Tillandsia myosura Grisebach ex Baker (1878: 240)

Rossado, Andrés J., Donadío, Sabina & Bonifacino, José M., 2024, A taxonomic revision of Tillandsia (Tillandsioideae, Bromeliaceae) from Uruguay, Phytotaxa 667 (1), pp. 1-90 : 1-90

10. Tillandsia myosura Grisebach ex Baker (1878: 240) ; Grisebach (1879: 333), isonym. Figs. 12D , 30 , and 31.

Type (lectotype designated by Smith 1935: 200 [as “type”]):— ARGENTINA. Córdoba: “Estancia Germania prope Córdoba ”, June– December 1874, Lorentz 122 ( BM [ BM000923949 , online image!]; isolectotypes: F0BN011516 [online image!], G00098172 [online image!], LE00006299 [online image!], M0111448 [online image!], NY00247359 [online image!], P00753256 [online image!], WRchb. 1889-0003279 [online image!], WU-Keck 0001377 [online image!]) .

Plants caulescent, few to much branched from the base and then forming sublax clumps up to 20 cm in diameter. Roots 0.4–0.7 mm diameter, only present at the base of the stem. Stems up to 12 cm long. Leaves distichously arranged, distributed along the stem, gray to grayish-green; sheaths 12–20 × 11–16 mm, distinct from the blade, widely elliptic to widely elliptic-ovate, glabrous in the quarter or basal half, lepidote towards the apex; blades (40–)50–110 × 3–5 mm, spreading to reflexed, very narrowly triangular, channelled from the base to almost the apex, fleshy, densely tomentose-lepidote throughout, trichomes asymmetric. Inflorescences simple; peduncles bractless (sometimes with one basal foliaceous bract), 5–15 cm long, 0.9–1.3 mm in diameter, densely lepidote. Spike distichously and densely (1)2 to 4-flowered, 10–35 mm long, very narrowly elliptic in outline; rachis partially exposed, grayish-green, densely lepidote. Floral bracts densely distributed (2–3 times longer than the internodes), imbricate, clasping the flower, ovate to elliptic-ovate, ecarinate, grayish to grayish-green, densely lepidote; basal floral bracts 11–24 × 7.5–10.5 mm, surpassing or seldom equaling the sepals, bladeless or sometimes with a very short blade (less than 5 mm long and not clearly differentiated from the sheath); upper floral bracts shorter, bladeless. Flowers cleistogamous; sepals wholly hidden by the floral bract to apically exposed, 9–14 × 3–4.2 mm, unequally connate, narrowly elliptic-ovate to narrowly elliptic-oblong, or narrowly oblong-ovate, yellowish-green with vinaceous tinges, and stramineous towards the apex, lepidote, trichomes over the middle nerve and towards the apex; abaxial sepal ecarinate, connate with the adaxial sepals for 0.8–1.1 mm; adaxial sepals somewhat carinate at the base and middle part, connate to each other for 1.3–3.9(–5.6) mm; petals 10–15(–18) × 2.9–5.7 mm, lingulate, yellowish (whitish towards the base), margins entire; stamens 7.5–10 mm long, deeply included, a little longer than the pistil; filaments 5–8.5 mm long, straight (not plicate); pollen yellow; pistil 5–7 mm long, deeply included; ovary 2.6–4.5 × 1.5–2 mm, cylindrical-obovoid, clearly differentiated from the style; style 1.5–2.4 mm long, shorter than the ovary, whitish; stigmas simple-truncate. Capsules (27–)31–41 × 2–2.8 mm, much exceeding its respective floral bract, cylindrical-prismatic, apex rounded and short-beaked.

Vernacular names:— Clavel del aire blanco amarillento ( Brito & Llano 2008).

Classification:— Tillandsia myosura belongs to T. subg. Diaphoranthema , based on its morphological characteristics ( Smith & Downs 1977), and the morphological and molecular phylogenies by Donadío et al. (2015), Granados Mendoza et al. (2017), and Donadío et al. (2022).

Distribution and habitat:— Bolivia, Argentina, and Uruguay. In Argentina, Tillandsia myosura is widely distributed from the northwest to the central region of the country. Additionally, there are some remote populations in central-eastern Argentina (eastern Entre Ríos and northeastern Buenos Aires provinces).

In Uruguay, Tillandsia myosura occurs in the south, southwest, and west-central regions ( Fig. 31 ). The distribution of this species in Uruguay is by no means continuous within its range, commonly with populations often widely spaced. In this sense, it should be noted that in some localities T. myosura consists of large populations (e.g. Laguna de los Cuervos, Lavalleja), while in others, it is a rare species with only a few individuals present (e.g. Pan de Azúcar Hill, Maldonado).

Tillandsia myosura inhabits open thorn forests, xeric hillside forests, psammophilous forests, and sparse riverside forests or if dense, then in forest margins or openings. Tillandsia myosura grows exclusively as epiphyte, mainly on small branches from fully sun-exposed to part-shaded conditions, together with T. aëranthos and T. recurvata , and occasionally with T. bandensis .

