Swamp rose-mallow (Hibiscus moscheutos) COSEWIC assessment and status report: chapter 6
Vegetative reproduction appears to be important in Hibiscus moscheutos, with clumps able to produce new flowering stems yearly. Clumps may also become fragmented and dispersed by wind and wave action, facilitating the colonization of new sites. Most pollination is accomplished by a single species of non-social bee, Ptilothrix bombiformis, and the appearance and disappearance of adults is largely coincidental with the flowering of H. moscheutos, with much of the bee's activity centres around these plants (Blanchard, 1976). Other visitors to flowers noted are several species of moths, butterflies, small bees and flies, but none appear to be effective pollinators. It is important to note that P. bombiformis has not been reported to occur in Canada. It may be that at the edge its range, pollinators such as P. bombiformis are not present. Two beetles, Althaeus hibisci and Conotrachelus fissunguis are known to parasitize Hibiscus seeds (Blanchard, 1976).
The seeds of H. moscheutos can float for an extended period of time, and since plants occur in coastal marshes, seeds could be carried for some distance by water currents, particularly on high storm tides. The seeds are known to be eaten by Northern Bobwhite (Colinus virginianus), Blue-winged Teal (Anas discors), Northern Pintail (Anas acuta) and Wood Ducks (Aix sponsa) and have limited food value for waterfowl (Blanchard, 1976). Hibiscus moscheutos is found in open wetlands and is probably dependent upon periodic burning, flooding, drought, or anthropogenic disturbance to decrease shading from trees and shrubs and create open habitat.
At the time of the original status report, little was known about the reproductive biology of H. moscheutos. However, this has been remedied by a number of recent studies including those by Spira (1989), Snow and Spira (1991a), Snow and Spira (1991b), Spira et al. (1992), Snow and Spira (1993), Snow and Spira (1996), Snow et al. (1996), Spira et al. (1996), and Snow et al. (2000).
Populations of Hibiscus moscheutos consist of ramets (vegetative stems with the potential for independent existence), and genets (a plant that originates from a seed). Genets can be propagated vegetatively, but they are not clonal under natural conditions (Snow et al. 2000). Individual plants (genets) produce multiple shoots from a dense, fibrous, perennial rootstock. These ‘clumps’ are 1-2 metres in diameter (no lateral spreading), with a few to 70 flowering stems per clump, 1 to 2 m tall, and at peak flowering, up to 20 or more flowers can be open at once on the largest plants (Ford, 1985; Spira, 1989; Snow et al., 1996). Vegetative reproduction appears to be important, with clumps able to produce new flowering stems yearly. Clumps may also become fragmented and dispersed by wind and wave action, facilitating the colonization of new sites.
Hibiscus moscheutos has chasmogamous flowers with a breeding system that tends to favour outcrossing (Blanchard, 1976) (Chasmogamy is the production of flowers that open to expose the reproductive organs. This allows cross pollination but does not preclude self pollination). However, H. moscheutos is clearly self-compatible, and seeds can be sired by inbreeding (Spira, 1989). Automatic self-pollination is prevented by the stigmas protruding well beyond the uppermost anthers. Wind pollination is unlikely as the pollen grains are sticky and tend to clump together. Flowers covered by Spira with a single layer of cheesecloth (porous to pollen but not to insect pollinators) accumulated few to no pollen grains on stigmas and failed to mature fruits, indicating that flowers were not apomictic and that a vector other than wind was needed for successful pollination (Spira, 1989). Spira (1989) concluded that, “Spatial separation of anthers from stigmas (herkogamy) effectively prevents self-pollination in this self-compatible species. It does not, however, prevent pollination between flowers on the same plant (geitonogamy). Even though Hibiscus clones are multi-stemmed, the number of open flowers each day is generally less than five. Most stems do not produce an open flower on a given day and those that do tend to have only one or two open flowers. The relatively small number of open flowers per genet at a given time should decrease geitonogamous pollination and promote outcrossing in this species.” Snow et al. (1996) concluded that geitonogamy can lead to higher selfing rates.
