A nomenclator for the frailejones (Espeletiinae Cuatrec., Asteraceae)

Abstract The páramos and high Andean forests of the tropical Andes are largely dominated by frailejones (Nomen nudum Cuatrec., Nomen nudum). These plants are ecologically and culturally essential for both ecosystems and local inhabitants. The frailejones have been studied for over two centuries, but the taxonomic knowledge is still sparse and incomplete. The inedited monograph by Cuatrecasas contains only ca. 70% of the species known today, and publications in the last decade disagree regarding the number of taxa within the group, with estimates ranging from 3 genera and 90 species to 8 genera and 154 species. Moreover the literature contains inexact information about their distribution. As part of a study of the phylogenetic and biogeographic relationships of the group, a thorough revision of the nomenclature was needed as a first step. Currently the subtribe has 8 recognized genera, 141 species, 17 subspecies, 22 varieties, 8 forms, 33 recognized hybrids, 142 synonyms and 5 invalid names, for a total of 368 names (autonyms not counted). The most current list of taxa is presented here, along with some notes and Spanish names. Tamananthus crinitus V.M.Badillo is not included within the subtribe. Various previous species or infraspecific taxa (i.e. Carramboa tachirensis (Aristeg.) Cuatrec., Espeletia algodonosa Aristeg., Espeletia aurantia Aristeg., Espeletia brassicoidea var. macroclada, Espeletia brassicoidea var. pedunculata, Espeletia garcibarrigae Cuatrec. and Espeletiopsis cristalinensis (Cuatrec.) Cuatrec.) are proposed or confirmed as hybrids. Two new records for Colombia are mentioned: Ruilopezia cardonae (Cuatrec.) Cuatrec., which is the first report of Ruilopezia for that country, and Espeletia steyermarkii Cuatrec. Observations regarding the frequency of hybrids in the subtribe are also given.


Introduction
José Celestino Mutis, founder and director of the "Expedición Botánica del Nuevo Reino de Granada" wrote two short diagnoses for the fi rst species of frailejones to be collected, described and illustrated; this material is currently preserved at the Royal Botanical Garden of Madrid, Spain. A manuscript with a four-page description, dated March 7 1792, was found in the Institute of France, Paris. Th is document was presumably written by Juan Bautista Aguilar, calligrapher for the Mutis expedition, and included detailed morphological information and uses. Th ese earliest works are evidence of the admiration that these plants caused in the members of the botanical expedition. Among all the plants collected in over a decade in the New Kingdom of Granada, Mutis chose these plants to honor the viceroy José de Ezpeleta, dedicating the genus Espeletia to him. Th e genus was published offi cially by Humboldt andBonpland in Plantae Aequinoctiales (1809 [1808]) de Cumana et de Barcelone, aux Andes de la Nouvelle Grenade, de Quito et du Pé rou, et sur les bords du rio-Negro de Oré noque et de la riviè re des Amazones. Th is was the beginning of two centuries of prolifi c but complex scientifi c research on this group.
Th e frailejones are very abundant and dominant in the páramos, subpáramos and high Andean forests (Cuatrecasas 1986). Th ey are of critical ecological importance because they contribute to regulating the hydrologic cycle, produce most of the biomass in these ecosystems, prevent soil erosion, and have key associations with more than 125 animal species (García et al. 2005). Much like the astonishing dendrosenecios of the African mountains, frailejones have intrigued naturalists and botanists, not just for their appealing beauty but also for their remarkable adaptations to the extremely harsh environmental conditions of the páramo. Th eir morphological diversity is impressive. Plant sizes range from diminutive fi st-sized species to plants more than 15 m high. Leaves vary from grass-like and glabrous to some of the most densely pubescent found in fl owering plants. Reproductive structures exhibit a similarly spectacular level of variation and hybridization occurs between almost all sympatric species.
In the last decades hundreds of works have been produced on the páramo ecosystem, mainly on ecology, evolution, ethnobotany, and phytochemistry. A quick search in Google Scholar (accessed in March 28 2012) yields 1290 biological papers mentioning one or more of the genera of the subtribe, and 104 papers include these names in their titles. However, the lack of clarity regarding the correct nomenclature and taxonomy of the Espeletiinae, the obligate point of reference in almost all studies on páramos, is a serious handicap. Cuatrecasas worked until his last day on the revision of the subtribe. Th is revision is based on his extremely detailed morphological observations and extensive fi eld experience of decades, and treats seven of the eight genera (excluding Espeletiopsis) and 105 species. Th e revision includes new taxa and combinations, and some of these unpublished names have been used in the literature for more than a decade. Th e manuscript, however, is still in press at the New York Botanical Garden, and valid publication of these names is forthcoming.
