Introduction

Carex L., with about 2000 species, is one of the largest genera of Angiosperms (e.g. Global Carex Group 2015, 2016). It has a cosmopolitan distribution, with more than two hundred species distributed in the Neotropics (Govaerts & al. 2018+). Despite its relative species richness, further study is still needed in Central and South America. The efforts of Reznicek (1986), Reznicek & González-Elizondo (1995) and Wheeler (1987, 1990, 1996, 2002, 2009, among many other works) have greatly clarified the taxonomy of several taxonomic groups in these regions. Yet, other groups remain understudied, as evidenced by a number of recently described new species (Jiménez-Mejías & Escudero 2016; Jiménez-Mejías & Roalson 2016; Poindexter & al. 2017).

This publication is a continuation of the revision work started with a first set of miscellaneous notes published by Jiménez-Mejías & al. (2016a). Here we continue to present new relevant taxonomic, nomenclatural, and chorological information, as a result of studying additional herbarium collections.

Material and methods

Material from the following herbaria was studied: A, BOZ, DUKE, HAL, MICH, MO, NY, SI, UPOS and US (abbreviations according to Thiers 2017+). High-resolution images available on the Internet from the herbaria CORD, F, G, K, LE, LPB and VEN were also carefully examined. Specimens were identified using the specialized taxonomic literature cited under each taxon. The species are presented in alphabetical order. The proposed synonymization is discussed at the end of the manuscript. Accepted names follow the World Checklist of Cyperaceae (Govaerts & al. 2018+). Terminology of the inflorescence prophylls (utricles and cladoprophylls) follows the suggestions in Jiménez-Mejías & al. (2016b).

Results and Discussion

Carex buxbaumii Wahlenb. in Kongl. Vetensk. Acad. Nya Handl. 24: 163. 1803. — Lectotype (designated by Moberg & Nilsson 1991: 291): Sweden, in Sueciae paludosis, s.d., G. Wahlenberg (UPS-THUNB 21771 [digital image!]).

Illustrations — Ball & Reznicek (2002: 407), Jermy & al. (2007: 451).

Remarks — First record of this Holarctic, predominantly boreal species for the South American continent, constituting the known absolute southernmost native limit of the species. Carex buxbaumii, C. leptalea Wahlenb., C. limosa L., C. livida (Wahlenb.) Willd., and C. lurida Wahlenb. display a remarkable trans-Caribbean disjunction, being the five species widely distributed in North America and appearing disjunct in the N Andes (Govaerts & al. 2018+), and also on the Island of Hispaniola in the case of C. leptalea, C. limosa and C. lurida. Such disjunction seems to be related to the bird migration via the American Atlantic flyway (Jiménez-Mejías & al. 2016a).

The coordinates provided in the voucher actually lies on the Colombian side of the border, although still pretty close to Venezuela. Since the coordinates are given as approximates and since we cannot authentically discern whether the mistake is in the coordinates or in the country citation, we cite the country as shown on the specimen label.

There is some confusion regarding the typification of the name Carex buxbaumii Wahlenb. and its relationship with the illegitimate C. polygama Schkuhr (Beschr. Riedgras. 1: 84. 1801, nom. illeg., non C. polygama J. F. Gmel., Syst. Nat. ed. 13[bis]: 145. 1791). While Moberg & Nilsson (1991) correctly typified C. buxbaumii Wahlenb. on a Wahlenberg's herbarium material, Egorova (1999) considered that C. buxbaumii a replacement name for C. polygama Schkuhr., and thus proposed a new lectotypification through the designation of the type material of C. polygama Schkuhr [Egorova (1999: 389): Denmark, Zealand, in uliginosis Siaellandiae, Jul 1799, M. Vahl s.n. (HAL 0103626!)]. However, the case is not as straight forward as it seems. While Wahlenberg (1803) cited Schkuhr's name in the protologue of C. buxbaumii, he also made reference to a Buxbaum's preLinnaean work. Since there is no evidence that Wahlenberg wanted to prioritize one citation over the other, and that he actually created the new name after Buxbaum, C. buxbaumii Wahlenb. cannot simply be treated as a replacement name for C. polygama Schkuhr. Moreover, Schkuhr (1801) does not make any reference in his protologue to Buxbaum's work, rejecting the possibility that this name could also be based on Buxbaum's plant. As a consequence, the earlier Moberg & Nilsson (1991) typification must be considered valid for C. buxbaumii Wahlenb., while Egorova's (1999) typification is in form an effective lectotypification of C. polygama Schkuhr, but without any effect on Wahlenberg's name.

