Cyclopogon guayanensis (Lindley) B.M.Carvalho & Meneguzzo

Cyclopogon guayanensis (Lindley) B.M.Carvalho & Meneguzzo , comb. nov.

Goodyera guayanensis Lindley (1840: 494) . Spiranthes guayanensis (Lindl.) Cogniaux (1895: 209) . Brachystele guayanensis (Lindl.) Schlechter (1920: 373) . ( Figure 1 View FIGURE 1 )

Type:— GUYANA. Berbice: sine loco acurato, 1838, R.H. Schomburgk 698 [lectotype designated by Szlachetko & Rutkowski (2008: 142): K-L barcode K000079771 !, isolectotypes BM barcode BM000077222 !, G barcode G00169152 !, K barcode K000079081 !, K barcode K001044093 !, TCD barcode TCD0007956 !, W-R not seen].

= Spiranthes aguacatensis Reichenbach (1855: 214) . Gyrostachys aguacatensis (Rchb.f.) Kuntze (1891: 664) . Brachystele aguacatensis (Rchb.f.) Schlechter (1920: 371) .

Type:— COSTA RICA. Alajuela: Monte de Aguacate, 1845–1848, A.S. Oersted 7214 [lectotype designated by Szlachetko & Rutkowski (2008: 142): C barcode C10016151!].

= Spiranthes hostmannii Reichenbach ex Grisebach in Grisebach (1864: 640). Gyrostachys hostmannii (Rchb.f. ex Griseb.) Kuntze (1891: 664).

Type:—SURINANE. s. loc., 1843, F.W. Hostman 386 [lectotype designated by Szlachetko & Rutkowski (2008: 142): W-R!, isolectotypes P!, W!].

= Spiranthes brenesii Schlechter (1912: 481) . Brachystele brenesii (Schltr.) Schlechter (1920: 372) .

Type:— COSTA RICA. Colinas de Piedades, 19 May 1901, A.M. Brenes 14282 [lectotype designated by Szlachetko & Rutkowski (2008: 142): US 579490, US 1083417, isolectotypes K barcode K000463265!, G barcode G00168545!].

Description:— Geophyte, erect, sympodial herb. Roots 0.3–0.6 cm diam., tuberoid, pilose. Stem inconspicuous, glabrous. Leaves rosulate, absent at anthesis, 9.0–14.0 × 0.9–3.2 cm, 2–4(–6) per stem, convolute, non-articulate, narrowly elliptic, sessile, glabrous, chartaceous, veining camptodromous. Inflorescence (9.0–)15.0–27.0(–38.5) cm long, terminal, simple racemose, green; peduncle (7.5–)9.5–15.5(–27.0) cm long, pilose towards rachis, floral bracts 0.3–0.5 × 0.2 cm, conspicuously pilose; rachis 4.5–12(–16) cm long, multiflorous, pilose. Flowers congested, (9–)30– 120(–160) per inflorescence, resupinate; ovary with pedicel 0.5–0.7 × 0.2 cm, non-articulate, pilose, white; sepals 0.3– 0.45 × 0.1–0.2 cm, oblong-lanceolate, abaxially pilose, lateral sepals slightly oblique and partly adnate at proximal part; petals 0.25–0.35 × 0.1–0.15 cm, oblong-oblanceolate, glabrous, partly adnate to the dorsal sepal; labellum unguiculate, glabrous, hypochile 0.15–0.18 × 0.25–0.28 cm, transversely elliptic, base with pair of short, corneous glans obliquely pointed inwards, canaliculate, center greenish, margin entire, epichile 0.2–0.25 × 0.1–0.15 cm, obovate, apex truncate, margin erose; column 0.2–0.25 cm long, clavate, foot 0.1–0.15 cm long, oblique, partly adnate to the column, anther bilobed, pair of pollinia oblong, viscidium dorsal, widely lacriform, rostellum transversally elliptic. Capsule 0.4–0.5 × 0.2–0.3 cm, septicidal, ellipsoid.

