Selaginella rhodostachya Baker

Selaginella rhodostachya Baker, Timehri 5: 221. 1887. Type: Guyana. [Venezuela]: Mt. Roraima, [Oct. 1884 to Jan. 1885], [E. F.] im Thurn s. n. (holotype: K! [K 000229471]; isotypes: BM! [as im Thurn 226, an erroneous number; most likely im Thurn s. n., BM 014638426], US! [as im Thurn 271, an erroneous number; most likely im Thurn s. n., US 00135738]. Fig. 3.

= Selaginella brevispicata Hieron. ex Bautista, Acta Amazonica 4: 19. 1974. Type: Brazil. Roraima: [Mt.] Roraima, alt. 2300 m, Dec. 1909, E. Ule 8491 (holotype: MG - digital image! [MG 013555]; isotypes: B! - 2 sheets [B 20 0095099 & B 20 0095100], BM! [BM 000905714], G! [G 00349430], K! [K 000589268], L! [L 0057418]).

= Selaginella valdepilosa Baker subsp. tricholoma Jermy & Rankin, Bull. Brit. Mus. (Nat. Hist.), Bot. 9: 292. 1981. Type: Venezuela: Bolívar, summit of Mt. Roraima, [on southern half of the summit between Summit Camp, Central Rift, Central Swamp, and pond at southern end], alt. 2700–2740 m, 28 Sep. 1944, J. A. Steyermark 58916 (holotype: BM! [as 58016] [BM 000905664]; isotypes: F! [Herb. No. 1208244], G!, GH p. p.!, L!, MICH! [MICH 1191442], MO! [MO -255519], NY! [NY 00020658], US - 2 sheets! [Herb. Nos. 1915037 & 1915105]). Fig. 4.

Description.

Plants terrestrial or epipetric. Stems creeping, stramineous, 10–15 cm long, 0.2–0.5 mm diam., non-articulate, not flagelliform or stoloniferous, 1–3 - branched, determinate. Rhizophores dorsal, ventro-axillary, or axillary, borne throughout the stem length, filiform, 0.1 or 0.2 mm diam. Leaves chartaceous to coriaceous, heteromorphic, of three types (ventral, median, and axillary). Lateral leaves distant, ascending and spreading at 45 ° to the stems, elliptic to elliptic-ovate, 0.7–2.0 × 0.5–1.2 mm; bases rounded, glabrous, the acroscopic bases strongly overlapping stems, the basiscopic bases free from the stems; acroscopic and basiscopic margins on the upper and lower surfaces hyaline or greenish-hyaline, continuously bordered by a band 1–3 cells wide of idioblasts, the idioblasts elongate, straight-walled and papillate, the papillae in a single row on each cell lumen, entire throughout or entire along proximal 3 / 4 and scarcely denticulate along distal 1 / 4, or long-ciliate throughout; apices obtuse, tipped by 6–8, short, tooth-like hairs; upper surfaces glabrous, comprised of rounded, sinuate-walled, and papillate cells, the papillae 2–6 on each cell lumen, without stomata, the lower surfaces comprising elongate, sinuate-walled, and laevigate cells and with some, sparsely distributed, elongate, sinuate-walled, and papillate idioblasts, the papillae 10–20 arranged in a single row on each cell lumen, with stomata along the distal 1 / 2 of the midribs in 1–3 rows along the midribs, without submarginal stomata or tooth-like hairs, the midribs straw-colored and prominently raised on side view. Median leaves distant or imbricate towards stem and branch apices, ascending and spreading at 70 ° to the stems, elliptic, elliptic-ovate, or broadly ovate, 0.9–1.4 × 0.5–1.0 mm; bases glabrous, rounded, without auricles; margins on the upper and lower surfaces hyaline, continuously bordered by a narrow band of idioblasts, the band 1 or 2 cells wide, the idioblasts as in the lateral leaves, entire throughout or entire along the proximal 3 / 4 and scarcely denticulate along the distal 1 / 4, or long-ciliate throughout; apices acute, broadly acute to obtuse, each tipped by 1–5 tooth-like, projections, often caducous, or by 2 divergent, long cilia; upper surfaces glabrous, comprise of rounded, sinuate-walled, and papillate cells, the papillae 2–6 on each cell lumen, without stomata, the lower surfaces comprising elongate, sinuate-walled, and laevigate cells and with some, sparsely distributed, elongate, sinuate-walled, and papillate idioblasts, the papillae 10–20 arranged in a single row on each cell lumen, the midribs green and prominently raised on side view throughout the leaf length or along distal 1 / 2 where the stomata are present along the distal 1 / 2 of the midribs in 1 or 3 rows, without submarginal stomata or tooth-like hairs. Axillary leaves elliptic or narrowly ovate to ovate-lanceolate, 1.0–1.4 × 0.5–0.8 mm; bases rounded; margins, apices, and leaf surfaces as in the lateral leaves. Strobili terminal on the branch tips, loosely quadrangular, 2–5 mm long. Sporophylls monomorphic, ascending, imbricate, keeled along the midribs, the keels glabrous on the upper surfaces, without laminar flaps, ovate to ovate-elliptic, broadly elliptic or broadly ovate, 1.0–1.4 × 0.6–0.8 mm; bases rounded; apices acute or obtuse, tipped by 2–5 short, short-teeth-like projections or 2 divergent, long-cilia; stomata found along the midribs; dorsal sporophylls with margins hyaline, bordered by 1 or 2 rows of elongate, straight-walled and papillate idioblasts similar to those of the median leaf margins, without marginal stomata, entire throughout or entire on proximal 2 / 3, otherwise sparsely denticulate distally; upper surfaces greenish, comprised of rounded to quadrangular cells, except for the leaf half that imbricate with the ventral sporophylls, where the cells are elongate and straight-walled, the lower surfaces comprised of elongate and straight-walled cells; ventral sporophylls with margins hyaline similarly bordered as in the dorsal sporophylls, entire on proximal 3 / 4 and sparsely denticulate along the distal 1 / 4, without marginal stomata; the upper and lower surfaces hyaline, composed of elongate, straight-walled and laevigate cells. Megasporangia in two ventral rows; megaspores light yellow, yellow, or cream, without an equatorial flange, proximal and distal faces psilate to slightly rugulate, the microstructure of proximal and distal faces not observed, 200–300 µm diam. Microsporangia in two dorsal rows; microspores orange, proximal and distal faces psilate, micro-ornamentation not observed, 20–40 µm diam.

