A Strobilanthes (Acanthaceae) miscellany

This paper presents miscellaneous, previously unpublished results arising from the authors’ research into Strobilanthes. S. moylaniae J.R.I.Wood & Scotland from New Guinea and S. dengii J.R.I.Wood & Scotland from Sumatra are proposed as new species, while an unnamed species is described from Sumba Island in Indonesia in the hope that it might be rediscovered. A new subspecies, subsp. subovata J.R.I.Wood & Scotland is described for S. timorensis Nees, while attention is drawn to its plietesial flowering pattern. The new combinations S. benculensis (Bremek.) J.R.I.Wood & Scotland, S. wetarensis (Bremek.) J.R.I.Wood & Scotland, S. javanica (Bremek.) J.R.I.Wood & Scotland and S. serpens (Nees) J.R.I. Wood & Scotland are made. Typification is clarified and where appropriate lectotypes are designated for Hemigraphis hispidula Craib, Lepidagathis setigera Blume, Ruellia crispa L., R. sabiniana Lindl., Strobilanthes duclouxii Benoist, S. erecta C.B.Clarke, S. flava Kurz, S. forrestii Diels, S. scabra Nees, S. secunda T.Anderson and S. thomsonii T.Anderson.


Introduction
Strobilanthes Blume is the second largest genus of Acanthaceae with around 400 -420 accepted species according to our estimate and if a broad concept of the genus is accepted. It is almost entirely Asian in its distribution, extending from Japan and Korea in the north, from Afghanistan and Pakistan in the west and just reaching northern Australia in the south. It is most diverse in the Indian subcontinent, south China and mainland South East Asia but many species also occur in island South East Asia. The genus is remarkable for the diversity of its pollen (Carine & Scotland 1998;Wood et al. 2003;Wood & Scotland 2009;Hu Chia-chi et al. 2011) and the plietesial flowering pattern of many species (Wood 1994b). Although several species are widely distributed (Bennett et al. 2008), most are rare and localised in their distribution (Wood & Scotland 2009). Most grow in hill forest in areas with a seasonally dry monsoon climate and the genus is noticeably less diverse in tropical rain forest. A few species reach altitudes above 3000 m in the Himalayas and SW China (Hu Chia-chi et al. 2011) but most are found at lower altitudes. Several species are cultivated both in temperate and tropical regions and are occasionally reported as naturalised outside their native range (Wood 2014).
Over the last 25 years, the authors and various colleagues have published a series of papers on the distribution, characteristics, systematics and taxonomy of Strobilanthes but there remain a number of matters that await publication. This paper aims to tidy up miscellaneous issues left over from these studies or which have arisen from the study of specimens received on loan. In particular, two descriptions resulting from Elizabeth Moylan's work on Hemigraphis Nees (Moylan et al. 2002(Moylan et al. , 2004 are still pending. A number of generic transfers are needed and the opportunity is taken to lectotypify several species whose typification is ambiguous. We accept a broad concept of Strobilanthes following Terao (1983) based on our own morphological (Wood 1994b;Carine & Scotland 2002;Wood & Scotland 2009) and molecular studies (Moylan et al. 2004). These results have been confirmed by other studies (Tripp et al. 2013) and are now generally accepted (Hu et al. 2011;Augustine 2018;Deng 2019). Essentially, this means that all genera placed in the Strobilanthinae sensu Bremekamp (1944) belong to the genus Strobilanthes, including species placed both in Hemigraphis and Sericocalyx Bremek.

