Dipterocarpaceae

Primary tabs

Dipterocarpaceae

Description

See , , , , , , , , , , . The bole can be circular in cross-section, or fluted or ribbed, and is usually buttressed. See , , , . The buttresses can be small or large, thin with a sharp edge as in Shorea sect. Shorea generally, or thick with a rounded edge; the edge, from the soil surface to the apex where the buttresses merge into the bole, is usually concave, but is sometimes straight or convex; the buttresses can terminate more or less abruptly at apex and base, or continue as ribs up the bole as in Anisoptera sect. Anisoptera or superficial roots over the ground. Flying buttresses are convex-edged buttresses originating from the bole above the soil surface, and are differentiated from the stilt roots by being flat, therefore elongate in cross-section, as opposed to being terete or spherical in cross-section; stilt roots can be adventitious to the buttresses or arise from the bole. Flying buttresses and stilt roots are a particular feature of Hopea, and are found at least occasionally in a majority of species with the exception of those in sect, subsect. Hopea. , .

In the seasonal parts of S.E. Asia (Thailand, etc.) various species of Dipterocarpaceae have a capacity for suckering and produce new stems from trunk-bases after cutting. However, the Malesian rain-forest species generally lack this capacity and regenerate merely from seed.

The mode of branching, and the leaf arrangement, changes ontogenetically. The seedling dipterocarp, after production of one or more pairs of opposite leaves, sends up a stem with spirally arranged leaves. From the axils of these leaves arise the lateral branches; the leaves on the branches are generally arranged distichously, consequently the branchlets are also alternately arranged and the branches occupy one plane (plagiotropy), ascending, descending, or horizontal. The sapling leader, or stem apex, grows continually at first, but soon adopts resting periods, growth proceeding in flushes that are not necessarily seasonal at this stage though often synchronised among members of a clump. The internodes are frequently longer towards the beginning of the growth flush than towards the end, so that the stem leaves in some species, particularly Hopea sect. Dryobalanoides, are largely bunched at the end of each growth stage of the leader; consequently the branches appear to arise in whorls and the sapling and pole sized tree assumes a 'pagoda shape' (Corner, 1940); each flush of growth by the leader is coordinated with the sprouting of some of the axillary buds of the previous flush to form new lateral branches.

Young dipterocarps therefore conform either to Massart's model, that is, with rhythmic growth and branching, or approach Roux's, with continuous growth (Halle, Oldeman & Tomlinson, 1978).

The leaders have spiral leaf arrangement; if a leader dies or its growth is otherwise arrested at this stage, new leaders emerge not from axillary but from tiny subsidiary buds (Ng, 1976); whether a single leader eventually achieves dominance over others varies with the species.

If the tree remains small, and never reaches the forest canopy, as in many Hopea species, it frequently remains monopodial with this form of branching, and the crown remains lanceolate or conical. , . If the leader reaches the canopy the subsequent lateral branching generally become orthotropic, apical dominance is lost, the earlier plagiotropic branches die and are lost and a hemispherical or dome-shaped crown is formed. .

Alternatively, plagiotropic branches become orthotropic towards their tips (Halle, in pass.). Many Vatica species, however, which never become tall enough to reach the forest canopy, become more or less sympodially branched and develop an irregular oblong crown; while Parashorea macrophylla, and some members of the Shorea sections Mutica and Anthoshorea remain monopodial for some time after they have emerged above the canopy and reached full height, and may even send up further vertical leaders, apparently from axillary buds from the plagiotropic side-branches. In most species with an emergent sympodial crown the leaves are spirally arranged on the branchlets, but in species that tend to branch low, such as the riverside species, those species in which some or all the branches become horizontal or pendant towards the apices, and the species with compressed twigs, the leaf arrangement is frequently or always alternate. If the branches are many and radiate from the bole apex, as in many emergent Hopea species, the crown of emergent species takes on an evenly hemispherical appearance. If the main branches are few and large but much branched towards the apices, the branchlets become bunched towards the ends of the large branches into more or less hemispherical groups and the crown assumes the appearance of a cauliflower head. If the twigs are stout and the leaves large the crown shape is uneven and the leaves tend not to be confined to the perimeter of the crown. Rarely, the branches become pendent at the apices and the crown becomes 'weeping' (e.g. Shorea inaequilateralis, S. quadrinervis). .

