Artocarpus (Moraceae)-Gall Midge Pollination Mutualism Mediated by a Male-Flower Parasitic Fungus

by Shoko Sakai, Makoto Kato, Hidetoshi Nagamasu
Artocarpus (Moraceae)-Gall Midge Pollination Mutualism Mediated by a Male-Flower Parasitic Fungus
Shoko Sakai, Makoto Kato, Hidetoshi Nagamasu
American Journal of Botany
Start Page: 
End Page: 
Select license: 
Select License





Center for Ecological Research, Kyoto University, Sakyo, Kyoto, Japan 606-8502; 'Graduate School of Human and
Environtnental Studies, Kyoto University, Sakyo, Kyoto, Japan 606-8501; and The Kyoto University Museum,
Kyoto University, Sakyo, Kyoto, Japan 606-8501

A previously undescribed pollination system involving a monoecious tree species, Artocarp~ls integer (Moraceae), polli- nator gall midges, and fungi is reported from a mixed dipterocarp forest in Sarawak, Borneo. The fungus Clzoanepkorn sp. (Choanephoraceae, Mucorales, Zygomycetes) infects male inflorescences of A. integer, and gall midges (Contarinin spp., Cecidomyiinae, Diptera) feed on the fungal tnycelia and oviposit on the inflorescence. Their larvae also feed on the mycelia and pupate in the inflorescence. The gall tnidges are also attracted by fetnale inflorescences lacking mycelia, probably due to a floral fragrance similar to that of tnale inflorescences. Because of the sticky pollen, dotninance of Co~~tnritzin

spp. in flower visitors, and pollen load observed on Contririnin spp. collected on both tnale and fetnale inflorescences, Artocarpzrs integer is thought to be pollinated by the gall midges. Although several pathogenic fungi have been reported to have interactions with pollinators, this is the first report on a pollination mutualism in which a fungus plays an indispensable role. The pollination systetn described here suggests that we should be more aware of the roles fungi can play in pollinations.

Keg w-ords: Borneo; gall midge; Lambir; monoecy; Moraceae; Mucorales; pollination.

Animal-pollinated flowers have evolved various re-(Choanephoraceae, Mucorales, Zygomycetes), and gall wards and attractants to entice pollinators and thus pro- midges (Contarinia spp., Cecidomyiinae, Diptera) and mote pollen transport. Whereas nectar and pollen are the their larvae feed on the fungal mycelia. The gall midges most common rewards, floral tissues may also be offered are attracted to both male and female inflorescences and as brood sites for pollinators in some plant groups. Plants thus pollen transport occurs. This is the first report on a such as Ficus (Moraceae) (Janzen, 1979; Wiebes, 1979; pollination mutualism in which a fungus plays an indis- Bronstein and McKey, 1989; Herre and West, 1997), Yuc-pensable role. ca (Agavaceae) (Baker, 1986), Trollius (Ranunculaceae) Moraceae are a family with 38 genera and 1100 spe- (Pellmyr, 1989), Heuchera (Saxflagaceae) (Pellmyr et al., cies distributed mainly in tropical regions (Mabberley, 1996) sacrifice a portion of their ovules for brood sites 1997). Throughout the family, unisexual flowers of both of their pollinators. On the other hand, other plants have sexes are produced on the same or separate plants and developed brood sites and nutrients for pollinators inde- are often arranged in heads to form spadices. Animal pol- pendent from ovules. Androecial parts of plants function lination has been confirmed only in Ficus and a few other as brood sites for pollinators in Peltandra (Araceae) (Patt genera of the family (Bawa et al., 1985; Momose et al., et al., 1995), Zamia (Zamiaceae) (Tang, 1987; Norstog 1998), whereas wind pollination seems to be common and Fawcett, 1989), and Eupomatia (Eupomatiaceae) (Zapata and Arroyo, 1978; Bawa and Crisp, 1980; Wil- (Armstrong and Irvine, 1990). In most of the cases, how- liams and Adam, 1993; Bullock, 1994). The genus Arever, plants and pollinators directly exchange service and tocarpus consists of 50 species including some cultivated rewards, and participation of fungi in pollination has rare- species such as jack-fruits and is distributed throughout ly been reported, although fungi may sometimes have the Indomalesia region (Primack, 1983; Mabberley, great effects on the pollination success of their host or 1997). Both animal-pollinated and wind-pollinated spe- nonhost plants (Patt, 1992; Roy, 1994, 1996). cies are thought to occur in the genus (Corner, 1952; van