Materials and Methods
This paper is based on the study of herbarium specimens, principally those received on loan from BM, K and L. Extensive use has been made of on-line resources, particularly the literature available through the Biodiversity Heritage library and the specimen images on Jstor. Images available on the websites of individual herbaria, especially those of Kew (K), Edinburgh Botanic Garden (E), Naturalis in the Netherlands (L, U) and The Natural History Museum, Paris (P) have also proved very useful. All cited specimens have been seen unless indicated to the contrary. The original work of Elizabeth Moylan as part of her DPhil studies of species placed originally in Hemigraphis has also provided valuable insights although she is not responsible for decisions we have made in this paper.
RECOGNITION. Strobilanthes moylaniae is superficially similar to S. reptans (G.Forst.) Y.F.Deng & J.R.I.Wood in its habit and the absence of bracteoles. However, the new species is easily distinguished from S. reptans by the glandular (not multicellular) hairs on the bracts, calyx and ovary, flowers with stamens held at the corolla mouth (not included) and longer filaments that are pubescent (not glabrous) in the lower half. DISTRIBUTION & HABITAT. Papua New Guinea. Light scrubby woodland at c. 90 m. CONSERVATION STATUS. This species is only known from the type collection made over 50 years ago. This may reflect genuine rarity but equally it could have been overlooked because of its strong superficial similarity to Strobilanthes reptans. Any attempt to categorise this species within IUCN (2012) categories would be premature so it must be treated as Data Deficient (DD) until careful search for it is made near the type locality in Papua New Guinea. ETYMOLOGY. This species is named after Elizabeth Moylan who made a major study of species treated as Hemigraphis (Moylan et al. 2002(Moylan et al. , 2004. She separated out this specimen as a putative new species but never described it. NOTE. The pollen of this species (Fig. 2) is clearly ellipsoid and similar in shape and ornamentation to that of Strobilanthes tomentosa (Nees) J.R.I.Wood illustrated by Wood &Scotland 2009: 8 andHu Chia-chi et al. 2011: 382 Subshrub c. 3 m in height; stems stout, woody, much branched, obscurely winged, somewhat pustulate, bifariously scurfy-pubescent, glabrescent when old. Leaves shortly petiolate, 1 -3.5 × 0.5 -1.8 cm, ovate, acuminate, base attenuate and decurrent onto the petiole, margin serrate, both surfaces glabrous except a few appressed hairs on midrib near base, abaxially paler; petioles 1 -5 mm. Inflorescence of numerous, axillary, bracteate cymes mostly with 3 -5 flowers; bracts resembling small leaves, mostly 5 -10 × 3 -5 mm; calyx subequally 5-lobed to the base, glabrous, lobes linear, obtuse, at anthesis c. 7 × 0.5 mm but strongly accrescent to 15 × 1 mm; corolla "mauvewhite", glabrous, ±funnel-shaped, the short basal cylindrical tube 2 -3 × 1.5 mm, then bent slightly and gradually widened for c. 18 mm to 10 mm at the mouth, the limb 5-lobed with rounded ovate lobes c. 2 -3 × 2 -3 mm; stamens didynamous, longer filaments pilose, c. 7 mm long, shorter filaments glabrous, c. 2 mm long; anthers 2 × 0.5 mm, included; style glabrous, 12 mm long. Capsule oblong, 12 -15 × 3 mm, glabrous, 4-seeded; seeds lenticular, pubescent. Fig. 3.

RECOGNITION. Resembling Strobilanthes atropurpurea
Nees in the leaf shape, subequally lobed calyx with linear, obtuse lobes and the glabrous corolla, ovary and capsule but readily distinguished by the very robust, shrubby habit, axillary cymose inflorescences (not flowers in 1-sided racemes, often reduced to solitary axillary flowers) and in the smaller, narrowly funnel-shaped corolla less than 10 mm wide (not strongly ventricose, up to 20 mm wide). DISTRIBUTION & HABITAT. Not known, presumably forest. CONSERVATION STATUS. This species is only known from the type collection made over 65 years ago. This may reflect genuine rarity or the destruction of the original habitat, but equally may reflect the lack of recent collecting in the area. There are reports of coffee plantations in the region but whether these affect the type locality is unknown. It must be classified as Data Deficient within IUCN (2012) guidelines until further information becomes available. ETYMOLOGY. This species is named after our colleague Deng Yunfei, who has published extensively on the Acanthaceae and other aspects of the Chinese flora. He has collaborated with the authors on various papers on Strobilanthes and drew our attention to the distinctiveness of this species. NOTE. This species was separated out as a novelty by Hiroshi Terao but was never published. We were inclined to think of it as a form of the variable Strobilanthes atropurpurea until Deng Yunfei drew our attention to its distinct characteristics. Not only is it far more robust but the inflorescence is not composed of one-sided racemes, nor reduced to single axillary flowers. Most distinct of all is the smaller infundibuliform corolla which lacks the distinct ventricose bulge of S. atropurpurea. It is possible that two other collections from North Sumatra, Bartlett 8494 (US) from Deleng Baroes, Karo and Van Steenis 9349 (L) from Gadjah-Blang Kedjeren in Gajolanden, belong to this species but they differ somewhat from Alston 14958 and a final decision must await further collections, which may help show the range of

