Cryptanthus univittatus Leme, 2020

Leme, Elton M. C., Ribeiro, Otávio B. C., Souza, Fernanda Vidigal D., Souza, Everton Hilo De, Kollmann, Ludovic J. C. & Fontana, André P., 2020, Miscellaneous new species in the “ Cryptanthoid complex ” (Bromeliaceae: Bromelioideae) from Eastern Brazil, Phytotaxa 430 (3), pp. 157-202 : 165-168

1.4. Cryptanthus univittatus Leme , sp. nov. ( Fig. 4 A–E )

Diagnosis:––This new species is morphologically close to C. bivittatus , differing from it by the narrower leaf blades (1–1.5 cm × 3–4 cm wide), bearing adaxially a single, broad reddish-rose median stripe (vs. bearing 2 narrow whitish or rose stripes, in contrast with a narrow central and the marginal darker green stripes), marginal spines larger (1–2.5 mm vs. ca. 0.5 mm long), sepals longer (12–13 mm vs. 8–10 mm long) and higher connate (ca. 8 mm vs. 3.5–5 mm).

Type:–– BRAZIL. Minas Gerais: São José do Divino, Serra dos Aimorés, at the base of Pedra Riscada, 427 m elevation, 18º22’19” S, 41º19’45” W, May 2012, R. Vasconcelos s.n., fl. cult. E. Leme 8654 (holotype RB!).

Description:–– Plants stemless, propagating by short basal shoots. Leaves ca. 8 in number, suberect to spreading at anthesis, laxly arranged and forming an open rosette; sheath inconspicuous, reddish, white lepidote abaxially near distal end, glabrous adaxially; blade narrowly sublinear-lanceolate, attenuate, apex caudate, inconspicuously if at all narrowed toward the base, 13–20 × 1–1.5 cm, coriaceous, slightly canaliculate, bearing a single, broad reddish-rose median stripe in contrast with the narrow, dark reddish-green marginal stripes, non-lustrous, abaxially densely and coarsely white lepidote, adaxially glabrous, margins slightly if at all undulate, densely spinose; spines triangular, retrorsely (basal ones) to straight and antrorsely uncinate (median to apical ones), dark red, 1–2.5 mm long, 2–3 mm apart. Inflorescence sessile, shortly corymbose, ca. 2.5 cm long, ca. 2.5 cm in diameter, once branched at the base and bearing a distinct simple central head of densely arranged flowers; primary bracts resembling the leaves; flower fascicles ca. 4 in number, inconspicuous, with ca. 2 flowers, ca. 22 × 10 mm, complanate; floral bracts narrowly triangular, acuminate, ca. 15 × 7 mm, membranaceous, greenish, inconspicuously castaneous lepidote, spinulose near the apex, distinctly exceeded by the sepals, those in the flower fascicles distinctly carinate. Flowers sessile, fragrance not detected, the perfect ones ca. 36 mm long (with the petals extended), the staminate ones shorter; sepals 12–13 mm long, connate for ca. 8 mm, castaneous toward the apex whitish-rose near the base, sparsely pale castaneous lepidote, lobes broadly ovate, acute and apiculate, 4–5 × 3–3.5 mm, symmetrical, obtusely carinate, margins densely and inconspicuously crenulate-spinulose; petals subspathulate, apex rounded, slightly cucullate, white except for the green apex and apical margins, exceeding the stamens but spreading-recurved at anthesis and exposing them; those of the perfect flowers ca. 25 × 5 mm, connate at the base for ca. 6 mm, without callosities; filaments ca. 18 mm long, adnate to the petal-tube and free above it; anthers ca. 3 mm long, dorsifixed near the middle, base distinctly bilobed, apex obtuse; stigma conduplicate-patent, lobes ca. 4 × 2 mm, white, margins scalloped, without papillae; ovary narrowly subclavate, trigonous, ca. 10 × 5 mm, whitish, glabrous; epigynous tube lacking; placentation apical; ovules few, obtuse. Fruits not seen.

Distribution, habitat and conservation:–– Cryptanthus univittatus was found growing as a terrestrial under shrubby vegetation of the small fragments of dry Atlantic Forest at the base of Pedra Riscada ( Fig. 4 A ), which is part of Serra dos Aimorés, in the County of São José do Divino, in Minas Gerais state, not far from the border with Espírito Santo state, at 427 m elevation. Plants were observed forming sparse groups of individuals ( Fig. 4 , B–C) in clearly anthropized condition, which makes surviving of the only known population uncertain. So, it must be considered a critically endangered species (CR) according to the “A.1. c + d” and “B. 1. a + b. i to iv” criteria adopted by IUCN (2012).

Etymology:––The name of C. univittatus is based on the combination of the Latin word uni, which is the prefixal form of unus, meaning ‘one’, ‘single’, combined with the Latin word vittatus, meaning ‘striped’, as a reference to its leaf blades bearing a single, broad reddish-rose median stripe contrasting with the dark reddish-green marginal stripes.

Observations:— Cryptanthus univittatus is morphologically close to the mysterious C. bivittatus ( Hooker 1861: 5270) Regel (1865: 2) , a taxonomically poorly known species. However, it differs from it mainly by the narrower leaf blades (1–1.5 cm × 3–4 cm wide), which are sublinear-lanceolate (vs. lanceolate), white-lepidote abaxially (vs. castaneous lepidote), bearing adaxially a single broad reddish-rose median stripe (vs. bearing 2 narrow whitish or rose stripes, in contrast with a narrow central and the marginal darker green stripes), marginal spines larger (1–2.5 mm vs. ca. 0.5 mm long), and sepals longer (12–13 mm vs. 8–10 mm long) and higher connate (ca. 8 mm vs. 3.5–5 mm).

