was situated on a 40 degree slope on shale-derived soil. Visitors were observed for a total of 36 h.
The series of plants visited by an individual pollinator within an array (hereafter termed "foraging bout") was recorded. Using only those bouts in which more than one plant was visited, we first tested for homogeneity of preference for flower color following the procedure outlined by Jones (1997). The assumption behind this test is that if pollinators foraged with consistent preference, then the foraging bouts belong to a single binomial distribution; that is, visitors behaved as one homogeneous group. If we further assume that the foraging bouts are independent and the transitions within bouts are random, then the test statisic (S) should have a chi-square distribution. If the test is significant, then the data should not be pooled for further analysis but should be split up into homogeneous subgroups. Sub-groups for which homogeneity of preference is not violated can then be analyzed to see if the sequence of visits exhibit constancy (i.e., whether the transitions in the sequences of plants visited were assortative).
We analyzed the combined data for each array and the homogeneous sub-groups to see if the order of visitation from one plant to the next with respect to flower color deviated from random. Expected frequencies for transition between plants of each flower type were calculated based on the observed proportion of all visits to the flower colors for the respective group or sub-group. The null hypothesis was that the flower color of a plant visited was independent of the flower color of the plant previously visited. The four transition categories were White-to-White (WW), White-to-Purple (WP), Purple-to-White (PW), and Purple-to-Purple (PP). The expected frequencies were compared to the observed frequencies in a chi-square test. W also calculated the Constancy Index (CI), using a version from Gegear (2005) modified from Jacobs (1974), and the Bateman Index (Bateman, 1951; Waser, 1986; Gegear and Laverty, 2005).
Pollinator observations in natural populations.--There was considerable overlap in the suites of insects observed visiting flowers in wild populations of Antirrhinum subcordatum and A. vexillo-calyculatum (Table 1). Although five of the visitors to A. vexillo-calyculatum flowers were not observed in any of the five A. subcordatum populations (Ashmeadiella, Hoplitis, Xylocopa tabaniformis orpifex, Bombyliidae, and Hesperiidae), only Ashmeadiella and Hoplitis appeared to be potential pollinators. Xylocopa robbed nectar from outside the flower, and the species of Bombyliidae and Hesperiidae observed were unlikely to have successfully obtained or transfered pollen given the petal-lip mechanism of the snapdragon flower. Furthermore, Hoplitis was only observed or collected at populations of A. vexillo-calyculatum subsp. intermedium, probably due to the disjunct distribution of this subspecies of A. vexillo-calyculatum in the Sierra Nevada foothills. A total of 56 voucher specimens representing 17 insect taxa were collected across the 15 observation sites.
Two species of Bombus are represented in the voucher specimens, B. californicus and B. vosnesenskii. At the location where the two plant taxa occurred sympatrically, which was also the locality for Array 1, a voucher identified as B. vosnesenskii was collected. At the locality corresponding to Array 2 (background population was Antirrhinum vexillo-calyculatum), a voucher identified as B. californicus was collected, and at Array 3 (background population of A. subcordatum), a voucher identified as B. vosnesenskii was collected. Together with the other Bombus vouchers, a potential correlation exists of B. californicus at A. vexillo-calyculatum localities and B. vosnesenskii at A. subcordatum localities, with the exception of the locality where the plant taxa are sympatric. However, since it was not possible to identify Bombus to the species level in the field, the relationship between the Bombus species and plant taxa requires more investigation.
Array experiments.--At both Array 1 and Array 2, foraging bouts for all visitors were significantly heterogeneous (i.e., did not fit a single binomial distribution), indicating that floral visitors are composed of minimally two sub-groups (Table 2). For Array 1, the obvious sub-groupings were Bombus and non-Bombus. Bumblebees (presumably Bombus vosnesenskii at Array 1) visited both white- and purple-flowered plants, whereas the four other types of bees visited purple-flowered plants almost exclusively. Indeed, 33 of the 34 mixed bouts were performed by Bombus. At Array 2, the sub-groups were not as obvious since all visitors were Bombus (presumably B. californicus), but sub-groups homogeneous for preference could be formed if we again separated those visitors that visited exclusively purple flowers from those which visited white or white and purple flowers (Table 2). Since all visitors to Array 3 visited purple flowers exclusively, we could not perform the heterogeneity test. Nevertheless, this first analytical step revealed heterogeneity of color preference among pollinators at Arrays 1 and 2.