Distribution and ecology: — Cyclopogon guayanensis was previously known to be widespread, occurring in Mexico, Central America, Trinidad, northern South America, and Bolivia ( Szlachetko et al. 2005, Baksh-Comeau et al. 2016, POWO 2022), with a single record from northern Brazil [voucher specimen Black & Camargo 50-8850 (IAN)]. The fact that virtually all collections from Central Brazil are from the 21 st century, and from man-altered habitats, such as pastures dominated by grasses such as Andropogon Linnaeus (1753: 1045) , Axonopus Palisot de Beauvois (1812: 12) and Urochloa Palisot de Beauvois (1812: 52) suggest that deforestation of the Cerrado domain, with grasslands and pastures becoming much more common, could be contributing to expand its distribution ( Figure 2 View FIGURE 2 ). On average, less than 60% of each 1° latitude × 1° longitude square in the Cerrado biome is now covered with natural vegetation (Diniz Filho et al. 2009). However, the van Glocker 386 collection from Bahia (exact location uncertain so not mapped in Figure 2 View FIGURE 2 ) dates from 1842, so perhaps this species has a history of sporadic occurrence in Central Brazil that has been intensified by recent urban and pasture expansion. Even with this geographic expansion, C. guayanensis is not sympatric with any of the other Brachystele species sampled by Salazar et al. (2018) [ Brachystele camporum , Brachystele cyclochila , Brachystele dilatata ( Lindley 1840: 474) Schlechter (1920: 373) , Brachystele subfiliformis , Brachystele ulaei ( Cogniaux 1895: 107) Schlechter (1920: 374) and Brachystele unilateralis , the type species]. These species of Brachystele (which can be considered Brachystele sensu stricto as they include the type) form a wellsupported clade (bootstrap support = 95%; Salazar et al. 2018) that also included Odontorrhynchus chilensis Richard in Gay (1852: 475) Garay (1982: 341) and was nested within the Pelexia clade ( Salazar et al., 2018). All these species of Brachystele s.s. have a more or less austral distribution in South America (eastern and southern Brazil, Peru, Paraguay, Uruguay, Argentina, and Chile) and occur in either subtropical or tropical montane environments (above 1000 m), while B. guayanensis occurs in the tropical zone, and at lower altitudes (usually below 1000 m). It grows in natural open vegetation, urban grasslands, and rural, anthropogenic grasslands (such as pastures).

