biologytitle.gif (3277 bytes)



1New York Botanical Garden, Bronx, NY, USA; 2Fairchild Tropical Garden, Miami, FL, USA

Two species of cycad in Florida and one in Mexico are known to be pollinated by beetles. The native Florida Zamia pumila is pollinated by a snout weevil, Rhopalotria slossoni and a clavicorn beetle, Pharaxonotha zamiae. An introduced cycad, Z. furfuracea, is in Florida pollinated by the snout weevil, R. mollis, apparently introduced from Mexico along with its host. Observations and experiments show that both zamias are dependent upon pollinating beetles for their reproduction and, in turn, the pollinating insects are dependent upon the cycads for such "rewards" as brood places, food, and shelter. Dioon caljfanoi, like Z. pumila, is pollinated by a weevil, Parallocorynus bicolor and a species of Pharaxonotha. Aspects of the pollination syndrome in D. califanoi are not completely worked out but are probably much like those in Zamia. The importance of these coevolutionary relationships for the ecology and conservation of these cycad species is unquestioned. Neither the insects nor the cycads are separately capable of long-term survival in nature, and unless both are protected and conserved, wild populations of Zamia and Dioon and no doubt many other cycads are in jeopardy.


Although cycads have only recently proven to be insect pollinated, evidence is accumulating that many cycads have symbiotic associations with host-specific insects, mostly beetles. In some way, not completely understood, such insects are attracted to cycad cones, sometimes by fragrances, sometimes for food, shelter, a breeding place, and probably combinations of some or all of these. Oberprieler (1989) remarked that the intimate association of cycads and insects, in this instance weevils, was first noticed and reported by Thunberg in 1773. In addition to pollination interactions, seed cones and seeds of cycads often are preyed upon by insects, including lepidopterans that are capable of damaging or destroying them should the infestation be heavy, as well as by birds, monkeys, etc. which may act as agents of seed dispersal. Involved in all of these plant-animal interactions are the cycad toxins in complex symbiotic relationships only now becoming understood.

C.J. Chamberlain, a widely recognised authority on cycad phenomena (Chamberlain, 1919, 1935) wrote in 1935, "there have been many reports of insect pollination, but in a rather extensive field study in which all the genera have been examined, nothing has been observed which would indicate anything but wind pollination so characteristic of the whole group of gymnosperms." Despite Chamberlain's rather authoritarian opinion to the contrary, considerable evidence for insect pollination of cycads had existed since 1893 (Pearson, 1906; Rattray, 1913) and no experimental data had been obtained for either wind or insect pollination.

At about the time Chamberlain and his students were exploring the morphology and anatomy of the Mexican cycads in the early 1900's, nearly all of the pieces of the cycad pollination puzzle had been examined and fitted together into a credible explanation for pollination by insects. Pearson (1906), observing beetles in cones of Encephalartos villosus, suspected they might in some way be involved in pollination of that species. Pearson wrote "it may be noted that the habitat of E. villosus and the position of its cones must, in most cases, render wind-pollination impossible, for the bush is usually so dense that the air a few feet above the ground can rarely move rapidly enough to carry even so light a substance as pollen." He credited Joseph Hooker as first suggesting that cycads must be entomophilous, and goes on to say that male cones of E. villosus give off a strong odour at the time pollen is released and that such cones swarm with weevils, which are covered with pollen and are found only on cycads. Pearson did not observe weevils actually transporting pollen to female cones and suggested only that it was probable.

Pearson's observations were subsequently enlarged upon by Rattray (1913) who worked out many of the remarkable interactions between weevils and pollen transport in E. villosus and E. altenstein ii. A weevil, probably the Cossonine Porthetes zamiae found in and around Durban by R.A. Crowson (pers. comm. to MN), was found in great numbers in male cones of E. altensteinii at the time pollen is shed in April. At this time the cones emitted a faint but distinct odour and visiting weevils, mating upon the cones and feeding on pollen, become so covered with pollen that much of it adhered to their bodies and remained there for 60 hours or so.

While Rattray failed to present experimental evidence showing that beetles were partially or exclusively responsible for the pollination of the cycad, his comments are worth quoting because they so accurately reflect present thinking on the subject. "The whole life-history [of the weevil] appears so dependent on Encephalartos that it would be only reasonable to find that the cycad had demanded a quid pro quo and had turned the visits to use."

Rattray also called attention to the major defect in the concept that cycads are wind- pollinated: "Although I have examined over a score of female cones during the time pollen is shed and have had three under daily observation in my own garden I have never noted any separation of the sporophylls. Although I still feel bound to believe that it must occur, yet it must be only to a very limited extent — a condition which, while not altogether excluding anemophily in this species, certainly renders it much less probable; for the exposed parts of the sporophylls stand almost horizontally and do not overlap, so that unless the separation were very marked only wind-borne pollen coming horizontally could effect entrance, and only those ovules on the side of the prevailing winds would be fertilised (see Fig. 1). 

Fig. 1. Pollen-receptive female cone of Zamia
. Receptivity is indicated by a
vertical separation of two adjacent rows of
microsporophylls giving the only access to the
interior. Bar = 5 mm. 

(Click on figure to enlarge)

But in all these cones which I have worked through, even from plants growing on the outside of a group, embryos had been formed in uniform proportions on all sides." One can add here that cycad pollen tends to clump and is not readily airborne.

Later investigators also implicated insects in some aspects of pollination; Marloth (1914) noted insects in cones of Encephalartos and Baird (1939) considered that insects might transport pollen within cones of Macrozamia, but thought that inter-plant pollen transport required wind. More recently, Chadwick (1993) and Ornduff (1993) presented circumstantial evidence indicating that Macrozamia communis and M. riedlei, respectively, are weevil pollinated. Connell and Ladd (1993) considered weevil pollination in M. riedlei to be a possibility but were unable to produce statistically significant evidence either pro or con. Other botanists, however, contended that cycads were wind-pollinated, for example Lawson (1926), found nothing to suggest that Bowenia required any agent other than wind for its pollination but remarked that pollination in some cases occurred between plants separated by "a great distance." More recently, Wilson (1993) questioned the importance of wind in the pollination of Bowenia and emphasised the role of a weevil, Tranes subopaca in the pollination of B. serrulata.