The results for the analyses of flower constancy at Arrays 1 and 2 are shown in Table 3. Array 3 was not analyzed for flower constancy since all visits were to purple flowered plants. The results of the chi-square tests on the combined data for each array revealed a significant departure from the expected frequencies for both arrays. At Array 1, there were more transitions between flower colors than expected, whereas at Array 2 there were fewer transitions between flower colors than expected. For the sub-groups (i.e., the homogeneous groups) at each array, the observed transitions were not significantly different from expected. At Array 1, the values for the Constancy Index (CI) and Bateman Index (BI) for the combined data are 0.143 and 0.174 respectively. For the sub-group Bombus, they are -0.003 and 0.002. For the other sub-group consisting of non-Bombus visitors, the CI is 0.085 and the BI is not calculatable. At Array 2, the CI and BI for the combined data are 0.339 and 0.490 respectively. For the sub-group of visitors at Array 2 that went to white and/or purple flowers, the CI is 0.124 and the BI is 0.128.
This study demonstrates that even when the plant-pollinator relationship is generalist at the species level, the interaction can still be specialized at the level of the individual foraging bout. Our observations revealed that Antirrhinum subcordatum and A. vexillo-calyculatum share the same suite of floral visitors (Table 1), although the relative importance of each visitor to overall pollination success must still be investigated (Reynolds and Fenster, 2008). The overlapping suites of pollinators for the two plant taxa indicate that the floral visitors do not have fixed preferences and suggests that, at the level of insect species, the plant taxa have generalist pollination systems. This is not surprising given the high similarity, except for color, of the two flower types. Among all pollinator species, however, our array experiments showed significant heterogeneity of preference with respect to flower color (Table 2). Thus, in areas where the two plant species are sympatric, the individual floral visitors behave in a specialized manner. We refer to the pattern of faithfulness to one flower color at the level of the individual foraging bout, despite an overall generalist relationship at the species level, as transient specialization.
Transient specialization exhibited by individual pollinators, in contrast to a fixed preference for one flower color by pollinator insects, is congruent with observations in other systems. For example, short-term specialization has been reported for honeybees pollinating Raphanus sativus (Stanton et al., 1989). Even when rewards are the same, constraints upon information processing in bees cause an individual to have a tendency to access the most recently acquired information in its memory, such as the characteristics of the flower just visited (Chittka et al., 1999; Gegear, 2001; Menzel, 2001). In our data, this can be seen in the slight bias toward same-color transitions in Array 2 (Table 3). Previous work with Antirrhinum majus has demonstrated that such heterogeneity in foraging causes assortative pollen transfer among floral types (Jones, 1997;Jones et al., 1998). Furthermore, assortative pollen transfer can lead to assortative mating (Jones and Reithel, 2001). Thus, even though A. subcordatum and A. vexillo-calyculatum occur in sympatry and share pollinator species, the transient specialization of the bees may reduce gene flow between these two forms much as Grant (1949) envisioned.
The lability of pollinator specialization can also be seen when comparing our observational and experimental data. For example, Osmia freely visited Antirrhinum subcordatum in natural populations but only A. vexillo-calyculatum in the arrays. Similarly, even though Bombus did visit plants of both flower colors, it exhibited a higher than expected number of transitions to purple flowers at Array 1 (Table 3). A one-way bias in ethological isolation, where barriers to pollen flow between the two species are not symmetrical, has also been observed in the relationship between Bombus and Rhinanthus (Kwak, 1978). In short, the behavior of these bees was different when presented with a choice of flower colors. Thus, floral traits such as color may have the most impact on the pattern of mating when plants are growing in close proximity to others visited by the same kinds of pollinators. Results from Trifolium suggests that this is especially important when the flowers are otherwise highly similar (Wilson and Stine, 1996), such as the case here.