The frequent visits to the showy flowers result in strong competition among pollen tubes for ovules (Snow et al., 1996). Snow and Spira (1993) concluded from their studies on relative pollen tube growth rates that the outcome of competition between pollen tubes from self and outcrossed individuals is variable, and that pollen tube competition does not appear to be a general mechanism for enhancing the proportion of progeny that results from outcrossing in H. moscheutos.
In H. moscheutos, the effective pollinators are solitary anthophorid bees (Ptilothrix bombiformis) and bumblebees (Bombus spp.) (Blanchard, 1976; Spira, 1989; Snow and Spira, 1993). Unlike many members of the Malvaceae, flowers are held with their axis of symmetry more or less horizontal. In large flowers, this orientation is able to restrict the available landing surfaces to the reproductive parts. In H. moscheutos, the style branches are upturned and the stigmas are large and flattened. This combination of features provides an especially attractive landing place for insects.
According to Blanchard (1976) and Spira (1989) most pollination in their studies of US populations is accomplished by a single species of non-social bee, Ptilothrix bombiformis, (family Apidae). The appearance and disappearance of adults is largely coincidental with the flowering of H. moscheutos and much of the bee's activity centres around these plants. Apparently the only pollen used by this bee comes from several species of Hibiscus. When visiting a flower, the bee lands on the upturned stigmas and then proceeds towards the base of the flower by wading through the numerous pollen-laden anthers. Departure from the flower is usually from the lower petals so that the stigmas are not touched again. The females are the primary pollen collectors, pollen being used for provisioning the nest. As the females forage for pollen, their ventral surfaces generally become covered with it. As well, bees foraging for nectar frequently crawl over the anthers to reach the nectaries at the base of the flower, and thus accumulate large amounts of pollen on their ventral surfaces. The males, on the other hand, may drink nectar but do not collect pollen. Most of their time is spent in searching flowers for females. At night and during inclement weather, males may be found sleeping within the flower, curled around the base of the staminal column (Blanchard, 1976; Spira, 1989; Spira et al., 1992).
Spira (1989) and Spira et al. (1992) concluded that most visits by P. bombiformis and Bombus did not result in pollination, as the visitor failed to make contact with any of the flower’s stigmas. In their studies only 27% of flower-foraging P. bombiformis and B. pennsylvanicus appeared to contact a stigma while foraging for nectar or pollen. However, when they did contact the stigma, they generally deposited large amounts of pollen on the stigmas (up to 889 grains), 14 times as many pollen grains as there were ovules in the ovaries.
The two pollinator species, Bombus and Ptilothrix, use petals as a cue to locate Hibiscus flowers, because flowers with 100% petal removal are almost completely ignored (Kudoh and Whigham, 1998).
Other visitors to flowers noted are several species of moths, butterflies, small bees and flies, but none appear to be effective pollinators (Spira, 1989; Spira et al., 1992).
It is important to note that P. bombiformis has not been reported to occur in Canada (Mitchell, 1962) and was not seen during this study. Apis mellifera and Bombus spp. were the only insects found visiting flowers during fieldwork for the 1985 report, however, pollinator activity was low at all stations visited. The dearth of pollinators seen during this study may simply be the result of investigation at times not conducive to insect activity. Alternatively, it may be that at the edge its range, pollinators such as P. bombiformis are not present.
Spira et al. (1992) concluded that seed production in H. moscheutos is clearly not pollen-limited. Between 65 and 97% of the flowers sampled in their study had excess pollen on their stigmas within 2 and 3 hours after exposure to pollinators, suggesting that pollen competition occurs frequently.
The fruit is a capsule that dehisces in late fall, yielding round, hard-coated seeds <3 mm. (Cahoon and Stevenson, 1986).
Seed set appeared to be high in many of the herbarium specimens examined during the original study, but the percentage of mature capsules produced in a population and the number of viable seeds found in a capsule are not known.
Schull and Tachibara (as cited in Blanchard, 1976) have shown that a high concentration of H2SO is required to bring about a high percentage of seed germination.