Additional specimen examined — Venezuela: Zulia: Distrito Perijá, Sierra de Perijá, Serranía de los Motilones, mesa below international boundary on main ridge, [10°00′13″N, 72°58′c.25″W], c. 3000 m, 27 Jun — 5 Jul 1974, Tillet & Honig 747-792 (NY).

Carex catharinensis Boeckeler in Allg. Bot. Z. Syst. 2: 191. 1896. — Lectotype (designated by Jiménez-Mejías & al. 2016a: 729): Brazil, Santa Catarina, Serra Geral, Dec 1890, E. H. G. Ule 1610 (K 000584646 [digital image!]).

Illustrations — Silveira & Longhi-Wagner (2012: 389).

Remarks — We hereby confirm the expected occurrence of this species in Uruguay, given that it was already known north and south of Uruguay in Brazil (Silveira & Longhi-Wagner, 2012) and Buenos Aires province in Argentina (Jiménez-Mejías & al. 2016a). Herter's (1953) report of Carex fuscula d'Urv. was presumably a misapplication of C. catharinensis, and thus C. fuscula is most likely absent in Uruguay.

Additional specimens examined — Uruguay: Canelones Department, Arroyo del Sauce, 1 km S of km 18 on road W of Solis, 26 Nov 1943, Bartlett 20857 (MICH, US).

Carex divisa Huds., Fl. Angl.: 384. 1762. — Lectotype (designated by Molina & al. 2006: 1010): Herb. Sloane 127: 47, second specimen from right, s.d., W. Sherard s.n. (BM-SL!). — Epitype (designated by Molina & al. 2006): United Kingdom, [Kent], Isle of Sheppey, s.d., S. Goodenough s.n. (K 000960405 [digital image!]).

Illustrations — Jermy & al. (2007: 255).

Remarks — This is the first citation of this presumably introduced species in South America, wherein it was apparently formerly misidentified as Carex praegracilis W. Boott in Río de la Plata region (Herter 1953; Myndel-Pedersen 1968; in both works named as C. marcida Boott). The Argentinian material unequivocally matches C. divisa after comparison with materials of this species and C. praegracilis. In addition, this voucher also matches C. divisa according to Flora of North America key (Ball & Reznicek 2002). The Uruguayan one, despite being immature, also falls much closer within the variation reported for C. divisa than for C. praegracilis. Another Eurasian Carex species, C. divulsa Stokes, is also reported as introduced in some Argentinian provinces of the Río de la Plata Region (Myndel-Pedersen 1968; Jiménez-Mejías & al. 2016a).

Additional specimens examined — Argentina: Buenos Aires: Pdo. de Magdalena, Reserva de la Biosfera “Parque Costero del Sur”, Reserva “El Destino”, 30 Oct 2002, Torres Robles & García 976 (MO). — Uruguay: Montevideo, Pajas Blancas, en pradera anegada marítima, 26 Nov 1937, Rosengurtt B2022 (NY).

Carex haematopus Jim.-Mejías & Roalson in Phytotaxa 266: 23. 2016. — Holotype: Colombia, Nariño, Municipio de Pasto, corregimiento del Encanto, Isla La Corola, 2660 m, 1 Nov 1977, O. de Benavides 1197 (NY!).

Illustrations — Jiménez-Mejías & Roalson (2016: 25).

