New species and nomenclatural changes in Angraecum (Orchidaceae) from Madagascar

Three new species: Angraecum archangelicum, A. polyphemus and A. rotundifolium are described for the first time. Pectinariella is evaluated; a case is made to keep it at subgeneric level and the necessary two new combinations are made. Angraecum dasycarpum is neotypified. Finally A. ochraceum and A. setipes are compared with the latter being reduced to a synonym of the former.

Recent studies on the genus at Kew, the Missouri Botanical Garden, the University of Vienna, and PBZT Antananarivo have resulted in the recognition of a number of new species and a better understanding of their nomenclature. Three new species are described and several nomenclatural changes are outlined here.
Because of the limitations of herbarium material available for completing some of the drawings, it was not always possible to represent all the detail of the gynostemium; more detailed notes from field observations were added to the text.

IUCN Red List assessments
The conservation status of the new species given in this paper are summaries of the full IUCN Red List assessments which will be completed and submitted for review and publication by IUCN once the species names are validly published and therefore available for assessment. All the assessments have been compiled based on current knowledge of these taxa, using the IUCN Red List Categories andCriteria (2012, 2017).
RECOGNITION. With its sub-acaulous habit and relatively small flowers on an elongate peduncle the new species belongs in section Boryangraecum Schltr. (Schlechter 1925: 308). Within the section it is distinguished by the combination of its relatively large habit with long narrow leaves and its pendent, long, rigid peduncle. It can also be recognised by its non-resupinate flowers, relatively large erect petals, the broad central vein within the lip and its clavate spur. It is closest to Angraecum aviceps Schltr. (Schlechter 1925: 335) known from Bemarivo in the NE of Madagascar. Schlechter's description is short and the type specimen has been missing from P for many years. From the description it is clear that its leaves are shorter, the inflorescence has more flowers (3 -5 vs 1 -2) and is produced earlier in the year, and the flowers are a little smaller, with smaller petals and a subfiliform spur (vs clavate). It is also similar to A. calceolus Thouars (1822: t. 78) but the habit and inflorescence of that species are very different, the spur is generally longer (much longer than the ovary vs about the same length), the dorsal sepal shorter (6 -8 mm vs over 10 mm), the petals not erect and the lip of a different shape. It also shows some similarities in habit and inflorescence to A. ochraceum (Ridl.) Schltr. (Schlechter 1915: 436) but that species has much smaller flowers of a different shape and a spur at least 3 times longer than the ovary. DISTRIBUTION. The new species is endemic to central Madagascar; it is known from the type specimen but also from photographs taken by Jean-Michel Hervouet and Gilles Grunenwald in Ambondrombe and Ankaratra respectively. SPECIMEN EXAMINED. MADAGASCAR. Fianarantsoa prov., Antoetra, Feb. 2015, 1727 Antananarivo and Fianarantsoa provinces. It occurs in three defined threat locations and is not recorded from any Protected Areas. The extent of occurrence EOO, the area of occupancy AOO and the habitat quality are currently not known but are very likely in continuing decline due to major threats such as frequent fires. Therefore, it is assessed as Endangered (EN) under criterion B1ab(i,ii,iii)+ 2ab(i,ii,iii) according to the IUCN Red List Categories and Criteria. FLOWERING TIME. February -April. ETYMOLOGY. Referring to the resemblance of the flowers to a group of descending archangels. NOTES. The species was briefly described and illustrated by Guérin & Hervouet (2011: 97) and Hervouet (2018: 153) as an unidentified species (Angraecum sp. 2). Semi-erect or pendent flabellate epiphytic herb up to 5 cm long, on a very short stem; roots wiry, cylindrical, glabrous, c. 1 mm in diam. Stem 3 -8 × 2 -4 mm, flattened, the base covered