A new saxicolous species of Amaryllidaceae tentatively assigned to the tribe Clinantheae, Pamianthe ecollis Silverst., Meerow & Sánchez-Taborda, is described from the western slope of the Cordillera Occidental in the department of Cauca, Colombia. The new species differs from the two hitherto known species of Pamianthe in its yellow flowers and in its nearly obsolete perianth tube. The near loss of the perianth tube may be correlated with a change in pollinator. The new species lacks a bulb; it produces a large number of winged seeds that are wind-dispersed. A key to the species of Pamianthe is provided. This is the first record of the genus Pamianthe for Colombia. The phylogenetic position of the genus Pamianthe is discussed.

Se describe una nueva especie de Amaryllidaceae tentativamente perteneciente a la tribu Clinantheae, Pamianthe ecollis Silverst., Meerow & Sánchez-Taborda, procedente de la vertiente occidental de la cordillera Occidental en el departamento del Cauca, Colombia. La nueva especie difiere de las dos especies conocidas de Pamianthe por su perianto amarillo que tiene un tubo casi ausente. La reducción del tubo del perianto probablemente tiene correlación con un cambio en el polinizador. La nueva especie carece de un bulbo, y produce numerosas semillas aladas que se dispersan por el viento. Se provee una clave a las especies de Pamianthe. Este es el primer registro del género Pamianthe para Colombia. Se discute la posición filogenética del género Pamianthe.

Amaryllidaceae , biodiversity, Cauca, Clinantheae , Colombia, Pamianthe, Andes

Amaryllidaceae J. St.-Hil. is a cosmopolitan family represented in Colombia by nine native genera and 26 native species, including a monotypic endemic genus, Plagiolirion Baker (Meerow and Silverstone-Sopkin 1995). Some of the Colombian species have restricted ranges and are in danger of extinction or may already be extinct (Silverstone-Sopkin 2011). Recent field work in the Cordillera Occidental of the Andes, in the department of Cauca, has resulted in the discovery of a new species of Amaryllidaceae that also seems to be narrowly distributed. Vegetative and floral morphology and nrDNA ITS sequences indicate that this species represents a novelty in the genus Pamianthe Stapf.

Stapf (1933a, 1933b) published the genus Pamianthe in honor of Major Albert Pam, who cultivated bulbs in England that he received from Peru in 1928. There are five published species names that have been assigned to this genus: P. andreana (Baker) Stapf, P. cardenasii Traub, P. parviflora Meerow, P. peruviana Stapf, and P. quitoensis (Herb.) Stapf. Pamianthe quitoensis was transferred to the genus Leptochiton Sealy, as L. quitoensis (Herb.) Sealy, and P. andreana is considered a synonym of this species. Pamianthe cardenasii has been placed in the synonymy of P. peruviana (Meerow 1984). Thus, the genus Pamianthe, as previously recognized, includes only two species, P. parviflora, known only from Ecuador (Meerow 1984), and P. peruviana (the type species), known from Perú and Bolivia. The new species described in this paper is the third species of the genus and the first record from Colombia. It is also the first species of the tribe Clinantheae, to which Pamianthe has been assigned (Meerow et al. 2000; Leiva and Meerow 2016), discovered north of Ecuador.

Photographs of the flower in alcohol and seeds of Pamianthe ecollis were taken with a Nikon model DS-Ri1U3 digital camera, using a Nikon model SMZ-1500 stereo dissecting microscope at the Laboratorio de Imágenes del Postgrado en Ciencias-Biología de la Universidad del Valle; floral and seed measurements were made with NIS Elements Br, version 4.20 software.

DNA extraction, amplification and sequencing protocols were as described in Meerow et al. (2000, 2006). The ITS sequence of P. ecollis was aligned with a previous ITS alignment of the tribe Clinantheae (Meerow et al. 2000; Meerow 2010) using the program MAFFT (Katoh and Standley 2013). A branch and bound parsimony analysis was run using PAUP v. 4.10 (Swofford 2002), followed by generation of Jackknife support percentages. The ITS sequence of P. ecollis is deposited in GenBank (Genbank Acc. MH979036).