Reproductive observations: —Plants from different populations of Tillandsia myosura from Uruguay were kept in cultivation for several years. Although they produced fruits with viable seeds, in none of them open flowers were observed. Additionally, no herbarium specimens or plants in their natural habitat with open flowers were registered. Based on these observations, we consider the populations of T. myosura from Uruguay (and possibly those in Entre Ríos and Buenos Aires provinces in Argentina) to be exclusively cleistogamous.

Phenology: —In Tillandsia myosura the onset of inflorescence emergence is evidenced by the bulging of the apex of the stem, which takes place in early autumn (March–April; Fig. 4 ). The inflorescence continues to grow for seven to eight months until the flower buds reach maximum development (pre-anthesis stage). This stage is challenging to determine since, generally, all the floral whorls are almost completely clasped and hidden by the floral bract. Nevertheless, the apex of the petals occasionally becomes visible at this stage, remaining strongly imbricated among each other ( Fig. 30F ). Fruits begin to develop as of late September and October, continuing to grow in length until December. Afterward, they remain without apparent growth for seven to nine months until seed dispersal, which occurs in winter (mainly August to September).

In each plant of T. myosura , a period of approximately one and a half years elapses between the onset of the inflorescence emergence and the dehiscence of the capsules, which makes of this, together with that of T. duratii , one of the longest reproductive cycles for Tillandsia in Uruguay.

Conservation:— We consider Tillandsia myosura as Not Threatened in Uruguay. This species has a scattered distribution within the country but it is found in several localities. Furthermore, the populations observed were generally well-preserved, despite none of them being located in national protected areas.

Diagnostic characters:— In Uruguay, Tillandsia myosura is highly distinctive, not having any morphologically similar species. It can be differentiated from the other species of Tillandsia occurring in Uruguay by the combination of distichous leaves and blades wider than 3 mm (alive). In addition, it is the only species in the country with ebracteate peduncles that are densely lepidote throughout their length. It is important to note that these characters are valid only within Uruguay, as the presence of peduncle bracts and the pubescence of the peduncles exhibit considerable variation throughout its entire geographic range.

Background in Uruguay:— Tillandsia myosura was first cited for the Uruguayan flora by André in 1905, although without specifying reference material. It was Smith, in 1933, the first to cite a herbarium specimen of T. myosura for Uruguay (under the name T. crocata ). The oldest gathering for this species in this country dates back to 1893 ( Osten 3055, MVM) from Soriano department.

Note I:— Tillandsia myosura shows noticeable morphological variation throughout its geographic range in several characters, such as the presence of trichomes in the leaf sheath, type of indument on the leaf blades, peduncle indument, presence of peduncle bracts, flower number and density in the spike, and fusion and indument of the sepals. The species seems to occur in two disjunct areas separated by more than 500 km, one in Bolivia and western and northwestern Argentina, and the other in eastern Argentina and neighboring Uruguay.

In Uruguay and eastern Argentina, the entity that we are here recognizing as Tillandsia myosura has traditionally been treated by several authors ( Smith 1933, 1972 p.p.; Herter 1949 –1956; Marchesi 1965; Smith & Downs 1977 p.p.; Zuloaga et al. 2019 p.p.) as T. crocata (É. Morren) Baker (1887: 241) . Tillandsia myosura is easily distinguished from the latter by the presence of inconspicuous flowers (vs. showy). Nevertheless, these two species are somewhat vegetatively and reproductively similar when not in anthesis, and therefore they can be confused with each other.

Furthermore, in the key presented by Smith & Downs (1977), T. myosura is set apart from T. crocata solely by the nature of the leaf blade indument [trichomes with lobes (if any) appressed or slightly divergent in T. myosura vs. trichomes spreading, thereby making the leaf tomentose in T. crocata ]. This detail is significant, as plants belonging to T. myosura from Uruguay and eastern Argentina (Buenos Aires and Entre Ríos provinces) show tomentose-lepidote leaf blades. This type of indument is more similar to that found in T. crocata than in most populations of T. myosura .

These facts, coupled with the exclusively cleistogamous nature of Uruguayan populations of T. myosura (no visible flowers), might explain the misidentification of many T. myosura specimens from Uruguay and western Argentina as T. crocata .

Note II:— In the protologue of Tillandsia myosura , two gatherings, Lorentz 122 and Mandon 1180, are cited after the description of the species, and therefore are considered original material, both with several duplicates in different herbaria. Smith (1935) considered Lorentz 122 at BM as “type” for T. myosura [cited as “ holotype ” in Smith & Downs (1977) and later works based on this], thus making an inadvertent lectotypification ( Prado et al. 2015; ICN Art. 9.10, Turland et al. 2018).

It should be noted that Mez (1896) described the new species T. mandonii É. Morren ex Mez (1896: 871) based on Mandon 1180. Therefore, this gathering is no longer considered as type material for T. myosura .