Remarks — These represent the first records of this recently described species for Ecuador. Two of the new records are placed close to, and both north and south, to Kükenthal's (1909) citation of Carex aematorhyncha É. Desv. in Ecuador (“auf Triften bei Otavalo in der Prov. Imbabura; Sodiro 199/64”). Carex aematorhyncha is similar to C. haematopus, but can be easily distinguished by utricle size (Jiménez-Mejías & Roalson 2016). Accordingly, we consider that the supposed record of the former very probably refers to C. haematopus, and that the reported presence of C. aematorhyncha in Ecuador (e.g. Jprgensen & León-Yáñez 1999; Govaerts & al. 2018+) should be dismissed.

Additional specimens examined — Ecuador: Azuay: vicinity of the lake in the valley of the río Surucuchu (a branch of the río Matadero), 12–20 km W of Cuenca, 9800–10300 ft, Camp 4213 (NY, US). — Carchi: Hacienda Ingüesa, 3000–3200 m, 9 Nov 1952, Acosta-Solís 21625 (US). — Pichincha: 3600 m, Oct 1887, Sodiro 344 (US).

Carex larensis Steyerm. in Fieldiana, Bot. 28: 66. 1951. — Lectotype (designated here): Venezuela, Lara, between Buenos Aires and Páramo de las Rosas, altitude 2285–3290 m, 11 Feb 1944, Steyermark 55470 (F 1263860 [digital image!]; isolectotypes: US!, VEN [digital image!]).

Illustrations — Steyermark (1951: 64).

Remarks — So far cited only from its type locality in the state of Lara in Venezuela, here we formally report its presence in Colombia and also in the Venezuelan state of Mérida.

Carex larensis is a distinct species from N South America and belongs to C. sect. Ceratocystis Dumort. (Steyermark 1951; Global Carex Group 2016). It has brown to chocolate-brown staminate and pistillate glumes and 2-3(4) mm long, elliptic to obovate, often inflated utricles, with beaks 0.3-0.7(-1.2) mm long, straight or slightly curved, all ascending-spreading, or the middle and lower ones spreading. Although these characteristics clearly distinguish C. larensis from the S South American C. sagei Phil., C. larensis approaches the N hemisphere C. viridula Michx. in these features. However, the latter has lighter glumes (pale brown, stramineous or hyaline), and the middle utricles of each spike are always spreading.

Steyermark (1951) cited two collection numbers as the type material of Carex larensis: “Type in herb. Chi. Nat. Hist. Mus., collected in swampy meadow, between Buenos Aires and Paramo de las Rosas, state of Lara, alt. 2285–3290 m., February 11, 1944, Julian A. Steyermark 55470; also same locality, 55467”. Although the semicolon separating numbers 55470 and 55467 might indicate an intention of designating the first as holotype, we understand that it is not sufficiently clear. Accordingly, we perform a formal lectotype designation on the specimen at F from the first collection.

Additional specimens examined — Colombia: Magdalena: Sierra Nevada de Santa Marta, alrededores de cabeceras de Río Ancho, Páramo de Macotama, c. 3730 m, 15 Feb 1959, Barclay & Juajibioy 6981 (MO, US); Sierra Nevada de Santa Marta, alrededores de cabeceras de Río Ancho, Páramo de Macotama, c. 3520 m, 15 Feb 1959, Barclay & Juajibioy 6970 (MO, US); Sierra Nevada de Santa Marta, alrededores de cabeceras de Río Ancho, Páramo de Macotama, c. 3490 m, 16 Feb 1959, Barclay & Juajibioy 7020 (MO); La Guajira, Sierra Nevada de Santa Marta, Laguna Sabaca, nacimiento Río San Miguel, 3700–3900 m, 16 Aug 1986, Cuadros & Gentry 2706 (MO). — Boyacá: Sierra Nevada del Cocuy, Alto Valle Lagunillas, 2750 m al NNE de la Laguna Pintada, 3920 m, 7 Oct 1972, Cleef & Florschütz 5944 (NY). — Venezuela: Mérida: Sierra de Santo Domingo, Páramo de Mucubají, alrededores de La Laguna Grande, 3560–3600 m, 19 Nov 1959, Barclay & Juajibioy 9593 (US); Sierra Nevada de Santo Domingo, Laguna Negra, 3750 m, Humbert 26414 (NY); E moraine above L[ake]. Mucubají, 3650 m, 17 Jul 1972, Loveless 1675 (DUKE); E moraine above L[ake]. Mucubají, 3650 m, 17 Jul 1972, Loveless 1828 (DUKE); c. 50 km NE of Mérida, S shore of lake Mucubají [70°49′34″S, 08°47′45″W], 19 Oct 2012, Hilpold & al. 1657 (BOZ, UPOS).

