Miconia obtusa (Griseb.) Triana

Bécquer, Eldis R., Bochorny, Thuane, Gavrutenko, Maria & Michelangeli, Fabián A., 2022, A revision of the “ basal-axile placentation clade ” of Miconieae, the newly erected Miconia sect. Liogieria (Melastomataceae: Miconieae) from the Greater Antilles, Willdenowia 52 (3), pp. 387-432 : 411-415

8. Miconia obtusa (Griseb.) Triana in Trans. Linn. Soc. London 28: 103. 1871 ≡ Graffenrieda obtusa Griseb., Cat. Pl. Cub. : 103. 1866 ≡ Acinodendron obtusum (Griseb.) Kuntze, Revis. Gen. Pl. 2: 952. 1891. – Lectotype (designated here): Cuba, Cuba oriental [Prov. Guantánamo], Pinal near Baracoa, 15 Jun 1861 [as appearing on specimen in GOET and field note of Wright on specimen in GH], C. Wright 2527 (GOET 7029!; dubious isolectotypes: BM 1008010 [photo!], BR 5212330 [photo!], G-DC 00316986 [1] and 00316987 [2] on same sheet [photo!], G-DC 00316994 [photo!], GH 72277!, GOET [C. Wright 709]!, K 535933!, MO 313842 [photo!], NY 99580!, NY 99581!, S 5-3633, S 9-39614 [a] and S 9-12537 [b] on same sheet [photo!], UC 936605!, US 120414!, YU 65038!). – Fig. 15. = Miconia confusa Cogn. in Candolle & Candolle,

Monogr. Phan. 7: 759. 1891. – Lectotype (designat-

ed here): Cuba oriental, without precise locality or date, C. Wright 2527 p.p. (BR 5190799 [photo!]). =

Miconia cerasiflora var. setulifera Urb., Symb. Antill. 9: 112. 1923. – Lectotype (designated here): Cuba, Prov. Oriente [Guantánamo], Baracoa in collibus pineti prope El Yunque, 20 Nov 1914, E. L. Ekman 3567 ( S 9-12764 [photo!]).

Morphological description — Shrubs 1–1.7 m tall, branched, deciduous. Indumentum of stellate and dendritic trichomes 0.1–0.2 mm long, light brown or ferruginous, and elongate glandular trichomes 1–3 mm long, reddish to purple, on young branches, leaves, inflorescences, flowers and fruits. Young branches terete, densely glandular tomentose, later glabrescent, glandular trichomes more persistent than other trichomes. Mature branches with conspicuous longitudinal lenticels, sometimes setulose. Petiole 0.7–1.7 cm long, terete, canaliculate adaxially, sparsely glandular tomentose; leaf blade 5.5–10(–13) × 3.5–6.5(–11) cm, ovate-elliptic to broadly elliptic, subcoriaceous, base broadly cuneate, rounded to slightly cordate, apex obtuse to rounded, sometimes acute to acuminate, margin flat to slightly revolute, entire; adaxial surface flat to bullate, sparsely to moderately glandular tomentose; abaxial surface sparsely tomentose mostly on veins, glandular trichomes scattered. Venation with 2(or 3) pairs of basal secondary veins, symmetric, innermost conspicuous, marginal pair sometimes inconspicuous; midvein and secondary veins slightly impressed adaxially and prominent abaxially, tertiary veins flat to slightly impressed adaxially and slightly prominent abaxially, quaternary veins mostly inconspicuous. Mite domatia present as a cavity between midvein and pairs of secondary veins. Inflorescence 6–9 × 5–8.5 cm, an open/ lax paniculate cyme, peduncle 0.3–2 cm long, with 5–19 flowers; each inflorescence with 2 or 3 pairs of major branches, usually reddish and densely glandular tomentose, pseudopedicel 0.4–1.5 cm long; bracts 1.7–2.5 mm long, subulate, caducous, bracteoles 0.5–1.2 mm long, subulate, caducous. Flowers 5- or 6-merous; pedicel c. 1 mm long. Hypanthium 2–2.5 mm long, campanulate, free portion c. 1 mm long, usually reddish to purple, external surface tuberculate and densely glandular tomentose with scattered stellate trichomes, mostly toward calyx, internal surface glabrous. Calyx tube 0.3–0.5 mm long; calyx lobes 0.5–0.7 × 1.6–1.7 mm, rounded, internal surface glabrous; calyx teeth 0.2–0.3 mm long, inconspicuous, ferruginous tomentose. Petals 6.4–8 × 4–4.5 mm, broadly obovate, pink, glabrous, apex obtuse to rounded with a lateral apical notch, base decurrent. Stamens 10, deflexed in 2 groups on both sides of flower at anthesis, 3 or 4 toward style and 6 or 7 on opposite side, deflexed backward and turning pink to red with age, falling together with petals. Filaments 4–5.5 mm long. Anthers 2–3.5 × 0.7–1 mm, ovate to shortly lanceolate, flattened toward apex; connective thinning out toward apex, as long as thecae and not projecting below; thecae oblong in lateral view, slightly wrinkled, truncate at base, with a ventral-apical pore, white. Ovary 3-locular, free portion rounded, glabrous; placentation reduced axile. Style 1–1.1 cm long, glabrous. Berries c. 6 × 6.5 mm, subglobose, 35–37-seeded. Seeds 2.3–2.5 mm long.

