Primary tabs



Parasitic plants, rootless, without chlorophyll, always rich in tannins, monoecious or dioecious. Endophytic body as strings (chains) or plates of cells inside host plant. Flowers uni- or bisexual; perianth single (perigone), often tubular or saucer-shaped at base, in Rafflesia and Sapria the apex of the tube partly closed by a diaphragm; stamens grouped around a central column representing part of the pistil in female and bisexual flowers, often joined with that column; ovary unilocular, in general inferior but in some cases semi-inferior or superior (Mitrastema), with 4-6 or numerous parietal placentas. Fruits berry-like. Seeds minute, surrounded by pulp, testa hard, often thickened and pitted, embryo few-celled, undifferentiated.


Africa: present America: present Asia-Temperate:, Iraq (Iraqpresent) Mexico, Guatemala, and Colombia: present SW Australia: present including Iran: present
As defined here the family is subdivided into two subfamilies with a total of nine genera and about 40 species, mainly in tropical regions, although some species occur in subtropical and even temperate parts of the world. The three genera of the tribe Rafflesieae of the subfamily Rafflesioideae are only known from Indomalesia, the other two tribes of that subfamily (Cytineae and Apodantheae) occur in America, Africa, the Mediterranean region (including Iran and Iraq), and SW Australia. The subfamily Mitrastemoideae has only one genus, Mitrastema, originally found in E and SE Asia, but later also recorded from Mexico, Guatemala, and Colombia (Meijer 1993).


The ecology of Rafflesiaceae is of course closely interwoven with the host-parasite relationship, the chances of survival and the specific ecological requirements of the host plants, pollination efficiency, seed dispersal and dormancy, and germination chances. Between the production of about 100-150 seeds in a peppercorn-sized fruit of Pilostyles and a fist-size fruit of Rafflesia with several millions of seeds exists an enormous range of seed production per flower.

Mitrastema is most consistently reported as parasitizing the roots of Fagaceae: Quercus, Lithocarpus, Trigonobalanus, and Castanopsis and, if collectors' notes are accurate, occasionally adjacent tree roots of other families. All three genera of the tribe Rafflesieae (Rhizanthes, Rafflesia, Sapria) are obligate parasites of the genus Tetrastigma in Vitaceae. Where species names are mentioned in the treatment, they must be considered as preliminary, since identification of herbarium specimens is difficult and nomenclature uncertain.
A. Bänziger, H. in Nat. Hist. Bull. Siam Soc. 39. 1991, B. Bänziger, H. in Nat. Hist. Bull. Siam Soc. 43. 1995


The foetid smell of the flowers of Rafflesia attracts carrion-flies of the genus Lucilia (Docters van Leeuwen 1929; Ross in Meijer 1985; Beaman et al. 1988; Bänziger 1991). The smell is produced by fresh flowers, especially during sunny warm periods of the day. Beaman et al. suggested they observed green flies in action in Rafflesia at the upper part of the gullies in the column leading to the anther cavities between rows of hairs (Meijer 1985). The actual act of pollination is well illustrated by Bänziger (1991), see . Still nobody has fully documented with a video camera what the flies do: laying eggs, collecting nectar (Haak 1885) or just using the carpet of ramentae as a mating ground (as insects do inside the odorous inflorescences of aroids: Croat, verbal comm.). The flies carry the pollen on their backs but have not been observed to use it as food. The actual source area (osmophore) of the bad odour is also still not yet known. Bänziger (1991) suggested from his own observations that the smell originates from the perigone lobes. Maybe the stomata described by Cammerloher (1920) have a function after all.

Douglas Warren Stevens noticed during his exploration of Nicaragua a faint smell from the flowers in Bdallophyton americanum (verbal comm.) and Rutherford (1966) reported this also from the masses of flowers of Pilostyles thurberi, both genera which like Cytinus and Mitrastemon also produce nectar. Honeyeater birds use flowers of Mitrastemon in New Guinea as nectar source (Beehler 1994).