In 1996 Cuatrecasas published a key for 20 species of Espeletiopsis, knowing that he would not be able to fi nish the treatment for this genus (Cuatrecasas 1996). Since then, nearly 19 new species and one variety have been published, one species was reestablished and one hybrid proposed (Díaz-Piedrahita and Obando 2004;Díaz-Piedrahita and Pedraza 2001;Díaz-Piedrahita and Rodríguez 2008;Díaz-Piedrahita et al. 2006;Díaz-Piedrahita and Rodriguez-Cabeza 2010;Morillo and Briceño 2007); the author is currently working on the description of 17 additional new species.
Furthermore, there is no agreement regarding Cuatrecasas' classifi cation, and many of the genera were not resolved as monophyletic in earlier studies. Even the number of published species is unclear in recent publications. Based on molecular evidence (nrITS), Rauscher (2002) concluded that the subtribe is monophyletic and that the closest relatives are Rumfordia DC., Ichthyothere Mart. and Smallanthus Mack.; he proposed to replace the subtribe Espeletiinae by the "Espeletia complex" and mentioned that the group has approximately 100 species. Díaz-Piedrahita et al. (2006) indicated a total of 154 species for Espeletiinae. Fagua and González (2007) mentioned ca. 100 species. Panero (2007) described the subtribe with only three genera (Carramboa, Espeletia and Tamananthus) and 90 species, and Baldwin (2009) reported 90 species as well. Th e Plant List (Missouri Botanical Garden and Royal Botanic Gardens in Kew 2010), however, shows in total 145 species (Carramboa 7 spp., Coespeletia 6 spp., Espeletia 69 spp., Espeletiopsis 24 spp., Libanothamnus 14 spp., Ruilopezia 24 spp. and Tamania 1 sp.), and mentions 4 hybrids, 88 synonyms, 36 scientifi c plant names of infraspecifi c rank, and 24 unassessed names. Th e subtribe has been recently circumscribed within the tribe Millerieae Lindl, as part of the Heliantheae Alliance (Baldwin 2009;Panero 2007).

Methods
Th e taxonomy of the subtribe was carefully examined. A total of 4408 specimens from personal collections and diff erent herbaria (A, AAU, ANDES, CAS, COL, CUVC, DS, ECON, F, FMB, G, GB, GH, H, HECASA, HUA, K, M, MA, MEDEL, MER, MERF, MIN, MO, MY, MYF, NEU, NY, P, PSO, QCA, QCNE, S, U, UC, US, VALLE, VEN and WAG) were studied. A database was built and 3408 totally iden-tifi ed specimens were georeferenced. Plants from 1685 photographs taken during fi eldwork by the author were identifi ed and georeferenced as well, for a total of 5093 mapped individuals. Maps were generated in ArcGIS10 (ESRI) and DIVA-GIS 7.5 (Hijmans et al. 2012).
A list of all valid names including infraspecifi c taxa is provided here, with synonyms and basionyms. Autonyms are not included here; these are automatically established with the fi rst instance of valid publication of a name of an infraspecifi c taxon under a legitimate species, in conformity with the Art. 26.3 of the ICBN (McNeill et al. 2006). To preclude valid publication in this paper of the new taxa and combinations from Cuatrecasas' manuscript, no type information, description, diagnosis or full and direct references were added for these names (in corformity with Art. 33.4 and 33.8 of the ICBN, in McNeill et. al. 2006). In addition, the abbreviation "ined." was added after the names, as well as a note about the forthcoming publication. Th e information for these names will be updated in the online version after after the publication of the Cuatrecasas manuscript.
Hybrids are listed at the end of each genus. Th e dominant morphology of the hybrid corresponds to the fi rst species listed in the hybrid formula. Th erefore hybrids are arranged alphabetically by the fi rst species in the formula.
Th ere is no English name for these plants. In Spanish they are usually called frailejones (pronounced fry-lay-ho-ness; in singular frailejón, pronounced fry-lay-hón), which means "big monk", due to the resemblance of these plants to medieval monks when seen in the typically foggy weather of the páramo. Th is name is applied to all the species of the group. Some of the smallest species are also called chijí. Th e arboreal species often receive equivocal (applied to more than one species) names like anime, carrambo and quiñon (Carramboa spp.), punta de lanza, tabaco and its derived names, trementino or incienso (Libanothamnus spp. and Tamania chardonii). Punta de lanza refers to the spear-like shape of the leaves. Tabaco and derived homonymy (tabaquero, tabaquillo, tabacote, tabacón, etc.) were probably given by the resemblance of the leaves of some species to tobacco leaves. Th e names trementino and incienso originated from a common use for these plants. Resin from all the frailejones is abundant and aromatic, and can be used as incense or to extract oil of turpentine, which is used as a solvent for paints and varnishes. Th is was a common practice in the eighteenth century, as reported in the fi rst description of the genus Espeletia by Mutis' expedition in 1792.