Here, we report on the pollination system of Artocar-der Pijl, 1953; Primack, 1983; Endress, 1994). pus integer (Thunb.) Men: (Moraceae) from Sarawak, Borneo. It appears that male inflorescences of A. integer MATERIALS AND METHODS On a tree are infected by the fungus Choanephora sp'

Study site nndplnitt-The study site was a primary lowland diptero- carp forest in Latnbir Hills National Park, Sarawak, Malaysia (4"201 N, ' Manuscript received 11 Septetnber 1998; revision accepted 18 June 113"50' E, altitude 100-200 m). In August 1992, a Canopy Biology 1999. Plot (CBP) was demarcated for long-term monitoring of plant phenol- The authors thank T. Inoue, K. Ogino, *.*. Hamid, and s. Lee ogy and for the observation of plant-animal interactions, by the Canopy

for providing us with the chance to work in CBPS; K. ~Momose for Biology Program of Kyoto University and Sarawak Forest Department,collection of samples; A. Kohzu, T. Tateishi, and T. Osono for identi- The CBP covered an area of ha (200 400 and had a canopyfication of and information about fungi. This study was partly supported observation system that consisted of tree towers and aerial walkways by Grants-in-Aid of the Japanese Ministry of Education, Science and et 1995).

Culture (numbers 04041067, 06041013, 10041 169, and 09NP1501) and by JSPS Research Fellowships for Young Scientists for S. Sakai.

Artocarpus integer, called "chempedak" by local people and often *Author for correspondence, culvent address: Slnithsonian Tropical cultivated for edible fruits, is a monoecious tree up to 30 tn in height Research Institute, Apartado 2072, Ancon, Balboa, Republic of Panama. with a widespread distribution in primary and secondary dipterocarp


forests in Sarawak. We observed six individuals of the species: three trees of -20 cm dbh (diameter at breast height) at the forest edge near the headquarters of the National Park (Kl, K2, K3) and three trees of 30-45 cm dbh in CBP (B187, B188, and B27.1).

Observation and collection of power visitors-The midges on the marked male and female inflorescences on K1, K2 and K3 were counted at 0600, 1200, 1800, 2000, 2200, and 2400 on 30 April 1998 (four tnale and four female inflorescences) and 11 May (14 tnale and three female). Data at 2400 on 30 April were not collected because of heavy rain. At each collection, behaviors of tnidges on inflorescences were observed for 10-60 min. In addition, 92 male inflorescences of various stages from the six trees were collected and examined under a binocular.

Flower T-isitors on male and female inflorescences of the six trees were collected with suction bottles from the tree crowns up to 2 m high except B274. Collection on B274, the largest tree, was tnade using a walkway at the height of 20 tn. Some of the sampled midges were used to exatnine pollen load under a stereoscopic microscope and an electron microscope. In addition, insect parasites inhabiting inflorescences were investigated by incubating inflorescences in plastic containers at our field laboratory.


Flowering of Artocarpus integer usually occurs once in several years in mature forests, and its flowering seems to be associated with general flowering as well as other Artocarpus species in the park, while more frequent re- production was found in cultivated plants (Sakai et al., 1999; Sakai, personal observation). We started to monitor the largest tree, B274, in April 1993 (Sakai et al., 1999) and observed flowering for the first time in May 1998 except for minor flowering in July 1996.

Although A. integer is monoecious, the production of female inflorescence varied among trees irrespective of tree size. Among the six trees we observed, K3 and B 188 produced more female inflorescences (-30) than the oth- ers, and K1 and B 187 rarely did. On the other hand, male inflorescences were found in all trees. Trees flow- ered for -2 mo.

Male inflorescences are cylindrical, measuring 10 X 50 mm (Fig. I), and are produced both on branches of var- ious sizes or on the trunk of the tree. While they are flowering, the protrusion of stamens followed by anther dehiscence occurs continuously for 3-5 d. Pollen is spherical with a rough surface and measures 11 pm in diameter. It is barely dispersed by winds. As male flowers age, their color changes from greenish yellow to brown, and they gradually become covered with mycelia of Choanephora sp. (Choanephoraceae, Mucorales, Zygomyce- tes) (Fig. 2). Between days 4 and 6, Choanephora on the inflorescence produces sporangia and releases spores (Fig. 3). The fungal spores are ellipsoid, 18 pm long X 10 pm in diameter, and are not sticky. The inflorescences are then shed between days 5 and 9.