In light of what now is known of the highly specialised relationships between several species of weevils as obligate pollinators of genera and species of cycads, and indirect evidence that others are similarly pollinated, it is difficult to understand why insect pollination was neglected for such a long time after Rattray's remarkable early work. The answer most likely lies in the absence of experimental procedures designed to test the hypotheses of insect or wind pollination and the purely anecdotal evidence that was then available.

Because of the long controversial history of non-conclusive evidence, Norstog and Stevenson (1980) and Norstog et al. (1986) designed experiments and established criteria to test both the wind and insect hypotheses of cycad pollination biology. Thus exclusion experiments were designed to test for seed set. The following conditions were established: (1) exclusion of all pollen entry by plastic bagging of prepollination ovulate cones; (2) exclusion of insects by net bagging with a mesh size excluding insects but allowing pollen entry by wind; and (3) exclusion of wind by covering prepollination ovulate cones with inverted cylindrical beakers so that insects could gain access to the ovulate strobilus. Other criteria that were used to determinate if insect pollination was supported included direct observations of potential pollinators and maturing strobili, whether or not the potential pollinators visited other plant genera, the life cycle of pollinators and the pollinated, and the direct observation of pollen being carried from microsporangiate strobili to the micropyles of the ovules of ovulate strobili. Similar exclusion experiments have also been conducted on two species of Encephalartos by Donaldson et al. (1995) and Donaldson (1997).

Pollination of Zamia furfuracea

Zamia furfuracea, a Mexican cycad grown at Fairchild Tropical Garden in scattered plantings, is perfectly fertile (i.e., produces viable seed) in cultivation. It was reported by Norstog and Stevenson (1980) to attract swarms of tiny weevils at the time male cones were maturing. The weevils breed and feed upon male cones and, later, adults feeding and sheltering amongst microsporophylls, leave the cones (newly metamorphosed adults chewing exit holes, Figs. 2, 3A-G) and fly to other Z. furfuracea cones, including nearby female cones (Fig. 3K). It is during such visitations that pollen is thought to be transferred to ovules. 

Fig. 2. Rhopalotria mollis, a pollinator of
Zamia furfuracea. A. Female pupa; B.
Adult female, ventral view; C. Adult female,
dorsal view; D. Male pupa; E. Adult male,
ventral view; F. Adult male, dorsal view. 
Bar = 1 mm. 

(Click on figure to enlarge.)

Though circumstantial, the authors considered their observations to be strong evidence that Z. furfuracea is pollinated by a specific insect. Also, citing observations that female cones of the native cycad Z. integrifolia were fertile despite being buried beneath surface litter, they thought that this species too was insect-pollinated. This set the stage for the implementation of experimental criteria as outlined above.

In the next five years, Norstog et al. (1986) succeeded in demonstrating by exclusion experiments, in which either wind or insects were prevented from carrying pollen into female cones, that this weevil, Rhopalotria mollis, probably imported inadvertently from Mexico, was the faithful pollinator of Z. furfuracea.

The female cone is tightly closed during most of its development, only opening by a narrow crack when it is pollen-receptive (Fig. 1). This crack forms by separation of megasporophylls of adjacent vertical orthostichies and along the phyllotaxic spiral in the apical region of the ovulate strobilus. The crack remains open for several days, then closes completely for the duration of cone maturation. If one opens such closed cones, one often finds live weevils that have been trapped inside. A similar action is described by Chadwick (1993) in female cones of Macrozamia communis, in which observations suggest a weevil, Tranes lyterioides may be a pollinator. This weevil completes its entire life cycle in male cones of M. communis in much the same way as R. mollis does in Z. furfuracea.

Fig. 3. The reproductive cycle of Rhopalotria
in Zamia furfuracea. A. Male cone with
swarming weevils; B. Weevils feed, breed, and
lay eggs in microsporophylls; C. Eggs are laid
in microsporophyll ends, and adults feed at the 
bases of sporphylls; D. Several larvae feeding
in and excavating a micosporophyll. When 
larvae meet, one eats the other. Commonly one
larva survives in a microsporophyll. E. Larva in
pupa case; F. Pupa; G. Emerging adult -- from
egg to adult usually encompasses about 7-9 days.
H.-J. Toward the end of the breeding season,
some larvae construct thick-walled pupa cases
and enter diapause, remaining dormant until the
following year, or longer. K. Pollen-receptive
female cone and visitation by weevils. 

(Click on figure to enlarge.)

The weevils that start the annual breeding cycle in cones of Z. furfuracea emerge from long-dormant larvae that "over-winter" in debris at the base of the cycads they later infest, and they put in an appearance at about the time the male cones approach maturity. The first breeding cycle of the weevil is short, a week or so, and involves just a few individuals, but increasingly large numbers of progeny in succeeding generations soon follow.

Insect attractants
Weevils are attracted to male cones of Z. furfuracea at about the time microsporangia are mature and just before they are ready to dehisce (Fig. 3A). At about this time also, both male and female cones become aromatic and warm, and a few of the sporophylls begin to loosen and separate. The heat and odours result from developmental and metabolic processes that are of wide occurrence among cycads and function both to cause male and female cones to open and to volatilise fragrances (Tang, 1987b; Tang, 1993; Tang et al., 1987). The male cone scent of Z. furfuracea was identified by Pelimyr et al. (1991) as consisting of two major components: 1,3-octadiene and linalool. Linalool is known to be an attractant of euglossine bees in orchid flowers. These would appear to have a pheromonal function. Both Pearson (1906) and Rattray (1913) remarked on the odours released by male Encephalartos cones at the time of pollination and one suspects these also have a pheromonal function.