According to Halle, Oldeman & Tomlinson (1978) who have classified tree architecture into a series of 'models', trees do not change from one model to another during life. They placed the species of Dipterocarpus and Shorea they examined in Massart's model in which the trunk is orthotropic, the branches plagiotropic and in false whorls through rhythmic growth, and the position of flowers various; and in Roux's, which differs from Massart's in that trunk growth is continuous. From the foregoing description it would appear that the model may be mutable in dipterocarps, for growth eventually always becomes rhythmic though it may not start thus, and because in most emergent species successive branches become increasingly orthotropic with a change from alternate to spiral leaf arrangement. This change in branching pattern would not imply a change in model were such orthotropic branches arising through reiteration, that is from supra-axillary buds following damage to the leader; but this does not seem to be the case.
Small or large resinous usually evergreen trees, usually buttressed, and often (if large trees) with flaky or fissured bark. , , , , , , , , , . Thorenaar (1926) first systematically examined bark morphology in the family; Symington (1943) used bark characters extensively for field diagnosis. Whitmore (1962) described how weathering processes and tangential strain during growth together act in conjunction with the growth of secondary phloem, expansion tissue developed from phloem rays, phloem proliferation tissue developed from the phloem parenchyma, and the periderms (each of which are laid down in several ways), to produce surface features distinctive of both species and higher taxa; these features change ontogenetically from an initial smooth surface. He thus rationalised bark description, which essentially comprises the surface pattern (the visual summation of surface configurations) and the slash appearance (the visual summation of the appearance of the inner and outer bark in oblique transverse section). Various degrees and types of Assuring develop through tangential strain, and of flaking on account of the disposition of the periderms. In a few taxa (e.g. most Vatica, and Shorea sect. Pachycarpae) the bark remains smooth; in others (e.g. Anisoptera, Dipterocarpus, Shorea sect. Shorea and sect. Richetioides) it almost always becomes distinctively flaky, while in yet others (e.g. Shorea subsect. Mutica) it becomes distinctively fissured. In one known instance (fissured-bark Hopeas) a distinctive bark configuration has no apparent taxonomic significance above the species level, and is not indeed even consistently developed within the species. Seven distinct bark types, in three groups differing in the amount of expansion tissue, were recognised; within each a number of categories exhibiting lesser structural differences were defined as bark manifestations; it is these that characterise taxa at generic and sectional level. In this account bark characters are summarised in generic and sectional descriptions where appropriate, but have been omitted from species descriptions; reliable bark descriptions of species can be found by reference to Symington and Whitmore and also Ashton (1964, 1968). In particular they contribute important definitive characters for the sections of Shorea. Resin exudations from wounds in living bark and wood also differ in frequency, rate and mode of crystallisation, translucency and colour and sometimes form useful subsidiary characters for identification.
Some or most parts with a tomentum of fascicled hairs, or sometimes single hairs, unicellular or multicellular glandular hairs, or multicellular, short or long lobed or peltate hairs. Leaves alternate, simple, margin entire or sinuate, not crenate, terminating +- abruptly at the ± prominent geniculate petiole, penninerved (in Dryobalanops and some Hopea nerves co, dense and slender), often with domatia in axils between nerves and midrib or along midrib and (rarely) nerves; Stipules paired, large or small, persistent or fugaceous, leaving small to amplexicaul scars. Inflorescence paniculate, racemose, rarely cymose, ± regularly, rarely irregularly, branched, terminal or axillary; The centrifugal stamens originate from a number of common bundles independent from the gynoecium. . The gynoecium bundles break away from the common supply with the stamens before the stamen supply begins to branch. In all genera but Upuna and some Stemonoporus (Monoporandra) of Ceylon the number of primary stamen bundles is 10. In Upuna they are independent of the perianth supply and continue between the perianth bundles in the pedicel; in other genera the 10 stamen bundles are associated with the 10 perianth members, though in some Shorea (e.g. S. geniculata) they appear not to unite within the length of the receptacle and pedicel. In all species where the number of stamens is more than 10 the stamen fascicles originating from the calyx supply are larger than those originating from the petals. In Shorea, Hopea, Vatica and Cotylelobium the petal and sepal bundles, together with the staminal bundles originating from them, are independent. Upuna, Anisoptera, Vateria, Stemonoporus and Dipterocarpus are alike in that a single lateral branch arises constantly from the left or right of each sepal bundle which supplies the petals; the staminal bundles from the sepal supply originate above this bifurcation.

