A TAXONOMIC REVISION OF MALLOTUS SECTION PHILIPPINENSES (FORMER SECTION ROTTLERA – EUPHORBIACEAE) IN MALESIA AND THAILAND

A revision of Mallotus section Philippinenses (former section Rottlera ) in Malesia and Thailand is given. Descriptions, distribution maps, habit drawings, and a key to the species are provided. The diagnostic characters for the section are briefly discussed. Five species ( M. kongkandae, M. leptostachyus , M. pallidus , M. philippensis , and M. repandus ) are recognized. Mallotus chromocarpus is excluded from the section because it has more shared characters with the monospecific genus Octo- spermum , like the presence of indehiscent fruits, absence of stipules, marginal extrafloral nectaries on the upper side of the leaf blade, broad connectives (shaped umbrella-like), and its occurrence in New Guinea, and is therefore probably closely related to it.

The genus Mallotus was first described by De Loureiro (1790). He included only one species (Mallotus cochinchinensis), which is a synonym of Mallotus paniculatus (Lam.) Müll.Arg. Morphologically Mallotus can be recognized by the presence of stellate hairs (often in combination with simple hairs), coloured glandular hairs, extrafloral nectaries on the leaves, leaves that are alternate or opposite (then unequal in size), a lack of petals, undivided stigmas, and capsules (often armed with spines). The large number of species in Mallotus together with variable morphology has resulted in three main subgeneric classifications. The first was proposed by Müller Argoviensis (1865,1866), recognizing a total of five sections. However, he classified some species in different sections, which are now considered to be synonyms. This shows the large morphological variability within some of the species and also his unclear sectional delimitations. Pax & Hoffmann (1914) proposed a new subdivision into ten sections, which was later refined by Airy Shaw (1968). Airy Shaw subdivided the genus into eight sections, Axenfeldia, Hancea, Mallotus, Oliganthae, Polyadenii, Rottlera, Rottleropsis, and Stylanthus. A taxonomic revision of Malesian species of the separate sections Hancea, Polyadenii, and Stylanthus has been published by Bollendorff et al. (2000) and Slik & Van Welzen (2001a). Phylogenetic studies of Mallotus based on morphology have also been performed by the latter two authors (Slik & Van Welzen, 2001b). Their results show that sections Hancea, Oliganthae, Polyadenii, and possibly Mallotus are monophyletic, and that all the other sections are either paraphyletic (Rottlera, Stylanthus) or polyphyletic (Axenfeldia, Rottleropsis). However, their phylogenetic hypotheses are not stable, because of weak support for most branches and incomplete taxon sampling: only the sections that had previously been taxonomically revised include all the representative species, while the remaining sections include few taxa.
Although the sectional delimitations based on morphological characters are unsatisfactory, we will follow the traditional generic delimitation as circumscribed by Airy Shaw (1968) until the phylogenetic studies of the sections of Mallotus based on molecular and morphological data are completed.
Nomenclature follows the rules as accepted by the Botanical Congress of St. Louis (Greuter et al., 1999).

MALLOTUS SECTION PHILIPPINENSES VERSUS ROTTLERA
The section Philippinenses was previously named Rottlera. However, nomenclatural problems with this name prohibit its further use and the younger name Philippinenses Pax & K. Hoffm. should be preferred.
The name Rottlera was first used by Willdenow (1798) when he described the species Rottlera indica Willd. However, this combination is superfluous: Willdenow cited the validly published name Tetragastris ossea Gaertn. (Gaertner, 1790) in the synonymy, meaning, that he should have made the combination Rottlera ossea. Therefore, Rottlera is a homotypic synonym of Tetragastris, which belongs to Burseraceae. Willdenow also mentioned that M. cochinchinensis might belong to the same genus, but this does not imply that he should have referred his plant to the genus Mallotus, as Airy Shaw (1968) suggested.
Soon after, Roxburgh (1802) described another species: Rottlera tinctoria Roxb. In 1806 Willdenow synonymized Rottlera (incl. Tetragastris!) with Trewia and erroneously thought that the name became available again for a different use. Under modern rules this is not allowed and he is considered to have created an illegitimate, later heterotypic homonym. He attributed the authorship to Roxburgh (1802), which is also incorrect, as Roxburgh merely added a species to the first species of Rottlera. Rottlera Roxb. ex Willd. was subdivided into three sections by Reichenbach & Zollinger (1857). Of these, Rottlera sect. Pseudorottlera and sect. Stylanthus are valid, but Eurottlera is invalid (Art. 21.3).
Finally, Airy Shaw (1968) proposed a new sectional subdivision of Mallotus. He discussed the nomenclatural relationships between sect. Philippinenses and Rottlera, rejecting the first name because it would be an erroneous spelling, and because of the earlier Eurottlera Rchb.f. & Zoll. (Airy Shaw, 1968;391, 392). He, therefore, erroneously proposed sect. Rottlera "((Willd.) Rchb.f. & Zoll., corr.) Airy Shaw". He did not realize that the first argument cannot be sustained because a sectional name can be spelled in any way that the authors please, and the presence of a 'philippensis' does not require one to call the section 'philippenses'. Furthermore, under the present code illegitimate names do not automatically generate autonyms (Art. 22.5). Therefore, sect. Rottlera Airy Shaw is a superfluous name for sect. Philippinenses.

Mallotus leptostachyus
Habitat & Ecology -Locally common in understorey of primary to secondary forests, scrub, mostly found on disturbed sites; on ridges, forest edges, road and river sides, steep slopes, marshy and savannah areas; in wet (riverine, swampy) to welldrained terrains; on a large variety of soil types, like granite, limestone, sandstone, sandy clay, sandy loam soil, volcanic rock, gravel, quartz, shale, and rock. Also cultivated (Hawaii and Miami, USA). Altitude: sea level up to 1600 m. Flowering and fruiting the whole year through.  Uses -Ornamental (red fruits). The wood is used for rafters, tool handles, matchboxes, and house-posts. The fruits and bark are used as an antihelminthic, to relieve constipation, and against cutaneous affections. The fruits and leaves are used against colds and to cure stings and bites of snakes and other poisonous animals. The leaves are used as a fodder. The glands of the fruits are used as a red dye. The roots are used for dissolving coagulated blood and contusions. The oil of the seeds is used as a substitute for tung oil (Vernicia Lour., Euphorbiaceae) in the formulation of rapiddrying paints, varnishes, hair fixers and ointments. Partly after Ambasta (1986) and Manilal (2003).
2. Croton punctatus and C. coccineus were based on different specimens in different herbaria, both with the same annotations of J. König. The name C. coccineus, being based on a different specimen, is not superfluous.
3. Under the number Wallich Numer. List 7772A there are two very distinct species as can be seen on the IDC microfiche nr. 7394 of the Wallich herbarium in K. The specimen found on the left hand base corner is M. philippensis. The one found on the right hand base corner is M. distans as it is generally known based on Müller Argoviense's description of the fruits (densely hairy mixed with punctuate glandular hairs).
4. The collections Ludwigs 247 and 465 (M) have on the label as locality 'Kamerun: Victoria', but it is likely that this is a mistake or the plants were cultivated exotics.
Distribution -Endemic to New Guinea.
Habitat & Ecology -Locally common in the canopy of young to old secondary forest, or on the grassland-edge of swamp-forest