A revision of the genus Gouania (Rhamnaceae) in the Philippines and Sundaland

The climbing genus Gouania (Rhamnaceae) is revised in western Malesia. Five species are recognised for the Philippines and Sundaland using morphological evidence, including a new species from the Philippines: Gouania longipedunculata. A taxonomic treatment, including a preliminary IUCN conservation status assessment, is presented for each species.


Introduction
Gouania Jacq. is a pantropical genus of over 50 species of woody climbers, recognised by having circinnate tendrils and 3-winged dry schizocarps separating into three 2-winged indehiscent mericarps (Medan & Schirarend 2004). The genus belongs to the monophyletic tribe Gouanieae, first formally described by Endlicher (1840) and confirmed to be monophyletic by Richardson et al. (2000a, b). Gouania and Helinus E.Mey. ex Endl. are the only members of the tribe occurring in the Old World. Helinus species are also climbers with tendrils, but their fruits are unwinged explosive capsules and are restricted to Eastern and Southern Africa, Madagascar, the southern end of the Arabian Peninsula and the Indian subcontinent. Gouania alone is found in the Philippines and Sundaland. As noted by Medan & Schrirarend (loc. cit.), the genus needs revision, but recent headway has been made with revisions for the Western Indian Ocean (Buerki et al. 2011) and North America (Pool 2014). Most of the c. 15 South-East Asian genera of Rhamnaceae, and the climbing genera in particular, have yet to be revised and included in contemporary floras such as Flora Malesiana or the Flora of Peninsular Malaysia, although a synopsis of Smythea (Cahen & Utteridge 2018) and a key to the Bornean species of Ventilago (Cahen & Utteridge 2017) were recently published. The most complete overview of Asian Gouania species to date is by Lauterbach (1922), with some minor changes by Suessenguth (1953) in the last attempt at a complete treatment at species level.
While Sundaland's phytogeographical boundaries differ according to sources and the distribution of a single species is often at odds with general patterns (van Steenis 1950), the area delimited by Woodruff (2003) shares the same northwestern and southeastern limits as the distribution of Gouania obtusifolia, with limits just north of the Isthmus of Kra on the Malay Peninsula and immediately east of Java, and is used here to define the extent of the study area for Sundaland. The taxonomic history of the genus in Sundaland is complex and best understood by starting with Java as more literature is available for this area. Blume (1826) recognised two species that had originally been described from Réunion: G. tiliaefolia Lam. [G. tiliifolia Lam.] and G. mauritiana Lam. The same year, Brongniart (1826) described G. obtusifolia Vent. ex Brongn., a new species based on material in Ventenats's herbarium collected in Java. Miquel (1856) seems to have been unaware of Brongniart's species when he described G. javanica Miq., and also treated G. tiliaefolia sensu Blume non Lam. as a synonym of G. leptostachya DC. and listed G. microcarpa DC. and G. retinaria DC. as additional species for the island. Gouania microcarpa however is restricted to India and Sri Lanka, as noted by King (1896), and G. retinaria is a synonym of G. scandens (Gaertn.) R.B.Drumm., which is only known to occur in the Western Indian Ocean (Buerki et al. 2011). Backer & Bakhuizen van der Brink (1965) recognised only G. javanica and G. leptostachya for Java. Study of herbarium material supports this view, although G. javanica is placed here in the synonymy of G. obtusifolia. The same two species found in Java occur in Peninsular Malaysia as was recognised by King (loc. cit.) and Ridley (1922) and are the only Gouania species known to occur there. Confusingly, the name G. javanica has also been extensively used to refer to plants observed in southern China and throughout SE Asia (Chen & Schirarend 2007;Pitard 1907Pitard -1912Wilson 1916). Plants identified as G. javanica collected north of the Malay Peninsula differ from G. obtusifolia but are best referred as G. brandisii Hassk. (discussed below in Notes for G. obtusifolia). Little could be found about the genus in Sumatra; Gouania is not included in Prodromus florae Sumatranae (Miquel 1860), but Suessengueth (1953 mentions both G. javanica (= G. obtusifolia) and G. leptostachya as occurring on the island.