Cycad toxicity and pollination by insects
At the time of pollination of both Z. integrifolia and Z. fuifuracea, the microsporophyll parenchyma is rich in starch, whereas megasporophyll tissue is poor in starch (Norstog and Fawcett, 1989). In addition, it can be seen that a type of storage cell (idioblast) appears intact in microsporophylls but modified in content in megasporophylls. Norstog and Fawcett proposed that idioblasts of both species contain sequestered toxins in male cone tissues, but not in female cone tissues in which toxins are liberated and dispersed at or prior to pollination. The evidence for this scheme is admittedly circumstantial. Idioblasts in cycad tissues are readily stained with ninhydrin and give the same color reaction as does the ninhydrin-positive neurotoxin BMAA, which was first discovered by Vega and Bell (1967) in tissues of Cycas circinalis.

At the time of pollination, female-cone idioblasts do not stain with ninhydrin, suggesting either degradation or mobilisation of toxin(s) (BMAA?). Other differences in male- and female-cone idioblasts also are apparent at the time of pollination; specifically, male-cone idioblasts are unstained with PAS (periodic-acid Schiff reagent, a general stain for reducing sugars) but those of female cones are PAS positive. This may only indicate an altered structural status for such idioblasts in male vs. female cones. However, another class of toxins, the MAM-glycosides (cycasin, neocycasin, macrozamin), all of which are highly toxic to animal tissues (Whiting, 1963), appear to be PAS positive. This brings up the possibility that female cones of at least some cycads have evolved a differential deployment of these toxins in their idioblasts as a defense against predators. Because male cones must be attractive to certain predatory insects, specifically pollinators, they may have evolved in the direction of lower toxicity levels (or possibly temporarily lowered concentrations of toxins). Female cones, on the other hand, while offering certain cues to attract visitation by pollinators (mimicry of male cone morphology and odours) may employ toxins to prevent seed predation.

Breeding and feeding site selection by weevils
Swarming weevils exhibit cone preferences, selecting only one or two male cones from among as many as 15-20, and remaining upon the chosen cone 1-3 days before abandoning it (Fig. 3A). Male cones of Z. furfuracea become mature sequentially, thus spreading out the period of pollen production for at least 4-6 weeks, while female cones, which are much fewer per plant, become mature more-or-less synchronously and all are pollen-receptive during a 2-3 week interval (Fig. 1). This obviates the difficulty of synchronising cone production that might be encountered were cones of both sexes to be fertile only during similarly short periods of time. A corollary observation is that a dozen or more small pollen cones ripening sequentially constitute evidence that insect pollination is the mode of reproduction in the species. Even so, male and female cone maturation must be synchronised, if only loosely, for pollination to succeed.

Adult male plants of Zamia furfuracea produce up to 20 or so male cones during the reproductive season. These mature in sequence to the point that of several cones visible on a plant, only 1-3 will be ready to shed pollen, a stage indicated by cone elongation, a slight separation of the microsporophylls, and an emanation of a musty fragrance.

Rhopalotria mollis individuals do not collect indiscriminately on male cycad cones, but are attracted only to those cones that are near to dehiscence and pollen release. The weevils swarm on to such cones (Fig. 3A) a day before the sporangia dehisce, and there is constant activity during which weevils probe for food, fight, mate and lay eggs in recently excavated feeding holes. Although the microsporophylls have not become loosely separated, they have begun to separate at the base of the cone. Sometimes the weevils probe into feeding holes which have already been made by other weevils, and sometime they start their own holes. When they start a hole of their own, they do not eat the stiff epidermal trichomes that cover the sporophyll endings, but they pull them out in tufts with their mandibles, which are clawed like hands. Trichomes of this cycad contain idioblasts and, presumably, also contain toxins, and this behaviour apparently constitutes toxin avoidance. When a female is going to lay an egg, she feeds very deeply, and her snout goes down into the sporophyll all the way to her eyes (Figs. 2C, 3B).

After the females have laid their eggs, both sexes go inside the cone between the separated sporophylls at the base of the cone. Weevil adults and larvae feed on male sporophyll tissue and ingest idioblasts with sequestered BMAA together with other parenchyma but the weevils and larvae do not digest the idioblasts. Rather, they accumulate and excrete them in their feces. The fecal pellets are used by Rhopalotria larvae to construct pupa cases (cocoons) and analysis of these cases reveals that they have a comparatively high content of the cycad neurotoxin, BMAA (Mark Duncan, pers. commun.). The toxins in the pupa cases give added protection to the larvae and pupae within them. While these actions are not direct evidence of the existence of specific toxins in idioblasts, they do constitute strong circumstantial evidence supporting this assumption.

Larvae of R. mollis are legless grubs which hatch within a day or so from eggs laid principally in microsporophyll tips and are quite active (Fig. 3C, D). As many as six eggs can be laid in the tip of a single microsporophyll by various individual females, although 2-4 eggs are more usual, and a corresponding number of larvae can be found in each microsporophyll where they feed voraciously (Norstog and Fawcett, 1989). When, inevitably, feeding larvae meet, one eats the other and eventually only one large and well-fed individual remains. With the exception of this cannibalism, larvae feed exclusively on parenchyma, as indicated by staining reactions of sections of the larval gut. They do not feed at all on pollen, either at this or any other stage of their cycle.

About 3-4 days after hatching, a larva constructs a chamber in a microsporophyll stalk within which it pupates (Fig. 3E, F), its head oriented toward the outer end of the microsporophyll. The pupa case (cocoon) of fecal matter held together with chitin secreted by the larva hardens into a tough, cylindrical capsule. In 2-3 days the adult emerges after metamorphosis, usually boring through the tip of the sporophyll (Fig. 3G). Except toward the end of the season, the entire process from egg to adult takes 7-9 days, and it is possible that during the season a half-dozen or so generations may result from a single initial mating. One more larval instar occurs before pupation. Then, after their metamorphosis, adult weevils escape from pupa cases by chewing through the end of the cases and boring distinctive exit holes in the ends of the sporophylls (Fig. 3G).