In a recent paper C. Woon & H. Keng (1979) gave details with very numerous figures of the stamens in Dipterocarpaceae.

Flowers are usually small, except in Dipterocarpus and Vateria. The shape and size of the bud and the size and shape of the perianth members is diagnostic for some sections of Shorea. The number of stamens in species with less than 20 is constant in normal flowers; in those with more than 20 the number varies about a mean, but provides a useful specific diagnostic character. In Shorea the shape of the filament is an important section character, and in many genera the shape of the connectival appendage is diagnostic. The presence or absence of a stylopodium, the shape of the ovary and stylopodium and the size of the style relative to that of the ovary are often diagnostic for sections and sometimes species, particularly in Shorea and Hopea.
Flowers secund or distichous, bisexual, actinomorphic, scented, nodding. Stamens 5-110, 1-3 verticillate or irregular, hypogynous or subperigynous, centrifugal; The number of dipterocarp stamens may vary from 5 to c. 105. The most widespread number is 15; 10 occur in some Hopea, some species of Shorea sect. Richetioides and sometimes in abnormal flowers in sect. Mutica. The two Stemonoporus species formerly in Monoporandra, and a single Bornean species of Vatica, possess 5 only. Genera with many stamens as a characteristic are Pakaraimaea, Marquesia, Monotes, Vateria, Upuna, Dryobalanops and Dipterocarpus; it is usual also in the type section of Anisoptera, and sect. Anthoshorea, Ovalis and Shorea of Shorea. Phytogeographical evidence within Asia lends support to the suggestion that the primitive type had numerous stamens, and that genera with 15, 10 or 5 stamens reached these numbers by reduction. Vateria occurs in southern India, Ceylon, and the Seychelles whereas Stemonoporus, with 5 or 15, is endemic to the wet zone of Ceylon, Anisoptera sect. Glabrae, with 15 stamens, ranges from Burma and Indo-China through Malaya to Borneo, whereas sect. Anisoptera with many stamens occurs from Burma and Indo-China through Malaya and Sumatra, Borneo, the Philippines, and across Wallace's line to New Guinea. Within Shorea, though the monotypic and probably advanced sect. Ovalis, with many stamens, occurs in Malaya, Sumatra and Borneo, the large sect. Shorea and Anthoshorea (with 15-∞ stamens) are considerably more widespread than the equally large sect. Mutica, Richetioides and Brachypterae (typically with 10-15 stamens with the exception of two species). The former occur east of Borneo and the Philippines, and through India to Ceylon; the latter are confined to the everwet region of western Malesia with the exception of one species in the Moluccas. In this case a clear pattern emerges of a reduction of staminal number among local endemic species of the everwet zone, especially in the geologically recent region of lowland Borneo. The widespread species, often of the seasonal tropics and including the American and African subfamilies, retain the primitive condition. This pattern is most clearly seen within those groups in which staminal number is variable.
Ovary superior or semi-inferior, 3-, rarely 2-, locular; Ovules 2(-3) in each loculus, axile, pendulous or laterally anatropous, bitegmatic with ventral raphe and superior micropyle. Fruit indehiscent, 1-seeded;