From their studies of H. moscheutos in Chesapeake Bay, Kudoh and Whigham (2001) concluded that fruit maturation takes 3 to 4 weeks, and most seeds are released from dehisced fruits in October and November. Each capsule produces approximately 120 seeds (Spira, 1989).
In their study of the seed bank of a freshwater tidal marsh in New Jersey, which included Hibiscus moscheutos, Leck and Graveline (1979) concluded that the lack of significance in numbers of seeds at 8 to 10 cm may support the contention that in saturated marsh soils dormancy is prolonged and longevity increased. Hall et al. (as cited in Blanchard, 1976) observed that the seeds of H. laevis remain dormant while submerged in an experimental pool, but when it was drained, they germinated rapidly and profusely. This adaptation is probably an important mechanism for the colonization of mud flats and recently drained areas by H. laevis and other related species.
Although H. moscheutos is a perennial, it is able to colonize newly created spoil banks and flower within a year of colonization.
It is presumed that as a plant matures, the root system grows and is able to produce more and more flowering stems every year. However, since large clumps can become fragmented, it is impossible to distinguish between old fragmented clumps and young plants.
There are numerous reports of cultivated crosses between H. moscheutos and other members of the genus Hibiscus. It is from some of these crosses that commercial cultivars are derived (Blanchard, 1976). Under cultivated conditions, crosses with H. moscheutos ssp. lasiocarpos, H. grandiflorus, H. laevis, H. coccineus, H. dasycalyx and H. mutabilis have been achieved with varying degrees of success. The only known naturally occurring hybrids have been with H. grandiflorus and H. laevis, but such hybrids are rare. Hibiscus grandiflorus has different pollinators and is largely allopatric with H. moscheutos, while H. laevis produces semi-lethal hybrids and is separated by ecological barriers (Blanchard, 1976). Wise and Menzel (1971) noted diminished fruit and seed set in crosses between members of southern United statespopulations of H. moscheutosand H. laevis. Stout (1917) reported from his trials that races of H. moscheutos hybridized readily with H. laevis, giving highly fertile F1 progeny. Klips (1999) has noted that the habitats of H. moscheutos (open marshes) and H. laevis (along slow-moving river banks) sometimes merge with one another, and where plants of both species are proximal enough to allow individual pollinators to visit flowers of both species, he believes a small amount of hybridization is likely to occur. He states that hybrids should be infrequent, due to the apparent pollen competition detected in his study, but cautions that given the readiness with which a few hybrid progeny were formed under his observation, that introgression of genes from one species into the populations of the other might occur. This had not been detected between H. moscheutos and H. laevis by the time of his work in 1999.
A dwarf race of H. moscheutos was reported at Long Island, New York (Stout, 1917), with no plant over 26 inches tall, and all plants evidently several years old. Stout transplanted 25 of these plants to experimental plots at the New York Botanical Garden. If the results of this transplant were ever published, they could not be located by the authors.
Spira (1989) has suggested that severe regional drought conditions can likely influence fruit set in H. moscheutos, and during the July and August of 1986, when plants were flowering and setting fruit, he noted symptoms of water stress such as drooping stems and wilted leaves. He found that set was extremely low in that drought year, as compared with the previous non-drought year, but that seed set within those fruits that did develop, remained quite high.
From a study site in Maryland, Snow and Spira (1996) studied the effects of salinity and high soil nutrients on pollen performance in H. moscheutos. They observed that their salinity and fertilization treatments resulted in reduced vegetative growth, fewer flowers, and smaller petals, as compared with the control treatment, but no change occurred in style length or paternal success following mixed-donor pollinations. The high nutrient treatment led to slightly improved growth and larger petals as compared with controls, yet this treatment also had no effect on style length or pollen competitive ability. They concluded that style length, and most importantly, the number of seeds sired, were buffered from the effects of environmental variation (they cautioned that other untested environmental conditions could influence this trait), whereas flower production and petal length were not.