The Philippine and Malaysian floras are closely related (van Steenis loc. cit.) and both areas are included in a recognised Sunda Shelf and Philippines bioregion (Wikramanayake et al. 2002). While two species included here are Philippine endemics, G. nematostachya straddles both Sundaland and the Philippines and the genus is advantageously studied for both areas together. Blanco (1837) originally presented G. domingensis (Jacq.) L. as the only species of the genus occurring in the Philippines. Gouania domingensis sensu Blanco, non L. was later treated as a synonym of G. leptostachya by Fernández-Villar (1880) in Novissima Appendix to Blanco's Flora. Gouania microcarpa was later recognised as an additional species of the Philippines by Rolfe (1885), who noted that nearly all plants included in the Novissima Appendix were based on descriptions of species from India and the Malay Archipelago, resulting in an underestimation of the number of endemic species. Following Rolfe's publication, G. domingensis sensu Blanco, non L. and G. leptostachya sensu Fern.-Vill., non DC. were placed as synonyms of G. microcarpa by Vidal (1886) and Merrill (1918), resulting in a single recognised species for the Philippines. Merrill (1923) then placed G. microcarpa in synonymy under G. tiliaefolia and recognised G. javanica as an additional species for Luzon, Polillo, Mindoro, Leyte and Mindanao. Lauterbach (1922) did not consider G. tiliaefolia and G. javanica to occur in the Philippines and described two endemic new species: G. fimbriata Reissek ex Lauterb. and G. nematostachya Reissek ex Lauterb., and both species were retained by Suessengueth (1953) and are recognised here as distinct species after study of herbarium material. Gouania nematostachya also occurs in Sulawesi and Borneo. It is the only species of Gouania known to grow in Borneo, where it was identified as G. microcarpa by several authors, e.g. Merrill 1921, Masamune 1942, Beaman & Anderson 2004, but differs from G. microcarpa in several morphological characters (discussed below in Notes for G. nematostachya). Specimens from Philippine islands south of Luzon, with densely hairy leaves and peduncles longer than in G. fimbriata, are described here as belonging to a new species: G. longipedunculata.

Material and Methods
Herbarium specimens from BKF, E, GH, K and L were studied and measurements with additional digital images studied from A, E, G, L, M, NY, P, U, S, W, WAG and WRSL; an exclamation mark (!) is used to show that a specimen has been seen and barcodes are given for type information where there could be confusion. An alphabetical index to all numbered collections examined is given at the end of the paper (Appendix 1). Material was examined under a Zeiss Stemi 1000 binocular microscope at magnifications of up ×350. Leaf anatomy terms used are from Hickey (1979). Leaf measurements were made on branchlet leaves inserted proximally to the racemiform inflorescences, rather than leaves inserted along the racemiform inflorescence rachis. The terminology used to describe stipules follows Pool (2014). Hair density terms are defined as follows: sparse when hairs are scattered enough not to touch when pressed towards each other, abundant when hairs are close enough to touch if pressed towards each other, dense when hairs are so close to each other that they hide the surface of the organ they grow on. Inflorescences are described here as racemiform thyrses because they superficially resemble racemes, with the cymes inserted along a well-developed, unbranched axis. Fruit measurements follow Pool (loc. cit.), but only on fruit mature enough to be at dispersal, i.e. splitting, or on free mericarps; pedicels were also measured at the mature fruit stage. Other morphology terms follow Beentje (2010). Climber length, basal diameter, flower colour and fragrance information were retrieved from herbarium specimen labels. Habitat names used in species accounts are based on the Ecoregions 2017 map (Dinerstein et al. 2017). Localities and collection dates were inferred where possible using the Cyclopaedia of Malesian Collectors (van Steenis-Kruseman 2017) but coordinates are not given for localities when uncertainty or imprecision is too great. Conservation assessments in species accounts apply IUCN Red List categories and criteria (IUCN 2012), using GeoCat (Bachman et al. 2011) to calculate Extent of Occurrence (EOO) metrics. Protected areas were located using the World Database on Protected Areas online interface (IUCN & UNEP-WCMC 2018).

