Short answer
Ctenanthe marantifolia (Vell.) J.M.A.Braga & H.Gomes is the accepted name for the plant long sold as Ctenanthe pilosa Eichler, now a synonym [6]. Vellozo named it Thalia marantifolia in 1829 [2], and that older epithet was restored in 2007 [1]. It is a stem-forming Marantaceae herb of Brazil's Atlantic Forest, from Espírito Santo to Paraná, with hairy leaf sheaths and plain green leaves [7].
Why is Ctenanthe pilosa now called Ctenanthe marantifolia?
Because Vellozo got there first. José Mariano da Conceição Vellozo described the plant as Thalia marantifolia in his Flora Fluminensis, the flora of the Rio de Janeiro region; IPNI dates the text page to 1829, noting that the volume is dated 1825 but was published between 7 September and 28 November 1829 [2]. The epithet means maranta-leaved. Flora Fluminensis appeared decades after Vellozo wrote it and his names were often overlooked, so later botanists described the same plant afresh.
The later names pile up in the World Checklist of Vascular Plants (WCVP), the Kew dataset behind Plants of the World Online: Renealmia marantifolia (Vell.) D.Dietr. in 1839, a transfer into a ginger genus; Maranta pilosa Schauer in 1843; Thalia steudneri K.Koch ex Körn. in 1862; and two names Eichler published in 1884 when he set up Ctenanthe, C. pilosa and C. steudneri [6]. IPNI records Ctenanthe pilosa Eichler at page 83 of the Berlin Academy's Abhandlungen for 1883, issued 4 February 1884 [3]. For more than a century C. pilosa was the name in use, and it survives in the trade.
Under the International Code of Nomenclature, the earliest validly published epithet for a species has priority once the species is placed in a given genus. João Marcelo Alvarenga Braga and H. Gomes applied that rule in December 2007, publishing Ctenanthe marantifolia (Vell.) J.M.A.Braga & H.Gomes in Kew Bulletin 62(4): 647, with Vellozo's Thalia marantifolia as basionym, the name on which the new combination rests [1]. WCVP accepts it and lists Ctenanthe pilosa Eichler as a synonym [6], and GBIF follows the same treatment [9]. The same Vellozo priority, and the same Braga, also explain why Stromanthe thalia replaced S. sanguinea.
Where does Ctenanthe marantifolia grow?
WCVP gives the native range as southeastern Brazil to Paraná [4], recorded in the Brazil Southeast and Brazil South regions, and lists the species as introduced in Colombia and in Trinidad and Tobago [5]. Flora e Funga do Brasil confirms C. marantifolia in Espírito Santo, Minas Gerais, Rio de Janeiro and São Paulo in the Southeast and Paraná in the South, endemic to Brazil and restricted to the Atlantic Rainforest [7]. That is the coastal forest belt from Espírito Santo to the Paraná coast, the same heartland as Vellozo's Rio de Janeiro flora.
The habitat record is broader than for most of its relatives. The Brazilian treatment records a terrestrial herb in two vegetation types: Seasonally Semideciduous Forest and Ombrophyllous Forest (Tropical Rain Forest) [7]. Ombrophilous forest is evergreen and rain-fed; semideciduous forest has a dry season in which part of the canopy loses its leaves, letting in more light and letting the topsoil dry. A Ctenanthe spanning both is adapted to fluctuating light and soil moisture rather than to a permanently wet forest floor.
How is Ctenanthe marantifolia identified?
Habit comes first. Flora e Funga do Brasil describes C. marantifolia as caulescent: from an underground rhizome it raises aerial stems that carry the leaves, rather than holding them in a basal rosette [7]. Luna and colleagues found that this stem-forming habit evolved independently several times in the group that includes Ctenanthe, Saranthe and Stromanthe, one of them in C. marantifolia itself [10]. The leaf sheath is densely pilose (hairy), which is the character the old name pilosa recorded, and there is no ligule [7].