with dried leaf and peduncle sheaths. Leaves up to 9, distichous, twisted at the base so as to lie in a single plane containing the stem axis, conduplicate at base at the transition to the leaf sheaths, ligulate to lanceolate, 10 -29 × 2.5 -6 mm, the margins a little thinner, flat canaliculate on top, roundly keeled below with the mid-vein ending in an apicule before the attenuate unequally bilobed leaf tip, on a short 2 -4 × 1 -1.2 mm petiole. Inflorescence axillary, short, diverging, 10 -15 mm long, single or rarely 2-flowered with an embryonic bud at the apex. Peduncle slender, 6 -9 mm, flattened, capitate at the tip, persistent, peduncle sheath attenuate, 2.5 -3.9 × 1 -1.5 mm. Floral bracts lanceolate, 2.5 -4 × 1.1 -1.8 mm, attenuate. Flowers non-resupinate with the lip uppermost, 8 -11 × 6 -8 mm, white except for the apex of the spur, which is green and, the anther which has yellowish green cross-shaped bands in the middle. Pedicel and ovary obconical, incurved, glabrous, 6.5 -8.3 × 1 -1.3 mm. Dorsal sepal convex, obovate to elliptic-obovate, 4.8 -5.1 × 1.9 -2.2 mm, apiculate at the apex. Lateral sepals convex, obovate to ellipticobovate, 4.6 -4.9 × 2.2 -2.4 mm, apiculate at the apex. Petals oblong-obovate, 4.5 -5 × 1.6 -1.8 mm, subapiculate at the apex. Lip 3-lobed with the lateral lobes erect and parallel to the column and the midlobe upturned, 4.9 -5.4 × 5.2 -5.6 mm, the lateral lobes sinuate at the anterior margin, the mid-lobe sinuate to undulate at the margin, emarginate to acute at the apex, the base with a few scattered thin hairs; spur almost parallel to the ovary, clavate, rounded at the apex, slightly arcuate, 4.3 -4.7 × 1.2 -1.9 mm. Column ovoid, 2 -2.2 × 1.7 -1.8 mm, lateral lobes of the rostellum roundly lobed, c. 1.5 mm long, midlobe lobular, stigma ovoid, c. 0.8 mm in diam.; anther cap elliptic, produced with a long bilobed part at the front, with two distinct chambers, 1.7 -1.9 × 1.2 -1.4 mm; pollinia 2, ovoid, c. 1 × 0.7 mm, with two lorate stipes and a small angular viscidium. Figs 3, 4.

Angraecum polyphemus
RECOGNITION. Angraecum polyphemus belongs in section Pectinaria (Benth.) Schltr. (Schlechter 1918b: 157) for its habit and usually single-flowered on a short peduncle. It is distinct by its flabellate habit on a very short stem with relatively wide ligulate to lanceolate flat leaves lacking a dorsal wing, and its comparatively large flower(s) with the lip uppermost, with broad rounded side lobes, a sinuate anterior margin with a few hairs at the base and a relatively long clavate spur and an anther with a large lobe at the front. It is  Hermans & Verlynde, including the short peduncle, 1 to rarely 2 flowers, thickened apicule at the apex of petals and sepals, 3-lobed lip with hairs at the base and relatively short thickened spur. It differs from all three by its very short stem, flat ligulate-lanceolate leaves that are roundly keeled (vs thin-flattened below), the different arrangements of the leaves which are generally fewer in number, twisted at the base and arranged in one plane containing the stem axis (vs not twisted at the base and arranged in more than one plane perpendicular to the stem axis), the larger flowers (twice the size of A. edmundi and A. scroticalcar, and about a quarter larger than A. pterophyllum), the lateral lobes of the lip less distinct with sinuate anterior margins (vs entire (except in A. edmundi)), fewer hairs at the base of the lip. The spur is similar but is more elongate than the other species (almost twice the length) and more arching. Table 1. polyphemus is estimated to be over 24,749 km 2 (which exceeds the limits for Vulnerable status under criterion B1) whereas its minimal area of occupancy (AOO) is estimated to be 24 km 2 (which falls within the limits for Endangered status under the criterion B2). This species is known from two specimens collected in distinct protected areas, one from the Parc National de la Montagne d'Ambre, and one from the Corridor Ankeniheny Zahamena. In addition to these officially protected sites, eight specimens of this species have also been collected in unprotected forests subjected to forest