Distribution

Africa present: Seychelles (Seychelles presentpresent), Asia-Temperate: China South-Central (Yunnan present); Hainan (Hainan present), Asia-Tropical: Bangladesh (Bangladesh present); Borneo presentpresent; Malaya presentpresent; Maluku (Maluku present); New Guinea present; Philippines (Philippines present); Sulawesi (Sulawesi present); Sumatera (Sumatera present); Thailand (Thailand present), Batan Is. north of Luzon, Philippines present, Burma present, Ceylon present present present, D'Entrecasteaux Is. off S.E. Papua present, East India present, East Malesia present, Kwangsi present, Madagascar present, New Britain absent, New Ireland absent, Pantropical present, Pattani adjoining Kelantan in N. Malaya present, S. Kwantung present, Southern India present, Sunda shelf present, Sunda shelf islands present, continental Asia present, continental S. China present, former British Guyana present, humid non-seasonal areas present, seasonal area encompassing the Lesser Sunda Is. east of Sumbawa as far as the Tenimber Is absent, south of Peninsular India present
The newly described monotypic genus Pakaraimaea Maguire & Ashton (1977), locally found in the south of former British Guyana, makes the family pantropical, confined to the lowlands and hills of the tropics below 1800 m. . This genus represents a distinct subfamily Pakaraimoideae.

The second subfamily, Monotoideae is represented in Africa and Madagascar, with some 36 spp. of Monotes A.DC. and a few species of Marquesia Gilg (cf. Bancroft, 1935).

Subfamily Dipterocarpoideae, comprising 13 genera and some 470 spp. ranges from the Seychelles through Ceylon (where a proportionally large diversification exists) to the south of Peninsular India, and then to East India, Bangladesh, Burma, Thailand, Indo-China, to continental S. China (Yunnan, Kwangsi, S. Kwantung, Hainan) and through Malesia southeastwards to the D'Entrecasteaux Is. off S.E. Papua (not in New Britain and New Ireland), and northwards to the Batan Is. north of Luzon, Philippines.

Fossils do not significantly extend subfamilial range in Asia and Malesia, but they do essentially so in East Africa. .

In Malesia 10 genera with 386 spp. occur, predominantly in the humid non-seasonal areas, absent only from the seasonal area encompassing the Lesser Sunda Is. east of Sumbawa as far as the Tenimber Is.

The local species diversity of these genera is very uneven, with a tendency to decline eastwards, as is shown by the total number of species per island. . See also , , , , .

Three of the 10 genera are endemic in Malesia, viz the monotypic genus Upuna in Borneo, Neobalanocarpus in Malaya (&Pattani adjoining Kelantan in N. Malaya), and the genus Dryobalanops (7 spp.) on the Sunda shelf (Sumatra, Borneo, Malaya); the 7 others Malesia shares with continental Asia, and Ceylon (except Anisoptera and Parashorea). A further three are endemic to Southern India, Ceylon and the Seychelles.

Four genera range widely through Malesia and also have species (mostly few) in East Malesia (Celebes, Moluccas, and New Guinea), viz Anisoptera (11 spp., 10 in Malesia), Vatica (65 spp., 55 in Malesia), Hopea (102 spp., 84 in Malesia), and Shorea (194 spp., 163 in Malesia).

Of the remaining three Cotylelobium (5 spp., 3 in Malesia) is known in Malesia only from the Sunda shelf islands, while Dipterocarpus (69 spp., 53 or 54 in Malesia) and Parashorea (14 spp., 10 in Malesia) occur on the Sunda shelf islands, but also in the Philippines.

Some of the Malesian genera formerly had in the Tertiary a wider distribution, e.g. Dryobalanops occurred in West Java and Southern India, Dipterocarpus in N.E. Africa, and Anisoptera (now only from Chittagong and Burma southeastwards) in India.

It is noteworthy that there are hardly any clear disjunctions in the generic ranges (apart of course from seas separating adjacent islands), the exception being that of Cotylelobium, with 1 sp. in Ceylon and further from S. Thailand to West Malesia which stems obviously from extinction. Valeria and its close ally Vateriopsis are confined to Ceylon and the Seychelles respectively; this huge oversea gap must be ascribed to ancient geomorphological processes.