Rewards to pollinators
Rhopalotria mollis
is an excellent flyer and when males and females are attracted to male cones, as related above, they mate and probe for food and the females lay eggs, but whether the emitted volatiles that seem to attract them may be considered as "rewards" is conjectural. A brood place and food thus constitute the chief "reward" the plant gives the weevil for the "service" of pollination.

Pollination by weevils
At the time of pollination some megasporophylls of female cones become separated slightly and give access to the interior via a longitudinal crack (Fig. 1). This is accompanied by generation of a considerable amount of heat as well as aroma (Tang et al., 1987) and the cones become attractive to weevils. Throughout the season some weevils, perhaps in error, visit female cones where they attempt to feed, but apparently find the tissues to be unpalatable. The female sporophylls only become separated for a few days when the ovules are receptive, as indicated by pollination droplets on the micropyles. We find a few weevils in such cones at this time though most escape.

After pupation, newly emerged adults re-enter male cones where they shelter and feed--a single male cone at this stage may contain a hundred or more weevils. In moving around in the cone they become covered with pollen and it is at this stage that visitation to female cones occurs (Fig. 3K). However, female cones are not fed upon, though interior surfaces sometimes exhibit minute scars, "nibble marks," which are probably made by weevils. The reason for the unpalatability of female cone tissue may be the release of toxin by idioblasts present in both male and female cone tissues. Idioblasts in female cones break down just prior to their pollination and as well as having eroded idioblasts, female cones contain very little starch (Norstog and Fawcett, 1989; Vovides et al., 1993). Weevils that enter female cones do not feed on female cone tissue. Either they are repelled by some noxious compound (toxin), or the fact that no starch is available, or both, lead to this absence of feeding. In any case, pollen is transported to the cone and at least some of it is deposited upon the micropyles of ovules. As will be noted later, pollination of this cycad by weevils is very efficient and cone fertility may approach 100%.

More recently, in an interesting series of experiments and observations, Donaldson et al. (1995) employed both screens and insecticides to exclude the presumed pollinator, an undescribed Langurid beetle, from female cones of Encephalartos cycadifolius in habitat in Cape Province, South Africa. The percentage of fertile seed set was drastically reduced from 95% to 30% when the weevils were excluded. One outstanding observation resulted from their use of fluorescent dye to coat pollen-carrying beetles. It was possible to show that beetles, after entering cones, deposited the dye on the micropyles of a large number of ovules. It is the first conclusive evidence for the actual deposition of pollen on cycad ovules by specific insect pollinators.

Dormancy of weevils
Although all the Rhopalotria larvae in a male cone construct pupa cases, not all larvae metamorphose directly into pupae and adults. Towards the end of the season, there is a change in the tissue at the ends of the sporophylls which becomes green and tends to be avoided by the weevils. Near the end of the coning season of the cycad, the feeding behaviour of the weevil larvae is, therefore, a bit different from that occurring earlier in the season. Larvae tend to feed nearer the base of the microsporophylls (Fig. 3H, I). Such late-instar weevil larvae develop thick, opaque cuticles, become fat and waxy, and build very thick-walled pupa cases in which they go into diapause until the next season (Fig. 3C). These constitute the breeding stock for the following year.

Diapause, which is a type of dormancy consisting of an arrested state of development, occurs in Rhopalotria and can last for 2 or more years in this species of weevil (Norstog and Fawcett, unpubl. observ.) This, no doubt, accounts for the long-term efficiency of the relationship between the cycad and its pollinator and its continuity even over several interrupted breeding seasons.

We wondered what precipitated the change in behaviour of the weevils at the end of the season, and whether it was the lower starch content of the axis, or exposure to substances in the green tissue at the ends of the sporophyll, that would account for the initiation of diapause. Because fewer cones are maturing at the end of the season, those remaining continue to attract weevils longer than is the case earlier in the season. It seems to us now that the most likely explanation for the onset of diapause is that, due to an increase in weevil numbers in late-season male cones and resulting consumption of all available food, competition among larvae in a microsporophyll becomes fierce. When looking at these late cones, we found larvae which were almost ready to pupate coming up from the base, and down from the end of the sporophyll, as well as younger larvae of all ages, and eggs (Fig. 3I). Often very little edible tissue was left but cannibalism continued apace. We think that the most likely cause of diapause is an increase of protein in the larval diet resulting from increased cannibalism. The greater thickness of the cuticles of diapausic larvae then may be due to increased levels of dietary proteins.

At the end of the season, we collected some late male cones of Zamia furfuracea and put 20 in one plastic box and 12 in another, then sealed the boxes with paraffin wax and masking tape. A few larvae metamorphosed at the end of December and in January, when we were working on the cones of Zamia pumila. By March 17th, 11 had metamorphosed in the box with 20 cones, and seven in the box with 12 cones, but a massive metamorphosis took place at the end of March and the beginning of April, starting about a week after a colleague, Andrew Vovides, came to work at Fairchild Tropical Garden, bringing with him cones of Dioon edule, D. califanoi, Ceratozamia mexicana, and Zomia furfuracea collected in Mexico in heavy plastic bags, pickled in 70% ethanol. The smell of the cycad cones and pickling agents was very strong all over the laboratory and persisted for several weeks and we think emergence from diapause in this instance probably had a pheromonal stimulus. Because live larvae in diapause continued to be present in the stored cones long after this massive emergence, diapause in R. mollis can last for more than one year. The end of diapause may either be nutritional and determined by the fat reserves carried by the larvae or the result of stimulants or inhibitors. These could be emitted by the plant which bore the cones containing larvae in diapause.

Pollination of Zamia integrifolia

Tang (1987a) conducted observations and exclusion experiments that demonstrated that the Florida native Zamia integrifolia also was pollinated by a weevil, in this case Rhopalotria slossoni, and also by a second beetle species, Pharaxonotha zamiae. Moreover, these observations demonstrated that R. slossoni, like R. mollis, is a species-specific pollinator as is P. zamiae. Neither of these two beetles is active upon Z. furfuracea, although they may pollinate Caribbean relatives of Z. integrifolia (e.g., Z. pumila). This finding has led to the suspicion that the insects that pollinate cycads tend to be host-specific, or at least are faithful to species within a narrow range of taxonomic affinities (Norstog and Fawcett, 1989; Vovides, 1991).

