Dichorisandra odorata Aona & M.C.E. Amaral, 2014

Aona-Pinheiro, Lidyanne Yuriko Saleme, Bittrich, Volker & Amaral, Maria Do Carmo E., 2014, Two new species of Dichorisandra (Commelinaceae) from Rio de Janeiro and comments on the two species included in Vellozo’s “ Flora Fluminensis ”, Phytotaxa 184 (4), pp. 223-234 : 224-227

Dichorisandra odorata Aona & M.C.E. Amaral , sp. nov. ( Figs. 1–3 )

Dichorisandra odorata is differentiated by its spirally arranged leaves, congested at the apex, densely pilose on both sides, inflorescence subsessile, with 2–3 cincinni and erect, fragrant flowers with six anthers each dehiscing from a single apical pore.

Type: — BRAZIL. Rio de Janeiro: Rio das Ostras, Reserva Biológica União , Rod. BR 101, km 185, Rocha Leão. 22°24’26”S, 42°01’34”W, 07 December 2001, fl., F. Feres 25, K. Matsumoto & R. Belinello (holotype: UEC!, isotype: US!)

Erect herbs, aerial stems branched at base or unbranched, 10–30 cm tall. Stem ca. 2–3 mm diameter at base, ca. 2mm diameter at apex, pilose or rarely glabrous, hairs ca. 2 mm long, brownish; internodes 2–7 cm long. Leaves spirally arranged, congested at the apex of the stem, sheaths 1.2–2 cm long, villous, hairs brownish, ca. 2 mm long, scarious, margin ciliate, cilia 1–2 mm long; leaves subsessile; leaf blades elliptic to ovate, 7.8–11 x 3.8–6 cm, symmetric at base, apex acuminate, margin ciliate at apex, pale green beneath, dark green above, densely pilose on both sides, hairs ca. 2 mm long, brownish. Thyrse terminal, erect, subsessile; basal bract leaf-like; cincinni 2–3, subsessile, strongly congested, each cincinnus 3–4 flowered; bracts of cincinni parallel to cincinni, scarious, triangular-acuminate, 7–10 x 3 mm, adaxial surface glabrous, abaxial surface pilose, margin ciliate, bracteoles scarious, 2–3 x 2–3 mm, pilose on both sides, margin ciliate. Flowers either bisexual or staminate, corolla actinomorphic, 1–1.5 cm diameter; pedicel 0.3–1 cm long in flower, to 2 cm long in fruit, pilose; floral buds ovoid, 6–7 x 3.2–4 mm; sepals elliptic to lanceolate, 8–9 x 3–4 mm, greenish, pilose, margin glabrous; petals elliptic to ovate,11–12 x 6–9 mm, blue-purplish, whitish at base; stamens 6, apically connivent, subequal, filaments 2–3 mm long, anthers whitish with apex purplish, 4–5 mm long, dehiscing by 1 apical pore, base cordate to subcordate, connective pilose; ovary ca. 2 x 2 mm, whitish, globose, pilose, smooth, style ca. 6–7 mm long, style with upper half bluish, apex erect, stigma truncate. Capsule cylindrical, 15–20 x 2–3 mm, 4–5 seeds per locule, epicarp pilose. Seeds 5–8 x 2–3 mm, elliptic, black; aril white, inconspicuous.

Distribution and habitat: —This species is only known from Rio das Ostras, in “restinga arbórea” (marine influenced vegetation on sandy soil), Silva Jardim and Santa Maria Madalena municipalities, in Atlantic Forest vegetation ( Fig. 3 ).

Phenology: —Flowering in December, March and April and collected with young fruits in December.

Etymology: —The epithet means “scented”, making reference to the scented flowers of this species, a character apparently unique for the genus.

Conservation status:—Endangered (EN B2 ab (ii, iii; IUCN 2001). Dichorisandra odorata has a very restricted distribution (area of occupation: 72,70 km 2) and a small number of collections in three municipalities. It should be remembered, however, that the collection density in the Atlantic forest is often below the desirable level. Thus the species might well occur in still other localities ( Figure 3 ).