The leaves are plainer than in most cultivated Ctenanthe. The blade is oblong, green and concolor, meaning both faces are the same colour, with a green underside [7]. That sets it apart from the banded, wine-backed leaves of C. setosa and C. oppenheimiana. Its closest look-alike is C. lubbersiana, also caulescent: Flora e Funga do Brasil separates the two by their leaves, variegated in C. lubbersiana and concolor in C. marantifolia [8]. One tension in the literature is worth recording. Luna and colleagues place C. marantifolia in a subclade whose leaves are generally greyish above and green or purple beneath [10]; the word generally covers it, but the national flora's description of a uniformly green blade is the more specific field mark.
The flowers settle any doubt. The synflorescence, the whole flowering system, is sessile, one-noded and compressed, with one to four florescences per node subtended by a frondose (leaf-like) bract [7]. Flowers are lilac, with spathulate outer staminodes, sterile stamens that act as petals, and the callose staminode carries its thickened callus at the base; the ovary is pilose [7]. The fruit is a papery ellipsoid capsule, green when mature, and each seed bears a yellow lamellate aril, a fleshy layered appendage [7].
What happened to the trade name 'Ctenanthe pilosa'?
Nursery labels change more slowly than checklists. The RHS still files the variegated trade plant as Ctenanthe pilosa 'Golden Mosaic' (v), with Calathea 'Golden Mosaic' and Ctenanthe lubbersiana 'Variegata' as synonyms and its name status marked Unresolved [11]. Read against the botany, that label is doubtful: C. pilosa is a synonym of C. marantifolia [9], whose wild leaves are plain green on both faces [7], while the yellow-mottled leaf of Golden Mosaic matches the wild description of C. lubbersiana [8]. A variegated mutation of plain-leaved C. marantifolia is possible in principle, and no originator, patent or registration record that would settle the parentage of 'Golden Mosaic' was located in USPTO full text, IPNI or the RHS database as of 2026-10-08.
A plant sold as Ctenanthe pilosa with plain green leaves on hairy-sheathed stems is most likely C. marantifolia under its old name. A variegated plant under the same label is better checked against the C. lubbersiana page before the label is trusted.
What does its habitat suggest for growing it?
No horticultural study or growing record specific to Ctenanthe marantifolia was located for this page, so cultivation points here are inferred from its ecology and stated as such. A caulescent herb of both evergreen and semideciduous Atlantic Forest is built for filtered light that brightens seasonally and for soil that dries between rains without staying dry for long. In a pot that translates to a free-draining, humus-rich mix allowed to dry slightly at the surface, and bright light without direct midday sun.
The stems matter for handling. Because the leaves of C. marantifolia are carried up on aerial shoots rather than from a basal crown, an old plant becomes tall and open; dividing the rhizome at repotting gives new clumps, each with its own stem bases. The dense hairs on the leaf sheaths are a stable feature of the species and a useful check on identity, not a sign of pests.
Frequently asked questions
Is Ctenanthe pilosa a valid name?
Not as an accepted name. WCVP lists Ctenanthe pilosa Eichler (1884) as a synonym of Ctenanthe marantifolia [6], because Vellozo had already named the species Thalia marantifolia in 1829 [2]. Braga and Gomes made the new combination in 2007 [1].
Where is Ctenanthe marantifolia native?
Southeastern Brazil to Paraná [4]. Flora e Funga do Brasil records it in Espírito Santo, Minas Gerais, Rio de Janeiro, São Paulo and Paraná, in both evergreen and semideciduous Atlantic Forest [7].
How do you tell Ctenanthe marantifolia from Ctenanthe lubbersiana?
Both form aerial stems, but the leaves differ: concolor and plain green in C. marantifolia, variegated in C. lubbersiana [8]. C. marantifolia also has a densely hairy leaf sheath and no ligule [7].
Is 'Golden Mosaic' a Ctenanthe marantifolia?
The RHS files it as Ctenanthe pilosa 'Golden Mosaic' with status Unresolved [11]. Since C. pilosa equals C. marantifolia [9] and that species has plain green leaves [7], the attribution is doubtful; its mottled leaf fits C. lubbersiana better [8].