Flower color within the California Antirrhinum clade in general is evolutionarily labile. Indeed, the flower color difference like that observed here between Antirrhinum subcordatum and A. vexillo-calyculatum is genetically simple to achieve (Schwinn et al., 2006). Of course, other floral features besides color could also be influencing pollinator behavior. We tried to assess nectar production but were not successful due to the extreme heat at the localities, small size of the flowers and low levels of nectar. It is known that A. majus flowers produce scent (Dudareva et al., 2005), but whether this is also true for the Californian species and whether there is a difference between the two plant taxa studied here remains an open question. Given that pollen and nectar rewards are probably similar between the two plant taxa, the transient specialization we observe here is likely a manifestation of flower constancy, as defined by Waser (1986). More data is needed to exclude labile preference (sensu Waser, 1986) as an explanation, but whatever the cause, the transient specialization of the pollinators observed is relevant for understanding the divergence of these sympatric taxa.
In a comprehensive study on the contribution of ethological isolation to the speciation of flowering plants, Ramsey et al. (2003) investigated many possible factors contributing to the reproductive isolation of two sister species of Mimulus. While they too found pollinator-mediated isolation to be an important component of isolation, even more important were ecogeographic factors. In contrast to the Antirrhinum species studied here, however, the Mimulus species used in their study seem to have already diverged and had specialized on different suites of pollinators. Still, ecogeographic factors also seem to be relevant in our system since A. subcordatum is found on shale-derived soils whereas A. vexillo-calyculatum is found growing on serpentine-derived soils, which are known to exert a strong selective force (Gardner and Macnair, 2000).
An association of ecogeographic factors and mechanisms of reproductive isolation finds resonance in some models of speciation. These demonstrate that lineage divergence can occur in the face of much more gene flow than previously supposed if the genes underlying reproductive isolation are in linkage disequilibrium with those for ecological specialization (Kondrashov, 1992; Kondrashov and Kondrashov, 1999; Doebeli and Dieckmann, 2003). Aspects of these models have gained empirical support in natural populations of Drosophila (Greenberg et al., 2003). Consequently, even low levels of assortative pollinator movements may suffice to promote speciation when assortative mating is selectively favored by ecological divergent selection (Jones, 2001). The discovery of a wild hybrid between Antirrhinum subcordatum and A. vexillo-calyculatum (it was found at a locality where the two species are sympatric and its flower color and leaf morphology was intermediate) means that ethological isolation between these two Antirrhinum species does not completely prevent gene-flow, a situation also observed in sympatric species of Rhinanthus (Kwak, 1979). In this context, it will be important to investigate the apparently divergent soil preferences of the two taxa and the genetic basis for the flower color difference between A. subcordatum and A. vexillo-calyculatum.
Conclusions.--The difference in flower color between Antirrhinum subcordatum and A. vexillo-calyculatum allows for assortative pollinator visitation to different flower color morphs. This transient specialization is significant given that the pollination system overall could be described as generalized. Considering that these two plant taxa form a monophyletic group, have overlapping ranges and incomplete post pollination barriers to hybridization, the pollination system may be crucial in allowing for their continued evolutionary divergence. This study represents a first look into the pollination ecology of this system. Further work on other floral characters as well as the soil ecology and population genetics is needed to understand how these incipient species were established and to predict their likely long-term fate.
Acknowledgments.--The authors thank: The Dawley family- Big Bluff Ranch; California Department of Fish and Game; Barbara Castro, Kristina Shierenbeck and Lawrence Janeway--California State University, Chico; Barbara Whitlock, Dianella Howarth, Lena Hileman and Janet Sherwood--Dept. Organismic & Evolutionary Biology, Harvard University; David Isle--Mendocino National Forest, USFS; Joe Callizo--Napa Valley Land Trust; Robin Thorp--University of California, Davis. This work was performed in part at the University of California Natural Reserve System--McLaughlin Natural Reserve, and funded in part by a Deland Award from the Arnold Arboretum, Harvard University to RKO.
BATEMAN, A.J. 1951. The taxonomic discrimination of bees. Heredity, 5:271-278.
BRADSHAW, H. D. AND D. W. SCHEMSKE. 2003. Allele substitution at a flower colour locus produces a pollinator shift in monkeyflowers. Nature, 426:176-178.
CHITTKA, L., J. D. THOMSON AND N. M. WASER. 1999. Flower constancy, insect psychology, and plant evolution. Naturwissenschaften, 86:361-377.