RECOGNITION. This undescribed species resembles
Strobilanthes linearifolia (Bremek.) Y.F.Deng (=Hemigraphis ciliata S.Moore) in the leaf shape, small, terminal inflorescence and absence of bracteoles but is a subshrub, the calyx is larger (7 mm long, not < 5.8 mm long) and all filaments are glabrous. NOTE. This species collected almost 150 years ago was separated out by Elizabeth Moylan and appears to be distinct. Although we have not formally described it as new, a full description is provided in the hopes that someone may be motivated to explore the island of Sumba and rediscover it. All the available information is provided above and in the accompanying Fig. 4. Some details are based on C. B. Clarke's sketch which is clearly visible in Fig. 4.
Perennial herb, stems hispid, becoming woody and erect, reaching at least 2 m (fide Schmutz 6165). Leaves 3 -12 × 0.5 -5 cm, oblong-lanceolate, acuminate, narrowed at base into a slightly winged petiole 1 -2 cm long, margin repand, lateral veins c. 7 pairs, both surfaces scabrid, adaxially shiny. Inflorescence terminal and axillary, often forming a loose, leafless branched thyrse, the branches up to 10 cm long, the flowers arranged in short bracteate spikes at the branch tips; spikes 1.5 -4 cm long; bracts 7 -12 × 2 -5 mm, oblong-lanceolate to ovate, acuminate, obtuse, asperous-pilose, the whitish hairs sometimes glandular apically; bracteoles c. 5 × 0.75 mm, linear; calyx c. 8 mm long, slightly unequally lobed to just above the base, lobes linear, 1 mm wide, longest lobe c. 8 mm, shorter lobes 6 mm, pilose with gland-tipped hairs; corolla c. 13 mm long, pubescent, yellow, the lobes c. 3.5 mm long; longer filaments 3.5 mm, shorter filaments 2 mm; anthers included. Capsule oblong in outline, 8.5 -10 × 2 mm, shortly comose with glandtipped hairs, 4 -6-seeded; seeds c. 2 × 1.5 mm, lenticular, pubescent. Bremekamp (1944) treated this species as belonging to Sericocalyx (as S. timorensis (Nees) Bremek.) and divided it into two varieties, var. quadriovulatus Bremek., a superfluous name for the type variety and var. sexovulatus Bremek., the varieties distinguished by the number of ovules, four in the first case but six in the second. As the number of ovules and seeds is quite variable in several species referred previously to Aechmanthera (Wood & Scotland 2009: 16), Hemigraphis (Moylan et al. 2002: 785, 790, 794, 812 etc.) or Sericocalyx (Kurz 1871: 74 in discussion of Strobilanthes flava), this character is of little value, particularly as it shows no correlation with other characters or any geographical or ecological patterning. However, examination of around 25 specimens of Strobilamthes timorensis indicates that this species can be divided into two geographically distinct subspecies based on the shape of the bracts, lanceolate to oblong in the type subspecies but ovate to elliptic in the subspecies ovata described below. Subsp. timorensis is more eastern in its distribution than subsp. ovata.  PHENOLOGY. Subspecies timorensis apparently flowers in the May to July period. NOTES. The field notes accompanying Schmutz 6165 comment: "On July 9 1987 there were thickets of this plant in flower. If not distressed by Eupatorium odoratum, (they) will grow up to as tall as a man and more, and as thick as a walking stick, suitable for firewood ……One year later, there are big masses of seedlings, sized between 5 and 22 cm…..(on) August 27 1988 in Pouggeok/Ulung at 700 m …. I met a few flowering specimens. They always reveal as being side shoots from early (before simultaneous flowering) cut main stems. Such often happens close to the road when people are cleaning up the road by cutting away all the weeds." He also noted seven-year flowering cycles and an unfounded folk legend that flowering announces a prolonged dry season (Schmutz 1988: 19). These field notes indicate quite clearly that this is a plietesial species, flowering gregariously in an approximately seven-year cycle. Although plietesial flowering has been known in Strobilanthes for two centuries (Bremekamp 1944;Wood 1994b), Schmutz' brief paper is only the second note of the phenomenon in the Malesian region after Van Steenis (1942
Although the lectotype has no original label, it was collected by Osbeck as indicated by Linnaeus (1753: 635). Osbeck is known to have visited Java and the collection number has been added as a result of studies by Hansen & Fox Maule (1973). Strobilanthes crispa is apparently native in Java and Madura but has been cultivated elsewhere. Clarke (in sched.) identified the Osbeck collection as S. scabra but that species differs in the lanceolate scabrid (rather than ovate, pilose) bracts and is not known to occur outside cultivation in Indonesia. Benoist (1922: 96). Type: China, Yunnan, Environs de Lan Ngui Tsi (?), 9 Aug. 1904, Pius Pi in Ducloux 2603 (lectotype P-00719313, selected here).