Habitat and distribution.

Selaginella rhodostachya is found on muddy soils, in rock crevices, and at the bases of wet bluffs within shrubland and forest at 1,350 –2,800 m. It is a typical species of the Guayana Highlands, recorded from the Mazaruni-Potaro region of Guyana, the Amazonas and Bolívar states of Venezuela, and Roraima in Brazil.

Conservation status.

Selaginella rhodostachya is known from multiple localities in the Guayana Highlands, with confirmed collections from Venezuela, Guyana, and Brazil. The available georeferenced records indicate an extent of occurrence (EOO) of 21,005.349 km 2 and an area of occupancy (AOO) of 28 km 2. According to the IUCN Standards and Petitions Committee (2022) guidelines, we recommend a preliminary conservation status of Vulnerable (VU B 2 ab (iii)), due to its limited AOO and occurrence at a few locations. Additionally, ongoing expansion of extractive activities, particularly gold mining, along with logging, agriculture, and infrastructure development in the Guiana Shield, suggests a continuing decline in habitat quality in parts of the Guayana Highlands (Bovolo et al. 2018).

Specimens examined.

Guyana • [Cuyuni-Mazaruni]: slopes near Roraima, G. S. Jenman s. n. (NY). Venezuela • Amazonas: Río Negro, Cerro Aracamuni summit, Proa Camp, 01°32'N, 65°49'W, alt. 1400 m, 30 Oct. 1987, R. L. Liesner & G. Carnevali 22659 (U-digital image, VEN-digital image), • Popa Camp, 01°26'N, 65°47'W, alt. 1550 m, 18 Oct. 1987, R. L. Liesner & F. Delascio 22129 (U-digital image, VEN-digital image); • Camp XI, Cerro de la Neblina, 6.2 km NNE, Pico Phelps (= Neblina), 20.5 km ENE Neblina Base Camp, 00°51'45"N, 65°58'52"W, alt. 1350–1450 m, 26 Feb. 1985, J. M. Beitel 85272 (NY, VEN-digital image). • Bolívar: Kukenan Tepuy, alt. 2600 m, Jan. 1977, F. Delascio & C. Brewer 4937 (MO, UC, US), 05°13'N, 60°18'W, alt. 2550 m, 11 Apr. 1988, R. L. Liesner 23111 (MO, UC, US, VEN), 12 Apr. 1988, R. L. Liesner 23181 (MO, NY, U, UC, VEN); • Mt. Roraima, ledge above Rondon Camp, alt. 6900 ft [2103.12 m], Dec. 1884, im Thurn 226 (BM), ledge above Rondon Camp, alt. 6900 ft [2103.12 m], 1 Dec. 1927, G. H. H. Tate 466 (BM, 466 b at NY), alt. 2255–2620 m, 27 Sep. 1944, J. A. Steyermark 58753 (mixed coll. 58753 a at F, mixed coll. 58753 a at GH, MO, NY, US), 05°12'N, 60°42'W, alt. 2750–2800, 26 Aug. – 2 Sep. 1976, J. A. Steyermark et al. 112530 (U), 26 Aug. 1976, J. A. Steyermark et al. 112556 (NY, US, VEN-digital image), • Cima del Roraima-tepui, lower part of Valle de los Cristales, 05°11'58"N, 60°43'58.08"W, alt. 2680 m, 24 Mar. 2012, Y. Vivas et al. 3077 B (UC, VEN). Roraima, Nov. – Dec. 1931, N. J. Abbensetts 1 A (BM - 2 sheets). Brazil • Roraima: Dec. 1909, E. Ule 8491 (BM - text & illustration from original paper, MG), E. Ule 8617 (B).