Bänziger (1995: 352-356) gave interesting new data on solidification of the liquid pollen of Rhizanthes on the back of pollinating insects and reliquefying by profuse stigmatic fluid and viability over a period of 3 weeks. He also gave details with more recent names of the pollinating carrion flies now classified as Calliphoridae.
C. Beaman, R., et al., in Amer. J. Bot. 75. 1988, D. Beehler, B.M. in Biotropica 26. 1994, E. Bänziger, H. in Nat. Hist. Bull. Siam Soc. 39. 1991, F. Bänziger, H. in Nat. Hist. Bull. Siam Soc. 43. 1995, G. Cammerloher, H. in Oesterr. Bot. Zeitschr. 69. 1920, H. Docters van Leeuwen, W. J. in Trop. Natuur 18. 1929, I. Haak, J. in Weekbl. Pharmacie 3. 1885, J. Meijer, W. in Nat. Geogr. 168. 1985, K. Rutherford, R.J. - in PhD Thesis Claremont Grad. School, USA. 1966


Watanabe (1933, 1936) assumed long range seed dispersal of Mitrastema by birds. Observations of jungle walkers working for Meijer (1958, 1983) at the sites of Rafflesia in West Sumatra showed that ground squirrels and tree shrews like to eat the contents of the white pulpy ripe fruits of these plants. These observations have been confirmed now by Emmons, Jamili Nais and Ali Briun (1991) and documented with colour photographs from a hide erected in August 1989 near a fruit of Rafflesia keithii on Mt Kinabalu. The fruits of Rhizanthes also have white pulp, but they are hidden under the old perigones and predation has never been observed yet.

In analogy with other root and stem parasitic plants it is assumed that only disturbance and damage to the organs of the host can deliver the signal for seeds in contact with them to germinate (Bouman & Meijer 1994).


The vegetative parts are a kind of tissue mass often compared with fungal mycelia. The cells have remarkable large nuclei and form the so-called endophytic system within the stems and roots of the host plants, especially inside the living bark close to the cambium of the host. From there, special strings (so-called sinkers) can penetrate through rays of the host into the deeper lying xylem, while these cell strings inside the host cambium in general keep pace with the secondary growth of the host. In Pilostyles thurberi the endophytic body stays at a uniform distance from the apical meristem of the host (Rutherford 1970), in the Cytineae generally near the ground level and in the Rafflesieae in the roots as well as high on stems, though in general close to the ground. When the host cambium is reactivated, it may succeed to put a xylem layer on top of the parasite tissues. In general the parasite can migrate through the rays into the new host tissues, as described by Forstmeier et al. (1983) for Cytinus.

The simple cell strings do not contain vascular tissues; in many genera these occur mainly in cell cushions from which flowering structures originate, connecting ultimately with the staminal or ovary structures.

Solms-Laubach (1869) described the situation in Cytinus where the parasite forms a kind of tissue sleeve between cambium and xylem of the Cistus host root. In a rhythmic way, parts of the parasite tissue are covered by xylem layers, apparently and for unknown reasons turning the cambium activity off from time to time. The resulting more mature closed cylinder sleeve of the parasite consists of a cambium and two surrounding layers of tissue: an inner medullary and outer cortical plate. In the latter many irregular weak vascular bundles occur. Flowering structures originate inside swollen parenchyma tissues and cause gall-like growth above connections with the vascular bundles of the host plant, in general at least 2-3 years after the first infection of the host. The host tissues are filled with starch at this place except in a few cell layers close to the parasite tissue. These and similar situations were observed by Cartellieri (1926) in Rhizanthes, Brown (1912) in Rafflesia manillana, Haak (1889), Schaar (1898) and Hunziker (1920) in Rafflesia patma, Endriss (1902) in Pilostyles ingae, Rutherford (1970) in Pilostyles thurberi and Meijer & Behnke (unpubl.) in Rhizanthes, while the old classical study by Solms-Laubach (1875) is still of value. If the family is polyphyletic we are dealing here with remarkable convergences in growth morphology.

Kuijt et al. (1985) discovered sieve elements in the endophytic body of Pilostyles thurberi, apparently in a discontinuous system which they considered to be a vestigial cell type. In older studies such elements were in general overlooked.

Weak vascular tissues are reported from the endophytic bodies of species of Pilostyles. Floral tissue cushions contain rings of concentric vascular bundles in Rafflesia and Rhizanthes.