Male cones provide an egg-laying site, as well as food for both the adults and larvae of Rhopalotria slossoni, a weevil resembling R. mollis except for having darker wing covers, and Pharaxonotha zamiae, a pollen-eating beetle also present along with R. slossoni. In this case, oil of wintergreen (methyl salicylate), also a known insect attractant in orchids, is the major volatile component in male Z. integrifolia cones (Pellmyr et al., 1991). Both beetles are able to live only in male cones of the native Florida Z. integrifolia or closely related Caribbean forms of this species. In its essentials, the life cycle of R. slossoni closely resembles that of R. mollis, as does its interactions with its host species of Zamia, and therefore will not be described further. However, the life cycle of Pharaxonotha zamiae is sufficiently different from that of Rhopalotria to merit separate treatment.

Life cycle of Pharaxonotha in Zamia
The place of origin of individuals of P. zamiae that first invade the cones of Z. integrjfolia is uncertain, but it is suspected that they overwinter in soil at the base of the cycad (Fawcett and Norstog, 1993). Pharaxonotha zamiae arrives at male cones when pollen is mature (Fig. 4A-C). 

FIG. 4. Reproductive cycle of Pharaxonotha
in Zamia pumila. A. Young adults 
leave the soil and visit male cones of Z. pumila
in which microsporangia are beginning to dehisce;
B. Beetles forage in the male cone; C. Beetles 
feed on pollen in microsporophylls and lay eggs 
between (arrow) and sometimes in sporangia; 
D. Eggs hatch and larvae feed on pollen; E. When
larvae are ready to moult, they eat a hole in the 
sporophyll into which they enter to moult. When 
they emerge in the next instar they return to 
feeding on pollen. Only the early instars eat pollen 
and by the third instar much of the pol1en is gone; 
F. Later instars eat only parenchyma tissue, 
entering the cone axis through a sporophyll; 
G. Larvae continue feeding until interior tissues of
the cone axis and peduncle are empty but do not 
feed on the stem itself; H. The last larval instar 
eats a hole in the peduncle and drops to the ground; 
I. Larvae burrow into soil, pupate (J.) and after 4-7
days emerge as mature beetles; J. Beetles leave the
soil in search of a new male cone to feed and breed
in; K. Pollination takes place when some individuals
enter a female cone instead of a male cone. They 
neither oviposit in female cones nor feed on them.

(Click on figure to enlarge.)

There they feed and lay eggs between the microsporangia, and sometimes inside dehisced sporangia. Their place of mating is unknown. It is possible that they mate underground, but it is more likely that they mate in the male cones when they go there to feed and lay their eggs.

The eggs are very large in proportion to the size of the female, which carries up to six eggs at a time at various stages of maturity. The adults feed only on pollen. The pollen becomes scattered everywhere, with all of the insects present within the cone, including R. slossoni, becoming covered. Unless the sporangia of Z. integrifolia are disturbed, pollen tends to remain in them in sticky clumps and is undispersed.

The oval eggs, which are greyish-white and translucent, hatch within hours. The larvae eat their way out of the egg and begin feeding on pollen. Larvae in all of the instars look much alike; each body segment has a pair of light-brown patches on the dorsal side, so that the larvae appear striped (Fig. 4D-F).

It is only larvae in the first instars that eat pollen and there is little pollen left in the cone after the first two instars. A larva that is ready to molt eats a hole in the back of one of the sporophylls and crawls inside, later emerging as the next instar (Fig. 4E). Larvae in later instars eat only tissue, entering the cone axis through a microsporophyll, and eating their way down through the cone into the cone peduncle, which becomes completely hollow (Fig. 4F, G). When they have finished feeding and are ready to leave the cone they eat a little hole in the peduncle and drop down into the ground where they burrow into the soil (Fig. 4H, I).

After several days in the ground, larvae become pupae and remain in this stage for 4-7 days (Fig. 4J). New adult beetles, and perhaps older dormant ones, leave the soil in search of male cones (Fig. 4A) and, possibly by chance, also visit female cones (Fig. 4K), but late in the season when no new cones are present, the young adults, instead of flying away to repeat the breeding cycle, burrow into the soil.

Pharaxonotha zamiae is found in female cones of Z. integrifolia in southern Florida, and it is assumed they blunder into them in error as appears to be the case with Rhopalotria. In northern Florida another species of Pharaxonotha appears to be the only pollinator of Zamia and possibly this specificity is related to the subtle genetic differences in allozymes in the northern forms described by Walters and Decker-Walters (1991). The life cycle of P. zamiae is not as tightly linked with its host as is that of R. slossoni in Z. integrifolia or that of R. mollis in Z. furfuracea.

The Case for Wind Pollination

Informative though the studies described above have been, they leave a partially unanswered question: does wind play any role in cycad pollination?

Most exotic cycads at Fairchild Tropical Garden and elsewhere, do not produce fertile seeds in cultivation unless artificially pollinated. This holds true even when pollen-producing male plants are only a meter or so distant from pollen-receptive females. These examples represent the "ultimate pollinator-exclusion experiment," and, at the same time, show that air currents are not pollen vectors for most cycads, otherwise the exotic species should be as fertile or at least have pollen present at the micropyle of the ovules as the native ones. The corollary question of course is that if wind is not the primary agent of pollination in such cycads, does it play any part in a cycad's reproductive cycle?

