| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
2 An der Windkunst 2, D-52134 Herzogenrath, Germany; 3 Department of Biology, Southwest Texas State University, San Marcos, Texas 78666 USA; and 4 Department of Biology, University of Missouri-St. Louis, St. Louis, Missouri 63121 USA and The Missouri Botanical Garden, St. Louis, Missouri 63166 USA
Received for publication October 26, 1999. Accepted for publication February 24, 2000.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONLITERATURE CITED |
|---|
Key Words: bee pollination • beetle pollination • fly pollination • Nuphar • Nymphaeaceae • pollinator spectra • self-compatibility
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONLITERATURE CITED |
|---|
; Padgett, Les, and Crow, 1999
). An earlier assessment (Beal, 1956
) that recognized only two species worldwide cannot be upheld (see also Wiersema and Hellquist, 1997
). Pollination has been studied in N. advena (Aiton) W. T. Aiton from the southeastern United States (Schneider and Moore, 1977
, and earlier work summarized therein), N. lutea (L.) Sm. from Europe and southern Eurasia (Ervik, Renner, and Johanson, 1995
; Lippok and Renner, 1997
, and earlier work summarized therein), and N. pumila (Timm) DC. from eastern Scandinavia and northern Eurasia (Lippok and Renner, 1997
). A few observations on the northwestern American species N. polysepala Engelmann by V. Grant and B. Meeuse were summarized by Schneider and Moore (1977)
. Below (Results and Discussion) we add new details of Grant's observations on N. polysepala based on correspondence with him (letters to SSR of 12 and 25 August, 1999).
These four species appear to have different pollinator spectra, with the European species pollinated by flies and bees, the American species by beetles, with flies and bees playing but a secondary role (Schneider and Moore, 1977
; Ervik, Renner, and Johanson, 1995
; Lippok and Renner, 1997
). The view that Nuphar as a genus is closely adapted for pollination by beetles comes from a study of N. advena in central Texas (Schneider and Moore, 1977
; following then prevailing broad species concepts [Beal, 1956], Schneider and Moore treated N. advena as ssp. macrophylla (Small) Beal. of the European species N. lutea). Schneider and Moore found an average of three long-horned leaf beetles, Donacia piscatrix Lac., in 32 N. advena 1st-d (female stage) flowers. The beetles carried copious amounts of pollen, and Schneider and Moore judged them more effective pollinators than the sweat bees, honey bees, and flies that also visited the flowers "because of [the beetles'] abundance and the length of time which they remain in the flowers, whereas the bees soon departed ... [ ]." Whether such an assessment holds would depend on the plant's mating system, especially on the long-term effects of selfing, but nothing is known about genetic neighborhood sizes in Nuphar.
Robertson (1889), studying the same species, N. advena, in Illinois and Florida, found it mainly pollinated by sweat bees (Halictus pectoralis) and shore flies (Notiphila sp.), and he regarded Donacia piscatrix as "worse than useless when it comes to pollination."
The fossil record of Donaciinae in North America goes back about 58 million years, with one of the fossils from Alberta belonging to subgenus Donacia (Donacia), the most derived group (ten Nearctic species and five Old World species), which feeds and oviposits exclusively on submerged portions of nymphaeaceous leaves, stems, and peduncles (Hoffman, 1940
; Askevold, 1988, 1990, 1991
). Perhaps influenced by the beetles' exclusive dependence on Nymphaeaceae, which in many areas translates as Nuphar, there has been a tendency to assume reciprocal dependence on the part of the flowers on the beetles for pollination. Thus, the numerous stamens, flat stigmatic surfaces, nocturnal flower closure, and intense sweet scent of Nuphar flower have been interpreted as "primary adaptations to assure pollination by beetles," and it was suggested that pollination by bees and flies developed later: "the evolution of other insects (e.g., bees) and other aquatic plants [...] has brought about the appearance of new pollinators of Nuphar and the adaptive radiation of Donacia" (Schneider and Moore, 1977
). That beetle pollination is the ancestral condition in Nuphar, as well as being the predominant pollination mode in the genus today, became widely accepted in the literature and is reflected in statements such as "the flowers of Nuphar have a close relationship with beetles of the genus Donacia, which complete their life cycle in association with the plant, during which time they facilitate pollination" (Schneider and Williamson, 1993
).