Most of the tissues are parenchymatous and these cells also never contain starch. In all genera the parenchymatous tissues are rich in tannins.

Stomata are known from scales of Pilostyles and Cytinus and also from Rafflesia and Rhizanthes (see Cammerloher 1920). Glandular hairs are known from the bracts of Cytinus and also in some species from the perigone. The function of the so-called ramenta, the outgrowths inside the flower tubes of species of Rafflesia and from the upper part of the diaphragm of Sapria is still unknown; their different forms have diagnostic taxonomic value (Winkler 1927). The osmophoric and nectarial structures of the family still need more comparative anatomical studies.
Only in the genus Bdallophyton do the flowering structures (inflorescences) have a well developed long axis. In all other genera they are more or less fascicles or very short uni-flowered shoots, surrounded by a few whorls of scaly dark brown or blackish leaves. In Rafflesia and Rhizanthes there are in general 3 whorls of 5 scales (bracts), in Pilostyles the whorls are 3-6-merous. Apodanthes has a regular structure of 2 outer bracts, 4 calyx-like bracts followed by 4 perigone lobes and an ovary with 4 placentas. Only in the tribus Cytineae inflorescences are developed like racemes, spikes or umbels, sometimes with bracts and bracteoles.

Flowers are in general unisexual. In Rhizanthes zippelii unisexual flowers occur besides bisexual ones. Bdallophyton oxylepis has bisexual flowers and B. americanum unisexual ones on different plants; Mitrastema is always bisexual and protandrous.

Perigones can be partly tubular at their bases and the lobes are imbricate (Rafflesia) or valvate (Rhizanthes). Rafflesia and Sapria possess as unique organ the diaphragma. Staminal structures always have the thecae interconnected in a ring like in Hydnoraceae, though never in phalanges, and are in various ways connected with the central column. In some cases in Pilostyles they occur in 2-4 rings. The detailed anatomy of the anthers in various genera shows also some variation (see ).
L. Brown, W. H. in Philipp. J. Sc., Bot. 7. 1912, M. Cammerloher, H. in Oesterr. Bot. Zeitschr. 69. 1920, N. Cartellieri, E. von in Bot. Archiv 14. 1926, O. Endriss, W. in Flora 91, Erg. Band. 1902, P. Forstmeier, L., F. Weberling & H.C. Weber in Beitr. Biol. Pflanzen 58. 1983, Q. Haak, J. - in Observ. Rafflesias. 1889, R. Hunziker, J. - in PhD Thesis Freiburg. 1920, S. Koorders, S.H. - in Bot. Overz. Raffles. Ned.-Indië. 1918, T. Kuijt, J., D. Bray & L.R. Olson in Canad. J. Bot. 63. 1985, U. Meijer, W. - Flow. PL, Dicot. in Families & Genera of Vascular Plants. 1993, V. Rutherford, R. J. in Aliso 7. 1970, W. Schaar, E in Sitzungsber. K. Akad. Wiss. Wien, Math.-Naturw. Kl. 107. 1898, X. Solms-Laubach, H. in Abh. Naturf. Ges. Halle 13, 3. 1875, Y. Solms-Laubach, H. in Bot. Zeitung 27. 1869, Z. Winkler, H. in Planta 4. 1927


Since Dumortier (1829) most systems of plant families of the world have treated Rafflesiaceae in a wide sense. Lindley (1836) and Richard (1838) considered Cytinaceae to be a separate family with Pilostyles as part of the Rafflesiaceae and the very closely related Apodanthus in the Cytinaceae. Van Tieghem (1890) elevated Apodanthaceae to the rank of family and Makino (1911) put his new genus Mitrastema in its own family. More recent accounts, as Thorne (1992), Beaman et al. (1992) and Meijer (1993) treat the family in a wide sense. Takhtajan et al. (1985) considered Rafflesiaceae sensu lato as a heterogeneous group, both macromorphologically and palynologically, and expressed the idea that the four tribes recognized by Harms (1935) deserve to be treated as separate families:
  • Rafflesiaceae: Rafflesia, Rhizanthes, Sapria
  • Apodanthaceae: Apodanthes, Pilostyles
  • Cytinaceae: Bdallophyton, Cytinus
  • Mitrastemaceae: Mitrastema.