Paratypes: — BRAZIL. Rio de Janeiro : Rio de Janeiro, Silva Jardim, poço d’Anta, floresta do morro à beira do Rio São João , 18 April 1977, fl. D. Araújo 1633 & R. F. Oliveira 376 ( GUA, RB); Santa Maria Madalena, águas paradas, fl., 5 March 1935, S. Lima & Brade 14183 ( B, RB) .

Discussion: — Dichorisandra odorata resembles D. acaulis Cogn. (1894:297) due to its small stature, subsessile cincinni, actinomorphic perianth and pilose ovary. However, D. odorata can be distinguished from the latter by its pilose leaves with brownish hairs on both sides (vs. leaves adaxially glabrous and abaxially setose with hyaline hairs), its anthers dehisce by one apical pore, (vs. dehiscing by two apical pores in D. acaulis ) and by the pilose connectives (vs. glabrous in D. acaulis ). The stems in D. odorata can be branched, giving the plant a shrubby appearance, while in D. acaulis the stems are always unbranched.

Habit and flower morphology of D. odorata are similar to D. nutabilis Aona & M.C.E. Amaral (2012: 21) , a species endemic to the state of Espírito Santo. D. odorata can be distinguished, however, by the densely pilose leaves (vs. sparsely pilose in D. nutabilis ), by the reduced erect inflorescence (vs. elongate and pendulous inflorescences in D. nutabilis ) and pilose connectives (vs. glabrous connectives in D. nutabilis ) ( Figure 1 ).

Dichorisandra acaulis and D. nutabilis seem to be restricted to the state of Espírito Santo, whereas D. odorata is apparently restricted to northwestern Rio de Janeiro.

Dichorisandra picta Lodd. (1830: 142) is a poorly known species, but apparently also similar to D. odorata . In the revision of the genus ( Aona, 2008), D. picta was excluded, considered a dubious name. Loddiges’s protologue (1830) includes a beautiful illustration of a D. picta (1830: 1667) plant showing some similarity in shape and color of leaves. It also illustrates, however, some differences between these species, notably in the stamens ( D. picta with 5 stamens, connective glabrous and D. odorata with 6 stamens, connective pilose) and sepal colour (sepals whitish in D. picta and sepals greenish in D. odorata ) showing that they should considered distinct species. The protologue does not include descriptions of flower parts. Hooker (1854: 1667), unaware of Loddiges’s work, published D. picta , a posterior homonym and thus illegitimate. That he was unaware of Loddiges´s publication is evident from his text: “... received from Mr. Low, of the Clapton Nursery, under the name we have here adopted, but which we can nowhere find published in books to which we have access, and the species does not appear to be described under any other name.” In this work, illustration and description seem to be from a species presumably very closely related to D. odorata , also sharing the presence of 6 stamens with a single apical pore. The species described by Hooker, however, is distinct by the presence of glabrous connective, blotched leaves with two broad brown longitudinal stripes on the adaxial side and by its long inflorescences and. As no herbarium specimens exist, the identity of D. picta Lodd. and D. picta Hook. (nom. illeg.) possibly never will be discovered. It should be remembered in this context that huge parts of the Atlantic forest, where many species of Dichorisandra are distributed, were destroyed during the last 150 years. It would thus be no surprise, if several species of the genus went extinct since the descriptions of Loddiges and Hooker were done.

Figures

FIGURE 1. Dichorisandra odorata. A. Habit. B. Detail of sheaths. C. Detail of leaf, adaxial surface. D. Detail of leaf, abaxial surface. E. Flower. F. Detail of stigma. G. Detail of a bisexual flower with part of the sepals and petals removed. H–J. Three stamens; H. adaxial view. I. Abaxial view; note hairs on connective. J. Lateral view. K. Bisexual flower. All from Feres 25 (UEC). (originally from Rio das Ostras; cultivated at the UNICAMP).

FIGURE 2. Dichorisandra odorata. Flower in front view.

FIGURE 3. Map of distribution of. D. nana and D. odorata in Rio de Janeiro state. Note fragments of the Atlantic rain forest.