clearing, selective logging, timber harvesting for small-scale subsistence and tavy (slash-and-burn farming). This species is thus known from ten specimens representing three existing and one partially extirpated subpopulation. These four subpopulations represent a total of four "locations" (sensu IUCN 2012) with respect to the main threat of habitat destruction, falling within the limit for Endangered status. We estimate that the past loss of its habitat, which will continue, has and will induce a strong continuous decline in habitat quality, number of subpopulations and mature individuals in the next ten years as well as a decline of its AOO. Angraecum polyphemus is therefore assigned a status of EN B2ab(i,ii,iii,iv,v). FLOWERING TIME. January to May. ETYMOLOGY. Referring to Polyphemus from Greek mythology, with the central column and anther cap resembling the eye of the monster and the clubshaped spur the weapon often carried by the same mythical creature. NOTES. Flowering plants of this species have recently (Jan. 2017) been observed in the Montagne d'Ambre RECOGNITION. A small branching plant, lying flat on the substrate, and with small elliptic to oblong leathery leaves, a short single-flowered inflorescence, and small flowers with an echinate to shortly hirsute ovary, an ovate lip roundly lobed at the base, a spur that is cylindrical towards the base, then scrotiform and flattened, and a column with falciform rostellum lobes folded at the apex. With its habit and solitary small flower on a short peduncle Angraecum rotundifolium is placed in section Pectinaria of the genus. Its flowers are somewhat similar to those of A. pectinatum Thouars (1822: t. 51) and A. humblotianum (Finet) Schltr. (Schlechter 1915: 434) but its habit is very different with broad, rounded leaves (vs elongate-linear or almost terete) and flowers with a spur that is also a little shorter and scrotiform (vs cylindrical). With its branching leafy stem and small flower it has frequently been confused with the superficially similar A. dasycarpum Schltr. (Schlechter 1918a: 337) circumscribed below, but it is different in several characteristics which are not always obvious in herbarium materials: the plant grows flat against the substrate (vs the somewhat erect growth of A. dasycarpum), and has flat, elliptic to oblong leaves (vs ovate-lanceolate slightly divergent to the stem), tepals that are consistently around half the size, a scrotiform spur (vs cylindrical narrowing towards the tip), an echinate to shortly hirsute ovary (vs longly hirsute) and a column with prominent angular ribs (vs rounded). Its distribution is also different, with A. rotundifolium recorded only from the eastern forest near Andasibe in Toamasina province at c. 600 -1000 m where it flowers from February to May, whereas A. dasycarpum is found in north-eastern Madagascar in lower elevation forest where it flowers from August to November.  less than 500 km 2 , the extent of occurrence EOO less than 5,000 km 2 and three threat locations defined. The species is threatened by habitat destruction due to logging forest and shifting agriculture that cause the continuing decline in the EOO, AOO and the habitat quality. This species is therefore assessed as   Kuntze (1903: 200) in Epidorchis sect. Pectinaria. Garay (1973: 499) in his review of the genus Angraecum defined section Pectinaria, based on Bentham's characteristics as follows: "stem elongate, loosely leafy, leaves alternate, fleshy, elongate, linear or with distinct blade, but never equitant, inflorescence one-flowered, peduncle very short or scarcely developed, sessile, flowers small". Micheneau et al. (2008) in their analysis of the angraecoid orchids, using a limited number of samples, concluded that the section does not appear to be natural and is represented in two clades. Cribb in Pridgeon et al. (2014: 362), based on Micheneau et al. (2008), also concluded that the section may be separated from Angraecum sensu stricto when further information became available. Szlachetko et al. (2013) in their re-classification of the Angraecum-alliance picked up on the comments by previous authors to