According to the protologue of C. bivittatus ( Hooker 1861) , this species arrived in Hooker’s hands from Linden, in 1859, under the name Billbergia bivittata (basionym of C. bivittatus ), without exact place of origin, but “(…) no doubt, a native of South America (…)”. Baker (1889) described C. bivittatus var. luddemannii (1889: 16) based on a Morren painting. The portrayed specimen came from Porto Seguro, Bahia, also introduced by Linden, cultivated in Kew Gardens since 1859, and flowering there in 1861. The only difference of this variety is its robustness—“more robust than the type ”, according to Baker (1889). However, Mez (1934 -35), who provided the most complete description of the species ‒ which is the base of the description adopted by Smith & Downs (1979) ‒ implicitly considered C. bivittatus var. luddemannii a synonym of the typical variety (making reference to the Porto Seguro origin), being followed by Smith & Downs (1979). Finally, C. bivittatus var. atropurpureus was described by Mez (1934 -35: 18), based on a specimen smaller in all its parts (“ a typo evidenter differ omnibus partibus satis minoribus ”), from Edinburgh, with leaves about 5 cm long and 1.5 cm wide, constantly red, with longitudinal bands less pronounced, but no specimen was cited and the type is unkown.

The typical Cryptanthus bivittatus , Hooker’s so called “Ribbanded Billbergia ”, was described as having “(…) foliis (…) fusco-viridibus lineis duabus albo-vittatis (…)”, or in his own English words, leaves “(…) upper green, with two broad, buff, longitudinal bands, which pass into dull red at the base of the leaf.” This characteristic is striking and clearly visible in the protologue, in the color plate 5270, as well as explicitly indicated in the line drawings in the holotype deposited in Kew Herbarium, which is accompanied by a single leaf, two longitudinal sections of the inflorescence (the apical part of one of it is separated from the basal part), including the primary bracts, two separate flowers, part of one corolla, and immature flower buds, some of them with attached floral bracts. When transferring Billbergia bivittata to Cryptanthus, Regel (1865) portrayed in his color plate 458 a greener specimen, despite the two longitudinal bands are still distinctly visible.

As far as our knowledge is concerned, C. bivittatus was not yet collected in its natural habitat and all available material was originating from cultivated plants from Europe and United States, which include popular selected cultivars and hybrids of recognized ornamental value. So establishing a new species morphologically close related to it is a true challenge mainly because only vegetative characters are well described so far: floral details are very poorly characterized in the protologue as well as in Baker (1889), Mez (1934 -35) and Smith & Downs (1979) descriptions, which are the basis of the current knowledge of the species. However, Ramírez (1996), when reviewing the genus, provided an expanded description, here considered, although it was based only on one specimens of cultivated origin.

If only leaves are considered, C. univittatus resembles in color and shape the dubious C. pratextus Baker (1889: 16) , which was “(…) described from a drawing of Professor Morren’s ” ( Baker 1889), available in Kew Library (despite not listed), but there is no known preserved specimen. Morren’s color drawing portrays a short caulescent specimen (vs. stemless in C. univittatus ), with distinctly undulate (vs. slightly if at all undulate) leaf blades bearing a single median paler colored stripe in contrast with the color the marginal area. It was in very late post-floral stage being attested by the presence of a basal, erect offshoot well visible in the drawing (vs. short basal shoots in C. univittatus ). Flower or flower parts are completely missing, which explains why there are no flower details described in the protologue. This situation is confirmed by Mez (1934 -35). The absence of any data on flower parts makes Ramírez (1996) to consider C. pratextus a doubtful taxa, which is confirmed here, as well as the impossibility to fully compare it to any existing taxon.

On the other hand, C. univittatus can be also compared with C. marginatus L.B. Smith (1955: 24) . It is differing by the narrower leaf blades (1–1.5 cm × ca. 3 cm), with larger marginal spines (1–2.5 mm × ca. 0.5 mm long), and by the sepals with densely and inconspicuously crenulate-spinulose margins (vs. entire). It also differ from C. ilhanus Leme (Leme & Kollman 2013: 16) , by the longer leaf blades (13–20 cm vs. 8–12 cm long), bearing a single, broad, reddish-rose median stripe (vs. median stripe narrow, green to yellowish-green), floral bracts narrowly triangular (vs. sublinear-oblong), larger (ca. 15 × 7 mm vs. 10 × 1–2 mm), spinulose near the apex (vs. entire), sepal lobes broadly elliptic (vs. elliptic to obovate), larger (4–5 × 3–5.5 mm vs. 2.5–3 × 2 mm), castaneous (vs. greenish-white), petals without callosities (vs. with distinct callosities), and larger ovary (ca. 10 × 5 mm vs. 5 × 3–4 mm).

Figures

FIGURE 4. Cryptanthus univittatus from the vegetated base of Pedra Riscada, Serra dos Aimorés, in São José do Divino, Minas Gerais state (Leme 8654). A. General view of Pedra Riscada. B–C. Small group of plants of C. univittatus at the type locality. D. Detail of the inflorescence highlighting the hermaphrodite flower of the lateral flower fascicle. E. Habit of a flowering specimen in cultivation. B–C. Photo R. Vasconcelos.