Donax canniformis (G.Forst.) Schumann (1892: 440)

Niissalo, Matti A., Khew, Gillian S., Webb, Edward L. & Leong-Škorničková, Jana, 2016, Notes on Singaporean native Zingiberales II: revision of Marantaceae, with a new generic record and notes on naturalised and commonly cultivated exotic species, Phytotaxa 289 (3), pp. 201-224 : 203-206

1. Donax canniformis (G.Forst.) Schumann (1892: 440)

Basionym:— Thalia canniformis Forster (1786: 1) , as ‘cannaeformis’ ≡ Actoplanes canniformis (G.Forst.) Schumann (1902: 34) ≡ Arundastrum canniforme (G.Forst.) Kuntze (1891: 683) ≡ Clinogyne canniformis (G.Forst.) Schumann (1897: 96) ≡ Ilythuria canniformis (G.Forst.) Rafinesque (1836: 51) ≡ Phrynium canniforme (G.Forst.) Körnicke (1858: 85) ≡ Phrynium canniforme (G.Forst.) Schrank (1824: 178) , nom. illeg., non Körnicke (1858: 85). ≡ Phrynium dichotomum Roxburgh (1810: 324) , nom. inval. ≡ Clinogyne dichotoma Salisbury (1812: 276) , nom. inval. ≡ Clinogyne dichotoma Bentham (1883: 651) , nom. inval. ≡ Maranta dichotoma Dietrich (1831: 17) .

Type:— VANUATU. Mallicollo, [Nov. Hebrid. Mallicolo], without date, J. G. A. Forster s.n. (lectotype BM! [ BM 000632825], first step designated by Rolfe 1907, second step designated by Veldkamp & Turner 2016).

= Donax arundastrum Loureiro (1790: 11) ≡ Maranta arundastrum (Lour.) Almeida (2009: 111) . Type: VIETNAM. Cochinchina [Southern Vietnam], without date, J. de Loureiro s.n. (lectotype BM! [BM000632823], designated by Rolfe 1907).

= Maranta tonchat Blume (1827: 36) . Type: INDONESIA. Java, without date C.V. Blume s.n. (lectotype L! [L1482450], here designated, isolectotypes L! [L1482442, L1482443, L1482444, L1482451, L1482452], BO?).

= Maranta arundinacea Blanco (1837: 5) , nom. illeg., non Linnaeus (1753a: 2) . Type: PHILIPPINES, Palawan, Tayao, May 1913, Merrill Species Blancoanae No. 279 (neotype US! [ US 00346027], here designated, isoneotypes L! [L1482266] P! [P02199296]).

= Maranta grandis Miquel (1860: 616) ≡ Actoplanes grandis (Miq.) Schumann (1902: 34) ≡ Arundastrum grande (Miq.) Kuntze (1891: 684) ≡ Clinogyne grandis (Miq.) Baker (1892: 258) ≡ Donax grandis (Miq) Schumann (1892: 440) ≡ Donax grandis (Miq.) Ridley (1899: 176) , nom. illeg., non Schumann (1892: 440). Type: INDONESIA. Sumatra, Palembang, without date, J.E. Teijsmann s.n. (lectotype K! [K000292262], here designated, isolectotype L! [U0282853], BO?).

= Actoplanes ridleyi Schumann (1902: 35) . Type: Burma and Malay Peninsula, without precise locality or date, W. Griffith 9784 (lectotype L! [L1482521], here designated).

= Donax parviflora Ridley 1910: 59 . Type: MALAYSIA. Selangor, woods at the base of Batu Caves, August 1908, H. N. Ridley 13393 (lectotype K! [K000292264], first step (“SING”) designated by Holttum (1951; see further notes below), second step lectotype here designated, isolectotypes BM! [BM000617230], K! [K000292263]).

