| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Reproductive Biology |
Department of Botany, Graduate School of Science, Kyoto University, Kyoto 606-8502, Japan
Received for publication September 14, 2006. Accepted for publication April 2, 2007.
ABSTRACT
Among the Cycadales (Cycadaceae and Zamiaceae), the Zamiaceae are known to be insect-pollinated. In contrast, the Cycadaceae are still considered wind-pollinated, although some doubt has been cast on several species, including Cycas revoluta. Using a large population of C. revoluta on Yonaguni Island (Okinawa, Japan), we performed exclusion experiments, documented insects from male and female cones, and analyzed the morphology of the apical part of the ovule to determine the pollination method of this species. Insect exclusion resulted in a notable reduction in seed set, except in a few individuals growing near male cones. The amount of airborne pollen was abundant within a 2-m radius of male cones but decreased markedly beyond this distance. Pollen grains of C. revoluta were found on the body of Carpophilus chalybeus (Nitidulidae, Coleoptera), one of a few species of insects collected from both male cones and female cones far from males. We conclude that C. revoluta relies on both wind (anemophily) and insect pollination (entomophily), although such anemophily is restricted to female trees growing within a 2-m radius of male trees. The nitidulids are not host specific to this cycad and primarily feed on plant tissue but serve as pollinators during pollen release. Cycas revoluta appears to be in an initial mode of animal pollination, as opposed to the host-specific insect pollination observed in most Zamiaceae.
Key Words: ambophily • anemophily • Cycadaceae • Cycas revoluta • entomophily • Nitidulidae • pollination
The Cycadophyta, an early diverging plant group among extant gymnosperms, comprises two families, the Cycadaceae (Cycas only, ca. 100 species) and the Zamiaceae (nine or 10 genera, ca. 200 species; Hill et al., 2003
, 2004
). This group has been considered wind-pollinated, or anemophilous, since Chamberlain (1935)
reported that wind-pollination was characteristic of all gymnosperms. However, Norstog et al. (1986)
showed by means of exclusion experiments that Zamia furfuracea (Zamiaceae) is pollinated by the weevil Rhopalotria mollis (Curculionidae). Pollination by weevils and thrips has since been reported for other species of Zamiaceae, including Z. pumila (Tang, 1987
), Macrozamia spp. (Connell and Ladd, 1993
; Mound and Terry, 2001
; Terry, 2001
; Terry et al., 2005
), Encephalartos spp. (Donaldson et al., 1995
; Donaldson, 1997
), Bowenia spp. (Wilson, 2002
), and Lepidozamia peroffskyana (Hall et al., 2004
). Thus, pollination by insects, which are all host-specific (for review see Norstog and Nicholls, 1997
; Hall et al., 2004
), is now widely reported among the Zamiaceae.
On the other hand, pollination in the Cycadaceae (Cycas) remains controversial. Several reports have discussed pollination mechanisms in different species of Cycas. By means of wind-tunnel experiments and showering cones with water, Niklas and Norstog (1984)
suggested that pollination in C. circinalis may have two phases: the transport of wind-borne pollen grains to megasporophylls and then the subsequent transport of adhering pollen to ovules by water and/or wind. For C. panzhihuaensis, Wang et al. (1997)
suggested that after pollen grains are first transported by wind from male to female cones, they may be carried toward the ovules in the female cones by both insects (ants and/or cockroaches) and "droplets" secreted on the megasporophylls. According to Keppel (2002), a female C. seemanni plant cultivated at the Sacred Heart Cathedral in Suva, Fiji, which was shielded from wind on three sides by buildings and thus was subject to wind currents only from one direction, bore the majority of seeds on the windward side of the cycad. Keppel further noted that seeds were not produced in a rainforest cycad population but were produced in a population in the adjacent coastal region. Based on these data, Keppel concluded that wind is the major pollination agent in C. seemanii.
In contrast, Vorster (1995a
, b) questioned wind pollination in several Cycas species. With respect to C. revoluta (based on three individuals, one male and two female, cultivated in a garden in South Africa), he reported that both male and female cones emitted a strong odor and had an increase in temperature at the time of pollination; pollen grains were sticky; no spontaneous pollination occurred among individuals that grew within 1.5 m of each other (Vorster, 1995b
). Both a strong odor and a temperature increase have also been observed in C. rumphii and C. thouarsii (Vorster, 1995a
). Vorster (1995a)
interpreted these phenomena as a mechanism to attract insects. Insect pollination was further suggested by Norstog and Fawcett (unpublished observation in Norstog and Nicholls, 1997
). They found that while about a dozen specimens of C. media that grew closely spaced near City Hall in Cairns, Australia, did not set seed, individuals growing in the adjacent woodlands produced perfectly fertile seeds. Furthermore, by examining weevil activity ("insect bore-holes in sporophylls, frass in male cones, etc.," p. 150) in those wild populations, they concluded that C. media is probably insect-pollinated. Various species of insects reported on male and/or female Cycas cones have been considered pollinators (Ornduff, 1991
; Forster et al., 1994
; Lindstrom, 1999
; Tang et al., 1999
) or predators (Kato, 2001
).