Cycad pollen grains are large (35-40 pm) and sticky and are not transported beyond a few meters by breezes and light winds. However, Norstog and Niklas (1984) used stroboscopic photography to track paths of pollen movements in wind tunnel experiments. Their results revealed that the shapes of the cones set up air-current vortices which tended to deposit pollen upon and in those parts of the cone that were open--such as at the base of Dioon (Fig. 5); a small vertical crack in Zamia (Fig. 1), and on the blades of the Cycas megasporophylls which at this stage are in erect, cone-like clusters. When these

FIG. 5. Pollination of Dioon califanoi,
hypothesis. Weevils, probably Rhopalotria
are shown landing on open, sterile
sporophylls at the base of a female cone. 
The basal sporophylls are glabrous and form
a kind of "landing platform" where weevils
pick up wind-blown pollen, or, more likely, 
bring it with them from male cones where 
they breed and feed. The construction of the
female cone resembles an old-fashioned fly-
trap, in which weevils move upward in the 
direction of dim light penetrating the felt of 
hairs that invests all but the basal sporophylls.
As they move upward, they rub off pollen on 
the ends of ovules made sticky by micropylar 
fluid. Eventually the weevils escape by 
squeezing past the apical sporophylls.

(Click on figure to enlarge.)

 cones were scored for pollen impact points it was found that pollen was not preferentially deposited on the windward side but more-or-less evenly distributed around the circumference, and it piled up on or in the cone openings. Inspection of the interiors of such cones showed that comparatively little pollen actually impacted on the ovules (20-43 ovules per cone in Z. integrifolia and Z. furfuracea, nil in Dioon edule in which only the basal most, sterile sporophylls are open during pollination). In the two species of Zamla the pollen tended to settle preferentially on ovules near the cracks between megasporophylls. Because these cones contain about 200 ovules per cone, this effect, at best, would have resulted in pollen striking only about 20% of the ovules assuming the unlikely circumstance that every pollen grain settled upon a micropyle. The area of the micropylar tip is <1 mm2 while the surface of the whole ovule is about 80 mm2, so it would have been highly unlikely that more than 1% of the ovules were actually pollinated. Even if such an experiment were continued so as to mimic pollination in nature, it seems unlikely that pollen deposition on micropyles would yield very high seed set. Remarkably, when developing ovules are dissected, pollination by single pollen grains is found to be rare; most ovules contain 3-4 and up to a dozen pollen tubes. In plantings of Z. furfuracea that were the sites for the observations reported here, fertility in control cones approached 98%.

The results of wind-tunnel experiments do not rule out wind as a cycad pollen vector, especially in Cycas. Pollen blown onto the blades of erect Cycas megasporophylls might be washed down to the ovules by dews or rain. However, adjacent male and female specimens of C. rumphii, C. circinalis, C. media, and C. revoluta at Fairchild Tropical Garden exhibit very low fertility unless artificially pollinated. In the case of cones of Dioon and Zamia, Niklas and Norstog (1984) proposed that a combination of wind and insects might function in pollen transport and pollination in such cycads, with air currents depositing masses of pollen upon certain receptive sporophylls and insect visitors distributing it inside the cone. This hypothesis is especially attractive in the case of female cones of Dioon species in which the only access is via a basal whorl of sterile megasporophylls (Fig. 5).

Small-scale experiments indicate that wind plays little or no part in the pollination of Zamia furfuracea in cultivation (Norstog et al., 1992). As noted, in a pollen-receptive cone the single vertical crack, measuring 3-5 cm2 (Fig. 1), provides the only access to the cone's interior. Such cones were monitored for airborne pollen during the period of pollen receptivity of the cones. It was calculated that about 40 pollen grains entered each female cone. Here as well, in the highly unlikely circumstance that every such pollen grain would have impacted directly on a micropyle of one of the 200 or so ovules in a cone, then 40 seeds might have been produced for a fertility on the order of 20%. Actually, the average fertility of cones in this planting in 1981 was 97%, and in 1988 the fertility of the monitored cones averaged 82% (range 63-93%). Coupled with data from exclusion experiments in this planting, we conclude that the case for a significant contribution to fertility by airborne pollen is very weak in this cycad.

One source of evidence in favour of wind pollination has been fertile seed set, although at a low level ranging from 5-40% in insect exclusion experiments of Norstog et al. (1986), Tang (1987a), Donaldson et al. (1995) and Donaldson (1997). However, all of these authors have noted that it is virtually impossible to totally exclude weevils while at the same time making wind pollination possible. The weevils will chew through mesh bags, burrow around fasteners encircling the peduncle that are coated with lanolin, and even borrow through plastic bags. Our interpretation of this is that total exclusion of insects while allowing wind pollination is not totally possible because of the strong attraction of the weevils to the female cones for weevil sexual reproduction.

Pollination biology of Dioon califanoi
Observations in the field on Dioon califanoi in its habitat near Teotilan del Camino, Oaxaca, Mexico, show that the female-cone life cycle of the cycad lasts about two years from initiation of the cone through seed set and cone disintegration. Pollination occurs near the end of the first year. At this stage, the sterile scales at the cone base separate giving rise to interscale apertures which last about 10 days, after which they close up tightly (Fawcett et al., 1995). Beetles enter by way of these apertures and exit by way of the apical sporophylls as shown in Fig. 5. Opening of the basal sterile sporophylls coincides with the activity of the pollinating weevil Parallocorynus bicolor (= Rhopalotria bicolor Oberprieler, 1995) that is emerging from dehiscent male cones as well as from diapause in old, dried cones of previous reproductive episodes of the cycad. Beetle pupation begins in early August and lasts through October to mid-November. The female cones mature through the second year and dehisce at the end of this period, or well into the third year.

There actually appear to be two pollinators in D. califanoi: Parallocorynus bicolor seems to predate the top of the male cone and a species of Pharaxonotha does so in its lower part; the former feeding solely on sporophyll tissue and the latter both on sporophyll tissue and pollen, as in the case of P. zamiae in Z. integrifolia (Fawcett et al., 1995). Dried male cones may persist on the plant well into the second or even third year and contain diapausic weevils. Persistent male cones are not likely to be found when Pharaxonotha is active in large numbers because they are very destructive and destroy the integrity of the cone, resulting in its fragmentation. Individuals of Parallocorynus bicolor that emerge from male cones of D. califanoi are covered with pollen (Vovides unpubl. observ.) and this pollen is carried by the weevils into the female cone via the open, sterile, basal scales. Thus, pollen on the basal cone scales of Dioon is not necessarily attributable to wind deposition as proposed by Niklas and Norstog (1984).