Key morphological characters
Leaves Gouania species all have simple, alternate, petiolate leaves. Glands and foliaceous appendages are often present along the petiole, especially near the junction with the leaf blade. Glands are also usually present along the leaf margin, with one gland associated to each serration when serrations are present. Leaves vary in shape on the same plant. A consistent pattern noted by Lauterbach (1922) is for proximal leaves to be cordate and distal leaves to be cuneate to rounded at the base. All Philippine and Sundaland species of Gouania can be recognised by observing leaf features only. Gouania leptostachya stands out by being the only species with 4 -5 pairs of secondary veins and sunken marginal glands (Fig. 1B). Other species have leaves with 5 -6 (-7) pairs of secondary veins and protruding marginal glands. Gouania leptostachya leaves are also completely glabrous except along veins. The only other species that sometimes has glabrous leaves is G. fimbriata, which is only known from Luzon and where G. leptostachya is unrecorded. Gouania nematostachya is unique in having up to 50 small serrations along the leaf margins (Fig. 1C). Other species have leaves with up to 30 serrations along the margin. A foliaceous appendage or tufts of hairs (domatia) are often present in axils of veins in G. fimbriata, G. longipedunculata and G. obtusifolia (Fig. 1D). Of the three, G. obtusifolia is the only species with conspicuous crenations all along the leaf margin. Gouania longipedunculata leaves are abundantly to densely hairy when mature (Fig. 2D & E), unlike the glabrous to sparsely hairy leaves of G. fimbriata. The leaf blade apex is rounded to acute in G. obtusifolia, acute to attenuate (-acuminate) in G. fimbriata and G. longipedunculata and attenuate to acuminate in G. leptostachya and G. nematostachya.

Stipules
Stipules are soon fugaceous for all Gouania species from the region and are visible in just a small proportion of herbarium specimens from that area. They are usually lost by anthesis and are most easily observed remaining near the apical meristem of developing shoots. Only in G. fimbriata, and to a certain extent G. longipedunculata, are stipules frequently retained at anthesis. Reissek (1861), and especially Pool (2014), recognised the diagnostic usefulness of stipules for identification of American Gouania species. Stipules are also useful to recognise Asian species. They are deeply pinnatifid and up to 8 mm long in G. fimbriata, pinnatifid and up to 7 mm in G. longipedunculata (Fig. 2F & G), entire to slightly pinnatifid and c. 6 mm long in G. obtusifolia, entire and c. 2.5 mm long in G. nematostachya. The only species of this region that has stipules with a lobule at the base is G. leptostachya. The upper portion of the stipule is entire and c. 4 mm long; the lobule diverges at c. 120°f rom the upper portion and is c. 1 mm long (Fig. 1A).

Disk and disk lobes
Disks of Gouania flowers provide especially useful diagnostic characters and have been extensively used in dichotomous keys (Reissek 1861;Lauterbach 1922;Suessenguth 1953;Buerki et al. 2011;Pool 2014) and are helpful to identify Philippine and Sundaland species. In all known Asian Gouania species, nectar disks are fleshy and fill most of the hypanthium above the ovary and around the exserting trifid style. The central portion of the disk, called the annulus (following Pool 2014), often forms a sheath around the style, especially in G. fimbriata, G. longipedunculata and G. obtusifolia. Weak ridges are sometimes visible on the disk surface, radiating from the annulus. The surface is glabrous in all Malesian species, smooth in G. fimbriata, G. longipedunculata and G. obtusifolia, smooth to rugulose (weakly wrinkled) in G. leptostachya and rugose (wrinkled) in G. nematostachya (Fig.  1E). Narrow disk lobes opposite the sepals are present in many species of Gouania worldwide, including all species of the Philippines and Sundaland except G. nematostachya. The disk lobes are tapering in G. fimbriata, G. longipedunculata (Fig. 2K) and G. obtusifolia, unlike in G. leptostachya, which has disk lobes with parallel sides as noted by Lauterbach (1922) and Suessenguth (1953). It is often possible to observe the presence of disk lobes and determine whether the disk is smooth or rugose by looking at the flower remains at the distal end of fruits.