Overall, Norstog and Nicholls (1997)
indicated that the evidence for and against wind vs. insect pollination in Cycas is primarily anecdotal or has not been experimentally verified. To date, no concrete evidence for wind or insect pollination has been provided. We present the results of exclusion experiments in C. revoluta similar to those that have been attempted in studies of pollination in the Zamiaceae (e.g., Norstog et al., 1986
; Tang, 1987
; Connell and Ladd, 1993
; Donaldson et al., 1995
; Donaldson, 1997
; Terry, 2001
; Wilson, 2002
; Hall et al., 2004
; Terry et al., 2005
).
Cycas revoluta is native to southern Japan and has been considered wind-pollinated (Norstog and Nicholls, 1997
). Like Vorster (1995a
, b
), however, we have also doubted anemophily in C. revoluta, due to the presence of several features unfavorable for wind pollination. First, because the stem is usually 0.5–2 m tall with large pinnate leaves persisting for several years, air currents are often prevented, particularly in dense populations. Second, although the total period of pollination spans approximately 2 mo (May and June on Yonaguni Island, Okinawa, Japan), ovules of each female are receptive only for 2 wk at the longest (M. Kono, personal observation). In other words, the duration of pollination is not necessarily synchronized among individuals. Moreover, in natural habitats, pollination occurs during the rainy season, which is unfavorable for transporting pollen grains by wind. These conditions contrast with those enumerated by Whitehead (1983)
as ecological aspects favoring wind pollination. Additionally, recent studies have shown that the strong odor emitted by male and female cones at the time of pollination contains an estragole as the main volatile compound, which acts as both an attractant and a deterrent to insects (Azuma and Kono, 2006
).
We also present results of counting airborne pollen grains and determining whether insects were present on male and female cones. We identified and counted all captured insects and determined whether cycad pollen grains were attached to their bodies. Based on the results, we clarify whether Cycas revoluta is wind- or insect-pollinated. Because Cycas represents one of the major lineages in the Cycadales (Hill et al., 2003
), understanding the pollination mode of one of the species in detail will help clarify how pollination methods developed in one of the early offshoots of the seed plants. In addition, because C. revoluta is a major landscaping species that has been planted all over the world, the study here may have some importance for the landscaping industry.
MATERIALS AND METHODS
Study site and plants
This study was conducted in a ca. 50 000-m2 southern meadow on Yonaguni Island, Okinawa, Japan (122°56''–123°2'' E and 24°28'' N; Fig. 1A–C), where several thousand C. revoluta grow densely in some areas and rather sparsely in others (Fig. 1D). The meadow extends from east to west with a steep limestone cliff to the north and faces the ocean to the south; it is windy and open, with no tall trees. Although it is uncertain whether cycads are native or introduced to the island, they often bear male or female cones and set fertile seed on female cones, thus providing a suitable site to study the pollination system in C. revoluta.
|
The male cone consists of numerous microsporophylls, each bearing 30–600 microsporangia, and it withers after releasing pollen grains for about a week. On the other hand, each megasporophyll bears one to eight ovules at both lateral margins of its lower half (Fig. 1G), and thus a female plant bears several hundreds of ovules in the "cabbage-like" assemblage of the megasporophylls. At the time of pollination, ovules are ovoid and slightly flattened, 7–10 mm wide, yellowish green, and heavily tomentose except around the micropyle. During the pollination period, the female cones are open and emit odor, with each megasporophyll recurved to expose the ovules outward, and the ovules secrete a pollination drop; they close again after pollination. By fertilization in late August, i.e., about 2 or 3 mo after pollination, the ovules are c. 25–35 mm wide and are orange with a glabrous surface.
Exclusion experiments
To determine pollen vectors of C. revoluta, two kinds of treatments were applied to select female cones, and the rate of seed set in those cones was compared to that of 21 naturally pollinated or untreated female cones. The rate of seed set here indicates the proportion of mature seeds among all ovules in each female cone and was used as an index of pollination success. Mature seeds were recognized by their size at the end of October (about 4 mo after pollination) because, with the exception of Encephalartos and Macrozamia, ovules will abort if they are not pollinated (Norstog and Nicholls, 1997
).