Female inflorescences are cauliflorous and ellipsoid, measuring 15 X 50 mrn, and are green with white thread- like pistils of 1.5 mm in length (Fig. 4). The flowering ~eriod of a single female inflorescence lasts for -10 d.


After the receptive stage of the stigma, the stigma turns brown and withers. Almost all female inflorescences grow into fruits, which measure 30 X 18 cm. No mycelia or sporangia were observed on female inflorescences. Both male and female inflorescences release a scent sim- ilar to ripe watermelon between 1800 and 2000 while flowering.

The predominant flower visitors were two species of gall midges (Contarinia spp., Cecidomyiidae, Diptera) (Figs. 1, 2, 4), among which 92% were Contarinia sp. 1 (Table I). The proportion of female midges on male in- florescences was significantly higher than that on female inflorescences (P < 0.001, x2test; Table 1). Pollen grains were not found in their stomachs. Most of the midges collected on both male (85%, N = 48) and female inflo- rescences (84%, N = 25) carried pollen, but no fungal spores were found on the midges (Figs. 5, 6). Rare visits by cockroaches and beetles were also recorded (Table 1).

Midges on male inflorescences were observed eating exudate of the mycelia. Frequent oviposition by female midges and numerous larvae and pupae of the gall midg- es (69 2 34 larvae and pupae per inflorescence, N = 6) were observed on male inflorescences (Fig. 7). Eggs and larvae of gall midges and fungal mycelia were found on most male inflorescences except for young ones just start- ing to flower (98%, N = 92). The midge larvae on a male inflorescence attached to the mycelia and sucked liquid from them. A larva remained on a single male inflores- cence and pupated inside the spadix, which had become soft and rotten because of the fungus. Both species of Contarinia emerged from male inflorescences. Female midges did not have mature eggs upon emergence. The period required to grow from an egg to an adult midge was estimated to be 8 d. Only a few eggs were found on female inflorescences, and the eggs on female inflores- cences did not develop, as evidenced by the lack of larvae and emerging adults from female inflorescences.

Although the gall midges remained on inflorescences throughout the day, density of gall midges on male inflo- rescences fluctuated considerably. They flew among the inflorescences most actively from 1800 to 2000, and re- corded the highest density from 2000 to 0600 on the fol- lowing day (Fig. 8). In contrast, density on female inflo- rescences was maintained at a lower level than that on male inflorescences (Fig. 8).


Although we did not perform exclusion experiments to confirm the contribution of the gall midges to pollination, sticky pollen, which is not easily dispersed by wind, dominance of two species of Contarinia in flower visi- tors, and pollen load observed on most Contarinia spp. collected on inflorescences of both sexes, indicate that Artocarpus integer is pollinated by gall midges, Contarirzia spp. The midges are attracted to both male and fe- male inflorescences by a fruit-like, somewhat unpleasant smell. On the male inflorescences, which are infected by the fungus Choanephora, female midges ingest nutrition from the mycelia for ovary maturation and oviposit, whereas they seem to leave a female inflorescence lack- ing mycelia soon after arriving (Fig. 9). Male gall midges possibly visit inflorescences for mating, while we could not observe mating behavior of gall midges. Male midges may not feed on mycelia as eagerly as the females and remain on a female inflorescence longer. The difference in their behavior can explain the higher ratio of female

TABLE1. The numbers of gall midges collected on inflorescences of six Artocarpus trees

Male inflorercence Female inflorescence

Cotiinrtnio FP. 1 Coninrinin sp. 2 Conin,-inia sp. 1 Co,iiarinia sp 2 Tree ID Male Female Male Female Others Male Female Male Female Otherr

B187 2 15
B188           17 22
B274   16   1      
K1 2 32   2 1    
K2   8     1 3 7   4
K3 14 3 7 1 7          
Total 18 108 1 10 2" 20 29 0 4 0

Toleoptera (1'1 and Blattodea (1'1.

midges on a male inflorescence compared to those found on female inflorescences (Table 1).