DOEBELI, M. AND U. DIECKMANN. 2003. Speciation along environmental gradients. Nature, 421:259-264.
DUDAREVA, N., S. ANDERSSON, I. ORLOVA, N. GATTO, M. REICHELT, D. RHODES, W. BOLAND AND J. GERSHENZON. 2005. The nonmevalonate pathway supports both monoterpene and sesquiterpene formation in snapdragon flowers. Proc. Natl. Acad. Sci. U. S. A., 102:933-938.
FENSTER, C. B., W. S. ARMBRUSTER, P. WILSON, M. R. DUDASH AND J. D. THOMSON. 2004. Pollination syndromes and floral specialization. Ann. Rev. Ecol. Evol. System., 35:375-403.
FULTON, M. AND S. A. HODGES. 1999. Floral isolation between Aquilegia formosa and Aquilegia pubescens. Proc. R. Soc. Lond. Ser. B-Biol. Sci., 266:2247-2252.
GARDNER, M. AND M. MacNAIR. 2000. Factors affecting the co-existence of the serpentine endemic Mimulus nudatus Curran and its presumed progenitor, Mimulus guttatus Fischer ex DC. Biol. J. Linnean Soc., 69:443-459.
GEGEAR, R.J. 2001. The effect of variation among floral traits on the flower constancy of pollinators, p. 1-20. In: L. Chittka and J. D. Thomson (eds.). Cognitive ecology of pollination: animal behavior and floral evolution. Cambridge University Press, Cambridge, England.
-- AND T. M. LAVERTY. 2005. Flower constancy in bumblebees: a test of the trait variability hypothesis. Anim. Behav., 69:939-949.
GRANT, V. 1949. Pollination Systems as Isolating Mechanisms in Angiosperms. Evolution, 3:82-97.
GREENBERG, A. J., J. R. MORAN, J. A. COYNE AND C. I. WU. 2003. Ecological adaptation during incipient speciation revealed by precise gene replacement. Science, 302:1754-1757.
JACOBS, J. 1974. Quantitative measurement of food selection--modification of forage ratio and Ivlev's electivity index. Oecologia, 14:413-417.
JOHNSON, S. D. AND K. E. STEINER. 2000. Generalization versus specialization in plant pollination systems. Trends Ecol. Evol., 15:140-143.
JONES, K. N. 1997. Analysis of pollinator foraging: Tests for non-random behaviour. Funct. Ecol., 11:255-259.
--. 2001. Pollinator-mediated assortative mating: causes and consequences, p. 259-273. In: L. Chittka and J. D. Thomson (eds.). Cognitive ecology of pollination: animal behavior and floral evolution. Cambridge University Press, Cambridge, England.
-- AND J. S. REITHEL. 2001. Pollinator-mediated selection on a flower color polymorphism in experimental populations of Antirrhinum (Scrophulariaceae). Am. J. Bot., 88:447-454.
--,-- AND R. E. IRWIN. 1998. A trade-off between the frequency and duration of bumblebee visits to flowers. Oecologia, 117:161-168.
KONDRASHOV, A. S. 1992. Species and Speciation. Nature, 356:752-752.
-- AND F. A. KONDRASHOV. 1999. Interactions among quantitative traits in the course of sympatric speciation. Nature, 400:351-354.
KWAK, M. M. 1978. Pollination, hybridization and ethological isolation of Rhinanthus minor and R. serotinus (Rhinanthoidead-Scrophulariaceae) by Bumblebees (Bombus latr). Taxon, 27:145-158.
--. 1979. Maintenance of species integrity in sympatrically occurring Rhinanthus minor and Rhinanthus serotinus in the Netherlands. Oecologia, 41:1-9.
MATHER, K. 1947. Species crosses in Antirrhinum. Heredity, 1:175-186.
MENZEL, R. 2001. Searching for the memory trace in a mini-brain, the honeybee. Learning & Memory, 8:53-62.
OYAMA, R. K. AND D. A. BAUM. 2004. Phylogenetic relationships of North American Antirrhinum (Veronicaceae). Am. J. Bot., 91:918-925.