First, to confirm that agamospermy does not occur during seed production in C. revoluta, we conducted experiments excluding both airborne pollen and insects using 28 female cones. To prevent any pollen vectors from reaching the ovules, the entire female cone was covered with a plastic bag and tied at the base with a plastic-covered wire (Fig. 2A). In addition, we covered 21 other female cones with plastic bags for artificial pollination. Artificial pollination (by lightly beating a male cone on the side of a female cone to flow pollen) was conducted one to three times during the pollination period after the plastic bags were removed. After artificial pollination, the cones were covered with plastic bags again. These female cones had 73% seed set on average, confirming that the plastic bags did not hinder the development of ovules in the cones.
|
) to reach the ovules but to keep insects out. To confirm that the mesh size was large enough to allow C. revoluta pollen grains to pass through, we poured pollen grains over another female cone covered with a mesh bag; indeed, 92% of the ovules reached maturity. To minimize the effects of the bags on cycad growth, we applied the two aforementioned treatments to the cones several days to a week before pollination began. The bags were kept for about 3–5 wk as they were. After the cones were pollinated and completely closed by shifting megasporophylls from a recurved to an incurved state, the bags were removed from the cones. During that period, for cones covered with plastic bags, we untied the plastic bags slightly at the base every 2 or 3 d to remove water that had accumulated at the bottom of the plastic bag due to transpiration. We checked both kinds of bags every 2 or 3 d to see if they were torn, blown off, or invaded by insects. We were unsuccessful in excluding insects from nine of the 28 female cones covered with plastic bags because several species of insects were able to enter through the knots in the bags. We excluded those nine cones from the statistical analysis, but because they had a significant rate of seed set, we used them to discuss the possibility of insect pollination.
We statistically tested the significance of differences in the rate of seed set among the naturally pollinated cones and the two types of exclusion experiments. Skewness in the data indicated that the distribution of mesh-covered samples was not normal (despite having been arcsine square-root transformed); therefore, we used the Mann-Whitney U-test (Wilcoxon two-sample test) for these nonparametric data.
Density of airborne pollen grains
Airborne pollen grains of Cycas revoluta were collected from 25 May to 18 June 2004 using Durham's sampling device (Durham, 1946
). Three sampling devices were set at a height of 1 m at three different distances (2, 10, and 60 m) from the nearest male cone. All sampling points were in the meadow, although two points (10 and 60 m) were adjacent to a dense copse of screw pine (Pandanus odoratissimus L. f.). A glass slide smeared with petroleum jelly to capture airborne pollen grains was placed horizontally on each device, and each slide was replaced daily with a fresh slide for 25 d.
An additional sampling device was placed at a height of 1 m at a distance of 50 cm from the nearest male cone from 7–18 June, which was almost at the end of the pollination season in the cycad population on Yonaguni Island. The distance of 50 cm corresponds to approximately half the length of a single leaf as well as to the distance maintained by two individuals of C. revoluta growing adjacent to each other. One male cone was present 50 cm away from the additional device.
Airborne pollen grains collected by the glass slides were stained with 1% safranin and counted under a light microscope (BX51, Olympus, Tokyo, Japan). Pollen grains were identified as C. revoluta based on their size and surface sculpturing (Dehgan and Dehgan, 1988
). The number of pollen grains per centimeter per day was calculated. The glass slides also captured pollen grains of grasses, such as Chrysopogon aciculatus, growing in the meadow. Their numbers were counted to confirm that the sampling devices successfully captured airborne pollen irrespective of the places they were set and thus were placed in suitable positions.
Capturing insects on male and female cones
We attempted to investigate insects visiting male and female cones by visual observation continuously for 48 h between 30 May and 1 June 2004, but were unable to find any insects except a few melon flies (Zeugodacus cucurbitae, Tephritidae, Diptera) that had not touched ovules. We thought that because the megasporophylls were densely arranged along the long pinnate vegetative leaves (ca. 1–2 m long), creating blind corners, we may have overlooked visiting insects, especially crawling species. However, removal of some vegetative leaves on one side to reduce blind corners resulted in a failure to produce fertile seeds on that side; therefore, this method was deemed inappropriate for investigating visiting insects.
Instead, in the evenings of 12 and 13 June, we tried to find female cones that were open and emitting odor with the ovules secreting pollination drop at night. We collected five such cones growing either near or far from male trees soon after covering the whole assemblage of megasporophylls with mesh bags, and subsequently with plastic bags, to secure the capture of insects. To kill insects, the plastic bag that covered the whole female cone was sealed with a wire at the base after cotton soaked with ethyl acetate was placed inside. One hour after bagging, we removed both bags. While removing the megasporophylls one by one (Fig. 4A), we collected all insects that were present. We also noted any signs of insect activity, such as bore-holes or excrement.
|
Insects found in male and/or female C. revoluta cones may be predators rather than pollinators. To determine how insects used the cycads, we observed their activities on 18 female cones from the end of October to the beginning of November 2004, about 1 mo after fertilization. Any damage on the megasporophylls was also recorded.
Pollen load on insects
Insects collected from male and female cones were individually covered with waxed paper, dried, and observed under a light microscope. If pollen grains were found, the whole insect was sputter-coated with gold (JFC-1200, JEOL, Tokyo, Japan) and observed with a scanning electron microscope (JSM-5800, JEOL) to confirm whether the pollen was from C. revoluta.