Additional specimens examined: — BOLIVIA. Beni: Trinidad, Missiones Guarayos , 250m, 15°53'24"S, 63°11'24" W, September 1926, E. Werdermann 2572 (MO) GoogleMaps ; Trinidad, Missiones Guarayos , 159m, 14°50'06" S, 64°54'16" W, December 1926, E. Werdermann s.n. (AMES) GoogleMaps . BRAZIL. Bahia: 1842, E. F. von Glocker 386 (BM). Distrito Federal: Brasília, área verde entre a SQN 416 e a L4 Norte , final da Asa Norte, 1012m, 15°44’34” S, 47°53’17” W, 1 December 2009, J.A.N. Batista & T.E.C. Meneguzzo 2927 (BHCB) GoogleMaps ; Brasília, Asa Norte, área verde entre SQN 416 e a L4 Norte , final da Asa Norte , 1012m, 15°46’47” S, 47°55’47” W, 3 January 2011, J.A.N. Batista & K. Proite 3117 (BHCB, CEN) GoogleMaps ; Asa Norte , final, no gramado da entrequadra 416-216, 1018m, 15°44’09” S, 47°53’20” W, 7 November 2009, T.E.C. Meneguzzo 456 (UB) GoogleMaps ; Brasília, Gramado no final da Asa Norte , onde terminam as SQNs 416 e 216 15°44’13” S, 47°53’33” W, 9 November 2009, F.M. Santos 240 (BHCB) GoogleMaps ; Brasília, Asa Norte, 13 November 2020, C.E.B. Proença 5579 (UB). Goiás: Alto Horizonte , Fazenda Cajás , próximo à sede da fazenda, 14°11’44” S, 49°20’19” W, 30 October 2010, J.E.Q. Faria 911 (CEN, HUEG, UB) GoogleMaps ; Alto Horizonte, Região da Sururuca, Fazenda Cajás , 385m, 14°11’47” S, 49°17’44” W, 22 October 2020, J.E.Q. Faria 10697 (UB) GoogleMaps ; Alvorada do Norte, Ponto 3 da fitossociologia, perto da gruta, afloramento calcário (muda proveniente da casa da vegetação), 400 m, 14°31’30” S, 46°43’43” W, 3 December 2003, G. Pereira-Silva 8108 (BHCB, CEN, HUEFS). GoogleMaps Mato Grosso: Colíder , praça central da cidade, 322m, 10°48’22” S, 55°27’16” W, 18 October 2014, M.E. Engels 2631 (CNMT, HERBAM, MBM, SP) GoogleMaps ; Colíder, Avenida Marechal Candido Rondon, região de ecótono entre a Floresta Amazônica e Cerrado , floresta do “Planalto dos Parecís’’, mosaico das fitofisionomias densa, estacional semidecidual e savana, 324m, 10°50’13” S, 55°27’21” W, 17 October 2014, M.E. Engels 2633 (CNMT, HERBAM, MBM, TANG); GoogleMaps Colíder, Avenida Marechal Candido Rondon, região de ecótono entre a Floresta Amazônica e Cerrado, floresta do “Planalto dos Parecís’’, mosaico das fitofisionomias densa, estacional semidecidual e savana, 320m, 10°48’56” S, 55°27’14” W, 19 October 2014, M.E. Engels 2634 (CNMT, HERBAM, MBM, TANG) GoogleMaps ; Colíder, Avenida Tancredo Neves , 322m, 10°48’56” S, 55°27’13” W, 8 October 2014, M.E. Engels 2686 (HERBAM, MBM, TANG) GoogleMaps ; Colíder, Avenida Marechal Candido Rondon , 320m, 10°50’35” S, 55°27’13” W, 7 October 2014, Engels, M. E. 2687 (CNMT, HERBAM, MBM, TANG) GoogleMaps ; Colíder, Cachoeira do Mercúrio , 10°49’04” S, 55°27’03” W, 5 October 2014, M.E. Engels 4168 (MBM, UNOP) GoogleMaps ; resgate de FLORA da UHE Colíder , estada de acesso, 10°49’04” S, 55°27’03” W, 6 