describe four new genera and elevated the section Pectinaria to generic rank as the new genus Pectinariella (misspelled as 'Pectianriella' in the abstract), together with another ten sections of Angraecum. All the species placed in the section Pectinaria by Garay were transferred without substantiation to Pectinariella. Angraecum pterophyllum H. Perrier (1938: 106) was not considered by Szlachetko et al. (2013), probably because it was listed as a synonym of A. hermannii (Cordem.) Schltr. (Schlechter 1915: 424) by Garay. Accordingly the same authors did not acknowledge that A. hermannii was transferred to section Lemurangis Garay (1973: 501) by Micheneau et al. (2008: 915) and that A. pterophyllum was recognised as a valid species by them. Szlachetko et al. (2013: 18) stated that the new genus is "easily distinguishable from all other Angraecinae by its peculiar habit that is fleshy leaves, basally twisted with almost sessile flowers forming single-flowered inflorescence and ecallose lip". In the accompanying key the leaves are said to be conduplicate, the spur filiform and mostly longer than the ovary, whilst in the description the spur is said by the same authors to be predominantly subequal in length to the ovary and the leaves linear to linear-lanceolate. These characters do not apply to several species that were included in Pectinariella: the leaves of A. dasycarpum are flat, ovate-lanceolate and not conduplicate, linear to linear-lanceolate; the spur of A. gabonense Summerh. (Summerhayes 1954: 587) and A. dasycarpum are about equal to the length of the ovary; A. subulatum Lindl. (Lindley 1837: 206) and several other recently described 'Pectinariella' have two-flowered inflorescences (vs single-flowered) (Verlynde et al. 2016: 228). In addition, the definition could apply to a number of species traditionally assigned to other sections, including A. baronii (Finet) Schltr. (Schlechter 1915: 431). It is also puzzling that Szlachetko et al. (2013) included A. dasycarpum in their new genus despite the evidence presented in their cladogram (2013: 6) placing it amongst a group of morphologically very different species in section Perrierangraecum Schltr. (Schlechter 1925: 309). This could be due to an error in identification. No voucher information is provided for the specimens used in the research to account for possible misidentifications. Simo-Droissart et al. (2013) confirmed the polyphyly of Angraecum section Pectinaria and found the continental African species to be closely related to section Dolabrifolia (Pfitzer) Garay (1973: 499). Verlynde et al. The acceptance of Pectinariella is here brought into doubt by a number of factors: the flawed delimitation of Pectinariella; the lack of voucher specimens cited in some of the research outlined above; the doubt over some identifications; the limited number of species included in the analyses; the lack of consistency of the morphological features cited; and the distribution of the different species included in the genus. It is likely that some of the species of Pectinariella belong in different sections or genera. In combination with this lack of evidence, we consider that it is desirable to maintain the large genus Angraecum, instead of a number of small genera with overlapping morphologies, especially as the phylogenetic analysis of Angraecum s.l. is still incomplete. Several new species have been added recently to the section and the relationship between the continental African species and those from Madagascar and surrounding islands, and therefore with the type species, is still unclear. It is therefore prudent to maintain section Pectinaria of Angraecum, excluding the species recently transferred to Afropectinariella, for the time being. In agreement with this treatment the necessary new combinations are made below.
It is likely that further research will show that the species allied to Angraecum pterophyllum with an uppermost lip, an inflorescence with one or two flowers and an anther cap with a lobular extension, need to be separated from those allied to A. pectinatum with single flowers and a lip similar to the petals and sepals.