Ecology


Taxonomy

Historical review. The first mention of the family of great trees that dominate the lowland and hill-forests of the Far East in European literature is in the diary of Marco Polo, who recorded visiting Fansur in Sumatra, considered to be the present Baros on the west coast, where the camphor tapped from the hinterland was at that time literally worth its weight in gold. Indeed, up till the mid-eighteenth century it was only the Sumatran camphor which attracted the attention of European travellers. It is mentioned in Dutch literature several times in the mid and late 17th and early 18th century. Rumphius (1755) described the Arbor Camphorifera II occidentalis and also mentioned an Arbor Koring (1741), which Merrill (1917) regards as Dipterocarpus hasseltii Bl. (the name keruing and variations of it being the Malay and Indonesian generic name); Merrill's evidence is meagre however. The camphor tree receives further mention from Charles Miller (1778) who sent a specimen to Banks from Sumatra.

Meanwhile Linnaeus had received material from India of a plant that he described as early as 1737 as the genus Vateria, and in the 1st edition of Species Plantarum (1753) and 5th edition of Genera Plantarum (1754) as Vateria indica. This he placed in his Class Polyandria Monogynia between Mesua and Thea, and shortly after Microcos, Tilia, and Elaeocarpus. In 1771 he described Vatica in Mantissa Plantarum, placing it in Dodecandria Monogynia with Befaria Mutis. De Jussieu (1789) placed Vatica and Vateria under 'Genera alternifolia, hinc Guttiferis, inde Aurantiis affinis', with two other genera of undecided affinity — Allophyllus (now in Sapindaceae), and Elaeocarpus. The two then known genera were thus brought together for the first time. In 1824 A. P. de Candolle placed Vatica under Tiliaceae, but omitted mention of Vateria.

In 1825 Sprengel included Shorea Gaertn. and Dipterocarpus Gaertn. under Polyandria Monogynia, citing Dipterocarpus in the Tiliaceae. Vatica and Hopea Roxb. he included in Dodecandria Monogynia following Linnaeus. In 1828 Reichenbach placed the then known genera in his Laurineae d. Pterigiae, adopting the latter name after Correa's genus Pterigium (1806), in which had been included Dipterocarpus and Dryobalanops Gaertn. described the year previously.

In 1825 however Blume had created the family Dipterocarpeae, stating that it bore affinities to the Tiliaceae in the contorted corolla, and to the Guttiferae in the resin ducts, superior ovary, many stamens, and single exalbuminous seed.

Lindley (1836) put the 'Dipterocarpeae' with Sterculiaceae, Malvaceae, Elaeocarpaceae, Tiliaceae and Lythraceae into his Alliance Malvales, commenting that Blume had noticed affinities with Guttiferae. So did Meisner (1837) who placed Dipterocarpaceae next to Sterculiaceae, Tiliaceae and the Madagascan Sarcolaenaceae (Chlaenaceae).

Endlicher (1840) on the contrary had put the family with the Class Guttiferae, with Chlaenaceae, Ternstroemiaceae, Clusiaceae, Marcgraviaceae, Elatineae, Reaumuriaceae and Tamaricaceae, thus far separated from Tiliaceae. He also erected the African genus Lophira Banks ex Gaertn. (now Ochnaceae), which Guillemin (1830) had considered a dipterocarp, into an order of its own, and reduced Shorea Roxb. ex. Gaertn. to Vatica L. This classification was subsequently followed by Lindley (1846). Bentham & Hooker (1862), besides accepting Guillemin's conclusions on Lophira, included also Ancistrocladus Wall. Planchon (1849) had previously put both genera in a group of their own allied to the Ochnaceae. Bentham & Hooker resurrected Shorea, and maintained the family in their Cohors Guttiferales, though stating its affinities to be with the Tiliaceae as well as the Ternstroemiaceae.

The first complete monograph of the family appeared in A. de Candolle's Prodromus (1868); he enumerated 126 species in 13 genera including Monotes; 27 years earlier Korthals (1841) had estimated the total known species at only 34. De Candolle again placed Ancistrocladus and Lophira in separate families; he further described the first known African dipterocarp as Monotes africanus, indicating by its name that the genus occupied an isolated place in the family. He claimed the affinities of Dipterocarpaceae to be with Chlaenaceae and Ternstroemiaceae, with Lophiraceae and Ancistrocladaceae as intermediate groups.