Studies of ribosomal RNA of parasitic plants carried out by Dr Daniel Nickrent in Carbondale (University of South Illinois) are still preliminary but already support the removal of Cytinaceae from the Rafflesiaceae and ultimately may support the family system of Takhtajan et al. (1985). The ribosomal DNA data (Nickrent & Duff, in press) make it clear that the tribe Cytineae deserves to become a family at its own. Results on Apodanthes and Pilostyles are not yet available but it is quite possible also that they cannot easily be compared with the Rafflesia tribe (or family sensu stricto).

Solms-Laubach (1901: 7) warned not to rush to conclusions about the taxonomy and relationships of parasitic plants, which could hide their real ancestry in their strong reductions and adaptations required to survive as parasites. We can safely restrict ourselves here to the tribe Rafflesieae and the subfamily Mitrastemoideae and leave it to future research as to how closely or distantly they might be related.
AA. Beaman, R., et al. - in Proc. Int. Conf. For. Biol. & Cons. Borneo. 1992, AB. Brown, R. in Trans. Linn. Soc. London 19. 1844, AC. Dumortier, B.C. J. - in Anal. Fam. PL. 1829, AD. Harms, H. - in Nat. Pflanzenfam. 1935, AE. Lindley, J. - in Nat. Syst. Bot. 1836, AF. Makino, T. in Bot. Mag. Tokyo 25. 1911, AG. Meijer, W. - Rafflesiaceae in Families & Genera of Vascular Plants, Flow. PL, Dicot. 1993, AH. Nickrent, D.L. & R.J. Duff - in Proc. 6th Parasitic Weed Symp. April 1996, Cordoba, Spain. (in press), AI. Richard, A. - in Nouv. Elem. Bot. 1838, AJ. Solms-Laubach, H. - in Nat. Pflanzenfam. 1889, AK. Solms-Laubach, H. - in Pflanzenr. 1901, AL. Takhtajan, A.L., N. Meyer, R. & V.N. Kosenko in Bot. J. Leningrad 70. 1985, AM. Thorne, R.F. in Aliso 13. 1992, AN. Van Tieghem, P.E.L. - in Traité Bot. 1890, AO. in Ann. Sc. Nat. 2, 1. 1834


This mainly tropical family of parasitic plants is badly known at present from a chemical point of view (see Hegnauer 1973, 1990). Cytinus hypocistis yielded isoterchebin, an ellagitannin, as its yellow pigment and pelargonidin 3-galactoside and petunidin 3-gluco-side as red flower pigments. The sole crystalline compound isolated from acetone extracts of Psilostyles thurberi turned out to be sucrose. Pollination biology of Rafflesiaceae is still poorly known. Recent observations with Rafflesia pricei suggest that optical and olfactory mimicry may offer the clue. Flower pigments and odorous principles emanated from flowers attract carrion-flies; they seem not to be compensated for their pollination activities by nurture or suitable breeding sites. Most authors are convinced that Rafflesiaceae belong to or are affiliated with Polycarpicae sensu Wettstein. However, synthesis and accumulation of ellagitannins do not favour such an assumption, because polymeric proanthocyanidins (i.e. 'condensed tannins') are the characteristic tannin-like metabolites of Polycarpicae. In this respect Rafflesiaceae are similar to Nymphaeaceae s.str. which produce ellagitannins and possibly Cabombaceae which produce gallic acid and probably gallotannins. Therefore it is noteworthy that Takhtajan (1980) has Aristolochiales, Rafflesiales, Nymphaeales (Cabombaceae, Nymphaeaceae s.str. and Ceratophyllaceae) and Nelumbonales as numbers 5-8 in his subclass Magnoliidae, and that in Thome's (1992) classification Aristolochiaceae are incorporated in Magnolianae-Magnoliales-Magnoliidae and the superorders Nymphaeanae and Rafflesianae immediately follow Magnolianae. Nelumbonales and Ceratophyllales are treated by Thorne as orders of Magnolianae. Synthesis and accumulation of gallo- and ellagitannins may have evolved in a number of taxa now considered as outgroups of true polycarps.
1. R. Hegnauer