Hibiscus moscheutos is a long flowering species, blooming from July 25 to September 25, with the height of flowering being reached in the second week of August (Botham, 1981). Up to eight blooms may be present on a stem with large clumps possessing hundreds of flowers. Single flowers arise in the axils of the upper leaves and are open only a few days before withering. In the course of development, flowers show a sequence of changes in orientation. In the bud stage they are erect, while at the time of anthesis (the period from flower opening to fruit seed set) the peduncles bend to direct the open flowers horizontally. Following anthesis, the peduncles thicken and lengthen with maturation of the fruit. Mature fruits are noticeable soon after flowering and remain on the plant. The peduncles, however, do have an abscission zone above the insertion to the leaf axil, and it is through this zone that aborted and unfertilized flowers are abscised (Blanchard, 1976).
The diurnal periodicity of Hibiscus has been the object of study for years. The world-renowned ornithologist Alexander F. Skutch studied H. moscheutos in Maryland 75 years ago (Skutch & Burwell, 1928). By mid-August, they found that most flowers unfold to practically their full extent by 7:30 or 8:00 a.m. They observed pollination, and then noted that by 4:00 p.m. the corollas had closed perceptibly, and by 6:00 they were completely closed. They found that flowers bloomed only a single day, and did not open the following morning, their period of full bloom being nine hours or less. Through their experiments, they realized a difference between the pollinated and unpollinated flowers of Hibiscus. If the pollination of the flowers was prevented, the flowers would remain open for two or more days. Through this work, they drew attention to the importance of considering pollination success when determining anthesis in flowers, a factor which had apparently been overlooked dating back to Linnaeus’s “floral clock”.
The seeds are known to be eaten by Northern Bobwhite, Blue-winged Teal, Pintail and Wood Ducks (Blanchard, 1976). The seeds are hard coated and may be expected to pass through the digestive tract intact (Blanchard, 1976). They have a limited food value for waterfowl (McCormick and Somes, 1982, as cited in Cahoon and Stevenson, 1986).
The seeds can float for an extended period of time, and since plants occur in coastal marshes, seeds could be carried for some distance by water currents, particularly on high storm tides (Cahoon and Stevenson, 1986). Spira (1989) has proven that the seeds are buoyant and appear to be dispersed by water. In a study of the seed bank of a freshwater tidal marsh in New Jersey, Leck and Graveline (1979) found that seedlings of H. moscheutos were abundant along a stream bank, suggesting effective dispersal by water (hydrochory). Kudoh and Whigham (2001) have demonstrated the importance of hydrochory to metapopulations of Hibiscus moscheuotos in the intertidal habitat of Chesapeake Bay. In their 1997 study, also at Chesapeake Bay, they concluded that pollinator behaviour cannot solely explain the almost complete panmixia (broad interchange of alleles) within H. moscheutos populations, and suggested that seeds are widely dispersed when sites are flooded. They felt that spatial mixing of genotypes by hydrochory probably reduces the effect of biparental inbreeding. Despite the evidence for gene flow between populations of H. moscheutos, significant genetic structuring among populations occurs, primarily when populations are somewhat isolated from the tidal creeks (Kudoh and Whigham, 1997).
Nutrition and Interspecific Interactions
As with most flowering plants H. moscheutos can have a close association with its pollinators and is dependent upon them for successful sexual reproduction. However, not all insects are positively associated with rose mallow. Two beetles, Althaeus hibisci and Conotrachelus fissunguis are known to parasitize Hibiscus seeds. Althaeus hibisci can be found feeding on pollen and nectar, and gathering in the space between the corolla and calyx. Hidden in this space, the females wait for the corolla to wither, indicating that pollination has occurred. The females then begin ovipositing on the newly fertilized ovary. Upon hatching the larvae burrow through the ovary wall in the locule and then into the developing seed. The seeds appear to go through a normal development, while the larvae devour the seeds' contents. The larvae pupate in the seed and the adults emerge at the time the capsule dehisces (Blanchard, 1976). Blanchard (1976) has found high levels of parasitism in the wild and implies that infestation can become high enough to adversely affect the reproductive success of a population.