However, investigations of N. lutea and N. pumila in the Netherlands, Norway, and Germany showed that these species are pollinated by bees and flies rather than by Donacia or other beetles. The flies include several apparent Nuphar specialists, such as the scatophagid Hydromyza livens and ephydrids of the genera Hydrellia and Notiphila (van der Velde, 1986
; Ervik, Renner, and Johanson, 1995
; Lippok and Renner, 1997
). Another species of Notiphila was found as a N. advena pollinator in Florida (Robertson, 1889).
The different role of Donacia beetles in the pollination of Nuphar in Europe and (apparently) different parts of North America led to the present study. By comparing beetle, bee, and fly abundances and behaviors on two American species with the European findings (Ervik, Renner, and Johanson, 1995
; Lippok and Renner, 1997
) we wanted to assess whether different floral morphologies of Old and New World species are associated with different pollinators as suggested by Padgett, Les, and Crow (1999)
. These authors have shown that Nuphar comprises two major clades. Of the four species whose pollination has been studied, N. advena and N. polysepala belong to the New World clade, and N. pumila and N. lutea to the Old World clade. A fifth species investigated here, N. ozarkana (G. S. Miller & Standley) Standley, also represents the New World clade. [Nuphar ozarkana is considered a synonym of N. advena by Wiersema and Hellquist (1997)
, whereas Padgett, Les, and Crow (1999)
accord it species status based on relative matK and internal transcribed spacer (ITS) sequence divergence among 13 species and two hybrid lineages of Nuphar.] All these species have bowl-shaped flowers with seemingly easily accessible pollen and nectar rewards. However, New World and Old World flowers differ markedly in anther lengths and stigmatic disk sizes (Padgett, Les, and Crow, 1999
). Different-sized reproductive parts or different amounts of nectar might be differentially attractive to flies, bees, or beetles. Given the widely different assessments of the role of Donacia beetles as Nuphar pollinators (Robertson, 1889
; Schneider and Moore, 1977
), we also investigated these beetles' effectiveness as pollinators experimentally at the same site as Schneider and Moore (1977)
.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONLITERATURE CITED |
|---|
). To evaluate the mating system of N. advena we bagged flowers just prior to opening, subjected them to one of the following treatments, and then rebagged them. (1) Flowers were emasculated prior to anther dehiscence to test for the presence of agamospermy. (2) Flower buds were bagged without further manipulation to test for spontaneous selfing. (3) First-day flowers were emasculated and cross-pollinated with pollen from a plant 200 m away. (4) Early 2nd-d flowers, still in the receptive stage, were emasculated and cross-pollinated with pollen from a plant 200 m away. (5) Early 2nd-d flowers were self-pollinated with pollen from their own freshly dehisced anthers. Seed set in naturally pollinated flowers was quantified for comparison with that in experimental flowers. Hand-pollinations were carried out by excising stamens and rubbing their anthers over the stigmatic disks. Because of the clonal nature of pond-lilies, widely spaced plants were used to ensure cross-pollination. Bags were removed after anthesis so as not to hinder fruit development. Developing fruits were collected ~10 d to 2 wk later and fixed in 70% ethanol. Developing seeds and aborted ovules were counted to determine percentage seed set.