PELLMYR, O. 2003. Yuccas, yucca moths, and coevolution: A review. Ann. Missouri Botan. Gar., 90:35-55.
RAMSEY, J., H. D. BRADSHAW AND D. W. SCHEMSKE. 2003. Components of reproductive isolation between the monkeyflowers Mimulus lewisii and M. cardinalis (Phrymaceae). Evolution, 57:1520-1534.
REYNOLDS, R.J. AND C. B. FENSTER. 2008. Point and interval estimation of pollinator importance: a study using pollination data of Silene caroliniana. Oecologia, 156:325-332.
RONSTED, N., G. D. WEIBLEN, J. M. COOK, N. SALAMIN, C. A. MacHADO AND V. SAVOLAINEN. 2005. 60 million years of co-divergence in the fig-wasp symbiosis. Proc. Roy. Soc. B-Biol. Sci., 272:2593-2599.
SCHEMSKE, D. W. AND H. D. BRADSHAW. 1999. Pollinator preference and the evolution of floral traits in monkeyflowers (Mimulus). Proc. Natl. Acad. Sci. U. S. A., 96:11910-11915.
SCHWINN, K., J. VENAIL, Y.J. SHANG, S. MacKAY, V. ALM, E. BUTELLI, R. OYAMA, P. BAILEY, K. DAVIES AND C. MARTIN. 2006. A small family of MYB-regulatory genes controls floral pigmentation intensity and patterning in the genus Antirrhinum, Plant Cell, 18:831-851.
STANTON, M. L. 1987. Reproductive biology of petal color variants in wild populations of Raphanus sativus: I. Pollinator response to color morphs. Am. J. Bot., 74:178-187.
--, A. A. SNOW, S. N. HANDEL AND J. BERECZKY. 1989. The impact of a flower-color polymorphism on mating patterns in experimental populations of wild radish (Rahapnus raphanistrum L.). Evolution, 43:335-346.
THOMPSON, D. 1988. Systematics of Antirrhinum (Scrophulariaceae) in the NewWorld. System. Bot. Monog., 22:1-142.
van der PIJL, L. 1961. Ecological aspects of flower evolution. II. Zoophilous flower classes. Evolution, 15:44-59.
VARGAS, P.,J. A. ROSSELLO, R. OYAMA AND J. GUEMES. 2004. Molecular evidence for naturalness of genera in the tribe Antirrhineae (Scrophulariaceae) and three independent evolutionary lineages from the New World and the Old. Plant System. Evol., 249:151-172.
WASER, N. M. 1986. Flower constancy: definition, cause, and measurement. Am. Nat., 127:593-603.
--. 1998. Pollination, angiosperm speciation, and the nature of species boundaries. Oikos, 82:198-201.
--, L. CHITTKA, M. V. PRICE, N. M. WILLIAMS AND J. OLLERTON. 1996. Generalization in pollination systems, and why it matters. Ecology, 77:1043-1060.
WILSON, P. AND M. STINE. 1996. Floral constancy in bumble bees: Handling efficiency or perceptual conditioning? Oecologia, 106:493-499.