The other beetle that is found to be destructive to H. moscheutos' seeds is the weevil C. fissunguis. Apparently the adults feed on the bases of the petals and deposit their eggs within the maturing capsule. The larvae feed on the seed contents and locule wall and at dehiscence drop to the ground. The larvae pupate beneath the soil surface (Blanchard, 1976).
In a two-year study (1985, 1986) conducted on Chesapeake Bay populations of H. moscheutos, Spira (1989) found that approximately 53% and 89% of the potential seeds (88% were viable) within fruits were destroyed by either A. hibisci or C. fissunguis. He concluded that pre-dispersal seed predators dramatically reduced reproductive output in H. moscheutos. From their predator studies at a Hibiscus marsh in Maryland, Kudoh and Whigham (1998) concluded that final seed set varied considerably depending on the larval densities of Althaeus hibisci and Conotrachelus fissunguis. And Bauman et al. (2001) found from a study on the Lake Erie shore of Ohio that damage by both A. hibisci and C. fissunguis was greatest for flowers that opened before peak flowering and decreased as the season progressed. With the high levels of fruit and seed damage they observed, they hypothesized that synchronized flowering may be strongly advantageous in H. moscheutos, and that damage by seed predators appears to have a greater effect on plant fecundity than pollinator service, because previous studies had shown that seed production was not pollinator-limited.
Other insects known to parasitize H. moscheutos are the stem-boring buprestid Paragrilus tenuis, the stem-nesting sphecid wasp Ectemnius paucimaculatus, the sawfly Atomacera decepta, a leafroller Chionodes hibiscella, and the moth Acontia delecta (Cahoon and Stevenson, 1986; Blanchard, 1976). From their study sites in New Jersey, Weiss and Dickerson, 1919, cited in Cahoon and Stevenson, 1986), concluded that H. moscheutos appears to provide insect populations with a highly palatable and diverse substrate situated above the catastrophic effects of tides and storms, and they surmised that this may account for the 30 insect species known to infest the leaves, stems, and flowers of the species. It is not known how prevalent these various parasites are in the Ontario populations of H. moscheutos.
Hibiscus moscheutos is found in open wetlands and is probably dependent upon periodic burning, flooding, drought, or anthropogenic disturbance to decrease shading from trees and shrubs and create open habitat. Farney and Bookout (1982) describe how high water levels in Lake Erie converted emergent vegetation into open water and virtually eliminated large common cover types, such as H. moscheutos. On the other hand, large areas of shallow water favoured expansion of rose mallow populations and plants did best under a management regime of controlled water levels with partial yearly drawdowns. High water levels, as well as total drawdowns, may have a detrimental effect on rose mallow populations. However, if these conditions are of a short duration they are probably beneficial in eliminating competing species and in creating open conditions.
The hardy rose-mallows were introduced into cultivation very early in the United States and Europe, with a listing of Hibiscus moscheutos and H. palustris by John Bartram and Son, Philadelphia, in 1807. From about 1850 the different species of Rose-mallows were regularly listed in European nursery catalogues, and at least one race of H. moscheutos has become naturalized there (Winters, 1970). Since the early 1900s, successful hybridization of H. moscheutos with the progeny of H. coccineus, H. laevis, and H. grandiflorus has produced several widely-used F1 hybrids – notably dixie belle, southern belle, and its semi-dwarf version disco belle. These hybrids are available as seed from catalogue companies, e.g. Thompson & Morgan, Chiltern, Park, Stokes, and Sakata, and are appropriate for USDA zones 5 to 10. The garden varieties bloom in 135 to 150 days from seed. They can be started in February to March in a warm greenhouse and grown in large individual pots. Soaking the seed for 24 hours prior to sowing is beneficial to germination. Germinate for 1 to 2 weeks at 21°C. The hybrid found for years in the nursery trade as Meehan’s Mallow Marvels, was obtained by crossing a hardy red-flowered hybrid of H. coccineus X H. militaris with H. moscheutos (Wise and Menzel, 1971; Vesterin, 1997).
H. moscheutos plants withstand transplanting easily when in full flower, and are also readily grown from seed and will flower the first season if sown early (Winters, 1970).
Report a problem or mistake on this page
- Date modified: