This species differs from all known species of Clinanthus by the uniquely inflated perigone, the shape and coloration of which is reminiscent of the genus Urceolina Rchb. (Eucharideae) or superficially certain species of Andean vaccinioid Ericaceae.

PERU. Ancash: Prov. Recuay, Dist. Cotaparaco, Sector Santa Cruz, 2450 m, 6 Mar 2007. M. Morales & E. Jara 767 (holotype: USM!).

Small terrestrial geophytic herb (Fig. 1A); bulb small, globose-ellipsoidal, ca. 2 cm diam., prolonged into neck 7.5–8 × 60–70 mm. Leaves 2–3, sessile, narrowly lorate, 4–6 × 90–110 mm, slightly canaliculate, obtuse at apex, with conspicuous midrib, glabrous, green, synanthous. Inflorescence pseudo-umbellate, borne at apex of naked scape; scape cylindrical, solid, 150–200 mm long (as observed), terminated by 2 spathe bracts, bracts green, ovate lanceolate. Flowers 2–3, pendulous, 35–40 mm long; pedicels slender, 15–20 mm long at anthesis; perigone consisting of 6 tepals in 2 whorls, fused into a tube that is narrowly cylindrical in proximal 15–17 mm and arcuate proximally, abruptly dilated distally to ca. 8 mm at throat; limb urceolate, abruptly inflated to ca. 33 mm in diam., then slightly constricted in distal 10 mm; tepals 6, bright orange-red proximally, concrescent, green in distal 10 mm with yellowish green margins, glabrous; outer tepals elliptical, 5–6 × 13–14 mm, apiculate; inner tepals elliptical, 5.5–6.5 × 12–13 mm, minutely apiculate. Stamens 6, basally connate into cylindrical staminal cup or corona, 7.4–7.6 × 11.0–11.5 mm, reaching to ca. 2 mm from the apex of the limb, salmon proximally, prominently 6-lobed, the lobes white distally and coarsely dentate along their edge, ca. 2 × 3 mm; free filaments short, inserted at sinus between the lobes; anthers grouped in the center of the flower (but not connivent), ca. 3 mm long, linear, dorsifixed, versatile, longitudinally dehiscent; pollen yellow. Style ca. 45 mm long, exerted ca. 5 mm past apex of the limb, white; stigma capitate, papillate, white, 2–2.3 mm wide. Ovary inferior, green, oblong, ca. 5 × 10 mm, 3-loculed, placentation axile, ovules oblong, flattened, ca. 20 per locule, superposed. Capsule and seed not seen.

Figure 1. 

A–C Clinanthus inflatus. Photos by anonymous source, used with permission. A Inflorescence in habitat B bouquet of cut inflorescences C habit in nature D–F Ismene parviflora, photos by Asunción Cano D–E flowers F habit in cultivation.

Figure 2. 

Clinanthus inflatus. A Habit B whole flower C whole flower cut open to show staminal corona and ovule number. Drawing by Klei Sousa.

Clinanthus inflatus is known only from the type locality (Fig. 3), in seasonally dry vegetation.

Figure 3. 

Map of northern Peru showing type localities of Clinanthus inflatus (white circle) and Ismene parviflora (black square). Inset: map of South America with Peru filled in black.

Plants were collected in flower in March.

The specific epithet is from Latin and refers to the abruptly inflated perigone.

Since nothing is known of the distribution of this species beyond the type locality, it is best to place it in the category Data Deficient (IUCN 2012, 2017). The type specimen label indicates that it was abundant at the collection site.

The urceolate perigone of C. inflatus is yet another example of the convergent evolution that characterizes the tetraploid Andean lineages (Meerow 2010). It seems to have affinity with the subclade of Clinanthus (Meerow et al. 2000) that includes C. campodensis (Ravenna) Meerow, C. humilis, C. recurvatus (Ruiz & Pav.) Meerow, and C. ruber (Herb.) Meerow & A. Cano, all with leaves < 1 cm wide, and relatively small flowers.

Pancratium rubrum (Herb.) Pav. ex Steud. was not included in Ruiz and Pavon (1802), even though a specimen from their expedition (FI!, FI011974) is labelled as the type of S. rubrum Herb. The name was validated by von Steudel (1841). Ravenna (1978) believed that the type specimen represented Stenomesson flavum (Ruiz & Pav.) Herb., a species that was originally illustrated and described in Ruiz and Pavon (1802) as Pancratium flavum Ruiz & Pav. We believe that Ravenna (1978) was mistaken in this regard. The specimen has a single scape that bears only three flowers, and the stamens are barely if at all exerted from the perigone. Stenomesson flavum always has > 3 flowers and the stamens are exerted at least 1 cm from the limb.

Herbert (1837) illustrated S. rubrum (Fig. 4A), which he seemed to consider distinct from Stenomesson coccineum (Ruiz. & Pav.) Herb. [≡ Clinanthus coccineus (Ruiz. & Pav.) Meerow], a decision with which we agree. Herbert (1837) also surmised that errors were made regarding Ruiz and Pavon types and plates in regard to both species. Four years later, Herbert (1841) published Coburgia coccinea Herb. (Fig. 4B). He made no reference to either S. rubrum or S. coccineum, though the flower in the plate (Fig. 4B) looks very similar to the flower illustrated for S. rubrum (Fig. 4A). The well-illustrated leaf morphology in the plate undoubtedly places the plant in Clinanthus. One year later, Herbert (1842) assigned both Pancratium coccineum Ruiz [sic] and Stenomesson coccineum to synonomy with Coburgia coccinea. We thus regard Coburgia coccinea Herb. (1841) as distinct from Coburgia coccinea (Ruiz & Pav.) Herb. (1842), nom. illeg., but conspecific to Clinanthus ruber, and include it in the synonymy of the latter taxon here.

Figure 4. 

A Plate 28 from Herbert (1837) with Stenomesson rubrum Herb. illustrated (number 6, red arrow). Stenomesson breviflorum Herb. is illustrated in the same plate (numbers 7 & 8, black arrows), evidence that Herbert considered these two different species B Painting of Coburgia coccinea Herb. [= Clinanthus ruber (Herb.) Meerow & A. Cano] that accompanied the description by Herbert (1841) C Clinanthus ruber, photographed in Peru. Photo by Oscar de Viveroscar D Clinanthus coccineus (Ruiz & Pav.) Meerow, loma form, photographed in Lima Dept., Peru. Photo by Norton Cuba Melly.

Ravenna (1978) diagnosed Clinanthus coccineus (Ruiz & Pav.) Meerow as belonging to his Stenomesson subgen. Fulgituba Ravenna (1974), along with Clinanthus incarnatus (Kunth) Meerow, C. microstephius (Ravenna) Meerow, C. variegatus (Ruiz & Pav.) Meerow and other species, all of which have varying apical zones of green on the tepals. Recent collections in Peru (N. C. Melly, S. Leiva, pers. comm.) confirm this, with Clinanthus coccineus (Fig. 4D) found both in coastal lomas and in the northwestern Andean slopes. Ravenna (1978) seems to have considered S. rubrum a synonym for S. breviflorum Herb. We disagree with this assessment. Herbert (1837) illustrated both in the same plate (Fig. 4A). The type specimen of S. breviflorum (Maclean s.n. K!, K000322395 & 396) clearly bears pseudo-petiolate leaves, which confirms its place in Stenomesson rather than Clinanthus. Clinanthus ruber is fairly widespread in northern Peru, and can have red or pink flowers, but never more than five in all the material we have seen. It is variable in the degree to which the limb spreads from the throat of the perigone. The flowers are without any green apical zone, and bear relatively narrow leaves < 1 cm wide.

Ismene parviflora differs from all other species of the genus by the small flowers (< 4 cm long), that are loosely funnelform, terminating with a ventricose limb, but not infundibular to the degree of I. vargasii and I. morrisonii (subgen. Pseudostenomesson Meerow), the latter species to which it approaches in size. All other species of Ismene have large, crateriform flowers (subgen. Ismene) or large, zygomorphic flowers (subgen. Elisena).

Perú, Ancash: Prov. Recuay, Dist. Cotaparaco, Sector Santa Cruz,, 2332 m, rocky slope with shrubs and silty soil, 6 Mar 2007. M. Morales & E. Jara 767 (holotype: USM!).

Terrestrial geophytic plant from subterranean bulb. Bulb not seen. Leaves 2, oblanceolate-lorate, ca. 15 × 150–170 mm, bright green, glabrous, slightly canaliculate, obtuse, tightly sheathing proximally and forming a green aerial pseudostem 40–60 mm long, synanthous. Inflorescence pseudo-umbellate, borne on a naked scape; scape ancipitous, solid, 200–250 × ca. 18 mm, terminated by 2 ovate-lanceolate green bracts that persist at anthesis, over-topping flowers by several cm, bracts 50–55 × 7.5–9.6 mm, acute. Flowers 3–5, mostly perpendicular to scape, loosely funnelform-tubular, shortly pedicellate, pedicels 2–3 mm long; perigone 35–37 mm long, consisting of 6 tepals in two whorls, fused proximally into a tube that is cylindrical for proximal 20 mm, ca. 1.5 mm wide, then dilating to 3.5 mm at throat; tube curved proximally; limb slightly ventricose, apically ca. 10 mm wide, tepals loosely concrescent, not spreading significantly; outer tepals elliptical, concave, mostly green abaxially, white towards apex, adaxially white with green veins in distal ½–2/3, green below, 13.5–14.0 × 4.5–5.0 mm, apiculate; inner tepals ovate-elliptical, less concave than outer, 13.0–13.5 × 5.0–5.2 mm, minutely apiculate, colored like outer. Stamens 6, fused into green staminal cup 6.5–7.5 × 3.5–3.7 mm, cylindrical in proximal 6 mm, abruptly ampliate in its distal 1 mm to ca. 4 mm, coarsely and unevenly dentate at rim; free filaments inserted ca 2 mm below rim of cup, filiform, white, incurved, ca. 1 mm long; anthers oblong, closely appressed to each other but not connivant, 3.5–4.5 mm long, pollen yellow. Style exerted 10–15 mm past the limb apex, 40–45 mm long, white, filiform; stigma capitate, papillose, white, 2–3 mm wide. Ovary globose, ovules 1 or 2 per locule, basal. Capsule and seed not seen.

Figure 5. 

Ismene parviflora. Drawing by Klei Sousa. A Habit B whole flower C whole flower cut open to show staminal corona and ovule number.

Ismene parviflora is known only from the type locality (Fig. 3), in seasonally dry vegetation.

Plants were collected in flower in March.

The specific epithet is from Latin and refers to the small size of the flowers.

Since nothing is known of the distribution of this species beyond the type locality, it is best to place it in the category Data Deficient (IUCN 2012, 2017). The type specimen label says it was abundant at the collection site.

Ismene parviflora has greatest affinity morphologically to the two members of Ismene subgen. Pseudostenomesson (Velarde) Meerow, I. vargasii (Velarde) Gereau & Meerow and I. morrisonii (Vargas) Velarde, but the unique morphology of the new species make it difficult to assign I. parviflora to I. subgen. Pseudostenomesson with confidence at this time. Ravenna (1988) disagreed with the transfer by Velarde (1949) of Stenomesson morrisonii Vargas to Pseudostenomesson Velarde, but based on the same information (Vargas 1943), we believe it belongs with I. vargasii in I. subgen. Pseudostenomesson. In particular, the photo in Vargas (1943) shows the typical foliar morphology of Ismene, i.e. the aerial pseudostem formed by the tightly sheathing leaf bases. The two species of subgen. Pseudostenomesson are found in the Departments of Junin (I. vargasii) and Apurimac (I. morrisonii), and, unlike I. parviflora, have larger, fully infundibular, and completely green perigones.