In 1874 Dyer monographed the genera Dipterocarpus and Dryobalanops. Burck (1887) made a study of East Indian dipterocarps based in large part on anatomical characters; he created no new genera but united the genera Pentacme DC., Monoporandra Thw. and Stemonoporus Thw. with Vateria and transferred part of the genus Hopea Roxb. to Doona Thw.

Heim's 'Recherches sur les Dipterocarpacees' (1892) remains to this day the most detailed study of the whole family. Though he frequently made anatomical studies of leaf, petiole, twig and fruit, he gave particular weight to the characters of the stamens, the embryo and to the 'caracteristique' (the arrangement of the vascular bundles as seen in transverse section in the petiole at the umbo). His system suffered because he was working at a time when herbarium collections were quite inadequate in quality and in numbers for his task. The result was unfortunate; whereas Burck five years previously had recognised only 10 genera, Heim maintained 30, in 8 series and 2 subseries; of these 13 genera were new though based altogether on only 17 species, of which 11 were described from single herbarium sheets. The genus Cotylelobiopsis Heim, for instance, was described from a single sterile sheet in which the anatomy of the petiole was found to be unique in the family; the specimen, which is lost, appears to represent fallen leaflets of Pseudosindora palustris Sym. in the Leguminosae. Cotylelobium melanoxylon was represented under three binomials, each based on a single specimen. Of the 4 species recognised by Heim in his genus Richetia, Symington (1933: 153) later correctly reduced 3 to a single species already described by Burck. The 4 genera united under Vatica by Burck were redivided and placed in 2 series; Monotes DC. was removed to the Tiliaceae.

Heim promised a more complete monograph at a later date, but in 1895 the Dipterocarpaceae were treated by Brandis and Gilg for the Pflanzenfamilien; later in the same year Brandis also published an 'Enumeration of the Dipterocarpaceae.', based on the specimens at Kew and the British Museum. Monotes was reunited with the Dipterocarpaceae, while none of Heim's genera were accepted and few of his species. They maintained but 16 genera, being a return to de Candolle's generic concept, with the addition of Balanocarpus, Cotylelobium, Parashorea and Isoptera, described subsequently to 1868; the reduction of Petalandra Hassk. to Hopea; and the maintenance of Stemonoporus separate from Vatica, under which name it had been reduced by de Candolle. They recognised 5 tribes in Dipterocarpaceae sens. str. following Brandis. Later Gilg (1899) placed the African dipterocarps in a separate subfamily. His account in the 2nd edition of the Pflanzenfamilien (1925), which is the most recent of the whole family, is with this exception mainly a reprint of the 1895 account. In 1941 Symington described the genus Upuna; this genus necessitates a redefinition of Brandis's tribes, and this has been discussed by me (1978). Two tribes are now recognised in Dipterocarpaceae on the basis of the calyx, resin canal distribution and basic chromosome number. One includes Brandis's tribe Shoreae and Dryobalanops, while the remaining genera are contained in the other.

Notes

1. The herbarium identification of dipterocarps must remain difficult, mainly for the reason that besides leaves also fruit and flowers are required.

The generic key is largely based on sterile characters and those of fruit; I have added an additional key to field groups which may be useful especially in the field.

Species of the genera Cotylelobium, Dipterocarpus, Dryobalanops, Parashorea and Vatica are also largely keyed out by means of sterile characters together with fruit. So are the species of the large genus Hopea, but in this key there are several leads, mostly for a few species, in which flowers are required.

In the largest genus, Shorea, leaf characters alone are only diagnostic at sectional level for Richetioides. Though bark and leaf characters together can provide a basis for keying out the vast majority of Malesian species, sections — and hence species — are impossible to key reliably without the valuable sectional characters provided by the former. This is because the combination of ontogenetic and phenotypic variability in leaf characters of these emergent trees, combined with the great number of species, makes it impossible to construct a key on this basis though the experienced taxonomist can still identify most by careful matching with named material, based on his knowledge of the intricate combination of subtle features by which the leaf of each species may be diagnosed at least when mature. In this genus flowers provide invaluable sectional, and in some sections species characters, but fruit are only of limited value particularly for distinguishing the species with short fruit sepals from the rest.

I do not believe, therefore, that the dichotomous key provides a practical means of identifying sterile material in the larger genera.

In the forest it is a different matter, as Symington so excellently demonstrated. His, and all subsequent, keys to sterile material include the field characters of bark and wood. Such keys are practicable mainly on a provincial basis; they already exist for West and East Malaysia (Symington, 1943; Meijer & Wood, 1964; Ashton, 1968); all but a handful of Bornean species are included in the latter two. They are not appropriate to a regional monograph, and this account therefore aims to provide the sound taxonomic base upon which forest botanists can build.

The species of Anisoptera and Shorea are here therefore mainly keyed out on sterile characters and flowers, though in Shorea there are a number of leads, mostly for a few species, for which fruit is required.

It would be an impossible task to frame two keys for each genus, one based on flower, the other on fruit characters: in Vatica and Dipterocarpus, for instance, keys based on sterile characters and flowers are impossible, while in Shorea reliance on either flower or fruit alone would be impossible (unless bark characters were included) though the key based on flowers would approach completeness. A reliable key based entirely on vegetative characters visible in the herbarium seems out of the question.

It is a rather unfortunate situation which we have to accept. Also local keys for the various islands would not bring much solution, since each of the three large Sunda islands harbours so many species.

2.As far as the vernacular names are concerned, I have selected only a limited number, as there is much overlapping.

3.Brandis, Dyer, Symington, van Slooten, I myself, and some other authors have entered in their works quite a number of unpublished names in the synonymy; they were not rarely taken up from herbarium labels where they were put from provisional identifications. They have here only been taken up if they were also taken up in the Index Kewensis.

4.As to the spelling of sectional and subsectional epithets I have adhered to priority of the original epithet given, which was either in the plural or singular.

Phytochemo

General chemical properties were summarised by Hegnauer (1966). Production of oleoresins (balms, resins) is characteristic of most members of the family. Their volatile portion consists mainly of sesquiterpenes such as humulenes, caryophyllenes, copaenes, elemenes and guajenes (e.g. gurjunenes, apitonene); in some instances monoterpenoids predominate (i.e. borneol in the so-called 'Borneo camphor' from Dryobalanops aromatica. The sesquiterpene alcohol spathulenol occurs in balms of species of three of the four subgenera of Shorea (Bisset et al., 1971). The resin fractions of the oleoresins are composed of triterpenoids and usually consist of neutral and acidic constituents. Dipterocarpol (= hydroxydammaradienone-II) is a ketonic tetracyclic triterpene alcohol having the so-called dammarane skeleton; together with similar compounds like dryobalanone it represents an outstanding feature of the subfamily Dipterocarpoideae. The dammarane skeleton is also present in a number of acidic resin constituents such as dipterocarpolic acid, dammarenolic acid (I) and shoreic acid (II). Other dipterocarpaceous resin triterpenoids possess the pentacyclic skeletons of ursolic acid (e.g. ursonic acid, asiatic acid, the lactonic compound B [III] etc.), oleanolic acid (e.g. oleanolic acid, hederagenin etc.) and betulinic acid (e.g. erythrodiol). Compounds I, II and III have an oxidatively cleaved A-ring; they represent so-called A-ring seco-triterpenes, which seem to be rather characteristic of dipterocarps. Many of the oleoresin constituents mentioned were described since 1966 (e.g. Chan, 1969; Cheung, 1967, 1968; Cheung & Feng, 1968; Cheung & Yan, 1972; Cheung & Wong, 1972; Gupta & Sukh Dev, 1971; Harrison et al, 1971; Lantz & Wolff, 1968). Some attention has been paid also to the phenolic constituents of leaves, barks, woods and seeds. Dipterocarps tend to produce proantho- cyanidins (i.e. oligomeric catechins formerly called leucoanthocyanidins) and gallic acid derivatives. These polyphenolic compounds are building stones of condensed and hydrolysable tannins; both types of tannins are present in taxon-characteristic ratios and amounts in many members of the family. Two derivatives of gallic acid deserve mentioning. Ellagic acid, the dilactone formed on hydrolysis of ellagitannins, was shown to occur in leaves and seeds of many species and bergenin, a striking derivative of gallic acid, has been isolated up to this day from members of Dipterocarpus, Shorea, Stemonoporus and Vateria (Bhrara & Seshadri, 1966; Desai et al., 1967, 1971; Bandaranayake et al., 1977). Most probably both compounds will turn out in future to represent good chemical characters of Dipterocarpaceae. The same may be true for Hopea-phenol, a phenolic constituent of barks and heartwoods, which is presently known from species of Hopea and Shorea (Coggon et al1965, 1966; Madhav et al., 1967). Hopea-phenol was shown to be a condensation product of four molecules of the trihydroxystilbene resveratrol. It is chemically similar to the viniferin-type phytoalexins of Vitis vinifera. With regard to phenolic leaf constituents Bate-Smith and Whitmore (ex Hegnauer, 1966) stressed the frequent occurrence of vicinal trihydroxylation (ellagic acid, gallic acid; B-ring in the flavonoids myricetin and prodelphinidin) in dipterocarps. As far as investigated, seed fats (oils) of Dipterocarpaceae are characterized by a strong predominance of stearic and oleic acid. Sal fat (oil of seed kernels of Shorea robusta) was shown recently to contain also small amounts (c. 4%) of 9,10-epoxystearic acid (Bringi, 1972).