RESULTS
Seed set by natural pollination
The rate of seed set in the 21 naturally pollinated, or uncontrolled, female cones varied from 15% to 73%, with an average of 40.4% (Fig. 2C). Among the trees bearing these female cones, a tree growing in relative isolation (without any other individual of C. revoluta within 10 m radius) and a tree growing within the dense foliage of other cycads had 31.3% and 29.2% seed set, respectively. Mature seeds did not have an apparent preference for any position or direction within a female cone.
Seed set in exclusion experiments
Exclusion experiments using plastic bags in 19 female individuals (excluding nine individuals that allowed insect visits) resulted in a mean 3.1% seed set (Fig. 2D). Of these 19 individuals, 15 contained 1–28 mature seeds, reaching 9.1% seed set at the highest rate, while the remaining four had no mature seeds. In those seeds, we found traces of pollen tubes in the pollen chamber and confirmed that embryogenesis proceeded normally. This indicates that the seeds were produced by the ordinary process of pollination and fertilization rather than by agamospermy and that pollen grains had already been transported to the cones before being bagged.
On the other hand, of the nine female cones from which we failed to exclude insects, two were invaded by fruit flies, and yielded 2.9% and 5.7% seed set, respectively. The remaining seven were invaded by nitidulid beetles and yielded 14.7–33.5% seed set, largely in June during the late season of pollination.
Exclusion experiments using mesh bags on 29 female cones resulted in a 10.0% seed set on average (Fig. 2E). In all the female cones covered with mesh bags, we did not find any insect or any sign of insect invasion such as bore holes or frass. There were two peaks in the rate of seed set, one at a lower rate of seed set (mean 5%) observed in 24 of the 29 plants and the other at a higher rate of seed set (about 30%, with about 100 mature seeds) observed in five plants. The higher rate of seed set almost agreed with the rate of seed set by natural pollination. The precise rate of seed set in the five trees was 25.7, 30.3, 32.7, 39.1, and 43.0%. These trees (except one with 39.1% seed set) all grew within a distance of 1–2 m from male trees, whereas the other 24 trees were located more than 2 m away from male individuals. Thus, trees growing close to male trees produced seeds at the high rate. The tree that had 39.1% seed set may have also grown close to male trees, but this was not confirmed because the tree grew at the edge of a population that was so dense we were unable to examine it closely.
On the other hand, one may wonder how trees growing more than 2 m away from male trees set seeds at a lower rate and wonder whether some ovules were pollinated before being bagged or some insects may have entered and exited the mesh bags. However, a few mature seeds in those trees were often restricted to several adjoined megasporophylls. Such a restricted distribution was not observed in cones covered with plastic bags. Thus, in case of mesh-bagged cones, a few ovules may have received airborne pollen via strong wind.
The statistical analysis showed that, compared to natural pollination, the rate of seed set significantly decreased (P < 0.001) in the exclusion experiments with the two types of bag. Moreover, we found a significant difference (P < 0.05) in the rate of seed set between the plastic bags and mesh bags in the exclusion experiment .
Density of airborne pollen
The daily density of airborne cycad pollen was calculated at three different sampling points: 2, 10, and 60 m from the nearest male cone. Figure 3A shows the changes in density (number of pollen grains·cm–2·d–1) for about 1 mo, and the average density at 2, 10, and 60 m was 12.24, 2.64, and 1.20 (number of pollen grains·cm–2·d–1), respectively. Although the density fluctuated each day, the number of pollen grains clearly decreased with distance from male cones. The highest density was obtained at 2 m from the nearest male cone. The density of cycad pollen grains at 2 m always exceeded those at the two other sampling points, fluctuating from 0.24 to 41.04 pollen grains·cm–2·d–1. The density at 2 m was more than five and 10 times as high as those at 10 m and 60 m, respectively. In contrast, the difference in the density was not as conspicuous between 10 m and 60 m.
|
Figure 3B shows the density of airborne pollen grains collected at 50 cm from one male cone that shed pollen for 12 d between 7 and 18 June. We collected 1428 pollen grains·cm–2·d–1 at 50 cm, which was >100-fold higher than the concentration at 2 m (mean = 11.28 pollen grains·cm–2·d–1). More than 4800 pollen grains·cm–2·d–1 were collected on 9 June, when more pollen than usual was dispersed. On this day, a large amount of pollen had dropped to the ground near the male cone.
By combining the data used for Fig. 3A with those used for Fig. 3B, we obtained Fig. 3C, which shows that the amount of airborne pollen grains greatly decreased with distance from a male cone. Within 50 cm, the amount of airborne pollen grains was much higher than beyond 50 cm.