Carex ownbeyi G. A. Wheeler in Darwiniana 40: 200. 2002. — Holotype: Bolivia, Cochabamba, Quillacollo Prov., “camino Sipe Sipe-Lipichi”, 3800 m, 9 Apr 1990, Hensen 731 (MIN; isotype: LPB [digital image!]).

Illustrations — Wheeler (2002: 202).

Remarks — This is the first citation of this poorly understood species from the N Andes. Until its recent collection from Tucumán province in Argentina (Jiménez-Mejías & al. 2016a), this species was formerly known only from the type locality in Bolivia (Wheeler 2002).

Additional specimen examined — Colombia: Boyacá: Sierra Nevada del Cocuy, 1 km N of Hacienda Retacuba, 3850 m, 8 Apr 1957, Grubb & al. 266 (US 2322338).

Carex phalaroides Kunth, Enum. Pl. 2: 482. 1837. — Lectotype (designated here): Brazil, Brasilia, Sellow s.n. (LE 00000787 [digital image!]; isolectotypes: F 0BN013398 [digital image!], K 000584673 [digital image!], LE 00000788 [digital image!]).

Illustrations — Barros (1947: Tab. 195); Silveira & Longhi-Wagner (2012: 395 [subsp. moesta], 398 [subsp. phalaroides]).

Remarks — This same collection was reported earlier by Chater (1994) in Flora mesoamericana as Carex planostachys Kunze. This is the first record of the species for Central America, representing the northernmost limit of the species, until now only known from South America (Govaerts & al. 2018+).

The Carex phalaroides group is much in need of revision. The specimen we cite from Guatemala approaches Silveira & Longhi-Wagner's (2012) concept of subsp. moesta (Kunth) Luceño & Alves, with the spikes lanceolate and the stem sides concave.

Additional specimen examined — Guatemala: Huehuetenango, Chiantla, Llano de Tsajuala, 3200 m, 26 Aug 1976, Smith 385 (MICH).

Carex setigluma Reznicek & S. Gonzalez in Contr. Univ. Michigan Herb. 20: 222. 1995. — Holotype: Ecuador, Carchi, near El Pun Páramo, 1 Mar 1953, G. W. Prescott 701 (MICH!; isotypes: CAS 0001340!, NY!).

Illustrations — Reznicek & González (1995: 223).

Remarks — This is the first record of the species from Central America. The Costa Rican specimen displays some achenes constricted on two sides, contrasting with the original description of Carex setigluma achenes as constricted at only one side. Otherwise, the specimen is typical of C. setigluma.

Additional specimen examined — Costa Rica: Limón, Cantón de Talamanca, Sabanas de Dúrika, margen izquierda del Río Kuk, entre los Ríos Kuk y Dipali [09°25′20″N, 83°18′10″W], 2400 m, 18 Oct 1989, Chacón 561 (MO).

Carex tachirensis Steyerm. in Fieldiana, Bot. 28: 68. 1951. — Holotype: Venezuela, Tachirá, Páramo de Tamá, near Colombia-Venezuelan boundary, 3045–3475 m, 15 Jul 1944, Steyermark 57367 (F [digital image!]; isotypes: US!, VEN [digital image!]).