Bate-Smith & Whitmore (1959) examined the phenols of fresh leaves in 28 species in 8 genera, giving attention to those compounds known to be of chemotaxonomic interest elsewhere. A grouping of genera was arrived at on the basis of the leucoanthocyanins present and their abundance which little reflected grouping established by traditional means; no clear grouping of Shorea species by their established sections was possible, though Neobalanocarpus was confirmed to closely resemble Hopea.

Recently taxonomic potentialities of chemical characters at an intrafamiliar level were discussed by several authors. Diaz et al. (1966) and Bisset et al. (1966, 1967, 1971) showed that the composition of the oleoresins (sesquiterpene-fractions, triterpene-fractions) is rather characteristic of the taxa Shorea sect. Doona, Anisoptera, Cotylelobium and Upuna and that in the genara Dryobalanops, Dipterocarpus and Shorea the chemistry of oleoresins might be helpful to classification beneath generic level. Subsequently Bandaranayake et al. (1975, 1977) stressed the systematic importance of resin composition in dipterocarps. According to these authors the significant differences in resin composition between representatives of Shorea sect. Doona and other sections do not agree with the proposition to merge Doona with Shorea. Very little work was performed with representatives of the African subfamily Monotoideae (Monotes, Marquesia) which lack resin ducts and seem to deviate chemically in several respects (Diaz et al., 1966) from the Asiatic subfamily Dipterocarpoideae. At present a chemotaxonomic discussion of relationships between these subfamilies seems to be premature. The same is true with regard to the recently described New World dipterocarpaceous genus Pakaraimaea (Maguire & Ashton, 1977) which according to Kostermans (1978) would belong to Tiliaceae. The chemical evidence (Giannasi & Niklas, 1977) given for dipterocarpaceous affinity is inadequate. Confirmation of the preliminary results reported by the authors mentioned as well as extension of phytochemical research are needed before chemical characters can make a serious contribution to the classification of the taxa concerned. Regarding relationships of Dipterocarpoideae with other plant families, the opinion held in 1966 by the present author is still valid; we are not yet in a position to discuss relationships in terms of chemical characters. On morphological arguments relationships with members of Malvales are often postulated. Presently known chemical characters do not convincingly contradict such a classification, but they form by no means strong evidence for such an affinity.
- R. Hegnauer