Chlorophytum ankarense H. Perrier

2. Chlorophytum ankarense H.Perrier Figs 4A, 11 Chlorophytum ankarense H.Perrier (Perrier de la Bâthie 1935: 47). Etymology The epithet refers to the Ankara Plateau (NW Madagascar). Type material MADAGASCAR – Boeny [Mahajanga Prov.] • Rochers (calcaires) boisés de Kamakama, causses d’Ankara; [16°55′ S, 46°28′ E]; Mar. 1901; fl; H. Perrier de La Bâthie 10948; holotype: P [P01046141]!. Other material examined MADAGASCAR – Androy [Toliara Prov.] • Ampandrandava; 1943; fl; Jardin botanique 5989; TAN! • Crête des Pisopiso, Nord d’Ampandrandava; [24°04′ S, 45°42′ E]; 900 m; Jan. 1943; fl; A. Seyrig 521; P [P02071689]! • same data as for preceding; 1943 (en culture); fl, fr; A. Seyrig 521B; P [P02071695]!. – Atsimo-Andrefana [Toliara Prov.] • Sud-ouest Manombo; [22°57′44″ S, 43°29′58″ E]; 26 Jan. 1947; fr; H. Humbert 19977; P [P02158247]!, MO [MO-3020335, MO-3020336] image!. – Boeny [Mahajanga Prov.] • Massif de l’Ankarafantsika; [16°19′ S, 46°48′ E]; 19 Jan 1938; fr; R. Decary 12828; P [P02157231]!, TAN! • RN 8, Soalala, Andranomava; [16°28′ S, 45°20′ E]; 27 Mar. 1954; fr; RN 6144 (Rakotovao); P [P02157237]! • same data as for preceding; [16°34′ S, 45°34′ E]; 10 Mar. 1955; fl, fr; RN 7115 (H. Randriamiera); P [P 02157232, P02157224]!. Description Plant herbaceous, 13–38 cm tall; roots fasciculated, fibrous, slightly thickened, 0.5–0.7 mm, bearing infrequent fusiform tubers, 8–10 × 1–2 mm; leaves 4–13, lanceolate, 13–38 × 1.5–3 cm, base attenuated into a pseudo-petiole of 4.5–10 cm long, approximately 1/3 to 1/4 of the total length of the leaf blade, leaf-base sheathing, surrounded by a fibrous sleeve composed of the remains of old leaves, leaf margins hyaline, mostly scabrous (except for RN 6144 where the leaf margin is smooth), leaf apex tapered, with a pseudo-midrib, comprising 24–34 evenly-spaced, conspicuous veins and with a network of transverse veinlets; inflorescence 1(–2) per plant, erect, branched, robust, 32–60 cm in height; central axis and lateral branches covered with hyaline, more or less papillose hairs up to 0.1 mm long, mixed with short excrescences of variable density differing between the specimens, peduncle 18–32 cm long, with 1–3(–4), triangular, often long-tapered bracts, 7–40 × 2.5–5 mm at the base, veins 8–16, rachis 16–30 cm long; flowers 2–5(–7) per node, floral bracts triangular, very slender, reddish, 2.5–6(–10) × 1.5–3 mm, veins 3–5, shorter than the pedicel, the latter (3–) 4–8 mm long, generally articulated below the middle, or occasionally ca 1.5 mm from the base (e.g., Seyrig 521 and 521B), pedicel glabrous; perianth white or very pale green; tepals narrowly elliptical, equal or subequal, 4.5–5.7 × 0.5–0.8 mm; veins 3, discreet; stamens unequal, shorter than the tepals, in two groups – 3 long (2.5–4 mm) and 3 short (1.8–3 mm), filaments flattened at the base, spindle-shaped, thickened, papillose above, anthers with a triangular base, sometimes curved at anthesis, 0.5–1 × 0.2–0.3 mm, divergent at the base, anthers dorsifixed, inserted on the filament at the lower 1/3; gynoecium: ovary, 0.5–0.8 × 0.6–1 mm; ovules 1(–2) per locule, style filiform, 4.2–6 mm long, exserted; stigma with thickened tip; capsules pendulous at maturity, distinctly wider than tall, 3–5 × 5–7 mm, often with one or two aborted locules, with 4–6 thick transverse ribs; seeds 1 per locule, black, somewhat discoid, 2–2.8 mm, with a convex surface and a small depression on the flat side behind the hilum, shiny, granulate, with a narrow notch, where the radicle is visible, but not protruding. Phenology Flowering has been recorded from December to March, and fruiting from January, and continues beyond the month of March. Distribution and ecology Chlorophytum ankarense has a highly disjunct known distribution. It has been recorded in the western parts of Madagascar in Boeny and Atsimo-Andrefana Regions, as well as in the Androy Region in the south-central part of the country, near Ampandrandava. It occurs at elevations ranging from 100 to 900 m above sea level, on both sandy and limestone substrates, but always in dry forest, and generally in well-shaded areas. The disjunct and climatically heterogeneous distribution of the species does raise questions, and furthermore the available material is limited, and no recent collections of C. ankarense are known. Close to half of the available collections consist of separate elements, including just a few leaves, inflorescences, or a single infructescence. Observing similarities in the hairiness of the floral axes, the size and arrangement of the floral parts, the appearance of the fruit and seeds, we group together the cited elements as a single species without absolute certainty, as not all these characteristics are consistently present together, and new material is needed to confirm the species delimitation. Provisional IUCN Red List assessment Chlorophytum ankarense is endemic to Madagascar. The species is known from a total of 9 collections that represent 5 occurrences. These show a highly disjunct distribution with one group of records from the south of the country, and a second group from the northeast. The species is not known to have been collected or observed since 1955. Furthemore, the known collection sites are mostly not considered to be badly degraded, and there are records of the species from the following three formally protected areas: the Parc National d’Ankarafantsika and Parc National de Namoroka in the west, and the Aire Protégée de Ranobe PK 32 in the southwest. The two other known records are from areas lacking formal protection, the Causse d’Ankara in the west – which nevertheless has relatively intact vegetation, and Ampandrandava in the south, where, in contrast, the vegetation is badly degraded. We therefore consider the 5 known subpopulations to represent 5 locations due to differing management levels of the corresponding sites, due to there diverse conditions and wide separation. With an AOO of only 20 km 2, but an EOO of over 150 000 km 2, Chlorophytum ankarense could be assessed as Endangered [EN B2ab(i,iii,iv,v)], however, given the lack of any recent material of the species, we propose an assessment of Data Deficient [DD]. Nevertheless, we suspect that this rather inconspicuous plant is likely to be more frequent across its broad range than the limited number of available collections would suggest, and we recommend a targeted search for the species in the vicinities of the known localities.