Donacia beetles' role as pollinator was tested by placing floating mesh cages over N. advena plants with buds ready to open and adding D. piscatrix and/or D. texana individuals to each cage; we could not securely distinguish these species in the field. Beetles were inspected for visible pollen loads, but it is nevertheless possible that they had pollen grains from earlier visits to Nuphar flowers on them. Because flowers at San Marcos usually had two or three beetles in them, we added 2–3 beetles per flower per cage; in total, 30 beetles were added to ten caged flowers. Beetles were placed inside cages, but not directly onto flowers, which resulted in some flowers being visited by several beetles, others by none. The cages completely excluded bees and flies.
|
RESULTS AND DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONLITERATURE CITED |
|---|
). The entity occurring in Missouri is considered a species, N. ozarkana, by Padgett, Les, and Crow (1999)
based on relative matK and ITS sequence divergence among 13 species and two hybrid lineages of Nuphar. Based on morphology it was considered a form of N. advena not worthy of taxonomic recognition (Wiersema and Hellquist, 1997
). Our measurements (below) and ecological observations suggest to us that the genetic distinctness of N. ozarkana from N. advena is paralleled by floral-morphological differences. Flowers of N. ozarkana measure ~2.5 cm in diameter and are borne on rigid peduncles ~2 cm above the water surface. They have six broadly spatulate sepals (N = 9 flowers), which are tinged green on the outside and dark red on the inside, and on average ten petals that are completely yellow. The ~119 stamens and petaloid stamens are oblong-linear, 7 mm long, and packed below the stigmatic disk before anthesis. They open centripetally. As is typical of Nymphaeaceae, there is a gradual transition from stamens to petals, with petaloid and staminoid intermediates. The ovary is ovoid and crowned by a circular flattened disk with typically 11 rays of stigmatic tissue. There are 25–90 ovules per ovary (N = 8 ovaries), and the mature fruits contain on average 41 seeds (N = 6 fruits; range 43–72).
Flowers of N. advena are distinctly larger than those of N. ozarkana, measuring ~3.5 cm in diameter (vs. 2.5 cm in N. ozarkana) and having on average 204 stamens and petaloid stamens (vs. 119). The sepals are slightly obovate, and there are 18 petals (vs. 10 in N. ozarkana; N = 5 flowers). Red-tinged sepal insides, common in N. ozarkana, are not seen in N. advena. The stigmatic disk has 16, rather than 11, rays. On average, there are 327 ovules per ovary (vs. 25–90 in N. ozarkana; N = 10 ovaries), and the mature fruits contain on average 252 seeds (N = 11 fruits; range 186–353).
First-day (female-stage) flowers of both species remain almost completely closed except for a small triangular opening just above the stigmatic disk. The only way for insects to enter these flowers and reach the petal nectaries is to land on, or crawl over, the stigma, which on the first day is covered with a sticky mucus. The mucus is devoid of glucose as assayed by diabetes test strips and does not appear to serve as a pollinator reward. Flowers of N. ozarkana emitted no perceptible odor, even after fresh flowers had been placed in a clean glass vial for ~20 min. By contrast, those of N. advena smelled like immature apricots. Fresh petal nectaries of N. ozarkana did not produce visible nectar drops. However, when the same glucose test strip was touched to all ten nectaries of a flower it stained weakly. In N. advena, by contrast, single nectaries produced visible drops that strongly stained the glucose test strip. In neither species did the nectaries stain with neutral red (as also reported by Schneider and Moore, 1977
).
Flowers of both species opened around 0800, depending on sun intensity. They were functionally female on the first day, because pollen sacs only open on the second day. First-day flowers closed completely between 1800 and 1900, while 2nd- and 3rd-d flowers remained partially open. On the second day, the outer row of stamens bent backwards, presenting pollen, while the mucus on the stigmatic rays gradually dried out. During each of the following three days, one or two rows of anthers matured and presented pollen, while the now completely dry stigma changed color from dark yellow to brown. Following the maturation of the last row of anthers, the petals and stamens withered, peduncles lost their rigidity, and fruit ripening proceeded.
Flowers of N. ozarkana were visited by three species of sweat bees, Lasioglossum (Evylaeus) nelumbonis Robertson, Dialictus bruneri (Crawford), and Augochlora pura pura (Say), that together made up 97% of all visits. The syrphid flies Paragus sp., Chalcosyrphus metallicus Wiedemann, and Toxomerus geminatus Say together accounted for 3% of the visits. There was a single visit by a species of Bombus. The bees and syrphids collected or ate pollen and occasionally sat on the stigmatic disk to groom themselves. Pollen-seeking bees predominantly visited older, pollen-presenting flowers. Donacia piscatrix Lac. was observed only on 2- or 3-d old flowers that were in late male stage, and of 38 flowers checked during one night, five contained one beetle each.