SUBMITTED 15 OCTOBER 2009
ACCEPTED 15 FEBRUARY 2010
RYAN K. OYAMA, (1) KRISTINA N. JONES (2) AND DAVID A. BAUM (3)
Harvard University Herbaria, 22 Divinity Avenue, Cambridge, Massachusetts 02138
(1) Corresponding author present address: Institute for Systematic Botany, Ludwig-Maximilians University, Menzinger Strasse 67, 80638 Munich Germany; Telephone: +49 (89) 17861 194; email: email@example.com
(2) Present address: Wellesley College, 106 Central Street, Wellesley, Massachusetts 02481
(3) Present address: Department of Botany, University of Wisconsin, 430 Lincoln Drive, Madison 53706
TABLE 1.--Checklist of visitors to flowers of Antirrhinum subcordatum (white flowers) and A. vexillo-calyculatum (purple flowers). Flower visitors in the fi.eld-observed and identified on the fly--could only be assigned to the level of genus. Notes: (1) Voucher specimens of Bombus vosnesenskii were collected at localities of A. subcordatum and A. vexillo-calyrulatum but B. californicus was only collected at localities of A. vexillo-calyculatum. (2) Xylocapa was only observed nectar robbing Flower visitor A. subcordatum A. vexillo- calyculatum Anthidium spp. X x Anthophora sp. x x Ashmeadiella .sp. -- x Bombus spp. (1) x x Ceratina spp. x x Dianthidium sp. x x Hoplitis sp. x x Osmia spp. x x Xylocopa tabaniformis orpifex (2) -- x Bombyliidae -- x Hesperiidae -- x other x x Total number of hours observed 66 89 Total number of visitors observed 202 206 TABLE 2.--Results of pollinator foraging on experimental randomized arrays. Columns, from left to right: Array identification number; visitor groups and combined data for all visitors; number of foraging bouts in which only white-flowered plants were visited, only purple-flowered plants were visited, or both ("mixed"); the average number of plants visited per foraging bout on the array ([+ or -] SD); the ratio of all plant visits that were to purple-flowered plants; the total number of observed plant visits; the test statistic for the test of homogeneity of the binomial (S); and the significance level for each S against a chi-square distribution (P). Only bouts in which two or more plants were visited were included in the analyses # of bouts that were: white purple # plants per Array Group only mixed only bout 1 Combined 11 25 31 5.0 [+ or -] 4.7 Bombus 11 24 1 6.9 [+ or -] 5.8 Others (1) 0 1 30 2.8 [+ or -] 1.1 2 Combined (2) 1 9 17 6.9 [+ or -] 4.1 Both 1 9 0 8.2 [+ or -] 5.3 Purple only 0 0 17 6.1 [+ or -] 3.1 3 Combined (3) 0 0 9 2.5 [+ or -] 0.6 ratio plant Array Group purple visits S P 1 Combined 0.61 339 124.3 <0.01 Bombus 0.47 249 46.2 0.097 Others (1) 0.99 87 43.0 0.059 2 Combined (2) 0.79 185 89.6 <0.01 Both 0.52 82 18.4 0.860 Purple only 1 103 n.a. n.a. 3 Combined (3) 1.0 23 n.a. n.a. (1) = Osmia sp. (n = 24), Anthidium sp. (n = 2), Anthophora urbana (n = 2), Dianthidium sp. (n = 2) (2) = All visitors observed at Array 2 were Bombus (3) = Anthophora urbana (n = 8) and Dianthidium sp. (n = 1) TABLE 3.--Tests of floral constancy for the combined data at each array and of the homogeneous groups identified in Table 2. Columns from left to right: the array, the groups tested (the combined data for the array is in bold), the observed and (expected) numbers for each transition category, the P-value for a chi-square test, the constancy index (CI) and the Bateman index (BI). P = purple, W = white. Values for Array 3 were not calculated since in all foraging bouts only purple flowered plants were visited Array Group P [right arrow] P P [right arrow] W 1 Combined 100 (97.7) 47 (623) Bombus 49 (47.2) 47 (52.4) Others 51 (50.8) 0 (0.6) 2 Combined 107 (97.8) 14 (26.1) Both 21 (19.5) 14 (17.7) Purple only 86 (86) 0 (0) Array Group W [right arrow] P W [right arrow] W 1 Combined 59 (62.3) 56 (39.7) Bombus 58 (52.4) 56 (58.1) Others 1 (0.6) 0 (0.01) 2 Combined 17 (26.1) 19 (7) Both 17 (17.7) 19 (16.1) Purple only 0 (0) 0 (0) Array Group [chi square] P CI BI 1 Combined 0.014 0.143 0.174 Bombus 0.728 -0.003 0.002 Others 0.830 0.085 n.a. 2 Combined <0.01 0.339 0.490 Both 0.659 0.124 0.128 Purple only n.a. n.a. n.a.
|Printer friendly Cite/link Email Feedback|
|Author:||Oyama, Ryan K.; Jones, Kristina N.; Baum, David A.|
|Publication:||The American Midland Naturalist|
|Date:||Oct 1, 2010|
|Previous Article:||Assessing plant community changes over sixteen years of restoration in a remnant Michigan tallgrass prairie.|
|Next Article:||Natural History Informing Research: A Review of The Freshwater Mussels of Ohio.|