Invertebrates collected from male and female cones
A total of 41 individuals of 11 species was collected from four male cones (Table 1). All were obtained from the three male cones collected on 30 May, except for a spider that was the only animal found in the male cone collected on 18 June. Among the insects found in male cones, Ornebius kanetataki (Orthoptera) was the most abundant, followed by Derelomus bicarinatus (Coleoptera) and nymphs of Onychostylus notulatus (Blattodea).
|
There were several signs of insect activity on female trees during the pollination period. These included insect excrement on megasporophylls and sometimes a small hole that appeared to have been chewed by an insect into the micropyle (Fig. 4A, B). Such holes did not appear to prevent the ovule from further development (Fig. 4C), because ovules with holes reached maturity.
One month after fertilization, which occurred approximately in late August, more insects were observed in female cones, particularly nitidulids and Larinus weevils, which were observed frequently. Longicorns, cockroaches, pill bugs, leafhoppers, centipedes, spiders, scorpions, and lizards were also observed and appeared to use the cycads as shelter. Of 18 naturally pollinated female cones, 15 contained nitidulid colonies comprising hundreds of adults and larvae per colony. They appeared to consume several megasporophylls and ruined seeds borne on those megasporophylls, although most megasporophylls on the same tree were not damaged and set fertile seeds. In these 15 female cones, which had 46–116 megasporophylls per cone, an average of 3.8% of megasporophylls was damaged by insects; in the worst case, 10.5% (10 of 95 megasporophylls) were damaged. However, even the most damaged cone achieved 30.6% seed set, producing 146 mature seeds.
Pollen load on insects
While Carpophilus chalybeus collected from male cones was covered with Cycas revoluta pollen grains, those collected from female cones also had some pollen on their body (Fig. 5A). Scanning electron micrographs (SEMs) confirmed that pollen grains identified as C. revoluta (Fig. 5E) were attached almost uniformly to the head, thorax, and abdomen (Fig. 5B–D). The five nitidulids examined by SEM bore several dozen cycad pollen grains, but no pollen grains of other plants.
|
Is Cycas revoluta wind-pollinated?
Insect exclusion experiments with mesh bags using 29 female cones resulted in 10% seed set on average. This rate of seed set is low compared to the 40.4% seed set achieved by natural pollination. However, of these 29 female cones, the five that grew within a distance of 1–2 m from male trees showed about 30% seed set and produced about 100 mature seeds, while the remaining 24 that grew >2 m away from male cones had 5% (mean) seed set. Because the mesh itself may prevent pollen from passing through and/or change the aerodynamic conditions around a cone, the seed-set proportion under mesh-bagged conditions can be considered a minimal value (Tang, 1987
; Dafni, 1992
; Donaldson, 1997
). With regard to the five trees, because about 30% of their seeds matured without insects and only with the use of airborne pollen grains for pollination, there is no doubt that C. revoluta can be pollinated by wind (anemophilous). However, because the amount of airborne pollen decreased markedly beyond a 2-m radius from male cones, wind pollination is limited unless female cones are located near male cones.
In the entomophilous species Lepidozamia peroffskyana, a large amount of pollen is in the air at a distance of 50 cm from males, and the amount of pollen decreases and becomes more sporadic beyond 2 m (Hall et al., 2004
). In the anemophilous cycad Cycas panzhihuaensis, the amount of pollen decreases substantially in the air within 2.55 m from male trees (Wang et al., 1997
). Thus, taken together, these data indicate that large amounts of cycad pollen are generally not dispersed by wind over long distances. Wind seems effective for pollination only when both males and females grow densely in windy, open areas.
Seed set significantly decreased in the mesh-bagged treatment. Naturally pollinated female trees consistently had a high rate of seed set (40.4%), irrespective of their distance from male trees. These included female trees that were isolated from male trees, as well as those that grew in poorly ventilated sites beneath the foliage of adjacent individuals. These data indicate that C. revoluta is also pollinated by a vector other than wind.
Is Cycas revoluta insect-pollinated?
Among the insects collected from male and female cones, the nitidulid beetle Carpophilus chalybeus bore cycad pollen grains on its body and was common on female cones both near and far from male trees. In other words, nitidulids did not carry these cycad pollen grains to the female cones after the pollen had been transported by wind, but carried them directly from male cones. In fact, the mesh-bagged female cones set very few seeds, except for those within 2 m of male cones. Moreover, the nine plastic-bagged female cones that were invaded by nitidulids had a high rate (14.7–33.5%) of seed set, indicating that C. revoluta is also pollinated by nitidulids. Thus, C. revoluta shows both wind (anemophily) and insect pollination (entomophily), i.e., it is ambophilous, although wind pollination is restricted to female trees located within a 2-m radius of male trees.