(Bentham 1883: 585). Pectinaria (Benth.) Cordem. (Cordemoy 1899: 412) nom. illeg. Epidorchis sect. Pectinaria (Benth.) Kuntze (1903: 200 Erect, caespitose, epiphytic or lithophytic herb, up to 14 cm high, on a short branching stem; roots wiry, branching, greyish-white, puberulent, 0.5 -1 mm in diam. Stem somewhat fleshy, branching, up to 12 cm × 3 -4 mm, flattened, the margins winged, carrying up to 20 leaves, pale green. Leaves fleshy, alternate, distichous, horizontal or slightly divergent to the stem, ovate-lanceolate, 7 -10 × 2.3 mm, flat with an indistinct mid-vein, with a central ridge underneath, almost sessile at the base, unequally bilobed at the tip, pale green, faveolate with dark pits on the upper and lower surface and a darker mid-vein. Inflorescence single-flowered, leaf-opposed, almost sessile with a 1.5 -3 mm peduncle. Sheaths 1 to rarely 2, amplexicaul, lanceolate, 5 -7 × 2 -3 mm, compressed, brown. Flowers stellate, 16 -21 × 13 -20 mm, pure white with the column and anther yellowing with age. Pedicel and ovary fusiform, 5 -7 × 1.1 -1.9 mm, densely pubescent (appearing shortly hirsute in dried specimens). Dorsal sepal lanceolate, acute, 7 -8.3 × 2 -4.1 mm, recurved toward the tip. Lateral sepals oblong, acute, 7 -10 × 2 -4 mm, divergent. Petals ligulate, 5 -8.5 × 1.2 -3 mm, recurved toward the tip, slightly concave. Lip narrowly elliptic, 5.5 -10 × 2 -3.6 mm, slightly narrowed in the middle, with long narrow wings towards the base, apex acute and a little recurved; spur parallel to the ovary, cylindrical, slightly narrowing towards the obtuse apex, 3.1 -4 × 0.8 -1 mm. Column very short, rounded, 1.1 -2 × 1. HABITAT. Lowland evergreen coastal forest, on rocks covered in moss, on Philippia, in shade. Altitude 0 -300 m. CONSERVATION STATUS. Angraecum dasycarpum is endemic to Madagascar, distributed in Analanjirofo region, Toamasina province. It does not occur in any Protected Area. The area of occupancy AOO is estimated to be less than 500 km 2 , the extent of occurrence EOO less than 5,000 km 2 and three threat locations defined. The species is threatened by habitat destruction due to logging forest and shifting agriculture that cause the continuing decline in the EOO, AOO and the habitat quality. This species is therefore assessed as Endangered (EN) under criterion B1ab(i,ii,iii)+ 2ab(i,ii,iii). FLOWERING TIME. August to November. ETYMOLOGY. Referring to the rough-pubescent seed capsule/ovary. NOTES. Angraecum dasycarpum, in section Pectinaria, is characterised by its caespitose branching plant with fleshy ovate-lanceolate distichous leaves, its stellate flowers that are relatively large for the section, with a pubescent ovary covered with soft hairs, an elliptic lip, a cylindrical spur narrowed towards the apex and a short column with an obtuse central lobe. It is a low-elevation species from the more tropical north-east of Madagascar. The flowers are similar to those of A. pectinatum and A. humblotianum but it has a different habit with broader and rounder leaves, and a distinct lip and spur shape. Other differences are discussed above under A. rotundifolium and shown in Table 2. Angraecum dasycarpum was described by Schlechter (1918a) together with a group of other orchids from Madagascar collected by Laggiara. Although he was involved in describing almost half of the orchid flora of Madagascar, Schlechter never visited the island. He relied on other explorers for his material; mostly from the Frenchman Henri Perrier de la Bâthie but also from a few others. One of these was Dr Paolo Ferko, an Italian orchid grower and occasional importer who in turn received orchids collected in Madagascar by a Mr Laggiara. Not much is known about the collector apart from the fact that he collected in north-eastern Madagascar at the beginning of the 20 th century (Dorr 1997: 251). Very little of the Schlechter herbarium has survived in Berlin (B) and the holotype specimen of A. dasycarpum must therefore be considered destroyed in the 1943 fire (Butzin 1978(Butzin , 1981; no Laggiara specimens have been found in other herbaria. The only surviving material associated with the type is the small illustration by Schlechter (1932: t. 88 no 352): the drawing is not detailed but it does show the main characteristics of the flower. It is possible that the illustration was based on material associated with the type but there is no direct evidence of this, furthermore, it was published well after the protologue and for these reasons cannot be considered as a lectotype. Schlechter's 1918a description is also limited in detail but, combined with the illustration, does provide sufficient information to identify the species with some certainty. A specimen collected by Bosser in 1963 (Bosser 17546 in P!) is designated here as the neotype of A. dasycarpum because it is well-documented, consists of several complete plants and flowers showing all the typical characteristics of the species. Other complete specimens were collected by Walter Rauh in northern Madagascar (Rauh M211 (HEID-spirit)) and illustrated in Senghas (1997: 103). Szlachetko et al. (2013) transferred the species to their new genus Pectinariella, which is here placed back in Angraecum; the reasons for this are discussed above under the evaluation of the status of Pectinariella.