A strict consensus tree cladogram (Fig. 4) based on ITS sequences of the tribe Clinantheae places the new species of Pamianthe as sister to P. peruviana with 92% jackknife support, in a subclade that is sister to a second subclade comprising Clinanthus Herb. and Paramongaia Velarde. However, with ITS there is no support for Pamianthe as part of Clinantheae (jackknife support = 42%; Fig. 4). Preliminary super matrix trees from sequence capture with anchored bait enrichment (Meerow, unpublished data) suggest that Pamianthe is in fact sister to the tribes Clinantheae, Eucharideae, and Hymenocallideae, rather than the first branch in Clinantheae.

Figure 4. 

Strict branch and bound parsimony consensus tree of the Clinantheae, based on ITS sequences, with jackknife support values.

Pamianthe ecollis resembles the two other species of Pamianthe in its staminal cup morphology, with the free portion of the staminal filaments attached to the rim of the cup (not below the rim), two lobes or teeth between each two staminal filaments, and the staminal filaments strongly curved inward, as well as numerous, biseriate, winged, wind-dispersed seeds. Leaf width and the conspicuous midvein are similar to that of P. peruviana. It differs from both of the two hitherto known species in having a yellow perianth and staminal cup (versus white in the other two species) and in its nearly obsolete perianth tube. Moreover, P. parviflora has a shorter ovary (10 mm versus 40 mm in P. ecollis) and fewer ovules per locule (about 20 versus about 100 in P. ecollis). Pamianthe peruviana additionally differs in having fewer flowers (2–4, usually 2, versus 9–10 in P. ecollis), shorter pedicels (1.5–3 cm long versus 7–9 cm long in P. ecollis), free tepals much longer (outer tepals 10–12 cm long, inner tepals 9–11 cm long, versus 3.2 and 2.8 cm long in P. ecollis), staminal cup 8 cm long and long-exserted (versus ca. 0.5 cm long and included in P. ecollis), and larger fruit (8 cm long, 5 cm wide, versus 3.8 cm long, 2.9 cm wide in P. ecollis).

The elongate (12–25 cm long) perianth tube in P. peruviana, which contains three nectar-bearing internal channels (Traub 1972), may be correlated with pollination by sphingid moths. The nearly obsolete perianth tube of P. ecollis may be associated with a change in pollinators; in a tubeless perianth, nectar would be available to short-tongued insects, such as bees. No flower visitors have been observed.

The glandular papillae (Fig. 2C) on the adaxial protuberance of the outer tepals apparently have a secretory function. They probably play a role in pollinator attraction; they may produce a substance that is gathered by insect visitors, or they may function as osmophores. Possible osmophores have been reported in the Chilean allioid amaryllid Gilliesia Lindl. (Rudall et al. 2002). The flat, alate seeds are most likely wind-dispersed, suggesting that these plants inhabit open areas within the cloud forest vegetation; seeds of Amaryllidaceae of closed lowland tropical forest, such as Eucharis Planch. & Lind., are relatively few per locule, subglobose, and wingless, and probably are bird-dispersed, and in one case possibly water-dispersed (Silverstone-Sopkin 2011).

The Clinantheae, which is sister to the tribe Hymenocallideae (Meerow et al. 2000), was not previously known to extend to Colombia. We hypothesize that the three rare species of Pamianthe may represent the remnants of a once more broadly distributed epiphytic and lithophytic lineage in the tribe that were isolated as the Andes rose to their present position, and moist forests contracted on the western slopes. We are confident that rigorous analysis of our next generation sequence data will successfully resolve the current ambiguous phylogenetic position of the genus.

Field work by Jhon A. Sánchez-Taborda was financed by the Fondo de Alianzas para los Ecosistemas Críticos (CEPF), Conservación Internacional, Fundación Ecohábitats, and the Corporación Autónoma Regional del Cauca (CRC). We thank the Laboratorio de Imágenes del Postgrado en Ciencias-Biología de la Universidad del Valle (Cali, Colombia) for permission to photograph the flower and seeds of the new species; photographs in this lab were taken by Juan Felipe Ortega-Giraldo. Martín Llano-Almario arranged Figures 1 and 2. Fredy Gómez-Ortiz collected a plant in fruit at the type locality. Laura Clavijo photographed the fruiting plant. Thanks to Liliana Paz y Luis Alfonso Ortega from the Fundación Ecohabitats for their work in designing and managing the project that allowed the collection of this new species, and to Asociación Agroambiental Santa Clara-El Pinche for their support during the field work.