Relationship between nitidulids and Cycas revoluta
We showed that female C. revoluta trees growing far from male trees are pollinated by nitidulids. This raises several questions, such as what attracts nitidulids to male and female cones, what is their reward, and why and/or how insects carry pollen grains to the ovule in the female cone. Nitidulids are attracted by fruity, spicy, or unpleasant odors, and/or by visual cues in tropical angiosperms such as the Annonaceae (Gottsberger, 1999
; Jürgens et al., 2000
; Tukada et al., 2005
), Arecaceae (Scariot et al., 1991
; Listabarth, 1996
; Ervik et al., 1999
; Henderson et al., 2000
; Consiglio and Bourne, 2001
; Voeks, 2002
; Núñez et al., 2005
), Araceae (García-Robledo et al., 2004
), and Proteaceae (Hemborg and Bond, 2005
). Because strong odors (with estragole as the main volatile compound) are emitted by both male and female C. revoluta cones (Azuma and Kono, 2006
), they are very likely to attract nitidulids during the pollination period.
Cycas revoluta pollen is probably one of the rewards for nitidulids although temporal variation appears to exist in their number because we did not observe any insects on male cones on 18 June but many (including one nitidulid) on 30 May. In angiosperms such as the Arecaceae (Ervik et al., 1999
; Henderson et al., 2000
; Consiglio and Bourne, 2001
; Nuñez et al., 2005) and the Araceae (Saibeh and Mansor, 1996
; García-Robledo et al., 2004
), nitidulids, after eating pollen and flower and inflorescence tissue, leave the flowers with pollen on their bodies. Nitidulids also formed colonies in female cones of C. revolute after the pollination period, and hundreds of adults and larvae were observed. Female cones may provide nitidulids food (e.g., megasporophylls and pollination drops secreted by the ovules), mating sites, brood/breeding sites, and/or heat. With regard to thermogenesis in female cones, Vorster (1995b
, p. 25) stated that "both male and female cones show a raise above ambient temperature at the time of pollination," but did not provide any evidence for it. Nitidulids consumed at most 10% of the megasporophylls of C. revoluta. Thus, male and female cones appear to provide ample food rewards to nitidulids.
With regard to the transportation of pollen grains to the micropyle, pollination drops attract insects to the micropyle in the Zamiaceae (Tang, 1987
; Donaldson, 1997
), and this may also be the case in C. revoluta. The small holes (Fig. 5B, C) that appeared to have been chewed by insects near the micropyle suggest another possibility for an attractant. In C. revoluta, the surface of each megasporophyll is densely covered with hard trichomes, except in the apical part of the ovule, i.e., around the micropyle. Insects may perceive such a naked part as edible, thus attracting them to the micropyle. After having attempted to eat tissue of the apical part of the ovule without success and leaving, an insect may leave some pollen grains near the micropyle along with the small hole caused by chewing.
In summary, several different insect species (11 on male cones and seven on female cones) visit cones of C. revoluta, probably attracted by odor. Among them, the nitidulid Carpophilus chalybeus, which may be attracted by pollination drops and/or the naked apical parts of the ovule, carries the pollen grains to the micropyle. Colonies of this species become established in female cones, rather than in male cones, as in most Zamiaceae (Norstog et al., 1986
; Tang, 1987
; Norstog and Fawcett, 1989
; Donaldson, 1997
; Hall et al., 2004
), suggesting that this nitidulid is a pollinator as well as a predator of female cones of C. revoluta. The insect may still be in early stages of pollination (Pellmyr, 2002
), where initially it started as a predator on female cones and now is providing a pollination service.
Ambophily in C. revoluta appears to be in marked contrast to pollination by host-specific insects in the Zamiaceae. It may occur in more species of Cycas, particularly in species thought to be insect-pollinated (e.g., C. media, Ornduff, 1991
; C. rumphii, C. thouarsii, Vorster, 1995a
; C. pectinata, Lindstrom, 1999
; several Cycas spp. in Thailand, Vietnam, and China, Tang et al., 1999
). We should carefully investigate these and other cycad species to understand how they are pollinated; this will lead to a better understanding of the functional aspects of ambophily in cycad evolution.
FOOTNOTES
1 The authors are grateful to M. Kato, H. Kojima, and M. Muramatsu for their assistance in identifying insects collected from cycad cones, and to R. Oberprieler for his suggestion on insects visiting cycads. This study was supported by a Grant-in-Aid for Scientific Research from the Japan Society for the Promotion of Science (no. 18370036) and a grant for Biodiversity Research of the 21st Century Center of Excellence (COE, A14). 
2 Author for correspondence (tobe{at}sys.bot.kyoto-u.ac.jp
)
LITERATURE CITED
Azuma H. Kono M.. 2006. Estragole (4-allylanisole) is the primary compound in volatiles emitted from the male and female cones of Cycas revoluta. Journal of Plant Research 119: 671-676.[CrossRef][ISI][Medline]
Chamberlain C. J.. 1935. Gymnosperms: structure and evolution University of Chicago Press, Chicago, Illinois, USA.