Fruit
The mature fruit provides diagnostic characters for several taxa from North America (Pool 2014) and the Western Indian Ocean (Buerki et al. 2011), and is also useful for the identification of Philippine and Sundaland species. What is referred to here as mature is when mericarps can be found splitting from each other (Fig. 1F). This definition is not completely satisfactory because some fruits with not fully developed wings can occasionally be seen splitting on herbarium specimens (Cahen & Stenn pers. obs.). When immature, fruits are obovoid and unwinged (Fig. 2H). The inferior ovary dries with a darker colour than the persisting sepal lobes, giving the impression it might mature into a fleshy fruit. As fruits develop however, the mericarp wings expand, both laterally and vertically (Fig. 2N). Gouania leptostachya has significantly bigger fruits than all other speciesoftheregion, withafruitbody(the medialportionof the mericarp where the seed is enclosed) c. 8 mm high and mericarps 10 -16 mm wide (Fig. 1F). Fruit bodies are c. 4 mm high in G. fimbriata, G. longipedunculata (Fig. 2N) and G. obtusifolia, and c. 5 mm high in G. nematostachya. Their mericarps are 9 -12 mm wide. The fruit body also tends to be sparsely hairy and to dry with a darker colour than the mericarp wings in these species, unlike G. leptostachya, which has a glabrous fruit body that dries with the same colour as the mericarp wings. Gouania nematostachya fruits are slightly smaller and have more of a "bowtie shape" (terminology from Pool 2014), with a smaller height/width ratio than all other species of the region. Note that peduncles and pedicels can easily be confused when examining Gouania specimens. Peduncles are persistent after fruit dispersal and bear scars where the pedicels were inserted. The only species in this region with strongly developed peduncles are G. fimbriata (to 4 mm long) and G. longipedunculata (to 8 mm long).
Woody climber with circinnate tendrils, to at least 3 m long; girth to at least 5 cm. Indumentum lacking on older branches to dense at distal end of branchlets; stem hairs 0.2 mm long, curved to tortuous, spreading to appressed-antrorse, mostly reddish-ferruginous, some whitish. Branchlets hollow in the centre, slender, subterete, smooth, with flush longitudinal striations.
Stipules fugaceous although often persisting to anthesis, densely hairy, slightly asymmetric, lanceolate, acuminate, pinnatisect, 7 (-9) mm long, medial undivided portion c. 1.3 mm wide, divisions narrow, c. 0.2 mm wide, to 3 mm long. Leaves: discolorous, dark green adaxially, glossy adaxially when fresh, paler abaxially, blade wide ovateovate, 3.0 -8.2 cm long, 2.0 -6.0 cm wide, chartaceous, apex attenuate (acuminate), base rounded to cordate, margins weakly serrate, usually 15 -20 serrations on either side, margin subentire in proximal 1 3, serrations more densely crowded at leaf apex, margin with one protruding gland at the apex of each serration, glands crowned with hairs, leaf margin often ciliate; primary vein densely hairy abaxially, hairs mostly appressed, antrorse; secondary veins (5 -) 6 -7 pairs, densely hairy abaxially, venation eucamptodromous, basal secondary veins reaching margin at c. 60% of leaf height, angle of divergence from primary vein 30 -40°; outer secondary veins branching off basal secondary veins conspicuous, 4 -5; lamina glabrous adaxially when leaf mature, including along veins; lamina glabrous to sparsely hairy abaxially when leaf mature, densely hairy when immature; domatia usually present in vein axils, a foliaceous appendage or tufts of hairs; petiole subterete, flattening distally, channelled, 8 -21 mm long, densely hairy, with glands and foliaceous appendages, especially along edges of channel and near junction with leaf blade, similar to appendages found in domatia. Inflorescence of congested cymes arranged along racemiform thyrses, the distal thyrses often arranged in a panicle, thyrses to 25 cm, with c. 6 flowers in each cyme; bracts narrow-lanceolate, acuminate, 