The thin single flowered-inflorescence and flower shape are similar to that of Angraecum rhynchoglossum Schltr. (Schlechter 1925: 339) but in that species the leaves are consistently shorter, smaller and oblong (vs ligulate), and the flowers are about a third larger. Other stemless species with long narrow leaves, a long thin peduncle and a single (rarely two) flower include: A. andasibeense H. Perrier (1938: 124), which has thinner and shorter, linear leaves (vs ligulate), flowers that are a third smaller with an ovate, acuminate lip (vs ovaterhombic) and an anther cap with a well-developed lobe at the front (vs entire); A. brachyrhopalon Schltr. (Schlechter 1925: 336) which is a slightly shorter plant, with narrowly linear leaves (vs ligulate), flowers that are a third smaller and have an incurved clavate spur (vs straight and slightly thickened at the end); and A. pumilio Schltr. (Schlechter 1916: 33) which is smaller by a third in all aspects and the 7 mm long spur is around half the size (vs at least 14 mm long).
Mystacidium ochraceum was described by Henry Ridley in 1885, based on a collection by the Rev. William Deans Cowan from 'Ankafana' in Madagascar. Ankafana is a general name for the missionary post in the Fianarantsoa area of Madagascar (Christenson 1932: 9). The type specimen at BM is sterile with an indication of a flower in spirit but this has not been located and presumed lost. However, a watercolour ('no 7') can be found in Deans Cowan's sketchbook at the London Natural History Museum, London (Cowan 1880: 7) (Fig. 9), showing a flowering plant of the species which corresponds well with one of the plants on the herbarium sheet, it is annotated 'Mystacidium ochraceum Ridley' in Ridley's hand. On the type specimen Ridley also indicated the number '7', referring to Deans Cowan's watercolour, it can therefore be considered part of the type material used by Ridley especially as it complements the limited herbarium material and description; the illustration is therefore designated here as an epitype. In 1907 Finet placed the species in the genus Macroplectrum but based it on a Baron specimen donated by K to P. Schlechter (1918b) transferred the species to Angraecum. It is unlikely that Schlechter (1925) was familiar with the Deans Cowan specimen, let alone the watercolour, when he described Angraecum setipes, based on a Perrier specimen from the Ankaratra mountains in central Madagascar. When comparing the two taxa there is no doubt that they are conspecific, with the name A. ochraceum having precedence. In the broad range of specimens examined, it was found that the lip shape is a little variable with some flowers having a more deeply concave lip with a more pronounced acuminate apex but this variation occurred in different flowers on the same plant and may be dependent on the maturity of the flowers. The length of the spur too was found to be variable ranging from 14 -22 mm: this was dependent on the stage of development of the flowers and also some variants consistently had a shorter spur but also smaller flowers. This variation is not unusual in other Angraecum species: on examination of over 25 specimens of the similar A. rhynchoglossum it was found that spur length ranged from 15 -28 mm while in A. calceolus Thouars (1822: t. 78) the range was 10 -18 mm. The thickening towards the end of the spur was also found to be variable but it was consistently present. Considering this variability it seems reasonable to treat A. setipes as a synonym of a widespread and variable A. ochraceum.
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/ .