Flowers of N. advena at San Marcos received fewer insect visits than those of N. ozarkana at Poplar Bluff (279 vs. 591 visits to a total of 26 flowers studied for comparable lengths of time). Sixty-nine percent of all visits were by Lasioglossum (Evylaeus) nelumbonis, the same bee that was the most abundant visitor to Nuphar at Poplar Bluff. Syrphids made up 3.9% of all visits and Xylocopa bees 11%. Two species of Donacia were found inside the flowers, D. piscatrix and D. texana Crotch. Together they accounted for 15% of all visits to flowers during the 20 h of observation. Sometimes, beetles were observed pushing their way through the sepals of N. advena and into a flower before it fully opened, apparently to feed on nectar. Of 22 flowers checked during one night (some being completely closed 1st-d flowers, others partially open 2nd- and 3rd-d flowers), each contained at least one beetle, and several contained copulating pairs (a total of 29 beetles were seen in these flowers). At both study sites, a few of the outer sepals were visibly gnawed by Donacia. One beetle gnawing on a petal of N. advena had pollen attached to its abdomen, and several dissected beetles had abundant N. advena pollen in their guts.
Results of breeding system experiments (Table 1) indicate that N. advena is incapable of agamospermy and does not spontaneously self-pollinate. Experimental selfing of 2nd-d flowers yielded 9% seed set; 1st-d flowers cannot be autogamously pollinated because of the flowers' strong protogyny. Experimental outcrossing of 2nd-d flowers resulted in a similarly low seed set (12%), indicating that stigmas are less receptive on the second day. Experimental cross-pollination of 1st-d flowers and natural pollination yielded high seed sets (62 and 78%, respectively). Six of ten flowers that bloomed inside the mesh cages with "beetles-only" as potential pollinators set no or few seeds and appeared not to have been visited. Four had an average set seed of 69% due to beetle pollination. Since experimental crossing and selfing of 2nd-d flowers resulted in much lower seed sets (9 and 12%), the 69% seed set in the cages likely results from beetle-mediated geitonogamous pollination of 1st-d flowers (flowers inside cages belonged to single plants). Taken together, these figures suggest that natural seed set in N. advena is mostly due to xenogamy or geitonogamy of 1st-d flowers. In European N. lutea and N. pumila, by contrast, experimental selfing and outcrossing of 2nd-d flowers yielded seed sets almost as high as those in 1st-d flowers (Lippok and Renner, 1997
; Table 1).
|
during a 12-mo study of N. advena found halictids (Halictus sp., which may be identical with Lasioglossum nelumbonis) and honey bees to be the most frequent visitors, together with Donacia piscatrix. They observed the bees to "visit more flowers in a given amount of time [than the Donacia beetles]." We found Donacia to be infrequent on N. ozarkana, where they were only seen in 2nd- or 3rd-d flowers whose stigmas had dried up and to be abundant on N. advena as earlier observed by Schneider and Moore (1977)
. Possibly the larger size of N. advena flowers, intense apricot smell, and more copious nectar (each flower has 18, rather than 10, petals and therefore nectaries) contribute to that species' higher attractivity for beetles. Also, two species of Donacia occur at San Marcos, but only one at Poplar Bluff, and this may explain why more beetles were found on San Marcos flowers. In a monograph of Donacia, Askevold (1990
, p.610) reports that D. piscatrix and D. texana have mouthparts adapted for feeding on plant tissue, including, but not restricted to Nymphaeaceae, and for cutting holes into leaf or peduncle surfaces into which they oviposit, rather than for feeding on pollen or nectar. However, D. piscatrix and D. texana individuals we dissected contained abundant Nuphar pollen. We also agree with Schneider and Moore (1977)
that beetles appear to be taking up nectar from the petal nectaries. In addition to feeding on nectar and pollen, D. piscatrix also mates in Nuphar flowers (Marx, 1957
, p.242).