4 -6 mm long, pinnatifid, densely hairy; cymes mostly pedunculate, peduncles 1 -4 mm long, pedicels to 1.5 mm long, densely hairy. Flower odour unknown; hypanthium densely hairy adaxially, indumentum often paler than on inflorescence rachis, sepals triangular, c. 1.0 mm long, greenish-white when fresh, glabrous and keeled adaxially; petals clawed, c. 1.0 mm long, rugulose adaxially, colour when fresh unknown; stamen filaments flat, subulate, c. 0.7 mm long, anthers c. 0.2 mm long, enclosed by petals, nectary disk stellate, glabrous, smooth, diameter excluding lobes c. 1.5 mm, annulus glabrous, raised, tightly sheathing style, disk lobes narrow, tapering, apex often notched, usually appressed to sepal lobe when dry, 0.5 -0.7 mm long ( 2 3 of sepal length); style arms 3, glabrous. Fruit green when fresh, fruit body glabrous to sparsely hairy, more densely near apex, drying darker brown than wings, c. 4.5 mm high, wings glabrous, c. 7 mm high, > 1.5 × height of fruit body, distance between highest points of wings 3 -5 mm, mericarp width 7 -9 mm, fruit body width 2 -3 mm, 1 3 mericarp width. Seeds 3 × 2 mm. obovate, dorsally convex, ventrally concave, shiny brown. Ramos BS 1749 (K!, WRSL!-image seen); Antipolo, Araneta, 25 Jan. 1953, Viola 34 (L!). HABITAT. Luzon rain forests; alt. 0 -450 m. CONSERVATION STATUS. Gouania fimbriata was collected in many parts of Luzon, from Albay in the south-east to Ilocos Norte in the north-west, including several collections in the 1990s from various parts of the island (Abra, Bulacan, Cagayan, Ilocos Norte). The species has been observed in less threatened parts of the island, along the Central Cordillera and Southern Sierra Madre. The EOO of G. fimbriata is of over 85,000 km 2 and it is assessed here as of Least Concern (LC). However, most collections available were made before World War II, including many made near Manila where the species could be severely threatened. Populations could also be threatened in the Central Luzon Plains and Cagayan River Valley, which have experienced extensive clearing for logging and agricultural use (Lamoreux 2001). PHENOLOGY. Mostly collected in flower from Dec. to Feb., but also Sept. and Oct.; collected in fruit (mature, splitting) from Jan. to March and Oct. VERNACULAR NAMES. Lituan (Tagalog, fide Viola 34); Sitiran or Fitiran (Tagalog, fide Manayon 69). NOTES. Gouania fimbriata is recognised by its deeply pinnatifid stipules and bracts and by the stipules often persisting at anthesis on herbarium specimens. The leaf lamina is glabrous to sparsely hairy abaxially when mature and cymes are borne on peduncles 1 -4 mm long. The other known species of the genus found in the Philippines are G. nematostachya, which has a rugose nectar disk with no narrow lobes opposite the sepals, and G. longipedunculata, which has abundantly to densely hairy mature leaves and peduncles to 8 mm long (Table 1).
Like Gouania nematostachya, this species was published by Lauterbach (1922) based on a manuscript name and notes by Reissek with no clear indication of type specimens and has hardly appeared in the taxonomic literature since (see Notes for G. nematostachya).
In his description of Gouania domingensis, Blanco (1837) specified that narrow disk lobes are present opposite the sepal lobes, which rules out that he was referring to G. nematostachya. In addition, he wrote that the leaves are somewhat hairy, unlike the densely hairy leaves of G. longipedunculata. Blanco observed the plant in Angat on Luzon where G. longipedunculata is not known to occur. It is therefore most likely that G. domingensis sensu Blanco is a synonym of G. fimbriata. Merrill (1918: 245) selected Merrill 250 a G. fimbriata specimenas illustrative of Blanco's G. domingensis. Merrill (loc. cit.) also reduced G. domingensis sensu Blanco to a synonym of G. microcarpa, a name commonly used to designate Gouania specimens from the Philippines despite being a species restricted to India and Sri Lanka (also see Notes for G. nematostachya).