Although strict exclusion experiments have not been made with Dioon cones, there is sufficient circumstantial evidence to propose that Parallocorynus bicolor is the pollinator of D. califanoi. This evidence includes the observations that P. bicolor feeding and reproductive habits are the same as reported by Norstog and Fawcett (1989) for R. mollis on Z. furfuracea, that the life cycle of P. bicolor is tied to D. califanoi, and that the diapause stage is associated with old male cones.

Dioecy in cycads
Cycads are dioecious plants and a function of dioecy may be specialisation of males and females for different functions relating to reproductive fitness. If in fact this is the case in the two species of Zamia discussed above, a clear-cut example is seen in the sex-related differential allocation of starch in cones which provides a reward for pollination by co-symbiotic insects. In addition, alternate pathways of idioblast development may have evolved so that toxins sequestered in male cones make them non-poisonous and hence useful as brood places, while mobilised toxins in female cones suppress predation. Thus, the value of dioecy in these species becomes even more apparent.

Species specificity
We have found that Rhopalotria mollis will live in Zomia loddigesii and Z. spartea, but not Z. integrjfolia, the cones of which contain less starch and are more mucilaginous. Rhopalotria slossoni has been found in other species similar to Zamia integrifolia, but not in Z. furfuracea or similar species. Zamia furfuracea, Z. loddigesii and Z. spartea all produce cones in the summer, whereas Z. integrifolia and the similar Caribbean zamias all produce cones in the winter and pollinator specificities may simply be phenological in origin.

Coevolution of Cycads and Insects

Aside from the value of cycad pollination studies in relation to reproduction and ecology, such studies may shed new light on origins of plant and animal interdependence, not only in pollination but also in protection of seeds during later development and seed dispersal mechanisms as well. Insect pollination is commonly associated with the evolution of flowering plants (Faegri and van der Pijl, 1979), but there seems little doubt that the coevolution of plants and insect pollinators predates the angiosperms (Crepet, 1979; Crowson, 1981, 1989, 1991).

Coevolution involves an interdependence of two species which have undergone morphological and behavioral modifications of mutual import (Ehrlich and Raven, 1964; Janzen, 1980). The mutualism exhibited by Rhopalotria mollis and Zamia furfuracea and R. slossoni and Z. integrifolia well may be an ancient coevolution, although there is no direct evidence that these organisms did indeed evolve together, and curculionid beetles may have parasitised angiosperms before going on to cycads (Crowson, 1991). Even so, these kinds of partnerships might have originated very early in the evolution of both beetles and cycads. Clearly, with each cycad genus having its own genus of curculionid pollinator and with each cycad species within a genus apparently having its own species within a genus of pollinators, coevolution between cycads and weevils has and is occurring.


Much remains to be discovered about the relationships of insects and cycads. All attempts at cultivation and maintenance of the plants with a view to their introduction into natural habitats will be completely pointless if the insect pollinators are lost, and it would be of great value if additional workers could be encouraged to study this important area, perhaps rather than concentrating on taxonomic questions. In particular we need to know the life history and environmental requirements of the pollinators for each cycad species. We also need to survey the effects that man-made insecticides used in agriculture and forestry are having on the reproductive capacity of adjacent or incorporated populations of cycads; this may be the primary reason why some species are failing to reproduce in the wild.


Baird, A.M. (1939). A contribution to the life history of Macrozamia riedlei. J. Roy. Soc. W. Australia 25: 153-169.

Chadwick, C.E. (1993). The role of Tranes lyterioides and T. sparsus Boh (Col., Curculionidae) in the pollination of Macrozamia communis (Zamiaceae). In: D.W. Stevenson and K. J. Norstog (eds). Proceedings of the Second International Conference on Cycad Biology, pp. 77-88, Palm and Cycad Societies of Australia, Milton, QId.

Chamberlain, C.J. (1919). The living cycads. University of Chicago Press, Chicago.

Chamberlain, C.J. (1935). Gymnosperms, structure and evolution. University of Chicago Press, Chicago.

Connell, S.W. and Ladd, PG. (1993). Pollination biology of Macrozamia riedlei &mdash; the role of insects. In: D.W. Stevenson and K.J. Norstog (eds), Proceedings of the Second International Conference on Cycad Biology, pp. 96-102. Palm and Cycad Societies of Australia, Milton, QId.

Crepet, W. (1979). Insect pollination: a paleontological perspective. Bioscience 29: 102-108. 

Crowson, R.A. (1981). The biology of the Coleoptera. Academic Press, London.

Crowson, R.A. (1989). Relationships to cycads. International Congress of Coleopterology, Barcelona. Sept 18-23, 1989 (Abstracts) pp. 13-15, European Association of Coleopterology, Barcelona.

Crowson, R.H. (1991). The relations of Coleoptera to Cycadales. In: M. Zunino, X. Beltes and M. Blas (eds), Advances in coleopterology, pp. 13-28, European Association of Coleopterology, Barcelona.

Donaldson, J.S. (1997). Is there a floral parasite mutualism in cycad pollination? The pollination biology of Encephalartos villosus (Zamiaceae). Amer. J. Bot. 84: 1398-1406.

Donaldson, 3.8., NMnni, I. and do Wet Bbsenberg, .1, (1995). Beetle pollination of Encephalartos cycadifolius. In: P. Vorster (ed.). Proceedings of the Third International Conference on Cycad Biology, pp. 423-434, The Cycad Society of South Africa, Stellenbosch.

Ehrlich, PR. and Raven, P.H. (1964). Butterflies and plants: a study in coevolution. Evolution 18: 403-416.

Faegri, K. and Pijl, L. van der (1979). The principles of pollination ecology, ed. 3. Pergamon Press, New York.