Rhizomatous, herbaceous, terrestrial shrub with tall, branching aerial stems 2.0– 4.5 m, each stem protected at the base by a papery bract ca. 20–30 cm, after which there are no internodes or leaves until the height of 1.0– 2.5 m, at which point the plant branches out to many secondary branches, which branch further to produce numerous leaves, the length of internodes decreases as the order of branches increases. Leaf sheath 7–20 cm, green, glabrous except for a few hairs at the very base and adjacent to the pulvinus; petiole absent; pulvinus 1–4 cm; lamina oblong-ovate, 15–38 × 8–21 cm, apex acuminate, acumen 0.5–1.0 cm, base rounded, upper leaf surface medium to dark green, glabrous, lower surface light green, glabrous except for a pubescent narrow strip close to the midrib (top of midrib glabrous). Inflorescence terminal, several (three to many) emerging at the base of a leaf near the apex of the stem, spreading, base obscured by an enveloping leaf sheath; peduncle absent or very short; bract at the base of the synflorescence 5–10 × 0.8–1.0 cm, often absent (deciduous?); synflorescence heavily branched, lax, 11–20 cm; with variable orders of branching, each branch producing 10–20 fertile bracts, each fertile bract with a single flower-pair or a new branch, fertile bracts linear, narrow, deciduous, acute, 2.5–4.0 × 0.5–0.8 cm, tomentose, light green, soon drying to straw-yellow and papery, prophyll 18–22 × 2–4 mm, interphyll absent; Flower pair axes 1.5–4.0 cm to the lowest flower, pedicel of individual flowers 3–6 mm, with a short and very thick bracteole (which functions as an extrafloral nectary), 2.5 × 2.5 mm. Flower 1.7–1.9 cm long, with a very faint scent of jasmine; sepals 3, free, subulate with a distinct basal thickening, glabrous except for a few long hairs at the thickening, 5 × 0.5 mm, white; floral tube 6–7 mm long, incompletely fused; petal lobes equal, elliptic, acute, 10–12 × 5–6 mm, white, with a translucent margin, reflexed; staminodial tube 2–3 mm longer than the floral tube; outer staminodes 2, almost equal, free part of both staminodes elliptic, white with a slightly yellow margin, rolled in bud, only apex slightly crumpled, 9–10 × 2.5–4.5 mm; hooded staminode cucullate, almost white, free part 4 × 2.7 mm, appendix recurved, sulphur-yellow, ca. 2 × 1 mm; fleshy staminode almost tubular, opening up towards the apex, white, sulphur yellow towards to apex, free part 7–8 × 5–7 mm, with a tall and thin, sail-like white appendix that curves along the margin, 3 × 2 mm; fertile stamen 4.5 × 1.7 mm (including the fertile anther and appendage), recurved, linear with a blunt tip, anther emerging at 2.0– 2.5 mm measured from the base of the free part of the stamen, 1.7 × 0.6 mm; style with a 4 mm free part, curved inwards, asymmetrical with several lobes towards the apex, stigmatic cavity ca. 1 mm diameter; ovary globose, 1.8–2.4 × 1.9–2.1 mm, light brown, sericeous. Fruits globose, 10–12 mm in diameter, sericeous when young, maturing almost glabrous; only one seed develops (the other two can be seen as aborted seeds in mature fruit), aril absent.

Provisional IUCN conservation assessment:— Least Concern (LC). This species is widespread in the Malesian region, West to the Andaman and Nicobar Islands, North to Indochina and Southeast to Oceania. The species is capable of growing in disturbed habitats ( Clausager & Borchsenius 2003) and the population is globally not known to be decreasing. There are hundreds of collections made during the past thirty years.

In Singapore D. canniformis is nationally Critically Endangered (CR), under criterion D, as only ca. 40 clumps (each currently best considered to consist of a single individual) are known in a small area (one location, five localities) in forested habitats in the boundaries of Bukit Timah Nature Reserve. Because the species is very robust, it may appear abundant and dominant in small areas. As the species roots easily at leaf nodes and is therefore capable of vegetative spread, it is not yet known how many genetically distinct individuals are present in Singapore. There are only two records supported by specimens from Pulau Ubin, both very recent (2011 and 2012) from a single locality which is in the proximity of a village, and therefore its native status remains highly doubtful, considering that this species is often planted.

Specimens examined:— SINGAPORE. Bukit Timah: without date, Mohd Noor 1803 ( SING) ; April 1885, R. W. Hullett 440 ( SING) ; Taban Valley , 24 June 1994, S. C. Lee BT 011 ( SINU) ; Taban Valley , 19 July 1994, S. C. Lee BT 027 ( SINU) ; Tiup Tiup Path , 12 October 1995, Eugene Tang & Hj. Sidek 997 ( SING) ; 23 May 2013, J. Leong-Škorničková & V. Gowda SNG-157 ( SING) ; Dairy farm Loop, 9 March 2016, M. A. Niissalo, A. Thame, D. Liew & S. Teo SNG-343 ( SING) ; Dairy farm Loop , 9 March 2016, M. A. Niissalo et al. SNG-342 ( SING) ; Tiup Tiup Path , 8 April 2016, M. A. Niissalo et al. SNG-333 ( SING) . Bukit Panjang: 10 January 1889, H. N. Ridley 86 ( SING) . Kranji: 1894, H. N. Ridley s.n. ( SING) . Without locality: 1883, R. W. Hullett s.n. ( SING) ; without date, N. Cantley’s Collection s.n. ( SING) .

Probably introduced material:—Pulau Ubin: 11 February 2011 H. D. Tran et al. SNG-041 ( SING) ; Bukit Jelutong, 2 October 2012, Ali Ibrahim & J. Lai SING 2012-271 ( SING) .