The lectotype of Gouania fimbriata selected here was collected in Albay according to the specimen's label. However, the collection location may have been Pangasinan according to Merrill (1915: 183 Woody climber with circinnate tendrils, to at least 20 m long; girth to 15 cm. Indumentum lacking on branchlets except on rachis of racemiform thyrses. Branchlets hollow in the centre, slender, subterete, often flattened distally by collapsing inside hollow section, smooth, with flush longitudinal striations. Stipules soon fugaceous, glabrous except at apex and along margins, asymmetric, upper lobe lanceolate, acuminate, c. 4 mm long, entire, lower lobe lanceolate, acuminate, diverging at a c. 120°angle from the upper lobe, c. 1.0 mm long. Leaves: discolorous, dark green adaxially, paler abaxially, blade (narrow-)ovate, 4.5 -10.0 cm long, 2.5 -6.2 cm wide, chartaceous, shiny when fresh, apex acuminate, base rounded to cordate, margins crenate, up to 30 serrations on either side, all along leaf margin, quite evenly spaced, serrations with glands submarginal, often in a small depression, leaf margin glabrous; primary vein sparsely to abundantly hairy, hairs mostly appressed, antrorse; secondary veins 4 -5 pairs, sparsely to abundantly hairy, venation eucamptodromous, basal secondary veins reaching margin at c. 60 % of leaf height. angle of divergence from primary vein 35 -50°; outer secondary veins branching off basal secondary veins often conspicuous, 3 -5; lamina glabrous to very sparsely hairy adaxially and abaxially; domatia lacking; petiole subterete, channelled, often laterally compressed, 7 -29 mm long, glabrous except along edges of channel, with glands and foliaceous appendages, especially near junction with leaf blade. Inflorescence of congested cymes arranged along racemiform thyrses, the distal thyrses often arranged in a panicle, thyrses to 35 cm, with c. 5 flowers in each cyme; bracts narrowlanceolate, acuminate, c. 2.5 mm long, entire, glabrous; cymes sessile, pedicels to 2.5 mm long, sparsely hairy. Flowers scented; hypanthium adaxially sparsely hairy at base, more densely distally; sepals triangular, c. 1.2 mm long, often rugulose adaxially, white when fresh, glabrous and keeled abaxially; petals clawed, c. 1.1 mm long, rugulose adaxially, whitish-cream when fresh; stamen filaments flat, subulate, c. 0.9 mm long, anthers c. 0.2 mm long, enclosed by petals, nectary disk stellate, glabrous, often rugulose, diameter excluding lobes c. 1.8 mm, annulus glabrous, barely raised, loose around style, disk lobes narrow, with parallel sides, usually bending away from sepal lobes when dry, apex entire to slightly notched, c. 0.6 mm long (½ of sepal length); style arms 3, glabrous, often undeveloped. Fruit green when fresh, glabrous, drying yellowish-brown, fruit body c. 8 mm high, wings 9 -12 mm high, c. 1.3 × height of fruit body, distance between highest points of wings c. 7 mm, mericarp width 10 -16 mm, fruit body width c. 4 mm, 1 3 mericarp width. Seeds 4 × 4 mm. obovate, dorsally convex, ventrally concave, shiny brown.  USES. Finely crushed roots, stems and leaves were used externally against headaches and as a treatment for scabies; shampoo made from the bark and applied to the scalp was used to kill parasites (Heyne 1922). NOTES. Gouania leptostachya is the only species in Sundaland to have stipules with a basal lobule pointing sideways, glabrous bracts, 4 -5 pairs of secondary veins, marginal glands sunken inside margin serrations, a fruit body c. 8 mm high and that has the same colour as the wings; mericarps 10 -16 mm wide and seeds 4 × 4 mm. All other species have unlobed stipules (entire to pinnatifid), hairy bracts, usually at least 5 pairs of secondary veins, protruding marginal glands, a fruit body to 5 mm high and that is usually darker than the wings, mericarps to 12 mm wide and seeds to 3 × 2.5 mm (Table 2). Gouania retinaria DC. is listed as a possible synonym of G. leptostachya by Suessenguth (1953). According to Buerki et al. (2011), G. retinaria is instead a synonym of G. scandens (Gaertn.) R.B.Drumm., which occurs in Mauritius, Réunion and Madagascar. Gouania leptostachya material from Sundaland was identified as G. tiliaefolia [G. tiliifolia Lam.] by Blume (1826), which is also synonym of G. scandens according to Buerki et al. (loc. cit.).
