Espeletia praesidentis, a new species of Espeletiinae (Millerieae, Asteraceae) from northeastern Colombia

Abstract A new species of Espeletia from the Páramo de Presidente in northeastern Colombia is described. The species is named Espeletia praesidentis after the name of the páramo, and it is dedicated to the President Juan Manuel Santos, for his persistent efforts in working for peace for Colombia. The new species is closely related to Espeletia dugandii, but differs in the shape and colour of the leaves and arrangements of the capitulescences. A large population was found, but its total extension is yet to be determine.


Introduction
In 1932 the renowned Spanish botanist José Cuatrecasas visited for the fi rst time a páramo in Colombia. From that moment he decided to study the frailejones (common name given to most of the Espeletia Mutis ex Bonpl. species), not knowing the taxonomic endeavor he was going to begin. Th irty-four years later he published the classifi cation of the subtribe Espeletiiane Cuatrec. (Asteraceae: Millerieae) (Cuatrecasas 1976), and continued working on the group until his death in 1996. His masterpiece, the systematic study of the subtribe, represents his work for about 64 years, and was finally published in 2013, after almost two decades of further additions. Th e treatment, however, did not include the genus Espeletiopsis Cuatrec., because Cuatrecasas was not able to complete the work on it (Cuatrecasas 2013). Th e main reason for the delay was the diffi culty of working on the group, defying species concepts and challenging collectors and curators to deepen their collecting techniques and studies.
In recent years, several new species have been published, and more new species are foreseen. At the moment the subtribe contains 8 genera, 144 species (including the new species described here), 17 subspecies, 22 varieties and 8 forms (Cuatrecasas 2013;Diazgranados 2012a;Diazgranados and Morillo 2013;Diazgranados and Sanchez 2013). It has been highlighted as one of the examples of the rapid radiations of the páramos (Madriñán et al. 2013), and phylogenetic approaches will bring further reorganizations within the subtribe (Diazgranados 2012b).
Th ere are a number of reasons making the subtribe a diffi cult group for taxonomist. It is easily identifi ed morphologically, with clear synapomorphies, unique to these plants. However, the variation between and within the species can be astonishing. Despite the notorious morphological variations, characters are often continuous rather than discrete, and species are identifi ed by combinations of character states. Th is is evident, for example, in Cuatrecasas' dichotomous keys, where paragraphs with various character states are needed to identify the species (Cuatrecasas 2013). Th ere is frequent inter-specifi c and inter-generic hybridization with introgression. Th e presence of hybrids with three parental species has even been suggested (Cuatrecasas 2013;Diazgranados 2012a). Hybrid speciation can be also important, as genomic analyses are recently suggesting (Mavárez, J., unpubl. data). Genetic variability between species is very low, and genetic markers are not very useful for separating species (Diazgranados 2012b). Also, populations are very large, dominating the landscape of the páramos. Th is challenges unexperienced collectors, who collect rare individuals that can be hybrids or just mutants in the population, without reporting the population. Also, collecting frailejones appropriately is time-consuming and samples are bulky. Due to all of this, herbarium samples are often incomplete and can be misclassifi ed and misidentifi ed. Finally, because of the island eff ect of the páramos and the limited dispersal capability of frailejones, local (geographic) endemism is common (Diazgranados 2013). Th e previous reasons form together the ingredients for a perfect 'taxonomic storm', inviting review of the species concept for Espeletiinae. In this work we follow the unifi ed species concept (De Queiroz 2007), applicable for rapid radiations in early states of divergence, where populations become phenetically distinguishable and diagnosable.
Th e genus Espeletia, as it is currently defi ned, contains 72 species (including the one described here), distributed from the páramos of Lara (Venezuela) to the páramos of Llanganates (Ecuador). Most of the species grow above the timberline, although some have been identifi ed as pioneer species after disturbance and can grow in ecotones between the high-Andean forest and the páramos, in azonal páramos or in areas with secondary paramisation (Diazgranados 2013).