Field observations in 1981 and 1983 in the Ulu Gadut area near Padang, West Sumatra have shown that ripening of fruits of Rafflesia gadutensis takes about 8 months. Two fruits were monitored for periods of 6-7 months and the eldest had ripe seeds in February 1984. Meijer's guide Satar, who was monitoring this fruit below his hut, mentioned that it was visited by a 'tupai kuning' (most likely a tree shrew), the same species which was eating his young chickens. In Sumatra as well as Borneo we noticed male buds being opened by squirrels or/and tree shrews and it looks as if one shrew sitting on the exit of such a vandalized bud was captured on the lens on Kinabalu near Poring; see Attenborough (1995). It sits on a bud not at the entrance of a flower as the legend in the book asserts. The holes made are too small for most squirrels.

From a very obscure report by R.H.C.C. Scheffer in a letter to Solms-Laubach it might appear that actual pollination is not necessary for fruit set. Scheffer wrote to Solms (translated): "Only one of the buds made it into a flower, only 3 years after the inoculation. Notwithstanding this was a female one, and without having a male flower nearby it still developed seeds. These were used to infect another plant of Cissus scariosa and this was successful."

Solms asserts that Scheffer suspected that the rudimentary anthers might have produced some pollen. That is very unlikely. Meijer's unpublished field observations showed that unfertilized female flowers, several months after flowering, had produced no seeds with embryos. Ovaries with very little pollination may be stimulated to grow into fruits, but they can be full of seeds without embryos as observed from a fruit of Rafflesia keithii outside the Kinabalu National Park, August 1993 (seeds tested by F. Bouman, Amsterdam). Possibly there were other flowers of Rafflesia open in the garden or on Mount Salak about 120 years ago, to account for Scheffer's observations. Real parthenocarpy of Rafflesia has never been proven.

Ovule and seed development stages have been studied for Rafflesia and Rhizanthes (Solms-Laubach 1898; Ernst & Schmid 1913). Pollen tetrads form according to the suc-cedaneous system; the tapetum of anthers consists of 2 or 3 layers of cells. The stigma is in Rafflesia and Rhizanthes situated in a ring around the rim of the column disk, often more or less papillose; the ovary shows in these genera a maze of placental plates which fill the ovary cavity. For the development of the embryosac see Ernst & Schmid (1913) and Olah (1960). The outer integument is reduced. Ovaries grow into berry-like fruits, with whitish pulpa around the seeds. The seeds () are small and with hard scales . Rafflesia patma has a 2-cell suspensor and an embryo consisting of 3-5 layers of 2-4 cells, often only 6 cells in total. Endosperm is nuclear and consists of 30-40 cells in R. patma. Endosperm and embryo mostly contain oil, see also Schuerhoff (1926: 526). Mitrastema also has cellular endosperm, but ovules with only one integument.
AP. Attenborough, D. - in The private life of Plants. 1995, AQ. Bouman, F. & W. Meijer in PI. Syst. Evol. 193. 1994, AR. Ernst, A. & E. Schmid in Ann. Jard. Bot. Buitenzorg 12. 1913, AS. Harms, H. - in Nat. Pflanzenfam. 1935, AT. Olah, L.V. in Bull. Torrey Bot. Club 87. 1960, AU. Schuerhoff, P.N. - in Zytol. Bliitenprl. 1926, AV. Solms-Laubach, H. in Abh. Naturf. Ges. Halle 13, 3. 1875, AW. Solms-Laubach, H. in Ann. Jard. Bot. Buitenzorg 9. 1891, AX. Solms-Laubach, H. in Ann. Jard. Bot. Buitenzorg Suppl. 2. 1898


Hansen - in Fl. Camb., Laos & Vietnam. 1973
Meijer - in Kubitzki, Fam. & Gen. Vase. Pl. 2. 1993
Dell - in Fl. Austral. 1984
Hansen - in Fl. Thailand. 1972
Koord., Bot. Overz. Raffles. Ned.-Indië. 1918
Kiu & Wing - in Fl. Reipubl. Pop. Sin. 1988