Notes:— Donax canniformis , as currently circumscribed, is an extremely widespread species. Due to the perishable nature of flowers, which are rarely present in herbarium material, previously published synonymies are mostly based on comparison of leaves, fruits and inflorescence branching that vary widely. We have reviewed all available type material associated with names currently linked to this taxon and largely agree with previous treatments ( Suksathan & Borchsenius 2005, The Plant List 2013), although it is possible that one synonym, Donax parviflora , might need to be resurrected after fresh flowering material is recollected. We have also excluded Arundastrum benthamianum Kuntze (1891: 684) . This name, here used for an unrelated taxon, was recently discussed by Veldkamp & Turner (2016).

In Singapore native populations of D. canniformis are restricted to Bukit Timah Nature Reserve and its immediate surroundings. The previously recorded habitats in Bukit Panjang and Kranji no longer contain primary forest, and although the species is known to also inhabit secondary forests populations these may be of introduced origin. It could possibly be found in Singapore in the Central Catchment Nature Reserve, the species has been often planted from unknown origin, sourced from local nurseries. Therefore, plants found in secondary habitat are best to be treated as introduced as they are probably genetically different from the local stock.

Two sterile collections of this species from Pulau Ubin were mistakenly identified as Schumannianthus benthamianus , but can be safely determined as belonging to Donax canniformis by their much larger laminae (over 20 cm long) and by the presence of a narrow pubescent line positioned immediately adjacent to the midrib abaxially. More details on these collections are mentioned under Schumannianthus benthamianus .

Typification:— Maranta arundinacea Blanco non L.: There are very few specimens by Blanco or other contemporary collections from the Philippines ( Veldkamp 1989). We have not found any of Blanco’s contemporary material of Maranta arundinacea Blanco at the most likely herbaria, MA, P, G or L, and it is unlikely that any were collected and survived. We agree with Merrill (1918) that Maranta arundinacea , illustrated in the third edition of Blanco’s flora, an illegitimate later homonym of M. arundinacea L., is a synonym of Donax canniformis . Merrill’s “illustrative specimens” are not considered valid neotypifications ( Nicolson & Arculus 2001) and we therefore formalise neotypification of the name using Merrill’s collection of Donax canniformis at US.

Actoplanes ridleyi K.Schum. : The material cited by Schumann for the description of Actoplanes ridleyi were most certainly at B, but as with all other Zingiberales specimens, these have been lost during WWII. We found very few specimens in other herbaria that could be original material or from the same locality. Schumann cited Kurz collections from Tenasserin and Andaman, and Scortechini, Ridley and Schottmüller collections from Peninsular Malaysia [“Halbinsel Malakka ”] in the protologue. No relevant material by Kurz, Scortechini, Ridley or Schottmüller survived at B. There are three specimens by Scortechini, one each at L, P, and SING, at least one pre-dating Schumann’s publication, and a few collections by Ridley from Peninsular Malaysia that pre-date Schumann’s work, all annotated as Donax parviflorum . It is not clear if any of these are from same collections as the material seen by Schumann. We have checked all material for Donax Loureiro (1790: 1) we could locate, and the only specimen determined in Schumann’s handwriting as Actoplanes ridleyi is a Griffith specimen that pre-dates the publication of A. ridleyi (it was distributed to K in 1963–1964), but which is not cited in the protologue. However, Schumann’s determination slip is labelled as “Bearbeitet für das ‘Pflanzenreich’”, which was the place of publication of Actoplanes ridleyi . The material is clearly part of the uncited original material, and unambiguously the best suitable lectotype available for this name. Schumann’s description may have been based on mixed material, as his description in general fits Donax canniformis , but he states that A. ridleyi only rarely has a single seed in the fruit, usually more. Some of the fruiting specimens Schumann saw may have been of Schumannianthus benthamianus , which usually has up to three seeds per fruit. However, we have been unable to find any Schumannianthus specimens identified by Schumann as A. ridleyi .

Donax parviflora Ridl. : Holttum (1951) selected the collection Ridley 13393 as the lectotype of this name, stating that all material seen of this species is in the Singapore Herbarium. However, no material of this collection number could be found at SING and there is no record of it in the card index at SING. Turner (2000) cites the location of the lectotype as K, where there are two specimens of this collection. One of the specimens has been transferred from what appears to be a SING herbarium sheet (still seen on the background) onto a K sheet. This collection might be an unreturned loan (or gift) from SING to K, and was most likely the specimen originally selected by Holttum who worked at K. We have therefore selected it as the lectotype of this name in a second-step lectotypification, since there are now two sheets of the same collection present at K. The other specimen is an isolectotype.

Donax grandis Miq. : While two collections (identifiable by location details and handwriting attributed to J.E. Teijsmann) at L and K are almost certainly duplicates, and the sheet at L (originally from Utrecht) also probably originates from Miquel’s herbarium, we have selected the collection at K as the lectotype, since it is explicitly collected by Teijsmann, and carries a label clearly identifying it as part of Miquel’s herbarium.

Maranta tonchat Blume : The most complete specimen was selected from known collections by Blume at L.