A. P. de Candolle cited material from the botanical garden of Calcutta in his protologue of Gouania leptostachya. The sheets found in G-DC include a specimen from Hort. Bot. Calcutt. designated here as the lectotype. Additional material of G. leptostachya from the botanical garden of Calcutta listed in the Wallich Catalogue (Catalogue no 4270.E) is present in herbaria (e.g. G!-image seen [G00378049], K-W! [K001038588], E!-image seen [E00777941]). Because De Candolle received material from Calcutta before Wallich started sorting the East India Company specimens, it is uncertain as to whether these specimens are duplicates of the type and they are not listed here as isolectotypes. The lectotype for Nagelia dubia is selected here at P where Java collections by Zollinger and received by Moritzi are held according to Stafleu & Cowan (1981: 589).

Gouania longipedunculata
RECOGNITION. Most like Gouania fimbriata in having narrow disk lobes opposite the sepals, deeply divided stipules often persistent to anthesis and pedunculate cymes, but differs by its abundantly to densely hairy leaves and longer peduncles to 8 mm long.  Pedicel to 2.5 mm long to 2 mm long to 1 mm long sparsely hairy densely hairy densely hairy Body c. 8 mm high c. 5 mm high c. 4 mm high same colour as wings darker than wings darker than wings glabrous glabrous to sparsely hairy sparsely hairy Wing height 9 -12 mm 6 -7 mm 7 -10 mm Mericarp width 10 -16 mm 9 -11 mm 9 -12 mm Seed size 4 × 4 mm 3 × 2 mm 2 × 2 mm PHENOLOGY. Collected in flower in April -July and Dec.; collected in fruit (mature, splitting) in Dec. -Feb. ETYMOLOGY. The specific epithet refers to the peduncles to 8 mm long, which are the longest of any known Gouania species of Malesia. NOTES. Gouania longipedunculata is unique in the genus with the following combination of characters: peduncles to 8 mm long, flower disks with five narrow lobes opposite the sepals and leaves abundantly to very densely hairy abaxially when mature. All specimens of this species were collected outside of Luzon, in various islands and habitats. There is significant variability in how deeply the bracts are divided and how long the peduncles are. This contrasts with the more morphologically consistent Luzonian Gouania fimbriata. All the specimens examined did however share the character of abundantly to densely hairy mature leaves and peduncles mostly longer than the maximum of 4 mm for G. fimbriata (Table 1). This suggests that G. longipedunculata could form a group distinct from G. fimbriata but possibly comprising of morphological subsets.
Cuming 1623 is filed as Gouania syringaefolia at W, an unpublished name presumably used by Reissek for this material, although Lauterbach (1922) did not mention such a species. The label on the herbarium specimen at K indicates that the specimen was collected in Mindanao, but it would have been collected in Mindoro according Merrill (1915: 183).
Specimens of Gouania longipedunculata were originally identified as G. microcarpa less often than specimens of G. fimbriata and G. nematostachya. Instead, they were more often identified as G. tiliaefolia [= G. tiliifolia Lam.], a species restricted to islands of the Western Indian Ocean (see G. leptostachya notes).