October 2014, M.E. Engels 169 2699 Copel 1 (HERBAM) GoogleMaps ; resgate da FLORA da UHE Colíder, Lote G de Supressão , 273m, 11°00’37” S, 55°23’50” W, 26 October 2015, M.E. Engels 3723 (CNMT, HCF, MBM, TANG) GoogleMaps ; resgate da Flora da UHE Colíder, no canteiro central da Avenida Marechal Candido Rondon, 320m, 10°48’43” S, 55°27’22” W, 14 November 2014, M.E. Engels 4074 (MBM). GoogleMaps Minas Gerais: Uberlândia, vereda situada a 22 km do centro da cidade, nas margens da estrada que leva a Campo Florido, 19°03’ S, 48°21’ W, 3 November 2013, J.C.F. Cardoso 7 (HUFU). GoogleMaps Pará: Ilha de Marajó , Fazendinha, 0°43’00” S, 48°31’24” W, 1 February 1950, G.A. Black 50-8850 (IAN) GoogleMaps . COSTA RICA. Las Concávas [Concávas, Cartago?], 1351m, 9°50’15” N, 83°53’25” W, 4 May 1926, C.H. Lankester 1126 (AMES). GoogleMaps Alajuela: 1061m, 10°05’19” N, 84°28’15” W, May 1913, A. Tonduz 17629 (P) GoogleMaps . FRENCH GUIANA. Locality uncertain, F.M.R. Leprieur s.n. (P) ; in Guiana gallica, F.M.R. Leprieur 105 (P) ; Locality uncertain, 1842, E.M. Mélinon 131 (P) ; Locality uncertain, 1821, G.S. Perrottet s.n. (P) ; GUATEMALA. El Quiché: 15°00'43" N, 91°02'18" W, 8 June 1933, C.L. Lundell 3704 (AMES) GoogleMaps ; GUYANA. Guiana Anglica: August 18??, R.H. Schomburgk 698 (L, MO) . HONDURAS. Comayagua: 14°26'05" N, 87°00'00" W, 13 June 1933, J.B. Edwards 432 (AMES) GoogleMaps . MEXICO. Chiapas: 1227m, 15°41’33” N, 92°37’12” W, 17 June 1945, E. Matuda 5998 (AMES). GoogleMaps Veracruz: 18m, 17°57’53” N, 94°43’04” W, 5 July 1910, C.R. Orcutt s.n. (AMES) GoogleMaps . PANAMA. Canal Zone : 883m, 8°47'30" N, 82°49'44" W, 8 June 1923, W.R. Maxon 6862 (AMES) GoogleMaps . SURINAM. August 18??, F.W.R. Hostmann 841 (paratype, AMES barcode 00104457 image!, CAS, G barcode G00169118 image!, K barcode K000079094 image!, P barcodes P00340990 image!, P00340991 image!) ; Unknown s.n. (L) ; Paramaribo: in pascuis arenosis prope Paramaribo, August 18??, H.A.H. Kegel 468 (K, P) . TRINIDAD AND TOBAGO. Trinidad : H. Krueger s.n. (paratype GOET barcode GOET013864 image!) ; Saint Ann's , 10°24'26" N, 61°23'26" W, 1931, W.E. Broadway s.n. (AMES, BM) GoogleMaps ; Governor's house, 10°39'26" N, 61°31'05" W, 26 June 1904, W.E. Broadway s.n. (NY) GoogleMaps . VENEZUELA. Bolívar: between Kamarata and Uruyen, savanna at south end of Auyan-tepui , 5°42'45" N, 62°24'31" W, 28 April 1964, J.A. Steyermark 93182 (AMES, F) GoogleMaps ; San Ignacio de Yuruaní , 850m, 5°00' N, 61°10' W, 8 May 1988, R.L. Liesner 24378 (MO) GoogleMaps ; Cuenca de Río Arabopó, Sabanas en el sector SE, aprox. 5 km. de la frontera con el Brasil , 4°55'48" N, 60°37'48" W, 14 June 1985, O. Huber 10502 (NY) GoogleMaps ; Mérida: between Mucuchachi and Canaguá, near stream above El Bado , 8°08'44" N, 71°24'36" W, 6 May 1944, J.A. Steyermark 56330 (AMES, F) GoogleMaps .