Connell S. W. Ladd P. G.. 1993. Pollination biology of Macrozamia riedlei—The role of insects. In D. W. Stevenson and K. J. Norstog [eds.], Proceedings of the Second International Conference on Cycad biology, Townsville, Queensland, Australia, 1990 96-102 Palm and Cycad Societies of Australia, Milton, Queensland, Australia.
Consiglio T. K. Bourne G. R.. 2001. Pollination and breeding system of a neotropical palm Astrocaryum vulgare in Guyana: a test of the predictability of syndromes. Journal of Tropical Ecology 17: 577-592.[CrossRef][ISI]
Dafni A.. 1992. Pollination ecology Oxford University Press, New York, New York, USA.
Dehgan B. Dehgan N. B.. 1988. Comparative pollen morphology and taxonomic affinities in Cycadales. American Journal of Botany 75: 1501-1516.[CrossRef][ISI]
Donaldson J. S.. 1997. Is there a floral parasite mutualism in cycad pollination? The pollination biology of Enchephalartos villosus (Zamiaceae). American Journal of Botany 84: 1398-1406.[Abstract]
Donaldson J. S. Nänni I. Bösenberg J. D.. 1995. The role of insects in the pollination of Encephalartos cycadifolius. In P. Voreter [ed.], Proceedings of the Third International Conference on Cycad biology, Pretoria, South Africa, 1993 423-434 Cycad Society of South Africa, Stellenbosch, South Africa.
Durham O. C.. 1946. The volumetric incidence of atmospheric allergens. Journal of Allergy 17: 79-86.[CrossRef][ISI]
Ervik F. Tollsten L. Knudsen J. T.. 1999. Floral scent chemistry and pollination ecology in phytelephantoid palms (Arecaceae). Plant Systematics and Evolution 217: 279-297.[CrossRef][ISI]
Forster P. I. Machin P. J. Mound L. Wilson G. W.. 1994. Insects associated with reproductive structures of cycads in Queensland and northeast New South Wales, Australia. Biotropica 26: 217-222.[CrossRef][ISI]
García-Robledo C. Kattan G. Murcia C. Quintero-Marín P.. 2004. Beetle pollination and fruit predation of Xanthosoma daguense (Araceae) in an Andean cloud forest in Colombia. Journal of Tropical Ecology 20: 459-469.[CrossRef][ISI]
Gottsberger G.. 1999. Pollination and evolution in neotropical Annonaceae. Plant Species Biology 14: 143-152.[CrossRef]
Hall J. A. Walter G. H. Bergstrom D. M. Machin P.. 2004. Pollination ecology of the Australian cycad Lepidozamia peroffskyana (Zamiaceae). Australian Journal of Botany 52: 333-343.[CrossRef][ISI]
Hemborg Å. M. Bond W. J.. 2005. Different rewards in female and male flowers can explain the evolution of sexual dimorphism in plants. Biological Journal of the Linnaean Society 85: 97-109.[CrossRef]
Henderson A. Pardini R. Robello J. F. Vanin S. Almeida D.. 2000. Pollination of Bactris (Palmae) in an Amazon forest. Brittonia 52: 160-171.[CrossRef][ISI]
Hill K. D. Chase M. W. Stevenson D. W. Hills H. G. Schutzman B.. 2003. The families and genera of cycads: a molecular phylogenetic analysis of Cycadophyta based on nuclear and plastid DNA sequences. International Journal of Plant Sciences 164: 933-948.[CrossRef][ISI]
Hill K. D. Stevenson D. W. Osborne R.. 2004. The world list of Cycads. In T. Walters and R. Osborne [eds.], Cycad classification 219-235 CABI publishing, Wallingford, UK.
Jones D. L.. 2002. Cycads of the world: ancient plants in today's landscape, 2nd ed Smithsonian Institution Press, Washington, D.C., USA.
Jürgens A. Webber A. C. Gottsberger G.. 2000. Floral scent compounds of Amazonian Annonaceae species pollinated by small beetles and thrips. Phytochemistry 55: 551-558.[CrossRef][ISI][Medline]
Kato M.. 2001. Insect fauna associated with Cycas revoluta (Cycadaceae), with a discovery of a cerambycid megasporophyll miner. Special publication of the Japan Coleopterological Society, Osaka 1: 73-78.
Keppel G.. 2001. Notes on the natural history of Cycas seemannii (Cycadaceae). South Pacific Journal of Natural Science 19: 35-41.
Lindstrom A.. 1999. Possible pollinator of Cycas pectinata. In C. J. Chen [ed.], Biology and conservation of cycads—Proceedings of the Fourth International Conference on Cycad biology, Panzhihua, Sichuan, China, 1996 306-308 International Academic Publishers, Beijing, China.