Midge larvae also grow up feeding on the mycelia and subsequently emerge from the inflorescence (Fig. 9). During the relatively long flowering period of a tree (2 mo), gall midges can increase their population dramati- cally. Although female inflorescences are not infected by the fungus and therefore do not function as a brooding place for gall midges, the midges are sometimes deceived by its odor and on occasions oviposit even on a female inflorescence. Many myophilous flowers work by imitat- ing oviposition substrate, and olfactory cue is thought to be important in these systems (Dafni, 1984; Roy and Ra- guso, 1997). This mimicry system, however, may seldom disadvantage the pollinators, because female inflores- cences occur much less frequently than male inflores- cences, and male inflorescences are easily found nearby on the same tree. No floral secretion was found in the species in a natural forest, while a secretion containing protein was reported in female inflorescences of cultivat- ed A. integer (Momose et al., 1998). Genetic differences between wild plants and plants under cultivation remain to be studied.

The fungus Choanephora sp. belongs to the family Choanephoraceae, in which all members appear to be re- stricted to the tropics. They are soil fungi, which can be facultative flowerlfruit parasites on a wide range of host plants (Issac et al., 1993). The genus Choanephora includes the well-known pathogen C. cucurbitarum, which infects floral parts of many plants after fertilization and invades the fruits causing soft and wet rot in a number of vegetable crops (Agrios, 1988). Choanephora sp. found on inflorescences of A. integer may also be more or less generalist parasites on plant reproductive organs. While some plant diseases are transmitted by pollinators of their host plants (Alexander, 1987, 1990; Roy, 1994), we could not find fungal spores from limited samples examined by scanning electron microscope. The spores may be predominantly dispersed by wind as those of many Mucolares are thought to be.

Contarinia is a large genus in the gall-midge family Cecidomyiidae with numerous species displaying a va- riety of living habits. In addition to gall-making forms, from which the common name gall midges arises, the family contains many species that are phytophagous on

0 0 0 0 0

0 0 0 0 0 Fig. 9. Diagram illustrating Artocrirp~~s iizteger-gall midge (Cotz-

a3 d CO


2 CU 0 2 tariizia spp.) pollination mutualism mediated by parasitic fungi (Cho- atzephor~sp.). The fungi grow on male inflorescences of A. integer,


and the midges are parasites on the parasitic fungus. Artocarprls and Fig. 8. Changes in the mean number of Contriritzia spp. on male the midges are mutualistic since the midges serve Artocarp~~s

as polli- and female inflorescences. Values are means t 1 SE. nation agents.

flower heads or stems without making galls; still others are found to be mycophagous, predaceous, or parasitoid (Skuhhrava, Skuhhravy, and Brewer, 1984). While gall midges are often observed visiting and laying eggs in flowers of different plants (Proctor and Yeo, 1973), pol- lination by gall midges is only known from Siparuna spp. (Monimiaceae) (Feil, 1992). In the case of Siparurza, gall midge eggs are laid within flowers and the larvae grow up in the flower probably feeding on floral tissues, where- as the emergence of adult gall midges has not been con- firmed. The Contarinin species found in A. integer flowers are the first cecidomyiid midges that parasitize and ~ollinate the flowers.

Pollination mutualism in which a fungus plays an es- sential and indispensable role has never been reported. It is quite interesting that both the pollinator of A. integer and the fungus are originally flower parasites. In this sys- tem, however, the midges are parasites on the parasitic fungus, and Artocarpus and the midges are mutualistic since the midges serve Artocarpus as pollination agents (Fig. 9). Several other pathogenic fungi have interactions with pollinators (Batra and Batra, 1985; Alexander, 1987, 1990; Roy, 1994). In Moraceae, for example, the patho- gen Fzisarizim, transmitted by the fig wasp and causing endosepsis (endo, inner; sepsis, rotting) of figs, is known to damage cultivated figs (Carter, 1973). However, only two other systems have been described that are very sim- ilar to the system reported here. Peltandra virginica is pollinated by a fly that breeds in the inflorescences and feeds on a rust fungus, which it also transmits (Patt, 1992; Patt et al., 1995). The other pollinator fungus mutualism involves the pollination of Epichloe fungi by flies. The flies feed on hyphae and transport spermatia of their host fungi, Epichloe (Bultman et al., 1995). However, in the former example the fungus does not seem to be essential for pollination, and pollination of a plant is not directly involved in the latter case.

There are some examples in which androecial parts of plants function as a brood site for pollinators. Peltandra virginica (Araceae) is pollinated by chrolophid flies. The adult flies feed on pollen and their larvae complete their development within the floral chamber (Patt et al., 1995). The flies are thought to breed only on inflorescences of

P. virginia. Pollinators of Zarnia furfuracea (Zamiaceae) are host-specific weevils, which feed and reproduce on male cones of the plant (Norstog and Fawcett, 1989). Pollination mutualism of weevil beetles and Euuoinatia (Eupomatiaceae) also seems to be species specific. In these cases, the pollinators are adapted for flowering phe- nology of host plants and may be inactive during the nonflowering season (Armstrong and Irvine, 1990).

