Voyria alvesiana E.F.Guim., T.S.Mendes & N.G.Silva, 1818

Gomes, Sofia I. F., Kikuchi, Izai A. B. S., Lachenaud, Olivier, Perdomo, Julian, Léotard, Guillaume, Maas, Paul J. M., Kamer, Hiltje Maas-van de & Merckx, Vincent S. F. T., 2022, Unravelling the species diversity, phylogeny and biogeography of the mycoheterotrophic Voyrieae (Gentianaceae) and the description of a new species, TAXON 71 (5), pp. 1013-1024 : 1018-1019

Voyria alvesiana E.F.Guim., T.S.Mendes & N.G.Silva in Rodriguésia 69(3): 1131. 2018.

During our investigations we came across several specimens from French Guiana and Surinam identified as Voyria tenella Hook. , which after closer investigation belong to V. alvesiana . This species was described by Guimarães & al. (2018) from a limited number of specimens from the Brazilian state of Pará. Based on our observations, we provide a description of this species with updated character dimensions.

Description. – Mycoheterotrophic herb, 5–20 cm high; roots radiating from the base of the stem, to 7 mm long and 0.3 mm diam., simple, numerous, white, sparsely to rather densely covered with erect hairs, proximally fleshy and narrowly ovoid, apically tapering and ending in a long, filiform and upwardly recurved apex; stems simple, white to brownish, internodes to 10 mm long. Leaves white, connate in their lower half, narrowly triangular, 4–5 × 1–1.5 mm, apex acuminate. Flowers solitary, (4–)5-merous, nodding in bud; pedicels 55–60 mm long; calyx whitish or very pale yellow, tube 1– 2 mm long, lobes triangular, c. 1 × 0.5 mm, apex acute, sinuses obtuse to acute, calyx scales present; corolla salverform, 12–14 mm long, tube whitish to yellowish, 7–10 mm long, lobes white, generally lemon yellow to orange yellow at the base, narrowly triangular, 3–7 × 1–2 mm, apex acute, overlapping each other at the base on one side, inner side densely papillate; stamens inserted c. 2 mm below the throat, anthers subsessile, dorsifixed, 1.2–1.3 mm long; corolla lobes and upper part of corolla tube together falling off after anthesis; ovary narrowly obovoid, c. 5 × 3 mm, provided with two distinctly stipitate, club-shaped glands about as long as the ovary, stipe c. 5 mm long, head of glands ellipsoid to subglobose and flattened against the ovary; style shorter than the ovary, c. 3 mm long, stigma capitate. Capsule narrowly obovoid, 5– 10 × 2.5 mm, medially longitudinally septicidal, margins of the septa thickened and rolled inwards, creating a lantern-like structure. Seeds filiform, c. 0.2 mm long.

Distribution and ecology. – Brazil (Pará), French Guiana, and Surinam. In understory of primary forests, on wet places along creeks, at higher elevations in more open cloud forests among rocks. In French Guiana at 50–500 m elevation. In Surinam at 950 m elevation.

Notes. – Although looking alike, Voyria alvesiana differs from V. tenella by several independent characters:

(1) Length of stipitate glands: in Voyria alvesiana , the head of the glands almost reaches the top of the ovary. In V. tenella , the head of the glands reaches about halfway up the ovary. Gland characteristics are probably the best diagnostic feature distinguishing both species. Reliable identification of herbarium specimens should primarily be based upon this character, despite sometimes requiring a small dissection and thus being a little bit destructive.

(2) Shape and color of corolla throat: in close frontal view, the corolla throats of both species are obviously distinct, i.e., open and 5-merous star-shaped in Voyria alvesiana , narrower and perfectly circular in V. tenella . This difference is mainly due to the form of the base of corolla lobes: in V. alvesiana , they are concave on the upper side, overlap dextrorsely, and usually bear a distinct yellow patch (honey mark) at the base; in V. tenella , the base of the corolla lobes is flat to convex, not overlapping, and always devoid of honey mark (but the floral tube is yellow orange!). Additionally, the abscising upper part of the corolla tube is slightly widened and more or less bulging in V. alvesiana , but not in V. tenella .

(3) Shape of corolla lobes: in Voyria alvesiana , corolla lobes are narrowly triangular with an acute tip and a tendency to be asymmetrical-falcate, whereas in V. tenella , corolla lobes are narrowly ovate to narrowly triangular with a symmetrical rounded tip. Easy to analyse in the field, this character can be much more difficult to see in herbarium specimens, especially in poorly preserved ones.

(4) Shape of corolla tube: tube with bulging part below the corolla lobes in Voyria alvesiana (corolla tube without bulging part in V. tenella ).