Gouania nematostachya
Reiss. ex Lauterb. (Lauterbach 1922: 339 Gouania javanica sensu Merr. (Merrill: 1923: 525), pro parte, non Miq. Gouania tiliaefolia sensu Merr. (Merrill: 1923: 526) NOTES. Gouania nematostachya is the only Gouania species in the region with nectar disks lacking narrow lobes opposite the sepals, and is unique in having consistently rugose (wrinkled) nectar disks; it is also recognised by having leaves with usually 40 -50 fine serrations along the margin. Gouania nematostachya is the only species of the genus known to occur in Borneo. The only other known species of Gouania in the Philippines are G. fimbriata and G. longipedunculata, which have narrow disk lobes opposite the sepals and a leaf margin with usually 15 -20 serrations on each side. Mature fruits of G. nematostachya have a smaller height to width ratio than other species of this region ( Table 1). Leaves of this species are acuminate, a character that can also be observed in G. leptostachya, which is not known to grow sympatrically with G. nematostachaya and is easily distinguished by its glabrous leaf surface except along veins, fewer secondary veins and crenate leaves with sunken marginal glands (Table 2). This species and Gouania fimbriata were published by Lauterbach (1922) based on manuscript names and notes by Reissek with no designation of a type specimen. Adherence to the type principle did not become mandatory until 1958, and because a diagnosis is offered for both species, they were validly published. Neither G. fimbriata nor G. nematostachya seem to have appeared in the taxonomic literature since Suessengueth's monographic studies of Rhamnaceae (1953). In addition, no other known herbarium specimens besides those examined by Reissek and Lauterbach had hitherto been identified as belonging to either species.
Cuming 1578 held in W is selected here as the lectotype because it was determined as Gouania nematostachya by both Reissek and Lauterbach, has some open flowers and duplicates are available in several herbaria. The collection year (1837) is not indicated on the label and was inferred using the Cyclopaedia of Malesian Collectors (van Steenis-Kruseman 2017).
About half of the Gouania nematostachya specimens from Borneo and the Philippines were determined as Gouania microcarpa DC. (the rest mostly determined as Gouania sp.). Both species have nectar disks without narrow lobes opposite the sepals, but G. microcarpa has a densely hairy nectar disk and narrower, subentire leaves that are nearly glabrous except along veins. It is only known to occur in India and Sri Lanka, as noted by King (1896). Confusingly, Bhandari & Bhansali (1990) correctly specify that the disk of G. microcarpa is villous in their key to species of Gouania, but note the disk is glabrous in the species description.
Specimens from the Philippines listed by Rolfe (1885) as Gouania microcarpa also belong to G. nematostachya. Elmer (1915) recognised that Philippine material of G. nematostachyathen not yet published by Lauterbach was distinct from typical G. microcarpa. He described it as G. microcarpa var. subglabra while noting that it should possibly be considered a new species. Merrill (1921), Masamune (1942) and Beaman & Anderson (2004) all list Gouania microcarpa as occurring in Borneo. Illustrative material used by Beaman could be traced in herbaria and is identified here as G. nematostachya. Lauterbach (1922), while agreeing with King (1896) that Gouania microcarpa s.s. is restricted to India and Sri Lanka, described several taxa without disk lobes opposite the sepals, from New Guinea as varieties of G. microcarpa. Nevertheless, as Lauterbach (loc. cit.) noted, none of the New Guinean taxa have hairy disks. Despite Lauterbach considering the New Guinean specimens to be varieties of G. microcarpa, he recognised G. nematostachya as a separate species and not an additional variety of G. microcarpa.
Although the scope of this revision is restricted to the Philippines and Sundaland, occurrences from Sulawesi are included here to give a complete list for this poorly known species. Sulawesi specimens usually have slightly narrower leaves and were collected in flower and fruit at earlier months of the year.
All but two Bornean specimens of Gouania nematostachya available were collected in Sabah. Amdjah 667 was collected nearby in North Kalimantan. Only Korthals s.n. was collected in the southern half of the islandits leaves are not as conspicuously serrate as other specimens from Borneo, but all other characters suggest it belongs to G. nematostachya. A sampling bias may explain why almost all specimens were collected in such a limited area of the island. More material is available from Sabah, especially collections from SAN, than from other parts of the island. BRUN does not hold any Gouania specimens (Joffre Haji Ali Ahmad, pers. comm.).