Listabarth C.. 1996. Pollination of Bactris by Phyllotrox and Epurea. Implication of the palm breeding beetles on pollination at the community level. Biotropica 28: 69-81.[CrossRef][ISI]
Mound L. A. Terry I.. 2001. Thrips pollination of the central Australian cycad, Macrozamia macdonnellii (Cycadales). International Journal of Plant Sciences 162: 147-154.[CrossRef][ISI]
Niklas K. J. Norstog K.. 1984. Aerodynamics and pollen grain depositional patterns on cycad megastrobili: implications on the reproduction of three cycad genera (Cycas, Dioon, and Zamia). Botanical Gazette 145: 92-104.
Norstog K. J. Fawcett P. K. S.. 1989. Insect-cycad symbiosis and its relation to the pollination of Zamia furfuracea (Zamiaceae) by Rhopalotria mollis (Curculionidae). American Journal of Botany 76: 1380-1394.[CrossRef][ISI]
Norstog K. J. Nicholls T. J.. 1997. The biology of the cycads Cornell University Press, Ithaca, New York, USA.
Norstog K. J. Stevenson D. W. Niklas K. J.. 1986. The role of beetles in the pollination of Zamia furfuracea L. fil. (Zamiaceae). Biotropica 18: 300-306.[CrossRef][ISI]
Núñez L. A. Bernal R. Knudsen J. T.. 2005. Diurnal palm pollination by mystropine beetles: is it weather-related?. Plant Systematics and Evolution 254: 149-171.[CrossRef][ISI]
Ornduff R.. 1991. Size classes, reproductive behavior, and insect associates of Cycas media (Cycadaceae) in Australia. Botanical Gazette 152: 203-207.
Pellmyr O.. 2002. Pollination by animals. In C. M. Herrera and O. Pellmyr [eds.], Plant–animal interactions 157-184 Blackwell Science, Malden, Massachusetts, USA.
Saibeh K. Mansor M.. 1996. Differential temporal patterns of insect populations visiting Cryptocoryne ciliate. Malaysian Applied Biology 25: 95-100.
Scariot A. O. Lleras E. Hay J. D.. 1991. Reproductive biology of the palm Acrocomia aculeate in central Brazil. Biotropica 23: 12-22.[CrossRef][ISI]
Tang W.. 1987. Insect pollination in the cycad Zamia pumila (Zamiaceae). American Journal of Botany 74: 90-99.[CrossRef][ISI]
Tang W. Oberprieler R. G. Yang S.. 1999. Beetles (Coleoptera) in cones of Asian Cycas: diversity, evolutionary patterns, and implications for Cycas taxonomy. In C. J. Chen [ed.], Biology and conservation of cycads, Proceedings of the Fourth International Conference on Cycad biology, Panzhihua, Sichuan, China, 1996 280-297 International Academic Publishers, Beijing, China.
Terry I.. 2001. Thrips and weevils as dual, specialist pollinators of the Australian cycad Macrozamia communis (Zamiaceae). International Journal of Plant Sciences 162: 1293-1305.[CrossRef][ISI]
Terry I. Walter G. H. Donaldson J. S. Snow E. Forster P. I. Machin P. J.. 2005. Pollination of Australian Macrozamia cycads (Zamiaceae): effectiveness and behavior of specialist vectors in a dependent mutualism. American Journal of Botany 92: 931-940.
Tukada M. Higuchi H. Furukawa T. Yoshida A.. 2005. Flower visitors to cherimoya, Annona cherimola (Magnoliales: Annonaceae) in Japan. Applied Entomology and Zoology 40: 317-324.[CrossRef][ISI]
Voeks R. A.. 2002. Reproductive ecology of the piassava palm (Attalea funifera) of Bahia, Brazil. Journal of Tropical Ecology 18: 121-136.[ISI]
Vorster P.. 1995a. Aspects of the reproduction of cycads. 1. Pollinating mechanisms and the association of Amorphocerus (Curculionidae) with Encephalartos. In P. Voreter [ed.], Proceedings of the Third International Conference on Cycad biology, Pretoria, South Africa, 1993 367-378 Cycad Society of South Africa, Stellenbosch, South Africa.
Vorster P.. 1995b. Comments on Cycas revoluta. Encephalartos 42: 25-26.
Wang Q. Li C.-L. Yang S.-Y. Huang R. Chen F.-L.. 1997. Pollination biology of Cycas panzhihuaensis. L. Zhou et and S. Y. Yang [eds.], Acta Botanica Sinica 39: 156-163 (in Chinese with English abstract).
Whitehead D. R.. 1983. Wind pollination: some ecological and evolutionary perspectives. In L. Real [ed.], Pollination biology 97-108 Academic Press, Orlando, Florida, USA.
Whitelock L. M.. 2002. The cycads Timber Press, Portland, Oregon, USA.
Wilson G. W.. 2002. Insect pollination in the cycad genus Bowenia Hook. ex Hook. f. (Stangeriaceae). Biotropica 34: 438-441.[ISI]
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