Illustrations — Steyermark (1951: 64); Fig. 1.

Remarks — Carex tachirensis was described by Steyermark (1951) on the basis of small-sized plants bearing inflorescences each consisting of a single terminal androgynous spike. Accordingly, Steyermark assigned C. tachirensis to C. subg. Primocarex Kük. (= C. subg. Psyllophora (Degl.) Peterm.), which is the Carex group that embraces the great majority of unispicate taxa (Kükenthal 1909; Egorova 1999; Ball & Reznicek 2002). He discussed the affinities of C. tachirensis with other Neotropical and North American species of that subgenus and concluded that the relationships of C. tachirensis were unclear. The only material of this species known to date was the type collection.

A study of herbarium materials has revealed the existence of a Carex species from Venezuela and Colombia of uncertain affinities. The relevant material consists mainly of plants with inflorescences each bearing a terminal androgynous spike and one to several lateral pistillate or shortly androgynous lateral spikes. Some stems, however, rarely show a single terminal androgynous spike (e.g. Cleef 6960). These unispicate inflorescences provided a key hint to the identity of these collections, pointing to possible affinities with C. tachirensis. The detailed comparison of the utricles and glumes of these specimens with the utricles from the isotype of C. tachirensis at US (Fig. 1), revealed that the morphology of both materials were similar, with clear affinities in beak shape and venation pattern. Consequently, these materials bearing several-spiked inflorescences were also identified as C. tachirensis.

The classification of these new materials implies a significant widening of the taxonomic understanding of Carex tachirensis. Although the most immediate affinities of this taxon are still unclear, the presence of non-unispicate specimens, sheathing bracts, and tubular cladoprophylls points to C. subg. Carex. The utricles of C. tachirensis showed strong morphological affinities with the Central American C. chiapensis F. J. Herm. (Reznicek, 1986), especially regarding the raised nerves and the bidentate beak. The new materials constitute a substantial broadening of the morphological circumscription of C. tachirensis, with specimens having stems c. 60 cm tall, leaves flattish, up to 3 mm wide, and inflorescences with up to 7 spikes, the lowermost bract 25 cm long, and the longest spikes 4 cm long on peduncles up to 10.5 cm long.

Fig. 1.

Carex tachirensis, representative variation of utricles; A: Steyermark 57367 (isotype US); B: Cleef 6960 (NY); C: Tillett & Hönig 746-738 (NY). — Scale bar = 1 cm.

Although certainly unusual, a few other species of Carex may sometimes bear one or several spikes in their inflorescences, such as C. exilis Dewey (Reznicek & Ball, 1980), C. malmei Kalela and C. monodynama (Griseb.) G. A. Wheeler (Wheeler, 1990; Gebauer & al., 2015), or C. subantarctica Speg. (Barros, 1969; Wheeler, 1987, 2009). Such variation of the inflorescence might be due, at least in part, to different growing conditions. Indeed, an overall reduction of the morphology is also known from high mountain populations of the C. flava group when compared with specimens inhabiting habitats with milder climatic conditions (Jiménez-Mejías & al. 2012, 2014, 2017). The majority of the labels in the studied C. tachirensis specimens refer to rocky outcrops and other poorly developed soils. Accordingly, the different degree of development of the different plants might be due to differences by growing on deeper versus poor shallow soils.

The specimen records provided here constitutes a considerable expansion of the range of Carex tachirensis, both north and south along the northernmost branch of the Andes. These localities also constitute the first citations of the species for Colombia, which, however, was expected since the species was described from Venezuela but “near the Colombia-Venezuela boundary” (Steyermark 1951).