Discussion.

Selaginella rhodostachya is a distinctive species that has often been misunderstood. It is characterized by its elliptic to elliptic-ovate median leaves with distinctly keeled midribs on the upper surfaces, with stomata along the keel, narrowly hyaline margins comprised of one or two rows of papillate idioblasts in a single row. Its margins are entire throughout or entire along proximal 3 / 4 and scarcely denticulate along the distal 1 / 4, or long-ciliate throughout, with obtuse to rounded apices, each tipped by 1–5 tooth-like projections, often caducous, or 2 divergent, long cilia. Additionally, it has elliptic to elliptic-ovate lateral leaves with margins and apices similar to those of the median leaves, as well as elliptic or narrowly ovate to ovate-lanceolate axillary leaves, with margins and tips also resembling the median leaves. Furthermore, S. rhodostachya has dorsal, ventro-axillary, or axillary rhizophores, short strobili measuring 2–5 mm in length, composed of ovate to ovate-elliptic, broadly elliptic, or broadly ovate sporophylls, with midribs, margins, and tips similar to those of the median leaves. It also has microsporangia arranged in two dorsal rows, megasporangia along two ventral rows, and megaspores that are light yellow, yellow, or cream.

Selaginella rhodostachya has been mistaken for S. brachyclada, which was included under the former by some authors (e. g., Alston et al. 1981; Cremers et al. 2007; Mostacero, 2008). The features used to distinguish these two species are described under S. brachyclada.

According to Jermy and Rankin (in Alston et al. 1981: 294), Alston was inclined to classify specimens with long-ciliate leaf margins and two divergent cilia at their tips as a distinct, new species previously identified as S. rhodostachya. However, Jermy and Rankin (in Alston et al. 1981: 292) considered these specimens more similar to S. valdepilosa (Fig. 5) and proposed the subspecies S. valdepilosa subsp. tricholoma to categorize them taxonomically. Additionally, in both S. rhodostachya and its segregate S. valdepilosa subsp. tricholoma, leaf marginal indument appears to be a variable morphological character. Some specimens, such as Steyermark 58753 at GH!, that were collected near the same locality as the type of S. valdepilosa subsp. tricholoma, match the typical S. rhodostachya. These specimens lack cilia on their lateral and axillary leaf margins; however, their median leaves have short- to long-ciliate margins along the distal 1 / 8 – 1 / 2 of the leaf laminae, with apices tipped by 1 or 2 divergent long cilia, which are characteristic of S. valdepilosa subsp. tricholoma. Similarly, specimens from Kukenan Tepui, Bolívar State, Venezuela, have sparsely ciliate leaf margins but match the types of both S. rhodostachya and S. valdepilosa subsp. tricholoma in all other features. It appears that in S. rhodostachya, the marginal leaf indument is often caducous even within geographically close populations. Accordingly, we agree with Smith (1995) and Cremers et al. (2007), who synonymized S. valdepilosa subsp. tricholoma under S. rhodostachya.

Specimens of S. rhodostachya described as S. valdepilosa subsp. tricholoma share characteristics with S. valdepilosa, such as notably long-ciliate leaf margins with apices tipped by long, divergent cilia. Nonetheless, S. valdepilosa (Fig. 5) is distinguished from the former by its more delicate, liverwort-like growth form, broad ovate-orbicular to orbicular lateral leaves with the lower surfaces completely covered by long idioblasts, and long cilia mostly along the acroscopic leaf margins with few on the proximal 1 / 4, as well as ovate-orbicular to orbicular median leaves. In fact, S. valdepilosa is quite similar morphologically to S. cyclophylla, but it differs in having long-ciliate (vs. entire) leaves, and median leaves with keeled (vs. plane) midribs on the upper surfaces of the laminae, which have conspicuous stomata throughout their length in 1–4 rows (vs. stomata absent or not visible on upper leaf surfaces). Additionally, the lateral leaf lower surfaces are shiny with long, papillate idioblasts throughout (vs. leaf lower surfaces dull, made up of seemingly roundish or elongate cells without papillate idioblasts, except on the proximal 1 / 3 and near the lamina bases, where they are present). Furthermore, the narrow habit and imbricate leaves of S. cyclophylla resemble those of some liverworts from the genus Porella L. [e. g., P. navicularis (Lehm. & Lindenb.) Pfeiff. and P. obtusata (Taylor) Trevis.].

