The Cycad Pages
Cycas rumphii

Cycas rumphii Miq., Bull. Sci. Phys. Nat. NĂ©erl. 2: 45 (1839).
"TYPE: the illustration: Rumphius 1749: t. 23. (lecto, fide Hill)."

Etymology: Honouring German-born Dutch naturalist Rumphius (Georg Eberhard Rumpf, 1628-1702), military officer with the Dutch East India Company in Ambon, 1652-1657, then with the civil merchant service of the Dutch East India Company.

Literature: Stapf 1916, Backer 1925, Ochse 1931, Schuster 1932, Kanehira 1938, Backer & Bakhuizen van den Brink 1963, Peekel 1984, Hill 1994c.

Illustrations: Miquel 1852, Warburg 1900, Schuster 1932, Kanehira 1938,Peekel 1984, Hill 1994.

Historical notes: Although first legitimately described in 1839 by Dutch botanist Miquel, the existence of a distinct Malesian taxon was first recognised by Roxburgh (1832: 747), on the basis of plants in cultivation in the Calcutta Botanic Gardens. He recognised that two taxa were present, treating one as C. circinalis and describing the other as a new species C. spherica. Roxburgh had, however, mistakenly applied the new name C. spherica to plants belonging to the C. circinalis group. The distinct Malesian taxon he then treated under the name C. circinalis, although he had also mixed material of the two taxa under each description.

This species as recognised by Miquel (1839: 45) was based on part of Olus Calappoides of Rumphius (1741). Miquel was never clear on the specific limits of this taxon, at first separating material from Sulawesi as C. celebica (1839: 45), and later (1868: 232), combining the two. He also separated material from Timor as C. rumphii var. timorensis (1841: 125), and from Java as C. circinalis var. javana (1842: 28). In addition, he at first recognised Roxburgh's C. spherica (1843: 693), apparently on the basis of Roxburgh's published account and without realising the confusion with C. circinalis. He later ( 1851: 32) correctly noted Roxburgh's confusion, and placed C. spherica in the synonymy of C. circinalis, although later still (1868: 230), he again recognised C. spherica at specific rank. At the same time (1868: 232), he placed C. celebica and C. circinalis var. javana in C. rumphii, with no mention of C. rumphii var. timorensis.

Recognition of C. rumphii as a distinct species has been widely argued, with early discussion very ably summarised by Stapf (1916). Schuster (1932: 74) provided an essentially correct treatment of this species (with the exception of the varieties and subspecies, and the New Caledonian material cited), in contrast to his confused treatment of most other species. The name has been generally accepted subsequently, although misapplications of the name C. circinalis to this and related taxa continue (eg. Jones 1993).

Warburg also recognised the presence of a coastal taxon in New Guinea, assigning it to C. rumphii. This coastal taxon in New Guinea has also been assigned to C. circinalis (White 1922).

Distinguishing features: Distinguished within the group of species with spongy encocarp by the broad, falcate, hard, glossy leaflets with relatively broad bases, present but discontinuous laminar hypodermis, the relatively long and usually wholly spinescent petiole, and the narrowly triangular megasporophyll lamina with a slender apical spine (10-23 mm long) and vestigial lateral spines. Adaxial mesophyll is usually continuous across the moderately broad and rounded midrib, but sometimes interrupted by the midrib in the east of the range. This may be due to genetic admixture of C. bougainvilleana from further to the east. The condition of reduced lateral spines on megasporophylls also occurs in related taxa from Philippines, Malaysia and the Indian Ocean. These taxa are, however, distinguished as a group by the lack of the apical wing on the seed.

Rumphius in one of the type illustrations showed (somewhat stylistically) rather distinct lateral teeth on the megasporophyll that do not accord with other collections from the type region, and more match C. scratchleyana, also known from the Moluccan region (see above). For this reason, the other Rumphius illustration hes been selected as lectotype.