(5) Root system: the roots are more numerous, narrowly ovoid and ending in a long, recurved, filiform apex in Voyria alvesiana , whereas in V. tenella , roots are fewer and swollen over their whole length, thus ending in a shorter and more obtuse apex.

(6) Corolla color: as far as known, the corolla is always white in Voyria alvesiana . On the other hand, in V. tenella , the normal color of the corolla is bright blue, but hypochromic individuals, although always rare, are occasionally found. In the large population of Mont Itoupé, we found on a daily basis individuals of V. tenella with a pink or white corolla. In the field, corolla color is the easier way to detect V. alvesiana , but to ascertain the identity, one should always rely on the shape of the corolla lobes, throat and tube. The corolla can rarely be pale lilac (Feuillet 9928).

(7) Phenology: in a syntopious situation, the optimum of anthesis was at least one month earlier in Voyria tenella than in V. alvesiana . In 2010, at Mont Itoupé, populations of both species were observed growing side by side during more than five weeks. Voyria tenella had already reached its optimum by the end of February, and one month later, it was even difficult to find one in flower. On the contrary, V. alvesiana , which was only occasionally seen at the beginning of March, was abundantly flowering by the end of this month. Indeed it is precisely this phenological difference between blue and white flowers that prompted more thorough investigations, and led to the preliminary conclusion that two species were involved. However, the overlapping of anthesis was important at this site, and even if we expect that the phenological difference holds at other sites, it should be noted that on a larger scale, flowering of both species is known to last through most of the rainy season in French Guiana. One label mentions that flowers are slightly perfumed (De Granville 846), a fact which is supported by the presence of a densely papillate epiderm on the inner side of the corolla lobes.

Specimens studied. – FRENCH GUIANA. Régina Region: Eastern plateau of Montagne Tortue , 11 km WNW of Approuague river, alt. 200–450 m, 12 Jun 1988, 04°18′N 52° 22′W, primary forest with many open places, Feuillet & al. 9928 ( CAY, U). Régina, Eastern plateau of Montagne Tortue , 11 km WNW of Approuague river, near left bank of Crique “Feuillet ”, alt. 400 m, 13 Jun 1988, 04°18′N 52°22′W, light forest, Feuillet & al. 10142 ( CAY, U). Monts Atachi Bacca, between top of table-mountain and top of 525 m, alt. 400 m, 8 Mar 1971, forest on ridge, De Granville 823 ( CAY). 500 m of Maroni River, E of island Assadam Tabiki , on track leading to Monts Atachi Bacca , alt. 500 m, 12 Mar 1971, forest on slope, De Granville 846 ( CAY, U). Western slope of table-mountain top , about 45 km SE of Saül, alt. 500 m, 29 Aug 1980, near stream between rocks, De Granville 3669 ( CAY). NW part of Montagne Bellevue de l’ Inini , alt. 550 m, 14 Aug 1985, forest on slope, De Granville & al. 7487 ( CAY). Basin of Sinnamary River , Drop Zone of Crique Jupiter , alt. 100 m, 25 Apr 1991, 04°04′N 53°09′W, on well-drained soil in understory, De Granville & al. 11512 ( CAY, spirit collection). Cacao, sentier Molokoï , alt. 53 m, 17 May 2014, wet forest on terra firme, Lachenaud 1818 ( CAY, BR, U, spirit collection, slides). Saül, Sentier Botanique , near crique about 2 hours from Les Eaux Claires , alt. 200 m, 11 Feb 1993, on dark colored sand on edge of stream in high forest, Maas & al. 8073 ( U). Route National 2 , km 72 , alt. 35 m, 27 Aug 2005, 04°27′N 52°20′W, wet open place along creek, holotype, Maas & al. 9680 (holo U, iso CAY). Castle , on road to Cacao , alt. 190 m alt., 20 Jul 2005, Moonen 343 ( U). Mt. Cacao, about 60 km S of Cayenne, alt. 250 m, 25 Feb 1965, very dark understory of forest, Oldeman 1161 ( CAY). Mt. Cacao , about 60 km S of Cayenne, alt. 50–400 m, 11 Jun 1965, high forest, Oldeman 1359 ( CAY). Mont Itoupé, near the base camp , alt. 600 m, 29 Mar 2010, 03°01′N 53°05′W, understory of primary forest on well-drained slopes, Tostain & Léotard 5828 ( CAY, spirit collection). SURINAM. Emmaketen, plateau just below Kleine Hendrik-top , alt. 950 m, 30 Jul 1959, cloud forest, Daniëls & Jonker 798 ( U). Ulemari River, 99 km upstream of confluence with Litani River, base camp 2 , alt. 170 m, 9 Apr 1998, 02°58′N 54°35′W, understory of primary forest on well-drained slopes, Raghoenandan & Pinas UVS17823 ( U) .