Strobilanthes duclouxii
There are at least two isolectotypes at P and several specimens of the syntypes, all corresponding to the same species but the only one with Benoist's handwriting is the lectotype. It should be noted that not all specimens were distributed with the numbers cited by Benoist. We can only assume that numbers were changed at some stage during labelling or distribution.
Strobilanthes duclouxii is a synonym of S. extensa Nees.
The typification of Strobilanthes flava is somewhat complicated. It was originally described as Ruellia flava by Roxburgh, but no specimen was preserved so the only possible type is the Roxburgh painting at Kew. Ruellia flava Roxb. is illegitimate, being a later homonym of R. flava Pers. Strobilanthes flava was first cited with Ruellia flava Roxb. as basionym by Kurz (1870: 78) where he treated it erroneously as the correct name for Strobilanthes scabra Nees. Kurz later provided a full description of S. flava in the Forest Flora of British Burma (Kurz 1877: 243). The specimen designated above as epitype (Kurz 2130) is from the same area (Pegu Yoma) as the Carey collection but was collected and annotated by Kurz. It should serve to resolve any ambiguity in the interpretation of the Roxburgh plate.
Unfortunately, when this was lectotypified by Moylan et al. (2002), it was not noted that there were three separate collections by Kloss mounted on the same sheet. In order to avoid ambiguity, a second step lectotypification is made above selecting the collection in the upper part of the sheet as the lectotype.  Candolle & Radcliffe-Smith 1981), even though he rarely annotated them. Contrastingly, the specimens in Graz are annotated by Nees but probably at a date after 1840, as Strobilanthes is treated as masculine in these annotations whereas in the 1830s he always treated the generic name as feminine, as did Blume, the original author (Wood 1994a).
There is an issue over the type location of Sylhet. This lies within the borders of Bangladesh. Most specimens of Strobilanthes labelled as from Sylhet originate from the Khasi Hills to the north, which lie within India. However, S. scabra is a plant of the Bengal plains so the type location probably lies within Bangladesh.
Anderson cited five different collections as syntypes but only two of these at Kew bear his annotation, a Griffith collection and Simons 23, which he specifically cites as having seen at Kew. In view of this and the fact that the Griffith collection lacks an inflorescence, Simons 23 is designated as lectotype.
Strobilanthes thomsonii T. Anderson (1867: 478 Anderson cited seven collections as syntypes of Stroboilanthes thomsonii and we have selected as lectotype the only specimen we have seen with Anderson's annotation. Anderson drew attention to the variation in Strobilanthes thomsonii as did Clarke (1884: 467). Very robust forms, such as Wood 6824 (E, FHO) and 7496 (E, FHO), both from Bhutan, may correspond with Anderson's var. serratifolius. However, the inflorescence can be very variable, sometimes lax and formed of branched axillary racemes as in Wood 6824 but sometimes condensed, unbranched and ±fasciculate as in Wood 7495 (E, FHO), also from Bhutan. The corolla of the last specimen is around 3 cm long whereas in Wood 5971 (E, FHO) from Bhutan the inflorescence is slender and the corolla a mere 2 cm long.

Strobilanthes yunnanensis
This specimen is chosen here as lectotype rather than the other cited specimen, Forrest (7)616, partly because it is the more ample specimen and partly because the number (7)616 was not cited correctly.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/ licenses/by/4.0/.