Distribution and habitat: Cycas rumphii has been poorly understood in the past. Recent recognition of a number of related species has allowed clarification of its identity, and clear delineation of its distribution. As now understood, C. rumphii has a distribution centred on the Moluccan island group (Maluku, or the Spice Islands) extending east into Indonesian Papua and a short way along the north coast of Papua New Guinea, and north to Sulawesi. In the west, it appears to extend to southern Bornea and north-eastern Java.

C. rumphii appears to share the ecological preferences of several other taxa in this group, being largely a species of closed woodland or forest on more or less calcareous substrates in near-shore environments.

This species is relatively common in cultivation in Fiji and, to a lesser extent, Vanuatu. In both cases, female plants only are known, all of which have been vegetatively propagated from stem offsets. The cultivated plants in both countries show distinctively yellow new growth that is not typical of the species, which suggests that they are all of the one clone (see Hill 1994c). These plants are apparently not interfertile with the native cycad in the region (C. seemannii), as evidenced by the lack of seed-set in plants growing with fertile male and female plants of C. seemannii. Their close relationship would suggest that they are unlikely to be truly genetically incompatible, and further study is required to evaluate possible reasons for the lack of hybridisation. Temporal separation is unlikely, both species having been observed to cone at almost any time of year. Pollen vectors are not known conclusively, and both wind and insect pollination have been suggested (see Pellmyr et al. 1991). The strong odour reported from both the male and female cones in the C. rumphii group has been regarded as both an insect attractant (Tang 1987). and deterrent (Pellmyr et al. 1991). A considerable body of evidence is now available to support the theory that most cycads are insect pollinated (Donaldson et al. 1993); Norstog and Fawcett 1987), and slight chemical differences could mean vector attraction to one species and not the other.

Conservation status: Unknown. The 1997 IUCN Red List of Threatened Plants lists C. celebica as category R, although very little is known of this or in fact most occurrences in this region.

Vernacular: Ftoi (Weda language, Halmahera), nufuès (Biak dialect, Amdoei village, Irian Jaya), sumo (Wapi language, Yakoi village).


Stems arborescent, to 3(-10) m tall, 11-20 cm diam. at narrowest point.

Leaves bright green, highly glossy, 150-250 cm long, flat (not keeled) in section (opposing leaflets inserted at 180° on rachis), with 150-200 leaflets, with orange tomentum shedding as leaf expands; rachis consistently terminated by paired leaflets. Petiole 35-60 cm long (20-30% of total leaf), petiole glabrous, spinescent for 80-100% of length. Basal leaflets not gradually reducing to spines, 190 mm long.

Median leaflets simple, strongly discolorous, 220-320 mm long, 12-16 mm wide, inserted at 70-85° to rachis, decurrent for 5-8 mm, narrowed to 4.5-7 mm at base (to 35-50% of maximum width), 15-19 mm apart on rachis; median leaflets section flat; margins slightly recurved; apex acute, not spinescent; midrib flat above, raised below.

Cataphylls narrowly triangular, soft, pilose.

Pollen cones fusiform, yellow to brown (pale); microsporophyll lamina firm, dorsiventrally thickened, apical spine rudimentary, sharply upturned, 2-5 mm long.

Megasporophylls 18-32 cm long, white-tomentose or yellow-tomentose; ovules 6, glabrous; lamina lanceolate, 50-75 mm long, 25-35 mm wide, obscurely dentate, with 12 soft lateral spines 0-1 mm long, 0-2 mm wide, apical spine distinct from lateral spines, 15-25 mm long.

Seeds flattened-ovoid, 45 mm long, 30 mm wide; sarcotesta orange-brown, not pruinose; fibrous layer absent; sclerotesta apically crested. Spongy endocarp present.

The Cycad Pages

© 1998-2012 Royal Botanic Gardens Sy dney
Written and maintained by Ken Hill 1998-2010
Maintained by Leonie Stanberg and Dennis Stevenson 2010-2012
This site is currently not being maintained