Fawcett. P.K.S. and Norstog, K.J. (1993). Zamia pumila in South Florida: a preliminary report on its pollinators R. slossoni, a snout weevil and P. zamiae, a clavicorn beetle. In: D.W. Stevenson and K.J. Norstog (eds). Proceedings of the Second International Conference on Cycad Biology, pp. 109-120, Palm and Cycad Societies of Australia, Milton, Qld.

Fawcett, P.K.S., Norstog, K.J. and Vovides, A.P. (1995). Pollination of two species of Zamia in south Florida and its implication for evolution of cycad strobili. In: D.D. Pant, A.N. Bhatnagar, K.R. Surange, M.N. Bose, and P.K. Khare (eds) Proceedings of the International Conference on Global Environment and Diversification of Plants through Geological Time, pp. 111-121, South Asian Publishers, Allahabad, India. 

Janzen, D.H. (1980). When is it coevolution? Evolution 34: 611-612. 

Lawson, A.A. (1926). A contribution to the life history of Bowenia. Proc. Roy. Soc. Edinburgh 54: 357-394.

Marloth, R. (1914). Notes on the entomophilous nature of Encephalartos. Trans. Roy. Soc. South Africa 4: 69-71.

Niklas, K.J. and Norstog, K. (1984). Aerodynamics and pollen grain depositional patterns on cycad megastrobili: implications on the reproduction of three cycad genera (Cycas, Dioon, and Zamia). Bot. Gaz. 145: 205-222.

Norstog, K.J. and Fawcett, P.K.S. (1989). Insect-cycad symbiosis and its relation to the pollination of Zamia furfuracea (Zamiaceae) by Rhopalotria mollis (Curculionidae). Amer J. Bot. 76: 1380-1394.

Norstog, K.J. and Nicholls, T.J. (1997). The biology of the cycads. Cornell University Press, Ithaca, NY.

Norstog, K and Stevenson, D.W. (1980). Wind? Or insects? The pollination of cycads. Fairchild Trop. Gard. Bull. 35(1): 28-30.

Norstog, K.J., Stevenson, D.W. and Niklas, K.J. (1986). The role of beetles in the pollination of Zamia furfuracea L. fil. (Zamiaceae). Biotropica 18: 300-306.

Norstog, K.J., Fawcett, P.K.S. and Vovides, A.P. (1992). Beetle pollination of two species of Zamia: evolutionary and ecological considerations. Palaeobotanist 41: 149-158.

Oberprieler, R.G. (1989). Platymerus, a forgotten weevil. Pellea no. 8: 50-54.

Oberprieler, R.G. (1995). The weevils (Coleoptera: Curculionoidea) associated with cycads 1. Classification, relationships, and biology. In: P. Vorster (ed.). Proceedings of the Third International Conference on Cycad Biology, pp. 295-324, The Cycad Society of South Africa, Stellenbosch.

Ornduff, R. (1993). Features of the reproductive biology of Macrozamia species and Cycas media in Australia. In: D.W. Stevenson and K.J. Norstog (eds). Proceedings of the Second International Conference on Cycad Biology, pp. 121-124. Palm and Cycad Societies of Australia, Milton, Qld.
Pearson, H.H.W. (1906). Note on the South African cycads. I. Trans. S African Philos. Soc. 16: 341-354.

Pellmyr., 0, Tang, W. Groth, I. Bergstrom, G. and Thien. LB. (1991). Cycad cone and angiosperm floral volatiles: inferences for the evolution of insect pollination. Biochem. Syst. & Ecol. 19: 623-627.

Rattray, G. (1913). Notes on the pollination of some South African cycads. Trans. Roy. Soc. South Africa 3: 259-270.

Tang, W. (1987a). Insect pollination in the cycad Zamia pumila (Zamiaceae). Amer J. Bot. 74:

Tang, W. (1987b). Heat production in cycad cones. Bot. Gaz. 148: 165-174.

Tang, W. (1993). Heat and odour production in cycad cones and their role in insect pollination. In: D.W. Stevenson and K.J. Norstog (eds). Proceedings of the Second International Conference on Cycad Biology, pp. 140-147, Palm and Cycad Societies of Australia, Milton, QId.

Tang, W., Steinberg, L. and Price, D. (1987). Metabolic aspects of thermogenesis in male cones of five cycad species. Amer J. Bot. 74: 1555-1559.

Vega, A. and Bell, A.E. (1967). a-amino-b-methylaminopropionic acid, a new amino acid from seeds of Cycas circinalis. Phytochemistry 6: 759-762.

Vovides, A.P. (1991). Cone idioblasts of eleven cycad genera: morphology, distribution and significance. Bot. Gaz. 152: 91-99.

Vovides, A.P., Norstog, K.J. Fawcett, P.K.S. Duncan, M.W. Nash, R.J. and Molsen, D.V. (1993). Histological changes during maturation in male and female cones of Zamia furfuracea L. fil. (Zamiaceae, Cycadales) and their significance in relation to pollination biology. J. Linn. Soc., Bot. 111: 241-252.

Walters, T.W. and Decker-Walters, D.S. (1992). Patterns of allozyme diversity in the West Indies cycad, Zamia pumila. Amer. J. Bot. 78: 436-445.

Whiting, M.G. (1963). Toxicity of cycads. Econ. Bot. 23: 270-302.

Wilson, G.W. (1993). Initial observations of the reproductive behaviour and an insect pollination agent of Bowenia serrulata (W. Bull) Chamberlain. Encephalartos no. 36 (Dec.): 13-18.

Note: This article was originally published as "Stevenson, D.W., Norstog, K.J. and Fawcet, P.K.S., (1998). Pollination biology of cycads. In: S.J. Owens and P.J. Rudall (Editors). Reproductive Biology, pp. 277-294. Royal Botanic Gardens, Kew." It is reprinted here with permission.

PACSOF Home Page

VCE Home Page

VCE Table of Contents

VCE Photo gallery

Virtual Cycad Encyclopedia Site Map
Powered by FreeFind.

This site is copyrighted 1998-2006, Palm & Cycad Societies of Florida, Inc.
For questions or comments, e-mail the webmaster.
Internet hosting provided by Zone 10, Inc.