We do not have definite information on host specificity of Chonrzephora sp. or Contnrinia spp. Rather low host specificity reported in other Choanephora species (Issac et al., 1993) suggests that the Cor~tariniasp. found on A. integer is also a more or less generalist species. Consid- ering the irregular and supra-annual flowering of Artocarpus and the aseasonal climate of the forest, it is un- likely that Contnrinia spp. are inactive or dormant during the nonflowering season. It is possible that Choarzephora sp. use fruits of various plants as substrates and that Contnrinia spp. usually feed and breed on fruits infected by Choanephora sp. or other fungi. When irregular flower- ing of A. integer starts, the fungi infect male inflores- cences and the inflorescences are visited by Contarinia spp. A large population of Contarinia bred by male in- florescences could supply enough pollination service, al- though flowering of A. integer occurs at long and irreg- ular intervals.

The pollination system of the genus Artocarpus still needs study. There are only brief reports on drosophilid fly pollination in A. heterophyllus (van der Pijl, 1953), moth pollination in A. odoratissinzus (Momose et al., 1998), and wind pollination of the other species (Corner, 1952; Primack, 1983; Endress, 1994). The pollination system described here suggests that the evolution of pol- lination mutualisms in the Moraceae may be more wide- spread than just the fig-fig-wasp mutualism and that we should be more aware of the roles fungi can play in pol- lination.


AGRIOS,G. N. 1988. Plant pathology, 3rd ed. Academic Press, San Diego, CA. ALEXANDER. 1987. Pollination limitation in a population of Si-

H. M. lene albn infected by the anther-smut fungus, Cr,~tilngoviolacen. Jo~rnzalof' Ecology 75: 77 1-780.

. 1990. Epidemiology of anther-smut infection of Silei7e alba caused by Ustilago 1,iolncea: patterns of spore deposition and dis- ease incidence. Jo~(rnn1of Ecology 78: 166-179.

AR~ISTRONG, 1990. Function of sta~ninodia in

J. E.. AND A. K. IRVINE. the beetle-pollinated flowers of Eupor~zntin Inurir7n. Biotropicn 22: 429-43 1.

BAKER,H. G. 1986. Yuccas and yucca moths-a historical commen- tary. A1zr7nls ofthe ~Wissoilri Botnrzicnl Garden 73: 556-564. BATRA,L. R.. AND S. W. T. BATRA. 1985. Floral mimicry induced by mummy-berry fungus exploits host's pollinators as vectors. Science

228: 1011-1013. BAWA,K. S., AND J. E. CRISP. 1980. Wind-pollination in the understory of a rainforest in Costa Rica. Joirrnal qf'Ecology 68: 871-876.

. S. H. BULLOCK, D. R. PERRY, R. E. COVILLE, AND M. H. GRA- yuxr. 1985. Reproductive biology of a rain forest in Costa Rica. Jo~(rr7nl of Ecology 68: 871-876.

BRONSTEIN,J. L., AND D. MCKEY. 1989. The fig-pollinator mutualism:

a model system for comparative biology. E.xl~erientin45: 601-604. BULLOCK, 1994. Wind pollination of neotropical dioecious trees.

S. H.

Biotropicn 26: 172-179. BULT~~AS, A. M. WELCH, AND J.

T. L., J. E WHITE, JR., T. I. BOWDISH. JOHNSTON.1995. Mutualistic transfer of Epichloe spermata by Phoi-l~iaflies. M~~cologin

87: 182-189. CARTER,W. 1973. Insects in relation to plant disease. Wiley, New York,

Nen. York, USA. CORNER,

E. J. H. 1952. Wayside trees of Malaya 11. Government Print- ing Office, Singapore. Malaysia. DAFNI,A. 1984. Mimicry and deception in pollination. Annun1 Review: of Ecology arzrl S?,steinntics 15: 259-278. ENDRESS,P. K. 1994. Diversity and evolutionary biology of tropical flowers. Cambridge University Press, Cambridge, UK.

FEIL,J. P. 1992. Reproductive ecology of dioecious Sipnruna (Monimiaceae) in Ecuador-a case of gall midge pollination. Biological Jo~(n7(11of Lit7nearl Society 110: 171-203.