Note III:— Tillandsia myosura was validly described as a new species by Baker (1878), using a name written on the labels of the specimens Lorentz 122 by Grisebach, something clearly stated in the protologue: “ T. myosura, Griseb. , in Lorentz Pl. Argent. Exsic. No. 122”. One year later, Grisebach (1879) published T. myosura as a new species with its own description and citing two plant materials: “O.: Tarija, in arboribus pr. S. Augustin ( forma pedunculis villosus)” and “C.: pr. Córdoba ( forma glabrescens )”. The first material corresponds unequivocally to the gathering Lorentz & Hieronymus 969, which later was cited as a paratype for Tillandsia caliginosa W. Till in Till & Hromandnik (1984: 282). The second material has been ascribed to Lorentz 122 ( Lorentz & Niederlein 1881, Smith 1935, Smith & Downs 1977, Till & Hromandnik 1984, and annotations in specimens) and to Hieronymus 347 (only annotations in specimens at CORD and GOET).

According to Hunziker (1960), the gatherings cited by Grisebach (1879) from Córdoba are based on collections of Lorentz, Hieronymus, or Galander, between November 1872 and December 1877. In this sense, Lorentz 122 is the only gathering identified as T. myosura that agrees with Hunziker (1960) and that has written on its label “prope Córdoba ” as it is stated in Grisebach’s publication (as “pr. Córdoba ”; the specimen Hieronymus 347 has “Sierra de Córdoba ” written on its label). Additionally, on the labels of Lorentz 122 there is an indication noting that the specimen has been determined by Grisebach (“Determ. Prof. Grisebach”). Thus, based on all this evidence, we only consider Lorentz 122 (and not Hieronymus 347) as the gathering cited by Grisebach (1879) as “C.: pr. Córdoba ( forma glabrescens )”. Therefore, T. myosura Grisebach ex Baker and T. myosura Grisebach , both based on the same type (Lorentz 122), should be considered isonyms. Only the earlier of the two names, i.e., T. myosura Grisebach ex Baker (1878) , has nomenclatural status (ICN Art. 6.3 note 2, Turland et al. 2018).

Representative specimens examined:— URUGUAY. Colonia: Barra Arroyo Rosario , 25 October 1964, Marchesi 1248 ( MVFA) . Lavalleja: Cerro Arequita , 20 October 1962, Rosengurtt & Del Puerto s.n. ( MVFA 8373 ) . Maldonado: Cerro Pan de Azúcar , 12 August 2013, Rossado 177 ( MVFA) . Rio Negro: Arroyo Negro, aprox. 1, 5 km al Noroeste de Paso de Las Cadenas , 27 December 2017, Rossado et al. 502 ( MVJB) . Salto: Unos 3.5 km al SE de Constitución, arroyo Ceibal Grande , 2 December 2014 (cultivated in Canelones, pressed 15 October 2017), Berazategui et al. s.n. ( MVJB 30514 ) . San José: Arazatí , 16 October 1965, Del Puerto s.n. ( MVFA 5328 ) . Soriano: Rincón del Cololó , 22 July 1893, Osten . 3055 ( MVM) .

Figures

FIGURE 4. Phenology of Tillandsia species occurring in Uruguay. Divisions within each month indicate intervals of approximately 10 days, identified as the beginning, middle, and end of each month. Tillandsia loliacea is excluded due to lack of enough information. * Denotes species with cleistogamous flowers. ** Denotes species with insufficient data to ascertain the timing of fruiting or fruit dehiscence.

FIGURE 12. Illustrations of floral parts. A. Tillandsia bandensis subsp. bandensis. B. T. bandensis subsp. grandipetala. C. T. duratii. D. T. myosura. From left to right: floral bract (stretched view), sepals (stretched view; adaxial sepals to the right), petal (stretched view), and pistil. A based on Rossado et al. 369 (MVFA); B based on Rossado et al. 408 (MVFA; floral bract and petal) and Rossado et al. 407 (MVJB; sepals and pistil); C based on Rossado et al. 365 (MVFA); D based on Mai 76 (MVJB) and on a cultivated plant from Arequita Hill (petal and pistil).

FIGURE 30. Tillandsia myosura. A. Fruiting plants in an open thorn forest near Paso de la Cadena (Río Negro, Uruguay). B. Habit (Laguna de los Cuervos, Lavalleja, Uruguay). C. Habit (Sierra de Mahoma, San José, Uruguay) D. Individuals bearing well-developed inflorescences (Cañada de la Tapera, San José, Uruguay). E. Close up of the fertile part of the inflorescence. F. Lateral view of two flowers and their floral bracts; on the left, flower with withered petals; on the right, apical flower with petals turgid. In this apical flower, pollination may be occurring (cleistogamous flowers). G. Infructescence. Photos by Andrés Rossado.

FIGURE 31. Distribution map of Tillandsia myosura in Uruguay and neighboring areas.