Additional specimens examined — Colombia: Boyacá: Páramo de Pisva, Carretera Socha-La Punta km 78, c. 1 km al NE de la Laguna Colorada, 3500 m, 17 Jun 1972, Cleef 4660 (NY); Páramo de Pisva, flanco SW de los Morros de S. Gabriel, 2 km al SW de la Laguna Batanera, 3670 m, Cleef 4702, 18 May 1972 (MO, NY); Páramo de La Rusia, NW-N de Duitama, Serranía Peña Negra, Hoya de la Laguna Agua Clara, lajas de arenisca, 800 m al S de esa laguna, 3935 m, 10 Dec 1972, Cleef 6990 (NY); Páramo de La Rusia, NW-N de Duitama, Alto de Avendaño, 2 km al SE de la Laguna Negra, 3870 m, 10 Dec 1972, Cleef 6960 (NY); Páramo de la Sarna entre Sogamoso y Vado Hondo, 5 km al NE de la Laguna de Tota, 3550 m, 29 Mar 1973, Cleef 9204 (NY). Páramos al NW de Belén, cabeceras Quebrada Minas, 600 m al N de la Laguna El Alcohol, 3915 m, 27 Feb 1972, Cleef 1920 (NY). — Cundinamarca: Páramo de Sumapaz, Chisacá, cerca de la cárcel destruida, 3630 m, 12 Jul 1972, Cleef 4933 (NY). — Venezuela: Merida: Municipio Arzobispo Chacón, Páramos de San José y Piedra Pirela, vía Laguna Pozo Negro, 8–10 km al S de Acequias [08°21′N, 71°14′W], 3000–3460 m, 12 Apr 2013, Gonto & al. 5254 (NY); Municipio Rangel, Páramo de Motumbo, SW of Laguna Larga (Trujillo State) along border of divide to Laguna Las Parias, Monumento Natural Teta de Niquitao, Guirigay, 3200–3500 m, 17 Sep 2003, Stergios & al. 20526 (MICH). — Lara: Dpto Moran, trail from Humocaro to Buenos Aires (Caserío) below Páramo Las Rosas [09°40′N, 70°05′W], 3300 m, 25 Jun 1979, Liesner & al. (MO). — Zulia: Distrito Perijá, Sierra de los Motilones, environs of “Campamento Frontera II”, mesa below international boundary on main ridge, headwaters of Río Negro [10°00′13″N, 72°58′c.25″W], c. 3000 m, 27 Jun – 5 Jul 1974, Tillett & Honig 746-738 (NY).

Synonymization of Carex tucumanensis under C. niederleiniana

Carex niederleiniana Boeckeler, Beitr. Cyper. 1: 50. 1888 ≡ Carex acutata var. hirtisquama Kük. in Bot. Jahrb. Syst. 27: 549. 1899 ≡ Carex acutata var. niederleiniana (Boeckeler) Kük. in Engler, Pflanzern. IV. 20 (Heft 38): 702. 1909, nom. superil. — Lectotype (designated here): Argentina, La Rioja, Sierra Famatina, Quebrada del Vallecito, 21 Jan 1879, Hieronymus & Niederlein 604 (CORD 00002100 [digital image!]; isolectotypes: CORD 00002101 [digital image!], G 00098281 [digital image!], K 000584758 [digital image!], SI 055592!). — Syntype: Argentina, La Rioja, Sierra Famatina, en las cercanías de La Cuesta, más arriba del Vallecito, 15–20 Jan 1879, Hieronymus & Niederlein 657 (SI 000225!).

= Carex tucumanensis G. A. Wheeler in Hickenia 2: 190. 1996. — Holotype: Argentina, Tucumán, Dpto. Chicligasta, Estancia Las Pavas, 2500 m, en la barranca del Río de la Cascada, 15 Mar 1924, Venturi 3193 (LIL 000013 [digital image!]; isotype: SI 000238!).

Illustrations — Wheeler (2009: 336).