The type collection of S. rhodostachya is based on im Thurn s. n., but as Baker (1886: 211) noted, the parentheses around the number (271 *) in the protologue refer to the species’ position in the sequence of the authors’ “ Synopsis Filicum ”. In fact, “ im Thurn 271 ” corresponds to Trichomanes macilentum Bosch (see Baker, 1886: 212). Furthermore, E. F. im Thurn (1887) published a verbatim reproduction of his “ Notes on the plants observed during the Roraima expedition of 1884, ” which Baker also treated, and reiterated the previous information (Baker, 1887 b: 288 and 296). Additionally, Baker (1887 a: 112), referencing his earlier publication (i. e., Baker 1887 b), listed the type of S. rhodostachya as “ im Thurn s. n. ” It is worth noting that in Baker (1887 a: 112), S. rhodostachya is listed as species number “ 292 ” rather than “ (271 *) ”. Therefore, the correct type citation for S. rhodostachya should be “ im Thurn s. n. ”

Furthermore, an isotype of S. rhodostachya at BM is labeled “ im Thurn 226, ” but this is also likely a mistake, as the type collection was not numbered, and the number “ 226 ” actually refers to the type collection of S. vernicosa Baker (Baker, 1886: 220). It seems that whoever created this label unintentionally misnumbered it. This is clear because a) the holotype at K lacks an original label and only has a later note by an unknown source; b) the isotype label at US differs from that at K, which was initially numbered in pencil as “ 227, ” then crossed out and replaced with “ 271 ” in pencil, matching Baker’s numbering system; and c) there is a penciled note on the isotype label at BM, probably by Alston, indicating that “ 271 ” is wrong. Therefore, it’s reasonable to assume that the specimens labeled “ 226, ” “ 227, ” and “ 271 ” at BM and “ 271 ” at US are mislabeled, mostly corresponding to the species’ position or number in the sequence of Baker’s “ Synopsis Filicum ”. Accordingly, “ 227, ” and “ 271 ” are considered duplicates of the original collection of S. rhodostachya (i. e., im Thurn s. n.) and, thus, isotypes. Sadly, this was likely missed by Alston et al. (1981: 297), probably because the first author’s original work on Selaginella was completed many years after his untimely death in 1958 (see Alston et al. 1981: 233–236).

A similar confusion exists regarding the correct collection number for the type collection of S. valdepilosa subsp. tricholoma. The label on the holotype reads “ J. A. Steyermark 58016, ” but this appears to be a mistake for 58916. The label attached to the holotype was typed with a typewriter likely at and sent by the US, while the labels attached to the isotypes at F, GH, MICH, MO, and US are printed (except for the label on a paratype at US, 58916 bis [a handwritten “ 9 ” superimposed on a typed “ 0 ”, which is almost identical to the one on the holotype]) and that also represents S. valdepilosa subsp. tricholoma. Originally, it seemed that there were two similar collection numbers presumably made on the same day, with a gap of “ 900 numbers ” between them, which represent the same new subspecies, but this is unlikely. Indeed, after the revision of Steyermark’s collection book at MO, it was determined that the collection number Steyermark 58016 corresponds to an angiosperm rather than a Selaginella (R. C. Moran and B. K. Holst, pers. com.). Therefore, it is most likely that, when the label for the specimen sent to the British Museum for Alston’s study was typed, a “ 0 ” was entered instead of a “ 9. ” Accordingly, it is concluded that the true collection number for the type collection of S. valdepilosa subsp. tricholoma should be “ 58916 ” rather than “ 58016 ”.

The isotypes at F, GH!, MICH!, MO!, NY!, and S! are mixed collections, which we separated and identified as follows: a = S. rhodostachya and b = S. vernicosa (Alston previously classified them this way). Additionally, J. A. Steyermark 58753 at F! was determined by us to be a = S. rhodostachya and b = S. vernicosa, while Tate 466 at NY! is also a mixed collection, which we identified as a = S. vernicosa and b = S. rhodostachya.

Finally, other collections (Guyana: Forest path to the top of Kaieteur Fall, Oct 1888, [G. S.] Jenman s. n., NY!; Kaieteur, Oct 1888, [G. S.] Jenman s. n., NY!) cited by Alston et al. (1981) under S. rhodostachya actually refer to S. sandwithii Alston.