Th ere are other common names used, mainly in Venezuela, where the group shows its maximum morphological variation. A list of all common names found on specimens and in the literature (Bernal et al. 2012) contained a total of 49 diff erent names used for all 141 species and some hybrids. Th ere are 16 univocal names (applied only to a single species), but there are no biunivocal names (applied only to a single species, which does not receive any other common name). One species receives six common names and four species are referred to by fi ve names.
Th e red list for the frailejones of Colombia (García et al. 2005) proposed new Spanish names for all the species included in the publication, with the aim of socializing the knowledge of these species and instilling a feeling of responsibility among local inhabitants to protect these resources. Th is strategy has been applied successfully in the past, and remarkable examples exist today, especially with birds, where some endangered species that are well-known today were virtually unknown to the public two decades ago. Frailejones are common motifs in posters, postcards and T-shirts, and most Colombians, northern Ecuadorians and Venezuelans from the mountain regions have seen or heard of frailejones. However, their knowledge about the diversity of this group is extremely low, and most people believe that there is only one species, the frailejón.
In this work, all common names reported for each taxon are included, and the respective countries of origin are indicated in parenthesis (C: Colombia, E: Ecuador, V: Venezuela). When names are in indigenous tongue, the name of the tongue is reported (e.g. Kamsá). Names for intraspecifi c taxa are reported only when they diff er from the names used for the species. For species that lack univocal names, new univocal Spanish names are proposed to create the consciousness, appropriateness and responsibility for conservation among locals (marked with *). Recommended names to use (univocal names) are emphasized in bold. Common names. Anime amarillo (V), anime montañero (V), carrambo (V), carrambo paramero (V), frailejón amarillo (V), frailejón anime (V).

Systematics
Note. C. badilloi includes now former C. littlei and C. pittieri was proposed as a variety of C. badilloi (below). However some individuals show contrasting morphological diff erences, especially in the morphology of the leaves. More ecological, morphological, anatomical and genetic studies are needed to determine if these entities are indeed diff erent taxa.
Th e specimen was poorly preserve and glued to the herbarium sheet, and does not allow a comprehensive examination. Th e following characteristics do not fi t within the synapomorphies of the Espeletiinae: herbaceous or shrubby habit, alternate leaves, bilabiate corolla with the 2-lobed-lip well developed, bidentate style, 7-8 ray fl owers in total, isomorphic uncompressed cypselae with prominent carpopodium, and fertile disk fl owers. Some species (Coespeletia spp. and a few Espeletia spp.) in Espeletiinae show bilabiate ray fl owers, sometimes with one or two lobes, although this seems to be an inconsistent character within the species. However, the lip in Espeletiinae species, when present, is never as developed as in T. crinitus, and lobes in Espeletiinae are in general shorter, deeper and more divergent from each other. Th e morphological characteristics in the specimen and in the description do not conclusively place this taxon with the Espeletiinae. DNA amplifi cations have not been successful until now, but AFLP data suggest that this species is closer to Ichthyothere Mart. and Smallanthus Mack. than to the Espeletiinae. I propose excluding this genus from the subtribe until more material can be studied morphologically and genetically.

Frequency of hybrids in the subtribe Espeletiinae
A total of 27 hybrids are included here. From the 4408 herbarium specimens examined, 122 (2.8%) were hybrids. However these numbers do not represent the frequency of hybridization events. It is unquestionable that hybridization is an important process in the subtribe. During my fi eld observations, I was able to collect and photograph dozens of putative hybrids, which occur typically in contact zones of two parapatric populations (e.g. between the usually low-elevation populations of Libanothamnus and the high-elevation populations of Espeletia), or in areas where sympatric species have large populations (e.g. páramos with E. grandifl ora and E. argentea, or with E. schultzii and E. weddellii). However, the frequency of hybridization is overemphasized in the botanical collections, because hybrids are the "diff erent ones" and are usually more often collected by inexpert collectors who believe they are novelties to science.