Remarks — Carex niederleiniana is a poorly understood species endemic to Argentina. It was described from the pre-Andean range known as Sierra Famatina, in the province of La Rioja in N Argentina. It is further known from the northern provinces of Catamarca and Tucumán, as well as from Río Negro (Wheeler 2009; Flora del Cono Sur 2017). Carex niederleiniana is morphologically closely related to C. acutata Boott (Boeckeler, 1888), which is distributed along the Pacific side of the Andes from Chile to Ecuador (Flora del Cono Sur 2017; Govaerts & al. 2018+). Carex niederleiniana was described as having the terminal spike androgynous (Boeckeler 1888). Later, the terminal spike was regarded to be either gynaecandrous (Wheeler 1996) or “con flores estaminadas entremezcladas entre numerosas flores pistiladas arriba y unas pocas debajo” [with staminate flowers intermingled among numerous pistillate flowers above and a few below] (Wheeler 2009). On the other hand, C. tucumanensis was described by Wheeler (1996) from the also pre-Andean range of Sierra del Aconquija, in the province of Tucumán. He considered that the main differences between C. tucumanensis and C. niederleiniana were the terminal spike, which was entirely staminate in C. tucumanensis (but occasionally with pistillate flowers at the middle), and the size of the achenes, which were slightly smaller (1.5-1.7 × 0.6-0.9 mm in C. tucumanensis vs. 1.6-1.8 × 0.9-1.2 mm in C. niederleiniana). Since then, no additional studies have compared the two taxa. Both, C. niederleiniana and C. tucumanensis were ascribed to C. sect. Pseudocypereae Tuck, ex Kük., today mostly considered part of a widened C. sect. Vesicariae (Heuff.) J. Carey (Ball & Reznicek 2002; Global Carex Group 2016).

The case of the description of Carex tucumanensis as a species different from C. niederleiniana seems to be a classical example of a misunderstood morphological variation. Flower sex distribution within the inflorescences of Carex is a key taxonomic character in a majority of species groups, distinguishing some species from their closest counterparts, and even characterizing entire taxonomic groups (i.e. sections, e.g. Egorova 1999; Ball & Reznicek 2002). However, C. sect. Vesicariae sensu latissimo is one of the groups where sex distribution within spikes can be rather variable, especially as regards the terminal spike, such that within the same taxon (e.g. C. pseudocyperus L., species of former C. sect. Lupulinae Tuck, ex J. Carey) the terminal spike can vary from entirely staminate, to androgynous or gynaecandrous (Ball & Reznicek 2002; Luceño & al. 2008; S. Gebauer, pers. obs.). It seems that Wheeler was misled by deviant specimens of C. niederleiniana bearing a terminal staminate spike, a feature rarely observed in this species. Given the very limited presence of C. niederleiniana in herbaria, he probably overemphasized the importance of the terminal spike as a taxonomic character, and considered that these plants deserved formal recognition. As explained above, the distribution of sex in the terminal spike of C. niederleiniana itself is indeed extremely variable (see Boeckeler 1888; Barros 1947, 1969; Wheeler 1996, 2009). In fact, the lectotype collection displays stems either with a terminal spike that is androgynous (G), gynaecandrous (CORD), or “androgynaecandrous” (i.e. with staminate flowers above and below, and staminate flowers at the middle; CORD, G, SI). Moreover, Wheeler himself cited as C. niederleiniana a collection from a place just 400 m lower than the type locality of C. tucumanensis (Venturi 4648; Wheeler 1996). While the specimen Wheeler studied from this collection (US) displayed a terminal gynaecandrous spike, and thus he classified it as C. niederleiniana, other duplicates that we have studied from the same collection indistinctly bear terminal spikes that are entirely staminate (A, SI), gynaecandrous (A), or “androgynaecandrous” (A, MO).

Accordingly, we conclude that Carex tucumanensis must be considered part of the variation within C. niederleiniana, with the terminal spike entirely staminate, and the achene size within the smaller threshold of the species.

Additional specimens examined — Argentina: Tucumán: Chicligasta Department, Estancia las Pavas, 2100m, 23 Nov 1926, Venturi 4648 (A, MO, SI [2 sheets], US); Chicligasta Department, Estancia las Pavas-Puesto El Bayo, 3000 m, 11 Mar 1924, Venturi 3060 (US).