Phenology — Flowering specimens have been collected in March, May, June and October, fruiting specimens in June and August.

Distribution and ecology — Miconia obtusa is endemic to E Cuba ( Fig. 16), where it is found in the Sierra de Moa (HolguÍn and Guantánamo), Sierra de Baracoa (Guantánamo) and the coastal area between Moa and Baracoa (HolguÍn and Guantánamo). It grows in semi-arid montane shrub woods, semi-arid montane rainforest and pine forest on serpentine ( Borhidi 1996) at 10– 800 m.

Informal conservation status — Miconia obtusa (as M. cerasiflora var. setulifera ) was preliminarily assessed in the Red List of Cuban flora as Least Concern (LC) (see González-Torres & al. 2016). We estimated the EOO of M. obtusa to be 306 km 2 (within the limit for Endangered status under sub-criterion B1) and its AOO to be 9 km 2 (within the limit for Endangered under sub-criterion B2). It is known from nine localities, seven of them includ- ed in the National Park Alejandro de Humboldt and the Protected Area with sustainable use of natural resources Cuchillas del Toa. These localities represent three locations (sensu IUCN 2012). The number of locations allows us to assess this species within the Endangered category using criterion B. The extent and quality of habitat and number of individuals is declining at Yamaniguey because of illegal logging of associated timber species and construction and maintenance of high-voltage towers. The population is not considered severely fragmented, but the species is rare. Therefore, we assess M. obtusa as Endangered (EN) under criteria B1ab(iii,v)+2ab(iii,v).

Discussion — Miconia obtusa is one of the members of M. sect. Liogieria with a distinctive indumentum of elongate glandular trichomes. See the identification key and discussion under M. lutgardae for comparisons with that species. It can also be distinguished from M. thomasiana , the only member of the deciduous species endemic to Puerto Rico, by the differences in leaf margin, venation and the distribution of elongate glandular trichomes on the plant (see the key).

Our concept of Miconia obtusa corresponds with what until now has been treated as M. cerasiflora var. setulifera (see taxonomic history above). Given the morphological variability across all other species in this clade, and between M. obtusa and what had been traditionally recognized as M. cerasiflora var. cerasiflora , we feel that these two entities are distinct and should both be recognized at specific rank.

In the protologue of Graffenrieda obtusa, Grisebach (1866) did not specify a herbarium for the type gathering. As we clarified above, the specimen of G. obtusa (Wright 2527) seen by Grisebach was deposited in GOET (7029), which is designated here as the lectotype. Specimens from Wright 2527 have five different labels (sensu Howard 1988). The lectotype in GOET in the only one with label type 9. The GOET specimen with the number “709” (handwritten by A. Gray) and “2527” (handwritten by Grisebach) has label type 5, while that in US (120414) has label type 6 and those in BM (1008010), G-DC (00316994), GH (72277), K (535933), MO (313842), NY (99581), S (9-39614 [a] and 9-12537 [b] on the same sheet), UC (936605) and YU (65038) all have label type 6. The specimens in NY (99580) and S (5-3633) have label type 8 and, finally, those in BR (5212330) and G-DC (00316986 [1] and 00316987 [2] on the same sheet) do not have Wright labels. None of the additional specimens has the same label type as the lectotype and they should therefore be considered as dubious isolectotypes. Moreover, the sheet deposited in GH (72277) contains a mixture of gatherings. On this sheet, inside the envelope, are two Wright field notes with different data on locations and dates: (1) “Edges of savannas near Sagna de Taramo”, “3 Apr”; (2) “Pinal near Baracoa, 15 Jun 1861 ”. The data on the second field note match those on the label of the lectotype in GOET, but it is impossible to discern which of the three specimens mounted on the GH sheet is the one that corresponds to that field note.

The protologue of Miconia confusa ( Cogniaux 1891) did not specify a herbarium for the type gathering. As we clarified above, only the specimen of Wright 2527 in BR (5190799) has a handwritten determination by Cogniaux, as M. confusa , and therefore it is selected as the lectotype. It should be noted that because under M. confusa Cogniaux (1891) cited Wright 2527 only in part, and likewise the synonyms Graffenrieda obtusa and M. obtusa in part, M. confusa was not nomenclaturally superfluous when published.