Verne Grant (personal communication in Schneider and Moore [1977]
and letters to SSR of 12 and 25 August, 1999), who studied two populations of N. polysepala Engelm. on two days in July 1949 near Echo Lake, El Dorado County, Sierra Nevada, California, found the flowers frequented by Donacia beetles, syrphids, and muscid-like flies; no bee visits were observed. Donacia beetles were abundant on flowers in one population, but only one beetle was found in the other population. Beetles, syrphids, and muscid-like flies had their venters covered with Nuphar pollen, and since they usually landed on the stigmas were effective pollinators.
The precise role of Donacia as pollinators in nine species of Nuphar remains to be investigated. The species occurring in China (N. sinensis Hand.-Mazz.) and Japan (N. oguraensis Miki, N. japonica DC., and N. pumila) have flowers very similar to those of N. lutea, and indeed Donacia gracilipes Jacoby and D. nitidor (Nakane) have been recorded as visitors to the last two species (Kimoto, 1981
). Of the New World species yet to be studied, N. ulvacea (Mill. & Standl.) Standl., N. orbiculata (Small) Mill. & Standl., N. variegata Durand, and N. sagittifolia (Walt.) Pursh are similar to N. advena, N. polysepala, and N. ozarkana (Padgett, Les, and Crow, 1999
). The North American N. microphylla (Pers.) Fern. is part of the Old World clade, and its flowers are similar to those of N. pumila. All these species possess petal nectaries and, on the second day, abundant pollen, and both rewards are accessible to flies, bees, and beetles.
Padgett, Les, and Crow (1999)
suggested that the different anther lengths of Old World and New World species may influence pollinator effectiveness and selection (i.e., result in shifts in pollinator spectra). Analyses of pollen deposition on pollinators' body surfaces (and subsequent deposition patterns on stigmas) would be necessary to substantiate a correlation between anther length and pollinator effectiveness in Nuphar. However, Padgett, Les, and Crow's hypothesis can be assessed preliminarily by comparing pollinator spectra in the two Old World and three New World species studied so far (keeping in mind the unequal and in the case of N. polysepala short observation periods). Nuphar lutea and N. pumila (three populations studied) are mainly visited by flies, with bees playing a secondary role; N. polysepala (two populations) is visited by flies and Donacia; N. advena (one population studied in detail) is mainly visited by halictids, flies, and Donacia; and N. ozarkana (one population) is mainly visited by halictids and flies. Flies are thus the only insects visiting all species, while visitation by bees and beetles varies among species and sites (for example, Donacia is absent from N. lutea populations studied in Germany, but is an occasional visitor in Norway; Lippok and Renner, 1997
). The relative contribution of flies, bees, and beetles in any one population of Nuphar thus appears to depend more on these insects' relative abundances (and in the case of Donacia, presence) and alternative food sources than on stamen length differences between Old World and New World pond-lilies.
An earlier study of N. advena found that bagged flowers produced some seeds (Schneider and Moore, 1977
; seed set was not quantified), while our bagged flowers set no seeds. All species of Nuphar are protogynous, with the only overlap between the female and male stages occurring during the morning of the second day when the stigmatic rays are still receptive (Ervik, Renner, and Johanson, 1995
; Lippok and Renner, 1997
; this report). This would be the time when spontaneous selfing might occur were it not for the position of the stamens below the stigmatic disk and their extrorse dehiscence, which together preclude pollen from reaching the stigma unless via some outside agent. Insect-mediated selfing of 2nd-d flowers is possible because, like other pond-lilies studied (Ervik, Renner, and Johanson, 1995
; Lippok and Renner, 1997
; Table 1), N. advena is self-compatible. Whether species of Nuphar, like many clonal plants, are adapted to constant selfing (in the form of geitonogamy) or maintain mixed-mating systems is unknown.
|
FOOTNOTES |
|---|
5 Author for correspondence (e-mail: Renner{at}admiral.umsl.edu
).
|
LITERATURE CITED |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONLITERATURE CITED |
|---|
———. 1990 Classification of Tertiary fossil Donaciinae of North America and their implications about evolution of Donaciinae (Coleoptera: Chrysomelidae). Canadian Journal of Zoology 68: 2135–2145.[CrossRef]
———. 1991 Classification, reconstructed phylogeny, and geographic history of the New World members of Plateumaris Thomson, 1859 (Coleoptera: Chrysomelidae: Donaciinae). Memoirs of the Entomological Society of Canada 157: 5–175.