It is interesting to note that hybridization appears to be more frequent at both extremes of the altitudinal range. In the superpáramos it is relatively frequent between Coespeletia species and other sympatric taxa of the subtribe (e.g. E. schultzii). In the subpáramos it is common between Carramboa, Espeletiopsis, lower-elevation Espeletia (e.g. E. aristeguietana, E. argentea, E. boyacensis, etc.), Libanothamnus and lower-elevation Ruilopezia species (e.g. R. marcescens, R. jabonensis, R. vergarae, etc.).
Even more intriguing is that when hybridization in these altitudinal extremes occurs, the synfl orescence of the hybrids are often more similar to the species that is adapted to lower elevations, while the leaves are more similar to the species adapted to higher elevations. For instance, Espeletiopsis ×cristalinensis is found in the upper range of Libanothamnus neriifolius and the lower range of Espeletia aristeguietana. It has very pubescent thick leaves like those of E. aristeguietana, whereas its synfl orescence has the branching patterns and shape of L. neriifolius. Th e same pattern can be seen when hybridization occurs between the lower-elevation E. brassicoidea and the higher-elevation E. conglomerata, or with other hybrids. A possible explanation is that the leaves of the higher-elevation parental species are more adapted to the colder conditions than those of a parental species adapted to warmer conditions, and therefore the expression of this character is selected. In the same way, synfl orescences of lower-elevation parental species may be more eff ective in attracting potential pollinators than those adapted to higher-elevation pollinators. What is clear is that there is a fertile fi eld of research on hybridization within the subtribe.
Based on my personal observations and collections of Espeletiinae, collectors should be always suspicious about a hybrid individual when any of the following conditions occurs: -Isolated individuals with no abundant populations: Espeletiinae have no pappus in their cypselae, and seeds are dispersed mainly by gravity; therefore collectors should expect to fi nd several individuals of the same species, all related in some degree, and a few sporadic hybrids of sympatric or parapatric species.
-Generally the hybrids have leaves with undulate margins, and often unhealthy appearance.
-Pubescence of hybrids frequently looks disheveled.
Note. Th e species had been reported in the Venezuelan slopes of the gran Páramo de Tama, especially along the basin of the Oirá river. One large population was found on the Colombian side. Th is is the fi rst record of Ruilopezia for Colombia.

Notes about the geographic distribution
Th e frailejones are widely distributed and abundant in the high Andean forest and páramos of Colombia, Venezuela and, to a lesser extent, Ecuador, where only one species occurs in the north of the country and in the Sierra de Llanganates (Fig. 2A). Th e northernmost population (-73.8°W, 11°N) grows in the Sierra Nevada de Santa Marta, at -73.8°N. Even though the author has been told about reports of L. neriifolius in the state of Sucre in Venezuela (approx. -63.5°W, 10.5°N), which would be the extreme NE of the geographic range, no herbarium specimens from this locality have been seen. Th e southernmost species is Espeletia pycnophylla, which grows in the border region between Colombia and Ecuador and has a separate isolated population far south in the Llanganates National Park (-78.5°W, 1.2°S) (Fig. 2B). Nearly 80% of the species occur between -70.5 and -73°W and 5.5 and 9.0°N, along the Eastern Cordillera in Colombia and the Venezuelan Andes. Th e frailejones grow in 21 states (64%) of Colombia, six in Ecuador (24%) and 12 in Venezuela (52%).  Most of the species of Espeletiinae are restricted to a single topographic páramo complex. Similar to island species, they have lost the ability for long-distance dispersal. Seeds lack pappus (except for the scale-like pappus of Tamania) and are dispersed by gravity, with no long-distance dispersers known. Only nine species are shared between Colombian and Venezuela, and a few species occur in multiple páramo localities, includ-Th ere are three apparent centers of radiation: Mérida (with 44 spp.) in Venezuela, Santander-Norte de Santander (39 spp. combined) and Boyacá (45 spp.) in Colombia (Fig. 2E-F). An analysis of richness by area (squares of 0.4° × 0.4° km 2 ) highlights two areas with the highest number of species (> 22 spp.), one in central Mérida and the other in Boyacá with limits of Santander (Fig. 2D). Táchira, a Venezuelan state bordering with Colombia, has the highest richness of genera, followed by Mérida (Fig. 2E). Overall, Colombia has the highest richness of species (86 spp.), followed by Venezuela (67 spp.) and Ecuador (1 spp.). Th ese numbers include one presumable species in Colombia (i.e. E. tillettii), found on the Sierra de Perijá in the Venezuelan border with Colombia.
Th e author is currently describing 17 new taxa of frailejones. Th e phylogeny will provide new insights into understanding the relationships between species and genera, and will generate further changes in the taxonomy of the group. Th e systematic list here provides a taxonomic base for all these changes and for any other study related to frailejones.