HERRE, E. A., AND S. A. WEST. 1997. Conflict of interest in a mutu- alism: documenting the elusive fig wasp-seed trade-off. Proceedings of'tlze Royal Socieh of' London Series B 264: 1501-1507.


T., Construction of a canopy observation system in a tropical rainforest of Sarawak. Selhyarza 16: 100-1 1 1.


R. WATLING, 1993. Aspects of tropical mycology. Cambridge University Press, Cam- bridge, UK.

JANZEN,D. H. 1979. How to be a fig. A~zn~tnl

Review qf'Ecology nnrl Systeinatics 10: 13-5 1.


D. J.

Press, Cambridge, UK. MOMOSE,K., A. HATADA, R. YAMAOKA,

AND T INOUE. 1998. Pollination biology of the genus Artociirpus, Moraceae. Tropics 7: 165-172.

NORSTOG.K. J., AND l? K. S. FAWCETT. 1989. Insect-cycad symbiosis and its relation to the pollination of Znn7in j'irifirrncen (Zamiaceae) by Rlzo~~alotricir~zollis(Cuclionidae). Ai~zericai~Jourrzal of' Botar2~.

76: 1380-1394.

PATT, J. M. 1992. A plant-pollinator obligate mutualism stabilized by a rust fungus. ~Lrjletii?-of rlie Ecolog~ccil Socieh of' Ai,zericc~73: 296-297. , J. C. FRENCH, C. SCHAL, J. LECH, AND T. G. HARTMAN.1995. The pollination biology of tuckahoe, Peltar2elrci virgiilica (Araceae). Arnericnr7 Jo1lr17al of' Botany 82: 1230-1 240.

PELLMYR,0. 1989. The cost of mutualism: interactions between Trollilts elrropaeus and its pollinating parasites. Oecologici 78: 53-59. , J. M. THOMPSON, AND R. G. HARRISON.

J. M. BROWN, 1996. Evolution of pollination and mutualism in the yucca moth lineage. Ailrericar2 !Vnturalist 148: 827-847.

PRIMACK, 1983. Forester's guide to the Moraceae of Sarawak.

R. B. Forest Department, Sarawak, Malaysia. PROCTOR,M., AND l? YEO. 1973. The pollination of plants. William Collins Sons 8: Co. Ltd., London, UK.

ROY, B. A. 1994. The use and abuse of pollinators by fungi. Trer2els in Ecology clnd Evolilrior2 9: 335-339.

. 1996. A plant pathogen influences pollinator behavior and may influence reproduction of nonhosts. Ecology 77: 2445-2457. . A~DR. A. RAGUSO. 1997. Olfactory versus visual cues in a floral mimicry system. Oecologici 109: 414-426. SAKAI,S., K. MOMOSE, H. NAGAMASU, A.



AND T. INOUE. 1999. Plant reproduc- tive phenology over four years including an episode of general flowering in a lowland dipterocarp forest, Sarawak. Malaysia. Ar17ericcii7 Jo~rrnnl of' Botnily 86: 1414-1436.


M., V. SKUHHRAVY, 1984. Biology of gall midges. Ii1 T. N. Ananthalwishnan [ed.]. Biology of gall insects. 169-222. Oxford & IBH Publishing Co., New Delhi, India.

TANG,W. 1987. Insect pollination in the cycad Ziir~iici plrr~iiln (Zamiaceae). Airier-icci Jo~ri.r~cilof Botnrly 67: 2066-2070.

\:AN DER PIJL,L. 1953. On the flower biology of some plants from Java-with general remarks on fly-traps (species of Annonc~, Ar- tocarpus, T\.~I~~oI~~LIII~,GIICJ~LIIII, Ai7r7nls Bo-

Arisner~znand Ahror17c1).goriei7ses 1: 77-99. WIEBES.J. T. 1979. Coevolution of figs and their insect pollinators. AI~I~LI~II 10: 1-12.

Re~~ieri:of' Ecology clr7d S>,ster~rntics

WILLIA~IS, 1993. Ballistic pollen release in Australian

G., AND P. ADA~I. members of Moraceae. Biotrol,ic~o25: 478-480. ZAPATA,T. R., AND M. T. K. ARROYO. 1978. Plant reproductive ecology of a secondary deciduous tropical forest in Venezuela. Biotropica

10: 221-230.

  • Recommend Us