Taxonomic discussion:— This species was first described by Lindley (1840) as Goodyera guayanensis Lindl. based on Robert Hermann Schomburgk’s collection from Berbice, at that time a part of British Guiana, currently Guyana. It has been treated in the taxonomic literature as Brachystele guayanensis (Lindl.) Schltr. for over a hundred years, ever since its combination in Schlechter’s taxonomic treatment of the Spiranthinae ( Schlechter 1920, Correa 1955, Burns-Balogh & Robinson 1983, Szlachetko & Rutkowski 2008). Szlachetko et al. (2005) and POWO (2022)

agree that the three heterotypic synonyms are conspecific to it, with types from Central America and the Guianas. The spelling error B. guyanensis is relatively frequent for this species (e.g. Schultes 1956, Pupulin 2002, IPNI 2024). Correction should not be applied to this epithet as it was not a misspelling but Lindley’s deliberate choice. He cited the species as “Hab. in Guayana” ( Lindley 1840: 494) and the toponym “Guayana” was (and still is) used interchangeably with Guyana ( Berry & Riina, 2005 and references therein).

The question of why this species has never been included in Cyclopogon naturally arises. This appears to be because the more obvious characters align it with Brachystele , such as the small, white, subsessile flowers, organized in dense, many-flowered inflorescences. Superficially, C. guayanensis is remarkable like several species of Brachystele , such as B. camporum ( Lindley 1840: 473) Schlechter (1920: 372) , and with such small flowers it is difficult to evaluate subtle differences in shape and texture from herbarium material. The characters that it has in common with Cyclopogon , on the other hand, are cryptic, such as the presence of glandular hairs on the outer surface of the calyx, curvature and position of floral parts and the position of the viscidium. Upon closer inspection of images of live plants that have only recently become available, it was obvious to us that all species of Brachystele sensu stricto that were sampled by Salazar et al. (2018) have similar floral characteristics that differ from those of Cyclopogon ( Figs. 1 View FIGURE 1 and 3 View FIGURE 3 ).

In Brachystele s.s. the dorsal sepal is rigidly protruding and is subtended and supported by the closely appressed, parallel lateral petals, forming a flattish hood that is the most prominent floral structure ( Figure 3 A–B View FIGURE 3 ). The labellum is somewhat rigid and down-pointing, and in B. unilateralis ( Poiret 1798: 569) Schlechter (1920: 374) , B. camporum and B. subfiliformis ( Cogniaux 1906: 544) Schlechter (1920: 374) it is rather flat, curls like a tongue, and has an obtuse to emarginate apex. In the B. ulei-dilatata-cyclochila clade sensu Salazar et al. (2018) it is fleshy and navicular, with an acute tip ( Figure 3 C, F View FIGURE 3 ), giving the general flower shape a certain degree of lateral compression, so that floweraccess is circular or vertically wide-elliptic. In Cyclopogon , on the other hand, the dorsal sepal is usually poorly developed, upturned at the tip, and flanked on either side by the small, thin lateral petals. Both dorsal and lateral sepals in Cyclopogon flowers have showy, upturned tips that provide a visual cue. It is the labellum that is the most prominent floral feature in Cyclopogon , thinly petaloid and expanded at the apex, sometimes to the point that the sides curl upwards or sideways. The Cyclopogon flower is thus strongly compressed in the dorsal-ventral plane so that floweraccess is horizontally elliptic. There is also apparently a difference in the function of the lateral sepals between the two genera. In Brachystele s.s., the lateral sepals are conspicuous, frequently white and aliform ( Figure 3 A–B View FIGURE 3 ), except in B. cyclochila ( Kraenzlin 1911: 36) Schlechter (1920: 373) and are part of the floral visual cue. In Cyclopogon , the lateral sepals are usually greenish and stiff, and subtend the showy labellum ( Figure 3 G–H View FIGURE 3 ), providing it with structural support.

Cyclopogon guayanensis thus follows the Cyclopogon floral bauplan but with shortened, small, densely aggregated, subsessile flowers ( Table 1 View TABLE 1 ) with petaloid sepals and petals. Although the sepals and petals in C. guayanensis are white and petaloid, as also occurs in some Brachystele s.s., there is no flattish hood, the dorsal sepal is slightly upturned, as are the lateral petals that flank the dorsal sepal on each side as in Cyclopogon , while the thin, truncate or emarginate labellum is the most prominent floral feature ( Figure 1 View FIGURE 1 ). Convergence can be seen in two features: 1) the alate-explanate lateral sepals of C. guayanensis that are similar to those of Brachystele s.s. (although it is possible to see that their bases subtend the labellum as in Cyclopogon ); and 2) the downward curve and shape of the labellum in C. guayanensis that emulates that of some Brachystele , B. camporum for example, although its thin texture and groove is as in Cyclopogon .