Methods
Th e Páramo de Presidente is located 28 km south from Chitagá (Norte de Santander), and it is considered part of the Páramos de Almorzadero-Santurbán complex. It can be reached on the road from Chitagá to Cerrito. For decades this area has been considered unsafe, and the fl ora of this páramo has not been studied well yet. Material of the new species was collected during an expedition of the authors in 2009, in which they met with left-wing armed members. Duplicates were distributed to COL, AN-DES and HECASA. Additional duplicates will be distributed to other Colombian herbaria. Micrographs were taken by the fi rst author at the Scanning Electron Microscopy Laboratory of the National Museum of Natural History, in Washington DC. Lauren Merchant from Saint Louis University provided the illustrations, which were funded by the Missouri Botanical Garden and the Smithsonian Institution. Diagnosis. Caulescent rosette of yellowish-whitish appearance, with leaf laminae linear or linear-obovate, naked scapes with long peduncles and 3(-5) capitula, small in diameter, disc paleae oblong, oblanceolate or narrowly obtrullate, very short yellow ray fl owers, and lobes of disc corollas with hairs. Similar to E. dugandii, but more yellowish, with much linear and narrower leaf laminae, much longer peduncles, smaller capitula and ray fl owers, and disc corolla lobes with more hairs. Description. Caulescent polycarpic rosette of yellowish-whitish appearance (not cinereous), 0.8-1.5 m tall (including capitulescences), growing in grassland of páramo proper. Excluding reproductive parts, rosette 40-60 cm in diameter, on stems 0-40 cm tall ( Fig. 1 A-B).
Ecology. A large population of several hundreds of individuals growing in the grasslands of the páramo proper was observed (Fig. 1, 6). Other Espeletia species found in the area are: Espeletia brassicoidea Cuatrec., E. canescens A.C.Sm., E. conglomerata A. C. Sm., E. dugandii Cuatrec., E. standleyana A. C. Sm., and E. steyermarkii Cuatrec. (Fig. 6). E praesidentis can be found in slightly humid plains and on relatively drained slopes.
Etymology. Th e specifi c epithet of this new species, "praesidentis", taken from the locality where the species is found, is dedicated also to the President of Colombia, Juan Manuel Santos Calderón, for his persistent eff orts to achieve peace with the guerillas FARC in Colombia, after 52 years of confl ict. Th e Páramo de Presidente has been one of those places that has been closed to researchers for decades. With the peace agreement this and other places will be open for fruitful botanical explorations during the post-confl ict times in Colombia. May this publication inspire the President to continue with further actions for the preservation of Colombian biodiversity.
Conservation status. Despite seeing a relatively large population, this páramo area is not under any sort of protection, and there are signs of grazing activity. Also, very close there are extensive potato plantations in areas that were covered by páramo vegetation in the past. Th is combination of elements suggests that the species is probably Critically Endangered (CR, according to the IUCN criteria: extent of occurrence estimated to be less than 100 km 2 , habitat fragmentation, and likely decline of the extent of the páramo; http://jr.iucnredlist.org/documents/redlist_cats_crit_en.pdf), or Critically Imperiled (G1, according to NatureServe; http://www.natureserve.org/ explorer/ranking.htm).
Discussion. Th e páramos of Santander and Norte de Santander (Colombia) are considered one of the three centres of radiation for the Espeletiinae (Cuatrecasas 2013;Diazgranados 2012b). Probably because of the topographic complexity of these mountains and the longer time for evolution of these plants in this area with respect to other Colombian cordilleras, the overall diversity in the Santanderes is remarkable: 36 species belonging to 7 genera (all but Carramboa). New species continue to appear as collectors reach previously unexplored páramos, whilst our taxonomic understanding of the group improves.
In 1926 -1927 two American botanists explored the vegetation of these mountains, Ellsworth Paine Killip (1890-1968) and Albert Charles Smith (1906-1999. Smith, who would became later the director of the National Museum of Natural History at the Smithsonian Institution, described years later 10 new species of Espeletia from those collections. Since then, no one has really visited the same places that these botanists explored, probably not even Cuatrecasas, who spent decades collecting the Espeletiinae in the páramos. Collectors of Espeletiinae know well that if they miss the slope or the mountain, they can totally miss the species they are looking for, because of the extreme local endemism of the group. As a consequence, Cuatrecasas (2013) made clear in his monographic work that the status of several taxa could be subject to changes with further collections.
Espeletia praesidentis exemplifi es the lack of collections throughout the páramos of the region, and the challenges taxonomists have to face when studying this group. Cuatrecasas's collections were often limited to the accessibility of roads in those years , and he never found the topolocality where Killip and Smith collected species such as E. conglomerata and E. canescens. In the remarks for E. canescens of his treatment he said "Sometimes I have been inclined to consider E. canescens as a local, extreme variation of E. conglomerata. However, the scanty, authentic material of E. canescens shows features that can justify its specifi c status […] On my 1973 trip, I did not have the time to walk from La Baja all the way to the highest spots at the opposite north end of the Páramo del Romeral, where Killip probably collected the type specimens of E. canescens. Additional collections from the extreme section of the Páramo del Romeral may clarify the taxonomic status of E. canescens" (Cuatrecasas, 2013, pag. 319). In the remarks for E. conglomerata he said "E. conglomerata as well as E. canescens were described with type specimens from Páramo del Romeral between "3800 and 4200" m of altitude. However, according to recent maps, this páramo generally does not exceed 3800 m […]. My own collections represent several minor variations, as well as the typical form" (Cuatrecasas, 2013, pag. 316). In that moment Cuatrecasas was 70-year old, and clearly did not have time or possibilities to explored close areas were in recent years various new species have been discovered (e.g. Espeletiopsis sanchezii S. Díaz & S. Obando or E. diazii Diazgr. & L. R. Sánchez). With no other material than his own collections, he fi rst described morphological variations of his specimens as varieties (Espeletia conglomerata var. macroclada Cuatrec. and E. conglomerata var. pedunculata Cuatrec.). Later, he decided to change the status to hybrids, both within E. conglomerata × E. brassicoidea, and synonymised E. brassicoidea f. contracta Cuatrec. with E. conglomerata. Also, he never published E. conglomerata var. lanceolata Cuatrec. [Nom. nud.,Carriker 34].
Espeletia praesidentis diff ers notably from the type of E. conglomerata (Killip E. P. and Smith A. C. 18635, see key below), and from the hybrids described from Cuatrecasas (2013). We believe E. praesidentis cannot be considered a local variation or hybrid of similar or neighboring species for two reasons: 1) there is a large population of several hundreds of individuals; and 2) there are remarkable morphological diff erences between E. praesidentis and the type collections of other species (as seen in Fig.  4). In this work we do not intend to propose a new categorization for hybrids and/or varieties of E. conglomerata, and we recognize that hybrids can be easily spotted when sympatric species occur, but this clearly was not the case.
Key to Espeletia praesidentis and other Espeletia species found in an area (E. brassicoidea, E. canescens, E. conglomerata, E. dugandii, E. standleyana and E. steyermarkii) Th e most complete key for the genus has been published by Cuatrecasas (2013)