Beal, D. E. 1956 Taxonomic revision of the genus Nuphar Sm. of North America and Europe. Journal of the Elisha Mitchell Scientific Society 72: 319–346.
Ervik, F., S. Renner, and K. A. Johanson. 1995 Breeding system and pollination of Nuphar lutea (L.) Smith (Nymphaeaceae) in Norway. Flora 190: 109–113.[ISI]
Hoffman, C. E. 1940 Limnological relationships of some northern Michigan Donaciini (Chrysomelidae: Coleoptera). Transactions of the American Microscopical Society 59: 259–274.[CrossRef]
Kimoto, S. 1981 New or little known Japanese Donaciinae (Coleoptera: Chrysomelidae). Bulletin of the Osaka Museum of Natural History 34: 27–46.
Lippok, B., and S. S. Renner. 1997 Pollination of Nuphar (Nymphaeaceae) in Europe: flies and bees rather than Donacia beetles. Plant Systematics and Evolution 207: 273–283.[CrossRef][ISI]
Marx. E. J. F. 1957 A review of the subgenus Donacia in the Western Hemisphere (Colepotera: Donacidae). Bulletin of the American Museum of Natural History 112: 191–278.
Padgett, D. J. 1997 Molecular study of yellow pond lilies: the study of Nuphar DNA confirms, and contradicts relationship theories in Nuphar. Water Garden Journal 13: 34–37.
———. 1998 Phenetic distinction between the dwarf yellow water-lilies: Nuphar microphylla and N. pumila (Nympheaeceae). Canadian Journal of Botany 76: 1755–1762.
———, D. H. Les, and G. E. Crow. 1999 Phylogenetic relationships in Nuphar (Nymphaeaceae): evidence from morphology, chloroplast DNA, and nuclear ribosomal DNA. American Journal of Botany 86: 1316–1324.
Robertson, C. 1889 Flowers and insects I. Botanical Gazette 14: 122–123.
Schneider, E. L., and L. A. Moore. 1977 Morphological studies of the Nymphaeaceae. VII: The floral biology of Nuphar lutea ssp. macrophylla. Brittonia 29: 88–99.[CrossRef][ISI]
———, and P. S. Williamson. 1993 Nymphaeaceae. In K. Kubitzki, J. G. Rohwer, and V. Bittrich [eds.], The families and genera of vascular plants, vol. 2, 486–493. Springer, Berlin, Germany.
van der Velde, G. 1986 Developmental stages in the floral biology of Dutch Nymphaeaceae (Nymphaea alba L., Nymphaea candida Presl, Nuphar lutea (L.) Sm.). Acta Botanica Neerlandica 35: 111–113.[ISI]
Vogel, S., 1990 The role of scent glands in pollination. Smithsonian Institution Libraries, Washington, D.C., USA.
Wiersema, J. H., and C. B. Hellquist. 1997 Nymphaeaceae. In Flora of North America, vol. 3, 66–77. Oxford University Press, New York, New York, USA.
This article has been cited by other articles:
|
|
 |
|
  P. S. Soltis and D. E. Soltis The origin and diversification of angiosperms Am. J. Botany, October 1, 2004; 91(10): 1614 - 1626. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. BERNHARDT, T. SAGE, P. WESTON, H. AZUMA, M. LAM, L. B. THIEN, and J. BRUHL The Pollination of Trimenia moorei (Trimeniaceae): Floral Volatiles, Insect/Wind Pollen Vectors and Stigmatic Self-incompatibility in a Basal Angiosperm Ann. Bot., September 1, 2003; 92(3): 445 - 458. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