The general similarity in both floral and vegetative morphology between Cyclopogon guayanensis and Brachystele s.s. is probably due to convergence resulting from the shift to grassland environments and to the different pollinators associated with these environments. Similar cases of convergence have been documented in other groups of Spiranthinae , such as the remarkable similarity in floral morphology of Sacoila Rafinesque (1837: 86) , Stenorrhynchos Richard ex Sprengel in Sprengel (1826: 677), Coccineorchis Schlechter (1920: 434) , Svenkoeltzia Burns-Balog (1989: 12) and some Dichromanthus Garay (1982: 313) , as a result of independent adaptation to pollination by hummingbirds (Salazar et al. 2011) and between Mesadenus Schlechter (1920: 367) , and Espinhassoa Salazar & Batista in Salazar et al. (2019: 291), whose tiny, greenish or whitish flowers, narrow floral tube, and nocturnal floral scent suggest pollination by small nocturnal insects, perhaps dipterans or lepidopterans ( Salazar et al. 2019). The major change in the circumscription of Cyclopogon with the inclusion of C. guayanensis is a dramatic increase in the number of flowers per inflorescence, a reduction in flower size, and the fact that sepals may be petaloid. The changes in the morphological circumscription of Brachystele and Cyclopogon with the transfer of B. guayanensis to the latter genus is presented in Table 1 View TABLE 1 .

Although the sampling of Brachystele in phylogenetic analyses is still small and some Mesoamerican species have not yet been sampled, the latter fit the geographic pattern of C. guayanensis as described above and we hypothesize that they do not belong to Brachystele s.s. Several Mesoamerican species previously attributed to Brachystele have now been transferred to other genera, such as Schiedeella Schlechter (1920: 379) , Mesadenus Schlechter (1920: 367) , Microthelys Garay (1982: 336) and Galeottiella Schlechter (1920: 36) based on morphology or molecular results. Additional molecular sampling of species currently positioned within Brachystele , as well as additional molecular markers to increase clade support, would also be necessary to confirm if Brachystele s.s. is confined to eastern and southern Brazil and adjacent southern cone countries ( Argentina, Uruguay, and Chile). The current sample that suggests this geographic distribution is admittedly insufficient (only c. 30% of the genus; Salazar et al. 2018). Except for B. guayanensis , other species of Brachystele from Mexico, Central America, and northwestern South America have not been sampled.

If Brachystele guayanensis and Veyretia were to be transferred to Cyclopogon the latter genus would become monophyletic, so there is arguably a case for Veyretia to be transferred to Cyclopogon along with B. guayanensis . The ongoing Ph.D. thesis on the phylogenetics and biogeography of Cyclopogon and Veyretia by one of us (BMC) using the same molecular markers as Salazar et al. (2018) but sampling an additional 21 species of these genera (78 terminals and 47 species) has confirmed that both Brachystele guayanensis and Veyretia are nested within Cyclopogon . Full results of this study will be published elsewhere, including taxonomic changes associated with Veyretia . Although the transfer of B. guayanensis to Cyclopogon does not solve the problem of the lack of monophyly of the large genus Cyclopogon (since Veyretia remains nested within it), it does solve the problem of the lack of monophyly of Brachystele . Under the present circumscription, Brachystele can be recognized as a monophyletic genus of rather limited southerly/montane distribution ( Figure 3 View FIGURE 3 ). The new combination Cyclopogon guayanensis (Schltr.) B.M.Machado & Meneguzzo , which we are proposing in advance of the probable new combinations of species of Veyretia in Cyclopogon , is made partially for this reason. Another reason is so that the recent, detailed information on its floral biology, pollinators, micromorphology, and nectar chemistry (Meneguzzo et al., unpublished results) is not spuriously associated to Brachystele s.s.